Human Decidua Is a Major Source of Renin
Kathryn J. Shaw, Yung S. Do, Siri Kjos, Pam W. Anderson, Tatsuo Shinagawa, Louis Dubeau, and Willa A. Hsueh
Departments ofMedicine, Obstetrics, and Gynecology, and Pathology, University ofSouthern California, School ofMedicine, and the
University ofSouthern California Comprehensive Cancer Center, Los Angeles, California 90033
Plasma prorenin levels are elevated in normal pregnant
women. Current evidence suggests renin production by tissues
of the uteroplacental unit contribute to this elevation. The pur-
pose of this investigation was to define the source of renin
biosynthesis within the human uteroplacental unit and to char-
acterize the renin produced. RNA extraction and Northern blot
analysis consistently demonstrated renin mRNA expression in
uterine lining both in the pregnant (decidua) and nonpregnant
states (endometrium) and in fetal chorion laeve, which is insep-
arable from the decidua. In contrast, renin mRNA expression
was not detected in basal plate and intertwin chorion (which is
separate from decidua), amnion, myometrium, or placental vil-
lae. The total renin content in decidual homogenates was two-
to threefold greater than in endometrial homogenates, and cul-
tured human decidual cells produced significantly more total
renin than cultured human endometrial cells, suggesting that
pregnancy enhanced renin production by the cells lining the
uterus. Immunoblot analysis and I3H]leucine incorporation
identified 47,000-mol wt prorenin as the major form of renin
produced by cultured human decidual cells. These studies indi-
cate that maternal decidua is the major source of prorenin in
the uteroplacental unit.
In normal human pregnancy circulating prorenin levels are
profoundly elevated compared with levels in the nonpregnant
state (1, 2). Early in pregnancy the ovary appears to be respon-
sible for the increase in plasma prorenin concentration (3, 4),
while the uterine-fetal-placental unit may contribute in part to
the prorenin rise in the third trimester (5). The exact source of
uteroplacental prorenin is unknown. Fetal chorion laeve has
been thought to be the major source (6). However, chorion
laeve abuts against maternal decidua, which forms the inner
lining of the uterus in pregnancy, so that decidua is also ob-
tained when chorion tissue or cells are isolated. Early studies
have suggested that decidua contains and secretes an enzyme
with reninlike activity that is enhanced after trypsin treatment
(7). Thus, studies ofchorion renin production have been com-
plicated by potential contributions from maternal decidua. In
addition, when renin activity of tissue homogenates and per-
Address correspondence to Dr. Willa A. Hsueh, University of South-
ern California, School ofMedicine, 2025 Zonal Avenue, Los Angeles,
Receivedforpublication I August 1988 and in revisedform 3 Jan-
fusates ofwhole organ cultures have been compared with renin
immunohistochemical studies, renin has variably been found
not only in chorion and decidua but also in neighboring tissues
such as uterine myometrium and amnion (7, summarized in
reference 8). Thus, the objective of the present investigation
was to define more precisely the source ofrenin biosynthesis in
the uterus, fetal membranes, and placenta, and to characterize
renin produced by human uterine lining. In particular, de-
cidua was sampled in the absence of chorion from ectopic
pregnancies, and chorion without decidua contamination was
sampled from the basal plate and from common chorion be-
tween opposing amniotic membranes in dichorionic-diamni-
otic twin pregnancies. Our results demonstrate that (a) both
endometrium and decidua contain and secrete renin; (b) the
major form of renin secreted by human decidual cells in cul-
ture is prorenin; and (c) decidua and endometrium express the
renin message while chorion not anatomically in contact with
decidua (i.e., basal plate chorion and intertwin chorion) dem-
onstrates no detectable expression of the renin gene. In addi-
tion, myometrium, placental villi, and amnion also demon-
strate no detectable renin gene expression. We conclude that
human decidua is a major source ofprorenin in the uteropla-
cental unit. In previous studies the demonstration of renin
production by cultured fetal chorion was probably the result of
contamination by maternal decidua. These results will be im-
portant in elucidating the role of local tissue renin production
in human pregnancy.
Measurements. Active renin concentration was determined by RIA of
angiotensin I (AI)' generated by incubation with excess semipurified
sheep angiotensinogen, final concentration 1 MM, at pH 7.4, 37°C for
10-30 min in the presence of 5 mM EDTA, 16 mM 1,2-dimercapto-
propane, and 3.4mM 8-hydroxyquinoline (9). In this assay, generation
of 1.2 X I05 ngAI/ml per h is equal to 1 Goldblatt unit (GU) ofrenin as
determined against Medical Research Council renal renin standard
(68-356) from the National Institute ofBiological Standards and Con-
trols, Holly Hill, London. Total renin was measured after acid dialysis
against 0.5 M glycine buffer, pH 3.3, at 4°C and titrated to pH 7.1 with
1.0 M phosphate buffer, pH 8.0. After this procedure plasma renin is
totally activated, but activation is reversible. Therefore, samples were
incubated with sheep angiotensinogen for five time periods between 0
and 0.5 h at pH 7.5, and the slope of AI concentration vs. time was
determined by linear regression analysis, with an r value of0.96 being
acceptable (10). Prorenin concentration is the difference between total
and active renin.
