REPRODUCTION-DEVELOPMENT

The Orphan Nuclear Receptor Nr5a2 Is Essential for

Luteinization in the Female Mouse Ovary

Kalyne Bertolin, Jan Gossen, Kristina Schoonjans, and Bruce D. Murphy

Centre de Recherche en Reproduction Animale (K.B., B.D.M.), Faculté de Médecine Vétérinaire,
Université de Montréal, Saint Hyacinthe, Québec, Canada J2S 7C6; Gossen Consultancy (J.G.), 5345 BA
Oss, The Netherlands; and Laboratory of Integrative and Systems Physiology (K.S.), École Polytechnique
Fédérale de Lausanne, CH-1015 Lausanne, Switzerland

In the ovary, the follicular granulosa cells express the nuclear receptor Nr5a2 (nuclear receptor
subfamily 5 group A member 2), also known as liver receptor homolog-1, and after ovulation, Nr5a2
expression persists in the corpus luteum. Previous studies demonstrated that Nr5a2 is required for
both ovulation and luteal steroid synthesis. Our objectives were to analyze the temporal sequence
in the regulatory effects of Nr5a2 in the ovary, with focus on its contribution to luteal function. We
developed a female mouse model of granulosa-specific targeted disruption from the formation of
the antral follicles forward (genotype Nr5a2Cyp19⫺/⫺). Mice lacking Nr5a2 in granulosa cells of
antral follicles are infertile. Although their cumulus cells undergo expansion after gonadotropin
stimulation, ovulation is disrupted in those mice, at least in part, due to the down-regulation of the
progesterone receptor (Pgr) gene. The depletion of Nr5a2 in antral follicles permits formation of
luteal-like structures but not functional corpora lutea, as evidenced by reduced progesterone levels
and failure to support pseudopregnancy. Progesterone synthesis is affected by depletion of Nr5a2
due to, among others, defects in the transport of cholesterol, evidenced by down-regulation of
Scarb1, Ldlr, and Star. Comparison of this mouse line with the models in which Nr5a2 is depleted
from the primary follicle forward (genotype Nr5a2Amhr2⫺/⫺) and after the ovulatory signal (ge-
notype Nr5a2Pgr⫺/⫺) demonstrates that Nr5a2 differentially regulates female fertility across the
trajectory of follicular development. (Endocrinology 155: 1931–1943, 2014)

ust before birth, the mouse ovary contains a cluster of restricted to granulosa cells, and the expression pattern of
J densely packed oocytes, soon to be transformed by the
process of follicle assembly (1). The product is the pri-
the gene encoding the Amh type II receptor (Amhr2) over-
laps with Amh in postnatal granulosa cells (4).
mordial follicle, formed in the first 2 days after birth, when The first antral follicles are seen by day 13 and are
flat squamous pregranulosa cells surround the oocytes. characterized by the presence of multiple layers of gran-
Cohorts of these follicles enter the growing pool, and the ulosa cells containing scattered areas of interstitial fluid.
presence of cuboidal and dividing granulosa cells defines Granulosa cells of antral follicles express aromatase, the
these as primary follicles (2). The growth of the oocyte in enzyme responsible for catalyzing the conversion of an-
this class of follicles correlates with the beginning of anti- drogens into estrogens. The aromatase enzyme is encoded
Müllerian hormone (Amh) expression (3). Amh expression is by the cytochrome P450 family 19 subfamily A polypep-

ISSN Print 0013-7227 ISSN Online 1945-7170 Abbreviations: Akr1c18, aldo-keto reductase family 1 member C18; Amh, anti-Müllerian
Printed in U.S.A. hormone; Amhr2, Amh type II receptor; BSA, bovine serum albumin; cKO, conditional
Copyright © 2014 by the Endocrine Society knockout; CL, corpus luteum; COC, cumulus-oocyte complex; Cyp11a1, cytochrome P450
Received August 14, 2013. Accepted February 10, 2014. side-chain cleavage enzyme; Cyp19a1, cytochrome P450 family 19 subfamily A polypep-
First Published Online February 19, 2014 tide 1; Cy3, cyanine 3; DAPI, 4⬘,6-diamidino-2-phenylindole; dpc, day postcoitum; eCG,
equine chorionic gonadotropin; hCG, human CG; Hmg-CoA, 3-hydroxy-3-methylglutaryl-
coenzyme A; Hsd3␤, 3-hydroxysteroid dehydrogenase; Ldlr, low-density lipoprotein re-
ceptor; LS, luteal-like structure; Nr5a2, nuclear receptor subfamily 5 group A member 2;
PBS, phosphate buffered saline; Pgr, progesterone receptor; PRKO, Pgr knockout; Ptgs2,
prostaglandin-endoperoxide synthase-2; qPCR, quantitative PCR; RT, room temperature;
Scarb1, scavenger receptor class B type I; SF-1, steroidogenic factor 1; Star, steroidogenic
acute regulatory protein; TBST, Tris-buffered saline containing 0.1% Triton X-100.

doi: 10.1210/en.2013-1765 Endocrinology, May 2014, 155(5):1931–1943 endo.endojournals.org 1931

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
1932 Bertolin et al The Role of Nr5a2 in Luteinization Endocrinology, May 2014, 155(5):1931–1943

