Estradiol, Progesterone, and Genistein Inhibit OocyteNest Breakdown and Primordial Follicle Assembly in theNeonatal Mouse Ovary in Vitro and in Vivo
Ying Chen, Wendy N. Jefferson, Retha R. Newbold, Elizabeth Padilla-Banks, and Melissa E. Pepling
Department of Biology (Y.C., M.E.P.), Syracuse University, Syracuse, New York 13244; and Developmental Endocrinologyand Endocrine Disruptor Section (W.N.J., R.R.N., E.P.-B.), Laboratory of Molecular Toxicology, National Institute ofEnvironmental Health Sciences, National Institutes of Health/Department of Health and Human Services, ResearchTriangle Park, North Carolina 27709
In developing mouse ovaries, oocytes develop as clusters ofcells called nests or germ cell cysts. Shortly after birth, oocytenests dissociate and granulosa cells surround individual oo-cytes forming primordial follicles. At the same time, twothirds of the oocytes die by apoptosis, but the link betweenoocyte nest breakdown and oocyte death is unclear. Althoughmechanisms controlling breakdown of nests into individualoocytes and selection of oocytes for survival are currentlyunknown, steroid hormones may play a role. Treatment ofneonatal mice with natural or synthetic estrogens results inabnormal multiple oocyte follicles in adult ovaries. Neonatalgenistein treatment inhibits nest breakdown suggesting mul-tiple oocyte follicles are nests that did not break down. Herewe investigated the role of estrogen signaling in nest break-down and oocyte survival. We characterized an ovary organ
culture system that recapitulates nest breakdown, reductionin oocyte number, primordial follicle assembly, and folliclegrowth in vitro. We found that estradiol, progesterone, andgenistein inhibit nest breakdown and primordial follicle as-sembly but have no effect on oocyte number both in organculture and in vivo. Fetal ovaries, removed from their normalenvironment of high levels of pregnancy hormones, under-went premature nest breakdown and oocyte loss that wasrescued by addition of estradiol or progesterone. Our resultsimplicate hormone signaling in ovarian differentiation withdecreased estrogen and progesterone at birth as the primarysignal to initiate oocyte nest breakdown and follicle assembly.These findings also provide insight into the mechanism ofmultiple oocyte follicle formation. (Endocrinology 148:3580–3590, 2007)
DIFFERENTIATION OF PRIMORDIAL germ cells intooocytes is essential for ovarian differentiation and
subsequent female fertility. In the embryonic mouse ovary,germ cells undergo several rounds of mitosis and are clas-sified as oogonia (1). Cytokinesis is incomplete during thesecell divisions so that the oogonia form clusters of connectedcells called nests or germ cell cysts similar to the germlinecysts of invertebrate females such as Drosophila (2). Afteroogonia stop dividing, they enter meiosis and become oo-cytes (3). The oocytes remain in clusters, which here will bereferred to as oocyte nests, for the rest of fetal development.After birth, oocytes separate and become enclosed in pri-mordial follicles consisting of one oocyte and several somaticgranulosa cells. Oocytes in primordial follicles are thought torepresent the entire pool of gametes available to a femalethroughout her reproductive life, although it has been re-cently suggested that there are postnatal germ line stem cellsin the adult ovary (4, 5). Thus, establishment of the primor-dial follicle pool is essential for mammalian reproduction.
The process by which oocyte nests break apart to form
primordial follicles is not understood. In mammalian speciesincluding the mouse, more than half of oocytes die early indevelopment before follicles are formed (3, 6–9). During thisprocess, some cells in each nest die by programmed celldeath, leaving only a third of the total number surviving. Inour model for mammalian oocyte nest breakdown (8), onecell of a nest dies causing the nest to break into smaller nests.This is repeated until a few individual oocytes remain. Thusprogrammed cell death of some oocytes would be requiredfor nests to break down.
Developmental exposure to estrogenic compounds ex-erts effects on reproductive organs including increasedoccurrence of multiple oocyte follicles (MOFs) in the adultovary (10). In normal adult female mouse ovaries, folliclesconsist of a single oocyte surrounded by layers of gran-ulosa cells; follicles with more than one oocyte are rarelyfound (11). In contrast, ovaries from adult female micetreated as neonates with natural estrogens or compoundswith estrogen-like activity have more MOFs (12–16).MOFs are postulated to be oocyte clusters that did notseparate, resulting in more than one oocyte becoming en-closed in a single follicle (12, 13, 17).
Genistein, a phytoestrogen, causes an increase in MOFs(16). Genistein has properties other than estrogenic activity,such as tyrosine kinase inhibition; however, the occurrenceof MOFs was not due to this property because another ty-rosine kinase inhibitor, lavendustin, did not have this effect(16). In addition, the primary mechanism by which estrogen
First Published Online April 19, 2007Abbreviations: DMSO, Dimethylsulfoxide; dpc, days postcoitum; ER,
estrogen receptor; Lfng, Lunatic fringe; MOF, multiple oocyte follicle;PND, postnatal day; STAT, signal transducer and activator oftranscription.Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.
elicits its action is through nuclear hormone receptors, es-trogen receptor (ER)-� and ER� (18, 19). To determinewhether the estrogenic activity of genistein was involved inthe formation of MOFs, mice lacking ER� or ER� weretreated neonatally with genistein. Mice lacking ER� did notdevelop MOFs, whereas mice lacking ER� did, further im-plicating genistein’s estrogenic activity and showing that theeffect on the ovary is mediated through ER� (16). Althougheffects of estrogenic compounds have been known for manyyears, the mechanism by which MOFs form is unknown (12,13). We hypothesize that MOFs are oocyte nests that did notcompletely break down during neonatal ovarian develop-ment. Our model is that normally, exposure of fetal oocytesto maternal estrogen keeps oocytes in nests and at birthestrogen levels drop resulting in nest breakdown. However,when neonatal oocytes are exposed to estrogens, nest break-down is inhibited.
Mechanisms involved in MOF formation were studied byexamining effects of neonatal genistein treatment on oocytedevelopment (20). At postnatal day (PND) 4, genistein-treated mice had significantly more unassembled oocytes,compared with control mice. Thus, genistein disrupted pri-mordial follicle assembly. In addition, genistein treatmentinhibited nest breakdown. These results support the idea thatMOFs observed in the genistein treated adult ovaries resultfrom incomplete breakdown of oocyte nests during neonataloocyte development. Genistein treatment also affected oo-cyte survival because significantly more oocytes were foundat PND 4 and 6 in treated mice.
