Binding affinity for AR
Binding of T and DHT to the androgen receptor (AR) stabilizes the AR and slows what would otherwise be rapid degradation. At low circulating androgen levels, DHT binding is favored over T but at higher relative T concentrations (e.g., eugonadal state), stabilization of the AR is driven by T more than DHT (20). Nonetheless, DHT is the most potent endogenous androgen based on four critical aspects of its binding to the AR. First, DHT has a relative binding affinity for the AR that is roughly 4 times that of T (21). Second, the rate of dissociation from the AR is about 3 times slower than T (22). Third, binding of DHT to the AR transforms the AR to its DNA-binding state (23). And lastly, DHT upregulates AR synthesis and reduces AR turnover (24). Collectively, these processes amplify the androgenic action of DHT and increase its potency compared with T. However, this may lead to the incorrect conclusion that binding of DHT to the AR always occurs preferentially over T. This is too simplistic a view and ignores the importance of intracellular control of T and DHT concentrations that are mediated by a host of local metabolic pathways. Organ differences in receptor binding of T and DHT result, in part, from relative differences in intracellular concentrations of these androgens rather than from differences in receptor affinities alone (22). Indeed, it has been clearly demonstrated that high concentrations of intracellular T can shift AR binding away from DHT by mass action (25). Moreover, despite there being a single AR, physiological differences in T and DHT action are well known and likely reflect variations in AR receptor distribution, ligand-induced conformational changes to AR that effect stabilization, local hormone synthesis and metabolism, AR-ligand interactions with chromatin, cooperativity of receptors with other transcription factors, and actions of coactivators and corepressors (26, 27). Thus, local tissue control of androgen levels in conjunction with numerous other factors drive AR-induced transcriptional responses. And as elucidated later in this review, tissue concentrations of androgens (particularly in the prostate) are partly distinct from circulating levels.
Protein binding
Like T, circulating DHT is principally bound to sex hormone–binding globulin (SHBG) and, more weakly, to albumin. In general, protein-bound DHT is inactive except in some reproductive tissues in which megalin, an endocytic receptor, acts as a pathway for cellular uptake of DHT when bound to SHBG (28). Studies of interactions between a wide array of natural and synthetic androgens and SHBG indicate that the molecular structure of DHT favors tight linkage to the steroid binding site on SHBG (29). Compared with T, DHT has roughly a fivefold greater binding affinity to SHBG (30). Binding of circulating DHT to SHBG is highest in young males 0.5 to 2 years of age (90%) and thereafter declines to about 70% at age 15 and to 40% in young adult men (age 18) (31). The increase of SHBG that occurs with aging (approximately 1% per year) increases DHT binding in older men (32–35). Dissociation rate constants from SHBG for DHT and T have been measured in human serum and correspond to half times of dissociation of 43 (DHT) and 12 (T) seconds, thus further demonstrating the tenacity to which DHT binds to SHBG (36). Accordingly, concentrations of free circulating DHT in eugonadal men are very low and would be expected to remain so even when total DHT levels increase in response to TRT.
This leads to an important question: Can an increase in circulating levels of SHBG-DHT give rise to DHT-mediated effects? It is well known that SHBG can bind to cell membranes and interact with the SHBG receptor (RSHBG), thus potentially providing a means for its bound ligand to enter the cell. In the case of SHBG-DHT, studies have shown that this complex does not bind to the RSHBG (37). However, once formed, the SHBG-RSHBG can be activated by an agonist steroid to initiate downstream events beginning with the activation of adenylyl cyclase and the generation of cyclic adenosine monophosphate (cAMP) (37). Generation of cAMP in this scenario has been shown to be steroid specific. For example, when DHT or estradiol were exposed to unbound SHBG in a human prostate cancer cell line (namely, LNCaP), rapid increases in intracellular cAMP were observed. However, when this experiment was conducted with human prostatic explants, estradiol caused a rise in cAMP but DHT did not (37).