tide 1 (Cyp19a1) gene (5). The preovulatory LH surge All mutant and control mice were of the C57BL/6 background.
triggers a series of events, including cumulus expansion Animals were maintained under a 14-hour light, 10-hour dark
cycle and provided food and water ad libitum. Euthanasia was
from the oocyte in an extracellular matrix (6), a critical
performed by anesthetizing the animals with isoflurane (PPC) in
process for ovulation (7). After the LH peak, mural gran- a closed container, followed by cervical dislocation. Nr5a2
ulosa cells of preovulatory follicles are induced to express floxed (Nr5a2f/f) mice, generated by Dr Johan Auwerx, have
the progesterone receptor (Pgr) gene, with the maximum been described previously (17). To generate a granulosa-specific
Pgr expression just before ovulation (8), which occurs at Nr5a2 mutation in antral follicles, animals expressing Cre-re-
12 hours after the LH peak in the mouse (9). combinase from Cyp19a1 (tgCyp19a1Cre/⫹) (18) were crossed
with Nr5a2f/f mice (Nr5a2Cyp19⫺/⫺). To compare 2 different
Granulosa and theca cells from the postovulatory fol-
granulosa-specific mouse models, we employed Nr5a2f/f mice
licle undergo cellular reorganization to form the corpus crossed with animals expressing Cre-recombinase from the Amh
luteum (CL), a transient endocrine organ that produces type II receptor (Amhr2Cre/⫹) (generous gift of Dr Richard Beh-
progesterone, essential for endometrial cell differentiation ringer) (15), generating a granulosa-specific Nr5a2 mutant in
and maintenance of pregnancy (2). The functional remod- primary follicles (Nr5a2Amhr2⫺/⫺). The litters were genotyped
(for primers, see Supplemental Table 1, published on The En-
eling of the follicle to form the CL entails a large scale
docrine Society’s Journals Online web site at http://endo.en-
increase in progesterone synthesis, requiring up-regula- dojournals.org) using DNA extracted from tail biopsies. The
tion of cholesterol substrate availability (10), brought nonmutant littermates were used as control groups, ie,
about by expression of cholesterol transport-related pro- Nr5a2f/ftgCyp19⫹/⫹ and Nr5a2f/fAmhr2⫹/⫹ (respectively for
teins, such as scavenger receptor class B type I (Scarb1) the Nr5a2Cyp19⫺/⫺ and Nr5a2Amhr2⫺/⫺).
(11) and the steroidogenic acute regulatory protein (Star).
Cholesterol is then converted to pregnenolone by the cy- Estrous cycle and superstimulation protocol
For estrous cycle staging, vaginal smears were collected daily
tochrome P450 side-chain cleavage enzyme (Cyp11a1),
(9 AM) with phosphate buffered saline (PBS) 1⫻; smears were
and 3-hydroxysteroid dehydrogenase (Hsd3␤) converts placed on glass slides, and cytology was evaluated under the
pregnenolone to progesterone. microscope. For superstimulation, 25- to 27-day-old imma-
The orphan nuclear receptor subfamily 5, group A, ture female mice, both Nr5a2Cyp19⫺/⫺ and control, were in-
member 2 (Nr5a2), also known as liver receptor ho- jected ip with 5 IU of equine chorionic gonadotropin (eCG) (Fol-
molog-1, is highly expressed in granulosa cells of primary ligon; Intervet) to stimulate follicular development, followed by
the injection of 5 IU of human CG (hCG) (Chorulon; Intervet)
to preovulatory follicles (12) as well as in luteal cells (13).
44 – 48 hours after eCG, to induce ovulation. Ovaries and blood
Germline deletion of Nr5a2 results in early embryo lethal- samples (collected by cardiac puncture) were taken at 12, 18, or
ity (14); thus, there is a need for a strategy of conditional 24 hours after hCG; ovaries were either snap frozen in liquid
depletion to explore its role in the ovary. By means of nitrogen and stored at ⫺80°C for RNA or fixed in paraformal-
ovary-specific depletion, we have previously shown that dehyde 4% (Sigma-Aldrich) and embedded in paraffin for his-
tological examination. Ovulated cumulus-oocyte complexes
Nr5a2 is required for multiple events essential for ovula-
(COCs) were collected from oviducts at 16 hours after hCG
tion (15). It was also reported that Nr5a2 is essential for injection and counted. Pure granulosa and cumulus cells were
pregnancy, because it regulates uterine decidualization collected from Nr5a2Cyp19⫺/⫺ and control at 12 hours after
(16). To date, these studies have focused either on the role hCG for the detection of Nr5a2 by quantitative PCR (qPCR)
of Nr5a2 in ovarian follicular development from primary and immunoblotting; these cells were isolated by puncturing
follicles (15) or after the luteinizing stimulus (16). The both ovaries with 25-G 5/8-inch needles in PBS 1⫻, treating
with hyaluronidase (from bovine testes; Sigma) and filtered
consequences of depletion at different times in the follicu-
(40-␮m Nylon; BD Falcon, Cell Strainer) to remove the
logenesis suggest that there is a temporal sequence in the oocyte. At 4 hours after hCG, pure granulosa and cumulus cells
regulatory effects of Nr5a2. In the present study, we ex- were isolated from Nr5a2Cyp19⫺/⫺, Nr5a2Amhr2⫺/⫺ and their re-
plored the role of Nr5a2, using a mouse model with gran- spective control females for RNA analysis.
ulosa-specific depletion in antral and preovulatory folli-
cles. This new model provides insight into the role of Breeding trials
Nr5a2 in the later stages of follicle development. For breeding assays, 8-week-old Nr5a2Cyp19⫺/⫺ and control
females were housed with reproductively proven C57BL/6 males
for 6 months. Cages were inspected daily, and parturition dates
and litter sizes were recorded. Another group of 8-week-old
Materials and Methods Nr5a2Cyp19⫺/⫺ females was housed with C57BL/6 males,
whereas control females were housed with vasectomized males
Animals and colony maintenance for 40 days to establish the frequency of mating. The occurrence
All animal experiments were approved by the University of of copulatory plugs was verified by visual examination at 9 AM
Montreal Animal Care Committee and were conducted accord- each day. For vasectomies, male mice were anesthetized by in-
ing to the guidelines of the Canadian Council on Animal Care. jection of ketamine (100 mg/kg; Bioniche Animal Health) and

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1765 endo.endojournals.org 1933

