CHAPTER 25

Steroid Receptors in the Uterus and Ovary April K. Binder, Wipawee Winuthayanon, Sylvia C. Hewitt Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, NC, USA John F. Couse Taconic Farms, Albany Operations, Rensselaer, NY, USA Kenneth S. Korach Laboratory of Reproductive and Developmental Toxicology, National Institute of Environmental Health Sciences, NC, USA

INTRODUCTION of other steroid hormones, exemplified by progesterone, glucocorticoid, and thyroid hormones. The Jensen and The ovarian-derived sex steroid hormones dictate the Gorski laboratories were then the first to show that the uterine estrous or menstrual cycle in mammals and are tissue retention of estrogen was due to the presence of therefore essential to the establishment and maintenance specific proteins contained within those tissues, which of pregnancy. The uterine response to the preovulatory became referred to as estrogen receptors (ER). Numer- rise in circulating estradiol (E2) is required to prepare ous research groups followed with studies showing that the tissue for the forthcoming rise in progesterone (P) the receptor proteins were hormone specific and their that accompanies ovulation and is critical to embryo presence was linked to some biological activity in cer- implantation. This physiological coordination between tain tissues. Subsequent development some 20 years the ovary and uterus is common to the large majority later of the first antibodies to the ER allowed for more of mammals studied and the spectrum of actions and direct detection and qualification of the protein levels effects of the sex steroids in uterine tissue are mediated in tissues under different physiological states or disease by their cognate NRs. Estrogen was first described almost conditions. Although initially the dramatic physiologi- 100 years ago as a substance that induced estrus. We now cal effects of estrogens on the mammalian uterus were know from many studies over the past 50 plus years that appreciated and well described, later investigations its physiological activity is broader than simply induc- began to elucidate the precise mechanisms and signaling ing a secretory response in the reproductive tract—estro- pathways involved. Our understanding of steroid gen also affects multiple organs that previously were not hormone action in the uterus has been greatly advanced considered estrogen responsive. The mechanistic theory by utilization of new technologies with the generation of for explaining the biological actions of steroid hormones knockout and knock-in models, including receptor or co- came in 1960 when Elwood Jensen and colleagues first activator null, tissue and cell selective null, and mutant described the uptake and retention of labeled E2 in certain knock-in models in mice. These models are also used in tissues that, because of this retention, became referred combination with microarray, RNA-seq, and ChIP-seq to as estrogen target tissues. The ovary and uterus were methods that allow for comprehensive mapping of inter- amongst the tissues showing this property. The hormone action of nuclear receptors (NRs) with chromatin and tissue retention concept was then shown to be reflective modulation of genomic response to steroids in uterine

Knobil and Neill’s Physiology of Reproduction, Fourth Edition 1099 http://dx.doi.org/10.1016/B978-0-12-397175-3.00025-9 © 2015 Elsevier Inc. All rights reserved. 1100 25. STEROID RECEPTORS IN THE UTERUS AND OVARY tissues. Together, these models and techniques have our understanding of the reproductive role of each led to better understanding of the molecular details of receptor signaling pathway in toto. Therefore, we steroid receptor roles in biological processes. include a detailed description of the ovarian and uter- Receptor-mediated actions of progestins, androgens, ine phenotypes for each of these models in their respec- and estrogens are central to reproduction. Perhaps tive sections. unique among the various categories of signaling pathways, the sex steroid hormones and their cognate receptors are surely mentioned, if not discussed in THE STEROID RECEPTORS detail, in every chapter of this volume. The sex steroid hormones are integrated into every aspect of mamma- The steroid receptors are part of the family of lian reproductive physiology in both sexes, including nuclear receptors (NRs) that is ubiquitous throughout sexual maturation and development, gametogenesis, the animal kingdom. The members of the NR fam- hypothalamic–pituitary control of gonadal function, ily fulfill a plethora of functions and are integral to sexual and maternal behavior, pregnancy, and lactation. the development and maintenance of multiple physi- Disruption of the signaling pathways for any one of the ological systems. Continued discovery of new NRs has three gonadal steroids leads to reduced fecundity, if not led to a considerable expansion of the NR family and infertility, due to aberrations in multiple organ systems. demanded a system of categorization. NRs have been Therefore, a thorough understanding of the sex steroid assigned to three classes.1–3 Class I includes the steroid receptors in terms of their expression in reproductive receptors (ER, PR, AR, GR), which interact with high and endocrine tissues, their mechanism of action, their specificity to their steroid ligand as well as to DNA role in reproductive processes, and their interaction with motifs comprised of inverted-repeat hexanucleotides nonsteroidal signaling pathways is instrumental to our separated by a 3-base pair (bp) spacer. The Nuclear ability to manage infertility and reproductive disease. Receptors Nomenclature Committee has developed This chapter principally covers the signaling pathways an NR nomenclature scheme that is based largely on for progestins, androgens, glucocorticoids, and estro- sequence homology and evolutionary comparisons and gens, and the role of each in ovarian and uterine func- divides the over 60 NRs into seven subfamilies and a tion in mammals. Because of the importance of the sex varied number of groups within each subfamily.4 The steroids and the likelihood that each is discussed in ERs (ERα and ERβ) are the sole members of the NR3A several other chapters of this volume, it is appropriate subgroup, ERα and ERβ being NR3A1 and NR3A2, to first discuss the structure, mechanism of action, and respectively.4 The PR and AR are the third and fourth signaling pathways for each of the sex steroid recep- members of the NR3C subgroup that also includes the tors. Sections then follow on the role of sex steroids in GR and the mineralocorticoid receptor.4 The official ovarian and uterine function. We chose to divide each of gene names for these steroid receptors are ESR1 and these sections according to the sex steroid receptors: the ESR2 for ERα and ERβ, respectively, PGR for proges- estrogen receptors (ERs), progesterone receptors (PRs), terone receptor, AR for androgen receptor, and GCR for androgen receptors (ARs), and glucocorticoid receptors glucocorticoid receptor. The general properties of the (GRs). The section Sex Steroid Receptors and Ovarian sex steroid receptor family and those mechanisms of Function discusses what is currently known concerning receptor action that are most relevant to ovarian and each particular sex steroid signaling pathway in terms of reproductive tract function are outlined following. its expression pattern and distribution within the ovary, and its intraovarian role in granulosa cell proliferation Genes, mRNA, and Regulation and differentiation, thecal cell function, and ovulation. Because a separate chapter (Chapter 23) is dedicated to The genes encoding the sex steroid receptors the corpus luteum, the role of the sex steroid receptors in exhibit a highly conserved structural organization. luteal function is not covered in detail. The section Sex The human PR (PGR or NR3C3), AR (AR or NR3C4), Steroid Receptors in Uterine Function discusses what is ERα (ESR1 or NR3A1), ERβ (ESR2 or NR3A2), and GR currently known concerning each sex steroid signaling (GCR or NR3C1) genes are each composed of eight pathway in terms of its expression pattern and distribu- coding exons, and vary in length from 40 kb (ERβ) to tion within the uterus and its role in uterine cell prolifera- >140 kb (ERα).5–8 In each case, the N-terminal domain tion and differentiation. New since the previous editions (NTD) of the receptor is usually encoded by a single of this volume is the development of tissue selective null exon, the two zinc fingers of the DNA binding domain, or “knock-in” mouse models for each of the sex ste- each encoded by separate exons (2 and 3), and the roids. The study of each of these models has already ligand binding domain (LBD), encoded by exons 4–8 and will continue to make enormous contributions to (Figure 25.1).

4. FEMALE REPRODUCTIVE SYSTEM THE STEROID RECEPTORS 1101

demonstrated that the proximal 0.4 kb of the mouse Estrogen Receptor Esr1 promoter is sufficient to provide for widespread The human ESR1 (ERα) cDNA was first cloned in but specific expression of a reporter construct in vivo. 19859–11 and has since been isolated from numerous addi- Regulatory elements that may provide for AP-1, Sp1, tional species.12,13 A second ER gene, termed ESR2 (ERβ), and ER autoregulation of the human ESR1 promoter was discovered in 1996 in rat14 and human15 tissues have been described.23,24 Decreased ESR1 expression and has since been cloned from many species.13 Unlike is linked to receptor-mediated actions of vitamin D25 the A and B forms of PR and AR, ERα and ERβ are not and increased intracellular cAMP or mitogen-activated isoforms but rather distinct receptors encoded by two protein kinase activity.23 Increased methylation of the genes on different chromosomes. The ERα proteins are ESR1 promoter is implicated in reducing ER levels, 595 and 599 amino acids in length in human and mice, especially in tumorigenic tissues.23 respectively, with an approximate molecular weight of The ESR2 genes of multiple species yield numerous 66 kDa (Figure 25.1).8,16–18 Multiple promoter and regu- transcripts that range from 1 to >9 kb in length, in contrast latory regions in the 5′-untranslated sequences of the to the single predominant transcript of ∼7 kb transcribed human and rat ESR1 gene have been described, yet all from the ESR1 gene.18 Initial descriptions of human possess the same single open reading frame.8 Numerous and rodent ERβ projected a protein of 485 amino acids. naturally occurring variants of the ESR1 mRNA in nor- However, it is now apparent that translation of the ESR2 mal and neoplastic tissues of several species have been mRNA initiates upstream of these original open reading described, but the existence of corresponding proteins frames and yields a receptor of 549 amino acids in rodents remains controversial.18 and 530 amino acids in humans, each with an approxi- The promoter region of the ESR1 gene has been mate molecular weight of 60–63 kDa (Figure 25.1).8 relatively well characterized and indicates a complex Therefore, ERβ is slightly smaller than ERα, and most of regulation of expression.8,19–21 Cicatiello et al.22 this difference lies within the N-terminus. A number of

FIGURE 25.1 The steroid family of receptors and endogenous ligands. (A) Schematic illustration of the structural organization of the sex steroid nuclear receptors. The more highly conserved C and E domains are depicted as open boxes and the less well-conserved A/B, D, and F domains as filled bars. The F domain is unique to the estrogen receptors (ER). The functions of the modular domains are indicated as are the two known transcriptional activation function domains, AF-1 and AF-2 (as well as AF-3 found only in PRB). The AF-1 domain is harbored in the A/B region and exhibits constitutive activity in vitro, whereas the AF-2 domain lies with the ligand binding (E) domain and is critical to ligand-induced receptor activation. Both domains synergize during ligand-activated receptor actions. NLS, nuclear localization signal. (B) Comparison of the human steroid receptors, the androgen receptor (AR), estrogen receptors α (ERα) and β (ERβ), progesterone receptor B (PRB) and A (PRA) and the glucocorticoid receptor (GR). The amino acids that compose each domain of the human forms are indicated. For PRA, the numbers refer to the amino acid residues in reference to PRB.

4. FEMALE REPRODUCTIVE SYSTEM 1102 25. STEROID RECEPTORS IN THE UTERUS AND OVARY variant transcripts of the ESR2 gene have been described; full A/B domain and first zinc finger of the C domain, however, unlike ERα, there is growing evidence that some has been described, but its expression as a protein is of these variants may coexist with the wild-type (WT) controversial.29,30 The level of PR expression is modu- receptor form in certain tissues (Figure 25.2). Portions of lated by ER-mediated effects of E2 in certain but not the mouse Esr2 gene have been characterized and pos- all tissues, and this is especially true in the female sess potential binding sites for numerous transcription reproductive tract.31–34 Several naturally existing PGR factors8,26 and the initiation of transcription from at least variant transcripts harboring exon deletions or distinct two distinct untranslated exons.27 5′-untranslated sequences have been described but are not well characterized in terms of protein expression or Progesterone Receptor functionality.27,30,35 Two distinct promoters in the PGR gene provide for the generation of two major PR isoforms, PRB (114 kDa) Androgen Receptor and PRA (94 kDa), in most species except rabbit, which The AR gene is located on the X chromosome, and possess PRB only. The two PR forms are identical in all therefore genetic males possess only a single copy.7,36,37 regions except the NTD, which is truncated by 128–164 Two distinct start sites in the AR gene are used to amino acids in PRA, depending on the species (Figure produce two isoforms, AR-B (110 kDa) and AR-A 25.1). Both PR isoforms are expressed at relatively equal (87 kDa), which differ only in the N-terminus ( Figure levels in tissues, although differences in the ratio have 25.1).12,38 However, in contrast to the PR, the two been noted in certain tissues. A third isoform, termed known AR isoforms exhibit only subtle functional dif- PR-C, an N-terminally truncated form that lacks the ferences. A unique feature of the AR gene relative to

FIGURE 25.2 Human (h) and mouse (m) estrogen receptor variant transcripts. ERs are shown, with the general modular structure of ste- roid hormone receptors illustrating the N-terminal (A/B) domain, DNA binding (C) domain, hinge (D) region, ligand binding (E) domain, and C-terminal (F) domain (see Figure 25.1 for details). Top: hERα full-length and truncated variants (hERα-46 and hERα-36), lacking N-terminal AF-1. Bottom: Full-length hERβ (or ERβ1) with % sequences similarity in comparison to ERα indicated above it. Reported hERβ isoforms are shown beneath. hERβ2 is also called ERβCx, and encodes an alternate C-terminal region, depicted by the striped fill, also seen in hERβ4 and 5. hERβcx is unable to bind E2 but acts as a dominant negative receptor when heterodimerized with ERα in vitro.28 hERβΔexon5 is produced by deletion of exon 5, and produced a truncated protein due to a frame shift leading to translational termination, and has been found in other mammals.18,28 mERβ2 contains a 54-bp insert (INS) that codes for an additional 18 amino acids in the ligand binding (E) domain and has only been found in rodents. mERβ2 exhibits an approximate 35-fold decreased affinity for E2 relative to WT mERβ and a significantly reduced transactivational activ- ity in vitro.18 mERβΔexon5, mERβΔexon6, and mERβΔexon5&6 lack parts of the LBD. Source: Redrawn with permission based on a figure from Ref. 18.

4. FEMALE REPRODUCTIVE SYSTEM THE STEROID RECEPTORS 1103 its sex steroid receptor counterparts is the presence of NTD or A/B Domain polymorphic repeats of glutamine and glycine in the The NTD or A/B domains of the NR family members NTD, the former of which has been linked to certain greatly vary in length and share little homology among chronic diseases and cancer in humans.39 There are the steroid receptors, although some structural features binding sites for a wide array of transcription factors are conserved (Figure 25.1). Crystallography studies in the AR promoter, including Sp1, cAMP-response ele- of the steroid receptor NTD have been largely unsuc- ment binding (CREB) protein, and c-myc, suggesting cessful because this portion of the receptor fluctuates complex tissue-specific regulation. Direct androgen- in aqueous solutions. However, evidence suggests that mediated autoregulation of AR expression has also intramolecular interactions between the A/B and other been shown.39 receptor domains are likely to induce a more structured NTD.1,37,45 Current models of NR signaling incorporate Glucocorticoid Receptor the flexibility of intrinsically disordered (ID) regions of Although not historically recognized for its roles in the receptor, including the NTD, into a mechanism of uterus and ovary, GR is ubiquitously expressed, and allosteric interaction and coordination of ligands, DNA more recent studies have demonstrated important inter- motifs, and NR domain functions.1,45,48 The NTD of each actions between GR and sex steroid receptor mediated of the sex steroid receptors harbors the transcriptional signaling.40,41 GR isoforms, are produced by alternate activation function-1 (AF-1) domain and provides for cell splicing leading to different c-terminal amino acids. and promoter-specific activity of the receptor as well as a GRα is the primary isoform, whereas GRβ is unable to site for interaction with co-receptor proteins (Table 25.1). bind GR ligands and can behave as a dominant negative The AF-1 domain alone can confer constitutive transcrip- inhibitor of GRα activity.42 tional (i.e., ligand-independent) activity when linked to a heterologous NR DNA binding domain in vitro, but this Receptor Structure function is largely overcome in the context of the whole receptor.12 Posttranslational modifications to the A/B Common to all members of the NR family is a mod- domain can dramatically affect the overall behavior of the ular structure of domains, each of which harbors an receptor and are thought to be an important mechanism autonomous function that is critical to total receptor for the modulation of AF-1 functions.17 Phosphoryla- action.1,2,12,43–47 The sex steroid receptors are composed tion of the A/B domain is the most well-characterized of five functional modules, an N-terminal domain (NTD) posttranslational modification and occurs in all three or A/B domain, the DNA binding (C) domain, a hinge sex steroid receptors via the actions of multiple intracel- (D) region, and an LBD (E) (Figure 25.1). The ERs also lular signaling pathways, including mitogen-activated possess a unique C-terminal F domain of unknown func- protein kinase pathways,49 the cAMP/protein kinase A tion. Our understanding of the NR functional domains (PKA) pathway, and cyclin-dependent kinases.12,37,39,43,50 and their importance to overall receptor activity is The extended N-terminal sequences that are unique largely derived from the in vitro study of artificially gen- to PRB (Figure 25.1) provide a third transcriptional AF-3 erated mutant receptors and more recently from X-ray domain in this isoform that is lacking in PRA. This may crystallography studies. allow for PRB-specific expression of certain P regulated

TABLE 25.1 Steroid Receptor Co-Regulator Complexes

Complex Functions Comments References

SRC1, SRC2, SRC3 Interact with Helix12 of agonist-bound NR, 51,52 interact with SWI/SNF, histone modifiers

Mediator “Bridges” NR and transcriptional “machinery” Made up of >20 subunits, MED 1-31, arranged in 3 53,54 (RNA Pol II) to control transcription modules (head, middle, tail)

SWI/SNF Regulate access to enhancer sequences via Made up of 9+ subunits, examples include BRG1, 55 chromatin remodeling, ATPase activity BRM, BAF subunits

Histone Modifiers Modify histones to increase or decrease Acetyltransferase (HAT; e.g., p300/CBP), deacetylase 56,57 transcription (HDAC; e.g., NCoR), Methyl transferase (e.g., PMRT/ CARM), de-methylase

26S Proteasome “Clears” transcriptional modulatory proteins Structure made up of 20S catalytic core particles (CP), 58,59 to facilitate subsequent transcription, 19S regulatory particles (RP) transcriptional termination

4. FEMALE REPRODUCTIVE SYSTEM 1104 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

FIGURE 25.3 The nuclear receptor response elements and primary structure of the DNA binding domain. Top: Shown is the canonical core hormone response element (HRE) for the estrogen receptors (ERE) and glucocorticoid receptors (GRE). These sequences are located in the regulatory regions of sex steroid target genes and provide a site for receptor binding and transactivational activity. A full HRE consists of two core palindromic sequences arranged as an inverted repeat, always separated by three bp (denoted by XXX). Note that a change from GT to AA at positions two and three of the core sequence modifies the ERE to a GRE. Bottom (A) The DNA binding domain (C) of the sex steroid receptors is highly conserved and composed of two zinc fingers and a C-terminal extension (CTE). The ERα DBD is shown. Each zinc finger is composed of four conserved cysteine residues that coordinate to chelate a single zinc ion. Helix 1 of the first zinc finger contains the P-box residues involved in the discrimination of the HRE. Residues in the second zinc finger (helix 2) form the D box that provides a dimerization interface. The CTE is critical

4. FEMALE REPRODUCTIVE SYSTEM THE STEROID RECEPTORS 1105 genes.29,50,60 In general, PRB is a stronger activator of Crystallography studies indicate a highly conserved transcription versus PRA when acting on a hormone structure consisting of dual α-helices positioned per- response element (HRE)–driven gene construct in vitro.60 pendicular to each other1,45,46 (Figure 25.3). Amino acids Furthermore, antagonist-bound PRB acts as a strong in the C-terminal “knuckle” of the first zinc finger form transcriptional activator in certain cell and promoter the “P box” (proximal box) of the DNA binding domain contexts, whereas antagonist-bound PRA is inactive.29 and confer sequence specificity to the receptor; hence, The NTD of the AR is especially autologous in vitro the proximal zinc finger is often referred as forming the and is likely more important to overall AR transactivation “recognition helix” (Figure 25.3).64 Amino acids at the than the C-terminal AF-2 domain (described later).1,37–39 N-terminal “knuckle” of the second zinc finger form the Furthermore, the first 30 residues of the AR NTD are “D-box” (distal box) and are more specifically involved highly conserved and important for interactions with in differentiating the “spacer” sequence within the HRE the LBD that provide for agonist-induced stabilization as well as providing an interface for receptor dimeriza- of the receptor.37 Unique to the AR among the sex steroid tion (Figure 25.3). receptors are a series of amino acid repeat sequences in The consensus HREs that sex steroid receptors bind the NTD, most notably the polymorphic tracts of gluta- are composed of two 6-bp palindromic sequences mine and glycine, of which certain lengths are thought to arranged as an inverted repeat and always separated by be linked to various human diseases or cancer.37,38 a 3-bp spacer (Figure 25.3). Because the P-box residues The greatest structural disparities between ERα are identical among the AR, PR, and GR, these receptors and ERβ lie within the A/B domain, which is approxi- bind a common consensus HRE consisting of two 6-bp mately 30 amino acids shorter in ERβ and exhibits only palindromic half-sites (5′-PuG[G/A]ACA) arranged >20% homology.8,61,62 This divergence likely accounts as an inverted repeat and always separated by a 3-bp for the many functional differences that have been spacer (Figure 25.3). The consensus ERE bears the same revealed from comparative studies of the two ER arrangement but is composed of a unique palindromic forms.8,17,18,62 In general, ERβ tends to be a less effec- half-site (5′-PuGGTCA) (Figure 25.3). The inverted- tive transcriptional activator compared with ERα when repeat arrangement of the palindromic sequences is acting in a classic estrogen response element (ERE)– unique to sex steroid receptor HREs and dictates that the driven mechanism in vitro.18 An ERα/ERβ chimera in receptors homodimerize in a “head-to-head” position which the A/B domain of ERβ has been replaced with when bound to DNA. Evidence indicates that the AR that of ERα is able to activate transcription much better can uniquely dimerize in a “head-to-tail” fashion on an than the native ERβ.61 Furthermore, certain antagonists HRE of the monomer 5′-GGTTCT. More recent structural (e.g., tamoxifen) that exhibit some agonist-like proper- analysis has revealed the importance of the 10-30 amino ties through AF-1 when bound to ERα exhibit no such acid carboxy terminal extension (CTE) of the DBD in activity with ERβ.61,63 DNA interaction.1,45,46,48 Although this region is variable between steroid receptors, it is crucial for DNA binding, DNA Binding or C Domain particularly for sequence selectivity of DNA binding, by The C domain of the NR family members is that extending the interaction surfaces between the receptor portion of the receptor that specifically functions to rec- and the DNA. ognize and bind to the cis-acting enhancer sequences, or HREs, that are located within the regulatory regions of Hinge Region or D Domain target genes (Figure 25.3). It is the most highly conserved Historically, the D domain was thought to primarily (55–80%) region among the NR family members.2,46,64 serve to connect the more highly conserved C and E The C domains of ERα and ERβ are practically identi- domains of the receptor.12 The previously described cal (>95% homology) in most species and are therefore CTE extends into the hinge region, which also harbors expected to exhibit a similar affinity for the same HREs.62 a nuclear localization signal, and influences cellular The functionality of the C domain is provided by a motif compartmentalization of receptors, as well as sites of of two zinc fingers, each composed of four cysteine posttranslation modifications.48 Current mechanisms residues that chelate a single Zn2 ion, and are always suggest this nonconserved and intrinsically disordered encoded by separate exons within the gene (Figure 25.3). (ID) domain is important for intramolecular allosteric

for monomeric DNA binding. Diagram includes amino acids found in ERα. (B) Amino acid sequences in the vicinity of the P box and CTE regions of the steroid receptors are compared. The underlined residues in the P-box sequences are critical for HRE selectivity. Amino acid substitutions made within the P box to create the NERKI and the EAAE DNA binding–deficient ERα mouse lines are also shown. The EAAE mutations also includes an amino acid change in the “tip” of the first finger that isn’t in the P box. (C) Shown is a ribbon diagram of the DNA binding (C) domain illustrating the putative arrangement of helix 1 and helix 2 crossing at right angles to form the core of the DNA binding domain that recognizes a hemi-site of the response element. Source: (A) reproduced with permission from Ref. 43 and (C) reproduced with permission from Ref. 65.

4. FEMALE REPRODUCTIVE SYSTEM 1106 25. STEROID RECEPTORS IN THE UTERUS AND OVARY interactions involving the N-terminal and ligand binding forms also bind the synthetic estrogen diethylstilbestrol domain (LBD)48 (described earlier in the section NTD or (DES) with relatively equal affinity.67 However, given the A/B Domain). This type of flexible structural interaction divergence in homology, it is not surprising that ERα and works to allow rapid response to diverse modulators ERβ exhibit measurable differences in their affinity for governing changes in biological environments.48 other endogenous steroids and xenoestrogens.8,17 Natural and synthetic steroidal and nonsteroidal ER agonists and LBD or E Domain antagonists have been described, some of which show The LBD or E domain of the steroid receptors is a specificity for one or the other ER subtype (Table 25.2), highly structured multifunctional region that primar- illustrating differences between the LBDs of ERα and ily serves to specifically bind ligand and provide for ERβ, and provide for powerful pharmacological tools to hormone-dependent transcriptional activity.12 An AF-2 discern the overall function of each ER (Table 25.2). The domain located in the C-terminus of the E domain most widely used ER subtype selective ligands currently mediates this latter function. The AF-2 domain is subject in use are propylpyrazole (PPT), an ERα selective ago- to posttranslational modifications12 and is an especially nist, and diarylpropionitrate (DPN), an agonist showing strong activator of transcription in the ER and PR but is preference, but not exclusive selectivity, towards ERβ.8 markedly weaker in the AR, where it is more involved in interactions with residues in the NTD.37,38 Also harbored F Domain within the E domain is a strong receptor dimerization Among the sex steroid receptors, only ERs possess interface, sites for interaction with heat shock proteins, a well-defined F domain (Figure 25.1). This region is and nuclear localization signals.12,48 Although there is relatively unstructured and harbors little known func- minimal homology in the primary sequence of the LBD tion, although some data indicate a role in co-activator for the sex steroid receptors, comparative studies of the recruitment, dimerization, and receptor stability.61,62,68–70 crystal structures of liganded and unliganded LBDs indi- cate a highly conserved structural arrangement. These Co-Regulatory Complexes structural studies indicate that the LBD is composed of All steroid receptors interact with co-regulatory mol- 11 α-helices (H1, and H3–H12) arranged in a three-layer ecules, co-activators, and co-repressors.51,71 The primary α-helical sandwich to create a hydrophobic ligand bind- co-activator interaction for steroid receptors (SR) is with ing pocket near the C-terminus of the receptor.48 The p160/SRC (steroid receptor co-activator) 1, 2, and 3.52,72,73 resulting shape and volume of the ligand binding pocket SRC1 (NCOA1), SRC2 (GRIP1, TIF2), and SRC3 (pCIP, is larger than necessary to accommodate the correspond- RAC3, ACTR, TRAM, A1B1) interact with helix 12 of SRs ing ligand, suggesting that key interactions between the via “LXXLL” motifs in their nuclear receptor interacting ligand and specific receptor residues are more critical to domain, which are leucine rich regions with “X” desig- conferring ligand specificity.61 Receptor binding to an nating any amino acid.52 SRCs also contain activation agonist ligand leads to rearrangement of the LBD such domains that recruit secondary molecules such as p300, that H11 is repositioned and H12 swings back toward the and a bHLH-PAS motif within the N-terminal region, core of domain to form a “lid” over the binding pocket. which can interact with other transcription factors.52 The This agonist-induced repositioning of H12 leads to the elegant complexity of co-activator composition and func- formation of a hydrophobic cleft, or “NR box”, by helices tion has been increasingly revealed. Steroid receptors 3, 4, and 5 on the receptor surface, constituting the AF-2, and SRCs function as a nexus interacting with massive which serves to recruit co-activators (Table 25.1) to the multimeric complexes, such as the SWI/SNF chroma- receptor complex. In contrast, receptor antagonists are tin remodeler, mediator complex, or proteasomes (Table unable to induce a similar repositioning of H12, leading to 25.1).73 These interactions facilitate coordination of the a receptor formation that is incompatible with co-activator specific functions necessary to allow appropriate gene recruitment and is therefore less likely to activate and cell selective access to chromatin, via modifications transcription. The LBDs of PRB and PRA are identical and of histones or members of co-regulatory complexes.74 In provide for high affinity binding to P.12,50 The LBD of the this way, co-activators dynamically mediate and coor- AR in humans, rats, and mice is identical and provides dinate processes necessary to accomplish transcription, for high affinity binding of two endogenous androgens, including initiation, elongation, termination, and clear- T and 5α-hydroxyT (DHT), the latter of which binds ing or turnover of the transcriptional modulators. with much greater affinity.38,66 Numerous high affinity synthetic ligands for PR and AR have been developed, Receptor Mechanisms of Action including the agonists R5020 and R1181, respectively (Table 25.2).12 The LBDs of ERα and ERβ exhibit less than Our understanding of the mechanisms by which 60% homology but bind the endogenous ligand, E2, with steroid hormones and their cognate intracellular recep- similar affinity (ERα, 0.1 nM; ERβ, 0.4 nM).8,12,17 Both ER tors influence cell function and behavior has expanded

4. FEMALE REPRODUCTIVE SYSTEM THE STEROID RECEPTORS 1107

TABLE 25.2 Structures of Endogenous and Synthetic Agonists and Antagonists of Steroid Receptors

profoundly since 1988 when Clark and Markaverich their promoter (Figure 25.4); plasma membrane ste- reviewed the field for the inaugural edition of this vol- roid signaling, often referred to as “nongenomic” ste- ume. Much of that earlier chapter remains contempo- roid actions; and ligand-independent “cross talk” with rary in reference to general receptor biochemistry and intracellular and second messenger systems that pro- ligand-dependent activation, which is now referred vide for sex steroid receptor activation in the absence of to as the “classical” or ligand-dependent direct-DNA the cognate steroid ligand (Figure 25.4). The existence binding model of receptor function (Figure 25.4). In of multiple receptor-mediated signaling pathways the years since, numerous discoveries have been made likely accounts for the refined control and plasticity that illuminate the complexity of sex steroid receptor of tissue responses to sex steroids. These modes of sex signaling in cells and tissues, such as the discovery of steroid receptor action as currently understood are dis- additional receptor forms and variants (e.g., ERβ) and cussed following. high-resolution crystallography of receptor domains. The entrée into the “omics” era has facilitated massive Ligand-Dependent Actions: Direct or Classical expansion of study of transcriptional regulation and The classic model of steroid receptor action states chromatin remodeling. In addition, several alterna- that the receptor resides in the nucleus or cytoplasm tive receptor signaling mechanisms that diverge from but is sequestered in a multiprotein inhibitory com- the classic model have become apparent, including plex in the absence of hormone (Figure 25.4). The lipo- “tethering” of the sex steroid receptor to heterologous philic steroid ligands are able to freely diffuse across DNA-bound transcription factors to provide for steroid the plasma and nuclear membranes and bind their regulation of genes that lack HRE sequences within cognate receptor. The binding of ligand results in a

4. FEMALE REPRODUCTIVE SYSTEM 1108 25. STEROID RECEPTORS IN THE UTERUS AND OVARY conformational change in the receptor, transforming it In general, PRB can better or uniquely activate to an “activated” state that is now available for homodi- transcription versus PRA when acting on progesterone merization, increased phosphorylation, and binding to response element–regulated genes in various cell an HRE within target gene promoters, and interaction types in vitro due to the presence of a third AF-3 in with chromatin and transcriptional mediators. NRs the NTD of PRB.50,60 Furthermore, PRA can repress seem to be preferentially recruited to open regions of the transcriptional activities of PRB in certain cell and chromatin.75 Studies using MCF7 breast cancer cells promoter contexts,50 suggesting that PRA may modulate indicate that FoxA1 acts as a pioneering factor, pro- P responsiveness in certain tissues. Interestingly, PRA viding accessible regions in the chromatin that recruit can also repress the actions of other heterologous NRs, ERα.76–79 The ligand-HRE-bound receptor complex including the AR, ER, and GR.12,50,60 The variant, PR-C, then engages co-activator molecules as described is primarily cytosolic and lacks the A/B domain and the before,52 leading to modulation of transcription rates 5′ portion of the C domain, and has been shown to bind P of responding genes. This classic steroid receptor and enhance the actions of PRB and PRA,80 but the pres- mechanism is dependent on the functions of both AF-1 ence of encoded protein and a potential biological role and AF-2 domains of the receptor, which synergize via remains controversial.30,81 Some evidence for a role in P the recruitment of co-activator proteins, most notably signal dampening during labor has been described for the p160 family members.52 Depending on the cell PR-C.82 When acting via a classic ERE-driven mechanism and target gene promoter context, the DNA-bound in vitro, ERα homodimers and ERα/ERβ heterodimers receptor complex may positively or negatively affect tend to be stronger activators of transcription com- expression of the downstream target gene. Initially, pared with ERβ homodimers.18,61 Corroborating in vivo study of NR-mediated gene regulation was carried evidence of differential regulation and heterodimer out on a gene-by-gene basis using a handful of known formation by the ER subtypes has been difficult to gen- hormone regulated transcripts, such as the ER target erate. However, a microarray study by Lindberg et al.83 genes PS2, lactotransferrin, and PR or the AR target using bone tissue from ovariectomized mice found that gene prostate specific antigen. Now, after numerous ERβ generally inhibits ERα-mediated gene expression, comprehensive analyses of hormonally regulated but in the absence of ERα, ERβ can partially provide transcriptional profiles, using microarray and more some E2-stimulated gene expression. No evidence for recently RNA-seq, thousands of NR targets have been ERα or ERβ preferential regulation of specific transcripts found in various cell lines and tissues. or promoters has been forthcoming; rather, it seems the

FIGURE 25.4 Ligand-dependent and ligand-independent nuclear receptor mechanisms. The direct “classic” model of sex steroid receptor (SR) action involves direct interaction between SR bound to hormone (triangles) and HRE; the tethered pathway utilizes indirect “tethering” of SR to genes via interactions with other transcription factors (TF). “Nongenomic” signaling is initiated by membrane-localized receptors modulating extranuclear second messenger (SM) signaling pathways. Ligand-independent responses occur as a result of transduction of membrane receptor signaling, such as growth factors, to nuclear SR. Source: Adapted with permission from Ref. 18.

4. FEMALE REPRODUCTIVE SYSTEM THE STEROID RECEPTORS 1109 important factor is which receptors are expressed in a An obstacle to our better understanding of responding tissue. membrane-associated effects of steroids is the lack of data on the nature of the cell-surface steroid “receptors”. Indirect/Tethered Actions (HRE Independent) Questions remain concerning whether the membrane- Ligand-activated steroid receptors can stimulate associated receptors are identical or variant forms of the the expression of genes that lack a conspicuous HRE sex steroid NR or instead distinct receptors altogether. P is within their promoter. This has been especially demon- known to rapidly induce maturation in Xenopus oocytes strated for ERα84–86 but is also known to occur for PR.87 that are arrested in the G2 phase of meiosis I,90 without This mechanism of HRE-independent steroid receptor requiring protein synthesis. Most evidence indicates this activation is postulated to involve a “tethering” of the process begins at plasma membrane progesterone bind- ligand-activated receptor to transcription factors that ing sites that are unique from the classic nuclear forms of are directly bound to DNA via their respective response PR, called progesterone receptor membrane component elements (Figure 25.4), This mechanism involves a 1 (PGRMC1),91 although cDNA homologs of the PR mediator component (e.g., p160) between ER and AP-1 were cloned from Xenopus and found to be associated versus direct interaction.85 Interestingly, ERβ is unable with P-induced oocyte maturation.89 In other studies, P to enhance the actions of a DNA-bound AP-1 complex is able to inhibit apoptosis of rat granulosa cells in vitro when bound to E2 but can do so in the presence of via a membrane-bound receptor that is immunoreactive certain selective estrogen receptor modulators (SERMs) to antisera directed against the LBD of the classic PR.89 such as tamoxifen, raloxifene, and ICI 164,384.18 A simi- Rapid effects of E2 have been described in a vast lar HRE-independent mechanism of sex steroid receptor array of tissues, including a rapid activation of endo- regulation has been documented for genes that possess a thelial nitric oxide synthase in endothelial cells, poten- GC-rich region or Sp1 binding site within the promoter, tiation of currents induced by the ion channel agonist upon which the actions of a bound Sp1 complex can be kainite in hippocampal neurons, and influx of calcium enhanced by ERα.84 These mechanisms have primarily in uterine endometrial cells.88,90 The evidence of an asso- been demonstrated using in vitro cell culture models. ciation between ERα and plasma membrane–bound components has increased, supporting the view that Nongenomic Actions this mechanism may indeed account for certain rapid All models of sex steroid receptor action described thus effects of E2.92,93 In this sense, the various nongenomic far influence cellular phenotypes by acting in the nucleus actions of E2 are often categorized according to their sus- and modulating target gene expression. The required ceptibility to inhibition by classical ER antagonists (e.g., gene transcription and mRNA translation is a relatively ICI type compounds).92 E2-induced activation of mem- slow process; microarray studies demonstrate that initial brane ion channels, endothelial nitric oxide synthase, transcriptional responses occur beginning about 60 min and mitogen-activated protein kinase kinase are all after administering hormones. Therefore, the numer- inhibited by ER antagonists and therefore likely involve ous examples described to date in which steroids affect membrane-associated ERα or ERβ.92 In contrast, exam- cellular dynamics within seconds to minutes of exposure ples of E2-induced PKA and protein kinase C (PKC) have remained difficult to explain within the context activation are not inhibited by ICI compounds and are of the previous models. Rapid cellular changes such as therefore not likely to involve the nuclear ER form.92 A increased intracellular calcium or cAMP or activation recent study utilized a mutated form of the mouse ERα of kinase signaling cascades have been attributed to engineered to remain sequestered outside the nucleus sex steroid exposure of varied cell types and tissues.88,89 (ERαH2NES), regardless of estrogen ligand. This form Because these steroid effects do not involve direct steroid of ER did not mediate transcriptional responses but receptor activation of gene transcription, they are often maintained estrogen-induced MAPK phosphorylation.94 collectively referred to as representing “nongenomic” Targeting steroid receptors to the membrane involves pathways of steroid action. Indeed, several of the docu- palmitoylation, which is facilitated by HSP27.88 The mented effects lack a conspicuous nuclear component palmitoylation promotes interaction with caveolin-1, or have even been demonstrated to occur in enucleated which then results in localization of the receptor in cells, yet others can ultimately affect gene expression. membrane caveolin rafts. Another potential mediator Therefore, rather than nongenomic, some have proposed of rapid membrane localized hormone response is the these types of steroid actions to be loosely categorized as G protein-coupled receptor GPER (originally referred plasma membrane–associated steroid signaling events. to as GPR30), which is activated by E2.95 GPER-null Although there are numerous examples of putative mice lack reproductive phenotypes,96 although effects membrane-associated steroid signaling effects in repro- on the degrees of uterine responses elicited by E2 have ductive tissues and processes, a thorough description of been observed with G15, a GPER selective antagonist, these mechanisms is beyond the scope of this chapter. suggesting a potential role for GPER in modulating

4. FEMALE REPRODUCTIVE SYSTEM 1110 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

ERα-mediated outcomes.97 A multiprotein complex of ligands possess “mixed” agonist/antagonist activity extracellular signal–related kinases, the inner plasma that is dependent on the receptor, cell, and promoter membrane component of tyrosine kinase receptors context, leading to the more descriptive terms of SERMs, (c-Src), and the nuclear sex steroid receptors92,98–101 has selective PR modulator, and selective AR modulator. also been described as a mediator of rapid response to E. As described in the subsection Receptor Mechanisms A different mechanism speculates that androgens or E2 of Action of the section Steroid Receptors, sex steroid may act as extracellular signaling hormones via binding receptor agonists interact with the LBD to generate an to sex hormone binding globulin that is already bound ordered array of 11 α-helices that is most conducive to to the cell membrane via a G protein–coupled-like sex receptor interaction with co-activator proteins and/or hormone binding globulin cell-surface receptor.102 promotes disassociation of co-repressor proteins. In turn, antagonist ligands generally fail to generate the Ligand-Independent Actions: Membrane Receptor necessary changes in receptor structure that induce full Cross Talk transcriptional activity.48,112–114 We now have ample evidence that the sex steroid Antiprogestins have a number of applications in repro- receptors can be activated via intracellular second ductive medicine, including use as contraceptives due messenger and signaling pathways, allowing for the to their ability to prevent implantation or ovulation.115 induction of sex steroid target genes in the absence of The currently available PR antagonists are all competi- steroid ligand, or the enhancement of steroid hormone tive inhibitors of P binding to the PR LBD.115 Type I anti- signaling (Figure 25.4).29,103–105 Polypeptide growth fac- progestins, such as ZK 98299 (onapristone), bind PR but tors are able to activate ERα-mediated gene expression fail to induce receptor phosphorylation and induce weak via mitogen-activated protein kinase activation of ERα receptor binding to HREs.115,116 In contrast, type II anti- in the absence of E2 (Figure 25.4). Similarly, interleu- progestins, such as RU 486 (mifepristone, Table 25.2), bind kin-6 stimulation of cells leads to increased AR-medi- PR and promote receptor phosphorylation, homodimer- ated gene expression,38,106 and cyclin-dependent kinases ization, and binding of the receptor complex to an HRE, have been shown to activate PR-regulated gene expres- but are unable to activate transcription.50 Crystallogra- sion.29 The intracellular signaling molecule cAMP, a phy data indicate that RU 486 binding to PR induces an common second messenger of G protein–coupled recep- arrangement of helix 12 that is not conducive to binding tors and an activator of the PKA pathway, can stimu- of steroid receptor co-activators and may even promote late increased PR-, AR-, and ER-mediated transcription recruitment of co-repressors such as NCoR.50 Type II anti- of target genes in the absence of steroid ligand.29,103 A progestins do possess some PR agonist activity when act- well-characterized example of ligand-independent ing in the context of the cAMP/PKA signaling pathway in cAMP activation of PR occurs after dopamine stimula- certain cell types.50 tion of dopaminergic neurons107 and is the mechanism The field of antiestrogens and SERMS has been an by which sexual behavior is induced in female rodents. area of intense research because the potential clinical Likewise, growth factors are able to mimic the effects of applications include treatments for fertility, cardiovas- E2 in the rodent uterus via E2 independent activation cular disease, osteoporosis, cognitive function, post- of ERα.109,110 Most interesting are recent studies indicat- menopausal symptoms, and breast and gynecological ing that the MAP kinase protein ERK is co-recruited to cancers.18 The best-characterized antiestrogens are com- chromatin with ERα.111 Ligand-independent activation petitive inhibitors of E2 binding to ER and have been of sex steroid receptors is believed to rely largely on cel- classified into two major groups. The type I class of ER lular kinase pathways that alter the phosphorylation antagonists includes the triphenylethlylene compounds, state of the receptor and/or its associated proteins (e.g., tamoxifen (Table 25.2), hydroxy-tamoxifen, and raloxi- co-activators, heat shock proteins) (Figure 25.4). As in the fene and are characterized by mixed agonist/antagonist classic ligand-dependent mechanism described earlier, activity depending on the receptor, cell, and promoter specific receptor domains are critical to ligand-indepen- context.18 Upon binding the ER, these compounds selec- dent activation as well. ER activation by peptide growth tively inhibit AF-2 function but leave AF-1–mediated factor–signaling pathways appears to be more depen- receptor functions intact, thus explaining the selective dent on AF-1 functions, whereas the effects of increased agonist activity. In contrast, the type II compounds intracellular cAMP are postulated to depend on the AF-2 are considered pure antagonists, the most well char- domain and do not require a functional AF-1.103 acterized being ICI 182,780 (Table 25.2), which is a 7α-substituted derivative of E2. This compound binds Receptor Antagonists the ER but possesses extended side chains that prevent Structural analyses of the sex steroid receptors co-activator association with the NR box of helix 12 and provide insight into the agonist/antagonist actions of may even promote interaction with co-repressors as endogenous or synthetic ligands (Table 25.2). Certain well as increased receptor degradation,18 resulting in

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1111 decreased ER protein. Compounds of this type have now ER Expression in the Uterus been designated as selective estrogen receptor degraders RODENTS (SERDs). ER-selective nonsteroidal ligands exploit dif- ERα is present in the female reproductive tract of ferences between the LBDs of ERα and ERβ and provide rodents throughout late fetal and neonatal development, powerful pharmacological tools to discern the overall puberty, and adulthood.120,121 In mice, ERα transcript function of each ER (Table 25.2). has been detected in the blastocyst, but immunoreac- Clinical use of antiandrogens has focused largely on the tivity first appears in the mesenchymal cells of the fetal treatment of prostate cancer.38,117 The best-characterized uterus as early as gestational day 15,120,121 whereas epi- AR antagonists include the steroidal compound cyprot- thelial expression is delayed until the late fetal period erone acetate (Table 25.2) and nonsteroidal compounds but increases substantially on postnatal day 4 and hydroxyflutamide (Table 25.2), casodex, and nilu- peaks on day 16.122,123 The level of ERα transcripts in tamide. Hydroxyflutamide is considered a pure AR the murine uterus during neonatal development closely antagonist that competitively binds AR and reduces AR mirrors the patterns indicated by earlier immunohisto- transcriptional activity by preventing interaction with chemical studies.124 There is reportedly no ERβ immu- co-activators.117,118 noreactivity in the neonatal mouse uterus, although transcripts are detected at a modest level.124 In adult mice, ERα transcripts and immunoreactivity are quite SEX STEROID RECEPTORS IN UTERINE high in uterine epithelial cells before ovulation125 but FUNCTION decrease during pregnancy, whereas expression in the stroma continues.126 In contrast, ERβ expression is rela- Estrogen Receptor Signaling in Uterine Function tively low in uteri of virgin and pregnant mice.124,127,128 Adult rats exhibit a similar uterine ER expression pat- The ovarian-derived sex steroid hormones dictate tern such that ERα transcripts and immunoreactivity are the uterine estrous or menstrual cycle in mammals easily detectable in the epithelium, stoma, and myome- and are therefore essential to the establishment and trium, and levels peak during the proestrous period of maintenance of pregnancy. The uterine response to the increased circulating E2 levels.129 The evidence of ERβ preovulatory rise in circulating E2 is required to pre- expression in the rat uterus is conflicting; whereas ERβ pare the tissue for the forthcoming rise in P that accom- expression has been reported to be in approximately panies ovulation and is critical to embryo implantation. 40% of rat uterine epithelial and stromal cells regardless This physiological coordination between the ovary and of the estrous cycle stage,129 others report a total lack of uterus is common to the large majority of mammals detectable ERβ.130 These observations are plagued by the studied, and the spectrum of actions and effects of the inconsistent quality of antibodies against ERβ. In ovari- sex steroids in uterine tissue are mediated by their cog- ectomized mice, ERα and ERβ proteins were detected in nate NRs. Estrogen was first described almost 100 years the epithelial, stromal, and myometrial cells,131–133 indi- ago as a substance that induced estrus.119 Although cating that removal of ovarian hormones alters the ER initially the dramatic physiological effects of estrogens expression pattern. Although ERβ protein is present in on the mammalian uterus have been appreciated and the mouse uterus, there is overwhelming functional evi- well described, later investigations begun to eluci- dence that ERα is responsible for mediating the many date the precise mechanisms and signaling pathways effects of estrogens in the uterus. As described follow- involved in estrogen responses. Our understanding of ing, only ERα-null female mice exhibit a severely attenu- steroid hormone action in the uterus has been greatly ated response to acute and chronic estrogen (e.g., E2 or advanced by utilization of new technologies with the DES) treatment in terms of uterine growth and altered generation of knockout and knock-in models, includ- gene expression, as well as exhibit impaired embryo ing receptor or co-activator null, tissue and cell selec- implantation.110,134,135 Furthermore, categorical studies tive null, and mutant knock-in models in mice. These have shown that the toxic effects of neonatal estrogen models are also used in combination with microarray, exposure in the neonatal female mouse uterus are clearly RNA-seq, and ChIP-seq methods that allow for com- mediated by ERα.136,137 prehensive mapping of interaction of NRs with chroma- tin and modulation of genomic response to steroids in uterine tissues. Together, these models and techniques DOMESTIC ANIMALS have led to better understanding of the molecular In the neonatal ovine uterus, ERα mRNA and protein details of steroid receptor roles in biological processes. levels are highest in the developing glandular structures; The following sections discuss what is currently known detectable but lower in the luminal epithelia, stroma, and hypothesized about sex steroid receptor actions in and vascular endothelial cells; and absent in the myome- the mammalian uterus. trium.138,139 In the adult ewe, ERα expression is detected

4. FEMALE REPRODUCTIVE SYSTEM 1112 25. STEROID RECEPTORS IN THE UTERUS AND OVARY in all uterine cell types and is highest on day 1 of the subendometrial myometrium during the proliferative 16- to 18-day cycle, concurrent with peak plasma E2 phase and are postulated to be involved in uterine peri- levels.138,140 During days 1–6, when P levels rise and E2 staltic activity.149 declines, ERα expression decreases accordingly and does ERβ expression in the human uterus is much lower not begin to rise again until the end of the cycle nears relative to ERα but is generally present in the same cell on days 11–15.138,140 A similar pattern of uterine ERα types and exhibits comparative changes during the expression is reported during the 21-day bovine cycle.141 menstrual cycle.149,150 There is one report of increased In situ hybridization and immunohistochemical analy- ERβ immunoreactivity in the uterine stroma during ses indicate maximum ERα levels on days 1–3, especially the secretory phase.150 ERβ immunoreactivity is espe- in the glandular and stromal cells; these levels generally cially detectable in vascular smooth muscle cells and decrease by day 6, although cells comprising the deep increases therein during the secretory phase, suggesting glands maintain ERα expression.142 The observed cycli- a role for this receptor form in vascularization.150 The cal changes in ERα expression in ovine and bovine uteri human-specific ERβ isoform, ERβcx, also called ERβ2 are consistent with steroid regulation of receptor lev- (Figure 25.2), is also detected in the uterus and is present els. Indeed, E2 treatment of ovariectomized cows leads along with full-length ERβ (ERβ1) in the functional and to increased ERα expression in the glands, stroma, and basal layer endometrial glands.151 Interestingly, ERβcx luminal epithelia, whereas P alone or in combination immunoreactivity decreases in the basal layer glandu- with E2 elicits a decrease in ERα levels.142 lar cells during the mid-secretory phase, whereas little change occurs in the level of full-length ERβ,151 suggest- PRIMATES ing that differential expression of ERβ isoforms over the Early investigations in human uterine tissue found course of the menstrual cycle may modify the cellular that levels of ER immunoreactivity or high-affinity E2 responses to E2. binding vary during the menstrual cycle such that peak Nonhuman primates exhibit uterine ER expression expression occurs in the mid- to late proliferative phase patterns similar to those described in humans. ERα and then decreases at ovulation and commencement of immunoreactivity in baboon uteri is highest in endome- the secretory phase.143–145 This peak in uterine ER levels trial glandular, stroma, and myometrial smooth muscle at the time of rising plasma E2 levels is similar to several cells during the proliferative phase and decreases in all other species and suggests that expression is autoregu- cell types upon commencement of the secretory phase.152 lated. However, maintenance of uterine ER expression A similar ERα expression pattern in epithelial, stromal, in postmenopausal women suggests that nonsteroidal and myometrial cells is reported in the cynomolgus mon- regulatory factors are also involved.145 The ovulatory key uterus.153 ERα immunoreactivity is also reported decline in ERα levels among all uterine cell types is con- in the vasculature, including the spinal arteries, of the current with the rise in circulating P levels and suggests baboon uterus.152 Ovariectomy followed by exogenous that PR-mediated P actions may act to downregulate ER E2 or E2 plus P treatments in baboons leads to changes expression. Such PR-mediated decrease in ERα has been in uterine ER expression that mirror the menstrual cycle, reported in experimental systems as well.146 A distinct supporting the direct role of ovarian steroids in the regu- gradient of ER levels among the functional components lation of uterine ER levels.152 Considerable levels of ERβ of the human uterus was also reported such that levels transcripts are also detected in the epithelial, stromal, are highest in the fundus and progressively decrease in and myometrial cells of cynomolgus monkey uteri.154 those tissues closer to the cervix.147,148 In contrast, ERβ immunoreactivity is reportedly low in The discovery of ERβ forced a reevaluation of uter- normal uterine tissue from baboons, although higher ine ER expression because the early studies discussed levels are observed in diseased glands associated with previously did not use methods that adequately distin- endometriosis.155 guished between the two ER forms. More recent studies definitively show that ERα is the predominant form in Uterine Phenotypes in Mouse Models of Disrupted the human uterus and accounts for the increased recep- Estrogen Signaling tor levels that occur during the proliferative phase.149 The murine models of disrupted estrogen signaling ERα transcripts and protein are especially high in glan- have proven invaluable to experimental investigation dular epithelial and stromal cells during the prolif- of estrogen actions in the uterus and the contribution of erative phase and exhibit a dramatic decrease among each ER form to these functions (Table 25.3). In addition uterine tissues upon entering the secretory phase of the to the ER-null models are two independently derived cycle.149 The levels of ERα transcripts and immunoreac- lines of mice that lack the capacity to synthesize E2 due tivity in the uterine myometrium are much lower and to disruption of the Cyp19 gene.131,156 Following, we will exhibit little change during the menstrual cycle.149 Still, describe how these different mouse models have helped comparatively higher ERα levels are described in the to delineate the role of ER in the uterus.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1113

TABLE 25.3 Uterine Phenotypes in Mice Null or Mutated for Sex Steroid Receptors and Signaling

Mutated or Null for Sex Steroid Receptors and Signaling Uterine Phenotypes References

Esr1−/− (homozygous null alleles for ERα: αERKO Normal uterine development but exhibits hypoplastic uteri 135,157–161 and Ex3αERKO) Insensitive to the proliferative and differentiating effects of endogenous growth factors and exogenous E2 Implantation defect *Exhibit decidualization Infertile

NERKI+/− (one mutated allele of two-point mutation Normal uterine development but exhibits hyperplastic uteri 132 in ERα DBD and one WT allele) Hypersensitive to estrogen Infertile

KIKO (ERAA/−) (one mutated allele of two-point Normal uterine development 162,163 mutation in DNA binding domain of ERα and Insensitive to the proliferative effects of exogenous E2 treatment one ERαKO allele) Infertile

ERαEAAE/EAAE (homozygous animal of 4-point Normal uterine development but exhibits hypoplastic uteri 164 mutation of DBD ERα) Loss of E2-induced uterine transcripts Infertile

ERαAF-10 (deletion of amino acids 2–128 on ERα) Normal uterine development and architecture 165,166 Blunted E2 response Infertile

ERαAF-20 (deletion of amino acids 543–549 on ERα) Normal uterine development but exhibits hypoplastic uteri 167 Insensitive to E2 treatment Infertile

ENERKI (ERαG525L) (homozygous animal of 1-point Normal uterine development but exhibits hypoplastic uteri 168 mutation in LBD of ERα) Insensitive to E2 treatment IGF1 induced slight uterine epithelial proliferation compared to control littermates (nonhomogenous pattern) Infertile

AF2ERKI/KI (homozygous knock-in of 2-point Normal uterine development but exhibits hypoplastic uteri 169 mutation in LBD of ERα) Insensitive to E2 treatment ER antagonists and partial agonist (ICI 182,780 and TAM) induced uterine epithelial proliferation Growth factor did not induce the uterine epithelial cell proliferation Infertile

Wnt7aCre+;Esr1f/f (uterine epithelial cell–specific Normal uterine development 170 deletion of ERα) Sensitive to E2- and growth factor–induced epithelial cell proliferation Selective loss of E2-target gene response Implantation defect Infertile

Esr2−/− (homozygous null alleles for ERβ: βERKO, Exhibit grossly normal uterine development and function 158,171–173 β β L−/L− Ex3 ERKO, and **ER ST ) Sensitive to E2 treatment Some Esr2−/− lines reported elevated uterine epithelial proliferation after E treatment compared to WT Some are completely sterile (due to ovarian phenotype)

αβERKO (homozygous null for both ERα and ERβ) Normal uterine development but exhibit hypoplastic uteri, similar to 158,174 those of Esr1−/− Insensitive to E2 Infertile

Cyp19a1−/− (homozygous null aromatase: ArKO) Normal uterine development but exhibits hypoplastic uteri 131,156,175 Sensitive to E2-induced epithelial cell proliferation Infertile

Pgr−/− (homozygous null alleles for PRA and PRB: Exhibit grossly normal uterine development 176–178 PRKO) Implantation and decidualization defects Exhibit epithelial hyperplasia in response to E2 Loss of P-induced proliferative switch in uterine epithelial and stromal cells Infertile

(Continued) 1114 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

TABLE 25.3 Uterine Phenotypes in Mice Null or Mutated for Sex Steroid Receptors and Signaling—cont’d

Mutated or Null for Sex Steroid Receptors and Signaling Uterine Phenotypes References

Wnt7aCre+;Pgrf/−(uterine epithelial cell–specific Normal uterine development 179 deletion of PR) Loss of P-induced uterine epithelial P target gene transcription Abnormal proliferative switch of uterine epithelial and stromal cells Implantation and decidualization defect Infertile

PRKO and PRd/d (homozygous null alleles for PR) Normal uterine development 176 Abnormal epithelial proliferation in response to P Implantation and decidualization defect Infertile

PRA−/− (homozygous null alleles for PRA: PRAKO) Normal uterine development 180 Resemble PRKO phenotypes Infertile

PRB−/− (homozygous null alleles for PRB: PRBKO) Normal uterine development 181,182 Normal implantation and decidualization responses Do not exhibit epithelial hyperplasia in response to E2 Normal fertility

AR−/− (homozygous null alleles for AR: ARKO) Normal uterine development, function, and maintenance of pregnancy 183,184 Slightly longer uterine horns compared to controls Slightly reduced uterine circumference Reduced uterine hypertrophic response to gonadotropins Increase placenta size (placentamegaly) Subfertile

AR−/− in-frame deletion of exon 3, loss of DNA No overt uterine phenotype 185 binding activity Subfertile

SPARKI (homozygous knock-in of mutated Normal female fertility, no uterine phenotype reported 186 DBD of AR)

* Ex3αERKO females have a similar uterine phenotype to the original αERKO except for exhibiting a decidualization defect, which may be due to the splice variants in the original αERKO that retains ER activities. β L−/L− β ** ER ST females are the only line of ER knockout animals that are reported to be completely sterile.

