LINEAGE SPECIFICATION OF OVARIAN THECA CELLS REQUIRES MULTI- CELLULAR INTERACTIONS VIA OOCYTE AND GRANULOSA CELLS
BY
CHANG LIU
DISSERTATION
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Animal Sciences in the Graduate College of the University of Illinois at Urbana-Champaign, 2015
Urbana, Illinois
Doctoral Committee: Professor Janice M. Bahr, Chair Professor David J. Miller Professor Romana A. Nowak Principle Investigator Humphrey H.-C. Yao, National Institute of Environmental Health Sciences
ABSTRACT
Organogenesis of the ovary is a highly orchestrated process involving multiple
lineage determinations of ovarian surface epithelium, granulosa cells, and theca cells. While the sources of ovarian surface epithelium and granulosa cells are known, the embryonic origin(s) of theca stem/progenitor cells have not been definitively identified. Here I show that theca cells derive from two sources: Wt1‐positive cells indigenous to the ovary and
Gli1‐positive mesenchymal cells that migrate from the mesonephros. These two theca progenitor populations are distinct in their cell lineage contributions, relative proximity to granulosa cells, and their gene expression profiles. The two sources of theca progenitors acquire theca lineage marker Gli1 in the ovary soon after birth, in response to paracrine signals Desert hedgehog (Dhh) and Indian hedgehog (Ihh) from granulosa cells. Ovaries lacking Dhh/Ihh exhibited a loss of theca layer, blunted steroid production, arrested folliculogenesis, and failure to form corpora lutea. Production of Dhh/Ihh in granulosa cells requires growth differentiation factor 9 (GDF9) from the oocyte. Gdf9 ablation resulted in diminished expression of Dhh, Ihh, and theca cell‐specific Gli1. Conversely, supplementation of GDF9 to oocyte‐depleted ovaries reactivated Dhh/Ihh production and subsequent expression of Gli1. My studies provide the first genetic evidence for the origins of theca
cells and reveal a novel multicellular interaction via GDF9 and Hedgehog signaling critical
for ovarian folliculogenesis and formation of a functional theca.
ii
To my wife Yanzhen Li
iii ACKNOWLEDGEMENTS
I own a great debt of gratitude to my mentor Dr. Humphrey Yao for his guidance, training, support, generosity and friendship. He is my mentor for life. Words cannot express my appreciation to him. I want to thank all the members of the Yao lab including
Karina Rodriguez, Paula Brown, Erica Ungewitter, Barbara Nicol and Fei Zhao as well as former lab members Heather Franco, Denise Archambeault, Jeff Chen‐Che Huang, Chia‐
Feng Liu for all your support and help over the years. I feel extremely blessed to have such a wonderful “family” and it is you all who have made my everyday work in the lab full of joy. It is my great honor to have worked with each and every one of you.
I would like to thank all my committee members including Dr. Janice Bahr, Dr.
Romana Nowak, and Dr. David Miller for their advice, encouragement and availability especially during the years I am doing my PhD thesis at NIEHS in NC. Not to mention that many of them also served my committee during my Master’s study. I cannot come to this far without all your support. I want to particularly thank my scientific grandma Dr. Janice
Bahr for her loving kindness, advice and support throughout my graduate study.
I am also thankful for all the support that I received from Reproductive and
Development Biology Laboratory at NIEHS. I thank Dr. Ken Korach, Dr. Mitch Eddy, Dr.
Carmen Williams and Dr. Masahiko Negishi for their advice and suggestions on my research projects and my career as well as our secretaries Juanita Roman and Marsha Johnston who had made my life so much easier here at NIEHS.
Finally, I would like to acknowledge my family and friends for their care and prayers over the years. I love you all.
iv Table of Contents
CHAPTER 1: Literature Review ...... 1 1.1 Overview of ovarian organogenesis ...... 1 1.2 Current understanding of the establishment of the ovarian somatic cell lineages ...... 3 1.3 Current understanding of the establishment of the female germline ...... 6 1.4 Regulation of folliculogenesis ...... 11 1.5 Theca cells and ovarian diseases ...... 21 1.6 The Hedgehog signaling Pathway in mammalian gonad development ...... 24 1.7 Summary ...... 26
CHAPTER 2: The Origins of Theca Cells: Divergent Sources to a Convergent End ...... 28 2.1 Abstract ...... 28 2.2 Introduction ...... 29 2.3 Materials and Methods ...... 31 2.4 Results ...... 36 2.5 Discussion...... 43 2.6 Figures and Tables ...... 50
CHAPTER 3: Role of Desert/Indian Hedgehog Signaling in Theca Cell Development in the Murine Ovary ...... 70 3.1 Abstract ...... 70 3.2 Introduction ...... 72 3.3 Materials and Methods ...... 74 3.4 Results ...... 80 3.5 Discussion...... 84 3.6 Figures and Tables ...... 90
CHAPTER 4: Expression of Theca Cell‐Specific Gli1 Requires Oocyte‐Derived GDF9 through Hedgehog Signaling from Granulosa Cells ...... 103 4.1 Abstract ...... 103
v 4.2 Introduction ...... 105 4.3 Material and Methods ...... 108 4.4 Results ...... 111 4.5 Discussion...... 114 4.6 Figures ...... 118
CHAPTER 5: Conclusions and Future Directions ...... 124 5.1 Figures ...... 130
BIBLIOGRAPHY ...... 131
vi CHAPTER 1: Literature Review
1.1 Overview of ovarian organogenesis
The word “ovary” derives from the Latin word “ovarium”, literally “egg” or “nut”.
The mammalian ovary plays two central roles in female reproduction by acting as both
gonads and reproductive glands. As gonads, the ovary supplies germ cells to produce the
next generation. While acting as glands, the ovary produces hormones controlling a variety
of aspects of female development and physiology. The structural precursor of the ovary is
the genital ridge (or gonadal primordium). Genital ridges emerge as paired thickenings of
the epithelial layer, the coelomic epithelium, that overlays the ventral surface of the
mesonephros at 10 days post coitum (dpc) in the mouse embryo. The mechanism by which
genital ridges are initiated is unclear, however several genes, such as Emx2 [1], Lhx9 [2], Sf1
[3], Wt1 [4], Pod1 [5], M33[6], Six1 and Six4 [7] have been identified in mice that play
critical roles in the formation or growth of the gonads.
The gonadal primordium or genital ridges initially contain no germ cells. In both
sexes in mammals, primordial germ cells (PGCs) arise from the proximal epiblast, and
migrate through the dorsal mesentery of the hindgut to colonize the genital ridge at around
10.5 dpc [8‐10]. Upon entering the genital ridge, PGCs proliferate and coalesce with somatic cells to form primitive sex cords. These primitive sex cords are delineated by the deposition of the basal lamina [11‐13]. Up until this point, the gonadal development is identical in male and female and the gonad is considered bipotential, or ambisexual stage of sex differentiation [14]. In further testis development, primitive sex cords differentiate into testis cords as a result of the action of Sertoli cells [15] as well as endothelial cells from the
1 mesonephros [16, 17]. In contrast in the fetal ovary, primitive cords remain as clusters of
female germ cells (Germ cell nests) that are surrounded loosely by pre‐granulosa cells. In
mice, the squamous pre‐granulosa cells break down the germ cell nests by enclosing
individual female germ cells or oocytes at around time of birth, leading to the formation of
primordial follicles [11, 18]. Primordial follicle formation represents the first stage of
folliculogenesis and in mammals, primordial follicles serve as a resting and finite source of
oocytes available throughout the female reproductive lifespan [19]. The supporting cell
precursors that comprise each primordial follicle appear to be a heterogeneous population:
The pre‐granulosa cells in the fetal gonad contribute to the first wave of primordial follicles
in the medulla of the ovary immediately after birth. This population of primordial follicles
has long been considered, for the most part, not contributing to fertility [20‐22]. On the
other hand, the supporting cells recruited in postnatal ovaries, likely derived from the
surface epithelium, form the second wave of primordial follicles in the ovarian cortical
region [20‐22]. These primordial follicles are expected to be the follicles that ovulate over
the entire female reproductive life [20‐23]. Once the growth of primordial follicle is
activated, the primordial follicles pass through distinct stages of follicular development
classified as primary, secondary, antral and pre‐ovulatory follicles. After ovulation, the
follicle becomes corpus luteum, secreting mostly progesterone for the maintenance of pregnancy [19, 24].
2 1.2 Current understanding of the establishment of the ovarian somatic cell lineages
1.2.1 Current understanding on the establishment of granulosa cells
The embryonic origins of ovarian granulosa cells have been a subject of debate for decades. Granulosa cells function to support the growth and development of the oocytes throughout multiple stages of folliculogenesis. Follicular development starts with a handful of granulosa cells surrounding the oocyte. These original granulosa cells then undergo at least ten rounds of cell division to produce more than 2000 cells of a full grown follicle [24].
Based on morphological and histological observations of diverse mammalian ovary models, granulosa cell precursors could arise from three potential sources: The somatic precursor cells in the genital ridge [25, 26]; rete ovarii at the ovary‐mesonephros border [27‐31], and the ovarian surface epithelium [29, 32‐34]. It is possible that granulosa cells derive from one or multiple sources, and species variation seems evident in regards to the cellular origin(s) of granulosa cell precursors. Because granulosa cells in the ovary and Sertoli cells in the testis are considered analogous to each other, their origins are thought to stem from a common somatic progenitor population present in the genital ridge, before the commitment to ovary or testis occurs [26, 35, 36]. Although it is commonly believed that granulosa cells are derived from the common progenitor cells in bipotential gonad, this assumption has not been proven.
One piece of evidence that granulosa cells and Sertoli cells stem from a common progenitor population came from the observations that granulosa cells and Sertoli cells have the potential to switch their cellular fate. In mice, loss of DMRT1 transcription factor in adult Sertoli cells activates FOXL2, and transdifferentiate Sertoli cells into granulosa cells
[37]. On the contrary, FOXL2 ablation in granulosa cells in adult ovary can reprogram
3 granulosa cells into Sertoli cells [38]. Foxl2 is a winged‐helix/forkhead transcription factor.
It first comes into picture due to its potential link to the ovary‐to‐testis sex reversal
phenotype associated with Polled Intersex Syndrome (PIS) in XX goats [39, 40]. Foxl2
exhibits a granulosa‐cell specific pattern in the ovary and is required for granulosa cell
differentiation [41‐43]. These findings altogether not only reveal the role of FOXL2 in
maintaining granulosa cell identity, but also lead to the speculation that Foxl2 can be used
as a lineage maker for granulosa cells. Indeed, lineage tracing of Foxl2‐expressing cells in
the fetal ovary give rise to granulosa cell in later stages of ovarian development [22].
Additionally, the Foxl2‐expressing cells in the ovary are likely derived from the surface
epithelium in two distinct and sequential waves: an initial fetal wave followed by a
neonatal wave [22]. Foxl2‐expressing cells from the initial wave contribute to granulosa
cells of the follicles in the medulla of the ovary that start to grow immediately after birth.
Cells from the second wave populate the primordial follicles in the cortex of the ovary and these follicles are likely activated later in adulthood [21, 22].
1.2.2 Current understanding of the establishment of theca cells
The embryonic origins of ovarian theca cells are poorly understood. Theca cells play diverse roles during ovarian development, such as synthesizing androgens, which are then converted to estrogen by granulosa cells [44‐48]; interacting with granulosa cells and oocytes during folliculogenesis [49‐51]; and providing structural support and blood supply for the developing follicles for ovulation [50, 52, 53]. In mice, theca cells are first observed around one week after birth in follicles with two or more layers of granulosa cells
(secondary follicles) [19] . As the ovarian follicles contribute to develop, the theca can be
4 divided into two layers: theca interna and theca externa. Cells of the theca interna form immediately adjacent to the basal lamina of the developing follicle and are considered highly vascular and steroidogenic. The theca externa on the other hand, is located just outside of theca interna and is composed of non‐steroidogenic cells such as fibroblast cells and smooth muscle cells, important for ovulation [19].
Theca cells have long been thought to originate from the mesenchymal cells within the ovarian stroma [48, 54‐56]. Using neonatal mouse ovaries, putative thecal stem cells were isolated, and induced to differentiate into androgen‐producing theca cells in culture
[57]. Because theca cells are only found in the developing follicle and are adjacent to granulosa cells, their differentiation is considered to be under the control of granulosa cells
[58, 59]. This concept is supported by several culture experiments in which cultured ovarian stromal cells differentiate into functional theca cells in the presence of granulosa cells, or in granulosa cell‐conditioned culture medium [58, 60, 61]. Gel filtration chromatography further indicates that factors stimulating theca differentiation appear to be small‐molecular‐weight proteins ranged from 20‐ to 24‐ kDa [60]. Other speculated sources for the precursor theca cells include the coelomic mesothelium which lines the surface of the ovary [62], and the mesonephros [63‐66]. However it remains uncertain whether these sources indeed contribute to steroidogenic tissue of the ovary, as these assumptions are mainly based on morphological examinations and the definitive origins of theca cells remain veiled.
5 1.3 Current understanding of the establishment of the female germline
Generation of the female germline (or oogenesis) is the one of the other key functions of the ovary in addition to hormone production. Between 1920 and 1950, the field of oogenesis was dominated by the doctrine that the germinal epithelium, or ovarian surface epithelium encapsulating the ovary, gave rise to oocytes during each estrous or menstrual cycle. This doctrine was later rejected based on evidence that a finite stock of meiotic oocytes is present in the ovary before birth and no new oocytes are generated during adult life [67]. However, the discovery of putative germline stem cells in postnatal ovaries led to the resurgence of the controversial idea of “neo‐oogenesis” [68‐70]. In this chapter, I focus only on the establishment of female germline during fetal life.
In the mouse fetal ovary, oocytes start to form around 13.5 dpc when female germ cells (or oogonia) stop proliferating and enter the first meiosis [35]. The oocytes progress through leptonema, zygonema, pachynema, and diplonema and eventually rest at the dictyate stage of meiotic prophase I around the time of birth [71, 72]. Oocytes do not
resume meiosis until ovulation when the female reaches sexual maturity. Once ovulated
from the ovary, oocytes complete the first meiotic division, enter the second meiotic
division and arrest again. The second meiotic division is completed after fertilization [73].
Germ cells in the mouse testis behave very differently from their female
counterparts. Instead of entering meiosis at 13.5 dpc, male germ cells arrest in mitosis in
fetal life and resume mitosis immediately after birth [74]. The male germ cells or
spermatogonia then undergo the first meiotic and second meiotic divisions to generate
spermatids. The process is repeated many times to constitute a renewing supply of mature
6 sperm [73]. In contrast to the finite stock of oocytes at the time of birth, spermatogonia in
the testis retain the ability to self‐renew throughout the entire reproductive life.
How germ cells make the decision to follow the female or male path has been a
central focus of study since 1970s. By creating an XX/XY chimeric embryo, researchers
observed that XY germ cells in the ovarian tissue enter meiosis and become functional Y‐
bearing oocytes [75‐77]. By contrast, XX germ cells avoid entering meiosis if they find
themselves in a testicular environment [78]. The conclusion was therefore drawn that all
germ cells, regardless of their sex chromosome constitution, are programmed to follow
male or female pattern of meiosis according to the surrounding somatic environment [79,
80].
Byskov and others proposed that germ cell meiosis is controlled by meiosis‐
inducing substance or/and a meiosis‐preventing substance produced by the somatic cells
in the gonads [81‐87]. When an undifferentiated fetal testis is cultured with ovaries
containing meiotic germ cells, the male germ cells in the testis are coaxed into entering meiosis [82]. On the other hand, when ovaries containing germ cells in meiosis are cultured with fetal testes with well‐formed testicular structure, the oocytes are prevented from reaching the diplotene stage of meiotic prophase. Thus it was hypothesized that the fetal ovary secretes a “meiosis‐inducing substance”, which triggers the meiotic entry of germ cells. The fetal testis instead produces a “meiosis‐inhibiting substance” that prevents germ cells from entering meiosis [82].
In contrast to the hypothesis that the ovarian somatic environment secretes factor(s) that induces germ cell meiosis, McLaren and others were in favor of the concept that germ cell entry into meiosis follows a cell‐autonomous or intrinsic program [88, 89].
7 This concept was based on the observation that when male germ cells lose their way during
migration and settle in nongonadal organs such as mesonephros and adrenal, these stray
germ cells enter meiosis following the same developmental time frame as their female
counterparts in the ovary [89‐93]. Likewise, XY germ cells enter and progress through
meiotic prophase after they are separated from Sertoli cells and then cultured with other
nongonadal somatic cells such as embryonic lung cells [89]. These findings led to the
hypothesis that germ cells in the fetal ovary enter meiosis spontaneously whereas meiosis
is inhibited in the testis by factors produced by the somatic cells, probably Sertoli cells [89,
94].
The debate on how sexually dimorphic pattern of germ cell meiosis is established
appear to be resolved by the findings of Bowles et al. and Koubova et al. in 2006. A
mechanism involving retinoic acid (RA) and its degrading enzyme CYP26B1 is in action,
with both meiosis‐inducing (RA) and meiosis‐inhibiting (CYP26B1) properties [95, 96]. The first clue that RA might play a role in meiotic entry of germ cells in the ovary came from an expression screen designed to identify sexually dimorphic genes in mouse fetal gonads. It was found that Cyp26b1, the gene encoding a P450 enzyme that degrades RA [97‐100], shows a testis‐specific expression pattern. Initially present in gonads of both sexes,
Cyp26b1 becomes undetectable in female gonads after 11.5 dpc. However in the testis,
Cyp26b1 expression is maintained and reaches its maximum at 13.5 dpc [101]. The
presence of RA‐degrading CYP26b1 in fetal testes is consistent with a significant lower
level of RA in the testis compared to the ovary at 13.5 dpc. This evidence suggests that a
low RA level is necessary for preventing germ cell meiosis in the testis [101]. In other
words, a high RA level in the fetal ovary is responsible for inducing germ cell entry into
8 meiosis. Indeed, when exogenous RA is given to fetal testes in culture, XY germ cells enter
meiosis [101, 102]. In addition, treatment of fetal testes with CYP26 inhibitors in culture
induces meiotic entry of XY germ cells and an up‐regulation of Stimulated by retinoic acid
gene 8 (Stra8), which is required for premeiotic DNA replication and the subsequent events
of meiotic prophase in germ cells of embryonic ovaries [101, 103]. Finally, exposure of fetal ovaries to RA antagonists prevents XX germ cells from entering meiosis [101]. These in vitro results were later substantiated by examination of the Cyp26b1 knockout mice. In the absence of functional Cyp26b1 genes, RA levels are increased in embryonic testes and XY germ cells enter meiosis prophase at 13.5 dpc, similar to germ cells in a normal fetal ovary
[101, 104]. Collectively, this evidence supports the concept that high RA in the fetal ovary due to lack of CYP26B1 is responsible for inducing germ cell meiosis [105]. Presence of
CYP26B1 in the fetal testis prevents accumulation of RA and consequent germ cell meiosis.
Intriguingly, fetal testes and ovaries are not the source of RA. Gonads apparently lack the ability to synthesize RA because they do not express Aldh1a2, the gene encoding the major enzyme for RA synthesis [101]. Instead, RA is secreted by mesonephroi, the mesoderm‐derived tissues immediately adjacent to the gonads [101]. In 1970s, Byskov proposed that rete ovarii, the extending mesonephric derivative that connects to the ovarian medulla, may be the source of “meiosis‐inducing factor” [27, 29, 81]. In the mouse gonads, mesonephric tubules connect to the anterior portion of the gonads. If indeed the
“meiosis‐inducing factor” (or RA) comes from the mesonephric tubules, one would expect that the RA level would be higher at the anterior end than at the posterior end of the gonad.
This anterior‐to‐posterior gradient of RA in the gonads was later confirmed [101]. This phenomenon is also supported by the fact that female germ cells in the anterior part of the
9 fetal ovary enter meiosis earlier than those in the posterior end of the fetal ovary [106‐108].
Germ cell meiosis in the fetal ovary is controlled by not only the availability of the
meiosis‐inducing RA, but also the competence of germ cells to respond to RA. Germline
specific RNA‐binding protein DAZL (Deleted in azoospermia‐like gene) and NANOS2 have
emerged as intrinsic factors in germ cells that define their ability to enter meiosis in
response to RA. When Dazl became nonfunctional, germ cells in the fetal ovary fail to enter
meiosis. In addition, male germ cells in the Dazl knockout testis lose their ability to enter
meiosis in response to exogenous RA [109]. This evidence implies that the presence of Dazl
is a prerequisite for germ cells to gain the ability to respond to RA, therefore becoming
meiosis‐competent. Nanos2, on the other hand, plays a role in suppressing germ cell
meiosis. Nanos2 is expressed exclusively in germ cells in the fetal testis whereas it is absent
in female germ cells [110]. Loss of Nanos2 in the fetal testis results in upregulation of Stra8
and meiosis of male germ cells. Ectopic expression of Nanos2 in the female germ cells
decreases Stra8 expression and inhibits germ cell meiosis [111]. NANOS2 probably inhibits
germ cell meiosis by decreasing Stra8 expression, the downstream target of RA [112].
Although the role of mesonephros‐derived RA in initiating the meiotic entry of
germ cells seems evident, this concept is later challenged in 2011 by Kumar et al, who
claimed that the sex‐specific timing of meiosis in mouse embryos is independent of RA
signaling, based on the fact that germ cells still enter meiosis in the RA‐deficient fetal ovary.
On the other hand, inhibition of CYP26B1 in RA‐deficient testis allows meiosis to occur, but only when a RA‐deficient mesonephros is attached. The authors therefore conclude that
CYP26B1 prevents the onset of meiosis by degrading a mesonephros‐derived substrate other than RA that initiates meiosis [113]. How can this seemingly contradictory study be
10 reconciled with previous findings? The discrepancies imply the possibility that multiple
pathways (RA and non‐RA) may play a role in controlling the onset of meiosis.
In summary, establishment of the female germline, characterized by entry into
meiosis in fetal life, requires a synchronized action of both extracellular and intracellular
factors. Extrinsic RA or non‐RA signals, derived from the mesonephros, serves as a meiosis‐
inducing agent in the fetal ovary. Female germ cells become competent to enter meiosis in
response to these signals only after they are primed by the presence of the intrinsic factor
DAZL. The fetal ovary also suppresses production of the meiosis‐inhibiting factors
including Cyp26b1 and Nanos2, clearing the path for female germ cells to differentiate into
oocytes. The observation of putative female germline stem cells in adult ovaries [68‐70]
raises the question of how these putative germline stem cells escape from the meiosis‐
inducing substrates during fetal development and re‐enter the meiosis path later in life.
Knowledge of sexually dimorphic regulation of the germline could have implications for
solving the controversy around the existence of female germline stem cells in adult ovaries.
1.4 Regulation of folliculogenesis
Formation of primordial follicle occurs when an individual female germ cell is encased by squamous pre‐granulosa cells. This event is accompanied with oocyte apoptosis, which results in two third of the perinatal reduction in germ cell number [114]. Deletion of anti‐apoptotic gene Bcl‐2 results in fewer oocytes and primordial follicles [115]. On the contrary, mutations of genes that promote cell death, such as Ahr [116] and Casp2 [117], resulted in an increase in primordial follicle number. These results altogether suggest the critical role of apoptosis in establishing primordial follicle pool. Another factor that appears
11 to promote primordial follicle formation is FIGLA (factor in the germline α), a germ cell‐
specific bHLH transcription factor. Ovaries lacking Figla fail to form primordial follicles and consequently, the female mice are sterile [118]. In addition to FIGLA, the role of Notch signaling during the formation of primordial follicle is evident. In the neonatal mouse ovary, the Notch ligand JAGGED1 is expressed in oocytes, and the NOTCH2 receptor as well as Notch target gene Hes1 is expressed in nearby granulosa cells. Inhibition of Notch signaling decreases primordial follicle formation, along with a corresponding increase in
the number of germ cells that remain in clusters [119]. Further, mice lacking Lunatic Fringe
(Lfng), a modulator of Notch signaling that expresses in granulosa cells and theca cells of
developing follicles, exhibited polyovular follicles, indicating a failure in germ cell nest
breakdown [120]. In addition to the locally produced factors, circulating steroid hormone
also appears to regulate germ cell cyst breakdown. Neonatal ovaries that exposed to
genistein, a phytoestrogen, exhibited delayed nest breakdown [121, 122] and developed
multi‐oocytes follicles later in adulthood [123]. In contrast, E16.5 ovaries, removed from
their normal environment (highly level of pregnancy hormones), underwent premature
nest breakdown that was otherwise prevented by exogenous estrogen [122].
The transition of primordial follicle to primary follicle is defined by morphological
changes in granulosa cell from squamous to cuboidal [19]. Several key germ cell‐specific
transcription factors, including NOBOX, SOHLH1 and SOHLH2 have been identified to be
critical in this process. Mice lacking any of these genes exhibit normal primordial follicle
number but primary follicles are missing, indicating the defects in primordial to primary
follicle transition [124‐127]. In addition to oocyte‐specific transcription factors, transition
from primordial to primary follicles seems also require proper interactions between KIT
12 ligand and the KIT tyrosine receptor. In the postnatal ovary, KIT is expressed in oocytes and KIT ligand is expressed in the granulosa cells of developing follicles [128‐132].
Newborn mice injected with ACK2, an antibody to KIT, resulting in follicle arrest at primordial stage. However this phenotype is not observed if ACK2 is injected after birth when formation of the primordial follicle already occurred [132]. In contrast, supplement of KIT ligand in neonatal ovary cultures promotes the transition from primordial to primary follicle, leading to an increase in the number of growing follicles [133]. KIT ligand/KIT signaling appears to act on early folliculogenesis through the inactivation of forkhead box 3 (FOXO3), a member of O subclass of the forkhead transcription factors [134,
135]. Mice lacking Foxo3 exhibit accelerated follicular initiation, evident by global primordial follicle activation that leads to early depletion of ovarian follicles and consequently infertility [136]. This result suggests that FOXO3 functions as an inhibitor of follicular activation. While FOXO3 is a key oocyte‐derived factor important for suppressing follicular activation, another member of forkhead family, forkhead box L2 (FOXL2), is critical for the morphological changes of granulosa cells from squamous to cuboidal. FOXL2 is expressed predominantly in pregranulosa cells in the fetal ovary and granulosa cells of the developing follicles [137]. Ovaries deficient for Foxl2 form primordial follicles, but further differentiation of granulosa cells is blocked, leading to the absence of secondary follicles and oocyte atresia [138, 139]. Loss of Foxl2 further results in reduced expression of AMH (anti‐Mullerian inhibiting substance), the inhibitor of primordial follicle activation
[138]. AMH is expressed in granulosa cells from the primary follicles onward [140, 141], and its expression from the growing follicles is thought to suppress primordial follicle activation. Female mice deficient for AMH are fertile, but the number of growing follicles
13 significantly increases by 4 months of age along with a decrease in primordial follicle number [142]. The role of AMH in repressing primordial follicle recruitment is further supported by in vitro studies, in which neonatal ovaries treated with AMH present fewer growing follicles [143]. Besides AMH, the chemoattractive cytokine CXCL12 is also shown to suppress primordial follicle activation. Both CXCL12 and its receptor CXCR4 are dominantly expressed in the oocytes of the primordial and growing follicles [144].
Neonatal ovaries cultured with recombinant CXCL12 exhibit lower percentage of activated follicles, in contrast to higher density of primordial follicles [144].
The progression of the follicular development is tightly controlled by communications between oocyte and the somatic compartment of the follicle [19]. The dominant role of oocyte in directing the rate of follicle growth was demonstrated by Eppig et al [145] in elegant reaggregation experiments, where mid‐sized oocytes isolated from secondary follicles mice were transferred back to primary follicles. The reaggregated follicles exhibited accelerated folliculogenesis with expedited differentiation of follicular cells [145]. Several growth factors have been identified to be critical for oocyte‐somatic cell interactions. In the absence of GDF9, the oocyte‐specific growth differentiation factor‐9, follicle development is arrested at primary stage [146]. Particularly, the differentiation of granulosa cells is disrupted and these granulosa cells eventually show an abnormal steroidogenic phenotype [147]. The defects in granulosa cell differentiation is likely due to up‐regulated expression of inhibin α, which could potentially prevent the proliferation of granulosa cells at the primary follicle stage [147]. In fact, ablation of both inhibin α and
GDF9 in mice leads to the progression of follicle development beyond primary stage [148].
In addition to GDF9, neurotrophins NTF5 and BDNF (Brain derived neurotrophic factor)
14 expressed in granulosa cells appear to be critical during preantral folliculogenesis. NTF5 and BDNF signal via NTRK2 on the oocyte [149, 150]. Loss of both NTF and BDNF results in a drastic reduction in the number of secondary follicles, which resembles the phenotypes in
Ntrk2 ‐/‐ ovaries [150]. Besides the paracrine regulations between oocyte and somatic cells, gap junction‐mediated intercellular communications is also crucial during folliculogenesis.
This notion is supported by studies of mice model lacking key connexins [151‐154]. For instance, ovaries deficient for connexin 43 (CX43) show follicle arrest at primary follicle stage [151, 153], with disrupted granulosa cell differentiation and delayed oocyte growth.
During the secondary stage of follicular development, a morphologically distinct layer of somatic cells, the theca, is seen just outside the basal membrane of the developing follicle [19]. Theca cells and granulosa cells interact through classical epithelial‐ mesenchymal interactions during the course of follicle development [155]. The main function of theca cells is to produce androgen. Upon the stimulation of luteinizing hormone
(LH) surge, theca cell up‐regulates the expression of steroidogenic enzymes such as STAR,
CYP11A1, CYP17A1, and 3β‐HSD to produce primarily androgens. Androgen produced by theca cells then diffuses across the basal membrane to granulosa cells. Granulosa cells on the other hand in respond to follicle‐stimulating hormone (FSH) by up‐regulating the expression of CYP19A1 and HSD17B1, which convert androgen into estrogen [156].
Preantral granulosa cells later give rise to two functionally distinct cell populations during antral folliculogenesis, the mural granulosa cells lining the follicle wall and the cumulus granulosa cells immediately surrounding the oocyte [19]. These two granulosa cell populations appear to be defined through opposing effect of oocyte‐secreted factors from within the follicle and FSH from outside, although the identities of the oocyte‐derived
15 factors remain to be determined [157].
FSH and LH play crucial role in many aspects of folliculogenesis, including
granulosa cell proliferation and apoptosis, follicle atresia, estradiol production, ovulation
and the formation of corpus luteum [158‐162]. Ovaries lacking LH‐β fail to form corpora
lutea in spite of the presence of large antral follicles, and these mice are sterile [162].
Likewise, FSH‐deficient mice are infertile due to the absence of large antral follicles and a
lack of corpora lutea (CL), an indicator of ovulation failure. However, the number of oocytes
and subsequent formation of corpora lutea are recovered when these mice are treated with
exogenous PMSG (pregnant mare serum gonadotropin) and hCG (human chorionic
gonadotropin), suggesting that the ovulation competence is not affected by FSH or LH
deficiency [158, 162]. In addition, mice lacking FSH receptor (FSHR) exhibit similar ovarian
phenotypes and are also sterile [159, 163].