xylazine (10 mg/kg; Bayer, Inc), then an incision of 1 cm through embedded. Ovarian cell-specific localization of Nr5a2, Nr5a1
the skin along the midline of the lower abdomen was made, and (steroidogenic factor 1 [SF-1]) and Scarb1 was verified by fluo-
1 cm of vas deferens on each side was removed. Two weeks later, rescence immunohistochemistry. Briefly, paraffin sections of
the sterility of the vasectomized males was confirmed by housing ovaries were rehydrated, boiled in 10mM sodium citrate (pH 6)
with fertile females and subsequent observation of the absence of for 30 minutes, and cooled to RT. Sections were blocked over-
pregnancy for 60 days. For analysis of progesterone, adult (⬎60 d night with 5% bovine serum albumin (BSA) (IgG-Free, Protease-
of age) Nr5a2Cyp19⫺/⫺ and control females were bred to males of Free; Jackson ImmunoResearch) at 4°C. Samples were then in-
proven fertility. Animals were examined daily at 9 AM for cop- cubated with Nr5a2 (ab18293; Abcam), rabbit polyclonal
ulatory plugs, and the day that a plug was found was designated antimouse Scarb1 (Novus Biologicals), or anti-SF-1 (ab65815;
1 day postcoitum (dpc). Animals were terminated at 5.5 dpc for Abcam), diluted, respectively, 1:50, 1:50, and 1:200 in PBS 1⫻
morphological analysis of the uterus and ovaries, and blood was with 5% BSA overnight at 4°C. As a negative control, some
collected for progesterone RIA. sections were incubated with 5% BSA only. Sections were
washed and incubated with cyanine 3 (Cy3)-conjugated goat
RNA isolation and real-time PCR (qPCR) antirabbit IgG (Jackson ImmunoResearch), diluted 1:400 in PBS
Total RNA from ovary, liver, and granulosa/cumulus cells 1⫻ for 1 hour at RT. Slides were then washed, and sections were
was extracted with RNeasy Mini kit (QIAGEN) as per manu- counterstained with 4⬘,6-diamidino-2-phenylindole (DAPI)
facturer’s instructions; adipose tissue total RNA was extracted (Sigma), diluted 1:1000 in PBS 1⫻, for 5 minutes. Slides were
by using the RNeasy Lipid Tissue Midi kit (QIAGEN). Reverse mounted in Permafluor (Thermo Fisher Scientific). Ovarian dis-
transcription was performed with the SuperScript II Reverse tribution of Nr5a2, Nr5a1, and Scarb1 was observed by Axio
Transcriptase (Invitrogen) and random hexamer primers. Real- imager M1 (Zeiss Microscopy).
time qPCR was performed using Power SYBR Green PCR Mas-
ter Mix (Applied Biosystems), with the 7300 Real-Time PCR Hormone assays
System (Applied Biosystems). A common thermal cycling pro- RIA for progesterone (serum) and estradiol-17␤ (serum and
gram (2 min at 50°C, 10 min at 95°C, then 40 cycles of 15 s at follicular fluid) concentrations was evaluated in single assays
95°C and 60 s at 60°C) was used to amplify each transcript. according to a protocol previously described (15). The intraassay
Melting curve analyses were performed to verify product iden- coefficients of variation, calculated between duplicates, ranged
tity, adding a dissociation step to the PCR run. The sequences of from 4% to 9%.
the primers used are listed in Supplemental Table 1, each was
used at a final concentration of 3␮M. Primers were validated by Statistical analyses
standard curves. Samples were run in triplicate, and results were All data were analyzed using JMP (version 5.0.1a; SAS Insti-
expressed relative to the ribosomal protein S18 levels. Data were tute, Inc) statistical software. Differences between mutant and
then normalized to a calibrator sample using ⌬⌬cycle threshold control mice for a single data point were determined by Student’s
method, as described by Pfaffl (19), with correction for ampli- t test. For data obtained at multiple time points or multiple ge-
fication efficiency by LinRegPCR 11.0 software (Academic notypes, two-way ANOVA was used and differences between
Medical Center, Amsterdam, the Netherlands). individual means were determined by the JMP LS-Means Con-
trast test. All numerical data are represented as mean ⫾ SEM.
Immunoblotting Significant difference was recognized at P ⬍ .05.
Total protein was extracted with lysis buffer 25 ␮g from gran-
ulosa/cumulus cells, separated on 10% SDS-PAGE, and subse-
quently transferred to a polyvinylidene difluoride membrane.
The membrane was blocked using 0.5% nonfat milk in Tris- Results
buffered saline containing 0.1% Triton X-100 (TBST) for 1 hour
The reproductive performance of female Nr5a2
at room temperature (RT), incubated with rabbit polyclonal
Nr5a2 (ab18293; Abcam), diluted 1:500 in blocking solution conditional knockout (cKO) mice in granulosa cells
overnight at 4°C, and then incubated in horseradish peroxidase- of antral follicles
conjugated goat antirabbit IgG (Santa Cruz Biotechnology, Inc) The cKO (Nr5a2Cyp19⫺/⫺) mouse model generated by
diluted 1:50 000 in TBST for 1 hour at RT. Proteins were visu- crossing Nr5a2 floxed mice with mice expressing the Cre-
alized by chemiluminescence (Immobilon Western; Millipore
recombinase driven by the aromatase promoter, ie, the
Corp). Subsequently, membranes were washed in TBST, blocked
as above, and incubated with horseradish peroxidase-conju- transgenic line tgCyp19a1Cre/⫹ (18), resulted in efficient
gated antimouse-␤-actin (Santa Cruz Biotechnology, Inc), di- and tissue-specific disruption of the Nr5a2 target gene as
luted 1:80 000 in blocking solution for 20 minutes at RT. The indicated by dramatically reduced mRNA abundance in
optical density of bands detected at the expected molecular ovarian granulosa cells from antral follicles but not in the
weight for Nr5a2 and ␤-actin was quantified by the Image Pro- other organs expressing Nr5a2 (Figure 1, A–C). Protein
cessing and Analysis in Java (ImageJ) software. The ratio of
levels of Nr5a2 were also reduced in granulosa cells of
Nr5a2 and ␤-actin densities was calculated for each sample.
the Nr5a2Cyp19⫺/⫺ ovaries, as shown by immunoblotting
Immunofluorescence (Figure 1, D and E) and by immunofluorescence (Figure 1,
For histological examination, ovaries were fixed in parafor- F–L). The temporal specificity of Nr5a2 down-regulation
maldehyde 4% overnight, then washed in PBS 1⫻ and paraffin in the Nr5a2Cyp19⫺/⫺ ovarian sections was evident by

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
1934 Bertolin et al The Role of Nr5a2 in Luteinization Endocrinology, May 2014, 155(5):1931–1943

Figure 1. Characterization of the granulosa-specific knockout mouse Nr5a2Cyp19⫺/⫺. A, Relative abundance of Nr5a2 mRNA in granulosa cells of
superstimulated immature females isolated by ovarian puncture (n ⫽ 4 per genotype) (P ⬍ .001). Relative abundance of Nr5a2 mRNA in the liver
(B) and adipose tissue (C) of adult females (n ⫽ 5 per genotype) (P ⬎ .1). D, Relative abundance of Nr5a2 protein in granulosa cells of
superstimulated immature females (n ⫽ 4 per genotype) by immunoblotting (P ⬍ .05). E, Representative immunoblot (F–L). Immunofluorescence
demonstrating the depletion of Nr5a2 protein in granulosa cells of the cKO ovary exclusively in the antral follicle (I) but not in preantral follicles (G).
Positive control for fluorescent immunohistochemistry staining of (F) preantral and (H) antral follicles from not-knockout ovaries (n ⫽ 5 per
genotype). J, Not-knockout ovarian histological section incubated with BSA 5% and stained with Cy3 as negative control. K, DAPI and Cy3
merged. L, Luteal cells as positive control. Scale bars, 60 ␮m (F–L). TC, theca cells; PAF, preantral follicle; GC, granulosa cells; O, oocyte; AF, antral
follicle; CC, cumulus cells; CON, control, solid bars (Nr5a2f/ftgCyp19⫹/⫹); cKO, open bars (Nr5a2Cyp19⫺/⫺).

depletion in granulosa cells of antral follicles (Figure 1I) model to produce pups is not due to an irregular estrous
but not in preantral follicles (Figure 1G), as expected. cycle.
The ovarian expression of Nr5a1 was not affected by
the lack of Nr5a2 in granulosa cells, because these nu- Antral follicles lacking Nr5a2 in granulosa cells are
clear receptors have a different pattern of distribution in anovulatory
the mouse ovary (Supplemental Figure 1, A and B). After Although preantral and antral follicle formation was
a 6-month breeding trial with reproductively proven observed in Nr5a2Cyp19⫺/⫺ ovaries (Supplemental Figure
males, the control group composed of intact littermates 1C), and cumulus expansion was stimulated by gonado-
(Nr5a2f/f tgCyp19a1⫹/⫹) generated the expected fre- tropin administration as seen at 12 hours after hCG
quency of litters per female and litter size, whereas no (Figure 3, A–C), almost no COCs were retrieved from the
litters were born from the Nr5a2Cyp19⫺/⫺ females (Figure Nr5a2Cyp19⫺/⫺ oviduct 16 hours after hCG injection, un-
2A). Daily consecutive evaluation of the vaginal epithe- like the superovulation induced in the control group (Fig-
lium evidenced a normal pattern of estrous cyclicity of 4 –5 ure 3, D–F). Analysis of ovarian granulosa and cumulus
days in Nr5a2Cyp19⫺/⫺ females compared with the control cells comparing the Nr5a2Cyp19⫺/⫺ with the previously
group (Figure 2B), with no abnormally prolonged inter- described Nr5a2Amhr2⫺/⫺ mouse revealed that the pattern
vals, indicating that the inability of the Nr5a2 cKO mouse of gene expression differed between the 2 knockout

