STEROID HORMONE METABOLISM
Division of Clinical Psychopharmacology, University Institute of Psychiatry,
2, ch. du Petit Bel-Air, 1225 Chêne-Bourg, Geneva, Switzerland
Steroids are lipophilic, low-molecular weight compounds derived from cholesterol that play a number of important physiological roles. The steroid hormones are synthesized mainly by endocrine glands such as the gonads (testis and ovary), the adrenals and (during gestation) by the fetoplacental unit, and are then released into the blood circulation. They act both on peripheral target tissues and the central nervous system (CNS). An important function of the steroid hormones is to coordinate physiological and behavioural responses for specific biological purposes, e.g. reproduction. Thus, gonadal steroids influence the sexual differentiation of the genitalia and of the brain, determine secondary sexual characteristics during development and sexual maturation, contribute to the maintenance of their functional state in adulthood and control or modulate sexual behaviour. It has been recently discovered (review in ref. 1) that, in addition to the endocrine glands, the CNS is also able to form a number of biologically active steroids directly from cholesterol (the so-called "neurosteroids"). These neurosteroids, however, are more likely to have "autocrine" or "paracrine" functions rather than true endocrine effects.
Despite their relatively simple chemical structure, steroids occur in a wide variety of biologically active forms. This variety is not only due to the large range of compounds secreted by steroid-synthesizing tissues, but also to the fact that circulating steroids are extensively metabolised peripherally, notably in the liver, and in their target tissues, where conversion to an active form is sometimes required before they can elicit their biological responses. Steroid metabolism is therefore important not only for the production of these hormones, but also for the regulation of their cellular and physiological actions. This chapter will consider both aspects of steroid metabolism. The emphasis will be on the functional and biological significance of metabolism in endocrine physiology rather than on an extensive description of the metabolic pathways and the enzymes involved.
Steroid hormones: Structure, nomenclature and classification
The parent compound from which all steroids are derived is cholesterol. As shown in Fig. 1a, cholesterol is made up of three hexagonal carbon rings (A,B,C) and a pentagonal carbon ring (D) to which a side-chain (carbons 20-27) is attached (at position 17 of the polycyclic hydrocarbon). Two angular methyl groups are also found at position 18 and 19. Removal of part of the side-chain gives rise to C21-compounds of the pregnane series (progestins and corticosteroids). Total removal produces C19-steroids of the androstane series (including the androgens), whereas loss of the 19-methyl group (usually after conversion of the A-ring to a phenolic structure, hence the term "aromatization") yields the estrane series, to which estrogens belong. Individual compounds are characterised by the presence or absence of specific functional groups (mainly hydroxy, keto(oxo) and aldehyde functions for the naturally occurring steroids) at certain positions of the carbon skeleton (particularly at positions 3,5,11,17,18,20 and 21).
Given that at most positions, the functional groups can be oriented either in equatorial or axial position (see Fig. 1b), this type of structure gives rise to a great number of possible stereoisomers (i.e. molecules having the same chemical formula, but a different three-dimensional conformation). Stereoisomerism is very important for biological activity (i.e. for steroid-protein interactions). Apart from a few exceptions, steroids are rather " flat " molecules: the hexagonal carbon rings (A to C) usually assume a " boat " rather than a " chair " form and are mostly fused in the trans- conformation (Fig. 1b, top). Substituent groups above the plane of the molecule are said to be in the " ß " position, whereas those situated under the plane of the molecule are said to be in the " α " position. A complete description of a steroid molecule must therefore include the name of the parent compound (pregnane, androstane or estrane series), and the name, number, position and orientation (α or ß) of all functional groups. Commonly occurring steroids are usually identified by a " trivial " name (e.g. cortisol, testosterone, etc.). However, this can lead to confusion in some cases, and the use of systematic names (based on IUPAC rules, i.e. including all the information required to fully characterize each molecule) is recommended. According to this nomenclature, double-bonds are indicated by the suffix -ene, with their position. Thus, testosterone (trivial name) becomes " 17ß-hydroxy-androst-4-ene-3-one " (further details can be found in refs. 3,4 and 9).