One key function of pituitary gonadotropins FSH and LH is to act on granulosa cells
and theca cells for the production of estradiol. Mice lacking Cyp19a1, a gene encodes
aromatase, are unable to produce estradiol [164]. This deficiency of estrogen leads to
female infertility secondary to abnormal development of antral follicles and the absence of
CLs [165, 166]. The action of estrogen on folliculogenesis relies on two estrogen receptors,
ERα (Esr1) and ERβ (Esr2) [165]. ERα is predominantly expressed in theca‐interstitial
cells in the developing ovary [167]. Ovaries from ERα‐deficient mice contain cystic,
hemorrhagic follicles and CLs are absent [168]. The cystic follicle phenotype observed in
ERα‐deficient mice is similar to those observed in Cyp19a1‐null mice and in polycystic ovarian syndrome [169]. ERβ, on the other hand, is expressed mostly in granulosa cells of the developing follicles [167]. Mice lacking ERβ are subfertile, producing smaller and fewer
16 litters due to reduced ovulation efficiency [170]. Interestingly, besides being sterile, mice lacking both ERα and ERβ (αβERKO) presented a distinct phenotype from each single ER knockout. In addition to the follicle arrest at small antral follicle stage and the absence of
CLs, the αβERKO mice exhibited sex‐reversed follicles, evident by the appearance of testis cords like structure and SOX9 positive Sertoli‐like cells [171, 172]. In addition to ER receptors, androgen receptor (AR) also plays a role in ovarian folliculogenesis. Young and sexually matured Ar‐/‐ female mice show reduced fertility with decreased ovulation rates and number of CLs [173, 174]. However at forty‐weeks of age, these mice become infertile due to complete absence of follicles, a phenotype resemble that of women with premature ovarian failure [174, 175].
Only a subset of follicles in the growing pool can proceed to the stage of ovulation, whereas a majority of follicles will undergo atresia instead. The mechanism of how a subset of follicles survives to ovulation is not clear; however, both in vivo and in vitro studies have suggested several factors/pathways critical for this process [19]. Preovulatory follicles highly express luteinizing hormone/choriogonadotropin receptor (LHCGR) in mural granulosa cells [157], enabling them to respond to LH surge, leading to ovulation and subsequence formation of CLs [19]. Mice deficient in Lhcgr are sterile [176, 177], and mirror the histological defects observed in LH‐β‐deficient mice. However, exogenous gonadotropin administration cannot rescue the follicular defects in Lhcgr null mice, differed from the recovery observed in LH‐β‐deficient mice [176]. Upon the initiation of LH surge, mural granulosa cells [157] produce EGF‐like factors including Areg, Ereg and Btc
[178], which act on cumulus granulosa cells that surround the oocyte [179] to induce cumulus expansion and oocyte maturation. The expansion of the cumulus oophorus is a
17 prerequisite for normal ovulation. Proper formation of the extracellular matrix of the
cumulus oophorus depends on the activation of MAPK signaling [180, 181], along with the
expression of critical genes including Has2, Ptgs2, Ptx3 and Tnfaip6 in cumulus cells [182‐
186]. Activation of MAPK signaling in cumulus cells requires one or more paracrine factors
from the oocyte [187, 188]. Oocytectomy of oocyte‐cumulus complexes prevents FSH
induced cumulus expansion, whereas co‐culture of FSH stimulated oocytectomized
complexes with denuded oocytes rescues the expansion of the oocytectomized complexes
[189]. Similarly, dissociated cumulus cells alone failed to synthesize a mucinous matrix in
response to FSH, but the matrix production is restored when denuded full‐grown oocytes
are supplemented in culture [190]. Thus, it is postulated that oocytes secrete a cumulus
expansion‐enabling factor (CEEF) that allowed granulosa cells to respond to FSH [191].
One such potential factor is GDF9, a member of the transforming growth factor‐beta (TGFβ) super family. Knockdown of Gdf9 in oocytes results in down‐regulated expression of Has2,
Ptgs2, and therefore suppresses cumulus expansion [192], whereas the supplement of recombinant GDF9 in granulosa cell culture increases levels of Has2, Ptgs2 and Ptx2 expression [193, 194]. Conversely, the response of granulosa cells to GDF9 signaling also requires the activation of MAPK pathways, suggesting that GDF9 and MAPK signaling are cooperatively required for follicle growth unto ovulation [195]. In addition to GDF9, another member of TGFβ family BMP15, also appear to be involved in the regulation of
cumulus expansion [196]. This is because first, BMP‐15 induces cumulus expansion in
mouse cumulus–oocyte complexes via stimulating the expression of EGF‐like growth
factors in vitro [196]. Second, Gdf9+/‐ Bmp15‐/‐ double mutant mice are infertile with
impaired cumulus expansion [197]. Third, co‐culture of Gdf9+/‐ Bmp15‐/‐ oocyte with
18 oocytectomized cumulus complexes fails to cause FSH‐induced cumulus expansion along
with decreased level of MAPK signaling [198]. These observations suggest that BMP15
signals synergistically with GDF9 in regulating cumulus expansions in mouse.
Several transcription factors downstream of LH receptor activation have been
demonstrated to be necessary for ovulation [19]. Progesterone receptor (PR), a member of
nuclear receptor subfamily, is expressed in the mural granulosa cells of the preovulatory
follicles [199, 200]. PR‐deficient female mice are anovulatory despite intact cumulus
expansion in unruptured follicles [201]. The single progesterone receptor gene (Pgr) uses
separate promoters and translational start sites to produce two PR isoforms, PR‐A and PR‐
B [202]. To further determine which PR isoform was responsible for the anovulatory
phenotype, selective ablation of each PR isoform is carried out and PR‐A, but not PR‐B,
regulated the ovulatory response to the LH surge [203, 204]. Other members of nuclear
receptor family involved in this process include LRH1 (Nr5a2), SF1 (Nr5a1) and TR4
(Nr2c2). Conditional deletion of either Nr5a2 or Nr5a1 leads to infertility secondary to
anovulation, in spite of different molecular and hormonal changes in each mutant [205,
206]. Mice deficient for TR4 are subfertile with declined number of preovulatory follicles
and CLs, likely due to an insufficient expression of LH receptor to respond to LH surge [207,
208]. In addition to these nuclear receptors aforementioned, co‐regulators of the nuclear
receptor superfamily are also implicated in the regulation of ovulation. One such factor is
nuclear‐receptor‐interacting protein 1 (NRIP1; encoded by the gene Nrip1). Nrip1‐/‐ mice are unable to ovulate; however, luteinization proceeds normally, resulting in the formation of CLs with trapped oocytes [209]. Further, Nrip1‐/‐ ovaries exhibit defects in cumulus expansion, indicating the role of Nrip1 in this process besides ovulation [209, 210]. Another
19 transcription factor necessary for the process of ovulation is C/EBPβ (CCAAT‐enhancer‐
binding protein β, encoded by Cebpb). Cebpb expression is induced specifically in granulosa
cells of preovulatory follicles following hCG administration. Mice carrying a targeted
deletion of Cebpb contain unruptured follicles and lack CLs [211].
The CL is a highly differentiated endocrine structure required for maternal
development during pregnancy and is an essential source of progesterone [212, 213].
Formation of CLs requires granulosa cells to exit from the cell cycle [214]. Granulosa cell
cycle arrest is regulated by LH surge‐activated MAPK signaling via up‐regulating inhibitors
of cell cycle progression [181]. One such inhibitor is Cyclin‐dependent kinase inhibitor 1B
(CDKN1B; also known as p27Kip1)[215]. Female mice deficient for CDKN1B are sterile due to granulosa cell hyperplasia and defects in CL formation [216‐218]. Administration of
gonadotropins induces ovulation; however, pregnancy is not maintained due to insufficient
progesterone level [216]. Although progesterone is considered “the pregnancy hormone”,
pituitary gland‐secreted prolactin (PRL) also plays a critical role in the formation of CLs
and the production of progesterone. Female mice deficient for prolactin receptor (PRLR)
are sterile, unable to support implantation and pregnancy because of insufficient
progesterone production [219]. The blunted progesterone production is likely due to
defects in the maintenance of CLs, as they are structurally disorganized and many of them
undergo apoptosis [220]. Because luteinization occurs only after ovulation, it is then
hypothesized that factors secreted from oocyte inhibit premature luteinization. It was
previously shown that oocyte‐stimulated SMAD2/3 signaling is involved in enabling
cumulus expansion in vitro [157], pointing out members of TGFβ family as potential
20 oocyte‐derived inhibitors for luteinization. Evidence for this is seen in ovaries lacking
Smad4, which exhibits severe cumulus cell defects and premature luteinization [221].
1.5 Theca cells and ovarian diseases
1.5.1 Epithelial ovarian cancer
Ovarian cancer is the fifth leading cause of cancer death among women in the
Western world and 100,000 women die of this cancer each year [222, 223]. Eighty percent of ovarian cancers are epithelial in origin [224]. Epithelial ovarian cancer is classified by their genetic characteristics and histopathologies as serous, endometrioid, clear cell and mucinous [225, 226]. High‐grade serous ovarian carcinoma is the most prevalent subtype and accounts for more than 75% of mortality associated with the malignancy [222, 227,
228]. In most cases, women with invasive ovarian cancer are diagnosed at an advanced‐ stage, when a cure is rare. Although epithelial ovarian cancer is considered the most lethal gynecological malignancy [229], its pathogenesis is poorly understood [222, 223, 230].
Epithelial ovarian cancers are thought to arise from the ovarian surface epithelium that comprises the ovarian surface [231]. Using mouse ovary as a model, Andrea et al. showed that ovarian surface epithelium at the hilum region of the ovary contains a cancer‐ prone stem cell niche positive for stem cell marker ALDH1 [232]. Interestingly, ALDH1 staining is also present in the theca cells [232]. In addition, disorders of androgen production have been implicated in epithelium ovarian cancer [233] and androgen receptor is expressed in the ovarian surface epithelial cells [234]. However, it is not clear whether theca cells play a role in the etiology and progression of these cancers. [233]
21 In addition to ovarian surface epithelium as a source of epithelial ovarian cancer,
recent studies suggest that many invasive cancers likely originate from non‐ovarian tissues
[235]. Possible sites of origins for high‐grade serous ovarian carcinomas include the distal
fallopian tube [236‐238] particularly in women with BRCA1 (breast cancer 1) or BRCA2
(breast cancer 2) mutations [239‐241], and the abdominal peritoneum [235]. Because a
considerable proportion of ovarian tumors do not arise from ovarian tissues, the term
ovarian cancer is therefore for invasive cancers collectively, and has evolved to reflect the
possibility of additional sites of origins from different tissues [227].
1.5.2 Polycystic ovary syndrome
Polycystic ovarian syndrome (PCOS) is a common endocrinopathies, affecting 5‐
10% women of reproductive age [169, 242, 243]. The symptoms associated with PCOS include hyperandrogenism, anovulation, and infertility as a result. The hallmark ovarian morphology of PCOS is the accumulation of multiple small subcortical follicular “cysts” embedded in enlarged ovaries, along with an increases in ovarian stromal volume [244‐
246]. Although there has been debate about the diagnostic criteria for PCOS, hyperandrogenism and its symptoms are widely believed to be a critical component of the syndrome [247‐250].
It is generally agreed that theca cells from the ovary are at least one primary source of excess androgen production in women with PCOS [169, 250, 251]. This notion is supported by studies on isolated theca tissue or cultures of theca cells from PCOS and normal women, which demonstrate that PCOS theca produces greater levels of androgens than that of normal women [251‐257]. The rise in ovarian androgen production is likely
22 due to increased expression of key enzymes involved in ovarian steroidogenesis including
LH receptor, StAR, P450scc, CYP11A1 and CYP17A1 [253‐255, 257, 258].
Polycystic ovary syndrome appears to be a familial condition and a genetic basis of
PCOS is strongly suggested [259‐261]. One line of evidence that links theca cells and PCOS
is a genome‐wide study that identifies DENND1A as a PCOS candidate locus. It was shown
that the spliced form of DENND1A, DENND1A.V2, is increasingly expressed in human PCOS
theca cells. Overexpression of DENND1A.V2 in theca cells increases the expression of
steroidogenic genes that leads to excess androgen production. On the other hand,
knockdown of DENND1A.V2 resulted in decreased expression of genes involved in
steroidogenesis, and consequently a reduction in androgen production [262]. Although the
association between theca cells and PCOS is apparent, it is yet unclear whether
dysfunctions of theca cells contribute to PCOS, or the consequence of it.
1.5.3 Premature ovarian failure
Premature ovarian failure (POF), also known as primary ovarian insufficiency or
premature menopause, is the early decline of ovarian function in women after seemingly
normal folliculogenesis [263]. This medical condition is affecting 1‐2% of women before
age 40 and 0.1% of women under the age of 30 [264]. Characteristic of POF include
amenorrhoea, hypo‐oestrogenism and elevated gonadotropin levels [265]. Consequently,
women with POF are infertile, which in most cases is due to follicle depletion [266], and in
other cases, follicle dysfunction [267].
How spontaneous POF arise is largely unknown but several causes have been
implicated, including mutation of autosomal genes, autoimmunity, irritation, chemotherapy
23 and X‐chromosome abnormalities such as Turner’s syndrome [265, 266, 268, 269].
Several lines of evidence have implicated the role of theca cells in the development of POF.
Mice deficient in X chromosome androgen receptor (Ar) gene develop POF phenotype as
they age [270]. This finding indicates a possible role of androgen signaling in the etiology of
POF. In the absence of Ar, expression of Gdf9, Bmp15 and Hgf are downregulated in an age‐
dependent manner. While Gdf9 and Bmp15 are oocyte‐specific factors critical for ovarian
folliculogenesis [271, 272], Hgf is produced by theca cells and is important for granulosa cell function [273]. Additionally, in cases associated with autoimmunity, theca cells appear
to be the primary target of the circulating autoantibodies in the initial stage, resulting
multifollicular development, reduced estrogen level, and elevated inhibin levels [274].
1.6 The Hedgehog signaling Pathway in mammalian gonad development
The origin of the name Hedgehog (Hh) stems from the “spiked” phenotype of the
cuticle of the Drosophila larvae. Mutations in the Hh gene result in the impairment of the development of the larval body patterning in flies [275]. Since its discovery, Hh signaling has been implicated in almost every aspect of the body plan, such as cell fate, growth and survival, and patterning throughout the vertebrate species. In addition to the role of Hh signaling in normal organogenesis, dysregulation of the signaling pathway is implied in many diseases in human such as cancers [276, 277]. In mammals, three Hh ligands have been identified: sonic Hedgehog, desert Hedgehog (DHH), indian Hedgehog (IHH) [276]. In the absence of the ligands, the hedgehog receptor Patched (PTCH1 or PTCH2) inhibits the activity of another membrane‐bound receptor Smoothened (SMO). Binding of Hh to PTCH suppresses its activity, releasing the action of SMO. SMO then transduces the Hh signals to
24 the cytoplasm, drives the activation of zinc‐finger transcription factors (GLI1‐3 in mammals), and ultimately leads to the expression of Hh target genes [276, 278].
The role of Hh signaling in gonad development was first indicated in studies investigating the action of DHH in the mouse embryo, as its expression is largely restricted to gonads. In the fetal testis, Dhh is the sole Hh ligand that is specifically expressed in the
Sertoli cells starting E11.5 [279]. On the other hand, the Hh downstream targets Gli1 and
Ptch1 are distinctly expressed in the interstitium of the fetal testis, including fetal Leydig cells and peritubular myoid cells [280‐282]. Consistent with the restricted expression pattern of Dhh in the testis, Dhh null mice are viable but males are infertile due to loss of mature sperm in adult testis [282], disorganized testis cords [283], or Leydig [281] and peritubular myoid cell defects [284], depending on the genetic background of the mice.
The role of Hh signaling on Leydig cell development is further supported by an in vitro study, in which the inhibition of the transcriptional activity of Hh downstream effectors
GLI1 and GLI2 results in the loss of fetal Leydig cells [280].
Differed from the fetal testis, the fetal ovary is devoid of Hh signaling based on the observation that the expression of Ptch1 and Gli1, the indicators of active Hh signaling, is absent in the ovary during fetal life [279, 280]. If Hh signaling is ectopically activated in the fetal ovary, it leads to the appearance of functional fetal Leydig cells [285]. Thus, inactive state of Hh signaling seems necessary for the fetal ovary [278]. In contrast to the absence of active Hh signaling in the fetal ovary, both Dhh and Ihh are detected in granulosa cells of the developing follicles in the adult ovary [286, 287]. The Hh downstream targets, such as
Ptch1, and Gli1, are localized specifically in the mesenchymal stromal cells surrounding the follicles [287]. This unique expression pattern suggests a potential paracrine regulation of
25 DHH/IHH signaling on theca cell development. Female Dhh‐deficient mice are fertile and exhibit normal ovarian development [282], suggesting that the other Hh ligands, such as
Ihh, may compensate for the loss of Dhh during folliculogenesis. Unfortunately, most of the
Ihh‐deficient embryos die before birth [288], therefore precluding the analysis of the ovary in the adult. Alternatively, short‐term culture of neonatal ovaries with an Hh inhibitor, cyclopamine, does not affect folliculogenesis and early secondary follicles are observed
[287]. This result indicates that Hh singling is likely not critical in early folliculogenesis. On the other hand, constitutive activation of the Hh signaling pathway in the ovary in vivo disrupts theca cell development, leading to ovulation failure and female infertility [289].
The role of Hh signaling in theca cell development is also supported by several in vitro studies in which the addition of recombinant SHH in culture stimulates androstenedione production of the theca cells [290] as well as the growth of the follicles [286].
1.7 Summary
Folliculogenesis in mammalian females is controlled by intricate communication between oocytes, granulosa, and theca cells. One primary role of theca cells is to produce androgens for ovarian function [44‐48]. Dysfunctions of theca cells are implicated in ovarian diseases such as polycystic ovarian syndrome, premature ovarian failure and ovarian cancers [233, 270, 291‐294]. Despite their involvement in ovarian functions and pathogenesis, theca cells received much less attention over the years and were poorly studied [50].
One key unanswered question in ovarian biology is wherefrom and how theca cells arise during ovarian development. Theca cells have long been thought to originate from
26 the mesenchymal cells within the ovarian stroma postnatally [48, 54‐56]. However, this assumption is mainly based on morphological examination and the definitive origins of theca cells remain unclear. Other possible sources for the precursor theca cells include the coelomic mesothelium which lines the surface of the ovary [62], and the mesonephros [63‐
66]. However, none of these sources have been tested conclusively. One reason for this is the lack of lineage markers exclusively for theca cells. Theca cell share similar steroidogenic activity and regulation with their testicular counterpart, the Leydig cells[137,
295, 296]. Thus, the knowledge gained from the Leydig cells studies might be applicable to theca cells in the ovary. The specification of the fetal Leydig cells is regulated by DHH signaling from Sertoli cells, and the downstream transcription targets involved in this process include Ptch1 and Gli1 [281, 283, 297]. In the postnatal and adult ovaries, theca cells express Gli1 and Ptch1, and the Hh ligands Dhh and Ihh are expressed in granulosa cells [287]. These observations implicate a potential role of Hh signaling in theca cell development and suggest that GLI1 may be a lineage marker for theca cells. I therefore hypothesize that Hh signaling is playing an important role in theca cell development. This central hypothesis will be tested in three specific objectives. The first specific objective of my doctoral studies is to determine whether GLI1 can be used as a lineage marker for theca cells and from where theca cells originate. The second objective is to investigate whether
Hh signaling is required for theca cell differentiation and ovarian folliculogenesis. My third objective is to explore the mechanisms in which Hh signaling is regulated in the ovary. My research provides valuable information regarding the origin(s) of theca cells, the mechanisms of their differentiation, and their impact on female fertility.
27 CHAPTER 2: The Origins of Theca Cells: Divergent Sources to a Convergent End
2.1 Abstract
Theca cells play essential roles in not only steroidogenesis but also folliculogenesis in the ovary. This ovarian specific cell type is thought to derive from the stroma surrounding the developing follicles; however, their definitive origin(s) has not yet been identified. Using Cre/loxP‐based lineage‐tracing mouse models, I revealed two distinct sources for theca cell progenitors: the Wt1‐positive cells intrinsic to the ovary and the mesenchymal Gli1‐positive cells that migrate from the mesonephros, an embryonic tissue immediately adjacent to the fetal ovary. In addition to the difference in their spatial localizations, fetal ovary‐derived Wt1‐positive cells give rise to the majority of theca‐ interstitial cells in the adult ovary and represent a broad mesenchymal cell population(s), whereas mesonephros‐derived Gli1‐positive cells only contribute to a small yet specific steroidogenic theca cell population. The fetal ovary‐ and mesonephros‐derived progenitor cells converge in the ovary after birth and commit to the lineage of theca cells by acquiring
Gli1 expression. To further investigate if these Gli1‐positive theca progenitor cells are important for ovarian functions, I developed a genetic model that induces cell death specifically in the Gli1‐positive cells after birth. Ablation of the Gli1‐positive cells resulted in loss of primary and secondary follicles, decreased ovarian mass and decreased steroidogenic activity. In summary, these studies uncover multi‐origins of the progenitors of theca cell lineage: the ovarian Wt1‐positive cells and the extraovarian Gli1‐positive cells in the mesonephros. Furthermore, the Gli1‐positive theca cells are indispensable for normal ovarian function.
28 2.2 Introduction
Ovarian morphogenesis is a highly orchestrated process in which the follicles, the basic unit of ovary, form and mature to produce functional eggs. The development of follicle is governed by intricate communication among oocyte, epithelial‐like granulosa cells and mesenchymal‐like theca cells. Defects in this process have dire consequences for female reproductive health and fertility. The process of folliculogenesis starts with germ cell nest breakdown, in which individual oocyte becomes encased by somatic granulosa cells. As the follicle continues to develop, it recruits precursors for the theca cell lineage and completes follicle assembly [19, 24]. Theca cells and granulosa cells interact through classical epithelial‐mesenchymal type interactions for the progression of follicle development [155]. Theca cells produce androgens in response to luteinizing hormone.
The androgens are subsequently converted to estrogens by granulosa cells in responses to follicle‐stimulating hormone [19], and the estrogens in turn provides a local feedback loop in regulating androgen production [298]. This “two‐cell two‐gonadotropin” theory establishes the significance of theca cells in follicular steroid production for ensured ovulation. Disorders in theca cell differentiation are implicated in ovarian diseases such as polycystic ovary syndrome (PCOS), premature ovarian failure, and ovarian cancers [233,
270, 292], therefore compromising the fertility. The long‐held assumption in the field is that theca cells are derived from the mesenchymal cells within the ovarian stroma [48, 50,
54, 55, 299]. However, this assumption is mainly based on morphological examinations and in vitro analysis, and the definitive origins of theca cells remain veiled. The lack of knowledge on theca cell precursors hinders the investigation on the molecular basis of theca cell recruitment and differentiation.
29 Theca cells in the mouse ovary become morphologically distinguishable around one week after birth in follicles with two layers of granulosa cells (secondary follicles) [19].
These specialized mesenchymal cells are located in the ovarian mesenchyme, which is separated from granulosa cells and oocytes by a basal membrane [19]. Their histological proximity to granulosa cells led to the hypothesis that recruitment of theca cells from the stroma is regulated by factor(s) produced by granulosa cells [137]. A candidate for this factor is Hedgehog (Hh) ligand, a morphogen responsible for lineage specification in many organs [276]. In adult ovaries, Indian hedgehog (Ihh) and Desert hedgehog (Dhh) are detected in granulosa cells, and their downstream target Gli1 is expressed specifically in the theca cell layer [287]. While the role of Hh signaling in the ovary is not clear, it is known that DHH derived from Sertoli cells in the testis regulates the appearance of fetal Leydig cells [281], the counterpart of putative theca progenitor cells in the ovary. In addition, one of the Hh effectors involved in the specification of fetal Leydig cells is Gli1 [280, 281]. With all these observations, I therefore hypothesize that Gli1 is expressed in the precursors of theca cell and might be used as a lineage marker for theca cells in the ovary.
30 2.3 Materials and Methods
2.3.1 Animals and experimental protocols
Gli1‐LacZ (#008211), Gli1‐CreERT2 (#007913) [300], Wt1‐CreERT2 (#010912), Rosa‐
LSL‐tdTomato (#007905), Rosa‐DTA (#006331) mice were purchased from the Jackson
Laboratory (Bar Harbor, ME). Female mice were housed with male mice overnight and
checked for the presence of vaginal plug the next morning. The day when the vaginal plug
was detected was considered embryonic day (E) 0.5. All animal procedures were approved
by the National Institute of Environmental Health Sciences (NIEHS) Animal Care and Use
Committee and are in compliance with a NIEHS‐approved animal study proposal. All experiments were performed on at least three animals for each genotype.
2.3.2 Tamoxifen‐induced lineage tracing and cell ablation experiments
For lineage tracing and cell ablation experiments, CreERT2 activity was induced by
IP injection of 1 mg and 5 mg of tamoxifen (T‐5648, Sigma‐Aldrich, St. Louis, MO) per mouse in 100 μl corn oil (C‐8267, Sigma‐Aldrich, St. Louis, MO), receptively. For the vehicle control, an equivalent volume of corn oil was injected. Lineage tracing experiments were conducted by crossing male Gli1‐CreERT2, or Wt1‐CreERT2 mice with female Rosa‐LSL‐
tdTomato reporter mice to generate pups carrying both CreERT2 and Rosa‐LSL‐tdTomato
alleles. Cell ablation experiments were conducted by crossing male Gli1‐CreERT2 mice with
female Rosa‐DTA mice to generate pups carrying both CreERT2 and Rosa‐DTA alleles. To
induce Cre‐mediated recombination in neonatal pups (cell ablation experiments),
tamoxifen was injected to the lactating dams. No overt teratological effects were observed
after tamoxifen administration under these conditions.
31 2.3.3 Immunohistochemistry and histological analysis
For immunohistochemistry on frozen sections, ovaries were fixed in 4% paraformaldehyde in PBS at 4oC overnight, dehydrated through a sucrose gradient, embedded, and cryosectioned at 10 μm increments. After preincubating with 5% normal donkey serum in
PBS for 1 hour, the sections were then incubated with primary antibodies (listed in the table below) in PBS-Triton X-100 solution with 5% normal donkey serum at 4oC overnight. The sections were then washed and incubated in the appropriate secondary antibody (1:500;
Invitrogen, NY, USA) before mounting in Vector Mount with DAPI (Vector Labs). Slides were imaged under a Leica DMI4000 confocal microscope. For paraffin-embedded tissues (regular samples or samples after LacZ staining), the sections were dewaxed and rehydrated in a series of alcohol to PBS. The slides were pretreated in 0.1 mM citric acid for 20 min in the microwave, then incubated with an antibody specific to anti-3HSD (listed in the table below) with the condition listed above. For histological analysis, the samples were fixed in 4% paraformaldehyde in PBS at 4oC overnight, dehydrated through an ethanol gradient, and embedded in paraffin wax. Sections were counterstained with hematoxylin and eosin.
Primary antibodies Species Dilution Source
FOXL2 Rabbit 1:500 Dr. Dagmar Wilhelm, Monash University, Australia
FOXL2 Goat 1:500 Imgenex, Littleton, CO, USA
PECAM-1 Rat 1:500 BD Biosciences, San Jose, CA, USA
3HSD Rabbit 1:500 CosmoBio Co.Ltd, Japan
β-galactosidase Chicken 1:1000 Abcam, San Francisco CA, USA
TRA98 Rat 1:1000 MBL international, Woburn, MA, USA
WT1 Rabbit 1:300 Abcam, San Francisco CA, USA
32 DsRed Rabbit 1:1000 Clontech, Mountain View, CA, USA
laminin Rabbit 1:500 Sigma-Aldrich, St. Louis, MO, USA
2.3.4 LacZ staining
Fresh tissues were stained in LacZ staining solution at 37oC for 3 hours. The LacZ
staining solution was prepared by dissolving X‐gal (Invitrogen, Carlsbad, CA) in
dimethylformamide and the tissue stain base solution (Chemicon, Billerica, MA) to make 1
mg/ml working solution as described elsewhere [301]. The stained samples were then
fixed in 4% paraformaldehyde/PBS at 4oC overnight. For further histological analysis,
samples after LacZ staining were embedded in paraffin following the sectioning procedure
listed above. The sections were counterstained with Fast Red (Sigma).
2.3.5 Ovarian transplantation
Ovaries from tamoxifen‐treated Gli1CreERT2; Rosa‐DTA or Rosa‐DTA pups were transplanted under the kidney capsule of ovariectomized wild type recipient mice. A small
hole was torn in the kidney capsule of the recipient mice using a 28G needle. A control
ovary (Rosa‐DTA) was inserted into one kidney capsule and the Gli1CreERT2; Rosa‐DTA
ovary was inserted into the kidney capsule of the contralateral ovary in the same recipient
mouse. In order to stimulate the growth of the transplants, both ovaries of the recipient
mice were removed [302]. The transplants were collected 2 weeks after transplantation,
followed by fixation in 4% paraformaldehyde in PBS and were processed for
immunohistochemistry and histological analysis.
33 2.3.6 Microarray analysis
Gene expression analysis was conducted using Affymetrix Mouse Genome 430 2.0
GeneChip® arrays (Affymetrix, Santa Clara, CA). Total RNA was extracted in 11 μl elution
buffer using Arcturus® PicoPure® RNA isolation kit (Applied Biosystems by Life
Technologies) following the manufacture’s protocol. Due to limited amount of RNA, the
eluted RNA was dried in a vacuum concentrator, and re‐suspended in 4 μl of RNase‐free
water. The 4 μl of total RNA was then amplified as directed in the WT‐Ovation Pico RNA
Amplification System protocol, and labeled with biotin following the Encore Biotin Module.
Amplified biotin‐aRNAs (4.6 μg) were fragmented and hybridized to each array for 18
hours at 45°C in a rotating hybridization. Array slides were stained with
streptavidin/phycoerythrin utilizing a double‐antibody staining procedure and then
washed for antibody amplification according to the GeneChip Hybridization, Wash and
Stain Kit and user manual following protocol FS450‐0004. Arrays were scanned in an
Affymetrix Scanner 3000 and data was obtained using the GeneChip® Command Console
Software (AGCC; Version 1.1). The microarray raw data were analyzed using the Partek
Genomic Suite Software (St. Louis, Missouri). Two‐way ANOVA was performed to
determine the statistical significance between the means. The gene list was filtered with
fold‐change cutoff of 1.4 and an unadjusted p<0.05.
2.3.7 Gonad dissociation and cell sorting
Gonadal cells from Gli1‐CreERT2; Rosa‐LSL‐tdTomato ovaries were minced and then enzymatically dissociated in Hank’s balanced salt solution (Sigma) with 1 mg/ml collagenase B (Roche), 1.2 U/ml Dispase II (Roche) and 5 U/ml DNAse I (Sigma) for 20 min
34 at 37 oC. To further aid tissue dissociation, the solution was mixed by repeated pipetting
(P1000 pipette) [303]. Dissociated cells were rinsed and re‐suspended in ice‐old PBS, and
were subject to fluorescence‐activated cell sorting (FACS) for tdTomato+ and tdTomato‐ cells. Cell counts were also obtained during FACS analysis. FACS was performed using a
FACSAria II flow cytometer from Becton Dickinson (San Jose, CA) and FACSDiVa software version 6.1.3.