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1765 endo.endojournals.org 1935

structures (LSs) in Nr5a2Cyp19⫺/⫺ ovarian sections. These

structures appeared to be comprised of theca cells invad-
ing the follicles, and the follicles often contained a trapped
oocyte (Figure 5, C and D). The number of LSs in
Nr5a2Cyp19⫺/⫺ ovaries was significantly lower than the
number of CLs found in the control ovarian sections (n ⫽
16 sections analyzed per group) (Figure 5E) at 12, 18, or
24 hours after hCG. We then asked whether the LSs
were functional. During a 40-day breeding trial, adult
Nr5a2Cyp19⫺/⫺ females were mated with reproductively
proven males, and control females were mated with va-
sectomized males to establish the frequency of mating by
examining copulatory plugs. The Nr5a2Cyp19⫺/⫺ females
showed frequent and irregular mating (mean interval,
3.67 ⫾ 0.21 d) compared with the control group (10.38 ⫾
0.51 d), indicating the absence of a persistent pseudopreg-
Figure 2. Reproductive performance of granulosa-specific nant CL in Nr5a2Cyp19⫺/⫺ females compared with the
Nr5a2Cyp19⫺/⫺ female mice. A, Six-month breeding trial employing control mice. To determine whether the Nr5a2Cyp19⫺/⫺
CON (n ⫽ 5) and cKO (n ⫽ 4) adult females housed with
reproductively proven C57BL/6 males; *, P ⬍ .001. B, Representative LSs could promote normal circulating levels of progester-
vaginal smear profiles from CON (solid line) and cKO (dotted line) one, adult Nr5a2Cyp19⫺/⫺ and control females were bred
female mice (n ⫽ 4 per genotype) showing a normal 4- to 5-day to fertile males, examined for copulatory plugs, and eu-
estrous cycle length for both genotypes. CON, control; P, proestrus; E,
estrus; M, metestrus; D, diestrus. thanized 5.5 dpc (Figure 5, G–J). By this time of expected
early postimplantation gestation, we found shorter uter-
models at 4 hours after hCG (Figure 4, A–F). The Pgr, ine horns in Nr5a2Cyp19⫺/⫺ (Figure 5H), in contrast to the
which is essential for ovulation, showed decreased elongated pregnant control uterus (Figure 5G). Although
mRNA levels in both models (Figure 4A), whereas Ad- LSs were observed in the Nr5a2Cyp19⫺/⫺ ovaries (Figure
amts4 was down-regulated in the Nr5a2Cyp19⫺/⫺ model 5J), they were fewer than the normal number of CLs in
but not in Nr5a2Amhr2⫺/⫺ ovary (Figure 4B). For Ad- the control ovaries (Figure 5I), and extremely low proges-
amts1, no difference in gene expression was seen between terone concentrations were detected in Nr5a2Cyp19⫺/⫺ se-
both cKO models relative to their respective controls (Fig- rum at 5.5 dpc (Figure 5K), whereas equivalent levels of se-
ure 4C). Although there was no statistically significant rum and intrafollicular estradiol-17␤ were identified in
difference for mRNA levels of the prostaglandin-endoper- Nr5a2Cyp19⫺/⫺ females compared with the control levels
oxide synthase-2 (Ptgs2) (Figure 4D) and the low-density (Supplemental Figure 1, D and E). Insufficiency of the
lipoprotein receptor (Ldlr) (Figure 4E), a tendency to- Nr5a2Cyp19⫺/⫺ LSs was confirmed by reduced circulating
wards down-regulation was observed in the granulosa concentrations of progesterone in the Nr5a2Cyp19⫺/⫺ se-
cells of both Nr5a2Cyp19⫺/⫺ and Nr5a2Amhr2⫺/⫺. Further, rum relative to control mice at 12, 18, or 24 hours after
at 4 hours after hCG, Cyp19a1 presented a tendency to-
hCG (Figure 5L). Based on the progesterone levels found
wards up-regulation in the Nr5a2Amh2⫺/⫺, whereas it was
after superstimulation, we then combined the eCG/hCG
down-regulated in Nr5a2Cyp19⫺/⫺ (Figure 4F). These data
protocol with the 5.5 dpc scheme to check whether su-
indicate that Nr5a2 expression in granulosa cells of antral
perstimulation would promote appropriate synthesis of
and preovulatory follicles is not essential for cumulus ex-
progesterone in adult females after mating. Although
pansion, but it is essential for ovulation.
control females were pregnant at 5.5 dpc (Supplemental
Antral expression of Nr5a2 is required for the Figure 2A), no implantation sites were observed in the
formation of a functional CL Nr5a2Cyp19⫺/⫺ uterine horns (Supplemental Figure 2B),
We then investigated the role of Nr5a2 in luteinization and progesterone was consistently low in Nr5a2Cyp19⫺/⫺
using standard hormonal superstimulation in immature serum compared with progesterone concentrations in con-
mice and analyzing the ovaries at later time points after the trol females (Supplemental Figure 2C). We conclude that
ovulatory stimulus (Figure 5, A–E). At 12, 18, and 24 depletion of Nr5a2 in granulosa cells from the antral fol-
hours after hCG, when CLs are normally abundant in the licle stage results in LSs that are insufficient to promote
control ovary (Figure 5, A and B), we identified luteal-like normal serum levels of progesterone.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
1936 Bertolin et al The Role of Nr5a2 in Luteinization Endocrinology, May 2014, 155(5):1931–1943

Figure 3. Ovarian superstimulation promotes cumulus expansion but not ovulation in granulosa-specific knockout female mice. A, Bright field
microscopy images of hematoxylin-eosin-stained sections of ovaries from immature cKO mice superstimulated with eCG and hCG and collected at
12 hours after hCG. Scale bar, 100 ␮m (n ⫽ 5 per genotype). PAF, preantral follicle; AF, antral follicle; CC, cumulus cells; O, oocyte. B, Arrow
indicates expanded cumulus cells at 12 hours after hCG in cKO follicle. Scale bar, 40 ␮m. C, Quantification of cumulus expansion by averaging the
number of follicles containing expanded COCs in CON and cKO ovarian sections at 12 hours after hCG (n ⫽ 6 sections, from the middle of the
ovary, per genotype) (P ⬎ .1). Ovulation rate (number of oocytes in the oviduct at 16 h after hCG) of CON (D; ⫻10 magnification) and cKO
(E; ⫻40 magnification) immature female mice in response to superstimulation. F, Mean number of oocytes in the oviduct in CON vs cKO mice (n ⫽
5 per genotype). Different superscripts indicate significant differences at P ⬍ .001. CON, control.

