All steroid hormones are of basically similar structure with relatively minor chemical differences leading to striking alterations in biochemical activity. The basic structure is the perhydrocyclopentanephenanthrene molecule. It is composed of three 6-carbon rings and one 5-carbon ring. One 6-carbon ring is benzene, two of the 6-carbon rings are naphthalene, and three 6-carbon rings are phenanthrene; add a cyclopentane (5-carbon ring), and you have the perhydrocyclopentanephenanthrene structure of the steroid nucleus.

The sex steroids are divided into three main groups according to the number of carbon atoms they possess. The 21-carbon series includes the corticoids as well as the progestins, and the basic structure is the pregnane nucleus. The 19-carbon series includes all the androgens and is based on the androstane nucleus, whereas the estrogens are 18-carbon steroids based on the estrane nucleus.

There are six centers of asymmetry on the basic ring structure, and there are 64 possible isomers. Almost all naturally occurring and active steroids are nearly flat, and substituents below and above the plane of the ring are designated alpha (a) (dotted line) and beta (β) (solid line), respectively. Changes in the position of only one substituent can lead to inactive isomers. For example, 17-epitestosterone is considerably weaker than testosterone; the only difference being a hydroxyl group in the a position at C-17 rather than in the β position.

The convention of naming steroids uses the number of carbon atoms to designate the basic name (e.g., pregnane, androstane, or estrane). The basic name is preceded by numbers that indicate the position of double bonds, and the name is altered as follows to indicate one, two, or three double bonds: -ene, -diene, and -triene. Following the basic name, hydroxyl groups are indicated by the number of the carbon attachment, and one, two, or
three hydroxyl groups are designated -ol, -diol, or -triol. Ketone groups are listed last with numbers of carbon attachments, and one, two, or three groups designated -one, -dione, or -trione. Special designations include dehydro, elimination of two hydrogens; deoxy, elimination of oxygen; nor, elimination of carbon; delta or δ, location of double bond.

The overall steroid biosynthesis pathway shown in the figure is based primarily on the pioneering work of Kenneth J. Ryan and his coworkers.^1,2 These pathways follow a fundamental pattern displayed by all steroid-producing endocrine organs. As a result, it should be no surprise that the normal human ovary produces all three classes of sex steroids: estrogens, progestins, and androgens. The importance of ovarian androgens is appreciated, not only as obligate precursors to estrogens, but also as clinically important secretory products. The ovary differs from the testis in its fundamental complement of critical enzymes and, hence, its distribution of secretory products. The ovary is distinguished from the adrenal gland in that it is deficient in 21-hydroxylase and 11β-hydroxylase reactions. Glucocorticoids and mineralocorticoids, therefore, are not produced in normal ovarian tissue.

During steroidogenesis, the number of carbon atoms in cholesterol or any other steroid molecule can be reduced but never increased. The following reactions can take place:

The traditional view of steroidogenesis was that each step was mediated by many enzymes, with differences from tissue to tissue. A fundamental simplicity to the system emerged when the responsible complementary DNAs and genes were cloned.^3,4 and ⁵

Steroidogenic enzymes are either dehydrogenases or members of the cytochrome P450 group of oxidases. Cytochrome P450 is a generic term for a family of oxidative enzymes, termed 450 because of a pigment (450) absorbance shift when reduced. P450 enzymes can metabolize many substrates; e.g., in the liver, P450 enzymes metabolize toxins and environmental pollutants. The human genome contains genes for 57 cytochrome P450 enzymes, 7 in mitochondria and 50 in the endoplasmic reticulum (the major site for metabolic clearance). The following distinct P450 enzymes are identified with steroidogenesis: P450scc is the cholesterol side chain cleavage enzyme; P450c11 mediates 11-hydroxylase, 18-hydroxylase, and 19-methyloxidase; P450c17 mediates 17-hydroxylase and 17,20-lyase; P450c21 mediates 21-hydroxylase; and P450arom mediates aromatization of androgens to estrogens. Marked differences in the exonintron organization of the P450 genes are compatible with an ancient origin; thus, the superfamily of P450 genes diverged more than 1.5 billion years ago.

The structural knowledge of the P450 enzymes that has been derived from amino acid and nucleotide sequencing studies demonstrated that all the steps between cholesterol and pregnenolone were mediated by a single protein, P450scc, bound to the inner mitochondrial membrane. Cloning data indicate the presence of a single, unique P450scc gene on chromosome 15. These experiments indicated that multiple steps did not require multiple enzymes. Differing activity in different tissues may reflect posttranslational modifications. In addition, these genes contain tissue-specific promoter sequences, which is another reason that regulatory mechanisms can differ in different tissues (e.g., placenta and ovary). P450scc mutations are very rare, producing impaired steroidogenesis in both the adrenal glands and the gonads, causing abnormal phenotypes and adrenal failure.⁶

Conversion of cholesterol to pregnenolone involves hydroxylation at the carbon 20 and 22 positions, with subsequent cleavage of the side chain. Conversion of cholesterol to pregnenolone by P450scc takes place within the mitochondria. It is one of the principal effects of tropic hormone stimulation, which also causes the uptake of the cholesterol substrate for this step in the ovary. The tropic hormones from the anterior pituitary bind to the cell surface receptor of the G protein system, activate adenylate cyclase, and increase intracellular cyclic AMP. Cyclic AMP activity leads to gene transcription that encodes the steroidogenic enzymes and accessory proteins. In a process that is faster than gene transcription, cyclic AMP stimulates the hydrolysis of cholesteryl esters and the transport of free cholesterol to the mitochondria.

