Ruolo dell'11β-idrossisteroidodeidrogenasi tipo 1 nella sindrome dell'ovaio policistico

HSD11B1 encodes the enzyme 11b-hydroxysteroid dehydrogenase type 1 (11b-HSD1) which catalyses regeneration of cortisol from its inactive metabolite cortisone, thereby amplifying glucocorticoid receptor activation, e.g. in liver and adipose tissue. Both increased and decreased 11b-HSD1 activity has been associated with common disease. Glucocorticoid excess, e.g. in Cushing’s syndrome, is associated with features of the Metabolic Syndrome, including central obesity, hypertension and glucose intolerance. Transgenic overexpression of 11b-HSD1 in liver or adipose tissue in mice induces elevated local glucocorticoid concentrations and features of Metabolic Syndrome, while inhibition or disruption of 11b-HSD1 ameliorates features of the Metabolic Syndrome. In human obesity, 11b-HSD1 expression is increased in adipose tissue. In contrast, decreased 11b-HSD1 results in impaired regeneration of cortisol and hence increased metabolic clearance rate for cortisol; the resulting compensatory activation of the hypothalamicpituitary- adrenal axis may be responsible for adrenal androgen excess in some patients with polycystic ovary syndrome (PCOS). The few studies performed until now to estimate 11b-HSD1 activity in vivo in PCOS rely on the ratio of metabolites of cortisol and cortisone in urine. There are many flaws to this, which can explain the lack of consistency between the studies. This assay, in fact, reflects not just 11b-HSD1 activity, but the balance between opposing activities of 11b-HSD1 and 11b-HSD type 2, and changes in A ring reductases. In addition, urinary ratios are not able to determine tissue-specific changes in 11b-HSD1, as has been recently observed in human obesity. A variety of hormonal and nutritional factors regulate 11b-HSD1 expression. In addition, common non-coding single nucleotide polymorphisms (SNP) in HSD11B1 exist, allowing investigation of the genetic contribution to inter-individual variation in cortisol regeneration. SNPs in the 5’ flanking region (rs846910, A to G, ‘SNP 1’) and in an enhancer region in intron 3 (rs12086634, G to T, ‘SNP5’) have been associated independently with insulin resistance, type 2 diabetes and/or hypertension, but not with obesity in several, but not all, populations. Conversely, the G allele of SNP5, which causes lower 11b-HSD1 expression in vitro, is associated with hyperandrogenism amongst lean women with PCOS, although it is not more common amongst PCOS cases as a whole. PCOS is characterized by a high prevalence of all features of the Metabolic Syndrome, but the role played by HSD11B1, although attractive, is unknown. In addition, the consequences of the nonexonic polymorphisms for 11b-HSD1 expression and function in vitro (with the exception of SNP5) or in vivo have not been determined. In subcutaneous adipose tissue biopsies, SNP5 G allele was associated with low 11b-HSD1 mRNA in one patient, but neither SNP1 nor SNP5 genotype predicted 11b-HSD1 mRNA in adipocytes in 61 subjects from a native N American cohort. Urinary ratios of cortisol:cortisone metabolites are not associated with SNP5 genotype. In the present study we aimed to establish the consequences of SNP1 and SNP5 polymorphisms on features of Metabolic Syndrome and of PCOS, and further to define their influence on whole-body and liver or adipose tissue 11b-HSD1 activities. We genotyped a cohort of 600 women in whom metabolic and endocrine variables were carefully assessed, half of whom have PCOS, and selected 40 participants for a nested study in which we undertook detailed measurements of 11b-HSD1 activity in vivo. Whole-body 11b-HSD1 activity was assessed by measuring rate of appearance of 9,12,12-2H3- cortisol (d3-cortisol) during 9,11,12,12-2H4-cortisol infusion. Hepatic 11b-HSD1 activity was quantified by measuring the generation of plasma cortisol following an oral dose of cortisone 25mg after overnight dexamethasone suppression, whereas subcutaneous adipose 11b-HSD1 activity was assessed, in vitro, by measuring the conversion of 1,2,6,7-3H-cortisol to cortisone. In the same homogenates, obtained by sc abdominal fat biopsy, 11b-HSD1 mRNA was quantified. In