Characterisation of the relationship between Heme-oxygenase-1 and adiponectin

Abstract

The prevalence of obesity is growing at an alarming rate worldwide, and current therapies have limited benefits. Therefore, it is essential to increase our understanding of the underlying mechanisms so that new strategies can be developed to reduce the incidence of obesity related diseases. Adiponectin is an adipocyte-derived hormone that regulates glucose and lipid metabolism via direct and indirect mechanisms and has anti-inflammatory, anti-diabetic, anti-atherogenic and cardioprotective properties. Hypoadiponectinemia is implicated in the aetiology of obesity-related diseases making strategies to increase circulating adiponectin levels therapeutically attractive. Emerging evidence, predominantly from preclinical studies, suggests induction of heme-oxygenase-1 (HO-1) increases adiponectin production concomitant with decreased inflammatory tone prompting the proposal of a “HO-1 – adiponectin axis.” However, the underlying mechanisms of increased adiponectin production via HO-1 induction are poorly defined; indeed a direct relationship has not been demonstrated. Thus, the overarching aim of this thesis is to investigate the effects of HO-1 induction on adiponectin production as well as adipocyte and adipose tissue remodelling and metabolic parameters using cellular and mouse models. Initial studies were designed to characterize the direct effect of HO-1 induction on adiponectin production. We found that treatment with the widely used HO-1 inducer cobalt protoporphyrin (CoPP) or hemin for 24-48 h increased HO-1 expression and activity without affecting adiponectin expression and secretion, in human mature adipocyte under a variety of experimental scenarios. Treatment of adipocytes with TNFα reduced adiponectin production and induced pro-inflammatory cytokines production. HO-1 induction failed to reverse these effects. These results do not support a direct HO-1 – adiponectin axis. However, literature suggests that chronic induction of HO-1 via CoPP administration throughout differentiation, promotes adiponectin secretion, albeit in the context of reduced adipogenesis. While our previous findings argued against the existence of a direct “HO-1 – adiponectin axis,” they did not address the potential effects of chronic induction of HO-1 on adiponectin production. Thus, we extended our study to characterise the effects of chronic HO-1 induction throughout differentiation of human preadipocytes. In this study we demonstrate that chronic induction of HO-1 with CoPP or hemin throughout differentiation results in dose-dependent inhibition of adipogenesis and adiponectin production as well as induction of additional NRF2 target genes. Co-treatment with SnMP (HO-1 activity inhibitor) and HO-1 siRNA did not prevent these effects. These findings suggest that the chronic treatment with CoPP or hemin inhibits adipogenesis and adiponectin production potentially by a HO-1-independent mechanism that may be downstream of NRF2. However, our results contradict the majority of reports in the literature. One possibility is that the literature that supports “HO-1 – adiponectin axis” used mesenchymal stem cells, thus we propose the effect of CoPP is different in different cell types. In fact, recent reports show caloric restriction increases adiponectin production from bone marrow adipose tissue but not white adipose tissue (WAT) supports our proposal. The above findings indicate that induction of HO-1 failed to promote adiponectin production in vitro, specifically in WAT. Many studies in pre-clinical models have also shown that CoPP administration in obese mice reduces body weight gain, improves insulin sensitivity, reduces adipocyte size and attenuates liver steatosis. However, it should be noted that reduction in body weight gain which, in-itself, is predicted to decrease adipose tissue as well as systemic inflammation and to increase adiponectin. Thus, we again extended our study to characterise the effects of administration of CoPP in lean and obese mice. We found that CoPP administration significantly increased inflammatory tone in white adipose tissue irrespective of obesity. In obese mice, circulating adiponectin levels were elevated by CoPP administration whereas the adiponectin gene expression in WAT was significantly reduced. Additional effects specific to obese mice included reduced food intake, weight gain, adipocyte size and liver steatosis as well as improved insulin responsiveness. Co-administration with SnMP in obese mice ameliorated CoPP effects on adipose tissue inflammation and blunted CoPP effects on circulating adiponectin, but not other parameters. This supports the hypothesis that the induction of HO-1 may, at least partly, increase circulating adiponectin levels. We propose the increase adiponectin production from bone marrow adipose tissue (BMAT) may contribute to this, which is consistent with our in vitro findings. Further studies are needed to investigate if BMAT plays a central role in the effects of HO-1 induction on adiponectin production. In summary, the work presented in this thesis establishes CoPP administration in vitro does not have benefit on adipocytes, evidence against a direct “HO-1 – adiponectin axis” in WAT. CoPP administration in vivo promotes divergent effects on adipose tissue, adiponectin and insulin responsiveness. However, the majority of these effects appear to be independent of HO-1 activity. These factors suggest there is no positive effect of HO-1 induction on adiponectin production in WAT, and emphasize the potential role of BMAT in increased adiponectin production via CoPP administration.