The cholesterol used for steroid synthesis is derived from circulating low-density lipoproteins (LDL), followed by the mobilization and transport of intracellular stores.^5,7,8 LDL cholesterol esters are incorporated into the cell by tropic hormone stimulation of endocytosis via clathrin-coated pits (a mechanism discussed later in this chapter). Cholesterol is stored in the cell in the ester form or as free cholesterol. Indeed, the rate-limiting step in steroidogenesis is the transfer of cholesterol from the outer mitochondrial membrane to the inner mitochondrial membrane where fully active P450scc waits for substrate. The rate-limiting transfer of hydrophobic cholesterol through the aqueous space between the outer and inner mitochondrial membranes is mediated by protein activation stimulated by the tropic hormone. Long-term, chronic steroidogenesis requires gene transcription and protein synthesis, but short-term, acute responses occur independently of new RNA synthesis, although protein synthesis is still necessary, specifically the proteins that regulate cholesterol transfer across the mitochondrial membrane.

Several proteins have been characterized and proposed as regulators of acute intracellular cholesterol transfer. Sterol carrier protein 2 (SCP2) is able to bind and transfer cholesterol between compartments within a cell. Another candidate is a small molecule, steroidogenesis activator polypeptide (SAP), and still another is peripheral benzodiazepine receptor (PBR), which affects cholesterol flux through a pore structure. But the most studied and favored protein as a regulator of acute cholesterol transfer is steroidogenic acute regulator (StAR) protein.^9,10,11,12 and ¹³ StAR messenger RNA and proteins are induced concomitantly with acute steroidogenesis in response to cyclic AMP stimulation. StAR protein increases steroid production and is imported and localized in the mitochondria. But most impressively, congenital lipoid adrenal hyperplasia (an autosomal-recessive disorder) is a failure in adrenal and gonadal steroidogenesis due to a mutation in the StAR gene that results in premature stop codons.^14,15 With this mutation, a low level of steroidogenesis is possible, even permitting feminization at puberty, but continuing tropic hormonal stimulation results in an accumulation of intracellular lipid deposits that destroy steroidogenic capability.¹⁶ StAR is required for adrenal and gonadal steroidogenesis, and, therefore, is necessary for normal male sexual differentiation.

StAR mediates the transport of cholesterol into mitochondria in adrenal and gonadal steroidogenesis, but not in the placenta and brain. StAR moves cholesterol from the outer mitochondrial membrane to the inner mitochondrial membrane where it can enter the steroidogenic pathway by being converted to prenenolone. A group of proteins structurally related to StAR have been identified, designated StARD4, StARD5, and StARD6. StARD4 is responsible for binding free cholesterol as it is produced in the cytoplasm and transporting it to the outer mitochondrial membrane.¹² Because steroid-producing cells do not store large amounts of hormones, acute increases in secretion depend on this system to produce rapid synthesis.

StAR is synthesized in a precursor form as a 285-amino acid protein that has a 25-residue sequence cleaved from the NH₂-terminal after transport into mitochondria.¹⁷ The mutant forms of StAR undergo premature truncation that prevents this proteolytic cleavage. Mutations of the StAR gene, located on chromosome 8p11.2, are the only inherited disorder of steroidogenesis not caused by a defect in one of the steroidogenic enzymes. The absence of StAR expression in placenta and brain indicates the presence of different mechanisms for cholesterol transport in those tissues.

Once pregnenolone is formed, further steroid synthesis in the ovary can proceed by one of two pathways, either via Δ⁵-3β-hydroxysteroids or via the Δ⁴-3-ketone pathway. The first
(the Δ⁵ pathway) proceeds by way of pregnenolone and dehydroepiandrosterone (DHEA) and the second (the Δ⁴ pathway) via progesterone and 17a-hydroxyprogesterone.

