Steroid hormone formation in the steroid-producing endocrine glands follows the same fundamental pathway and mainly relies on exogenous (or plasma) cholesterol, with the exception of the liver and intestinal mucosa, which are capable of synthesizing cholesterol endogenously from acetyl-coA. The primary source of cholesterol for steroidogenesis in the ovary is derived from the uptake of plasma LDL.⁴ The rate-limiting step in steroidogenesis is transfer of cholesterol from the cytosol to the inner membrane of the mitochondria.⁵ This is mediated by an LH-induced mitochondrial enzyme called steroidogenic acute regulatory (StAR) protein.⁶ The StAR gene is located on chromosome 8p11.2 and codes for a 285-amino acid precursor protein, of which 25 amino acids are cleaved off after transport to the mitochondria.^7,⁸ Nonsense mutations of the StAR gene that result in premature stop codons have been identified as a cause of congenital lipoid adrenal hyperplasia, which is characterized by the presence of intracellular lipid deposits that destroy steroidogenesis.⁷

Ovary steroid hormones are synthesized in both interstitial and follicular cells. The basic structure of cholesterol is three hexagonal carbon rings and a pentagonal carbon ring to which a side chain is attached (Fig. 2-1). Two important methyl groups are also attached at positions 18 and 19. Progestins and corticosteroids (pregnane series 21-carbon steroids) are produced by partial cleavage of the side chain (i.e., the desmolase reaction). Androgens (androstane series 19-carbon steroids) are produced by the total removal of the side chain. Estrogens (estrane series 18-carbon steroids) are produced by aromatization of one of the three hexagonal carbon rings to a phenolic structure with loss of the 19-methyl group.

Figure 2-1 Structures of cholesterol (27 carbons) and of the three major classes of ovarian steroids: the progestins (21 carbons), androgens (19 carbons), and estrogens (18 carbons).

The first step in steroidogenesis is the conversion of cholesterol to pregnenolone via hydroxylation at the carbon 20 and 22 positions, which is followed by cleavage of the side chain (Fig. 2-2). From pregnenolone, steroid hormones are produced by one of two general pathways.⁵ The pregnenolone (Δ⁵) pathway produces androgens and estrogens (pregnenolone→17OH-pregnenolone →dehydroepiandrosterone [DHEA]→testosterone→estrogen). The progesterone (Δ⁴) pathway produces androgens and estrogens (pregnenolone→progesterone→17OH-progesterone→androgen →estrogen). In the adrenal gland, the Δ⁴ pathway produces mineralocorticoids and glucocorticoids.

Figure 2-2 Steroid hormone biosynthesis pathways in the ovary. The Δ⁴ pathway converts pregnenolone to progesterone and is the main pathway in the corpus luteum. The Δ⁵ pathway converts pregnenolone to androgens and subsequently estrogens and is the preferred pathway in the thecal cells. Also shown are the conversion of 17OH-progesterone to aldosterone and cortisol as a result of CYP21 and CYP11, which occurs exclusively in the adrenal gland. Irreversible reactions are denoted by a single arrow, and reversible reactions are denoted by double arrows.

The enzymes involved in the intracellular synthesis of steroid hormones include five hydroxylases, two dehydrogenases, a reductase, and an aromatase. The hydroxylases and aromatase belong to the cytochrome P450 (CYP) supergene family (Table 2-1). These enzymes exist on both the mitochondria and endoplasmic reticulum.

Table 2-1 Enzyme Reaction and Cellular Location of Steroidogenic Enzymes

Enzyme Reaction	Gene (Enzyme)	Cellular Location/Tissue Location
Cholesterol side chain cleavage	CYP11A (P450scc)	Mitochondria (theca; granulosa)
17α-hydroxylase	CYP17 (P450c17)	ER (theca)
17,20-hydroxylase (lyase)	CYP17 (P450c17)	ER (theca)
Aromatase	CYP19 (P450arom)	ER (granulosa)
3β-hydroxysteroid dehydrogenase	3βHSD	ER (theca; granulosa)
17β-hydroxysteroid dehydrogenase	17βHSD	ER (granulosa)
21-hydroxylase	CYP21 (P450c21)	ER (adrenal)
11β-hydroxylase	CYP11B1 (P450c11)	Mitochondria (adrenal)

ER, endoplasmic reticulum

Of these nine enzymes, four key enzymes regulate the main steps of steroidogenesis (see Fig. 2-2): CYP11A (P450scc), a side chain cleavage enzyme that catalyzes the conversion of cholesterol to pregnenolone; 3βHSD, or 3 βa-hydroxysteroid dehydrogenase, which converts pregnenolone to progesterone; CYP17 (P450c17), an hydroxylase that converts pregnenolone to androgens; and CYP19 (P450arom), an aromatase that converts androgens to estrogens. Most reactions are irreversible (denoted by a single arrow in Fig. 2-2). The few reversible reactions (denoted by double arrows) are dependent on cofactor availability (e.g., NADP/NADPH ratio).

The kind of hormone produced depends on the nature of the cell and the presence or absence of the inherent steroidogenic enzymes in the tissue. The adrenal cortex lacks 17βHSD; hence, adrenal androgen production is limited to DHEA and androstenedione. In the testes, LH controls 17βHSD activity and testosterone production. The steroid-producing cells of the ovary (granulosa, theca, corpus luteum) contain the full enzymatic complement for steroid hormone synthesis. In the thecal cells, LH also controls 17βHSD activity and androstenedione production, whereas CYP19 (P450arom) activity in the granulosa cells is controlled by FSH and hence estradiol production. These relationships are the basis for the two-cell, two-gonadotropin system (Fig. 2-3). Aromatization occurs in the endoplasmic reticulum.

Figure 2-3 Differential regulation by luteinizing hormone (LH) and follicle-stimulating hormone (FSH) of ovarian estrogen, progesterone, and androgen production, the basis of the two-cell, two-gonadotropin system. LH acts on both thecal and granulosa cells; FSH acts only on granulosa cells. FSH and LH stimulate adenylate cyclase via G protein-coupled receptors. Cyclic adenosine monophosphate (cAMP) generated from adenosine triphosphate (ATP) activates protein kinase A, which in turn stimulates steroidogenic enzymes. Gs, G protein; GDP, guanosine diphosphate; GTP, guanosine triphosphate.

In each of the two cell types, the amount of the various enzymes differs depending on the stage of follicle development. CYP11A (P450scc) and 3βHSD are expressed in both thecal and granulosa cells of antral and preovulatory follicles and in the luteinized granulosa and thecal cells of the corpus luteum. In contrast, CYP17 (P450c17) is expressed only in the thecal cells of antral and preovulatory follicles and of the corpus luteum (see Fig. 2-3).

Steroid Hormones of the Ovary

On the basis of chemical structure and biologic function, the major steroid hormones synthesized and secreted by the ovaries can be classified into three major types: estrogens, progesterone, and androgens.

Estrogens

Physiologic Role

Estrogens are essential in the development and maintenance ofthe female phenotype, germ cell maturation, and pregnancy. In addition to their reproductive effects, estrogens also have many other nonreproductive systemic effects, such as bone metabolism/remodeling, nervous system maturation, and endothelial responsiveness.⁹

At puberty, estrogen stimulates breast development and enlargement and maturation of the uterus, ovaries, and vagina.^10,¹¹ Estrogen also works in concert with growth hormone and insulin-like growth factor I (IGF-I) to produce a growth spurt and stimulates the maturation of chondrocytes and osteoblasts, which ultimately leads to epiphyseal fusion.^12,¹³ After midpuberty, estrogen begins to exert a positive feedback on gonadotropin-releasing hormone (GnRH) secretion, leading to the progressive increase of LH and FSH production, culminating in the LH surge, ovulation, and the initiation of the menstrual cycle.

In the adult female, estrogen plays a critical role in maintaining the menstrual cycle.¹⁴ The cyclical changes in estradiol, progesterone, and pituitary hormones are illustrated in Figure 2-4. In the early follicular phase of the menstrual cycle, FSH stimulates granulosa cell aromatase activity, resulting in increased follicular concentrations of estrogen. The rising estrogen level further increases the sensitivity of the follicle to FSH and estrogen by increasing the number of estradiol receptors on the granulosa cells. Follicular growth and antral formation is also promoted by estrogen. This sets up a positive feedback cycle, which culminates in one dominant follicle producing an exponential rise in estrogen levels. This exerts a negative feedback on FSH so that falling FSH levels contribute to atresia of other nondominant follicles. The dominant follicle secretes large quantities of estrogen; estradiol levels must be greater than 200 pg/mL for approximately 50 hours before a positive feedback on LH release is achieved.^13,¹⁵ Once the LH surge is initiated, luteinization of the granulosa cells and progesterone production occurs. In pregnancy, estrogen augments uterine blood flow, although it is not required in itself for the maintenance of pregnancy.¹⁶

Figure 2-4 Plasma hormone concentrations (mean ± standard error) during the female menstrual cycle. Graph A, inhibins; Graph B, progesterone and estradiol; Graph C, LH and FSH.

(Data from Groome NP, et al: Measurement of dimeric inhibin B throughout the human menstrual cycle. J Clin Endocrinol Metab 81:1401–1405, 1996.)

In the central nervous system, estrogen withdrawal at menopause has been associated with reduced libido, altered mood, and cognitive disturbances. These effects have been attributed to estrogen’s ability to modulate the synthesis, release, and metabolism of many neuropeptides and neurotransmitters.¹⁷ Estrogen acts as a serotoninergic agonist by increasing serotonin synthesis in the brain, which may positively influence mood.¹⁸ Although prospective observational studies in postmenopausal women have suggested that estrogen replacement therapy might protect against cognitive decline¹⁹ and the development of dementia,²⁰ randomized trials of estrogen in the treatment of Alzheimer’s disease have shown no evidence of benefit.²¹^–²⁴

In the skeletal system, estrogen antagonizes the effect of parathyroid hormone by directly inhibiting the function of osteoclasts, which decreases the rate of bone resorption and diminishes bone loss. The Postmenopausal Estrogen/Progestin Interventions (PEPI) trial was a prospective, placebo-controlled trial designed to study the effects of hormone replacement on bone density in postmenopausal women. After 12 months of treatment with estrogen, bone mineral density increased by 1.8% at the hip and by 3% to 5% at the spine.²⁵ The Women’s Health Initiative (WHI) showed that estrogen reduced the risk of both hip and vertebral fractures by 30% to 39%.²⁶

In the cardiovascular system, there is strong evidence that estrogen has a natural vasoprotective role. At a cellular level, estrogen receptors are found on the smooth muscle cells of coronary arteries ²⁷ and the endothelial cells of various sites.²⁸ Estrogen causes short-term vasodilation by increasing nitric oxide and prostacyclin release in endothelial cells.²⁹ Several large observational studies, including the Framingham study and the Nurses Health Study, have shown that cardiovascular incidence rates are lower in premenopausal than postmenopausal women.³⁰ There was also a significant association between a younger age at menopause and a higher risk of coronary artery disease.³¹ These studies led to the conviction that estrogen replacement therapy would consequently prevent the progression of atherosclerosis and coronary heart disease. However, the WHI study and the Heart and Estrogen/progestin Replacement Study (HERS), both large randomized, prospective trials designed to specifically address this issue, have not shown any benefit of estrogen for either the primary or secondary prevention of coronary artery disease, respectively.^26,³²

Biosynthesis and Metabolism

Estrogens are 18-carbon steroids derived from cholesterol (see Fig. 2-1). The three forms of naturally occurring estrogen include estrone, 17β-estradiol, and estriol. In nonpregnant females, estrone and estradiol are the main biologically active estrogens secreted by the ovary. Estradiol is almost 2 to 5 times more potent than estrone.³³ The circulating levels of estradiol are 2 to 4 times higher than those of estrone in premenopausal women. Estradiol concentrations in postmenopausal women are one tenth of those in premenopausal women. Estrone concentrations do not differ with menopausal status; thus, over time, the premenopausal estradiol-to-estrone ratio is reversed.³⁴ In contrast, estriol is not the secretory product of the ovary but is the peripheral metabolite of estrone and estradiol.

The main estrogen in premenopausal women is estradiol, which is produced primarily by the granulosa cells of the ovary. Androstenedione is converted to testosterone via 17βHSD, which is rapidly demethylated at the C-19 position and aromatized to estradiol. Estradiol is also generated to some degree from androstenedione via estrone. Estrone is also a secreted product of the ovary. It constitutes the remaining circulating estrogen (40%) and is mainly derived from the extragonadal peripheral aromatization of adrenal androstenedione.³⁵ Peripheral conversion of androgens to estrogens occurs in skin, muscle, and adipose tissue and in the endometrium.³⁶

In the normal adult female, the production of estradiol varies according to the phase of the menstrual cycle. During the mid luteal phase, for example, the production rate is about 100 to 270 μg/day. In comparison, the production rate for androstenedione is about 3 mg/day, and with its peripheral conversion rate to estrone of about 1.5%, it accounts roughly for about 10% to 30% of estrone production per day. Secondary increases in estrone formation occur in patients with polycystic ovaries or with ovarian cancer characterized by increased androgen production. In such patients, the increased estrogen can disturb the menstrual cycle. In postmenopausal women, the ovarian contribution shrinks, leaving estrone, derived from adrenal androstenedione, as the main source of circulating estrogen.³⁷

In the pregnant woman, the placenta becomes the main source of estrogen in the form of estriol. The placenta is unable to synthesize steroids de novo and depends on circulating precursors from both fetal and maternal steroids. Most of the placental estrogens are derived from fetal androgens (e.g., DHEA sulfate), produced by the fetal adrenal gland.³⁸ Fetal DHEA sulfate is converted to free DHEA by placental sulfatase and then to androstenedione and testosterone before being aromatized to estrone and estradiol. Finally, it is hydroxylated to form estriol.

