January 27, 2001 BRIEF COMMUNICATIONS 331
REFERENCES Transplantation in miniature swine. I. Fixation of the major
1. Calne RY, Yoffa DE, White HJO, Maginn RR. A technique of
orthotopic liver transplantation in the pig. Br J Surg 1968; 55:
203.
2. Terblanche J, Peacock JH, Bowes J, Hobbs KEF. The technique
of orthotopic liver homotransplantation in the pig. J Surg Res
1968; 8: 151.
3. Lerut J. Orthotopic liver transplantation: experimental and clin-
ical study. (Thesis) University of Berne, 1987.
4. Kamada N, Calne RY. Orthotopic liver transplantation in the
rat: technique using cuff for portal vein anastomosis and bili-
ary drainage. Transplantation 1979; 28: 47.
5. Sachs DH, Leight G, Cone J, Schwarz S, Stuart L, Rosenberg S.
Downloaded from https://journals.lww.com/transplantjournal by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 02/16/2021
VITAMIN E INHIBITS RENAL mRNA EXPRESSION OF COX II,
HO I, TGF␤, AND OSTEOPONTIN IN THE RAT MODEL OF
CYCLOSPORINE NEPHROTOXICITY
JOHN K. JENKINS,1 HONG HUANG, KENNETH NDEBELE,
Department of Medicine, University of Mississippi Medical Center, Jackson, Mississippi
Background. In a rat model of cyclosporine (CsA)
nephrotoxicity, vitamin E preserves renal function
and reduces free radicals, vasoconstrictive thrombox-
anes, and tubulointerstitial fibrosis. We examined the
effect of vitamin E on tubule gene expression in this
model.
Methods. In two of three groups, rats were treated
with either CsA, or CsA plus vitamin E, whereas the
control group received vehicles. We pooled purified
tubules or whole kidney tissue in a novel manner to
represent each treatment group, harvested RNA, and
performed rigorously controlled qualitative reverse
transcription-polymerase chain reaction.
Results. Cyclooxygenase (COX) I mRNA was detect-
able in control animals, was increased by CsA, but was
unchanged by vitamin E. COX II mRNA was detected
in controls, was inhibited in the CsA group, and was
further inhibited with vitamin E. Hemeoxygenase I
and TGF-␤ and osteopontin mRNA were increased in
the CsA-treated group and were inhibited by vitamin
E.
Conclusions. Our data support the involvement of
free radicals, COX pathways, and pro-fibrotic genes in
cyclosporine nephrotoxicity and suggest that the sal-
utary effect of vitamin E involves the suppression of
some of these genes.
Cyclosporine A (CsA) is a potent and selective immunosup-
pressive agent, and its introduction has permitted a steady
increase in kidney and other organ transplantation. How-
ever, its clinical use has not been fully realized because of its
frequent and at times severe renal toxicity. Several studies
have been undertaken to understand the mechanism of this
unwanted side effect (1, 2). Two main components of CsA
1
Address correspondence to: John K. Jenkins, MD, Division of
Rheumatology/Molecular Immunology, Department of Medicine,
University of Mississippi Medical Center, 2500 North State Street,
Jackson, MS 39216-4505.
histocompatibility complex. Transplantation 1976; 22: 559.
6. Monden M, Barters RH, Fortner JG. A simple method of ortho-
topic liver transplantation in dogs. Ann Surg 1982; 195: 110.
7. Flye MW, Pennington L, Kirkman R, Weber B, Sindelar W,
Sachs DH. Spontaneous acceptance or rejection of orthotopic
liver transplants in outbred and partially inbred miniature
swine. Transplantation 1999; 68: 599.
8. Goto S. Orthotopic liver transplantation in miniature pigs: some
technical considerations in recipient. J Jpn Surg Soc 1988; 89:
215.
Received 17 February 2000.
Accepted 21 May 2000.
