You are on page 1of 227

MOLECULARAND CELLULAREFFECTS OF NUTRITION ON DISEASE PROCESSES

Molecular and Cellular Effects of


Nutrition on Disease Processes

Edited by

GRANT N. PIERCE HEINZRUPP


Institute of Cardiovascular Sciences Medizinische Forschung
St. Boniface General Hospital Philipps University of Marburg
Research Centre, 351 Tache Avenue Karl-von-Frisch-Str. 1
Winnipeg, Manitoba, Canada, R2H 2A6 35033 Marburg, Germany

TOHRUIZUMI ALAIN GRYNBERG


Professor and Chairman INRA
Department of Internal Medicine Lipides Membranaires et Fonctions Cardiovasculaires
Kitasato University School of Medicine Faculte des Sciences Pharmaceutiques et Biologiques
1-15-1 Kitasato, Sagamihara 4 Av de I 'Observatoire
Kanagawa 228, Japan 75270 Paris Cedex 06, France

Reprinted from Molecular and Cellular Biochemistry, Volume 188 (1998)

Springer-Science+Business Media, B.V.


Library of Congress Cataloging-in-Publication Data

Molecular and cellular effects of nutrition on disease process/edited by Grant N. Pierce


... [etal.].
p. cm. -- (Developments in molecular and cellular biochemistry)
ISBN 978-1-4613-7641-5
1. Nutritionally induced diseases- Molecular aspects - Congresses.
2. Diet therapy - Congresses. 3. Pathology, Molecular - Congresses.
4. Functional foods-Congresses. I. Pierce, Grant N. 11. Series.
RC622.M64 1998
616.07 - dc21 98-3954
CIP

ISBN 978-1-4613-7641-5 ISBN 978-1-4615-5763-0 (eBook)


DOI 10.1007/978-1-4615-5763-0

Printed on acid-free paper

All rights reserved

© Springer Science+Business Media Dordrecht 1998


Originally published by Kluwer Academic Publishers in 1998
Softcover reprint of the hardcover 1st edition 1998

No part ofthe material protected by this copyright notice may be reproduced or


utilized in any form or by any means, electronic or mechanical,
including photocopying, recording or by any information storage and
retrieval system, without written permission from the copyright owner
Molecular and Cellular Biochemistry:
An International Journal for Chemie al Biology in Health and Disease
CONTENTS VOLUME 188, Nos. 1 & 2, November 1998

MOLECULAR AND CELLULAR EFFECTS OF NUTRITION ON DISEASE PROCESSES


Drs. Grant N. Pierce, Heinz Rupp, Tohru Izumi andAlain Grynberg

Preface

Part I: Cancer
L. Hilakivi-Clarke and R. Clarke: Timing of dietary fat exposure and mammary tumorigenesis: Role of estrogen receptor and protein
kinase C activity 5--12
E. Nolan, M. Donepudi, K. VanWeelden, L. Flanagan and J. Welsh: Dissociation ofvitamin D3 and anti-estrogen mediated growth
regulation in MCF -7 breast cancer cells 13--20
B. Schwartz, C.Avivi-Green and S. Polak-Charcon: Sodium butyrate induces retinoblastoma protein dephosphOlylation, pl6 expression
and growth arrest of colon cancer cells 21-30

Part II: Cell growth and development


A. Vogel Hertzel and D.A Bernlohr: Regulation of adipocyte gene expression by polyunsaturated fatty acids 33--39
K.H. Falchuk: The molecular basis for the role of zinc in developmental biology 41-48
A.-ü. Makinde, P.F. Kantor and G.D. Lopaschuk: Maturation offatty acid and carbohydrate metabolism in the newborn heart 49--56
E.D. Harris, Y. Qian and M.C.M. Reddy: Genes regulating copper metabolism 57--62
A.S. Prasad: Zinc and immunity 63--69

Part III: Diabetes


P. Poucheret, S. Verma, M.D. Grynpas and J.H. McNeill: Vanadium and diabetes 73--80
L.J. McCargar, S.M. Innis, E. Bowron, J. Leichter, K. Dawson, E. Toth and K. Wall: Effect of enteral nutritional products differing
in carbohydrate and fat on indices of carbohydrate and lipid metabolism in patients with NIDDM 8H!9
L. Golfinan, I.M.C. Dixon, N. Takeda,A Lukas, K. Dakshinamurti and N.S. Dhalla: Cardiac sarcolemmal Na+-Ca'+ exchange and
Na+-K+ ATPase activities and gene expression in alloxan-induced diabetes in rats 91-101
P. Rösen, X. Du and D. Tschöpe: Role of oxygen derived radicals for vascular dysfunction in the diabetic heart: Prevention bya-
tocopherol? 1O3--111
J.C. RusselI, S.E. Graham and M. Richardson: Cardiovascular disease in the JCR:LA-cp rat 113--126

Part IV: Vascular dysfunction


M.B. Zernel: Nutritional and endocrine modulation of intracellular calcium: Implications in obesity, insulin resistance and hypertension 129--136
K. Dakshinamurti, K.J. Lai and P.K. Ganguly: Hypertension, calcium channel and pyridoxine (vitamin B6) 137-148
M. Aviram and B. Fuhrman: LDLoxidation by arterial wall macrophages depends on the oxidative status in the lipoprotein and in the
cells: Role ofprooxidants vs. antioxidants 149-159
N. Iliskovic, T. Li, N. Khaper, V. Palace and P.K. Singal: Modulation of adriamycin-induced changes in serum free fatty acids,
albumin and cardiac oxidative stress 161-166

Part V: Heart disease


K.A. Detiliieux,AF.A. Meyers, J.T.A Meij and PA Cattini:AnNG-rich motifin the rat fibroblast growth factor-2 gene confers
enhancer activity on a heterologous promoter in neonatal rat cardiac myocytes 169-176
R. Vetter, M. Kott, W. Schulze and H. Rupp: Influence of different culture conditions on sarcoplasmic reticular calcium transport in
isolated neonatal rat cardiomyocytes 177-185
G. Bkaily, D. Jaalouk, S. Sader, H. Shbaklo, P. Pothier, D. Jacques, P. D'ürleans-Juste, E.J. Cragoe Jr. and R. Bose: Taurine indirectly
increases [Cal; by inducing Ca'+ influx through the Na+-Ca2 + exchanger 187-197
S. Takeo, Y. Nasa, K. Tanonaka, K-i. Yabe, M. Nojiri, M. Hayashi, H. Sasaki, K. Ida and K. Yanai: Effects oflong-term treatment
with eicosapentaenoic acid on the heart subjected to ischemia/reperfusion and hypoxia/reoxygenation in rats 199--208
H. Rupp, V. Elimban and N.S. Dhalla: Differential influence of fasting and BM 13.907 treatment on growth and phenotype of pressure
overloaded rat heart 209--215
F.Y. Xu, S.L. Kelly, W.A. Taylor and G.M. Hatch: On the mechanism ofthe phospholipase C-mediated attenuation of cardiolipin
biosynthesis in H9c2 cardiac myoblast cells 217-223
M. Turcani and H. Rupp: Development ofpressure overload induced cardiac hypertrophy is unaffected by long-term treatment with
losartan 225-233

Index to Volume 188 235-237


Molecular and Cellular Biochemistry 188: 1, 1998.

Preface

The papers ofthis Special Issue of Molecular and Cellular Nutraceuticals are also known as 'health foods' but they also
Biochemistry are invited, peer-reviewed submissions from include any product derived from animal or plant sources
speakers who attended the 2nd World Conference of the which ultimately can provide a medical benefit. Four symposia
International Society for MolecuIar Nutrition & Therapy. This on various aspects concerning nutraceutical products and the
Conference was held fromAugust 2--4th, 1997 in Winnipeg, marketing ofthese compounds was addressed. It is hoped that
Canada. The goal of the Conference was to advance our these interactions represent the start of a new, highly interesting
knowledge concerning the molecular events which link scientific venture uniting the nutraceutical industry with the
nutrition to various disease processes in the body. Attending main stream academic medical research community. The
the meeting were scientists from 18 countries ineluding nutraceutical industry is expected to be a multibillion dollar
Canada, the United States, Japan, the United Kingdom, a year business world-wide by the turn of the century. The
Franee, the Czeeh Republie, Germany, the Netherlands, identification of the active ingredients in any number of
Bahrain, Hong Kong, India, Israel, China, Kuwait,Argentina, health foods reputed to be medieally beneficial would be just
Sweden, Australia and Mexieo. There were 56 Distinguished one of the goals of such a scientific interaction. This Con-
Speakers invited to present leetures as part of 16 different ferenee represented an important starting point for initiating
symposia. There were two distinetive aspeets of the Con- these potentially significant ventures in the future.
ferenee worthy of note. First, it represented a starting point Overall, this was a highly enjoyable, rewarding Con-
for the establishment of an American Seetion of the Inter- ference which appears to have succeeded as a spring board
national Soeiety for Moleeular Nutrition & Therapy. The goal for advancing science in this area in the future. VITA HEALTH
ofthis Seetion and the Soeiety as a whole is to bring together has played a leading role in advancing products in the
scientists who are interested in the moleeular and eellular nutraceutical industry and the Conferenee Organizers would
events whieh underlie nutritional therapy of a variety of like to thank VITA HEALTR for thcir help in bringing thc
disease proeesses. The Ameriean Seetion of ISMNT was Conference to fruition. The volume of papers assembled
successfully launehed and now has asolid eore of 15 out- here in this Special Issue represent an important compilation
standing scientists from all over North America on its of unique articles addressing the molecular and cellular
Executive Committee. The second unique feature of the basis for the nutritional and therapeutical treatment of five
Conference was its emphasis on the nutraceutical industry. general disease processes. We hope that you enjoy them!

Grant N. Pierce, Winnipeg, Canada;


Heinz Rupp, Marburg, Germany;
Tohru Izumi, Kitasato, Japan;
Alain Grynberg, Paris, Franee
PART I

CANCER
Molecular and Cellular Biochemistry 188: 5--12, 1998.
© 1998 Kluwer Academic Publishers.

Timing of dietary fat exposure and mammary


tumorigenesis: Role of estrogen receptor and
protein kinase C activity
Leena Hilakivi-Clarke l ,2 and Robert Clarke1,3
lLombardi Cancer Center; 2Department of Psychiatry, and 3Department ofPhysiology, Georgetown University,
Washington, DC, USA

Abstract
The possible association between a high fat diet and increased breast cancer risk has remained controversial. This largely reflects
the conflicting data obtained from migrant, case control and animal studies, which generally support this association, and cohort
studies which often fail to show a link between fat and breast cancer. The mammary gland is particularly sensitive to estrogens
during fctal dcvclopmcnt, leading us to hypothesize that dietary fat levels during this period may significantly influencc breast
cancer risk. Using chemically-induced mammary tumors in rats as our experimental model, we have demonstrated the ability
ofa matemal diet, high in the polyunsaturated fatty acid (PUFA) linoleic acid, to alter mammary gland differentiation, accelerate
the onset of sexual maturation, and increase breast cancer risk. The mammary glands of female rats exposed to a high-fat diet
in utero have more ofthe undifferentiated structures (terminal end buds) and fewer ofthe differentiated structures (alveolar
buds) than the glands ofrats exposed to a low-fat diet in utero. Furthermore, these mammary glands contain lower levels of
total estrogen receptors and have reduced total protein kinase C activity. These effects appear to be mediated by an increase in
the serum estradiollevels ofpregnancy, which are elevated at least 30% in pregnant dams fed a high-fat diet. Furthermore, the
administration of estradiol to pregnant dams produces effects on mammary gland development, onset of puberty and sensitivity
to chemical carcinogenesis comparable to those seen in the offspring ofrats fed a high fat diet during pregnancy. Our results,
thus, support the hypothesis based on epidemiological data that high matemal estrogen levels increase daughters' breast cancer
risk. The results also suggest that a high-fat diet may be an important factor in increasing pregnancy estrogenic activity. (Mol
Cell Biochem 188: 5-12, 1998)

Key words: breast cancer, estrogen, dietary fat, polyunsaturated fatty acid, estradiol, pregnancy

Dietary fat and breast cancer The initial observation linking dietary fat to breast cancer
was the high correlation noted between national fat intake and
The greatest potential for future success in reducing breast breast cancer risk across different countries [2]. This
cancer mortality is to identify new prevention strategies for observation led to aseries of case control studies, the majority
this disease. Diet is suggested to be involved in the multistep of which implicated that cases consumed more fat than
process that leads to clinical manifestation of breast cancer. controls [3,4]. Numerous animal studies also showed that
It has been estimated that 35% of breast cancers could be feeding a diet high in polyunsaturated fatty acids (PUFAs)
prevented hy appropriate dietary modifications [I]. However, promoted carcinogen-induced mammary tumorigenesis [5,
despite extensive human and animal research, controversies 6]. Other types of fats, inc1uding saturated fats, were less
exist concerning the possible link between nutritional effective in animals [5]. In humans, total or saturated fat
components, including fat, fiber and vitamins, and hreast intake appeared to be most closely linked to hreast cancer [4].
cancer risk. Additional controversy arose when it became apparent that

AddressJor ojJPrints: L. Hilakivi-Clarke, Lombardi Cancer Center, Room W405, Georgetown University, 3970 Reservoir Rd, NW, Washington, DC 20007-
2197, USA
6

the majority, but not all, cohort studies did not establish an levels (Table 2). Several clinical studies show that both a
increased breast cancer risk associated with a high-fat diet [7]. high fat and/or total caloric intake increase the levels of
circulating free estrogens, whereas a low-fat diet is associated
with low plasma estrogen levels [15-18]. Results obtained
Sensitive periods and a high-fat in animal studies are less clear, but support the link between
an isocaloric high-fat diet and high serum estrogen levels
exposure [19,20]. Our findings clearly indicate that a diet high in corn
oil increases serum estradiol (E2) levels in pregnant rats [13,
We have proposed that a high-fat diet increases breast 21]. This increase does not persist in the offspring after
cancer risk if consumed during periods when the mammary birth, which is consistent with the rapid clearance of
gland is sensitive to endogenous estrogens, such as fetallife maternal estrogens from neonates.
[8]. Evidence in support of the hypothesis includes the At least three possible mechanisms exist for a high-fat
observation that Asian women, who consume a low-fat diet diet to increase circulating estrogens. Firstly, a high fat
in their horne country and have a low breast cancer incidence intake tends to lead to accumulation of adipose tissue, which
[9], when immigrating to the United States reach the high is an important site for the conversation of androstenedione
breast cancer incidence of Western countries within a few to estrone [25]. Secondly, arachidonic acid, a metabolite of
succeeding generations [10, 11]. Breast cancer risk among PUFAs, activates P450 aromatase that then increases
Asian immigrants is already 80% higher after a decade of conversion of androstenedione to estrone [26]. Finally,
living in the US, when compared with the risk of Asian PUFAs can reduce the binding of estrogens to serum binding
women living in the East [11]. The most dramatic increase proteins, including both sex-hormone binding globulin
in risk occurs between Asian-Americans born in the West (SHBG) and albumin, thereby increasing the circulating
and those born in the East [11]. A higher dietary fat intake levels ofbiologically potent estrogens [27].
during pregnancy in the West than in the East may be
responsible for the transition towards higher breast cancer
risk between Asian generations living in the West. Estrogen receptor and breast cancer
Animal studies provide more direct evidence that a maternal
intake of a high-fat diet during pregnancy increases sponta- The fat-induced elevation in circulating estrogens is likely to
neous or carcinogen-induced mammary tumor incidence in indirectly affect mammary glands and tumors by influencing
the female offspring [12, 13]. The dietary fat source in these
studies was corn oil that is high in PUFAs, particularly Table 1. Putative mechanisms of action of dietary fat.
linoleic acid. The earlier studies have generally explored the
effect of fat on promotion or progression of breast cancer. Generation of active lipid peroxides [22]
None of these studies were constructed/performed in a
lndirect effects on signal transduction and gene expression through
manner that would enable them to address our hypothesis,
alterations in cellular membrane structure and TImction [23]
that dietary exposure during early life can significantly
increase breast cancer risk in later life. Thus, an important Direct effeets on lipid-mediated signal transduction pathways. e.g., by
period during which the mammary gland may be sensitive to altering the levels of araehidonie acid and eieosanoids [24]
dietary fat, might have been overlooked.
Alterations in the levels/bioavailability of sex steroid hormones and their
reeeptors [15-21]

Mechanisms mediating the effects of


Table 2. Observations demonstrating potential associations betwcen dietary
dietary fat fat consumption, perturbations in serum estrogen levels and breast cancer
risk.
Diet and estrogens
Lifetime exposure to estrogens is lower in Asian warnen, who also
consume a low fat diet and have a lower breast cancer risk [9]
The mechanisms through which dietary fat alters mammary
tumorigenesis, are unclear. Several possible mechanisms A low fat diet can reduce serum estrogen levels [18]
exist, including direct effects on lipid signaling pathways,
Elevated serum estrogen levels are associated with increased breast
and indirect effects mediated through perturbations in sex
cancer risk in same women [73]
steroid levels (Table 1). Since estrogens have been extensively
implicated in affecting breast cancer risk [14], one likely Obese women have elevated serum estrogen levels, and postmenopausal
factor is the apparent ability of fat to alter serum estrogen obesity is associated with increased breast cancer risk [74]
7

estrogen' sinteractions with its nuc1ear estrogen receptors ER may be required for fat to promote tumor growth and,
(ER). The ER is present in utero [28], and its concentrations perhaps, elevate the levels of this receptor. There is no
in the mammary gland increase between birth and the first correlation between the recurrence of ER negative breast
week of life [29]. The ER content remains at a relatively tumors and dietary fat intake, while ER positive tumors are
constant level after the prepubertal period, but appears to vary more likely to metastasize if a woman is consuming a high-
in concert with the estrus cyc1e [30). A change in mammary fat diet [46].
ER levels also occurs during other periods when the levels Where associations between dietary fat and increased
of estrogens vary. For example, during pregnancy when breast cancer risk have been reported, these are almost
circulating estrogen levels are high, the ER concentrations exc1usively found in postmenopausal patients [47,48]. A
are low, but detectable [31]. A marked increase in mammary higher proportion of ER-positive tumors arise in post-
ER content and a decrease in serum estrogens occurs during menopausal women [43]. However, a significant proportion
lactation [32]. Thus, the changes in ER levels during ofhuman breast tumors are ER-negative [34], whereas the
pregnancy and lactation reflect the downregulation of ER carcinogen-induced rodent mammary tumor models are ER-
by estrogens [33). positive and strongly estrogen-dependent [49]. An exposure
ER-positive breast tumors, which account for about 60% to a high-fat diet prornotes the growth ofmammary tumors
ofall breast cancers [34] probably arise from within the ER in animal models [5]. These observations are consistent with
positive epithelial cell populations of normal breast tissue the hypothesis that a high-fat diet preferentially affects
[14]. Some 90% ofthese are associated with adjacent, ER- breast cancer in ER-positive mammary glands. Thus, the
positive, non-neoplastic tissue [35]. ER-positive tumors tend effect of dietary fat on mammary gland ER content may
to be more differentiated [36-38], exhibit a slower growth explain some ofthe difference between animal studies and
pattern [39, 40J and a better overall prognosis [41,42]. A lower human cohort studies concerning a high-fat diet and breast
proportion ofER-positive tumors are found in premenopausal cancer. In humans, a high-fat diet would preferentially
than postmenopausal breast cancer patients [43]. These increase breast cancer growth among the approximately
findings could suggest that low mammary ER levels may be 60% ofthe study population with ER-positive tumors. This
associated with an increased risk to develop premenopausal could impact the power of some cohort studies to detect a
breast cancer. Since the data are based on ERa measurements, significant overall trend for increased breast cancer risk in
the role of ERß in affecting breast cancer risk remains to be women consuming a high fat diet [7].
determined.

Estrogen receptor and dietary Jat exposure during


Estrogen receptor and dietary Jat in take sensitive periods in mammary gland development

Since a high-fat intake elevates serum estrogens, a logical The results in humans and animals suggest that high ER levels
consequence of a high dietary fat consumption is an alteration in the mammary gland are associated with fat-induced
in mammary ER status. Therefore, we have explored whether promotion and progression ofbreast cancer. However, high
a diet high in n-6 PUFA affects ER protein levels in the ER levels prior to the carcinogen exposure may not affect the
mammary gland. Surprisingly, female mice consuming a susceptibility to develop carcinogen-induced mammary
high-fat diet exhibit a 6-fold increase in the mammary ER tumors. An exposure to a high-fat diet prior to breast cancer
content [44]. A similar increase has been seen in DMBA- initiation in adult animals does not appear to have significant
induced manunary tumors in female rats fed with a high corn effects on mammary tumorigenesis [5]. However, fat is likely
oil diet [20]. These results would suggest that PUFAs may to increase mammary ER content both before and after
up-regulate ER, and this increase cannot be reversed with a carcinogen exposure. Thus, high mammary ER levels in
simultaneous increase in circulating estrogens. normal adult mammary glands do not increase the risk to
The effect of fat intake on ER in the breast has not been develop carcinogen-induced tumors.
directly explored in human populations. Indirect evidence It was of interest to determine whether in utero exposure
to suggest that a high-fat diet may induce ER is available to a high-fat diet, that affects preinitiation events that increase
from studies showing that either obese women, or women susceptibility to develop mammary tumors, also affects
consuming a high-fat diet, are more likely to develop ER- mammary ER levels. Our data show that a maternal exposure
positive mammary tumors than women consuming a low- to a high-fat diet induces a 4-fold reduction in the ER content
fat diet [45]. Thus, a dietary fat appears to increase in the offspring's mammary gland, when compared with ER
mammary ER content hoth in animals and humans. This content in the low-fat offspring [44]. This finding is similar
increase is not likely to be caused by a fat-induced elevation to that showing that early postnatal treatment with DES
in serum estrogens. However, it appears that the presence causes a permanent reduction in the concentrations of ER
8

in the mouse mammary gland and DMBA induced tumors high-fat diet may induce different responses in the mammary
in rats [50, 51]. The decrease in the mammary ER content gland. Whether this is due to differential expression ofERß
both in the high-fat and DES-exposed offspring suggest that vs. ERa remains to be established. Nevertheless, the potential
the factor(s) responsible may be a high maternal estrogenic for the mammary gland to exhibit different responses to
environment. There are several clinical observations that estrogens at different times could have considerable impact
support the link between high maternal estrogen levels on how we think about estrogenic exposure and breast cancer
during pregnancy and increased breast cancer risk among risk, and how to modulate risk through chemoprevention or
daughters (Table 3). Thus, elevated circulating E21evels during other dietary means.
pregnancy in the mothers consuming a high-fat diet as well as
maternal DES exposure could have resulted in a permanent
down-regulation ofER in the offspring's mammary gland. Pro tein kin ase C and breast cancer
We do not have a clear explanation for why a high-fat diet
increases ER levels when consumed by adult animals (and Hormones can regulate cellular functions by activating some
in breast tumors in humans), and why it reduces ER levels if ofthe several isoforms ofprotein kinase C (PKC) [55, 56].
the exposure occurs through a pregnant mother. In both cases, For example, E2 increases PKCÖ expression in the uterus
a high-fat diet increases serum E2 levels. However, the [57]. In human breast cancer celllines PKC isoenzymes
developmental state ofthe mammary gland, and the respective down-regulate ERmRNA expression [58-60]. Proteinkinase
endocrinologic environments, are quite different in fetal and C also is dependent on diacylglycerol (a fatty acid metabolite)
adult life. This suggests that the response ofthe gland to fatf and calcium for activation.
E2 may vary during lifetime. While PKC is linked to breast cancer [55, 61], its role is
Another explanation for differences between adult and fetal not fully understood. PKC activity is high er in the malignant
exposure may be that in utero dietary fat may predominantly than benign breast tissues [61], and higher in the more
affect ERß receptors. These may be the predominant ER form aggressive than in the less aggressive phenotype of human
during fetallife. We have used a ligand binding assay, which breast cancer celllines [62]. However, reduced expression
does not distinguish between ERa and ERß. The ERß cDNA of some PKC isoforms, such as PKCll, is associated with
was recently cJoned from the rat prostate [52]. The DNA- increased neoplastic transformation in the mammary gland,
binding domain of this ERß is highly homologous to that in while expression of other isoforms, such as PKCEl is linked
the 'c1assical' ERa protein, while the C-terminalligand binding to a more aggressive neoplastic process [62, 63].
domain bears only 55% homology. Most estrogenic substances
or estrogenic antagonists compete with E2 for binding to both
Protein kinase C and dietary Ja! in take
ER subtypes with identical preference and potency, but tran-
script tissue distribution is quite different for the ERa and ERß Diet influences PKC activity. Ahigh-fat diet increases PKC
mRNA [53]. ERß mRNA expression is high in the ovary and activity [64], whereas caloric restriction inhibits PKC
prostate, moderate in uterus and testis, and low in the brain activity [65] in epidermal cells in Sencar mice. A high corn oil
[53]. The distribution of ER subtypes in the breast is still diet also enhances PKC activity in the colon and carcinogen-
unclear. However, human breast cancer celllines appear to induced colon tumors in male rats [66]. Further, a diet high
express less ERß mRNA than normal breast tissues [53, 54]. in corn oil can block the inhibition of skin carcinogenesis, and
Taken together, the observations described above strongly reverse the reduction in PKC activity induced by moderate
suggest that both the timing and duration of exposure to a energy restrietion [67]. Our data indicate that similarly to skin
and colon, a diet high in n-6 PUFA increases PKC activity in
Tahte 3. Observations associating breast cancer risk with in ulero estrogenic the mouse mammary gland [44]. Since a high-fat intake
exposure. increases mammary tumor incidence in animal models [5], the
higher PKC activity in the mammary glands of the high-fat fed
Dizygotic twins have a high estrogenic pregnancy environment. The mice paralleJs the observed association between malignant
daughters ofthese pregnancies have an increased breast cancer risk [75-
progression and high PKC activity.
77].

High birth weigh is associated with high levels of estrogens during


pregnancy. The daughters ofthese pregnancies have an incrcascd brcast Pro tein kinase C and dietary Jat exposure during sensitive
cancer risk [78, 79]. periods
Preeclamptic and eclamptic pregnancies are accompanied by low serum
estrogen levels. The daughters of these pregnancies have a lower breast In contrast to the results obtained in adult mice fed with a high
cancer risk [80,81]. n-6 PUFA diet, PKC activity appears to be reduced in the
9

mammary gland in the offspring of mothers that were kept Carcinogen administration
on a high-fat diet during pregnancy [44]. It is possible that
the reduced PKC activity reflects specific morphological I
changes in the mammary gland that increase susceptibility
to neoplastic transformation. In utero exposure to a high-fat
diet alters the normal development of a mammary gland [13,
44]. The number of terminal end buds, structures that are the
targets of malignant transformation in the rodent mammary
gland and possibly in the human breast [68, 69], is higher in
the high-fat offspring than in a low-fat offspring. Thus, these
structures that are sensitive to neoplastic changes, persist in
the mammary glands offemale mice exposed to a high-fat
diet in utero. Persistent TEBs also have been reported in
transgenic mice [70] and in rats exposed to either a high-fat Adult Mammary Gland
diet or estradiol (E2) in utero, or during early postnatal period - increased epithelial proliferation
- increased ER levels
[13, 71]. All these groups exhibit an increased incidence of - increased protein kinase C activity
malignant growth in the mammary glands. These findings - increased promotion of transformed cells
suggest that low PKC activity (as well as low ER content) in
the mammary gland may increase the subsequent susceptibility
to develop mammary tumors. Increased breast
cancer incidence

Conclusions
In conclusion, a diet high in PUFAs, when exposure occurs
during adult life (mammary gland is weil developed)
increases the amounts of ER and PKC in the mammary
gland. These events may be linked to the fat-induced
increase in mammary tumorigenesis in animal models [5]
and the effects (or lack of them) on breast cancer risk in
humans [3, 7, 72]. In marked contrast, matemal intake of a Birth

/
high-fat diet during pregnancy reduces the ER content and
PKC activity in the offspring's mammary gland and alters
the gland's state of differentiation (Fig. 1). Since these
offspring are at an increased risk to develop spontaneous Developing Mammary Gland
and carcinogen-induced mammary tumors [12, 13], low - reduced epithelial differentiation (more TEBS. fewer ABs)
amounts ofER and low PKC activity in the mammary gland - lower ER levels
- lower protein kinase C activity
may predict an increased breast cancer risk. In addition, the - increased susceptibility to transformation
expression of ERß may be higher in the fetal than adult
mammary gland [82], and therefore a matemal high-fat diet
may affect offsprings' breast cancer risk by affecting this
novel ER sub type. Our future studies will determine whether Carcinogen administration ----i.~ Increased breast
cancer incidence
a matemal high-fat intake specifically affects the ERa or ERß
in the offspring. We also plan to investigate whether the low
PKC activity re fleets areduction in activity of a specifie Fig. 1. Plausible mechanisms mediating the effects of an exposure to a
isoform ofPKC family of genes. diet high in polyunsaturated fatly acids (a) after a carcinogen administration
or (b) in utero on breast cancer risk. TEB - terminal end bud; AB - alveolar
bud; ER - estrogen receptor (ERa and ER~); PKC - Protein kinase C.

Acknowledgments
Health Service (P50-CA58185 to R.C.) and the Lombardi
This work was supported by grants from the American Cancer Center Shared Animal Resource Facility (P30-
Cancer Society (CN80420 to L.H.-C.), and the Public CA51008).
10

References hormone changes and differences between Sprague-Dawley and F344


rats. J Nat! Cancer Inst 59: 1279-1283, 1977
20. Ip C, Ip MM: Serum estrogens and estrogen responsiveness in 7,12-
1. Boyd NF, Cousins M, Lockwood G, Tritchler D: The feasibility of dimethylbenz(a) anthracene-induced mammary tumors as influenced
testing experimentally the dietary fat-breast cancer hypothesis. Br J by dietary fat. J Natl Cancer Inst 66: 291-295,1981
Cancer 62: 878--881, 1990 21. Hilakivi-Clarke L, Onojafe I, Raygada M, Cho E, Clarke R, Lippman
2. Carroll KK, Khor HT: Dietary fat in relation to tumorigenesis: In: M: High-fat diet during pregnancy increases breast cancer risk in rats.
Progress in Biochemistry and Pharmacology. Basel, Karger, 1975, pp J Nat! Cancer Inst 88: 1821-1827,1996
308--353 22. Welsch CW: Review of the effects of dietary fat on experimental
3. Howe GR, Hirohata T, Hislop TG, Iscovich JM, Yuan JM, Katsoyanni mammary gland tumorigenesis: Role oflipid peroxidation. Free Radic
K, Lubin F, Marubini E, Modan B, Rohan T: Dietary factors and risk Biol Med 18: 757-773, 1995
ofbreast cancer: combined analysis of 12 case-control studies. J Nat 23. Stubbs CD, Smith AD: Modification of mammahan membrane
Cancer Inst 82: 561-569,1990 polyunsaturated fatty acid composition in relation to membrane fluidity
4. Boyd NF, Martin LJ, Noffel M, Lockwood GA, Tritchler DL: Ameta- and function. Biochim Biophys Acta 779: 89-137, 1984
analysis of studies of dietary at and breast cancerrisk. Br J Cancer 68: 24. Rose DP: Effects of dietary fatty acids on breast and prostate cancers:
627--{)36, 1993 Evidence from in vitro experiments and animal studies. Am J Clin
5. Welsch CW: Relationship between dietary fat and experimental Nutr 66: 1513S-1522S, 1997
mammary tumorigenesis: A review and critique. Cancer Res 52: 2040- 25. Greenblatt RB, Colle ML, Mahesh VB: Ovarian and adrenal steroid
2048,1992 production on the postmenopausal women. Obstet Gynecol47: 383-
6. Fay MP, Freedman LS:Analysis of dietary fats and mammaryneoplasms 387,1976
in rodent experiments. Breast Cancer Res Treat 46: 215-223,1997 26. Noble LS, Simpson ER, Johns A, Bulun SE: Aromatase expression in
7. Hunter DJ, Spiegelman D,Adami HO, Beeson L, van den Brandt PA, endometriosis. J Clin Endocrin Metab 81: 179 1996
Folsom AR, Fraser GE, Goldbohm RA, Graham S, Howe GR, Kushi 27. Bruning PF, Bonfrer JMG: Possible relevance of steroid availability
LH, Marshall JR, McDermottA, MillerAB, SpeizerFE, WolkA, Yuan and breast cancer. Arm NY Acad Sci 257-264, 1990
S, Willett W: Cohort studies offat intake and the risk ofbreast cancer 28. Holderegger C, Keefer D: The ontogeny of the mouse estrogen
- a pooled analysis. N Engl J Med 334: 356--361, 1996 receptor: The pelvic region. Am J Anatomy 177: 285-297,1986
8. Hilakivi-Clarke LA: Mechanisms of high matemal fat intake during 29. Muldoon TG: Steroid hormone receptor regulation by various
pregnancy on increased breast cancer risk in female rodent offspring. hormonal factors during mammary development and growth in the
Breast Cancer Res Treat 46: 119-214, 1997 normal mouse. Ann NY Acad Sci 464: 17-36, 1986
9. Adlercreutz H, Gorbach SL, Goldin BR, Woods MN, Dwyer JT, 30. Anderson E, Clarke RB, Howell A: Changes in the normal human breast
Hamalainen E: Estrogen metabo1ism and excretion in oriental and throughout the menstrual cycle: Relevance to breast carcinogenesis.
caucasian women. J Nat! Cancer Inst 86: 1076--1082, 1994 Endocrine-related Cancer 4: 23--33, 1997
10. Thomas DB, Karagas MR: Cancer in first and second generation 31. Kelly PA, Dijiane J, Malancon R: Characterization of estrogen,
Americans. Cancer Res 47: 5771-5776, 1987 progesterone and glucocorticoid receptors in rabbit mammary glands
11. Ziegler RG, Hoover RN, Pike MC, HildesheimA, NomuraAMY, West and their measurement during pregnancy and lactation. J Steroid
DW, Wu-Williams AH, Kolonel LN, Horn-Ross PL, RosenthaI JF, Bioehern 18: 215-221,1983
Hyer MB: Migration patterns and breast cancer risk in Asian-American 32. Yu WC, Leung BS: Variation of cytoplasmic content of estrogen and
women. J Natl Cancer Inst 85: 1819-1827, 1993 progesterone receptors in the mammary gland and uterus of rats at
12. Walker BE: Tumors in female offspring of control and diethylstilbestrol- time ofparturition. Biol Reprod 27: 658--{)64, 1982
exposed mice fed high-fat diets. J Nat Cancer Inst 82: 50-54, 1990 33. Haslam SZ, Nummy KA: The ontogeny and cellular distribution of
13. Hilakivi-Clarke L, Clarke R, Onojafe I, Raygada M, Cho E, Lippman estrogen receptors in normal mouse mammary gland. J Steroid Biochem
ME: A matemal die! high in n-6 polyunsaturated fats alters mammary Mol Bio142: 589-595, 1992
gland development, puberty onset, and breast cancer risk among female 34. McGuire WL, Carbone PP, Sears ME, Escher GC: Estrogen receptors
rat ol/spring. Proc Nat! Acad Sei USA 94: 9372-9377, 1997 in human breast cancer: An overview. In: W.L. McGuire, P.P. Carbone,
14. Clarke R, Dickson RB, Lippman ME: Hormonal aspeets of breast M.E. Sears, G.C. Escher (eds). Estrogen Receptors in Human Breast
cancer. Growth factors, drugs and stromal interactions. Crit Rev Oncol Cancer. Raven Press, New York, 1975, pp 1-7
Hematol12: 1-23, 1992 35. Netto GJ, Cheek JH, Zachariah NY, Romer JC, Lee CF, Schoenfeld
15. Goldin BR, Adelercreutz H, Gorbach SL, Warram JH: Estrogcn R, Chakmakjian ZH: Steraid receptors in benign mastectomy tissue.
excrelion patterns and plasma levels in vegetarian and omnivorous Am J Clin Pathol94: 14-17, 1990
women. N Engl J Med 307: 1452-1547, 1982 36. Blancho G, Alavaikko M, Ojala A, Collan Y, Hietanen T, Aine R,
16. Bennett FC, Ingram DM: Diet and female sex hormone concentrations: Taskinen PJ: Estrogen and progesterone receptos in breast cancer:
an intervention study forthe type offat consumed. Am J Clin Nutr 52: Relationships to tumor histopathology and survival of patients.
808-812,1990 Anticancer Res 4: 383-390, 1984
17. Rose DP, Boyar AP, Cohen C, Strong LE: Effect of a low-fat diet on 37. Singh L, Wilson AJ, Baum M, Whimster WF, Birch /H, Jackson IM,
hormone levels in women with cystic breast disease. J Nat! Cancer Lowrey C, Palmer MK: The relationship between histological grade,
Inst 78: 623--{)26, 1987 oestrogen receptor status, events and survival at 8 years in the NATO
18. Rose DP, Connolly JM, Chlebowski RT, Buzzard IM, Wynder EL: (Nolvadex) trial. Br J Cancer 57: 612--{j14, 1988
The effects of a low-fat dietary intervention and tamoxifen adjuvant 38. Henry JA, Nicholson JA, Farndon JR, Westley BR, May F: Measurement
therapy on the serum estrogen and sex hormone-binding globulin of oestrogen mRNA levels in human breast tumors. Br J Cancer 58:
concentrations of postrnenopausal breast cancer patients. Breast Cancer 600--{)05, 1988
Res Treat 27: 253-262, 1993 39. Anatnades K, Spector H: Quantitative estrogen receptor values and
19. Chan PC, Head JF, Cohen LA, Wynder EL: Influence of dietary fat on growth of carcinoma ofthe breast before surgieal intervention. Cancer
the induction of mammary tumors by nitrosomethylurea: Associated 50: 793-796, 1982
11

40. McycrsJS, RaoBR, Stevens SC, White WL: Lowineidenceofestrogen 61. O'Brian CA, Ward N: Bio10gy ofthe protein kinase C farnily. Cancer
reeeptor in breast careinornas with rapid rates of eellular proliferation. Metast Rev 8: 199--214, 1989
Cancer 40: 2290-2298, 1977 62. Ways DK, Kukoly CA, de Vente J, Hooker JL, Bryant WO, Posekany
41. Aarnda1 S, Bormer 0, Jorgensen 0, Host H, Ei1assen G, Kaa1hus 0, KJ, Fleteher DJ, Cook PP, Parker PP: MCF-7 breast cancer cells
Pih1 A: Estrogen receptor and 10ng term prognosis in breast cancer. transfected with PKC-alpha exhibit altered expression of otherprotein
Cancer 53: 2525-2529,1984 kinase C isoforms and displayamore aggressive neoplastic phenotype.
42. Clark GM, McGuire WL: Seroid receptors and other prognostic factors J Clin Invest 95: 1906-1915, 1995
in prirnary breast cancer. Sernin Oncol15: 20-25,1988 63. Manni A, Buckalter E, Etindi R, Kunse1man S, Rossini A, Mauger D,
43. Honig SF: Ireatment ofrnetastatie disease: hormonal therapy and cherno- Dabbs D, Demers 1: Induetion of less aggressive breast cancer
therapy. In: J.R. Harris, M.E. Lippman, M. Morrow, S. Hellman (eds). phenotype by protein kinase-alpha and beta overexpression. Cell
Diseases ofthe Breas!. Lipponcott-Ravcn, Philadelphia, 6, pp 669--734 GrowthDiff7: 1187-1198,1996
44. Hilakivi-Clarke LA, Raygada M, Stoica A, Martin M-B: Conswnption 64. Choe M, Kris ES, Luthra R, Copenhaver J, Pelling JC, Donnelly TE,
of a high-fat diet during pregnancy alters estrogen receptor content, Birt DE: Protein kinase C is aetivated and diaeylg1yeero1 is elevated
protein kinase C activity and morphology of marnmary gland in the in epidermal eells from SENCAR mice fed high fat diets. J Nutr 122:
mother and her female offspring. Cancer Res 58: 654-660, 1998 2322-2329, 1992
45. Potter JD, Cerhan JR, Seilers IA, McGovern PG, Drinkard C, Kushi 65. Kris E, Choe M, Luthra R, Conway H, Barnett I, Yakine A, Birt DE:
LR, Fo1sorn AR: Progcstcronc and cstrogen receptors and rnarnmary Protein kinase C activity is reduced in epidermal cells from energy
neop1asia in the lewa Wornen' s Health Study: How rnany kinds of restricted SENCAR mice. J Nutr 124: 485-492, 1994
breast cancer are there. J Natl Cancer Inst 85: 32-36, 1995 66. Reddy BS, Simi B, Patel N, Aliaga C, Rao CV: Effect of amount and
46. Holm L-E, Nordevang E, Hjalrnar ML, Lidbrink E, Callrner E, Nilsson types of dietary fat on intestinal bacterial 7 a1pha-dehydroxylase and
B: Ireatrnent failure and dietary habits in wornen with breast cancer. J phosphatidylinosito1-specific phospholipase C and colonic mucosa1
Natl Cancer Inst 85: 32-36,1993 diacy1g1ycero1 kinase and PKC activities during different stages of
47. Willett WC, Hunter DJ: Prospective studies of diet and breast cancer. colon tumor promotion. Cancer Res 56: 2314-2320, 1996
Cancer 74: 1085-1089, 1994 67. Birt DE, Barnett T, Pour PM, Copenhaver J: High-fat diet blocks the
48. Wilte JS, Ursin G, Sierniatycki J, Ihornpson WD, Paganini-Hill A, inhibition of skin carcinogenesis and reductions in protein kinase C by
Haile RW: Diet and premenopausa1 bilateral breast cancer: A case- moderate energy restrietion. Mol Carcinogenesis 16: 115--120, 1996
contro1 study. Breast Cancer Res Treat 42: 243-251, 1997 68. Russo J, Gusterson BA, RogersAK, Russo IR, Wellings SR, van Zwieten
49. Russo IR, Medado J, Russo J: Endocrine influences on the rnamrnary MJ: Comparative study ofhuman and rat mammary tumorigenesis. Lab
gland. In: T.C. Jones, U. Mohr, R.D. Hunt (eds). Integument and Invest 62: 244-278, 1990
Mammary Glands. Springer-Verlag, Berlin, 1994, pp 252-266,1994 69. Russo J, Romero AL, Russo IR: Architectural pattern of the normal
50. Bern HA, Mills KT, Edery M: Estrogen-associated defects in rodent and caneerous breast under the influence of parity. Cancer Epidemiol
rnammary gland developrnent. In: J.A. McLach1an (ed.). Estrogens in Biomarkers Prev 3: 219-224, 1994
the Environment. Elsevier, The Netherlands, 1985, pp 319--326, 70. Krane IM, Leder P: NDF/heregulin induces persistenee of terminal
51. Verhoeven G, Vandoren G, Heyns W, Kuhn ER, Janssens JP, Teuwen end buds and adenocareinomas in the mammay glands of transgenie
D, Goddeeris E, Lesaffie E, DeMoor P: J Endocrino1 95: 357-368, 1982 mice. Oncogcne 12: 1781-1788, 1996
52. Kuiper GGJM, Enmark E, Pelto-Huikko M, Ni1sson S, Gustafsson 71. Hilakivi-Clarke L, Cho E, Raygada M, Kenney N: Alterations in
JA: Cloning of a novet estrogen reeeptor expressed in rat prostate and marnmary gland development following neonata1 exposure to estradio1,
ovary. Proc Natl Acad Sci USA 93: 5925-5930, 1996 transforming growth factor alpha, and estrogen reeeptor antagonist
53. Kuiper GGJM, Carlsson B, Grandien K, Enmark E, Haggb1ad J, ICI 182,780. J Cell Physio1170: 279--289,1997
Nilsson S, Gustafsson JA: Comparison oftbe ligand binding specificity 72. Wolk A, Bergström R, Hunter D, Willett W, Ljung H, Holmberg L,
and transeript tissue distribution of estrogen reeeptors alpha and beta. Bergkvist L, BruceA, Adami HO: A prospective study of association
Endoerinology 138: 863--870, 1997 of monounsaturated fat and other types of fat with risk ofbreast cancer.
54 V1adusie EA, Hornby AK, Guerra FK, Lupu R: Expression ofERbeta Arch Intern Med 158: 41-45,1998
in breast cancer cells: Does it p1ay an improtant role? Proc Am Assoc 73. Tonio1o PG, Levitz M, Ze1eniuh-JaequetteA, Banerjee S, Koenig KL,
Cancer Res 88, Abstract 1995, 1997 Shore RE: A prospective study of endogenous estrogens and breast
55. Nishizuka Y: Intracellu1ar signalling by hydrolysis of phospholipids cancer in postmenopausal women. J Natl Cancer Inst 87: 190-197,
and activation ofprotein kinase C. Science 258: 607--{i14, 1992 1995
56. Dekker LV, Parker PJ: Protein kinase C - a question of specifieity. 74. Ingram D, Nottage E, Ng S, Sparrow L, RobertsA, Willcox D: Obesity
Trends Bioehern Sei 19: 73-77, 1994 and breast cancer. Cancer 64: 1049--1053, 1989
57. Cutler RE, Maize1s ET, Hunzicker-Dunn M: delta protein kinase-C in 75. Weiss HA, Potischman N, Brinton LA, Brogan D, Coates RJ,
the rat ovary: Estrogen regulation and localization. Endocrino1ogy 135: Gammon MD, Malone KE, Schoenberg JB: Prenata1 and perinatal
1669-1678, 1994 risk factors forbreast cancer in young women. Epidemiology 8: 181-
58. Tzukerman M, Zhang XK, Pfahl M: Inhibition of estrogen reeeptor 187, 1997
aetivity by the tumor promoter 12-0-Tetradecanylphorbol- 13 -Acetate: 76. Braun MM, Ahlbom A, Floderus B, Brinton LA, Hoover RN: Effect
A molecular analysis. Mol Endocrino1 5: 1893-1992, 1991 oftwinship on incidence of cancer ofthe testis, breast, and other sites.
59. Saeeda M, Knabbe C, Diekson RB, Lippman ME, Bronzert D, Lindsey Cancer Causes Control 6: 519-524, 1995
RL, Gottardis M, Martin M-B: Post transcriptiona1 destabilization of 77. Swetdlow AJ, Stavola BLD, Swanwick MA, Maconochie NES: Risk
estrogen receptor mRNA in MCF -7 eells by 12-0-tetradecano1phorbo1- of breast and testicu1ar cancers in young adult twins in England and
13-aeetate. J Biol Chem 266: 17809--17814, 1991 Wales: evidenee on prenatal and genetie aetio10gy. Laneet 350: 1723-
60. Martin M-B, Garcia-Morales P, Stoica A, Harrison BS, Pierce M, In~lm .
Katz D, Zhang S, Danielson M, Saceda M: Effects of 12-0- 78. Michels KB, Trichopoulos D, Robins JM, Rosner BA, Manson JE,
tetradecanoylphorbol-13-acetate on estrogen receptor activity in Hunter D, Colditz GA, Hankinson SE, Speizer FE, Willett WC: Birth-
MCF-7 cells. J Bio1 Chem 270: 25244-25251,1995 weight as a risk factor for breast cancer. Lancet 348: 1542-1546, 1996
12

79. Sanderson M, Williarns M, Malone KE, Stanford JL, Ernanuel I, White 81. Ekborn A, Hsieh C-c, Lipworth L, Adarni H-O, Trichopoulos D:
E, Daling JR: Perinatal factors and risk ofbreast cancer. Epiderniology Intrauterine environment and breast cancer risk in warnen: A population-
7: 34-37,1996 based study. J Natl Cancer Inst 89: 71-76, 1997
80. Ekborn A, Trichopoulos 0, Adarni HO, Hsieh CC, Lan SJ: Evidence 82. Brandenberger AW, Tee MK, Lee JY, Chao V, Jaffe RB: Tissue
ofprenatal influences on breast cancer risk. Lancet 340: 101S-1018, distribution of estrogen receptor a (ERa) and ß (ERß) rnRNA in the
1992 rnid gestational human fetus. J Clin Endocrinol Metab 82: 3509-3512
Molecular and Cellular Biochernistry 188: ß-20, 1998.
© 1998 Kluwer Acadernic Publishers.

Dissociation ofvitamin D3 and anti-estrogen


mediated growth regulation in MCF-7 breast
cancer ceUs
Elizabeth Nolan, 1 Manjula Donepudi, 1 Kathryn VanWeelden,I,2 Louise
Flanagan1,3 and JoEllen Welsh 1
] W. Alton Jones Cell Science Center, Lake Placid, New York; 2Clarkson University, Potsdam, New York, USA and
3University College Dublin, Ireland

Abstract
Our studies have identified 1,25(OH)P3 as a coordinate regulator ofproliferation and apoptosis in breast cancer cells. In MCF-
7 cells, 1,25(OH)2DJ down regulates the estrogen receptor (ER), suggesting that the effects of 1,25(OH)2D3 may be linked to
disruption of estrogen regulated survival signals. Although studies have demonstrated that 1,25(OH)2D3 inhibits growth ofER
negative breast cancer cells, previous data were generated by comparison of celllines derived from heterogeneous human tumors
and harboring diverse genetic alterations. To provide more conclusive evidence for independent growth regulatory pathways
mediated by antiestrogens and 1,25(OH)P3' we examined vitamin D J sensitivity in MCF-7 cells selected for resistance to ICI
182,780 (Zencca, Macclesfie1d, UK). The clones we selected for resistance to ICI 182,780 retain functional VDR and undergo
1,25(OH)PJ mediated growth arrest and apoptosis, in vitro and in vivo, despite loss of estrogen dependence. Cell cycle data
indicate that treatment ofparental or anti-estrogen resistant MCF-7 clones with 1,25(OH)P3' in the presence or absence ofICI
182,780, increases the percentage of cells in GiG] while reducing the number of cells in S phase. In addition, 1,25(OH)P3
induces characteristic features ofapoptosis, including DNA fragmentation, in both parental and anti-estrogen resistant MCF-
7 cells. Furthermore, we report that cells selected for vitamin DJ resistance retain sensitivity to ICI 182,780 mediated growth
arrest and apoptosis. This work emphasizes that vitamin D3compounds and anti-estrogens trigger growth arrest and apoptosis
in breast cancer cells by distinct mechanisms, and that breast cancer cell sensitivity to 1,25(OH)2D3 is not diminished during
the progression to estrogen independence. (Mol CeH Biochem 188: 13-20, 1998)

Key words: vitamin D, anti-estrogens, apoptosis, MCF -7 cells, cell cycle

Introduction induces cell cycle arrest in GJG" up-regulation ofthe cell cycle
inhibitors p53 and p21, and de-phosphorylation ofthe retino-
1,25(OH)P3 is the biologically active form ofvitamin D3, a blastoma protein [4--7]. Furthermore, we and others have
fat soluble vitamin identified as an anti-rachitic factor in the demonstrated that 1,25(OH)P3 induces morphological and
early 1920s. Although initially identified as a hormonal biochemical features of apoptosis, up-regulates apoptosis
regulator of calcium homeostasis and bone remodeling, it has associated proteins and down regulates the anti-apoptosis
recently become apparent that 1,25(OH)PJ is also a potent pro tein bcl-2, in MCF-7 cells [4--7]. These studies have
regulator of cell proliferation and differentiation. In breast identified 1,25(OH)P3 as a co ordinate regulator of apoptosis
cancer cells, 1,25(OH)P3 has been identified as a potent and proliferation in MCF-7 cells [8].
negative growth regulator [1-3]. Treatment of estrogen In estrogen receptor (ER) positive breast cancer cells, the
dependent MCF-7 human breast cancer cells with 1,25(OH)P3 effects of I ,25(OH)2D3 are similar to those induced by anti-

Addressfor ojJprints: J. Welsh, W. Alton Jones Cen Seienee Center, 10 Old Barn Rd, Lake P1acid, NY 12946, USA
14

estrogens [9] and 1,25(OH)P3 mediated apoptosis in MCF- Cell growth, apoptosis and cell cycle
7 cells is preceded by down regulation ofER and decreased
expression of estrogen dependent genes such as the proge- For quantitation ofviable cell number, cells were grown in
sterone receptor and pS" [6, 12]. These findings have led to 24 well plates (4000 cells/well), treated with 1,25(OH)P3'
the hypothesis that, at least in MCF -7 cells, the effects of ICI 182,780 or ethanol and stained with crystal violet as
vitamin DJ may in part be related to disruption of estrogen previously described [15]. For flow cytometric analysis of
regulated proliferation and survival signals. If so, then cell cycle kinetics, cells were trypsinized, pelleted and fixed
sensitivity to 1,25(OH)P3 mediated growth arrest/apoptosis with ethanol prior to staining with propidium iodide as
could be reduced in estrogen independent breast cancer cells. described [6]. At least 10,000 cells per treatment were
Although studies with ER negative breast cancer celllines analyzed on a Coulter Epics XL cytometer and cell cycle
have indicated that 1,25(OH)2D3 effectively inhibits growth modeling was achieved with the Pheonix Flow Systems
independently of estrogen signaling [1, 13], previous data MultiCycle software.
were generated by comparison of distinct celllines derived For morphological assessment of apoptosis, cells were
from heterogeneous human tumors and harboring diverse plated on coverslips and treated with I ,25(OH)P3' ICI 182,780
genetic alterations. In the following studies, we have or both on the day aftcr plating. Aftcr 96 h, cclls were fixed
examined the relationship between sensitivity to 1,25(OH)P3 and incubated with Hoescht 33258, a fluorescent dye which
and estrogen independence in breast cancer cells using intercalates with DNA, to examine nuclei for characteristic
variants of MCF -7 cells selected for resistance to the pure apoptotic morphology. Cells were photographed under UV
steroidal anti-estrogen ICI 182,780 [14]. Our results fluorescence and phase contrast.
demonstrate that sensitivity to vitamin D3 mediated growth
arrest/apoptosis is maintained in human breast cancer cells
which have progressed to anti-estrogen resistance, and Western blottingfor steroid receptors
conversely, that cells selected for vitamin D3resistance retain
sensitivity to anti-estrogen mediated apoptosis. These data For immunoblotting ofVDR and ER, nuclear extracts were
support a model in which vitamin D 3 and anti-estrogens prepared as described [15], separated on SDS-PAGE and
mcdiatc thcir cffccts on proliferation and apoptosis ofbrcast transferred to nitrocellulose. Blots were incubated with
cancer cells via independent pathways. monoclonal antibodies directed against the VDR (clone 9A7y,
Neomarkers, Fremont, CA, USA) or ER (clone AER 320,
Neomarkers, Freemont, CA, USA) followed by horseradish
peroxidase conjugated anti-rat (for VDR) or anti-mouse (for
Materials and methods ER) secondary antibody and chemiluminescent detection.

Cell culture
In vivo studies
MCF-7 cells, obtained from ATCC, were routinely plated in
MEM media with 5% FBS and passaged every three days. The Ovariectomized ncr nu/nu mice (Taconic Fanns, Germantown,
selection and characterization of a vitamin D3 resistant MCF- NY, USA) were fed a low calcium (0.1 %) purified rodent
7 cellline has already been described [15]. Briefly, MCF -7 cells chow (Purina Test Diets, Richmond, IN) for one week prior
were selected for growth in 100 11M 1,25(OH)P3' and a stable to inoculation of MCF_7ICIRes cells (2 x 106 cells suspended
cellline [MCF -7D3Re,] which is resistant to the growth inhibitory in Matrigel) into the mammary fad pad. Tumor volumes were
effects of 1,25(OH)P3 was generated. The anti-estrogen monitored weekly by caliper measurement of the length,
resistant variant was similarly developed by plating MCF-7 width, and height, and the tumor volume was calculated using
cells in MEM media with 5% FBS and 10 11M ICI 182,780 the formula for a semi-ellipsoid: 4/3nr3/2. After three weeks,
(Zeneca, Macclesfield, UK). Continued culture of surviving tumor volumes averaged approximately 50 mm 3 and mice
cells in 10 11M IC1182,780 for over one year has resulted in a were randomly divided into control or vitamin D3 treatment
stable cellline [MCF _7ICIReS] which grows equally weil in the groups. Due to the in vivo side effects of 1,25(OH)P3'
presence or absence oflCI 182,780. MCF-7 variants are vitamin D3 treatment was achieved with the low calcemic
routinely grown in media containing 100 nM 1,25(OH)P3(for synthetic analog, EB 1089 (LEO Pharmaceuticals, Ballerup,
MCF_7DJRes cells) or 10 11M ICI 182,780 (for MCF_7 IC1Rcs cells). Denmark). EB 1089 was suspended in 80% propylene
For experiments, cells were plated in media containing 5% FBS glycol/20% PBS and administered via subcutaneous injection
without I ,25(OH)P3' ICI 182,780 or ethanol, and treatments at a dose of 60 pmol/day. This dose of EBI089 has been
as indicated in the figure legends were initiated on the day shown by our lab to induce apoptotic regression ofMCF-7wT
following plating. tumors [16]. Control mice received daily injections of the
15

vehicle alone. Body weights and tumor volumes were both ER and VDR, are presented in the top panel of Fig. 1.
monitored weekly for five weeks. Treatment with either 100 nM 1,25(OH)P3 or 10 11M ICI
182,780 significantly decreased MCF _7 WT cell number (40-
50% of control values) after 96 h, and co-incubation with both
Results compounds had an additive effect (20% of control). In MCF-
71CIRes cells (Fig. 1, bottom panel), treatment with ICI
EjJect 0/ anti-estrogens and vitamin D) on growth 0/ 182,780 had minor effects on cell number (85% of control)
parental and anti-estrogen resistant MCF-7 cells but cell number was significantly lower after treatment with
1,25(OH)2D) (36% of control). Co-incubation with both
The interactions between 1,25(OH)P3 and ICI 182,780 in 1,25(OH)2D) and ICI 182,780 had additive effects on cell
modulating growth of parental MCF _7 WT cells, which express number in MCF71CIRes cells, similar to that observed in MCF-
7WT cells.

Cell cycle analysis


MCF-7 WT

The percentage of cells in each phase of the cell cyc1e was


determined by flow cytometry in both cell lines after 48 h
treatment with 100 nM 1,25(OH)Pl' 10 uM ICI 182,780 or
both compounds. As presented in Table I, treatment ofMCF-
7WT cells with I ,25(OH)P3 induced an accumulation of cells
in GiG, and reduced the percentage of cells in S phase.
Treatment ofMCF-7 wT cells with ICI 182,780 also increased
the percentage of cells in G/G 1 and reduced the percentage
of cells in S phase, mimicking the effects of 1,25(OH)Pl'
In MCF _71C1Res cells, treatment with ICI 182,780 induced a
slight increase in the percentage of cells in GiG I' but had no
effect on S phase distribution. Interestingly, treatment with
ICI 182,780 decreased the percentage ofcells in G/M to the
same extent in both MCF_7ICIRes and MCF_7 wT cells. Despite
the loss of sensitivity to anti-estrogen mediated growth arrest,

Table I. Effects of 1,25(OH),D, and ICI 182,780 on cell cycle kinetics of


parental and antiestrogen resistant MCF-7 eell variants.
MCF-7 ICIRB.
%G I %S %G,

MCF_7 w TCELLS
Ethanol 54.0 34.4 11.6
100 nM D, 63.8 26.9 9.3
10 11M ICI 182,780 81.1 11.3 7.6
D, +ICI 84 7.4 8.6

MCF_7 ICIR o< CELLS


Ethanol 58.6 20.4 21.0
100 nM D) 66.5 15.0 18.5
ICI&D3 10 11M ICI 182,780 63.3 22.1 14.6
D, + ICI 77.7 8.6 13.7

Cultures of MCF-7 WT and MCF_7 ICIRe< eells were harvested after 48 h


Fig. 1. Effects of 1,25(OH),D, and ICI 182,780 on viable cell number of treatment with 100 nM I ,2 5(OH),oJ' in the presence or absence of 10 11M
MCF-7 wT and MCF_7 ICIR" cells. MCF-7 wT (Top) and MCF_7,cIR" (Bottom) ICI 182,780, or ethanol vehicle. Cells were trypsinized, fixed in ethanol
cells were plated in 24 well plates (4000 cells/well) and treated the day and suspended in propidium iodide/RNAse at a concentration of 106 cells/
after plating with ethanol vehide, 100 nM 1,25(OH),D3 , 10 (lM ICI 182,780 ml. At least 10,000 cells were analyzed by flow cytometry for each
or both 100 nM 1,25(OH),D, and 10 (lM ICI 182,780. Cell number was treatment, and eell eyde parameters were genera ted using the Multi-Cyde
quantitated after 96 h by crystal violet staining as described in Materials AV software (Phoenix Flow Systems). Similar data were obtained in two
and methods. independent experiments.
16

the effeets of 1,25(OH)P3 on eell eycle ofMCF_7 IClRe , eells Table 2. Effects of 1,25(OH),D, and ICI 182,780 on cell cyclc kinetics of
were eomparable to those observed in MCF-7 wT cells. MCF-7 DR" cultures.

In MCF-7 WT cells, treatment with both 1,25(OH)2D3 and 0/0 GI %S %G z


ICI 182,780 exerted an additive effect on cell eycle kinetics
MCF_7 DRES CELLS
compared to eithertreatment alone. Surprisingly, co-ineubation Ethanol 55.3 31.2 13.5
of MCF _7ICIRes cells with 1,25(OH)2D3 and ICI 182,780 IOD nM 0] 59.6 30.0 10.5
resulted in a further decrease in the percentage of cells in S 10 flM ICI 182,780 84.8 6.3 8.8
phase compared to either treatment alone, suggesting that D,+ICI 85 4.1 10.9
1,25(OH)2D3 may sensitize MCF_7ICIRes cells to anti-estrogen MCF_7 DR " cells were harvested after 48 h treatment with 100 nM
mediated growth arrest. 1,25(OH),D" in the presence or absence of 10 flM ICI 182,780, or ethanol
vehicle. Cells were analyzed by flow cytometry as described in legend for
Table I. Similar data were obtained in lwo independent experiments.

Effect of [CI 182, 780 on growth ofMCF_7D3R" cells


1,25(OH)P3 suggesting that ICI 182,780 did not sensitize
We have previously characterized a vitamin D 3 resistant MCF_7 DJRos cells to 1,25(OH)P3'
variant ofMCF-7 cells [MCF_7 D3Re<] and reported that these
cells retained sensitivity to growth arrest/apoptosis mediated
by the non-steroidal anti -estrogen tamoxifen [15]. In the Morphology ofMCF-7 cell variants treated with
current studies, we examined whether MCF_7DJRes cells 1,25(OH)P3 and [CI 182,780
retained sensitivity to the effects of ICI 182,780, using
approaches similar to those described above. As presented The morphology ofthe MCF-7 cell variants was examined
in Fig. 2, 1,25(OH)2D3 had minimal effects on viable cell by phase contrast and epi-fluorescence after staining with
number in MCF_7 D3Re< cultures (90% of control), but ICI the DNA interactive dye, Hoechst 33258, which enables
182,780 significantly redueed cell number (30% of control). identification ofboth mitotic and apoptotic cells. In ethanol
No potentiating effects were observed in MCF_7 D3Res cultures treated control cultures of MCF-7wT cells, the majority of
treated with both ICI 182,780 and 1,25(OH)PJ' cells were quiescent, with normal nuc1ear morphology
Analysis of cell cycle (Table 2) indicated that 1,25(OH)PJ visualized by Hoechst staining (Fig. 3A). Mitotic eells were
did not induce G,jG I arrest or reduee the percentage of cells detected by their characteristic chromatin patterns in Hoechst
in S phase in MCF_7 D3Re> cultures. However, the effects ofICI stained control cultures (arrows). After 96 h treatment with
182,780 on cell cycle kinetics of MCF _7D3Res cells were 1,25(OH)PJ' mitotic cells were no longer detectable, and
comparable to that observed in MCF7 wT cells (refer to Table cells began to display characteristic apoptotic morphology,
I). In MCF_7 D3Res cells, cell cycle kinetics were similar in cells inc1uding condensation, retraction from neighboring cells
treated with lCI 182,780 in the presence or absence of and pyknotic nuclei (arrows). A distinet morphology was
observed in MCF _7 WT cultures treated with ICI 182,780,
with cells growing in small clusters and exhibiting apoptotic
0.75 nuc1ei (arrows). MCF-7 WT cultures co-incubated with both
MCF-7 D3Res
1,25(OH)2DJ and ICI 182,780 consisted almost entirely of
cells exhibiting apoptotic morphology.
In the absence oftreatment, the morphology ofMCF_71ClRcs
cultures (Fig. 3B) was similar to that ofthe MCF-7 wT cells
from which they were derived, consisting of quiescent and
mitotic (arrows) eells. After treatment with ICI 182,780,
however, MCF_7 IC1Re< cells increased in size and exhibited a
flattened, rounded morphology, but no apoptotic cells were
present. In contrast, treatment of MCF_7ICIRos cultures with
I ,25(OH)P3 reduced cell number and induced characteristic
apoptotic morphology (arrows) and after co-incubation with
both 1,25(OH)P3 and ICI 182,780, the fewremaining MCF-
Fig.2. Effects of 1,25(OH),D] and IC1182,780 on viable cell nurnber of 7 ICIRos cells had apoptotic morphology.
MCF-7DR" cells. MCF-7 DR " cells were plated in 24 weil plates (4000 cells!
There were no obvious differences in morphology between
weil) and treated the day after plating with ethanol vehicle, 100 nM
1,25(OH),D" 10 flM ICI 182,780 orboth 100 nM 1,25(OH),D] and 10 flM
MCF_7 wT and MCF_7 DJRos cells (Fig. 3C) after vehicle
ICI 182,780. Cell nurnber was quantitated after 96 h by crystal violet staining treatment. In MCF_7 D3Re; cultures treated with 1,25(OH)2D3'
as described in Materials .nd methods. cells continued to proliferate (mitotic cells marked by
17

arrows), and no apoptotic cells were visible. However,


A treatment with IC1182,780 induced characteristic apoptotic
morphology in MCF_7 D3R'" cuItures. After treatment with
both 1,25(OH)P3 and lCI 182,780 MCF _7 D3 !<o; cultures
were similar in morphology to those treated with ICI 182,780
alone.
Analysis of DNA fragmentation by in situ end-Iabeling
(TUNEL) indicated that I ,25(OH)P, induced DNA fragmen-
tation in MCF _7 WT and MCF _7ICIRes cells but not in MCF7 D3R,"
cells [15 and data not shown]. Treatment with ICI 182,780
increased the percentage of TUNEL positive cells in MCF-
7WT and MCF-7 D3R'" cultures but not in MCF -TelRes cultures.
These data are consistcnt with the morphological features of
MCF-7 apoptosis evident in Fig. 3.

B
Expression of ER and VDR in MCF-7 variant cel/fines.

To explore the possihility that se]eclion for resistance to


vitamin D3 Of anti-estrogens induced changcs in stcroid
receptor expression, untreated cells were subjectcd to
Western bJotting with specific monoclonal antibodies which
recognizc VDR or ER. As demorrstrated in Fig. 4, MCF-7 wT
cells expressed both VDR and ER. MCF-7 D3R" cells expressed
higher levels ofboth VDR and ER than did lvICF _7 WT cells and
evcn higher expression ofVDR was detcctcd in the MCF-
7ICIRes variant. In contrast, the lowest levels ofER expression
were observed in the MCF_7ICIRes cells.

c In vivo effect ofvitamin D, analog treatment on growth of


tumors derived from antiestrogen reststant cells.

Ovariectomized mice inoculated in the mammary fat pad with


MCF7ICIRes cells developed tumors with a 100% take rate.

43JL.-__-__-_---'I~ VOR

MCF_7D3Aes
58 1--- - I~ ER
~
Fig. 3 . Morphology of MCF-7 variant cell lines after treatment with
1,25(OH),D3 and ICI 182,780. MCF-7\VT cells (A), MCF_7 IC1Ro; cells (B) MCF-7 Cell Variant
and MCF_7 DRe> eells (C) were plated on coverslips and treated the day after
plating with ethanol vellicle, lOO nM 1,25(OH)P" 10 flM IC1182,780 or Fig. 4. VDR and ER expression in MCF· 7 variant cell lines. Nuc1ear
bolh 100 nM 1,25(OH)P, and 10 flM ICI182,780. After 96 h, cells were extracts derived from untreated MCF _7 WT , MCF _7ICIR~ and MCF -7DRoo cells
washed with PES, fixed alld illcubated with Hoechst 33258 (0.5 flg/ml) for were solubilized in Laemlli buffer, separated on SDS-PAGE (50 ug protein
one hour. After washing, coverslips were mounted and viewed under phase per lane), transferred onto nitrocellulose, and innl1unoblotted wilh antibodies
contrast and epillnorescence. against the VDR or ER as described in Materials and methods.
18

400,------------------------------------, retained comparable sensitivity to 1,25(OH)2D3 mediated


---Control
growth inhibition as the parental MCF_7 wT cells. Furthermore,
-+-EB1089
60 pmol in several ofthe ICI 182,780 resistantMCF-7 clonal celliines,
"'E' 300 sensitivity to 1,25(OH)P3 was higher than that ofMCF-7 wT
.§. cells (data not shown). Further, the growth oftumors derived
.,
E from MCF_7ICIRes cells was significantly reduced in mice treated
:::l 200

..
with the vitamin DJ analog EB 1089, an effect which we have
~
o
previously observed with tumors derived from the parental
E MCF _7 WT cells [16]. These studies indicate that disruption of
? 100
the estrogen signaling pathway in estrogen dependent breast
cancer cells does not necessarily disturb the growth inhibitory
pathways triggered by activation ofthc VDR.
O+-------r------.------.-------.-----~
o 4 5 Although ICI 182,780 did not alter growth of MCF _7ICIRes
Weeks of Treatment cells, this anti-estrogen appeared to sensitize MCF _7ICIRes
cells to the effects of 1,25(OH)P3' as cell number and the
Fig.5. Effect ofthe vitamin D3 analog EBI089 on growth ofMCF-7 IC1R" percentage of cells in S phase was lower in MCF _7 IC1R" cells
lumors in ovarieclomized nude mice. Ovarieclomized nude mice bearing treated with 1,25(OH)P3 in the presence ofICI 182,780 than
tumors derived from MCF7 IC1R" cells were Irealed by daily subcutaneous in the absence. In addition, morphological features of
injeclions of EB 1089 (60 pmol/day) or vehicle and tumor volume was
apoptosis and DNA fragmentation were enhanced in MCF-
monilored weekly by caliper measurement. Data poinls represent mean ±
S.E.M ofal leasl 11 mice per group per time point.'p < 0.05; '*p < 0.01; 7 JC1R " cultures after treatment with both ICI 182,780 and
***p< 0.001, control vs. EB1089 as delermined by Student"s unpaired t- 1,25(OH)P3 compared to 1,25(OH)P3 alone. These results
lesl. warrant additional studies to assess whether synergistic effects
ofvitamin D 3 analogs and ICI 182,780 can be demonstrated
in xenografts derived from MCF_7 IClRes cells.
Since MCF-7 wT tumors are dependent on the presence of The potential mechanism by which ICI 182,780 and
estrogen for growth in nude mice [16, 17], growth ofMCF- 1,25(OH)2D3 synergize to retard growth of MCF_7ICIRes cells
7lCIRes cells in ovariectomized mice confirms their estrogen is unclear, but may involve changes in expression or function
independence. Treatment oftumors derived from MCF_7ICIRes of the VDR. Western blotting of nuclear extracts indicated
cells with the vitamin D 3 analog EB 1089 significantly that despite very low levels of ER, the MCF -7lCIRes cells
reduced tumor volume within two weeks, with a two fold retained expression ofthe VDR. In fact, VDR expression was
reduction in final tumor size (Fig. 5). After five weeks, mice higher in MCF7lCIRes cells than in either MCF-7 WT or MCF-
treated with EB I 089 exhibited an elevation of serum calcium 7D3Res cells. These are the first studies to demonstrate that
(control, 8.6 ± 0.3 mg/100 ml, n = 11; EB1089 treated, 12.0 expression ofthe VDR is enhanced in cells selected for anti-
± 0.2 mg/100 ml, n = 16; P < 0_001) but only a slight loss of cstrogen resistance, and support the concept that a functional
body weight (data not shown). Thus, despite estrogen estrogen signaling pathway may down-regulate vitamin D3
independence and anti-estrogen resistance, MCF7lCIRes cells signaling in MCF-7 cells. Assessment of the effects of lCI
retained sensitivity to vitamin D J mediated regression when 182,780 on VDR expression in MCF-7 wT and MCF_?,CIRes
grown as tumors in nude mice. cells will be necessary to determine if the synergy between
antiestrogens and vitamin D J involves further up-regulation
ofthe VDR.
Discussion Having demonstrated that anti-estrogen resistant cells
retained sensitivity to vitamin D 3mediated growth arrest and
In these in vitro and in vivo studies, we have demonstrated apoptosis, we next examined whether MCF -7 cells selected
that vitamin D 3 and antiestrogens mediate their growth for 1,25(OH)2DJ resistance retained sensitivity to ICI
regulatory effects on MCF-7 cells via independent signalling 182,780. We have previously reported [15, 18] that MCF-
pathways. This conclusion is based on experiments in which 7D3Res cells are rcsistant to 1,25(OH)2D3 mediated growth
we compared growth, morphology and cell cycle of parental arrest and apoptosis yet express the estrogen receptor and
MCF-7 cells (MCF_7 WT cells) with variant celllines selected retain sensitivity to the anti-estrogen tamoxifen, which can
for resistance to the pure steroidal anti-estrogen ICI 182,780 exhibit both estrogenic and anti-estrogenic effects. In the
(MCF_7ICIRes cells) or to the biologically active vitamin D 3 current studies, we have extended those findings by demon-
metabolite 1,25(OH)2D3 (MCF_7 DJRes cells). strating that MCF _7 DJRes cells undergo cell cyc1e arrest and
Our studies clearly demonstrate that, despite estrogen exhibit morphologie al features of apoptosis after treatment
independence, cells selected for resistance to ICI 182,780 with ICI 182,780, a pure anti-estrogen [14]. Similar results
19

have been obtained with several distinct vitamin D j resistant in response to 1,25(OH)P3' This working model suggests
clones generated from two different MCF -7 stock cultures that, for patients with breast tumors containing estrogen
(data not shown), suggesting that the ability to dissociate dependent and estrogen independent cells, a distinct thera-
sensitivity to vitamin D j compounds from that of anti- peutic advantage may be achieved by combining therapies
estrogens is a general phenomenon. In addition, we have that activate VDR with those that disrupt estrogen signaling.
previously reported that tumors derived from the MCF- This concept is supported by studies demonstrating comple-
7D3Re , cells are resistant to EB 1089 treatment in vivo, yet mentary effects ofanti-estrogens and vitamin D3compounds
exhibit comparable regression upon removal of estrogen on growth ofER positive breast cancer cells and tumors [12,
supplementation as tumors derived from MCF _7 WT cells 20-22]. Perhaps most importantly, we have shown that breast
[16]. Combined with the studies on the MCF_7 IC1R.. tumors cancer cell sensitivity to 1,25(OH)P3 mediated growth
described here, these data support the hypothesis that arrest/apoptosis is not diminished during the progression to
dissociation between these two nuclear receptor signalling estrogen independence, indicating that vitamin D 3 analogs
pathways can also be demonstrated in vivo. may offer an effective therapeutic approach for aggressive,
Consistent with cross-talk between estrogen and vitamin late stage tumors which exhibit resistance to the standard anti-
Dj signaling, western blotting indicated that the ER was up- estrogen regimens.
regulated in MCF_7 D3Re< cells compared to either MCF-7 wT
or MCF _7 IC1Re< cells. This data suggests that disruption ofthe
vitamin D3 signaling pathway enhances estrogen sensitivity, Acknowledgements
and would imply that activation or up-regulation ofthe VDR
would downregulate estrogen signaling. This suggestion is This work was funded by operating grants from the NIH
consistent with reports that treatment of MCF-7 cells with (#CA69700) and the American Institute for Cancer Re-
1,25(OH)2D3 decreases ER and estrogen dependent gene search (#95B068). We would like to thank Dr. Lise Binderup
expression [6, 12]. Ofnote, expression ofthe VDR in MCF- (LEO Pharmaceuticals, Ballerup, Denmark) for supplying
7D3Rc. cells is not reduced relative to MCF-7wT cells. In previous EB 1089, and Dr. Alan Wakeling (Zeneca Pharmaceuticals,
studies, we have demonstrated that the VDR from MCF_7 D3Re, Macclesfield, UK) for supplying ICI 182,780.
cells binds 1,25(OH)2D3 and the vitamin D response element
[15]. Furthermore, no mutations in the nucleic acid sequence
ofthe VDR coding region from MCF_7 D3Re, cells has been
identified (Byrne and Welsh, in preparation). This data
References
indicates that lack of 1,25(OH)P3 responsiveness in MCF-
1. Chouvet C, Berger U, Coombes RC: 1,25 Dihydroxyvitamin D J
7DjRe, cells is not secondary to mutation or loss of the VDR, inhibitory effect on the growth of two human breast cancer celllines
implicating a post-receptor defect in development ofvitamin (MCF-7, BT-20). J Ster Biochem 24: 373-376,1986
Dj resistance. 2. Colston K. Berger U, Coombes RC: Possible role for vitamin D in
controlling breast cancer eell proliferation. Laneet I: 188--191, 1989
3. Eisman J, Sutherland RL, McMenemy ML, Fragonas JC, Musgrove
EA, Pang G: Effeets of 1,25-dihydroxyvitamin 0) on eell eycle kineties
Conclusions ofT47D human breast eaneereells. J Cell Physiol138: 611-{j16, 1989
4. Welsh JE: Induclion of apoptosis in breast cancer cells in response to
The data presented in this study emphasize that vitamin Dj vitamin 0 and antiestrogens. Biochem Cell Biol 72: 537-545, 1994
compounds and anti-estrogens regulate growth and apoptosis 5. Simboli-Campbell M, Narvaez CJ, Tenniswood M, Welsh JE:
1,25(OH)2D) induees morphologieal and biochemie al indices of
ofbreast cancer cells by independent mechanisms. Analyses
apoptosis in MCF-7 breast cancer eells. J Ster Bioehem Mol Biol 58:
ofhuman breast tumor biopsies have demonstrated that the ER 367-376, 1996
and VDR do not necessarily co-Iocalize within the same cell 6. Simboli-Campbell M, Narvaez CJ, VanWeelden K, Tenniswood M,
[19], indicating that these two nuclear receptors are presen! in Welsh JE: Comparative effeets of 1,25(OH),D) and EBI089 on eell
distinct target cells within a mixed tumor. Our working eycle kineties and apoptosis in MCF-7 cells. Breast Cancer Res Treat
42: 31-41, 1997
model suggests that in cells expressing both ER and VDR,
7. James SY, Maekay AG, Colston K: Effects of 1,25 dihydroxyvitamin
1,25(OH)P3' through the VDR, activcly induccs growth DJ and its analogues on induetion of apoptosis in breast cancer eells. J
arrest and facilitates apoptosis, whereas treatment with anti- Ster Bioehem Mol Bio158: 395-401,1996
estrogens disrupts ER signaling, negating the protective effect 8. Simboli-Campbell M, Welsh JE: 1,25 Dihydroxyvitamin D 3: Coordinate
of estrogen. In estrogen independent cells, the estrogen regulator of aelive eell death and proliferation in MCF-7 breast cancer
eells. In: M. Tenniswood and H. Michna (eds). Apoptosis in Honnone
signaling pathway is disabled, allowing for escape from
Dependent Cancers. Springer Verlag, Berlin, 1995 pp 181-200
estrogen dependence and resistance to anti-estrogen mediated 9. Welsh JE, Simboli-Campbell M, Narvaez CJ, Tenniswood M: Role of
apoptosis. However, the presence ofthe VDR in anti-estrogen apoplosis in the growlh inhibitory effects ofvitamin D in MCF-7 eells.
resistant cells renders them sensitive to activation of apoptosis Adv Exper Biol Med 375: 45-52,1995
20

10. Bursch W, EllingerA, Kienzl H, Torok L, Pandey S, Sikorska M, Walker 17. Kyprianou N, English H, Davidson N, Isaacs J: Prograrmned cell death
R, Schulte Hermann R: Active cell death induced by the anti-estrogens during regression ofthe MCF -7 human breast cancer following estrogen
tamoxifen and ICI 164 384 in human mammary carcinoma cells (MCF- ablation. Cancer Res 51: 162-166,1991
7) in culture: The role of autophagy. Carcinogenesis 17: 1595-1607, 1996 18. Navaez CJ, Welsh JE: Differential effects of 1,25-Dihydroxyvitamin
11. WarriA, Huovinen R, LaineA, Martikainen P, Harkonen P:Apoptosis 0) and TPA on cell cycle and apoptosis in MCF-7 cells and a vitamin
in toremifene-induced growth inhibition ofhuman breast cancer cells 0 3 resistant variant. Endocrinology 138: 4690-4698, 1997
in viva and in vitra. J Nat Cancer Inst 85: 1412-1418, 1993 19. Berger U, McClelland R, Wilson P, Greene G, HaussIer M, Pike J,
12. James SY, Mackay A, Binderup L, Colston K: Effects of a new synthetic Colston K, Easton D, Coombes RC: Immunoctyochemical determination
vitamin D analogue, EB 1089, on the oestrogen responsive growth of of estrogen receptor, progesterone receptor and 1,25(OH)2-vitamin D3
human breast cancer cells. J Endocrinol141: 555-563, 1994 receptor in breast cancer and relationship to prognosis. Cancer Res
13. Elstner E, Linker-Israeli M, Said J, Umiel T, deVos S, Shintaku IP, 51: 239-244,1991
Heber D, Binderup L, Uskokovic M, Koeffler HP: 20-epi Vitamin 0) 20. Abe-Hashimoto J, Kikuchi T, Matsumoto T, Nishii Y, Ogata E, Ikeda
analogues: A novel class of potent inhibitors of proliferation and K: Antitumor effect of 22-oxa-calcitriol, a noncalcemic analogue of
inducers of differentiation of human breast cancer cells. Cancer Res calcitriol, in athymic mice implanted with human breast carcinoma
55: 2822-2830,1995 and its synergism with tamoxifen. Cancer Res 53: 2534--2537,1993
14. Wake/ing AK, Dukes M and Bowler J: A potent specific pure anti- 21. Vink-van Wijngaarden T, Pols H, Buurman CJ, van den Bemd, G,
estrogen with c1inical potential. Cancer Res 51: 3867-3873, 1991 Dorssers L, Birkenhager J, van Leeuwen J: Inhibition ofbreast cancer
15. Narvaez CJ, VanWeelden K, Byrne I, Welsh JE: Characterization ofa cell growth by combined treatment with vitamin D) analogues and
vitaminD3 resistant MCF-7 cellline. Endocrinology 137: 400-409,1996 tamoxifen. Cancer Res 54: 5711-5717,1994
16. Van Weelden K, Flanagan L, Binderup L, Tenniswood M, Welsh J: 22. Love-Schimenti C, Gibson 0, Ratnam A and Bikle D: Antiestrogen
Apoptotic regression ofMCF-7 xenografts in nude mice treated with potentiation of antiproliferative effects of vitamin 0 3 analogues in
the vitamin D3 analog, EB 1089. Endocrinology 139: 2102-2110, 1998 breast cancer cells. Cancer Res 56: 2789-2794, 1996
Molecular and Cellular Biochemistry 188: 21-30, 1998.
© 1998 Kluwer Academic Publishers.

Sodium butyrate induces retinoblastoma protein


dephosphorylation, p16 expression and growth
arrest of colon cancer cells
Bertha Schwartz, 1,2 Carmel Avivi-Green l and Sylvie Polak-Charcon3
IJnstitute of Biochemistry, Food Science and Nutrition, Faculty ofAgriculture, Food and 2Environmental Quality Sciences,
The Hebrew University oflerusalem, Campus Rehovot and 3Pathology Department, Tel-Hashomer Medical Center, Tel-
Hashomer, Israel

Abstract
Sodium butyrate causes alteration of colon cancer cell morphology and biology towards that of a more differentiated phenotype.
The retinoblastoma gene encodes a nuelear phosphoprotein (pRb) present in a wide range ofhuman cancer celllines including
colon canccr cell lines. pRB is synthcsizcd throughout the cell cyele and phosphorylated in a phase speeifie manner: the
predominant proteins in GO/G 1 are the unphosphorylated species (110 kD) whereas phosphorylated pRb (112-114 kD) are in
Sand G2. 110 kD pRb binds transcription factors and prevents transcription of responsive genes such as the gene for thymidine
kinase, which are expressed in late GI. The precise mechanisms controlling cell arrest are unknown, but recent data suggest
that cyelin-dependent kinase inhibitors such as pl6 may playa role. The aim ofthe present study was to assess the effect of
sodium butyrate on cell cyc1e staging, thymidine kinase activity, phosphorylation ofthe pRb protein and expression ofp16.
We show that sodium butyrate treatment induces differentiation ofLS l74T colon cancer cells, inhibits thymidine kinase aetivity
coneomitantly with induction of pRb dephosphorylation, p 16 transcription and cell cyele arrest at GO/G 1. Initial dephos-
phorylation was observed 24 h after treatment of LS l74T cells with sodium butyrate, whereas complete shift to the
dephosphorylated form was observed 3 days after treatment. Induetion of pRb dephosphorylation by sodium butyrate preceded
inhibition of growth and the specific cell cyc1e arrest. RNase proteetion assay with a p 16 specific riboprobe showed undetectable
levels in proliferating cells to several fold increase in differentiated colonocytes.
In conelusion, the results provide evidence for a specifie cellular mechanism of butyrate induced growth arrest and
differentiation ofa colon cancer cellline. (Mol Cell Biochem 188: 21-30,1998)

Key words: Colon cancer, retinoblastoma, differentiation, proliferation, cyc1in kinases, cyc1in kinase inhibitors

Introduction expression of a variety of genes either at a transeriptional


level, through speeific promoter regions, or via posttrans-
Butyric acid is a short chain fatty acid and provides the most eriptional meehanisms [3,4].
important end product of eolonie bacterial fermentation of Treatment of colon cancer celllines with sodium butyrate
fiber and stareh. Butyrate along with other short chain fatty or short ehain fatty acids induces the expression of alkaline
acids released in the colon are rapidly taken up by luminal phosphatase and causes cell polarization and dome formation,
epithelial eolonic cells, where they serve as central energy established markers of intestinal eellular differentiation [1,
providers and additionally they affect proliferation and 2,5,6]. Differentiation is associated with withdrawal from
differentiation [I, 2]. In a vast number ofanimal cells butyrate the cell eyele to the GO stage. A number of proteins have been
has been shown to affect gene expression and cell growth. identified which playa role deterrnining progression through-
Recent studies suggest that sodium butyrate ean modulate the out the eell eyc1e, one ofthese protcins is the retinoblastoma

Addressfor offPrints: B. Sehwartz, Institute ofBioehemistry, Food Seienee and Nutrition, Faeulty ofAgrieulture, Food and Environmental Quality Seiences,
The Hehrew University of Jerusalem, Campus Rehovot 76100, Israel
22

protein (PRb). The retinoblastoma gene encodes the 105-114 Materials and methods
kD nuclear phosphoprotein pRb [7-10]. The presence of pRb
was demonstrated in a wide range ofhuman cancer celliines Chemieals
including those derived from colorectal carcinomas [11, 12].
Loss of pRb function, as for other tumor suppressor gene Sodium butyrate, RNaseA, propidium iodide were purchased
products, is associated with uncontrolled cell growth [7-10]. from Sigma Chemical Co., St. Louis, MO, USA Mouse
pRb inhibits cell cycle progression during a specific time monoclonal antibody isotype IgG 1 clone # Rb I to the re-
window in the GI phase and also is able to regulate entry into combinant pRb gene product was a product of Zymed
S-phase [7-10]. Progression through GI stage involves the Laboratories (South San Francisco, CA, USA). The antibody
inactivation ofpRb and this is achieved by phosphorylation recognizes both the phosphorylated and nonphosphorylated
of the pRb protein with specific GI cell cycle kinases. forms of the pRb protein. A horseradish peroxidase (HRP)-
Phosphorylation of pRb leads to the consequent release of a labeled secondary antimurine-antibody (Amersham) was
number of factors which this protein binds in its unphos- used. All other chemicals were of reagent grade. TAQ
phorylated stage. Once these transcription factors are free, polymerase was from Bcit-Ha'cmck Laboratories (Bcit-
they are able to activate the transcription of genes whose Ha'emek, Israel). Oligonucleotides used for PCR primers
products are required for S phase progression. Altematively, were synthesized in the Biotechnology Unit, KyriatWeizman
unphosphorylated pRb binds to specific transcription factors (Rehovot, Israel).
such as E2F family. Unphosphorylated pRb functions as a
negative growth regulator by preventing the transcription of
their responsive genes, for example, the binding of pRb to Cell culture
the transcription factor E2F prevents the transcription of
enzymes required for DNA replication and nucleotide LS 174 T colon cancer cell lines, were obtained from the
synthesis such as dihydrofolate reductase and thymidine American Type Culture Collection (Rockville, MD). The
kinase [7]. culture media and specified supplements were purchased
Phosphorylation of pRb is mediated by the action of the from Beit Haemek, Biological Industries, Israel: Dulbecco
GI phase Cyclin D kinases (CDK4 and CDK5), which are Modified Eagle's Medium (DMEM), Foetal Calf Serum
activated following binding of Cyclin D protein [13-17]. This (FCS), glutamine and antibiotics. LS174T cells were main-
tight regulation of phosphorylation during the cell cycle tained in continuous logarithmic growth in DMEM supp-
suggests that pRb itself may act as a cell cycle regulator and lernented with 10% FCS, 100 units/ml penicillin and 100 Ilg/
its inactivation may lead to unrestrained cell growth. In ml streptomycin sulfate at 37°C under a humidified atmo-
several mammalian primary cells and celliines the unphos- sphere at 7% CO, . The culture medium was changed every
phorylated species of 105-110 kD pRb are detected only in other day. Cultures were seeded at 0.2 x 10 6 ceIIs/ml in a
GO/GI phase, whereas in Sand G2 phases the 112-114 kD medium containing test compounds. Stock solutions were
pRb are the predominant species [17, 18]. Phosphorylation prepared in PBS.
of pRb is associated with the subsequent passage to the S Cell density was determined by electronic particle count-
phase of the cell cycle whereas cell proliferation may be ing. Growth curves were obtained by plotting the cell density
stopped in early GI by controlling cyclin D kinase activity. versus days in culture. Doubling times were determined
The precise mechanisms controlling cell arrest are largely graphically on the growth curves between days 0 and 4.
unknown, however recent findings suggest that specific CDK
inhibitors may playa central role [14, 19,20]. In mammalian
cells two families of CDK inhibitors have been character- Determination of enzymatic activities
ized;'the p21 and p 16 families. Recent data obtained in studies
in primary celliines [21], in established celliines [22] and a) Alkaline phosphatase
in the pl6 gene knockout mice [23], suggest that increased Alkaline phosphatase activity was measured in cytosolic
dosage of this CDK inhibitor is associated with cellular preparations by a standard spectrophotometric method using
senescence whereas decreased expression may be associated paranitrophenylphosphate as substrate [24].
with unrestrained cell growth.
In the present study we have determined whether the b) Thymidine kin ase activity
induced terminal differentiation in colon cancer celliines by The cytosolic enzyme was extracted from the colon cancer
sodium butyrate is associated with stable withdrawal from the cells as previously reported [25] and the enzymatic activity
cell cycle by means of regulating pRb phosphorylation and determined. Briefly, [3H]-thymidine served as the substrate
the expressionlactivity of other associated cell cycle proteins. to be phosphorylated by the extracted enzyme in the presence
23

of ATP. The phosphorylated tritriated products were isolated acetate and lead citrate and finally examined with Jeol 100
by means oftheir specific binding to a DEAE-81 Cellulose CX or 1200 EX transmission electron microscope (TEM).
filter. Radioactivity was determined on the cellulose discs and
the enzyme activity expressed as pmoles [3H]-TMP /mg
proteinlmin. RNA isolation and Ribonuclease Protection Assay

Total RNA was extracted from control and sodium butyrate-


Analysis 0/ cellular DNA content and cell-cycle by flow treated cells according to a protocol for single-step RNA
cytometry isolation, based on acid guanidium-thiocyanate-phenol-
chloroform extraction, usingTri-reagent solution (Molecular
Cells (5 x l0 5-2 x 106) were fixed in 70% ethanol. On the day Research Center, lnc., USA). Total RNA was quantified and
of the experiment, the cells were centrifuged, resuspended aliquoted identically in sterile and kept frozen at -70°C until
in phosphate buffered saline (PBS) containing calcium (O.lg/ assay.
L), and treated with 250 ml ofRNase solution (0.1 mg/mi in The ribonuclease protection assay (RPA) was performed
PBS containing calcium) for 1 h. They were then stained with with the Ambion RPA II™ kit, Ambion, lnc., USA. The in
250 ml of a freshly prepared propidium iodide solution (0.1 vitro transcription kit MAXIscript™ SP6/T7 and the pSP6
mg/mi with PBS containing calcium). Flow cytometric G3PDH- human antisense control template were also pur-
analysis was performed on a FACScan flow cytometer chased from Ambion. A 212-bp PCR-amplified fragment of
(Becton Dickinson, San Jose, CA, USA). Cell cycle distrib- the p 16 cDNA as previously described by LoisAF et al. [27]
ution was measured from 10,000 acquired single cells, using was inserted into pGEWM3z vector atAval sites (nucleotides
a doublet discriminator for analysis. The percentage of cells 251-465 from the 5' end). The resulting plasmid was linear-
in each phase was calculated using the Beckton Dickinson ized before producing anti sense riboprobes. p16 and GPDH
SFIT cell cycle algorithm. radiolabeled antisense riboprobes were prepared by in vitro
transcription with SP6 or T7 according to the in vitro kit
MAXIscript™ instructions. Total RNA (5-8 Ilg) isolated
SDS and Western blotting analysis from control or sodium butyrate-treated cells was resus-
pended in 20 111 hybridization buffer and 1 x 105 cpm [a32 _
SDS and Western blotting procedures were performed P]UTP-Iabeled antisense pl6 or GPDH riboprobe, denatured
essentially as previously described by DeCaprio et al. [26] at 95°C for 4 min and incubated overnight at 55°C to permit
to discriminate the phosphorylation ofthe pRb protein. Whole hybridization ofthe probes and the complementary mRNA
ceillysates were prepared by boiling loosed methanol-fixed in the sampie RNA.
cell pellets in a loading buffer containing 4M Urea, 6% SDS, Unhybridized RNA was digested in a mixture ofRNaseAl
4mMEDTA, 125 mMTris,pH 6.9, 3.5% (VN)ß-mercapto- RNase Tl, incubated for 30 min at 37°C. The reaction was
ethanol and 0.25% bromphenolblue. Electrophoresis was terminated following addition of300 111 RNase inactivationl
performed in 7% SDS-polyacrylamide gel, and then trans- precipitation mixture, then 100 111 ethanol was added to
ferred to PVDF membranes (Schleicher and Schuell) in improve precipitation efficiency, the sampies were pelleted
transfer buffer containing 25 mMTris-HCl, 192 mM glycine, and resuspended in 8 111 of loading buffer, heated for 3 min
2mM SDS and 20% v/v methanol. Membranes were rinsed at 90°C and loaded on a 6% polyacrilamide gel. The gel was
briefly in TBS (25 mM Tris-HCl, pH 8.0, 150 mM NaCl) run at approx. 250 volts. The gels were dried, covered with
containing Tween-20 (0.05%): TTBS, blocked in TTBS plastic wrap and exposed at -70°C on X-ray film (Kodak
containing 5% BSA. Western pRb biotted proteins were XRP).
detected by the enhanced chemiluminiscence detection
system (ECL).
Results
Light microscopy and transmission electron microscopy
Effect 0/ sodium butyrate on cell growth
Cells grown in flasks were fixed in 1% glutaraldehyde in 0.1
M phosphate buffer (PBS pH 7.4) for 1 hat room temperature. LS 174T colorectal cancer cells were grown in the absence
Fixed cells were then removed with a rubber policeman, or the presence of sodium butyrate (2mM). LS 174T cells
washed in PBS and post fixed in 1% OsO4 for 1 h. Sampies grew exponentially from the time ofseeding. LSl74T cells
were dehydrated in graded ethanol solutions and embedded reached a density of 2.45 x 106 after 4 days in culture. Lags
in Epon. Ultra thin sections in grids were stained with uranyl in LS 174T cell proliferation in the presence of 2mM sodium
24

0.6 - . - - - - - - - - - - - - - - - - - - - ,

- Contro!
- + - Sodium butyrate 0.5 -G-- Control
2.5
~ Sodium butyrate

0.4
2
"'0 0.3
><:
1.5
<l
E::I
Z

"
U

0.5
2 4

Days in Culture
0
0 2 3 4 5 Fig. 2. Alkaline phosphatase activity in eytosols prepared from LSI74T
cells eultured for various time periods in control media and in media
Days in Culture containing sodium butyrate (2mM). Treatment with sodium butyrate for
2-4 days caused a statistically significant enbancement in the phosphatase
Fig. 1. Growth curve of control and sodium butyrate-treated (2mM) LS !74T activity when compared to control cultures. The results represent the mean
cells. The results represent the mean ± S.E.M. often independent culturcs. ± S.E.M. offour independent cuItures.

butyrate were observed already at 1 day and onwards. changes were evident which became most pronounced day
However, a significant growth reduction was evident 3 and after day. After only one day of culture with sodium butyrate
4 days after sodium butyrate treatment. A representative cyst formations are already evident by light microscopy (Fig.
growth curve for the effect of sodium butyrate on LS174T 3-2). The cysts were randomly distributed all over the culture.
ccll prolifcration is presented in Fig. I. Electron microscopy of thin sections from LS l74T treated
cells revealed the presence ofICL and ITCL. The ICL were
observed in many cells (Fig. 3-4, 3-5). They appear as
Effect of sodium butyrate on alkaline phosphatase activity vascular cavities inside the cytoplasm and lined with brush
border microvilli. The ITCL appeared as large lumina
Alkaline phosphatase activity determined in the cytosolic embedded in the multilayer and surrounded by several cells.
fraction of control LS 174T cells showed a feeble activity (Fig. The ITeL were either lined with brush border microvilli or
2). However, 2--4 days oftreatment with the short chain fatty with smooth contour lined by few microvilli (Fig. 3-6). The
acid aforementioned caused a significant rise in the enzyme cells surrounding the ITCL were polarized and linked
activity. together with tight junctions (Fig. 3-5).

Light microscopy and ultrastructural studies Effect of sodium butyrate on the cell cycle of colon cancer
cells
LS 174T cells grown in DMEM media revealed essentially
undifferentiated cells ordered in multilayers (Fig. 3-1). Cells Figures 4 a-d show representative cell-cycle distribution of
facing the medium are covered with few and irregularly LS 174 T cells cultured under control conditions or after
distributed microvilli (Fig. 3-3). Iunctional complexes such sodium-butyrate treatment for several time periods. Figure
as tight junctions, intermediate junctions and desmosomes are 4a shows a representative cell cycle distribution diagram of
absent between neighbor apical cells. No cyst formation, LS l74T cells grown in control media for four days. A high
either inside the cells (intracellular lurnina-ICL) or between percent of cells populating the S phase of the cell cycle is
the cells (intercellular lurnina-ITCL) were found through all observed. Figure 4b shows a representative cell cycle
the time period (4 days) that LS174T colon cancer cells were distribution diagram of LS 174T ceHs treated one day with
cultured in DMEM control media (Fig. 3-3). sodium butyrate. A reduction in S-phase cyc1in ceHs. Fig.
Following addition ofbutyrate to the media, morphological 4c show representative cell cycle distribution ofLS 174T cells
25

Fig. 3. (I) Phase contrast micrograph of control LSl74T cells (x 400). (2) Phase contrast micrograph ofLS 174T cells treated with sodium butyrate for 4
days. Note the intracellular lumina (small arrow) and the intercellular lumina (long arrow) (x 400). (3) Electron microscope (EM) micrograph LS 174T cells
of contro! undifferentiated LS 174T cells. Note the irregularly distributed microvilli (x 5,200). (4) EM micrograph of LS 174T cells treated with sodium
butyrate (2mM) for 4 days. Note the intracellular lumina (IeL) inside the cytoplasm (x 15,000). (5) EM micrograph ofLSl74T cells treated with sodium
butyrate for 4 days. Note the intercellular lumina (ITCL) bonded by brush border microvilli and surrounded by many cells. The arrows indicate junctions
between cells (x 7,500). (6) EM ofLSl74T cells treated with sodium butyrate (2mM) for 4 days. Note the polarization and differentiation of cells exhibiting
brush border microvilli at their apical surface. (x 20,000).

treated for 3 days in media with sodium butyrate. Char- 1 shows that LS 174T colon cancer cells grown for 1 day in
acteristics of growth retardation are evident. In Fig. 4d a control media actively proliferate and 67.4 ± 7% of the cells
typical diagram indicating growth arrest in GI is evident. The populated the S phase whereas only 23.1 ± 3% ofthe cells
values presented in Table 1 of different cell cycle fraction GI, populated the G 1 phase of the cell cycle. 60.3 ± 6.5% of
Sand G2M are an average of3 different experiments. Table LS174T cells cultured 4 days in control media populated the
26

a b

c d

'0»

Fig. 4. Cell cycle analysis (DNA histograms: abscissa-relative DNA content, ordinate-relative number of cells.). Cell cycle distribution was determined on
control or sodium butyrate-treated LSl74T cells fixed in ethanol, treated with 100 /lg/ml RNase, and stained with 50 /lg/ml propidium iodide. Representative
cell cyc1e distribution of Control LSl74T cells 4 days in culture (60.4% of cells in S phase, Fig. 4a); sodium butyrate-treated cells, I day (47.8% of cells in
S phase, Fig. 4b); sodium butyrate-treated cells, 3 days (9.5% of cells in S phase, Fig. 4c); sodium butyrate-treated cells, 4 days (9.5% of cells in S phase, Fig.
4d); are shown. Surface colonocytes were mostly in G l. In LSI74T cells treated with sodium butyrate for 4 days a significant fraction ofthe cells were in
the G I phase of the cell cycle. In contrast, LS 174T cells culturcd under control conditions most of the cells were in the Sand G2/M phases of the cell cycle
(Fig.4a).

S phase of the cell cycle. A consistent and significant ulation ofthe cell cycle was observed due to sodium butyrate
reduction in the percentage of LS 174T cells populating the treatment. 35.8 ± 4% of LS 174T cells were in GI after the
S phase was consistently observed from day 2--4 after first day of sodium butyrate treatment, 52.1 ± 0.5% of the
initiating the treatment with sodium butyrate. Concomitantly, cells were at GI after 3 days oftreatment and 89.9 ± 0.9% of
a time-dependent progressive enhancement of the GI pop- the cells arrested at GI after 4 days of treatment.
27

1 2 3 4 5 6

_1 : 114
110
Fig. 5. Western analysis ofretinoblastoma protein (pRb) in LSI74T cells cultured under control conditions (lanes 1 and 2; land 4 days respectively) or
treated with sodium butyrate (lanes 3, 4, 5 and 6; 1 day, 2 days, 3 days and 4 days respectively). The hypophosphorylated pRb (range of -110 kDa), deteeted
as an intense band following 3 and 4 days of sodium butyrate treatment (Ianes 5 and 6), appears as a faster migrating dis(inc( molecular species, as compared
to the slower migration ofthe phosphorylated (range of -114 kDa) retinoblastoma protein deteeted in LSl74T cells eultured under eontrol conditions (1 and
4 days, lanes 1 and 2) or treated with sodium butyrate (1, 2 days; lanes 3 and 4).

Effect of the difjerentiation-inducing agent on alkaline 1 Z 34567


phosphatase activity

Feeble cytosolic alkaline phosphatase activity was measured


inLS174Tcontrol cells (Fig. 4). However, 3 days oftreatment
with sodium butyrate and onwards caused a significant rise
in the enzyme activity. p 16

EjJect of the difjerentiation-inducing agents on thymidine


kinase activity

The enzyme activity was assessed in LS174T cells cultured


in control media or exposed to the differentiation inducing
agents sodium butyrate. Thymidine kinase activity was
determined in cytosolic fractions and results summarized in
Table 2. The highest activity was measured in LS174T control GAPDH
cellular preparations (7.2 pmol 3HdTMP/mg proteinimin).
Thymidine kinase activity remained constant during a11 the
period (4 days) that LS174T cells were cultured in untreated
control conditions (range: 6.8-7.2 pmol 3HdTMP Ima protein/
min). A time-dependent progressive decrease in thymidine
kinase activity was observed in cancer cells treated with
sodium butyrate. The maximal and the most statistically
Fig. 6. RNase protcetion assay of RNA harvested from LS174T eells
significant decline was consistently noted on day 4 of
cultured under control conditions (I, 2 and 4 days) or treated with sodium
treatment, reaching to the lowest level ofO.Ol pmol JHdTMP butyrate (I, 2, 3 and 4 days). RNA was hybridized with pl6 and GPDH
Img protein/min. specific riboprobes.
Lane l-LSI74T eells cultured under eontrol eonditions for I day.
Lane 2-LSl74Teells cultured under control eonditions for 2 days.
Lane 3-LS174T eells cultured under control conditions for 4 days.
Effect of the difjerentiation-inducing agents on pRB
Lane 4-LS174T eells treated with sodium butyrate for I day.
phosphorylation Lane 5-LS174T cells treated with sodium butyrate for 2 days.
Lane 6-LS 174T cells treated with sodium butyrate for 3 days.
We found that retinoblastoma protein is constitutively Lane 7-LS174T eells treated with sodium butyrate for 4 days.
expressed in LS174T cells actively growing in DMEM media
whether the cells were grown for one day or four days in daily conditions (Fig. 5). When sodium butyrate (2mM) was added
added fresh media. Under those conditions, the protein to the culture media lower molecular weight (110 kDa) or
expressed was mostly in the phosphorylated form (114 kDa), unphosphorylated pRb species were clearly detected follow-
either after one or after four days in culture under control ing 3 days of treatment. After the fourth day of treatment,
28

when a significant growth inhibition was observed the Table 1. Thymidine kinase activity in contra! and sodium butyrate-treated
unphosphorylated species were predominant and their LSI74T cells
expression were significantly up-regulated as a result of the LS174T cells, Treatment, Days Thymidine kinase activity
differentiation agent treatment (Fig. 5). pmol 'HdTMP/mg pratein/min

Control, 1 day 7.2 ± 0.6


Contral, 4 days 6.8 ± 0.5
Detection ofp16 transeripts by Ribonuclease Protection Sodium butyrate, 1 day 5.8 ± 0.4
Assay Sodium butyrate,2 days 4.2 ± 0.4
Sodium butyrate,3 days 1.1 ± 0.1
Most commercial antibodies to pl6 either recognize multiple Sodium butyrate,4 days 0.01 ± 0.005
members of the family of cyclin kinase activity and in our
Thymidine kinase activity in cytosolic preparations ofLSI74T cells cu!tured
system, in repeated experiments this specific protein ex- under contro! and sodium butyrate (2mM) treatment conditions. The resu!ts
pression in LS 174T cancer celliine was shown to be bellow rcpresent the average ± S.E.M. offive independent determinations.
the limit of detection. Also Northem blot determinations of
p16 mRNA transcripts were undetectable, therefore we used intestine, specifically in the 10th week old embryo [30-32].
ribonuclease protection assay (RPA) methodology which has One ofthe roles attributed to ITCL (secondary lumen) during
a higher sensitivity to detect specific mRNA transcripts since morphogenesis ofthe intestine is in the remodulation of the
hybridization is performed in solution and with a riboprobe mucosa from the stratified to simple structures [30-32].
and not a cDNA probe. Therefore, ICL' sand ITCL' s formation may be envisaged as
RPA with a pl6 specific riboprobe showed undetectable specific morphological features observed during the normal
mRNA transcript levels in proliferating cells to several-fold process of differentiation. This observation supports the
increase in sodium-butyrate induced differentiated colono- notion that the in vitro model of differentiation described in
cytes (Fig. 6). Following one day oftreatment with the short our study is not only relevant to cancer-associated mechan-
chain fatty acid a significant increase in p 16 expression was isms but also to normal intestinal morphogenesis-related
observed, as compared to control conditions. After 3 days of systems.
treatment the maximal expression was achieved and the Following four days treatment with butyrate of LSl74T
intensity was similar to that observed after four days of colon cancer cells a GO/G I arrest in cell cycle progression
treatment. was measured (Fig. 3 and Table 2). The ability of the colon
cancer eel1s to undergo colonocyte differentiation, as deter-
mined by enzymatic and morphological criteria, in response
Discussion to the above mentioned differentiation inducer paralleled their
ability to growth arrest (Tables land 2).
The results reported in this study indicate that treatment of In the present study sodium butyrate treatment alongside
the unpolarized colon cancer cellline LSI74T with sodium with the morphologie al and biochemie al changes, induced
butyrate promoted the expression of a more differentiated also pRb dephosphorylation. Initial dephosphorylation was
phenotype according to similar findings reported previously observed 3 days after treatment with the short chain fatty acid
by us and others [4, 5, 24]. The treatment was consistently whereas complete shift to the dephospharylated protein
associated with enhanced alkaline phosphatase activity, a species was observed 4 days after the treatment was initiated
marker of colonic cell differentiation and the marked reduc-
tion in thymidine kinase activity, a marker of cell prolifer-
Table 2. Cell cyc!e distribution of contra! and sodium butyrate-treated
ation, measured in cytosolic preparations from colon cancer LS174Tcells
celliines in control conditions and treated far various time
periods with sodium butyrate. LSI74Tcells GI S G2M
Treatment, days
From the morphologic point of view, the treated colon
cancer cells showed clear formation oflumens expressed as Contral, I day 23.1 ± 3 67.4 ± 7 9.5 ± I
intercellular lumens (ICL' s) and intracellular lumens (ITCL' s). Contral, 4 days 36.1 ± 4 60.3 ± 6.5 3.6 ± 0.5
Sodium butyrate, I day 35.8 ± 4 56.4 ± 6 7.8 ± 0.9
Cells surrounding the lumen of ITCL were polarized and
Sodium butyrate, 2 days 41.2 ± 4 49.3 ± 6 9.5 ± I
linked by a weil developed junctional complex. Similar Sodium butyrate, 3 days 52.1 ± 0.5 45.7 ± 0.5 2.2 ± 0.3
structures were previously reported for differentiated polar- Sodium butyrate, 4 days 89.9 ± 0.9 4.3 ± 0.4 5.8 ± 0.7
ized cells grown in media enriched with galactose [28], in
Contral and sodium butyrate-treated LSI74T cells were prepared for cell
media devoid of glucose or glutamine [29], or when cells cycle distribution as described in Materials and methods seetion. The table
were transferred from DMEM to RPMI [24]. These features shows average ± S.E.M. cell cycle distribution of five independent
were also observed during morphogenesis of the human experiments.
29

and corresponded to cell cycle arrest measured by flow down-regulation of this enzyme is associated to the pRb
cytometry. The biochemical and morphological expression phosphorylation status induced by sodium butyrate, per-
of the differentiated phenotype was also associated with the mitting binding ofE2F species and inhibiting their mediator
appearance ofthe hypophosphorylated pRb species. Alkaline activity in the transcription of the thymidine kinase gene.
phosphatase activity and morphological changes were In the present study, we also addressed the possibility
observcd when pRb dephosphoprylated specics were evident. whether sodium butyrate is able to interfere with the cyclin
The frequent inactivation of retinoblastoma gene in various D-dependent kinase inhibitor p 16, whose overexpression in
tumors indicates that this gene and its protein product pRb mammalian cells has been associated with cell cycle arrest
are important components ofthe growth control machinery [19]. pl6 has been shown in other systems to specifically
in normal cells. The retinoblastoma gene is conserved among inhibit the cyclin D-dependent kineses edk4 and edk6, thus
vertebrates. Human and mouse retinoblastoma shares a 91 % preventing retinoblastoma protein phosphorylation, a post-
homology at the aminoacid level. This homology extends into translational modification that allows entry of the cells into
the promoter region which contains potential binding sites S phase [20]. In the present study expression of p 16 transcripts
for the transcription factors ATF, Sp 1, and E2F [33]. It has was minimal in untreated LS 174T colon cancer eells whereas
been reported that pRb is phosphorylated in an undulating in cells following I day of sodium butyrate treatment a strong
fashion during the eell eycle in many mammalian cells. It message was detected by RNase proteetion assay, the express-
becomes hyperphosphorylated in late GI phase and remains ion of p 16 transcripts increased from undetectable levels in
hyperphosphorylated in S, G2 and M. Cells in GO/G 1 express actively cycling proliferating LSI74T cells (control media)
pRb in an un- or hypophosphorylated form. pRb protein has to several fold increase during differentiation-induced
been implicated in regulating the transition from GO/G 1 to treatment with sodium butyrate. The expression ofthe CDK
S phase in a number of cellular systems [7-10,33]. Moreover, inhibitor p 16 preceded G I arrest during terminal differ-
the phosphorylation status of pRb has been correlated with entiation. This finding suggest that one ofthe earliest events
this activity such as the dephosphorylated form of pRb in the differentiation process of LS 174T cells indueed by
predominates in Sand G2/M. This has led to the hypothesis sodium butyrate activates p 16 transcription. This enhaneed
that phosphorylation of pRb in late GO/G 1 results in release CDK-inhibitor expression was later on associated with
of a block and allows the cells to progress through cell cycle. decreased pRb phosphorylation aetivity.
The tight regulation of phosphorylation during the cell cycle One of the effects of butyrate treatment is thought to be
suggests that the pRb protein itself may act as a cell regulator, hyperacetylation of histones and non-histones nuclear
the inactivation ofwhich may lead to unbridled cell growth. proteins [35]. Histone deacetylase is inhibited, both in vitra
The results from this study shed light on a possible and in viva, with no change in histone acetyltransferase
mechanistic approach to the role ofbutyrate on cell growth activity [1,35]. Hyperacetylation of core histones, HMG-14
arrest. It appears that butyrate mediates pRb dephosphory- and HM G-17 ean account for ehanges in chromatin strueture
lation, and thus inhibits further cell cycle progression. As whieh, in turn, ean affect gene transcription and block the
alluded to in the previous paragraph, butyrate-mediated pRb eells in GO/G 1 phase.
dephosphorylation may allow the binding of specific trans- Aberrant hypermethylation of 5' CpG islands within
cription factors resulting in the inhibition of aetivation of proximal promoter regions has been implicated as mechan-
responsive genes. isms of gene inactivation [36, 37]. Examples for this mechan-
Terminal differentiation is coupled to withdrawal from the ism of gene inactivation has been described for the p 16 gene
cell eycle. Withdrawal from the cell cycle is an essential [38]. It can be surmised that histone deacetylation activity
aspect of mammalian differentiation and requires the pRb induced by sodium butyrate treatment may let the free
protein in an hypophosphorylated state able to bind trans- approach of intracellular DNA-demethylating enzymes to
cription factors needed to activate genes required for cell hypermethylated 5' CpG islands within proximal promoter
cycle reentry. Dephosphorylated pRb is able to interact with regions ofthe pI6 tumor suppressor gene, activity which is
several transcription factors such as E2F, MyoD, Elf-I, and associated with restoration ofpl6 RNA levels. The validity
others, whereas the phosphorylated species do not bind this ofthis putative proposed mechanism of action remains to be
gene activating factors. E2F, originally discovered through determined.
its role in the activation of the adenovirus E2 promoter has Additional studies suggest that sodium butyrate can
been shown to transcriptionally activate important genes for modulate the expression of a variety of genes either at the
cell division such as DNA polymerase a, dihydrofolate transcriptionallevel, through spccific promoter regions, or
reductase, thymidine kinase and others [7-10, 34]. In the via posttranscriptional mechanisms. It can be surmised that
present study we indeed observed that sodium butyrate sodium butyrate directly controls pI6 gene expression either
significantly down-regulates the activity ofthe proliferation- directly following binding to butyrate DNA responsive
associated enzyme thymidine kinase. It can be surmised that elements, by activating other transcription faetors, by
30

stabilizing mRNA transcripts, or alternatively by other Biol136: 155-165, 1997


mechanism. The exact mechanism of action remains to be 19. Sherr CJ, Roberts 1M: Inhibitors ofmammalian GI cyclin-dependent
kinases. Genes Dev 9: 1149--1163, 1995
investigated.
20. Harper JW, Elledge SJ: Cdk inhibitors in development and cancer.
Curr Opin Genet Dev 6: 56-64, 1996
21. Yeager T, Stadler W, Belair C, Puthenveettil J, Olopade 0, Reznikoff
Conclusion C: Increased pl6 levels correlate with pRb alterations in human
urothelial cells. Cancer Res 55: 493-497, 1995
22. Ragione FD, Russo GL, Oliva A, Mercurio C, Mastropietro S, Pietra
The. results from this study provide evidence for a specific VD, Zappia V: Biochemical characterization ofpl6NK4- and p18-
cellular mechanism whereby butyrate induces to colon cancer containing complexes in human celllines. J Biol Chem 271: 15942-
cells to undergo cell cycle arrest and differentiation. 15949, 1996
23. Serrano M, Lee HW, Chin L, Cordon-Cardo C, Beach D, DePinho
RA: Role ofthe INK4a locus in tumor suppression and cell mortality.
Ce1l85: 27-37, 1996
References 24. Schwartz B, Polak-Charcon S, Lamprecht SA, Niv Y, Kim YS: The
induction of a differentiated phenotype in human colon cancer cells
1. Scheppach W, Bartram HP, Richter F: Role of short-chain fatty acids in requires the attenuation ofcytoskeletal tyrosine phosphorylation. Oncol
the prevention of colorectal cancer. Eur J Cancer 31A: 1077-1080, 1995 Res 7: 277-287, 1995
2. Hague A, Butt AJ, Paraskeva C: Tbe role ofbutyrate in human colonie 25. Schwartz B,Avivi C, Lamprecht SA: Isolation and characterization of
epithelial cells: an energy source or inducer of differentiation and normal and neoplastic colonic epithelial cell populations. Gastro-
apoptosis? Proc Nut Soc 55: 937-943, 1996 enterology 100: 692-702, 1991
3. Deng G, Liu G, Hu L, Gum JR, Kim YS: Transcriptional regulation of 26. DeCaprio JA, Ludlow JW, Lynch D, Furukawa Y, Griffin J, Plwnica-
the human placental-like alkaline phosphatase gene and mechanisms Worms H, Huang CM, Livingston DM: Tbe product of the retino-
involved in its induction by sodium butyrate. Cancer Res 52: 337&- blastoma susceptibility gene has properties of a cell cycle regulatory
3383, 1992 element. Ce1l58: 1085-1095, 1989
4. Kruh J, Tichonicky L, Defer N: Effect ofbutyrate on gene expression. 27. Lois AF, Cooper LT, Geng Y, Nobori T, Carson D: Expression of the
In: HJ Binder, JH Cummings, K Soergel (eds.) Short Chain Fatty Acids. p 16 and p 15 cyclin-dependent kinase inhibitors in Iymphocyte activation
Dordrecht, Kluwer, pp 135-147, 1994 and neuronal differentiation. Cancer Res 55: 4010-4013,1995
5. Gamet L, Daviaud D, Denis -Pouxviel C, Remesy C, MuratJ-C: Effect 28. Pinto M,Appay MD, Simon-Assman P, Cavalier G, Dracopoli N: Fogh
of short-chain fatty acids on growth and differentiation of the human J, et al.: Enterocytic differentiation in cultured human colon cancer
colon cancer cellline HT-29. Int J Cancer 52: 286-289, 1992 cells by replacement of glucose by galactose in the medium. Biol Cell
6. Gum JR, Kam JC, Hicks JW, Sleisenger MH, Kim YS: Effects of sodium 44: 193-196, 1982
butyrate on human colonic adenocarcinoma cells. Induction ofplacental- 29. ZweibaumA, Pinto M, Chevalier G, Dussaulx E, Triadou N, Lacroix
like alkaline phosphatase. J Biol Chem 262: 1092-1097, 1987 B, et al.: Enterocytic differentiation of a subpopulation of the human
7. Wiman KG: Tbe retinoblastoma gene: role in cell cycle control and colon cancer cellline HT-29 selected for growth in sugar-free medium
cell differentiation. Faseb J 7: 841--845, 1993 and its inhibition by glucose. J Cell Physiol122: 21-29, 1985
8. Riley DJ, Lee KY, Lee WH: Tbe retinoblastoma protein: more than a 30. Simons SK, Fuller SD: Cell surface polarity in epithelia. Ann Rev
tumor suppresor. Annu Rev Cell BiollO: 1-29,1995 Cell BiolI: 243-288, 1985
9. Weinberg RA: Tbe retinoblastoma protein and cell cycle controL Cell 31. Rodriguez-Boulon E, Nelson WJ: Morphogenesis of the polarized
81: 323-330, 1995 epithelial cell phenotype. Science 245: 718--725, 1989
10. Bartek J, Bartkova J, Lukas J: Tbe retinoblastoma protein pathway 32. Cereijido M, Contreras RG, Gonzalez-Mariscal L: Development and
and the restrietion point. Current Opin Cell Biol 8: 805--814, 1996 alteration ofpolarity. Ann Rev Physiol51: 585--595, 1989
11. Gope R, Gope ML: Abundance and state of phosphorylation of the 33. Wang JYJ, Knudsen ES, Welch PJ: The retinoblastoma tumor
retinoblastoma susceptibility gene product in human colon cancer. Mol suppressor protein. Adv Cancer Res 64: 25--85, 1994
Cell Bioehern 110: 123-133, 1992 34. Ahnasan A, Yin Y, Kelly RE, EP Lee, Bradley A, Li W, Bertino JR,
12. Ali AA, Harvey JP, Wildrick DM, Boman BM: Retinoblastoma gene Wahl GM: Deficiency ofretinoblastoma protein leads to inappropriate
productassociated proteins in human colon cancer celllines. Bioehern S-phase entry, activation ofE2F-responsive genes, and apoptosis. Proc
Biophys Res Commun 194: 848--854, 1993 NatlAcad Sei 92: 5436-5440,1995
13. Sherr CJ: Mammalian GI cyclins. Cell 73: 1059--1065, 1993 35. Csordas A: On the biological role of histone acetylation. Bioehern J
14. HunterT, Pines J: Cyclins and cancer 11: Cyclin D and CDK inhibitors 265: 23-38, 1990
come ofage. Ce1l79: 573-582,1994 36. Herman JG, LatifF, Weng Y, Lerman MI, Zbar B, Liu S, Samid D, et
15. Nigg EA: Cyclin-dependent protein kinases: Key regulators of the al.: Silencingofthe VHL tumor-suppressor gene by DNA methylation
eukaryotic cell cycle. Bio Essays 17: 471-480, 1995 in renal carcinoma. Proc Natl Acad Sci 91: 9700--9704, 1994
16. Resnitzky D, Reed SI: Different roles for cyclins D1 and E in regulation 37. Yoshiura K, Kanai Y, Ochiai A, Shimoyama Y, Sugimura T, Hirohashi
ofthe GI-to-S transition. Mol Cell Bio115: 3463-3469, 1995 S: Silencing ofthe E-cadherin invasion-suppressor gene by CpGmethyl-
17. Kranenburg 0, Van der EB AJ, ZantemaA: Cyclin D kinases and pRb: ation in human carcinomas. Proc Natl Acad Sci 92: 7416-7419, 1995
Regulators of the proliferation-differentiation switch. 367: 103-106, 38. Merlo A, Herman JG, Lee DJ, Gabrielson E, Burger PC, Baylin SB,
1995 Sidransky D: 5' -CpG island methylation is associated with trans-
18. Tanaka EM, GannAAF, Gates PB, Brockes JP: Newt myotobes reenter criptional silencing of the tumor suppressor pI6/CDKN2IMTSI in
the cell cycle by phosphorylation ofthe retinoblastoma protein. J Cell human cancers. Nat Med I: 686-692, 1995
PART II

CELL GROWTH AND DEVELOPMENT


Molecular and Cellular Biochernistry 188: 33-39, 1998.
© 1998 Kluwer Acadernic Publishers.

