Download Special Issue

Journal of Diabetes Research

Inflammatory Regulation in Diabetes and Metabolic Dysfunction

Guest Editors: Jixin Zhong, Quan Gong, and Akira Mima Inflammatory Regulation in Diabetes and Metabolic Dysfunction Journal of Diabetes Research Inflammatory Regulation in Diabetes and Metabolic Dysfunction

Guest Editors: Jixin Zhong, Quan Gong, and Akira Mima Copyright © 2017 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in “Journal of Diabetes Research.” All articles are open access articles distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Editorial Board

Steven F. Abcouwer, USA Thomas Haak, Germany Ulrike Rothe, Germany Reza Abdi, USA Thomas J. Hawke, Canada Toralph Ruge, Sweden Abdelaziz Amrani, Canada Ole Kristian Hejlesen, Denmark Christoph H. Saely, Austria Fabrizio Barbetti, Italy Dario Iafusco, Italy Ponnusamy Saravanan, UK I. D. Blackberry, Australia K. Kantartzis, Germany Toshiyasu Sasaoka, Japan Simona Bo, Italy Daisuke Koya, Japan Andrea Scaramuzza, Italy Sihem Boudina, USA Frida Leonetti, Italy Yael Segev, Israel Monica Bullo, Spain Afshan Malik, UK Suat Simsek, Netherlands Stefania Camastra, Italy Roberto Mallone, France Marco Songini, Italy Norman Cameron, UK Raffaele Marfella, Italy Janet H. Southerland, USA Ilaria Campesi, Italy C. Martinez Salgado, Spain David Strain, UK Riccardo Candido, Italy Lucy Marzban, Canada Kiyoshi Suzuma, Japan Brunella Capaldo, Italy Raffaella Mastrocola, Italy Giovanni Targher, Italy Danila Capoccia, Italy David Meyre, Canada Patrizio Tatti, Italy Sergiu Catrina, Sweden Maria G. Montez, USA Farook Thameem, USA Subrata Chakrabarti, Canada Stephan Morbach, Germany Michael J. Theodorakis, UK Munmun Chattopadhyay, USA Jiro Nakamura, Japan Peter Thule, USA Eusebio Chiefari, Italy Pratibha V. Nerurkar, USA Andrea Tura, Italy Secundino Cigarran, Spain Monika A. Niewczas, USA R. Varela-Calvino, Spain Kim Connelly, Canada F. Javier Nóvoa, Spain Christian Wadsack, Austria Laurent Crenier, Belgium Craig S. Nunemaker, USA Matthias Weck, Germany Christophe De Block, Belgium Hiroshi Okamoto, Japan Per Westermark, Sweden Devon A. Dobrosielski, USA Ike S. Okosun, USA J. L. Wilkinson-Berka, Australia Khalid M. Elased, USA Fernando Ovalle, USA Dane K. Wukich, USA Ulf J. Eriksson, Sweden Jun Panee, USA Kazuya Yamagata, Japan Paolo Fiorina, USA Cesare Patrone, Sweden Shi Fang Yan, USA Andrea Flex, Italy Subramaniam Pennathur, USA Mark A. Yorek, USA Daniela Foti, Italy Andreas Pfützner, Germany Liping Yu, USA Georgia Fousteri, Italy Bernard Portha, France David Zangen, Israel Maria Pia Francescato, Italy Ed Randell, Canada Thomas J. Zgonis, USA Pedro M. Geraldes, Canada Jordi Lluis Reverter, Spain Dan Ziegler, Germany Margalit D. Goldfracht, Israel Ute Christine Rogner, France Contents

Inflammatory Regulation in Diabetes and Metabolic Dysfunction JixinZhong,QuanGong,andAkiraMima Volume 2017, Article ID 5165268, 2 pages

Relationship between Immunological Abnormalities in Rat Models of Diabetes Mellitus and the Amplification Circuits for Diabetes Yuji Takeda, Tomoko Shimomura, Hironobu Asao, and Ichiro Wakabayashi Volume 2017, Article ID 4275851, 9 pages

Role of T Lymphocytes in Type 2 Diabetes and Diabetes-Associated Inflammation Chang Xia, Xiaoquan Rao, and Jixin Zhong Volume 2017, Article ID 6494795, 6 pages

The Role of HMGB1 in the Pathogenesis of Type 2 Diabetes Yanan Wang, Jixin Zhong, Xiangzhi Zhang, Ziwei Liu, Yuan Yang, Quan Gong, and Boxu Ren Volume 2016, Article ID 2543268, 11 pages

IL-6 Promotes Islet 𝛽-Cell Dysfunction in Rat Collagen-Induced Arthritis Huan Jin, Yaogui Ning, Haotong Zhou, and Youlian Wang Volume 2016, Article ID 7592931, 6 pages

Considerations for Defining Cytokine Dose, Duration, and Milieu That Are Appropriate for Modeling Chronic Low-Grade Inflammation in Type 2 Diabetes Craig S. Nunemaker Volume 2016, Article ID 2846570, 9 pages

Altered Macrophage and Dendritic Cell Response in Mif −/− Mice Reveals a Role of Mif for Inflammatory-Th1 Response in Type 1 Diabetes Yuriko Itzel Sánchez-Zamora, Imelda Juarez-Avelar, Alicia Vazquez-Mendoza, Marcia Hiriart, and Miriam Rodriguez-Sosa Volume 2016, Article ID 7053963, 19 pages

Potential of IL-33 for Preventing the Kidney Injury via Regulating the Lipid Metabolism in Gout Patients Lihua Duan, Yan Huang, Qun Su, Qingyan Lin, Wen Liu, Jiao Luo, Bing Yu, Yan He, Hongyan Qian, Yuan Liu, Jie Chen, and Guixiu Shi Volume 2016, Article ID 1028401, 7 pages

Does Hsp60 Provide a Link between Mitochondrial Stress and Inflammation in Diabetes Mellitus? Joshua Juwono and Ryan D. Martinus Volume 2016, Article ID 8017571, 6 pages

Analysis of Inflammatory Mediators in Prediabetes and Newly Diagnosed Type 2 Diabetes Patients Zhen Wang, Xu-Hui Shen, Wen-Ming Feng, Guo-fen Ye, Wei Qiu, and Bo Li Volume 2016, Article ID 7965317, 10 pages Hindawi Journal of Diabetes Research Volume 2017, Article ID 5165268, 2 pages http://dx.doi.org/10.1155/2017/5165268

Editorial Inflammatory Regulation in Diabetes and Metabolic Dysfunction

Jixin Zhong,1,2 Quan Gong,3 and Akira Mima4

1 Department of Endocrinology, Central Hospital of Wuhan, Wuhan, Hubei 430061, China 2Cardiovascular Research Institute, Case Western Reserve University, Cleveland, OH 44106, USA 3DepartmentofImmunology,SchoolofMedicine,YangtzeUniversity,Jingzhou,Hubei434023,China 4Department of Nephrology, Kindai University Nara Hospital, Faculty of Medicine, Kindai University, Nara 630-0293, Japan

Correspondence should be addressed to Jixin Zhong; [email protected]

Received 5 February 2017; Accepted 8 February 2017; Published 15 March 2017

Copyright © 2017 Jixin Zhong et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Theresearchinthepastdecadeshasrevealedacriticallink immune responses in different diabetes models may help the between metabolic disorders and inflammation, which leads researchers choose an appropriate model in their studies. to a concept called “metaflammation.” Metaflammation is Innate immune response has long been recognized to a form of low-grade systemic and chronic inflammation play an essential role in the development of diabetes and related to excess nutrients and energy [1, 2]. There has been metabolicregulation[2].Y.I.Sanchez-Zamora´ et al. reported increasing evidence showing diabetes is an inflammatory dis- a novel role for macrophage migration inhibitory factor (Mif) ease. Type 1 diabetes (T1DM), characterized by autoimmune- in regulating macrophage/dendritic cell response and T1DM. mediated destruction of pancreatic 𝛽 cells and insufficient Using a STZ-induced diabetes model and Mif knockout mice, secretion of insulin, has long been considered as an inflam- their study suggested that Mif upregulates the expressions of matory disease. Not until the early 1990s, however, was type MHC-II, costimulatory molecules CD80, CD86, and CD40, 2diabetes(T2DM)linkedtoinflammatoryresponse[3]. Toll-like receptor- (TLR-) 2, and TLR-4 and thus promotes T2DM is manifested by peripheral insulin resistance and the activation of macrophage/dendritic cell response and Th1 aberrant production of insulin, accompanied by chronic low response in T1DM. J. Juwono and R. D. Martinus reviewed grade inflammation in peripheral tissues such as adipose theroleofheatshockprotein60(Hsp60),amitochondrial tissue, liver, and muscle. In the last decades, there has been stress protein, in the pathogenesis of T1DM and T2DM. Y. growing evidence linking obesity and insulin resistance to Wang et al. summarized the role of HMGB1, an endogenous inflammation [2–4]. Given the significant roles inflammation danger signal molecule in the development of T2DM. In plays in its pathogenesis, T2DM is now being redefined as recent years, the role of adaptive immunity in T2DM has an immune disorder [5–7]. In addition to diabetes, many also been emphasized [4, 8–10]. In this issue, C. Xia et al. other metabolic disorders have also been associated with reviewed the recent progress in understanding the role of inflammation [2]. This special issue showcases a number T lymphocyte-mediated adaptive immune response in the original research articles and review papers on the topic of inflammatory regulation of T2DM. inflammatory regulation in metabolic dysfunction. Cytokines are important regulators of inflammation in Animal models are useful tools to study pathogenic a number of disease conditions. Z. Wang et al. examined mechanisms including inflammation in diabetes. In this the inflammatory markers in prediabetes by analyzing 215 special issue, Y. Takeda et al. introduced a number of rat patients with prediabetes, 126 cases of newly diagnosed models of diabetes and discussed the immune features in T2DM, and 219 controls with normal glucose tolerance. those diabetic rat models. A better understanding of the Plasma Hs-CRP, IgE, IL-4, IL-10, and tryptase levels were 2 Journal of Diabetes Research identified to increase in prediabetes and early T2DM. C. S. [8]S.Nishimura,I.Manabe,M.Nagasakietal.,“CD8+effectorT Nunemaker addressed the differences in the inflammatory cells contribute to macrophage recruitment and adipose tissue cytokines from the perspective of the pancreatic islet in type inflammation in obesity,” Nature Medicine,vol.15,no.8,pp. 1 versus type 2 diabetes and proposed to consider alternative 914–920, 2009. models of cytokine exposure to reflect the pancreatic environ- [9] S. Winer, Y. Chan, G. Paltser et al., “Normalization of obesity- ment in T2DM more accurately. H. Jin et al. demonstrated associated insulin resistance through immunotherapy,” Nature in an original research article that increased IL-6 level in Medicine,vol.15,no.8,pp.921–929,2009. collagen-induced arthritis was associated with elevated fast- [10] M. Feuerer, L. Herrero, D. Cipolletta et al., “Lean, but not obese, ing blood glucose and reduced fasting insulin level. This may fat is enriched for a unique population of regulatory T cells that be caused by IL-6-induced activation of caspase-3 and loss affect metabolic parameters,” Nature Medicine,vol.15,no.8,pp. of pancreatic islet cell. L. Duan et al. reported an association 930–939, 2009. between serum IL-33 and lipid metabolism dysfunction, a common reason for kidney injury in gout. In addition, IL-33 was higher in gout patients without kidney injury compared to patients with kidney injury, suggesting a potential role of IL-33 in lipid metabolism and kidney injury in gout patients. Collectively, all research papers and review articles in this special issue highlight the current status of research in the area of inflammatory regulation in metabolic disorders. Despitetherapidgrowthinthisarea,westillhavealongway to go to fully understand the inflammatory mechanisms of metabolic disorders.

Acknowledgments We would like to thank all contributors in this special issue for their participation. We believe that the articles in this special issue may help our understanding of the mechanisms of inflammatory regulation in metabolic disorders.

Jixin Zhong Quan Gong Akira Mima

References

[1] M. F. Gregor and G. S. Hotamisligil, “Inflammatory mecha- nisms in obesity,” Annual Review of Immunology,vol.29,pp. 415–445, 2011. [2] G. S. Hotamisligil, “Inflammation and metabolic disorders,” Nature,vol.444,no.7121,pp.860–867,2006. [3] G. S. Hotamisligil, N. S. Shargill, and B. M. Spiegelman, “Adi- pose expression of tumor necrosis factor-𝛼: direct role in obesity-linked insulin resistance,” Science,vol.259,no.5091,pp. 87–91, 1993. [4]D.A.Winer,S.Winer,L.Shenetal.,“Bcellspromoteinsulin resistance through modulation of T cells and production of pathogenic IgG antibodies,” Nature Medicine,vol.17,no.5,pp. 610–617, 2011. [5]S.Tsai,X.Clemente-Casares,X.S.Revelo,S.Winer,andD. A. Winer, “Are obesity-related insulin resistance and type 2 diabetes autoimmune diseases?” Diabetes,vol.64,no.6,pp. 1886–1897, 2015. [6] L. A. Velloso, D. L. Eizirik, and M. Cnop, “Type 2 diabetes mellitus—an autoimmune disease?” Nature Reviews Endocri- nology,vol.9,no.12,pp.750–755,2013. [7] M. Y. Donath and S. E. Shoelson, “Type 2 diabetes as an inflam- matory disease,” Nature Reviews Immunology,vol.11,no.2,pp. 98–107, 2011. Hindawi Journal of Diabetes Research Volume 2017, Article ID 4275851, 9 pages http://dx.doi.org/10.1155/2017/4275851

Review Article Relationship between Immunological Abnormalities in Rat Models of Diabetes Mellitus and the Amplification Circuits for Diabetes

Yuji Takeda,1,2 Tomoko Shimomura,1 Hironobu Asao,2 and Ichiro Wakabayashi1

1 Department of Environmental and Preventive Medicine, Hyogo College of Medicine, Nishinomiya, Japan 2Department of Immunology, Faculty of Medicine, Yamagata University, Yamagata, Japan

Correspondence should be addressed to Yuji Takeda; [email protected]

Received 20 July 2016; Revised 13 December 2016; Accepted 26 January 2017; Published 19 February 2017

Academic Editor: Akira Mima

Copyright © 2017 Yuji Takeda et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

A better understanding of pathogenic mechanisms is required in order to treat diseases. However, the mechanisms of diabetes mellitus and diabetic complications are extremely complex. Immune reactions are involved in the pathogenesis of diabetes and its complications, while diabetes influences immune reactions. Furthermore, both diabetes and immune reactions are influenced by genetic and environmental factors. To address these issues, animal models are useful tools. So far, various animal models of diabetes have been developed in rats, which have advantages over mice models in terms of the larger volume of tissue samples and the variety of type 2 diabetes models. In this review, we introduce rat models of diabetes and summarize the immune reactions in diabetic rat models. Finally, we speculate on the relationship between immune reactions and diabetic episodes. For example, diabetes- prone Biobreeding rats, type 1 diabetes model rats, exhibit increased autoreactive cellular and inflammatory immune reactions, while Goto-Kakizaki rats, type 2 diabetes model rats, exhibit increased Th2 reactions and attenuation of phagocytic activity. Investigation of immunological abnormalities in various diabetic rat models is useful for elucidating complicated mechanisms in the pathophysiology of diabetes. Studying immunological alterations, such as predominance of Th1/17 or Th2 cells, humoral immunity, and innate immune reactions, may improve understanding the structure of amplification circuits for diabetes in future studies.