AND ABDULLA K. SALAHUDEEN
nephrotoxicity have been recognized: (1) an acute and revers-
ible renal vasoconstriction leading to acute reductions in
renal blood flow and glomerular filtration rate, and (2) a
chronic and irreversible tubulointerstitial fibrosis leading to
chronic renal failure. Mediators and mechanisms postulated
to be involved include thromboxanes and prostaglandins,
fibrogenic growth factors including TGF␤ as well as reactive
oxygen species (ROS) (2– 4). CsA-induced renal production of
vasoconstrictive thromboxanes may result from lipid hy-
droperoxidation and/or induction of cyclooxygenase. The his-
tologic findings in CsA nephrotoxicity are primarily those of
tubulointerstitial fibrosis. Recent reports suggest a role of
renal expression of fibrogenic growth factors such as TGF␤
and extracellular matrix proteins including osteopontin, col-
lagens, and fibronectin (3–5).
Regarding the possible role of cyclooxygenase in the pro-
duction of vasoconstrictive prostanoids, the effects of CsA on
tubule expression of either isoform of cyclooxygenase (COX)
is not clear. Non-specific COX inhibitors have been shown to
worsen CsA-induced nephrotoxicity (6). Differential up-reg-
ulation of COX II has therapeutic implication in CsA-neph-
rotoxicity, because specific COX II inhibitors are now avail-
able and could inhibit renal vasoconstrictor production
without inhibiting the vasodilatory prostanoids elsewhere
downstream of COX I.
We previously hypothesized that the antioxidant vitamin E
suppresses CsA-induced lipid hydroperoxides and renal pro-
duction of thromboxanes (1). In an earlier study, we found
that administration of vitamin E suppressed tubulointersti-
tial fibrosis and limited renal dysfunction when coadminis-
tered with CsA in the rat model of CsA-induced nephrotox-
icity. There was essentially no glomerular damage. Increase
in urinary metabolites of prostanoid vasoconstrictors was
also associated with CsA nephrotoxicity and similarly was
reduced with vitamin E. In this report we have further ex-
plored the mechanism of CsA-induced tubulointerstitial in-
332 TRANSPLANTATION
flammatory and fibrogenic processes and their inhibition by
vitamin E by examining in vivo the renal tubule expression of
genes thought to be involved in the pathologic processes of
fibrosis, response to oxidative stress, and vasoconstrictor pro-
duction. Understanding these processes may help us to de-
vise specific clinical strategies to minimize CsA toxicity and
expand its use in both organ transplantation and autoimmu-
nity.
Male Sprague-Dawley rats were matched for body weight
(275–300 g) and assigned to three groups consisting of five to
six rats each. Animal protocols were approved by the insti-
tutional animal use committee. The three groups of animals
were given the following daily injections: (1) control rats, 0.3
ml/kg olive oil (CsA diluent) subcutaneously and 0.5 ml/kg of
castor oil (vitamin E diluent) intraperitoneally; (2) CsA
group, CsA 25 mg/kg subcutaneously and olive oil intraperi-
toneally; (3) CsA plus vitamin E group, vitamin E (dL-␣-
tocopherol) 25 mg/kg intraperitoneally and castor oil subcu-
taneously. CsA was generously donated by Sandoz (East
Hanover, NJ), and vitamin E was purchased from Sigma (St.
Louis, MO). Rats were cared for as previously described (1).
At study end, renal and histological data were determined
by methods we previously reported. Data were analyzed by
ANOVA for comparison between the three groups. Results
are presented as mean⫾SEM in Table 1. Body weight and
hematocrit were unchanged between the groups. Serum vi-
tamin E levels were threefold higher in the vitamin E-treated
group, 14 versus 4.9 ␮g/ml. The administration of CsA caused
significant renal functional impairment, as shown by eleva-
tion of the serum creatinine from 0.38 to 0.64 mg/dl and
reduction in the creatinine clearance from 2.1 to 1.04 ml/min.
The histologically determined fibrosis score increased from
0.4 to 2.4 with CsA. Vitamin E prevented fibrosis, reversing
the score to near normal: score⫽0.6. Vitamin E reduced the
elevated serum creatinine from 0.64 to 0.47, or 65%, and
inhibited the reduction in creatinine clearance from 1.04 to
1.5, or 51%.