MICE LACKING ERα and ERα−/−) exon 3158,160 of the murine Esr1 (ERα) gene. α Unlike the AR, which is mutated in the naturally The uterine estrogenic response in ERKO females dif- α occurring Tfm rat and mouse, no known mutations of fers from the latter two lines, as ERKO animals have α the ERα have been reported, thus the deletion of ERα was a low level of truncated ER protein produced from a thought to be lethal. Similarly, only one male patient and splice variant, which preserves some residual biological 189 α one female patient with ERα mutation have been discov- functions, but all ER -null female mice are infertile. α ered.187,188 The male patient’s mutation results in severe Recently, an ER -null rat has been derived using zinc truncation of the ERα protein due to a stop codon in the finger nuclease (ZFN) genome editing. All phenotypes α A/B domain. The female patient has a point mutation examined thus far were previously seen in the ER -null in her ERα LBD that decreases activity by reducing E mice, including infertility due to hypoplastic uteri, poly- binding more than 200-fold. ERα-null mice were the first cystic ovaries, and ovulation defects.190 Interestingly, experimentally generated steroid-receptor-null animal, the recently described female patient with homozygous α reported in 1993, and preceded the discovery of ERβ in ER mutation also has cystic ovaries and a small uterus 1996. There are currently numerous reported lines of ERα- despite elevated circulating E.188 null mice and additional lines of mice with mutations in β functional domains of ERα. Three separate lines of ERα- MICE LACKING ER null mice were generated: the αERKO, first described The first description of ERβ-null mice followed the by Lubahn et al. in 1993,157 the ERαKO (or Ex3αERKO), discovery of ERβ by only two years. Prior to generation described by Dupont et al. in 2000158 and by Hewitt in of the ERβ-null model, very little was known as to the 2010,159 and ERα−/−, described by Antonson et al. in role of ERβ in biological processes, making it difficult 2012.160 Homologous recombination was employed to to predict possible phenotypes. Therefore, the ERβ-null either disrupt (αERKO) or completely excise (ERαKO mice have served a tremendous role in providing insight

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1115 into the importance of the newly discovered ER form to exons 1 and 2 of the Cyp19 gene.175 Toda et al. generated female fertility, and studies to date indicate ERβ plays a the most recent mouse line of Cyp19-null in 2001 with a particularly important role in ovarian function. Four dif- targeted disruption of exon 9 of the Cyp19 gene.131 These ferent lines of ERβ-null mice have been described. The ArKO female phenotypes are indistinguishable.131,156,175 βERKO was first described by Krege et al. in 1998,171 and Briefly, these mice are infertile due to ovarian dysfunc- the ERβKO, or Ex3βERKO, was described by Dupont tion marked by cystic follicles and a failure to respond to et al. in 2000.158 Homologous recombination was exogenous gonadotropins. This phenotype is described employed in both lines to disrupt exon 3158 of the murine in detail in the subsection Mice lacking Cyp19 within Esr2 (ERβ) gene. As described to date, the reproductive, the section Ovarian Phenotypes in Mouse Models of endocrine, and ovarian phenotypes of both lines are Disrupted Estrogen Signaling of this chapter. indistinguishable, with both exhibiting female subfertil- ity. Then, in 2002, Shughrue et al. reported the third line UTERINE PHENOTYPES IN GLOBAL ER- OR of ERβKO animals, however, no uterine or ovarian CYP19-NULL MICE 191 β L−/ phenotypes were reported. Recently, ER KOST Females within each respective model exhibit a simi- L− animals, which contain LoxP sites flanking exon 3 of lar phenotypic syndrome. Female mice lacking ERα or Esr2, were generated using the Cre/loxP recombination aromatase are infertile due to dysfunction of numer- system.172 Interestingly, female mice from this recently ous physiological systems, including the ovary (see the β L−/L− described ER KOST colony were reported to be ster- section Estrogen Receptor Signaling in Ovarian Func- ile due to an ovarian defect. tion) and uterus, whereas ERβ-null females exhibit reduced or lost fecundity that is largely attributable to α β MICE LACKING ER AND ovarian dysfunction. A level of caution is warranted Much insight into the physiological roles of E2 has when making phenotypic comparisons between the been gained from the study of mice lacking one or the ER-null and Cyp19-null models because sensitivity other known ER forms. Still, conclusions drawn from to maternally derived estrogens may provide a more these models are confounded by possible compensatory normal developmental environment during gestation mechanisms provided by the opposite ER in each respec- in Cyp19-null mice, and sensitivity to dietary estrogens tive ERKO model. Furthermore, mice lacking only one ER during adulthood is able to abate several phenotypes in form will lose possible ERα/ERβ heterodimer-mediated Cyp19-null mice.194 The reported uterine phenotypes are actions.192,193 Therefore, compound ER-null mice (i.e., summarized in Table 25.3. αβERKO) represent an opportunity to elucidate potential All lines of ER-null females exhibit uteri that possess compensatory and cooperative actions of ERα and ERβ. the expected tissue compartments, myometrium, endo- The two reported lines of compound ER-null mice are metrial stroma, and epithelium134,159,174 (Figure 25.5). the αβERKO, described by Couse et al. in 1999,174 and the However, in females lacking functional ERα or Cyp19, ERαβKO, described by Dupont et al. in 2000.158 Both were uteri are overtly hypoplastic and exhibit severely reduced generated by crossbreeding animals heterozygous for weights relative to WT littermates,131,134,156,195 whereas the respective individual ER-null mice and, as described ERβ-null females possess uteri that are grossly normal to date, exhibit comparable reproductive, endocrine, and and responsive to ovarian-derived steroids134 (Figure ovarian phenotypes as discussed in the subsection Mice 25.5). The uterine endometrium of ERα-null females is Lacking ER α and β within the section Ovarian Pheno- severely hypotrophic, poorly organized, and possesses types in Mouse Models of Disrupted Estrogen Signaling a paucity of glandular structures (Figure 25.5(B)).159,196 of this chapter. Shughrue et al. generated an additional The luminal and glandular epithelial cells in ERα-null ERαβKO line in 2002, however, their reproductive func- uteri are severely immature with fewer glands and con- tion was not described, as the study focused on ER-medi- sistently exhibit a cuboidal morphology, versus the tall ated estrogen responses in the brain.191 columnar morphology and basal location of the nucleus of an “estrogenized” epithelium in WT uteri (Figure MICE LACKING CYP19 25.5(B)). Therefore, fetal, neonatal, and perinatal devel- Estrogens are produced by aromatase cytochrome opment of the female reproductive tract in mice is largely P450, the product of Cyp19 gene. Female mice with independent of ERα- and ERβ-mediated actions, but disruption of circulating estrogen production exhibit estrogen responsiveness and sexual maturation of the altered female reproduction.131,156,175 There are three ani- adult uterus are ablated after the loss of functional ERα. mal models of Cyp19-null (called ArKO). Fisher et al. The totality of the ERα-null phenotype and lack of any reported the first mouse line in 1998, which disrupted overt uterine abnormalities in ERβ-null females suggest exon 9 of Cyp19 gene, as the region is a highly conserved that ERβ has little function in mediating estrogen actions region among all aromatase.156 Later, in 1998, Honda in the uterus. Moreover, ERαβ-null also demonstrated a et al. reported a mouse line with targeted disruption of similar uterine phenotype as ERα-null (Figure 25.5).197

4. FEMALE REPRODUCTIVE SYSTEM 1116 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

FIGURE 25.5 Gross uterine morphology of adult female estrogen receptor knockout mouse models. (A) Female reproductive phenotype of wild-type (WT) control, ERα-null, ERβ-null, and ERαβ-null. (B) Histology of uterine and vaginal tissue of WT, ERα-null and ERβ-null females. Top: Uterine cross-sections from representative adult WT, ERα-null and ERβ-null adult females illustrating the presence of a normal uterine architecture in all. The WT uterine section illustrates a normal myometrium (Myo), endometrial stroma (End), and columnar luminal epithe- lium (arrowhead). The ERα-null uterine section illustrates the characteristic hypoplasia of each anatomical compartment, a slightly disorganized endometrial stroma, and cuboidal luminal (arrowhead) and glandular epithelium. The ERβ-null uterine section is indistinguishable from that of the WT. Bottom: Vaginal cross-sections from tissue adult WT, ERα-null, and ERβ-null female mice. The WT and ERβ-null vaginal tissues illustrate a normal stroma (Str) and hypertrophied epithelium (Epi) showing estrogen-induced stratification and cornification (arrow). In contrast, the ERα-null vaginal tissue indicates complete estrogen insensitivity, as illustrated by a thin epithelium (Epi) and total lack of cornification. Source: Reproduced with permission from Refs 197 and 134.

Weihua et al. reported that ERβ-null females exhibited deletion of ERα, using the Cre/loxP recombination sys- a slightly aberrant uterine growth response after estro- tem, by crossing Wnt7aCre+198 with Esr1f/f animals159 to gen replacement; however, the uterine bioassay was generate mice with deletion of ERα in uterine epithe- conducted in immature intact, not ovariectomized adult, lial cells (Wnt7aCre+;Esr1f/f). The expression of ERα in animals.133 In addition, Wada-Hiraike et al. showed that the uterine luminal and glandular epithelium of these in immature females, loss of ERβ leads to increased uter- animals was ablated, while the ERα expression in the ine epithelial proliferation induced by E2 compared to stromal cells and other uterine cells remains intact.170 WT uteri.173 Although ERβ-null females are subfertile, The epithelial ERα was ablated not only in the uterus in when pregnancies are established they are sustained this mouse line170 but also in the oviduct (unpublished to term,171 indicating uterine competence. More recent data). As expected, based on findings in the global ERα findings suggest that loss of ERβ leads to complete ste- knockouts, loss of uterine epithelial ERα has no effect rility due to a defect in ovarian function,158,172 described on female reproductive tract development. Uterine his- in more detail in the subsection Mice Lacking ERβ of tological analysis showed a similar uterine morphology the section Ovarian Phenotypes in Mouse Models of as wild-type control.170 The Wnt7aCre+;Esr1f/f uteri are Disrupted Estrogen Signaling of this chapter. sensitive to 24 h treatment of E2, as the uterine epithe- lial proliferation is preserved (Figure 25.6). However, α MICE WITH UTERINE-SPECIFIC DELETION OF ER Wnt7aCre+;Esr1f/f uteri lack a complete uterine response Currently, only a single mouse model with to E2, apparent following a 3-day uterine bioassay, which tissue-specific deletion of ERα in the uterus has been demonstrated a blunted growth response and increased published. Our laboratory has described selective apoptotic activity in Wnt7aCre+;Esr1f/f compared to the

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1117

FIGURE 25.6 Estrogen-induced epithelial DNA synthesis in the uterus of adult ovariectomized mice is ERα dependent but occurs indepen- dent of epithelial ERα. Shown are cross-sections of uterine tissues from ovariectomized WT control, ERα-null, or uterine epithelial ERαKO mice treated for 24 h with vehicle or E2. DNA synthesis as indicated by nuclear EdU incorporation is detected in the luminal epithelia (arrowheads) of vehicle-treated WT uteri, and this is dramatically increased after E2 treatment of WT or epithelial ERαKO uteri. In contrast, ERα-null uteri exhibit a basal level of DNA synthesis that is unchanged after E2 treatment, indicating that E2-induced proliferation of the uterine epithelium is dependent on functional ERα (Scale bar = 100 μM). Hoeschst was used as a counterstain to visualize the tissue. Source: Adapted with permission from Ref. 170 control uteri. Additionally, a lack of ERα expression 201 with E, and in the critical P box (Figure 25.3(B)), K in the uterine epithelial cells contributes to complete at position 210 with A, K at position 214 with A, and R infertility, partly due to an implantation defect.170 This at position 215 with E (Figure 25.3(B)) as described by suggests that uterine epithelial ERα is dispensable for Ahlbory-Dieker et al. in 2009 (called ERαEAAE/EAAE).164 early uterine proliferative responses but crucial for com- The NERKI+/− females have normal uterine plete biological responses induced by E2, as well as for development but exhibit hyperplastic uteri and are hyper- embryo implantation. sensitive to estrogen.132 This suggests that disruption of the ERα binding to ERE while maintaining the presence MICE WITH MUTATED DNA BINDING DOMAINS OF of a normal WT allele leads to an aberrant overstimu- α ER lated uterine response to estrogen. Interestingly, these The generation of mouse lines with mutation in the NERKI+/− are infertile and exhibit a uterine abnormality specific domains of ERα allows us to dissect the biologi- of enlarged hyperplastic endometrial glands despite cal functionality of individual domains of the receptor possessing normal levels of circulating sex steroids.132 in vivo. To date, there are two mouse lines with mutations ERαAA/− females have normal uterine development. that are designed to disrupt the DNA binding function of Initially, O’Brien et al. reported that ERαAA/− females, the ERα that have been “knocked-in” (KI) at the ERα gene with mutation of the DNA binding domain, maintained locus. The first line was generated by replacing critical proliferative responses—induced by E2.162 However, P-box amino acids E207 and G208 with alanines (Figure in subsequent studies, no uterine proliferation was 25.3(B)). This line was named “Nongenomic ER knock- observed.163,199 Ahlbory-Dieker et al. showed that, unlike in” (NERKI), as these mutations were intended to restrict the NERKI+/−, females heterozygous for the ERαEAAE ERα signaling to the nongenomic and tethered mecha- mutation are fertile. The homozygous ERαEAAE/EAAE nisms (Figure 25.4). Female NERKI+/− animals that have females have normal reproductive tract development, one mutated allele and one WT allele132 were infertile, but uteri are severely hypoplastic (Figure 25.7(A)), simi- exhibiting a highly novel hyperplastic uterine phenotype, lar to global ERα-null uteri. Additionally, ERαEAAE/EAAE so NERKI+/− males were crossed with ERα-null hetero- uteri do not respond to E treatment, as the estrogen- zygous (WT/KO) females to produce mice with one regulated genes (such as Pgr, Cdkn1a, and Igf1) failed to NERKI mutated allele and one deleted Esr1 allele, called be upregulated in ERαEAAE/EAAE compared to control ERα KIKO or ERαAA/− as described by O’Brien et al. in uteri. The females from these two mouse lines with point 2006.162 The second line of DNA binding domain knock- mutations in the DNA binding domain of ERα are infer- in animals possessed mutation of four amino acids in the tile. Thus the physiological function of the DNA binding first zinc finger of the Esr1 gene, substituting Y at position domain of ERα is crucial for female reproduction.

4. FEMALE REPRODUCTIVE SYSTEM 1118 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

FIGURE 25.7 Gross uterine morphology and histological analysis of mice with knock-in of mutated ERα. (A) Female homozygous ERαEAAE/EAAE mice develop hypoplastic uteri (indicated by arrow) compared to heterozygous control (ERαEAAE/+). (B) Uterine weight of ERα+/+, ERα−/−, and ERαAF- 10 or (C) ERαAF-2+/+ and ERαAF-20 mice treated with placebo control or E2 (80 μg/kg per day for 2 weeks). (D) Adult reproduc- tive tract histology from WT control and ERαG525L mice. H&E staining demonstrated that ERαG525L animals exhibit hypoplastic uteri compared to WT control uteri. Insets depict gross repro- ductive tract morphology. Scale bar = 100 μm. (E) H&E staining of uterine tissues from WT control (+/+) and AF2ERKI/KI mice. Insets depict gross reproductive tract morphology. Scale bar: 100 μm. Source: Reproduced and modified with permission from Refs. 164,165,167–169.

MICE WITH MUTATED AF-1 OR AF-2 DOMAINS OF ERα reported the third line of animals with mutations within As discussed in the section Receptor Structure, AF-1 LDB of ERα169 with two point mutations of L at position and AF-2 are important for ER transcriptional activ- 543 and 544 to A (called AF2ERKI/KI animals). All females ity (Figure 25.1). A single mouse line with amino acids from the ERαAF-10, ERαG525L, ERαAF-20, and AF2ERKI/KI 2–128 deleted from exon 1 of Esr1, which removes the AF mouse lines are sterile.165,167–169 domain, called ERαAF-10, was described by Billon-Gales ERαAF-10 females exhibited minimal uterine wet et al. in 2009.165 There are three reported mouse lines with weight gain compared to ER+/+ uteri after treatment mutation in AF-2 domain of ERα. In the first line, the ani- with E2 pellets for two consecutive weeks ( Figure mals have a single point mutation in ERα of G at position 25.7(B)), while ERαAF-20 females did not respond 525 to L in the ligand binding domain (LBD), and this ( Figure 25.7(C)).165–167 This indicates that the ERαAF-2 model was described by Sinkevicius et al. in 2008 (called functional domain contributes to minimal uterine weight “Estrogen-nonresponsive ERα Knock-in or ENERKI” increase induced by E2 in the absence of AF-1. Both lines or ERαG525L).168 Billon-Gales et al. generated a second of AF-2-mutated animals (ERαG525L and AF2ERKI/KI) line of LBD mutation animals in 2011, with amino acids display severely hypoplastic uteri (Figure 25.7(D,E)) and 543–549 deleted from the LBD of ERα, removing helix lack uterine growth response to E2 treatment.167–169 Inter- 12 and thus AF-2 functionality (called ERαAF-20).167 One estingly, uterine wet weight can be increased by using month after the second line was published, Arao et al. the synthetic ERα agonist PPT (Table 25.2) in ERαG525L

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1119 or by using the ER antagonists ICI 182,780 or tamoxifen REA or REA+/−).208 Gross uterine morphology of REA+/− (Table 25.2) in AF2ERKI/KI females.168,169 The ability of appeared normal compared to control littermates. How- the antagonists to mediate responses seems to be due ever, the REA+/− females exhibited hypersensitivity to to a unique conformation of the LBD of the AF2ER that E2 treatment, as uterine weight and luminal epithelial leads to AF-1-dependent transcriptional activity.169,200 cell height and proliferation were increased compared to Arao et al. also demonstrated that the uterine response E2-treated control uteri. In addition, E2-induced uterine to ICI or tamoxifen includes increased DNA synthesis genes (such as C3 and Ltf) were also elevated in E2-treated in the uterine epithelial cells of AF2ERKI/KI. The growth REA+/− compared to E2-treated control uteri.208 factor IGF1 (insulin-like growth factor 1) induced mini- Due to the embryonic lethality of REA−/−,208 uterine mal uterine epithelial proliferation in ERαG525L and was deletion of REA was generated using PgrCre+ crossed with ineffective in AF2ERKI/KI uteri.168,169 Together, these REAf/f, producing REAf/d (heterozygous deletion) and findings indicated that both AF-1 and AF-2 activation REAd/d (homozygous deletion) animal models. Uterine domains of ERα contribute to a normal regulation of epithelial cell hyperproliferation after E2 treatment was uterine growth and reproductive functions. As the AF seen in REAf/d, similar to the E2 response of REA+/− uteri.207 domains mediate ER-co-regulator interaction (Table The enhanced uterine weight increase by E2 in REAf/d 25.1), this emphasizes the importance of effective ERα females is in part due to increased levels of the aquaporin co-activator protein recruitment for successful uterine water transport gene (Aqp4). The levels of Aqp3 and Aqp5 E2 response. Similarly, mice lacking sufficient SRC-1 induced by E2 treatment were similar between control co-activator (Src1−/−) exhibit measurably diminished and REAf/d uteri, but higher levels of Aqp4 were reached uterine response to E2.201 in REAf/d than control uteri after E treatment.207 REAf/d females were subfertile, while REAd/d females were com- Uterine Phenotypes in Mouse Models with pletely infertile. No ovulation defect was found in REAd/d Mutations in Transcriptional Proteins that Affect females, rather REAd/d females showed altered uterine ER Function growth and maturation, as well as implantation, decidual- MICE WITH UTERINE-SPECIFIC DELETION OF ization, and placentation defects.207 Aberrant apoptosis is COUP-TFII observed in REAd/d uteri, as reflected by drastic increases Chicken ovalbumin upstream promoter transcription in both TUNEL and active caspase-3 levels in REAd/d uteri factor II (COUP-TFII), encoded from the Nr2f2 gene, is a compared to wild-type uteri.207 These findings suggest that transcription factor, belonging to nuclear receptor super- REA regulates proper ER responsiveness in vivo and is cru- family. COUP-TFII is expressed in uterine stromal cells and cial for ER-mediated female uterine function. is crucial for female reproduction.202–204 The tissue-specific deletion of COUP-TFII (COUP-TFIId/d) in the uterus using MICE WITH UTERINE-SPECIFIC DELETION OF MSX1 PgrCre+ crossed with COUP-TFIIf/f animals demonstrated AND MSX2 that COUP-TFIId/d females exhibited normal uterine devel- Mammalian homeobox genes, Msx1 and Msx2, are opment but were completely infertile, in part, due to lack of crucial for organogenesis and embryo development. implantation.202 Kurihara et al. also showed an enhanced Nallasamy et al. reported that the selective ablation of estrogenic response in the epithelial cells of COUP-TFIId/d either Msx1 (PgrCre+;Msx1f/f, called Msx1d/d) or Msx2 uteri, as E-responsive genes lactotransferrin (Ltf) and (PgrCre+;Msx2f/f, called Msx2d/d) in the uterus lead to sub- mucin 1 (Muc1) were elevated during the receptive win- fertility in females, however, deletion of both Msx1 and dow, which suggested that COUP-TFII modulates uterine Msx2 in the uterus (Msx1d/dMsx2d/d), resulted in com- epithelial ERα activity during pregnancy.202 Inhibition of plete infertility.210 This indicates that Msx1 and Msx2 func- ERα activity using ER antagonist (ICI 182,789) partially res- tions can partially compensate for each other. The female cued implantation in COUP-TFIId/d females and corrected infertility in Msx1d/dMsx2d/d females is in part due to an the expression level of Ltf and Muc1.203 These results sug- implantation defect.210 Deletion of Msx1 and Msx2 in the gest COUP-TFII plays an essential role in regulating uter- uterus causes uterine hypersensitivity to estrogen due to ine ERα activity during early pregnancy in rodents. an increase in uterine ERα activity. Increased ER phos- phorylation in uterine epithelial and subepithelial stromal MICE WITH UTERINE-SPECIFIC DELETION OF REA cells, resulting from high fibroblast growth factor (FGF) Repressor of estrogen receptor activity (REA) is an ER expression, were observed.210 Female mice with uterine co-regulator that represses ER activity both in vitro and Msx1 and Msx2 deletion exhibited not only an altered ERα in vivo.205–209 In Chinese hamster ovarian (CHO) cells, activity but also showed aberrant expression of Wnt/β- REA is shown to be a direct co-regulator of ERα and catenin signaling molecules.210 This indicates that homeo- ERβ, but not PR or RAR.205 In 2005, Park et al. gener- box genes Msx1 and Msx2 in the uterus are mediators of ated a mouse line with a targeted deletion of REA exon physiological ERα activity, which subsequently play piv- sequence encoding amino acids 12–201 (heterozygous otal roles in establishing successful pregnancy.

4. FEMALE REPRODUCTIVE SYSTEM 1120 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

MICE WITH UTERINE-SPECIFIC DELETION OF NCOA6 genes, lactoferrin (Ltf), chloride channel calcium–activated Nuclear receptor co-activator-6 (NCOA-6) is an ERα co- 3 (Clca3), and complement component 3 (C3) were also activator. Kawagoe et al. reported that NCOA-6 regulates elevated in Mig-6d/d uteri in the presence of both E and uterine ER activity by using selective deletion of Ncoa6 in P, however, the PR-targeted genes amphiregulin (Areg) the uterus (PgrCre+;Ncoa6f/f, called Ncoa6d/d).211 Ovulation, and follistatin (Fst) remained comparable to wild-type fertilization rates, and blastocyst development are all nor- uteri. Ovariectomized Mig-6d/d females that were treated mal in Ncoa6d/d females; however, embryo implantation with E for 3 months exhibited endometrial cancer. How- fails. The implantation defect is in part due to an increased ever, co-treatment with E and P attenuated the pathology level of the ERα co-activator, SRC3, in the uterine epithe- in the uterus of E treated Mig-6d/d, although the uterine lial cells of Ncoa6d/d, which causes elevated uterine epithe- hyperplasia was still observed in Mig-6d/d females. Addi- lial ERα activity compared to wild-type uteri. Moreover, a tionally, the expression level of MIG-6 in women with diminished PR activity, particularly during the implan- endometrioid carcinoma was elevated compared to nor- tation period, was observed. Implantation was partially mal endometrium.108 These findings together suggest that rescued and the level of MUC1 was corrected in uterine Mig-6, as a tumor suppressor in both rodents and human, luminal epithelia by treating Ncoa6d/d females with the plays a crucial role in uterine growth in response to E in ER antagonist ICI 182 780.211 These findings suggest that part by mediating the protective action of P.108 loss of uterine NCOA-6 causes implantation failure due to MICE WITH UTERINE-SPECIFIC DELETION OF HAND2 an aberrant ER-estrogen hypersensitivity. Heart and neural crest derivatives expressed 2 (Hand2) MICE WITH UTERINE-SPECIFIC DELETION OF MIG6 is regulated by P in the uterus, as protein expression of Mitogen-inducible gene 6 (Mig-6, encoded from the HAND2 is induced in the uterine stromal cells after P treat- Errfi1 gene) is a downstream target of P-PR and SRC1 ment.212 PRKO uteri showed loss of Hand2 expression, action in the uterus.108 However, uterine-specific dele- indicating the direct regulation of Hand2 by PR.212 Selec- tion of Mig-6 (Mig-6d/d or PgrCre+;Mig-6f/f) demonstrated tive ablation of Hand2 in the uterus (Hand2d/d) using the altered ERα signaling resulting in uterine hyperplasia.108 PgrCre+;Hand2f/f animal model demonstrated an elevated Mig-6d/d uteri are enlarged, with weights approximately uterine ER activity.212 Loss of uterine Hand2 expression in 3-fold more than wild-type (Mig-6f/f) uteri. Immunohis- the stromal cells leads to aberrant fibroblast growth factor tochemical analysis demonstrated that Mig-6d/d uteri receptor (FGFR) signaling, which subsequently induced α exhibit hyperproliferation of the endometrium, with ERK signaling and ER activity in the uterine epithelium increased phospho-H3 staining, as a result of increased (Figure 25.8), resulting in complete infertility. The infer- ERα expression and phosphorylation.108 The ERα targeted tility in Hand2d/d females is in part due to an implanta- tion defect as a result of enhanced uterine epithelial ERα

FIGURE 25.8 Schematic representation of paracrine and autocrine mechanisms of uterine proliferation in response to hormones. E2 stimulation of proliferation of the uterine epithelium requires the presence of functional ERα in the under- lying stroma independent from epithelium, indicating that E2/ ERα actions in the stroma induce the secretion of paracrine factors that then act on the epithelium to stimulate prolifera- tion. Progesterone (P) acts through epithelial and stromal PRs to inhibit the proliferative response of the epithelium to E, while inducing proliferation of the underlying stroma. BM, basement membrane. (For detail see text.)

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1121 activity during the implantation period.212 This suggests of ovariectomized mice.128,223 This is an ERα dependent that Hand2 expression in the uterus, orchestrated by an regulation, as PR is found only in epithelial cells, regard- action of PR, mediates a proper physiological ERα activity less of E2 treatment in αERKO. Additionally, epithelial ERα (Figure 25.8) in response to ovarian hormones, leading to is dispensable for this response as increased PR expression successful pregnancy establishment. in the stromal cells is preserved in the absence of epithelial ERα.170 These observations have been largely confirmed in Vaginal and Cervical Phenotypes in Mouse Models studies using a PR-lacZ reporter mouse that represents Pgr of Disrupted Estrogen Signaling gene promoter activity as β-galactosidase activity in situ.224 The vaginal mucosa is composed of an epithelium and In these studies, β-galactosidase activity representative of underlying stroma, possesses significant levels of ER,213 PR expression is observed in luminal and glandular epi- and is highly sensitive to estrogens.214 Endogenous or thelium after ovariectomy and decreased after estrogen exogenous estrogens induce stromal differentiation, epi- treatment.224 P treatments also lead to decreased PR-lacZ thelial proliferation, and epithelial keratin synthesis to expression in all cell types.224 produce the stratified layer of cornified cells that lines the In the adult ovine and bovine uterus, PR expression vaginal lumen.214–216 These changes in the vaginal mucosa is highest during the early proliferative phase and pres- are used experimentally to estimate the level of circulat- ent in all uterine cell types but is most abundant in the ing gonadal steroids and approximate the estrous cycle stroma and myometrium.138,140–142 PR levels decline as stage. Despite exposure to elevated endogenous E2 levels, the cycle progresses to the secretory phase, initially in the vaginal tissue of ERα-null females lacks any indica- the stroma and myometrial cells.138,140–142 tions of estrogenization (Figure 25.5(B)).134 Exogenous In the human uterus, PR immunoreactivity is increased administration of E2 or DES also elicits no discernible in all cell types during the proliferative phase but mea- vaginal response in ERα-null females.196 In contrast, the surably lower than ER levels.143,145 During the secretory vaginal mucosa of ERβ-null females undergoes the nor- phase, PR levels are maintained in the stroma and myo- mal cyclic changes as dictated by ovarian sex steroids metrium but decrease in the glandular epithelium such (Figure 25.5(B)). An effect on the vaginal mucosa simi- that levels are eventually undetectable.143,145 Studies lar to that observed in ERα-null females is produced in using PR isoform selective antibodies have shown that rodents after prolonged ovariectomy or exposure to ER the decrease in PR levels that occurs in the glandular epi- antagonists.217–219 Therefore, estrogenization of the vagi- thelium during the mid-secretory phase is largely repre- nal mucosa in rodents, which is critical to mating, is ERα sentative of PRA, leading to an increased PRB/PRA ratio dependent. Using a human papilloma virus (HPV) trans- in these cells.225 In stroma, where fewer cells expressed genic model of cervical cancer, Chung et al. demonstrated PR relative to glandular epithelium, PRA is more abun- that estrogen promotes development of HPV-induced dant.225 During the late secretory phase, PRA levels in the cervical cancer through ERα, as αERKO HPV–transgenic stroma decrease but remain detectable, whereas stromal animals were protected from developing disease.220,221 PRB expression decreases occur earlier and are signifi- cantly diminished by the late secretory phase.225 The pre- Progesterone Receptor Expression in the Uterus dominance of PRA in the glandular epithelium during the early secretory phase may serve to inhibit further estro- In the murine uterus, PR transcript and protein levels gen-mediated proliferation, as the uteri in PRA-null mice are low on days 1–2 of pregnancy with the highest expres- exhibit an exaggerated proliferative response to estro- sion occurring in the epithelia and subepithelial stroma.126 gens.181 As P levels rise during the mid-secretory phase, PR levels rise in these same tissues on days 3 and 4 of preg- PRB is concentrated in nuclear foci, suggesting active nancy, leading up to implantation.126,222 Receptor-medi- transcriptional activity.226 A similar finding of focal-con- ated E2 actions are a primary regulator of PR expression in centrated PRB is reported in endometrial cancer cells in the rodent uterus, but these actions are complex and spe- postmenopausal women, suggesting that inappropriate cific to the different uterine compartments. Ovariectomy PRB activity may be associated with uterine cancer.226 The leads to increased PR levels in the uterine luminal epithe- pattern of PR expression in uteri of nonhuman primates lium that can be reduced upon exogenous E2 treatments, is similar to that described in humans. In female baboons, indicating E2 acts to repress PR expression in this uterine PR levels are increased during the proliferative phase and compartment.128,223 Recombination experiments of stro- then decrease accordingly during the secretory phase.152 mal and epithelial tissues similar to those described earlier, Uterine Phenotypes in Mouse Models with as well as studies using epithelial selective ERα deletion, Disrupted Progesterone Signaling indicate this to be a paracrine-mediated effect that requires functional ERα in the stroma.125,170 Simultaneous with MICE LACKING PR causing decreased PR expression in the epithelium, E2 There are several reported lines of PR-null mice. Mice treatments lead to increased PR levels in the uterine stroma lacking both nuclear PR isoforms (PRKO) were first

4. FEMALE REPRODUCTIVE SYSTEM 1122 25. STEROID RECEPTORS IN THE UTERUS AND OVARY described by Lydon et al. in 1995 and were generated via mice demonstrated the spatiotemporal expression pat- targeted disruption of exon 1 of the murine Pgr gene.176 tern of SRC1 and SRC2 in the luminal epithelial, stromal In 2006, Hashimoto-Partyka et al. generated the PR as well as secondary decidual zone. The uterine expres- (Pgrf/f)-targeted mutation in exon 1 of PR using EIIa-Cre sion pattern of SRC1 and SRC2 was similar to the pat- promoter, which expressed Cre recombinase in the pre- tern of PR expression during pregnancy.232 In uteri from implantation embryo in order to create a deletion of exon ovariectomized animals, the majority of SRC1 and SRC2 1 and loss of PR protein.177 In 2010, Fernandez-Valdivia was expressed in the luminal and glandular epithelium, et al. described a mouse line that had a deletion of PR much like PR expression.232 Both E and P, either alone or exon 2 and loss of PR protein using Pgrf/f and ZP3-Cre administered together, increased the expression of SRC1 promoter expressing mice (called PRd/d).178 and SRC2 in the uterine stromal cells.232 Mice lacking only PRA (PRAKO)180 or PRB There are several reports describing SRC-null mice and (PRBKO)182,227 were generated via a Cre/loxP-based resulting effects on female reproductive functions.201,232–234 gene targeting strategy that mutated the respective start Global disruption of the Src1 gene demonstrated that both codons for each isoform while preserving the transla- males and females were fertile, however, females exhibited tional reading frame of the other. Comparative studies a blunted response to ovarian hormones.201 Treatment of in all models have greatly furthered our understand- Src1−/− females with E2 for 3 consecutive days resulted ing of the divergent functions for the two PR isoforms. in less uterine weight increase than in WT, similarly, arti- Recently, Franco et al. generated an epithelial cell spe- ficial induction of decidualization resulted in less stromal cific deletion of both PRA and PRB in the uterus (called differentiation of Src1−/− uteri.201 Due to a lethality of Src2 Wnt7a-Cre+PRf/−) using Wnt7a-Cre expressing promoter global knockout,235 uterine-specific SRC2 deletion (Src2d/d) crossed with Pgrf/− animals.179 was generated using PgrCre+ crossed with Src2f/f animals.233 PR-null mice exhibit normally developed uteri that pos- Src2d/d females were infertile, in part, due to an implanta- sess the expected uterine architecture (Table 25.3).181,182 tion and decidualization defect.232,233,236 E2-induced uter- During the secretory phase of the murine estrous cycle, ine epithelial cell proliferation was preserved in Src2d/d decreasing circulating levels of E2 concurrent with rising uteri.233 The decidualization markers such as Bmp2, Ptgs2 P cause a shift in uterine proliferation from the epithelial to (Cox-2), and Fst were significantly decreased in Src2d/d stroma cells.228 Estrogen, through ERα, is known to stim- when compared to wild-type uteri.233 A complete absence ulate uterine epithelial cell proliferation in adult mouse of decidual response was observed in Src2d/d uteri in a uteri,125,159,229 and P exhibits the antiproliferative effect on Src1−/− background.233 However, Src3−/− females had no the action of E2 (Figure 25.8).230,231 Therefore, PR-mediated overt uterine phenotype.232 These findings indicate that of P actions negatively modulate E2-induced proliferation of the three SRCs, SRC2 is most crucial for P-dependent uter- the uterine epithelium.228 In addition, ovariectomized PR- ine function. null females treated with daily injections of E2 and P for 3 weeks exhibit abnormally large fluid-filled uteri that are MICE WITH UTERINE-SPECIFIC DELETION OF characterized by a thickened uterine wall due to extensive COUP-TFII extracellular edema, hyperproliferation of the glandular COUP-TFII regulates uterine ER function during early epithelia resulting in a disordered cellular arrangement, pregnancy as mentioned earlier. Additionally, the COUP- and acute inflammation of the endometrium.176 The TFIId/d animal model also demonstrated that COUP-TFII hyperestrogenic response in PR-null uteri is similar to that is a downstream effector for PR-mediated decidualiza- observed in ovariectomized mice after prolonged exposure tion.202 Loss of uterine COUP-TFII resulted in a defect to unopposed estrogens229 and provides strong support in decidualization, as reflected by lack of expression of of the modulatory actions of PR in the uterus.176 Similar decidual cell markers (WNT4 and BMP2).203 Treatment experiments in isoform-specific PR-null mice indicate this of COUP-TFIId/d females with ICI 182,780 restores the phenotype is reproduced in PRA-null mice only, whereas expression of WNT4 and BMP2 in the decidual cells.203 PRB-null mice exhibit a normal uterine response, indicat- Together, the findings from uterine-specific deletion ing that PRA is primarily responsible for negatively modu- of COUP-TFII suggest that COUP-TFII regulates P lating the uterine response to E2.180–182 signaling during implantation and decidualization by controlling uterine ER activity. Uterine Phenotypes in Mouse Model with Mutations in Transcriptional Proteins That MICE WITH UTERINE-SPECIFIC DELETION OF REA Mediate PR Function As discussed earlier, REA is crucial for uterine MICE WITH UTERINE-SPECIFIC DELETION OF P160 responses to ovarian hormones and establishment of (SRCs) pregnancy. Using REAd/d animals also demonstrated that The SRCs mediate P-dependent action in the uterus.232 loss of uterine REA led to aberrant PR function, which Immunohistochemical analysis in the uteri of pregnant resulted in a decidualization defect.207 The decidual cell

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1123 markers, Wnt4 and Bmp2, were not increased in REAd/d uteri after artificial decidualization. The defect in decidu- alization is due to loss of PR expression after uterine REA ablation.207 These findings indicate that uterine REA expression is not only crucial for ER activity but also for PR function to establish and maintain pregnancy.

MICE WITH UTERINE-SPECIFIC DELETION OF FKBPs FK506 Binding Proteins 4 (FKBP4; encoded by Fkbp52 gene) and 5 (FKBP5; encoded by Fkbp51 gene) are immu- nophilin family co-chaperones that interact with PR in the absence of the ligands.237 Fkbp4 is expressed in the luminal and glandular epithelial cells during implanta- tion, whereas a minimal level was detected in the uter- ine stromal cells, similar to PR expression.237 During the decidual response, Fkbp4 was highly expressed in both primary and secondary decidual zones. Although the expression of Fkbp5 also showed a similar pattern, the level of expression was much lower than that of Fkbp4.237 FIGURE 25.9 The global knockout of FKBP5 was fertile and showed Estrogenic response in immature uteri in mice (21-day-old). DNA synthesis as indicated by nuclear EdU incorpora- 237,238 no impairment of uterine functions. However, tion is detected in vehicle-treated uterine epithelial cells (arrowhead). the global knockout of FKBP4 caused complete female However, DNA synthesis is dramatically increased in the luminal and infertility,237,238 due to implantation and decidualization glandular epithelia as well as stromal cells in the uteri 16 h after E2 defects,237,238 suggesting that FKBP4 not only binds to PR treatment. Hoescht was used as a counterstain to visualize the tissue. μ but is crucial for PR action during pregnancy. Scale bar = 100 m. Arrowhead indicates EdU positive epithelial cell.

model to mimic the uterine events that occur during the Uterine Response to Estradiol and Progesterone estrous phase of the rodent cycle or immediately after the preovulatory E2 surge. Morphological and biochemical Changes in Physiology changes occur in the rodent uterus after estrogen stimu- Estrogens stimulate a complex process of epithelial lation following an established biphasic temporal pat- proliferation and differentiation in the sexually mature tern241,242 (Table 25.4). Estrogen-stimulated changes in uterus that leads to the formation of a multilayered secre- the rodent uterus that occur early, within the first 6 h after tory endometrium. This effect is critical to provide a suit- treatment, include increases in nuclear ER occupancy, able intrauterine environment for the establishment and water imbibition, vascular permeability and hyperemia, maintenance of pregnancy. Early studies in neonatal and prostaglandin release, glucose metabolism, eosinophil prepubertal rodents found that both the uterine stroma infiltration, gene expression (e.g., c-fos), and lipid and and epithelium proliferate in response to estrogens protein synthesis (Table 25.4). Recent ERα ChIP-Seq pro- ( Figure 25.9);239 however, estrogen-induced mitogenesis files from uterine tissues showed that the receptor preoc- in uteri of sexually mature rodents is limited to the epithe- cupies chromatin sites and that E2 treatment increases lium.229 Therefore, sexual maturation of the rodent uterus ERα recruitment.243 These processes are then accompa- is not simply marked by the presence of ER or an estrogen nied by a delayed response that peaks after 24–72 h and response but is rather the acquired capacity to undergo includes dramatic increases in RNA and DNA synthe- synchronized phases of proliferation and differentiation sis, epithelial proliferation, and differentiation toward a as dictated by the ovarian-derived sex steroids. The lack of more columnar secretory phenotype, dramatic increases estrogen-induced uterine epithelial proliferation in ERα- in uterine weight, and continued gene expression (e.g., null uteri indicates the essential role of ERα in this process lactotransferrin) (Table 25.4). Ovariectomized mice (Figure 25.6). Although ERα is present in both epithelial exhibit a three- to four-fold increase in uterine weight and stromal compartments, tissue recombination stud- after three daily treatments with E2 or DES, whereas ies and study of the uterine epithelial-specific ERα null no such response is observed in the uteri of ERα-null indicate the proliferative epithelial response to estrogens females.157,159,244 The early phase effects of water imbibi- is indirect and dependent on stromal ERα actions (Figure tion and hyperemia as well as the late-phase effects of 25.6 and a working model in Figure 25.8).170,240 increased DNA synthesis and epithelial proliferation are Treatment of ovariectomized mice with estrogens absent in ERα-null uteri159,189,196 (Figure 25.6). Interest- (e.g., E2 or DES) has long served as an experimental ingly, females that are heterozygous for the Esr1 gene

4. FEMALE REPRODUCTIVE SYSTEM 1124 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

TABLE 25.4 Biphasic Response of the Rodent Uterus to and stromal cells.245,246 ICI 182,786 (ER antagonist) Estrogens strongly inhibited E2-induced Cebpb transcript in the EARLY UTEROTROPIC RESPONSES (WITHIN 6 H) uterus, suggesting an ER-dependent expression of C/ EBPβ.247 In addition, loss of epithelial ERα in the uterus Nuclear localization of estrogen receptor/recruitment to chromatin did not alter E2-induced Cebpb expression, indicating Activation of receptor-tyrosine kinase pathways that Cebpb expression is independent of epithelial ER.170 α Changes in gene expression (induction/repression of “early” genes) This points to the action of estrogen through ER as the major mediator of C/EBPβ expression in the uterus. Increased vascular permeability Indeed, the deletion of C/EBPβ (C/EBPβ−/−) leads to a Increased protein synthesis lack of the E-induced uterine proliferative response245 Water imbibition as reflected by the absence of mitotic activity, S-phase activity, and an increase in apoptotic activity in the Hyperemia uterine epithelial cells.246 In addition to a blunted uterine Eosinophil infiltration growth response to hormones, the C/EBPβ−/− females 247 Albumin accumulation also exhibit complete infertility due to implantation and decidualization defects.245 Increased electrolytes Pan et al. demonstrated that the uterine expression Lysozyme labilization of minichromosome maintenance proteins (MCMs), Increased cyclic nucleotides, prostaglandins, and associated enzyme a complex required for DNA synthesis initiation, is activity induced after E2 treatment, specifically MCM2 and MCM3248 (Figure 25.8). MCM2 activity is crucial and Increased glucose metabolism and associated enzyme activity required for DNA synthesis in the uterine epithelial Calcium influx cells.249 The DNA replication of uterine epithelial cells Increased lipid synthesis induced by E2 is attenuated by P action, which will be discussed further in a later section. LATE UTEROTROPIC RESPONSES (WITHIN 24 H)

Second peak of nuclear localization of estrogen receptor Estrogen–Growth Factor Cross Talk in the Uterus

Changes in gene expression (induction/repression of “late” genes) The autocrine and paracrine actions of polypeptide growth factors are an integral component of the uter- Increased protein synthesis ine response to estrogens. The uterine response to E2 Increased DNA synthesis is modulated by stromal factors, such as IGF1, that are 250 Epithelial proliferation in “waves” induced by E2 and then impact epithelial responses. Igf1 transcript is increased with concomitant decrease of Cellular hypertrophy Igfbp3242 and activation of the IGF1 receptor and down- Overall increase in uterine dry weight stream effectors following E2 treatment.251 Igf1 transcript is increased in both stromal and epithelial compartments of the uterus by E2, with greater signal apparent in the disruption possess approximately one-half the normal stroma (Figure 25.10(A)).252 Igf1 has been demonstrated level of ERα in the uterus, but their response to estro- to play an essential role in the uterine growth response, gen treatment is comparable with wild-type females. as Igf1 null mice lack a full uterine proliferative response, The total lack of response to estrogens in ERα-null uteri and more specifically, lack G2/M progression of the epi- as well as a lack of late biological response in epithelial thelial cells following E2 stimulation.253 Additionally, ERα knockout uteri provide strong evidence that ERα is transgenic mice overexpressing Igfbp1, which sequesters, required to mediate the full biochemical and biological and therefore decreases, the amount of available IGF1, uterine response to estrogens.159,170 have an attenuated uterine response to E2.254 Uterine response is restored by transplanting Igf1KO uterine tis- Molecular Mechanisms of Estrogen-Induced sue into a WT host,255 which demonstrated the paracrine Uterine Proliferation effect of the host Igf1. Further, E2 treatment results in the Numerous studies have elucidated the molecular activation of downstream mediators of Igf1 signaling, mechanisms of E2-induced uterine epithelial cell pro- including the Igf1 receptor, IRS1,251 AKT, and inhibition liferative responses in animal models. For example, the of GSK3β,252 leading to nuclear translocation of CCND1 transcription factor CCAAT enhancer binding protein and epithelial proliferation. Additionally, picropodo- beta (C/EBPβ) is involved in hormone-induced uter- phyllin (PPP), an IGF1R inhibitor, blunted the prolifer- ine proliferation.245 Maximum uterine expression of C/ ative effect of E2. This effect of PPP on E2-stimulated EBPβ is induced 1 h after E2 treatment in both epithelial uterine proliferation could be reversed, however, by

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1125

ERα-null females.259 Cunha and colleagues125,250,255 used the ERα-null mice in a series of tissue recombination experiments to further demonstrate interaction between ER and growth factor signaling in the murine reproduc- tive tract. In these studies, uterine stoma and epithelium are enzymatically disassociated from prepubertal mice and recombined with corresponding tissue from animals of different treatments or genotypes. The resulting chi- meric stromal-epithelial unit is then implanted under the kidney capsule of an ovariectomized nude mouse that is then treated with various hormonal combina- tions. A caveat to these studies is that neonatal tissues are used and may not accurately reflect the uterine phys- iology of sexually mature females. Nonetheless, these methods have been effectively used to demonstrate that estrogen-induced proliferation of uterine and vaginal epithelium requires functional ERα in the underlying stroma only,250 whereas estrogen-induced increases in secretory products (e.g., LTF, complement C3, keratins) in the uterine or vaginal epithelium requires functional FIGURE 25.10 E2 treatment stimulates uterine epithelial prolif- α eration through IGF1 signaling. (A) In situ hybridization of transverse ER in both uterine compartments. These findings sections of uteri of control (1 and 4) and E2-treated (2, 3, 5, and 6) mice have now been confirmed in sexually mature intact probed by using antisense (1, 2, 4, and 5) or sense (3 and 6) probes. uterine tissue using uterine epithelial cell–selective Dark precipitate represents the hybridization signal and indicates an ERα-null mice.170 In addition, treatment of uterine epi- upregulation of Igf1 after E2 treatment principally in the stroma but thelial ERα-null females with IGF1 or EGF mimics the also in the luminal and glandular epithelium. (Magnification: 1–3, 10×; 4–6, 40×.) (B) E2 signals through IGF1 and GSK3β to induce DNA syn- uterine epithelial cell DNA synthesis stimulated by E2 thesis in the uterine epithelium. BrdU incorporation in uterine section (Figure 25.11). These studies strongly support a paracrine from ovariectomized mice 15 h after the following treatments: (1) con- mechanism of estrogen-mediated epithelial proliferation trol, (2) E2 (subcutaneous, s.c.), (3) E2 (s.c.) + PPP (picropodophyllin; that requires ERα in the stroma (working model in IGF1R inhibitor) given intraluminally, and (4) E2 (s.c.) + PPP + SB415286 Figures 25.8 and 25.11). Interestingly, microarray stud- (GSK3β inhibitor) given intraluminally. Source: Reproduced and modified with permission from Ref. 252. Copyright (2007) National Academy of Sci- ies indicate that the growth factor–induced genomic ences, USA. response is intact in ERα-null uteri, although epithelial proliferation is absent,256 suggesting a greater complex- ity to the estrogen–growth factor signaling “cross talk” co-treatment with PPP and the GSK3β inhibitor, SB415286 than was originally perceived. GF signaling is thought (Figure 25.10(B)).252 These findings elegantly illustrate a to mainly utilize AF-1 via MAPK phosphorylation sites direct role for IGF1R-initiated signaling through AKT and in the N-terminus. Thus is was expected that IGF1 or GSK3β on uterine response to E2. EGF treatment of AF2ERKI/KI mice would lead to uterine In addition, IGF1 or EGF treatment of ovariecto- growth, however, no response was seen,169 indicating mized wild-type mice elicits a pattern of global gene functional AF-2 is also involved in this mechanism. expression similar to that induced by E2, although some estrogen-specific genes are revealed.256 IGF1 treatment Antiproliferative Functions of PR in the was shown to increase the expression of an ERE-driven Endometrium luciferase reporter gene in transgenic mice, providing the Tong and Pollard demonstrated that P inhibits first in vivo evidence of E2-independent ER activation E-induced uterine proliferation via the inhibition of by growth factor.257 ER-null mice provide an excellent DNA synthesis and PCNA expression (a component of in vivo model for the study of ER–growth factor “cross DNA polymerase gamma, which is required for S-phase talk” mechanisms in the uterus. ERα-null uteri express entry).228 Additionally, Ray and Pollard demonstrated wild-type levels of functional EGFR but are unrespon- that Krupple-like transcription factor 4 (KLF4) expres- sive to the mitogenic actions of EGF, confirming the sion is induced whereas KLF15 is suppressed by E2 interaction of these two signaling systems.258 However, treatment in the uterine epithelial cells.249 Conversely, not all EGF responses are lacking in the uteri of αERKO co-treatment with P and E decreased KLF4 expression females, as upregulation of the c-fos gene by this growth and increased KLF15 in the uterus.249 In addition, over- factor remains intact. Similar studies have demonstrated expression of uterine KLF15 using an adenovirus leads that the uterine response to IGF1 is compromised in to a lack of E2-induced uterine epithelial proliferative

4. FEMALE REPRODUCTIVE SYSTEM 1126 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

FIGURE 25.11 Growth factors are able to mimic the proliferative effects of E2 in the mouse uterus. Shown are uterine cross-sections from ovariectomized WT, ERα-null, and uterine epithe- lial ERα-knockout (uterine epithelial ERαKO) mice treated for 24 h with vehicle, E2, epidermal growth factor (EGF), or insulin-like growth factor 1 (IGF1). Uterine tissue was then subjected to immunohis- tochemistry for the antigen Ki67, a marker of cell proliferation. As expected, E2 treatment induces a dramatic increase in Ki67 immunoreactivity in uter- ine luminal and glandular epithelium, and EGF and IGF1 mimic this effect in the absence of endogenous E2. Source: Reproduced and modified with permission from Refs 159,170.

stromal cell proliferation but required for the P-medi- ated inhibition of epithelial cell proliferative response induced by E2.179,228 Wnt7a-Cre+PRf/− animals exhibit an aberrant uterine expression of MCM3 and Cyclin D1,179 which are important mediators of E2-induced uterine epithelial proliferative responses. Moreover, P treatment also inhibits the pro-inflam- matory activity induced by estrogen.263 The inflamma- tory response observed in PR-null uteri after prolonged estrogen/P treatment is extensive and consists of FIGURE 25.12 Progesterone inhibits E2-induced uterine epithe- “marked” infiltration of polymorphonuclear leukocytes lial cell proliferation and stimulates stromal cell proliferation. BrdU into the endometrial stroma, mucosal epithelium, and incorporation in the uterine tissues from ovariectomized mice treated uterine luminal fluid.176 In vitro studies have shown with control (no treatment), E2 (15 h), P (daily for 4 days), and PE2 (P that P inhibits the expression of chemotactic cytokines for 4 days and E2 on the fourth day for 15 h). Source: Reproduced and for neutrophil and lymphocyte infiltration264,265 and may modified with permission from Ref. 228. reduce prostaglandin E levels in uterine decidual and chorionic tissue during early pregnancy.266 Therefore, response due to the loss of MCM2 expression,249 dem- PR-mediated P actions may be critical to suppressing onstrating that KLF15, through the action of P, inhibits the uterine inflammatory response that may accompany E2-induced uterine epithelial DNA synthesis by inhibi- embryo implantation.176,267 The uterine levels of macro- tion of MCM2 expression. Moreover, P also diminished phages, neutrophils, and Ltf are increased in both WT E2 induction of Cyclin D1 nuclear translocalization and and PRKO uteri after unopposed E treatment. The pro- expression of Cyclin A, which consequently decreased inflammatory activity is decreased by co-treatment of E the phosphorylation of pRb and p107.228,260,261 Together, with P in WT, but not in PRKO uteri,223,263 thus PR func- these findings described the mechanism by which P tion is crucial to control uterine inflammatory response inhibits cell cycle activity induced by E2 in the uterine during the ovarian cycle. cells (see Figure 25.8). P not only inhibits the uterine epithelial pro- Roles of PR in the Rodent Uterus during Pregnancy liferation but also contributes to an induction of PR-mediated P actions in the uterus are critical to pre- stromal cell proliferation in the presence of E2. Recom- paring the uterine endometrium for pregnancy. Embryo bination experiments using uterine tissues from implantation is a highly complex process that requires PR-null mice demonstrate that the antiproliferative synchronized cooperation between the blastocyst and functions of P in the uterine epithelium are para- uterine endometrium (see Chapter 38). Circulating E2 crine mediated and require PRA in the stroma125,262 levels peak at ovulation and elicit a cascade of prolif- (Figure 25.12). However, selective deletion of PR in the eration and differentiation in the luminal and glandu- uterine epithelial cells (Wnt7a-Cre+;PRf/− animal model) lar epithelium of the uterus, including induction of demonstrated that epithelial PR expression is dispensable PR expression in the endometrial stroma and myome- for E2-induced uterine weight increase and induction of trium.223 Postovulatory increases in circulating P then

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1127 cause decidualization, a complex process that involves in mice.270 Studies have shown that the ER antagonist massive proliferation and differentiation of the endome- ICI 182,780 prevents artificially induced uterine decid- trial stroma along with localized increases in vascular ualization in wild-type mice, indicating involvement permeability and edema.34,268,269 This process involves of ER signaling.161 Therefore, it is surprising that ERα- the synthesis and interaction of numerous hormones and null mice exhibit a uterine response when exposed to signaling pathways, including prolactin, cytokines, pros- an artificial model of hormonal and mechanical induc- taglandins, and extracellular matrix components.268,269 tion of uterine decidualization.161,271 However, this may The result is a remarkable swelling of the uterine stroma partly be due to the residual ERα protein produced by that is thought to be necessary for implantation by forc- the splice variant found in the uterus of the αERKO ing the uterus to close down on the blastocyst.268,269 The line made by targeted disruption.189 Nevertheless, final stage of apposition is a grasping of the blastocyst these findings suggest that an altered “organization” and ultimate attachment to the uterine wall, a process of the uterine tissue or compensatory pathways might thought to be dependent on secondary rises in ovarian- provide for hormonally driven uterine decidualization derived E2.268,269 Therefore, preparation of the uterus in ERα-null mice. Alternatively, some genes detected at for blastocyst implantation is dependent on the multi- the time of implantation, including the gap junction pro- functional and sometimes opposing effects of E2 and P. tein connexin 26272 and the cytokine leukemia inhibitory In addition, PR-null models suggest that PRA is crucial factor,135 are regulated by dual pathways. One involves not only for implantation but also the decidualization estrogen- stimulated ERα, and the ability to induce via this process as female mice lacking only PRA fail to exhibit mechanism is lost in the αERKO. The second pathway is a uterine decidual response, illustrating the importance initiated by decidualization-associated signals and is ERα of PRA for establishing successful pregnancy.176,178,180 independent and retained in the αERKO. This indicates a Recently, Franco et al. demonstrated that the epithelial redundancy of regulatory mechanisms that allows reten- PR is responsible for establishing successful pregnancy tion of gene regulation in the αERKO and accounts for as the deletion of uterine epithelial PR contributes to an the ERα-independent uterine decidualization response impaired embryo attachment, implantation, and decid- observed in ERα-null mice. Recent studies demonstrated ual response.179 Details of implantation and decidualiza- that de novo E2 synthesis occurs locally during decidual- tion mechanisms are discussed in Chapter 38. ization, especially on day 6 and 7 of pregnancy in mice,273 indicating an important role for local E2 synthesis for Maintenance of Progesterone Action in the Absence establishment of successful pregnancy. of ER The Pgr gene is a well-described target of estrogen- Androgen Receptor Signaling in Uterine induced expression via the classic model of ER action, Function especially in the uterus.32,33 The lack of E2-induced increases in Pgr expression in ERα-null uteri confirms AR Expression in the Uterus the regulatory dependence on ERα action.189 Therefore, AR is present in uterine tissues of multiple spe- it was hypothesized that disruption of the ERα gene cies,130,274–278 although the function of androgen sig- may subsequently result in abnormally low levels of PR naling in the uterus remains unclear. In rodents, ARs in ERα-null uteri and thereby render this tissue refrac- are present in all uterine cell types but most highly tory to P as well. However, assays for Pgr expression and expressed in the myometrium, where expression may be P binding in ERα-null uteri indicate PR levels are only positively regulated by estrogens.276,279 In humans, ARs reduced by approximately half.110,161,189 Furthermore, a are also detected in the myometrial and endometrial greater proportion of PR is nuclear localized in ERα-null uterine tissues, and levels increase during the prolifera- uteri (≈25%) relative to wild-type (≈5%).161 Western blots tive phase.280 indicate no difference in the relative levels of PRA and PRB between genotypes, with PRA consistently pres- MICE LACKING AR ent in greater amounts in both.161 Therefore, a loss of There are several reported lines of AR-null as well as ERα action in the uterus has not led to a complete lack mice with knock-in of mutated forms of AR (reviewed in of PR or to altered and preferential transcription from Refs 281,282). Tfm mice were first described in 1970 and one of the two Pgr gene promoters. Furthermore, sev- are a naturally existing androgen-resistant mutant283 eral P-dependent actions are preserved in ERα-null uteri, with an inactivating mutation of the Ar gene.284,285 including P-induced expression of amphiregulin and Comparable inactivation mutations of the AR gene and calcitonin and the uterine decidual response,161 but not resulting phenotypes are well described in rats and embryo attachment or implanation.135 humans.286 Because the Ar gene is located on the X chro- The postovulatory nadir in circulating E2 levels has mosome and Tfm males are infertile, it is impossible to long been known to be critical to uterine decidualization breed for XX female mice that are homozygous for the

4. FEMALE REPRODUCTIVE SYSTEM 1128 25. STEROID RECEPTORS IN THE UTERUS AND OVARY mutation. Lyon and Glenister overcame this challenge E2 synthesis in the ovaries rather than a role for AR in more than 30 years ago by utilizing embryo aggrega- maintaining uterine weight (Table 25.3).183 Furthermore, tion to generate a limited number of Tfm chimeric males AR-null females are able to establish and maintain preg- that were fertile and carried germ cells harboring the Ar nancies to term.183 Microarray studies have indicated mutation.287 More recently, AR-null female mice were that DHT leads to a pattern of uterine gene regulation generated via a Cre/loxP targeting scheme that allows for that is remarkably similar to that elicited by estrogens tissue and temporal specific disruption of the Ar gene but less robust,292,293 with 86% of the DHT response and therefore the generation of fertile male carriers of consisting of a subset of estrogen responses. The fold the targeted Ar allele.185,288–290 Exon 2 of the murine Ar response of the overlapping genes was in general more gene was targeted for deletion in nearly all lines that robust after estrogen treatment versus DHT. Thus, have been described in the literature.183,184,289,290 Addi- despite differences in biological outcomes, the global tionally, one line of mice with in-frame deletion of exon genomic patterns elicited by estrogens and nonaroma- 3 of the Ar gene has been generated, however, the uter- tizable androgens largely overlap. Genes noted included ine phenotype of this line was has not been described.185 those involved in metabolism, tissue growth and remod- The exon 2 AR-null females have normal uterine devel- eling, transcription, protein synthesis and processing, opment, although the uterine circumference is smaller and signal transduction.293 Suppression of AR expres- than that of wild-type uteri (Table 25.3).183,184 Hu et al. sion in human endometrial stromal cells (HESCs) using demonstrated that the AR-null uteri had increased uter- siRNA led to a decrease in decidual cell proliferation and ine horn length, but smaller uterine diameter, as a result differentiation.295 These findings suggest that AR plays a of decreased endometrial and myometrial areas dur- minimal role for uterine growth and differentiation but ing diestrus, when compared to wild-type uteri.183,184 is not essential for uterine development. These AR-null females were subfertile, due to impaired folliculogenesis (discussed later in the section Ovarian Glucocorticoid Receptor Signaling in Uterine Phenotypes in Mouse Models of Disrupted Androgen Function Signaling). Still, studies to date indicate that AR-null females exhibit reproductive phenotypes that are quite GR Expression in the Uterus similar to those originally described in Tfm/Tfm females. The uterus is not considered a classic target tissue of Recently, a mouse line was developed in which muta- glucocorticoid action, although GRs are present through- tion of the AR DNA binding domain alters the ability out the cell types of the rodent uterus, including uterine of AR to bind to selective androgen response elements. natural killer cells (uNK cells).40,296 This mouse line was generated by Schauwaers et al. with a targeted mutation of 12 amino acids in the second zinc- MICE LACKING GR finger of the DNA binding domain at the exon 3 of AR Homozygous GR-null animals generated via homolo- (called SPARKI).186 The heterozygous or homozygous gous recombination in embryonic stem cells die at birth SPARKI females do not have an overt phenotype and due to respiratory failure.297 Therefore, the generation have normal fertility,186 indicating selective AR DNA of tissue-specific GR deletion in female reproductive tis- binding is not critical for female reproductive organ sues is crucial for studying the role of GR in reproductive development and function. functions. Treatment of hypophysectomized rats with DHT is known to cause increased uterine weight in rodents, Uterine Functions of GR indicating that ligand-dependent AR actions can affect The limited experimental data available indicate that uterine responses.291 Several studies also showed that GR-mediated glucocorticoid actions are present in the DHT induced uterine weight increase in αERKO ani- uterus.40 Treatment with the GR agonist, dexametha- mals.292,293 Upregulation of AR expression is observed in sone, did not increase the weight of immature rat uterus αERKO uteri after DHT treatment, suggesting that DHT compared to E.40,298 However, dexamethasone inhib- exerts its action in the uterus through AR in the absence its the uterine weight increase induced by estrogen in of ERα.292 In addition, AR agonists increased myometrial the rodent uteri.40,298,299 Surprisingly, stimulation of the thickness in the uteri of ovariectomized rats but inhib- GR signaling pathway by dexamethasone treatment in ited estrogen-induced epithelial proliferation.264 Similar immature rats elicits a pattern of gene expression that observations were made in female mice, which exhib- is remarkably similar to, but not as robust as, estro- ited not only uterine epithelial cell proliferation but also gen.40 However, the molecular mechanism by which myometrial smooth muscle cell proliferation after DHT dexamethasone inhibits uterotropic action of estrogen treatment.294 AR-null female mice exhibit relatively nor- remains unclear. A recent study reported that E-induced mal uteri that are somewhat hypoplastic, although this human uterine leiomyoma cell proliferation is sup- phenotype may be more representative of decreased pressed by GR activation.41 Gilz (glucocorticoid-induced

