Physical Address

304 North Cardinal St.
Dorchester Center, MA 02124

Steroid Hormone Action


Steroid Hormone Receptors Act as Ligand-Dependent Transcription Activators or Repressors

  • Steroid hormones are derived from the metabolic conversion of cholesterol into biologically active steroid products that bind to intracellular receptors with high specificity.

  • The canonical mechanism of action for the steroid hormone receptors involves regulating gene transcription through transactivation and transrepression.

  • Precise regulation of gene transcription is essential for development, physiology, and homeostasis.

Steroids are small, lipophilic hormones synthesized from a common precursor molecule, cholesterol, through a complex biosynthetic process in specific tissues and glands throughout the body (see Chapter 4 ). Despite their shared molecular origin and basic structural similarities, the steroids mineralocorticoids , glucocorticoids , estrogens , progestins , and androgens are distinct classes of hormones that interact with specific, high-affinity receptors to exert their unique biological effects (mineralocorticoid [MR], glucocorticoid [GR], estrogen [ER], progestin [PR], and androgen [AR]). These hormones control diverse physiological and cellular processes and affect almost all aspects of eukaryotic physiology, from sexual differentiation, growth, and reproductive functions to immunity, metabolism, behavior, and learning. Consequently, a clear and complete understanding of the general mechanisms of steroid hormone action, as well as those activities that occur in a tissue- and cell-type-specific manner, is of critical importance for the promotion of health and the understanding of disease processes.

This chapter reviews the structural similarities and differences among the steroid hormone receptors, as well as what is known regarding their mechanisms of action. The classic mode of action entails simple diffusion of steroid hormones into the cell, where they interact with cognate receptors and stimulate or inhibit transcription of target genes ( Fig. 5.1 ). The hormone-dependent changes in receptor conformation can drive both the transactivation and transrepression of gene expression by (1) altering interactions with molecular chaperones that keep the receptor in a ligand-independent state; (2) inducing posttranslational modifications that alter receptor activity; (3) promoting the formation of receptor dimers; (4) enhancing interactions with specific DNA sequences ( hormone response elements ); and (5) facilitating recruitment of coactivator or corepressor proteins that alter chromatin structure and contact with the basal transcription machinery. Recent mechanistic studies have built on the classic mode of action to reveal a complex regulatory network of interacting factors and chromatin state. , For example, steroid hormone receptors have been shown to interact with closed chromatin through the recognition of partial DNA sequence motifs to enable other transcription factors to engage chromatin and form regulatory complexes (e.g., pioneer factors) or work through receptor cooperation to initiate the opening of chromatin and allow for the binding of other steroid hormone receptors or secondary transcription factors. , In addition to reviewing the various mechanisms of steroid hormone action, this chapter will discuss factors that regulate hormone activity, as well as nonclassical modes of action for steroid hormones and their receptors.

Fig. 5.1, General mechanism of action for cytoplasmic steroid receptors as described in the text.

Steroid Hormone Receptor Structure and the Evolution of Specificity

  • The steroid hormone receptors belong to a large family of transcription factors called nuclear receptors.

  • The structure of the steroid hormone receptors is modular, with distinct domains. The steroid hormone receptors contain a highly conserved DNA-binding domain (DBD), a moderately conserved ligand-binding domain (LBD), and less well-conserved amino- and carboxy-terminal domains.

  • Phylogenetic analysis determined that the steroid hormone receptors cluster separately from other ligand-dependent transcription factors. The evolution of the steroid hormone receptors is controversial and may reflect gene duplication, mutation, and functional divergence.

The steroid hormone receptors belong to a larger family of structurally and evolutionarily related proteins called nuclear receptors , encoded by 48 genes in the human genome. All nuclear receptors, including the steroid hormone receptors, exhibit a modular structure composed of distinct domains ( Fig. 5.2 ). In general, nuclear receptors contain a variable amino-terminal region (A/B), a highly conserved DBD (C), a highly variable hinge region (D), and a moderately conserved hormone- or ligand-binding domain (LBD) (E). , Some receptors also contain a carboxy-terminal F domain. The primary structure of each human steroid hormone receptor is shown, along with its physiological ligand ( Fig. 5.2 ). Specific residues within the DBD and LBD play an important part in receptor dimerization, which is critical because most nuclear receptors are only transcriptionally active as homo- or heterodimers. Finally, nuclear receptors have regions called activation function 1 and 2 (AF1 and AF2) that are required to transactivate gene expression. Whereas the activity of AF1 is usually ligand-independent and located in the A/B domain, AF2 is found in the LBD and is predominantly regulated by hormone binding.