For RNA analysis, we first isolated tubules from whole
kidney tissue and glomeruli. At sacrifice, 12⁄3 decapsulated
kidneys from each rat were minced and forced through con-
secutively smaller wire sieves (220 mesh, 250 ␮m; then 120
mesh, 150 ␮m; Fisher Scientific, Atlanta, GA) with a pestle.
Material remaining on the smaller sieve was 98% pure tu-
bules by light microscopy. Fractions from at least three ani-
mals were pooled to ensure an adequate RNA yield. Whole
kidney tissue was prepared also. Fractions were lysed with 4
M guanidinium thiocyanate and homogenized cold. One-
TABLE 1. General measurements, renal function, and renal
fibrosis score in three groups of rats at the end of the study
CsA ⫹
Control CsA
vitamin E
Body wt (g) 360⫾10 333⫾10 340⫾9
Hematocrit 51⫾1 49⫾3 49⫾2
Serum vitamin E (␮g/ml) 4.9⫾0.6 4.3⫾0.6 14⫾1.8a
Serum creatinine (mg/dl) 0.38⫾0.02 0.64⫾.02a 0.47⫾0.06
Creatinine clearance 2.10⫾0.20 1.04⫾0.09a 1.50⫾0.12b
(ml/min)
Tubulointerstitial fibrosis 0.40⫾0.30 2.40⫾0.25a 0.60⫾0.25
score
a
P⬍0.05 vs. all other groups.
b
P⬍0.05 vs. control only.
Vol. 71, No. 2
tenth volume of 20% sarkosyl was added, and the mixture
was heated (65°C, 5⬘). These pooled fractions were centri-
fuged through a CsCl2 cushion, and RNA was prepared as
described (7). Spectrophotometric quantification and qualita-
tive analysis were performed to determine the quality and
ensure that no gross miscalculations were made.
Reverse transcription-polymerase chain reaction (RT-PCR)
analysis of kidney RNA was performed in a standard fashion
similar to a published protocol (8), and is described briefly here.
Standard reverse transcription (RT) of RNA was performed by
using equal ␮g input RNA from the sample from each animal
group: 1X reaction buffer, 1X hexanucleotide mix, 1 mM dNTP,
5 U AMV reverse transcriptase, 1 U RNasin (all Promega,
Madison, WI) in a 20-␮l reaction for 50⬘ at 42°C, followed by
95°C for 5⬘. Five microliters of this reaction were used for PCR
amplification in a 25-␮l reaction, representing amplification of
cDNA generated from 0.25 ␮g of total cellular RNA. The PCR
reaction was performed under the following conditions in a
Perkin-Elmer DNA Thermal Cycler: 10 pmol each primer, 200
␮M dNTP, 1.25 U Taq, 1X Taq reaction buffer (all Promega), as
previously described (7). PCR reactions were optimized for an-
nealing temperature and were performed at various cycle num-
bers to determine the linear range of the amplification curve.
Densitometry on Polaroid negatives confirmed the linearity of
the amplification range. Results shown here are from amplifi-
cations falling in the middle of the linear curve. ␤-Actin was
used as a control and bactin amplification was performed at
least once for each transcription reaction. Polaroid photos of
ethidium-stained gels are shown. Primers were rat specific:
␤-actin, sense 5⬘ CCAACCGTGAAAAGATGACC, antisense 5⬘
CAGGAGGAGCAATGATCTTG; COX I, sense 5⬘ TGCATGTG-
GCTGTGGATGTCATCAA, antisense 5⬘ CACTAAGACAGAC
CCGTCATCTCAA; COX II, sense 5⬘ ATCCACTCAGTTTGTT-
GAGTCATTC, antisense 5⬘ TTTGATTAGTACTGTAGGG-
TTAATG; HO I, sense 5⬘ TGGAAGAGGAGATAGAGCGA,
antisense 5⬘ TGTTGAGCAGGAAGGCGGTC; HO II, sense 5⬘
CAGTTGAGCAGGAAGGCGGTC, antisense 5⬘ CTTCATACT-
CAGGTCCAAGGC; TGF␤1 , sense 5⬘ CCTGAGTGGCT-
GTCTTTTGA, antisense 5⬘ GCGCACAATCATGTTGGACA; os-
teopontin, sense 5⬘ CAGCCAAGGACCAACTACAACCAT,
antisense 5⬘ GCACAGAAAGAACAGAAGCGAAAT. RNA anal-
ysis performed in this manner represents the average gene
expression in each treatment group. The results are highly
qualitative when performed in this fairly rigorous fashion.
