Background
The prevalence of metabolic diseases such as obesity, type 2 diabetes and obesity-associated hypertension is increasing gradually[
1]. Given that IR was shown to correlate with reduction of insulin-mediated glucose uptake in skeletal muscle and adipose tissue, IR was recently recognized as a key etiological factor of those metabolic disorders[
2]. The molecular mechanisms underlying the development of obesity-directed IR are not well understood. Several lines of evidence support the thesis, that the first step towards the development of autonomous insulin resistance in adipose tissue, as well as in the liver, is inflammation[
3‐
5].
Pro-inflammatory cytokines produced by adipocytes in fat tissue such as TNFalpha and IL-6 accelerate inflammatory responses of the surrounding tissue, and recruit pro-inflammatory cells. Recently we were able to show that the first inflammatory cells recruited to the adipose tissue during development of obesity-induced IR are CD4-positive T-lymphocytes[
6]. The recruitment of T-lymphocytes into fat tissue is likely mediated through resident adipocytes expressing stromal cell-derived factor-1 alpha (SDF-1alpha), a known attractant molecule for T-cells. In parallel, adipocytes produce monocyte chemoatractant protein-1 (MCP-1) for subsequent attraction of macrophages. Then, CD4-positive T-lymphocytes are able to induce pro-inflammatory responses in macrophages by the release of interferon gamma (IFNgamma)[
6,
7]. It is well known that macrophages which are recruited during the development of the obesity-related IR to fat tissue belong to the "pro-inflammatory" M1-phenotype[
3,
8]. Several independent research groups have demonstrated, that M1 macrophages are highly activated, sensitive to lipopolysaccharide (LPS) and free fatty acids (FFA), express F4/80, CD11b and CD11c markers as well as toll like receptors (TLR) 2 and 4, and produce a wide range of pro-inflammatory cytokines, such as TNFalpha and IL-6. This is in contrast to resident anti-inflammatory M2 macrophages, with a low sensitivity to LPS and FFA, lack of CD11c marker, and production of anti-inflammatory cytokines such as IL-4 and IL-10[
3,
8,
9]. Recently Stienstra and colleagues[
10] reported the M2- to M1-transition of resident adipose tissue specific macrophages (ATM) in rodents fed with HFD. Additionally, the authors showed that the transition could be inverted in HFD-fed mice treated with the PPARgamma agonist rosiglitazone.
PPARgamma belongs to the nuclear hormone receptor family of transcription factors, which are activated upon binding of specific ligands or agonists. PPARgamma is a key regulator of glucose and lipid metabolism by controlling energy homeostasis in adipose tissue, liver and skeletal muscle[
11]. Glitazones or Thiazolidinediones (TZDs), such as pioglitazone and rosiglitazone, are potent synthetic PPARgamma agonists used in clinic to treat type 2 diabetes[
12]. The activation of PPARgamma leads to improvement of systemic IR/glucose tolerance and the metabolic function of adipose tissue, liver and skeletal muscle. Recently we could demonstrate that a subgroup of angiotensin type 1 receptor (AT1R) blockers (ARBs) such as telmisartan act as a partial PPARgamma agonists, and show - similar to full agonists-beneficial metabolic effects in mouse model of HFD-induced IR[
13].
The aim of the present study was to elucidate the role of ligand-activated PPARgamma activation on T-lymphocyte-derived adipose tissue inflammation. Our data indicate that PPARgamma plays a central role in the development of IR, acting as an anti-inflammatory factor on T-lymphocyte activation and infiltration into fat tissue, and by this mean contributing to the attenuation of the systemic development of IR.
Discussion
The present study demonstrates that activation of PPARgamma reduces HFD-induced T-lymphocyte infiltration into adipose tissue. Both PPARgamma agonists, rosiglitazone and telmisartan, were able to reduce diet-induced T-lymphocyte accumulation in fat tissue. Subsequently, PPARgamma activation was linked to reduced macrophage infiltration into adipose tissue of HFD-fed mice. Anti-inflammatory actions of PPARgamma in adipose tissue were associated with an improved metabolic phenotype of mice treated with PPARgamma agonists for 5- and 10-weeks parallel to the HFD-intervention. T-lymphocytes were recently shown to initiate the inflammatory response associated with obesity-related IR. Our data demonstrate for the first time a new anti-inflammatory action of PPARgamma agonists targeting T-cells, which may contribute to the beneficial metabolic actions of PPARgamma ligands.