Fig. 5.2, Schematic diagram of the primary structure of a generic steroid receptor and its functional domains.

Some nuclear receptors have defined natural ligands, such as the steroid hormones, thyroid hormones, retinoids, or vitamin D, but the ligand for other nuclear receptors has not yet been identified and are therefore termed “orphan receptors.” The finding that diverse compounds act as ligands for nuclear receptors and that some receptors have no apparent ligand led to the hypothesis that ancestral nuclear receptors were constitutive transcription factors that independently evolved the ability to bind ligands. , However, a second hypothesis posits that ancestral nuclear receptors were ligand-dependent transcription factors that evolved specificity for different ligands by gene duplication, mutation, and functional divergence. Under this hypothesis, the yet-to-be-identified ligands for the orphan receptors are likely intermediates in the synthesis of ligands for related receptors. There are several lines of evidence that favors the latter hypothesis for the evolution of ligand binding in the steroid hormone receptor family. For example, the primary, secondary, and tertiary structures of the LBD from different steroid hormone receptors are highly similar. Furthermore, detailed sequence, structural, and functional analyses support the hypothesis that the ancestral steroid hormone receptor-bound estrogens and specificity for other steroids evolved by serial and parallel duplications of the ancestral gene, mutation of nucleotides coding for specific amino acids, and structural and functional divergence of the paralogs. , , Indeed, reconstructing ancestral sequences, using a maximum likelihood approach, suggests that the first steroid hormone receptor was an ER-like molecule ( Fig. 5.3 ). When this gene was duplicated, one copy was constrained by natural selection and retained its function as an estrogen-binding receptor, while the other copy evolved specificity for 3-ketosteroid-like ligands ( Fig. 5.3 ). Duplication of the latter gene then produced a corticoid receptor-like protein and a receptor for 3-ketogonadal steroid-like molecules (e.g., androgens, progestins, or both). This final duplication then led to the evolution of the true androgen and progesterone receptors from the ancestral 3-ketogonadal steroid hormone receptor.

Fig. 5.3, Phylogeny of the steroid receptor gene family.

Crystal structures of the various steroid hormone receptors have shown that the LBD folds into a highly homologous three-layered structure with a small ligand-binding pocket in the center, while domain swapping of the entire LBD between steroid hormone receptors has shown that this region determines specificity for particular classes of steroid hormones. The ligand-binding pocket is composed of roughly 30 amino acids that are in close proximity or make direct contact with hormones when bound to their cognate receptors. In agreement with structural studies, experiments using site-directed mutagenesis indicate that specific, but minor, changes of amino acids within the ligand-binding pocket can lead to dramatic changes in the binding specificity of the steroid hormone receptors. For example, the human PR, GR, and MR contain a conserved cysteine residue that appears to be critical for contacting the C20 keto group found in progestins, glucocorticoids, and mineralocorticoids. , Mutation of the corresponding threonine to cysteine in the AR reduces its affinity for androgens and allows the receptor to transactivate in the presence of progesterone and corticoids. Based on structure-function studies of this sort and phylogenetic analyses, Thornton proposed a series of relatively minor amino acid changes that may account for broad changes in hormone specificity during the evolution of the steroid hormone receptors.

Similar studies of the DBD have defined the molecular basis for interactions between steroid hormone receptors and particular DNA sequences. , The DBD of nuclear receptors contains two zinc fingers. The first zinc finger interacts with the major groove of DNA, whereas the second is involved in receptor dimerization. Mutation of three residues within a five-residue motif known as the Proximal box (P box) of the first zinc finger of the GR to the corresponding residues in the ER changes the binding specificity from DNA sequences called glucocorticoid-response elements (GREs) to estrogen-response elements (EREs) and vice versa ( Fig. 5.4 ). Although the DBD is very highly conserved among nuclear receptors, the hormone response elements are variable among different receptors, which may in part account for receptor-specific regulation of distinct sets of genes. These examples illustrate how site-directed mutagenesis and fine-scale comparison of amino acid sequences among receptors, in the context of the tertiary structure of the DBD and LBD, facilitate testable hypotheses regarding the evolution of signaling and regulation of gene expression by different classes of steroids.