Induced COX products may be responsible for the produc-
tion of vasoconstrictive thromboxanes leading to the reduc-
tion in renal blood flow in CsA-induced nephrotoxicity. We
first examined the expression of COX I and II mRNA in the
tubules, because prostanoids may be involved in the inflam-
mation/fibrosis, as well as vasoconstriction. As can be seen in
the upper panel of Figure 1, there is an increase in the renal
tubule steady state level of COX I mRNA in the CsA-treated
group compared with the control group. Vitamin E did not
have any significant effect on CsA-induced COX I. Surpris-
ingly, COX II mRNA expression was readily detectable in the
tubule fraction (first lane, C), though at a moderately high-
amplification cycle number, 35 cycles. CsA-treated animals
had a significantly lower expression of COX II mRNA than
did controls. However, animals treated concomitantly with
vitamin E did not have detectable COX II mRNA. We verified
these results in combined tubule tissue from two more rats,
which suggests that our PCR technique is quite sensitive.
January 27, 2001 BRIEF COMMUNICATIONS
FIGURE 1. COX mRNA expression. Total cellular RNA from
pooled whole kidneys or purified tubules were analyzed for
COX I and COX II mRNA expression by RT-PCR: cycle num-
bers were 28 and 35, respectively; annealing temperatures
were 56°C and 58°C, respectively. ␤-Actin was amplified as a
control from each different reverse transcription reaction (27
cycles, annealing temperature 60°C). C, control (untreated)
rats; CsA, rats treated with cyclosporine A; CsA/E, rats
treated concomitantly with vitamin E and CsA.
To confirm these results, however, we performed COX I
and II RT-PCR on RNA prepared from pooled whole kidneys
(lower panel, Fig. 1). A control sample to verify the basal
COX II expression was not available. The whole kidney prep-
aration is mostly tubules and therefore likely to reflect gene
expression in the tubule fraction, unless the gene examined
is predominantly expressed elsewhere. As in the upper panel,
rats with CsA nephrotoxicity given vitamin E had no signif-
icant change in the COX I expression, compared with the
CsA-treated group not given vitamin E. However, COX II
mRNA was reduced with vitamin E, as in the tubule-only
fraction. It has previously been reported that CsA alone does
not induce COX I or II but does inhibit IL-1␤-, calcium
ionophore- or LPS-induced COX II (but not COX I) mRNA
and protein expression in rat mesangial cells in vitro (8). Our
data show that CsA inhibits COX II but not COX I mRNA in
vivo in rat tubules.
Based on our previous report of the effects of vitamin E on
prostanoid metabolites, we hypothesized that CsA would in-
crease cyclooxygenase and that vitamin E would decrease it
(1). The latter hypothesis was confirmed. COX II might also
be down regulated by its products. Detectable expression of
COX II mRNA in the control group is probably due to the
sensitivity of our PCR technique, and we have confirmed in
vivo that CsA inhibits COX II. It is possible that COX I plays
a significant role, because its mRNA was expressed at rela-
tively higher steady state levels compared with COX II in
this study. Based on amplification cycle differences and band
intensity, we estimate an 18-fold higher basal expression of
COX I mRNA over COX II. The role of cyclooxygenase and
prostanoids in CsA nephrotoxicity is likely to be complex, and
the potential role of specific antiinflammatory COX II inhib-
itors is unclear.
333
Free radicals may be involved in the inflammatory or fibrotic
process in CsA nephrotoxicity. One mechanism for handling
these potential mediators of renal injury is through the enzyme
induction responsible for scavenging these molecules.
Hemeoxygenase (HO) has been shown to be up-regulated in
macrophages in glomerular disease (9). Therefore, we examined
the expression of HO I and II (inducible and constitutively
expressed isoforms, respectively) by RT-PCR in tubules of these
groups of animals, as described above. Control rats had no
detectable HO I mRNA (Fig. 2). CsA-treated animals expressed
easily detectable HO I mRNA in tubules. Vitamin E slightly
inhibited CsA-induced HO I mRNA compared with the group
treated only with CsA. There was no change in the steady state
HO II mRNA levels in animals in the three treatment groups.