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS IN UTERINE FUNCTION 1129 leucine zipper) is crucial for immune-related function of granulosa cells of Dicer1fl/fl;Amhr2Cre/+ compared to glucocorticoid activity.300 Whirledge and Cidlowski also wild-type ovaries.312 Moreover, Dicer1fl/fl;Amhr2Cre/+ demonstrated that E suppressed Gilz gene expression females exhibit an aberrant oviductal morphology induced by dexamethasone in human uterine epithelial with enlarged and fluid-filled cystic oviducts.311–313 No (ECC1) cells.300 This suggests that estrogen and gluco- embryos were found in Dicer1fl/fl;Amhr2Cre/+ uteri, but corticoid exert interplay between immune-responsive were discovered to be retained within the oviduct.312,313 functions in the uterus. The oviducts from Dicer1fl/fl;Amhr2Cre/+ females have an GR activity not only exhibits inhibitory effects augmented inflammatory response seen as an increase on estrogenic action in the uterus, but may also play in recruitment of lymphocytes and macrophages into important roles in the decidual response, as it is known the oviductal tissues.313 Moreover, Dicer1fl/fl;Amhr2Cre/+ that GR is expressed in uterine natural killer (uNK) females are also defective in embryo transport as trans- cells,296 and that the uNK cells play important roles ferred blue beads were mostly retained in the Dicer1fl/fl; for establishment and maintenance of successful preg- Amhr2Cre/+ oviduct, whereas in WT females the beads nancy (reviewed in Ref. 301). However, the role of GR moved through the oviduct and were found to be in the in uNK cell activity in the uterus remains unclear. uterine lumen.313 A lack of a proper embryo transport in Dicer1fl/fl;Amhr2Cre/+ appears to be due to a uterotubal MicroRNAs in Uterine Reproductive Functions junction defect, resulting in retrograde uterine flow into the oviduct.313 Additionally, the expression of Wnt/β- MicroRNA (miRNA) are very small RNA molecules catenin signaling molecules was altered in the Dicer1fl/fl; (18–24 nt) that are transcribed as larger primary RNAs Amhr2Cre/+ oviduct.311,313 A recent finding from a uter- (priRNA), either within introns of encoding transcripts ine-specific Dicer deletion using PgrCre/+ and Dicerfl/fl or from miRNA encoding promoters. priRNA is pro- animals demonstrated that the females are infertile.314 cessed to a shorter stem and loop pre-miRNA form by Loss of uterine dicer leads to uterine developmental the activity of the DGCR8/DROSHA microprocessor defects, including absence of glandular epithelium and complex. The pre-miRNA is then exported from the increased apoptosis in the stromal cells.314 Stromal cell nucleus by exportin 5, to the cytoplasm where DICER proliferation could not be induced in Dicer1fl/fl;PgrCre/+ cleaves the loop, leaving miRNA, which then functions uteri treated with estrogen and P.314 A decidualization to interact with mRNA, targeting it for degradation or defect was also found in Dicer1fl/fl;PgrCre/+, indicat- repressing translation.302,303 miRNA is particularly criti- ing loss of uterine response to P.314 Additionally, loss of cal to female reproductive tract function, as mouse mod- uterine dicer expression also leads to a dysregulation in els with deletion of DICER in these tissues lose fertility. Wnt/β-catenin signaling molecules, similar to the find- Estrogen has been demonstrated to regulate levels of ings reported in Dicer1fl/fl;Amhr2Cre/+ females. These miRNA in uterine tissue.304,305 findings together suggest that miRNAs, produced by The biological roles of Dicer are crucial for a nor- Dicer processing, play pivotal roles in female repro- mal development of limbs,306 muscle,307 and lung.308 In ductive functions needed for establishing successful addition, recent findings demonstrated that expression of pregnancy. Dicer is not only critical for female germ line biologi- 309 310 cal functions but also for embryo implantation. Changes in Uterine Gene Expression Although miRNAs are expressed in the female reproduc- tive tract, their precise physiological function remains The dramatic physiological changes that occur in the unclear. To study the role of dicer and miRNAs dur- uterus in response to steroid hormones are presumably ing female reproduction, several studies used Dicer1fl/ the ultimate effects of equally dramatic changes in gene fl crossed with Amhr2Cre/+ animals to generate a specific expression among the uterine cells. It is unlikely that the deletion of dicer in the mesenchymal layer of female E2–ER complex is directly involved in mediating the reproductive tract (called Dicer1fl/fl;Amhr2Cre/+).311–313 whole genomic response in the uterus but more plau- Female mice with a lack of Dicer1 expression in the mes- sibly serves to stimulate a cascade of downstream sig- enchyme exhibit a complete infertility.311–313 Loss of Dicer1 naling pathways that act to amplify the estrogen action. has no effect on uterine development as the uterine epi- However, early investigations of the genomic response thelial cells, glands, stroma, and myometrium are pres- to estrogens in the rodent uterus discovered a handful ent in Dicer1fl/fl;Amhr2Cre/+ females.311,313 Uterine horns of genes that are directly regulated via the classic ER of Dicer1fl/fl;Amhr2Cre/+ females are shorter and hypo- mode of action, including Pgr and Ltf. The uteri of ERα- plastic when compared to control (Dicer1fl/fl) uteri.311,313 null females fail to exhibit estrogen-induced increases in Additionally, the Dicer1fl/fl;Amhr2Cre/+ uteri respond nor- Pgr and Ltf expression,174 indicating the importance of mally to an artificial decidual stimulation.312 An ovarian ERα to this response. Furthermore, estrogen- stimulated defect was observed as increased apoptosis within the increases in PR in the rodent uterus are localized to

4. FEMALE REPRODUCTIVE SYSTEM 1130 25. STEROID RECEPTORS IN THE UTERUS AND OVARY the stromal and myometrial compartments, whereas Microarray analyses have also been used to map the increased Ltf is limited to luminal and glandular epi- uterine response to P and have identified numerous tar- thelium,223 indicating that ERα functions are critical to get genes, including those involved in immune function, induced gene expression in multiple uterine cell types. metabolism, growth factor regulation, signal transduc- Further complexity of estrogen action in the uterus is tion, and extra- and intracellular structure.326–328 These illustrated by the simultaneous induction of PR expres- observations have identified numerous signals and sion in the myometrium and stroma while eliciting a mechanisms important for normal uterine function and decrease in PR levels in the luminal epithelium.223 Simi- involved in uterine cancer (Table 25.5).34 larly, the c-fos gene is rapidly induced by estrogen in the Microarray approaches have been utilized to inves- uterus, and this is ERα dependent.159,242 tigate the molecular responses to xenoestrogens, such Microarray analysis of gene expression has signifi- as the industrial chemical bisphenol A (BPA) and the cantly advanced understanding of genomic response of pesticide metabolite 2,2-bis(p-Hydroxyphenyl)-1,1,1- the rodent uterus to E2. Numerous studies have used trichloroethane (HPTE). Microarray transcript profiles microarray techniques to map the global gene expression reveal that both xenoestrogens induce responses that are patterns after estrogen exposure in the uterus and largely highly correlated to E2 response early (2 h), but less cor- demonstrate that the biphasic uterine response to estro- related later (24 h), similar to a pattern seen with a weak gens, so well characterized by physiological indicators estrogen such as estriol.336 Accordingly, BPA and HPTE, discussed earlier (Table 25.4), is mirrored by the global like estriol, are unable to induce uterine growth to the changes in gene expression (Figure 25.13).242,256,315–320 same degree as E2.336 The clearly defined patterns of early and late response genes found in mouse uterine tissues are completely lacking in ERα-null uteri.159,242 The identified genes fall into functional groupings, including signal transduc- TABLE 25.5 Steroid Receptor Uterine Target Genes from Microarray Studies tion, gene transcription, metabolism, protein synthesis and processing, immune function, and cell cycle. Sur- Steroid Targets References prisingly, the expression levels of a striking number of P Ihh, Hoxa10, FoxO1, Bmp2, FKBP52, Cebpb, 232,245,326,328–331 genes are repressed by estrogen in the mouse uterus, Wnt pathway, Calbindin 9k, Pparg, Lox and these effects were either absent in ERα-null uteri or 12/15,Mig-6, Ihh, Klk5, Klk6, Bcl2, Fst relieved by co-treatment with ER antagonists, indicating E cell cycle regulators CcnB1+2, Cdc2a, p21, 199,242,248,332–334 α 159,242 that ER is also actively involved in this process. Mad2l2, Cdc25c, Gadd45g, DNA replication Comparative gene array analyses have also been con- licensing MCM2, Aqp, thioredoxin ducted on human endometrial tissues during the prolif- pathway, Mig-6, Wnt signaling, Cyr61, erative and secretory phases of the menstrual cycle and RAMP3, Aqp5, Bop1/Scx, Muc1, Inhbb, have revealed gene expression patterns that are similar T ATM/Gadd45g pathway 335 to those described in rodents.321–325 Dex Ad1, Mgla, Ykt6, Tis11, Id3, Gilz 40,300

FIGURE 25.13 Top: Schematic representation of uterine biphasic responses that occur following a single injection of E into an ovariectomized mouse (see also Table 25.4). Bottom: Transcriptional profile of uterine RNA, fold changes of transcripts in comparison of vehicle-treated animals at indicated times after E injec- tion. White and black boxes highlight transcripts that typify early and late phases, respectively.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1131

Whole transcriptome analyses are now routinely incor- cells351 and LnCAP prostate cancer cells.352 GR binding porated into studies of disruptions in signaling pathways has been evaluated in mouse liver tissue,353 mouse epi- underlying uterine phenotypes of mouse models such as thelial cells,354 and A549 lung cancer cells.355 We have those described in Table 25.3. Thus, microarray compari- learned that NRs bind to thousands of sites within the sons have now become just one of many tools employed cellular chromatin, and that not all potential HREs in for investigation of uterine functions. The data are rou- every cell demonstrate cognate NR binding. Rather, it is tinely deposited in the publically accessible site GEO apparent that chromatin exhibits “pre-opened” regions (Gene Expression Omnibus; http://www.ncbi.nlm.nih. destined to recruit NR.356 For ER in MCF7, FoxA1 gov/geo/). Table 25.5 lists examples of steroid-regulated establishes ER and AR accessible regions; for other cells uterine transcripts discovered in microarray studies. or other NRs this function might be mediated by other “pioneers”, such as AP1 for GR.354 The accessible chro- Chip-seq matin regions are co-localized within nuclear “hubs” that seem to optimize frequency of interaction with More recently, technologies to facilitate evaluation of NR.356 ChIP-seq is also used to locate other molecules sites of transcription factor interaction with chromatin involved in chromatin remodeling and transcriptional (by enriching a DNA binding protein, such as ERα, that regulation, and to examine activating or repressive his- has been cross-linked in situ to chromatin, with immu- tone modifications or “marks”. These maps of relative noprecipitation [chromatin immunoprecipitation or locations and dynamics of NR and chromatin compo- ChIP], followed by hybridizing the associated DNA to a nents greatly enhance our understanding of hormone chip tiled with promoter region sequences [ChIP-Chip] response mechanisms.339–342,345 or by “next generation” massively parallel sequencing [ChIP-seq]) have been developed and widely utilized to study sites of nuclear receptor interaction (Figure SEX STEROID RECEPTORS AND 25.14).75,337–341 Initial studies focused on ERα binding in OVARIAN FUNCTION MCF7 breast cancer cells, and several similar studies fol- lowed, which are summarized and compared in several The functions of the sex steroids and their cognate review articles,97,342–345 and reported that most sites were receptors in ovarian function are especially complex distal from transcriptional start sites (TSS) or were in and multifaceted. P, T, and E2 are all synthesized intronic regions, rather than adjacent to TSS, as models and secreted by the ovary during folliculogenesis of NR regulation of target transcripts had hypothesized. and act via both extraovarian (i.e., endocrine) and These comprehensive maps of cis-acting transcriptional intraovarian (i.e., para-/autocrine) pathways to pro- regulators have been dubbed “cistromes”. The initial foundly influence all aspects of ovarian function. The ERα cistrome–associated sequences were evaluated for endocrine actions of the sex steroids in the hypotha- enrichment of transcription factor motifs, and confirmed lamic–pituitary axis are critical to the regulation of binding to the experimentally defined “ERE” sequence. gonadotropin secretion and the ovarian cycle and are In the case of the MCF7 cells, enrichment of motifs for described in more detail in other chapters of this book. forkhead binding factors (Fox) was apparent. Owing to In turn, our appreciation of the extent and importance the abundant expression of the FoxA1 member of the Fox of the para-/autocrine actions of sex steroids within family, a potential role for FoxA1 in estrogen response the ovary has increased substantially over the past was pursued with an arsenal of bioinformatic, Next Gen years. Herein, we specifically review the literature sequencing, and biological studies that demonstrated concerning the expression of the sex steroid receptors the ability of FoxA1 to “pioneer”, and thus make acces- (Table 25.6) in the mammalian ovary and employ the sible, regions of the chromatin that were subsequently described null and mutated mouse models that have targeted by ERα.76 been developed as a platform to focus on recent rev- Most ERα cistromes reported have utilized in vitro elations concerning the intraovarian roles of steroid cultured cell models, although ERα profiles from signaling in ovarian function. A caveat to experimen- mouse liver tissue have been obtained.346 A ChIP-seq tal testing of steroid action and steroid receptor stud- study by Hewitt et al. examining ERα binding sites ies in a tissue such as the ovary is that the organ is in mouse uterine tissue indicated that, much like the also producing the steroid being studied. Androgen, MCF7 breast cancer study, most ERα sites were not estrogen, and progesterone are all synthesized by proximal to TSS.243 Subsequent comprehensive ChIP- the ovary; therefore, removal and replacement stud- Chip or ChIP-seq studies have been published for the ies to analyze response are impossible. Thus use of PR in mouse uterus,347 and T47D breast cancer and leio- mutant or null mice allows evaluation of the pheno- myoma cells348 and for the AR in mouse epididymis349 type reflecting loss of functional interactions between skeletal muscle myoblasts,350 ZR-75-1 breast cancer steroids and their receptors.

4. FEMALE REPRODUCTIVE SYSTEM 1132 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

FIGURE 25.14 (A) Schematic (Source: Modified with permission from Ref. 337), shows method to evaluate the “cistrome” of a steroid recep- tor or other chromatin interacting factors using ChIP–Chip or ChIP-seq. Cells or tissues are treated with a chemical, often formaldehyde, to cross-link DNA binding proteins to DNA. Then, chromatin (DNA and attached proteins) is isolated and fragmented with sonication or another method to break the chromatin into small pieces (approx. 500 bp). The protein of interest is immunoprecipitated with an antibody to enrich DNA fragments bound to the protein. Cross-linking is reversed to allow the enriched DNA to be purified. For ChIP–chip, DNA is amplified and labeled and hybridized to a chip spotted with promoter regions. For ChIP-seq, the DNA is sequenced using massively parallel (Deep) sequencing. The sequence “reads” are mapped onto known genomic regions, resulting in peaks indicating regions of binding of the immuno- precipitated DNA binding protein. Both methods are normalized in comparison to “input” chromatin that has not been immunoprecipitated. (B) A screenshot showing an example from ERα and RNA polymerase II cistromes in the vicinity of the Fos transcript from a mouse uterine dataset243 is shown.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1133

Estrogen Receptor Signaling in Ovarian Function us to consider a greater potential role for ERα/ERβ het- erodimers in the ovaries of large animals and primates Estrogen Receptor Expression in the Ovary that may not occur in rodents. The more restricted There is a plethora of evidence for ER expression in expression pattern of ERα and ERβ in rodent ovaries somatic cell types of the mammalian ovary, and numer- does not exclude possible cooperative action between ous reviews describe the importance of ERα and ERβ in the two receptors, as suggested by the unique ovar- ovarian function. As early as 1969, Stumpf demonstrated ian phenotypes in compound ER-null mice compared the specific uptake of [3H]-E2 in rat ovaries using dry- to each single knockout ovarian phenotype (Table mount autoradiography.357 Richards later demonstrated 25.7).158,174 Although much has been learned concern- that the predominance of high-affinity E2 binding sites ing ER localization, it remains difficult to delineate in the rat ovary is found in the granulosa cell compo- which of the incongruent findings among species are nent and that these levels are regulated by estrogen truly representative of divergent expression patterns or and gonadotropins.358,359 Successful cloning of the are more attributable to disparities in techniques and individual ER genes from multiple species and contin- immunoreagents. This is further confounded by evi- ued development of better ER-specific immunoglobu- dence that levels of ER mRNA do not always correspond lin allow for more detailed characterization of ERα with the levels of immunoreactive protein within the and ERβ expression and regulation in the mammalian ovarian compartments.360 The following section reviews ovary. It should be pointed out that immunodetection the literature concerning the expression patterns of ERα and expression patterns of steroid receptors are depen- and ERβ in the ovaries of rodents, domestic animals, and dent on the quality of the antibodies used. Such anti- primates, and the changes in expression patterns that bodies often vary between studies of ovarian tissue as occur during folliculogenesis. Altered expression has described following and may explain both species and been implicated in diseases affecting ovarian function tissue differences. (including polycystic ovarian syndrome and ovarian ER expression patterns are well conserved among cancer); however, this will not be addressed. mammalian species, although distinct differences are α apparent in late stage follicles. A high level of ERβ Estrogen Receptor expression in granulosa cells of growing follicles is RODENTS conserved among rodents, large animals, and primates. ERα is localized to the ovarian interstitial/stroma, ERα expression is exclusive to thecal/interstitial cells thecal cells of growing follicles, and surface epithe- in the stroma of the ovary in rats and mice (Table 25.6, lium in rat129,130,274,366–372 and mouse124,373 ovaries. This mice) but this compartmental expression pattern does expression pattern is not evident in hamsters374 where not hold true in hamsters, domestic animals, monkeys, both receptors are detectable to varied degrees through- or humans (Figure 25.15). In fact, current data indi- out the different somatic cell types of the ovary. In the cate that granulosa cells of preovulatory follicles in fetal rat, ERα expression is detectable shortly after the large animals and primates express comparable levels first indications of gonadal differentiation.375 Postnatal of both ERα and ERβ, and, in humans, ERα may pre- rat and mouse ovaries exhibit ERα mRNA shortly after dominate.360 Similarly, thecal cells in large animal and birth, but levels remain relatively constant; whereas human ovaries possess both ER forms. These data force ERβ levels rise substantially during this period.124,376 Immunoreactivity for ERα in neonatal rat ovaries indi- α TABLE 25.6 Localization of Steroid Receptor Expression in the cates ER expression is limited to thecal/interstitial Mouse Ovary cells and the ovarian surface epithelia and absent in granulosa cells and oocytes.366,375 A similar pattern of Steroid Receptor Granulosa Theca Oocyte Stroma ERα expression occurs in fetal hamster ovaries, which ERα + +++ – ++ exhibit immunoreactivity as early as gestational day 14 and significant increases thereafter during the neonatal ERβ +++ – – – period.377 Yang et al. report that a noticeable increase PRA +++ + – ++ in ERα immunoreactivity occurs in the granulosa cells PRB + + – ++ and oocytes of primordial follicles during neonatal days 8–15.377 AR +++ ++ + +/− In adult rats and mice, ERα expression continues to GR +++ ++ – ND be exclusively localized to thecal cells of growing fol- Aromatase +++ – – – licles, interstitial/stromal cells, and the ovarian surface epithelium129,130,367,369,370,372,378,379 (Figure 25.15). Electro- ND: Not determined. –: Not expressed. phoresis mobility shift assays (EMSA) of nuclear extracts +/−, +, ++, +++: Intensity of signal indicated by number of + symbols. from rat granulosa cells indicate that virtually all of the

4. FEMALE REPRODUCTIVE SYSTEM 1134 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

TABLE 25.7 Ovarian Phenotypes in E Signaling Mutant and Null Mouse Models

Mutated or Null for Sex Steroid Signaling Ovarian Phenotypes Hormone Levels References

Esr1−/− (homozygous null alleles for Anovulatory and infertile Elevated T, E2 and LH 157–160 ERα animal) Hemorrhagic and cystic pathology Normal FSH & P No CLs present in ovarian sections Increased expression of steroidogenic enzymes Reduced response or failure to respond to exogenous gonadotropins

NERKI+/− (one mutated allele with Anovulatory and infertile Normal LH, FSH and E2 132 2-point mutation in DNA binding No plugs observed in NERKI females after Reduced P zinc finger of ERα and one WT allele) superovulation and natural mating Superovulation was able to partially restore ovulation in NERKI mice, however also increased hemorrhagic follicles in the ovaries

KIKO (ERAA/−) (one mutated allele Anovulatory and infertile Normal E2 and P 162 of 2-point mutation in DNA binding No CLs present in ovarian sections domain of ERα and one ERαKO allele)

ENERKI (ERαG525L) (homozygous Anovulatory Elevated serum E2, T and LH 168 animal of 1-point mutation in LBD Hemorrhagic and cystic ovarian pathology Normal FSH of ERα) Increased number of atretic antral follicles and no CLs Hyperplastic theca cells in response to LH (data not shown)

ERαEAAE/EAAE (homozygous animal Infertile Not reported 164 of 4-point mutation of DBD ERα) Hemorrhagic ovaries

AF2ERKI/KI (homozygous animal of Anovulatory and infertile Elevated serum LH and E2 169 2-point mutation in LBD of ERα) Hemorrhagic and cystic ovarian pathology No CLs present

Cyp17cre;ERaflox/flox Fertility normal in young mice, but 6 month old Elevated T at both 2 & 361,362 (Theca cell specific ERα knockout) animals have reduced fertility and longer estrous 6 months cycle Normal FSH Decreased LH at 2 months with further decrease at 6 months

Esr2−/− (homozygous null alleles Subfertile although some lines are infertile Normal LH and FSH 158,171–173,363 for ERβ: βERKO, Ex3βERKO, and Reduced (or lack of) response to exogenous β L−/L− **ER ST ) gonadotropins Approximately 1/5 the number of large preovulatory follicles after superovulation compared to WT mice Lack of COC expansion

ERαAF-10 Ovarian phenotype not reported Unknown 165

ERαAF-20 Ovarian phenotype not reported Unknown 167

αβERKO Anovulatory and infertile Elevated LH and T 158,174 No CLs and few large follicles Normal FSH and P Ovarian transdifferentiation to Sertoli-like cells that express Sox9 Altered expression of steroidogenic enzymes

Cyp19a1−/− Anovulatory and infertile No E2 131,156,175,194,364,365 Hemorrhagic and cystic pathology Elevated LH, FSH and T No CLs present in ovarian sections Poor response to exogenous gonadotropins, although treatment with E2 in addition to gonadotropins improves ovulatory response Ovarian transdifferentiation to Sertoli-like cells that express Sox9

β L−/L− β **ER ST females are the only line of ER knockout animals that are sterile.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1135 specific ERE-bound complexes are supershifted by anti- is strongest among those in proximity to the forming ERβ but not anti-ERα immunoglobulin, providing fur- antrum.374 FSH treatment (twice a day for 2 consecutive ther support that rat granulosa cells possess very little days) or a single injection of E2 elicits a significant induc- ERα.367,371 Western blot analyses of whole cell or nuclear tion of ERα expression in both thecal and granulosa cells extracts from isolated rat granulosa cells do indicate a in hypophysectomized hamsters.374 Still, Western blot low level of ERα protein of 61 kDa,371 and use of laser analyses of whole ovarian homogenates from hamsters capture microdissection (LCM) shows some expres- indicate an ERβ:ERα ratio of 14:1 during the follicular sion of ERα mRNA in granulosa cells (Figure 25.15).363 phase, indicating that ERβ predominates. However, the Unlike ERβ, ERα expression remains relatively constant rapid decline in ERβ and concurrent increase in ERα that throughout the rat estrous cycle.380 Furthermore, no occurs just prior to and shortly after the gonadotropin change in ERα levels is detected in granulosa cells cul- surge adjusts this ratio to 2:1, suggesting that ERβ plays tured in FSH/T over a period of 72 h; however, a 24-h a more predominant role in follicular growth and ERα treatment with forskolin induces a marked increase in is more important during luteinization374 in the hamster ERα immunoreactivity in the nuclei.371 ovary. ERα localization in the adult hamster ovary is some- what divergent from that in rats and mice. Yang et al. DOMESTIC ANIMALS report moderate clear ERα immunoreactivity in the- In fetal bovine ovaries, ERα was shown to be local- cal/interstitial cells, but also appreciable immunoreac- ized in all cell types involved with follicle development, tive granulosa cells of small preantral follicles.374 In the including granulosa cells, oocytes, and epithelial cells granulosa cells of antral follicles, ERα immunoreactivity throughout development, with expression becoming

FIGURE 25.15 (A) Immunohistochemistry for ERα and ERβ in adult mouse ovary. Immunohistochemistry for ERα (top) indicates specific nuclear immunoreactivity in thecal/interstitial cells and thecal cells (TC) in and around a preantral follicle. Granulosa cells lack any measurable immunoreactivity for ERα. The staining around the outer surface of the oocyte is nonspecific and not representative of ERα immunoreactivity. Immunohistochemistry for ERβ (bottom) indicates specific nuclear immunoreactivity in the granulosa cells (GC) throughout a preantral follicle. Thecal cells (TC) lack any measurable immunoreactivity for ERβ, but the surrounding thecal/interstitial cells exhibit some cytoplasmic staining. Immunohistochemistry was done by Dr. Madhabananda Sar (see Ref. 366 for protocol). (B) Pure populations of granulosa and theca cells were isolated from large preovulatory follicles using laser capture microdissection (LCM), and RNA was reverse transcribed and real time PCR was performed using primers specific for ERα and ERβ.

4. FEMALE REPRODUCTIVE SYSTEM 1136 25. STEROID RECEPTORS IN THE UTERUS AND OVARY more localized to thecal and some stromal cells during detectable in the granulosa cells of less mature follicles. the later stages of development.381 In adult bovine ova- In support of this explanation is a report by Taylor and ries, ERα immunoreactivity is relatively high in thecal Al-Azzawi in which ERα immunoreactivity is illustrated cells of secondary and tertiary follicles, substantially in the granulosa cells of what is clearly a large antral weaker in granulosa cells of tertiary follicles, and totally follicle when using antisera raised against the bovine absent in primordial, primary, and secondary follicles.382 ERα.394 Indeed, Saunders et al. described ERα immuno- In agreement, Barisha et al. found ERα transcripts are reactivity that is limited to granulosa cells of large antral detectable in both granulosa and thecal cell fractions, but follicles and undetectable in smaller follicles in the levels are substantially higher in the latter cell type.383 human ovary.395 Also, in agreement with Pelletier and Furthermore, thecal ERα expression appears to be dif- El-Alfy,393 Saunders et al. found ERα immunoreactivity ferentially regulated as it is almost four-fold higher in in thecal cells of preantral and antral follicles, and ovar- follicles of 20–180 mm versus those <0.5 mm, whereas ian surface epithelia.395 ERβ expression in thecal cells remains constant among ERα transcripts are detectable in human nonluteinized follicles of different sizes.383 ERα expression in granu- granulosa cells396 and granulosa-luteal cells collected at losa cells, although much lower relative to theca, also the time of oocyte retrieval during in vitro fertilization increased with advanced follicle size.383 In porcine ova- (IVF) procedures.396,397 In fact, Jakimiuk et al. found ERα ries, ERα immunoreactivity is detected in both theca mRNA and protein levels are two- to four-fold higher interna and granulosa cells, but only in large follicles; in granulosa versus thecal cells of small antral follicles and even then ERα immunoreactivity is considerably in human ovaries.360 Furthermore, ERα protein levels moderate relative to ERβ.384 In ovine ovaries, ERα mRNA remain elevated in granulosa cells but decrease in the- and immunoreactivity is detected in both cell types but cal cells of dominant follicles.360 In luteinized human predominates in granulosa cells and is particularly high granulosa cells, Chiang et al. found ERα levels to remain in cumulus oophorus cells of small antral follicles.385 relatively constant over a period of 10 days in culture but always much lower than ERα mRNA levels.398 Recent PRIMATES reports in human ovarian samples collected during elec- The first reported study of ERα immunoreactivity rhe- tive hysterectomy demonstrate that ERα is expressed in sus or cynomolgus monkey ovaries found that ERα was granulosa cells of antral follicles, with weak to no posi- undetectable in all ovarian cell types except the surface tive immunohistochemical staining in luteinized granu- epithelia, regardless of menstrual stage.386 However, a losa cells399 supporting previous data. later study in baboon ovaries using the same antibody β found nuclear ERα immunoreactivity in 30–40% of the Estrogen Receptor granulosa cells of healthy antral follicles, and detectable In the nonpregnant ovaries of most mammalian but less intense staining in granulosa cells of preantral species, ERβ is clearly the predominant ER form and follicles.387 This same study found the stroma, inter- is most often exclusively localized to granulosa cells stitial, and thecal cells to be largely unlabeled.387 Pau of follicles from the primary to preovulatory stage. et al. produced similar findings of ERα expression in This expression pattern is documented in the ovaries the granulosa cells of rhesus monkey ovaries by in situ of rats,14,130,366,367,370–372,380,400–402 mice,124,373,379,402 ham- hybridization.388 sters,374 cows,383,403,404 sheep,405,406 pigs,402,407 nonhu- High-affinity E2 binding sites in normal human ovar- man primates,274,388 and humans.393,394,402,408,409 ian tissue were first described several decades ago,389–391 but these techniques were not able to differentiate RODENTS between the two ER forms. Iwai et al. used anti-ERα In developing rat ovaries, ERβ expression is detectable antisera similar to that in the nonhuman primate stud- shortly after the first indications of gonadal differentia- ies to demonstrate immunoreactivity in the granulosa tion on gestational day 14, and levels increase thereafter cells of antral and preovulatory follicles in human ovary during prenatal development.375 A second, more robust but a total absence in primordial, preantral follicles, increase in ERβ expression occurs during days 10–15 of atretic follicles, and thecal/interstitial cells.392 Pelletier neonatal development in rat ovaries,376 and immunore- and El-Alfy393 used a different antihuman ERα antisera activity is largely localized to granulosa cells of growing and reported results that are in contrast to those of Iwai follicles and notably absent in primordial follicles and et al.,392 i.e., a total lack of ERα immunoreactivity in oocytes.366,375 Developing mouse ovaries exhibit a com- granulosa cells but clear nuclear staining of theca interna parable pattern ERβ expression.124 In contrast, ERα is cells, interstitial gland cells, and ovarian surface epithe- limited to thecal and interstitial cells366,375 and exhibits lia. A possible explanation for this discrepancy may be a little change in levels with increased age124,376 in devel- lack of large antral follicles in samples evaluated by Pel- oping rat and mouse ovaries. The hamster ovary exhib- leteir and El-Alfy since both studies agree that ERα is not its a fairly similar ontogeny as ERβ is detectable as early

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1137 as gestational day 13 and increases thereafter to peak on However, not all granulosa cells of healthy growing fol- postnatal day 10, throughout which ERβ immunoreac- licles in the rodent ovary express ERβ, as negative cells tivity is predominantly localized to granulosa cells but are randomly distributed throughout the follicle.366,367 detectable in some interstitial cells and oocytes.377 Data from Western blots of nuclear extracts from In the ovaries of prepubertal and adult rats and mice, whole ovaries or isolated granulosa cells support the ERβ immunoreactivity is nearly exclusive to the nuclei immunohistochemical data that ERβ is the predominant of granulosa cells of healthy follicles at all advanced receptor form in rodent granulosa cells.367,371 Electropho- stages of folliculogenesis124,129,130,366–370,374,402 and is nota- resis mobility shift assays in which whole cell extracts bly decreased or absent in atretic follicles129,366 (Figure from ovaries or isolated granulosa cells of rats provide 25.15). Primordial follicles, thecal cells, oocytes, ovarian further evidence of the predominance of ERβ.367,371,410 surface epithelium, and luteal cells lack ERβ immunore- Sharma et al. found that antisera raised against resi- activity in adult rat and mouse ovaries,129,130,366–369,374,402 dues 54–71 of rat ERβ produces the optimum results on although some cytoplasmic staining in the latter cell Western blot and detects four immunoreactive bands of type is reported when using certain antisera.129,366 Gran- 58/52 to 46/44 kDa in size, the 58 kDa protein presum- ulosa cell specific expression of ERβ in the rat ovary was ably being full-length ERβ (ERβ1 in Figure 25.2).371 Anti- reproduced using three separate anti-rat ERβ antisera, sera raised against the AF-2 (residues 182–485) domain two raised against residues 467–485 and a third raised of rat ERβ detects a single specific protein of 60 kDa in against residues 54–71.369 Furthermore, independent whole-cell extracts from both rat and mouse ovaries that studies using in situ hybridization produced results con- comigrates with the putative full-length ERβ.367 Choi gruent with the immunohistochemical findings of strict et al.402 and Hiroi et al.368 produced comparable findings localization of ERβ mRNA to granulosa cells of healthy of immunoreactive proteins at 60 and 55 kDa in rat and growing follicles,14,129,375,401 although Bao et al. reported mouse ovarian extracts when using either a monoclonal scattered but specific hybridization for ERβ mRNA in antibody mapped to residues 272–285 of human ERβ thecal cells of healthy medium and large follicles in the or antisera raised against C-terminal residues 467–485 rat ovary.400 of rat ERβ, respectively. In the hamster, a single ERβ- A study by Saunders et al.370 stands in contrast to those immunoreactive protein of 54 kDa is detected.374 To date, just discussed. In this study, strong nuclear immunoreac- ERβ antibodies are not as specific as those manufactured tivity is described throughout the different somatic cell for ERα, and produce bands in the ERβ-null tissues simi- types of the rat ovary, including thecal, interstitial, and lar to those observed in WT tissues (unpublished data luteal cells, as well granulosa cells of maturing follicles from Korach laboratory) such that further characteriza- when using an anti-rat ERβ antisera raised against resi- tion of ERβ localization is challenging. dues 196–213. Yang et al. reported similar findings of Several different ERβ isoforms are present in rodent ERβ immunoreactivity in thecal/interstitial cell prepa- ovaries, notably ERβ2 (Figure 25.2), ERβ-Δ3, and ERβ2-Δ3, rations from hamster ovary using different ERβ-specific the latter being a compound form of the two former vari- antisera.374 Still, the results of Saunders et al. are prob- ants. ERβ-Δ3 is an exon 3 deletion that results in a recep- lematic because they include reports of substantial ERβ tor lacking the C-terminal zinc-finger of the DNA binding immunoreactivity in certain reproductive tissues of the domain and therefore unable to bind an ERE.193 ERβ2 is rat, such as the oviduct and uterus,370 that are otherwise an especially interesting variant because it possesses an thought to possess relatively low or null levels of ERβ insert of 18 amino acids in the N-terminal region of the mRNA and immunoreactivity. Two obvious differences LBD (Figure 25.2) that causes a 35-fold reduction in affin- between the immunohistochemical study of Saunders ity for E2 and a 1000-fold decrease in E2-induced trans- et al.370 and others are the aforementioned use of differ- activational activity in vitro.193 This isoform is not found ent antisera as well as different tissue fixative. in human ovary (Figure 25.2). Pettersson et al.193 specu- The previous discrepant findings notwithstanding, late that ERβ2 may be especially attuned to mediating E2 ERβ is clearly the predominant ER form present in gran- actions within healthy growing follicles where intrafol- ulosa cells of healthy growing follicle in the rodent ova- licular E2 levels greatly exceed that required to activate ries. Recent isolation of pure granulosa and theca cells wild-type ER forms. Differential RT-PCR indicates a 1:1 from mouse ovaries using laser capture microdissection ratio of ERβ1:ERβ2 mRNA in the ovary of adult371,410,411 demonstrates that ERβ mRNA expression is solely in and neonatal rats,323 while transcripts encoding ERβ1-Δ3 granulosa cells (Figure 25.15). Any specific distribution or ERβ2-Δ3 variants are detectable but at substantially of ERβ throughout the subpopulations of granulosa cells lower levels.410,411 Pettersen et al. demonstrated that of growing follicles has been difficult to ascertain. In FLAG-tagged clones of ERβ1 and ERβ2 expressed in general, ERβ immunoreactivity is continuous and strong 293T cells co-migrate at approximately 60 kDa,411 sug- among the mural cells and relatively uniform among gesting that resolution of the two isoforms by Western those layers closer to the antrum and oocyte.366,367,369 blot may be difficult. However, several studies illustrate

4. FEMALE REPRODUCTIVE SYSTEM 1138 25. STEROID RECEPTORS IN THE UTERUS AND OVARY two distinct ERβ-specific bands appear on Western blots in medium sized (275–450 mm) follicles possessing clearly from rat367,368,371,410 and mouse402 ovarian extracts. Fur- defined antrum.400 Most studies agree on the precipitous thermore, O’Brien et al. fractionated rat granulosa cell decline in ERβ levels that occur shortly after the ovulatory preparations in a 7.5% SDS-PAGE from which immu- gonadotropin surge and the low levels that remain during noreactive protein bands presumed to be ERβ1 (60 kDa) estrus in rats380 and hamsters374 (Figure 25.16). In retro- and ERβ2 (62 kDa) were resolved; they further remarked spect, it is evident that the decreasing effect of the gonado- that an equal ratio of ERβ1:ERβ2 transcripts does not tropin surge on ERβ levels was first indicated by Richards correlate to actual protein amount by Western blot, as in 1975, in which a 75% decrease in high-affinity E2 bind- much greater amounts of ERβ1 protein are present.410 In ing sites in granulosa cells of hypophysectomized, FSH/ retrospect, previous [3H]-E2 binding data also supports E2-primed rats was found to occur following a single LH a predominance of ERβ1 versus ERβ2 in rat granulosa treatment.358 This effect of LH on ERβ expression has been cells. Because recombinant ERβ1 and ERβ2 differentially reproduced in vivo in gonadotropin-primed rats367,380 bind E2 with affinities of 0.14 and 5.1 nM, respectively,411 and mice415 in which a single hCG injection led to a rapid one would expect differentiation of the two forms when decrease in ERβ mRNA and protein levels of more than assessed by Scatchard plot analysis. Therefore, the ear- 50% within 6–9 h (Figure 25.16). Similarly, granulosa lier studies using [3H]-E2412 and more recent studies cells isolated from pregnant mares serum gonadotropin using [125I]-17α-iodovinyl-11β-methoxyE2367 that report (PMSG)-stimulated rats and exposed to hCG in vitro a single, saturable, high-affinity binding factor with a exhibit a comparable loss of ERβ protein when assessed 367 371 Kd = 0.4 nM in rat ovarian or granulosa cell extracts are by western blot or EMSA. congruent with ERβ1 as the predominant form present. Therefore, induction of the LH-signaling pathway in Similar data of a single E2 binding component with a differentiated granulosa cells is the primary stimulus for 413,414 β Kd = 1–1.4 nM in hamster ovaries is also reported. rapidly decreased ER expression. Evidence of a direct The majority of studies on the regulation of β expres- role of LH comes from findings that only preovulatory sion during folliculogenesis focus on the rat ovary. While follicles that co-express LH receptor exhibit a decline some report high but relatively constant levels of ERβ in ERβ expression; while smaller, LH-receptor negative mRNA leading up to proestrus,380 more detailed studies follicles maintain ERβ expression as late as 24 h post- indicate a substantial increase in ERβ expression that peaks hCG exposure.367,380 The inhibitory effect of LH on ERβ

FIGURE 25.16 Regulation of ERβ expression in adult rat ovaries during the estrous cycle or after exogenous gonadotropin treatments. (A) Quantification of gene expression from Northern blot analysis (not shown) for ERα and ERβ in adult rat ovaries during the estrous cycle. Band intensities were measured on a phosphorimager and normalized to an S16 internal control for each time point. mRNA levels are shown relative to the level at E1100 (set to 1.0). Sera from animals were used to determine LH concentrations; the onset of the LH surge was observed at 1600 h of proestrous, and the peak was observed at 1800 h of proestrous. Annotation at bottom indicates the estrous cycle stage and hour of tissue collec- tion: E, estrous; M, metestrous; D, diestrous; P, proestrous. (Source: Reproduced with permission from Ref. 380.) (B) Reverse transcriptase polymerase chain reaction was used to quantify ERβ mRNA levels in the ovaries of immature rats after treatment with PMSG alone or PMSG for 48 h followed by hCG. The ratio of ERβ/S16 of control rats with no hormonal treatment was set to 1.0. Shown are the mean ± SE (n = 4). (Source: Reproduced with permission from Ref. 380) (C) ERβ immunoreactivity in rat ovaries treated with vehicle (Veh), PMSG for 48 h, or PMSG for 48 h followed by hCG. ERβ immunoreactivity is detected in granulosa cells of small and large antral follicles in Veh PMSG animals. Stars indicate the location of antra in antral follicles. PMSG-treated rats were injected with an ovulatory dose of hCG and ovaries isolated after 3, 9, 12, or 24 h. The expression of ERβ protein in granulosa cells 9 h after hCG is reduced in large antral follicles (left), greatly reduced in preovulatory follicles (center), but does not change in small antral follicles (right). Similar expression is observed 12 h after hCG administration. One day (24 h) after hCG treatment, ERβ expres- sion is not detected in corpora lutea (small arrowheads) but is highly expressed in preantral (right) and small antral follicles. Magnification 250×. Source: Reproduced with permission from Ref. 367.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1139 expression can be reproduced by exposing differentiated once again this is more likely due to an increased gran- granulosa cells to either forskolin or TPA,380,410 activa- ulosa cell population. However, Tonetta et al. demon- tors of the protein kinase-A and protein kinase-C path- strated that FSH-induced increases in E2 binding sites in ways, respectively, and thought to mimic the effects of granulosa cells of hypophysectomized rats are blocked LH stimulation.416 Furthermore, forskolin or TPA treat- by co-administration of an ER antagonist, suggesting an ment reproduce the same temporal pattern of decreased autoregulatory element for ER expression.418 In contrast, ERβ expression that follows LH or hCG exposure,417 sug- Kim and Greenwald found little change in the number gesting a common mechanism. Guo et al. demonstrated of E2 binding sites in hamster ovaries following E2 or both activators and mimics of the LH-signaling pathway DES treatment, although animals were exposed for only elicit the rapid decline in granulosa cell ERβ levels by 1–2 days, and therefore significant gains in granulosa decreasing the stability of ERβ transcripts rather than via cell number may not have been achieved.413 Drummond a repression of gene expression.417 et al. reported that 1–4 days of treatment with DES has In contrast to the peri-ovulatory decrease in ERβ no discernible effect on the levels of ERβ1 or ERβ2 tran- expression that occurs, there is little known about the scripts in immature rat ovaries.376 Yang et al. found that positive regulation of ERβ expression in rat granulosa hypophysectomized hamsters exhibit an almost six- cells. Induction of folliculogenesis by PMSG or FSH in fold increase in ERβ expression as detected by immu- hypophysectomized rats leads to a significant rise in nohistochemistry 24 h after a single injection of 0.1 mg E2 binding among granulosa cells, presumably repre- E2- valerate.374 Furthermore, rat granulosa cells isolated sentative of ERβ,358,418 but this is likely due more to an from E2-primed versus untreated animals exhibit a four- increased number of granulosa cell population rather to six-fold higher basal estrogenic activity on an ERE- than direct gonadotropin-stimulated ERβ expression. luciferase reporter construct, suggesting that individual The presence of ERβ in the granulosa cells of primary cellular levels of ERβ are increased by E2 treatment.371 follicles, considered to be insensitive to direct gonado- Also, isolated rat granulosa cells exhibit an increased tropin stimulation, supports a minor role for FSH in ERβ nuclear intensity for ERβ immunoreactivity 1.5 h after expression.366,375 Furthermore, ERβ levels exhibit little E2 exposure that is maintained for 24 h but totally lost change in immature rat or mouse ovaries 48 h after a by 48 h.371 Still, the failure of granulosa cells to totally single injection with PMSG,367,380,415 and no reports exist sustain ERβ expression when maintained in FSH and T, of altered ERβ expression in the ovaries of mice null for and therefore capable of synthesizing endogenous E2, is FSH signaling.419–421 Sharma et al. found that ERβ pro- puzzling. It is plausible that E2 regulation of ERβ expres- tein levels in gonadotropin-primed rat granulosa cells sion in granulosa cells may require the actions of ERα in evaluated are highest if assessed after isolation and thecal cells, which cannot be mimicked when granulosa decrease steadily to undetectable levels within 72 h in cells are isolated in culture. Preservation of ERβ expres- culture, even in the presence of FSH and T.371 However, sion in preantral follicles of ERα-null ovaries argues this pattern was not totally reproduced at the level of against this hypothesis.127,373 Recent work in KGN cells ERβ mRNA, which drops by only 55% when cultured suggest that GIOT-4, a cofactor regulated by FSH, may for the first 24 h in the absence of hormones but stabilizes act to increase ERβ expression as well as work in com- upon the addition of FSH and T,371 suggesting that FSH bination with ERβ to regulate target genes in follicles.423 may be more important to the maintenance rather than Although this work was done primarily in KGN ovar- stimulation of granulosa cell ERβ expression. In contrast ian cells, it was also reported that GIOT-4 and ERβ are to the previous findings, preovulatory rat granulosa co-localized in growing follicles providing support for cells allowed to lose ERβ expression following long-term the co-regulation. The expression of GIOT-4 has not been (6 day) culture exhibit a significant rise in levels 48 h examined in in vitro–cultured granulosa cells. Therefore, after exposure to forskolin, a direct PKA activator and although the signaling factors that may be most impor- thought to mimic FSH signaling.371 Interestingly, FSH tant to inducing ERβ expression remain unknown, it is may have a greater positive influence on ERβ expres- clear that the level of ERβ is highly dependent upon the sion in the hamster ovary where a substantial increase in state of granulosa cell differentiation as cells from pri- expression in preantral and antral follicles is observed in mary, growing, and preovulatory follicles exhibit diver- hypophysectomized females following 1–2 days of treat- gent expression patterns and mechanisms of regulation. ment with ovine FSH;374 and ovaries of neonatal ham- sters in which prenatal FSH action is inhibited exhibit DOMESTIC ANIMALS significant decreases in ERβ expression.377 Descriptions of ERβ expression in the ovaries of The existing evidence of E2 regulation of ERβ expres- domestic animals are limited but generally indicate pat- sion is conflicting. Treatment of hypophysectomized terns that are congruent with rodent ovaries. In the fetal rats with E2 for 3–4 days results in steady and dramatic bovine ovary, ERβ is localized in the cell types associated increases in E2 binding sites in granulosa cells,358,422 but with follicle growth, predominantly in pregranulosa

4. FEMALE REPRODUCTIVE SYSTEM 1140 25. STEROID RECEPTORS IN THE UTERUS AND OVARY and granulosa cells. In early development, epithelial hybridization to granulosa cells of follicles at all differ- cells show expression; however, this becomes punctate ent stages of development, as well as substantial labeling as follicles begin to develop.381 In adult bovine ovaries, in theca interna and ovarian surface epithelia.154 Similar Rosenfeld et al. found that ERβ mRNA and immunore- findings are described by Pau et al.388 in granulosa cells activity is restricted to granulosa cells of small and large of large follicles in rhesus monkey ovary. In the baboon antral follicles with no detectable levels in thecal cells.403 ovary, Pepe et al. found ERβ immunoreactivity is abun- A follow-up study further demonstrated substantial ERβ dant in granulosa cells of follicles at all stages, including immunoreactivity in cumulus oophorus and antral gran- large antral follicles but notably lacking in thecal cells.425 ulosa cells relative to mural granulosa, as well as men- Western blots of baboon ovarian homogenates indicate tioned significant ERβ expression in oocytes,404 the latter two specific ERβ protein bands at 55 and 63 kDa.425 An finding being common only to hamsters.374,377 In con- extensive immunohistochemical study in marmoset trast to the immunohistochemical data, Walther et al.424 ovaries using antisera raised against human ERβ found and Berisha et al.383 found ERβ transcripts are detectable extensive immunoreactivity in granulosa cells of pri- in both granulosa and thecal cells. In ovine ovaries, ERβ mary through to mid- and late follicular stage follicles, is highly expressed in granulosa cells of growing folli- including those with a large antrum, but a clear lack of cles,405,406 and levels exhibit little change over the course staining in atretic follicles.395 No specific pattern of ERβ of the estrous cycle.406 However, Jansen et al. report via immunoreactivity was apparent among subpopulations in situ hybridization that ERβ mRNA is greatest in fol- of granulosa cells within any one follicle.395 In contrast licles ≤3 mm and declines in larger follicles during the to baboon ovaries but in agreement with cynomolgus early follicular phase.405 Both reports describe low but monkey ovaries, thecal cells of large growing follicles as detectable ERβ expression in thecal cells of growing folli- well as the ovarian surface epithelia in marmoset ovaries cles in the sheep ovary405,406 as well as in ovarian surface possess appreciable levels of ERβ immunoreactivity.395 epithelia.406 In porcine ovaries, ERβ is almost equally High-affinity E2 binding sites in normal human ovarian expressed among the granulosa and theca interna of tissue were described several decades ago.389–391 Al-Timimi medium and large follicles, as well appreciable detection et al. found 45 of 89 normal ovaries from premenopausal in the oocyte and ovarian surface epithelia.384 Like the women to possess detectable E2 binding sites; whereas rodent, atretic follicles in porcine ovaries exhibit reduced all of the (n = 10) postmenopausal ovaries assayed were ERβ expression.384 LaVoie et al. found that ERβ mRNA devoid of detectable binding.390 Vierikko et al. reported levels in whole ovarian lysates exhibit little change dur- high- affinity E2 binding in a similar percentage of pre- ing the porcine estrous cycle and are relatively equal menopausal ovaries but did find 67% of postmeno- when comparing follicles of 1–5 mm in size.407 pausal samples were also positive, although the levels of An interesting commonality among bovine, ovine, and E2 binding in both are notably lower than that for P.391 The porcine ovaries is the presence of a truncated ERβ iso- advent of better reagents has allowed for differentiation form lacking most of the ligand binding domain, termed of the two ER forms within human ovaries. ERβ expres- ERβΔLBD in the cow424 and ERβ-Δ5 in sheep and pig406,407 sion was first detected in human ovary by Northern blot (Figure 25.2). In all three species, ERβ-Δ5 is a deletion of analysis.15,408,426,427 Follow-up studies using RNase-protec- exon 5 and encodes a receptor isoform that is truncated tion assays or RT-PCR indicate a relatively equal ratio of 15 residues into the LBD and terminates with four to nine ERα:ERβ transcripts in mixed-cell homogenates from ova- heterologous residues depending on the species.406,407,424 ries of pre-and postmenopausal women but exclusively LaVoie et al.407 and Choi et al.402 produced Western ERβ transcripts in luteinized granulosa cells isolated from blot evidence that ERβ-Δ5 may indeed be translated as women undergoing IVF.408,427 Hillier et al. report equal a 35 kDa protein in porcine ovaries. Given that ERβ-Δ5 levels of ERα and ERβ mRNA in both nonluteinized and lacks the LBD, it predictably exhibits little estrogen- luteinized granulosa cells from normal human ovaries, induced transactivational activity when overexpressed although levels of both ER forms are reduced in the latter.396 in vitro but does possess measurable ligand-independent Another study comparing ERβ mRNA and protein levels transactivational activity that is three-fold lower than within isolated thecal and granulosa cells of small antral full-length ERβ, suggesting preservation of AF-1 func- follicles found mRNA levels are two-fold higher in thecal tions.407 ERβ-Δ5 is also able to reduce the in vitro activity cells but protein levels are similar in both cell types.360 of full-length ERβ by approximately 40% when present in Three reports describing immunohistochemical local- excess.407,424 Thus ERβ-Δ5 might modulate ERβ activity in ization of ERβ in human ovaries produced comparable an environment of high hormone concentrations, such as findings of specific immunoreactivity among granulosa is found with estrogens in granulosa cells. cells of follicles from primary to late antral stage, and specific, but considerably lower, levels in thecal cells of PRIMATES preantral and antral follicles.393–395 Furthermore, all three In ovaries of known fertile cynomolgus monkeys, demonstrate substantial ERβ immunoreactivity in human Pelletier et al. localized ERβ expression via in situ ovarian surface epithelia.393–395 Hillier et al. also found

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1141

ERβ transcripts are detectable in primary cultures of mice exhibit relatively normal ovaries except for signs of human ovarian surface epithelial cells;396 however, simi- premature folliculogenesis as indicated by the sporadic lar evaluations in cell lines derived from human ovarian presence of large antral follicles.158,373 Adult ERα-null surface epithelium indicate no detectable ERβ mRNA.427 females are anovulatory and hence infertile, and exhibit There is currently little known about ERβ regulation ovaries that possess normal pre- and small antral-stage in human ovaries. Jakimiuk et al. found ERβ expression follicles, multiple hemorrhagic cysts, absence of corpora is significantly decreased in thecal and granulosa cells of lutea, reduced number of interstitial glandular cells, dominant follicles relative to those at earlier stages, yet ERβ protein levels did not mirror this difference.360 In primary cultures of human granulosa–luteal cells, ERβ is more highly expressed relative to ERα and levels of the former increase 1.6-fold over a 10-day period dur- ing which spontaneous luteinization is believed to occur, while ERα levels remain constant throughout.398 Treat- ment of human granulosa-luteal cells maintained in cul- ture for 7 days with hCG reduces ERβ and ERα expression by almost 50% within 24 h, and this effect is mimicked by activators of protein kinase A (forskolin) or protein kinase C (TPA).398 Recent examination of human ovaries found that ERβ1 and ERβ2 are expressed in granulosa-luteal cells as well as endothelial cells at all stages of the luteal phase. This study also used in vitro luteinized granulosa cells, treated with hCG or E2. Treatment of the lutein- ized granulosa cells with hCG reduced expression of ERα, ERβ1, and ERβ2 by approximately 50%, while treat- ment with E2 reduced expression of ERα and ERβ1 with no significant change in ERβ2,399 suggesting that human granulosa cells may respond to LH by reducing ERβ in a manner similar to that found in rodent ovaries. Studies to date indicate that human tissues express β β 426 a unique C-terminal variant of ER termed ER cx or β 409,428 β human ER 2 (Figure 25.2). ER cx is a 495–amino acid–long receptor form in which the 61 C-terminal resi- dues are replaced by a heterologous string of 26 amino FIGURE 25.17 Ovarian phenotypes in ER-null mice. (A–C) Shown 426,428 acids encoded by an alternative downstream exon. are cross-sections from representative adult ovaries of wild-type (A), ERβ- β α β Transcripts encoding ER cx are detectable in multi- null (B), and ER -null (C) female mice. Wild-type and ER -null ovaries ple human tissues but most especially in ovary, testis, each exhibit all stages of folliculogenesis except corpora lutea, although β α thymus, and spleen and are often at levels equal to or large antral follicles are sparse in the ER -null ovary. In contrast, ER - null ovaries are characterized by several large, hemorrhagic, and cystic greater than wild-type ERβ.426,428 Immunohistochem- β follicles; a sparse number of follicles at the early stages of proliferation; istry employing ER cx-specific antisera indicate the and a lack of corpora lutea. (D) Cross-section of a representative ERα-null variant is nuclear localized in granulosa cells of early ovary after prolonged treatment with a GnRH antagonist. Circulating follicles.409 The physiological function of ERβ remains LH levels were reduced in ERβ-null females by treatment with a GnRH cx μ unclear. It is unable to bind E2, but in vitro studies indi- antagonist (60 g Antide) every 48 h from the age of 28–53 days. Ovaries were collected within 24 h of the final treatment. The characteristic ovar- cate the isoform may preferentially heterodimerize with α α α ian phenotypes of ER -null mice (shown in C) are prevented by reduc- ER and act as a dominant negative modulator of ER ing circulating LH levels, indicating these more dramatic phenotypes are action. Rodents also possess an ERβ variant (ERβ2) that secondary to the loss of ERα actions in the hypothalamic pituitary axis poorly binds E2, but this receptor form does not inhibit that are necessary to maintain proper LH levels and not due to the loss α wild-type ERα or ERβ activity.193 Furthermore, categori- of ER within the ovary. (Source: Reproduced from Ref. 415.) (E) Ribonucle- β ase protection assays for the steroidogenic enzymes in adult wild-type cal studies have shown that the rodent ER 2 variant is (WT) and ERα-null ovaries treated either with vehicle (V) or a GnRH 409,429 not detected in human tissues. antagonist (A) every 48 h for 12 days (as described above). ERα-null ova- ries exhibit increased expression of Cyp17 and Cyp19, both of which are Ovarian Phenotypes in Mouse Models of Disrupted reduced to normal after reduction of gonadotropin levels. ERα-null ova- Estrogen Signaling ries also uniquely express Hsd17b3, a Leydig cell–specific enzyme, and MICE LACKING ERα this expression is dependent on gonadotropin stimulation as it is ablated α after the reduction of circulating LH levels with GnRH-antagonist treat- Several lines of ER -null mice have been characterized ments. All samples are normalized to β-actin (Actb) mRNA levels. Source: as listed in Table 25.7. Neonatal and prepubertal ERα-null Reproduced with permission from Ref. 431.

