Regulation of the sodium pump in pregnancy-related tissues in

preeclampsia
Cynthia V. Maxwell, MD, Qing-Feng Tao, MD, Ellen W. Seely, MD, John T. Repke, MD, and Steven
W. Graves, PhD
Boston, Massachusetts

OBJECTIVE: This study examined the expression of the three -isoforms of the sodium pump in preeclamp-
sia. Reductions in sodium pump number and activity in smooth muscle may underlie hypertension in
preeclampsia.
STUDY DESIGN: Northern and Western analyses were used to determine whether sodium pump -isoform
regulation in myometrium, placenta, and umbilical artery of women with preeclampsia differed from those
with normotensive pregnancies.
RESULTS: Levels of 1 and 3 messenger ribonucleic acid were reduced in myometrium of women with
preeclampsia compared with normotensive pregnancies, as was 2 messenger ribonucleic acid in
preeclamptic placenta. Protein expression of the -isoforms was unaltered in placenta and umbilical artery
from women with preeclampsia versus those with normotensive pregnancies, but myometrial 2 protein lev-
els were reduced significantly in women with preeclampsia. Moreover, myometrial 1 protein expression was
undetectable.
CONCLUSIONS: Reduced smooth muscle sodium pump expression in preeclampsia may raise cell sodium,
increase pressor sensitivity, or increase tone directly, which may contribute to hypertension in preeclampsia.
(Am J Obstet Gynecol 1998;179:28-34.)

Key words: Preeclampsia, sodium and potassiumactivated adenosine triphosphatase, my-

ometrium, placenta, umbilical artery, messenger ribonucleic acid

Preeclampsia is a potentially severe form of pregnancy-
induced hypertension characterized by vascular endothe-
lial damage, hypertension, and increased vascular tone.1, 2
The mechanisms mediating the increased vascular tone
are incompletely understood.1, 2 One cellular transport
system affecting smooth muscle tone is the sodium and
potassiumactivated adenosine triphosphatase (Na,K-
ATPase) or sodium pump. Reductions in its activity can
result in increased muscle contraction.3 Studies of blood
cells from patients with essential hypertension have gen-
erally demonstrated reductions in Na,K-ATPase activity.4
Most similar studies in pregnant women carefully defined
as preeclamptic have found similar reductions.5-8
Reductions in sodium pump activity increase intracellu-
lar sodium and calcium levels,3 and similar increases are
seen in blood cells from women with preeclampsia.9 Such
alterations in sodium pump activity might be achieved in
From the Endocrine-Hypertension Division, Department of Medicine,
and the Department of Obstetrics and Gynecology, Harvard Medical
School, Brigham and Womens Hospital.
Supported in part by Howard Hughes Institute.
Received for publication June 23, 1997; revised September 15, 1997; ac-
cepted December 29, 1997.
Reprint requests: Steven W. Graves, PhD, Endocrine-Hypertension
Division, Brigham and Womens Hospital, 221 Longwood Ave, Boston,
MA 02115.
Copyright 1998 by Mosby, Inc.
0002-9378/98 $5.00 + 0 6/1/88530
two general ways: a reduction in the number of sodium
pump units, for example, Miyamoto et al6 found reduced
sodium pump sites in erythrocytes from pregnant
women with preeclampsia compared with normotensive
pregnancy or an increase in levels of circulating sodium
pump inhibitors, as also reported in women with
preeclampsia.10 Sodium pump inhibitors can directly re-
duce Na,K-ATPase activity but may also reduce sodium
pump number in the cell membrane through a process
similar to endocytosis.11, 12
The sodium pump contains two subunits: an -sub-
unit, which is catalytically active, and a -subunit, which
may be instrumental in protein assembly and membrane
insertion.13 The -subunit occurs as three isoforms that
differ enzymatically, for example, different turnover
rates, and in their tissue distribution.14 Recent studies of
experimental hypertension have suggested that alter-
ations in the -isoforms may contribute to reduced
sodium pump activity in some tissues.15 Our hypotheses
for this study were two. First, we predicted that sodium
pump -isoform expression would be modified in re-
sponse to pregnancy. Second, we hypothesized that
sodium pump -isoform expression would be altered in
critical tissues in preeclamptic pregnancies compared
with normotensive pregnancies. Reductions in -isoform
number or alterations in their distribution in preeclamp-
sia, whether a primary or compensatory event, may con-