The expression pattern of luteinization-related losa cells of antral follicles at 18 and 24 hours (Figure 6,
genes driven by Nr5a2 in preovulatory follicles E and F). The expression of the aldo-keto reductase family
To explore the abundance of transcripts of genes es- 1 member C18 (Akr1c18) (Figure 6G) in the cKO ovaries
sential for cholesterol supply and progesterone synthesis was decreased at 12 hours and less evidently decreased at
in granulosa cells in response to LH stimulation, immature 18 hours after hCG; at 24 hours, a significant increase in
females were treated with eCG/hCG, and whole ovaries its expression was observed. The relative mRNA abun-
were collected at 12, 18, or 24 hours after hCG (Figure 6, dance of 3-hydroxy-3-methylglutaryl-coenzyme A (Hmg-
A–I). The relative abundance of Scarb1 mRNA was CoA) (Figure 6H) was increased in cKO ovaries at 12, 18,
markedly reduced beginning at 18 hours and continuing and 24 hours after hCG. Inversely, the vascular endothe-
through 24 hours after hCG in the Nr5a2Cyp19⫺/⫺ group lial growth factor A (Vegfa) (Figure 6I) had a decreased
compared with the controls (Figure 6A). The abundance expression in the cKO in the 3 time points. We then ex-
of Ldlr, Star, and Cyp11a1 was lower in Nr5a2Cyp19⫺/⫺ amined the expression of Scarb1 in ovarian sections by
ovaries at 18 hours after hCG, whereas at 12 and 24 hours, immunofluorescence to determine whether the luteal-like
there was a tendency of down-regulation for these genes cells were indeed luteinized granulosa cells (Figure 6, J and
(Figure 6, B–D). This notwithstanding, the strength of the K). In Nr5a2Cyp19⫺/⫺ ovaries, LSs have an apparent higher
mRNA signal for Hsd3␤ and Cyp19a1 showed an increase signal of Scarb1 than granulosa cells in the same sections
in control mice over 12 hours after hCG, but the expres- (Figure 6K), but the protein expression appears substan-
sion was not altered by the depletion of Nr5a2 in granu- tially lower than in the normal CL of control animals

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1765 endo.endojournals.org 1937

Figure 4. Abundance of mRNA as determined by qPCR for Pgr (A), Adamts4 (B), Adamts1 (C), Ptgs2 (D), Ldlr (E), and Cyp19a1 (F) relative to
RPS18 comparing 2 granulosa-specific knockout mouse lines distinguished by the timing of Nr5a2 depletion during the follicular development.
Superstimulated ovaries from immature females were collected at 4 hours after hCG (n ⫽ 5 per group) and granulosa cells were isolated by
ovarian puncture. Amhr2 (Cre-recombinase): left bars, mutation in primary follicles, genotype Nr5a2Amhr2⫺/⫺; Cyp19 (Cre-recombinase), right bars,
mutation in antral follicles, genotype Nr5a2Cyp19⫺/⫺. CON, control, black bars (Nr5a2f/f tgCyp19⫹/⫹; or Nr5a2f/f Amhr2⫹/⫹); cKO, gray bars
(Nr5a2Cyp19⫺/⫺ or Nr5a2Amhr2⫺/⫺), respectively. Differing superscripts indicated differences at P ⬍ .05.

(Figure 6J). These data demonstrate that the deficient tor Nr5a2 has not only been designated by multiple names,
luteinization resulting from the absence of Nr5a2 in liver receptor homolog-1, ␣-fetoprotein transcription fac-
preovulatory follicles is related to the compromised ex- tor, Fushi tarazu factor 1, and cholesterol 7 alpha-hydrox-
pression of the gene program, and those involved in ylase promoter binding factor, among others, but it has
cholesterol transport, but not cholesterol synthesis, are been recognized to have an immense range of functions in
affected. both vertebrate and invertebrate models. Nr5a2 is ex-
pressed in steroidogenic tissues (22), and its strong gran-
ulosa-specific expression has led to investigation of its role
Discussion
in reproduction, especially in the ovaries (13, 23, 24).
From the time of its discovery in Drosophila (25, 26) to More recently, it has been shown to be an important reg-
our present state of knowledge, the orphan nuclear recep- ulator in the mouse endometrium, to be essential for suc-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
1938 Bertolin et al The Role of Nr5a2 in Luteinization Endocrinology, May 2014, 155(5):1931–1943

Figure 5. Nr5a2 is important for the formation of a normal CL. A–D, Bright field microscopy images of hematoxylin-eosin-stained sections of
ovaries from immature control and Nr5a2Cyp19⫺/⫺ mice superstimulated with eCG/hCG and collected at 18 hours after hCG (n ⫽ 5 per genotype).
A, Overview of the control ovary containing several CLs. Scale bar, 100 ␮m. B, A single CL from a control ovarian section. Scale bar, 40 ␮m. C,
General overview of the Nr5a2Cyp19⫺/⫺ ovary containing 1 LS. Scale bar, 100 ␮m. D, Nr5a2Cyp19⫺/⫺ LS showing an entrapped oocyte (O) indicated
by the arrow. Scale bar, 40 ␮m. E, Quantification of luteinization by averaging the frequency of luteal formation in control and Nr5a2Cyp19⫺/⫺
ovarian sections at 12, 18, and 24 hours after hCG (n ⫽ 5 sections per genotype) (P ⬍ .001). F, Mating frequency comparing adult control (n ⫽ 7)
and Nr5a2Cyp19⫺/⫺ (n ⫽ 5) females during a 40-day breeding trial. Control females were housed with vasectomized C57BL/6 males, and
Nr5a2Cyp19⫺/⫺ females were housed with reproductively proven C57BL/6 males. Black squares indicate the observation of a copulatory plug, with
short and irregular interval between plugs for Nr5a2Cyp19⫺/⫺ females, whereas the control group averaged 10- to 13-day interval between plugs;
P ⬍ .0001. Ovaries (OV) and uteri of control (G) and Nr5a2Cyp19⫺/⫺ (H) adult females housed with a reproductively proven C57BL/6 male and
collected at 5.5 dpc (n ⫽ 4 per genotype). G, Elongated uterine horns containing implantation sites (indicated by the arrows) and the presence of
several CLs in the hematoxylin-eosin-stained ovary (I) represents the control female reproductive system under the influence of the progesterone.
A few LSs are found in the hematoxylin-eosin-stained sections of the Nr5a2Cyp19⫺/⫺ ovaries (J). Scale bars, 4 mm (G and H) and 40 ␮m (I and J).
Plasma progesterone concentrations in control and Nr5a2Cyp19⫺/⫺ adult female mice at 5.5 dpc (n ⫽ 4 per genotype) (K) and in immature
superstimulated female mice at 12, 18, and 24 hours after hCG (n ⫽ 5 per genotype) (L) (P ⬍ .001).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1765 endo.endojournals.org 1939

Figure 6. Pattern of expression as determined by qPCR for genes related to steroidogenesis at intervals around the time of and after expected
ovulation. Relative abundance of mRNA of Scarb1 (A), Ldlr (B), Star (C), Cyp11a1 (D), Hsd3␤ (E), Cyp19a1 (F), Akr1c18 (G), Hmg-CoA (H), and
Vegfa (I) in whole ovaries of superstimulated immature females collected at 12, 18, or 24 hours after hCG (n ⫽ 5 per group) comparing control
and Nr5a2Cyp19⫺/⫺ (P ⬍ .05) groups. Merged microscopy images of fluorescent immunohistochemistry of Scarb1 in ovarian sections of control (J)
and Nr5a2Cyp19⫺/⫺ (K) adult females. AF, antral follicle. Scale bar, 50 ␮m. J, inset, Negative control with 5% BSA using DAPI to stain the nuclei.

cessful gestation (16), and it appears to be upstream to The SF-1 (also known as Nr5a1) is the closest mam-
regulation of gonadotropin secretion in the hypothalamus malian homolog to Nr5a2. Nr5a1 is expressed in ovarian
(25). theca cells, with lower levels present in granulosa cells and