Fig. 5.4, Amino acid sequence of the DNA-binding domain of human glucocorticoid receptor showing two zinc fingers and the P box, which is outlined.

Steroid Hormone Receptor Function

  • The steroid hormone receptors have well-described functions in the reproductive tract but also critically regulate general physiology.

  • Transgenic animal models and mutations identified in humans have provided insight into the cell-specific activities of the steroid hormones and their receptors.

  • ER, PR, and AR are indispensable for reproductive function but also regulate aspects of physiology outside of the reproductive system.

  • GR signaling is essential for life after birth, maintaining general physiologic homeostasis, and integrating the hypothalamic-pituitary-adrenal (HPA) axis with reproductive functions mediated by the hypothalamic-pituitary-gonadal (HPG) axis.

  • MR is required for electrolyte balance and fluid transport, where insufficiency or deficiency can lead to mortality.

Steroid hormone receptors have taken on distinct physiological roles during evolution, which has been demonstrated by transgenic animal models and case reports in humans with genetic mutations/deletions. For example, the sex hormone receptors demonstrate a pattern of expression primarily restricted to cells and organs of the HPG axis and play an essential role in sexual differentiation and reproduction. However, these steroid hormone receptors are not sex-limited and also have some functions outside of the reproductive system in both sexes. In contrast, the GR is expressed ubiquitously and plays a role in a wide range of physiological processes, including, development, immune function, cognition and behavior, cardiovascular health, metabolic homeostasis, and reproduction. , Outside of their roles in reproductive functions and overall physiology, the steroid hormone receptors can drive disease pathology when signaling becomes unregulated. Functions of ER, PR, AR, GR, and MR in physiology and pathophysiology are described briefly below.

Estrogen Receptor

The biological effects of estrogens were originally believed to be mediated by a single receptor (ERα) until the cloning of a second receptor (ERβ) and reports of estrogen binding to a membrane-bound G protein-coupled receptor (GPR30/GPER) several decades later. , The estrogen receptors are products of different genes and demonstrate distinct expression patterns across tissues and cell types ( Table 5.1 ). Along with their tissue type-specific distribution, the different estrogen receptors regulate unique biological functions, which have been characterized in transgenic animal models. For example, the ovarian phenotypes of ERα and ERβ knockout mice are distinct, likely due to the cellular distribution of these two receptors in the ovary, where ovarian granulosa cells express ERβ, and ERα is primarily found in the germinal epithelial cells, interstitial cells, and theca cells of the ovary. Subsequently, the ERα knockout mouse is anovulatory and accumulates cystic and hemorrhagic follicles, whereas the ERβ knockout mouse contains histologically normal ovaries but still displays impaired ovulation. In the uterus, the absence of ERα results in a significant infertile phenotype due to a hypoplastic uterus and defects in implantation and decidualization; however, the loss of ERβ does not alter uterine development, function, or cellular responses to estradiol. , Subsequently, cell-specific deletion models have dissected the biological effects of ERα in the uterine stroma and epithelial cells. The mammary gland is also a key organ for ERα signaling. ERα knockout mice demonstrate impaired ductal growth and differentiation following the onset of endogenous estrogen production during puberty. , The loss of ERβ does not alter ductal growth but appears to play a role in differentiation of the mammary gland during pregnancy. In males, loss of ERα results in infertility related to defects in fluid resorption in the efferent ducts and epididymis with progressive deterioration of testicular morphology, while the ERβ knockout males are fertile with phenotypically normal testis. , Gene targeting technology has also led to the creation of mouse models with mutations in the functional domains of ERα, which have allowed the functional domains to be associated with specific physiological functions. For example, the functional domain mediating the metabolic functions of ERα have been discriminated using mouse models that selectively target the AF1 and AF2 domains. ,