HO is responsible for degrading heme to free iron, CO, and
bilirubin, the latter being a potent scavenger of free radicals.
Free iron induces ferritin, another potent scavenger of free
radicals. Induction of HO I may be a protective mechanism
against free radical-induced tubular damage in CsA-induced
nephrotoxicity. The inhibition HO I induction with vitamin E
may merely be reflective of the reduction of free radical-induced
damage via vitamin E or a direct antagonist effect of vitamin E
on the induction of HO I.
Next we were interested in genes involved in the fibrotic
process, inasmuch as vitamin E reduces histologically evident
fibrosis in CsA nephrotoxicity. TGF␤ has been implicated in
numerous fibrotic diseases, and TGF␤ protein reportedly is
associated with CsA-induced nephritis, although the effects of
vitamin E are unknown (3, 5). We examined TGF␤1 mRNA
expression by RT-PCR of tubule RNA (Fig. 2, middle panel).
TGF␤ mRNA was increased in animals with CsA-induced ne-
phritis, compared with controls. Vitamin E inhibited CsA-in-
duced steady state levels of TGF␤ mRNA to below that seen in
control rats. We next examined a TGF␤-inducible gene thought
to be involved in fibrosis, osteopontin. Osteopontin is a constit-
FIGURE 2. Hemeoxygenase, TGF␤, and osteopontin mRNA ex-
pression in tubule cells. Total cellular RNA from pooled kid-
ney tubules was analyzed for HO I and II; TGF␤ and os-
teopontin mRNA expression were analyzed by RT-PCR. Cycle
numbers and annealing temperatures were: HO I, 35 cycles
and 58°C; HO II, 31 cycles and 58°C; TGF␤1, 35 cycles and 60°C;
and osteopontin, 30 cycles and 57°C. ␤-Actin served as a con-
trol, and lanes are labeled as in Figure 1.
334 TRANSPLANTATION
uent of the increased connective tissue in fibrotic processes (10).
As seen in the bottom panel of Figure 2, steady state levels of
osteopontin mRNA were easily detectable in controls and were
increased in the tubule fractions from animals treated with
CsA. Vitamin E also dramatically inhibited CsA-induced os-
teopontin mRNA (Figure 2, bottom panel).
In summary, we have examined the effects of the antioxi-
dant vitamin E on gene expression in the rat model of CsA
nephropathy. Vitamin E significantly inhibits CsA-induced
renal damage, including the tubulointerstitial fibrosis score
and prostanoids (ref. 1 and this report). CsA induces the
expression of COX I, HO I, TGF␤, and osteopontin in vivo in
the rat, and this expression is inhibited by concomitant treat-
ment of the animals with vitamin E. COX II was also inhib-
ited by vitamin E treatment, although we did not find that it
was induced by CsA. Whereas the full scope of the mecha-
nism of the renal effects of CsA are not yet known, they
appear to include multiple arms of the vasoconstrictive, in-
flammatory, and fibrotic pathways. We have not yet deter-
mined the relative contribution of the various processes, but
vitamin E appears to strongly affect fibrosis from CsA, re-
gardless of whether the gene expression effects are direct or
indirect. CsA nephrotoxicity is inhibited by the antioxidant
vitamin E, which has a broad and specific effect on the
pathologic processes initiated by CsA. Our findings suggest
that vitamin E should be studied clinically as a possible
preventative agent against the development of nephrotoxic-
ity in those patients in whom CsA is prescribed.