Metabolism
DHT formed in peripheral tissues is extensively metabolized before its metabolites appear in the circulation (38, 39). Metabolism of DHT to inactive steroids occurs primarily via the initial actions of 3α-17β-hydroxysteroid dehydrogenase (3α-HSD) and 3β-17β-hydroxysteroid dehydrogenase (3β-HSD) in liver, intestine, skin, and androgen-sensitive tissues. Subsequent conjugation by uridine 5′-diphospho (UDP)-glucuronyltransferase (UGT) is the major pathway for urinary and biliary elimination of DHT metabolites and, locally, is the principal irreversible step to protect tissues from high concentrations of this potent androgen (Fig. 1). Of the UGTs, only UGT2 isozymes participate in DHT metabolism. In this regard, UGT2B7, B15, and B17 have remarkable capacities to conjugate androgens and are abundant in androgen-sensitive tissues (6). Differential expression of UGT2 isozymes has been reported and likely plays a role in tissue DHT concentrations independent of circulating androgen levels, particularly in androgen-sensitive tissue. For example, transcripts of UGT2B7, B15, and B17 have been identified in liver, intestine, skin, breast, uterus, and ovary, but adipose tissue expresses only UGT2B15, whereas in prostate, UGT2B15 and B17 are expressed only in luminal and basal cells, respectively. This differential localization combined with other local differences in androgen-metabolizing enzymes provides a finely tuned mechanism for control of intracellular androgen concentrations (7). Polymorphisms of UGT2B15 (that is highly effective in conjugating DHT and its metabolites) have been identified (40) and are postulated to protect prostate tissue from high DHT concentrations and thus lower prostate cancer risk (41, 42). Conversely, increased prostate cancer risk had been observed in white but not African American men with UGT2B17 deletion polymorphism (43). So although it is generally true that DHT concentration in tissue is finely regulated (and, as discussed later, probably not effected to any relevant degree by circulating levels observed in response to androgen therapy), it is equally true that polymorphisms in genes responsible for androgen metabolism may perturb this homeostatic mechanism, thus leading to clinically relevant consequences—both positive and negative.
Finally, the metabolism of DHT must also be considered in light of its metabolic clearance. The overall metabolic clearance of DHT and its metabolism in muscle and adipose tissue of normal men were evaluated in response to intravenously infused DHT (15, 44). The overall mean metabolic clearance of DHT was roughly 70% that of T, thus indicating a modestly longer residence time for DHT. Metabolism of DHT was substantially greater in adipose tissue compared with T, and there was little conversion of T to DHT in muscle. Metabolism of intravenously administered DHT compared with transdermally applied DHT revealed that skin is a major site of peripheral DHT metabolism to 3α-androstanediol, whereas intravenously-administered DHT yielded greater concentrations of 3α-androstanediol-glucuronide (45). Splanchnic tissues have a high capacity to metabolize DHT to DHT-glucuronide, which has importance when oral androgens like T undecanoate (TU) are administered (46). A large fraction of DHT produced in the liver is metabolized to DHT-glucuronide prior to subsequent entry into the circulation (17).
Analytical methods for DHT quantification
In adult eugonadal men, serum DHT concentrations are most accurately measured by liquid chromatography tandem-mass spectrometry (LC-MS/MS), and consistent normal ranges based on this assay platform have been reported across several studies of men spanning a wide age range. A DHT reference range of 14 to 77 ng/dL (0.47 to 2.65 nmol/L) for healthy adult men (18 to 59 years; n = 113) has been reported by Shiraishi et al. (47). Handelsman et al. (48) evaluated age-specific population profiles of circulating DHT in community-dwelling men (<65 years; n = 2606) and observed a serum DHT range of 23 to 102 ng/dL (0.8 to 35 nmol/L). In a cohort of healthy older men (71 to 87 years; n = 394), a DHT reference range of 14 to 92 ng/dL (0.49 to 3.2 nmol/L) has been reported (49). Finally, a normal DHT range of 11 to 95 ng/dL (0.38 to 3.27 nmol/L) has been published by a well-regarded commercial clinical laboratory that utilizes LC-MS/MS for the assay of DHT (Mayo Clinical Medical Laboratory, Rochester, MN). In eugonadal men, DHT concentrations are roughly 7- to 10-fold lower than circulating concentrations of T. Also of note is that plasma T and DHT tend to be highly correlated with a correlation coefficient of 0.7 (49).
Prior to the advent of LC-MS/MS for measurement of DHT, less-precise direct DHT immunoassay methods were used in older studies [e.g., direct radioimmunoassay (RIA), enzyme-linked immunosorbent assay (ELISA)]. We now know that these older assays yielded consistently higher T and DHT values compared with LC-MS/MS by up to 25% (50), particularly at low hormone levels. Others have reported that serum DHT measured by RIA overestimated DHT based on LC-MS/MS by as much as 40% (47). These discrepancies are likely due to lack of specificity of the DHT antibody used in the RIA and failure to remove T from the assay that contributes to cross-reactivity. Because of this, some caution must be exercised in the interpretation of DHT values not measured by LC-MS/MS or by RIA in the absence of Celite column chromatography or other methods to remove T prior to DHT immunoassay. However, when DHT is administered exogenously in pharmacologic amounts, circulating DHT levels increase dramatically, whereas there is a parallel drop in luteinizing hormone and T. Consequently, the use of older RIA methods in situations where DHT levels were high likely yielded reasonably accurate measures of DHT and DHT/T ratios because the mass excess of DHT would have minimized the impact of cross-reactivity with T. In this review, we have noted how T and DHT were measured in each of the studies considered. Findings from studies in which DHT and DHT/T ratios were reported based on LC-MS/MS are more informative and should be afforded more weight.