4. FEMALE REPRODUCTIVE SYSTEM 1142 25. STEROID RECEPTORS IN THE UTERUS AND OVARY and mast cell infiltration in the interstitium134,158,373,430 cross the blood–brain barrier (e.g., ZM-189,154, EM-800) (Figure 25.17) (Table 25.7). The cystic and hemorrhagic and produce an αERKO-like gonadotropin profile lead follicles likely originate from antral follicles that fail to to a similar ovarian phenotype; whereas treatments ovulate due to acyclicity (i.e., lack of an LH surge) and with tamoxifen, a receptor antagonist that does not alter therefore become atretic and accumulate within the gonadotropin secretion generates no such effects in the ovary. They are characterized by a mural granulosa cell ovary.217–219 layer of one to several cells thick surrounding an enor- A prominent phenotype in αERKO ovaries that may be mous fluid-filled antrum containing blood and immune indicative of an intraovarian role for ERα is their elevated cells; a degenerating ovum if visible at all; elevated FSH- capacity to synthesize androgens.431 Relative to their receptor and LH-receptor expression in the granulosa wild-type littermates, αERKO females possess plasma cell layer, and a hypertrophied theca (Figure 25.17) that levels of androstenedione and T that are increased 3 and also exhibits elevated LH-receptor levels.373,415 There- 40-fold, respectively.431 Indeed, LH is the primary stim- fore, ERα is not required for the recruitment and early ulus of thecal cell androgen synthesis and plasma LH growth of follicles or the induction of gonadotropin and thecal cell LH-receptor levels are both significantly receptors in thecal and granulosa cells, but is vital to the increased in αERKO females (as discussed previously). later stages of folliculogenesis in the mouse ovary.158,373 Therefore, it is not totally unexpected that αERKO ova- ERα-null (αERKO) females exhibit a severely dis- ries exhibit increased expression of the enzymes neces- rupted reproductive hormonal milieu, which in toto sary for androgen synthesis, most notably a remarkable represents the cause and effect of the overt ovarian phe- increase of Cyp17 expression,431 the enzyme necessary notypes415,431 (Table 25.7). Plasma LH level in αERKO for the final step of androstenedione synthesis (Figure females is elevated three to eight-fold relative to their 25.17). However, the elevated plasma androgens found in wild-type littermates, while FSH levels remain nor- αERKO females is likely not due solely to LH-mediated mal.431 This is interesting in light of the fact that ovariec- hyperstimulation of the theca. E2 from granulosa cells tomy of normal WT mice leads to elevation of circulating is proposed to mediate an intraovarian short feedback FSH and suggests that the αERKO differs from ovari- loop upon thecal cells to negatively modulate androgen ectomized WT females. Therefore the ovarian pheno- synthesis during the later stages of folliculogenesis by types of thecal hypertrophy,373 elevated steroidogenic primarily targeting CYP17 activity.438 Given that ERα is enzyme expression, and increased sex steroid synthesis the dominant ER form expressed in rodent thecal cells, in αERKO females are congruent with hypergonado- the elevated Cyp17 expression found in αERKO ovaries tropic-hypergonadism.431 A synopsis of evidence that suggests ERα mediates this effect of E2 on androgen syn- supports acyclicity and chronically elevated LH as a thesis. Additional support for a specific ERα-mediated primary cause of the other ovarian phenotypes in ERα- effect on thecal cell steroidogenesis comes from our find- null females includes: (1) the cystic and hemorrhagic fol- ings that chronically elevated LH in wild-type and ERβ- licles appear after the onset of puberty (approximately null females leads to a much more moderate increase 40 days of age),373 (2) the cystic follicles and elevated ste- in Cyp17 expression and androgen synthesis relative to roidogenesis (Figure 25.17) are prevented when plasma αERKO females.439 Furthermore, individually cultured LH levels are reduced to normal via treatments with a ERα-null follicles in vitro secrete substantially more GnRH antagonist,415,432 (3) transgenic mice possessing androgens relative to similarly propagated wild-type elevated LH but functional ERα exhibit a comparable follicles even though the level of gonadotropin stimula- ovarian phenotype433–435 that is rescued by periodic tion is held constant.440 Therefore, it may be concluded ovulatory doses of exogenous hCG to induce luteiniza- that ERα is paramount to maintaining proper androgen tion,436,437 and (4) immature αERKO females successfully synthesis in rodent females via: (1) endocrine actions in ovulate, albeit with a reduced number of oocytes com- the hypothalamic–pituitary axis that negatively modu- pared to wild-type mice, and form some corpora lutea late LH secretion, and (2) intraovarian actions on thecal when treated with exogenous gonadotropins prior to cells to negatively modulate Cyp17 expression. the onset of the cystic phenotype.415,430,432 Others pro- The granulosa cell–specific enzymes necessary for pose that aberrantly high intraovarian histamine levels E2 synthesis, Cyp19 and Hsd17b1, are also expressed at due to infiltration of mast cells in the interstitium may elevated levels in αERKO ovaries.431 As a result, plasma also contribute to cyst formation in ERα-null ovaries.158 E2 levels are typically increased almost 10-fold relative Therefore, the previous findings indicate that a primary to wild-type littermates.431 It is likely that the plasma role of ERα in murine ovarian function may be extra- androstenedione levels discussed earlier would be even ovarian, i.e., as an essential mediator of the endocrine greater in αERKO females if ovarian aromatase activity actions of E2 in the hypothalamic–pituitary axis that are was not also equally elevated and providing for effi- critical to gonadotropin regulation. Indeed, prolonged cient conversion.431 Although these findings in αERKO treatment of female rodents with antiestrogens that females suggest that ERα may also negatively modulate

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1143

E2 synthesis in granulosa cells, there is little precedent mice were found to have expressed a splice variant in for this hypothesis. The chronically elevated plasma some tissues, and to circumvent this, a new ERα-null LH in αERKO females is also not likely to positively mouse model was made by Cre/loxP-mediated recombi- influence granulosa cell E2 synthesis as most evidence nation (Ex3αERKO).159 The ovarian phenotype and hor- indicates that LH stimulation of granulosa cells leads to monal milieu of these mice is similar to that observed decreased aromatase activity.441–443 Furthermore, indi- in the αERKO mice,134,157 however the ability of these vidually cultured ERα-null follicles in vitro continue animals to respond to exogenous gonadotropins has to exhibit heightened E2 synthesis relative to similarly not been fully studied to date. Preliminary data suggest propagated wild-type follicles in an environment of con- reduced response to exogenous gonadotropins, as less trolled FSH and LH stimulation.440 Instead, increased than 50% ovulate and those that do respond ovulate a aromatase activity in ERα-null ovaries is likely due to significantly lower number of oocytes (preliminary data the positive actions of estrogens on FSH-induced granu- Korach laboratory); however, further work is necessary losa cell steroidogenesis442,444–447 which are presumably to confirm this finding and fully characterize the ovula- mediated by ERβ and therefore remain intact in ERα-null tory response of the Ex3αERKO mice. ovaries. Androgens have also been shown to augment Recently, a woman with homozygous mutation of FSH induction of E2 synthesis442,448,449 and therefore the ERα leading to severe E resistance was reported.188 It is elevated androstenedione and T levels characteristic of interesting to note that, similar to the ERα null mouse αERKO females may also contribute to increased E2 syn- lines, she had cystic ovaries and elevated E, but unlike thesis in granulosa cells. the mouse models, her LH and T levels were normal The several attempts to induce ovulation of ERα- or nominally elevated. This is perplexing in light of the null females collectively indicate an age-dependent mechanisms discussed previously proposing the key role effect of the loss of ERα. Schomberg et al. report that of LH elevation in causing the hemorrhagic cystic ovar- 4-month-old αERKO females do not successfully ovu- ian phenotype, and suggests notable interspecies differ- late following exogenous treatments with PMSG and ences in the mechanisms underlying these processes. hCG, although this conclusion was based solely on the α absence of corpora lutea rather than actual oocyte num- MICE WITH OVARIAN-SPECIFIC DELETION OF ER bers.373 Two later studies focused on younger αERKO In rodent species, ERα is expressed predominately females (3–5 weeks) and produced comparable results in theca cells, while ERβ is expressed in granulosa cells. of successful ovulation (oocytes in the oviduct) and for- Global loss of ERα in mice leads to a severe ovarian phe- mation of functional corpora lutea, although the oocyte notype with large hemorrhagic and cystic phenotype yield was reduced compared to age-matched wild-type due increased serum LH from loss of negative feedback females.415,430 Oocytes harvested from ERα-null females in the hypothalamic–pituitary axis. While examination successfully undergo in vitro fertilization, indicating of ovarian function in these global ERα-null mice has that ERα may not be important to oocyte function.415 In provided important data on the necessity of ERα for contrast, Dupont et al. report that the other line of ERα- ovarian function, the models make it difficult to delin- null females (ERαKO) fail to ovulate or exhibit corpora eate the importance of ERα specifically in the theca cells lutea following exogenous gonadotropin treatments of the ovary. To circumvent this, Bridges et al. generated even at 21–25 days of age.158 Studies in immature αERKO a mouse model with Cre-recombinase under control of the females of pure C57BL6 background (versus the mixed Cyp17 promoter so that it is expressed specifically in the- background of earlier studies) found that these mice can cal cells in the female mouse, and when crossed with a elicit a response to gonadotropin-induced ovulation.432 ERαf/f mouse deletes expression of ERα in thecal cells in This discrepancy in in vivo ovulatory success between the ovary.361 Further characterization of this mouse and the αERKO and ERαKO mice may be due to differences ovarian function in the absence of ERα demonstrated in genetic strain, the considerable disparity in the doses that these mice prematurely lose fertility.362 of gonadotropin used, or the small sample size in the Theca cell–specific ERα knockout (thEsr1KO) mice Dupont et al. (n = 3). In further support of a minor role did not have an altered estrous cycle, fertility as mea- for ERα in ovulation, Emmen et al. demonstrated that sured by the number of offspring born, or ovarian individual ERα-null follicles propagated and induced to response to exogenous gonadotropins (superovulation) ovulate in vitro behave no differently than similarly cul- in animals aged 2 months; however, these measures were tured wild-type follicles.440 significantly reduced in thEsr1KO females at 6 months The discrepancies in the ovulatory response described of age (Table 25.7).362 In response to exogenous gonado- herein may also be due to the splice variant of ERα tropins the thEsr1KO animals release similar numbers expressed in the αERKO that was able to respond to of oocytes to WT at 2 months of age (27 in WT versus exogenous gonadotropins (albeit reduced number [∼15 22 in thEsr1KO); however, by 6 months there are signifi- oocytes] compared to wild-type [∼41 oocytes]).415 These cant reductions in oocytes released (22 in WT versus 6 in

4. FEMALE REPRODUCTIVE SYSTEM 1144 25. STEROID RECEPTORS IN THE UTERUS AND OVARY thEsr1KO) and the ovaries show an increase in the num- alternative mechanism of ER action, often referred to as ber of cystic/hemorrhagic follicles.362 Additionally, the “tethering”, is thought to occur via interaction with tran- oocytes collected from the thEsr1KO ampulla appeared scription factors that are in fact bound directly to DNA to be more degenerated than WT oocytes, however, (Figure 25.4). To develop an in vivo model for the study of further studies are required to examine the viability of ERα signaling via ERE-independent pathways, Jakacka oocytes collected from thEsr1KO mice. The age-related et al. generated mice that possess a mutant form of the reduction in fertility coincides with an increase in the zinc-finger of the ERα DBD.132 Because this mutation length of estrous cycle these animals have at 6 months of was incorporated into the endogenous Esr1 gene (i.e., age,362 demonstrating that ERα is important for mainte- a “knock-in”), transcriptional expression of the mutant nance of fertility, but its expression in thecal cells is dis- ERα is presumably no different than that of an undis- pensable for normal ovarian response in younger mice. rupted Esr1 gene. Female animals that are heterozygous Global ERα-null mice have increased serum LH, for the ERα mutation, termed NERKI+/− or ERαAA/+ which contributes to the cystic and hemorrhagic phe- mice, are infertile and exhibit distinct ovarian pheno- notype presumably due to loss of negative feedback in types characterized by follicles of all stages of growth the hypothalamic–pituitary axis. Loss of ERα specifically but no corpora lutea and lipid-filled cells throughout the in the theca cell layer leads to a reduction of serum LH stroma (Table 25.7).132 The latter phenotype is similar in mice at 2 months of age, and a further reduction at to that of female mice lacking functional steroid acute 6 months.362 This reduction in LH in the thEsr1KO ani- regulatory protein (STAR) and correlates with reduced mals was not due to reduced gonadotropes in the ante- StAR expression in NERKI+/− ovaries.132 Upon stimula- rior pituitary, as there was actually an increase in the tion with exogenous gonadotropins, NERKI+/− ovaries number of gonadotrope cells. Moreover, no difference is exhibit multiple large hemorrhagic follicles similar to seen in the amount of LH stored in these cells as exam- those found in ERα-null ovaries.132 NERKI+/− females ined by immunofluorescence.362 This confirms that the exhibit normal plasma gonadotropin and E2 levels excess LH in the global ERα-null models is due to loss and proper expression of steroidogenic enzymes in the of ER in the hypothalamic–pituitary axis, and suggests ovary.132 Still, induced ovulation in NERKI+/− females that ERα expression in the ovary may be important for is only partially successful in causing some oocytes to positive feedback and/or normal LH secretory patterns, be released and the formation of corpora lutea, as sev- although this needs to be tested experimentally. eral preovulatory follicles fail to rupture and become Previous work has suggested that androgen synthe- hemorrhagic.132 sis is negatively regulated by estrogen through para- Without further data it is difficult to determine crine signaling actions, which can be examined in the the exact cause of the NERKI+/− ovarian phenotypes. thEsr1KO mouse model. Examination of serum hormone Jakacka et al. postulate that the mutant ERα may act as levels at both 2 and 6 months shows that T is increased,362 a dominant-negative form of ER, however, the mutant confirming that ERα is necessary for maintenance of nor- ERα does not exhibit this property in vitro.132 Further- mal T levels. Examination of the expression of several more, prolactin (Prl) is a known ERE-regulated gene and steroidogenic enzymes found that Cyp17 expression was exhibits no change in expression in NERKI+/− females.132 increased at both 2 and 6 months following PMSG and An imbalance between ERE-dependent and -indepen- hCG stimulation, while Cyp19 expression was not signif- dent pathways or tissue-specific inhibitory effects of icantly different between WT and thEsr1KO animals.362 the mutant ERα may also explain some of the resulting While expression of Cyp17 is increased, the protein con- phenotypes.132 centration of CYP17 was also increased in thEsr1KO To further explore the role of the DNA binding domain ovaries in both aged animals and younger mice. CYP17 in ERα function in the ovary, both the ERAA/− (also called levels were also examined via immunohistochemistry “KIKO”) and EREAAE mouse models (described in the during diestrus, when CYP17 activity and expression section Mice with Mutated DNA Binding Domains of should be minimal, and increased CYP17 was observed ERα) were developed,162,168 both expressing only ERα in interstitial cells in the thEsr1KO ovary compared to that is unable to bind directly to ERE DNA motifs. wild-type ovaries.362 The collective data support the KIKO were found to be anovulatory and infertile (Table notion that ERα signaling is necessary to support proper 25.7).162 Further characterization of the ovaries of these androgen synthesis in the ovary. mice found follicles at most stages of development, how- ever, there was a lack of corpora lutea in the ovary sec- MICE WITH MUTATED DNA BINDING DOMAINS OF tions. The lack of corpora lutea and an estrous cycle in α ER these mice suggests that they do not ovulate, however, As discussed earlier in this chapter, the ERs may act the ability of these mice to respond to exogenous gonad- as transcription factors on certain genes in the absence otropins was not examined. These animals have some of a classical ERE within the regulatory sequences. This of the ovarian phenotypes identified in the NERKI+/−,

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1145 demonstrating that the DNA binding activity of ERα is reported (Table 25.7).165,167 To examine the importance important for normal ovarian function. ERαEAAE/EAAE of the AF-2 domain, a knock-in mutation in the AF-2 females are infertile and have a hemorrhagic and cys- domain of ERα was made such that E2 is no longer able tic ovarian pathology (Table 25.7)164 similar to the ovar- to transactivate E-dependent transcription.169 This two- ian phenotypes observed in the other models that carry point mutation in the AF-2 domain was introduced into a mutation in the DNA binding domain. While further a knock-in mouse referred to as AF2ERKI/KI. The mice characterization of the ovarian phenotype in these ani- are anovulatory and infertile (Table 25.7),169 similar to mals is necessary to determine the exact cause of the the global αERKO knockout mouse models developed to ovarian phenotypes, the data to date suggest that DNA date.134,157,159 The ovaries of these mice also resemble the binding is an essential activity of ERα in normal ovarian ovaries in ERα-null mouse models, in that they are hem- function. orrhagic and cystic and lack presence of corpora lutea.169 The KIKO animals have normal levels of E2 and P Furthermore, the endocrine milieu of these mice is simi- (Table 25.7),162 and hormone levels were not examined lar to the global ERα-null animals; the AF2ERKI/KI mice in the ERαEAAE/EAAE mouse model although ovarian have elevated serum LH and E2 (Table 25.7)169 presum- phenotypes were similar between the two mouse mod- ably due to lack of negative feedback in the hypotha- els,164 suggesting that regulation of serum steroid levels lamic–pituitary axis. The ovarian functions of other mice are not dependent on ERα DNA binding activities. Fur- with AF-1 or AF-2 disruptions (ERαAF-10 and ERαAF-20) thermore, the role of the DNA binding domain in ERα in were not reported (Table 25.7). These models suggest tissues other than the ovary (such as the pituitary) could that the ligand binding domain of ERα is required for contribute to some of the observed phenotypes. Tissue- normal ovarian function; however, the results differ specific knock-in mutations to examine the role of direct slightly from the tissue-specific ERα knockout mouse DNA binding of ERα in ovarian function experiments described earlier. Therefore, further work using a tissue- could provide more mechanistic properties of the role of specific knock-in mouse model would be useful in deter- ERα, however, these experiments have not been done to mining the mechanistic role of the different domains of date. ERα in the ovary.

MICE WITH MUTATED AF-1 OR AF-2 DOMAINS OF ERα MICE LACKING ERβ While the models presented to date discuss the role of Several facets of ovarian physiology thought to be the DNA binding domain of ERα, there are other impor- dependent on the paracrine actions of E2 are clearly tant domains of the nuclear receptor that may also con- maintained in ERα-null and ERα-mutated ovaries, tribute to the role of ERα in the ovary. To examine this, including granulosa cell proliferation, LH-receptor two different mouse models were developed with muta- and aromatase expression, antrum formation, and the tions in the ligand binding domain of ERα. In the first attenuation of atresia. The preservation of these estro- model, a single point mutation was made in the LBD gen actions in ERα-null ovaries strongly suggests their (Figure 25.1) where a leucine was substituted in place dependence on ERβ, which continues to be expressed of residue 525 where normally a glycine resides. This in αERKO granulosa cells.373,431 Neonatal ovaries mutation alters the binding pocket, preventing normal from ERβ-null mice exhibit no gross abnormali- ligand binding due to the increased side chain present ties,134,158,171,450 however, there are some differences in in leucine compared to the single methyl group present the expression of extracellular matrix proteins at this in glycine. The knock-in mouse model ERαG525L, hereaf- time of development.451 Adult ERβ-null ovaries pos- ter referred to as ENERKI, is anovulatory with a hem- sess all stages of folliculogenesis,158,171,440,450 a slight orrhagic and cystic ovarian pathology (Table 25.7).168 but perceptible increase in atretic follicles,171,440,450 The animals were found to have an increased number and a paucity of corpora lutea171,440 (Table 25.7; Fig- of atretic antral follicles, which presumably became the ure 25.17). Cheng et al. found that among 5-month- hemorrhagic/cystic follicles observed due to lack of old ERβ-null females, less than 30% exhibit corpora rupture. Furthermore, these animals did not have any lutea compared to 100% incidence in wild-type litter- corpora lutea present, similar to the ERα suggesting that mates.450 Furthermore, when corpora lutea are pres- they do not ovulate,168 demonstrating the importance of ent in ERβ-null ovaries, there are rarely more than two ligand binding ERα for normal ovarian function. versus upwards of seven in wild-type females.450 Con- The AF-1 and AF-2 domains of ERα are impor- tinuous mating studies with ERβ-null females indicate tant for ER function as demonstrated in vitro reporter a severe impairment in fecundity, but there is discern- assays. Mouse models were developed with each AF ible variability in this phenotype among age-matched domain deleted, and uterine function was determined animals.158,171 Consistent among βERKO pregnancies as described earlier in this chapter; however, the ovar- is reduced litter size of approximately one-third rela- ian phenotype or hormone levels of these mice were not tive to wild-type.158,171

4. FEMALE REPRODUCTIVE SYSTEM 1146 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

However, whereas some βERKO females exhibit the phenotypes that are associated with chronically elevated expected number of pregnancies over a 4 month period, LH.415,435 Therefore, ERα is clearly the principal receptor others become pregnant only once while still others involved in mediating the negative feedback effects of exhibit total infertility.171,172 This variability in fecun- E2 on gonadotropin secretion from the hypothalamic– dity among ERβ-null females remains puzzling but is pituitary axis. Sex steroid hormone levels in ERβ-null reported in both independently generated lines.158,171 females are also comparable to wild-type littermates.431 Given that the ERβ-null lines represent distinct target- Additionally, the overtly estrogenized uterine and ing schemes for the Esr2 gene on different genetic back- vaginal tissues observed in ERβ-null females indicate grounds, the observed variability in fertility is likely sufficient ovarian estrogen synthesis in the absence of related to the loss of ERβ functions. A similar variabil- functional ERβ.134 This preservation of normal E2 levels ity in oocyte yield is found when immature ERβ-null in βERKO females is surprising given the evidence that females are induced to ovulate by exogenous gonado- estrogens are necessary to maximize FSH stimulation of tropin treatments.158,171,172,432 A third ERβ-null mouse aromatase activity in granulosa cells.442,444–447 model was recently developed in which all mice were Recent studies into the role of ERβ in normal ovula- found to be infertile and unable to respond to exogenous tory function have found that ERβ is necessary in gran- gonadotropins.172 These animals also had a significant ulosa cells for normal response to both gonadotropins reduction in the number of corpora lutea present in the FSH and LH. Granulosa cells isolated from ERβ-null ovaries of the ERβ-null mice.172 While subfertile ver- ovaries and grown in culture had reduced levels of sus infertile phenotypes are observed in the different cAMP in response to FSH stimulation compared to gran- ERβ-null mouse lines, it should be noted that although ulosa cells isolated from WT ovaries.454 The reduced sig- some mice did produce pups and respond to exogenous naling response to FSH was also observed in granulosa gonadotropins, the number was significantly lower than cells from in vitro–stimulated ovaries, and whole fol- that observed in WT mice, demonstrating the impor- licles grown in vitro.454,455 The reduced cAMP accumu- tance of ERβ in normal ovarian function. Therefore, the lation after FSH stimulation coincided with a reduction observed sub/infertility in ERβ-null females appears to of LH-receptor (officially Lhcgr) mRNA accumulation.455 originate from disrupted ovarian function that is best This suggests that granulosa cells in ERβ-null ovaries are characterized to date as infrequent and inefficient spon- unable to stimulate increased LH-receptor expression, taneous ovulation. A lack of evidence indicating embryo which could contribute to the inability of these cells to resorption during gestation in mated βERKO females171 respond to LH intracellular signaling components363,455 rules out an extraovarian contribution, although such and ultimately ovulate. Still, βERKO females clearly do phenotypes have not been categorically studied. not exhibit a normal ovarian cycle, as evidenced by vagi- Histological evaluation of ERβ-null ovaries follow- nal smears indicating animals in persistent estrous and ing induced ovulation reveals multiple preovulatory the paucity of corpora lutea in the ovary. This suggests follicles possessing underdeveloped antra and minimal the βERKO female mounts a preovulatory rise in E2 that cumulus expansion.158,171,172,432 Krege et al.171 report is insufficient to induce the gonadotropin surge from the obvious corpora lutea in immature βERKO ovaries 20 h hypothalamic–pituitary axis, and/or a diminished abil- after hCG treatment, but Dupont et al. remark that no ity of the βERKO ovary to respond to the LH surge. As luteal cells were observed in similarly treated ERβKO mentioned before, androgens also augment FSH-induc- ovaries that did not ovulate.158 However, immature tion of E2 synthesis442,448,449 and provide for compensa- βERKO females do indeed exhibit functional corpora tory actions in ERβ-null granulosa cells. lutea following induced ovulation, as indicated by histo- α β logical evaluation of the ovaries and a substantial rise in MICE LACKING ER AND plasma P levels following induced ovulation.432 Lutein- Congruent with the loss of ERα, αβERKO females are ization and increased P synthesis is also observed when infertile and do not spontaneously ovulate.158,174 Adult individually propagated ERβ-null follicles are exposed αβERKO ovaries possess structures appropriate to the to an hCG bolus in vitro.440 Furthermore, luteinization normal ovary, including primordial and growing fol- occurs in unruptured preovulatory follicles both in vivo licles, although the latter possess an underdeveloped and in vitro,134,171,440 resulting in “trapped” oocytes simi- antrum, reduced granulosa number and a thin poorly lar to those observed in other null mouse models.176,452 structured theca (Table 25.7).158,174 Cystic follicles that ERβ-null females exhibit a relatively normal repro- are somewhat characteristic of those found in ERα-null ductive endocrine milieu.431 Basal gonadotropin lev- ovaries are also present but are not as large or as hemor- els are within the range of control values,431 although rhagic in αβERKO ovaries and are more apt to exhibit one report remarks that plasma LH levels are slightly only a thin layer of granulosa cells.158,174 We have used increased in βERKO females.453 This notwithstand- mice with chronically elevated expression of LH from a ing, ERβ-null females clearly do not exhibit the ovarian transgene (α-LHβCTP mice) to show that the formation

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1147 of hemorrhagic and cystic follicles in the mouse ovary indicate these structures are the “ghosts” of atretic are due to ovarian acyclicity compounded by chroni- follicles as they possess an intact basal lamina, par- cally elevated plasma LH in the presence of normal FSH tial layers of granulosa cells, and an invariably degen- levels.435,437 The occurrence of hemorrhagic follicles in erating oocyte174 (Figure 25.18). Furthermore, these α-LHβCTP transgenic ERα-null mice and the absence structures are postpubertal as Couse et al. report of hemorrhagic follicles in α-LHβCTP transgenic ERβ- no such structures in αβERKO ovaries at 10 days of null mice439 demonstrates a role of ERβ in development age174 and Dupont et al. concur in another line of of the cysts. Therefore, the absence of hemorrhagic and ERαβKO ovary as late as 23 days of age.158 The defining cystic follicles in αβERKO ovaries indicates an impor- characteristic of the “ghost” follicles that led to their ini- tant intraovarian role for ERβ in this pathology. A recent tial description as testis-like seminiferous tubules is the compound αβERKO mouse model that does not express overt presence of Sertoli-like cells in the lumen (Table the splice variant for ERα159 has elevated plasma LH and 25.7 and Figure 25.18).158,174 Several morphological fea- hemorrhagic and cystic ovarian phenotype observed tures of these cells in both αβERKO and ERαβKO ovaries in the ERα-null mice.456 The discrepancy observed in are strongly indicative of a Sertoli cell phenotype (Figure these mice suggests that the presence of the splice vari- 25.18), including: (1) alignment with the basal lamina ant, genetic background, or mode of knockout in the of the follicle wall, (2) a tripartite nucleolus, (3) numer- mouse models may contribute to the ovarian pathology ous veil-like cytoplasmic processes extending inward observed. Further studies are necessary to determine the toward the lumen, (4) ectoplasmic specializations that role of ERα or ERβ in development of hemorrhagic fol- are unique to Sertoli cells in prepubertal testis, and (5) licles within the ovary. immunoreactivity for Müllerian-inhibiting substance A most remarkable feature of adult αβERKO ova- and sulfated glycoprotein-2. ries is the presence of seminiferous tubule-like struc- Granulosa and Sertoli cells fulfill analogous game- tures that often occupy large portions of the gonad togenic roles as “nurse” cells in the ovary and testis, but are conspicuously absent in ERα- or ERβ-null ova- respectively, and are postulated to derive from a com- ries158,174 (Figure 25.18). Morphological observations mon embryological precursor cell during early gonadal

FIGURE 25.18 Ovarian phenotype in adult mice lacking both ERα and ERβ. (A, B) Low-power magnification of repre- sentative adult wild-type (A) and ERαβ–null (B) ovaries. The wild-type ovary exhibits all stages of folliculogenesis, including a preovulatory follicle (right) and corpus luteum (right, bottom). The ERαβ-null ovary exhibits a few relatively healthy maturing follicles but is overwhelmed by multiple “sex-reversed” follicles. (C–F) High-power magnification of representative follicles from adult ERαβ-null ovaries (C, 66×; D, 330×). A healthy follicle shows a single oocyte (O), several layers of granulosa cells (GC), an intact basal lamina, and thecal cells (TC) (E, 66×; F, 330×). A representative “sex-reversed” follicle in which there is no lon- ger evidence of an oocyte and the somatic cells have undergone redifferentiation to a Sertoli-like cell (SC) phenotype. Preserva- tion of the basement membrane (BM) of the follicles provides for the “tubular” appearance that assimilates testicular cords. The Sertoli-like cells (SC) possess the characteristic tripartite nucleolus and veil-like cytoplasmic extensions (F). Scale bars: E ×100 μm; D, F ×10 μm. Source: Reproduced with permission from Ref. 174.

4. FEMALE REPRODUCTIVE SYSTEM 1148 25. STEROID RECEPTORS IN THE UTERUS AND OVARY differentiation.457 Thus, the origin of the Sertoli-like throughout life in the mouse. Recently, postnatal repro- cells in αβERKO ovaries remains a perplexing question. gramming of ovaries to testis with cells that are positive Do they originate from a bi-potent precursor popula- for SOX9 expression and cells resembling Sertoli cells tion present in the αβERKO ovary at birth, or are they was reported in adult mice lacking FOXL2.468 This pro- the result of granulosa cell transdifferentiation fol- vided a second postnatal model for this ovarian transdif- lowing oocyte death and follicle atresia? A report of ferentiation, and suggested that FOXL2 is an important extra-follicular Sertoli cells in the ERαβKO ovary158,458 transcription factor necessary for granulosa cell differenti- supports the existence of an embryological precursor cell ation and ovarian function. Interestingly, the Ex3αβERKO population, yet similar interstitial cell populations are mouse model shows loss of Foxl2/FOXL2 expression fol- not found in ovaries from the other αβERKO.174 In turn, lowed by increased Sox9/SOX9 expression and appear- the following findings that are common to both lines ance of Sertoli-like cells in the ovary of adult mice.456,469 support a pathway of granulosa cell transdifferentiation: A causal link between the loss of all ER function and (1) the Sertoli-like cells are not apparent prior to 30 days postnatal morphological ovarian sex reversal is unclear. of age, (2) the spherical shape of the “tubules” suggested Dupont et al. report that female mice possessing only by two-dimensional histology suggests they originate one functional ERα allele but lacking functional ERβ from a once-healthy follicle in which only the basement also exhibit the αβERKO ovarian phenotype, suggesting membrane remains, and (3) the appearance of Sertoli a gene-dosage effect for ERα.158 The majority of docu- cells is strongly correlated with oocyte death and fol- mented cases of similar phenotypes in the mammalian licle atresia.158,174 Furthermore, Sox9 is highly expressed ovary are preceded by massive germ cell loss.470 The in both αβERKO and ERαβKO ovaries174 and localized remarkable oocyte attrition that occurs in the αβERKO to the Sertoli-like cells.458 SOX9 is an Sry-related tran- ovaries suggests a similar mechanism occurs in this scription factor that is critical to normal Sertoli cell dif- model; yet it remains unclear if the progressive loss ferentiation during testis development in rodents and of germ cells in αβERKO ovaries illustrates the loss of humans.459,460 Transgenic overexpression of Sox9 leads critical intraovarian estrogen signaling that is perhaps to phenotypic male gonadal development in XX mice dependent upon ERα/ERβ cooperation. A role of direct and therefore acts “downstream” of the Y-linked male- estrogen/ER actions in gonadal differentiation is sup- determining gene, Sry.461,462 In strong favor of a pathway ported by reports of sex reversal in turtles and whiptail of granulosa cell transdifferentiation, Sox9 expression in lizards following developmental exposure to aromatase ERαβKO ovaries is restricted to the granulosa cells of inhibitors.471,472 The mouse Sox9 promoter lacks an obvi- atretic follicles and precedes the overt appearance of ous ERE but does possess two consensus binding sites Sertoli-like cells but is not detectable in the granulosa for GATA-1,460 a transcription factor expressed in Sertoli cells of primordial follicles.458 Furthermore, those cells cells during differentiation of mouse testis but repressed expressing Sox9 are initially present with the confines by germ cells in adult testis.473 Furthermore, ERα can of the follicle “ghosts” and appear in the intrafollicular repress the transcriptional activities of GATA-1,474 sug- spaces of the ovary only after the phenotype has pro- gesting that aberrant Sox9 expression leading to Sertoli- gressed to an advanced stage,458 supporting a follicular cell differentiation in αβERKO ovaries may be due to (i.e., granulosa) origin. increased GATA-1 activity following the loss of ERα and Similar phenotypes of “sex-reversal” are reported in other potential oocyte-derived inhibitory factors. As a fetal rodent ovaries following: (1) transplantation of the similar phenotype is observed in FOXL2 knockout ova- fetal ovary to an adult host,463,464 (2) in vitro exposure ries, it has been hypothesized that Foxl2/FOXL2 is regu- to purified Müllerian inhibiting substance (MIS),465 (3) lated by ERE-dependent transcription,468 and loss of this transgenic MIS overexpression in vivo,453 (4) transgenic regulation contributes to the Sertoli-cell phenotype in Sox9 overexpression in vivo,461,462 and (5) following tar- the ovaries of knockout mice. This does not account for geted disruption of the Wnt4 gene in mice.466 Although the lack of transdifferentiation during ovarian develop- the αβERKO ovarian phenotype shares a number of ment or in neonatal animals,158,174,468 which suggests that morphological similarities with the previous findings, this “sex-reversal” or transdifferentiation phenotype is a including aberrant expression of the MIS, SGP-2 and complex trait that requires aberrant regulation/expres- Sox9 genes, a remarkable distinction in the αβERKO sion of multiple genes. Current studies are underway ovary is the postnatal onset of the phenotype. The previ- with the Ex3αβERKO mouse model to look at ovarian ous descriptions of ovarian sex reversal are reported to gene expression and identify possible candidate genes occur or even require fetal ovarian tissue.465,467 Therefore, that may contribute to the granulosa cell transdifferen- the observed “sex-reversal” adult αβERKO ovaries is the tiation observed.456,469 first to be described in the postnatal mouse gonad, indi- Female αβERKO animals exhibit a gonadotropin pro- cating that the potential of female ovarian somatic cells file that is similar to αERKO females, although plasma to redifferentiate into Sertoli-like cells may be present LH levels are noticeably higher. Therefore, the hormonal

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1149 milieu in αβERKO females may impact the ovarian “sex females. However, Britt et al. later demonstrated that reversal” observed.174,431 Still, transgenic mice possess- mutant females maintained on a diet free of phytoes- ing chronically elevated plasma LH also display a pro- trogens exhibit a clearly exacerbated ovarian phenotype gressive loss of oocytes but lack any evidence of ovarian by 6 weeks of age that included hemorrhagic and cystic “sex reversal”.435,439,475 Interestingly, αβERKO ovaries do follicles, seminiferous tubule–like structures possessing not exhibit αERKO-like increases in ovarian aromatase Sertoli-like cells, and an almost 40-fold increase in Sox9 or E2 synthesis.431 This finding provides further support expression.194 By 16 weeks of age, the tubule-like struc- that elevated LH and precursor substrates, combined tures and Sertoli cells occupy up to 80% of the ovary and with the positive actions of ERβ, lead to enhanced E2 are strikingly similar to those found in αβERKO ovaries synthesis in ERα-null ovaries—a scenario that would be in both morphology and gene expression.194 Additional disrupted in αβERKO ovaries. The dramatic loss of germ supporting evidence of Sertoli-like cells in Cyp19-null cells in the αβERKO ovary may also impact gonadotropin ovaries is documentation of Sertoli cell-specific ectoplas- responsiveness of granulosa cells as oocyte-derived fac- mic specializations and immunoreactivity for Espin, an tors are known to positively influence steroidogenesis; 476 important component of Sertoli cell junctions in testis.194 further studies are necessary to determine how germ cell Recent microarray analysis has identified 450 genes that loss contributes to the observed phenotype. are differentially expressed in Cyp19-null ovaries com- pared to wild-type ovaries, with 291 showing increased MICE LACKING CYP19 expression and 159 showing reduced expression.478 In the ovary, CYP19, or aromatase, is expressed exclu- Several of these genes have been implicated in testis sively in the granulosa cells (Table 25.6). Initial reports of development, however, further studies and confirma- Cyp19-null ovaries indicated no phenotypic difference tion are necessary to determine potential candidates that from WT, suggesting estrogen-independent signaling may contribute to ovarian transdifferentiation. Interest- mechanisms were involved in eliminating the pheno- ingly, a similar phenotype of “sex reversal” has not been types of the ERKO mice.477 It was not until Cyp19-null described in the ovaries of a second line of Cyp19-null mice were fed a soy-free diet that phenotypes appeared. females.131 These observations were some of the first to directly As expected, the reproductive hormonal milieu in demonstrate dietary estrogen contamination in feed. female Cyp19-null mice is severely disrupted. Plasma E2 Therefore, to study the phenotypes associated with levels131 and ovarian aromatase activity156 are below the loss of estrogens, Cyp19 null are kept on soy-free diets. level of detection, supporting the existence of a single Under these conditions, Cyp19-null mice have ovaries aromatase-encoding gene in mice. In turn, plasma T lev- with all stages of folliculogenesis, including multiple els in Cyp19-null females are 10-fold that found in wild- large antral follicles but no corpora lutea at 10–14 weeks type littermates.131,156 Increased plasma androgens in of age (Table 25.7).131,156,364 At 21 weeks of age, Cyp19- Cyp19-null females are likely the combined effect of an null ovaries continue to exhibit all stages of folliculo- accumulation of androgen precursors that occurs in the genesis, but larger follicles possess a reduced number of absence of aromatization to E2, and increased synthesis granulosa cells and become enlarged, cystic, and hem- due to hyperstimulation of the ovarian theca by five-fold orrhagic, similar to ERα-null follicles.364 By 1 year of increased plasma LH.156,364 Thus, increased plasma LH age, Cyp19-null ovaries are characterized by a paucity and androgens are characteristic of mice lacking ERα- of normal follicles, an increased number of cystic and mediated estrogen actions following the loss of recep- hemorrhagic follicles, significant collagen deposition, tor (ERα-null animals) or ligand (Cyp19-null animals). and massive macrophage infiltration into the intersti- However, one striking difference between Cyp19-null tium (Table 25.7).364 Several findings indicate infertility and ERα-null females is the four- to six-fold increase among Cyp19-null females due to an inability to spon- in plasma FSH that is unique to the former.364 Interest- taneously ovulate, including: (1) absence of corpora lutea ingly, Cyp19-null females exhibit the above abnormal at all ages,131,156,364 (2) lack of pregnancies during con- hormone milieu regardless of phytoestrogen content in tinuous mating,131 and (3) total failure to ovulate follow- the diet,194 suggesting that dietary phytoestrogens may ing exogenous gonadotropin treatments.131 Recently, it attenuate some but not all of the effects that follow the has been demonstrated that co-treatment of Cyp19-null loss of endogenous E2 synthesis. mice with E2 and PMSG followed by hCG can rescue the Studies have indeed shown that Cyp19-null mice still anovulatory phenotype,365 demonstrating the necessity possess functional ER signaling.479 Tonic E2 replacement of E2 signaling in addition to the gonadotropins for nor- therapy in Cyp19-null females over a period of 21 days mal ovulatory function. leads to restored gonadotropin homeostasis, improved Surprisingly, initial descriptions of the ovarian phe- follicular development in the ovary, a reduced presence notypes in Cyp19-null females did not include remarks of Sertoli-like cells and Sox9 expression, and restora- of Sertoli-like cells similar to those found in αβERKO tion of ovulation and corpora lutea formation.479,480 In

4. FEMALE REPRODUCTIVE SYSTEM 1150 25. STEROID RECEPTORS IN THE UTERUS AND OVARY a similar study, Toda et al. report that E2 administration This section employs the contributions from studies of to Cyp19-null females every fourth day for 4 weeks also the pertinent null- or mutated-mouse models as a plat- ameliorates the ovarian phenotypes but did not restore form to discuss the current view on intraovarian estro- ovulation.131 Ovulation was partially restored when gen/ER signaling. the mice were treated with E2 in addition to exogenous gonadotropins,365 although oocyte numbers were still GRANULOSA CELL PROLIFERATION significantly lower than those observed in WT mice. A number of studies in immature hypophysectomized Furthermore, treatment of the Cyp19-null mice with a rodents have demonstrated that DES or E2 treatments GnRH antagonist and E2 replacement was able to restore over 2–4 days leads to a marked increase in ovarian normal gonadotropin levels, while treatment with an weight489 a significant rise in the number of medium- antiandrogen had no effect. This demonstrates that loss sized preantral follicles and increased DNA synthesis of E2 in these animals contributes to the ovarian pheno- in both thecal and granulosa cells,422,490–492 indicating type observed and also indicates that an indirect effect a direct and gonadotropin-independent effect of E2 from excess androgens does not directly contribute to on the ovary. Similar data have been reported follow- the excess gonadotropins, hemorrhagic follicles, or Ser- ing DES treatment of intact immature rats.493 However, toli-like cells present in the Cyp19-null mice.481 there is equally convincing data that sequential treat- ment of immature hypophysectomized rodents with Intraovarian Roles of Estradiol in Ovarian DES followed by FSH results in a synergistic response Function in terms of increased ovarian weight and DNA syn- The importance of the extraovarian or endocrine thesis, and elicits several major indices of granulosa roles of estrogen signaling in ovarian function has long cell differentiation, i.e., antrum formation and acquisi- been recognized and is the precedent for steroid-based tion of LH-receptor expression, that are not observed oral contraceptives.482,483 In contrast, the significance of with either hormone alone.359,490,491,494 Furthermore, estrogen signaling within the ovary is not well under- co-administration of an antiestrogen (cis-clomiphene) stood due to the inherent difficulties associated with the inhibits FSH-induced increases in ovarian weight and study of a particular hormone’s action within the very follicular growth in hypophysectomized rats, suggest- tissue from which it is synthesized and secreted. Past ing that ER-mediated effects are necessary for a full in vivo investigations employing ER antagonists217–219,484 response to gonadotropin.495 The synergistic actions of or aromatase inhibitors484–486 to inhibit hormone recep- E2 on FSH-induced granulosa cell proliferation have tor actions within the ovary have provided informative been replicated in vitro,496 although others report that but often equivocal findings217–219 due to the following FSH-mediated growth of isolated murine follicles is limitations: (1) the enormous levels of endogenous E2 in not affected by anti-E2 antisera or ER antagonist (ICI the ovary are difficult to overcome by pharmacological 182,780).497 The capacity of E2 to enhance the granulosa administration of a receptor antagonist, (2) the enormous cell response to FSH is not believed to rely on estrogen- levels of aromatase activity in the ovary are difficult induced increases in FSH-receptor levels, although some to overcome by pharmacological administration of an evidence indicates such an effect may play a greater enzyme inhibitor, (3) the effects of a pharmacological role in mice versus rats.359,492,494 These differences will ER antagonist due to actions directly in the ovary versus be easier to study with the recently described ERα-null actions in the hypothalamus or pituitary are difficult to rat,190 which can be used to compare the roles of ERα in discern, (4) the well-characterized agonists (e.g., E2 and rat versus mouse ovarian functions. Still, pretreatment DES) comparably activate both ERα and ERβ, (5) early of hypophysectomized rats with antiestrogen (CI628) antagonists were not selective for the two known ERs, inhibits FSH-induced increases in granulosa cell FSH and (6) the two ERs may respond differently to agonist receptor levels.418 or antagonist ligands. The generation of ER and Cyp19- It is generally agreed that the synergistic response null mice, and development of improved ER-selective of granulosa cells to estrogens and discovery of ERβ ligands or SERMs61,487,488 have contributed to better and its substantial expression in the granulosa cells understanding over the years. of growing follicles and absence in atretic or lutein- The abundance of data on the localization of ERα and ized follicles strongly suggests that the trophic actions ERβ within rodent follicles allows for a simplified model of estrogens are direct and ERβ mediated. Support- of estrogen actions within the ovary. For example, given ing data of a greater importance of ERβ versus ERα in that ERα predominates in thecal cells, does it mediate estrogen-induced granulosa cell proliferation comes the inhibitory effects of E2 on LH-stimulated androgen from a report of Hegele-Hartung et al. in which the ERβ- synthesis? In turn, as ERβ clearly is the predominant selective agonist 8-vinylestra-1,3,5-(10)-triene-3,17β-diol receptor in granulosa cells, is it responsible for mediat- induces increased ovarian weight and granulosa cell ing the known estrogen augmentation of FSH actions? proliferation to levels comparable to that elicited by

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1151

17β-E2 in hypophysectomized rats; whereas the ERα- and E2 work in parallel. Therefore, these observations selective agonist 3,17-dihydroxy-19-nor-17α-pregna-1,3,5 and the absence of any gross deficiency in granulosa (10)-triene-21,16α-lactone has little effect.488 However, cells in ERβ-null ovaries suggests that FSH may play a although these data indicate that E2/ERβ actions may greater role in inducing Ccnd2 expression and granulosa indeed enhance FSH-induced granulosa cell prolifera- cell proliferation. Interestingly, E2 specifically induces tion, several findings question the overall importance Ccnd2 expression in keratinocytes via the cAMP/PKA of this intraovarian role. First, ERβ-null ovaries exhibit signaling pathway that leads to activation of CREB,502 a relatively normal number of growing preantral fol- the same intracellular pathway employed by FSH sig- licles with no obvious reduction in the granulosa cell naling in granulosa cells. These findings suggest that E2 population.158,171,498 Furthermore, immature ERβ-null may also function within the FSH signaling pathway to females treated with PMSG exhibit the expected increase induce Ccnd2 in granulosa cells. in the number of small and large preantral follicles.158,440 A second mechanism by which E2 may amplify FSH- In addition, granulosa cell tumors, due to the loss of induced granulosa cell proliferation is via interactions inhibin-α (Inha) in mice continue to proliferate regard- with the ovarian IGF1 axis. Like Ccnd2-null mice, Igf1- less of the presence or absence of functional ERβ.499 It null mice exhibit a total failure of follicles to progress is possible that ERα may provide some compensatory beyond the small preantral stage.503,504 This arrest in actions as ERβ-null follicles show increased expression folliculogenesis in Igf1-null ovaries is primarily due to of ERα mRNA.455 Furthermore, Cyp19-null mice are pre- a >50% reduction in FSH receptor levels in the granu- sumably devoid of E2-dependent ER action but exhibit losa cells, severely hampering their response to FSH.504 ovaries possessing all stages of follicle growth, each with However, Kadakia et al. demonstrated that Igf1-null a normal complement of granulosa cells.364 Therefore, granulosa cells are also refractory to E2-induced prolif- the data suggest that although estrogen-induced granu- eration and exhibit a 50% lower rate of DNA synthesis losa cell proliferation is likely mediated by ERβ, it is not and decreased Ccnd2 induction following E2 treatment obligatory to gonadotropin-induced follicle growth in in vivo.505 The poor response of Igf1-null granulosa cells rodent ovaries. to E2 may be due to a likewise reduction in ERβ levels as ERβ may facilitate gonadotropin-induced granulosa IGF1 is reportedly necessary for the maintenance of ERβ cell proliferation through the induction of cyclin-D2 expression in granulosa cells in vitro.506 There are cur- (Ccnd2), which enhances progression of the cell cycle rently no reports describing the levels of ERβ expression from the G0 to S phase. Targeted disruption of the Ccnd2 in the ovaries of Igf1–null mice. Interestingly, Richards in mice indicates this cyclin protein is obligatory to gran- et al. demonstrated that E2 significantly upregulates the ulosa cell proliferation as Ccnd2-null ovaries lack fol- expression of several components of the IGF1 signal- licles beyond the small preantral stage.500 Furthermore, ing pathway in rat granulosa cells, including the IGF1 the arrest in folliculogenesis observed in Ccnd2-null receptor-β subunit, the glucose transporter, Glut-1, and mice is not rescued by exogenous FSH treatment in vivo the forkhead family member, FKHR (Foxo1).506 These or in vitro, indicating the critical downstream role of findings suggest that E2 may enhance IGF1 actions in cyclin-D2 in gonadotropin-mediated granulosa cell pro- granulosa cells by regulating targets important for cel- liferation.500 In rats, Ccnd2 expression is localized to the lular energy flow, glucose metabolism, and cell sur- granulosa cells of early growing follicles and is signifi- vival.506 In summary, IGF1 actions are clearly necessary cantly induced by PMSG or FSH treatments.501 How- to provide for sufficient FSH signaling in granulosa cells, ever, E2 also induces Ccnd2 expression in rat granulosa which in turn, is necessary for the induction of aroma- cells, and although the response is less rapid relative to tase activity and E2 synthesis, thereby providing ligand that elicited by FSH, the overall induction by E2 is higher for ERβ to enhance the effects of IGF1, hence forming an and better sustained over 24 h.501 Furthermore, E2 elicits autocrine regulatory pathway to promote granulosa cell a concomitant decrease in the expression of Cdkn1b (p27), proliferation.506 The extent to which these findings may a cyclin-dependent kinase inhibitor.501 E2 induction of translate to the human ovary remains to be determined Ccnd2 in undifferentiated rat granulosa cells in vitro is since IGFII, rather than IGF1, is the predominant and inhibited by the ER antagonist ICI 164,384, indicating gonadotropin-regulated IGF form in the human ovary.507 this to be an ER-mediated event, presumably ERβ.501 The previous discussion provides a compelling argu- Therefore, parallel pathways of FSH and E2 converge to ment that the mitogenic actions of E2 on granulosa cells induce and maintain Ccnd2 expression in granulosa cells, are direct and ERβ mediated. However, thecal cell ER thereby forcing cell cycle progression and subsequent actions may also be involved. For example, ERα in the proliferation.501 Interestingly, adult ERβ-null ovaries thecal cells may respond to granulosa cell–derived E2 by possess normal granulosa cell expression of Ccnd2450 as inducing the synthesis and secretion of a paracrine-acting well as exhibit the expected induction following PMSG growth factor(s) that then becomes the principal stimulus exposure,432 further supporting the hypothesis that FSH for granulosa cell proliferation. Indeed, Dorrington and

4. FEMALE REPRODUCTIVE SYSTEM 1152 25. STEROID RECEPTORS IN THE UTERUS AND OVARY colleagues have proposed that thecal-derived transform- to the capacity of E2 to both amplify the level of FSH- ing growth factor-β (TGFβ) is estrogen induced and stim- stimulated intracellular cAMP359,441,494,515,516 and aug- ulates granulosa cell proliferation.508,509 Similar examples ment the downstream actions of cAMP itself.494,515 The of growth factor–mediated cooperative actions between mechanisms involved remain poorly understood and theca and granulosa cells are postulated to regulate fol- are unlikely to involve estrogen-induced increases of licle steroid production.510,511 Still, ERα-null follicles do FSH receptor levels but appear more likely to rely on not exhibit any gross deficits in granulosa cell number E2’s capacity to positively modulate the granulosa cell and possess a normal number of preovulatory follicles adenylyl cyclase system.441,494 following gonadotropin stimulation.415 ERα-null follicles ER-mediated estrogen activation of the adenylyl in vitro show normal growth, however, they do not reach cyclase system and subsequent cAMP-regulated gene the same size as wild-type follicles.440 It is plausible that expression has been demonstrated in uterine and breast a dramatic effect on granulosa cell proliferation may cancer cells.517 Furthermore, in vitro and in vivo stud- only become apparent following the loss of both putative ies indicate that ERα or ERβ can activate ERE-dependent estrogen pathways, i.e., the direct (ERβ-mediated) and gene transcription via the cAMP/PKA signaling path- indirect (ERα-mediated) actions—hence explaining the way in the absence of E2.93,103,518 This pathway involves unique somatic cell phenotypes observed in the ovaries the CREB protein,519 which is also known to be critical of compound ER-null158,174 and older Cyp19-null mice364 to FSH induction of gene expression.442,520,521 There- on a soy-free diet in absence of estrogen ligand. fore, given the existing evidence that E2 is required to maximize the FSH response, combined with the known GRANULOSA CELL DIFFERENTIATION cooperative activity between the ER and PKA signaling The process of granulosa cell differentiation that pathways, it may be reasonable to expect that a loss of ER occurs during progression from a preantral to preovula- action in the ovary would cause severe deficits in granu- tory follicle has been an area of intense research. Fully losa cell differentiation. Indeed, the ovarian phenotypes differentiated preovulatory follicles in mammalian in ERβ-null females in terms of antrum formation, aro- ovaries are distinctly characterized by: (1) a large fluid- matase expression, and LH responsiveness collectively filled antrum, (2) acquisition of LH responsiveness, (3) represent an attenuated response to FSH-induced granu- significant increases in aromatase (CYP19) activity, and losa cell differentiation in the absence of ERβ. Addition- (4) increased inhibin synthesis. Although FSH is the pri- ally, ERβ-null granulosa cells and follicles have a reduced mary stimulus for both granulosa cell proliferation and response to FSH marked by reduced cAMP accumula- differentiation, these two processes appear to be inde- tion in vivo and in vitro.454,455 The reduction in cAMP can pendent as the undersized follicles in Ccnd2-null ova- be overcome when cells are stimulated with forskolin ries still develop an antrum, possess sufficient aromatase to activate cAMP accumulation,455 demonstrating that activity, and are responsive to exogenous LH stimula- altered cAMP/PKA signaling contributes to the reduced tion.502 Therefore, granulosa cell proliferation and dif- granulosa cell differentiation observed in ERβ-null ova- ferentiation are separate gonadotropin-induced events ries and suggests that maximal activation requires cross during folliculogenesis. talk between FSH and ER-mediated signaling pathways. The absolute requirement of FSH signaling in the Possible downstream mediators were recently identified process of granulosa cell differentiation is illustrated in a microarray study,363 providing key targets to further by the lack of antral follicles, aromatase and E2 synthe- study in the future. sis, and granulosa cell LH receptor in mice null for FSH action.420,512,513 Igf1-null follicles exhibit a similar failure ANTRUM FORMATION to differentiate; this is postulated to be due to a lack of Goldenberg et al. first demonstrated that both FSH sufficient FSH-receptor expression.503 However, numer- and estrogen are required for full antrum formation in ous studies in rat ovaries have demonstrated that FSH preovulatory follicles of hypophysectomized rat ova- induction of antrum formation,490 aromatase expres- ries.490 Wang and Greenwald produced similar data sion and activity,442,444–447 and LH-receptor expres- in hypophysectomized mice, reporting that FSH plus sion484,514,515 require E2 for maximum effect. FSH acts E2 leads to an average of >70 large antral follicles per via a classic heterotrimeric G protein–coupled receptor ovary versus an average of six or none when treated pathway that may activate multiple intracellular second with FSH or E2 alone, respectively.492 The requirement messenger systems, the most well characterized being of FSH plus E2 for antrum formation holds true when activation of adenylyl cyclase, which leads to intracel- rat preantral follicles are grown in vitro.522 Others have lular accumulation of adenosine 3′,5′-monophosphate shown that exposure to aromatase inhibitors during (cAMP) and activation of protein kinase A (PKA).416,494 gonadotropin-induced folliculogenesis also reduces the The long-recognized synergism between E2 and FSH number of healthy antral follicles in rats,485,486 although on granulosa cell differentiation is believed to be due contradictory findings are reported when follicles are

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1153 similarly exposed in vitro.523 Independent evaluations antrum development follows gonadotropin-stimulated of ovaries from three separate lines of ERβ-null mice increases in granulosa cell proteoglycan secretion, indicate a paucity of follicles with fully developed which act to increase osmotic pressure within the fol- antra, even following standard gonadotropin stimula- licle and cause the influx of water from the ovarian tion.158,172,440 These data strongly suggest that the vital vasculature.526,527 McConnell et al. report that water role of E2 in FSH-induced antrum formation is mediated movement into mouse follicles in vitro occurs via a by ERβ. A role for ERα in antrum formation cannot be transcellular pathway and is likely mediated by aqua- excluded but is unlikely given the hallmark phenotype porins-7, -8, or -9 on the surface of granulosa cells.528 in ERα-null ovaries is the invariable presence of severely Although there is currently no evidence of estrogen oversized antral follicles, lending to their description as regulation of aquaporin expression in the ovary, E2 is “cystic”.157–159,373,415 Furthermore, a role of aberrant LH known to influence fluid transport in the uterus via action cannot be discounted, as mice overexpressing a regulation of aquaporins -1,332 -2, and -3,529 and 5 and mutant LH receptor have large cystic and hemorrhagic 8.242,530 Furthermore, ERα is principally involved in the follicles similar to those observed in ERα-null mice.524 regulation of Na+ and water transport across the effer- While it is possible to hypothesize that ERβ may con- ent ductules in the rodent testis,531 suggesting ERα tribute to the formation of exaggerated antrum due to may play a similar role in follicles. Antrum formation hyperstimulation of the receptor, the data supporting may not rely totally on hydromechanical mechanisms this idea are not clear. For example, mice that lack both but is likely to require cell–cell interactions among the ER forms (αβERKO),158,171 or transgenic βERKO females granulosa cell layer. Gore-Langton and Daniel dem- that have elevated LH,439 do not have large cystic fol- onstrated in vitro that rat preantral follicles lacking licles, although a new line of compound αβERKO mice an intact basement membrane and therefore unable to (Ex3αβERKO) do have large cystic follicles456 as do the provide a sealed environment exhibit a reorganization Cyp19-null mice.480 This contradictory data suggests of granulosa cells that is unmistakably antrum-like in that the mechanism behind antrum formation may not response to FSH plus E2, supporting a role for cell–cell be due solely to overstimulation of ERβ, and multiple interaction during antrum formation.522 A lack of antral mechanisms may exist. Furthermore, mice lacking ERα follicles in the ovaries of mice lacking the gap junction specifically in the theca cells develop cystic follicles as component connexin-37 (Cx37) further supports a need they age in response to excess gonadotropins, which for cell–cell interactions among granulosa cells.532 E2 is not observed in younger mice,362 suggesting a more is known to dramatically increase the number of gap complex mechanism contributes to antrum formation. junctions between granulosa cells533 as well as regulate As mentioned, FSH alone will induce the formation ovarian expression of certain component proteins.534–536 of small antra in growing follicles whereas E2 has no Additionally, the extracellular matrix may be important such effect in vivo and in vitro, but both hormones are to antrum formation. Mice lacking ERβ exhibit altered required for full antrum development.490,492,522 Com- expression of several extracellular matrix proteins, pared to the total failure of antrum formation among including increased expression of Collagen 11a1 and follicles in the murine models of disrupted FSH signal- Nidogen 2 at both the mRNA and protein level in both ing,420,503,512,513,525 similarly sized follicles in ERβ-null prepubertal and adult ovaries,451,500 which may alter the ovaries exhibit an initiation of antrum formation but fail follicle composition and contribute to loss of antrum in to develop to a full preovulatory stage.158 Quantitative these mice. Future studies of a potential regulatory role analyses of immature ERβ-null ovaries 48 h after PMSG for E2 in the expression of the aquaporins, connexin, stimulation indicate an increased number of small antral and extracellular matrix proteins may reveal a precise follicles but fewer large (preovulatory) follicles relative role for ERβ in antrum formation. to wild-type females, suggesting an arrest in antrum for- mation.440 Likewise, a large percentage of ERβ-null fol- AROMATASE AND ESTRADIOL SYNTHESIS licles fail to reach maximum size when grown in culture Acquisition of aromatase activity and E2 synthesis is with FSH and E2.440 These follicles also fail to accumu- a hallmark of healthy preovulatory follicles in mamma- late maximal cAMP levels after FSH stimulation,455 sug- lian ovaries. In monotocous species, including women, gesting that multiple signaling pathways may contribute 90% of the circulating E2 is estimated to originate from to antrum formation. These data support the role of FSH the dominant follicle in the ovary. Aromatase (CYP19) actions in the initiation of antrum formation, but in the expression in rat, mouse, bovine, marmoset, and human absence of ERβ-mediated E2 augmentation, maximum ovaries, via activation of the gonad-specific promoter antrum development is prohibited. II,537 is exclusive to and highest in the granulosa cells of Because so little is known about the physiology of fully differentiated healthy follicles (Table 25.6).400,538– antrum formation, the role of ERβ-mediated estrogen 540 Likewise, E2 levels are consistently highest in the actions is speculative. It has long been proposed that follicular fluid of healthy preovulatory, and substantially

4. FEMALE REPRODUCTIVE SYSTEM 1154 25. STEROID RECEPTORS IN THE UTERUS AND OVARY lower in overtly atretic, follicles. The endocrine actions granulosa cells and thereby positively influences its own of ovarian-derived E2 are critical to preparing the female rate of synthesis (Figure 25.19). This pattern of CYP19 reproductive tract for a forthcoming pregnancy and prim- expression described herein strongly resembles the pro- ing the hypothalamic–pituitary axis for generation of the file of ERβ during folliculogenesis, suggesting that the gonadotropin surge. However, there is equally abundant E2-induced enhancement of aromatase activity in granu- evidence that E2 acts in a paracrine or autocrine man- losa cells is ERβ mediated.440 ner to augment FSH stimulation of aromatase activity in

FIGURE 25.19 Synergistic action of sex steroids and FSH in the induction of Cyp19 expression and aromatase activity in rodent granulosa cells. (A) Shown is the effect of in vitro treatment with the synthetic estrogen DES on FSH-induced aromatase activity in granu- losa cells isolated from immature hypophysectomized rats. Granu- losa cells were cultured for 3 days in the presence or absence of FSH (10 ng/ml) with or without increasing concentrations of DES. Three days later, the cells were washed with medium and reincubated for a 5-h test interval in medium supplemented with androstenedione. The accumulation of estrogen (as measured by radioimmunoassay) during this test period is taken as a measure of the level of aromatase activ- ity. C, controls. (Source: Reproduced with permission from Ref. 445.) (B, C) Shown is the effect of FSH and sex steroids on Cyp19 expression and aromatase activity in cultured granulosa cells. Granulosa cells were isolated from 26-day-old rats that had been untreated (unprimed) or treated with E2 (1.5 mg/day) for 3 days (E-primed). Cells were incu- bated in serum-free medium with 5α-dihydroT (DHT) at 20 nM, FSH (50 ng/ml), or FSH in combination with DHT (20 nM), T (20 nM), or E2 (20 nM). Cells were also incubated in medium containing 5% fetal bovine serum (FBS), with FSH alone (50 ng/ml), or in combination with T (20 nM). After 48 h in culture, cells were collected for RNA extraction and evaluation of Cyp19 expression by Northern Blot analy- sis (B) or for measurement of aromatase activity (C). Source: Reproduced with permission from Ref. 442.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1155

FSH is clearly the principal stimulus for the acquisi- primary stimulus of LH-receptor (Lhcgr) expression in tion of aromatase activity in granulosa cells. Once again, preovulatory granulosa cells. However, FSH exposure female mice lacking FSH (Fshb-null mice)419,541 or FSH alone leads to a minimal increase of LH-receptor lev- receptor (Fshr-null mice)421,513 exhibit phenotypes that are els in the ovaries of hypophysectomized rats, and then consistent with a severe reduction in circulating estrogens, only after 4 days of treatment; whereas pretreatment of including uterine hypoplasia and elevated plasma levels animals with E2 for 4 days prior to FSH exposure leads of LH. Furthermore, the ovaries of Fshb-null mice exhibit to an enormous induction of LH-receptor levels that are a six-fold reduction in Cyp19 transcripts,419 although limited to the granulosa cells of preovulatory follicles these data are not corroborated in Fshr-null mice.421 (Figure 25.20).359 This requirement for FSH plus E2 and FSH regulation of Cyp19 expression in granulosa cells the temporal pattern of LH-receptor expression has been is believed to require the actions of multiple transcrip- reproduced in rat granulosa cells in vitro, with maxi- tions factors, including steroidogenic factor-1 (SF-1) and mum levels reached after 48–72 h.515,548,549 Lhcgr mRNA β-catenin,542 CREB protein, GATA-4 and CBP, the former levels mirror the LH-receptor protein levels and exhibit a three of which possess cognate response elements within comparable FSH/E2 regulation in rat granulosa cells.550 the rodent and human CYP19 promoter II.441,520,521,543 Further confirmation that estrogens are required to aug- The mechanism by which E2 augments FSH stimula- ment FSH-induced LH-receptor expression in granulosa tion of aromatase activity or Cyp19 expression in granulosa cells comes from studies that have collectively shown the cells is unclear but shown to occur in isolated granulosa following: (1) only estrogens (e.g., E2, diethylstilbestrol) cell cultures,442,444–447 ruling out the influence of a the- or substrates for aromatization (e.g., androstenedione, cal cell–derived factor and supporting a direct role for T, estrone) are effective,515,548,549 (2) nonaromatizable ERβ. Coadministration of FSH and an ER antagonist androgens (e.g., DHT) and P have little effect (Figure (tamoxifen) to granulosa cells either in vivo or in vitro 25.20),515,549 (3) co-treatment with aromatase inhibitors, inhibits the expected induction of aromatase activity;544 ER antagonists, or 17β-E2-specific antisera block FSH although others report conflicting data using a different induction of LH receptor484,514,548 and hCG-induced ovu- antagonist (ICI 182,780) in whole follicle culture experi- lation,485 and (4) Lhcgr levels are decreased five-fold in ments.497 More recently, Emmen et al.440 and Rodriguez the ovaries of Cyp19-null mice.479 The poor ovulatory et al. found that ERβ-null follicles grown in vitro do response of ERβ-null ovaries to an ovulatory dose of the indeed secrete significantly less E2 relative to wild-type LH analog, hCG,171,432 strongly suggests that ERβ medi- follicles.455 The absence of a consensus ERE within the rat ates the synergistic actions of E2 during FSH induction or human Cyp19 promoter II makes it unlikely that the of the Lhcgr gene in granulosa cells.134,171 In contrast, the effect of E2 is via a direct ERβ mechanism (Figure 25.4). granulosa cells of large antral follicles in ERα-null ova- However, ERβ may interact with the cAMP/PKA path- ries exhibit abnormally high levels of LH receptor, even way, SF-1, CREB, GATA-4, CBP, or other unknown tran- after becoming enlarged and cystic.134,373 scription factors that act to promote Cyp19 expression. The complexity of the Lhcgr promoter is illustrated by Indeed, ERβ has been shown to associate with both the differential regulation between thecal and preovula- CREB519 and CBP545,546 in heterologous in vitro systems. tory granulosa cells. Not surprisingly, FSH induction of The ERs are also known to influence GATA-directed gene LH-receptor expression in granulosa cells occurs via the expression via direct interactions with multiple members cAMP/PKA pathway.551–554 However, the proximal pro- of the GATA family of transcription factors.474,547 There- moter of the Lhcgr gene in both humans and rats lacks fore, there is ample evidence to support ER interaction a consensus cAMP response element (CRE) similar to with the multitude of transcription factors known to those that confer FSH regulation of the Cyp19 promoter influence Cyp19 expression, but the exact nature by which II.555,556 Studies indicate that a GC-rich region possess- ERβ is involved requires further study. ing a cluster of Sp1 binding sites and an ERE half site contributes to FSH and 8-bromo-cAMP induction of the ACQUISITION OF LUTEINIZING HORMONE rat Lhcgr promoter in vitro.557–559 Furthermore, multiple RECEPTOR nonconsensus cAMP-response elements within this In contrast to the constitutive expression of LH GC-rich region of the rat Lhcgr promoter provide a plat- receptor in thecal cells, granulosa cells express LH form for assembling an undefined complex of nuclear receptor only in follicles in the late preovulatory stage. proteins from granulosa cell extracts that confer cAMP This limited expression of LH receptor to the granulosa induction.556,560 With such limited data concerning Lhcgr cells of healthy preovulatory follicles provides for an regulation, it is difficult to speculate where the actions of intraovarian mechanism that ensures only those follicles ERβ may impact LH responsiveness in granulosa cells. that are suitable for ovulation acquire the capacity to ERβ-null granulosa cells have reduced cAMP accumu- respond to the LH surge. As with the other markers of lation after FSH signaling and also show reduced Lhcgr follicle differentiation discussed previously, FSH is the accumulation compared to wild-type granulosa cells.454

4. FEMALE REPRODUCTIVE SYSTEM 1156 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

The reduced expression suggests it may be downstream of cAMP, and in vitro follicle culture studies found increased Lhcgr expression when follicles were treated with forskolin to induce cAMP accumulation.455 As discussed before, E2 enhances the level and action of FSH-stimulated increases in intracellular cAMP to maximize Cyp19 expression, and it is likely that a similar mechanism is employed for induction of LH- receptor levels. However, there are notable differences in the E2/FSH regulation of Cyp19 versus Lhcgr expres- sion. FSH induction of Lhcgr expression in preovulatory granulosa cells is augmented by estrogens only,515,548,549 whereas nonaromatizable androgens or estrogens act through their respective receptor pathways to enhance FSH induction of Cyp19 expression.442 This divergence in regulation is critical since androgen augmentation of FSH on Cyp19 expression allows for FSH/AR-mediated initiation of E2 synthesis in preantral follicles shortly after thecal cells begin synthesizing aromatizable pre- cursors; whereas the specificity for ER/E2 actions in the FSH induction of the Lhcgr gene provides that LH receptors are acquired only by those follicles that have sufficient E2 synthesis, and are hence healthy and suit- able for ovulation. Secondly, Farookhi and Desjardins549 demonstrated that FSH induction of Lhcgr but not Cyp19 expression is lost when estrogen-pretreated granulosa cells are dispersed in culture by EGTA-disruption of cell– cell contacts.549 A requirement for intact cell–cell contacts for E2/FSH-mediated Lhcgr expression may ensure that atretic follicles, which often exhibit a fragmented and less compact granulosa cell population, do not acquire the capacity to respond to an LH surge and ovulate an unhealthy oocyte.

