Steroid secretion is an essential part of human physiology. Steroids are involved in a number of different important processes, such as inflammation (Critchley et al., 1999; Maia and Casoy, 2008), control of circulating volume (Stewart, 2008) and psychological well-being (Weber, 1998), as well as having critical significance in aspects of human sexual development (Geissler et al., 1994; Can et al., 1998). There are a number of steroid hormones that play a major role in reproductive functioning in the human body. Particularly important are the sex hormones, including the androgens testosterone and its 5a-reduced form, dihydrotestosterone (DHT), which are essential for virilisation in males. All the major steroid sex hormones are made through a common set of interconnected pathways referred to collectively as steroidogenesis (Kronenberg et al., 2008).
Steroidogenesis is a complex cascade of enzymatic reactions. All of the steroid hormones of the adrenal cortex, placenta, ovaries and testes originate from steroidogenesis, including mineralocorticoids, glucocorticoids and gonadal steroids (Hanley and Arlt, 2006). The products of steroidogenesis all originate from the common steroid precursor cholesterol, which is mainly obtained from circulating low-density lipoprotein and from local conversion of acetate. The final steroid product generated is determined by the precise complement of enzymes involved (Christenson and Devoto, 2003). The coordinated expression of steroidogenic enzymes delineates the steroidogenic phenotype of a particular organ or tissue within the endocrine system, with each steroidogenic organ expressing its own characteristic complement of enzymes.
In steroidogenic cells, cholesterol is transported from the outer mitochondrial membrane to the inner mitochondrial membrane, where steroidogenesis begins. This transportation event is mediated by the steroidogenic acute regulatory protein (StAR). StAR plays a key role in the acute regulation of steroid hormone synthesis. Within the inner membrane of mitochondria, the initial conversion step of steroidogenesis is the side-chain cleavage of cholesterol, which occurs universally in all steroidogenic cells, performed by CYP11A1. Following this, there is an intrinsic tissue-specific and/or cell-specific restriction on what steroids can be synthesised, which is predetermined by the expression of the appropriate sequence of enzymes in the tissue or cells in question (Miller, 2005). For example, in the adult adrenal cortex, following conversion of cholesterol to pregnenolone, expression of CYP21A2 in both the zona glomerulosa and zona fasciculata directs steroidogenesis towards mineralocorticoid and glucocorticoid synthesis, respectively. However, it is the expression of CYP11B2 exclusively in the zona glomerulosa that leads to the production of aldosterone. In contrast, the lack of CYP11B2 in the zona fasciculata prevents aldosterone production, whereas the expression of CYP11B1 in this location permits cortisol instead (Stewart, 2002). Likewise, the presence of HSD17B3 activity in the gonads directs steroid biosynthesis towards the production of testosterone in the testes, which is further converted to estradiol in the ovary by CYP19A1 (aromatase) expression (Bashin et al., 2002). All conversion steps following pregnenolone synthesis are performed either on the inner surface of the mitochondria (e.g. CYP11B1 and CYP11B2) or in the endoplasmic reticulum (e.g. CYP21A2, CYP17A1, CYP19A1 and HSD3B2) of steroidogenic cells.
Androgen biosynthesis in the human testis
One of the key factors produced by the developing testis that promotes male differentiation and virilisation of tissues is testosterone. Following testes determination and the initial establishment of the male morphological sex, the biosynthesis of testosterone by foetal testicular Leydig cells at approximately eight weeks gestation, is crucial to continuing the cascade of events that lead to the male phenotype (Siiteri and Wilson, 1974). Like all steroid hormones, testosterone is produced from cholesterol. In the so-called ‘classical pathway’, the sequence of steroidogenic reactions that leads to testosterone biosynthesis util ises the steroidogenic enzymes CYP11A1, HSD3B2, CYP17A1 and HSD17B3, to ultimately produce testosterone from cholesterol. Testosterone has a major role in initiating male differentiation of the reproductive system in the foetus. As such, it is termed an androgen.
Testosterone can be converted into the more potent androgen, DHT (Kim et al., 2002). Following secretion of testosterone from the testis, DHT is produced locally in tissues by the action of the enzyme, type 2 5α-reductase, which is a membrane-bound enzyme encoded by the gene SRD5A2 (Ntais et al., 2003). Tissues where DHT is produced typically express SRD5A2, allowing for conversion of testosterone in target cells (Berman et al., 1995). Interestingly however, it has been demonstrated in humans that SRD5A2 is active from a very early stage in development and is able to convert testosterone to DHT even before the foetal testis acquires the ability to produce testosterone (Fisher, 2008). DHT plays a vital role in normal male sexual differentiation during embryogenesis, and is also important for many androgen-mediated events during male puberty. Studies in mice indicate that DHT acts during embryonic life to amplify hormonal signals that can only be mediated by testosterone at higher concentrations (Wilson et al., 2002); indicating that DHT is a more potent androgen than testosterone. This is consistent with reports of inactivating mutations in SRD5A2 that have caused micropenis, hypospadias and partial sex reversal in males (Nordenskjold and Ivarsson, 1998; Chavez et al., 2000; Sasaki et al., 2003).
