Modelling altered glucocorticoid sensitivity : from HPA axis to metabolic abnormalities in mice and humans

• Main Author: Michailidou, Zoi
• Other Authors: Chapman, Karen
• Published: University of Edinburgh 2008
• Subjects: 612.6; Endocrinology
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.562206

The primary determinants of tissue glucocorticoid action are glucocorticoid receptor (GR) density and intracellular levels of ligand, the latter determined both by activity of the hypothalamic-pituitary-adrenal (HPA) axis and cellular activity of 11beta- hydroxysteroid dehydrogenase (11beta-HSD) enzymes that interconvert active 11- hydroxy (corticosterone, cortisol) and inactive 11-keto (11-dehydrocorticosterone, cortisone) glucocorticoids. Here, the contribution of GR density and ligand levels in determining body composition and metabolic phenotype have been investigated in mice and in humans. Genetic evidence in humans implicates variations in the GR gene in the regulation of the HPA axis as well as the control of body fat distribution, metabolic parameters and blood pressure. Although GR deficient mouse models have been previously generated (with homozygous nulls dying at birth), the effects of altered GR density upon fat distribution and blood pressure have not been described. This study addresses the relationship between GR density and metabolic parameters, including body fat distribution, insulin resistance and hypertension. A novel line of mice harbouring a null mutation in the GR gene (GR+/-) was generated from an ES cell line in which a beta-galactosidase-neomycin phosphotransferase (beta geo) reporter cassette was fused with GR. The resulting fusion protein lacks part of the DNA binding domain and the entire ligand binding domain and is transcriptionally inactive. In addition, the beta-galactosidase enzyme activity “reports” activity of the GR gene promoter. GR-/- mice are present in a normal Mendelian ratio before birth. Intriguingly, 1 (of 36/146 expected if null allele not lethal) survived to adulthood suggesting this might be a hypomorphic rather than a null allele. Heterozygous 15 (GR+/-) mice showed 40-45% reductions in GR mRNA levels in the hippocampus, paraventricular nucleus of the hypothalamus, pituitary gland and adipose tissue, 30% in liver, 56% in muscle and 67% in adrenals. X-gal staining of GR+/- brain sections showed that GR-beta gal is present throughout, mirroring GR mRNA expression. Adult GR+/- mice had larger adrenals, higher evening plasma corticosterone levels and greater corticosterone responses following 10 minute restraint suggesting a hyperactive HPA axis. Compared to GR+/+ littermates, GR+/- mice had similar body weight gain on normal chow or high fat diet, with unaltered fat depot (inguinal, epididymal, mesenteric) weights and similar glucose and insulin tolerance. However, GR+/ - mice had higher (10%) systolic blood pressure, associated with activation of the renin-angiotensin system. Thus GR haploinsufficiency in mice causes increased blood pressure and accords with data associating GR polymorphisms with hypertension in humans. The role of altered GC sensitivity was also investigated in a mouse model of HPA axis hypoactivity pro-opiomelanocortin null (POMC) mice. POMC-null mice are obese due to central melanocortin deficiency. In contrast to most rodent models of obesity, POMC-null mice are also glucocorticoid deficient due to ACTH deficiency. Previous data have shown that glucocorticoid replacement in POMC-null mice exaggerated hyperphagia, obesity and insulin resistance and caused hypertension. Here, the contribution of peripheral glucocorticoid sensitivity was investigated. POMC-null mice have increased liver and retroperitoneal fat GR mRNA levels but, specifically in adipose tissue, decreased levels of mRNA encoding 11beta-HSD1, a reductase which regenerates active glucocorticoids, thus amplifying their action.