ESTRADIOL MEDIATED NEGATIVE FEEDBACK ON THECAL CELL FUNCTION Like granulosa cells, follicular thecal cells also undergo a process of differentiation during progression from a preantral to preovulatory follicle.561 Fully differenti- FIGURE 25.20 Synergistic action of E2 and FSH in the induction ated thecal cells are best characterized by their unique of LH receptor in rodent granulosa cells. (A) Induction of Lhcgr (LH/ capacity to convert cholesterol to C19 steroids, primarily CG-R) mRNA in granulosa cells of hypophysectomized rats after treat- androstenedione.561 A principal element of the two-cell, ment with FSH and/or E2. Animals were untreated (H) or injected with two-gonadotropin paradigm of steroidogenesis in maturing 1.5 mg E2/day for 3 days (HE), 1 mg FSH twice daily for 2 days (HF), or a follicles (Figure 25.21) is that thecal cells exclusively pos- sequential combination of E then FSH (HEF). Animals were euthanized β after treatment and granulosa cells isolated from the ovaries for total sess the 17 -hydroxylase:C17,20-lyase activity necessary for RNA extraction. Top: Northern blot of 20 μg RNA/lane probed for Lhcgr synthesizing androgens, forcing the granulosa cell layer to (LH/CG-R) mRNA. (PO GC: pro-ovulatory granulosa cell RNA) (Source: be dependent upon the theca as a source of aromatizable Reproduced with permission from Ref. 550.) (B) Effect of antiestrogens on precursors.562 Thecal cell synthesis of androstenedione the induction of LH-receptor levels by FSH, forskolin, or 8-bromo-cAMP. is considered the rate-limiting step in E2 production by Granulosa cells were isolated from immature rats implanted with DES pellets for 4 days. 2 × 105 cells were cultured for 48 h with FSH (100 ng), preovulatory follicles and involves the sequential enzy- forskolin (10 μM), or 8-bromo-cAMP (5 nM), with or without E2, 10−8 M matic actions of P450scc (CYP11), P45017α/17,20-lyase (CYP17) in the absence or presence of antiestrogens, keoxifene (K, 1 μM), or and Δ5-3β-hydroxysteroid dehydrogenase (3β-HSD 1).562 tamoxifen (T, 1 μM). Media were then removed and cells were analyzed It is generally accepted that LH regulates thecal cell ste- for LH-receptor content by radiolabeled binding assay. The first bar on roidogenesis via the induction of CYP17 and CYP11 activ- the left side of the figure is FSH (A), forskolin (B), or 8-bromo-cAMP (C) 561 β 562 alone. Control cells bound less than 10 pg hCG/2 × 105 cells, and the data ity, while 3 -HSD 1 is constitutive. The requirement are not shown. Source: Reproduced with permission from Ref. 514. of LH stimulation to induce CYP17 activity in thecal cells

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1157

FIGURE 25.21 Model of sex steroid receptor actions in the ovary during follicular differentiation. The process of follicle differentiation from preantral to preovulatory follicle is marked in granulosa cells by the induction of aromatase (CYP19) activity for the synthesis of E2 and LH- receptor (LH-R) expression. This process is principally dependent on stimulation of thecal cells by LH and granulosa cells by FSH. However, there is substantial evidence of a modulatory role for the sex steroid receptors in this process. The above model illustrates the putative expression pat- tern and actions of the sex steroid receptors during follicle differentiation. Some aspects of the above model are supported by experimental data, whereas others are more speculation at this time (see text for details). (I) Preantral Follicle: In undifferentiated preantral follicles, LH stimulation of thecal cells (which constitutively express LH-R) leads to de novo synthesis of androstenedione (A4) and T. A4 and T diffuse across the base- ment membrane of the follicle and into the granulosa cells. A4 may be converted to T in granulosa cells. Thecal-cell derived or locally synthesized T activates the androgen receptor (AR), and this action synergizes with FSH to induce aromatase (CYP19) expression in the granulosa cells. The induction of E2 synthesis commences the process of follicle differentiation. (II) Preovulatory Follicle: Increasing E2 synthesis in the granulosa cells leads to activation of ERβ and induces differentiation by synergizing with FSH to further increase CYP19 expression and E2 synthesis and induce LH-R expression in the granulosa cells. Ligand-dependent ERβ actions may also lead to decreased AR expression, allowing the role of A4 and T to shift from that of a ligand for AR-mediated actions to that as substrates for E2 synthesis. Increasing E2 levels eventually activate ERα in thecal cells to reduce further androgen synthesis by decreasing CYP17 activity, forming a negative feedback loop to ensure the proper estrogen- to-androgen ratio in the follicle. LH Surge: Climaxing E2 levels in the circulation facilitate the release of the LH surge from the pituitary, which stimulates ovulation and terminal differentiation of the granulosa cells. Only those follicles that possess properly differentiated granulosa cells (i.e., have acquired sufficient LH-R) are able to respond to the LH surge, which causes a dramatic rise in PR expression that occurs in the preovula- tory granulosa cells within hours of the LH surge, whereas ERβ and CYP19 levels are both dramatically decreased. is effectively illustrated by the 12-fold reduction in Cyp17 There is ample evidence that granulosa cell–derived E2 expression in ovaries of LH-receptor (Lhcgr)-null mice,563 may mediate a feedback loop on thecal cells to decrease as well as the five-fold increase in expression in the ova- further androgen synthesis in thecal cells,566 however, ries of transgenic mice that possess chronically elevated the mechanism remains unclear. Magoffin posed the LH levels.439 Likewise, chronically elevated plasma LH question several years ago as to whether E2 inhibition of leads to a comparable increase in Cyp17 expression and androgen synthesis occurs at the level of steroidogenic androstenedione synthesis in both ERα-null and ERβ- enzyme expression or via direct inhibition of enzymatic null ovaries and cultured follicles, indicating that neither activity.561 Indeed, estrogens have been shown to inhibit ER is obligatory to the positive regulation of thecal cell CYP17 activity in both thecal and Leydig cell prepara- steroidogenesis.431,439,564 tions while having no effect on activities of the upstream Therefore, androgens are necessary to serve as both steroidogenic enzymes.566–571 Early studies found that E2 a stimulus of and substrate for aromatase activity in can compete with androstenedione for CYP17 enzymatic granulosa cells. However, a proper androgen:estrogen activity;568 however, this effect could not be reproduced ratio is critical as an intrafollicular milieu of increased in ovarian lysates in vitro.572 Additional studies in dis- androgens is strongly associated with atresia.565 It is gen- persed rat ovarian cells found that E2 has no effect on erally believed that an intraovarian mechanism exists to the number of LH receptors or level of hCG-stimulated negatively modulate thecal cell androgen synthesis so cAMP synthesis, further supporting a direct E2 inhibition that levels do not surpass the aromatase capacity of the of CYP17 activity as a mechanism for regulating andro- granulosa cell layer. Indeed, defects in this intraovarian gen synthesis.570 However, other evidence supports a mechanism are postulated to be an underlying cause of role for ER-mediated regulation of Cyp17 expression the excess thecal cell androgen synthesis that is a hall- at the transcriptional level, including demonstrations mark of polycystic ovarian syndrome (PCOS) in women. that an antiestrogen blocks the expected drop in CYP17

4. FEMALE REPRODUCTIVE SYSTEM 1158 25. STEROID RECEPTORS IN THE UTERUS AND OVARY activity that occurs just prior to ovulation in rats,573 and detrimental effects of aromatase inhibition such as those that E2 significantly reduces Cyp17 expression in the tes- reported in the rat ovary are not observed in similarly tes of rats574 and fish.575 Our findings that ERα-null ova- treated hamsters, rabbits, or monkeys.486 ries exhibit increased Cyp17 expression and enzymatic Multiple investigations of the ovulatory capacity of activity despite a milieu of significantly elevated E2 also ER-null and CYP19-null mice have been conducted. supports a mechanism of ERα-mediated suppression of As described before, ERα-null females are anovulatory Cyp17 transcription in thecal cells versus direct inhibi- throughout life and gonadotropin-induced ovulation is tion of CYP17 activity.431 Furthermore, Taniguchi dem- unsuccessful in ERα-null females at 4 months of age when onstrated that in vitro–cultured ERα-null follicles have the ovarian cystic phenotype is advanced.373 However, increased Cyp17 expression similar to that observed in immature (≤28 days of age) ERα-null (αERKO) females WT follicles cultured in the presence of an aromatase do respond to induced ovulation and yield recoverable inhibitor.564 Treatment of ERα-null follicles with E2 or oocytes in the oviduct at an average yield of 14.5 oocytes/ an ERα-specific agonist as well as an aromatase inhibitor ERα-null female versus 40.6 oocytes/wild-type female.415 reduced Cyp17 expression to that observed in wild-type Similar data has been reported in ERα-null females as old vehicle treated follicles.564 Therefore, the excess andro- as 5 weeks of age.415,430 Interestingly, Dupont et al. report gen synthesis observed in ERα-null ovaries is due to the that immature females from their ERα-null (ERαKO) loss of ERα-mediated suppression of Cyp17 expression, line exhibit a total failure to ovulate in response to exog- indicating the existence of intraovarian feedback actions enous gonadotropin treatments, although this study of E2 that are mediated by ERα and critical to maintain- involved only three females.158 Preliminary studies sug- ing homeostasis of androgen levels in the female mouse. gest Ex3αERKO mice159 have a significantly reduced ovu- lation response, where less than 50% of mice ovulate in OVULATION response to exogenous gonadotropins, with few oocytes E2 clearly facilitates the acquisition of several gonado- produced from those that do ovulate (Korach laboratory, tropin-dependent effects during folliculogenesis, includ- unpublished data). These observations suggest that the ing: (1) the proper cellular organization of the follicle, e.g., response observed previously415 was possibly due to the antrum and cumulus oocyte complex; (2) the necessary presence of residual ERα splice variant expressed in the enzymatic activity, e.g., E2 synthesis; and (3) the essen- αERKO mice,159,189 thus further studies are underway to tial receptor signaling pathways; e.g., LH receptor. All of confirm these preliminary studies. these characteristics define a healthy differentiated follicle The cause of the reduced ovulatory response to exog- and are necessary for a proper response to the gonadotro- enous gonadotropins in ERα-null females is unclear and pin surge and expulsion of a competent oocyte. However, suggests a facilitatory role for ERα in follicular rupture. a need for estrogen signaling in the ovulating follicle or However, the expected induction of several genes that following the postgonadotropin surge remains unclear. are critical to follicle rupture, such as P receptor, prosta- Selvaraj et al.485 found that coadministration of immature glandin-synthase-2,576 occurs in gonadotropin-primed rats with PMSG and an aromatase inhibitor (Fadrazole) ERα-null ovaries shortly after hCG treatment.415,432 Fur- leads to a reduced number of healthy antral follicles that thermore, ERα-null ovaries form functional corpora lutea culminates with a severely reduced ovulatory response to following induced ovulation, indicating that granulosa a bolus of hCG. These data support the intrafollicular role cells of ERα-null preovulatory follicles possess the capac- of E2 in preparing competent preovulatory follicles for ity to respond to LH. Therefore, the reduced ovulatory ovulation. However, in an additional experiment, Selvaraj response in ERα-null ovaries is more likely attributable et al. once again treated immature rats with PMSG but to premature increases in endogenous plasma LH415 and delayed administration of the aromatase inhibitor by 40 h subsequent ovarian androgen synthesis.415,430 Although such that exposure occurred just 6 h prior to hCG induced immature ERα-null ovaries do not yet manifest the overt ovulation.485 These animals exhibit an equally poor morphological effects of LH-hyperstimulation, e.g., cys- response in terms of follicular rupture and recoverable tic follicles, they do exhibit elevated LH-receptor lev- oocytes from the oviduct; and this effect is abated by coad- els.415 Cumulative damage in the ovary from chronic LH ministration of E2 just prior to induced ovulation. These hyperstimulation likely causes the age-related reduction data in turn suggest that E2 does facilitate the ovulatory in gonadotropin-induced ovulatory capacity in ERα-null process. However, inhibition of E2 synthesis is known to females discussed previously. Indeed, Armstrong and col- cause severe intraovarian485 and intrafollicular523 accu- leagues demonstrated in rats that treatment with PMSG mulations of P and androgens that could also lead to a preparations possessing increased LH activity lead to a poor ovulatory response. In contrast to the above in vivo reduced ovulatory response to subsequent hCG stimu- findings, Hu et al. reports that isolated mouse follicles lation577,578 and that this effect is likely due to inappro- grown exposed to an aromatase inhibitor in vitro exhibit priately high stimulation of androgen synthesis during a normal ovulatory response to hCG, including mucifi- follicle maturation.578 Studies showing ERα-null follicles cation of the cumulus–oocyte complex.523 Furthermore, allowed to differentiate in vitro under controlled hormonal

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1159 stimulation exhibit a rate of hCG-induced rupture that is to increase the ovulation of these follicles in vitro,455 comparable to similarly cultured wild-type follicles fur- suggesting that ERβ is necessary for maximal ovulatory ther indicating that any role for ERα in the ovulatory pro- response downstream of gonadotropins, although ERβ cess is minimal.440 The ERα was found to be necessary for is not directly contributing to follicle rupture. ovulatory response in aging animals, as thEsr1KO mice Evidence of any cooperative action between ERα and had normal ovulatory response at 2 months that was ERβ in ovulation may be derived from studies of the severely dampened by the time the mice were 6 months compound ER-null and CYP19-null mice. The single old.362 The premature loss of fertility in these mice sug- report of induced ovulation in immature ERα/ERβ-null gests that ERα is important for normal ovulatory function; (ERαβKO) females describes a total failure of ovulation.158 however, the precise mechanism has yet to be elucidated. Not surprisingly, ovaries from these females exhibited The loss of functional ERβ clearly leads to ovarian underdeveloped preovulatory follicles that failed to defects that are inhibitory to ovulation. Both lines of ERβ- exhibit cumulus–oocyte complex expansion or luteiniza- null females exhibit severely reduced fertility due to ovar- tion,158 indicating an attenuated response to hCG similar ian defects; however, unlike ERα-null females, ERβ-null to that in ERβ-null ovaries. Investigations in both lines animals do spontaneously ovulate.158,171 Krege et al. found of Cyp19-null females produced comparable results of a that immature ERβ-null (βERKO) females exhibit a severely total failure to ovulate and luteinize.131,580 Interestingly, reduced oocyte yield of 6 oocytes/female versus 34 Huynh et al. remarked that cumulus–oocyte complexes oocytes/female among age-matched wild-types following lacking oocytes were recovered from the oviducts of gonadotropin-induced ovulation.171 Dupont et al. report 4-week-old Cyp19-null females following induced ovu- a similar average yield in immature ERβKO females (17.6 lation, indicating that some elements of follicle rupture versus 37 oocytes/female), even when excluding those may occur after degradation of the oocyte.580 The more females that do not ovulate at all.158 A third line of ERβ-null severe ovulatory phenotypes in compound ER-null and mice was found to be unresponsive to exogenous gonado- Cyp19-null females suggest that the intraovarian func- tropins, and no animals were able to ovulate (n = 19).172 tions of both ERα and ERβ are necessary for ovulation Collectively, even though some animals are able to respond but not necessarily for the process of follicular rupture. with some oocytes ovulated, the models demonstrate the Recent studies in Cyp19-null mice suggest that treat- necessity of ERβ for optimal ovulatory response. ment with E2 as well as exogenous gonadotropins could Immature ERβ-null ovaries collected 20 h after stimulate some ovulation,365 suggesting the importance hCG-induced ovulation reveal multiple preovulatory of the steroid and gonadotropin hormones in regulating follicles that retain their oocyte and fail to undergo ovulation. expansion of the cumulus–oocyte complex.158,171,432 α Furthermore, hCG induction of prostaglandin-syn- ER MEDIATED REPRESSION OF LEYDIG CELL thase 2 and PR, two events that are critical to follicular DEVELOPMENT IN THE OVARY rupture, are significantly compromised in immature As discussed in earlier sections, female mice lacking βERKO ovaries.432 ERβ-null follicles induced to ovu- ERα action due either to the loss of receptor (αERKO, late in vitro exhibit an equally poor rate of rupture and thEsr1KO, and αβERKO) or ligand (Cyp19-null) exhibit gene induction following hCG exposure.440 The sum of abnormally high levels of plasma T. For example, the in vivo and in vitro data from ERβ-null mice strongly average plasma T level in adult αERKO and αβERKO indicates an inability to fully respond to an endogenous females is 5.5 ng/ml and 2.1 ng/ml, respectively, versus LH surge or an exogenous bolus of hCG. This is likely 0.14 ng/ml in wild-type females.431 In females of the two due to insufficient acquisition of LH-receptor expres- existing Cyp19-null lines, Fisher et al. report average sion among granulosa cells of ERβ-null preovulatory plasma T levels in mutant females to be 2.3 ng/ml ver- follicles.432,440,455 Indeed, Clemens et al. found that the sus 0.24 ng/ml in wild-type;156 while Toda et al. report ER-antagonist ICI 164,384 inhibits FSH-induced differ- 1.4 ng/ml in the mutant females versus 0.13 ng/ml in entiation of rat granulosa cells in vitro, as illustrated wild-type.131 These levels of plasma T in ERα-null and by a severely compromised induction of PR expres- Cyp19-null females surpass the lower limits character- sion following stimulation of the cAMP/PKA signaling istic of normal male mice (2–10 ng/ml)131,156,581 and are pathway.579 However, if exposure to the ER antagonist obviously very uncharacteristic of normal females. The is delayed until the time of ovulatory stimulation, lit- mouse ovary possesses the enzymatic machinery nec- tle effect is observed,579 further supporting a role for essary for T synthesis, although it is not a major prod- ERβ in granulosa cell differentiation and LH-receptor uct. The two-cell, two-gonadotropin paradigm of ovarian expression rather than during ovulation. ERβ-null fol- steroidogenesis (Figure 25.21) postulates that thecal licles cultured in vitro also have a reduced ovulatory cell–derived androstenedione serves as a substrate for response and reduced cAMP accumulations after FSH two sequential reactions in granulosa cells, CYP19- stimulation.455 Interestingly, treatment of these fol- mediated aromatization to estrone followed by a reduc- licles with forskolin to increase cAMP levels was able tion to E2 by 17β-hydroxysteroid dehydrogenase type I

4. FEMALE REPRODUCTIVE SYSTEM 1160 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

(HSD17β1).561 The murine form of 17β-HSD 1 is reported of Leydig-like cells in the ovaries.479 Therefore, ERα may to efficiently reduce androstenedione to T as well as be critical to promoting thecal cell differentiation and estrone to E2; whereas the human form may not possess inhibiting the Leydig cell phenotype during ovarian dif- this activity.582 Therefore, the accumulation of andro- ferentiation. A similar phenotype of Leydig cell differen- stenedione that occurs in ERα-null and Cyp19-null ova- tiation, Hsd17b3 expression, and “masculinization” of the ries due to LH hyperstimulation of the theca in both, and ovary and reproductive tract is reported in female mice the absence of aromatization in the latter,156,431 provides lacking Wnt4 function.466 for sufficient substrate for 17β-HSD 1 conversion to T. However, further investigations have discovered that DISRUPTED ESTROGEN SIGNALING IN HUMANS ERα-null431 and Cyp19-null479 ovaries uniquely possess There is one case of a human female that is autosomal substantial levels of 17β-hydroxysteroid dehydrogenase recessive for an inactivating mutation of the ESR1 (ERα) type III (Hsd17b3), a Leydig cell–specific enzyme in tes- but no patients with ESR2 (ERβ) mutation reported. tes that specifically functions to reduce androstenedione There are cases of human females that are autosomal to T.583–587 Remarkably, the level of Hsd17b3 expression recessive for inactivating mutations of the CYP19 gene in αERKO, αβERKO,431 and Cyp19-null479 ovaries is and therefore lack the ability to synthesize E2.591–597 The comparable to that found in testes of wild-type males. CYP19 mutated females present with androgen-induced Therefore, the abnormally high capacity of ERα-null and pseudohermaphrodism at birth,591–593 and amenarche, Cyp19-null ovaries to synthesize T may be attributed to polycystic ovaries, and an absence of female second- the accumulation of androstenedione and its conversion ary sex characteristics at puberty.591,593,594 Three reports to T by 17β-HSD 3 rather than 17β-HSD 1. describe elevated plasma FSH and LH levels, leading Hsd17b3 expression in Leydig cells is primarily LH to hypergonadism that manifests as increased plasma regulated.585 The ectopic expression in ERα-null ovaries androgens and virulization.591,593,594 The female with is also dependent on LH hyperstimulation as expression Esr1 mutation has polycystic ovaries and elevated E2. is eradicated following prolonged treatment of animals The ovarian pathology exhibited is somewhat similar to with a GnRH antagonist.431 In ERα-null females, Hsd17b3 that observed in ERα-null females and compatible with a expression is limited to the ovarian interstitium and possi- diagnosis of PCOS.593 To date there are no indications of bly the theca.588 However, it is unclear whether these cells “sex reversal” in terms of male cell types or gene expres- are Leydig-like and suggestive of an “organizational” sion in the ovaries of CYP19-null human females. Estro- defect in ERα-null ovaries or are thecal/interstitial type gen and P replacement effectively alleviates all of the cells within which the loss of ERα leads to loss of Hsd17b3 above phenotypes.591,593,594 repression, i.e., an “activational” defect. Evidence to sup- Polymorphisms of the human ESR1 gene have been port the latter hypothesis includes: (1) testes from ERα- linked to breast cancer, cardiovascular disease, osteo- null males also exhibit elevated Hsd17b3 expression and porosis cognitive function, and multiple reproductive 589 598–603 T synthesis, (2) female mice possessing chronically anomalies. A polymorphic (TA)n repeat within the elevated LH but intact ERα signaling do not exhibit ovar- ESR1 promoter region has been linked to premature ian Hsd17b3 expression,439 and (3) prolonged E2 therapy ovarian failure604 but has no effect on plasma levels of in Cyp19-null females eradicates Hsd17b3 expression.479 E2, sex-hormone binding globulin (SHBG), T, or FSH in In turn, there exists supporting evidence for the former premenopausal women during the follicular phase.605 hypothesis that a defect during ovarian differentiation Two PvuII single nucleotide polymorphisms (SNP) in the may occur in the absence of ERα. Thecal and Leydig cells ERS1 gene have been described to date. The first of these are believed to derive from a common bi-potent embryo- is a glutamine-to-glycine transformation at amino acid logical precursor cell during gonadal differentiation. In 400606 and has received little investigative attention. The Cyp19-null ovaries, Britt et al. have described the pres- second is an anonymous SNP located in intron 1, 400 bp ence of cells resembling the mature Leydig cells of tes- upstream of exon 2,607 and is found in approximately 30% tes, in that they possess a tubulovesicular arrangement of women.608–612 Although the intronic ESR1 PvuII SNP of mitochondrial cristae, an annular nucleolus, and an does not affect ERα receptor levels,607 it may be associated abundance of smooth endoplasmic reticulum in whorl- with: (1) poor performance in women undergoing in vitro like formations.194 These Leydig cell–like ultrastructural fertilization,608,612 (2) increased plasma E2 and andro- features were also observed in ERα-null ovaries.588 These stenedione levels,610,612 and (3) late onset of menarche Leydig-like cells in Cyp19-null ovaries are located in the and/or menopause.609,611,613,614 Two SNP of the ESR2 gene interstitial regions and often proximal to the basement have been described to date, an RsaI SNP leads to a silent membrane of atretic follicles that exhibit Sertoli-like cells nucleotide change in exon 5, and an AluI SNP occurs in in the follicle lumen.194 Indeed, E2 is a potent inhibitor of the 3′-untranslated region of exon 8.615 Sundarrajan et al. Leydig cell proliferation in vitro and in vivo,590 and E2 found that Chinese women that are homozygous for replacement in Cyp19-null females reduces the presence either the RsaI or AluI SNP of the ESR2 gene are more

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1161 prone to exhibit ovulatory defects, menstrual disorders, relatively well characterized. Over the course of the and elevated LH; while women that are homozygous for estrous cycle in rats, Pgr mRNA levels are increased >30- both exhibit more severe idiopathic ovulatory defects.616 fold during a 4 h period (1800–2200 h) on the evening of proestrus and then return to baseline shortly after 628 Progesterone Receptor Signaling in Ovarian (2400 h). The increased expression is almost exclusive to the mural granulosa cells of healthy preovulatory fol- Function licles in rats,628,633 indicating that only fully differenti- PR Expression in the Ovary ated granulosa cells possess the capacity for such rapid Pgr The first report of saturable, high-affinity P binding induction of the gene (Figure 25.22). Although PR sites in the rat ovary is that of Schreiber and colleagues expression climaxes within hours after plasma E2 and LH levels climax, the following evidence indicates LH is in 1979617,618 and has since been followed by similar most directly involved: (1) expression peaks just 2 h after descriptions in human389–391,619 and cow619,620 ovaries. plasma LH levels climax versus >8 h after peak plasma Later immunohistochemical studies primarily localized E2 levels,628 (2) hCG treatment 48 h after PMSG exposure ovarian PR expression to the theca of large preovulatory leads to a comparable increase in PR expression, whereas follicles, the surface epithelia, and stromal/interstitium, PMSG and concomitant increases in E2 have a minimal and found this expression pattern is well conserved effect,415,621,628 among species, including mice,621,622 pig,623 mon- (3) treatment of hypophysectomized or immature rats, or differentiated rat granulosa cells with keys,386,624 and humans.392,625,626 Numerous studies in E2 has no marked effect on ovarian PR expression,422,634 species ranging from rodents to large domestic animals (4) inhibition of the LH surge by pentobarbital abates the and primates agree that granulosa cells possess mini- increase in PR expression despite having no effect on the mal PR expression throughout folliculogenesis except rise in plasma E2 levels,628 (5) LH but not E2 elicits a dra- for the period 2–4 h after the preovulatory gonadotropin matic PR induction in differentiated granulosa cells from surge, whereupon an enormous induction of PR expres- rat634,635 629 sion occurs in preovulatory granulosa cells and peaks or porcine ovaries in vitro, (6) an ER antago- just prior to follicle rupture386,392,415,621–624,627–630 (Figure nist (ICI 164,384) does not inhibit forskolin-induced PR 579 (7) 25.22). This dramatic increase in PR expression in the expression in differentiated rat granulosa cells, Pgr granulosa cells of ovulatory follicles is absolutely essen- the putative ERE-like region of the rat promoter α β 579 182,631 does not bind ER or ER in ovarian extracts, and tial to follicle rupture. This induction of PR expres- β sion in ovulating follicles is quite transient and lasts less (8) ER (mainly ER ) levels are rapidly decreasing at the 415,621,628 time of rising PR expression in preovulatory granulosa than 12 h in rodent ovaries but is maintained 367,371,380,415 throughout luteinization and corpus luteum formation cells. in porcine623 and primate386,392,624 ovaries. The above data provide a convincing argument that The regulation of ovarian PR expression prior to the the rapid and transient induction of PR expression in preovulatory surge is poorly characterized. PMSG treat- preovulatory granulosa cells is directly mediated by the ment of immature rats leads to a four-fold increase in cyto- LH surge. Clearly only granulosa cells of preovulatory follicles in vivo or fully differentiated granulosa cells solic P binding sites632 and a modest two-fold increase in vitro possess the capacity to respond to the LH surge in Pgr mRNA levels628 after 48 h, indicating that FSH or and increase PR expression, presumably due to their subsequent E2 signaling may stimulate these basal levels unique acquisition of LH receptor.579,634 Indeed, granu- of PR expression. In contrast, the mechanisms underly- losa cells isolated from estrogen-primed immature rats ing the dramatic induction of PR expression in preovu- require a 48-h exposure to physiological concentrations latory granulosa following the gonadotropin surge are of FSH and T to acquire LH induction of PR expression.579 LH-mediated induction of the intracellular cAMP/pro- tein kinase-A pathway is likely the principal stimulus for increased PR expression in preovulatory granulosa cells as large doses of FSH, forskolin, or (Bu)2cAMP also elicit a comparable response in rat ovaries in vivo634 and dif- ferentiated rat634,635 or porcine629 granulosa cells in vitro. However, GnRH, PMA, or EGF exposure of differenti- ated rat granulosa cells in vitro elicits a comparable FIGURE 25.22 Regulation of PR expression in mouse ovaries induction and temporal pattern of PR expression, sug- and granulosa cells during gonadotropin-induced ovulation. (A, B) gesting that stimulation of protein kinase C or tyrosine Immunohistochemistry for PR in the ovary of a PMSG-primed mouse 635 4 h after hCG treatment, illustrating the dramatic induction of PR kinase signaling cascades may also be involved. In fact, immunoreactivity in the granulosa cells of preovulatory follicles (POF) Sriraman et al. report that forskolin and PMA synergize but not surrounding preantral follicles. (B, higher magnification of A.) to induce increased PR expression in undifferentiated

4. FEMALE REPRODUCTIVE SYSTEM 1162 25. STEROID RECEPTORS IN THE UTERUS AND OVARY primary rat granulosa cells.636 However, studies in fully immunoreactivity is restricted to the thecal cells of prean- differentiated rat granulosa cells demonstrate that for- tral and antral follicles prior to the gonadotropin surge; skolin-induction of PR expression is fully blocked by whereas PRB immunoreactivity is detected in both thecal co-treatment with a PKA inhibitor (H89) but only mini- and mural granulosa cells throughout the stages of follicu- mally affected by a PKC inhibitor (Cal-C),579 indicating logenesis and shows less variation over the course of the a greater importance for the former signaling pathway. estrous cycle.622 The enormous induction of PR expression The mechanisms by which LH activation of the PKA that occurs in preovulatory granulosa cells on the evening pathway leads to increased PR expression in preovula- of proestrus is primarily representative of PRA, resulting tory granulosa cells remain unclear. The Pgr promoter is in a 2:1 ratio of PRA:PRB in these cells.622 In turn, the more complex, consisting of distal and proximal regions and moderate LH-induced increases in PR expression in the- able to give rise to multiple transcripts.31,33 Initial stud- cal cells is predominantly PRB.622 During the metestrous ies concluded that an estrogen-response element-like stage and rising P synthesis that follow ovulation, PRB is region (ERE3) within the proximal rat Pgr promoter is the sole isoform in thecal cells of preantral and antral fol- important to LH induction,579 but later studies that better licles.622 The differential expression of PRA and PRB sug- modeled the structure of the rat Pgr promoter indicated gests that specific roles exist for each isoform. these sequences are dispensable.636 Sriraman et al.636 To examine this possibility, Sriraman et al. recently demonstrated that two Sp1/Sp3 binding sites within the examined genes that were induced by PRA or PRB in proximal promoter of the mouse Pgr gene bind Sp1/Sp3 granulosa cells in vitro. In the absence of P ligand, PRA and are essential to forskolin/PMA-induced expression. regulated 41 genes, while PRB only showed differential Furthermore, LH induction of Pgr also involves MAPK expression of seven genes compared to control cells.639 activation as mice lacking ERK1/2 in granulosa cells do When the granulosa cells were treated with a PR ligand not induce expression of Pgr in response to an hCG ovula- in vitro, PRA-transduced cells showed an increase in 42 tory signal.637 While these mice show reduced expression, genes, while PRB-transduced cells only showed 27 tran- the direct targets of ERK1/2 remain unknown at this time. scripts that were differentially expressed, while only Recent evidence in porcine granulosa suggests that andro- three transcripts were shared between the two PR iso- gens may regulate Pgr expression, where treatment with forms.639 The differential expression of genes in granu- either T or an AR antagonist alone reduces Pgr expression, losa cells transduced with either PR isoform supports while co-treatment leads to increased Pgr expression.638 the hypothesis that specific roles exist for each isoform Further characterization into the molecular mechanisms and provide target genes for future study. involved in this regulation is necessary. Therefore, unlike the uterus where Pgr is estrogen regulated, estrogen does Ovarian Phenotypes in Mouse Models of Disrupted not have the same effect on ovarian Pgr. Progesterone Signaling Differential expression of the two PR isoforms, PRA and Several mouse lines have been developed that are null PRB, among thecal and granulosa cells further illustrates for both PR forms and exhibit grossly normal ovaries but the complexity of PR expression in the ovary. Western blot are anovulatory and therefore infertile176–178 (Table 25.8). analyses following LH induction indicate a PRA:PRB ratio Furthermore, gonadotropin-induced ovulation of PR-null of 3:1 in differentiated granulosa cells from rats635 and 2:1 females is unable to rescue the anovulatory phenotype as ratio in whole ovaries from mice.621 In rat ovaries, PRA no oocytes could be recovered following treatment.176,178

TABLE 25.8 Ovarian Phenotypes in PR-Null Mouse Models

Mutated or Null for Sex Steroid Signaling Ovarian Phenotypes Hormone Levels References

Pgr−/− (homozygous null alleles for Anovulatory and infertile Elevated LH 176–178 PRA and PRB: PRKO) Failure to respond to superovulation— Normal FSH large follicles present in ovary with trapped oocytes No follicle rupture No CL

PRA−/− (homozygous null alleles for Anovulatory and infertile Not reported 180 PRA: PRAKO) Reduced to no response to superovulation Large follicles present with trapped oocytes similar to PRKO No follicle rupture

PRB−/− (homozygous null alleles for Fertile Not reported 182,227 PRB: PRBKO) Ovaries appear normal Respond to superovulation similar to wild-type

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1163

Instead, PR-null ovaries following induced ovulation females lacking PRA ovulate an average of 9 oocytes/ exhibit multiple large, preovulatory follicles, each pos- female versus 32 oocytes/female among wild-type litter- sessing a healthy, competent oocyte but a total failure mates; whereas PRB females exhibit no deficits in oocyte to rupture176,178 (Figure 25.23). The follicular cells of the yield.180,182,227 This finding that the PRA isoform is more unruptured preovulatory follicles undergo luteinization critical to ovulation180 is congruent with the above studies and form functional corpora lutea within the expected time of Gava et al. that demonstrate this isoform is much more frame, suggesting that some elements of an LH response dramatically induced by the LH surge in preovulatory are intact177,178,182,633 (Figure 25.23). Similar experiments granulosa cells. This also supports the finding that PRA in isoform-specific PR-null females indicate a more severe in the presence of ligand regulates more genes in granu- phenotype following the loss of PRA relative to PRB, as losa cells in vitro compared to cells expressing PRB.639 Still, the absolute failure of ovulation even after gonadotropin induction in females lacking both PR forms indicates a level of cooperation between PRA and PRB that is vital to follicular rupture.180 Recently, a mouse model was devel- oped to use for tissue-specific deletion of PR,177 however to date only global knockout models have been published that show a similar phenotype as previously reported.176,178 Intraovarian Role of Progesterone in Ovarian Function FOLLICLE GROWTH AND DIFFERENTIATION Although immunohistochemical studies indicate that both PR isoforms are expressed at basal levels through- out thecal and granulosa cells,622 there is little evidence that P action profoundly influences follicle growth and differentiation in the rodent ovary. Administration of P or synthetic progestins to immature or hypophysecto- mized rats has no obvious effect in the ovary in terms of estrogen receptor levels358,422 or gonadotropin/estrogen- induced follicle growth and maturation.358,640 In addi- tion, P has no measurable influence on FSH-induced granulosa cell proliferation496 or growth and differentia- tion of whole follicles497 in vitro. Perhaps most indicative of a minor role for P signaling during folliculogenesis prior to ovulation is the lack of any overt phenotype in follicle growth, maturation, or steroidogenesis in the ovaries of PR-null mice (Table 25.8).176–178,641 Indeed, FIGURE 25.23 Ovarian response to gonadotropin-induced ovula- Robker et al. demonstrated that the expected FSH- or tion in PR-null mice. (A) Transverse section of a typical ovary isolated PMSG-induced increases in Cyp19 and Lhcgr expression from a 6-week-old wild-type mouse (+/+) treated with an intraperito- in mature follicles occurs in PR-null ovaries.633 neal injection of 5 IU of PMSG, followed 48 h later with 5 IU of hCG, and killed 24 h later. Note the presence of numerous corpora lutea (CL). OVULATION Scale bar = 100 μm. (B) A representative cross-section of an ovary from an age-matched PR-null (−/−) female that was hormonally treated PR expression in the ovary is relatively low through- exactly as the wild-type described above. Note the unusual presence out folliculogenesis except during the 4–6 h period imme- of several unruptured follicles (UF) as indicated. Scale bar = 100 μm. (C) diately after the ovulatory gonadotropin surge, during High magnification of a corpus luteum in the wild-type ovary, exhibit- which an enormous induction of PR expression occurs μ ing the characteristic hypertrophied luteal cells (LC). Scale bar = 50 m. in granulosa cells of ovulating follicles628,634,635 and is (D) High magnification of an unruptured follicle present in a PR-null ovary, exhibiting an intact oocyte (O) with a zona pellucida and granu- synchronized with an equally acute increase in ovar- losa cells (GC) that have undergone cumulus expansion. Note the lack ian P synthesis. These phenomena alone are indicative of luteinization among the GCs in the PR-null follicle compared to of an important role for PR-mediated P signaling dur- those in the wild-type (C). Scale bar = 50 μm. (Source: (A–D) reproduced ing ovulation. Experimental evidence of a critical role with permission from Ref. 176.) (E) Average number of oocytes (±SEM) for PR in ovulation comes from studies in which ovula- released per mouse after gonadotropin-induced ovulation in wild-type (WT), PRA-null (PRAKO−/−), PRB-null (PRBKO−/−), and PR-null tion is blocked when peri-ovulatory ovaries or follicles 642 (PRKO−/−) mice; n = 8 per test group. Source: Reproduced with permis- are exposed to: (1) anti-P antisera, (2) inhibitors of P sion from Ref. 182. synthesis,643–645 or (3) PR antagonists.621,646,647 Finally, the

4. FEMALE REPRODUCTIVE SYSTEM 1164 25. STEROID RECEPTORS IN THE UTERUS AND OVARY anovulatory phenotype of PR-null mice discussed before coinciding with a period of follicular rupture.644,649 A provides definitive evidence that PR actions are required similar pattern of peri-ovulatory ADAMTS-1 expression in the mammalian ovary during the period just prior is documented in preovulatory follicles of porcine,650 to ovulation.176–178,180 PR-null females exhibit multiple equine,630 and primate651 ovaries. Cathepsin-L (Ctsl) is large, preovulatory but unruptured follicles following also significantly increased following hCG exposure in induced ovulation.176 Therefore, follicle growth and dif- granulosa cells of preovulatory follicles and also peaks ferentiation are unaffected by the loss of PR functions as 12 h after treatment.633,652 PR-null females,633 as well as PR-null follicles exhibit fully formed antra and increased wild-type female rats exposed to a P synthesis inhibitor Cyp19 and Lhcgr expression following FSH or PMSG near the time of hCG treatment644,653 fail to exhibit hCG- treatment.176,633 Furthermore, although follicle rupture stimulated increases in Adamts-1 expression, indicating is impaired, PR-null ovaries exhibit several indications this effect of LH is primarily dependent on PR-mediated that other LH- or hCG-induced responses are intact, P action. ADAMTS-1-null female mice exhibit a severely including: (1) rapidly decreased Cyp19 expression,633 (2) compromised ovulation and a phenotype of large, pre- dramatic induction of Ptgs2 expression,633 (3) expan- ovulatory but unruptured, follicles following gonado- sion of the cumulus–oocyte complex,176 and (4) lutein- tropin-induced ovulation.654,655 ization and formation of functional corpora lutea.182,633 Interestingly, PRA-null mice exhibit a severe ovula- These findings in PR-null females are congruent with tory phenotype similar to total PR-null females but pos- earlier studies using P synthesis inhibitors645,648 or PR sess normal LH-stimulated induction of Adamts-1 and antagonists,646 which also report that ovulatory levels Ctsl during induced ovulation,182 suggesting possible of prostaglandins and steroids are largely unaffected by compensatory actions by PRB but also questioning the inhibition of P signaling. importance of ADAMTS-1 and cathepsin-L to follicle The previous data clearly indicate that PR actions rupture. ADAM8 was recently shown to be regulated are highly specific and critical for follicle rupture. Stud- specifically by PRA in reporter assays, while PRB was ies have discovered that PR-null ovaries fail to exhibit unable to regulate the promoter of ADAM8 in vitro.649 LH-induced expression of several proteases, includ- This could explain why PRA-null mice show LH-stim- ing ADAMTS-1, ADAM8, and cathepsin-L, which are ulated induction of Adamts-1 and Ctsl yet are unable to postulated to be involved in degradation of the follicle ovulate, although further studies are needed to confirm wall and extracellular matrix that is necessary for oocyte this. extrusion (Figure 25.24). ADAMTS-1 (Adamts1) and Several other genes have been identified downstream ADAM8 are members of the A disintegrin and metal- of PR and suggested to be important in follicle rupture loproteinase family of proteases and are dramatically and ultimately ovulation. Since ovulation is discussed increased in granulosa cells of preovulatory follicles more fully in another chapter (Chapter 22) of this vol- 12 h after hCG treatment, after peak PR expression and ume, we will not provide in-depth description, although

FIGURE 25.24 Ovarian expression of ADAMTS-1 in PR-null ovaries during gonadotropin-induced ovula- tion. Immature heterozygous (PR+/−) and homozygous PR-null (PRKO) mice were left untreated or injected with PMSG and euthanized 46 h later, or were treated with PMSG for 46 h followed by hCG and euthanized 7, 12, and 24 h later. Reverse transcriptase polymerase chain reac- tion analysis of total RNA from whole ovaries indicates that ADAMTS-1 expression is dramatically increased by hCG in PMSG-primed mice, but this induction is lacking in PR-null ovaries, indicating dependence on PR function. ADAMTS-1 expression was normalized to expression of ribosomal protein L19. Source: Reproduced with permission from Ref. 633.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1165 we will briefly describe what is known to date. These In the ovaries of multiple species, Ar/AR expression include SNAP25, synaptosomal-associated protein 25, among growing follicles is inversely correlated with the an LH-regulated gene that has decreased expression in extent of granulosa cell differentiation (Figure 25.25). A PR-null mice but not Ptgs2 mice.656 Furthermore, Snap25 mechanism to decrease AR levels during follicle matura- was PR regulated in in vitro reporter assays, and the PR tion is consistent with the need to reduce sensitivity to antagonist RU486 was able to inhibit this regulation. intrafollicular androgens, which rise to levels sufficient SNAP25 acts to alter cytokine and chemokine secretion for aromatization to E2, as activating AR ligands are det- in granulosa cells, which may act to facilitate follicle rup- rimental to the health of preovulatory follicles.680 In the ture.656 Endothelin 2 (Edn2) was identified to be an LH- ovaries of gonadotropin-stimulated rats, Testsuka and regulated gene with undetectable levels in PR-null mice Hillier found that large antral follicles (>400 μm) exhibit that plays a role in follicle rupture.657,658 Recent studies a 2.75-fold higher level of Cyp19 expression relative to have shown that Edn2 is regulated by PRA,639 and stud- small follicles (<200 μm) but a 51% decrease in Ar expres- ies in rats have shown that it acts to stimulate smooth sion.665 In rat ovaries, Ar expression is first apparent in muscle contraction and follicle constriction.659 EDN2 early postnatal development in follicles of the preantral provides the mechanical actions that help to facilitate fol- stage.681 In adult ovaries, decreasing AR immunoreac- licle rupture and successful ovulation. This is supported tivity occurs during follicle differentiation; a decrease by its regulation specifically by PRA, which is essential is first apparent in mural granulosa cells and then pro- for ovulation, whereas loss of PRB doesn’t reduce ovula- gresses in those cells closest to the antrum.667 Interest- tory response as described previously. ingly, cells composing the cumulus–oocyte complex of Numerous transcription factors have also been iden- preovulatory follicles are believed to be the last to dif- tified as PR targets, and knockout mouse models have ferentiate682 and maintain Ar expression throughout.667 demonstrated that these factors are necessary for ovula- This gradient of AR expression suggests that AR signal- tion in mice. Hypoxia inducible factor 1 (Hif1) is induced ing correlates with the differentiation state of granulosa by FSH in rat granulosa cells cultured in vitro,660,661 and cells and follicle development. have also been found to be regulated by PR.662 Further- In primate and human ovaries, AR expression is also more, the expression of several HIF family members, highest in the granulosa cells of small growing follicles, including Hif1a, Hif2a, and Hif1b, are reduced in PR-null yet evidence of an inverse correlation with follicle dif- ovaries662,663 suggesting an important role in follicle rup- ferentiation is conflicting.625,674,675,678,684 Hillier et al. ture. To further explore this possibility, pharmacological found AR immunoreactivity in the marmoset ovary is inhibitor of HIF activity was shown to block ovulation most abundant in granulosa cells of healthy pre- to small and inhibit follicle rupture,662 demonstrating the impor- antral follicles and low or absent in preovulatory folli- tance of HIF activity in regulating ovulation. Another cles of late follicular stage.674 In contrast, minimal differ- nuclear receptor, PPARG, is also necessary for follicle ences in granulosa cell AR expression between preantral rupture as demonstrated by targeted deletion in granu- and large antral follicles are reported in rhesus monkey losa cells in mice that had no response to exogenous ovaries.675 Furthermore, healthy follicles at late stages of gonadotropins.664 The action of PPARG was found to be maturation in human ovaries are described to possess mediated by several PR target genes, including Il6 and significant AR immunoreactivity.625,678,679 A recent report Edn2,664 demonstrating that several targets are regulated has suggested that AR expression is highest in small downstream of PR and LH signaling to coordinate follicle antral follicles and decreases as the follicle matures,685 rupture and ovulation. suggesting that message levels may not correspond with AR protein levels in human follicles. Androgen Receptor Signaling in Ovarian The regulatory factors for AR expression in the ovary Function are poorly understood. Small preantral follicles in rats maintain high AR levels after hypophysectomy indicat- AR Expression in the Ovary ing that gonadotropins are not required to induce AR AR expression has been documented in the ovaries expression.668 In fact, most evidence indicates that FSH of multiple species, including mouse,373,450 rat,130,665–668 or PMSG-induced differentiation of granulosa cells is pig,669,670 sheep,671 cow,672 monkey,274,395,673–675 and the primary cause of reduced AR expression; Campo human.395,625,676–679 The ovarian pattern of AR expres- et al. demonstrated that PMSG treatment of rats leads to sion is well conserved among species, with levels being the replacement of high-affinity, low-capacity androgen detectable throughout the stages of folliculogenesis binding sites by nonsaturable, low-affinity binding sites except primordial follicles, with the highest expression in the ovary.686 Similarly, immature rats treated with in the granulosa cells of small preantral follicles, detect- recombinant FSH over 48 h exhibit a 65% reduction in able expression in thecal/interstitial cells, and little to no ovarian Ar mRNA levels and a further decrease when expression in luteal cells (Table 25.6). LH is included.668 However, neither FSH nor (BR)-cAMP

4. FEMALE REPRODUCTIVE SYSTEM 1166 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

FIGURE 25.25 Relationship between expression of androgen receptor and aromatase (CYP19) during folliculogenesis. (A) Expres- sion of androgen receptor (Ar) and aromatase (Cyp19) expression in gran- ulosa cells of small (∼200 μm), medium (200–400 μm), and large (>400 μm) follicles from immature rats after treatment with PMSG. Expression was quantified by RNase-protection assay and normalized to 18S rRNA (not shown); data are expressed as percentage of control values (SEM of three separate trials, each consisting of 8–10 animals). a, b: P <0.05; a, c: P <0.01 (by ANOVA with the Newman–Keuls test). (Source: Reproduced with permission from Ref. 491.) (B) Hypothetical model of androgen uti- lization during folliculogenesis. As follicular development progresses, thecal androgen production gradually increases. During the early stages of follicular differentiation, androgens act via androgen receptor (AR) to enhance FSH-induced differentiation, including the stimulation of Cyp19 expression. During the final stages of follicular development, androgens primarily serve as a substrate for CYP19-mediated E2 syn- thesis under stimulation by FSH and LH. This differential regulation of AR and CYP19 may be important in shifting androgen utilization from action to metabolism, thereby ensuring a healthy transition of a follicle from the early maturation to full maturation stage. Source: Reproduced with permission from Ref. 683.

affect Ar mRNA levels in cultured rat granulosa cells increasing evidence that E2 may play a role in decreas- in vitro.665 This suggests that a paracrine interaction ing granulosa cell AR levels during follicle differentia- with theca and/or stromal cells is necessary for reduced tion. Like DHT, E2 exposure of rat granulosa cells also Ar expression observed in granulosa cells. Interestingly, leads to a 20% reduction in Ar transcripts, but this DHT also elicits a 20% reduction in Ar expression in rat effect is unabated by FSH.665 Furthermore, granulosa granulosa cells in vitro that is prevented by co-treatment cells retrieved from untreated hypophysectomized rats with FSH.665 In contrast, T reportedly has little effect respond well to DHT plus FSH and exhibit increased on AR expression in rhesus monkey ovaries.675 There is Cyp19 expression accordingly; but the effect of DHT is

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1167 lost in granulosa cells isolated from E2-primed hypoph- suggesting that folliculogenesis is preserved but ovulation ysectomized rats, suggesting that estrogen pretreatment and luteinization are disrupted. Gonadotropin-induced decreases AR expression.442 Indeed, mature follicles in ovulation of immature AR-null females leads to a ERβ-null ovaries are reported to possess aberrantly high severely reduced ovulatory yield and oocytes that often AR expression relative to similarly staged follicles in exhibit a less condensed cumulus–oocyte complex.183 wild-type ovaries,450 supporting the idea that ER may This latter finding is especially interesting in light of act to decrease Ar expression. While the hormonal regu- the preserved Ar expression in the cumulus oophorus lation of Ar expression is unclear, recent reports suggest cells of preovulatory follicles. Additional evidence that that a member of the orphan nuclear receptor family, AR-null ovaries fail to fully respond to an ovulatory Nur77, is required for maximal Ar expression in murine dose of hCG is an inadequate induction of Pgr, hyaluro- ovaries. Nur77 regulates several steroidogenic enzymes nyl synthetase-2 (Has2), and tumor necrosis factor-α- in the ovary, including those involved in androgen syn- stimulated gene 6 (Tnfaip6),183 all factors that are critical thesis,687 and recent reports demonstrate 20% reduction to expansion of the cumulus–oocyte complex and fol- in Ar expression in the granulosa cells from Nur77- licle rupture. Furthermore, granulosa cells of ovulatory null mice.688 Dia and colleagues were able to show follicles in AR-null ovaries fail to terminally differenti- that NUR77 bound directly to the promoter regulatory ate and luteinize following hCG-induced ovulation, as region of the Ar gene in mouse granulosa cells, provid- indicated by decreased expression of cyclin-dependent ing insight into one of the transcriptional factors neces- kinase inhibitor 1A (Cdkn1a) and cytochrome P450scc sary for maximal AR expression in the ovary.688 (Cyp11).183 These animals were generated by targeted deletion of exon 2 and demonstrate that AR signaling is Ovarian Phenotypes in Mouse Models of Disrupted important for normal ovarian function and to prevent Androgen Signaling POF in rodents. MICE LACKING AR A second global knockout mouse model was devel- Tfm mice were first described in 1970 and are a natu- oped by Shiina et al., where AR was deleted through Cre/ rally existing androgen-resistant mutant283 due to an loxP mediated removal of exon 1.690 This animal model inactivating mutation of the Ar gene.284,285 With the is similar to that described before in that the female mice advent of gene targeting techniques, comparable inac- show an age-dependent decrease in fecundity, however, tivation mutations of the AR gene and resulting phe- these mice also have a decrease in number of offspring notypes have been described in rats and humans.286 per litter in animals aged 8 weeks old (4.5 versus 8.3 Tfm/Tfm female mice are fertile but exhibit a noticeably in wild-type). At 8 weeks of age, these mice also have shortened reproductive life span that becomes appar- decreased number of CL and an increased number of ent after approximately four litters compared to control unhealthy and atretic follicles.690 Knockout of AR in exon (Tfm/X or X/X) females; however, individual litter sizes 1 also causes a POF phenotype where the number of off- were not different.287 Histological evaluation of ovaries spring decreases over time concomitant with a decrease from the Tfm/Tfm breeding females indicated a sparse in the number of healthy follicles, suggesting defects in number of healthy follicles and a hypertrophied inter- folliculogenesis. By 32 weeks of age, 40% of knockout stitium, both illustrative of premature ovarian failure females are infertile, similar to that observed in Tfm/Tfm (POF).287 The difficulties in generating Tfm/Tfm female females and the exon 2–deleted AR-null mice,183,283 and mice limited more extensive studies. at 40 weeks of age all of the mice are infertile.690 No dif- Several global AR-null mouse models have been gen- ference was observed in genes involved in steroidogen- erated via a Cre/loxP targeting scheme that allows for tis- esis within the ovary, and serum hormone levels for E2, sue and temporal specific disruption of the Ar gene and P, and T are within normal rages similar to WT controls, therefore the generation of fertile male carriers of the providing evidence that AR is not necessary for normal targeted Ar allele183,185,288,290 (Table 25.9). In one study, steroidogenesis in the ovary (Table 25.9)690 in contrast global AR-null females made through deletion of exon 1 to reduced P observed in the exon 2–deleted AR-null exhibit normal fecundity only during the first 12 weeks female.689 Microarray analysis was done on ovaries from of continuous breeding, after which they produce a these animals to find potentially altered transcriptional reduced number of litters per female (2.3 versus 3.3 in networks that might contribute to the POF phenotype. wild-type) and less offspring per litter (4.5 versus 9.8 in Several genes implicated in oocyte-granulosa cell com- wild-type), while others become totally infertile.183 The munication were found to be increased,690 including POF observed in AR-null females is remarkably similar kit ligand (Kitl). Kitl is a downstream target of AR as to that described before in Tfm/Tfm females.287 Adult demonstrated by induction of Kitl expression following AR-null females possess ovaries exhibiting a normal DHT treatment, while co-treatment with DHT and the number of growing and antral follicles but notably fewer antiandrogen flutamide attenuated the increased Kitl corpora lutea183 and reduced circulating P (Table 25.9),689 expression.690 This report suggests that AR signaling has

4. FEMALE REPRODUCTIVE SYSTEM 1168 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

TABLE 25.9 Ovarian Phenotypes in AR-Mutant or Null Mouse Models

Mutated or Null for Sex Steroid Signaling Ovarian Phenotypes Hormone Levels References

Tfm/Tfm mice Fertile but have a reduced number of pups/litter as they age Not reported 283 Shorted reproductive life span

AR−/− (exon 2) Reduced fertility and a shortened reproductive life span Reduced P 183,290 Adults exhibit grossly normal ovaries with reduced number of CL Other hormones not reported Abnormal estrous cycle Immature females have reduced response to gonadotropin-induced ovulation

AR−/− (exon 1) Reduced fertility and POF phenotype where animals have a Normal serum E2, P, T, LH, 690 shortened reproductive life span and FSH Increased number of atretic follicles and reduced number of CL Altered ovarian gene expression in immature and adult animals

AR−/− (inframe deletion Delayed production of first litter Normal serum E2, T, LH and 185 of exon 3, loss of DNA Decreased pups/litter and reduced number of CL FSH binding activity) AR+/− animals had age-dependent reduction in pups/litter Increased intraovarian T in Increased number of unhealthy antral follicles 10–12 week old mice Normal response to exogenous gonadotropins, but a reduced number of oocytes ovulated during natural mating

AMHCre+Arf/+;Arf/f Subfertile due to reduced number of litters and age dependent Normal serum LH and FSH 691 (exon 3 floxed, loss of decrease in total number of pups born DNA binding activity in Decreased fertility over time with a cumulative decrease in pups large preantral to antral born per dam follicles) 3-month-old animals have increased large preantral and antral follicle numbers 6-month-old animals have increased length of estrous cycle Reduced cumulus expansion and oocyte/embryo viability

Amhr2Cre+;Arf/f (exon 2 Reduced fertility Not reported 692 floxed) GCARcKO Altered follicle progression and development 8–9 week mice: normal estrous cycle, reduced pups/litter, reduced number of oocytes from natural mating, however, no differences observed when treated with exogenous gonadotropins 24 weeks: increased length of estrous cycle, reduced number of pups born per dam, reduced number of oocytes from natural mating and treatment with endogenous gonadotropins

Gdf9Cre+;Arf/f (exon 2 Normal female phenotype except androgens were unable to Not reported 692 floxed, oocyte specific promote oocyte maturation in vitro loss of AR)

SPARKI (homozygous KI Normal female phenotype Not reported 186 of mutated DBD of AR)

an important role in the oocyte-granulosa cell regula- in human patients with androgen insensitivity syn- tory loop,690 although future work is needed to identify drome (AIS), demonstrating the importance of binding potential mechanisms of action. DNA to AR activities in normal physiological state.693–695 These knockout mouse models demonstrated an This mouse model allows for analysis of the direct effects important role for AR in normal ovarian function and of classical genomic AR signaling in the ovary compared suggest that loss of AR contributes to POF (Table 25.9). to the AR-null models that lack all AR protein. Dele- Both AR-null models lack all AR protein due to a trun- tion of the DNA binding domain in female mice led to cated message and premature stop codon. A third AR-null a subfertile phenotype where the females homozygous mouse model was developed by Walters et al. where an for the deletion had less offspring per litter compared in-frame deletion of exon 3 was made that led to produc- to both heterozygous and WT females.185 Interestingly, tion of an AR protein lacking the second zinc finger and the AR(exon 3)-null females had reduced numbers of is therefore truncated and unable to bind DNA.185 Muta- CL similar to other AR-null mice,290,690 demonstrating tions in this region of the AR gene have been identified the importance for intact AR signaling for this process.