28
Volume 179, Number 1
Am J Obstet Gynecol
tribute to increased smooth muscle tone and hyperten-
sion in preeclampsia. Studies of sodium pump -isoforms
have not been thorough in human pregnancy, especially
when complicated by preeclampsia.
Material and methods
Subjects. Women in this study were patients at
Brigham and Womens Hospital in Boston,
Massachusetts. The protocol had been approved by the
Human Subjects Committee, and all participants gave in-
formed signed consent. All normotensive and
preeclamptic pregnant women enrolled in the study
whether undergoing elective or emergency cesarean de-
livery (normotensive 19/19, preeclamptic 13/16) had
their medical records reviewed to confirm diagnosis and
obtain other relevant data. None of these women was in
labor or had undergone induction. Two preeclamptic
women who underwent vaginal delivery provided speci-
mens of placenta only. Both were in spontaneous labor.
Normotensive pregnant women had no history of hyper-
tension and no evidence of hypertension or proteinuria
during the current pregnancy. First-, second-, and third-
trimester ambulatory blood pressure readings were veri-
fied as normal (blood pressure <140/90 mm Hg).
American College of Obstetricians and Gynecologists cri-
teria were used to define women as preeclamptic.16 All
preeclamptic women had normal first-trimester blood
pressures. No pregnant subject had other diseases, in-
cluding diabetes and renal disease.
In an attempt to model human vascular smooth mus-
cle, we used uterine smooth muscle.17 Additional tissues
were also obtained: placenta as a mechanically inactive
control and umbilical artery to represent fetal smooth
muscle. Tissue specimens obtained at cesarean section
included an ~1-cm3 specimen of myometrium taken
close to the incision and placenta and umbilical cord.
The placental tissue excluded the chorion and amnion
but may have included a small amount of decidual tissue.
Myometrial samples were collected from nonpregnant
women undergoing hysterectomy. Only normal tissue
was used. Tissue samples were immediately frozen on dry
ice and stored at 70C until processed.
Preparation of tissue microsomes. Tissue samples were
minced with a razor, homogenized with 6 strokes at
medium speed with a glass-Teflon homogenizer in ice-
cold homogenization buffer (320 mmol/L sucrose, 1
mmol/L ethylenediaminetetra-acetic acid, 50 mmol/L
Trishydrochloric acid; pH 7.4), and then centrifuged ac-
cording to published methods.18
Northern blot analysis. Equal amounts of total ribonu-
cleic acid (RNA) (8 to 10 g) were separated with elec-
trophoresis on a 1% agarose and formamide (0.47
mmol/L) gel, transferred to nylon membrane (Duralon-
UV Membranes, Stratagene, La Jolla, Calif), and probed
with labeled complementary deoxyribonucleic acid
Maxwell et al 29
(cDNA) fragments of human 1 (2.2 kb), 2 (125 bp),
or 3 (100 bp) isoforms of Na,K-ATPase (gifts of Dr Paul
Allen, Childrens Hospital, Boston) and cDNA fragments
of mouse -actin (to assess potential loss of messenger
RNA (mRNA), 500 bp, Ambion, Austin, Tex) as de-
scribed elsewhere.19 Signal density was linear over the
range of 1 to 16 g RNA. After the intensive dissection re-
quired to isolate umbilical arteries was performed, -iso-
form mRNA was reduced to undetectable levels in almost
all specimens.
Western blot analysis. Microsomal [Na,K]ATPase (20
g) was separated by 7.5% sodium dodecyl sulfatepoly-
acrylamide gel electrophoresis, transferred to nitrocellu-
lose membranes (Schleicher and Schuell, Keene, NH),
and probed only once with well-characterized, isoform-
specific murine monoclonal antibodies, 1: McK1, 1:50
dilution; 2: McB2, 1:40 dilution (gifts of Dr Kathleen
Sweadner, Massachusetts General Hospital, Boston)20;
3: 1:1000 dilution (MA3-915, Affinity Bioreagents Inc,
Neshanic Station, NJ), as described previously.21 Signal
response was linear over the range of 5 to 20 g protein.
Statistical analysis. Results were expressed as the mean
SEM. Comparisons of demographic data or myometrial
-isoform abundance among nonpregnant and nor-
motensive women and normotensive pregnant and
preeclamptic women were tested by analysis of variance
with post hoc Newman-Keuls analysis. Comparisons of
placental or umbilical arterial -isoform abundance for
preeclamptic and normotensive pregnant subjects were
made by unpaired Student t test. Interrelationships of
two parameters were evaluated by Pearsons product mo-
ment correlation analysis. A P < .05 was considered statis-
tically significant.
Results
Patient data. Of the 50 patients sampled, 38 met crite-
ria (Table I). Of the 27 pregnant subjects, 19 were nor-
motensive, whereas 8 were preeclamptic. Eleven patients
were nonpregnant and normotensive. Reasons for the 12
exclusions included chronic hypertension (n = 8), dia-
betes (n = 2), hyperthyroidism (n = 1), and transient hy-
pertension of pregnancy (n = 1). By study design
preeclamptic women had significantly higher systolic and
diastolic blood pressures than normotensive pregnant or
nonpregnant women. As expected, these same women
had lower parity (P = .01) and gestational age (P = .0008).
Urine protein, measured in the preeclamptic women
only, was 1534 mg. Normotensive nonpregnant women
were significantly older, typical of women undergoing
hysterectomy. Racial composition and gravidity were not
different among the groups.
Myometrial sodium pump -isoform mRNA abundance
and membrane protein expression
Myometrial -isoform gene expression in nonpregnant
compared with normal pregnant women. mRNA was de-
30 Maxwell et al July 1998
Am J Obstet Gynecol