Table 5.1
Tissue-Specific Patterns of ERα and ERβ mRNA Expression in the Rat
From Kuiper GGJM, Carlsson B, Grandien K, et al. Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors α and β. Endocrinology. 1997;138:863–870.
Receptor ∗
Tissue ERα ERβ
Epididymis +++ +
Prostate + +++
Testis +++ +
Pituitary ++ +
Ovary +++ +++
Uterus +++ ++
Bladder + ++
Lung 0 +
Liver + 0
Kidney ++ 0
Thymus + +
Adrenal ++ 0
Olfactory lobe 0 +
Cerebellum 0 +
Brain stem 0 +
Spinal cord 0 +
Heart + 0

∗ Relative levels of expression are indicated by the number of plus signs: 0, not detected; +, low; ++, medium; +++, high.

Although both ERα and ERβ bind estradiol with similar affinities ( Table 5.2 ), in vitro studies and computational modeling have determined that the relative affinity for exogenous estrogenic compounds differ from the receptor subtypes. The differences in ligand binding have been exploited for the benefit of various therapies (selective estrogen receptor modulators). However, some exogenous estrogen-like compounds are also able to interfere with endogenous ligand binding, resulting in aberrant activation or repression of ERα and ERβ functions and endocrine-disrupting activities. One of the first examples of endocrine disruption by a synthetic estrogenic compound occurred when women were treated with diethylstilbestrol (DES) during pregnancy. Originally prescribed to pregnant women from the 1940s to 1970s to help prevent miscarriage, it was later discovered that DES did not work to prevent miscarriage, and in fact, was linked to vaginal cancer in female offspring exposed prenatally to DES. The long-term effects of in utero DES exposure were ultimately linked to other reproductive pathologies, including vaginal adenosis, uterine malformations, polycystic ovary, cryptorchidism, epididymal cysts, and hypoplastic testes, as well as an increased risk for infertility, ectopic pregnancy, miscarriage, preterm birth, stillbirth, and early menopause in the female offspring of DES-treated women. Subsequently, several other estrogenic endocrine disruptors have been studied with reported impacts on ER signaling that vary in severity and duration. ,

Table 5.2
Binding Affinity of Various Ligands for ERα and ERβ Relative to Binding Affinity of E2
From Kuiper GGJM, Carlsson B, Grandien K, et al. Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors α and β. Endocrinology. 1997;138:863–870.
Relative Binding Affinity ∗
Ligand ERα ERβ
E2 100 100
Diethylstilbestrol 468 295
Hexestrol 302 234
Dienestrol 223 404
Estrone 60 37
17α-Estradiol 58 11
Moxestrol 43 5
Estriol 14 21
4-OH-Estradiol 13 7
2-OH-Estradiol 7 11
Estrone-3-sulfate <1 <1
4-OH-Tamoxifen 178 339
ICI-164384 85 166
Nafoxidine 44 16
Clomifene 25 12
Tamoxifen 7 6
Coumestrol 94 185
Genestein 5 36
Bisphenol A 0.05 0.33
Methoxychlor 0.01 0.13

∗ Relative binding affinity is the ratio of concentrations of E2 and competitor required to displace 50% of specific radioligand binding. Relative binding affinity was set to 100% for E2.

In addition to their physiological role in regulating reproductive tract function, estrogens are a contributing factor to the initiation and progression of breast cancer. However, ER antagonists can slow or stop the growth of breast cancer by blocking estrogen-regulated gene expression. The carcinogenic effects of estrogen in breast tissue may, in part, be mediated by the induction of vascular endothelial growth factor (VEGF), as tumor growth is generally dependent on angiogenesis and a steady blood supply. Interestingly, the breast cancer susceptibility gene ( BRCA1 ) can directly interact with ER and block the induction of VEGF. Mutations that inactivate BRCA1 and confer an increased risk for breast cancer may be due to a loss of the ability to antagonize the actions of estrogen.