2.3.8 Gene expression analysis
Gene expression analysis was performed as described with modifications [304].
Total RNA was isolated from ovaries using the PicoPure RNA isolation kit (Arcturus,
Mountain View, CA) according to the manufacturer’s protocol. The cDNA preparation was synthesized using random hexamers and the Superscript II cDNA synthesis system
(Invitrogen Corp., Carlsbad, CA) following the manufacturer’s instruction. Gene expression was analyzed by real‐time PCR using Bio‐Rad CFX96TM Real‐Time PCR Detection system.
Taqman gene‐expression probes (Gli1: Mm00494654_m1, Star: Mm00441558_m1,
Cyp11a1: Mm00490735_m1, Cyp17a1: Mm00484040_m1, Lhcgr: Mm00442931_m1) were
used to detect the fold changes of the transcripts. Fold changes in gene expression were
determined by quantitation of cDNA from target (treated) samples relative to a calibrator sample (control). All real‐time PCR analyses were performed in duplicate, and the results were reported from at least three independent experiments. The relative fold change of
transcript was calculated using the mathematical model of Pfaffl as previously described
[305] and was normalized to 18S rRNA (Mm03928990_g1) as an endogenous reference.
35 2.4 Results
2.4.1 Migrating Gli1+ cells from the mesonephros contribute to the ovarian theca cell
lineage
The expression of Gli1, indicated by the presence of beta‐galactosidase in the Gli1‐
LacZ reporter mouse, was absent in the ovary before birth (E17.5) and did not appear in
the ovary until postnatal day 2 (P2) (Fig. 2.1). Before P2, Gli1‐LacZ expression was
restricted to the mesenchyme of the mesonephros immediately adjacent to the ovary (Fig.
2.1m & o), particularly in the mesonephric tubules that connect the rete ovarii of the ovary
(Fig. 2.1b). Between birth and P2, Gli1‐LacZ‐positive cells spread from mesonephric tubules
into ovarian interstitium (Fig. 2.1b, c, g and k). By P6, Gli1‐LacZ‐positive cells were found in
the mesenchyme surrounding the FOXL2‐positive granulosa cells (Fig. 2.1d, h and l) of
primary follicles (Fig. 2.3n & p). This pattern persisted in the ovary at 2 months of age (Fig.
2.2a) and some Gli1‐positive cells are positive for steroidogenic marker 3βHSD (Fig. 2.2b)
and steroidogenic factor‐1 (SF‐1) (Fig. 2.2c). The close association of Gli1‐positive cells in
the mesonephros and the ovary suggests that Gli1‐positive cells from the mesonephros
could be a source of theca progenitors. To test this hypothesis, I utilized a tamoxifen‐
induced Rosa‐LSL‐tdTomato lineage‐tracing model, in which Gli1‐positive cells in the
mesonephros were labeled exclusively during embryogenesis [306]. When tamoxifen is
injected to the Gli1‐CreERT2; Rosa‐LSL‐tdTomato animals, it activates CreER recombinase, which then induces permanent expression of tdTomato fluorescence in the Gli1‐positive
cells and their progeny [300, 307]. In addition, the lineage labeling of the Gli1‐positive cells
lasts for approximately 36 hr after tamoxifen injection [308, 309]. I administered a single
tamoxifen injection to pregnant mice carrying Gli1‐CreERT2; Rosa‐LSL‐tdTomato embryos at
36 E12.5, when Gli1 expression was restricted in the mesonephros [280]. The specificity of
this model was confirmed by the lack of fluorescence in the tamoxifen treated Cre‐negative
ovaries, and the oil‐treated Gli1‐CreERT2; Rosa‐LSL‐tdTomato controls (Fig. 2.3). During fetal life, the tdTomato‐positive cells were present specifically in the mesonephros but absent in the ovary (Fig. 2.4a & b), consistent with endogenous Gli1 expression pattern (Fig.
2.1a, e and i). The tdTomato‐positive cells first appeared in the ovary just before birth at
E18.5 (Fig. 2.4c) and a significant number of these cells were observed in the neonatal ovary (Fig. 2.4d). This result demonstrates that the tdTomato‐positive cells in the neonatal ovary were derived from the Gli1‐positive cells in the fetal mesonephros. At two months of age, the mesonephros‐derived Gli1‐positive cells in the ovarian interstitium became steroidogenic theca cells positive for 3HSD (Fig. 2.4e).
When comparing the Gli1‐positive cells in the ovary (Fig. 2.1d & h) with the contribution of Gli1‐positive cells from the mesonephros (Fig. 2.4e), it became apparent that the mesonephros‐derived Gli1‐positive cells represent a small fraction of the ovarian
Gli1‐positive cell population. To exclude the possibility that the low number of mesonephros‐derived Gli1‐positive cells is due to inefficient labeling of Gli1‐positive cells in
the mesonephros, I increased tamoxifen‐induced lineage labeling by administrating
tamoxifen to embryos from E12.5 to E14.5 for 3 consecutive days. The efficiency of lineage
labeling was validated in E15.5 mesonephros, and the majority of mesenchymal cells in the
mesonephros were positive for tdTomato (Fig. 2.5a‐c). Although plenty of Gli1‐positive
cells in the mesonephros were lineage labeled during fetal life, it seems that only a small
number of them actually migrated into the ovary after birth (Fig. 2.5d‐f). The majority of
theca cells therefore seem to come from the ovarian mesenchymal cells that become Gli1‐
37 positive (Fig. 2.1g & h). This hypothesis was supported by the findings that lineage‐labeled
ovarian Gli1‐positive cells at P2 (Fig. 2.1c, g, k) became the majority of the 3HSD‐positive
cells in the theca compartment (Fig. 2.4f) in the adult ovary. Notably, ovarian Gli1‐positive
cells at P2 gave rise to nearly all cells in the theca cell layer (Fig. 2.4f). Further quantitative
analysis of the lineage‐traced cell number revealed that the mesonephros‐derived Gli1‐
positive cells only contribute to about 1% of the total cells in the adult ovary, whereas the
neonatal ovary‐derived Gli1‐positive cells give rise to around 41% of the total cells in the
ovary (Table 2.1). These results suggest that a small percentage of the theca cells in the
adult ovary are derived from the mesonephros whereas the majority of the theca cells must
come from progenitor cells that originated in the ovarian mesenchyme.
2.4.2 Fetal ovary‐derived Wt1+ cells are the main source of theca cell population
The progenitor cells in the gonadal primordium are thought to be the bona fide source of gonadal somatic cells [310]. When the somatic cell progenitors first appear in the gonad, they express Wilms’ tumor 1 (Wt1), a transcription factor essential for gonadal formation [4, 311]. To determine if the gonad‐derived Wt1‐positive cells contribute to the theca cell lineage, I performed similar tamoxifen‐induced lineage‐tracing experiments by labeling the Wt1‐postive cells in Wt1‐CreERT2; Rosa‐LSL‐tdTomato gonads at the onset of sex determination (E10.5). Twenty‐four hours after tamoxifen treatment (E11.5), the lineage‐labeled Wt1‐positive cells were present in the somatic compartment of fetal ovaries, overlapping with endogenous WT1 protein (Fig. 2.6a & d). At E13.5 and E18.5, fetal ovary‐ derived Wt1‐positive cells gave rise to both FOXL2‐positive cells and FOXL2‐negative cells in the developing ovary (Fig. 2.6b, c, e and f), and were located mostly in the ovarian
38 interstitium delineated by laminin staining (Fig. 2.7a, b, d and e). At one month of age, the
ovary‐derived Wt1‐positive cells were located in the ovarian interstitium surrounding the
follicles and positive for 3HSD (Fig. 2.7c, f). These results indicate that Wt1 marks a
specific pool of steroidogenic precursors in the gonadal primordium before sexual
differentiation occurs.
To further investigate if these fetal ovary‐derived Wt1‐positive cells become Gli1‐ positive theca progenitor cells after birth (Fig. 2.1g & h), I examined the induction of Gli1‐
LacZ expression in the Wt1‐positive cells in postnatal ovaries. When the fetal ovary‐derived
Wt1‐positive cells began to surround the primary follicles in the medulla of the postnatal ovary, they acquired the expression of Gli1, indicating that they have committed to the theca cell lineage (Fig. 2.8). In ovaries from 2‐month old mice, the fetal ovary‐derived Wt1‐ positive cells occupied almost the entire theca cell layer and they were positive for endogenous Gli1‐LacZ (Fig. 2.9).
One thing to note is that the fetal‐ovary derived Wt1‐positive cells give rise to not only FOXL2‐negtive population but also FOXL2‐positive population at fetal and neonatal stages (Fig. 2.6 and Fig. 2.8b). This observation suggests that the fetal‐ovary derived Wt1‐ positive cells should also give rise to adult granulosa cells, however, endogenous tdTomato signals somehow were not observed in granulosa cells in adult ovaries (Fig. 2.7c and Fig.
2.9d). To enhance the tdTomato signals, I used an antibody specific to tdTomato and demonstrated the presence of tdTomato signals in granulosa cells of the developing follicles in adult ovary (Fig. 2.10). This result indicates that first, in addition to theca cells, fetal ovary‐derived Wt1‐positive cells also give rise to granulosa cells in the adult ovary.
Second, it suggests that both granulosa cells and theca cells are derived from the same
39 progenitor population (Wt1‐positive) in the fetal ovary. However, the mechanisms that
specify the two different ovarian somatic cell lineages remain to be explored.
One potential concern regarding the Wt1‐ lineage‐tracing experiments is that Wt1
is not only expressed in the fetal ovary, but also expressed in the fetal mesonephros and
might therefore contribute to adult theca cells as well. To further investigate this
possibility, I examined the Gli1‐LacZ expression in the lineage‐labeled Wt1‐positive cells in
the mesonephros. It was apparent that Gli1‐postive cells and the lineage‐labeled Wt1‐
positive cells in the mesonephros represent almost the same population (Fig. 2.11). This
observation suggests that, even if the lineage‐labeled Wt1‐positive cells in the
mesonephros were to migrate into ovary and became theca cells, they would represent the
same population as the mesonephros‐derived Gli1‐positive cells, and only contribute to a
small percentage of adult theca cell population.
2.4.3 Characterization of the mesonephros‐derived Gli1+ cells
Fetal ovary‐derived Wt1‐positive cells and mesonephros‐derived Gli1‐positive cells appear to represent two different populations of theca cells. Fetal ovary‐derived Wt1‐
positive cells were found in the entire theca in the adult ovary (Fig. 2.9), whereas
mesonephros‐derived Gli1‐positive cells were located preferentially near the basal lamina
(Fig. 2.12), a region that contains predominantly steroidogenic cells [24]. This observation leads to the hypothesis that these mesonephros‐derived Gli1‐positive cells represent a pure population of steroidogenic cells. I isolated mesonephros‐derived Gli1‐positive cells (Fig.
2.4e) and neonatal ovary‐derived Gli1‐positive cells (Fig. 2.4f) by FACS from 2 month‐old mice, and compared their transcriptomes (Fig. 2.13a). The mesonephros‐derived theca
40 cells and the ovary‐derived theca cells shared a common transcriptome profile yet they
exhibit a gene expression profiling distinct from each other. Among 45,101 tested probes,
4,936 were differentially expressed (Fig. 2.13a). Among the 4,936 probes, 2,223 probes
were higher in the ovary‐derived Gli1‐positive cells, including Esr1 (Estrogen receptor 1),
Wt1, and genes implicated in cell growth and proliferation. Many of the signaling pathways
that were activated are not only involved in cell growth and proliferation, but also
associated with progression of PCOS and cancers such as RhoA signaling [312, 313], MAKP
signaling [314, 315], and Wnt/β‐catenin signaling [316‐318] (Table 2.2 and Fig. 2.13a). On the other hand, 2,716 probes were enriched in the mesonephros‐derived Gli1‐positive cells including the genes associated with steroidogenesis, such as Star, Cyp11a1, Cyp17a1, Lhcgr and HSD17b3 (Fig. 2.13b and table 2.2), indicating that the mesonephros‐derived Gli1‐ positive cells are more steroidogenically active than the ovary‐derived Gli1‐positive cells.
The steroidogenic activity of these cells was also confirmed by immunostaining with
3βHSD (Fig. 2.14). These results altogether demonstrate that ovary‐derived Gli1‐positive cells represent a broad mesenchymal cell population(s), whereas the mesonephros‐derived
Gli1‐positive cells contribute to the steroidogenic theca cell population.
2.4.4 Neonatal Gli1+ cells in the ovary are functionally important for ovarian
development
The findings that two sources of theca progenitor pools converge and become Gli1‐ positive cells in the ovary prompted us to examine the functional significance of these Gli1‐ positive cells. I developed a genetic model (Gli1‐CreERT2; Rosa‐DTA) that induces cell death
specifically in the Gli1‐positive cells of the ovary after birth. Because ovarian Gli1‐positive
41 cells consist of cells both from the mesonephros (Gli1‐positive) and the fetal ovary (Wt1‐ positive), this cell ablation model induces cell death in both sources of theca progenitor cells. Tamoxifen treatment induces the expression of DTA (diphtheria toxin fragment A) in the Gli1‐positive cells, which subsequently undergo apoptosis [319]. I administered tamoxifen to the Gli1‐CreERT2; Rosa‐DTA pups and their littermate controls via the milk by injecting lactating dam from PD1 to PD4, when Gli1 expression first appears in the ovary
(Fig. 2.1c, d, g and h). Because Gli1 is also expressed in essential organs such as brain, ablation of Gli1‐positive cells could negatively impact the general development of the animals. To circumvent this potential complication, I removed the ovaries from the pups after 4 days of tamoxifen treatment, transplanted them under the kidney capsule of wild type ovariectomized mice, and examined the ovaries 2 weeks after transplantation (Fig.
2.15a). Ablation of the Gli1‐positive cells resulted in a decrease in ovarian mass and follicle numbers and a complete loss of primary follicles compared to the controls (Fig. 2.15b & c).
Immunostaining for the steroidogenic marker 3βHSD shows decreased expression of steroidogenic enzymes (Fig. 2.15d & e). These results altogether indicate that the Gli1‐ positive theca progenitor cells in the neonatal ovary are indispensable for ovarian development.
42 2.5 Discussion
My study provided the first genetic evidence for the two origins of theca progenitor cells in the mouse ovary: mesonephros (Gli1‐positive) and somatic precursors within the fetal ovary (Wt1‐positive). Theca progenitor cells from divergent sources acquire the expression of Gli1 during their commitment to the theca cell lineage (Fig. 2.16). Further, genetic ablation of the neonatal Gli1‐positive theca progenitor cells results in fewer follicle numbers and renders the ovary smaller in size.
Before P2, Gli1‐LacZ expression was restricted in the mesonephric tubules that connect the rete ovarii of the ovary. The rete ovarii is anastomosing tubule‐like structure lined by cuboidal or columnar epithelium [320], and its formation is thought to be contributed by cells from the mesonephros [321]. This specialized structure interests us because rete ovarii has been suggested an important source of granulosa cells of the developing follicles in mammals such as mink, cat, ferret, mouse and rat [29, 320‐322].
Because Gli1‐LacZ expression seems to spread from mesonephros into ovary via rete‐ovarii,
I therefore hypothesized that Gli1‐positive cells from mesonephros could contribute to somatic cells in the ovary.
In contrast to the long‐standing assumption that theca cells only derive from within the ovarian stroma [48, 54, 55, 57, 299], my results reveal an extra‐ovarian source of Gli1‐positive cells from the mesonephros that contributes to the adult theca cell lineage.
For a long time, the mesonephric contribution to gonadal morphogenesis has been considered a male‐specific event [295], as the migration of mesonephric cells into the XY gonads occurs from E11.5 to E16.5 but this phenomenon is never observed in XX gonads
[323]. The migrating cells in XY gonads appear to be exclusively endothelial cells, which are
43 essential for the formation of testis cords [295, 324]. My finding of the mesonephros‐
derived Gli1‐positive cells contributing to the theca cell lineage in the ovary suggests that such migration event also happens in the female. While the mesonephric cells migrate into testis at E11.5 for the formation of testis cords, the migration of mesonephric cells into ovary does not occur until right before birth, when the process of folliculogenesis is just initiated. One seemingly conflicting fact is that the mesonephros‐derived Gli1‐positive cells migrate into the ovary prior to birth, whereas the endogenous Gli1 expression in the ovary is not observed until after birth. This is because first, lineage labeling of tdTomato to the
Gli1‐expressing cells only occurs at the time of tamoxifein treatment (E12.5), therefore the expression of tdToamto does not reflect the endogenous Gli1 expression later on in the ovary. Second, Hh signaling is known to be inactive in the ovary during fetal stage [279,
280]. This inactive status of Hh signaling is necessary for fetal ovary development, as ectopic activation of Hh signaling in the fetal ovary leads to the appearance of fetal Leydig cells [285].
My present finding that theca cell precursors originate from multiple‐sources raises several important questions. For instance, why, and for what purpose, are two sources of theca progenitor pools needed for ovarian function? Furthermore, what induces the migration of the Gli1‐positive cells from the mesonephros? Because theca cells are only associated with developing follicles, their recruitment is probably regulated by factors from the activated primary follicles [50, 137]. In agreement with this, the mesonephros‐derived
Gli1‐positive cells were seen first in the medulla of the ovary at birth, where the formation of primary follicle occurs. It is not clear why two theca progenitor cell populations are needed for ovarian development. Because theca cells are only associated with developing
44 follicles, one would assume that the presence of two theca progenitor pools is involved in follicular development. Interestingly, two classes of primordial follicles seem to exist in the ovary: the primordial follicles (first wave) activated immediately after birth in the medulla of the ovary, and the primordial follicles (second wave) that are gradually activated later in adulthood in ovarian cortex [21]. The first wave of primordial follicles has long been considered, for the most part, not contributing to fertility because they are depleted from the ovary via atresia [20‐22], whereas the second wave of primordial follicles are expected to be the follicles that ovulate over the entire female reproductive life [20‐23]. Similarly, the granulosa cell precursors that comprise each primordial follicle also appear to be a heterogeneous population. The pre‐granulosa cells specified during fetal stage contribute to the first wave of primordial follicles in the medulla of the ovary immediately after birth.
On the other hand, the granulosa cells recruited in the cortex of postnatal ovaries forming the second wave of primordial follicles later in adulthood [21, 22]. Along the line of evidence, my finding demonstrated that theca cells also arise from two distinct sources. It is therefore possible that the two sources of theca progenitor cells are involved in the two waves of follicular development. In fact, the mesonephros‐derived Gli1‐positive cells are first seen in the medulla of the neonatal ovary where the first wave of follicular development occurs. This observation suggests that the mesonephros‐derived theca progenitor cells may be implicated in the initiation of the first wave of folliculogenesis.
There are a few interesting things to note regarding the mesonephros‐derived Gli1‐ positive cells. First, they migrate into ovary at around time of birth, concomitant with first wave of follicular development. Second, these cells start in few numbers. Third, they give rise to almost exclusively steroidogenic theca cells in the adult ovary. These observations
45 lead me to revisit the idea of the existence of theca stem cells. Previous work by Honda et al
[57] indicated that newborn mouse ovaries contain “putative theca stem cells”. These putative stem cells, when isolated from newborn ovaries, have the capacity to differentiate into androgen‐producing cells in culture, with molecular and cellular characteristics that resemble steroidogenic theca cells. When injected into host ovaries, these cells colonized in the interstitium of the ovary where theca cells are located [57]. Although these neonatal
“theca stem cells” demonstrated great capacity to differentiate into functional theca cells, their true identify was never defined, as these cells were unexpectedly obtained using an isolation procedure optimized to male germline stem cells from neonatal testes [57]. Do the mesonephros‐derived Gli1‐positive cells in the neonatal ovaries represent the “theca stem cells” in vivo? Do they exhibit stem cell properties when cultured in vitro? It would be interesting to investigate the stem cell aspects of the mesonephros‐derived Gli1‐positive cells in future studies.
It is not clear whether mesonephros‐derived Gli1‐positive cells are functionally important for ovarian function. Studies dealing with fetal ovary transplantations provided us some clues. When the fetal ovaries without mesonephroi attached were transplanted under the kidney capsule of recipient female mice, the ovarian grafts exhibited the progression of follicle development till antral stage [325]. In another study, the reaggregated ovarian cells from fetal ovaries, when transplanted into the recipient mice, reorganized into ovarian tissue with different stages of follicle development [325, 326].
These observations suggest that follicle development does not require the presence of mesonephros, although whether long‐term folliculogenesis is contributed by cell from mesonephros is not known. Interestingly, when fetal ovaries without mesonephroi were
46 transplanted into female hosts, they develop into ovotestis at a much higher frequency than the ovaries transplanted with intact mesonephroi attached [327]. This result suggests that fetal mesonephros somehow prevents the sex‐reversal of the ovarian grafts in female hosts.
Based on these observations, I suspect that mesonephros‐derived Gli1‐positive cells are dispensable for ovarian folliculogenesis for at least short‐term fertility; however, they may also be critically involved in other aspects of ovarian functions other than follicular development.
When compare the mesonephros‐derived Gli1‐positive cells and the neonatal ovary‐derived Gli1‐positive cells, their transcriptome profile appear to be distinct from each other. Genes involved in steroidogenesis such as Cyp17a1 are highly expressed in the mesonephros‐derived Gli1‐positive cells. On the other hand, the neonatal ovary‐derived
Gli1‐positive cells enrich for Esr1, Wt1 and genes likely associated with cell proliferation
(Table 2.2). Notably, Esr1 and Cyp17a1 are both specifically expressed in the theca cell layer and are considered theca cell markers. My finding of the differential expression of
Cyp17a1 and Esr1 in two theca cell populations suggests that the heterogeneous theca cell populations might be responsive to different regulatory signaling. The difference in the gene expression profiling likely results from the distinct contributions of the two theca progenitor populations: the mesonephros‐derived Gli1‐positive cells only contribute to steroidogenic theca cells, but the ovary‐derived Gli1‐positive cells give rise to almost all cells in the interstitium of the adult ovaries including steroidogenic theca cells and non‐ steroidogenic cells.
One remain question is why do some ovary‐derived Gli1‐positive cells give rise to steroidogenic cells and some do not? This phenomenon would make more sense if we
47 revisit the endogenous Gli1 expression pattern in the adult ovary. Although Gli1 is specifically expressed in the theca cell layer in the adult ovary, Gli1‐positive cells are nevertheless a heterogeneous population containing both 3βHSD‐positive cells and 3βHSD‐ negative cells. In a similar manner, the lineage‐traced Gli1‐positive cells from neonatal ovary give rise to both steroidogenic positive cells and non‐steroidogenic cells in adult ovary. My interpretation is that Gli1 marks a board theca cell population including both steroidogenic cells and non‐steroidogenic cells. While the steroidogenic cells are mostly derived from the mesonephros‐derived Gli1‐positive cells that locate immediately adjacent to the developing follicles, the non‐steroidogenic cells are predominantly originated from the ovary‐derived Wt1‐positive cells that occupy almost the entire theca layer. This finding is in tune with pre‐existing knowledge that theca cell layer consists of theca interna and theca externa. Theca interna locates right next to the follicles and is highly vascular and steroidogenic, whereas theca externa sits outside the theca interna and is composed of mostly non‐steroidogenic cells such as fibroblast cells and smooth muscle cells important for ovulation [19].
When the two sources of theca progenitor cells are ablated in the neonatal ovaries, the ovaries become smaller in size and exhibit the lack of primary follicles. The smaller size of the ovary is expected because the Gli1‐positive theca progenitor cells are removed from the ovary at birth. These theca progenitor cells acquire the expression of Gli1 when they start to surround the primary follicles. In this sense, the ablation of the Gli1‐positive cells likely disrupt the development of the primary follicles so that primary follicles are missing in the ovarian grafts two weeks after transplantation. Because the tamoxifen administration to the pups is potentially affected by multiple factors such as the dams’
48 metabolism and their lactating ability, and how much milk each pup actually obtained, the cell ablation in the neonatal ovaries is not 100% and that may explain the presence of remaining follicles in the ovarian grafts. It is not yet certain whether the phenotypes of the ovarian grafts are permanent, or it will be restored over time. However, I expect that if given enough time, the Gli1‐positive cell ablated ovaries should recover their folliculogenesis, mostly because 1) the primordial follicles are still present in the Gli1‐ positive cells ablated ovaries; 2) the ovarian surface epithelium can give rise to granulosa cells that comprise new follicles to compensate the loss of follicles in the Gli1‐positive cells ablated ovaries (see Chapter 1); 3) the ablation is not 100% so that the remaining theca progenitor cells might re‐populate over time and resume follicular development.
In summary, my findings of two origins of theca cells and their distinct transcriptomes provide important information regarding theca cell development. I reveal that theca cells stem from two sources during development: precursor cells intrinsic to fetal ovary and the mesenchymal cells in the fetal mesonephros. Their differences in origins and gene expression profiling provide developmental basis and genetic evidence for the heterogeneity and complex of the theca cell population. Disregualtion of theca cells are linked to many ovarian diseases such as polycystic ovarian syndrome and ovarian cancers
[233, 270, 292]. My discovery of theca cell origins not only answers a fundamental question in ovarian biology, but also assists in our understanding the normal process of development and how diseases such as infertility and ovarian cancer may arise.
49 2.6 Figures and Tables
Figure 2.1
50 Figure 2.1: Time course analyses of Gli1 expression in the ovary and mesonephros. a‐ l, Expression of Gli1‐LacZ in the developing ovaries was detected by whole mount beta‐ cfrgalactosidase staining (a‐d) or fluorescent immunohistochemistry for Gli1‐positive cell marker LacZ, granulosa cell marker FOXL2, and nuclear counterstain DAPI in lower magnification (e‐h) and higher magnification images (i‐l). Arrow, mesonephric tubules; E, embryonic day; Ov, ovary; Ms, mesonephros; P, postnatal day. Scale bar: a‐d, 500 m; e‐l,
25 m. m‐p, Transverse sections of E17.5 (m, o) and P6 ovary (n, p) and mesonephros from
Gli1‐LacZ reporter embryos after LacZ staining (blue) in lower magnification (m, n) and higher magnification (o, p) images. The sections were counterstained with nuclear FastRed
(red). Ov, ovary; Ms, mesonephros. Scale bar: 100 m.
51
Figure 2.2: Expression of Gli1‐LacZ in the adult ovary. a‐c, Fluorescent immunohistochemistry for LacZ (green) with granulosa cell marker FOXL2 (red) (a), or steroidogenic cell marker 3HSD (red) (b), or SF1 (red), a regulator of steroidogenesis in
52 Gli1‐LacZ ovaries at 2 months of age. The inset is higher magnification of the outlined area.
Scale bar. 25 m.
53
Figure 2.3: Cre‐mediated tdTomato expression is not observed in the absence of Cre or tamoxifen. a‐c, Tamoxifen or oil was administered at E12.5 and the ovaries were analyzed for tdTomato (red) at birth. tdTomato expression is only detected in Gli1‐CreERT2;
Rosa‐LSL‐tdTomato ovaries that were treated with tamoxifen (c), but not in the Cre‐ negative control (b), nor the oil‐treated control (a). Mesonephroi were removed from all ovaries for imaging. Ov, ovary; Ms, mesonephros. Scale bar: 500 m.
54
Figure 2.4: Lineage‐tracing experiments for the Gli1‐positive cells in the mesonephros and ovary. a‐f, Lineage‐tracing of the Gli1‐positive cells in the Gli1‐CreERT2;
Rosa‐LSL‐tdTomato embryos was induced by tamoxifen (TM) administration at E12.5
(mesonephros‐derived, a‐e), or at P2 (neonatal ovary‐derived, f). The ovaries were examined at different stages of development for tdTomato, steroidogenic marker 3HSD, and DAPI. From E17.5 to birth, mesonephroi were removed from the ovaries for better imaging (b‐d). The staining for 3HSD was observed not only in theca cell layer but also in granulosa cells in large antral follicles (e & f). The insets in (e) and (f) are higher magnification of the outlined areas. Ov, ovary; Ms, mesonephros. Scale bar: a‐d, 500 m; e‐f,
25 m.
55
Figure 2.5: Lineage‐tracing experiments for the mesonephros‐derived Gli1‐positive cells in the mesonephros and ovary. a‐f, The lineage‐tracing of mesonephros‐derived
Gli1‐positive cells in the Gli1‐CreERT2; Rosa‐LSL‐tdTomato embryos was induced by tamoxifen administration from E12.5‐E14.5, and the mesonephroi and ovaries were analyzed at E15.5 and P5 for tdTomato (red), granulosa cell marker FOXL2 (cyan) and
DAPI (Grey). Ov, ovary; Ms, mesonephros. Scale bar: 25 m.
56
Figure 2.6: Lineage‐tracing experiments for the Wt1‐positive cells in the ovary. a‐d,
Lineage‐tracing of the fetal ovary‐derived Wt1‐positive cells in the Wt1‐CreERT2; Rosa‐LSL‐ tdTomato embryos was induced by tamoxifen (TM) administration at E10.5. The ovaries were analyzed at different stages of development by fluorescent immunohistochemistry for tdTomato, WT1, germ cell marker PECAM‐1, granulosa cell marker FOXL2, steroidogenic cell marker 3HSD, and DAPI. E= embryonic day. e‐h, higher magnification of the outlined areas in (a‐d). Scale bar: 25 m. Arrows indicate fetal ovary‐derived Wt1‐positive cells.
57
Figure 2.7: Lineage‐tracing experiments for the Wt1‐positive cells in the ovary. a‐b,
Lineage‐tracing of the fetal ovary‐derived Wt1‐positive cells in the Wt1‐CreERT2; Rosa‐LSL‐ tdTomato embryos was induced by tamoxifen (TM) administration at E10.5. The ovaries were analyzed at different stages of development by fluorescent immunohistochemistry for tdTomato (red), germ cell marker TRA98 (cyan), and laminin that marks basal membrane
(green). E= embryonic day. c‐d, higher magnification of the outlined areas in (a‐b). Scale bar: 25 m.
58
Figure 2.8: Fetal ovary‐derived Wt1‐positive cells acquire the expression of Gli1 in
the perinatal ovary. a, Lineage tracing of the fetal ovary‐derived Wt1‐positive cells in the
Wt1‐CreERT2; Rosa‐LSL‐tdTomato; Gli1‐LacZ embryos was induced by tamoxifen (TM)
administration at E10.5. The ovaries were analyzed at P4 by fluorescent
immunohistochemistry for tdTomato, LacZ for Gli1‐positive cells, granulosa cell marker
FOXL2, and DAPI. b‐e, higher magnification of the outlined area in (a). E= embryonic day;
P= postnatal day. Scale bar: 25 m.
