Adiponectin and HIV-Lipodystrophy: Taking HAART

Infection with HIV and consequent AIDS is a major public health problem affecting more than 40 million people worldwide. Successful combination therapy based on HIV protease inhibitors, reverse transcriptase inhibitors (both nucleoside analogs and nonnucleoside analogs), and fusion inhibitors, with the goal of complete suppression of viral replication and meaningful immune restoration [highly active antiretroviral therapy (HAART)], has improved the prognosis of patients to the point that long-term complications of treatment have become an important clinical concern. Such major complications that have emerged are metabolic and have been termed the “HIV-lipodystrophy syndrome,” although it is not clear that all of the described metabolic abnormalities are truly part of a single syndrome. These metabolic abnormalities were first associated with protease inhibitors, but both nonnucleoside analogs and at least one of the nucleoside analogs (stavudine) have also been implicated. The picture includes lipodystrophy, hypertriglyceride- and cholesterol-emia, and insulin resistance, with or without hyperglycemia (1, 2). This raises two major concerns. First is the specter of increased risk for cardiovascular disease, which could mitigate the prospect of longevity offered by HAART. Second is that cosmetic aspects as well as heightened concern about cardiovascular disease could impel patients to reject effective therapy. In this issue of Endocrinology, Xu et al. (3) describe a promising therapeutic approach in a mouse model of HIV-lipodystrophy. The lipodystrophy includes both the loss of facial and limb fat and central obesity, one or both of which may occur in a given individual (2). Atrophic adipose tissue contains small, poorly developed adipocytes and striking alterations in gene expression (see What Mechanism Might Underlie these Effects?) (4). The prevalence of lipodystrophy in patients on HAART is approximately 40% (2). Lipid levels are differentially affected by HIV-protease inhibitors, with ritonovir producing the most pronounced elevations. In a Swiss study, ritonavir caused increases in low-density lipoprotein cholesterol (50%), fasting triglycerides (201%), and, unexpectedly, Lp(a) (88%) (1). Frank diabetes is much more prevalent in HIV subjects with lipodystrophy than in those without it (5). Lipodystrophy patients have a 2-fold reduction in insulin sensitivity in clamp studies, and muscle is the primary locus of the defect (6). These alarming metabolic derangements, coupled with inflammatory markers related to AIDS per se, fuel concerns regarding cardiovascular risk from HAART, and raise the general hypothesis that a hormonal disturbance underlies them. Similarities between HIV lipodystrophy and Cushing’s syndrome suggested excess glucocorticoid action, but this has not been found (7). Investigative focus on hormones that are secreted from adipose tissue to influence caloric intake and disposition (adipocytokines) has produced intriguing clues, however. Among these, adiponectin has been implicated most convincingly in HIV lipodystrophy. Adiponectin is a peptide hormone secreted from adipocytes (8), and blood levels correlate with adipose content (9, 10). Two typical seven-transmembrane domain-type receptors (AdipoR1 and -R2) have been cloned (11). AdipoR1 predominates in muscle, whereas AdipoR2 is mostly hepatic. Administration of adiponectin to mice enhances fatty acid oxidation by muscle (12) and augments insulin action in the liver (13). Importantly, giving adiponectin to mice with insulin resistance associated with lipoatrophy or obesity improved insulin action and reduced lipid levels (14). On the other hand, adiponectin −/− mice exhibit insulin resistance and glucose intolerance (10) as well as evidence of vascular disease (15). Adiponectin levels are low in type II diabetics and are lower still in the subgroup with coronary artery disease (16). Peroxisome proliferator-activated receptor-γ transduces signals for genes that govern adipocyte differentiation and insulin-induced metabolic alterations (17). Thiazolidinediones, synthetic peroxisome proliferator-activated receptor-γ agonists that are useful in the treatment of diabetes, enhance the expression of adiponectin in obese mice (18) and in type II diabetic patients (19). Similar effects are seen in normal subjects, obese or not (20). Adiponectin is unique among hormones in that it acts primarily to promote the efficient action of another hormone, namely insulin. Correlative studies have implicated adiponectin in the