ER loss-of-function mutations in humans are rare but have been reported in a small number of patients. The clinical phenotypes of these patients were comparable to the ER knockout mouse model. Females presented with amenorrhea, cystic ovaries, thin endometrium, and no breast development. The predominant phenotype of affected males was delayed bone maturation. Somatic mutations and genetic polymorphisms in the human ER are also associated with various disease states. For example, many ER mutations have been detected in breast cancer. A high-frequency somatic mutation between D and E regions in ERα was identified in breast hyperplasia and was found to enhance estrogen sensitivity when evaluated in vitro . Breast cancer cell models of two other common ERα mutations, Y537S and D538G, demonstrate that some mutations bestow ligand-independent gene regulation, growth, and resistance to ER modulators. Polymorphisms in ERα and ERβ have also been associated with breast cancer severity. Finally, various polymorphisms identified in genomic ER binding sites are predicted to alter ER activity at those sites by increasing or decreasing ER-DNA interactions and can contribute to breast cancer pathogenesis. Thus, the presence of ER mutations and polymorphisms mediates the pathogenesis of estrogen-driven disease as well as the response to treatment.

Progesterone Receptor

Like estrogens, progestins are necessary for female fertility. In fact, the etymology of the word progesterone is “ progest ational ster oid horm one .” Progesterone action is primarily mediated by two PR isoforms, PR-A and PR-B, derived from the same gene. The necessity of progesterone signaling for female fertility was demonstrated in a mouse model devoid of both PR isoforms. The female PR knockout mice are infertile due to defects in mating behavior, ovarian follicle rupture, implantation, and uterine function. Knockout females also displayed disrupted mammary gland development. In contrast, PR deficient male mice appear normal, and fertility was comparable to wild-type males.

The isoforms PR-A and PR-B are alike except for the addition of 164 amino acids at the amino-terminal end of PR-B due to the utilization of different promoters and translational start sites. These 164 amino acids confer an ancillary activation function domain (AF3), which contributes to the functional differences between the two isoforms. The AF3 domain prevents the activity of a transcription inhibitory domain in the AF1 of PR-B, while the absence of the AF3 in PR-A allows the inhibitory domain to function and renders this isoform more “transcriptionally repressive.” Moreover, at sites of transcriptional activity, PR-B preferentially recruits coactivators, while PR-A recruits corepressors. Mice with targeted deletion of either PR-A or PR-B have allowed the functions of the PR isoforms to be characterized in different tissues in vivo . , Deletion of the PR-A isoform phenocopied the fertility defects described in PR null mice, whereas selective ablation of PR-B only affected mammary gland development. Overexpression of PR-A in the uterus also resulted in infertility due to disrupted embryo attachment, indicating that precise regulation over the spatiotemporal expression of PR-A is required for the establishment of pregnancy. A third PR isoform, PR-C, has been reported and results from the initiation of translation of 430 amino acids downstream of the PR-A start site in the PR gene. Although PR-C lacks the AF1 and DBDs, it has been shown to heterodimerize with PR-B and alter its transcriptional activity.

One of the primary functions of progesterone signaling in the uterus prior to implantation is to inhibit the early actions of estradiol. For example, estrogen signaling drives uterine epithelial cell proliferation. However, estrogen signaling also promotes the expression of the PR in the uterus, which subsequently inhibits epithelial proliferation through paracrine signaling to allow embryo implantation. Likewise, mucin 1 (MUC1), a cell surface glycoprotein that provides a protective barrier to the uterine epithelium, is induced by estradiol in the uterus. However, the preimplantation rise in progesterone levels leads to the down-regulation of MUC1, which is necessary for the embryo to adhere, implant, and invade. PR also critically coordinates paracrine signaling across the stroma and epithelium during early pregnancy. Progesterone target genes in the epithelium mediate downstream signaling events in the stroma, which are critical during embryo implantation and for postimplantation maintenance of pregnancy.

The antiproliferative properties of progesterone in the uterus have been targeted for treatment of endometrial cancer. The standard treatment for endometrial carcinomas is often surgical. However, patient desire to preserve fertility has prompted the use of progestin hormonal therapy as an alternative treatment option with varied success. Interestingly, progesterone promotes the growth of uterine fibroids, benign tumors of the uterine smooth muscle, by stimulating proliferation and hypertrophy. For this reason, selective progesterone receptor modulators with antagonistic activity have been evaluated as therapeutic agents for fibroids.