Steroid hormones can be grouped in various classes according to a number of criteria. Based on their chemical structure they can belong to one of the classes (series) mentioned above (e.g. " a pregnane derivative "). If their site of production is considered to be more important, one can distinguish for example between " ovarian " or " adrenal " steroids. If their biological function is essential, terms like " a glucocorticoid " or " sex steroids " can be used. Finally, classification can also be based on their molecular actions (" an estrogen-receptor agonist "), or biochemical effects.
Steroid hormone biosynthesis
A general outline of the major biosynthetic pathways
The adrenals produce both androgens and corticosteroids (mineralo- and glucocorticoids), the ovaries (depending on the stage of the ovarian cycle) can secrete estrogens and progestins, and the testis mainly androgens. However, the biochemical pathways involved are strikingly similar in all tissues, the difference in secretory capacity being mostly due to the presence or absence of specific enzymes. It is therefore possible to give a general outline of the major biosynthetic pathways which is applicable to all steroid-secreting glands, as shown in Fig. 2.
From acetate to cholesterol.
Cholesterol can be synthetised in all steroid-producing tissues from acetate, but the main production sites are the liver, the skin and the intestinal mucosa. Steroid hormone formation in endocrine glands probably relies mostly on exogenous cholesterol (plasma cholesterol). The 27-carbon skeleton of cholesterol is derived from acetyl-CoA through a series of reactions which involve the following intermediate products: (1) Mevalonate (by condensation of 3 molecules of acetyl-CoA), which requires the enzyme HMG-CoA-reductase, an important enzyme in the control of cholesterol biosynthesis; (2) Squalene, a 30-carbon linear structure which undergoes cyclization to yield (3) Lanosterol; and (4) after removal of 3 carbons, cholesterol. In addition to the relevant enzymes, the final steps of cholesterol biosynthesis (from squalene onwards) require sterol carrier proteins (SCP) to ensure the solubilisation of these highly lipophilic molecules (see ref. 9, pp. 67-79, for more details on the acetate-cholesterol pathway).
From cholesterol to progestins, androgens and estrogens.
The first committed step in steroid biosynthesis is the conversion of the 27-carbon skeleton of cholesterol to a C21-compound, pregnenolone (Fig. 2). This critical step, which is subject to hormonal control by the adrenocorticotropic hormone (ACTH) in the adrenals and by the luteinizing hormone (LH) in the gonads, is catalysed by a P-450 enzyme, the cholesterol side-chain cleavage enzyme P-450scc (also called 20,22-desmolase, or 20,22-lyase). Pregnenolone can be converted either to progesterone, which branches to the glucocorticoid and androgen/estrogen pathways, or to 17α-hydroxypregnenolone, which is another route for the formation of androgens and estrogens (Fig. 2, top-left part). Androgen formation in the adrenals is limited to dehydroepiandrosterone and androstenedione, whereas in the testes the presence of 17ß-hydroxysteroid dehydrogenase (17HSD) in Leydig cells (under the control of LH) ensures the formation of testosterone, the principal " male " hormone. Estrogen formation requires another P-450 enzyme, the aromatase complex (P-450Arom). The substrate is either androstenedione (for estrone) or testosterone (for estradiol). Estrone and estradiol are interconvertible through a reversible reaction involving another 17ß-hydroxysteroid dehydrogenase, as in the androstenedione-testosterone conversion. Aromatase activity is present in the ovary and the placenta (see below). In the ovary, aromatase activity and estrogen formation occur in granulosa cells and are controlled by the follicle-stimulating hormone (FSH), whereas production of the androgenic substrates (testosterone, 4-androstenedione) requires LH stimulation of the theca cells (5).
From progesterone and 17α-hydroxyprogesterone to gluco- and mineralocorticoids.
Hydroxylation of progesterone at carbon 21 yields 11-deoxycorticosterone (DOC), and corticosterone after another hydroxylation step at carbon 11. Corticosterone is a major glucocorticoid in rats and other species which do not produce cortisol. Two further steps (hydroxylation and oxydoreduction at carbon 18) result in the formation of aldosterone.
