Osteoprotegerin increases in metabolic syndrome and promotes adipose tissue proinflammatory changes

Molecular and Cellular Endocrinology 394 (2014) 13–20

Contents lists available at ScienceDirect

Molecular and Cellular Endocrinology journal homepage: www.elsevier.com/locate/mce

Osteoprotegerin increases in metabolic syndrome and promotes adipose tissue proinflammatory changes Stella Bernardi a,b,⇑,1, Bruno Fabris c,1, Merlin Thomas b, Barbara Toffoli d, Christos Tikellis b, Riccardo Candido e, Cristiana Catena g, Paolo Mulatero h, Fabio Barbone f, Oriano Radillo d, Giorgio Zauli d, Paola Secchiero a a

Department of Morphology, Surgery and Experimental Medicine, LTTA Centre, University of Ferrara, Via Fossato di Mortara 66, 44100 Ferrara, Italy Baker IDI, Heart and Diabetes Research Institute, 75 Commercial Road, Melbourne, VIC 3004, Australia Department of Medical, Surgical and Health Sciences, University of Trieste, Ospedale di Cattinara, Strada di Fiume 447, 34149 Trieste, Italy d Institute for Maternal and Child Health, IRCCS Burlo Garofolo, 34100 Trieste, Italy e Diabetological Centre, via Puccini 48/50, 34148 Trieste, Italy f Department of Medical, Experimental and Clinical Sciences, University of Udine, Ospedale Santa Maria della Misericordia, Udine, Italy g Department of Medical and Biological Sciences, University of Udine, Ospedale Santa Maria della Misericordia, Udine, Italy h Division of Internal Medicine and Hypertension, University of Torino, Ospedale San Giovanni Battista, 10126 Torino, Italy b c

a r t i c l e

i n f o

Article history: Received 6 October 2013 Received in revised form 8 June 2014 Accepted 9 June 2014 Available online 3 July 2014 Keywords: Metabolic syndrome High-fat diet Adipose tissue Osteoprotegerin Inflammation

a b s t r a c t Background: Inflammation is believed to link obesity to insulin resistance, as in the setting of metabolic syndrome (MetS). Osteoprotegerin (OPG) is a soluble protein that seems to exert proatherogenic and prodiabetogenic effects. This study aims at determining OPG levels in MetS and whether OPG might contribute to MetS development and progression. Methodology/principal findings: Circulating OPG was measured in 46 patients with MetS and 63 controls, and was found significantly elevated in those with MetS. In addition, circulating and tissue OPG was significantly increased in high-fat diet (HFD) fed C57BL6 mice, which is one of the animal models for the study of MetS. To evaluate the consequences of OPG elevation, we delivered this protein to C57BL6 mice, finding that it promoted systemic and adipose tissue proinflammatory changes in association with metabolic abnormalities. Conclusions/significance: These data suggest that OPG may trigger adipose tissue proinflammatory changes in MetS/HFD-induced obesity. Ó 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Recent studies show that type 2 diabetes mellitus (T2DM) is reaching epidemic proportions and that its prevalence is strongly associated with overall obesity and central obesity, mainly due to unhealthy changes in lifestyle (Yang et al., 2010). It has been shown that obesity is in fact an independent predictor of insulin

⇑ Corresponding author at: Department of Morphology, Surgery and Experimental Medicine and LTTA Centre, University of Ferrara, Via Fossato di Mortara 66, 44100 Ferrara, Italy. Tel.: +39 3339534214. E-mail addresses: [email protected] (S. Bernardi), b.fabris@fmc. units.it (B. Fabris), [email protected] (M. Thomas), [email protected] (B. Toffoli), [email protected] (C. Tikellis), [email protected] (R. Candido), [email protected] (C. Catena), [email protected] (P. Mulatero), [email protected] (F. Barbone), [email protected] (O. Radillo), zla. [email protected] (G. Zauli), [email protected] (P. Secchiero). 1 The first two authors have equally contributed to this work. http://dx.doi.org/10.1016/j.mce.2014.06.004 0303-7207/Ó 2014 Elsevier Ireland Ltd. All rights reserved.

resistance and T2DM (Collins et al., 2011), which on the other hand is prevented by weight loss (Tuomilehto et al., 2001). Among the multiple mechanisms linking obesity to T2DM, inflammation is a common feature that has been implicated in the pathology of both diseases. Several evidences (Esposito et al., 2003; Bastard et al., 2000; Ryan and Nicklas, 2004) have in fact proven the existence of an association between obesity, low-grade inflammation, and metabolic disturbances, such as insulin resistance and T2DM. Metabolic syndrome (MetS) is a condition that clusters obesity, low-grade inflammation, and insulin resistance and predicts the future risk of diabetes and cardiovascular diseases (CVD) (Kahn et al., 2005). Although insulin resistance was initially believed to be the major underlying pathological process (Reaven, 1988), it has then been shown that the etiology of MetS was related to abnormalities in adipose tissue, or to an altered inflammatory state (Carr et al., 2004; Alberti et al., 2006); consequently obesity is now considered the first step in the etiological cascade leading to the