Regulation of adipocyte gene expression by


polyunsaturated fatty acids
Ann Vogel Hertzel and David A. Bemlohr
Department 0/ Biochemistry, University 0/Minnesota, SI. Paul, Minnesota, USA

Abstract
A wide number of adipocyte genes are regulated by exogenous polyunsaturated fatty acids (PUFA) through the actions ofthe
peroxisome proliferator activated receptor. Such genes include the adipocyte lipid-binding protein (ALBP or aP2) which plays
a central role in facilitating the traflicking of fatty acids within adipocytes. Work from a number oflaboratories has suggested
the key elements of the lipid signal transduction pathway include: (1) the transport of exogenous PUFAs across the plasma
membrane, (2) metabolism of polyunsaturated fatty acids to second messengers including 15-deoxy d 12 ,14 prostagiandin J2
(15dPGJ2), (3) traflicking of 15dPGJ2 and other second messengers from the smooth ER to the nucleus for association with
peroxisome proliferator activated receptory (PPARy), and (4) dimerization ofPPARywith retinoid X receptor (RXR) permitting
regulation of transcription via association with any of several nuclear co-activators or repressors. In addition to the aP2 gene
being a target of activation by fatty acids, at the protein level ALBP/aP2 plays a role in traflicking of fatty acids and/or their
metabolises. We report here that in a heterologous system using CV-I cells transiently transfected with PPARy2, co-expression
of ALBP/aP2 enhances the PPAR-dependent activation of gene transcription. These results suggest that ALBP/aP2 functions
as a positive factor in fatty acid signalling by directly targetting and delivering fatty acids metabolites to the lipid signal
transduction pathway. (Mol Cell Biochem 188: 33-39, 1998)

Key words: fatty acids, adipocytes, gene regulation

Introduction similarly unc1ear process, cross the epithelial cell boundary


to merge with the circulating pool of albumin. Therefore,
adipocytes carry out bidirectional fatty acid trafficking,
Polyunsaturated fatty acids serve many important functions. inward during nutrient abundance and outward during
Besides being constituents of all phospholipids, fatty acids nutrient depletion.
are stored in triacylglycerol droplets of adipose cells which It has been hypothesized that the metabolic rationale for
serve as a readily metabolizable source of reducedhydrocarbon cytosolic fatty acid-binding proteins (FABP) is inherently
[I]. During nutrient abundance, adipose-derived lipoprotein connected to the low solubility of fatty acids in aqueous
lipase hydrolyzes chylomicra triacylglycerol, liberating solutions. The solubilization of fatty acids within these
fatty acids which, in an ill-defined process, transit the proteins would thereby create a large intracellular pool of
capillary celllayer to the extracellular space where they can unesterified fatty acids for utilization [2]. FABP-bound fatty
be intemalized by the adipocyte. Such nutritionally derived acids could then be delivered to specific sites within
fatty acids are condensed with a-glycerol phosphate and adipocytes for metabolism and/or signaling. The adipose
stored as triacyglycerol in large droplets. During aperiod of member of the lipid-binding protein multigene family is
energy deficit, adipose triacylglycerol stores are released termed the adipocyte lipid-binding protein (ALBP or aP2).
through activation ofhormone sensitive lipase to liberate free As with most members ofthis family, ALBP/aP2 forms a I: I
fatty acids which are transported out of the adipocyte, bound complex with unesterified fatty acids, binding them in a large
by serum albumin within the extracellular space, and by a interior cavity of the protein [2].

Address Jor offPrints: D.A. Bcrn1ohr, Department ofBiochemistry, University of Minnesota, 1479 Gortner Avenne, St. Paut, MN 55108, USA
34

While our understanding of polyunsaturated fatty acids In this study we transiently transfected ALBP/aP2 into CV-
as structural molecules or as an energy source has been 1 cells in the presence ofa PPAR-responsive reporter system
appreciated for decades, polyenoic fatty acids have been to address whether the protein functions in the lipid signalling
shown more recently to be precursors of signaling molecules pathway. We present evidence suggesting thatALBP/aP2 is
which alter gene transcription in response to changes in a positive effector of gene transcription by PUFAs and that
metabolic status [3, 4]. In this regard, polyunsaturated fatty an autoregulatory loop exists to increase the expression of
acids act as prohormones, being metabolized to second lipid-binding proteins in response to dietary fatty acids.
messengers affecting critical regulatory functions. Such
bioactive lipid products include, but are not Iimited to, the
prostaglandins, leukotrienes and thromboxanes [5]. One Materials and methods
example includes an arachidonic acid metabolite, 15-deoxy-
ßI2,14_prostaglandinJ2 (15d-PGJ 2), which has been shown to Plasmids
bind to peroxisome proliferator activated receptory (PPAR)
and activate transcription [6-8]. In adipocytes, two of the Thc Gal DNA-binding domain (amino acids 1-147) was
three murine PPAR subtypes have been identified, PPARy amplified by PCR and subcloned into pcDNA3 (Invitrogen
and PPARo [9, 10]. These ligand activated receptors Corp., San Diego, CA, USA) downstream of the CMV
heterodimerize with the 9-cis retinoic acid-binding retinoid promoter. PPARy21igand-binding domain (amino acids 203-
X receptors (RXRa) [11] and preferentially bind to peroxisome 505) was translationally fused to the end ofthe Gal sequences,
proliferator response sequences, termed DR- 1elements, These creating a Gal DNA-binding domain-PPARyligand-binding
elements contain adegenerate six base-pair direct repeat domain chimeric construct. The firefly luciferase reporter
separated by a single nucleotide. Recent results suggest construct was generated by PCR amplification of four repeats
nuc1ear heterodimers associate with either co-activators ofthe Gal upstream activating sequence cloned 5' ofthe SV40
(p300) or co-repressors (SMRT/NCo-R) and other auxiliary promoter in the vector, pGL3-promoter (Promega Corp.,
factors to alter the level of histone acetylation by regulating Madison, WI, USA). Transfection efficiency was monitored
the activities of histone acetylase and/or histone deacetylase by co-transfecting a pRL-CMV plasmid (Promega Corp.,
[12-16]. Histone modification presumably affects nucleo- Madison, WI, USA), measuring dualluciferase activities
some positioning within chromatin and the eventual activationl and normalizing the firefly luciferase activity to the pRL
repression of genes. In adipocytes, several target genes of luciferase activity. The ALBP/aP2 promoter was PCR
PPARy action have been identified, two of the most thoroughly amplified (5.4 kb) and c10ned into pGL2-basic (Promega
characterized are those encoding phosphoenol pyruvate Corp., Madison, WI, USA) creating an ALBP-luciferase
carboxykinase and ALBP/aP2 [17-20]. Both genes contain reporter plasmid. Fulliength PPARy2 was c10ned into the
a functional DR-l element which directs expression in BamHI site of pcDNA3 resulting in a constitutively expressed
response to PUFA, most likely via the 15d-PGJ2 pathway. PPARy2 construct. Plasmid containing mPPARy2 was a
Recent work from this laboratory has shown that ALBPI generous gift from Dr, Bruce M. Spiegelman, Dana-Farber
aP2 binds with high affinity and specificity to 15d-PGJ2 [21]. Cancer Institute, Harvard Medical School, Boston, USA. A
ALBP/aP2 binds to 15d-PGJ, with an apparent dissociation vector containing hRXRa was a gift from Dr. Ronald M.
constant of 1.9~M, about 6-fold weaker than the binding of Evans, Salk Institute, La Jolla CA, USA. Pioglitazone was a
the lipid to PPARy. However, the concentration ofALBP/aP2 generous gift from Dr. Steven D. Clarke, University ofTexas,
is approximately 1 mM in the cell while the nuc1ear PPARy Austin TX, USA.
is at least two-orders of magnitude 10wer [21]. Consequently,
ALBP/aP2 represents a high capacity, low affinity binding
site for 15d-PGJ2 while PPARyis a high affinity, low capacity Cellsltransfection
binding protein. Due to its binding properties, ALBP/aP2
could playapositive role in 15d-PGJ 2-mediated gene CV -I cells were maintained in DMEM with 10% fetal bovine
expression by solubilizing the lipid making it more available serum at 37°C in 5% CO2. For transient transfections, the cells
to PPARy. Alternatively, ALBP/aP2 could playanegative were plated in phenol-red free DMEM with 10% charcoal-
role in 15d-PGJ,-mediated gene expression by sequestering treated fetal bovine serum in 12-well plates the day prior to
the lipid, making it less available to PPARy. An additional transfection, The cells were transfected in quadruplicate via
factor to consider is that ALBP/aP2 also binds PUFAs with a calcium phosphate DNA precipitate [22]. Typically, 0.5 ~g
high affinity and specificity. By similar arguments, ALBPI Gal DNA-binding domain-PPARy ligand-binding domain
aP2 could either enhance or reduce gene expression in construct, 0.5 ~g Gal upstream activating sequence-Iuciferase
response to PUFAs by making the fatty acid more or less construct and 10 ng pRL-CMV were used per transfection.
available to the 15d-PGJ2 biosynthetic pathway, respectively. Where indicated, 0.5 ~g ALBP/aP2 promoter-Iuciferase,
35

pCMV-PPARy2, RXRa expression vector, and 10 ng pRL- 8~-------------------------.


CMV were used. The total mass of DNA transfected in
• Linoleate
comparable sampies was normalized by the addition of 7 [] Arachidonate
pUC119. After 18 h, the cells were washed three times with
PBS and refed, followed by treatment with either ll!l ethanol 6
or fatty acid (I OOOX stocks delivered in ethanol). Fatty acids
were purchased from NuChek Prep, Elysian MN. 9-cis
5
retinoic acid was purchased from Biomol, Plymouth Meeting,
PA. Twenty-four hours following treatment, the cells were
4
washed, harvested and luciferase activity was assayed using
a luminometer and the Dual Luciferase Reporter Assay
System (Promega Corp., Madison, WI, USA).
3

2
Results
In order to study the fatty acid regulation of gene expression
by PPARy2, CV-l cells were transiently transfected with o
three plasmids: (I) A construct consisting of 4 copies ofthe ETOH 5 10 25 50 100
Gal upstream activating sequence inserted upstream of a SV40
promoter-frrefly luciferase gene (2) a plasmid containing a Fig. 1. Activation of Gal-PPARy by polynnsaturated fatty acids. CV-l
CMV promoter driving expression of Gal DNA-binding cells were plated and transfected with three plasmids: l-Gal-PPARr, 2-
uas-Iuciferase; 3-pRL-CMV (see Materials and methods). The cells were
domain translationally fused to PPARy2 ligand-binding washed, refed, and treated with the following conditions: ethanol (EtOH);
domain; (3) the pCMV-RL luciferase plasmid included as a linoleic acidorarachidonic acid 5,10,25, SO, or 100 11M for24 h. Following
control oftransfection efficiency. In this system, ligand binding treatment, the cells were washed, harvested,lysed and assayed for luciferase
by the PPARy2 ligand-binding domain allows DNA binding activity. Each transfection and treatment were perfonned in quadruplicate.
The results shown represent one quadruplicate experiment which was
by Gal4 and in turn activates luciferase expression. Such a
repeated at least three times. The firefly luciferase activity was normalized
heterologous system is useful in order to observe maximal to the transfection controlluciferase levels. The resuIts are expressed as
responsiveness to fatty acids since adipocytes preferentially fold induction relative to the EtOH contrals. Error bars represent the
metabolize the lipid to triacylglycerol. Figure 1 shows the standard deviation of the quadruplicate sampies.
concentration dependence ofPPARy2 activation by linoleate
and arachidonate in CV-l cells. The activation of PPARy2
increased as the concentration ofligand added increased, with A similar three plasmid system was devised to examine the
a 1.5-fold increase in luciferase reporterlevels when 10 !-IM fatty acid regulation ofthe aP2 gene: (1) a plasmid containing
linoleic acid was added and a maximum of 6-fold activation 5.4 kb of ALBP/aP2 upstream sequence (including the DR-
by 100 I!M linoleic acid (with respect to the ethanol treated 1 element [24]) fused to firefly luciferase sequence; (2) a
sampies). Arachidonate stimulated PPARy2-dependent second containing the CMV promoter controlling expression
luciferase expression to similar levels. offulliength PPARy2; (3) the pCMV-RL luciferase plasmid
Members ofthe thiazolidinedione family ofhypolipidemic included as a control of transfection efficiency. Similar to
drugs have been demonstrated to be potent activators of previous reports, transient transfections showed that the aP2
PPARr and have been shown to directly bind this receptor promoter is regulated by fatty acids through PPARy2 (Fig.
with nanomolar affinity [23]. In similar experiments to those 3). Although CV-l cells contain RXRa at concentrations
previously described, transient transfections of CV- I cells sufficient for activation of DR-I containing genes, an
treated with 10 I!M pioglitazone reached maximum activation enhancement of expression from the aP2 promoter was seen
of approximately 30-40-fold that of the control (ethanol with the addition of a pCMV-RXRa plasmid and 9-cis
treated) sampies (Fig. 2). The maximal activation was retinoic acid. These results suggest that RXRa is limiting in
dependent on time, requiring 24 h of treatment. These results the CV-l system. Moreover, the synergy of activation by fatty
are consistent with pioglitazone activated PPARy2-dependent acids and retinoids in the presence of both PPARy2 and
regulation of gene expression similar to that seen with the RXRa is consistent with a heterodimer ofPPARy2IRXRa as
ALBP/aP2 gene. The greater level of gene activation observed the activating protein complex.
with pioglita~one than PUFAs is likely due to the lack of To address the effects of ALBP/aP2 on the PPARrl
intracellular metabolism ofthe thiazolidinedione relative to RXRa-dependent activation ofthe 5.4 kb ALBP-luciferase,
the fatty acid which is converted to triacylglycerols. the three plasmids (ALBP promoter-Iuciferase, pCMV-
36

40 PPARy2, pCMV-RL) were cotransfected along with varying


concentrations of a pCMV-ALBP construct. Under these
35
conditions, the ALBP promoter is available for PPARy-
·foB 30 dependent activation as previously described. However, the
efficiency ofthis activation as influenced by ALBP/aP2 could
'"
11)
25
be determined. ALBP/aP2 association with fatty acids and/
'"
S 20 or 15d-PGJ2' could yield three likely scenarios: the first is
...!l that ALBP/aP2 would have no effect on PPAR-dependent

..a 15 activation, the second is that ALBP/aP2 would simply
:sl sequester the fatty acids and prevent them from being
~ metabolized and targeted to the PPARs resulting in a decrease
5 in PPARdependent gene activation, and the third is that
ALBP/aP2 would bind and deliver the ligand either to a
0
metabolie pathway or to the PPARs directly. To evaluate these
possibilities, increasing concentrations of pCMV-ALBP
PPARr + + plasmid were transfected with the other three plasrnids and
Piog. + + the luciferase reporter levels derived from the ALBP/aP2
promoter-Iuciferase construct were measured. By transfecting
Fig. 2. Activation of Gal-PPARy by pioglitazone. CV-I cells were plated 0.05-1.0 ~g of the ALBP/aP2 expressing plasmid, the
and transfected as in Fig. I. Minus (-) or plus (+) indicates absence or resulting luciferase levels increased from 1.7-2.5-fold (Fig.
presence ofPPARy, respectively. Tbe cells were washed, refed and treated
4). Since the total mass ofDNA transfected was constant due
with 10 ~ pioglitazone (+) or DMSO (-) for 24 h. Following treatment,
the cells were washed, harvested, lysed and assayed for luciferase activity. to the addition ofvarious amounts ofpUC119 DNA, this
Each transfection and treatment were perforrned in quadruplicate. Tbe firefly effect is specific for the presence of ALBP/aP2 and not due
luciferase activity was norrnalized to the transfection controlluciferase levels.
The results are expressed as fold induction relative to the DMSO controls.
Error bars represent the staodard deviation ofthe quadruplicate sampies. 3.5

C
25 'S;
'.z:j
u
2.5
«I
11)

e 2
CI}
20
.~
'l:I ~
al 'ü 1.5
15 ..a"0
~
o!l
'0
&
.e 10
0.5
~
&!
pCMV-ALBP 0 0.05 0.25 0.5
0 pUC1l9 0.95 0.75 0.5 o
PPARy + + ~g DNA transfected
RXRa + +
Linoleate + +
Fig.4. Effect of ALBP/aP2 on the regulation ofthe ALBP/aP2 promoter
9-cis RA + + by ppARy. CV-I cells were plated and transfected with the following
plasmids: 1- ALBP-Iuciferase; 2- pCMV-PPARy, 3- pRL-CMV and 4-
Fig. 3. Fatty acid regulation of the ALBP/aP2 promoter by PPARyand various amounts of a pCMV-ALBP. pUC1l9 DNA was added to norrnalize
RXRa. CV-I cells were plated and transfected with the following plasmids: the total mass of DNA transfected in each case. After 24 h, the cells were
1- ALBP-Iuciferase; 2- pCMV-PPARy, 3- pCMV-RXRa; 4pRL-CMV. washed, harvested, Iysed and assayed for luciferase activity. Each
Minus (-) or plus (+) indicates absence or presence of each nuclear receptor transfection and treatment were perforrned in quadruplicate. The firefly
(PPARy or RXRa). Tbe cells were washed, refed and treated with 100 ~M luciferase activity was norrnalized to the transfection control (renilla
linoleate and/or I I'M 9-cis retinoic acid (RA) as indicated. Following luciferase) levels. The results are expressed as fold induction relative to
treatment, the cells were washed, harvested, Iysed and assayed for luciferase the transfeetion without pCMV-ALBP. Error bars represent the standard
activity. Each transfeetion and treatment were perforrned in quadruplicate deviation of the quadruplicate sampies. (Student (-test values: p = 0.003,
and analyzed as in Figs I and 2. 0.059, 0.014, and 0.012 for transfections 0.05--1 ~g, respectively).
37

to a variation in the mass of DNA transfected. Since no abundance of endogenous fatty acids fluxing in and out of
exogenous ligands are added under these conditions, this result an adipocyte causing near maximal expression of fatty acid
demonstrates that the presence of ALBP/aP2 significantly responsive genes, such as ALBP/aP2. The addition of
enhances the activation of gene expression by PPARy under exogenous fatty acids may not significantly affect the total
conditions oflimited ligands, thereby sensitizing the system concentration of fatty acids available for signalling in the
to small changes in available activators. adipocyte. Altematively, since adipocytes function as lipid
storage centers, the exogenously added fatty acids may be
rapidly converted into triacylglycerol, effectively removing
Discussion them from potential PPAR activation. Because ofthe small
responses to fatty acids in adipocytes, most ofthe current data
The key elements in the lipid signal transduction pathway has been collected in either a heterologous system (i.e.; CV-
include: the transport of exogenous fatty acids into adipo- I cells) or in preadipocytes where much larger changes (3-
cytes, metabolism of polyunsaturated fatty acids to second 50 fold increases) can be detected.
messengers, trafficking ofthe messengers to the nucleus and In the experiments presented here, two different plasmid
activation of PPARy/RXRa resulting in the regulation of systems were used to demonstrate gene activation by fatty
transcription. Ofthe adipocyte genes examined to date which acids in cultured cells. Both of these were dependent on a
are controlled by this pathway, by far the most thoroughly ligand activated PPARr receptor. The first system utilized
characterized gene is that for ALBP/aP2. Regulation of the Gal DNA-binding domain and the corresponding
ALBP/aP2 gene expression by fatty acids has been studied upstream activating DNA sequences to which it binds.
in several systems and at several levels. In committed Administration of the potent PPARy agonist pioglitazone,
preadipocytes, but not differentiated adipocytes, linolenate resulted in the dramatic increases in gene expression,
treatment leads to a 12-fold increase in ALBP/aP2 message exceeding that observed with fatty acids alone. Although
[19]. This activation was prevented by actinomycin D not entirely clear, the hyperactivation observed with pioglita-
suggesting a transcriptional mechanism of regulation, with zone is thought to be due to its lack ofmetabolism in cells,
protein levels paralleling message levels. This phenomenon thereby allowing for an accumulation ofhigh intracellular
is not simply a result of differentiation since a similar trend levels. Fatty acid activation of gene expression in this
was not seen with other adipose related genes. The ability of system was also demonstrated, but to a quantitatively lower
fatty acids to regulate ALBP/aP2 is strongly dependent on extent.
the fatty acid length, with less influence from the degree of The second transcription system monitored luciferase
saturation [19]. Similar experiments in preadipocytes activity driven by theALBP/aP2 promoter. These experiments
determined that the non-metabolizable bromopalrnitate is a did not require chimeric proteins as in the previous system, but
more potent inducer of ALBP/aP2 than the metabolizable rather simulated more physiological conditions utilizing full
palmitate [20]. Consistent with PPAR regulation ofALBPI length native protein receptors. Results ofthese experiments
aP2, the thiazoladinedione class of hypolipidemic drugs demonstrated a fatty acid regulatory etTect in the presence of
known to be synthetic ligands ofPPARy induce expression fulliength PPARy2. Mechanistically PPARs have been shown
of ALBP/aP2. In 3T3-LI adipocytes, one ofthese compounds, to heterodimerize with RXRa creating a protein complex
pioglitazone, produced a 2-fold increase in ALBP/aP2 competent for DNA-binding to PPREs. Although CV-I cells
message [25]. However, treatment of preadipocytes with have been considered to contain endogenous levels ofRXRa
pioglitazone resulted in the formation ofthe PPARy/RXRa sufficient for maximal gene activation, our experiments
heterodimer and a lO-fold increase in ALBP/aP2 expression indicated that the addition of an RXRa expression plasrnid
[26]. In a heterologous system using NIH3T3 fibroblasts in the presence of9-cis retinoic acid increased the expression
which normally do not express ALBP/aP2, transient levels by another 3-fold. This result is consistent with fatty
transfection with PPARy and treatment with linoleic acid acids and retinoids activating gene expression synergistically
resulted in the expression of ALBP/aP2 mRNA [27]. through the nuclear receptors PPARyand RXRa.
Therefore, although activation of ALBP/aP2 in adipocytes Having established PPAR-dependent activation of the
is modest (even with the very potent PPARy activator, ALBP/aP2 promoter by fatty acids, we were then able to test
pioglitazone), in preadipocytes or fibroblast systems, the efficiency of this activation in the presence of the
ALBP/aP2 expression is extremely sensitive to levels of adipocyte lipid-binding protein. In theory, three results were
fatty acids. possible. The first is that the presence of ALBP/aP2 would
Regulation of genes by fatty acids in adipocytes has been have no effect on the activation of gene expression by
difficult to ex amine within fat cells. In differentiated PPARs. The second is thatALBP/aP2 could have a negative
adipocytes, the addition offatty acids has caused relatively effect by binding and sequestering endogenous fatty acids,
small increases in message levels. This could be due to the preventing their metabolism into activating compounds,
38

thereby reducing the total concentration of available PPAR of this manuscript. This work was supported by American
activating ligands. Additionally, ALBP/aP2 could be capable Institute ofCancer Research grant 95B039 to DAB.AVH was
of binding the metabolites of fatty acids (including the supported by National Institutes ofHealth NRSA 1 F32 DK
activating ligands ), again resulting in their removal. Either 09599-01 ZRG2 01. In addition, the assistance of the
of these would result in a decrease in the activation of gene University of Minnesota Obesity Center is gratefully
transcription by fatty acids in the presence of ALBP/aP2. acknowledged.
A third possibility is that ALBP/aP2 eould have a positive
effeet on gene activation by binding a fatty acid or metabolite
and delivering it to a fatty acid regulatory pathway.
References
Alternatively, ALBP/aP2 could bind to an inhibitory
compound which interferes with PPAR aetivation. The
I. Bemlohr DA. Simpson MA: Adiposc tissuc and lipid metabolism. In:
results presented here are consistent with a positive effect
D.E. Vance, I.E. Vanee (eds). Biochemistry of Lipids, Lipoproteins
of ALBP/aP2 on PPAR-dependent activation of gene and Membranes. Elsevier Science, The Netherlands, 1996, pp 257-
transcription. 281
The involvement of ALBP/aP2 in the increased PPAR 2. Bemlohr DA, Simpson MA, Hertzel AV, Banaszak LJ: Intracellular
activation could occur either through a direct or indirect lipid-binding proteins and their genes. Annu Rev Nutr 17: 277-303,
1997
mechanism and could manifest itself at various points ofthe 3. Keller H, Dreyer C, Medin 1, Mahfoudi A, Ozato K, Wahli W: Fatty
fatty acid regulatory pathway. ALBP/aP2 could exert its effect acids and retinoids control lipid metabolism through activation of
at an early stage in the signaling pathway, initiating the peroxisome proliferator-activated receptor-retinoid X receptor
metabolism of PUFA through delivery to the metabolie heterodimers. Proc Nat! Acad Sei 90: 2160--2164, 1993
enzyme responsible for producing an active PPAR ligand. 4. Clarke SO, lump OB: Dietary polyunsaturated fally acid regulation
of gene transcription. Annu Rev Nutr 14: 83-98, 1994
Additionally, ALBP/aP2 might traffiek the fatty aeid or its 5. Lemberger T, Desvergne B, Wahli W: Peroxisome proliferator-
metabolite directly to PPARy, thereby requiring direet activated receptors: A nuclear receptor signaling pathway in lipid
protein-protein interaetions. This process could increase the physiology. Annu Rev Cell Dev Bio112: 335-363, 1996
half-life of such a compound and/or increase the rate of 6. Forman BM, Tontonoz P, Chen 1, Brun RP, Spiegelman BM, Evans
ligand binding to the receptor, cither of which would rcsult in RM: 15-Deoxy-delta 12, 14-prostaglandin 12 is a ligand for the
adipocyte determination factor PPAR gamma. Cell 83: 803-812,
a greater efficiency oftranseriptional activation.Altematively 1995
this may be an indirect effect mediated through binding of 7. Kliewer SA, Lenhard 1M, Willson TM, Patel I, Morris DC, Lehmann
endogenous lipophilic PPARyinhibitors, although currently 1M: A prostagIandin 12 metabolise binds peroxisome proliferator-
there is no evidence that such an inhibitor exists. Independent aotivated receptor gamma and prornotes adipocyte differentiation. Cell
ofthe mechanism of action, ALBP/aP2 was shown to be a 83: 813-819, 1995
8. Issemann I, Green S: Activation ofa member ofthe steroid hormone
positive effector of gene transcription which suggests a receptor superfamily by peroxisome proliferators. Nature 347: 645-
central role for ALBP/aP2 in the fatty acid control of gene 650,1990
expression. This effect was seen in CV- 1 fibroblasts where 9. Mandrup S, Lane MD: Regu1ating Adipogenesis. 1 Biol Chem 272:
concentrations of activating ligands for PPARs would be 5367-5370,1997
low, implying a sensitizing function of ALBP/aP2 with 10. Kliewer SA, Forman BM, Blumberg B, Ong ES, Borgmeyer U:
Differential expression and activalion of a family of mutine peroxisome
respect to the PPAR-dependent regulation of gene expression. proliferator-activated receptors. Proc Nat! Acad Sei 91: 7355-7359,
The presenee of ALBP/aP2 may ensure the completion of 1994
the lipid signalling pathway even when concentrations of 11. Kliewer SA, Umesono K, Noonan DJ, Heyman RA, Evans RM:
the signal may be somewhat limiting. Interestingly, since Convergence of 9-cis retinoic acid and pcroxisomc prolifcrator
this enhancement by ALBP/aP2 acts on its own gene, a signalling pathways through heterodimer formation oftheir receplors.
Nature 358: 771-774, 1992
positive autoregulatory loop is ereated to inerease the 12. DiRenzo 1, Soderstrom M, Kurokawa R, Ogliastro M-H, Ricote M,
expression of a lipid-binding protein in response to dietary Ingrey S, Horlein A, Rosenfeld MG, Glass CK: Peroxisome
fatty acids. Experiments are currently underway to meehanisti- proliferator-activated receplors and retinoic acid receptors differentially
eally characterize the lipid-binding protein stimulation of control the interaclions of retinoid X receplOT helerodimers with
transeription. ligands, coactivators, and corepressors. Mol Cell Bio117: 2166--2176,
1997
13. Hörlein Al, Näär AM, Heinzel T, Torchia 1, Gloss B, Kurokawa R,
Ryan A, Kamei Y, Söderström M, Glass CK, Rosenfcld MG: Ligand-
Acknowledgments independent repression by the thyroid hormone receptor mediated by
a nuclear receptor co-repressor. Nature 377: 397-404,1995
14. Yang X-J, Ogryzko VV, Nishikawa I-i, Howard BH, Nakatani Y: A
We would like to thank members ofthe Bemlohr laboratory p300lCBP-associated factor that competes with the adenoviral
for helpful suggestions and comrnents during the preparation oncoprotein ETA. Nature 382: 319-324, 1996
39

15. Nagy L, Kao H-Y, Chakravarti 0, Lin RJ, Hassig CA, Ayer OE, of Adipoeytes. Mol Cell Bioehern. (in press) 1997
Schreiber SL, Evans RM: Nucle.r receptor repression mediated by • 22. Sambrook J, Fritsch EF, Maniatis T: Moleeular Cloning:ALaboratory
complex eontaining SMRT, mSin3A, and histone deacetylase. Cell Manual, 2nd edition. Cold Spring Harbor Laborntory Press, Cold
89: 373-380, 1997 Spring Harbor, NY 1982
16. Spencer TE, Jenster G, Burein MM, Allis CD, Zhou JX, Mizzen CA, 23. Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson
Mckenn. NJ, Onate SA, Tsai SY, Tsai MJ, Omalley BW: Steroid reeeptor TM, Kliewer SA: An antidiabetie thiazolidinedione is a high affinity
eoactivator-I is a histone .cetyltransferase. Nature 389: 194--198, 1997 ligand for peroxisome proliferator-activated receptor gamma (PPAR
17. Tontonoz P, Hu E, Devine J, Beale EG, Spiegelman BM: PPAR-y2 gamma). J Biol Chem 270: 12953-12956, 1995
regulates adipose expression ofthe phosphoenolpyruvate carboxykinase 24. Ross SR, Graves RA, Greenstein A, Platt KA, Shyu HL: A fat-speeifie
gene. Mol Cell Bio115: 351-357, 1995 enhaneer is the primary determinant of gene expression far adipocyte
18. Tontonoz P, Hu E, Graves RA, BudavariAI, Spiegelman BM: mPPAR P2 in vivo. Proc Nat! Acad Sei 87: 9590-9594, 1990
gamma 2: Tissue-specifie regulator of an adipoeyte enhaneer. Genes 25. Kletzien RF, Foellmi LA, Harris PKW, Wyse BM, Clarke SO:
Dev 8: 1224--1234, 1994 Adipocyte fatty acid-binding protein: Regulation of gene expression
19. Amri E-Z, Bertrand B, Ailhaud G, Grimaldi P: Regulation of adipose in vivo and in vitro by an insulin-sensitizing agent. Mol Pharm 42:
cell differentiation. 1. Fatty acids .re indueers of the aP2 gene 558--562, 1992
expression. J Lipid Res 32: 1449-1456, 1991 26. Harris PK, Kletzien RF: Loealiz.tion of a pioglit.zone response
20. Grimaldi PA, Knobel SM, Whitcsell RR, Abumrad NA: Induetion of element in thc adipocytc fatty acid binding protein gene. Mol Pharm
aP2 gene expression by nonmetabolized long-ehain fatty acids. Proe 45: 439-445,1994
NatlAead Sei 89: 10930-10934, 1992 27. Tontonoz P, Hu E, Spiegelman BM: Stimulation of adipogenesis in
21. Simpson MA, LiCata V; Ribarik-Coe R, Bemlohr DA: Bioehemieal fibroblasts by PPARyl, a lipid activated transcription factor. Cell 79:
and Biophysical Analysis of the Intraeellular Lipid Binding Proteins 1147-1156, 1994
Molecular and Cellular Biochernistry 188: 41-48, 1998.
© 1998 Kluwer Acadernic Publishers.

The molecular basis for the role of zine in


developmental biology
Kenneth H. Falchuk
Department oj"Medicine, Brigham and Women 's Hospital and Center for Biochemical and Biophysical Sciences and
Medicine, Harvard Medical School, Boston, MA, USA

Abstract

Zine regulates the gene expression maehinery. It affects the structure of chromatin, the template function of its DNA, the activity
of numerous transcription factors and ofRNA polymerases. Hence, it determines both the types of mRNA transcripts synthesized
and the rate of transcription itself. Alterations in one or more of these zinc dependent processes have been proposed to account
for the proliferative arrest and teratology induced by zinc deficiency. To examine this proposal, studies ofzine duringX laevis
development have been initiated. The kinetics of X laevis ooeyte zinc uptake and storage and of zinc utilization during
embryogenesis have been examined first. Vitellogenin carries zinc into the ooeyte. Ten % ofthe total zinc (10 ng/egg) remains
within the cytosol while 90% (90 ng/egg) is stored in the yolk platelets associated with lipovitellin. The cytosolic pool is the
souree of the zinc for all newly formed metalloproteins involved in embryo development. The yolk platelet zinc pool is stored
for later use during early metamorphosis. It is now possible to exarnine zinc transfer to moleeules, such as e.g. transcription
factors, and the role of the metal in their function in development and organogenesis. (Mol Cell Biochem 188: 41-48,1998)

Key words: zinc, transeription factors, gene expression, organogenesis, Xenopus laevis

Introduction Zinc as a determinant of chromatin composition, structure


and function
There is a large body of information available on the effects
ofzinc and its deficiency on the growth and development of A major function ofzinc in biology is its role in chromatin
diverse types of cells and whole organisms, including humans biochemistry [I]. Most ofthe available information has been
[1). In all cases, zinc is required for cell division while its obtained from studies ofthe effects of zinc and its deficiency
deficiency is known to arrest proliferation, suppress growth on E. gracilis, though sparse, but consistent, data on mam-
and cause congenital mal formations in offspring of zinc malian organisms havc bcen reported [3, 4, 8-12]. In the
deprived animals [1, 2). Recently, intracellular zinc has been chromatin of zinc sufficient eells, the principal molecules
shown to have a role in determining the eomposition and associated with genomic DNA are the standard histone and
structure of chromatin as weil as in selective gene expression non-histone proteins. The ratio ofhistone protein to DNA is
controlling quantitative and qualitative aspects of trans- one. These proteins organize chromatin into an ordered
cription [3-7). Ultimately, therefore, the metal determines the repeating structure of approximately 150-250 base pairs, the
types of proteins formed by cells. This direct involvement in basic nucleosome unit that is readily generated following
processes fundamental to gene expression has provided the standard nuclease digestion [8]. In contrast, the composition
basis for understanding the molecular events that underlie its and structure of chromatin of zine deficient organisms is
essential requirement for cellular proliferation, differ- radicallyaltercd. In E. gracilis, DNA content doubles! [4].
entiation, growth and organogenesis. The types and amounts of proteins bound to this nucleic acid

Addressfor ojJprints: K.H. Falchuk, Depaliment ofMedicine, Brigham and Women's Hospital and Center for Biochemical and Biophysical Sciences and
Medicine, Harvard Medical Schoo1, Seeley G. Mudd Building, 250 Longwood Avenue, Boston, MA, 02Jl5, USA
lWhen cell division ceases in zinc deficient E. gracilis organisms the cell cycle is blocked at the premitotic stage ofS/G2 • Deprivation of other metals block
the cell cyc1e at other stages. Specifically, deficiency of iron blocks the cell cyc1e at GI' of magnesium at cytokinesis while of manganese at both karyokinesis
and cytokinesis [14-16].
42

and their state of chemical modification differ from control the 24 hinterval that follows zinc replenishment. The
[8-10, 13]. The ratio ofhistone protein to DNA is reduced arginine-rich DNA binding polypeptides and the novel u-
from I to about OA. The histones that are present bind more amanitin-resistant RNA polymerase disappear while new
tightly to DNA due to an increase in the extent to which they histones and the three types ofRNA polymerases are fonned.
are acetylated. The total amount ofbasic proteins relative to Ultimately, chromatin structure, composition and function
DNA is still unity, however, since a set of arginine rich return to nonnal and proliferation then resumes. The cycle
polypeptides of3-5 kD in size replace histones on a weight can be repeated as often as desired first by reducing intra-
basis. These polypeptides, together with the chemically cellular zinc content and then bringing it back to levels
modified histones, exert a major influence on the structure characteristic of nonnal cells.
and function of chromatin. The size of the digests detected
when zinc deficient chromatin is treated with standard
concentrations of nucleases is over 2000 base pairs compared Formulation ofhypothesis that zinc regulates selective
with the 150--250 of the basic nucleosomal unit. Ten-fold gene expression
higher eoncentrations of nuclease are required to produce
units ofDNA ofthe size ofnonnal nuclesomes. This major Taken together, the results demonstrate that in the presence
increase in the state of chromatin condensation reduces the of adequate intracellular zinc content genes are activated and
ability ofDNA to serve as a template for RNA polymerase. proteins are formed that allow for cell division and other
As a result, the rate of3H-uridine incorporation into RNA is normal cellular processes to take place. In contrast, in
decreased in zinc deficient cells [3 V response to a decrease in intracellular zinc, E. gracilis
The effects of zinc deficiency on the transcription process transcription is limited only to a selected set of genes that
itself, however, are not only quantitative but also qualitative, generate unique proteins. In turn, the proteins condense its
indicating participation of other zinc dependent molecular chromatin and shut down major components of their meta-
processes acting at the level ofselective gene activation and/ bolism, a process that is reversed by raising the intracellular
ar repression. Thus, while total RNA synthesis is deereased, zinc content. These findings led to the proposal, described
the types of messenger RNA found in zinc deficient cells are fonnally in 1981, that one ofthe fundamental biological roles
distinct from those ofthe corresponding zinc sufficient ones. of intracellular zinc is to regulate the expression of a defined
Their base composition as weil as their translation products number of genes [3,4]. The proposal fonnulated the concept
differ from those of control cells [17]. From among the unique that there existed zinc dependent gene regulatory proteins
translation products fonned by zine deficient cells, two have acting at the level of transcription. Nuclear zinc interacted
been identified: the 3-5 k Da basic DNA binding poly-
peptides described above as weil as a novel type 11 DNA
dependent RNA polymerase [3, 18]. Zinc sufficient cells
contain the three standard RNA polymerases responsible far
ActivatorslRepressors
synthesizing the three cJasses ofRNA. Each polymerase is a
zinc enzyme containing 2 gram atoms of Zn per mole of +Zn I -Zn ~
enzyme. In contrast, zinc deficient chromatin contains not
three but only a single, novel RNA polymerase that is also a
zinc enzyme and is fully functional in a standard trans- Transcription Transcription

criptional assay. The chromatographic behavior of the


mRNAs mRNAs
enzyme as weil as the typical inhibition constant observed
in response to incubation with 1,1 O-phenanthroline are
Translation Translation
characteristic of type 11 polymerases. However, the RNA
polymerase 11 from zinc deficient organisms is insensitive to
u-amanitin, a property that identifies it as a species distinct Enzymes, Inhibitory Absent,
from the correspondingone present in zinc sufficient eells. Regulatory Polypeptides, Non-Functional
Proteins Novel Enzymes Proteins
That zinc is the detenninant of the characteristic pattern
of pro teins formed during zinc deficiency as weil as the Proliferation
associated biochemical and biological consequences is Differentiation
Organogenesis +
continned by the reversal of all ofthe above findings during
Fig. 1. Zine aetivator and repressor moleeules determine the expression of
'Thc size, high eontent of basic amino acids and nuclear loealization of seleeted genes. The pattern of genes expressed (ar repressed) during zine
these polypeptides, together with their effeet on the state of chromatin deficieney differs from eontrol and leads to arrest of proliferation, altered
eondensation and on transeription, are analogous to the protamines in sperm. differentiation and organogenesis.
43

with these regulatory proteins to activate (or in some cases Table I. Zine transeription regulatory proteins
repress) transcription of speeific genes, whieh in turn,
Protien Zine, g at/mole Referenee
detennine the types of proteins fonned. In the absence ofthe
metal, a number of gene regulatory proteins either might not X laevis TFIIIA 2 6
X laevis TFlllA 7-11 7
be fonned at a11 or, iffonned, might remain as apomoleeules
Glueoeorticoid Receptor
that would lack funetion. On the other hand, other genes (407-556) Fragment 2 19
might be aetivated and their transcripts translate into inhibit- Estrogen Reeeptor
ory polypeptides or novel proteins. Together, such effeets on (185-250) Fragment 2 20
synthesis of funetional proteins and gene repression and/or YeastGAL4
(1-147) Fragment 2 21
activation could produee the phenotype ofthe zine defieient
HIV tat protein 2 22
ee11s and organisms (Fig. 1). A. nidulans ALCR
(7-58) Fragment 2 23
Yeast CYPl(HAP1)
Zinc in transcription Jactors: Confirmation oJ hypothesis (49-126) Fragment 2 24
Yeast PPRI
(1-118) Fragment 2 25
The model was eonfinned in prineipal by subsequent work LIMDomain
showing that a gene regulatory protein, the Xenopus laevis (lin-lI, RBTN, ISI-I) 2 26,27
transcription faetor TFIIIA, is a zine metalloprotein [6, 7]. HeIa eell SPI
The metal is essential to its ability to bind speeifieally to the (614-778) Fragment 28
K.laclis LAC9
5S RNA gene and aetivate its transeription; in its apofonn,
(1-228) Fragment 2 29
TFIIlAis devoid offunetion and 5S RNAis not fonned. The
availability of eDNA teehnology led other investigators to
identify the zine binding sites of TFITIA. The primary c10ned and expressed in a variety of veetors. The resultant
strueture of TFIIIA eontains highly eonserved sequenees polypeptides, seleeted to eontain their DNA binding do-
eomprised of two eysteine and two histidine residues sep- main(s), have been shown to bind zine and require the metal
arated by variable numbers of amino acids in 9 repeat units for interaetion with speeifie DNA. Zine binding has been
of about 30 amino acids. The Cys and His residues in eaeh demonstrated either by direet analysis by atomie absorption
of the eonserved repeat units of TFIIlA eould serve as zine speetrometry, X-ray absorption fine stmeture, XAFS, or by
ligands fonning tetrahedral eoordination eomplexes with one the zine blot teehnique. The list of the few transeription
zine atom [7]. The presenee of zine bound to these ligands faetors (inc1uding the c10ned and expressed DNA domain
in TFIIIA generates an intervening, compaet loop strueture fragments) that to date have been demonstrated to eontain
eontaining the DNA binding domain ofthe protein in the zine by direet analysis is shown in Table 1.
sequenee intervening between the pairs of Cys and His The zine bioehemistry of these transeription faetors,
residues, now reeognized as the c1assie 'zine finger' DNA inc1uding a detailed diseussion of the strueture/funetion
binding motif [7). relationship of the zine-protein interaetion, has been des-
An explosion of reports identifying hundreds of other eribed elsewhere [6, 7, 19-29,30]. Only seleeted, pertinent
putative zine binding domains and the establishment of an aspeets will be reviewed briefly here. The initial report on the
entire c1ass of zine transeription faetors, a11 c1assified under metal eontent of TFIIIA revealed the pro tein eontained 2
the tenn 'zine finger' proteins, resulted when homologous gram-atoms of zine/mole [6]. In a subsequent publieation the
zine bin ding sites were sought in other transcription faetors. 7S intaet protein-RNA particles were shown to bind from 7-
The putative zine binding sites in these transeription factors 11 zine atoms [7]. The presenee of 9 repeat units in the
are believed to be eomprised of different eombinations of four primary sequenee of TFIIIA would imply that the protein
ligands from eysteine and/or histidine residues. These could eontain up to 9 zinc atoms. However, not a11 of the
proteins are presumed to eontain zine on the basis of the TFIIIA 'zine fingers' units may be required far DNA binding
presenee of consensus sequenees in their predicted primary with high affinity and sequenee speeifieity. Thus, a peptide
struetures. Unlike TFIIIA, however, most of these trans- eontaining 'fingers' 1-3 binds almost as tightly to specifie
cription faetors, have never been isolated to homogeneity and DNA fragments as the nine 'finger' mole eule [31]. The aetual
eharaeterized as zine proteins by standard analytieal and number of zine atoms needed by TFIIIA to earry out its
funetional methods sinee they are present in exeeedingly nonnal aetivity, therefore, still needs to be rcsolvcd expcri-
sm all quantities in the eell. Henee, they should be eonsidered menta11y. It should be noted that two or three zine atoms
only as putative zine proteins until direet information is appears to be the most frequent number required for function
available on their zine eontent. However, far a number of by a11 of the other transeription faetors examined thus far
these transcription faetors, fragments ofthe protein have been (Table 1).
44

The glueoeortieoid and estrogen reeeptors are members of of the metallothionein gene is also zine dependent [40].
the multigene family that includes reeeptors for retinoie acid, Transeription ofhistone 4 requires the interaction ofH4TF-
thyroid hormone and vitamin D3. The primary strueture of land H4TF-2 to an H4 promoter upstream of the trans-
the DNA binding domain of the glueoeortieoid reeeptor eription start site [41]. Gel shift assays show that binding of
eneompasses I His and 9 Cys residues found within amino both ofthese faetors to the H4promoter is abolished by 1,10-
acid residues 440-525 [19]. The eysteines, but not the His, phenanthroline or EDTAand is restored by zine [42]. The zine
serve as the ligands for the 2 gram-atoms of zine/mole requirement ofthe histone gene aetivator is eonsistent with data
associated with the reeeptor. Eaeh metal is eoordinated to four showing that the histone contentIDNA is markedly redueed in
isolated -S-ligands. There are no bridging sulfur ligands. The zine deficient E. gracilis [8-10] and rat liver [11, 12].
intervening DNA binding sequenee is loeated between and
anehored by the two zine atom eomplexes. The DNA binding
motif is helical and does not have a 'finger' strueture [32]. Transcription Jactors involved in development that could
The GAUprotein from yeast eontains two zine atoms, the require zinc Jor Junction
metal is required for aetivation of genes utilized for galactose
metabolism [21, 33]. The two metal atoms of GAL4 are As deseribed above, zine defieieney results in a pattern of
eoordinated to 6 Cys, two of whieh form bridging ligands aetivation andlor repression of a set of genes that is quite
between the metal atoms. The resultant strueture, therefore, distinet from that of the zine sufficient state [3]. This pattern
differs from that ofTFIIIA and of the glueoeortieoid reeeptor. manifests in the formation of partieular gene produets
The binuclear metal clusters in the GAL4 protein are similar together with the lack of synthesis of others and has been
to those observed in metallothionein [21, 34]. proposed to aeeount for the arrest of proliferation and
The binding domain of the transaetivating tat protein from teratology observed with zine defieient organisms. Failure to
HIV also differs from that of the 'zine finger' struetures. express appropriate maeromoleeules required for develop-
Optieal absorption speetroseopy reveals it binds 2 zine atoms ment of an embryo at a critical juneture in the proeess of
per monomer and forms a metal linked dimer. Cysteine
residues are involved as the metalligands [22].
Table 2. Putative zine transeription faetors involved in development
The struetural information available from the limited
number of zine transeription proteins that have been eharae- Gene [*] Tissues expressing Effect of absence or
terized, therefore, makes it possible to eonelude that the gene mutation

ligands that bind zine and the resultant DNA binding motifs scratch [43] neural precursor eells Decrease eye
generated when the metal associates with the protein are quite photoreeeptors and
varied [35]. These proteins have been eategorized into neuronalloss
castar [44] Delaminated CNS Reduction in CNS
different strueturally distinet groups that likely represent a
neuroblasts, ventral axonal network
family of molecules related by their ability to bind to specifie midline glial
DNA sequenees and aetivate their transcription. They are precursors
eategorized as 'zine fingers' (TFIIIA), 'zine twists' (gluco- spelt [45] Terminal pattern Absence of terminal
eorticoid and estrogen reeeptors) and 'zine clusters' (GAL4 elements elements
Krox20 [46] Hindbrain Loss of rhombomeres 3
protein) [35]. The limited struetural data already available on
and 5, Fusion of
such proteins as the HIV tat protein, UM domain, RING trigeminal, facial and
finger and PKC CRRD suggest the existenee of yet other vestibular ganglia
distinet motifs [36]. The funetional signifieanee of such Kiz-l [47] Olfactory epitbeleal ?Malformation
diverse DNA binding domains will need to be elucidated. eells
cKr2 [48] Schwan eells, eephalie ?Malformation
A requirement for zine has been reported for other gene
and neural erest
regulatory proteins not listed in Table I. While the metal is derived tissues
needed for aetivity, zine analyses have not been earried out. zic [49] Early embryonic stage Neural Tube defects
The transcription faetor ADRI does not bind to DNAandior neural tube and
aetivate transcription either in the absence of added metal or granule cells of
developing cerebellum
in the presenee of ehelating agents [37]. Zine addition, but
zfh-4 [50) Midbrain ?Malformation
not of other metals, enhanees the interaction of the 4.5 S ova [51] Female garnetes ?Altered gametogenesis
dihydrotestosterone-reeeptor with rat prostate tumor nuclei Kr [52) Abdominal Segment Absence of tboraeie
[38]. The metal regulatory element(s), MRE, binding proteins and abdominal segments
of the metallothionein gene will not bind to specifie DNA MZFI [53] Hematopoetic cells Altered hematopoesis
Egr-l [54]
regions in the presenee ofEDTA or I, I O-phenanthroline [39].
The binding of a rat liver protein to the mouse MRE-a region [*] Reference
45

organogenesis or to supply and ineorporate a metal into


proteins that require it for funetion eould result in abnormal
phenotypes. I-
Z
• • • . -..
X. laevis (ng)
The identifieation of the speeifie zine transeription faetors W
that eould be affeeted and whieh eould be responsible for the
pathology of zine deficieney have yet to be earried out. A
I-
Z
o • w·_·_·_·_·_·_·.·_·_JL·_·_·-.-·_~
Sea urchin (pg) •
ü
number of transcription faetors, known to be expressed in
eells and tissues of developing organisms, are eandidates for
c:
N ......._.......................................................
Mouse (pg)
eonsideration. They have been eategorized as belonging to
the class of 'zine fingers, zine clusters or zine twists'
moleeules sinee they contain sequenees homologous to those I I

of the known zine transeription factors. While these have not 0.0 0.5 2 5 10 20 50
been isolated or eharaeterized as aetual zine proteins, their HOURS AFTER FERTILIZATION
funetions in developmental proeess are of partieular interest
here. The ones deseribed below (Table 2) are seleeted from Fig. 2. The zine eontent of Xenopus laevis, sea urehin and mouse ooeytes
among many others on the basis of their apparent involve- and embryos. Each embryo is a 'closed system' in tenns ofits zine during
study period, henee, the metal content remains constant during major
ment in the formation of tissues, such as the nervous,
developmental processes.
reproduetive, museuloskeletal and hematologieal systems, all
known to be targets of zinc deficiency [2]. In some instances,
gene mutation or knoek out experiments have been earried while at the other is the mouse that proeeeds to early blastula
out and resulted in the lack of expression of normal gene stage. Therefore, in eaeh ofthese embryos any zine needed
produets aeeompanied by either abnormal or even absent to form aetive zine trascription faetors (or other zinc depend-
anatomieal struetures and/or organs (Table 2). ent molecules) from the eorresponding apoproteins during
embryogenesis must be aequired from stores already existing
within the mature ooeyte (in the ease of X laevis and sea
A biological system to study zinc transcription factors in urchin) or one eelled-embryo (in the ease ofthe mouse).
developing organisms Of the three organisms, Xenopus laevis ooeytes and
embryos are the most suitable system for study ofthese zine
The gene produets deseribed in Table 2 serve to generate an stores and the kineties oftheir re-distribution to newly formed
initial list of speeifie examples with which to study the apoproteins. Thus, their zine contents are 50 and 100 X
bioehemieal funetions and eharacteristies of their trans- greater than that of sea urchin and mouse eggs and embryos,
eription products in both zine sufficient and defieient organ- respeetively. In addition, an extensive body of information
isms. The objeetive of such studies would be to link a regarding the maeromoleeular events for the induetion and
functional alteration in a putative zine transcription faetor and formation of its organs has emerged in the past decade.
a specifie developmental abnormality to zine defieieney. Finally, large numbers ofXenopus laevis ooeytes at eaeh stage
Towards this end, a biological system is required to examine of oogenesis and embryos at different points of development
the relationship between zine, transeription faetors, differ- ean be obtained [56]. The first steps are to understand the
entiation, organogenesis and other aspects of development. metabolism of zine in this biologieal system.
Xenopous laevis, sea urchin and mouse ooeytes and embryos The frog oocyte requires up to three years to develop into
have been explored for their utility in this regard. From the a mature egg. During that period of time it forms proteins,
perspeetive of zinc content, mature Xenopous laevis eggs DNA, RNA and aceumulates and stores nutrients and pre-
contain about 100 ng of zine while sea urchin and mouse have cursors to be used for embryogenesis, a proeess it aehieves
20 and 2 pg ofzine, respeetively [1, 55]. Following fertili- within less than 48 h [55]. Zine is one ofthe essential nutrients
zation, the zine eontent of Xenopous laevis and sea urchin that is aequired from matemal sources during oogenesis [56].
eggs remains unehanged while that of the mouse inereases To insure that zine dependent steps essential for eell division
to 7 pglembryo prior to the first cleavage (Fig. 2). Beyond and differentiation are earried out in the short time interval
this change in the mouse oocyte following fertilization, all encompassed by embryogenesis, the frog has evolved sys-
three embryos remain 'closed systems' relative to zine. In the tems that regulate zine transport, uptake and distribution both
case of sea urchin and mouse, the eontent of the metal is within the ooeyte and the embryo. A number of the eom-
constant for at least 50 h post fertilization, and in the frog for ponents of the zine uptake and storage system in the egg have
over 75 h (Fig. 2). This period eneompasses major develop- now been identified.
mental events. At one extreme is the frog, which undergoes Ooeyte zine uptake is aehieved through endoeytosis of
the entire process of embryogenesis and early metamorphosis vitellogenin [57], a protein that contains 2 gram atoms ofzinel
46

mole of dimer [58]. Moreover, onee vitellogenin is taken up, of ocular bud regions. The spine and tail regions evidenee
it is proeessed into at least two major proteins, Iipovitellin blebs. Somites and heart are absent. The malformed embryos
and phosvitin, that are stored in yolk platelets to be used after survive for approximately 24--48 h after hatehing. The same
the ooeyte is fertilized. Lipovitellin is the ooeyte protein that eoneentration of non-chelating analogues does not induee any
eontains the zine [58]. Onee the ooeytes are fully mature, they teratology.3
attain a zine eoneentration of about 1 mM [56]. Exposure of the embryos to the ehelating agent for only
InX laevis, the zine in those stores is distributed into two limited amounts of time and for only particular stages of
eompartments (Fig. 3). The first eompartment eontains over development (stages 1-6 or 7-15 or 16-25) identifies the
90% ofthe total and is in the yolk platelets. The seeond, with most suseeptible developmental stages . Thus, the time
the remainder ofthe zine, is in the eytosolie fraetion [57]. The required to proeeed through the c1eavage (stages 1-6) and
larger of the two pools remains sequestered in the platelets gastrulation steps (stages 16-25) are prolonged by exposure
and is inaeeessible to eytosolic apoproteins throughout the to the chelating agent but ectoderm, endoderm and mesoderm
entire period of embryogenesis. The smaller one, on the other are induced and the neural fold and plate proeeed unhindered.
hand, is already present in the cytosol and is postulated to The resultant embryos are relatively normal by histological
constitute the sole source ofzinc required for embryogenesis. and gross morphologie al eriteria. In contrast, when embryos
In quantitative terms, this cytosolic pool (eontaining about 10 progressing through steps involved with the migration of
ng of zine) is sufficient to provide the required metal for the germ line cells and early stages of organization of future
formation ofthe entire embryo. The yolk platelet zinc begins organs (stages 7-15) are exposed to the chelating agent its
to be redistributed to the cytosol several days after the tadpole final appearance is markedly altered and is characteristic of
has hatched and early metamorphosis has begun (Fig. 3). the typical teratology 01' zine defieieney. Hence, stages (7-
Curtailment of the zine available from the putative cyto- 15) are the most sensitive ones to reduetions in zine availa-
solic stores can be aeeomplished by use of ehelating agents, bility in the presenee of a ehelating agent. This eonc1usions
such as 1, lO-phenanthroline [59]. Embryos ineubated in the frames the developmental time period in whieh to search for
presenee of the agent at 10- 5 M do not develop normally. the eorresponding zinc proteins whose funetion(s) are
About 74% of the embryos hateh, the remainder do not perturbed.
survive to that stage of development. The gross external This work now sets the stage to determine the meehanism( s)
appearanee ofthe survivors is markedly affected, exhibiting 1'or zine exchange in embryo eells and the identifieation and
classic teratology ofzine defieieney [2, 59]. They are smaller study of the zine proteins involved in developmental proeesses.
in size compared to control, manifest craniofacial mal form-

,
ations, including mieroeephaly. They do not form head
structures, the brain and eyes are absent and there is extrusion Acknowledgment

This work was supported, in part, by the Endowment for


100 r1j~--------------------------~
Research in Human Biology, Inc.
-
~.,....' - - - -.....
...J
80
---=-
8lASTVLA J
References
~ 60 <=>
GASTRULA
~ 1. Vallee BL, Falchuk KH: The biochemieal basis of zine physiology.
<=J
ORGANOOENESIS Physiol Rev 73: 79--111 , 1993
2. Keen CL, Hurley LS: Zine and reproduetion: Elfeets of deticiency on
MerAMORPHOSIS foetal and postnatal development. In: CF Mills (ed). Zinc in Human

,,•
......._.........................-.
Biology, Springer-Verlag, London, pp 183-220, 1989
20 3. Vallee BL, Falchuk KH: Zine and gene expresson. Philos Trans R Soc
~, Lond B Biol SCI 294: 185-197,1981
4. Falchuk KH: Zinc defieieney and the Euglena gracilis chromatin. In:
O~J--~~~~--~--~~~~--~
o 2 3 5 10 20 30 50 100
HOURS POST FERTILIZATION
' While the ehelating agent ean bind to other metals, zine is the most Iikely
target meta!. The teratology observed (see text) is similar to that noted in
Fig. 3. Changes in zinc content of cytosol (dotted !ine) and yolk platelet the olfspring of other animals rendered zine defieient [21 and differs from
(solid !ine) fractions in Xenopus laevis during different stages ofembryo- any reported for iron or copper, other metals that could interae! with the
genesis and metamorphosis. Each fraction remains unchanged until the phenanthroline. The stability eonstants of zine phenanthro!inates exeeeds
tadpole initiates the proeess of early metamorphosis and the yolk platelet those ofiron [59]. finally, in the frog oocyte there is twiec as much zinc as
fraction is redistributed to the eytoso!. iron and one hundred fold more as copper [56].
47
A. S. Prasad (cd). Essential and Toxie Traec Elements in Human Health 25. Ball LJ, Diakun GP, Gadhavi PL, Young NA, Annstrong EM, Garner
and Disease, New York: Liss, pp 75-91, 1988 CD, Laue ED: Zine eoordination in the DNA-binding domain ofthe
5. Falehuk KH: Zine in developmental biology: the role of metal yeast transeription activator PPR 1. FEBS Lett 358: 278--282, 1995
dependent transeriptional regulation. In: A. S. Prasad (ed). Essential 26. Li PM, Reiehert J, Freyd G, Horvitz HR, Walsh CT: The UM region
and Toxie Trace Elements in Human Health and Disease: An Update. of a presumptive Caenorhabdilis elegans transeription faetor is an
NewYork: Liss,pp91-11I, 1993 iron-sulfur and zine eontaining metallodomain. Proe Nat! Aead Sei
6. Hanas JS, Hazuda 0, Bogenhagen DF, Wu FY-H, Wu C-W: Xenopus USA 88: 9210-9213,1991
lranseription faetor A requires zine for binding to the 5S gene. J Biol 27. ArcherVE, Breton J, Sanehez-Gareia I, Osada H, ForsterA, Thomson
Chem 258: 14120-14125, 1983 AJ, Rabbitts TH: Cysteine-rich UM domains of Lim-homeodomain
7. Miller J, McLaehlanAD, KlugA: Repetitive zinc binding domains in and UM-only proteins eontain zine but not iron. Proe Nat! Acad Sei
the protein transcription faetor IIIA from Xenopus ooeytes. EMBO J USA 91: 316-320, 1994
4: 1609-1614,1985 28. Kuwahara J, Coleman JE: Role ofzine (I1) ions in the strueture ofthe
8. StankiewiczA, Falchnk KH, Vallee BL: Composition and structure of three-finger DNA binding domain of the SPI transcription faetor.
zine deficient Euglena gracilis chromatin. Bioehemistry 22: 5150- Bioehemistry 29: 8628--8631, 1990
5156,1983 29. Halvorsen YC, Nandabaln K, Dickson RC: LAC 9 DNA-binding
9. Mazus B, Fa1chnk KH, Vallee BL: Histone formation, gene expression domain coordinates two zine atoms per monomer and eontaets DNA
and zine defieieney in Euglenagracilis. Bioehemistry 23: 42-44, 1984 as a dimer. J Biol Chem 265: 13283-13289, 1990
10. Czupryn M, Falehuk KR, Vallee BL: Zine defieiency and metabolism 30. Falchuk KH: Transcription faetors in cellular differentiation and
ofhistones and nonhistone proteins inEuglena gracilis. Bioehemistry organogenesis. J Traee Eiern Exp Med (in press)
26: 8263-8269, 1987 31. Liao X, Clemens KR, Tennant PE, Wright JM, Gottesfeld M: Speeifie
11. Castro CE, Alvares OF, Sevall JS: Diet-mediatcd alteration o[ inleraction of the first three zine fingers of TFIIIA with the internal
chromatin strueture. Federation Proe 45; 2394-2398,1986 eontrol region oftheXenapus 5S RNAgene. J Mol Bio1223: 857, 1992
12. Caslro CE, Alvares OF, Seval JS: Zine defieiency deereases histone 32. Pan T, Freedman LP, Coleman JE: Cadmium-I 13 NMR studies ofthe
III * in rat liver. Nutr Rep Int 34: 67-75,1986 DNA binding domain of the mammalian glueoeortieoid reeeptor.
13. Falchnk KH, Gordon PR, StankiewiezA, Hilt KL, Vallee BL: The E. Bioehemistry 29: 9218-9225, 1990
Gradlis chromatin: A eomparison of the effeets of zine-, iron-, 33. Johnston M: Genetie evidenee that zine is an essential eofaetor in the
magnesium-, or manganese-dcfieieney and cold shoek. Bioehemistry DNA binding domain of GAL4 protein. Nature Lond 328: 353-355,
25: 5388-5391, 1986 1987
14. Falchuk KR, Faweett 0, Vallee BL: Role of zine in eell division of 34. Kagi JHR, Kojima Y: Nomenclature of metallothionein. Experientia
Euglenagracilis. JCell Sei 17: 57--68,1975 Suppl Basel 52: 19-22, 1987
15. Falehuk KH, KrishamA, Vallee BL: DNAdistribution in the eell eycle 35. Vallee BL, Coleman JE, Auld OS: Zine fingers, 7ine clusters, and zine
of Euglena gracilis: Cytofluorometry of zine defieient eells. Bio- twists in DNA-binding protein domains. Proe Nat! Aead Sci USA 88:
ehemistry 14: 3439-3444, 1975b 999-1003,1991
16. Waekei WEC: Nucleie acids and metals III. Changes innucleie acids, 36. Schwabe JWR, KlugA: Zine mining for protein domains. Nature Struet
protein and metal content as a eonsequenee of zine defieiency in Biol 1:345-349, 1994
Euglena gracilis. Bioehemistry I: 859-865, 1962 37. Eisen A, Taylor WE, Blumberg H, Young ET: The yeast regulatory
17. Cross1cy LL, Falchuk KR, Vallee BL: Messenger ribonucleie acid fune- protein ADRI binds in a zine dependent manner to the upstream
tion and protein synthesis in zine defieient E. gracilis. Bioehemistry aetivating sequenee of ADH2. Cell Biol 8: 4552-4556, 1988
21: 5359-5363, 1982 38. Colvard OS, Wilson EM: Zinc potentiation of androgen reeeptor
18. Fa1chuk KR, Mazus B, Ber E, Ulpino-Lobb L, Vallee BL: Zine defieieney binding to nuclei in vitra. Bioehemistry 23: 3471-3478, 1984
and the Euglena gracilis chromatin: Formation of an alpha-amanitin- 39. Scguin CA: Nuclcar factor rcquires Zn 2+ to bind a regulatory MRE
resistant RNA polymerase II. Bioehemistry 24: 2576-2580, 1985 element ofthe mouse gene eneoding metallothionein-l. Gene 97 :295-
19. Freedman LP, Luisi BF, Korszun ZR, Basavappa R, Sigler PB, 300, 1991.
Yamamoto KR: The funetion and strueture of the metal eoordination 40. Searle, PF: Zine dependent binding of a liver nuclear faetor to metal
siles within the glueoeortieoid receptor DNA binding domain. Nature response element MRE-a of the mouse metallothionein gene and
Lond 334: 543-546, 1988 variant sequenees. NucleieAcid Res 18: 4683-4690,1990
20. Schwabe JWR, Neuhaus 0, Rhodes 0: Solution strueture ofthe DNA- 41. Dailey L, Boseman-Roberts S, Heintz N: RNA polymerase II
binding domain ofthe oestrogen receptor. Nature Lond 348: 458-460, transeription factors H4TF-I and H4TF-2 require metal to bind speeific
1990 DNA sequenees. Mol Cell Biol 7: 4582-4584, 1987
21. Pan T, Colcman JE: GAL4 transeiption faetor is not a 'zine finger' but 42. Dailey L, Hanly SM, Roeder RG, Hinotz N: Distinet transcription
forms a Zn(I1),Cys6 binuclcar cluster. Proe Nat! Aead Sei USA 87: faetors bind speeifieally to two regions of the human histone 4
2977-2981, 1990 promotor. Proe Nat! Aead Sei USA 83: 7241-7245, 1986
22. Frankel AD, Chen L, Cotter RJ, Pabo CO: Tat protein from human 43. Roark M, Sturtevant MA, Emery J, Vaessin H, Grell E, Bier E: scratch,
immunodefieiency virus forms a metal-linked dimer. Seienee Wash a pan-neural gene eneoding a zinc finger protein related to snail,
DC 240: 70-73, 1988 prornotes neuronal development. Genes Dev 9: 2384-2390, 1995
23. Sequeval 0, FelenbokB: Relationship between zine eontent and DNA- 44. Melleriek DM, Kassis JA, Zhang SO, Odenwald WF: castor eneodes
binding aetivity ofthe DNA-binding motif ofthe transcription faetor a novel zine fingerprotein required for the development of a subset of
ALCR in Aspergillus nidulans. Mol Gen Genet 242: 33-39, 1994 CNS neurons in Drosophila. Neuron 9: 789-803, 1992
24. Timmerman JE, Gniard B, Shechter E, Delsue MA, Lallemand JY, 45. Kuhnlein RP, Frommer G, Friedrieh M, Gonzalez-Gaitan M, Weber
Gervais M: The DNA-binding domain of the yeast Saccharomyces A, Wagner-Bemholz JF, Gehring WJ, lackle H, SehuhR: spalt eneodes
cervisiae CYPI(HAPI) transeription faetor possesses two zine ions an evolutionary eonserved zine finger protein of novel strueture whieh
whieh are complexed in a zine cluster. Fur J Biochem 225: 593-599, provides homeotie gene fimetion in the head and lai! region of Ihc
1994 Drosophila embryo. EMBO J 13: 168-179, 1994
48

46. Swiatek PJ, Gridely T: Perinatal lethality and defeets in hindbrain 53. Perrotti D, Melotti P, Skorski T, Casella I, Pesehle C, Calabretta B:
development in mice homozygous for a targeted mutation of the zine Overexpression of the zine finger protein MZI inhibits hematopoietie
finger gene Krox20. Genes Dev 7: 2071-2084, 1993 development from embryonie stem eells: Correlation with negative
47. Bernard 0, Ganiatisas S, Kannourakis G, Dringer R: Kizl, a protein regulation of CD34 and emye promoter aetivity. Mol Cell Biol 15:
with UM zine finger and kinase domains, is expressed mainly in 6075-6087,1995
neurons. Cell Growth Differ 5: 1159-1171, 1994 54. Krishnaraju K, Nguyen HQ, Liebermann DA, Hoffman B: The zine
48. Schutz B, Niessing J: Cloning and structure of a chicken zine finger finger transeription faetor Egr-l potentiates maerophage differentiation
eDNA: Restrieted expression in developing neural erest eells. Gene ofhematopoietie eells. Mol Cell Bio115: 5499-55-7,1995
148: 227-236, 1994 55. Hansen P, Riebessei M: The early development of Xenopus laevis.
49. Nagai T, Aruga J, Takada S, Gunther T, Sporle R, Schugart K, Springer-Verlag, Berlin, pp 1-18, 1991
Mikoshiba K: The expression of the mouse Ziel, Zie2 and Zic3 gene 56. Nomizu T, Falchuk KH, Vallee BL: Zine, iron, and eopper contents
suggests an essential role for Zie genes in body pattern formation. of Xenopus laevis ooeytes and embryos. Mol Reprod Dev 36: 419,
DevBiol182: 299--313,1997 1993
50. Kostieh WA, Sanes JR: Expression ofzjh-4, a new memberofthe zine 57. Falehnk KH, Montorzi M, Vallee BL: Zine uptake and distribution
finger-homeodomain family, in developing brain and muscles. Dev in Xenopus laevis ooeytes and embryos. Biochemistry 34: 16524,
Dyn 202: 145-152, 1995 1995
51. Mevel-Ninio M, Terraeol R, Kafatos FC: The ovo gene of Drosophila 58. Montorsi M, Fa1chuk KH, Vallee BL: Vitellogenin and !ipovitellin:
eneodes a zine finger protein required for female germ !ine develop- Zine proteins of Xenopus laevis ooeytes. Bioehemistry 34: 10851,
ment. EMBO J 10: 2259--2266,1991 1995
52. Redemann N, Gaul U, laekle H: Disruption of a putative Cys-zine 59. Jornvall H, Fa1chuk KH, Geraei G, Vallee BL: 1,IO-phenanthroline
interaction e!iminates the biologieal aetivity 01' the Krüppel finger and Xenopus laevis teratology. Bioehem Biophys Res Comm 200:
protein. Nature Lond 332: 90--92, 1988 1398, 1993
Mulecular and Cellular Biachemistry 188: 49--56, 1998.
© 1998 Kluwer Academic Publishers.

Maturation of fatty acid and carbohydrate


metabolism in the newborn heart
A-Olufemi Makinde, Paul F. Kantor and Gary D. Lopaschuk
Cardiovascular Research Group, Lipid and Lipoprotein Research Group, Departments of Pediatrics and Pharmacology,
F aculty oiMedicine, The University oiAlberta, Canada

Abstract
During fetallife, myocardial ATP is derived predominantly from glycolysis and lactate oxidation. Following birth, a rapid
maturational increase in fatty acid oxidation occurs along with a decline in glycolytic and lactate oxidative rates, thus changing
the major source of myocardial ATP production. This shift in energy substrate preference occurs in response to changes in the
circulating substrate content of newbom plasma with the onset of suckling, and is also due to alterations in circulating levels
of hormones, such as insulin and glucagon. Important changes in subcellular regulatory mechanisms of both fatty acid and
carbohydrate metabolism in the heart also characterize this response. This review deals with recent advances in the understanding
ofthese subcellular mechanisms which regulate this important shift in myocardial energy metabolism, with particular emphasis
on thc molccular cvcnts occurring in the heart during the transition from fetal to newbom life. (Mol Ccll Biochcm 188: 49-56,
1998)

Key words: 5' AMP activated protein kinase, acetyl CoA carboxylase, newbom, fatty acid oxidation, glycolysis, glucose oxidation,
fatty acid oxidation

Abbreviations: AMPK - 5' AMP activated protein kinase;ACC - acetyl CoA carboxylase; CPT I-camitine palmitoyltransferase
I; CPT 2 - carnitine palmitoyltransferase 2; LDH -lactate dehydrogenase; T3 - triiodothyronine; PFK - phosphofructokinase;
TCA - tricarboxylic acid; PD H - pyruvate dehydrogenase

Introduction and the greater efficiency of peripheral tissue oxygen


extraction from the circulation. This decrease in MV0 2 is due
The he art has a high demand for energy, which is supplied in part to unique features of fetal hemoglobin.
primarily from the metabolism of fatty acids and carbo- In contrast, the newbom heart is confronted with the need
hydrates. At birth, major changes occur in the fetal milieu to increase cardiac output to both the pulmonary and systemic
that require the heart to adapt to the newbom environment. circulation, to operate against an increased vascular re-
One change is a dramatic alteration in energy substrate sistance, and to adapt to variations in total metabolic rate, and
supply and the hormonal environment to which the newborn in the supply of it's vital energy substrate requirements.
heart is exposed. The heart quickly adapts to these changes In the fetal heart, lactate oxidation and glycolysis are the
to ensure a constant supply of energy to meet it's contractile preferred source for ATP production [I, 4-7]. However,
demands. The molecular and integrative changes that occur following birth fatty acid oxidation becomes the predominant
in the heart during the perinatal period are the subj ect of this source of ATP production for the heart [8-11]. As the new-
review. born matures, fatty acids remain the dominant oxidative
The fetal milieu is characterized by a low oxygen/substrate substrate in the heart and the major source of high energy
respiratory coefficient [see 1,2 and 3 for review], a euthermic phosphate production under normal aerobic conditions. The
and euglycemic environment. In this setting, the developing mechanisms responsible for this switch in energy substrate
heart has both the decreased circulatory workload of fetallife, preference are due to a combination of events, including

Address Jar ajJprints: G.D. Lopaschuk, 423 Heritage Medical Research Bldg., The University of Alberta, Edmonton, Alherta, Canada T6G 2S2
50

changes in energy substrate supply to the heart, changes in Table 2. Overview of Carbohydrate metabolism in the fetal, newbom and
hormonal control of energy metabolism in the heart, and mature myocardium
direct subcellular changes within the myocardium. Fetal Newbom Mature

Dominant metabolie Lactate Glucose/Glycogen Glucose


pathway
Energy substrate supply to the fetal and neonatal heart
% total contribution to ATP 40--<i0 40 10

Following conception, the premature heart starts to beat at Dominant glucose GLUT-l GLUT-l GLUT-4
approximately the third week of gestation in humans. The transporter GLUT-4
increasing work requirement of the fetal heart during develop- LDH isofonn LDHI LDHlILDH4 LDH4
ment would then appear to create the necessity for increased
Enzyme activity
encrgy substrate mctabolism to support not only the energy
(relative to the mature heart)
demands of the contractile machinery, but also the growth PFK high very high
requirement of the developing heart. Carbohydrates, especially LDH-M high high
glucose and lactate, are important energy substrates in the fetal PDH high ±
he art [I]. While circulating levels of glucose are not that TCAcyc1e low same
different between the fetus, newbom, and adult, the fetal heart
is in an environment highly enriched in lactate (Table I). it is yet unclear whether fetal heart really has a lower
Studies by Comline and Silver [12] have shown that the oxidative capacity than adults. A number of studies have
placenta is a highly enriched source oflactate during gestation. shown that the fetal heart is low in total mitochondrial content
In late gestation, plasma lactate levels in the fetus can be 10 and has both a decreased tricarboxylic acid cycle activity and
mmol/I or even higher (Table I). This compares to levels that electron transport chain enzymes [17-22]. This is consistent
are 1-2 mmol/I in the newbom and even less in the adult. with a heart that has a decreased oxygen availability and
Thcsc high Icvels of lactatc in the fetus are an important oxidative metabolism. However, Fisher et al. [6] showed that
source of energy for the fetal heart, resulting in over 60% of myocardial oxygen consumption in the fetal lamb is not
the hearts total oxygen consumption being accounted for by different from that observed in the adult lamb. This may be
lactate oxidation. due in part to the increased myocardial blood flow coupled
Opposite to what is seen with lactate levels, blood levels with an increased myocardial oxygen extraction efficiency
offatty acids are very low in the fetal circulation as compared in the fetallamb. In fact myocardial oxygen consumption per
to newbom or adult heart [13-16]. This contributes to a low unit weight of tissue is similar in both fetal and adult sheep
fatty acid use in the fetal heart. hearts. Moreover, studies in the fetal pig heart (about 90%
Another factor that may contribute to differences in ge station) showed that myocardial oxygen consumption is
myocardial energy substrate preference in the fetal hearts is about 23.5 !lmol·min-1.g dry weighr1. This is only slightly
a relative decrease in oxygen availability to the heart in the lower when compared to values of 32.4, Ilmol.min-l.g dry
fetallife compared to li fe ex utero. This favors the heart's weighr' obtained in the adult pig heart. High lactate oxidation
ability to utilize glycolysis as an energy source. However, rates in the fetal heart are also consistent with a substantial
although many studies have demonstrated that the arterial oxidative capacity in this setting.
oxygen content in fetallife is lower than that seen in the adult, In contrast to these studies, Rolph and Jones [7] found that
there exists a marked difference in oxygen consumption in the
Table J. Overview of changes in circulating substrate and honnone levels fetal and adult guinea pig hearts. Oxygen consumption rates
in the fetus, newbom and adult were as low as 1.5 Ilmol/min l/g dry weighr1 in fetal guinea
Parameter Fetal Newbom Adult pig heart, which is less than one half of values expected in the
adult. In view of the above, it would appear that the low
Substrate Levels
oxidative capacity attributed to the fetal heart may not be
Glucose (mM) -5 -6 -5 generalized to all species of experimental animals, and that
Lactate (mM) -10 -2 -0.5
given appropriate substrate, the fetal heart is indeed capable
Falty acids (mM) -0.02 -0.3 -0.4
of substantial oxygen consumption and substrate oxidation.
Honnonallevels As mentioned, fatty acids are not an important source of
Insulin ([lU/mI) -95 -10 -40 myocardial ATP production in the fetal heart. This is partly
Glucagon (ng/ml) -0.18 -1.0 -0.11 due to the fact that they are present in such low concentration
Triiodothyronine (ng/dl) -7 -18 -90 (Table I). Whereas the reasons for very low concentrations
Corticosterone (mg/dl) N.D -1 -5
of free fatty acid in the fetal circulation are not yet fully
'Values for the substrate levels are from ref. 14 and 15. elucidated, a transplacental free fatty acid gradient (such as
51

is present in humans) may be a contributing factor. Following conditions prevailing during fetal, newbom transitional and
birth, plasma free fatty acid concentrations increase dra- postnatallife.
matically to levels similar to that usually observed in the adult.
It is therefore not surprising that fatty acid oxidation rates Carbohydrate metabolism
increases dramatically in the early post-natal life (see Conceptually, the subcellular control mechanisms of carbo-
following seetions). hydrate metabolism can be grouped into those pertaining to
membrane transporters and signaling, cytoplasmic enzymes,
and mitochondrial enzymes:
Hormonal changes in the neonatal heart
Membrane transporters and signaling. Specific transporter
The importance ofhormones in influencing metabolism and mechanisms for the influx of glucose and lactate into the
metabolie rate cannot be overemphasized. During fetallife, myocyte exist forming a mechanism for the regulation of
circulating insulin levels are high and glucagon levels are substrate utilization. Glucose utilization by embryonie rat
quite low. Immediately following birth however, insulin hearts is of the order of 60 ng-l/min/mg-1, while that of the
levels decreases precipitously and glucagon levels rises (see entire embryo is on average only 13 ng-1/min/mg-l (attesting
Table 1). It is possible that the high insulin levels function to to the metabolie activity ofthe developing myocardium) [25].
repress high oxygen demanding metabolic process in favor Hyperglycemic conditions alone are able to increase the
of biosynthetic activities within the cell in fetal and early glucose metabolic index by 30-100 % in all tissues except
newbom period. Moreover, the high insulin levels in fetallife the brain. Increasing fetal plasma insulin will also result in a
may be expected, as the fetus nccds to bc able to cfficiently small increase in myocardial glucose uptake. This suggests
transport glucose from matern al circulation into its own that during fetallife, glucose uptake is regulated by a low
circulation to meet glucose use in a variety ofbiosynthetic affinity membrane transport mechanism. GLUT-I, the 'basal'
processes going on in the developing fetus. The alterations glucose transporter is regulated in its expression in the fetal
in the levels of many key substrates and enzyme activities in heart mainly by plasma glucose levels, while the levels of
the neonate coupled with changes in hormonal levels have insulin remain constant [26]. Glucose transporter molecules
very important consequences for the energy substrate utiliz- expressed on the sarcolemma switch from the GLUT-l to the
ation in the newbom heart. In parallel with this changes in GLUT-4 isoform shortly after birth during normal suckling
circulating hormones is a switch in energy substrate utiliz- [27]. At this point a greater degree of insulin sensitivity
ation for ATP production in the early newborn period. (higher glucose affinity) is demonstrable. Thus GLUT-4 may
Both the levels of triiodothyronine (T3) and corticosterone be considered the insulin responsive iso form of the glucose
increases significantly in the early newbom period (Table 1). transporters, and following birth it undergoes translocation
Plasma corticosterone decreases to very 10w levels within 24 to sarcolemmal sites by mechanisms not yet described. As a
h ofbirth. There is, however, a significant increase in the level result glucose transport becomes the rate-limiting step in
ofthis hormone in the immediate neonatal period [15, 23, 24], glucose utilization by the myocardium.
but then the levels begin to decrease again. Triiodothyronine Lactate influx and efflux from the myocardial cell is
(T3) has been shown to significantly increase basal metabolic controlled by the monocarboxylate co-transporter [28]. This
rates, and induce the expression of somc kcy cnzymcs rcccntly dcscribcd transporter also facilitatcs thc movcmcnt
involved in oxidative metabolism. The importance of increas- of alpha-ketoacids and pyruvate across cell membranes and
ed T3 and plasma corticosterone levels and its effect on mitochondrial membranes in many tissues. In cardiac myo-
energy substrate metabolism in the immediate newborn cytes it is a 45 kDa protein with some features distinguishing
period remains to be determined. it from the transporter found in other tissues and red cells.
There may be two populations of transporter isoforms in
cardiac tissue, but this along with the nature of any matur-
Changes in subcellular control of energy metabolism in ational changes in expression remains to be described. The
the newborn heart function of the co-transporter is dependent (as it's name
suggests) on a trans-membrane gradient ofH+. This feature
The maturing heart has a number of integrative control suggests that the fetal heart is able to exploit such a gradient
mechanisms with which to regulate it's use of energy sub- to it's advantage in the prolific uptake and aerobic oxidation
strate. Tt is thus responsive to the extemal milieu of oxygen of lactate during fetallife.
and substrate availability, to circulating hormonal regulators,
and also to it's own requirements for tissue growth and Cytosolic enzymes. Phosphofructokinase (PFK) is the rate
contractile performance. These mechanisms have been shown limiting enzyme of the glycolytic disposition of activated
to be versatile, and able to accommodate a wide variation in intracellular glucose. Existing in two isoforms, PFK 1 and
52

PFK2, this enzyme catalyzes the phosphorylation of fructose period probably oecurs due to the plentiful fatty acid supply
6-phosphate, to form a 1,6-diphosphate intermediary in the and low carbohydrate supply during the suckling period.
ca se of PFK j and a 2,6-diphosphate moiety in the case of
PFK2 • This 2,6 diphosphate residue serves to further activate Fatty acid metabolism
PFK j and therefore increase glycolytic flux. Typically PFK In fetallife, myoeardial fatty acid oxidation rates are low, and
activity is closely regulatcd in cardiac tissuc by cytosolicATP provide only a small portion of the myocardial energy
and citrate concentrations, preventing any excessive flux of requirements [8-11]. Low circulating fatty acid levels in the
glycolysis and uncoupling of the glycolytic/oxidative ratio fetal circulation are one important contributing factor to these
for glucose utilization. However, in fetallife there is evidence low fatty acid oxidation rates (see section on 'Energy
that PFK is less sensitive to these inhibitory regulators, but substrate supply to the fetal and neonatal heart'). However,
more sensitive to fructose 2,6, diphosphate activation, thus subcellular differences in the neonatal and mature heart also
potentially allowing for a rapid activation of glycolysis in appear to be a major contributing factor. This is evident
times ofhypoxic stress [29] . because despite a rapid rise in blood levels of fatty acids
When exogenous lactate enters the cytosol, it undergoes within hours ofbirth, the heart remains unable to oxidize fatty
reversible oxidation to form pyruvate prior to entry into the acids. For instance, if isolated hearts obtained from rabbits
TCA cycle. This reaction requires the supply ofNADH as an in the immediate newborn period are perfused with normal
electron donor, and is thus dependent on the cytosolic redox physiologicallevels of fatty acids, fatty acid oxidation rates
state . The reaction is catalyzed by lactate dehydrogenase are much lower than rates in hearts from 7 day old rabbits
(LDH) a tetrameric complex comprised of varying comb- perfused under identical conditions [9, 10]. Similar ob-
inations ofthe two constitutive monomers designated LDH- servations have also been made in newbom pig hearts [8, 11].
H (predominant in he art) and LDH-M (predominant in Although myocardial fatty acid oxidation rates are low in
skeletal museie ). The dominant isoform of LDH expressed the fetal and immediate newborn period, fatty acid oxidation
in any specific tissue is then determined by the ratio ofLDH- rates increase shortly after birth. Within days of birth, the
Hto LDH-M. newbom rabbit heart rapidly acquires the ability to oxidize
Early descriptions of the newborn maturation of LDH fatty acids, such that fatty acids quickly becomes the major
expression in guinea pig myocardium from the 1970's energy substrate ofthe heart. As shown in Fig. 1, fatty acid

-
suggested that the LDH-HlLDH-M ratio was low in the fetal
heart, and began to rise in the first month after birth [30].
Overall myocardial LDH activity has also been shown to be E:I Glycolysis
higher in mature than in fetal sheep LV, while in long term D Glucose Oxidation
prenatal hypoxemia LDH activity increases further. Whether Lactate Oxidation
Palmitate Oxidation
this is associated with an increase of oxidative metabolism Z IS
of lactate or of glycolytic production of lactate is however 0
unclear [31]. ~
U
--: 12
·s=
~
Q
Mitochondrial enzymes. Pyruvate decarboxylation is the key 0
=:
..
irreversible step in carbohydrate oxidation and is catalyzed
by PDH, a multi-enzyme complex located within the mito-
~
' 9
chondrial inner membrane. Increased concentrations of TCA
cycle intermediaries thus reduce the flux of pyruvate into the
~
-<
."
..
~
~

mitochondrion. PDH activity is regulated by a specific kinase ~ ~ 6


-<
E-<
mediated inactivation process, which is responsive to acetyl CI)
Q
CoA/CoA and NADH/NAD+ ratios, and a phosphatase ;.. E
CI
::L 3
mediated acti vation process which responds to mitochondrial
[Ca 2+] and [Mg 2+]. Overall, the rate of pyruvate oxidation
-<
~
E-<
correlates weil with the degree ofphosphorylation ofPDH. CI)
0
In fetal rabbit myocardium PDH eomplex activity is 1 Day 14Day
maximal towards the end of gestation, and is 3--8 fold higher
in heart than in lung or livcr tissuc from corresponding Fig. J. Profile of energy substrate preference in the newborn heart. Values
for glyeolysis. glucose oxidation, lactate oxidation and palmitate oxidation
animals [28]. In the early newborn period, flux through PDH
in 1 dayand 14 day old hearts were taken from referenees 9 and 10. Isolated
is low (based on glucose oxidation measurements) [9, 10], working hearts were perfused wilh physiologieallevels of glucose, lactate
and probably does not mature until following weaning. This and fatty acids, as deseribed in referenee 9. ATP produetion rates were
decrease in carbohydrate oxidation in the early newborn ealculated as described in referenee 10.
53

oxidation provides only a small portion of the hearts energy myocytes. Using cultured neonatal rat myocytes, these
requirements in 1 day old rabbits. However, in 14 day old investigators showed that CPT 1 iso form expression can be
rabbits, fatty acid is the predominant source of energy. This changed in response to electrical stimulation. Electrical
increase in fatty acid oxidation rates ensures that the high fatty stimulation results in an increase mRNA for the 82 kDa CPT
acid intake of the suckling newborn are utilized as an energy 1 isoform.
substrate by the heart. While CPT I isoform expression can change in the new-
The rapid increase in the ability of the heart to oxidize fatty born period, the significance of this to the maturation of fatty
acid oxidation despite similar extracellular fatty acid concent- acid oxidation following birth is not completely understood.
rations, demonstrates that subcellular control of fatty acid For instance, increased expression of the 82 kDa isoform
oxidation is altered in the newborn heart. We have shown that should·make the mitochondrial uptake offatty acids more
one major alteration in control offatty acid uptake occurs at sensitive to inhibition by malonyl CoA. This scenario would
the level of the mitochondrial uptake of fatty acids [32]. In not be expected in a maturing heart in which fatty acid uptake
order to be metabolized, fatty acids that are taken up by the by the mitochondria increases. Moreover, the role of lower
heart and activated in the cytoplasm to fatty acyl CoA must L-carnitine levels in the newborn heart, especially on the
first be converted to long chain acy1carnitine by carnitine rapid maturation of fatty acid oxidation, remains unclear,
palmitoyltransferase 1 (CPT I). This enzyme, which is since an increase in the proportion ofthe 88 kDa isoform of
located on the inner surface of the outer mitochondrial CPT I would be expected (this isoform is less sensitive to
membrane, is the rate limiting enzyme for mitochondrial fatty malonyl CoA inhibition). Further studies are necessary to
acid uptake [33]. Long chain acy1carnitine produced by CPT clarify the relationship between fatty acid oxidation, CPT I
I is transported into the mitochondria via a carnitine isoform expression, malonyl CoA levels, and L-carnitine
acyltranslocase, where it is acted upon by carnitine palmitoyl- levels in the newborn heart.
transferase 2 (CPT 2) to produce fatty acyl CoA, the first While the role of CPT I isoform expression in regulating
substrate for fatty acid ß-oxidation. An important consider- the increase in fatty acid oxidation post-birth is not clear,
ation in this uptake process is that CPT 1 is very sensitive recent studies have shown that a dramatic decrease in malonyl
to malonyl CoA inhibition. CoA levels in the newborn period are important in the
increase in fatty acid oxidation in the newborn period. We
Malonyl CoA and CPT 1 in the neonatal heart have recently shown in newborn rabbit hearts that myocardial
In both the adult and newborn heart, CPT 1 is inhibited by levels ofmalonyl CoA decrease dramatically in the newborn
nanomolar concentrations ofmalonyl CoA [32]. Malonyl period, and that this drop in malonyl CoA is accompanied by
CoA is the product of the enzymatic carboxylation of acetyl an increase in fatty acid oxidation rates [33]. This decrease
CoA. This carboxylation of acetyl CoA is catalyzed by a in malonyl CoA levels is due to a decrease in ACC activity
cytoplasmic enzyme, acetyl CoA carboxylase (ACC) [33, in the newborn period.
34]. It has now been established that malonyl CoA production
by ACC is an important regulatory mechanisms controlling Acetyl CoA carboxylase activity in the newborn heart
the entry of long chain fatty acyl CoA. An overview des- In I day old rabbits, ACC activity in the heart is very high,
cribing this proposed pathway is shown in Fig. I. resulting in very high levels of malonyl CoA [33, 37]. We
In addition to malonyl CoA, L-carnitine has an important propose that these high levels of malonyl CoA are responsible
role in facilitating mitochondrial fatty acid uptake. In the for the low rates of fatty acid oxidation in the newborn heart,
newborn, L-carnitine levels are quite low, and increase as the secondary to an inhibition ofCPT I (see Fig. 2). Ifhearts are
animal matures. McGarry et al. recently demonstrated a CPT perfused in the absence of insulin or in the presence of
I isoform switch in the newborn mitochondria. In the first glucagon, ACC activity decreases, malonyl CoA levels
few days oflife, a 82 kDa isoform with a low Km for carnitine decrease, and fatty acid oxidation rates increase [33, 37, 38].
predominates, and accounts for almost 70% of total CPT I In 7 day old rabbits, which have high rates of fatty acid
activity with the remainder arising from a 88 kDa CPT 1 oxidation bothACC activity and malonyl CoA levels are very
isoform [35]. This 88 kDa isoform has a higher affinity for low, confirming that a decrease in malonyl CoA production
L-carnitine, but is less sensitive to inhibition by malonyl CoA. by ACC is an important mechanism responsible for the
As a result, because of the CPT I isoform expressed in the increase in fatty acid oxidation following birth.
immediate newborn period, flux through CPT 1 is not only Dur recent studies have suggested that the decrease inACC
dependent on the altered isoform expression, but also on the activity observed following birth can be explained by a
cytoplasmic concentration of both malonyl CoA and L- phosphorylation and inhibition ofACC. The newborn rabbit
carnitine. heart expresses both a 265 kDa and a 280 kDa iso form ofthe
Recent studies from McMillinet al. [36] also demonstrates ACC in almost equal proportion [33]. Both isoforms ofACC
that CPT I isoform expression can change in neonatal can be phosphorylated and inhibited by a novel AMP-
54

CYTOSOL MITOCHONDRIAL
MATRIX

:
TRIACYLGLYCEROL

FATTY ACIDS _ FATTY ACYL COA

MALONYL COA ---=--I~~


e j carnitine palmitoyl-
transferase 1

I
carnitine
acyltrans/ocase
FATTY ACYLCARNITINE FATTY ACYLCARNITINE

acetyl CoA
CARNITINE ~

ß.,.-j
carboxylase

e/
FATTY ACYL COA

GTE
co,
S' AMP activated
protein kinase

PYRUVATE --+--+ ACETYL COA--


ACETYL COA
pyruvate
dehydrogenase
complex

Fig. 2. AMPK and ACC involvement in the regulation of fatty acid oxidation. Camitine palmitoyltransferase (CPT) I is the rate limiting enzyme in the
mitochondrial uptake offatty acids. Cardiac CPT I is very sensitive to inhibition by malonyl CoA. This malonyl CoA is produced in the cytoplasm by acetyl
CoA carboxylase (ACC). ACC can be phosphorylated and inhibited by AMP-activated protein kinase (AMPK) which is very active in heart musele.
Through this pathway, an increase in AMPK activity can result in an increase in fatty acid oxidation rates.

activated protein kinase (AMPK) which is highly expressed that an endogenous inhibitor of the AMPK activity may be
and active in both the adult and newbom heart [37--40]. degraded, thus facilitating the activity of the activated
enzyme. Ihis possibility has yet to be established.
AMP activated protein kinase in the newborn heart
We have recently shown thatAMPK activity increases in the Insulin and glucagon effects on AMPK, ACC andfatty
neonatal rabbit heart and that this increase inAMPK activity acid oxidation in the newborn heart
can explain the decrease in ACC activity and malonyl CoA While the molecular mechanisms responsible for the increase
levels seen in the newbom heart [37]. Furthermore, AMPK in AMPK activity in the newbom period are yet to be fully
protein expression and activity increase significantly in the elucidated, it is clear that hormonal change following birth
rabbit heart between 1 day and 7 days following birth. The contribute to changes inAMPK activity. We recently showed
potential significance of this is that an increase in AMPK that high levels of insulin inhibitAMPK activity, suggesting
protein level and activity in vivo may result in an increase that a decrease in insulin levels in vivo has the potential to
phosphorylation and inactivation of ACC in the newbom stimulate the observed AMPK activity. Moreover, AMPK
myocardium (see Fig. 2). Ihis decreased ACC activity would activity can be stimulated pharmacologically with 5-amino-
result in lower levels of malonyl CoA, and therefore increased 4-imidazole carboxamide ribotide in 7 day old rabbit hearts
CPT 1 activity. Ihis could then partly account for the increase [42]. Ihis results in a decrease inACC activity and an increase
in fatty acid oxidation rates seen in the newbom period. in fatty acid oxidation rates. It thereforc appears thatAMPK
At present, the molecular mechanism responsible for in the newbom heart not only regulates fatty acid oxidation,
increased AMPK activity is still unknown. One possibility but that the AMPK activity itself is under hormonal contro!.
is that a recently characterized AMPK kinase is responsible Figure 3 describes a proposed pathway by which changes in
for phosphorylation and activation of AMPK. While an hormonal levels may regulateAMPK activity in the immed-
AMPKkinase has been characterized in liver [41], it has yet iate newbom period. We believe that a decrease in insulin
to be determined ifthis enzyme is active in the heart.Another levels following birth is accompanied by an increase in
possible mechanism by whichAMPK activity is increased is AMPK activity (i.e. insulin inhibition ofAMPK is relieved).
55

I Insulin levels
t following birth
Glucagon levels
following birth
t cagon levels post-birth is associated with increases in
myocardial cAMP levels. This would be expected to activate
protein kinase A activity. Protein kinase A is capable of

~
phosphorylating and inhibiting both isoforms of cardiacACC

Y
[43]. In intact heart we have shown that glucagon perfusion

t of ncwbom hcarts will also inhibitACC activity and increasc

t Protein fatty acid oxidation. We therefore propose that both the


AMPK
Activity KinaseA decrease in insulin and increase in glucagon levels act in
Activity concert 10 inhibil ACC activity in the newbom hearts (Fig.
3). This proposed pathway provides an attractive mechanism
as to how acule hormonal changes in the immediate newbom
period can increase falty acid oxidation in the newbom heart.

~ ACC Activity Conclusion


1
ACC Activity
(Iow) (high)
In the fetal heart, lactate oxidation is the predominant source
of energy. Immediately following birth, glycolysis and
I Malonyl CoA glucose oxidation becomes the preferred substrate for ATP
+ Levels production. However, within the first week of Ii fe the
newbom heart rapidly switches energy substrate preference

~
to fatty acids. This increase is due to an increase in AMPK
activity, resulting in an increased phosphorylation and
CPT 1 Activity1 CPT 1 Activity
inhibition of ACe. The result is a decrease in malonyl CoA
levels thus relieving its inhibitory effect on CPT I enzyme
(Iow) (high)
system. This condition would be expected to increase trans-

t Fatty Acid
Oxidation
location of activated fatty acid into the mitochondria with
subsequent increase in ß-oxidation of fatty acids.

Fig. 3. Hypothetieal seheme as to how a deerease in plasma levels of insulin


and a rise in glucagon levels result in an inerease in fatty acid oxidation in
Acknowledgements
the newborn heart. Following birth. insulin levels in the blood deerease
while glucagon levels inerease. The decrease in insulin levels results in an This study was supported by a grant from the Heart and
aetivation of AMPK in the heart. AMPK then phosphorylates and inhibits Stroke Foundation of Alberta. GDL is a Medical Research
ACC aetivity. The inerease in glucagon levels inereases cAMP levels. This Council ofCanada Scientist and anAlberta Heritage Found-
results in an inerease in protein kinase A activity, which Ihen phosphoryJates
and inhibits ACC aelivity. Therefore, Ihe deerease in insulin and inerease
ation for Medical Research Senior Scholar. AOM is a grad-
in glucagon resull in a decrease in ACC and malonyl CoA levels. This uate student trainee of the Alberta Heritage Foundation for
relieves inhibition of CPT I, resulting in CPT I being in the more aetive Medical Research and the Heart and Stroke Foundation of
form. As a resul! of this inerease in CPT 1 aetivity, fatly acid oxidation Canada. PFK is a post-doctoral fellow oftheAlberta Heritage
rates inerease. This allows the heart to switch to the use of fatty acids as a Foundation for Medical Research and the Heart and Stroke
source of energy in the newborn period.
Foundation of Canada.

This results in an increased phosphorylation and inhibition


of ACC activity. The ultimate consequence is a decrease in References
myocardial malonyl CoA levels. The resulting increase in
CPT I activity is then responsible for the increase in fatty acid I. Fisher DJ: Oxygenation and metabolism in the developing heart. Semin
oxidation seen in these hearts. Prenatal8: 217-225,1984
Not only do circulating levels ofinsulin decreases follow- 2. Rolph TP, Jones CT: Metabolism during fetal life: A functional
assessment ofmetabolie development. Physiol Rev 65: 357-430, 1985
ing birth, there is also an increase in circulating glucagon
3. Lopaschuk GD, Collins-Nakai RL, ltoi T: Developmental changes in
levels (Table 1). These changes in glucagon do not appear to energy substrate use in the heart. Cardiovasc Res 26: 1172-1180, 1992
directly alter AMPK activity (A.-O. Makinde and G.D. 4. Wemer JC. Sieard, RE: Lactate metabolism of isolated perfused felal
Lopaschuk, unpublished data). However, increases in glu- and newbom pig hearts. Pediatr Res 22: 552-556, 1987
56
5. Fisher DJ, Heymann MA, Rudolph AM: Myoeardial oxygen and and hyperinsulinemia inerease glucose utilization in fetal rat tissue.
earbohydrate eonsumption in fetallambs in utero and in adult sheep. Am J Physiol260: E588--E593, 1991
Am J Physiol238 (Heart Cire Physiol 7): H399-H405, 1980 26. Sehroeder RE, Doma Medina CL, Das UG, Sivitz WI, Devaskar SU:
6. Fisher DJ, Heymann MA, Rudolph AM: Myoeardial eonsumption of Effeet of matemal diabetes on fetal rat myoeardial and skeletal museIe
oxygen and earbohydrates in newbom sheep. Pediatr Res 15: 843- glucose transporters. Pediatr Res 41: 11-19, 1997
846, 1981 27. Postie C, Leturque A, Prinz RL, Maulard P, Loizeau M, Granner DK,
7. Rolph TP, Jones CT: Regulation of glyeolytie flux in the heart ofthe Girard J: Development and regulation of glucose transporter and
fetal guinea pig. J Dev Physiol5: 31-49, 1983 hexokinase expression in rat. Am J Physiol 266: E548--E559, 1994
8. Wemer JC, Sieard RE, Sehuler HG: Palmitate oxidation by isolated 28. Poole RC, Haiestrap AP: Transport of lactate and other mono-
working fetal newbom pig hearts. Am J Physiol256 (Endoerine Metab earboxylates across mammalian plasma membranes.Am J PhysioI264:
19): E315--E321, 1989 C761-{;782, 1993
9. Itoi T, Lopasehuk GD: Calcium improves meehanieal funetion and 29. Bristow J, Bie DM, Langer LG: Regulation of adult and fetal
earbohydrate metabolism following isehemia in isolated bi-ventrieular myoeardial phosphofruetokinase. J Biol Chem 262: 2172-2175, 1987
working hearts from immature rabbits. J Mol Cell Cardiol28: 1501- 30. Bamie SE, Harris P: Myoeardial enzyme aetivities in guinea pigs
1514, 1996 during development. Am J Physiol233 (6) H707-H7I0, 1977
10. Lopasehuk GD, Spafford MA, Marsh DR: G1yeolysis in predominant 31. Ohtsuka T, Gilbert RD: Cardiae enzyme aetivities in fetal and adult
souree of myoeardial ATP produetion immediately after birth. Am J pregnant and non pregnant sheep exposed to high-altitude isehemia. J
Physiol261: HI698--HI705, 1991 Appl Physiol 79 (4): 1286-1289, 1995
11. Aseuitto RJ, Ross-Aseuitto NT, Chen V, Downing SE: Ventrieular 32 MeGarry 10, Foster DW: Regulation of hepatie fatty acid oxidation
funetion and fatty acid metabolism in neonatal piglet heart. Am J and ketone body produetion. Annu Rev Bioehem 49: 395-420, 1980
Physiol256: H9-HI5, 1989 33. Lopasehuk GD, Witters LA, Itoi T, Barr R, Barr A: Acetyl CoA
12. Comline RS, Silver M: Some aspeets of fetal and uteroplaeental earboxylase involvement in the rapid maturation of fatty acid oxidation
metabolism in eows with indwelling umbilieal and uterine vaseular in the newbom rabbit heart. J Biol Chem 269: 25871-25878, 1994
eatheters. J Physiol (Lond) 260: 571-1l6, 1976 34. Saddik M, Gamble J, Willers LA, Lopasehuk GD: Aeetyl-CoA
13. Phelps RL, Metzger BE, Freinkel N: Carbohydrate metabolism in earboxylase regulation offatty acid oxidation in the heart. J Biol Chem
pregnaney. XVII. Diurnal profiles of plasma glucose, insulin, free fatty 268: 25836-25845, 1993
acids, triglycerides, eholesterol and individual acids in the late normal 35. Brown NF, Weis BC, Husti JE, Foster DW, MeGarry JD: Mitoehondria
pregnaney. Am J Obstet Gyneeol140: 730--736, 1969 eamitine palmitoyltransferase I isoform switehing in the developing
14. Medina JM: The role of lactate as an energy substrate for the brain rat heart. J Biol Chem 270 (15) 8952-1l957, 1995
during the early neonatal period. Biol Neonate 48: 237-244, 1985 36. Xia Y, Buja M, MeMillin JB: Change in expression ofheart earnitine
15. Girard J, Ferre P, Pegorier JP, Duee PH: Adaptations of glucose and palmitoyltransferase I isoforms with eleetrieal stimulation of eultured
fatty acid metabolism during perinatal period and suekling-weaning rat neonatal eardiae myoeytes. J Biol Chem 271: 12082-12087, 1996
transition. Physiol Rev 72: 507-562, 1992 37. Makinde AO, Gamble J, Lopasehuk GD: Upregulation of 5'AMP
16. Knopp RH, Warth MR, Charles 0, et al. Lipoprotein metabolism in activated protein kinase is responsible for the inerease in myoeardial
pregnaney, fat transport to the fetus and e!feets of diabetes. Biol fatty acid oxidation rates following birth in the newbom rabbit. Cire
Neonate 50: 297-317, 1986 Res 80 (4) 482-489,1997
17. Warshaw JB, Terry ML: Cellular energy metabolism during fetal 38. Kudo N, Barr AJ, Barr RL, Desai S, Lopasehuk GD: High rates of
development. II Fatty acid oxidation by the developing heart. J Cell fatty acid oxidation following reperfusion of isehemie hearts are
Bio144: 354-360, 1970 assoeiated with a deerease in malonyl CoA levels due to an inerease
18. Warshaw JB: Cellular energy metabolism during fetal development. in 5' AMP-activated protein kinase inhibition of acetyl CoA earboxy-
N. Fatty acid aetivation, acyl tranferase and fatty acid oxidation during lase, J Biol Chern 270: 17513-17520, 1995
development ofthe ehiek and rat. Dev Bio128: 537-544, 1972 39. Beri RK, Marley AK, See CG, Sopwith WF, Aguan K, Carling D,
19. Goodwin CW, Mela L, Deutsch C, Forster RE, Miller LD, Delivoria- Seoll J, Carey F: Moleeular cloning, expression and chromosomal
Papadopoulous M: Development and adaptation ofheart mitoehondria loealisation ofhuman AMP-aetivated protein kinase. FEBS Lett 356:
respiratory ehain funetion in fetus and newbom. Adv Exp Biol 75: 117-121, 1994
713-719, 1976 40. Aguan K, Seoll J, See CG, Sarkar NH: Charaeterization and ehrom-
20. Dallman PR, Schwart HC: Cytoehrome eoneentration during rat and osomalloealization of the human homologue of a rat AMP-aetivated
guinea-pig development. Pediatries 33: 106-110, 1964 protein kinase-eneoding gene: A major regulator oflipid metabolism
21. Glatz JFC, Veerkarnp lli: Postnatal development of paimitate oxidation inmammals. Gene 149: 345--350,1994
and mitochondrial enzyme aetivities in rat eardiae and skeletal museies. 41. Hawley SA, Davidson M, Woods A, Davies SP, Beri RJ, Carling D,
BioehimBiphysAeta 711: 327-335,1982 Hardie DG: Charaeterization of the AMP aetivated protein kinase
22. Wemer JC, Whitman V, Musselman J, Sehuler HG: Perinatal ehanges kinase from rat liver and identifieation of threonine 172 as the major
in mitochondrial respiration ofthe rabbit heart. Biol Neonate 42: 208-- site at whieh it phosphorylates AMP aetivated protein kinase. J Biol
216, 1982 Chem 271: 27879-27887, 1996
23. Fisher DA, Dussault JH, Sack J, Chopra J: Ontogenesis of hypot- 42. Makinde AO, Lopasehuk GD: Stimulating 5'AMP aetivated protein
thalamie-pituitary-thyroid funetion and metabolism in man, sheep and kinase inereases fatty acid oxidation in newbom rabbit hearts. J Mol
rat. Reeent Prog Horm Res 33: 59-107, 1977 Cell Cardiol28 (6): A179, 1996
24. BassettJM,Alexander G: Insulin, growth hormone and eortieosteroids 43. Haystead TAJ, Moore F, Cohen P, Hardie G: Roles of the AMP-
in neonatallambs. Normal eoneentrations and the e!feet of cold. Biol aetivated and eyelie-AMP-dependent protein kineses in the adrenaline-
Neonate 17: 112-125, 1971 indueed inaetivation of aeetyl-CoA earboxylase in rat adipoeytes. Eur
25. LeturqueA, ReveIli JP, Haugwel S, Kande J, GirardJ: Hyperglyeemia J Bioehem 187: 199-205, 1990
Moleeular and Cellular Biochemistry 188: 57---{)2, 1998.
© 1998 Kluwer Academic Publishers.

