Metabolic syndrome is marked by perturbed glucocorticoid (GC) signaling, systemic inflammation, and altered immune status. Dehydroepiandrosterone (DHEA), a major circulating adrenal steroid and dietary supplement, demonstrates antiobesity, anti-inflammatory, GC-opposing and immune-modulating activity when administered to rodents. However, plasma DHEA levels failed to correlate with metabolic syndrome and oral replacement therapy provided only mild benefits to patients. Androstene-3β,7β,17β-triol (β-AET) an anti-inflammatory metabolite of DHEA, also exhibits GC-opposing and immune-modulating activity when administered to rodents. We hypothesized a role for β-AET in obesity. We now report that plasma levels of β-AET positively correlate with BMI in healthy men and women. Together with previous studies, the observations reported here may suggest a compensatory role for β-AET in preventing the development of metabolic syndrome. The β-AET structural core may provide the basis for novel pharmaceuticals to treat this disease.
Metabolic syndrome, characterized by abdominal obesity as well as atherogenic dyslipidemia, insulin resistance, elevated blood pressure and a prothrombotic, proinflammatory propensity, has been steadily increasing in the United States and Europe and is now common in developing nations as well (1,2,3). It is highly correlated with age, BMI, and adrenocortical dysregulation (4,5). The adrenals, endocrine glands that lie at the superior poles of the two kidneys, produce steroid hormones that regulate mammalian metabolism. These include glucocorticoids (GC) such as cortisol and copious amounts of dehydroepiandrosterone (DHEA) (6). Interestingly, increased GC signaling and aberrant steroidogenesis have both been described in patients with metabolic syndrome (7,8).
DHEA has purported antiobesity effects (9,10,11) and opposes certain activities of endogenous cortisol (12,13,14). It reduces the cortisol/cortisone ratio in adipose tissue by effects on GC receptor transcriptional activity (15) and by decreasing the expression of hydroxysteroid dehydrogenase-1 (11β-HSD1) (7,8). 11β-HSD1 is the enzyme responsible (in humans) for the tissue-specific interconversion of cortisol to it's inactive metabolite 11-cortisone (16). Elevated 11β-HSD1 has emerged as a key pharmaceutical target for therapeutic agents designed to treat metabolic syndrome (17). Such reports help to explain why DHEA is widely used as a dietary supplement (18). However, many of these observations were made either in vitro, or in rodents. In humans, plasma DHEA levels failed to correlate with metabolic syndrome (19,20) and in well-controlled clinical trials DHEA replacement therapy provided little benefit to patients (21,22,23). Profound differences in DHEA metabolism between rodents and humans suggest that the desirable activities of DHEA may reside in one or more of it's highly oxidized metabolites. These are readily formed in rodents, but to a lesser extent in humans (24,25,26,27,28,29).
Androstene-3β, 7β, 17β-triol (β-AET) appears to be a biologically relevant anti-inflammatory (30) DHEA metabolite (31) with multifaceted properties that include GC-opposing functions (32,33,34) and immune modulation (35,36). Oxidation of DHEA occurs at C-7 in certain tissues, such as brain, liver, gut, lymphoid organs, joints and, importantly, adipose via the CYP7 enzymes (37,38,39). Although the human metabolic pathways do not favor formation of highly oxidized metabolites from exogenous DHEA (40), these 7-hydroxy metabolites are collectively present in the low nmol/l concentrations in human circulation. Interestingly, the products, 3β, 7α-dihydroxy-androst-5-en-17-one and 3β, 7α, 17β-trihydroxy-androst-5-ene are converted to 7-oxo and 7-β-hydroxy forms by 11β-HSD1 (41).
Taken together, these observations led us to hypothesize a potential role for β-AET in limiting inflammation and GC-induced features of metabolic syndrome. In support of this role, we here report correlation of elevated plasma levels of β-AET with BMI in healthy men and women and suggest a role for β-AET in limiting metabolic syndrome (42). Our results imply that in healthy individuals, β-AET may act in a compensatory fashion, to oppose GC action and that perturbations to this balance may be associated with the onset of metabolic syndrome.
Methods and Procedures
Human plasma analysis
Subjects. Plasma samples were obtained from 252 volunteers (102 males and 150 females; ages 20–80). BMI was calculated as weight (kg)/height (m2). Informed consent was obtained from all subjects and approved by the institutional review boards that approved sample acquisitions. Samples were residual to phase I studies and were obtained before any drug or placebo exposure. All volunteers were healthy individuals as defined by their lipid profiles and their vital signs were within normal ranges.