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1169

Furthermore, these mice also had an age-dependent mouse models have been developed to remove AR spe- reduction in fecundity similar to the other AR-null mod- cifically in granulosa cells of the ovary.691,692 These mod- els,183,690 and females heterozygous for the AR(exon3) els provide a unique way to explore the role(s) of AR in deletion also showed an age-dependent reduction in female fertility and ovarian function. offspring born at 6 months of age, suggesting a dosage Sen and Hammes used Amhr2-cre to delete exon 2 of effect.185 An increase was observed in the number of AR in the ovary and found that the animals had reduced unhealthy antral follicles, although AR(exon3)-null mice numbers of offspring born to young dams (2.8 pups in did not have significant changes in overall follicle devel- KO versus 7 in WT mice),692 similar to that observed in opment even in animals aged 10 months185 in contradic- the global knockout animals.183,290 In these granulosa tion to reports in AR-null ovaries with reduced follicle cell AR conditional knockout (GCARcKO) animals, numbers.183 While these models have indicated some altered follicle progression was observed in animals differences in the role of AR in normal folliculogenesis, aged 9 weeks as well as in animals aged 24 weeks (Table all demonstrate an increase in the number of atretic fol- 25.9),692 supporting observations in global knockout ani- licles, demonstrating the importance of AR signaling mals,183,290 where numbers of corpora lutea were reduced in maintaining ovarian health over time. AR-null mice and atretic follicles were increased. These patterns per- have reduced response to gonadotropin-induced ovula- sisted in older animals, which also showed an increased tion,183 and examination of the AR(exon3)-null animals length of estrous cycle at 24 weeks of age compared to found that they respond to exogenous gonadotropin mice aged 9 weeks old that had a normal estrous cycle. stimulation with similar numbers of oocytes ovulated; Several other differences were observed in animals however, a decrease was observed in oocytes ovulated at the two ages examined (9 weeks versus 24 weeks), when examined after natural mating.185 This suggests including differences in ovulation. Mice aged 9 weeks that AR is necessary for normal ovulatory response; had a reduced number of oocytes ovulated during natu- however, excess gonadotropins are able to override this ral mating, but treatment with exogenous gonadotropins defect in the presence of a truncated AR protein (lack- was able to circumvent this defect and cause normal ing the entire DNA binding domain) compared to ova- numbers of oocytes to be ovulated from these animals. ries that are void of all AR protein. Mutation of the DNA As the animals aged, however, the ability of exogenous binding domain that alters the ability of AR to bind to gonadotropins to override this defect was lost, and selective androgen response elements in the SPARKI reduced oocyte numbers were observed from natural mouse model presented by Schauwaers et al. did not mating and superovulation paradigms.692 Failure of the lead to a female reproductive defect, suggesting that the ovary to respond to exogenous gonadotropins concur- ovary does not require selective AR binding to response rent with a reduction in the total number of offspring elements as compared to the male SPARKI mice that born over a long-term fertility study (19 in KO versus have a reproductive phenotype and will not be further 119 in WT mice) supports the POF phenotype observed discussed herein.186 in the global AR-knockout models.183,185,290 The identification of genes necessary for oocyte-gran- The GCARcKO presented by Sen and Hammes was ulosa communication in the ovary as being AR targets690 made using the Amhr2-cre mouse, which has been found provides a possible mechanism that is aberrant in AR- to be active in tissues other than granulosa cells includ- null ovaries, but further studies are necessary to confirm ing the pituitary,696 which may complicate the findings. this difference and explore the pathways involved. Loss While reduced expression of Ar was not reported in the of the DNA binding domain in AR did not affect circulat- hypothalamus or the pituitary,692 the heterogeneous ing hormone levels, as the mice had normal serum levels population of cells present could mask a reduction in of E2, T, LH, and FSH; however, an increase in intraovar- gonadotrope cells that are required for normal female ian T levels was observed in mice aged 10–12 weeks,185 fertility. To eliminate this possibility and explore the suggesting alterations in the steroid environment locally classical role of AR in large preantral to antral follicles, may have paracrine effects on ovarian function that was Walters et al. used the Amh-cre mouse model to delete not measured in other AR-null models.183,690 exon 3 in the ovary.691 Amh is the ligand for Amhr2 and is expressed specifically in large preantral to antral fol- MICE WITH OVARIAN-SPECIFIC DELETION OF AR licles during folliculogenesis demonstrated by crossing The global AR knockout mouse models described the Amh-cre mouse to R26R (ROSA) reporter mice.691 herein have offered many insights into the important role This mouse expresses β-galactosidase activity where of AR in female reproduction, specifically in the ovary. Cre recombinase is expressed, and the X-Gal blue stain However, these models lack AR in all tissues, includ- can be used to localize cells expressing the β-gal indica- ing all parts of the hypothalamic–pituitary–gonadal tor within tissues. The Amh-cre mouse was crossed with axis, which could contribute to the ovarian phenotype exon3 floxed AR mouse that creates a truncated protein observed. To circumvent this, two conditional knockout lacking the DNA binding domain.185 The use of this

4. FEMALE REPRODUCTIVE SYSTEM 1170 25. STEROID RECEPTORS IN THE UTERUS AND OVARY mouse model allowed Ar/AR to be deleted in a tempo- Loss of AR in granulosa cells contributes to altered ral pattern and to examine whether DNA binding activi- ovarian function including a POF phenotype and loss ties are necessary during this stage of folliculogenesis for of ability of the ovary to ovulate oocytes through both normal ovarian function. natural mating or superovulation depending on the age The AMHCre+Ar(exon3)f/f females are subfertile, and the animals were tested.183,185,692 While differences were while the number of offspring born per litter was not observed depending on the AR-null mouse model exam- significantly altered, an age-dependent decrease in the ined and the ovulation paradigm presented, the data total number of pups born was observed185 similar to provide evidence that AR is necessary for normal ovula- the POF phenotype observed in other AR-null models tory function of the ovary. Interestingly, a decrease was (Table 25.9).183,185,283,690,692 This provides evidence that observed in the number of oocytes naturally ovulated AR binding to direct AR target genes in large prean- by one granulosa cell–specific AR-null model,692 while tral follicles is necessary for normal ovarian function no differences were observed in a second model (Table over time. The AMHCre+Ar(exon3)f/f females also have an 25.9).691 These differences may be due to the timing of increased length of estrous cycle at the age of 6 months AR deletion or the different flox animals used (i.e., exon that is not observed in younger mice,691 supporting the 2 versus exon 3). Further study was done on oocytes POF phenotype observed and the importance of AR in ovulated from natural mating in the AMHCre+Ar(exon3)f/f maintaining normal ovarian health. Altered follicle pro- females to see if these oocytes could be fertilized and gression was observed in these mice, and at 3 months of progress through early embryo development in vitro.691 age fewer large preantral and small antral follicles were While the number of oocytes ovulated was not differ- observed; however, this difference was not noted in older ent, the number of fertilized oocytes was reduced (38.6% animals.691 While no difference is observed in follicle pro- in KO versus 94.5% in WT), as was the number of the gression during folliculogenesis in AMHCre+Ar(exon3)f/f embryos able to process to the two-cell stage of embryo females aged 6 months, a significant increase in unhealthy development (35.7% in KO versus 89.5% in WT).691 This follicles was observed,691 supporting the notation that AR supports the necessity of AR expression in granulosa signaling is necessary for normal ovarian health and pre- cells to promote oocyte maturation and development, vention of POF in the ovary. and indicates that DHT-mediated oocyte maturation Communication between granulosa cells and the could be mediated through expression of AR in granu- oocyte are necessary for oocyte health and viability, and losa cells. It also provides insight into the inability of disruption of this communication can lead to infertil- some oocytes to respond to in vitro fertilization, which ity. Animals lacking functional AR protein have altered has implications in human health in cases where some expression of several genes implicated in this commu- women fail to respond to assisted reproductive tech- nication network, including the AR-dependent gene nologies. The variety of AR-null mouse models to date Kitl.690 Pharmacological studies have suggested that provides evidence of the importance of AR signaling in T can induce mouse697,698 and porcine389,699,700 oocyte normal ovarian development, female fertility, and fol- maturation presumably through classical AR signaling licle health. in vitro. To directly examine the role of AR expression in oocytes, a conditional knockout mouse model was Intraovarian Role of Androgen in Ovarian developed using an oocyte-selective Gdf9-cre crossed Function with exon 2–floxed AR.692 These mice had normal fer- The intraovarian roles of androgens can be catego- tility and no overt ovarian phenotype compared to rized into three distinct functions: (1) as substrates for WT controls, demonstrating that AR expression within E2 synthesis, (2) as an enhancer of follicle differentiation, oocytes is not necessary for female fertility, while granu- and (3) as a stage-specific inhibitor of follicle growth.683 losa cell specific expression of the AR nuclear receptor The role of thecal-derived androgens, most notably is necessary in vivo.692 Interestingly, when oocytes were androstenedione, as substrates for E2 synthesis in granu- removed from immature unprimed females and grown losa cells is obviously critical given the profound impor- in vitro, DHT was unable to induce oocyte maturation tance of the latter hormone to reproductive function. in the absence of AR expression in oocytes.692 However, While ovarian steroidogenesis is essential for function P-mediated oocyte maturation was not affected by the of the ovary and female fertility, this topic will not be loss of AR in vitro, demonstrating that these oocytes covered here; instead we focus on the additional intra- could still mature, albeit not through an androgen- ovarian actions of androgens, which include their role as mediated mechanism. This suggests that AR expression activating ligands for AR-mediated effects. in oocytes is necessary for in vitro maturation in the absence of granulosa cells, while in vivo expression of GRANULOSA CELL PROLIFERATION AR in the granulosa cells is sufficient for oocyte matura- The follicular response to androgens is dependent tion, possibly due to P-mediated mechanisms. upon the stage of growth and differentiation. In preantral,

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1171 undifferentiated follicles that are unable to synthesize mice, Wang and Greenwald found that T or DHT, in E2, androgens promote gonadotropin-induced granu- combination with FSH, induces greater levels of DNA losa cell proliferation and maturation. In large, differen- synthesis in antral follicles than FSH alone; a similar tiated follicles that have acquired aromatase activity, E2 response is not observed in small- and medium-sized assumes the role of enhancing FSH-induced granulosa follicles.492 In preantral mouse follicles grown in vitro, cell proliferation and differentiation, and androgens in antiandrogen antisera reduce FSH-stimulated growth excess of that required for conversion to E2 are detri- and DNA synthesis; however, these data must be inter- mental to follicle health, leading to atresia. This paradox preted with caution since antiandrogen antisera may of androgen action during follicle growth and differen- also remove androgens from the pool of substrates for tiation explains the contradictory results when compar- E2 synthesis. Additional evidence that androgens may ing experimental studies. promote granulosa cell proliferation in mice comes from Several studies in hypophysectomized rats have the findings of Burns et al. that the AR-antagonist flu- shown that coadministration of T inhibits estrogen or tamide reduces the growth rate of granulosa cell tumors gonadotropin-induced granulosa cell proliferation and in animals lacking functional inhibins.499 In vitro follicle increased ovarian weight, and promotes degeneration culture studies demonstrated the importance of AR sig- in more mature follicles.422,489,701–703 In contrast, Arm- naling in folliculogenesis, where treating follicles with strong and Papkoff found aromatizable androgens antiandrogenic compounds reduced follicle growth dur- (e.g., T or androstenedione) enhance the promotional ing the preantral stage of folliculogenesis,709 demonstrat- effects of FSH in the ovaries of hypophysectomized rats ing the importance of androgens in early follicle growth. but that the nonaromatizable androgen DHT is inhibi- In cumulus cells of large antral follicles from porcine tory.291 Similar treatment with hCG instead of T has a ovaries, DHT augments FSH or IGF1-induced prolifera- comparable effect on gonadotropin or E2-induced folli- tion, and this effect is inhibited by flutamide.710 Rhesus cle growth in hypophysectomized rats, but is prevented monkey ovaries exhibit a strong correlation (r = 0.91) by AR-antagonists, indicating the detrimental effects of between AR expression and cell proliferation675 and hCG are due to stimulation of thecal cell androgen syn- exhibit an increased number of follicles of all stages thesis.702,704,705 However, Bley et al. found that granulosa except large antral following T or DHT treatments.711 cells from preantral follicles of estrogen-primed rats pro- Yang and Fortune reported that T was able to stimu- liferate at comparable rates in response to FSH plus E2 late early follicle maturation in bovine follicles.712 The or DHT in vitro496 and that the promotional effects of the collective data support the notion that androgens are latter steroid are blocked by the antiandrogen hydroxy- important in granulosa cell proliferation during early flutamide.496 Investigations toward the mechanism by stages, but at latter stages the effect appears to be species which androgens inhibit granulosa cell proliferation dependent. indicate that DHT exposure to estrogen-primed rats prior to granulosa cell isolation causes a blunted response to GRANULOSA CELL DIFFERENTIATION forskolin-induced cyclin-D2 (Ccnd2) expression in vitro, Fully differentiated preovulatory follicles in mamma- leading to cell-cycle arrest and reduced proliferation.706 lian ovaries are distinctly characterized by a large fluid- DHT is able to exert its inhibitory effect on Ccnd2 expres- filled antrum, significantly increased aromatase (CYP19) sion through AMP-activated protein kinase (AMPK) activity, and acquisition of LH responsiveness. The activation, which in turn inhibits phosphorylation of absolute requirement of FSH signaling in this process MAPK and ultimately impacts granulosa cell prolifera- is illustrated by the absence of all three phenotypes in tion downstream of FSH.707,708 AMPK has a role in inhib- mice null for FSH action.420,512,513 E2 is also required for iting cell proliferation, and pharmacological activation full FSH-induced follicle differentiation. However, there of the kinase can lead to cell cycle arrest, while blocking is substantial evidence that androgens may be equally AMPK activity prior to treatment with DHT was able efficient as E2 in augmenting certain responses to FSH, to overcome the inhibitory effect normally observed,708 most especially the induction of aromatase activity and providing insights into the mechanism of action of DHT antrum formation.683 Both T and DHT enhance FSH inhibition on cell cycle progression in estrogen-primed induction of aromatase activity in hypophysectomized rat granulosa cells. This work also highlights how andro- rat ovaries or isolated granulosa cells in vitro in a dose- gens are able to be stimulatory during early stages of fol- dependent fashion.291,713 Furthermore, pharmacological liculogenesis and then inhibitory during the latter stages inhibition of ovarian androgen synthesis in rats simulta- when estrogen signaling is the main driver of follicle neously reduces aromatase activity that can be restored development. by exogenous T.714 The stage-specific effects of androgens during follicu- Fitzpatrick et al. found that T, and DHT to a lesser logenesis described herein may be more apparent in rat extent, significantly enhances FSH induction of Cyp19 ovaries relative to other species. In hypophysectomized expression and aromatase activity in granulosa cells from

4. FEMALE REPRODUCTIVE SYSTEM 1172 25. STEROID RECEPTORS IN THE UTERUS AND OVARY untreated hypophysectomized rats442 (Figure 25.19). suggests that androgens may act at a site prior to adeny- This finding was supported by work done in in vitro lyl cyclase. For example, neither T nor DHT synergize cultured granulosa cells isolated from rats where T was with (Bu)2-cAMP to induce aromatase activity in rat able to increase Cyp19 and P450scc expression. This granulosa cells719 whereas E2 does.442,720 Instead, AR- regulation was found to be dependent on liver receptor mediated androgen actions may largely act by increasing homolog-1 (Lrh1) expression that is induced by T and the level of FSH receptor in granulosa cells of preantral not DHT through AR binding to Lrh1 promoter regu- follicles. T can also restore the losses in FSH receptor and latory region715 and provides evidence that T, and not responsiveness that occur in rat granulosa cells cultured DHT, affects the expression of genes necessary for estro- without gonadotropins, and this effect is inhibited by AR gen production during folliculogenes. Furthermore, Lrh1 antagonists.719 Evidence that androgens increase granu- expression is also induced by FSH in granulosa cells,715 losa FSH-receptor levels comes from reports that ovaries providing a potential mechanism of action for synergy of 10-day-old and prepubertal AR-null females exhibit necessary for differentiation of granulosa cells in the significantly reduced Fshr mRNA levels, even 48 h after ovary. PMSG treatment in the latter age group.183 Similarly, Interestingly, this effect is lost in granulosa cells iso- Weil et al. found that 3 or 10 days of T treatment in adult lated from estrogen-primed hypophysectomized rats,442 rhesus monkeys increases FSH-receptor expression in supporting the hypothesis that early growing follicles follicles of all stages except primary.684 In human small are more sensitive to androgen action and that estrogens antral follicles cultured in vitro, a positive correlation was downregulate AR expression as part of the process of observed in AR expression, androgen concentrations in follicle differentiation. Similar data has been produced follicular fluid, and FSHR expression721 in small follicles in mice. For example, AR antagonists are able to inhibit demonstrating an association with AR and androgens in FSH-induced E2 synthesis in individually cultured immature granulosa cells, and suggests that in human murine follicles if included at the beginning of the culture ovaries this association is important for normal follicle period prior to any indications of differentiation.716 Simi- development. larly, E2 output by large antral murine follicles in vitro is AR-mediated androgen actions may also be involved unaffected by DHT.492 In marmoset ovaries, granulosa in antrum formation. Murray et al. found DHT poten- cells isolated from small (0.5–1 mm) follicles are highly tiates suboptimal doses of FSH to stimulate increased responsive to the enhancing effects of T or DHT on FSH- follicle diameter of murine follicles in culture, and this induced aromatase activity, but this response is lost in promotional effect is inhibited by AR antagonist.716 T granulosa cells from larger (>2 mm) follicles whereupon or DHT are also reported to be almost as effective as DHT exposure becomes inhibitory.717,718 In contrast, 3 or E2 in augmenting FSH-induced increases in the num- 10 days of T treatment in rhesus monkeys has no effect ber of antral follicles in the ovaries of hypophysecto- on ovarian CYP19 expression.684 mized mice.492 However, the ovaries of global AR-null The majority of data indicate that healthy preantral mice do not show differences in the number of antral follicles are responsive to T and require the steroid for follicles,183,185,690 while in the granulosa-specific AR- the initiation of Cyp19 expression by FSH in granulosa null mice a reduction in the number of antral follicles cells. Since androgen synthesis in thecal cells precedes was observed at all ages,692 or only in younger mice the acquisition of aromatase activity in the accompany- (3 months).691 The conflicting reports of antral forma- ing granulosa cells, this mechanism of androgen/FSH tion in AR-null mice demonstrate the complexity of synergism allows for more efficient induction of Cyp19 this developmental process in the ovary. Therefore, expression than would occur with FSH alone. Once the FSH-dependent antrum formation may be enhanced by granulosa cells acquire sufficient aromatase activity, either receptor-mediated androgen or estrogen actions, thecal-derived androgens become more important as congruent with the process being part of the transition substrates and ER-mediated E2 actions continue the role period during which follicles become less sensitive to of synergizing with FSH to promote follicle differentia- androgens and more dependent on estrogens. tion (Figure 25.25). Hence, the hallmark of a healthy pre- Not all FSH-dependent processes during follicle dif- ovulatory follicle in rodent ovaries may be the presence ferentiation are enhanced by androgens. LH-receptor of substantial aromatase activity and E2 output, and expression and LH responsiveness by granulosa cells decreased AR expression and androgen sensitivity. occurs only in preovulatory follicles during the final The mechanism by which androgens augment FSH stages of folliculogenesis, just prior to the LH surge. action on granulosa cells is unclear but may differ from Unlike the above processes of follicle differentiation, that hypothesized for E2. Whereas ER-mediated E2 FSH stimulation of LH-receptor expression in pre- actions are believed to enhance the amount and effec- ovulatory granulosa cells is facilitated by estrogen or tiveness of FSH-stimulated intracellular cAMP with- aromatizable androgens only515,548,549 while DHT has out obvious changes in FSH-receptor levels, evidence no effect or may even be inhibitory.515,549 These data

4. FEMALE REPRODUCTIVE SYSTEM SEX STEROID RECEPTORS AND OVARIAN FUNCTION 1173 are consistent with the mutually exclusive pattern of OVULATION AR and LH receptor in preovulatory follicles of rodent Over the past decade there have been few studies ovaries, as AR expression is limited to the cumulus aimed at determining a role for androgen signaling in cells667 while LH receptor is predominantly localized follicle rupture and ovulation, the majority of these from to mural and antral granulosa cells.722–724 In addition, the AR-null mouse models (Table 25.9). This is espe- in vivo studies demonstrating that AR antagonists cially surprising since Mori et al. demonstrated in 1977 have no effect on LH-receptor expression in the ovaries that inhibition of androgen action at the time of ovula- of FSH-stimulated, DES-primed hypophysectomized tion is detrimental to follicle rupture in the rat.728 These rats suggest that AR-mediated actions are not required studies employed anti-T and anti-P antisera during to promote or maintain LH-receptor expression in induced ovulation in immature or hypophysectomized preovulatory follicles.725 The specific requirement of rats to demonstrate that: (1) acute treatment with anti-T ER-mediated E2 actions to augment FSH-stimulated antisera at the time or shortly after hCG-induced ovula- LH-receptor expression in preovulatory granulosa tion leads to a dose-dependent reduction in the number cells provides a mechanism by which only those fol- of ovulated oocytes, and (2) inhibition of ovulation by licles that acquire sufficient aromatase activity and are acute treatment with an anti-P antisera at the time of suitable to ovulate acquire the capacity to respond to hCG treatment is rescued by concurrent treatment with the LH surge. T or DHT but not E2.728 In a similar study, Peluso et al. demonstrated that treatment of immature rats with the THECAL CELL STEROIDOGENESIS AR antagonists, cyproterone acetate or flutamide, 3 h There are few studies on the role of androgen sig- prior to hCG-induced ovulation drastically reduces the naling in thecal cell function and steroidogenesis number of ovulated oocytes and prevents expansion of despite these cells being the primary source of andro- the cumulus–oocyte complex.729 Similar findings are gen synthesis. Several nonsteroidal paracrine factors reported in mice.730 The development of AR-null mouse are known to modulate thecal cell steroidogenesis and models has provided some insight into the role of have been more thoroughly studied.476,566 The limited androgen signaling in ovulation. An ovulatory defect in available evidence suggests that AR-mediated andro- AR-null mice is suggested by the presence of a normal gen actions may negatively regulate thecal cell ste- number of preovulatory follicles but few corpora lutea roidogenesis, thereby forming an autocrine regulatory in the ovaries,183 as well as reduced circulating P lev- feedback loop. Mahesh and colleagues have shown in els.689 Furthermore, gonadotropin-induced ovulation of enriched thecal/interstitial cell cultures from rats that immature AR-null females indicates a severely reduced AR activation by nonaromatizable agonists attenuates response in terms of recoverable oocytes in the ovi- hCG or hCG/IFG-1–stimulated increases in andro- ducts, and may be attributable to their failure to exhibit stenedione synthesis by 32% and 40%, respectively, hCG-induced expression of P receptor, hyaluronyl via selective inhibition of Cyp17 expression.726 Simi- synthetase-2, and tumor necrosis factor-α-stimulated lar experiments demonstrated that AR-antagonists gene 6.183 Reduced ovulatory rates were also observed enhance hCG-induced androstenedione synthesis, fur- in another AR-null mouse model lacking the DNA bind- ther indicating a receptor-mediated effect of androgens ing domain after normal mating,185 however ovulation on thecal cell steroidogenesis.726 The majority of recent rates after treatment with exogenous gonadotropins studies published to date focus on the transcriptional were normal. A similar finding was observed in granu- regulation of genes involved in androgen biosynthesis losa cell–specific AR-null mouse model where oocytes in theca cells,687,688 but not on the autocrine action of collected after natural mating are reduced but animals the synthesized androgen(s) within the cells. In fact, respond to exogenous gonadotropins. Interestingly, a recent review entitled “Theca: The forgotten cell of when these animals were 6 months of age, they didn’t the ovarian follicle” described the various factors that ovulate at all.692 These mouse models suggest that AR are known to regulate steroidogenesis and androgen plays a role in normal ovulation, but stimulation with production,727 which also highlights the need for more excess gonadotropins can overcome these deficiencies focused experiments looking at theca cell functions of only in young mice. Further work to study the role of AR. The use of conditional knockout mouse models to androgen signaling in other mammalian species includ- explore the role of AR specifically in theca cells could ing primates and humans could provide insight into also shine light on this important question. Therefore, possible fertility treatments. ER-mediated E2 actions (see the section Estrogen Receptor Signaling in Ovarian Function) and AR- FOLLICLE ATRESIA mediated androgen actions may both contribute to Follicle atresia in the mammalian ovary has been maintaining homeostasis of androgen synthesis in the thoroughly described in several reviews.565,680,731 Fol- thecal cells of developing follicles. licular atresia is a complex hormonally driven process

4. FEMALE REPRODUCTIVE SYSTEM 1174 25. STEROID RECEPTORS IN THE UTERUS AND OVARY involving both putative “survival” factors, e.g., insu- Glucocorticoid signaling in the ovary acts to inhibit lin-like growth factor-1, epidermal growth factor, basic LH action and steroid biosynthesis during folliculogene- fibroblast growth factor, E2, and activin; and putative sis.736 Treatment of granulosa cells with cortisol or dexa- atretogenic factors, e.g., tumor necrosis factor-α, GnRH, methasone was able to inhibit FSH-stimulated increases and androgens.565,680,731 The respective action of each of in aromatase activity in vitro, while P production was these factors on atresia is dependent on the follicular increased compared to control-treated cells. This finding stage,731 for example, epidermal growth factor and basic suggests that GR signaling may affect steroidogenesis fibroblast growth factor begin to act early on primor- of the ovarian steroids in a differential manner.737 Fur- dial stage follicles, whereas insulin-like growth factor-1 thermore, this finding suggests that GR signaling in the affects early antral follicles. ovary is important for P function and ovulation. Dexa- Circumstantial evidence of the actions of androgen methasone was also found to stimulate FSH-mediated includes the numerous observations that atretic follicles increases in P production in porcine granulosa cells,738 consistently exhibit an increased androgen:estrogen suggesting a key role for GR signaling during late stages ratio within the intrafollicular fluid.565 Supporting of ovulation and for corpora lutea function. experimental evidence includes reports that low doses The expression of GR in granulosa cells does not of hCG (to induce endogenous androgen production), T, change significantly during early follicle development; or DHT following 4 days of E2 treatment in hypophy- however, the treatment of cells with agonist shows dif- sectomized rats lead to increased numbers of atretic fol- ferential activity depending on the stage of differentiation licles within 24 h.705,732 The detrimental effects of hCG of the cells. To circumvent the action of glucocorticoids, or exogenous androgens are inhibited by co-treatment expression of two metabolizing enzymes has been found with anti-T antisera or AR antagonists, indicating the to have stage-specific expression in the ovary.736 The involvement of androgen receptor in this process.705,732 expression of the type I and type II 11β-hydroxysteroid Interestingly, large (>150 μm) follicles are more sus- dehydrogenase (11HSD 1 and 11HSD 2) modulates the ceptible to the atretogenic effects of androgens than action of glucocorticoids by regulating their concentra- small (<150 μm) follicles, indicating that androgens are tions within tissues.739 This is important in the ovary largely atretogenic in late-stage follicles.705,732 In simi- during folliculogenesis since glucocorticoids are not syn- lar experiments on hypophysectomized rats, Billig et al. thesized within the tissue itself. During FSH-stimulated demonstrated that the follicular atresia that follows the follicle growth, 11Hsd1 is expressed and acts to suppress withdrawal of 2 days of continuous estrogen (DES) treat- glucocorticoid action, then as the granulosa cells become ment is enhanced by subsequent T exposure but pre- luteinized after LH stimulation, 11Hsd2 expression is vented by E2.733 Conversely, Flaws et al. show follicular increased.735,740 This acts to increase the action of gluco- atresia induced by the endocrine-disrupting chemical corticoid action on the ovary during luteinization and for- methoxychlor.734 mation of the corpora lutea, which also shows an increase Additionally, AR-null mouse models also have in GR expression, demonstrating the importance of GR increased atretic follicles,183,185,690 demonstrating that signaling during this process. Interestingly, expression AR signaling plays a role in follicle atresia; how- of 11Hsd2 is reduced during luteal regression,741 demon- ever, further studies are needed to understand the strating the importance of GR activity specifically in the mechanism. corpora lutea. The evidence to date suggests that GR sig- naling is important in ovarian function, although further Glucocorticoid Receptor Signaling in Ovarian work is needed to define the precise activity and neces- Function sity of this signaling mechanism specifically in the ovary. While the ovary is not considered a classic target tis- sue of glucocorticoids, altered glucocorticoid production SUMMARY can affect reproduction and fertility throughout the HPG axis. Within the ovary, glucocorticoid signaling through The salient aspects of the previous discussion are sum- GR can modulate steroidogenesis/gametogenesis and marized in a model of sex steroid and steroid receptor will briefly be discussed here. GR is expressed in the action during folliculogenesis. Although certain postu- rat ovary, and examination of whole ovary expression lated mechanisms and pathways portrayed in this model shows that levels do not change upon treatment with are supported by substantial evidence, other aspects are exogenous gonadotropins.735 Further examination of GR much more speculative and remain to be demonstrated. expression in isolated granulosa cells shows no signifi- Regardless, the evidence is solid that the receptor-medi- cant alterations in expression within these cells, although ated steroid pathways exist in the ovary and their intra- GR expression appears to be elevated in corpora lutea.735 ovarian functions are critical to female fertility.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1175

CONCLUSION 11. Walter P, Green S, Greene G, et al. Cloning of the human estrogen receptor cDNA. Proc Natl Acad Sci USA 1985;82:7889–93. 12. Laudet V, Gronemeyer H. The nuclear receptor factsbook. London: Within this fairly extensive review we have endeav- Academic Press; 2001. ored to describe as comprehensively as possible what 13. Katsu Y, Taniguchi E, Urushitani H, et al. Gen Comp Endocrinol is currently known regarding the important functions 2010;168:220–30. of estrogen, progesterone, androgen, and glucocorti- 14. Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA. coid receptors in ovarian and uterine tissues. Clearly, Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA 1996;93:5925–30. an abundance of data has been obtained that has greatly 15. Mosselman S, Polman J, Dijkema R. Er beta: Identification and advanced our understanding of the roles of these recep- characterization of a novel human estrogen receptor. FEBS Lett tors since early descriptions more than five decades ago. 1996;392:49–53. Nevertheless, many open questions remain for further 16. White R, Lees JA, Needham M, Ham J, Parker M. Structural elaboration. At the level of the steroid receptors them- organization and expression of the mouse estrogen receptor. Mol Endocrinol 1987;1:735–44. selves, questions remain regarding how the receptors 17. Le Romancer M, Poulard C, Cohen P, Sentis S, Renoir JM, Corbo mediate the spectrum of responses, and what roles tissue L. Cracking the estrogen receptor’s posttranslational code in and gene specificity of distinct activation domains on the breast tumors. Endocr Rev 2011;32:597–622. receptors play. Additionally, better understanding of the 18. Heldring N, Pike A, Andersson S, et al. Estrogen receptors: interactions of receptors with chromatin, which are often how do they signal and what are their targets. Physiol Rev 2007;87:905–31. quite far from regulated transcripts, and how the inter- 19. Kos M, Denger S, Reid G, Gannon F. Upstream open read- actions lead to changes in transcription, are challeng- ing frames regulate the translation of the multiple mRNA ing questions to approach. We can hope to learn more variants of the estrogen receptor alpha. J Biol Chem 2002;277: about temporal patterns of changes in the organization 37131–8. of nuclear structures initiated by steroid receptors. At 20. Kos M, Reid G, Denger S, Gannon F. Minireview: genomic orga- nization of the human ER alpha gene promoter region. Mol Endo- the whole animal level, what remains is to piece together crinol 2001;15:2057–63. advances gleaned from diverse models and methods 21. Wilson ME, Westberry JM. Regulation of oestrogen receptor gene into a comprehensive description of the coordinated sig- expression: new insights and novel mechanisms. J Neuroendocrinol naling between receptors for different steroids in cells 2009;21:238–42. and tissues that culminate in an optimal environment 22. Cicatiello L, Cobellis G, Addeo R, et al. In vivo functional analy- sis of the mouse estrogen receptor gene promoter: a transgenic for fertilization, implantation, gestation, and parturition. mouse model to study tissue-specific and developmental regu- lation of estrogen receptor gene transcription. Mol Endocrinol 1995;9:1077–90. References 23. Pinzone JJ, Stevenson H, Strobl JS, Berg PE. Molecular and cel- lular determinants of estrogen receptor alpha expression. Mol Cell 1. Aagaard MM, Siersbaek R, Mandrup S. Molecular basis for gene- Biol 2004;24:4605–12. specific transactivation by nuclear receptors. Biochim Biophys Acta 24. Donaghue C, Westley BR, May FEB. Selective promoter usage of 2011;1812:824–35. the human estrogen receptor-alpha gene and its regulation by 2. Mangelsdorf DJ, Thummel C, Beato M, et al. The nuclear receptor estrogen. Mol Endocrinol 1999;13:1934–50. superfamily: the second decade. Cell 1995;83:835–9. 25. Stoica A, Saceda M, Fakhro A, Solomon HB, Fenster BD, Mar- 3. Bain DL, Heneghan AF, Connaghan-Jones KD, Miura MT. tin MB. Regulation of estrogen receptor-alpha gene expres- Nuclear receptor structure: implications for function. Annu Rev sion by 1,25-dihydroxyvitamin D in MCF-7 cells. J Cell Biochem Physiol 2007;69:201–20. 1999;75:640–51. 4. Committee NRN. A unified nomenclature system for the nuclear 26. Ishibashi O, Kawashima H. Cloning and characterization of the receptor superfamily. Cell 1999;97:161–3. functional promoter of mouse estrogen receptor beta gene. Bio- 5. Ponglikitmongkol M, Green S, Chambon P. Genomic organiza- chim Biophys Acta 2001;1519:223–9. tion of the human oestrogen receptor gene. EMBO J 1988;7: 27. Hirata S, Shoda T, Kato J, Hoshi K. Isoform/variant mRNAs for 3385–8. sex steroid hormone receptors in humans. Trends Endocrinol Metab 6. Misrahi M, Venencie PY, Saugier-Veber P, Sar S, Dessen P, Milgrom 2003;14:124–9. E. Structure of the human progesterone receptor gene. Biochim 28. Lewandowski S, Kalita K, Kaczmarek L. Estrogen receptor beta. Biophys Acta 1993;1216:289–92. Potential functional significance of a variety of mRNA isoforms. 7. Kuiper GG, Faber PW, van Rooij HC, et al. Structural organi- FEBS Lett 2002;524:1–5. zation of the human androgen receptor gene. J Mol Endocrinol 29. Li XT, O’Malley BW. Unfolding the action of progesterone recep- 1989;2:R1–4. tors. J Biol Chem 2003;278:39261–4. 8. Gibson DA, Saunders PT. Estrogen dependent signaling in repro- 30. Scarpin KM, Graham JD, Mote PA, Clarke CL. Progesterone ductive tissues - a role for estrogen receptors and estrogen related action in human tissues: regulation by progesterone receptor (PR) receptors. Mol Cell Endocrinol 2012;348:361–72. isoform expression, nuclear positioning and coregulator expres- 9. Green S, Walter P, Greene G, et al. Cloning of the human oestro- sion. Nucl Recept Signal 2009;7:e009. gen receptor cDNA. J Steroid Biochem 1986;24:77–83. 31. Kastner P, Krust A, Turcotte B, et al. Two distinct estrogen-regu- 10. Green S, Walter P, Kumar V, et al. Human oestrogen receptor lated promoters generate transcripts encoding the two function- cDNA: sequence, expression and homology to v-erb-A. Nature ally different human progesterone receptor forms A and B. EMBO 1986;320:134–9. J 1990;9:1603–14.

4. FEMALE REPRODUCTIVE SYSTEM 1176 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

32. Kraus WL, Katzenellenbogen BS. Regulation of progesterone 53. Conaway RC, Conaway JW. Function and regulation of the medi- receptor gene expression and growth in the rat uterus: modula- ator complex. Curr Opin Genet Dev 2011;21:225–30. tion of estrogen actions by progesterone and sex steroid hormone 54. Malik S, Roeder RG. The metazoan mediator co-activator com- antagonists. Endocrinology 1993;132:2371–9. plex as an integrative hub for transcriptional regulation. Nat Rev 33. Kraus WL, Montano MM, Katzenellenbogen BS. Cloning of the Genet 2010;11:761–72. rat progesterone receptor gene 5′-region and identification of two 55. Roberts CW, Orkin SH. The SWI/SNF complex – chromatin and functionally distinct promoters. Mol Endocrinol 1993;7:1603–16. cancer. Nat Rev Cancer 2004;4:133–42. 34. Wetendorf M, DeMayo FJ. The progesterone receptor regulates 56. Wu SC, Zhang Y. Minireview: role of protein methylation and implantation, decidualization, and glandular development demethylation in nuclear hormone signaling. Mol Endocrinol via a complex paracrine signaling network. Mol Cell Endocrinol 2009;23:1323–34. 2012;357:108–18. 57. Barnes PJ, Adcock IM, Ito K. Histone acetylation and deacety- 35. Richer JK, Lange CA, Wierman AM, et al. Progesterone receptor lation: importance in inflammatory lung diseases. Eur Respir J 2005; variants found in breast cells repress transcription by wild-type 25:552–63. receptors. Breast Cancer Res Treat 1998;48:231–41. 58. Keppler BR, Archer TK, Kinyamu HK. Emerging roles of the 26S 36. Lubahn DB, Brown TR, Simental JA, et al. Sequence of the intron/ proteasome in nuclear hormone receptor-regulated transcription. exon junctions of the coding region of the human androgen Biochim Biophys Acta 2011;1809:109–18. receptor gene and identification of a point mutation in a family 59. Kim HM, Yu Y, Cheng Y. Structure characterization of the 26S with complete androgen insensitivity. Proc Natl Acad Sci USA proteasome. Biochim Biophys Acta 2011;1809:67–79. 1989;86:9534–8. 60. Giangrande PH, McDonnell DP. The A and B isoforms of the 37. McEwan IJ. Molecular mechanisms of androgen receptor- human progesterone receptor: two functionally different tran- mediated gene regulation: structure-function analysis of the AF-1 scription factors encoded by a single gene. Recent Prog Horm Res domain. Endocr Relat Cancer 2004;11:281–93. 1999;54:291–314. 38. Bennett NC, Gardiner RA, Hooper JD, Johnson DW, Gobe GC. 61. Katzenellenbogen BS, Montano MM, Ediger TR, et al. Estrogen Molecular cell biology of androgen receptor signalling. Int J receptors: selective ligands, partners, and distinctive pharmacol- Biochem Cell Biol 2010;42:813–27. ogy. Recent Prog Horm Res 2000;55:163–93. discussion 194–5. 39. Gelmann EP. Molecular biology of the androgen receptor. J Clin 62. Kumar R, Zakharov MN, Khan SH, et al. The dynamic structure Oncol 2002;20:3001–15. of the estrogen receptor. J Amino Acids 2011;2011:812540. 40. Rhen T, Grissom S, Afshari C, Cidlowski JA. Dexamethasone 63. Watanabe T, Inoue S, Ogawa S, et al. Agonistic effect of tamoxifen blocks the rapid biological effects of 17beta-estradiol in the rat is dependent on cell type, ERE-promoter context, and estrogen uterus without antagonizing its global genomic actions. FASEB J receptor subtype: functional difference between estrogen recep- 2003;17:1849–70. tors alpha and beta. Biochem Biophys Res Commun 1997;236:140–5. 41. Whirledge S, Dixon D, Cidlowski JA. Glucocorticoids regulate 64. Tsai MJ, O’Malley BW. Molecular mechanisms of action of steroid/ gene expression and repress cellular proliferation in human uter- thyroid receptor superfamily members. Annu Rev Biochem 1994; ine leiomyoma cells. Horm Cancer 2012;3:79–92. 63:451–86. 42. Lu NZ, Cidlowski JA. Glucocorticoid receptor isoforms generate 65. Glass CK. Differential recognition of target genes by nuclear transcription specificity. Trends Cell Biol 2006;16:301–7. receptor monomers, dimers, and heterodimers. Endocr Rev 1994; 43. Aranda A, Pascual A. Nuclear hormone receptors and gene 15:391–407. expression. Physiol Rev 2001;81:1269–304. 66. Hiipakka RA, Liao S. Molecular mechanism of androgen action. 44. Brelivet Y, Rochel N, Moras D. Structural analysis of nuclear Trends Endocrinol Metab 1998;9:317–24. receptors: from isolated domains to integral proteins. Mol Cell 67. Kuiper GG, Carlsson B, Grandien K, et al. Comparison of Endocrinol 2012;348:466–73. the ligand binding specificity and transcript tissue distri- 45. Hilser VJ, Thompson EB. Structural dynamics, intrinsic disorder, bution of estrogen receptors alpha and beta. Endocrinology and allostery in nuclear receptors as transcription factors. J Biol 1997;138:863–70. Chem 2011;286:39675–82. 68. Koide A, Zhao C, Naganuma M, et al. Identification of regions within 46. Helsen C, Kerkhofs S, Clinckemalie L, et al. Structural basis for the F domain of the human estrogen receptor alpha that are impor- nuclear hormone receptor DNA binding. Mol Cell Endocrinol tant for modulating transactivation and protein-protein interactions. 2012;348:411–7. Mol Endocrinol 2007;21:829–42. 47. McEwan IJ. Nuclear hormone receptors: allosteric switches. Mol 69. Yang J, Singleton DW, Shaughnessy EA, Khan SA. The F-domain Cell Endocrinol 2012;348:345–7. of estrogen receptor-alpha inhibits ligand induced receptor 48. Kumar R, McEwan IJ. Allosteric modulators of steroid hormone dimerization. Mol Cell Endocrinol 2008;295:94–100. receptors: structural dynamics and gene regulation. Endocr Rev 70. Montano MM, Muller V, Trobaugh A, Katzenellenbogen BS. 2012;33:271–99. The carboxy-terminal F domain of the human estrogen receptor: 49. Zassadowski F, Rochette-Egly C, Chomienne C, Cassinat B. Regu- role in the transcriptional activity of the receptor and the effec- lation of the transcriptional activity of nuclear receptors by the tiveness of antiestrogens as estrogen antagonists. Mol Endocrinol MEK/ERK1/2 pathway. Cell Signal 2012;24:2369–77. 1995;9:814–25. 50. Hill KK, Roemer SC, Churchill MEA, Edwards DP. Structural 71. George CL, Lightman SL, Biddie SC. Transcription factor interactions and functional analysis of domains of the progesterone receptor. in genomic nuclear receptor function. Epigenomics 2011;3:471–85. Mol Cell Endocrinol 2012;348:418–29. 72. Lonard DM, O’Malley BW. Expanding functional diversity of the 51. Hsia EY, Goodson ML, Zou JX, Privalsky ML, Chen HW. Nuclear coactivators. Trends Biochem Sci 2005;30:126–32. receptor coregulators as a new paradigm for therapeutic target- 73. Bulynko YA, O’Malley BW. Nuclear receptor coactivators: struc- ing. Adv Drug Delivery Rev 2010;62:1227–37. tural and functional biochemistry. Biochemistry (Mosc) 2011; 52. Johnson AB, O’Malley BW. Steroid receptor coactivators 1, 2, and 50:313–28. 3: critical regulators of nuclear receptor activity and steroid recep- 74. O’Malley BW, Malovannaya A, Qin J. Minireview: nuclear recep- tor modulator (SRM)-based cancer therapy. Mol Cell Endocrinol tor and coregulator proteomics–2012 and beyond. Mol Endocrinol 2012;348:430–9. 2012;26:1646–50.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1177

75. Biddie SC, John S, Hager GL. Genome-wide mechanisms of 98. Migliaccio A, Castoria G, Di Domenico M, et al. Steroid- nuclear receptor action. Trends Endocrinol Metab 2010;21:3–9. induced androgen receptor-oestradiol receptor beta-Src com- 76. Zaret KS, Carroll JS. Pioneer transcription factors: establishing plex triggers prostate cancer cell proliferation. EMBO J 2000;19: competence for gene expression. Genes Dev 2011;25:2227–41. 5406–17. 77. Fu X, Huang C, Schiff R. More on fox news: foxa1 on the hori- 99. Kousteni S, Bellido T, Plotkin LI, et al. Nongenotropic, sex- zon of estrogen receptor function and endocrine response. Breast nonspecific signaling through the estrogen or androgen Cancer Res 2011;13:307. receptors: dissociation from transcriptional activity. Cell 78. Carroll JS, Liu XS, Brodsky AS, et al. Chromosome-wide mapping 2001;104:719–30. of estrogen receptor binding reveals long-range regulation requir- 100. Boonyaratanakornkit V, Scott MP, Ribon V, et al. Progesterone ing the forkhead protein foxa1. Cell 2005;122:33–43. receptor contains a proline-rich motif that directly interacts with 79. Carroll JS, Brown M. Estrogen receptor target gene: an evolving SH3 domains and activates c-Src family tyrosine kinases. Mol Cell concept. Mol Endocrinol 2006;20:1707–14. 2001;8:269–80. 80. Ogle TF. Progesterone-action in the decidual mesometrium of 101. Boonyaratanakornkit V, Edwards DP. Receptor mechanisms pregnancy. Steroids 2002;67:1–14. of rapid extranuclear signalling initiated by steroid hormones. 81. Samalecos A, Gellersen B. Systematic expression analysis and In: Essays in biochemistry: nuclear receptor superfamily. London: antibody screening do not support the existence of naturally Portland Press on behalf of the Biochemical Society; 2004. occurring progesterone receptor (PR)-C, PR-M, or other truncated pp. 105–20. PR isoforms. Endocrinology 2008;149:5872–87. 102. Heinlein CA, Chang C. The roles of androgen receptors and 82. Mendelson CR. Minireview: fetal-maternal hormonal signaling in androgen-binding proteins in nongenomic androgen actions. Mol pregnancy and labor. Mol Endocrinol 2009;23:947–54. Endocrinol 2002;16:2181–7. 83. Lindberg MK, Moverare S, Skrtic S, et al. Estrogen receptor (ER)- 103. Coleman KM, Smith CL. Intracellular signaling pathways: nonge- beta reduces ER alpha-regulated gene transcription, supporting a nomic actions of estrogens and ligand-independent activation of “ying yang” relationship between ER alpha and ER beta in mice. estrogen receptors. Front Biosci 2001;6:D1379–91. Mol Endocrinol 2003;17:203–8. 104. Weigel NL, Zhang Y. Ligand-independent activation of steroid 84. Safe S, Kim K. Non-classical genomic estrogen receptor (ER)/ hormone receptors. J Mol Med 1998;76:469–79. specificity protein and ER/activating protein-1 signaling path- 105. Lupien M, Meyer CA, Bailey ST, et al. Growth factor stimulation ways. J Mol Endocrinol 2008;41:263–75. induces a distinct ER(alpha) cistrome underlying breast cancer 85. Kushner PJ, Agard DA, Greene GL, et al. Estrogen receptor path- endocrine resistance. Genes Dev 2010;24:2219–27. ways to AP-1. J Steroid Biochem Mol Biol 2000;74:311–7. 106. Culig Z. Androgen receptor cross-talk with cell signalling path- 86. Safe S, Kim K. Nuclear receptor-mediated transactivation through ways. Growth Factors 2004;22:179–84. interaction with Sp proteins. Prog Nucleic Acid Res Mol Biol 107. Power RF, Conneely OM, O’Malley BW. New insights into activa- 2004;77:1–36. tion of the steroid hormone receptor superfamily. Trends Pharma- 87. Owen GI, Richer JK, Tung L, Takimoto G, Horwitz KB. Proges- col Sci 1992;13:318–23. terone regulates transcription of the p21(WAF1) cyclin-dependent 108. Mani SK, Oyola MG. Progesterone signaling mechanisms in brain kinase inhibitor gene through Sp1 and CBP/p300. J Biol Chem and behavior. Frontiers Endocrinol 2012;3(7):1–8. 1998;273:10696–701. 109. Fox EM, Andrade J, Shupnik MA. Novel actions of estrogen to 88. Levin ER. Minireview: extranuclear steroid receptors: roles in promote proliferation: integration of cytoplasmic and nuclear modulation of cell functions. Mol Endocrinol 2011;25:377–84. pathways. Steroids 2009;74:622–7. 89. Hammes SR, Levin ER. Minireview: recent advances in 110. Curtis SH, Korach KS. Steroid receptor knockout models: pheno- extranuclear steroid receptor actions. Endocrinology 2011;152: types and responses illustrate interactions between receptor sig- 4489–95. naling pathways in vivo. In: O’Malley BW, editor. Hormones and 90. Losel RM, Falkenstein E, Feuring M, et al. Nongenomic ste- signaling. San Diego: Academic Press; 1999. pp. 357–80. roid action: controversies, questions, and answers. Physiol Rev 111. Madak-Erdogan Z, Lupien M, Stossi F, Brown M, Katzenellenbo- 2003;83:965–1016. gen BS. Genomic collaboration of estrogen receptor alpha and 91. Losel RM, Besong D, Peluso JJ, Wehling M. Progesterone recep- extracellular signal-regulated kinase 2 in regulating gene and tor membrane component 1–many tasks for a versatile protein. proliferation programs. Mol Cell Biol 2011;31:226–36. Steroids 2008;73:929–34. 112. Nettles KW, Greene GL. Ligand control of coregulator recruit- 92. Segars JH, Driggers PH. Estrogen action and cytoplasmic signal- ment to nuclear receptors. Annu Rev Physiol 2005;67:309–33. ing cascades. Part I: membrane-associated signaling complexes. 113. Huang PX, Chandra V, Rastinejad F. Structural overview of the Trends Endocrinol Metab 2002;13:349–54. nuclear receptor superfamily: insights into physiology and thera- 93. Driggers PH, Segars JH. Estrogen action and cytoplasmic peutics. Annu Rev Physiol 2010;72:247–72. signaling pathways. Part II: the role of growth factors and 114. Renaud JP, Moras D. Structural studies on nuclear receptors. Cell phosphorylation in estrogen signaling. Trends Endocrinol Metab Mol Life Sci 2000;57:1748–69. 2002;13:422–7. 115. Bouchard P, Chabbert-Buffet N, Fauser BC. Selective progester- 94. Burns KA, Li Y, Arao Y, Petrovich RM, Korach KS. Selective one receptor modulators in reproductive medicine: pharmacol- mutations in estrogen receptor alpha D-domain alters nuclear ogy, clinical efficacy and safety. Fertil Steril 2011;96:1175–89. translocation and non-estrogen response element gene regulatory 116. Leonhardt SA, Edwards DP. Mechanism of action of progesterone mechanisms. J Biol Chem 2011;286:12640–9. antagonists. Exp Biol Med (Maywood) 2002;227:969–80. 95. Prossnitz ER, Barton M. The G-protein-coupled estrogen receptor 117. Dehm SM, Tindall DJ. Androgen receptor structural and func- GPER in health and disease. Nat Rev Endocrinol 2011;7:715–26. tional elements: role and regulation in prostate cancer. Mol 96. Langer G, Bader B, Meoli L, et al. A critical review of funda- Endocrinol 2007;21:2855–63. mental controversies in the field of GPR30 research. Steroids 118. Gobinet J, Poujol N, Sultan C. Molecular action of androgens. Mol 2010;75:603–10. Cell Endocrinol 2002;198:15–24. 97. Gao H, Dahlman-Wright K. The gene regulatory networks con- 119. Stockard CR, Papanicolaou GN. A rhythmical “heat period” in trolled by estrogens. Mol Cell Endocrinol 2011;334:83–90. the guinea-pig. Science 1917;46:42–4.

4. FEMALE REPRODUCTIVE SYSTEM 1178 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

120. Greco TL, Duello TM, Gorski J. Estrogen receptors, estradiol, and 139. Taylor KM, Gray CA, Joyce MM, Stewart MD, Bazer FW, Spencer diethylstilbestrol in early development: the mouse as a model for TE. Neonatal ovine uterine development involves alterations in the study of estrogen receptors and estrogen sensitivity in embry- expression of receptors for estrogen, progesterone, and prolactin. onic development of male and female reproductive tracts. Endocr Biol Reprod 2000;63:1192–204. Rev 1993;14:59–71. 140. Spencer TE, Bazer FW. Temporal and spatial alterations in uterine 121. Cunha GR, Cooke PS, Bigsby RM, Brody JR. Ontogeny of sex estrogen receptor and progesterone receptor gene expression dur- steroid receptors in mammals. In: Parker MG, editor. Nuclear ing the estrous cycle and early pregnancy in the ewe. Biol Reprod hormone receptors: molecular mechanisms, cellular, functions, clinical 1995;53:1527–43. abnormalities. London: Academic Press; 1991. pp. 235–68. 141. Meikle A, Sahlin L, Ferraris A, et al. Endometrial mRNA expres- 122. Korach KS, Horigome T, Tomooka Y, Yamashita S, Newbold RR, sion of oestrogen receptor alpha, progesterone receptor and insu- McLachlan JA. Immunodetection of estrogen receptor in epithe- lin-like growth factor-I (IGF-I) throughout the bovine oestrous lial and stromal tissues of neonatal mouse uterus. Proc Natl Acad cycle. Anim Reprod Sci 2001;68:45–56. Sci USA 1988;85:3334–7. 142. Kimmins S, MacLaren LA. Oestrous cycle and pregnancy effects 123. Yamashita S, Newbold RR, McLachlan JA, Korach KS. The role of on the distribution of oestrogen and progesterone receptors in the estrogen receptor in uterine epithelial proliferation and cytodif- bovine endometrium. Placenta 2001;22:742–8. ferentiation in neonatal mice. Endocrinology 1990;127:2456–63. 143. Lessey BA, Killam AP, Metzger DA, Haney AF, Greene GL, 124. Jefferson WN, Couse JF, Banks EP, Korach KS, Newbold RR. McCarty Jr KS. Immunohistochemical analysis of human uterine Expression of estrogen receptor beta is developmentally regu- estrogen and progesterone receptors throughout the menstrual lated in reproductive tissues of male and female mice. Biol Reprod cycle. J Clin Endocrinol Metab 1988;67:334–40. 2000;62:310–7. 144. Lessey BA, Metzger DA, Haney AF, McCarty Jr KS. Immunohisto- 125. Kurita T, Lee K, Cooke PS, Taylor JA, Lubahn DB, Cunha GR. Para- chemical analysis of estrogen and progesterone receptors in endome- crine regulation of epithelial progesterone receptor by estradiol in triosis: comparison with normal endometrium during the menstrual the mouse female reproductive tract. Biol Reprod 2000;62:821–30. cycle and the effect of medical therapy. Fertil Steril 1989;51:409–15. 126. Tan J, Paria BC, Dey SK, Das SK. Differential uterine expression 145. Noe M, Kunz G, Herbertz M, Mall G, Leyendecker G. The cyclic of estrogen and progesterone receptors correlates with uterine pattern of the immunocytochemical expression of oestrogen and preparation for implantation and decidualization in the mouse. progesterone receptors in human myometrial and endometrial Endocrinology 1999;140:5310–21. layers: characterization of the endometrial-subendometrial unit. 127. Couse JF, Lindzey J, Grandien K, Gustafsson JA, Korach KS. Tis- Hum Reprod 1999;14:190–7. sue distribution and quantitative analysis of estrogen receptor- 146. Bhakoo HS, Katzenellenbogen BS. Progesterone modulation of alpha (ERalpha) and estrogen receptor-beta (ERbeta) messenger estrogen-stimulated uterine biosynthetic events and estrogen ribonucleic acid in the wild-type and ERalpha-knockout mouse. receptor levels. Mol Cell Endocrinol 1977;8:121–34. Endocrinology 1997;138:4613–21. 147. Tsibris JC, Fort FL, Cazenave CR, et al. The uneven distribution 128. Kurita T, Cooke PS, Cunha GR. Epithelial-stromal tissue interac- of estrogen and progesterone receptors in human endometrium. J tion in paramesonephric (Mullerian) epithelial differentiation. Steroid Biochem 1981;14:997–1003. Dev Biol 2001;240:194–211. 148. Kauppila A, Janne O, Stenback F, Vihko R. Cytosolic estrogen 129. Wang H, Eriksson H, Sahlin L. Estrogen receptors alpha and beta and progestin receptors in human endometrium from different in the female reproductive tract of the rat during the estrous cycle. regions of the uterus. Gynecol Oncol 1982;14:225–9. Biol Reprod 2000;63:1331–40. 149. Matsuzaki S, Fukaya T, Suzuki T, Murakami T, Sasano H, Yajima 130. Pelletier G, Labrie C, Labrie F. Localization of oestrogen receptor A. Oestrogen receptor alpha and beta mRNA expression in alpha, oestrogen receptor beta and androgen receptors in the rat human endometrium throughout the menstrual cycle. Mol Hum reproductive organs. J Endocrinol 2000;165:359–70. Reprod 1999;5:559–64. 131. Toda K, Takeda K, Okada T, et al. Targeted disruption of the aro- 150. Lecce G, Meduri G, Ancelin M, Bergeron C, Perrot-Applanat M. matase P450 gene (Cyp19) in mice and their ovarian and uterine Presence of estrogen receptor beta in the human endometrium responses to 17beta-oestradiol. J Endocrinol 2001;170:99–111. through the cycle: expression in glandular, stromal, and vascular 132. Jakacka M, Ito M, Martinson F, Ishikawa T, Lee EJ, Jameson JL. cells. J Clin Endocrinol Metab 2001;86:1379–86. An estrogen receptor (ER)alpha deoxyribonucleic acid-binding 151. Saunders PT, Millar MR, Macpherson S, et al. ERbeta1 and the domain knock-in mutation provides evidence for nonclassical ER ERbeta2 splice variant (ERbetacx/beta2) are expressed in dis- pathway signaling in vivo. Mol Endocrinol 2002;16:2188–201. tinct cell populations in the adult human testis. J Clin Endocri- 133. Weihua Z, Saji S, Makinen S, et al. Estrogen receptor (ER) beta, nol Metab 2002;87:2706–15. a modulator of eralpha in the uterus. Proc Natl Acad Sci USA 152. Hild-Petito S, Verhage HG, Fazleabas AT. Immunocytochemical 2000;97:5936–41. localization of estrogen and progestin receptors in the baboon 134. Couse JF, Korach KS. Estrogen receptor null mice: what have we (papio anubis) uterus during implantation and pregnancy. Endocri- learned and where will they lead us? Endocr Rev 1999;20:358–417. nology 1992;130:2343–53. 135. Curtis Hewitt S, Goulding EH, Eddy EM, Korach KS. Studies 153. Koji T, Brenner RM. Localization of estrogen receptor messenger using the estrogen receptor alpha knockout uterus demonstrate ribonucleic acid in rhesus monkey uterus by nonradioactive in that implantation but not decidualization-associated signaling is situ hybridization with digoxigenin-labeled oligodeoxynucleo- estrogen dependent. Biol Reprod 2002;67:1268–77. tides. Endocrinology 1993;132:382–92. 136. Couse JF, Dixon D, Yates M, et al. Estrogen receptor-alpha knock- 154. Pelletier G, Luu-The V, Charbonneau A, Labrie F. Cellular out mice exhibit resistance to the developmental effects of neona- localization of estrogen receptor beta messenger ribonucleic tal diethylstilbestrol exposure on the female reproductive tract. acid in cynomolgus monkey reproductive organs. Biol Reprod Dev Biol 2001;238:224–38. 1999;61:1249–55. 137. Couse JF, Korach KS. Estrogen receptor-alpha mediates the detri- 155. Fazleabas AT, Brudney A, Chai D, Langoi D, Bulun SE. Steroid mental effects of neonatal diethylstilbestrol (DES) exposure in the receptor and aromatase expression in baboon endometriotic murine reproductive tract. Toxicology 2004;205:55–63. lesions. Fertil Steril 2003;80:820–7. 138. Spencer TE, Bazer FW. Biology of progesterone action during 156. Fisher CR, Graves KH, Parlow AF, Simpson ER. Characterization pregnancy recognition and maintenance of pregnancy. Front Biosci of mice deficient in aromatase (ArKO) because of targeted disrup- 2002;7:d1879–98. tion of the cyp19 gene. Proc Natl Acad Sci USA 1998;95:6965–70.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1179

157. Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, Smithies 175. Honda S, Harada N, Ito S, Takagi Y, Maeda S. Disruption of sexual O. Alteration of reproductive function but not prenatal sexual behavior in male aromatase-deficient mice lacking exons 1 and 2 development after insertional disruption of the mouse estrogen of the cyp19 gene. Biochem Biophys Res Commun 1998;252:445–9. receptor gene. Proc Natl Acad Sci USA 1993;90:11162–6. 176. Lydon JP, DeMayo FJ, Funk CR, et al. Mice lacking progesterone 158. Dupont S, Krust A, Gansmuller A, Dierich A, Chambon P, Mark M. receptor exhibit pleiotropic reproductive abnormalities. Genes Dev Effect of single and compound knockouts of estrogen receptors 1995;9:2266–78. alpha (ERalpha) and beta (ERbeta) on mouse reproductive phe- 177. Hashimoto-Partyka MK, Lydon JP, Iruela-Arispe ML. Generation notypes. Development 2000;127:4277–91. of a mouse for conditional excision of progesterone receptor. Genesis 159. Hewitt SC, Kissling GE, Fieselman KE, Jayes FL, Gerrish KE, 2006;44:391–5. Korach KS. Biological and biochemical consequences of global 178. Fernandez-Valdivia R, Jeong J, Mukherjee A, et al. A mouse model deletion of exon 3 from the ER alpha gene. FASEB J 2010;24:4660–7. to dissect progesterone signaling in the female reproductive tract 160. Antonson P, Omoto Y, Humire P, Gustafsson JA. Generation of and mammary gland. Genesis 2010;48:106–13. ERalpha-floxed and knockout mice using the Cre/LoxP system. 179. Franco HL, Rubel CA, Large MJ, et al. Epithelial progesterone Biochem Biophys Res Commun 2012;424:710–6. receptor exhibits pleiotropic roles in uterine development and 161. Curtis SW, Clark J, Myers P, Korach KS. Disruption of estrogen function. FASEB J 2012;26:1218–27. signaling does not prevent progesterone action in the estro- 180. Mulac-Jericevic B, Mullinax RA, DeMayo FJ, Lydon JP, Con- gen receptor or knockout mouse uterus. Proc Natl Acad Sci USA neely OM. Subgroup of reproductive functions of progesterone 1999;96:3646–51. mediated by progesterone receptor-B isoform. Science 2000;289: 162. O’Brien JE, Peterson TJ, Tong MH, et al. Estrogen-induced prolif- 1751–4. eration of uterine epithelial cells is independent of estrogen recep- 181. Conneely OM, Mulac-Jericevic B, Lydon JP. Progesterone-depen- tor alpha binding to classical estrogen response elements. J Biol dent regulation of female reproductive activity by two distinct Chem 2006;281:26683–92. progesterone receptor isoforms. Steroids 2003;68:771–8. 163. Hewitt SC, Li Y, Li L, Korach KS. Estrogen-mediated regulation 182. Conneely OM, Mulac-Jericevic B, DeMayo F, Lydon JP, O’Malley of IGF1 transcription and uterine growth involves direct binding BW. Reproductive functions of progesterone receptors. Recent of estrogen receptor alpha to estrogen-responsive elements. J Biol Prog Horm Res 2002;57:339–55. Chem 2010;285:2676–85. 183. Hu YC, Wang PH, Yeh S, et al. Subfertility and defective folliculo- 164. Ahlbory-Dieker DL, Stride BD, Leder G, et al. DNA binding genesis in female mice lacking androgen receptor. Proc Natl Acad by estrogen receptor-alpha is essential for the transcriptional Sci USA 2004;101:11209–14. response to estrogen in the liver and the uterus. Mol Endocrinol 184. Walters KA, McTavish KJ, Seneviratne MG, et al. Subfertile female 2009;23:1544–55. androgen receptor knockout mice exhibit defects in neuroendo- 165. Billon-Gales A, Fontaine C, Filipe C, et al. The transactivating function crine signaling, intraovarian function, and uterine development 1 of estrogen receptor alpha is dispensable for the vasculoprotective but not uterine function. Endocrinology 2009;150:3274–82. actions of 17beta-estradiol. Proc Natl Acad Sci USA 2009;106:2053–8. 185. Walters KA, Allan CM, Jimenez M, et al. Female mice haplo- 166. Abot A, Fontaine C, Raymond-Letron I, et al. The AF-1 activation insufficient for an inactivated androgen receptor (AR) exhibit function of estrogen receptor alpha is necessary and sufficient age-dependent defects that resemble the AR null phenotype of for uterine epithelial cell proliferation in vivo. Endocrinology dysfunctional late follicle development, ovulation, and fertility. 2013;154:2222–33. Endocrinology 2007;148:3674–84. 167. Billon-Gales A, Krust A, Fontaine C, et al. Activation function 2 186. Schauwaers K, De Gendt K, Saunders PT, et al. Loss of androgen (AF2) of estrogen receptor-alpha is required for the atheroprotec- receptor binding to selective androgen response elements causes tive action of estradiol but not to accelerate endothelial healing. a reproductive phenotype in a knockin mouse model. Proc Natl Proc Natl Acad Sci USA 2011;108:13311–6. Acad Sci USA 2007;104:4961–6. 168. Sinkevicius KW, Burdette JE, Woloszyn K, et al. An estrogen 187. Smith EP, Boyd J, Frank GR, et al. Estrogen resistance caused by receptor-alpha knock-in mutation provides evidence of ligand- a mutation in the estrogen-receptor gene in a man. N Engl J Med independent signaling and allows modulation of ligand-induced 1994;331:1056–61. pathways in vivo. Endocrinology 2008;149:2970–9. 188. Quaynor SD, Stradtman Jr EW, Kim HG, et al. Delayed puberty 169. Arao Y, Hamilton KJ, Ray MK, Scott G, Mishina Y, Korach KS. and estrogen resistance in a woman with estrogen receptor alpha Estrogen receptor alpha AF-2 mutation results in antagonist variant. N Engl J Med 2013;369:164–71. reversal and reveals tissue selective function of estrogen recep- 189. Couse JF, Curtis SW, Washburn TF, et al. Analysis of transcrip- tor modulators. Proc Natl Acad Sci USA 2011;108:14986–91. tion and estrogen insensitivity in the female mouse after tar- 170. Winuthayanon W, Hewitt SC, Orvis GD, Behringer RR, Korach KS. geted disruption of the estrogen receptor gene. Mol Endocrinol Uterine epithelial estrogen receptor alpha is dispensable for pro- 1995;9:1441–54. liferation but essential for complete biological and biochemical 190. Rumi MAK, Dhakal P, Kubota K, et al. Generation of esr1 knockout responses. Proc Natl Acad Sci USA 2010;107:19272–7. rats using zinc finger nuclease-mediated genome editing. Endocrinol- 171. Krege JH, Hodgin JB, Couse JF, et al. Generation and reproductive ogy 2014. http://dx.doi.org/10.1210/en.2013-2150. phenotypes of mice lacking estrogen receptor beta. Proc Natl Acad 191. Shughrue PJ, Askew GR, Dellovade TL, Merchenthaler I. Estrogen- Sci USA 1998;95:15677–82. binding sites and their functional capacity in estrogen receptor 172. Antal MC, Krust A, Chambon P, Mark M. Sterility and double knockout mouse brain. Endocrinology 2002;143:1643–50. absence of histopathological defects in nonreproductive 192. Cowley SM, Hoare S, Mosselman S, Parker MG. Estrogen recep- organs of a mouse ERbeta-null mutant. Proc Natl Acad Sci USA tors alpha and beta form heterodimers on DNA. J Biol Chem 1997; 2008;105:2433–8. 272:19858–62. 173. Wada-Hiraike O, Hiraike H, Okinaga H, et al. Role of estro- 193. Pettersson K, Grandien K, Kuiper GG, Gustafsson JA. Mouse gen receptor beta in uterine stroma and epithelium: insights estrogen receptor beta forms estrogen response element- binding from estrogen receptor beta-/- mice. Proc Natl Acad Sci USA heterodimers with estrogen receptor alpha. Mol Endocrinol 2006;103:18350–5. 1997;11:1486–96. 174. Couse JF, Hewitt SC, Bunch DO, et al. Postnatal sex reversal of the 194. Britt KL, Kerr J, O’Donnell L, et al. Estrogen regulates development ovaries in mice lacking estrogen receptors alpha and beta. Science of the somatic cell phenotype in the eutherian ovary. FASEB J 1999;286:2328–31. 2002;16:1389–97.