The conversion of pregnenolone to progesterone involves two steps: the 3β-hydroxysteroid dehydrogenase and Δ⁴ and ⁵ isomerase reactions that convert the 3-hydroxyl group to a ketone and transfer the double bond from the 5-6 position to the 4-5 position. The 3β-hydroxysteroid dehydrogenase enzyme catalyzes both the dehydrogenation and isomerization reactions, and exists in two forms (type I and type II), encoded by two separate genes on chromosome 1 (the type I gene is expressed in the placenta, breast, and other non-glandular tissues, the type II gene is expressed in the gonads and the adrenal glands). Once the Δ⁴ and ⁵ ketone is formed, progesterone is hydroxylated at the 17 position to form 17a-hydroxyprogesterone. 17a-Hydroxyprogesterone is the immediate precursor of the C-19 (19 carbons) series of androgens in this pathway. By peroxide formation at C-20, followed by epoxidation of the C-17, C-20 carbons, the side chain is split off, forming androstenedione. The 17-ketone may be reduced to a 17β-hydroxyl to form testosterone by the 17β-hydroxysteroid dehydrogenase reaction. Both C-19 steroids (androstenedione and testosterone) can be converted to corresponding C-18 phenolic steroid estrogens (estrone and estradiol) by microsomal reactions in a process referred to as aromatization. This process includes hydroxylation of the angular 19-methyl group, followed by oxidation, loss of the 19-carbon as formaldehyde, and ring A aromatization (dehydrogenation).

As an alternative, pregnenolone can be directly converted to the Δ⁵-3β-hydroxy C-19 steroid, dehydroepiandrosterone (DHEA), by 17a-hydroxylation followed by cleavage of the side chain. With formation of the Δ⁴-3-ketone, DHEA is converted into androstenedione. The four reactions involved in converting pregnenolone and progesterone to their 17-hydroxylated products are mediated by a single enzyme, P450c17, bound to smooth endoplasmic reticulum, regulated by a gene on chromosome 10q24.3. 17-Hydroxylase and 17,20-lyase were traditionally regarded as separate enzymes. These two different functions of a single enzyme, P450c17, are not genetic or structural but represent the effect of posttranslational influencing factors.¹⁸ In the adrenal gland pathway to cortisol, very little 17,20-lyase activity is expressed. In the ovarian theca cells, the testicular Leydig cells, and the adrenal reticularis, both 17-hydroxylase and 17,20-lyase activities are expressed, directing the steroidogenic pathway via dehydroepiandrosterone (DHEA). In the corpus luteum, the principal pathway is via progesterone.

Characterization of the P450c21 protein and gene cloning indicate that the 21-hydroxylase gene, CYP21, is located on chromosome 6p21.3. An inactive pseudogene, CYP21P, is nearby. Many of the mutations that affect CYP21 and cause congenital adrenal hyperplasia are gene conversions involving recombinations between CYP21 and inactivating mutations in CYP21P.

Aromatization is mediated by P450arom found in the endoplasmic reticulum.^19,20 Aromatase cytochrome P450 is derived from chromosome 15q21.1, at a site designated as the CYP19A1 (cytochrome P450, family 19, subfamily A, polypeptide 1) gene, denoting oxidation of the C-19 methyl group. Aromatization in different tissues with different substrates is the result of the single P450arom enzyme encoded by the single gene. A specific haplotype of genetic polymorphisms in CYP19 may be linked to endometrial cancer, a known consequence of excessive estrogenic stimulation of the endometrium.²¹ Aromatase deficiency because of an inactivating mutation of CYP19A1 is very rare; only a handful of cases have been reported.²² Affected females present at birth with virilization because the placenta cannot convert fetal adrenal androgens to estrogens; thus, maternal virilization during the pregnancy is usually also present.

Aromatase transcription is regulated by several promoter sites that respond to cytokines, cyclic nucleotides, gonadotropins, glucocorticoids, and growth factors.²³ Tissue-specific expression is regulated by tissue-specific promoters at the 5′ end of the gene. Thus, this gene has alternative promoters that allow the extremes of highly regulated expression in the
ovary in response to cyclic AMP and gonadotropins, expression in adipose tissue stimulated by prostaglandin E₂, and nonregulated expression in the placenta and adipose. Very specific inhibitors of P450arom have been developed, called “aromatase inhibitors,” that allow intense blockage of estrogen production, with clinical applications that include the treatment of breast cancer (e.g., anastrozole and letrozole) and dysfunctional uterine bleeding. The aromatase complex also includes NADPH-cytochrome P450 reductase, a ubiquitous flavoprotein involved in reduction reactions.

The 17β-hydroxysteroid dehydrogenase and 5a-reductase reactions are mediated by non-P450 enzymes. The 17β-hydroxysteroid dehydrogenase is bound to the endoplasmic reticulum and the 5a-reductase to the nuclear membrane. The 17β-hydroxysteroid dehydrogenase enzymes convert estrone to estradiol, androstenedione to testosterone, and DHEA to androstenediol, and vice versa. Eight different isozymes have been cloned and characterized.²⁴ The type 1 enzyme is active in the placenta and granulosa cells, converting estrone to estradiol. The type 2 and 4 enzymes, found in many tissues, form androstenedione and estrone from testosterone and estradiol, respectively. The type 3 and 5 enzymes in the testis reduce androstenedione to testosterone. The type 6 enzyme may be found only in rodents, and the type 7 and 8 enzymes are widespread, but have limited activity. Thus types 1,3, and 5 form active estrogens and androgens, whereas types 2 and 4 produce weaker products, a form of inactivation, important, for example, in protecting the fetus against testosterone and estradiol in the maternal circulation. The cell-specific production of each of these isoforms is a method for regulating the local concentration of estrogens and androgens.