14

S. Bernardi et al. / Molecular and Cellular Endocrinology 394 (2014) 13–20

other MetS disturbances. In particular, the hypothesis of what links obesity to MetS development relies on the understanding that the white adipose tissue (WAT) is an endocrine organ, whose secretory pattern changes in obese subjects as it gets inflamed (Guilherme et al., 2008) and releases more proinflammatory molecules that would impair insulin sensitivity. Experimental evidence shows that in obesity WAT contains an increased number of macrophages, which are obviously a potential source of proinflammatory factors that influence adipocyte biology and systemic insulin resistance (Weisberg et al., 2003; Xu et al., 2003; Di Gregorio et al., 2005). However, whilst it is becoming clear that obesity-related insulin resistance is, at least in part, a chronic inflammatory disease initiated in the WAT, the molecular signals that turn the increased adiposity into an inflamed adiposity, thereby triggering macrophage infiltration and promoting metabolic disturbances, are less clear (Neels and Olefsky, 2006). Osteoprotegerin (OPG) is a soluble protein acting as decoy receptor of RANKL (receptor activator for nuclear factor jB ligand) and TRAIL (TNF-related apoptosis-inducing ligand) (Zauli et al., 2009) and it is exactly for its ability to block RANKL that OPG was initially identified as a key regulator in bone turnover (Boyle et al., 2003; Simonet et al., 1997). Not only is OPG secreted by osteoblasts (Hofbauer and Schoppet, 2004) but it is also produced by a wide range of tissues, such as the endocrine pancreas (Schrader et al., 2007), as well as different types of cells, including endothelial (Malyankar et al., 2000), smooth muscle cells (Zhang et al., 2002), and adipocytes (An et al., 2007). Interestingly, experimental evidence would suggest that the RANKL-OPG-TRAIL pathway is implicated in the regulation of glucose homeostasis (Bernardi et al., 2012b; Browner et al., 2001; Kiechl et al., 2013; Secchiero et al., 2006). Nevertheless, conflicting data have been found so far on the relationship between OPG and MetS. In this study, we aimed at evaluating whether OPG circulating levels change in MetS. We hypothesized that OPG could increase in the setting of MetS contributing to its development and progression. Therefore, a further aim of this study was to evaluate whether an increase in OPG could contribute to adipose tissue proinflammatory changes, which seem to be fundamental to the pathogenesis of MetS. 2. Material and methods 2.1. Clinical study 2.1.1. Subject selection To evaluate OPG levels in MetS, 46 patients with newly diagnosed MetS (cases) along with 63 healthy subjects matched by age and sex (controls) were consecutively selected from the subjects referring to three hospital-based specialized Internal Medicine Clinics, over a period of 18 months. The exclusion criteria were: age below 18 or above 65 years, history or clinical evidence of cardiopulmonary, renal, or hepatic diseases. MetS was diagnosed according to the International Diabetes Federation (IDF) definition (Alberti et al., 2006). After the initial screening visit at our Clinics and before blood sampling, all the subjects selected were asked to sign a written informed consent for participating in this study, whose protocol had been previously approved by the Institutional Ethics Committee of the University of Trieste (AOUTS – Azienda Ospedaliero Universitaria di Trieste). 2.1.2. Laboratory tests Blood samples were collected at 08.00 a.m., after overnight fasting. Glucose, insulin, total cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides and C-reactive protein (CRP) were measured by autoanalyzer. Low-density lipoprotein (LDL)

cholesterol was calculated by the Friedwald’s formula. Insulin sensitivity was calculated according to the formula of the homeostasis model assessment (HOMA) index method: insulin resistance = fasting plasma insulin (lUI/ml)  fasting plasma glucose (mmol/ l)/22. Plasma OPG was measured by ELISA, according to the manufacturer’s instructions (Alexis biochemical distributed by Axxora.com; Cat#Alx-850-280A-KI01) 2.2. Experimental studies 2.2.1. Animals Study 1. High-fat diet (HFD) fed mice represent an useful animal model for the study of MetS (Fraulob et al., 2010; Gallou-Kabani et al., 2007). So, in order to confirm OPG increase in the setting of MetS, 18 adult (8-wk-old) male C57BL6 mice were randomly allocated to a standard chow diet (C57 chow = 9), or a HFD (C57 HF = 9) for 12 weeks. The animals were kept in a temperature-controlled room (22 ± 1 °C) on a 12-h light/dark cycle with free access to food and water and they were fed ad libitum for the length of the study. The standard chow diet had 19.6% of protein, 4.6% of fat, and 4.5% of crude fibre, providing a digestible energy of 14.3 MJ/kg. The HFD had 22.6% of protein, 23.5% of fat, and 5.4% of crude fibre, providing a digestible energy of 19 MJ/kg. Intraperitoneal (ip) glucose and insulin tolerance tests (IPGTT and IPITT) were performed at week 6 and 12. As for the IPGTT, glucose (2 g/kg of body weight) was injected intraperitoneally after an overnight fast and bloods were collected from the tail tip at baseline, 15, 60, and 120 min after glucose injection. Blood glucose was measured by an automatic glucometer. Then blood samples were centrifuged at 6000 rpm for 6 min and serum insulin was measured by ELISA. As for the IPITT, insulin (1 unit/kg of body weight) was injected intraperitoneally after a 6-h fast, bloods were collected as before, and blood glucose was measured by an automatic glucometer. At the end of the study, total body mass and fat mass were measured by EchoMRI (Echo Medical Systems), and systolic blood pressure by tail-cuff pletismography. Then, the animals were anethestized by an ip injection of pentobarbitone (Euthal, Delvet, NSW, Australia) at a dose of 100 mg per kg of body weight. Blood was collected from the left ventricle, centrifuged and plasma was stored at 20 °C for analysis. Epididymal white adipose tissue (WAT), pancreases, and livers were collected and either snap-frozen and stored at 70 °C or fixed for histological analysis. This study was carried out at the Baker IDI Institute and was approved by the AMREP Animal Ethic Committee (ID 0796/2009). Study 2. In order to determine the significance of OPG increase, 18 adult (8-wk-old) male C57BL6 mice were randomized to receive either human recombinant full-length OPG (C57 OPG, n = 9) or vehicle (C57 veh, n = 9) every 3 weeks for 12 weeks. Human recombinant full-length OPG (R&D Systems, Minneapolis) was delivered intraperitoneally at a dose of 1 lg per mouse in a total of 200 ll of HEPES-buffered saline. The animals were kept in a temperature-controlled room (22 ± 1 °C) on a 12-h light/dark cycle with free access to food (standard chow diet) and water and they were fed ad libitum for the length of the study. An IPGTT was performed at week 12, as before. At the end of the study, after measuring body weight and systolic blood pressure, all the animals were anesthetized and sacrificed for collecting their plasma and WAT, as above. This study was carried out at Cattinara University Hospital and was approved by its Animal Ethic Committee (ID 28.0.2008). In both studies, principles of laboratory animal care as well as specific national laws were followed where applicable. 2.2.2. Glucose, OPG, proinflammatory cytokine, and lipid measurement Glucose was measured using an automatic glucometer (AccuCheck II; Roche) during the IPGTT, the IPITT, and at the end of both studies, at fasting. Insulin was measured by ELISA (Millipore,