1. Introduction can be obtained from rats. For example, peripheral blood from rats can be collected in more than 10-fold volume The prevalence of diabetes mellitus (DM) is increasing as compared to blood from mice. For clinical examination, in developed countries, including Japan. Obesity, which is blood specimens from humans usually consist of peripheral caused by abnormalities in lifestyles, such as diet, nutrition, blood, and therefore, the diabetic rat is a valuable experimen- andphysicalactivity,isthemostimportantriskfactorfortype tal model for application to laboratory test. In this review, 2 diabetes mellitus. Obesity, a determinant in the pathogen- we summarize the diverse diabetic rat models and discuss esis of diabetes and diabetic complications, is also associated the relationship between diabetic episodes and immune with various immune reactions in patients with diabetes. An reactions. accumulation of information on immune reactions associ- ated with diabetic episodes is required in order to achieve a 2. Drug-Induced DM better understanding of its pathogenic mechanisms. Animal models of DM are useful for investigating the relationship Streptozotocin (STZ) and alloxan are often used for the between immune reactions and diabetic episodes. induction of type 1 DM in animal models. These reagents In recent years, various type 2 DM models have been cause the production of reactive oxygen species in the 𝛽 cells developedinrats.Ratmodelshavetheadvantageovermice of the pancreas, resulting in 𝛽 cell death [1, 2]. Although these modelsthatahighervolumeofbloodandtissuesamples models of type 1 diabetes are useful for investigating the effect 2 Journal of Diabetes Research of high glucose conditions on immune responses without and glycolysis increase rapidly after antigen rechallenge and obesity, they are not suitable for investigation of autoimmune facilitate their recall responses. On the other hand, chronic reactivity to 𝛽 cells, which is known to be a major mechanism hyperglycemic conditions result in reduction of cellular ATP in the causation of type 1 diabetes in humans. Furthermore, levels by O-GlcNAcylation of protein subunits of mito- Muller et al. clearly demonstrated that STZ both directly and chondrial respiratory complexes I, III, and IV, intimately indirectly induces suppressive effect on lymphocytes in mice associated with normative cellular bioenergetics and ATP [3]. The evaluation of immune responses in the drug-induced production [10]. Therefore, hyperglycemic conditions induce DM models must be carefully considered. the dysfunction of mitochondrial respiration. These reports Alloxan also has a suppressive effect on lymphocytes [4]. allow us to speculate that hyperglycemic condition with However, Gaulton et al. reported that the effect of alloxan on mitochondrial dysfunction may inhibit the primary T cell lymphocytes is transient and concluded that alloxan-induced expansion and the memory T cell maintenance in the diabetic diabetes, but not STZ-induced diabetes, provides a useful model rats. model for evaluating immunological changes associated with STZ doses of 65∼70 mg/kg are often used for inducing hyperglycemia and diabetes [4]. Therefore, we focus on the type 1 DM in rats. In some studies, lower doses of STZ, alloxan-induced diabetes model in relation to immunity in 30∼35 mg/kg, were employed to create a type 2 DM model this review. [11, 12]. Obese and hyperglycemic rats, such as high-fat diet- The reason why only alloxan shows transient effect on fed Sprague-Dawley (SD) rats and Zucker fatty rats, were lymphocyteremainsunclear.Themostlikelyreasonisthat treated with low doses of STZ in order to reduce insulin the source of generated reactive oxygen species is different secretion. The exhaustion of 𝛽 cells was artificially induced in in the pharmacological action of alloxan and STZ. Alloxan these type 2 diabetes models. To our knowledge, there have is more unstable than STZ and is rapidly taken up into been no reports on immunological observations using these insulin-producing cells (𝛽 cells) and liver. The highly reac- low dose-STZ-induced diabetes models. tive hydroxyl radicals are formed by the Fenton reaction. However, the liver and other tissues are more resistant to 3. Type 1 DM Models reactive oxygen species in comparison to pancreatic 𝛽 cells [1]. On the other hand, STZ enters into the cells via glucose Type 1 DM, with a prevalence that is <5% of the prevalence transporter and then causes alkylation of DNA. DNA damage of total DM cases, is considered an autoimmune disease. induces activation of poly-ADP-ribosylation, which leads However, type 1 DM is known to occur through various + to depletion of cellular NAD and ATP. Enhanced ATP pathological processes. For example, fulminant type 1 DM, dephosphorylation supplies a substrate for xanthine oxidase, which represents 20% of acute-onset type 1 DM, is not related resulting in the formation of superoxide radicals [1]. It is to autoreactive immune responses [13]. Other pathological well known that T lymphocytes (T cells) also express the processes, such as slowly progressive insulin-dependent (type glucose transporter during proliferation [5] and that the cells 1) diabetes mellitus (SPIDDM), or latent autoimmune dia- are always accompanied with poly-ADP-ribosylation dur- betes in adults (LADA), which result from gradual distraction ing proliferation [6, 7]. Therefore, inappropriate poly-ADP- of islets of Langerhans in pancreas through an autoimmune ribosylation induced by STZ may be involved in irreversible response, are difficult to distinguish from type 2 DM [14, 15]. impairment of T cell function. Thus, the pathological processes of type 1 DM are variable in In alloxan-induced diabetic rats, both cellular immunity humans. and humoral immunity are impaired by hyperglycemia [4, 8]. Diabetes-prone Biobreeding (DP-BB) rats are a known Phagocytosis by alveolar macrophages from alloxan-DM rats type 1 DM-rat model. Th1 and Th17 responses are domi- is also impaired [9]. These findings are summarized in Table 1. nant, and moreover, the T cell receptor (TCR) repertoire Impairment of both cellular and humoral immunity in is decreased in DP-BB rats due to T-lymphocytopenia [16]. DM was induced by dysfunction of lymphocyte proliferation, These abnormalities trigger the onset of DM, as summarized which is suppressed under high glucose conditions [8]. in Table 1. u Impairment of phagocytosis is also induced by deficient RT 1 haplotype and a mutation in the Gimap5 gene coupling of the leukotriene receptor with the Fc𝛾 receptor areknowntobethegeneticfactorsinvolvedintheonsetof signaling cascade [9]. Stimulation with insulin to augment diabetes in DP-BB rats [17, 18]. Gimap5 plays a key role in cell membrane kinetics plays a key role in this coupling antiapoptosis and calcium signaling in T cells. Dysfunction of [9]. These previous reports make it possible to propose that Gimap5 induces lymphopenia and inhibits the accumulation acquired immunity is regulated by glucose concentration and of a normal T cell pool, including regulatory subsets [19, innate immunity is maintained by the secretion of insulin. 20]. Lymphopenia is considered a cause of an autoimmune Recently, accumulating evidences have demonstrated that reaction in DP-BB rats. On the other hand, diabetes resistant- glucose is a key metabolic substrate for T cells as below [5]: BB (DR-BB) rats have a normal Gimap5 gene. When T cells na¨ıveTcellsmatureandexitfromthethymusprimarily are modulated by irradiation, or when Th1 immune reactions relying on oxidize glucose-derived pyruvate in their mito- are activated by virus infection, DR-BB rats also develop type chondria via oxidative phosphorylation (OXPHOS) for their 1DM[21–23]. metabolic needs to generate cellular ATP. In the secondary Recently, the inhibition of costimulatory signaling for lymphoid organs, memory T cells are a quiescent population T cell activation or transplantation of regulatory T cells of the cells that primarily use OXPHOS, but both OXPHOS hasbeenshowntopreventtheonsetofDMusingDP-BB Journal of Diabetes Research 3

Table 1: Summary of immunological abnormalities in rat models of diabetes mellitus.

Cell type (subset) Abnormality Effect Alloxan-induced diabetic rats (type 1 diabetes mellitus) Downregulation of cellular Tcells Reduced T cell proliferation immunity and humoral immunity Monocyte/macrophage Impaired phagocytosis (Susceptibility to infection?) DP-BB rat Tcelllymphopenia CD4 helper T cells Increase of autoreactivity (Reduction of TCR repertoire) Augmentation of cellular Th1 Increase of ratio immunity Th17 Increase of ratio Augmentation of inflammation Decrease in cell number CD8 cytotoxic T cells Increase of autoreactivity (Reduction in TCR repertoire) Zucker fatty (ZF) rats Decrease in cell number Unknown Tcells + + (CD4 and CD8 ) (Impaired response?) Bcells Augmentation of immunoglobulin (IgM, Unknown IgA, and IgG) production (Autoreactive?) Increase in cell number Macrophages, (monocytes, dendritic cells) Augmentation of nitric oxide production Augmentation of inflammatory Augmentation of inflammatory cytokine responses production G-K rats Increase in cell ratio Tcells Unknown Th2 dominant Bcells Decrease in cell ratio Unknown Augmentation of natural IgM production (Autoreactive?) Unknown Monocytes Attenuation of phagocytic activity (Susceptibility to infection) rats [24–26]. In the future, repertoire or autoreactive antigen IL-1𝛽) is augmented in ZF rats [31, 32]. These findings are analysis of autoreactive T cells and regulatory T cells in summarized in Table 1. Overall, in ZF rats, the cellular DP-BB or DR-BB rats will be useful for the investigation component of acquired immunity is impaired, while innate of various pathological mechanisms in type 1 DM, such as immunity is augmented. fulminant type 1 DM, acute-onset type 1 DM, SPIDDM, and A characteristic of OLETF rats is an age-dependent LADA type 1 DM. onsetofdiabetes[33].OLETFratsalsoshowaugmented production of proinflammatory cytokines [34]. However, to the best of our knowledge, there have been no reports on 4. Type 2 DM Model acquired immune responses in OLETF rats. 4.1.Type2DMModelwithObesity. Obesity is a major risk Zucker strain rats were used for establishing Zucker lean factorfortype2DM.Theestablishmentandbreedingof (ZL) rats and Zucker diabetic fatty (ZDF) rats [35]. Insulin obeseratshavebeenperformedforalongtime.Currently, secretion is impaired in ZDF rats. Furthermore, ZDF rats the two main strains of obese model rats available are were used for breeding with Wister rats and WBN/Kob leptin receptor-deficient rats (Zucker fatty, ZF) and rats with rats, which were established to be Wistar fatty rats and a spontaneous lack of cholecystokinin 1 receptor (Otsuka WBN/Kob-Lepr (fa) rats, respectively [36–39]. These newly Long-Evans Tokushima Fatty, OLETF) [27, 28]. established rat strains are used in the investigation of diabetes Immunological investigations of type 2 diabetic rats have complications. been performed extensively using Zucker rat strains. ZF rats exhibit susceptibility to infection, although phagocytosis 4.2. Type 2 DM without Obesity. Several type 2 DM model by neutrophils and macrophages is not impaired [29, 30]. rats exhibit glucose tolerance without obesity. These model + + Furthermore, T-lymphocytopenia (CD4 and CD8 Tcells) rats exhibit mild hyposecretion (impaired secretion) of is observed in ZF rats [31, 32]. On the other hand, the insulin from birth. Thus, energy metabolic efficiency is poor, production of immunoglobulins (IgA, IgG, and IgM), nitric andbodyweightisratherlessthanthatofnormalrats[40]. oxide, and proinflammatory cytokines (such as TNF-𝛼 and These models are termed nonobese type 2 DM model rats. 4 Journal of Diabetes Research

The Goto-Kakizaki (G-K) rat is a typical model rat of were lower than those of ZF rats (300∼400 mg/dL), and the nonobese type 2 DM. G-K rats were established by selection body weights of ZDF rats were lower than those of ZL rats of bred individuals exhibiting reduced glucose tolerance in and Wister rats. Thus, both G-K rats and ZDF rats showed Wistar rats [41, 42]. G-K rats already exhibit mild hyper- hyposecretion of insulin in response to blood glucose at 16 glycemia and impaired insulin secretion in response to a weeks old; however, there are great differences in the age- glucose load at 4 weeks of age. In G-K rats, reduction and dependent changes of body weight, insulin level, and blood dysplasia of 𝛽 cells in the islets of Langerhans are already glucose level between G-K rats and ZDF rats. + observed at embryonic stages [43], and inflammation and The T cell ratios (% of CD3 ) in ZDF rats were not fibrosis of the pancreas progress with age [44]. significantly different from those of ZL or ZF rats. TheT Presently, Rosengren et al. have demonstrated that over- cell ratios of G-K rats were highest of the rat strains tested. expression of alpha2A adrenergic receptor (𝛼(2A)AR) is Interestingly, the reaction to the pathogen (increased ERK the cause of hyposecretion of insulin in G-K rats [45]. activity upon lipopolysaccharide stimulation) of ZDF rats Overexpression of 𝛼(2A)AR on the cell membrane has been was the highest among the rat strains tested [57]. shown to prevent membrane fusion during exocytosis in We analyzed the relationships between CD11b/c and insulin secretion. In humans, a polymorphism of 𝛼(2A)AR phagocytic activity and between CD3 and body weight, gene (ADRA2A) was reportedly associated with an increased in Zucker strains and G-K/Wistar strains of diabetic rat risk for type 2 DM [46]. models (Figure 1). Phagocytic activity is associated with the G-K rats exhibit mild diabetic complications, such as expression level of CD11b/c (Figure 1(a)). The T cell ratio + retinopathy, nephropathy, and peripheral neuropathy [47]. (% CD3 ) in lymphocytes is inversely correlated with body Spontaneously Diabetic Torii (SDT) rats have severe diabetic weight (Figure 1(b)). Therefore, the measurements of CD3 complications without obesity [48–51]. Recently, mating of on lymphocytes and CD11b/c on monocytes are useful for ZF rats with SDT rats has also been carried out: the mated examining the relationship between immunity and diabetic rats exhibited an increased risk of hypertension caused by episodes. For example, a case with CD3 upregulation and DM with obesity [52, 53]. To our knowledge, there have CD11b/c downregulation may be consistent with the patho- been no reports on immunological abnormality in SDT logic status of G-K rats, and conversely, a case with CD3 rats. The comparison between G-K rats and SDT rats is of downregulation and CD11b/c upregulation may be analogous interest for revealing the differences between them in terms to that of ZF rats. Further, a case with upregulation of CD11b/c of susceptibility to diabetic complications. and increased ERK activation corresponds to the pathologic We investigated the immunological abnormalities of G-K status of ZDF rats. rats, which are described in the following section. 5. Measurement of Monocyte 4.3. Immunological Abnormalities of G-K Rats. We investi- Response and Amplification Circuits gated immunological disorders in G-K rats, in comparison with Wistar rats [54–56]. The results are summarized in Monocytes are important in the initiation of various immune Table 1. There was no difference in the white blood cell counts responses [58, 59]. Monocytes migrate into tissues from the between the G-K rats and Wistar rats. However, the T cell peripheral blood and are differentiated into inflammatory ratios in the white blood cells were increased, and B cell ratios or regulatory macrophages depending on the environmental were decreased in G-K rats, as compared with Wistar rats. conditions [60–62]. Resident macrophages in various tissues The distinguishing feature of immunological abnormality in migrate during the fetal period and proliferate in each G-Kratsistheimpairmentofthephagocyticactivityof tissue. However, up to half of the resident macrophages are monocytes. Furthermore, natural (or innate) IgM production derived from peripheral blood monocytes after birth [60]. and Th2 immune reactivity were augmented in G-K rats, as Furthermore, infections or tissue injury augment monocyte compared with Wistar rats. These increases are diametrically migration into the tissue. Particularly, the intestine is known opposedtothechangesobservedinDP-BBrats. as “monocyte fall” [63]. The inflammatory response to enteric To clarify the abnormality of G-K rats, immunological bacteria is regulated by monocyte migration into the intes- parameters of peripheral blood from G-K rats were compared tine, but regulatory (anti-inflammatory) monocytes must to the parameters of the blood from not only the Wister rat also maintain intestinal homeostasis [63]. strain but also the Zucker rat strains, such as Zucker lean It is well known that proinflammatory responses are rats (ZL), Zucker fatty rats (ZF), and Zucker diabetic fatty involved in the acceleration of arteriosclerosis and toler- rats (ZDF) (Table 2). Although the body weight of G-K rats ance to insulin [64, 65]. In DM, high glucose blood is gradually increased depending on their ages (∼400 g), the considered to indicate a mild inflammatory status [66, 67]. insulin levels of G-K rats (1∼3 ng/mL) at the nonfasting status Indeed, the differentiation to M1 macrophages (inflammatory were slightly higher than the levels of Wister rats. The insulin type macrophages) is augmented in adipose tissue or DM level of ZDF peaked at 8 weeks old (15∼25 ng/mL) and then modelmice[68,69].Incontrast,chronicinflammation,such becamelowerafter16weeksold(1∼4 ng/mL), although the as tumor progression, induces immunosuppressive myeloid body weight of ZDF was saturated during 15 to 24 weeks cells [70]. Thus, the physiological or pathological inflamma- old (350∼450 g). After 16 weeks old, ZDF exhibited the most tory status is important for predicting disease progression. severe hyperglycemia (>600 mg/dL) among the Zucker rat Monocyte status is regulated not only in the tissues, strains. The blood glucose levels of G-K rats (200∼300 mg/dL) but also in the peripheral blood [71]. The interaction of Journal of Diabetes Research 5

Table 2: Comparison of various parameters between the Zucker strains and G-K/Wistar strains.