Estradiol is rapidly converted in the liver to estrone by 17βHSD. Estrone can be further metabolized via three pathways. First, it can be hydroxylated to 16α-hydroxyestrone, which is then converted to estriol. Estriol is further metabolized by sulfation and glucuronidation, and the conjugates are excreted into the bile or urine. Secondly, estrone can be conjugated to form estrone sulfate, which occurs primarily in the liver. Estrone sulfate is biologically inactive and is present in concentrations that are 10-fold to 20-fold higher than concentrations of estrone or estradiol.³⁹ Estrone sulfate can be hydrolyzed by sulfatases present in various tissues to estrone and may serve as a reserve of estrogen in an inactive form. Estrone sulfate may be of some importance in assessing estrogenicity in women and can be detected in serum as well as in urine.⁴⁰ Thirdly, estrone can also be metabolized by hydroxylation to form 2-hydroxyestrone and 4-hydroxyestrone, which are known as catechol estrogens. These are then converted to the 2-methoxy and 4-methoxy compounds by catechol-O-methyltransferase.

Progesterone

Physiologic Role

Progesterone plays a critical role in reproduction. It inhibits further endometrial proliferation mediated by estrogen and converts the endometrium into a secretory type, preparing it not only to receive the blastocyst for implantation, but also to maintain the pregnancy. It inhibits uterine contractions and increases the viscosity of cervical mucus. Progesterone also inhibits the action of prolactin so that lactation occurs only after delivery. It raises the basal body temperature by about 0.5°C (0.9°F) and increases the sensitivity of the respiratory center to carbon dioxide (CO₂), leading to hyperventilation.

During pregnancy, progesterone increases insulin resistance in concert with the production of the other placental counterregulatory hormones, including placental growth hormone, placental lactogens, placental corticotropin-releasing hormone, and cortisol.⁴¹

Biosynthesis and Metabolism

Progesterone is part of the group of 21-carbon steroids, which also includes pregnenolone and 17OH-progesterone. Progesterone is responsible for all the progestational effects, whereas pregnenolone is the precursor for all steroid hormones. 17OH-progesterone has little biologic activity. Progesterone and 17OH-progesterone are mainly produced by the corpus luteum in the luteal phase of the menstrual cycle and by the placenta if pregnancy occurs. Circulating levels of progesterone at concentrations greater than 4 to 5 ng/mL (12.7 to 15.9 nmol/L) are indicative of ovulation.⁴²

Progesterone is rapidly metabolized by the liver and has a half-life of approximately 5 minutes. It is converted to pregnanediol and conjugated to glucuronic acid in the liver. Pregnanediol glucuronide is excreted in the urine. Pregnanetriol is the main urinary metabolite of 17OH-progesterone.

Androgens

Physiologic Role

In women, androgens originate as 19-carbon steroids from the adrenals and ovaries. The major androgens produced in the ovary, primarily by the thecal cells and to a lesser degree by the ovarian stroma, include DHEA, androstenedione, and a small amount of testosterone. Both DHEA and androstenedione serve as precursors to estrogen synthesis and have little, if any, androgenic activity. However, these biologically inactive androgens are converted by extraglandular metabolism to biologically active androgens such as testosterone and dihydrotestosterone (DHT). Normally, the levels of these potent androgens are low in females and have no significant physiologic function. Excessive production of androgens by the ovary or adrenals has been implicated as the cause of hirsutism and virilization in women.⁴³ In contrast, in the male the androgens, of which testosterone and DHT are the most crucial, are of primary importance.

Biosynthesis and Metabolism

Androgens are 19-carbon steroids derived from cholesterol. The rate-limiting step in androgen synthesis is the conversion of cholesterol to pregnenolone, which is mediated by the action of LH on the ovary and testes. In a normal ovulatory woman, the ovaries secrete approximately 1 to 2 mg of androstenedione, 1 mg of DHEA, and approximately 0.1 mg of testosterone. The majority (≈0.2 mg) of circulating testosterone is derived from peripheral metabolism of DHEA and androstenedione. Overall, testosterone production in women is about 0.3 mg/day; roughly 50% of this is derived from peripheral conversion whereas the remaining 50% is secreted equally by the ovary and the adrenals.⁴⁴

In the male, more than 95% of circulating testosterone is secreted by the testicular Leydig cells. The testes also secrete small amounts of DHT and the weak androgen DHEA and androstenedione. In most androgen target cells, testosterone is converted to the biologically more potent DHT by the enzyme 5α-reductase. In the female, androgens are derived either from the adrenal cortex in the form of DHEA and androstenedione or from the peripheral conversion of these androgen precursors to testosterone and DHT.

Most of the circulating testosterone is metabolized in the liver into androsterone and etiocholanolone, which are conjugated with glucuronic acid or sulfuric acid and excreted in the urine as 17-ketosteroids. Of note, only 20% to 30% of urinary 17-ketosteroids are derived from testosterone metabolism; the rest originate from the metabolism of adrenal steroids.

Transport of Ovarian Steroid Hormones in Plasma

Steroid hormones are not water-soluble and require transport proteins to be carried to their target tissues. The two types of transport proteins are general carrier proteins such as albumin and transthyretin and specific carrier proteins such as thyroxine-binding globulin, sex hormone-binding globulin (SHBG), and transcortin. Both types of proteins are produced in the liver. Less than 2% of ovarian steroid hormones are free in the circulation; the remainder are mostly bound to SHBG and albumin.^45,⁴⁶

Sex hormone-binding globulin, a β-globulin of 95 kDa, is synthesized in the liver. Its gene is localized on the short arm of chromosome 17 (p12-13).⁴⁷ It is a homodimer composed of two polypeptide chains and has a single binding site for androgens and estrogens. Dimerization is a necessary step in the binding process.⁴⁸ The bound and free fractions appear to exist in a steady state of equilibrium. The amount of free fraction depends on the concentration of steroid hormone and on the levels and binding affinities of the binding proteins.

Of all the steroid hormones, DHT has the highest affinity for SHBG. Approximately 98% of testosterone circulates bound to SHBG (≈65%) and albumin (≈33%). Estradiol is primarily bound to albumin (≈60%) but also to SHBG (38%); about 2% circulates as the free fraction.⁴⁹ Progesterone, on the other hand, is mainly bound to albumin (≈80%) but also to transcortin (≈18%). Only approximately 0.6% of progesterone is bound to SHBG and about 2% exists in the free state.

The metabolic clearance of these steroids is inversely related to their binding affinity to SHBG. Thus, conditions that affect levels of SHBG (e.g., pregnancy, oral contraceptives) directly affect the levels of free hormone. Because estrogens increase SHBG synthesis and androgens decrease its synthesis, SHBG levels are twice as high in women compared to men. Several other hormones and other factors are known to influence SHBG levels. Thyroid hormones increase its synthesis and release by the liver.⁴⁶ Insulin, IGF-I, and prolactin have been shown to inhibit SHBG production in cultured cells.^50,⁵¹ Furthermore, serum concentrations of SHBG are increased in many disease states, including hyperthyroidism and liver cirrhosis. Certain drugs, including estrogen, tamoxifen, and phenytoin, can also increase serum SHBG concentrations. Carrier protein levels are decreased by hypothyroidism, obesity, and acromegaly and by administration of exogenous androgens, glucocorticoids, and growth hormones.

For many years, only the free fraction of testosterone was regarded as the biologically active component. However, researchers noted that steroid hormones bind with greater affinity to their specific carrier proteins and with much less affinity to albumin. In addition, studies of tissue delivery in vivo showed that the dissociation of albumin-bound testosterone can occur rapidly in a capillary bed so that the active fraction can be larger than the free fraction measured under equilibrium conditions in vitro.⁵² Thus, unconjugated steroids that are bound to albumin may be treated as free and biologically available.^53,⁵⁴

As mentioned above, SHBG levels can be influenced by numerous disease states. As such, changes in SHBG concentrations can result in large shifts in the free and SHBG-bound fractions. Hence, measurement of SHBG is of great clinical interest because it allows more accurate assessment of free hormones. SHBG is measured by a technique called saturation analysis, in which specific binding capacity of ³H-labeled testosterone is detected.^55,⁵⁶ With modifications, this method can also measure the non-SHBG bound fraction (bioavailable).⁵⁷ Recently, specific nonisotopic two-site immunoassays for SHBG have become available and are used in most clinical laboratories.

Measurement of Steroid Hormones in Circulation

The technique responsible for the accurate measurement of low concentrations of various steroid hormones and metabolites is competitive inhibition immunoassay or radioimmunoassay (RIA), which was originally described in 1960 by Yalow and Berson.⁵⁸ However, the development of steroid immunoassays presented several problems. First, they are not immunogenic and have a similar structure—they all have a same cyclopentahaptene nucleus with only minor structural variations—which makes it difficult to generate specific antibodies. Steroids can be made immunogenic via chemical coupling to a carrier protein known as hapten, and antibodies can be raised by immunization with haptens.⁵⁹ However, the site of the steroid where the protein is conjugated has a significant impact on the specificity of the resulting antibody.⁵⁹ Antibodies raised to conjugates of BSA coupled at the 19th position show higher specificity than those coupled at the 3rd or 17th positions.⁶⁰ For accurate clinical interpretation, it is important to know the cross-reactivity data on each antibody that is selected for a given assay. Most commercial assay reagent manufacturers provide cross-reactivity data, but it may not always be reliable and must be evaluated in the clinical laboratories performing the assay.⁶¹

Second, the high-affinity binding proteins such as SHBG in the serum compete with the antibody and thus interfere with the measurement of steroid molecules by RIA. This makes direct measurement difficult and necessitates a preassay extraction procedure with organic solvents and often a chromatographic separation of the steroid. Alternatively, the use of certain chemicals, such as 8-anilinonaphthalene sulfonic acid can inhibit the binding of steroids to proteins, which allows the direct measurement of steroid hormones without the extraction step. Direct assays are fast and can be automated. Several automated platforms for measuring estradiol, progesterone, and testosterone are commercially available and are used in most clinical laboratories. These assays, however, have a low sensitivity and, when used to measure very low concentrations, have poor reliability.⁶² Therefore, they may not be the best choice for clinical applications that require the ability to measure very low hormone concentrations such as estradiol measurement in children and in men (<100 pg/mL),⁶³ testosterone measurement in children and women(<1.5 ng/mL),⁶⁴ or progesterone measurement during ovarian stimulation (<1 ng/mL).⁶⁵ To overcome this problem, serum can be extracted with hexane-ethyl acetate (3:2 by volume), dried, and reconstituted in steroid-free serum, which can then be assayed in the automated platform.⁶⁶

Gas chromatography combined with mass spectrometry (GC-MS) addresses many of the shortcomings of immunoassays and is considered more reliable and accurate than immunoassays. However, this technique requires multiple steps, including solvent extraction, chromatographic fractionation, and chemical dramatization before instrumental analysis, and it is often less sensitive than some immunoassays. This technique has now been superseded by liquid chromatography combined with tamdem mass spectrometry. This newer technique has a higher sensitivity and throughput than GC-MS and is considered a reference methodology.^67,⁶⁸ The technique has been used to simultaneously measure estradiol and estrone in human plasma with no cross-reactivity.⁶⁹ These methods are recommended as reference methods and can be used to standardize and validate immunoassays, which provide the simplicity and rapid throughput needed for clinical use.

Estrogens

Measurement of estradiol is important in the assessment of female reproductive function. It can be used as an aid in the diagnosis of infertility and oligomenorrhea and to determine menopausal status (Table 2-2). In addition, measurement of estradiol is widely used to monitor ovulation induction and in vitro fertilization protocols.^70,⁷¹ Estrone levels are of limited clinical value in nonpregnant women because their levels closely parallel those of estradiol, except in the postmenopausal woman, in whom estrone becomes the main form of circulating estrogen. Also, as mentioned above, a specific RIA for the measurement of estrone sulfate has been described and made available commercially. Because estrone sulfate has a large circulating pool, it can serve as a marker of estrogenicity, especially in women on estrogen replacement therapy, in whom estradiol measurements are of little value due to the variable cross-reactivity of conjugated estrogens in the estradiol assays.^72,⁷³ In pregnant women, estriol is the main form of estrogen produced, and the amount of estrogen secreted increases from microgram quantities to milligram quantities.

Table 2-2 Assay Techniques and Clinical Applications of Ovarian Steroid Hormones

Hormone	Assay Techniques	Clinical Application
Estradiol	Radioimmunoassay Nonisotopic immunoassays

Evaluation of ovarian function

Assessment of menopausal status

Monitoring ovulation induction and in vitro fertilization cycle

Estrone sulfate Radioimmunoassay Marker of estrogenicity in women on hormone replacement therapy Progesterone

Radioimmunoassay

Nonisotopic immunoassays

Marker for ovulation

Detection of luteal phase defects

Marker for threatened abortion in early pregnancy

Total testosterone

Radioimmunoassay

Nonisotopic immunoassays

Evaluation of patients with hirsutism.

Male infertility

Free testosterone

Equilibrium-dialysis

Diafiltration

Evaluation of patients with hirsutism.