59
Figure 2.9: Fetal ovary‐derived Wt1‐positive cells acquire the expression of Gli1 in
the adult ovary. (a‐d) Lineage tracing of the fetal ovary‐derived Wt1‐positive cells in the
Wt1‐CreERT2; Rosa‐LSL‐tdTomato; Gli1‐LacZ embryos was induced by tamoxifen (TM)
administration at E10.5. The ovaries were analyzed at 2 months of age by fluorescent
immunohistochemistry for tdTomato (b), LacZ for Gli1‐positive cells (c), and DAPI. E= embryonic day. Scale bar: 25 m.
60
Figure 2.10: Fetal ovary‐derived Wt1‐positive cells give rise to granulosa cells in the adult ovary. a‐c, Lineage tracing of the fetal ovary‐derived Wt1‐positive cells in the Wt1‐
CreERT2; Rosa‐LSL‐tdTomato embryos was induced by tamoxifen (TM) administration at
E10.5. The ovaries were analyzed at 1 month of age by fluorescent immunohistochemistry for endogenous tdTomato (red), DsRed for tdTomato‐positive cells (green). d‐f, higher magnification of a‐c. E= embryonic day. Scale bar: 25 m.
61
Figure 2.11: Fetal ovary‐derived Wt1‐positive cells are positive for Gli1 in the fetal mesonephros. a‐e, Lineage tracing of the fetal ovary‐derived Wt1‐positive cells in the Wt1‐
CreERT2; Rosa‐LSL‐tdTomato; Gli1‐LacZ embryos was induced by tamoxifen (TM) administration at E10.5. The mesonephroi were analyzed at E15.5 by fluorescent immunohistochemistry for tdTomato (b), LacZ for Gli1‐positive cells (c), germ cell marker
TRA98 (d), and DAPI (e). E= embryonic day. Ov, ovary; Ms, mesonephros. Scale bar: 25 m.
62
Figure 2.12: The mesonephros‐derived Gli1‐positive cells are located immediately
adjacent to the basal membrane in adult ovary. Lineage tracing of the Gli1‐positive cells in the Gli1‐CreERT2; Rosa‐LSL‐tdTomato embryos were induced by tamoxifen (TM)
administration at E12.5. The ovaries were examined at 2 months of age for tdTomato,
3HSD, and DAPI. Yellow dotted line indicates the basal membrane that separates the theca
cell layer from granulosa cells.
63
Figure 2.13: Characterization of the mesonephros‐derived Gli1‐positive cells. a, Gene expression profiles of 3 independent pools of the mesonephros‐derived (Tamoxifen at
E12.5, E14.5 and E16.5) and the neonatal ovary‐derived Gli1‐positive cells (Tamoxifen at
P1‐3 via lactating dams). All cells were isolated from the adult ovaries of Gli1‐CreERT2;
Rosa‐LSL‐tdTomato mice at 2 months of age. b, qPCR analysis of Gli1 and markers of theca cell steroidogenesis (Star, Cyp11a1, Cyp17a1, and Lhcgr) in the mesonephros‐derived and the neonatal ovary‐derived Gli1‐positive cells. *P < 0.05.
64
Figure 2.14: Mesonephros‐derived Gli1‐positive cells are steroidogenically active in the adult ovary. a‐c, The lineage‐tracing of the mesonephros‐derived Gli1‐positive cells in
Gli1‐CreERT2; Rosa‐LSL‐tdTomato embryos was induced by tamoxifen administration from
E12.5‐14.5 and the ovaries were analyzed at 2 months of age for tdTomato (red) and steroidogenic cell marker 3HSD (green). Scale bar: 25 m.
65
Figure 2.15: Effects of genetic ablation of neonatal Gli1‐positive theca progenitor
cells on development of the ovary. a, Timeline of cell ablation experiment with time
points for tamoxifen (TM) administration, transplantation, and tissue collection.
PD=Postnatal day. b‐c, Hematoxylin and eosin staining on the sections of control (b) or Gli1‐
positive cells ablated ovarian transplants (c). d‐e, Immunofluorescence analysis was
performed on the sections of the transplants for steroidogenic cell marker 3HSD (green) and DAPI (grey). Scale bar: 100 m.
66
Figure 2.16: Proposed model for the origins of theca cells in the mouse ovary. Theca
cells derive from two sources: Wt1+ cells indigenous to the ovary and Gli1+ mesenchymal cells migrated from the mesonephros around time of birth.
67
Timing'of'lineage1 Percentage'of'tdTomato+'' Percentage'of'tdTomato1' Target'cell'type' Time'of'collection' labeling' cells'in'the'ovary' cells'in'the'ovary'
Mesonephros)derived- TM-at-E12.5,-E14.5,-E16.5- 2-months- 1.03%- 98.97%- Gli1)positive-cells-
Ovarian-Gli1)positive- TM-at-P1,-P2,-P3- 2-months- 41.07%- 58.93%- cells- E: Embryonic day. TM: Tamoxifen
Table 2.1: Comparison of theca cell contributions between mesonephros-derived and neonatal ovary-derived Gli1-positive cells. Lineage-labeling of the Gli1-positive cells in the Gli1-CreERT2; Rosa-LSL-tdTomato embryos was induced by tamoxifen (TM) administration from E12.5-E16.5 (mesonephros-derived), or from P1-3 (neonatal ovary- derived). The ovaries from 2 months of mice were subjected to fluorescence-activated cell sorting for the percentage of tdTomato+ and tdTomato- cells in the ovary.
68
Fold change Genes of interest (Ovary derived cells vs. mesonephros‐derived cells)
Pgr 4.113
Wt1 2.374
Esr1 2.881
Bmp2 2.029
Cxcl2 3.800
Cxcl3 4.890
Cxcl12 2.487
Foxo3 1.869
HSD17b3 ‐1.549
Table 2.2. Fold changes of genes of interest from microarray analysis of ovary‐derived
Gli1‐positive cells versus mesonephros‐derived Gli1‐positive cells from 2 months of age mice.
69 CHAPTER 3: Role of Desert/Indian Hedgehog Signaling in Theca Cell Development in
the Murine Ovary
3.1 Abstract
Establishment of the androgen‐producing theca cell lineage is critical for folliculogenesis and ovulation of the mammalian ovaries. In Chapter 2, I have found that theca cells derived from two origins: the precursor cells (Wt1‐positive) intrinsic to the fetal ovary and the mesenchymal cells (Gli1‐positive) in the mesonephros. However, the signaling pathway that directly specify theca cell lineage is not clear. One clue is that when the two sources of theca progenitor cells come to surround the primary follicles in the postnatal ovary, they acquire the expression of Gli1, indicating that Gli1 is involved in the specification of theca cell lineage. To determine what regulates Gli1 expression in the ovary, neonatal ovaries were cultured in the presence of a Hedgehog (Hh) signaling inhibitor cyclopamine. Cyclopamine treatment suppressed the expression of Gli1 in the ovary, indicating that Gli1 expression in the neonatal ovary is activated by the canonical Hh pathway. The Hh ligands that were produced in the neonatal ovaries appeared to be Dhh
and Ihh. To further examine what cell type(s) is responsible for Hh ligands production, I
isolated granulosa cells and theca progenitor cells from postnatal ovaries. Dhh and Ihh were highly expressed in granulosa cells, whereas their downstream targets Gli1 and Ptch1 were enriched in theca progenitor cells. This observation suggests a novel regulation from granulosa cells to theca progenitor cells via Hh signaling. To investigate whether Dhh and
Ihh play a functional role in theca cell development, I generated Dhh/Ihh double knockout
(DKO) mice in which the expression of Dhh and Ihh are specifically ablated in gonadal
70 somatic cells. Ovaries lacking Dhh/Ihh were smaller in size, exhibiting follicle arrest at pre‐ antral stage and the absence of corpora lutea. In addition, steroidogenic theca cells were completely lost, along with a significant down‐regulation of steroidogenic genes. Consistent with the lack of Gli1‐positive theca cells and defects in follicular development, the levels of circulating hormones including dehydroepiandrosterone (DHEA), testosterone, and progesterone in the DKO female are decreased. In summary, these studies demonstrate a critical role of granulosa cell‐derived DHH/IHH signaling in regulating theca cell differentiation in the developing ovary.
71 3.2 Introduction
Theca cells in the mouse ovary do not become morphologically distinguishable
until one week after birth when the follicles develop to secondary stage [19]. These cells
play diverse roles during ovarian development, such as synthesizing androgens, which are
then converted to estrogen by granulosa cells [44‐48]; interacting with granulosa cells and
oocytes during folliculogenesis [49‐51]; and providing structural support and blood supply
for the developing follicles for ovulation [50, 52, 53]. In addition to the function of theca
cells in normal ovarian development, dysfunction of androgen production are implicated in
pathogenesis of many ovarian diseases, such as polycystic ovarian syndrome (PCOS),
premature ovarian failure and ovarian cancers [233, 262, 270, 292, 328]. Although theca
cells play such a critical role in ovarian development, the mechanisms that regulate their
differentiation remain poorly understood. Because theca precursor cells do not express LH
receptors [329], and theca layer still forms in FSH‐deficient female mice [158], the initial
theca cell differentiation/recruitment is likely gonadotropin‐independent.
The factors that regulate theca cell differentiation appear to be small‐molecular‐ weight proteins ranged from 20‐ to 24‐ kDa secreted from granulosa cells of the growing follicles [60]. This is because that theca cells are only associated with developing follicles, and their histological proximity to granulosa cells led to the hypothesis that differentiation of theca cells is regulated by paracrine factor(s) produced by granulosa cells [137]. In fact, when undifferentiated theca‐stromal cells (or putative theca stem cells), were cultured with granulosa cells, or granulosa cell‐conditioned culture medium, markers of theca differentiation, such as LH receptors and steroidogenic enzymes were expressed and androgens produced [57, 58, 60, 61]. Candidate factors that may contribute to theca cell
72 differentiation include IGF, KIT ligand, GDF9 and Hh ligand. Addition of IGF‐I or KIT ligand in rat theca cell culture is able to increase expression of some but not all theca cell makers
[330]. In Gdf9 knockout ovary, the theca layer is not formed, yet it is unclear whether GDF9 regulates theca cell differentiation directly or indirectly through regulation of granulosa cell development [146]. Among those potential factors, Hh ligand interests me the most because of the role of DHH in the development of fetal Leydig cells in mouse testis [281], as
Leydig cells are considered the male counterpart of theca cells in the ovary. Both in vivo and in vitro studies have demonstrated that DHH derived from Sertoli cells regulates the specification and appearance of fetal Leydig cells [281], and one of the downstream effectors involved in this process is Gli1 [280, 281]. In a similar pattern, Dhh and Ihh are expressed in granulosa cells of the developing follicles in adult mouse ovary [286, 287], whereas their downstream targets Ptch1 and Gli1, are localized specifically in the mesenchymal stromal cells surrounding the follicles [287]. Whether Dhh and Ihh together play a functional role during ovarian development is not known [282, 288]. Constitute activation of Hh signaling in the ovary results in defects in smooth muscle cell development and subsequent failure in ovulation, indicating that appropriate Hh signaling is required for proper ovarian function. In addition, recombinant SHH is able to stimulate androstenedione production of the bovine theca cells in culture [290]. With all these observations mentioned above, I therefore hypothesize that Hh signaling from granuosa cells is important for theca cell development and proper ovarian functions.
73 3.3 Materials and Methods
3.3.1 Animals breeding
Gli1‐LacZ (#008211), Gli1‐CreERT2 (#007913), Foxl2‐CreERT2 (#015854), Rosa‐LSL‐
tdTomato (#007905), Dhh+/‐ (#002784), Ihh+/‐ (#004290) mice were purchased from the
Jackson Laboratory (Bar Harbor, ME). Sf1‐Cre and Ihh floxed/floxed mice were kind gifts from
Dr. Keith Parker (UT Southwestern Medical Center) and Dr. Francesco DeMayo (Baylor
College of Medicine), respectively. Female mice were housed with male mice overnight and checked for the presence of vaginal plug the next morning. The day when the vaginal plug was detected was considered embryonic day (E) 0.5. All animal procedures were approved by the National Institute of Environmental Health Sciences (NIEHS) Animal Care and Use
Committee and are in compliance with a NIEHS‐approved animal study proposal. All experiments were performed on at least three animals for each genotype.
3.3.2 Isolation of granulosa cells and theca progenitor cells
Cell lineage labeling was conducted by administrating tamoxifen (T‐5648, Sigma‐
Aldrich, St. Louis, MO) to Gli1‐CreERT2; Rosa‐LSL‐tdTomato or Foxl2‐CreERT2; Rosa‐LSL‐
tdTomato neonatal pups through lactating dams. The pups were fostered to CD‐1 dams
immediately after birth followed by tamoxifen administration. 5 mg tamoxifen dissolved in
100 μl corn oil (C‐8267, Sigma‐Aldrich, St. Louis, MO) was injected to the foster dam once
per day for 2 consecutive days (PND1‐2). No overt teratological effects were observed after
tamoxifen administration under these conditions.
Neonatal ovaries from Gli1‐CreERT2; Rosa‐LSL‐tdTomato or Foxl2‐CreERT2; Rosa‐
LSL‐tdTomato pups were minced and enzymatically dissociated in Hank’s balanced salt
74 solution (Sigma) with 1 mg/ml collagenase B (Roche), 1.2 U/ml Dispase II (Roche) and 5
U/ml DNAse I (Sigma) for 20 min at 37 oC. To further aid tissue dissociation, the solution
was mixed by repeated pipetting (P1000 pipette) [303]. Dissociated cells were rinsed and
re‐suspended in ice‐old PBS and were subject to fluorescence‐activated cell sorting (FACS)
for tdTomato+ and tdTomato‐ cells. FACS was performed using a FACSAria II flow cytometer
from Becton Dickinson (San Jose, CA) and FACSDiVa software version 6.1.3.
3.3.3 Immunohistochemistry and histological analysis
Immunohistochemistry and histological analysis were performed as described with modifications[331]. Tissues were paraffin‐embedded and the sections were dewaxed and rehydrated in a series of alcohol to PBS. The slides were pretreated in 0.1 mM citric acid for 20 min in the microwave, incubated with primary antibodies (listed in the table below) in PBS‐Triton X‐100 solution with 5% normal donkey serum at 4°C overnight. The sections were then washed and incubated in the appropriate secondary antibody (1:500;
Invitrogen, USA) before mounting in Vector Mount with DAPI (Vector Labs). For anti‐
COUPTFII antibody, a M.O.M kit was used following the manufacturer protocol (BMK‐2202,
Vector labs). Slides were imaged under a Leica DMI4000 confocal microscope. For histological analysis, the samples were fixed in 4% paraformaldehyde in PBS at 4°C overnight, dehydrated through an ethanol gradient, and embedded in paraffin wax.
Sections were stained with hematoxylin/eosin or PAS/ hematoxylin, and were scanned using Aperio ScanScope XT Scanner (Aperio Technologies, Inc., CA, USA) for digital image analysis.
75 Primary antibodies Species Dilution Source
FOXL2 Rabbit 1:500 Dr. Dagmar Wilhelm, Monash University, Australia
FOXL2 Goat 1:500 Imgenex, Littleton, CO, USA
αSMA Rabbit 1:500 Abcam, San Francisco CA, USA
CYP17A1 Goat 1:100 Santa Cruz Biotechnology, Texas, USA
COUPTFII Mouse 1:200 R&D systems. Minneapolis, MN, USA
3.3.4 Gene expression analysis
Gene expression analysis was performed as described with modifications [304].
Total RNA was isolated from ovaries using the PicoPure RNA isolation kit (Arcturus,
Mountain View, CA) according to the manufacturer’s protocol. The cDNA preparation was synthesized using random hexamers and the Superscript II cDNA synthesis system
(Invitrogen Corp., Carlsbad, CA) following the manufacturer’s instruction. Gene expression was analyzed by real‐time PCR using Bio‐Rad CFX96TM Real‐Time PCR Detection system.
Taqman gene‐expression probes (Gli1: Mm00494654_m1, Ptch1: Mm00436026_m1, Ihh:
Mm00439613_m1, Shh: Mm00436528_m1, Dhh: Mm01310203_m1, Foxl2:
Mm00843544_m1, Star: Mm00441558_m1, Cyp11a1: Mm00490735_m1, Hsd3b1:
Mm01261921_mH, Nr5a1: Hs01018738_m1, Cyp19a1: Mm00484049_m1) were used to detect the fold changes of the transcripts. Fold changes in gene expression were determined by quantitation of cDNA from target (treated) samples relative to a calibrator sample (control). All real‐time PCR analyses were performed in duplicate, and the results were reported from at least three independent experiments. The relative fold change of transcript was calculated using the mathematical model of Pfaffl as previously described
76 [305] and was normalized to 18S rRNA (Mm03928990_g1) as an endogenous reference.
3.3.5 Microarray analysis
Gene expression analysis was conducted using Affymetrix Mouse Genome 430 2.0
GeneChip® arrays (Affymetrix, Santa Clara, CA). Fifty (50) ng of total RNA was amplified as directed in the WT‐Ovation Pico RNA Amplification System protocol, and labeling with biotin following the Encore Biotin Module. 4.6 μg of amplified biotin‐aRNAs were fragmented and hybridized to each array for 18 hours at 45°C in a rotating hybridization.
Array slides were stained with streptavidin/phycoerythrin utilizing a double‐antibody staining procedure and then washed for antibody amplification according to the GeneChip
Hybridization, Wash and Stain Kit and user manual following protocol FS450‐0004. Arrays were scanned in an Affymetrix Scanner 3000 and data was obtained using the GeneChip®
Command Console Software (AGCC; Version 1.1) using the MAS5 algorithm to generate.
CHP files. The microarray raw data were analyzed using the Partek Genomic Suite Software
(St. Louis, Missouri). Two‐way Anova was performed to determine the statistical significance between the means. The gene list was filtered with fold‐change cutoff of 1.5 and an unadjusted p<0.05.
3.3.6 Organ culture
Ovaries with mesonephroi attached from E18.5 Gli1‐LacZ embryos were cultured for 3 days on Millicell‐PC membrane inserts (0.4 μm pore size; 30 mm diameter; Millipore corp., Medford, MA) with medium filling only the lower chamber [332]. 8‐10 ovaries were placed on each membrane insert, with one drop of medium on each ovary. For culture with
77 the Hh inhibitor, cyclopamine (25 μM, Sigma) [333], or an equal volume of vehicle,
dimethyl sulfoxide (DMSO), was added to the culture medium at the beginning of the
culture for 3 days and the culture medium was changed every other day. The medium for
organ culture was DMEM/F12 (1:1) + L‐Glutamine + 15mM HEPES (Invitrogen) supplemented with 50 μg/ml penicillin G/streptomycin sulfate, and 5% (vol/vol) fetal bovine serum (FBS). After 3 days of culture, the ovaries were collected and processed for
LacZ staining as described in Chapter 2.
3.3.7 Steroid Hormone Multiplex Immunoassay
Serum samples were collected at various times and stored at ‐80°C until further analyzed. Mice were not staged for estrus cycle at the time of sample collection. Serum samples were assayed for DHEA, Estradiol, Progesterone, and Testosterone using the
Steroid Hormone Panel kit (cat # N45CB‐1) from MSD (Meso Scale Discovery, Gaithersburg,
Maryland, USA) according to manufacturer’s protocols. The Steroid Hormone Panel utilizes a competitive immunoassay format. Briefly, a 96 well custom designed plate that had been precoated with capture antibodies on spatially distinct spots was blocked with blocking solution for 1 hr. at room temperature with constant shaking at 700 rpm. After washing 3x with PBS‐T buffer, samples and standards were added to appropriate wells of the MSD plate and incubated for 2 hrs at RT with constant shaking. After incubation of the samples, a mixture of MSD SULFOTAGTM labeled tracers are added to each well. The tracers compete
with analytes in the samples for binding on the immobilized antibodies generating a signal
for each assay this is inversely proportional to the analyte concentration. After incubating
with the tracers, the plate was washed again 3x with PBS‐T, pat dried, and 150 μl of 1x
78 Read Buffer was added to each well. The plate was immediately analyzed on the Sector
Imager 2400 System (MSD). The instrument measures intensity of emitted light to afford a
quantitative measure of DHEA, Estradiol, Progesterone, and Testosterone in the sample.
Testosterone standard is not supplied with the MSD kit, it was purchased separately from
Steraloids (cat # A6950) (Newport, RI, USA) and used in the range from 16 ng/mL to 0.1
ng/mL. All the samples were run on the same plate. The intra‐array %CV (n=7) is 3.99 for
DHEA, 7.63 for testosterone and 13.32 for progesterone.
3.3.8 Statistical analysis
Data were analyzed using Prism (Version 6, GraphPad Software) by two‐tailed
Student’s t‐test. Values are expressed as mean±s.e.m. A minimal of 3 biological replicates was used for each experiment.
79 3.4 Results
3.4.1 Expression of Gli1 in the neonatal ovary is under the control of canonical Hh pathway
As presented in Chapter 2, Gli1 was a lineage marker for adult theca cells and the expression of Gli1 was not present in the fetal ovary until right after birth. The identification of Gli1 as a theca cell lineage marker raises the question: what signal(s)
specify theca cell lineage and induce Gli1 expression? Gli1 is a known downstream target of
the Hh pathway [334]. I first examined whether activation of the Hh pathway is responsible for inducing Gli1 expression by culturing fetal ovaries in the presence of the Hh inhibitor cyclopamine [281, 333]. Cyclopamine inhibits Hh signaling by inactivating SMO, the downstream Hh signal transducer [333]. The ovaries were cultured for 3 days starting
E18.5, before Gli1 expression occurred in the ovary (Fig, 3.1a). Gli1‐LacZ was expressed in the ovaries after 3 days culture as expected in the absence of cyclopamine (Fig. 3.1b). When cyclopamine was added in culture, the expression of Gli1‐LacZ in the ovaries was completely abolished (Fig. 3.1c). These results indicate that Gli1 expression is induced by the canonical Hh pathway.
3.4.2 Dhh and Ihh are produced primarily in granulosa cells in the ovary
In search of the Hh ligands that may be responsible for activating the Hh pathway
in the neonatal ovaries, I examined mRNA expression of Dhh, Ihh, and Shh in the ovary before (E17.5) and after (PND3) the appearance of Gli1. Expression of Gli1 was at basal level before birth but dramatically up‐regulated at PND3 (Fig. 3.2a). In the similar pattern, expression of Dhh and Ihh was low in the ovary before birth and significantly increased at
80 PND3 (Fig. 3.2a). Notably, expression of Shh was undetectable in ovaries at both stages (Fig.
3.2b), suggesting that Dhh and Ihh, but not Shh, are responsible for the onset of Gli1
expression in the neonatal ovary. The specificity of the probes was validated using different
tissues: testis (negative for Shh and positive for Dhh), adrenal (positive for Shh), limb bud
(positive for Shh and Ihh and negative for Dhh) (Fig. 3.2b).
To identify which somatic cell type(s) produce Dhh and Ihh, I isolated theca progenitor cells and granulosa cells from neonatal ovaries and examined the expression of
Dhh and Ihh in the isolated cell populations. This experiment was accomplished by using
Gli1‐CreERT2; Rosa‐LSL‐tdTomato and Foxl2‐CreERT2; Rosa‐LSL‐tdTomato lineage tracing genetic models, in which ovarian Gli1‐positive cells were obtained for theca progenitor cells and Foxl2‐posiitve cells for granulosa cells, respectively. These ovarian somatic cells were labeled with tdTomato via tamoxifen administration from PND1 to PND2, and the ovaries were collected and processed for FACS at PND5 (Fig. 3.3a). The purity of the isolated tdTomato‐positive cells was validated with their high expression for Gli1 and Foxl2, respectively (Fig. 3.3a). Notably, tdTomato‐negative cells also expressed Gli1 and Foxl2 but at a much lower level compare to that of tdTomato‐positive cells. This was probably due to the fact that only a portion of Gli1‐ and Foxl2‐positive cells were labeled with tdTomato through 2 days of tamoxifen treatment. The remaining unlabeled Gli1‐ and Foxl2‐positive cells therefore contributed to the Gli1 and Foxl2 expression in the tdTomato‐negative cells
(Fig. 3.3a). Further analysis of the isolated ovarian somatic cells revealed that, the Hh downstream targets Gli1 and Ptch1 were highly enriched in theca progenitor cells, whereas
Hh ligands Dhh and Ihh were predominantly expressed in granulosa cells (Fig. 3.3b). These
81 results implicate a novel regulation of theca cell differentiation by granulosa cells through
the Hh signaling pathway.
3.4.3 Loss of Ihh and Dhh results in defective folliculogenesis in the ovary
Female Dhh‐deficient mice are fertile and exhibit normal ovarian development
[282], suggesting that the other Hh ligands, such as Ihh, may compensate for the loss of Dhh during folliculogenesis. Most of the Ihh‐deficient embryos die before birth [288], therefore precluding the analysis of the ovary in the adult. To investigate whether Dhh and Ihh together regulate theca cell development, I generated Dhh/Ihh double knockout mice
(Hereafter referred as Dhh/Ihh DKO) in which Dhh and Ihh was ablated from Sf1‐positive gonadal somatic cells (Sf1‐Cre; Ihh f/‐; Dhh ‐/‐) [335]. Loss of Dhh, Ihh, and Gli1 expression in
the single and double KO ovaries confirmed the efficiency of the knockouts and the
inactivation of the Hedgehog pathway (Fig. 3.4). Ovaries deficient in either Dhh or Ihh alone
appeared to exhibit normal folliculogenesis, evident by different stages of follicle
development as well as the presence of corpora lutea (CLs) (Fig. 3.5a‐f). In addition, the
theca cell layer is formed, as determined by immunostaining with α‐SMA, a marker for
smooth muscle cells in theca cell layer (Fig. 3.5g‐i) [289]. However, ovaries from Dhh/Ihh
DKO female mice appeared significantly smaller in size and irregular in shape compared to
the control (Fig. 3.6a & b). In contrast to the normal progression of folliculogenesis and the
presence of corpora lutea in adult control ovaries (Fig. 3.6c), Dhh/Ihh DKO ovaries lacked
corpora lutea, and the follicles failed to progress beyond preantral follicle stage, suggesting
that ovulation did not occur (Fig. 3.6d, g and h). Antral follicles were rarely found in the
DKO ovaries and if they were present, they appeared cystic and hemorrhagic (Fig. 3.6e & f).
82 The theca layer was almost nonexistent in ovaries lacking Dhh and Ihh, as determined by
immunostaining with α‐SMA, a marker for smooth muscle cells in theca cell layer (Fig. 3.7a
& b) [289]. Furthermore, the expression of 3βHSD and CYP17A1, the key enzymes in
androgen biosynthesis, was undetectable in the ovarian mesenchyme, confirming the
absence of androgen‐producing theca cells (Fig. 3.7c‐f). Consistent with these findings,
levels of circulating testosterone and dehydroepiandrosterone (DHEA) were reduced in the
absence of Dhh and Ihh CLs (Fig. 3.8). In addition, the level of circulating progesterone was also decreased in Dhh/Ihh DKO female, likely as a result of the absence of CLs (Fig. 3.8).
Although the loss of Dhh and Ihh abolished a large number of cells in the theca layer, there
were still some interstitial cells survived the ablation of Hh signaling. To examine their
identify, I stained the ovarian section with COUP‐TFII, an orphan nuclear receptor with
suggested roles in adult Leydig cell differentiation in testis [336, 337]. COUP‐TFII is
normally expressed in the ovarian interstitium including the cells that immediately
adjacent to the follicles (Fig. 3.9a & c). In the absence of Dhh and Ihh, COUP‐TFII remained
expressed in the interstitium of the ovary and in cells adjoining the granulosa cells (Fig.
3.9b & d).
To further investigate the transcriptome changes in the DKO ovaries, I performed
microarray analysis with DKO ovaries and control ovaries (Dhh het/Ihh het) (Fig. 3.7a).
Among 1393 differentially expressed probes, 636 probes were up‐regulated in the DKO ovary including genes involved in follicular growth and granulosa cell differentiation such as Gata6, Wt1, Igf1 (table 3.1). On the other hand, 757 probes were down‐regulated in the
DKO ovary and many of the them were associated with Hedgehog pathway and ovarian steroidogenesis, such as Nr5a1, Star, Cyp11a1, Hsd3b1 and Cyp19a1 (Fig. 3.10b). The
83 significant decrease of genes directly involved in androgen production including Star,
Cyp11a1 and Hsd3b1 further confirms the defects of theca cell steroidogenesis at the molecular level (Fig. 3.10b).
3.5 Discussion
My studies demonstrated the critical role of Dhh and Ihh during theca cell differentiation. Dhh and Ihh are produced from granulosa cells of developing follicles soon
after birth, and the Hh ligand production in turn induces Gli1 expression in the interstitial
cells of the neonatal ovary. In the absence of Dhh and Ihh, androgen‐producing theca cells are missing, follicle development are arrested at pre‐antral stage along with blunted steroid production.
It is not clear why two Hh ligands are needed in granulosa cells for theca cell differentiation. Ovaries deficient for either Dhh or Ihh exhibit normal folliculogenesis,
whereas ovaries deficient for both Dhh and Ihh show completely loss of functional theca
cells and follicle arrest. These observations indicate that Dhh and Ihh from granulosa cells
play a redundant role during folliculogenesis. Although the role of Dhh and Ihh during follicle development seems redundant, they appear to have different capacity in regulating
Gli1 expression in the ovary. In the absence of Dhh alone, Gli1 expression was unaffected, whereas loss of Ihh alone significantly decreased the expression of Gli1 in the ovaries.
These results suggest that even though Dhh and Ihh play a redundant role during theca cell differentiation, Ihh might be a predominant factor in this case. In chapter 2, I demonstrated that theca cells surrounding the developing follicles are heterogeneous as they stem from
84 two sources: the somatic precursor cells in the fetal ovary and the mesenchymal cells in the fetal mesonephros. Do the two Hh ligands play different roles in regulating the two theca cell populations? Because the two theca cell populations (ovary‐ and mesonephros‐derived) exhibit distinct transcriptomes in the adult ovary, a further microarray comparison between their gene expression profiling and the transcriptome changes in the absence of
Dhh and Ihh might shed light on the interactions between the two Hh ligands and the two theca cell populations.