Neonatal treatment of rodents with testosterone and pro-gesterone also results in increased MOFs (13, 21). However,progesterone and testosterone can be converted to estrogen,so it is not known whether these effects are direct or indirectdue to conversion to estrogen. Another recent study showedthat neonatal progesterone treatment reduced oocyte celldeath and primordial follicle assembly in rats supporting theprevious mouse studies (22).
We postulate that estrogen signaling plays a role in nestbreakdown based on the effects of the plant estrogen,genistein, on developing ovaries. MOFs have been observedin adult females that were treated as neonates with estradiolon PND 1–5 (13). Here we show that the levels of estrogenand other hormones influence nest breakdown but not oo-cyte death, providing insight into the mechanism by whichthese hormones induce MOFs.
Materials and MethodsAnimals
Adult CD-1 female mice were obtained from Charles River Labora-tories (Wilmington, MA) and bred to male mice of the same strain.Vaginal plug detection was considered d 0.5 of pregnancy. A group ofpregnant mice was killed at 16.5 d postcoitum (dpc) for the in vitro ovaryorgan culture experiments. The remainder of the pregnant mice deliv-ered pups at 19.5 dpc designated PND 1. Pregnant mice were housedunder controlled lighting (12 h light, 12 h dark) and temperature (21–22C) conditions. All animal procedures complied with an approved Na-tional Institute of Environmental Health Sciences/National Institutes ofHealth animal care protocol and the Syracuse University InstitutionalAnimal Care and Use Committee.
In vitro ovary organ culture
Ovaries were collected either at 16.5 dpc or PND 1 and placed intoculture. Ovaries were cultured in drops of media on 0.4 �m floatingfilters (Millicell-CM; Millipore Corp., Bedford, MA) in 0.4 ml DMEM-Ham’s F-12 media supplemented with penicillin-streptomycin, 5�ITS-X (Life Technologies, Inc., Grand Island, NY), 0.1% BSA, 0.1% al-bumax, and 0.05 mg/ml l-ascorbic acid in 4-well culture plates. Ovariesfrom an additional group of mice were collected at PND 1 and PND 8and not cultured for in vivo comparisons.
Chemicals used in the in vitro studies were as follows: estradiol(Sigma Chemical Co., St. Louis, MO); progesterone (Sigma); promege-stone (PerkinElmer, Waltham, MA); and genistein (Sigma). All chemi-cals were dissolved in dimethylsulfoxide (DMSO) at a concentration of0.1 m and then added to culture media to achieve the desired finalconcentration. DMSO was added to media at the same percent as com-pound treatment (�0.1%) to serve as a vehicle control.
Ovaries collected at 16.5 dpc were placed in culture and exposed dailyto estradiol at 10�6 m or DMSO alone as a vehicle control (n � 5–8 ovariesper treatment group). One group of ovaries collected at PND 1 wascultured for 1–7 d in media without hormones and collected after 2, 3,4, 5, 6, and 7 d of culture (n � 3–5 ovaries per time point). Another groupof ovaries collected at PND 1 was cultured for 7 d in the presence ofgenistein, estradiol, progesterone, or promegestone in media at concen-trations ranging from 10�9 to 10�4 m (n � 3–8 ovaries per treatmentgroup). Another group of ovaries collected at PND 1 was cultured for7 d in the presence of 10�6 m estradiol alone, 10�6 m progesterone alone,or both 10�6 m estradiol � 10�6 m progesterone (n � 6–8 ovaries pertreatment group). All ovaries were processed for whole-mountimmunohistochemistry.
In vivo estradiol treatment
Pregnant mice delivered pups at 19.5 d gestation (PND 1); pups wereseparated according to sex, pooled together, and then randomly stan-dardized to eight female pups per litter. Male pups were used in anotherexperiment. To examine the effects of hormone treatment at PND 4,female pups were treated on d 1–4 with estradiol by sc injection at a doseof 5 mg/kg�d in corn oil or left untreated as controls (eight mice pertreatment group). Ovaries were collected and processed for whole-mount immunohistochemistry on PND 4. To examine effects of hor-mones on occurrence of MOFs, another group of female pups wastreated on d 1–5 with estradiol at 5 mg/kg�d by sc injection in corn oilor left untreated as controls (eight mice per treatment group). Ovarieswere collected and processed for histological and morphological eval-uation on PND 19.
Histology and morphological evaluation
Ovaries collected at PND 19 after neonatal exposure to estradiol orleft untreated as controls were fixed in 10% neutral buffered formalin for6 h at 4 C, transferred to 70% ethyl alcohol, and embedded in paraffin.For each animal, three sections (6 �m) were prepared from both ovariesfrom different depths and stained with hematoxylin and eosin accordingto standard laboratory procedures (35). Ovaries from eight mice pertreatment group were analyzed for the presence of MOFs using lightmicroscopy. The presence of one MOF in a single ovarian section cat-egorized a mouse as positive for MOFs.
Whole-mount immunohistochemistry and fluorescencemicroscopy
Ovaries collected from in vivo and in vitro experiments were fixed in5% EM-grade paraformaldehyde (Ted Pella, Inc., Redding, CA) in PBSfor 1 h followed by several washes in 5% BSA and 0.1% Triton X-100 inPBS. Whole ovaries were immunostained as previously described (8, 23).The signal transducer and activator of transcription (STAT)-3 (C20)antibody (Santa Cruz Biotechnology, La Jolla, CA) was used at a dilutionof 1:500 (23). Propidium iodide or Toto-3 (Molecular Probes, now partof Invitrogen, Carlsbad, CA) was used to label nuclei. Samples wereimaged on a Zeiss Pascal Confocal microscope (Carl Zeiss MicroImag-ing, Inc., Thornwood, NY).
Chen et al. • Estrogen and Oocyte Development Endocrinology, August 2007, 148(8):3580–3590 3581
Analysis of oocyte nest breakdown, primordial follicleassembly, and follicle development
Whole ovaries were labeled with an antibody against STAT3, a spe-cific marker for germ cells (23). Ovaries were examined for percent singleoocytes relative to the total number of oocytes to assess oocyte nestbreakdown (8, 20). The number of individual oocytes relative to thenumber of oocytes in nests was determined by examining eight opticalregions per ovary. These regions were obtained by examining two areasof the ovary and taking four representative, confocal sections at least 20�m apart in each area. For each region, a single confocal section wasexamined. In addition, for each region, a stack of 10 sections, 1 �m apartcentered around the single section was obtained. This stack of sectionswas used to determine whether oocytes in the center section were as-sociated with oocyte nests above or below the plane of focus. For pri-mordial follicle assembly and development, the number of each type offollicle per region was determined. For primordial follicle assembly,oocytes were considered unassembled if granulosa cells did not com-pletely surround them or if STAT3 antibody labeling showed the oocyteswere associated. Follicles were classified as follows: primordial (oocytesurrounded by several granulosa cells with flattened nuclei), primary(oocyte surrounded by one layer of granulosa cells with cuboidal nuclei),or secondary (oocyte surrounded by more than one layer of granulosacells).