Table I. Patient data

Statistical
Parameter NNP NTP PE significance

No. 11 19 8
Age (y) 44.2 2.3 35.8 1.0 32.4 3.1 P = .0007
Gravidity 3.1 0.7 2.9 0.3 1.8 0.3 P = .14
Parity 2.4 0.5 1.3 0.3 0.4 0.2 P = .013
Nulliparity (%) 9 32 63 P < .01*
Gestational age (wk) 39.1 0.1 33.8 1.0 P = .0008
Race (black/white/Asian) (%) 18/82/0 11/84/5 13/87/0 NS
SBP (mm Hg) 113 2 110 2 147 4 P < .0001
DBP (mm Hg) 70 1 69 1 91 2 P < .0001
Urine protein (mg/24 h) 1534 544

NNP, Normotensive nonpregnant; NTP, normotensive third-trimester pregnant; PE, preeclamptic third trimester; NS, not significant;
SBP, systolic blood pressure; DBP, diastolic blood pressure.
*2 analysis.

Table II. Myometrial changes with pregnancy and with preeclampsia

Statistical
Isoform NNP NTP PE significance

Absolute mRNA (No.)* 8 6 6

-Actin 75.8 16.3 78.6 13.8 30.8 6.7 P = .06
1 9.0 2.4 12.0 1.4 1.1 0.3 P = .0006
2 90.5 12.8 118.0 12.3 58.7 9.8 P = .02
3 15.9 3.1 17.2 2.0 2.9 1.2 P = .007
Ratio /actin mRNA
1 0.12 0.02 0.18 0.03 0.05 0.01 P = .03
2 1.5 0.2 1.9 0.6 1.8 0.1 P = .62
3 0.29 0.07 0.31 0.13 0.12 0.02 P = .39
Protein (No.) 11 11 8
1 ND ND ND
2 19.0 3.1 9.2 1.4 3.0 1.7 P = .0002
3 10.6 2.1 15.5 4.6 15.5 4.3 P = .54