When Dhh and Ihh were absent, morphologically evident preantral follicles were present yet the theca cell differentiation failed to occur. This result indicates that growth from primordial to secondary follicle does not require steroidogenic theca cells. In addition to the absence of functional theca cells, Dhh/Ihh DKO ovaries also exhibited a dramatic loss of interstitial‐stroma cells. In particularly, vacuoles‐like structure was found in the medulla of the adult ovary and in the intra‐ovarian rete ovarii. These vacuoles appeared to be degenerated follicles, as some of them still contained oocytes but the surrounding somatic cells were absent. The degeneration of follicles in the medulla of the ovary is reminiscent of the Gli1 expression pattern in the neonatal ovary, as Gli1 first appears in the center of the ovary where the primary follicles start to form. As I reviewed in Chapter 1, two waves of follicular development have been suggested during ovarian development: the first wave of follicle development initiates in the medulla of the ovary at time of birth, and the second wave of follicle development occurs in the ovarian cortex postnatally with the cells derived from surface epithelium [20‐22]. Are these degenerated follicles in the DKO ovary the ones from the first wave of follicle development? Are the remaining pre‐antral follicles in the ovarian cortex derived from the surface epithelium? Because the expression of Gli1 is
85 almost abolished in the Dhh/Ihh DKO ovaries, one would assume that Hh signaling is
completely inactivated in the DKO ovary. This observation raises up two possibilities
concerning the remaining follicles in the cortex of the ovary: first, these follicles do not
express Dhh and Ihh under normal physiological circumstances and therefore their growth is independent of Hh signaling. However, because of the lack of theca‐stromal cells, these follicles are unable to develop further. Second, the remaining follicles in ovarian cortex do express Dhh and Ihh normally, but the Hh signaling do not become critical for folliculogenesis until these follicles proceed beyond pre‐antral stage. One following question is that if the animals were allowed to live longer, would the continual degeneration of follicles lead to follicular exhaustion and the ovaries exhibit a phenotype that resembles premature ovarian failure? Further analysis of the DKO ovaries at different developmental stages (e.g. P5, P21 and 6 months of age) should be able to answer at least some of these questions.
The defective folliculogenesis shown in the Dhh/Ihh DKO ovaries somewhat resemble what was observed in Cyp19a1 KO and αβER KO (mice lacking both ERα and ERβ) ovaries [338, 339], in which most follicle only reach small antral stage, large antral follicles that remain are cystic and hemorrhagic, and CLs are absent. The ovarian phenotypes in
Cyp19a1 KO and αβER KO ovaries suggest a role of estrogen signaling on folliculogenesis.
In line with these findings, the expression of Cyp19a1 (necessary for converting androgens into estrogens) in the Dhh/Ihh DKO ovaries is also dramatically decreased compared to that of control ovaries. This result suggests the possibility that the follicular defects in DKO ovary is at least partly due to insufficient estrogen production. This speculation is probably true, given that the level of circulating androgens in the Dhh/Ihh DKO females is also
86 significantly reduced. Considering the low androgen level and reduced expression of
Cyp19a1, it is likely that the DKO ovary cannot produce enough estradiol necessary for the
maintenance of antral follicle and ovulation [19]. Further measurement of the levels of
estradiol, FSH and LH in the DKO female mice might help us to identify the causes for the
defects in follicular development. In addition to the follicular arrest and lack of CLs, sex‐
reversed follicles were observed with the appearance of Sertoli‐like cells in both Cyp19a1
KO and αβER KO ovaries [171, 340]. Interestingly, the expression of Sox9, a male specific gene and a marker for Sertoli cells in testis [14], was also up‐regulated in the DKO ovaries by microarray analysis. It would be interesting to see if SOX9 protein is actually present in the DKO ovaries, and if seminiferous tubule‐like structures eventually develop in the ovaries as the mice age.
Ablation of Dhh and Ihh resulted in a loss of a large number of interstitial cells but
not all of them. It was shown that the cells remained in the interstitium of DKO ovary were
positive for orphan nuclear receptor COUP‐TFII. This observation is interesting to me
because disruption of COUP‐TFII in prepubertal male mice results in arrested adult Leydig
cell differentiation at progenitor cell stage [337]. Additionally, in the fetal testis, COUP‐TFII
marks an interstitial cell population other than fetal Leydig cells [336]. These results
suggest that COUP‐TFII‐positive cells in the fetal testis might be the stem cells for adult
Leydig cells. Although androgen‐producing theca cells were lost in the DKO ovary, it is
possible that theca progenitor/stem cells were still present in the ovarian interstitium. Are
these remaining COUP‐TFII‐positive cells the stem cells for theca cells? Do these COUP‐
TFII‐positive cells differentiate into functional theca cells when isolated in culture? Or, do
these cells represent the “third” sub‐population of theca cells? Because the cells responsive
87 to Hh signaling should express Gli1, one would assume that these remaining COUP‐TFII‐
positive cells are negative for Gli1 and are independent of Hh signaling. In Chapter 2, Gli1
was shown to be a lineage maker for differentiating theca cells. If this is the case, are the
remaining COUP‐TFII‐positive cells in an undifferentiated state? This hypothesis brings up
the possibilities that the ovarian interstitial cells might have different developmental status
in terms of their differentiation pathway. The first status is a naïve state and the cells are
not responsive to extracellular cues. The second status is that the cells become responsive
to cellular signals, for example, they acquire Gli1 in respond to Hh signaling in this case. The third status is that the cells take on a differentiation trajectories and becoming terminally
differentiated cells, such as the steroidogenic theca cells. One concern with my current
study is that, although I have shown that androgen‐producing theca cells were lost in the
absence of Hh signaling, I cannot distinguish whether this is because Hh signaling directly
regulates the process of steroidogenesis, or it is simply because the inactivation of Hh
signaling leads to the cells being not responsive to cellular signals for steroidogenesis.
Potential strategies to test if Hh signaling directly regulates theca‐steroidogenesis are: 1) to
inactivate Hh signaling after theca progenitor cells already acquires Gli1 expression, and
examine if further steroidogenic differentiation of the cells require Hh signaling. 2) Treat
the DKO female mice with exogenous gonadotropins to examine if the remaining interstitial
cells differentiate into steroidogenic theca cells in the absence of Hh signaling.
In the absence of Dhh and Ihh, steroidogenic theca cells were lost and the follicles
were arrested at pre‐antral stage. Consistent with the histological observations, microarray
analysis of Dhh/Ihh DKO and the control ovaries showed down‐regulation of genes
associated with theca‐steroidogenesis. Along the line with the lack of CLs phenotype, genes
88 involved in follicle development and ovulation were also down‐regulated, such as Lepr
(Leptin receptor) [341], vefga (Vascular endothelial growth factor A) [342‐344], and Cebpb
(CCAAT/enhancer‐binding protein‐β) [181, 211]. Conversely, genes that are implicated in
granulosa cell differentiation appeared to be up‐regulated in the DKO ovaries, including
Wt1 [345], Igf1 (Insulin‐like growth factor 1) [346], Bmp6 (Bone morphogenetic protein 6)
[347], Gata6 (GATA‐binding factor 6) [348]. Consistent with these observations, granulosa cells of the developing follicles in the DKO ovaries appeared to exhibit high expression of
3βHSD, suggesting that granulosa cells are compensating the ovarian steroidogenesis by the loss of theca cells. A relevant phenomenon was observed in Gdf9‐/‐ ovaries. In the
absence of Gdf9, the follicles were arrested at primary stage along with a lack of theca cell
layer, and the granulosa cells of the primary follicles eventually developed abnormal
steroidogenic phenotype due to defects in cell differentiation [349].
In summary, my finding provides the first genetic evidence on the critical role of
Hh signaling during ovarian development using in vivo knockout models. While loss of either Dhh or Ihh alone does not seem to affect follicle development, ablation of both Dhh and Ihh in the ovary results in the complete loss of steroidogenic theca cells, follicular arrest and subsequent lack of CLs. Given that disorders in theca cell differentiation are implicated in ovarian diseases such as polycystic ovary syndrome, premature ovarian failure and ovarian cancers [233, 262, 270, 292, 328], my discovery of the mechanism underlying theca cell development not only fill a critical void in basic ovarian biology, but also serve as novel entry points to understand how theca cell‐related pathology affects female reproductive health.
89 3.6 Figures and Tables
Figure 3.1. Effects of the Hedgehog inhibitor cyclopamine on expression of Gli1‐LacZ in the ovary. a‐c, Gli1‐LacZ ovaries (E18.5) were cultured in the presence of cyclopamine
[333] (25 μM), or an equal volume of vehicle, for 3 days followed by LacZ staining (blue).
Ov= ovary; Ms= mesonephros. Scale bar: 500 m.
90
Figure 3.2. Dhh and Ihh, but not Shh, are produced in the ovary right after birth. a, qPCR analysis of Gli1, Ihh and Dhh mRNA expression in E17.5 and PND3 ovaries. b, qPCR analysis of Shh, Ihh, and Dhh in the perinatal ovaries, testis (P3), adrenal (P3) and limb bud
(E17.5). Results were normalized to 18S, and the expression levels in the testis were set as
1 . *P < 0.05; **P < 0.01; ***P < 0.001.
91
Figure 3.3. Dhh and Ihh are highly expressed in granulosa cells and Gli1 and Ptch1
are enriched in theca progenitor cells. a, qPCR validation of Gli1+ theca progenitor cells and Foxl2+ granulosa cells. Results were normalized to 18S, and the expression levels in
tdTomato+ cells were set as 1. b, qPCR analysis of Gli1, Ptch1, Ihh, and Dhh mRNA expression in isolated theca progenitor cells and granulosa cells. *P < 0.05; **P < 0.01; ***P
< 0.001; ****P< 0.0001.
92
Figure 3.4. qPCR analysis of Dhh, Ihh, and Gli1 expression in ovaries deficient for Dhh
and/or Ihh. Ihh/Dhh Dhet: Sf1‐Cre; Ihh f/+; Dhh+/‐; Ihh het/Dhh KO: Sf1‐Cre; Ihh f/+; Dhh‐/‐;
Ihh cKO/Dhh het: Sf1‐Cre; Ihh f/‐; Dhh+/‐; Dhh/Ihh DKO: Sf1‐Cre; Ihh f/‐; Dhh ‐/‐. Results were normalized to 18S, and the expression levels in the Ihh/Dhh Dhet ovaries were set as 1.
93
Figure 3.5. Folliculogenesis appears normal in the absence of either Dhh or Ihh in the adult ovary. a‐c, PAS/ Hematoxylin staining of ovarian sections from control (Sf1‐Cre; Ihh
+/‐; Dhh +/‐), Ihh het/Dhh KO (Sf1‐Cre; Ihh f/+; Dhh ‐/‐) and Ihh cKO/Dhh het (Sf1‐Cre; Ihh f/‐;
Dhh +/‐) ovaries at 2 months of age. Asterisks indicate the presence of corpora lutea. d‐f,
Higher magnification images of corpus luteum. g‐i, Immunofluorescence of α‐SMA in red,
and FOXL2 in green in Dhh/Ihh DKO and control ovaries. α‐SMA is a marker for smooth
muscle cell. FOXL2 marks granulosa cells. Scale bar: 200 m.
94
Figure 3.6. Loss of Dhh and Ihh results in disrupted folliculogenesis in the ovary. a‐b,
Whole mount light‐field microscopic images. c‐h, histological analysis of the ovaries from
95 control (Sf1‐Cre; Ihh f/+; Dhh+/‐) and Dhh/Ihh DKO mice (Sf1‐Cre; Ihh f/‐; Dhh ‐/‐) at 2 month of age. (c, d) whole ovarian sections with different stages of follicle development. (e, f) antral follicles from control and DKO ovaries (hemorrhagic). (g, h) the most advanced stage of follicle development in control (corpus luteum) and DKO ovaries (Pre‐antral follicle).
Arrow indicates intra‐ovarian rete ovarii. Arrowhead indicates the degenerating follicles.
The inset is higher magnification of the outlined area in (d). Asterisks indicate the presence of corpora lutea. Scale bar: 200 m.
96
Figure 3.7. Female mice deficient for Dhh and Ihh show defective theca cell differentiation in the ovary. a‐f, Immunofluorescence of α‐SMA in red (a, b), CYP17A1 in
97 magenta (c, d), FOXL2 in green (a‐d), 3βHSD in red (e, f) and DAPI in grey (e, f) in Dhh/Ihh
DKO (Sf1‐Cre; Ihh f/‐; Dhh ‐/‐) and control (Sf1‐Cre; Ihh f/+; Dhh+/‐) ovaries. α‐SMA is a marker for smooth muscle cell. CYP17A1 is a marker for androgen‐producing theca cells, 3βHSD marks steroidogenic cells and FOXL2 marks granulosa cells. Asterisks indicate the presence of corpora lutea. Scale bar: 200 m.
98
Figure 3.8. Female mice deficient for Dhh and Ihh exhibit blunted steroid production.
Serum levels of DHEA (didehydroepiandrosterone), testosterone and progesterone in
control (Sf1‐Cre; Ihh f/+; Dhh+/‐) and Dhh/Ihh DKO (Sf1‐Cre; Ihh f/‐; Dhh ‐/‐) female mice. . *P <
0.05.
99
Figure 3.9. The interstitial cells in the DKO ovary are positive for COUP‐TFII. a‐d,
Immunofluorescence of COUP‐TFII in red and FOXL2 in green in Dhh/Ihh DKO (Sf1‐Cre; Ihh
f/‐; Dhh ‐/‐) and control (Sf1‐Cre; Ihh f/+; Dhh+/‐) ovaries. FOXL2 marks granulosa cells.
Arrows indicate COUPTFII‐positive interstitial cells that are immediately adjacent to the granulosa cells. Scale bar: 100 m.
100
Figure 3.10. Transcriptome changes in the absence of Dhh and Ihh in the adult ovary.
a, Microarray analysis of control (Sf1‐Cre; Ihh f/+; Dhh+/‐, n=4) and the Dhh/Ihh DKO (Sf1‐Cre;
Ihh f/‐; Dhh ‐/‐, n=3) ovaries. b, qPCR analysis of gene expression for Hedgehog signaling
components (Dhh, Ihh, and Gli1), and genes associated with steroidogenesis (Nr5a1, Star,
Cyp11a1, Hsd3b1, Cyp19a1) in control and DKO ovaries. Results were normalized to 18S,
and the expression levels in the control ovaries were set as 1. *P < 0.05; **P < 0.01; ***P <
0.001; ****P< 0.0001.
101 Fold change Genes of interest (Dhh/Ihh DKO vs. Control)
Ehf 3.216
Ppargc1a 1.926
Bmp6 1.842
Wt1 1.665
Igf1 1.594
Gata6 1.550
Sox9 1.548
Lepr ‐4.425
Nr2f2 ‐2.807
Vegfa ‐1.695
Cebpb ‐1.645
Table 3.1. Fold changes of genes of interest from microarray analysis of Dhh/Ihh DKO (Sf1‐
Cre; Ihh f/‐; Dhh ‐/‐) ovaries versus control (Sf1‐Cre; Ihh f/+; Dhh+/‐) ovaries from 2 months of age mice.
102 CHAPTER 4: Expression of Theca Cell‐Specific Gli1 Requires Oocyte‐Derived GDF9
through Hedgehog Signaling from Granulosa Cells
4.1 Abstract
Folliculogenesis is controlled by intricate communication between oocytes, granulosa, and theca cells. In Chapter 2 and 3, I have demonstrated that theca cells derive from two sources: Wt1+ cells indigenous to the ovary and Gli1+ mesenchymal cells migrated
from the mesonephros. These progenitors acquire theca lineage marker Gli1 in response to
paracrine signals Desert hedgehog (Dhh) and Indian hedgehog (Ihh) from granulosa cells.
One remaining question is what factor(s) regulate the production of Dhh and Ihh in the granulosa cells. Because the differentiation of granulosa cells is tightly controlled by factors from the oocytes, I therefore hypothesize that the oocyte plays a role in regulating
Hh ligand production in granulosa cells. In the absence of oocytes, the expression of the components of Hh signaling including Dhh, Ihh and Gli1 in the ovary was significantly decreased, indicating that oocytes are necessary for the production of Hh ligands in the ovary. Next I investigated what factor(s) from the oocyte are responsible for Hh ligand production in granulosa cells. One such oocyte‐specific candidate is growth differentiation factor‐9 (Gdf9), a TGF‐beta family member with known roles in the formation of theca cell layer. In the absence of Gdf9, Dhh and Ihh expressions in the granulosa cells as well as expression of Gli1 in the theca cells were significantly down‐regulated. However, when
GDF9 recombinant protein was supplemented to an oocyte‐depleted ovary in culture, the expression of Dhh, Ihh were significantly induced as well as the subsequent expression of
Gli1, further supporting the role of oocyte‐specific factor GDF9 in regulating Ihh and Dhh in
103 the granulosa cells. In conclusion, these studies reveal novel multicellular interactions via
GDF9 and Hh signaling critical for ovarian folliculogenesis and formation of a functional theca.
104 4.2 Introduction
In Chapter 3, I have demonstrated the mechanisms underlying theca cell development, in which the theca progenitor cells in the ovary acquired lineage specific Gli1 in response to paracrine DHH and IHH signaling from granulosa cells. In the absence of
Dhh and Ihh, theca cells failed to differentiate, a functional theca layer was lacking and the follicles were unable to develop beyond pre‐antral stage. These results establish the role of
Hedgehog signaling in regulating the differentiation of theca cells. Additionally, it was shown that constitutive activation of Hh signaling in the ovary disrupted the development of smooth muscle cell in the theca cell layer and caused subsequent failure in ovulation
[289]. These observations altogether suggest that although Hh signaling is critical for ovarian folliculogenesis, it needs to be properly regulated for ovarian functions.
Candidate factor(s) that regulate Hh signaling in the ovary are likely to be originated from pituitary, and/or oocytes within the ovary through regulation of granulosa cells [19]. On one hand, pituitary‐derived LH and FSH control many aspects of granulosa cell development in various events such as cell proliferation and apoptosis, follicle atresia, estrogen production and the formation of CLs [158‐162]. On the other hand, oocyte is considered the central regulator of granulosa cell function and thereby play critical roles during the progression of folliculogenesis [350]. Many of the studies regarding the role of oocyte on ovarian somatic cells were conducted in chicken and mice. In chicken, the germinal disc (GD) located on the surface of the female gamete contains the genetic material and cellular organelles of the egg, and is considered equivalent to the oocyte of the mammal [351]. Granulosa cells surrounding the oocyte exhibit different phenotypes based on their locations relative to the GD: granulosa cells that directly adjacent to the GD are
105 more proliferative, whereas the ones distal to the GD are more differentiated [352‐354].
This phenomenon is further explained by the discovery of epidermal growth factor (EGF) derived from the GD [355‐357]. GD‐derived EGF acts on proximal granulosa cells and renders the cells more proliferative by suppressing the expression of LH receptors [356‐
359]. In contrast, the distal granulosa cells are less exposed to EGF and therefore exhibit a more differentiated phenotype promoted by LH [358]. In mammals, it was first proposed in
1970 that the oocyte has the capacity to prevent follicular luteinization in rabbit [360, 361].
Subsequent studies dealing with co‐cultures of ovarian granulosa cells and denuded oocytes further establish the dominant role of oocytes in regulating granulosa cells throughout follicular development [362‐364]. For instances, oocytectomy of oocyte‐ cumulus complexes prevents FSH induced cumulus expansion, whereas co‐culture of the oocytectomized complexes with denuded oocytes rescues the expansion of the cumulus cells [189]. Similarly, dissociated cumulus cells alone failed to synthesize a mucinous matrix in response to FSH, but the matrix production is restored when denuded full‐grown oocytes are supplemented in culture [190].
Two oocyte‐derived factors critical for follicular cell function in vivo are growth‐ differentiation factor 9 (GDF9) and bone morphogenetic protein 15 (BMP15), as mutation of these genes causes female infertility [146, 365]. Both of GDF9 and BMP15 are closely related members of TGFβ family and they activate signaling pathways in granulosa cells to regulate cellular process required for ovarian functions [272, 350, 363]. In addition to the regulation of granulosa cells differentiation, GDF9 also appears to play critical roles in theca cell recruitment/differentiation. In the absence Gdf9, a theca layer failed to develop, accompanied by follicle arrest at primary stage [146]. However, whether GDF9 directly
106 regulates theca cell development, or indirectly through regulation of granulosa cells is not known. With all the observations aforementioned, I hypothesize that oocyte‐specific GDF9 indirectly regulates theca cell development though inducing the Hh signaling in the granulosa cells.
107 4.3 Material and Methods
4.3.1 Animal breeding
Gli1‐LacZ (#008211) mice were purchased from the Jackson Laboratory (Bar
Harbor, ME). Gdf9 knockout ovaries were provided by Dr. Martin Matzuk (Baylor College of
Medicine). Female mice were housed with male mice overnight and checked for the
presence of vaginal plug the next morning. The day when the vaginal plug was detected
was considered embryonic day (E) 0.5. All animal procedures were approved by the
National Institute of Environmental Health Sciences (NIEHS) Animal Care and Use
Committee and are in compliance with a NIEHS‐approved animal study proposal. All experiments were performed on at least three animals for each genotype.
4.3.2 Busulfan treatment
Pregnant CD‐1 females were injected intraperitoneally (IP) at E10.5 with 40 mg/kg of busulfan (Sigma) dissolved in 50% DMSO, or an equivalent volume of 50% DMSO as the vehicle control. Busulfan‐treated pregnant female often undergo dystocia. In order to
examine the expression of theca progenitor cells‐specific Gli1 in postnatal ovaries (Chapter
2), the oocyte‐depleted ovaries were collected just prior to birth and cultured for 3 days.
4.3.3 Immunohistochemistry and histological analysis
Immunohistochemistry and histological analysis were performed as described with
modifications[331]. For immunohistochemistry on frozen sections, ovaries were fixed in 4%
paraformaldehyde in PBS at 4°C overnight, dehydrated through a sucrose gradient, embedded,
and cryosectioned at 10 μm increments. After preincubating with 5% normal donkey serum in
108 PBS for 1 hour, the sections were then incubated with either anti-FOXL2 (1:500, a gift from Dr.
Dagmar Wilhelm, Monash University, Australia) and anti-TRA98 (1:1000, MBL international,
USA) in PBS-Triton X-100 solution with 5% normal donkey serum at 4°C overnight. The sections were then washed and incubated in the appropriate secondary antibody (1:500;
Invitrogen, USA) before mounting in Vector Mount with DAPI (Vector Labs). Slides were imaged under a Leica DMI4000 confocal microscope.
4.3.4 LacZ staining
Fresh tissues were stained in LacZ staining solution at 37oC for 3 hours. The LacZ
staining solution was prepared by dissolving X‐gal (Invitrogen, Carlsbad, CA) in
dimethylformamide and the tissue stain base solution (Chemicon, Billerica, MA) to make 1
mg/ml working solution as described elsewhere [301]. The stained samples were then
fixed in 4% paraformaldehyde/PBS at 4oC overnight. For further histological analysis,
samples after LacZ staining were embedded in paraffin following the sectioning procedure
listed above. The sections were counterstained with Fast Red (Sigma).
4.3.5 Gene expression analysis
Gene expression analysis was performed as described with modifications [304].
Total RNA was isolated from ovaries using the PicoPure RNA isolation kit (Arcturus,
Mountain View, CA) according to the manufacturer’s protocol. The cDNA preparation was
synthesized using random hexamers and the Superscript II cDNA synthesis system
(Invitrogen Corp., Carlsbad, CA) following the manufacturer’s instruction. Gene expression
was analyzed by real‐time PCR using Bio‐Rad CFX96TM Real‐Time PCR Detection system.
109 Taqman gene‐expression probes (Gli1: Mm00494654_m1, Ihh: Mm00439613_m1, Dhh:
Mm01310203_m1, Gdf9: Mm00433565_m1) were used to detect the fold changes of the
transcripts. Fold changes in gene expression were determined by quantitation of cDNA
from target (treated) samples relative to a calibrator sample (control). All real‐time PCR
analyses were performed in duplicate, and the results were reported from at least three
independent experiments. The relative fold change of transcript was calculated using the
mathematical model of Pfaffl as previously described [305] and was normalized to 18S
rRNA (Mm03928990_g1) or Gata4 (Mm00484689_m1) as an endogenous reference.
4.3.6 Organ culture
Ovaries from E18.5 Gli1‐LacZ, or CD‐1 embryos were cultured for 3 days on
Millicell‐PC membrane inserts (0.4 μm pore size; 30 mm diameter; Millipore corp., Medford,
MA) with medium filling only the lower chamber [332]. 8‐10 ovaries were placed on each
membrane, with one drop of medium on each ovary. For culture with GDF9 recombinant
protein (Peng et al. PNAS 2013), the protein (60 ng/ml), or an equal volume of vehicle (TBS
(pH8.0) with 1mg/ml BSA), was added to the last day of culture 24 hours prior to the collection. The medium for organ culture was DMEM/F12 (1:1) + L‐Glutamine + 15mM
HEPES (Invitrogen) supplemented with 50 μg/ml penicillin G/streptomycin sulfate, and
5% (vol/vol) fetal bovine serum (FBS). The culture medium was changed every other day,
and after 3 days of culture, the ovaries were collected and subject to RNA extraction, or
processed for LacZ staining as described above.
110 4.3.7 Statistical analysis
Data were analyzed using Prism (Version 6, GraphPad Software) by two‐tailed
Student’s t‐test. Values are expressed as mean±s.e.m. A minimal of 3 biological replicates was used for each experiment.
4.4 Results
4.4.1 Dhh and Ihh production in the granulosa cells is regulated by factor(s) from the oocyte
I have demonstrated that DHH and IHH from granulosa cells were responsible for theca cell differentiation in the ovarian development (Chapter 2 & 3). In the absence of Dhh and Ihh, theca cell differentiation was abolished and the follicles were unable to develop beyond pre‐antral stage. Next I investigated what signal induces the production of Hh ligands in the granulosa cells. Functions of granulosa cells are regulated not only systemically by hormone signals from the pituitary, but also locally via factors produced in the oocytes. Without the oocyte, follicle formation never occur [366]. I therefore hypothesized that oocytes may play a role in regulating Hh ligand production in granulosa cells. To test this hypothesis, I treated pregnant females with busulfan, a chemotherapeutical drug known to induce germ cell death in the embryos [367, 368].
Busulfan was injected to pregnant Gli1‐LacZ female mice at E10.5, when the germ cells just arrive in genital ridge and start proliferating [8‐10]. The specificity of in utero busulfan treatment on oocytes was confirmed by a complete abolishment of oocytes at E18.5 prior to birth, coincident with normal establishment of FOXL2‐positive granulosa cell population in ovaries of busulfan‐treated pups (Fig. 4.1a & b). This result confirms previous finding
111 that busulfan‐induced germ cell depletion does not alter the prenatal establishment of
ovarian somatic cell types [369]. After birth, Gli1‐LacZ staining appeared in the medulla of
the control ovaries as expected. However, the busulfan‐treated ovaries were devoid of any
LacZ staining and the ovaries were much smaller in size, likely due to the loss of female
germ cells (Fig. 4.1c & d). Further analysis of the ovarian sections revealed an absence of
follicle formation in the busulfan‐treated ovaries (Fig. 4.1e & f). Gli1‐LacZ staining in the adjacent mesonephros was unaffected (Fig. 4.1e & f), indicating that the loss of Gli1 expression in the ovary is due to the depletion of oocytes, not a general effect of busulfan on Gli1 expression. To further investigate whether the Hh ligand production is affected by the loss of oocytes, I examined the expression of Dhh, Ihh, and Gli1 in the busulfan‐treated ovaries. In the absence of oocytes, the expression of Gli1 mRNA was significantly down‐ regulated, consistent with the histological results and LacZ staining (Fig. 4.1). Similar to the
down‐regulated Gli1 expression, expression of Dhh and Ihh were also significantly
decreased (Fig. 4.2). These results indicate that oocyte is playing a role in stimulating Dhh
and Ihh production in granulosa cells.
4.4.2 Oocyte‐specific GDF9 is responsible for Dhh and Ihh production in granulosa
cells
Next I investigated what factor(s) from oocyte are responsible for Hh production in
granulosa cells. One potential factor is GDF9, an oocyte‐specific factor essential for
folliculogenesis and formation of theca layer [146]. In the absence of Gdf9, follicle
development is arrested at primary stage and the theca cell layer is missing. This
observation suggests that GDF9 is involved in theca cell development. I therefore
112 hypothesize that GDF9 regulates theca cell development via inducing Hh ligand production
in the granulosa cells. To testis this hypothesis, I first examined the expression of Gdf9 in the perinatal ovaries. Gdf9 expression was absent in the ovary before birth but significantly increased at PND3 (Fig. 4.3a), in parallel with the expression pattern of Dhh and Ihh
(Chapter 3). In the PND3 Gdf9 knockout ovaries, expression of Dhh and Ihh was significantly reduced compared to that of Gdf9 heterozygous ovaries with normal follicle development [146] (Fig. 4.3b). These results indicate that GDF9 stimulates Dhh and Ihh production in granulosa cells. Additionally, when Gdf9 expression was absent, Gli1 expression was also decreased in parallel with the diminished ligand production (Fig. 4.3b), further confirming the inactivation of Hh signaling.
4.4.3 Exogenous GDF9 protein reactivates Hh signaling in the absence of oocytes.
While it is apparent that Gdf9 plays a role in regulating Dhh and Ihh production in granulosa cells, it is not clear whether GDF9 is capable of rescuing Hh signaling in the absence of the oocytes. To test this possibility, I cultured the busulfan‐treated ovaries with or without mouse recombinant GDF9 protein [272]. When recombinant GDF9 was supplemented to the oocyte‐depleted ovaries (Busulfan‐treated) in culture, it significantly increased mRNA expression of Ihh, Dhh, and Gli1 (Fig. 4.3c), confirming the critical role of oocyte‐derived GDF9 in stimulating Dhh and Ihh production in granulosa cells, and subsequent theca progenitor cell‐specific Gli1. Additionally, the expression of Gdf9 in the
ovary was not affected by the absence of Dhh and Ihh, not did its co‐operative factor Bmp15
[196, 272] (Fig. 4.4). These results not only support that GDF9 functions as an upstream
113 regulator of Hh signaling, but also affirms that the lack of androgen‐producing theca cells was specifically due to loss of Hh ligands, not decreased Gdf9 expression.
4.5 Discussion
My findings demonstrate that oocyte‐specific factor GDF9 stimulates Dhh and Ihh production in the granulosa cells. The Hh ligand production in granulosa cells then induces the expression of Gli1, a maker for theca cell lineages that derived from both ovarian and extraovarian origins.
Although Dhh and Ihh appear to be downstream targets of GDF9, the ovarian phenotypes of Gdf9 KO ovaries differ from those observed in Dhh/Ihh DKO ovaries. The follicles in the Gdf9 KO ovaries arrest at primary stage, whereas follicles in Dhh/Ihh DKO ovaries develop beyond the primary stage and arrest at the preantral stage. The difference in phenotypes reveals two important findings: First, Dhh/Ihh signaling pathway plays a specific role in theca cell differentiation and second, GDF9, the factor responsible for the production of Dhh and Ihh, has a broader function in follicle development. The identities of other players that act downstream of GDF9 remains to be identified. A possible approached to address that question is to perform a microarray analysis on Gdf9 knockout adult ovaries, and compare their transcriptome changes with that of the Dhh/Ihh DKO ovaries.