Plasma analysis. The concentrations of β-AET and DHEA were measured in plasma by liquid chromatography-mass spectrometry. For β-AET, plasma samples (0.2 ml) were spiked with β-AET-d3 as an internal standard (1 ng/ml), steroids were extracted with 10 ml ethyl acetate and concentrated by evaporation under nitrogen. The dried residue was derivatized with 0.5 ml of 50 mg/ml nicotinyl chloride anhydrous pyridine solution for 1 h at 80 °C, cooled to room temperature, and quenched with 1 ml 5% sodium bicarbonate. The derivatized steroids were extracted with 10 ml methyl-tert-butyl ether, dried under nitrogen, dissolved in high-performance liquid chromatography mobile phase, and analyzed by reverse phase liquid chromatography-mass spectrometry (ESI+ mode). Tri-nicotinyl-β-AET was identified by its liquid chromatography retention time and a 622 > 376 amu transition and quantified using a standard curve from 10 to 1,000 pg/ml in water (10 pg/ml was the lower limit of detection for this assay). For DHEA, plasma samples (100 µl) were spiked with DHEA-d2 and extracted with 4 ml methyl t-butyl ether. The organic phase was evaporated to dryness, and the residue was dissolved in 1 mol/l aqueous hydroxylamine hydrochloride, and incubated for 1 h at 60 °C. The resulting steroid-oxime derivatives were extracted with 2.5 ml methyl t-butyl ether, dried, and reconstituted in 200 µl of water/acetonitrile (75:25, vol:vol), and analyzed by reverse phase liquid chromatography-mass spectrometry (ESI + mode). DHEA was identified by its retention time and 304 >253 amu transition, and quantified with a 50–125,000 pg/ml standard curve (50 pg/ml was the lower limit of detection for this assay).
Statistical analysis. Data were analyzed using Graphpad Prism 4 software (Prism, San Diego, CA).
Human plasma correlates
To ascertain the relationship between endogenous plasma levels of β-AET and obesity, BMI was calculated and plasma β-AET measured in serum samples taken from healthy subjects, 102 males (ages 20–80) and 150 females (ages 20–73). The concentration of β˜-AET in plasma ranged from 2 to 162 pg/ml in males and from 6 to 249 pg/ml in females (Figure 1, top). The BMI ranged from 18 to 49 in males and from 21 to 56 in females. Females had significantly (P < 0.05) higher serum levels of β-AET than males. Females also had significantly (P = 0.0002) higher BMIs than males (32.0 ± 6.4 vs. 28.7 ± 5.6; Figure 1, middle). There was no significant difference in age between males (46.5 ± 13.6) and females (45.6 ± 12.5) (Figure 1, bottom).
Linear regression analysis of β-AET vs. BMI revealed significant nonzero slope for males (P = 0.005; r2 = 0.076) and females (P < 0.0001; r2 = 0.172) (Figure 2). Serum levels of DHEA were available for 75 males and 68 females. Serum levels of β-AET were highly correlated in both males (P = 0.0008; r2 = 0.15) and females (P < 0.0001; r2 = 0.51) (Figure 3). In contrast to β˜-AET, there was no correlation with serum levels of DHEA and BMI in either male (P = 0.16; r2 = 0.027) or female (P = 0.84; r2 = 0.0006) subjects (data not shown). In these healthy volunteers, we detected no trends or significant correlations between serum levels of β-AET and other metabolic syndrome manifestations including serum triglycerides, hemoglobin A1c, C-reactive protein, or blood pressure (data not shown).
This is the first report that serum levels of β-AET are positively correlated with increasing BMI in healthy men and women. We hypothesized a role for β-AET in obesity as compensatory to increasing GC signaling based on several reports in the literature indicating it's GC-opposing functions and our own observations of it's anti-inflammatory activity. A population of healthy individuals (male and female) with BMI ranging from 18 to 56, were assayed for endogenous levels of β-AET and a significant positive correlation was found for both males and females, which is consistent with (but does not prove) this hypothesis. No such correlation was found with DHEA, which is in agreement with previous studies (19,20). Our observation that β-AET levels are significantly lower in males (among whom metabolic syndrome is more prevalent (43)) than in age-matched females, is also consistent with this hypothesis. Since 11 β-HSD1 is involved in the formation of β-AET (41), the endocrinology behind increased β-AET levels in obese subjects may relate to the increased levels of this enzyme in obesity (44). In the context of high GC levels β-AET may act to counter regulate aspects of GC signaling and restore homeostasis. Accordingly, perturbations in adrenocortical signaling relevant to the development of metabolic syndrome (5) may extend to steroid metabolism and effect signaling pathways within peripheral tissues (4). However, our observations do not necessarily imply an etiologic role of β-AET in metabolic syndrome.