Genes regulating copper metabolism


Edward D. Harris,1,2 Yongchang Qian 1 and M.C.M. Reddyl
IDepartment of Biochemistry and Biophysics and the 2Faculty ofNutrition, Texas A&M University, College Station Texas,
USA

Abstract
The metabolism of Cu is intimately linked with its nutrition. From gut to enzymes, Cu bioavailability to key enzymes and other
components operates through a complex mechanism that uses transport proteins as weil as small molecular weight ligands.
Steps in Cu transport through the blood, absorption by cells, and incorporation into enzymes are slowly being understood.
Cloning and sequencing of the genes for Menkes disease and Wilson disease has shown that membrane-bound enzymes analogous
to Cu-ATPases in prokaryotes are equally important to Cu transport and homeostasis in mammalian cells. The primary structurc
ofthe mammalian Cu-ATPases has been deduced from cDNAs from tissues and organs. It now appears that mammalian Cu-
ATPase have tissue and developmental specificity. In this review, we will focus on the Cu-ATPase that has been identified with
Menkes disease. An emphasis will be placed on the existence of multiple forms of the ATPase and some indication as to how
the different isoforms befit their role in the normal physiology of copper, specifically transmembrane transport and maintenance
of a favorable internal cellular environment. (Mol Cell Biochem 188: 57-62, 1998)

Key words: copper transport, Menkes disease, Wilson disease, copper absorption

Introduction protein (ATP7A) and the Wilson protein (ATP7B) ATPase


have paralleIs to the heavy metal binding proteins in bacteria
Menkes disease is an X-linked disorder of infants resulting and show a 57% sequence homology to one another [9]. The
from a failure to pass Cu ions completely across the intestinal Menkes gene mRNA has a single open reading frame that
mucosa. The entrapment of Cu within intestinal cells leads codes for a protein of exactly 1,500 amino acids. Strong
to lcss Cu being transported to peripheral organs and tissues expression ofATP7A mRNA is seen in muscle, kidney, lung,
[1,2] and the gross symptoms ofthe disease resemble a severe brain; placenta and pancreas are weaker, and liver shows
Cu deficiency. In contrast, Wilson disease or hepatolenticular only a traces [3,4,10]. The Wilson transcript is 7.5 kb and
degeneration, results from a pathological accumulation ofCu encodes a protein of 1411 amino acids. In contrast to the
primarily in the liver and brain. Wilson disease, an autosomal Menkes gene, the Wilson gene is strongly expressed in liver
recessive disorder, affects males and females generally at and kidney [5, 11].
puberty or beyond. The major defect lies in a failure to release A cartoon of the Menkes protein is shown in Fig. 1. Its
liver Cu into the bile as well as give bidirectional movement essential structural features include a heavy metal binding
to its transport in brain. Both diseases have provided valuable domain (Hmb) of about 650 residues with 6 metal-binding
insights into the genetic factors that regulate Cu transport in cysteine residues in tandem as part of the structural motif
mammalian organs.As a first approximation, one can assurne GMT/HCXSC. The bacterial protein CopA, a P-type Cu-
that the respective genes for Menkes and Wilson diseases ATPase, contains only one Cu binding site [12] and other
encode similar biochemical components. Within certain bacteria as weil as yeast have two metal sequestering CXXC
boundaries, this interpretation has proven to be true. With the motifs, leading one to speculate whether Cu binding is the
isolation and sequencing ofthe genes for Menkes [3,4] and only function associated with this region oftheATPase. The
Wilson [5-8] has come the realization that both encode core ofthe ATPase has eight transmembrane (Tm) regions
membrane-bound P-type ATPases. Moreover, the Menkes that anchor and define the channel through which Cu ions

Addressfor offprints: E.D. Harris, Biochemistry and Biophysics, Texas A&M University, College Station, TX 77843-2128, USA
58

1 2 3 4 5

ATPase Domain

Heavy Metal Binding


650 residues

Fig. J. Schematic diagram ofthe Menkes P-type ATPase. The basic stmcture consists of a heavy metal binding region for binding Cu ions, 8 transmembrane
domains, and 2 cytoplasrnic loops. Shown are key residues in the ATPase domain that eomprise the catalytic uni!.

pass. A small flexible loop between Tm domains 4 and 5, The Menkes ATPase and Cu transport
and a much larger one between Tm 6 and 7 extend into the
cytosol and encompass the major catalytic and transport The discovery ofthe gene for Menkes or ATP7A was followed
motif' s of the protein. Included in the larger loop is the by Northem analysis to identify the organs and tissues that
DKTG sequence (Asp-Lys-Thr-Gly) with the critical aspartate express ATP7A mRNA. Those data are summarized in Table
(D) residue that is phosphorylated transiently during I where + signs designate relative staining intensity. Strong
transport. This sequence is a signature for all P-type expression is seen in lung, brain, musc1e, spleen and testes.
ATPases and was instrumental in identifying the Menkes Weaker amounts are shown in the kidney and heart tissue. Liver
ATP7 Aas a member of that c1ass. Adjacent to the region is is the only major organ that shows no expression. The lack of
an SEPHL motifthat serves to regulate cation flow through ATP7A mRNAin liver is of interest for two reasons. First, liver
the channel and a TGDN that is believed to be the site where strongly expresses the Wilson gene (ATP7AB) and a homo-
ATP binds. A CysPro-Cys motif in Tm 6 (not shown in the logue ofthis gene in yeast, CCC2, appears to be essential for
figure ) is believed to be part of the ion channel through the incorporating Cu into FET3, a copper oxidase in yeast that
membrane [3]. X -linked Menkes disease has its counter part plays a major role in iron transport [20]. Secondly, we have
in several murine models, the brindled mouse and the found that HepG2 cells, a human liver hepatoma celliine,
macular mouse being the most popular [13, 14]. The express a modified transcript ofthe Menkes gene (see below).
macular mutant and heterozygotes have very high Cu in the Table 2 summarizes our findings ofATP7A expression in eight
kidney and small intestine while maintaining lower amounts
in the liver [15]. This mutant is also impaired in transporting
Table j. Expression of ATP7 A mRNA in Organs and Tissues
Cu from the cytosol to organelles such as mitochondria,
nuclei, and lysosomes [16]. Reduced Cu accumulation in Tissue or Organ
CHO cells [17] and enhanced Cu accumulation in rat C6
Hear! +
glioma cells [18] have been linked to the Menkes gene and Brain ++
gene product. Enhanced accumulation of Cu is seen in C6 Plaeenta +
cells when the Cu-ATPase is blocked by sulfhydryl reagents Lung +++
[18] whereas over expression of the ATPase allows CHO Liver Not detected
Muscle ++
cells to tolerate toxic levels [17]. The mutation in both the
Kidney +
brindled and macular mouse impairs Cu transport across the Pancreas +
intestine resulting in less Cu being absorbed into the system Spleen ++
[ 19]. Testes ++
59

Table 2. Detection of ATP7A mRNA in Different Cell Lines overlook that protecting against toxic levels of Cu mayaiso
be a function ofthisATPase. Thus, cells require a Cu-ATPase
CRITERIA
Cells PCR Sequence Western Efflux
not only to retain manageable levels of Cu but also to respond
to potential toxicity. The latter point has been emphasized
Caco-2 + + + ND
by aseries of experiments performed by Camakaris and
BeWo (uninduced) Not detected
BeWo (induced) + + + +
colleagues [17]. In searching for causes of superior Cu
Human Fibroblasts + + ND + tolerance in certain strains of CHO cells, these investigators
RatC6glioma + + ND + were able to correlate resistance with the expression of the
Rat PCl2 + + ND + Menkes gene. Cells more prone to succumb to toxicity
Rat IEC6 cells + + ND ND weakly expressed the ATP7 A gene product. Since Menkes
Mouse capillarily + + ND ND
cells phenotypically are more sensitive to medium Cu than
ND - not determined nonnal cells [23], the data further support the correlation
between release of Cu and ATP7A function. In essence, the
data support a defect in the efflux system.
celilines. RT-PCR analysis and sequence data have con-
finned the presence of the ATP7A transcript. lt is important
to note that in BeWo cells, a placental choriocarcinoma cell Evidence lor alternative lorms 01 the Menkes gene
line, there is no evidence ofATP7A expression in cells that
are grown on plastic (petri dish) surfaces [21]. Expression The Menkes gene has been localized to Xq13.3 [10], isolated
is seen, however, when the cells are grown on microporous and cloned by positional cloning [4] and sequenced [3]. The
filter surfaces or treated with sodium butyrate [22]. In two Menkes transcript as detected by Northern analysis is 8.3-
celilines, viz., Caco-2 and BeWo, it has been possible to 8.5 kb and codes for a membrane bound Cu-transporting
obtain evidence for ATP7A protein using Western analysis. ATPase classified as a P-typeATPase of exactly 1500 amino
As ofyet, no data has been obtained by direct sequencing of acid residues [3] as determined from cDNA sequence data.
the protein. The Menkes transcript has been observed primarily by RT-
The genetic evidence backed up by experimental observa- PCR analysis in numerous cell types, including fibroblasts,
tions suggests that a functionalATP7 Ais required to maintain epithelial cells ofthe intestine, placenta, and brain (Table 2).
cellular Cu homeostasis in cells. Homeostasis is the outcome In our studies we have observed what appear to be alternative
of two competing transport processes, influx and effux. forms of the Menkes transcript. The forms differ by having
Which process is more keenly affected in Menkes disease is base insertions at the 5' end or skipped exons within the
ofparamount importance in deciding the function ofATP7 A. coding region, suggesting the forms represent alternative
In our studies with fibroblasts we found that the influx of Cu spliced transcripts. Importantly, the alternate forms appear to
as measured by uptake of 67 CuC12 was basically unimpeded be cell-type specific. Using primers designed to generate a
in Menkes cells as compared to normal fibroblasts. In fact, the 5.8 kb fragment, we attempted to amplify by RT-PCR a
Menkes fibroblasts appeared to take up Cu more aggressively transcript containing the complete open reading from BeWo
than normal fibroblasts [21]. The apparentenhanced efficiency cells, a human choriocarcinoma placental cell line. In addition
in Cu uptake, however, was later shown to be due to Cu to finding a cDNA of 5.8 kb, there was also evidence for at
accumulation which in turn resulted from a failure to release least four other transcripts, one a 1.9 kb fragment (Fig. 2).
the absorbed Cu. This point was made clearer when we When the amplified products were subcloned into pCR3. 1
examined the effux of 67CU from normal and Menkes vectors and sequenced, the 1.9 band revealed an unusual
fibroblasts. Once the Cu ions had been absorbed, more than fusion between exons 2 and 16. This fragment was missing
90% ofthe Cu was retained by the Menkes cells as compared exons 3-15 but otherwise was in frame with exons 16-23 at
to about 60% for normal cells. Such retention has been the 3' end. A second clone, however, upon sequencing
interpreted to reflect an avid internal sequestering of Cu by revealed a 45 bp insert in the 5' region upstream from the
metallothionein, a heavy metal binding protein in the cytosol published ATG start site. The insert contained an ATG at its
[23]. A prevailing opinion was to consider metallothionein 5' end that was in-frame with the down streamATG (Fig. 3).
in Menkes cells to be hypersensitive to induction by Cu [24]. Assuming the upstream ATG is a second start site, the
This proved not to be the case, however, when it was shown intervening sequence ofbases would code for an addendum
that Menkes cells respond to the same intracellular level of peptide of22 amino acids atthe N-terrninal. The added bases
Cu that induced metallothionein in normal cells [25,26]. suggested that the N -terminal of some ATP7As may extend
The discovery of a Cu-ATPase as the defective factor in 22 amino acids further than reported and could represent a
Menkes cells has emphasized membrane transport as the isoform ofATP7 A. We have tentatively called this new fonn
process controlling Cu homeostasis in cells. One should not ATP7A-1.
60

M F F F information far ATP7 A. Cloning each of the fragments


revealed that at least two had unique sequences at the 5' end.
This was apparent when an upper primer complementary to
the 45 bp insert identified in BeWo cells was used to generate
PCR products. One fragment had a 45 bp insert as expected,
but a slightly larger fragment contained a second insert of 192
21.2 bp that was 3' to the 45 bp insert. Both inserts were upstream
5.8 from the ATG start si te and were localized in what would
5.14 correspond to the exon I region of the transcript. Thus, like
4.27 3.7
2.8 Be Wo cells, Caco-2 cells also express a 45 bp insert.
1.8 Although not seen in BeWo cells, the 192 bp would code
l:~~ for a 66 amino acid polypeptide. ABLAST search of the
amino acid sequence was positive for two proteins in the
data bank. One referred to as p 115 has been classified as a
Golgi anchoring protein originally found on the trans-Golgi
membrane and believed to function in cis to medialcompart-
mental transport within the Golgi stacks [27]. The second
protein, referred to as TAP (transcytosis associated protein),
Fig. 2. Multiple cDNA Fragments of ATP7AmRNA in BeWo Cells. The like p 115, was involved in intravesicular transport within the
fragments were generated by PCR using primers that were designed to
cytosol [28]. Both proteins mediate a stable attachment of
cover the open reading frame. M - Lambda HindlIIfEeoRi molecular weight
marker; F - filter-grown BeWo cells. vesicles to their target membranes. This latter observation is
particularly important in view of the recent demonstration that
the Menkes protein could be associated with vesicles that
Caco-2 cells display both 45 bp and 192 bp insertions move from the Golgi to the plasma membrane in response to
a Cu challenge [29]. We have referred to theATP7 A with the
The Caco-2 cellline has been used as a cell model emulating two inserts as ATP7A-2.
the intestinal mucosa. Our interest in this cellline sterns from
its colonic origin and appropriateness to studies of absorption
and bioavailability of Cu in the gut. In Menkes disease, Cu Conclusions
is trapped in the intestine and fails to penetrate the intestinal
mucosa, indicating the final stage in intestinal transport is In summary, there is now compelling evidence that internal
interrupted. Stoppage is believed to occur at the release of Cu metabolism is subject to the action of Cu-ATPases that
Cu ions from the intestinal cells which precludes absorption operate at the membrane surface and conduct anATP-driven
by capillarics on the serosal side. Predictably, the Caco-2 cells movement ofCu ions across the membrane. The nmction of
should express the Menkes gene and ATP7 A protein. In these CuATPAses is not unlike those of Ca-ATPases that
searching far the gene, it was noted that the Caco-2 cells pump Ca2+ ions from cells. Cu-ATPaseperform no less ofan
displayed at least three transeripts that encoded genetic important function in maintaining a favorable cellular
environment of Cu. In Menkes disease, Cu ions fail to be
released from secretory cells. Therefore, it is logical to place
5' 45 Base Insert the Menkes ATPase on the excretory surfaces of polarized
5' M K
cells where a vectorial dis charge of Cu would favor its
AACCAG ATGAGG AAACTGAGCATCAGAAAGAGACCA
passage out of one side of the cello In the intestine, the
AACCAG . . . . . . . . . . . . . , . . . . . . . . . . . . columnar epithelial cells discharge Cu from basal surfaces
NNLLKEC EE K
giving absorbing capillaries in the serosa access to the Cu
AACAATTTGCTGAAGGAATGTAATGAGGAAATCAAAATG ions. In the brain, Cu is discharged from the glia to be taken
. . . . . . . . . . . AGGAATGTAATGAGGAAATCAAAATG up by the neurons. These scenarios perhaps explain thc
function of the Menkes ATPase and provide a rationale as to
its plasma membrane location.
Fig. 3. Sequence 01' the 45 Base Insert in BeWo Cells. Shown are the Table 3 surnmarizes the current data regarding isoforms of
published sequenee of ATP7A in the region 5' to the start site (underlined,
far right). The modified sequenee is shown above the reported sequenee the primary Menkes transcript. Research to characterize these
and above that are the amino acids that would eorrespond to the base transeripts is currently underway in the labaratory. It is
sequence. interesting to note that a 2-16 transcript has been observed
61

Table 3. Occurrence of Modified ATP7A Transcripts in Cell Lines concentrations. Further research is needed to reveal the
functions of ATP7 A in light of these variants and if each
ATP7A Variant
Cell Line 2-16 45 bp 192 bp
ATP7 Aform is subject to tissue and developmental specificity,
ultimately to leam how both work in regulating Cu homeo-
BeWo 1 + (pCR, Seq) + (pCR, Seq) -(PCR) stasis. Eventually this research must confront the factors that
Caco-2 + (PCR) + (PCR) + (PCR)
Fibroblasts + (PCR) -(PCR) -(PCR)
activate as weil as inhibit the function ofthe Cu-ATPases.
HepG2 + (PCR) -(PCR) -(PCR)
C6glioma -(PCR) ND ND
Brain Capillary -(PCR) ND ND
Acknowledgements
IGrown on filter surfaces to induce expression of ATP7 A PCR, as
determined by polymerase chain reaction; Seq - sequence determined; ND This research was supported in part by grants from the
- not determined. See text for meaning of terms.
National Institute ofHealth, HD29952 We are grateful to Mr.
lohn Nelson for providing technical assistance.
in a number of different cell types and is not restricted to any
one particular tissue. Only rat glioma and brain capillary cells
have thus far not shown the 2-16 variant. The actual detection References
of a fusion between exons 2 and 16, however, has been verified
by sequence analysis only for BeWo cells, Caco-2 cells, and I. Danks DM, Campbell PE, Walker-Smith J, et al.: Menkes' kinky-hair
syndrome. Lancet I: 11 00--11 03, 1972
HepG2 cells. Its occurrence in HepG2 cells is particularly
2. Danks DM: Copper Deficiency in Humans. Ann Rev Nutr 8: 235--
important because Northern blot analysis has failed to detect 257, 1988
a Menkes mRNA in liver [4, 10]. The data, therefore, suggest 3. Vulpe C, Levinson B, Whitney S, Packman S, Gitschier J: Isolation of
that liver cells may have the capacity to synthesize ATP7 A a candidate gene for Menkes disease and evidence that it encodes a
mRNA but fail to produce a translation product. One may copper-transporting ATPase. Nature Genet 3: 7-13, 1993
4. Mercer JFB, Livingston J, Hall B, et al.: Isolation ofa partial candidate
assume that a posttranscriptional modification giving rise to
gene for Menkes disease by positional cloning. Nature Genet 3: 20--
a 2-16 variant is one way ATP7 AmRNA is regulated in cells. 25,1993
Table 3 also shows that the 45 and 192 bp inserts are seen 5. Bull PC, Tbomas GR, Rommens JM, Forbes JR, Cox DW: The Wilson
only in BeWo and Caco-2 cells, but not fibroblasts or HepG2 disease gene is a putative copper transporting P-type ATPase similar
cells. Not finding either in fibroblasts provides an explanation to the Menkes gene. Nature Genet 5: 327-336,1993
6. Tanzi RE, Petrukhin K, Chernov I, et al.: The Wilson disease gene is
for why the 45 and 192 sequences have not been reported
a copper transporting ATPase with homology to the Menkes disease
previously since much of the early sequence work onATP7A gene. Nature Genet 5: 344--350, 1993
was performed in fibroblasts. The meaning of these inserts 7. Petrukhin K, Fischer SG, Pirastu M, et al.: Mapping, cloning and
at the 5' end is still subject to scrutiny and a more detailed genetic characterization of the region containing the Wilson disease
analysis. Recently, imrnunohistochemical data suggests that gene. Nature Genet 5: 338--343,1993
8. Thomas GR, Forbes JR, Roberts EA, Walshe JM, Cox DW: The Wilson
the Menkes protein (perhaps one form of this protein) is
disease gene: Spectrum of mutations and their consequences. Nat Genet
associated with vesicles-like structures that take harbor 9: 210--217,1995
around or within the Golgi ofthe cell [29]. These vesicles are 9. Vulpe CD, Packman S: Cellular copper transport. Ann Rev Nutr 15:
mobile and relocate to the plasma membrane in response to 293-322, 1995
high extracellular Cu. To be associated with the Golgi, 10. Chelly J, Tümer Z, Tonnesen T, et al.: Isolation of a candidate gene
for Menkes disease that encodes a potential heavy metal binding
ATP7 A or some other vesicular protein must express a Golgi protein. Nature Genet 3: 14--19, 1993
targeting or docking signal. What we have referred to as 11. Yamaguchi Y, Heiny ME, Gitlin JD: Isolation and characterization of
ATP7A-2 may have such a signal appended to the N-terminus a human liver cDNA as a candidate gene for Wilson' s disease. Biochem
of the peptide that allows a Golgi location, i.e., anchored Biophys Res Commun 197: 271-277,1993
temporarily to proteins in the Golgi stacks. Moreover, since 12. Solioz M, OdermattA, KrapfR: Copper pumpingATPases: Common
concepts in bacteria and man. FEBS LeU 346: 44--47, 1994
the anchoring signal is near the Cu-binding region onATP7A,
13. Shiraishi N, Aono K, Taguchi T: Copper metabolism in the macular
it is conceivable that Cu ions create a situation that disengages mutant mouse: An animal model ofMenkes' kinky-hair disease. Biol
the targeting signal from its Golgi receptor. The protein referred Neonate 54: 173-180, 1988
to asATP7Amay be a form that is expressed constitutively and 14. Shiraishi N, Kondoh S, Hiraki Y, Aono K, Taguchi T: Metallothionein
takes part in Cu transport under non-stressed conditions. On in kidney and liver of the macular mouse as an animal model of
Menkes' kinky hair disease. Physiol Chem Phys Med NMR 19: 227-
the other hand, ATP7 A-2 may be a form within the Golgi 233, 1987
region that is designed to absorb Cu ions into vesicles prior to 15. Shiraishi N, Taguchi T, Kinebuchi H. Copper-induced toxicity in
their discharge from the cell and could be responsible for macular mutant mouse: An animal model for Menkes' kinky-hair
maintaining Cu homeostasis in the face of elevated Cu disease. Toxicol Appl Pharmacol 11 0: 89--96, 1991
62

16. Kodama H, Abe T, Takama M, Takahashi I, Kodama M, Nishimura 23. Riordan JR, Jolicoeur-Paquet L: Metallothionein accumulation may
M: Histochemicallocalization of copper in the intestine and kidney of account for intracellular copper retention in Menkes' disease. J Biol
macular mice: Light and electron microscopic study. J Histochem Chem 257: 4639-4645, 1982
Cytochem41: 1529--1535, 1993 24. Lcone A, Pavlakis GN, Hamer DH: Menkes' discase: Abnormal
17. Camakaris J, Petris MJ, Bailey L, et a1.: Gene amplification of the metallothionein gene regulation in response to copper. Ce1l40: 301-
Menkes (MNK; ATP7 A) P-type ATPase gene of CHO cells is 309, 1985
associated with copper resistance and enhanced copper effiux. Hum 25. Sone T, Yamaoka K, Minami Y, Tsunoo H: Induction of metallothionein
Mol Genet 4: 2117-2123, 1995 synthesis in Menkes' and normal lymphoblastoid cells is controlled
18. Qian YC, Tiffany-Castiglioni E, Harris ED: Copper transport and by the level of intracellular copper. J Biol Chem 262: 5878-5885,
kinetics in cultured C6 rat glioma cells. Am J Physiol269 (Cell Physiol 1987
39): C892--C898, 1995 26. Herzberg NH, Wolterman RA, van den Berg GJ, Barth PG, Bolhuis
19. Mann JR, Camakaris J, Danks DM: Copper metabolism in mottled PA: Metallothionein in Menkes' disease: Induetion in eultured museIe
mouse mutants. Defective placental transfer of 64Cu to foetal brindled ceUs. J Neur Sci 100: 50-56, 1990
(Mobr) mice. Biochem J 186: 629--{j31, 1980 27. Waters MG, Clary DO, Rothman JE: A novel 115-kD peripheral
20. Yuan DS, Stearman R, Dancis A, Dunn T, Beeler T, Klausner RD: membrane protein is required for intereisternal transport in the Golgi
The Menkes/Wilson gene homologue in ye.st provides copper to a stack. J CeU Bio1118: 1015-1026, 1992
ceruloplasmin-like oxidase required for iron uptake. Proc Nat! Acad 28. Barroso M, Nelson DS, Sztul E: Transcytosis-associated protein (TAP)/
Sci (USA) 92: 2632-2636, 1995 p 115 is a general fusion factor required for binding of vesicles to
21. Qian YC, Tiffany-Castiglioni E, Harris ED. Functional analysis of a acceptor membranes. Proc Nat Acad Sci (USA) 92: 527-531, 1995
genetic defect of copper transport (Menkes disease) in different cell 29. Petris MJ, Mereer JFB, Culvenor JG, Gleeson PA, Camakaris J:
lines. Am J Physiol 271 (Cell PhysioI40): C378-C384, 1995 Ligand-regulated transport of the Menkes copper P-type ATPase
22. Qian YC, Tiffany-Castiglioni E, Harris ED: Coincident expression of emux pump from the Golgi apparatus to thc plasma membrane: A
the Menkes gene with copper effiux in human placental cells. Am J novel mechanism ofregulated trafficking. EMBO J 15: 6084--{i095,
Physiol270 (Cell PhysioI39): CI 88O--CI 884, 1995 1996
Moleeular and Cellular Biochemistry 188: 63--{i9, 1998.
© 1998 Kluwer Academic Publishers.

Zine and immunity


Ananda s. Prasad
Department ofInternal Medicine, Division ofHematology-Oncology, Wayne State University School ofMedicine, Detroit,
Michigan, USA

Abstract
Nutritional defieieney of zine is widespread throughout the developing countries and a eonditioned defieieney of zine is known
to oeeur in many diseased states. Zine is known to play an important role in the immune system and zine defieient subjeets
may experienee inereased suseeptibility to a variety of pathogens. We have studied the effeets of a mild defieieney of zine on
T eells in an experimental model ofhuman zine defieieney. We showed that T eell funetions were affeeted adversely even when
the defieieney of zine was mild in humans. Charaeteristieally during zine defieieney, the serum thymulin aetivity (a thymie
hormone) was deereased whieh was restored following zine supplementation. Our studies also showed that zine defieieney
eaused an imbalanee between THI and TH2 funetions. The produetion ofIFN-g, IL-2, TNF-a (produets ofTHI eells) were
deereased, whereas the production ofIL-4, IL-6 and IL-lO (produets of TH2) were not affeeted during zine defieieney. T eell
subpopulation studies revealed that the CD4+ CD45RA + to CD4+ CD45RO+ ratio was deereased as a result of zine deficieney,
suggesting that zine may be required for the regeneration of new CD4+ T eells. We further doeumented that zine defieieney
deereased NK eelllytie aetivity and eaused a deerease in the percentage of CD8+ CD73+ T eells whieh are known to be
predominantly preeursors of eytotoxie T eells. In a suitable eell eulture model our studies revealed that the gene expression of
a DNA synthesizing enzyme TK was affeeted adversely whieh resulted in delayed eell cycle and decreased cell growth. The
above immunological eonsequences of zine deficiency may be responsible far deereased eell mediated immune functions in
zine defieient subjeets. (Mol Cell Biochem 188: 63--69, 1998)

Key words: zine, immune system, eell growth, cell eycle, T cell

Introduction Both primary and seeondary antibody responses were


reported to be depressed in zinc deficient mice. An investig-
It has been known for many years that zinc deficiency in ation into the influenee of suboptimal zine nutrition on the
experimental animals results in atrophy of thymie and lym- memory eells in mice also showed an impairment which was
phoid tissue [1]. Later studies in young adult zine defieient mice only partially eorreeted by zine repletion [7]. A deerease in
by other investigators showed thymic atrophy, reductions in in vivo-generated eytotoxie T -killer aetivity to allogeneic
the absolute number of splenoeytes, and depressed responses tumar eelJs in zine defieient mice was observed by Femandes
to both T-eell-dependent (TD) and T-cell-independent (TI) et al., [2] and Good and Fernandes [8]. Frost et al., [9]
antigens [2-5]. In response to sheep red blood cells (SRBC), reported an impairment in cell-mediated response to non-H2
a TD antigen, the zine defieient mice produeed only 40 % as allogeneie tumor eells in zine defieient mice. Animals
many IgM and IgG plaque-forming eells (PFC) per spleen maintained on a zine defieient diet for as little as 2 weeks
as did the zine-adequate mice. Although the ratio oft and B developed a severe impairment in their ability to generate a
eells were not ehanged, the defieient mice had nearly double eytotoxie response to the tumor challenge. This phenomenon
the proportion ofB eells bearing surfaee IgM in eomparison was totally reversible by zine supplementation.
to the eontrol mice. It was suggested that immature B eells Zine defieieney was shown to affeet the development of
were aeeumulating in the spleens of zine defieient mice [3]. immune responses adversely if it oeeurred during the eritieal
Nash et al. [6] had observed earlier that greater numbers of periods of ontogeny [10]. At 4 weeks of age, direet splenie
immature T eells were present in zine deficient mice. PFC responses to SRBC were dramatieally deereased in

Addressfor of!prints: A.S. Prasad, University Health Center, 5-C, 4201 SI. Antoine, Detroit, Michigan, 48201, USA
64

outbred N: NIH (S) mice that were moderately or severely reported by Minniehet al. later [15]. The anemia was eorrected
deprived of zine during postnatal period. The zine defieient by administration of oral iron only in every instanee.
mice also exhibited a disordered serum immunologieal profile Lemann [16] had deseribed similar clinieal features in
with absence of deteetable IgM, IgG2a, and IgA, along with patients with hookworm infeetion in the USA in 1910 but he
an elevated serum levels ofIgG. These findings suggest that did not relate these to a nutritional defieiency. In Turkey,
zine defieieney during the period of rapid growth and Reirnann [17] described similar cases and he eonsidered a
development may predispose higher animals to an aequired genetie defect as a possible explanation for eertain aspects
immune defieieney and predispose the anirnals to oppor- of this clinical syndrome.
tunistie infeetions [11]. In another experiment, progeny of Our detailed description of this elinical syndrome in Iran
zine deprived dams showed signifieant growth retardation was published in 1961 [18]. Although we had no data to
and preferentially deereased growth of spleen and thymus document zine deficieney in our patients from Iran, we
relative to eontrols [12,13]. Cross-fostering ofzine deprived speeulated that a deficieney of zine may have caused growth
pups on eontrol dams improved the growth of organs but retardation, gonadal failure, skin changes, and mental
failed to improve the growth of the spleen and thymus. lethargy [18].
In this review, the effeet of zine defieieney on eell mediated The Iranian patients were treated with ferrous sulphate (Ja
immunity in humans will be presented and possible mechan- daily) and a weil balanced nutritional diet containing animal
isms of zine action on T eell funetions will be diseussed. The protein. Following this, the anemia was eorrected, the
effeet of zine defieieney on eell proliferation in a suitable eell hepatosplenomegaly improved, pubic hair grew, and the size
eulture model will be also presented briefly. Although zine of their genitalia inereased [18]. A marked inerease in serum
deficieney in animals have been known to oeeur for many alkaline phosphatase was observed following treatment.
deeades [14], zine deficieney in humans was deseribed only Retrospeetively this observation was related to two faetors:
in the early sixties. I will also briefly include the historieal (1) the ordinary pharmaeeutiea1 preparation of iron also
aspeet ofthe diseovery ofhurnan zine defieieney in this review. eontained appreciable quantities ofzine as a contarninant and
(2) the animal protein provided available zinc, thus inducing
the aetivity of alkaline phosphatase, now known to be a zine
Discovery of human zinc deficiency metalloenzyme.
Growth retardation and testieular atrophy are not observed
The essentiality of zine for animals has been known sinee in iron deficient experimental animals, on the other hand zine
1934 [14], however its ubiquity made it seem unlikely that deficiency was known to produce growth retardation and
alterations in zine metabolism, eould lead to signifieant testicular atrophy in anirnals. Inasmuch as heavy metals form
problems in human nutrition or elinieal medieine. This insoluble complexes with phosphates, we speeulated that
attitude is now ehanged. some unknown dietary factors were responsible for complex-
While I was a visiting Professor in Shiraz, Iran, in the fall ing both iron and zine, thus affeeting adversely the availa-
of 1958, Dr. J.A. Halsted brought to my attention a 21 year bility of these elements for absorption. Phytate (inositol
old male at Saadi Hospital, who looked like a 10 year old boy. hexaphosphate) which is present in cereal proteins is now
He was severely growth retarded. The other elinieal mani- known to affect bioavailibility of zine and impair its absorp-
festations in this patient included hypogonadism, severe tion [19]. Thus, in our Iranian patients growth retardation,
hypoehromie microeytic anemia, hepatosplenomegaly, rough testieular atrophy, and ehanges in serum alkaline phosphatase
and dry skin, mentallethargy, and geophagia (habit of eating following treatment could have been explained on the basis
clay). The patient ate only bread made ofunleavened wheat of zine deficieney.
flour and the intake of animal protein was negligible. He ate Subsequently I moved to Egypt. There I encountered
one pound of clay per day. Later we learned that geophagia patients who resembled the Iranian dwarfs. Their elinical
was fairly prevalent in the villages around Shiraz. The anemia features were remarkably similar exeept for the following:
was due to iron deficieney but there was no blood loss. the Iranian patients had more pronouneed hepatospleno-
Hypopituitarism as an explanation of growth retardation was megaly; they had a history of geophagia; and none had
ruled out inasmueh as we observed ten additional similar hookworm infeetion. In contrast, the Egyptian patients had
cases within a short period of time. both sehistosomiasis and hookworm infections, and none had
We eonsidered three probable faetors responsible for a history of geophagia.
anemia in these patients: (a) The total amount of available Our team (HH Sandstead,A Sehulert,AMiale, Z Farid, and
iron in the diet was insufficient: (b) an exeessive sweating in myselt) earried out in depth study of the Egyptian patients
a hot climate probably eaused greater iron loss from the skin at the US Navel Medieal Research unit No. 3, Cairo Egypt
than would oeeur in temperate elimate; and (e) geophagia [20]. Animal protein intake ofthese subjeets was negligible.
may have further deereased the iron availability as was Their diet eonsisted of bread and beans (vicia fava). We
65

doeumented that these dwarfs were zine deficient. This Further studies in Egypt demonstrated that the rate of
eoncIusion was based on the fo11owing: Zine levels in plasma, growth was greater in subjeets who reeeived zine supp-
red ee11s, and hair were deereased; and 65 Zn studies revealed lementation as eompared with those who reeeived iron
that the plasma zine turn over rate was greater, the 24-h instead, or those who reeeived only an adequate animal
exehangeable pool was sma11er, and the exeretion of 65 Zn in protein diet [21, 24]. The zine supplemented group gained
stool and urine was less in these patients than in the eontrols approximately 5 inches in height on an annual basis. Pubie
[20]. This was the first demonstration that zine defieieney hair appeared in a11 eases within 7-12 weeks fo11owing zine
oeeurred in humans. supplementation. Genitalia beeame normal and seeondary
We carefu11y excIuded chronie debilitative disorders and sexual eharaeteristies developed within 12-24 weeks in
ehronie liver disease in our subjeets. Hyperzineuria in patients reeeiving zine. In eontrast no such ehanges were
humans, in the absence of advaneed eirrhosis ofthe liver, had observed in iron supplemented group or in the group reeeiv-
never been deseribed before. Furthermore, in eontrast to ing an animal protein diet only during a eomparable 1ength
eirrhoties, who exerete abnorma11y high quantity of zine in oftime. We, therefore, eoncIuded that growth retardation and
urine [21], our patients exereted less zine in urine than did gonadal hypofunetion in this syndrome were due to def-
the eontrol subjeets. ieieney of zine.
The physieians in Iran related the growth retardation and In the Uni ted States, Caggiano et al. [25] in 1969 were the
hypogonadism in these patients to viseeralleishmaniasis and first to report a ease of zine defieieney in a Puerto Riean
geophagia. We, however, failed to find any evidenee for subjeet with dwarfism, hypogonadism, hypogammaglobulin-
viseeralleishmaniasis in our patients in Iran. The role of emia, giardiasis, strongyloidosis, and sehistosomiasis. The
geophagia was not c1ear, however, we suspeeted that the patient responded to zine administration whieh improved
exeess amount of phosphate in the cIay may have prevented growth and development in this patient.
absorption ofboth dietary iron and zine. The predominantly In 1972 Hambidge et al. [26] observed that a number of
wheat diet in the Middle East known to eontain high quant- ehildren from middle-cIass families in Denver, Co10rado
ities of phytate and fiber, mayaiso have further redueed the showed evidenee of nutritional zine defieieney. Growth
availability of these elements. retardation, poor appetite, and impaired taste aeuity were
In Egypt, the physicians related the growth retardation in re1ated to zine defieieney in these ehildren. Zine supp1ement-
these patients to sehistosomiasis. Chinese investigators also ation eorreeted these cIinieal problems.
implieated sehistosomiasis as a eausative faetor for growth Halsted et al. [23] in 1972 in a study ofzine supp1ement-
retardation [21]. However, the existcnee of dwarfism and ation in Iran involving a group of 15 men and 2 women,
hypogonadism in areas such as Iran and Kharga Oasis [22], eonfirmed the oeeurrenee of zine-responsive growth and
where sehistosomiasis was not present, indieated that this gonadal dysfunetions in the Middle East. Their results were
parasitie infeetion was not solely responsible for these cIinieal similar to those reported by Prasad and Sandstead et al. from
findings. Furthermore, as mentioned earlier, iron defieieney Egypt [21, 24].
in animals and humans is not known to eause growth retard- Zine defieieney is now known to be prevalent in human
ation and hypogonadism. In view of the above and the population throughout the world. It is believed that zine
similarity between these cIinieal features of our subjeets and defieieney should be present in countries where primarily
those seen in animals with zine defieieney, we related the eereal proteins are eonsumed by the population. It is our
growth retardation and hypogonadism in this syndrome to a estimate that zine defieieney is as prevalent as iron defieieney
defieieney of zine. in the world.
Anemia in a11 eases was due to iron defieieney and this was In 1974, a landmark deeision to estab1ish reeommended
eorreeted by oral iron therapy. The explanation for hapato- dietary allowanee for humans for zine was made by the
splenomegaly was not obvious. We eonsidered three possibil- National Research Couneil, Food and Nutrition Board ofthe
ities: (I) anemia, (2) zine defieiency, or (3) a eombination of National Aeademy of Seienees.
these. Later we observed that following zine supplement-
ation, the size ofthe liver and spleen deereased signifieantly
in all the subjeets. The exaet meehanism as to how zine Conditioned dejiciency of zinc in humans
deficieney may have increased the size ofthe liver and spleen
remains unknown. It is now evident that nutritional as weil as eonditioned
Our studics in the Middle East were limited to males, defieieney ofzine may eomp1ieate many diseases. MaeMahon
inasmuch as females refused to participate in our studies. et al. [27] demonstrated for the first time the oeeurrenee of
Later, Halsted et al. [23] demonstrated in Iran that zine zine defieiency in patients with malabsorption. Since then,
defieieney in females manifesting growth retardation was many examples of zine defieieney in patients with maIabsorp-
also prevalent. tion have been reported [28].
66

In 1973, Bames and Moynahan [29] studied a 2 year-old on immune funetions [38]. A semipurified diet based on
girl with severe aerodermatitis enteropathiea (a genetie texturized soy protein was developed for eonsumption by
disorder) who was being treated with diiodohydroxyquino- human volunteers. This diet supplied ealories, protein, macro
line and a laetose-defieient synthetie diet. The elinieal and miero elements, and vitamins aeeording to reeommended
response to this treatment was unsatisfaetory. They also dietary allowanees (NationalAeademy ofSeience, Food and
observed that the serum zine in this patient was deereased, Nutrition Board) exeept for zine, whieh varied as desired.
therefore, they administrated oral zine sulphate. The skin Male volunteers aged 20-45 years were seleeted for these
lesions, and gastrointestinal manifestations cleared eompletely. studies. The experimental protoeol for our studies was
It soon beeame clear that zine might be fundamental to the reviewed and approved by the Human and animal Investi-
pathogenesis of this rare genetie disorder and zine was indeed gation Committee ofWayne State University and informed
eurative. This original observation was quiekly eonfirmed in eonsents were obtained in each case. Before the study, a
other eases throughout the world. The underlying path- thorough history, physical examination, and routine labor-
ogenesis of zinc deficiency in these patients is most likely, atory test (including complete blood count, liver funetion,
related to malabsorption of zine, the mechanism of which is sequential multiple analyzer-12, and serum electrolytes) were
not known. performed and found normal. The volunteers were ambul-
Kay and Tasman-Jones [30] reported the occurrenee of atory and were encouraged to do daily moderate exercise
severe zine defieieney in subjects who reeeived total parent- throughout the study period.
eral nutrition for prolonged periods without zine. This The subjects were given a hospital diet eontaining animal
observation was eonfirmed by many investigators [31, 32] protein daily for 4 weeks. This diet averaged 12 mg Zn/d,
and now zine is being routinely included in total parenteral eonsistent with the RDA. After that they reeeived a semi-
fluids for subjeets who are likely to reeeive such therapy for purified soy-protein-based experimental diet that supplied
long periods. 3.0-5.0 mg Zn/d. The details for preparation of the experi-
An example of severe parakeratosis in humans related to mental diet were published elsewhere [38]. This regime was
defieieney ofzine was first reported by Klingberg et al. [33] eontinued for 28 weeks, after whieh the subjeets reeeived 27
in a patient who reeeived penicillamine therapy for Wilson' s mg Zn/d supplement for 12 weeks while still eonsuming
disease. Zine supplementation eompletely reversed the experimental diet.
clinical manifestations. Throughout the study, the amounts of all nutrients, includ-
Several studies have suggested that various elinical ing protein, amino acids, vitamins, and minerals (both maero-
manifestations in patients with sickle-cell disease such as and mieroelements), were kept constant, meeting the RDAs,
growth retardation, hypogonadism in the males, lack of exeept for zine, whieh was varied as outlined above. By this
prompt healing of chronie leg ulcers, abnormal dark adapt- teehnique we were able to induee a speeifie zine defieieney
ation, and abnormality in cell mediated immunity are related in human volunteers.
to a deficieney of zine [34-37]. The exaet pathogenesis of zine The peripheral blood eells (lymphoeytes, granuloeytes, and
defieieney in si ekle eell disease is not weil understood, platelets) for zine assay were isolated by a modification of a
however, hyperzineuria has been observed in these patients and previously published method [39]. Special eare was taken to
this may be a eontributing faetor in eausing zine defieieney. remove red eells from the granuloeytes, platelets from the
granuloeytes and lymphoeytes, and trapped plasma from the
platelets. Extreme eare was exercised to avoid exogenous
Studies of immune functions in experimental human model zine eontamination throughout the assay proeedure. Zine was
assayed in the sampies by means of an atomie-absorption
During our studies in the Middle East, we observed that most speetrophotometer [39].
of the zine defieient dwarfs did not live beyond the age of When zine defieieney was very mild (5.0 mg Zn intake
25 years. The eause of death appeared to be infeetions during the zine-restrieted period), the plasma zine eoneent-
although the exaet nature of infeetive organisms was not ration remained more or less within the normal range and it
known. Parasitie infeetions were common, however, viral and decreased only after 4-5 months of zine restrietion. On the
baeterial infections rcmained undoeumented. The possibility other hand, zine eoneentrations in lymphoeytes, granuloeytes,
that zine defieiency may have played a role in immune and platelets deereased within 8---12 weeks, suggesting that
dysfunetions in the zine defieient dwarfs was eonsidered but the assay of eellular zine may provide a sensitive eriterion
lack of proper faeilities prevented us from gathering meaning- for diagnosing mild defieiency ofzine [40].
ful data on immune functions in the patients from the Middle In subjeets who reeeived 3.0 mg dietary zine during zine-
East. restrieted period, the plasma zine eoneentration remained >
We have developed an experimental model whieh allowed 100 Ilg/dl for 2 months on the zine-restrieted diet. In the third
us to study speeifie effeets of mild zine defieieney in humans month a signifieant deerease in the plasma zine eoneent-
67

rations was observed [41]. In these volunteers, the zine in affeeted due to zine deficiency [45]. In this study the effeet
platelets deereased within 1 months, and zine eoneentrations of zine restrietion on IL2 production showed only a border-
in lymphoeytes and granuloeytes deereased within 2 months line signifieanee, probably due to a small 11. Our earlier
after institution ofthe zine-restrieted diet [41]. The maximum studies have shown a signifieant effeet of zine defieieney on
decline in eellular zine eoneentrations was observed at the TL-2 aetivity and production in experimental human model
end of 6 months of restrieted dietary zine intake. subjeets and in patients with siekle eell disease and head and
We [41] assayed serum thymulin aetivity in mildly zine- neck cancer patients [41, 46]. Taken together, our studies
defieient human subjeets. Thymulin is a thymus-speeifie suggest that zine affeets mainly the funetions ofTHI eells.
hormone and it requires the presenee ofzine for its biological We did not assay TNF-ß, surprisingly however, we ob-
aetivity to be expressed [42, 43]. Thymulin binds to high- served that the produetion of TNF -<X by peripheral blood
affinity reeeptors on T eells, induees several T-eell markers, mononuclear eells, was deereased in zine defieieney. We
and prornotes T-eell funetion, including allogenie eyto- speeulate that THI eells in humans also produec TNF-<x,
toxieity, suppressor funetions, and interleukin-2 (IL-2) inasmueh, as onlyTH 1 fimetions were affeeted in human zine
produetion. defici eney.
The serum level of biologieally aetive thymulin was IFN-y is known to down regulate the TH2 clone, and IL-
evaluated by a rosette assay deseribed earlier [42, 43]. The 10 may down regulate theTHI clone [49, 50].An imbalanee
assay analyzes the eonversion of relatively Azathioprine between THI and TH2 responses in patients with human
(Az)-resistant spleen of adult thymeetomized mice to theta immunodefieieney virus infeetion has been implieated in the
positive rosette-forming eells that are more sensitive to Az. immune dysregulation in these patients and it has been
As a result of mild defieieney of zine, the aetivity of proposed that resistanee to infeetion andlor progression to
thymulin in serum was signifieantly deereased and was aequired immunodefieieney syndrome is dependent on a THI
eorreeted by both in viva and in vitra zine supplementation. > TH2 dominance [48]. Our data in experimental human
The in vitra supplementation studies indieated that the model suggest that eell-mediated immune dysfunetions in
inaetive thymulin peptide was present in the serum in zine- human zine defieieney may be due to an imbalanee between
defieient subjeets and was aetivated by addition of zine [41]. THI and TH2 eell funetions.
The assay of serum thymulin aetivity with or without zine THI eells are known to promote maerophage aetivation
addition in vitra thus may be used as sensitive eriterion for and produetion of eomplement fixing and opsonizing anti-
the diagnosis of mild zine defieieney in humans. bodies [49,50]. IFN-yis the major eomponent ofTHl response
An inerease in T I01 -, slg-eells, a deerease in the ratio ofT4+ panel, and it upregulates major histoeompatibility eomplex
to T8+, and deereased IL-2 aetivity were observed in the class I antigen expression. Our studies, therefore, provide a
experimental human model during the zinedepletion phase, possible meehanism of zine on eellmediated immunity.
all of whieh were eorreeted after repletion with zine [41]. We T eell subpopulation studies revealed that CD4+ to CD8+
[44] had previously reported that natural-killer (NK)-eells ratio was signifieantly related to zine status [45, 46]. A
activity was also sensitive to zinc restriction, thus it appears deerease in this ratio was observed during zine defieieney and
that zinc may playa very important and eritical role in the this was eorreeted by zine supplementation. A borderline
funetions of T eells in humans. signifieant effeet of zine status on the ratio of CD4+
In our reeent studies we have shown that a mild defieieney CD45RA+ to CD4+ CD45RO+ eells was seen in the human
of zine leads to an imbalanee of THI and TH2 funetions, volunteers. The newly produeed CD4+ T Iymphoeytes
deereases the reeruitment ofT naive eells (CD4+CD45RA+), express CD45 isoforms, whieh are designated CD45RA+,
and deereases the pereentage of CD73+ eells in the CD8+ and onee these cells encounter a speeifie antigen, they
subset that are preeursors to eytotoxie T Iymphoeytes (CTL) beeome 'memory' T Iymphoeytes, expressing a small isoform
[45-47]. In mice, IL-2, IFN-y, and TNF-ß are eonsidered to ofc1eaved CD45 designated CD45 RO+ eells [51]. It appears
be produets ofTHI eells, whereas IL-4, IL-6, and IL-lO are that zine is required for the regeneration ofnew CD4+ T eells.
produets ofTH2 eells. THI and TH2 subsets are well eharae- Inasmueh as zine is essential for the aetivity of thymulin, a
terized in the murine system, however, these eategories are thymie hormone, it is possible that zine may be intrinsieally
not as clear cut in humans. Nonetheless, the separation ofT involved in the devc10pment ofhematopoietic stern eells to
helper eells into THI and TH2 aeeording to their fimetions T lymphoeytes in the thymie microenvironment [42, 43].
in eell-mediated (THI) and humoral immunity (TH2) is Earlier we had observed that NK eell aetivity is zine
beeoming very useful in the understanding of the immune dependent [44]. In patients with zine defieiency, NK eell
meehanisms in humans [48-50]. aetivity is deereased and this is correetable by zine supp-
Our studies in the experimental human model showed for lementation [44]. Our reeent studies in experimental human
the first time that the produetion of IFN-y was deereased, model showed that the percentage of CD8+ CD73+ T Iymph-
whereas the production of IL-4, IL-6 and IL-I 0 was not oeyte are deereased in zine deficieney and this is correeted
68

by zine supplementation [47]. CD8+ CD73+ lymphoeytes are References


predominantly preeursors to eytotoxie T lymphoeytes (CTL),
and the presenee of CD73 moleeule on CTL is required for 1. Prasad AS, Oberleas 0: Changes in aetivities of zine-dependent
antigen reeognition, the proliferative proeess, and for gener- enzymes in zine-deficient tissues of rats. J Appl Physiol31: 842-846,
ation of eytolytie proeess [47]. It is, therefore, likely that the 1971
inereased frequeney of infeetions seen in zine defieient 2. Femandes G, Nair M, Onoe K, Tanaka T, FLoyd R, Good RA:
Impairment of eell-mediated immunity funetions by dietary zine
patients may be related to a deerease in NK eelilytie aetivity deficieney in mice. Proe NatlAead Sei USA 76: 457-461, 1979
and deereased CTL aetivity. 3. Fraker PJ: Zine defieieney: a eommon immunodefieieney stale. Surv
Our studies in the experimental human model, thus show lmmunol Res 2: 155-163, 1983
that even a mild defieieney of zine in humans may be 4. Fraker PJ, OePasquale-Jardieu P, Zwiekl CM, Lueeke RW: Regener-
aeeompanied by an imbalanee of THI and TH2 eells, de- ation ofT-eell helper funetion in zine-deficient adult mice. Proe Natl
Aead Sei USA 75: 5660-5664, 1978
ereased serum thymulin aetivity, deereased reeruitment ofT 5. Fraker PJ, Haas S, Lueeke RW: Effect ofzine defieieney on the immune
naive eells, deereased percent ofT eytolytie eells and deereas- response ofthe young adultNJ mouse. JNutr 107: 1889-1895, 1977
ed NK eelllytie aetivity. These immunologie eonsequenees 6. Nash L, Iwata T, Femandes G, Good RA, Ineefy G: Effeet of zine
of zine defieieney may be responsible for deereased eell defieieney on autologous rosette-forming eells. Celllmmunol48: 238-
mediated immune funetions in zine-defieient subjeets. 243,1979
7. Fraker PJ, Gershwin ME, Good RA, Prasad AS: Interrelationships
between zine and immune funetion. Fed Proe 45: 1475-1479, 1986
8. Good RA, Femandes G: Nutrition, immunity, and eaneer- a review.
Effect 01 zinc on cell cycle and deoxythymidine kinase Part I: influenee of protein or protein-ealorie malnutrition and zine
gene expression in HUT-78 cells deficieney on immunity. Clin Bull 9: 3-12, 1979
9. Frost P, Rabbani P, Smith J, Prasad AS: Cell mediated eytotoxicity
and tumor growth in zine-defieient mice. Proe Soe Exp Biol Med 167:
Although zine is known to be involved in eell proliferation 333-337, 1981
and DNA synthesis, the meehanism by whieh zine may 10. Beaeh RS, Gershwin ME, Makishima RK, Hurley LS: Impaired
regulate these processes is not understood. We have studied immunologie ontogeny in postnatal zine deprivation. J Nutr 110: 805-
the role of zine on eell proliferation and gene expression of 815, 1980
DNA synthesizing enzyme, deoxythymidine kinase (TK), in 11. Beaeh RS, Gershwin ME, Hurley LS: Growth and development of
postnatally zine-deprived mice. J Nutr IlO: 201-2Il, 1980
T helper human malignant lymphoblastoid eelliine (HUT- 12. Beaeh RS, Gershwin ME, Hurley LS: The reversibility of develop-
78) [52]. In zine defieient and zine suffieient media, the eell mental retardation following murine fetal zine deprivation. J Nutr 112:
doubling time (mean ± 8.D.) ofHUT-78 was 59 ± 8 hand 1169-1181,1982
32.6 ± 6 h respeetively. The effeet ofzine was T eell speeifie, 13. Beaeh RS, Gershwin ME, Hurley LS: Nutritional faetors and auto·
inasmueh as the eell growth of anotherT malignant lympho- immunity. H. Prolongation of survival in zine-deprived NZBIW mice.
JImmuno1128: 308-313, 1982
blastoid eeliline, Molt-3 (immature T eells), was not affeeted 14. Todd WR, Elvehjem CA, Hart EB: Zine in the nutrition ofthe rat. Am
by zine deficieney. Iron, copper, or mangane se did not J Physiol107: 146-156, 1934
eompletely eorreet eell growth of zine defieient HUT-78 eells. 15. Minnieh V, Okevog1a A, Tareon Y, Arcasoy A, Yorukoglu 0, Renda F,
TK aetivity and the relative aeeumulation ofTK-mRNA were Demirag B: The e!feet of clay on iron absorption as a possible eause for
signifieantly deereased in zine defieient eells during GI phase anemia ofTurkish subjeets with piea. Am J Clin Nutr 21: 78-86, 1968
16. Lemann 11: A study of the type of infantilism in hookworm disease.
of eell eycle in eomparison to zine sufficient eells. Nuelear ArehlntemMed6: 139-146, 1910
run-on experiments and aetinomyein-D studies showed that 17. Reimann F: Growth anomalies and malformations in iron-deficient
the transcription ofTK-mRNA was affeeted adversely by zine states. Proeeedings of the 5th kongr Eur Gasellschaft Haematol.
defieieney. Cell eycle studies showed that more zine deficient Freiburg, FRG: HM Keller, 546-550,1955 (in Gerrnan)
eells remained in 8 phase and did not undergo mitosis in 18. Prasad AS, Halsted JA, Nadimi M: Syndrome of iron defieieney
anemia, hepatosplenomegaly, hypogonadism, dwarflsm and geophagia.
eomparison to zine suffieient eells. In eonclusion, our data Am J Med 31: 532-546, 1961
show that zine is a T eell speeifie growth factor and that a 19. O'Oell BL, Savage JE: Effeet ofphytie acid on zine availability. Proe
deereased gene expression ofDNA synthesizing enzymeTK Soe Exp Biol Med 103: 304-306, 1960
in zine defieient HUT-78 eells in GI phase, affeeted adversely 20. Prasad AS, Miale A, Farid Z, Sehulert A, Sandstead HH: Zine
the DNA synthesis in 8 phase and delayed eell eycle. metabolism in patients with the syndrome of iron defieieney anemia,
hypogonadism, and dwarfism. J Lab Clin Med 61: 537-549, 1963
21. Prasad AS: Metabolism of zine and its defieieney in human subjeets.
In: Prasad AS (ed). Zine Metabolism. Springfield, IL: Char1es W
Acknowledgement Thomas, pp 250-302, 1966
22. PrasadAS, SehulertAR, MialeA Jr, Farid Z, Sandstead HH: Zine and
iron deficiencies in male subjeets with dwarfism and hypogonadism
8upported by GER, George Eby Research, Austin Texas, but without aneylostomiasis and sehistosomiasis or severe anemia.
78704, and Labeatal Laboratories, Paris, Franee. Am J Clin Nutr 12: 437-444, 1963
69
23. Halstcd JA, Ronaghy HA,Abadi P, Haghshenass M,Amirhakin:ü GH, 39. Wang H, Prasad AS, DuMouehelle E: Zine in plate1ets, 1yrnphoeytes,
Barakat RM, Reinhold SG: Zinc deficiency in man: the Shiraz and granuloeytes by flameless atomie absorption spectrophotometry.
experiment. Am J Med 53: 277-284, 1972 J Micronutrient Anal5: 181-190, 1989
24. Sandstead HH, Prasad AS, Schulert AR, Farid Z, Miale A, Bassily S, 40. Meftah S, Prasad AS, Lee DY, Brcwer GJ: Ecto nucJeotidase (5 'NT)
Darby WJ: Human zinc deficiency, endocrine manifestations and as a sensitive indieator ofhuman zine deficiency. J Lab Clin Med 118:
response to treatment. Am J Clin Nutr 20: 422--442, 1967 309--316,1991
25. Caggiano V, Schnitzier R, Strauss W, Baker RK, Carter AC, Josephson 41. Prasad AS, Meftah S, Abdallah J, Kaplan J, Brewer GJ, Bach JF,
AS, Wallech S: Zine deficieney in a patient with retarded growth, Dardenne M: Serum thymulin in human zine defieiency. J Clin Invest
hypogonadism, hypogammaglobulinemia, and ehronie infeetion. Am 82: 1202-1210, 1988
J Med Sei 257: 305-319,1969 42. Dardenne M, Pleau JM, Nabarra B, Le francier P, Derrien M, Choay
26. Hambidge KM, Hambidge C, Jacobs M, Baum JD: Low levels ofzinc J, Bach JF: Contribution of zinc and other metals to the biologieal
in hair, anorexia, poor growth, and hypogeusia in children. Pediatr aetivity ofthe serum thymie factor. Proe NatlAcad Sci USA 79: 5370--
Res 6: 868--874, 1972 5373,1982
27. MacMahon RA, Parker ML, MeKinnon M: Zine treatment in mal- 43. Pleau JM, Fuentes V, Morgat JL, Bach JF: Speeific receptor for the
absorption. Med JAust 2: 210--212, 1968 serum thymic factor (FTS) in Iymphoblastoid cultured celliines. Proe
28. McClain CJ, Adams L, Shedlofsky S: Zine and the gastrointestinal NatlAead Sei USA 77: 2861-2865,1980
system. In: Prasad AS (ed). Essential and Toxie Trace Elements in 44. Tapazoglou E, Prasad AS, Hill G, Brewer GJ, Kaplan J: Deereased
Human Health and Disease. New York: Alan R Liss, pp 55-73, 1988 natural killer eell aetivity in zine defieient subjeets with siekle cell
29. Barnes PM, Moynahan EJ: Zinc deficiency in acrodermatitis entero- disease. J Lab Clin Med 105: 19-22, 1985
pathica: multiple dietary intolerance treated with synthetic zinc. Proc 45. Beck FWJ, PrasadAS, Kaplan J, Fitzgera1d JT, Brewer GJ: Changes
R Soc Med 66: 327-329, 1973 in cytokine produetion and T cell subpopulations in experimentally
30. Kay RG, Tasman-Jones C: Zinc defieieney and intravenous feeding. induced zine-defieient humans.Am J PhysioI272: EI 002-E I 007, 1997
Laneet 2: 605--606, 1975 46. Prasad AS, Beek FWJ, Grabowski SM, Kaplan J, Mathog RH: Zine
31. OkadaA, Takagi Y, Itakura T, Satani M, Manabe H, Iida Y, Tanigaki T, defieieney: ehanges in eytokine production and T-cell subpopulations
Iwasaki M, Kasaham N: Skiniesions during intravenous hyper- in patients with head and neck cancer in non-cancer subjects. Proe
alimentation: zinc deficieney. Surgery 80: 629--635, 1976 Assn Am Phys 109: 68-77, 1997
32. Arakawa T, Tamura T, IgarashiY: Zine defieieney in two infants during 47. Beck FWJ, Kaplan J, Fine N, Handschu W, Prasad AS: Deereased
parenteral alimentation for diarrhea.Am J Clin Nutr 29: 197-204, 1976 expression of CD73 (eeto-5' -nuc1eotidase in the CD8+ subset is
33. Klingberg WG, Prasad AS, Oberleas D: Zine defieiency following assoeiated with zine defieieney in human patients. J Lab Clin Med
penicillamine therapy. In: PrasadAS (ed). Traee Elements in Human 130: 147-156, 1997
Health and Disease. Voll New York: Aeademie Press, pp 51--65, 48. Clerici M, Shearer GM: A TH---.TH2 switeh is a eritiea1 step in the
1976 etiology ofHiV infeetion. Immunol Today 14: 107-111, 1993
34. Prasad AS, Schoomaker EB, Ortega J, Brewer GJ, Oberlease D, 49. Foster K: TH1ITh2 eytokine responses, and disease association.
Oelshlegel FJ: Zine deficieney in siekle eell disease. Clin Chem 21: Immunovations 7: 1--8, 1995
582-587, 1975 50. Mosmann TR, Coffman RL: THI and TH2 cells: Different patterns of
35. Prasad AS, Abbasi AA, Rabbani P, DuMouehelle E: Effeet of zine lyrnphokine secretion lead to different funetiona1 properties. Annu Rev
supplementation on serum testosterone level in adult male siek1e eell Immunol7: 145-173, 1989
anemia subjeets.AmJHematoI19: 119--127,1981 51. MacKalI CL, Fleisher TA, Brown MR, Andrich MP, Chen CC,
36. Prasad AS, Cossaek ZT: Zine supplementation and growth in siekle Feuerstien IM, Horowitz MD, Magrath IT, Shad AT, Steinberg SM,
eell disease. Ann Intern Med 100: 367-371, 1984 Wexler LH, Gress RE: Age, thymopoiesis, and CD4+ T-lyrnphocyte
37. Warth JA, Prasad AS, Zwas F, Frank RN: Abnormal dark adaptation regeneration after intensive ehemotherapy. N Engl J Med 332: 143-
in siekle cell anemia. J Lab Clin Med 98: 189--194,1981 149,1995
38. Rabbani PI, PrasadAS, Tsai R, Harland BF, Fox MRS: Dietary model 52. Prasad AS, Beck FWJ, Endre L, Handschu W, Kukuruga M, Kumar
for production of experimental zine deficiency in man. Am J Clin Nutr G: Zinc deficiency affeets cell cycle and deoxythmidmic kinase (TK)
45: 1514--1525, 1987 gene expression in HUT-78 cell. J Lab Clin Med 128: 51--60, 1996
PART III

DIABETES
Molecular and Cellular Bioclzemistry 188: 73--80, 1998.
© 1998 Kluwer Academic Publislzers.

Vanadium and diabetes


Patrick Poucheret,l Subodh Verma,l Mare D. Grynpas2 and lohn H.
MeNeill l
IFaculty of Pharmaceutical Sciences, The University of British Columbia, Vancouver; 2Samuel Lunenfeld Research
Institute, Mount Sinai Hospital, Toronto, Canada

Abstract
We demonstrated in 1985 that vanadium administered in the drinking water to streptozotocin (STZ) diabetic rats restored elevated
blood glucose to normal. Subsequent studies have shown that vanadyl sulfate can lower elevated blood glucose, cholesterol
and triglycerides in a variety of diabetic models including the STZ diabetic rat, the Zucker fatty rat and the Zucker diabetic
fatty rat. Long-term studies ofup to one year did not show toxicity in control or STZ rats administered vanadyl sulfate in doses
that lowered elevated blood glucose. In the BB diabetic rat, a model of insulin-dependent diabetes, vanadyl sulfate lowered the
insulin requirement by up to 75%. Vanadyl sulfate is effective orally when administered by either single dose or chronic doses.
It is also effective by the intraperitoneal route. We have also been able to demonstrate marked long-term effects ofvanadyl
sulfate in diabetic animals following treatment and withdrawal ofvanadyl sulfate. Because vanadyl sulfate is not weil absorbed
we have synthesized and tested a number of organic vanadium compounds. One ofthese, bismaltolato-oxovanadium IV (BMOV),
has shown promise as a therapeutic agent. BMOV is 2-3x more potent than vanadyl sulfate and has shown less toxicity. Recent
studies from OUf laboratory have shown that the effects of vanadium are not due to a decrease in food intake and that while
vanadium is deposited in bone it does not appear to affect bone strength or architecture. The mechanism of action of vanadium
is currently under investigation. Several studies indicate that vanadium is a phosphatase inhibitor and that vanadium can activate
serine/threonine kinases distal to the insulin receptor presumably by preventing dephosphorylation due to inhibition of
phosphatases Short-term clinical trials using inorganic vanadium compounds in diabetic patients have been promising. (Mol
Cell Biochem 188: 73-80, 1998)

Key words: vanadium, diabetes, glucose lowering, insulin-mimetic

Introduction treatment of diabetes?'. Key aspects ofthe findings with this


unique trace element in relation to its insulin-mimetic,
As we approach the 21 st century, diabetes research has never antihyperglycemic and antihypertensive effects are reviewed
been so intense. Along with important improvements in the in the following pages.
scientific tools and techniques, deep insights into the complex
inter-relationship between insulin action, insulin resistance,
lipid and carbohydrate metabolism have been made. Twelve Background
years ago, a very surprising discovery was made in vivo
demonstrating that vanadium can enhance and/or mimic the Vanadium was discovered in 1813 by the mineralogist Dei
physiological effects of insulin in an experimental model of Rio who gave it the name panchromium because of its color
diabetes [1]. Since this first observation, a great deal of work changes as a function of its oxidation state. This transitional
has been done and recently data demonstrating beneficial element was then rediscovered in 1831 by the Swedish
effects ofvanadium in humans have been published leading chemist Nils Gabriel Sefstom who named the compound
to the question, 'could vanadium be a useful adjunct in the vanadis, a nickname ofthe Germanic goddess ofbeauty. In

Addressfor offprints: J.H. McNeill, Faculty ofPharmaceutica1 Seiences, The University ofBritish Columbia, 2146 East Mall, Vaneouver, B.C. V6T IZ3,
Canada
74

common with most transitional metals, vanadium exists in Table 1. In vivo effeets ofvanadium in STZ-diabetes
several valence states (-3, -1,0, + 1 to +5) and the expression
Amelioration of insulin resistanee refleeted by a greater glucose
of a given form is highly pH dependent. In biological systems lowering effeet of vanadium treated rats to insulin [25].
vanadium is found predominantly as the vanadate (+5) and Normalization ofboth basal and stimulated hepatie glucose production
vanadyl (+4) forms. In plasma, vanadium exists in both by chronic vanadium administration [59).
oxidation states. Approximately 90% is bound to proteins Enhanced insulin sensitivity in vanadium treated rats correlates with
(predominantly transferrin)[2]. Most ingested vanadium is restoration of insulin stimulated MAP and S6 kinase activities in
skeletal musele [47).
transformed in the stomach to V02+ and remains in this form Chronic vanadium treatment eorreets abnormalilies in glycolytic
as it passes through the duodenum. Vanadium has been shown enzymes i.e. phosphofructokinase-2 and glucokinase [60).
to be stored in the bone (main storage depot), kidney and liver Restoration of glycogen synthase and phosphorylase activies [61).
following i.p. injection [3]. In humans, the total body pool Aberrations in the tissue specific expression of2 isoforms ofglucose
of vanadium is estimated to be 100--200 j.1g [4]. transporter in STZ-diabetes are normalized by vanadium treatment
[62].
Amelioration of oxidative stress [63].
Long term effects on glucose metabolism following oral treatment
In-vitro insulin mimetic actions of und withdrawal [24).
Prevention of secondary cardiac eomplications such as cardiomyopathy
vanadium and eataracts [1, 23, 27].
Prevents long-term secondary complications
Decreases elevated triglyceride and eholesterollevels [23, 27]
In vitra, the insulin-like effects of vanadium extend to
Prevents plasma elevations ofurea and creatinine [29]
numerous processes involved in carbohydrate, lipid and Restores thyroid hormone levels to normal [23, 27]
protein metabolism: (i) carbohydrate metabolism through
glucose transport, glucose transport translocation, glycolysis
and glycogen synthesis [5-11]; (ii) lipid metabolism primarily
by inhibition of lipolysis, and (iii) protein metabolism and corrected plasma glucose, lipids, creatinine and thyroid
mitogenesis [12-16]. hormone blood levels. These improvement were accompanied
by correction of abnormalities in isolated heart function and
glycerol output from adipose tissue [22].
Ramanadham et al. in 1989 [23] published a very in-
In-vivo insulin mimetic actions of teresting result involving vanadyl sulfate. In the study, STZ
vanadium diabetic animals were orally treated for 3 weeks with
vanadyl sulfate and then the treatment was withdrawn for
The first report of vanadium's insulin mimetic and anti- 13 weeks. At the end of this period the authors recorded a
diabetic potential in viva was published by Heyliger et al. sustained prevention of myocardial and metabolic aberra-
[1]. This study was performed on streptozotocin (STZ) tions, along with normal blood levels of glucose, insulin,
diabetic rats chronically treated with sodium orthovanadatc. lipid and thyroid hormones. In addition, corrected glycerol
Normalization of hyperglycemia and improvement of output from adipose tissue and no evidence of cataract
cardiac depressed function were both recorded without an development could be observed. Together, these findings
increase in plasma insulin levels in the diabetic rats. This strongly suggested that vanadium could produce a sustained
observation demonstrated the ability ofvanadium to improve insulin-like effect on these animals despite the fact that
insulin sensitivity. Hyperinsulinemic clamp studies later vanadium was no longer present in the body. These animals
confirmed a decreased insulin resistance fol1owing vanadium were not completely normal in that achallenge with glucose
treatment [17]. In 1987, Meyerovitch showed that in did not produce further insulin release and resulted in an
addition to lowering plasma glucose levels, chronic sodium abnormal glucose tolerance curve.
metavanadate administration could also enhance basal In order to better understand the basis of the antidiabetic
hexose transport in both liver and muscle [18]. Brichard et al. effects of vanadyl sulfate, Cam et al. conducted an experiment
subsequently described a dose-response relationship between to test the hypothesis of a putative prophylactic action of
vanadate and its glucose lowering effects [19]. A summary vanadium against the cytotoxic destruction ofthe pancreatic
ofthe most prominent effects ofvanadium in STZ diabetes beta ceHs by STZ [24]. Diabetes was induced by STZ
is shown in Table l. injection and vanadyl sulfate treatment started 3, 10 or 17
Since vanadyl sulfate was reported to be 6-10 times less days later and lasting for a 5 month period. Irrespective of
toxic than vanadate [20] this form of vanadium was extensively the delay separating diabetes induction and beginning ofthe
investigated for its insulin-like cffccts. STZ-diabetic animals treatment, parameters such as glucose tolerance and adipose
responded to vanadyl sulfate givcn in drinking water with tissue function were normalized in diabetic-treated animals.
75

These observations are therefore not in favor ofthe concept be administered orally and absorbed by passive diffusion. It
that vanadium efficaey as an insulin mimetie is due to a is water soluble, electrically neutral and has a low molecular
proteetive effeet ofthe endoerine beta eells from the deleterious weight [26, 27]. BMOV and vanadyl sulfate have been
effects of STZ. compared for their effects by both oral administration and
Concentration dependent effects and the in vivo interaction intraperitoneal (i.p.) administration using a single dose [28].
ofvanadyl with insulin was studied by Battell et al. [21] and The ED so following oral administration indicated that the
Ramanadham et al. [25]. Using the diabetic BB rat, a spon- ED so for BMOV (0.5 mmollkg) was twice as potent as
taneous Type I model of diabetes which produces no insulin, vanadyl sulfate (ED so - 0.92 mmollkg). Following i.p.
vanadyl sulfate produeed dose-dependent effeets which administration, BMOV was 3 times more potent than vanadyl
reduced the exogenous insulin requirement necessary to sulfate. In addition, it was interesting to note that the highest
prevent glycosuria by 75%. Two important findings were BMOV dose produced euglycemia in 100% of the treated
obtained from this set of results: (1) vanadium requires the animals whereas vanadyl sulfate produced the effect in only
presence of so me insulin to produce its in vivo effects, 80-90% ofthe animals.
therefore, (2) vanadium in vivo is an insulin enhancer [21,25]. Chronic BMOV treatment (0.75 mg /ml in drinking water)
Vanadium effects have also been demonstrated in other over aperiod of 6 months in STZ-diabetic rats restored
models of both Type I and Type 11 diabetes and these are plasma glucose levels to normal (8/12 animals) as weil as
summarized in Table 2. heart function in all diabetic-treated rats. However, BMOV
did not affect body weight gain in control rats for the first
10 weeks when compared to vanadyl sulfate [28]. This last
The use of organic vanadium complexes observation is important regarding some concems expressed
in the literature about vanadium toxicity. Dai et al. in our
to increase potency laboratory evaluated the effects of long-term BMOV and
vanadyl sulfate treatments on several pathological deter-
In order to overeome poor absorption of vanadate and minants of STZ-induced diabetes [29]. Chronic BMOV
vanadyl from the gastrointestinal (GI) tract and GI toxicity, treatment completely prevented elevations in plasma urea,
i.e. diarrhea, our laboratory and others have synthesized creatinine, alanine aminotransferase (ALT) and improved
various organic vanadium compounds which were designed histological abnormalities in the kidney and liver from STZ-
to improve absorption, potency and therapeutic safety. diabetic rats.
Bis(maltolato)oxovanadium (IV) (BMOV), a maltoll BMOV was also uscd to assess organic vanadium cftcctive-
vanadyl compound, was developed in collaboration with Dr. ness in a Type II model of diabetes, the fa/fa Zucker rat. BMOV
e. Orvig in the Department of Chemistry at the University at a maximal concentration of 0.5 mg/mi for 14 weeks of
ofBritish Columbia, Vancouver, Canada [26]. BMOV is an treatment did not affect body weight gain in lean controls but
example of aseries of compounds specifically designed to did significantly reduce body weight gain in the fatty-treated
group. At this concentration, BMOV also significantly
reduced plasma insulin levels. A lower concentration ofO.2
Table 2. In vivo effects ofvanadium in other models of diabetes mg/mI BMOV did not affect food or fluid intake and did not
Spontaneouslv Diabetic BB rat
decrease body weight or plasma cholesterollevels in fatty-
Reduces the dose of insulin required by 75% [21,25] treated animals. The lower dose also significantly reduced
Partiallv Pancreatectomized Rat plasma, insulin and triglyceride levels. Oral glucose tolerance
Improves insulin-sensitivity towards peripheral glucose uptake in 90% was also improved by BMOV trcatment in thc fatty animals
pancreatectomized rats predominantly through correction of glycogen
even at this lower dose [30].
synthesis [64]
Neonatal STZ-Diabetic Rats
Vanadium treatment corrects basal and stimulated hepatic and glucose
production and peripheral glucose utilization [65]
Genetically obese fa/fa (Zucker) Rats
Vanadium compounds ameliorate
Attcnuatcs hyperinsulinemia and impaired glucose tolerance [30, 66] insulin resistance hyperinsulinemia and
Lowers blood lipids, dose dependent decrease in weight gain [30]
Obese ob/ob Mice
hypertension
Attenuates hyperglycemia, improves glucose tolerance and hepatic
glycogen content. Prevenls pancreatic exhaustion of insulin [67] Extensive epidemiologieal, clinical and experimental data lend
Zucker Diabetic Fatty Rats
Attenuates hyperglycemia, hyperinsulinemia and hyperlipidemia.
credence to the association between essential hypertension and
Restores glucosc tolerancc and decreases pancreatic insulin abnormalities in carbohydrate and lipid metabolism [31-
depletion (Yuen and McNeill, unpublished observations). 33]. Hyperinsulinemia and insulin resistance are glucose
76

metabolism related defects frequently found associated with observed while no IR kinase activity change could be
hypertension. They are also linked to a high atherogenic risk recorded [43 J. These last two observations and others [44,
profile, dyslipidemia and atherosclerosis. It would then be 45J are in favor of potential post -receptor effects of vanadium,
expected that drugs interventions that corrected these defects further downstream in the insulin signaling cascade.
would also decrease blood pressure if there is a correlation Knowing that intracellular vanadium is primarily in the
between the events. In order to test the hypothesis of a vanadyl form (which is not a potent phosphatase inhibitor),
potential relationship between hyperinsulinemia, insulin it is reasonable to speculate that other signaling molecules
resistance and hypertension, we undertook aseries of in the insulin signaling pathways might be targets for
experiments, using vanadyl sulfate and BMOV. The study vanadium action. In addition, the effects of vanadium on
was conducted in both a genetic and an acquired model of intracellular calcium flux, intracellular and intravesicular
hypertension, respectively: the spontaneously hypertensive pR should not be underestimated in studying the insulin-
rat (8RR rat) and the fructose hypertensive rat. Vanadium mimetic effects of vanadium.
compounds reduced plasma insulin levels and blood pressure A study was conducted in intact rat adipocytes where
in both types of animals. Moreover, administration of vanadium was shown to activate a staurosporine sensitive
exogenous insulin to match the level ofthe untreated animals cytosolic protein tyrosine kinase (Cyt PTK) distinct from IR
reversed the beneficial effects ofvanadium on blood pressure. tyrosine kinase. This activation was linked to glucose
These results strengthen the hypothesis of the link between oxidation and lipid synthesis but dissociated from glucose
hyperinsulinemia/insulin resistance and hypertension and uptake and inhibition of lipolysis. Cyt PTK would then be
demonstrate the antihypertensive potential of vanadium in implicated only in specific cellular response. Furthermore,
vivo [17, 34--36J. Cyt PTK would be highly selective for vanadium since
neither insulin, isoproterenol, dibutyryl cAMP, okadaic acid,
hydrogen peroxide or phorbol ester TPA did affect Cyt PTK
Mechanism of action activity. In addition to vanadium, other PTPase inhibitors
have also been shown to activate Cyt PTK in adipocytes. It
The mechanism of action of vanadium in producing its should be noted that insulin mimetic effects ofvanadium on
antidiabetic effects in vivo is poorly understood and is hexose uptake and inhibition of lipolysis are not blocked by
currently the subject of much investigation. In vitro and in staurosporine (a blocker of Cyt PTK) indicating that this
vivo data demonstrate that vanadium does affect various pathway is dehmitely not the only means by which vanadium
aspects ofthe insulin signaling pathway (Table 3). influences cellular physiology [46].
It was postulated that vanadium' s insulin mimetic effects Insulin signal transduction is mediated intracellularly
would be the result of vanadium behaving as a phosphate through a complex network of cascades of reversible protein
analog and stimulating pro tein tyrosine phosphorylation phosphorylations and dephosphorylations. Among the
numerous kineses involved in insulin signaling, MAP and 86
through inhibition of protein tyrosine phosphatases (PTPases)
[37, 38J. Thus, it was not surprising to re cord vanadium kinases have been demonstrated to be defective in both the
induction of autophosphorylation of solubilized insulin basal and insulin stimulated state in 8TZ-diabetic rats. We
receptor (IR) in a fashion analogous to insulin [39, 40J with found that chronic vanadium treatment corrected the insulin-
stimulation of the tyrosine kinase activity of the IR beta induced activation of these kinases [47, 48]. Thus, insulin
subunit [41]. Other studies however demonstrated that resistance associated with long-term diabetes may be linked
vanadium was equally effective in stimulating glucose with altered signaling through these kinases and vanadium
metabolism in rat fat cells when half the IR had been could rectify the observed defects. Preliminary results trom
inactivated by insulin over stimulation [42]. Furthermore, recent experiments in our laboratory indicate that the picture
glucose-Iowering effects with oral vanadium treatment were may be more complex regarding the relative importance of
different kinases as potential vanadium targets in order to
explain its insulin mimetic effects and its ability to correct
insulin resistance.
Table 3. Suggested effeets ofvanadium on insulin-signaling pathways in
The long-term effects ofvanadium treatment were further
vitra
investigated in our laboratory. Since a persistent euglycemic
Stimulates autophosphorylation of insulin receptars [39, 40] state can be observed following vanadium treatment with-
Increases insulin receptor tyrosine kinase activity [41] drawal with only minor improvements in pancreatic secretory
Stimulates down regulation of insulin receptors [46] function, several hypothesis have been proposed. The
Increases insulin receplor binding [46]
Increases protein tyrosine kinase activily [44]
vanadium-treated rats could sustain an increased sensitivity
lncreases Ser/Thr protein kinase activity [47, 48] to circulating insulin even after the treatment has been
Inhibits phospho~yrosine phosphatase activity [37, 38] stopped. A second possibility could be that vanadium is
77

released from potential tissue storage sites producing the anti- attributable to vanadium alone or to food restriction alone.
hyperglycemic effects although this seems highly unlikely. Two main parameters were used to assess these influences:
Alternatively, Cametal. [49] suggestthatvanadium-induced plasma glucose levels and lipid levels [55]. STZ-diabetic rats
amelioration of the diabetic state may be partially due to were treated daily over a 6 week period with BMOV dissolved
preservation of a functional portion of pancreatic beta cells in drinking water. Pair-fed groups were fed basedon the intake
in the STZ animals. This study showed that a modest increase of their respective counterparts from the previous day.
in ß cell content was crucial to the long-term effect of Decreases in plasma glucose, triglycerides and cholesterol
vanadium even though the total insulin content was still much levels in diabetic-treated rats were recorded with no effect
less than normal. It remains clear that the absence ofnorrnal on the plasma insulin level. None ofthese parameters were
plasma insulin levels strongly suggests the presence of affected in pair-fed animals. In addition, prevention of cardiac
additional actions ofvanadium, perhaps at the level ofinsulin function impairment was observed in STZ-diabetic rats
sensitive tissues [49]. treated with BMOV but not in pair-fed diabetic animals. The
experimental design used in Malabu ' s study may explain the
different results since food was given 10 pair-fed animals only
Clinical studies once a day. Our observation is that hyperphagic diabetic
animals consume this small amount of food in a very short
Some very elegant studies have recently been conducted on period oftime after provided. The animals are therefore left
both Type I and Type II human diabetic subjects. Sodium fasting for a long period oftime since blood was not collected
metavanadate was administered for 2 weeks at 125 mg daily until the morning. This factor is crucial since reduction in
in divided doses. In Type I diabetic patients, vanadium plasma glucose levels in their study for pair-fed diabetic groups
lowered insulin requirements without an effect on C-peptide was similar to what we have observed after a prolonged (20
levels suggesting the absence of an influence on insulin h) period offasting.
release. In addition, 2 out ofthe 5 insulin-dependent subjects
showed improved glucose utilization. In Type II diabetic
patients, improved insulin sensitivity, enhancement ofnon- Vanadium in bone
oxidative glucose disposal rates and higher basal MAP and
S6 kinases activity in monocytes were recorded. Hepatic Concern has been raised by the fact that vanadium deposits
glucose production was not affected. Diarrhea was the main in bone and could be potentially toxic [56]. Indeed studies
have shown that chronic vanadium administration to rats
side effect observed [50].
Vanadyl sulfate was also studied in 6 non-insulin-dependent results in bone concentrations ofvanadium of 10-26 jlg/g
diabetic patients at a dose ofl 00 mg!day for 3 weeks. Reduced [57, 58]. These concentrations were twice that of the
fasting plasma glucose and HbAic were recorded without an accumulation in kidney and 6-1 0 times higher than those found
effect on plasma insulin levels. An interesting finding was in liver [57, 58]. In order to test the effects ofvanadium on bone
the persistence of improved insulin sensitivity for up to 2 strength, we examined the tibia and vertebrae from Zucker
weeks following cessation of the treatment [51]. This last Diabetic Fatty rats treated with BMOV at concentrations in
observation agrees with experimental studies described earlier. the drinking water beginning with 0.2 mg/mi increasing
Additional recent reports tend to validate the observation of incrementally to 0.8 mg/mi over a 10 week period. The
beneficial effects of vanadyl and vanadate treatment in non- treatment with BMOV improved blood glucose in the animals
insulin-dependent human subjects [52, 53]. from 28.7 mM to 13.3 mM and modestly improved plasma
triglyceride levels. Polydypsia was improved and there was
a slight (8%) decrease in body weight. Vanadium content
oftreated rats was 9.42 ± 1.72 (S.D.) jlg/g fortibia and 6.60
Vanadium glucose lowering effects and ± 1.1 0 jlg!g for vertebrae, the difference is likely due to the
food restriction fact that vertebrae have a lower content of bone than tibia.
This is evident ifwe consider the vanadium/phosphate ratio
Vanadium compounds have now been extensively demon- (0.79 ± 0.15 fortibia, 1.07 ± 0.19 forvertebrae) is higher for
strated to normalize the hyperphagia associated with experi- vertebrae thus reflecting the higher turnover of trabecular
mental diabetes. In 1994, Malabu et al. stated some concerns bone (vertebrae). Vanadium treatment did not affect the
claiming that the decrease in plasma glucose levels observed content of other minerals in the bone (K, Mg, Na, Ca, P)
after vanadate administration was entirely attributable to a nor did it affect bone crystal size as determined by X-ray
reduction in food intake [54]. diffraction. The mechanical properties of the tibia as
In our laboratory, Yuen et al. conducted a detailed study reflected by the 3-point bending test and of vertebrae as
on STZ-diabetic rats to precisely defme the respective effects determined by compression testing were not affected by
78

vanadium administration. Finally image analysis of a stained References


thin section of vertebrae showed no changes in bone
architecture in vertebrae of rats treated with BMOV. Thus I. Heyliger CE, TahilianiAG, McNeiJI JH: Effect ofvanadate on elevated
the architecture, density and mechanical properties ofbone blood glucose and depressed cardiac performance of diabetic rats.
were not affected by the treatment with vanadium in doses Science 227: 1474-1477, 1985
2. Nechay BR: Mechanism ofaction ofvanadium. Ann Rev Pharmacol
which produced insulin-like effects. Further studies over Toxicol24: 501-524,1984
longer periods oftime will be required to fully determine the 3. Talvite NA, Wagner WD: Studies in vanadium toxicology. Arch Ind
effects of vanadium on bone. Hyg Occup Med 9: 414--422, 1954
4. Byme AR, Kosta L: Vanadium in food and human body fluids and
tissues. Sei Total Environ 10: 17-30, 1993
5. Nakai M, Watanabe H, Fujiwara C, Kakegawa H, Satoh T, Takeda J,
Conclusion Matsushita R, Sakurai H: Mechanism ofinsulin-like action ofvanadyl
sulfate: studies on interaetion between rat adipocytes and vanadium
compounds. J Pharm Soc Jap 18: 119-125,1995
Since our initial demonstration of the anti-diabetic effects of
6. Duckworth WC, Solomon SS, Liepneks J, Hamel FG, Hand S, Peavy
vanadium in vivo, significant advances have been made in DE: Insulin like effects of vanadate in isolated rat adipocytes.
understanding the glucose-lowering properties and the Endocrinology 122: 2285-2289, 1988
mechanism(s) ofaction ofvanadium compounds. The exact 7. Schechter Y, Karlish SJD: Insulin-like stimulation of glucose
intracellular mechanism and/or mediators involved in oxidation in rat adipocytes by vanadyl (IV) ions. Nature 284: 556-
558, 1980
vanadium actions remain unknown but vanadium effects may
8. Tamura S, Brown TA, WhippleJH, Fujita-Yamaguchi Y, Dubler RE,
be mediated by a synergy between several post-receptor Cheng K, Lamer J: A novel mechanism far the insulin-like effect of
events of the insulin signaling cascade in the target tissues vanadate on glycogen synthase in rat adipocytes. J Biol Chem 259:
of the hormone. Design and development of organic ligands 6650-6658, 1984
with improved absorption, tissue uptake, potency, and having 9. Tamura S, Brown TA, Dubler RE, Lamer J: Insulin-Iike effect of
vanadate on adipocytc glycogcn synthasc and on phosphorylation of
decreased toxicity is important. BMOV exemplifies one 95,000 Dalton subunit ofinsulin receptor. Biochem Biophys Res Comm
such organically chelated complex that appears to be a 113: 8042-8048, 1983
potent insulin-mimetic and insulin enhancer. BMOV demon- 10. McNeill JH, Yuen VG, Dai S, Orvig C: Increased potency ofvanadium
strates less gastrointestinal side effects and does not affect using organie ligands. Mol Cell Biochem 153: 175-180, 1995
body weight gain and food and fluid intake in control- 11. Orvig C, Thompson KH, Battell M, McNeill JH: Vanadium compounds
as insulin mimics in 'Metal Ions in Biologieal Systems' [no H. Sigel
treated rats. and A. Sigel (eds). Marcel Dekker, Inc, 1995
Vanadium research demonstrating the antihypertensive 12. Jackson T, Salhanick AL, Sparks JD, Sparks CE, Bolognino M,
effects of vanadium compounds in hyperinsulinemic and Amatrude JM: Insulin-mimetic effects ofvanadate in primary cultures
insulin-resistant models of hypertension mayaiso prove to ohat hepatocytes. Diabetes 37: 1234-1240, 1988
13. Morita T, Imagawa T, Kanagawa A, Ueki H: Sodium orthovanadate
be important. Early trials with vanadium in diabetic human
increases phospholipase A2 activity in isolated rat fat pats: A role of
volunteers have shown promising results consistent with phospholipase A2 in the vanadate stimulated release of lipoprotein
experimental studies. Within the next few years the possible lipase activity. Biol Pharm Bull [8: 347-349, 1995
therapeutic roles of vanadium should be more clearly 14. Maher PA: Stimulation of endothelial eell proliferation by vanadate
established. specific for microvascular endothelial eells. J Cell Physiol 151: 549-
554,1992
15. Bames DM, Sykes DB, Schechter Y, Miller DS: Multiple sites of
vanadate and peroxovanadate action in xenopus oocytes. J Cell Physiol
Acknowledgements 162: 154-161, 1995
16. Hajjar JJ, Fucci Je, Rowe WA, Tomieie TK: Elfect of vanadate on
amino aeid transport in rat jejunum. Proc Natl Acad Sei USA 184:
Studies quoted in this paper from our laboratory have been 403-409, 1987
supported by the Canadian Diabetes Association, the Medical 17. Bhanot S, Bryer-Ash M, Cheung A, McNeill JH: Bis(maltolato)
Research Council ofCanada, the Heart and Stroke Foundation oxovanadium (IV) attenuates hyperinsulinemia and hypertension in
ofB.C. andYukon, and the Natural Sciences andEngineering spontaneously hypertensive rats. Diabetes 43: 857-861, 1994
18. Meyerovitch J, Farfel Z, Sack J, Schechter Y: Oral administration of
Research Council of Canada. The studies on bone were vanadate normalizes blood glucose levels in streptozotocin treated rats.
supported by a grant from the UBC University-Industry J Biol Chem 262: 6658-6662, 1987
Office. The administrative and technical assistance ofMary 19. Brichard SM, Okitolonda W, Henquin JC: Long term improvement
Battell and Erika Vera is greatly appreciated. We thank Sylvia of glucose homeostasis by vanadate treatment in diabetic rats.
Chan far expert secretarial assistance. Patrick Poucheret was Endocrinology 123: 2048--2053, 1988
20. Hudson TGF: Vanadium, Toxicology and Biological Significance. New
supported by a Canadian Diabetes Association Fellowship,
York, Elsevier, 1964
Subodh Verma by a Fellowship from the Medical Research 21. Battell ML, Yuen VG, McNeill JH: Treatment ofBB rats with vanady1
Council of Canada. sulfate. Pharmacol Commun I: 291-301,1992
79

22. Ramanadham S, Mongold H, Brownsey RW, Cros GH, MeNeill JH: 42. Green A: The insulin like effeet of sodium vanadate on adipoeyte
Oral vanadyl sulfate treatment of diabetes mellitus in rats. Am J Physiol transport is mediated at a post insulin reeeptor level. Bioehern J 238:
257: H904-H911, 1989 663-{j69, 1993
23. Ramanadham S, Brownsey RW, Cros GH, Mongold JJ, MeNeill JH: 43. Mooney RA, BordweIl KL, Luhowskyj S, Casnelle JE: The insulin
Sustained prevention of myoeardial abnormalities in diabetie rats like effeet of vanadate on lipolysis in rat adipoeytes is not accompanied
following withdrawal from oral vanadyl treatment. Metabolism 38: by an insulin like effect on tyrosine phosphorylation. Endoerinology
1022-1028, 1989 124: 422-429, 1989
24. Cam MC, Pederson RA, Brownsey RW, MeNeill JH: Long term 44. ShishevaA, Sheehter Y: Role of eytosolie tyrosinc kinase in mediating
effeetiveness of oral vanadyl sulfate in streptozotoein-diabetie rats. the insulin like aetions of vanadate in rat adipoeytes. J Biol Chem
Diabetologia 36: 218-224, 1993 268: 6463-{j469, 1993
25. Ramanadham S, Cros GH, Mongold H, Serrano H, MeNeill JH: 45. SheehterY, ShishcvaA, LazarR, LibmanJ, ShanzerA: Hydrophobie
Enhaneedin vivo sensitivity ofvanadyl treated diabetie rats to insulin. carriers ofvanadyl ions augmentthe insulinomimetie aetions ofvanadyl
Can J Physiol Pharmacol68: 486-491,1990 ions in rat adipoeytes. Bioehemistry 31: 2063-2068, 1982
26. McNeill JH, Yuen VG, Hoveyda HR, Orvig C: Bis(maltolato) 46. Sekar N, Li J, Seheehter Y: Vanadium salts as insulin substitutes:
oxovanadium (IV) is a potent insulin mimic. J Med Chem 35: 1489-- Meehanism of action, a seientific and therapeutie tool in diabetes mellitus
1491,1992 research. Critieal Rev Bioehern Mol Biol 31: 339--359, 1996
27. Yuen VG, Orvig C, MeNeill JH: Glucose lowering effeets of a new 47. Bei YJ, Chen X, Pclcch SL, Diamond J, McNeill JH: Skeletal museIe
organic vanadium eomplex bis(maltolato)oxovanadium (IV). Can J mitogen aetivated protein kinases and ribosomal S6 kinases suppression
Physiol Pharmacol 71: 263-269, 1993 in ehronie diabetes and reversal by vanadium. Diabetes 44: 1147-1155,
28. Yuen VG, Orvig C, MeNeill JH: Comparison ofthe glucose 10wering 1995
properties uf vanadyl sulfate and bis(maltolato )oxovanadium (IV) 48. Hei YJ, Diamond J, MeNeill JH: Distribution ofMAP kinase, S6 kinase
following acute and ehronie administration. Can J Physiol Pharmaeol and casein kinase 2 in rat tissues: activation by insulin in spleen.
73: 55--64, 1995 Bioehern Cell Biol 72: 49-53, 1994
29. Dai S, Yuen VG, Orvig C, MeNeill JH: Prevention of diabetes-indueed 49. Cam MC, Li WM, McNeill JH: Partial preservation ofpanereatie beta
pathology in STZ-diabetic rats by bis(maltolato)oxovanadium (IV). cells by vanadium: Evidenee for ehronie amelioration of diabetes.
Pharmaeol Communie 3: 311-321,1993 Metabolism 46: 1-11, 1997
30. Yuen VG, Pederson RA, Dai S, Orvig C, MeNeill JH: The effects of 50. Goldfine AB, Simonson DC, Folli F, Patti ME, Khan CR: Metabolie
low and high dose administration ofbis(maltolato)oxovanadium (IV) effeets of sodium mctavanadate in humans with insulin dependent and
on fa/fa Zucker rats. Can J Physiol Pharmaeol 74: 1001-1009, 1996 non insulin dependent diabetes mellitus in vivo and in vitra studies. J
31. Bhanot S, MeNeill JH: Insulin and hypertension: A causal relationship? Clin Endocrinol Metab 80: 3311-3320, 1995
Cardiovasc Res 31: 212-221,1996 51. Cohen N, Halberstarn M, Shlimovieh P, Chang CJ, Shamoon H,
32. DeFronzo RA, Ferrannini E: Insulin resistance: A multifaceted syndrome Rossetti L: Oral vanadyl sulfate improves hepatie and peripheral insulin
responsible for NIDDM, obesity, hypertension, dyslipidemia and sensitivity in patients with non insulin dependent diabetes mellitus. J
atherosclerotic eardiovascular disease. Diabetes Care 14: 173--194, 1991 Clin Inves! 95: 2501-2509,1995
33. Ferrannini E, Natali A: Essential hypertension, metabolie disorders 52. Boden G, Chen S, Ruiz J, George DV, Rossum V, Tureo S: Effects of
and insulin resistanee. Am Heart J 121: 1274-1282, 1991 vanadyl sulfate on earbohydrate and lipid metabolism in patients with
34. Bhanot S, Michoulas A, MeNeill JH: Antihypertensivc effcets of non insulin dependent diabetes mellitus. Metabolism 45: 1130--1135,
vanadium eompounds in hyperinsulinemie, hypertensive rats. Mole 1996
Cell Bioehern 153: 205-209,1995 53. Halberstarn M, Cohen N, Shlimovieh P, Rossetti L, Shamoon H: Oral
35. Bhanot S, McNeill JH: Vanadyl sulfate lowers plasma insulin levels vanadyl sulfate improves insulin sensitivity in NIDDM but not in obese
and blood pressure in spontaneollsly hypertensive rats. Hypertension non diabetie subjeets. Diabetes 45: 659-{j66, 1996
24: 625-{j32, 1994 54. Malabu US, Dryden S, MeCarthy HO, Kilpatriek A, Williams G:
36. Bhanot S, MeNeill JH, Bryer-Ash M: Vanadyl sulfate prevents fructose Effeets of ehronie vanadate administration in the STZ diabetic rat.
indueed hyperinsulinemia and hypertension in rats. Hypertension 23: The antihyperglyeemie action of vanadate is attributable entirely to
308--312, 1994 its suppression offeeding. Diabetes 43: 9--15, 1994
37. Swarup G, Cohen S, Garbers DI: Inhibition of membrane phospho- 55. Yucn VG, Orvig C, MeNeill JH: Effeets ofbis(maltolato) oxovanadium
tyrosyl-protein phosphatase aetivity by vanadate. BioehemBiophys Res (IV) are distinct from food restrietion in STZ-diabetie rats. Am J Physiol
Com 107: 1104-1109, 1982 272: E30--E35, 1997
38. Traeey AS, Gresser MJ: Interaction of vanadate with phenol and 56. MravoeaA, Jirova D, Janci Hand Lener J: Effeet of orally administered
tyrosine: implieations for the effeets ofvanadate on systems regulated vanadium on the immune system and bone metabolism in experimental
by tyrosine phosphorylation. Proe Nat! Aead Sei USA 83: 609-{i 13, animals. The Seienee oftheTotal Environment (Supplement), Elsevier
1986 Publishers, Amsterdam, 1993, pp 663-{j69
39. Ueno A, Arakaki N, Takeda Y, Fujio H: Inhibition of tyrosine 57. Yueo VG, Orvig C, Thompson KH, McNeill JH: Improvement in
autophosphorylation of the solubilized insulin reeeptor by an insulin cardiac dysfunetion in STZ-indueed diabetie rats following ehronie
stimulating peptide derived from bovine serum albumin. Bioehern oral administration ofbis(maltolato )oxovanadium (IV). Can J Physiol
Biophys Res Com 144: 11-18, 1987 Pharmaeol71: 270--276,1993
40. Gherzi R, Caratti G, Andraghetti G, Bertolini S, Montemurro A, Sesto 58. Mongold H, Cros GH, Vian L, Tep A, Ramanadham S, Siou G, Diaz
G, Cordera R: Direct modulation of insulin reeeptor proteiD tyrosine J, McNeill JB, Seffano JJ; Toxieologieal aspeets of vanadyl sulfate on
kinase by vanadate and antiinsulin reeeptor mODoelonal antibodies. diabetie rats - Effeets on vanadium levels and panereatie B-eell
Bioehern Biophys Res Com 152: 1474-1480,1988 morphology. Pharmaeol Toxieol67: 192-198, 1990
41. Smith DM, Sale GJ: Evidenee that a novel serine kinase eatalyses 59. Blondel 0, Bailbe D, Purtha B: In vivo insulin resistanee in
phosphorylation of the insulin reeeptor in an insulin dependent and streptozotoein-diabetie rats - evidenee for reversal following vanadate
tyrosine kinase dependent manner. Bioehern J 256: 903-909, 1988 treatment. Diabetologia 32: 185--190, 1989
80

60. Gil J, Miralpeix M, Carreras J, Bartrons R: Insulin like effects of 64. Rossetti L, McLaughlin MR: Correction ofchronic hyperglycemia by
vanadate on glucokinase activity and fructose 2,6 biphosphate levels vanadate but not with phlorizin, normalizes in vivo glycogen repletion
in livers ofdiabetic rats. J Biol Chem 263: 186&--1871,1988 and in vitro glycogen synthase activity in diabetic skeletal musc1e. J
61. Pugazhenthi S, Khandelwal R: Insulin like effects of vanadate on Clin Invest 84: 892--899, 1989
hepatic glycogen metabolism in non-diabetic and streptozotocin- 65. BlondelO, Simon J, Chevalier B, Portha B: Impaired insulin action
induced diabetic rats. Diabetes 39: 821--827, 1990 but normal insulin receptor activity in diabetic rat liver: Effect of
62. Brichard SM, Debuquois B, Girard J: Vanadate treatment of diabetic vanadate. Am J Physiol258: E459-E467, 1990
rals reverses the impaired expression of genes involved in hepatic glucose 66. Brichard SM, Pottier AM, Henquin Je: Lang term improvement of
metabolism: effects on glycolytic and gluconeogenic enzymes and on glucose homeostasis by vanadate in obese hyperinsulinemic fa/fa rats.
glucose transporter GLUT2. Mol Cell Endocrinol91: 91-97, 1993 Endocrinology 125: 2510--2516,1989
63. Thompson KH and McNeill JH: Effect ofvanadyl sulfate feeding on 67. Brichard SM, Bailey CJ, Henquin JC: Marked improvement af glucose
susceptibility to peroxidative change in diabetic rats. Res Comm Chem homeostasis in diabetic ob/ob mice given oral vanadate. Diabetes 39:
Path Pharmacol 80: 187-200, 1993 1326--1332,1990
Moleeular and Cellular Biochemistry 188: 81-89, 1998.
© 1998 Kluwer Academic Publishers.

Effect of enteral nutritional products differing in


carbohydrate and fat on indices of carbohydrate
and lipid metabolism in patients with NIDDM
Linda J. McCargar, 1 Sheila M. Innis,3 Elaine Bowron,2 Joseph
Leichter,2 Keith Dawson,4 Ellen Toth5 and Katherine WalP
IDepartment 0/ Agricultural Food & Nutritional Science, University 0/ Alberta; 2School 0/ Family and Nutritional
Sciences; 3Department 0/ Pediatrics; 4Department 0/ Medicine, University 0/ British Columbia and 5Department 0/
Medicine, University 0/ Alberta, Canada

Abstract
Non-insulin dependent diabetes mellitus (NIDDM) is associated with chronic hyperglycemia, which increases the risk of
developing microvascular and macrovascular complications. Elevated triglyceride (TG) and VLDL cholesterollevels and low
levels of HDL cholesterol have also been frequently reported in NIDDM patients. A diet high in complex carbohydrate and
low in fat is typically recommended for management of NIDDM, however, this has recently been challenged by scientific
reports of the benefits of dietary intakes high in monounsaturated fat. Thirty-two individuals with NIDDM were randomized
to receive either Ensure with Fibre® (30% fat) or a high monounsaturated fatty acid product, Glucema® (50% fat). These
products were consumed for 28 days at >80% of daily energy intake. Post-treatment, dietary compliance was verified by a
higher plasma TG 18:1 n-9 (p < 0.001) in the Glucema® group and a higher plasma TG 18:2 n-6 (p < 0.001) in the Ensure
with Fibre® group. The postprandial rise in blood glucose levels, determined by fingerprick sampies, was significantly lower
(p < 0.01) in the Glucema® group. Trends of clinical interest were greater mean decreases in the Glucema® group compared
10 the Ensure with Fibre® group in: fmctosamine, 9.l3 umol/L vs 0.l4 umol/L; glucose, 1.61 mmol/L vs 0.63 mmollL; and
insulin, 46.0 pmol/L vs 12.6 pmol/L; respectively. However, overall, fasting plasma glucose, fmctosamine, TG and cholesterol
levels were not significantly different between groups. Thus, in these patients, the high monounsaturated fat diet and the standard
diet were similar with regard to usual indicators of carbohydrate and lipid metabolism. A high monounsaturated fat diet appears
to pose no risk to lipoprotein metabolism in NIDDM patients. (Mol Cell Biochem 188: 81-89, 1998)

Key words: NIDDM, glucose metabolism, lipid metabolism, enteral formulas

Introduction uptake in target tissues, especially skeletal muscle [2] and


impaired postprandial insulin secretion [3].
Non-insulin dependent diabetes mellitus (NIDDM) is a NIDDM patients also have an altered lipid profile which
common disorder of carbohydrate metabolism. It is includes elevated triglycerides (TG) and very-Iow density
characterized by reduced insulin sensitivity and/or reduced lipoprotein (VLDL) cholesterol, and reduced high density
secretion of insulin. Most individuals with NIDDM have lipoprotein (HDL) cholesterol [I]. The hypertriglyceridemia
chronically high blood glucose levels [1]. This arises from is mainly due to increased VLDL-TG as a result of over-
several contributing factors including increased hepatic production of TG [4], possibly due to increased uptake of
glucose production [2], decreased insulin-mediated glucose glucose and free fatty acids in the liver. There is some

Address jor ofIPrints: L. McCargar, Departrnent of Agrieultural F ood & Nutritional Seience, 4-10 Agrieulture Forestry Centre University of Alberta, Edmonton,
AB T6G 2P5, Canada
82

evidence that the usual suppressive effect of insulin on the Laboratories, Columbus, OH, USA) is a liquid diet formulated
release of free fatty acids from adipose tissue is not optimally specifically for the nutritional support of patients with
effective in NIDDM [5], allowing increased flow of fatty acids abnormal glucose tolerance. lt provides 33% of energy from
to the liver. Although acute insulin release has an inhibitory carbohydrate (glucose polymers and fructose) and 50% of
effect on VLDL production due to increased degradation of energy from fat (64.7% ofthe fatty acids are oleic acid, 18: I
apo B, chronic hyperinsulinemia, as seen in insulin resistant n-9). Soy polysaccharide is also the source of dietary fibre
NIDDM patients, might override this effect [6] such that apo in this supplement. The macronutrient distribution and the high
B production is not inhibited. This, combined with the lipo- monounsaturated fatty acid content of Glucema® is similar
genic action of insulin and the increased availability of lipo- to some ofthe experimental diet treatments previously tested
genie substrates, may allow increased synthesis and secretion with NIDDM patients.
of VLDL resulting in increased plasma TG. In hypertri- Thus, the purpose ofthis study was to assess the short term
glyceridemia, the cholesterol ester transfer protein (CETP)- (7 days) and long term (28 days) effects of consuming a
mediated exchange of cholesterol and TG between HDL and high carbohydrate diet (Ensure with Fibre®) and a high
VLDL results in increased TG concentration in the HDL monounsaturated fat diet (Glucema®), at a minimum of 80%
particles and decreased HDL cholesterol concentration [7]. of total energy requirements, on indices of carbohydrate and
Thus, hypertriglyceridemia is often associated with increased lipid metabolism in outpatients with NIDDM. It was of
VLDL cholesterol and decreased HDL cholesterol. interest to investigate the impact of a high monounsaturated
As a result of hyperglycemia and hyperlipidemia, the fat diet, which has been proposed by others for diabetes
diabetic population is at increased risk for microvascular management. It was also of interest to determine whether
(nephropathy, retinopathy and neuropathy) and macrovascular the nutritional formula specifical1y designed for those with
(coronary heart disease) complications [8]. Thus, control of impaired glucose tolerance resulted in a different response
blood glucose is a primary goal ofnutritional management. than the standard nutritional formula.
However, because coronary heart disease is also a major
concern in diabetes, the optimal diabetic diet should aid in
minimizing cardiovascular risk factors. In NIDDM patients, Materials and methods
as stated above, these risk factors include high TG and VLDL
cholesterol, and low HDL cholesterol [9]. Design
The current Canadian dietary recommendation for people
with NIDDM is a diet high in complex carbohydrate and low The study was a prospective, randomized, two group, parallel
in fat, especially saturated fat [10]. However, there has been design. Subjects were randomized to receive either Ensure
recent interest in diets high in monounsaturated fatty acids for with Fibre® or Glucema® as their main source of energy
nutritional management ofNIDDM [11]. High mono- for 28 days. The subjects recorded measurements of blood
unsaturated fatty acid diets resulted in lower postprandial glucose using their own blood glucose monitoring equipment
blood glucose levels, lower fasting TG and lower VLDL just before, and two hours after each meal (a total of 6
cholesterol levels when compared to diets high in carbo- measures/day) for two days each week ofthe 28 day study.
hydrate [12-15]. This was found even when the diets were Fasting blood sampIes were taken by venipuncture on day
as high as 50% of energy from fat. Higher HDL cholesterol I, 8 and 29 for analysis of parameters of carbohydrate
levels have also been reported after consumption of high metabolism (glucose, fructosamine, insulin, glucagon) and
monounsaturated fatty acid diets [12-14]. As a result of lipid metabolism (TG, total cholesterol, apo B-lipoprotein
these studies and others, the revised American Diabetes cholesterol, HDL cholesterol and levels of 18:2 n-6 and
Association guidelines have added a recommendation for 18: 1 n-9 in TG). The subjects also recorded their liquid
increased monounsaturated fat for individuals with elevated formula and food consumption each day throughout the
TG and VLDL cholesterol [16]. study in daily diaries. They reported to the study centre
Liquid diets are formulated to provide complete nutrition once per week for follow-up and to pick up additional
and they can be used for enteral fee ding or as oral supple- formula.
ments to the diet. Ensure with Fibre® (Abbott Laboratories, Individuals were eligible to participate if they met the
Columbus, OH, USA) is a nutritional formula designed forthe following inclusion criteria: they had NIDDM which had
. general population. It has a composition similar to that been diagnosed by a physician; they were not taking insulin
currently recommended for NIDDM patients [10]; it provides and were between the ages of 18 and 70 years; their body
55% of energy from carbohydrate (hydrolyzedcomstarch and weight was stable (less than 5% change in weight in the last
sucrose) and 30.5% of energy from fat (56.6% of the fatty 3 months) with a BMI between 19 and 36 (only to exclude
acids are Iinoleic acid, 18:2 n-6). Soy polysaccharide those with morbid obesity); they had no history of ketosis
provides the source of dietary fibre. Glucema® (Abbolt and no other significant medical conditions. They were
83

required to do self blood glucose monitoring and have a 235 ml offormula). Up to 20% ofthe subjects' dailyenergy
support person with them throughout the study. The diabetic intake could be from regular food (referred to herein as
population typically has elevated plasma lipid and HgbAIC 'mini-meals'). The mini-meals were specifically designed to
levels, thus normal values could not be used for inclusion provide a similar macronutrient composition and a similar
criteria. Upper limits were set to exclude individuals with very ratio of saturated, monounsaturated and polyunsaturated fatty
high values (TG > 4.5 mmollL, cholesterol > 6.5 mmol/L, acids to the respective formula. Thus, the formulas were
HgbAIC > 0.09%). Subjects also could not be taking steroids, consumed at a minimum level of 80% of total daily energy
~-blockers or any lipid lowering medication. All medical intake. For example, if an individual' s energy requirement was
information was verified by the subject's physician. All 2350 kcal/day, they would be required to consume at least 8
subjects were required to do a taste test (by consuming 4 cans/day of Glucema. Detailed individualized counseling with
cans of each formula over aperiod of 8 h on 2 separate regards to the diet protocol was provided by a Registered
days) prior to enrollment. Ifthey tolerated both formulas, Dietitian.
they were then randomized to receive one of the two
formulas. The study was carried out at two sites and ethical
approval was received from both the University of British Athropometric measurements
Columbia and the University of Alberta. All subjects
provided their written consent. Anthropometric measurements were taken according to
published techniques [17]. Body mass index (BMI) was
calculated from weight and height (kg/m2). Weight was
Study diets recorded at each weekly visit. At the beginning and the end
of the study, circumference and skinfold measurements were
The study diets were different in macronutrient composition done. Circumference measurements were taken at the waist
as shown in Table 1. In general, they represented a high and the hip for determination of the waist-to-hip ratio
carbohydrate (Ensure with Fibre®) versus a high fat diet (WHR). Skinfold measurements were taken at four sites:
(Glucema®), with a majority ofthe fat in the latter comprised triceps, biceps, subscapular and suprailiac as an indirect
of monounsaturated fatty acids. Both diets provided similar measure of subcutaneous fat. Measurements were taken in
amounts of protein (9 .4g/23 5 ml can and 9. 9g123 5 ml triplicate and the mean of the closest two values was used,
respectively) and similar amounts of dietary fiber (3.3g1 as recommended by Gibson [17]. Percent body fat was

Table 1. Study diets'

High carbohydrate diet High fat diet

Diet composition:
(% oftotal energy)
Carbohydrate 55.0"10 33.3%
Fat 30.5% 50.0"10
Protein 14.5% 16.7%

Fatty acid profile:


(% oftotal fatty acids)
Polyunsatnrated 58% 26%
Monounsaturated 26% 64%
Saturated 16% 10"10

Sources of nutrients:
Carbohydrate(% oftotal) 58% maltodextrin 53.3% glucose polymers
32% sucrose 21.3% fructose
10% soy polysaccharide 25.4% soy polysaccharide

Fat (% oftotal) 100%comoil 85% high oleic saffiower oil


15% nonhydrogenated soy oil

Protein (% oftotill) 84% Na/Ca caseinates 100% Na/Ca caseinates


16% soy protein

'Source: Ross Laboratories (refs 35, 36)


84

estimated from the sum of skinfolds at four sites, using were a few instances where Tukey's HSD test failed to
published tables [18]. separate the means even though there was a significant
overall F test in the RBD ANOVA. For weekly intakc
parameters, calculated means were analyzed similar to the
Biochemical analysis other continuous parameters. All categorical parameters were
analyzed with a Cochran-Mantel-Haenszel for General
Serum glucose was measured using an enzymatic method Association. This tests the association between two para-
with a test kit from Kodak Ektachem (Rochester, NY, USA). meters (diet group and one of the categorical parameters)
Fructosamine was determined by a kinetic reduction test while controlling for a third, which in this case was study site.
using a test kit from Boehringer Mannheim Canada (Laval,
QC, Canada). Insulin and glucagon was determined by RlA
with kits from Imrnunocorp. Sciences Inc. (Montreal, QC, Results
Canada) and ICN Biomedicals Inc (Carson, CA, USA)
respectively. Thirty-four subjects volunteered to participate in the study.
Total cholesterol was determined by an enzymatic method Two subjects withdrew (one due to personal reasons and
using a kit from Diagnostic Chemicals Ltd (Charlottetown, the other was prescribed beta-blocker medication just
PE!, Canada). For determination ofHOL cholesterol, the apo prior to Day I of the study). Thus, data from 32 subjects
B containing lipoproteins in plasma were precipitated using (Vancouver site: n = 12; Edmonton site: n = 20) are pre-
the heparin manganese method [19]. In the presence of sented. There were 16 subjects in each diet group. Subject
heparin and manganese ions, apo B containing proteins form characteristics are shown in Table 2. There were no dif-
a crosslinked matrix which precipitates from plasma. The ferences between the groups at baseline for age, gender
cholesterol remaining in the sampIe is associated with the distribution, most indicators of health status and social
high density lipoprotein and was determined by the same kit history. The only exception was a group difference in history
as for total cholesterol. Cholesterol in the apo B containing of gastrointestinal disturbances (p < 0.05) which were more
lipoprotein (VLOL and LOL) was estimated by subtracting the frequent in the high fat diet group [6/16 reported a history
HOL cholesterol from the total cholesterol. TG determination of heartburn, diarrhea (2 subjects), gastric u1cer, hiatus
was done by an enzymatic method using a kit from Diagnostic hernia and colitis] than the high carbohydrate diet group.
Chemicals Ltd. (Charlottetown, PEI). The levels of 18: I n- The mean intake from formula per week (as a % of total
9 and 18:2 n-6 in plasma TG were determined by capillary energy) over the four weeks ofthe study was 87 ± 2,86 ± 2,
colunm gas liquid chromatography following extraction and 86 ± 2, 87 ± 2% for the high carbohydrate diet group (total
separation ofplasma TG. energy intake of2183 ± 111 kcal/day in week 4); and 87 ±
3,84 ± 3, 84 ± 3, 84 ± 3% for the high fat diet group (total
energy intake of 2127 ± 116 kcal/day in week 4). This
Statistical analysis resulted in consumption of 7-8 cans of formula and 1-2
mini-meals per day, which was similar between groups.
All biochemical measurements were done on days 1,8 and Compliance was determined from daily records kept by the
29 except fructosamine (not day 8). For these parameters, subjects, which were completed by all subjects for the 28
the baseline data and all changes from baseline (ie. day 8 - days. Also, subjects were asked to return labels from all of
baseline and day 29 - baseline) were analyzed with a
Randomized Complete Block Design (RBD) analysis of
variance model. This allows for comparison of the diet Table 2. Subjeet eharaeteristies at baseline
groups (high carbohydrate or high fat) while blocking on
High High
the two study sites. The residuals from fitting this model
earbohydrale fal diet
were examined for evidence of a non-normal distribution diet
with the Shapiro-Wilks test. Those parameters for which the (n = 16) (n= 16)
residuals showed evidence ofbeing non-normally distributed
Age (y) 59 ±2' 55 ±3
(using p < 0.05 for the Shapiro-Wilk test) at one or more 171.3 ±2.3
Height (em) 171.7± 1.4
time points (Baseline or one of the calculated changes from Weight(kg) 84.7±3.5 83.9 ±2.3
baseline) were analyzed with nonparametric methods. This Duralion ofNIDDM (y) 4.8± 1 4.1 ±O.6
consisted of first ranking the data and then analyzing the Gender (mlf) 13/3 12/4
ranks with the RBDANOVA. Significant findings for either On medieation far NIDDM (y/n) 1115 12/4
Smoking status (y/n) 1115 4/12
study site or feeding group were examined with theTukey's
HSD test to determine where the differences occurred. There 'Mean ± S.E.M.
85

the cans of forrnula they consumed. There was approximately Table 5. Self blood glucose monitoring values'
a 95% return rate. The reason labels were missing was
Breakfast Lunch Dinner Totalb
because the subject forgot to save them or they lost them,
not because of non-compliance. High carbohydrate diet (n= 120) (n= 123) (n= 122) (n=365)
Anthropometrie measures are shown in Table 3. There Pre-meal glucose
(mmoVI) 7,0± 0.2' 6,7±0.2 6.8 ±0.2 6,8 ±0.1
were no significant differences between groups in weight
Postprandial glucose (2 h)
(kg), BMI, WHR or body fat (%). (mmoVI) 8.3 ±0.2 8.1 ± 0.2 8.7±0.2 8,4±0.1
Indicators of glucose metabolism are shown in Table 4. Postprandial rise
There were no statistically significant differences between (2 hpc-pre) 1.4±0.S' I.S±0.4 2.0±0.9* 1.6 ± 0.4"
groups. However, of c1inical interest, both groups showed
High fat diet (n=124) (n= 124) (n= 123) (n=371)
decreases from baseline to day 29 in fasting glucose, 7 and
Pre-meal glucose
18%, insulin, 7 and 26%, and glucagon, 28 and 35% in the (mmolll) 7.2 ±0.2 7.1±0.2 7.4 ± 0.4 7.1 ±O,I
high carbohydrate and high fat diet groups respectively. Postprandial glucose (2 h)
Table 5 shows the results of the finger-prick blood (mmoVI) 7,9±0.2 8.3±0.2 8,S ± 0.2 8.2±0,1
glucose valucs takcn by the subjects at horne. The post- Postprandial rise
(2 hpc-pre) 0.7±0.3' l.3±0.4 1.4 ± 0.4* 1.l±0.2'*
prandial rise in blood glucose was lower (p < 0.05) in the
high fat diet group than the high carbohydrate group after 'Values were determined twice per week (6x/day) for the four weeks ofthe
study for a possible 128 measurements per group, There were a few missing
values in each group. bAll pre-meal measures combined; all postprandial
Table 3. Anthropometrie measurements measures combined. 'Mean ± S.E.M. 'p < 0.05 '*p < 0.01 rostprandial rise
in glucose: High Carbohydrate Diet group > High Fat Diet group,
Day! Day29

Weight(kg)
breakfast and dinner meals, but not after the lunch meal. Taken
High carbohydrate diet 84.7 ± 3.5'b 83.2 ± 3.S
High fat diet 83.9±4.1 83.0±4.l as an average over thc whole study, the rise in blood glucose
after a meal was significantly lower for the high fat diet group
BMI(kg/m2) than the high carbohydrate diet group (p < 0.01),
High carbohydrate diet 28.7±I.1 28.1 ± I.1 Indicators oflipid metabolism are shown in Table 6, Levels
High fat diet 28.4± I.1 28.4± I.1
of 18:1 n-9 in TG were significantly higher (p < 0,001) in the
WHR(cmlcm)
high fat diet group (who were consuming higher levels of 18: 1
High carbohydrate diet 0.92±2.2 0.92±2.6 n-9) than in the high carbohydrate diet group at the end ofthe
High fat diet 0.91±2.2 0.91±2.8 study. Similarly, levels of 18:2 n-6 in TG were significantly
higher (p < 0,001) in the high carbohydrate diet group (who
BodyFat(%)
were consuming higher levels of 18:2 n-6) than in the high
High carbohydrate diet 32.7± 1.7 31.8±1.9
High fat die! 32.2± 1.4 32.9± 1.6 fat diet group at the end of the study. This also suggests good
compliance to the study diets and incorporation of dietary fatty
"Mean ± S.E.M. ben = 16/group). No significant differences between groups.
acids into plasma TG over the one month study period. There
Table 4. Measures of glucose metabolism
were no other significant differences between the diet
groups. However, of clinical interest, both groups showed a
Day! Day8 Day29 20% decrease in HDL cholesterol from baseline to day 29.
Glucose (mmol/l)
Only day 1 and 29 data are discussed further. As shown in
High carbohydrate diet 8.73± 0.46",b 8.73±0.S3 7.97±OS3 Tables 4 and 6, changes in some ofthe parameters did occur
High fat diet 9.16±0.59 8.07±0.41 7.54±037 by day 8, However, in most cases, the day 8 value was either
interrnediate between days I and 29 or similar to day I or 29.
Insulin (pmol/l) There were some differences between the two study sites, but
High carbohydrate diet 180±37 15S±24 l78±4S
175±36 131 ±12 l29± 11
none that would influence the overall interpretation of the
High fat diet
results. Thus, site comparisons are not discussed further.
Glucagon (pg/ml)
High carbohydrate diet 387±84 303±79 297±76
High fat diet 248±38 193±37 163±22
Discussion
Fructosamine Ütmol/l)
High carbohydrate diet 296± 9 297± 13 The major findings ofthis study were that the high fat, high
High fat diet 293± 10 284± 12 monounsaturated fatty acid diet resulted in lower post-
"Mean ± S,E,M. ben = 16/group) No sigoificant differences between groups. prandial increases in blood glucose than the high carbo-
86

Table 6. Measures of lipid metabolism central adiposity (ie. a WHR > 1.0 for males and > 0.8 for
females) has been associated with insulin insensitivity [20]
Dayl Day8 Day29
and increased risk of coronary heart disease [21]. The graup
Triglycerides (mmol/l) means for this study were not at the risk level (.91 and .92),
High carbohydrate diet 1.81 ±0.27,·b 1.74±0.17 1.86 ± 0.28 however some individual subjects had significant central
High fat diet 1.46 ± 0.29 1.43±0.28 1.33±0.24
adiposity.
Total cholesterol (mmol/l)
High carbohydrate diet 5.06±0.21 5.22±0.41 4.58 ± 0.27
High fat diet 5.24 ± 0.19 5.45±0.31 5.05 ± 0.28 Carbohydrate metabolism

VLDL + LDL cholesterol (mmol/l) , Glucerna®, the high monounsaturated fatty acid formu-
High carbohydrate diet 3.59 ± 0.20 3.94±0.35 3.48 ± 0.28 lated diet, resulted in a lower rise in postprandial blood
High fat diet 3.75±O.l9 4.23±0.34 3.88 ± 0.26 glucose levels than the high carbohydrate diet (Ensure with
HDL cholesterol (mmol/l)
Fibre®. This is consistent with other studies that have
High carbohydrate diet 1.37±O.l9 1.29 ± 0.24 1.08±0.07 evaluated postprandial blood glucose when subjects
High fat diet 1.50±O.l6 1.22±O.l1 1.20±0.1O consumed a high carbohydrate or a high monounsaturated
fatty acid diet [11,12,22] andmore specifically, evaluation
18:2 n-6 in Triglycerides (% total fally acids) of a high carbohydrate formula or Glucerna® [23-26].
High carbohydrate diet 15.l±0.9 23.6±1.4* 26.1 ± 1.4*
High fat diet 14.1 ± 1.0 17.0±0.5 16.7±0.5
Glucerna® contains 7% of total energy from fructose.
Fructose pro duces less of an acute blood glucose response
18: 1 n-9 in Triglycerides (% total fally acids) than sucrase [27-30]. Thus, the lower rise in postprandial
High carbohydrate diet 43.8±0.8 39.0± 1.0 37.4±1.0·· blood glucose resulting from the Glucerna® formulated
High fat diet 42.1 ± 0.8 47.5 ± 1.0 46.3±1.4 diet is likely due to both the low carbohydrate to fat ratio
'Mean ± S.E.M. b(n = 16/group). 'The apo-B cholesterol components were and the presence of fructose instead of sucrose as a sugar
combined. *(p < 0.00 I) High carbohydrate diet group > High fat diet group source.
(Day8 - baseline andDay 29- baseline) **(p < 0.001) High fatdietgroup > Fasting blood glucose levels were not different between
High carbohydrate diet group (Day 29 - baseline)
graups. This is in agreement with most of the recent studies
comparing high carbohydrate vs. high monounsaturated fatty
acid diets [12, 14, 15,24,31]. One study did show a lower
hydrate diet. There were no other differences between fasting blood glucose level after a high monounsaturated
groups for the indicators of carbohydrate or lipid metabolism fatty acid diet [16]. In the present study, mean fasting glucose
except fatty acid incorporation into TG. As such, the graup values decreased in both diet graups (7% decrease in the
consuming the high monounsaturated fatty acid diet (18: I high carbohydrate diet group; 18% decrease in the high fat
n-9) had a higher percentage of 18: 1 n-9 in plasma TG, and diet group) over the 28 day study period. This may be partly
the group consuming the diet higher in 18:2 n-6 (high explained by the decrease in glucagon of 28 and 35%
carbohydrate diet) had a higher percentage of 18:2 n-6 in respectively in the two diet groups. Also, decreased fasting
plasma TG. These results suggest a high level of dietary blood glucose may be a result ofthe day-to-day consistency
compliance. Positive changes in some parameters occurred of the diet, regular fiber intake and more evenly distributed
in both groups, suggesting a possible influence of diet food consumption patterns thraughout the day.
consistency. Fructosamine determination, representing average blood
The randomization pracedure resulted in relatively weH- glucose levels for the previous 1-4 weeks, revealed no
matched groups with no differences in most characteristics. differences between graups. One study which examined
BMI is a measure of relative weight for height whereby a BMI fructosamine levels reported values of (mean ± S.D.) 227 ±
< 20 indicates underweight, 20-25 indicates a healthy 35 ~mol/l for nondiabetic patients, 299 ± 56 ~mol/l far
weight, 25-27 may be associated with some health risks for individuals with diabetes in good-moderate contral and 408
some people, > 27 is associated with increased risk ofhealth ± 95 j.l.mol/l far individuals with diabetes in poor contral
problems and a BM! > 30 is considered obese [17]. The BMI [32]. Thus, for the subjects in the present study, their mean
for the groups in this study was 28. Most NIDDM patients values of284 ± 297 ~mol/I, were indicative of good contral
are overweight, thus it is not surprising that these subjects with minimal change over the study period.
were not in the healthy weight range of a BMI of20-25. Initial Similar to the glucose results, a trend was observed in
body weight was maintained by the subjects throughout the serum insulin values whereby the high fat diet graup showed
study. This was a primary objective so as not to confound the a greater decrease from baseline (26%) than the high
biochemical results. Waist to hip ratio was assessed because carbohydrate group (7%). Because both groups showed
87

improvement, again this may be attributed to the consistency of energy in some investigations, making the difference in
of the diets. carbohydrate between the high fat and high carbohydrate
Blood glucagon levels of both groups were elevated diets larger than the present study. Subjects in the present
above the normal range of 50-100 pg/ml. It has been study were free-living outpatients, compared to other studies
speculated that persistent hyperglycemia perpetuates insulin in which subjects ate pre-prepared meals at a study centre
resistance and increases glucagon levels in NIDDM patients or were in a metabolic ward. Despite concerns of decreased
[13]. The deereased blood glucagon levels resulting from dietary eompliance among outpatients, the ease and con-
both diets (by 28-35%) suggest a decreased need for the venience of consurning cans of formula in the current study
counter-regulatory hormone and perhaps better metabolie resulted in a high level of compliance among the participants.
contro!. This study also included subjeets who were on medication for
diabetes, whereas others did not. This may be a confounding
factor. Lastly, the most important difference was that subjects
Lipid metabolism in this study were not hyperlipidemic and did not have
extremely high HgbAIC values. Those that have higher
The evaluation ofthe plasma lipid response to these two diet
values at the beginning of an intervention often have the
treatments was important for two reasons. First, there has
greatest relative improvements. The lowering effect of the
been debate recently about dietary guidelines for diabetes
high monounsaturated fatty acid diet on plasma TG levels
management, with eonsideration being given to a high
may only be readily apparent when baseline values are high.
complex carbohydrate, low fat diet [10] or a high fat, high
Lerman-Garberet al. [33] have reported that individuals with
monounsaturated fatty aeid diet [22]. Many have questioned
higher baseline TG levels had the larger decrements in their
the implications of the latter on plasma lipids. Second,
values.
Glucerna® had been well-evaluated in terms of its effects
The two lipid abnormalities that are consistently reported
on carbohydrate metabolism, however, little information was
in individuals with NIDDM, are elevated TG and reduced
available on its effect on lipid metabolism. For individuals
HDL cholesterol. In the present study, TG levels were within
with NIDDM, who may require a nutritional supplement for
the normal range and there was a slight decrease (9%)
an extended period of time, it was important to assess the
observed in the high fat diet group with no change in the high
longer term effects on plasma lipids.
carbohydrate diet group. Thus, either diet composition
Overall, none ofthe main lipid plasma components (TG,
would appear reasonable for individuals with normal TG
cholesterol, apo B lipoprotein cholesterol, HDL chole-
levels.
steroi) were different between groups. Individuals with
Of concern, HDL cholesterol levels decreased by 20%
NIDDM typically have elevated TG and VLDL-cholesterol
in both groups. The reason for this is unknown. For these
and reduced HDL cholesterol. Other studies have reported
individuals, exercise, weight loss or perhaps increased
that a high monounsaturated fatty acid diet (compared to a
intake of n-3 fatty acids may need to be considered as a
high carbohydrate diet) results in decreased TG [11-14, 22,
means ofmaintaining HDL levels.
33], decreased VLDL-cholesterol [12-14, 22], and in-
creased HDL-cholesterol [12, 14]. Only one found a de-
crease in total cholesterol [13] and none of the above studies
reported a significant change in LDL-cholesterol. Conclusions
The significant differences in response to high carbo-
hydrate and high monounsaturated fatty acid diets reported In this study both nutritional formulas were well-tolerated.
in other studies were not observed in the present study. This Although diets of very different carbohydrate to fat ratios
may be due, in part, to differences in experimental conditions were consumed, major parameters of carbohydrate and
among the various studies. This study had a relatively small lipid metabolism were not different between groups after
sampie size. There were many different trends observed in one month. The body appears to adapt to a wide range of
each diet group which may have reached significance with nutrient sourees. The observation of a lower postprandial
greater statistical power. The main source of energy in the rise in blood glucose presents a defmite advantage of the
present study was provided as liquid nutritional formulas, high monounsaturated fatty acid formulated diet, since
rather than solid food as in other studies. Although the use lowering day-Iong blood glucose levels may help to prevent
of liquid formulas influences gastric emptying and perhaps the chronie microvasular complications ofNIDDM [34].
satiety signals, it should not have a significant influence on The high monounsaturated fatty acid diet did not result in
nutrient metabolism. The macronutrient content ofthe diets the improvements in lipid parameters to the magnitude of
varied somewhat between studies. Specifically, the high that reported in other studies. This is likely due, in part, to
carbohydrate diet provided 60 [12,14,31,33] to 65% [13] the fact that the subjects in this study were normolipidemic
88

at baseline. In spite of its high fat content, the high mono- 14. Garg A, Grundy SM, Unger RH: Comparison ofeffects ofhigh and
unsaturated fatty diet did not appear to pose any risks in low carbohydrate diets on plasma lipoproteins and insulin sensitivity
in patients with NIDDM. Diabetes 41: 1278-1285, 1992
terms of either carbohydrate or lipid metabolism compared
15. Rasmussen OW, Thomsen C, Hansen KW, Vesterlund M, Winther E,
to the standard formulated high carbohydrate diet. Hennansen K: Effects on blood pressure, glucose, and lipid levels of
a high-monounsaturatcd fat diet compared with a high-carbohydrate
diet in NIDDM subjects. Diabetes Care 16: 1563-1571, 1993
16. American Diabetes Association: Nutritional recommendations and
Acknowledgements principles for individuals with diabetes mellitus (Position Statement).
Diabetes Care 20(Supp!. I): S 15-S 18, 1996
Financial support for the research was provided by Ross 17. Gibson R: Principles of Nutritional Assessment. Oxford University
Products Division, Abbott Canada (Montreal, QC) and Abbott Press, NewYork, NY, USA 1990
18. Dumin JVGA, Womersley J: Body fat assessed from total body density
US (Columbus,OH). The following dietitians are gratefully
and its estimation from skinfold thickness: measurements on 481 men
acknowledged for their counseling, monitoring and guidance and warnen aged from 16 to 72 years. Brit JNutr 32: 77-97,1974
ofthe participants in the study: Janet LePatourel MSc RD, 19. Wamick GR, Albers H: A comprehensive evaluation of the heparin-
Charitini Orphanidou MSc RD, Wendy King MSc RD, and manganese precipitation procedure for estimating high density
Arlene Nadon MSc RD. lipoprotein cholestero!. J Lipid Research 19: 65-76, 1978
20. Pederson SB, Borglum JD, Schmitz 0, Bak JF, Sorensen NS, Riche1sen
B: Abdominal obesity is assoeiatcd with insulin resistancc and rcduccd
glycogen synthetasc activity in skeletal musc1e. Metabolism: Clin Exper
References 42: 998-1005, 1993
21. Oshaug A, Bugge KR, Bjonnes CH, Ryg M: Use of anthropometrie
measuremenls in assessing risk for coronary heart disease: A useful
I. Reaven GM: Abnonnallipoprotein metabolism in non-insulin dependent tool in worksite health screening? Int Arch Oecupat Environ Health
diabetes mellitus-pathogenesis and treatment. Am J Med 83 (Suppl 67: 35~366, 1995
3A): 31-40, 1987 22. Garg A, Bantic JP, Henry RR, Coulston AM, Griver KA, Raatz SK,
2. Koltennan OG, Gray TS, Griffin J, Bumstein P, Insel J, Searlett JA, Brinkley L, Chen YDI, Grundy SM, Huet BA, Reaven GM: Effects of
Olefsky JM: Reeeptor and post-reeeptor defects eontribute to the insulin varying carbohydrate content of diet in patients with non-insulin-
resistance in non-insulin dependent diabetes mellitus. J Clin Invest 68: dependent diabetes mellitus. J Am Med Assoe 271: 1421-1423, 1994
957-969.1981 23. Davidson MV, PetersAL, Isaac RM: Lack of glucose rise after simulated
3. Reaven GM: Role ofinsulin resistanee in human disease. Diabetes 37: tube feeding with a low earbohydrate, high fat enteral fonnula in type
1595-1607,1988 I diabetic patients. (Abstr.) ClinRes 37: 140, 1989
4. Abrams H, Ginsberg H, Grnndy S: Metabolism of eholesterol and 24. Peters AL, Davidson MB: Effects ofvarious enteral feeding products
plasma triglyceride in nonketotic diabetes mellitus. Diabetes 31 : 903- on postprandial blood glucose response in patients with type I diabetes.
910,1982 J Parent and Ent Nutr 16: 6~74, 1992
5. Krentz AJ, Nattrass M: Insulin resistanee: a multifaceted metabolie 25. Harley JR, Pohl SL, Isaac RM: Low carbohydrate with fiber versus high
syndrome. Insights gained using a low-dose insulin infusion teehnique. carbohydrate without fiber enteral fonnulas. Effeet on blood glucose
DiabeticMed 13: 30-39,1995 excursion in patients with type II diabetes. (Abstr.) ClinRes 37: 141, 1989
6. Lewis GF, Uffelman KD, Azeto LW, Steiner G: Effeets of aeute 26. Galkowski J, Silverstone FA, Brod M, Isaac RM: Use 01' a low
hyperinsulinemia on VLDL apo B production in nonnal weight and carbohydrate with fiber enteral fonnula as a snack for elderly patients
obese individuals. Diabetes 42: 833-842,1993 with type 2 diabetes. (Abstr.) ClinRes 37: 89, 1989
7. TallAR: Plasma CETP. J Lipid Research 34: l255-1274, 1993 27. Crapo PA, Koltennan OG, Olefsky J: Elfects of oral fructose in normal,
8. Ross SA, Zinman B, Leiter LA, MacDonald F: Physieians Guide to Non- diabetic and impaired glucose tolerance subjects. Diabetes Care 3:
Insulin Dependent Diabetes Mellitus. STA Communications Inc, 582-584,1980
Montreal QC, Canada 1992 28. Bantle JP, Laine PC, Thomas JW: Metabolie effects of dietary fructose
9. Laakso M, Lehto S, Pentila I, Pyorala K: Lipids and lipoproteins and sucrose in types land II diabetic subjects. J Am Med Assoc 256:
predicting coronary heart disease mortality in patients with non-insulin 324-346, 1986
dependent diabetes. Circulation 88: 1421-1430, 1993 29. Crapo PA, Kolterman OG, Henry RR: Metabolic consequence oftwo-
10. Canadian Diabetes Association: Guidelines for the nutritional manage- week fructose feeding in diabetic subjects. Diabetes Care 9: 111-119,
ment of diabetes mellitus in the 1990's. A position statement by the 1986
Canadian Diabetes Association. Beta Release 13: 8-17, 1989 30. MeAteer EJ, O'Reilly G, Hadden DR: The effects of one month high
11. Campbell LV, Mannot PE, Dyer JA, Borkman M, Starlien LH: The fructose intake on plasma glucose and lipid levels in non-insulin
highmonounsaturated fat diet as a practica1 alternative far NIDDM. dependent diabetes. Diabetic Med 4: 62--<i4, 1987
Diabetes Care 17: 177-182,1994 31. Lennan-Garber I, Gulias-HerreroA, Palma ME, Valles VE, Guerrero LA,
12. Garg A, Bonanome A, Grundy SM, Zhang Al, Unger RH: Comparison Garcia EG, Gomez-Perez FJ, RuH JA: Response to high carbohydrate
of a high carbohydrate diet with a high monounsaturated fat diet in and high monounsaturated fat diet in hypertriglyceridemic non-insulin
patients with non-insulindependent diabetes mellitus. N Engl J Med dependent diabetic patients with poor glycemie contra!. Diab Nutr
319: 829-834, 1988 Metab 8: 33~345, 1995
13. Garg A, Grundy SM, Koffler M: Effect ofhigh carbohydrate intake on 32. Cefalu WT, Bell-Farrow AD, Petty M, Iz1ar C, Smith JA: Clinical
hyperglycemia, islet function, and plasma lipoproteins in NIDDM. validation ofa second-generation fructosamine assay. Clin Chem 37:
Diabetes Care 15: 1572-1580, 1992 1252-1256,1991
89

33. Lennan-Garber I, Ishazo-Cerro S, Zamora-Gonalez J, Cardosa-Saldana G, EnglJ Med 329: 977-986, 1993
PosadasRomero C: Effects of a high-monunsaturated fat diet enriched 35. Ross Laboratories: Ross Nutritionals. Ross Laboratories Publishers,
with avocado in NIDDM patients. DiabetesCare 17: 311-315, 1994 Montreal, QC, Canada, Jan 1990
34. Diabetes Control and Complications Trial Research Group: The effect 36. Ross Laboratories: Specialized enteral nutritional formula for patients
of intensive treatment of diabetes on the development and progression with abnormal glucose tolerance. Ross Laboratories Monograph,
of long-term complications in insulin-dependent diabetes mellitus. N Columbus, OH, USA May 1989
Moleeular and Cellular Biochemistry 188: 91-10 1, 1998.
© 1998 Kluwer Academic Publishers.