Polycystic ovarian syndrome

Progesterone

Although GC-MS has been recommended as the reference method for the measurement of progesterone,⁷⁴ immunoassays using steroid-specific antibodies are again the preferred mode of measurement in most clinical laboratories.^75,⁷⁶ Both RIA and nonisotopic immunoassays for progesterone are available for commercial use. Progesterone measurement is routinely used to detect ovulation and luteal phase defects (see Table 2-2).⁷⁷ In the follicular phase, progesterone levels are low (<5 nmol/L or 1.5 ng/mL). In the luteal phase, they range between 9 and 79 nmol/L (3 to 25 ng/mL). As noted, progesterone is required to maintain pregnancy, and progesterone measurement in early pregnancy can be valuable in the diagnosis of defects or threatened abortion.⁷⁸

Androgens

The measurement of androgens, including androstenedione and total testosterone or DHT, can also be accomplished using immunoassays.⁷⁹^–⁸¹ The one main drawback of these assays is the cross reactivity with other steroid hormones. Interference with cross-reactive testosterone in the androstenedione assay has been overcome by the use of specific testosterone antiserum to neutralize plasma testosterone.⁸¹ Direct measurement of androstenedione and testosterone without extraction is now possible with new commercial assays. However, commercial assays demonstrate high variability, which is greatest in samples from females.⁸² Total testosterone in hirsute women overlaps significantly with levels seen in normal women, and measurement of free testosterone correlates better with disease.⁸³ Free testosterone has been measured by equilibrium-dialysis, which is a time-consuming and difficult technique for most clinical laboratories to perform. An ultrafiltration technique can be used instead, which depends on MPS-1 centrifugal gel filtration devices and correlates well with the equilibrium-dialysis method^83,⁸⁴ as well as with GC-MS.⁸⁵ Direct, single-step, nonextraction immunoassay methods using ¹²⁵I-labeled testosterone analog have been developed commercially and are used in a number of laboratories. The accuracy and validity of this direct assay has been questioned.^86,⁸⁷ Alternatively, an indirect parameter of free testosterone—FAI—can be calculated as a ratio of testosterone to SHBG.⁸⁸ FAI is a better discriminator of hirsutism than either total testosterone or SHBG levels.⁸⁷^–⁸⁹ When bound with albumin and transcortin, testosterone dissociates more quickly than when bound with SHBG. This loosely bound teatosterone may be biologically available through dissociation during capillary transit. Cumming and Wall provided evidence for this hypothesis and suggested that this non-SHBG bound testosterone is a marker of hyperandrogenism.^57,⁹⁰

Saliva Measurements

SHBG is either undetectable or minimally present in saliva. Thus, this biologic fluid may reflect the free fraction of plasma steroids. Therefore, measurement of steroid hormones in saliva has attracted considerable attention.⁹¹^–⁹³ The ease of noninvasive collection combined with the simplicity of measurement makes salivary measurement a promising and attractive alternative to measurement of steroids in plasma. In the future, salivary assays may become useful adjuncts to those performed in plasma.⁹⁴^–⁹⁷

PEPTIDE HORMONES AND INTRAOVARIAN GROWTH FACTORS OF OVARY

The role of gonadotropins and gonadal steroids in the process of folliculogenesis in the ovary is well recognized. However, multiple other phenomena within the ovary suggest the presence of other intraovarian factors that may fine-tune the effects of gonadotropins and gonadal steroids. For example, the initiation and arrest of meiosis and the different rates of follicular growth leading to the selection of a dominant follicle point toward the existence of an intraovarian modulatory system. The concept of a gonadal factor with endocrine action at the pituitary level can be traced back more than 70 years to Mottram and Cramer.⁹⁸ The first moiety was identified and named inhibin for its inhibitory effect on the pituitary.⁹⁹ Since then, there has been an explosion of information regarding multiple and potential intraovarian regulators and their physiology, biochemistry, and biosynthesis as well as the identification of their receptors. Some of these compounds, which have been the subject of intense investigation, include peptide hormones/growth factors, cytokines, and neuropeptides. These factors may act in either an endocrine, autocrine, or paracrine fashion.

Peptide Hormones of the Ovary: Inhibins, Activins, and Follistatins

The first water-soluble peptide hormone in testicular extracts that exerted selective inhibitory activity at the pituitary level was described in 1932 and termed inhibin.⁹⁹ It was not until 1985 (almost 50 years later) that inhibin was isolated and characterized.^100,¹⁰¹ This was followed by the identification and characterization of two other related peptide factors (i.e., activin and follistatin).^102,¹⁰³ With the cloning of inhibin α and β subunit genes,¹⁰⁴ it was recognized that inhibin and activin belonged to the transforming growth factor-β (TGFβ) superfamily of growth and differentiation factors. Both inhibin and activin are of particular clinical interest and have been extensively reviewed^3,^105,¹⁰⁶ in recent years. They have been found to exert numerous different regulatory functions in a wide variety of both normal and neoplastic cells. Follistatin, a glycoprotein that is structurally distinct but functionally closely linked to inhibin and activin, is also discussed here. The site of production and areas of clinical interest of these peptide hormones are listed in Table 2-3.

Table 2-3 Role of Inhibins and Activins in Reproduction

Inhibins

The primary sources of inhibin production are the granulosa cells of the ovary and the Sertoli cells of the testis. Inhibin is also produced during pregnancy by the fetus, placenta, decidua, and fetal membranes. The main role of inhibin is to selectively suppress the production of FSH by the pituitary.^107,¹⁰⁸ This is achieved by modulating FSH biosynthesis through two main mechanisms: by reducing steady-state FSH mRNA in pituitary gonadotropes¹⁰⁹ and decreasing the stability of FSH mRNA.¹¹⁰

Inhibin is a 32 kDa glycoprotein heterodimer consisting of 2 subunits, α and β, linked by disulfide bonds.¹¹¹ There is a common α subunit but also two types of β subunits known as β_A or β_B. Thus, the two isoforms of inhibin are denoted inhibin A and inhibin B. As shown in Figure 2-5, each subunit derives from a separate precursor molecule called prepro-inhibin α (364 amino acid residues), prepro-inhibin β_A (424 amino acid residues) and prepro-inhibin β_B (407 amino acid residues). These are processed by proteolytic cleavage to yield the mature forms.¹⁰⁴ In addition to the fully mature forms (αβ dimers, Mr∼32,000), larger forms of dimeric inhibins with amino terminally extended α and/or β subunits have been identified in follicular fluid, which also possess FSH-suppressing bioactivity.^112,¹¹³ Furthermore, monomeric forms of both α and β subunits and certain fragments (αN and proαN-αC) generated during subunit processing are present in follicular fluid (see Fig. 2-1) and have intrinsic biologic activities distinct from classical inhibin-like bioactivity.^114,¹¹⁵

Figure 2-5 Activins, inhibins, follistatin, and TGFβ. A schematic representation of the formation of different dimeric proteins from three basic subunits, α, β_A, and β_B. Inhibins are heterodimers of α and β subunits (α-β_A, α-β_B); activins are homodimers of the β subunits (β_Aβ_A, β_Aβ_B, β_Bβ_B). Links between the units represent disulfide bridges.

Because the circulation contains multiple molecular forms of inhibin, it can be difficult to accurately measure inhibin levels using conventional RIAs. Most lack specificity for dimeric inhibin due to the variable cross-reactivity of monomeric forms and of various fragments with the antibody used in the assay.^112,^116,¹¹⁷ Also, conventional inhibin bioassays based on FSH suppression or release by cultured pituitary cells lack specificity due to the FSH-regulating activities of follistatin and activins. In addition, bioassays lack the sensitivity necessary to measure inhibin levels in the circulation. The development of two-site immunoassays utilizing αβ dimer specific antibodies overcame these problems and allowed specific measurement of the two forms of inhibin dimers (A and B).¹¹⁸ Use of these novel two-site assays allows measurement of inhibin levels throughout the normal menstrual cycle.¹¹⁹

Synthesis of the two isoforms of inhibin differ during the various phases of the menstrual cycle (see Fig. 2-4). Inhibin B levels are highest during the luteal–follicular transition and the early follicular phase, and studies on its presence in follicular fluid and basal granulosa cell secretion suggest that it is secreted by small developing antral follicles.¹²⁰ In contrast, inhibin A levels in the early and midfollicular phase reflect the sum of FSH- and LH-stimulated inhibin A secretion from all antral follicles. Levels of inhibin A during the late follicular phase mainly reflect secretion from the dominant follicle. Hence, inhibin A and B levels can be used as markers of follicular development. Inhibins have also been investigated as prognostic markers for women undergoing assisted reproductive technologies. In particular, it was suggested that measuring inhibin B levels during the early stages of FSH stimulation for ovulation induction could predict the number of oocytes retrieved and may be useful in monitoring ovarian stimulation for in vitro fertilization. However, because there is a large overlap between normal and subnormal ovarian responses in terms of inhibin B levels, it may be just as effective to obtain Day 3 FSH or perform a clomiphene challenge test.^121,¹²² Women in early perimenopause show significant decrease in inhibin B (no significant change in inhibin A and estradiol), which correlates with a mild increase in FSH levels. This perimenopausal decrease in inhibin B precedes a decrease in inhibin A, suggesting that inhibin B may serve as a sensitive marker for the onset of menopause.¹²³ Studies investigating the role of inhibins in the pathophysiology of polycystic ovary syndrome (PCOS) show conflicting results. Serum inhibin B levels in early follicular phase show significant increase in some studies or no change in others^124,¹²⁵—this awaits future studies to confirm its role in PCOS.

During pregnancy, inhibin A is produced primarily by the fetoplacental unit, whereas inhibin B levels remain low throughout the pregnancy.¹²⁶ Because there is a twofold elevation of circulating inhibin A levels in the second trimester of Down syndrome pregnancies, measurement of inhibin A is a clinically useful and important test.¹²⁷ When measurement of inhibin A level is added to alpha-fetoprotein, maternal age, and β-hCG (quad screen), the detection rate for Down syndrome increases from 53% to 75%.¹²⁸

Inhibin A has also recently been used as a tumor marker for ovarian sex cord tumors. This heterogeneous group of tumors accounts for 7% of all malignant primary ovarian neoplasms. They are composed of granulosa cells, thecal cells, Sertoli cells, Leydig cells, and other nonspecific stromal cells. It is important to distinguish this group of tumors from carcinomas and sarcomas because the former are low-grade tumors with a better prognosis. Inhibin A and its α-subunit have been found to be sensitive immunohistochemical markers of most ovarian sex cord-stromal tumors.¹²⁹ Inhibin B, however, has been found to be elevated in both sex cord-stromal and epithelial tumors and hence is of limited value in differentiating between the two entities. In addition, low levels of inhibin A in the cyst fluid of epithelial ovarian tumors has recently been reported to be associated with a worse prognosis.¹³⁰

Inhibin B has been found to be the predominant inhibin secreted in males and is produced by the Sertoli cells of the testes. It also has a negative feedback role on FSH from the pituitary, and its production is regulated by spermatogenesis. Inhibin B levels correlate with sperm count and testicular volume ¹³¹^–¹³³ but cannot distinguish between spermatid arrest and obstructive azoospermia—a condition in which sperm counts are normal.¹³⁴ As such, it is unlikely to replace testicular biopsy in the evaluation of male infertility.

Activins

Activins are made up of dimers of the inhibin β subunit (β_Aβ_A, β_Aβ_B, or β_Bβ_B) and have a molecular weight of approximately 25 kDa (see Fig. 2-5).¹³⁵ They are predominantly produced by the granulosa cells of the ovary. Activin/inhibin mRNA and protein have also been detected in extragonadal sources, including the placental trophoblast and decidua, testes, adrenal cortex, brain, spinal cord, and anterior pituitary. This implies that activin has diverse physiologic roles that are not confined to the reproductive system.

Acting either alone or with FSH, activin exerts an autocrine effect on granulosa cells to promote and maintain granulosa cell differentiation. It promotes FSH receptor expression on small undifferentiated granulosa cells ¹³⁶, enhances their response to FSH and LH, and hence increases aromatase activity and estrogen production.¹³⁷ This may explain how small preantral follicles progress from a gonadotropin-independent to a gonadotropin-dependent stage of development. After the granulosa cells acquire FSH receptors, further growth and differentiation of those cells to a preovulatory stage would be driven by activin acting in concert with FSH. Activin also inhibits both spontaneous and LH/hCG-induced increases in progesterone output by human follicles,¹³⁷ implying that it plays a role in delaying the onset of premature luteinization.

Activin also has a paracrine effect on thecal steroidogenesis by inhibiting thecal androgen output. It has been proposed that at earlier stages of follicular development, when the androgen requirements are low, thecal androgen synthesis is kept in check due to the relative excess of activin over inhibin and follistatin. However, as dominant follicles approach preovulatory status, increasing granulosa cell expression of inhibin and follistatin upregulates thecal androgen synthesis and ensures that the granulosa cells receive an adequate supply of aromatase substrate for conversion to estradiol. Activin stimulates pituitary FSH production, acting as a functional antagonist to inhibin (Fig. 2-6).¹⁰² It achieves this effect by increasing FSH mRNA synthesis as well as by increasing the stability of produced mRNA. Its actions are intimately modulated by intrapituitary concentrations of follistatin, which bind to activin to limit its bioavailability.

Figure 2-6 A diagrammatic presentation of the hypothalamic-pituitary-ovarian axis. Developed follicles secrete steroid hormones (estradiol and progesterone) and peptide hormones (inhibin, activin, and follistatin); all collectively control secretion of gonadotropins. Estradiol and progesterone, depending on concentration, have either positive or negative feedback and can alter the frequency and/or amplitudes of pulses at the level of both the hypothalamus and pituitary.

Free activin levels as measured by competitive protein binding assay, using follistatin as binding protein, show very little change over the menstrual cycle.¹³⁸ However, activin levels were elevated throughout the cycle in older versus younger women, suggesting that activin may play an endocrine role in maintaining FSH elevation in reproductive aging.¹³⁹ Lower levels of activin are detected in PCOS patients with a simultaneous increase in inhibins and follistatin, suggesting that an imbalance in these hormones may contribute to an abnormal LH/FSH ratio.^125,¹⁴⁰

Follistatin

In the ovary, follistatin is produced by the granulosa cells in antral follicles as well as by luteinized granulosa cells, which are under the positive regulation of FSH. It modulates the function of granulosa cells in favor of luteinization or atresia by neutralizing the effects of activin. It may also directly modulate progesterone metabolism by granulosa cells.¹⁴¹

Follistatin is a single-chain polypeptide (315 amino acids) that functions as the binding protein for activin, thereby neutralizing it. Most of its biology is explained by its antagonism with activin. Follistatin exists in two forms; as full-length follistatin (FS 315) in the circulation and as processed follistatin (FS 288) in follicular fluid ¹⁴² and the pituitary. It is part of an intrapituitary negative feedback loop where activin promotes FSH biosynthesis and the increased expression of follistatin limits its bioavailability for binding to the activin receptor on target cell membranes. It has been shown that the processed isoform of follistatin (FS 288) binds to cell-surface heparan sulfate proteoglycans with a higher affinity than FS 315.¹⁴³ Because proteoglycans are anchored to cell membranes, this suggests that this would limit follistatin diffusion from the site of release, leading to high local concentrations of follistatin being maintained. The membrane-anchored follistatin would be able to compete with activin receptors on nearby cells, thus modulating the biologic effects of activin. Once bound to activin, follistatin is able to accelerate the endocytotic internalization and lysosomal degradation of activin by pituitary cells.¹⁴⁴

Follistatin levels as measured by sensitive and specific two-site enzyme immunoassay are significantly higher in women with PCOS in comparison to controls. Higher follistatin levels combined with lower activin levels in these patients suggest its role in the lack of pre-ovular follicle development and FSH suppression.¹²⁵

Growth Factors/Cytokines As Intraovarian Regulators

Certain growth factors have been implicated in cellular communications within the ovary, including insulin-like growth factors IGF-I and IGF-II, epidermal growth factor (EGF), transforming growth factor-α (TGFα), basic fibroblast growth factor (bFGF), cytokines such as interleukins (IL-1 and IL-6), and tumor necrosis factor (TNFα). Growth factors and cytokines that are important as intraovarian regulators are outlined in Table 2-4.