NNP, Normotensive nonpregnant; NTP, normotensive third-trimester pregnant; PE, preeclamptic third trimester; ND, not detectable.
*Values are 104 optical density units.

tected for all 3 -isoforms of the sodium pump in
human myometrium (Table II). Neither the absolute
abundance of any -isoform mRNA nor the -isoform
abundance normalized to -actin mRNA (Table II) or to
18s ribosomal RNA (not shown) differed significantly be-
tween nonpregnant and pregnant myometrium. -Actin
mRNA levels did not differ significantly between the nor-
motensive pregnant and nonpregnant myometrium. The
presence of 2 and 3 mRNA in human myometrium
has not been previously reported. Although not quantita-
tive, the data suggest markedly greater 2 isoform abun-
dance.
Myometrial -isoform protein expression in normal preg-
nancy. In contrast to the mRNA, only 2 and 3 isoform
protein could be detected in the cell membrane of
human myometrium (Table II), even after the anti-1 an-
tibody titer was increased to 1:40. Positive controls, rat
kidney, human placental, and umbilical artery microso-
mal samples clearly demonstrated the 1 isoform with
comparable or lower antibody titers (data not shown), in-
dicating that the antibody was not at issue. Myometrial
specimens run in the absence of nonspecific blocking
agent showed a weak 1 band, indicating it was present
but at very low levels.
In contrast to the mRNA data for the -isoforms in
nonpregnant and pregnant myometrium, levels of ex-
pressed 2 protein in the myometrial membrane prepa-
rations were significantly decreased in normotensive
pregnant compared with normotensive nonpregnant tis-
sue (Fig 1). No significant correlation was found between
the gestational age and 2 protein expression (r = .44, P
= .18), maternal age, and 2 protein expression (r = .32,
P = .34) or parity and 2 protein expression (r = .20, P =
.56). Myometrial 3 protein levels were unchanged by
pregnancy.
Myometrial -isoform gene expression in preeclampsia.
Absolute mRNA levels of all 3 -isoforms of the sodium
pump mRNA were significantly decreased in preeclamp-
tic compared with normotensive pregnant myometrium
(Table II). Absolute mRNA levels of -actin were not sig-
Volume 179, Number 1
Am J Obstet Gynecol
Fig 1. Comparison of amount of 2 isoform protein expressed in
plasma membrane of myometrium from normotensive nonpreg-
nant women (NNP) and normotensive third-trimester pregnant
women (NTP). Normotensive pregnant women had 52% reduc-
tion in 2 expression (P = .01).
nificantly decreased in preeclamptic myometrium but
were nearly so. Consequently, we compared myometrial
-isoform mRNA levels with -actin mRNA and 18s
rRNA. The -isoform/-actin mRNA levels described a
significant reduction in 1 message (Table II), whereas
the 1 and 3 to 18s RNA ratio demonstrated significant
reductions in abundance in preeclamptic compared with
normotensive pregnant myometrium (Table III). The
18s rRNA levels were slightly reduced in preeclampsia;
however, this would actually reduce the differences in the
-isoform ratios in preeclamptic compared with nor-
motensive pregnant women.
Myometrial -isoform protein expression in preeclampsia.
As was the case for normotensive nonpregnant and nor-
motensive pregnant women, no 1 protein was detected
in myometrium of preeclamptic women. A further sub-
stantial reduction in 2 protein expression occurred in
preeclamptic compared with normotensive pregnant my-
ometrium (Fig 2). Again, no significant correlation was
seen between 2 protein expression and gestational age
(r = .21, P = .62), maternal age (r = .43, P = .29), or par-
ity (r = .13, P = .76). In contrast to the reductions in 3
mRNA abundance in preeclamptic myometrium, protein
levels of 3 were similar in normotensive pregnant and
preeclamptic myometrium.
Placental sodium pump -isoform mRNA abundance
and membrane protein expression
Placental -isoform gene expression in preeclampsia. As
was found for myometrium, mRNA for all 3 -isoforms
was identified in human placenta, a previously unre-
ported finding (Table IV). Placental -actin mRNA ex-
pression did not differ significantly between normoten-
sive pregnant and preeclamptic tissues (Table IV).
Placental 1 and 2 mRNA levels were significantly de-
creased in preeclamptic compared with normotensive
Maxwell et al 31
Fig 2. Comparison of amount of 2 isoform protein expressed in
plasma membrane of myometrium from normotensive third-
trimester pregnant women (NTP) and preeclamptic third-
trimester women (PE). Myometrium from preeclamptic women
demonstrated 65% reduction in 2 expression compared with
myometrium from normotensive pregnant women (P = .023)
and 83% reduction compared with normotensive nonpregnant
women (P = .0009).
pregnant tissue, when expressed as absolute mRNA
abundance, but 2 mRNA abundance was no longer sig-
nificantly reduced (P = .078) when normalized to -actin
mRNA. 3 mRNA levels did not differ between the two
groups.
Placental -isoform protein expression in preeclampsia.
Despite easily detectable mRNA bands for all 3 of the
-isoforms, membrane protein expression of the 2 iso-
form was undetectable in placenta (Table IV). Rat skele-
tal muscle, used as a control, and human myometrial and
umbilical arterial 2 protein were consistently positive.
Although significant decreases in 1 mRNA abundance
have been found in placenta from preeclamptic com-
pared with normotensive pregnant women, 1 protein
expression was not significantly different, nor was placen-
tal 3 protein expression.
Umbilical artery sodium pump -isoform expression
Umbilical artery -isoform mRNA expression in preeclamp-
sia. Because of the difficulty of dissection of umbilical
artery, most arterial RNA extracts were degraded. Only 2
samples (1 normotensive, 1 preeclamptic) were success-
fully processed and revealed expression of all 3 -iso-
forms.
Umbilical artery -isoform protein expression in preeclamp-
sia. Expressed protein levels of all 3 -isoforms of the
sodium pump in the umbilical artery were not signifi-
cantly different in preeclamptic versus normotensive
pregnant tissues (Table V).
Comment
Changes with pregnancy. Our first hypothesis, that
sodium pump -isoform expression would be altered in
32 Maxwell et al July 1998
Am J Obstet Gynecol