Cortisol is formed from 17α-hydroxyprogesterone, with 11-deoxycortisol as an intermediate. Cortisol is the main glucocorticoid secreted by human adrenal glands.
Steroid biosynthesis in the fetoplacental unit.
Progesterone is produced by the corpus luteum during the first 6-8 weeks of gestation, but during pregnancy the main source for this steroid is the placenta. Estrogen levels (and that of its metabolite estriol) rise markedly during gestation. The substrate for estrogen biosynthesis in the fetoplacental unit is dehydroepiandrosterone sulfate (DHEAS) which is obtained from the fetomaternal bloodstream (see ref. 5, pp. 191-194, for more details).
Enzymes involved in steroid biosynthesis
The reactions shown in Fig. 2 are catalysed by a limited number of enzymes (listed on the left-hand side of the figure), which belong to four main classes:
- Desmolases (or lyases): these enzymes catalyse reactions which result in the removal of parts of the original cholesterol side-chain. This involves sequential hydroxylation of adjacent C (e.g., of C-20 and C-22 for P-450scc) and requires a cytochrome P-450, molecular oxygen (O2) and nicotinamide dinucleotide phosphate, reduced form (NADPH) as a cofactor. These enzymes are located in the mitochondria and are linked to an electron transport system (9).
- Hydroxylases: these enzymes are membrane-bound and are present either in the mitochondrial or in the microsomal fraction of the cell. They also require a cytochrome P-450, molecular oxygen and NADPH, as for lyases.
- Hydroxysteroid dehydrogenases (oxido-reductases): these enzymes catalyse reversible reactions and depend either on NADP(H) or NAD(H). They are found both in the cell cytosol and in the microsomal fraction.
- Aromatase: conversion of the A-ring to a phenolic structure (i.e. with a phenolic HO-group at C-3), a process known as " aromatization ", involves a complex sequence of hydroxylation reactions and loss of the angular C-19 methyl group (10). Aromatase activity is mainly found in the ovary, the placenta and the brain, and is also membrane-bound. Its substrate is either 4-androstenedione or testosterone.
The gene encoding human aromatase cytochrome P-450 has been cloned recently and its expression has been shown to be regulated by tissue-specific promoters (8).
Disorders resulting from defects in steroid biosynthesis
A number of endocrine disorders can be attributed to specific enzyme defects. Thus, inability to secrete normal levels of adrenals steroids may result in congenital adrenal hyperplasia (CAH) following hyperstimulation by ACTH (the negative steroid feed-back controlling adrenal activity being lost). In the majority of cases, this syndrome is due to 21-hydroxylase deficiency, and is associated with increased adrenal androgen secretion and partial virilization in girls (5). Less common adrenal enzyme deficiencies involve either 17-hydroxylase (with a possible increase in mineralocorticoid levels) or 18-hydroxylase (aldosterone may be deficient with normal levels of cortisol). Defects in testicular androgen synthesis (17,20-desmolase or 17ß-hydroxysteroid dehydrogenase deficiency) can lead to male pseudohermaphroditism. However, in more than 80% of the cases, male pseudohermaphroditism is assumed to result from abnormalities in androgen action at the target cell level (see the possible role of metabolic defects in the relevant section below).
Steroid hormones in the blood
It is generally assumed that steroids are released into the blood circulation as soon as they are formed, i.e. there are no active transport and/or release mechanisms. Secretion rates are therefore directly related to the biosynthetic activity of the gland and to the blood flow rate.