15

S. Bernardi et al. / Molecular and Cellular Endocrinology 394 (2014) 13–20

Cat#EZRMI-13K) in the sera collected during the IPGTT. Circulating mouse OPG was measured by ELISA (R&D System; Cat#DY459) in study 1 and study 2, circulating human OPG was measured by ELISA (R&D System; Cat#DY805) in study 2. Interleukin-6 (IL-6), macrophage chemoattractant protein-1 (MCP-1), tumor necrosis factor-a (TNF-a) were measured by a multiplex kit (Millipore; Cat#MADPK-71K). Total cholesterol, LDL cholesterol, HDL cholesterol as well as triglycerides were measured by COBAS Integra 200 in study 1, and by autoanalyzer in study 2. 2.2.3. Gene expression quantification by Real-time PCR For gene expression analysis 3 lg of total RNA extracted from the pancreas, WAT, and liver were used to synthesize cDNA with Superscript First Strand synthesis system for RT-PCR (Gibco BRL, Invitrogen). IL-6, MCP-1, OPG, and TNF-a gene expression was analysed by real-time quantitative RT-PCR using the TaqMan system based on real-time detection of accumulated fluorescence (ABI Prism 7900 HT, Perkin-Elmer Inc). Fluorescence for each cycle was quantitatively analysed by an ABI Prism 7900 HT Sequence Detection System. To control for variation in the amount of cDNA available for PCR in the samples, gene expression of the target sequence was normalized in relation to the expression of an endogenous control, 18S ribosomal RNA (rRNA) (18s rRNA TaqMan Control Reagent kit; ABI Prism 7900 HT, Perkin-Elmer Inc). Primers and TaqMan probes for IL-6, MCP-1, OPG, TNF-a and the endogenous reference 18S rRNA were constructed with the help of Primer Express (ABI Prism 7900 HT, Perkin-Elmer Inc.). Every amplification was performed with the following time course: 2 min at 50 °C, 10 min at 95 °C, 40 cycles of 20 s at 95 °C, and 1 min at 60 °C. Results were expressed as arbitrary units relative to the respective control tissue values, which were arbitrarily assigned a value of 1. 2.2.4. Immunostainings After neutralization of endogenous peroxidase, 20-lm frozen sections of WAT were incubated overnight with a rat anti-mouse CD68 (Serotec; Cat# MCA1957S) at a dilution of 1:50. Biotinylated rabbit anti-rat antibodies diluted 1:200 were then applied as secondary antibodies (Vector Laboratories), followed by the amplification of the signal by ABC kit (Vector Laboratories), according to the manufacturer’s instructions. The staining was then visualized by reaction with 3,30 -diaminobenzidine tetrahydrochloride (Sigma Chemicals). After counterstaining with hematoxylin, all the sections were examined by light microscopy and digitized using a high-resolution camera. Macrophages were counted as total cells positive for CD68 per frame. A total of 45 frames (5 frames per mouse) per each group were analysed. 2.3. Statistical analysis Results were expressed as mean ± standard error of the mean (SEM). A p value <0.05 was considered statistically significant. The Kolmogorov–Smirnov test was applied to continuous variables to check for distribution normality. In clinical studies, cases of MetS and controls were compared using either the t-test for independent samples or the Mann– Whitney U test, where appropriate. The Spearman rank correlation coefficient was calculated to evaluate the correlation between OPG and age, sex, CRP, body mass index (BMI), waist circumference, fasting glucose, insulin, HOMA, total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides, systolic blood pressure (SBP), and diastolic blood pressure (DBP), one at the time. To evaluate the association of high serum levels of OPG with metabolic syndrome controlled for variables initially associated with metabolic syndrome and with OPG in the bivariate analyses, we tested a combination of multiple logistic regression models using SAS 9.3. The

final multivariate model will be shown which includes all the individual variables that remained statistically significant (p < 0.10). In animal studies, analysis of variance (ANOVA) was performed to evaluate the differences in continuous variables among groups. Comparisons of group means were performed by Fisher’s least significant differences (LSD) method unless otherwise specified. 3. Results 3.1. Circulating levels of OPG are significantly associated with MetS Clinical and metabolic parameters of patients with MetS and their matched controls are shown in Table 1. Consistent with the definition of the syndrome, patients with MetS were obese, displayed significantly increased waist circumference and had higher levels of diastolic blood pressure; in addition, they presented significantly higher levels of glucose and triglycerides as well as lower levels of HDL cholesterol at fasting, as compared to the controls. CRP levels were also significantly increased. Circulating OPG was significantly higher in patients with MetS (p < 0.0001) as compared to the controls (Table 1). In this study, OPG was found to correlate with HDL cholesterol (F test = 7.03; p < 0.05) and CRP (F test = 13.27; p < 0.001). The difference in OPG levels between patients with MetS and controls was tested for potential confounders such as age, sex, BMI, waist circumference, glucose, insulin, HOMA index, total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides, SBP, DBP, and CRP. In Table 2 the output of the multivariate logistic regression model is presented. After disregarding other weaker associations (data not shown), and after adjustment for CRP and HDL cholesterol, OPG was still significantly associated with MetS (p < 0.0377). 3.2. HFD promotes the development of obesity, impaired glucose tolerance and low-grade inflammation in mice At the end of the study HFD fed mice developed obesity, impaired glucose tolerance/insulin resistance, and low-grade inflammation, which are not only some of the typical features of MetS (Alberti et al., 2006), but also the main underlying pathological processes of this syndrome (Kahn et al., 2005), confirming the view that this model can be used for the study of MetS (Fraulob et al., 2010; Gallou-Kabani et al., 2007). First of all, C57 HF mice became obese, weighing 45.4 ± 1.1 g while C57 chow mice

Table 1 General parameters and metabolic features of the study groups. Variable

CNT (n = 63)

MetS (n = 46)

Age (years) Sex (men/women) BMI (kg/m2) Waist circumference (cm) Glucose (mM) Insulin (lU/ml) HOMA index Total cholesterol (mM) LDL cholesterol (mM) HDL cholesterol (mM) Triglycerides (mM) Clinic SBP (mm Hg) Clinic DBP (mm Hg) CRP (mg/L) OPG (ng/L)

49.31 ± 1.62 29/34 24.97 ± 0.47 87.10 ± 1.41 5.05 ± 0.13 10.26 ± 0.64 2.26 ± 0.14 5.53 ± 0.12 3.45 ± 0.11 1.56 ± 0.05 1.14 ± 0.06 138.72 ± 1.72 84.09 ± 1.04 2.16 ± 0.19 68.10 ± 2.55

50.69 ± 1.61 21/25 30.82 ± 0.65§ 103.20 ± 1.76§ 5.60 ± 0.17§ 17.30 ± 1.89§ 3.87 ± 0.36§ 5.59 ± 0.14 3.60 ± 0.13 1.18 ± 0.04§ 1.89 ± 0.19§ 146.07 ± 2.01** 90.87 ± 1.35§ 3.56 ± 0.33§ 80.62 ± 4.21§

Data are expressed as mean ± SEM. BMI, body mass index; HOMA, homeostasis model assessment; LDL, low-density lipoprotein; HDL, high-density lipoprotein; SBP, systolic blood pressure; DBP, diastolic blood pressure. ** p < 0.01 vs CNT; § p < 0.0001 vs CNT.