The two-cell system is a logical explanation of the events involved in ovarian follicular steroidogenesis.²⁵ This explanation, first proposed by Falck in 1959,²⁶ brings together information on the site of specific steroid production, along with the appearance and importance of hormone receptors. The following facts are important:

These facts combine into the two-cell system to explain the sequence of events in ovarian follicular growth and steroidogenesis. The initial change from a primordial follicle to a preantral follicle is independent of hormones, and the stimulus governing this initial step in growth is unknown. Continued growth, however, depends on FSH stimulation. As the granulosa responds to FSH, proliferation and growth are associated with an increase in FSH receptors, a specific effect of FSH itself, but an action that is enhanced very significantly by the autocrine and paracrine peptides. The theca cells are characterized by steroidogenic activity in response to LH, specifically resulting in androgen production, by transcription of the P450scc, P450c17, and 3β-hydroxysteroid dehydrogenase genes. Aromatization of
androgens to estrogens is a distinct activity within the granulosa layer induced by FSH by activation of the P450arom gene. Androgens produced in the theca layer, therefore, must diffuse into the granulosa layer. In the granulosa layer they are converted to estrogens, and the increasing level of estradiol in the peripheral circulation reflects release of the estrogen back toward the theca layer and into blood vessels.

The theca and granulosa cells secrete peptides that operate as both autocrine and paracrine factors.²⁷ Insulin-like growth factor (IGF) is secreted by the theca and enhances the LH stimulation of androgen production in the theca cells as well as FSH-mediated aromatization in the granulosa. Evidence indicates that the endogenous insulin-like growth factor in the human ovarian follicle is IGF-II in both the granulosa and the theca cells.²⁸ Studies indicating activity of IGF-I with human ovarian tissue can be explained by the fact that both IGF-I and IGF-II activities can be mediated by the type I IGF receptor, which is structurally similar to the insulin receptor. The regulation of FSH receptors on granulosa cells is relatively complex. Although FSH increases the activity of its own receptor gene in a cyclic AMP-mediated mechanism, this action is influenced by inhibitory agents, such as epidermal growth factor, fibroblast growth factor, and even a gonadotropin-releasing hormone (GnRH)-like protein. Inhibin and activin are produced in the granulosa in response to FSH, and activin has the important autocrine role of enhancing FSH actions, especially the production of FSH receptors. Inhibin enhances LH stimulation of androgen synthesis in the
theca to provide a substrate for aromatization to estrogen in the granulosa, whereas activin suppresses androgen synthesis. This important paracrine regulation of androgen production in thecal cells by inhibin and activin, discussed in Chapter 6, is exerted primarily through modification of the expression of steroidogenic enzymes, especially P450c17.²⁹

After ovulation, the dominance of the luteinized granulosa layer is dependent on preovulatory induction of an adequate number of LH receptors, and, therefore, dependent on adequate FSH action. Prior to ovulation the granulosa layer is characterized by aromatization activity and conversion of the theca androgens to estrogens, an FSH-mediated activity. After ovulation the granulosa layer secretes progesterone and estrogens directly into the bloodstream, an LH-mediated activity.

Granulosa and theca cells each have an androgen aromatase system that can be demonstrated in vitro. However, in vivo, the activity of the granulosa layer in the follicular phase is several hundred times greater than the activity of the theca layer, and, therefore, the granulosa is the main biosynthetic source of estrogen in the growing follicle.³⁰ Because granulosa cells lack P450c17, the rate of aromatization in the granulosa layer is directly related to and dependent on the androgen substrate made available by the theca cells. Hence, estrogen secretion by the follicle prior to ovulation is the result of combined LH and FSH stimulation of the two cell types, the theca and the granulosa. After ovulation, it is believed the two cell types continue to function as a two-cell system; luteal cells derived from theca produce androgens for aromatization into estrogens by luteal cells derived from granulosa.

While circulating in the blood, a majority of the principal sex steroids, estradiol and testosterone, is bound to a protein carrier, known as sex hormone-binding globulin (SHBG) produced mainly in the liver. Another 30% is loosely bound to albumin, leaving only about 1% unbound and free. A very small percentage also binds to corticosteroid-binding globulin. Hyperthyroidism, pregnancy, and estrogen administration all increase SHBG levels, whereas corticoids, androgens, progestins, growth hormone, insulin, and IGF-I decrease SHBG.