HIV-lipodystrophy syndrome. Adiponectin levels in HIV-infected men with lipodystrophy are approximately 50% of those without lipodystrophy, and approximately 25% of those in uninfected controls (21). Adiponectin, but not leptin, correlates with insulin sensitivity, and the abnormal metabolic parameters in a plethora of recent studies (21–24), including one that showed reduced adiponectin mRNA in dystrophic fat (25). In this issue of Endocrinology, Xu et al. (3) directly address the hypothesis that adiponectin participates in the pathogenesis of the HIV-lipodystrophy syndrome by intervening in a mouse model. Administration of protease inhibitors caused a dose- and compound-related reduction in adiponectin and its mRNA in cultured fat cells in proportion to the propensity of each drug to produce lipodystrophy (ritonavir > indinavir > nelfinavir), and ritonavir also reduced adiponectin levels in mice. The mice also showed increased triglycerides and fatty acids, and enhanced hepatic fatty acid and cholesterol synthesis, thus simulating in the dyslipidemia associated with HAART in patients. Xu et al. (3) advanced beyond the correlative level by administering adiponectin to the mice. In their words, “a major novel finding of our present studies is that adiponectin replacement therapy can improve dyslipidemia induced by ritonavir.” This was manifest by reduced serum levels of triglyceride and fatty acids, associated with commensurate reversal of accelerated hepatic synthesis. Cholesterol metabolism, however, was unaffected. The HIV-lipodystrophy phenotype suggests a defect in both the control of adipocyte differentiation and lipid metabolism. Sterol response element-binding protein-1c (SREBP-1c, also known as adipocyte determination- and differentiation-dependent factor 1) plays a prominent role in both arenas (26, 27). SREBP-1c is an unusual transcription factor in that it is tethered to the endoplasmic reticulum by a hydrophobic tail. Proteolytic removal of the tail releases active SREBP-1c, which may travel to the nucleus, bind to DNA, and activate genes involved in long chain fatty acid synthesis and adipocyte differentiation (reviewed in Ref. 28). Activation of SREBP-1 is mediated by insulin (29). Recent microarray analysis showed that SREBP influences a wide spectrum of genes involved in the production of fatty acids. Indeed, mice overexpressing an active SREBP-1c exhibit peripheral lipoatrophy, severe insulin resistance, and hypertrigyceridemia, as is seen in patients taking HAART (30). Effects of HIV protease inhibitors on SREBP have been studied in mice and in humans. In mice, ritonavir causes hypertriglyceridemia, hypercholesterolemia, and fatty liver, with enhanced rates of hepatic sterol and fatty acid synthesis. This is associated with increased nuclear accumulation of active SREBPs in liver and fat cells, and reduced SREBP clearance via an HIV-protease inhibitor-sensitive proteosomal pathway appears to be the mechanism (31). Likewise, analysis of peripheral adipose tissue from patients receiving HAART shows increased SREBP-1c content despite a 93% reduction in SREBP-1c mRNA, again indicating marked stabilization of the protein (4). Thus, in addition to reducing the production of adiponectin, HAART may enhance the action of the critical transcription factor SREBP-1c, leading to increased fatty acid synthesis by activating genes in the liver. Moreover, SREBP excess induces morphological abnormalities in adipose depots, as seen in the SREBP-1c transgenic mice (30). HIV lipodystrophy syndrome is therefore a situation of reduced overall insulin action because of adiponectin deficiency, whereas lipogenesis, normally driven by insulin, is paradoxically enhanced because of SREBP-1c accumulation. Presumably both effects are ultimately related to the inhibition of proteosome function by the HIV drugs. Importantly, the findings of Xu et al. (3) indicate that adiponectin replacement therapy may functionally trump any direct effect of ritonavir on hepatic SREBP degradation pathways. The staging of a debate as to whether antiretroviral therapy should be modified based on cardiovascular risk at the recent International AIDS Conference highlights the practical challenge posed by the HIV-lipodystrophy syndrome (32). If further investigation confirms the utility of adiponectin replacement therapy, it will be reasonable to undertake trials in patients taking HAART, and in other insulin-resistant states as well. Two typical seven-transmembrane domain-type receptors; highly active antiretroviral therapy; sterol response element-binding protein-1c.