16

S. Bernardi et al. / Molecular and Cellular Endocrinology 394 (2014) 13–20

Table 2 Association of OPG, CRP and HDL with metabolic syndrome from a mutually adjusted multiple logistic regression model. Dependent variable: metabolic syndrome Predictive variables

DF

b Estimate

Standard error

Wald Chi-square

p-Value

OPG CRP HDL

1 1 1

0.0283 0.3200 0.0992

0.0136 0.1773 0.0255

4.3174 3.2564 15.0851

0.0377 0.0711 0.0001

Model R-square = 0.3364. CRP, C-reactive protein; HDL, high-density lipoprotein; OPG, osteoprotegerin.

weighed 37.1 ± 1.4 g (p < 0.005). In particular, C57 HF mice had 15.0 ± 0.9 g of fat while C57 chow mice had 6.7 ± 0.7 g of fat (p < 0.005). Secondly, the high dietary content of fat led to hyperglycemia, hyperinsulinemia and reduced insulin sensitivity. The IPGTT (Fig. 1A and B) showed that the HFD resulted in a significant impairment of glucose clearance, leading to hyperglycemia at 15, 60, and 120 min after a glucose load, which could already be seen 6 weeks after the start of the HFD, while insulinemia

increased significantly 12 weeks after (Fig. 1C and D). The IPITT (Fig. 1E and F) showed that the HFD impaired insulin sensitivity, causing insulin resistance. Thirdly, the HFD promoted the development of a systemic low-grade inflammatory response, as C57 HF mice displayed an elevation in circulating IL-6, MCP-1 and TNF-a (Fig. 2A) with a significant upregulation of the same molecules in their WAT (Fig. 2B). On the other hand, the HFD did not change BP and, although it caused dyslipidemia, the changes observed in

*

*

*

* *

*

*

*

*

*

*

*

*

*

*

*

Fig. 1. Glucose and insulin in HFD fed mice. Blood glucose (mM) at 0, 15, 60, and 120 min during an IPGTT performed after 6 weeks (A) and 12 weeks (B) of HFD; Serum insulin (mM) at 0, 15, 60, and 120 min during an IPGTT performed after 6 weeks (C) and 12 weeks (D) of HFD; Blood glucose (mM) at 0, 15, 60, and 120 min during an IPITT performed after 6 weeks (E) and 12 weeks (F) of HFD. Data are reported as mean ± SEM; *p < 0.05 vs C57 chow. Chow, chow diet fed; HF, high-fat diet fed; HFD, high-fat diet; IPGTT, intraperitoneal glucose tolerance test; IPITT, intraperitoneal insulin tolerance test.

17

S. Bernardi et al. / Molecular and Cellular Endocrinology 394 (2014) 13–20

*

*

*

*

*

† *

Fig. 2. HFD-induced proinflammatory changes and OPG gene expression. (A) Circulating levels of IL-6, MCP-1, and TNF-a in C57 chow (n = 9) and C57 HF (n = 9) mice; (B) IL-6, MCP-1, and TNF-a mRNA expression in white adipose tissue, reported as relative gene units; (C) Pancreatic, white adipose tissue, hepatic, and skeletal muscle OPG mRNA expression is reported as relative gene units; data is expressed as mean ± SEM. Data is expressed as mean ± SEM. *p < 0.05 vs C57 chow;  p < 0.005 vs C57 chow. Chow, chow diet fed; HF, high-fat diet fed; IL-6, interleukin-6; MCP-1, monocyte chemotactic protein-1; OPG, osteoprotegerin; TNF-a, tumor necrosis factor-a.

the lipid profile did not resemble those of human MetS, given that HDL cholesterol increased significantly while LDL cholesterol and triglycerides remained the same (Table 3). 3.3. OPG increases in HFD fed mice In C57 HF mice, which represent one of the experimental models for the study of MetS (Fraulob et al., 2010), circulating levels of OPG were significantly elevated as compared to their controls (Table 3). Consistent with this, the HFD significantly increased OPG gene expression in the pancreas, WAT, and liver (Fig. 2C). 3.4. Repeated OPG injections induce metabolic and proinflammatory changes in mice To investigate the significance of such an increase of OPG, we analysed the metabolic and proinflammatory effects of OPG delivery in chow fed C57 mice. First of all repeated OPG injections increased circulating human OPG, although not significantly, being human OPG 217.2 ± 22.3 ng/L in the treatment group and 190.2 ± 10.5 in the control group. On the other hand, the same treatment induced a significant increase of murine OPG, as reported in Table 3. Secondly, OPG delivery significantly increased

glucose at fasting (Table 3) and at all time points after a glucose load. The IPGTT showed that in OPG-injected mice glucose increased from 10.6 ± 0.6 mM at baseline to 16.5 ± 1.0 mM at 15 min, 13.8 ± 0.6 mM at 60 min, and 16.8 ± 1.1 mM at 120 min after a glucose load. These values were significantly higher than those of C57 veh mice, where glucose increased from 8.5 ± 0.7 mM at baseline to 13.1 ± 0.9 mM at 15 min, 10.2 ± 1.2 mM at 60 min, and 8.3 ± 0.8 mM at 120 min after the same glucose load (p < 0.05 vs C57 OPG mice). In the third place, OPG delivery led to circulating and peripheral proinflammatory changes. In OPG-injected mice, both circulating levels and gene expression of MCP-1 and TNF-a from WAT increased significantly (Fig. 3A and B). Such changes were associated with a significant increase in the number of macrophages infiltrating the WAT of OPG-injected mice (Fig. 3C and E). As previously reported (Candido et al., 2010), OPG delivery did not change BP levels nor the lipid profile (Table 3). 4. Discussion This work demonstrates that circulating OPG is not only significantly associated with MetS in humans and but also significantly higher in HFD fed mice, which represent an experimental model

Table 3 General parameters and metabolic features of the experimental study groups. Variable

C57 chow (n = 9)

C57 HF (n = 9)

C57 veh (n = 9)

C57 OPG (n = 9)

Body weight (g) Glucose (mM) Total cholesterol (mM) LDL cholesterol (mM) HDL cholesterol(mM) Triglycerides (mM) Clinic SBP (mmHg) OPG (ng/L)

37.1 ± 1.4 10.1 ± 0.7 1.6 ± 0.3 0.4 ± 0.1 0.5 ± 0.2 1.3 ± 0.1 88.4 ± 1.5 364.1 ± 38.2

45.4 ± 1.1* 11.7 ± 0.6 3.7 ± 0.2* 1.0 ± 0.1 2.3 ± 0.1* 0.9 ± 0.1 85.3 ± 3.1 411.9 ± 26.7*

31.0 ± 1.0 8.7 ± 0.4 2.0 ± 0.2 0.5 ± 0.1 1.3 ± 0.2 0.6 ± 0.1 86.3 ± 10.5 431.6 ± 33.6

33.0 ± 1.5 10.4 ± 0.7** 2.3 ± 0.5 0.4 ± 0.1 1.5 ± 0.1 0.8 ± 0.2 94.2 ± 8.7 556.4 ± 47.1*

Data are expressed as mean ± SEM; Chow, chow diet; HF, high-fat diet; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SBP, systolic blood pressure; veh, vehicle. * p < 0.05 vs C57 chow. ** p < 0.01 vs C57 veh.