Tissue collection. Tissues were obtained after written informed
consent. Human endometrium was collected by curetting the uterus of
seven healthy premenopausal women undergoing abdominal hyster-
ectomy for uterine leiomyomata or vaginal hysterectomy for pelvic
relaxation. Their mean age was 36±4 yr. Pregnancy tissue was ob-
1. Abbreviations used in this paper: AI, angiotensin I; C-section, cesar-
ean section; CHO, Chinese hamster ovary; GU, Goldblatt unit.
Human Decidua Is a Major Source ofRenin
J. Clin. Invest.
© The American Society for Clinical Investigation, Inc.
Volume 83, June 1989, 2085-2092
tained from healthy pregnant women undergoing cesarean section (C-
section) for dysfunctional labor or fetal distress at 36-42 wk gestation.
Pregnancy tissue was also obtained from two women who required
cesarean hysterectomy for postpartum hemorrhage. The mean age of
the pregnant women was 32±2 yr. The different tissues collected are
depicted in Fig. 1 A. Decidua, placenta, and fetal membranes were
obtained from a total of 12 pregnant women with a single fetus and 6
pregnant women with twin fetuses. Human decidua was collected by
curettage of the uterine lining after removal of the placenta and fetal
membrane fragments after delivery ofthe placenta: (a) chorion laeve,
lining the uterus, was separated from amnion by careful peeling; (b)
basal plate chorion was peeled as a visibly separate layer from the fetal
side ofthe placental surface with care taken to avoid blood vessels; and
(c) in the case of twin pregnancies, intertwin chorion was carefully
peeled away from the two layers ofamnion between each fetus. Basal
plate chorion and intertwin chorion were used to obtain samples of
chorion not in contact with decidua. Histologic examination demon-
strated that basal plate chorion and intertwin chorion contained viable
trophoblastic cells and no maternal decidua. In contrast, chorion laeve
had large amounts of decidua attached to the membrane (Fig. 1 B).
Placental villi were obtained as placental tissue separate from the fetal
membranes. Myometrium was dissected from the uterine fundus of
patients requiring cesarean hysterectomy. It represented the middle
portion ofthe uterine wall and was carefully excised to avoid decidual
and serosal contamination. Human decidua was also obtained from
patients with ectopic pregnancy at 6-10 wk gestation during curettage
of the uterus. Renal cortex was obtained at the time of surgery from
normotensive patients requiring nephrectomy for renal tumors; amac-
roscopically normal area of kidney as determined by pathological ex-
amination was used. Tissues were either used immediately or immedi-
ately frozen at -700C.
Homogenization. I g of tissue was homogenized in 2 vol buffer (5
mM Na phosphate, pH 7.1, containing 10mM EDTA, 2 mM PMSF,
8-hydroxyquinoline, aprotinin, 20,000 kallikrein inhibitor units/l and
20 mM benzamidine) with a Tissumizer (Tekmar Co., Cincinnati,
OH) at 4°C as previously reported for human kidney tissue (9). After
centrifugation of the homogenate at 10,000 rpm for 30 min, the su-
pernate was assayed for renin. Tissue homogenate renin levels were
expressed as nanograms Al generated/hour per gram wet weight of
tissue. Renin in endometrium vs. decidua was compared by Wilcoxon
rank sum analysis using the University ofSouthern California Clinical
Research Center CLINFO system. Studies comparing tissue renin
from C-section vs. twin pregnancy were performed using the paired t
RNA extraction and hybridization. Tissues obtained at the time of
surgery were immediately placed on dry ice and stored in 4 M guani-
dinium isothiocyanate. RNA was isolated using two techniques. One
was a modification of the technique ofChirgwin et al. (11). All solu-
tions were treated with 0.1% diethylpyrocarbonate to remove RNAase.
Tissue washomogenized in 9 vol ofguanidinium isothiocyanate (4 M),
saturated with cesium chloride (CsCl; 1 g/2.5 ml), and layered onto 5.7
M CsCl. Ultracentrifugation was carried out at 30,000 rpm at 20°C for
16 h in an SW40 rotor (Beckman Instruments, Inc., Palo Alto, CA).
The pellet oftotal RNA that sedimented to the bottom ofthe tube was
resuspended in Tris buffer, extracted with phenol and chloroform-
isoamyl alcohol, and precipitated in ethanol. A second method of
RNA extraction was that ofChomczynski et al. (12). 1 g oftissue was
homogenized in 10 ml ofa denaturing solution containing 4 M guani-
dinium isothiocyanate, 25 mM sodium citrate, pH 7, 0.5% sarkosyl,
and 0.1 M 2-mercaptoethanol. This tissue solution was then extracted
twice with 2 M sodium acetate, pH 4.0, phenol, and chloroform-iso-
amyl alcohol. RNA present in the aqueous phase resulting from cen-
trifugation was precipitated with isopropanol. Total RNA was quanti-
tated by ultraviolet absorbance measurements at 260 nm using a pho-
tometer (Roy V Spectronic 1201; Milton Roy Spectronic, Rochester,
NY). RNA was stored in ethanol at -70°C.
Northern blot analysis. Total cellular RNA (20 Ag/lane)was elec-
trophoresed on 1.0% agaroseunderdenaturingconditionsaccordingto
the method ofLehrach et al. (13) and transferred to nylon membranes
(MSI Magnagraph; Fisher Scientific Co., Pittsburgh, PA). RNA was
crosslinked to the membrane by ultraviolet irradiation and hybridized
to specific probes using the procedure of Church and Gilbert (14).