the whole cohort, allele frequencies were 4% for SNP1 A and 15% for SNP5 G. No participants were homozygous for SNP1 A. There were no differences in the frequency of either SNP, or of haplotypes of both SNPs, between PCOS cases and controls. Very few participants had haplotypes GAGG (n=1) or GGGG (n=11) so they were not included in further analyses. Strikingly, neither SNP had independent effects on phenotype. However, there were differences between individuals of different haplotype. Individuals heterozygous for SNP1 A and homozygous for SNP5 T (GATT) had the highest waist circumference and, particularly amongst the PCOS group, higher plasma transaminase and triglyceride levels, and higher insulin response to the oral glucose tolerance test, as AUC. The proportion meeting the NCEP-ATPIII criteria for Metabolic Syndrome was also substantially higher in women with GATT haplotype (42.3% versus 18% in the whole cohort). The association between GATT haplotype and the Metabolic Syndrome was also observed by analysing PCOS women and controls separately. There were weaker associations between HSD11B1 genotype and hormonal features. Cortisol metabolites were measured in urine of the entire cohort of 600 women to assess relationships between HSD11B1 haplotypes and in vivo 11b-HSD1 activity. As previously reporter, individual SNPs did not predict urinary (5a-THF + 5b-THF)/5b-THE ratio. However, the GATT haplotype was associated with a higher ratio amongst women with PCOS, suggesting higher conversion of cortisone to cortisol. In the nested study, the group with the GATT haplotype had higher whole body rates of appearance of cortisol, reflecting the combination of adrenal cortisol secretion and net regeneration of cortisol by 11b-HSD1, and d3-cortisol, reflecting exclusively the contribution of 11b-HSD1. Rate of appearance of cortisol and d3-cortisol did not differ between PCOS and controls. Women with the GATT haplotype had higher transcript levels of 11b-HSD1 in subcutaneous adipose tissue only if they also had PCOS. Liver 11b-HSD1, measured as appearance of cortisol on first pass conversion after an oral dose of cortisone, was not significantly different between haplotypes, but was significantly lower in PCOS with respect to controls. Subcutaneous adipose tissue 11b-HSD1 trascript levels were positively and significantly correlated with insulin-resistance state and insulin levels. In conclusion, these data demonstrate that the GATT haplotype of HSD11B1 is associated with the Metabolic Syndrome and with increased whole body activity of 11b-HSD1, which appears to be independent of PCOS status. In addition, the whole-body activity is not altered in PCOS, suggesting that any differences which occur in 11b- HSD1 in one tissue are cancelled out by differences in other tissues. The role of tissue-specific dysregulation of 11b-HSD1 in PCOS must be clarified. However, we can speculate that upregulation of adipose 11b-HSD1 can be a key contributor to metabolic dysfunction in women with PCOS. In addition, in the present study we investigated the effect of two polymorphisms upon expression of 11b-HSD1; SNP1 and rs13306421, a rare polymorphism identified in a Japanese population, situated at -2 with respect to the translation start site (SNP-2). We did not find any effect of SNP1, thus suggesting that this SNP can be in linkage disequilibrium with another, functional, polymorphism. At variance, we found that SNP-2 modulates translation of the enzyme, with the A allele resulting in greater enzyme activity and a higher proportion of full-length enzyme synthesised in vitro. However, whether the SNP-2 is truly a polymorphism or may more properly be called a mutation is an open question. In fact, we did not find the A allele in our population of 300 PCOS patients, nor in the case controls. Furthemore, although the A allele was reported at 5% in the original Japanese population, in a further study, genotyping of a total of 210 individuals in four different populations detected only the G allele, as we did. Thus, further studies are required to confirm whether the A allele is indeed a polymorphism. Nevertheless, our studies show the clear potential to influence levels of active 11b-HSD1.