The organs or tissues potentially subject to the action of β-AET on 11β-HSD1-mediated GC signaling include liver, muscle, brain, and adipose tissue (44,45,46) and may also involve other steroidogenic enzymes such as CYP7A, 11β-HSD2, and tissue-specific transcription factors (39,47). For example, DHEA-mediated inhibition of 11β-HSD1 activity has been reported in adipose tissue and involves a switch in CCAAT/enhancer-binding protein (C/EBP) expression. C/EBPα, a potent activator of 11β-HSD1 gene transcription, was downregulated in 3T3-L1 adipocytes and in liver and adipose tissue of DHEA-treated mice, whereas C/EBPβ and C/EBPΔ were unchanged or elevated (48). In contrast, DHEA induces 11β-HSD2 gene expression via action on this same transcription factor (49). Those authors observed a steroid-mediated toggle between 11β-HSD1 and 11β-HSD2 gene expression and suggested a PI3k/Akt-dependent mechanism for modulation of GC signaling.
The relevant β-AET-protein interactions potentially involved in regulating GC signaling also remain unknown. Preliminary studies in our laboratory support the hypothesis that like DHEA, β-AET modulates 11β-HSD1 gene expression (Harbor Biosciences, San Diego, CA, unpublished data). And while elevated 11β-HSD1 has emerged as a key pharmaceutical target for therapeutic agents designed to treat metabolic syndrome (17), it is possible that β-AET may also modulate other genes. β-AET, or a metabolite, may interact with proteins integral to tissue-specific transcription complexes such as C/EBPα, and/or related cofactors. It is also possible that β-AET may affect enzymatic activity. While our own studies using human microsomes could not demonstrate an effect of β-AET on the enzymatic activity of either 11β-HSD1 or CYP7B (Harbor Biosciences, unpublished data), certain 7-oxygenated steroids are known to be involved with 11β-HSD1 enzymatic activity in cellular assays (50). Three dimensional-modeling analyses indicated that the 7-keto and 7-hydroxy metabolites of DHEA occupy the same binding site in 11β-HSD1 as cortisone and cortisol, respectively (51). Thus, we cannot strictly rule out a potential effect of β-AET or a metabolite on enzymatic activity. Although no dedicated nuclear hormone receptor has yet been identified for 7-hydroxylated steroids (52), activity may still reasonably be expected to involve signaling via an orphan or other as-yet-uncharacterized nuclear receptor. An interaction with surface receptors that exerts nongenomic activities is also a possibility as has been suggested for DHEA (53), progesterone, estrogen (54), and testosterone (55). Kinases, phosphatases (56), as well as other enzymes (10), may also be involved. These possibilities are not mutually exclusive, and multiple interactions may be involved, as has been demonstrated for DHEA (57).
We have speculated that β-AET may be responsible for the GC-opposing activity initially ascribed to DHEA in rodents (30). And while endogenous levels of β-AET in patients with metabolic syndrome is the subject of forthcoming studies from our group, simply restoring homeostatic β-AET concentrations in patients may not correct the underlying cause of disease, as was repeatedly found to be true with DHEA. Altered steroid metabolism and/or signal resistance in specific tissues may result from reregulation of steroidogenesis associated with disease (5). This was confirmed by contrasting findings in our own clinical trials. In phase I studies, treatment with β-AET decreased serum cholesterol and triglyceride levels in healthy subjects, but not in phase II studies in dysregulated hyperlipidemic patients, regardless of the route of administration (manuscript in preparation). This outcome prompted the search for synthetic β-AET analogues with improved pharmaceutical properties that might deliver the compensatory signal even in dysregulated hormonal environments.
HE3286 (17α-ethynyl-5-androstene-3β,7β,17β-triol) was identified as a synthetic analogue of β-AET that demonstrated similar biological activities of the parent compound in rodents and improved pharmaceutical properties in humans, including increased metabolic stability and oral bioavailability (58). In preclinical studies, HE3286 ameliorated glucose intolerance, improved insulin sensitivity, and delayed disease in rodent models of type II diabetes (59). In those studies, the observed therapeutic effects appeared to result from attenuation of proinflammatory pathways. Orally administered HE3286 also demonstrated activity similar to or greater than β-AET in murine models of colitis, multiple sclerosis (36), and rheumatoid arthritis (60). In those studies, HE3286 attenuated nuclear factor-κB activation, increased regulatory T cells, decreased interleukin-17 signaling and limited other key Th1-associated proinflammatory cytokines including interferon-γ and tumor necrosis factor-α. These features support the potential for this compound to treat metabolic syndrome. Because low circulating levels of DHEA and increased Th1-associated inflammatory cytokines have been reported in patients with nonalcoholic fatty liver disease (61,62), the hepatic manifestation of metabolic syndrome, HE3286 may provide particular benefit to these patients.
Our observations support and expand the potential for roles of 7-hydroxy C-19 steroids in the evolution of metabolic syndrome and suggest that the β-AET structural core may be fundamental to the development of a new series of synthetic therapeutic agents based on findings in the DHEA metabolome.
The authors wish to acknowledge Mr Kevin Liu for help in creating tables and figures and formatting.
Authors are employed by and hold equity positions in Harbor Biosciences.