Cardiac sarcolemmal Na+-Ca2+exchange and Na+-


K+ ATPase activities and gene expression in
aUoxan-induced diabetes in rats
Leonard Golfman,l lan M.C. Dixon,l Nobuakira Takeda,2 Anton
Lukas,l Krishnamurti Dakshinamurti l and Naranjan S. Dhallal
lInstitute of Cardiovascular Sciences, St. Boniface General Hospital Research Centre, Departments of Physiology and
Biochemistry, Faculty ofMedicine, University of Manitoba, Winnipeg, Manitoba, Canada and 2Department of Internal
Medicine, Aoto Hospital, Jikei University, Tokyo, Japan

Abstract
To determine the sequence of alterations in cardiac sarcolemmal (SL) Na+-Ca2+ exchange, Na+-K+ ATPase and Ca2+-transport
activities during the development of diabetes, rats were made diabetic by an intravenous injection of 65 mg/kg alloxan. SL
membranes were prepared from control and experimental hearts 1-12 weeks after induction of diabetes. Aseparate group of
4 week diabetic animals were injected with insulin (3 U/day) for an additional 4 weeks. Both Na+ -K+ ATPase and Ca2+-stimulated
ATPase activities were depressed as early as 10 days after alloxan administration; Mg 2+ATPase activity was not depressed
throughout the experimental periods. Both Na+ -Ca'+ exchange andATPdependent Ca'+ -uptake activities were depressed in diabetic
hearts 2 weeks after diabetes induction. These defects in SL Na+ -K+ ATPase and Ca-transport activities were normalized upon
treatment of diabetic animals with insulin. Northem blot analysis was employed to compare the relative mRNA abundances of
ul-subunit ofNa+ -K+ ATPase and Na+ -Ca2+exchanger in diabetic ventricular tissue vs. control samples.At 6 weeks after alloxan
administration, a significant depression of the Na+-K+ ATPase u 1- subunit mRNA was noted in diabetic heart. A significant
increase in the Na+-Ca2+ exchanger mRNA abundance was observed at 3 weeks which retumed to control by 5 weeks. The
results from the alloxan-rat model of diabetes support the view that SL membrane abnormalities in Na+-K+ ATPase, Na+Ca2+
exchange and Ca'+-pump activities may lead to the occurrence of intracellular Ca2+overload during the development of diabetic
cardiomyopathy but these defects may not be the consequence of depressed expression of genes specific for those SL proteins.
(Mol Cell Biochem 188: 91-101, 1998)

Key words: diabetic cardiomyopathy, cardiac Na+-K+ ATPase, cardiac Na+-Ca 2+ exchange, sarcolemmal Ca 2+ transport,
sarcolemmal gene expression

Introduction membrane systems become defective with respect to their


Ca'+-transport properties in the diabetic heart [6-10]. These
It is now clear that the cardiac dysfunction commonly membrane changes are considered to cause the occurrence
observed in chronic diabetes is associated with defects in of intracellular Ca2+-overload and subsequent ultrastructural
Ca'+ -handling in cardiac myocytes within the myocardium damage and contractile dysfunction in diabetic cardio-
[1-5]. Two membrane systems which are intimately involved myopathy [1, 2]. This view is supported by the fact that
in the maintenance and regulation of intracellular Ca2+levels verapamil, a Ca'+ channel antagonist, prevents the functional,
for normal cardiac contraction-relaxation are the sarcolemma ultrastructural and metabolic abnormalities in the diabetic
(SL) and sarcoplasmic rcticulum (SR); however, these heart [11, 12]. It should be pointed out that this concept is

AddressJor ojjjJrints: N.S. Dhalla, Institute ofCardiovascular Sciences, St. Boniface General Hospital Research Centre, 351 Tache Avenue, Winnipeg,
Manitoba, R2H 2A6, Canada
92

primarily based on observations made using streptozotocin this experimental model producing 100% diabetogenesis.
(STZ) induced diabetes in rats as an experimental model. Three days after alloxan injection, rats displaying glycosuria
Although alloxan-induced diabetes in rabbits and rats also (> 2%) and elevated plasma glucose (> 300 mgl1 00 ml) were
produces cardiac dysfunction and ultrastructural damage used as the diabetic group. Age-matched control rats
[13-15], relatively little infonnation regarding the sequence received an injection of only the citrate-buffered saline
of changes in cardiac Ca 2+-transport in either of these two solution. Throughout the course ofthe experimental period,
experimental models is available in the literature. In this food and water were provided ad libitum. Some of the
regard, depressed SR Ca 2+-uptake activity occurs in both the randomly selected 4 week diabetic animals were given daily
alloxan-rat model and alloxanrabbit models of diabetes [10, subcutaneous injections of ultra\ente insulin for a further
16,17]. Furthermore, SL Na+-Ca2+ exchange activity is period of 4 weeks. The dose of insulin was adjusted to
decreased in myocardium from alloxan-diabetic rabbits achieve blood glucose levels in the range of 100-200 mg/
without any change in SL Ca2+-pump activity compared to 100 ml; the approximate dose of insulin was 3 U/day. All
non-diabetic control sampIes [17]. The Ca2+-transporting diabetic rats, insulin-treated diabetic rats and the age-
activities of the cardiac SL Na+-Ca 2+ exchanger and Ca 2+- matched control rats were killed at the desired time point
pump (ATP dependent Ca2+-uptake and Ca 2+-stimulated by decapitation. Trunk blood was collected at time of death
ATPase) are also markedly decreased at 4 weeks after in heparinized tubes and plasma was prepared from the
induction of diabetes by alloxan in rats; these defects were blood sampies by centrifugation at 3000 x g. Plasma was
prevented by insulin treatment [10]. However, important stored at -20°C for analysis of glucose (Sigma Glucose
differences exist in alloxan- and STZ -induced diabetes with Reagent Kit) and for RIA analysis of insulin (Lineo Rat-
respect to kctosis and ketonuria as well as plasma and Insulin RIA Kit). Hearts were imrnediately removed and the
cardiac lipid profiles [18-20], and the sequence of SL atria, connective tissue, as well as major blood vessels were
changes in alloxan-induced diabetes has not been examined removed. The ventricular tissue was weighed and then
to date. In view ofthe limited information available and the processed for the isolation ofthe SL membranes. Ventricular
differences among different experimental models of tissue used for molecular studies was washed and rinsed in
diabetes, this study was undertaken to examine the status a 10 mM EDTA solution to remove any adhering blood;
and sequence of changes in SL Na+-Ca 2+ exchange, Ca2+- these tissues were then placed and stored in liquid nitrogen
pump (ATP-dependent Ca2+-uptake and Ca 2+-stimulated for RNA extraction.
ATPase) and Na+-K+ ATPase activities in hearts ofrats with
alloxan-induced diabetes. In addition, the abundance ofthe
cardiac Na+-Ca 2+ exchanger and Na+-K+ ATPase mRNAs Isolation of sarcolemmal membranes
was examined in this experimental model of chronic
Purified light sarcolemrnal membrane fraction was isolated
diabetes.
from ventricular tissue according to the method ofPitts [25].
Briefly, the ventricles were washed, minced and then
homogenized in 0.6 M sucrose, 10 mM imidazole-RCl, pR
Materials and methods 7.0 (3.5 mVgtissue) witha polytron PT-20 (5 x20 sec, setting
5). The resulting homogenate was centrifuged at 12,000 g for
Experimental model: Alloxan-diabetic rats 30 min and the pellet was discarded. After diluting (5 ml/g
tissue) with 140 mM KCl 20 mM 3-(N-morpholino)-
Male Sprague-Dawley rats (200-250 g) were made diabetic propanesulphonic acid (MOPS), pH 7.4 at 37°C, the supematant
with a single injection of alloxan monohydrate (65 mg/kg) was centrifuged at 95,000 g for 60 min. The resultant pellet
into the tail vein. Rats were anesthetized with 2% halothane was suspended in 140 mM KCI, 20 mM MOPS, pH 7.4
and alloxan was dissolved in a 0.05 M citrate-buffer (pH buffer and layered over a 30% sucrose solution containing
4.5) saline solutionjust prior to injection. Because alloxan 0.3 M KCI - 50 mM Na4 P0 4 07 , and 0.1 M Tris (hydroxy-
can produce nonspecific kidney damage [21, 22], 5-10 ml methyl) aminomethane; (Tris)-HCl, pH 8.3. After centri-
of 0.9% NaCI were given intraperitoneally immediately fugation at 95,000 g for 90 min (using a Beckman swinging
following alloxan administration as suggested by Heimberg bucket rotor) the band at the sucrose-buffer interface was
et al. [23, 24] to diminish kidney damage. Rats given removed and diluted with 3 vol of 140 mM KCl, 20 mM
alloxan undergo a transient and dangerously fatal hypo- MOPS, pH 7.4 (at 37°C). A final centrifugation at 95,000
glycemic period. Thus, a 50% dextrose-saline solution was g for 30 min resulted in a pellet enriched in sarcolemrna.
administered subcutaneously within 12-24 h after alloxan All isolation steps were carried out at 0-4°C. The final pellet
administration to minimize and prevent mortality. These was suspended in 0.25 M sucrose - 10 mM histidine, pH 7.2
precautions resulted in a mortality rate of less than 2% in (3.5 mg/mi) and then quickly frozen in liquid N 2 • Protein
93

concentration in SL membranes was determined by the incubated at 37°C for various times in 0.20 ml ofmedium
method ofLowry et al. [26]. containing 140 mM KCl, 10 mM MOPS, pR 7.4, 2 mM
MgCI 2, and 45CaCI2 -EGTA (which contained 5 x 10--6 M free
Ca2+). Ca2+ accumulation was initiated by the addition of 4
Sarcolemmal Ca 2+-uptake activities mM Tris-ATP, pR 7.4. 180 ~l aliquots were immediately
filtered through Millipore filters (pore sizc = 0.45 ~M),
Na+ -dependent Ca 2+-uptake washed twice with 2.5 ml ice-cold 140 mM KCI, 20 mM
Na+ -dependent Ca2+-uptake measurement was carried out by MOPS, and 0.1 mM LaCI J pR 7.4, dried, and radioactivity
the method described previously [27]. In short, 5 I.d of determined for calculating the total Ca 2+ accumulation.
sarcolemmal vesicles (1.5 mg/mi; 7.5 ~g protein/tube) were Nonspecific Ca 2+-binding was measured in the absence of
preloaded with NaCI/MOPS buffer at 37°C for 30 min. ATP for each set of experiments and the ATP-dependent
These vesicles were rapidly diluted 50 times with Ca2+- Ca2+uptake was ca1culated by subtracting nonspecific Ca2+-
uptake medium containing 140 mM KCl, 20 mM MOPS, binding from the total Ca2+ accumulation.
0.4 ~M valinomycin, 0.3 ~Ci45Ca2+ and various Ca 2 +
concentrations (5-80 ~M), pR 7.4 at 37°C. After aperiod Na+-K+ ATPase activity
of 2 sec, the re action was stopped by the addition ofO.03 ml Estimation ofNa+-K+ ATPase activity was carried out by a
ice-cold solution containing 140 mM KCl, 1 mM LaCl J , 20 previously described method [32] with some moditications.
mM MOPS, pR 7.4 (at 37°C). Sampies (0.25 ml from 0.28 Briefly, sarcolemmal vesicles (10 ~g) were preincubated at
ml of the total reaction mixture ) were filtered through 37°C, 1.0 mM EGTA-Tris, 50 mM histidine-RCI, pR 7.4 at
Millipore filters (pore size = 0.45 um) and washed twice 37°C, 5 mM NaN3, 2.5 mM PEP, 100 mM NaCl, 10 mM
with 2.5 ml of the ice-cold stopping solution. The radio- KCl, 6 mM MgCI 2, and 10 I.U./ml pyruvate kinase. The
activity offilters was determined using a Beckman LS 1701 reaction was started by the addition ofO.025 ml 80 mM Na2
Counter. In parallel to these sampies, nonspecific Ca 2+- ATP, pR 7.4, and terminated after 10 min with 0.5 ml ice-cold
uptake was measured by placing the Na+ -loaded sarcolemmal 12% TCA. The liberated phosphate was measured by the
vesicles in Ca 2+-uptake medium which contained 140 mM method ofTaussky and Shorr [33]. In parallel experiments,
NaCI instead of KCI. The Na+-dependent Ca 2+-uptake either Na+plus K+ or Mg2+ was omitted from thc incubation
activity was corrected by subtraction of the non-specific medium. Na+-K+ ATPase activity was calculated as the
Ca2+-uptake values. difference between activities with and without Na+ plus K+.
Mg2+ATPase activity was estimated as the difference between
Sarcolemmal calcium pump activity the ac ti vi ti es registered with and without Mg2+ in the
For the determination ofMg2+-ATPase and Ca2+-stimulated absence ofNa+ and K+ in the medium. The degree of cross
ATPase activities, experimental conditions were the same contamination in SL membranes was assessed by determining
as reported elsewhere [28, 29] with some modifications. digitoxigenin-sensitive and ouabain-sensitive Na+K+ ATPase
Sarcolemmal vesic1es (resuspended in 140 mM KCl, 20 mM activities as weil as cytochrome C oxidase and rotenone-
MOPS, pR 7.4 at 37°C; 25 ~g protein/tube) were pre- insensitive cytochrome C reductase activities according to the
incubated at 37°C for 5 min in 0.5 ml ofmedium containing methods used earlier [8, 27].
140 mM KCl, 20 mM MOPS, pR 7.4 at 37°C, 2 mM MgCl"
5 mM NaN 3 , 0.1 mM EGTA, 2.5 mM phosphoenolpyruvate
(PEP), and 10 LU./ml pyruvate kinase. The reaction for RNA extraction and Northern blot analysis
Mg2+-ATPase was started by the addition of 4 mM Tris-ATP,
pR 7.4 at 37°C and terminated 5 min later with 0.5 ml of Total cardiac RNA was isolated from experimental and
12% ice-cold trichloracetic acid (TCA). The liberated sham control animals by the method of Chomczynski and
phosphate was measured as be fore [30]. Estimation oftotal Sacchi [34]. Recovered RNA was dissolved in diethyl pyro-
Ca 2+ + Mg 2+-ATPase was made in the above mentioned carbonate (DEPC)-treated water and the concentration of
medium containing 5 x 10--6 M free Ca2+. Mg2+-ATPase and nucleic acid was ca1culated from the absorbance at 260 nm
free Ca 2+ concentration in the incubation medium was prior to size fractionation. Twenty micrograms oftotal RNA
calculated using the 'SPECS' FORTRAN pro gram of was electrophoresed in a 1.2% agarose/formaldehyde gel and
Fabiato [31]. The Ca2+-stimulated ATPase activity was the the fractionated RNA was transferred to a 0.45 ~m positive
difference between the total ATPase and the Mg2+-ATPase charge-modified nylon membrane (NYTRAN Plus, Schleicher
activities. & Schuell). The RNA was covalently cross-linked to the
In order to measure ATP-dependent Ca 2+ uptake [28], membrane using UV radiation (UV Stratalinker 2400,
sarcolemmal vesic\es (22.5 ~g protein/tube) were pre- Stratagene). Blots were prehybridized at 42°C for 16 h. Each
94

membrane was hybridized with cDNA probes labelled (32P) the SL changes in the diabetic animals, activities of some
using a random primer labelling kit (specific activity >109 cpm selected marker enzymes were detemrined in the SL membrane
per Ilg DNA) at 42°C for 16-20 h. After washing, the fractions from control, diabetic and insulin treated diabetic rat
membranes were exposed to X-ray film (Kodak X-OMAT) at hearts (Table I). The SL protein yield was not different among
--80°C with intensifying screens. The cDNA for glyceralde- these groups and the membrane preparations employed in this
hyde-3-phosphate dehydrogenase (GAPDH) was obtained study were enriched by about 18 fold with respect to the heart
from the American Type Culture CoIlection (Rockville, MD, homogenate Na+-K+ ATPase activities. Because ouabain-
USA). Results of autoradiographs from Northem blot analysis sensitive Na+-K+ ATPase activities in the sarcolemmal
were quantified by densitometry (Bio-Rad, GS 670, Hercules, vesic1es was on average 15-18% ofthe total Na+-K+ ATPase
CA, USA). The signals ofmRNA specific for SL proteins were activity in all three preparations, the inside-outside populations
normalized to those of GAPDH to correct for differences in (82-85%) of the control, diabetic and insulin-treated SL
loading and/or transfer ofmRNA. vesic1es were similar. The cytochrome c oxidase and rotenone-
insensitive NADPH cytochrome c reductase activities ofthe
SL preparations shown in Table 1 suggest minimal contamina-
Statistical analysis tion by mitochondria and SR in control, diabetic and insulin
treated diabetic hearts.
All values are expressed as mean ± S.E.M. One way analysis
ofvariance (ANOVA) foIlowed by Student-Newrnan-Keuls
test was used for comparing the differences among multiple
Na+-K+ ATPase activity
groups. Significant differences among groups were defined
by a probability ofless than 0.05.
Table 2 indicates that SL Na+-K+ ATPase activity was
significantly depressed in diabetic rats 10 days after alloxan
administration. This depression persisted throughout the 12
Results weeks duration of the diabetic state but was corrected in
diabetic rats after treatment for 4 weeks with insulin. The
Sarcolemmal marker enzymes and characterization SL Na+-K+ ATPase activity from control, diabetic and
insulin treated diabetic rats was also studied by varying the
The presence of diabetes in our model was verified by concentration of Mg-ATP in the incubation medium (Fig.
elevated levels of plasma glucose and decreased ventricular 1). Increasing the concentration ofMg-ATP increased Na+-
to body wt ratio as weIl as depressed plasma insulin level and K+ ATPase activity in all groups, but Na+ -K+ -ATPase activity
loss of body wt in rats injected with alloxan (Table 1). All in diabetic heart remained depressed compared with controls.
changes were fully reversible with insulin treatment except Insulin administration normalized the Na+-K+ ATPase
that the loss of body wt was only partially reversible. To activity in the presence of varying concentrations of Mg-
assess whether differential contamination could contribute to ATP.

Table 1. General characteristics of control, diabetic and insulin-treated diabetic rats, as weil as of cardiac sarcolemmal membrane preparations.

Control Diabetic Insulin-treated diabetic

Bodywt(g) 465 ± 17 231 ± 12" 364 ± 21"


Ventricular wtJbody wt ratio (mg/g) 2.3 ± 0.15 3.2 ± 0.12' 2.4 ± 0.11
Plasma glucose (mg/dL) 147 ± 12 501 ± 18" 154 ±11
Plasma insulin (ng/mI) 3.0 ± 0.2 0.6 ± 0.2' 3.2 ± 0.3
Sarcolemmal yield 1.15 ± 0.20 1.21 ± 0.15 1.10 ± 0.14
(ma proteinlg tissue)
Digitoxigenin-sensitive 24.6 ± 1.4 16.3 ± 0.9 24.0 ± 1.2
Na+-K+ ATPase Ülmollmg/min) (J 7.5) (18.7) (18.2)
Cytochrome C oxidase 51 ± 5.1 49 ± 4.5 47 ± 5.0
(nmol/mg/min) (0.06) (0.51) (0.54)
Rotenone-insensitive 4.5 ± 0.4 4.1 ± 0.3 4.2 ± 0.5
cytochrome C reductase (U) (1.1) (1.1)
(nmollmglmin)

Values are means ± S.E. of 8 preparations per treatment group. Diabetes in rats was induced by alloxan and were used 8 weeks later. For the insulin treated
group, 4 weeks diabetic anima1s were treated daily with insulin for 4 weeks. Values in parenthesis represent the ratio of activities for marker enzymes in the
sarcolemma and respective homogenate. " Significantly different from control, p < 0.05.
95

Table 2. Influence of diabetes on Na+-K+ ATPase and ouabain-sensitive Na+-K+ ATPase as weil as Mg'+-ATPase and Ca'+stimulatedATPase activities in rat
heart sarcolemmal membranes obtained at different intervals after alloxan injection.

Na+-K+ ATPase Ouabain-sensitive Mg'+-dependent Ca'+-stimulated


Group activity Na+-K+ ATPase activity ATPase activity ATPase activity

1 Week
Control 23.7 ± 1.1 4.20 ± 0.S51 120.5 ± 10.6 19.15 ± I.S5
Diabetic 20.1 ± 0.9 3.55 ± 0.90 115.7 ± S.7 17.IS ± 1.19

10 Days
Control 22.1 ± 0.7 2.51 ± 0.50 99.7 ± 12.2 IS.65 ± 1.54
Diabetic IS.0 ± O.S* 3.25 ± 0.77 103.5 ± 11.6 15.49 ± 1.15"

2 Weeks
Control 23.2 ± 1.2 2.95 ± 0.54 117.1 ± 13.2 19.25 ± 1.07
Diabetic 16.S ± 1.3* 1.99 ± 0.79 105.3 ± 10.4 13.10 ± 0.99"

4 Weeks
Control 25.6 ± 0.9 4.41 ± 0.83 95.1 ± 12.6 IS.91 ± 1.43
Diabetic 17.1 ± 0.7' 3.44 ± O.SI IOS.4 ± 11.4 12.95 ± 1.21'

S Weeks
Control 25.7 ± 1.0 4.01 ± 0.69 112.0 ± 9.9 19.35 ± 1.01
Diabetic 16.5 ± 1.2* 3.20 ± 0.74 102.6 ± 13.1 13.41 ± 1.22*

12 Weeks
Control 24.S ± 1.5 3.95 ± 0.S4 101.5 ± 12.5 IS.51 ± 1.27
Diabetic 17.3±1.I* 3.56 ± 0.95 9S.4 ± S.5 12.67 ± 0.9S*

4 Weeks Diabetic and


4 Weeks Insulin Treated
Control 25.1 ± 1.1 3.97 ± 0.77 110.0 ± 12.5 20.6 ± 1.15
Treated Diabetic 25.0 ± 1.3 4.2S ± 0.65 120.5 ± 14.2 20.2 ± 1.57

Values are means ± S.E. of 5--7 experiments. The ATPase activities are expressed as ~ol Pilmglh. Sarcolemma treated with 0.2 mg deoxycholate/mg
sarcolemmal protein to expose all sites for Na+-K+ ATPase reaction; the activity was completely inhibited by 2 mM ouabain. Ouabain sensitive without any
pretreatment refers to enzyme activity inhibited by 2 mM ouabain. *Significantly different from control, p < 0.05.

30

25

-<J-Control
- e - Diabetic
- ... - lnsulin-treated

[MgATP] (mM)

Fig. 1. Cardiac sarcolemmal Na+-K+ ATPase activity at different concentrations ofMgATP in control, diabetic (S weeks) and insulin-treated diabetic rats.
Values are means ± S.E. of 4--6 experiments. *Significantly different (p < 0.05) from contro!.
96

Ca 2+-pump activities Na+-Ca 2+exchange activity

The SL Ca2+-stimulated ATPase activity, unlike the Mg2+_ Another set of experiments measured Na+-dependent Ca2+-
ATPase activity, was significantly depressed (17%) in uptake in sarcolemmal vesic1es isolated from hearts after
diabetic sarcolemma preparations 10 days after alloxan different times of inducing diabetes. Table 3 shows that the
administration (Table 2) and remained depressed throughout SL Na+-Ca2+ exchange activity in diabetics was depressed
the 12 weeks period. Insulin administration completely by -40% compared to controls 2 weeks after alloxan
reversed the depression in Ca2+-stimulated ATPase activity administration; the decrease in SL Na+ -dependent Ca2+uptake
associated with the diabetic state. It.should be pointed out that was progressive during the 12 week period of diabetes. Na+-
unlike SR, the Ca2+-stimulated activities ofthe SL preparations dependent Ca2+-uptake was also determined at different
were depressed by low concentrations (0.5--2 11M) ofvanadate. concentrations of Ca2+ in the incubation medium and the
ATP-dependent SL Ca2+-uptake was depressed by about 25% results shown in Fig. 2 reveal a depression in the activity at
in diabetic rats 14 days after the administration of alloxan all concentrations ofCa2+. There were no differences in the
(Table 3). The decrease in SL ATP-dependent Ca2+-uptake non-specific Ca2+- binding between sarcolemmal preparations
progressed during the 12 week period of diabetes. Insulin in these experiments. Insulin administration reversed the
administration to diabetic animals reversed the depression observed depression in SL Na+-dependent Ca2+-uptake
seen in ATP-dependent Ca2+ uptake (Table 3). activity in diabetic heart (Fig. 2 and Table 3).

Table 3. Influence of diabetes on ATP-dependent Ca2+-uptake and Na+-dependent Ca2+-uptake activities in heart sarcolemmal membranes at different time
intervals after inducing diabetes in rats with alloxan.

Ca2+-pump experiments Na+ -Ca2+exchange experiments


ATP-dependent Non-specific Na+-dependent Non-specific
Group Ca"-uptake Ca"-uptake Ca2+-uptake Ca2+-uptake

1 Week
Control 19.1 ± 0.94 3.05 ± 0.74 5.04 ± 0.95 1.59 ± 0.61
Diabetic 17.4 ± 0.90 2.85 ± 0.83 4.15 ± 0.81 1.35 ± 0.55

\0 Days
Control 18.9± 1.0 2.85 ± 0.59 4.95 ± 0.89 1.71 ± 0.75
Diabetic 16.0 ± 0.80 2.94 ± 0.71 3.65 ± 0.97 1.58 ± 0.49

2 Weeks
Control 18.5 ± 0.85 3.25 ± 0.87 4.95 ± 0.71 1.50 ± 0.62
Diabetic 14.0 ± 0.90· 2.25 ± 0.75 2.80 ± 0.55 1.29 ± 0.57

4 Weeks
Control 20.1 ± l.l 2.58 ± 0.71 5.20 ± 0.85 1.85 ± 0.77
Diabetic 13.2 ± 1.8· 1.99 ± 0.80 2.99 ± 0.67· 1.58 ± 0.81

8 Weeks
Control 18.1 ± 1.5 2.11 ± 0.91 4.97 ± 0.80 1.95 ± 0.41
Diabetic 12.1 ± 1.7· 1.55 ± 0.55 2.75 ± 0.63· 1.77 ± 0,45

12 Weeks
Control 18.7 ± 1.2 2.78 ± 0.99 4.91 ± 0.76 1.78 ± 0.71
Diabetic13.2 ± 0.9* 1.95 ± 0.66 2.69 ± 0.58 1.90 ± 0.80

4 Weeks Diabetic and


4 Weeks Insulin Treated
Control 19.0 ± 0.72 2.51 ± 0.86 4.93 ± 0.90 1.83 ± 0.62
Treated Diabetic 18.8 ± 1.0 2.47 ± 0.77 5.14 ± 0.85 1.80 ± 0.64
Values are means ± S.E. of5--7 experiments and are expressed as nmol Ca2+/mglmin for ATP-dependent Ca"-pump experiments and as nmol Ca2+/mgl2 sec
for Na+-Ca2+exchange experiments. ATP-dependent Ca2+-uptake was measured in the presence of \0 11M free calcium whereas Na+-dependent Ca2+-uptake
was measured in the presence of \0 11M free calcium. ·Significantly different from control, p < 0.05.
97

12
-O-Control
Ci - e - Diabetic
-'"" ~ 10
'"
Ce ~
:::> c
+
"
u 2
~'"
Q.
.,
C Cl)
E
-0
C +~
[ N(Ij
., u
9 ö
+", E
2: 5

1 I I I
20 40 60 80 100 120 140 160

Fig.2. Cardiae sareolemmal Na+-dependent Ca'+ -uptake aetivity at different eoneentrations ofCa'" in eontrol, diabetie (8 weeks) and insulin-treated diabetic
rats. Values are means ± S.E. ofH experiments. *Signifieantly different (p < 0.05) from eontrol.

mRNA abundancefor Na+-K+ ATPase and Na+-Ca 2+ diabetie hearts 6 weeks after alloxan administration. We also
exchanger examined the Na+-Ca2+exehanger/GAPDH mRNA abundanee
ratio 2--6 weeks after alloxan administration (Fig. 4). In these
Figure 3 illustrates that 2-5 weeks after alloxan administration studies, we could not detect any statistically significant
there was no significant change in the <x1-subunit ofNa+-K+ decrease in the expression ofthis mRNA between the control
ATPase / GAPDH mRNA abundance ratio between the and diabetic ventricles except that the ratio of Na+-Ca 2+
diabetic ventric1es with respect to the control ventric1e. exchanger/GAPDH mRNA ratio was markedly increased in
However, a significant depression was observed in the diabetic rats at 3 wecks. No effort was made to cxamine
abundanee ofthe mRNA for Na+-K+ ATPase <XI-subunit in changes in the expression of SL Ca2+-pump ATPase in the
diabetic heart since Northern blot analysis was not sensitive
enough to detect mRNA levels ofSL Ca2+-pump in the control
Ul subunit Na+-K+ A TPase mRNA heart.

0.5
o
.~
:r: 0.4
Discussion
Cl
d 0.3
0.-

The Na+-K+ ATPase is an integral membrane transporter and


~ ion pump that is crucial for the regulation of membrane
0:: 0.3
potential and ion transport. By transporting Na+ out ofthe cell
I-
...: and K+ into the cell against their respective concentration
+~ 0.2
gradients, the sodium-pump maintains the osmotic balance
Z 0.1 of the cell, transports nutrients into the cell and contributes
to the electrical activity in excitable cells [35]. The Na+-K+
C OB C OB C OB C OB ATPase consists of a transmembrane <x-subunit (MW =
2 weeks 3 weeks 5 weeks 6 weeks
112,000) and ß-subunit (MW =35,000) that is glycosylated.
The Na+-K+ ATPase subunits are known to belong to a multi-
gene family and three isoforms of the catalytic <x-subunit,
Fig.3. Steady-state mRNA abundanee ofu 1-subunitofeardiae sareolemmal namely <XI' <X2 ' and <X3 ' as weil as three isoforms of the ß-
Na+-K+ ATPase gene in hearts of alloxan-diabetie and age-matehed eontrol subunit, (ß I, ß2 and ß 3), have been cloned and sequenced [36-
rats. The relative intensity of mRNA bands was expressed as the ratio of
densitometrie intensity oful-Na+-K+ ATPase signal/GAPDH signal. Values
38]. Of the <x-isoforms, <XI is most commonly expressed,
are means ± S.E. of 4-6 experiments. C - eontrol; DB - diabetic. whereas <X2 and <X3 are mainly localized to excitable tissues
*Signifieantly different (p < 0.05) from eontrol. [39]. In rat cardiac tissue, the Na+-K+ ATPase <X and ß
98

Na+-Ca 2+ exchanger mRNA therapy was associated with a partial reversal in az-isoform
depression, without a significant effect on a j levels. These
0 results differ from those of Horowitz et al. [39] in which
:;::; 0.16
~ hypothyroidism was associated with decreases in both
I 0.14 immunoreactive a j and az- isoform cardiac Na+-K+ ATPase
0
a. 0.12 proteins. These reports suggest that the effects of diabetes and
<t:
<.9 hypothyroidism differ with respect to mRNA and protein
0::: 0.10
Q) expression vs. activity ofNa+-K+ ATPase. Isoform ratio shifts
Cl
ro
c 0.08 (of a/az) in response to hypertension are known to occur in
.s::.
() 0.06 the adult rat heart [44, 45] with no changes in aj-isoform
x
,
Q) levels but depressed az-isoform levels. One study [42]
0.04
N
ro reported elevated expression of the a 3-isoform in human
<1 0.02 failing hearts, whereas a decreased expression of all three a-
'ro
Z isoforms has been demonstrated in pressure-overloaded right
C OB C OB C OB C OB ventric1e. Thus, the significance ofprotein expressionlactivity
2 weeks 3 weeks 5 weeks 6 weeks in various pathologies is undefined.
In our study, we did not quantify the immunoreactive Na+-
Fig.4. Steady-state mRNA abundance ofcardiac sarcolemmal Na'-Ca" K+ ATPase protein concentration and therefore, reduced
exchange gene from hearts of alloxan-diabetic and age-matched control numbers ofthis enzyme at a later stage of diabetes (6 weeks
rats. The relative intensity of the autoradiographic signal specific for the or longer) may or may not be present. Since it is not clear
mRNA bands was expressed as the ratio of densitometric intensity of whether isoform mRNA levels may be proportional to
sarcolemmal Na'-Ca" exchange (target gene) signal/glyceraldehyde-3-
phosphate dehydrogenase (GAPDH - housekeeping gene) signal. Values
amounts ofNa+-K+-ATPase isoform proteins or the numbers
are means ± S.E. of 4-6 experiments. C - control; DB - diabetic. of functional pump units, further work will be needed to
*Significantly different (p < 0.05) from control. examine both the molecular and biochemical regulation of
this enzyme and its various a- and ß-isoforms in diabetes.
Nonetheless, our observations regarding depressed cardiac
isoforms are differentially expressed in a complex pattern Na+-K+ ATPase activity in alloxan-induced diabetes is in
during development [36-38] as weH as by various trophic agreement with a previous report employing STZ-induced
factors in pathological states [39-43]. The aj-mRNA is the diabetes in rats [46]. Dur finding ofno difference between
major isoform transcript (70-75% of the total a-subunit the expression ofthe Na+ -Ca2+exchanger mRNA in diabetes
mRNA abundance) expressed at aH developmental stages in vs. control hearts except an increase at 3 weeks of inducing
animal tissues [36-38] and in the human heart [42]. Changes diabetes is not surprising. Membrane factors, phosphorylation
in thyroid status are known to affect the sensitivity ofthe heart status, lipid alterations and metabolic state may likely be more
to cardiac glycosides and this hormone has been shown to important regulators of the SL Na+-Ca2+ exchanger than
increase the Na+-K+ ATPase activity in various types of molecular regulation of its function. Xiang and McNeill [47]
tissues, inc1uding the heart [43]. The hypothyroid state has reported that cytosolic protein kinase C (PKC) activity was
been shown to significantly decrease the levels ofaz mRNA reduced in diabetic rat myocardium and other investigators [48]
and a z protein for Na+-K+ ATPase [38]. As diabetes is observed that the insulin-stimulated Na+-Ca2+ exchanger
associated with both depressed thyroid and insulin levels [1, activity was blocked by a protein kinase inhibitor. In that
18], we exarnined whether changes occur in the expression study, the authors proposed that insulin may act through a
of a j subunit of the cardiac Na+ -K+ ATPase in alloxan- combination of G-protein coupling and protein phos-
induced diabetes. Dur results indicate a significant depression phorylation to enhance Na+-Caz+exchanger activity. Magyar
in the steady state a j subunit mRNA abundance at 6 weeks et al. [45] hypothesized that cardiac Na+-Caz+ exchanger
after alloxan administration. Since the cardiac SL Na+-K+ expression is regulated in a reciprocal fashion to Na+-K+
ATPase activity was depressed as early as 10-14 days after ATPase expression. Although we did not quantify Na+- Ca2+-
induction of diabetes, it can be concluded that there is no exchanger pro tein expression or N a+-K+ ATPase a 2-iso form
direct association between changes in steady-state aj-isoform subunit mRNA and protein levels, reciprocal regulation of
mRNAabundance and those ofNa+-K+ ATPase activity. Ng the Na+-Ca2+-exchanger and ofNa+-K+ ATPase a z expression
et al. [41], using 6 week STZ-induced diabetic rats, reported may occur. This could provide evidence for a homeostatic
altered az-protein levels but not a j levels in cardiac musc1e mechanism that would oppose the changes in cellular Ca-
wherein N a+ -K+ ATPase activity was significantly depressed. stores driven by the changes in Na+-Caz+ exchange Na+-K+
In that study, the truncated aj-isoform could not be detected ATPase activities. Zahler et al. [49] showed that the a j-
due to gel separation techniques. As well, 7 days of insulin isoform ofthe Na+-K+ ATPase was the only isoform detectable
99

in the adult T-tubule system of the myocyte and immuno- cytoplasmic levels of free Ca 2+ by sequestration into the
localization studies suggest that the Na+-Ca 2+- exchanger, lumen ofthe SR tubules [53]. Thus, the observed depression
which is governed by Na+ gradients in part generated by the in SL Ca 2+-transport activities in the alloxan-induced
Na-pump, is also concentrated in the T-tubule region [50). diabetic heart in our study may act to raise the intracellular
Thus, as suggested by Zahler et al. [49], it is possible that concentration of Ca2+. Intracellular Ca2+-overload is considered
some aspects of the iso form structure are better suited for to produce myocardial cell damage and contractile dysfunction
interaction with the Na+-Ca2+-exchanger than Na+ -K+ ATPase [53). In fact, 20 alloxan-induced diabetes in rats has been
u 2 or u J subunits. shown to exert adverse effects on cardiac performance [7,15].
Changes in cardiac Ca2+metabolism may underlie impaired It is therefore likely that SL membrane abnormalities with
contractile functioning. It is possible that depression ofCa2+ respect to its Ca 2+-handling may contribute towards the
transport not only via SR sequestration but also by the SL occurrence of intracellular Ca 2+-overload and diabetic
removal of Ca2+ in diastole could lead to impaired cardiac cardiomyopathy in rats induced by alloxan. Although cxact
relaxation. An excessive entry of Ca2+ in diabetic myo- mechanisms involving several biochemical and molecular
cardium has also been suggested to occur through changes approaches for explaining SL membrane defects in the
in SL stores for superficial Ca 2+ [51). In this study, we diabetic heart remain to be investigated in the future, our
determined the sequence ofalterations ofthe cardiac SLNa+- results demonstrate that depressions in cardiac SL Na+-K+-
K+ ATPase activity and SL Ca2+-transport activities in both ATPase and Ca2+-stimulated ATPase activities precede that
acute and chronic diabetes induced by alloxan. Depression ofthe SL Na+ -Ca2+-exchange activity in the alloxan-rat model
ofboth Na+-K+ ATPase and Ca2+-stimulatedATPase activities of diabetes.
preceded the decreases ofthe Na+-Ca 2+- exchanger and the
ATP-dependent Ca2 +-uptake activities. M g2+-dependent
ATPase activity remained unchanged throughout the course Acknowledgements
of this study; this finding is also in agreement with the study
ofMakino et al. [8] in STZ-induced diabetic rats. The results This study was supported by a grant from the Medical Research
presented here with respect to depression in SL Ca2+handling Council ofCanada (MRC Group in Experimental Cardiology)
and depressed Na+-pump activity occur consistently in the and the RacingAutomobile Memorial Foundation, Tokyo. Dr.
STZ-diabetic rat model [8] and in alloxan diabetic dogs Leonard Golfman was a recipient of an MRC Traineeship, Dr.
[52). The detailed time-course study by Makino et al. [8] lan M.C. Dixon is a Scholar of the MRC/PMAC Health
indicated that changes in the sarcolemma with respect to Program with funds provided by Astra Pharma Inc., and Dr.
Na+-Ca 2+ exchanger and Ca2+-pump activities occurred 4 Anton Lukas is aMyles Robinson Heart Scholar.
weeks after STZ administration, whereas other studies
examined Na+-pump activity and Ca2+-pump activity at 6-
8 weeks after STZ administration. Takeda et al. [9] using
STZ-induced diabetic rats demonstrated that depressions in
References
SLATP-dependent Ca2 +-uptake and Ca2 +-stimulatedATPase
1. Dhalla NS, Pieree GN, Innes IR, Beamish RE: Pathogenesis of
activities were both significantly decreased 18 days after eardiae dysfunetion in diabetes mellitus. Can J Cardiol I: 263-281,
induction of diabetes; these changes occurred prior to the 1985
depression of sarcolemmal Na+ -Ca2+- exchange activity. In 2. Schaffer SW: Cardiomyopathy associated with non-insulin-dependent
contrast to our study, these authors reported adepression diabetes. Mol Cell Bioehern 107: 1-20, 1991
3. Yu Z, Quanme GA, MeNeil JH: Depressed [Ca2+]; responses to
in Na+-K+ ATPase activity (24 days after inducing diabetes)
isoproterenol andeAMP in isolated eardiomyoeytes from experimental
after the depression in Na+- Ca 2 +- exchange activity. The diabetie rats. Am J Physiol266: H2334-H2342. 1994
significance ofthis difference between these two models of 4. Ren J, Davidoff AJ: Diabetes rapidly induees eontraetile dysfunetions
diabetes with respect to sequential changes of the Na+-K+ in isolated ventrieular myoeytes. Am J PhysioI272: HI48-HI58, 1997
ATPase and Na+-Ca2+ exchange activities are not clear. In 5. Xu YJ, Botsford MW, Panagia V, Dhalla NS: Responses of heart
funetion and intraeellular free Ca" to phosphatidie acid in chronie
common with both models, however, insulin administration
diabetes. Can J Cardiol12: 1092-1098,1996
reversed all ofthese sarcolemmal changcs. 6. Ganguly PK, Pieree GN, Dhalla NS: Defeetive sareoplasmie retieular
Decreased SL Na+-K+ ATPase activity is believed to affect calcium transport in diabetie eardiomyopathy. Am J Physiol244: E528-
the Na+-Ca2+- exchanger systemindirectly [53], and this rnay E535, 1983
favour steady balance [Ca2+]j beginning 10 days after the 7. Lopasehuk GD, Tahiliani AG, Vadlamudi RVSV, Katz S, MeNeil JH:
Cardiae sareoplasmie retieulum function in insulin-or camitine-treated
indication of diabetes with alloxan. Both SL Ca2+-pump and
diabelic rals. Am J Physiol245: H969--H976, 1983
the Na+-Ca 2 + exchanger in the heart are considered to be 8. Makino N, Dhalla KS, Elimban V, Dhalla NS: Sarco1emmal Ca 2'_
intimately involved in the efflux of Ca 2+ from the cardio- transport in streplozotoein-indueed diabetie eardiomyopalhy in rats.
myocyte whereas SR Ca2+-pump has been shown to lower the Am J Physiol253: E202-E207, 1987
100

9. Takeda N, Dixon IMC, HataT, Elimban V, Shab K, Dhalla NS: Sequence 30. Pieree GN, Dhalla NS: Mechanisms of the defect in cardiac myo-
of alterations in subcellular organelles during the development ofheart fibrillar function during diabetes. Am J Physiol 248: E170--E175,
dysfunction in diabetes. Diabetes Res Clin Prac 30 (Suppl.): SI13-S 1985
122, 1996 31. Fabiato A: Computer programs far calculating total from specified
10. Golfinan LS, Takeda N, Dhalla NS: Cardiac membrane Ca2+-transport free or free from specified total ionic concentrations in aqueous
in alloxaninduced diabetes in rats. Diabetes Res Clin Prae 31 (Suppl.): solutions containing multiple metals and ligands. In: S. Fleischer, B.
573-577, 1996 Fleischer (eds). Methods in Enzymology. Academic Press: NY, 1988,
11. Afzal N, Ganguly PK, Dhalla KS, Pieree GN, Singal PK, Dhalla NS: pp 378-417
Benefieial effeets of verapamil in diabetic cardiomyopathy. Diabetes 32. Norby JG: Coupled assay ofNa+-K+ ATPase activity. In: S. Fleischer,
37: 936--942, 1988 B. Fleischer (eds). Methods in Enzymology. Academic Press: NY,
12. Afzal N, Pierce GN, Elimban V, Beamish RE, Dhalla NS: Influence 1988, pp 116--119
ofverapamil on some subcellular defects in diabetic cardiomyopathy. 33. Taussky H, Shorr E: A microcaloric method for the estimation of
Am J Physiol 256: E453-E458, 1989 inorganic phosphorous. J Biol Chem 202: 678'-'{'85, 1953
13. Fein FS, Miller-Green B, Sonnenblick EH: Altered myocardial 34. Chomczynski P, Sacchi N: Single-step method of RNA isolation by
mechanisms in diabetic rabbit. Am J Physiol248: H729--H736, 1985 acid guanidinium thiocyanate-phenol-chloroform extraction. Anal
14. Bhimji S, Godin DV, McNeill JH: Myocardia1 ultrastructural Biochem 162: 156-159, 1987
changes in alloxaninduced diabetes in rabbits. Acta Anat 125: 195- 35. Skou JC, Esmann M: The Na, K-ATPase. J Biocnerg Biomcmbr 24:
200, 1986 249--261,1992
15. Vadlamudi RVSV, Rogers RL, Neill JH: The effect of chronic alloxan 36. Good PJ, Richter K, David IB: A nervous system-specific isotype of
and streptozotocin induced diabetes on isolated rat heart performance. the beta subunitofNa+-K+ ATPase expressed during early development
Can J Physiol Pharmacol60: 902-911, 1982 ofXenopus laevis. Proc Natl Acad Sci USA 87: 9088-9092, 1990
16. Lopaschuk GD, Katz S, McNeill JH: The effeet of alloxan and 37. Lingrel JB, Orlowski J, Shull MM, Price EM: Molecular genetics of
streptozotocin induced diabetes on calcium transport in rat cardiac Na, K-ATPase. Prog Nueleic Acid Res Mol Bio138: 37-89, 1990
sarcoplasmic reticu1um. The possible involvement of long chain 38. Sweadner KJ: Isozymes of the Na+-K+ ATPase. Bioehim Biophys
acylcamitines. Can J Physiol Pharmacol61: 439-448,1984 Acta 988: 185--220, 1989
17. Golfman LS, Takeda N, Beamish RE, Dhalla NS: Cardiac contractile 39. Horowitz B, Hensley CB, Quintero M, Azuma KK, Putman D,
failure and ultrastruetural abnormalities during the development of McDonough AA: Differential regulation of Na, K-ATPase alpha,
diabetic cardiomyopathy. In: N.S. Dhalla, R.E. Beamish, N. Takeda, alpha2 and beta subunit mRNA and protein levels by thyroid hormone.
M. Nagano (eds). The Failing Heart. Raven Press: NY, 1995, pp 131- J Biol Chem 265: 14308-14314, 1990
161 40. Ng Y-C, Yao AZ, Akera T: Tissue-specific isoform regulation ofNa+-
18. Pieree GN, Beamish RE, Dhalla NS: Heart Dysfunetion in Diabetes. K+ ATPase by thyroid hormone in ferrets. Am J Physiol257: H534-
CRC Press: Boca Raton, 1988, pp 23-50 H539,1989
19. Veleminsky J, Burr IM, Stauffacher W: Comparative study of early 41. Ng Y-C, Tolerico PH, Book C-BS: Alterations in levels of Na+-K+
metabolie events resulting from the administration of the two ATPase isoforms in heart, skeletal musele and kidney of diabetic rats.
diabetogenie agents alloxan and streptozotocin. Eur J Clin Invest I: Am J Physiol 265: E243-E251 , 1993
104-108,1970 42. Zahler R, Gilmore HM, Baldwin JC, Franee K, Benz EJ, Jr: Expression
20. Mansford KRL, Opie LH: Comparison of metabolic abnormalities in of alpha isoforms of the Na, K-ATPase in human heart. Biochim
diabetes mellitus induced by streptozotoeinor byalloxan. Laneet I: Biophys Acta 1149: 189--194, 1993
670-671, 1968 43. Chaudhury S, Ismail-Beigi FI, Gick GG, Levenson R, Edelman IS: Effect
21. Bruckman G, Wertheimer E: The action of alloxan homologues and of thyroid hormone on the abundance of Na, K adenosine triphosphate
relatedeompounds: Alloxan studies. JBiol ehern 168: 241-256,1947 alpha-subunit messenger. Mol Endocrinoll: 83--89, 1987
22. Vargus L, Friederiei HHR, Maibeno HC: Cortieal sponge kidneys 44. Charlemagne D, Orlowski J, Oliviero P, Rannou F, Beuve CS,
indueed in rats by alloxan. Diabetes 19: 33-37, 1970 Swynghedanw B, Lane LK: Alteration of Na, K-ATPase subunit
23. Heimberg M, Dunkerley A, Brown TO: The effeet on insulin and mRNA and protein levels in hypertrophied rat hearts. J Bio1 Chem
alloxan diabetes on hepatie transport of triglycerides and fatty acids. 269: 1541-1547, 1994
Biochem Pharmaeol14: 890--893, 1965 45. Magyar CE, Wang J, Azuma KK, MeDonough AA: Reciprocal
24. Heimberg M, Meng HC, Park CR: Effeet of sex, fasting and alloxan regulation of cardiac Na, K-ATPase and Na/Ca exchanger: Hyper-
diabetes on the uptake of neutral fat by isolated perfused rat liver. Am tension, thyroid hormone, development. Am J Physiol 269: C675-
J Physiol195: 673'-'{'77, 1958 C682, 1995
25. Pitts BJR: Stoichiometry of sodium-calcium exchange in eardiac 46. Pierce GN, Dhalla NS: Sarcolemmal Na+-K+ ATPase aetivity in diabetic
sarcolemmal vesicles. J Biol Chem 254: 6232'-'{'235, 1979 rat heart. Am J Physiol 245: C241-C24 7, 1983
26. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ: Protein measure- 47. Xiang H, McNeill JH: Protein kinase C activity is altered in diabetic
ment with the Folin phenol·reagent. J Biol Chem 193: 265-275, rat hearts. Biochem Biophys Res Commun 187: 703-710,1992
1951 48. Ballard C, Mozaffari M, Schaffer S: Signal transduction mechanism
27. Dixon IMC, Hata T, Eyolfson DA, Dhalla NS: Sareolemmal Na+-Ca'+ fur the stimulation ofthe sarcolemma Na+-Ca2+ exchanger by insulin.
exchange activity in hearts subject to hypoxia reoxygenation. Am J Mol Cell Biochem 135: 113-119, 1994
Physiol253: HI026-Hl034, 1987 49. Zahler R, Sun W, Ardito T, Kash-Garian M: Na, K-ATPase a-isoforrn
28. Kaneko M, Beamish RE, Dhalla NS: Depression ofheart sarcolemmal expression in heart and vascular endothelia: cellular and developmental
Ca'+-pump activity by oxygen free radicals. Am J Physiol256: H368- regulation. Am J Physiol 70: C361--C371, 1996
H374,1989 50. Frank JS, Mottino G, Reid D, Molday REt, Philipson KD: Distribution
29. Seppet EK, Dhalla NS: Characteristies ofCa'+-stimulated ATPase in of the Na+-Ca2+ exchange protein in mammalian cardiac myocytes:
rat heart sarcolemma in the presence of dithiothreitol and alamethicin. An immunofluorescence and immunocolloidal gold-Iabelling study. J
Mol Cell Bioehern 91: 137-147,1989 Cell Bio11l7: 337-345,1992
101

51. Pierce ON, Kutryk MJB, Dhalla NS: Alterations in calcium binding adenosine triphosphate enzyme system in dog hearts. Biochem Biophys
and composition of the cardiac sarcolemmal membrane in chronic Res Commun 96: 799-804, 1980
diabetes. Proe Natl Aead Sei USA 8: 5412-5416, 1983 53. Dhalla NS, Pierce ON, Panagia V, Singal PK, Bearnish RE: Ca1ciummove-
52. Onji L, Liu M-S: Effects ofalloxan-diabetes on the sodium potassium ments in relation to heart function. Basic Res Cardiol 77: 117-139, 1982
Molecular and Cellular Biochemistry 188: 103-111, 1998.
© 1998 Kluwer Academic Publishers.

Role of oxygen derived radicals for vascular


dysfunction in the diabetic heart: Prevention
bya-tocopherol?
Peter Rösen, Xue1iang Du and Diethelm Tschöpe
Diabetes Research Institute at the Heinrich-Heine-University, Düsseldorf, Germany

Abstract
The evidence that the generation of reactive oxygen intermediates (ROI) plays an important role for the increased cardiovascular
risk in diabetes is summarised. In addition to the weil known parameters of oxidative stress as lipid hydroperoxides and
thiobarbituric acid substances (TBARS), recent observations indicate that isoprostanes which can be taken as a more specific
parameter of oxidative stress, are generated in higher amounts by diabetic patients. This increased formation of isoprostanes
ean be inhibited by an instalment of a elose metabolie control or the supplementation with tocopherol. The cause for the
elevated oxidative stress is not yet fully understood, however the autoxidation of glucose, the formation of advanced glyeation
endproduets and the activation ofNADPH-oxidase seem to be relevant processes. Sinee ROI are able to quench nitric oxide
and to inhibit the synthesis of prostaeyclin, the antithrombotic, vasodilating and antiatheroselerotie properties of endothelium
are impaired in diabetes. Additionally, the balance of endothelial mediators released by endothelium is shifted to angiotensin
II and endothelin, compounds which enhance the proliferation of smooth museIe cells and may limit the coronary reserve of
myocardium. The activation ofthe transcription factor NF-KB by glucose and its autoxidative products is regarded as a key
event in the transformation ofthe vasculature in diabetes. Epidemiological observations and very recent clinical studies underlie
the impact ofROI for the development of cardiovascular complieations in diabetes and suggest that an antioxidative treatment
might be helpful to reduce the cardiae risk in diabetes. (Mol Cell Biochem 188: 103~ 111, 1998)

Key words: diabetes, myocardium, oxygen derived radicals, vitamin E, transcription factor NF-KB, endothelium, nitric oxide,
lipid peroxidation

Introduction by the instalment of an intensified metabolie eontrol [11].


These observations suggest that the cardiovascular risk is not
There is much evidence that the cardiovascular risk is direetly related to the degree of hyperglycaemia, and that
increased in diabetes. It is estimated that the incidenee for the relationship between the cardiovascular risk and the
eardiovaseular death is 3~7 fold elevated in diabeties as metabolie defect is more eomplex than that observed for
compared to the non-diabetic population. In addition, the sex diabetic microangiopathy. We and others have suggested that
differenees typically observed in non-diabetics are not seen it is not the hyperglycaemia itself, but a related factor such
in diabetic patients [1~1O]. The DCCT data indicate thatthe as the increased formation of oxygen derived radicals whieh
diabctic microangiopathy is strictly related to the extent is causing the increase in the cardiovascular risk [13~15].
of the metabolic defeet. The development of diabetic In line with this hypothesis, the amount of lipid peroxides
nephropathy and retinopathy could be prevented or at least and isoprostanes taken as speeifie parameters for an elevated
delayed by an improvement of the metabolie control [11, oxidative stress have been shown to be inereased in both
12]. A similar, elose relationship has not been demonstrated types of diabetic patients. Furthermore, much evidence has
between the cardiovascular risk and the metabolie state. Even been presented that reactive oxygen species ean cause severe
in the DCCT study, the cardiovascular risk was not affeeted disturbances in the regulation of coronary flow and cellular

Addressfor ojJprints: P. Rösen, Klinisch-BiochemischeAbt., Diabetesforschungsinstitut, Aufm Hennekamp 65, D-40225 Düsseldorf, Germany
104

hemostasis, which may finally lead to severe macrovascular 1,20 -r-----------------..,


lesions typically seen in diabetic patients after a diabetes
duration of one or two decades [I, 8, 9]. In the following we
will discuss in which way the generation of oxygen derived
radicals may be related to the metabolie defects of diabetic 1,00
patients and contribute to the development of an increased
cardiovascular risk in diabetes. E
c
0:1'
C') 0,80
N
What is the evidence that the oxidative stress is 0
increased in diabetic patients? 0

Oxidative stress is defined as an imbalance between the 0,60


formation of oxygen derived radicals and the antioxidant
capacity. Thus, oxidative stress can be derived either from
an increased formation of reactive oxygen species and/or by
a diminished ability to inactivate reactive oxygen species. 0,40 -'-,------r-----r----.,.-----!
Until now, there is no general accepted marker for o 50 100 150 200
oxidative stress. Therefore, evidence for an increased Time (mln)
oxidative stress is mainly derived from indirect parameters
such as lipid peroxidation. In various studies using manifold Fig. 1. Inhibition ofthe copper-induced oxidation oflow density lipoproteins
techniques an increased rate of lipid peroxidation has been (LDL) by treatment of diabetic patients with vitamin E. LDL was isolated
shown (Table I): increased levels oflipid peroxides and lipid from the plasma of a type II diabetic patient before (e) and after treatment
with DL-a-tocopherol (800 IU/day) for 24 weeks C.). The LDL oxidation
hydroperoxides, oxidised low density lipoproteins (LDL),
was tested in vitra by the addition of CuSO4 and the amount of diene
and thiobarbituric acid reactive compounds (TBARS) have conjugated lipids fonned was measured at 234 nm as described [25, 26]. As
been observed in the plasma ofboth types of diabetic patients can be seen, the lag phase before the start of LDL oxidation is prolonged
[16-20]. Very recently, the formation of isoprostanes has and the rate for oxidation prolonged by treatment with vitamin E.

Table 1. Evidence for oxidative stress in diabetes

Marker Observation Reference

Lipid peroxides elevated in plasma Nishigaka et al. 1981


Nishigakaetal.1981
Kajietal.1985
Jain et al. 1989
Chiriccco et al.1993
Nourooz-Zadehetal.1995
TBARS elevated in plasma Tomiyamaetal.1976
Yagi 1976
Conjugated dienes elevated in nerve tissue Low and Nickander 1991
Superoxide Anion fonnation in plasma elevated Cerielloetal.1991
Isoprostanes elcvated in plasma Gopaul et al. 1995
increased urinary excretion Davi et al. 1997
nonnalised by vitamin E
LDL amount of oxidised and Bellomo et al. 1995
glycoxidised LDL is elevated Reavan et al. 1995
nonnalised by vitamin E Leonhardt et al. 1997
VitaminE reduced in platelets Karpenetal.1985
reduced diet dependent diet Simon-Schnass et al. 1997
uptake ofvitamin E
Antioxidative capacity reduced in plasma Asayama et al. 1993
Tsai et al. 1994
Sundarametal. 1996
Maxwell et al. 1997
lOS

3000~----------------------------------------~

>: 2500
CU
~
m
-
2000
CJ
.lII::

-
G)
~ 1500
C.
::::I
>- 1000
~
G)
c:
UJ 500

O~----~----~~----~-----r----~----~
o 4 B 12 16 20 24

Vitamin E (mg/day)

Fig. 2. Uptake of vitamin E by diabetic patients dependent upon the caloric intake. Thc uptake of vitamin E was calculated as described [27] from the
nutritional protocols. There is a strong relationship between the total energy and the vitamin E uptake. A significant number ofpatients (43%) took up less
vitamin Ethan the recommended daily allowance of 12 mg/day.

been suggested as a more specific parameter of oxidative contained high amounts of polyunsaturated fatty acids ([27],
stress [21]. Isoprostanes such as the 8-epi-isoprostaglandin Fig. 2). On this background it is not surprising that the
F2a are synthesised non-enzymatically from arachidonic acid intracellular concentrations of tocopherol and reduced
by auto xi dative processes. In two recent studies evidence has glutathione have been reported to be diminished [28, 29]. In
been presented that the formation ofthese specific markers addition the total antioxidative capacity in the plasma of
of oxidative stress is increased in diabetic patients [22]. diabetic patients is reduced [30].
Furthermore, the levels of isoprostanes measured in the Whereas the evidence for an increased oxidative stress
urine became reduced ifthe metabolie control was improved has convincingly demonstrated, less is known about the
or if the patients were treated with antioxidative vitamins sources and the mechanisms where and by which these
(22). Thus, these and other data suggest a clear dependency reactive oxygen radicals are generated in diabetes. In
ofthe oxidative stress on the metabolie control as originally hypercholesterolemie as weil as hypertonie animal models,
postulated from experimental observations [23]. several groups have shown that the endothelium is one ofthe
It is important to note that the glycation of proteins major sources for the generation of reactive oxygen species.
increased their susceptibility for oxidative modifications. In parallel with the vascular dysfunction the formation of
Thus, glycated LDL are much more easily oxidised than non- superoxide anions became augmented and removal of endo-
glycated LDL. Both fractions of LDL - oxidised and thelium abolished the production of reactive oxygen
glycoxidised LDL - are increased in diabetic patients [24- species completely [31-33]. That high glucose stimulates
26]. This high susceptibility can be reduced by treatment of the generation of superoxide anions by human umbilical
diabetic patients with tocopherol [24-26]. A typical example vein endothelial cells (HUVEC) has been shown very
is shown in Fig. I. recently [34]. This process is initiated by the release of
Another point may be of importance. We have recently calcium from intracellular stores [35]. These data and the
observed that the provision of patients with vitamin E might observation that the production of superoxide anions is not
be insufficient in diabetes, especially in older patients and prevented by inhibitors of nitric oxide synthases, P-450-
in those who are on a calorie restricted diet. Thc daily uptake dependent oxygenases, and lipo- and cyclooxygenases
of vitamin E was partly lower than the recommended daily suggest that the autoxidation of glucose may be one of the
allowance. About 55% of the patients older than 55 years major sources of reactive oxygen species in diabetes [36].
took up vitamin E in an amount lower than 12 mg/day. This In addition, Giordino et al. [37] have recently presented
situation became further worsened if the patient's diet evidence in elegant experiments that the formation of
106

Mechanisms involved in the generation of


oxygen derived radicals in hyperglycemia

Autoxidation Advanced glycation


of glucose
~ endproducts (AGE)

lfI °2-
Activation of receptor
NOS-Synthase for AGE (RAGE)
NADPH-Oxidase

Fig.3. Possible mechanisms responsible for the increased generation of oxygen derived radicals in diabetes.

advanced glycation end products (AGE) is c10sely associated gesting the autoxidative process plays an important role in
with the formation of reactive oxygen species. Inhibition of the complex reaction cascade leading finally to AGE. In
ROS also prevented the generation of AGE-products sug- addition, AGE products can stimulate the release of ROS in
endothelium by a receptor mediated (RAGE) process, the
intracellular signalling is not yet known [38].
In summary, we can state that there is much evidence for
140 an increased oxidative stress in diabetes. Several hypothesis
have been suggested to link diabetes with the increased
120 generation of ROS. These hypothesis are mainly based on
experimental data using cultured cells, conc1usive c1inical
..- evidence is, however, missing (Fig. 3) .
100
~
411
Ö
E 80 How is the increased oxidative stress related to the
.s elevated cardiovascular risk?
0
u;> 60
U Numerous studies have shown that ROS or oxidised LDL are
W
ab1e to significant1y alter the vascular function and to disturb
40
cellular hemostasis [20, 39].
The impaired endothelium dependent regulation of
20 vasomotion can be taken as one of the earliest indications of
vascular dysfunction in diabetes [13,40-43]. In experimental
0 studies the endothelium dependent vasodilatation by 5-
hydroxytryptamin was impaired dependent on the duration
c OB +SOO +Vit-E
of diabetes in coronary vessels, but also in mesenteric ones
[13,40]. Nitenberg et al. observed a reduced acetylcholine,
Fig. 4. Influence ofsuperoxide dismutase and vitamin E on the endothelium
endothelium mediated vasodilatation in coronaries of both
dependent stimulation of coronary flow in streptozotocin-diabetic rats.
Diabetes was induced in rats by streptozotocin. After a diabetes duration
type of diabetic patients [44]. Since perfusion ofthe isolated
of 16 weeks, the stimulation of coronary flow by 5-hydroxytryptaminon was rat heart by superoxide dismutase or pre-treatment of the
measured in the isolated heart preparation as described previously [13]. diabetic rats with high amounts of a-tocopherol prevented the
The half maximal concentrations (Ee\o) were determined and represent a impairment of endothelial function (Fig. 4), we concluded that
measure for thc sensitivity of endothelium to dilate the coronary vaseulature.
the release of superoxide anions is responsible for the
As can be seen, in diabetes (DB) the sensitivity of endothelium is impaired
as compared to healthy controls (C), but perfusion with SOD (50 f'U/ml) or
impairment of endothelium mediated vasodilatation. Similar
pre-treatment ofthe animals with tocopherol (1000 U1kg body wt/day) was conclusions have been drawn from several studies using
able to improve or restore the endothelium dependent vasodilatation. different types of vessels [40-43].
107

In addition, oxidised LDL may additionally promote the 1. Inactivation of the action and synthesis of nitric oxide
development of vascular lesions by several mechanisms and prostacyclin which are not only the most potent
such as quenching of NO, inhibition of the synthesis of vasodilators, but also significantly inhibit the expression
prostacyclin, and the induction of tissue factor and other of adhesion molecules and the proliferation of smooth
mediators which reduce the thromboresistance by the musc1e cells [48]. Two major mediators of the
expression of adhesion molecules (ICAM-I, VCAM, P- antithrombotic and antiathcrosclcrotic defcnce become
Selectin) in endothelium and blood cells [39,45,46]. defective by oxidative stress. A shift in the balance be-
The mechanism by which these vascular changes are tween these vasodilating mediators and the counteracting
occurring by oxygen derived radicals , are not yet weil under- endothelin and angiotensin II may be a further consequence
stood. We assume from preliminary data that high glucose which has important implications for the proliferation of
induces the generation of superoxide anions and thereby the smooth muscle cells and the remodelling of the heart in
activation oftranscription factors as NF-KB. A variety of diabetes [13, 49].
genes has been shown to be regulated by NF -KB proteins (Fig. 2. Activation ofNF-KB and presumably other transcription
5) including genes encoding TNF-a, IL-I, VCAM, CAM-I, factors which would induce a proinflammatory state, ac-
ICAM -I, c-myc, macrophage colony stimulating factor tivation of the immune response and cell growth, proc-
(MCSF), monocyte chemotractant factor-I (MCP-l), tissue esses which are characteristic in the pathogenesis of
factor and others [47]. Thus, it is obvious that NF-KB can atherosclerosis [50].
play an important role in the pathogenesis of athero-
sclerosis and other defects in cellular hemostasis. In line
with this assumption, we observed that high glucose, but also Can a-tocopherol reduce the cardiovascular risk in
the non-metabolisable glucose derivative 3-0-methyl- diabetes?
glucose is able to activate the transcription factor NF-KBlRel
in human umbilical vein endothelial cells (HlNEC, Fig. 6). Epidemiological evidence
The activation ofNF-KB was prevented by incubation ofthe Descriptive epidemiological evidence has been demonstrated
ceHs with antioxidants as tocopherol and thioctic acid that ischemic heart disease mortality rates and lipid stan-
indicating that the glucose mediated activation ofNF-KB is dardised vitamin E levels are inversely related [51,52]. The
caused by oxygen derived radicals. relationship between cardiovascular disease and vitamin E
Thus, there are severallines of evidence that in diabetes and was also studied in large scale prospective cohort studies,.
even in intermediate hyperglycaemic episodes endothelial In the largest prospective study - the Nurses ' Health Study
ceHs and presumably also other vascular cells generate with 121,700 U.S. female nurses [53]- the relative risk for
superoxide anions which together with oxidised LDL have those with the highest intake quintile compared with the
two major consequences (Fig. 7): lowest was 0.66, a 34% reduction in risk. In a similar

endothelial ceHs TNFa,IL-l


smooth muscle cells (NF-KB-activation)
0-
2 M-CSF, GM·CSF,
c-myc
(proliferation)

MCP-l
(Chemotaxis)

ox-LDL macrophages
VCAM-l,ICAM-l
(Adhesion)
monocytes
Tcells
TF
(Thrombogenesis)

Fig. 5. Activation ofthe transcriptions factor NF-KB which may be involved in changes in gene expression and cellular hemostasis that may be of importance
in the pathogenesis of atherosc1erosis.
108

Fig. 6. Activation ofNF-KB by high glucose in endothelial cells. Human umbilical vein endothelial cells (HUVEC) were isolated and incubated with various
concentrations of glucose as described. After 4 h the activated NF -kB was detected by immunofluarescence using an antibody specific for the activated form
ofNF-KB which is translocated into the nucleus: (al controls incubated with low glucose (5 mM) far 4 h; (b) HUVECS incubated with high glucose (30 mM)
for4 h.

analysis, in the Health Professionals Follow-up the relative Clinical evidence derived from primary and secondary
risk of coronary heart disease or death for those in the prevention trials
highest vitamin Eintake quintile was 0.6 [54]. Similar results At this time, two trials with respect to primary prevention
were obtained in further studies such as the Iowa Women's have been competed, one amongst a population of middle-
Health Study [55]. The postulated benefits are most likely aged male smokers in Finland [56] and the other among a
to be small to moderate in size, in the order of 20-40% general, but poorly nourished population in Linxian, China
reduction in cardiac risk, but nevertheless of large sig- [57]. In both studies no significant reduction of cardiac risk
nificance in such a common and serious disease as coronary could be obtained. However, it has to taken into account that
heart disease. For diabetics who represent a vascular high the doses of vitamin E (50 and 30 mg, respectively) were
risk population, a similar benefit is to be expected, but presumably too small to confer a benefit on cardiovascular
specific data are not available at this time. risk. In addition, coronary heart disease is rare in the
Chinese population and it is difficult to compare effects
of supplementation with antioxidants in poorly nourished
Hyperglycemia population with those in the Western societies. Specific data
for the diabetic population are still missing.
~ With respect to secondary prevention controversial
observations have been reported [58-60]. However recently,
the results of the Cambridge Heart Antioxidant Study
(CHAOS, [61]) have been given some new insight. In a placebo
controlled study, 2002 patients with angiographically proven
coronary atherosclerosis werc trcated with vitamin E (400
and 800 IU/day) or placebo far up to 2 years. A statistically
significant reduction in the risk of the primary endpoint of
non-fatal myocardial infarction plus cardiovascular death
was achieved. Since the number of cardiac deaths was not
diminished by vitamin E, the 77% reduction in risk is entirely
due to a reduction of the number of non-fatal myocardial
infarcts. Again, specific data for the diabetic population are
missing.
Fig. 7. Impact of oxidative stress on the regulation of coronary tlow and Thus, there is some reliable, but not conclusive evidence
cellular hemostasis. that treatment with vitamin E can reduce the cardiovascular
109

risk in the non-diabetic population. A similar protection is regulation of coronary flow and cellular hemostasis and
expected for diabetic patients, but not yet proven. Ihis finally in the pathogenesis of atherosc1erosis. Ihis hy-
conelusion is supported by two further lines of evidence: pothesis is also in line with epidemiological data based
on the observation of populations of remarkable size.
I. Clinical studies with other endpoints than the reduction
Lastly, but very importantly, c1inical trials like tlJe CHAOS
of the cardiovascular risk: Hodis et al. [62] coneluded
study with elinical end points such as re-infarct incidence
from their studies on 156 men with previous coronary
give a strong indication that antioxidant treatment, al-
artery bypass graft that there is elose association between
though only partly understood, may have a large impact
supplementary vitamin E and the angiographically dem-
for the treatment of cardiovascular diseases in the future.
onstrated reduction in coronary lesion progression. Simi-
larly, De Maoi et al. [63] demonstrated a reduction in the
number of restenosis in patients in whom the coronaries
had been dilated by PICA in dependence of the plasma.
Acknowledgement
levels ofvitamin E.
Ihis work was supported by the Ministerium für Frauen,
2. In addition to the epidemiological and clinical evidence
Familie und Gesundheit der Bundesrepublik Deutschland
much supporting evidence has been presented with respect
and the Wissenschaftsministerium des Landes NRW, the
to the various discussed mechanisms contributing to the
Deutsche Forschungsgemeinschaft, Bonn, und the 'Klinische
development of coronary heart disease (Iable 2). Iaken
Zellbiologie und Biophysik' e.v., Düsseldorf.
together, we begin to understand why oxygen derived radi-
cals are generated in diabetes and by which mechanisms
the highly reactive compounds contribute to the damage References
of the myocardium and especially of the coronary sys-
tem in diabetes. Ihis hypothesis is strongly supported by I. Schemthaner G: Cardiovascular mortality and morbidity in type-2
a substantial amount of experimental and clinical obser- diabetes mellitus. Diab Res Clin Pract 31 (suppl): S3-S 14, 1996
vations demonstrating that antioxidants are able to inhibit 2. Laakso M, Lehtoo S, Penttilä I, Pyörälä K: Lipids and lipoproteins
specific pathophysiological mechanisms involved in the predieting coronary heart disease mortality and morbidity in patients
with non-insulin-dependent diabetes. Circulation 88: 1421-1430, 1993
3. Kannel WB, McGee DL: Diabetes and cardiovascular risk factors. The
Table 2. Effects ofvitamin E on reactions whichmay be ofimportance for Framingham Study. Circulation 1: 8--13, 1979
prevention and inhibition of the development of macrovascular 4. Pyörälä K, Laakso M, Uusitupa M: Diabetes and atherosclerosis: An
complications (reviews [20, 38, 63]) epidemiologieal view. Diabetes Metab Rev 3: 463-524,1987
5. Uusitupa MIJ, Mustonen IN, Airaksinen KEJ: Diabetic heart muscle
Observation Reference disease. Ann Med 22: 377-386, 1990
6. Crepaldi G, Nosadini R: Diabetic cardiopathy: Is it areal entity. Diabetes
Inhibition ofLDL oxidation Estabaueretal.1991
Metab Rev 4: 273-288, 1988
Reaven et al. 1995
7. Rösen P, Pogatsa G, Tschöpe D, Addicks K, Reinauer H: Diabelische
Leonhardt 1997
Kardiopathie: Pathophysiologisehe Konzepte and therapeutische
Reduced lipid peroxidation Burton and Ingold 1986
Ansätze. Klin Wochensehr 69( suppl XXIX): 3-15, 1992
Ingold et al. 1987
8. Regan TI, Lyons MM, Ahmed S, Levinson G, Oldewurtel H, Ahmed
Diminished generation ofthromboxane A2 Karpenetal.1981,1985
M, Haider B: Evidence for cardiomyopathy in familial diabetes mellitus.
Inhibition of platelet aggregation and Jandaletal. 1988, 1989
JClin Invest 60: 885--899,1977
adhesion
9. Hamby R, Zoneraieh S, Sherman S. Diabetie eardiomyopathy. JAMA
Redueed expression of adhesion proteins Cominacini et al. 1997
229: 1749-1754,1974
(ICAM-I, VCAM-l, P-seleetin)
10. Shapiro LM, Leatterdale BA, Mackinnon J, Fleteher RF: Left ventricular
Diminished adhesion of monocytes Faruqietal.1994
funetion in diabetes mellitus. BrHeart J 45: 129-132,1981
Erletal.1997
11. The Diabetes Control and Complication Trial Research. The effect of
Martini et al. 1997
intensive treatment of diabetes on the development and progression
Protection of endothelium: increased Karpenetal.1981
oflong-term complications in insulin-dependent diabetes mellitus. N
formation of prostacylin and NO Keegan et al. 1995
Engl J Med 329: 977-986, 1993
Rösenetal.1995
12. The Diabetes Control and Complicalion Trial Research. Effect of
Anderson et al. 1995
intensive therapy on the development and progression of diabetic
Inhibition ofSMC proliferation in SMC Azzietal.1993
nephropathy in the Diabetes Control and Complication Trial. Kidney
Boscoboinik et al. 1994
Int47: 1703-1720, 1995
Fazzioetal.1997
13. Rösen P, Ballhausen T, Bloch W, Addicks K. Endothelial relaxation is
Inhibition ofPKC activation in SMC Azzietal.1993
disturbed by oxidative stress in the diabetie rat heart: Influence of
Boscoboinik et al. 1994
tocopherol as antioxidant. Diabetologia 8: 1157-1168, 1995
Fazzioetal.1997
14. Giugliano D, Ceriello A, Paolisso G: Diabetes mellitus, hypertension,
Inhibition ofthe activation ofNF-KB Suzuki and Packer 1993
and cardiovaseular disease: Which role for oxidative stress? Metabolism
Erletal. 1997
44: 363-368,1995
110

15. Baynes JE: Role of oxidative stress in the development of complications 35. Graier W, Simececk S, Kukovetz WR, Kostner GM: High D-glucose
in diabetes. Diabetes 40: 405-412, 1991 induced changes in endothelial Ca'+/EDRF signalling are due to
16. Nourooz-Zadeh J, Tajaddini-Sannadi J, McCarthy S, Betteridge DJ, generation ofsuperoxide anions. Diabetes 45: 1386-1395,1996
Wolff SP: Elevated levels of authentie plasma hydroperoxides in 36. Hunt JA, Smith CCT, WolffSP. Autoxidative glycosylation and possible
NIDDM.Diabetes44: 1054--1058, 1995 involvement of peroxides and free radicals in LDL modification by
17. Nourooz-Zadeh J, Halliwell B, Tritschler H, Betteridge DJ: Decreased glucose. Diabetes 39: 1420-1424, 1990
lipid standardised plasma a-tocopherol in non-insulin-dependent 37. Giardino I, Edelsteine 0, Brownlee M: BCL-2 expression or antioxidants
diabetes mellitus. Diabetes & Stoffw 6(suppI2): 20-23,1997 prevent hyperglycemia induced formation of intracellular advanced
18. Nishigaka I, Hagihara M, Tsunekawa H, Maseki M, Yagi K: Lipid glycation endproducts in bovine endothelial cells. J Clin luvest 97:
peroxide levels of serum lipoprotein fractions of diabetic patients. 1422-1428, 1996
Biochem Med 25: 373--378, 1981 38. Yan SO, SchmidtAM,Anderson GM, Zhang J, Brett J, Zou YS, Pinsky
19. Bellomo G, Maggi E, Polli M, Agosta FG, Bollati P, Finardi G: 0, Stern 0: Enhanced cellular oxidant stress by the interaction of
Autoantibodies against oxidatively modified low density lipoproteins advanced glycation end products with their receptors/binding proteins.
in NIDDM. Diabetes 44: 60-66, 1995 JBiol Chem269: 9889-9897,1994
20. Rösen P, Tchöpe D. Vitamin E and Diabetes. Fat Sci Technol 11: 25- 39. Lyons TJ: Oxidised low density lipoproteins: A role in the pathogenesis
431,1991 of atheroselerosis in diabetes? Diabetes Med 8: 411-419, 1991
21. Gopaul NK,Anggard EE, MalletAI, Betteridge DJ, WolffSP, Nourooz- 40. Dietrich 0, Skepec J, Diederich A, Dai FX: Endothelial dysfunction in
Zadeh J: Plasma 8-epi-PGF2a levels are elevated in individuals with mesenteric resistance arteries of diabetie rats: Role offree radicals. Am
N1DDM FEBS Let! 368: 225-229, 1995 J Physiol 266: H1153-H1161, 1994
22. Davi G, MezzettiA, Vitacolonna E, Costantini F, Pennese E, FalcoA, 41. Pieper GM, Gross GJ: Oxygen free radicals abolish endothelium
Ciabattoni G, Patrono C, Consoli A: In vivo formation of 8-epi- dependent relaxation in diabetic rat aorta. Am J Phyiol 255: H825-
prostaglandin F2a in diabetes mellitus. Effects of tight control and H833,1988
vitamin E supplementation. Diabetes 46(suppll): 113A, 1997 42. Tesfamariam B. Free radieals in diabetic endothelial cell dysfunction.
23. Jain SK, McVie R, Duett J, Herbst JJ: Erythrocyte membrane lipid Free Radical Biol Med 16: 383--391, 1994
peroxidation and glycosylated hemoglobin in diabetes. Diabetes 38: 43. Cohen RA: Dysfunction ofvascular endothelium in diabetes mellitus.
1539--1543,1989 Circulation 87(suppl V): V67-V76, 1993
24. Bellomo G, Maggi E, Palladini G, Perugini, Seccia M: Oxidation oflow 44. Nitenberg A, Valensi P, Sachs R, Dali M, Aptecar E, Attali JR:
density lipoproteins and vitamin E status in non insulin dependent Impairment of coronary vascular reserve and Ach induced coronary
diabetes mellitus (NIDOM). Diabetes & Stoffw 6(suppI2): 29--33, 1997 vasodilation in diabetic patients with angiographically normal coronary
25. Leonhardt W, Hanefeld M, Lattke P, Jaroß W: Vitamin E Mangel und arteries and normalleft ventricular systolic function. Diabetes 42: 10 17-
Oxidierbarkeit der Low-Density-Lipoproteine bei Typ I und Typ II 1025,1993
Diabetes: Einfluß der Qualität der Stoffwechselkontrolle. Diabetes & 45. Jessup W: Oxidised lipoproteins and nitric oxide. Curr Opin Lipidology
Stoffw 6(suppI2): 24--28, 1997 7: 274--280,19%
26. Reavan PD, Herold DA, Barnett J, Edelmann S: Effects ofvitamin E on 46. Brown AJ: Atheroselerosis: cell biology and lipoproteins. Curr Opin
susceptibility ofLDL and LDL subfractions to oxidation and on protein Lipidology7: Ul62-U167, 1996
glycation inNIDDM. Diabetes Care 18: 807-S16, 1995 47. Brand K, Page S, Rogler G, Bartsch A, Brand! R, Knuechel R, Page M,
27. Simon-Schnass I, Rosak Ch, Tritschler HJ, Rösen P: Alpha-Toco- Kaltschmidt C, Baeuerle PA, Neumeier D:Activated transcription factor
pherolaufuahme und - zufuhr bei Typ II Diabetikern. Diabetes & Stoffw nuelear factor-kappa B is present in atherosclerotic lesion. J Clin Invest
6(suppI2): 16-19, 1997 97: 1715-1722,1996
28. Karpen CW, Cataland S, D"Orisiso D, Panaganamal RV: Production of 48. SchmidtHH, Walter U: NO at work. Ce1l78: 919-925, 1994
12-HETE and vitamin E status in platelets oftype I diabetes subjects. 49. Rösen P, Hönack C, Müssig K, Bloch W, Addicks K: Influence ofATI
Diabetes 34: 526-531 receptor inhibition on cardiac function and structure of diabetic rats.
29. Kashiwagi A, Asahina T, Ikebuchi M, Tanaka Y, Takagi Y, Nishio Y, In: NS Dhalla, P Zabradka, IMC Dixon, RE Beamish (eds). Angiotensin
Kikkawa R, ShigetaY: Abnormal glutathione metabolism and increased II Receptor Blockade: Physiological and Clinical Implications. Kluwer
cytotoxicity caused by H,ü, in human umbilical vein endothelial cells Academic Press, Boston, (in press)
cultured in high glucose. Diabetologia 37: 264--269, 1994 50. Ross R: The pathogenesis of atherosclerosis: A perspective for the
30. Maxwell SRJ, Thomason H, Sandler D, Leguen C, Baxter MA, Thorpe 1990s. Nature 362: 801-S09, 1993
GHG, Jones AF, Bamett AH: Antioxidant status in patients with 51. Gey KF, Puska P, Jordan P, Moser UK: Inverse correlation between
uncomplicated insulin-dependent and non-insulin-dependent diabetes plasma vitamin E and mortality from ischemic heart disease in cross-
mellitus. Eur J Clin !nvest 27: 484-490, 1997 cultura1 epidemiology.AmJ Clin Nutr 53: 326S-334S, 1991
31. Minor RL, Myers PR, Guerra R, Bates JN, Harrison DG: Diet-induced 52. Riemersma RA, Oliver M, Elton RA,Alfthan G, Vartiainen E, Salo M,
atherosclerosis increases the release of vascular relaxing factor. J Clin Rubba P, Mancini M, Georgi H, Vuillewnier JP et al.: Plasma antioxidants
!nvest 86: 2109--2116,1990 and coronary heart disease: Vitamins C and E, and selenium. Eur J Clin
32. Keaney JF, Gaziano JM, Xu A, Frei B, Curran-Celentano J, Shwaery Nutr44: 143--150, 1990
GT, Loscalzo J, Vita JA: Low dose a-tocopherol improves and high 53. Stampfer MJ, Hennekens CH, Martison JE, Colditz GA, Rosner B, Willet
dose a-tocopherol worsens endothelial vasodilator function in WC: Vitamin E consumption and the risk of coronary disease in woman.
cholesterol-fed rabbits. J Clinlnvest 93: 844-S5 I, 1994 NEng1JMed328: 1444--1449,1993
33. OharaY, Peterson TE, Harrison DG: Hypercholesterolemia increases endo- 54. Rimm EB, Stampfer MJ,AscherioA, Giovannuci E, Colditz GA, Willet
thelial superoxide anion production. J Clin Invest 91: 2546-2551, 1993 WC: Vitamin E consumption and the risk of coronary heart disease in
34. Du XL, Sui GZ, Stockklauser-Färber K, Weiß J, Zink S, Schwippert B, men. NEnglJ Med 328: 1450-1456, 1993
Wu QX, Tscböpe 0, Rösen P: Induction of apoptosis GI high pro 55. Kushi LH, FolsomAR, Prineas RJ, Mink PJ, Wu Y, Bostick RM: Dietary
insulin and glucose in human umbilical vein endothelial cells is mediated antioxidant vitamins and death from coronary heart disease in
by reactive oxygen species. Diabetologia 41 : 249--256, 1998 postmenopausal women. N EnglJ Med 334: 1156-1162, 1996
111

56. The Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study Group: 61. Stephens NG, Parsons A, Schofield PM, Kelly F, Cheeseman K,
Tbe effect of vitamin E and beta carotene on the incidence oflung cancer Mitchinson M, Brown MI: Randomised controlled trial ofvitamin E in
and other cancers in male smokers. N EnglJ Med 330: 1029--1035, 1994 patients with cornnary disease: Cambridge Heart Antioxidant Study
57. Blot WI, Li IY, Taylor PRet al.: Nutrition intervention trials in Linxian, (CHAOS). Lancet347: 781-786, 1996
China: Supplementation with specific vitamin/mineral combinations, 62. Hodis HN, Mack WJ, LaBree L et al.: Serial coronary angiographic
cancer incidence, and disease specific mortality in the general evidence that antioxidant vitamin intake reduced progression of
population. JNatl Cancer Inst 85: 1483- 1492,1993 coronary artery atherosclerosis. JAMA 273: 18491854, 1995
58. Haeger K: Long-time treatment ofintermittent claumcation with vitamin 63. DeMaio SJ, King SB, Lembo NJ, Roubin GS, Hearn JA, Bhagavan HN,
E.AmJClinNutr27: 1179-1181, 1974 Sgoutas OS: Vitamin E supplementation, plasma lipids and incidence
59. Livingston PD, Jones C: Treatment of intermittent claudication with of restenosis after percutaneous transluminal coronary angioplasty
vitamin E. Lancet2: 602--{'i04, 1958 (PTCA).J AmCollNutr 11: 68--73,1992
60. Williarns HTG, Fenna D, MacBeth RA: Alpha-tocopherol in the treatment 64. Traber MG, Packer L: Vitamin E: Beyond antioxidant function. Free
ofintermittent claudication. Surg Gynecol Obstet 132: 662--{'i66, 1971 Rad Biol Med 19: 227-250,1995
Molecular and Cellular Biochemistry 188: 113-126, 1998.
© 1998 Kluwer Academic Publishers.

Cardiovascular disease in the JCR:LA-cp rat


James C. Russell,l Sandra E. Graham l and Mary Richardson2
lDepartment 0/ Surgery, University 0/Alberta, Edmonton, Alberta; 2Department 0/ Pathology, McMaster University,
Hamilton, On ta rio, Canada

Abstract
The JCR:LA-cp rat is one of a nwnber of strains that carry the mutant autosomal recessive cp gene.Animals, of all strains, that
are homozygous, for the gene (cp/cp) become obese, insulin resistant, and hypertriglyceridemic. Heterozygotes or homozygous
normal rats (+/+)are lean and metabolically normal. The JCR:LA-cp rat is unique in the development of a frank vasculopathy
with atherosclerotic lesions and associated ischemic myocardiallesions. The cardiovascular disease is strongly correlated with
the hyperinsulinemia, which develops as the animals mature from 4 to 8 weeks of age. The hyperinsulinemia can be decreased
by marked food restrietion, ethanol conswnption, or reduction of the postprandial glucose and insulin responses through the
use ofa-glucosidase inhibitors.Any treatment that reduces plasma insulin levels is associated with a reduction in cardiovascular
disease. In contrast, a reduction in plasma triglyceride concentrations, alone, has no effect on end-stage lesions. JCR:LA-cp
rats, particularly those that are cp/cp, are, however, sensitive to cholesterol in the diet, unlike other strains that are highly resistant.
Further, the rats have abnormal vascular smooth muscle cells that, especially in the cp/cp animals, are hyperplastic and activated
and migrate into the intimal space. Dur findings suggest that susceptibility to cardiovascular disease requires hypermsulinemic
stress coupled with excessive dietary intake and the presence of one or more other necessary, but not sufficient, genetic factors.
One of these may be a genetic abnormality of vascular smooth muscle cells. A similar situation may occur in humans. (Mol
Cell Biochem 188: 113-126, 1998)

Key words: insulin resistance, hyperlipidemia, cardiovascular disease, smooth muscle cells, JCR:LA-cp rat

factorial and undoubtedly polygenetic in nature, with im-


Introduction portant components of interaction between genetic pre-
disposition and the environment. The elucidation of the
The syndrome characterized by obesity, insulin resistance, components of the complex processes leading to insulin
and hypertriglyceridemia is strongly associated with cardio- resistance and vascular disease is an important scientific
vascular disease. This 'Metabolie Syndrome' is apparently problem. Real progress will undoubtedly require basic studies
increasing in frequency in prosperous societies, with the on appropriate animal models in order to identify the factors
underlying mechanisms beingpoorly understood [1-3]. The and mechanisms, followed by confirmatory work on human
development of a marked insulin resistance with resultant subjects.
hyperinsulinemia is a central part of the syndrome, yet the The JCR:LA-cp rat is a unique strain that develops obesity,
causal roles of abnormal lipid metabolism with hyper- insulin resistance, hyperlipidemia, vasculopathy, and athero-
triglyceridemia and elevated insulin levels are not clear. sclerosis ifhomozygous, for the autosomal recessive cp gene
Further, not all obese individuals develop the Metabolie (cp/cp) [5-10]. Rats that are heterozygous (cp/+) or homo-
Syndrome, although many do; and not all those with the zygous normal (+/+) are lean and metabolically normal.
syndrome, even in the extreme form with frank type II These unusual animals provide a model of the Metabolie
diabetes, develop cardiovascular disease [4]. These apparent- Syndrome in hwnans that is characterized by abdominal
ly puzzling characteristics indicate that the syndrome and the obesity, mild type II diabetes, and high risk for cardiovascular
associated vascular disease are not simple, single cause-, disease. They exhibit all aspects of the syndrome seen in
perhaps single gene-, determined. Rather, they are multi- hwnans, including vasculopathy, intimaliesions, and myo-

Address for ofJj;rints: J.C. Russen, Department of Surgery, 275 Heritage Medica1 Research Centre, University of Alberta, Edmonton, Alberta T6G 2S2,
Canada
114

cardial damage. Ofthese, the development of cardiovascular weight (to that ofthe +/+ animals) up to 10 weeks ofage, a
disease is unique among rodent models and represents the normalization ofplasma triacylglycerol concentrations, and
critical complication of insulin resistance and type II diabetes. skeletal muscle triacylglycerollevels identical to those ofthe
We present evidence that the profound insulin resistance +/+ rats (Russell et al., unpublished observations). In these
ofthe cp/cp male rat is due to metabolic changes consequent treated rats plasma insulin concentrations rose only modestly
to the presence ofthe cp gene. There is, however, an inter- (to 150 mUll), with an age at half-maximum rise of8.5 weeks.
action with the rest of the genome, as females are less severely These fIndings demonstrate a clear link between abnormal
affected than male rats, and the effects of the cp gene are not lipid metabolism and the development of the insulin re-
identical in even closely related strains of rats [11-14]. The sistance. We are not yet able to show a causal relationship at
development of the vascular abnormalities and dysfunction this stage, and the ultimate defect may lie elsewhere.
is clearly multifactorial as related strains of rats that in- The hyperinsulinemia of the cp/cp rat is a direct con-
corporate the cp gene are obese, insulin resistant, and hyper- sequence of the profound hypersecretion of insulin. In
triglyceridemic [15-17], but do not show the vasculopathy comparison to the fatty (fa/fa) Zucker rat, the cp/cp rat shows
and ischemic damage that are characteristic ofthe JCR:LA-cp a more extreme basal hyperinsulinemia, but, unlike the fa/fa
rat [12-14]. Thus, there are factors otherthan thecp gene that rat, does not respond to insulin secretegogs, such as arginine
lead the JCR:LA-cp atiimals to be sensitive to the hyper- and gastric inhibitory polypeptide, at low glucose con-
insulinemic, hyperlipidemic state. While these factors are centrations. The cp/cp rat does respond to the secretegogs at
presently unknown, an abnormality of the vascular smooth high glucose levels with an extreme secretion of insulin
muscle cells (SMC) does appear to be important [18]. (approximately 4-fold greater than in the fa/fa rat) [11]. The
two strains of rat have different mutations in the gene for the
leptin receptor (ObR), the Ja mutation being a substitution
Abnormal metabolism oJ the JCR:LA -cp rat leading to an amino acid change at codon 269 in the extra-
cellular domain ofthe receptor, and the cp mutation leading
The insulin resistance of the cp/cp rats is not present at 3 to a stop codon at 763, also in the extracellular domain [23,
weeks of age, when the animals are weaned, and plasma 24]. The Ja mutation gives rise to a leptin resistant state due
insulin concentrations are only slightly elevated at tbis age. to impaired receptor binding of the leptin, while the cp
At 4 weeks of age the skeletal muscle of the cp/cp rats has mutation may be expected to lead to an absence of receptors.
an effective insulin-mediated glucose uptake essentially Leptin has been reported to inhibit the pancreatic release of
identical to that ofthe +/+ rats. By 8 weeks of age, however, insulin through a receptor-mediated mechanism [25]. Thus,
the skeletal muscle no longer has any insulin-mediated the two strains of rat may be expected to exhibit different
glucose uptake, while basal (non-insulin-mediated) uptake abnormalities ofboth food intake (a central nervous system
is unchanged and is not different from that of the lean rats. effect) and insulin metabolism (a visceral organ effect). The
This change is reflected in plasma insulin levels in the normal cp/cp rat has a markedly greater ß-cell activity of the
fed state which are only slightly elevated (about 50 mUll) in glucokinase that is believed to be the glucosensor of the
the cp/cp rats at 4 weeks of age. There is a rapid rise in insulin pancreas [26]. The reduction of plasma insulin levels through
concentration to 375 mUll by 7 weeks of age, with an age at partial pancreatectomy leads to markedly increased plasma
half-maximum increase of5.5 weeks. This makes it plain that glucose concentrations and glucosuria, a frankly diabetic state
the insulin resistance is not due to a simple, fIxed, genetically (Russell and Grabam, unpublished observations). The rate of
determined defect, but is a metabolic response to another clearance of insulin from the circulation is reduced in the cp/
dysfunction that is determined by the cp gene. cp rat [6] and its binding to liver and muscle is reduced about
Plasma triacylglycerol concentrations are already 4-fold 40% (Hopkins, Williams and RusselI, unpublished observa-
greater in the cp/cp males than in the +/+ rats at 4 weeks of tions). In some way, the cp/cp rat is able to compensate for
age and continue to rise thereafter to very high values (6 mM the profound insulin resistance of the peripheral tissues
at 12 weeks of age) [19, 20]. The triacylglycerol content of through the hyperinsulinemia. This is accomplished at the
skeletal muscle is 2-fold greater in the cp/cp rats at 4 weeks expense of a hyperplasia of the islets ofLangerhans such that
of age, preceding the development of the insulinresistant state up to 20% of the pancreatic volume is occupied by islet tissue
(Shillabeer et al., unpublished observations). Treatment of the [27]. It is also at the expense of other secondary metabolic
cp/cp rats with the highly effective hypotriglyceridemic complications and a pathophysiology ofthe cardiovascular
agent, MEDICA 16 (ß,ß' -tetramethylhexadecanedioic acid), system.
even when the animals are fully insulin resistant; reduces The hypertriglyceridemia ofthe cp/cp rat is due to a marked
plasma triglycerides by 80%; halves fasting insulin levels; hepatic hypersecretion of very-low-density lipoprotein
and is antiatherosclerotic and cardioprotective [21, 22]. Rats (VLDL), but without any increase in the secretion ofhigh-
treated from 3 weeks of age show a normalization of body density lipoprotein (HDL) [20, 28]. The excessive VLDL
115

secretion is a consequence ofthe diversion of the large flow Pathophysiology ofthe vessel wall
of glucose from the increased intake of a carbohydrate diet
by the hyperphagic animals to fatty acid and triacylglycerol The most serious dysfunction in the cp/cp rat lies in the
synthesis [29]. With the essentially totally insulin resistant vascular system. This is evident at various levels, ftom the
peripheral tissues unable to respond to the high insulin levels medial SMC to the intima, to the fibrinolytic system, and to
with the needed uptake of glucose, there is no real alternative the contractile/relaxant function of both large and small
to hepatic uptake of the glucose. In addition, insulin resistance arteries. The dysfunctional characteristics are such as to
leads to a failure of the inhibition of VLDL synthesis in predispose to vasospasm, thrombosis, and ischemic injury.
response to the signal ofhigh insulin levels [30]. The resulting Both arterial intimal arterial injury and cardiac ischemic
triacylglycerol, in turn, must be secreted as VLDL by the liver lesions are seen in these animals, confirming the importance
as it does not have the capacity to store the resulting fat. The of the vascular dysfunction.
VLDL is hydrolyzed and the triacylglycerol taken up by the At the level ofthe vessel wall cells, we have found that the
various fat depots, leading to the severe obesity of the cp/cp SMC appear to be abnormal in ultrastructure. As shown in
rat. The liver shows some mild fatty accumulation and a Fig. 1, the SMC are highly activated and of an irregular shape
marked disruption of the hepatocyte mitochondria, pre- and are present in significant numbers in the intimal space.
sumably in response to the abnormal metabolic drive. When grown ftom aortic explants, these SMC remain hyper-
The cp/cp rats are hyperphagic at weaning, consistent with plastic for at least 9 passages. They are also hyperresponsive
our finding of increased levels of neuropeptide Y (NPY) in to the cytokines, TGF-ß, bFGF, and IGF-l [18]. At a ftmct-
the arcuate nuc1eus of the hypothalamus [31]. This, in turn, ional level, both aortic and mesenteric resistance artery
is consistent with an inability ofleptin to downregulate NPY, exhibit hypercontractility in response to phenylephrine and
and thus food intake, in these animals. However, the under- norepinephrine (McKendrick, O'Brien, Radomski and
lying mechanisms are more complex, as even severely RusselI, unpublished observations). The reduction of nitric
food-restricted cp/cp rats maintained at 60% ofthe intake of oxide synthase (NOS) activity by N-nitro-L-arginine methyl
+/+ control animals retain the obese phenotype as weil as both ester (L-NAME) administered in vivo results in a greater
their hyperlipidemia and hyperinsulinemia, albeit much inhibition of acety1choline-mediated vascular relaxation in
reduced [7]. Even very intensive exercise with marked food the cp/cp rats than in the +/+ animals. Thus, in addition to
restriction, while beneficial, does not normalize the cp/cp rat the abnormal vasospastic characteristics of the vessel wall,
[32, 33]. cp/cp rats also have an impaired endothelial function. A

Fig. 1. Transmission electron micrograph of a section from the aortic arch of a 9·month-old male cp/cp male rat showing intemal elastic lamina (iel) and
highlyabnormal SMC (arrowheads) and abundant connective tissue in the intimal space. Bar; 50 ~m.
116

further indication of endothelial dysfunetion is the increased Fig. 3, advaneed lesions exhibit all the elements seen in
expression ofmRNA for plasminogen activator inhibitor-I atherosclerotie lesions in humans. These raised intimal
(PAI-I) in the aortas of ep/ep rats, together with signifieantly lesions eontain lipid and abundant eonneetive tissue, includ-
elevated plasma eoneentrations ofPAI-1 (Schneider, Absher, ing collagen and proteoglyean; disturbed endothelium; and
Russell and Sobel, unpublished observations). Plasma aetivated SMC (Fig. 3). Unlike other strains of rats, the
eoneentrations ofPAI-l were strongly eorrelated with plasma JCR:LA-ep rat is highly sensitive to dietary eholesterol,
eoneentrations of both insulin and triaeylglyeerides. These developing exaeerbated intimaliesions when fed a standard
vaseular dysfunetions may be expeeted to lead to increased rat ehow diet supplemented with 0.25% eholesterol from 6
vaseular eontraetility, a propensity to vasospasm in response to 26 weeks of age.A transmission eleetron mieroseope image
to stress, and increased fibrin aeeumulation within the of a typieallesion in the aorta of a eholesterol-fed rat whieh
eirculation, all ofwhieh have been observed in the ep/ep rat includes extensive medial smooth muscle involvement with
[8, 13,34]. intra- and extraeellular lipid deposition is shown in Fig. 4.
The vessels of the ep/ep rats also show extensive areas of
adherent maerophages, many of whieh are clearly aetivated
Vascular and myocardial lesions and engaged in entering the intimal space through the
endothelium (Fig. 5).Adherent thrombi are also seen, ranging
The male ep/ep rat ofthe JCR:LA-cp strain is unique among from large, organized struetures (Fig. 6) to smalI, micro-
rats in the development ofboth intimal vascular lesions and thrombi (Fig. 5), eonsistent with the known elevation ofPAI-l
lesions in the myocardium that are evidently of isehemie activity in the cp/ep rat (Fig. 6).
origin. The lesions of the vessel wall appear to initially eonsist The oceurrence of extensive fibrin deposits in the lumen
of abnormal SMC that are aetive and migrate to the intimal ofthe major arteries would imply that occlusive thrombotie
spaee in both +/+ and ep/ep rats from an early age (Fig. 1). episodes occur in smaller vessels, such as the coronary
The phenomenon is considerably more severe in the ep/ep arteries. Clearly, such episodes must be relatively infrequent,
rats than in the +/+ animals. Figure 2 shows a seanning or ep/cp males would not survive for very long. Nonetheless,
eleetron mierograph ofthe aortie areh with extensive raised occlusive thrombi have been found in the pancreatie artery
intimaliesions on the lesser eurve of the areh. As shown in [34] and Fig. 7 shows an occlusive thrombus in the eoronary

Fig. 2. Scanning electron micrograph ofthe aortic arch of a 9-month-old male cp/cp rat. The cut edge ofthe lesser curve is evident at the bottom of the image.
Above this is an extensive area of raised intimaliesions covered with abnormal, but intact, endothelium.
117

Fig. 3. Transmission eleetron micrographs ofa seetion from the aortie areh ofa 9-month-old male cp/ep rat, illuslraling (a) an intimalIesion with lipid-eontaining
foam cells (fe), collagen, and proteoglycan matrix; and (b) an area ofdegenerating endothelium. (e) adjacent to a foam eell with a platelet and red blood eell
in the luminal spaee. Bar = 1 ~m; iel, intemal elastie lamina.
118

Fig. 4. (al Transmission eleetron mierograph of a typical sec!ion from the aortic areh male ep/ep rat, 6 months of age, that was fed a standard rat chow
supplemented with 0.25% cholesterol from 3 weeks. The intimal space is enlarged and contains a necrotic foam ce 11. There is intra- and extracellular lipid
(arrowheadsl in the media, and the character ofthe vessel wall is markedly degraded in comparison to that ofrats fed control food or ofage-matched lean
animals fed the eholesterol-enriched diet, an example of which is shown in Fig. 4b. Bar =50 Ilm; iel, internal elastic lamina.
119

Fig. 5. Scanning electron micrograph of an area of the aortic arch of a 9-month-old male ep/cp rat. There are numerous macrophages adherent to the
endothelial surface, which is uneven and appears to overlie the region of an intimaliesion. The macrophages are activated and some are in the process of
penetrating the endothelium into the intimal space. There is also a mature microthrombus visible in the lower centre ofthe field whieh includes trapped blood
cells.

Fig. 6. Scanning electron micrograph showing the proximal portion of one of the branches of the aortic arch of a 9-month-old male cplcp rat. The arch has
been cut along the greater curvature and the branchiocephalic artery has been spli!. The Jeading edge ofthe ostium is evident on the right side of the image,
with the branch extending away, cut edges upward. There is a large, mature thrombus adherent to the surface ofthe branch. This is a stable structure with
some macrophages, platelets, and erythrocytes visible on the surface and at higher magnification.
120

Fig. 7. Thrombus in the eoronary artery ofa 9-month-old male ep/ep rat. The eoronary arteries ofthe rat lie intramurally, unlike those ofdogs, pigs, and
humans, that lie on the surfaee ofthe myocardium. The lumen is completely occ1uded by an organized thrombus. Hematoxylin and eosin stain, x50.

artery of a 9-month-old male ep/ep rat. Sueh a thrombus must lesions with prominent collagen bands and represent the
have led to a signifieant distal region of isehemia and end-stage of the reparative proeess at ages from 4 weeks
infaretion. and greater.