The circulating level of SHBG is inversely related to body weight, and, thus, significant weight gain can decrease SHBG and produce important changes in the unbound levels of the sex steroids. Another important mechanism for a reduction in circulating SHBG levels is insulin resistance and hyperinsulinemia.^32,33 Thus, increased insulin levels in the circulation lower SHBG levels, and this may be the major mechanism that mediates the impact of increased body weight on SHBG. This relationship between the levels of insulin and SHBG is so strong that SHBG concentrations are a marker for hyperinsulinemic insulin resistance, and a low level of SHBG is a predictor for the development of type 2 diabetes mellitus.³⁴

The distribution of body fat has a strong influence on SHBG levels. Android or central fat is located in the abdominal wall and visceral-mesenteric locations. This fat distribution
is associated with hyperinsulinemia, hyperandrogenism, and decreased levels of SHBG.³⁵ The common mechanism for these changes is probably the hyperinsulinemia.

SHBG is a glycoprotein that contains a single binding site for androgens and estrogens, even though it is a homodimer composed of two monomers. Its gene has been localized to the short arm (p12-13) of chromosome 17.³⁶ Genetic studies have revealed that the SHBG gene also encodes the androgen-binding protein present in the seminiferous tubules, synthesized by the Sertoli cells.^37,38 Dimerization is believed to be necessary to form the single steroidbinding site. Specific chromosomal abnormalities with decreased or abnormal SHBG have not been reported. SHBG gene expression has now been identified in other tissues (brain, placenta, and endometrium), although a biologic significance has not been determined.

Transcortin, also called corticosteroid-binding globulin, is a plasma glycoprotein that binds cortisol, progesterone, deoxycorticosterone, corticosterone, and some of the other minor corticoid compounds. Normally about 75% of circulating cortisol is bound to transcortin, 15% is loosely bound to albumin, and 10% is unbound or free. Progesterone circulates in the following percentages: less than 2% unbound, 80% bound to albumin, 18% bound to transcortin, and less than 1% bound to SHBG. Binding in the circulation follows the law of mass action: the amount of the free, unbound hormone is in equilibrium with the bound hormone. Thus, the total binding capacity of a binding globulin will influence the amount that is free and unbound.

The biologic effects of the major sex steroids are largely determined by the unbound portion, known as the free hormone. In other words, the active hormone is unbound and free, whereas the bound hormone is relatively inactive. This concept is not without controversy. The hormone-protein complex may be involved in an active uptake process at the target cell plasma membrane.^39,40 and ⁴¹ The albumin-bound fraction of steroids may also be available for cellular action because this binding has low affinity. Because the concentration of albumin in plasma is manyfold greater than that of SHBG, the contribution of the albumin-bound fraction can be significant. Routine assays determine the total hormone concentration, bound plus free, and special steps are required to measure the active free level of testosterone, estradiol, and cortisol.

Androgens are the precursors of estrogens. 17β-Hydroxysteroid dehydrogenase activity converts androstenedione to testosterone, which is not a major secretory product of the normal ovary. It is rapidly demethylated at the C-19 position and aromatized to estradiol, the major estrogen secreted by the human ovary. Estradiol also arises to a major degree from androstenedione via estrone, and estrone itself is secreted in significant daily amounts. Estriol is the peripheral metabolite of estrone and estradiol and not a secretory product of the ovary. The formation of estriol is typical of general metabolic “detoxification,” conversion of biologically active material to less active forms.

The conversion of steroids in peripheral tissues is not always a form of inactivation. Free androgens are peripherally converted to free estrogens, for example, in skin and adipose cells. The location of the adipose cells influences their activity. Women with central obesity (the abdominal area) produce more androgens.⁴² The work of Siiteri and MacDonald⁴³ demonstrated that enough estrogen can be derived from circulating androgens to produce bleeding in the postmenopausal woman. In the female the adrenal gland remains the major source of circulating androgens, in particular androstenedione. In the male, almost all of the circulating estrogens are derived from peripheral conversion of androgens. The precursor
androgens consist principally of androstenedione, dehydroepiandrosterone, and dehydroepiandrosterone sulfate.

It can be seen, therefore, that the pattern of circulating steroids in the female is influenced by the activity of various processes outside the ovary. Because of the peripheral contribution to steroid levels, the term secretion rate is reserved for direct organ secretion, whereas production rate includes organ secretion plus peripheral contribution via conversion of precursors. The metabolic clearance rate (MCR) equals the volume of blood that is cleared of the hormone per unit of time. The blood production rate (PR) then equals the metabolic clearance rate multiplied by the concentration of the hormone in the blood.

In the normal nonpregnant female, estradiol is produced at the rate of 100-300 μg/day. The production of androstenedione is about 3 mg/day, and the peripheral conversion (about 1%) of androstenedione to estrone accounts for about 20-30% of the estrone produced per day. Because androstenedione is secreted in milligram amounts, even a small percent conversion to estrogen results in a significant contribution to estrogens, which exist and function in the circulation in picogram amounts. Thus, the circulating estrogens in the female are the sum of direct ovarian secretion of estradiol and estrone, plus peripheral conversion of C-19 precursors. Whereas estradiol is produced in μg amounts, it circulates and functions within cells in concentrations of pg/mL.