Determination of germ cell number
The number of oocytes per section was determined by counting thenumber of oocytes in the eight representative, confocal sections thatwere collected for analysis of cyst breakdown and determining theaverage number of oocytes per section.
Statistical analysis
One-way ANOVA was conducted to look at treatment effects onoocyte number, percent single oocytes, follicle assembly, and follicledevelopment. PROC GLM of SAS 9.1 (SAS Institute Inc., Cary, NC) wasused to calculate the least-squares means and test specific hypotheses foreffects. P � 0.05 was considered significant.
ResultsCharacterization of neonatal oocyte development in an invitro organ culture system
We characterized nest breakdown, oocyte survival, andprimordial follicle development in ovary organ culture usingovaries from newborn CD-1 female mice. Before culture(PND 1 or d 0 of culture) and on d 1–7 of culture, ovaries werefixed and labeled for the oocyte marker STAT3 (23). Thenumber of oocytes per section, percent of single oocytes, andrelative numbers of different stage follicles were determined.The number of oocytes per section dropped significantlyafter 1 d of culture (Fig. 1A) similar to nest breakdown in vivo(8). In addition, by d 2, most nests had broken apart (greaterthan 60%) as in vivo (Fig. 1A). Next, we determined thenumber of unassembled oocytes and different stages of de-veloping follicles. Before culture, the majority of oocyteswere not enclosed in follicles (over 80%, Fig. 1B) but after 7 din organ culture, most oocytes were in follicles (almost 90%),and some had started growing as evidenced by the appear-ance of primary and secondary follicles. We also examinedfollicle development in vivo at PND 8, which would be com-parable with 7 d of culture and found that as in culture themajority of oocytes were enclosed in follicles and some hadstarted progressing to the primary and secondary stage (Fig.1C). However, there were significantly fewer secondary fol-licles and more primary follicles in vitro, compared with in
vivo. This difference may reflect the lack of some factor in theorgan culture media that is necessary for optimal follicledevelopment. Finally, during the culture period the ovariesgrew in size. Thus, early events of nest breakdown, primor-dial follicle assembly, and follicle growth proceed similar toin vivo in our organ culture system, but the transition of thefirst wave of developing follicles from primary to secondaryfollicles is delayed.
Effects of genistein on oocyte development in ovaryorgan culture
We have previously shown that treatment of neonatal micewith genistein results in inhibition of nest breakdown and an
FIG. 1. Oocyte development and primordial follicle assembly in organculture. A, Numbers of oocytes per confocal section and percent singleoocytes in newborn ovaries before and after 1–7 d of organ culture. B,Primordial follicle assembly and development in newborn ovaries andafter 7 d of organ culture. C, Primordial follicle development in new-born ovaries after 7 d of organ culture, compared with 7 d in vivo (PND8). Data are presented as the mean � SEM. *, Significant differencebetween percentage of follicles at the same stage of developmentbefore, after 7 d of culture, or in vivo (one-way ANOVA, P � 0.05; n �3–5 ovaries per group).
3582 Endocrinology, August 2007, 148(8):3580–3590 Chen et al. • Estrogen and Oocyte Development
increase in oocyte survival (20). Therefore, in the currentstudy, we determined whether genistein would have thesame effect on ovaries in organ culture as it did in vivo.Ovaries were collected on PND 1 and placed in culture for7 d. Ovaries were treated daily with concentrations from 10�9
to 10�4 m genistein, and nest breakdown, oocyte survival,and follicle development were analyzed after whole-mountimmunohistochemistry. Control ovaries had approximately90% single oocytes, whereas genistein-treated ovaries hadfewer single oocytes (60–70%) at concentrations of 10�8 mand higher (Fig. 2A); thus, as in vivo, nest breakdown wasinhibited. Examples of a cultured control ovary and an ovarytreated with 10�6 m genistein are shown in Fig. 2, D and E,illustrating the large nests that are present in the treatedovaries. Ovaries treated with the highest concentration ofgenistein, 10�4 m, did not grow during the culture period andthe structure of the ovary appeared disorganized, suggestingthat the tissue did not survive the culture period. There wasno significant difference between the number of oocytes persection in control and treated ovaries (Fig. 2B). We also ex-amined follicle assembly and found an increase in the num-ber of unassembled oocytes from 13% to more than 30% (Fig.2C). Thus, like its effect in vivo, genistein inhibited nest break-down and follicle assembly in organ culture. However, oo-cyte survival was not altered in vitro.
Estradiol alters ovarian differentiation in vivo and in vitro
To confirm that the alterations in ovarian differentiationare due to estrogen activity, the natural mammalian hor-
mone, estradiol was studied in vivo as well as in vitro. Neo-natal mice were treated on d 1–4 with 5 mg/kg�d estradiolor vehicle alone, and nest breakdown, oocyte survival, andfollicle development were measured after whole-mount im-munohistochemistry. We found that the percent of singleoocytes was reduced from 74% in control animals to 46% intreated animals at PND4 (Fig. 3A), whereas oocyte numberwas not affected (Fig. 3B). Primordial follicle assembly wasalso affected by estradiol treatment of neonates with treatedanimals having 54% unassembled oocytes, compared withcontrol animals, with only 26% unassembled follicles (Fig.3C). Thus, estradiol treatment perturbs neonatal ovarian de-velopment in vivo.
To determine whether this effect persisted into puberty,mice were again treated with 5 mg/kg�d estradiol on d 1–5.Ovaries from mice treated with estradiol exhibited an in-creased incidence of MOFs at PND 19, compared with controlmice (five of eight mice exhibited MOFs in the treated groupvs. none of eight in the control), similar to our previousfinding of increased MOFs after neonatal genistein treatment(20).