Hybridization was carried out for 72 h at 420C. After hybridization,
membranes underwent three 15-min washes. The first was done at
room temperature in 2X standard saline citrate (0.03 M sodium chlo-
ride, 0.003 M sodium citrate), 1% SDS. The second was performed at
room temperature in 0.2X standard saline citrate, 1.0% SDS. The final
wash was done at 550C in the latter solution. Membranes were then
subjected to autoradiography for 24-48 h at -70'C.
Source and labeling ofDNA probes. A 600-bp cDNA probe ofthe
carboxy terminus (3') of intact human renin cDNA was kindly pro-
vided by Dr. Tim Reudelheuber (University of California, San Fran-
cisco). Mouse a-skeletal actin cDNA (1,000 bp) was obtained from Dr.
Amy Lee (University of Southern California Comprehensive Cancer
Center). Probes were labeled with a-[32P]d CTP (ICN Radiochemicals,
Irvine, CA) to specific activities >108/1gby random priming as de-
scribed by Feinberg and Vogelstein (15).
Culture. Decidua orendometrium was digested in 0.1% collagenase
and 0.2% hyaluronidase in DME (Sigma Chemical Co., St. Louis, MO)
for 3 h at 370C. Suspensions were centrifuged (600 g, 280C) for 5 min
and cells were then dispersed in 0.9% NH4CI for 15 min at 370C to
cause hemolysis. Suspensions were again centrifuged and cells were
washed twice in PBS. Cells were resuspended (DME + 10% newborn
calf serum; Irvine Scientific, Santa Ana, CA), plated (105 cells/ml) in
35-mm wells, and incubated at 37°C in 10% C02, 90% air. Media were
changed every 2-3 d and collected at each change for renin assay. Cells
were then subjected to trypsin digestion and counted (CoulterCounter,
model ZBI analyzer; Coulter Electronics Inc., Hialeah, FL). Trypan
blue exclusion was used as an index ofthe viability ofcells in culture.
Immunostaining ofthe cultured cells using the avidin-biotin complex
technique (16) with antibodies against vimentin, keratin, and Factor
VIII was used to assess whether the cells were predominantly stromal,
epithelial, or endothelial, respectively.
Renin characterization. Immunoblot analysis was performed as
described (17). Renin from decidual culture media was compared with
pure human renal renin (17) and recombinant prorenin (18). Recom-
binant prorenin was obtained from the culture media ofChinese ham-
ster ovary (CHO) cells transfected with the human renin gene and
partially purified by ion exchange and Cibacron blue chromatogra-
phies (19). Rabbit antiserum developed against pure human renal
renin (17) was used to detect immunoreactive forms of renin after
Western blotting. Antiserum developed against the carboxy-terminal
12 amino acids of the prosegment of human prorenin was used to
detect forms of renin with this propeptide attached (18, 20). Renin
forms were immunoprecipitated after incubation of decidual culture
media (10-3 GU) and recombinant prorenin (10-3 GU) with varying
dilutions of the prosegment antibody or the human renal renin anti-
body for 24 h, pH 7.5, at 4°C. Antigen-antibody complexes were
precipitated with protein A Sepharose 4B (Pharmacia Fine Chemicals,
Piscataway, NJ). Samples were centrifuged and the supernate assayed
for active and total renin.
For [3H]leucine incorporation, cells in culture were incubated for 2
h in medium lacking leucine and then for 6 h in fresh medium con-
taining [3H]leucine (I mCi/ml; ICN Biomedicals, Inc. (Costa Mesa,
CA). Radiolabeled renin was isolatedby immunoprecipitation and run
on SDS gels that were subjected to autoradiography.
Renin in human decidua and endometrium
Tissue homogenate. The supernate of both decidual and en-
dometrial homogenates contained active and prorenin. How-
ever, the concentrations of active and total renin were higher
in decidua than in endometrium when tissues were processed
identically and measured in the same renin assay (Fig. 2). The
Shaw, Do, Kjos, Anderson, Shinagawa, Dubeau, and Hsueh
if,@&~~~~' ai F
Figure 1. (A) Human uteroplacental unit. Chorion laeve (chorion) lies in direct contact with maternal decidua, whereas the basal chorion plate
is separated from decidua by the intervening placental cotyledons (viifi). (From Pritchard, J. A., P. C. McDonald, and N. F. Gant. 1985. Wil-
liams Obstetrics. Appleton-Century-Crofts, Norwalk, CT. 65-66, 98-108, with permission.) (B) Hematoxylin and eosin staining of human
chorion laeve. Note the large amount of maternal decidua attached to chorion.