The alternative pathway to dihydrotestosterone biosynthesis
Alongside the classical pathway is the more recently discovered alternative pathway to potent androgen (Auchus, 2004; Arlt et al., 2004; Shaw et al., 2006). In the classical pathway of androgen production, the end product is testosterone, which can serve as a precursor for 5a-reduction to the more potent DHT at the sites of androgen action. However, it has been demonstrated in animal models that DHT can be formed by another pathway. In the testis of the tammar wallaby (Macropus eugenii) pouch young and in immature mouse testis, it is thought that the accumulation of 17-hydroxyprogesterone (17-OHP) prompts the conversion of 17-OHP into DHT via a testosterone-independent route of steroidogenesis (Wilson et al., 2003; Mahendroo et al., 2004). This is thought to occur through the conversion of 17-OHP first to 5a-pregnane-17a-ol-3,20-dione by 5a reductase activity (presumed to be the type 1 enzyme, SRD5A1), which is followed by a sequential series of enzyme reactions converting it to 5a-pregnane-3a,17a-diol-20-one (5-pdiol) by aldo-keto reductase (HSD3A) activity, then androsterone by CYP17A1 action; androstanediol (5-adiol) by 17 b-hydroxysteroid dehydrogenase (HSD17B) activity; and finally to DHT by HSD3A activity.
In the tammar wallaby, 5-adiol plays a key role in the formation of the male urogenital system, and copious amounts have been detected in the testes of pouch young (Shaw et al., 2000). If female pouch young are treated with 5-adiol, the urogenital tract and the external genitalia become virilised (Leihy et al., 2001). Furthermore, in several rodent species, 5-adiol is the major androgen formed in immature male testes (Ge et al., 1999; Frungieri et al., 1999) and is readily secreted in regenerating Leydig cells of adult rats following treatment with cytotoxic drugs (Risbridger and Davies, 1994). In rodents it has also been demonstrated that the capacity to form 5-adiol declines during late foetal development, when there is a shift towards testosterone production (Chase and Payne, 1983; Ge and Hardy, 1998), which can be accelerated by chorionic gonadotrophin (Moger, 1977).
The role of the alternative pathway in male sexual differentiation in humans remains unclear and it is unknown to what extent the alternative pathway is in operation within the human foetal testis (Auchus, 2004). However, there is a possibility that it could represent an important pathway to DHT during early development, especially when we consider that the steps of the alternative pathway leading from 5-pdiol to 5-adiol avoid two energy inefficient steps used to synthesise testosterone. Since 17-OHP is superior to testosterone as a substrate for 5-reductase (Pratis et al., 2000), and 5-pdiol appears to be the best substrate for the 17, 20-lyase activity of CYP17A1 (Wilson et al., 2003), the use of 5-pdiol in the alternative pathway suggests a more efficient route to production of 19-carbon steroids. Hence the formation of 5-adiol in foetal testes could represent the most efficient route to DHT due to the inherent properties of the steroidogenic enzymes involved. However, the fact that the foetal testis secretes high amounts of testosterone (Siiteri and Wilson, 1974) and that mutations in HSD17B3 and SRD5A2 (both clear parts of the classical pathway) cause profound under-virilisation (Imperato-McGinley and Zhu, 2002), suggests the classical pathway predominates in humans during normal male differentiation, although a role for the alternative pathway cannot be fully excluded.
Regulation of androgen steroidogenesis in adults
The pathways and associated enzymes involved in androgen biosynthesis are typically subject to regulation consisting of trophic hormone stimuli and negative feedback mechanisms. In the normal human adult testis, regulation of testosterone biosynthesis is primarily controlled by the hypothalamic-pituitary-gonadal (HPG) axis, via secretion of lutein ising hormone (LH). The HPG axis comprises of the hypothalamus, anterior pituitary gland and testes/ovaries, and represents a ‘closed-loop’ endocrine system subject to negative feedback control (Low, 2008). The hypothalamus is located at the base of the forebrain and is directly connected to the pituitary stalk, which is in turn continuous with the posterior lobe of the pituitary gland. The hypothalamus is also very closely associated with the anterior pituitary through a dense capillary network, the median eminence capillary plexus, which lies at the interface between the hypothalamus and anterior pituitary. The central function of the hypothalamus is to organ ise the secretion of different hormones in response to multiple inputs from the circulatory and nervous systems regarding the homeostatic state of the body (Mastorakos et al., 2006). In terms of the HPG axis, hypothalamic- releasing hormones are secreted in response to relevant inputs, and carried locally in the bloodstream where they are detected by specialised secretory cells (gonadotrophs) in the anterior pituitary gland (Melmed and Kleinberg, 2008). In response to the releasing hormones, the anterior pituitary is prompted to release its own hormones, but in much larger quantities, which go on to regulate hormone biosynthesis and secretion from the gonads.