Steroid binding proteins
Because of their lipophilic properties, free steroid molecules are only sparingly soluble in water. In biological fluids, they are usually found either in a conjugated form, i.e. linked to a hydrophilic moiety (e.g. as sulfate or glucuronide derivatives) or bound to proteins (non-covalent, reversible binding). In the plasma, unconjugated steroids are found mostly bound to carrier proteins (6). Binding to plasma albumin (which accounts for 20-50% of the bound fraction) is rather unspecific, whereas binding to either corticosteroid-binding globulin (CBG) or the sex hormone-binding globulin (SHBG) [sometimes called " sex steroid-binding protein ", or SBP] is based on more stringent stereospecific criteria. The " free fraction " (1-10% of total plasma concentration) is usually considered to represent the biologically active fraction (i.e. hormone that is directly available for action), although this idea has been challenged by recent evidence that, in some cases at least, the specific binding proteins may facilitate steroid entry into target tissues. Apart from the two functions mentioned above, the major roles of plasma binding proteins seem to be (a) to act as a " buffer " or reservoir for active hormones (because of the non-covalent nature of the binding, protein-bound steroids are released into the plasma in free form as soon as the free concentration drops according to the law of mass action) and (b) to protect the hormone from peripheral metabolism (notably by liver enzymes) and increase the half-life of biologically active forms.
Peripheral metabolism of circulating steroids
Because steroids are lipophilic, they diffuse easily through the cell membranes, and therefore have a very large distribution volume. In their target tissues, steroids are concentrated by an uptake mechanism which relies on their binding to intracellular proteins (or " receptors ", see below). High concentration of steroids are also found in adipose tissue, although this is not a target for hormone action. In the human male, adipose tissue contains aromatase activity, and seems to be the main source of androgen-derived estrogens found in the circulation. But most of the peripheral metabolism occurs in the liver and to some extent in the kidneys, which are the major sites of hormone inactivation and elimination, or catabolism (see below).
Steroid interaction with target tissues
Genomic versus non-genomic action of steroids
Steroids have both short- and long-term effects. Long-term effects (lasting from hours to days) usually involve interaction of the hormone with a specific intracellular steroid-binding protein called a receptor. These receptors are DNA-binding proteins of the steroid/thyroid hormone receptor superfamily (2). The steroid-receptor complex binds to hormone-responsive elements on the chromatin and regulates gene transcription (12). Steroid receptor genes are only expressed in target tissues, where their presence determines accumulation of the hormone in the cell nucleus and facilitates steroid entry into the target cell by the law of mass action. This mode of cellular action is generally referred to as a genomic action. Non-genomic action, on the other hand, is any mode of action for which gene transcription is not directly implicated, e.g. rapid steroid effects on the electrical activity of nerve cells, or interaction with the γ-aminobutyric acid (GABA A) receptor (11). In contrast to the genomic effects, non-genomic effects require the continued presence of the hormone. Some of these effects may involve specific receptors located on the cell membrane (11).
Formation of active metabolites in target tissues
For certain classes of hormones and particular target tissues, steroids must be converted in situ to an active form before they can interact with their specific receptor(s). This metabolic activation step is either an absolute prerequisite or a way of achieving a range of complex effects which involve interaction with more than one type of receptor. In the first case, metabolism has a permissive role, whereas in the second case it modulates steroid action, usually depending on particular physiological and/or environmental conditions. Two examples are shown in Fig. 3. Conversion of testosterone to 5α-DHT (Fig. 3, top) is required for its action on prostate growth and function, whereas aromatization to estradiol-17ß in the brain is mandatory for some of its developmental, neuroendocrine and behavioural effects. Unlike its parent compound, the progesterone metabolite 5α-DHP (Fig. 3, bottom) has no effect on the uterus, but is more effective than progesterone itself regarding the facilitation and/or inhibition of GnRH-induced LH release in vitro (7). It also has a barbiturate-like action on brain GABA A receptors, as the other metabolite shown in Fig. 3 (bottom right), pregnanolone (5ß-pregnane-3ß-ol-20-one). The two main classes of hormones for which metabolic activation has been shown to play a role are the progestins and the androgens, but catecholestrogens (2- or 4-OH derivatives of estrogens) may also constitute another class of biologically active compounds resulting from target organ metabolism. When conversion of the circulating hormone is required for its action, the original compound is sometime called a prehormone.
Enzymes involved in metabolic activation usually catalyse irreversible conversion steps and are often rate-limiting for steroid action, i.e. the effect depends more on the conversion rate than on the number of steroid-binding sites (receptors) available. Steroid metabolism in target tissues may be critical for determining both the specificity and the magnitude of hormone effects.