Human Decidua Is a Major Source ofRenin 2087
Figure 2. Total and ac-
tive renin in decidual
and endometrial ho-
mogenates. Results are
expressed as nanograms
AI/hour per gram wet
tissue weight. Total
renin was higher in de-
cidua (1,692±423) com-
pared with endome-
trium (544±85; P
< 0.01). Active renin
content was similarly
greater in decidua
(58±17; P < 0.01).
mean total renin concentration was 1,692±423 ng AI/h per g
(wet weight tissue) for decidua from five normal pregnant
women and 544±85 ng AI/h per g for endometrium from six
premenopausal women (P < 0.01); active renin was 648 and
53 ng AI/ml per h per g (P < 0.01), respectively, and prorenin
was 1,052±313 and 486±88 ng Al/ml per h per g (P = NS),
respectively. In decidua, active renin contributed to
the total renin, while in endometrium it contributed to 10% of
Renin concentration in culture media. Decidua was cul-
tured from four normal pregnant women requiring C-section
and three women with ectopic pregnancies. Endometrium was
cultured from four nonpregnant women. Cultures ofendome-
trium and decidua had a similar microscopic appearance. The
majority of cells were elongated and spindly; these immuno-
stained positively for vimentin. There were occasional clusters
of crescent-shaped vacuolated cells, which immunostained
positively for keratin. No cells had positive staining for Factor
VIII. Thus, stroma appeared to represent the major cell type in
the decidua and endometrial primary cultures, although areas
(<10% ofthe total cells) ofepithelial cells were present. From
4 d to up to 3 wk ofculture the major form ofrenin produced
by decidua or endometrium was inactive prorenin. The pat-
tern of renin production is illustrated in Fig. 3. Total renin
secreted into the media by decidua or endometrium peaked at
-2 wk of culture. When expressed as total renin/104 cells,
cultured decidua from both intrauterine and ectopic pregnan-
cies consistently produced higher levels oftotal renin than did
endometrium. Active renin constituted
renin in the culture media ofboth decidua and endometrium.
The time course of active renin production paralleled that of
prorenin. After 2 wk of culture, cell number continued to
increase and cell morphology remained the same. Thus, nei-
ther of these factors contributed to the declining media renin
levels after this time.
Characterization of renin produced by cultured decidua.
Immunoblot analysis demonstrated that renin in decidual
media crossreacted with the anti-human active renal renin
antibody (Fig. 4) and had a mol wt of 47,000. The molecular
weight was the same as that determined for recombinant pro-
renin (18). It was larger than that of pure active human renal
renin which has a mol wt of 44,000 and also has 22,000 and
18,000 mol wt subunits (17).
The antibody directed against the prosegment ofrenin sim-
ilarly precipitated renin in medium conditioned by either de-
- 10% of the total
DAYS IN CULTURE
12 1620 24
Figure 3. Total renin
production by cultured
human decidua cells
from term pregnancy (n
= 4) (o), from ectopic
pregnancy (n = 3) (.),
and by cultured human
endometrial cells (n
= 4) (- - -). Each sym-
bol represents the mean
and SEM ofthree wells
from each patient as-
sayed in duplicate.
Total renin levels are
similar in the media of
decidua from term and
ectopic pregnancy, but
lower in media from
cidua or CHO cells (Fig. 5). 50% ofthe renin was precipitated
renin from decidua was not precipitated by this antibody.
Renin from decidua or CHO cells also precipitated the anti-
human renal renin antibody. 50% inhibition occurred with a
- 1/2,000 dilution of this antibody.
Immunoprecipitation of radiolabeled renin with anti-
human renal renin antibody demonstrated that cultured de-
cidual cells produced a single protein that crossreacted with the
anti-human renin antibody. It had a mol wt of 47,000, in
agreement with data obtained by immunoblot analysis of un-
labeled culture media.
- 1/500 dilution of the prosegment antibody. Active
Localization ofrenin production in the uteroplacental unit
Renin concentration in decidua vs. chorion. Total and active
renin concentrations were measured in tissue homogenates of
full-term decidua, chorion laeve, and basal plate chorion ob-
tained from five women undergoing C-section (Fig. 6). In each
patient the decidua and chorion laeve homogenates had simi-
lar levels of total renin, while the homogenate of basal plate
chorion had significantly lower levels oftotal renin (P< 0.05,
paired t test). The total renin concentration in the decidua
homogenates in this group of patients was not significantly
different from the group sampled in Fig. 1. In five other pa-
7 5 , 0 00 -
Figure 4. Immunoblot
analysis ofhuman de-
cidua renin. Lane 1 is
pure human renal renin
with a mol wt of
44,000, with 22,000,
and 18,000 mol wt sub-
units. Lane 2 is culture
media containing re-
with a mol wt of
47,000. Lane 3 is media
of human decidua cells
in culture. Lane 4 is pu-
r.fied recombinant pro-
Shaw, Do, Kjos, Anderson, Shinagawa, Dubeau, and Hsueh
Figure 5. Inhibition of
decidual (.) and recom-
Both decidual and re-
produced in culture
were precipitated simi-
larly by the prosegment
antibody (50% at a
1:500 antibody dilu-
tion), whereas active
renin from decidual
0 vOcultures (o) was unaf-
tients with twin pregnancies similar levels of total renin were
found in chorion laeve homogenates (Fig. 6). However, in
these patients intertwin chorion levels of total renin were sig-
nificantly lower (P < 0.01, paired t test) than in chorion laeve.
Active renin concentration in these tissues ranged from 30 to
40% of the total renin concentration in decidua and chorion
laeve and 10 to 20% of the total renin in basal plate chorion
and intertwin chorion. Histologic evaluation of all of these
tissues demonstrated decidua was present in chorion laeve but
not in basal plate chorion or intertwin chorion.