To establish a direct effect of estradiol on ovarian differ-entiation, ovaries were again collected from CD-1 mice onPND 1 and placed in culture for 7 d. Ovaries were treateddaily with concentrations from 10�9 to 10�4 m estradiol, andnest breakdown, oocyte survival, and follicle developmentwere analyzed after whole-mount immunohistochemistry.Control ovaries had approximately 90% single oocytes,whereas estradiol-treated ovaries had fewer single oocytes
FIG. 2. Genistein treatment inhibits nest breakdown and primordial follicle assembly in organ culture. A, Percent single oocytes in newbornovaries after 7 d in culture with increasing concentrations of genistein. B, Number of oocytes per confocal section in newborn ovaries after 7 din culture with increasing concentrations of genistein. C, Percent unassembled, primordial, primary, and secondary follicles in newborn ovariesafter 7 d in culture with increasing concentrations of genistein. Data are presented as the mean � SEM. Different letter superscripts representa significant difference between control and genistein-treated ovaries (one-way ANOVA, P � 0.05; n � 4–7 ovaries per group). D, Confocal sectionof a control ovary grown in organ culture for 7 d labeled with STAT3 antibody to visualize oocytes (green) and propidium iodide to visualizenuclei (red). E, Confocal section of an ovary cultured in 10�6 M genistein for 7 d labeled with STAT3 antibody (green) and propidium iodide (red).Scale bar, 10 �m.
Chen et al. • Estrogen and Oocyte Development Endocrinology, August 2007, 148(8):3580–3590 3583
(60–70%), illustrating that estradiol inhibits nest breakdown(Fig. 4A). An example of the effects of treatment with 10�6
m estradiol is shown in Fig. 4, D and E. Strikingly, at thehighest concentration of estradiol (10�4 m), there was furtherreduction in nest breakdown (30% single oocytes). Althoughestradiol slightly decreased germ cell number with increas-ing concentration, there was no significant difference be-tween control and treated ovaries except at the highest es-tradiol concentration (Fig. 4B). We also examined follicleassembly and found that even at the lowest dose (10�9 m), thenumber of oocytes not enclosed in follicles increased from 13to between 30 and 40% and was more than 70% at the highest
concentration (Fig. 4C). Thus, the natural estrogen, estradiol,inhibits nest breakdown and primordial follicle formation,supporting our model of ovarian differentiation that the dropof estrogen at birth causes oocyte nests to disassemble intosingle oocytes.
Ovaries were collected from CD-1 mice on PND 1, placedin culture, and treated daily as control or with concentrationsfrom 10�9 to 10�4 m progesterone. Ovaries were collectedafter 7 d of culture, and nest breakdown, oocyte survival, andfollicle development were analyzed after whole-mount im-munohistochemistry. Control ovaries had 90% single oo-cytes, whereas progesterone-treated ovaries had fewer singleoocytes (70%) (Fig. 5A). Examples of cultured control (Fig.5D) and ovaries treated with 10�6 m progesterone (Fig. 5E)are shown to illustrate the large nests that are present in thetreated ovaries. Like the genistein-treated ovaries, ovariestreated with the highest concentration of progesterone (10�4
m) did not grow and appeared disorganized. There was nosignificant difference in the number of oocytes between con-trol and treated ovaries (Fig. 5B). We also examined follicleassembly and found that primordial follicle assembly anddevelopment were altered (Fig. 5C). Like estradiol andgenistein, progesterone inhibited nest breakdown and folli-cle assembly in organ culture but did not affect oocytesurvival.
Because progesterone can be converted to estrogen, itseffect could be exerted either directly or via conversion toestrogen. To test whether progesterone can directly interferewith oocyte development, we treated cultured ovaries witha nonmetabolizable version of progesterone, promegestonefor 7 d. Promegestone, like progesterone, inhibited nestbreakdown and follicle assembly but did not affect oocytenumber (Fig. 6). Thus, progesterone has a direct effect in-dependent of estrogen on oocyte development.
Both estradiol and progesterone inhibited nest breakdownand primordial follicle assembly in ovary organ culture. Totest for additive effects of estradiol and progesterone, wetreated cultured PND 1 ovaries with 10�6 m estradiol only,10�6 m progesterone only, or both 10�6 m estradiol and10�6 m progesterone daily for 7 d. Control ovaries had ap-proximately 86% single oocytes (Fig. 7A). Progesterone-treated ovaries had fewer single oocytes (67%), estradiol-treated ovaries had still fewer single oocytes (53%), andovaries treated with both hormones had the lowest percentof single oocytes (34%). There was no significant differencebetween the number of oocytes per section in control andtreated ovaries (Fig. 7B). Like nest breakdown, follicle as-sembly was inhibited further when ovaries were treated withboth hormones (control, 14% unassembled; progesteronetreated, 33%; estradiol, 47%; both hormones, 66%) (Fig. 7C).Thus, progesterone and estradiol inhibited nest breakdownand follicle assembly in an additive manner.
Prenatal organ culture triggers premature oocytedevelopment that is rescued by estradiol treatment
Ovaries from 16.5 dpc fetuses were grown in organ culturein the absence of maternal hormones for 3 d, which would
FIG. 3. In vivo estradiol treatment inhibits nest breakdown and pri-mordial follicle assembly. Neonatal mice were injected with 5 mg/kg�destradiol or left untreated as controls. A, Percentage of single oocytesis plotted at PND 4 in control and estradiol (E2)-treated mice. B,Number of oocytes per section at PND 4 in control and E2-treatedmice. C, Percent of unassembled, primordial, primary, and secondaryfollicles at PND 4 in control and estradiol-treated mice. Data arepresented as the mean � SEM. *, Significant difference between con-trol and estradiol-treated ovaries (one-way ANOVA, P � 0.05; n � 6–7ovaries per group).
3584 Endocrinology, August 2007, 148(8):3580–3590 Chen et al. • Estrogen and Oocyte Development
be equivalent to the day of birth. We found that the majorityof nests had broken down (63% single oocytes) in fetal ova-ries cultured for 3 d as compared with the equivalent in vivotime point of PND 1 (19.5 dpc) (only 16% single oocytes) (Fig.8A). We also found that the number of oocytes was reducedto 28 per section, compared with 39 in vivo (Fig. 8B) andunassembled oocytes were reduced from 84 to 37% (Fig. 8C).In vivo, ovaries at PND 1 had some primordial follicles (16%)and no primary or secondary follicles. Cultured ovaries in theabsence of maternal hormones had significantly more pri-mordial follicles (57%) and some primary follicles (6%).