Subject ZL ZF ZDF Wistar G-K ∗ Weight (g) 461.8 ± 31.2 620.8 ± 30.6§ 387.6 ± 6.1 411.0 ± 7.9 340.2 ± 10.4 ∗ Glucose (mg/dL) 163.0 ± 27.8 399.0 ± 63.6§ 636.6 ± 63.5§ 111.2 ± 15.2 227.2 ± 31.8 ∗ CD3 (%) 57.5 ± 3.8 40.6 ± 5.2§ 52.1 ± 1.8 50.1 ± 1.8 67.0 ± 1.4 ∗ CD45RA (%) 28.7 ± 3.1 33.9 ± 6.1 22.4 ± 1.7 32.7 ± 1.7 17.3 ± 1.5 ∗ CD11b/c (%) 90.9 ± 3.8 92.4 ± 1.6 86.4 ± 2.7 67.1 ± 4.3 51.6 ± 4.7 CD11b (%) 85.5 ± 3.8 70.8 ± 10.4 59.6 ± 10.3 67.8 ± 4.5 58.5 ± 4.3 ∗ Phagocytosis (MFI) 49.4 ± 20.6 68.1 ± 19.0 51.5 ± 8.8 25.7 ± 14.1 9.4 ± 0.5 pERK (ratio) 4.0 ± 0.9 3.7 ± 0.4 7.1 ± 0.7§ 3.8 ± 0.4 2.9 ± 0.2 All data were obtained from 16-week-old, male rats. ZL, Zucker lean rats; ZF, Zucker fatty rats; ZDF, Zucker diabetic fatty rats. §𝑝 <0.05(versus ZL rats, by one-way ANOVA, n =4∼5) ∗ 𝑝 <0.05(versus Wistar rats, by Mann–Whitney test, n =8∼11) CD3 (%) and CD45RA (%) indicate the corresponding ratios in lymphocytes. Phagocytosis (MFI) indicates the phagocytic activity of monocytes using FITC-labeled BioParticle. MFI, mean of fluorescence intensity. pERK (ratio) shows the ratio of phosphorylated ERK to total ERK in monocytes after stimulation with lipopolysaccharide for 10 min, as described inext. thet

ZFF 68 635 ZF

585 56 ZDF 535 44 ZL

485 ZL

Phagocytosis 32 Wistar Body weight 435 Wistar 21 385 ZDF G-K G-K 9 335 51 59 67 76 84 92 40 45 50 55 60 65 70 CD11b/c CD3 (a) (b)

Figure 1: Relationships between phagocytosis and CD11b/c and between body weight and CD3 in the G-K/Wistar strains and Zucker strains. (a) Relationship between phagocytic ability and CD11b/c expression levels in G-K/Wistar and Zucker strain rats. (b) Relationship between body weight and CD3 ratio in lymphocytes in G-K/Wistar and Zucker strain rats. Bubble size indicates standard error, and bubble color shows each rat strain as follows: deep blue, ZL; light blue, ZF; green, ZDF; yellow, Wistar; red, G-K. The bubble charts were drawn using Graph-R software, version 2.19.

monocytes with activated platelets induces proinflammatory immunological disorders. Thus, we measured signal trans- responses [72], while the interaction of monocytes with ducers in peripheral blood monocytes. The assay to detect “silent-platelets” accelerates anti-inflammatory responses phosphorylated signal transducers in monocytes was con- [71]. Therefore, monocytes in the peripheral blood monitor ducted using heparinized blood and cell purification and systemic immunological conditions and regulate immune long-term culture were not required. Additionally, the sam- ∘ responses in various tissues. Thus, inflammatory monocytes ples were stored at −20 Cinmethanoluntiltheflowcytom- or anti-inflammatory monocytes may become problematic. etry measurements were obtained. These advances simplify The measurement of cytokine levels in the serum or the procedure [55, 57, 73]. cytokine mRNA in peripheral blood leukocytes is use- Immune systems are under the control of the NF-𝜅B ful method for detection of immunological responses, but and extracellular signal-regulated kinase families [65, 74]. cytokine modulation in DM is much lower than in other Previously, we measured phosphorylated NF-𝜅B-p65 and 6 Journal of Diabetes Research

DP-BB rats Disappearance of insulin

Susceptibility Obesity to infection TCR repertoire ↓ Hyperglycemia Regulatory subset ↓ Hyperlipidemia Acute (Autoimmune reaction ↑) ZF rats Ig production ↑

Proinflammatory cytokines↑

Damage to 𝛽 cells ZDF rats

Unresponsiveness to insulin Slow (chronic) Complication of DM

Low insulin Phagocytosis↓

Susceptibility G-K rats to infection

Figure 2: Pathogenesis of diabetes mellitus in rat models. Dash lines and gray squares indicate the pathogenesis of diabetes in each rat model. DM, diabetes mellitus; Ig, immunoglobulin; TCR, T cell receptor. extracellular signal-regulated kinase 1/2 (ERK) in peripheral receptor 4 (TLR4) [76], and hyperglycemia has been shown monocytes from Zucker rats (Figure 3). Interestingly, the to enhance IL-1𝛽 secretion [77]. Furthermore, increasing basal levels of phosphorylated NF-𝜅B-p65 in monocytes from knowledge of IL-1 family members leads to the clarification ZF and ZDF were lower than that in monocytes from ZL of the association between adipose tissue inflammation and (Figure 3(a)). Furthermore, the early induction of phospho- insulin resistance [75]. IL-1𝛽 also directly contributes to rylated ERK by lipopolysaccharide at 10 min was significantly dysfunction of 𝛽 cells in islets [75]. These findings reveal augmented in monocytes from ZDF compared to those from the pathogenetic cascade, from obesity to DM, mediated by ZL and ZF (Figure 3(b)). According to the introduction inflammation. of the Zucker strains from Charles River Laboratories, the Reduction and dysplasia of 𝛽 cells in the islets are already insulin levels of ZL, ZF, and ZDF at 15∼16 weeks old were observed at embryonic stages [43], based on data from G- 2.4 ± 0.7, 65.4 ± 37.1,and6.3 ± 5.0 (ng/mL), respectively. K rats, suggesting that there is an amplification circuit in On the other hand, the summated blood glucose levels in DM, as summarized in Figure 2. The background of each the glucose tolerance test (GTT) of ZL, ZF, and ZDF at 16 DM-rat model, shown as the circuit schema in Figure 2, is weeks old were 329 ± 15, 524 ± 64,and2916 ± 594 (mg/dL), associated with immune responses. Autoimmune responses respectively. Thus, the insulin response is strongly attenuated and lymphopenia are detected in DP-BB rats, lymphopenia in ZDF. These results suggest that deficiency of the leptin and augmentation of humoral immunity are found in ZF signal reduces basal NF-𝜅B activation and that insulin signal rats, enhancement of the inflammatory response is detected reduction augments ERK activation. in ZDF, and increased humoral immunity and reduction of Collation of the signals to the amplification circuit indi- phagocytic activity are observed in G-K rats. These rat models catedthatthereducedbasallevelofNF-𝜅Bwascorrelated clearly illustrate that immunological phenomena are linked with the reduction in the TCR repertoire and that augmen- with the amplification circuit in DM. tation of the ERK response was involved in the production In the future, the structure of DM-circuit consisting of of proinflammatory cytokines. These abnormalities of signal diabetic episode and immune response will be clarified more transduction in monocytes will be useful for clinical testing in detail using DM-rat models. to evaluate the progression of DM related to immunological disorders. Competing Interests

6. Conclusion The authors declare that there are no competing interests. Obesity is characterized by chronic and low-grade inflam- Authors’ Contributions mation[75].Inrecentyears,freefattyacid-boundfetuin- Ahasbeenshowntobeanendogenousligandfortoll-like All authors read and approved the paper for submission. Journal of Diabetes Research 7

∗∗∗ ∗∗∗ 5 ∗∗∗ 10 ∗∗∗ ∗ ∗∗∗ ∗∗ ∗∗ 4 8 min) 0 min) 10

3 6 B MFI (at 𝜅 NF-

65 2 4

1 MFI (at total-ERK Phospho: 2 Phospho: total-p Phospho:

0 0 ZL ZF ZDF Wistar G-K ZL ZF ZDF Wistar G-K (a) (b)

Figure 3: Difference in phosphorylation of signal transducers (NF-𝜅B and ERK) in monocytes dependent on diabetes status. The phosphorylation levels of NF-𝜅B and ERK in peripheral blood monocytes were measured by flow cytometry. Peripheral blood was collected from Zucker lean (ZL), Zucker fatty (ZF), Zucker diabetic fatty (ZDF), Wister, or G-K and stimulated with lipopolysaccharide (1 𝜇g/mL). ∘ After stimulation for 0 or 10 min, the blood was immediately fixed in Lyse/Fix Buffer (BD Biosciences) and treated with methanol at −20 C for membrane-permeabilization. Next, whole leukocytes in blood were stained with specific antibodies (anti-NF-𝜅B-p60 Abs and anti-ERK Abs from Cell Signaling Technologies).Thecellsweremeasuredbyflowcytometry,andthelevelsoftotalorphosphorylatedsignaltransducers in monocytes were analyzed by monocyte-gating. The ratio of phosphorylation to the total was calculated based on the mean of fluorescence intensity. Changes in the phosphorylation levels of NF-𝜅B or ERK are shown in (a) or (b), respectively. Data are the mean ± SE (𝑛=3). ∗ ∗∗ ∗∗∗ 𝑝 < 0.05; 𝑝 < 0.01; 𝑝 < 0.001 (one-way ANOVA with post hoc test using Bonferroni).

Acknowledgments [7] D. Lautier, J. Lagueux, J. Thibodeau, L. Menard,´ and G. G. Poirier, “Molecular and biochemical features of poly (ADP- This work was supported by a Grant-in-Aid for Scientific ribose) metabolism,” Molecular and Cellular Biochemistry,vol. Research from the Japan Society for the Promotion of Science 122, no. 2, pp. 171–193, 1993. (no. 22590560 and no. 16K16294) and Grant-in-Aid for [8] R.Rubinstein,A.M.Genaro,A.Motta,G.Cremaschi,andM.R. Researchers, Hyogo College of Medicine, 2009. Wald, “Impaired immune responses in streptozotocin-induced type I diabetes in mice. Involvement of high glucose,” Clinical and Experimental Immunology,vol.154,no.2,pp.235–246, References 2008. [1] T. Szkudelski, “The mechanism of alloxan and streptozotocin [9]M.Ferracini,J.O.Martins,M.R.M.Campos,D.B.C.Anger, action in B cells of the rat pancreas,” Physiological Research,vol. and S. Jancar, “Impaired phagocytosis by alveolar macrophages 50, no. 6, pp. 537–546, 2001. from diabetic rats is related to the deficient coupling of LTs to 𝛾 [2] S. Lenzen, “The mechanisms of alloxan- and streptozotocin- the Fc R signaling cascade,” Molecular Immunology,vol.47,no. induced diabetes,” Diabetologia,vol.51,no.2,pp.216–226,2008. 11-12, pp. 1974–1980, 2010. [3]Y.D.Muller,D.Golshayan,D.Ehirchiouetal.,“Immuno- [10] G. B. Stefano, S. Challenger, and R. M. Kream, “Hyperglycemia- suppressive effects of streptozotocin-induced diabetes result in associated alterations in cellular signaling and dysregulated absolute lymphopenia and a relative increase of T regulatory mitochondrial bioenergetics in human metabolic disorders,” cells,” Diabetes,vol.60,no.9,pp.2331–2340,2011. European Journal of Nutrition,vol.55,no.8,pp.2339–2345, [4]G.N.Gaulton,J.L.Schwartz,andD.D.Eardley,“Assessment 2016. of the diabetogenic drugs alloxan and streptozotocin as models [11] K. Srinivasan, B. Viswanad, L. Asrat, C. L. Kaul, and P.Ramarao, for the study of immune defects in diabetic mice,” Diabetologia, “Combination of high-fat diet-fed and low-dose streptozotocin- vol. 28, no. 10, pp. 769–775, 1985. treatedrat:amodelfortype2diabetesandpharmacological [5] M. D. Buck, D. O’Sullivan, and E. L. Pearce, “T cell metabolism screening,” Pharmacological Research,vol.52,no.4,pp.313–320, drives immunity,” Journal of Experimental Medicine,vol.212,no. 2005. 9, pp. 1345–1360, 2015. [12] L. J. Slieker, E. F. Roberts, W.N. Shaw, and W.T. Johnson, “Effect [6] T. Boulikas, B. Bastin, P. Boulikas, and G. Dupuis, “Increase in of streptozocin-induced diabetes on insulin-receptor tyrosine histone poly(ADP-ribosylation) in mitogen-activated lymphoid kinase activity in obese Zucker rats,” Diabetes,vol.39,no.5,pp. cells,” Experimental Cell Research,vol.187,no.1,pp.77–84,1990. 619–625, 1990. 8 Journal of Diabetes Research