The HPG axis is central to both the development and maintenance of the reproductive system in males and females. In humans the hypothalamic secretion of gonadotrophin-releasing hormone (GnRH; also known as Lutein ising hormone-releasing hormone, LHRH) stimulates gonadotroph cells in the anterior pituitary to release large quantities of glycoprotein gonadotrophins, specifically lutein ising hormone (LH) and follicle- stimulating hormone (FSH; Huirne and Lambalk, 2001). These are secreted into the body’s general circulation and soon reach the target gonad tissues.
Regulation of androgen biosynthesis during human foetal development
The regulatory mechanisms underlying androgen steroidogenesis in adult males are well documented. However, despite its critical importance during early development of the reproductive tract, the regulation of testicular androgen production during the first trimester has remained unclear (Sobel et al., 2004). Generally, three candidates for regulation have been considered: (i) LH from the HPG axis (i.e. similar control mechanism to adult regulation); (ii) hCG from the placenta; or (iii) constitutive secretion of androgens from the foetal testis.
Current understanding of androgen biosynthesis during early development has been achieved primarily through biochemical studies where steroid intermediaries are traced from initial substrate to final products, and by the phenotypes observed in individuals with mutations in key biosynthetic enzymes. The importance of LHR, inferred from similar mutational analysis, suggests its ligands LH and/or hCG could be important for regulation of foetal testicular androgen biosynthesis (Ascoli et al., 2002). Severe inactivating LHR mutations causes a disorder of sex development (DSD) in 46,XY males, which can include complete failure to virilise and the appearance of female-type external genitalia at birth (Themmen and Huhtaniemi, 2000). LHR belongs to a family of G protein- coupled receptors with seven transmembrane helices and structurally resembles the thyrotrophin receptor (Ji and Ji, 1991). It is encoded by a single gene on chromosome two, which has 11 exons and 10 introns. A number of tissue- specific splice variants of LHR have been identified, although only two well- characterised splice products exist: the full- length protein composed of 699 amino acid residues, and the smaller splice variant composed of 685 amino acid residues (Ryu et al., 1998; Minegishi et al., 2007). Despite the apparent importance of LHR, it appears that the gonadotrophs that produce LH are functionally immature during the first trimester, making LH production unlikely at this stage of sexual differentiation.
An obvious alternative candidate to LH regulation is hCG. Although hCG is synthes ised abundantly during the first trimester, it would seem unlikely that it is responsible for androgen regulation, as it would be required to cross the foetal-placental barrier. Furthermore, testosterone secretion has been illustrated prior to LHR expression in the developing foetus, which could suggest constitutive secretion of androgens and the potential importance of Leydig cell nuclear transcription factors in the regulation of testosterone biosynthesis (e.g., SF1). It has been documented that inactivating mutations in SF1 result in a lack of gonadogenesis, and sex reversal in 46,XY males, including ambiguous internal genitalia (Achermann et al., 1999; Lin et al., 2007). It has also been shown that mutations causing constitutive activation of LHR lead to a normal male phenotype, although precocious puberty occurs due to androgen excess (Min et al., 1998a).
The lack of strong evidence to establish LH and/or hCG as key regulators of early testosterone biosynthesis during foetal development has led to the assumption that the testes themselves are the main source hormonal influence on testicular testosterone biosynthesis at this time. This could take the form of paracrine/autocrine regulation, or alternatively, constitutive testosterone secretion from the foetal testes. Putative paracrine/autocrine regulators include adenylate cyclase-activating polypeptide 1 (Matsumoto et al., 2008) and the natriuretic peptide hormones (Khurana and Pandey, 1993; El-Gehani et al., 2001). However, little research has been performed to measure the influence of these on early foetal testosterone and thus the significance of their influence is unknown. The final possibility is that testosterone is simply produced constitutively during early foetal development. Constitutive expression is exhibited by many different cell types (e.g., chondrocytes (Recklies et al., 2005) and osteoblasts (Rifas et al., 1989)) and can be driven by constitutively active G protein-coupled receptors, similar to LHR (Olesnicky et al., 1999). Given that regulation of early testosterone production can still not yet be attributed to a major hormone, despite much research in this field, it is possible that this unknown factor does not exist and that testosterone production is an intrinsic property of early foetal Leydig cells in humans, regulated by transcription factors such as SF-1 or DAX1.