4. FEMALE REPRODUCTIVE SYSTEM 1180 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

195. Simpson ER, Clyne C, Rubin G, et al. Aromatase–a brief over- 214. Galand P, Leroy F, Chretien J. Effect of oestradiol on cell prolifera- view. Annu Rev Physiol 2002;64:93–127. tion and histological changes in the uterus and vagina of mice. 196. Korach KS, Couse JF, Curtis SW, et al. Estrogen receptor gene dis- J Endocrinol 1971;49:243–52. ruption: molecular characterization and experimental and clinical 215. Kronenberg MS, Clark JH. Changes in keratin expression during phenotypes. Recent Prog Horm Res 1996;51:159–86; discussion 86–8. the estrogen-mediated differentiation of rat vaginal epithelium. 197. Walker VR, Korach KS. Estrogen receptor knockout mice as a Endocrinology 1985;117:1480–9. model for endocrine research. ILAR J 2004;45:455–61. 216. Gimenez-Conti IB, Lynch M, Roop D, Bhowmik S, Majeski P, 198. Huang CC, Orvis GD, Wang Y, Behringer RR. Stromal-to- epithelial Conti CJ. Expression of keratins in mouse vaginal epithelium. transition during postpartum endometrial regeneration. PLoS Differentiation 1994;56:143–51. One 2012;7:e44285. 217. Sourla A, Luo S, Labrie C, Belanger A, Labrie F. Morphological 199. Hewitt SC, O’Brien JE, Jameson JL, Kissling GE, Korach KS. Selec- changes induced by 6-month treatment of intact and ovariecto- tive disruption of ER alpha DNA-binding activity alters uterine mized mice with tamoxifen and the pure antiestrogen EM-800. responsiveness to estradiol. Mol Endocrinol 2009;23:2111–6. Endocrinology 1997;138:5605–17. 200. Arao Y, Hamilton KJ, Coons LA, Korach KS. Estrogen recep- 218. Luo S, Martel C, Sourla A, et al. Comparative effects of 28-day tor alpha L543A, L544A mutation changes antagonists to treatment with the new anti-estrogen EM-800 and tamoxifen agonists which correlates with the ligand binding domain on estrogen-sensitive parameters in intact mice. Int J Cancer dimerization associated with DNA binding activity. J Biol Chem 1997;73:381–91. 2013;288:21105–16. 219. Dukes M, Chester R, Yarwood L, Wakeling AE. Effects of a non- 201. Xu J, Qiu Y, DeMayo FJ, Tsai SY, Tsai MJ, O’Malley BW. Partial steroidal pure antioestrogen, ZM 189,154, on oestrogen target hormone resistance in mice with disruption of the steroid recep- organs of the rat including bones. J Endocrinol 1994;141:335–41. tor coactivator-1 (SRC-1) gene. Science 1998;279:1922–5. 220. Chung SH, Franceschi S, Lambert PF. Estrogen and ER alpha: cul- 202. Kurihara I, Lee DK, Petit FG, et al. COUP-TFII mediates proges- prits in cervical cancer? Trends Endocrinol Metab 2010;21:504–11. terone regulation of uterine implantation by controlling ER activ- 221. Chung SH, Wiedmeyer K, Shai A, Korach KS, Lambert PF. ity. PLoS Genet 2007;3:e102. Requirement for estrogen receptor alpha in a mouse model for 203. Lee DK, Kurihara I, Jeong JW, et al. Suppression of ERalpha activ- human papillomavirus-associated cervical cancer. Cancer Res ity by COUP-TFII is essential for successful implantation and 2008;68:9928–34. decidualization. Mol Endocrinol 2010;24:930–40. 222. Bazer FW, Slayden OD. Progesterone-induced gene expression in 204. Petit FG, Jamin SP, Kurihara I, et al. Deletion of the orphan nuclear uterine epithelia: a myth perpetuated by conventional wisdom. receptor COUP-TFII in uterus leads to placental deficiency. Proc Biol Reprod 2008;79:1008–9. Natl Acad Sci USA 2007;104:6293–8. 223. Tibbetts TA, Mendoza-Meneses M, O’Malley BW, Conneely OM. 205. Montano MM, Ekena K, Delage-Mourroux R, Chang W, Mutual and intercompartmental regulation of estrogen recep- Martini P, Katzenellenbogen BS. An estrogen receptor-selective tor and progesterone receptor expression in the mouse uterus. coregulator that potentiates the effectiveness of antiestrogens Biol Reprod 1998;59:1143–52. and represses the activity of estrogens. Proc Natl Acad Sci USA 224. Ismail PM, Li J, DeMayo FJ, O’Malley BW, Lydon JP. A novel 1999;96:6947–52. LacZ reporter mouse reveals complex regulation of the proges- 206. Mussi P, Liao L, Park SE, et al. Haploinsufficiency of the core- terone receptor promoter during mammary gland development. pressor of estrogen receptor activity (REA) enhances estrogen Mol Endocrinol 2002;16:2475–89. receptor function in the mammary gland. Proc Natl Acad Sci USA 225. Mote PA, Johnston JF, Manninen T, Tuohimaa P, Clarke CL. 2006;103:16716–21. Detection of progesterone receptor forms A and B by immunohis- 207. Park S, Yoon S, Zhao Y, et al. Uterine development and fertility tochemical analysis. J Clin Pathol 2001;54:624–30. are dependent on gene dosage of the nuclear receptor coregulator 226. Arnett-Mansfield RL, DeFazio A, Mote PA, Clarke CL. Subnu- REA. Endocrinology 2012;153:3982–94. clear distribution of progesterone receptors A and B in normal 208. Park SE, Xu J, Frolova A, Liao L, O’Malley BW, Katzenellenbogen BS. and malignant endometrium. J Clin Endocrinol Metab 2004;89: Genetic deletion of the repressor of estrogen receptor activity 1429–42. (REA) enhances the response to estrogen in target tissues in vivo. 227. Mulac-Jericevic B, Lydon JP, DeMayo FJ, Conneely OM. Defective Mol Cell Biol 2005;25:1989–99. mammary gland morphogenesis in mice lacking the progesterone 209. Delage-Mourroux R, Martini PG, Choi I, Kraichely DM, receptor B isoform. Proc Natl Acad Sci USA 2003;100:9744–9. Hoeksema J, Katzenellenbogen BS. Analysis of estrogen receptor 228. Tong W, Pollard JW. Progesterone inhibits estrogen-induced interaction with a repressor of estrogen receptor activity (REA) cyclin D1 and cdk4 nuclear translocation, cyclin E- and cyclin and the regulation of estrogen receptor transcriptional activity by A-cdk2 kinase activation, and cell proliferation in uterine epithe- REA. J Biol Chem 2000;275:35848–56. lial cells in mice. Mol Cell Biol 1999;19:2251–64. 210. Nallasamy S, Li Q, Bagchi MK, Bagchi IC. Msx homeobox genes 229. Quarmby VE, Korach KS. The influence of 17 beta-estradiol on pat- critically regulate embryo implantation by controlling paracrine terns of cell division in the uterus. Endocrinology 1984;114:694–702. signaling between uterine stroma and epithelium. PLoS Genet 230. Das RM, Martin L. Progesterone inhibition of mouse uterine 2012;8:e1002500. epithelial proliferation. J Endocrinol 1973;59:205–6. 211. Kawagoe J, Li Q, Mussi P, et al. Nuclear receptor coactivator-6 231. Martin L, Das RM, Finn CA. The inhibition by progesterone attenuates uterine estrogen sensitivity to permit embryo implan- of uterine epithelial proliferation in the mouse. J Endocrinol tation. Dev Cell 2012;23:858–65. 1973;57:549–54. 212. Li Q, Kannan A, DeMayo FJ, et al. The antiproliferative action of 232. Jeong JW, Lee KY, Han SJ, et al. The P160 steroid receptor coacti- progesterone in uterine epithelium is mediated by Hand2. Science vator 2, Src-2, regulates murine endometrial function and regu- 2011;331:912–6. lates progesterone-independent and -dependent gene expression. 213. Stumpf W, Sar M. Autoradiographic localization of estrogen, Endocrinology 2007;148:4238–50. androgen, progestin, and glucocorticoid in “target tissues” and 233. Mukherjee A, Amato P, Allred DC, et al. Steroid receptor coactiva- “non-target” tissues. In: Pasqualini J, editor. Receptors and mecha- tor 2 is essential for progesterone-dependent uterine function and nisms of action of steroid hormones. New York: Marcel Dekker; 1976. mammary morphogenesis: insights from the mouse–implications pp. 41–84. for the human. J Steroid Biochem Mol Biol 2006;102:22–31.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1181

234. Han SJ, DeMayo FJ, Xu J, Tsai SY, Tsai MJ, O’Malley BW. Steroid 254. Rajkumar K, Dheen T, Krsek M, Murphy LJ. Impaired estrogen receptor coactivator (SRC)-1 and SRC-3 differentially modulate action in the uterus of insulin-like growth factor binding protein-1 tissue-specific activation functions of the progesterone receptor. transgenic mice. Endocrinology 1996;137:1258–64. Mol Endocrinol 2006;20:45–55. 255. Sato T, Wang G, Hardy MP, Kurita T, Cunha GR, Cooke PS. Role of 235. Gehin M, Mark M, Dennefeld C, Dierich A, Gronemeyer H, Cham- systemic and local IGF-I in the effects of estrogen on growth and epi- bon P. The function of TIF2/GRIP1 in mouse reproduction is dis- thelial proliferation of mouse uterus. Endocrinology 2002;143:2673–9. tinct from those of SRC-1 and p/CIP. Mol Cell Biol 2002;22:5923–37. 256. Hewitt SC, Collins J, Grissom S, Deroo B, Korach KS. Global 236. Fernandez-Valdivia R, Mukherjee A, Amato P, et al. Progester- uterine genomics in vivo: microarray evaluation of the estro- one-action in the murine uterus and mammary gland requires gen receptor alpha-growth factor cross-talk mechanism. Mol steroid receptor coactivator 2: relevance to the human. Front Biosci Endocrinol 2005;19:657–68. 2007;12:3640–7. 257. Klotz DM, Hewitt SC, Ciana P, et al. Requirement of estrogen 237. Tranguch S, Cheung-Flynn J, Daikoku T, et al. Cochaperone receptor-alpha in insulin-like growth factor-1 (IGF-1)-induced immunophilin FKBP52 is critical to uterine receptivity for embryo uterine responses and in vivo evidence for IGF-1/estrogen recep- implantation. Proc Natl Acad Sci USA 2005;102:14326–31. tor cross-talk. J Biol Chem 2002;277:8531–7. 238. Yang Z, Wolf IM, Chen H, et al. FK506-binding protein 52 is essen- 258. Curtis SW, Washburn T, Sewall C, et al. Physiological coupling of tial to uterine reproductive physiology controlled by the proges- growth factor and steroid receptor signaling pathways: estrogen terone receptor A isoform. Mol Endocrinol 2006;20:2682–94. receptor knockout mice lack estrogen-like response to epidermal 239. Katzenellenbogen BS, Greger NG. Ontogeny of uterine respon- growth factor. Proc Natl Acad Sci USA 1996;93:12626–30. siveness to estrogen during early development in the rat. Mol Cell 259. Klotz DM, Hewitt SC, Korach KS, Diaugustine RP. Activation of a Endocrinol 1974;2:31–42. uterine insulin-like growth factor I signaling pathway by clinical 240. Cooke PS, Buchanan DL, Young P, et al. Stromal estrogen receptors and environmental estrogens: requirement of estrogen receptor- mediate mitogenic effects of estradiol on uterine epithelium. Proc alpha. Endocrinology 2000;141:3430–9. Natl Acad Sci USA 1997;94:6535–40. 260. Chen B, Pan H, Zhu L, Deng Y, Pollard JW. Progesterone inhibits 241. Clark JH, Markaverich BM. Actions of ovarian steroid hormones. the estrogen-induced phosphoinositide 3-kinase-->AKT-->GSK- In: Knobil E, Neill JD, editors. The physiology of reproduction. 3beta-->cyclin D1-->pRB pathway to block uterine epithelial cell New York: Raven Press; 1988. pp. 675–724. proliferation. Mol Endocrinol 2005;19:1978–90. 242. Hewitt SC, Deroo BJ, Hansen K, et al. Estrogen receptor-dependent 261. Wang Y, Feng H, Bi C, Zhu L, Pollard JW, Chen B. GSK-3beta genomic responses in the uterus mirror the biphasic physiological mediates in the progesterone inhibition of estrogen induced response to estrogen. Mol Endocrinol 2003;17:2070–83. cyclin D2 nuclear localization and cell proliferation in cyclin 243. Hewitt SC, Li L, Grimm SA, et al. Research resource: whole- D1-/- mouse uterine epithelium. FEBS Lett 2007;581:3069–75. genome estrogen receptor alpha binding in mouse uterine tissue 262. Buchanan DL, Kurita T, Taylor JA, Lubahn DB, Cunha GR, revealed by chip-seq. Mol Endocrinol 2012;26:887–98. Cooke PS. Role of stromal and epithelial estrogen receptors in 244. Korach KS. Insights from the study of animals lacking functional vaginal epithelial proliferation, stratification, and cornification. estrogen receptor. Science 1994;266:1524–7. Endocrinology 1998;139:4345–52. 245. Mantena SR, Kannan A, Cheon YP, et al. C/EBPbeta is a critical 263. Tibbetts TA, Conneely OM, O’Malley BW. Progesterone via its mediator of steroid hormone-regulated cell proliferation and dif- receptor antagonizes the pro-inflammatory activity of estrogen in ferentiation in the uterine epithelium and stroma. Proc Natl Acad the mouse uterus. Biol Reprod 1999;60:1158–65. Sci USA 2006;103:1870–5. 264. Kelly RW, Leask R, Calder AA. Choriodecidual production of 246. Ramathal C, Bagchi IC, Bagchi MK. Lack of CCAAT enhancer interleukin-8 and mechanism of parturition. Lancet 1992;339:776–7. binding protein beta (C/EBPbeta) in uterine epithelial cells 265. Ito A, Imada K, Sato T, Kubo T, Matsushima K, Mori Y. Suppres- impairs estrogen-induced DNA replication, induces DNA dam- sion of interleukin 8 production by progesterone in rabbit uterine age response pathways, and promotes apoptosis. Mol Cell Biol cervix. Biochem J 1994;301:183–6. 2010;30:1607–19. 266. Cheng L, Kelly RW, Thong KJ, Hume R, Baird DT. The effect of 247. Bagchi MK, Mantena SR, Kannan A, Bagchi IC. Role of C/EBP mifepristone (RU486) on the immunohistochemical distribution beta in steroid-induced cell proliferation and differentiation in the of prostaglandin E and its metabolite in decidual and chorionic uterus: functional implications for establishment of early preg- tissue in early pregnancy. J Clin Endocrinol Metab 1993;77:873–7. nancy. Placenta 2006;27:A13–A. 267. Lydon JP, DeMayo FJ, Conneely OM, O’Malley BW. Reproduc- 248. Pan H, Deng Y, Pollard JW. Progesterone blocks estrogen-induced tive phenotypes of the progesterone receptor null mutant mouse. DNA synthesis through the inhibition of replication licensing. J Steroid Biochem Mol Biol 1996;56:67–77. Proc Natl Acad Sci USA 2006;103:14021–6. 268. Wang H, Dey SK. Roadmap to embryo implantation: clues from 249. Ray S, Pollard JW. KLF15 negatively regulates estrogen-induced mouse models. Nat Rev Genet 2006;7:185–99. epithelial cell proliferation by inhibition of DNA replication 269. Bazer FW, Spencer TE, Johnson GA, Burkhart RC, Wu G. Comparative licensing. Proc Natl Acad Sci USA 2012;109:E1334–43. aspects of implantation. Reproduction (Cambridge) 2009;138:195–209. 250. Cunha GR, Cooke PS, Kurita T. Role of stromal-epithelial interac- 270. Finn CA. Oestrogen and the decidual cell reaction of implantation tions in hormonal responses. Arch Histol Cytol 2004;67:417–34. in mice. J Endocrinol 1965;32:223–9. 251. Richards RG, DiAugustine RP, Petrusz P, Clark GC, Sebastian J. 271. Paria BC, Tan J, Lubahn DB, Dey SK, Das SK. Uterine decidual Estradiol stimulates tyrosine phosphorylation of the insulin-like response occurs in estrogen receptor-alpha-deficient mice. Endo- growth factor-1 receptor and insulin receptor substrate-1 in the crinology 1999;140:2704–10. uterus. Proc Natl Acad Sci USA 1996;93:12002–7. 272. Grummer R, Hewitt SW, Traub O, Korach KS, Winterhager E. Dif- 252. Zhu L, Pollard JW. Estradiol-17 beta regulates mouse uterine ferent regulatory pathways of endometrial connexin expression: epithelial cell proliferation through insulin-like growth factor 1 preimplantation hormonal-mediated pathway versus embryo signaling. Proc Natl Acad Sci USA 2007;104:15847–51. implantation-initiated pathway. Biol Reprod 2004;71:273–81. 253. Adesanya OO, Zhou J, Samathanam C, Powell-Braxton L, Bondy CA. 273. Das A, Mantena SR, Kannan A, Evans DB, Bagchi MK, Bagchi IC. Insulin-like growth factor 1 is required for G2 progression in the De novo synthesis of estrogen in pregnant uterus is critical for estradiol-induced mitotic cycle. Proc Natl Acad Sci USA 1999;96 stromal decidualization and angiogenesis. Proc Natl Acad Sci USA :3287–91. 2009;106:12542–7.

4. FEMALE REPRODUCTIVE SYSTEM 1182 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

274. Pelletier G. Localization of androgen and estrogen receptors in rat 294. Zhang J, Sun Y, Liu Y, Liao DJ. Synergistic effects of androgen and primate tissues. Histol Histopathol 2000;15:1261–70. and estrogen on the mouse uterus and mammary gland. Oncol 275. Hirai M, Hirata S, Osada T, Hagihara K, Kato J. Androgen recep- Rep 2004;12:709–16. tor mRNA in the rat ovary and uterus. J Steroid Biochem Mol Biol 295. Cloke B, Huhtinen K, Fusi L, et al. The androgen and progesterone 1994;49:1–7. receptors regulate distinct gene networks and cellular functions 276. Pelletier G, Luu-The V, Li S, Labrie F. Localization and estrogenic in decidualizing endometrium. Endocrinology 2008;149:4462–74. regulation of androgen receptor mRNA expression in the mouse 296. Henderson TA, Saunders PT, Moffett-King A, Groome NP, uterus and vagina. J Endocrinol 2004;180:77–85. Critchley HO. Steroid receptor expression in uterine natural killer 277. Kimura N, Mizokami A, Oonuma T, Sasano H, Nagura H. Immu- cells. J Clin Endocrinol Metab 2003;88:440–9. nocytochemical localization of androgen receptor with polyclonal 297. Cole TJ, Blendy JA, Monaghan AP, et al. Targeted disruption of antibody in paraffin-embedded human tissues. J Histochem Cyto- the glucocorticoid receptor gene blocks adrenergic chromaffin chem 1993;41:671–8. cell development and severely retards lung maturation. Genes Dev 278. Slomczynska M, Duda M, Burek M, Knapczyk K, Czaplicki D, 1995;9:1608–21. Koziorowski M. Distribution of androgen receptor in the por- 298. Rhen T, Cidlowski JA. Estrogens and glucocorticoids have oppos- cine uterus throughout pregnancy. Reprod Domest Anim 2008;43: ing effects on the amount and latent activity of complement pro- 35–41. teins in the rat uterus. Biol Reprod 2006;74:265–74. 279. Weihua Z, Ekman J, Almkvist A, et al. Involvement of andro- 299. Rabin DS, Johnson EO, Brandon DD, Liapi C, Chrousos GP. Glu- gen receptor in 17beta-estradiol-induced cell proliferation in rat cocorticoids inhibit estradiol-mediated uterine growth: possible uterus. Biol Reprod 2002;67:616–23. role of the uterine estradiol receptor. Biol Reprod 1990;42:74–80. 280. Mertens HJ, Heineman MJ, Theunissen PH, de Jong FH, 300. Whirledge S, Cidlowski JA. Estradiol antagonism of glucocorti- Evers JL. Androgen, estrogen and progesterone receptor expres- coid-induced GILZ expression in human uterine epithelial cells sion in the human uterus during the menstrual cycle. Eur J Obstet and murine uterus. Endocrinology 2013;154:499–510. Gynecol Reprod Biol 2001;98:58–65. 301. Dosiou C, Giudice LC. Natural killer cells in pregnancy and 281. Kerkhofs S, Denayer S, Haelens A, Claessens F. Androgen recep- recurrent pregnancy loss: endocrine and immunologic perspec- tor knockout and knock-in mouse models. J Mol Endocrinol tives. Endocr Rev 2005;26:44–62. 2009;42:11–7. 302. Nothnick WB. The role of micro-RNAs in the female reproductive 282. Zhou X. Roles of androgen receptor in male and female repro- tract. Reproduction 2012;143:559–76. duction: lessons from global and cell-specific androgen receptor 303. Hawkins SM, Buchold GM, Matzuk MM. Minireview: the roles knockout (ARKO) mice. J Androl 2010;31:235–43. of small RNA pathways in reproductive medicine. Mol Endocrinol 283. Lyon MF, Hawkes SG. X-linked gene for testicular feminization in 2011;25:1257–79. the mouse. Nature 1970;227:1217–9. 304. Nothnick WB, Healy C. Estrogen induces distinct patterns 284. Young CY, Johnson MP, Prescott JL, Tindall DJ. The androgen of microRNA expression within the mouse uterus. Reprod Sci receptor of the testicular-feminized (Tfm) mutant mouse is smaller 2010;17:987–94. than the wild-type receptor. Endocrinology 1989;124:771–5. 305. Klinge CM. miRNAs and estrogen action. Trends Endocrinol Metab 285. Gaspar ML, Meo T, Bourgarel P, Guenet JL, Tosi M. A single base 2012;23:223–33. deletion in the Tfm androgen receptor gene creates a short-lived 306. Harfe BD, McManus MT, Mansfield JH, Hornstein E, Tabin CJ. messenger RNA that directs internal translation initiation. Proc The RNAseIII enzyme Dicer is required for morphogenesis but Natl Acad Sci USA 1991;88:8606–10. not patterning of the vertebrate limb. Proc Natl Acad Sci USA 286. Patterson MN, McPhaul MJ, Hughes IA. Androgen insensitivity 2005;102:10898–903. syndrome. In: Sheppard MC, Stewart PM, editors. Bailliere’s clini- 307. O’Rourke JR, Georges SA, Seay HR, et al. Essential role for dicer cal endocrinology and metabolism: hormones, enzymes and receptors. during skeletal muscle development. Dev Biol 2007;311:359–68. London: Bailliere Tindall; 1994. pp. 379–404. 308. Harris KS, Zhang Z, McManus MT, Harfe BD, Sun X. Dicer func- 287. Lyon MF, Glenister PH. Reduced reproductive performance in tion is essential for lung epithelium morphogenesis. Proc Natl androgen-resistant Tfm/Tfm female mice. Proc R Soc Lond B Biol Acad Sci USA 2006;103:2208–13. Sci 1980;208:1–12. 309. Murchison EP, Stein P, Xuan Z, et al. Critical roles for Dicer in the 288. Kato S, Matsumoto T, Kawano H, Sato T, Takeyama K. Function female germline. Genes Dev 2007;21:682–93. of androgen receptor in gene regulations. J Steroid Biochem Mol 310. Chakrabarty A, Tranguch S, Daikoku T, Jensen K, Furneaux H, Biol 2004;89–90:627–33. Dey SK. MicroRNA regulation of cyclooxygenase-2 during 289. Sato T, Kawano H, Kato S. Study of androgen action in bone by embryo implantation. Proc Natl Acad Sci USA 2007;104:15144–9. analysis of androgen-receptor deficient mice. J Bone Miner Metab 311. Hong X, Luense LJ, McGinnis LK, Nothnick WB, Christenson LK. 2002;20:326–30. Dicer1 is essential for female fertility and normal development 290. Yeh S, Tsai MY, Xu Q, et al. Generation and characterization of of the female reproductive system. Endocrinology 2008;149: androgen receptor knockout (ARKO) mice: an in vivo model for 6207–12. the study of androgen functions in selective tissues. Proc Natl Acad 312. Nagaraja AK, Andreu-Vieyra C, Franco HL, et al. Deletion of Sci USA 2002;99:13498–503. Dicer in somatic cells of the female reproductive tract causes ste- 291. Armstrong DT, Papkoff H. Stimulation of aromatization of exog- rility. Mol Endocrinol 2008;22:2336–52. enous and endogenous androgens in ovaries of hypophysecto- 313. Gonzalez G, Behringer RR. Dicer is required for female reproduc- mized rats in vivo by follicle-stimulating hormone. Endocrinology tive tract development and fertility in the mouse. Mol Reprod Dev 1976;99:1144–51. 2009;76:678–88. 292. Nantermet PV, Masarachia P, Gentile MA, et al. Androgenic 314. Hawkins SM, Andreu-Vieyra CV, Kim TH, et al. Dysregulation of induction of growth and differentiation in the rodent uterus uterine signaling pathways in progesterone receptor-Cre knock- involves the modulation of estro gen- regulated genetic pathways. out of dicer. Mol Endocrinol 2012;26:1552–66. Endocrinology 2005;146:564–78. 315. Ho Hong S, Young Nah H, Yoon Lee J, Chan Gye M, Hoon Kim C, 293. Hewitt SC, Collins J, Grissom S, Hamilton K, Korach KS. Estren Kyoo Kim M. Analysis of estrogen-regulated genes in mouse behaves as a weak estrogen rather than a nongenomic selective uterus using cDNA microarray and laser capture microdissection. activator in the mouse uterus. Endocrinology 2006;147:2203–14. J Endocrinol 2004;181:157–67.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1183

316. Watanabe H, Suzuki A, Kobayashi M, et al. Analysis of tempo- 335. Ivanga M, Labrie Y, Calvo E, et al. Fine temporal analysis of DHT ral changes in the expression of estrogen-regulated genes in the transcriptional modulation of the ATM/Gadd45g signaling path- uterus. J Mol Endocrinol 2003;30:347–58. ways in the mouse uterus. Mol Reprod Dev 2009;76:278–88. 317. Fertuck KC, Eckel JE, Gennings C, Zacharewski TR. Identification 336. Hewitt SC, Korach KS. Estrogenic activity of bisphenol A and of temporal patterns of gene expression in the uteri of immature, 2,2-bis(p-hydroxyphenyl)-1,1,1-trichloroethane (HPTE) demon- ovariectomized mice following exposure to ethynylestradiol. strated in mouse uterine gene profiles. Environ Health Perspect Physiol Genomics 2003;15:127–41. 2011;119:63–70. 318. Andrade PM, Silva I, Borra RC, de LGR, Baracat EC. Estrogen 337. Farnham PJ. Insights from genomic profiling of transcription regulation of uterine genes in vivo detected by complementary factors. Nat Rev Genet 2009;10:605–16. DNA array. Horm Metab Res 2002;34:238–44. 338. Park PJ. Chip-seq: advantages and challenges of a maturing tech- 319. Moggs JG, Ashby J, Tinwell H, et al. The need to decide if nology. Nat Rev Genet 2009;10:669–80. all estrogens are intrinsically similar. Environ Health Perspect 339. Martens JH, Rao NA, Stunnenberg HG. Genome-wide interplay 2004;112:1137–42. of nuclear receptors with the epigenome. Biochim Biophys Acta 320. Hong EJ, Park SH, Choi KC, Leung PC, Jeung EB. Identification of 2011;1812:818–23. estrogen-regulated genes by microarray analysis of the uterus of 340. Green CD, Han JDJ. Epigenetic regulation by nuclear receptors. immature rats exposed to endocrine disrupting chemicals. Reprod Epigenomics 2011;3:59–72. Biol Endocrinol 2006;4:49. 341. Meyer CA, Tang Q, Liu XS. Minireview: applications of next- 321. Groothuis PG, Dassen HH, Romano A, Punyadeera C. Estrogen generation sequencing on studies of nuclear receptor regulation and the endometrium: lessons learned from gene expression and function. Mol Endocrinol 2012;26:1651–9. profiling in rodents and human. Hum Reprod Update 2007; 342. Deblois G, Giguere V. Nuclear receptor location analyses in mam- 13:405–17. malian genomes: from gene regulation to regulatory networks. 322. Kuokkanen S, Chen B, Ojalvo L, Benard L, Santoro N, Pollard JW. Mol Endocrinol 2008;22:1999–2011. Genomic profiling of microRNAs and messenger RNAs reveals 343. Tang Q, Chen Y, Meyer C, et al. A comprehensive view of nuclear hormonal regulation in microRNA expression in human endome- receptor cancer cistromes. Cancer Res 2011;71:6940–7. trium. Biol Reprod 2010;82:791–801. 344. Cheung E, Kraus WL. Genomic analyses of hormone signaling 323. Aghajanova L, Hamilton AE, Giudice LC. Uterine receptivity to and gene regulation. Annu Rev Physiol 2010;72:191–218. human embryonic implantation: histology, biomarkers, and tran- 345. Gilfillan S, Fiorito E, Hurtado A. Functional genomic methods to scriptomics. Semin Cell Dev Biol 2008;19:204–11. study estrogen receptor activity. J Mammary Gland Biol Neoplasia 324. Kao LC, Tulac S, Lobo S, et al. Global gene profiling in human 2012;17:147–53. endometrium during the window of implantation. Endocrinology 346. Gao H, Falt S, Sandelin A, Gustafsson JA, Dahlman-Wright K. 2002;143:2119–38. Genome-wide identification of estrogen receptor alpha-binding 325. Borthwick JM, Charnock-Jones DS, Tom BD, et al. Determination sites in mouse liver. Mol Endocrinol 2008;22:10–22. of the transcript profile of human endometrium. Mol Hum Reprod 347. Rubel CA, Lanz RB, Kommagani R, Franco HL, Lydon JP, 2003;9:19–33. Demayo FJ. Research resource: genome-wide profiling of pro- 326. Cheon YP, Li Q, Xu X, DeMayo FJ, Bagchi IC, Bagchi MK. A gesterone receptor binding in the mouse uterus. Mol Endocrinol genomic approach to identify novel progesterone receptor regu- 2012;26:1428–42. lated pathways in the uterus during implantation. Mol Endocrinol 348. Yin P, Roqueiro D, Huang L, et al. Genome-wide progesterone 2002;16:2853–71. receptor binding: cell type-specific and shared mechanisms in 327. Yao MW, Lim H, Schust DJ, et al. Gene expression profiling T47D breast cancer cells and primary leiomyoma cells. PLoS One reveals progesterone-mediated cell cycle and immunoregulatory 2012;7:e29021. roles of Hoxa-10 in the preimplantation uterus. Mol Endocrinol 349. Hu S, Yao G, Guan X, et al. Research resource: genome-wide map- 2003;17:610–27. ping of in vivo androgen receptor binding sites in mouse epididy- 328. Jeong JW, Lee KY, Kwak I, et al. Identification of murine uterine mis. Mol Endocrinol 2010;24:2392–405. genes regulated in a ligand-dependent manner by the progester- 350. Wyce A, Bai Y, Nagpal S, Thompson CC. Research resource: the one receptor. Endocrinology 2005;146:3490–505. androgen receptor modulates expression of genes with criti- 329. An BS, Choi KC, Kang SK, et al. Mouse calbindin-D-9k gene cal roles in muscle development and function. Mol Endocrinol expression in the uterus during late pregnancy and lactation. Mol 2010;24:1665–74. Cell Endocrinol 2003;205:79–88. 351. Need EF, Selth LA, Harris TJ, Birrell SN, Tilley WD, Buchanan G. 330. Li Q, Cheon YP, Kannan A, Shanker S, Bagchi IC, Bagchi MK. A Research resource: interplay between the genomic and tran- novel pathway involving progesterone receptor, 12/15-lipoxygen- scriptional networks of androgen receptor and estrogen ase-derived eicosanoids, and peroxisome proliferator-activated receptor alpha in luminal breast cancer cells. Mol Endocrinol receptor gamma regulates implantation in mice. J Biol Chem 2004; 2012;26:1941–52. 279:11570–81. 352. He HH, Meyer CA, Chen MW, Jordan VC, Brown M, Liu XS. Dif- 331. Ramathal CY, Bagchi IC, Taylor RN, Bagchi MK. Endometrial ferential DNAse I hypersensitivity reveals factor-dependent chro- decidualization: of mice and men. Semin Reprod Med 2010;28: matin dynamics. Genome Res 2012;22:1015–25. 17–26. 353. Phuc Le P, Friedman JR, Schug J, et al. Glucocorticoid receptor- 332. Richard C, Gao J, Brown N, Reese J. Aquaporin water channel dependent gene regulatory networks. PLoS Genet 2005;1:e16. genes are differentially expressed and regulated by ovarian ste- 354. Biddie SC, John S, Sabo PJ, et al. Transcription factor AP1 potenti- roids during the periimplantation period in the mouse. Endocri- ates chromatin accessibility and glucocorticoid receptor binding. nology 2003;144:1533–41. Mol Cell 2011;43:145–55. 333. Deroo BJ, Hewitt SC, Peddada SD, Korach KS. Estradiol regulates 355. Reddy TE, Pauli F, Sprouse RO, et al. Genomic determination of the thioredoxin antioxidant system in the mouse uterus. Endocri- the glucocorticoid response reveals unexpected mechanisms of nology 2004;145:5485–92. gene regulation. Genome Res 2009;19:2163–71. 334. Arao Y, Carpenter K, Hewitt S, Korach KS. Estrogen down-reg- 356. Grontved L, Hager GL. Impact of chromatin structure on PR sig- ulation of the Scx gene is mediated by the opposing strand-over- naling: transition from local to global analysis. Mol Cell Endocrinol lapping gene Bop1. J Biol Chem 2010;285:4806–14. 2012;357:30–6.

4. FEMALE REPRODUCTIVE SYSTEM 1184 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

357. Stumpf WE. Nuclear concentration of 3h-estradiol in target tis- 376. Drummond AE, Baillie AJ, Findlay JK. Ovarian estrogen receptor sues. Dry-mount autoradiography of vagina, oviduct, ovary, testis, alpha and beta mRNA expression: Impact of development and mammary tumor, liver and adrenal. Endocrinology 1969;85:31–7. estrogen. Mol Cell Endocrinol 1999;149:153–61. 358. Richards JS. Estradiol receptor content in rat granulosa cells 377. Yang P, Wang J, Shen Y, Roy SK. Developmental expression of during follicular development: modification by estradiol and estrogen receptor (ER) alpha and ERbeta in the hamster ovary: gonadotropins. Endocrinology 1975;97:1174–84. regulation by follicle-stimulating hormone. Endocrinology 2004;145 359. Richards JS, Ireland JJ, Rao MC, Bernath GA, Midgley Jr AR, :5757–66. Reichert Jr LE. Ovarian follicular development in the rat: hor- 378. Sar M, Parikh I. Immunohistochemical localization of estrogen mone receptor regulation by estradiol, follicle stimulating hor- receptor in rat brain, pituitary and uterus with monoclonal anti- mone and luteinizing hormone. Endocrinology 1976;99:1562–70. bodies. J Steroid Biochem 1986;24:497–503. 360. Jakimiuk AJ, Weitsman SR, Yen HW, Bogusiewicz M, Magoffin 379. Hishikawa Y, Damavandi E, Izumi S, Koji T. Molecular histo- DA. Estrogen receptor alpha and beta expression in theca and chemical analysis of estrogen receptor alpha and beta expressions granulosa cells from women with polycystic ovary syndrome. in the mouse ovary: in situ hybridization and southwestern histo- J Clin Endocrinol Metab 2002;87:5532–8. chemistry. Med Electron Microsc 2003;36:67–73. 361. Bridges PJ, Koo Y, Kang D-W, et al. Generation of Cyp17iCre 380. Byers M, Kuiper GG, Gustafsson JA, Park-Sarge OK. Estrogen transgenic mice and their application to conditionally delete receptor-beta mRNA expression in rat ovary: down-regulation by estrogen receptor alpha (Esr1) from the ovary and testis. Genesis gonadotropins. Mol Endocrinol 1997;11:172–82. 2008;46:499–505. 381. Burkhart MN, Juengel JL, Smith PR, et al. Morphological devel- 362. Lee S, Kang DW, Hudgins-Spivey S, et al. Theca-specific estrogen opment and characterization of aromatase and estrogen recep- receptor-alpha knockout mice lose fertility prematurely. Endocri- tors alpha and beta in fetal ovaries of cattle from days 110 to 250. nology 2009;150:3855–62. Anim Reprod Sci 2010;117:43–54. 363. Binder AK, Rodriguez KF, Stockton PS, Hamilton KJ, Reed CE, 382. Van Den Broeck W, Coryn M, Simoens P, Lauwers H. Cell-specific Korach KS. The absence of ERβ results in altered gene expression distribution of oestrogen receptor-alpha in the bovine ovary. Reprod in ovarian granulosa cells from in vivo preovulatory follicles. Domest Anim 2002;37:291–3. Endocrinology 2013;154:2174–87. 383. Berisha B, Pfaffl MW, Schams D. Expression of estrogen and pro- 364. Britt KL, Drummond AE, Cox VA, et al. An age-related ovarian gesterone receptors in the bovine ovary during estrous cycle and phenotype in mice with targeted disruption of the Cyp 19 (aroma- pregnancy. Endocrine 2002;17:207–14. tase) gene. Endocrinology 2000;141:2614–23. 384. Slomczynska M, Wozniak J. Differential distribution of estrogen 365. Toda K, Hayashi Y, Ono M, Saibara T. Impact of ovarian sex ste- receptor-beta and estrogen receptor-alpha in the porcine ovary. roids on ovulation and ovulatory gene induction in aromatase- Exp Clin Endocrinol Diabetes 2001;109:238–44. null mice. Endocrinology 2012;153:386–94. 385. Tomanek M, Pisselet C, Monget P, Madigou T, Thieulant ML, 366. Sar M, Welsch F. Differential expression of estrogen receptor- Monniaux D. Estrogen receptor protein and mRNA expression in beta and estrogen receptor-alpha in the rat ovary. Endocrinology the ovary of sheep. Mol Reprod Dev 1997;48:53–62. 1999;140:963–71. 386. Hild-Petito S, Stouffer RL, Brenner RM. Immunocytochemical 367. Fitzpatrick SL, Funkhouser JM, Sindoni DM, et al. Expression localization of estradiol and progesterone receptors in the mon- of estrogen receptor-beta protein in rodent ovary. Endocrinology key ovary throughout the menstrual cycle. Endocrinology 1988; 1999;140:2581–91. 123:2896–905. 368. Hiroi H, Inoue S, Watanabe T, et al. Differential immunolocaliza- 387. Billiar RB, Loukides JA, Miller MM. Evidence for the presence of tion of estrogen receptor alpha and beta in rat ovary and uterus. J the estrogen receptor in the ovary of the baboon (papio anubis). Mol Endocrinol 1999;22:37–44. J Clin Endocrinol Metab 1992;75:1159–65. 369. Okada A, Ohta Y, Buchanan DL, et al. Changes in ontogenetic 388. Pau CY, Pau KY, Spies HG. Putative estrogen receptor beta and expression of estrogen receptor alpha and not of estrogen recep- alpha mRNA expression in male and female rhesus macaques. tor beta in the female rat reproductive tract. J Mol Endocrinol Mol Cell Endocrinol 1998;146:59–68. 2002;28:87–97. 389. Lantta M. Estradiol and progesterone receptors in normal ovary 370. Saunders PT, Maguire SM, Gaughan J, Millar MR. Expression of and ovarian tumors. Acta Obstet Gynecol Scand 1984;63:497–503. oestrogen receptor beta (ER beta) in multiple rat tissues visualised 390. Al-Timimi A, Buckley CH, Fox H. An immunohistochemical by immunohistochemistry. J Endocrinol 1997;154:R13–6. study of the incidence and significance of sex steroid hormone 371. Sharma SC, Clemens JW, Pisarska MD, Richards JS. Expres- binding sites in normal and neoplastic human ovarian tissue. Int J sion and function of estrogen receptor subtypes in granulosa Gynecol Pathol 1985;4:24–41. cells: regulation by estradiol and forskolin. Endocrinology 1999; 391. Vierikko P, Kauppila A, Vihko R. Cytosol and nuclear estrogen 140:4320–34. and progestin receptors and 17 beta-hydroxysteroid dehydroge- 372. Mowa CN, Iwanaga T. Differential distribution of oestrogen nase activity in non-diseased tissue and in benign and malignant receptor-alpha and -beta mRNAs in the female reproductive tumors of the human ovary. Int J Cancer 1983;32:413–22. organ of rats as revealed by in situ hybridization. J Endocrinol 392. Iwai T, Nanbu Y, Iwai M, Taii S, Fujii S, Mori T. Immunohisto- 2000;165:59–66. chemical localization of oestrogen receptors and progesterone 373. Schomberg DW, Couse JF, Mukherjee A, et al. Targeted disruption receptors in the human ovary throughout the menstrual cycle. of the estrogen receptor-alpha gene in female mice: characteriza- Virchows Arch A Pathol Anat Histopathol 1990;417:369–75. tion of ovarian responses and phenotype in the adult. Endocrinol- 393. Pelletier G, El-Alfy M. Immunocytochemical localization of estro- ogy 1999;140:2733–44. gen receptors alpha and beta in the human reproductive organs. 374. Yang P, Kriatchko A, Roy SK. Expression of ER-alpha and ER-beta J Clin Endocrinol Metab 2000;85:4835–40. in the hamster ovary: differential regulation by gonadotropins 394. Taylor AH, Al-Azzawi F. Immunolocalisation of oestrogen recep- and ovarian steroid hormones. Endocrinology 2002;143:2385–98. tor beta in human tissues. J Mol Endocrinol 2000;24:145–55. 375. Mowa CN, Iwanaga T. Developmental changes of the oestrogen 395. Saunders PT, Millar MR, Williams K, et al. Differential expres- receptor-alpha and -beta mRNAs in the female reproductive sion of estrogen receptor-alpha and -beta and androgen recep- organ of the rat–an analysis by in situ hybridization. J Endocrinol tor in the ovaries of marmosets and humans. Biol Reprod 2000;167:363–9. 2000;63:1098–105.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1185

396. Hillier SG, Anderson RA, Williams AR, Tetsuka M. Expression 415. Couse JF, Bunch DO, Lindzey J, Schomberg DW, Korach KS. Pre- of oestrogen receptor alpha and beta in cultured human ovarian vention of the polycystic ovarian phenotype and characterization surface epithelial cells. Mol Hum Reprod 1998;4:811–5. of ovulatory capacity in the estrogen receptor-alpha knockout 397. Hurst BS, Zilberstein M, Chou JY, Litman B, Stephens J, Leslie KK. mouse. Endocrinology 1999;140:5855–65. Estrogen receptors are present in human granulosa cells. J Clin 416. Conti M. Specificity of the cyclic adenosine 3′,5′-monophosphate Endocrinol Metab 1995;80:229–32. signal in granulosa cell function. Biol Reprod 2002;67:1653–61. 398. Chiang CH, Cheng KW, Igarashi S, Nathwani PS, Leung PC. 417. Guo C, Savage L, Sarge KD, Park-Sarge OK. Gonadotropins Hormonal regulation of estrogen receptor alpha and beta gene decrease estrogen receptor-beta messenger ribonucleic acid sta- expression in human granulosa-luteal cells in vitro. J Clin Endocri- bility in rat granulosa cells. Endocrinology 2001;142:2230–7. nol Metab 2000;85:3828–39. 418. Tonetta SA, Spicer LJ, Ireland JJ. Ci628 inhibits follicle-stimulating 399. van den Driesche S, Smith VM, Myers M, Duncan WC. Expres- hormone (FSH)-induced increases in FSH receptors of the rat sion and regulation of oestrogen receptors in the human corpus ovary: requirement of estradiol for FSH action. Endocrinology 1985; luteum. Reproduction 2008;135:509–17. 116:715–22. 400. Bao B, Kumar N, Karp RM, Garverick HA, Sundaram K. Estrogen 419. Burns KH, Yan C, Kumar TR, Matzuk MM. Analysis of ovarian receptor-beta expression in relation to the expression of luteiniz- gene expression in follicle-stimulating hormone beta knockout ing hormone receptor and cytochrome P450 enzymes in rat ovar- mice. Endocrinology 2001;142:2742–51. ian follicles. Biol Reprod 2000;63:1747–55. 420. Dierich A, Sairam MR, Monaco L, et al. Impairing follicle-stim- 401. Shughrue PJ, Lane MV, Scrimo PJ, Merchenthaler I. Comparative ulating hormone (FSH) signaling in vivo: targeted disruption of distribution of estrogen receptor-alpha (ER-alpha) and beta (ER- the FSH receptor leads to aberrant gametogenesis and hormonal beta) mRNA in the rat pituitary, gonad, and reproductive tract. imbalance. Proc Natl Acad Sci USA 1998;95:13612–7. Steroids 1998;63:498–504. 421. Danilovich N, Babu PS, Xing W, Gerdes M, Krishnamurthy H, 402. Choi I, Ko C, Park-Sarge OK, et al. Human estrogen receptor beta- Sairam MR. Estrogen deficiency, obesity, and skeletal abnormali- specific monoclonal antibodies: characterization and use in stud- ties in follicle-stimulating hormone receptor knockout (FORKO) ies of estrogen receptor beta protein expression in reproductive female mice. Endocrinology 2000;141:4295–308. tissues. Mol Cell Endocrinol 2001;181:139–50. 422. Saiduddin S, Zassenhaus HP. Effect of testosterone and proges- 403. Rosenfeld CS, Yuan X, Manikkam M, Calder MD, Garverick HA, terone on the estradiol receptor in the immature rat ovary. Endo- Lubahn DB. Cloning, sequencing, and localization of bovine estrogen crinology 1978;102:1069–76. receptor-beta within the ovarian follicle. Biol Reprod 1999;60:691–7. 423. Kouzu-Fujita M, Mezaki Y, Sawatsubashi S, et al. Coactivation of 404. Manikkam M, Bao B, Rosenfeld CS, et al. Expression of the estrogen receptor beta by gonadotropin-induced cofactor GIOT-4. bovine oestrogen receptor-beta (bERbeta) messenger ribonu- Mol Cell Biol 2009;29:83–92. cleic acid (mRNA) during the first ovarian follicular wave and 424. Walther N, Lioutas C, Tillmann G, Ivell R. Cloning of lack of change in the expression of bERbeta mRNA of second bovine estrogen receptor beta (ERbeta): expression of novel wave follicles after LH infusion into cows. Anim Reprod Sci deleted isoforms in reproductive tissues. Mol Cell Endocrinol 2001;67:159–69. 1999;152:37–45. 405. Jansen HT, West C, Lehman MN, Padmanabhan V. Ovarian estro- 425. Pepe GJ, Billiar RB, Leavitt MG, Zachos NC, Gustafsson JA, gen receptor-beta (ERbeta) regulation: I. Changes in ERbeta mes- Albrecht ED. Expression of estrogen receptors alpha and beta in senger RNA expression prior to ovulation in the ewe. Biol Reprod the baboon fetal ovary. Biol Reprod 2002;66:1054–60. 2001;65:866–72. 426. Ogawa S, Inoue S, Watanabe T, et al. Molecular cloning and char- 406. Cardenas H, Burke KA, Bigsby RM, Pope WF, Nephew KP. Estro- acterization of human estrogen receptor betacx: a potential inhibi- gen receptor beta in the sheep ovary during the estrous cycle and tor of estrogen action in human. Nucleic Acids Res 1998;26:3505–12. early pregnancy. Biol Reprod 2001;65:128–34. 427. Brandenberger AW, Tee MK, Jaffe RB. Estrogen receptor alpha 407. LaVoie HA, DeSimone DC, Gillio-Meina C, Hui YY. Cloning and (ER-alpha) and beta (ER-beta) mRNAs in normal ovary, ovarian characterization of porcine ovarian estrogen receptor beta iso- serous cystadenocarcinoma and ovarian cancer cell lines: down- forms. Biol Reprod 2002;66:616–23. regulation of ER-beta in neoplastic tissues. J Clin Endocrinol Metab 408. Enmark E, Pelto-Huikko M, Grandien K, et al. Human estrogen 1998;83:1025–8. receptor beta-gene structure, chromosomal localization, and 428. Moore JT, McKee DD, Slentz-Kesler K, et al. Cloning and char- expression pattern. J Clin Endocrinol Metab 1997;82:4258–65. acterization of human estrogen receptor beta isoforms. Biochem 409. Scobie GA, Macpherson S, Millar MR, Groome NP, Romana PG, Biophys Res Commun 1998;247:75–8. Saunders PT. Human oestrogen receptors: differential expression 429. Lu B, Leygue E, Dotzlaw H, Murphy LJ, Murphy LC, Watson PH. of ER alpha and beta and the identification of ER beta variants. Estrogen receptor-beta mRNA variants in human and murine Steroids 2002;67:985–92. tissues. Mol Cell Endocrinol 1998;138:199–203. 410. O’Brien ML, Park K, In Y, Park-Sarge OK. Characterization of 430. Rosenfeld CS, Murray AA, Simmer G, et al. Gonadotropin induc- estrogen receptor-beta (ERbeta) messenger ribonucleic acid and tion of ovulation and corpus luteum formation in young estrogen protein expression in rat granulosa cells. Endocrinology 1999;140 receptor-alpha knockout mice. Biol Reprod 2000;62:599–605. :4530–41. 431. Couse JF, Yates MM, Walker VR, Korach KS. Characterization 411. Petersen DN, Tkalcevic GT, Koza-Taylor PH, Turi TG, Brown TA. of the hypothalamic-pituitary-gonadal axis in estrogen receptor Identification of estrogen receptor beta2, a functional variant of (ER) null mice reveals hypergonadism and endocrine sex rever- estrogen receptor beta expressed in normal rat tissues. Endocrinol- sal in females lacking ERalpha but not ERbeta. Mol Endocrinol ogy 1998;139:1082–92. 2003;17:1039–53. 412. Saiduddin S, Zassenhaus HP. Estradiol-17beta receptors in the 432. Couse JF, Yates MM, Deroo BJ, Korach KS. Estrogen receptor- immature rat ovary. Steroids 1977;29:197–213. beta is critical to granulosa cell differentiation and the ovulatory 413. Kim I, Greenwald GS. Estrogen receptors in ovary and uterus response to gonadotropins. Endocrinology 2005;146:3247–62. of immature hamster and rat: effects of estrogens. Endocrinol Jpn 433. Risma KA, Clay CM, Nett TM, Wagner T, Yun J, Nilson JH. Tar- 1987;34:45–53. geted overexpression of luteinizing hormone in transgenic mice 414. Kawashima M, Greenwald GS. Comparison of follicular estrogen leads to infertility, polycystic ovaries, and ovarian tumors. Proc receptors in rat, hamster, and pig. Biol Reprod 1993;48:172–9. Natl Acad Sci USA 1995;92:1322–6.