[13] A. Imagawa, T. Hanafusa, Y. Uchigata et al., “Fulminant type 1 [29] B. J. Plotkin, D. Paulson, A. Chelich et al., “Immune respon- diabetes: a nationwide survey in Japan,” Diabetes Care,vol.26, siveness in a rat model for type II diabetes (Zucker rat, no. 8, pp. 2345–2352, 2003. fa/fa): susceptibility to Candida albicans infection and leucocyte [14] T. Kobayashi, “Subtype of insulin-dependent diabetes mellitus function,” Journal of Medical Microbiology,vol.44,no.4,pp. (IDDM) in Japan: slowly progressive IDDM—the clinical char- 277–283, 1996. acteristics and pathogenesis of the syndrome,” Diabetes Research [30] B. J. Plotkin and D. Paulson, “Zucker rat (fa/fa), a model for the and Clinical Practice, vol. 24, Supplement, pp. S95–S99, 1994. study of immune function in Type-II diabetes mellitus: effect [15] P. Zimmet, R. Turner, D. McCarty, M. Rowley, and I. Mackay, of exercise and caloric restriction on the phagocytic activity of “Crucial points at diagnosis. Type 2 diabetes or slow type 1 macrophages,” Laboratory Animal Science,vol.46,no.6,pp. diabetes,” Diabetes Care, vol. 22, supplement 2, pp. B59–B64, 682–684, 1996. 1999. [31] S.-I. Tanaka, F. Isoda, T. Yamakawa, M. Ishihara, and H. [16]J.VanDenBrandt,H.J.Fischer,L.Walter,T.Hunig,¨ I. Kloting,¨ Sekihara, “T lymphopenia in genetically obese rats,” Clinical andH.M.Reichardt,“Type1diabetesinbiobreedingratsis Immunology and Immunopathology,vol.86,no.2,pp.219–225, critically linked to an imbalance between Th17 and regulatory 1998. T cells and an altered TCR repertoire,” Journal of Immunology, [32] M.R.Ruth,C.G.Taylor,P.Zahradka,andC.J.Field,“Abnormal vol.185,no.4,pp.2285–2294,2010. immune responses in fa/fa Zucker rats and effects of feeding [17] A.-M. Martin, E. P. Blankenhorn, M. N. Maxson et al., “Non- conjugated linoleic acid,” Obesity,vol.16,no.8,pp.1770–1779, major histocompatibility complex-linked diabetes susceptibil- 2008. ity loci on chromosomes 4 and 13 in a backcross of the DP- [33] T. Hirashima, K. Kawano, S. Mori, and T. Natori, “A diabeto- BB/Wor rat to the WF rat,” Diabetes,vol.48,no.1,pp.50–58, genic gene, ODB2, identified on chromosome 14 of the OLETF 1999. rat and its synergistic action with ODB1,” Biochemical and [18] R. H. Wallis, K. Wang, L. Marandi et al., “Type 1 diabetes in the Biophysical Research Communications,vol.224,no.2,pp.420– BB rat: a polygenic disease,” Diabetes,vol.58,no.4,pp.1007– 425, 1996. 1017, 2009. [34] N. Nagai, T. Murao, Y. Ito, N. Okamoto, and H. Okamura, [19] N. N. Iwakoshi, I. Goldschneider, F. Tausche, J. P. Mordes, A. A. “Involvement of interleukin 18 in lens opacification of otsuka Rossini, and D. L. Greiner, “High frequency apoptosis of recent long-evans tokushima fatty rats, a model of human type 2 thymic emigrants in the liver of lymphopenic diabetes-prone diabetes,” Current Eye Research,vol.36,no.6,pp.497–506,2011. BioBreeding rats,” The Journal of Immunology,vol.160,no.12, [35] J. B. Clark, C. J. Palmer, and W. N. Shaw, “The diabetic zucker pp. 5838–5850, 1998. fatty rat,” Proceedings of the Society for Experimental Biology and [20] G. Hernandez-Hoyos,S.Joseph,N.G.A.Miller,andG.W.´ Medicine,vol.173,no.1,pp.68–75,1983. Butcher, “The lymphopenia mutation of the BB rat causes inap- [36] H. Ikeda, A. Shino, T. Matsuo, H. Iwatsuka, and Z. Suzuoki, propriate apoptosis of mature thymocytes,” European Journal of “A new genetically obese-hyperglycemic rat (Wistar fatty),” Immunology,vol.29,no.6,pp.1832–1841,1999. Diabetes,vol.30,no.12,pp.1045–1050,1981. [21] I. Kloting,¨ S. Berg, P. Kovacs,´ B. Voigt, L. Vogt, and S. Schmidt, [37] N. Kaji, A. Okuno, K. Ohno-Ichiki et al., “Plasma profiles of “Diabetes and hypertension in rodent models,” Annals of the glucose, insulin and lipids in the male WBN/Kob-leprfa rat, a New York Academy of Sciences,vol.827,pp.64–84,1997. new model of type 2 diabetes with obesity,” Journal of Veterinary [22]S.Ramanathan,M.-T.Bihoreau,A.D.Paterson,L.Marandi,D. Medical Science,vol.74,no.9,pp.1185–1189,2012. Gauguier, and P.Poussier, “Thymectomy and radiation-induced [38] H. Matsui, M. Suzuki, R. Tsukuda, K. Iida, M. Miyasaka, and type 1 diabetes in nonlymphopenic BB rats,” Diabetes,vol.51,no. H. Ikeda, “Expression of ICAM-1 on glomeruli is associated 10, pp. 2975–2981, 2002. with progression of diabetic nephropathy in a genetically [23] I. I. Mendez, Y.-H. Chung, H.-S. Jun, and J.-W. Yoon, obese diabetic rat, Wistar fatty,” Diabetes Research and Clinical “Immunoregulatory role of nitric oxide in Kilham rat virus- Practice,vol.32,no.1-2,pp.1–9,1996. induced autoimmune diabetes in DR-BB rats,” The Journal of [39] T. Noguchi, T. Matsuda, Y.Tomari et al., “The regulation of gene Immunology,vol.173,no.2,pp.1327–1335,2004. expression by insulin is differentially impaired in the liver of the [24] K. L. Kover, Z. Geng, D. M. Hess, C. D. Benjamin, and W. V. genetically obese-hyperglycemic Wistar fatty rat,” FEBS Letters, Moore, “Anti-CD154 (CD40L) prevents recurrence of diabetes vol. 328, no. 1-2, pp. 145–148, 1993. inisletisograftsintheDR-BBrat,”Diabetes,vol.49,no.10,pp. [40] C. B. Landersdorfer, D. C. DuBois, R. R. Almon, and W. J. 1666–1670, 2000. Jusko, “Mechanism-based modeling of nutritional and leptin [25] D. Lundsgaard, T. L. Holm, L. Hornum, and H. Markholst, “In influences on growth in normal and type 2 diabetic rats,” Journal + + vivo control of diabetogenic T-cells by regulatory CD4 CD25 of Pharmacology and Experimental Therapeutics,vol.328,no.2, T-cells expressing Foxp3,” Diabetes,vol.54,no.4,pp.1040–1047, pp.644–651,2009. 2005. [41] A. Kitahara, T. Toyota, M. Kakizaki, and Y. Goto, “Activities [26]P.Poussier,T.Ning,T.Murphy,D.Dabrowski,andS. of hepatic enzymes in spontaneous diabetes rats produced by Ramanathan, “Impaired post-thymic development of regula- selective breeding of normal wistar rats,” Tohoku Journal of tory CD4+25 + T cells contributes to diabetes pathogenesis in Experimental Medicine, vol. 126, no. 1, pp. 7–11, 1978. BB rats,” Journal of Immunology,vol.174,no.7,pp.4081–4089, [42]Y.Goto,M.Kakizaki,andN.Masaki,“Productionofsponta- 2005. neous diabetic rats by repetition of selective breeding,” Tohoku [27] G. A. Bray, “The Zucker fatty rat: a review,” Federation Proceed- Journal of Experimental Medicine,vol.119,no.1,pp.85–90,1976. ings,vol.36,no.2,pp.148–153,1977. [43] N. Suzuki, T. Aizawa, N. Asanuma et al., “An early insulin [28] K. Kawano, T. Hirashima, S. Mori, Y. Saitoh, M. Kurosumi, intervention accelerates pancreatic 𝛽-cell dysfunction in young and T. Natori, “Spontaneous long-term hyperglycemic rat with goto-kakizaki rats, a model of naturally occurring noninsulin- diabetic complications: Otsuka Long-Evans Tokushima Fatty dependent diabetes,” Endocrinology,vol.138,no.3,pp.1106– (OLETF) strain,” Diabetes,vol.41,no.11,pp.1422–1428,1992. 1110, 1997. Journal of Diabetes Research 9

[44] F. Homo-Delarche, S. Calderari, J. C. Irminger et al., “Islet [60] E. G. Perdiguero and F. Geissmann, “The development and inflammation and fibrosis in a spontaneous model of type 2 maintenance of resident macrophages,” Nature Immunology, diabetes, the GK rat,” Diabetes,vol.55,no.6,pp.1625–1633, vol.17,no.1,pp.2–8,2016. 2006. [61] Y. Okabe and R. Medzhitov, “Tissue biology perspective on [45] A. H. Rosengren, R. Jokubka, D. Tojjar et al., “Overexpression of macrophages,” Nature Immunology,vol.17,no.1,pp.9–17,2016. Alpha2A-Adrenergic receptors contributes to type 2 diabetes,” [62] C. K. Glass and G. Natoli, “Molecular control of activation and Science,vol.327,no.5962,pp.217–220,2010. priming in macrophages,” Nature Immunology,vol.17,no.1,pp. [46] F. M. Gribble, “𝛼2𝐴-adrenergic receptors and type 2 diabetes,” 26–33, 2016. New England Journal of Medicine,vol.362,no.4,pp.361–362, [63] C. C. Bain, A. Bravo-Blas, C. L. Scott et al., “Constant replen- 2010. ishment from circulating monocytes maintains the macrophage [47]Y.Murakawa,W.Zhang,C.R.Pierson,T.Brismar,C.-G. pool in the intestine of adult mice,” Nature Immunology,vol.15, Ostenson,¨ and A. A. F. Sima, “Impaired glucose tolerance and no.10,pp.929–937,2014. insulinopenia in the GK-rat causes peripheral neuropathy,” [64] F. Rivas, “In this issue: inflammation,” Cell,vol.140,no.6,p.755, Diabetes/Metabolism Research and Reviews,vol.18,no.6,pp. 757, 2010. 473–483, 2002. [65] M. F. Gregor and G. S. Hotamisligil, “Inflammatory mecha- [48] M. Shinohara, T. Masuyama, T. Shoda et al., “A new sponta- nisms in obesity,” Annual Review of Immunology,vol.29,pp. neously diabetic non-obese torii rat strain with severe ocular 415–445, 2011. complications,” Experimental Diabesity Research,vol.1,no.2, [66] M. R. Dasu, S. Devaraj, L. Zhao, D. H. Hwang, and I. Jialal, pp. 89–100, 2000. “High glucose induces toll-like receptor expression in human [49] T. Sasase, “Pathophysiological characteristics of diabetic ocular monocytes mechanism of activation,” Diabetes,vol.57,no.11, complications in spontaneously diabetic torii rat,” Journal of pp. 3090–3098, 2008. Ophthalmology, vol. 2010, Article ID 615641, 7 pages, 2010. [67] M. R. Dasu, S. Devaraj, and I. Jialal, “High glucose induces 𝛽 [50]Y.Matsumoto,K.Torimoto,H.Matsuyoshietal.,“Long-term IL-1 expression in human monocytes: mechanistic insights,” effects of diabetes mellitus on voiding function in a new model American Journal of Physiology - Endocrinology and Metabolism, of type 2 diabetes mellitus, the Spontaneously Diabetic Torii vol. 293, no. 1, pp. E337–E346, 2007. (SDT) rat,” Biomedical Research,vol.30,no.6,pp.331–335,2009. [68] M. Ichioka, T. Suganami, N. Tsuda et al., “Increased expression [51] T. Sasase, H. Morinaga, T. Abe et al., “Protein kinase C of macrophage-inducible C-type lectin in adipose tissue of beta inhibitor prevents diabetic peripheral neuropathy, but obese mice and humans,” Diabetes, vol. 60, no. 3, pp. 819–826, not histopathological abnormalities of retina in Spontaneously 2011. Diabetic Torii rat,” Diabetes,ObesityandMetabolism, vol. 11, no. [69] Y. Pang, Y. Li, Y. Lv et al., “Intermedin restores hyperhomo- 11, pp. 1084–1087, 2009. cysteinemia-induced macrophage polarization and improves insulin resistance in mice,” Journal of Biological Chemistry,vol. [52] T. Masuyama, Y. Katsuda, and M. Shinohara, “A novel model 𝑓𝑎 291, no. 23, pp. 12336–12345, 2016. of obesity-related diabetes: introgression of the Lepr allele of [70] Y. Takeda, T. Kato, H. Ito et al., “The pattern of GPI-80 the zucker fatty rat into nonobese spontaneously diabetic torii expression is a useful marker for unusual myeloid maturation (SDT) rats,” Experimental Animals,vol.54,no.1,pp.13–20, in peripheral blood,” Clinical & Experimental Immunology,vol. 2005. 186, no. 3, pp. 373–386, 2016. [53] Y. Ishii, M. Maki, H. Yamamoto, T. Sasase, M. Kakutani, and T. [71] Y. Takeda, M. Marumo, H. Nara, Z.-G. Feng, H. Asao, and Ohta, “Blood pressure characteristics of female spontaneously I. Wakabayashi, “Selective induction of anti-inflammatory diabetic Torii-Leprfa rats,” Journal of Veterinary Medical Science, monocyte-platelet aggregates in a model of pulsatile blood flow vol. 73, no. 4, pp. 501–505, 2011. at low shear rates,” Platelets,vol.27,no.6,pp.583–592,2016. [54] I. Wakabayashi, T. Nakano, and Y. Takahashi, “Enhancement 𝛽 [72]J.W.Semple,J.E.ItalianoJr.,andJ.Freedman,“Plateletsand of interleukin-1 -induced iNOS expression in cultured vascular the immune continuum,” Nature Reviews Immunology, vol. 11, smooth muscle cells of Goto-Kakizaki diabetes rats,” European no. 4, pp. 264–274, 2011. Journal of Pharmacology,vol.629,no.1–3,pp.1–6,2010. [73] Y. Takeda, H. Nara, A. Araki, and H. Asao, “Human peripheral [55] Y.Takeda,M.Marumo,andI.Wakabayashi,“Attenuatedphago- neutrophils express functional IL-21 receptors,” Inflammation, cyticactivityofmonocytesintype2diabeticGoto-Kakizaki vol.37,no.5,pp.1521–1532,2014. rats,” Immunobiology, vol. 216, no. 10, pp. 1094–1102, 2011. [74] M. S. Hayden and S. Ghosh, “NF-𝜅B in immunobiology,” Cell [56] Y. Takeda and I. Wakabayashi, “Alteration in lymphocyte Research, vol. 21, no. 2, pp. 223–244, 2011. population and humoral immune response in type 2 diabetic [75] D. B. Ballak, R. Stienstra, C. J. Tack, C. A. Dinarello, and Goto-Kakizaki rats,” Life Sciences,vol.90,no.13-14,pp.545–552, J. A. van Diepen, “IL-1 family members in the pathogenesis 2012. and treatment of metabolic disease: focus on adipose tissue [57] Y. Takeda, T. Shimomura, and I. Wakabayashi, “Immunological inflammation and insulin resistance,” Cytokine,vol.75,no.2, disorders of diabetes mellitus in experimental rat models,” pp.280–290,2015. Nihon eiseigaku zasshi. Japanese journal of hygiene,vol.69,no. [76] D. Pal, S. Dasgupta, R. Kundu et al., “Fetuin-A acts as an 3, pp. 166–176, 2014. endogenous ligand of TLR4 to promote lipid-induced insulin [58] T. A. Wynn, A. Chawla, and J. W.Pollard, “Macrophage biology resistance,” Nature Medicine,vol.18,no.8,pp.1279–1285,2012. in development, homeostasis and disease,” Nature,vol.496,no. [77] T. B. Koenen, R. Stienstra, L. J. Van Tits et al., “Hyperglycemia 7446, pp. 445–455, 2013. activates caspase-1 and TXNIP-mediated IL-1𝛽 transcription in [59] F. Ginhoux and S. Jung, “Monocytes and macrophages: devel- human adipose tissue,” Diabetes,vol.60,no.2,pp.517–524,2011. opmental pathways and tissue homeostasis,” Nature Reviews Immunology,vol.14,no.6,pp.392–404,2014. Hindawi Journal of Diabetes Research Volume 2017, Article ID 6494795, 6 pages http://dx.doi.org/10.1155/2017/6494795