Peripheral conversion of steroids to progesterone is not seen in the nonpregnant female; rather, the progesterone production rate is a combination of secretion from the adrenal and the ovaries. Including the small contribution from the adrenal, the blood production rate of progesterone in the preovulatory phase is less than 1 mg/day. During the luteal phase, production increases to 20-30 mg/day. The metabolic fate of progesterone, as expressed by its many excretion products, is more complex than estrogen. About 10-20% of progesterone is excreted as pregnanediol.

Pregnanediol glucuronide is present in the urine in concentrations less than 1 mg/day until ovulation. Postovulation pregnanediol excretion reaches a peak of 3-6 mg/day, which is maintained until 2 days prior to menses. The assay of pregnanediol in the urine now has little clinical use.

In the preovulatory phase in adult females, in all prepubertal females, and in the normal male, the blood levels of progesterone are at the lower limits of immunoassay sensitivity: less than 1 ng/mL. After ovulation, i.e., during the luteal phase, progesterone ranges from 3 to 15 ng/mL. In congenital adrenal hyperplasia, progesterone blood levels can be as high as 50 times above normal.

Pregnanetriol is the chief urinary metabolite of 17a-hydroxyprogesterone and has clinical significance in the adrenogenital syndrome, a syndrome of virilizing adrenal hyperplasia in which an enzyme defect results in accumulation of 17a-hydroxyprogesterone and increased excretion of pregnanetriol. The inherited abnormality in virilizing adrenal hyperplasia results in an inability to synthesize glucocorticoids. The hypothalamic-pituitary axis reacts to the low level of cortisol by elevated ACTH secretion in a homeostatic response to achieve normal levels of cortisol production. This stimulation induces a hyperplastic adrenal cortex that produces androgens as well as corticoid precursors in abnormal quantities. The plasma or serum assay of 17a-hydroxyprogesterone is a more sensitive and accurate index of this enzyme deficiency than measurement of pregnanetriol. Normally, the blood level of 17a-hydroxyprogesterone is less than 100 ng/dL, although after ovulation and during the luteal phase of a normal menstrual cycle, a peak of 200 ng/dL can be reached. In syndromes of adrenal hyperplasia, values can be 10-400 times normal.

The major androgen products of the ovary are dehydroepiandrosterone (DHEA) and androstenedione (and only a little testosterone), which are secreted mainly by stromal tissue derived from theca cells. With excessive accumulation of stromal tissue or in the presence of an androgen-producing tumor, testosterone becomes a significant secretory product. Occasionally, a nonfunctioning tumor can induce stromal proliferation and increased androgen production. The normal accumulation of stromal tissue at midcycle
results in a rise in circulating levels of androstenedione and testosterone at the time of ovulation.

The adrenal cortex produces three groups of steroid hormones: the glucocorticoids, the mineralocorticoids, and the sex steroids. The adrenal sex steroids represent intermediate byproducts in the synthesis of glucocorticoids and mineralocorticoids, and excessive secretion of the sex steroids occurs only with neoplastic cells or in association with enzyme deficiencies. Under normal circumstances, adrenal gland production of the sex steroids is less significant than gonadal production of androgens and estrogens. About one-half of the daily production of DHEA and androstenedione comes from the adrenal gland; the other half of androstenedione is secreted by the ovary, but the other half of DHEA is split almost equally between the ovary and peripheral tissues. The production rate of testosterone in the normal female is 0.2-0.3 mg/day, and approximately 50% arises from peripheral conversion of androstenedione (and a small amount from DHEA) to testosterone, whereas 25% is secreted by the ovary and 25% by the adrenal. The major androgens are excreted in the urine as 17-ketosteroids.

There is no circadian cycle of the major sex steroids in the female. However, short-term variations in the blood levels due to episodic secretion require multiple sampling for absolutely accurate assessment. Although frequent sampling is necessary for a high degree of accuracy, a random sample is sufficient for clinical purposes to determine whether a level is within a normal range.

The testosterone-binding capacity is decreased by androgens; hence, the binding capacity in men is lower than that in normal women. The binding globulin level in women with increased androgen production is also depressed. Androgenic effects are dependent on the unbound fraction that can move freely from the vascular compartment into the target cells. Routine assays determine the total hormone concentration, bound plus free. Thus, a total testosterone concentration can be in the normal range in a woman who is hirsute or even virilized, but because the binding globulin level is depressed by the androgen effects, the percent free and active testosterone is elevated. The need for a specific assay for the free portion of testosterone can be questioned because the very presence of hirsutism or virilism indicates increased androgen effects. In the face of hirsutism, one can reliably interpret a normal testosterone level as compatible with decreased binding capacity and increased active free testosterone.

Both total and unbound testosterone are normal in only a few women with hirsutism. In these cases, the hirsutism, heretofore regarded as idiopathic, most likely results from excessive intracellular androgen effects (specifically increased intracellular conversion of testosterone to dihydrotestosterone).

