5α-Reductase type 1 (5αR1) catalyses A-ring reduction of androgens and glucocorticoids in liver, potentially influencing hepatic manifestations of the metabolic syndrome. Male mice, homozygous for a disrupted 5αR1 allele (5αR1 knockout [KO] mice), were studied after metabolic (high-fat diet) and fibrotic (carbon tetrachloride [CCl4]) challenge. The effect of the 5α-reductase inhibitor finasteride on metabolism was investigated in male obese Zucker rats. While eating a high-fat diet, male 5αR1-KO mice demonstrated greater mean weight gain (21.6 ± 1.4 vs 16.2 ± 2.4 g), hyperinsulinemia (insulin area under the curve during glucose tolerance test 609 ± 103 vs. 313 ± 66 ng ⋅ mL−1 ⋅ min), and hepatic steatosis (liver triglycerides 136.1 ± 17.0 vs. 89.3 ± 12.1 μmol ⋅ g−1). mRNA transcript profiles in liver were consistent with decreased fatty acid β-oxidation and increased triglyceride storage. 5αR1-KO male mice were more susceptible to fibrosis after CCl4 administration (37% increase in collagen staining). The nonselective 5α-reductase inhibitor finasteride induced hyperinsulinemia and hepatic steatosis (10.6 ± 1.2 vs. 7.0 ± 1.0 μmol ⋅ g−1) in obese male Zucker rats, both intact and castrated. 5αR1 deficiency induces insulin resistance and hepatic steatosis, consistent with the intrahepatic accumulation of glucocorticoids, and predisposes to hepatic fibrosis. Hepatic steatosis is independent of androgens in rats. Variations in 5αR1 activity in obesity and with nonselective 5α-reductase inhibition in men with prostate disease may have important consequences for the onset and progression of metabolic liver disease.

BackgroundObesity and metabolic syndrome results from a complex interaction between genetic and environmental factors. In addition to brain-regulated processes, recent genome wide association studies have indicated that genes highly expressed in adipose tissue affect the distribution and function of fat and thus contribute to obesity. Using a stratified transcriptome gene enrichment approach we attempted to identify adipose tissue-specific obesity genes in the unique polygenic Fat (F) mouse strain generated by selective breeding over 60 generations for divergent adiposity from a comparator Lean (L) strain.ResultsTo enrich for adipose tissue obesity genes a ‘snap-shot’ pooled-sample transcriptome comparison of key fat depots and non adipose tissues (muscle, liver, kidney) was performed. Known obesity quantitative trait loci (QTL) information for the model allowed us to further filter genes for increased likelihood of being causal or secondary for obesity. This successfully identified several genes previously linked to obesity (C1qr1, and Np3r) as positional QTL candidate genes elevated specifically in F line adipose tissue. A number of novel obesity candidate genes were also identified (Thbs1, Ppp1r3d, Tmepai, Trp53inp2, Ttc7b, Tuba1a, Fgf13, Fmr) that have inferred roles in fat cell function. Quantitative microarray analysis was then applied to the most phenotypically divergent adipose depot after exaggerating F and L strain differences with chronic high fat feeding which revealed a distinct gene expression profile of line, fat depot and diet-responsive inflammatory, angiogenic and metabolic pathways. Selected candidate genes Npr3 and Thbs1, as well as Gys2, a non-QTL gene that otherwise passed our enrichment criteria were characterised, revealing novel functional effects consistent with a contribution to obesity.ConclusionsA focussed candidate gene enrichment strategy in the unique F and L model has identified novel adipose tissue-enriched genes contributing to obesity.

Patients with critical illness or hepatic failure exhibit impaired cortisol responses to ACTH, a phenomenon known as ‘relative adrenal insufficiency’. A putative mechanism is that elevated bile acids inhibit inactivation of cortisol in liver by 5α-reductases type 1 and type 2 and 5β-reductase, resulting in compensatory downregulation of the hypothalamic–pituitary–adrenal axis and adrenocortical atrophy. To test the hypothesis that impaired glucocorticoid clearance can cause relative adrenal insufficiency, we investigated the consequences of 5α-reductase type 1 deficiency in mice. In adrenalectomised male mice with targeted disruption of 5α-reductase type 1, clearance of corticosterone was lower after acute or chronic (eightfold, P<0.05) administration, compared with WT control mice. In intact 5α-reductase-deficient male mice, although resting plasma corticosterone levels were maintained, corticosterone responses were impaired after ACTH administration (26% lower, P<0.05), handling stress (2.5-fold lower, P<0.05) and restraint stress (43% lower, P<0.05) compared with WT mice. mRNA levels of Nr3c1 (glucocorticoid receptor), Crh and Avp in pituitary or hypothalamus were altered, consistent with enhanced negative feedback. These findings confirm that impaired peripheral clearance of glucocorticoids can cause ‘relative adrenal insufficiency’ in mice, an observation with important implications for patients with critical illness or hepatic failure, and for patients receiving 5α-reductase inhibitors for prostatic disease.

5α-Reductases irreversibly catalyse A-ring reduction of pregnene steroids, including glucocorticoids and androgens. Genetic disruption of 5α-reductase 1 in male mice impairs glucocorticoid clearance and predisposes to glucose intolerance and hepatic steatosis upon metabolic challenge. However, it is unclear whether this is driven by changes in androgen and/or glucocorticoid action. Female mice with transgenic disruption of 5α-reductase 1 (5αR1-KO) were studied, representing a ‘low androgen’ state. Glucocorticoid clearance and stress responses were studied in mice aged 6 months. Metabolism was assessed in mice on normal chow (aged 6 and 12 m) and also in a separate cohort following 1-month high-fat diet (aged 3 m). Female 5αR1-KO mice had adrenal suppression (44% lower AUC corticosterone after stress), and upon corticosterone infusion, accumulated hepatic glucocorticoids (~27% increased corticosterone). Female 5αR1-KO mice aged 6 m fed normal chow demonstrated insulin resistance (~35% increased area under curve (AUC) for insulin upon glucose tolerance testing) and hepatic steatosis (~33% increased hepatic triglycerides) compared with controls. This progressed to obesity (~12% increased body weight) and sustained insulin resistance (~38% increased AUC insulin) by age 12 m. Hepatic transcript profiles supported impaired lipid β-oxidation and increased triglyceride storage. Female 5αR1-KO mice were also predisposed to develop high-fat diet-induced insulin resistance. Exaggerated predisposition to metabolic disorders in female mice, compared with that seen in male mice, after disruption of 5αR1 suggests phenotypic changes may be underpinned by altered metabolism of glucocorticoids rather than androgens.