Table III. Myometrial changes with preeclampsia corrected for 18s RNA
Isoform NTP PE Statistical significance

No. 6 6
Absolute 18s rRNA* 8.5 0.2 6.8 0.8 P = .04
Ratio mRNA/18s rRNA
1 142.2 18.3 28.0 12.1 P = .005
2 1389 133 1115 244 P = .31
3 204.3 27.3 39.5 16.5 P = .017

NTP, Normotensive third-trimester pregnant; PE, preeclamptic third trimester.

18s rRNA values are 102 optical density units.
*All

Table IV. Placental changes with preeclampsia

Isoform NTP PE Statistical significance

Absolute mRNA (No.) 10 6

-Actin 6.6 0.9 5.5 1.0 P = .40
1 22.7 4.3 9.8 3.9 P = .05
2 101.9 16.4 44.8 18.1 P = .037
3 25.9 2.9 21.7 5.9 P = .49
Ratio /actin mRNA
1 4.0 0.8 1.8 0.6 P = .078
2 17.3 3.0 7.7 2.4 P = .036
3 4.4 0.6 3.0 1.0 P = .61
Protein (No.) 11 7
1 12.3 1.5 13.3 2.0 P = .68
2 ND ND
3 12.2 1.6 13.0 1.9 P = .76

NTP, Normotensive third-trimester pregnant; PE, preeclamptic third trimester; ND, not detectable.
values are 104 optical density units.
*All