The ARB telmisartan has been shown to improve metabolic outcome of DIO-mice [
13]. Although the molecular mechanism of telmisartan´s action is not fully understood, its beneficial metabolic properties are linked to the ability to activate PPARgamma as a partial agonist, similarly to potent anti-diabetic drugs such as TZDs [
13,
16]. Telmisartan has also been demonstrated to modulate hepatic mitochondrial fatty acid oxidation due to the activation of hepatic PPARalpha[
17]. Along that line, a recently published study from Rong. X and colleagues on AT1R-deficient mice indicates for the first time that beneficial effects of telmisartan on diet-induced obesity, insulin resistance and hepatic triglyceride accumulation in mice appear to be AT1-R independent [
18].
Recent work published by Hernandez-Trujillo and colleagues [
19] demonstrated that the administration of rosiglitazone or losartan to mice fed a high-fat, high-cholesterol diet increases PPARgamma activity in both cases. Interestingly, treatment with rosiglitazone showed anti-atherogenic properties, such as a significant reduction of eNOS and CD36 expression, when compared to littermates treated with losartan. Furthermore, rosiglitazone has been shown to improve endothelial dysfunction in femoral arteries of Zucker diabetic fatty rats, mainly due to the regulation of eNOS expression and enhancement of endothelium-dependent vasorelaxation. Rosiglitazone treatment did not affect the mechanical properties of the artery as well as an expression of alpha-adrenoceptor (alpha-AR), MMP9 and elastase in this model [
20].
Importantly, treatment with both rosiglitazone and telmisartan led to the reduction of BW after 10 weeks of HFD, when compared to vehicle-treated littermates. Pair-feeding protocols reduced but not fully eliminated the differences in BW observed in HFD-fed mice treated with rosiglitazone and telmisartan. The BW difference can potentially have an impact on the anti-inflammatory and anti-diabetic properties of those two PPARgamma agonists.
In previous studies we characterized the role of T-lymphocytes in adipose tissue inflammation[
6]. Lymphocyte infiltration occurs during the early phases of IR development in HFD-induced obesity and precedes adipose tissue macrophage accumulation. Furthermore, it was demonstrated by us and the others, that activation of PPARgamma by TZDs or telmisartan reduces inflammatory T-cell activity, C-peptide-induced T-cell chemotaxis, and SDF-1- directed T-cell migration in vitro[
21‐
24], indicating an important function of PPARgamma in T-cells. Along this line, our present study shows that ligand-mediated PPARgamma activation prevents T-cell infiltration into adipose tissue which was associated with an improvement of diet-induced IR.
In their recent study Winer and colleagues identified, similar to our data, a residential population of T4-lymphocytes that controlled insulin resistance in visceral adipose tissue of HFD-fed mice. In elegant transplantation experiments with CD4
+ T-lymphocyte transfer into lymphocyte-free Rag1-null- and HFD-fed mice the authors showed reversed weight gain and improved IR of these mice in comparison to control littermates predominantly mediated through anti-inflammatory T
H2 cells[
15]. In addition, a distinct set of the CD4
+ and Foxp3
+- T-cells (T-regs) were recently identified in adipose tissue from lean rodents. These cells have been also suggested to play a regulatory role of the metabolic phenotype modulating an anti-inflammatory response in adipose tissue[
25]. Accordingly, the population of T-reg cells was dramatically reduced in adipose tissue of mice with obesity-related IR confirming a beneficial function of this T-cell population. It appears that a pro- and anti-inflammatory subset of CD4
+ T-lymphocytes does exist in adipose tissue, and the balance betweens these populations likely determines the phenotype. In our study we did not investigate CD4
+ T-lymphocyte subsets; however, as previously described in other studies using the HFD model, pro-inflammatory T
H1 cells are the predominant subset during weight gain. Therefore it is likely that PPARgamma ligands mainly target T
H1 cells in our model to mediate their anti-inflammatory action.