Review Article Role of T Lymphocytes in Type 2 Diabetes and Diabetes-Associated Inflammation

Chang Xia,1,2 Xiaoquan Rao,2 and Jixin Zhong2

1 College of Health Science & Nursing, Wuhan Polytechnic University, Wuhan, Hubei, China 2Cardiovascular Research Institute, Case Western Reserve University, Cleveland, OH, USA

Correspondence should be addressed to Jixin Zhong; [email protected]

Received 23 July 2016; Revised 30 December 2016; Accepted 12 January 2017; Published 31 January 2017

Academic Editor: Monica Bullo

Copyright © 2017 Chang Xia et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Although a critical role of adaptive immune system has been confirmed in driving local and systemic inflammation in type2 diabetes and promoting insulin resistance, the underlying mechanism is not completely understood. Inflammatory regulation has been focused on innate immunity especially macrophage for a long time, while increasing evidence suggests T cells are crucial for the development of metabolic inflammation and insulin resistance since 2009. There was growing evidence supporting the critical implication of T cells in the pathogenesis of type 2 diabetes.WewilldiscusstheavailableeffectofTcellssubsetsinadaptive immune system associated with the procession of T2DM, which may unveil several potential strategies that could provide successful therapies in the future.

1. Introduction ∼5% in lean subjects to a level of up to 50% of all adipose tissue cells in obese individuals [8, 9]. Therefore, early studies Type 2 diabetes mellitus (T2DM) is characterized by impaired on inflammatory regulation of diabetes have been focused insulin secretion, glucose intolerance, and hyperglycemia. on innate immune function. Recent studies, however, suggest T2DM is widely viewed as a chronic, low-grade inflammatory adaptive immune system, especially T lymphocyte, also plays disease caused by long-term immune system imbalance, a pivotal role in the pathogenesis of T2DM. There has been metabolic syndrome, or nutrient excess associated with obe- a rapid growth in our understanding of the role of T cell sity [1, 2]. In addition, T2DM associated complications in the in the pathogenesis of T2DM in recent years, necessitating kidneys, arteries, and eyes are also manifested by inflamma- a timely review on this topic. In the review, we will discuss tory process [3]. Therefore, inflammation is considered asa the functional involvement of T cells in the pathogenesis of major driving force in T2DM and associated complications. T2DM, especially the regulatory effects of T cells on chronic Inflammation was first linked to insulin resistance and inflammation. diabetes in the early 1990s. Hotamisligil et al. reported an increase of TNF-𝛼 inadiposetissuefromdifferentanimal 2. Th1 and Th2 Cells models of obesity and diabetes. Neutralization of TNF-𝛼 in obese rats improved peripheral glucose uptake [4]. Studies Increasing evidence suggests a pathological role for CD4+ have indicated that other inflammatory cytokines such as IL- T cells in obesity and insulin resistance. Obesity is a 1𝛽 and IFN-𝛾 which are increased in obesity and diabetes also major critical risk factor for T2DM. A recent study by modulate insulin signaling [5, 6]. On the other hand, a num- Shirakawa et al. demonstrated that activated CD4+ T cells + hi lo ber of anti-inflammatory cytokines, such as IL-4 and IL-10, (CD4 CD44 CD62L ) increased in the visceral adipose have been associated with the protection of insulin sensitivity tissue of obese mice. These cells also expressed PD-1 and [7]. Macrophages are the major inflammatory cell type in the CD153 and displayed characteristics of cellular senescence glucose-utilizing tissues such as adipose tissue and liver. For [10]. It has also been shown that obesity induces MHC class example,macrophageintheadiposetissueincreasedfrom II expression on adipocytes and thus activates CD4+ T cells 2 Journal of Diabetes Research to initiate adipose tissue inflammation [11]. These studies observed in lymphocyte-deficient mice, which displayed a suggest CD4+ T cells may play an important role in obesity less severe phenotype of insulin resistance on short-term and obesity-induced insulin resistance. high-fat diet [25]. Th1 and CD8+ lymphocytes in the adipose CD4+ effector T cells can be further divided into tissue were significantly increased in response to a high- proinflammatory Th1, Th17, and anti-inflammatory Th2and fat diet, while anti-inflammatory Th2 and Treg cells were Foxp3+ regulatory T cell (Treg) subtypes based on their func- decreased [26]. tionality and cytokine production [12]. Once activated, Th1 CD4+ helper T cells also play a critical role in a and Th2 cells show many of the significant signs of inflam- series of complications associated with T2DM. Activated T mation, such as releasing large amount of cytokines. Th1 cells lymphocytes and the inflammatory cytokines increased in could produce interferon- (IFN-) gamma, interleukin-2 (IL- the kidneys in patients with T2DM [27, 28]. Experimental 2), and tumor necrosis factor- (TNF-) beta, triggering cell- evidence indicates that the activation of Th2 cell-mediated mediated immunity and phagocyte-dependent inflammation immunity is delayed and impaired in diabetes [29]. It has [12]. Th2 cells, in contrast, produce IL-4, IL-5, IL-6, IL-9, IL- been shown that Th1-associated cytokines induce hyperin- 10, and IL-13 to regulate antibody responses [13]. Studies have flammatory response and subsequently lead to progressive shown that Th1 and Th2 cells had a key functional role in innate immune response [30]. It is reported that the circu- regulating inflammatory process, although they are activated later than macrophages in inflammation [14–16]. lating levels of Th1-associated cytokines increased in diabetic It is widely accepted that macrophage activation is patients [30]. The levels of T cell-related cytokines such as affected by T cells. T cells play a dominant role in promoting IL-10 and IL-17 were significantly higher in patients with or sustaining inflammatory processes and insulin resistance T2DM, suggesting an involvement of T cells in diabetes through inducing proinflammatory cytokines in metabolic [31]. In consistency, oral anti-CD3 plus glucosylceramide (an organs, such as the adipose tissue, liver, muscle, and pancreas. NKT cell target antigen) treatment was shown to induce the Macrophages are the major inflammatory cells in the adipose production of IL-10 and TGF-𝛽,whichwasassociatedwith tissue in obesity, increasing from 5–10% in lean subjects improved fasting glucose, visceral adipose tissue inflamma- to up to 50% in obese individuals [8, 9]. Based on their tion, liver enzymes, and hepatic steatosis in ob/ob mice [32]. cytokine production and activation conditions, adipose tissue These findings suggest both Th1 and Th2 are closely associated macrophages are categorized into two populations, proin- with insulin resistance and chronic inflammation in T2DM. flammatory M1 (classically activated macrophage) and anti- inflammatory M2 (alternatively activated macrophage). M1 3. Th17 Cells macrophages release proinflammatory cytokines including TNF-𝛼, IL-6, and IL-1𝛽, which contribute to the local and Th17, an important proinflammatory CD4+ T cell subtype systemic inflammation [16]. In contrast, M2 macrophages secreting IL-17, has also been associated with T2DM [33, 34]. secrete IL-10, which could inhibit the activity of most proin- A recent study examined the differentiation of different CD4+ flammatory cell types including M1 macrophages [16, 17]. IL- T cell subsets in T2DM by analyzing the cytokine production 10 has been shown to suppress TNF-𝛼 by interacting with the by PBMCs [35]. The results demonstrated that Th17 cells p38/MAPKpathway[17,18].Th1cellscouldsecretIFN-𝛾 to increased in T2DM patients and might be associated with promote M1 polarization and enhance its proinflammatory dysregulated lipid metabolism [35]. Studies have shown that functions by inducing the release of IL-1, IL-6, and TNF- IL-17 could stimulate the production of TNF-𝛼,thefirst 𝛼. In contrast, Th2 cells could produce anti-inflammatory cytokine being associated with obesity and insulin resistance. IL-4 and IL-13 to skew the differentiation of macrophage It is reported that intestinal Th17 cells may induce AMPs, towards M2 [19, 20]. In diet-induced obesity, CD4+T cells specifically Reg3b and Reg3g, contributing to increased were increased in the adipose tissue and induced the recruit- intestinal permeability in high-fat diet-fed mice [35, 36]. Th17 ment and differentiation of TNF-𝛼-secreting M1, while the has also been shown to increase in diabetic complications number of IL-10-secreting M2 reduced in the adipose tissue such as diabetic nephropathy [28]. The function of Th17 cells [21]. These findings suggested the macrophage polarization could not be ignored in the progress of complications asso- of proinflammatory M1 versus anti-inflammatory M2 was ciatedwithT2DM.TheresultshowedthatTh17countsand viewed as a key factor in the development of obesity and Th17/Treg ratio increased in diabetic nephropathy patients T2DM [17]. Therefore, Th1 and Th2 responses, which are compared with healthy individuals and diabetic patients closely related to M1/M2 polarization, may also have a critical without nephropathy [28]. The adipose tissue (AT) DCs role in obesity and T2DM. isolatedfromobeseanimalsandhumanswereassociatedwith SeveralclinicalstudieshaveconfirmedthatTh1was the differentiation of Th17 cells in vitro [37]. However, the upregulated in the adipose tissue and peripheral blood from interaction between DCs and Th17 cells have not yet been the individuals with prediabetes or T2DM [22]. Zeng and identified in AT from lean or obese animals [38]. Whether coworkers reported an imbalance of CD4+ T helper cell CD103 DCs can regulate obesity-induced inflammation in AT subsets including Treg, Th1, and Th17 in the patients with by influencing the balance of Treg and Th17 cell differentia- T2DM [23]. In contrast, naive CD4+ T cells were reported to tion in vivo remains unknown [38]. decline in the patients with T2DM, which may be associated IL-22 is produced by Th17, the second product down- with adaptive immune activation and chronic inflammation stream of IL-6 and IL-23. It was shown that the microphages during the pathogenesis of T2DM [24]. Similar findings were from AT could express the IL-22 receptor (IL-22R) and Journal of Diabetes Research 3

hi respond to IL-22 to secret more IL-1b, promoting AT inflam- (CTLA-4) levels were higher on the surface of CD39 Tregs, low hi mation [39, 40]. It is interesting whether IL-22 associated compared with CD39 Tregs [51]. In addition, CD39 low metaboliceffectsinmicecouldapplytohumans.BothIL-17 Tregs could produce more IL-10 than CD39 Tregs [51]. and IL- 22 indicate that Th17 may play an alternative role of Adenosine and its analogues have also been demonstrated “protective” and “destructive” in T2DM development [41]. to promote lymphocyte apoptosis through the A2A receptor, indicating Tregs in diabetic patients may be more susceptible 4. Regulatory T Cells to apoptosis [52].