Table V. Umbilical artery changes with preeclampsia near term to prepare the uterus for labor, because reduc-
tions in sodium pump activity render smooth muscle
Isoform Statistical
protein NTP PE significance more susceptible to contraction.3 Perhaps this reduction
reflects an adaptive response in some tissues to the high
No. 8 7 metabolic demand of pregnancy. Perhaps the reduction
1 12.7 2.0 14.0 4.0 P = .77
2 12.8 2.3 16.3 4.8 P = .50 simply reflects the milieu of pregnancy.
3 11.2 2.2 15.7 3.0 P = .25 The undetectable levels of the 1 isoform in human
myometrium may not be surprising. Muscle of all types
Values are in optical density units. NTP, Normotensive third-
frequently has more 2 than 1, sometimes much
trimester pregnant; PE, preeclamptic third trimester.
more.19 Low 1 may reflect decreased 1 production,
transport or membrane insertion, or increased turnover,
human pregnancy, was correct in that a significant reduc- and consequently it will be important to study the -sub-
tion occurred in 2 isoform protein in the membrane of units subcellular distribution in myometrium.
pregnant myometrium despite no observable effect on Decreased 2 protein in normotensive third-trimester
2 mRNA abundance. pregnant myometrium accompanying unaltered 2
Two questions arise: what purpose might a reduction mRNA expression was not likely caused by technical
in myometrial 2 protein in the membrane serve, and problems with protein or mRNA stability or extraction,
how is this decrease accomplished without a decrease in because 3 mRNA and protein in these same tissues re-
mRNA? In myometrium the 2 isoform appeared to be mained unchanged. Changes in membrane -isoform
the most abundant. Levels of 3 protein appeared to be levels independent of mRNA changes have been re-
markedly lower than 2 abundance, and 1 was unde- ported and reflect poorly understood regulated differ-
tectable. We can only speculate as to why the changes oc- ences.14 The changes here do not appear to reflect the
curred: Most of these women were near term (gestational subjects age. The 2 reductions may reflect the observa-
age of 39.1 weeks), and reductions in 2 might occur tion that when an active sodium pump unit binds cardiac
Volume 179, Number 1
Am J Obstet Gynecol
glycosides, the enzyme-inhibitor complex is removed
from the cell membrane.11,12 Increased serum levels of
cardiac glycoside-like compounds have been described in
pregnancy including such a factor with 2 selectivity.21
Hence the increased presence of such an inhibitor may
lead to increased 2 pump turnover.11
Changes with preeclampsia. Our second hypothesis
was that reductions in sodium pump activity reported in
preeclampsia may reflect reductions in sodium pump
number or distribution. The hypothesis was found to be
correct in that preeclampsia brought about marked re-
ductions in mRNA abundance for the 1, perhaps the
3, but not the 2 isoform in myometrium. In addition,
significant reductions in 2 and near significant reduc-
tions in 1 mRNA in placenta were seen. The disposition
of the -isoform mRNAs in umbilical artery was not de-
termined because of technical problems resulting from
the difficult and slow dissection. Myometrial levels of -
actin fell, although not significantly, whereas 18s rRNA
fell slightly but significantly. For this reason myometrial
mRNA results were expressed as absolute levels and also
as the level normalized to either -actin (the most com-
mon way of expressing these data in the literature) or 18s
rRNA. The reduction in mRNA of several isoforms in
these two tissues with preeclampsia suggests that the
changes resulted from some feature of the disease that
down-regulated the sodium pump. What might do this is
unknown. Local ischemia may alter pH and in turn influ-
ence intracellular volume or sodium ion concentration,
which can influence sodium pump regulation.22
However, acidosis, increased cell volume, and higher cell
sodium concentrations, which might be anticipated in is-
chemia, are reported to up-regulate the sodium pump in
vitro.22