Table 2-4 Autocrine and Paracrine Growth Factors and Cell–cell Signaling in the Ovary

Insulin-like Growth Factors

IGF-I and IGF-II promote cellular mitosis and differentiation in a variety of systems and play an important role in modulating folliculogenesis in an autocrine/paracrine fashion.¹⁴⁵^–¹⁴⁷ Insulin-like growth factors consist of two single-chain polypeptide growth factors that are structurally and functionally similar to proinsulin. The IGF autocrine/paracrine system includes IGFs, their specific receptors in target cells, and a family of IGF-binding proteins that regulate their bioavailability. Both IGF-I and IGF-II are produced in the ovary and augment the effects of the gonadotropins, although the main IGF in human follicles is IGF-II.¹⁴⁸ In small antral follicles, both IGF-I and IGF-II are expressed but restricted to thecal cells. However, IGF-I receptor mRNA has been detected only in granulosa cells.¹⁴⁶ It serves as an autocrine regulator in thecal cells and as a paracrine regulator in granulosa cells. In dominant follicles, however, no IGF-I mRNA has been detected either in thecal or in granulosa cells, and IGF-II expression is restricted to granulosa cells only. IGF-I receptors are present in granulosa cells only, and IGF-II receptors are expressed in both cell types.¹⁴⁶ Thus, in the dominant follicle, IGF-I functions mainly as a paracrine regulator^149,¹⁵⁰ and IGF-II acts as an autocrine factor. This suggests that IGF-II has an important role in coordinating differential follicular development within the ovary.

Epidermal Growth Factor (EGF), Transforming Growth Factor-α (TGFα), and Basic Fibroblast Growth Factor (bFGF)

A number of other peptide growth factors have been implicated as regulators of follicular development and steroidogenesis (see Table 2-4).¹⁵¹ These include EGF, TGFα, and bFGF. EGF is a single-chain polypeptide of 53 amino acids with three disulfide bonds and has mitogenic effects in a variety of ectodermal and mesodermal tissues. TGFα is a 50-amino acid peptide with 30% to 40% homology with EGF. The EGF receptor is a 170-kDa glycoprotein with tyrosine kinase activity. TGFα binds to EGF receptors with the same affinity as EGF.

The presence of EGF/TGFα and bFGF as well as their receptors in the ovary has been shown both at the protein and mRNA levels. The presence of immunoreactive EGF as well as EGF receptors has been identified in the preovulatory follicles ¹⁵² and corpus luteum.^153,¹⁵⁴ Furthermore, studies have shown the presence of TGFα¹⁵⁵ and bFGF¹⁵⁶ mRNAs in follicular cells; studies have also shown that the TGFα message is upregulated by FSH in vivo. FSH plus TGFα or FSH plus EGF resulted in significantly elevated progesterone and 20α-hydroxyprogesterone levels in granulosa cells in culture.¹⁵⁷ The presence of TGFα message in cultured granulosa cells and the fact that it mediates its action via binding to EGF receptor all point to its autocrine role in granulosa cell differentiation, follicle development, and selection. Expression of bFGF and its receptor mRNA has been detected in fetal ovaries and in granulosa cells.^158,¹⁵⁹ bFGF has been shown to be mitogenic for granulosa cells and to cause an inhibitory action on granulosa cell differentiation and thecal cell steroidogenesis.^160,¹⁶¹ It also has potent angiogenic activity.¹⁶²

Cytokines

Cytokines primarily produced by white blood cells modulate various cellular functions. The cytokines in the ovary are secreted both by the immune cells that are recruited from the circulation in the ovarian stroma and by the thecal and granulosa cells. A number of cytokines have been linked to modulation of ovarian function, including interleukins IL-1 and IL-6 and TNFα. Both IL-1 and IL-6 have been found in significant amounts in follicular fluid.^163,¹⁶⁴ Granulosa cells accounted for the majority of immunostaining for IL-1 and IL-6 in follicular aspirates, which suggests that these cytokines are produced in granulosa cells^164,¹⁶⁵ and that they affect granulosa function.¹⁶⁶ During folliculogenesis, IL-I promotes proliferation and suppresses differentiation. In the ovulatory process, it promotes ovulation by increasing production of chemokines, steroids, ecosanoids, and vasoactive substances.¹⁶⁷ IL-6 demonstrates inhibitory effects on both estradiol and progesterone secretion by FSH-stimulated granulosa cells.¹⁶⁸ Elevated IL-6 levels during genital infections may provide a possible link to reproductive dysfunction. TNFα expression has also been detected in granulosa cells of human antral and atretic follicles by immunohistochemistry.¹⁶⁹ Also, in vitro treatment with TNFα enhanced steroidogenesis in both healthy and atretic follicles,¹⁷⁰ suggesting that TNFα has a paracrine and/or autocrine role. Nonetheless, the physiologic implications of these actions remain unclear and require further investigation.

Neuropeptides

Some evidence suggests that an independent ovarian-central nervous system axis exists.^171,¹⁷² Electrical stimulation of the hypothalamus in hypophysectomized and adrenalectomized rats produced a change in ovarian steroidal synthesis that was independent of changes in ovarian blood flow.¹⁷³ In addition, murine thecal cells can produce androgens under adrenergic stimulation.¹⁷⁴ Adrenergic innervation of the ovary acts primarily on the thecal-interstitial cells through β₂ receptors, synergizing with the effects of gonadotropins in the production of ovarian androgens.¹⁷⁴ This may in turn play a role in the regulation of estrogen production by granulosa cells, thereby influencing follicle recruitment and selection.

REGULATION OF OVARIAN HORMONES

The hypothalamic-pituitary axis plays a key role in the regulation of hormonal synthesis by the ovaries. The hypothalamus is connected to the pituitary gland via a portal vascular system that allows transport of hypothalamic releasing factors from the brain to the pituitary (see Fig. 2-6). The hypothalamus, being the coordinating center, provides precise signals via pusatile release of GnRH to the gonadotrophs, which in turn secrete LH and FSH. Any interruption in this connection results in low gonadotropin levels, leading to failure of ovarian hormone secretion.

Hypothalamic Regulation

Gonadotropin-releasing Hormone

The gonadotropin-releasing activity in hypothalamic extracts was first demonstrated in the late 1950s. In 1971, almost 15 years later, GnRH was first isolated and characterized from a hypothalamic extract. This was followed by its synthesis and its clinical availability. However, early clinical studies proved disappointing until the nature of its pulsatile secretion was recognized. In their classic experiment in the rhesus monkey, Knobil and colleagues showed that normal LH and FSH release required the pulsatile infusion of GnRH at approximately 60-minute intervals ¹⁷⁵ Furthermore, a change in pulse frequency (lesser or greater) or continuous infusion resulted in failure of LH and FSH secretion and release (Fig. 2-7).¹⁷⁶ These observations were confirmed in humans. Pulsatile GnRH infusion reproduces appropriate patterns of hormonal changes, resulting in ovulation and fertility in women with hypothalamic amenorrhea.

Figure 2-7 Change in luteinizing hormone (LH) and follicle-stimulating hormone (FSH) levels in gonadotropin-releasing hormone (GnRH) deficient monkeys, according to the mode of delivery of GnRH. A continuous infusion decreased both LH and FSH levels, which are restored by pulsatile GnRH administration.

(Data from Belchetz PE, et al: Hypophysial responses to continuous and intermittent delivery of hypopthalamic gonadotropin-releasing hormone. Science 202:631–633, 1978.)

Physiologic Role of GnRH

GnRH is produced by secretory neurons located in the arcuate nucleus of the medial basal hypothalamus and the preoptic area of the anterior hypothalamus. The nerve terminals are found in the lateral portions of the external layer of the median eminence adjacent to the pituitary stalk.¹⁷⁷ GnRH has an intrinsically pulsatile pattern of secretion, which is under the control of a hypothalamic pulse generator in the arcuate nucleus.^178,¹⁷⁹ The frequency and amplitude of the pulsatile rhythm of GnRH secretion are crucial in regulating gonadotropin secretion and hence gonadal activity.^180,¹⁸¹ Physiologic frequency (approximately hourly pulses) tends to upregulate GnRH receptors, enhancing pituitary responsiveness to subsequent stimulation by GnRH. This leads to a “self-priming” effect, whereby LH levels have been shown to increase sequentially with sequential GnRH pulses. A longer frequency causes anovulation and amenorrhea; a shorter frequency or constant exposure to GnRH downregulates the GnRH receptors, inducing refractory gonadotropin responses.^176,¹⁸²^–¹⁸⁴

The “pulse generator” is subject to modification by two main inputs: (1) hormone-mediated signals and (2) neural signals. Hormone signals include negative and positive feedback from the gonadal steroids (e.g., estrogen and progesterone) as well as gonadal protein hormones. Neural signals may come from a wide variety of sources and are mediated by neurotransmitters, including acetylcholine, catecholamines, serotonin, opioids, and γ-aminobutyric acid.¹⁸⁵ Norepinephrine is believed to stimulate GnRH release whereas opioids exert inhibitory effects. Dopamine can produce both inhibitory and excitatory GnRH responses, depending on the physiologic state.¹⁸⁶

Biochemistry and Biosynthesis

GnRH is a linear decapeptide, derived from the post-translational processing of a large precursor molecule, prepro-GnRH. The prepro-GnRH molecule consists of 92 amino acids in a tripartite structure (Fig. 2-8). It begins with a signal peptide of 23 amino acids, which is followed by the decapeptide and then a Gly-Lys-Arg sequence needed for proteolytic processing and C-terminal amidation of GnRH molecules. The last 56 amino acid residues are collectively known as the GnRH-associated peptide (GAP), which may have prolactin-inhibiting activity.^187,¹⁸⁸ Knowledge of GnRH structure led to the development of many clinically important long-acting GnRH agonists, including buserelin, leuprolide, and nafareline (see Fig. 2-8).

Figure 2-8 The structures of native gonadotropin-releasing hormone (GnRH) and some analogs currently in clinical use.

Gonadotropin-releasing hormone is encoded from a single gene on the short arm of chromosome 8p21-p11. The human gene contains 4 exons; exon 2 encodes pro-GnRH, exon 3 and part of exon 2 and 4 encode the GAP protein, and exon 4 encodes a long 3′ untranslated region. Molecular processing occurs primarily in the nucleus of the cell body (soma). After transcription, the mRNA is transported to the cytoplasm where translation takes place and it is converted into the decapeptide. GnRH and its cleavage products, GAP and pro-GnRH, are then transported to the nerve terminals where they are secreted in tandem into the portal circulation.^187,^189,¹⁹⁰

Measurement

GnRH has a short half-life (2–4 min) due to its degradation by peptidases in the hypothalamus and pituitary gland. These peptidases cleave the molecule at the Gly-Leu bond and at position 10. Due to the large dilutional effect, peripheral GnRH serum levels are too low for pulse characteristics to be determined reliably in humans. However, the simultaneous measurement of circulating levels of LH in hypophysial-portal and peripheral blood samples in other species has been shown to correlate closely with GnRH release.^191,¹⁹² As such, frequent measurements of LH pulses can be used as an accurate indicator of GnRH secretion patterns. FSH also correlates with GnRH secretion, but is less useful due to its long half-life.

Regulation by Pituitary Hormones

As mentioned above, GnRH action on gonadotrophs stimulates gonadotropin production and release in a pulsatile fashion. In addition, LH and FSH release from the pituitary is also affected in both a positive and negative manner by estrogen and progesterone as well as by the protein hormones secretion by the ovaries (see Fig. 2-6). The positive effect of estrogen and the negative effect of progesterone on the gonadotropins depend on the level of steroid hormone and the duration of exposure to the gonadotrophs. On the other hand, both LH and FSH are required for ovarian estrogen synthesis and the level of estrogen production depends on the time of exposure and the level of gonadotropins.^193,¹⁹⁴ Nonetheless, a disordered signal from the pituitary gland may result in infrequent ovulation (oligo-ovulation) or absent ovulation (anovulation).

Regulation by Luteinizing Hormone and Follicle-stimulating Hormone

Physiologic Role

As discussed above, LH regulates ovarian steroidogenesis. The surge of LH in the middle of the menstrual cycle is also responsible for inducing ovulation. The surge occurs as a result of a dramatic rise in estradiol produced by the preovulatory follicle, which produces a positive feedback on LH. The midcycle surge stimulates the resumption of meiosis and the completion of reduction division in the oocyte with release of the first polar body. Proteolytic enzymes and prostaglandins are increased in response to LH, leading to a release of the oocyte from the ovary.¹⁹⁵ Finally, the continued secretion of LH after ovulation converts the remaining follicular cells in the ovary to the corpus luteum and stimulates the corpus luteum to produce progesterone by enhancing the conversion of cholesterol to pregnenolone.

Follicle-stimulating hormone regulates ovarian estrogen synthesis by binding to the FSH receptor on the surface of the granulosa cell and is required for follicular maturation and growth.^196,¹⁹⁷ This results in elevated cyclic adenosine monophosphate (cAMP) levels and the induction of aromatase, which converts androstenedione from the neighboring thecal cells to estrone. FSH also induces expression of 17βHSD type 1, which converts estrone to estradiol. Increased secretion of estradiol leads to further proliferation of granulosa cells and follicular growth and an increase in the number of estradiol receptors.^196,^198,¹⁹⁹ In the mature follicle, FSH and estradiol increase the LH receptors’ expression in granulosa cells, making these cells responsive to LH and augmenting progesterone secretion. Progesterone then increases FSH release in midcycle.