Correlation between structure and function: the role of metabolism
The biological activity of a steroid molecule depends on its ability to interact with a specific binding site on the corresponding receptor. In most cases, biological activity can be directly correlated with binding affinity. The affinity (usually characterised by the binding constant KD, which is the molar concentration required to saturate half of the available binding sites) of a steroid for its specific receptor is dependent upon the presence or absence of particular functional groups and the overall three-dimensional structure of the molecule. Stereoisomerism may play an important role in this respect: molecules with the same chemical composition but a different spatial orientation of their substituents at critical points (e.g. at C-5) may have totally different binding properties and biological effects. Thus, 5α-reduced dihydrotestosterone (DHT) is a potent androgen, with a strong affinity for intracellular androgen receptors, whereas its 5ß-epimers does not bind to these receptors and is totally devoid of androgenic properties, but is an effective inducer of hematopoiesis. Isomerisation can therefore lead either to inactivation or to a change in the specific biological properties of the original molecule.
The importance of even minor changes in the structure of a steroid molecule for its biological activity explains why target tissue metabolism may play such a critical role in modulating hormone action at the cell level. Since the activity of most enzymes is regulated by a number of factors (in particular hormonal factors related to the endocrine status), and since this activity is often rate-limiting for steroid action, target tissue metabolism provides an additional degree of control over steroid hormone action. It should be mentioned here that target tissue metabolism is not limited to the local production of active metabolites: inactivation can also occur within the target cell, and this mechanism can contribute to the regulation of the intracellular concentration of biologically active molecules. Thus, the hormonal " micro environment " of a steroid-target cell is determined by a complex interplay between activating and inactivating mechanisms.
Disorders resulting from defects in target tissue metabolism
Various disorders can result from a genetic defect in target tissue metabolism. The best known example is male pseudohermaphroditism (i.e. 46,XY individuals with a feminised phenotype) due to 5α-reductase deficiency. Individuals with this autosomal recessive disorder have normal plasma levels of testosterone, but this hormone cannot be converted to its active metabolite, 5α-DHT, in the target tissues and is ineffective. This type of androgen resistance syndrome results notably in an abnormal sexual differentiation of the male genitalia.
Steroid inactivation and catabolism
Inactivation refers to the metabolic conversion of a biologically active compound into an inactive one. Inactivation can occur at various stages of hormone action. Peripheral inactivation (e.g. by liver enzymes) is required to ensure steady-state levels of plasma hormones as steroids are more or less continuously secreted into the bloodstream. Moreover, if a hormone is to act as a " chemical signal ", its half-life in the circulation must be limited, so that any change in secretion rate is immediately reflected by a change in its plasma concentration (particularly when secretion rates are decreased). But hormone inactivation can also occur in target tissues, notably after the hormone has triggered the relevant biological effects in order to ensure termination of hormone action.
The main site of peripheral steroid inactivation and catabolism is the liver, but some catabolic activity also occurs in the kidneys. Inactive hormones are mainly eliminated as urinary (mostly conjugated) metabolites. Usually, steroids are eliminated once they have been inactivated (i.e., they are not " recycled "). This elimination (e.g. as a urinary excretion products) requires conversion to hydrophilic compounds in order to ensure their solubility in biological fluids at rather high concentrations. Depending on the structure of the starting steroid, the following reactions may be involved (4):
- Reduction of a double bond at C-4 and reduction of an oxo(keto) group at C-3 to a secondary alcoholic group.
- Reduction of an oxo group at C-20 to a secondary alcoholic group.
- Oxidation of a 17ß-hydroxyl group.
- Further hydroxylations at various positions of the steroid nucleus (e.g. 7-hydroxylation of 5α-reduced androgens).
- Conjugation (sulphate and/or glucuronide derivatives).
A few examples of steroid excretion products are shown in Table 1.