The results for the isoform protein expression in the
cell membranes of preeclamptic tissues are equally inter-
esting but challenging to explain. Despite the drop in
mRNA levels for the 1 and 2 isoforms in placenta, no
change occurred in the 1 protein expression, whereas
2 was undetectable. These findings might suggest that
the 1 protein may be recruited from other sites within
the cell to maintain membrane levels. Detailed studies of
translation rates, subcellular localization, and sodium
pump insertion now become relevant given these find-
ings.
Of greatest potential importance was the marked de-
crease in 2 protein levels in the myometrium of
preeclamptic compared with normotensive pregnant
women. This result has added impact because of the con-
comitant low levels of 1 protein in this same tissue
membrane. This change appears unrelated to gestational
age, maternal age, and parity. Reduced [Na,K]ATPase ac-
tivity can lead to smooth muscle contraction, although
none of these women was in labor. If this reduction in 2
could be generalized to the maternal vasculature, it
Maxwell et al 33
might contribute to the increased cell sodium, vascular
tone, and hypertension that characterize preeclampsia.3
This idea requires further investigation.
The mechanism for the reduction of 2 protein in the
cell membrane is unknown. Preeclamptic women have
greater levels of an endogenous sodium pump in-
hibitor,10 which in turn might reduce membrane levels of
the sodium pump. Others have described a selective re-
duction in 2 isoform mRNA of the sodium pump in car-
diac muscle of rats with deoxycorticosterone-salt hyper-
tension,23 rats with heart failure, rats with cardiac
hypertrophy, and in failing human heart, but the mecha-
nism is unclear.24 Some have proposed that the changes
reflected a reduction in serum potassium, a situation un-
likely to pertain in preeclampsia.25
In conclusion, this study suggests that human iso-
forms of the sodium pump are highly regulated systems,
demonstrating tissue-specific distribution and response
to pregnancy and preeclampsia. However, the factors
that confer this selective regulation remain elusive.
Differences in the functional quantity of isoform ex-
pressed in the cell membrane require refinements in
technique to clarify mechanism. Whatever the mecha-
nism is, the reductions in 2 coupled with the very low
levels of 1 protein expression diminish sodium pump
reserve and could limit the preeclamptic myometriums
ability to rid itself of sodium. Similar abnormalities in
maternal vascular smooth muscle of preeclamptic
women might explain the increased cell sodium and cal-
cium, increased vascular responsiveness to pressors, and
increased peripheral vascular resistance that underlies
much of preeclampsia.
We thank Drs Paul D. Allen and Kathleen J. Sweadner
for the molecular biologic probes that made this re-
search possible.
REFERENCES
1. Lindheimer MD. Hypertension in pregnancy. Hypertens
1993;22:127-37.
2. Lyall F, Greer IA. Pre-eclampsia: a multifaceted vascular disor-
der of pregnancy. J Hypertens 1994;12:1339-45.
3. Stewart L, Hamilton D, Ingwall J, Naomi S, Graves SW, Canessa
M, et al. Vascular smooth muscle response to ouabain: relation
of tissue sodium to the contractile response. Circ Res
1993;73:1113-20.
4. Poston L, Sewell RB, Wilkinson SP, Richardson PJ, Williams R,
Clarkson EM, et al. Evidence for a circulating sodium transport
inhibitor in essential hypertension. BMJ 1981;282:847-9.
5. Mimura O, Morikawa H, Yamasali M, Mochizuki M. A study on
the membrane Na+ efflux of pregnancy-induced hypertension.
Folia Endocrinol 1988;64:463-81.
6. Miyamoto S, Makino N, Shimokawa K, Wake N, Nakano H. The
characteristics of eukaryotic Na+ transport systems in normal
pregnancy and pregnancy-induced hypertension. J Hypertens
1992;10:367-72.
7. Sultan B, Wennergren M, Jonsson O, Herlitz H. Cellular sodium
transport in severe preeclampsia and chronic hypertension dur-
ing pregnancy. Hypertens Preg 1994;13:135-43
34 Maxwell et al July 1998
Am J Obstet Gynecol