4. FEMALE REPRODUCTIVE SYSTEM 1186 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

434. Risma KA, Hirshfield AN, Nilson JH. Elevated luteinizing hor- 452. Lim H, Paria BC, Das SK, et al. Multiple female reproductive fail- mone in prepubertal transgenic mice causes hyperandrogenemia, ures in cyclooxygenase 2-deficient mice. Cell 1997;91:197–208. precocious puberty, and substantial ovarian pathology. Endocri- 453. Behringer RR, Cate RL, Froelick GJ, Palmiter RD, Brin- nology 1997;138:3540–7. ster RL. Abnormal sexual development in transgenic mice 435. Nilson JH, Abbud RA, Keri RA, Quirk CC. Chronic hypersecre- chronically expressing Mullerian inhibiting substance. Nature tion of luteinizing hormone in transgenic mice disrupts both 1990;345:167–70. ovarian and pituitary function, with some effects modified by the 454. Deroo BJ, Rodriguez KF, Couse JF, et al. Estrogen receptor beta is genetic background. Recent Prog Horm Res 2000;55:69–89; discus- required for optimal cAMP production in mouse granulosa cells. sion 91. Mol Endocrinol 2009;23:955–65. 436. Owens GE, Keri RA, Nilson JH. Ovulatory surges of human CG 455. Rodriguez KF, Couse JF, Jayes FL, et al. Insufficient luteinizing prevent hormone-induced granulosa cell tumor formation lead- hormone-induced intracellular signaling disrupts ovulation in ing to the identification of tumor-associated changes in the tran- preovulatory follicles lacking estrogen receptor-{beta}. Endocrinol- scriptome. Mol Endocrinol 2002;16:1230–42. ogy 2010;151:2826–34. 437. Mann RJ, Keri RA, Nilson JH. Consequences of elevated luteinizing 456. Binder AK, Burns KA, Rodriguez KF, Korach KS. Transdifferen- hormone on diverse physiological systems: use of the LHbetaCTP tiation of granulosa cells to Sertoli cells in ovaries from ex3αβ estro- transgenic mouse as a model of ovarian hyperstimulation-induced gen receptor double knockout animals. Portland (OR): Society for the pathophysiology. Recent Prog Horm Res 2003;58:343–75. Study of Reproduction; 2011. 438. Magoffin DA, Weitsman SR. Differentiation of ovarian theca- 457. Swain A, Lovell-Badge R. Mammalian sex determination: a interstitial cells in vitro: regulation of 17 alpha-hydroxylase mes- molecular drama. Genes Dev 1999;13:755–67. senger ribonucleic acid expression by luteinizing hormone and 458. Dupont S, Dennefeld C, Krust A, Chambon P, Mark M. Expres- insulin-like growth factor-I. Endocrinology 1993;132:1945–51. sion of Sox9 in granulosa cells lacking the estrogen receptors, 439. Couse JF, Yates MM, Sanford R, Nyska A, Nilson JH, Korach KS. ERalpha and ERbeta. Dev Dyn 2003;226:103–6. Formation of cystic ovarian follicles associated with elevated 459. Morais da Silva S, Hacker A, Harley V, Goodfellow P, Swain A, luteinizing hormone requires estrogen receptor-beta. Endocrinol- Lovell-Badge R. Sox9 expression during gonadal development ogy 2004;145:4693–702. implies a conserved role for the gene in testis differentiation in 440. Emmen JM, Couse JF, Elmore SA, Yates MM, Kissling GE, Korach mammals and birds. Nat Genet 1996;14:62–8. KS. In vitro growth and ovulation of follicles from ovaries of estro- 460. Kanai Y, Koopman P. Structural and functional characterization gen receptor (ER){alpha} and ER{beta} null mice indicate a role for of the mouse Sox9 promoter: implications for campomelic dyspla- ER{beta} in follicular maturation. Endocrinology 2005;146:2817–26. sia. Hum Mol Genet 1999;8:691–6. 441. Richards JS. Hormonal control of gene expression in the ovary. 461. Vidal VP, Chaboissier MC, de Rooij DG, Schedl A. Sox9 Endocr Rev 1994;15:725–51. induces testis development in XX transgenic mice. Nat Genet 442. Fitzpatrick SL, Richards JS. Regulation of cytochrome P450 2001;28:216–7. aromatase messenger ribonucleic acid and activity by ste- 462. Bishop CE, Whitworth DJ, Qin Y, et al. A transgenic insertion roids and gonadotropins in rat granulosa cells. Endocrinology upstream of Sox9 is associated with dominant XX sex reversal in 1991;129:1452–62. the mouse. Nat Genet 2000;26:490–4. 443. Fitzpatrick SL, Carlone DL, Robker RL, Richards JS. Expression of 463. Taketo-Hosotani T. Factors involved in the testicular develop- aromatase in the ovary: down-regulation of mRNA by the ovula- ment from fetal mouse ovaries following transplantation. J Exp tory luteinizing hormone surge. Steroids 1997;62:197–206. Zool 1987;241:95–100. 444. Zhuang LZ, Adashi EY, Hsuch AJ. Direct enhancement of 464. Whitworth DJ, Shaw G, Renfree MB. Gonadal sex reversal of the gonadotropin-stimulated ovarian estrogen biosynthesis by estro- developing marsupial ovary in vivo and in vitro. Development gen and clomiphene citrate. Endocrinology 1982;110:2219–21. 1996;122:4057–63. 445. Adashi EY, Hsueh AJ. Estrogens augment the stimulation of ovar- 465. Vigier B, Forest MG, Eychenne B, et al. Anti-Mullerian hormone ian aromatase activity by follicle-stimulating hormone in cultured produces endocrine sex reversal of fetal ovaries. Proc Natl Acad Sci rat granulosa cells. J Biol Chem 1982;257:6077–83. USA 1989;86:3684–8. 446. Daniel SA, Armstrong DT. Involvement of estrogens in the regu- 466. Vainio S, Heikkila M, Kispert A, Chin N, McMahon AP. Female lation of granulosa cell aromatase activity. Can J Physiol Pharmacol development in mammals is regulated by Wnt-4 signalling. 1983;61:507–11. Nature 1999;397:405–9. 447. Welsh Jr TH, Jia XC, Jones PB, Zhuang LZ, Hsueh AJ. Disparate 467. Hashimoto N, Kubokawa R, Yamazaki K, Noguchi M, Kato Y. effects of triphenylethylene antiestrogens on estrogen and pro- Germ cell deficiency causes testis cord differentiation in reconsti- gestin biosyntheses by cultured rat granulosa cells. Endocrinology tuted mouse fetal ovaries. J Exp Zool 1990;253:61–70. 1984;115:1275–82. 468. Uhlenhaut NH, Jakob S, Anlag K, et al. Somatic sex repro- 448. Ghersevich S, Nokelainen P, Poutanen M, et al. Rat 17 beta- gramming of adult ovaries to testes by FOXL2 ablation. Cell hydroxysteroid dehydrogenase type 1: primary structure and 2009;139:1130–42. regulation of enzyme expression in rat ovary by diethylstilbestrol 469. Binder AK, Burns KA, Rodriguez KF, Korach KS. Aberrant expres- and gonadotropins in vivo. Endocrinology 1994;135:1477–87. sion of Dmrt1 and Foxl2 in transdifferentiated Sertoli-like cells in the 449. Ghersevich S, Poutanen M, Tapanainen J, Vihko R. Hormonal ovaries of Ex3αβERKO double knockout mice. State College (PA): regulation of rat 17 beta-hydroxysteroid dehydrogenase type Society for the Study of Reproduction; 2012. 1 in cultured rat granulosa cells: effects of recombinant follicle- 470. Whitworth DJ. XX germ cells: the difference between an ovary stimulating hormone, estrogens, androgens, and epidermal and a testis. Trends Endocrinol Metab 1998;9:2–6. growth factor. Endocrinology 1994;135:1963–71. 471. Richard-Mercier N, Dorizzi M, Desvages G, Girondot M, Pieau C. 450. Cheng G, Weihua Z, Makinen S, et al. A role for the androgen Endocrine sex reversal of gonads by the aromatase inhibitor letro- receptor in follicular atresia of estrogen receptor beta knockout zole (CGS 20267) in Emys orbicularis, a turtle with temperature- mouse ovary. Biol Reprod 2002;66:77–84. dependent sex determination. Gen Comp Endocrinol 1995;100:314–26. 451. Zalewski A, Cecchini EL, Deroo BJ. Expression of extracellular 472. Wennstrom KL, Crews D. Making males from females: the effects matrix components is disrupted in the immature and adult estro- of aromatase inhibitors on a parthenogenetic species of whiptail gen receptor beta-null mouse ovary. PLoS One 2012;7:e29937. lizard. Gen Comp Endocrinol 1995;99:316–22.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1187

473. Yomogida K, Ohtani H, Harigae H, et al. Developmental stage- 495. Nakano R, Nakayama T, Iwao M. Inhibition of ovarian follicle and spermatogenic cycle-specific expression of transcription fac- growth by a chemical antiestrogen. Horm Res 1982;16:230–6. tor GATA-1 in mouse Sertoli cells. Development 1994;120:1759–66. 496. Bley MA, Saragueta PE, Baranao JL. Concerted stimulation of 474. Blobel GA, Sieff CA, Orkin SH. Ligand-dependent repression of rat granulosa cell deoxyribonucleic acid synthesis by sex ste- the erythroid transcription factor GATA-1 by the estrogen recep- roids and follicle-stimulating hormone. J Steroid Biochem Mol Biol tor. Mol Cell Biol 1995;15:3147–53. 1997;62:11–9. 475. Flaws JA, Abbud R, Mann RJ, Nilson JH, Hirshfield AN. Chroni- 497. Spears N, Murray AA, Allison V, Boland NI, Gosden RG. Role cally elevated luteinizing hormone depletes primordial follicles of gonadotrophins and ovarian steroids in the development of in the mouse ovary. Biol Reprod 1997;57:1233–7. mouse follicles in vitro. J Reprod Fertil 1998;113:19–26. 476. Magoffin DA. The ovarian androgen-producing cells: a 2001 per- 498. Couse JF, Korach KS. Exploring the role of sex steroids through spective. Rev Endocr Metab Disord 2002;3:47–53. studies of receptor deficient mice. J Mol Med (Berl) 1998;76:497–511. 477. Britt KL, Simpson ER, Findlay JK. Effects of phytoestrogens on 499. Burns KH, Agno JE, Chen L, et al. Sexually dimorphic roles of ste- the ovarian and pituitary phenotypes of estrogen-deficient female roid hormone receptor signaling in gonadal tumorigenesis. Mol aromatase knockout mice. Menopause 2005;12:174–85. Endocrinol 2003;17:2039–52. 478. Liew SH, Sarraj MA, Drummond AE, Findlay JK. Estrogen-depen- 500. Sicinski P, Donaher JL, Geng Y, et al. Cyclin D2 is an FSH-respon- dent gene expression in the mouse ovary. PLoS One 2011;6: sive gene involved in gonadal cell proliferation and oncogenesis. e14672. Nature 1996;384:470–4. 479. Britt KL, Stanton PG, Misso M, Simpson ER, Findlay JK. The 501. Robker RL, Richards JS. Hormone-induced proliferation and effects of estrogen on the expression of genes underlying the dif- differentiation of granulosa cells: a coordinated balance of the ferentiation of somatic cells in the murine gonad. Endocrinology cell cycle regulators cyclin D2 and p27Kip1. Mol Endocrinol 2004;145:3950–60. 1998;12:924–40. 480. Britt KL, Saunders PK, McPherson SJ, Misso ML, Simpson ER, 502. Kanda N, Watanabe S. 17beta-estradiol stimulates the growth of Findlay JK. Estrogen actions on follicle formation and early fol- human keratinocytes by inducing cyclin D2 expression. J Invest licle development. Biol Reprod 2004;71:1712–23. Dermatol 2004;123:319–28. 481. Liew SH, Drummond AE, Jones ME, Findlay JK. The lack of estro- 503. Baker J, Hardy MP, Zhou J, et al. Effects of an IGF1 gene null gen and excess luteinizing hormone are responsible for the female mutation on mouse reproduction. Mol Endocrinol 1996;10:903–18. ArKO mouse phenotype. Mol Cell Endocrinol 2010;327:56–64. 504. Zhou J, Kumar TR, Matzuk MM, Bondy C. Insulin-like growth 482. Pincus G. Reproduction. Annu Rev Physiol 1962;24:57–84. factor I regulates gonadotropin responsiveness in the murine 483. Pincus G. Control of conception by hormonal steroids. Science ovary. Mol Endocrinol 1997;11:1924–33. 1966;153:493–500. 505. Kadakia R, Arraztoa JA, Bondy C, Zhou J. Granulosa cell prolif- 484. Knecht M, Tsai-Morris CH, Catt KJ. Estrogen dependence of eration is impaired in the IGF1 null ovary. Growth Horm IGF Res luteinizing hormone receptor expression in cultured rat granu- 2001;11:220–4. losa cells. Inhibition of granulosa cell development by the anties- 506. Richards JS, Sharma SC, Falender AE, Lo YH. Expression of trogens tamoxifen and keoxifene. Endocrinology 1985;116:1771–7. FKHR, FKHRL1, and AFX genes in the rodent ovary: evidence for 485. Selvaraj N, Shetty G, Vijayalakshmi K, Bhatnagar AS, Moudgal regulation by IGF-I, estrogen, and the gonadotropins. Mol Endo- NR. Effect of blocking oestrogen synthesis with a new generation crinol 2002;16:580–99. aromatase inhibitor CGS 16949A on follicular maturation induced 507. Poretsky L, Cataldo NA, Rosenwaks Z, Giudice LC. The insulin- by pregnant mare serum gonadotrophin in the immature rat. related ovarian regulatory system in health and disease. Endocr J Endocrinol 1994;142:563–70. Rev 1999;20:535–82. 486. Moudgal NR, Shetty G, Selvaraj N, Bhatnagar AS. Use of a specific 508. Bendell JJ, Dorrington J. Rat thecal/interstitial cells secrete aromatase inhibitor for determining whether there is a role for a transforming growth factor-beta-like factor that promotes oestrogen in follicle/oocyte maturation, ovulation and preimplan- growth and differentiation in rat granulosa cells. Endocrinology tation embryo development. J Reprod Fertil Suppl 1996;50:69–81. 1988;123:941–8. 487. Harris HA, Bapat AR, Gonder DS, Frail DE. The ligand binding 509. Dorrington JH, Bendell JJ, Khan SA. Interactions between FSH, profiles of estrogen receptors alpha and beta are species depen- estradiol-17 beta and transforming growth factor-beta regulate dent. Steroids 2002;67:379–84. growth and differentiation in the rat gonad. J Steroid Biochem Mol 488. Hegele-Hartung C, Siebel P, Peters O, et al. Impact of isotype- Biol 1993;44:441–7. selective estrogen receptor agonists on ovarian function. Proc Natl 510. Roberts AJ, Skinner MK. Estrogen regulation of thecal cell ste- Acad Sci USA 2004;101:5129–34. roidogenesis and differentiation: thecal cell-granulosa cell inter- 489. Pencharz RI. Effect of estrogens and androgens alone and in actions. Endocrinology 1990;127:2918–29. combination with chorionic gonadotropin on the ovary of the 511. Parrott JA, Skinner MK. Developmental and hormonal regulation hypophysectomized rat. Science 1940;91:554–5. of keratinocyte growth factor expression and action in the ovarian 490. Goldenberg RL, Vaitukaitis JL, Ross GT. Estrogen and follicle follicle. Endocrinology 1998;139:228–35. stimulation hormone interactions on follicle growth in rats. Endo- 512. Kumar TR, Wang Y, Lu N, Matzuk MM. Follicle stimulating hor- crinology 1972;90:1492–8. mone is required for ovarian follicle maturation but not male 491. Rao MC, Midgley Jr AR, Richards JS. Hormonal regulation of fertility. Nat Genet 1997;15:201–4. ovarian cellular proliferation. Cell 1978;14:71–8. 513. Danilovich N, Roy I, Sairam MR. Ovarian pathology and high 492. Wang XN, Greenwald GS. Synergistic effects of steroids with FSH incidence of sex cord tumors in follitropin receptor knockout on folliculogenesis, steroidogenesis and FSH- and hCG-receptors (FORKO) mice. Endocrinology 2001;142:3673–84. in hypophysectomized mice. J Reprod Fertil 1993;99:403–13. 514. Knecht M, Brodie AM, Catt KJ. Aromatase inhibitors prevent gran- 493. Chakravorty A, Mahesh VB, Mills TM. Regulation of follicular ulosa cell differentiation: an obligatory role for estrogens in lutein- development by diethylstilboestrol in ovaries of immature rats. izing hormone receptor expression. Endocrinology 1985;117:1156–61. J Reprod Fertil 1991;92:307–21. 515. Knecht M, Darbon JM, Ranta T, Baukal AJ, Catt KJ. Estrogens 494. Richards JS. Maturation of ovarian follicles: actions and interac- enhance the adenosine 3′,5′-monophosphate-mediated induction tions of pituitary and ovarian hormones on follicular cell differ- of follicle-stimulating hormone and luteinizing hormone recep- entiation. Physiol Rev 1980;60:51–89. tors in rat granulosa cells. Endocrinology 1984;115:41–9.

4. FEMALE REPRODUCTIVE SYSTEM 1188 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

516. Richards JS, Jonassen JA, Rolfes AI, Kersey K, Reichert Jr LE. 536. Risek B, Klier FG, Phillips A, Hahn DW, Gilula NB. Gap junction Adenosine 3′,5′-monophosphate, luteinizing hormone receptor, regulation in the uterus and ovaries of immature rats by estrogen and progesterone during granulosa cell differentiation: effects of and progesterone. J Cell Sci 1995;108:1017–32. estradiol and follicle-stimulating hormone. Endocrinology 1979; 537. Mendelson CR, Kamat A. Mechanisms in the regulation of aroma- 104:765–73. tase in developing ovary and placenta. J Steroid Biochem Mol Biol 517. Aronica SM, Kraus WL, Katzenellenbogen BS. Estrogen action via 2007;106:62–70. the cAMP signaling pathway: stimulation of adenylate cyclase 538. Turner KJ, Macpherson S, Millar MR, et al. Development and vali- and cAMP-regulated gene transcription. Proc Natl Acad Sci USA dation of a new monoclonal antibody to mammalian aromatase. J 1994;91:8517–21. Endocrinol 2002;172:21–30. 518. Coleman KM, Dutertre M, El-Gharbawy A, Rowan BG, Weigel 539. Garrett WM, Guthrie HD. Steroidogenic enzyme expression NL, Smith CL. Mechanistic differences in the activation of estro- during preovulatory follicle maturation in pigs. Biol Reprod gen receptor-alpha (ER alpha)- and ER beta-dependent gene 1997;56:1424–31. expression by cAMP signaling pathway(s). J Biol Chem 2003;278: 540. Bao B, Garverick HA. Expression of steroidogenic enzyme and 12834–45. gonadotropin receptor genes in bovine follicles during ovarian 519. Lazennec G, Thomas JA, Katzenellenbogen BS. Involvement of follicular waves: a review. J Anim Sci 1998;76:1903–21. cyclic AMP response element binding protein (CREB) and estro- 541. Kumar TR, Kelly M, Mortrud M, Low MJ, Matzuk MM. Cloning gen receptor phosphorylation in the synergistic activation of the of the mouse gonadotropin beta-subunit-encoding genes, I. Struc- estrogen receptor by estradiol and protein kinase activators. J Ste- ture of the follicle-stimulating hormone beta-subunit-encoding roid Biochem Mol Biol 2001;77:193–203. gene. Gene 1995;166:333–4. 520. Carlone DL, Richards JS. Evidence that functional interactions of 542. Parakh TN, Hernandez JA, Grammer JC, et al. Follicle-stimulating CREB and SF-1 mediate hormone regulated expression of the aro- hormone/cAMP regulation of aromatase gene expression matase gene in granulosa cells and constitutive expression in R2C requires beta-catenin. Proc Natl Acad Sci USA 2006;103:12435–40. cells. J Steroid Biochem Mol Biol 1997;61:223–31. 543. Tremblay JJ, Viger RS. GATA factors differentially activate 521. Young M, McPhaul MJ. Definition of the elements required for the multiple gonadal promoters through conserved GATA regulatory activity of the rat aromatase promoter in steroidogenic cell lines. elements. Endocrinology 2001;142:977–86. J Steroid Biochem Mol Biol 1997;61:341–8. 544. Watson J, Howson JW. Inhibition of tamoxifen of the stimula- 522. Gore-Langton RE, Daniel SA. Follicle-stimulating hormone and tory action of FSH on oestradiol-17beta synthesis by rat ovaries estradiol regulate antrum-like reorganization of granulosa cells in in vitro. J Reprod Fertil 1977;49:375–6. rat preantral follicle cultures. Biol Reprod 1990;43:65–72. 545. Tremblay A, Tremblay GB, Labrie F, Giguere V. Ligand- 523. Hu Y, Cortvrindt R, Smitz J. Effects of aromatase inhibi- independent recruitment of SRC-1 to estrogen receptor beta tion on in vitro follicle and oocyte development analyzed by through phosphorylation of activation function AF-1. Mol Cell early preantral mouse follicle culture. Mol Reprod Dev 2002; 1999;3:513–9. 61:549–59. 546. Tremblay A, Giguere V. Contribution of steroid receptor coacti- 524. Meehan TP, Narayan P. Constitutively active luteinizing hor- vator-1 and CREB binding protein in ligand-independent mone receptors: consequences of in vivo expression. Mol Cell activity of estrogen receptor beta. J Steroid Biochem Mol Biol Endocrinol 2007;260–262:294–300. 2001;77:19–27. 525. Mason AJ, Hayflick JS, Zoeller RT, et al. A deletion truncating the 547. Blobel GA, Orkin SH. Estrogen-induced apoptosis by inhibi- gonadotropin-releasing hormone gene is responsible for hypogo- tion of the erythroid transcription factor GATA-1. Mol Cell Biol nadism in the hpg mouse. Science 1986;234:1366–71. 1996;16:1687–94. 526. Guraya SS. Biology of ovarian follicles in mammals. Berlin (New York): 548. Kessel B, Liu YX, Jia XC, Hsueh AJ. Autocrine role of estrogens in Springer-Verlag; 1985. the augmentation of luteinizing hormone receptor formation in 527. Edwards RG. Follicular fluid. J Reprod Fertil Suppl 1974;37: cultured rat granulosa cells. Biol Reprod 1985;32:1038–50. 189–219. 549. Farookhi R, Desjardins J. Luteinizing hormone receptor induction 528. McConnell NA, Yunus RS, Gross SA, Bost KL, Clemens MG, in dispersed granulosa cells requires estrogen. Mol Cell Endocrinol Hughes Jr FM. Water permeability of an ovarian antral follicle 1986;47:13–24. is predominantly transcellular and mediated by aquaporins. 550. Segaloff DL, Wang HY, Richards JS. Hormonal regulation of Endocrinology 2002;143:2905–12. luteinizing hormone/chorionic gonadotropin receptor mRNA in 529. Jablonski EM, McConnell NA, Hughes Jr FM, Huet-Hudson YM. rat ovarian cells during follicular development and luteinization. Estrogen regulation of aquaporins in the mouse uterus: potential Mol Endocrinol 1990;4:1856–65. roles in uterine water movement. Biol Reprod 2003;69:1481–7. 551. Knecht M, Catt KJ. Induction of luteinizing hormone receptors by 530. Kobayashi M, Takahashi E, Miyagawa S, Watanabe H, Iguchi adenosine 3′,5′-monophosphate in cultured granulosa cells. Endo- T. Chromatin immunoprecipitation-mediated target identifica- crinology 1982;111:1192–200. tion proved aquaporin 5 is regulated directly by estrogen in the 552. Erickson GF, Wang C, Casper R, Mattson G, Hofeditz C. Studies uterus. Genes Cells 2006;11:1133–43. on the mechanisms of LH receptor control by FSH. Mol Cell Endo- 531. Hess RA. Estrogen in the adult male reproductive tract: a review. crinol 1982;27:17–30. Reprod Biol Endocrinol 2003;1:52. 553. Nimrod A. The induction of ovarian LH-receptors by FSH is 532. Simon AM, Goodenough DA, Li E, Paul DL. Female infertility in mediated by cyclic AMP. FEBS Lett 1981;131:31–3. mice lacking connexin 37. Nature 1997;385:525–9. 554. Segaloff DL, Limbird LE. Luteinizing hormone receptor appear- 533. Burghardt RC, Anderson E. Hormonal modulation of gap junc- ance in cultured porcine granulosa cells requires continual pres- tions in rat ovarian follicles. Cell Tissue Res 1981;214:181–93. ence of follicle-stimulating hormone. Proc Natl Acad Sci USA 534. Kidder GM, Mhawi AA. Gap junctions and ovarian folliculogen- 1983;80:5631–5. esis. Reproduction 2002;123:613–20. 555. Wang H, Nelson S, Ascoli M, Segaloff DL. The 5′-flanking region 535. Wright CS, Becker DL, Lin JS, Warner AE, Hardy K. Stage-specific of the rat luteinizing hormone/chorionic gonadotropin receptor and differential expression of gap junctions in the mouse ovary: gene confers Leydig cell expression and negative regulation of connexin-specific roles in follicular regulation. Reproduction 2001; gene transcription by 3′,5′-cyclic adenosine monophosphate. Mol 121:77–88. Endocrinol 1992;6:320–6.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1189

556. Chen S, Liu X, Segaloff DL. A novel cyclic adenosine 576. Richards JS, Russell DL, Ochsner S, Espey LL. Ovulation: 3′,5′- monophosphate-responsive element involved in the tran- new dimensions and new regulators of the inflammatory-like scriptional regulation of the lutropin receptor gene in granulosa response. Annu Rev Physiol 2002;64:69–92. cells. Mol Endocrinol 2000;14:1498–508. 577. Opavsky MA, Armstrong DT. Effects of luteinizing hormone on 557. Chen S, Shi H, Liu X, Segaloff DL. Multiple elements and protein superovulatory and steroidogenic responses of rat ovaries to infu- factors coordinate the basal and cyclic adenosine 3′,5′-monophos- sion with follicle-stimulating hormone. Biol Reprod 1989;41:15–25. phate-induced transcription of the lutropin receptor gene in rat 578. Armstrong DT, Siuda A, Opavsky MA, Chandrasekhar Y. Bimodal granulosa cells. Endocrinology 1999;140:2100–9. effects of luteinizing hormone and role of androgens in modifying 558. Geng Y, Tsai-Morris CH, Zhang Y, Dufau ML. The human luteiniz- superovulatory responses of rats to infusion with purified por- ing hormone receptor gene promoter: activation by Sp1 and Sp3 and cine follicle-stimulating hormone. Biol Reprod 1989;41:54–62. inhibitory regulation. Biochem Biophys Res Commun 1999;263:366–71. 579. Clemens JW, Robker RL, Kraus WL, Katzenellenbogen BS, 559. Zhang Y, Dufau ML. Dual mechanisms of regulation of transcrip- Richards JS. Hormone induction of progesterone receptor (PR) tion of luteinizing hormone receptor gene by nuclear orphan messenger ribonucleic acid and activation of PR promoter receptors and histone deacetylase complexes. J Steroid Biochem regions in ovarian granulosa cells: evidence for a role of cyclic Mol Biol 2003;85:401–14. adenosine 3′,5′-monophosphate but not estradiol. Mol Endocrinol 560. Chen S, Liu X, Segaloff DL. Identification of an SAS (Sp1c adjacent 1998;12:1201–14. site)-like element in the distal 5′-flanking region of the rat lutropin 580. Huynh K, Jones G, Thouas G, Britt KL, Simpson ER, Jones ME. receptor gene essential for cyclic adenosine 3′,5′-monophosphate Estrogen is not directly required for oocyte developmental com- responsiveness. Endocrinology 2001;142:2013–21. petence. Biol Reprod 2004;70:1263–9. 561. Magoffin DA. Regulation of differentiated functions in ovarian 581. Eddy EM, Washburn TF, Bunch DO, et al. Targeted disruption of theca cells. Sem Reprod Endocrinol 1991;9:321–31. the estrogen receptor gene in male mice causes alteration of sper- 562. Gore-Langton RE, Armstrong DT. The physiology of reproduction. matogenesis and infertility. Endocrinology 1996;137:4796–805. New York: Raven Press; 1994. 582. Nokelainen P, Puranen T, Peltoketo H, Orava M, Vihko P, Vihko R. 563. Zhang FP, Poutanen M, Wilbertz J, Huhtaniemi I. Normal prenatal Molecular cloning of mouse 17 beta-hydroxysteroid dehydroge- but arrested postnatal sexual development of luteinizing hormone nase type 1 and characterization of enzyme activity. Eur J Biochem receptor knockout (LuRKO) mice. Mol Endocrinol 2001;15:172–83. 1996;236:482–90. 564. Taniguchi F, Couse JF, Rodriguez KF, Emmen JM, Poirier D, 583. Andersson S, Moghrabi N. Physiology and molecular genetics of Korach KS. Estrogen receptor-alpha mediates an intraovarian 17 beta-hydroxysteroid dehydrogenases. Steroids 1997;62:143–7. negative feedback loop on thecal cell steroidogenesis via modu- 584. Sha J, Baker P, O’Shaughnessy PJ. Both reductive forms of 17 beta- lation of Cyp17a1 (cytochrome P450, steroid 17alpha-hydroxy- hydroxysteroid dehydrogenase (types 1 and 3) are expressed dur- lase/17,20 lyase) expression. FASEB J 2007;21:586–95. ing development in the mouse testis. Biochem Biophys Res Commun 565. Hsueh AJ, Billig H, Tsafriri A. Ovarian follicle atresia: a hormon- 1996;222:90–4. ally controlled apoptotic process. Endocr Rev 1994;15:707–24. 585. Baker PJ, Sha JH, O’Shaughnessy PJ. Localisation and regula- 566. Erickson GF, Magoffin DA, Dyer CA, Hofeditz C. The ovarian tion of 17beta-hydroxysteroid dehydrogenase type 3 mRNA androgen producing cells: a review of structure/function rela- during development in the mouse testis. Mol Cell Endocrinol tionships. Endocr Rev 1985;6:371–99. 1997;133:127–33. 567. Samuels LT, Uchikawa T, Zain-ul-Abedin M, Huseby RA. Effect 586. Mustonen MV, Poutanen MH, Isomaa VV, Vihko PT, Vihko RK. of diethylstilbestrol on enzymes of cryptochid mouse testes of Cloning of mouse 17beta-hydroxysteroid dehydrogenase type BALB-c mice. Endocrinology 1969;85:96–102. 2, and analysing expression of the mRNAs for types 1, 2, 3, 4 568. Onoda M, Hall PF. Inhibition of testicular microsomal cyto- and 5 in mouse embryos and adult tissues. Biochem J 1997;325: chrome P-450 (17 alpha-hydroxylase/C-17,20-lyase) by estrogens. 199–205. Endocrinology 1981;109:763–7. 587. Tsai-Morris CH, Khanum A, Tang PZ, Dufau ML. The rat 17beta- 569. Samuels LT, Bussmann L, Matsumoto K, Huseby RA. Organi- hydroxysteroid dehydrogenase type III: molecular cloning and zation of androgen biosynthesis in the testis. J Steroid Biochem gonadotropin regulation. Endocrinology 1999;140:3534–42. 1975;6:291–6. 588. Couse JF, Yates MM, Rodriguez KF, Johnson JA, Poirier D, 570. Magoffin DA, Erickson GF. Mechanism by which 17 beta- estradiol Korach KS. The intraovarian actions of estrogen receptor-alpha inhibits ovarian androgen production in the rat. Endocrinology are necessary to repress the formation of morphological and 1981;108:962–9. functional Leydig-like cells in the female gonad. Endocrinology 571. Magoffin DA, Erickson GF. Direct inhibitory effect of estrogen on 2006;147:3666–78. LH-stimulated androgen synthesis by ovarian cells cultured in 589. Akingbemi BT, Ge R, Rosenfeld CS, et al. Estrogen receptor-alpha defined medium. Mol Cell Endocrinol 1982;28:81–9. gene deficiency enhances androgen biosynthesis in the mouse 572. Johnson DC, Martin H, Tsai-Morris CH. The in vitro and in vivo Leydig cell. Endocrinology 2003;144:84–93. effect of estradiol upon the 17 alpha-hydroxylase and C17,20- 590. Abney TO. The potential roles of estrogens in regulating Leydig lyase activity in the ovaries of immature hypophysectomized rats. cell development and function: a review. Steroids 1999;64:610–7. Mol Cell Endocrinol 1984;35:199–204. 591. Conte FA, Grumbach MM, Ito Y, Fisher CR, Simpson ER. A syn- 573. Banks PK, Meyer K, Brodie AM. Regulation of ovarian steroid drome of female pseudohermaphrodism, hypergonadotropic biosynthesis by estrogen during proestrus in the rat. Endocrinol- hypogonadism, and multicystic ovaries associated with missense ogy 1991;129:1295–304. mutations in the gene encoding aromatase (P450arom). J Clin 574. Sakaue M, Ishimura R, Kurosawa S, et al. Administration of estra- Endocrinol Metab 1994;78:1287–92. diol-3-benzoate down-regulates the expression of testicular ste- 592. Shozu M, Akasofu K, Harada T, Kubota Y. A new cause of female roidogenic enzyme genes for testosterone production in the adult pseudohermaphroditism: placental aromatase deficiency. J Clin rat. J Vet Med Sci 2002;64:107–13. Endocrinol Metab 1991;72:560–6. 575. Govoroun M, McMeel OM, Mecherouki H, Smith TJ, Guiguen Y. 593. Morishima A, Grumbach MM, Simpson ER, Fisher C, Qin K. Aro- 17beta-estradiol treatment decreases steroidogenic enzyme mes- matase deficiency in male and female siblings caused by a novel senger ribonucleic acid levels in the rainbow trout testis. Endocri- mutation and the physiological role of estrogens. J Clin Endocrinol nology 2001;142:1841–8. Metab 1995;80:3689–98.

4. FEMALE REPRODUCTIVE SYSTEM 1190 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

594. Mullis PE, Yoshimura N, Kuhlmann B, Lippuner K, Jaeger P, 614. Kok HS, Onland-Moret NC, van Asselt KM, et al. No associa- Harada H. Aromatase deficiency in a female who is compound tion of estrogen receptor alpha and cytochrome P450c17alpha heterozygote for two new point mutations in the P450arom polymorphisms with age at menopause in a Dutch cohort. Hum gene: impact of estrogens on hypergonadotropic hypogonadism, Reprod 2005;20:536–42. multicystic ovaries, and bone densitometry in childhood. J Clin 615. Rosenkranz K, Hinney A, Ziegler A, et al. Systematic mutation Endocrinol Metab 1997;82:1739–45. screening of the estrogen receptor beta gene in probands of dif- 595. Ludwig M, Beck A, Wickert L, et al. Female pseudohermaph- ferent weight extremes: identification of several genetic variants. roditism associated with a novel homozygous G-to-A (V370-to- J Clin Endocrinol Metab 1998;83:4524–7. M) substitution in the P-450 aromatase gene. J Pediatr Endocrinol 616. Sundarrajan C, Liao WX, Roy AC, Ng SC. Association between Metab 1998;11:657–64. estrogen receptor-beta gene polymorphisms and ovulatory dys- 596. Grumbach MM, Auchus RJ. Estrogen: consequences and implica- functions in patients with menstrual disorders. J Clin Endocrinol tions of human mutations in synthesis and action. J Clin Endocri- Metab 2001;86:135–9. nol Metab 1999;84:4677–94. 617. Schreiber JR, Erickson GF. Progesterone receptor in the rat ovary: 597. Simpson ER. Models of aromatase insufficiency. Semin Reprod Med further characterization and localization in the granulosa cell. 2004;22:25–30. Steroids 1979;34:459–69. 598. Ongphiphadhanakul B. Genetic polymorphisms of estrogen 618. Schreiber JR, Hsueh JW. Progesterone “receptor” in rat ovary. receptor-alpha: possible implications for targeted osteoporosis Endocrinology 1979;105:915–9. therapy. Am J Pharmacogenomics 2003;3:5–9. 619. Jacobs BR, Suchocki S, Smith RG. Evidence for a human ovarian 599. Andersen TI, Heimdal KR, Skrede M, Tveit K, Berg K, progesterone receptor. Am J Obstet Gynecol 1980;138:332–6. Borresen AL. Oestrogen receptor (ESR) polymorphisms and 620. Jacobs BR, Smith RG. A comparison of progesterone and R5020 breast cancer susceptibility. Hum Genet 1994;94:665–70. binding in endometrium, ovary, pituitary, and hypothalamus. 600. Herrington DM, Howard TD. ER-alpha variants and the cardiovas- Fertil Steril 1981;35:438–41. cular effects of hormone replacement therapy. Pharmacogenomics 621. Shao R, Markstrom E, Friberg PA, Johansson M, Billig H. Expres- 2003;4:269–77. sion of progesterone receptor (PR) A and B isoforms in mouse 601. Herrington DM. Role of estrogen receptor-alpha in pharmacoge- granulosa cells: stage-dependent PR-mediated regulation of netics of estrogen action. Curr Opin Lipidol 2003;14:145–50. apoptosis and cell proliferation. Biol Reprod 2003;68:914–21. 602. Liu YZ, Liu YJ, Recker RR, Deng HW. Molecular studies of iden- 622. Gava N, Clarke CL, Byth K, Arnett-Mansfield RL, deFazio A. tification of genes for osteoporosis: the 2002 update. J Endocrinol Expression of progesterone receptors A and B in the mouse ovary 2003;177:147–96. during the estrous cycle. Endocrinology 2004;145:3487–94. 603. Tempfer CB, Schneeberger C, Huber JC. Applications of poly- 623. Slomczynska M, Krok M, Pierscinski A. Localization of the pro- morphisms and pharmacogenomics in obstetrics and gynecology. gesterone receptor in the porcine ovary. Acta Histochem 2000; Pharmacogenomics 2004;5:57–65. 102:183–91. 604. Syrrou M, Georgiou I, Patsalis PC, Bouba I, Adonakis G, Pagoula- 624. Chandrasekher YA, Melner MH, Nagalla SR, Stouffer RL. Proges- tos GN. Fragile X premutations and (TA)n estrogen receptor poly- terone receptor, but not estradiol receptor, messenger ribonucleic morphism in women with ovarian dysfunction. Am J Med Genet acid is expressed in luteinizing granulosa cells and the corpus 1999;84:306–8. luteum in rhesus monkeys. Endocrinology 1994;135:307–14. 605. Westberg L, Baghaei F, Rosmond R, et al. Polymorphisms of the 625. Suzuki T, Sasano H, Kimura N, et al. Immunohistochemical androgen receptor gene and the estrogen receptor beta gene are distribution of progesterone, androgen and oestrogen receptors associated with androgen levels in women. J Clin Endocrinol Metab in the human ovary during the menstrual cycle: relationship to 2001;86:2562–8. expression of steroidogenic enzymes. Hum Reprod 1994;9:1589–95. 606. Wang Y, Miksicek RJ. Characterization of estrogen receptor 626. Revelli A, Pacchioni D, Cassoni P, Bussolati G, Massobrio M. In cDNAs from human uterus: Identification of a novel PvuII poly- situ hybridization study of messenger RNA for estrogen recep- morphism. Mol Cell Endocrinol 1994;101:101–10. tor and immunohistochemical detection of estrogen and proges- 607. Yaich L, Dupont WD, Cavener DR, Parl FF. Analysis of the PvuII terone receptors in the human ovary. Gynecol Endocrinol 1996; restriction fragment-length polymorphism and exon structure 10:177–86. of the estrogen receptor gene in breast cancer and peripheral 627. Iwai T, Fujii S, Nanbu Y, et al. Effect of human chorionic gonado- blood. Cancer Res 1992;52:77–83. tropin on the expression of progesterone receptors and estrogen 608. Georgiou I, Konstantelli M, Syrrou M, Messinis IE, Lolis DE. Oes- receptors in rabbit ovarian granulosa cells and the uterus. Endocri- trogen receptor gene polymorphisms and ovarian stimulation for nology 1991;129:1840–8. in-vitro fertilization. Hum Reprod 1997;12:1430–3. 628. Park OK, Mayo KE. Transient expression of progesterone receptor 609. Weel AE, Uitterlinden AG, Westendorp IC, et al. Estrogen recep- messenger RNA in ovarian granulosa cells after the preovulatory tor polymorphism predicts the onset of natural and surgical luteinizing hormone surge. Mol Endocrinol 1991;5:967–78. menopause. J Clin Endocrinol Metab 1999;84:3146–50. 629. Iwai M, Yasuda K, Fukuoka M, et al. Luteinizing hormone induces 610. Zofkova I, Zajickova K, Hill M. The estrogen receptor alpha gene progesterone receptor gene expression in cultured porcine granu- determines serum androstenedione levels in postmenopausal losa cells. Endocrinology 1991;129:1621–7. women. Steroids 2002;67:815–9. 630. Boerboom D, Russell DL, Richards JS, Sirois J. Regulation of tran- 611. Gorai I, Tanaka K, Inada M, et al. Estrogen-metabolizing gene poly- scripts encoding ADAMTS-1 (a disintegrin and metalloproteinase morphisms, but not estrogen receptor-alpha gene polymorphisms, with thrombospondin-like motifs-1) and progesterone receptor are associated with the onset of menarche in healthy postmeno- by human chorionic gonadotropin in equine preovulatory fol- pausal Japanese women. J Clin Endocrinol Metab 2003;88:799–803. licles. J Mol Endocrinol 2003;31:473–85. 612. Sundarrajan C, Liao W, Roy AC, Ng SC. Association of oestrogen 631. Conneely OM, Lydon JP. Progesterone receptors in reproduc- receptor gene polymorphisms with outcome of ovarian stimula- tion: functional impact of the A and B isoforms. Steroids 2000; tion in patients undergoing IVF. Mol Hum Reprod 1999;5:797–802. 65:571–7. 613. Stavrou I, Zois C, Ioannidis JP, Tsatsoulis A. Association of poly- 632. Arakawa S, Iyo M, Ohkawa R, Kambegawa A, Okinaga S, Arai K. morphisms of the oestrogen receptor alpha gene with the age of Steroid hormone receptors in the uterus and ovary of immature menarche. Hum Reprod 2002;17:1101–5. rats treated with gonadotropins. Endocrinol Jpn 1989;36:219–28.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1191

633. Robker RL, Russell DL, Espey LL, Lydon JP, O’Malley BW, 652. Sriraman V, Richards JS. Cathepsin L gene expression and pro- Richards JS. Progesterone-regulated genes in the ovulation pro- moter activation in rodent granulosa cells. Endocrinology 2004; cess: ADAMTS-1 and cathepsin L proteases. Proc Natl Acad Sci USA 145:582–91. 2000;97:4689–94. 653. Doyle KM, Russell DL, Sriraman V, Richards JS. Coordinate tran- 634. Park-Sarge OK, Mayo KE. Regulation of the progesterone receptor scription of the ADAMTS-1 gene by luteinizing hormone and pro- gene by gonadotropins and cyclic adenosine 3′,5′-monophosphate gesterone receptor. Mol Endocrinol 2004;18:2463–78. in rat granulosa cells. Endocrinology 1994;134:709–18. 654. Brown HM, Dunning KR, Robker RL, et al. ADAMTS1 cleavage of 635. Natraj U, Richards JS. Hormonal regulation, localization, and versican mediates essential structural remodeling of the ovarian functional activity of the progesterone receptor in granulosa cells follicle and cumulus-oocyte matrix during ovulation in mice. Biol of rat preovulatory follicles. Endocrinology 1993;133:761–9. Reprod 2010;83:549–57. 636. Sriraman V, Sharma SC, Richards JS. Transactivation of the pro- 655. Mittaz L, Russell DL, Wilson T, et al. Adamts-1 is essential for the gesterone receptor gene in granulosa cells: evidence that Sp1/Sp3 development and function of the urogenital system. Biol Reprod binding sites in the proximal promoter play a key role in lutein- 2004;70:1096–105. izing hormone inducibility. Mol Endocrinol 2003;17:436–49. 656. Shimada M, Yanai Y, Okazaki T, et al. Synaptosomal-associated 637. Fan HY, Liu Z, Shimada M, et al. MAPK3/1 (ERK1/2) in ovar- protein 25 gene expression is hormonally regulated during ovu- ian granulosa cells are essential for female fertility. Science lation and is involved in cytokine/chemokine exocytosis from 2009;324:938–41. granulosa cells. Mol Endocrinol 2007;21:2487–502. 638. Duda M, Durlej-Grzesiak M, Tabarowski Z, Slomczynska M. 657. Palanisamy GS, Cheon YP, Kim J, et al. A novel pathway involv- Effects of testosterone and 2-hydroxyflutamide on progesterone ing progesterone receptor, endothelin-2, and endothelin receptor receptor expression in porcine ovarian follicles in vitro. Reprod B controls ovulation in mice. Mol Endocrinol 2006;20:2784–95. Biol 2012;12:333–40. 658. Ko C, Gieske MC, Al-Alem L, et al. Endothelin-2 in ovarian fol- 639. Sriraman V, Sinha M, Richards JS. Progesterone receptor-induced licle rupture. Endocrinology 2006;147:1770–9. gene expression in primary mouse granulosa cell cultures. Biol 659. Bridges PJ, Jo M, Al Alem L, et al. Production and binding of Reprod 2010;82:402–12. endothelin-2 (EDN2) in the rat ovary: endothelin receptor sub- 640. Smith BD, Bradbury JT. Influence of progestins on ovarian type A (EDNRA)-mediated contraction. Reprod Fertil Dev 2010; responses to estrogen and gonadotrophins in immature rats. 22:780–7. Endocrinology 1966;78:297–301. 660. Alam H, Maizels ET, Park Y, et al. Follicle-stimulating hormone 641. Chappell PE, Lydon JP, Conneely OM, O’Malley BW, Levine JE. activation of hypoxia-inducible factor-1 by the phosphatidylino- Endocrine defects in mice carrying a null mutation for the proges- sitol 3-kinase/AKT/Ras homolog enriched in brain (Rheb)/ terone receptor gene. Endocrinology 1997;138:4147–52. mammalian target of rapamycin (mTOR) pathway is necessary 642. Mori T, Suzuki A, Nishimura T, Kambegawa A. Inhibition of ovu- for induction of select protein markers of follicular differentiation. lation in immature rats by anti-progesterone antiserum. J Endocri- J Biol Chem 2004;279:19431–40. nol 1977;73:185–6. 661. Alam H, Weck J, Maizels E, et al. Role of the phosphatidylinosi- 643. Brannstrom M, Janson PO. Progesterone is a mediator in the tol-3-kinase and extracellular regulated kinase pathways in the ovulatory process of the in vitro-perfused rat ovary. Biol Reprod induction of hypoxia-inducible factor (HIF)-1 activity and the 1989;40:1170–8. HIF-1 target vascular endothelial growth factor in ovarian granu- 644. Espey LL, Yoshioka S, Russell DL, Robker RL, Fujii S, Richards JS. losa cells in response to follicle-stimulating hormone. Endocrinol- Ovarian expression of a disintegrin and metalloproteinase with ogy 2009;150:915–28. thrombospondin motifs during ovulation in the gonadotropin- 662. Kim J, Bagchi IC, Bagchi MK. Signaling by hypoxia-induc- primed immature rat. Biol Reprod 2000;62:1090–5. ible factors is critical for ovulation in mice. Endocrinology 645. Tanaka N, Espey LL, Kawano T, Okamura H. Comparison of 2009;150:3392–400. inhibitory actions of indomethacin and epostane on ovulation in 663. Kim J, Bagchi IC, Bagchi MK. Control of ovulation in mice by rats. Am J Physiol 1991;260:E170–4. progesterone receptor-regulated gene networks. Mol Hum Reprod 646. Brannstrom M. Inhibitory effect of mifepristone (RU 486) on ovula- 2009;15:821–8. tion in the isolated perfused rat ovary. Contraception 1993;48:393–402. 664. Kim J, Sato M, Li Q, et al. Peroxisome proliferator-activated 647. Loutradis D, Bletsa R, Aravantinos L, Kallianidis K, Michalas S, receptor gamma is a target of progesterone regulation in the pre- Psychoyos A. Preovulatory effects of the progesterone antagonist ovulatory follicles and controls ovulation in mice. Mol Cell Biol mifepristone (RU486) in mice. Hum Reprod 1991;6:1238–40. 2008;28:1770–82. 648. Espey LL, Adams RF, Tanaka N, Okamura H. Effects of epostane on 665. Tetsuka M, Hillier SG. Androgen receptor gene expression in rat ovarian levels of progesterone, 17 beta-estradiol, prostaglandin E2, granulosa cells: the role of follicle-stimulating hormone and ste- and prostaglandin F2 alpha during ovulation in the gonadotropin- roid hormones. Endocrinology 1996;137:4392–7. primed immature rat. Endocrinology 1990;127:259–63. 666. Slomczynska M, Szoltys M. Immunohistochemical localization 649. Sriraman V, Eichenlaub-Ritter U, Bartsch JW, Rittger A, Mulders of androgen receptor (AR) in rat ovary. Folia Histochem Cytobiol SM, Richards JS. Regulated expression of ADAM8 (a disintegrin and 1997;35:101–2. metalloprotease domain 8) in the mouse ovary: evidence for a regula- 667. Szoltys M, Slomczynska M. Changes in distribution of androgen tory role of luteinizing hormone, progesterone receptor, and epider- receptor during maturation of rat ovarian follicles. Exp Clin Endo- mal growth factor-like growth factors. Biol Reprod 2008;78:1038–48. crinol Diabetes 2000;108:228–34. 650. Shimada M, Nishibori M, Yamashita Y, Ito J, Mori T, Richards JS. 668. Tetsuka M, Whitelaw PF, Bremner WJ, Millar MR, Smyth CD, Down-regulated expression of A disintegrin and metalloprotein- Hillier SG. Developmental regulation of androgen receptor in rat ase with thrombospondin-like repeats-1 by progesterone recep- ovary. J Endocrinol 1995;145:535–43. tor antagonist is associated with impaired expansion of porcine 669. Cardenas H, Pope WF. Androgen receptor and follicle-stimulating cumulus-oocyte complexes. Endocrinology 2004;145:4603–14. hormone receptor in the pig ovary during the follicular phase of 651. Young KA, Tumlinson B, Stouffer RL. ADAMTS-1/METH-1 and the estrous cycle. Mol Reprod Dev 2002;62:92–8. TIMP-3 expression in the primate corpus luteum: divergent pat- 670. Garrett WM, Guthrie HD. Expression of androgen receptors and terns and stage-dependent regulation during the natural men- steroidogenic enzymes in relation to follicular growth and atresia strual cycle. Mol Hum Reprod 2004;10:559–65. following ovulation in pigs. Biol Reprod 1996;55:949–55.

4. FEMALE REPRODUCTIVE SYSTEM 1192 25. STEROID RECEPTORS IN THE UTERUS AND OVARY

671. Campo SM, Carson RS, Findlay JK. Distribution of specific andro- 691. Walters KA, Middleton LJ, Joseph SR, et al. Targeted loss of gen binding sites within the ovine ovarian follicle. Mol Cell Endo- androgen receptor signaling in murine granulosa cells of pre- crinol 1985;39:255–65. antral and antral follicles causes female subfertility. Biol Reprod 672. Hampton JH, Manikkam M, Lubahn DB, Smith MF, Garverick 2012;87:151. HA. Androgen receptor mRNA expression in the bovine ovary. 692. Sen A, Hammes SR. Granulosa cell-specific androgen receptors Domest Anim Endocrinol 2004;27:81–8. are critical regulators of ovarian development and function. Mol 673. Duffy DM, Abdelgadir SE, Stott KR, Resko JA, Stouffer RL, Endocrinol 2010;24:1393–403. Zelinski-Wooten MB. Androgen receptor mRNA expression in the 693. Mowszowicz I, Lee HJ, Chen HT, et al. A point mutation in the rhesus monkey ovary. Endocrine 1999;11:23–30. second zinc finger of the DNA-binding domain of the androgen 674. Hillier SG, Tetsuka M, Fraser HM. Location and developmental receptor gene causes complete androgen insensitivity in two sib- regulation of androgen receptor in primate ovary. Hum Reprod lings with receptor-positive androgen resistance. Mol Endocrinol 1997;12:107–11. 1993;7:861–9. 675. Weil SJ, Vendola K, Zhou J, et al. Androgen receptor gene expres- 694. De Leo V, Lanzetta D, D’Antona D, la Marca A, Morgante G. Hor- sion in the primate ovary: cellular localization, regulation, and monal effects of flutamide in young women with polycystic ovary functional correlations. J Clin Endocrinol Metab 1998;83:2479–85. syndrome. J Clin Endocrinol Metab 1998;83:99–102. 676. Milwidsky A, Younes MA, Besch NF, Besch PK, Kaufman RH. 695. Quigley CA, Evans BA, Simental JA, et al. Complete androgen Receptor-like binding proteins for testosterone and progesterone insensitivity due to deletion of exon C of the androgen receptor in the human ovary. Am J Obstet Gynecol 1980;138:93–8. gene highlights the functional importance of the second zinc finger 677. Chadha S, Pache TD, Huikeshoven JM, Brinkmann AO, van der of the androgen receptor in vivo. Mol Endocrinol 1992;6:1103–12. Kwast TH. Androgen receptor expression in human ovarian and 696. Hernandez Gifford JA, Hunzicker-Dunn ME, Nilson JH. Con- uterine tissue of long-term androgen-treated transsexual women. ditional deletion of beta-catenin mediated by Amhr2cre in mice Hum Pathol 1994;25:1198–204. causes female infertility. Biol Reprod 2009;80:1282–92. 678. Horie K, Takakura K, Fujiwara H, Suginami H, Liao S, Mori 697. Gill A, Jamnongjit M, Hammes SR. Androgens promote matura- T. Immunohistochemical localization of androgen receptor in tion and signaling in mouse oocytes independent of transcription: the human ovary throughout the menstrual cycle in relation to a release of inhibition model for mammalian oocyte meiosis. Mol oestrogen and progesterone receptor expression. Hum Reprod Endocrinol 2004;18:97–104. 1992;7:184–90. 698. Jamnongjit M, Gill A, Hammes SR. Epidermal growth factor 679. Takayama K, Fukaya T, Sasano H, et al. Immunohistochemical receptor signaling is required for normal ovarian steroidogenesis study of steroidogenesis and cell proliferation in polycystic ovar- and oocyte maturation. Proc Natl Acad Sci USA 2005;102:16257–62. ian syndrome. Hum Reprod 1996;11:1387–92. 699. Li M, Ai JS, Xu BZ, et al. Testosterone potentially triggers meiotic 680. Kaipia A, Hsueh AJ. Regulation of ovarian follicle atresia. Annu resumption by activation of intra-oocyte SRC and MAPK in por- Rev Physiol 1997;59:349–63. cine oocytes. Biol Reprod 2008;79:897–905. 681. Galas J, Słomczyńska M, Knapczyk-Stwora K, et al. Steroid levels 700. Li M, Schatten H, Sun QY. Androgen receptor’s destiny in and the spatiotemporal expression of steroidogenic enzymes and mammalian oocytes: a new hypothesis. Mol Hum Reprod 2009; androgen receptor in developing ovaries of immature rats. Acta 15:149–54. Histochem 2012;114:207–16. 701. Hellbaum AA, Owens Jr JN, Payne RW. The effect of androgens 682. Hirshfield AN. Patterns of [3H] thymidine incorporation dif- on the ovaries and uterus of the estrogen treated hypophysecto- fer in immature rats and mature, cycling rats. Biol Reprod 1986; mized immature rat. Endocrinology 1956;59:306–16. 34:229–35. 702. Payne RW, Runser RH. The influence of estrogen and androgen 683. Tetsuka M, Hillier SG. Differential regulation of aromatase and on the ovarian response of hypophysectomized immature rats to androgen receptor in granulosa cells. J Steroid Biochem Mol Biol gonadotropins. Endocrinology 1958;62:313–21. 1997;61:233–9. 703. Conway BA, Mahesh VB, Mills TM. Effect of dihydrotestosterone 684. Weil S, Vendola K, Zhou J, Bondy CA. Androgen and follicle- on the growth and function of ovarian follicles in intact immature stimulating hormone interactions in primate ovarian follicle female rats primed with PMSG. J Reprod Fertil 1990;90:267–77. development. J Clin Endocrinol Metab 1999;84:2951–6. 704. Hillier SG, Ross GT. Effects of exogenous testosterone on ovar- 685. Jeppesen JV, Kristensen SG, Nielsen ME, et al. LH-receptor gene ian weight, follicular morphology and intraovarian progesterone expression in human granulosa and cumulus cells from antral and concentration in estrogen-primed hypophysectomized immature preovulatory follicles. J Clin Endocrinol Metab 2012;97:E1524–31. female rats. Biol Reprod 1979;20:261–8. 686. Campo S, Carson RS, Findlay JK. Acute effect of PMSG on ovarian 705. Louvet JP, Harman SM, Schrieber JR, Ross GT. Evidence of a androgen-binding sites in the intact immature female rat. Reprod role of androgens in follicular maturation. Endocrinology 1975; Fertil Dev 1992;4:55–65. 97:366–72. 687. Li M, Xue K, Ling J, Diao FY, Cui YG, Liu JY. The orphan nuclear 706. Pradeep PK, Li X, Peegel H, Menon KM. Dihydrotestosterone receptor NR4A1 regulates transcription of key steroidogenic inhibits granulosa cell proliferation by decreasing the cyclin D2 enzymes in ovarian theca cells. Mol Cell Endocrinol 2010;319:39–46. mRNA expression and cell cycle arrest at G1 phase. Endocrinology 688. Dai A, Yan G, He Q, et al. Orphan nuclear receptor Nur77 regu- 2002;143:2930–5. lates androgen receptor gene expression in mouse ovary. PLoS 707. Kayampilly PP, Menon KMJ. Inhibition of extracellular signal- One 2012;7:e39950. regulated protein kinase-2 phosphorylation by dihydrotestos- 689. Yeh S, Hu YC, Wang PH, et al. Abnormal mammary gland terone reduces follicle-stimulating hormone-mediated cyclin development and growth retardation in female mice and D2 messenger ribonucleic acid expression in rat granulosa cells. MCF7 breast cancer cells lacking androgen receptor. J Exp Med Endocrinology 2004;145:1786–93. 2003;198:1899–908. 708. Kayampilly PP, Menon KMJ. AMPK activation by dihydrotes- 690. Shiina H, Matsumoto T, Sato T, et al. Premature ovarian fail- tosterone reduces FSH-stimulated cell proliferation in rat gran- ure in androgen receptor-deficient mice. Proc Natl Acad Sci USA ulosa cells by inhibiting ERK signaling pathway. Endocrinology 2006;103:224–9. 2012;153:2831–8.

4. FEMALE REPRODUCTIVE SYSTEM REFERENCES 1193

709. Lenie S, Smitz J. Functional AR signaling is evident in an in vitro 725. Zeleznik AJ, Hillier SG, Ross GT. Follicle stimulating hormone- mouse follicle culture bioassay that encompasses most stages of induced follicular development: an examination of the role of folliculogenesis. Biol Reprod 2009;80:685–95. androgens. Biol Reprod 1979;21:673–81. 710. Hickey TE, Marrocco DL, Gilchrist RB, Norman RJ, Armstrong DT. 726. Simone DA, Chorich LP, Mahesh VB. Mechanisms of action Interactions between androgen and growth factors in granulosa for an androgen-mediated autoregulatory process in rat thecal- cell subtypes of porcine antral follicles. Biol Reprod 2004;71:45–52. interstitial cells. Biol Reprod 1993;49:1190–201. 711. Vendola KA, Zhou J, Adesanya OO, Weil SJ, Bondy CA. Andro- 727. Young JM, McNeilly AS. Theca: the forgotten cell of the ovarian gens stimulate early stages of follicular growth in the primate follicle. Reproduction 2010;140:489–504. ovary. J Clin Invest 1998;101:2622–9. 728. Mori T, Suzuki A, Nishimura T, Kambegawa A. Evidence for 712. Yang MY, Fortune JE. Testosterone stimulates the primary to sec- androgen participation in induced ovulation in immature rats. ondary follicle transition in bovine follicles in vitro. Biol Reprod Endocrinology 1977;101:623–6. 2006;75:924–32. 729. Peluso JJ, Stude D, Steger RW. Role of androgens in hCG-induced 713. Daniel SA, Armstrong DT. Enhancement of follicle-stimulating ovulation in PMSG-primed immature rats. Acta Endocrinol hormone-induced aromatase activity by androgens in cultured (Copenh) 1980;93:505–12. rat granulosa cells. Endocrinology 1980;107:1027–33. 730. Ware VC. The role of androgens in follicular development in the 714. Katz Y, Leung PC, Armstrong DT. Testosterone restores ovarian ovary. I. A quantitative analysis of oocyte ovulation. J Exp Zool 1982; aromatase activity in rats treated with a 17,20-lyase inhibitor. Mol 222:155–67. Cell Endocrinol 1979;14:37–44. 731. Markstrom E, Svensson E, Shao R, Svanberg B, Billig H. Survival 715. Wu YG, Bennett J, Talla D, Stocco C. Testosterone, not 5alpha- factors regulating ovarian apoptosis – dependence on follicle dihydrotestosterone, stimulates LRH-1 leading to FSH-indepen- differentiation. Reproduction 2002;123:23–30. dent expression of Cyp19 and P450scc in granulosa cells. Mol 732. Azzolin GC, Saiduddin S. Effect of androgens on the ovarian Endocrinol 2011;25:656–68. morphology of the hypophysectomized rat. Proc Soc Exp Biol Med 716. Murray AA, Gosden RG, Allison V, Spears N. Effect of androgens 1983;172:70–3. on the development of mouse follicles growing in vitro. J Reprod 733. Billig H, Furuta I, Hsueh AJ. Estrogens inhibit and androgens Fertil 1998;113:27–33. enhance ovarian granulosa cell apoptosis. Endocrinology 1993; 717. Harlow CR, Hillier SG, Hodges JK. Androgen modulation of 133:2204–12. follicle-stimulating hormone-induced granulosa cell steroidogen- 734. Gupta RK, Aberdeen G, Babus JK, Albrecht ED, Flaws JA. esis in the primate ovary. Endocrinology 1986;119:1403–5. Methoxychlor and its metabolites inhibit growth and induce atre- 718. Harlow CR, Shaw HJ, Hillier SG, Hodges JK. Factors influenc- sia of baboon antral follicles. Toxicol Pathol 2007;35:649–56. ing follicle-stimulating hormone-responsive steroidogenesis in 735. Tetsuka M, Milne M, Simpson GE, Hillier SG. Expression of marmoset granulosa cells: effects of androgens and the stage of 11beta-hydroxysteroid dehydrogenase, glucocorticoid receptor, follicular maturity. Endocrinology 1988;122:2780–7. and mineralocorticoid receptor genes in rat ovary. Biol Reprod 719. Daniel SA, Armstrong DT. Site of action of androgens on follicle- 1999;60:330–5. stimulating hormone-induced aromatase activity in cultured rat 736. Michael AE, Pester LA, Curtis P, Shaw RW, Edwards CR, granulosa cells. Endocrinology 1984;114:1975–82. Cooke BA. Direct inhibition of ovarian steroidogenesis by cortisol 720. Fitzpatrick SL, Richards JS. Identification of a cyclic adenosine and the modulatory role of 11 beta-hydroxysteroid dehydroge- 3′,5′-monophosphate-response element in the rat aromatase pro- nase. Clin Endocrinol (Oxf) 1993;38:641–4. moter that is required for transcriptional activation in rat granu- 737. Hsueh AJ, Erickson GF. Glucocorticoid inhibition of FSH-induced losa cells and R2C Leydig cells. Mol Endocrinol 1994;8:1309–19. estrogen production in cultured rat granulosa cells. Steroids 721. Nielsen ME, Rasmussen IA, Kristensen SG, et al. In human gran- 1978;32:639–48. ulosa cells from small antral follicles, androgen receptor mRNA 738. Yang JG, Yu CC, Li PS. Dexamethasone enhances follicle stimu- and androgen levels in follicular fluid correlate with FSH receptor lating hormone-induced P450scc mRNA expression and pro- mRNA. Mol Hum Reprod 2011;17:63–70. gesterone production in pig granulosa cells. Chin J Physiol 2001; 722. Wang XN, Greenwald GS. Hypophysectomy of the cyclic 44:111–9. mouse. I. Effects on folliculogenesis, oocyte growth, and follicle- 739. Krozowski Z, Li KX, Koyama K, et al. The type I and type II stimulating hormone and human chorionic gonadotropin recep- 11beta- hydroxysteroid dehydrogenase enzymes. J Steroid Biochem tors. Biol Reprod 1993;48:585–94. Mol Biol 1999;69:391–401. 723. Wang XN, Greenwald GS. Hypophysectomy of the cyclic mouse. 740. Michael AE, Evagelatou M, Norgate DP, et al. Isoforms of 11beta- II. Effects of follicle-stimulating hormone (FSH) and luteinizing hydroxysteroid dehydrogenase in human granulosa-lutein cells. hormone on folliculogenesis, FSH and human chorionic gonado- Mol Cell Endocrinol 1997;132:43–52. tropin receptors, and steroidogenesis. Biol Reprod 1993;48:595–605. 741. Waddell BJ, Benediktsson R, Seckl JR. 11beta-hydroxysteroid 724. Whitelaw PF, Smyth CD, Howles CM, Hillier SG. Cell-specific dehydrogenase type 2 in the rat corpus luteum: induction of mes- expression of aromatase and LH receptor mRNAs in rat ovary. senger ribonucleic acid expression and bioactivity coincident J Mol Endocrinol 1992;9:309–12. with luteal regression. Endocrinology 1996;137:5386–91.

4. FEMALE REPRODUCTIVE SYSTEM