18

S. Bernardi et al. / Molecular and Cellular Endocrinology 394 (2014) 13–20

Fig. 3. OPG-induced proinflammatory changes and white adipose tissue macrophages infiltration. (A) Circulating levels of IL-6, MCP-1, and TNF-a in C57 veh (n = 9) and C57 OPG (n = 9) mice; (B) IL-6, MCP-1, and TNF-a mRNA expression in white adipose tissue, reported as relative gene units; (C) Semi-quantitative analysis of macrophages infiltration into white adipose tissue expressed as number of macrophages per field. (D and E) Representative sections of adipose tissue stained for CD68 (original magnifications 20 X); (D) C57 veh; (E) C57 OPG. Data is expressed as mean ± SEM. *p < 0.05 vs C57 veh; àp < 0.001 vs C57 veh. IL-6, interleukin-6; MCP-1, monocyte chemotactic protein-1; OPG, osteoprotegerin; TNF-a, tumor necrosis factor-a; Veh, vehicle.

for the study of MetS. The possibility of an existing causal link between circulating OPG and MetS is suggested by our finding that OPG administration led to metabolic and adipose tissue proinflammatory changes that resemble those observed in HFD fed mice. Moreover, here we demonstrate that OPG delivery is associated with macrophage accumulation and increased gene expression of inflammation markers in the adipose tissue. Although initially it seemed that there was no correlation between OPG and MetS (Nabipour et al., 2010), this correlation was then found in women with history of gestational diabetes (Akinci et al., 2011), which would be consistent with our study, showing that OPG is higher in patients with MetS. In addition, in our study there was also a positive and independent association between OPG, HDL cholesterol, and CRP, which has already been described by Gannagé-Yared in two separate works (GannageYared et al., 2006; Gannage-Yared et al., 2008). Nevertheless, we did not find any correlation between OPG and the remaining components of MetS, where also the literature is conflicting (GannagéYared et al., 2006; Gannage-Yared et al., 2008; Dallmeier et al., 2012; Nabipour et al., 2010; Pepene et al., 2011; Ashley et al., 2011). This heterogeneity can be partly ascribed to the fact that the majority of the studies investigating the relationship between OPG and MetS have focussed on too much specific cohorts of patients differing in terms of sex, age, and associated diseases. On the other hand, another reason explaining the inconsistencies between the studies is that there are several ways to diagnose MetS. While some authors have in fact used the well-known ATPIII criteria (Carr et al., 2004), others have followed the IDF definition (Alberti et al., 2006). Here we selected the patients according to the IDF criteria, where central obesity should always be present, consistent with the concept that obesity is an early step in the etiological cascade leading to all the metabolic disturbances that characterize this syndrome. In keeping with our clinical data, circulating and tissue OPG was significantly higher in HFD fed mice, which represent an experimental model for the study of MetS (Fraulob et al., 2010;

Gallou-Kabani et al., 2007). High-fat diets are in fact one of the main reasons accounting for the incidence of MetS in the world (Buettner et al., 2007). Therefore, given that this syndrome is largely dietary in origin, high-fat diets have been extensively used to cause an experimental condition that closely resembles human MetS, which is the obesity-induced insulin resistance in rodents (Rossmeisl et al., 2003; Surwit et al., 1995). As expected from previous studies (Bernardi et al., 2012b), HFD fed C57BL6 mice gained body weight, increased fat storage, and developed glucose intolerance with insulin resistance. In our study, glucose clearance was significantly delayed in HFD fed mice, given that glucose levels remained elevated 120 min after glucose administration. This was associated and likely to be due to the insulin resistance induced by the HFD. It has been suggested that obesity-related insulin resistance is at least in part a chronic inflammatory disease initiated in the WAT (Xu et al., 2003), given that this condition ameliorates following adipose tissue removal (Pitombo et al., 2006). In particular, one of the primary pathological processes causing obesity-related insulin resistance and MetS development would be the secretion of proinflammatory adipokines by the WAT. Hotamisligil and coworkers have in fact shown that in the plasma and WAT of obese rodents there is an elevation of TNF-a, whose blockade ameliorates insulin sensitivity (Hotamisligil et al., 1993). Likewise, HFD fed mice display increased levels of IL-6, which would cause insulin resistance (Sabio et al., 2008). Experimental evidences would suggest that obesity is associated with macrophage accumulation in the WAT (Weisberg et al., 2003), and that such macrophage infiltration is responsible for the secretion of proinflammatory mediators that are implicated in the development of obesity-related insulin resistance (Xu et al., 2003). Consistent with this view, HFD fed mice lacking the C–C motif chemokine receptor-2, which binds to MCP-1, exhibited fewer macrophages, a lower inflammatory gene profile in WAT, and reduced insulin resistance as compared to the HFD fed wild-type (Weisberg et al., 2006). Conversely, mice overexpressing MCP-1 in their WAT showed the

S. Bernardi et al. / Molecular and Cellular Endocrinology 394 (2014) 13–20

opposite phenotype (Kanda et al., 2006). In keeping with the concept that obesity-induced insulin resistance is caused by WAT inflammation, in this work we found that HFD fed mice, which became obese and insulin resistant, developed systemic and adipose tissue proinflammatory changes. The HFD fed mice presented higher levels of circulating proinfammatory cytokines that were mirrored by similar changes in WAT, where the gene expression of IL-6, MCP-1, and TNF-a was upregulated. It remains unclear whether circulating OPG is causally linked to the pathology of MetS. In different settings, such as CVD and DM, experimental evidence would suggest that OPG is not simply a marker of risk but rather a risk factor for developing CVD and DM. With respect to CVD, circulating OPG has been found to correlate with the risk of developing silent myocardial ischemia, acute myocardial infarction, left ventricular dysfunction, as well as with an increased risk of cardiovascular mortality in different cohorts of patients (Venuraju et al., 2010; Bernardi et al., 2012a). Further works have successively shed light on a possible causative role of OPG in CVD development and progression, given that it promotes leukocyte adhesion to endothelial cells (Mangan et al., 2007; Zauli et al., 2007), stimulates vascular smooth muscle cell proliferation, increases atherosclerotic plaque area (Candido et al., 2010), and induces fibrogenesis (Toffoli et al., 2011b). On the other hand, circulating OPG levels have also been found to correlate with T1DM (Rasmussen et al., 2006) and T2DM (Browner et al., 2001; Secchiero et al., 2006; Toffoli et al., 2011a), especially in those patients with poor glycemic control and complicated disease course (Augoulea et al., 2013). As above, experimental evidence would suggest that OPG plays a causative role in DM. For instance, we have recently shown that OPG administration promotes significant structural changes in pancreatic islets, including selective loss of b-cells by apoptosis, increased infiltration of monocytes/ macrophages, and fibrosis (Toffoli et al., 2011a,b). Similarly to what has been reported for CVD and DM, here we not only show that OPG increases in the setting of Mets but that OPG may also contribute to the development and progression of this syndrome. The second important finding of this study is in fact that the administration of OPG led to similar metabolic and proinflammatory changes to those observed in HFD fed mice. In our study, circulating human OPG was found higher, although not significantly, in the mice treated with it. This is consistent with the literature (Candido et al., 2010), it may be ascribed to the strategy of delivering OPG intraperitoneally, and it may be explained by OPG rapid adhesion to the vascular wall/other tissues. Therefore, it is conceivable that the levels of circulating OPG do not reflect the real amount of OPG in the tissues (An et al., 2007). However, murine OPG significantly increased in the mice treated with human OPG, possibly due to a stimulatory loop (Toffoli et al., 2011b). Having said that, OPG delivery increased glucose levels as well as circulating and tissue proinflammatory molecules, such as IL-6, MCP-1 and TNF-a, leading to an infiltration of macrophages in the WAT. Although high OPG levels have already been linked to a proinflammatory state characteristic of insulin resistance (Harith et al., 2013), this is the first report, to our knowledge, showing that OPG delivery is associated with increased macrophage accumulation and inflammatory marker expression in the WAT. Consequently, it is possible to speculate that beyond pancreatic damage (Toffoli et al., 2011a) OPG affects glucose homeostasis by promoting insulin resistance through fat inflammation. It has been argued that OPG proinflammatory effects do not only depend on TRAIL and RANKL blockade but also on OPG ability to interact with cell surface (Toffoli et al., 2011a,b). As for TRAIL blockade, we have recently shown that recombinant TRAIL delivery significantly attenuates metabolic abnormalities in HFD fed mice by reducing adiposity and systemic inflammation (Bernardi et al., 2012b). Therefore it is possible to speculate that OPG might pro-