8. Seon R, Forrester T. Relationship between leukocyte sodium
content and high blood pressure during development and reso-
lution of pre-eclampsia. Clin Sci 1989;76:199-203.
9. Kilby MK, Broughton-Pipkin F, Symonds EM. Calcium and
platelets in normotensive and hypertensive human pregnancy. J
Hypertens 1992;10:997-1003.
10. Seely EW, Williams GH, Graves SW. Markers of sodium and vol-
ume homeostasis in pregnancy-induced hypertension. J Clin
Endocrinol Metab 1992;74:150-6.
11. Will PC, Longworth JW, Brake ET, Cook JS. Analysis of intracel-
lular drug (ouabain) sequestration as a mechanism of detoxifi-
cation. Mol Pharmacol 1976;13:161-71.
12. Rayson BM. Rates of synthesis and degradation of Na+,K+-
ATPase during chronic ouabain treatment. Am J Physiol
1989;256:C75-C80.
13. Geering K. Subunit assembly and functional maturation of
Na,K-ATPase. J Membr Biol 1990;115:109-21.
14. Lingrel J, Orlowski J, Shull MM, Price EM. Molecular genetics of
Na,K-ATPase. In: Progress in nucleic acids research and molecu-
lar biology. New York: Academic Press; 1990. p. 37-89.
15. Herrera VLM, Chobanian AV, Ruiz-Opazo N. Isoform-specific
modulation of Na+,K+- ATPase -subunit gene expression in hy-
pertension. Science 1988;241:221-3.
16. National Institutes of Health Working Group on Hypertension
in Pregnancy. Classification of hypertensive disorders of preg-
nancy. Bethesda (MD): Dept. of Health and Human Services
(US); 1991.
17. Mendlowitz M. Toxemia of pregnancy and eclampsia. Obstet
Gynecol Surv 1980;35:327-31.
18. Sun GY, Huang HM, Kelleher JA, Stubbs EB, Sun AY. Marker en-
zymes, phospholipids and acyl group composition of a somal
plasma membrane fraction isolated from rat cerebral cortex: a
comparison with microsomes and synaptic plasma membranes.
Neurochem Int 1988;12:64-77.
19. Soszynski PA, Tao Q-F, Hollenberg NK, Graves SW. Tissue spe-
cific regulation of the sodium pump in DOCA-salt hypertension.
Am J Hypertens 1997;10:1132-9.
20. Urayama O, Shutt H, Sweadner KJ. Identification of three
isozyme proteins of the catalytic subunit of the Na,K-ATPase in
rat brain. J Biol Chem 1989;264:8271-80.
21. Tao Q-F, Hollenberg NK, Price DA, Graves SW. Sodium pump
isoform specificity for the digitalis-like factor isolated from
human peritoneal dialysate. Hypertension 1997;29:815-21.
22. Yamamoto K, Ikeda U, Seino Y, Tsuruya Y, Oguchi A, Okada K,
et al. Regulation of Na,K-adenosine triphosphatase gene expres-
sion by sodium ions in cultured neonatal rat cardiocytes. J Clin
Invest 1993;92:1889-95.
23. Sweadner KJ, Herrera VL, Amato S, Moellmann A, Gibbons DK,
Repke KRH. Immunologic identification of Na+,K(+)-ATPase
isoforms in myocardium. Isoform change in deoxycorticos-
terone acetate-salt hypertension. Circ Res 1994;74:669-78.
24. Ruiz-Opazo N, Xiang XH, Herrera VL. Pressure overload dein-
duction of human alpha 2 Na,K-ATPase gene expression in
transgenic rats. Hypertension 1997;29:606-12.
25. Handwerker SM, Altura BT, Altura BM. Ionized serum magne-
sium and potassium levels in pregnant women with preeclamp-
sia and eclampsia. J Reprod Med 1995;40:201-8.

Availability of Journal back issues

As a service to our subscribers, copies of back issues of the American Journal of Obstetrics
and Gynecology for the preceding 5 years are maintained and are available for purchase
from Mosby until inventory is depleted at a cost of $15.00 per issue. The following quantity
discounts are available: 25% off on quantities of 12 to 23 and one third off on quantities of
24 or more. Please write to Mosby, Inc, Subscription Services, 11830 Westline Industrial
Drive, St. Louis, MO 63146-3318, or call (800)453-4351 or (314)453-4351 for information on
availability of particular issues. If back issues are unavailable from the publisher, photocopies
of complete issues may be purchased from UMI, 300 N Zeeb Rd, Ann Arbor, MI 48106-1346.
Tel: (313)761-4700.