Reduction of the Δ⁴ unsaturation (an irreversible pathway) in testosterone is very significant, producing derivatives very different in their spatial configuration and activity. The 5β-derivatives are not androgenic, and this is not an important pathway; however, the 5a-derivative (a very active pathway) is extremely potent. Indeed, dihydrotestosterone (DHT), the 5a-derivative, is the principal androgenic hormone in a variety of target tissues and is formed within the target tissue itself.

In men, the majority of circulating DHT is derived from testosterone that enters a target cell and is converted by means of 5a-reductase to DHT. In women, because the production rate of androstenedione is greater than testosterone, blood DHT is primarily derived from androstenedione and partly from dehydroepiandrosterone.⁴⁴ Thus, in women, the skin production of DHT is predominantly influenced by androstenedione. DHT is by definition an intracrine hormone, formed and acting within target tissues.⁴⁵ The 5a-reductase enzyme exists in two forms, type I and II, each encoded by a separate gene, with the type I enzyme found in skin and the type II reductase predominantly expressed in reproductive tissues.⁴⁶

DHT is largely metabolized intracellularly; hence, the blood DHT is only about one-tenth the level of circulating testosterone, and it is clear that testosterone is the major circulating androgen. In tissues sensitive to DHT (which includes hair follicles), only DHT enters the nucleus to provide the androgen message. DHT also can perform androgenic actions within cells that do not possess the ability to convert testosterone to DHT. DHT is further reduced by a 3a-keto-reductase to androstanediol, which is relatively inactive. The metabolite of androstanediol, 3a-androstanediol glucuronide, is the major metabolite of DHT and can be measured in the plasma, indicating the level of activity of target tissue conversion of testosterone to DHT.

Not all androgen-sensitive tissues require the prior conversion of testosterone to DHT. In the process of masculine differentiation, the development of the wolffian duct structures (epididymis, the vas deferens, and the seminal vesicle) is dependent on testosterone as the intracellular mediator, whereas development of the urogenital sinus and urogenital tubercle into the male external genitalia, urethra, and prostate requires the conversion of testosterone to DHT.⁴⁷ Muscle development is under the direct control of testosterone. Testosterone is also aromatized to a significant extent in the brain, liver, and breast; and in some circumstances (e.g., in the brain) androgenic messages can be transmitted via estrogen.

The characteristics of sex steroid metabolism reviewed above contribute to an important clinical concept: circulating levels of sex hormones do not always reflect concentrations in target cells. In premenopausal women, target tissues synthesize and metabolize most of the testosterone produced. Thus, in women testosterone functions as a paracrine and intracrine hormone. In men, abundant secretion of testosterone creates circulating levels that are sufficient to allow testosterone to function as a classic hormone. In women, the same description applies to estradiol. Estradiol functions as a classical circulating hormone until menopause, after which both estradiol and testosterone activities are due to local target tissue synthesis, using precursors derived from the circulation. Clinical interventions after menopause, therefore, are directed to local hormone production; for example, the use of aromatase inhibitors to treat breast cancer.

Active steroids and metabolites are excreted as sulfo and glucuro conjugates. Conjugation of a steroid converts a hydrophobic compound into a hydrophilic one and generally reduces or eliminates the activity of a steroid. This is not completely true, however, because hydrolysis of the ester linkage can occur in target tissues and restore the active form. Furthermore, estrogen conjugates can have biologic activity, and it is known that sulfated conjugates are actively secreted and may serve as precursors, present in the circulation in relatively high concentrations because of binding to serum proteins. Ordinarily, however, conjugation by liver and intestinal mucosa is a step in deactivation preliminary to, and essential for, excretion into urine and bile.

Hormones circulate in extremely low concentrations and, in order to respond with specific and effective actions, target cells require the presence of special mechanisms. There are two major types of hormone action at target tissues. One mediates the action of tropic hormones (peptide and glycoprotein hormones) with receptors at the cell membrane level. In contrast, the smaller steroid hormones enter cells readily, and the basic mechanism of action involves specific receptor molecules within the cells. It is the affinity, specificity, and activity of the receptors, together with the large concentration of receptors in cells, that allow a small amount of hormone to produce a biologic response.

The many different types of receptors can be organized into the following basic categories:

The specificity of the reaction of tissues to sex steroid hormones is due to the presence of intracellular receptor proteins. Different types of tissues, such as liver, kidney, and uterus, respond in a similar manner. The mechanism includes: (1) steroid hormone diffusion across the cell membrane, (2) steroid hormone binding to a receptor protein, (3) interaction of a hormone-receptor complex with nuclear DNA, (4) synthesis of messenger RNA (mRNA), (5) transport of the mRNA to the ribosomes, and finally, (6) protein synthesis in the cytoplasm that results in specific cellular activity. The steroid hormone receptors primarily affect gene transcription, but also regulate posttranscriptional events and nongenomic events. Steroid receptors regulate gene transcription through multiple mechanisms, not all of which require direct interactions with DNA.