Northern blot analysis ofuteroplacental tissues. Total RNA
was extracted from decidua, chorion laeve, intertwin chorion,
basal plate chorion, amnion, myometrium, and placental villi
obtained from normal pregnant women. In addition, uterine
lining without chorion was obtained from patients with ec-
topic pregnancies and from nonpregnant women. Total RNA
extracted from human kidney cortex was used as a positive
control. The total RNA yield ranged from 100 to 400 gg/g
tissue (wet weight). The A260/A280 were similar in the various
tissues, ranging from 1.7 to 1.9. Representative blots are
shown in Fig. 7. Positive renin mRNA was defined as a single
band at 1.5 kb, which is the known size of human renin
mRNA (21-24). Some RNA degradation is evidenced by tail-
ing after the band but this did not interfere with the interpre-
Figure 6. Total renin
levels in decidua, cho-
rion laeve (chorion-de-
cidua), and basal plate
All three samples were
obtained from the same
patients. Basal plate
chorion had signifi-
cantly lower total renin
n rion laeve (chorion-de-
from a separate group
chorion total renin
levels were significantly
lower than chorion
patients with twin
C..arean sectln (n-5)
Twhi pronancy (n-5)
tation of the results or constitute a positive result. Renin
mRNA expression was detected in kidney, decidua, and cho-
rion laeve from normal pregnant women, in decidua from
women with ectopic pregnancy, and in endometrium. In con-
trast, renin mRNA expression was not detected in basal plate
chorion or intertwin chorion. In addition, amnion, myome-
trium, and placental villi did not demonstrate detectable renin
mRNA. Reprobing the same membrane with actin cDNA
showed that the actin message was present in all lanes in ap-
proximately equal quantities (results not shown). Table I sum-
marizes our results from tissues sampled from different pa-
Investigation of gene expression is a powerful tool to identify
tissue biosynthesis of proteins and to study their regulation.
Detection of renin gene expression in the present study ulti-
mately identified human decidua as a major source ofrenin in
the uteroplacental unit. Previous studies have suggested a
seemingly ubiquitous production ofrenin by uterine and pla-
cental tissues, leading to confusion regarding the source and
physiologic role oftissue renin in pregnancy (6, 8, 25, 26). We
examined renin gene expression in various tissues in the
uteroplacental unit to address this issue. Particular attention
was given to obtaining decidua without chorion (i.e., decidua
from ectopic pregnancies) and conversely, chorion without
decidua (i.e., intertwin and basal plate chorion ). Addi-
tionally, endometrium was studied since it represented the
uterine lining in the nonpregnant state. In this study we dem-
onstrated consistent renin gene expression in decidua from
both intrauterine and ectopic pregnancies and in endome-
trium. Chorion laeve, shown histologically to contain decidua,
also demonstrated renin gene expression. In contrast, renin
mRNA was nondetectable in basal plate and intertwin chor-
ion, as well as in amnion, placental villi, and myometrium.
These findings indicate that decidua is a primary source of
renin. This conclusion is supported by measurement oftissue
renin content, demonstrating decidual content to be signifi-
cantly greater than that of basal plate and intertwin chorion.
This is further confirmed by demonstration of renin produc-
tion by cultured decidual cells from both intrauterine and ec-
This conclusion conflicts, in part, with previous reports;
yet, several explanations may be offered. Tissue culture studies
have described chorion laeve as a primary source ofrenin (6, 7,
26). One report appropriately termed this tissue "choriode-
cidua," having noted the significant amount of decidua at-
tached to chorion (7). Once chorion laeve is cultured, either
chorionic, decidual, or both cell types could be responsible for
renin produced. Since cultured decidua not in contact with
chorion (i.e., from ectopic pregnancies) produces renin in
quantities similar to cultured decidua from normal intrauter-
ine pregnancies, decidual production of renin is independent
of contact with the chorion in culture. Intertwin and basal
plate chorion, which do not touch decidua, grow very poorly
in culture (Shaw, K., Y. S. Do, and W. Hsueh, unpublished
observations); furthermore, renin mRNA was not detected in
either of these chorions. Taken together, these results suggest
that renin produced by cultures ofchorion laeve may be due to
decidual contamination. Immunohistochemistry using a spe-
cific anti-renin antibody demonstrated positive staining in the
Human Decidua Is a Major Source ofRenin
Figure 7. Human renin
mRNA expression in
Total RNA (20 ug/lane)
wasseparated on a I%
nylon membranes, and
then hybridized with a
renin cDNA probe. 18S
and 28S refer to ribo-
somal RNA species. (A)
metrium (lane 2), pla-
cental villi (lane 3),
myometrium (lane 4),
amnion (lane 5), kidney
(lane 6), dedidua (lane
7), myometrium (lane
8), villi (lane 9), cho-
laeve (lane iO),
basal plate chorion
(lane 11), and ectopic
.f t .*
*; i fl w w :
RNA from endo-
3 4 5
6 7 8 9 10 1112
decidua (lane 12). Lane 1 is a blank. (B) Total RNA was extracted from tissues obtained from a twin gestation. Decidua (lane 1), chorion laeve
(lane 2), intertwin chorion (lane 3), basal plate chorion (lane 4), placental villi (lane 5), and amnion (lane 6).
cytotrophoblast layer ofthe chorion (25). Indeed, in our study
we detected reninlike activity in homogenates of basal plate
and intertwin chorion. These results do not necessarily indi-
cate production ofrenin by these cells, but may reflect uptake
ofrenin by the chorion. Amniotic fluid renin concentration is
10 times the plasma concentration (1, 28), and although con-
tribution from the fetus may be important, amniotic fluid
renin may be of decidual origin. Thus, transport across both
chorion and amnion would be necessary. An analogous situa-
tion has been described for amniotic fluid prolactin that arises
from maternal decidua (29, 30). The presence of renin in
chorion without detectable renin mRNA expression supports
the possibility ofrenin uptake by this tissue.