REFERENCES
1. Kanji VK, Wang C, Salahudeen AK. Vitamin E suppresses cy-
closporine A-induced increase in the urinary excretion of ara-
CHRONIC DIARRHEA AS A RESULT OF INTESTINAL
MICROSPOSIDIOSIS IN A LIVER TRANSPLANT RECIPIENT
MARTIN GOETZ,1,2 SUSANNA EICHENLAUB,3 GERD R. PAPE,1
Klinikum Grosshadern, Second Medical Department, University of Munich, 81377 Munich, Germany; and Department of
Infectious Diseases and Tropical Medicine, University of Munich, 80802 Munich, Germany
Background. Microsporidia are common pathogens
among patients infected with human immunodefi-
ciency virus. They account for a substantial propor-
tion of chronic diarrhea and malabsorption in ac-
quired immune deficiency syndrome, but their
appearance after solid organ transplantation has only
rarely been reported.
Methods. We report what we believe is the first case
of documented Enterocytozoon bieneusi infection in a
liver transplant recipient.
Results. Our patient presented with chronic diar-
rhea and colicky abdominal pain. Although symptoms
were severe, only mild microscopical mucosal changes
1
Klinikum Grosshadern, Second Medical Department.
2
Address correspondence to: Martin Goetz, M.D., I. Med. Dept.,
Universitaetsklinik Mainz, Langenbeckstr. 1, 55131 Mainz, Germany.
3
Department of Infectious Diseases and Tropical Medicine, Uni-
versity of Munich.
Vol. 71, No. 2
chidonic acid metabolites including F2-isoprostanes in the rat
model. Transplant Proc 1999; 31(3): 1724.
2. Wang C., Salahudeen AK. Lipid peroxidation accompanies cyclo-
sporine nephrotoxicity: effect of vitamin E. Kidney Int 1995;
47(3): 927.
3. Shihab FS, Andoh TF, Tanner AM, Bennett, WM. Sodium de-
pletion enhances fibrosis and the expression of TGF-beta 1 and
matrix proteins in experimental chronic cyclosporine nephrop-
athy. Am J Kidney Dis 1997; 30(1): 71.
4. Shihab FS, Andoh TF, Tanner AM, et al. Role of transforming
growth factor-beta 1 in experimental chronic cyclosporine ne-
phropathy. Kidney Int 1996; 49(4): 1141.
5. Wolf G, Thaiss F, Stahl RA. Cyclosporine stimulates expression
of transforming growth factor-beta in renal cells: possible
mechanism of cyclosporines antiproliferative effects. Trans-
plantation 1995; 60(3): 237.
6. Williamson HE. Interaction of cyclosporine and indomethacin in
the rat. Res Commun Chem Pathol Pharmacol 1988; 61: 141.
7. Jenkins JK, Drong RF, Shuck ME, et al. Intracellular interleu-
kin-1 receptor antagonist promoter: cell type-specificity and
inducible regulatory regions. J Immunol 1997; 158(2): 748.
8. Martin M, Neumann D, Hoff T, Resch K, DeWitt D, Goppelt-
Struebe M. Interleukin-1-induced cyclooxygenase 2 is sup-
pressed by cyclosporin A in rat mesangial cells. Kidney Int
1994. 45: 150.
9. Mosley K, Wembridge DE, Cattell V, Cook T. Heme oxygenase is
induced in nephrotoxic nephritis, and hemin, a stimulator of
heme oxygenase synthesis, ameliorates disease. Kidney Int
1998; 53: 672.
10. Denhardt DT, X Guo. Osteopontin: a protein with diverse func-
tions. FASEB J 1993; 7: 1745.
Received 12 April 2000.
Accepted 6 June 2000.
AND ROBERT M. HOFFMANN1
were found in the intestinal tract. A modified
trichrome stain of stool specimens revealed microspo-
ridial spores, and species differentiation by restric-
tion fragment length polymorphism polymerase chain
reaction identified Enterocytozoon bieneusi. Albenda-
zole therapy brought symptomatic relief but no micro-
biological clearance.
Conclusions. Enterocytozoon bieneusi may cause
chronic diarrhea not only in immunosuppression as a
result of human immunodeficiency virus infection but
also among patients with therapeutic immunosup-
pression after organ transplantation. Therefore, mi-
crosporidial infection should be considered in immu-
nosuppressed patients with otherwise unexplained
diarrhea.
Microsporidia are obligate intracellular protozoa of the
phylum Microspora. Of the more than 1000 species only
Enterocytozoon, Encephalitozoon, Nosema, Pleistophora, Tra-