19

mote proinflammatory changes by inhibiting TRAIL actions. On the other hand, OPG can act directly through its heparin-binding domain (Theoleyre et al., 2006), whereby it stimulates leukocyte adhesion to endothelial cells (Mangan et al., 2007; Zauli et al., 2007), and promote macrophage recruitment to the adipose tissue. 5. Conclusions This work shows that circulating OPG increases in the setting of MetS. Circulating OPG is in fact higher in patients with MetS, where it is significantly associated with this syndrome and also positively and independently associated with HDL cholesterol and CRP. In HFD fed rodents, which represent an experimental model for the study of MetS, OPG does not only increase in the circulation but also at a tissue level, where its gene expression is significantly upregulated. To understand the significance of this finding, we studied the effects of OPG delivery and found that OPG administration significantly increases glucose levels and enhances systemic and tissue proinflammatory changes. In particular, OPG delivery is associated with macrophage accumulation and increased inflammatory marker gene expression in the WAT. Therefore, by damaging the pancreas (Toffoli et al., 2011a) as well as turning fat mass into an inflamed mass, OPG seems to exert prodiabetogenic effects and it should be considered a risk factor for the development of metabolic abnormalities rather than simply a marker of them. In conclusion, our current demonstration that OPG stimulates the proinflammatory changes classically seen in the setting of obesity-related insulin-resistance suggests that OPG may be one of the missing links between a western diet, obesity, metabolic syndrome and insulin-resistance. 6. Conflict of Interest All authors declare no conflict of interests, including any financial, personal or other relationships with other people or organizations within three years of beginning the submitted work that could inappropriately influence, or be perceived to influence their work. Acknowledgements S. Bernardi received a scholarship from the SIIA (Societa’ Italiana per lo studio dell’Ipertensione Arteriosa). References Akinci, B., Celtik, A., Yuksel, F., Genc, S., Yener, S., Secil, M., Ozcan, M.A., Yesil, S., 2011. Increased osteoprotegerin levels in women with previous gestational diabetes developing metabolic syndrome. Diabetes Res. Clin. Pract. 91, 26–31. Alberti, K.G., Zimmet, P., Shaw, J., 2006. Metabolic syndrome – a new world-wide definition. A Consensus statement from the international diabetes federation. Diabet. Med. 23, 469–480. An, J.J., Han, D.H., Kim, D.M., Kim, S.H., Rhee, Y., Lee, E.J., Lim, S.K., 2007. Expression and regulation of osteoprotegerin in adipose tissue. Yonsei Med. J. 48, 765–772. Ashley, D.T., O’Sullivan, E.P., Davenport, C., Devlin, N., Crowley, R.K., McCaffrey, N., Moyna, N.M., Smith, D., O’Gorman, D.J., 2011. Similar to adiponectin, serum levels of osteoprotegerin are associated with obesity in healthy subjects. Metabolism 60, 994–1000. Augoulea, A., Vrachnis, N., Lambrinoudaki, I., Dafopoulos, K., Iliodromiti, Z., Daniilidis, A., Varras, M., Alexandrou, A., Deligeoroglou, E., Creatsas, G., 2013. Osteoprotegerin as a marker of atherosclerosis in diabetic patients. Int. J. Endocrinol. 2013, 182060. Bastard, J.P., Jardel, C., Bruckert, E., Blondy, P., Capeau, J., Laville, M., Vidal, H., Hainque, B., 2000. Elevated levels of interleukin 6 are reduced in serum and subcutaneous adipose tissue of obese women after weight loss. J. Clin. Endocrinol. Metab. 85, 3338–3342. Bernardi, S., Milani, D., Fabris, B., Secchiero, P., Zauli, G., 2012a. TRAIL as biomarker and potential therapeutic tool for cardiovascular diseases. Curr. Drug Targets 13, 1215–1221. Bernardi, S., Zauli, G., Tikellis, C., Candido, R., Fabris, B., Secchiero, P., Cooper, M.E., Thomas, M.C., 2012b. TNF-related apoptosis-inducing ligand significantly attenuates metabolic abnormalities in high-fat-fed mice reducing adiposity and systemic inflammation. Clin .Sci. (Lond.) 123, 547–555.