Table L Tissue mRNA Expression
Tissue renin gene expression: renin mRNA was consistently detected
in uterine lining both in the nonpregnant (endometrium) and preg-
nant (decidua) state, as well as in chorion laeve. In other forms of
chorion (basal plate and intertwin) renin mRNA was nondetectable,
as was the case for amnion, placenta villi, and myometrium. Actin
mRNA was detected in all tissues.
Myometrium has also been described as a source ofrenin,
particularly in the rabbit (31, 32). However, immunohisto-
chemical studies ofhuman myometrium demonstrated renin
staining only in small clusters ofcellsaround blood vessels and
not in smooth muscle cells (33). Human myometrial cells in
culture were shown to produce renin, but only after 10-14 d of
culture. This suggests that renin-producing cells constitute a
small fraction of the initial cells cultured so that renin was
detected only after these cells multiplied. Alternatively, as sug-
gested by one author (26), renin production by cultured myo-
metrial cells may be repressed in vivo and may represent an in
vitro phenomenon. These data are consistent with our finding
ofno detectable renin mRNA in human myometrium.
Tritiated leucine incorporation in cultured decidua con-
firmed translation ofrenin mRNA. Immunoprecipitation and
gel electrophoresis of the 3H-labeled proteins demonstrated a
single renin band with a mol wt of 47,000, which was con-
firmed by immunoblot analysis. Precipitation with an anti-
body generated against the carboxy-terminal third ofthe pro-
segment demonstrated that at least this portion ofthe proseg-
ment was present on renin produced by cultured decidual
cells. Thus, prorenin is the major form of renin produced by
cultured human decidual cells. This system may be an effec-
tive tool in which to study renin production and processing in
vitro, since isolated cultured human juxtaglomerular cells are
not yet available. In fact, in cultured human cells found to
produce renin, human decidual cells secrete about 6-fold less
renin than CHO cells transfected with the human renin gene
(19), but secrete 100-fold more renin than cultured human
ovarian theca cells (34) and 5-fold more renin than primary
cultures ofan ovarian human renin-secreting tumor (35).
Endometrium and decidua both produce renin, yet our
results suggest that pregnancy enhances renin production. The
total renin in decidual homogenates was threefold greater than
in similarly processed endometrial homogenates. When grown
in culture under identical conditions, decidua produced three
to four times the amount of prorenin per cell compared with
Shaw, Do, Kjos, Anderson, Shinagawa, Dubeau, andHsueh
endometrium. In pregnancy, the uterine lining hypertrophies
and the cellular histology changes (36). After this decidualiza-
tion process, hormonal or local hemodynamic alterations
could enhance endometrial renin production either by in-
creasing renin transcription or by enhancing the number of
renin-producing cells. Further studies of the regulation of
renin production by cultured decidual and endometrial cells
will be important in assessing the role ofpregnancy hormones
on uterine renin production.
An additional observation is the discrepancy between the
percent of the total renin that is active in tissue homogenates
vs. cell culture. In our study active renin comprised 38% ofthe
total renin in homogenized decidual tissue, compared with
10% for homogenized endometrium, while in culture the ac-
tive percentage of both cell types was 10%. Thawing and ho-
mogenization of tissue could conceivably inadvertently acti-
vate prorenin. However, a cocktail of protease inhibitors and
rapid processing was used to prevent activation. Furthermore,
decidua and endometrium were handled identically. Whether
discrepancies between the percent active renin in decidual or
endometrial homogenates are real or artifact remains to be
determined. Similar findings were reported in a renin-secreting
renal tumor (37) in which tumor tissue homogenates con-
tained 72% active renin, while in cultured tumor cells the per-
cent active renin decreased to as low as 3% after 2 wk of
culture. Hence, decidua may produce relatively larger quanti-
ties ofactive renin in vivo than when cultured in vitro.
The role of renin in human decidua and endometrium is
unknown. Several potential roles exist, most ofwhich are me-
diated by generation ofangiotensin II. Thus, it will be impor-
tant to determine (a) if the large quantities of prorenin pro-
duced by decidua are activated and (b) whether a complete
renin-angiotensin system exists locally in the decidua and en-
dometrium. Pure recombinant prorenin exists in equilibrium
in an inactive and active form (38), so ifrelatively large quan-
tities ofprorenin were produced locally, a significant amount
ofactive renin could be available to generate AI. Furthermore,
a variety ofenzymes can activate prorenin (39) so that activa-
tion could potentially occur extracellularly by enzymes re-
leased from tissues adjacent to decidua. Angiotensin II has
been implicated in steroidogenesis, angiogenesis (40), growth
(41), smooth muscle (vascular and uterine) contraction (42),
and regulation ofamniotic fluid volume and electrolytes (43).