Treg cells, a small subset of T lymphocytes constituting 5. CD8+ Cytotoxic T Cells only 5–20% of the CD4+ compartment, are thought to be important to prevent excessive inflammatory responses and The inflammation of adipose tissue is considered as a key limittissueimpairment[42,43].Typically,theyregulatethe event leading to the metabolic syndrome, diabetes, and response of other T cell subtypes but can also influence atherosclerotic cardiovascular disease. It was reported that, theactivitiesofinnateimmunecells[44,45].Tregcellsare within two weeks of high-fat diet feeding, the CD8+CD4− T characterized by high-level expression of a fork head/winged- cells were significantly increased in C57BL/6 mice [21]. The helix transcription factor, Foxp3. Some studies showed that amount of CD8+CD4− T cells kept increasing until 15 weeks. the relative proportion of Th1 to Foxp3+ cells was positively Incontrast,TregcellsandCD4+Tcellswerereducedupon correlated with body mass index (BMI) [46]. A higher high-fatdiet[21].Unlikeinwildtypemice,therewereno proportion of Treg was observed in lean mice [19, 46]. significant increases in the M1 or M2 macrophage fraction Fat-resident regulatory T cells were commonly observed in in CD8-deficient mice under high-fat diet although both close proximity to monocytes or macrophages, which imply body weight and epididymal fat mass significantly increased potential links with fat associated monocytes or macrophage on high-fat diet [21]. More importantly, it was shown that [19]. The CD62L+ CD44− naive CD4+ T cells were increased CD8+Tcellswereessentialforinductionofmacrophage in aged fat-specific Treg knockout mice, but there were no activation and migration to adipose tissue by secreting MCP- 1,MCP-3,andRANTES(regulatedonactivation,normalT significant differences in the levels of inflammatory cytokines cell expressed and secreted) [21, 53]. These studies indicate in serum, such as TNF-𝛼,IL1𝛽,IL6,IFN-𝛾,andIL17,com- that CD8+ T cells were crucially involved in evoking inflam- paredwithwildtypemice[19].Tregcellswerehigherinthe matory cascades in obese adipose tissue. fat versus spleen and lung detected by intracellular staining for IL-10 protein [19]. Because the expression of IL-10 receptor Recent clinical studies also reported that the level of 𝛾 was upregulated in fat Treg cells, it seems the Treg cells not IFN- produced by CD3+ T cells was positively correlated only produced IL-10 but also responded to it [19]. withBMIinpatientswithT2DM.Anotherresearchindicates InT2DM,TregcellscouldsuppressTh1,Th2,andTh17 that CD8+ T cells increased in both the small bowel and response to improve insulin resistance. Treg could inhibit the colon in some obese individuals, compared with lean humans 𝛾 inflammatory response by various pathways, such as surpass- [54]. This study also showed the level of IFN- produced ing cytokine secretion, modulating the microenvironment, by CD8+ T cells increased in obese individuals [54]. The and changing the expression of surface receptors [47]. The increased intraepithelial CD8+ T cells may have the potential appropriate balance between proinflammatory (Th17 or Th1) to modulate the insulin sensitivity of enterocyte [54]. and anti-inflammatory (Treg) subsets of T cells is vital to CD137-CD137L interaction contributes to CD8+ cyto- maintain host immunity and control inflammatory damage toxic T cell proliferation and secretion of IFN-𝛾,TNF-𝛼,IL- [48]. It was found that the amount of Treg cells decreased 2, and IL-4 [55, 56]. The expression of CD137 was upregu- in patients with T2DM [46]. Zeng et al. also reported that lated in obese humans and mice [56]. CD137-CD137L could Treg/Th17 ratio and Treg/Th1 ratio decreased in patients with promote monocyte and T cell recruitment to the adipose T2DM [49]. Further analysis suggests peripheral induced tissue [57]. The inflammation in adipose tissue decreased in −/− Treg but not natural Treg produced in the thymus reduced CD137 mice, but glucose tolerance was strengthened [58]. in T2DM, possibly due to decreased bcl2/bax and low HDL Therefore, CD137 contributes to abnormal glucose and lipid level [49]. metabolism, which may involve the expansion and activation Previous study has demonstrated that the expression of CD8+ T cells. of CD39, an ectoenzyme highly expressed in Treg cells, increased in CD4+ lymphocytes from T2DM and obese 6. 𝛾𝛿 T Cells patients [50]. Adenosine triphosphate could be hydrolyzed by CD39 and adenosine diphosphate into adenosine monophos- T cells are subdivided into two major populations according phate (AMP), which subsequently converted into a T cell sup- to their surface expression of 𝛼𝛽 and 𝛾𝛿 T cell receptors pressive metabolite adenosine by CD73 [50]. CD39 was con- (TCR). Activated 𝛾𝛿 T cells could produce a vast variety firmedtobeassociatedwiththestabilityandthesuppressive ofcytokinessuchasIFN-𝛾 and TNF-𝛼 to regulate the function of Foxp3+Treg cells. T2DM patients with obesity function of other immune cells. In recent years, there is hi showed a significantly lower level of CD39 Treg compared increasing evidence indicating that 𝛾𝛿 T cells may interact with overweight controls [51]. It was reported that PD- with macrophages, CTLs, Th1/Th2 cells, Treg, and Th17 cells L1, CD25, and cytotoxic T lymphocyte-associated protein-4 depending on specific microenvironment, bridging innate 4 Journal of Diabetes Research and adaptive immunity. 𝛾𝛿 T cells could produce IL-10 and Acknowledgments IL-17 [59]. In addition, they also secret TNF-𝛼 to regulate the activation of CD8+ T cells [60]. This work was supported by grants from NIH (K01 Obesity is an important risk factor for chronic inflam- DK105108), AHA (15SDG25700381 and 13POST17210033), Na- matory diseases, such as T2DM and cardiovascular disease tional Natural and Scientific Foundation of China (81670431), [61]. 𝛾𝛿 T cells have been suggested to play a critical role Mid-Atlantic Nutrition Obesity Research Center (NORC) in chronic inflammatory disease. It was found that obese under NIH Award no. P30DK072488, and BoehringerIngel- patients exhibited a decreased level of V𝛾9V𝛿2Tcellsand heim (IIS2015-10485). Xiaoquan Rao was supported by the level of secreting IFN-𝛾 was also declined. These V𝛾9V𝛿2 K99ES026241 and T32DK098107. Chang Xia was supported T cells could prefer to differentiate mature T effecter memory by grants from National Natural and Scientific Foundation cluster of differentiation 45RA (CD45RA+) cells, indicating of China (81101247) and Zhejiang Provincial Natural Science 𝛾𝛿 T cells are also involved in inflammation in obesity and Foundation of China (Y2110580). diabetes [62]. A recent study reported that IL-2 treatment rescued V𝛾9V𝛿2 T cell cytokine production [63], which References provides a novel insight into the pathogenic role 𝛾𝛿 Tcells in the development of T2DM. [1] J. M. Guzman-Flores´ and S. Lopez-Briones,´ “Cells of the innate and adaptive immunity in type 2 diabetes mellitus and obesity,” Gaceta Medica de Mexico,vol.148,no.4,pp.381–389,2012. 7. iNKT Cells [2]C.J.Shu,C.Benoist,andD.Mathis,“Theimmunesystem’s involvement in obesity-driven type 2 diabetes,” Seminars in Previous studies have demonstrated that circulating natural Immunology,vol.24,no.6,pp.436–442,2012. killer cells (iNKT cells) were decreased and their function was suppressed in obese individuals, compared with lean [3] V.Guarner and M. E. Rubio-Ruiz, “Low-grade systemic inflam- mation connects aging, metabolic syndrome and cardiovascular individuals [64]. It has been identified that iNKT cells are disease,” Aging and Health—A Systems Biology Perspective,vol. enrichedinhumanandmurineadiposetissue[65,66].iNKT 40, pp. 99–106, 2014. cells could recognize lipid antigens presented by CD1d but [4] G. S. Hotamisligil, N. S. Shargill, and B. M. Spiegelman, not peptide through MHC molecules in the innate immune “Adipose expression of tumor necrosis factor-𝛼: direct role in system [67, 68]. Further investigation has been shown that obesity-linked insulin resistance,” Science,vol.259,no.5091,pp. adipose iNKT cells are a tissue resident with little influx from 87–91, 1993. the circulation and play an important anti-inflammatory [5] D. Mathis, “Immunological goings-on in visceral adipose tis- regulatory role in congenic parabiotic mice [69]. However, sue,” Cell Metabolism,vol.17,no.6,pp.851–859,2013. adiposeiNKTcellsareobviouslydecreasedinobesity[65, [6] N. Ouchi, J. L. Parker, J. J. Lugus, and K. Walsh, “Adipokines in 66, 70]. iNKT cells can regulate the cross talk between inflammation and metabolic disease,” Nature Reviews Immunol- innate and adaptive immunity as potent transactivators. Most ogy, vol. 11, no. 2, pp. 85–97, 2011. studies reported that iNKT cell-deficient mice have worse [7] J. I. Odegaard, R. R. Ricardo-Gonzalez, M. H. Goforth et al., metabolic disorder and weight gain [65, 66, 70, 71]. 𝛼- 𝛾 𝛼 “Macrophage-specific PPAR controls alternative activation Galactosylceramide ( GalCer), a potent lipid ligand, could and improves insulin resistance,” Nature,vol.447,no.7148,pp. 𝛼 activate iNKT cell. It is shown that injection of GalCer in 1116–1120, 2007. obesity increases the iNKT cell amount and induces weight [8] S. P. Weisberg, D. McCann, M. Desai, M. Rosenbaum, R. L. loss rapidly, anti-inflammatory macrophage differentiation, Leibel, and A. W. Ferrante Jr., “Obesity is associated with ma- and reversal of glucose and insulin sensitivity without hypo- crophage accumulation in adipose tissue,” Journal of Clinical glycemia [72, 73]. The majority of findings suggest that iNKT Investigation,vol.112,no.12,pp.1796–1808,2003. cells in AT play a critical role in regulating local inflammation [9] L. Boutens and R. Stienstra, “Adipose tissue macrophages: going and protecting against metabolic disorder in obesity [74]. off track during obesity,” Diabetologia,vol.59,no.5,pp.879– 894, 2016. 8. Conclusion [10] K. Shirakawa, X. Yan, K. Shinmura et al., “Obesity accelerates T cell senescence in murine visceral adipose tissue,” Journal of In conclusion, there is increasing evidence linking the acti- Clinical Investigation,vol.126,no.12,pp.4626–4639,2016. vation of T lymphocytes to T2DM. Due to the importance [11] T. Deng, C. J. Lyon, L. J. Minze et al., “Class II major histo- of inflammation in the development of insulin resistance, compatibility complex plays an essential role in obesity-induced T2DM is now considered as an autoimmune disease [75]. adipose inflammation,” Cell Metabolism,vol.17,no.3,pp.411– Understanding the role of specific immune cells and proin- 422, 2013. flammatory molecules in obesity and diabetes is required [12]I.Raphael,S.Nalawade,T.N.Eagar,andT.G.Forsthuber,“T for developing novel therapeutic approaches to modulate cell subsets and their signature cytokines in autoimmune and metabolic inflammation and insulin resistance. inflammatory diseases,” Cytokine,vol.74,no.1,pp.5–17,2015. [13] S. E. Kahn, R. L. Hull, and K. M. Utzschneider, “Mechanisms Competing Interests linking obesity to insulin resistance and type 2 diabetes,” Nature, vol.444,no.7121,pp.840–846,2006. The authors declare that there is no conflict of interests [14] F. O. Martinez, “Macrophage activation and polarization,” regarding the publication of this paper. Frontiers in Bioscience, vol. 13, no. 13, p. 453, 2008. Journal of Diabetes Research 5

[15]D.E.Cintra,J.R.Pauli,E.P.Araujo´ et al., “Interleukin-10 is a [32] Y. Ilan, R. Maron, A.-M. Tukpah et al., “Induction of regulatory protective factor against diet-induced insulin resistance in liver,” T cells decreases adipose inflammation and alleviates insulin Journal of Hepatology,vol.48,no.4,pp.628–637,2008. resistance in ob/ob mice,” Proceedings of the National Academy [16]C.N.Lumeng,J.L.Bodzin,andA.R.Saltiel,“Obesityinducesa of Sciences of the United States of America,vol.107,no.21,pp. phenotypic switch in adipose tissue macrophage polarization,” 9765–9770, 2010. Journal of Clinical Investigation,vol.117,no.1,pp.175–184,2007. [33] C. Zhang, C. Xiao, P. Wang et al., “The alteration of Th1/Th2/ [17] S. Winer, Y. Chan, G. Paltser et al., “Normalization of obesity- Th17/Treg paradigm in patients with type 2 diabetes mellitus: associated insulin resistance through immunotherapy,” Nature relationship with diabetic nephropathy,” Human Immunology, Medicine,vol.15,no.8,pp.921–929,2009. vol.75,no.4,pp.289–296,2014. [18] F. C. McGillicuddy, E. H. Chiquoine, C. C. Hinkle et al., [34] L. A. Zuniga, W. Shen, B. Joyce-Shaikh et al., “IL-17 regulates “Interferon 𝛾 attenuates insulin signaling, lipid storage, and adipogenesis, glucose homeostasis, and obesity,” The Journal of differentiation in human adipocytes via activation of the JAK/ Immunology,vol.185,no.11,pp.6947–6959,2010. STAT pathway,” TheJournalofBiologicalChemistry,vol.284,no. [35] L. Garidou, C. Pomie,´ P. Klopp et al., “The gut microbiota 46, pp. 31936–31944, 2009. regulates intestinal CD4 T cells expressing ROR𝛾tandcontrols [19] M. Feuerer, L. Herrero, D. Cipolletta et al., “Lean, but not obese, metabolic disease,” Cell Metabolism,vol.22,no.1,pp.100–112, fat is enriched for a unique population of regulatory T cells that 2015. affect metabolic parameters,” Nature Medicine,vol.15,no.8,pp. [36]M.K.Hamilton,G.Boudry,D.G.Lemay,andH.E.Raybould, 930–939, 2009. “Changes in intestinal barrier function and gut microbiota [20]A.Chatzigeorgiou,K.P.Karalis,S.R.Bornstein,andT. in high-fat diet-fed rats are dynamic and region dependent,” Chavakis, “Lymphocytes in obesity-related adipose tissue in- American Journal of Physiology—Gastrointestinal and Liver flammation,” Diabetologia, vol. 55, no. 10, pp. 2583–2592, 2012. Physiology,vol.308,no.10,pp.G840–G851,2015. [21] S. Nishimura, I. Manabe, M. Nagasaki et al., “CD8+ effector T [37] A. Bertola, T. Ciucci, D. Rousseau et al., “Identification of cells contribute to macrophage recruitment and adipose tissue adipose tissue dendritic cells correlated with obesity-associated inflammation in obesity,” Nature Medicine,vol.15,no.8,pp. insulin-resistance and inducing Th17 responses in mice and 914–920, 2009. patients,” Diabetes,vol.61,no.9,pp.2238–2247,2012. [22] T. McLaughlin, L. Liu, C. Lamendola et al., “T-cell profile in [38] B.-C. Lee and J. Lee, “Cellular and molecular players in adipose adipose tissue is associated with insulin resistance and systemic tissue inflammation in the development of obesity-induced inflammation in humans,” Arteriosclerosis, Thrombosis, and insulin resistance,” Biochimica et Biophysica Acta—Molecular Vascular Biology,vol.34,no.12,pp.2637–2643,2014. Basis of Disease,vol.1842,no.3,pp.446–462,2014. [23] C. Zeng, X. Shi, B. Zhang et al., “The imbalance of Th17/Th1/ [39] E. Dalmas, N. Venteclef, C. Caer et al., “T cell-derived IL-22 Tregs in patients with type 2 diabetes: relationship with meta- amplifies IL-1𝛽-driven inflammation in human adipose tissue: bolic factors and complications,” Journal of Molecular Medicine, relevance to obesity and type 2 diabetes,” Diabetes,vol.63,no. vol.90,no.2,pp.175–186,2012. 6, pp. 1966–1977, 2014. [24]M.P.Nekoua,R.Fachinan,A.K.Atchamouetal.,“Modulation [40] R. Zhao, D. Tang, S. Yi et al., “Elevated peripheral frequencies of immune cells and Th1/Th2 cytokines in insulin-treated type of Th22 cells: a novel potent participant in obesity and type 2 2diabetesmellitus,”African Health Sciences,vol.16,no.3,pp. diabetes,” PLoS ONE,vol.9,no.1,ArticleIDe85770,2014. 712–724, 2016. [41]B.C.Ip,A.E.Hogan,andB.S.Nikolajczyk,“Lymphocyte [25] Y. S. Lee, P. Li, J. Y. Huh et al., “Inflammation is necessary roles in metabolic dysfunction: of men and mice,” Trends in for long-term but not short-term high-fat diet-induced insulin Endocrinology and Metabolism,vol.26,no.2,pp.91–100,2015. resistance,” Diabetes,vol.60,no.10,pp.2474–2483,2011. [26] K. A. Harford, C. M. Reynolds, F. C. McGillicuddy, and H. [42] Y.Zheng and A. Y.Rudensky, “Foxp3 in control of the regulatory M. Roche, “Fats, inflammation and insulin resistance: insights Tcelllineage,”Nature Immunology,vol.8,no.5,pp.457–462, to the role of macrophage and T-cell accumulation in adipose 2007. tissue,” Proceedings of the Nutrition Society,vol.70,no.4,pp. [43] S. Sakaguchi, T. Yamaguchi, T. Nomura, and M. Ono, “Regu- 408–417, 2011. latory T cells and immune tolerance,” Cell,vol.133,no.5,pp. [27] J. F. Navarro and C. Mora, “Diabetes, inflammation, proin- 775–787, 2008. flammatory cytokines, and diabetic nephropathy,” TheScientific- [44] T. J. Murphy, N. N. Choileain, Y. Zang, J. A. Mannick, and + + WorldJournal,vol.6,pp.908–917,2006. J. A. Lederer, “CD4 CD25 regulatory T cells control innate [28] S. Abouzeid and N. Sherif, “Role of alteration in Treg/Th17 cells’ immune reactivity after injury,” Journal of Immunology,vol.174, balance in nephropathic patients with Type 2 diabetes mellitus,” no. 5, pp. 2957–2963, 2005. Electronic physician,vol.7,no.8,pp.1613–1618,2015. [45] L. T. Nguyen, J. Jacobs, D. Mathis, and C. Benoist, “Where [29] D. Wu, A. B. Molofsky, H.-E. Liang et al., “Eosinophils sustain FoxP3-dependent regulatory T cells impinge on the develop- adipose alternatively activated macrophages associated with ment of inflammatory arthritis,” Arthritis & Rheumatism,vol. glucose homeostasis,” Science,vol.332,no.6026,pp.243–247, 56,no.2,pp.509–520,2007. 2011. [46] M. Jagannathan-Bogdan, M. E. McDonnell, H. Shin et al., “Ele- [30] C. Francisco, A. Catai, S. Moura-Tonello et al., “Cytokine profile vated proinflammatory cytokine production by a skewed T cell and lymphocyte subsets in type 2 diabetes,” Brazilian Journal of compartment requires monocytes and promotes inflammation Medical and Biological Research,vol.49,no.4,2016. in type 2 diabetes,” Journal of Immunology,vol.186,no.2,pp. [31] S. Mocellin, F. Marincola, C. R. Rossi, D. Nitti, and M. Lise, 1162–1172, 2011. “The multifaceted relationship between IL-10 and adaptive [47] J. M. Guzman-Flores´ and D. P. Portales-Perez,´ “Mechanisms immunity: putting together the pieces of a puzzle,” Cytokine and of suppression by regulatory T cells (Treg),” Gaceta Medica de Growth Factor Reviews,vol.15,no.1,pp.61–76,2004. Mexico,vol.149,no.6,pp.630–638,2013. 6 Journal of Diabetes Research