Biochemistry and Biosynthesis

The anterior pituitary produces three glycoprotein hormones: LH, FSH, and thyrotropin-stimulating hormone (TSH), all of which share a similar biochemical structure. Each consists of a heterodimer of two noncovalently linked protein subunits, αβ. Each subunit is cysteine-rich and contains multiple disulfide linkages (Fig. 2-9). There are also multiple carbohydrate moieties on both subunits that are important in the metabolism and biologic activity of the hormones.^200,²⁰¹ The α subunit is common to all three hormones, whereas the β subunit is unique.

Figure 2-9 A schematic presentation of the gonadotropin subunits, showing the sizes, locations of carbohydrate side chains, and currently known mutations and polymorphisms: The symbols Y and O represent the locations of the N-linked and O-linked carbohydrate side chains, respectively. The symbol S represents disulfide bonds. The arrows indicate the positions of point mutations/polymorphisms. The numbers below the right end of the bars indicate the number of amino acids in the mature subunit protein. Cα, common α subunit; LHβ, luteinizing hormone β subunit; FSHβ, follicle-stimulating hormone β subunit; HCGβ, human chorionic gonadotropin β subunit.

(Adapted from Huhtaniemi I: Functional consequences of mutations and polymorphisms in gonadotropin and gonadotropin resistant genes. In Leung PCK, Adashi E (eds): The Ovary, 2nd ed. London, Elsevier Academic Press, 2004, p 56.)

α Subunit

The human α subunit gene is located on chromosome 6p21.1-23 and is composed of 4 exons. The first exon is noncoding. The gene encodes a 14-kDa polypeptide that consists of a 24-amino acid signal peptide and the mature α subunit of 92 amino acids with 10 cysteine residues and two N-linked oligosaccharide groups.²⁰² The cysteine residues participate in the intrasubunit disulfide linkages. The α subunit is more abundant than the β subunit, and unassociated, or “free,” α subunits are present in the serum and pituitary. They have little known biologic activity. Hence, only the αβ heterodimer possesses biologic activity.

β Subunit

The β subunits for LH and FSH are encoded by separate genes and are located on different chromosomes. The gene coding for the LHβ subunit consists of three exons and is present in a complex gene cluster on human chromosome 19q13.3.²⁰³ The cluster includes six chorionic gonadotropin β (CGβ) genes that are presumably derived from single ancestral LHβ gene by gene duplication.²⁰⁴ Both LHβ and CGβ proteins are structurally and functionally similar. They are approximately 80% homologous in amino acid sequence. Both propeptides contain a 20-amino acid signal sequence. The mature LHβ and CGβ subunits consist of 121 and 145 amino acids, respectively. The major difference between LHβ and CGβ proteins is the presence of a 24-amino acid C-terminal peptide in CGβ, which is heavily glycosylated, with four O-linked carbohydrate moieties (Fig. 2-10).

Figure 2-10 Basic structure and general mechanism of action for cytoplasmic/nuclear steroid hormone receptors. A, Basic structure of steroid hormone receptor. B, Mechanism of action involving steps 1–6 as described in the text. HSP90, heat shock protein; HRE, hormone response element; CA, coactivators.

The gene for the FSHβ subunit is located on the short arm of chromosome 11p13 and, like the gene for LH, it consists of three exons.²⁰⁵ The molecular size of FSH is 33 kDa, consisting of a signal peptide of 18 amino acids and the mature FSHβ protein of 111 amino acids. Like α subunits, both LHβ and FSHβ subunits contain 10 cysteines for disulfide formation. Compared to FSHβ and CGβ subunits, LH does not have a terminal sialic acid on its carbohydrate side chain. This results in a shorter metabolic clearance time for LH, compared to FSH and HCGβ subunits. HCGβ has the highest content of sialic acid and has the longest half-life.^206,²⁰⁷ Deglycosylation of gonadotropins has no effect on receptor binding but abolishes signal transduction.²⁰⁸

Mutations in Gonadotropin Genes

Although rare, mutations in gonadotropin genes can result in clinical disorders, which have been described in the literature. In fact, the subject has been recently reviewed extensively.^201,²⁰⁹ There are several reports of neutral polymorphisms, but no activating mutations have been reported so far in the α subunit gene.²⁰⁹

Examples of mutations in the LHβ subunit include a single amino acid substitution (Glu to Arg) at codon 54 ²¹⁰ that is associated with hypogonadism in homozygous males and with a high incidence of infertility in heterozygotes. In addition two point mutations—Trp to Arg at codon 8 and Ile to Thr at codon 15—have been described in five women with immunologically anomalous LH but with hyperbioactivity.²¹¹ However the direct relationship of these mutations with a specific pathogenesis remained elusive. The other variant of LHβ (polymorphism) with substitution of Serine to Glycine at codon 102 in exon 3 has been found to be population-specific in 4% of women in Singapore with menstrual disorders.

Several inactivating mutations have been described in the FSHβ subunit gene (see Fig. 2-9). The first mutation reported was a homozygous 2bp deletion in a codon 61 (Valine) in a woman who presented with delayed puberty, amenorrhea, and infertility.²¹² The mutation was a nonsense mutation leading to altered amino acid sequence after codon 60, leading to a stop codon at residue 87. The mutated protein was altered and was thus unable to associate with the α subunit and was nonfunctional. The patient was treated with exogenous FSH, leading to follicle maturation and pregnancy. Another case report on FSH mutation with a similar phenotype was found to be heterozygote for two FSH mutations: the first was a nonsense mutation at codon 61 and the second was a substitution of Cysteine to Glycine at codon 51. The loss of the cysteine residue was likely responsible for the altered conformation change and loss of function leading to the phenotype.²¹³ An inactivating mutation leading to the substitution of Cysteine to Arginine residue at codon 82 was described in an infertile man.²¹⁴

Measurement of LH and FSH

It has been difficult to develop highly specific immunoassays for the gonadotropins due to the high homology of the glycoprotein hormones as well as the need to distinguish between free α subunits and intact hormones. Cross-reactivity with the α subunit has made it difficult to accurately measure LH and FSH with RIAs that use polyclonal antibodies.²¹⁵^–²¹⁷ In addition, there is microheterogeneity in both pituitary and circulating gonadotropins due to the degree of glycosylation. Variation in oligosaccharide content and structure also causes charge microheterogeneity, resulting in serum isoforms that separate over the pH range of 6.5 to 10 on electrofocusing.²¹⁸ Although the clinical significance of this microheterogeneity of circulating isoforms remains unknown, it affects immunoreactivity as detected by various antibodies, all of which lead to variability among different immunoassays. Microheterogeneity also affects the biologic activity and may be primarily responsible for the discrepant results generated by immunoassays versus bioassays.

The other problem stems from a lack of a suitable reference preparation for both immunoassays and bioassays. The international reference preparation (IRP) of purified human urinary menopausal gonadotropins (2nd IRP-hMG) was established by the World Health Organization (WHO) and has been widely used in most immunoassays and bioassays. The unitage assigned to IRP was defined by bioassays and formed the basis of all subsequent purified pituitary preparations as, for example, the WHO international standard (2nd IS). Most commercial assays are calibrated against 2nd IS. Although the biologic potency estimates are given for the IRP or IS, the immunoreactivity varies in different immunoassays depending on the antibody specificity.

In recent years, two-site directed immunoradiometric (IRMA) or immunochemiluminometric (ICMA) assays have been developed and are based on the use of two monoclonal antibodies. These have helped overcome most limitations of RIAs.²¹⁹ These assays are automated and show exquisite sensitivity approaching 0.1 mIU/mL. Furthermore, these assays are not affected by the presence of free α subunits and correlate better with bioassays. The high sensitivity of these assays also allows detection of low levels seen in early puberty.

Luteinizing hormone bioassays utilize dispersed mouse or rat Leydig cells or a Leydig cell tumor cell line (MA-10) in culture ^220,²²¹ and measure testosterone production in vitro. These assays measure the biologic activity of circulating LH under physiologic conditions. This is important because the biologic activity of LH changes with alterations in glycosylation and the tertiary structure of the molecule. The combination of bioassay and RIA permits the calculation of bioactive/immunoreactive ratios, which can provide a useful index of qualitative changes of the LH molecule.^219,²²² Although the bioactive and immunoactive LH profiles are generally well correlated during physiologic changes, significant discrepancies can occur in some pathologic states. For example, inactivating mutations in the LHβ gene result in elevated levels of immunoreactive LH but a marked loss of bioactivity. In vitro bioassays for FSH activity using rat granulosa cells or Sertoli cells measure production of cAMP or aromatase activity in response to FSH. The sensitivity of these assays is about 2.5 mIU/mL. Alhough in vitro bioassays have been valuable in elucidating the physiology, they remain cumbersome and time consuming and are not practical for routine clinical use.

Measurement of LH and FSH is useful in the diagnosis of gonadal function disorders (Table 2-5). Elevated levels of FSH generally indicate ovarian failure but may be seen in some patients with viable ovarian follicles.²²³ Although rare, high gonadotropin levels associated with gonadotropin-secreting pituitary tumors or ectopic gonadotropin-producing tumors. In an amenorrheic patient, an elevated LH level with normal FSH and LH-to-FSH ratio typically (but not invariably) of greater than 2 is suggestive of PCOS. Low levels of these hormones are indicative of pituitary or hypothalamic dysfunction and occur together with low serum estradiol levels. For further assessment of pituitary reserve, provocative GnRH testing is required. LH and FSH responses to intravenous injection of 100 μg GnRH are measured at 20 and 60 minutes. Lack of response may suggest the likely diagnosis of hypogonadotropic hypogonadism; however, its sensitivity and specificity are low in patients receiving exogenous sex steroids.²²⁴

Table 2-5 Role of Pituitary Hormones in Assessment of Female Infertility

Hormone	Hormone Levels	Interpretation
Prolactin	↑Prolactin	Evaluate for prolactinoma after excluding hypothyroidism, pregnancy, macroprolactinemia as cause.
LH and FSH	↓ LH, ↓ FSH	Hypothalamic or pituitary disease
	↑ LH, ↑ FSH	Premature ovarian failure
	↑ LH, ↓ or normal FSH	Polycystic ovary syndrome

Prolactin

Physiologic Role of Prolactin

The main role of prolactin is the stimulation of lactation in the postpartum period. Its effect is blunted by estrogen and progesterone; the decrease in both these hormones after parturition allows lactation to occur. Prolactin also acts at the hypothalamus to suppress gonadotropin production, primarily by inhibiting the pulsatile secretion of GnRH. Thus, conditions associated with hyperprolactinemia can cause hypogonadism, which leads to shortened luteal phases, decreased numbers of granulosa cells, anovulation, and amenorrhea in women. In men, excess prolactin can lead to decreased spermatogenesis, decreased libido, impotence, and infertility.

Biochemistry and Biosynthesis

Prolactin is a single polypeptide with a molecular weight of 22 kDa. It consists of 198 amino acids and is folded into a globular shape connected by disulfide bonds. It is remarkably homologous to human growth hormone (hGH) and human placental lactogen (hPL). The gene for prolactin is found on chromosome 6, and seems to have been evolutionarily derived from a common somatomammotropic (hGH-hPRL-hPL) precursor.

It is produced by lactotrophs in the pituitary, which make up almost 50% of the total pituitary cell population. Its production is under tonic inhibition by dopamine, produced by the tuberoinfundibular cells, and the hypothalamic tuberohypophyseal dopaminergic system. Prolactin is extremely heterogeneous and exists in at least four different molecular forms ²²⁵^–²²⁷: (1) little prolactin, molecular weight (MW) 23 kDa, a nonglycosylated monomeric hormone with high receptor binding and bioactivity; (2) G, or glycosylated prolactin, MW 25 kDa, which has reduced immunoreactivity; (3) big prolactin, MW 50 kDa, consisting of a mixture of both dimeric and trimeric forms of G prolactin; and (4) big-big prolactin, MW 100 kDa, consisting of G prolactin covalently coupled with an immunoglobulin, also known as macroprolactin. The big and big-big forms have lower receptor-binding affinity, but may be converted to little prolactin by reduction of the disulfide bonds. As such, discrepancies can exist between measured prolactin levels and the clinical effects.

Measurement

Prolactin levels are measured by use of IRMA and ICMA. These methods give excellent reproducibility, sensitivity, and assay efficiency; however, they vary in their abilty to react with biologically inactive macroprolactin. Hence, a measured serum immunoreactive prolactin level often does not correlate with expected clinical effects. A polyethylene glycol precipitation method should be used to detect the macroprolactinemia.²²⁸ The other caveat is that these samples are usually assayed at a single dilution. As such, extremely high levels of prolactin (≈1000 ng/mL), such as those seen in macroprolactinomas, may saturate both capture and localizing antibodies, leading to a falsely low value in some one-step sandwich immunoassays. This has been known as the hook effect. Thus, in patients with macroadenomas, a 1:100 serial dilution should be performed if the assay is prone to this hook effect.

MECHANISMS OF ACTION OF HORMONES

Ovarian/pituitary hormones circulate in very low concentrations in the extracellular fluid, generally in the range of 10⁻¹⁵ to 10⁻⁹ mol/L. To exert their biologic effects on the target cells, special recognition mechanisms are required. Target cells are able to discriminate between the different hormones at low concentrations by depending on cell-associated recognition molecules called receptors.

Hormones exert their biologic effects by interacting with these high-affinity receptors, which in turn trigger one or more effector systems within the cell. The high affinity, specificity, and receptor expression level together define the nature and degree of biologic response of a hormone. All receptors have at least two functional domains, a recognition domain and a signal-generating domain. The recognition domain binds to the hormone, and the second domain generates a signal that couples hormone recognition to some intracellular function. This coupling of hormone binding to signal transduction, or receptor–effector coupling, provides the first step in the amplification of a hormonal response and distinguishes the target cell receptor from the plasma carrier proteins that bind hormone but do not generate a signal.