In addition to the reduction of T-lymphocyte accumulation by PPARgamma ligands we demonstrate that the amount of macrophages is reduced in groups treated with rosiglitazone or telmisartan. Accordingly to the characterization of CD4
+ T-lymphocyte subsets in adipose tissue, distinct subpopulations (pro-inflammatory M1 and anti-inflammatory M2) of macrophages have been identified in adipose tissue[
3,
10]. PPARgamma has been recently demonstrated to promote infiltration of M2 (anti-inflammatory) macrophages into adipose tissue in DIO model in mice[
10]. In this work mice were fed with HFD over 20-weeks, followed by 1-week treatment with rosiglitazone (0.01% w/w) the authors reported a striking up-regulation of the MCP-1 expression in the adipose tissue of rosiglitazone-treated mice, which is in contrast to our data, as well as to the results in human and murine adipose tissue published by others[
5,
26]. This discrepancy between our data and results published by Stienstra and colleagues could be explained, at least in part, by the different DIO model used in the study. Stienstra and colleagues investigated chronic fat tissue inflammation followed by a short pharmacological intervention with rosiglitazone. We were interested in the early stages of IR development, and mice used in our work were set on HFD parallel to the rosiglitazone/telmisartan intervention for over 10-weeks. In our mice model we could observe a clear reduction of adipose tissue MCP-1 by PPARgamma agonists, which is in agreement with previously published data[
5]. Stienstra and collaborators observed also enhanced infiltration of M2 macrophages after rosiglitazone intervention, when compared to vehicle-treated animals. This is in agreement with data published by Bassaganya-Riera and others[
9], showing that deficiency of PPARgamma in immune cells results in significant upregulation of the M1 markers and repression of M2 markers characteristic for macrophages in white adipose tissue of mice with HFD-induced obesity. Lack of PPARgamma led to the dominance of M1 differentiation in this model. Following this line, reduction of MCP-1 and macrophage markers by PPARgamma ligands in our study likely reflect the diminution of pro-inflammatory M1-macrophages during the early development of obesity/IR. Enhancement of M2-macrophage responses by PPARgamma activation might be a feature of late phase during HFD feeding.
Although our study clearly indicate that macrophage infiltration and the progression of fat tissue inflammation and IR could be specifically reduced by the intervention with PPARgamma agonists, one has to take into account that conclusion derived form animal studies in general cannot be automatically extrapolated to human or to clinical studies, as treatment with rosiglitazone was recently linked with increased cardiovascular risk [
27].
Acknowledgements
This work was supported by GlaxoSmithKline GMBH Co.KG, Munich, Germany and Boeringer Ingelhein Pharma. A.F.L. is supported by the Deutsche Forschungsgemeinschaft (DFG-KI 712/5-1, KFO 218/1). N.M. is supported by Deutsche Forschungsgemeinschaft (MA 2047/4-1), the European Foundation for the Study of Diabetes, and the Novartis-Stiftung für therapeutische Forschung, T.U. is supported by the Deutsche Forschungsgemeinschaft (DFG-GK 754III, DFG-GK 865-II) U.K. is supported by the Deutsche Forschungsgemeinschaft (DFG-KI 712/3-2, DFG-KI 712/5-1 (FG 1054/1), DFG-KI 712/6-1 (KFO 218/1). The authors thank A. Hausauer and B. Thalke for help with immunohistochemical analysis.
Competing interests
NM, TU, UK have received research grants and/ or speaker fees from GSK, BI, Bayer.
Authors' contributions
AFL contributed to research data, contributed to discussion, wrote and edited MS, MH contributed to research data and edited MS, MC contributed to research data, CS contributed to research data, KH contributed to research data, NM contributed to discussion, reviewed and edited manuscript, TU contributed to discussion, reviewed and edited manuscript and UK designed the study, wrote the MS, contributed to discussion, reviewed and edited manuscript. All authors read and approved the final manuscript.