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
1940 Bertolin et al The Role of Nr5a2 in Luteinization Endocrinology, May 2014, 155(5):1931–1943

no expression in luteal cells (26). Despite their structural This phenotype of cumulus expansion without ovulation
conservation, there is a remarkable specificity in both their is similar to that found in the Pgr knockout (PRKO)
specific activation profiles and their selective induction of mouse. Granulosa cells from preovulatory follicles of mice
specific target genes and signaling pathways (26). Nr5a1 and in the PRKO are responsive to the LH surge, demonstrated
Nr5a2 clearly derive from separate genes and have a different by the presence of expanded cumulus cells (29), but Pgr is
pattern of distribution in the mammalian tissues (27); they essential for LH-induced follicular rupture leading to ovu-
also have distinct structural ligand binding domains and lation (8). In the present investigation, Nr5a2 depletion in
regions that recruit specific cofactors to regulate tran- either the Nr5a2Cyp19⫺/⫺ model or the Nr5a2Amhr2⫺/⫺ is
scription. The integration of the variant domains within accompanied by significant down-regulation of Pgr. This
the receptor results in the specificity of their activities, suggests that Nr5a2 is upstream of the Pgr gene and reg-
contributing to their different biological functions (28). ulates ovulation as a transcription factor to activate the
Nr5a2 is necessary for proper embryonic primitive critical Pgr pathway.
streak formation, thus preventing early embryo survival in The Pgr, in turn, is an important regulator of the
the mouse (14), precluding studies of its germline mutation transcription of other genes linked to ovulatory success,
in postnatal animals. Employing a mouse line with loxP sites including Adamts1 (30). At 4 hours after hCG, the Ad-
flanking the third and fourth exons of Nr5a2 gene, and tak- amts1 levels are normal for both Nr5a2Amhr2⫺/⫺ and
ing advantage of the Cre-loxP system has allowed in depth Nr5a2Cyp19⫺/⫺ ovaries. However, at 12 hours after hCG,
exploration of Nr5a2 in the ovary in recent years. By the a down-regulation of Adamts1 occurs in Nr5a2Cyp19⫺/⫺
use of a Cre-recombinase mouse line guided by Amhr2 (data not shown), whereas in the Nr5a2Amhr2⫺/⫺ ovaries,
(Nr5a2Amhr2⫺/⫺), it was shown that Nr5a2 is essential for normal levels are still observed (15). Although this could be
cumulus expansion, ovulation, and female fertility (15). A related to the different cumulus expansion profiles in these 2
later study on the role of Nr5a2 in reproduction applied mouse models, other pathways may also be involved. The
the Cre-recombinase mouse line guided by the Pgr pattern of abundance of the Adamts4 transcript also
(Nr5a2Pgr⫺/⫺), which was responsible for depleting Nr5a2 differs between the 2 granulosa-specific knockout mod-
during the late periovulatory period in follicles and in the els; although it is down-regulated in Nr5a2Cyp19⫺/⫺ ova-
uterus, resulting in successful ovulation, formation of ries, it is normally expressed at 4 hours after hCG in the
small but functional CLs, and infertility due to compro- Nr5a2Amhr2⫺/⫺ ovaries. Although Adamts4 is closely re-
mised endometrial decidualization (16). lated to Adamts1, with some distinct and overlapping
In the present study, we employed a third model with functions, it does not appear to be a target for the Pgr (31).
a different temporal pattern of depletion of Nr5a2. Follicular prostaglandins are mandatorily increased be-
Mating with the Cyp19a1 Cre-recombinase mouse line fore ovulation, induced by selective induction of the Ptgs2
(Nr5a2Cyp19⫺/⫺) resulted in excision of Nr5a2 from the expression in granulosa cells (32). Although the progester-
antral follicle forward in the follicle development trajectory one pathway seems to be regulated by Nr5a2, the expression
(18). Females were infertile with an ovarian phenotype dis- of Ptgs2, a marker of follicular commitment to ovulation
tinct from the 2 previous models, demonstrating another after the LH peak (32), appears to be normal in the PRKO
variant reproductive profile for cKO mice lines in these stud- mice (33), whereas Ptgs2 tends to be down-regulated in both
ies. Our Nr5a2Cyp19⫺/⫺ mouse model confirms that Nr5a2 Nr5a2Cyp19⫺/⫺ and Nr5a2Amhr2⫺/⫺ models. Thus, the com-
is essential for ovulation with the almost completely absent plex phenotypes resulting from depletion of Nr5a2 in gran-
COCs found in Nr5a2Cyp19⫺/⫺ female oviducts after treat- ulosa cells at different stages of follicular maturation strongly
ment with an ovulation induction protocol. This indicates indicate that Nr5a2 regulates multiple pathways of female
that expression of Nr5a2 in antral follicles is required to fertility. Further evidence that Nr5a2 regulates multiple
provide the necessary ovulatory signals. The Nr5a2Pgr⫺/⫺ pathways can be found in the observation that, despite the
mouse model demonstrates that ovulation is independent of failure to ovulate, luteinization proceeds normally in the
Nr5a2 after the ovulatory stimulus, because these animals PRKO mice (33) but not in our models lacking Nr5a2 in
showed normal ovulatory rates (16). granulosa cells.
Although the nonovulatory phenotype is present in In our previous studies using Nr5a2Amhr2⫺/⫺ and
both the Nr5a2Cyp19⫺/⫺ and the Nr5a2Amhr2⫺/⫺ mouse Nr5a2Pgr⫺/⫺ models, although they suggested that Nr5a2
models, an interesting difference between them is that cu- has a role in ovarian luteinization (15, 16), the time of
mulus expansion is absent when Nr5a2 is depleted in pri- depletion of the gene was either very early in the follicular
mary follicles in the Nr5a2Amhr2⫺/⫺ mouse (15), whereas development or after the LH peak. In the Nr5a2Amhr2⫺/⫺
the presence of Nr5a2 in preantral follicles is able to pro- model, the lack of Nr5a2 in granulosa cells in primary
mote cumulus expansion in the Nr5a2Cyp19⫺/⫺ females. follicles prevents CL formation (15). In the Nr5a2Pgr⫺/⫺