20

S. Bernardi et al. / Molecular and Cellular Endocrinology 394 (2014) 13–20

Boyle, W.J., Simonet, W.S., Lacey, D.L., 2003. Osteoclast differentiation and activation. Nature 423, 337–342. Browner, W.S., Lui, L.Y., Cummings, S.R., 2001. Associations of serum osteoprotegerin levels with diabetes, stroke, bone density, fractures, and mortality in elderly women. J. Clin. Endocrinol. Metab. 86, 631–637. Buettner, R., Scholmerich, J., Bollheimer, L.C., 2007. High-fat diets: modeling the metabolic disorders of human obesity in rodents. Obesity (Silver Spring) 15, 798–808. Candido, R., Toffoli, B., Corallini, F., Bernardi, S., Zella, D., Voltan, R., Grill, V., Celeghini, C., Fabris, B., 2010. Human full-length osteoprotegerin induces the proliferation of rodent vascular smooth muscle cells both in vitro and in vivo. J. Vasc. Res. 47, 252–261. Carr, D.B., Utzschneider, K.M., Hull, R.L., Kodama, K., Retzlaff, B.M., Brunzell, J.D., Shofer, J.B., Fish, B.E., Knopp, R.H., Kahn, S.E., 2004. Intra-abdominal fat is a major determinant of the National Cholesterol Education Program Adult Treatment Panel III criteria for the metabolic syndrome. Diabetes 53, 2087– 2094. Collins, G.S., Mallett, S., Omar, O., Yu, L.M., 2011. Developing risk prediction models for type 2 diabetes: a systematic review of methodology and reporting. BMC Med. 9, 103. Dallmeier, D., Larson, M.G., Vasan, R.S., Keaney Jr., J.F., Fontes, J.D., Meigs, J.B., Fox, C.S., Benjamin, E.J., 2012. Metabolic syndrome and inflammatory biomarkers: a community-based cross-sectional study at the Framingham Heart Study. Diabetol. Metab. Syndr. 4, 28. Di Gregorio, G.B., Yao-Borengasser, A., Rasouli, N., Varma, V., Lu, T., Miles, L.M., Ranganathan, G., Peterson, C.A., McGehee, R.E., Kern, P.A., 2005. Expression of CD68 and macrophage chemoattractant protein-1 genes in human adipose and muscle tissues: association with cytokine expression, insulin resistance, and reduction by pioglitazone. Diabetes 54, 2305–2313. Esposito, K., Pontillo, A., Di Palo, C., Giugliano, G., Masella, M., Marfella, R., Giugliano, D., 2003. Effect of weight loss and lifestyle changes on vascular inflammatory markers in obese women: a randomized trial. JAMA 289, 1799–1804. Fraulob, J.C., Ogg-Diamantino, R., Fernandes-Santos, C., Aguila, M.B., Mandarim-deLacerda, C.A., 2010. A mouse model of metabolic syndrome: insulin resistance, fatty liver and non-alcoholic fatty pancreas disease (NAFPD) in C57BL/6 mice fed a high fat diet. J. Clin. Biochem. Nutr. 46, 212–223. Gallou-Kabani, C., Vige, A., Gross, M.S., Rabes, J.P., Boileau, C., Larue-Achagiotis, C., Tome, D., Jais, J.P., Junien, C., 2007. C57BL/6J and A/J mice fed a high-fat diet delineate components of metabolic syndrome. Obesity (Silver Spring) 15, 1996–2005. Gannage-Yared, M.H., Fares, F., Semaan, M., Khalife, S., Jambart, S., 2006. Circulating osteoprotegerin is correlated with lipid profile, insulin sensitivity, adiponectin and sex steroids in an ageing male population. Clin. Endocrinol. (Oxf.) 64, 652–658. Gannage-Yared, M.H., Yaghi, C., Habre, B., Khalife, S., Noun, R., Germanos-Haddad, M., Trak-Smayra, V., 2008. Osteoprotegerin in relation to body weight, lipid parameters insulin sensitivity, adipocytokines, and C-reactive protein in obese and non-obese young individuals: results from both cross-sectional and interventional study. Eur. J. Endocrinol. 158, 353–359. Guilherme, A., Virbasius, J.V., Puri, V., Czech, M.P., 2008. Adipocyte dysfunctions linking obesity to insulin resistance and type 2 diabetes. Nat. Rev. Mol. Cell Biol. 9, 367–377. Harith, H.H., Morris, M.J., Kavurma, M.M., 2013. On the TRAIL of obesity and diabetes. Trends Endocrinol. Metab. 24, 578–587. Hofbauer, L.C., Schoppet, M., 2004. Clinical implications of the osteoprotegerin/ RANKL/RANK system for bone and vascular diseases. JAMA 292, 490–495. Hotamisligil, G.S., Shargill, N.S., Spiegelman, B.M., 1993. Adipose expression of tumor necrosis factor-alpha: direct role in obesity-linked insulin resistance. Science 259, 87–91. Kahn, R., Buse, J., Ferrannini, E., Stern, M., American Diabetes Association and European Association for the Study of Diabetes, 2005. The Metabolic Syndrome: Time for a Critical Appraisal: Joint Statement from the American Diabetes Association and the European Association for the Study of Diabetes, Diabetes Care, vol. 28, pp. 2289–2304. Kanda, H., Tateya, S., Tamori, Y., Kotani, K., Hiasa, K., Kitazawa, R., Kitazawa, S., Miyachi, H., Maeda, S., Egashira, K., Kasuga, M., 2006. MCP-1 contributes to macrophage infiltration into adipose tissue, insulin resistance, and hepatic steatosis in obesity. J. Clin. Invest. 116, 1494–1505. Kiechl, S., Wittmann, J., Giaccari, A., Knoflach, M., Willeit, P., Bozec, A., Moschen, A.R., Muscogiuri, G., Sorice, G.P., Kireva, T., Summerer, M., Wirtz, S., Luther, J., Mielenz, D., Billmeier, U., Egger, G., Mayr, A., Oberhollenzer, F., Kronenberg, F., Orthofer, M., Penninger, J.M., Meigs, J.B., Bonora, E., Tilg, H., Willeit, J., Schett, G., 2013. Blockade of receptor activator of nuclear factor-kappaB (RANKL) signaling improves hepatic insulin resistance and prevents development of diabetes mellitus. Nat. Med. 19, 358–363. Malyankar, U.M., Scatena, M., Suchland, K.L., Yun, T.J., Clark, E.A., Giachelli, C.M., 2000. Osteoprotegerin is an alpha vbeta 3-induced, NF-kappa B-dependent survival factor for endothelial cells. J. Biol. Chem. 275, 20959–20962. Mangan, S.H., Van Campenhout, A., Rush, C., Golledge, J., 2007. Osteoprotegerin upregulates endothelial cell adhesion molecule response to tumor necrosis factor-alpha associated with induction of angiopoietin-2. Cardiovasc. Res. 76 (3), 494–505. Nabipour, I., Kalantarhormozi, M., Larijani, B., Assadi, M., Sanjdideh, Z., 2010. Osteoprotegerin in relation to type 2 diabetes mellitus and the metabolic syndrome in postmenopausal women. Metabolism 59, 742–747.