Just on the basis of the location of the hormone receptor (i.e., intracellular/nuclear or cell surface), two distinct mechanisms of hormone actions can be classified. These mechanisms further differ by the nature of the signal transduction pathway or second messenger responsible for mediating hormone action (Table 2-6). Examples of nuclear receptors include steroid hormones that are lipophilic and pass through the cell membrane to interact with receptors located either within the cytoplasm or the nucleus. This in turn affects gene transcription within the nuclear compartment. Polypeptide hormones (i.e., LH, FSH, hCG, GnRH, inhibins, and activins) and growth factors that are hydrophilic interact with cell-surface receptors that are located on the plasma membrane. They trigger a plethora of signaling activity in the membrane and cytoplasmic compartments as well as exerting parallel effects on the transcriptional apparatus in the nuclear compartment. These cell surface receptors can be further classified based on the second messenger into four major subgroups, as listed in Table 2-6.

Table 2-6 Classification of Receptors for Steroid and Peptide Hormones

Hormones that Bind to Intracellular/Nuclear Receptors	Hormones that Bind to Cell Surface Receptors
Type 1 receptor, classical steroid hormone receptor Requires ligand binding for activation Homodimerization pattern Estrogen receptor Progesterone receptor Androgen receptor Glucocorticoid receptor Mineralocorticoid receptor

Type 2 receptors

Able to bind DNA in the absence of ligand, exerting repressive effect

Heterodimerization pattern

Thyroid Hormone

Retinoic Acid Receptor

Steroid Hormone Action

Most of the early knowledge about steroid hormone mechanism of action has been derived from in vitro and in vivo binding studies that used radiolabeled estradiol as a ligand.²³¹ The role of ERs in the regulation of breast cancer growth has been well studied over the past four decades; it was the first steroid receptor to be discovered in the early 1960s. This led to the subsequent elucidation of a general pathway for steroid hormone action. In general, nuclear receptors share a common protein architecture consisting of five functional domains²³² as illustrated in Figure 2-10A.

Amino-Terminal Transactivation Domain (A/B Region)

This is the most variable domain in terms of sequence and length in the family of nuclear receptors. It can range in size from 20 amino acids in the vitamin D₃ receptor to 600 amino acids in the mineralocorticoid receptor. It usually contains a transcription activation function (TAF-1), which interacts with other core transcriptional machinery (e.g., coactivators) to activate target gene transcription.

Central DNA-binding Domain (C Region)

This region is essential for activation of transcription. It encodes two zinc finger motifs and has a very high degree of homology among all types of nuclear receptors. Hormone binding to the receptor induces specific conformational changes in this region, allowing receptor binding to the hormone-responsive element (HRE) of the target gene. The amino acid sequence that lies between the first and second zinc fingers (i.e., the recognition helix) is responsible for establishing specific contact with the DNA. The second zinc finger stabilizes the contact and increases the affinity of the receptor for the DNA.

Hinge Region (D Region)

As the designation hinge indicates, this region is the site of rotation that allows the protein to alter its conformation after ligand binding. It provides the localization signal that is important for moving the recptor to the nucleus in the absence of ligand and contains a nuclear localization domain (glucocorticoid and progesterone receptors) and/or a transactivation domain (thyroid hormone or glucocorticoid receptors).

Carboxy-Terminal Ligand-Binding Domain (E Region)

This region is responsible for binding of the relevant ligand and is responsible for receptor dimerization or heterodimerization. It also contains the binding site for heat shock proteins and has a transactivation function (TAF-2) that can drive transcriptional activity. Unlike TAF-1, its transcriptional activity depends on hormone binding. Conformational change that occurs after ligand binding is responsible for interaction with coactivators or corepressors.

Cellular Mechanism of Action of Steroid Hormones

The steroid hormone nuclear receptors are known as ligand-dependent transcription factors. Binding with their ligands is a necessary step for their function as transcriptional regulators. Unliganded receptors may be localized to either the cytoplasm (e.g., glucocorticoid receptor) or the nucleus (e.g., estrogen, progesterone, and thyroid hormone receptors). For most steroid hormones the unliganded receptors exist in the cell nucleus as large molecular weight oligomers (≈300K; 7-10S sedimentation rate)²³³ and can be isolated in cytosolic fraction from cells or tissues disrupted in hypotonic media. The oligomers are formed by noncovalent association of a monomeric receptor protein with a dimer of heat shock protein (HSP90, HSP70, or HSP56).²³⁴

The general features of the mechanism of action of these hormones are depicted in Figure 2-10B. Steroid hormones that freely diffuse through the cell membrane bind to the specific receptors in the nucleus. Ligand binding to receptor initiates the receptor transformation, or so-called activation process. During this process the receptor undergoes conformational changes that primarily occur as a result of its dissociation from HSP, which exposes the DNA-binding site. Nuclear translocation and dimerization of the activated receptor then occurs. Most evidence suggests that this process is thermodynamically irreversible. The hormone receptor complex then binds to a specific region of DNA, the HRE, which is located upstream of the gene. The first HRE was identified for the glucocorticoid receptor. Later the HREs for the progesterone, androgen, estrogen, and mineralocorticoid receptors were shown to be similar to that of the glucocorticoid receptor.²³⁵^–²³⁷ The steroid HREs in the target genes are a palindromic (inverted-repeat) DNA sequence of 15 base pairs (Table 2-7). This interaction leads to the recruitment of a host of ancillary factors known as coregulators (coactivators or corepressors), creating a transcriptionally permissive or nonpermissive environment at the promoter, as well as communicating with other general transcription factors and RNA polymerase II. Coactivators function as adaptors in a signal transduction pathway. The binding of these coregulators modulates the resulting transcription (i.e., the activation and inactivation of specific genes). Hormone antagonists, for example, induce a different conformation in the TAF-2 that hinders the coactivator-binding site and recruits a corepressor instead, and inhibits gene expression. The availability of these coregulators in different tissues plays an important role in defining the biologic response to both steroid hormone agonists and antagonists.²³⁸

Table 2-7 Sequence of DNA Recognition Elements for Steroid Hormone Receptors

Steroid Hormone Receptor	Element	DNA Recognition Sequence
Estrogen receptor	ERE	AGGTCAsssTGACCT

Progesterone receptor

Androgen receptor

Glucocorticoid receptor

Mineralocorticoid receptor

HRE

AGAACAsssTGTTCT

Sequences read 5′-3′ direction indicated by arrows. S indicates spacer nucleotides (A, G, C, or T).

The biologic activity of a hormone is determined by at least four factors intimately related to the structure of the hormone receptor in question. The first factor is the affinity of the hormone for the hormone-binding domain of the receptor. The second factor is the differential expression of receptor subtypes in the target tissue, altering the response to the same hormone. The third factor is the conformational shape of the ligand–receptor complex and its consequent effects on dimerization and the modulation of adaptor proteins. The last factor is the differential expression of target tissue adaptor proteins and phosphorylation. A higher concentration of coactivators or corepressors in the target tissue can affect the cellular response of that tissue to the same ligand. Phosphorylation of the receptor by protein kinases increases the transcriptional activity of the receptor.

The Estrogen Receptor

Structure and Function

The structure of ER (now known as ERα) was reported in 1986.²³⁹ It consists of five components or domains that are divided into six regions, referred to as A-F (Fig. 2-11), instead of the five regions seen in most steroid receptors. The F region is a C-terminal segment of 42 amino acids that influences the conformational changes that occur after estrogen/antiestrogen binding. Thus, it modulates the level of transcriptional activities, most likely by affecting the interaction with coregulator proteins. It has a molecular weight of 66,000 and contains 595 amino acids. ER mRNA is 6.8 kilobases and contains 8 exons derived from a gene located on the long arm of chromosome 6. More recently a second form of ER has been discovered and named ERβ; it is encoded by a gene located on chromosome 14²⁴⁰ and is in close proximity to the genes that are related to Alzheimer’s disease.²⁴¹

Figure 2-11 A schematic diagram of two estrogen receptor (ER) isoforms. Different domains (A-F) and their corresponding functions are illustrated. ER-α, estrogen receptor α; ER-β, estrogen receptor β.

The two receptors show a high degree of homology in the DNA-binding domain (97%) and ligand-binding domain (59%) but less so in hinge (30%), regulatory (17%), and F regions (17.9%)^241,²⁴² (see Fig. 2-11). Hence, the binding characteristics of these two receptors are similar, although they differ significantly in their ability to activate gene transcription by regulatory domain TAF-1, which is minimal or absent in ERβ. Both ERα and ERβ are required for normal ovarian function, as shown by specific receptor knockout studies in mice.¹⁶ ERα is primarily responsible for estrogenic effects in other tissues, including the uterus.

17β-estradiol binds to the estrogen receptor with a much higher affinity than estrone or estriol. In addition, the binding of estradiol to its receptor and its subsequent activation also enhances cooperativity, meaning that the action of estradiol binding to one site increases the affinity for it to bind to another site, enabling the receptors to respond to small changes in hormone concentration. Estrogen’s relatively long duration of action is also due in part to the high-affinity state achieved by its receptor. On the other hand, clomiphene exerts its antiestrogenic effects by negative cooperativity, preventing the transition of the estrogen receptor from its low-affinity state to its high-affinity state.

The two receptors (ERα and ERβ) are differentially expressed in different tissues, leading to the differences in the response to the same hormone subtypes.^243,²⁴⁴ The α receptors are predominantly expressed in breast cancer tissue, ovarian stroma, and endometrium. ERβ receptors, on the other hand, are expressed in several nonclassic target tissues, including the kidney, intestinal mucosa, lung, bone, brain, endothelial cells, and the prostate gland. 17β-estradiol and estrone have a higher affinity for α receptors and thus exert their effects predominantly on target tissue with α receptor expression. Phytoestrogens such as genistein and coumestrol, on the other hand, bind predominantly to β receptors²⁴⁵ and would be expected to exert their effects on target tissues expressing these receptors.

The conformational change of the ligand-binding domain also differs in both α and β estrogen receptors, depending on which ligand has been bound to the receptor.²⁴⁴ This distinct conformational change is the major factor that determines the receptor’s ability to interact with coactivators or corepressors. For example, estradiol activates transcription when it binds to ERα but inhibits transcription when bound to ERβ. Raloxifene and tamoxifen, on the other hand, inhibit transcription when forming complexes with ERα and activate transcription when bound to ERβ.

The differential expression of target tissue adaptor proteins and phosphorylation also affect gene transcription. A higher concentration of coactivators or corepressors in the target tissue can affect the cellular response of that tissue to the same ligand. Phosphorylation of the receptor by protein kinases increases the transcriptional activity of the receptor. For example, growth factors such as epidermal growth factor and IGF-I can stimulate protein kinase phosphorylation, activating the estrogen receptor—even in the absence of estrogen.

Somatic mutations in ERα are described and may be associated with certain disease states. A nonsense mutation (premature stop codon) in ERα has been described in a patient with decreased bone mineral density, increased bone turnover, and incomplete closure of bone epiphyses, which demonstrates the role of the ERα in bone growth and homeostasis.²⁴⁶ ER mutations have also been detected in patients with breast cancer. Such mutations include exon 5 deletion within the ligand-binding domain, leading to a constitutively active receptor, and exon 7 deletion, displaying dominant negative activity and inhibition of ER function.²⁴⁷

Antiestrogens and ER

Compounds with antiestrogenic activity can be classified into two categories: those with pure antiestrogenic activity and those with both agonist and antagonist properties. Tamoxifen—an antiestrogen that is used both as a chemopreventive agent and as a hormonal therapeutic agent for breast cancer—inhibits ER action.²⁴⁸

Paradoxically, tamoxifen acts as an estrogen in uterine tissue, and this tissue-specific estrogenic effect is the reason why prolonged tamoxifen therapy increases the risk of uterine cancer.²⁴⁹ Raloxifene, a related benzothiophene analog, retains its antiestrogenic effect in breast and uterine tissue. Both tamoxifen and raloxifene have estrogen-like effects on nonreproductive tissues such as bone and heart and lung tissue.²⁵⁰

Tamoxifen acts by competing with estrogen for receptor binding. Because estrogen-binding affinity is of higher magnitude than tamoxifen, severalfold higher concentrations of tamoxifen are required to inhibit estrogen action. The agonistic or antagonistic effect of tamoxifen is determined by the presence of different promoter elements in the specific cell type.²⁵¹ Estrogen binding to receptors activates both transcription domains (i.e., TAF-1 and TAF-2). Tamoxifen’s agonistic activity is due to activation of TAF-1, and its antagonistic activity is due to its ability to inhibit the estrogen-dependent activation of TAF-2. The ligand-binding sites for estrogen and antiestrogen are not identical, and tamoxifen binding on receptors induces conformational changes that alter interaction with estrogen-associated proteins and modulates transcriptional activity.^251,²⁵² Tamoxifen also activates ER-mediated induction of promoters that are regulated by the TAF-1 site, which explains why it has estrogenic effects on the endometrium—a tissue with significant TAF-1 transcription function. In other cell types, such as those in the breast, TAF-1 has weak transcriptional activity. Hence, antiestrogens have no effect on TAF-1 mediated transcription.²⁵³ Raloxifene, on the other hand, may activate estrogen-responsive genes through a response element that is distinct from HRE.²⁵⁴

Pure antiestrogens are derivatives of estradiol that have a long hydrophobic side chain at position 7. Examples of pure antiestrogens include ICI 164,384 and ICI 182,780 (Fulvestrant). Binding of pure antiestrogens may sterically interfere with the dimerization process and thus inhibit DNA binding. Furthermore, these compounds increase the rate of receptor degradation and also inhibit ER-mediated transcription by preferential binding to corepressors, which contributes to their antiestrogenic activity.^255,²⁵⁶

Progesterone Receptor

Structure and Function

As with the estrogen receptor, there are two major forms of progesterone receptors— PR-A and PR-B—which are derived from the same gene. PR-A and PR-B are identical except that PR-B contains an additional 164-amino acid sequence at the N-terminal end, which is referred to as the B-upstream segment (BUS) (Fig. 2-12). PR-A has a molecular weight of 94 kDa and contains 768 amino acids; PR-B is 114 kDa with 933 amino acids. The two forms derive from two distinct estrogen-regulated promoters.²⁵⁷ The transcription function domain (TAF-1) in the progesterone receptor is located in the 91-amino acid segment of the regulatory region, and TAF-2 is located in the hormone-binding domain. In PR-B, the BUS contains a third activation domain, TAF-3, which can synergize the actions of other TAFs or autonomously activate transcription.²⁵⁸ TAF-3 recruits and allows a separate subset of coactivators to bind with PR-B that do not interact efficiently with PR-A. Thus, PR-A and PR-B display different transactivation properties that are both cell specific and target gene promoter specific.²⁵⁹ The two isoforms PR-A and PR-B have distinct cellular localization and in the absence of ligand binding, PR-A is predominantly localized in the nucleus and PR-B is present in the cytoplasm.²⁶⁰

Figure 2-12 A schematic diagram of progesterone receptors isoforms. Different domains (A-E) and their corresponding functions are shown. PR-A, progesterone receptor-A; PR-B, progesterone receptor-B.