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1765 endo.endojournals.org 1941

model, CLs, although at reduced size, were formed in the portance of Nr5a2 up to later stages of the luteinization,
Nr5a2Pgr⫺/⫺ ovaries (16), and this formation is probably but this model does not provide information on the role of
due to the persistence of Nr5a2 expression in luteal cells Nr5a2 in the early period of luteinization, due to persis-
for at least 4 hours after hCG, the time when Cre-recom- tence of Nr5a2 (16). Our Nr5a2Cyp19⫺/⫺ model appears to
binase induces Nr5a2 depletion in this model. The com- be the best currently available to allow insight into the early
promised CL in the Nr5a2Pgr⫺/⫺ ovaries indicates the im- luteal regulation of Nr5a2.
Therefore, with the Nr5a2Cyp19⫺/⫺
model, we provide new information on
the role of Nr5a2 in luteal formation,
because the time of Nr5a2 excision in
granulosa cells of antral follicles is
guided by the Cyp19a1 expression,
which appears before the LH peak in
the ovary (18). However, our granu-
losa-specific Nr5a2Cyp19⫺/⫺ knock-
out mice present a luteal-like forma-
tion in the ovaries, which are
apparently comprised of luteinized
theca cells surrounding undifferen-
tiated granulosa cells. These struc-
tures were present in adult and im-
mature superstimulated females.
Although the number of these struc-
tures is reduced if compared with the
normal number of CLs in the control
group, they appear able to generate
some circulating levels of progester-
one after hormonal superstimula-
tion. This secretion did not persist in
adult females to 5.5 dpc. During a
40-day breeding trial, control fe-
males became pseudopregnant after
a sterile mating. In mouse pseudo-
pregnancy, the surge of prolactin
and the progesterone synthesized by
the CLs prevent the LH peak; with
increased secretion of prostaglandin
(prostaglandin F2 ␣) in the absence of
embryos, progesterone levels de-
cline, and females resume the estrous
Figure 7. Summary of the Nr5a2 functions in the ovary. Comparison of 3 different cKO mouse cycle (34), with a consequent inter-
models. A, Diagram of normal follicular maturation in the mouse ovary. Primary follicles (1) have val of 10 –13 days between copula-
1 layer of granulosa cells around the oocyte, acquiring new layers during follicular development; tions. The frequent and irregular in-
tertiary follicles (2) will develop to antral follicles (3), when granulosa cells differentiate into mural
and cumulus cells. When the LH peak occurs (4), immediate changes happen in the follicle (5), tervals between copulations in the
including cumulus expansion (6), ovulation (7), and luteinization (8). B, By the Cre-loxP system, Nr5a2Cyp19⫺/⫺, indicating the re-
flanking the third and fourth Nr5a2 exons and using the Amhr2-Cre recombinase, Nr5a2 is turn to estrus, confirm the nonfunc-
excised at the primary follicle stage, which does not affect folliculogenesis; on the other hand,
tionality of the LSs in producing the
cumulus expansion, ovulation, and luteinization are absent. C, Excising Nr5a2 at the antral follicle
stage before the LH peak, by the use of the Cyp19-Cre recombinase, both folliculogenesis and necessary amount of progesterone to
cumulus expansion are observed, but ovulation is impaired; luteinization is abnormal, and promote pseudopregnancy.
oocytes are trapped in the luteal-like formations. D, Excising Nr5a2 at the antral follicle stage At the molecular level, luteiniza-
after the LH peak, by the use of the Pgr-Cre recombinase, the aberrant ovarian phenotype is
restricted to the formation of reduced size CLs, but folliculogenesis, cumulus expansion, and tion in Nr5a2Cyp19⫺/⫺ mice appears
ovulation occur normally. to be impaired by disruptions in the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
1942 Bertolin et al The Role of Nr5a2 in Luteinization Endocrinology, May 2014, 155(5):1931–1943

cholesterol supply and transport of proteins. Of impor- technical assistance; Dr Jane Fenelon for carefully reading this
tance is the down-regulation of Scarb1. This factor is re- manuscript; and Dr Richard Behringer for the Amhr2Cre/⫹
sponsible for mediating the selective uptake of cholesterol mouse line.
from high-density lipoprotein, the principal source of cho-
Address all correspondence and requests for reprints to: Dr
lesterol supply in the mouse ovary (11). We have shown
Bruce D. Murphy, PhD, Centre de Recherche en Reproduction
that ablation of Scarb1 results in defective intake of the
Animale, Université de Montréal, 3200 Rue Sicotte, Saint
precursor cholesterol into the granulosa-derived luteal
Hyacinthe, Québec, Canada J2S 7C6. E-mail: bruce.d.
cells (35). A second cholesterol import pathway initiated
murphy@umontreal.ca.
by Ldlr is also down-regulated in the Nr5a2Cyp19⫺/⫺ ova- This work was supported by the Canadian Institutes of
ries, while there is modest up-regulation of Hmg-CoA, the Health Research Operating Grant FRN11018 (to B.D.M.). K.B.
rate-limiting enzyme in de novo cholesterol synthesis. The was supported by the Fonds Québécois de la Recherche sur la
expression of other genes known to be necessary for ste- Nature et les Technologies (FQRNT)–Merit Scholarship Pro-
roidogenesis was also disrupted, including Star, the pro- gram for Foreign Students (PBEEE).
tein that governs the transport of cholesterol from the Disclosure Summary: The authors have nothing to disclose.
outer to inner mitochondrial membrane (36), and
Cyp11a1, which is involved in the conversion of choles-
terol to pregnenolone. The transcript for the next enzyme
References
in the synthetic pathway, the Hsd3␤, is up-regulated ear-
lier after ovulation, perhaps to compensate for the down- 1. Peters H. The development of the mouse ovary from birth to ma-
regulation of the previous steroidogenic steps. Compari- turity. Acta Endocrinol (Copenh). 1969;62(1):98 –116.
2. Rajkovic A, Pangas SA, Matzuk MM. Follicular development:
son of the Nr5a2Cyp19⫺/⫺ ovary with that found in the
mouse, sheep, and human models. In: Knobil and Neill’s Physiology
Nr5a2Pgr⫺/⫺ model mouse, where ovulation and lutein- of Reproduction. 3rd ed. Los Angeles, CA: Elsevier; 2006:383– 423.
ization occur but CLs are deficient in progesterone syn- 3. Münsterberg A, Lovell-Badge R. Expression of the mouse anti-mül-
thesis (21), indicates both similarities and differences in lerian hormone gene suggests a role in both male and female sexual
differentiation. Development. 1991;113(2):613– 624.
the gene expression patterns. In the Nr5a2Pgr⫺/⫺, the de-
4. Jamin SP, Arango NA, Mishina Y, Hanks MC, Behringer RR. Ge-
cline in the expression of Star and Cyp11a1 phenocopies netic studies of the AMH/MIS signaling pathway for Müllerian duct
the Nr5a2Cyp19⫺/⫺ model, whereas Scarb1 and Ldlr re- regression. Mol Cell Endocrinol. 2003;211:15–19.
ductions are not present until substantially later. The tran- 5. Stocco C. Aromatase expression in the ovary: hormonal and mo-
script Akr1c18, the enzyme that converts progesterone to lecular regulation. Steroids. 2008;73(5):473– 487.
6. Espey LL, Richards JS. Ovulation. In: Knobil and Neill’s Physiology
its less potent 20a hydroxyprogesterone metabolite, in- of Reproduction. 3rd ed. Los Angeles, CA: Elsevier; 2006:425– 474.
creases earlier in the Nr5a2Cyp19⫺/⫺ model, whereas abun- 7. Elvin JA, Clark AT, Wang P, Wolfman NM, Matzuk MM. Paracrine
dance of Hsd3␤ declines in the Nr5a2Pgr⫺/⫺ model and is actions of growth differentiation factor-9 in the mammalian ovary.
unaffected in the Nr5a2Cyp19⫺/⫺ ovary. These differences Mol Endocrinol. 1999;13(6):1035–1048.
8. Akison LK, Robker RL. The critical roles of progesterone receptor
are consistent with the successful ovulation in Nr5a2Pgr⫺/⫺ (PGR) in ovulation, oocyte developmental competence and oviduc-
mouse and the formation of CLs from the expected gran- tal transport in mammalian reproduction. Reprod Domest Anim.
ulosa and theca contributions, compared with the small 2012;47(suppl 4):288 –296.
luteal structure where the granulosa cell component has 9. Tullet JM, Pocock V, Steel JH, White R, Milligan S, Parker MG.
Multiple signaling defects in the absence of RIP140 impair both
not differentiated to its luteal counterpart in the Nr5a2Cyp19⫺/⫺
cumulus expansion and follicle rupture. Endocrinology. 2005;
line. 146(9):4127– 4137.
In conclusion, our findings demonstrate that Nr5a2 10. Murphy BD. Models of luteinization. Biol Reprod. 2000;63(1):
regulates ovulation and female fertility differently in gran- 2–11.
11. Jiménez LM, Binelli M, Bertolin K, Pelletier RM, Murphy BD. Scav-
ulosa cells across the trajectory of follicular development
enger receptor-B1 and luteal function in mice. J Lipid Res. 2010;
(Figure 7). The Pgr pathway seems to be the immediate 51(8):2362–2371.
downstream target for Nr5a2 activity in ovulation. The 12. Mendelson CR, Kamat A. Mechanisms in the regulation of aroma-
inadequate luteal function can be attributed, at least in tase in developing ovary and placenta. J Steroid Biochem Mol Biol.
part, to the disruption of the cholesterol transport and 2007;106(1–5):62–70.
13. Hinshelwood MM, Repa JJ, Shelton JM, Richardson JA, Mangels-
steroidogenic pathways, although other pathways are also dorf DJ, Mendelson CR. Expression of LRH-1 and SF-1 in the mouse
affected. ovary: localization in different cell types correlates with differing
function Mol Cell Endocrinol. 2003;207(1–2):39 – 45.
14. Labelle-Dumais C, Jacob-Wagner M, Paré JF, Bélanger L, Dufort D.
Acknowledgments Nuclear receptor NR5A2 is required for proper primitive streak
morphogenesis. Dev Dyn. 2006;235(12):3359 –3369.
We thank Anne-Marie Bellefleur for valuable scientific discus- 15. Duggavathi R, Volle DH, Mataki C, et al. Liver receptor homolog
sions; Mira Dobias Goff, Vickie Roussel, and Fanny Morin for 1 is essential for ovulation. Genes Dev. 2008;22:1871–1876.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/en.2013-1765 endo.endojournals.org 1943