[48] M. F. Gregor and G. S. Hotamisligil, “Inflammatory mecha- [63]A.E.Costanzo,K.R.Taylor,S.Dutt,P.P.Han,K.Fujioka, nisms in obesity,” Annual Review of Immunology,vol.29,pp. and J. M. Jameson, “Obesity impairs 𝛾𝛿 Tcellhomeostasisand 415–445, 2011. antiviral function in humans,” PLoS ONE,vol.10,no.3,Article [49] L. Zeng, H. Lu, H. Deng, P. Mu, X. Li, and M. Wang, “Non- ID e0120918, 2015. inferiority effects on glycemic control and 𝛽-cell function [64] L. A. Lynch, J. M. O’Connell, A. K. Kwasnik, T. J. Cawood, C. improvement in newly diagnosed type 2 diabetes patients: basal O’Farrelly, and D. B. O’Shea, “Are natural killer cells protecting insulin monotherapy versus continuous subcutaneous insulin the metabolically healthy obese patient?” Obesity,vol.17,no.3, infusion treatment,” Diabetes Technology and Therapeutics,vol. pp.601–605,2009. 14,no.1,pp.35–42,2012. [65] L. Lynch, D. O’Shea, D. C. Winter, J. Geoghegan, D. G. Doherty, + [50]V.DellaLatta,M.Cabiati,S.Rocchiccioli,S.DelRy,andM.-A. and C. O’Farrelly, “Invariant NKT cells and CD1d cells amass Morales, “The role of the adenosinergic system in lung fibrosis,” in human omentum and are depleted in patients with cancer Pharmacological Research,vol.76,pp.182–189,2013. and obesity,” European Journal of Immunology,vol.39,no.7,pp. [51] J. Gu, X. Ni, X. Pan et al., “Human CD39hi regulatory T cells 1893–1901, 2009. present stronger stability and function under inflammatory [66] L. Lynch, M. Nowak, B. Varghese et al., “Adipose tissue invariant conditions,” Cellular & Molecular Immunology,vol.13,pp.1–8, NKT cells protect against diet-induced obesity and metabolic 2016. disorder through regulatory cytokine production,” Immunity, [52]Y.Sun,W.-Z.Liu,T.Liu,X.Feng,N.Yang,andH.-F.Zhou, vol. 37, no. 3, pp. 574–587, 2012. “Signaling pathway of MAPK/ERK in cell proliferation, dif- [67] M. Brigl and M. B. Brenner, “CD1: antigen presentation and T ferentiation, migration, senescence and apoptosis,” Journal of cell function,” Annual Review of Immunology,vol.22,pp.817– Receptors and Signal Transduction,vol.35,no.6,pp.600–604, 890, 2004. 2015. [68] E. M. Beckman, S. A. Porcelli, C. T. Morita, S. M. Behar, S. T. [53] U. Kintscher, M. Hartge, K. Hess et al., “T-lymphocyte infiltra- Furlong, and M. B. Brenner, “Recognition of a lipid antigen by tion in visceral adipose tissue: a primary event in adipose tissue GDI-restricted 𝛼𝛽+Tcells,”Nature,vol.372,no.6507,pp.691– inflammation and the development of obesity-mediated insulin 694, 1994. resistance,” Arteriosclerosis, Thrombosis, and Vascular Biology, [69] L. Lynch, X. Michelet, S. Zhang et al., “Regulatory iNKT cells vol.28,no.7,pp.1304–1310,2008. lack expression of the transcription factor PLZF and control the [54]H.Luck,S.Tsai,J.Chungetal.,“Regulationofobesity-relat- homeostasis of T𝑟𝑒𝑔 cells and macrophages in adipose tissue,” ed insulin resistance with gut anti-inflammatory agents,” Cell Nature Immunology,vol.16,no.1,pp.85–95,2015. Metabolism,vol.21,no.4,pp.527–542,2015. [70] H. S. Schipper, M. Rakhshandehroo, S. F. J. Van De Graaf et al., [55] D. S. Vinay and B. S. Kwon, “Role of 4-1BB in immune re- “Natural killer T cells in adipose tissue prevent insulin resis- sponses,” Seminars in Immunology,vol.10,no.6,pp.481–489, tance,” Journal of Clinical Investigation,vol.122,no.9,pp.3343– 1998. 3354, 2012. [56] T.H. Tu, C.-S. Kim, J.-H. Kang et al., “Levels of 4-1BB transcripts [71] J. Y. Huh, J. I. Kim, Y. J. Park et al., “A novel function of and soluble 4-1BB protein are elevated in the adipose tissue of adipocytes in lipid antigen presentation to iNKT cells,” Molecu- human obese subjects and are associated with inflammatory lar and Cellular Biology,vol.33,no.2,pp.328–339,2013. and metabolic parameters,” International Journal of Obesity,vol. [72]E.Hams,R.M.Locksley,A.N.J.McKenzie,andP.G.Fallon, 38, no. 8, pp. 1075–1082, 2014. “Cutting edge: IL-25 elicits innate lymphoid type 2 and type [57]T.H.Tu,C.-S.Kim,T.Goto,T.Kawada,B.-S.Kim,andR.Yu, II NKT cells that regulate obesity in mice,” The Journal of “4-1BB/4-1BBL interaction promotes obesity-induced adipose Immunology,vol.191,no.11,pp.5349–5353,2013. inflammation by triggering bidirectional inflammatory signal- ing in adipocytes/macrophages,” Mediators of Inflammation, [73] Y. Ji, S. Sun, A. Xu et al., “Activation of natural killer T vol. 2012, Article ID 972629, 10 pages, 2012. cells promotes M2 macrophage polarization in adipose tissue and improves systemic glucose tolerance via interleukin-4 (IL- [58] C. S. Kim, J. G. Kim, B. J. Lee et al., “Deficiency for costimulatory 4)/STAT6 protein signaling axis in obesity,” Journal of Biological receptor 4-1BB protects against obesity-induced inflammation Chemistry, vol. 287, no. 17, pp. 13561–13571, 2012. and metabolic disorders,” Diabetes,vol.60,no.12,pp.3159–3168, 2011. [74] A. Kohlgruber and L. Lynch, “Adipose tissue inflammation in the pathogenesis of type 2 diabetes,” Current Diabetes Reports, [59]K.A.Rhodes,E.M.Andrew,D.J.Newton,D.Tramonti,and vol. 15, no. 11, article no. 92, 2015. S. R. Carding, “A subset of IL-10-producing 𝛾𝛿 Tcellsprotect + the liver from Listeria-elicited, CD8 T cell-mediated injury,” [75]S.Tsai,X.Clemente-Casares,X.S.Revelo,S.Winer,andD. European Journal of Immunology,vol.38,no.8,pp.2274–2283, A. Winer, “Are obesity-related insulin resistance and type 2 2008. diabetes autoimmune diseases?” Diabetes,vol.64,no.6,pp. 1886–1897, 2015. [60] K. D. C. Jensen, X. Su, S. Shin et al., “Thymic selection determines 𝛾𝛿 T cell effector fate: antigen-naive cells make interleukin-17 and antigen-experienced cells make interferon 𝛾,” Immunity,vol.29,no.1,pp.90–100,2008. [61] D. E. Lackey and J. M. Olefsky, “Regulation of metabolism by the innate immune system,” Nature Reviews Endocrinology,vol. 12,no.1,pp.15–20,2016. [62] T. Dimova, M. Brouwer, F. Gosselin et al., “Effector𝛾 V 9V𝛿2T cells dominate the human fetal 𝛾𝛿 T-cell repertoire,” Proceedings of the National Academy of Sciences,vol.112,no.6,pp.E556– E565, 2015. Hindawi Publishing Corporation Journal of Diabetes Research Volume 2016, Article ID 2543268, 11 pages http://dx.doi.org/10.1155/2016/2543268

Review Article TheRoleofHMGB1inthePathogenesisofType2Diabetes

Yanan Wang,1 Jixin Zhong,1,2 Xiangzhi Zhang,3 Ziwei Liu,1 Yuan Yang,1 Quan Gong,1 and Boxu Ren1

1 Department of Immunology, Medical School, Yangtze University, Jingzhou 434023, China 2Department of Medicine, University of Maryland School of Medicine, Baltimore, MD 21201, USA 3Department of Medicine, Hospital of Yangtze University, Jingzhou 434000, China

Correspondence should be addressed to Quan Gong; [email protected] and Boxu Ren; [email protected]

Received 15 July 2016; Revised 8 November 2016; Accepted 29 November 2016

Academic Editor: Konstantinos Kantartzis

Copyright © 2016 Yanan Wang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Significance. With an alarming increase in recent years, diabetes mellitus has become a global challenge. Despite advances in treatment of diabetes mellitus, currently, medications available are unable to control the progression of diabetes and its complications. Growing evidence suggests that inflammation is an important pathogenic mediator in the development of diabetes mellitus. The perspectives including suggestions for new therapies involving the shift from metabolic stress to inflammation should be taken into account. Critical Issues. High-mobility group box 1 (HMGB1), a nonhistone nuclear protein regulating gene expression, was rediscovered as an endogenous danger signal molecule to trigger inflammatory responses when released into extracellular milieu in the late 1990s. Given the similarities of inflammatory response in the development of T2D, we will discuss the potential implication of HMGB1 in the pathogenesis of T2D. Importantly, we will summarize and renovate the role of HMGB1 and HMGB1- mediated inflammatory pathways in adipose tissue inflammation, insulin resistance, and islet dysfunction. Future Directions. HMGB1 and its downstream receptors RAGE and TLRs may serve as potential antidiabetic targets. Current and forthcoming projects in this territory will pave the way for prospective approaches targeting the center of HMGB1-mediated inflammation to improve T2D and its complications.

1. Introduction T2D has been associated with chronic low-grade inflam- mation for decades. Given the importance of innate immu- It is reported that there are approximately 10 percent of the nity in inflammation, innate immune mediators may play adult population suffering from diabetes in the world. More a vital role in the development of this disease. Emerging importantly, the incidence of diabetes mellitus is increasing at evidence has indicated that HMGB1, a highly conserved an alarming rate [1]. T2D, a metabolic disorder formed after a nonhistone nuclear protein that serves as a damage asso- long and complicate pathological process, is characterized by ciated molecule pattern molecule, is associated with the decreased insulin sensitivity and following pancreatic 𝛽-cell pathogenesis of T2D. HMGB1 can signal through receptor dysfunction [2, 3]. After long-term overnutrition, metabolic for advanced glycation end products (RAGE) and Toll-like balance of body is broken and becomes an origination of 𝜅 𝜅 insulin resistance. Insulin resistance leads to compensatory receptors (TLRs) to activate nuclear factor- B(NF- B) sig- increase of insulin secretion and 𝛽-cell hypertrophy. Long- naling pathway [5, 6] thus contributing to the inflammatory term overload of work may result in islet 𝛽-cell dysfunction responses in T2D. even death. Several metabolic processes, such as endoplasmic In this review we will focus on the pathophysiological reticulum stress, hypoxia, and lipotoxicity, are all involved in connections between HMGB1 and obesity, insulin resistance, overnutrient-induced metabolic inflammation. Importantly, and islet dysfunction. The understanding of the role of obesity and obesity-associated inflammation have long been HMGB1 may provide a new insight into anti-inflammatory proposed to be responsible for insulin resistance and T2D [4]. therapeutic strategies for T2D. 2 Journal of Diabetes Research