Further studies are needed to evaluate these potential roles and
should include evaluation of the renin-angiotensin system in
various pathologic states ofpregnancy as well as investigations
into factors regulating decidual production ofrenin.
This work was supported by grants from the American Heart Associa-
tion, Florida Affiliate, and the National Institutes of Health
(AM-30254). Dr. Hsueh has an NIH Research Center Development
Award (AM-01035). Dr. Shaw has an NIH Fellowship (HL-07523).
Dr. Anderson is a Research Fellow ofthe American Heart Association,
Greater Los Angeles affiliate.
1. Hsueh, W. A., J. A. Luetscher, E. J. Carlson, G. Grislis, E. Fraze,
and A. McHargue. 1982. Changes in active and inactive renin
throughout pregnancy. J. Clin. Endocrinol. Metab. 54:1010-1015.
2. Sealey, J. E,, M. Wilson, A. A. Morganti, I. Zervoudakis, and
J. H. Laragh. 1982. Changes in active and inactive renin throughout
normal pregnancy. Clin. Exp. Hypertens. A4:2373-2384.
3. Itskovitz, J., and J. E. Sealey. 1987. Ovarian prorenin-renin-an-
giotensin system. Obstet. Gynecol. Surv. 42:545-551.
4. Derkx, F. H. M., A. T. Alberda, F. H. DeJong, F. H. Zeilmaker,
J. W. Makovitz, and M. A. D. H. Schalekamp. 1987. Source ofplasma
prorenin in early and late pregnancy: observations in a patient with
primary ovarian failure. J. Clin. Endocrinol. Metab. 65:349-353.
5. Brar, H. S., Y. S. Do, H. B. Tam, G. J. Valenzuela, R. D. Murray,
L. D. Longo, M. L. Yonekura, and W. A. Hsueh. 1986. Uteroplacental
unit as a source of elevated circulating prorenin levels in normal preg-
nancy. Am. J. Obstet. Gynecol. 155:1223-1226.
6. Acker, G. M., F. X. Galen, C. DeVaux, S. Foote, E. Papernik, A.
Pesty, J. Menard, and P. Corvol. 1982. Human chorionic cells in
primary culture: a model for renin biosynthesis. J. Clin. Endocrinol.
7. Warren, A. Y., D. J. Craven, and E. M. Symonds. 1982. Produc-
tion of active and inactive renin by cultured explants from human
female genital tract. Br. J. Obstet. Gynaecol. 628-632.
8. Symonds, E. M. 1988. Renin and reproduction. Am. J. Obstet.
9. Hsueh, W. A., E. J. Carlson, and V. Dzau. 1983. Characteriza-
tion' ofinactive renin from human kidney and plasma. Evidence for a
renal source of circulating inactive renin. J. Clin. Invest. 71:506-517.
10. Hsueh, W. A., E. J. Carlson, and M. Israel-Hagman. 1981.
Mechanism ofacid-activation ofrenin: role ofkallikrein in renin acti-
vation. Hypertension. 3(Suppl. I):I-22, I-29.
11. Chirgwin, J. M., A. E. Pryzbyla, R. J. MacDonald, and W. J.
Rutter. 1979. Isolation of biologically active ribonucleic acid from
sources enriched in ribonuclease. Biochemistry. 5294-5299.
12. Chomczynski, P., and S. Nicoletta. 1987. Single-step method of
RNH isolation by acid quanidinium thiocyanate-phenol-chloroform
extraction. Anal. Biochem. 162:156-159.
13. Lehrach, M., D. Diamond, J. M. Wozmey, and H. Boedtker.
1977. RNA molecular weight determination by gel electrophoresis
under denaturing conditions: a critical re-examination. J. Biochem.
14. Church, G. M., and W. Gilbert. 1984. Genomic sequencing.
Proc. Nat!. Acad. Sci. USA. 81:1991-1995.
15. Feinberg, A. P., and B. Vogelstein. 1983. A technique for ra-
diolabeling DNA restriction endonuclease fragments to high specific
activity. Anal. Biochem. 132:6-13.
16. Hsu, S. M., L. Raine, and H. Fanger. 1981. Use ofavidin-biotin
complex (ABC) in immunoperoxidase techniques. J. Histochem. Cy-
17. Do, Y. S., T. Shinagawa, H. Tam, and W. Hsueh. 1987. Charac-
terization of pure human renin: evidence for the subunit structure of
human renal renin. J. Biol. Chem. 262:1037-1043.
18. Hsueh, W., Y. Do, T. Shinagawa, H. Tam, P. Ponte, and L.
Fritz. Biochemical similarity ofexpressed human prorenin and native
inactive renin. Hypertension. 8(Suppl. II):II-78-II-83.
19. Fritz, L. C., A. E. Arfsten, V. J. Dzau, S. A. Atlas, J. D. Baxter,
J. C. Fiddes, J. Shine, C. L. Cofer, P. Kushner, and P. A. Ponta. 1986.
Characterization of human prorenin expressed in mammalian cells
from cloned cDNA. Proc. Natl. Acad. Sci. USA. 83:4114-4118.
20. Atlas, S. A., P. Christofalo, T. Hesson, J. E. Sealey, and L. C.
Fritz. 1985. Immunological evidence that inactive renin is prorenin.