In the absence of Gdf9, expression of Dhh, Ihh and Gli1 was significantly down‐ regulated but not completely abolished. This observation raises the question as whether
GDF9 is the only upstream regulator of the Hh signaling. In fact, it has been shown that
GDF9 often signals synergistically with BMP15 to regulate granulosa cell function [196,
114 370‐372]. Although GDF9 homodimers could modulate ovarian pathways in vitro [373,
374], GDF9:BMP15 heterodimers are nevertheless much more potent [272]. Thus, in spite
the fact that GDF9 alone is capable of inducing Hh signaling in the ovary, one cannot
exclude the possibility that GDF9 also forms a more potent heterodimer with BMP15 to
regulate Hh ligand production [272]. When mouse GDF9 recombinant protein was added to
the oocyte‐depleted ovary in culture, it significantly induced the production of Hh ligands
in granulosa cells. If GDF9 indeed works synergistically with BMP15 to control Hh
signaling in vivo, I would expect to see an even more pronounced increase of Hh ligand
production, when both GDF9 and BMP15 are added to the oocyte‐depleted ovaries in
culture. Although the upstream regulators of Hh signaling might be multiple, GDF9 is likely
to be the dominant growth factor in this case at least in mice. This is because 1) BMP15
homodimers do not affect granulosa cells differentiation in vitro [272]; 2) the level of Dhh
and Ihh expression are significantly decreased in the absence of Gdf9; 3) Gdf9‐/‐ female mice exhibit major defects in ovarian folliculogenesis and they are infertile [146], whereas
Bmp15‐/‐ female mice are subfertile with minimal histopathological defects [197]; 4)
Ovaries from Gdf9+/‐ Bmp15‐/‐ exhibited more severe fertility defects than Bmp15‐/‐ mice, while ovaries from Gdf9‐/‐ Bmp15‐/‐ mice are phenotypically resemble Gdf9‐/‐ mutants [197].
Although I have shown that GDF9 acts on granulosa cells and indirectly induces
Gli1 expression in the theca cells, this result does not exclude the possibility that GDF9, or
GDF9 and BMP15 together, might also directly regulate theca cells. Work to date suggests
that GDF9 and BMP15 use the same type II receptor BMPRII [375‐377], but different type I
receptors. While GDF9 activates TGF‐β type I receptor TGFBR1 (ALK‐5) [378, 379], BMP15
appears to utilize BMP type 1b receptor BMPR1B (ALK‐6) [376]. As a result, GDF9 activates
115 the intracellular SMAD2/3 pathway [378, 380], whereas BMP15 induces the BMP pathway via intracellular SMAD1/5/8 signaling [376, 377]. The GDF9:BMP15 heterodimers signal through BMPRII‐ALK4/5/7‐ALK/6 receptor complex to stimulate phosphorylation of
SMAD2/3 in mouse and human granulosa cells [272]. The receptors of GDF9 and BMP15 are expressed in sheep, rat, pig and mouse granulosa cells throughout follicular development [381‐387], rendering granulosa cells a primary target for GDF9 and BMP15 signaling [19]. In contrast to granulosa cells, the direct roles of GDF9 and BMP15 on theca cell development is far less clear. Several pieces of evidence suggest that TGFBR1 (ALK‐5)
[387‐390] and TGFBR2 [387, 390, 391] are expressed in the theca‐interstitial cells in species such as pig, sheep, cow and mouse. The presence of TGFBR1 in the theca compartment suggests that theca‐interstitial cells might be a direct target of GDF9 signaling. This notion is supported by several in vitro studies, in which GDF9 suppresses
androgen production of human [392], bovine [389] and mouse [393] theca‐interstitial cells
whereas promotes androgen biosynthesis in theca cells in rat [394], revealing important
species‐specific differences. In the absence of Dhh and Ihh, steroidogenic theca cells were
missing yet the expression of Gdf9 and Bmp15 were unaffected, indicating that the loss of
functional theca cells was not due to blunted Gdf9 and/or Bmp15 expression. However, it is
not yet clear whether the expression of the GDF9/BMP15 receptors such as TGFBR1 was
altered by the loss of Hh signaling. If the expression of the GDF9 receptors in the
interstitium of the ovary were not affected in the absence of Hh signaling, it indicates that
GDF9 does not directly regulate theca cells in vivo. However, if the expression of GDF9
receptors in ovarian interstitium were lost, it would be difficult to further distinguish
whether the decreased receptor expression is the result of loss of theca cells, or it is a
116 decrease in expression itself. Alternatively, cultures of theca‐interstitial cells in the presence of a Hh signaling inhibitor might be able to answer whether inhibition of Hh signaling affects the expression of GDF9 receptors in theca‐interstitial cells.
The importance of GDF9 in folliculogenesis was first described by Dong et al in
1996 [146]. When the expression of Gdf9 was lost in mice, folliculogenesis was arrested at primary stage and the theca cell layer failed to form. One key question that remains puzzling for the past two decades is whether GDF9 directly regulates theca cell recruitment and/or differentiation, or indirectly through regulation of pre‐antral granulosa cell development. My findings of GDF9 regulating Hh signaling in granulosa cells bridge the gap between loss of Gdf9 in the oocyte and the lack of theca layer in the ovarian interstitium.
Although whether GDF9 directly regulates theca cells is not clear, the GDF9/Hh signaling appears to be the main regulatory pathway that governs theca cell differentiation, due to the fact that ablation of Hh signaling resulted in the loss of androgen‐producing theca cells despite the presence of Gdf9. In addition, this study indicates that the process of theca cell differentiation is accomplished through intricate communications among three cell types: oocytes, granulosa cells and the mesenchymal cells in ovarian interstitium. These findings not only answer a fundamental question in the field, but also provide a model system that might be applicable to study morphogenesis in other organs.
117 4.6 Figures
Figure 4.1: Effects of in utero busulfan treatment on oocytes and the expression of
Gli1 in the ovary. a‐b, E18.5 control and busulfan‐treated ovaries were analyzed by
118 immunofluorescence detection for granulosa cell marker FOXL2 (green), germ cell marker
TRA98 (red), and nuclear counterstain DAPI (blue). Scale bar: 100 m. c‐f, Gli1‐LacZ expression (blue) in the control and busulfan‐treated ovaries (n=8‐15) after 3 days of culture (c & d are whole mount) and (e & f are sections counterstained with FastRed).
Scale bar: 500 m.
119
Figure 4.2: Loss of oocytes in the ovary suppresses the expression of the components of Hh signaling. a, qPCR analysis of Gli1, Ihh and Dhh mRNA expression in the control and busulfan‐treated ovaries. Every two ovaries were pooled to ensure enough RNA yield for
PCR analysis (n=3‐5). Results were normalized to a somatic cell gene Gata4. *P < 0.05; **P
< 0.01; ***P < 0.001.
120
Figure 4.3: Hh signaling in granulosa cells is regulated by oocyte‐derived factor GDF9. a, qPCR analysis of Gdf9 mRNA expression in the E17.5 and P3 ovaries (n=5). b, qPCR
121 analysis of Gli1, Ihh and Dhh mRNA expression in control and Gdf9 knockout ovaries (n=4).
c, qPCR analysis of Gli1, Ihh and Dhh mRNA expression in the E18.5 oocyte‐depleted ovaries
(busulfan‐treated) cultured with or without recombinant mouse GDF9 protein (60 ng/ml).
Every two ovaries were pooled to ensure enough RNA yield for PCR analysis (n=5). The expression level in oocyte‐depleted ovaries without GDF9 (Control) was set as 1. Results were normalized to 18S. *P < 0.05; **P < 0.01; ***P < 0.001.
122 Control Dhh/Ihh DKO 2.5
2.0
1.5
1.0
0.5
Relative mRNA expression Relative 0.0
Gdf9 Bmp15
Figure 4.4: qPCR analysis of Gdf9 and Bmp15 expression in control and Dhh/Ihh DKO
(Sf1‐Cre; Ihh f/‐; Dhh ‐/‐) ovaries. Results were normalized to 18S, and the expression levels in the control (Sf1‐Cre; Ihh f/+; Dhh+/‐) ovaries were set as 1. No statistical significance was detected by two‐tailed Student’s t‐test.
123 CHAPTER 5: Conclusions and Future Directions
My doctoral research has uncovered the embryonic origins of theca cells and the
mechanism underlying their differentiation (Fig. 5.1). Theca cells are derived from two
sources: the Wt1‐positive precursor cells in the fetal ovary and the Gli1‐positive
mesenchymal cells in the fetal mesonephros, an embryonic organ immediately adjacent to
the fetal ovary. The Gli1‐positive cells in the fetal mesonephros migrate into the ovary just
prior to birth, and give rise to a small percentage of theca cells in the adult ovary, whereas
the Wt1‐positive cells derived from fetal ovary contribute to the majority of theca cells in
adulthood. While mesonephros‐derived Gli1‐positive cells give rise to mostly steroidogenic
theca cell located proximal to granulosa cells, fetal ovary‐derived Wt1‐positive cells
contribute to both steroidogenic theca cells and non‐steroidogenic cells spanning the entire
theca layer. In addition to their differences in cell‐lineage contributions and relative
proximity to granulosa cells, the two sources of theca cells exhibit distinct gene expression
profiling. These findings not only support the historical observation that theca cells are a
mixed population consisting of theca interna and theca externa, but also demonstrate that
the heterogeneity of this population is a reflection of disparate cellular origins and
transcriptomes. Although originated from different embryonic sources, the two pools of
theca progenitor cells acquire the expression of Gli1 when they commit to theca cell lineage
in the ovary soon after birth. The expression of Gli1 in theca progenitor cells is induced by
Dhh and Ihh produced by granulosa cells, and the Hh ligand production in granulosa cells is
then regulated by oocyte‐specific factor GDF9. In the absence of Dhh and Ihh, theca progenitor cells fail to differentiate into androgen‐producing theca cells, and the
124 folliculogenesis is arrested at preantral stage along with blunted steroid productions and
lack of CLs. Although I have identified the origins of theca cells and characterized the
mechanism underlying theca cell differentiation, there remain many aspects of theca cell
development to be explored. In order to gain a deeper insight into theca cell development
and regulation, I propose the following future experiments:
1. What is the source of signal(s) that trigger the migration of Gli1‐positive cells from
the mesonephros into the ovary?
I have found that theca cell progenitors originate from two sources: Wt1‐positive cells in the fetal ovary and Gli1‐positve cells in the mesonephros. The appearance of mesonephros‐derived Gli1‐positive cells in the ovary suggests that signal(s) from the ovary induce the migration of the Gli1‐positive cells into the ovary. Based on the fact that the mesonephros‐derived theca cells first appear in the medulla of the neonatal ovary and are located adjacent to primary follicles, I therefore hypothesize that factor(s) from the primary follicles trigger the migration. To test if the developing follicles play a role in triggering the migration of mesonephric Gli1‐positive cells, I plan to disrupt the formation of the follicle in the ovary and examine if the Gli1‐positive cells from the mesonephros still
migrate. I expect that the mesonephros‐derived Gli1‐positive cells fail to migrate into the
ovary in the absence of the primary follicle. If so, I will then conduct experiments to further
identify what factor(s) from the primary follicle triggers migration. Based on my findings
that theca cell differentiation is regulated indirectly by oocyte‐specific factor GDF9 through
Dhh and Ihh in the granulosa cells, I speculate that oocyte‐derived GDF9 and/or granulosa
cell‐derived Hh ligands might be the migration‐inducing agents. To test if this is the case, I
125 plan to culture busulfan‐treated Gli1‐lineage tracing ovaries (with mesonephros attached)
with recombinant GDF9 and/or DHH/IHH proteins. If neither GDF9 nor Dhh/Ihh is able to
induce migration, I will use mass spectrometry to screen for potential chemoattractants
produced by the follicles and perform the migration assay.
2. Does loss of Dhh and Ihh impair the fertility of female mice?
I have shown that in the absence of Dhh and Ihh, the theca layer is barely present, and the androgen‐producing theca cells were missing along with blunted circulating androgens and progesterone. In addition, the ovaries exhibited many defects including follicle arrest at the preantral stage and a lack of corpora lutea, indicative of ovulation failure. Together these results prompt me to examine whether the Dhh/Ihh DKO female mice are infertile, and whether the defective folliculogenesis is a result of loss of functional theca cells. I will generate Dhh/Ihh DKO mice and their littermate control. I will examine the fertility of the female mice by addressing the following questions: a. Are the Dhh/Ihh female mice able to produce offspring?
Control and Dhh/Ihh DKO female mice will be bred with fertile males to examine the fertility of the female. Because the DKO females exhibited ovarian follicle arrest in the
preantral follicle stage, I expect that the estrous cycle in the Dhh/Ihh DKO female is
abnormal, which leads to the absence of vaginal plug. In addition, I expect that the females
will be completely infertile because of the absence of corpora lutea.
b. Do exogenous gonadotropins induce ovulation in the Dhh/Ihh DKO ovaries?
Ovaries lacking Dhh and Ihh exhibited disrupted folliculogenesis with the absence of corpora lutea, indicating that ovulation did not occur. To gain further insight into the
126 fertility problems as well as the defects in folliculogenesis, I will investigate the ability of
the DKO female to ovulate by performing the superovulation procedure. I expect that the
exogenous gonadotropins will have very minimal effects on follicle development and
ovulation since functional theca cells are absent. Because exogenous gonadotropins will
have minimal effects on ovulation, I therefore expect that no corpus lutea, or very few
corpora lutea will be present in the DKO ovaries.
c. Does exogenous testosterone stimulate preantral follicles to progress to ovulation?
The most dramatic defect in the Dhh/Ihh DKO ovaries was a lack of theca layer. The
defects in theca cell differentiation result in reduced level of circulating testosterone. These
observation leads to the hypothesis that the arrest in follicular development in the DKO
ovary is due to defects in theca cell development and possibly insufficient testosterone production. To test this hypothesis, I will provide testosterone supplement to prepubertal
Dhh/Ihh DKO female mice to see if it rescues follicle development. I expect that the testosterone supplement will allow the follicles to grow to at least antral follicle stage; however, the follicles won’t be ovulated as a result of a lack of theca externa, which contains smooth muscle layer important for ovulation [289]. Consistently with these, I expect that the level of circulating testosterone and estradiol will be increased but the progesterone level will remain low due to a lack of ovulation and production of CL.
3. Identifying signaling pathways downstream of GDF9 in folliculogenesis
My findings suggest that theca cell differentiation is indirectly regulated by oocyte‐ derived factor GDF9 through Hedgehog signaling from granulosa cells. Although Dhh and
Ihh appear to be downstream targets of GDF9, the ovarian phenotypes in the Gdf9 KO and
127 Ihh/Dhh DKO are quite different. Both Gdf9 KO ovaries and Dhh/Ihh DKO ovaries exhibited
a lack of theca cell layer. However, follicles in the Gdf9 KO ovaries arrest at primary stage in
contrast to the preantral stage in the Dhh/Ihh DKO ovaries. The differences in phenotypes
reveal two important findings: first, Dhh/Ihh signaling pathway plays a specific role in
theca cell differentiation and second, GDF9, the factor responsible for the production of
Dhh and Ihh, has a broader function in follicle development. In this aim, I will use our
knowledge on Dhh/Ihh to gain insight into folliculogenesis under the control of GDF9.
1. Does activation of Hedgehog signaling ameliorate the phenotypes in Gdf9 KO ovaries?
Loss of Gdf9 leads to defects in folliculogenesis including an arrest of follicle
development at primary stage, inactivation of the Hh pathway, and consequent failure in
the formation of theca layer. To investigate whether the inactivation of the Hh pathway is
completely or partially responsible for the follicular defects in the absence of Gdf9, I plan to
ectopically activate the Hh pathway in the Gdf9 KO ovary by culturing pieces of Gdf9 KO
ovary [395] with IHH and DHH, and examine if the activation of Hedgehog pathway
facilitates further development of the follicles in the absence of Gdf9. I expect that the
presence of IHH/DHH will allow the Gdf9 KO follicles to develop beyond the primary stage.
Because GDF9 appears to have a boarder function in follicle development than DHH and
IHH, I do not expect the presence of DHH/IHH is able to fully rescue the phenotypes of Gdf9
KO ovaries.
2. What other factors, in addition to DHH and IHH, act downstream of GDF9?
To answer this question, I plan to perform microarray analyses on Gdf9 KO ovaries and Dhh/Ihh DKO ovaries. By comparing the mRNA transcriptomes from the two mutant models at different time points, I will be able to obtain a list of genes that are differentially
128 expressed in the absence of Gdf9 but are unaffected by the loss of Dhh and Ihh. I will confirm the gene list by real time PCR and search the literature for known functions of the genes on the list. I will then exclude the genes with known functions and focus on the novel genes in the future.
129 5.1 Figures
Figure 5.1 Proposed model for the origins and differentiation of theca cells in the mouse ovary.
130 BIBLIOGRAPHY
1. Miyamoto, N., et al., Defects of urogenital development in mice lacking Emx2. Development, 1997. 124(9): p. 1653‐64. 2. Birk, O.S., et al., The LIM homeobox gene Lhx9 is essential for mouse gonad formation. Nature, 2000. 403(6772): p. 909‐13. 3. Luo, X., Y. Ikeda, and K.L. Parker, A cell‐specific nuclear receptor is essential for adrenal and gonadal development and sexual differentiation. Cell, 1994. 77(4): p. 481‐90. 4. Kreidberg, J.A., et al., WT‐1 is required for early kidney development. Cell, 1993. 74(4): p. 679‐91. 5. Cui, S., et al., Disrupted gonadogenesis and male‐to‐female sex reversal in Pod1 knockout mice. Development, 2004. 131(16): p. 4095‐105. 6. Katoh‐Fukui, Y., et al., Male‐to‐female sex reversal in M33 mutant mice. Nature, 1998. 393(6686): p. 688‐92. 7. Fujimoto, Y., et al., Homeoproteins Six1 and Six4 regulate male sex determination and mouse gonadal development. Dev Cell, 2013. 26(4): p. 416‐30. 8. Tam, P.P. and M.H. Snow, Proliferation and migration of primordial germ cells during compensatory growth in mouse embryos. Journal of embryology and experimental morphology, 1981. 64: p. 133‐147. 9. Anderson, R., et al., The onset of germ cell migration in the mouse embryo. Mechanisms of development, 2000. 91(1‐2): p. 61‐68. 10. Molyneaux, K.A., et al., Time‐lapse analysis of living mouse germ cell migration. Developmental biology, 2001. 240(2): p. 488‐498. 11. Pepling, M.E. and A.C. Spradling, Female mouse germ cells form synchronously dividing cysts. Development, 1998. 125(17): p. 3323‐8. 12. Merchant, H., Rat gonadal and ovarioan organogenesis with and without germ cells. An ultrastructural study. Dev Biol, 1975. 44(1): p. 1‐21. 13. Ruby, J.R., R.F. Dyer, and R.G. Skalko, The occurrence of intercellular bridges during oogenesis in the mouse. J Morphol, 1969. 127(3): p. 307‐39. 14. Brennan, J. and B. Capel, One tissue, two fates: molecular genetic events that underlie testis versus ovary development. Nat Rev Genet, 2004. 5(7): p. 509‐21. 15. Gassei, K. and S. Schlatt, Testicular morphogenesis: comparison of in vivo and in vitro models to study male gonadal development. Ann N Y Acad Sci, 2007. 1120: p. 152‐67. 16. Combes, A.N., et al., Endothelial cell migration directs testis cord formation. Developmental biology, 2009. 326(1): p. 112‐20. 17. Tilmann, C. and B. Capel, Mesonephric cell migration induces testis cord formation and Sertoli cell differentiation in the mammalian gonad. Development, 1999. 126(13): p. 2883‐90. 18. Merchant‐Larios, H. and J. Chimal‐Monroy, The ontogeny of primordial follicles in the mouse ovary. Prog Clin Biol Res, 1989. 296: p. 55‐63. 19. Edson, M.A., A.K. Nagaraja, and M.M. Matzuk, The mammalian ovary from genesis to revelation. Endocrine reviews, 2009. 30(6): p. 624‐712. 20. Hirshfield, A.N., Heterogeneity of cell populations that contribute to the formation of primordial follicles in rats. Biol Reprod, 1992. 47(3): p. 466‐72.
131 21. Zheng, W., et al., Two classes of ovarian primordial follicles exhibit distinct developmental dynamics and physiological functions. Hum Mol Genet, 2013. 22. Mork, L., et al., Temporal differences in granulosa cell specification in the ovary reflect distinct follicle fates in mice. Biol Reprod, 2012. 86(2): p. 37. 23. Eppig, J.J. and M.A. Handel, Origins of granulosa cells clarified and complexified by waves. Biology of reproduction, 2012. 86(2): p. 34. 24. Hirshfield, A.N., Development of follicles in the mammalian ovary. International review of cytology, 1991. 124: p. 43‐101. 25. Pinkerton, J.H., et al., Development of the human ovary‐‐a study using histochemical technics. Obstet Gynecol, 1961. 18: p. 152‐81. 26. Albrecht, K.H. and E.M. Eicher, Evidence that Sry is expressed in pre‐Sertoli cells and Sertoli and granulosa cells have a common precursor. Dev Biol, 2001. 240(1): p. 92‐ 107. 27. Byskov, A.G. and S. Lintern‐Moore, Follicle formation in the immature mouse ovary: the role of the rete ovarii. J Anat, 1973. 116(Pt 2): p. 207‐17. 28. Byskov, A.G. and G. Rasmussen, Ultrastructural studies of the developing follicle. In The Development and Maturation of the Ovary and its Functions. Excerpta Medica International Congress Series, 1973. 267: p. 55‐62. 29. Byskov, A.G., The role of the rete ovarii in meiosis and follicle formation in the cat, mink and ferret. J Reprod Fertil, 1975. 45(2): p. 201‐9. 30. Zamboni L, Bezard J, and M. P., The role of the mesonephros in the development of the sheep fetal ovary. Ann Biol Anim Biochem Biophys, 1975. 19: p. 1153‐1178. 31. Byskov, A.G., The anatomy and ultrastructure of the rete system in the fetal mouse ovary. Biol Reprod, 1978. 19(4): p. 720‐35. 32. Sawyer, H.R., et al., Formation of ovarian follicles during fetal development in sheep. Biol Reprod, 2002. 66(4): p. 1134‐50. 33. Motta, P.M. and S. Makabe, Development of the ovarian surface and associated germ cells in the human fetus. A correlated study by scanning and transmission electron microscopy. Cell Tissue Res, 1982. 226(3): p. 493‐510. 34. Gondos, B., Surface epithelium of the developing ovary. Possible correlation with ovarian neoplasia. Am J Pathol, 1975. 81(2): p. 303‐21. 35. McLaren, A., Germ and somatic cell lineages in the developing gonad. Mol Cell Endocrinol, 2000. 163(1‐2): p. 3‐9. 36. McLaren, A., Development of the mammalian gonad: the fate of the supporting cell lineage. Bioessays, 1991. 13(4): p. 151‐6. 37. Matson, C.K., et al., DMRT1 prevents female reprogramming in the postnatal mammalian testis. Nature, 2011. 476(7358): p. 101‐4. 38. Uhlenhaut, N.H., et al., Somatic sex reprogramming of adult ovaries to testes by FOXL2 ablation. Cell, 2009. 139(6): p. 1130‐42. 39. Pailhoux, E., et al., Ontogenesis of female‐to‐male sex‐reversal in XX polled goats. Dev Dyn, 2002. 224(1): p. 39‐50. 40. Pailhoux, E., et al., A 11.7‐kb deletion triggers intersexuality and polledness in goats. Nat Genet, 2001. 29(4): p. 453‐8. 41. Wang, D., et al., Molecular cloning and gene expression of Foxl2 in the Nile tilapia, Oreochromis niloticus. Biochem Biophys Res Commun, 2004. 320(1): p. 83‐9.
132 42. Loffler, K.A., D. Zarkower, and P. Koopman, Etiology of ovarian failure in blepharophimosis ptosis epicanthus inversus syndrome: FOXL2 is a conserved, early‐ acting gene in vertebrate ovarian development. Endocrinology, 2003. 144(7): p. 3237‐43. 43. Schmidt, D., et al., The murine winged‐helix transcription factor Foxl2 is required for granulosa cell differentiation and ovary maintenance. Development, 2004. 131(4): p. 933‐42. 44. Suzuki, K. and B.I. Tamaoki, Postovulatory decrease in estrogen production is caused by the diminished supply of aromatizable androgen to ovarian aromatase. Endocrinology, 1980. 107(6): p. 2115‐6. 45. Ryan, K.J. and Z. Petro, Steroid biosynthesis by human ovarian granulosa and thecal cells. J Clin Endocrinol Metab, 1966. 26(1): p. 46‐52. 46. Bjersing, L., On the morphology and endocrine function of granulosa cells in ovarian follicles and corpora lutea. Biochemical, histochemical, and ultrastructural studies on the porcine ovary with special reference to steroid hormone synthesis. Acta Endocrinol (Copenh), 1968: p. Suppl 125:1‐23. 47. Uilenbroek, J.T., P.J. Woutersen, and P. van der Schoot, Atresia of preovulatory follicles: gonadotropin binding and steroidogenic activity. Biol Reprod, 1980. 23(1): p. 219‐29. 48. Erickson, G.F., et al., The ovarian androgen producing cells: a review of structure/function relationships. Endocr Rev, 1985. 6(3): p. 371‐99. 49. Orisaka, M., et al., Oocyte‐granulosa‐theca cell interactions during preantral follicular development. J Ovarian Res, 2009. 2(1): p. 9. 50. Young, J.M. and A.S. McNeilly, Theca: the forgotten cell of the ovarian follicle. Reproduction, 2010. 140(4): p. 489‐504. 51. Orisaka, M., et al., Effects of ovarian theca cells on granulosa cell differentiation during gonadotropin‐independent follicular growth in cattle. Mol Reprod Dev, 2006. 73(6): p. 737‐44. 52. Gougeon, A., Regulation of ovarian follicular development in primates: facts and hypotheses. Endocr Rev, 1996. 17(2): p. 121‐55. 53. Braw‐Tal, R. and S. Yossefi, Studies in vivo and in vitro on the initiation of follicle growth in the bovine ovary. J Reprod Fertil, 1997. 109(1): p. 165‐71. 54. Quattropani, S.L., Morphogenesis of the ovarian interstitial tissue in the neonatal mouse. Anat Rec, 1973. 177(4): p. 569‐83. 55. Hirshfield, A., Theca cells may be present at the outset of follicular growth. Biology of reproduction, 1991. 44(6): p. 1157‐1162. 56. Peters, H. and T. Pedersen, Origin of follicle cells in the infant mouse ovary. Fertil Steril, 1967. 18(3): p. 309‐13. 57. Honda, A., et al., Isolation, characterization, and in vitro and in vivo differentiation of putative thecal stem cells. Proceedings of the National Academy of Sciences of the United States of America, 2007. 104(30): p. 12389‐94. 58. Orisaka, M., et al., Granulosa cells promote differentiation of cortical stromal cells into theca cells in the bovine ovary. Biol Reprod, 2006. 75(5): p. 734‐40. 59. Kotsuji, F., et al., Bovine theca and granulosa cell interactions modulate their growth, morphology, and function. Biol Reprod, 1990. 43(5): p. 726‐32.
133 60. Magoffin, D.A. and P.C. Magarelli, Preantral follicles stimulate luteinizing hormone independent differentiation of ovarian theca‐interstitial cells by an intrafollicular paracrine mechanism. Endocrine, 1995. 3(2): p. 107‐12. 61. Magarelli, P.C., R.J. Zachow, and D.A. Magoffin, Developmental and hormonal regulation of rat theca‐cell differentiation factor secretion in ovarian follicles. Biology of Reproduction, 1996. 55(2): p. 416‐420. 62. Everett, N.B., Observational and experimental evidences relating to the origin and differentiation of the definitive germ cells in mice. J. exp. Zool., 1943. 92: p. 49‐91. 63. Upadhyay, S., J. Luciani, and L. Zamboni, Role of the mesonephros in the development of indifferent gonads and ovaries of the mouse. Annales de biologie animale, biochimie, biophysique, 1979. 19(NB4): p. 1179‐1196. 64. Grinsted, J., INFLUENCE OF MESONEPHROS ON FETAL AND NEONATAL RABBIT GONADS .2. SEX‐STEROID RELEASE BY THE OVARY INVITRO. Acta endocrinologica, 1982. 99(2): p. 281‐287. 65. Bezard, J., P. Mauleon, and L. Zamboni, ROLE OF THE MESONEPHROS IN THE DEVELOPMENT OF THE SHEEP FETAL OVARY. Annales de biologie animale, biochimie, biophysique, 1979. 19(NB4): p. 1153‐1178. 66. Wartenberg, H., Development of the early human ovary and role of the mesonephros in the differentiation of the cortex. Anat Embryol (Berl), 1982. 165(2): p. 253‐80. 67. Zuckerman, S., The Number of Oocytes in the Mature Ovary. Recent Progress in Hormone Research, 1951. 6: p. 63‐109. 68. Johnson, J., et al., Oocyte generation in adult mammalian ovaries by putative germ cells in bone marrow and peripheral blood. Cell, 2005. 122(2): p. 303‐15. 69. Johnson, J., et al., Germline stem cells and follicular renewal in the postnatal mammalian ovary. Nature, 2004. 428(6979): p. 145‐50. 70. Zou, K., et al., Production of offspring from a germline stem cell line derived from neonatal ovaries. Nat Cell Biol, 2009. 11(5): p. 631‐6. 71. Speed, R.M., Meiosis in the foetal mouse ovary. I. An analysis at the light microscope level using surface‐spreading. Chromosoma, 1982. 85(3): p. 427‐37. 72. Borum, K., Oogenesis in the mouse. A study of the meiotic prophase. Exp Cell Res, 1961. 24: p. 495‐507. 73. Lewis, W., et al., Principles of Development2006: Oxford University Press. 74. Hilscher, B., et al., Kinetics of gametogenesis. I. Comparative histological and autoradiographic studies of oocytes and transitional prospermatogonia during oogenesis and prespermatogenesis. Cell and tissue research, 1974. 154(4): p. 443‐ 470. 75. Evans, E.P., C.E. Ford, and M.F. Lyon, Direct evidence of the capacity of the XY germ cell in the mouse to become an oocyte. Nature, 1977. 267(5610): p. 430‐431. 76. Mystkowska, E.T. and A.K. Tarkowski, Behaviour of germ cells and sexual differentiation in late embryonic and early postnatal mouse chimeras. J Embryol Exp Morphol, 1970. 23(2): p. 395‐405. 77. McLaren, A., A.C. Chandley, and S. Kofman‐Alfaro, A study of meiotic germ cells in the gonads of foetal mouse chimaeras. J Embryol Exp Morphol, 1972. 27(3): p. 515‐24. 78. Palmer, S.J. and P.S. Burgoyne, In situ analysis of fetal, prepuberal and adult XX‐‐‐‐XY chimaeric mouse testes: Sertoli cells are predominantly, but not exclusively, XY. Development (Cambridge, England), 1991. 112(1): p. 265‐268.