Examples of stage 2 and 4 lesions from 9-month-old ep/ep


Myocardial infarction male rats are shown in Fig. 8.
The observation of isehemie damage to the heart suggests
In light ofthe vaseular abnormalities seen in the ep/ep rat, it that the rats should exhibit 'elinieal' evidenee ofmyocardial
may be expeeted that some isehemic damage would follow infaretion episodes. This is diffieult to quantitate as a rapidly
and such is the ease. We have reported the presenee of fatal event leaves insuffieient time for histologieally evident
isehemie myoeardiallesions in ep/ep rats that are present at lesions to develop in the heart and there is no ready meeh-
3 months of age and inerease in frequeney up to 9 months of anism to elieit a eomplaint ofpain from a rat. Nonetheless,
age [10]. These lesions have been eategorized as: oeeasional spontaneous deaths do oeeur in the rats, espeeially
during or immediately following stressful proeedures, and a
Stage 1: areas of neerosis without signifieant ehronie few sueh oeeurrenees have been doeumented. Figure 9A
inflammatory eell infiltration. These are lesions with an shows an arterial blood pressure reeording from a ehronically
age of hours to one day. eannulated rat taken I min after handling. The animal
Stage 2: areas of eell lysis with aetive ehronic inflam- exhibited sudden distress and assumed a hunehed posture
matory eell infiltration. These lesions are days to a week with piloereetion. The pressure traee shows a serious arrhy-
old. thmia with aperiod of falling blood pressure after missed
Stage 3: nodules of ehronie inflammatory eells, without beats. Over aperiod of 2 h the eardiae rhythm gradually
any visible dead myoeytes. These may represent a loeation improved (Fig. 9B) and the rat survived. After 4 days the rat
where one or a few myoeytes have been Iysed and have was saerifieed and the heart was removed for histologie
disappeared. examination. Figure 10 shows an infareted area in the heart,
Stage 4: old, searred lesions with no signifieant remaining with a clear, large stage 2 lesion with extensive chronic
inflammatory eell aetivity. These are mature, eontraeted inflammatory cell infiltration. While this phenomenon has not
121

Fig. 8. Representative ischemic lesions ofthe heart, as described in the text. These lesions are from 9-month-old male cp/cp rats. (A), Stage 21esion showing
three different areas of varying ages, all with chronic inflammatory cell infiltration and celllysis. Hematoxylin and eosin stain, x 100. (B), an old, mature
stage 4 lesion, showing striated collagen bands and some isolated surviving myocytes. Masson's trichrome stain, x40.
122

2.5 5.0 7.5 10.0


time seconds

2.5 5.0 7.5 10.0


time seconds

100

2.5 5.0 7.5 10.0


time seconds

Fig. 9. B100d pressure recordings from the aorta of a female cp/cp rat, 12 weeks of age, previously fitted with a chronic indwelling cannula through the right
carolid artery. On Ihis day, the animal's cannula was connecled 10 apressure transducer and flushed with 100 111 of saline. The rat became distressed,
hyperventilated, and, as shown in (A), developed an arrhythmia. Three minutes laler the arrhythmia had worsened (B), bul then began 10 stabilize (C), and
over aperiod of 2 h resolved. The pressure also stabilized at approximately 120/90 mm Hg. The rat survived and was kept for 4 days.
123

Fig. 10. Section ofthe heart ofthe rat described in Fig. 9, showing arecent, active stage 21esion in the left ventricle. Hematoxylin and eosin stain, xIOO.

yet been made reproducible on demand, it does indieate that ethanol (0.5 or 4% in the drinking water) prevented the
the cp/cp rat is subject to myocardial infarction that strongly myoeardiallesions in ep/cp rats (37). While there were no
mimics that seen in some human patients. accompanying decreases in plasma lipids, plasma insulin
levels were significantly reduced and the marked hyperplasia
ofthe islets ofLangerhans was halved. Severe food restriction
Effects 01 dietary in take (to 60% ofthe intake of +/+ animals) markedly reduced the
frequency of the myocardial lesions, and food restriction
We have examined the effects of a range of dietary modifi- combined with prolonged running provided even greater
eations on the metabolism and pathophysiology ofthe ep/ep eardioprotection [7, 33). These treatments lowered lipid
rat and find that some are highly benefieial and others have levels and greatly redueed plasma insulin levels. The findings,
no effeet. Omega-3 fatty acids have been widely suggested taken together, strongly suggest that the critieal element in
to have eardioproteetive effeets in humans. Dietary suppl- the metabolie disorder ofthe cp/cp rat is the hyperinsulinemia
ementation of male ep/cp rats with 10% (w/w) of olive oil or some other factor related to the insulin resistance, with
lowered the plasma free cholesterol by 38% and triglycerides hyperlipidemia playing a secondary role.
by 28% [35). Redfish oil, a marine oil with a eomposition
heavily weighted to long chain monounsaturated fatty acids,
caused a very pronounced decrease in all lipid classes, with Pharmacological interventions
the exeeption of free cholesterol. The largest reduction was
in triglycerides, which were reduced approximately 50%. Pharmacological agents may be directed toward any of the
Despite the signifieant reduetion in plasma lipids induced by elements ofthe insulin resistancelhyperlipidemic syndrome,
the redfish oil supplement, there was no effect on insulin or to the vascular dysfunetion, andlor to the lesions. We have
glucose eoncentrations and no reduction in the frequency of found that interventions that improve insulin and glucose
myocardiallesions at 9 months of age (35). In contrast, metabolism are effective, while those that only lower plasma
supplementation ofthe rat diet with fructose (20% of caloric lipid concentrations are not. Agents that reduce vasospasm
content) resulted in a virtual prevention of the myocardial or arterial contraction are also protective.
lesions without any change in lipid, insulin, or glucose The anorectic compounds, benfluorex andD-fenfluramine,
concentrations (36). Similarly, chronic eonsumption of reduce food intake and lower body weight, with accom-
124

panying reductions in plasma lipids and insulin concent- plasma cholesterol levels have lesion-promoting effects,
rations [29, 38]. Benfluorex treatment delayed and reduced especially in light ofthe demonstrated cholesterol sensitivity
the hyperinsulinemia, improved insulin sensitivity, and of this strain. These results emphasize the importance of
reduced the size and histologie abnormalities of the islets of careful characterization of both animal models and study
Langerhans [39]. Very similar effects were seen with D-fen- populations in clinical studies.
fluramine treatment, and both agents induced a virtual It is not necessary to significantly improve the vas-
absence of myocardiallesions in treated animals [39] (RusseIl culopathy or arterial lesions in order to protect the heart
et al., unpublished observations). against ischemic damage. Angiotensin-converting enzyme
Amelioration of the effects of established insulin resistance (ACE) inhibitors are effective blood pressure lowering
can be accomplished through either food restrietion or a delay agents, through inhibition of the renin angiotensin system.
in absorption, both serving to reduce the postprandial rise in ACE inhibitors, however, also inhibit the degradation of
plasma glucose concentration. The a-glucosidase inhibitor, kinins, and that ofbradykinin, in particular. They have been
acarbose (Bay-g-542l), was found to reduce fasting and shown to confer cardioprotection following ischemia in dog
postprandial plasma insulin levels and to increase the rate of models through the enhancement of bradykinin levels [44].
clearance of glucose in an intravenous glucose tolerance test The ACE inhibitor, captopril, similarly has strong cardio-
[36]. Similarly, the newer a-glucosidase inhibitor, miglitol protective effects in the cp/cp male rat. The marked (79%)
(Bay-m-l099), very strongly inhibited the postprandial decrease in myocardiallesion frequency was not associated
insulin response to a test meal (RusseIl et al., unpublished with a significant improvement in lipid levels, insulin or
observations). This effect persisted when the test meal no glucose metabolism, or with any reduction in the severity of
longer ineluded the drug. Acarbose, in long-term treatment, lesions of the aortic arch (RusseIl et al., unpublished obser-
was cardioprotective while causing little change in the plasma vations). Such resuIts illustrate that, in a complicated and
lipids. These resuIts again confirm the importance of the multifactorial disease, intervention at any of a number of
insulin metabolism in the genesis of cardiovascular disease critical stages ofthe pathophysiological process mayprevent
in this animal model. manifestation of the end-stage disease.
The most strikingly effective agent studied to date has been
MEDICA 16 (ß,ß' tetramethylhexadecanedioic acid). Ifthe
animals are treated from 6 weeks of age, when the insulin Conclusions
resistance is essentially established, lipid levels are reduced
by some 80% - elose, but not quite, to normal concentration The JCR:LA-cp shows the elements ofthe Metabolic Synd-
-, insulin levels are moderated, and at 9 months of age the rome in a very pronounced form. This is accompanied by a
treated rats show more normal islets ofLangerhans, reduced development of end-stage vascular dysfunction and lesions
severity of aortic lesions, and strong cardioprotection against that is unique among small animal models. Even more
ischemic lesions [22]. These resuIts emphasize the strong striking is the occurrence of ischemic lesions of the heart-
roles ofboth insulin resistance/hyperinsulinemia and hyper- the prominent elinical problem in humans with the obesityl
triglyceridemia in the pathophysiological processes that lead insulin resistance syndrome. The rats develop the metabolie
to end-stage disease in this animal model. and pathophysiological characteristics while eating the rodent
Calcium channel antagonists have been used as an effective equivalent ofa 'health food' diet, a very-Iow-fatgrain-based
treatment for hypertension, and patients highly likely to be rat chow. The accumulated evidence strongly suggests that
so treated are those with the Metabolie Syndrome as hyper- the disease-prone character ofthe JCR:LA-cp rat is not just
tension is strongly associated [40]. In the male cp/cp rat of a consequence ofthe cp gene itself, but is multifactorial and
the JCR:LA-cp strain, both benzothiazepine and dihydro- polygenetic. The disease-prone status requires that the animal
pyridine calcium channel antagonists lower plasma triacyl- be both cp/cp and male. However this, alone, is not sufficient,
glycerol concentrations by 50%, with an accompanying and there are clearly other elements ofthis rat's genome that
similar reduction in cholesterollevels [41]. In contrast, female are not present in other strains that are necessary for the
cp/cp rats show no significant effect on the very high plasma development of cardiovascular disease. One of these appears
triacylglycerol concentrations and, in fact, exhibit a 50% rise to be a strain-specific abnormality of SMC function that
in esterified and total cholesterol. In the male rats we have makes the cells hyperplastic, hyperresponsive, and aggress-
shown that nifedipine is also cardioprotective and that ively migratory. We suspect that there are other factors at
nisoldipine both improves the status ofthe aortic arch and is work that have not yet been identified. These findings would
cardioprotective [42, 43]. It is not known what long-term be merely interesting on their own, except that it is becoming
effects these agents may have onfemale cp/cp rats. While the elear that cardiovascular disease in the human population is
females do not develop significant vasculopathy or myo- also multifactorial and polygenetic in nature. The background
cardiallesions at early ages, it is entirely possible that raised genome in humans is as variable from individual to individual
125

as it is between different strains of rats. The development of 14. Amy RM, Dolphin PJ, Pederson RA, RusseIl JC: Atherogenesis in
cardiovascular disease appears to require a susceptible two strains of obese rats: The fatty Zucker and LAIN-corpulent.
Atheroselerosis 69: 199-209, 1988
background, and then exposure to an environment or diet or
15. Greenhouse DD, Michaelis OE, Peterson RG: The development of
genetic abnormality, such as the cp gene, to lead to overt fatty and corpulentrat strains. In: CT Hansen and OE Michaelis (eds).
disease. The use of animal models of complex diseases, such New Models ofGeneticaIlyObese Rats for Studies in Diabetes, Heart
as the Metabolie Syndrome and its pathophysiologie al Disease, and Complications of Obesity. National Institutes ofHealth,
sequelae, is an essential first step to unravelling the under- Bethesda, 1988, pp 3-6
16. Elwood KC, Michaelis OE, Emberland JJ, Bhathena SJ: Hormonal
lying mechanisms and multiple causative factors. Animal
and lipogenic and gluconeogenic enzyme responses in LAIN-corpulent
models like the JCR:LA-cp rat are also very valuable tools rats. Proc Soc Exp Biol Med 179: 163-167,1985
for the next step, the development of new pharmaceutical or 17. Michaelis OE, Elwood KC, Judge JM, Schoene NW, Hansen CT: Effect
dietary approaches to prevention and treatment of human of dietary sucrose on the SHRIN-corpulent rat - A new model for
disease. insulin resistance. Amer J Clin Nutr 39: 612--618, 1983
18. Absher PM, Schneider DJ, RusseIl JC, Sobel BE: Increased prolif-
eration of explanted vascular smooth museIe cells: A marker presaging
atherogenesis. Atheroselerosis 131: 187-194, 1997
Acknowledgements 19. Dolphin PJ, Amy RM, Russell JC: The effects of age and sex on the
serum lipids and lipoproteins ofthe male and fernale JCR:LA-corpulent
This work has been supported by the Heart and Stroke rat. Biochim BiophysActa 1042: 99-106,1990
20. RusseII JC, Koeslag DG,Amy RM, Dolphin PJ: Plasma lipid secretion
Foundations of Alberta and the Northwest Territories and and elearance in the hyperlipidemic JCR:LA-corpulent rat. Arterio-
Ontario and by the Medical Research Council of Canada. selerosis 9: 869-876, 1989
21. Russell JC, Dolphin PJ, Hameed M, Stewart B, Koeslag DG,
Rose-Kahn G, Bar-Tana J: The hypolipidemic effect of ~,ß"-tetra­

References methyl hexadecanedioic acid (MEDICA 16) in hyperlipidemic


JCR:LA-corpulent rats. Arterioselerosis II: 602--609, 1991
22. RusseII JC,Amy RM, Graham SE, Dolphin PJ, Bar-Tana J: Inhibition
I. Reaven GM, Chen Y-DI: Insulin resistance, its consequences, and of atherosclerosis and myocardial lesions in the JCR:LA-cp rat by
coronary heart disease. Circulation 93: 1780-1783, 1996 ~,~. -tetramethylhexadecanedioic acid (MEDICA 16). ArterioseIer
2. Steiner G: Hyperinsulinernia and hypertriglyceridemia. J Int Med Thromb Vase Biol 15: 918--923, 1995
736(Suppl): 23-26, 1994 23. Iida M, Murskami T, Ishida K, Mizuno A, Kuwajima M, Shima K:
3. Taegtmeyer H: Insulin resistance and atheroselerosis. Common roots Substitution at codon 260 (glutamine ~ proline) ofthe leptin receptor
fortwo common diseases. Science 93: 17771-11779, 1996 (OB-R) cDNA is the only mutation found in the Zucker fatty (fa/fa)
4. Stern MP: Diabetes and cardiovascular disease: The 'common soil' rat. Biochem Biophys Res Comm 224: 597--604, 1996
hypothesis. Diabetes 44: 369-374, 1995 24. Wu-Peng, XS, Chua SC Jr, Okada N, Liu S-M, Nicolson M, Leibel
5. Russell JC, Ahuja SK, Manickavel V, Rajolte RV, Amy RM: Insulin RL: Phenotype ofthe obese Koletsky (j) rat due to Tyr763Stop mutation
resistance and impaired glucose tolerance in the atheroselerosis prone in the extraceIIular domain of the leptin receptor (Lepr). Diabetes 46:
LAIN-corpulent rat. Arteriosclerosis 7: 620-626, 1987 513-518, 1997
6. RusseIl JC, Graham S, Hameed M: Abnormal insulin and glucose 25. EmiIsson V, Liu YL, Cawthorne MA, Morton NM, Davenport M:
metabolism in the JCR:LA-corpulentrat. Metabolism 43: 538--543, 1994 Expression of the functionalleptin receptor mRNA in pancreatic islets
7. Russell JC, Manickavel V, Koeslag DG,Amy RM: Effects ofadvancing and direct inhibitory action of leptin on insulin secretion. Diabetes
age and severe food restriction on pathological processes in the insulin 46: 313-316,1997
resistant JCR:LA-corpulent rat. Diabetes Res 15: 53-62, 1990 26. Chan CB, MacPhail RM, Kibenge MT, Russell JC: Increased glucose
8. McNamee CJ, Kappagoda CT, Kunjara R, Russell JC: Defective phosphorylating activity correlates with insulin secreting capacity of
endothelium-dependent relaxation in the JCR:LA-corpulent rat. Circ male JCR:LA-corpulent rat islets. Can J Physiol Pharmacol 73: 501-
Res 74: 1126-1132, 1994 508,1995
9. Russell JC, Amy RM: Early atheroselerotic lesions in a susceptible 27. Ahuja S, Manickavel V, Amy RM, Russell JC: Age-related qualitative
rat model: The LAIN-corpulent rat. Atheroselerosis 60: 119-129,1986 and quantitative changes in the endocrine pancreas of the LAI
10. RusseIl JC, Amy RM: Myocardial and vascular lesions in the LA! N-corpulent rat. Diabetes Res 6: 137-144, 1987
N-corpulent rat. Can J Physiol Pharmacol64: 1272-1280, 1986 28. Vance JE, Russell JC: Hypersecretion of VLDL, but not HDL, by
11. Pederson RA, Campos RV, Buchan AMJ, Chisholm CB, RusseIl JC, hepatocytes from the JCR:LA-corpulent rat. J Lipid Res 31: 1491-
Brown JC: Comparison of the enteroinsular axis in two strains of obese 1501, 1990
rats: Tbe fatty Zucker and JCR:LA-corpulent. Int J Obes 15: 461- 29. Russell JC, Graham SE, Dolphin PJ, Brindley DN: Effects of
470,1991 benfluorex on serum triacylglycerols and insulin sensitivity in the
12. RusseIl JC, Amy RM, Michaelis OE, McCune SM, Abraham AA: corpulent rat. Can J Physiol Pharmacol 74: 879-886, 1996
Myocardial disease in the corpulent strains ofrats. In: E Shaftir (ed). 30. Bougeois CS, Wiggins D, Hems R, Gibbons GR: VLDL output by
Frontiers in Diabetes Research: Lessons from Animal Diabetes III. hepatocytes from obese Zucker rats is resistant to the inhibitory effects
Smith-Gordon, London, 1990, pp 402-407 ofinsulin. Am J Physiol 269: E208--E215, 1995
13. Russell JC: The atherosclerosis-prone JCR:LA-corpulent rat. In: FP 31. Williams G, Shellard L, Lewis DA, McKibbin PE, McCarthy HD,
Woodford, J Davignon and A Sniderman (eds). Atherosclerosis X Koeslag DG, Russell JC: Hypothalamic neuropeptide Y distur-
Proceedings-ofthe 10th International Symposium önAtheroselerosis. bances in the obese cp/cp JCR:LA-corpulent rat. Peptides 13: 537-
Elsevier Science, Amsterdam, 1995, pp 121-125 540,1992
126

32. Williams G, Cardoso H, Domin J, Ghatei MA, Russen JC, Bloom SR: 39. Russen JC, Graham SE, Dolphin PJ, Amy RM, Wood GO, Brindley
Disturbances of reguJatory peptides in the hypothalamus of the ON: Antiatherogenic effects of long-term benfluorex treatment in
JCR:LA-corpulent rat. Diabetes Res 15: 1-7,1990 male insulin resistant JCR:LA-ep rats. Atherosclerosis 132: 187-
33. Russen JC, Amy RM, Manickavel V, Dolphin PJ, Epling WF, Pierce 197, 1997
0, Boer 0: Prevention of myocardial disease in the JCR:LA-corpulent 40. Sowers JR, Sowers PS, Peuler JO: Role of insulin resistance and
rat by running. J Appl Physiol66: 1649, 1989 hyperinsulinemia in development ofhypertension and atherosc1erosis.
34. Russen JC, Amy RM: Early atherosclerotic lesions in a susceptible J Lab Clin Med 123: 647--{j52, 1994
rat model: The LA/N-corpulent rat. Atherosc1erosis 60: 119-129, 1988 41. Russen JC, Graham S, Stewart B, Oolphin PJ: Sexual dimorphism in
35. Russen JC, Amy RM, Dolphin PJ: Effect ofn-3 fatty acids on athero- the metabolie response to the calcium channel antagonists, diltiazem
sclerosis proneJCR:LA-corpuJentrats. Exp MolPathol55: 28~293,1991 and clentiazem, by hyperlipidemic JCR:LA-cp rats. Bioehim Biophys
36. Russen JC, Koeslag DG, Dolphin PJ, Amy RM: Beneficial effects of Acta 1258: 199-205, 1995
acarbose in the atherosclerosis prone JCR:LA-corpulent rat. Meta- 42. Russen JC, Koeslag DG, Dolphin PJ, Amy RM: Prevention of
bolism 43: 218-223, 1993 myoeardiallesions in JCR:LA-corpulent rats by nifedipine. Arterio-
37. Russen JC, Amy RM, Maniekavel V, Dolphin PJ: Effeets ofchronie sclerosis 10: 658--{i64, 1990
ethanol eonsumption in atheroselerosis-prone JCR:LA-corpulent rat. 43. Russen JC, Dolphin PJ, Graham SE, Amy RM: Cardioproteetive and
Arteriosclerosis 9: 1221-1228,1989 hypolipidemie effects ofnisoldipine in the JCR:LA-eprat. J Cardiovase
38. Brindley ON, Haies P, Al-SieniAI, Russen JC: Sustained decreases in Pharrnacol29: 586--592, 1997
body weight and circulating insulin, glucose, triaeylglycerol and 44. Linz W, Martorana PA, Schölkens BA. Local inhibition ofbradykinin
cholesterol in JCR:LA-eorpulent rats chronieany treated with degradation in isehemic hearts. J Cardiovase Pharmacol 15(6): S99-
o-fenfluramine. Br J Pharmacoll05: 679-685, 1992 SI09, 1990
PART IV

VASCULAR DYSFUNCTION
Molecular and Cellular Biochemistry 188: 129-136, 1998.
© 1998 Kluwer Academic Publishers.

Nutritional and endocrine modulation of


intracellular calcium: Implications in obesity,
insulin resistance and hypertension
Michael B. Zemel
Departments ofNutrition and Medicine, The University ufTennessee, Knoxville, TN, USA

Abstract
Regulation of intracellular Ca2+([Ca2+l) plays a key role in obesity, insulin resistance and hypertension, and [Ca2+]j disorders
may represent a fundamental factor linking these three conditions. We have shown insulin to be a direct vasodilator, attenuating
voltage-gated Ca2+ influx and stimulating Ca2+-ATPase transcription via a glucose-6-phosphate response element These result in
a net decrease in [Ca2+l and thereby decrease vascular resistance, while these effects are blunted in insulin resistance, leading to
increased vascular resistance. Consistent with this concept, pharmacological amplification of peripheral insulin sensitivity results
in reduced arterial pressure. While insulin regulates [Ca2+l, Ca2+also regulates insulin signaling, as increasing [Ca2+l impairs insulin
signaling in some systems, possibly due to Ca2+ inhibition of insulin-regulated dephosphorylation. Finally, in recent studies ofthe
mouse agouti gene, we have also demonstrated increased [Ca2+]j to playa key role in adipocyte lipogenesis, as folIows. We have
found dominant agouti mutants to exhibit increased [Ca2+l in most tissues, leading to increased vascular reactivity and insulin
resistance in vascular smooth muscle and skeletal muscle cells, respectively. Further, we have found recombinant agouti protein
to directly increase [Ca2+l in a variety of cells, including murine and human adipocytes, and to stimulate both the expression
and activity of adipocyte fatty acid synthase and increase triglyceride accumulation in a Ca2+-dependent manner. These effects
can be mimicked by stimulation of Ca2+ influx and blocked by Ca2+ channel inhibition, while treatment of mice with a Ca2+
antagonist attenuates agouti-induced obesity. Since humans express agouti in adipose tissue, it may similarly exert paracrine
effects on [Ca2+]j and thereby stimulate de novo lipogenesis and promote obesity. Thus, Ca2+ signaling represents a target for
therapeutic intervention in obesity as weil as hypertension and insulin resistance. (Mol Cell Biochem 188: 129--136, 1998)

Key words: agouti calcium, insulin resistance, hypertension, obesity

Introduction insulinemia, hypertension and dyslipidemias. The operational


premise of this review is that insulin resistance and hyper-
Regulation of intracellular Ca2+([Ca2+l) is well recognized tension are not merely comorbid factors which occur as a
to playa primary role in hypertension, and emerging result of obesity, but rather that these three conditions are
evidence suggests a corresponding role for alterations in the consequence of an underlying impairment in [Ca2+]i'
[Ca 2+]i regulation in the pathophysiology of obesity and which may therefore be a fundamental factor in the patho-
insulin resistance as wel\. Insulin resistance and hyper- physiology of all three.
tension are co-morbid factors of obesity, and the three
conditions are often present in what has been termed the
deadly quartet of obesity, hypertriglyceridemia, hyper- Hypertension in insulin resistance
tension, and insulin resistance/hyperinsulinemia. The term
Syndrome X was used by Reaven [ I]) to define a similar It is now weil established that hypertension is an intrinsically
syndrome of insulin resistance with compensatory hyper- insulin resistant state, as relative insulin resistance occurs

Address Jor offPrints: M.B. Zemel, Departments ofNutrition and Medicine, The University ofTennessee, 1215 W. Cumberland Ave - Room 229, Knoxville,
TN 37996-1900. USA
130

even in non-obese, non-glucose-intolerant hypertensive may be impaired in insulin-resistant states, similar to the
individuals compared to carefully matched controls [reviewed impairments found in 'classical' insulin target tissues.
in 2]. Similarly, both nonobese and obese animal models of
hypertension exhibit a degree of insulin resistance compared
to littermate controls. However, the underlying cellular link Insulin regulation of vascular smooth
between insulin resistance and hypertension has been unclear
and somewhat controversial. muscle [Ca2+].I
Hypertension in insulin resistant states is frequently
attributed to selective insulin resistance [3], chiefly by Increases in peripheral vascular resistance result from
skeletal muscle, with preservation of the renal and sym- agonist- and voltage-mediated increases in vascular smooth
pathetic neural responses to insulin. This selective insulin musc1e [Ca2+]., and several lines of evidence, reviewed
resistance results in hyperinsulinemia, which in turn stimu- below, indicate that insulin regulation of vascular smooth
lates sympathetic neural output and renal sodium retention musc1e [Ca2+]; is responsible for insulin regulation of
and thereby increasing peripheral vascular resistance [1, 2]. vascular tone [21]. Insulin attenuates both voltage- and
Consistent with this, insulin does exert an antinatriuretic receptor-mediated Ca 2+ influx in cultured rat vascular
effect [1-3], and hyperinsulinemia during euglycemic smooth muscle cells [22], and physiological concentrations
clamp studies results in acute increases in norepinephrine of insulin attenuate [Ca 2+]. responses to arginine vaso-
release [4] and in directly measured sympathetic neural pressin, angiotensin II and norepinephrine [23]. Further,
activity [5, 6]. insulin attenuation of vasoconstriction in canine femoral
However, despite this potential for insulin to increase artery cells is dependent upon inhibition of Ca2+influx [24].
peripheral vascular resistance, insulin also exerts a direct However, since insulin attenuates Ca 2+responses to angio-
vasodilatory role; consequently, resistance of the vaso- tensin II [23], which is dependent upon sarcoplasmic Ca2+
dilatory response to insulin, rather than hyperinsulinemia, release rather than by transmembrane flux, the effects of
may result in hypertension. Indeed, we found that hyper- insulin on plasmalemmal Ca2+channe1s can not be sufficient
tension and exaggerated in vivo pressor sensitivity in Zucker to explain insulin attenuation of Ca 2+ responses and of
obese rats persist during ganglionic blockade and are vasoconstriction.
therefore independent of sympathetic neural support [7]. Consequently, we have examined whether insulin may
Moreover, insulin exerts direct vasodilatory effects in attenuate vasoconstrictor responses via stimulation of Ca2+ -
vascular smooth muscle from several vascular beds, including ATPase-mediated Ca2+effiux. Insulin stimulates Ca2+-ATPase
rat aorta, human pulmonary artery, canine femoral artery and in a variety of tissues [25], while Ca2+-ATPase activity is
rabbit efferent renal arteriole [8-13], and this vasodilation is reduced in type II diabetes [21, 26]. Further, physiological
significantly atlenuated in insulin resistant rats. Further, insulin concentrations of insulin stimulate vascular relaxation and
infusion in both dogs and normotensive and hypertensive vascular smooth musc1e Ca2+-ATPase-mediated Ca2+ effiux
humans during euglycernic clamp studies results in decreases [10], and vascular smooth muscle cell Ca2+effiux is impaired
in peripheral vascular resistance which occur despite increases in both insulin resistant (Zucker obese) and insulinopenic
in muscle sympathetic nerve activity [14,15]. (streptozotocin-treated) rats [27-29]. This suggests that lack
Finally, amplification of peripheral insulin sensitivity with of cellular insulin action, whethcr duc to insulinopcnia or
biguanides (e.g. metformin) or thiazolidinediones (e.g. insulin resistance, results in impaired vascular smooth musc1e
troglitazone) results in reduced arterial pressure [16-18]. Taken Ca2+effiux and, consequently, exaggerated vasoconstriction.
together, these data suggest that peripheral vascular mani- These effects are dependent upon insulin stimulation of
festations of insulin resistance, rather than the accompanying Ca2+-ATPase gene transcription, as we have demonstrated
hyperinsulinemia, appear to be responsible for hypertension reduced Ca2+ expression in vascular smooth musc1e from
in insulin resistant states and that correction of vascular Zucker obese rats [30], as well as insulin stimulation ofCa2+-
manifestations of insulin resistance will correct or attenuate ATPase expression in both rat and human vascular smooth
the accompanying hypertension. Although vascular smooth musc1e cells [11, 21, 31]. These effects are associated with
muscle is not commonly thought of as an insulin target tissue, insulin-induced acceleration in the rate of [ea2+]. recovery
the insulin-sensitive glucose transporter (GLUT 4) mRNA from agonist-induced stimulation and a consequent reduction
and protein, and insulin-sensitive glucose transport, has been in the integrated area under the [Ca2+]. response curve [11,
reported in the A-I0 vascular smooth muscle celliine [19]. 31]. Interestingly, these effects are dependent upon the
Similarly, we have reported that rat aortic tissue contains high 'classical' actions of insulin in regulating glucose transport.
concentrations of GLUT 4 and exhibits insulin-dependent Replacement of glucose with 3-0-methylglucose, which is
glucose transport that is impaired in obese Zucker rats [20]. transported but not metabolized, resulted in complete loss of
Thus, vascular smooth muscle exhibits insulin sensitivity that insulin regulation ofCa2+-ATPase expression and [Ca2+]; [31].
131

However, replacement of glucose with 2-deoxyglucose, Ca2+ antagonism with nitrendipine significantly improved
which is transported and metabolized to the glucose-6- insulin sensitivity and glucose tolerance in spontaneously
phosphate analogue, 2-deoxyglucose-6-phosphate, resulted hypertensive rats [36], while Ca2+ antagonism with either
in preservation of this regulation [31]. Thus, glucose transport nitrendipine, diltiazem or amlodipine improved glucose
and metabolism to glucose-6-phosphate is essential for tolerance and lowered fasting and glucose-stimulated
insulin regulation of vascular smooth muscle cell [Ca2+l. This circulating insulin levels in insulin-resistant obese and
regulation is likely to occur via a glucose-6-phosphate hypertensive patients [37-40].
response element in the Ca 2+-ATPase promoter, as both Although the mechanism(s) whereby increased [Ca2+]j
human and rat Ca2+-ATPase genes contain motifs similarto the regulates insulin signaling is poody understood, there are
carbohydrate response elements of other insulin-responsive several candidate steps which have been addressed. [Ca2+]j
genes [31]. These mechanisms are summarized in Fig. 1. appears to have little or no effect on insulin-receptor binding
or receptor tyrosine kinase activity [41]. Modest decreases
in GLUT 4 levels in skeletal muscle from type TI diabetics
Ca2+ regulation of insulin sensitivity have been reported [42], although others report no change
[43,44]. However, [Ca2+]j may contribute to insulin resistance
Although insulin regulates [Ca2+l, there is also a corresponding without changing GLUT4 levels by inhibiting the dephos-
regulation of calcium sensitivity by insulin. For example, phorylation ofGLUT 4 and thereby decreasing it's activity.
Draznin et al. [32] demonstrated an optimal range of [Ca2+l In support of this concept, increasing Ca2+ influx via either
for maximizing insulin-stimulated glucose transport in depolarization or parathyroid hormone incubation increased
isolated adipocytes, with increases beyond this range causing GLUT4 phosphorylation in isolated adipocytes, while Ca2+
marked decreases in insulin sensitivity. Increasing [Ca2+]j in antagonism with nifedipine restored GLUT4 to its more
several cell types inhibits insulin signaling, while decreasing active, dephosphorylated state [45].
Ca2+improves cellular insulin sensitivity [21, 33, 34]. Thus, Maintenance of GLUT4 in a dephosphorylated state is
although [Ca2+]j has been suggested to playa role in insulin dependent upon phosphoserine phosphatase I (PP 1). Insulin
signal transduction [e.g. 35], inappropriate increases normally activates PPI via an insulin-stimulated pro tein
nonetheless appear to result in loss of signal transduction. kinase (ISPK- 1), which phosphorylates site 1 ofPPl, thereby
Consistent with this concept, Ca2+ antagonism in obese activating the enzyme. In contrast, protein kinase A mediates
elderly patients significantly increased insulin sensitivity of fuH phosphorylation ofboth serine phosphorylation sites of
adipocytes obtained from abdominal biopsies [33]. Similarly, pp 1 and also phosphorylates and activates the specific

Insulin
~
Ci) Glucose

Glucose - - Glucose-®

PVR

i
~ [Ca2+]i

Fig. 1. Insulin Regulation of Pcriphcral Vaseuhr Rcsistance (PYR). Insulin inhibits Ca'+ influx, contributing to reduced intracellular free Ca" ([Ca"]),
thereby reducing PVR. Insulin stimulation of glucose transport and phosphorylation to glucose-6-phosphate results in increased Ca" -A TPase transcription
and activity; this results in increased cellular Ca" effiux, thereby reducing ([Ca"], and, consequently, PVR.
132

inhibitor ofPP 1, inhibitor 1. When activated, inhibitor-1 binds


to the catalytic subunit ofPPl. Increasing [Ca2+J; results in
phosphorylation of inhibitor-I, while Ca 2+ antagonism In recent studies ofthe mechanisms of agouti gene-induced
prevents this phosphorylation and activation [46]. Reduced obesity, we have found that [Ca 2+l; modulates de novo
basal and insulin-stimulated activity of pp I has been lipogenesis in both rodent and human adipocytes and may
reported in skeletal muscle of insulin-resistant Pima Indians thereby represent a key therapeutic target for the control of
[47], and sustained elevation of [Ca 2+]j induced by either obesity. The agouti gene is normally involved in coat color
depolarization or parathyroid hormone was accompanied by regulation in wild-type mice [51]. The agouti gene is
inhibition of PPI in insulin-target cells [46,48]. This normally expressed only transiently, and specifically in skin,
decrease in PPI results in decreased activity of insulin- during early neonatal development, where it causes a switch
sensitive substrates, including GLUT4 and glycogen syn- from black (eumelanin) to yellow (phaeomelanin) pigment,
thase [45,49]. resulting in a subapical yellow band on an otherwise dark
It is also possible, however, that increased [Ca2+l contributes hair shaft [52]. However, in several dominant mutations at
to insulin resistance via protein kinase C activation, with a the murine agouti locus (a), such as viable yellow (A'v),
consequent phosphorylation and inactivation of the insulin regulation of agouti expression is disrupted, resulting in
receptor ~-subunit tyrosine kinase. An additional mechanism ectopic expression in almost all tissues throughout the life
whereby [Ca2+l may modulate insulin signaling is via direct ofthe mouse. In addition to a distinctive yellow pellage, these
interaction between Ca2+-calmodulin and IRS-I. Calmodulin mice are characterized by progressive obesity and insulin
has recently been reported to bind IRS-l, and this interaction resistance, developing body fat levels which are 35-50%
is regulated by Ca 2+, as the affinity for IRS-l for Ca2+- greater than wild-type mice [53].
calmodulin was markedly greater than for Ca2+-free calmodulin The agouti gene was the first obesity gene to be cloned [54];
[50]. Potential mechanisms for Ca2+ inhibition of insulin the gene encodes a 131 amino acid protein with a consensus
sensitivity are summarized in Fig. 2. signal peptide [54]. Molecular analysis demonstrates that

, - - - - - - . t PKC - - - + . TK-0

[Ca2+ ] i - - - - . tCaCM_____.. tCaCM - IRS-I

1
~
Inhibitor-l Inhibitor-l- ®

(;)
1
PPI (inactive) PPI (active)

1
Insulin-sensitive ~ Insulin-sensitive
substrate - ® substrate (active)
(inactive)

Fig. 2. Intraeellular Ca'+ ([Ca'+].) Modulation ofInsulin Sensitivity. [Ca'+J; aetivation ofprotein kinase C (PKC) results in phosphorylation and inaelivalion
of insulin reeeptor ~-subunit tyr~sine kinase (TK). Inereasing [Ca'+J; also inereases Ca2-ealmodulin (CaCM) binding, resulting in binding ofCaCM to IRS-
I; effeets ofCaCM binding to IRS-l on insulin signaling are presently unknown. Increasing [Ca'+), also results in phosphoryl.tion _nd activ_tion ofinhibitor
I. Inhibitor I binds to the catalytic subunit of phosphoserine phosphatase I (PP I), resulting in inactivation of PP I. Inaetivation of PP I results in impaired
dephosphorylation ofinsulin-sensitive substrates, such as GLUT4 and glycogen synthase. Consequently, a gre.ter proportion ofthese substrates rem.ins in
the phosphorylated (inaetive) state.
133

dominant obese yellow mice exhibit mutations in the promoter found that FAS expression and activity are markedly increased
region of the gene, resulting in ectopic expression of agouti in AVY relative to control mice [62], and that nanomolar
transeripts which contain the entire normal protein-coding concentrations of recombinant agouti protein stimulate -2-
region in multiple tissues [55, 56]. Transgenic mice in which fold increases in FAS gene expression and triglyceride
the wild-type agouti cDNA is placed under transcriptional accumulation in 3T3-L 1 adipocytes [62], as well as in human
regulation ofubiquitous promoters not only express agouti in adipocytes, similar to maximal increases stimulated by
multiple tissues, but also develop the syndrome of obesity, insulin. Further, we have found that agouti protein and insulin
hyperinsulinemia and yellow coat color, demonstrating that exert additive effects on FAS transcription, with a 6-fold
ectopic expression of agouti per se is responsible for the increase resulting from their combination versus a 2-3-fold
'yellow mouse' syndrome [57]. Accordingly, it is clear that increase from each independently [63]; FAS responses to
expression of agouti in novel target tissue( s) triggers the agouti are mediated by a response sequence distinct from the
development of obesity. insulin-response sequence. The agouti response sequence is
We have demonstrated that A'Y mice exhibit increases in localized between -500 and -300 upstream of the FAS
both steady state [Ca2']i and Ca" influx in several tissues [58, transcription start site, while the insulin-response site is located
59]. This increase in [Ca'+]. was closely correlated with both between --fJ7 and -52 upstream ofthe FAS transcription start
the degree of ectopic agouti expression and body weight [58], site [63].
suggesting the possibility of a causal mechanism between Agouti modulation of FAS transcription and lipogenesis
[Ca"]. and the obesity mechanism in these animals. Our appears to be [Ca2']i-mediated. We have found recombinant
investigations of this mechanism have focused on adipose murine and human agouti proteins to cause dose-dependent
tissue. A'Y mice exhibit elevated rates of adipocyte lipogenesis increases in Ca'+ influx and [Ca2+]. in a variety ofcells [58, 59],
and increased adipocyte size relative to lean controls [60, 61]. inc1uding both murine and human adipocytes. Further, agouti
It is, therefore, reasonable to speculate that regulatory modulation ofFAS transcription is dependent upon Ca'+ influx,
enzymes in lipid metabolism may be regulated by agouti and as it can be inhibited by Ca'+ antagonism [62, 64]. Moreover,
[Ca"] .. agouti regulation ofFAS and lipogenesis can be recapitulated
We have focused on fatty acid synthase (FAS), as this in the absence of agouti via Ca" -channel activation [64,65].
mutifunctional enzyme is highly regulated by nutrients and In addition, we have recently reported that treatment of
hormones and is rate limiting in de novo lipogenesis. We have transgenic mice ubiquitously expressing agouti under the

Agouti ~

Ca" G)

t [Ca"]i
Insulin ---------__\__ insulin
/'
Agouti
~(j)
Pancreatic ß cell

t lipogenesis
de novo

Adipocyte

Fig. 3. Agouti Modulation of Ca" Stimulates Fatty Acid Synthase (FAS) Transeription. Agouti protein stimulates Ca" influx in panereatie ~-eell, resulting
in inereased insulin release. Agouti inereases adipoeyte Ca" influx whieh, eoupled with the elevated insulin, inereases FAS transeription and activity,
resulting in inereased de nova lipogenesis.
134

control ofthe ß-actin promoter with a Ca2+-channel antagonist 9. Zemel MB, Reddy S, Sowers JR: Insulin a!tenualion ofvasoeonstrictor
resulted in an 18% reduction in fat pad mass over a 4 week responses to phenylephrine in Zucker lean and obese rats. Am J
Hypertension 4: 537-539, 1991
treatment period [67]. Since humans also express agouti,
10. Zemel MB, Johnson BA, Ambrozy SA: Insulin-stimulated vascular
primarily in adipose tissue [66], it may exert similar relaxation: Role of Ca 2·-ATPase. Am J Hypertension 5: 637--{)41,
paracrine effects on [Ca2+]j and thereby stimulate de novo 1992
lipogenesis. 11. Kim YC, Zemel MB: Insulin increases vascular smooth muscle
Finally, it should be noted that agouti interaction with recovery from intracellular calcium loads. Hypertension 22: 74-77,
1993
insulin appears to be required for full expression of agouti-
12. Kahn AM, Seidel CL, Allen JC, O'Neil RG: Insulin reduces contraction
induced obesity. As noted above, agouti protein and insulin and intracellular calcium coneentration in vascular smooth muscle.
exert independent, additive effects on FAS transcription Hypertension 22: 735-742, 1993
[63]. Since increased [Ca2+l is the proximate signal for 13. Juncos LA, ito S, Carretero OA: Disparate effects of insulin on
insulin release, and we have demonstrated agouti regulation isolated rabbit afferent and efferent arterioles. Hypertension 20: 403,
1992
of [Ca2+]j signaling in several cell types [59], it is reasonable
14. Hall JE, Coleman TG, Mizelle HL: Does chronic hyperinsulincmia
to speculate that agouti may stimulate insulin release as cause hypertension. Am J Hypertension 2: 171-173, 1989
well. Indeed, we have recently found that agouti stimulates 15. Brands MW, Mizelle HL, Gaillard CA, Hildebrandt DA, Hall JE: The
Ca2+ signaling in cultured pancreatic ß-cells and, consequently, hemodynamic response to chronie hyperinsulinemia in conscious dogs.
causes a marked increase in insulin release [68]. Further, Am J Hypertension 4: 164-168, 1991
16. Morgan DA, Ray CA, Balon TW, Mark AL: Metformin increases
hyperplasia ofß-cells precedes the development of obesity in insulin sensitivity and 10wers arterial pressure in spontaneously
agouti mutant mice, suggesting that hyperinsulinemia may be hypertensive rats. Hypertension 20: 421,1992
a direct effect of agouti acting on the pancreas and that the 17. Dubey RK, Zhang HY, Reddy SR, Buegehold MA, Kotchen TA:
combination of this hyperinsulinemia and adipocyte agouti Pioglitazone attenuates hypertension and growth of renal arteriolar
expression may lead to obesity. Indeed, we have recently smooth muscle cells in rats. Am J Phsyiol265: R276-R732, 1993
18. Pershadsingh HA, Szollosi J, Bensom S, Hyun WC, Feurstein BG,
reported that transgenic mice expressing agouti at high levels
Kurtz TW: Effect of eiglitazone on blood pressure and intraeellular
specifically in adipose tissue under the control of the aP2 calcium metabolism. Hypertension 21: 1020-1023, 1993
promoter become obese only if exogenous insulin is provided, 19. Cooper DR, Foote J, Petch C, Schofield PM: Chronie effects of glucose
while supplemental insulin was without effect on body weight on insulin signaling in A-IO vascular smooth muscle cells. Areh
in non-transgenic littermate controls [69]. Taken together, these Biochem Biophys 302: 490-498, 1993

data indicate that regulation of adipocyte [Ca 2+l, possibly


20. Ban, WJ, Abel MA, Zemel MB: Insulin regulation ofvascular smooth
muscle glucose transport in insulin-sensitive and resistant rats. Hor
coupled with pancreatic [Ca2+l and insulin release, may be an Metab Res 28: 271-275, 1996
important target for the development of therapeutic strategies 21. Zemel MB: Insulin resistance vs. Hypcrinsulinemia in hypertension:
in the prevention and treatment of obesity (Fig. 3). Insulin regulation of Ca2• transport and Ca" regulation of insulin
sensitivity. J Nutr 125: 1738S-1743S, 1995
22. Standley PR, Zhang F, Ram JL, Zemel MB, Sowers JR: Insulin
attenuates vasopressin-induced calcium transients and voltage-
References dependent calcium CUITent in rat vascular smooth museIe cells. J Clin
Invest 88: 1230-1236, 1991
I. Reaven G: Role ofinsulin resistanee in human disease. Diabetes 37: 23. Touyz RM, Tolloczko B, Schiffrin EL: Insulin a!tenuates agonist-evoked
1595--1607, 1988 calcium transients in vascular smooth muscle cells. Hypertension 23
2. Zemel MB: Insulin resistanee, obesity and hypertension: An overview. (Suppt. I): 125--128, 1994
JNutrl25: 1715S-1717S, 1995 24. Kahn AM, Seidel CL, Allen JC, O'Neil RG: Insulin reduces contraction
3. Roeehini AP: Insulin resistanee and blood pressure regulation in obese and intracellular calcium concentration in vascular smooth muscle.
and non-obese subjects. Hypertension 17: 837-1l42, 1991 Hypertension 22: 735--742, 1993
4. Rowe JW, Young JB, Minaker KL, Stevens AL, Palotta, Landsberg L: 25. Levy J, Zemel MB, Sowers JR: Role of cellular calcium metabolism
Effcct of insulin and glucose infusion on sympathetic nervous system in abnormal glucose metabolism and diabetic hypertension. Am J Med
activity in normal man. Diabetes 30: 219-225,1981 87 (Suppt. 6A): 155-165
5. Anderson EA, Hoffman RP, Balar TW, Sinkey CA, Mark AL: 26. Zemel MB, Bedford BA, Zemel PC, Marwah 0, Sowers JR: Altered
Hyperinsulinemia produces bath sympathelic neural activalion and cation transport in non-insulin-dependen! diabetics: Effeets of dietary
vasodilation in normal humans. J Clin Invest 87: 2246-2252, 1991 calcium. J Hypertension 6 (Suppt. 4): S228-5230
6. Anderson EA, BalorTW, Hoffman RP, Sinkey CA, MarkAL: Increases 27. Shchin SE, Sowers JR, Zemel MB: Impaired vascular smooth muscle
sympathetic aclivity but not blood pressure in borderline hypertensive 45Ca effiux and hypertension in Zucker obese rats. J Vase Med Biol
humans. J Clin Invest 89: 621--{)27, 1992 1: 278-282,1989
7. Zemel MB, Penler JD, Sowers JR, Simpson L: Hypertension in insulin- 28. Abel MA, Zemel MB: Impaired recovery ofvaseular smooth muscle
resistant Zucker obese rats is independent of sympathelic neural intraeellular calcium following agonist stimulation in insulin resistant
support. Am J Physiol262: E368-E371, 1992 (Zucker obese) rats. Am J Hypertension 6: 500-504, 1993
8. Zemel MB, Reddy S, Shehin S, Lockette W, Sowers JR: Vascular 29. Reddy S, Shehin S, Sowers IR, Dardas G, Zemel MB: Aortic "Ca
reaetivity in Zucker obese rats: Role ofinsulin resistanee. JVase Med flux and blood pressure regulation in streptozotocin-induced diabetic
Biol2: 81-85, 1990 rats. J Vase Med Bio12: 47-50, 1990
135

30. Zemel MB, lannucci A, Moore JW: Role of insulin in regulating 49. Dent P, Lavoinne A, Nakielny S, Caudwell FB: The molecular
vascular smooth museie Ca2+-ATPase expression. J Vase Med Biol4: meehanism by whieh insulin stimulates glycogen synthesis in
79-84 mammalian skeletal musc1e. Nature 348: 302-308, 1990
31. Kim YC, Zemel MB: Insulin stimulation of intracellular free Ca'+ 50. Munshi HG, Burks DJ, Joyal JL, White MF, Sacks DB: Ca'+ rcgulatcs
recovery and Ca'+-ATPase gene expression in cultured vascular smooth calmodulin binding to IQ motifs in IRS-1. Bioehemistry 35: 15883-
musc1e cells: Role of glucose-6-phosphate. Biochem J 311: 555-559, 15889, 1996
1995 51. Yen TT, GilIAM, Frigeri LG, Barsh GS, WolffGL: Obesity, diabetes
32. Draznin B, Sussman K. Kao M, Lewis D, Sherman N: The existence of and neoplasia in yellow A"Y/- mice: Ectopic expression ofthe agouti
an optimal range of cytosolic free calcium for insulin-stimulated glucose gene. FASEB J 8: 479-488,1994
transport in rat adipocytes. J Biol Chem 262: 14385-14388,1987 52. Zemel MB, Moore WJ, Rahman MK, Moustaid N: Diazoxide
33. Draznin B: Cytosolic calcium and insulin resistance. Am J Kidney antagonism of glybenclamide-induced Ca'+ signaling and lipogenic
Dis 21 (Suppl. 3): 32-38, 1993 activity in 3T3-L 1 adipocytes. Obesity Res 4: 28s, 1996
34. Segal S, Lloyd S, Sherman N, Sussman KE, Draznin B: Postprandial 53. Draznin B, Sussman KE, Eckel RH, Kao M, YOSI T, Sherman NA:
changes in cytosolic free calcium and glucose uptake in adipocytes in Possible role of cytosolie free calcium concentrations in mediating
obesity and non-insulindependent diabetes mellitus. HormRes 34: 39- insulin resistance of obesity and hyperinsulinemia. J Clin lnvest 82:
44,1990 1848-1852,1988
35. Epstein M, Sowers JR: Diabetes mellitus and hypertension. Hyper- 54. Bultman SJ, Michaud EJ, Woyehik RP. Molecular charaeterization of
tension 19: 403-418,1992 the mouse agouti locus. Cell 71: 1195-1204, 1992
36. Bursztyn M, Raz I, Mekler J, Ben-Ishay D: Nitrendipine improves 55. Michaud EJ, Bultman SJ, Stubbs LJ, Woychik RP: The embryonic
glucose tolerance and deoxyglucose uptake in hypertensive rats. letality of homozygous lethaI yellow mice (AY/AY) is associated with
Hypertension 23: 1051-1053, 1994 the disruption of a nove1 RNA-binding protein. Genes Dev 7: 1203-
37. Beer NA, Jakubowicz DJ, Beer RM,Arocha IR, Nestler JE: Effects of 1213,1993
nitrendipine on glucose tolerance and serum insulin and dehydro- 56. Michaud EJ, Vugt MJ, Bultman SJ, Sweet HO, Davisson MT, Woychik
epiandrosterone sulfate levels in insulin-resistant obese and hyper- RP: Differential expression of a new dominant mutant agouti allele
tensive men. J Clin Endocrinology 76: 178-183, 1993 (Ai',,) is eOITelated with methylation state and is influenced by parental
38. Beer NA, Jakubowiez DJ, Beer RM: Disparate effeets of insulin lineage. Genes Dev 8: 1463-1472, 1994
reduction with diltiazem on serum dehydroepiandrosterone sulfate 57. Kiebig ML, Wilkinson JE, Geisler JG, Woychik RP: Eetopie expression
levels in obese hypertensive men and women. J Clin Endoerinology ofthe agouti gene in transgenic mice causes obesity, features of type
79: 1077-1081, 1994 II, and yellow fur. Proc NatAcad Sei USA 92: 4728-4732, 1995
39. Byyny RL, LoVerde M, Mitehell W, Draznin B: Cytosolie calcium 58. Zemel MB, Kim JH, Woychik RP, Michaud EJ, Kadwell SH, Patel IR,
and insulin resistance in elderly patients. Am J Hypertension 5: 459- Wilkison WO: Agouti regulation ofintracellular calcium: Role in the
464,1992 insulin resistance ofviable yellow mice. Proc Nat Acad Sci USA 92:
40. Harano Y, Dageyama A,Hirose J, Asakura Y, Yokota T, Ikebuchi M, 4733-4737, 1995
Suzuki M, Omae T: Improvement of insulin sensitivity for glucose 59. Kim JH, Kiefcr LL, Woychik RP, Wilkison WO, Trucsdale A, Ittoop
metabolism with the longacting Ca-channel blocker amlodipine in 0, WilIard D, Nichols J, Zemel MB: Agouti regulation ofintracellular
essential hypertensive subjeets. Metabolism 44: 315-319, 1995 calcium: role of melanocortin receptors. Am J Physiol 272: E379-
41. Draznin B, Lewis D, HoulderN, Sherman N,Adamo M, GarveyWT, E384, 1997
LeRoigh D, Sussman K: Meehanism ofinsulin resistanee induced by 60. WolffGL: Growth ofinbred yellow (AY/a) and non-yellow (a/a) mice
sustained levels of cytosolie free calcium in rat adipocytes. Endo- in parabiosis. Genetics 48: 1041-1058, 1963
crinology 125: 2341-2349, 1989 61. Johnson PR, Hirsch J: Cellularity of adipose depot in six strains of
42. Dohm GL, Elton CW, Friedman JE, Pileh PF, Pories WJ, Atkinson genetically obese mice. J Lipid Res 13: 2-11,1972
SM Jr, Caro JF: Decreased expression of glucose transporter in muscle 62. Jones BH, Kim JH, Zemel MB, Woychik RP, Michaud EJ, Wilkison
from insulin resistant patients. Am J Physiol 260: E459-E463, 1991 WO, Moustaid N: Upregulation of adipocyte metabolism by agouti
43. Pederson 0, Bak JF, Andersen PH, Lund S, Moller DE, Flier JS, Kahn pro tein: Possible paraerine actions in yellow mouse obesity. Am J
BB: Evidence .Itered expression ofGLUTl or GLUT$ in skeletal muscle Physiol270: EI92-EI96, 1996
ofpatients with obesity or NIDDM. Diabetes 39: 865-870, 1990 63. Claycombe KH, Jones BH, Standridge MK, Wilkison WO, Zemel MB,
44. HandbergA, VaagA, Damsba P, Beck-Nielsen H, Vinten J: Expression Guo YS, Moustaid N: Transcriptional regulation ofthe adipocyte fatty
of insuilnregulatable glucose transporters in skelet.l musclc from type acid synthase gene by thc agouti gene product: Interaction with insulin.
2 (non-insulindependent) diabetic patients. Diabetologia 33: 625-627, FASEB J 11: A352, 1997
1990 64. Zemel MB, Kim JH, Jones BH, Moore JW, Woyehik RP, Moustaid N,
45. Begum N, Leitner W, Reuseh J E-B, Sussman KE, Draznin B: Glut-4 Wilkison WO: Agouti gene product regulation of intracellular free
phosphorylation and its intrinsic aetivity. J Biol Chem 268: 3352- calcium results in stimulation of fatty acid synthase. Obesity Res 3:
3356, 1993 338s, 1995
46. Begum N, Sussman KE, DrazninB: Calcium-induced inhibition ofphos- 65. Moore JW, Willard D, Moustaid N, Wilkison WO, Zemel MB: Role
phoserine phosphatase in insuiln target eells is mediated by the phos- of intracellular calcium in agouti and insulin modulation of fatty acid
phorylation and activation of inhibitor 1. J Biol Chem 267: 5959-5963 synthase. FASEB J 10: A187, 1996
47. Kida Y, Esposito-Del Puente A, Bogardus C, Mott DM: Insulin 66. Kwon HY, Bultman SJ, Loffler C, Chen WJ, Furdon PJ, Powell JG,
resistanee is associated with redueed fasting and insulin-stimulated U salaAL, Wilkison W, Hansmann I, Woychik RP: Molecular strueture
glycogen synthase phosphatase activity in human skeletal muscle. J and chromosomal mapping of the human homologue of the agouti
Clin lnvest 85: 476-481, 1990 gene. Proc Nat Acad Sci USA 91: 9760-9764,1994
48. Freidenberg GR, Henry RR, Klein HH, Reichart DR, Olefsky JM: 67. Kim JH, Mynatt RL, Moore JW, Woychik RP, Moustaid N, Wilkison
Decreased kinase activity of insulin receptors from adipocytes of non- WO and Zemel MB: The effects of calcium channel blockade on
insulin-dependent diabetie subjets. J Clin Invest 79: 240-250, 1987 agouti-induced obesity. FASEB J 10: 1646-1652
136

68. Xue B. Wilkison WO, Mynatt RL, Moustaid N, Zemel MB: Theagouti 69. Mynatt RL, Miltenberger RJ, Kiebig ML, Zeme1 MB, Wilkinson JE,
gene product stimulates pancreatic ß cell Ca'+ signaling and insulin Wilkison WO, Woychik RP: Combined elfects of insulin treatment
release. FASEB J 11: A320, 1997 and adipose tissue-speeific agouti expression on the development of
obesity. Proc Nat Acad Sei USA 94: 919-922, 1997
Molecular and Cellular Biochemistry 188: 137-148, 1998.
© 1998 Kluwer Academic Publishers.

Hypertension, calcium channel and pyridoxine


(vitamin B6)
Krishnamurti Dakshinamurti,I,2 Kovvuri Jawahar LaF and Pallab K.
Ganguly3
lInstitute of Cardiovascular Sciences, St. Boniface General Hospital Research Centre; 2Departments of Biochemistry and
Molecular Biology, and 3Anatomy, Faculty ofMedicine, University ofManitoba, Winnipeg, Canada

Abstract
The moderately pyridoxine (vitamin B6)-deficient male rat was introduced by us as an animal model (B6DHT) for the study of
hypertension. Hypertension in this rat is associated with increased sympathetic stimulation. Arterial segments from B6DHT
rats maintained a higher resting tone. The influx of 45 calcium into intracellular compartment of the vascular smooth musc1e of
the caudate artery ofB6DHT rats was also enhanced. Administration ofpyridoxine attenuated the hypertension in B6DHT rats
as weH as in genetic or dietary-induced moderately hypertensive conditions such as in the Zucker obese rat and sucrose or low
calcium-fed rats. However, pyridoxine did not have any effect on the spontaneously hypertensive rat. All c1asses of calcium
channel blockers were effective in lowering the systolic blood pressure ofB6DHT rats. The increased in vitro influx of 45 calcium
into intracellular compartment of artery segments ofB6DHT rats as well as the BAY K 8644-induced influx of 45 calcium into
artery segments from normal rats were blocked by pyridoxal phosphate as well as by dihydropyridine-sensitive calcium channel
blockers (DHP). Pyridoxal phosphate (PLP) in vitro enhances the binding of calcium channel antagonists to membrane
preparations from vascular tissue. PLP corrects the membrane abnormality in responsive hypertensive conditions and thus,
could be an endogenous modulator ofDHP - sensitive calcium channels. (Mol Cell Bioehern 188: 137-148, 1998)

Key words: Ca2+ channels, hypertension, vitamin B6 , vitamin deficiency, Zucker rat, vascular biology

Introduction acetate (DOCA) - saline treated rats, adrenal regeneration in


rats, Goldblatt renal hypertension in dogs, post-DOCA
Hypertension is one of the major causes of chronic illness in hypertension in rats and spontaneously hypertensive rats.
Western societies. About 20-30% of the adult population More recently, there has been considerable interest in
have some degree of elevation of blood pressure (diastolic genetic obesity-induced or sucrose (or fructose )induced
blood pressure 90-100 mmHg). Hypertension is associated metaboJically aberrant conditions which are also associated
with conditions such as hypersecretion of aldosterone, with mild hypertension. The Zucker obese (fa/fa) rat,
mineralocorticoids or glucocorticoid, catecholamine secreting originally studied as a model of obesity, had found extensive
tumors of adrenal medullary or paraganglionic origin, renal use for the study of diabetes mellitus [1,2]. Various reports
disease and narrowing of the renal arteries or ofthe aorta. It have shown that the Zucker obese rat also develops hyper-
is also associated with toxemia of pregnancy. However, in tension [3]. The high blood pressure was not dependent on
over 90% of patients with elevated blood pressure the hyperphagia or the increase in body weight per se. Acute or
causative has not been recognized yet. This is referred to as chronic ingestion of simple carbohydrates such as sucrose or
essential hypertension. fructose has been shown to cause increase in systolic pressure
Rat and dog models used in the study of experimental (SBP) in several strains inc1uding the spontaneously hyper-
hypertension include uninephrectomized deoxycorticosterone tensive rat [4, 5]. We have introduced the moderately vitamin

Addressfor offprints: K. Dakshinamurti, Department ofBiochemistry and Molecular Biology, Faculty ofMedicine, University ofManitoba, Winnipeg, MB,
Canada R3G OW3
138

B6 (pyridoxine) deficient male rat as an animal model to study 400


hypertension [6] and propose that pyridoxal phosphate could
be an endogenous modulator of dihydropyridine-sensitive 350
calcium channels. c;,
We investigated the effect of vitamin B6 -deficiency on
blood pressure regulation. Male Sprague-Dawley rats (95- -
. &;
Cl
CI>
300

250
110 g body weight) purchased from Charles River, St. :;:
Constant, PQ, Canada, were used in all the experiments. They >-
"C 200
were housed in animal quarters under constant temperature 0
D:l
(23°C) and illumination (12 h light/dark cycle). The rats were
randomly divided into three groups. Group I was fed a 150
vitamin B6-deficient diet prepared as described previously by
us [7]. Group 2 was pair-fed with group 1, a vitamin B 6 (as 100
pyridoxine )-supplemented diet (control) and group 3 was fed
ad libitum a commercial rat chow. Water was provided ad c;,
J:
libitum to all rats. Systolic blood pressure and heart rate were 170
E
recorded weekly. Systolic blood pressure was measured .E 160
indirectly in conscious restrained animals by tail-cuff
...
..
CI> 150
plethysmogaraphy using MOD a 59 Amplifier connected to ::J
T
140
a MOD 38 Channel recorder (IlTC, Inc., Woodland Hills, CA, CI>

USA) as described [8]. Heart rate was calculated from pulse Q. 130
pressure. In order to reduce the effect of stress, rats were trained "C
0
120
0
to the procedure before the actual start of the experiment. ä5 110
Experiments were performed during the same time of the day 100

.
.~
in order to avoid the effect of circadian rhythm. In some Ci
90
experiments blood pressure was measured directly from canula >-

oI
(/)
placed in the right carotid artery in anesthetized animals [6]. c:
The values for systolic blood pressure obtained by tail cuff 'CI>" 0 2 4 9 10 11 12
plethysmography were comparable to those obtained by :::i:
direct measurement using carotid artery cathetarization. Time on Die! !Weeksl
The systolic blood pressure ± S.E.M. was 96 ± 3 mm Hg
and the body weight was 106 ± 3 g in all the groups at the Fig. 1. Effect of feeding ad libitum a vitamin B6 -deficient diet (T) or pair-
start ofthe experiment. As shown in Fig. I the systolic blood feeding avitamin B6 -supplemented (control) diet (A) or a commercial rat
pressure increased to 125 ± 3 mm hg by fifth week on the chow (.) on systolic blood pressure and body weight in conscious rats.
Blood pressure was measured by tail cuff plethysmography. Each value
vitamin B 6-deficient diet (p < 0.01). There was a further represents the mean ± S.E. M. of 15-20 rats. (From LaI and Dakshinamurti
increase in systolic blood pressure in the subsequent weeks [23]: reprinted with pennission ofthe publisher).
when it reached a peak value of about 152 mm Hg (8-10
weeks). The systolic blood pressure began to decline starting
in the tenth week on the deficient diet and reached 110 ± 5 mm pressure of rats fed ad libitum the commercial rat diet started
Hg by the twelfth week although the rats were still on the to increase and reached a value of 120 mm Hg.
deficient diet. These values for systolic blood pressure obtained The blood pressure changes in the vitamin B6-deficient rat
by tail cuff plethysmogaraphy were comparable to those can be classified into three phases: (i) pre-hypertensive (1-4)
reported earlier by direct measurements in the anaesthetized weeks): (ii) hypertensive (5-11 weeks) and (iii) post-hyper-
rat [6]. The body weight ofthese rats increased to 262 ± 5 tensive (starting from 12th week). Vitamin B6-deficient rats
g during the experimental period whereas rats eating a during the hypertensive phase were only moderately pyri-
commercial rat chow diet were nearly 100 g more in body doxine deficient. They did not have any clinical signs of
weight. Rats pair-fed the vitamin B6-supplemented (control) deficiency. This moderately vitamin B6-deficient hypertensive
diet weighed 276 ± 6 g at the end ofthe experimental period. rat has been biochemically characterized [9] in terms oftissue
This was comparable to the body weight of vitamin B 6 - vitamin B 6 levels and as functionally deficient in neuro-
deficient rats at 12 weeks on the deficient diet. The systolic transmitters serotonin and y-aminobutyric acid (GABA). Brain
blood pressures of pair-fed control rats and rats fed ad libitum regional pyridoxal phosphate (PLP) levels were significantly
a commercial rat chow were comparable until the tenth week (p < 0.05) reduced after 8 weeks ofvitamin B6 depletion. PLP
on the respective diets. From then on, the systolic blood levels (mean ± S.E.M. of five determinations) in cerebral
139

cortex, hippocampus and thalamus were 5.5 ± 0.39, 7.45 ± 0.45 of the rats with pyridoxine restored the blood pressure and
and 8.70 ± 0.36 nmol!g, respectively, in controls (pair-fed a catecholamine levels to normal within 24 h. Pyridoxine
vitamin B6-supplemented diet) and 3.36 ± 0.40, 5.20 ± 0.32 administration to control rats had no significant effect on
and 6.3 ± 0.34 nrnol!g in vitamin B6-deficient rats, respectively. either of these parameters. The complete reversibility of
We refer to this as the 'moderately deficient' condition. After hypertension in such a short time would preclude a permanent
11 weeks ofvitamin B6 depletion, the PLP levels were further structural damagc to thc vcssel wall of the dcficicnt rat. The
reduced to 2.78 ± 0.28, 4.42 ± 0.35 and 5.10 ± 0.32 nrnol/g in lesion might possibly be at thc level of neurotransmitter
cerebral cortex, hippocampus and thalamus respectively At this regulation. We also determined NE turnover in the hearts
stage of vitamin B6 deficiency the rats were no longer hyper- of deficient and control rats. There was no difference in
tensive. They were normotensive or even hypotensive. We myocardial NE content between the two groups. However,
refer to this is as the advanced vitamin B6-depleted state. It is NE turnover was significantly increased in deficient rats as
the 'moderately vitamin B6-deficient rat' which we have used compared to controls (Table 2), thus supporting the contention
as an animal model of moderate hypertension. that peripheral sympathetic activity is increased in the
pyridoxine-deficient hypertensive anima!.

Characterization 0/pyridoxine-dejiciency hypertension


Cardiovascular ejJects 0/ serotonin
Tbe nature of the hypertension that developed in the pyrido-
xine-deficient animal (B6DHT) needed to be characterized in Serotonin is involved in a wide variety of functions of the
an effort to identify the causative factor(s). Using drugs such central nervous system. Serotonergic cell bodies occur mainly
as phenytoin, valproic acid and diazepam, it was shown that in the raphe nuclei of the brain stern. However the nerve
the hypertension was not the result of a hyper-excitable state axons project into virtually all parts of the brain and spinal
in these animals. Although pyridoxine treatment reversed both cord and thus control a variety of functions such as blood
the hypothyroidism and hypertension there was no indication pressure, emotional behaviour, endocrine function, perception
that the hypothyroid condition initiated hypertension. of pain and sleep. In addition, there are effects on the peripheral
An association between hypertension and sympathetic neurons and non-neural tissues. Serotonin, when administered
simulation has been observed in both hypertensive animals into the brain, elicits complex cardiovascular responses [12].
and humans. Therefore, the possibility that the reversible Depressor, pressor or biphasic responses were reported which
hypertension seen in pyridoxine-deficient rats was related to reflects the non-homogeneous nature ofbrain 5-HT neurons
sympathetic stimulation was considered. The concentration subserving different functions [13]. The effects ofserotonin
of norepinephrine (NE) in plasma is a valid reflection of on the cardiovascular system vary from species to species.
sympathetic activity. However, blood sampies have to be The receptors that mediate these effects are different and have
withdrawn from the conscious animal without trauma. Such been categorized into major families. Each 'family' consists
a system by implanting a vascular-access port (VAP) with of multiple receptor subtypes that share similarities in their
cathetarization to the jugular vein was developed by us [10]. molecular biological, pharmacological, biochemical and
We showed [11] that both epinephrine and norepinephrine physiological properties. These receptors are present
levels in the plasma of pyridoxine-deficient rats were nearly throughout the central and peripheral nervous system [14,
three fold higher compared with controls (Table 1). Treatment 15]. The development of specific agonists and antagonists

Table I. Elfect of pyridoxine on plasma norepinephrine and epinephrine Table 2. Myocardial norepinephrine (NE) content and turnover rates in
in control and pyridoxine-deficient adult rats pyridoxine-supplemented and pyridoxine-deficient adult rat

Animal status Norepinephrine (nmol/L) Epinephrine (nmol/L) NE content Ne turnover rate


(ng/g) (ng/g/h)
Group I: pyridoxine-
supplemented (control) 3.06 ± 0.28 1.89 ± 0.28 Pyridoxine-supplemented 1661.8 ± 241.8 30 ± 4.4
Group 2: pyriodoxine- Pyridoxine-deficient 1955.0 ± 260.8 106.6 ± 14.2*
treated (Control) 3.44 ± 0.27 1.52 ± 0.16
Alpha-methyl tyrosine (200 mg/kg) was injected into the tail vein of
Group 3: pyridoxine-
deficient (experimental) 9.04 ± 0.21" 4.39 ± 0.32" rats. At various time intervals thereafler rats were killed by cervical
Group 4: pyridoxine- dislocation, hearts rernoved, rinsed free ofblood and extracted with ice
treated (experimental) 3.97 ± 0.32 2.73 ± 0.24 cold 0.4 N perchloric acid. NE was assayed using HPLC with
electrochemical detection. Siope of decline was calculated by the method
Values are means ± S.E.M. of8-12 separate determinations in each group. of least squares. Values represent mean ± S.E.M. of 8 separate
"p < 0.01, compared with Groups 1,2, and 4 by Duncan's multiple range experiments. *p < 0.05. (From Viswanathan el al. [11]; reprinted with
test. (From Paulose el al. [6]; reprinted with permission ofthe publisher.) perm iss ion ofthe publisher).
140

that exert activity at different receptor subtypes has permitted


the elucidation ofthe roles ofvarious 5-HT receptor subtypes

~~
on cardiovascular function. That serotonergic receptors 01 -10
subserving vascular smooth musc1e contraction are of the 5- :r
HT2 type has been recognized [16]. The introduction of E
E
ketanserin as a selective 5-HTz antagonist in the treatment of -20

hypertension [17, 18] was followed by a controversy whether a.


a:l
the anti-hypertensive effect is caused by serotonergic blockade CI)
-30
or by a a-adrenoreceptor blockade or by a combination of ~
both [19]. It is now recognized that pure 5-HTz antagonists ..I
..I
do not duplicate the effects of ketanserin and its effect seems <
u..
-.0
to be related to the functions of the vascular endothelium.
The development of centrally acting 5-HT agonists such as
-50~-r---r--~--~--~--~--~--~--~-
8-hydroxy-2-(di-n-propylamino) tetralin (8-0H-DPAT), 0.001 0.003 0.01 0.03 0.1 0.3 1.0 3.0 10
ipsapirone and flesinoxan with specificity to 5-HT1A subtype
receptors has lead to the recognition that 5-HT1A receptors Log Dose lumoles/kg)
are involved in the central control of autonomie flow [20-
22]. Fig. 2. Effect of i.p. injection ofvarious doses of 5-HT1A receptor agonists
It is possible that the decrease in neuronal5-HT and the on systolic blood pressure of conscious vitamin B,-deficient hypertensive
(145 mm Hg) rats. (.) 8-0H-DPAT; (A) f1esinoxan; (+) urapidil and (T)
consequent changes its receptors, particularly 5-HT1A may
5-methyl urapidiI. Each value represents the mean :!: S.E.M. Of 8-10 rats.
cause hypertension in these animals. Hence, we investigated (From LaI and Dakshinamurti [23 ]; reprinted with permission of the
the effect of serotonergic 5-HT1A receptor agonists such as publisher).
8-0H DPAT, flesinoxan, urapidil or methyl urapidil on the
systolic blood pressure of conscious vitamin B6 -deficient
hypertensive rats. After hypertension developed and reached 20 mm Hg were compared, the following rank order was
its peak (8-10 weeks on the deficient diet) the rats were used established: 8-0H-DPAT (0.0046 ,umol/kg»flesinoxan
for assessing the effect ofthe drugs under investigation. They (0.034,umoUkg) > 5-methyl urapidil (l,umol!kg) > urapidil
were injected in a volume of 1 ml!kg body weight i.p. The (2.5 ,umol!kg). None ofthese drugs however, had a consistent
drug doses used are indicated in Fig. 2. Each study was done effect on heart rate. The affinity ofthe agonists for the 5-HT1A
using aseparate group of ~ rats. Systolic blood pressure receptor site [24] correlates with the order of their anti-
and heart rate were recorded at 0.5, 1,2,3,4 h respectively after hypertensive activity indieating that this effect is mediated
injection. The observed effect was compared with pre- through the 5-HT'A receptor site. The selective 5-HT1A receptor
injection values for possible changes in blood pressure and antagonist, spiraxatrine [25] dose dependently antagonized the
heart rate. Deviation from these values represented the hypotensive activity of 5-HT'A receptor agonists.
changes due to the drug under investigation. As shown in The moderately pyridoxine-deficient hypertensive rats
Fig. 2, the i.p. injection of 8-0H-DPAT (0.001-10 ,umol!kg), have a low concentration of serotonin in various brain areas,
flesinoxan (0.03-10 ,umol/kg). Urapidil (0.1-10 ,umol!kg), reflected in the increased 5-HT1A receptor number in
or 5-methyl urapidil (0.1-10 ,umol!kg) caused a significant membrane preparations. Lesioning of central serotonergic
fall in systolic blood pressure ofvitamin B6 -deficient hyper- tracts with 5,7 dihydroxytryptamine results a similar increase
tensive rats [23]. The effect c1early depended on the dose used in the 5-HT'A receptor numbers. These results indicates that
and was seen within 30 min after drug administration. The central serotonergic depletion is one of the contributors to the
peak time period and duration of effect varied depending on development of hypertension. What is the mechanism of the
the drug and its dose. Adose ofO.01,umol/kg of 8-0H-DPAT hypertensive action of 5-HT'A agonists? The hypertension of
caused a fall in systolic blood pressure of 24 ± 1 mm Hg for the moderately pyridoxine-deficient rat is characterized by
only 30 min after its injection whereas a higher dose of 0.1 central sympathetie stimulation as in other hypertensive
,umo!kg caused a similar decrease but the effect lasted for 60 animal models. When O<:z adrenoreceptor in the nuc1eus
min. The highest dose of flesinoxan decreased systolic blood tractus solitarii (NTS) are stimulated inhibitory neurons of
pressure to 44 ± 3 mm Hg and the effect las ted for 3 h. the vasomotor center are activated. Sympathetic out flow,
However, the duration of the hypotensive effect was com- which originates from the vasomotor center and innervates
parable for both 8-0H-DPAT and flesinoxan. The hypo- the peripheral vasculature, heart and kidney is reduced. As a
tensive effects of urapidil and 5-methyl urapidil were c1early result, peripheral vascular tone, heart rate and renin release
much smaller than those of 8-0H-DPAT and flesinoxan. are decreased resulting in a decrease in total peripheral
When doses that caused a fall in systolic blood pressure of resistance and cardiovascular output. Drugs such as c1onidine,
141

an =2 agonist, exert their cardiovascular effect through physical stimuli [31]. Calcium influx essentially occurs
stimulation of the =2 adrenoreceptors in the brain stern. through plasma membrane Ca2+ channels, receptor-operated
Activation of central =2 adrenoreceptors in the NTS required andvoltage-operated [32]. Voltage sensitive calcium channels
a serotonergic input through 5-HT lA receptor [26]. Hence, the participate in action potential generation bymediating voltage-
hypotensive action of 5-HT 1A receptor agonists. activated inward movement of calcium ions which depolarize
The end result of centrally mediated sympathetic stimulation the cell [33]. These channels couple cell surface electrical
is an increase in peripheral resistance. This is reflected in signals to the physiological response by mediating voltage-
elevation ofboth resting and stimulated vascular tone in the dependent increase in the cytosolic concentration of calcium,
resistance arteries of the moderately pyridoxine-deficient which is a key intracellular second messenger. The slow
hypertensive rats as seen in Fig. 3. Elevated peripheral channel (L-type) is the major pathway by which calcium ions
resistance is the hall mark of hypertension as seen in other enter the cell during excitation for initiation and regulation
models of hypertension [27, 28]. The increase in tone of of the force of contraction of cardiac and skeletal musc1e.
caudal artery segments from the hypertensive deficient rats Vascular smooth muscle also contains the (L-type) slow
is calcium dependent as seen by the decrease in unstimulated channe!. We evaluated the possibility that in the pyridoxine-
isometrie tension observed in response to the presence ofthe deficient hypertensive rat a higher concentration of cytosolic
specific calcium chelator EGTA in the medium and also the free Ca2+ might be responsible for the higher tension in the
response to the addition of calcium to the medium [29]. The vascular smooth musc1e. The higher cytosolic free Ca2+could
decrease in tone following the addition to the medium ofthe be caused by an abnormal increase in the permeability of
calcium channel antagonist, nifedipine, indicates increased dihydropyridine sensitive Ca2+channel ofthe vascular smooth
peripheral resistance resulting from increased permeability muscle plasma membrane. We determined the intracellular
of smooth muscle plasma membrane to Ca2+might be central calcium uptake by caudal artery segments of normal control
to the development ofhypertension [30]. and pyridoxine - deficient rats during progressive depletion
using lanthanum resistant calcium [45 Ca] uptake as an index
[23]. As seen in Fig. 4, in the prehypertensive vitamin B6-
Calcium channels deficient phase (weeks 3-4 on the deficient diet) the 45Ca2+
influx into vascular smooth musc1e was not significantly
The initiation of smooth musc1e contraction is principally different from control values. In the hypertensive (SBP, 145
dependent on the short term increase in cytosolic freecalcium. ± 2 mmHg) the 4lCa+2influx into the vascular smooth musc1e
Calcium moves in and out of the cell and intracellular storage was significantly increased to twice that of the contro!. As the
sites in response to chemical, electrical, pressure and other degree ofvitamin B6-deficiency increased (post-hypertensive

CONTROl. 1.5

z I

Q t EGTA I ca2 +
(j) CI>

zW z
x:::" 1.0
~;;
I- ...e ...J-
u..CI>
~,.
U
ce + ~ '"

PYRIDOXINE·ceFICIENT N .<=

M~V\~.'Il1i~v.v\,jJ.0 Ü~0.5
., E
:1E .. E
o I
(j)
~ Nlfedlplne ~ EGT A. c.2.

z
~'~,~ 0-'--.......-
(3-4) (8-9) (12-14)
...e
WEEKS ON DIET
S mln. TIME
Fig. 4. Basal calcium influx into caudal artery segments of rats fed D
Fig. 3. Time course of isometrie tension changes in isolated segments of control (vitamin B.-supplemented) or • vitamin B.-deficient diets for
eaudal artery from eontrol (top two Iraces) and vitamin B6-deficient rats various time periods. Data are mean values for 9 rats in each group. Bars
(bottom two Iraees). Effeet of addition of I IlmollL nifedipine or/and 5 represent standard error ofmean;'" p <0.05 with respeet to control;. p <
nmol/L EGTA-Na or 2.5 mmollL calcium is indieated by arrows. (From 0.0 I with respect to control (student's t-test). (From Lai et al. [47]; reprinted
Viswanathan et al. [29]; reprinted with perrnission of the publisher). with perrnission ofthe publisher).
142

phase, after 12 weeks on the deficient diet) therewas a further -30

large increase in 45Ca2+ uptake. The mechanisms leading to a


decrease in SBP during this advanced phase of deficiency are 0.
::t:
not under-stood yet. The possibility of altered sensitivity of
E -20
the contractile apparatus to increased intracellular calcium .§
and calmodulin cannot be discounted [34]. It is significant c..
that changing the diet of vitamin B 6-deficient rats in the ce
Vl
posthypertensive phase (after 12-14 weeks on the deficient ~
-10
diet) to that of a vitamin B 6 containing normal diet led to a --'
--'
significant increased in their SBP within 2 weeks. oe(
LI.. ***
In a further study we explored the defects in the calcium
channel of the vitamin B 6 -deficient hypertensive rat by
0
studying the effect of calcium channel antagonists on the C D
systolic blood pressure of conscious hypertensive rats [35].
They were on the deficient diet for 7-10 weeks and had a Fig.6. Effect ofintraperitioneal injection ofvehicle (A), 100 ng/kg BAY
SBP of 145-150 mm Hg. The drugs to be tested were K 8644 (B), 100 ng/kg nifedipine (e) and 100 ng/kg BAY K 8644 + ] 00
injected intraperitoeally. SBP and heart were recorded at ng/kg nifcdipinc (D) on systolic blood pressure (SBP) in conscious rats in
which hypertension (SBP, 145--150 mrnHg) had been induced by vitamin
time intervals up to 4 hiater. The SBP was compared with
B6 -deficient diet for 7-10 weeks. Values are expressed as means ± S.E.M.
pre-injection values for possible changes in blood pressure (n = 8-10). ***p < 0.001, vs. (C). (From Lai and Dakshinamurti [35];
and heart rate. As seen in Fig. 5 the intraperitoneal injection reprinted with permission nfthe publisher.
of nifedipine produced a dose-dependent decline in SBP. All
of the other calcium channel antagonists used (results not
shown here) were also effective in lowering the SBP of the deficient rat. However, when it was co-administered with
vitamin B 6- deficient hypertensive rats. The rank order of the dihydropyridine calcium channel antagonist, nifedipine
potency was: nifedipine> (-) 202-791> (±) verpamil > at equimolar doses, BAY K 8644 significantly antagonized
diltiazem. The specificity ofthe effect was seen in the effect the hypotensive effect of nifedipine. BAY K 8644 is known
of the dihydropyridine agonist, BAY k 8644 on the SBP of to prolong the open state of calcium channels during activation
the vitamin B6-deficient hypertensive rats. As seen in Fig. 6 [36,37], thereby promoting calcium entry into the cello The
injection of BAY K 8644 did not alter the SBP of the failure ofBAYK 8644 to do this in the deficient hypertensive
rats suggests that the dihydropyridine sensitive calcium
channel is probably maximally open, suggesting that the
10 - vitamin B6 status might be an important contributor to the
variation in calcium channel function.

/
CJ>
In further experiments we investigated the relationship
0
I between the vitamin B6 levels in the diet and some dietarily

//
E induced hypertensive conditions. Dietary manipulations such
E
-10 as decrease in the calcium content of the diet or increase in
c.. the sucrose or fructose content ofthe diet lead to a consistent,
co
Vl although modest, increase in systolic blood pressure. The
-20
?; effect of altering the level of calcium in the diet at different
;0:--;0:--
--' phases - prehypertensive, hypertensive and post hypertensive
--'
«
u... -30 - ofvitamin B6 deficiency was studied [38], As seen in Fig. 7,
lowering dietary calcium, from 1.()...{).1 % caused a significant
increased in the SBP in rats on a vitamin B6-sufficient diet.
-40
This occurred during weeks 3- 4 on the 10w calcium diet.
I I I I i I Similar efl'ects of a low calcium diet on blood pressure have
o 30 60 90 120 150 180 been reported by others [39, 40]. Low levels of calcium in
the diet potentiated the hypertension induced by the vitamin
Fig. 5. Effect ofintraperitoneal injection of 1 C.), 3 CA), 10 C'f'), 100 C+) B6-deficient diet when both deficiencies were present from
and 1000 (*) nmol/kg nifedipine on systolic blood pressure (SBP) in the beginning ofthe experiment. Feeding a low-calcium diet
conscious rats in which hypertension (SBP 145-150 mrnHg) had been
induced by feeding a vitamin B6-deficient diet for 7-10 weeks. Values are
during the hypertensive or post hypertensive phase failed
expressed as means ± S.E.M. [n = ]0-12,]. (From La] and Dakshinamurti to raise the SBP in these rats. Normalizing the vitamin B6
[35]; reprinted with permission of the publisher). status ofpost hypertensive vitamin B6-deficient rats restored
143

rats resulted in a modest elevation ofSBP. This was attenuated


by the inc1usion of a vitamin B6 supplement (five times the
normal intake) in their diet [52] The results show that
hypertension induced by dietary means such as low calcium
or increase in simple carbohydrates in the diet of rats
receiving normal amounts of vitamin B6 in their diet respond
to a dietary supplement ofvitamin B 6 (five times the normal
intake).
In further work we investigated whether a dietary supple-
ment ofvitamin B6 could attenuate the elevation of systolic
blood pressures in genetically hypertensive animal models
such as the Zucker obese or spontaneously hypertensive rats.
, I i , The Zucker obese (fa/fa) rat was originally studied as a model
4 6 10 11 12
WEEKS ON DIET
of obesity and atherosclerosis and has found extensive use
in the study of diabetes mellitus [I, 2]. Various reports have
shown that the Zucker obese rat also develops hypertension,
Fig. 7. Effecl offeeding a vilamin B 6 -sufficient or a vilamin B 6 -deficient
diet containing either normal calcium (1 %) or low calcium (0.1 %) on the which is specifically associated with the obese genotype (fa/
systolic blood pressure of rats. Blood pressure was measured by lail-cuff fa). In contrast, Zucker lean rats are normal in al1 parameters.
plethysmogaraphy in conscious rats. Values are expressed as means ± Metabolic alterations associated with obesity are believed to
S.E.M. for 12 rals .• Vitamin B6 -sufficient (normal diet) + normal calcium; be the pathogenetic determinants of hypertension. Caloric
0, vitamin B6-sufficient (normal diel) + low calcium; ß, vitamin B6-deficient
restriction of Zucker obese rats reduced the weight gain but
diet + normal calcium; Ä, vitamin B6-deficient diet + low calcium. (From
LaI and Dakshinamurti [38]; reprinled with permission of the publisher). did not attenuate the hypertension. The spontaneously
hypertensive rats (SHR) have been used extensively as an
experimental model for the study of essential hypertension
the ability of low dietary calcium to increase SBP in these in humans [53].
rats. Zucker obese rats (fa/fa), spontaneously hypertensive rats
It has been suggested that reduced dietary calcium depletes and their corresponding controls were tested for the effects
calcium from membrane storage sites, causing a less stable ofvitamin B6 ingestion in different ways: (1) Vitamin B6 was
membrane ofthe vascular smooth muscle [41,42]. This inc1uded as a supplement (five times the normal intake) from
results in enhanced calcium influx, increased tone and the start ofthe experiment until development ofhypertension;
reactivity. Peripheral resistance is elevated leading to (2) vitamin B6 supplement was removed from the diet of
hypertension. Stabilizing the membrane abnormality in Zucker obese and Zucker lean control groups after 16 weeks
vascular smooth muscle by using high dietary calcium has on the dietary supplements; (3) a diet deficient in vitamin B6
been demonstrated [43--45]. We have shown [46] that high was instituted in SHRs and control Wister-Kyoto (WKY) rats.
dietary calcium also reduces hypertension in rats with vitamin The SBP of rats in all groups was monitored in the conscious
B6 deficiency-induced hypertension, as has been shown in animal by tail-cuff plethysmography. The effect of the various
other models of hypertension such as the one-kidney treatments on the uptake of calcium by caudal artery segments
deoxycorticosterone - sodium chloride hypertensive Wister were also examined.
rat [47]. A low-calcium diet decreases ionic serum calcium. Male Zucker obese rats (fa/fa) of age 6 weeks fed a
Vitamin B 6 deficiency also appears to cause a similar commercial rat chow developed hypertension in 3-4 weeks,
abnormality. Calcium uptake by enterocytes is reduced in whereas their lean controls (Fa/Fa) did not. Similar increases
vitamin B 6 deficiency. An interesting finding was that an in the SBP ofthe Zucker obese rat have been reported using
increased concentration of vitamin B6 in the diet attenuated direct [54] and indirect [55] measurements. As seen in Fig.
the blood pressure increasing effect of low dietary calcium. 8 the inclusion of a vitamin B6 supplement (five times the
Vitamin B6 might correct the membrane abnormality by a normal intake) resulted in a complete attenuation of the
mechanism similar to that ofthe calcium channel antagonists hypertension in the obese strain. Infact, the SBP ofthe obese
[45,48]. rats after 16 weeks of supplementation was lower than their
Acute or chronic ingestion of simple carbohydrates such initial level. Heart rate was also lowered as a result offeeding
as sucrose or fructose has been shown to cause an increase the high vitamin B6 diet. The age associated increase in the
in systolic blood pressure ofvarying degrees in several strains Zucker lean rat was also attenuated by the high vitamin B6
ofrat [4, 49]. Anenhancement ofSBP due to sucrose feeding diet. When the high vitamin B6 diet of the Zucker obese rats
has also been shown in the spontaneously hypertensive rat was changed to a normal vitamin B6 dict the SBP ofthc rats
[50,51]. The ingestion of sucrose by male Sprague-Dawley increased by about 30 mmHg in 12 days. Similar switch of
144

180 190

170 180

Oi
~ 160 :r: 170
E
E
.s .s
.
!'! 150
~ ...
!!!
~

!!!
160

!'! 0..
~ 140 '0
0 150
o 0
o iD
äi .!!
~ 130 '0 140
'0
'0 >-
>-
rn rn
120 130

120 ~---r---,----~---r---,----.----,----,
o 2 4 6 8 10 12, 14 16 18 o 2 3 4 5 6 7
Weeks On Die! Weeks On Die!

Fig.8. Effect offeeding a diet containing nonnal or high (five times nonnal) Fig. 9. Effect of feeding a nonnal or a high-vitamin B 6 diet on systolic
levels ofvitamin B 6 to Zucker obese and Zucker lean rats. Systolic blood blood pressure in spontaneously hypertensive (SHR) and control (WKY)
pressure was measured by tail-cuff plethysmogaraphy in conscious animals. rats. Blood pressure was measured by tail-cuff plethsmography in
Each value represents the mean ± S.E.M. of \0 rats. 0, Zucker obese rat conscious animals. Each value represents the mean ± S.E.M. of 10 rats.
fed a normal vitamin B 6 diet; ., Zucker obese rat fed a high vitamin B6 0, SHRs on a nonnal vitamin B, diet; ., SHRs on a high-vitamin B 6
diet; 0, Zucker lean rat fed a nonnal vitamin B6 diet; ., Zucker lean rat diet; 0, WKY rats on a nonnal vitamin B 6 diet; ., WKY rats on a high-
fed a high vitamin B 6 diet. (From LaI and Dakshinamurti [52]; reprinted vitamin B6 diet. (From Lai and Dakshinamurti [52]; reprinted with
with pennission ofpublisher). pennission ofpublisher).

the diet in the Zucker lean rat also resulted in a rise in the SBP 08
within 2 days to the same level as shown by Zucker lean rats
fed the normal vitamin diet for the entire experimental period.
...
., 0.7
In contrast to the effects seen in the Zucker obese rats, there ~

was no response to the inclusion or removal of dietary vitamin t


o
0.6
B 6 supplement int he SHRs (Fig. 9). However the WKY rats
responded to both these conditions in a manner similar to that
:c
~ 0.5
seen in the Spraque-Dawley strain. Thus, the SBP ofSHRs, ~

..
0>
unlike that of WKY rats was insensitive to the vitamin B. ~ 0 .4
concentration in the diet. The changes in SBP in the Zucker '0
as weil as in the sucrose-fed rats correlated with changes in
the uptake of calcium by caudal artery segments in all these
.sE 0.3
~
groups (Fig. 10). This is the first observation that animal
ä 0 .2
models ofhypertension can be classified on the basis oftheir
.
~

response to a vitamin B 6 supplement. On this basis, the .f' 0.1


etiology ofhypertension in SHRs is quite distinct from that
in Zucker obese rats. 0 .0
In view ofthe results showing increased calcium uptake by
caudal artery segments from vitamin B 6-deficient hypertensive
rats and their attenuation by in vitra addition of nifedipine to Fig.10. Influx of"Ca'+ into caudal artery segments. Values are expressed
as means ± S.E.M. (n = 6). Zucker rats: ZOB - Zucker obese rats on a
the incubation medium [29] we investigated the possibility
nonnal vitamin B6 diet; ZOBHB - Zucker obese rats on a high-vitamin B6
that pyridoxine or more particularly, pyridoxal phosphate diet; ZL - Zucker lean rats on a nonnal vitamin B6 diet; ZLHB - Zucker
could directly modulate the cellular calcium uptake process. lean rats on a high-vitamin B, diet. (From Lai and Dakshinamurti [52];
Cold lanthanum resistant 45 Ca2+ uptake by segments of caudal reprinted with pennission ofpublisher).
145

artery was determined as described [29). The effect of 0.600


pyridoxal phosphate on the BAYK 8644-induced 45Ca2+ influx
was examined in artery segments from control (normal) rats. 0.500
Q)
The DHP-sensitive calcium channel agonist [36] was :l

ineffective in increasing further the basal calcium uptake by ~ ~ 0.200


...J_
u.. Q)
caudal artery segments from vitamin B6-deficient hyper- z ;:
tensive rats. However, BAY K 8644 stimulated 45Ca2+ entry i ~ 0.300
•• ••
os.J::;
into artery segments from control (normal) rats (Fig. 11). ü::.
~ ~ 0.200
Pyridoxal phosphate dose dependently (0.1-10 11M) reduced
the BAYK 8644 - stimulated calcium uptake by control artery .s
0.200
segments. As seen earlier [29] the basal uptake of 45Ca2+ by
caudal artery segments from vitamin B6-deficient hypertensive
rats was at least twice the uptake by artery segments from
control (normal) rats. Pyridoxal phosphate or nifedipine
added to the incubation medium reduced significantly the
45Ca2+ uptake by artery segments from the deficient hyper-
Fig. 12. Effect of in vilra addition of dihydropyridine and/ar pyridoxal
tensive rats (Fig. 12). However, in presence ofBAY K 8644 phosphate on 45Ca'+ influx into eaudal artery segments from vitamin B 6 -
(which by itselfhad no effect) in the incubation medium both defieient hypertensive rats. Basal- no addition to incubation medium; PLP
pyridoxal phosphate and nifedipine were much less effective - pyridoxal phosphate; NFP - Nifedipine. Bars represent standard error of
in attenuating the 45Ca2+uptake by artery segments from the mean values far 9 rats in each group .• p < 0.05 with respect to basal;'" p
deficient hypertensive rats. These in vitro direct antagonisms < O. 5 with respeet to BAY K 8644 alone; • p < 0.05 with respeet to
presence ofBAY K 8644, in addition. (From Laielai. [48]; reprinted with
indicate the possibility that the calcium channel agonist BAY perrnission ofpublisher.)
K 8644, the calcium channel antagonist, nifedipine and
pyridoxal phosphate might all act at the same site on the
calcium channel. pyridoxal phosphate for 30 min at 30°C. Figure 13 andTable
We have also examined the effect ofpyridoxal phosphate 3 show the Scatchard plots of the specific binding of [5-
on the binding of tritiated nitrendipine, a dihydropyridine methyI3H]-nitrendipine to crude membrane preparations from
calcium channel antagonist, to membrane preparations from caudal artery in the presence or absence of pyridoxal phos-
caudal artery ofnormal rats. Crude membrane preparations phate. The Scatchard plot analysis of the data revealed that
were preincubated in presence (15 11M) or absence of maximal number of binding sites (B m, ) was increased by
pyridoxal phosphate. Pyridoxal phosphate also increased the
affinity of the antagonist ligand to the membrane preparation.
0.300 Pyridoxal phosphate in vitro attenuates the influx of
• extracellular calcium. This effect is achieved through
modulation ofligand binding. This is a demonstration ofthe
Q)
:l action of pyridoxal phosphate other than as a cofactor of
~ ~ 0.200


...J_ pyridoxal phosphate - dependent enzymes. This is analogues
u.. Q) to the effect of pyridoxal phosphate on steroid hormone
Z ;:
i~ activity [56, 57]. The addition ofpyridoxal phosphate to the
os:;::
ü::.
~ ~ 0.100 • medium containing mouse mammary gland explants resulted
in significant decrease of both dexamethasone binding to
.s nuclear steroid receptor and dexamethasone-stimulated
casein mRNA synthesis [58]. Tbe possibility that pyridoxal

o...l..-...JBA-S...JA-L--~ BÄYK 864'4(1~M)'=-­


Table 3. Effeet of pyridoxal phosphate (PLP) on the binding of [3H]
PLP (~M) - - - . nitrendipine to membrane preparations from rat tai! artery
o 0.1 0.3 1.0 10
Kd* Bmax*
(nM) (fmol mg- l
Fig. J J. In vitra effeet of pyridoxal phosphate addition to the ineubation
medium on the BAY K 8644 indueed 45Ca 2+ influx into eaudal artery protein)
segments from control rats. Basal, no addition to the incubation medium. Control 0.69 ± 0.04 98 ± 6
Bars represent standard error of mean values for 9 rats in eaeh group .• p Control + PLP (15 J.lm) 0.57 ± 0.03+ 150±?+
< 0.05 withrespeet to basal;'" p < 0.05,. P < 0.01 with respectto BAY K
8644 alone. (From Lai el al. [48]; reprinted with perrnission ofpublisher). *mean ± S.E.M. 6 experiment; +p< 0.05 with respeet to contro!.
146

300

250
-::;
c:
-0>
E
200
ci
....,E
W 150
W
a:
LI..
Cl
Z 110
:::)
0
m
50

o 20 40 60 80 100 120 140 160 180 200

SOUND (f mol. mg' )

Fig. 13. Scatchard plot of [5-methyl JH] nitrendipine binding to crude membrane preparations from caudal artery of normal rats: . , in the absence and_,
in the presence of pyridoxal phosphate (15 flM).

phosphate might act as a modulator of protein DNA interaction channel [67]. The search of endogenous ligands for the calcium
has been suggested [59]. channel is in its infancy [68]. The evidence presented here
The existence of high affinity binding sites for calcium indicates the possibility that pyridoxal phosphate might be
channel agonists and antagonists has prompted a search for an endogenous modulator of calcium transport.
endogenous ligands for these sites. The characterization of
endogenous ligands for opiate and benzodiazepine [60]
receptors have been weH characterized. Voltage-sensitive Acknowledgment
calcium channels undergo long-term modulation by neuro-
transmitters and a variety of second messengers. Activation This work was supported by grant from the Heart and Stroke
of the channels is enhanced by cAMP and AMP-dependent Foundation of Canada.
protein kinase [61]. In common with the pharmacological
receptors, calcium channel are regulated by homologous and
heterologous factors. Chronic channel activation, chronic
drug exposure, hormonal influence and specific diseases are
References
all associated with alte red expression of calcium channel
I. Zucker LM, Zucker TF: Fatty, a new mutation in the rat. J Hered 62:
fimction and numbers [62]. The action of drugs at the calcium
275-278, 1967
channels would indicate that endogenous factors or ligands 2. Bray GA: The Zucker-falty ratA review. Fed Proc 36: 148--153, 1977
might serve as physiological regulators, a function which is 3. Paradise NF, Pilorti CF, Payne WR and Finkelstein JA: Left ventricular
mimicked by calcium channel agonists or antagonists. This functin of the isolated, genetically obese ra!'s heart. Am J Physiol
does not include regulation via second messengers or coupling 248: H438-H444, 1985
4. Preuss HG, Zein M, Knapka J, MacCarthy P, Yousufi AK, Gleim
proteins. High and low molecular weight factors - both peptide
GW, Glace B, Zukowska-Grojec Z: Blood pressure response to
and non-peptide - isolated from rat brain seem to inhibit suerose ingestion in four rat strains. Am J Hypertens 5: 244-250,
dihydropyridine binding with tissue specificity [63, 64]. The 1992
contents of absorbate and iron in brain extract was sufficient 5. Zein M, Areas JL, Knapka J, MacCarthy P, Yousufi AK, Di Pette D,
to explain the inhibition of [3H] PN 200-100 binding [65]. Holland B, Goel R, Preuss M: Excess sucrose and glucose ingestion
acutely elevate blood pressure in spontaneously hypertensive rats. Am
Factors isolated from erythrocytes and blood ofhypertensive
J Hypertens 3: 380-386, 1990
rats also seem to enhance calcium influx into vascular smooth 6. Paulose CS, Dakshinamurti K, Packer S and Stephens NL: Sympathetic
muscle [66]. Palmitoyl camitine has been shown to inhibit the stimulation and hypertension in the pyridoxine-deficient adult rat.
bin ding of nitrendipine, verpamil and diltiazem to the calcium Hypertension 11: 387-391, 1988
147

7. Dakshinamurti K, Stephens MC: Pyridoxine deficiency in the neonate 29. Viswanathan M, Bose R, Dakshinamurti K: Increased calcium influx
rat. JNeurochem 16: 1515--1522, 1969 in caudal artery of rats made hypertensive with pyridoxine deficiency.
8. Bunang RD, ButterfieId J: Tail cuff blood pressure measurements Am J Hypertens 4: 252-255,1991
without external preheating in the awake rat. Hypertension 4: 898- 30. Rapp JP, Nghiem CX, Oniwochei MO: Aortic calcium uptake and effux
903, 1982 in spontaneously hypertensive and inbred Dahl rats. 1 Hypertens 4:
9. Dakshinamurti K, Paulose, CS, Viswanathan M, Siow YL, Sharma 493-499, 1986
SK, Bolster B: Neurobiology of pyridoxine. Ann NY Acad Sci 585: 31. Bean HP: Classes of calcium channcls in vertebrate cells. Ann Rev
129-144,1990 Physiol 51: 367-384, 1989
10. Paulose CS, Dakshinamurti K J: Chronic catheterization using 32. Glossman H, Streissnig J: Calcium channels. Vitamins Hormones 44:
vascular-access-port in rats: Blood sampling with minimal stress for 155-328, 1988
plasma catecholamine determination. J Neurosci Methods 22: 141- 33. Catterall WA, Seager MJ, Takahashi M: Molecular properties of
146,1987 dihydropyridine-sensitive calcium channels in skeletal muscle. J Biol
11. Viswanathan M, Paulose KJ, LaI SK, Dakshinamurti, K: Alterations Chern 263: 3535-3538, 1988
in brain stern U, adrenoreceptor activity in pyridoxine-deficient rat 34. Ngheim CX, Rapp JP: Responses to calcium of chemically skinned
model ofhypertension. Neurosci Lett 111: 201-205,1990 vascular smooth muscle from spontaneously hypertensive rats. Chn
12. Dalton DW: The cardiovascular effects of centrally administered 5- Exp Hypertens [Al 4: 849-856, 1983
hydroxytryptamine in the conscious normotensive and hypertensive 35. LaI KJ, Dakshinamurti KJ: Calcium channels in vitamin B6 deficiency-
rat. J Auton Pharmacol 6: 67-75, 1986 induced hypertension. Hypertension 11: 1357-1362, 1993
13. WolfWA, Kuhn DM, Lovenberg W: Serotonin and central regulation 36. Schram M, Thomas G, Towart G, Francowiak, G: Novel dihydro-
of arterial blood pressure. In: P.M. Vanhoute (ed.). Serotonin and the pyridines with positive inotropic action through aetivation of Ca"
Cardiovascular System, Raven Press, New York, NY, pp 63-73, 1985 channels. Nature 303: 535-537, 1983
14. Peroutka SJ: Receptor 'families' for 5-hydroxytryptamine. J Cardiovasc 37. Kokbun S, Reuter H: Dihydropyridine derivatives prolong the open
Pharmacol16: S8-S14, 1990 state of calcium channels in cultured cardiac cells. Proc Nat! Acad Sei
15. Bonate PL: Serotonin receptor subtypes: functional, physiological and (USA) 81: 4824-4827,1984
clinical cOITelates. Clin Neuropharmacol14: 1-16, 1991 38. LaI KJ, Dakshinamurti K: Relationship between low calcium induced
16. Cohen ML, Mason N, Wiley KS, Fuller RW: Further evidence that hypertension and vitamin B6 status. J Hypertension 13: 327-332, 1995
vascular serotoninreceptors are ofthe 5HT, type. Biochem Pharmacol 39. Schleiffer R, Pornot 1', Berthelot A, Gairard A: Low calcium diet
32: 567-570, 1983 enhances development of hypertension in the spontaneously hyper-
17. Van Nueten JM, Leyson JE, Schuurkes JAJ, Vanhoute PM: Keanserin: tensive rat. Cl in Exp Hypertens [Al 6: 783-793, 1984
a selective antagonist of5HT, serotonergic receptors. Lancet 1: 297- 40. Baksi SN, Abhold R H, Speth RC: Low calcium diet increases blood
298, 1983 pressure and alters peripheral but not cenlral angiotensin II binding
18. Van Neuten JM, Leyson JE, de Clark F, Vanoute PM: Serotonergic sites in rats. J Hypertens 7: 423-427,1989
receptor subtypes and vascular reactivity. J Cardiovasc Pharmaacol 6: 41. Bolton JB: Mechanisms ofaction oftransmitters and other substances
S564-S574, 1984 on smooth muscle [review]. Physiol Rev 59: 606-718,1979
19. Schalekamp MADH: Serotonergic blockade and hyptertension. In: 42. Brickman AS, Nyby MD, Von Hungen K, Eggena P, Tuck ML:
P.M. Vanhoute (ed.). Serotonin and the Cardiovascular System, Raven Calcitropic hormones, platelet calcium and blood pressure in essential
Press, New York, NY, pp 135 145,1985 hypertension. Hypertension 16: 515-522, 1990
20. Arvidsson LE, HackseH V, Nilsson ILG, Hjorth S, Carlsson A, 43. Bukoski RD, McCaITon DA: Altered aortic reactivity and lowered
Lindberg P, Sanchez 0, Wikstrom H: 8-hydroxy-2-(di-n-propylamino) blood pressure associated with high calcium intake. Am J Physiol25:
tetralin, a new centrally acting 5-hydroxytryptamine receptor agonist. H 978-H983, 1986
J Med Chem 24: 921-923,1981 44. Porsti I: Arterial smooth muscle contractions in spontaneously
21. Wouters W, Tulp MT, Bevan P: Flexinoxan lowers blood pressure and hypertensive rats on a high calcium die!. 1 Hypertens 10: 255-263, 1992
heart rate in cats via 5HT'A receptors. Eur 1 Pharmacol149: 213-223, 45. Dominiczak AI', Bohr DF: Cell membrane abnormalities and the
1988 regulation ofintracellularcalcium concentration in hypertension. Clin
22. Schoeffier P, Hoyer 0: Centrally acting hypotensive agents with affinity Sci Colch 79: 415-423, 1990
for 5HT 'A binding sites inhibit forskolin-stimulated adenylate cyclase 46. Dakshninamurti K, LaI KJ: Vitamins and Hypertension. In: A.P.
activity in calfhippocampus. Br J Pharmacol95: 975-985, 1988 Simopoulas (ed.), World Rev Nutr Diat 69: 40-73, 1992
23. LaI KJ, Dakshinamurti K: Hypotensive action of 5-HT receptor 47. Arvola P, Ruskoaho H, Porsti, I: Effects ofhigh calcium diet on arterial
agonists in the vitamin B 6 deficient hypertensive rat. Eur J Pharmacol smooth muscle functions and electrolyte balance in mineralocorticoid-
234: 183-189, 1993 salt hypertensive rats. Br J Pharmacol 106: 948-958, 1993
24. Groz G, Hantt G, Kolassa N: Urapidil and some analogues with hypo- 48. LaI KJ, Sharma SK, Dakshinamurti K: Regulation of calcium influx
tensive properties show high affinities for 5-hydroxytryptamine (5-HT) into vascular smooth musde byvitamin B6. Clin and Exp Hypertension
binding sites ofthe 5-HT'A subtype and for u,-adrenoreceptorbinding 15: 489-500, 1993
sites. Naunyn-Schmiedberg'sArch Pharrnacol336: 597....{i01, 1987 49. Zein M, Areas JL, Preus HG: Long-term effects of excess sucrose
25. Nelson 0, Taylor WE: Spiroxatrine: A selective serotonin'A receptor ingestion on three strains ofrats. Am J Hypertens 3: 560-562,1990
antagonist. Eur J Pharmacol 124: 207-208, 1986 50. Bunag RD, Tomita T, Sasaki S: Chronic sucrose ingestion induces mild
26. Rapaport A, Sturtz F, Guicheney P: Regulation of central u-adreno- hypertension and tachycardia in rats. Hypertension 13: 896-901,1983
receptor by serotonergic denervation. Brain Res 344: 158-161, 1985 51. roumier RD, Chiueh CC, Kopin U, Knapka JJ, Di Pettte 0, Preuss
27. Noon IP, Rich PJ, Baldessarini RI: Calcium leakage as a cause ofthe HG: Refined carbohydrate increases blood pressure and catecholamine
high resting tension in vascular smooth muscle from the spontaneously excretion in SHR and WKY. Am J Physiol250: E381-E385, 1986
hypertensive rat. Proc :-1atl Acad Sci (USA) 75: 1605-1607,1978 52. LaI KJ, Dakshinamurti K, Thhveris J: The effect ofvitamin B 6 on the
28. Postnov YP, Orlov SN: Ion transport across plasma membrane in systolic blood pressure of rats in various animal models of hyper-
primary hypertnsion. Physiol Rev 65: 904-945, 1985 tension. 1 Hypertension 14: 355-363, 1996
148

53. Okamoto K, Aoki K: Development of astrain of spontaneously of ß andrenergie effectors and cyc1ic AMP on nilrendipine-sensitive
hypertensive rats. Jpn Cire J 27: 282-293, 1963 voltage-dependent Ca'+ channels of skeletal museIe. J Biol Chem 260:
54. Zemel MB, Peuler JD, Sowers JR, Simpson L: Hypertension in insnlin- 13041-13046, 1985
resistant Zucker obese rats is independent of sympathetie neural 62. Fcrrante J, Triggle DJ: Drug and disease-indueed regulation ofvoltage-
support. Am J Physiol262: E368-E371, 1992 dependent calcium channels. Pharmacol Rev 42: 29-44, 1990
55. Yoshioka S, Nishino H, Shiraki T, Ikeda K, Koike H, Okuno A: 63. Lewanczuk RZ, Resnick LM, Ho M-S, Benishin CG, Shan J, Pang
Antihypertensive effeets of CS-045 treatment in obese Zucker rats. PKT: Clinical aspects of parathyroid hypertensive factor. J Hypertens
Metabolism 42: 75-S0, 1993 12 (Supp!.): S11-S16, 1994
56. Litwaek G: The glueoeortieoid reeeptor at the protein leve!. Cancer 64. Lewanezuk RZ, Resnick LM, Blumemfeld JD, Laiagh m, Pany PK:
Res 48: 2636-2640, 1988 A new circulating hypertensive faetor in the plasma of essential
57. Compton MM, Cidlowski JA: Vitamin B 6 and glueoeortieoid action. hypertensive subjeets. J Hypertens 8: 105-108, 1 990
Endoerine Rev 7: 140-148, 1986 65. Ebersoie BJ, MolinoffPB: Identifieation of ascorbate as an endogenous
58. Majumdar PK, Joshi JB, Banerjee MR: Correlation between nuelear substance that inhibits binding of dihydropyridine calcium channel
glueocortieoid reeeptor levels and casein gene expression in murine blockers. J Neurochem 58: 1300-1309, 1992
mammary glands in vitra. J Biol Chem 258: 6793-6798, 1983 66. Simmons MA? Johnson EC, Becker JB, Todd DG, ReichenbecherVE,
59. Meisler NT, Thanassi JW: Pyridoxine-derived B 6 vitamers and McCumbee WD, Wright GL: An endogenous 'hypertensive factor'
pyridoxal 5'-phosphatc-binding protcin in cytosolic and nuelear cnhances the voltage-dependent calcium eurren!. FEBS Lett 254: 137-
fraetions ofHTC cells. J Biol Chem 265: 1193-1198, 1990 140,1989
60. Guidetti A, Forchetti CM, Corda MG, Konkel D, Bennett CD, Cossta 67. Spedding M, Mir AK: Direet activations ofCa'+ channels by pamitoyl
E: Isolation characterization and purification to homogeneity of an camitine? A putative endogenous ligand. Br J Pharmacol 92: 457-
endogenous polypeptide with agonistic action on benzodiazepine 468,1987
receptors. Proc Nat! Acad Sei (USA) 80: 3531-3535, 1983 68. Triggle DJ: Calcium, calcium ehannels and calcium ehannel antagonists.
61. Sehmid A, Renaud J, Luxdunski M: Short term and long term effeets Can J Physiol Pharmacol68: 1474--1481, 1989
Molecular and Cellular Biochemistry 188: 149-159, 1998.
© 1998 Kluwer Academic Publishers.

LDL oxidation by arte rial waD macrophages


depends on the oxidative status in the lipoprotein
and in the ceDs: Role of prooxidants vs.
antioxidants
Michael Aviram and Bianca Fuhrman
The Lipid Research Laboratory, Technion Faculty ofMedicine, The Rappaport Family Institutefor Research in the Medical
Sciences and Rambam Medical Center, Halfa, Israel

Abstract
Oxidized LDL is highly atherogenic as it stimulates macrophage cholesterol accumulation and foam cell formation, it is cytotoxic
to cells ofthe arterial wall and it stimulates inflammatory and thrombotic processes. LDL oxidation can lead to its subsequent
aggregation, which further increases cellular cholesterol accumulation.
All major cells in the arterial wall including endothelial cells, smooth museie cells and monocyte derived macrophages can
oxidize LDL. Macrophage-mediated oxidation ofLDL is probably a hallmark in early atherosclerosis, and it depends on the
oxidative state ofthe LDL and that ofthe macrophages. The LDL oxidative stateis elevated by increased ratio ofpoly/mono
unsaturated faUy acids, and it is reduced by elevation ofLDL-associated antioxidants such as vitamin E, ß-carotene, lycopene,
and polyphenolic flavonoids.
The macrophage oxidative state depends on the balance between cellular NADPH -oxidase and the glutathione system. LDL-
associated polyphenolic flavonoids which inhibit its oxidation, can also reduce macrophage oxidative state, and subsequently
the cell-mediated oxidation ofLDL. Oxidation ofthemacrophage lipids, which occurs under oxidative stress, can lead to cell-
mediated oxidation ofLDL even in the absence oftransition metal ions, and may be operable in vivo.
Finally, elimination ofOx-LDL from extracellular spaces, after it was formed under excessive oxidative stress, can possibly
be achieved by the hydrolytic action ofHDL-associated paraoxonase on lipoprotein's lipid peroxides. The present review article
summarizes the above issues with an emphasis on our own data. (Mol Cell Biochem 188: 149-159, 1998)

Key words: atherosclerosis, lipoproteins, lipid peroxidation, macrophages, antioxidants, polyphenols, carotenoids

LDL oxidation and atherosclerosis (SR-A), the CD-36 molecule and an additional, but hitherto
unidentified binding site [6, 7]. Because cell-surface proteo-
Foam cell formation, the hallmark of early atherosclerosis is glycans were previously shown to be involved in the cellular
characterized by accumulation of cholesterol (esterified and uptake of native LDL [8-10], we questioned whether heparan
unesterified), as weil as of oxidized derivatives of cholesterol sulfate or chondroitin sulfate proteoglycans (HSPG or CSPG)
in macrophages [1-5]. A major contributor to the macrophage on the macrophage surface could also mediate Ox-LDL
loading with cholesterol and cholesterol oxides, is oxidized binding. The addition of either heparinase or chondroitinase,
LDL (Ox-LDL), which can be formed by cclls ofthc arterial or both enzymes simultaneously, to macrophages that were
wall including macrophages. Macrophagc uptake ofOx-LDL pre-incubated with excess concentration of Ac-LDL, together
is mediated by several receptors. Macrophage binding sites with anti CD-36 antibodies (to block both the Ac-LDL
for oxidized LDL include the acetyl LDL (Ac-LDL) receptor receptor and the CD-36 binding sites), further reduced

Address Jor oJJprints: M. Aviram, The Lipid Research Laboratory, Rambam Medical Center, Haifa, 31096, Israel
ISO
cellular uptake ofOx-LDL by 19,29 or 42% respectively, in hydroxy-ß-methyl-glutaryl-CoA reductase inhibitors in
comparison to cells that were pretreated only with excess Ac- hypercholesterolemic patients [25-30], ACE inhibitors in
LDL and anti CD-36 antibodies. These results suggest that hypertensives, [I, 18] selenium in renal fai!ure patients [20],
cellular uptake of Ox-LDL is partially mediated via macro- and carotenoids in diabetic patients, were shown to reduce the
phage proteoglycans [11]. susceptibi!ity oftheir LDL to oxidation. (Table I). These effects
In glycosaminoglycans-enriched macrophages, excess were associated in several studies with areduction in the size
concentration of Ac-LDL, but not of anti CD-36 antibodies, of the atherosclerotic lesion suggesting that these treatments
further reduced cellular uptake ofOx-LDL by 1.5 fold over may contribute to the attenuation ofthe atherosclerotic process
control cells. These results suggest that the added glyco- [29, 31]. (3) Supplementation of nutrients, which are rich in
saminoglycans can also interact with the Ac-LDL receptor, polyphenols (red wine, licorice, olive oi!, ginger, orange peel,
leading to an enhanced macrophage uptake ofthe Ox-LDL [32-38]), or of selenium to humans [20] and to the EO mice
that binds to these glycosaminoglycans. This study thus [39], reduced LDL oxidation as weil as accelerated develop-
demonstrated that macrophage proteoglycans represent a ment of atherosclerotic lesions. Macrophages are activated
unique binding site for Ox-LDL, and this pathway can under oxidative stress [40, 41] and such activation can further
participate in the cellular uptake of Ox-LDL, leading to contribute to LDL modifications by the cells, including
macrophage cholesterol accumulation. Recently we have lipoprotein oxidation (and aggregation), with the consequent
observed, the ability of macrophages to lay down an extra- foam cell formation and acceleration of atherosclerosis
cellular matrix (ECM) to which Ox- LDL could bind, and this development.
ECM-bound Ox- LDL could be taken up and degraded by Recently, we have questioned the relationships between
macrophages [12]. the oxidation and aggregation states of plasma LDL [42,
Evidence for the occurrence of Ox-LDL in vivo is as 43]. For this purpose we studied LDL oxidizability and
follows: (1) The LDL in the atherosclerotic lesion (in humans, aggregability in the EO mice during their aging (and the
as weH as in the apolipoprotein E-deficient (EO) mice) is development of atherosclerosis), in comparison to plasma
oxidized, in comparison to plasma LDL which normally is LDL from control mice. During EO mice aging, oxidation of
not oxidized [13-16]. (2) Plasma LDL frompatients at high theirplasma LDL was found to precede its aggregation [44].
risk for atherosclerosis (such as hypercholesterolemics, To further assess the possible effect ofLDL oxidation on its
hypertensives, diabetics, renal failure patients, obese subsequent aggregation, LDL oxidation was induced by
subjects, and patients with xanthelasma palpebrarum, [17- either copper ions, or by the free radical generator, 2,2-azobis,
24], as well as from the EO mice [14], demonstratesincreased 2-amidinopropane hydrochloride (AAPH), or by hypochlorite.
susceptibility to oxidation, in comparison to normal LDL. All these oxidative systems led to different degrees ofLDL
In some groups of these patients, and in EO mice, LDL is oxidation and resulted in a substantial LDL aggregation. These
minimally oxidized already in the plasma [13, 14, 24]. ~- oxidation systems also enhanced the susceptibility of LDL

Table 1. The effect of drugs aod dietary antioxidaots on the susceptibility of LDL 10 oxidation.