To further test the effects of hormone signaling on nestbreakdown, estradiol, progesterone, or both hormones wereadded to the culture media daily for 3 d and nest breakdownrate and oocyte survival were compared with the ovariesgrown without added hormone and to PND 1 ovaries. Inestradiol-treated ovaries, nest breakdown was reduced withonly 18% single oocytes, which is comparable with PND 1 invivo (Fig. 8A). The number of oocytes was increased from 28to 34 oocytes per section, although this is still lower than PND1 in vivo (39 oocytes per section) (Fig. 8B). Most oocytes wereunassembled (82%) and the number of all follicle types wasreduced (17% primordial, 1% primary, and no secondaryfollicles), compared with ovaries cultured in media alone.Progesterone partially reduced nest breakdown, comparedwith ovaries cultured in medium without added hormone(29% single oocytes, compared with 63% in the culturedcontrol ovaries) but did not completely rescue to in vivo
numbers (16% single oocytes). Premature follicle formationwas also reduced, resulting in a higher percentage of unas-sembled oocytes (71%, compared with 37% in the in vitrocontrol) and did not increase the number of oocytes (Fig. 8,A and B). When estradiol and progesterone were both added,oocyte development was similar to added estradiol alone.Thus, premature oocyte development including nest break-down is triggered by removal of the ovaries from the ma-ternal hormonal environment, and this premature ovariandifferentiation can be partially rescued by addition of estra-diol or progesterone to organ culture supporting our modelof ovarian differentiation.
Discussion
It is a challenge to control the maternal environment of thefetus and to directly observe oocyte development in the fetaland neonatal ovary. To facilitate an analysis of hormonaleffects on fetal and neonatal oocyte development, we ex-plored an in vitro organ culture system that has previouslybeen used by others (22, 24–26). This permitted us to directlyvisualize the effect of hormone levels on nest breakdown andprimordial follicle formation. Previous work showed thattreatment of neonates with estrogenic compounds caused anincrease in MOFs in the adult ovary (12–16). Genistein treat-ment inhibits nest breakdown supporting the idea that MOFsare nests that did not break down (20). Our results indicate
FIG. 4. Estradiol treatment inhibits nest breakdown and follicle assembly in organ culture. A, Percent single oocytes in newborn ovaries after7 d in culture with increasing concentrations of estradiol. B, Number of oocytes per confocal section in newborn ovaries after 7 d in culture withincreasing concentrations of estradiol. C, Percent unassembled, primordial, primary, and secondary follicles in newborn ovaries after 7 d inculture with increasing concentrations of estradiol. Data are presented as the mean � SEM. Different letter superscripts represent a significantdifference (one-way ANOVA, P � 0.05; n � 3–7 ovaries per group). D, Confocal section of a control ovary grown in organ culture for 7 d labeledwith STAT3 antibody to visualize oocytes (green) and propidium iodide to visualize nuclei (red). E, Confocal section of an ovary cultured in 10�5
M estradiol for 7 d labeled with STAT3 antibody (green) and propidium iodide (red). Scale bar, 10 �m.
Chen et al. • Estrogen and Oocyte Development Endocrinology, August 2007, 148(8):3580–3590 3585
that genistein, estradiol, and progesterone inhibit nest break-down and primordial follicle assembly in organ culture.
Oocytes from prenatal ovaries grown in culture begin nestbreakdown prematurely, and this is partially prevented byaddition of estradiol or progesterone to the culture medium.These findings support our current model of the role ofestrogen in regulation of nest breakdown and primordialfollicle assembly. According to this model, before birth, ma-ternal hormone levels are high in the fetus, generating asignal that inhibits nest breakdown and oocyte apoptosis(20). At birth, estrogen and progesterone levels drop remov-ing the inhibitory signal and nests begin to break apart. Wedo not understand why it is necessary for nests to break apartat birth. It may be that if this process occurs earlier, theoocytes are not yet in the right stage of meiosis. In a recentstudy, inhibition of synaptonemal complex protein-1 causedpremature arrival in the diplotene stage of meiosis and ac-celeration of primordial follicle assembly, suggesting a linkbetween cell cycle stage and primordial follicle development(28).
Recently progesterone and estrogen were found to affectneonatal oocyte development in rats (22). Neonatal proges-terone treatment reduced primordial follicle assembly,whereas both progesterone and estrogen treatment reducedthe primordial to primary follicle transition in the initialwave of follicles that begin to develop. However, nest break-down was not examined, only whether primordial follicleshad formed. Progesterone also reduced the amount of pro-gramed cell death in the neonatal ovary but estrogen was not
tested. In our studies, both progesterone and estrogen in-hibited nest breakdown and primordial follicle assembly butnot subsequent follicle development. In addition, neitherhormone had an effect on oocyte survival in organ culture.One possibility for the observed differences between the twostudies is that certain aspects of hormonal effects on oocytedevelopment differ between rats and mice. Another possi-bility is that different methods of analysis and counting re-sulted in the apparent differences. One should note here thatpublished counts of oocyte number in developing mouseovaries vary widely, and care should be taken when com-paring studies that use different methods (8, 29).
There are several possible ways that abnormal MOFsfound in adult ovaries from mice that were treated withestrogenic compounds could form. Granulosa cells couldimproperly enclose more than one oocyte in a follicle oroocyte nest breakdown could be disrupted. Alternatively,two follicles could fuse. Studies with genistein treatmentshow there are more unassembled oocytes after genisteintreatment and more oocytes still in nests. In addition, inter-cellular bridges still connected oocytes in genistein-treatedmice, whereas none were apparent in control mice at PND 4(20), suggesting that MOFs in adult ovaries result from in-complete breakdown of oocyte nests during neonatal devel-opment leaving pregranulosa cells with multiple oocytes tosurround. Our results here show that estradiol and proges-terone disrupt nest breakdown, further supporting the ideathat MOFs in adult ovaries are oocyte nests that did not breakapart.
FIG. 5. Progesterone inhibits nest breakdown and primordial follicle assembly. A, Percent single oocytes in newborn ovaries after 7 d in culturewith increasing concentrations of progesterone. B, Number of oocytes per confocal section in newborn ovaries after 7 d in culture with increasingconcentrations of progesterone. C, Percent unassembled, primordial, primary, and secondary follicles in newborn ovaries after 7 d in culturewith increasing concentrations of progesterone. Data are presented as the mean � SEM. Different letter superscripts represent a significantdifference between different treatment groups (ANOVA, P � 0.05; n � 6–8 ovaries per group). D, Confocal section of a control ovary grownin organ culture for 7 d labeled with STAT3 antibody to visualize oocytes (green) and propidium iodide to visualize nuclei (red). E, Confocalsection of an ovary cultured in 10�6 M progesterone for 7 d labeled with STAT3 antibody (green) and propidium iodide (red). Scale bar, 10 �m.