Each of the major classes of the sex steroid hormones, including estrogens, progestins, and androgens, act according to this general mechanism. Glucocorticoid and mineralocorticoid receptors, when in the unbound state, reside in the cytoplasm and move into the nucleus after hormone-receptor binding. Estrogens, progestins, and androgens are transferred across the nuclear membrane and bind to their receptors within the nucleus.

Steroid hormones are rapidly transported across the cell membrane by simple diffusion. The factors responsible for this transfer are unknown, but the concentration of free (unbound) hormone in the bloodstream seems to be an important and influential determinant of cellular function. Once in the cell, the sex steroid hormones bind to their individual receptors. During this process, transformation or activation of the receptor occurs. Transformation refers to a conformational change of the hormone-receptor complex revealing or producing a binding site that is necessary in order for the complex to bind to the chromatin. In the unbound state, the receptor is associated with heat shock proteins that stabilize and protect the receptor and maintain a conformational shape that keeps the DNA binding region in an inactive state. Activation of the receptor is driven by hormone binding that causes a dissociation of the receptor-heat shock protein complex.

The hormone-receptor complex binds to specific DNA sites (hormone-responsive elements) that are located upstream of the gene. The specific binding of the hormone-receptor complex with DNA results in RNA polymerase initiation of transcription. Transcription leads to translation, mRNA-mediated protein synthesis on the ribosomes. The principal action of steroid hormones is the regulation of intracellular protein synthesis by means of the receptor mechanism.

Biologic activity is maintained only while the nuclear site is occupied with the hormonereceptor complex. The dissociation rate of the hormone and its receptor as well as the halflife of the nuclear chromatin-bound complex are factors in the biologic response because the hormone response elements are abundant and, under normal conditions, are occupied only to a small extent.⁴⁸ Thus, an important clinical principle is the following: duration of exposure to a hormone is as important as dose. One reason only small amounts of estrogen need be present in the circulation is the long half-life of the estrogen hormone-receptor complex. Indeed, a major factor in the potency differences among the various estrogens (estradiol, estrone, estriol) is the length of time the estrogen-receptor complex occupies the nucleus. The higher rate of dissociation with the weak estrogen (estriol) can be compensated for by continuous application to allow prolonged nuclear binding and activity. Cortisol and progesterone must circulate in large concentrations because their receptor complexes have short half-lives in the nucleus.

An important action of estrogen is the modification of its own and other steroid hormone activity by affecting receptor concentrations. Estrogen increases target tissue responsiveness to itself and to progestins and androgens by increasing the concentration of its own receptor and that of the intracellular progestin and androgen receptors. Progesterone and clomiphene, on the other hand, limit tissue response to estrogen by blocking this mechanism, thus decreasing over time the concentration of estrogen receptors. Small amounts of receptor depletion and small amounts of steroid in the blood activate the mechanism.

The synthesis of the sex steroid receptors obviously takes place in the cytoplasm, but with estrogen and progestin receptors, synthesis must be quickly followed by transportation into the nucleus. There is an amazingly extensive nuclear traffic.⁴⁹ The nuclear membrane contains 3,000 to 4,000 pores. A cell synthesizing DNA imports about one million histone molecules from the cytoplasm every 3 minutes. If the cell is growing rapidly, about three newly assembled ribosomes will be transported every minute in the other direction. The typical cell can synthesize 10,000 to 20,000 different proteins. How do they know where to go? The answer is that these proteins have localization signals. In the case of steroid hormone receptor proteins, the signal sequences are in the hinge region.

Estrogen and progestin receptors exit continuously from the nucleus to the cytoplasm and are actively transported back to the nucleus. This is a constant shuttle; diffusion into the cytoplasm is balanced by the active transport into the nucleus. This raises the possibility that some diseases are due to poor traffic control. This can be true of some acquired diseases as well, e.g., Reye’s syndrome, an acquired disorder of mitochondrial enzyme function.

The fate of the hormone-receptor complex after gene activation is referred to as hormone-receptor processing. In the case of estrogen receptors, processing involves the rapid degradation of receptors unbound with estrogen, and a much slower degradation of bound receptors after gene transcription. The rapid turnover of estrogen receptors has clinical significance. The continuous presence of estrogen is an important factor for continuing response.

The best example of the importance of these factors is the difference between estradiol and estriol. Estriol has only 20-30% affinity for the estrogen receptor compared with estradiol; therefore, it is rapidly cleared from a cell. But if the effective concentration is kept equivalent to that of estradiol, it can produce a similar biologic response.⁵⁰ In pregnancy, where the concentration of estriol is very great, it can be an important hormone, not just a metabolite.

The depletion of estrogen receptors in the endometrium by progestational agents is the fundamental reason for adding progestins to estrogen treatment programs. The progestins
accelerate the turnover of preexisting receptors, and this is followed by inhibition of estrogen-induced receptor synthesis. Using monoclonal antibody immunocytochemistry, this action has been pinpointed to the interruption of transcription in estrogen-regulated genes. The mechanism is different for androgen antiestrogen effects. Androgens also decrease estrogen receptors within target tissues, especially in the uterus.^51,52