134 79. McLaren, A., Germ cells and germ cell sex. Philosophical transactions of the Royal Society of London.Series B, Biological sciences, 1995. 350(1333): p. 229‐233. 80. McLaren, A., Primordial germ cells in the mouse. Developmental biology, 2003. 262(1): p. 1‐15. 81. Byskov, A.G., Does the rete ovarii act as a trigger for the onset of meiosis? Nature, 1974. 252(5482): p. 396‐397. 82. Byskov, A.G. and L. Saxen, Induction of meiosis in fetal mouse testis in vitro. Developmental biology, 1976. 52(2): p. 193‐200. 83. Grinsted, J. and A.G. Byskov, Meiosis‐inducing and meiosis‐preventing substances in human male reproductive organs. Fertil Steril, 1981. 35(2): p. 199‐204. 84. Byskov, A.G., et al., Chemical structure of sterols that activate oocyte meiosis. Nature, 1995. 374(6522): p. 559‐562. 85. Byskov, A.G., et al., Forskolin and the meiosis inducing substance synergistically initiate meiosis in fetal male germ cells. Molecular reproduction and development, 1993. 34(1): p. 47‐52. 86. Byskov, A.G., M. Baltsen, and C.Y. Andersen, Meiosis‐activating sterols: background, discovery, and possible use. J Mol Med, 1998. 76(12): p. 818‐23. 87. Gondos, B., A.G. Byskov, and J.L. Hansen, Regulation of the onset of meiosis in the developing testis. Ann Clin Lab Sci, 1996. 26(5): p. 421‐5. 88. McLaren, A., Meiosis and differentiation of mouse germ cells. Symp Soc Exp Biol, 1984. 38: p. 7‐23. 89. McLaren, A. and D. Southee, Entry of mouse embryonic germ cells into meiosis. Developmental biology, 1997. 187(1): p. 107‐113. 90. Upadhyay, S. and L. Zamboni, Ectopic germ cells: natural model for the study of germ cell sexual differentiation. Proceedings of the National Academy of Sciences of the United States of America, 1982. 79(21): p. 6584‐6588. 91. Chuma, S. and N. Nakatsuji, Autonomous transition into meiosis of mouse fetal germ cells in vitro and its inhibition by gp130‐mediated signaling. Developmental biology, 2001. 229(2): p. 468‐479. 92. McLaren, A., Studies on mouse germ cells inside and outside the gonad. The Journal of experimental zoology, 1983. 228(2): p. 167‐171. 93. Zamboni, L. and S. Upadhyay, Germ cell differentiation in mouse adrenal glands. J Exp Zool, 1983. 228(2): p. 173‐93. 94. Donovan, P.J., et al., Migratory and postmigratory mouse primordial germ cells behave differently in culture. Cell, 1986. 44(6): p. 831‐838. 95. Koubova, J., et al., Retinoic acid regulates sex‐specific timing of meiotic initiation in mice. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(8): p. 2474‐9. 96. Bowles, J., et al., Retinoid signaling determines germ cell fate in mice. Science, 2006. 312(5773): p. 596‐600. 97. Yashiro, K., et al., Regulation of retinoic acid distribution is required for proximodistal patterning and outgrowth of the developing mouse limb. Dev Cell, 2004. 6(3): p. 411‐ 22. 98. Romand, R., P. Dolle, and E. Hashino, Retinoid signaling in inner ear development. J Neurobiol, 2006. 66(7): p. 687‐704.
135 99. Hernandez, R.E., et al., Cyp26 enzymes generate the retinoic acid response pattern necessary for hindbrain development. Development, 2007. 134(1): p. 177‐87. 100. White, J.A., et al., Identification of the human cytochrome P450, P450RAI‐2, which is predominantly expressed in the adult cerebellum and is responsible for all‐trans‐ retinoic acid metabolism. Proceedings of the National Academy of Sciences of the United States of America, 2000. 97(12): p. 6403‐6408. 101. Bowles, J., et al., Retinoid signaling determines germ cell fate in mice. Science (New York, N.Y.), 2006. 312(5773): p. 596‐600. 102. Koubova, J., et al., Retinoic acid regulates sex‐specific timing of meiotic initiation in mice. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(8): p. 2474‐2479. 103. Baltus, A.E., et al., In germ cells of mouse embryonic ovaries, the decision to enter meiosis precedes premeiotic DNA replication. Nature genetics, 2006. 38(12): p. 1430‐ 1434. 104. MacLean, G., et al., Apoptotic extinction of germ cells in testes of Cyp26b1 knockout mice. Endocrinology, 2007. 148(10): p. 4560‐4567. 105. Bowles, J. and P. Koopman, Retinoic acid, meiosis and germ cell fate in mammals. Development (Cambridge, England), 2007. 134(19): p. 3401‐3411. 106. Bullejos, M. and P. Koopman, Germ cells enter meiosis in a rostro‐caudal wave during development of the mouse ovary. Mol Reprod Dev, 2004. 68(4): p. 422‐8. 107. Menke, D.B. and D.C. Page, Sexually dimorphic gene expression in the developing mouse gonad. Gene Expr Patterns, 2002. 2(3‐4): p. 359‐67. 108. Yao, H.H., L. DiNapoli, and B. Capel, Meiotic germ cells antagonize mesonephric cell migration and testis cord formation in mouse gonads. Development, 2003. 130(24): p. 5895‐902. 109. Lin, Y., et al., Germ cell‐intrinsic and ‐extrinsic factors govern meiotic initiation in mouse embryos. Science (New York, N.Y.), 2008. 322(5908): p. 1685‐1687. 110. Tsuda, M., et al., Conserved role of nanos proteins in germ cell development. Science (New York, N.Y.), 2003. 301(5637): p. 1239‐1241. 111. Suzuki, A. and Y. Saga, Nanos2 suppresses meiosis and promotes male germ cell differentiation. Genes & development, 2008. 22(4): p. 430‐435. 112. Suzuki, A., et al., NANOS2 interacts with the CCR4‐NOT deadenylation complex and leads to suppression of specific RNAs. Proc Natl Acad Sci USA, 2010. 107(8): p. 3594‐9. 113. Kumar, S., et al., Sex‐specific timing of meiotic initiation is regulated by Cyp26b1 independent of retinoic acid signalling. Nat Commun, 2011. 2: p. 151. 114. Pepling, M.E. and A.C. Spradling, Mouse ovarian germ cell cysts undergo programmed breakdown to form primordial follicles. Developmental biology, 2001. 234(2): p. 339‐ 51. 115. Ratts, V.S., et al., Ablation of bcl‐2 gene expression decreases the numbers of oocytes and primordial follicles established in the post‐natal female mouse gonad. Endocrinology, 1995. 136(8): p. 3665‐8. 116. Benedict, J.C., et al., Physiological role of the aryl hydrocarbon receptor in mouse ovary development. Toxicol Sci, 2000. 56(2): p. 382‐8. 117. Bergeron, L., et al., Defects in regulation of apoptosis in caspase‐2‐deficient mice. Genes & development, 1998. 12(9): p. 1304‐14.
136 118. Soyal, S.M., A. Amleh, and J. Dean, FIGalpha, a germ cell‐specific transcription factor required for ovarian follicle formation. Development, 2000. 127(21): p. 4645‐54. 119. Trombly, D.J., T.K. Woodruff, and K.E. Mayo, Suppression of Notch signaling in the neonatal mouse ovary decreases primordial follicle formation. Endocrinology, 2009. 150(2): p. 1014‐24. 120. Hahn, K.L., et al., Lunatic fringe null female mice are infertile due to defects in meiotic maturation. Development, 2005. 132(4): p. 817‐828. 121. Jefferson, W., et al., Neonatal genistein treatment alters ovarian differentiation in the mouse: Inhibition of oocyte nest breakdown and increased oocyte survival. Biology of Reproduction, 2006. 74(1): p. 161‐168. 122. Chen, Y., et al., Estradiol, progesterone, and genistein inhibit oocyte nest breakdown and primordial follicle assembly in the neonatal mouse ovary in vitro and in vivo. Endocrinology, 2007. 148(8): p. 3580‐3590. 123. Jefferson, W.N., et al., Neonatal exposure to genistein induces estrogen receptor (ER)alpha expression and multioocyte follicles in the maturing mouse ovary: Evidence for ER beta‐mediated and nonestrogenic actions. Biology of Reproduction, 2002. 67(4): p. 1285‐1296. 124. Choi, Y., D. Yuan, and A. Rajkovic, Germ Cell‐Specific Transcriptional Regulator Sohlh2 Is Essential for Early Mouse Folliculogenesis and Oocyte‐Specific Gene Expression. Biology of Reproduction, 2008. 79(6): p. 1176‐1182. 125. Hao, J., et al., Sohlh2 knockout mice are male‐sterile because of degeneration of differentiating type A spermatogonia. Stem Cells, 2008. 26(6): p. 1587‐1597. 126. Pangas, S.A., et al., Oogenesis requires germ cell‐specific transcriptional regulators Sohlh1 and Lhx8. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(21): p. 8090‐8095. 127. Rajkovic, A., et al., NOBOX deficiency disrupts early folliculogenesis and oocyte‐specific gene expression. Science, 2004. 305(5687): p. 1157‐1159. 128. Horie, K., et al., The Expression of C‐Kit Protein during Oogenesis and Early Embryonic‐Development. Biology of Reproduction, 1991. 45(4): p. 547‐552. 129. Joyce, I.M., et al., Oocyte regulation of kit ligand expression in mouse ovarian follicles. Developmental Biology, 1999. 214(2): p. 342‐353. 130. Manova, K., et al., The Expression Pattern of the C‐Kit Ligand in Gonads of Mice Supports a Role for the C‐Kit Receptor in Oocyte Growth and in Proliferation of Spermatogonia. Developmental Biology, 1993. 157(1): p. 85‐99. 131. Manova, K., et al., Gonadal Expression of C‐Kit Encoded at the W‐Locus of the Mouse. Development, 1990. 110(4): p. 1057‐1069. 132. Yoshida, H., et al., Stepwise requirement of c‐kit tyrosine kinase in mouse ovarian follicle development. Developmental Biology, 1997. 184(1): p. 122‐137. 133. Parrott, J.A. and M.K. Skinner, Kit‐ligand/stem cell factor induces primordial follicle development and initiates folliculogenesis. Endocrinology, 1999. 140(9): p. 4262‐ 4271. 134. Reddy, P., et al., Activation of Akt (PKB) and suppression of FKHRL1 in mouse and rat oocytes by stem cell factor during follicular activation and development. Developmental Biology, 2005. 281(2): p. 160‐170. 135. Accili, D. and K.C. Arden, FoxOs at the crossroads of cellular metabolism, differentiation, and transformation. Cell, 2004. 117(4): p. 421‐426.
137 136. Castrillon, D.H., et al., Suppression of ovarian follicle activation in mice by the transcription factor Foxo3a. Science, 2003. 301(5630): p. 215‐218. 137. Liu, C.F., C. Liu, and H.H. Yao, Building pathways for ovary organogenesis in the mouse embryo. Curr Top Dev Biol, 2010. 90: p. 263‐90. 138. Schmidt, D., et al., The murine winged‐helix transcription factor Foxl2 is required for granulosa cell differentiation and ovary maintenance. Development, 2004. 131(4): p. 933‐942. 139. Uda, M., et al., Foxl2 disruption causes mouse ovarian failure by pervasive blockage of follicle development. Hum Mol Genet, 2004. 13(11): p. 1171‐1181. 140. Visser, J.A., et al., Anti‐Mullerian hormone: a new marker for ovarian function. Reproduction, 2006. 131(1): p. 1‐9. 141. Mishina, Y., et al., Genetic analysis of the Mullerian‐inhibiting substance signal transduction pathway in mammalian sexual differentiation. Genes & development, 1996. 10(20): p. 2577‐2587. 142. Durlinger, A.L.L., et al., Control of primordial follicle recruitment by anti‐Mullerian hormone in the mouse ovary. Endocrinology, 1999. 140(12): p. 5789‐5796. 143. Durlinger, A.L.L., et al., Anti‐Mullerian hormone inhibits initiation of primordial follicle growth in the mouse ovary. Endocrinology, 2002. 143(3): p. 1076‐1084. 144. Holt, J.E., et al., CXCR4/SDF1 interaction inhibits the primordial to primary follicle transition in the neonatal mouse ovary. Developmental Biology, 2006. 293(2): p. 449‐460. 145. Eppig, J.J., K. Wigglesworth, and F.L. Pendola, The mammalian oocyte orchestrates the rate of ovarian follicular development. Proceedings of the National Academy of Sciences of the United States of America, 2002. 99(5): p. 2890‐4. 146. Dong, J., et al., Growth differentiation factor‐9 is required during early ovarian folliculogenesis. Nature, 1996. 383(6600): p. 531‐5. 147. Elvin, J.A., et al., Molecular characterization of the follicle defects in the growth differentiation factor 9‐deficient ovary. Molecular endocrinology, 1999. 13(6): p. 1018‐34. 148. Wu, X., et al., Interrelationship of growth differentiation factor 9 and inhibin in early folliculogenesis and ovarian tumorigenesis in mice. Molecular endocrinology, 2004. 18(6): p. 1509‐19. 149. Paredes, A., et al., TrkB receptors are required for follicular growth and oocyte survival in the mammalian ovary (vol 267, pg 430, 2004). Developmental Biology, 2004. 270(1): p. 273‐273. 150. Anderson, R.A., et al., Neurotropins and their receptors are expressed in the human fetal ovary. Journal of Clinical Endocrinology & Metabolism, 2002. 87(2): p. 890‐897. 151. Juneja, S.C., et al., Defects in the germ line and gonads of mice lacking connexin43. Biology of Reproduction, 1999. 60(5): p. 1263‐1270. 152. Simon, A.M., et al., Female infertility in mice lacking connexin 37. Nature, 1997. 385(6616): p. 525‐529. 153. Ackert, C.L., et al., Intercellular communication via connexin43 gap junctions is required for ovarian folliculogenesis in the mouse. Developmental Biology, 2001. 233(2): p. 258‐270.
138 154. Carabatsos, M.J., et al., Oocyte‐granulosa cell heterologous gap junctions are required for the coordination of nuclear and cytoplasmic meiotic competence. Developmental Biology, 2000. 226(2): p. 167‐179. 155. Nilsson, E. and M.K. Skinner, Cellular Interactions That Control Primordial Follicle Development and Folliculogenesis. Journal of the Society for Gynecologic Investigation, 2001. 8(1 suppl): p. S17‐S20. 156. Magoffin, D.A., Ovarian theca cell. Int J Biochem Cell Biol, 2005. 37(7): p. 1344‐9. 157. Diaz, F.J., K. Wigglesworth, and J.J. Eppig, Oocytes determine cumulus cell lineage in mouse ovarian follicles. J Cell Sci, 2007. 120(8): p. 1330‐1340. 158. Kumar, T.R., et al., Follicle stimulating hormone is required for ovarian follicle maturation but not male fertility. Nat Genet, 1997. 15(2): p. 201‐204. 159. Abel, M.H., et al., The effect of a null mutation in the follicle‐stimulating hormone receptor gene on mouse reproduction. Endocrinology, 2000. 141(5): p. 1795‐1803. 160. Chun, S.Y., et al., Hormonal regulation of apoptosis in early antral follicles: Follicle‐ stimulating hormone as a major survival factor. Endocrinology, 1996. 137(4): p. 1447‐1456. 161. Richards, J.S., Hormonal‐Control of Gene‐Expression in the Ovary. Endocrine reviews, 1994. 15(6): p. 725‐751. 162. Lee, S.L., et al., Luteinizing hormone deficiency and female infertility in mice lacking the transcription factor NGFI‐A (Egr‐1). Science, 1996. 273(5279): p. 1219‐1221. 163. Dierich, A., et al., Impairing follicle‐stimulating hormone (FSH) signaling in vivo: Targeted disruption of the FSH receptor leads to aberrant gametogenesis and hormonal imbalance. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(23): p. 13612‐13617. 164. Fisher, C.R., et al., Characterization of mice deficient in aromatase (ArKO) because of targeted disruption of the cyp19 gene. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(12): p. 6965‐6970. 165. Hewitt, S.C., B.J. Deroo, and K.S. Korach, A new mediator for an old hormone? Science, 2005. 307(5715): p. 1572‐1573. 166. Britt, K.L., et al., An age‐related ovarian phenotype in mice with targeted disruption of the Cyp 19 (aromatase) gene. Endocrinology, 2000. 141(7): p. 2614‐2623. 167. Britt, K.L. and J.K. Findlay, Regulation of the phenotype of ovarian somatic cells by estrogen. Molecular and Cellular Endocrinology, 2003. 202(1‐2): p. 11‐17. 168. Lubahn, D.B., et al., Alteration of Reproductive Function but Not Prenatal Sexual Development after Insertional Disruption of the Mouse Estrogen‐Receptor Gene. Proceedings of the National Academy of Sciences of the United States of America, 1993. 90(23): p. 11162‐11166. 169. Dunaif, A., Polycystic ovary syndrome in 2011: Genes, aging and sleep apnea in polycystic ovary syndrome. Nature reviews. Endocrinology, 2012. 8(2): p. 72‐4. 170. Krege, J.H., et al., Generation and reproductive phenotypes of mice lacking estrogen receptor beta. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(26): p. 15677‐15682. 171. Couse, J.F., et al., Postnatal sex reversal of the ovaries in mice lacking estrogen receptors alpha and beta. Science, 1999. 286(5448): p. 2328‐2331.
139 172. Dupont, S., et al., Effect of single and compound knockouts of estrogen receptors alpha (ER alpha) and beta (ER beta) on mouse reproductive phenotypes. Development, 2000. 127(19): p. 4277‐4291. 173. Hu, Y.C., et al., Subfertility and defective folliculogenesis in female mice lacking androgen receptor. Proceedings of the National Academy of Sciences of the United States of America, 2004. 101(31): p. 11209‐11214. 174. Shiina, H., et al., Premature ovarian failure in androgen receptor‐deficient mice. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(1): p. 224‐229. 175. Kimura, S., et al., Androgen receptor function in folliculogenesis and its clinical implication in premature ovarian failure. Trends in Endocrinology and Metabolism, 2007. 18(5): p. 183‐189. 176. Lei, Z.M., et al., Targeted disruption of luteinizing hormone/human chorionic gonadotropin receptor gene. Molecular endocrinology, 2001. 15(1): p. 184‐200. 177. Zhang, F.P., et al., Normal prenatal but arrested postnatal sexual development of luteinizing hormone receptor knockout (LuRKO) mice. Molecular endocrinology, 2001. 15(1): p. 172‐183. 178. Park, J.Y., et al., EGF‐like growth factors as mediators of LH action in the ovulatory follicle. Science, 2004. 303(5658): p. 682‐684. 179. Peng, X.R., et al., Localization of Luteinizing‐Hormone Receptor Messenger‐ Ribonucleic‐Acid Expression in Ovarian Cell‐Types during Follicle Development and Ovulation. Endocrinology, 1991. 129(6): p. 3200‐3207. 180. Su, Y.Q., et al., Mitogen‐activated protein kinase activity in cumulus cells is essential for gonadotropin‐induced oocyte meiotic resumption and cumulus expansion in the mouse. Endocrinology, 2002. 143(6): p. 2221‐2232. 181. Fan, H.Y., et al., MAPK3/1 (ERK1/2) in Ovarian Granulosa Cells Are Essential for Female Fertility. Science, 2009. 324(5929): p. 938‐941. 182. Dinchuk, J.E., et al., Renal Abnormalities and an Altered Inflammatory Response in Mice Lacking Cyclooxygenase‐Ii. Nature, 1995. 378(6555): p. 406‐409. 183. Lim, H., et al., Multiple female reproductive failures in cyclooxygenase 2‐deficient mice. Cell, 1997. 91(2): p. 197‐208. 184. Sugiura, K., Y.Q. Su, and J.J. Eppig, Targeted Suppression of Has2 mRNA in Mouse Cumulus Cell‐Oocyte Complexes by Adenovirus‐Mediated Short‐Hairpin RNA Expression. Molecular reproduction and development, 2009. 76(6): p. 537‐547. 185. Scarchilli, L., et al., PTX3 interacts with inter‐alpha‐trypsin inhibitor ‐ Implications for hyaluronan organization and cumulus oophorus expansion. Journal of Biological Chemistry, 2007. 282(41): p. 30161‐30170. 186. Fulop, C., et al., Impaired cumulus mucification and female sterility in tumor necrosis factor‐induced protein‐6 deficient mice. Development, 2003. 130(10): p. 2253‐2261. 187. Su, Y.Q., et al., Oocyte‐dependent activation of mitogen‐activated protein kinase (ERK1/2) in cumulus cells is required for the maturation of the mouse oocyte‐cumulus cell complex. Dev Biol, 2003. 263(1): p. 126‐38. 188. Buccione, R., et al., Fsh‐Induced Expansion of the Mouse Cumulus‐Oophorus Invitro Is Dependent Upon a Specific Factor(S) Secreted by the Oocyte. Developmental Biology, 1990. 138(1): p. 16‐25.
140 189. Buccione, R., et al., FSH‐induced expansion of the mouse cumulus oophorus in vitro is dependent upon a specific factor(s) secreted by the oocyte. Dev Biol, 1990. 138(1): p. 16‐25. 190. Salustri, A., M. Yanagishita, and V.C. Hascall, Mouse Oocytes Regulate Hyaluronic‐Acid Synthesis and Mucification by Fsh‐Stimulated Cumulus Cells. Developmental Biology, 1990. 138(1): p. 26‐32. 191. Eppig, J.J., Oocyte control of ovarian follicular development and function in mammals. Reproduction, 2001. 122(6): p. 829‐38. 192. Gui, U.M. and L.A. Joyce, RNA interference evidence that growth differentiation factor‐ 9 mediates clocyte regulation of cumulus expansion in mice. Biology of Reproduction, 2005. 72(1): p. 195‐199. 193. Elvin, J.A., et al., Paracrine actions of growth differentiation factor‐9 in the mammalian ovary. Molecular endocrinology, 1999. 13(6): p. 1035‐1048. 194. Varani, S., et al., Knockout of pentraxin 3, a downstream target of growth differentiation factor‐9, causes female subfertility. Molecular endocrinology, 2002. 16(6): p. 1154‐1167. 195. Sasseville, M., et al., Growth differentiation factor 9 signaling requires ERK1/2 activity in mouse granulosa and cumulus cells. J Cell Sci, 2010. 123(Pt 18): p. 3166‐76. 196. Yoshino, O., et al., A unique preovulatory expression pattern plays a key role in the physiological functions of BMP‐15 in the mouse. Proceedings of the National Academy of Sciences of the United States of America, 2006. 103(28): p. 10678‐10683. 197. Yan, C.N., et al., Synergistic roles of bone morphogenetic protein 15 and growth differentiation factor 9 in ovarian function. Molecular endocrinology, 2001. 15(6): p. 854‐866. 198. Su, Y.Q., et al., Synergistic roles of BMP15 and GDF9 in the development and function of the oocyte‐cumulus cell complex in mice: genetic evidence for an oocyte‐granulosa cell regulatory loop. Developmental Biology, 2004. 276(1): p. 64‐73. 199. Natraj, U. and J.S. Richards, Hormonal‐Regulation, Localization, and Functional‐ Activity of the Progesterone‐Receptor in Granulosa‐Cells of Rat Preovulatory Follicles. Endocrinology, 1993. 133(2): p. 761‐769. 200. Park, O.K. and K.E. Mayo, Transient Expression of Progesterone‐Receptor Messenger‐ Rna in Ovarian Granulosa‐Cells after the Preovulatory Luteinizing‐Hormone Surge. Molecular endocrinology, 1991. 5(7): p. 967‐978. 201. Lydon, J.P., et al., Mice Lacking Progesterone‐Receptor Exhibit Pleiotropic Reproductive Abnormalities. Genes & development, 1995. 9(18): p. 2266‐2278. 202. Mulac‐Jericevic, B. and O.M. Conneely, Reproductive tissue selective actions of progesterone receptors. Reproduction, 2004. 128(2): p. 139‐146. 203. Mulac‐Jericevic, B., et al., Defective mammary gland morphogenesis in mice lacking the progesterone receptor B isoform. Proceedings of the National Academy of Sciences of the United States of America, 2003. 100(17): p. 9744‐9749. 204. Mulac‐Jericevic, B., et al., Subgroup of reproductive functions of progesterone mediated by progesterone receptor‐B isoform. Science, 2000. 289(5485): p. 1751‐ 1754. 205. Duggavathi, R., et al., Liver receptor homolog 1 is essential for ovulation. Genes & development, 2008. 22(14): p. 1871‐1876.
141 206. Pelusi, C., et al., Impaired Follicle Development and Infertility in Female Mice Lacking Steroidogenic Factor 1 in Ovarian Granulosa Cells. Biology of Reproduction, 2008. 79(6): p. 1074‐1083. 207. Collins, L.L., et al., Growth retardation and abnormal maternal behavior in mice lacking testicular orphan nuclear receptor 4. Proceedings of the National Academy of Sciences of the United States of America, 2004. 101(42): p. 15058‐15063. 208. Chen, L.M., et al., Subfertility with defective folliculogenesis in female mice lacking testicular orphan nuclear receptor 4. Molecular endocrinology, 2008. 22(4): p. 858‐ 867. 209. White, R., et al., The nuclear receptor co‐repressor Nrip1 (RIP140) is essential for female fertility. Nature medicine, 2000. 6(12): p. 1368‐1374. 210. Burkart, A.D., et al., Repression of the inhibin alpha‐subunit gene by the transcription factor CCAAT/enhancer‐binding protein‐beta. Endocrinology, 2005. 146(4): p. 1909‐ 21. 211. Sterneck, E., L. Tessarollo, and P.F. Johnson, An essential role for C/EBPbeta in female reproduction. Genes & development, 1997. 11(17): p. 2153‐62. 212. Stocco, C., C. Telleria, and G. Gibori, The molecular control of corpus luteum formation, function, and regression. Endocrine reviews, 2007. 28(1): p. 117‐149. 213. Devoto, L., et al., The human corpus luteum: life cycle and function in natural cycles. Fertility and sterility, 2009. 92(3): p. 1067‐1079. 214. Murphy, B.D., Models of luteinization. Biology of Reproduction, 2000. 63(1): p. 2‐11. 215. Brenkman, A.B. and B.M.T. Burgering, FoxO3a eggs on fertility and aging. Trends in Molecular Medicine, 2003. 9(11): p. 464‐467. 216. Fero, M.L., et al., A syndrome of multiorgan hyperplasia with features of gigantism, tumorigenesis, and female sterility in p27(Kip1)‐deficient mice. Cell, 1996. 85(5): p. 733‐744. 217. Kiyokawa, H., et al., Enhanced growth of mice lacking the cyclin‐dependent kinase inhibitor function of p27(Kip1). Cell, 1996. 85(5): p. 721‐732. 218. Nakayama, K., et al., Mice lacking p27(Kip1) display increased body size, multiple organ hyperplasia, retinal dysplasia, and pituitary tumors. Cell, 1996. 85(5): p. 707‐ 720. 219. Ormandy, C.J., et al., Null mutation of the prolactin receptor gene produces multiple reproductive defects in the mouse. Genes & development, 1997. 11(2): p. 167‐178. 220. Grosdemouge, I., et al., Effects of deletion of the prolactin receptor on ovarian gene expression. Reproductive biology and endocrinology : RB&E, 2003. 1: p. 12. 221. Pangas, S.A., et al., Premature luteinization and cumulus cell defects in ovarian‐specific Smad4 knockout mice. Mol Endocrinol, 2006. 20(6): p. 1406‐22. 222. Jemal, A., et al., Cancer statistics, 2008. CA: a cancer journal for clinicians, 2008. 58(2): p. 71‐96. 223. Bowtell, D.D., The genesis and evolution of high‐grade serous ovarian cancer. Nature reviews. Cancer, 2010. 10(11): p. 803‐8. 224. Bocker, W., [WHO classification of breast tumors and tumors of the female genital organs: pathology and genetics]. Verhandlungen der Deutschen Gesellschaft fur Pathologie, 2002. 86: p. 116‐9. 225. Seidman, J.D., et al., The histologic type and stage distribution of ovarian carcinomas of surface epithelial origin. International journal of gynecological pathology : official
142 journal of the International Society of Gynecological Pathologists, 2004. 23(1): p. 41‐4. 226. Bell, D.A., Origins and molecular pathology of ovarian cancer. Modern pathology : an official journal of the United States and Canadian Academy of Pathology, Inc, 2005. 18 Suppl 2: p. S19‐32. 227. Vaughan, S., et al., Rethinking ovarian cancer: recommendations for improving outcomes. Nature reviews. Cancer, 2011. 11(10): p. 719‐25. 228. Berns, E.M. and D.D. Bowtell, The changing view of high‐grade serous ovarian cancer. Cancer Res, 2012. 72(11): p. 2701‐4. 229. Stewart, B.W. and P. Kleihues, The World Cancer Report‐‐the major findings. Central European journal of public health, 2003. 11(3): p. 177‐9. 230. Lengyel, E., Ovarian cancer development and metastasis. The American journal of pathology, 2010. 177(3): p. 1053‐64. 231. Auersperg, N., et al., Ovarian surface epithelium: biology, endocrinology, and pathology. Endocrine reviews, 2001. 22(2): p. 255‐88. 232. Flesken‐Nikitin, A., et al., Ovarian surface epithelium at the junction area contains a cancer‐prone stem cell niche. Nature, 2013. 495(7440): p. 241‐245. 233. Wang, P.H. and C. Chang, Androgens and ovarian cancers. Eur J Gynaecol Oncol, 2004. 25(2): p. 157‐63. 234. Elattar, A., et al., Androgen receptor expression is a biological marker for androgen sensitivity in high grade serous epithelial ovarian cancer. Gynecologic oncology, 2012. 124(1): p. 142‐147. 235. Kurman, R.J. and M. Shih Ie, The origin and pathogenesis of epithelial ovarian cancer: a proposed unifying theory. The American journal of surgical pathology, 2010. 34(3): p. 433‐43. 236. Piek, J.M., et al., Dysplastic changes in prophylactically removed Fallopian tubes of women predisposed to developing ovarian cancer. The Journal of pathology, 2001. 195(4): p. 451‐6. 237. Lee, Y., et al., A candidate precursor to serous carcinoma that originates in the distal fallopian tube. The Journal of pathology, 2007. 211(1): p. 26‐35. 238. Levanon, K., et al., Primary ex vivo cultures of human fallopian tube epithelium as a model for serous ovarian carcinogenesis. Oncogene, 2010. 29(8): p. 1103‐13. 239. Finch, A., et al., Clinical and pathologic findings of prophylactic salpingo‐ oophorectomies in 159 BRCA1 and BRCA2 carriers. Gynecologic oncology, 2006. 100(1): p. 58‐64. 240. Kindelberger, D.W., et al., Intraepithelial carcinoma of the fimbria and pelvic serous carcinoma: Evidence for a causal relationship. The American journal of surgical pathology, 2007. 31(2): p. 161‐9. 241. Crum, C.P., et al., Lessons from BRCA: the tubal fimbria emerges as an origin for pelvic serous cancer. Clinical medicine & research, 2007. 5(1): p. 35‐44. 242. Knochenhauer, E.S., et al., Prevalence of the polycystic ovary syndrome in unselected black and white women of the southeastern United States: a prospective study. The Journal of clinical endocrinology and metabolism, 1998. 83(9): p. 3078‐82. 243. Legro, R.S. and J.F. Strauss, Molecular progress in infertility: polycystic ovary syndrome. Fertility and sterility, 2002. 78(3): p. 569‐76.