Disease Treatment involved Inhibition of LDL oxidation Reference


(% reduction in TBARS)

Hypercholesterolemia Fluvastatin 47% 28


Atorvastatin hydroxy
metabolites 67% 30
Lovastatin 21% 26
Pravastatin 22% 27
Gemfibrozil metabolite 1 96% 30
Bezafibrate 40% 27
Cholestyramine 41% 27
Hypertension Captopril (ACE inhibitor) 32% 18
Enalapril (ACEinhibitor) 44% 18
Diabetes ~-Carotene 25%
Renal Failure Selenium 51% 20
Atherosclerotic
Pmice Losartan (angiotensin II receptor aotagonist) 55% 31
Captopril (ACE inhibitor) 25% 31
VitaminE 21% 43
Quercetin 54% 35
Glabridin 22% 37
151

to aggregation (induced by vortexing) by up to 23, 28 or susceptibility of their LDL to oxidation is also related to
40% respectively (Fig. I). Vitamin E supplementation to EO increased levels of arachidonic acid in their lipoprotein
mice resulted in a reduction (by 35%) in the LDL oxidation cholesteryl ester moiety [17). Similarly, LDL from subjects
state, and in parallel, LDL aggregation state was also reduced supplemented with fish oi!, which is rich in w-3 poly-
(by 23%). These reductions in LDL oxidation and aggregation unsaturated fatty acids, showed a two fold increased
states were accompanied by a 3 fold reduction in the aortic susccptibility to oxidation [45]. There was no significant
lesion area, in comparison to non-treated EO mice [44]. We alteration in the content oftheir LDL-associated antioxidant
conclude that in EO mice, LDL oxidation, which takes place vitaminsA, vitamin E and ß-carotene. However, although w-
already in plasma, leads to lipoprotein aggregation. These two 3 PUFA render the LDL more sensitive to oxidation, it also
modified forms ofLDL were shown to be taken up by macro- has beneficial effects on the production of some prostaglandins,
phages at enhanced rate, leading to foam cell formation. Thus, which oppose the oxidant characteristic.
the use of appropriate antioxidants can inhibit the formation The monounsaturated fatty acid oleic acid (C-18: 1), found
ofboth ofthese atherogenic forms ofLDL. Recently we have in abundance in olive oil, acts as an antioxidant. We tested
found that rnacrophages were able to induce LDL aggregation the effect of olive oil supplementation (50g/day) to the diet
and this phenomenon could be related to cellular activation of 10 healthy male subjects during a two weeks period, on the
and the release of proteoglycans. propensity oftheir LDL to oxidation. Olive oil supplementation
As macrophage-mediated oxidation ofLDL is considered to the diet modified the LDL lipid composition, and enriched
to be a key event during early atherogenesis, we have the lipoprotein with oleic acid, hydroxytyrosol, and sitosterol.
concentrated our studies on the elucidation of the mechanisms The olive oil induced-modified lipoprotein was found to be
involved in LDL oxidation by arterial wall macrophages. For more resistant to in vitro lipid peroxidation, and showed
this purpose we studied the effects of pro-oxidants and of reduced uptake by macrophages [32]. The LDL particle is
antioxidants in the LDL particle, as weH as in themacrophages. protected from oxidation by several antioxidants which are
associated to the lipoprotein (vitamin E, ß-carotene, lycopene,
ubiquinol, and severallipophylic polyphenols ), as weil as by
LDL-associated pro-oxidants and anti- antioxidants in the plasma environment (vitamin C, uric acid,
oxidants albumin and several hydrophilie polyphenols).
The carotenoids are lipid soluble antioxidants which are
LDL, the major cholesterol carrier in human plasma, contains located in the core ofthe LDL and their inhibitory effect on
in its cholesteryl ester mainly polyunsaturated fatty acids such LDL oxidation is controversial [46-49]. We have previously
as linoleic acid (C-18:2) and arachidonic acid (C-20:4) shown that both the all-trans and the 9-cis isomers of ß-
which are prone to oxidation. In hypercholesterolemic carotene can bind to plasma lipoproteins and affect LDL
patients we have previously shown that the enhanced oxidation, with the all-trans isomer ofß-carotene being more
effective than the 9-cis isomer in reducing the susceptibility
of the lipoproteins to lipid peroxidation, and in reducing the
LDL oxidation (by different oxidative mechanisms) cellular uptake of the oxidized LDL by macrophages [50].
enhances its susceptibility to aggregation Furthermore, in healthy subjects we have demonstrated an
inhibitory effect on the susceptibility of LDL to oxidative
c modification by both ofthese isomers ofß-carotene [51, 52].
o 0.8
LDl + NaDCl However, we found that not all subjects responded to ß-
+- LDl + AAPH
O~ LDl +Cu50 4 carotene supplementation by reduced ex vivo LDL oxidation.
CJ) ~ 0.6
Q)O lOl Comparison ofthe antioxidant status in 'responders' and 'non-
!....oo
OlU004
responders' LDLs, revealed that the vitamin E content in the
CJ) .
~
'responders LDLs' was significantly higher than that found
«C!
o in the 'non-responders LDLs'. We thus analyzed the effect
.....J '-' 0.2
o of carotenoids in combination with vitamin E, on the
.....J susceptibility of LDL to copper ions-induced oxidation. A
40 80 120 synergistic antioxidative effect against LDL oxidation was
obtained when a combination of the carotenoids together
Time of Vortexing (seconds) with vitamin E was used, instead of using the individual
antioxidant separately [53). Recently, we have also analyzed
Fig.1. The effect ofLDL (0. 1 mg ofproteinlml) oxidation by copper ions
the antioxidative capabilities of Iycopene against LDL
(5 !IM), by the free radieal generator AAPH (2 mM), and by hypochlorite, oxidation. We have demonstrated a protective effect of
on the susceptibility of the lipoprotein to aggregation. tomato's Iycopene against oxidative modification of LDL.
152

This protection ofLOL by lycopene exceeded the proteetion


exhibited by ß-carotene, was selective only to LOLs with ANTIOXIDATIVE EFTECT OF" UCORICE AND OF" GLABRIDIN

high vitamin E content, and was shown when the carotenoids AGAINST LDL OXIDATION IN APO-E DEF"ICIENT toIlCE

were present in combination with vitamin E, but not when the


carotenoids were supplemented alone [53]. Supplementation
ofvitamin E alone (25 Ilg1 mouse/day for 3 months) to the
P mice was recently found to be a potent antioxidant against
LOL oxidation and also reduced the atherosclerotic lesion
size by 35% [44].
Polyphenolic flavonoids are very potent antioxidants. In
olive oil, hydroxytyrosol was shown to contribute to its
inhibitory effect on LOLoxidation (in addition to a major effect
of the oleic acid). Other nutrient sources for polyphenols Contra I Ucorlc. Glabrldln
include licorice root ethanolic extract (rich with the isoflavan
glabridin), red wine (rich with the flavonol quercetin and the B. In vlvo
flavanol catechin), ginger, and orange peels [32-38]. The
Asian plant licorice is a source of polyphenols antioxidants,
with the isoflavan glabridin being the major one [54, 55].
Licorice root ethanolic extract was shown to inhibit human
LOL oxidation in a dose-dependent mann er. The mechanism
responsible for the antioxidative effects of licorice and
glabridin was shown to involve mainly their free radicals
scavenging capacity. LOL isolated from the plasma of ten
normolipidemic subjects who were supplemented for a
period of 2 weeks with 100 mg of licorice root extract per
day, was more resistant to copper ions-induced oxidation,
Fig. 2. The inhibitory effect of licorice root ethanolic extract and of its
as weil as to AAPH-induced oxidation, by 44 and 36%, major polyphenol glabridin, on LDL oxidation (induced by LDL incubation
respectively, in comparison to LOL isolated before licorice with 5 11M CuSO, far 4 h) in vitra (3 Ilg/m1 for 4 h) or in viva, in the EO
supplementation [38]. In EO mice, dietary supplementation mice (20 Ilg of glabridin equivalents/day /rnouse, for 6 weeks).
oflicorice (200 Ilg/day/mouse), or ofpure glabridin (20 Ilg1
day/mouse), for 6 weeks resulted in a 68% and 22% reduction
in the susceptibility of their LOL to copper ions-induced LOL. Thus, some phenolic substances that exist in red wine
oxidation, respectively (Fig. 2). This treatment also resulted are absorbed, bind to plasma LOL and thus, could be
in a significant reduction in the atherosclerotic lesion area responsible for the antioxidant properties ofred wine [33,
[38]. These results show that glabridin, the polyphenol with 34]. In EO mice that were supplemented with 50 f!g of
lipophilic characteristics which is present in licorice polyphenols/day/mouse for 6 weeks, plasma LOL isolated
ethanolic extract, is absorbed, binds to the LOL particle, and after red wine or quercetin, was less susceptible (30-80%)
subsequently protects the LOL from oxidation in multiple to oxidation induced by either copper ions, or by the free
modes of oxidative stress, as shown in humans and in the radical initiator AAPH, or by J-774 A.l macrophages in
EO mice. culture, in comparison to LOL isolated from the placebo -
The effect of consuming red wine (11 % alcohol) with treated EO mice group [36]. Cellular uptake of LOL from
meals, on the propensity of plasma and LOL to lipid EO mice that consumed catechin, quercetin or red wine was
peroxidation was studied in 17 healthy men. Red wine found to be reduced by 31, 40, and 52% respectively, in
consumption reduced the propensity of the volunteers LOL comparison to the cellular uptake of LOL derived from the
to copper ions-induced lipid peroxidation as determined by placebo group. In agreement with these results we found
a 46, 72 and 54% decrement in the content of the lipo- that the atherosclerotic lesion areas in EO mice that were
protein-associated aldehydes, lipid peroxides, and conjugated treated with red wine, quercetin or catechin were significantly
dienes, respectively, as weil as by a substantial prolongation reduced, by 40%,38% and 32%, in comparison to the lesion
ofthe lag phase required for the initiation ofLOL oxidation. areas in EO mice treated with placebo (Fig. 3). We thus
The antioxidant effect of dietary red wine on plasma lipid conclude that dietary consumption by EO mice, of red wine
peroxidation was not secondary to changes in plasma or of its polyphenolic flavonoids quercetin or catechin, leads
vitamin E or ß-carotene content, but could be related to the to reduced susceptibility oftheir LOL to oxidation and to
elevation in polyphenols concentration in plasma and in attenuation in the development of atherosclerosis.
153

RED WINE POLYPHENOLS CONSUMPTION BY EO MICE Macrophage binding ofLDL initiates the activation of cellular
REDUCES THE ATHEROSCLEROTIC LESIONS AREA oxygenases [56]. LDL oxidation by arterial wall cells, a key
event during early atherogenesis, was suggested to involve
12 the activation of macrophage 15-lipoxygenase and of
nicotinamide adenine dinucleotide phosphate (NADPH)
oxidase. We sought to analyze the role of these oxygenases
9
in macrophage-mediated oxidation ofLDL under oxidative
stress. Upon incubation of LDL with the J-774 A.1 macro-
phage-like cell line, or with human monocyte-derived
macrophages (HMDM), in the presence of 1 ~M CuS04 , the
release of superoxide anions to the medium was demonstrated
[57].
Under these conditions, the cytosolic protein components
Placebo Catechin Quercetin Red wine of the NADPH oxidase complex, P-47 and P-67, trans-
located to the plasma membrane, indicating LDL-mediated
Fig. 3. The effect of red wine derived polyphenols on the progression of activation of the NADPH oxidase complex. Under the
atherosclerosis in the atherosclerotic apolipoprotein E deficient mice, after above-mentioned experimental conditions, macrophage 15-
6 weeks of polyphenols (50 lIg polyphenols equivalents/day/mouse)
lipoxygenase was also activated, as determined by the
consumption.
release ofl5-hydroxy-5,8, 11, 13-eicosatetraenoic acid (15-
HETE) and 13-hydroxyoctadecadienoic acid (13-HÜDE) to
the medium. Inhibition ofthe macrophage NADPH oxidase
Macrophage-associated pro-oxidants with apocynin or dismutation of superoxide anions, the
and anti-oxidants products of NADPH oxidase activation, with superoxide
dismutase (SüD), significantly inhibited macrophage-
LDL oxidation is affected not only by the lipoprotein oxidative mediated oxidation ofLDL (by 61-89%, respectively) under
state, but macrophage-mediated oxidation of LDL is also these conditions [57]. Phorbol myristate acetate (PMA),
considerably affected by the oxidative state in the cells. This which causes NADPH oxidase activation in J-774 A.l
oxidative state depends on the balance between cellular macrophages, had no significant effect on 15-lipoxygenase
oxygenases and macrophage-associated antioxidants (Fig. 4). activity, but still resulted in cell-mediated oxidation ofLDL.

Fig. 4. The effect of the oxidative states of LDL and ofmacrophages on cell-mediated oxidation ofLDL. C- 18:2 -linoleic acid; Ox-FA - oxidized fatty
acids; CE - cholesteryl ester; UC - unesterified cholesterol; GSH - glutathione; SOD - superoxide dismutase; LPO -lipoxygenase.
154

Finally, HMDM from two patients with chronic granulomatous dependent on extracellular calcium ions. We conclude that
disease (CGD), that were shown to lack active NADPH phospholipases A2 and D can induce macrophage NADPH
oxidase, but to possess almost normal 15-lipoxygenase oxidase-dependent oxidation ofLDL, and thus can contribute
activity, failed to oxidize LDL (Fig. 5). We thus conclude that to the formation ofatherogenic oxidized lipoprotein I58].
LDL-induced NADPH oxidase activation (under oxidative Macrophage-mediated oxidation of LDL can also result
stress) is required for macrophage-mediated oxidation of from an initial peroxidation ofthe celllipids. When cultured
LDL, whereas activation of 15-lipoxygenase may not be macrophages were exposed to ferrous ions (50 IJM FeSOJ
sufficient for LDL oxidation under these conditions [57]. for 4 h at 37°C, cellular lipid peroxidation [measured by
On using J -774 A.I macrophages, we have demonstrated analyses of malondialdehyde (MDA), conjugated dienes
that phospholipase A2 as weil as phospholipase D, are involved (CD), and lipid peroxides (PD)], increased by 2--4 fold in
in macrophage NADPH oxidase-mediated oxidation ofLDL comparison with non-treated cells. Incubation of LDL (0.2
[58]. Furthermore, the products of these phospholipases, mg ofprotein/ml) with these oxidized macrophages resulted
arachidonic acid and phosphatidic acid respectively, can induce in LDL lipids peroxidation, as evidenced by an 8 fold increase
NADPH oxidase activation, followed by cell-mediated in LDL-associated MDA, in comparison with LDL that was
oxidation of LDL. This LDL oxidation was shown to be incubated under similar conditions with non-oxidized
macrophages [59]. Furthermore, oxidation ofLDL bymacro-
phages that were oxidized by incubation with deoxycholic
REDUCED CAPABIUTY OF' HMDM DERIVED F'ROM CGD PArIENTS acid (DCA) or angiotensin 11 (ANG-II), also induced
TO OXIDIZE lDL IN SPITE OF' THIER NORMAL
llPOXYGENASE ACTIVITY oxidative modification of macrophages, via metal ions-
A. LDL OXIDATION independent mechanisms [60-62]. Incubation ofLDL (200
...... IJg ofprotein/ml) for 24 h at 37°C with DCA, ANG-II or
....... .5
K!~ with FeS04-induced oxidized macrophages, resulted in a
iIt-....IA- substantial oxidative modification of the lipoprotein. The
'09 oxidative modification ofLDL by oxidized macrophages was
E 111
..sE found to be a progressive process. Incubation of LDL with
oxidized macrophages for increasing periods oftime (up to
24 h) resulted in a progressive increment in the electrophoretic
mobility ofLDL, the MDA formation in LDL, and the cellular
B.NADPH-Oxldase actlvlty uptake ofLDL by the oxidized macrophages via the Ox-LDL
receptor. The increased uptake of LDL by oxidized macro-
Oll
o .....
0::
phages thus results from two routes: (I) enhanced uptake via
the LDL receptor due to increased LDL receptor activity in
" E
Icf' ~
oxidized macrophages, and (2) enhanced lipoprotein uptake
Ci ~
=
.sE 11
via the Ox-LDL receptor [62].
As macrophage antioxidants mayaiso contribute to the
extent of cell-mediated oxidation of LDL, and since the
• • glutathione is an important cellular antioxidant system [63-
65], we analyzed the role of cellular reduced glutathione
C.15-LPO actlvlty
(GSH) content, and of glutathione peroxidase (GPx) activity
in this process. Upon incubation of J-774 A.I macrophages
for 20 h at 37°C with 50 IJM of buthionine sulfoximine
(BSO), an inhibitor of glutathione synthesis, cellular GSH
content and GPx activity were reduced by 89 and 50%
respectively [39], and this effect was associated with a 2 fold
elevation in macrophage-mediated oxidation ofLDL (Fig. 6).
The BSO-treated cells contained high levels ofperoxides,
o and released 39% more superoxide anions than non-treated
Control Patient 1 Patient 2
cells in response to their stimulation with phorbol myristate
acetate. In order to increase macrophage GSH content and
Fig.5. The involvement ofmonocyte-macrophage NADPH oxidase, but GPx activity we have used L-2-oxothiazolidine-4-carb-
not of 15-lipoxygenase (LPO) in eell-mediated oxidation ofLDL: Study of
oxylic acid (OTC) which delivers cysteine residues to the
HMDM from patients with chronie granulomatous disease (CGD). 0,--
superoxide anions;TBARS --thiobarbiturie acid reaetive substances; 15- cells for GSH synthesis, and also selenium which activates
HETE - 15-hydroxy -5,8,11,13 eieosateraenoie acid. GPx and increases cellular glutathione synthesis. GSH
155

REDUCED MACROPHAGE GLUTATHIONE CONTENT AND oxidation of LDL [39]. Intervention means to enhance the
INCREASED CELL-MEDIATED OXIDATION OF LDL macrophage GSH-GPx status may thus contribute to the
CD...... A. attenuation ofthe atherosclerotic process. The relationship
I: I:
00; 4 between the glutathione system and plasma lipid peroxida-
......
:Co
o
tion was also analyzed in six renal transplanted patients
....::Jo;-0- (which are under oxidative stress, and thus, at high risk for
c:::; " atherosclerosis), by using dietary selenium in order to
~ E 2 activate the glutathione system [18]. AAPH-induced plasma
~~ lipid peroxidation was increased (by 60%) in all six patients,
Gi E
('),5 in comparison to normal subjects. Dietary selenium supple-

9
mentation (0.2 mg/day for aperiod of 3 months) resulted in a
o1-----~------~--Ji~~It------l 50% reduction in AAPH-induced plasma lipid peroxidation.
't- 'C B. The susceptibility of the patients plasma to lipid peroxidation
00;

5 e 20
._ 0-
retumed toward baseline values 3 months after termination of
oB
1J...J
the selenium treatment. Analyses ofthe patients red blood cell
(RBC) glutathione system revealed low levels of reduced
XCI
o E glutathione (GSH) and decreased activity ofRBC glutathione
-g ':c- 10
.... 0
peroxidase by 23 and 20% respectively in comparison to
0::' normal RBC. Selenium treatment resulted in a significant
'C ö
E~
I~OL--------'--------'---
elevation in red blood cells GPx activity and GSH content by
64 and 11 % respectively [20].
CD Control +850 We next questioned whether macrophage enrichment with
(,)
nutritional antioxidants such as ß carotene, lycopene, vitamin
E, or polyphenolic flavonoids can affect their ability to oxidize
Fig. 6. The effect of macrophage reduced glutathione (GSH) content on
LDL. We investigated the effect of dietary supplementation of
cell-mediated oxidation of LDL. J-774A.1 macrophages were incubated
without (Contrai), or with butionine sulfoximine (50 11M BSO) for 20 h at ß-carotene on plasma lipid peroxidation (induced by AAPH)
37°C, followed by cell incubation with LDL, for a further 5 h, prior to and on cell-mediated oxidation ofLDL by human monocyte
analysis of LDL oxidation derived macrophages (HMDM). Significant enrichment with
ß-carotene was noted in plasma (2 fold), in LDL (2.6 fold) and
in HMDM (1.6 fold) two weeks after dietary supplementation
content and GPx activity in J-774 A.I macrophages were with 180 mg/day of ß carotene. In these subjects, plasma
increased by 80 and 50% respectively, following cells lipids peroxidation decreased by 22% and LDL oxidation
incubation with 2 mM of OTC for 20 h at 37°C, and this was decreased by 40% in AAPH- induced oxidation system [52].
paralleled by a 47% inhibition in LDL oxidation by these cells After this ß-carotene supplementation, LOL oxidation by
[39]. An inverse correlation was found between the extent macrophages was found to be reduced as a result of LDL
of macrophage-mediated oxidation of LDL and cellular enrichment with ß-carotene but was not affected by enrich-
GSH content (r = 0.97), or GPx activity (r = 0.95). Upon ment of the cells with ß-carotene. Thus, we suggest that ß-
incubation of J-774 A.I macrophages with selenomethionine carotene content ofLDL, but not that ofthe macrophages, is
(10 ng/ml) for one week, cellular GSH content and GPx responsible for the inhibition of LDL oxidation by the cells.
activity were increased by about 2 fold as compared to Similarly, enrichment of mouse peritoneal macrophages with
control cells, and this effect was associated with a 30% Iycopene or with ß-carotene did not affect cell-mediated
reduction in cell-mediated oxidation of LDL. Dietary oxidation of LDL (Fig. 7A). On the contrary, upon macro-
selenium supplementation (illg/day/mouse) to the athero- phage enrichment with vitamin E, cell-mediated oxidation of
sclerotic apolipoprotein E deficient mice for a 6 months LOL was significantly inhibited. Following 18 h ofmacro-
period, increased GSH content and GPx activity in the mice phage incubation with 75 IlM of vitamin E, macrophage-
peritoneal macrophages by 36 and 30% respectively, and mediated oxidation of LDL was reduccd by 59% (Fig. 7A).
this effect was associated with a 46% reduction in cell- Similarly, the polyphenols glabridin, catechin or quercetin
mediated oxidation of LDL. Finally, the atherosclerotic accumulated in macrophages upon cell incubation with these
lesion area in the aortas derived from these mice after purified polyphenols. Upon incubation ofthese antioxidants-
selenium supplementation was found to be reduced by 30% enriched macrophages with native LDL under oxidative
as compared to the lesion area found in non-treated mice. stress (IIlM Cu SO4)' cells enrichment with these antioxidants
Our results thus demonstrate an inverse relationship between resulted in up to 70, 45 and 90% inhibition in cell-mediated
macrophage GSH content! GPx activity and cell-mediated oxidation of LDL (Fig. 7B).
156

ErrECT or MACROPHAGE ENRICHMENT WITH ANTIOXIDANTS elucidated [66]. It was suggested that human apolipoprotein
ON CELL-MEDIATED OXIDATION or LDL A-I possesses antioxidant properties which might affect LDL
A. lipid peroxidation. Plasma HDL from both normal mice and
from the human apolipoprotein A-I transgenic mice at similar
concentrations, inhibits LDL lipid peroxidation, but the effect
-' ofthe HDL from the human apolipoprotein A-I transgenic
c mice was two fold greater than that of HDL derived from
-'
'0:5111 control mice [67]. An additional possibility for the inhibitory
~
+=
e
c..
effect of HDL on LDL oxidation is the presence of the
enzyme paraoxonase (PON) in plasma HDL [66, 68-73].
0-'
"09 Serum PON activity was shown to be lower in atherosclerotic
'x CI 0 -'--_.JLU
patients such as patients after myocardial infarction [68],
o E Control Vitamin E Lycopen~ p-Carotene
patients with familial hypercholesterolemia [74] or patients
"0'
_ID c-<
with diabetes mellitus [73-75]. Although the physiol
o ::I
:e-
ID 0
pathological role of serum PON is not known yet, evidence
:::I! E B. for a protective effect ofPON against oxidative damage was
15 shown [68--75]. As PON was suggested to be involed in
6
Gi atherogenesis via an inhibitory effect on lipoprotein oxidation,
u
we have used the apolipoprotein E deficient mice (which
develop accelerated atherosc1erosis) to analyze the association
among their atherosc1erosis lesion size, their serum lipid
peroxidation and their serum paraoxonase activity. Whereas
both serum lipids peroxidation and the extent ofthe mice lesion
area increased with age, serum PON activity significantly
Control Cot.chln Ouercetln Globrldln decreased [36]. However PON's activities were preserve in
serum derived from EO mice after consumption of polyhenols
Fig. 7. The effect ofmacrophage enrichment with nutritional antioxidants
for 2 weeks as 14, 113, and 75% higher PON activities were
on cell-mediated oxidation ofLDL. J-774A.l macrophages were incubated obtained after the consumption of catechin quercetin or red
for 20 h with the pure antioxidants at a concentration of 20 ~M for the wine, respectively, in comparison to PON activity in serum
polyphenols, 75 ~M for vitamin E and 50 ~M for the carotenoids. LDL from control mice [36]. We have recently analyzed the effect
(0.1 mg ofproteinlml) was then added and furtherincubated (in the presence ofHDL - associatedPON, as well as ofpurifiedPON, onHDL
of 1 ~M CuSO,) for 6 h at 37°C, prior to analyses of its oxidation.
and on LDL oxidation, by using specific PON inhibitors. A
significant inhibitory effect of PON on the oxidation of the
lipoproteins in several oxidative systems was noted. Titration
Possible mecbanisms for tbe removal of of human HDL with purified PON inhibited copper ions-
Ox-LDL induced HDL oxidation in a concentration dependent manner.
PON addition prolonged the oxidation lag phase and reduced
The atherogenicity ofOx-LDL involves its stimulatory effect the HDL peroxide and aldehyde formation by up to 95% [76].
on macrophage cholesterol accumulation, as well as its In contrast, in the presence ofPON specific inhibitors, HDL
atherogenic effects on blood cells and on arterial wall cells. oxidation induced either by copper ions or by AAPH was
Thus, removal of Ox-LDL from plasma or from the extra- markedly enhanced. Furthermore, both HDL-associated
cellular space by cells ofthe arterial wall may be beneficial PON, as weil as purified PON, were potent inhibitors ofLDL
as long as it does not cause massive cellular cholesterol oxidation. Finaly, PON was found to directly act upon and
accumulation and foam cell formation. Other mechanisms hydrolyze lipoprotein- associated cholesteryl linoleate
which can contribute to the elimination of atherogenic Ox- hydroperoxides in oxidized LDL and HDL, and to hydrolyze
LDL from entering the arterial wall cells are related to the hydrogene peroxides. We have recently demonstrated a PON-
inhibition of cell-mediated oxidation ofLDL by affecting the mediated hydrolysis of cholesteryllinoleate hydroperoxides
balance among LDL-associated and cell-associated pro-and (by up to 60%) in oxidized-LDL, and this effect could be
anti-oxidants. In addition, mechanisms for the elimination of reversed by pretreatment ofPON with the PON inhibitor PD
Ox-LDL from the extracellular space may inc1ude hydrolysis - 65950 (Fig. 8). These results suggest that paraoxonase may
of lipoprotein-associated peroxides. have an important role in the removal and elimination of
Plasma HDL was previously shown to inhibit LDL atherogenic oxidized lipoproteins and rnay thus be considered
oxidation, but the mechanism for this effect was not yet anti-atherogenic.
157

PARAOXONASE CAN HYDROLYZE CHOLESTERYL 11. Kaplan M, Williams JK, Mandel H, Aviram M: Role of macrophage
LlNOLEATE HYDROPEROXIDES IN OX-LDL glyeosaminoglyeans in the cellular catabolism of oxidized LDL by
macrophages. Arterioscler Thromb Vase Bio118: 542-553,1998
12. Kaplan M,Aviram M: Oxidized LDL binding to a maerophage-seereted
extracellular matrix. Bioehern Biophys Res Commun 237: 271-276,
1997
13. Aviram M, Maor I, Kcidar S, Hayek T, Oiknine J, Bar-EI Y, Adler Z,
Kertzman V, Milo S: Lesioned low density lipoprotein in athero-


sclerotic apolipoprotein E defieient trans genie mice and in humans is
oxidized and aggregatcd. Bioehern Biophys Res Commun 216: 501-
513,1995
14. HayekT, Oiknine J, Brook JG,Aviram M: Increased plasma lipoprotein
lipid peroxidation in apo E-defieient mice. Bioehern Biophys Res
Commun 201: 1567-1574, 1994
0-'--------'----'-- 15. Yla-Herttuala S, Palinski W, Rosenfeld ME, Pharthasarathy S, Carew
Control +PON +PON TE, Buttler S, Witzturn JL, Steinberg 0: Evidence for the presence of
+PD-65950 oxidatively modified low density lipoprotein in atherosclerotic lesions
ofrabbit and man. J CI in Invest 84: 10867-1095, 1989
16. Haberland ME, Fogelman AM: The role ofaltered lipoprotein in the
Fig. 8. Decomposition of cholesteryllinoleate hydroperoxides in Ox-LDL pathogenesis of atheroselerosis. Am Heart J 113: 573-517, 1987
by paraoxonase. Ox-LDL (1 mg/ml) was incubated without (control) or 17. Lavy A, Brook JG, Dankner G, Ben Amotz A, Aviram M: Enhanced
with 10 U/ml ofpurified paraoxonase (+PON). or PON that was pretreated in vitra oxidation of plasma lipoproteins derived from hyper-
with 100 11M ofthe PON inhibitor PD-65950 (+PON+PD-65950) for 3 h cholesterolemic patients. Metabolism 40: 794-799,1991
at 37°C. At the end of the incubation cholesteryllinileate hydroperoxides 18. Keidar S, Kaplan M, Shapira H, Brook JG, Aviram M: Low density
were determined by HPLC. Results expressed as mean ± S.O. (n.=.3). • p lipoprotein isolated from patients with essential hypertension exhibits
< 0.01 vs. Contro!. increased propensity for oxidation and enhanced uptake by maero-
phages: A possible role for angiotensin 11. Atheroselerosis 104: 71-
84,1994

References 19. Bergman R, Kasif Y, Aviram M, Maor I, Ullman Y, Gdal-on M,


Friedman-Birnbaum R: Normolipidemie xanthelasma palpebrarum:
Lipid composition, cholesterol metabolism in monocyte-derived
I. Aviram M: Oxidized low density lipoprotein (Ox-LDL) interaction maerophages, and plasma lipid peroxidation. Acta Oerm Venereol 76:
with macrophages in atherosclerosis and the antiatherogenieity of 107-110, 1996
antioxidants. Eur J Clin Chem Clin Bioehern 34: 599-608, 1996 20. Hussein 0, Rosenblat M, Refael G, Aviram M: Oietary selenium
2. Aviram M: Oxidative modifieation of low density lipoprotein and inereases eellular glutahtione peroxidase activity and reduces the
atheroselerosis.lsr J Med Sei 31: 241-249,1995 enbaneed suseeptibility to lipid peroxidation of plasma and low density
3. Aviram M: LDL-Platelet interaction under oxidative stress induees lipoprotein in kidney transplanted recipients. Transplantation 63: 679-
maerophage foam eell formation. Thromb Haemost 74(1): 560-564, 685, 1997
1995 21. Pipek R, Dankner G, Ben-Amotz A, Aviram M, Levy Y: Inereased
4. Steinberg 0, Parthasarathy S, Carew TE, Khaa JC, Witzturn JL: plasma oxidizability in subjects with severe obesity. J Nutr Environ
Beyond cholesterol: Modifications of low-dcnsity lipoprotein that Med 6: 267-272, 1996
inerease its atherogenicity. N Engl J Med 320: 915-924,1989 22. Lavy A, Ben-Amotz A, Aviram M: Inereased susceptibility to lipid
5. Berliner JA, Navab M, Fogelman AM, Frank JS, Demer LL, Edwards peroxidation of ehylomicrons and low density lipoprotein in eeliae
PA, et al.: Atherosclerosis: Basic mechanisms. Oxidation, inflammation disease. Ann Nutr Metabol 37: 68-74, 1993
and genctics. Cireulation 91: 2488-2498,1995 23. Maggi E, Bellazzi R, GazoA, Secciam T, Bellomo G: Autoantibodies
6. Aviram M: The contribution ofthe macrophage reeeptor for oxidized against oxidatively-modified LDL in chronic patients undergoing
LDL to its cellular uptake. Bioehern Biophys Res Commun 179: 359- dialysis. Kidney Intern 46: 869-876, 1994
365,1991 24. Nishigaki I, Hagihara M, Tsunekana H, Maseki M, Yagi K: Lipid
7. Keidar S, Brook JG, Rasenblat M, Fuhrman B, Dankner G, Aviram peroxide levels of serum lipoprotein fractions of diabetic patients.
M: Involvement ofthe maerophage LDL reeeptor binding domains in Bioehern Med 25: 373-378, 1981
the uptake of oxidized LDL. ArterioseIer Thromb Vase Biol 12(4): 25. Aviram M, Keidar S, Brook JG: Dual effeet of lovastatin and
484-493, 1992 simvastatin on LDL-maerophage interaction. Europ J Clin Chem Clin
8. Eisenberg S, Shayek E, Oliveerona T, Vlodavsky I: Lipoprotein lipase Bioehern 29(10): 657-664, 1991
enhances binding of lipoproteins to heparan sulfate on eell surfaces 26. Aviram M, Dankner G, Cogan U, HochgrafE, Brook JG: Lovastatin
and extracellular matrix. J Clin Invest 90: 2013-2021, 1992 inhibits LDL oxidation and alters its fluidity and uptake by m.cro-
9. Williams KJ, Fless GM, Petrie KA, Snyder ML, Broeia RW, Swenson phages:In vitra andin vivo studies. Metabolism41(3): 229-235,1992
TL: Mechanisms by which lipoprotein lipase alters cellular metabolism 27. Hoffman R, Brook JG, Aviram M: Hypolipidemic therapy reduces
of LP(a) LDL and nascent Iipoporoteins. J Biol Chem 267: 13284- lipoprotein susceptibility to undergo lipid peroxidation: In vitra and
13292, 1992 ex viva studies. Atheroselerosis 93: 105-113, 1992
10. AuerbachBJ, Bisgaier CL, Wolle J, Saxena U: Oxidation oflow density 28. Hussein 0, Schlezinger S, Rosenblat M, Keidar S,Aviram M: Reduced
lipoproteins greatly enhances their assoeiation with lipoprotein lipase suseeptibility of LDL to lipid peroxid'lion after fluvastatin therapy is
anehored to endothelial eell matrix. J Biol Chem 271: 1329-1335, assoeiated with the hypocholesterolemic effect of thc drug and its
1996 binding to the LDL. Atherosclerosis 128: 11-18, 1997
158

29. Aviram M, Hussein 0, Rosenblat M, Schlezinger S, Hayek I, Keidar 47. Hennekens CH, Buring JE, Manson JE, Stampfer M, Rosner B, Cook
S: Interactions of platelets, macrophages and lipoproteins in hyper- NR, Belanger C, LaMotte F, Gaziano JM, Ridker PM, Willett W, Peto
cholesterolemia: Antiatherogenic effects of HMG-CoA reductase R: Lack of effect of long-term supplementation with beta carotene on
inhibitor therapy. J Cardiovasc Pharmacol31: 39--45, 1998 the incidence of malignant neoplasms and eardiovascular disease. N
30. Aviram M, Rosenblat M, Bisgaier CL, Newton RS: Atorvastatin and Engl J Med 334: 1145-1149, 1996
gemfibrozil metabolises, but not the parent drugs are potent anti- 48. Jiala1 I, Norkus EP, Cristor L, Grundy SM: ß-Carotene inhibits the
oxidants against lipoproteins oxidation. Atherosc1erosis (in press) 1998 oxidative modification of low-density lipoprotein. Biochim Biophys
31. Keidar S, Attias J, Smith J, Breslow JL, Hayek T: The angiotensin-II Acta 134-138, 1991
receptor antagonist, losartan, inhibits LDL lipid peroxidation and 49. Sies W, Stahl W, SundquistAR: Antioxidant functions ofvitamins; vit-
atherosclerosis in apo lipoprotein E-deficient mice. Biochem Biophys amin E and C, Beta-carotene, and other carotenoids. In: H.E. Savberlich,
Res Commun 236(3): 622--D25, 1997 L.V Machlin (eds). Beyond defieiency: New views on the function and
32. Aviram M, Kasem E: Dietary olive oil reduces the susceptibility of health effects ofvitamins. Ann NY Acad Sci 669: 7-20, 1992
10w density lipoprotein to lipid peroxidation and inhibits lipoprotein 50. Lavy A, Ben Amotz A, Aviram M: Preferential inhibition of LDL
uptake by macrophages. Ann Nutr Metabol 37: 75-84, 1993 oxidation by the all-trans isomer ofß-carotene in comparison to the
33. Fuhrman B, Lavy A, Aviram M: Consumption of red wine with meals 9-cis-carotene. Eur J Clin Chern Clin Biochem 3 \: 83-90, 1993
reduces the susceptibility ofhuman plasma and LDL to undergo lipid 51. Levy Y, Ben-Amotz A, Aviram M: Effect of dietary supplementation
peroxidation. Am J Clin Nutr 61: 549-554,1995 of ß-Carotene to humans on its binding to plasma LDL and on the
34. Fuhrman B, Aviram M: White wine reduces LDL susceptibility to lipoprotein susceptibility to undergo oxidative modification: Com-
oxidation in vitra but not in viva. Am J Clin Nutr 63: 403-404, 1996 parison of the synthetic all trans isomer with the natural algae ß-
35. Lavy A, FuhrmanB, Marke! A, Dankner G, BenAmotz A, Presser D carotene. J Nutr Env Med 5: 13-22, 1995
Aviram M: Ihe effect of dietary supplementation of red or white wine 52. Levy Y, Kaplan M, Ben Amotz A, Aviram M: The effeet of dietary
on human blood chemistry, hematology and coagulation pattern; supplementation of ß-carotene on human monocyte-macrophage-
favorable effect of red wine on plasma high density lipoprotein. Ann mediated oxidation of low density lipoprotein. Isr J Med Sci 32(6):
Nutr Metabol 38: 287-294, 1994 473-478, 1996
36. HayekI, Fuhrman B, Via J, Rosenblat M, Belinky P, Coleman R, Elis 53. Fuhrman B, Ben-Yaish L, Attias J, Hayek T, Aviram M: Tomato's
A, Aviram M: Reduced progression of atherosc1erosis in the apolipo- Iycopene and ß-carotene inhibit 10w density lipoprotein oxidation and
protein E defieient mice following consumption of red wine, or its this effect depends on the lipoprotein vitamin E conten!. Nutr Metab
polyphenols quercetin, or catechin, is associated with redueed Cardiovasc Dis 7: 433-443,1997
susceptibility of LDL to oxidation and to aggregation. Arteriosc1er 54. Chandler RF: Licarice, more thanjust a flavour. Can Pharm J 118:
Thromb Vasc Biol 17: 2744-2752, 1997 420-424, 1985
37. Vaya J, Belinky P, Aviram M: Antioxidant constituents from licoriee 55. Demizu S, Kajiyama K, Takahashi K, Hiraga Y, Yamamoto S, Tamura
roots: Isolation, structure elucidation and antioxidative capacity Y, Okada K, Kinoshita T: Antioxidant and antimierobial constituents
towards LDL oxidation. Free Radic Biol Med 23: 302-313, 1997 of Licoriee: Isolation and structure elucidation of a new benzofuran
38. Fuhrman B, Buch S, Vaya J, BelinkyPA, Coleman R, Hayek T, Aviram derivative. Chem Pharm BuB 36: 3474-3479, 1988
M: Licorice extract and its major polyphenol glabridin protect LDL 56. Aviram M, Rosenblat M: Macrophage-mediated oxidation of extra-
against lipid peroxidation: In vitra and ex-vivo studies in humans and ceBular low density lipoprotein requires an initial binding of the
in the atherosclerotic apo lipoprotein E deficient mice. Am J Clin Nutr lipoprotein to its receptor. J Lipid Res 35: 385--398, 1994
66: 267-275, 1997 57. Aviram M, Rosenblat M, Etzioni A, Levy R: Activation of NADPH
39. Rosenb1at M, Aviram M: Macrophage glutathione content and oxidase is required for maerophage-mediated oxidation oflow density
glutathione peroxidase activity are inversely related to cell-mediated lipoprotein. Metabolism45(9): 1069-1079, 1996
oxidation ofLDL. Free Radic Biol Med 24: 305-317, 1997 58. Aviram M, Rosenblat M: Phospholipase A 2 and phospholipase D are
40. Oiknine J, Aviram M: Increased susceptibility to activation and involved in macrophage NADPH oxidase-mediated oxidation ofLDL.
increased uptake of 10w density lipoprotein by cholesterol-10aded Isr J Med Sei 32: 749-756, 1996
macrophages. Arteriosc1er Thromb Vasc Bio112: 745--753, 1992 59. Fuhrman B, Oiknine J, Aviram M: !ron induces lipid peroxidation in
41. Carter D, Fuhrman B,Aviram M: Macrophage activation with phorbo1 cultured maerophages inereases their ability to oxidative1y modifY LDL
myristate acetate is associated with cellular lipid peroxidation. Isr J and affect their secretory properties. Atherosclerosis 111: 65--78, 1994
Med Sci 32(6): 479--4851996 60. Keidar S, Kaplan M, Hoffman A, Brook JG, Aviram M: Angiotensin
42. Hoff HF, O'Neii J: Lesion-derived low density lipoprotein and oxidized II stimulates macrophage-mediated lipid peroxidation oflow density
low density lipoprotein share a lability for aggregation, leading to lipoprotein. Atherosc1erosis \15: 201-215,1995
enbanced macrophage degradation. Arteriosc1er Ihromb 11: 1209- 61. Ljubuncic P, Fuhrman B, Oiknine J,Aviram M, BomzonA: The effect
1222,1991 of deoxycholic acid and ursodeoxycholic acid on lipid peroxidation
43. Meyer DF, Mayans MO, Groot PH, Suckling KE, Bruckdorfer KR, in cu1tured macrophages. Gut 39: 475-478, 1996
Perkins SJ: Time-course studies by neutron solution scattering and 62. Fuhrman B, Oiknine J, Keidar S, Ben-Yaish L, Kaplan M, Aviram M:
biochemical assays of the aggregation of human low-density lipo- Increased uptake of LDL by oxidized macrophages is the result of an
protein during Cu2+-indueed oxidation. Biochem J 310: 417-426, 1995 initial enhanced LDL receptor activity and of further progressive LDL
44. Maar I, HayekT, Coleman R,Aviram M: Plasma LDL oxidation leads oxidation. Free Radic Biol Med 23: 34-46, 1997
to its aggregation in the atherosc1erotic apo lipoprotein E deficient mice. 63. Baker RD, Baker SS, LaRosa K: Selenium regulation of glutathione
Arteriosc1er Thromb Vasc Biol 17: 2995--3005, 1997 peroxidase in human hepatoma cell line Hep 3B. Arch Biochem
45. Levy Y, Ben Amotz A, Dankner G, Brook JG, Aviram M: Enhanced Biophys 304: 53-57, 1993
lipid peroxidation of low density lipoprotein by fish oil. J Optimo1 64. Kuzuya M, Natio M, Funakui C, Hayashi T, Asai K, Kuzuya F:
Nutr 2: 6-9, 1993 Protective role of intracellular glutathione against oxidized low density
46. Krinsky NI: Antioxidant functions of carotenoids. Free Radic Biol lipoprotein in cultured endothelial eells. Biochem Biophys Res
Med 7: 617--D35, 1989 Cornrnun 163(3): 1466-1472, 1989
159

65. Gotoh N, Graham A, Niki E, Darly-Usmar UM; Inhibition of glutahtione Fogelman AM: The Yin and Yang of oxidation in the development
synthesis increases the toxicity of oxidized low-density lipoprotein to of the fatty streak. Arterioscler Thromb Vase Biol 16: 831-842,
human monocytes end macrophages. Biochem J 296: 151-154, 1994 1996
66. Maekness MI, Abbot! C, Arrol S, Durrington PN: The role of high 72. La Du BN: Struetural and funetional diversity of paraoxonases. Nature
density lipoprotein and lipid-soluble antioxidant vitamins in inhibiting Medieine 2(11): 1186-1187,1996
low-density lipoprotein oxidation. Bioehem J 294: 829--834, 1993 73. Mackness MI, Maekness B, Durrington PN, Connelly PW, Hegele
67. Hayek T, Oiknine J, Dankner G, Brook JG, Aviram M: HDL apolipo- RA: Paraoxonase: Bioehemistry, geneties and relationship to plasma
protein A-I altenuates oxidative modifieation of low density lipo- lipoproteins. Curr Opin Lipidol 7: 69--76, 1996
protein: Studies in transgenie mice. Eur J Clin Chem Clin Bioehem 74. Maekness MI, Harty D, Bhatangar D, Winoeour PH, Arrol S, Ishola
33: 721-725, 1995 M, Durrington PN: Serum Paraoxonase aetivity in familial hyper-
68. Maekness MI, Durrington PN: HDL, its enzymes and its potential to eholesterolemia and insulin-dependent diabetes mellitus. Atherosclerosis
influenee lipid peroxidation. Atheroselerosis 115: 243-253, 1995 86: 193-199, 1991
69. Maekness MI, Arrol S, Abbot! C, Durrington PN: Proteetion of low- 75. Abbott CA, Maekness MI, Kurnar S, Boulton AJ, Durrington PN:
density lipoprotein against oxidative modifieation by high-density Serum paraoxonase activity, eoneentration, and phenotype distribu-
lipoprotein assoeiated paraoxonase. Atherosclerosis 104: 129--135, 1993 tion in diabetes mellitus and its relationship to serum lipids and
70. Maekness MI, Arrol S: Durrington PN Paraoxonase prevents lipoproteins. Arterioscler Thromb Vase Biol11: 1812-1818, 1995
aeeurnnlation oflipoperoxides in low density lipoprotein. FEBS Lett 76. Aviram M, Rosenblat M, Bisgaier CL, Newton RS, Primo-Parmo SL,
286: 152-154, 1991 La Du BN: Paraoxonase inhibits high density lipoprotein (HOL)
71. Navab M, Berliner JA, Watson AD, Hama SY, Teritto MC, Lusis oxidation and preserves its funetions: A possible peroxidative role for
AJ, Shih DM, Van Lenten DJ, Frank JS, Demer LL, Edwards PA, paraoxonase. J Clin Invest 101: 1581-1590, 1998
Malecular and Cellular Biachemistry 188: 161-166, 1998.
© 1998 Kluwer Academic Publishers.

Modulation of adriamycin-induced changes in


serum free fatty acids, albumin and cardiac
oxidative stress
Natasha Iliskovic, Timao Li, Neelam Khaper, Vince Palace and Pawan
K. Singal
Institute of Cardiovascular Sciences, St. Boniface General Hospital Research Centre and Department of Physiology,
Faculty ofMedicine, University of Manitoba, Winnipeg, Canada

Abstract
Adriamyein-indueed eardiomyopathic ehanges are prevented by eombination therapy with probucol. These beneficial effects
are suggested to be due to a combination of antioxidant as weil as lipid-lowering effects of probucol. In the present study, we
compared the effeets ofprobucol (PROB) with that oflovastatin (LOV), a lipid-Iowering drug, and trolox (TRO), an antioxidant,
on adriamycin (ADR)-indueed subchronic in vivo changes in serum free fatty acids (FFA), serum albumin and myocardial redueed
(GSH) and oxidized (GSSG) glutathione in rats.ADR caused a signifieant increase in FFA, decrease in albumin, and an increase
in FFAialbumin. PROB and LOV modulated the increases in FFA and FFAialbumin, while TRO was without any effect. ADR
reduced myocardial GSH, increased GSSG and decreased GSH/GSSG. Only PROB caused significant improvement in GSH
and normalized GSSG levels. It is suggested that these modulatory effects ofprobucol mayaIso contribute in the beneficial
effects ofthis drug against adriamycin-induced cardiomyopathy and congestive heart failure. (Mol Cell Biochem 188: 161-
166, 1998)

Key words: probueol, lovastatin, trolox, glutathione

Introduction and loss of myofibrils [2, 3]. The cause of adriamycin-


induced cardiomyopathy is probably multifaetorial and
Adriamycin, also known as doxorubicin, is a potent anti- complex, but free oxygen radicals and lipid peroxidation
tumor antibiotic used for the treatment of a variety of soft and appear to play an important role [4-9].Aceordingly, different
solid human malignancies. However, treatment may be free oxygen radical scavengers and antioxidants have been
complicated by its acute and chronic side-effeets. One major used to prevent or mitigate these adverse effects [3, 7-9].
chronic side-effect is the development of cardiomyopathy and Probucol, an antioxidant as weil as a lipid-Iowering drug,
ultimately congestive heart failure [1-3]. Several different has been reported to completely prevent the adriamycin-
mechanisms have been suggested to explain the development induced cardiomyopathy in rats [3, 9]. Protection was seen
of adriamycin-induced cardiomyopathy [3, 4]. Cardio- with respect to mortality, ultrastruetural changes, hemo-
myopathie changes in the heart, found after chronic treatment dynamic ftmetion and oxidative stress [3, 9]. As adriamycin
with adriamycin, include dilation ofthe heart, focal degenera- depressed myocardial antioxidants and probucol eountered
tion, atrophy of myocytes and fibrosis [2, 3]. Typical this effect, the observed protection was suggested to be due to
morphological changes of eardiac eells include cytoplasmic the enhancement of endogenous antioxidants [8]. Since
vacuolization due to distention ofthe sarcoplasmic reticulum adriamyein is known to increase serum free fatty acids [10, 11]

Address Jar offprints: P .K. Singal, Institute of Cardiovascular Sciences, St. Boniface General Hospital Research Centre, 351 rache Avenue, Room R3022,
Winnipeg, Manitoba, R2H 2A6 Canada
162

and since free fatty acids and reduced albumin are known to and ventricles were weighed and used to study glutathione
adversely effect the cardiac function [12], adriamycin- (GSH and GSSG) levels.
induced changes in serum free fatty acids and albumin wcrc
studied. Myocardiallevels of reduced and oxidized glutathione
were also examined. The effects ofprobucol were compared Free fatty acids and serum albumin assays
with another lipid-lowering drug, lovastatin, as weH as with
another antioxidant, trolox. Serum free fatty acids (FFA) were determined by standard
kit from WAKO (NEFA C), and serum albumin by standard
Sigma kit for albumin determination (#625-2).
Materials and methods
Animal model and treatment protocols Glutathione assays

Male Sprague-Dawley rats, body weight 250 ± 25 g, were Concentrations of total glutathione i.e., oxidized (GSSG) +
maintained on a normal rat chow and a regular light and dark reduced (GSH), were measured in the myocardium by the
cycle. Animals were given water and food ad libitum and glutathione reductase/5,5' -dithiobis-(2-nitrobenzoic acid)
divided into eight groups: CONT (control), ADR (adria- recycling assay [13]. The rate of dithiobis-nitrobenzoic acid
mycin-treated), PROB (probucol-treated), PROB + ADR formation was recorded at 412 nm which was proportional
(probucol + adriamycin-treated), LOV (Iovastatin-treated), to the sum ofGSH and GSSG present. Myocardial tissue was
LOV + ADR (lovastatin + adriamycin-treated), TRO (trolox- homogenized in 5% sulfosalicylic acid. The tissue homogenate
treated) and TRO + ADR (trolox + adriamycin-treated). was centrifuged for 10 min at 10,000 x g. GSSG alone was
Adriamycin (doxorubicin hydrochloride ) was administered measured by treating the sulfosalicylic acid supematant with
intraperitoneaHy (Monday, Wednesday and Friday) in six 2-vinylpyridine and triethanolamine. The solution was
equal injections (each containing 2.5 mg/kg ADR) to animals vigorously mixed and the final pH of the solution was
in ADR, PROB + ADR, LOV + ADR and TRO + ADR groups checked to be between 6 and 7. After 60 min, the derivatized
over aperiod of2 weeks for a cumulative dose of 15 mglkg sampies were assayed as described above in the glutathione
body weight. Probucol (cumulative dose, 120 mg/kg body reductase/5,5'-dithiobis-(2-nitrobenzoic acid)-GSSG reductase
weight) was also administered intraperitoneally to PROB and recycling assay. GSH values were calculated as the difference
PROB + ADR groups in twelve equal injections (each between total (GSSG + GSH) and GSSG concentrations.
treatment containing 10 mg/kg) over aperiod of 4 weeks, Values are reported in GSH equivalents, micromoies per gram
2 weeks prior and 2 weeks alternating with adriamycin of tissue weight.
injections (Tuesday, Thursday and Saturday). Lovastatin
was given (i.p.) to the LOV and LOV + ADR groups in 12
equal injections (each injection containing 4 mg/kg of Statistical analysis
lovastatin, total cumulative dose of 48 mg/kg), over aperiod
of 4 weeks, 2 weeks before adriamycin administration and 2 Data are expressed as mean ± S.E.M. For a statistical analysis
weeks altemating with adriamycin injections. Trolox was ofthe data, group means were compared by one-way ANOVA,
given (i.p.) to the rats in TRO and TRO + ADR groupS in 12 and Bonferroni' s test was used to identify differences between
equal injections (48 mg/kg total cumulative dose). Probucol, groups. Statistical significance was acceptable to a level of
lovastatin and trolox were dissolved in coconut oil prior to p < 0.05.
injection. CONT animals were injected with the vehicle
alone (lactose, 75 mg/kg in saline) in the same regimen as
ADR. In all instances, the volume ofvehicle injected was Results
1.25 ml/kg.
All animals were observed for as long as 3 weeks after the General observations, mortality and heart weight
last injection for general appearance, behaviour and mortality.
At the end ofthe 3 week post-treatment period, animals were General appearance of the animals in all 8 groups was
sacrificed by decapitation, and blood was collected in non- monitored throughout the study. Animals in the ADR group
heparinized tubes ('Vacutainer' brand SST). Serum was developed scruffy, yellowish fur and red exudate around the
immediately separated by centrifugation and was used for the eyes within I week ofthe last adriamycin injection. Similar
assessment offree fatty acids (FFA) and albumin. Atria and changes were observed in the LOV + ADR group. Animals
other connective tissue from hearts were dissected away, in the PROB + ADR group did not show any ofthese changes.
163

Tab!e 1. Effects ofprobucol, lovastatin and trolox on adriamycin-induced Serum free fatty acids (FFA), serum albumin and FFA/
changes in heart weight, heart weightIbody weight and mortality. albumin ratio
Animal Heart Weight Heart Weight/ Mortality
Group (g) Body Weight (%) Levels of serum FFA were measured in all groups, and data
Ratio X 10' are shown in Table 2. Adriamycin treatment caused a
CONT 1.63 ± 0.08 2.93 ± 0.07
significant increase in FFA levels in ADR group. Probucol,
0
ADR 1.10 ± 0.05' 2.34 ± 0.03* 45 lovastatin and trolox by themselves did not have any effect
PROB 1.52 ± 0.03 2.78 ± 0.09 0 on the FFA levels in the control animals. Both probucol and
PROB+ADR 1.46 ± 0.06 2.66 ± 0.12 0 lovastatin returned FFA back to controllevels in PROB +
LOV 1.66 ± 0.09 2.89 ± 0.11 0 ADR and LOV + ADR groups, while trolox had no influence.
LOV+ADR 1.14 ± 0.02* 2.68 ± 0.05 20
TRO 1.62 ± 0.08 3.05 ± 0.12
FFA levels in TRO + ADR group were not significantly
0
TRO+ ADR 1.20 ± 0.04* 3.06 ± 0.06 Not Available different from ADR group.
Serum albumin levels were significantly and comparably
CONT - contral; ADR - adriamycin; PROB - probucal; PROB+ADR-
decreased in ADR, PROB + ADR, LOV + ADR and TRO +
probucol + adriamycin; LOV - lovastatin; LOV + ADR Jovastatin +
adriamycin; TRO - troJox and TRO + ADR - trolox + adriamycin. Data ADR groups (Table 2). FFAialbumin ratio was significantly
are mean ± S.E.M. of8--10 animals for all studies except for mortality. For increased in ADR and TRO + ADR groups compared to all
determining mortality, 20 animals were used in each of the CONT, PROB, other groups (Fig. I). Both probucol and lovastatin treatments
PROB + ADR, LOV and TRO groups, and 40 animals each in the ADR normalized FFA/albumin ratio in PROB +ADR and LOV +
and LOV + ADR groups. Mortality data were not available for TRO +
ADR groups (Fig. I).
ADR group (see text for details). *Signiftcantly different from its own non-
adriamycin treated control.

Myocardial glutathione
The most noticeable characteristic ofrats in the ADR group
was the development of an enlarged abdomen due to Myocardiallevels of reduced (GSH) as weil as oxidized
accumulation of fluid in peritoneal cavity (ascites). Ab- glutathione (GSSG) were measured in all groups (Table 3).
dominal distention became apparent within a week after the GSH content was significantly decreased in all groups
adriamycin treatment was completed. Some abdominal exposed to adriamycin treatment including ADR, PROB +
distention due to ascites was noticed in LOV + ADR and ADR, LOV + ADR and TRO + ADR groups. Trolox and
TRO + ADR groups, while in the PROB + ADR group, only lovastatin treatment had no effect on adriamycin-induced
2 out of 14 rats had insignificant amounts of ascites (8 and decrease in GSH levels, while probucol treatment increased
15 ml respectively). GSH levels in PROB + ADR group compared to ADR, TRO
There were no deaths in CONT, PROB, LOV and TRO + ADR and LOV + ADR groups. However, GSH in the
groups (Table 1). Mortality rate was 45% in ADR group. PROB + ADR group still remained significantly decreased
Probucol treatment, in the PRO + ADR group, reduced compared to CONT, PROB, LOV and TRO groups. GSH
mortality to zero, while lovastatin in the LOV + ADR group
reduced mortality to 20%. Mortality data for TRO +ADR
group could not be obtained because of our commitment to Tab!e 2. Effects ofprobucol, lovastatin and trolox on adriamycin-induced
changes in serum free fatty acids and serum albumin.
comply with the new animal care regulations. Under these
regulations, animals showing any signs of distress or Animal Serum Albumin
significant abdominal distention must be removed from the Group Free Fatty Acids (mmol/L)
study even if the observation period cannot be completed. (mmol/L)
Death is no longer acceptable as a study endpoint at this CONT 0.44 ± 0.03 0.55 ± 0.28
university. ADR 0.79 ± 0.03' 0.45 ± 0.02*
Treatment with adriamycin resulted in a significant decrease PROB 0.44 ± 0.02 0.53 ± 0.00
in the heart weight in the ADR group compared to CONT, PROB + ADR 0.49 ± 0.13 0.45 ± 0.00'
LOV 0.43 ± 0.02 0.54 ± 0.01
PROB, LOV and TRO groups (Table 1). Probucol in the PRO LOV+ ADR 0.51 ± 0.04 0.46 ± 0.02'
+ ADR 6 group completely prevented the loss in heart weight TRO 0.35 ± 0.02 0.49 ± 0.00
due to adriamycin, while no such proteetion was seen in the TRO+ ADR 0.68 ± 0.02t 0.42 ± 0.00'
LOV + ADR and TRO + ADR groups. Heart to body weight
Data are mean ± S.E.M. of 6-8 hearts. All abbreviations are the same as in
ratio was significantly decreased only in ADR group, while Table 1. 'Signiftcantly different from CONT, LOV, PROB and TRO
it was normalized in the PROB + ADR, LOV + ADR and (P<0.05). 'Signiftcantly different from CONT, PROB, PROB + ADR, LOV,
TRO + ADR groups (Table 1). LOV + ADR and TRO (p < 0.05).
164

GSSG, but the differenee was not statistieally signifieant from


2.0 the ADR and LOV + ADR groups.
* Redox state was determined by calculating GSH/GSSG
ratio (Fig. 2). This ratio in hearts from the ADR group was
1.6 deereased by about 75% compared to controls. Lovastatin and
trolox treatment did not have any significant influence on the
....0:
.~
adriamyein-induced decrease in GSHIGSSG ratio. This ratio
~

·s=
in LOV + ADR and TRO + ADR groups stayed at depressed
1.2
levels as seen in the ADR group. However, significant
::I
.CI improvement in GSH/GSSG ratio was found in the PROB +
~ ADR group. In spite of the ameliorative effects of probucol
-<r.. 0.8
on adriamycin-induced changes in the glutathione system,
r.. the GSH/GSSG ratio in PROB + ADR group was still
significantly lower than in CONT, PROB, LOV and TRO
0.4 groups. Observed improvement in the redox state ofPROB
+ ADR group hearts was mainly due to a normalization of
the GSSG levels.

A P P+A L L+A T T+A

Fig. 1. Effects of probucol (P). lovstatin (L) and trolox (T) on adriamycin
Discussion
(A)·induced changes in serum free fatly acid (FFA)/albumin ratio. Values
are mean ± S.E.M. of 6--8 experiments. 'Significantly different from the Adriamyein (doxorubiein) is an exeellent anti-tumor anti-
C, P, P+A, L, L+A and T groups (p < 0.05). biotie, and is very effeetive against a large number ofhuman
malignaneies. Potential usefulness ofthis drug is very limited
by the development of a dose-related eardiomyopathie
levels in ADR, TRO + ADR and LOV + ADR groups were process terminating in severe heart failure which is refractory
decreased about 45% compared to controls, while in PROB
+ ADR group, this decrease was about 30% compared to
controllevels. There was no difference in GSH levels among 10,---------____________________--,
CONT, PROB, LOV and TRO groups.
GSSG levels were significantly increased in ADR, LOV
+ ADR and TRO + ADR groups (Table 3). Probucol treatment
retumed GSSG levels to control values in PROB + ADR 8
group. Lovastatin treatment had no influenee on GSSG levels,
while trolox treatment showed a trend towards a deerease of
6
Table 3. Effects ofprobucol, lovastatin and trolox on adriamycin-induced
changes in myocardial reduced (GSH) and oxidized (GSSG) glutathione.

Animal GSH GSSG


Group (~mol/g tissue) (~mol/g tissue)

CONT 56.16 ± 6.65 10.75 ± 1.03


ADR 26.33 ± 2A8' 16.75 ± 0.55" 2
PROB 54.00 ± 5.55 10.00 ± 0.93
PROB+ADR 39.00 ± 1.94'" 10AO ± 0.74
LOV 50.00 ± 4.56 8.66 ± 0.80
LOV+ ADR 28.20 ± 2.87' 16.40 ± 1.03"
TRO 58.16 ± 4.54 10.28 ± 1.80
TRO+ADR 24.80 ± 1.59' 14.00 ± 0.51"

Data are mean ± S.E.M from 6--8 experiments. Abbreviations are the same Fig.2. Effects ofprobucol (P), lovastatin (L) and trolox (T) on adriamycin
as in Table I. 'Significantly different from CONT, PROB, LOV and TRO (A)-induced changes in myocardial ratio of GSH/GSSG (redox state).
(p < 0.05). **Significantly different from CONT, PROB, PROB + ADR, Values are mean ± S.E.M. of 6-8 experiments. 'Significantly different
LOV and TRO (p < 0.05). "'Significantly different from all other groups from the C, P, P+A, Land T groups (p < 0.05). 'Significantly different
(p< 0.05). from the C, A, P, L, L+A, T and T+A groups (p < 0.05).
165

to any therapeutic approach [2, 4, 14, 15]. Combination study, probucol had no effect on adriamycin-induced decrease
therapy with probucol established in our laboratory has offered in serum albumin.
complete protection against the development of adriamycin Adriamycin treatment increased serum FFA and decreased
cardiomyopathy in rats [3, 9]. In the present study, the serum albumin levels, as weil as increased FFAIalbumin ratio.
chronic administration of adriamycin produced a typical Adverse effects ofhigh concentration offree fatty acids on
drug-induced cardiomyopathy. Combination treatment with cardiac contractility have been described before [12].
probucol prevented these changes as weil as the develop- Perfusion of isolated rat hearts with solutions containing
ment of congestive heart failure (data not shown) and also different free fatty acids/albumin ratios showed not only that
reduced mortality. Normalized cardiac structure due to high concentrations offree fatty acids have deleterious effects
probucol and adriamycin combination therapy has been on cardiac contractility, but that an increase in free fatty acids/
shown be fore [9]. albumin ratio could also be an important negative factor [12].
Adriamycin, due to its unique chemical structure, is very These serum changes due to adriamycin may contributc to
prone to generation of free radicals [6, 16]. There is an the reported depressed cardiac function due to this drug.
impressive list of evidence in the literature in support of the Both probucol and lovastatin treatment returned FFA/
role of free radicals in the pathogenesis of adriamycin- albumin ratio to the controllevels, as a result of a decrease
induced cardiomyopathy [3, 4]. Hence, antioxidant properties in serum FFA. Trolox had no effect on serum FFAlevels and
of probucol as weil as promotion of endogenous antioxidants FFA/albumin ratio. Reported decreases in carnitine levels
have been suggested to playa role in the protection against in the heart due to adriamycin [22] mayaIso promote the
adriamycin-induced cardiomyopathy [3, 9]. In addition, data production of free fatty acids. In this regard, carnitine plays
from the present study suggest that improved FFAIalbumin a central role in the transfer of long-chain fatty acids into
ratio as weil as redox ratio mayaiso have a role in this the mitochondrial matrix where ß-oxidation takes place
protection. [23]. Adriamycin-induced hyperlipidemia reported by us
In our study, adriamycin treatment decreased the levels of [24] as weil as by others [25,26] not only increases free fatty
reduced glutathione (GSH) and increased oxidized (GSSG) acids concentration, but it is also accompanied by hypo-
glutathione, resulting in a decreased GSH/GSSG ratio. albuminemia [10, 11]. Thus, adriamycin treatment increases
Different studies have described varying changes in glutathione free fatty acids/albumin ratio by affecting both componcnts
levels due to adriamycin administration [17-19]. While acute adversely. Modulation of the hyperlipidemia both by
studies described both increases [19] and decreases [17] in probucol and lovastatin improves free fatty acids/albumin
GSH levels, chronic administration of adriamycin caused ratio which might have a favourable effect on the cardiac
increase in cardiac glutathione levels [19]. When discussing function.
and comparing these glutathione data with those described Trolox has been shown to prevent cumene hydroperoxide-
in the literature, the experimental model and stage of the induced lipid peroxidation in cultured neonatal rat heart cells
cardiac dysfunction should be considered. It appears that [27]. Trolox reduced the area ofinfarction in a canine model
glutathione changes due to adriamycin administration of coronary artery occlusion, and it was more effective than
depend on the animal species, dose of the drug, length of superoxide dismutase or catalase in the protection of myocyte
the treatment and the post-treatment duration. Probucol cell cultures from the free radical damage [28]. In the present
treatment improved adriamycin-induced decrease in the study, trolox treatment did not have any effect on adriamycin-
GSH/GSG ratio mainly through normalization of GSSG induced decrease in GSH, increase in GSSG and decrease in
levels. This observation is in accord with the previously GSH/GSSG ratio. These changes might be explained by the
described beneficial effect ofprobucol on glutathione levels finding that trolox actually accelerates glutathione peroxidase
in the renal cortex of rats with bilateral urethral obstruction re action and GSH utilization, leading to lowering of GSH
[20]. levels as weil as GSH/GSSG ratio in myocytes exposed to
Comparable effects of probucol and lovastatin on total free radicals [27].
cholesterol have been described in rats with bilateral urethral
obstruction, while both drugs had very limited effect on
tri glyceride levels [20]. Probucol improved renal function in Conclusion
rats with bilateral urethral obstruction and lovastatin did not
[20]. Patients with hyperlipidemia due to nephrotic syndrome, Protection by probucol against adriamycin-induced cardio-
caused by a variety of renal pathologies, are benefited from myopathy is accompanied by modulation ofthe increase in
probucol treatment through the significant lowering of serum FFAIalbumin ratio and decrease in GSHIGSSG ratio due to
total cholesterol, triglycerides, HDL and LDL levels [21]. adriamycin. Both changes mayaIso contribute to the beneficial
However, probucol had no effect on urine and serum protein, effects of probucol against adriamycin-induced cardio-
serum albumin levels and renal function [21]. In the present myopathy and heart failure.
166

Acknowledgements 14. Minow RA, Benjamin RS, Lee EF, Gottlieb JA: Adriamycin cardio-
myopathy - risk factors. Cancer 39: 1397-1402, 1977
15. Praga C, Beretta G, Vigo PL, Lenaz GR, Pollini C, Bonadonna G,
Study supported by a grant from the Manitoba Heart and Connetta R, CasteIlani R, Villa E, Gallagher CG, von Melchner H,
Stroke Foundation. Dr. Singal is the recipient of a Career Hayat M, Ribaud P, de Wasch G, Mattsson R, Waldner HR, Kolavic
Award from the Medical Research Council. Drs. Iliskovic and K, Buchner R, Bokkel-Huyninck WT, Perevodchikova NI, Manziuk
LA, Senn HJ, Mayr AC: Adriamycin cardiotoxicity: A survey of1273
Palace were supported by the Manitoba Health Research
patients. Cancer Treat Rep 53: 827--834, 1979
Council. Ms. Khaper was supported by a Studentship from 16. Kalyanaraman B, Perez-Reyes E, Mason RP: Spin-trapping and direct
the University ofManitoba. electron spin resonance investigations of the redox metabolism of
quinone anticancer drugs. Biochim BiophysActa 630: 119-130, 1980
17. Boor PJ: Cardiac glutathione: Diurnal rhythrn and variation in drug-
induced cardiomyopathy. Res Commun Chem Path Pharmac 24: 27-
References 36, 1979
18. Julicher RHM, Sterrenberg L, Bast A, Riksen ROWM, Koomen JM,
I. Buja LM, Ferrans VJ, Mayer RJ, Roherts WC, Henderson ES: Cardiac Noordhoek J: The role of lipid peroxidation in acute doxorubicin-
ultrastructural changes induced by daunorubicin therapy. Cancer 32: induced cardiotoxicity as studied in rat isolated hear!. J Pharm
771-788, 1973 Pharmacol 38: 277-282, 1986
2. Lefrak EA, Pitha J, Rosenheim S, Gottleib JA: A clinicopathologic 19. Jackson JA, Reeves JP, Muntz KH, Kruk D, Prough RA, Willerson
analysis of adriamycin cardiotoxicity. Cancer 32: 302-314, 1973 JT, Buja LM: Evaluation of free radical effects and catecholamine
3. Singal PK, Siveski-Iliskovic N, Hill M, Thomas TP, Li T: Combination alterations in adriamycin cardiotoxicity. Am J Pathol 117: 140--153,
therapy with probucol prevents adriamycin-induced cardiomyopathy 1984
J Mol Cell Cardiol27: 1055-1063, 1995 20. Modi KS, Morrissey J, Shah SV, Schreiner GF, Klahr S: Effects of
4. Singal PK, Deally CMR, Weinberg LE: Subcellular effects of probucol on renal function in rats with bilateral ureteral obstruction.
adriamycin in the heart: A concise review. J Mol Cell Cardiol 19: Kidney International 38: 843--850, 1990
817--828,1987 21. Iida H, Izumino K,Asaka M, Fujita M, NishinoA, Sasayama S: Effect
5. Myers CE, McGuire WP, Liss RH, Ifrim I, Grotzinger K, Young RC: of probucol on hyperlipidemia in patients with nephrotic syndrome.
Adriamycin: the role oflipid peroxidation in cardiac toxicity and tumor Nephron 47: 280--283, 1987
response. Science 19: 165-167, 1977 22. Senekowitsch R, Lohninger A, Kriegel H, Staniek H, Kreiglsteiner
6. Doroshow JH: Effect of anthracycline antibiotics on oxygen radical H-P, Kaiser E: Protective effects of carnitine on adriamycin toxicity
formation in rat heart. Cancer Res 43: 460-472, 1983 to heart. In: H. Kaiser, A. Lohninger (eds). Carnitine - Its Role in
7. KaulN, Siveski-Iliskovic N,Hill M, SlezakJ, SingalPK: Free radicals Lung and Heart Disorders. Proc Satell Symp ZAK, Graz, 1985, pp
and the heart. J Pharmacol Toxicol Meth 30: 55--67, 1993 126--137
8. Siveski-Iliskovic N, Kaul N, Singal PK: Probucol prornotes endogenous 23. Bremer J: Carnitine metabolism and function. Physio1 Rev 63: 1420--
antioxidants and provides protection against adriamycin-induced 1480,1983
cardiomyopathy in rats. Circulation 89: 2829-2835, 1994 24. Iliskovic N, Singal PK: Lipid lowering: an important factor in
9. Siveski-Iliskovic N, Hill M, Chow D, Singal PK: Probucol prolects preventing adriamycin-induced heart failure. Am J Pathol 150: 727-
against adriamycin cardiomyopathy without interfering with its anti- 734,1997
tumorproperties. Circulation91: 10--15, 1995 25. Kunitomo M, Yamaguchi Y, Matsushima K, Futagawa Y, Bando Y:
10. Joles JA, van TolA, Jansen EHJM, Doomans HA, RabeiinkTJ, Grond Hyperlipidemic effects of adriamycin in rats. Japan J Pharmacol 39:
J, van Goor H: Plasma lipoproteins and renal apolipoproteins in rats 323-329, 1985
with chronic adriamycin nephrosis. Nephrol Dial Transplant 8: 831- 26. Washio M, Nanishi F, Okuda S, Onoyama K, Fujishima M: Alpha
838,1993 tocopherol improves focal glomerulosc1erosis in rats with adriamycin-
11. Skutelsky E, Hartzan S, Socher R, Gafter U: Modifications in induced progressive renal failure. Nephron 68: 347-352, 1994
glomerular polyanion distribution in adriamycin nephrosis. J Am Soc 27. Le CT, HoJlaar L, Van der Valk EJM, Franken NAPK, Van Ravels
Nephrol5: 1799-1805, 1995 FJM, Wondergem J, Van der LaarseA: Protection ofmyocytes against
12. Willebrands AF, Ter Welle HF, Tasseron SJA: The effect of a high free radical-induced damage by accelerated turnover ofthe glutathione
molar FFAlalbumin ratio in the perfusion medium on rhythm and redox cycle. Eur Heart J 16: 553-562, 1995
contractility ofthe isolated rat heart. J Mol Cell Cardiol5: 259-273, 28. Mickle DAG, Li RK, Weisel RD, Birnbaum PL, Wu TW, Jackowski
1979 G, Madonik MM, Burton GW, Ingold KU: Myocardial salvage with
13. Anderson ME: Determination of glutathione and glutathione disulfide trolox and ascorbic acid for an acute evolving infarction. Ann Thorac
in biological sampies. Methods Enzymo1113: 548--555, 1985 Surg 47: 553-557, 1989
PART V

HEART DISEASE
Moleeular and Cellular Biochemistry 188: 169-176, 1998.
© 1998 Kluwer Academic Publishers.

An AlG-rich motif in the rat fibroblast growth


factor-2 gene confers enhancer activity on a
heterologous promoter in neonatal rat cardiac
myocytes
Karen A. Detillieux, Adrienne EA. Meyers, Johanna T.A. Meij and
Peter A. Cattini
Gene Technology Group and Departments of Physiology and Pharmacology and Therapeutics, University of Manitoba,
Winnipeg, Manitoba, Canada

Abstract
We have cloned the rat fibroblast growth factor-2 (FGF-2) promoter region including 1058 base pairs (bp) of 5'-flanking DNA.
Complete sequencing ofthis promoter region revealed a 74 bp domain between nucleotides -793 and -720 that was greater than
97% AlG-rich. Arepeat ofthe sequence 5'-AGGGAGGG-3' separated by 11 bp was located at the core ofthis domain. A 37 bp
AlG-rich oligonucleotide containing these AGGG-repeat sequences was synthesised, and tested for function on a minimal herpes
simplex virus thymidine kinase (TK) promoter, fused to the firefly luciferase gene (TKp.luc), in transiently transfected neonatal
rat cardiac myocytes. Promoter activity was stimulated -3 fold in the presence ofAGGG-repeat sequences. This effect was neither
tissue or species-specific since TK promoter activity was increased -11 fold in both rat and human glial tumor cells. Four specific
complexes CC 1--4) were detected between neonatal rat heart nuc1ear proteins and the 37 bp AlG-rich oligonuc1eotide by gel mobility
shift assay. Competition with excess unlabelled 37 bp AlG-rich oligonucleotide revealed that two complexes represented very
high affinity/specificity interactions (C2 > C4) while CI and C3 were of lower affinity. As a result, competition with up to a 25
fold molar excess of37 bp AlG-rich oligonucleotide led to the loss ofC2 and C4, and a corresponding and transient increase in
the levels ofCl and C3, which themselves were reduced with more competitor oligonucleotide. The AGGG-repeat resembles the
5,_gGGGAGGG_3' sequence previously implicated in the response ofthe atrial natriuretic factor promoter to the a-adrenergic
agonist, phenylephrine. Although an additional 1.5 fold increase in TK promoter activity was detected in the presence ofthe
37 bp A/G-rieh oligonucleotide with phenylephrine treatment of transfected myocytes, this effect was not statistically
significant. Furthennore, there was no difference in the gel mobility shift (CI--4) pattern obtained with the 37 bp A/G-rieh
oligonucleotide and nuclear protein isolated from neonatal rat cardiac myocytes grown in the presence or absence of
norepinephrine. These data suggest that the AlG rich sequences in the rat FGF -2 gene 5'-flanking DNA, including the AGGG-
repeat, are able to confer stimulatory activity on a promoter in a tissue- and species-independent manner, but alone are not
able to induce a significant phenylephrine response in neonatal rat cardiac myocytes. (Mol Cell Biochem 188: 169-176, 1998)

Key words: FGF-2, transcription, gene transfer, HSV-thymidine kinase promoter

Introduction member of the tyrosine kinase family, are essential for the
nonnal growth and development ofthe myocardium [I, 2].
Fibroblast growth factor-2 (FGF-2) or basic fibroblast growth With regard to the post natal heart, there is increasing evidence
factor (bFGF) and its high affinity receptor FGFR-l, a that FGF-2, which is released with contraction [3, 4], is

Address Jor o.fJprints: P.A. Cattini. Gene Techno1ogy Group and Departrnent of Physio1ogy. University of Manitoba, 730 Williarn Avenue, Winnipeg,
Manitoba, Canada R3E 3J7
170

involved in the maintenance of a healthy myocardium gradient and plated on collagen coated plates at a density
through its angiogenic [5-7] as weil as cardioprotective of I x 10 6 cells per 35 mm dish. Cells were initially plated
properties [8, 9]. Many ofthe studies regarding FGF-2 have in Harn' s F 10 medium containing 10% fetal bovine serum
made use of exogenous addition ofthis factor, and focussed (FBS), 10% horse serum, antibiotic (l000 units/ml penicillin,
on its interaction with its high affinity cell surface receptor, 1 mg/mI streptomycin) and calcium chloride supplemented
which signals its mitogenic and angiogenic responses. In to 1.05 mM. Rat glioma C6 and human astrocytoma U87-
contrast, there have been relatively few studies on the MG cells were obtained from the American Type Culture
control ofFGF -2 synthesis, particularly at the transcriptional Collection and grown in monolayer culture in Dulbecco's
level, and specifically in the heart. The human and rat FGF- modified Eagle's medium (DMEM) supplemented with 10%
2 promoter sequences have been reported [10, 11], and (v/v) FBSwith antibiotic at 37°C in the presence of5% CO 2•
although they do not share extensive structural similarity, Plating densities for C6 and U87-MG cells were 0.5 x 10 6
short (-20 base pair) stretches ofhighly conserved sequences and 1.0 x 10 6 per 100 mm dish, rcspectively. All cultureware
were identified in the proximal promoter region [11]. Also, was purchased from Corning (Fisher Scientific, Nepean,
both promoters lacked a TATA box but contained G/C rich ON, Canada), and all media and reagents from Gibco-BRL
regions and Spl sites [10,11]. The human FGF-2 promoter (Life Technologies, Burlington, ON, Canada).
was shown to be regulated by mutant and wild type p53 in
a positive and negative manner, respectively, in reporter
gene/transfection studies using glioblastoma and hepato- Nuclear protein preparations
cellular carcinoma cells [12]. In addition, neuropeptides
were shown to influence the human FGF-2 promoter via the Heart tissue was dissccted aseptically from neonatal (1-2
zinc finger transcription factor Egr-I in glial cells [13]. A days) Sprague Dawley rats, and nuclear protein was prepared
consensus Egr-I site is also contained in the rat FGF-2 essentially as described previously 116], with final dialysis
promoter region [11]. against20 mM Hepes pH 7.9, 20% v/v glycerol, 0.1 M KCI,
Based on an alignment ofreported human and rat FGF-2 0.2 mM EDTA, 0.5 mM DTT and I mM PMSF. The nuc1ear
genomic sequences [10, 11], the rat fragment appears to extract was mixed with heparin-agarose and washed with
extend further upstream [11]. Here we report the detection 0.1 M KCl before elution of pro tein with 0.6 M KCI.
of an A/G rich domain in this extended DNA region containing Following determination ofprotein concentration using the
a repeat ofthe sequence 5'-AGGGAGGG-3'. This sequence Bradford Assay (Bio-Rad Laboratories, Mississauga, ON,
shows some similarity to a previously reported phenylephrine Canada), nuclear extracts were aliquoted and maintained at
responsive element (5'-gGGGAGGG-3') in the rat atrial -70°C. A modified, rapid extraction protocol [17] was used
natriuretic factor promoter, which is functional in transfected for isolation of nuclear protein from cultured neonatal rat
neonatal rat cardiac myocytes [14]. We have assessed these cardiac ventricular myocytes grown in the presence or
sequences for activity using the firefly luciferase reporter absence ofO.O I mM norepinephrine for 6 hours under serum-
gene assay and transient gene transfer into neonatal cardiac free conditions.
myocytes and non cardiac cells, as weil as their ability to
bind nuclear protein. Our data indicate that these A/G rich
sequences can confer stimulatory activity on a promoter, Hybrid gene constructions
without tissue- or species-specificity, and make high affinity/
specificity interactions with nuclear protein. These sequences The plasmid TKp.luc (pT81.1uc, 118]) contains a portion of
did not, however, show a significant response to phenylephrine the herpes simplex virus thymidine kinase (TK) promoter
in neonatal rat cardiac myocytes after gene transfer. (positions -81 to +52) fused to the firefly luciferase gene.
A 37 bp A/G rich (double-stranded) oligonucleotide,
corresponding to nucleotide positions -785/-749 [11], and
Materials and methods containing the putative AGGG-repeat sequences, 5'-
GGGAAAGGGAGGGGGMGGAAAGGAGGGAGGGMGGA-
Cell culture 3', was synthesised and inserted upstream of the TK promoter
in TKp.luc to generate A/G-TKp.luc. The 5 bp of sequence
Neonatal rat cardiac myocyte cultures were prepared upstream and downstream ofthe AGGG-repeat was included
essentially as previously described [15]. Briefly, ventricles in an attempt to limit any end effect that might interfere with
from Sprague-Dawley rat pups (at 1-2 days after birth) were specific protein binding. The promoterless firefly luciferase
dissected and the cells dissociated in aspinner flask using gene (-p.luc, contained in the vector pXPI) was described
a combination of trypsin and DNase I. Myocytes were previously [11]. The pRL-CMV vector, containing theRenilla
separated from non muscle cells on a discontinuous Percoll luciferase gene under the control of the cytomegalovirus
171

promoter was obtained from Promega Corporation (Fis her was 32P-end-Iabelled, in the presence of2 J!g ofpoly dI-dC.
Scientific). Incubation ofthe reaction (in 10 mM Hepes-NaOH pH 7.9,
50 mM KCI, 0.5 mM EDTA, 10% glycerol, 1 mM DTTwith
5 mM MgCI 2) at room temperature for 30 min was followed
Gene transfer by electrophoresis in non-denaturing 4% polyacrylamide
gels. For competition, specific unlabelled 37 bp A/G-rich
Cardiac myocytes and glial tumor cells were transfected by oligonucleotide or non specific RF -I element [21 ] (at a l-
the calcium phosphate-DNA precipitation method. Briefly, 100 fold molar excess over radiolabelIed oligonucleotide)
30 J!g of test plasmid (hybrid firefly luciferase gene) and 3 were added to nuclear extracts and incubated at room
J!g of control plasmid (pRL-CMV) was made up to a volume temperature for 10 min prior to the addition of the radio-
ofO.5 ml in 252 mM CaCl2 and added gradually to an equal labelled DNA and incubation at room temperature for a
volume of aerated 2 x HEBS buffer (280 mM NaCl, 50 mM further 20 min.
HEPES-KOH, pH 7.1, and 1.5 mM NalOJ Precipitate
was allowed to form at room temperature for 30 min, and
310 J!I was added to each of 3 culture dishes of cardiac Stalistical analysis
myocytes (35 mm) or glial tumor cells (100 mm) containing
4 ml or 8 ml ofDMEM/lO% FBS, respectively. Cells were Data presented in the text and figures represent the mean ±
transfected for 16 hand then washed thoroughly with S.E. from at least 2 independent experiments each done in
calcium- and magnesium-free phosphate-buffered saline. triplicate. Statistical analysis of the results was done using the
Cells were refed with DMEM/l 0% FBS and maintained for Mann-Whitney nonparametric test. Results were considered
48 h before processing. For phenylephrine treatment, significant ifp was determined to be < 0.05.
transfected cardiac myocytes were refed with DMEM-FI2
containing I x Redu-Ser II (Upstate Biotechnology, Lake
Placid, NY, U.S.A), 0.02 mg/mi ascorbic acid andantibiotics Results
with or without 0.1 mM phenylephrine, and maintained for
48 h before processing. Cells were also transfected with Sequence analysis reveals an A/G-rich domain in the rat
-p.luc as a control for random transcription initiation. Co- FGF-2 gene
transfection with pRL-CMV was used as a control for DNA
uptake and values were used subsequently to normalize the We have reported the cloning of a fragment of the rat FGF-
firefly luciferase activity (firefly luciferase/Renilla luciferase) 2 gene, including a promoter region of 1058 bp based on a
from the 'test' genes. For experiments with phenylephrine primary transcription start site (+ I) determined using brain
treatment, firefly luciferase activity was normalized RNA [11]. Inspection of these sequences reveals a 74 bp
against lysate protein content (luciferase activity/ng region which is 97% adenine/guanine (A/G-rich, located
protein) which was determined using the Bradford Assay bctween nucleotide positions-793 and -720 (Fig. I). Indeed,
(Bio-Rad Laboratories). a 65 bp stretch within this region (-793/-729) contains only
A or G residues. Located at the core of this region are two

Reporter gene assays


gctaccacagagaaaccctgtctagaaac
Both firefly and Renilla luciferase activities were measured,
following 'active lysis of cells by scraping', using the 'Dual- -793
Luciferase™ Reporter Assay System' (Promega) and a aCCAGAGAAGGGGGAAIAGGGAGGGIGGA
luminometer (ILA900 Luminometer, Tropix Inc., Bedford,
MA, USA) according to the manufacturer's instructions. The
protein content of cell Iysates generated in this manner was
AGGAAAGGIAGGGAGGGIAAGGAGGGAGA
determined by the bicinchoninic acid protein assay [19]. -729 -720
GGGGGAGGAAAGcAGGAcAGGtgttctca
Gel mobility shift assay
Fig. 1. Region eontaining AIG-rieh sequenees in the rat FGF-2 5-flanking
DNA. The A/G-rich region corresponds to nucleotides 793/-720, and its
The gel mobility shift assay was performed essentially as position is based on the major transcription initiation site dctcctcd in rat
described by Baldwin [20]. Nuclear extract (5 J!g) was brain RNA [11]. The two eopies of the AGGG-repeat sequenee, 5'-
incubated with the 37 bp AlG-rich oligonucleotide which AGGGAGGG-3' are boxed.
172

copies ofthe an AGGG-repeat sequence, 5'-AGGGAGGG- (-11 fold) versus primary eardiac myoeytes (-3 fold; Fig.
3'. This sequence shows a high degree of similarity with a 2B).
sequence, 5'-gGGGAGGG-3', found in the atrial natriuretic
factor (ANF) promoter and implicated in its response to (Xl-
specific adrenergic activation by phenylephrine [14]. A 37 bp Neonatal rat heart nuclear proteins make high affinityl
double-stranded AlG-rich oligonuc1eotide, eorresponding to specijicity interactions with the AIG-rich sequences
nuc1eotides -785/-749 ofthe FGF-2 promoter and eontaining
both copies of5'-AGGGAGGG-3', was synthesized (Fig. I, The gel mobility shift assay was used to investigate the
boxed sequenee). presenee of specific neonatal rat heart nuclear protein
interactions with the 37 bp AIG-rich oligonuc1eotide. The
radiolabelled DNA (0.5 ng) was incubated with 5 ~g of
An AIG-rich region containing an AGGG-repeat neonatal rat heart nuc1ear protein in the absence or presence
stimulates heterologous promoter activity in a tissue- and of a 25, 50 or 100 fold molar excess of unlabelled A/G-rich
:,pecies-independent manner oligonucleotide. As a further control, an RF-I DNA element,
containing an unrelated sequence, was also used at a 25, 50
To assess any effect of the AGGG-repeat sequenees on or 100 fold molar excess as a non speeifie oligonuc1eotide
promoter aetivity, the 37 bp AlG-rich oligonuc1eotide was competitor. Four specific complexes (CI--4) were identified
inserted upstream of a minimal viral TK gene promoter, (Fig. 3). Both C2 and C4 were competed eompletely with a
which was fused to the firefly luciferase gene (TKp.luc) to 25 fold molar exeess of specific (AlG-rieh oligonuc1eotide)
generate AlG-TKp.luc. Both TKp.luc and AlG-TKp.luc. were but not non specific (RF-I element) eompetitor. A slight
co-transfected with the hybrid Renilla luciferase gene (pRL increase in the amount ofCI and C3 complexes was detected
CMV) into neonatal rat eardiae myocytes as weil as rat C6 corresponding to the complete eompetition ofC2 and C4 with
and human U87 glial tumor cells, and then tested for activity a 25 (and to a lesser extent with a 50) fold molar exeess of
after 48 h. The results (firefly luciferase/Renilla luciferase specific competitor. The CI and C3 eomplexes required a 100
activity) are shown in Fig. 2A. A significant increase in TK fold molar excess of specific AlG-rieh oligonuc1eotide to be
promoter activity was observed in the presence ofthe AlG- competed efficiently (Fig. 3).
rich sequences in neonatal rat eardiac myoeytes (p < 0.005,
n =6), rat C6 glioma cells (p < 0.01, n = 5) and human U87-
MG astrocytoma cells (p < 0.05, n = 5). However, the level AlG-rlch RF-1
of stimulation was greater in rat or human glial tumor cells (specific) (non specific)
.....;-=--....;...-
FP OX 25x SOx l00x 25x SOx 100x

A B
15.---.---.---.
.TKp/uc 12] A/G-TKp.luc 0 -p.luc -C1
Q) 0.05
«
z
~ o -C2
T: -C3
E0.04 -/: '510 -
iI:
:fi ~'" 0.03 - /:
~3 -C4
.-
«&! ~
$ ! 0.02 -/
~ T
Ö
Ö 5 i--
~~ ~
CI)

~
~
w
a:
o ~
" 0.01
;: tl .v~~ I
'0
'0
~ u.
~ 0 '/ h o a b c d e f 9 h
ratCM rat C6 hum. U87 CM C6 U87
Fig.3. Gel mobility shift assay ofneonatal rat heart nuc\ear proteins and
Fig. 2. (A) Effeet ofthe 37 bp AlG-rich oligonuc\eotide on TK promoter the 37 bp A/G-rich oligonuc1eotide. Speeifieity was determined by
aetivity (firefly luciferaselRenilla lueiferase) in neonatal rat cardiae competition with 'speeifie' unlabelled AlG-rich or 'non specifie' RF-l
myoeytes (CM) as well as rat C6 and human U87 glial tumor eells after element oligonuc\eotide competitors. The radiolabelled AlG-rieh fragment
transient gene transfer. Results are expressed as the mean from at least two (free probe, FP) was ineubated in the (a) absence or (b-h) presence of (5
independent experiments. The bars represent the standard error ofthe mean. ~g) nuclear protein. For eompetition, (b) none or (c) 25, (d) 50, or (e)100
(B) The results from (A) are presented to show fold effeet ofthe A/G-rieh fold molar exeess of specific competitor, or (f) 25, (g) 50, or (h) 100 fold
sequenees on TK promoter aetivity (A/G-TKp.luclTKp.luc) in the various molar exeess of non specifie eompetitor was used. The positions/mobilities
eell types. offour speeifie eomplexes (Cl-C4) are indicated.
173

AlG-rlch ollgonucleotlde competltor • -PE 123 +PE


(molar excess) 120+-------.-------~
FP Ox Ix 2x 5x 10x 15x

100+-------~---=±=~

a-~
.>.~ 80+--------4----V
~~
~ ~ 60 +--------+----/'.
... ,E
S~
o ~ 40+--------+--~'l/

e
E~
g 20
a..2
+-----r.,;;..,,-+---O

O+----<....<....<L.....f-
a b c d e f 9 TKp.luc AlG-TKp.luc

Fig. 4. Detection of very high affinity interactions between neonatal rat Fig.5. Effect ofO.1 mM phenylephrine treatment for 48 h on TK promoter
heart nuclear protein and the 37 bp AIG-rich DNA fragment. Affinity was activity in the presence (A!G-TKp./uc) or absence (TKp./uc) ofthe 37 bp
determined by gel mobility shift assay and competition with low amounts AlG-rich oligonucleotide in transiently transfected neonatal rat cardiac
of 'specific' unlabelled AlG-rich oligonucleotide competitor. The radio- myocytes. Results are expressed as mean promoter activity (firefly
labelIed AlG-rich fragment (FP) was incubated in the (a) absence or (b-g) luciferase/ng protein). Basal levels for --p.luc in the presence and absence
presence of (5 flg) nuclear protein, (b) without or with a (c) I, (d) 2, (e) 5, ofphenylephrine were 0.078 ± 0.007 and 0.030 ± 0.004, respectively. Bars
(f) 10, or (g) 15 fold molar excess of specific competitor. The positions! represent standard error of the mean.
mobilities ofthe four specific complexes (CI-C4) are indicated.

To assess the relative affinity of nuelear protein for DNA sequences, respectively. However, although there was an
in the C2 vs. C4 eomplexes, gel mobility shift assays were additional -1.5 fold increase in activity observed in the
done using lesser amounts (1, 2, 5, 10, and 15 fold molar presence of the 5'-AGGGAGGG-3' sequence with phenyle-
exeess) of 37 bp AIG-rieh oligonuc1eotide for competition phrine treatment, this was not considered statistically
(Fig. 4). Complex C2 represents a very high affinity/ significant (p = O. 19).
specificity interaction since it was competed efficiently with To eomplement this study, we eompared the gel mobility
only a 2 fold molar excess of specifie AlG-rieh oligo- shift assay patterns obtained using the 37 bp AIG-rieh
nucleotide. C4 was competed eompletely with a 10 fold oligonuc1eotide, containing two copies ofthe 5'-AGGGAGGG-
molar exeess of speeific eompetitor. The transient inerease 3' sequence, with nuclear protein isolated from neonatal rat
in the amount of C 1 and C3 with competition of C2 and C4 cardiac myocytes grown in the absence versus presence of
was also apparent. 0.01 mM norepinephrine. Four complexes (CI-4) were
observed (Fig. 6). This pattern was not altered by norepine-
phrine stimulation. There was also no difference in the
EjJect ojphenylephrine treatment on the pattern oj degree of competition with a 50 or 100 fold molar excess
interaction between cardiac myocyte nuclear pro tein and of specifie AlG-rieh oligonuc1eotide, suggesting no change
DNA containing the J'-AGGGAGGG-J' sequence in affinity of these complexes with adrenergic stimulation
(Fig.6).
To assess the effect of phenylephrine on TK promoter
activity in the absence orpresenee ofthe 5'-AGGGAGGG-
3' sequenees, neonatal rat cardiac myocytes were transfected Discussion
with TKp./uc or AlG-TKp.luc. Transfected cells were treated
without or with 0.1 mM phenylephrine for 48 h, harvested We have used both functional and structural approaches in
and firefly lueiferase aetivity per ng protein was determined the form of transient gene transfer and gel mobility shift
(Fig. 5). Phenylephrine treatment inereased TK promoter assays to characterise the repeat elements 5'-AGGGAGGG-
activity 4.6 ± 0.8 (p < 0.0001, n =17) and 6.7 ± 1.1 fold (p < 3' contained in a 37 bp DNA fragment eomprised totally of
0.008, n = 5) in the absence and presence of the AlG-rich adenosine (A) or guanosine (G) monophosphates. These
174

control cell norepinephrine-trealed AGGG-repeats, alone, in the regulation of the FGF-2


nuclear protein ceU nuclear protein promoter in cardiac myocytes or glial cells [Detillieux,
FP 2.5 5 5 5 5 2.5 5 5 5 5Jl9 Meyers and Cattini, unpublished observations]. Of course,
- 0 50100 100 - 0 50100 100 xmolar this does not rule out the possibility that additional elements
--;::jG AF-1 -zr;- F;1 8XceSS
comp.
and/or factors participate in common complexes with those
associated with the AIG-rich sequences, thereby modifying
their action and, thus, relative importance.
-C1 The data obtained from gel mobility shift assays indicate
-C2 the presence of multiple proteinlDNA interactions. Four
-C3 complexes (C 1--4) between neonatal rat heart nuclear protein
-C4 and the A/G-rich oligonucIeotide were identified (Figs 3 and
4). The same pattern of complexes was also suggested when
nucIear protein isolated f,om neonatal rat cardiac myocytes
was used (Fig 5B), indicating that these interactions are not
only the product of non muscIe cell pro teins present in the
heart. All four complexes were specific as they were
competed by increasing amounts (25- 100 fold molar
abcdef ghlj k excess) ofunlabelled NG-rich oligonucleotide, but not by
equivalent doses ofthe RF-I DNA element [21] containing
Fig. 6. Comparison of gel mobility shin patterns seen with the 37 bp AI
unrelated sequences (Fig. 3). The complete competition of
G-rich oligonucleotide and nuclearprotein from isolated cardiac myocytes C2 and C4 with a 25 fold molar excess indicates that these
grown in the absence (b-t) or presence (e-k) ofO.OI mM norepinephrine complexes possess a higher affinity/specificity than CI and
for 6 h. Affinitylspecifieity was assessed by eompetition with 'speeifie' C3, which required 100 fold molar excess of specific
unlabelled AlG-rich or 'non specific' RF-I oligonucleodde competitors. oligonuc1eotide to see efficient competition. Interestingly,
The radiolabelled AlG-rieh fragment (FP) was incubated in the (a) absence
or (b, g) presence of2.5 ±g or (c-f, h-k) 5 I'g ofnuclear protein, (c, h)
an increase in CI and C3 was suggested with the complete
without or (d, i) with aSO, or (e, j) 100 fold molar excess of specific competition of C2 and C4 in the presence of a 25 fold molar
competitor, or (f, k) 100 fold molar exeess of non specific competitor. excess of AIG-rich oligonucleotide (Fig. 3). The experiment
The positions offour specific complexes (C l-C4) are indicated. was repeated with a reduced level of specific competitor
DNA fragment (1- 15 fold molar excess) to further dissect
this observation as well as determine the relative affinity of
sequences represent the core of a 74 bp region located at the C2 and C4 complexes (Fig. 4). C2 and C4 were effi-
nuc1eotide position -793/-720 that is 97% NG residues, ciently competed with a 2 and 10 fold molar excess, respect-
and ofwhich a 65 bp stretch contains only A or G nucleotides ively, confirming that these are both very high affinity/
(Fig. I). NG-rich regions have been linked to gene regulation specificity interactions (C2 > C4). These results also show
[14] as weil as alterations in DNA structure, specifically that there is a transient increase in Cl and C3 binding which
with regard to the human ß-globin locus control region [22]. corresponds to the competition ofC2 and C4. This suggests
The ability ofthe NG-rich oligonucleotide to increase the the possibility that the protein/DNA interactions reflected
activity of a minimal 81 bp viral thymidine kinase promoter by C2/C4 and C lIC3 are mutually exclusive with a preference,
in neonatal rat cardiac myocytes as weil as rat C6 and human under the experimental conditions used, for C2/C4 to form.
U87 glial tumor cells (Fig. 2) suggests that these sequences, When the higher affinity C2/C4 events are made invisible by
and more specifically the A/G-rich elements, can regulate competition (with a low molar excess of A/G-rich oligo-
gene expression. Furthermore, the stimulation of promoter nucleotide) , there is an increased opportunity for the
activity in cells of different tissues (cardiac myocytes vs. proteins associated with Cl and C3 to bind. The ability to
brain glial cells) or species (rat versus human glial cells) detect all four complexes (CI4) simultaneously suggests
origin correlates with the ubiquitous pattern of FGF-2 that the proteins associated with lower affinity CI and C3
synthesis. FGF -2 has been found in every tissue examined are present in excess in the neonatal rat heart. This scenario
so far [23, 24] and its structure is highly conserved between offers a possible mechanism for regulation of nuclear protein
species, such that mouse and human FGF-2 share 94% binding and hence function .
sequence similarity at the amino acid level [25]. However, The AGGG-repeat sequence 5'-AGGGAGGG-3' present
although the A/G oligonuc1eotide was able to stimulate in the FGF-2 gene is related structurally to the sequence 5'-
promoter activity in the context of a heterologous promoter gGGGAGGG-3', wh ich was reported previously to be the
and reporter gene (Fig. 2), a deletion analysis ofthe rat FGF- phenylephrine responsive element in the ANF promoter
2 5'-flanking DNA does not support a major role for the [14] . Indeed, the sequence 5' AGGGAGGG-3' exists in the
175

promoter of human a-skeletal actin, and was suggested to 4, Kaye D, Pimental D, Prasad S, Maki S, Berger HJ, McNeil PL, Smith
playa role in its phenylephrine responsiveness [14]. A TW, Kelly RA: Role of transient!y altered sareolemmal membrane
permeability and basic fibroblast growth factor release in the
subsequent investigation demonstrated the requirement of at
hypertrophic response of adult rat ventricu1ar myoeytes to increased
least one additional element, resembling a serum response mechanical activity in vitra. J Clin Invest 97: 281-291,1996
element, to produce an efficient response of theANF promoter 5, Unger EF, Banai S, Shou M, Lazarous DF, Jaklitsch MT, Scheinowitz
to phenylephrine treatment [26]. Although we detected an M, Correa R, Ningbeil C, Epstein SE: Basic fibroblast growth faetor
additional -1.5 fold increase in viral thymidine kinase enhances myocardial collateral flow in a canine model. Am J Physiol
266 (Heart Circ PhysioI35): HI58-HI595, 1994
promoter activity that could be attributed to the presence of
6. Lazarous DF, Scheinowitz M, Shou M, Hodge E, Majanayagam MAS,
AGGG-repeat sequences after phenylephrine treatment of Hunsberger S, Robinson Jr WG, Stiber JA, Correa R, Epstein SE,
transfected neonatal rat cardiac myocytes, this effect was not Unger EF: Effeets of ehronie systcmieadministration ofbasic fibroblast
statistically significant (Fig. 5). Furthermore, there was no growth faetor on collateral development in the canine heart. Circulation
change in the pattern of DNA interaction with nuclear 91: 145-153, 1995
7. Fox JC, Shanley JR: Antisense inhibition ofbasic fibroblast growth
proteins (CI-4) isolated from neonatal rat cardiac myocytes factor induces apoptosis in vascular smooth museIe cells. J Bio1 Chem
grown in the absence or presence of norepinephrine (Fig. 6). 271: 12578-12584, 1996
Thus, it is possible that the duplicated 5'-AGGGAGGG-3' 8. Padua RR, Sethi R, Dhalla NS, Kardami E: Basic fibroblast growth
sequences present in the FGF-2 gene constitute elements factor is cardioprotective in ischemia-reperfusion injury. Mol Cell
that contribute to phenylephrine regulation of the FGF-2 Biochem 143: 129-135,1995
9. Padua RR, Sethi R, Davey-Forgie SE, Lui L, Dhalla N, Kardami E:
promoter, but are themselves insufficient for a complete
Cardioprotection and basic fibroblast growth factor. In: N.S. Dhalla,
response. P.K. Singal, R.E. Beamish (eds). Heart Hypertrophy and Failure.
In summary, examination ofsequences upstream ofthe rat Kluwer Academic Publishers, Boston, 1996, pp 501-518.
FGF-2 co ding region revealed a 65 bp domain containing 10. Shibata F, Baird A, Florkiewicz RZ: Functional characterization of
only A or G nucleotides. Located at the core of this region the human basic fibroblast growth factor promoter. Growth Factors 4:
277-287,1991
are two copies of the repeat sequence 5'-AGGGAGGG-3', 11. Pasumarthi KBS, Jin Y, Cattini PA: Cloning ofthe rat fibroblast growth
which shows a strong similarity to a phenylephrine response factor-2 promoter region and its response to mitogenic stimuli in glioma
element. These A/G-rich sequences bind cardiac nuclear C6 cells. J Neurochem 68: 898-908, 1997
protein with high affinity and are able to confer stimulatory 12. Ueba T, Nosaka T, Takahashi JA, Shibata F, Florkiewicz RZ, Vogelstein
B, Oda Y, Kikuchi H, Hatanaka M: Transcriptional regulation ofbasic
activity on a viral promoter in a tissue- and species-independent
fibroblast growth factor gene by p53 in human glioblastma and
manner, but do not respond significantly to phenylephrine in hepatocellular careinoma cells. Proc Nat! Acad Sci USA 91: 9009-
transfected neonatal rat cardiac myocytes. 9013, 1994
13. Biesiada E, Razandi M, Levin ER: Egr-I aclivates basic fibroblast
growth factor transcription. J Biol Chern 271: 18576--18581, 1996
14. Ardati A, Nemer M A nuclear pathway for u,-adrenergic receptor
Acknowledgements signaling in cardiac cells. EMBO J 12: 5131-5139, 1993
15. Pasumarthi KBS, Kardami E, Cattini PA: High and low molecular
weight fibroblast growth faclor-2 increase proliferation 01' neonatal
The authors would like to thank Rama Mohan Surabhi far his
rat cardiac myocytes but have differential effects on binuclealion and
careful review of the manuscript. This work was supported nuclear morphology: Evidence for both paracrine and intracrine actions
by a grant from the Medical Research Council of Canada offibroblast growth faetor-2. Circ Res 78: 126--136, 1996
(MT-13398). KAD is the recipient of a Heart and Stroke 16. Dignam JD, Lebovitz RM, Roeder RG: Accurate transcription initiation
Foundation Studentship and AFAM is the recipient of a by RNA polymerase II in a soluble extract from isolated rnammalian
nuclei. Nucl Acids Res 11: 1475-1489, 1983
Medical Research Council Studentship Award.
17. Andrews Ne, Faller DV: A rapid micropreparation technique for
extraetion ofDNAbinding proteins from Iimiting numbers ofmammalian
cells. Nucl Acids Res 19: 2499, 1991
References 18. Nordeen SK: Luciferase reporter gene veetors for analysis of promoters
and enhaneers. Biotechniques 6: 454-458, 1988
19. Smith PK, Krohn RI, Hermansen GT, Mallia AK, Gartner FH,
1. Sugi Y, Sasse J, Lough J: Inhibition of preeardiae mesoderm eell Fugimoto EK, Gocke NM, Olson BJ, Klenk DC: Measurement of
proliferation by anti sense oligodeoxynucleotide eomplementary to protein using bieinchoninic acid. Anal Biochern 150: 76--85, 1985
fibroblast growth factor-2 (FGF-2). Dev Bio1157: 28-37, 1993 20. Baldwin AS: Analysis of sequence-specific DNA-binding proteins by
2. Mima T, Ueno H, Fischman DA, Williams LT, Mikawa T: Fibroblast the gel mobility shift assay. DNA Protein Eng Tech 2: 73--76, 1990
growth faetor receptor is required for in vive eardiae myocyte 21. Lytras A, Cattini PA: Human chorionic somatomammotropin gene
proliferation at early embryonic stages of heart development. Proc enhancer activity is dependent on the blockade of a repressor
Nat! Acad Sci USA 92: 467-471, 1995 rnechanism. Mol Endocrinol8: 478-489, 1994
3. Clarke MS, Caldwell RW, Chiao H, Miyake K, MeNeil PL: Contraction- 22. Boulikas T: Homeodomain protein binding sitos, invcrtcd repeats, and
induced cell wounding and release of basic fibroblast growth factor. nuclear matrix attachment regions along the human ß-globin gene
Circ Res 76: 927-~34, 1995 complex. J Cell Biochem 52: 23-36, 1993
176

23. Kardami E, Liu L, Padua RR, Fandrich RR, Pasumarthi SKB, Caltini Gao G, Goldfarb M: Reeeptor speeificity ofthe fibroblasl growth faclor
PA: Regulation of basic fibroblast growth factor (bFGF) and FGF family. J Biol Chem 271: 15292~15297, 1996
receptors in the heart. Ann NY Acad Sci 752: 353~369, 1995 26. Sprenkle AB, Murray SF, Glembotski CC: Involvement of multiple
24. Bikfalvi A, Klein S, Pintueci G, Rifkin OB: Biologieal roles of cis-elements in basal and a-adrenergie agonist-indueible ANF
fibroblast growth faetor-2. Endoerine Rev 18: 26--45,1997 transcription. Roles for serum response elements and an Sp-I-hke
25. Omitz DM, Xu J, Colvin JS, MeEwen DG, MaeArthur CA, Coulier F, element. Cire Res 77: 1060-1069, 1995
Molecular and Cellular Biochemistry 188: 177-185, 1998.
© 1998 Kluwer Academic Publishers.