3586 Endocrinology, August 2007, 148(8):3580–3590 Chen et al. • Estrogen and Oocyte Development
Programmed cell death of oocytes occurs at the time ofoocyte nest breakdown, but the relationship between thesetwo processes is not known (8). Currently it is thought that
programmed cell death is required for nests to break apart.Supporting this idea, mutants in the cell death regulator,Bcl-2-associated X protein, have more oocytes and a delay innest breakdown (30). In addition to inhibiting nest break-down, neonatal genistein treatment increased oocyte sur-vival in vivo and reduced apoptosis (20). However, here inorgan culture, genistein, estradiol, or progesterone treatmentinhibited nest breakdown, but most concentrations of hor-mone had no effect on oocyte death. At the highest level ofestradiol (10�4 m), there was an increase in oocytes surviving,whereas somatic cells appeared to be dying (data not shown).These responses may be due to which receptors are ex-
FIG. 6. A nonhydrolyzable form of progesterone, promegestone (NH-P4), inhibits nest breakdown and primordial follicle assembly. A,Percent single oocytes in newborn ovaries after 7 d of culture with noadded hormone or treated with 10�6 M promegestone. B, Number ofoocytes per confocal section in newborn ovaries after 7 d of culturewith no added hormone or treated with 10�6 M promegestone. C,Percent unassembled, primordial, primary, and secondary follicles innewborn ovaries after 7 d of culture with no added hormone or treatedwith 10�6 M promegestone. Data are presented as the mean � SEM.*, Significant difference between control and treated ovaries(ANOVA, P � 0.05; n � 4–5 ovaries per group).
FIG. 7. Estradiol and progesterone inhibit nest breakdown and pri-mordial follicle assembly. A, Percent single oocytes in newborn ova-ries after 7 d in culture with no hormone, 10�6 M progesterone (P4)alone, 10�6 M estradiol (E2) alone, or both estradiol and progesterone(E2 � P4) added. B, Number of oocytes per confocal section in newbornovaries after 7 d in culture with no hormone, 10�6 M progesterone (P4)alone, 10�6 M estradiol (E2) alone, or both estradiol and progesterone(E2 � P4) added. C, Percent unassembled, primordial, primary, andsecondary follicles in newborn ovaries after 7 d in culture with nohormone, 10�6 M progesterone (P4) alone, 10�6 M estradiol (E2) alone,or both estradiol and progesterone (E2 � P4) added. Data are pre-sented as the mean � SEM. Different letter superscripts represent asignificant difference between different treatment groups (ANOVA,P � 0.05; n � 5–6 ovaries per group).
Chen et al. • Estrogen and Oocyte Development Endocrinology, August 2007, 148(8):3580–3590 3587
pressed in different cells types. This level of estradiol mayhave a toxic effect on the ovary and somatic cells are just moresensitive to estradiol than oocytes. The oocytes may havesome protective mechanism so that they can survive an oth-erwise toxic insult. Ovaries did not survive the organ cultureperiod when treated with 10�4 m genistein or progesterone,again suggesting that extremely high levels of hormoneshave a general toxic effect on the ovary. Interestingly, in vivogenistein injections protected more oocytes from dying,whereas there was no change in oocyte number in organculture. This may reflect differences in the mechanism of
action of genistein, depending on how the ovary is receivingthe hormone. However, neither estradiol injections in vivonor estradiol in culture altered the number of oocytes sur-viving through the nest breakdown period. Based on count-ing numbers of oocytes, estradiol does not influence oocytesurvival. This suggests that programmed cell death is not thedriving force causing nests to break apart or that the pro-cesses of nest breakdown and oocyte cell death can be sep-arated at some level.
There are two mammalian estrogen receptors, ER� andER�, with multiple roles in female reproduction (31). MOFsresult from genistein treatment of ER� mutant mice but notER� mutant mice, suggesting that effects of genistein aremediated primarily through ER� (16). However, genisteinhas higher affinity for ER�, whereas estradiol has equal af-finity for ER� or ER� (32). Using neonatal ovary culture andin vivo injections, we find that estradiol, like genistein, in-hibits nest breakdown and primordial follicle assembly. Al-though it is likely that maternal estradiol maintains oocytesin nests through ER�, there are possible alternative mecha-nisms. Estradiol may exert its effects through ER� or bothER� and ER�. Another possibility is that estradiol may notexert its effects through either ER. Recently steroid hormoneshave been found to mediate effects through membranebound receptors (33). ER� knockout mice have reduced fer-tility with fewer overall litters and fewer pups per litter (19).Mutant adult ovaries have more atretic follicles and fewercorpora lutea, suggesting that more oocytes may be dying.However, immature ER� knockout mice (PND 28–30) do nothave fewer primordial follicles than wild-type mice (34). ER�mutants are sterile with ovaries lacking corpora lutea andcontaining cystic and hemorrhagic follicles (35). Nest break-down has not been examined in either ER mutant.
Several genes have mutant phenotypes, suggesting po-tential involvement in nest breakdown. For example, micelacking bone morphogenetic factor 15 or growth differenti-ation factor 9 have more MOFs than wild-type mice as wellas other defects of ovarian differentiation (36). The increasednumber of MOFs in these mutant animals suggests that theseproteins function in nest breakdown. Both proteins are mem-bers of the TGF� superfamily; they are oocyte-secreted fac-tors expressed early in ovarian differentiation (37). AnotherTGF� family member, activin, has also been implicated inearly oocyte differentiation. Interfering with the activin sig-naling pathway results in abnormal ovarian developmentincluding ovaries with MOFs (38, 39). Furthermore, treat-ment of neonatal mice with estradiol or diethylstilbestrolreduces activin mRNA and protein levels, supporting theidea that activin is a target of estrogen signaling (40). Lunaticfringe (Lfng) mutants are infertile and also have MOFS (41).Lfng is a member of the fringe family of proteins that functionby either stimulating or inhibiting Notch signaling (42, 43).Lfng is expressed in granulosa and theca cells of developingfollicles (41). Notch signaling may play a role in nest break-down or follicle assembly. Interestingly, in Drosophila,fringe mutants have follicles with more than one oocyte (44).Estrogen may affect nest breakdown by regulating some ofthese genes. Estrogen may also up-regulate cell adhesiongenes that would keep oocytes in nests. In addition, somaticcells need to migrate around oocytes to form follicles, and
FIG. 8. Prenatal cultured ovaries undergo nest breakdown and pri-mordial follicle assembly that is inhibited by estradiol and proges-terone treatment. A, Percentage of single oocytes is plotted in controlPND 1 ovaries (in vivo) and ovaries dissected at 16.5 dpc and grownin culture for 3 d without hormone added (in vitro control) or with 10�6
M estradiol (in vitro E2), 10�6 M progesterone (in vitro P4), or both 10�6
M estradiol and 10�6 M progesterone (in vitro E2 � P4) added. B,Number of oocytes per section in control PND 1 ovaries and 16.5 dpcovaries grown in culture with and without hormone added as in A. C,Percent of unassembled, primordial, and primary follicles in controlPND 1 ovaries and 16.5 dpc ovaries grown in culture with and withouthormone added as in A. Data are presented as the mean � SEM.Different letter superscripts represent a significant difference betweendifferent treatment groups (one-way ANOVA, P � 0.001; n � 5–8ovaries per group).