The role of progesterone receptor isoforms is not yet fully elucidated. Selective ablation of PR-A expression in mice resulted in severe abnormalities in ovarian and uterine function that led to infertility but did not affect progesterone responses in the mammary gland or thymus.²⁶¹ In contrast PR-B ablation does not affect ovarian, uterine, and thymic responses to progesterone and manifests as reduced mammary ductal morphogenesis. Thus, PR-A is essential for female fertility; in the absence of PR-A, PR-B functions in a tissue-specific manner and mediates some of the progesterone receptor actions in the mammary gland.²⁶¹ However, transgenic mice carring an extra copy of the PR-A gene have abnormal mammary gland development, indicating that overexpression of PR-A may have physiologic significance.

The relative levels of the two isoforms differ in the endometrium during the menstrual cycle.²⁶² In the uterus, progesterone action downregulates cell-cycle arrest proteins but upregulates growth factors and their receptors and other regulators²⁶³ and is also essential for the initiation and maintenance of pregnancy. Progesterone antagonist RU-486 (mifepristone) initially was synthesized as a glucocorticoid receptor antagonist²⁶⁴ but later was found to display marked antiprogesterone activity.²⁶⁵ It binds to the glucocorticoid recptor with threefold higher affinity than dexamethasone and to the progesterone receptor with a fivefold higher affinity than natural progesterone.²⁶⁶ Unlike progesterone, the RU-486–progesterone receptor complex inhibits transcription because it has a slightly different conformational change in the TAF-2 domain.²⁶⁷ If implantation occurs, it downregulates progesterone-induced genes and results in decidual necrosis and detachment of the conception products.^251,²⁶⁶

Androgen Receptor

Structure and Function

The androgen receptor gene was cloned in 1988 and was localized on the human X chromosome between the centromere and q13.²⁶⁸ Like the progesterone receptor, the androgen receptor also exists in two forms: a full-length B form and a shorter A form (molecular weights ≈110 kDa and 87 kDa); both are encoded by the same gene.²⁶⁹ The 87-kDa isoform (AR-A) contains the intact C-terminus but lacks 188 amino acid residues in the N-terminus of the 110-kDa isoform (AR-B) (Fig. 2-13). The ratio of AR-B to AR-A in genital skin fibroblasts from healthy subjects is 10:1. It is not known whether there are functional differences between these isoforms.²⁶⁹ The DNA-binding domain or TAF-2 of the androgen receptor is similar to the TAF-2 regions of other steroid hormone receptors (progesterone, estrogen, glucocorticoid, and mineralocorticoid receptors) but is related most closely to the progesterone receptor.²⁷⁰ Progesterone shows cross-reactivity with androgen receptors, to a degree that becomes clinically relevant only at pharmacologic doses.

Figure 2-13 A schematic diagram of androgen receptor isoforms. The androgen receptor is very similar to the progesterone receptor and exists in a shorter A form and a full-length B form.

In most tissues, testosterone is converted to DHT by the action of the enzyme 5α-reductase. DHT binds to the androgen receptor with a higher affinity than testosterone, leading to greater stabilization of the receptor and more efficient signaling, which amplifies the androgen action. Hence, the efficiency of local conversion of testosterone to DHT is an important intracellular step in the androgen response.

A large number of androgen receptor mutations that can alter receptor function have also been described ^271,²⁷²; for example inactivating point mutations in the hormone-binding domain of the androgen receptor can generate different phenotypes that lead to partial to complete androgen insensitivity. A point mutation at residue 689 that substitutes Proline to Histidine may alter the conformation of the ligand-binding domain, which reduces the androgen receptor’s affinity for DHT and abolishes its ability to transactivate the HRE.²⁷³ A Serine to Proline substitution at residue 865 eliminates androgen binding and transactivation; thus it also causes complete androgen insensitivity.²⁷⁴ At residue 807, a substitution of Threonine for Methionine causes partial androgen insensitivity by reducing—but not abrogating—the androgen binding to the receptor. However, a Valine or Arginine substitution at the same site totally abrogates androgen binding and causes complete androgen insensitivity syndrome.²⁷⁵

Nongenomic Actions of Steroids

Some steroid hormone actions are independent of their classic genomic actions, as mediated via nuclear receptors—these are called nongenomic actions.²⁷⁶ These effects are rapid, occur within seconds, and are not affected by inhibitors of gene transcription such as dactinomycin or inhibitors of protein synthesis such as cycloheximide. These rapid actions include sodium and calcium ion transport and some neural and cardiovascular effects.²⁷⁷ The messenger and effector system vary with the steroid and the cell type. Thus far, studies have suggested that specific binding sites or receptors are present on the cell membrane and that steroid binding triggers rapid changes in the electrolyte transport system. For example, estrogens have been shown to modulate some cardiovascular effects by inducing Ca²⁺ flux and vasodilation in the coronary artery.²⁷⁸ Furthermore, steroids can activate second-messenger pathways, generating second messengers capable of altering gene transcription independently of their classical receptor-mediated gene transcription.^279,²⁸⁰ Thus, steroid hormones have nongenomic effects as well as genomic effects that are mediated by both classical steroid receptor-mediated pathways and second-messenger pathways.

Tropic Hormone Action

Tropic hormones (i.e., gonadotropins and GnRH) are primarily hydrophilic and depend on their interactions with receptors scattered on the plasma membrane at the cell surface. These receptors, in turn, can be classified into four groups based on their structural homology and type of intracellular messenger involved in activating the signal transduction pathway (see Table 2-6).

G Protein-Coupled Receptors

The gonadotropin and GnRH receptors belong to the large family of receptors that are known as serpentine or seven-transmembrane domain receptors because each contains three domains: an amino-terminal extracellular domain (ectodomain) followed by seven hydrophobic amino acid segments; the transmembrane domains, which span the membrane bilayer (or endodomain); and a hydrophilic carboxyl-terminal domain, which resides within the cytoplasmic compartment. Because they depend on G protein transducers to execute their biologic effects, they are also known as G protein-coupled receptors (GPCRs). G proteins are heterotrimeric proteins associated with these receptors and are so-called because they bind the guanine nucleotides guanosine diphosphate (GDP) and guanosine triphosphate (GTP). Each G protein consists of three subunits: α, β, and γ. They are a heterogeneous family, and at least 16 α subunit genes, 6 β subunit genes, and 12 γ subunit genes have been identified. Various combinations of these provide a large number of possible αβγ complexes. The identity of the individual G protein is determined by the nature of the α subunit. As such, G proteins can be divided into four subfamilies (G_s, G_i, G_q, and G_i2) based on their protein sequence homology. They act as transducers, linking the receptors with the effector proteins that are responsible for producing changes in cellular function. The large variety of G proteins allows for flexibility and diversity in the response of the target cell, depending on the level of G protein expression.

There are many types of possible effector molecules, including adenylyl cyclase, calcium channels, potassium channels, cGMP, and phospholipase. Each effector molecule in turn produces a large quantity of second messengers such as cAMP, Ca²⁺, and phosphatidylinositols, which activate an even larger number of downstream molecules. This is an important factor that explains why the endocrine system is sensitive to such low concentrations of circulating ligand and why only a small percentage of cell membrane receptors need to be occupied to generate a response.

Other examples of receptors and their ligands that fall into this family include the TSH receptor and ACTH receptor. The structure of the LH receptor and FSH receptor are similar to each other and show close homology with the TSH receptor. A comparison of the LH receptor with the FSH receptor shows about 70% homology in the transmembrane domain, but only 42% in the ectodomain and 48% in the endodomain regions.

Gonadotropin Receptors

The gonadotropin receptors are primarily expressed in the gonadal tissues and demonstrate a unique expression pattern that determines their cell-specific roles. LH receptors are present in ovarian thecal and luteal cells and in testicular Leydig cells. FSH receptors are found in ovarian granulosa cells and in testicular Sertoli cells. Furthermore, FSH induces LH receptors in the granulosa cells that are expressed in mature follicles. These receptors are essential in mediating the gonadotropin actions in steroidogenesis and in gonadal growth and differentiation. Their structure, function, and regulation as well as molecular biology have been extensively covered in recent reviews.^281,²⁸²

Structure and Function

Genes for both LH and FSH receptors are located on chromosome 2p21. The LH receptor gene is about 70 kb and consists of 11 exons and 10 intervening introns.²⁸¹ The FSH receptor gene is about 54 kb and contains 10 exons and 9 intervening introns.²⁸² As shown in Figure 2-14, the structural similarities between the two receptors are remarkable with the exception of an additional exon in the LH receptor. The long exon 11 in LH receptor and exon 10 in FSH receptor encodes for the seven-transmembrane domains and the intracellular tail as well as the C-terminal end of the hinge region of ectodomain. The remaining 9 exons in the FSH receptor and 10 exons in the LH receptor encode for the entire ectodomain. Both receptors have multiple splice sites, leading to transcription of multiple mRNA splice variants and expression in the ovary and testis.²⁸¹

Figure 2-14 A schematic representation of the FSH (A)/LH (B) receptor gene and localization of identified mutations. The structural organization of the FSH/LH receptor protein is provided below it. The extracellular domain of the protein is shown as a straight line, followed by the seven-transmembrane domains and the intracellular domain. In the FSH receptor gene exon 10 and in the LH receptor gene exon 11 codes for the seven-transmembrane signal transduction domains.

The LH receptor protein contains a 24-amino acid signal peptide (17 in FSH receptor) and the mature protein consists of 675 amino acid residues (678 in FSH receptor) (see Fig. 2-14). The molecular mass is approximately 85 to 90 kDa for both glycosylated LH and FSH receptor proteins. The ectodomains of these receptors are composed of several leucine-rich repeats (LRRs) of about 24 amino acid residues each (9 in LH receptor and 10 in FSH receptor) that form a half donut-shaped structure and are essential for the binding of gonadotropins. The ectodomain is connected to the transmembrane region by the hinge region with conserved sequences. The transmembrane region consists of seven helical structures and is followed by the cytoplasmic C-terminal tail, both of which are important for interaction with intracellular proteins. The binding of LH and FSH to their specific receptors leads to receptor activation and downstream signaling.

Mechanism of Action

The intracellular messenger for LH and FSH is cAMP, which is derived from adenosine triphosphate (ATP) through the action of the enzyme adenylate cyclase. The basic steps involved in the receptor-mediated action of gonadotropins are illustrated in Figure 2-15. The regulation of adenylate cyclase is mediated by a GTP-dependent regulatory G-protein, each of which is composed of three subunits—α, β, and γ—and is associated with the receptor in an inactive GDP-bound form. Gonadotropic binding to their respective receptors induces a conformational change in the receptor and activation of the G-protein complex that leads to release of GDP and binding of the α subunit to GTP.²⁸³ Subsequently, the GTP-bound form of the α subunit disassociates from the receptor as well as from the stable β/γ dimer and activates the adenylate cyclase. This in turn leads to an increase in the levels of intracellular cAMP, which in turn activates protein kinase A (PKA) by binding to the inhibitory regulatory subunit of PKA and causing it to dissociate from the complex. Activated PKA phosphorylates many cellular substrates, including several nuclear transcription factors (e.g., cAMP response binding protein [CREB]).²⁸¹ The binding of CREB to the cAMP response element activates many genes. The fact that PKA activation does not account for all the actions of gonadotropins and that LH can stimulate steroid hormone synthesis without significant changes in cAMP indicated that another pathway may be activated. There is now evidence that the phosphatidylinositol 3,4,5-triphosphate (IP₃) pathway is also activated by the LH receptor²⁸⁴ as well as by the FSH receptor,²⁸⁵ although it is not clear whether the IP₃ pathway activates the same or different responses in the target cells. The third signaling pathway, the MAPK pathway, has also been shown to be activated by the LH receptor²⁸⁶ (see Fig. 2-15).

Figure 2-15 A diagrammatic representation of the interlinked signaling pathways between FSH/LH G-protein coupled receptors (GPCRs) and insulin/IGF-I receptor tyrosine kinases in ovarian cells. Activation of GPCR by hormone binding stimulates the Gα subunit to bind guanosine triphosphate (GTP) instead of guanosine diphosphate (GDP), leading to its dissociation from β/γ subunits to activate downstream signaling factors such as adenylyl cyclase that synthesizes second-messenger cyclic adenosine monophosphate (cAMP). Binding of cAMP in turn activates protein kinase A (PKA), leading to DNA binding and downstream cellular response. Also illustrated here is the IGF-I receptor signaling pathway. IGF-I/Insulin binding to the receptor initiates autophosphorylation and tyrosine phosphorylation of insulin receptor substrates (IRS), leading to activation of phosphatidylinositol-3-kinase (PI3K) and generation of 3-phosphorylated-inositol (IP3) from phosphoinositol (PIP2), which activates PI-dependent protein kinase-1 (PDK-1). PDK-1 in turn activates Akt/protein kinase B (Akt/PKB), leading to biologic effects. The activation of the insulin receptor substrates (IRS) also allows the docking and activation of small adaptor molecules with SH-2 domains (e.g., growth factor receptor binding protein-2 [Grb-2[] and Shp2). Activated Grb-2 recruits SOS-1, which activates the RAS pathway and gene transcription. (Main pathways are in bold, interlinked signaling pathways are in broken lines). SOS-1, son-of-sevenless; ERK, extracellular signal regulated kinase; FSH, follicle stimulating hormone; LH, luteinizing hormone.