Influence of different culture conditions on


sarcoplasmic reticular calcium transport in isolated
neonatalratcardiomyocytes
Roland Vetter,I,3 Monika Kott, 1 Wolfgang Schulze 1 and Heinz Rupp2
IMax Delbrück Center tor Molecular Medicine (MDC), Berlin-Buch; 2Molecular Cardiology Laboratory, Philipps
University Marburg, Marburg; l Institute 0/ Clinical Pharmacology and Toxicology, Free University 0/Berlin, Berlin,
Germany

Abstract
This study investigates sarcoplasmic reticulum (SR) calcium-(Ca2+) transportATPase (SERCA2a) and phospholamban (PLB)
in cultured spontaneously contracting neonatal rat cardiomyocytes (CM) to ascertain the function ofboth SR proteins under
various culture conditions. The two major SR proteins were readily detectable in cultured CM by immunofluorescent microscopy
using specific anti-SERCA2 and anti-PLB antibodies. Double labeling technique revealed that PLB-positive CM also labeled
with anti-SERCA2. Coexpression of SERCA2 and PLB in CM was supported by measurement of cell homogenate oxalate-
supported Ca 2+ uptake which was completely inhibited by thapsigargin and stimulated by protein kinase A-catalyzed
phosphorylation. Under serum-free conditions, incubation ofCM with the SERCA2a expression modulator 3,3 ',5-triiodo-L-
thyronine (100 nM, 72 h) resulted in elevated Ca2+ uptake of +33%. Specific Ca2+ uptake activity was not altered ifinsulin was
omitted from the serum-free culture medium but total SR Ca2+ transport activity was reduced under this culture condition. The
results indicate that primary culture of spontaneously contracting neonatal rat CM can be employed as a useful model system
for investigating both short- and long-term mechanisms determining the Ca2+ re-uptake function ofthe SR under defined culture
conditions. (Mol Cell Biochem 188: 177-185, 1998)

Key words: neonatal rat cardiomyocyte culture, sarcoplasmic reticulum, phospholamban, calciumATPase, calcium transport,
thyroid hormone

Introduction mental and disease-related alterations in the expression ofthe


SR Ca2+ ATPase iso form SERCA2a in cardiomyocytes as an
The sarcoplasmic reticulum (SR) in cardiac muscle controls important contributing mechanism to these changes [7, 9-12,
the relaxation of the heart muscle cell by removing Ca2+ from 14]. In addition, altered levels and state ofphosphorylation
the cytosol and acts as a source of Ca2+ fOT myofilament of the Ca 2+-ATPase modulatory protein phospholamban
activation during the excitation-contraction coupling process (PLB) could be another contributing factor [15-18]. Phos-
[I]. Several previous studies have shown that the SR Ca2+ re- pholamban, a homopentameric SR protein has been shown
uptake capability and expression of SR pro teins undergo to regulate the rate of SR Ca2+ transport through changes in
changes during developmental heart growth and aging [2- the affinity ofthe Ca2+-ATPase for Ca2+. The Ca2+ transporting
7]. Alterations of SR Ca2+ transport also occur in human and activity of this enzyme is suppressed while PLB is dephos-
experimental overload hypertrophy and congestive heart phorylated. This inhibition is relieved upon phosphorylation
failure [7-14]. However, the underlying mechanisms and of this pro tein by cyclic AMP( cAMP)-dependent protein
involved neuro-endocrine and metabolic signals are not kinase (at serine 16) which in intact tissue occurs in response
completely understood. Recent evidence points to develop- to ß-adrenergic stimulation factor [15-18]. Phospholamban

Address Jor ojJPrints: R. Vetter, Institute of C1inical Pharmaco1ogy and Toxico1ogy, Free University of Berlin, Garystrasse 5, D-14195, Berlin (Dah1em),
Germany
178

can also become phosphorylated at a distinct amino acid resi- sedimented at 300 x g for 8 min using a non-refrigerated
due by Ca2+/calmodulin-dependent proteinkinase [16, 17, 19]. centrifuge. The sedimented cells were resuspended in cell
Enzymatically isolated neonatal rat cardiomyocytes cul- growth medium CMRL 1415-ATM which was supplemented
tured in chemically defined culture medium are frequently with 10% FCS, 10% horse serum (HS), and 0.02 mg/mi
used as a cell model for studying the expression and function gentamicin (all constituents from Biochrom KG, Berlin,
of various functional important proteins [20-23]. With Germany) and adjusted to pH 7.40 using I N NaOH.
respect to cellular Ca'+ handling, spontaneously contracting For enrichment of cardiomyocytes, 20 ml ofthis suspen-
rat cardiomyocytes are expected to express functionally active sion containing approx 2 x 108 cardiac cells were incubated
SR proteins such as SERCA2a and PLB. It is of partieular in 175 cm' plastic Coming culture flask in a water-saturated
interest whether both proteins are coexpressed in this isolated atmosphere for 90 min at 37°C. During this incubation most
cell culture system in detectable amounts and which factors do of the nonmuscle cells comprising 30-40 % of total cell
determine their expression and function in cultured cardio- number attached to the bottom surface of the culture flask
myocytes that lack vegetative innervation and are not exposed [23]. The cardiomyocyte-enriched supematant was carefully
to a number ofhumoral factors present in vivo. removed and pooled in an Erlenmeyer flask. The number of
As the short- and long-term regulation ofthe cardiac SR cells was counted using a light microscope and a hemo-
Ca2+transport function is thought to depend on various factors cytometer, and cell density was adjusted with additional
and mechanisms that may be altered with the physiological growth medium to 1.2 x 106 cells/ml. Volumes of 5 ml final
and diseased state of the organism, we initiated a comparative cell suspension were then incubated in 25 cm' Coming culture
study on the Ca'+ transport function of this intracellular flasks in a water-saturated atmosphere at 37°C. Culture flasks
organelle in primary cultures of neonatal rat cardiomyocytes were unsealed for free gas exchange with air. Incubation with
that were exposed to growth medium with different com- 5% CO, was not necessary due to the high buffer capacity of
position. Furthermore, the question was addressed as to bicarbonate-free CMRL 1415 ATM medium [24]. After 24 h
whether the SR Ca'+ ATPase SERCA2a and its modulator ofincubation, the FCS/HS-supplemented growth medium was
protein PLB are coexpressed in cultured cardiomyocytes. replaced either by a 10% FCS-supplemented CMRL 1415-
ATM or serum-free, hormone-supplemented CMRL 1415-
ATM medium containing in addition 0.1 11M dexamethasone,
Materials and methods 0.4 11M ferrum-saturated transferrin, 5 11M bovine insulin
(Sigma-Aldrich Chemie GmbH, Deisenhoven, Germany), and
Preparation and primary culture of neonatal rat 0.4 11M bovine serum albumin (SERVA Feinbiochemica
cardiomyocytes GmbH, Heidelberg, Germany) if not otherwise indicated. In
part ofthe experiments growth medium also contained different
Primary neonatal heart cell cultures were prepared from concentrations of 3,3' ,5-triiodo-L-thyronine (T 3: Sigma-
ventricular tissue of I to 3 day-old Sprague-Dawley rat pups Aldrich Chemie GmbH, Deisenhofen, Germany). To inhibit
(Tierzucht Schönwalde GmbH, Schönwalde, Germany) by prolifcration of contaminating nonmuscle cells 2 11M fluoro-
a modification ofthe procedures described earlier [21, 23]. deoxyuridine (Sigma-Aldrich Chemie GmbH, Steinheim,
The removed ventricles of 30-50 animals were placed into Germany) was present in both serum-containing and serum-
ice-cold calcium ion-free phosphate-buffered solutionAcon- free culture medium throughout the whole culture period.
taining 120 mM NaCI, 4.56 mM KCl, 0.44 mM KHl04, 0.42 Growth medium was exchanged daily and biochemical
mM Na2 HP04 , 25 mMNaHC0 3 and 5.55 mM glucose, pH 7.5, analyses were performed at culture day 4 when cardio-
as weil as 0.5 mg/mi streptomycin and 5000 lE/ml penicillin myocytes had formed a synchronously contracting mono-
G (Biochrom KG, Berlin, Germany). Ventricular tissue was layer. The spontaneous contraction rate and photographic
transferred to a Petri dish and minced into pieces of approx I documentation of the growing cardiomyocytes were monitored
mm3 in size using two steril scalpels. Stepwise disaggregation before each medium exchange using an inverse phase contrast
for 15 min each of tissue pieces into single cells was performed microscope (Olympus IMT-2, Olympus, Hamburg, Germany).
at 37°C under continuous mixing with a magnetic stirrer at 150
rpm in a 50 ml Erlenmeyer flask containing 15 ml solution A
supplemented with 0.12% trypsin (Biochrom KG, Berlin, Fluorescence microscopy
Germany). The first tissue digest that consisted mainly of cell
debris and mesenchymal cells was discarded. The following For immunofluorescent detection ofthe SR Ca2+ ATPase and
3-6 supematants obtained after each 15 min digestion period PLB isolated neonatal rat cardiomyocytes were plated on
were poured into single sterile glass centrifuge tubes each glass cover slips (10 x 10 mm) andcultured for 96 h in serum-
containing 4 ml of ice-cold, heat-inactivated fetal calf serum supplemented CMRL-1415ATM medium as described above.
(FCS; Biochrom KG, Berlin, Germany). Cells were gently Sampies were shortly rinsed in phosphate-buffered saline
179

(PBS), fIXed for 10 min in PBS containing 2%paraformalde- Sarcoplasmic reticulum oxalate-supported Ca 2+ uptake
hyde and blocked against nonspecific labeling in PBS
containing 1% bovine serum albumin, fraction V (Sigma, As the amount of cell material obtained from cultures was
Deisenhofen, Germany) and 0.2% gelatine (Merck, Darm- limited and procedures for isolation ofpurified SR vesicles
stadt, Germany). A subsequent wash in PBS was performed from intact tissue are known to result in low and variable
prior to incubation with the respective primary antibody. For yields of material [27, 28], we decided to use cellular
detecting the SR Ca2+-ATPase SERCA2a, incubation was homogenates for the determination of Ca2+uptake. Oxalate-
performed with a rabbit anti-SERCA2 antiserum (dilution supported Ca2+ uptake into SR vesicles was measured at
1:500, kind gift of Dr. Dillmann, San Diego, USA) for 16 h 37°C in 40 mM imidazole-HCI buffer, pH 7.0, 100 mM KCl,
at +4°C in detergent-free PBS. For detecting immunoreactive 5 mM MgC12, 5 mM tris(hydroxymethyl)-aminomethane
PLB a 1:50 diluted monoclonal mouse anti-PLB antibody (Tris)-ATP, 6 mM phosphocreatine, 10 mM K-oxalate, 10
(Biomol, Hamburg, Germany) that recognized both phos- mM NaN 3 , 0.2 mM ethylene glycol-bis-(ß-aminoethyl
phorylated and nonphosphorylated PLB was used. Sampies ether)-N,N,N',N' -tetraacedic acid (EGTA), 0.1 mM 45 CaCl2
were then washed three times for 10 min in PBS containing (Amersham Buchler GmbH & Co KG, Braunschweig,
50 mM glycine and incubated for 2 h with dichlorotriazinyl Germany; sp act 12 dpm/pmol; 0.21 11M free Ca2+ con-
aminofluorescein(DTAF)-conjugated goat anti-rabbit IgG centration) and 10--1511g total cell homogenate protein per
(H+L) (1 mg dissolved in 100 ml Hp; Dianova, Hamburg, 0.35 ml uptake medium employing a method used pre-
Germany) or for 1 h with Cy3-conjugated goat anti-mouse viously for measurement of Ca 2+ uptake in ventricular
IgG (H+L) (0.75 mg dissolved in 100 ml Hp; Dianova, homogenates from intact neonatal rat hearts [29]. After
Hamburg, Germany). For double labeling, sampies were preincubation of the reaction mixture for 2 min in the
incubated first with anti-SERCA2 and then with anti-PLB absence ofhomogenate, the Ca2+uptake was started by the
followed by the sequential treatment with the solutions addition ofhomogenate. At selected time intervals, sampies
containing both secondary IgG conjugates. After mounting were taken and filtered by suction through 0.45 /!ID HAWP
on slides, stained sampies were examined and photographed Millipore filters (Millipore, Eschborn, Germany). Filters were
using a Zeiss Axioplan fluorescence microscope equipped washed twice with 3 m1 of ice-cold 40 mM imidazole, pH 7.0,
with appropriate filters. SERCA2 and PLB were visualized 100 mM KCI and 2 mM EGTA. Radioactivity associated with
by green and red fluorescence, respectively. Specificity of dry filters was deterrnined by liquid scintillation counting.
used primary antibodies was checked by Western blotting Reaction mixtures contained either 10 11M synthetic protein
analysis as described elsewhere [25]. kinase A inhibitor peptide [PKI( 6-22)amide] (GIBCO BRL,
Life Technologies GmbH, Eggenstein, Germany), 2 11M
catalytic (C) subunit of adenosine 3' -5'-cyclic mono-
Cell homogenates phosphate (cAMP)-dependent protein kinase (protein
At culture day 4 medium was removed from culture flasks kinase A), 10 11M thapsigargin or no additions. Transport
and the cell monolayer was washed twice with 10 ml ice-cold rates were calculated by the linear regression of data points
depolarisation buffer B containing 120 mM NaCl, 30 mM KCl at 1.0,2.0, 3.0,4.0 and 5.0 min measured in duplicate or
5 mM TrisIHCI, pH 7.4. The cell monolayer was then removed triplicate. The reaction mixtures allowed for Ca2+ uptake
from the bottom of culture flasks in the presence of 5 m1 ice- into SR vesicles only, with ATP-dependent Ca 2+ transport
cold buffer B using a self-made scraper made of silicon rubber into mitochondrial vesicles being inhibited by NaN 3 [30).
and a glass rod. The content of the flask was transferred to a Specific Ca2+uptake activity expressed in nmol Ca2+Imin/mg
conic 11 m1 plastic centrifugation tube and combined with protein is defined as the rate of oxalate-supported Ca2+uptake
additional 5 m1 ofbuffer B that were used forrinsing the culture related to milligram of cell homogenate protein. Total Ca2+
flask. Cells were then sedimented by centrifugation (300 x g, uptake activity expressed in Ca2+/min/flask is defined as the
10 min, 4°C). Obtained pellets were homogenized in 0.4 ml rate of uptake related to the amount of cell protein of 6 x 106
ice-cold buffer C containing 250 mM sucrose, 50 mM cells that are seeded in one culture flask.
NaHl04, 10 mM NaF, 1 mM EDTA and 10 mM histidine,
pH 7.4, using a Polytron PT-1200 (Kinematica, GmbH,
Luzern, Switzerland). Homogenization was done 6 times for Statistical analysis
10 sec each at a setting of 6. A sampie of the homogenate was
immediately used for Ca2+ uptake experiments, others were Statistical comparisons between different tissue cultures were
shock-frozen -in liquid nitrogen and stored at--80°C until use performed by unpaired t-test or nonparametrie Mann-Whit-
for other biochemical analysis. Protein was determined by ney test. Values are mean ± S.E.M. ifnot indicated otherwise.
Lowry's method using ovalbumin as standard [26]. Statistical significance was assumed at p < 0.05.
180

a b

Fig. I. Immunofluoreseenee mierographs of neonalal rat heart eells at eullure day 4 labeled with monoclonal anti-PLB and polyclonal anti-SERCA2
antibodics. Anligen-bound primary PLB and SERCA2 anlibodies were visualized with Cy3-eonjugated goat anti-mouse IgG anlibodies (a, red fluorescenee
of three eells) and DTAF-eonjugated goat anti-rabbil IgG antibodies Ce, green fluorescenee), respectively. Colocalization of PLB with SERCA2 in
eardiomyocytes is indicated by yellow color obtained after double labeling with anti-PLB and anti-SERCA2 anlibodies (b, same three cells as in micrograph
a). Anli-PLB slained parts of cardiomyoeytes only (a, b). Nelwork like fluorescenee in anti-SERCA2 labeled cardiomyocytes tbroughout the whole cell
revealed an apparently close conneelion of sareoplasmie reticulum and myofibrils. Both sareoplasmie retieulum proteins are most abundani in tbe perinuclear
region of eardiomyoeytes (CM). In noncardiomyocytes (NCM) week green flnorescenee indieates low abundance of SERCA in the cytoplasm and around
the nucleus. All bars = 10 11m. See also Materials and methods.
181

Results Table 1. Spontaneous contraction rate, total cellular protein and sarcoplasmic
reticulum calcium ion transport activity of neonatal rat cardiomyoctes
cultured in CMRL1415-ATM medium under serum-supplemented (+serum)
Application of immunofluorescence microscopy shows that and serum-free (-serum) conditions
the two major SR proteins PLB and SERCA2a were abundant
Parameter + serum -serum Difference
in spontaneously contracting neonatal rat cardiomyocytes at
culture day 4 (Fig. 1). By contrast, non-muscle cells were not Rate of contraction (l/min) 142 ±40 71 ±47* -50%
labeled with anti-PLB (Fig. 1 a). They were weakly labeled Protein (mg/jlask) 0.54 ± 0.04 0.09 ± 0.02* --80%
with anti-SERCA2 (Fig. Ib). Double labeling experiments Cell homogenate oxalate-supported SR Ca2+ uptake:
revealed that those cardiomyocytes labeled with anti-PLB Specific activity
also labeled with anti-SERCA2 (Fig. I a, b). However, not all (nmol Ca"/mgprotein/min) 3.79 ± 0.58 3.97 ± 1.12 +5%
cardiomyocytes labeled with SERCA2a also labeled with Total activity
anti-PLB (data not shown). Immunofluorescent SERCA2 (nmol Ca"/jlask/min) 1.72 ± 0.30 0.33 ± 0.09* --81%
staining appeared as a fme network-like pattern throughout Stimulation by catalytic
subunit of protein
the cardiomyocyte. A striation pattern resembling a myofib- kinase A (x{old) 1.6 ± 0.1 3.0 ± 0.3* +84%
rillar-like structure was observed frequently in certain
locations of the cardiomyocytes (Fig. 1 c). Most intense Values are means ± S.E.M. for 6 different cell isolations in each group and
were detennined at cu1ture day 4. Seeding density was 6 x 10'cellslculture
fluorescence for both antibodies was detected in the peri- fiasko Serum-supplemented medium contained 10% fetal calf serum. *p <
nuclear region. Thus, enzymatically isolated neonatal rat 0.05, significantly different from the value obtained in serum-supplemented
cardiomyocytes cultured in CMRL-1415ATM express the SR culture medium. See also Materials and methods.
Ca2+-ATPase and its modulator protein PLB. It indicates that
the SR ofthese cultured cells pocesses an apparently abun- In order to examine the functional capability ofthe SR Ca2+
dant protein machinery for ATP-dependent removal of transport in cultured neonatal cardiomyocytes oxalate-
cytosolic Ca2+ during each spontaneous contraction-relax- supported SR 45Ca2+ uptake was measured in freshly prepared
ation cycle.
25r------------------------,
.--------~8I o no addition

-S
100
o no addition
• 30 nM T3


80 b. + thapsigargin c:
20 100 nM T3
30 e0-
e)

20 ~ 15

10 -~c: 10

0
A if . b.
:zs: 6 ~
5

0 2 4 6 30
Time (min) o~--~--~--~--~--~~
o 1 2 3 4 5
Fig. 2. Time dependence of oxalate-supported Ca" uptake of cell
homogenates of cultured neonatal rat cardiomyocytes in the absence and Time (min)
presence of the specific SERCA inhibitor thapsigargin. Single values of
duplicate detenninations are expressed in nmoles Ca" taken up per mg Hg. 3. Infiuence of triiodo-L-thyronine (T,) on oxalate-supported Ca"
cellular protein. For examination of specific Ca" pump inhibition by the uptake ofneonatal rat cardiomyocytes cultured under serum-me conditions.
drug, homogenates were preincubated in ice for 30 min in the presence of T,-treatment was started at culture day 2 and Ca" uptake measurements
10 IJM thapsigargin; the same inhibitor concentration was also present in were perfonned 72 hlater. Mean values of duplicate determinations of a
the uptake medium. Cardiomyocytes were cultured for 96 h under serum- typical experiment are shown. Values are expressed in nmoles Ca" taken
containing conditions. See also Materials and methods. up per mg cellular protein. See also Materials and methods.
182

whole cell homogenates under nonphosporylating and To ex amine the functional response of SR Ca" transport
phosphorylating conditions. As shown in Fig. 2 cell homo- system to hormones that are known to alter the expression
genate oxalate-supported Ca'+ uptake was linear within the SERCA2 in vivo [32-34] the influence ofthyroid hormone
first 5 min of uptake. It was completely blocked in the T3 was investigated in cardiomyocytes. Figure 3 shows that
presence of the specific SR Ca'+ pump inhibitor thapsigargin exposure of serum-free cultures to T3 for 72 h resulted in a
[31]. The SR Ca'+ transport rate related per mg of total dose-dependent increase in the rate of cell homogenate SR
cellular protein was not significantly altered if myocytes Ca2+uptake.At 100 nM T3 , Ca2+uptake rate increased by 33%
were cultured under serum-free conditions (Table I). By (3.53 ± 0.19 vs. 4.70 ± 0.02 nmol Ca'+/mg protein/min, p <
contrast, there was a 5-fold reduction of the total cellular 0.05). This was accompanied by an elevation oftotal cellular
SR Ca2+ uptake activity under the latter conditions. It was protein by 45%. The effects of T3 could not be observed if
paralleled by a similar decline in total cellular protein and a cardiomyocytes were cultured in serum-supplemented
marked reduction in the spontaneous contraction rate (Table CMRL-1415ATM medium.
1). Removal ofinsulin from the serum-free CMRL-141 5ATM
The SR Ca2+ transport of cultured neonatal rat heart cells did not alter the specific SR Ca2+ uptake activity. However,
was sensitive to short-tenn metabolic regulation by protein this resulted in a decrease of total SR Ca'+ transport activity
phosphorylation. Addition of exogenous C subunit of protein (Fig. 4). Lack of insulin also caused a reduction ofboth total
kinase A to the Ca'+ uptake assay resulted in a strong stimu- cellular protein and the rate of spontaneous contraction by
lation of the SR Ca2+ uptake in cell homogenates prepared approximately 40% (data not shown).
from cardiomyocytes both cultured under serum-supple-
mented and serum-free medium conditions (Table I). The
protein kinaseA-induced increase ofCa'+ uptake was signifi- Discussion
cantly higher if cells were cultured under serum-free con-
ditions (Table 1).
Primary ccll culturcs of nconatal rat cardiomyocytcs havc
become a useful model system for cellular and molecular
studies ofthe myocardium [20, 23]. This is particularly due
o + insulin to the fact that these cells maintain many ofthe differentiated
6 • - insulin properties of cardiac muscle cells typical for the intact
myocardium. The data ofthe present study demonstrate that
cultured neonatal rat cardiomyocytes coexpress the two
sarcoplasmic reticular proteins SERCA2a and PLB which are
known to playa critical role in cellular Ca'+ handling in intact
myocardium [1, 18]. Evidence for expression of either
SERCA2a mRNA [35] or PLB [36] in primary cultures of
neonatal rat hearts has been reported earlier. However, these
studies havc not presented evidence for expression of both
functionally related SR proteins in a single cultured cardio-
myocyte. Coexpression of both SR proteins appears to be
valid for most ofthe cardiomyocytes ofthe primary culture
as demonstrated by double labeling of single cardiomyocytes
using anti-SERCA2 and anti-PLB antibodies. Interestingly,
0'-----'-----'---'-----'---.1...-------' we also observed some cardiomyocytes with abundant anti-
o 2 3 4 5 SERCA2 staining which apparently did not contain detect-
able amounts of PLB. This would suggests that SERCA2a-
Time (min) catalyzed Ca 2+re-uptake ofthese cardiomyocytes cannot be
modulated by reversible phosphorylation of PLB. Whether
Fig. 4. Rcduction of total oxalate-supported Ca" uptake of neonatal rat
this phenomenon is due to the culture conditions or reflects
eardiomyocytcs cultured under serum-free conditions in the absence of
insulin. Values are means ± S.E.M. far three different cultures in eaeh
a certain heterogeneity with respect to PLB expression in
group. *p < 0.05, significantly different from values obtained from eells cardiomyocytes that is characteristic for the intact neonatal
grown in the presenee of insulin. Seeding density was 6 x 106 eellslt1ask. rat ventricular myocardium needs further investigation.
Note: Ca" transport related to milligram of eell protein did not differ Colocalization of both coexpressed proteins in the mem-
between the two groups apparcntly duc to matched dechne oftolal cellular
branes of the SR of cultured cardiomyocytes is indirectly
and SERCA2a protein in the absence of insulin. See also Materials and
methods.
supported by oxalate-supported Ca'+ uptake data. In whole
183

cell homogenates, SR Ca2+uptake was found to be stimulated primary culture is thought to be constant due to inclusion of
by protein kinase A-catalyzed in vitro phosphorylation. the inhibitor fluorodeoxyuridine in the serum-free culture
Although in vitro formation ofphosphorylated PLB [17, 19, medium. As the Ca'+ uptake values in the T3 experiments
29, 37J has not been directly studied in this work, it seems were related to milli gram of total cellular protein, a T3-
reasonable to assurne that the observed protein kinase A- induced elevation ofthe laUer would blunt the T1 -induced
induced stimulation of Ca2+ uptake was due to phosphory- increase in specific SR Ca 2+ transport. F or example, a
lation ofPLB. An association ofthe degree ofphosphorylation matched increase or decrease in total cellular protein and
of PLB and SR Ca 2+ uptake activity has been frequently SERCA2a protein is expected to result in unchanged Ca2+
observed in whole tissue homogenates and purified frag- uptake values per milligram of total cellular protein. Thus,
mented SR prepared from intact myocardium [3, 18,37]. the observed T)-induced increase ofthe specific SR Ca2+uptake
Phospholamban phosphorylation in cultured neonatal rat apparently indicates a stronger activation of SERCA2a
heart cardiomyocytes has also been described [36]. Interest- expression than the T)-induction of other proteins that
ingly, the degree ofprotein kinaseA-induced stimulation of contribute to total cellular protein.
Ca2+ uptake described here was high suggesting a low level Under serum-free conditions cultured cardiomyocytes
of PLB phosphorylation under culture conditions [16, 37]. atrophied if insulin was omitted from the culture medium
This seems to be mainly due to the lack of sympathetic (data not shown). This was accompanied by a decline in total
innervation and circulating catecholamines in the primary cellular protein and total Ca2+uptake per culture flask whereas
culture [36J which are major determinants of in vivo PLB specific Ca'+ uptake activity related per milli gram of total
phosphorylation [16-18J. The higher degree of protein protein remaincd unchanged. This might suggest that the
kinase A-induced stimulation of Ca 2+ uptake in homo- insulin-dependent decrease in the overall cellular protein was
genates prepared from cardiomyocytes that were cultured matched by a similar decline in Ca'+ pump protein level.
under serum-free conditions supports this view. Under the However, a conclusive proof ofthis hypothesis would need
latter conditions, any serum-related influences on phos- additional experiments. These results differ from data
phorylation of PLB in cultured cardiomyocytes can be obtained in animals with insulin deficiency due to experi-
defini tely excluded. mental diabetes [41,42]. In the latter model specific cardiac
It is weil established that the SR Ca2+re-uptake activity in SR Ca 2+ transport activity was found to be markedly de-
hearts of intact animals is highly influenced by thyroid pressed. This suggests that insulin effects on the SR Ca2+
hormone [32, 38~O]. Alteration of SERCA2a mRNA ex- pump function may depend on a variety of other factors and
pression by thyroid hormone has been identified as one of influences that act on the heart in the animal model but are
the underlying mechanisms [33, 34]. It has also been reported lacking in the cell culture model system.
that the level of mRNA encoding the SR Ca 2+ ATPase in Taken together, the data of previous work [35, 36J and
primary neonatal rat cardiomyocytes is highly dependent on results presented in this study demonstrate that primary
thyroid hormone due to the presence of a thyroid hormone eulture of neonatal rat cardiomyocytes can be employed as a
responsive element in the promotor region of the SERCA2 useful model system to study mechanisms of short- and long-
gene [35,39]. In extension ofthis previous work with primary term regulation of the SERCA2/PLB system at the gene,
cultures of neonatal rat cardiomyocytes, we could demon- protein and functionallevel.
strate here rar this model system that the thyroid hormone
effect on expression ofthe SERCA2 can also be detected at
the functionallevel. A thyroid hormone-dependent increase Acknowledgments
in the SR Ca2+ ATPase-catalyzed oxalate-supparted Ca'+
uptake could be observed in cardiomyocytes cultured under The work was supported by grants from Deutsche Fors-
serum-free conditions; the effect was missing in cultures chungsgemeinschaft to R. V. (Ve-13611-3) and to H. R. (Ru-
grown in serum-supplemented medium. The latter finding is 245/6-2) as weil as by BMBF grant Kan HM5 to R. V. We are
in agreement with the observation ofRohrer et al. [35J that grateful to Dr. Wolfgang Dillmann far kind gift of SERCA2-
added T3 did cause a SR Ca2+ATPase induction only ifthe antiserum.
culture medium was supplemented with thyroidectomized
serum instead of normal calf serum that is known to contain
considerable amounts of endogenous T) [35].
Thyroid hormone treatment of serum-free cultures also
References
resulted in an increase of total cellular protein indicating
1. Katz AM, Takenaka H, Watras J: The sarcoplasmic reticulum. In: HA
hypertrophie growth of cardiomyocytes. Although we did not Fozzard, E Haber, RB Jennings,AM Katz, HE Morgan (eds). The Heart
measure the proteinlDNA ratio or cell area this assurnption and Cardiovascular System. Raven Press, New York, 1986, pp 731-
seems to be justified since the DNA concentration in the 746
184

2. Lompre A-M, Lambert F, Lakatta EG, Schwartz K: Expression of 19. Lindemann JP, WatanabeAM: Phosphorylation ofphospholamban in
sarcoplasmic reticulum Ca'+-ATPase and calsequestrin genes in rat intact myocardium. Role ofCa'+-calmodulin-dependent mechanisms.
heart during ontogenic development and aging. Cire Res 69: 1380- J Biol Chem 260: 4516-4525,1985
1388, 1991 20. Engelmann GL, McTiernan C, Gerrity RG, SarnareIAM: Serum-free
3. Freestone N, Singh J, Krause E-G, Vetter R: Early postoatal changes primary cultures of neonatal rat cardiomyocytes: Cellular and
in sarcoplasmic reticulum calcium transport funetion in spontaneously molecular applications. Technique 2: 279-291, 1990
hypertensive rats. Mol Cell Biochem 1641164: 57-66, 1996 21. Halle W, Wollenberger A: Differentiation and behavior of isolated
4. Vetter R, Studer R, Reinecke H, Kolar F, Ostadalova I, Drexler H: embryonic and neonatal heart cells in a chemically defined medium.
Reciprocal changes in the postoatal expression of the sarcolemmal Am J Cardiol25: 292-299, 1970
Na+-Ca'+-exchanger and SERCA2 in rat heart. J Mol Cell Cardiol27: 22. Wallukat G, Wollenberger A: Supersensitivity to ~-adrenoceptor
1689-1701,1995 stimulation evoked in eultured neonatal rat heart myocytes by L(+)-
5. Maciel LM, Polikar R, Rohrer D, Popovich BK, Dillmann WH: Age- lactate and pyruvate. J Auton Pharmacol13: 1-14, 1993
induced decreases in the messenger RNA coding for the sarcoplasmic 23. Werdan K, Erdmann E: Preparation and culture of embryonic and
reticulum Ca'+-ATPase ofthe rat heart. Cire Res 67: 230-234, 1990 neonatal heart muscle cells: Modification of transport activity. Meth
6. Arai M, Otsu K, Maclennan DH, Periasarny M: Regulation of Enzymo1173: 634-662, 1989
sarcoplasmic reticulum gene expression during cardiac and skeletal 24. Healy GM, Parker RC:An improved chemically defined basal medium
musc1e development. Am J Physiol262: C614-C620, 1992 (CMRL-1415) for newly explanted mouse embryo cells. J Cell Biol
7. Komuro I, Kumbayashi M, ShibazakiY, Takaku F, YazakiY: Molecular 30: 531-538, 1966
cloning and charaeterization of a Ca'+ + Mg'+-dependent adenosine 25. Kaasik A, Paju K, Vetter R, Seppet EK: Thyroid hormones increase
triphosphatase from rat cardiac sarcoplasmic reticulum. Regulation the contractility but suppress the effect of ~-adrenergic agonist by
of its expression by pressure overload and deyelopmental stage. J Clin decreasing phospholamban expression in rat atria. Cardiovasc Res 35:
Invest 83: 1102-1108, 1989 106-112, 1997
8. Levitsky D, de la Bastie D, Schwartz I(, Lompre A-M: Ca'+-ATPase 26. Lowry OH, Rosebrough NJ, FarrAL, Randali RI: Proteinmeasurement
and function of sarcoplasmic reticulum during eardiac hypertrophy. with the folin phenol reagent. J Biol Chem 193: 265-275, 1951
Am J Physiol261: 23-26, 1991 27. Feher JJ, Briggs FN, Hess ML: Characterization of cardiac sarco-
9. Sehwartz K, Carrier L, Lompre A-M, Mereadier JJ, Boheler KR: plasmic reticulum from ischemic myocardium: Comparison of isolated
Contractile proteins and sarcoplasmie retieulum ealeium-ATPase gene sarcoplasmic reticulum with unftactionated homogenates. J Mol Cell
expression in the hypertrophied and failing heart. Basic Res Cardiol Cardiol12: 427-432, 1980
87: 285-290, 1992 28. Solaro RI, Briggs FN: Estimating the functional capabilities of
10. de la Bastie D, Levitsky D, Rappaport L, Mereadier J-J, Marotte F, sarcoplasmic reticulum in cardiac muscle. Circ Res 34: 531-540, 1974
Wisnewsky C, Brovkovieh V, Sehwartz I(, LompreA-M: Function of 29. Vetter R, Rupp H: CPT-I inhibition by etomoxir has a chamber-related
the sareoplasmie retieulum and expression of its Ca'+-ATPase gene in action on cardiac sarcop1asmic reticulum and isomyosins. Am J Physiol
pressure overload-induced cardiac hypertrophy in the rat. Circ Res 267: H2091-H2099, 1994
66: 554-564, 1990 30. LompreA-M, de la Bastie D, Boheler KR, Schwartz K: Chacterization
11. Mereadier JJ, LompreA-M, Duc P, Boheler KR, Fraysse m, Wisnew- and expression of the rat heart sarcoplasmic reticulum Ca'+-ATPase
sky C, Allen PD, Komajda M, Schwartz K: Altered sareoplasmie mRNA. FEBS Lett 249: 35-41, 1989
reticulum Ca'+-ATPase gene expression in the human ventric1e during 31. Lytton J, Westlin M, Han1ey MR: Thapsigargin inhibits the sarco-
end-stage heart failure. J Clin Invest 85: 305-309, 1990 plasmic or endoplasmic reticulum Ca-ATPase farnily of calcium
12. Arai M, Matsui H, Periasarny M: Sareoplasmic reticulum gene pumps. J Biol Chem 266: 17067-17071, 1991
expression in cardiac hypertrophy and heart failure. Cire Res 74: 555- 32. Seppet EI(, Kadaya LY, Hata T, Kailikorm AP, Saks VA, Vetter R,
564,1994 Dhalla NS: Thyroid control over membrane processes in rat heart.
13. LarnmerichA, Günther J, Pfitzer G, Storch E, Vetter R:Alterations of Am J Physiol261: 66-71, 1991
cardiac contractile function are related to changes in membrane calcium 33. Rohrer D, Dillmann WH: Thyroid hormone markedly increases the
transport in spontaneously hypertensive rats. J Hypertens 13: 1313- mRNAcoding for sarcoplasmic reticulum Ca2+-ATPase in the ratheart.
1324,1995 J Biol Chem 263: 6941-6944, 1988
14. Hasenfuss G, Reineeke H, Studer R, Meyer M, Pieske B, Holtz J, 34. Arai M, Otsu K, Mac1ennan DH, Alpert NR, Periasarny M: Effect of
Holubarsch C, Posival H, Just H, Drexler H: Relation between thyroid hormone on the expression of messenger RNA encoding
myoeardial funetion and expression of sarcoplasmie retieulum Ca'+- sarcoplasmic reticulum proteins. Circ Res 69: 266-276, 1991
ATPase in failing and nonfailing human myoeardium. Cire Res 75: 35. Rohrer DK, Hartong R, Dillmann WH: Influence ofthyroid hormone
434-442, 1994 and retinoic acid on slow sarcoplasmic reticulum Ca'+ ATPase and
15. Vetter R, Kott M, Rupp H: Differential influenees of carnitine myosin heavy chain IX gene expression in cardiac myocytes.
palmitoyltransferase-I inhibition and hyperthyroidism on eardiae Delineation of cis-active DNA elements that confer responsiveness
growth and sarcoplasmic reticulum phosphorylation. Eur Heart J 15 to thyroid hormone but not to retinoic acid. J Biol Chem 266: 8638-
(Suppl D): 31-37, 1994 8646,1991
16. Karczewski P, Bartel S, Krause EG: Differential sensitivity to 36. Bartel S, Willenbrock R, Haase H, Karczewski P, Wallukat G, Dietz R,
isoprenaline of troponin I and phospholarnban phosphorylation in Krause EG: Cyc1ic GMP-mediated phospholamban phosphorylation in
isolated rat hearts. Biochem J 266: 115-122, 1990 intact cardiomyocytes. Biochem Biophys Res Commun 214: 75-80, 1995
17. Wegener AD, Simmerman HK, Lindemann JP, Jones LR: Phosphol- 37. Karczewski P, Vetter R, Holtzhauer M, Krause EG: Indirect technique
amban phosphorylation in intact ventricies. Phosphorylation of serine for the estimation of cAMP-dependent and Ca'+/caimoduIin-dependent
16 and threonine 17 in response to j3-adrenergic stimulation. J Biol phospholamban phosphorylation state in canine heartin vivo. Biomed
Chern 264: 11468-11474, 1989 BiochimActa 45: S227-S23l, 1986
18. Tada M, KatzAM: Phosphorylation ofthe sarcoplasmic retieulum and 38. Kolar F, Seppet EK, Vetter R, Prochazka J, Grunermel J, Zilmer K,
sarcolemrna. Annu Rev Physiol44: 401-423, 1982 Ostadal B: Thyruid control of contractile function and calcium handling
185

in neonatal rat heart. Pflügers Arch 421: 26--31, 1992 41. Penpargkul S, FeinF, SonnenblickEH, Scheurer J: Depressedcardiac
39. Dillmann WH: Biochemical basis of thyroid hormone action in the sarcoplasmic reticular funetion from diahetie rats. J Mol Cel1 Cardiol
heart. Am J Med 88: 626-630, 1990 13: 303-309, 1981
40. Beekman RE, Van Hardeveld C, Simonides WS: On the mechanism 42. Ganguly PK, Pieree GN, Dhalla KS, Dhalla NS: Defective sareo-
of the reduetion by thyroid hormone of ß-adrenergie relaxation rate plasmie retieular calcium transport in diahetie eardiomyopathy. Am J
stimulation in rat heart. Bioehem J 259: 229-236, 1989 Physiol 244: E528-E535, 1983
Molecular and Cellular Biachemistry 188: 187-197, 1998.
© 1998 Kluwer Academic Publishers.

Taurine indirectly increases [Ca]i by inducing Ca2+


influx through the Na+-Ca2+exchanger
Ghassan Bkaily, l Doris Jaalouk, l Sawsan Sader, l Hadia Shbaklo, l
Pierre Pothier,l Danielle Jacques,l Pedro D'Orleans-Juste,l Edward J.
Cragoe Jr. 3 and Ratna Bose2
lMRCC Group on immuno-cardiovascular interaction, Department ofAnatomy and Cell Biology, Faculty of Medicine,
University of Sherbrooke, Sherbrooke, QUI?bec; 2Departmenf of Pharmacology and Therapeutics, University ofManitoba,
Winnipeg, Manitoba, Canada; 3Nacogdoches, Texas, USA

Abstract
Recent studies in heart cells have shown taurine to induce a sustained increase ofboth intracellular Ca2+and Na+. These results led
us to believe that the increase in Na+ by taurine could be due to Na+ entry through the taurine-Na+ cotransporter which in turn
favours transarcolenunal Ca2+ influx through Na+_Ca2+ exchange. Therefore, we investigated the effect of ß-alanine, ablocker of
the taurine-Na+ cotransporter and low concentrations of CBDMB (a pyrazine derivative, 5-(N-4chlorobenzyl)-2',4'-
dimethylbenzamil), a Na+-Ca2+ exchanger blocker on taurine-induced [Ca], increase in embryonie chick heart cells.
Using Fura-2 Ca2+imaging and Fluo-3 Ca2+ confocal microscopy techniques, taurine (20 mM) as expected, induced a sustained
increase in [Ca]j at both the cytosolic and the nuclear levels. Preexposure to 500 JlM of the blocker of the taurine-Na+
cotransporter, ß-alanine, prevented the amino acid-induced increase oftotal [Ca],. On the other hand, application ofß-alanine
did not reverse the action oftaurine on total [Ca]j" However, low concentrations ofthe Na+-Ca2+ exchanger blocker, CBDMB,
reversed the taurine-induced sustained increase of cytosolic and nuclear free calcium (in presence or absence of ß-alanine).
Thus, the effect oftaurine on [Cal, in heart cells appears to be due to Na+ entry through the taurine-Na+ cotransporter which in
turn favours transarcolemmal Ca2+ influx through the Na+-Ca2+ exchanger. (Mol Cell Bioehern 188: 187-197, 1998)

Key words: heart cells, taurine, ß-alanine, taurine-Na+ cotransport, CBDMB, Na+-Ca2+ exchanger, calcium, nuc1eus, confocal

Introduction Insight into the possible physiological roles for taurine


comes from its well-established actions on the heart.
Taurine (2-aminoethane sulphonic acid) is the second most Through the modification of calcium metabolism, taurine
abundant amino acid after glutamate and one of the most has been reported to (a) produce a positive inotropic effect
intriguing amino acids in the body. It is found in very high in heart museIe [9-10], (b) reverse the negative inotropic
concentrations in excitable tissues, particularly in the heart effect ofreduced perfusate calcium or exposure to calcium
where it constitutes more than 50% of the total free amino antagonists [11-14], (c) prevent myocardial necrotic lesions
acid pool [1-5]. associated with calcium overload [15], (d) protect against
Vast data have been accumulating in the broad field of Ca2+ overload in a variety of conditions including cardio-
taurine research. There are numerous reports on the physio- myopathy [16], Ca2+-paradox [17,18], hypoxie injury [19,20],
10gical roles oftaurine in nutrition and development [6] and and isoprenaline toxicity [21], and (e) improve c1inical
on the roles oftaurine in the central nervous system [7], retina symptoms of congestive heart failure [22]. Taurine-deficiency
[8], liver, cardiovascular system and in both skeletal and in the heart was also reported to produce dilated cardio-
smooth musc1es. myopathy [23, 24], electrophysiological abnormalities [25-

Address far offprints: G. Bkaily, Department of Anatomy and Cell Biology, Faculty of Medicine, University of Sherbrooke, Sherbrooke, Quebec, Canada
JlH 5N4
188

26] and myofibrillar loss [27]. Yet, despite the numerous 34]. Later, using the whole-cell voltage clamp technique,
phenomena with which taurine has been associated, most of taurine was found to modulate several ionic currents in
these phenomena have yet to provide a clear understanding myocardial cells ofvarious species [29, 32, 3 4].
ofthe mechanism(s) involved. Taurine was reported to have no effect on Na+/K+-pump
A protective effect of taurine (exposure) against a decline current [31]. The sarcolemmal Na+ICa2+-exchanger has a high
in cardiac slow action potentials during hypoxia has been capacity to transport calcium and is considered the primary
reported [19, 20]. Moreover, it has been proposed that transporter excluding calcium from the cel!. Taurine was
myocardial taurine is one factor implicated in the protection reported to decrease or not to affect the Na+/Ca2+ exchange
against the calcium paradox [17, 18, 28]. current in cardiomyocytes [29, 32, 34--36].
Thus, it seems apparent that taurine plays a role in the Nevertheless, taurine seems to be capable of indirectly
regulation of intracellular calcium homeostasis through modulating Na+/Ca2+-exchanger activity. This indirect effect
modulation of calcium fluxes [29,30]. Taurine has the ability oftaurine was attributed to its ability to alter the phospholipid
to increase calcium availability for contraction and at the microenvironment ofthe Na+/Ca2+-exchanger. Panagiaet al.
same time protect against calcium overIoad injury. This [37] have shown that the Na+/Ca2+- exchanger is very sensitive
ability of taurine is potentially important for the long-term to the phosphatidylethanolamine and phosphatidylcholine
therapeutic application in treating patients with congestive content of the sarcolemmal membrane. In this regard, it is
heart failure. The modulation of myocardial contraction by interesting that taurine-deficient rats following treatment with
taurine has largely been attributed to alterations in calcium the taurine-Na+ cotransporter inhibitor, ß-alanine, exhibit a
transport. Earlier studies examining the effects of taurine on reduction in the capacity ofthe Na+/Ca2+-exchanger, an effect
the heart focused on the sarcolemma as a site oftaurine action consistent with the regulation of sarcolemmal phospholipid N-
and suggested that taurine stimulated myocardial contraction methylation and Na+/Ca2+-exchanger activity by taurine [38].
by enhancing Ca2 +binding to the sarcolemma [12]. However, In addition, taurine is capable of indirecdy modulating Na+1
as additional information became available regarding the Ca2+-exchanger activity by altering intracellular levels of
regulation of calcium homeostasis in the heart, recent studies sodium. Recently, it was suggested that taurine stimulates the
focused on the effects oftaurine on various sarcolemmal Ca2+ fast component of the slow inward Na+ current thereby
transporters. Attempts to clarify the mechanism and site of elevating intracellular Na+ concentrations and promoting Ca2+
taurine action have been thus far inconclusive. influx via the Na+/Ca2+-exchanger [29, 32]. Another mech-
It has been shown that the positive inotropic effect of anism whereby taurine may modulate intracellular Na+ level
taurine is not mediated by an incrcase in cAMP [20], or and thus Na+/Ca2+-cxchangcr activity has also been proposcd:
inhibition ofNa+/K+-ATPase activity [31]. This indicates that as taurine is cotransported with sodium across the cell
the mechanism of action of taurine is different from that of membrane [4, 35], then rapid rates oftaurine influx are thus
well-known positive inotropic agents, such as ß-agonists, associated with sodium entry into the cell, which promote
phosphodiesterase inhibitors and digitalis glycosides [32]. Ca2+ influx via the Na+/Ca2+-exchanger.
Also, the effects induced by taurine are independent of cGMP Hence, this work was undertaken in order to test the
levels and calmodulin-dependent protein kinase activity hypothesis that the taurine-induced increase of[Cal in heart
although taurine may inhibit protein kinase C activity [33]. cells is due to Na+ entry through the taurine-Na+ cotransporter
Published data on taurine effects in cardiac cells using which in turn favours transarcolemmal Ca2+ influx through
conventional intracellular microelectrode recordingtechniques Na+/Ca2+ exchanger.
and contraction recordings suggested that the positive inotropic
effect of taurine is not mediatcd through an increase in the slow
inward Ca2+ current since taurine was reported to exert a Materials and methods
positive inotropic effect without greatly affecting the slow
Ca2+-dependent action potentials (APs) accompanying contrac- Isolation of chick ventricular myocytes
tions in perfused 10-20 day old embryonic chick hearts [11].
Although taurine does not stimulate the slow APs under Ventricular myocytes were obtained from the lower third of
normal conditions, it was reported to exert a stimulant effect the heart of 10-day-old embryonic chicks as described
on guinea pig papillary muscle under hypoxie conditions [11]. previously [29, 30, 39, 40]. Briefly, single cell preparations
Since the function ofthe slow Ca2+ charmel was shown to be were harvested by repeated dispersions in sterile Hank's
dependent on metabolism and is regulated by cyclic nucleo- minimum essential medium (HMEM) containing 0.1 %
tides and phosphorylation [34], the protective/restorative trypsin and 1.8 mM Ca 2+. Following centrifugation and
effect of taurine was suggested to be mediated either in- washing, cells were resuspended in a culture medium con-
directly by stimulation of metabolism or phosphorylation or sisting ofHMEM supplemented with 5% fetal bovine serum
by a direct effect on one or more types of ion channe1s [32, and 50 lU/ml penicillin-G-potassium (Ayerst, Toronto, USA).
189

All other culture solutions were purchased from Gibco (St. scence ratio ofFura-2 in the presence ofsaturating Ca2+, and
Louis, MO, USA). Cultured cells were maintained at 37°C Rmin is the minimal fluorescence ratio in the presence of
in 5% CO2 , 95% air and utilized after 1-24 h in culture. minimal Ca2+. Rand
max
Rmm. were determined at the end of
each experiment using the divalent cation ionophore iono-
mycin (2 x 10-5 M) to permeabilize the cell (R ) and 30 mM
Loading oJFura 2/AMJor microfluorometry of the Ca2+ chelator EGTA (R .). max
mm

Isolated myocytes were cultured on 25 mm glass coverslips


which formed the bottom of the experimental bath chamber Loading o{Fluo-3-AMJor conJocal microscopy
[30, 31, 40]. Cells were loaded with the fluorescent ratiometric
calcium indicator Fura-2/AM (Calbiochem, La Jolla, CA, Cells were loaded according to the method described pre-
USA) according to the method described previously [7, 8]. viously [41]. Isolated ventricular myocytes were cultured and
Prior to loading, coverslips were washed 3 times in Tyrode's mounted in the same manner as cells for Fura-2/AM measure-
solution (Sigma, St. Louis, MO, USA) containing 10 mM ments. Frozen stocks ofFlu03-AM (Molecular Probes OR)
NaHC0 3, 5 mM HEPES, 1.8 mM CaCI2• 1 mM MgClz' 2.7 mM wcre reconstituted in DMSO and diluted to a final con-
KCl, 137mMNaCl, 3.6mMNaHl04 and5.5 mMD-glucose centration of 13.5 11M in Tyrode's-BSA. The cells were
(buffered to pH 7.4 with Tris base). Thc osmolarity ofthc incubated for 45 min at room temperature, washed, and
Tyrode's solution with or without 0.1 % bovine serum further incubated for 15 min at room temperature to complete
albumin (BSA) was adjusted to 310 mOsM wit,h sucrose. hydrolysis of acetoxymethyl ester groups.
Myocytes were incubated with freshly prepared Fura-2/AM
(111M final concentration in Tyrode-BSA solution) for 30
min at 28°C [40]. Stock solutions ofthe probe were prepared Ca 2+ imaging using conJocal microscopy
by addition of DMSO to 1 mM frozen aliquots. After
loading, the cells were washed with Tyrode's solution and ConJocal microscopy
incubated for another 30 min at 28°C in order to ensure Fluo-3 loaded cells were examined with a Molecular Dynarnics
complete hydrolysis of acetoxyinethyl ester groups. The (Surmyvale, CA, USA) Multi Probe 2001 confocal argon
cells were again washed in Tyrode' s solution prior to micro- laser scanning (CSLM) system equipped with a Nikon
fluorometric measurements. The N a+-Ca2+exchanger blocker, Diaphot epifluorescence inverted microscope and a 60 x (1.4
CBDMB was reported to interfere with Fura-2 [40] but it NA) Nikon Oil Plan achromat objective. The 488 nm argon
is unlikely to interfere with the long wave-length excitation laser line (9.0 mV) was directed to the sampIe via a 510 nm
fluorescent dye, Fluo-3. Thus, the effect ofCBDMB (10 primary dichroie filter and attenuated with a 3% neutral
11M) was tested in intact single heart cells using Ca 2 + density filter to reduce photobleaching. Pinhole size was set
imaging technique. At this concentration, CBDMB did not at 100 11m. The image size was 512 x 512 pixels with a pixel
affect at a11 the intrace11ular Ca2+ fluorescence of Fura-2 size of 0 .11 11m. Laser line intensity, photometrie gain, PMT
signal in intact heart cells. Thus, this compound at 10 11M settings and filter attenuation were kept constant throughout
could be used without any major problem in intact ventricular the experimental procedures [29, 30,42].
heart cells loaded with Fura-2.

Calcium fluorescence studies


Microfluorometric measurements
Changes in cytosolic and intranuclear calcium fluorescence
In order to monitor cytosolic intracellular Ca2+ changes in upon addition of increasing concentrations of taurine to thc
chick myocytes, fluorescence measurements were made external Tyrode's medium was measured in Fluo-3 (for Ca2+)
following the dual excitatory wavelength method (340/380 loaded myocytes [42]. For short-term treatment with taurine,
ratio) using a Deltascan and Imagescan microfluorometer cells were scanned prior to and after addition of taurine to
(Photon Technology International Inc., Princeton, NJ, USA) monitor cell response to the drug. Serial optical scans were
equipped with a NEC Power Mate 386/20 and an accom- performed 2-10 min after addition of each taurine con-
panying PTI software enabling instrument control, data centration. A total of 12-15 scans (512 x 512) were per-
acquisition and analysis [41]. Calculation of cytosolic Ca2+ formed for each series with a step size of 0.8-1.0 11m,
was performed using the standard equation included in the although the number of seetions and step size were rigorously
software [41], in which [Ca]i =~ (R - RminlRmax -R), where maintained during the course of each experiment in order to
the ~ for Fura-2 is 224 nM, R is the experimental fluore- 10calize calcium variations within the cytosol and the boun-
scence ratio (340/3 80) value, R max is the maximal fluore- daries ofthe nucleus.
190

Nuclear staining P values less than 0.05 were considered as significant. The n
value refers to the number of different cen preparations.
At the end of each experiment, the nucleus was stained with
100 nM of live cell nucleic acid stain Syto 11 (Molecular
Probes, Oregon, USA) according to the method described Results
previously [42]. In brief, serial optical scans were taken
immediately after development ofthe stain (approximately Previous work has shown that the actions oftaurine on the
8-10 min) while maintaining positioning, number of sections heart are associated with the regulation of intracellular
and step size identical to that used for calcium staining. 3-D calcium homeostasis through the modulation of calcium and
reconstructions ofthe nucleus [42] were performed through sodium fluxes [29, 30]. Attempts to determine the taurine-
volume rendering and used as templates to delineate nuclear induced the effects as weil as their mechanisms on [Ca] i and
from cytosolic free ion [29, 30, 42). [Na). in heart cells were mainly directed towards measuring
the effect oftaurine on both Ca2+ and Na+ ionic currents in
heart cells [32,34). These studies were almost all done using
Valurne rendering and nuclear calcium measurements the whole-cell voltage-clamp technique. More importantly,
these studies on taurine action on ionic currents reveal that the
Scanned images were transferred onto a Silicon Graphics
action of taurine on Ca'+ and Na+ channeIs is complex;
Indy 4000 workstation equipped with Molecular Dynamics'
however, they do not correlate between taurine-induced
Imagespace analysis and volume workbench software
multiple effects in heart cells and thus do not provide a cIear
modules. Reconstruction of3-D images were performed on
understanding ofthc positivc inotropic cffcct oftaurine nor its
Gaussian-filtered serial sections and are represented as cIosest
cardioprotective effect against intracellular Ca2+overload.
intensity projections for calcium distribution and look-
Recently, taurine was reported to modulate cytosolic and
through extended-focus projections for both nucleus and
nuclear [Ca]i and [Na]i in heart cens depending on time of
calcium co-Iocalisation studies [42]. Images are represented
exposure in the heart [29, 30]. It was suggested that the
as pseudo-coloured representations according to an intensity
increase in cytosolic and nucIear Na+ by short-term exposure
scale of 0-255 with lowest intensity in black and highest
to taurine could be due to Na+ influx through the taurine-Na+
intensity in white (Fig. 2).
cotransporter at the sarcolemmal membrane and probably
Measurement of calcium uptake within the nucIeus was
also at the nuclear membrane [29, 30). However, the increase
performed on 3-D reconstructs (section series). The nucIear
in [Ca], by taurine was suggested to be due to an increase in
area following Syto 11 staining was isolated from the rest of
[Na], which favors Ca'+ influx via the Na+-Ca2+exchanger
the cell by setting a lower intensity threshold filter to confine
[30]. Since there are no studies showing the effect of ablocker
relevant pixels. A 3-D binary image series of thc nuclcar
ofthe taurine-Na+ cotransporter on taurine-induced sustained
volume was then generated for each celI using the exact
increase of[Ca]i' we measured ina first series ofexperiments
same x, y and z set planes as those used during calcium
the dose-dependent short-term effect of extracellular taurine
uptake. By then applying these binary image patterns to the
(1--80 mM), on cytosolic [Ca). in single I O-day-old embryonic
same cen but labeled for calcium (the binary image serves
chick heart cells.
as a 'cookie cutter'), a new 3-D projection was created
To perform these measurements, we tumed to Ca2+fluore-
depicting fluorescence intensity levels excIusively within
scence measurement techniques that render it possible to
the nucleus. Hence, by 'removing' the nucleus from the
quantitatively and qualitatively measure intracellular levels
surrounding cytoplasm, we were then able to measure mean
of free Ca'+ in single cells, as weil as to determine the
calcium intensity values in the entire nuclear volume while
localization and distribution of these ions in isolated single
eliminating possible contribution of perinuclear calcium to our
cells [41, 42--44). The techniques employed were Fura-2
mcasurements [42).
Ca2+ fluorescence digital imaging of cytosolic [Cal and
three-dimensional cytosolic and nuclear Ca 2+ fluorcscencc
Statistics measurcmcnts using confocal microscopy [41, 42--44). Due
to its specific localisation within thc cytosol, the Fura-2 Ca2+
Intranuclear free calcium intensity levels are represented fluorescence digital imaging techniquc was used to monitor
either as mean fluorescence intensity values or as the per- cytosolic [Ca]i changes in single isolatcd heart cells. Fura-2/
centage of increase relative to control levels. Values are AM was also chosen due to its high affinity for Ca'+ and its
expressed as means ± S.E.M. Statistical significance was ratiometric property that permits more accurate determination
determined using the analysis of variance (ANOVA) test of cytosolic Ca'+ concentration. Another reason was its ease
followed by either a Tukey-Kramer or Dunnett's multiple of loading in the preparations and under the experimental
comparison test to assess statistical significance of the results. conditions used. On the other hand, Fluo-3 Ca'+ fluorescence
191

confocal imaging enables the measurement and visualization acting on Ca2+ and Na+ channels or transporters from both
of spatial changes in Ca2+fluorescence within both the cytosol extracellular and intracellu1ar surfaces of the plasma mem-
and the nucleus [29, 30, 42--45]. brane [29, 32].
Very recently it has been demonstrated by our laboratory Our recent [29, 30] and present results reveal that short-
that various factors that modulate [Ca]j in heart cells act via term exposure to taurine induces a steady-state increase in
modulating cytosolic and/or nuclear Ca2+ distribution levels both cytosolic and nuclear [Ca ]j' Thus, it is possible that the
[29, 30, 42--45]. Thus, to determine whether taurine may increase in cytoso1ic [Ca]j due to extracellular short-term
modulate the distribution of cytosolic and nuclear Ca 2+ in exposure to taurine could be due to sarcolemmal Na+ influx
heart cells, three-dimensional (3-D) measurement of Ca2+ through the taurine-Na+ cotransporter and this increase in
levels were also performed using Fluo-3. cytosolic [Nal would in turn induce Na+ outflow and Ca'+
influx through the sarcolemmal Na+ICa2+-exchanger [29, 30].
Thus, in the next series of experiments, we set out to de-
Short-term effect oftaurine on cytosolic [Ca], in iso la ted termine whether the short-term taurine-induced effect on total
ventricular heart cells using Fura-2 Ca 2+jluorescence [Ca l (cytosolic and nuclear) in single isolated heart cells is
microjluorometry indirect1y due to Na+ influx through the sarcolemmal taurine-
Na+ cotransporter. Forthis purpose, using Fluo-3 3D confocal
In the first series of experiments, Fura-2 Ca2+ fluorescence microscopy and Fura-2 2D Ca2+measurement techniques, the
microfluorometry imaging was used to measure the con- effect ofß-alanine, ablocker oftaurine-Na+ cotransporter [4,
centration-dependent short-term effect of taurine on cyto- 43] was tested on the short-term effect oftaurine on total [Cal
solic [Cal in isolated ventricular heart cells. In this series in single heart cells [29, 30].
of experiments, taurine (1-80 mM) was added sequentially Isolated heart cells were exposed to a single concentration
to the extracellular milieu after a steady-state period was of ß-alanine (500 11M) known to block the taurine-Na+
reached, and changes in cytosolic [Cal were monitored. cotransporter in many cell types [4,6,43] for aperiod of 10-
Figure 1 shows a two-dimensional view of two isolated 20 min after inducing an increase in steady-state cytosolic and
ventricular myocytes illustrating the effect of short-term nuclear [Ca L by 20 mM of taurine using Fura-2 micro-
exposure to various concentrations of taurine on cytosolic fluorometry (Fig. IG-I) and Fluo-3 confocal microscopy (Fig.
[Ca]j measured using Fura-2 Ca2+fluorescence imaging. As 3A-E). As seen in Figs lD, IH and 3A-C, short-term
can be seen in Fig. I A-F, thc application of cxtraccllular cxposurc of isolated vcntricular hcart cells to 20 mM extra-
taurine induced in a concentration-dependent manner an cellular taurine induced an increase in steady-state cytosolic
increase in steady-state cytosolic [Ca], when compared to [Ca]j (from 41.1 ± 0.9 up to 138.8 ± 7.3 nM, n = 9) in these
, basal resting cytosolic [Ca]e As seen in Fig. lA and G, the cells (when compared to controllevels) and in total (cytosolic
mean basal steady-state cytosolic [Cal ofthese two single and nuclear) intracellular mean relative Ca 2+ fluorescence
ventricular 10-day-old chick embryonie heart cells was intensity (from 68.2 ± 8 up to 97.6 ± 10.7, n =4, Fig. 3A and
approximately 50 nM. The application oflow concentrations B). In the presence of 20 mM taurine, addition of ß-alanine
oftaurine (1 mM), near that of normal circulating taurine levels, (500 11M), ablocker of taurine-Na+ cotransporter did not
increased cytosolic [Ca], by 10 nM within 10 min (Fig. IB). reverse the taurine increase of steady-state cytosolic [Cal
Taurine at 10 mM further increased the sustained level of (using cytosolic Fura-2 techniques, Fig. 1 H-I) and total [Ca],
cytosolic [Ca]j by 13 nM (Fig. lC). Further increases of (cytosolic and nuclear using Fluo-3 measurement, Fig. 3B-C).
extracellular taurine up to 20, 40 and 80 mM induced in a Like for20 mM oftaurine, ß-alanine did not block the 40 mM
concentration-dependent manner a further rise in the sustained taurine-induced sustained increase in [CaL (Fig. 4A~).
level of cytosolic [Ca]j (Fig. lD-F) (up to 202 nM [Ca]j with Thus, ß-alanine had no effect on total [Ca]j when heart cells
the addition of 80 mM taurine). These results are similar to were preexposed to taurine.
those recently reported in the same cell types [29, 30]. In another series of experiments we verified the effect of
preexposure to ß-alanine (500 11M) on short-term taurine-
induced increase of [CaL and Fig. 5 surnmarizes the results.
Effect of ß-alanine, ablocker of taurine- Na+ As seen in this figure, by contrast, when single heart cells
cotransporter, on the short-term taurine-induced sustained were preexposed to the taurine-Na+ cotransporter blocker,
increase of total [Ca], in single heart cel/s short-term exposure to high concentrations oftaurine (40 and
80 mM) failed to induce a sustained steady-state increase in
It has been shown that extracellular taurine is cotransported [Ca). in isolated ventricular heart cells (Fig. 5A-D). Thus,
with Na+ across the plasma membrane via a specific carrier ß-al~nine, ablocker of taurine-Na+ symporter, could only
that can be blocked by ß-alanine [4, 6]. Taurine was also prevent the short-term concentration-dependent taurine-
reported to induce its multiple effects on [Cal and [Na]j by induced increase in steady-state [Ca]j in single heart cells
A • CONTROl B • TAURINE 1 mM C· TAURINE 10 mM
-
521

:au.
10 min . 10 min.
'90

~~ 117
o. TAURINE 20 mM E • TAURINE 40 mM F • TAUR INE 80 mM
71.7

. . ..
4 • . 8

15 min. 15 mln. 10 mln. 12 . 3

4 . "7
CHICK HEART CELl. OCP 1898
3 . 40

G • CONTROl H • TAUR INE 20 mM I . j} - AlANINE 500 ~M

1187

52'

3'~
20 mln. 20 min
'90

117
J • CBDMB IOI'M K - CBDMB 10l'M L - CBDMB IOI'M
?t.S

.1 .•

-
in . ?

1 min. 5 min 10 m ln. 'e .e

CH ICK HEART CEL l # OCP 2078


..
3 . 40
c-.c.
lC'O"'-tI

A B c

- Control Taurine
20mM
+
CBDMB
10 uM

o 255
193

150 A B c o E F
~
'in
cQ)
***
~ 120 **
Q)
(J
cQ)
(J

'"0
Q)
'-
90
:J
I;::
+
N

u
<11 60
Q)
.~
(ij (4) (4)
Q)
'- 30
C
<11
Q)
::
0
Control Taurine Taurine Control Taurine Taurine
(20 mM) + +
ß-alanlne CBDMB
(500 11M) (10!-,M)

Fig. 3. Effect of ~-alanine, a taurine-Na' cotransporter blocker and CBDMB, a Na'/Ca" exchanger blocker, on short-tenn induced effect of20 mM taurine
on total intracellular Ca" levels in isolated ventricular heart cells measured by 3-D Fluo-3 confoeal microseopy. (A) Basal relative intracellular Ca"
fluorescence in single heart cells. (B and C) Taurine at 20 mM induced a significant increase in steady-state total intracellular Ca" fluorescenee intensity that
was not affected by the addition of~-alanine (500 11M) to the extracellular medium. (D and E) Taurine at 20 mM induced a significant increase in steady-state
total Ca" fluorescence intensity. (F) Application of 10 11M CBDMB significantly attenuated the increase in total intracellular Ca" fluorescence level
induced by taurine down to the controllevel. The data are expressed as mean relative Ca" fluorescence intensity. Values are given as means ± S.E.M. witb
the number of experiments indicated in brackets. "p < 0.01 and "'p < 0.001. The numerical va lues in A- F are respectively: 68.2 ± 8, 97.5 ± 10.7, 110.9
± 10.3, 70.4 ± 10.2, 103.2 ± 8.2 and 76.3 ± 8.4.

Fig. J. Two-dimcnsional view of isolated ventrieular myoeytes illustrating the dose-dependent short-tenn effect of taurine on eytosolie [Ca]; (A-D) and the
effeet of ß-alanine and CBDMB (H-L) on short-term taurine-induced increase in [Ca);. (A and G) Steady-state basal resting [Cal; in 10-day-old embryonie
ehiek heart eells using Fura-2 Ca" fluorcscencc imaging measurement. Addition of normal coneentrations oftaurine (1 mM) increased within 10 min the
steady-state basal [Ca), from 37.2-47.0 nM. (B-F) Inereasing the taurine concentration to 10, 20, 40 and 80 mM indueed in a eoneentration-dependent
manner a highly visible inerease in steady-state basal [Ca); within 10-15 min after applieation oftaurine (respeetively 59.6, 82.3 , 120 and 202 nM). (H) 20
mM 01' taurine-indueed a large increase in steady-state [Ca); in both heart eells. (I) Application of 500 11M ß-alanine did not reverse the [Ca); inerease
induced by 20 mM taurine. (J-L) Addition of CBDMB at 10 11M reversed the taurine-indueed [Ca); inerease as early as Imin, with further attenuation
reaehed after 10min. The time indieated below eaeh image represents the aetual time elapsed between addition of the drug and the recording of the image.
The vertical colored bar represents calibration scales for [Ca] from 3.4--3246 nM. The substanees were added sequentially after a lengthy steady-state effeet
was reaehcd. DCPI898 and DCP2078 are the identifieation numbers ofthe experiment.

Fig.2. Sagittal and cross-seetional view ofa 3-D reeonstructed isolated ventricular myocyte illustrating the effecl ofCBDMB on taurine-induced increase
in total intracellular Ca" level. (A) Basal resting Fluo-3 Ca" fluorescence intensity distribution in 10-day-old embryonie chick beart cell measured using
Fluo-3 3-D confocal microscopy. (B) Exposure to 10 mM taurine added to the extracellular medium for 10 min induced a highly visible increase in
steady-state Fluo-3 Ca" fluorescence intensity when compared to basal. (C) Application oflow concenlralion ofCBDMB, a speeific Na'ICa"-exchanger
blocker (10 !lM), reversed the taurine-induced increase in total Fluo-3 Ca" fluorescence intensity down to basal controllevels within 15 min. The white
scale bar is in !lm. The color scale below represents pseudocolor intensity levels ofFluo-3 Ca" fluorescence from 0-255. Cell number is CP293-Fluo-2.
194

450 A B C 0 E F G
***
.[1,
375

......... 300
:2
c
.......,
..-.
+ 225
N
cu
~
150

75

0 Control Taurine +
+ Contn;l: Taurir,e +
(40 mM) f3-alanine CBDMB (40 mM) CBDMB
(500I1M) (10 11M) (10 11M)

Hg. 4. A-G Effect of f3-alanine, a taurine-Na+ cotransporter blocker and CBOMB, a Na+ICa'+ exchanger blocker, on short-term taurine-induced effect on
cytosolic [Cal, in isolated ventricular heart cells measured by Fura-2 Ca'+ imaging. (A) Mean basal cytosolic [Cal, in lO-day-old embryonie chick heart cells
(41.1 ± 0.9 nM, n = 9). (B) High concentrations oftaurine (40 mM) induced a significant increase in steady-state cytosolic [Cal (up to 225.7 ± 20.2 nM, n
= 9). (C) Application off3-alanine (500 ~M) to the extracellular milieu did not block the taurine-induced increase in steady-state [Cal, in single heart cells (n
= 9). (0) Application ofCBOMB (10 11ffi) in the presence of f3-alanine significantly reversed the taurine-induced increase in [Cal, (0 vs. Cl, but at a level
still abovethe control (167.3 ± 6.9 nM, n = 9). (E-G) Application oftaurine CBOMB (10~) in the absence off3-alanine had the same effect on taurine-induced
increase in steady-state [Cal,. The results in terms of intracellular [Cal. values are given as mean ± S.E.M. with the number of experiments indicated in
brackets. 'p < 0.05, "'p < 0.001.

when cells were preexposed to it prior to the addition of taurine-induced [Cal- Thus, CBDMB (a pyrazine derivative),
extracellular taurine, whereas ß-alanine did not block the was synthesized specifically for this study by a previously
taurine-Na+ symporter in a therapeutic manner. described method [45] and in this series of experiments we
determined the effect of a low concentration of this N a+ICa2+-
exchanger blocker (10 11M), on 20 (Figs IG--L and 3D-F)
Effect ofthe Na+ICa2+-exchanger blocker, CBDMB (a and40 mM (Fig. 4A-G) taurine-induced increase in [Ca]j in
pyrazine derivative), on short-term induced effect of single heart cells using Fura-2 Ca2+ fluorescence imaging
taurine on [Cal; in single heart cells (Figs IG-L and 4A-G) as weH as 3-D Fluo-3 Ca2+ fluore-
scence confocal microscopy (Figs 2 and 3D-F) in absence
We have recently shown that short-term exposure to high (Figs 3D-F and 4E-G) or presence (Figs I G-L and 4A-D)
concentrations oftaurine (20 to 40 mM) induced a significant ofthe taurine-Na+ cotransporter blocker, ß-alanine.
increase in both total intracellular Na+ and Ca2+ levels in Similar to the results obtained using Fura-2 microfluoro-
single heart cells using Na+-green and Fluo-3 confocal metry and Fluo-3 confocal microscopy to detennine the short-
microscopy technique [29, 30]. Our present results also term induced effect oftaurine on [Cal; in isolated heart cells
showed that when cells were preexposed to ß-alanine, a [30], high concentrations oftaurine (10-40 mM) were found
blocker oftaurine- Na+ cotransport, the addition of even high to induce a significant increase in steady-state total [Ca1when
concentrations of taurine did not induce any expected effect compared to controllevels (Figs 1-5). As can be seen in Figs
on [Cal; in single heart cells. 1fthis increase in [Cal; induced 2 and 3D-F, using Fluo-3 Ca2+measurement technique, and
by taurine is due to an increase in [Na]j which in turn induces Fig. 4E-G using Fura-2 techniques, superfusion with high
Ca2+ influx through the Na+/Ca2+-exchanger, then ablocker concentrations oftaurine (20-40 mM) induced, as expected,
of this exchanger should inhibit the increase in short-term a sustained increase in [Ca1. within 10 min while the addition
195

120 A B c o

90
..-..
~
c
.........
,.....:..-
+ 60
N
co (7)
~

30

o + +
ß-alanine
Control Taurire Ta'Jrine
(500 JiM) (40 mM) (80 mM)

Fig.5. Effect ofpretreatment with ß-alanine, ablocker oft"aurine-Na+ cotransporter, on short-term taurine-induced effect on [Cal; in single heart cells. (A)
Basal [Cal; in single lO-day-old chick embryonie heart cells measured by Fura-2 microfluorometry (88.6 ± 3.3 nM). (B) Preexposure of heart cells to
ß-alanine (500 ~M) for 10--20 min had no effect on basal [Cal; (90.9 ± 3.7 nM). (C-D) High eoncentrations oftaurine (40 and 80 mM) also had no effeet on
steady-state total [Cal, when single heart cells were preexposed to ß-alanine (C = 9I.l ± 4.3 and D = 90.6 ± 3.3 nM). The resuIts are expressed in terms of
[Ca I;. Values are expressed as means ± S.E.M. nM. The number of experiments is indicated in brackets.

ofthe Na+(Ca2+-exchanger blocker, CBDMB, (10 I!M), was return [Cal to the near controllevel in the presence of 40 mM
found to significantly (p < 0.05) decrease this short-term taurine (Fig. 4D). Thus, using 2-D Fura-2 Ca2+ imaging and
taurine-induced increase in steady-state [Cal to ne ar the 3-D Fluo-3 confocal microscopy, the application ofthe Na+(
controllevel (Figs 2, 3D-F, 4E--G). As summarized in Fig. Ca 2+-exchanger blocker, CBDMB, blocked the taurine-
3D-E, 20 mM taurine increased the mean relative Ca 2+ induced steady-state increase in total [Cal and this effect of
fluorescence intensity from 70.4 ± 10.2 up to 103.2 ± 8.2 (n the exchanger blocker was found to be dependent mainly on
= 4) and CBDMB (10 I!M) reduced [CaJ i down to near the extracellular taurine concentrations.
controllevel (76.3 ± 8.4; Fig. 3F). However, increasing the
concentration of taurine up to 40 mM (n =6) increased [Cal
from 30.0 ± 0.8 up to 180.3 ± 18.7 nM and 10 I!M ofthe Discussion
exchanger blocker decreased [Cal to 128.2 ± 33.9 nM (Fig.
4E--G). This decrease ofthe 40 nM taurine-induced increase Earlier studies revealed that in single 3-, 10-, and 17 -day-old
of [Cal by the exchanger blocker was not significantly embryonic chick heart cells in culture, short-term exposure
different from control, but was significant (p < 0.05) when to taurine (5 mM) activated two transient inward voltage
compared to the level of [Cal induced by this very high dependent Ca2+ and Na+ currents [29, 32, 34]. These earlier
concentration ofthe amino acid (Fig. 4E--G). studies showed that taurine blocks the L-type Ca 2+ channel
In the same series of experiments, whereas ß-alanine failed whereas it activates the T-type Ca2+ channel [29]. Recently,
to reverse the sustained increase of [Ca]i induced by 20 mM short-term exposure of heart cells to taurine was reported to
(Fig. I G-I) and 40 mM (Fig. 4A-D) taurine, in the presence induce a dose-dependent sustained increase in [Ca]i [30]. This
ofthe taurine-Na+ cotransporter blocker, addition ofthe Na+( sustained increase in [Cal could accelerate the inactivation of
Ca2+-exchanger blocker CBDMB (10 I!M) significantly (p < the L-type Ca2+channel in working heart cells, thus explaining
0.01) decreased taurine-induced increase of[Cal (trom 170.3 in part the observed decrease ofthis type of current by taurine.
± 16.5 down to 111.2 ± 8.6, n = 9); however, as in the absence Similar to Ca2+, short-terrn exposure to high concentrations
of the taurine-Na+ cotransporter blocker, CBDMB did not (compared to normal 1 mM circulating taurine) of taurine (20
196

mM) was reported to induce a sustained increase in basal sustained free Ca2+ could be due to cytosolic Ca2+bulfering
[Nal. in quiescent single ventricular heart cells [30]. This by the nuc1eus [29, 30, 42--45] as weil as possibly to Ca2+
sustained increase ofbasal [Na]j in quiescent heart cells can influx through nuc1ear membrane Na+-Ca2+ exchanger in a
not be due to activation oftransient influx through a voltage- condition where nuclear Na+ overload is present [30].
dependent Na+ channel, but was suggested to be due to resting In summary, the fact that ß-alanine prevents taurine from
potential Na+ influx via the taurine-Na+ cotransporter [30]. inducing a sustained increase of [Cal. suggests that the
Our results show that short-term exposure (10--20 min) taurine-Na+ co transporter is present in 10-day-old chick
to normal physiological concentration oftaurine (1 mM) had embryonic heart cells and this pathway may contribute to
no significant effect on the steady-state basal intracellular cytosolic and nuc1ear Na+ homeostasis in heart muscle cells.
[Ca] in quiescent isolated ventricular heart cells, as mea- The fact that CBDMB blocks the taurine-induced sustained
sured by Fura-2 microfluorometry. The same results were increase of[Cal. highly suggests that this increase is mainly
also confirmed using Fluo-3 Ca2+three-dimensional confocal due to an elevation ofbasal resting [Na]. which in turn favors
microscopy. However, short-term exposure to increasing con- Ca2+ influx through the Na+ICa2+-exch~nger.
centrations of taurine added extracellularly from 10--80 mM
induced a significant concentration-dependent increase in
steady-state basal total [Ca]j of quiescent ventricular heart Acknowledgements
cells. These results are in accordance with those recently
reported results by our group [29, 30].
This study was supported by the Medical Research Council
It was suggested that short-term exposure to relatively high
ofCanada, grant no. MT-I2882 to Dr. Bkaily. Miss Jaalouk
concentrations oftaurine actually elevated the total sustained
is a Ph.D. Fellow ofthe Canadian International Development
basal [Ca]j presumably resulting from the stimulation ofthe
Agency (CIDA). The authors thank Ms. Susann Topping for
steady-state voltage dependent resting (R)-type Ca2+channel
her secretarial assistance.
[29,30] as weil as the Na+ICa2+-exchanger in noncontracting
resting ventricular heart cells, and to stimulation of Ca2+influx
through the T- and R-type Ca2+ channels and the Na+ICa2+-
exchanger in working heart cells [30]. Thus, short-term References
exposure to taurine was suggested to exert a positive inotropic
effect by such mechanisms [29, 30]. However, our finding 1. Jacohsen J, Smith LH: Biochemistry and physiology of taurine and
that pretrcatment with ß-alanine, ablocker of taurine-Na+ taurine derivatives. Physiol Rev 48: 427-511,1968
2. Pcterson MB, Mead RJ, Welty JD: Free amino acids in congestive
cotransporter, prevented (but did not reverse) taurine-induced
heart failure. J Mol Cen Cardiol5: 139-147,1973
dose-dependent increase in [Cal. in quiescent ventricular 3. Baskin SI, Finney CM: Effects of taurine and taurine analogues on
heart cells and that CBD MB (a specific blocker the N a+ICa 2+- the cardiovascular system. Ganryu Aminosan (Sulfur-containing
exchanger), blocked the increase in basal [Cal. induced by Amino Acids) 2: 1-18, 1979
short-term treatment with taurine leads us to conclude that 4. Suleiman MS, ChapmanRA: Changes in the principal free intracellular
amino acids in the Langerdoffperfused guinea pig heart during arrest
in quiescent and working ventricular heart cells: (1) the
with calcium-free or high potassium media. Cardiovas Res 27: 1810-
sustained increase ofbasal resting total [Nal. by short-term 1814,1993
treatment with taurine is largely due to Na+ entry through the 5. Michalk DV, Tittor F, Ringeisen P, Deeg KR Bohles H: The deve1op-
taurine-Na+ cotransporter [29, 30,44]. (2) the increase in ment of heart and brain function in low-birth weight infants red with
basal resting steady-state [Ca]. could then be due to an taurine-supplemented formula. Adv Exp Med Bio1217: 139-145, 1987
6. Huxtable RJ, Sebring LA: Towards a unifying theory for the actions
increase in [Na]. which in turn fav~rs sarcolcmmal Ca2+influx
oftaurine. Trends Pharmaeol Sei 7: 481-485, 1986
through the N~+ICa2+-exchanger [29, 30] and exc1udes a 7. Michalk DV, Ringeisen P, Tittor F, Lanffer H, Deeg KH, Bohles H:
possible taurine activation of resting potential Ca2+influx via Development of the nervous and cardiovaseular systems in low-birth
stimulation of the sarcolemmal voltage dependent R-type weight infant, red a taurine-supplemented formula. Eur J Pediatrics
Ca 2+ channel [29, 30]. 147: 296-299, 1988
8. Lima L, Matus P, Drujan B: The trophic role oftaurine in the retina.A
Using 3-D confocal Ca2+ and Na+ fluorescence measure-
possible mechanism of action. Adv Exp Med Bio1315: 287-294, 1992
ments, recent results showed that the increase ofbasal resting 9. Dietrich J, Diacono J: Comparison between ouabain and taurine effects
intracellular free Ca2+ and Na+ by short-term treatment with on isolated rat and guinea pig hearts in low calcium medium. Life Sci
high concentration of taurine (~ 5 mM) was mainly nuclear 10: 499-507, 1971
[29, 30]. The increase ofbasal resting nuclear sustained free 10. Schaffer SW, Chovan JP, Werkman RF: Dissociation ofcAMP changes
and myocardial contractility in taurine perfused rat heart. Biochem
Na+ by taurine could be due to cytosolic Na+ buffering by the
Biophys Res Commun 81: 248-253,1978
nuc1eus (via an unknown mechanism) and/or to possible 11. Sawamura A, Sperelakis N, Azuma J: Protective effect of taurine
presence of a taurine-Na+ cotransporter on the nuclear against decline of cardiac slow action potentials during hypoxia. Eur
membrane [30]. However, the observed increase of nuclear J Pharmacol120: 235-239, 1986
197

12. Chovan JP, Kulakowski EC, Sheakowski S, Schaffer SW: Calcium R. Huxtable,A. Barbeau (cds). Taurine. RavenPress, NewYork, 1976,
regulation by the low-affinity taurine binding sites of eardiac pp 121-134
sarcolemma. Mol Pharmacol 17: 295-300, 1980 32. Sperelakis N, Katsube Y, Kusaka M: Some actions oftaurine on ionic
13. Franeoni F, Martini F, Stenardi I, Matucci R, Ziletti L, Giotti A: Effeet eurrents of myoeardial cells and myometrial eells. Adv Exp Med Biol
of taurine on calcium level and eontraetility in guinea pig ventrieular 403: 275-284,1996
strips. Bioehern Pharmaeol31: 3181-3186, 1982 33. Lombardini JB: The inhibitory effeets of taurine on protein phos-
14. Franeoni F, Stenardi I, Martini F, Ziletti L, GiottiA: Interaetion between phorylation: Comparison of various eharaeteristies of the taurine
organic calcium channel blockers and taurine in vitra and in viva. J affeeted phosphoproteins present in rat retina, brain and heart. Adv
Pharm Pharmacol 34: 329-330, 1982 Exp Med Bio1359: 9-17,1994
15. Popovieh MJ, Kobets VA, Kostin SI, Kapc\ko VI: Protective effect of 34. Bkaily G, Perron N, Wang S, Seulptoreanu A, Jaeques 0, Menard 0:
taurine on the myocardial effects of prolonged treatment with Atrial natriuretie faetor blocks the high-threshold Ca2+ CUITent and
norepinephrine in rats. Cardiosci 31: 61-{i6, 1992 increasedK+ eurrent in fetal single ventrieulareells. J Mol Cell Cardiol
16. McBroom MJ, Welty JD: Effect of taurine on heart calcium in the 25: 1305-1316. 1993
eardiomyopathie hamster. J Mol Cell Cardiol9: 853-859, 1977 35. Schaffer SW, Ballard C, Azuma J: Mechanisms underlying physio-
17. Krarner JH, Chovan JP, Schaffer SW: The effect oftaurine on calcium logieal and pharmaeologieal aetions of taudne on myocardial calcium
paradox and isehernic heart failure. Am JPhysiol240: H238-H246, 1981 transport. Adv Exp Med Biol 359: 171-180, 1994
18. Takihara K, Azuma J, Kishimoto S, ünishi S, Sperelakis N: Taurine 36. Earm YE, Ho WX, So IS, Lcen CH: Effect oftaurine on the aetivation
prevention of calcium paradox-related damage in cardiac musele. of background eurrcnt in eardiae myoeytes. In: YE. Earm, D. Noble
Bioehern Pharmaeol37: 2651-2658, 1988 (eds). Cardiac Ion Channels and Effects of Taurine on the Hearl.
19. Franeoni F, Stenardi I, Failli P: The protective effects of taurine on Kluwer Academie Publishers, Boston, 1992
hypoxia and reoxygenation in guinea-pig heart. Bioehem Pharmaeol 37. Panagia V, Makimo N, Ganguly PK, Dhalla NS: Inhibition of Na+/
34: 2611-2615, 1985 Ca2+-exehange in heart sareolemmal vesieles by phosphatidyl-
20. SawamuraA, Sperelakis N,Azuma J, Kishimoto S: Effects oftaurine ethanolamine N-methylation. Eur J Biochem 166: 597-{i03, 1987
on the eleetrieal and rneehanieal aetivities of embryonie ehiek heart. 38. Harada H, Allo S, Viyuoh N, Azurna J, Takahashi K, Schaffer SW:
Can J Physiol Pharmaeol 64: 649-{i55, 1986 Regulation of calcium transport in drug-indueed taurine-depleted
21. ühta H, Junichi N, Awata N: Meehanism of the proteetive effeet of hearts. Biochem Biophys Acta 944: 273-278, 1988
taurine against isoprenaline indueed myoeardial damage. Cardiovas 39. Bkaily G: Single heart cells as model tor studying cardiae toxieology.
Res 22: 407-413,1988 In: G. Jolles, A. Cordier (cds). In Vitra Methods in Toxieology.
22. Azuma J: Long-term effeet of taurine in eongestive heart failure: Academie Press, London, 1992, pp 289-334
Preliminary report. Adv Exp Med Biol 359: 425-433, 1994 40. Kraut RP, Greenberg AR Cragoc EJ Jr, Bose R: Pyrazine eompounds
23. Pion PD, Kittleson MD, Skiles ML, Rogers QR, Morris JG: Dilated and the measurement of eytosolic Ca2+. Analytical Biochem214: 413-
cardiomyopathy assoeiated with taurine deficiency in the domestie 419,1993
cat: Re\ationship to diet and myocardial taurine content. Adv Exp Med 41. Bkaily G, Economos 0, Potvin L, Ardilouze JL, Marriott C, Coreos J,
Bio1315: 63-73, 1992 Bonneau 0, Fong CN: Blockade of insulin steady-state R-type Ca'+
24. Moise NS, Pacioretty LM, Kallfelz FA, Shipanuk MH, King JM, channel by PN 200-110 in heart and vascular smooth musele. Mol
Gilmour RF Jr.: Dietarytaurine deficiency and dilated eardiomyopathy Cell Bioehern 117: 93-\06, 1992
in the fox. Am Heart J 121: 541-547,1991 42. Bkaily G, Gros-Louis N, Naik R, Jaalouk 0, Pothier P: Implieation of
25. Lake N: Effeets of taurine defieieney on arrhythmogenesis and the nuelcus in exeilalion eontraction coupling ofheart eells. Mol Cell
excitation-eontraetion coupling in eardiae tissue. Adv Exp Med Biol Biocheml54: 113-121, 1996
315: 173-179, 1992 43. Jones DP, Miller LA, Budreau A, Chesney RW: Characteristics of
26. Eley DW, Lake N, ter Keurs HEDJ: Taurine depletion and exeitation- taurine transport in cultured renal epithelial cell lines: Asymetrie
eonlraetion eoupling in rat myoeardium. Cire Res 74: 1210--1219, 1994 polarity of proximal and distal eelliines. Adv Exp Med Biol 315:
27. Lake N: Alterations ofventrieular contraetility and myofibrilloss in 405-411, 1992
taurine-deficient hearts. Adv Exp Med Bio1359: 335-342, 1994 44. Bkaily G, POlhier P, D'ürleans-Juste P, Simaan M, Jaeques 0, Jaalouk
28. Takahashi K, Harada H, Schaffer SW, Azuma J: Effeet oftaurine on 0, Belzile F, Boutin C, IIaddad G, NeugebauerW: The use of confocal
intraeellular calcium dynamies of eultured myoeardial eells during the mieroseopy in the investigation of eell structure and function in heart,
calcium paradox. AdvExp MedBio1315: 153-161, 1992 vaseular endothelium and smooth muscle eells. Mol Cell Bioehem
29. Bkaily G, Haddad G, Jaalouk 0, Gros-Louis N, Taoudi-Beriehekroun 172: 171 194,1997
M, Nalk R, Pothier P, D'ürleans-Juste P, Bui M, Wang S, Sperelakis 45. Bkaily G, Jaalouk 0, Jaeques 0, Economos 0, Hassan G, Simaan
N: Modulation of Ca 2+ and Na+ transport by taurine in heart and M, Regoli 0, Pothier P: Bradykinin aetivates R-, T -, and L-type Ca'+
vaseular smooth musele. Adv Exp Med Bio1403: 263-273,1996 ehannels and induees a sustained inerease of nudear Ca'+ in aortic
30. Bkaily G, Jaalouk 0, Haddad G, Gros-Louis N, Simaan N, Naik R, vaseular srnooth rnuscle eells. Can J Physiol Pharmacol 75: 652-
Pothier P: Modulation of cytosolie and nuelear Ca2+and Na+ transport 660,1997
by taurine in heart eells. Mol Cell Bioehem 170: 1-8, 1997 46. Cragoe EJ Jr, Wultersdorlow Jr, Bieking JB, Kwong SE, Jones JH:
31. Akera T, Ku 0, Brody TM: Attenuation of ion movements as a Pyrazine diureties.I!. N-amidino-3-amino-5-substitutcd-6-halopyrazinc-
meehanism of drug-indueed arrhythmias and inotropie responses. In: carboxamides. J Med Chem 10: 66-75, 1997
Molecular and Cellular Biochemistry 188: 199--208, 1998.
© 1998 Kluwer Academic Publishers.

Effects of long-term treatment with


eicosapentaenoic acid on the heart subjected to
ischemia/reperfusion and hypoxia/reoxygenation in
rats
Satoshi Takeo, Yoshihisa Nasa, Kouichi Tanonaka, Ken-ichi Yabe,
Michiko Nojiri, Michihiko Hayashi, Hideo Sasaki, Kumiko Ida and
Kyoko Yanai
Department of Pharmacology, Tokyo University of Pharmacy & Life Science, Hachioji, Japan

Abstract
The effects of eicosapentaenoic acid (EPA) and long-term treatment with EPA-ethylester (EPA-E) were examined in perfused
rat hearts subjected to ischemialreperfusion and adult rat cardiomyocytes subjected to hypoxialreoxygenation. EPA (0.1 flM)
improved postischmic contractile dysfunction of the ischemic/reperfused heart. EPA (10 flM) attenuated hypoxialreoxygenation-
induced morphological deterioration of cardiomyocytes. The results suggest the presence of direct cardioprotective effects of
EPA. Rats were orally treated for 4 weeks with I g/kglday ofEPA-E to elucidate ex vivo effects ofEPA, and the fatty acid
composition of cardiac phospholipids was determined. The percent ratio of EPA in total fatty acids of cardiac phospholipids
increased whereas that of arachidonic acid decreased. The percent ratio ofn-3/n-6 fatty acid did not increase. Treatment with
EPA-E did not improve the post-ischemic contractile function, but attenuated the ischemialreperfusion-induced release of
prostaglandins during reperfusion. Treatment with EPA-E preserved a better morphological appearance of the cardiomyocytes
subjected to hypoxia/reoxygenation. The results suggest that the mechanisms responsible for cytoprotective effects ofhypoxicl
reoxygenated cardiomyocytes or inhibition of metabolic alterations of the ischemic/reperfused heart by long-term EPA-E
treatment did not contribute substantially to recovery of post-ischemic contractile dysfunction. The direct in vitro effects of
EPA may playa role in the protection ofthe heart from ischemialreperfusion or hypoxialreoxygenation injury. (Mol Cell Biochem
188: 199-208, 1998)

Key words: cardiac contraction, cardiomyocyte, encosapentaenoic acid, ischemia and reperfusion injury, perfused heart,
polyunsaturated fatty acid, prostaglandin

Introduction benefit of dietary fish oil to ischemia- or reperfusion-induced


arrhythmias [6-8]. However, the benefit of dietary fish oil to
Epidemiological studies have shown that dietary intake of ischemia/reperfusion-induced contractile dysfunction is
fish oil decreases incidence ofischemic heart disease [1, 2] controversial;effective [9, 10] orineffective [11,12]. Possible
and lowers the morbidity from coronary heart disease [3]. mechanisms underlying the benefit of dietary fish oil in
This effect is considered to be attributed to prolongation of prevention against coronary and ischemic heart diseases and
blood coagulation time and inhibition ofplatelet aggregation ischemic injury are considered to be due to intake of n-3
[4,5]. Several reports in animal studies have shown the polyunsaturated fatty acids (pUFAs) such as eicosapentaenoic

Addressfor offprints: S. Takeo, Department ofpharmaeology, Tokyo University ofPhannaey and Life Seienee, 1432-1, Horinouehi, Haehioji 192-0392, Japan
200

acid (EPA) and docosahexaenoic acid (DHA), and the intake- approximately 20 g diet/day, it is considered that animals
induced changes in ratios of n-3/n-6 fatty acids in cardiac cell generally take approximately 50 mg ofn-3 PUFAs from diet
membrane phospholipids [12, 13]. Despite such hypotheses, ofthe standard rat chow. This diet contained approximately
there is little information conceming cardioprotection ofEPA 1.2% EPA to the daily gavage of 1 g/kg/day EPA-E. After
or DHAper se against ischemic injury. The present study was treatment for 4 weeks, hearts were isolated from EPA-E-
undertaken to determine whether EPA may exert cardio- treated and vehic1e-treated animals, and their fatty acid
protective effects on perfused hearts subjected to ischemial compositions in phospholipids were determined by the HPLC
reperfusion or cytoprotective effects on cardiomyocytes method described previously[ 14]. Fatty acid composition of
subjected to hypoxia/reoxygenation. For this purpose, the in plasma lipids was also determined. The treated animals were
vitro effects ofEPAand the ex vivo effects oflong-termEPA- also used for studies ofperfused hearts and cardiomyocytes
ethylester (EPA-E) treatment on ischemic/reperfused hearts in another set of experiments.
and hypoxic/reoxygenated cardiomyocytes of rats were
examined.
Perfused heart study

Materials and methods Ischemia/reperfusion ofisolated hearts


Perfusion of isolated rat hearts was carried out in a Langendorff
Methods mode according to the method described previously [15]. The
hearts were perfusedat 37C at a constant flow rate of9.0 mll
Male Sprague-Dowley rats (SLC, Shizuoka, Japan) were min with the Krebs-Henseleit solution of the following
used for studies of the ex vivo and in vitro effects of EPA. composition (mM): NaC1120, KCI4.8, KHl04 1.2, MgS0 4
The experimental protocol was designed according to the 1.2, NaHC0 3 25, CaCl 2 1.25 and glucose 11. The perfusion
Guideline ofExperimental Animal Care issued by the Prime buffer was equilibrated with agas mixture of95% 02 + 5%
Minister's Office of Japan, and approved by the University CO 2 (P0 2 , > 600 mmHg), the pH of which maintained
Committee of Animal Care and Welfare. between 7.40-7.42. A latex balloon with an uninflated
diameter of 3.7 mm, connected to apressure transducer
(model TP-200T, Nihon Kohden, Tokyo, Japan), was inserted
Treatment with EPA-E into the left ventricularcavity through the mitral opening and
secured with a ligature that inc1uded the left arterial remnants.
Rats weighing 90-100 g were treated orally with 1 glkglday The hearts were loaded with 5 mmHg of the initial left
of EPA-E in a volume of 0.4 ml of saline for 4 weeks. As ventricular end-diastolic pressure. Left ventricular developed
control, rats were treated with vehicle. During treatment the pressure (LVDP) was measured by an electronic manometer
animals were fed with standard rat chow. This standard rat (model TP-400T, Nihon Kohden) and left ventricular end-
chow contained 0.26% n-3 PUFAs in which 0.6Img/g EPA, diastolic pressure (LVEDP) was triggered via the development
0.93mg/g DHA and 1.03 mg/g linolenic acid were present on ofLVDP. Heart rate measurements were triggered from the
the basis of the analysis of diet (Table 1). Since rats take development of LVDP by a heart rate counter (model AT-
601G, Nihon Kohden). Rate-pressure product (RPP) was
determined by multiplying heart rate and LVDP. Perfusion
Table 1. Fatty acid composition oftotallipids in rat chow pressure was measured through a branch ofthe aortic cannula
amount percent by an electronic manometer (model TP-400T, Nihon Kohden).
(j.lmollg) All hemodynamic parameters were recorded on athermal pen
recorder (model WT-645G, Nihon Kohden).
C12:0 0.1 0.1
22.5 26.2
Ischemia was induced by stopping the perfusion and
C16:0
C18:0 2.4 2.6 submerging the heart in an organ bath filled with glucose free-
C18:1 20.8 22.4 Krebs-Henseleit buffer equilibrated with 95% N2 + 5% CO2
CI8:2(n-6) 35.8 38.5 at 37°C to avoid hypothermia-induced cardioprotection.
CI8:3(n-3) 3.7 4.0 Reperfusion was carried out by perfusing the heart with
C20:4(n-6) 0.9 1.0
C20:5(n-3) 2.0 2.2
Krebs-Henseleit buffer equilibrated with agas mixture of
C22:6(n-3) 2.8 3.1 95% 02 + 5% CO2 after draining the submerging buffer. When
the direct effects ofEPA was examined, isolated hearts ofthe
Abbreviations; CI2:0-lauric acid; C16:0 -palmitic acid; CI8:0-stearic
non-treated rat, weighing 250-300g, were treated for the last
acid; C18:1 - oleic acid; C18:2 -linoleic acid; C18:3 - linolenic acid;
C20:4 arachidonic acid; C20:5 - eicosapentaenoic acid (EPA); C22:6 - 20 min of pre-ischemia with 0.1 mM EPA (free form)
docosahexaenoic acid (DHA). dissolved in the Krebs-Henseleit buffer and 0.001 % DMSO.
201

After pretreatment, the hearts were subjected to global containing coIlagenase and the tissue suspension was
ischemia for 30 min and subsequent reperfusion for 30 min. further reacted with collagenase for 3 min. The suspension
EPA at the same concentration as above was also present in was filtered through a mesh with a pore size ofO.4 mm and
the reperfusion buffer. then centrifuged at 22 x g for 1 min. The sediments were
When the ex vivo effects of EPA was examined, the rats resuspended in a calcium-free buffer containing 1% BSA.
were treated for 4 weeks with I g/kg/day of EPA-E as Calcium concentration for suspension of cells was gradually
described above. The hearts isolated from EPA-E-treated or elevated up to 1 mM. The isolated cardiomyocytes were
vehic1e-treated rats were subjected to global ischemia for 30 centrifuged at 22 x g for Imin to sediment viable cells.
min and subsequent reperfusion for 30 min as described Finally the cells (calcium-tolerant cardiomyocytes) were
above. resuspended in minimum essential medium (MEM, GIB CO
BRI, Grand Island, NY, USA) containing 10 mM HEPES (pH
7.4). Approximately 2-3 million calcium-tolerant rod-shaped
Measurement ofprostaglandins, norepinephrine and cardiomyocytes were iso la ted from each heart.
creatine kinase in efJluent
Hypoxialreoxygenation of iso la ted cardiomyocytes
To elucidate the contribution of prostaglandins (PGs) and The isolated cardiomyocytes were placed onto the bottom of
norepinephrine (NE) to recovery of contractile function of cultured dishes with grid lines (Inter Med., Nunc Inc.,
the perfused heart, the effiuent from the reperfused heart was Naperville, IL, USA), which wascoated with laminin (Becton
collected and the concentrations of PGs and NE in effiuent Dickinson, Parsippany, NJ, USA) to stick the cells onto the
were determined. 6-Keto-PGF 1a and TXB 2, whieh are derived bottom. The cells were incubated at 37°C for 60 min. The dish
from P0I2 and TXA2, respectively, were determined by the was rinsed with MEM medium to remove unattached, calcium-
method ofthe enzyme immunoassay commercially available intolerant cells. More than 90% of cells in a dish revealed rod-
(Cayman Chem. Ann Arbor, MI, USA). NE concentration shape in the present study. The cardiomyocytes in dishes were
was determined according to the HPLC method described placed in a tightly sealed hand-made chamber. Hypoxia was
previously[16]. Creatine kinase (CK) activity in effiuent was induced by incubating the cardiomyocytes in glucose-free
also determined by the enzymatic method described previously HEPES buffer (pH 7.4) with 0.1% BSA (hypoxie medium)
[15]. in the chamber at 37°C und er passage of 100% nitrogen gas
for 2 h. Reoxygenation was performed for 15 min by
exchanging nitrogen gas with 100% oxygen gas. Glucose at
Isolated cardiomyocyte study the final dish concentration of 11 mM was also supplemented
at the onset ofreoxygenation. The p02 ofthe medium in the
Isolation of cardiomyocytes dishes during hypoxia and reoxygenation, when determined
The rats used for cardiomyocyte study were treated with EPA- in a preliminary study, were 18--19 mmHg and > 500 mmHg,
E similarlyas described for the perfused heart study. Cardio- respectively. A preliminary study also showed that morpho-
myocytes were isolated according to the method described 10gical outcome of cardiomyocytes at 30 and 60 min of
previously [17], which was a modified method ofPiperet al. reoxygenation did not differ from that at 15 min of re-
[18]. In brief, after anesthesia with 50 mg/kg pentobarbital oxygenation.
sodium, i.p., their hearts wereisolated. The isolated hearts Morphological outcome was assessed by appearance of
were perfused with the buffer containing the following cardiomyocytes exposed to hypoxia and reoxygenation. After
composition (mM): NaCl 140.0, KCI 4.8, KH 2 P0 4 1.2, hypoxia or reoxygenation, cardiomyocytes were fixed with
MgS04 1.2, CaCl2 1.25, glucose 11, sodium pyruvate 5 and 2% glutaraldehyde and then were counted in the area in which
N-2-hydroxyethylpiperadine- N'-2-ethanesulfonic acid 10 the cells had been counted before hypoxie incubation (numbers
(HEPES), pH 7.4. The buffer was equilibrated with 100% of cells: > 500). Photographs were also taken to confirm the
oxygen gas. The hearts were at first perfused with the HEPES estimation of cell shapes. To evaluate morphological
buffer for 5 min in a non-recirculating manner, and then with changes, the shapes of cardiomyocytes were determined by
calcium-free HEPES buffer for 5 min. After perfusion, the a microscope (MT-2, Olympus Inc., Tokyo, Japan). We
hearts were further perfused in a recirculating manner for 30 c1assified three types of cells, rod-, square- and round-shaped
min with calcium-free HEPES buffer, pH 7.1, supplemented cells according to the criterion of Harworth et al. [19]. The
with 25 ~M CaCI 2, 0.05% collagenase (Type 11, Worthington, ratios of ceIl length to cell width were more than 3 for the
USA) and 0.1 % bovine serum albumin (BSA, Biocell Labora- rod shaped cell, and less than 3 for the square shaped cell.
tories, USA). The hearts were removed from the apparatus and The round cells were c1early round. The rod- and square-
chopped in a calcium-free HEPES buffer supplemented with shaped cells had an intact sarcolemma and were considered
1% BSA. Chopped tissue was incubated in the buffer to be viable, whereas the round shaped cells to be dead. This
202

was determined on the basis ofthe morphological appearance Statistics


and the ability to exc1ude trypan blue. The validity of this
criterion was also confirmed by the staining of cells with Values are expressed as the means ± S.E.M. Statistical
calcein-AM [20]. In the present study, the percentage of cells significance was evaluated by one-way or two-way analysis
of each type compared to the initial number of the cells was ofvariance (ANOVA). Student's t-test was also applied for
determined. Survival of cardiomyocytes after hypoxia and comparison ofmean values oftwo groups. Differences with
reoxygenation was estimated as apercentage of the sum of a probability of 5% or less were considered to be statistical
rod- and square-shaped cells. significant.
When the in vitro, direct effects ofEPA was examined, 10
11M EPA dissolved in 0.1 % DMSO was present in the buffer
of the dishes as above. EPA was present in the buffer Results
throughout the experiments. In a preliminary study, we
examined the effects of different concentrations of EPA In vitro effects
ranging from 0.1- 100 11M and found that the number ofrod-
shaped cells was the largest in the group treated with 10 11M Direct effects oj EPA on ischernic/reperjused hearts
EPA among groups. In the first set of experiments, the direct effects of EPA on
ischemic/reperfused hearts were examined. The recovery of
RPP, LVDP, LVEDP, heart rate and perfusion pressure ofthe
Materials heart at the end of reperfusion was determined. The recovery
ofRPP (45 ± 8%, n = 8) ofthe heart isolated from pretreatment
The following agents and substances were used in the present with EPA was smalI, but significantly, higher than that (24 ±
study: collagenase (type II, Worthington, USA), EPA-E 4%, n = 6) ofthe EPA-untreated (control) heart (Fig I). It
(Epadil®, Mochida Pharm. Co. Japan), bovine serum albumin should be noted that this recovery mainly depended on that
(Biocell Lab. USA), laminin (Becton Dickinson, USA), and ofLVDP (58 ± 12 and 28 ± 3 mmHg for EPA-treated and
HEPES (Dojin Chem., Japan) . control groups, respectively) ofthe ischemic/reperfusedheart,

.....
c30000 RPP 100
E2S000
..... lt.........
CI) 20000 ischemia ..... 75
iii aI
:::t:
.8 15000 E50
%
m10000 E
--25
E SOOO

--
E 0
-20 -10 0 10 20 30 40 5060
0

400
-20 -10 0 10 20 30 40 5060
150 LVEDP HR
125
~300
aI 100
:::t: E
..... 200
E 75
f
-- E
50
25
.8 100
0 0
-20 -10 0 10 20 30 40 50 60 ·20 -10 0 10 20 30 40 50 60
perfusion time (mln) perfusion time ( mln )

Fig. I. The time course of changes in rate·pressure product (RPP), left ventricular developed pressure (L VDP), left ventricular end·diastolic pressure
(LVEDP) and heart rate (HR) ofthe ischemic/reprfused heart treated with (closed triangles) and without (control; open circles) O.II'M eicosapentaenoic
acid (EPA). The isolated rat hearts were subjected to 30 min ofischemia and 30 min ofreperfusion. Treatment with EPAwas carried out during the last 20
min ofpre·ischemia and during 30 min ofreperfusion. Values represent the means+S.E.M. of6 (control) and 8 (EPA·treated) experiments.
203

since heart rate of the ischemic/reperfused heart did not differ


between EPA-treated and control groups. An appreciable EPA 10llM
100
attenuation of the increase in LDEVP (63 ± 10 and 88 ± 9 vlable cells EPA el: 1'=0.13
mmHg at the end ofreprfusion for EPA-treated and control 80 ,..-.- HIR eI: P<O.05
groups) of ischemic/reperfused heart was seen in hearts 60 r=-
treated with 0.1 ~M EPA (Fig I). There were no differences -==-
40
in changes ofperfusion pressure during ischemia/reperfusion
between EPA-treated and control groups. The release ofCK 20
between EPA-treated or control hearts during 30 min ~ 0
reperfusion did not differ (22 ± 4 and 23 ± 3 nmol NADPHI
~100~----------------------~
min/g wet tissue for EPA-treated and control groups). In a ~ 80 Rod-shaped cells
c:
preliminary study we examined the effects of different
concentrations ofEPA on post-ischemic recovery ofRPP of ~ 60 *
EPA el: P<O.02
HIR el: P::O.68 *
the heart. The recovery rates of the hearts treated with 0.01,
-s 40
-
'.',::.,::.,::.,:.:,: ,' ...,'.,
..,'..,'.,
..,'." :::;:::::;
0.1 and 1 ~M EPA were 28, 48 and 36% (mean values of 2 ....................

experiments). It should also be noted that the post-ischemic .e 20 .::.:,:.::.{;.:


. ...
. ~,~,:,~,~
:':':':':'
~.:~.:;:.;:.~,'.
. .. . . ~:~:;:;~;;
.. -
':'.:':'.:,:.::,.:;,,:

recovery ofRPP by in vitro treatment with EPA was smaller o O~~~~·~


~~~
,~~
~ ------~,~
·~~
~·~
,, ~~
-- ~
--~
- u

(31 ± 9%, n = 4) when the heart was treated only far the last
20 min ofpre-ischemia.
c:1~DT-----------------------------'
CD
...oCD EPA ef: P::O.13
11. HIR el: P<O.05
Direct ejJects on hypoxic/reoxygenated cardiomyocytes
In the second set of experiments, the direct effects of EPA
on hypoxic/reoxygenated cardiomyocytes isolated from
non-treated rats were examined (Fig 2). Hypoxia induced
a decrease in the number of rod-shaped cells and an increase
control EPA control EPA
in the number of square-shaped ceHs. Reoxygenation
resulted in a retention of the number of rod-shaped ceHs, a Hypoxla Reoxygenatlon
decrease in the number of square-shaped ceHs and an
increase in the number of round-shaped ceHs. EPA at a
concentration of 10 ~M slightly but significantly attenuated Fig. 2. Survival of cardiomyocytes following hypoxialreoxygenation
hypoxia/reoxygenation-induced morphological deteriora- isolatcd from EPA-treatcd (EPA) and untrcatcd (control) hearts. Survival
was determined as percentages of rod-shaped and square-shaped cells to their
tion of the cardiomyocytes, particularly the presence of
initial cell numbers. Values represent the mean ± S.E.M. of 5 experiments.
larger number of rod-shaped ceHs after hypoxia and re- EPA ef- EPA effect; HlR ef- hypoxia/reoxygenation effect (2-way repeated
oxygenation in EPA-treated cells than that of untreated measures ANOVA). A significant difference (p < 0.05) was seen in the
(control) cardiomyocytes. nurnbers of rod-shaped cells between in vitra EPA-treated and control groups
(2-way ANOV A). ·Significantly different from contra I graup (p < 0.05).

Ex vivo ejJects oj EPA Hemodynamics ojischemic/reperjused hearts ofEPA-E


treated rat
Fatty acid composition In the fourth set of experiments, hearts of the rats treated with
In the third set of experiments, EPA-E at a dose of I g/kgl EPA-E for 4 weeks were subjected to 30 min ischemia and 30
day was oraHy administered into rats for 4 weeks. Fatty min reperfusion and changes in hemodynamics of the heart
acid compositions of plasma lipids and myocardial phos- were determined (Fig 3). There were no appreciable differences
pholipids were determined by HPLC. As shown in Table 2, in the changes in RPP, LVDP, LVEDP and HR ofthe ischemicl
the ratio ofEPA of myocardial phospholipids was increased reperfused hearts between EPA-E-treated and vehicle-treated
whereas that of DHA was decreased by long-term treat- animals. Changes in perfusion pressure of the ischemicl
ment with EPA-E. The ratio ofarachidonic acid to the total reperfused heart were similar between EPA-E-treated and
fatty acid in the myocardium and in plasma was decreased vehicle-treated animals. The release ofCK ofthe heart ofthe
by treatment with EPA-E. The ratio ofn-3/n-6 fatty acid of EPA-E-treated anima 1 during reperfusion (19 ± 2 nmol
total lipids in plasma increased whereas that of the myo- NADPH/minlg wet tissue, n = 8) did not differ from that of
cardium decreased in rats following long-term treatment the vehicle-treated animal (21 ± I nmol NADPH/minig wet
with EPA-E. tissue, n = 7).
204

Table 2. Fatty acid composilion oflolal phospholipids in hearts and oflolallipid in plasma oflhe EPA-E-trealed and vehicle-Ireated rats

myocardium plasma
vehicle-treated EPA-E treated vehiele-treated EPA-treated

C12:0 0.03 ± 0.01 0.01 ± 0.01 0.15 ± 0.02 0.12 ± 0.03


C16:0 13.70 ± 0.35 13.60 ± 0.28 19.86 ± 4.22 30.13 ± 1.26'
C18:0 21.21 ± 1.29 21.63 ± 0.63 11.83 ± 1.61 10.21 ± 0.62
CI8:I 6.44 ± 0.24 8.05 ± 0.12' 18.36 ± 1.53 16.79 ± 0.85
CI8:2(n-6) 20.18 ± 1.16 27.50 ± 0.59* 23.66 ± 0.78 24.41 ± 0.23
C18:3(n-3) 0.83 ± 0.67 0.17 ± 0.01 1.23 ± 0.07 1.09 ± 0.03
C20:4(n-6) 15.51 ± 0.80 12.02 ± 0.31* 16.68 ± 1.52 6.69 ± 0.79'
C20:5(n-3) 0.49 ± 0.19 2.34 ± 0.06' 2.15 ± 0.08 4.51 ± 0.14'
C22:6(n-3) 21.61 ± 1.12 14.68 ± 0.46' 6.07 ± 0.21 6.05 ± 0.19

% ofn-3 FA 22.93 ± 1.48 17.19±0.48' 9.45 ± 0.23 11.64 ± 0.32


% ofn-6 FA 35.69 ± 1.61 39.52 ± 0.66 40.34 ± 1.32 31.11 ± 0.80'
n-3/n-6 FA 0.66 ± 0.07 0.44 ± 0.01* 0.24 ± 0.01 0.38 ± 0.02'

Total phospholipid fatty acid contents ofthe myoeardium ofvehiele-treated and EPA-E-treated animals were 453.72 ± 29.77 and 437.98 ± 11.40 mol/g dry
tissue (n = 6), and total fatty acid eomposition ofplasma 4604.6 ± 694.8 and 4523.2 ± 291.3 mmollml (n = 6), respeetively. Values (n = 6) are expressed as
mol%. Abbreviations; C 12:0 -laurie acid; C16:0-palmitie acid; C18:0 - stearie acid; C18: 1 - oleie acid; C18:2 - linoleie acid; C 18:3 linolenie acid; C20:4
- araehidonie acid; C20:5 - eicosapentaenoie acid; C22:6 - docosahexaenoic acid; FA - fatty acid. *Significantly different from vehicle-Irealed group (p <
0.05).

Effects on release of PGs and NE treated or vehicle-treated rats were determined (Fig 4). 6-
To elucidate the effects on PG synthesis in the heart, PG keto-PGF,a and TXB 2 were released from the hearts with
metabolites released from the perfused heart of EPA-E- EPA-E-treated and vehic1e-treated animals during pre-

.- 30000 100
I:
E
.....
25000
20000 .- 75
S
=15000
.CI
01
:Z::so
E
Q10000 E
:z:: - 25
E 5000
E 0 0
-10 0 10 20 30 40 SO 60 ·10 10 20 30 40 SO 60
1SO 400
LVEDP
125
..... ~300
01100
:z:: E

-
E 75 ...... 200
CD
E 1iS

-
50
,8100
25
0 0
·10 0 10 20 30 40 50 60 ·10 0 10 20 30 40 SO 60
perfusion time (mln) perfusion time (mln)

Fig. 3. The time course of changes in ralc-pressure producl (RPP), left venlricular developed pressure (L VDP), left venlricular end-diaslolie press ure
(LVEDP) and heart rale (HR) of Ihe ischemic/reprfused heart of rals wilh EPA-E-trealed (closed triangles) and vehiele-Ireated (open eireles) rats. The
isolaled rat hearts were subjeeted 10 30 min of ischemia and 30 min of reperfusion. Values represent the means+S.E.M. of 6 (vehiele- Irealed) and 8 (EPA-
Irealed) experimenls.
205

6-keto-PGFla TXB2 NE
0.20 100

-C 0.15
80

0
E
Cf)
60 *
...... 0.10
C)
...... 40

-
C)
C 1 0.05
* 20

0.00 0
vehlcle EPA-E vehlcle EPA-E vehlcle EPA-E

Fig.4. Release of6-keto-PGF 1a , TXB, and norepinephrine (NE) during 30 min ofreperfusion in the hearts with long-term EPA-E or vehicle treatment.
Values represents the means ± S.E.M. of 6 experiments. The initial values of the release of 6-keto-PGF '" (B and TXB of EPA-E treated animals were 0.53 ±
0.07 and 0.03 ± 0.01 ng/min/g wet tissue, and those ofvehicle-treated animal 0.30 ± 0.05 and 0.01 ± 0.01 ng/min/g wet tissue. The initial values for the
release ofNE ofEPA-E treated and vehicle-treated animals were 0.091 ± 0.03 and 0.095 ± 0.03 ng/minlg wet tissue, respectively. 'Significantly different
from vehicle-treated group (p < 0.05).

isehemia to a minor degree. The release of 6-keto-PGF la and was the largest in cardiomyocytes treated with 1 g/kg/day
TXB 2, stable metabolites of PGI 2 and TXA 2 respeetively, among these groups.
was increased when the heart was reperfused: particularly,
the release was prominent during the first 10 min of reper-
fusion. Long-term EPA-E treatment significantly attenuated Discussion
the release of both PG metabolites from the reperfused
heart. In the present study, we attempted to elucidate the possible
To elucidate the effects of EPA-E on NE release of the role ofEPA in eardioprotection against ischemialreperfusion
perfused heart, NE in the emuent from the reperfused heart and hypoxialreoxygenation injury. At first, the direct effects
was determined (Fig 4). NE was minimally released during of EPA on the ischemic/reperfused heart of non-treated rats
pre-ischemia from the heart. The release ofNE increased were examined. We found that in vitra treatment with EPA
upon reperfusion: particularly, the release was prominent appreciably improved the postischemic RPP (LVDP x heart
during the first 5 min of reperfusion. Long-term treatment rate) and LVDP of the isehemie/reperfused heart. This
with EPA-E resulted in a slight, but significant, attenuation observation may implicate the improvement of cardiac
ofthe release ofNE from thc rcperfused heart. contractile dysfunction since no effect of EPA on heart rate
was seen in the ischemie/reperfused heart. We also observed
Ex vivo effects on hypoxic/reoxygenated cardiomyacytes a better preservation of morphologieal appearanee of eardio-
In the fifth set of experiments, rats were treated orally with I myocytes subjected to hypoxia/reoxygenation by in vitra
glkg/day ofEPA-E for 4 weeks and their hearts were isolated. treatment with EPA. It should be noted that this cytoprotective
Cardiomyocytes isolated from the hearts were subjected to effect was seen in cardiomyocytes that did not contract
hypoxia/reoxygenation as deseribed in the Materials and spontaneously or artificially. Since the heart is an organ to
methods section. Morphological appearance of eardio- primarily reveal contraetile aetivity, deteriorating faetors
myoeytes at the end of reoxygenation was preserved by involving in the meehanieal funetion of cardiomyoeytes
treatment with EPA-E to a greater extent than that of the would be unlikely related to the effeet of EPA. Rather, the
vehicle-treated rat. That is, the number of square-shaped eells results imply that the protection of the eardiomyoeyte
after reoxygenation, but not after hypoxia, was greater in EPA- appearance is attributed to prevention of membrane function
E-treated rat cardiomyoeytes than that ofthe vehicle-treated and/or biochemieal events whieh cause hypoxialreoxygenation
rat cardiomyocytes (Fig 5). In a preliminary study, we injury.
examined the effects of different doses ofEPA-E ranging from There is, so far, no eoncrete evidenee to support the benefit
0.1 to 1 glkglday and found thatthe number ofrod-shaped cells of direct effeets ofEPA on isehemialreperfusion or hypoxial
206

as EPA may contribute to the improvement of ischemial


EPA-E 1 g/kg/day
100 reperfusion- or hypoxia/reoxygenation-induced injury to an
vlable cells appreciable degree.
BO In contrast to the in vitra effects, no appreciable improve-
EPA el: P<O.05
60 HIR el: P<O.05 ment of contractile function was seen in the ischemic/reper-
* fused heart isolated from the rat with long-term treatment

..
40
ofEPA-E. Several investigators have shown that improve-
20 ment of contractile dysfunction [9, 10] and mitochondrial
GI
.a 0 oxidative metabolism [12] ofthe ischemic/reperfused hearts
E 100 was associated with an increase in the ratio ofn-3/n-6 fatty
::s Rod-shaped cells
c: 80
acids ofthe myocardium following long-term intake offish
oil or PUFAs. We did not find any increase in this ratio
Gi 60 EPA el: P::O.35
U HIR el: P=O.16 following long-term treatment with EPA-E in the present

---
study. Gryberg et al. [24] have shown that docosapentaenoic
S0 40
acid (DPA), a n-3 fatty acid, is an elongation product ofEPA
20 and is increased when cardiomyocytes are incubated in
0
0 EPA-enriched medium. Thus, we tried to detect this compo-

..
c: nent ofthe heart and plasma in the present study, but it was
GI
U found that DPA and linoleic acid were inseparable under
GI the present HPLC conditions. Since we observed a signi-
11. EPA el: P<O.05
HIR el: P<O.05 ficant increase in linoleic acid in the myocardium of rats
treated with EPA-E, the amount of linoleic acid detected
might be overestimated in the present study. Further
examination ofthe role ofn-3/n-6 fatty acids in protection
ofthe heart against hypoxia/reoxygenation injury must be
vehlcle EPA-E vehlcle EPA-E
performed.
Hypoxla Reoxygenatlon In the present study, arachidonic acid composition ofthe
membrane phospholipid was reduced by long-term treatment
with EPA-E, as shown in Tablc 1. Several reports have
shown that ischemia or hypoxia elicited the reduced ability
Fig. 5. Survival of eardiomyoeytes following hypoxia/reoxygenation to acylate fatty acids due to lack of energy and accumulation
isolated from EPA-E-treated and vehicle-treated animals. Survival was
determined as a pereentage ofthe number of rod-shaped and square-shaped
offree arachidonic acid in the heart [25]. Such accumulation
cells to their initial eell numbers. Values represent the means ± S.E.M. of 8 of arachidonate may alter myocardial function through
experiments. EPA ef, [PA effeet; IIIR ef, hypoxia/reoxygenation effeet. arachindonic acid-mediated mechanisms including free radical
*Significantly different from vehiele-treated group (p < 0.05). formation [26, 27] and inhibition ofboth mitochondrial Ca2+
transport [28] and Na+/K+ ATPase activity [26]. Thus, long-
term treatment with EPA-E appears to be beneficial in terms
reoxygenation injury. Hallaq et al. [21,22] have shown that of reduction in arachidonic acid composition in the cardiac
n-3 PUFAs, such as EPA and DHA, attenuated ouabain- phospholipid. We did not, however, observe any appreciable
induced toxicity of neonatal rat myocytes possibly through effect oflong-term treatment with EPA-E on cardiac function
the mechanism by which calcium currents can be modulated of the ischemic/reperfused heart. Thus, it is likely that
through dihyropyridine-sensitive L-type calcium channel in although alteration in membrane fatty acid composition in
the sarcolemma. EPA per se has been shown to block sodium cardiac muscle cells is induced by long-term treatment with
channel ofneonatal ratcardiomyocytes [23]. It is also known EPA-E, it does not elicit any beneficial effect on contracti1e
that fatty acids are capable ofbeing utilized as substrates for function ofthe heart subjected to ischemia/reperfusion.
energy production und er aerobic conditions. Thus, the It should be noted that the results on examination of the
presence of EPA in the buffer might contribute to enhance perfusate suggested that production of PGI2 and TXA, was
energy production in the reoxygenated or reperfused heart. attenuated in the ischemic/reperfused heart of rats with long-
Although it would be premature to correlate these electro- term EPA-E treatment despite no improvement of cardiac
physiological effects of n-3 PUFAs with recovery of post- function. This reduction in the PG metabolites may be due
ischemic contractile dysfunction ofthe ischemic/reperfused to the decrease in arachidonic acid in cardiac phospholipids
heart and protection of the cardiomyocytes against hypoxial following long-term treatment with EPA-E. The results are
reoxygenation injury, the direct actions of n-3 PUFAs such in agreement with the findings that modulation of myocardial
207

eicosanoid production induced by dietary supplementalion relatively high as compared with the therapeutic dose ofEPA,
with n-3 PUFAs resulted in changes ofsynthesis ofPGI 2 and 30 mg/kg, in human. Thus, the observed effects of EPA-E
TXA2 and formation oftheir metabolites in myocytes [29, 30]. treatment on the attenuation ofthe release ofmetabolites and
Endogenous PGs, irrespective of types of PGs, have been protection of cardiac membrane may playa minor role in
shown to accompany reperfusion-induced cardiac dysfunction prevention of the development of cardiac contractile
in perfused hearts [31]. Ifthis were the case, reduction in the dysfunction in the ischemic/reperfused heart. Rather, direct
formation ofPGs in animals with long-term intake ofEPA-E action of an appropriate concentration ofEPAmay playa role
would be beneficial. However, reduction of such alterations in prevention ofthe heart from ischemia/reperfusion injury.
in PG synthesis did not accompany improvement of cardiac In a preliminary study, we also examined the ex vivo effects
function of the ischemic/reperfused heart in the present ofDHA on hypoxia/reoxygenation injury of isolated cardio-
study. myocytes according to the same experimental protocol as that
We observed the attenuation of NE release from the ofEPA-E. We observed theex vivo effects ofDHA to a similar
reperfused heart of the rats treated with EPA-E. It is weil degree to those of EPA-E. This mayaIso conclude that the
recognized that NE is released from the myocardium and/or ex vivo effects of n-3 fatty acids per se on the myocardium
coronary artery through excitotoxic mechanism when the would be limited.
tissues are ischemic [32]. NE release during sustained
ischemia occurs by nonexocitoxic mechanism [33]. Such
increase in NE release during ischemia has been shown to References
acce1erate the development of myocardial injury [34].
However, the attenuation ofNE release by long-term treatment
1. Bang HO, Dyerberg J: Plasma lipids and lipoproteins in Greenlandic
with EPA-E did not accompany the recovery of cardiac west coast Eskimos. Acta Med Scand 192: 85-94, 1972
contractile function in the present study, suggesting that 2. Bang HO, Dyerberg JE: A hypothesis on the development of acute
suppression ofNE release did not contribute to recovery of myocardial infarction in Greenlanders. Scand J Clin Lab Invest 42:
the post-ischemic contractile function ofthe heart under the 7-13, 1978
3. Kromhout 0, Bosschieter EB, Coulander CL: The inverse relation
present experimental conditions.
betwecn fish consumption and 20-year mortality from coronary heart
The cardiomyocyte study showed appreciable improvement diseasc. New Eng J Med 312: 1205-1209,1985
of morphologie al deterioration of hypoxic/reoxygenated 4. Bang HO, Dyerberg J: The bleeding tendency in Greenland Eskimos.
cardiomyocytes of rats with long-term EPA-E treatment. Danish Med BuH 27: 202-205, 1980
Possibly, this cytoprotection represents protection of 5. Zhu B-Q, Sievers RE, Surr Y-P, Morse-Fisher N, Parmley ww, Wolfe
membrane function of myocardial eells exposed to oxygen CL: Is the reduction of myocardial infarct size by dietary fish oil the
result of altered platelet function? Am Heart J 127: 744-755, 1994
deficiency since we observed in a prcliminary study that 6. McLennan DL, Abeywardena MY, Charnock JS: Dietary fish oil
morphological alterations of cardiomyocytes subjected to prevents ventricular fibriHation foHowing coronary artery occlusion
hypoxia and reoxygenation were always associated with and reperfusion. Am J Physiol 116: H709-H717, 1988
release of creatine kinase, a marker of cell death or altera- 7. Hock CE, Beck LD, Bodine RC, ReibelOK: Influence of dietary n-3
fatty acids on myocardial ischemia and rcperfusion. Am J Physiol259:
tions in membrane permeability [35], and purine metabolites
HI518-HI526,1990
and bases, a marker ofloss of ATP metabolites necessary to 8. Pepe S, McLennan PL: Oietary fish oil confers direct antiarrhythmic
synthesis of ATP in cardiac cells [36]. This suggests that properties on the myocardium ofrats. J Nutr 126: 34-42, 1996
eardiac membrane integrity of quiescent eardiomyocytes 9. Yang BC, Saldeen TGP, Bryant JL, Nichols WW, Mehta JL: Long-
against hypoxia/reoxygenation injury is preserved to an term dietary fish oil supplementation protects against ischemia-
appreciable degree by long-term treatment with EPA-E. reperfusion-inducedmyocardial dysfunction in isolated rat hearts. Am
Heart J 126: 1287-1292, 1993
In the present study, we observed the dissociation ofthe 10. Yang BC, Saldeen TGP, Nichols WW, Mehta JL: Oietary fish oil
recovery of cardiac contractile function of the ischemicl supplementation attenuates myocardial dysfunction and injury caused
rcperfused heart with the suppression in the release of PG by global ischemia and reperfusion in isolated rat hearts. J Nutr 123:
metabolites as weIl as NE, and with cytoprotection of 2067-2074, 1993
11. Karrnazyn M, Horackova M, Murphy MG: Effects of dietary cod liver
hypoxic/reoxygenated cardiomyocytes in terms of the
oil on fatty-acid composition and calcium transport in isolated adult
effects of EPA. The results suggest that functional failure rat ventricular myocytes and on the response of isolated hearts to
of ischemic/reperfused hearts may be caused not only by ischemia andreperfusion. CanJPhysiolPharrnacol65: 201-209, 1987
changes in membrane function or metabolic alterations but 12. Oemaison L, Sergiel JP, Moreau 0, Grynberg A: Tnfluence of the
also by other determinants which are more profoundly phospholipid n-6/n-3 polyunsaturated fatty acid ratio on the mito-
chondrial oxidative metabolism before and after myocardial ischemia.
affected including sodium and calcium overload [37, 38],
Biochim Biophys Acta 1227: 53-59,1994
formation of free-radicals [39], depletion of high-energy 13. Hock CE, Holahan MA, ReibelOK: Effect of dietary fish oil on
phosphatcs [40], and no reflow phenomenon [41]. In addition, myocardial phospholipids and myocardial ischemic damage. Am J
the dose of EPA-E employed in the present ex vivo study Physiol 252: H554-H560, 1987
208

14. Nasa Y, Sakamoto Y, Sanbe A, Sasaki H, Yamaguehi F, Takeo S: 27. Basu HO, Karmazyn M, Injury to rat hearts produced by an exogenous
Changes in fatty acid eompositions of myoeardiallipids in rats with free radical generating system. Study into role of .raehidonic acid
heart failure followingmyoeardial infaretion. Mol Cell Bioehern 176: and eieosapenlaenoids. J Pharmacol Exp Ther 242: 673-{i83, 1987
179-189,1997 28. De Villiers M, Loehner A: Miloehondrial Ca" fluxes: Role of free
15. Takeo S, Liu J-X, Tanonaka K, Nasa Y, Yabe K, Tanahashi H, Sudo fatty acids, acyl-CoA and aeylcamitine. Bioehim Biophys Acta 876:
H: Reperfusion at redueed rates enhanees postisehemie eontraetile 309-317,1986
recovery ofperfused heart. Am J Physiol268: H2384-H2395, 1995 29. Abeywardena MY, McLennan PL, Chamoek IS: Differential effeets
16. Tuda K, Tsuda S, Masuyama Y: Enhaneed neuroinhibitory effeet of of dielary fish oil on myoeardial prostagIandin 12 and Ihromboxane
diltiazem in blood vessels of spontaneously hypertensive rats. Am J A2 produetion. Am I Physiol260: H379-H385, 1991
Hypertension 555-559, 1990 30. Ahumda GG, Sobel BE, Needleman P: Synthesis of prostagIandin by
17. Hayashi M, Nasa Y, Tanonaka K, Sasaki H, Miyake R, Hayashi J, cultured rat heart myocytes and cardiae mesenehymal eells. I Mol Cell
Takeo S: The effeet of long-term treatment with eieosapentaenoie Cardiol 12: 685-700, 1980
acid and doeosahexaenoie acid on hypoxia/reoxygenation injury of 31. Karmazyn M: Contribution ofprostaglandins to reperfusion-induced
isolated cardiac cells in adult rats. J Mol Cell Cardiol27: 2031-2041, ventricular failure in isolaled rat hearts. Am I Physiol 251: H133-
1995 H140, 1986
18. Piper HM, Probsl I, Schwartz P, Hutter FI, Spickermann PG: Culturing 32. Dar! AM, Sehomig A, Dielz R, Mayer E, Kubler W: Release of
of calcium stable adult eardiac myoeytes. J Mol Cell Cardiol 14: 397- endogenous eateeholamines in the isehemie myocardium. Cire Res
412,1982 55: 702-706, 1984
19. Harworth RA, Hunter DR, Berkoff HA: Contraclure in isolated adult 33. Sehomig A, Dart AM, Dielz R, Mayer E, Kubler W: Release of
rat hcart eells. Ro1c ofCa", ATP, and eompartmentation. Cire Res 49: endogenous eateeholamines in the isehemie myoeardium of the rat.
1119-1129, 1981 Part A: Loeally mediated release. Cire Res 55: 689-701,1984
20. Nakano M, Mann DL, Knowlton AA: Blocking the endogenous inerease 34. Rona G: Cateeholamine eardioloxicity I Mol Cell Cardiol 17: 291-
in HSP 72 increases suseeptibility to hypoxia and reoxygenation in 306, 1985
isolaled adult feline cardiomyocYles. Circulalion 95: 1523-1521, 1997 35. Ganote CE, Kaltenbaeh IP: Oxygen-indueed enzyme release: Early
21. Hallaq H, Sellmayer A, Smith TW, Leaf A: Proteetive effeet of events and a proposed mechanism. I Mol Cell Cardiol 11: 387-406,
eicosapenlaenoie acid on ouabain toxicily in neonatal rat eardiae 1979
myoeytes. Proc Natl Aead Sei USA 87: 7834-7838, 1990 36. Takeo S, Tanonaka K, Miyake K, Fukumoto T: Role ofATP metabolites
22. Hallaq H, Smith TW, LeafA: Modulation of dihydropyridine-sensitive in induetion of ineomplete recovery of eardiac eontraetile force after
calcium ehannels in heart eells by fish oil fatty acids. Proc Natl Aead hypoxia. Can I Cardiol 4: 193-200, 1988
Sei USA: 89: 1760--1764, 1992 37. Shen AC, Iennings RB: Kinetics of calcium aecumulation in acute
23. Xiao Y-F, Kano JX, Morgan JP, Leaf A: Bloeking effeets of poly- myocardial isehemic injury. Am I Pathol67: 441-452,1972
unsaturated fatty acids on Na' ehannels of neonatal rat ventricular 38. Steenbergen C, Murphy E, Levy L, London RE: Elevation in cytosolic
myoeytes. Proe Natl Aead Sei USA 92: 11000--11004, 1995 free calcium concentration early in myoeardial isehemia in perfused
24. Grynberg A, Foumier A, Sergiel IP, Athias P: Effeet of doeosa- ral heart. Circ Res 60: 700--707, 1987
hexaenoic acid and eieosapentaenoie acid in the phospholipids of rat 39. Zweier JL, Flaherty JT, Weisfeldt ML: Direet measurement of free
heart musele eells on adrenoeeptor responsiveness and mechanism. J radical generation following reperfusion ofisehemie myoeardium. Proc
Mol Cell Cardiol27: 2507-2520, 1995 Natl Aead Sci USA 84: 1404-1407, 1987
25. Gunn MD, ChangASA, Willerson JT, Buja LM, Chien KR: Mechanisms 40. Hearse DI, Garliek PB, Humphrey SM: Ischemie eontraemre of the
of aeeumulation of arachidonic acid in culmred myoeardial eells during myocardium: Mechanisms and prevention. Am J Cardiol 39: 986-993,
ATP depletion. Am J Physiol249: H1l88-H1l94, 1985 1977
26. Karmayzn M, Moff.t MP: Toxie properties of araehidonie acid on 41. Kloner RA, Ganole CE, Jennings RB: The 'no reflow' phenomenon
normal, isehaemie and reperfused hearts. Indirect evidenee for free after temporary eoronary oeclusion in the dog. I Clin Invest 54: 1496-
radieal involvement. Prost Leueo Med 17: 251-264, 1985 1508, 1974
Molecular and Cellular Biochemistry 188: 209-215,1998.
© 1998 Kluwer Academic Publishers.

Differential influence of fasting and BM13.907


treatment on growth and phenotype of pressure
overloaded rat heart
Heinz Rupp, Vijayan Elimban and Naranjan s. Dhalla
Institute 0/ Cardiovascular Sciences, St. Boniface General Hospital Research Centre, and Department 0/ Physiology,
Faculty 0/ Medicine, University 0/Manitoba, Winnipeg, Canada

Abstract
To examine metabolie influences on markers of the fetal phenotype of pressure overloaded rat heart, rats with stenosis of the
abdominal aorta were intermittently fasted for 10-12 weeks. Although intermittent fasting, which reduces insulin mediated
glucose uptake in the heart and moderate pressure overload of the left ventricle increased the proportion of myosin ß-heavy
chains (ß-MHC) and reduced the Ca2+-stimulated ATPase activity of sarcoplasmic reticulum (SR) to a similar extent, these
interventions had no additive effects when combined. Furthermore, addition of sucrose (0.8%) to the drinking water prevented
the changes in both the ß-MHC proportion and SR Ca2+-stimulated ATPase activity due to pressure overload or fasting. To
assess the effects of a drug which stimulates glucose-carrier translocation, rats with stenosis of the abdominal aorta were treated
with BM13.907 (50 and 100 mglkg daily for 10-12 weeks). This treatment amplified the left ventricular hypertrophy (+43 vs.
21 % of untreated rats) and increased the ß-MHC proportion. The SR Ca2+-stimulated ATPase activity of pressure overloaded
rats treated with BM13.907 (100 mglkg/day) was, however, not reduced compared with sham operated control rats. Thus, an
intervention which is known to stimulate insulin-mediated glucose-carrier translocation, but not glucose-carrier activation,
partially prevented the characteristic phenotype ofpressure overloaded hearts. These data provide further evidence in favor of
metabolie influences linked to glucose uptake on growth and phenotype ofthe pressure overloaded heart. (Mol Cell Biochem
188:209-215,1998)

Key words: hypertrophy, fasting, myosin, sarcoplasmic reticulum, BM13.907

Introduction isozyme V 3 • Although an increased adrenergic activity or


catecholamine agonists increase the proportion of myosin V I
Prolonged pressure-overload of the adult heart results in [2-4], such an approach would be limited by the reduced
hypertrophy of the terrninally differentiated cardiomyocytes. adrenergic responsiveness of failing hearts. Based on the
Although increased muscle mass represents a physiological finding that myosin VI can be increased by thyroid hormones,
response to the overload, there is increasing evidence that the thyroid analogues were developed which improve heart
hypertrophy process is associated with the expression of function shortly after myocardial infarction [5]. An alternative
genes which are not optimal for cardiac performance. In view approach has also been indicated from the observation that
ofthe 'fetal' genes which are normally expressed during fetal a high carbohydrate diet increased myosin V, in thyro-
and neonatal development [I], attempts have been made to idectomized rats [6] and that drinking water containing low
develop drugs which selectively prevent changes in markers amounts of sucrose increased myosin V I in normal rats [7].
of the fetal phenotype exemplified by a reduced proportion To enhance glucose oxidation by a drug intervention, a
of myosin isozyme VI and increased proportion of myosin carnitine palmitoyltransferase-l (CPT-l) inhibitor, etomoxir,

Addressfor o./JjJrints: N.S. Dhalla, Institute ofCardiovascular Seiences, SI. Boniface General Hospital Research Centre, 351 Tache Avenue, Winnipeg,
Manitoba, Canada R2H 2A6
210

was used; this treatment reduced mitochondrial uptake and constricted to 0.9 mm diameter and the other 16 rats in each
oxidation of long-chain fatty acids but increased glucose group were sham operated. The 16 rats of each group were
oxidation. In fact, treatment ofrats with etomoxir increased further divided into two groups that received either tap water
the proportion ofmyosin VI' sarcoplasmic reticulum (SR) (8 rats) or tap water containing 0.8% (w/v) sucrose.
Ca2+-pump activity [&--12], phosphoenzyme level ofSR Ca2+-
pump [11, 12] and mRNA abundance of SR Ca2+-pump
protein (SERCA2a) [13]. These molecular changes were Administration 01 BM13.907
associated with increased rate of contraction and rate of
relaxation of pressure overloaded hearts [ 14]. Since etomoxir In24 male Sprague-Dawleyrats (6 weeks old), the abdominal
improved heart function in patients with chronic heart failure aorta was constricted to 0.9 mm diameter and 24 rats were sham
at NYHA II-III stage [15], the rat can serve as a valid operated. In each group, 8 rats were treated with 50 mglkg body
screening model for metabolically active compounds despite weight BM 13.907, 8 rats were treated with 100 mglkg body
species dependent phenotype markers. No conclusion can weight BM 13.907 and 8 rats remained untreated. BM 13.907
currently be drawn on the cellular mechanisms leading to the was mixed with powdered rat chow and the dosage was
altered myocyte phenotype because a partial inhibition of adjusted on the basis that adult rat consumed 15 g food per day.
CPT -1 reduced fatty acid oxidation and increased glucose BM 13.907 [(±)-5-(4-Chlorophenyl)-2-(4-methylphenyl-
oxidation in a compensatory manner. In the present approach, sulfonyl)-4-pentynoic acid] was kindly provided by Dr. Peter
interventions were examined which affect glucoseuptake in Freund, Boehringer Mannheim, Germany. All rats were fedad
the heart. In the first series of experiments, rats with pressure libitum and received tap water. Animal care and experimental
overloaded left ventricles were intermittently fasted. It procedures were according to the institutional guidelines.
should be noted that both myosin heavy chain (MHC)
expression and SR Ca2+-stimulated ATPase activity were
affected similarly due to pressure overload [16,17]. Further- Determination 01 myosin isozymes
more, during fasting, the reduced insulin level has been shown
to result in a greatly reduced whole-body glucose utilization After 10-12 weeks, rats were killed by decapitation, hearts
including by the heart muscle [18]. It was, therefore, the removed and 20 mg portions of the left ventricle were frozen
purpose of this study to assess whether fasting and moderate in liquid nitrogen. The proportion of myosin isozymes was
pressure overload have additive effects when combined. determined by non-dissociating polyacrylamide gel electro-
Since we have previously shown that sucrose-supplemented phoresis in the presence of pyrophosphate [20]. Myosin
drinking water prevented changes in MHC expression and isozymes were stained with Coomassie brilliant blue R250
SR Ca2+-stimulated ATPase activity arising from inter- and gels were scanned using an Ultroscan laser densitometer
mittent fasting [17] and moderate pressure overload [16], (LKB; Bromma, Sweden). The isozymes were quantitated by
the question was also addressed whether sucrose exerts measuring peak heights. If an isozyme was less than 5%, the
similar effects when intermittent fasting and moderate isozyme profile was fitted to an envelope consisting ofthree
pressure overload were combined. In the second series of Gaussian curves which were integrated [20].
experiments, rats with pressure overloaded hearts were fed
ad libitum and treated with the anti-diabetic compound, BM
13.907, which is known to increase the translocation of Determination 01 sareoplasmie retieulum Ca 2 +-stimulated
glucose-carriers and lowers blood glucose in animal models ATPase aetivity
ofnon-insulin dependent diabetes mellitus [19]. This study
provides further evidence that signals linked to glucose Fragmented SR was isolated from fresh ventricles as described
uptake are determinants of growth and phenotype of the [16]. The total ATPase activity was assayed in a medium
pressure overloaded rat heart. containing 100 mM KCI, 5 mM NaN3, 5 mM MgCI2 , 0.01
mM free Ca2+, 5 mM ATP and 20 mM Tris-Cl, pH 6.8. To
obtain Ca2+-stimulated ATPase activity, the basal activity
Materials and methods measured in the presence of2 mM EGTA was subtracted from
the total activity.
Effeet 01 intermittent lasting and suerose administration

Thirty two male Sprague Dawley rats (6 weeks old) received Determination 01 serum parameters
a regular chow ad libitum and 32 rats received the chow every
other day with one day fasting in between (intermittent The first 5-7 ml ofblood from the trunk was collected and
fasting). In 16 animals of each group, the abdominal aorta was the serum was used for the determinations. Insulin was
211

measured using the Sigma kit no. 16-Uv, triglycerides using Regular Regular
Drinking water Drinking water
the Sigma kit no. 336 and total eholesterol using the Sigma 3.0 ~---"---~
kit no. 352 (Sigma, St. Louis, MO, USA). Thyroxine and 0' A 60 c
triiodothyronine were determined by fluoroimmunoassays ~2.5
~
(Delfia of Pharmacia, Fairfield, NJ, USA). Insulin was CD
'-..
* ~ 40
measured using a rat insulin RIA kit (Lineo Research, St. o;z.O I
Louis, MO, USA). 5:;;: 1.5
2
I 20
<>:l.
~

Statistical evaluation
300~------~ ~250r-------~
Equality of variances was ehecked by Coehran's C test. In E B *
E
'di ZOO
0
ease of unequal variances, a BoxCox transformation was $Z50
....J E
performed before analysis ofvariance (STATGRAPHICS of f- ~150
STSC, Roekville, MD). Multiple eomparisons were made by ): ZOO o
Dunean's new multiple range test. Values are mean ± S.D.
0'
E ~ 100
~

~ 150
Statistieal signifieance was assumed at p < 0.05.
~
Results
Fig. I. Effect of intennittent fasting on hearts of rats with stenosis of thc
EfJect 01 comhined intermittent lasting and pressure abdominal aorta receiving normal drinking water. 1\ - normal (sham-
overload operated) rats; AS - rats with abdominal aortic stenosis; rats were either
fed ad libitum (fed) or fasted for one day (fasted) followed by one day ad
Intermittent fasting for 10-12 weeks resulted in a reduced libitum food intake. *p < 0.05, rats with aortic stenosis vs. nonnal rats; **p
body weight. Tibia length which represents a marker of < 0.05, fasted vs. red rats. For comparison, data of ad libitum fed rats with
aortic constriction [16] and intermittently fasted rats [17] were included in
general body growth was not signifieantly reduced (not
the figure. The combination of intcnnittcnt fasting and aortic constriction
shown). The ratio ofleft ventricular-to-body weight was not has not been reported previously.
signifieantly affeeted by intermittent fasting. Left ventricular
weight was, however, signifieantly redueed when referred to
the length of the tibia (Fig. 1A, B). Constrietion of the eonstrietion fed ad libitum (Fig. 2A,B), it prevented the left
abdominal aorta indueed left ventrieular hypertrophy both in ventrieular weight inerease of intermittently fasted rats when
intermittently fasted and ad libitum fed rats irrespeetive of referred to tibia length (Fig. 2B). Thus, left ventrieular weight
whether left ventricular weight was compared with body of intermittently fasted rats with pressure overload was 597
weight or tibia length (Fig. IA, B). ± 30 g as compared with 651 ± 106 g of sham operated
Intermittent fasting signifieantl y inereased the proportion animals. Tibia length was 3.9 ± 0.1 emin pressure overloaded
of myosin V3 and eoncomitantly redueed the proportion of rats and 4.0 ± 0.2 em in sham operated rats.
myosin VI (not shown). Correspondingly, the proportion of The sucrose feeding prevented the increased ~-MHC
~-MHC [myosin V3 (%) + 0.5 myosin V2 (%)] was inereased proportion of intermittently fasted rats, of pressure overloaded
(Fig. IC). A eomparable increase in ~-MHC was observed rats and of rats where fasting was eombined with pressure
in pressure overloaded ventricles (Fig. I C). When pressure overload (Fig. 2C). Similarly, the suerose feeding prevented
overload was eombined with intermittent fasting, ~-MHC the depression in SR Ca 2+-stimulated ATPase aetivity in all
was, however, not further inereased (Fig. I C). Thus, the three groups (Fig. 2D). When sucrose was provided in the
pressure overload altered the proportion of~-MHC expression drinking water, left ventricles exhibited a similar MHC
only in rats fedad libitum. SR Ca2+-stimulatedATPase aetivity expression and SR Ca2+-stimulatedATPase aetivity irrespeetive
was depressed in intermittently fasled rats and rats with left of left ventrieular weight.
ventrieular pressure overload (Fig. ID). When both regimens
were eombined, no further deerease in the SR Ca2+-stimulated
ATPase activity was observed (Fig. lD). EfJect 01 BM 13.907 administration
To examine the effeets of additional suerose intake, rats
were provided 0.8 g/l suerose in the drinking water. While In the second series of experiments, rats with moderate
the suerose feeding had only a slight effect on left ventrieular eonstrietion of the abdominal aorta were treated daily with
weight referred to body weight in rats with abdominal aorta 50 or 100 mg/kg body weight BM 13.907. In sham operated
212

0.8% sucrose 0.8% sucrose was compared with body weight (Fig. 3A) or tibia length (Fig.
Drinking water Drinking water
3.0 , - - - - - = - - - - - - - , 3B).
~
(J1 A 60 c In the moderately hypertrophied left ventric1es of untreated
--; 2.5 rats, the proportion ofß-MHC was not significantly increased
[D
* E (Fig. 3C), but SR Ca2+-ATPase activity was significantly
"-
0;2.0
u 40
:r: reduced (Fig. 3D). The BM13.907 treatment had no significant
53: 1.5 :::<
~20 effect on MH C expression in sham operated rats but increased
significantly the ß-MHC proportion in rats treated with 50
:'J
or 100 mg/kg (Fig. 3C). The proportion ofmyosin isozyme
V3 was significantly increased and of myosin isozyme V 1 was
3 0 0 , - - - - - - - - , ~250,---------, significantly reduced in rats treated with 100 mg/kg body
E B 'E
';;; 200
weight BM13.907 (Table 2). The proportion of myosin V3
~250
--' E was linearly correlated with left ventricular weight of normal
I- "- 150
(5 rats and rats with stenosis ofthe abdominal aorta (Fig. 4). Rats
::::: 200
(J1

E ** ~ 100 with pressure overloaded left ventric1es treated with BM


'--" 150
'--"
13.907 exhibited a weight-proportional increase in myosin
~ 50
~ o
0..
V J (Fig. 4). In contrast, the depressed SR Ca 2+-ATPase
activity of ventric1es exhibiting 42% hypertrophy was not
~ O~~~~~~~~
reduced further by the BM13.907 treatment (Fig. 3D). In
pressure overloaded rats treated with 100 mg/kg BM 13.907,
Fig. 2. Effect of intermittent fasting on hearts of rats with stenosis of the the SR Ca 2+-ATPase activity did not differ from untreated
abdominal aorta reeeiving drinking water containing 0.8% (w/v) suerose. sham operated rats.
N - normal (sham-operated) rats; AS - rats with abdominal aortic stenosis; At 100 mg/kg body weight, BM13.907 significantly
rats were either fed ad libitum (fed) or fasted for one day (fed) followcd by
reduced serum triglyceride and cholesterol concentrations in
one day ad libitum food intake. *p < 0.05. rats with aortic stenosis vs.
normal rats; *'p < 0.05, fasted vs. fed rats. For eomparison, data of ad rats with left ventricular pressure overload (Table I). Serum
libitum red rals wilh aortic conslriclion [16] and intermillently [asled rals insulin and glucose concentrations were not significantly
[17] were included in the figure. The combination of intermittent fasting affected (Table I). Although serum thyroxine concentration
and aortic constriction has not been reported previously. was significantly reduced in sham-operated and pressure
overloaded rats by BMI3.907, serum triiodothyronine
concentrations were not significantly affected (Table 2).
rats, BM 13.907 had no significant effect on body weight and
left ventricular weight (Table 1). In rats with abdominal aortic
constriction (21 % left ventricular hypertrophy), the BM Discussion
13.907 treatment amplified the ventricular weight increase.
The degree of left ventricular hypertrophy was 43% in the In the present study, the question was addressed whether
50 mg/kg group and 42% in the 100 mg/kg group (Table I). interventions associated with an altered glucose uptake have
Similar changes were observed when left ventricular weight specific effects on the phenotype of pressure overloaded rat

Table 1. Growth characteristies and serum parameters ofrats treated with BM 13.907.

Body Tibia LVW Glucose Insulin Cholesterol Triglyceride


weight length
(g) (ern) (mg) (mg/dl) (pmo1/l) (mmo1/l) (mmo1/l)

Normal 614 ± 60 4.26 ± 0.15 1002 ± 86 121 ± 16 739 ± 187 2.0 ± 0.3 2.0 ± 0.7
AS 586 ± 57 4.27 ± 0.11 1209 ± 223* 117 ± 14 666 ± 178 1.8 ± 0.1 1.8 ± 0.7
Normal + 50mglkg
BM13.907 599 ± 72 4.31 ± 0.14 1058 ± 167 123 ± 9 689 ± 91 1.8 ± 0.6 IA±OA
AS + 50mg/kg
BM13.907 621 ± 67 4.30 ± 0.14 1435 ± 211*" 119± 7 598 ± 73 1.7 ± 0.7 1.5 ± 0.8
Normal + 100 mglkg
BM13.907 624 ± 52 4.32±O.12 1076 ± 122 118 ± 8 725 ± 151 1.6 ± 0.5 1.4±0.9
AS + 100 mglkg BM13.907 578 ± 38 4.28 ± 0.12 1423 ± 147**' 118 ± 6 607 ± 137 1.3 ± 0.3*- 0.8 ± 0.2*'

AS - rats with stenosis of the abdominal aorta; LVW - left ventricle weight. *p < 0.05, AS vs. sham operated normal rats; **p < 0.05, treated AS vs.
untreated AS; 'p < 0.05, treated AS vs. treated (same dosage) sham operated normal rats.
lW2II + 100 mg/kg BM 13.907 60~O~N----------------------------~
o Untreated
3.0,-----------,
I?ZZZJ + 50 mg/kg BM13.907 t::. N + 50 mg/kg
,.....
A * o N + 100 mg/kg
* C *
;2.5
rn
0' 60
*
40
• AS
.. AS
• AS
+ 50 mgjkg
+ 100 mg/kg
U
::c 40
........
0;2.0 :::2i
I
g
5 ""- 20 >'"
~ 1.5
20 o
~

1.0 0
,.....
c
E400 B 'E D
OL---------~ __________L -_ _ _ _ _ _ _ _ ~

~
* ~200 0.5 1.0
~ E Left ventricular weight (g)
$300 -::::,.
0

.~~
~
0'
E 100
....... 200 ....... Fig. 4. Correlation ofleft ventricular weight and proportion ofmyosin V,
Cl)
fornormal rats and rats with constriction ofthe abdontinal aorta. Rats were
~ [L
Ul
0
treated with 50 or 100 mglkg body weight BMI3.907. A linear regression
100
!;i: 0
NASNASNAS was calculated based on al! data points (y=-2.966 + 3.958 x, r= 0.73). N:
normal (sharn-operated) rats; AS-rats with abdominal aortic stenosis.
Fig. 3. Effect of BM13.907 administration in normal rats and rats with
stenosis of the abdominal aorta. Rats were treated with 50 or 100 mglkg
bodyweightBMB.907. 'p < 0.05, rats with aortic stenosis vs. the respective oxidation and the concomitant reduction in glucose oxidation
untreated or treated sharn operated normal rats. N - normal (sharn-operated) via a glucose-fatty acid cycle [21]. Since serum fatty acids
rats; AS - rats with abdontinal aortic stenosis.
and associated levels in triglycerides and cholesterol were
correlated with changes in the ß-MHC proportion, it was
left ventric1e. We have previously shown that insulin concluded that in streptozotocin-induced diabetes mellitus the
deficiency arising from streptozotocininduced diabetes main effect of etomoxir had to arise from lowering the
mellitus markedly increased the ß-MHC expression which depressant effect of increased fatty acids on cardiac glucose
was closely related to increased serum lipid levels [21]. oxidation [21].
When diabetic rats were treated with etomoxir, the partial There is increasing evidence that the MHC expression
normalization of MHC expression did not arise directly responds not only to prolonged but also intermittently reduced
from the partial CPT - 1 inhibition since bypassing the CPT- insulin levels. Interrnittent fasting for one day, which is
1 inhibition by feeding a medium-chain fatty acid diet did not associated with intermittently reduced insulin level increased
blunt the effect of etomoxir. Etomoxir 10wered the high serum ß-MHC expression [17, 23]. Increasing the durationoffasting
fatty acid and triglyceride concentrations probably via CPT- to 2 days induced only a small further increase in the V J
1 independent mechanisms. The reduced fatty acid supply proportion [24]. Low amounts of sucrose in the drinking
was expected to relieve the heart from an excessive fatty acid waterprevented the changes in MHC expression [17]. Since

Table 2. Myosin isozymes (VI' V2 and V,) and serum thyroid hormones (T3 and T4) ofrats treated with BM13.907.

VI V2 V, T, T,
(%) (%) (%) (nmoVI) (nm01l1)

Normal 65.5 ± 11.1 22.5 ± 6.1 12.0 ± 5.4 0.92 ± 0.22 117 ± 20
AS 57.6 ± 22.3 25.4 ± 7.7 17.0 ± 15.7 0.84 ± 0.23 112 ± 15
Normal + 50 mglkg
BM13.907 57.8 ± 12.6 25.6 ± 4.5 16.6 ± 8.3 0.85 ± 0.33 99 ± 13'
AS + 50mglkg 42.0 ± 19.9 29.8 ± 7.7 28.3 ± 15.4 0.77 ± 0.07 90 ± 11"
Normal + 100 mglkg
BM13.907 68.8 ± 12.9 21.7 ± 8.6 9.5 ± 4.6 0.70 ± 0.17 91 ± 12'
AS + 100 mglkll
BMB.907 42.1 ± 14.8' 31.0 ± 3.9' 26.9 ± B.4' 0.66 ± 0.13 91 ± 19"

AS - rats with stenosis of the abdontinai aorta; T3 - triiodothyronine; T4 - thyroxine. *p < 0.05, treated sharn operated normal rats vs. untreated sharn
operated normal rats; "p < 0.05, treated AS vs. untreated AS; 'p < 0.05, treated (100 mglkg) AS vs. treated (100 mglkg) sharn operated normal rats.
214

the serum glucose concentration in sucrose-fed rats was 10]. Lower doses (8 mg/kg body weight) did not increase
reduced on the day of refeeding, it appears that the sucrose cardiac growth significantly irrespective of the presence of
feeding had increased glucose utilization of the body [17]. press ure overload [14]. Differential influences of etomoxir
Possible candidates of a sucrose-mediated increase in glucose and BM 13.907 were observed also with respect to MHC
utilization are the incretins gastric inhibitory polypeptide and expression and SR Ca2 +-stimulated ATPase activity. In the
the truncated glucagon-like peptide-l [25]. The present case of BM 13.907, the increase in left ventricular weight was
finding that left ventricular hypertrophy was reduced in rats associated with a parallel increase in p-MHC. However, the
with aortic constriction when fasting was combined with SR Ca2+-stimulated ATPase activity was not reduced. The
sucrose administration demonstrates that cardiac hypertrophy pressure overloaded rats treated with 100 mg/kg BM13.907
is sensitive to various metabolie interventions. More work is exhibited an SR Ca2+-stimulated ATPase activity which did
required to characterize the suerose dependent reduetion in not differ from that of sham-operated rats. In the case of
the degree of left ventrieular hypertrophy. etomoxir, the ß-MHC expression was redueed [9,14] and SR
It was an intriguing observation that the MHC expression Ca2+-stimulatedATPase was increased at a dose which reduced
of pressure overloaded heart with moderate left ventricular the ß-MHC expression [9]. Furthermore, a dissociation
hypertrophy responded also to feeding of low amounts of between cardiac growth and changes in MHC expression and
sucrose [16]. In the present eomparative study it is shown that SR Ca 2+-stimulated ATPase activity was observed after
intermittent fasting and constriction of the abdominal aorta treatment with 8 mglkg body weight enantiomeric etomoxir.
had similar effects on MHC expression and the assoeiated Although left ventricular weight was not affected, rate of SR
changes in SR Ca 2+-stimulatedATPase activity. When both Ca2+-uptake ofnormal rats [8] as weil as rate parameters of
interventions were combined, the changes were, however, contraetion and relaxation of rats with asccnding aorta
not additive. If intermittent fasting and pressure overload constriction were increased [14]. Compared with CPT-I
of the left ventricle affected the MHC expression by inhibition, the BM13.907 treatment was thus less effective
different mechanisms, one would have expected additive in blunting the appearance ofthe fetal phenotype ofpressure
effects. It thus appears that in pressure overloaded heart overloaded heart.
metabolie signals are changed in a manner that blunt any It appears that the treatment with BMI3 .907 induced some
further effect of intermittent fasting. In favor of this effects observed for growth hormone and the associated
contention is the finding that feeding of low amounts of insulin-Iike growth factor-I. In rats bearing a growth hormone
sucrose had similar effects in moderate pressure overload, secreting tumor, the enlarged heart exhibited a markedly
intermittent fasting and in rats where both interventions increased proportion ofmyosin V3 [26,27]. Furthermore, the
were combined. In support ofmetabolic signals affected by insulin-like growth factor-I [28] or insulin-like growth factor-
the ventricular pressure load is the observation that the CPT- II [29] induced hypertrophy ofneonatal rat cardiomyoeytes
I inhibitor etomoxir had a more pronounced action in the and increased mRNA level of skeletal <X-actin. The expression
left vs. the less loaded right ventricle [11] and in the pressure of the skeletal specific isoform of <x-actin is another eharacter-
overloaded left ventricle versus the normalleft ventricle [13, istic feature of pressure overload hypertrophy which can be
14]. prevented by etomoxir treatment [13]. Although the serum
To further characterize metabolie signals, BM13.907 was lipid levels in pressure overloaded animals, unlike the control
administered which is expected to enhance certain insulin group, are decreased upon treatment with BMI3.907, the
influences. BM13.907 represents an a-activated carbonic exact reason for these changes in pressure overloaded
acid, which exhibits a blood glucose lowering action in animals is not clear at present. In conclusion, the present
experimental models of non-insulin dependent diabetes study demonstrates that both cardiac growth and molecular
mellitus [19]. BM13.907 induced glucose-carrier translocation phenotype of the pressure overloaded rat heart react sensitively
in fat cells to the same extent as insulin but did not stimulate to interventions that are known to modify glucose uptake.
glucose carrier activation [19]. The glucose-transport activity Whereas cardiac growth was not significantly affected by
was 40-50% ofthe insulin effect [19]. In the present study, intermittently reducing insulin influences, it was increased
the BM13.907 treatment had no significant effect on left by BM 13 .907 known to increase glucose transporter transloca-
ventricular weight in sham operated rats, but increased left tion. The finding that a combination of pressure overload
ventricular weight in rats with stenosis of the abdominal and intermittent fasting had no additive effects on MHC
aorta. It thus appears that any growth-promoting action of expression and SR Ca2+-stimulatedATPase activity supports
BM 13.907 required the presence of growth stimuli associated the view that metabolic signals linked to glucose metabolism
with pressure overload. In this respect, it should be mentioned are already compromised in the pressure overloaded heart of
that a treatment with the biologically enantiomeric etomoxir ad libitum fed rats. It is evident, however, that the compound
(15 mg/kg body weight) increased heart weight to a similar BM13.907 could only partially induce changes which are
degree in sham-operated and pressure overloaded hearts [9, opposite to those observed in pressure overloaded hearts.
215

Thus, BM13.907 blunted the decrease in SR Ca2+-stimulated 12. Rupp H, Schulze W, Vetter R: Dietary medium-ehain triglycerides can
ATPase but had no effect on the weight proportional increase prevent ehanges in myosin and SR due to CPT-I inhibition by etomoxir.
Am J Physiol 269: R630--R640, 1995
in ß-MHC expression. Further work is required to examine
13. Zarain-Herzberg A, Rupp H, Elimban V. Dhalla NS: Modification of
the differential action of BM 13 .907 and etomoxir on the fetal sarcoplasmic reticulum gene expression in pressure overload cardiac
phenotype of pressure overloaded rat heart. hypertrophy by etomoxir. FASEB J 10: 1303-1309,1996
14. Tureani M, Rupp H: Etomoxir improves the function of pressure
overloaded rat heart. Circulation (in press) 1997
15. Schmidt-Schweda S, Holubarsch Ch: First clinical trial with etomoxir
Acknowledgements in patients with chronic heart failure NYHA 1I-1Il. J Mol Cell Cardiol
29: ASS, 1997
The study was supported by the Medical Research Council 16. Rupp H, Elimban V, Dhalla NS: Sucrose feeding prevents ehanges in
myosin isoenzymes and sarcoplasmic retieulum Ca2+-pump ATPase
of Canada (MRC Group in Experimental Cardiology). HR
in pressure-Ioadedrat heart. BiochemBiophys Res Commun 156: 917-
was supported by the Deutsche Forschungsgemeinschaft 923,1988
(Ru 24517-1) and the Science & Technology Cooperation 17. Rupp H, Elimban V, Dhalla NS: Diabetes-like action of intermittent
Germany/Canada (BMBF/HM4). We acknowledge helpful fasting on sarcoplasmic reticulum Ca2+-pump ATPase and myosin
discussions on this project with Dr. P. Freund ofBoehringer iso enzymes can be prevented by sucrose. Biochem Biophys Res
Commun 164: 319-325, 1989
Mannheim, Germany.
18. Isaad T, Penicaud L, Ferre P, Kande J, Baudon MA, Girard J: Effects
of fasting on tissue glucose utilization in eonscious resting rats. Major
glucose-sparing effect in working musc1cs. Biochem J 246: 241-244,
References 1987
19. Obermaier-Kusser B, Muhlbaeher C, Mushack J, Seffer E, Ermel B,
Machicao F, Schmidt F, Häring RU: Further evidence far a two-step model
1. Penniea D, King KL, Shaw KJ, Luis E, Rullamas J, Luoh SM, of glucose-transport regulation. Inositol phosphate-oligosaceharides
Darbonne WC, Knutzon DS, Yen R, Chien KR: Expression cloning of regulate glucose-carrier activity. Bioehem J 261: 699-705, 1989
eardiotrophin I, a eytokine that induees eardiae myocyte hypertrophy. 20. Rupp H, Dictz K: Mathematieal models of myosin heterodimer
Proc Nat! Acad Sei USA 92: 1142-1146, 1995 formation in the rat heart during thyroid hormone alterations. Cire
2. Rupp H: The adaptive changes in the isoenzyme pattern of myosin Res 68: 27-37, 1991
from hypertrophied rat myoeardium as a resul! of pressure overload 21. Rupp H, Elimban V, Dhalla NS: Modifieation of myosin isozymes
and physical training. Basic Res Cardiol 76: 79-88, 1981 and SR Ca'+ -pump ATPase of the diabetic rat heart by lipid-Iowering
3. Rupp H, Berger HJ, Pfeifer A, Werdan K: Effeet of positive interventions. Mol Cell Biochem 132: 69-80,1994
inotropic agents on myosin isozyme population and mechanical 22. Zhou YP, Priestman DA, Randle P J, Grill VE: Fasting and decreased
activity of eultured rat heart myocytes. Circ Res 68: 1164-1173, B cell sensitivity: Important role for fatty acid-induced inhibition of
1991 PDH activity. Am J Physiol 270: E988-E994, 1996
4. Gupta MP, Gupta M, Stewart A, Zak R: Aetivation of alpha-myosin 23. Rupp H, Wahl R: Influence of thyroid hormones and eatecholamines
heavy chain gene expression by cAMP in eultured fetal rat heart on myosin of swim-exereised rats. J Appl Physiol 68: 973-978, 1990
myoeytes. Bioehem Biophys Res Commun 174: 1196--1203, 1991 24. Rupp H, Maisch B, Brilla CG: Schedule-induced psychological stress
5. Morkin E, Pennoek GD, Raya TE, Bahl JJ, Goldman S: Development and moleeular structures of cardiomyocytes. Am J PhysioI272: R 776--
of a thyroid hormone analogue for the treatment of congestive heart R782, 1997
failure. Thyroid 6: 521-526, 1996 25. Fukase N, Takahashi H, Manaka H, Igarashi M, Yamatani K, Daimon
6. Sheer D, Morkin E: Myosin isoenzyme expression in rat ventricle: M, Sugiyama K, Tominaga M, Sasaki H: Differences in glucagon-like
effeets of thyroid hormone analogs, eateeholamines, glucocorticoids peptide- land GIP responses following sucrose ingestion. Diabetes
and high earbohydrate die!. J Pharmaeol Exp Ther 229: 872-879, Res Clin Pract 15: 187-195, 1992
1984 26. Timsit J, Riou B, Bertherat J, Wisnewsky C, Kato NS, Weisberg AS,
7. Rupp H, Wahl R, Jaeob R: Remodelling ofthe myocyte at amolecular Lubetzki J, Lecarpentier Y, Winegrad S, Mercadier JJ: Effects of
level - Relationship between myosin isoenzyme population and chronic growth hormone hypersecretion on intrinsic eontractility,
sarcoplasmie reticulum. In: N.S. Dhalla, G.N. Pierce, R.E. Beamish energetics, isomyosin pattern, and myosin adenosine triphosphalase
(eds). Heart Function and Metabolism. Martinus NijhoffPubl., Boston, activity of rat left ventricle. J CI in Invest 86: 507-515, 1990
1987, pp 307-318 27. Mayoux E, Ventura Clapier R, Timsit J, Bchar Cohen F, Hoffmann C,
8. Rupp H, Wahl R, Hansen M: Influence of diet and camitine palmitoyl- Mereadier JJ: Mechanical properties of rat cardiae skinned fibers are
transferase I inhibition on myosin and sarcoplasmic reticulum. J Appl altered by ehronic growth hormone hyperseeretion. Circ Res 72: 57-
Physiol72: 352-360, 1992 64,1993
9. Rupp H, Elimban V, Dhalla NS: Modifieation of subcellular organelles 28. Ito H, Hiroe M, Hirata Y, Tsujino M, Adaehi S, Shichiri M, Koike A,
in pressure-overloaded heart by etomoxir, a carnitine palmitoyl Nogarni A, Mammo F: Insulin-hke growth factor-I induces hypertrophy
transferase I inhibitor. FASEB J 6: 2349-2353, 1992 with enhanced expression of musc1e specifie genes in eultured rat
10. Rupp H, Jacob R: Metabolieally-modulated growth and phenotype of cardiomyocytes. Circulation 87: 1715-1721, 1993
the rat heart. Eur Heart J 13 (Supp!. D): 56--61, 1992 29. Adachi S, Ito H,Akimoto H, Tanaka M, Fujisaki H, Mammo F, Hiroe
11. Vetter R, Rupp H: CPT-I inhibition by etomoxir has a ehamber-related M: Illsulill-like growth factor-lI induees hypertrophy with increased
action on cardiae sarcoplasmic reticulum and isomyosins. Am J Physiol expression of muscle specific genes in cultured rat cardiomyocytes. J
267: H2091-H2099, 1994 Mol Cell Cardiol26: 789-795, 1994
Molecular and Cellular Biochemistry 188: 217-223, 1998.
© 1998 Kluwer Academic Publishers.

On the mechanism of the phospholipase C-


mediated attenuation of cardiolipin biosynthesis in
H9c2 cardiac myoblast cells
Fred Y Xu, Sherrie L. Kelly, William A. Taylor and Grant M. Hatch
Department 0/ Pharmacology and Therapeutics, University 0/ Manitoha, Winnipeg, Manitoha, Canada

Abstract
The effect of phospholipase C treatment on eardiolipin biosynthesis was investigated in intaet H9c2 cardiac myoblasts. Treatment
of cells with phosphatidylcholine-specific Clostridium welchii phospholipase C reduced the pool size of phosphatidy1choline
compared with controls whereas the pool size of cardiolipin and phosphatidylglycerol were unaffected. Pulse labeling
experiments with [1 ,VH]glycerol and pulse-chase labeling experiments with [1 ,VH]glycerol were performed in cells incubated
or pre-incubated in the absence or presence of phospholipase C. In all experiments, radioactivity incorporated into cardiolipin
and phosphatidylglycerol were reduced in phospholipase C-treated cells with time compared with controls indicating attenuated
de novo biosynthesis ofthese phospholipids. Addition of 1,2-dioctanoyl-sn-glycerol, a cell permeable 1,2-diacyl-sn-glycerol
analog, to cells mimicked the inhibitory effect ofphospholipase C on cardiolipin and phosphatidylglycerol biosynthesis from
[1,3- 3H]glycerol indicating the involvement of 1,2-diacyl-sn-glycerol. The mechanism for the reduction in cardiolipin and
phosphatidylglycerol biosynthesis in phospholipase C-treated cells appeared to be a decrease in the activities of phosphatidic
acid:cytidine-5'triphosphate cytidylyltransferase and phosphatidylglycerolphosphate synthase, mediated by elevated 1,2-diacyl-
sn-glycerol levels. Upon removal of phospholipase C from the incubation medium, phosphatidylcholine biosynthesis from
[methyPH]choline was markedly stimulated. These data suggest thatde novo phosphatidylglycerol and cardiolipin biosynthesis
may be regulated by 1,2-diacyl-sn-glycerol and support the notion that phosphatidylglycerol and cardiolipin biosynthesis may
be coordinated with phosphatidylcholine biosynthesis in H9c2 cardiac myoblast cells. (Mol Cell Biochem 188: 217-223, 1998)

Key words: cardiolipin biosynthesis, phospholipase C, H9c2 cardiac myoblast cells

Abbreviations: CL - cardiolipin; PLC - phospholipase C; DG - 1,2-diacyl-sn-glycerol; PG - phosphatidylglycerol; PGP -


phosphatidylglycerolphosphate; CDP-DG - cytidine-5' diphosphate-l ,2-diacyl-sn-glycerol; PA - phosphatidic acid; PC -
phosphatidylcholine; PE - phosphatidylethanolamine; PS - phosphatidylserine; PI - phosphatidylinositol; SM - sphingomyelin;
PLA2- phospholipase A2

Introduction by the mitochondrial CL synthase [5]. The CL synthase was


recently purified to homogeneity from rat liver by Schlame and
Cardiolipin (CV) is an important structural and functional Hostetler [6]. Mammalian PG and CDP-DG are synthesized
phospholipid localized in the mitochondria (for reviews see via the Kennedy pathway first described in rat liver [7]. In the
[1,2]). In the rat heart, CL comprises approximately 15% of first step ofthis pathway, PA and CTP are converted to CDP-
the entire phospholipid mass of this organ [3, 4J. PG and DG via the enzyme PA:CTP cytidylyltransferase. In the heart,
CDP-DG are the immediate precursors of CL and the pulse labeling and pulse-chase labeling experiments with
condensation of these two compounds to form CL is catalyzed [1,3- 3H]glycerol or [32PJorthophosphate indicated that the

Address for ofJprints: G.M. Hateh, Department of Phannacology and Therapeutics, Faculty of Medicine, University of Manitoba, Room A307 Chown
Building, 770 Bannatyne Avenue, Winnipeg, Manitoba, Canada R3E OW3
218

conversion of PA to CDP-DG was the rate-limiting step of Rat heart myoblastic H9c2 cells
CL biosynthesis [4]. This was later confirmed for H9c2
cardiac myoblast cells [8]. When cells were depleted ofCTP, H9c2 rat heart myoblast cells [17] were obtained from the
CL biosynthesis from [1,J-JH]glycerol was dramatically American Type Culture Collection. They were cultured in
reduced as a result of a lowered in vivo conversion ofPA to Dulbecco's modified eagle medium supplemented with 10%
CDP-DG. Finally, CDP-DG and glycerol-3-phosphate (by vol) heat-inactivated newborn calf serum, 100 !lg/ml
condense to form PGP which is then rapidly converted to penicillin G, 10 !lg/ml streptomycin and 0.25 !lg/ml ampho-
PG [7]. tericin B. Cell cultures were maintained at 37°C saturated
The regulatory mechanisms that govern CL biosynthesis with humidified air air/5% carbon dioxide. Each dish of cells
in the heart and mammalian cells remain largely unknown. was subcultured at 1:5 ratio and confluence was usually
The modulation of PA:CTP cytidylyltransferase and PGP obtained after 4 days of incubation. In all experiments, cells
synthase activities in vitro by PC is weil documented [9- at 70-80% confluence were made quicsccnt by incubation
11]. For example, exogenous PC addition to PA:CTP with a serumfree medium for 12 h (overnight) prior to
cytidylyltransferase solubilized from bovine brain micro- addition of PLC. All cell incubation procedures were
somes stimulated enzyme activity 1.7-fold compared with performed at 3TC.
controls [9]. In rat liver, PC addition to solubilized prepara-
tions ofmitochondrial PA:CTP cytidylyltransferase stimulated
enzyme activity 2-fold compared with controls [10]. Addition, PLC treatment and celliabeling studies
of egg PC to partially purified preparations of rat liver
mitochondrial PGP synthase increased enzyme activity 26% PLC was dissolved in a serum-free medium to obtain a stock
[11]. There was no information on whether alterations in PC solution of 25 Ulml. Aliquots of the stock were added to
levels in vivo affected PA:CTP cytidylyltransferase or PGP quiescent cells maintained in serum-free Dulbecco's modified
synthase activities. Furthermore, it was unknown if altera- eagle medium. In pulse-chase experiments, H9c2 cells in 60
tion of the cellular PC content would effect the de novo mm petri -dishes were incubated for 2 h with 2 ml of medium
biosynthesis of PG and CL. We investigated this in H9c2 containing 0.1 !lM [1 ,J-3H]glycerol (5 !lCi/ml). The medium
cardiac myoblast cells treated with PC-specific PLC. We was removed and the cells were washed twice with 3 ml of
demonstrate that the biosynthesis of new PG and CL are fresh medium and then incubated with 0.1 !lM glycerol for
attenuated when cells are treated with PC-specific PLC. In 30 min in the absence or presence of PLC (0.1 U/ml). In
addition, we provide evidence that the mechanism for this pulse-Iabeling experiments, H9c2 cells were incubated for 30
is a reduction in the activity ofPA:CTP cytidylyltransferase min with 2 ml of medium containing 0.1 !lM [1 ,VH]glycerol
and PGP synthase activities mediated by elevated cellular (10 !lCi/ml) in the absence or presence of PLC (0.1 U/ml).
DG levels. Our results support the notion that CL bio- In choline labeling experiments, the cells were incubated for
synthesis in H9c2 cells may be coordinated with that ofPC 30 min with 2 ml of medium in the absence or presence of
biosynthcsis. PLC (0.1 U/ml) for 30 min. The medium was removed and
the cells were washed twice with 3 ml of fresh medium. Cells
were then incubated with for 30 min with 2 ml of medium
Materials and methods containing 28!lM [methyPH]choline chloride (2 !lCi/dish).
The medium was removed and the cells were washed twice
Materials with 3 ml of fresh medium and subsequently incubated for
30 min with 2 ml of medium containing 28 !lM choline
[MethyPH]choline chloride was obtained from New England chloride.
Nuclear Division ofDupont (Mississauga, Ontario, Canada).
[1,J-JH]Glycerol and [ 14 C-U]glycerol-3-phosphate were
obtained from Amersham Canada Limited (Oakville, Ontario, Harvesting 0/ H9c2 cells, extraction and analysis 0/ lipids
Canada). PLC (Clostridium welchii), bovine serum albumin,
1,2-dioctanoyl-sn-glycerol (DiC8) and all lipid standards The cells were washed with 3 ml of ice cold phosphate
were obtained from Sigma Chemical Company (St. Louis, buffered saline. Two ml ofmethanol:water (1:1, v/v) was
MO., U.S.A.). Cell-culture media and reagents were products added to the dish and the cells were removed, using a rubber
of Canadian Life Technologies Inc. (GIBCO) (Burlington, policeman, into screw cap tubes. The suspension was vortexed
Ontario, Canada). Thin-Iayer chromatographie plates (silica twice and a 25 !ll aliquot taken for the determination of protein.
gel G) were from Fisher Scientific (Edmonton, Alberta, Lipids were extracted by a modified method ofFo1ch et al.
Canada). All other biochemicals were of analytical grade and [12]. Two ml of chloroform and 0.5 ml of 0.9% NaCI was
obtained from Sigma Chemie al Co. added to the suspension which was then vortexed and
219

centrifuged at 1,000 g for 10 min. The aqueous phase was described [17]. Student's t-testwas used forthe determination
removed and the organic phase was washed twice with 2 ml of significance. The level of significance was defined as p <
of chloroform:methanol:water (3 :48:47, by vol.). The aqueous 0.05.
phase was removed and the organic phase dried under nitrogen
and resuspended in 100 111 of chloroform:methanol (2: 1, v/v).
An aliquot ofthe lipid suspension was placed onto a thin-Iayer Results
chromatography plate and phospholipids were separated in
two-dimensions as described [3]. DG was separated in a To determine if PLC-treatment of H9c2 cells affected the
solvent system containing petroleum ether: ethyl ether:acetic CL pool size, cells were incubated for 30 min in the absence
acid (80:20:1, by vol.). Lipid spots on the thin-Iayer plates or presence of 0.1 Ulml PC-specific PLC and the pool size
were visualized by iodine vapor and the silica gel removed of phospholipids determined. Greater than 98% of the cells
and placed into scintillation vials. Five ml of Ecolite® treated with PLC in this manner excluded Trypan blue. In
scintillant was added and the radioactivity determined after addition, treatment of cells with PLC did not significantly
a 24 h period. In some experiments the silica gel was enhance medium lactate dehydrogenase activity (data not
removed for the determination of phospholipid phosphorus. shown).ATP and CTP content in these cells were 3.9 ± 0.5 mM
In other experiments several dishes of H9c2 cells were and 0.32 ± 0.07 mM (average ofthree experiments), respec-
pooled and CL from these cells isolated by two-dimensional tively, and were unaltered by PLC-treatment. Thus, the H9c2
thin-Iayer chromatography as described above. The silica cells remained viable throughout the experiment. Treatment of
gel corresponding to CL was removed and CL was extracted cells with 0.1 O/ml PLC for 30 min caused a 32% reduction
from the gel by washing three times with 3 ml of chloro- in the pool size ofPC compared with controls (Table 1). The
form:methanol (2: I, v/v). pool sizes of the other major membrane phospholipids CL,
PG, PE, PS, PI and SM were not affected. In addition, the
pool size of CDP-DG was 4.9 nmoles/mg and unaltered by
Subcellular fractionation and assay of enzymes PLC-treatrnent. Thus, Clostridium welchiiPLC incubation of
H9c2 cells specifically reduced PC content only. In some
All isolation procedures were performed at 4OC. H9c2 cells experiments the medium containing 0.1 Ulml PLC was
were incubated for 30 min in the absence or presence ofO.1 removed from the cells and the cells were further incubated
Ulml PLC and subsequently a 10% homogenate in 10 mM for another 30 min with fresh medium in the absence ofPLC.
Tris-HCI, pH 7.4, 0.25 M sucrose, 0.145 M NaCI and 1 mM In these experiments, the PC content was restored to control
EDTA was prepared using a Dounce A homogenizer. The levels. Thus, after removal of PLC the cellular levels of PC
homogenate was centrifuged at 1,000 x g for 5 min (Sorvall recovered indicating increased resynthesis to the parent
RC-5 Superspeed Refrigerated Centrifuge with SS-34 rotor) molecule to restore depleted cellular pools of PC.
to pellet debris and the resulting supematant designated the To determine ifPLC-treatrnent ofH9c2 cells effected CL
post homogenate fraction. The post homogenatc fraction was biosynthcsis, cells were incubated for 30 min with [1,3-
used for assay of PA:CTP cytidylyltransferase since this 3H]glycerol in the absence or presence ofO.l O/ml PCspecific
activity occurs in both mitochondrial and microsomal fractions. PLC and the radioactivity incorporated into PG and CL
In some experiments, the post homogenate fraction was determined. The presence ofPLC in the medium resulted in
centrifuged at 10,000 x g for 15 min. The resulting pellet was a 30 and 32% decrease in radioactivity incorporated into PG
resuspended in 50 mM Tris-maleate, pH 6.5, 0.1 M KC1, 10 and CL, respectively, compared to controls (Fig. IA). Cells
mM MgCI 2 , 0.5% Triton X-IOO and 10% glycerol and were then incubated for 2 h with [1,VH]glycerol and then
designated the mitochondrial fraction. PA:CTP cytidylyl- incubated for a further 30 min in the absence or presence of
transferase, PGP synthase, PGP phosphatase and CL synthase PLC. The presence of PLC in the medium resulted in a 30
were assayed as described [8]. and 29% decrease in radioactivity incorporated into PG and
CL, respectively, compared to controls (Fig. lB). Thus, in the
presence ofPLC, radioactivity incorporated into PG and CL
Other procedures from [1,3- 3H]glycerol was decreased in both continuous-
pulse and pulse-chase labeling experiments indicating
Lactate dehydrogenase activity in the medium was measured reduced de nova biosynthesis of these phospholipids. Cells
as described [13]. Cell protein was determined by the were incubated for 2 h with [I ,3- 3 H]glycerol and then
method ofLowry et al. [14]. Enzyme protein was determined incubated for a further 30 min in the absence or presence of
by the method ofBradford [15]. Phospholipid phosphorus PLC. The presence ofPLC in the medium resulted in a 1.7-
content was measured by the method ofRouser et al. [16]. fold increase in radioactivity incorporated into DG and an
ATP and CTP content in cells was determined as previously 27% decrease in radioactivity incorporated into PC compared
220

t'
3T····.... ·······················.. ···.. ·1 7,---~~---·············_···,,·,,·..--~--_···········--,

1'1 I
1
A
"!S : .),,!
c.Y
:.........L.~.... L ........ .

CL PG C~ PG

U+~~~~-' ...... ·.... ·;

Fig. 1. The effect of PLC-treatment of H9c2 cells on [1,3-3H]glycerol


incorporated into PG, CL, OG and PC. (A) H9c2 eells were ineubated
with [I ,3- 3H]glyeerol for 30 min in the absence (open bars) or presenee
(hatched bars) ofO.1 Ulml PLC and radioactivity incorporated into PG and
CL determined. B, C and O. H9c2 cells were incubated with [I ,3- 3H]glycerol iO
for 2 h and then incubated in the absence (open bars) or presenee (hatched
bars) ofO.1 Ulml PLC for 30 min and the radioaetivity incorporated into DG{uM}
PG and CL (B), OG (C) and PC (0) determined. Results represent the
mean ± S.O. of 4 experiments. *p < 0 05.
Fig. 2. Tbe effeet of exogenous OG and PLC-treatment of Hge2 cells on
PA:CTP cytidylyltransferase activity. (A) PA:CTP cytidylyltransferase
to controls (Fig. IC, lD). The presence of PLC in the activity was determined in post-homogenate fractions prepared from Hge2
cells ineubated for 30 min in Ihe absence (open bars) or presenee (hatehed
medium did not effect radioactivity incorporated into PE, PS
bars) of 0.1 V/mi PLC. B. PA:CTP eytidylyltransferase aetivity was
or PI (data not shown). Thus, in H9c2 cells PLC generated determined in post-homogenate fractions incubated in the presence of
DO fromPC. various concentralions ofOG. Results represent the mean of2 experiments
The mechanism for the reduction in PG and CL biosynthesis with results varying by less than 15%.
in H9c2 cells was examined. Previous studies had indicated
that in vitra PA:CTP cytidylyltransferase and POP synthase
activities could be modulated by lipids [9-11]. We examined DO to mitochondrial fractions prepared from H9c2 cells
if the products of PC-specific PLC hydrolysis, DO and resulted in a concentration-dependent decrease in POP
phosphocholine, effected the activities of these enzymes. synthase activity compared to controls (Fig. 3B). In contrast,
Cells were incubated with PLC for 30 min and post- addition of phosphocholine to subcellular fractions prepared
homogenate and mitochondrial fractions prepared and the from H9c2 cells did not effect enzyme activities (data not
activities of PA:CTP eytidylyltransferase, POP synthase, shown). Thus, in vi va generated DO inhibited PA:CTP
POP phosphatase and CL synthase determined. Treatment cytidylyltransferase and POP synthase activities and the
of cells with PLC resulted in a 38 and 57% decrease in inhibition ofthese enzymes activities by exogenous DO was
PA:CTP cytidylyltransferase and POP synthase activities, conccntration-dcpendent.
respectively, compared to controls (Figs 2A and 3A). POP The above studies suggested that elevated cellular DO
phosphatase and CL synthase activities were unaltered by levels inhibited PG and CL biosynthesis. To test this directly,
PLC-treatment (data not shown). Addition of various we examined whether DG addition itselfto H9c2 cel1s would
concentrations ofDO to post-homogenate fractions prepared effect PO and CL biosynthesis. We incubated cells with [1,3-
from H9c2 cells resulted in a concentration-dependent 3H]glycerol for 4 h in the absence or presence ofDiC8, a cell
decrease in PA:CTP cytidylyltransferase activity compared permeable DG analog. Treatment of cells with DiC8 was
to controls (Fig. 2B). Addition of various concentrations of shown to cause an increase in the cellular level of this lipid
221

120,-------------------------------~
IH

".ll

-fM
C
e IJ.S
100
'\

-
1
0.3
\
\
\
C
O.Z \
..
'"
tl.:
~
tl.tl

40
f.l.®

20~--_r----_.----~----~------ __--~
.S ~,ß~
o 25 50 75 100
5
--S
~
9.:>0
DieB (uM)

Fig. 4. The effect of DiC8 on PG and CL biosynthesis in H9c2 cells.


~s:< G.45
H9c2 cells were incubated with [1,3- 3H]glycerol for 4 h in the absence or
presence of various concentrations of DiC8 and the radioactivity
incorporated into PG and CL determined. Results are expressed as a
percentage of contra!. Contrals: CL ~ 5141 dpm/mg pratein; PG ~ 9882
IJ.4I.l dpm/mg protein. Results represent the mean of 2 experiments with results
varying by less than 15%.

cells with PLC inhibits PG and CL biosynthesis and that this


is due to a decrease in activities of PA:CTP cytidylyl-
Fig. 3. The effect of exogenous DG and PLC-treatment ofH9c2 cells on
transferase and PGP synthase mediated by elevated cellular
PGP synthase activity. (A) PGP synthase activity was determined in
mitochondrial fractions prepared from H9c2 cells incubated for 30 min in DG.
the absence (open bars) or presence (hatched bars) ofO.1 U/ml PLC. B.
PGP synthase activity was determined in mitochondrial fractions incubated
in the presence of various concentrations ofDG. Results represent the mean
of 2 experiments with results varying by less than 15%.
Tab!e 1. Elfect ofPLC-treatment on the pool-size of phospholipids in H9c2
cardiac myoblast cells.

[26,27]. Treatment of cells with DiC8 inhibited PG and CL Contral PLC-treated


A B
biosynthesis from [1,3- 3H]glycerol in a concentration-
dependent manner (Fig. 4). Thus, it was likely that DG nmol/mg pratein
generated from cellular PC was inhibiting de novo CL CL 7.49 ± 0.23 7.39 ± 0.21 7.52 ± 0.46
PC 81.3 ± 3.67 54.9 ± 2.81 * 75.9 ± 3.19
biosyntheis.
PE 51.1±2.13 52.3 ± 2.38 50.5 ± 2.26
The studies from Table 1 indicated elevated resynthesis PG 6.84 ± 0.22 6.59 ± 1.22 6.25 ± 1.59
of PC upon removal of PLC from H9c2 cells. We tested if PA 4.62 ± OAI 4.34 ± 0.51 4.12 ± 0.80
removal ofPLC from H9c2 cells resulted in an elevation in PS 8.25 ± 0.33 8.11 ± 1.54 8.34 ± 1.34
actual de novo PC biosynthesis. H9c2 cells were incubated PI 6.29 ± 0.58 6.11 ± 0.29 5.89 ± 0.44
in the absence or presence of PLC for 30 min and sub- SM 23.9 ± 3.20 21.8 ± 2.24 22.5 ± 2.90

sequently incubated with [rnethyPH]choline. When PLC H9c2 cardiac myoblasts were incubated for 30 min in the absence or
was removed from the medium ofH9c2 cells radioactivity presence of 0.1 olml PLC. The cells were washed twice with fresh
incorporated into PC from [rnethyPH]choline was elevated medium, then incubated in PLC-free medium for another 30 min and the
pool-size of phospholipids determined. (A) Incubated for 30 min with
2.3-fold compared to controls (Fig. 5). Thus, aur findings 0.1 Ulml PLC. (B) Incubated for 30 min with 0.1 Ulml PLC then 30 min
in H9c2 cells agree with other studies (18,19). Taken in PLC-free medium. Results represent the mean of3 determinations. *p
together all the above results suggest that treatment ofH9c2 < 0.05.
222

the amount of mitochondrial membrane contamination of


a these preparations as estimated by cytochrome c oxidase
* activities. Since CL is synthesized exclusively in the mito-

,.:s
.c
UI chondria [5] and may playa central role in mitochondrial
8 oxidative phosphorylation [22], it is hard to imagine why CL
would be localized outside this organelle. It could be argued
0
~ that alteration of the plasma membrane and hence cellular
>< 4
phospholipid composition would not etfect the biosynthesis
E ofCL due to its absence in the plasma membrane. Previously
Co we demonstrated that PLAz-treatment ofH9c2 cells resulted
~ 2
in areduction in cellular PC levels and an inhibition of
intracellular CL biosynthesis [23]. The mechanism was
o --'----------'--- determined to be areduction in PA:CTP cytidylyltransferase
activity mediated by elevated levels of cellular Iysophospha-
tidy1choline. However, in that study we could not role out
Fig. 5. The effect ofPLC on de nova PC biosynthesis in H9c2 cells. H9c2
cells were incubated in the absence (open bars) orpresence (hatched bars) the possibility that the cellular level of DG itself could
of 0.1 Ulml ofPLC for 30 min. The cells were then incubated with 28 11M attenuate CL biosynthesis. Fatty acids are known activators
[methy 1-3H]choline chloride for 30 min and the radioactivity incorporated ofPA phosphohydrolase [24]. Treatment ofH9c2 cells with
into PC determined. Results represent the mean ± S.O. of 4 experiments. PLAz would likely produce fatty acid which could stimulate
*p< 0.05.
the production of DG by an acceleration of the reaction
catalyzed by PA phosphohydrolase. Indeed, in H9c2 cells
treated with PLAz radioactive [1 ,3- 3H]glycerol incorporated
Discussion into DG was rapidly increased [23]. The current observation
that PLC treatment of H9c2 cells increased cellular DG
The objective of this study was to determine if alterations in production from PC lend further support to the hypothesis
cellular PC levels in vivo would etfect the de novo biosyn- that CL biosynthcsis may be regulated by the cellular level
thesis of CL. We took advantage of the ability to specifically ofDG.
reduce cellular PC levels using PC-specific PLC. The major The effcct of DG and PLC treatment of cclls on PC bio-
findings are as folIows. (I) Treatment of cells with PC-specific synthesis is weil documented. Treatment of rat liver
PLC resulted in a reduction in PC content and a decrease hepatocytes with C. welchii PLC reduced cellular PC levels
in the de novo biosynthesis of PG and CL from [1,VH] 27% and elevated cellular DG levels 2.7-fold [18]. This
glycerol. (2) Treatment of cells with DiC8, a cell permeable resulted in a translocation of CTP:phosphocholine cytidylyl-
DG analog, decreased de novo biosynthesis ofPG and CL from transferase, the rate-Iimiting enzyme ofPC biosynthesis, from
[1,3- 3H]glycerol in a similar manner to that of PLC-treated cytosol to membrancs and a subscqucnt increasc in PC
cells. (3) Treatment of cells with PLC caused a reduction in biosynthesis. The mechanism for enzyme translocation was a
PA:CTP cytidylyltransferase and PGP synthase activities (4) change in the ratio ofbilayer to non-bilayer forming lipids in
Exogenous addition of DG inhibited PA:CTP cytidylyl- the membrane. A model for PC biosynthesis in PLC-treated
transferase and PGP synthase activities in vitro and (5) The chick embryonic myoblast cells had been proposed in which
data indicate that de novo mitochondrial CL biosynthesis may treatment of these cells with PLC resulted in reduction in
be regulated by DG. membrane PC content and increased binding of CTP:
The plasma membrane provides an important structural phosphocholine cytidylyltransferase to membranes with a
and functional barrier for the cello Careful analysis of the subsequent increase in PC synthesis [19]. Phorbol ester
plasma membrane phospholipid composition reveals an treatment ofHeLa cells resulted in an increase in cellular DG
absence ofCL [2]. In one reported case, Zadjela hepatoma, levels and CTP:phosphocholine cytidylyltransferase activity
CL was observed in the plasma membrane [20]. This CL was on membranes [25]. Finally, DiC8 addition to HeLa cells
estimated to ac count for approximately 1% of the total increased cellular DiC8 levels and CTP:phosphocholine
phospholipid mass. However, the authors of this study cytidylyltransferase activity on membranes [26, 27]. Our
indicated that the silica gel removed from the thin layer plates observations that either PLAz-treatment [23] or PLC-treatment
used for the estimation of CL also contained PA. More of H9c2 cells reduced cellular de novo CL biosynthesis, and
recently, CL was shown to account for 0.8-1.8% of the this was followed by elevated de novo PC biosynthesis,
plasma membrane phospholipid composition ofHT29-Glc, support the hypo thesis that CL biosynthesis may be regulated
HT29-Ino and HT29-N cells [21]. However, in these cells the by the cellular level ofPC and provide indirect evidence that
CL content in the plasma membrane correlated directly with PC and CL biosynthesis in H9c2 cellS may be coordinated. At
223

the very least the data support the notion that CL biosynthesis 12. Folch J, Lees M, Sioane-Stanley GA: A simple method for the isolation
in H9c2 cells may be regulated by plasma membrane cell and purification oftotallipides from animal tissue. J Biol Chem 226:
497- 509, 1957
signaling events involving PC.
13. Saggerson ED, Greenbaum AL: The effect of dietary and hormonal
conditions on the activities of glycolytic enzymes in rat epididymal
adipose tissue. Bioehern J 115: 405-417, 1969
Acknowledgements 14. Lowry OH, Rosebrough NJ, Faff AL, Randall, RJ: Protein measurement
with the Folin phenol reagent. J Biol Chem 193: 265--275, 1951
15. Bradford MM: A rapid and sensitive method far the quantitation of
This work was supported by a grant from the Heart and Stroke microgram quantities of protein utilizing the principle of dye binding.
Foundation of Manitoba. G.M.H. is a Heart and Stroke Anal Biochem 72: 248-254, 1976
16. Rouser G, Siakotos AN, Fleischer S: Quantitative analysis of
Foundation ofCanada Scholar.
phospholipids by thinlayer chromatography and phosphorus analysis
of spots. Lipids 1: 85-86, 1966
17. Kimes BW, Brandt BL: Properties of a eloned museie cell line from
References rat heart. Exp Cell Res 98: 367-381, 1976
18. Iamil H, Hatch GM, Vance DE: Evidence that binding of CTP:
phosphocholine cytidylyltransferase to membranes in rat hepatocytes
1. Hatch GM: Regulation of cardiolipin biosynthesis in the heart. Mol is modulated by the ratio of bilayerto non-bilayer-forming lipids.
Cell Bioehern 159: 139-148, 1996 Bioehern J 291 : 419-427, 1993
2. Hostetler KY: Polyglycerophospholipids: Phosphatidylglycerol, 19. Sleight R, Kent C: Regulation of phosphatidylcholine biosynthesis in
diphosphatidy1glycero1 and bis(monoacylglycerol)phosphate. In J.N. cultured chick embryonic musele treated with phospholipase C. J Biol
Hawthorne, G.ß. Ansell (eds). Phospholipids. Elsevier, Amsterdam, Chem 255: 10644-10650, 1980
1982, pp 215--261 20. Bergelson LD, Dyatlovitskaya EV, Torkhovskaya TI, Sarokina IB and
3. Poorthuis BJ, Yazaki PJ, Hostetler KY: An improved!wo dimensional Gorkova NP: Phospholipid composition of the tumor cell. Biochim
thin-layer chromatography system far the separation ofphosphatidyl- Biophys Acta 210: 287-298, 1970
glycerol and its derivatives. J Lipid Res 17: 433-437, 1976 21. Raynier M, Sari H, d'Anglebermes M,Ahkye E and Pasero L: Differences
4. Hatch GM: Cardiolipin biosynthesis in the isolated heart. Bioehern J in lipid characteristics of undifferentiated and enterocytic differentiated
297: 201-208, 1994 HT29 human colonic cells. Cancer Res 51: 1270-1277, 1991
5. Hostetler KY, Van Den Bosch H, Van Deenen LLM: Biosynthesis of 22. Rusnak A, Mangat R, Xu F, McClarty G and Hatch GM: Cardiolipin
cardiolipin in liver mitochondria. Biochim Biophys Acta 239: 113- remodeling in a chinese hamster lung fibroblast cellline deficient in
119,1971 oxidative energy production. J Bioenerg Biomemb 29: 291-298, 1997
6. Schlame M, Hostetler KY: Solubilization, puriffcation and characteriza- 23. Xu F, Taylor Wand Hatch GM: Lysophosphatidylcholine inhibits
tion of cardiolipin synthase from rat liver mitochondria. J Biol Chem cardiolipin biosynthesis in H9c2 cardiac myoblast cells. Arch Biochem
266: 22398-22403, 1991 Biophys 349: 341-348,1998
7. Kiyasu lY, Pieringer RA, Paulus H, Kennedy EP: The biosynthesis of 24. Brindley DN: Metabolism of triacylglycerols. In: D.E. Vance, lE.
phosphatidylglycerol. J Biol Chem 238: 2293-2298, 1963 Vancc (eds). Biochemistry of Lipids Lipoproteins and Membranes.
8. Hatch GM, McClarty G: Regulation of cardiolipin biosynthesis by Elsevier, Amsterdam, 1991, pp 171-204
cytidine-5' -triphosphate in H9c2 cardiac myoblast cells. J Biol Chem 25. UtalAK, Jami! H, Vance DE: Diacylglycerol signals the translocation
271: 25810-25816,1996 of CTP:cholinephosphate cytidylyltransferase in HeLa cells treated
9. Lin CH. Lin J. Strickland KP: Bovine brain microsomal CDP- with 12-0-tetradecanoylphorbol-13-acetate. J Biol Chem 266: 24084-
diacylglycerol synthetase: solubilization and properties. Bioehern Int 24091, 1991
25: 299-306, 1991 26. Kolesnick RN, Hemer MR: Physiologie 1,2-diacylglycerol levels
10. Mok AYP, McDougall GE, McMurray WC: Comparative studies of induce protein kinase C-independent translocation of a regulatory
CDPdiacylglycerol synthase in rat liver mitochondria and microsomes. enzyme. J Biol Chem 265: 10900-10904, 1990
Biochem. Cell Bio171: 181--189, 1993 27. Slack BE, Breu J, Wurtman RJ: Production of diacylglycerol by
11. McMurray WC, Jarvis EC: Purification and properties ofphosphatidyl- exogenous phospholipase C stimulates CTP:phosphocholine eytidylyl-
glycerolphosphate synthetase from mammalian liver mitochondria. Can transferase activity and phosphatidylcholine biosynthesis in human
J Biochem 56: 414-419,1978 neuroblastoma cells. J Biol Chem 266: 24503-24508,1991
Molecular and Cellular Biochemistry 188: 225--233, 1998.
© 1998 Kluwer Academic Publishers.

Development of pressure overload induced cardiac


hypertrophy is unaffected by long-term treatment
with losartan
Marian Turcani 1 and Heinz Rupp2
lInstitute of Pathophysiology, Medical School, Comenius University, Bratislava, Slovak Republic and 2Molecular
Cardiology Laboratory, Division of Cardiology, Philipps University ofMarburg, Marburg, Germany

Abstract
Left ventricular hypertrophy with adequate wall thickness, preserved adult phenotype and extracellular matrix may be useful
in the prevention ofheart failure. Because activation ofsubtype 1 ofangiotensin II (AT) receptors is thought to be involved in
the hypertrophie response of cardiomyocytes, we tested the potential of systemic AT l blockade to modify the development of
left ventricular hypertrophy due to pressure overload.
Sham-operated rats and rats with ascending aorta constriction were treated with losartan (30 mg/kg/day) for 8 weeks. Left
ventricular geometry, dynamics of isovolumic contractions, hydroxyproline concentration as well as myosin isozymes (marker
offetal phenotype) were assessed. Rats with aortic constriction exhibited a marked increase in left ventricular weight and the
diastolic pressure-volume relationship was shifted to smaller volumes. An enlarged ventricular pressure-volume area and
increased (p < 0.05) peak values of +dP/dtmax and -dP/dtmax demonstrated an enhanced overall ventricular performance. Signs
of congestive heart failure were not apparent. In contrast, parameters of myocardial function (normalized length-stress area,
+da/dt and -da/dt ) were depressed (p < 0.05), indicating an impaired myocardial contractility. The hydroxyproline
concent;~tion remain~ad unaltered. However, the proportion of ß-myosin heavy chains (MHC) was increased (p < 0.05).
Administration of losartan decreased (p < 0.05) blood pressure and body weight in sham operated and pressure overloaded
rats. By contrast, neither the concentric left ventricular hypertrophy or depressed myocardial function nor thc increased ß-MHC
expression were significantly altered. Thus, activation of AT 1 receptars appears not to be involved in the initial expression of
the fetal phenotype ofpressure overloaded heart which may be responsible for the progressive functional deterioration ofthe
hypertrophied ventricle. (Mol Cell Biochem 188: 225-233, 1998)

Key words: heart hypertrophy, pressure overload, angiotensin II receptor antagonist, losartan, ventricular function, myosin
isozymes

Introduction phenotype [1-3]. Attempts were, therefore, made to develop


interventions that interfere with this phenotype which is
Despite recent progress in the treatment of pressure overload exemplified by an increased expression of ß-myosin heavy
induced left ventricular hypertrophy, heart failure cannot be chains (MHC) and skeletal a-actin as weil as a failing
prevented. Although an adverse remodeling ofthe extracellular upregulation of SERCA2a expression [4-6]. During our
matrix contributes to the progression of heart failure, there screening of potential interventions far reducing this fetal
is increasing evidence that an impaired performance of the phenotype, rats with constriction ofthe ascending aorta were
cardiomyocytes represents an early trigger far the disease. A treated with losartan, an antagonist of the subtype I of the
key feature of pressure overload cardiac hypertrophy is a angiotensin Ir (AT!) receptors. Treatment with losartan is
redirection of gene expression involving the so-called fetal known to increase systemic renin and angiotensin (Ang) II

Address Jor offprints: M. Turcani, Institute of Pathophysiology, Medical School, Comenius University, Sasinkova 4, 81108 Bratislava, Slovak Republic
226

concentration [7-10] and to activate subtype 2 of the angio- from a-MHC to ß-MHC represents a weil characterized
tensin II (AT) receptors. Via a negative feedback, tissue marker of the characteristic gene expression of pressure
angiotensin converting enzyme (ACE) expression is down- overloaded rat hearts [1,2], changes in MHC expression were
regulated [11]. Because tissueACE expression during cardiac followed at the protein level. Furthennore, the hydroxyproline
hypertrophy might be linked to unfavorable changes in the concentration was examined for monitoring interstitial or
phenotype of cardiomyocytes or the associated impaired perivascular fibrosis. Cardiac function was assessed in
function [12-14], losartan may positively influence the isovolumically beating hearts [29]. This approach pennits a
development of cardiac hypertrophy. load-independent evaluation and comparison ofleft ventricular
In favor of this concept is the finding that long-term and myocardial function taking into account influences that
treatment with losartan nonnalized the level ofleft ventricular arise from ventricular mass, geometry and myocardial
ACE mRNA although it did not induce regression of left contractility.
ventricular hypertrophy in rats with persistent systolic pressure
overload [10]. Iflosartan is administered simultaneouslywith
the pressure overload, possible effects on hypertrophy Materials and methods
development remain, however, unresolved. Apart from the
uncertainty about the biological role of AT 2 receptors in Animal model
cardiac tissues, the expression of Ang II receptors maybe
influenced by pressure overload [15, 16]. Thus, in left Male 3 week-old Wistar/WU rats were purchased from
ventricular hypertrophy secondary to ascending aortic Charles-River (Kissleg, Gennany) and were housed at 21-
stenosis, expression of AT2 receptors increases whercas AT[ 23°C on 12: 12 h light-dark cycle. The rats had free access to
receptors are down regulatcd. Thcphannacological action of tap water and regular chow (Ssniff ofPlange, Soest, Gennany).
losartan may, thcrefore, depend on the course ofhypertrophy Handling of animals and all experimental procedures were in
development. This mayaiso account for the finding that accordance with the Institutional Guidelines ofthe University
induction but not maintenance ofpressure overload hyper- of Tübingen.
trophy [17-19] or isoproterenol-mediated hypertrophy [20], The ascending aorta was banded with a 3-0 silk suture
appears not to be Ang II dependent. Moreover, in vascular ligature tied against a 0.8 mm blunt steel wire at a body
smooth muscle cells trophic effects ofAng II may be mediated weight of90-100 gunder Hypnonn® (fluanisone/fentanyl-
by AT2 receptors whereas changes in vascular smooth muscle dihydrogencitrate) anaesthesia (I ml/kg, i.p.). The wire was
cells phenotype may be mainly due to AT[ receptors [21]. removed, whereby the aorta was constricted to 60-70% ofthe
We hypothesized that the AT[ blockade and its conse- original diameter. This produced a mean left ventricular-aortic
quences, i.e. stimulation of AT 2 receptors and depression of pressure gradient of 93 ± 7 mm Hg at the left ventricular
local ACE synthesis, during ventricularadaptation to pressure systolic pressure 218 ± 11 mm Hg with no significant
overload may ensure an adequate hypertrophy characterized difference between losartan treated and non-treated animals.
by a nonnalized wall tension and improved ventricular and Ascending aortic constriction was created in 44 rats, 22 rats
myocardial function due to a preserved adult phenotype. remained untreated and 22 rats were daily treated with 30 mg/
Clinical experience as weil as experimental evidence [22, 23] kg body weight losartan starting one day post surgery.
suggests that cardiac hypertrophy represents an important Age-matched control animals underwent a right thoracotomy
compensatory mechanism because it nonnalizes ventricular and ascending aorta was isolated but not constricted. Starting
wall stress. Furthennore, a functional improvement is possible one day post surgery, 12 sham-operated rats received 30 mg/
if the fetal phenotype of the hypertrophied cardiomyocytes can kg/day losartan and 12 sham-operated rats remained without
be attenuated [24, 25]. The possibility that increased systemic treatment. Losartan provided generously by Du Pont Merck
concentrations of Ang II augment the proliferation of Phannaceutical Co., Wilmington was given in the drinking
interstitial tissue and thus negatively influence ventricular water. The dose was maintained by monitoring the daily water
pump function has also to be taken into account [26, 27]. consumption and body weight. In a preliminary experiment,
To test possible effects of blockade of AT[ receptors, the the effectiveness ofAng II receptor blockade by losartan (30
associated stimulation ofAT2 receptors and inhibition ofACE mg/kg/day) was tested in 6 treated and 6 untreated animals.
expression on hypertrophy development due to persistent After two weeks oftreatment, the pressor elfect of Ang II (I
systolic pressure overload, rats with constriction of the Ilg/kg, i. v.) was significantly attenuated (pressure increase by
ascending aorta were treated with theAT [ antagonist losartan. 35 ± 7 mm Hg in untreated animals vs. 5 ± 3 mm Hg in treated
In this model, the severe pressure overload developed animals).
gradually during maturation ofthe animals. Left ventricular Systolic arterial blood pressure and heart rate were
mass and geometry were assessed after 8 weeks when a measured in conscious rats by the tail-cuff method twice one
marked hypertrophy had developed [28]. Because the switch week before the hemodynamic measurements started.
227

Measurement ofleft ventricular isovolumic contraction possibility that any inhibitory effect ofAT1-receptors blockade
on hypertrophy development may become apparent with a
Animals were studied 8 weeks after surgery. The measurements greater delay in time [14].
were performed in open-ehest rats underurethane anaesthesia
(1.2 g/kg body weight, i.p.) as previously described [29, 30].
A mediosteral thoracotomy was performed and the left Data analysis
ventricle was pierced at the apex with a steel cannula no. I
connected to a Gould-Statham P23XL pressure transducer The present approach permitted the construction of complete
(Gould Electronics, Bilthoven, Netherlands). The right ventricular pressure-volume and stress-Iength diagrams.
carotid artery was cannulated with a polyethylene tubing Ventricular and myocardial function were assessed taking into
(0.5 mm inner diameter) and forwarded to the aortic arch. The account ventricular mass, geometry and loading conditions
tubing was connected to a second Gould-Statham P23XL [29]. Systolic peak pressures measured under isovolumic
pressure transducer. Left ventricular pressure, left ventricular conditions were plotted against end-diastolic volumes
diastolic pressure (high amplification of left ventricular resulting in isovolumic pressure-volume curves. The area
pressure), time derivative ofventricular pressure (dP/dt) and between systolic and end-diastolic pressure-volume curves
aortic pressure were recorded simultaneously on a Heilige up to maximal developed systolic pressure was used as an
Recomed recording system (Heilige, Freiburg, Germany). index of ventricular working capacity. All auxotonic
For monitoring isovolumic contractions, the ascending pressure-volmne values resided within this isovolumically
aorta above the aortic valve was clamped for 6-8 sec with a determined pressure-volume area.
forceps as verified by the absence of pulsatile pressure in the Transformation of the ventricular prcssurc-volume dia-
aortic arch. Small end-diastolic volumes, i.e. low preload gram to the stress-Iength relation permitted the evaluation
values, were achieved by the short initial tightening of a string of myocardial function in the presence of altered ventricular
around the inferior vena cava. The preload increased gradually mass or geometry. Myocardial contractility was evaluated
after relieving the inferior vena cava flow and clamping the on the basis of normalized stress-Iength area, i.e. the area
aorta. This procedure was repeated 4-6 times. The analysis between the systolic and end-diastolic mean wall stress vs.
was based on the recording showing the highest systolic normalized midwall circumference (length) curves. By
pressure development. analogy with papilIary musc1e, the nonnalized midwall
Left ventricular passive pressure-volume relations were circumference was ca1culated as the ratio between a given
assessed after recording isovolumic contractions. The midwall circumference and the midwall circumference
atrial-ventricular groove was ligated with a silk string and associated with peak developed wall stress [29, 31].
the right ventricle was emptied by incision. The left ventricle Pressure-volume data were transformed into slress-Iength
was filled with a defined volume of saline and emptied in data using a thick-walled spherical shell. The ca1culations
50111 steps while recording the passive ventricularpressure. assumed that the specific density of myocardium was I g/cm3,
Three reproducible pressure-volume curves were generated so the ventricular weight in grams equals the ventricular
within 3-4 min after the ligation and within this time no effects wall volume in cm3 •
of anoxia on the pressure-volume relation could be detected. Mean (systolic or diastolic) wall stress (a) was derived from
Using the passive pressure-volume relation, end-diastolic the following formula [32]: a= PI {[(V+W)IVF/3~1}, where P
cavity volumes required for the further analysis were derived is intraventricular pressure, V is ventricular cavity volume, W
from the measured end-diastolic pressures. is ventricular wall volume. Since the contraction was iso-
volumic, end-diastolic volume derived from the passive
pressure-volume curve was identical to the ventricular cavity
Heart weights volmne (V) during the respective beat. Midwall circumference
(C ) was ca1culatedaccording: CR=n{(3/4n)1/3[(VI/3+(V+W)I!
The heart was excised after hemodynamic evaluation. The atria 3]) ~ To evaluate the velocity of contraction and relaxation at
and major vessels were first trimmed and right ventricular free the myocardiallevel, rate of mean wall stress rise (+da/dt)
wall was dissected along the septum. The left ventricle plus or decline (--da/dt) was ca1culated using the recorded ± dPI
septum and right ventricle free wall were bIotted and weighed. dt values: ± da/dt = ± (dP/dt)1 {[(V+W)1V2i3 ~l} [32].
After we had determined that 8 weeks treatment with losartan
did not attenuate the left ventricular hypertrophy, we continued
losartan treatment in 6 rats with aortic constriction for Myosin isozymes
additional 4 weeks. Six non-treated rats with aortic constriction
served as controls. In these rats only the body and heart weights For determination ofmyosin isozymes ofthe same cardiac
were assessed. We started this experiment to exclude the region by non-dissociating polyacrylamide gel electro-
228

phoresis in the presence of pyrophosphate [33], a portion ship was shifted to smaller volumes. The enlarged ventricular
ofthe left ventricular free wall (about 100 mg) was cut from pressure-volume area (Fig lA, Table 1) and increased peak
the apex to the base and stored in liquid nitrogen. The values of +dP/dtrnax and -dP/dtrnox (Fig 2A, 2B) documented
myosin isozymes were stained with Coomassie brilliant blue augmented overall ventricular performance. The increase
R250 and the gels were scanned using a Quick Scan densito- in ventricular pressure-volume area was accompanied by
meter (Helena Laboratories, Beaumont, TX, USA). The concentric hypertrophy (decreased chamber volume/wall
isozymes VI, V2, V3 were quantitated by measuring peak volume ratio) maintaining the developed wall stress in
heights. normal ranges. The operational range of end-diastolic
Collagen content of ventricular tissue was assessed by volumes was shifted to smaller volumes but the hyper-
determining the hydroxyproline concentration. A 70--100 mg trophied ventricles developed, at comparable end-diastolic
portion of the left ventricular free wall was cut from the apex volumes, higher intraventricular pressures compared with
to the base, freeze-dried and processed essentially as given control ventriclcs (Fig IA). In contrast, calculated parameters
by Stegemann [34]. of myocardial fimction (normalized length-stress area, +dal
Statistical analysis was performed with Statistica/w dtrnax and-da/dtma) were depressed (Table 1, Fig IB, Fig 2C,
(Statsoft, Tulsa, OK, USA). Normality of distribution was 2D) indicating a decreased myocardial contractility.
checked by the Kolmogoroff-Smimoff test and equality of The losartan treatment did not prevent the depression of
variances according to Cochran. The statistical analysis of the functional parameters (Table 1, Fig 1 B, Fig 2C, 2D).
differences observed between losartan treated and untreated The unchanged hydroxyproline concentration (Table 2)
sham operated rats and rats with aortic stenosis were made suggests that alterations in cardiomyocytes were responsible
by one-way analysis of variance and Tukey HSD post-hoc for the depressed contractility rather than interstitial or peri-
test for unequal sampIe sizes (Spjotvoll & Stoline). Statistical vascular fibrosis. At the molecular level, the characteristic
significance was assumed at p < 0.05. shift in favor of an enhanced expression of ß-MHC was
observed (Table 2). Apparently, the increased left ventri-
cular mass can compensate for the depressed myocardial
Results performance associated with unfavorable molecular changes
in the cardiomyocytes. Also a prolonged administration of
Treatment with theAT 1 receptor antagonist losartan lowered losartan up to 12 weeks was without any effect on left
(p< 0.05) systolic blood pressure in rats with constriction of ventricular weight or left ventricular weight-to-body weight
the ascending aorta and in sham-operated rats (Table I). ratio. Thus, average left ventricular weight was 1168 ± 91
Heart rate was not affected and body weight was slightly mg in rats with ascending aortic stenosis without losartan
decreased (Table 1). After 8 weeks of pressure overload, the treatment and 1117 ± 95 mg with losartan treatment. Values
left ventricular weight-to-body weight ratio was signifi- for the left ventricular weight-to-body weight ratio were
cantly increased and the diastolic pressure-volume relation- 3.04 ± 0.21 mg/g and 3.03 ± 0.3 mg/g for untreated and

Table 1. Body weight, right and left ventricular weights, left ventricular passive diastolic dimensions, heart rate, systolic arterial pressure, left ventricular
pressure-volume area and normalized stress-Iength area.

Sham Aortic constriction


Vntreated +Losart.n Vntreated +Losartan
(n ~ 12) (0 ~ 12) (0 ~ 16) (0 ~ 16)

Body weight (g) 324 ±14 303 ± 10' 318 ± 16 301 ±21 *'
Right ventricular weight (mg) 201 ±22 196 ±II 211 ± 28 197 ±26
Right ventricul.r weightlbody weight (mg/g) 0.61 ± 0.06 0.64 ± 0.03 0.66 ± 0.1 0.66 ± 0.09
Left veotricular weight (mg) 690 ±40 640 ±48 1056 ± 102' 989 ±84*
Left ventricular weightlbody weight (mglg) 2.\3 ± 0.07 2.08 ± 0.14 3.32 ± 0.23' 3.29± 0.25'
Left ventricular cavity volume (mm3) 441 ±65 437 ±34 325 ± 29' 320 ±46*
Heart rate (beatsiminute ) 383 ±19 394 ±27 375 ± 20 378 ±12
Systolic arterial pressure (mm Hg) 134 ± 5 113 ± 7' 131 ± 12 \06 ± 8"
Left ventricul.r cavity voIumeiwall volume 0.64 ± 0.07 0.68 ± 0.06 0.31 ± 0.04' 0.33± 0.05'
Left ventricular pressure-volume area (mJ) 5.56 ± 0.96 5.50 ± 0.80 9.54± 1.45' 9.91 ± 1.33'
Left ventricular normalized stress-Iength area (kPa) 5.5 ± 0.44 5.28 ± 0.42 3.91 ± 0.48' 4.13 ± 0.58*

Left ventricular cavity volume, and left ventricular cavity volumel wall volume are giveo for the common intraventricular pressure 6 mm Hg. Heart rate and
systolic arterial pressure were measured usiog a tail-cuffmethod in unanesthetized rats. Values are means ± S.D. Pa - pascal; J - joul. 'p< 0.05 compared
with sham-oper.ted untre.ted rats; 'p < 0.05 compared with untreated rats with aortic coostriction.
229

A B
400 40

300 30
...... tu
~
C)
l:
E '"
§. 200
i!! ~ 20
::J

~
1c
D.. nI
Q)
100 ::::i: 10

0 I) ..... -.-~~---.! 0

100 200 300 400 0,7 0,8 0,9 1,0


Volume (mm') Normalized miclwall circumference

Fig. 1. Graphs showing effeets of aortie eonstrietion and losanan treatment on left ventrieular pressure-volume and stress-length relations. Panel A:
Isovolumie end-systolie and passive (diastolie) left ventrieular pressure-volume relations. Panel B: Pressure-volume relations transformed into normalized
stress-length relations. Values are me ans ± S.D. Diastolie volumes and passive diastolic properties at the intravenlrieular pressure of 6 mmHg and left
venlrieular pressure-volume and stress-lenglh areas for all experimental groups are shown in Table I. (0) Vntreated sham-operated conlrol rals, (e) losartan
Ireated sham-operated rals, (0) unlreated rals wilh aortie conslrietion, (.) losartan Ireated rats with aortie eonslrietion, (-) end-syslolie pressure-volume
relations, (- - -) passive (diaslolie) pressure-volume relalions.·p < 0.05 eompared wilh sham-operated unlreated rats.

treated rats with pressure overIoaded left ventricles, dilatation were responsible for the hypertrophy regression
respectively. due to losartan [14] as suggested by Weinberg et al. [10],
the inhibitory effect ofACE inhibition on cardiac growth had
to be attributed to load reduction or increased concentrations
Discussion of bradykinin and other kin ins rather than to a decreased
stimulation of AT, receptors.
AIthough there is increasing evidence that hypertensive Results ofthe present study suggest that not only mainte-
heart disease can be improved by ACE inhibition and AT) nance but also development of pressure overload hyper-
receptor blockade, the relative influences arising from trophy is load dependent and does not represent an AT,
blockade of cardiac Ang II receptors or ventricular un- receptor mediated process. In the case of ascending aortic
loading remains unclear. Thus, in rats with ascending aortic constriction with a severe fixed increase in afterload, the
stenosis with a persistent systolic overload chronic inhibition mechanical stimulation appears to dominate at least at later
of the ACE resuIted in regression of the developed left stages of the hypertrophy. Immediately after the load
ventricular hypertrophy [14, 35]. By contrast, a long-term induction, the role of AT) receptors might, however, be
AT) receptor blockade did not regress left ventricular different. Thus, losartan significantly attenuated the left
hypertrophy in rats with ascending aortic stenosis and did ventricular hypertrophy due to transverse aortic constriction
thus not confirm the contention that activation of AT) in mice during the first 7 days of pressure overload [36].
receptors is critical for the maintenance of cardiac hyper- Furthermore, AT, receptor blockade was similarly effective
trophy independently of thc hemodynamic load [10]. in reducing left ventricular hypertrophy after 3 days in rats
However, Bruckschlegel et al. [14] observed in the same with abdominal aortic coarctation [37] and after 2 weeks in
model of left ventricular pressure overload a significant rats with aortic insufficiency [38]. Blockade ofAT 2 receptors
regression of left ventricular hypertrophy after losartan had no effect. These short-term studies were in accordance
treatment. If the slightly higher degree of peripheral vaso- with in vitro effects ofAng 11 on cardiac myocytes [39]. An
230

18 A .-. B
:? 15
.!!! 8
*
CI
J:
CI
J: *
E 12 E 6
E E
'"....
0
9 "'0
~ :s. 4
1.1 1.1
..ß 6
~E
32
a. 3 a. 2
"C
+ "li'
0,0 0,2 0,4 0,6 0,8 1,0 0,0 0,2 0,4 0,6 0,8 1,0
Enddiastolic mean wall stress (kPa) Enddiastolic mean wall stress (kPa)

1800 700

1500 600
~
a.
~ 500
1200 ~
~
'"i 400
't!
1.1 900
..ß 300
32 "'6
b 600
"C
+ "li' 200
300 100

0,0 0,2 0,4 0,6 0,8 1,0 0,0 0,2 0,4 0,6 0,8 1,0

Enddiastolic mean wall stress (kPa) Enddiastolic mean wall stress (kPa)

Fig. 2. Graphs showing effects of aortic constriction and losartan treatment on preload dependence of maximal rate of intraventricular pressure rise (A),
maximal rate of intraventricular pressure decline (B), maximal rate of waU stress rise (C) and maximal rate of waU stress decline (D). The overaU frequency
ofisovolumic contractions was 329 ± 22 (beats/min) with no significant differences between experimental groups. Values are means ± S.D. CO) Untreated
sham-operated control rats, ce) losartan treated sham-operated rats, (0) untreated rats with aortic constriction, C.) losartan treated rats with aortic constriction.
*p < 0.05 compared with sham-operated untreated rats.

increased transcription of proto-oncogenes, transfonning with the hemodynamic overload or after the hypertrophy has
growth factor ß, and fetal genes represent a common developed, led to the concept that Ang II affects differen-
immediate response to stimulation by either Ang II or tially the induction and maintenance of cardiac hypertrophy.
mechanicalload. The restriction of these effects to an early Ang 11 may participate in maintenance but not development
phase ofthe hypertrophy development is suggested not only of isoproterenol induced cardiac hypertrophy [20] and
by the present study. Ito et al. [40] demonstrated that pressure overload hypertrophy [14, 18,35]. On the other
endothelin, which may mediate some trophic effects ofAng hand, a major role has been suggested for Ang II in the
II, affects cardiac hypertrophy also only during the early remodeling of a volume overloaded heart, but Ang TI may
phase (up to I week) ofpressure overload. also not directly contribute to the maintenance of volume
The marked differences in action of ACE inhibitors on overload induced cardiac hypertrophy [41]. The present data
cardiac hypertrophy when applied either simultaneously are in accordance with the concept of an Ang II independent

Table 2. Myosin isozyme population and hydroxyproline concentration.

Sbam Aortic constriction


Untreated +Losartan Untreated +Losartan
(n = 12) Cn = 12) (n= 16) Cn= 16)

Myosin VI 62 ±5 66 ±6 48 ± 7' 46 ±6*


Myosin V2 23 ±3 22 ±2 27 ±2 27 ±3
Myosin V3 15 ±3 12 ±3 25 ±5* 27 ±5*
Hydroxyproline C).lg/mg dry weigbt) 2.75 ± 0.20 2.78 ± 0.29 2.78 ± 0.27 2.59 ± 0.29

Values are means ± S.D. *p < 0.05 compared with sham-operated untreated rats.
231

development of hypertrophy under conditions of severe Conclusion


persistent pressure overload. Moreover, the unchanged left
ventricular weight after 12 weeks oflosartan administration
argues against a role of Ang 11 in maintaining this type of The present study demonstrates that a long-term systemic
hypertrophy. AT I-receptor blockade with losartan and associated changes
Besides a time-dependent role ofAT I receptors in induction in renin and Ang 11 concentration does not interfere with
of cardiac hypertrophy, a partial AT I receptor blockade may the pattern of adaptation of the left ventricle to a persistent
account for the lack of effect on pressure overload left systolic pressure overload due to ascending aortic constric-
ventricular hypertrophy. However, a long-term application tion. The data support the contention that development and
of 30 mg/kg body weight per day losartan caused a high maintenance of pressure overload hypertrophy does not
level of AT I receptor blockade as confirmed by an attenua- critically depend on trophic effects of Ang 11. The beneficial
tion of the pressor response to Ang Ir infusion and a effects ofAT) receptor blockers in the prevention and treatment
sustained moderate decrease in arterial systolic pressure [10, ofventrlcular dysfunction may, therefore, be attributed to the
and this study]. Modest but statistically significant reduction diminution ofventricular load and to the interference with the
in body weight was observed in losartan treated rats also. renin-angiotensin-aldosterone-system which becomes strongly
The lower body weight was not due to decreased food activated when heart failure progresses.
consumption. Chronic ACE inhibition in comparable young
rats similarly retarded body growth [35]. Interference with
the postulated involvement of Ang 11 in the growth regula- Acknowledgements
tion of neonatal and young animals [42] may be responsible
for this effect. The study was supported by the Deutsche Forschungs-
A systemicAT I receptor blockade increased plasma renin, gemeinschaft (Ja 172/14-1, Ru 245/7-1), the Alfred-Teufel-
and Ang 11 levels as weil as Ang 11 concentration in heart and Stiftung and the Siovak Grant Agency No. 1/4114/97 and 1/
other organs except kidney [7, 8]. Increased Ang 11 would 4133/97. The experimental part ofthis study was performed
be expected to stimuIate AT 2 receptors and together with at the Institute of Physiology TI, University of Tübingen,
increased renin to attenuate ACE gene expression by a Germany. The support of Prof. R. Jacob is greatly acknow-
negative feedback [11]. Despite increasing evidence for an ledged.
important role of AT 2 receptors in cellular growth and
differentiation processes [21] the present results do not
suggest a role of AT 2 receptors or the postulated down
regulation of the ACE gene either in the development of References
cardiac hypertrophy or in fetal MHC expression. The
unaltered MHC expression would be consistent with the I. Schwartz K, Bouvcrct P, Bercovici J, Swynghedauw B: An immuno-
chemical difference between myosins from normal and hypertrophied
observation that the switch in myosin isoform mRNA from
ral hcarts. FEBS Lett 93: 137-139, 1978
adult to the fetal pattern in Ang II induced hypertrophy 2. Izumo S, Lompre AM, Matsuoka R, Koren G, Schwartz K: Myosin
depends on pressure overload and not on activation of heavy chain messenger RNA and pro tein isoform transitions during
ATI-receptors [43]. Thus, the action ofetomoxir, an inhibi- cardiac hypertrophy: Interaction between hemodynamic and thyroid
tor of mitochondrial camitine palmitoyltransferase-l which hormone-induced signals. J Clin Invest 79: 970-977,1987
3. Chien KR, Knowlton KU, Zhu H, Chien S: Regulation of cardiac gene
increased the a-MHC expression, calcium transport activity
expression during myocardial growth and hyperttophy: Molecularstudies
of sarcoplasmic reticulum [44-46] and left ventricular ofan adaptive physiologie response. FASEB J 5: 3037-3046,1991
performance [25] cannot be associated by signals affected by 4. Schwartz K, Carrier L, Mercadier 11, Lompre AM, Boheler KR:
Ang 11. Molecular phenotype of the hyperttophied and failing myocardium.
Ang 11 induces an acute fibrotic response within the heart Circulation 87 (Supp!. VII): VII-5-VII-l0, 1993
5. Lompre AM, Anger M, Levitsky D: Sarco(endo)plasmic reticulum
[47] and vessels [21]. The absence of increased hydroxy-
calcium pumps in the cardiovascular system: Function and gene
proline concentration after losartan treatment which is expression. J Mol Cell Cardiol26: 1109-1121, 1994
associated with increased Ang 11 levels suggests that the Ang 6. FeldmanAM, Weinberg EO, Ray PE, Lorell BH: Selective changes in
11 stimulation of nonmyocyte cardiac cells is mediated by AT I cardiac gene expression during compensated hypertrophy and the
and notAT2 receptors. Although excellent correlations have transition to eardiac deeompensation in rats with ehronic aortic
banding. Cire Res 73: 184-192, 1993
been reported for histochemical techniques and hydroxy-
7. Goldberg MR, Tanaka W, Barchowsky A, Bradstreet TE, MeCrea J,
proline assay, a slight localized increase in interstitial or peri- Lo MW, MeWilliams EJ, Bjomsson TD: Effeets oflosartan on blood
vascular collagen may not have been detected by the present pressure, plasma renin activity, and angiotensin II in volunteers.
biochcmical mcthod [48]. Hypertension 21: 704-713,1993
232

8. Campbell DJ, KladisA, ValentijnAJ: Effeets oflosartan on angiotensin 25. Turcani M, Rupp H: Etomoxir improves left ventricular performance
and bradykinin peptides and angiotensin-eonverting enzyme. J ofpressure-overloaded rat heart. Circulation 96: 3691-3686, 1997
Cardiovase Pharm 26: 233-240, 1995 26. Weber KT, Brilla CG: Pathological hypertrophy and eardiac interstitium.
9. Levy BI, Benessian J, Henrion D, Caputo L, Heymes C, Diurez M, Fibrosis and renin-angiotensin-aldosteron system. Circulation 83: 1849-
Poitevin P, Samuel JL: Chronic blockade of AT2-subtype receptors 1865,1991
prevents the effect of angiotensin II on the rat vaseular strueture. J 27. Brilla CG, Rupp H, Funck R, Maisch B: The renin-angiotensin-
Clin Invest 98: 418-425, 1996 aldosterone system and myocardial collagen matrix remodeling in
10. Weinberg EO, Lee NU, Weigner M, Lindpaintner K, Bishop SP, congestive heart failure. Eur Heart J 16 (Supp!. 0): 107-109, 1995
Benediet CR, Ho KKL, Douglas PS, Chafizadeh E, Lorell BH: 28. Litwin SE, Katz SE, Weinberg EO, Lorell BH, Aurigemma GP,
Angiotensin AT, reeeptor inhibition. Effeets on hypertrophie Douglas PS: Serial echocardiographic-Doppler assessment of left
remodelling and ACE expression in rats with pressure-overload ventricular geometry and function in rats with pressure-overload
hypertrophy due to aseending aortie stenosis. Cireulation 95: 1592- hypertrophy. Chronic angiotensin-converting enzyme inhibition
1600,1997 attenuates the transition to heart failure. Circulation 91: 2642-2654,
11. Schieffer B, Wirger A, Meybrunn M, Seitz S, Holtz J, Drexler H: 1995
Comparative effects of ehronie angiotensin-eonverting enzyme 29. Jacob R, Brändle M, Dierberger B, Ross Ch: Conceptional and
inhibition and angiotensin II type I reeeptor blockade on eardiae methodological basis for the evaluation ofventricular and myocardial
remodelling after myoeardial infaretion in rats. Circulation 89: 2273- mechanies. In: R. Jaeob (ed.). Evaluation of Cardiac Contraetility.
2282,1994 Gustav Fischer, StuttgartlNew York, NY, 1987, pp 75-100
12. Schunkert H, Jackson B, Tang SS, Schoen FJ, Smits JFM, Apstein 30. Hepp A, Hansis M, Gülch R, Jacob R: Left ventricular isovolu-
CS, Lorell BH: Distribution and functional significance of eardiae metrie pressure-volume relations, 'diastolic tone', and contractility
ACE in hypertrophied rat hearts. Circulation 87: 1328-1339, 1993 in the heart after physieal training. Basic Res Cardiol 69: 517-
13. Hirsch AT, Talsness CE, Sehunkert H, Paul M, Dzau VJ: Tissue- 531,1974
speeific aetivation of cardiae angiotensin eonverting enzyme in 31. Jacob R, Dieberger B, Gülch RW, Noma K: Adaptive alterations in
experimental heart failure. Cire Res 69: 475-482, 1991 cardiac mass and configuration: beneficial or detrimental? A quantitative
14. Brueksehlegel G, Holmer SR, Jandeleit K, Grimm D, Muders F, analysis based on the finite-element method.ln: M. Nagano, N. Takeda,
Kromer EP, Riegger GAJ, Sehunkert H: Blockade of the renin- and N.S. Dhalla (eds). Tbe Adapted Heart. Raven Press, New York,
angiotensin system in eardiae pressure-overload hypertrophy in rats. NY, 1994, pp 111-123
Hypertension 25: 250--259, 1995 32. Sandler H, Dodge HT: Left ventricular tension and stress in man. Cire
15. Suzuki J, Matsubara H, Urakami M, lnada M. Rat angiotensin II (type Res 13: 91-104, 1963
Al) reeeptor mRNA regulation and subtype expression in myocardial 33. Rupp H, Wahl R, Hansen M: Influence of diet and earnitine palmitoyl-
growth and hypertrophy. Cire Res 73: 439-447, 1993 transferase I inhibition on myosin and sarcoplasmic reticulum. J Appl
16. Kabour A, Henegar JR, Janieki JS: Angiotensin II (AII)-indueed Physiol 72: 352-360, 1992
myoeyte necrosis: Role of the All reeeptor. J Cardiovase Pharm 23: 34. Stegemann, H: Mikrobestimmung von Hydroxyprolin mit Chlorarnin
547-553, 1994 T und p-Dimethylarninobenzaldehyd. Hoppe-Seylers Z Physiol Chem
17. Kromer EP, Riegger GAl: Effeets oflong-term angiotensin converting 311: 41-45, 1958
enzyme inhibition on myoeardial hypertrophy in experimental aortic 35. Weinberg EO, Sehoen FJ, George D, Kagaya Y, Douglas PS, Litwin
stenosis in the rat. Am J Cardiol62: 161-163, 1988 SE, Schunkert H, Benedict CR, Lorell BH: Angiotensin-eonverting
18. Kromer EP, Elsner D, Riegger GAJ: Role of neurohumoral systems for enzyme inhibition prolongs survival andmodifies the transition 10 heart
pressure induced left ventricular hypertrophy in experimental supra- failure in rats with pressure overload hypertrophy due to aseending
valvular aortic stenosis in rats. Am J Hypertens 4: 521-524, 1991 aortie stenosis. Circulation 90: 1410--1422, 1994
19. Zierhut W, Zimmer HG, GerdesAM: Effect of angiotensin converting 36. Roekroan HA, Wachhorst SP, Mao L, Ross J Jr: ANG II reeeptor
enzyme inhibition on pressure-induced left ventrieular hypertrophy blockade prevents ventricular hypertrophy and ANF gene expression
in rats. Cire Res 69: 609-617, 1991 with pressure overload in mice. Am J Physiol 266: H2468-H247S,
20. Golomb E,Abassi ZA, Cuda G, Stylianou M, Panchal VR, Trachevsky 1994
D, Keiser HR: Angiotensin II maintains, but does not mediate 37. Everett AD, Tufro-McReddie A, Fisher A, Gomez RA: Angiotensin
isoproterenol-induced cardiac hypertrophy in rats. Am J Physiol267: reeeptor regulates eardiae hypertrophy and transforming growth
HI496-HI506, 1994 factor-ß, expression. Hypertension 23: 587-592, 1994
21. Sabri A, Levy BI, Poitevin P, Caputo L, Faggin E, Marotte F, 38. Ishiye M, Umemura K, Uematsu T, Nakashima M: Angiotensin AT,
Rappaport L, Samuel JJ: Differential roles of AT, andAT 2 reeeptor reeeptor-mediated attenuation of eardiae hypertrophy due to volume
subtypes in vaseular trophic and phenotypic ehanges in response to overload: involvement of endothelin. Eur J Phannacol. 280: 11-17,
stimulation with angiotensin Ir. Arterioseler Thromb Vase Biol 17: 1995
257-264, 1997 39. Sadoshima J, Izumo S: Molecular eharacterization of angiotensin
22. Grossman W: Cardiac hypertrophy: useful adaptation or pathologie II-induced hypertrophy of cardiae myocytes an hyperplasia of cardiac
process? Am J Med 69: 576-584, 1980 fibroblasts: Critieal role of the AT, receptor subtype. Cire Res 73:
23. Koide M, Nagatsu M, Zile M, Harnawaki M, Swindle NIM, Keeeh G, 413-423, 1993
DeFreyte G, Tagawa H, Cooper IV G, Carabello BA: Premorbid 40. Ito H, Hirata Y, Adachi S, Tanaka M, Tsujino M, KoikeA, NogamiA,
determinants of left ventrieular dysfunetion in a novel model of Marumo F, Hiroe M: Endothelin-l is an autocrine/paracrine factor in
gradually induced pressure overload in the adult canine. Circulation the mechanism of angiotensin II-induced hypertrophy in cultured rat
95: 1601-1610,1997 eardiomyocytes. J Clin lnvest 92: 398-403, 1993
24. Penneoek GD, Raya TE, Bahl JJ, Goldman S, Morkin E: Combination 41. Ruzicka M, Leenen FHH: Renin-angiotensin system and volume
treatment with captopril and the thyroid hormone analogue 3,5-diiodo- overload-induced cardiac remodeling. In: N.S. Dhalla, R.E. Beamish,
thyropropionie acid (DITPA): A new approach to improving left ven- N. Takeda and M. Nagano (eds). The Failing Heart. Lippincott-Raven
trieular performance in heart failure. Circulation 88: 1289-1298, 1993 Publishers, Philadelphia, 1995, pp 305-325
233

42. Lee KR, MacFadyen RJ, Doing JK, Reid JL: Role of angiotensin in 267: H209 I-H2099, 1994
the extravascular system. J Hum Hypertens 7 (Supp!. 2): S7-S 12, 1993 46. Zarain-Herzberg A, Rupp H, Elimban V, Dhalia NS: Modification of
43. Sucic 0, Nunez E, Frohlich Ed, Prakash 0: Angiotensin II increases sarcoplasmic reticulum gene expression in pressure overloaded cardiac
left ventricular mass without affecting myosin isoform mRNAs. hypertrophy by etomoxir. FASEB J 10: 1303-1309, 1996
Hypertension 28: 265--268, 1996 47. Crawford DC, Chobanian AV, Brecher P: Angiotensin II induces
44. Rupp H, Wahl R, Hansen M: Influence of diet and carnitine palmitoyl- fibonectin expression associated with cardiac fibrosis in the rat. Circ
transferase I inhibition on myosin and sarcoplasmic reticulum. J Appl Res 74: 727 739, 1994
Physiol 72: 352-360, 1992 48. Pickering JG, Bougimer DR: Fibrosis in the transplanted heart and its
45. Vetter R, Rupp H: CPT- I inhibition by etomoxir has a chamber-re1atcd relation to donar ischemic time. Assessment with polarized light
action on cardiac sarcoplasmic reticulurn and isomyosins. Am J Physiol microscopy and digital image analysis. Circulation 81: 949-958, 1990
Molecular and Cellular Biochemistry 188: 235-237, 1998.

Index to Volume 188


Aviram M, Fuhnnan B: LDL oxidation by arterial wall macrophages depends on the oxidative status in the
lipoprotein and in the cells: Role of prooxidants vs. antioxidants 149-159
Avivi-Green C, see Schwartz B, Polak-Charcon S

Bernlohr DA, see Vogel Hertzel A


Bkaily G, Jaalouk D, Sader S, Shbaklo H, Pothier P, Jacques D, D'Orleans-Juste P, Cragoe Jr EJ, Bose R:
Taurine indirectly increases [Ca]i by inducing Ca2+ influx through the Na+-Ca2+ exchanger 187-197
Bose R, see Bkaily G et al.
Bowron E, see McCargar LJ et al.

Cattini PA, see Detillieux KA et al.


Clarke R, see Hilakivi-Clarke L
Cragoe Jr EJ, see Bkaily G et al.

D'Orleans-Juste P, see Bkaily G et al.


Dakshinamurti K, LaI KJ, Ganguly PK: Hypertension, calcium channel and pyridoxine (vitamin B6) 137-148
Dakshinamurti K, see Golfman L et al.
Dawson K, see McCargar LJ et al.
Detillieux KA, Meyers AFA, Meij JTA, Cattini PA: An AlG-rich motif in the rat fibroblast growth factor-2
gene confers enhancer activity on a heterologous promoter in neonatal rat cardiac myocytes 169-176
Dhalla NS, see Golfman L et al. and see Rupp H, Elimban V
Dixon IMC, see Golfman L et al.
Donepudi M, see Nolan E et al.
Du X, see Rösen P, Tschöpe D

Elimban V, see Rupp H, Dhalla NS

Falchuk KR: The molecular basis for the role of zine in developmental biology 41-48
Flanagan L, see Nolan E et al.
Fuhnnan B, see Aviram M

Ganguly PK, see Dakshinamurti K, LaI KJ


Golfman L, Dixon IMC, Takeda N, Lukas A, Dakshinamurti K, Dhalla NS: Cardiac sareolemmal Na+-Ca2+
exchange and Na+-K+ ATPase activities and gene expression in alloxan-indueed diabetes in rats 91-101
Graham SE, see Russell JC, Richardson M
Grynpas MD, see Poucheret P et al.

Harris ED, Qian Y, Reddy MCM: Genes regulating copper metabolism 57-fJ2
Hatch GM, see Xu FY et al.
Hayashi M, see Takeo S et al.
Hilakivi-Clarke L, Clarke R: Timing of dietary fat exposure and mammary turnorigenesis: Role of estrogen
receptor and protein kinase C activity 5-12

Ida K, see Takeo S et al.


Iliskovie N, Li T, Khaper N, Palaee V, Singal PK: Modulation of adriamycin-indueed changes in serum free
fatty acids, albumin and cardiae oxidative stress 161-166
Innis SM, see McCargar LJ et al.
236

Jaalouk D, see Bkaily G et al.


Jacques D, see Bkaily G et al.

Kantor PF, see Makinde AO, Lopaschuk GD


Kelly SL, see Xu FY et al.
Khaper N, see Iliskovic N et al.
Kott M, see Vetter R et al.

Lai KJ, see Dakshinamurti K, Ganguly PK


Leichter J, see McCargar LJ et al.
Li T, see Iliskovic N et al.
Lopaschuk GD, see Makinde AO, Kantor PF
Lukas A, see Golfman L et al.

MakindeAO, Kantor PF, Lopaschuk GD: Maturation offatty acid and carbohydrate metabolism in the newborn 49-56
heart
McCargar LJ, Innis SM, Bowron E, Leichter J, Dawson K, Toth E, Wall K: Effect of enteral nutritional products
differing in carbohydrate and fat on indices of carbohydrate and lipid metabolism in patients with NIDDM 81--89
McNeill JH, see Poucheret P et al.
Meij JTA, see Detillieux KA et al.
Meyers AFA, see Detillieux KA et al.

Nasa Y, see Takeo S et al.


Nojiri M, see Takeo S et al.
Nolan E, Donepudi M, VanWeelden K, Flanagan L, Welsh J: Dissociation ofvitamin D 3 and anti-estrogen
mediated growth regulation in MCF-7 breast cancer cells 13-20

Palace V, see Iliskovic N et al.


Polak-Charcon S, see Schwartz B, Avivi-Green C
Pothier P, see Bkaily G et al.
Poucheret P, Verma S, Grynpas MD, McNeill JH: Vanadium and diabetes 73--80
Prasad AS: Zinc and immunity 63-69

Qian Y, see Harris ED et al.

Reddy MCM, see Harris ED et al.


Richardson M, see Russell JC, Graham SE
Rösen P, Du X, Tschöpe D: Role of oxygen derived radicals for vascular dysfunction in the diabetic heart:
Prevention by a-tocopherol? 103-111
Rupp H, Elimban V, Dhalla NS: Differential influence of fasting and BM13.907 treatment on growth and
phenotype of pressure overloaded rat heart 209-215
Rupp H, see Turcani M and see Vetter R et al.
Russell JC, Graham SE, Richardson M: Cardiovascular disease in the JCR:LA-cp rat 113-126

Sader S, see Bkaily G et al.


Sasaki H, see Takeo S et al.
Schulze W, see Vetter R et al.
Schwartz B, Avivi-Green C, Polak-Charcon S: Sodium butyrate induces retinoblastoma protein dephos-
phorylation, p 16 expression and growth arrest of colon cancer cells 21-30
Shbaklo H, see Bkaily G et al.
Singal PK, see Iliskovic N et al.
237

Takeda N, see Golfman L et al.


Takeo S, Nasa Y, Tanonaka K, Yabe K-i, Nojiri M, Hayashi M, Sasaki H, Ida K, Yanai K: Effects oflong-term
treatment with eicosapentaenoic acid on the heart subjected to ischemialreperfusion and hypoxia/
reoxygenation in rats 199-208
Tanonaka K, see Takeo S et al.
Taylor WA, see Xu FY et al.
Toth E, see McCargar LJ et al.
Tschöpe D, see Rösen P, Du X
Turcani M, Rupp H: Development ofpressure overload induced cardiac hypertrophy is unaffected by long-term
treatment with losartan 225-233

VanWeelden K, see Nolan E et al.


Verma S, see Poucheret P et al.
Vetter R, Kott M, Schulze W, Rupp H: Influence of different culture conditions on sarcoplasmic reticular calcium
transport in isolated neonatal rat cardiomyocytes 177-185
Vogel Hertzel A, Bemlohr DA: Regulation of adipocyte gene expression by polyunsaturated fatty acids 33-39

Wall K, see McCargar LJ et al.


Welsh J, see Nolan E et al.

Xu FY, Kelly SL, Taylor WA, Hatch GM: On the mechanism of the phospholipase C-mediated attenuation of
cardiolipin biosynthesis in H9c2 cardiac myoblast cells 217-223

Yabe K-i, see Takeo S et al.


Yanai K, see Takeo S et al.

Zemel MB: Nutritional and endocrine modulation of intracellular calcium: Implications in obesity, insulin
resistance and hypertension 129-136
Developments in Molecular and Cellular Biochemistry

Series Editor: Naranjan S. Dhalla, Ph.D., M.D. (Hon.), FACC

1. VA Najjar (ed.): Biological Effects ofGlutamicAcid andlts Derivatives. 1981 ISBN 90-6193-841-4
2. VA Najjar (ed.): ImmunologicallyActive Peptides. 1981 ISBN 90-6193-842-2
3. VA Najjar (ed.): Enzyme Induction and Modulation. 1983 ISBN 0-89838-583-0
4. VA Najjar and L. Lorand (eds.): Transglutaminase. 1984 ISBN 0-89838-593-8
5. GJ. van der Vusse (ed.): Lipid Metabolism in Normoxic andIschemic Heart. 1989 ISBN 0-7923-0479-9
6. J.F.c. Glatz and GJ. van der Vusse (eds.): Cellular Fatty Acid-Binding Proteins. 1990
ISBN 0-7923-0896-4
7. H.E. Morgan (ed.): Molecular Mechanisms of Cellular Growth. 1991 ISBN 0-7923-1183-3
8. GJ. van der Vusse and H. Stam (eds.): Lipid Metabolism in the Healthy and Diseased Heart. 1992
ISBN 0-7923-1850-1
9. Y. Yazaki and S. Mochizuki (eds.): Cellular Function and Metabolism. 1993 ISBN 0-7923-2158-8
10. J.F.C. Glatz and GJ. van der Vusse (eds.): Cellular Fatty-Acid-Binding Proteins, II. 1993
ISBN 0-7923-2395-5
11. R.L. Khandelwal and lH. Wang (eds.): Reversible Pro tein Phosphorylation in Cell Regulation. 1993
ISBN 0-7923-2637-7
12. J. Moss and P. Zahradka (eds.): ADP-Ribosylation: Metabolic Effects and Regulatory Functions. 1994
ISBN 0-7923-2951-1
13. VA Saks and R. Ventura-Clapier (eds.): Cellular Bioenergetics: Role ofCoupled Creatine Kinases. 1994
ISBN 0-7923-2952-X
14. J. Slezäk and A Ziegelhöffer (eds.): Cellular Interactions in Cardiac Pathophysiology. 1995
ISBN 0-7923-3573-2
15. J.A Barnes, H.G. Coore, AH. Mohammed and R.K. Sharma (eds.): Signal TransductionMechanisms. 1995
ISBN 0-7923-3663-1
16. A.K. Srivastava and J.-L. Chiasson (eds.): Vanadium Compounds: Biochemical and TherapeuticApplications.
1995 ISBN 0-7923-3763-8
17. J .M.J. Lamers and P.D. Verdouw (eds.): Biochemistry of Signal Transduction in Myocardium. 1996
ISBN 0-7923-4067-1
18. E.-G. Krause and R. Vetter (eds.): Biochemical Mechanisms in Heart Function. 1996
ISBN 0-7923-4118-X
19. R. Vetter and E.-G. Krause (eds.): Biochemical Regulation of Myocardium. 1996 ISBN 0-7923-4259-3
20. G.N. Pierce and w.c. Claycomb (eds.): Novel Methods in Molecular and Cellular Biochemistry of Muscle.
1997 ISBN 0-7923-4387-5
21. F.N. GelIerich and S. Zierz (eds.): Detection ofMitochondrial Diseases. 1997 ISBN 0-7923-9925-0
22. P.K. Singal, V. Panagia and G.N. Pierce (eds.): The Cellular Basis ofCardiovascular Function in Health and
Disease. 1997 ISBN 0-7923-9974-9
23. S. Abdel-aleem and J.E. Lowe (eds.): Cardiac Metabolism in Health and Disease. 1998
ISBN 0-7923-8104-1
24. A.K. Srivastava and B. Posner (eds.): Insulin Action. 1998 ISBN 0-7923-8113-0
Developments in Molecular and Cellular Biochemistry

25. V.A. Saks, R. Ventura-Clapier, X. Leverve, A. Rossi and M. Rigoulet (eds.): Bioenergetics 01 the Cell:
Quantitative Aspects. 1998 ISBN 0-7923-8118-1
26. G.N. Pierce, H. Rupp, T. Izumi and A. Grynberg (eds.): Molecular and Cellular Effects 01 Nutrition on
Disease Processes. 1998 ISBN 0-7923-8171-8
27. K. Ahmed, E. Chambaz and O.G. Issinger (eds.): Molecular and Cellular View 01 Protein Kinase CK2. 1998
ISBN 0-7923-8208-0
28. M.Y. Cohen, J.M. Downey, RJ. Gelpi and J. Slezak (eds.): Myocardial Ischemia and Reperfusion. 1998
ISBN 0-7923-8173-4
29. D.A. Bernlohr and L. Banaszak (eds.): Lipid Binding Proteins within Molecular and Cellular Biochemistry.
1998 ISBN 0-7923-8223-4
30. R. Alvarez-Gonzalez (ed.): ADP-Ribosylation Reactions: From Bacterial Pathogenesis to Cancer. 1998
ISBN 0-7923-8235-8
31. S. Imai and M. Endo (eds.): Muscle Physiology and Biochemistry. 1998
ISBN 0-7923-8264-1

KLUWER ACADEMIC PUBLISHERS - DORDRECHT / BOSTON / LONDON

You might also like