16. Zhang C, Large MJ, Duggavathi R, et al. Liver receptor homolog-1
is essential for pregnancy. Nat Med. 2013;19(8):1061–1066.
17. Botrugno OA, Fayard E, Annicotte JS, et al. Synergy between LRH-1
and ␤-catenin induces G1 cyclin-mediated cell proliferation. Mol
Cell. 2004;15(4):499 –509.
18. Fan HY, Shimada M, Liu Z, et al. Selective expression of KrasG12D
in granulosa cells of the mouse ovary causes defects in follicle de-
velopment and ovulation. Development. 2008;135(12):2127–
2137.
19. Hellemans J, Mortier G, De Paepe A, Speleman F, Vandesompele J.
qBase relative quantification framework and software for manage-
ment and automated analysis of real-time quantitative PCR data.
Genome Biol. 2007;8(2):R19.
20. Lavorgna G, Ueda H, Clos J, Wu C. FTZ-F1, a steroid hormone
receptor-like protein implicated in the activation of fushi tarazu.
Science. 1991;252(5007):848 – 851.
21. Broadus J, McCabe JR, Endrizzi B, Thummel CS, Woodard CT. The
Drosophila ␤ FTZ-F1 orphan nuclear receptor provides competence
for stage-specific responses to the steroid hormone ecdysone. Mol
Cell. 1999;3(2):143–149.
22. Clyne CD, Speed CJ, Zhou J, Simpson ER. Liver receptor homo-
logue-1 (LRH-1) regulates expression of aromatase in preadi-
pocytes. J Biol Chem. 2002;277(23):20591–20597.
23. Sirianni R, Seely JB, Attia G, et al. Liver receptor homologue-1 is
expressed in human steroidogenic tissues and activates transcription
of genes encoding steroidogenic enzymes. J Endocrinol. 2002;
174(3):R13–R17.
24. Boerboom D, Pilon N, Behdjani R, Silversides DW, Sirois J. Ex-
pression and regulation of transcripts encoding two members of the
NR5A nuclear receptor subfamily of orphan nuclear receptors, ste-
roidogenic factor-1 and NR5A2, in equine ovarian cells during the
ovulatory process. Endocrinology. 2000;141(12):4647– 4656.
25. Atkin SD, Owen BM, Bookout AL, et al. Nuclear receptor LRH-1
induces the reproductive neuropeptide kisspeptin in the hypothal-
amus. Mol Endocrinol. 2013;27(4):598 – 605.
26. Fayard E, Auwerx J, Schoonjans K. LRH-1: an orphan nuclear re-
ceptor involved in development, metabolism and steroidogenesis.
Trends Cell Biol. 2004;14(5):250 –260.
27. Parker KL, Schimmer BP. Steroidogenic factor 1: a key determinant
of endocrine development and function. Endocr Rev. 1997;18(3):
361–377.
28. Sablin EP, Krylova IN, Fletterick RJ, Ingraham HA. Structural basis
for ligand-independent activation of the orphan nuclear receptor
LRH-1. Mol Cell. 2003;11(6):1575–1585.
29. Lydon JP, DeMayo FJ, Funk CR, et al. Mice lacking progesterone
receptor exhibit pleiotropic reproductive abnormalities. Genes Dev.
1995;9(18):2266 –2278.
30. Espey LL, Yoshioka S, Russell DL, Robker RL, Fujii S, Richards JS.
Ovarian expression of a disintegrin and metalloproteinase with
thrombospondin motifs during ovulation in the gonadotropin-
primed immature rat. Biol Reprod. 2000;62(4):1090 –1095.
31. Richards JS. Ovulation: new factors that prepare the oocyte for
fertilization. Mol Cell Endocrinol. 2005;234:75–79.
32. Sirois J, Sayasith K, Brown KA, Stock AE, Bouchard N, Doré M.
Cyclooxygenase-2 and its role in ovulation: a 2004 account. Hum
Reprod Update. 2004;10(5):373–385.
33. Robker RL, Russell DL, Espey LL, Lydon JP, O’Malley BW, Rich-
ards JS. Progesterone-regulated genes in the ovulation process: AD-
AMTS-1 and cathepsin L proteases. Proc Natl Acad Sci USA. 2000;
97(9):4689 – 4694.
34. Dewar AD. Observations on pseudopregnancy in the mouse. J En-
docrinol. 1959;18(2):186 –190.
35. Acton S, Rigotti A, Landschulz KT, Xu S, Hobbs HH, Krieger M.
Identification of scavenger receptor SR-BI as a high density lipo-
protein receptor. Science. 1996;271(5248):518 –520.
36. Christenson LK, Devoto L. Cholesterol transport and steroidogen-
esis by the corpus luteum. Reprod Biol Endocrinol. 2003;1:90.

Join The Endocrine Society and network

with endocrine thought leaders from around the world.
www.endocrine.org/join

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 20 March 2015. at 06:07 For personal use only. No other uses without permission. . All rights reserved.