Continuous stimulation of receptors can lead to a decrease in response, a phenomenon known as downregulation. For example, pulsatile LH secretion maintains LH receptors and steroidogenesis in the gonads. However, persistent endogenous elevation of LH or hCG levels downregulates LH receptors and leads to desensitization to the hormonal signal.²⁸⁷ This involves receptor phosphorylation, which uncouples the receptor from the G protein, ending the response. The LH/hCG receptor undergoes desensitization in response to LH or hCG by the phosphorylation of the C-terminal cytoplasmic tail of the receptor.²⁸⁸ Another mechanism of downregulation is the uncoupling of the regulatory and catalytic subunits of the adenylate cyclase enzymes. For example, LH stimulates steroidogenesis by coupling the stimulatory to the catalytic units of adenylate cyclase. Prostaglandin F_2α, on the other hand, inhibits the action of LH by an inhibitory regulatory unit that uncouples the catalytic unit to interfere with gonadotropin action.

Mutations in Gonadotropin Receptors

Both activating and inactivating gene mutations have been described in LH and FSH receptors and have been the subject of recent review.²⁰⁹ Interestingly, the activating mutations in the LH receptor gene result in altered testosterone production, and these mutations are known to affect the phenotype of men only, whereas inactivating mutations affect sexual differentiation and fertility in both men and women.

To date, a total of 15 activating mutations have been identified in the LH receptor, all localized in the transmembrane region or in the intracellular loops (see Fig. 2-14A). Transmembrane 6 and the third intracellular loop are recognized as mutational hot spots, and 10 of 14 mutations are localized to these regions. These gain-of-function mutations lead to male-limited precocious puberty characterized by Leydig cell hyperplasia, premature puberty, and onset of spermatogenesis in boys as young as age 3 years.^289,²⁹⁰ These patients have blood testosterone levels in the pubertal range, with low or undetectable LH levels. A novel activating LH receptor mutation associated with Leydig cell tumors and nonfamilial precocious puberty has also been described.²⁹¹

An equal number of inactivating mutations in LHR that cause partial to complete inactivation of receptor have been described.²⁰⁹ These inactivating mutations in LH receptors cause Leydig cell hypoplasia in men and amenorrhea in women. In contrast, loss-of-function mutations are scattered throughout the LH receptor protein. Depending on the mutation, the severity of the male phenotype varies from primary hypogonadism to male pseudohermaphroditism with sexual ambiguity.

Conversely, activating mutations of the FSH receptor are rare. Only one mutation has been identified this far (located in the third intracellular domain of exon 10) in a hypophysectomized patient who was fertile despite the lack of FSH and LH production.²⁹² Likewise, fewer inactivating mutations have been identified in the FSH receptor (see Fig. 2-14B). Most of these are located in the extracellular domain and show reduced ligand binding and signal transduction. They are implicated in hypergonadotropic ovarian dysgenesis in females and variable degrees of spermatogenic failure in males. Recently, two heterozygous missense mutations in the exon 10 encoding the transmembrane region have been described, and both were associated with the familial spontaneous ovarian hyperstimulation syndrome. Although there were no significant differences in the cAMP responses between the wild-type and mutated receptor, the mutated receptor gained sensitivity to hCG.^293,²⁹⁴ Low-affinity binding of hCG to the FSH receptor most likely triggered signal transduction that led to hyperstimulation of ovaries.

Gonadotropin-releasing Hormone (GnRH) Receptor

GnRH binds to specific membrane receptors in pituitary gonadotrophs and stimulates LH and FSH secretion (Fig. 2-16). Besides gonadotrophs, GnRH receptors have also been expressed on the gonads, placenta, and brain²⁹⁵^–²⁹⁷ Specific GnRH receptors are also characterized in immortalized αT3-1 gonadotroph cells as well as in GnRH-secreting GT1 hypothalamic cells.^296,²⁹⁸ In the ovary, GnRH receptors are expressed in granulosa and luteal cells. In the testis, these receptors are expressed in Leydig cells but not Sertoli cells.²⁹⁹ In cultured granulosa cells, the receptor activation stimulates progesterone and prostaglandin synthesis, oocyte maturation, and ovulation^300,³⁰¹ but inhibits FSH-induced steroidogenesis, follicular development, and maturation as well as inhibin secretion.³⁰²^–³⁰⁴ In luteal cells, it inhibits LH receptor expression and thus LH action.³⁰⁵

Figure 2-16 A schematic representation of the gonadotropin-releasing hormone (GnRH) receptor gene and localization of identified mutations. The structural organization of the GnRH receptor protein is provided below it. The short extracellular domain of the protein is indicated by the straight line, followed by the seven-transmembrane domains. The GnRH receptor lacks a carboxyterminal cytoplasmic tail. Exon 1 codes for transmembrane domains 1-3, exon 2 codes for transmembrane domains 4-5, and exon 3 codes for transmembrane domains 6-7. 5′UTR, 5′’ untranslated region.

The structure function and intracellular signaling pathways involved in GnRH action have been extensively reviewed in recent years.³⁰⁶^–³⁰⁸ The GnRH receptor gene, located on chromosome 4q21.2, consists of three exons³⁰⁹ and encodes a 327-amino acid protein with an approximate molecular weight of 50 to 60 kDa. The receptor has the characteristic features of GPCRs (Fig. 2-17). It consists of an aminoterminal domain, followed by seven transmembrane domains that are connected by three extracellular and three intracellular loops. Unlike other GPCRs, the GnRH receptor lacks the characteristic C-terminal cytoplasmic domain. GnRH binds to the transmembrane domain and the extracellular loop in the hairpin structure and requires partial entry of its N- and C-terminal regions into the transmembrane core of the receptor.³¹⁰

Figure 2-17 A schematic representation of GnRH signaling through GPCR. Binding of GnRH to its receptor acts through G_q protein. Activated G_q binds ATP and activates its downstream target phospholipase C (PLC) and hydrolyzes PIP2, producing IP₃ and diacylglycerol (DAG). IP₃ diffuses through the cytoplasm to the endoplasmic reticulum (ER) and opens a calcium channel releasing calcium stores. Calcium and DAG activate protein kinase C (PKC), leading to cellular response.

The nature of the intracellular signaling mechanism by the GnRH receptor has been primarily studied in a gonadotroph-derived αT3-1 cell line, as depicted in Figure 2-18. The effector system in the GnRH receptor is the phospholipase C β (PLCβ) and the second messengers are IP₃ and 1,2 diacylglycerol (DAG). Its mechanism of action is dependent on calcium. Activation of the receptor by ligand binding initiates a series of steps that lead to transduction of signals. The first step is G protein (G_q/11)-mediated activation of enzyme PLCβ, leading to hydrolysis of phosphoinositides (PIP₂) and resulting in the production of IP₃ and DAG. The IP₃ interacts with a receptor on the endoplasmic reticulum to promote the oscillatory or biphasic release of Ca²⁺ from intracellular stores, which is known to be an important trigger for gonadotropin secretion. The increased Ca²⁺ and DAG in turn activate a series of protein kinase C subspecies (PKC). PKC then induce various downstream signal transduction cascades, including the extracellular signal regulated kinase and Jun N-terminal kinase signaling pathways.

Figure 2-18 The TGF-β/activin signaling pathway. Binding of a TGF-β family member (e.g., activin) to its type 2 receptor forms a ligand–receptor complex and activation of the type 1 receptor by phosphorylation. The activated type 1 receptor then phosphorylates receptor-regulated SMAD (R-SMAD). This allows R-SMAD to associate with Co-SMAD and move into the nucleus. In the nucleus, the SMAD complex associates with a DNA-binding partner, such as FAST-1. This complex then binds to specific enhancers in the target gene, activating gene transcription. (In the TGF-β/activin pathway, R-SMAD is formed by SMAD2 and SMAD3. Co-SMAD is formed from SMAD4.)

Mutations in GnRH Receptor Gene

A functional GnRH receptor is a prerequisite for normal LH and FSH production. Hence, it is critical for normal pubertal development and reproduction. Several inactivating mutations have been identified in the GnRH receptor gene that are associated with variable phenotypes ranging from modest to complete insensitivity to GnRH. These mutations are present in a number of patients with idiopathic hypogonadotropic hypogonadism (IHH) with an autosomal recessive pattern of inheritance as well as in some patients with sporadic IHH.³¹¹ Most identified mutations are heterozygous missense mutations and are scattered throughout the GnRH receptor gene. A total of 14 mutations have been identified to date³⁰⁸ and are listed in Figure 2-16. In vitro studies have shown that some of these mutations make the receptor totally nonfunctional; others have some ability to elicit a response to GnRH.

Peptide Hormone/Growth Factor Action

Pituitary gonadotropin signaling plays a key role in follicular growth, ovulation, and luteinization. However, it is increasingly recognized that their actions are also dependent on their interaction with peptide/growth factor signaling pathways, including IGFs, EGF, and members of the TGFβ family (see Fig. 2-15). Understanding these pathways provides insight into how these factors interact and complement the FSH/LH pathway to control follicular growth.

Family of Receptors with Tyrosine Kinase Activity

Insulin, IGF, and EGF receptors belong to a distinct group of receptors and differ from the GPCRs both structurally and functionally. Unlike the GPCRs, they only span the membrane once and acquire their signaling ability through the activation of tyrosine kinase, which is intrinsic to these individual receptor molecules. Hence, they are known commonly as tyrosine kinases. Their main ligands include hormones such as insulin and IGF, as well as paracrine and autocrine regulators such as platelet-derived growth factor (PDGF), bFGF, and EGF. Thus, via tyrosine phosphorylation, a number of physiologic processes such as cell proliferation, cell migration, cell differentiation, and apoptosis are mediated by these receptors. This explains why this group of receptors has been the target of much oncogenic research.

All tyrosine kinase receptors have a similar structure: an extracellular domain for ligand binding, a single transmembrane domain, and a cytoplasmic domain. Ligand specificity is determined by the unique amino acid sequences making up the extracellular domain, which determines the three-dimensional conformation of the receptor. The transmembrane domains are heterogeneous, whereas the cytoplasmic domains are fairly homologous. They respond to ligand binding by undergoing conformational changes and autophosphorylation. The structure of the IGF-I receptor is strikingly similar to that of the insulin receptor, with two transmembrane domains linked by disulfide bridges, formed by two α and β subunits.³¹² The IGF-I receptor gene is located on chromosome 15 at bands q25-26 and contains 21 exons.³¹³

The steps involved in signal transduction by IGF-I/insulin have been extensively reviewed in recent years ^314,³¹⁵ and are illustrated in Figure 2-15. The association of the ligand (e.g., insulin) with the receptor’s extracellular domain triggers receptor dimerization. This results in the phosphorylation of tyrosine residues on both the receptor and nonreceptor substrates. The phosphorylation of receptor tyrosine residues occurs at specific locations, which causes these sites to associate with a variety of accessory proteins that have independent signaling capabilities. These accessory proteins include phospholipase Cγ, PI3 kinase (PI3K), GAP, and growth factor receptor-bound protein-2 (GRB2). These interactions are mediated by the presence of highly conserved type 2 src homology domains (SH2) in each accessory molecule, thus named based on their sequence homology to the src proto-oncogene. Each SH2 domain is specific for the amino acids surrounding the phosphotyrosine residues in the receptor molecule. PI3K also produces a second messenger, such as IP₃, which in turn activates kinase AKT, also known as protein kinase B (PKB). Proteins phosphorylated by PKB promote cell survival.

Although these associations may trigger immediate signaling events, other accessory proteins (e.g., GRB2) may serve to construct the scaffolding for a more complex signaling apparatus, such as that seen in the RAS-RAF-MEK pathway, This pathway recruits multiple other proteins, resulting in the activation of nuclear transcription and protein synthesis (see Fig. 2-15).

Growth hormone and prolactin fall into another group of receptors that also have a single transmembrane-spanning segment and a short cytoplasmic tail. However unlike IGF-I/insulin receptors they do not possess intrinsic tyrosine kinase activity but interact with other soluble transducer molecules (e.g., Janus kinase 2) that do have tyrosine kinase activity.

Mechanism of Action of Activins and Inhibins

Activins and inhibins are members of the TGFβ superfamily and use a common general mechanism for signal transduction through serine/threonine-specific protein kinases rather than tyrosine kinases. The mechanisms involved in signal transduction by the serine/threonine kinases has been extensively reviewed in recent years.^316,³¹⁷ The activin receptor was first to be cloned and was later followed by other receptors.^318,³¹⁹ They are glycoproteins of approximately 55 kDa and consist of a 500-amino acid sequence. Two types of activin membrane receptors have been identified: type I (ActR-I) and type II receptors (ActR-II). Each contains an intracytoplasmic serine/threonine kinase domain. The steps involved in the mechanism of action of activin are depicted in Figure 2-18. Activin directly interacts with and binds to the relatively short extracellular region of ActR-II when expressed alone or in concert with ActR-I. It can also bind to other TGFβ family members (e.g., bone morphogenic proteins [BMPs] 2, 4, and 7) in concert with BMP type I receptor, which suggests that these receptors have the ability to cross-talk. Activin binding brings together two type II receptors and two type I receptors, which form a receptor complex. One of the receptor kinases phosphorylates and activates the other, which in turn phosphorylates their substrates—the SMAD proteins [the term SMAD is derived from the combination of names of two genes, the C. elegans gene called Sma and the Drosiphila gene Mad]. SMADs are a novel family of signal transducers that can be divided into three groups that include receptor-regulated SMADs (R-SMAD) and a single common SMAD (C-SMAD or SMAD4). The third group includes inhibitory SMADS that antagonize signaling. In activin/TGFβ signaling, the activated type I receptor phosphorylates ligand-specific R-SMADs (SMAD2 and SMAD3), allowing these proteins to associate with SMAD4. The complex then translocates to the nucleus as transcription cofactor. In the nucleus, the action of SMAD complex is modulated by a variety of transcription factors (DNA-binding proteins) at target DNA promoters, which leads to gene transcription.

Inhibin is a heterodimer formed between an inhibin α chain and an activin β chain, and its biologic activity is opposite that of activin. A separate receptor for inhibin has not been identified. The mechanisms by which inhibin antagonizes activin actions are not clearly understood. It has been shown that inhibin competes with activin for binding to the activin receptors but is unable to trigger signaling.³²⁰ This may be one of the mechanisms by which inhibins antagonize activin actions. Other actions of inhibin may be mediated by as-yet unidentified inhibin receptors.