Modulation of 11beta-hydroxysteroid dehydrogenase isozymes by growth hormone and insulin-like growth factor: in vivo...

The interconversion of hormonally active cortisol (F) and inactive cortisone (E) is catalyzed by two isozymes of 11beta-hydroxysteroid dehydrogenase (11betaHSD), an oxo-reductase converting E to F (11betaHSD1) and a dehydrogenase (11betaHSD2) converting F to E. 11betaHSD1 is important in mediating glucocorticoid-regulated glucose homeostasis and regional adipocyte differentiation. Earlier studies conducted with GH-deficient subjects treated with replacement GH suggested that GH may modulate 11betaHSD1 activity. In 7 acromegalic subjects withdrawing from medical therapy (Sandostatin-LAR; 20-40 mg/month for at least 12 months), GH rose from 7.1 +/- 1.5 to 17.5 +/- 4.3 mU/L (mean +/- SE), and insulin-like growth factor I (IGF-I) rose from 43.0 +/- 8.8 to 82.1 +/- 13.7 nmol/L (both P < 0.05) 4 months after treatment. There was a significant alteration in the normal set-point of F to E interconversion toward E. The fall in the urinary tetrahydrocortisols/tetrahydocortisone ratio (THF+allo-THF/THE; 0.82 +/- 0.06 to 0.60 +/- 0.06; P < 0.02) but unaltered urinary free F/urinary free E ratio (a marker for 11betaHSD2 activity) suggested that this was due to inhibition of 11betaHSD1 activity. An inverse correlation between GH and the THF+allo-THF/THE ratio was observed (r = -0.422; P < 0.05). Conversely, in 12 acromegalic patients treated by transsphenoidal surgery (GH falling from 124 +/- 49.2 to 29.3 +/- 15.4 mU/L; P < 0.01), the THF+allo-THF/THE ratio rose from 0.53 +/- 0.06 to 0.63 +/- 0.07 (P < 0.05). Patients from either group who failed to demonstrate a change in GH levels showed no change in the THF+allo-THF/THE ratio. In vitro studies conducted on cells stably transfected with either the human 11betaHSD1 or 11betaHSD2 complementary DNA and primary cultures of human omental adipose stromal cells expressing only the 11betaHSD1 isozyme indicated a dose-dependent inhibition of 11betaHSD1 oxo-reductase activity with IGF-I, but not GH. Neither IGF-I nor GH had any effect on 11betaHSD2 activity. GH, through an IGF-I-mediated effect, inhibits 11betaHSD1 activity. This reduction in E to F conversion will increase the MCR of F, and care should be taken to monitor the adequacy of function of the hypothalamo-pituitary-adrenal axis in acromegalic subjects and in GH-deficient, hypopituitary patients commencing replacement GH therapy. Conversely, enhanced E to F conversion occurs with a reduction in GH levels; in liver and adipose tissue this would result in increased hepatic glucose output and visceral adiposity, suggesting that part of the phenotype currently attributable to adult GH deficiency may be an indirect consequence of its effect on tissue F metabolism via 11betaHSD1 expression.