In accord with the finding that PR is necessary for development of the mammary ducts, progesterone signaling has also been shown to play a role in the pathogenesis of breast cancer. Progesterone regulates proliferation of the mammary epithelium through autocrine and paracrine pathways. One of the main paracrine factors mediating progesterone-induced proliferation is the receptor activator of nuclear factor kappa-B ligand (RANKL). The induction of mitogenic factors by progesterone also contributes to the maintenance and expansion of the mammary gland stem cells. In the Women’s Health Initiative trial, combined estrogen plus progesterone therapy as part of a hormone replacement regimen in postmenopausal women increased the risk of breast cancer. However, studies have also demonstrated that PR activation in certain types of breast cancer improved progression-free survival. The differential actions of progesterone may reflect the distinct cellular context of healthy and malignant tissue or the dose, timing, and specificity of the progestogen utilized.

Androgen Receptor

Androgens, the main male sex steroids, drive the differentiation of the male reproductive tract during development and are crucial for reproductive functions in the adult. During development, the androgen-producing cells of the mammalian gonad arise from precursor cells in the interstitium to become Leydig cells in males and theca cells in females. Androgens produced by the fetal Leydig cells direct the development of the male external genitalia, vas deferens, and related structures from the Wolffian ducts. In mouse models of disrupted androgen signaling, the Wolffian ducts regress and the external female genitalia develop. Likewise, mutations in the human AR, causing androgen insensitivity syndrome, result in a female phenotype in individuals who are genotypically male. Mutations in AR have also been identified that lead to partial androgen insensitivity by disrupting AR-cofactor interactions. Furthermore, studies with the AR deficient mouse model have demonstrated a role for androgen signaling in female reproduction. AR null female mice undergo premature ovarian senescence due to follicle atresia. Subsequent analysis of the AR null mice revealed defective luteinization of the preovulatory follicles. In agreement with the animal studies, AR gene mutations have been characterized in female patients with premature ovarian failure.

Following reproductive tract development, androgens, primarily testosterone and 5α-dihydrotestosterone, induce and maintain the secondary sex characteristics, stimulate male reproductive behavior, and support spermatogenesis. In the testis, the primary target of androgen action is the Sertoli cell. Mice with a targeted deletion of AR in Sertoli cells are infertile due to spermatogenic arrest at meiosis and impaired Sertoli cell barrier formation (blood-testis-barrier). , AR has also been conditionally deleted from the Leydig cells, which caused spermatogenesis arrest at the round spermatid stage, empty and atrophied epididymides, and infertility. The Leydig cell-specific AR knockout mice also exhibited higher serum gonadotropin levels and lower serum testosterone levels. These observations indicate that testicular AR signaling is required to maintain spermatogenesis. During normal spermatogenesis, germ cell apoptosis is a constant feature of the adult testis and is required to maintain homeostasis within the seminiferous tubule. Exposure to excess and insufficient levels of testosterone can cause testicular cell death. In vitro culture of testicular tissue samples with testosterone improved measures of apoptosis. The mechanism by which testosterone acts as a pro-survival factor has been partially attributed to the inhibition of pro-death signals. Conversely, excess testosterone can cause cell death through the induction of pro-death signals, namely Fas and FasL.

The prostate epithelium is also a target of androgen action during development, and androgen signaling is required to maintain cellular homeostasis into adulthood. Castration results in prostate involution due to epithelial cell apoptosis, which is reversible by androgen supplementation. Cell-specific knockout models of AR in the prostate indicate that paracrine signaling by stromal AR regulates epithelial cell proliferation. The pro-survival actions of AR in the prostate play a fundamental role in the pathogenesis of prostate cancer. In fact, over 75 years ago, it was shown that androgen deprivation by castration caused regression of prostate cancer. Androgen-initiated prostate cancer growth is believed to occur through crosstalk with several signaling pathways, namely pathways involved in metabolic regulation, cell cycle progression, and growth factor signaling. Moreover, AR signaling in the epithelium, independent of paracrine AR signaling in the stroma, has been implicated in the “switch” from nonmalignant to a cancerous prostate. AR also contributes to prostate cancer progression by promoting angiogenesis through the induction of VEGF. Therefore, AR is the primary therapeutic target in prostate cancer. It is now clear that AR signaling contributes to cancers in organs other than the prostate, including the breast, bladder, pancreas, ovary, and endometrium.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here