143 244. Balen, A., R. Homburg, and S. Franks, Defining polycystic ovary syndrome. BMJ, 2009. 338: p. a2968. 245. Franks, S., et al., Ovarian morphology is a marker of heritable biochemical traits in sisters with polycystic ovaries. The Journal of clinical endocrinology and metabolism, 2008. 93(9): p. 3396‐402. 246. Franks, S., J. Stark, and K. Hardy, Follicle dynamics and anovulation in polycystic ovary syndrome. Human reproduction update, 2008. 14(4): p. 367‐78. 247. Azziz, R., et al., Criteria for defining polycystic ovary syndrome as a predominantly hyperandrogenic syndrome: An Androgen Excess Society guideline. Journal of Clinical Endocrinology & Metabolism, 2006. 91(11): p. 4237‐4245. 248. Chang, J., et al., Revised 2003 consensus on diagnostic criteria and long‐term health risks related to polycystic ovary syndrome. Fertility and sterility, 2004. 81(1): p. 19‐ 25. 249. Franks, S., et al., Pathogenesis of polycystic ovary syndrome: evidence for a genetically determined disorder of ovarian androgen production. Human fertility, 2000. 3(2): p. 77‐79. 250. Goodarzi, M.O., et al., Polycystic ovary syndrome: etiology, pathogenesis and diagnosis. Nature reviews. Endocrinology, 2011. 7(4): p. 219‐31. 251. Gilling‐Smith, C., et al., Evidence for a primary abnormality of thecal cell steroidogenesis in the polycystic ovary syndrome. Clinical endocrinology, 1997. 47(1): p. 93‐9. 252. Gilling‐Smith, C., et al., Hypersecretion of androstenedione by isolated thecal cells from polycystic ovaries. The Journal of clinical endocrinology and metabolism, 1994. 79(4): p. 1158‐65. 253. Nelson, V.L., et al., Augmented androgen production is a stable steroidogenic phenotype of propagated theca cells from polycystic ovaries. Molecular endocrinology, 1999. 13(6): p. 946‐957. 254. Nelson, V.L., et al., The biochemical basis for increased testosterone production in theca cells propagated from patients with polycystic ovary syndrome. Journal of Clinical Endocrinology & Metabolism, 2001. 86(12): p. 5925‐5933. 255. Wickenheisser, J.K., et al., Differential activity of the cytochrome P450 17alpha‐ hydroxylase and steroidogenic acute regulatory protein gene promoters in normal and polycystic ovary syndrome theca cells. The Journal of clinical endocrinology and metabolism, 2000. 85(6): p. 2304‐11. 256. Magoffin, D.A., Ovarian enzyme activities in women with polycystic ovary syndrome. Fertility and sterility, 2006. 86 Suppl 1: p. S9‐S11. 257. Jakimiuk, A.J., et al., Luteinizing hormone receptor, steroidogenesis acute regulatory protein, and steroidogenic enzyme messenger ribonucleic acids are overexpressed in thecal and granulosa cells from polycystic ovaries. The Journal of clinical endocrinology and metabolism, 2001. 86(3): p. 1318‐23. 258. Wickenheisser, J.K., et al., Cholesterol side‐chain cleavage gene expression in theca cells: augmented transcriptional regulation and mRNA stability in polycystic ovary syndrome. PLoS One, 2012. 7(11): p. e48963. 259. Strauss, J.F., Some New Thoughts on the Pathophysiology and Genetics of Polycystic Ovary Syndrome. Annals of the New York Academy of Sciences, 2003. 997(1): p. 42‐ 48.
144 260. Prapas, N., et al., Genetics of polycystic ovary syndrome. Hippokratia, 2009. 13(4): p. 216‐23. 261. Wood, J.R., et al., The molecular phenotype of polycystic ovary syndrome (PCOS) theca cells and new candidate PCOS genes defined by microarray analysis. J Biol Chem, 2003. 278(29): p. 26380‐90. 262. McAllister, J.M., et al., Overexpression of a DENND1A isoform produces a polycystic ovary syndrome theca phenotype. Proc Natl Acad Sci USA, 2014. 111(15): p. E1519‐ 27. 263. Laml, T., et al., Genetic disorders in premature ovarian failure. Human reproduction update, 2002. 8(5): p. 483‐91. 264. Coulam, C.B., S.C. Adamson, and J.F. Annegers, Incidence of premature ovarian failure. Obstetrics and gynecology, 1986. 67(4): p. 604‐6. 265. Shelling, A.N., Premature ovarian failure. Reproduction, 2010. 140(5): p. 633‐41. 266. Nelson, L.M., Clinical practice. Primary ovarian insufficiency. The New England journal of medicine, 2009. 360(6): p. 606‐14. 267. Aittomaki, K., et al., Mutation in the follicle‐stimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell, 1995. 82(6): p. 959‐68. 268. Shelling, A.N., X chromosome defects and premature ovarian failure. Australian and New Zealand journal of medicine, 2000. 30(1): p. 5‐7. 269. Husebye, E.S. and K. Lovas, Immunology of Addison's disease and premature ovarian failure. Endocrinol Metab Clin North Am, 2009. 38(2): p. 389‐405, ix. 270. Shiina, H., et al., Premature ovarian failure in androgen receptor‐deficient mice. Proc Natl Acad Sci USA, 2006. 103(1): p. 224‐9. 271. Mottershead, D.G., et al., Growth differentiation factor 9:bone morphogenetic protein 15 (GDF9:BMP15) synergism and protein heterodimerization. Proc Natl Acad Sci USA, 2013. 110(25): p. E2257. 272. Peng, J., et al., Growth differentiation factor 9:bone morphogenetic protein 15 heterodimers are potent regulators of ovarian functions. Proc Natl Acad Sci USA, 2013. 110(8): p. E776‐85. 273. Parrott, J.A. and M.K. Skinner, Thecal cell‐granulosa cell interactions involve a positive feedback loop among keratinocyte growth factor, hepatocyte growth factor, and Kit ligand during ovarian follicular development. Endocrinology, 1998. 139(5): p. 2240‐ 5. 274. Welt, C.K., et al., Selective theca cell dysfunction in autoimmune oophoritis results in multifollicular development, decreased estradiol, and elevated inhibin B levels. J Clin Endocrinol Metab, 2005. 90(5): p. 3069‐76. 275. Nusslein‐Volhard, C. and E. Wieschaus, Mutations affecting segment number and polarity in Drosophila. Nature, 1980. 287(5785): p. 795‐801. 276. Varjosalo, M. and J. Taipale, Hedgehog: functions and mechanisms. Genes Dev, 2008. 22(18): p. 2454‐72. 277. Jiang, J. and C.C. Hui, Hedgehog signaling in development and cancer. Developmental cell, 2008. 15(6): p. 801‐12. 278. Franco, H.L. and H.H. Yao, Sex and hedgehog: roles of genes in the hedgehog signaling pathway in mammalian sexual differentiation. Chromosome Res, 2012. 20(1): p. 247‐ 58.
145 279. Bitgood, M.J. and A.P. McMahon, Hedgehog and Bmp genes are coexpressed at many diverse sites of cell‐cell interaction in the mouse embryo. Developmental biology, 1995. 172(1): p. 126‐38. 280. Barsoum, I. and H.H. Yao, Redundant and differential roles of transcription factors Gli1 and Gli2 in the development of mouse fetal Leydig cells. Biol Reprod, 2011. 84(5): p. 894‐9. 281. Yao, H.H., W. Whoriskey, and B. Capel, Desert Hedgehog/Patched 1 signaling specifies fetal Leydig cell fate in testis organogenesis. Genes & development, 2002. 16(11): p. 1433‐40. 282. Bitgood, M.J., L. Shen, and A.P. McMahon, Sertoli cell signaling by Desert hedgehog regulates the male germline. Current biology : CB, 1996. 6(3): p. 298‐304. 283. Pierucci‐Alves, F., A.M. Clark, and L.D. Russell, A developmental study of the Desert hedgehog‐null mouse testis. Biology of reproduction, 2001. 65(5): p. 1392‐402. 284. Clark, A.M., K.K. Garland, and L.D. Russell, Desert hedgehog (Dhh) gene is required in the mouse testis for formation of adult‐type Leydig cells and normal development of peritubular cells and seminiferous tubules. Biology of reproduction, 2000. 63(6): p. 1825‐38. 285. Barsoum, I.B., et al., Activation of the Hedgehog pathway in the mouse fetal ovary leads to ectopic appearance of fetal Leydig cells and female pseudohermaphroditism. Developmental biology, 2009. 329(1): p. 96‐103. 286. Russell, M.C., et al., The hedgehog signaling pathway in the mouse ovary. Biology of reproduction, 2007. 77(2): p. 226‐36. 287. Wijgerde, M., et al., Hedgehog signaling in mouse ovary: Indian hedgehog and desert hedgehog from granulosa cells induce target gene expression in developing theca cells. Endocrinology, 2005. 146(8): p. 3558‐66. 288. St‐Jacques, B., M. Hammerschmidt, and A.P. McMahon, Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes & development, 1999. 13(16): p. 2072‐86. 289. Ren, Y., et al., Dominant activation of the hedgehog signaling pathway in the ovary alters theca development and prevents ovulation. Mol Endocrinol, 2009. 23(5): p. 711‐23. 290. Spicer, L.J., et al., The hedgehog‐patched signaling pathway and function in the mammalian ovary: a novel role for hedgehog proteins in stimulating proliferation and steroidogenesis of theca cells. Reproduction, 2009. 138(2): p. 329‐39. 291. Strauss, J. and Strauss, Some new thoughts on the pathophysiology and genetics of polycystic ovary syndrome WOMEN'S HEALTH AND DISEASE: GYNECOLOGIC AND REPRODUCTIVE ISSUES. Annals of the New York Academy of Sciences, 2003. 997(1): p. 42‐48. 292. Magoffin, D.A., The ovarian androgen‐producing cells: a 2001 perspective. Rev Endocr Metab Disord, 2002. 3(1): p. 47‐53. 293. Rinaldi, S., et al., Endogenous androgens and risk of epithelial ovarian cancer: results from the European Prospective Investigation into Cancer and Nutrition (EPIC). Cancer Epidemiol Biomarkers Prev, 2007. 16(1): p. 23‐9. 294. Modugno, F., Ovarian cancer and polymorphisms in the androgen and progesterone receptor genes: a HuGE review. Am J Epidemiol, 2004. 159(4): p. 319‐35.
146 295. Brennan, J. and B. Capel, One tissue, two fates: molecular genetic events that underlie testis versus ovary development. Nature reviews.Genetics, 2004. 5(7): p. 509‐521. 296. Barsoum, I.B., et al., Activation of the Hedgehog pathway in the mouse fetal ovary leads to ectopic appearance of fetal Leydig cells and female pseudohermaphroditism. Dev Biol, 2009. 329(1): p. 96‐103. 297. Kawai, Y., et al., A missense mutation of the Dhh gene is associated with male pseudohermaphroditic rats showing impaired Leydig cell development. Reproduction, 2011. 141(2): p. 217‐25. 298. Roberts, A.J. and M.K. Skinner, Mesenchymal‐epithelial cell interactions in the ovary: estrogen‐induced theca cell steroidogenesis. Mol Cell Endocrinol, 1990. 72(1): p. R1‐5. 299. Peters, H. and T. Pedersen, Origin of follicle cells in the infant mouse ovary. Fertility and sterility, 1967. 18(3): p. 309‐13. 300. Ahn, S. and A.L. Joyner, Dynamic changes in the response of cells to positive hedgehog signaling during mouse limb patterning. Cell, 2004. 118(4): p. 505‐16. 301. Franco, H.L., et al., Constitutive activation of smoothened leads to female infertility and altered uterine differentiation in the mouse. Biology of reproduction, 2010. 82(5): p. 991‐9. 302. Adhikari, D., et al., The safe use of a PTEN inhibitor for the activation of dormant mouse primordial follicles and generation of fertilizable eggs. PLoS One. 7(6): p. e39034. 303. Beverdam, A. and P. Koopman, Expression profiling of purified mouse gonadal somatic cells during the critical time window of sex determination reveals novel candidate genes for human sexual dysgenesis syndromes. Hum Mol Genet, 2006. 15(3): p. 417‐431. 304. Rodriguez, K.F., et al., Insufficient luteinizing hormone‐induced intracellular signaling disrupts ovulation in preovulatory follicles lacking estrogen receptor‐{beta}. Endocrinology, 2010. 151(6): p. 2826‐34. 305. Pfaffl, M.W., A new mathematical model for relative quantification in real‐time RT‐ PCR. Nucleic Acids Res, 2001. 29(9): p. e45. 306. Liu, C., et al., Investigating the Origins of Somatic Cell Populations in the Perinatal Mouse Ovaries Using Genetic Lineage Tracing and Immunohistochemistry. Germline Development: Methods and Protocols, 2012. 825: p. 211‐221. 307. Madisen, L., et al., A robust and high‐throughput Cre reporting and characterization system for the whole mouse brain. Nature neuroscience, 2010. 13(1): p. 133‐40. 308. Zervas, M., et al., Cell behaviors and genetic lineages of the mesencephalon and rhombomere 1. Neuron, 2004. 43(3): p. 345‐57. 309. Robinson, S.P., et al., Metabolites, pharmacodynamics, and pharmacokinetics of tamoxifen in rats and mice compared to the breast cancer patient. Drug metabolism and disposition: the biological fate of chemicals, 1991. 19(1): p. 36‐43. 310. Gillman, J., The development of the gonads in man, with a consideration of the role of fetal endocrines and the histogenesis of ovarian tumors. In Carnegie Institution of Washington Contributions to embryology1948, Washington,. v. 32, p. 81 ‐131. 311. Armstrong, J.F., et al., The expression of the Wilms' tumour gene, WT1, in the developing mammalian embryo. Mechanisms of development, 1993. 40(1‐2): p. 85‐ 97.
147 312. Horiuchi, A., et al., Up‐regulation of small GTPases, RhoA and RhoC, is associated with tumor progression in ovarian carcinoma. Laboratory Investigation, 2003. 83(6): p. 861‐870. 313. Pang, X.Y., et al., The Interaction of Adrenomedullin and Macrophages Induces Ovarian Cancer Cell Migration via Activation of RhoA Signaling Pathway. International Journal of Molecular Sciences, 2013. 14(2): p. 2774‐2787. 314. Makker, A., et al., PI3K‐Akt‐mTOR and MAPK signaling pathways in polycystic ovarian syndrome, uterine leiomyomas and endometriosis: an update. Gynecological Endocrinology, 2012. 28(3): p. 175‐181. 315. Leung, P.C.K. and J.H. Choi, Endocrine signaling in ovarian surface epithelium and cancer. Human reproduction update, 2007. 13(2): p. 143‐162. 316. Arend, R.C., et al., The Wnt/beta‐catenin pathway in ovarian cancer: A review. Gynecologic oncology, 2013. 131(3): p. 772‐779. 317. Boyer, A., A.K. Goff, and D. Boerboom, WNT signaling in ovarian follicle biology and tumorigenesis. Trends in Endocrinology and Metabolism, 2010. 21(1): p. 25‐32. 318. Mostowska, A., et al., An Analysis of Polymorphisms Within the Wnt Signaling Pathway in Relation to Ovarian Cancer Risk in a Polish Population. Molecular Diagnosis & Therapy, 2014. 18(1): p. 85‐91. 319. Ivanova, A., et al., In vivo genetic ablation by Cre‐mediated expression of diphtheria toxin fragment A. Genesis, 2005. 43(3): p. 129‐35. 320. Wenzel, J.G. and S. Odend'hal, The mammalian rete ovarii: a literature review. Cornell Vet, 1985. 75(3): p. 411‐25. 321. Jagiello, G.V.H.J., Bioregulators of reproduction1981, New York: Academic Press. 322. Stein, L.E. and C.H. Anderson, A qualitative and quantitative study of rete ovarii development in the fetal rat: correlation with the onset of meiosis and follicle cell appearance. Anat Rec, 1979. 193(2): p. 197‐211. 323. Martineau, J., et al., Male‐specific cell migration into the developing gonad. Curr Biol, 1997. 7(12): p. 958‐68. 324. Ungewitter, E.K. and H.H. Yao, How to make a gonad: cellular mechanisms governing formation of the testes and ovaries. Sex Dev, 2013. 7(1‐3): p. 7‐20. 325. Qing, T., et al., Mature oocytes derived from purified mouse fetal germ cells. Human reproduction, 2008. 23(1): p. 54‐61. 326. Zhang, H., et al., Experimental evidence showing that no mitotically active female germline progenitors exist in postnatal mouse ovaries. Proceedings of the National Academy of Sciences of the United States of America, 2012. 109(31): p. 12580‐ 12585. 327. Taketo‐Hosotani, T. and E. Sinclair‐Thompson, Influence of the mesonephros on the development of fetal mouse ovaries following transplantation into adult male and female mice. Dev Biol, 1987. 124(2): p. 423‐30. 328. Husebye, E.S. and K. Lovas, Immunology of Addison's disease and premature ovarian failure. Endocrinology and metabolism clinics of North America, 2009. 38(2): p. 389‐405, ix. 329. Magoffin, D.A. and S.R. Weitsman, Insulin‐like growth factor‐I regulation of luteinizing hormone (LH) receptor messenger ribonucleic acid expression and LH‐ stimulated signal transduction in rat ovarian theca‐interstitial cells. Biol Reprod, 1994. 51(4): p. 766‐75.
148 330. Huang, C.T., et al., Stem cell factor and insulin‐like growth factor‐I stimulate luteinizing hormone‐independent differentiation of rat ovarian theca cells. Biol Reprod, 2001. 64(2): p. 451‐6. 331. Huang, C.C., C. Liu, and H.H. Yao, Investigating the role of adrenal cortex in organization and differentiation of the adrenal medulla in mice. Mol Cell Endocrinol, 2012. 361(1‐2): p. 165‐71. 332. O'Brien, M.J., J.K. Pendola, and J.J. Eppig, A revised protocol for in vitro development of mouse oocytes from primordial follicles dramatically improves their developmental competence. Biology of reproduction, 2003. 68(5): p. 1682‐6. 333. Taipale, J., et al., Effects of oncogenic mutations in Smoothened and Patched can be reversed by cyclopamine. Nature, 2000. 406(6799): p. 1005‐9. 334. Ingham, P.W. and A.P. McMahon, Hedgehog signaling in animal development: paradigms and principles. Genes Dev, 2001. 15(23): p. 3059‐87. 335. Bingham, N.C., et al., Development of a steroidogenic factor 1/Cre transgenic mouse line. Genesis, 2006. 44(9): p. 419‐24. 336. Kilcoyne, K.R., et al., Fetal programming of adult Leydig cell function by androgenic effects on stem/progenitor cells. Proceedings of the National Academy of Sciences of the United States of America, 2014. 111(18): p. E1924‐32. 337. Qin, J., M.J. Tsai, and S.Y. Tsai, Essential roles of COUP‐TFII in Leydig cell differentiation and male fertility. PLoS One, 2008. 3(9): p. e3285. 338. Fisher, C.R., et al., Characterization of mice deficient in aromatase (ArKO) because of targeted disruption of the cyp19 gene. Proceedings of the National Academy of Sciences of the United States of America, 1998. 95(12): p. 6965‐70. 339. Britt, K.L., et al., An age‐related ovarian phenotype in mice with targeted disruption of the Cyp 19 (aromatase) gene. Endocrinology, 2000. 141(7): p. 2614‐23. 340. Britt, K.L., et al., Estrogen regulates development of the somatic cell phenotype in the eutherian ovary. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 2002. 16(11): p. 1389‐97. 341. Marino, J.S., et al., Adipocyte Dysfunction in a Mouse Model of Polycystic Ovary Syndrome (PCOS): Evidence of Adipocyte Hypertrophy and Tissue‐Specific Inflammation. PLoS One, 2012. 7(10). 342. Harada, M., H. Peegel, and K.M.J. Menon, Expression of vascular endothelial growth factor A during ligand‐induced down‐regulation of luteinizing hormone receptor in the ovary. Molecular and Cellular Endocrinology, 2010. 328(1‐2): p. 28‐33. 343. McFee, R.M. and A.S. Cupp, Vascular contributions to early ovarian development: potential roles of VEGFA isoforms. Reproduction Fertility and Development, 2013. 25(2): p. 333‐342. 344. Mcfee, R.M., T.G. Rozell, and A.S. Cupp, The balance of proangiogenic and antiangiogenic VEGFA isoforms regulate follicle development. Cell and Tissue Research, 2012. 349(3): p. 635‐647. 345. Gao, F., et al., Wt1 functions in ovarian follicle development by regulating granulosa cell differentiation. Hum Mol Genet, 2014. 23(2): p. 333‐341. 346. Kadakia, R., et al., Granulosa cell proliferation is impaired in the Igf1 null ovary. Growth Hormone & Igf Research, 2001. 11(4): p. 220‐224.
149 347. Otsuka, F., R.K. Moore, and S. Shimasaki, Biological function and cellular mechanism of bone morphogenetic protein‐6 in the ovary. Journal of Biological Chemistry, 2001. 276(35): p. 32889‐32895. 348. Padua, M.B., et al., Simultaneous Gene Deletion of Gata4 and Gata6 Leads to Early Disruption of Follicular Development and Germ Cell Loss in the Murine Ovary. Biology of Reproduction, 2014. 91(1). 349. Elvin, J.A., et al., Molecular characterization of the follicle defects in the growth differentiation factor 9‐deficient ovary. Molecular endocrinology, 1999. 13(6): p. 1018‐1034. 350. Matzuk, M.M., et al., Intercellular communication in the mammalian ovary: oocytes carry the conversation. Science, 2002. 296(5576): p. 2178‐80. 351. Conn, P.M., Sourcebook of Models for Biomedical Research2007: Humana Press. 352. Marrone, B.L., M. Jamaluddin, and F. Hertelendy, Regional Pattern of Cell Maturation and Progesterone Biosynthesis in the Avian Granulosa‐Cell Layer. Biology of Reproduction, 1990. 42(3): p. 405‐412. 353. Tischkau, S.A., et al., Characterization of the growth center of the avian preovulatory follicle. Biology of Reproduction, 1997. 56(2): p. 469‐474. 354. Perry, M.M., A.B. Gilbert, and A.J. Evans, The structure of the germinal disc region of the hen's ovarian follicle during the rapid growth phase. Journal of anatomy, 1978. 127(Pt 2): p. 379‐92. 355. Bahr, J., et al., Regulation of ovarian functions by the germinal disc region in the chicken. Advances in Comparative Endocrinology, Tomes 1 and 2, 1997: p. 1451‐ 1455. 356. Humphrey, H.H.C. and J.M. Bahr, Germinal disc‐derived epidermal growth factor: A paracrine factor to stimulate proliferation of granulosa cells. Biology of Reproduction, 2001. 64(1): p. 390‐395. 357. Volentine, K.K., H.H.C. Yao, and J.M. Bahr, Epidermal growth factor in the germinal disc and its potential role in follicular development in the chicken. Biology of Reproduction, 1998. 59(3): p. 522‐526. 358. Yao, H.H.C. and J.M. Bahr, Chicken granulosa cells show differential expression of epidermal growth factor (EGF) and luteinizing hormone (LH) receptor messenger RNA and differential responsiveness to EGF and LH dependent upon location of granulosa cells to the germinal disc. Biology of Reproduction, 2001. 64(6): p. 1790‐1796. 359. Wang, Y.J., et al., Epidermal growth factor (EGF) receptor ligands in the chicken ovary: I. Evidence for heparin‐binding EGF‐like growth factor (HB‐EGF) as a potential oocyte‐ derived signal to control granulosa cell proliferation and HB‐EGF and kit ligand expression. Endocrinology, 2007. 148(7): p. 3426‐3440. 360. el‐Fouly, M.A., et al., Role of the ovum in follicular luteinization. Endocrinology, 1970. 87(2): p. 286‐93. 361. Nekola, M.V. and Nalbando.Av, Morphological Changes of Rat Follicular Cells as Influenced by Oocytes. Biology of Reproduction, 1971. 4(2): p. 154‐&. 362. Eppig, J.J., Oocyte control of ovarian follicular development and function in mammals. Reproduction, 2001. 122(6): p. 829‐838. 363. Gilchrist, R.B., M. Lane, and J.G. Thompson, Oocyte‐secreted factors: regulators of cumulus cell function and oocyte quality. Human reproduction update, 2008. 14(2): p. 159‐77.
150 364. Vanderhyden, B.C., Oocyte‐secreted factors regulate granulosa cell steroidogenesis. Zygote, 1996. 4(4): p. 317‐21. 365. Galloway, S.M., et al., Mutations in an oocyte‐derived growth factor gene (BMP15) cause increased ovulation rate and infertility in a dosage‐sensitive manner. Nat Genet, 2000. 25(3): p. 279‐83. 366. McLaren, A., Development of the mammalian gonad: the fate of the supporting cell lineage. BioEssays : news and reviews in molecular, cellular and developmental biology, 1991. 13(4): p. 151‐6. 367. Bucci, L.R. and M.L. Meistrich, Effects of busulfan on murine spermatogenesis: cytotoxicity, sterility, sperm abnormalities, and dominant lethal mutations. Mutation research, 1987. 176(2): p. 259‐68. 368. Hemsworth, B.N. and H. Jackson, Effect of 'busulphan' on the foetal gonad. Nature, 1962. 195: p. 816‐7. 369. Maatouk, D.M., et al., Germ cells are not required to establish the female pathway in mouse fetal gonads. PLoS One, 2012. 7(10): p. e47238. 370. Reader, K.L., et al., Signalling pathways involved in the cooperative effects of ovine and murine GDF9+BMP15‐stimulated thymidine uptake by rat granulosa cells. Reproduction, 2011. 142(1): p. 123‐31. 371. McNatty, K.P., et al., Bone morphogenetic protein 15 and growth differentiation factor 9 co‐operate to regulate granulosa cell function in ruminants. Reproduction, 2005. 129(4): p. 481‐7. 372. Mottershead, D.G., L.J. Ritter, and R.B. Gilchrist, Signalling pathways mediating specific synergistic interactions between GDF9 and BMP15. Mol Hum Reprod, 2012. 18(3): p. 121‐8. 373. Elvin, J.A., et al., Paracrine actions of growth differentiation factor‐9 in the mammalian ovary. Mol Endocrinol, 1999. 13(6): p. 1035‐48. 374. Hayashi, M., et al., Recombinant growth differentiation factor‐9 (GDF‐9) enhances growth and differentiation of cultured early ovarian follicles. Endocrinology, 1999. 140(3): p. 1236‐44. 375. Vitt, U.A., et al., Bone morphogenetic protein receptor type II is a receptor for growth differentiation factor‐9. Biol Reprod, 2002. 67(2): p. 473‐80. 376. Moore, R.K., F. Otsuka, and S. Shimasaki, Molecular basis of bone morphogenetic protein‐15 signaling in granulosa cells. J Biol Chem, 2003. 278(1): p. 304‐10. 377. Pulkki, M.M., et al., The bioactivity of human bone morphogenetic protein‐15 is sensitive to C‐terminal modification: characterization of the purified untagged processed mature region. Mol Cell Endocrinol, 2011. 332(1‐2): p. 106‐15. 378. Kaivo‐Oja, N., et al., Adenoviral gene transfer allows Smad‐responsive gene promoter analyses and delineation of type I receptor usage of transforming growth factor‐beta family ligands in cultured human granulosa luteal cells. J Clin Endocrinol Metab, 2005. 90(1): p. 271‐8. 379. Mazerbourg, S., et al., Growth differentiation factor‐9 signaling is mediated by the type I receptor, activin receptor‐like kinase 5. Mol Endocrinol, 2004. 18(3): p. 653‐65. 380. Kaivo‐Oja, N., et al., Growth differentiation factor‐9 induces Smad2 activation and inhibin B production in cultured human granulosa‐luteal cells. J Clin Endocrinol Metab, 2003. 88(2): p. 755‐62.
151 381. Yi, S.E., et al., The type I BMP receptor BmprIB is essential for female reproductive function. Proc Natl Acad Sci USA, 2001. 98(14): p. 7994‐9. 382. Erickson, G.F. and S. Shimasaki, The spatiotemporal expression pattern of the bone morphogenetic protein family in rat ovary cell types during the estrous cycle. Reprod Biol Endocrinol, 2003. 1: p. 9. 383. Gueripel, X., M. Benahmed, and A. Gougeon, Sequential gonadotropin treatment of immature mice leads to amplification of transforming growth factor beta action, via upregulation of receptor‐type 1, Smad 2 and 4, and downregulation of Smad 6. Biol Reprod, 2004. 70(3): p. 640‐8. 384. Gilchrist, R.B., et al., Molecular basis of oocyte‐paracrine signalling that promotes granulosa cell proliferation. J Cell Sci, 2006. 119(Pt 18): p. 3811‐21. 385. Feary, E.S., et al., Patterns of expression of messenger RNAs encoding GDF9, BMP15, TGFBR1, BMPR1B, and BMPR2 during follicular development and characterization of ovarian follicular populations in ewes carrying the Woodlands FecX2W mutation. Biol Reprod, 2007. 77(6): p. 990‐8. 386. Rosairo, D., et al., Transforming growth factor‐beta: its role in ovarian follicle development. Reproduction, 2008. 136(6): p. 799‐809. 387. Paradis, F., et al., Temporal regulation of BMP2, BMP6, BMP15, GDF9, BMPR1A, BMPR1B, BMPR2 and TGFBR1 mRNA expression in the oocyte, granulosa and theca cells of developing preovulatory follicles in the pig. Reproduction, 2009. 138(1): p. 115‐29. 388. Li, Q., et al., Transforming growth factor beta receptor type 1 is essential for female reproductive tract integrity and function. PLoS Genet, 2011. 7(10): p. e1002320. 389. Spicer, L.J., et al., Growth differentiation factor 9 (GDF9) stimulates proliferation and inhibits steroidogenesis by bovine theca cells: influence of follicle size on responses to GDF9. Biol Reprod, 2008. 78(2): p. 243‐53. 390. Juengel, J.L., et al., The role of transforming growth factor‐beta (TGF‐beta) during ovarian follicular development in sheep. Reprod Biol Endocrinol, 2004. 2: p. 78. 391. Schmid, P., et al., Expression of TGF‐beta s and TGF‐beta type II receptor mRNAs in mouse folliculogenesis: stored maternal TGF‐beta 2 message in oocytes. Biochem Biophys Res Commun, 1994. 201(2): p. 649‐56. 392. Yamamoto, N., et al., Growth differentiation factor‐9 inhibits 3'5'‐adenosine monophosphate‐stimulated steroidogenesis in human granulosa and theca cells. J Clin Endocrinol Metab, 2002. 87(6): p. 2849‐56. 393. Chen, M.H., et al., Growth differential factor‐9 inhibits testosterone production in mouse theca interstitial cells. Fertil Steril, 2013. 100(5): p. 1444‐50. 394. Solovyeva, E.V., et al., Growth differentiation factor‐9 stimulates rat theca‐interstitial cell androgen biosynthesis. Biol Reprod, 2000. 63(4): p. 1214‐8. 395. Qvist, R., et al., Development of mouse ovarian follicles from primary to preovulatory stages in vitro. Journal of reproduction and fertility, 1990. 89(1): p. 169‐80.
152