3588 Endocrinology, August 2007, 148(8):3580–3590 Chen et al. • Estrogen and Oocyte Development
estrogen signaling may inhibit genes involved in thismigration.
It remains unclear what cell types respond to estrogenduring ovarian differentiation. Estrogen could signal directlyto the oocytes to regulate nest breakdown or to somatic cellsthat could in turn signal the oocytes. For example, culturedprimordial germ cells are stimulated to proliferate by treat-ment with estrogen but only in the presence of somatic go-nadal cells, suggesting that estrogen can mediate this effecton germ cells indirectly by signaling through somatic cells(45). Using immunocytochemistry, we detect ER� in oocytesat birth (Pepling, M. E., unpublished data). However, ER� isexpressed in granulosa cells by PND 5, continuing into adult-hood, and has been detected as early as PND 1 by RNaseprotection in whole ovaries (46). ER� is expressed in thecacells in adult ovary and is also detected in neonatal ovaries(46). Cell-specific knockouts of estrogen receptors will berequired to determine which receptor and which cells typesare required for oocyte development.
Our model of ovarian differentiation is that in the devel-oping mouse, maternal estrogen and progesterone keep oo-cytes in nests and that after birth the drop in these hormonescauses nests to break apart and follicles to assemble. How-ever, in humans, follicle formation occurs at about 4.5 monthsgestation, not at birth (47–49). It may be that humans do notuse the same mechanism as rodents. However, another pos-sibility is that even though total progesterone and estrogenlevels are high late in human pregnancy, the amount thatfetal tissues are exposed to is reduced at this time. Thisinterpretation is supported by a study in monkeys, whichshowed progesterone was reduced in fetal tissues during latepregnancy (50). It is possible that steroid hormone bindingproteins such as �-fetoprotein, which binds estrogen, couldsequester hormones causing the levels in ovarian tissue todrop (27).
We have shown that estradiol, progesterone, and genisteindisrupt nest breakdown and primordial follicle formation.The mechanism by which these hormones act on the ovaryis still not known. Investigation into how these signals arereceived and what cell types receive these signals is ongoing.The basic mechanisms underlying normal primordial follicleformation as well as how disorders disrupt these normalprocesses are not well understood. Elucidation of the pro-cesses involved in establishment of the primordial folliclepool will provide a better understanding of oocyte devel-opment. This understanding will give insight into prematureovarian failure, reproductive life span, menopause, andovarian cancer and contribute to potential treatments of fe-male infertility. In addition, this knowledge will aid in theunderstanding of toxic effects of exogenous estrogenexposure.
Acknowledgments
We thank Brian Calvi and Grant Gephardt for helpful comments onthe manuscript. We also thank Zhongan Chen for technical assistanceand especially thank Tom Starmer for assistance with statistical analysis.
Received January 22, 2007. Accepted April 11, 2007.Address all correspondence and requests for reprints to: Melissa
Pepling, Department of Biology, Syracuse University, 130 College Place,Syracuse, New York 13244. E-mail: [email protected].
This work was supported by the Intramural Research Program of theNational Institutes of Health, National Institute of EnvironmentalHealth Sciences and National Science Foundation Grant IOB-0613895 (toM.E.P.).
Disclosure Statement: The authors have nothing to disclose.
References
1. Monk M, McLaren A 1981 X-chromosome activity in foetal germ cells of themouse. J Embryol Exp Morphol 63:75–84
2. Pepling ME, Spradling AC 1998 Female mouse germ cells form synchronouslydividing cysts. Development 125:3323–3328
3. Borum K 1961 Oogenesis in the mouse. A study of the origin of the mature ova.Exp Cell Res 45:39–47
4. Johnson J, Bagley J, Skaznik-Wikiel M, Lee HJ, Adams GB, Niikura Y,Tschudy KS, Tilly JC, Cortes ML, Forkert R, Spitzer T, Iacomini J, ScaddenDT, Tilly JL 2005 Oocyte generation in adult mammalian ovaries by putativegerm cells in bone marrow and peripheral blood. Cell 122:303–315
5. Johnson J, Canning J, Kaneko T, Pru JK, Tilly JL 2004 Germline stem cells andfollicular renewal in the postnatal mammalian ovary. Nature 428:145–150
1993 Evidence that the mechanism of prenatal germ cell death in the mouseis apoptosis. Exp Cell Res 209:238–247
8. Pepling ME, Spradling AC 2001 The mouse ovary contains germ cell cysts thatundergo programmed breakdown to form follicles. Dev Biol 234:339–351
9. Ratts VS, Flaws JA, Klop R, Sorenson CM, Tilly JL 1995 Ablation of bcl-2 geneexpression decreases the number of oocytes and primordial follicles estab-lished in the postnatal female mouse gonad. Endocrinology 136:3665–3668
10. Iguchi T, Watanabe H, Katsu Y 2001 Developmental effects of estrogenicagents on mice, fish, and frogs: a mini-review. Horm Beh 40:248–251
11. Kent HA 1960 Polyovular follicles and multinucleate ova in the ovaries ofyoung mice. Anat Rec 137:521–524
12. Iguchi T, Takasugi N 1986 Polyovular follicles in the ovary of immature miceexposed prenatally to diethylstilbestrol. Anat Embryol (Berl) 175:53–55
13. Iguchi T, Takasugi N, Bern HA, Mills KT 1986 Frequent occurrence ofpolyovular follicles in ovaries of mice exposed neonatally to diethylstilbestrol.Teratology 34:29–35
14. Iguchi T, Fukazawa Y, Uesugi Y, Taksugi N 1990 Polyovular follicles in mouseovaries exposed neonatally to diethylstibestrol in vivo and in vitro. Biol Reprod43:478–484
15. Suzuki A, Sugihara A, Uchida K, Sato T, Ohta Y, Katsu Y, Watanabe H,Iguchi T 2002 Developmental effects of perinatal exposure to bisphenol-A anddiiethylstilbestrol on reproductive organs in female mice. Reprod Toxicol16:107–116