Biochem. Biophys. Res. Commun. 132:1038-1045.
21. Imai, T., H. Miyazaki, S. Hirose, H. Hori, T. Hayashi, R.
Kageyama, H. Ohkubo, S. Nakenishi, and K. Murakemi. 1983. Clon-
ing and sequence analysis ofcDNA for human renin precursor. Proc.
Nat!. Acad. Sci. USA. 80:7405-7409.
22. Hardman, J. A., Y. J. Hort, D. F. Catanzaro, J. T. Tellam, J. D.
Baxter, B. J. Morris, and J. Shine. 1984. Primary structure of the
human renin gene. DNA (NY). 3:457-468.
23. Hobart, P. M., M. Fogliano, B. A. O'Connor,I. M. Schaefer,
Human Decidua Is a Major Source ofRenin
and J. M. Chirgwin. 1984. Human renin gene: structure and sequence Download full-text
analysis. Proc. Nail. Acad. Sci. USA. 81:5026-5030.
24. Soubrier, F., J. J. Panthier, P. Corvol, and F. Rougeon. 1983.
Molecular cloning and nucleotide sequence of a renin cDNA frag-
ment. Nucleic Acids Res. 11:7181-7190.
25. Poisner, A. M., G. W. Wood, R. Poisner, and T. Inagami. 1981.
Localization of renin in trophoblasts in human chorion laeve at term
pregnancy. Endocrinology. 109:1150-1155.
26. Symonds, E. M., M. A. Stanley, and S. L. Skinner. 1968. Pro-
duction of renin by in vitro cultures of human chorion and uterine
muscle. Nature (Lond.). 217:1152-1154.
27. Lopez-Bernal, A., A. B. M. Anderson, D. M. Parry, and A. C.
Turnbull. 1980. Evidence that fetal membranes are not involved in
cortisol metabolism: study of dichorionic twin pregnancies. Am. J.
Obstet. Gynecol. 138:1168-1172.
28. Lumbers, E. R. 1971. Activation ofrenin in amniotic fluid by
low pH. Enzymologia. 40:329-333.
29. McCoshey, J. A., 0. Y. Tagger, A. Wodzicki, and J. E. Tyson.
1982. Choriodecidual adhesion promotes decidual prolactin transport
by human fetal membrane. Am. J. Physiol. 243:R552-R557.
30. Riddick, D. H., and I. A. Moslar. 1981. The transport of pro-
lactin by human fetal membranes. J. Clin. Endocrinol. Metab.
31. Ferris, T. F., P. Gorden, and P. J. Mulrow. Rabbit uterus as a
source ofrenin. Am. J. Physiol. 212:698-702.
32. Dzau, V. J., D. Gonzalez, K. Ellison, S. Churchill, and N.
Emmett. 1987. Characterization ofpurified rabbit uterine renin: influ-
ence ofpregnancy on uterine inactive renin. Endocrinology. 120:358-
33. Johnson, J., I. R. Johnson, and J. E. Ronan. 1984; The site of
renin in the human uterus. Histopathology (04). 8:273-278.
34. Do, Y. S., A. Sherrod, R. A. Lobo, R. J. Paulson, T. Shinagawa,
S. Chen, S. Kjos, and W. A. Hsueh. 1988. Human ovarian theca cells
are a source ofrenin. Proc. Natl. Acad. Sci. USA. 85:1957-1961.
35. Anderson, P. W., L. Macaulay, M. Koss, Y. S. Do, T. Shina-
gawa, and W. A. Hsueh. 1988. Morphology of renin processing by a
human renin-secreting tumor. Clin. Res. 36:423. (Abstr.)
36. Pritchard, J. A., P. C. McDonald, and N. F. Gant. 1985. Wil-
liams Obstetrics. Appleton-Century-Crofts, Norwalk, CT. 65-66,
37. Galen, F. X., M. Corvol, and C. Devaux. 1984. Renin biosyn-
thesis by human tumoral juxtaglomerular cells. Evidence for a renin
precursor. J. Clin. Invest. 73:1144-1155.
38. Heinrikson, R. L. 1988. The prorenin-renin interaction. Am.
Soc. Hyperten. Symp. Ser. 18.
39. Sealey, J. E., S. A. Atlas, and J. H. Laragh. 1980. Prorenin and
other large molecular weight forms of renin. Endocrinol. Rev. 1:265-
40. Fernandez, L. A., J. Twickler, and A. Mead. 1985. Neovascu-
larization produced by angiotensin II. J. Lab. Clin. Med. 105:141-145.
41. Normal, J., B. Badie-Dezfooly, E. P. Nord, I. Kurtz, J.
Schlosser, A. Chaudhari, and L. G. Fine. 1987. EGF-induced mito-
genesis in proximal tubular cells: potentiation by angiotensin II. Am. J.
42. Re, R. N. 1984. Cellular biology of the renin-angiotensin sys-
tems. Arch. Intern. Med. 144:2037-2041.
43. Cooke, S. F., D. J. Craven, and E. M. Symonds. 1985. Osmo-
larity changes can enhance the release ofrenin from human chorion.
Am. J. Obstet. Gynecol. 151:819-821.
Shaw, Do, Kjos, Anderson, Shinagawa, Dubeau, and Hsueh