Neels, J.G., Olefsky, J.M., 2006. Inflamed fat: what starts the fire? J. Clin. Invest. 116, 33–35. Pepene, C.E., Ilie, I.R., Marian, I., Duncea, I., 2011. Circulating osteoprotegerin and soluble receptor activator of nuclear factor kappaB ligand in polycystic ovary syndrome: relationships to insulin resistance and endothelial dysfunction. Eur. J. Endocrinol. 164, 61–68. Pitombo, C., Araujo, E.P., De Souza, C.T., Pareja, J.C., Geloneze, B., Velloso, L.A., 2006. Amelioration of diet-induced diabetes mellitus by removal of visceral fat. J. Endocrinol. 191, 699–706. Rasmussen, L.M., Tarnow, L., Hansen, T.K., Parving, H.H., Flyvbjerg, A., 2006. Plasma osteoprotegerin levels are associated with glycaemic status, systolic blood pressure, kidney function and cardiovascular morbidity in type 1 diabetic patients. Eur. J. Endocrinol. 154, 75–81. Reaven, G.M., 1988. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 37, 1595–1607. Rossmeisl, M., Rim, J.S., Koza, R.A., Kozak, L.P., 2003. Variation in type 2 diabetesrelated traits in mouse strains susceptible to diet-induced obesity. Diabetes 52, 1958–1966. Ryan, A.S., Nicklas, B.J., 2004. Reductions in plasma cytokine levels with weight loss improve insulin sensitivity in overweight and obese postmenopausal women. Diabetes Care 27, 1699–1705. Sabio, G., Das, M., Mora, A., Zhang, Z., Jun, J.Y., Ko, H.J., Barrett, T., Kim, J.K., Davis, R.J., 2008. A stress signaling pathway in adipose tissue regulates hepatic insulin resistance. Science 322, 1539–1543. Schrader, J., Rennekamp, W., Niebergall, U., Schoppet, M., Jahr, H., Brendel, M.D., Horsch, D., Hofbauer, L.C., 2007. Cytokine-induced osteoprotegerin expression protects pancreatic beta cells through p38 mitogen-activated protein kinase signalling against cell death. Diabetologia 50, 1243–1247. Secchiero, P., Corallini, F., Pandolfi, A., Consoli, A., Candido, R., Fabris, B., Celeghini, C., Capitani, S., Zauli, G., 2006a. An increased osteoprotegerin serum release characterizes the early onset of diabetes mellitus and may contribute to endothelial cell dysfunction. Am. J. Pathol. 169, 2236–2244. Simonet, W.S., Lacey, D.L., Dunstan, C.R., Kelley, M., Chang, M.S., Luthy, R., Nguyen, H.Q., Wooden, S., Bennett, L., Boone, T., Shimamoto, G., DeRose, M., Elliott, R., Colombero, A., Tan, H.L., Trail, G., Sullivan, J., Davy, E., Bucay, N., Renshaw-Gegg, L., Hughes, T.M., Hill, D., Pattison, W., Campbell, P., Sander, S., Van, G., Tarpley, J., Derby, P., Lee, R., Boyle, W.J., 1997. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 89, 309–319. Surwit, R.S., Feinglos, M.N., Rodin, J., Sutherland, A., Petro, A.E., Opara, E.C., Kuhn, C.M., Rebuffe-Scrive, M., 1995. Differential effects of fat and sucrose on the development of obesity and diabetes in C57BL/6J and A/J mice. Metabolism 44, 645–651. Theoleyre, S., Kwan Tat, S., Vusio, P., Blanchard, F., Gallagher, J., Ricard-Blum, S., Fortun, Y., Padrines, M., Redini, F., Heymann, D., 2006. Characterization of osteoprotegerin binding to glycosaminoglycans by surface plasmon resonance: role in the interactions with receptor activator of nuclear factor kappaB ligand (RANKL) and RANK. Biochem. Biophys. Res. Commun. 347, 460–467. Toffoli, B., Bernardi, S., Candido, R., Sabato, N., Carretta, R., Corallini, F., Secchiero, P., Zauli, G., Fabris, B., 2011a. Osteoprotegerin induces morphological and functional alterations in mouse pancreatic islets. Mol. Cell. Endocrinol. 331, 136–142. Toffoli, B., Pickering, R.J., Tsorotes, D., Wang, B., Bernardi, S., Kantharidis, P., Fabris, B., Zauli, G., Secchiero, P., Thomas, M.C., 2011b. Osteoprotegerin promotes vascular fibrosis via a TGF-beta1 autocrine loop. Atherosclerosis 218, 61–68. Tuomilehto, J., Lindstrom, J., Eriksson, J.G., Valle, T.T., Hamalainen, H., IlanneParikka, P., Keinanen-Kiukaanniemi, S., Laakso, M., Louheranta, A., Rastas, M., Salminen, V., Uusitupa, M., Finnish Diabetes Prevention Study Group, 2001. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. New Engl. J. Med. 344, 1343–1350. Venuraju, S.M., Yerramasu, A., Corder, R., Lahiri, A., 2010. Osteoprotegerin as a predictor of coronary artery disease and cardiovascular mortality and morbidity. J. Am. Coll. Cardiol. 55, 2049–2061. Weisberg, S.P., McCann, D., Desai, M., Rosenbaum, M., Leibel, R.L., Ferrante Jr., A.W., 2003. Obesity is associated with macrophage accumulation in adipose tissue. J. Clin. Invest. 112, 1796–1808. Weisberg, S.P., Hunter, D., Huber, R., Lemieux, J., Slaymaker, S., Vaddi, K., Charo, I., Leibel, R.L., Ferrante Jr., A.W., 2006. CCR2 modulates inflammatory and metabolic effects of high-fat feeding. J. Clin. Invest. 116 (1), 115–124. Xu, H., Barnes, G.T., Yang, Q., Tan, G., Yang, D., Chou, C.J., Sole, J., Nichols, A., Ross, J.S., Tartaglia, L.A., Chen, H., 2003. Chronic inflammation in fat plays a crucial role in the development of obesity-related insulin resistance. J. Clin. Invest. 112, 1821–1830. Yang, W., Lu, J., Weng, J., Jia, W., Ji, L., Xiao, J., Shan, Z., Liu, J., Tian, H., Ji, Q., Zhu, D., Ge, J., Lin, L., Chen, L., Guo, X., Zhao, Z., Li, Q., Zhou, Z., Shan, G., He, J., China National Diabetes and Metabolic Disorders Study Group, 2010. Prevalence of diabetes among men and women in China. New Engl. J. Med. 362, 1090–1101. Zauli, G., Corallini, F., Bossi, F., Fischetti, F., Durigutto, P., Celeghini, C., Tedesco, F., Secchiero, P., 2007. Osteoprotegerin increases leukocyte adhesion to endothelial cells both in vitro and in vivo. Blood. 110 (2), 536–543. Zauli, G., Melloni, E., Capitani, S., Secchiero, P., 2009. Role of full-length osteoprotegerin in tumor cell biology. Cell. Mol. Life Sci. 66, 841–851. Zhang, J., Fu, M., Myles, D., Zhu, X., Du, J., Cao, X., Chen, Y.E., 2002. PDGF induces osteoprotegerin expression in vascular smooth muscle cells by multiple signal pathways. FEBS Lett. 521, 180–184.