Formation of steroid conjugates
Conjugation (formation of hydrophilic molecules) is an important step in steroid catabolism. Most excretory products are in conjugated form. Two major pathways are used:
(1) Formation of glucuronides. This reaction requires uridine diphosphoglucuronic acid (UDPGA) and a glucuronyl transferase. Glucuronic acid is attached to a HO-group on the steroid molecule:
Steroid-OH + UDPGA — glucuronyl transferase→ Steroid glucuronide
(2) Formation of sulphates. This conversion is catalysed by sulphokinases, which occur in the cytosol of liver, testicular, adrenal and fetal tissues. The substrates are steroids with an HO-group and phosphoadenosine-5’-phosphosulphate (PAPS). This is a three-step reaction which requires magnesium (Mg++) ions:
(1) SO42- + ATP — ATP sulphurylase→ APS + P-Pi
(2) APS + ATP — ATP kinase→ PPAPS
(3) Steroid-OH + PAPS — ATP sulphokinase→ Steroid-O-SO3- + PAP + H+
where ATP is adenosine triphosphate, APS adenosine-5’-phosphate, PAP 3’,5’-phosphoadenosine and P-Pi pyrophosphate.
Two examples of conjugated derivatives are shown in Fig. 4. In addition to being excretory products, sulphates are also found in endocrine tissues and/or the plasma as precursors for hormone synthesis. This is the case of dehydroepiandrosterone sulphate (DHEAS), which is used notably for estrogen biosynthesis in the fetoplacental unit (see above). Sulphatases occurring in the microsomal fraction of liver, testis, ovary, adrenal and placenta catalyse the hydrolysis of sulphated steroids to free steroids. The digestive juice of the snail Helix pomatia contains both sulphatase and glucuronidase activity, and extracts from this source are used to hydrolyse urinary conjugates in vitro for clinical assessment of total and conjugated excretion products.
Metabolism plays many important roles in steroid hormone action. Various biosynthetic pathways occurring in endocrine glands such as the gonads, the adrenals and the fetoplacental unit are required to produce and secrete circulating hormones. These hormones are partly metabolised in the periphery, either before reaching their target tissues (to control plasma levels of active compounds), or after termination of their action (inactivation and elimination). But many of them (" prehormones ") are also metabolised within their target tissues, where a complex interplay between activation and inactivation mechanisms serves to regulate the specificity and the amplitude of the hormonal response.
- Baulieu, E.E. (1991): Biol. Cell. 71:3-10.
- Carlstedt-Duke, J., Eriksson, H., and Gustafsson, J.-A. (1989): The Steroid/Thyroid Hormone Receptor Family and Gene Regulation. Birkhäuser Verlag, Basel.
- Feder, H.H. (1981): In: Neuroendocrinology of Reproduction, edited by N.T. Adler, pp. 19-63. Plenum Press, London.
- Gower, D.B. (1979): Steroid Hormones. Croom Helm, London.
- Griffin, J.E., and Ojeda, S.R. (1988): Textbook of Endocrine Physiology. Oxford University Press, New York.
- Johnson, M., and Everitt, B. (1980): Essential Reproduction. Blackwell Scientific Publications, Oxford.
- Karavolas, H.J., and Hodges, D.R. (1990): In: Steroids and Neuronal Activity, edited by M.A. Simmonds, pp. 22-44. Ciba Foundation Symposium 34, Wiley, Chichester.
- Means, G.D., Kilgore, M.W., Mahendroo, M.S., Mendelson, C.R., and Simpson, E.R. (1991): Mol. Endocrinol. 5:2005-2013.
- Norman, A.W., and Litwack, G., editors (1987): Hormones. Academic Press, Orlando.
- Schulster, D., Burstein, S., and Cooke, B.A. (1976): Molecular Endocrinology of the Steroid Hormones. John Wiley & Sons, London.
- Schumacher, M. (1990): Trends Neurosci. 13:359-362.
- Spelsberg, T.C., Rories, C., Rejman, J.J., Goldberger, A., Fink, K., Lau, C.K., Clvard, D.S., and Wiseman, G. (1989): Biol. Reprod., 40:54-69.