Leptin, IL-6, and suPAR reflect distinct inflammatory changes associated with adiposity, lipodystrophy and low muscle mass in HIV-infected patients and controls

Inflammation has been intensively studied in age-associated complications. However, it is not clear whether it is the same inflammatory changes that are central in all age-associated complications. We therefore studied leptin, IL-6 and suPAR that we hypothesised would be associated distinctively with lipodystrophy, adiposity and sarcopenia. Leptin levels were significantly elevated in lipodystrophy of the Hypertrophic or Mixed type (Hyper/Mix) compared to HIV-infected patients without lipodystrophy (Non-Lipo). In contrast, IL-6 or suPAR levels were not changed in lipodystrophy, but tended to be higher in HIV-infected patients than Controls. Leptin was a strong marker of fat mass. In contrast, suPAR showed strong associations with low lean mass, and weak associations with fat mass. IL-6 was a marker of both increased fat and low lean mass. Thus we found leptin, IL-6 and suPAR to be associated with lipodystrophy, adiposity and sarcopenia distinctively.

For the first time we show that leptin levels were significantly elevated in patients with Hyper/Mix type lipodystrophy compared to Non-Lipo, not only as a result of the fat mass or the insulin sensitivity. In contrast, leptin levels were not significantly lower in patients with lipoatrophy, than what was expected from their low fat mass. Results from other studies suggest that the elevated leptin levels do not appear to be a general feature of age-associated AT re-distribution [19]. The elevated leptin levels in Hyper/Mix type lipodystrophy could be a physiological response to counteract ectopic fat-mediated lipotoxicity, as leptin exhibits lipotoxicity-protective effects [20]. However, the higher leptin levels could also represent a state of leptin resistance contributing to insulin resistance.

IL-6 and suPAR levels were not altered in lipodystrophy of either subtype, in agreement with most studies [18, 27]. In contrast, IL-6 and suPAR levels were elevated in Non-Lipo compared to Controls, though not significantly in multiple adjusted analyses. Along with the highly elevated CD8+ T cell counts and the lower CD4+/CD8+ T cell ratio, these findings support that the immune system has not been fully normalised in HIV-infected patients despite successful treatment, in agreement with previous studies [8]. The chronic immune activation could be involved in the earlier onset of comorbidities in HIV-infected patients.

FMI and tLMI and lLMI were positively associated, in agreement with the increase in lean mass required to carry an increased fat mass previously reported [31]. However, in Hyper/Mix the association appeared disturbed, also when removing an influential outlier. Whether this is an artefact, or whether Hyper/Mix type lipodystrophy is associated with dysregulation of muscle homeostasis as in sarcopenic obesity, is not possible to determine in this study.

Leptin and IL-6 were significantly associated with FMI and VAT in both univariate and multiple regression analyses. In contrast, suPAR was only significantly associated with FMI and VAT in multiple regression analyses, in agreement with previous studies [1, 19]. This did not appear to be a result of insulin resistance, since adjusting for HOMA-IR did not change the estimates. When adjusting for fat mass, IL-6 was independently associated with low lean mass, in contrast to the univariate analysis. Moreover, we confirmed our previous finding that suPAR reflects low lean mass in HIV-infected patients and extended this to healthy Controls [29]. Neither the association with suPAR nor IL-6 and low lean mass was explained by self-reported physical activity (data not shown). The association of suPAR with low lean mass may provide new insight into why suPAR is a risk marker for mortality.

Our findings, of inflammatory biomarkers being associated distinctively with age-related processes, are in agreement with other studies, where it was reported that inflammatory biomarkers were distinctively associated with disease severity, and with risk factors leading to disease development [30, 32]. This indicates that different inflammatory processes can be active simultaneously contributing to the systemic low grade inflammatory state. CCAAT-enhancer-binding proteins (C/EBP) and specificity protein (SP)-1 are transcription factors for both the leptin and IL-6 genes. Thus, on the molecular level, the association between leptin and IL-6 and adiposity could be the result of pathways leading to C/EBP and SP-1 binding to the promoter regions of the leptin and IL-6 genes [33‐35]. However, the different inflammatory processes also affect each other, as leptin and IL-6 affect the production of each other. It is not well understood, whether suPAR is regulated by, or regulates other inflammatory mediators [36, 37].

On the cellular level, adiposity-associated inflammation may result from an increased adipocyte size leading to AT stress. AT stress could induce adipocyte death, inflammation, and macrophage infiltration [38‐40]. However, inflammation can also induce AT stress, thereby initiating a vicious cycle. More specifically, enlarged adipocytes may contribute significantly to the elevated leptin levels, since leptin is produced in amounts proportional to fat cell size [41]. AT-derived IL-6 can be produced by stressed adipocytes or by infiltrating macrophages. Elevated IL-6 levels may contribute to disease development, by increasing the expression of cytokines like TNF-α known to induce insulin resistance, or by inducing insulin resistance itself [5, 42, 43].

Muscle leptin expression is very small, thus leptin does not appear to be involved in muscle pathology [33]. Inflammation could cause muscle atrophy by increasing Nuclear Factor-κB expression that is known to activate atrogen gene expression in muscle [26]. IL-6 may promote muscle catabolism directly through signal transducer and activator of transcription 3 resulting in Atrogin-1 expression. However, IL-6 anabolic effects have also been reported [24]. IL-6 catabolic effects may first ensue in a persistent inflammatory state and thus be dependent on other inflammatory mediators.

The uPAR ligand urokinase plasminogen activator (uPA) exhibits muscle anabolic properties, independently of its binding to uPAR [44‐47]. suPAR may directly affect muscle homeostatis by binding uPA. suPAR is cleaved from the cell surface by proteases involved in extracellular matrix degradation. The increased suPAR levels associated with low muscle mass, could therefore also be a result of the tissue re-modulation taken place during muscle atrophy [26, 28]. However, underlying pathophysiological processes could give rise to elevated suPAR and IL-6 levels, and low muscle mass, without any direct muscle atrophic effects of suPAR and IL-6.

There are some limitations to our study. The study was cross-sectional, and we could therefore not examine the role of leptin, IL-6, and suPAR in the development of complications. Moreover, our Control and lipodystrophy subgroups were small, and our findings should therefore be confirmed in larger studies. Furthermore, to understand whether the inflammatory changes identified here are indeed part of the pathophysiologic process, studies at the molecular level should be carried out. Our findings for patients with lipoatrophy are representative for relatively normo-metabolic patients, since inclusion criteria were no lipid-lowering or anti-diabetic treatment. Furthermore, our study was designed to assess lipodystrophy, and not lipodystrophy subtypes, and few patients with the lipohypertrophic phenotype were included. We therefore grouped patients with lipohypertrophy and mixed type lipodystrophy. Our findings support those from other studies, in that lipodystrophy phenotypes should be analysed separately [21, 48]. Finally, we could not determine the impact of the HIV infection itself, since no cART naïve HIV-infected group was included.