16. Jefferson WN, Couse JF, Padilla-Banks E, Korach KS, Newbold RR 2002Neonatal exposure to genistein induces estrogen receptor (ER)� expressionand multioocyte follicles in the maturing mouse ovary: evidence for ER�-mediated and nonestrogenic actions. Biol Reprod 67:1285–1296
17. Gougeon A 1981 Frequent occurrence of multiovular follicles and multinuclearoocytes in the adult human ovary. Fertil Steril 35:417–422
18. Couse JF, Korach KS 1999 Estrogen receptor null mice: what have we learnedand where will they lead us? Endocr Rev 20:358–417
19. Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler JF, Sar M,Korach KS, Gustafsson JA, Smithies O 1998 Generation and reproductivephenotypes of mice lacking estrogen receptor �. Proc Natl Acad Sci USA95:15677–15682
20. Jefferson W, Newbold R, Padilla-Banks E, Pepling M 2006 Neonatal genisteintreatment alters ovarian differentiation in the mouse: inhibition of oocyte nestbreakdown and increased oocyte survival. Biol Reprod 74:161–168
21. Iguchi T, Todoroki R, Takasugi N, Petrow V 1988 The effects of an aromataseinhibitor and a 5�-reductase inhibitor upon the occurrence of polyovularfollicles, persistent anovulation, and permanent vaginal stratification in micetreated neonatally with testosterone. Biol Reprod 39:689–697
22. Kezele P, Skinner MK 2003 Regulation of ovarian primordial follicle assemblyand development by estrogen and progesterone: endocrine model of follicleassembly. Endocrinology 144:3329–3337
23. Murphy K, Carvajal L, Medico L, Pepling ME 2005 Expression of Stat3 in germcells of developing and adult mouse ovaries and testes. Gene Expr Patterns5:475–482
24. Buehr M, McLaren A 1993 Isolation and culture of primordial germ cells.Methods Enzymol 225:58–77
25. Devine PJ, Rajapaksa KS, Hoyer PB 2002 In vitro ovarian tissue and organculture: a review. Front Biosci 7:d1979–d1989
26. Parrott JA, Skinner MK 1999 Kit-ligand/stem cell factor induces primordialfollicle development and initiates folliculogenesis. Endocrinology 140:4262–4271
27. Mizejewski GJ 2004 Biological roles of �-fetoprotein during pregnancy andperinatal development. Exp Biol Med (Maywood) 229:439–463
28. Paredes A, Garcia-Rudaz C, Kerr B, Tapia V, Dissen GA, Costa ME, CorneaA, Ojeda SR 2005 Loss of synaptonemal complex protein-1, a synaptonemal
Chen et al. • Estrogen and Oocyte Development Endocrinology, August 2007, 148(8):3580–3590 3589
complex protein, contributes to the initiation of follicular assembly in thedeveloping rat ovary. Endocrinology 146:5267–5277
29. Tilly JL 2003 Ovarian follicle counts—not as simple as 1, 2, 3. Reprod BiolEndocrinol 1:11
30. Greenfeld CR, Pepling ME, Babus JK, Furth PA, Flaws JA, BAX is involvedin regulating follicular endowment in mice. Reproduction, in press
31. Britt KL, Findlay JK 2002 Estrogen actions in the ovary revisited. J Endocrinol175:269–276
32. Kuiper GG, Lemmen JG, Carlsson B, Corton JC, Safe SH, van der Saag PT,van der Burg B, Gustafsson JA 1998 Interaction of estrogenic chemicals andphytoestrogens with estrogen receptor �. Endocrinology 139:4252–4263
33. Norman AW, Mizwicki MT, Norman DP 2004 Steroid-hormone rapid actions,membrane receptors and a conformational ensemble model. Nat Rev DrugDiscov 3:27–41
34. Emmen JM, Couse JF, Elmore SA, Yates MM, Kissling GE, Korach KS 2005In vitro growth and ovulation of follicles from ovaries of estrogen receptor(ER)� and ER� null mice indicate a role for ER� in follicular maturation.Endocrinology 146:2817–2826
35. Newbold RR, Bullock BC, McLachlan JA 1983 Exposure to diethylstilbestrolduring pregnancy permanently alters the ovary and oviduct. Biol Reprod28:735–744
36. Yan C, Wang P, DeMayo J, DeMayo FJ, Elvin JA, Carino C, Prasad SV,Skinner SS, Dunbar BS, Dube JL, Celeste AJ, Matzuk MM 2001 Synergisticroles of bone morphogenetic protein 15 and growth differentiation factor 9 inovarian function. Mol Endocrinol 15:854–866
37. Elvin JA, Yan C, Matzuk MM 2000 Oocyte-expressed TGF-� superfamilymembers in female fertility. Mol Cell Endocrinol 159:1–5
38. Bristol-Gould SK, Hutten CG, Sturgis C, Kilen SM, Mayo KE, Woodruff TK2005 The development of a mouse model of ovarian endosalpingiosis. Endo-crinology 146:5228–5236
39. McMullen ML, Cho BN, Yates CJ, Mayo KE 2001 Gonadal pathologies in
transgenic mice expressing the rat inhibin �-subunit. Endocrinology 142:5005–5014
40. Kipp JL, Kilen SM, Bristol-Gould S, Woodruff TK, Mayo KE 2007 Neonatalexposure to estrogens suppresses activin expression and signaling in themouse ovary. Endocrinology 148:1968–1976
41. Hahn KL, Johnson J, Beres BJ, Howard S, Wilson-Rawls J 2005 Lunatic fringenull female mice are infertile due to defects in meiotic maturation. Develop-ment 132:817–828
46. Jefferson WN, Couse JF, Banks EP, Korach KS, Newbold RR 2000 Expressionof estrogen receptor � is developmentally regulated in reproductive tissues ofmale and female mice. Biol Reprod 62:310–317
47. Gillman J 1948 The development of the gonads in man, with consideration ofthe role of fetal endocrines and the histogenesis of ovarian tumors. ContribEmbryol 32:81–131
48. Gondos B, Bhiraleus P, Hobel CJ 1971 Ultrastructural observations on germcells in human fetal ovaries. Am J Obstet Gynecol 110:644–652
49. Witschi E 1948 Migration of the germ cells of human embryos from the yolksac to the primitive gonadal folds. Contrib Embryol 32:67–80
50. Thau R, Lanman JT, Brinson A 1976 Declining plasma progesterone concen-tration with advancing gestation in blood from umbilical and uterine veins andfetal heart in monkeys. Biol Reprod 14:507–509
Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.
3590 Endocrinology, August 2007, 148(8):3580–3590 Chen et al. • Estrogen and Oocyte Development