2. Type 2 Diabetes the first time in 1973 by Goodwin and Johns [15] and was named because of its high migration ability in polyacrylamide T2D, also called non-insulin-dependent diabetes mellitus, is gel electrophoresis. According to the molecular weight, 𝛽 characterized by insulin resistance and pancreatic -cell sequence similarity, and DNA structure, HMG can be further dysfunction, resulting from an unsettled hyperglycemia con- divided into three families: HMGA, HMGB, and HMGN. dition [2, 3]. Insulin resistance runs through the whole HMGB1 is the most abundant HMG protein. HMGB1 is process of diabetes. The liver, muscle, and adipose tissue can a nonhistone chromosomal binding protein mainly located act as sites of insulin resistance. In order to compensate for in the nucleus of most tissues, wherein it binds to DNA insulin resistance, islet 𝛽 cells produce more insulin which 𝛽 and regulates chromatin remodeling, DNA damage repair, may exceed the maximum capability and results in cells and gene transcription [16, 17]. It is highly evolutionarily failure [7]. conserved in vertebrate animals and is widely distributed in Chronic, low-grade adipose tissue inflammation links lymphoid tissue, brain, liver, lung, heart, spleen, kidney, and obesity and insulin resistance, thereby playing a key role in other tissues. the early phase of T2D. In recent years, the role of inflam- HMGB1 molecule is expressed as a single polypeptide mation in the pathogenesis of T2D has been extensively chain of 215 amino acids (AA) and is composed of three studied. It has been shown that the peroxisome proliferator 𝛼/𝛾 distinct structural domains: A-box (AA 1–79), B-box (AA 89– activated receptor (PPAR) agonists attenuated insulin 162), and the acidic C tail (AA 186–215) [18]. Both A-box resistance in human adipocytes via reducing proinflamma- and B-box are able to bind to DNA and participate in the tory mediators including interleukin- (IL-) 6, CXC-L10, and folding and twisting of the DNA. The B-box is the functional monocyte chemoattractant protein (MCP)-1 [8]. Another region of inflammation. It consists of two crucial binding research reported that insulin significantly reduced several sites for TLR4 and RAGE and thus plays an important key mediators of inflammatory stress in humans [9]. These role in promoting inflammation. In comparison, the A-box studies indicated that anti-inflammatory mechanism might competes with full-length HMGB1 for binding sites and thus play a role in antidiabetic action. In other words, T2D is an induces anti-inflammatory effects [19–21]. One of our studies inflammatory disease. using recombinant A-box confirmed that the HMGB1 A-box As the two major features of T2D, both insulin resistance was able to alleviate LPS-induced inflammation in the lung 𝛽 and -cell dysfunction are closely related to inflammation. and modulate acute lung injury [22]. The acidic C terminus is Proinflammatory macrophages in obese adipose tissue and enriched with negatively charged aspartic acid and glutamic insulin resistance state are main immune cells, and there is acid for transcription stimulation. a detailed description of this process in an excellent recent HMGB1 also undergoes posttranslational modification review [4]. Proinflammatory factor, tumor necrosis factor- 𝛼 which determines its bioactivity. For example, there are 3 (TNF-) , was the first inflammatory maker that was been conserved redox-sensitive cysteines (C23, C45, and C106). suggested to play a role in the development of obesity- The disulfide linkage of C23 and C45 is required for the induced insulin resistance in the 1990s [10, 11]. In 1993, cytokine-stimulating activity of HMGB1 and C106 must Hotamisligil et al. [12] demonstrated that adipose tissue of remain in its reduced form as a thiol at the same time [23]. In 𝛼 obese rodents secreted TNF- , a potent negative regulator addition, both the acetylation of HMGB1 within its nuclear 𝛼 of insulin signaling. TNF- stimulates adipocyte lipolysis localization sequences [24] and the methylation at lysine 42 contributing to elevated serum free fatty acids (FFAs) con- [25] promote HMGB1 release by mobilizing HMGB1 from the centrations, which can lead to decreased insulin sensitivity. nucleus to the cytoplasm. Now, there are many lines of researches illustrating that obesity and obesity-induced insulin resistance are closely 3.2. HMGB1 Physiologic Function. HMGB1 is widely expre linked to inflammation. Recent studies have also identified a ssed in mammalian tissues and present in all vertebrate nuclei number of cellular and molecular players participating in the [26]. It was first discovered as a nuclear protein. Nuclear development of T2D. HMGB1 binds to DNA and regulates a number of key DNA events including V(D)J recombination, gene transcription, 3. HMGB1 DNA replication, and DNA repair [27–31]. It was considered HMGB1, a nuclear protein, was first known for its role in exclusively as a nuclear protein until 1999 that Wang et al. [32] the regulation of gene expression. Recent advances implicated first reported HMGB1 acted as a late inflammatory mediator that HMGB1 had alarming activities via activating proinflam- and was involved in the pathogenesis of sepsis. Afterwards, matory responses after being passively released by necrotic people began to recognize that HMGB1 played a significant cells or actively secreted by activated immune cells into role in the process of inflammation. the extracellular milieu [13, 14]. HMGB1, as an endogenous HMGB1 can be either actively secreted by stimulated danger signal triggering inflammatory responses, appears immune cells including activated monocytes, macrophages, to play an important role in the pathogenesis of several mature dendritic cells, natural killer cells, and endothe- inflammatory conditions, including sepsis, arthritis, cancer, lial cells, or passively released by necrotic and damaged and autoimmunity diseases. cells [33–36]. When secreted or released into extracellular environment, HMGB1 participates in several processes such 3.1. HMGB1 Biochemistry, Tissue Distribution, and Structure. as immune response, cell migration, cell differentiation, HMGB1 was extracted and identified in bovine thymus for proliferation, and tissue regeneration. HMGB1 is found to Journal of Diabetes Research 3 be associated with many inflammatory diseases such as and cytoplasmic Toll/interleukin-1 receptor (TIR) domain cancer, trauma, arthritis, ischemia reperfusion injury, sepsis, [70].TLRsarecharacterizedasonespeciesofpatternrecogni- cardiovascular shock, diabetes, and autoimmune diseases tion receptors to recognize several danger signals, including [37–41]. In addition, HMGB1 has been shown to be a crucial pathogen associated molecular patterns (PAMPs) and dam- coordinator of acute inflammation in various stress models age associated molecular patterns (DAMPs), and thus are [42]. involved in innate immune responses against infection and injury [71]. So far, 13 members of TLRs have been discovered. 4. HMGB1 Mediated Signaling Pathways Among those, HMGB1 can interact with TLR2, TLR4, and TLR9 [72–74]. HMGB1 signals through these three TLRs The proinflammatory signaling of HMGB1 is mainly trans- and activates myeloid differentiation factor 88- (MyD88-)- duced by RAGE and TLRs. RAGE was the first receptor iden- dependent pathway to activate the NF-𝜅Bandinterferon tified to be able to bind HMGB1. Nevertheless, the interaction regulatory factor (IRF) pathways. Studies suggested an with RAGE alone could not fully explain the complicated intramolecular Cys23–Cys45 fragment and Cys106 residue effects of HMGB1. Subsequent studies demonstrated that within the HMGB1 were required for TLR4 binding and HMGB1 might promote inflammation via interacting with activation [75, 76], while HMGB1-TLR2 structural basis needs TLRs (TLR2/4/9) and activating their downstream signaling to be further investigated. It was well elucidated that HMGB1 pathways. HMGB1, via interacting with its receptors, ulti- binding to lipopolysaccharide contributed to the transfer mately results in the activation of NF-𝜅B and production of of the complex to CD14 and, subsequently, to TLR4/MD2 proinflammatory cytokines including IL-6, IL-1𝛽,andTNF- complex [77–79]. Furthermore, it was researched that the 𝛼. Interestingly, the activation of NF-𝜅Bpathwaycouldin TLR4/MD2 receptor system specially recognized the disul- turn induce the expression of HMGB1 and its receptors, fide HMGB1, and this interaction was of great significance forming a positive feedback loop to sustain inflammatory for HMGB1-mediated inflammation [80]. TLR9 is a member conditions [43]. of the TLR family located in the endoplasmic reticulum- In recent studies, HMGB1 has also been suggested to Golgi intermediate compartment [81]. It was reported that signal through integrin [44], CXCR4 [45, 46], and triggering TLR9 was a receptor for unmethylated CpG-DNA [82]. receptor expressed on myeloid cell 1 (TREM1) [47]. However, HMGB1 acted as a CpG-ODN-binding protein that promoted the effects of these receptors and their downstream signaling cytokine release through interacting with TLR9 and signaling pathways are not completely understood. In this section, we through downstream pathways involving MyD88 and NF-𝜅B will briefly describe RAGE and Toll2/4/9-mediated signaling. molecules [83]. Moreover, HMGB1-CpG-DNA complex was responsible for the redistribution of TLR9 to early endosomes 4.1. RAGE. RAGE was first discovered as a cell surface and TLR9-mediated cytokine production [73, 83]. receptor for advanced glycation end products (AGEs) and is now recognized as a receptor for diverse ligands, including 5. Role of HMGB1 in Type 2 Diabetes HMGB1, 𝛽-amyloids [48], and S100 proteins [49, 50]. It is a primary receptor for HMGB1 [51]. HMGB1, as a late mediator of inflammation, has been RAGE is expressed on a various types of cells including proposed to be a significant mediator in the pathogenesis of a monocytes, macrophages, neurons, endothelial cells, and variety of diseases including T2D (Figure 1). The correlation many tumor cells [52, 53]. Knockout of RAGE reduced tumor between level of HMGB1 and diabetic complications has been growth and metastasis [54, 55] and augmented chemother- reported. Increased levels of HMGB1 have been reported in apy resistance [56]. HMGB1 and RAGE interaction mainly both diabetic patients and animal models. Pachydaki et al. induces CDC42/Rac and mitogen-activated protein kinase [84] and Yu et al. [85] observed an elevated expression of this (MAPKs) activation, which finally result in NF-𝜅Bactivation proteinintheretinasofdiabeticpatientsandratmodelswith [49, 52, 57]. The activation of CDC42/Rac and MAPKs path- retinopathy. Moreover, Dasu et al. [86] showed that circulat- ways also promotes chemotaxis, cytokines production [58], ing levels of HMGB1 were higher in type 2 diabetic patients endothelial cells activation [59], immune cells maturation, than control, and the same phenomenon was observed by and migration [60, 61]. Skrhaˇ et al. [87] Hagiwara et al. [88] demonstrated that hyper- HMGB1/RAGE binding activates the two canonical glycemia, induced by infusion of glucose in a rat model, was MAPKs, extracellular regulated protein kinases (ERK1/2) and associated with elevated serum HMGB1 levels. In addition, p38 MAPK, resulting in the nuclear translocation of NF-𝜅B metformin, a first-line antidiabetic drug, was also known and transcription of proinflammation cytokine [52, 62, 63]. to have anti-inflammatory effects. Tsoyi et al. [89] indicated The HMGB1/RAGE axis has been suggested to contribute to that metformin significantly decreased HMGB1 expression the pathogenesis of many diseases such as acute lung injury in LPS-treated RAW264.7 cells. Another report by Zhang [64], preeclampsia [65], diabetes [66], cancer [67, 68], and et al. [90] demonstrated that metformin protected against autoimmune diseases [69]. hyperglycemia-induced cardiomyocyte injury by inhibiting the expression of RAGE and HMGB1. 4.2. TLRs. TLRs are members of the type-1 transmembrane glycoprotein family which can bind extracellular ligands and 5.1.HMGB1andAdiposeTissueInflammation. Clinical and act as a transducer to sponsor proinflammatory signaling experimental researches showed that inflammation with the through ectodomain leucine-rich repeated (LRR) sequences infiltration of macrophages occurred in the adipose tissue 4 Journal of Diabetes Research

HMGB1 Necrosis Necrosis Islet

Secrete T cell

RAGE TLRs Adipose-derived stromal cells

NF-𝜅B NF-𝜅B

Proinflammatory cytokine (e.g., TNF-𝛼)

MΦ/DC

Figure 1: Potential involvement of HMGB1 in type 2 diabetes. Early inflammation in adipose tissue and pancreatic islets leads to the necrosis of adipose-derived stromal cells and islet cells. Necrotic cells release HMGB1, activating TLRs and RAGE on macrophages and dendritic cells. Activation of TLRs and RAGE leads to the translocation of NF-𝜅B into nucleus to promote the expression of inflammatory gene, which contributes to the secretion of proinflammatory cytokine, including HMGB1. In addition, activated macrophages and dendritic cells actively secrete HMGB1, which, in turn, exacerbate the necrosis of adipose tissue and pancreatic islets.

[91]. Although each cell type within the adipose tissue (such A growing number of studies elucidated a role of as preadipocyte, adipocyte, T cell, dendritic cell, and macro- HMGB1 in adipose tissue, such as activating proinflamma- phage) has a contribution to obesity-induced inflammation tory macrophages. Recently, a study among obese children and insulin resistance [92], macrophage is the primary source showed that obesity was positively correlated with increased of inflammatory effectors [93, 94]. Macrophages in the HMGB1 serum levels and was related to a number of adipose tissue can switch from an anti-inflammatory “M2” proinflammatory cytokines, such as IL-6 and TNF-𝛼 [102]. (alternatively activated) state to a proinflammatory “M1” Another study indicated that mice receiving the anti-HMGB1 (classically activated) state [95, 96]. M1 macrophages exert antibody gained less weight than the control animals [103]. proinflammatory effects through the secretion of cytokines However, anti-HMGB1 treatment only reduced the expres- TNF-𝛼,IL-1𝛽,andIL-6[97],whereasM2macrophagespro- sion of TNF-𝛼 and MCP-1 in the liver, but not in the adipose duce anti-inflammation cytokines such as IL-10 in contrast tissue [103]. They were unable to detect an improvement [98]. The imbalance of M1/M2 leads to an increased secretion of inflammation in high fat-induced adipose tissue after of proinflammatory factors such as TNF-𝛼 [99]. Moreover, anti-HMGB1 antibody treatment. Kanellakis and colleagues Vandanmagsar et al. [100] showed that macrophage inflam- demonstrated that administration of anti-HMGB1 antibody mation was also mediated by an excess of fatty acids (FAs) via led to a reduction in macrophage content of atherosclerotic TLR or NOD-like receptor family, one of which is the NOD- lesions in Apo-E deficient mice [104]. Indeed, in a recent like receptor family, the pyrin domain containing 3 (NLRP3) study, Song [105] and colleagues found that high fat feeding dependent pathway. However, there was one study showing induced expression of HMGB1 in the liver and adipose tissue, that endotoxin-free, free fatty acids did not have the ability while genetic deficiency of HMGB1 receptor RAGE prevented to activate TLR4 and thereby cause inflammation [101]. More the effects of high fat diet on energy expenditure, weight gain, importantly, they showed that some polyunsaturated fatty adipose tissue inflammation, and insulin resistance. acids exerted an anti-inflammatory action in human adipose The role of HMGB1 on macrophages has been clearly des- tissue and adipocytes. That might mean that HMGB1 was an cribed and stimulation of macrophages with HMGB1 induced evenmoreimportantTLR4ligandifFFAsindeed wereunable production of proinflammatory cytokines [53] which may to signal via TLR4. lead to an increase of adipose tissue inflammation and insulin Journal of Diabetes Research 5 resistance. Yu et al. [72] showed that anti-TLR2 antibodies (HOMA-IR) [102, 118–120]. In a further research of Dasu could reduce the binding of HMGB1 on the cell surface of and colleagues, the amount of HMGB1 was elevated in murine macrophage-like RAW 264.7 cells, leading to a reduc- human type 2 diabetic subjects and positively correlated tion in inflammation. Moreover, Yang et al. [75] reported that with TLR2 and TLR4, BMI, and HOMA-IR. In addition, the HMGB1 bound specifically to TLR4, leading to TNF release expression of MyD88 and NF-𝜅Bp65wasincreased[86]. from macrophages. Inhibition of HMGB1/TLR4 binding with The activation of TLR-MyD88-NF-𝜅B signaling resulted in neutralizing anti-HMGB1 mAb prevented HMGB1-induced elevated levels of cytokines [86]. Soon afterwards, Chen et cytokine release. These findings suggested that HMGB1 might al. [121] found that the expression of HMGB1 and NF-𝜅B have an important role in the initiation and development of and TNF-𝛼/vascular endothelial growth factor (VEGF) were adipose tissue inflammation. significantly upregulated in type 2 diabetic retinas and in high glucose treated ARPE-19 cells, compared to their respective 5.2. HMGB1 and Insulin Resistance. Insulin resistance refers counterparts. HMGB1 blockade significantly alleviated NF- to decreased capability of insulin to promote glucose uptake. 𝜅B activity and VEGF secretion in ARPE-19 cells stimulated In simple terms, insulin resistance is a condition in which with high glucose. Chen et al. [122] indicated that HMGB1 cells fail to appropriately respond to circulating insulin [106, was significantly upregulated by high glucose via NF-𝜅B 107]. All of the peripheral tissues such as liver, muscle, and signaling in vivo and in vitro, in an association with an adipose tissue can act as a site of insulin resistance. It has been elevation of proinflammatory cytokines. HMGB1 inhibition shown that adipose tissue macrophage infiltration was asso- reduced the upregulation of proinflammatory cytokines in ciated with an elevated level of serum insulin [108]. Therefore, response to high glucose. Hence, HMGB1 might be involved macrophage-mediated inflammatory response might be a in the development of insulin resistance via activating NF- critical contributor of insulin resistance. 𝜅B signaling and associated with a raised expression of The occurrence and development of insulin resistance are proinflammation mediators. closely related to the two signaling pathways: Jun N-terminal kinase (JNK) and I𝜅B kinase complex (IKK)𝛽/NF-𝜅B[109]. 5.3. HMGB1 and Pancreatic Islet Inflammation. Recently, a The JNK is known to play a role in regulating the development number of studies suggested that there was a sustained of obesity-induced insulin resistance. The activation of the inflammatory milieu in pancreatic islets in the process JNK decreased insulin sensitivity and this related in part to of T2D, owing to increased cytokine or chemokine pro- the phosphorylation of insulin receptor substrate- (IRS-) 1 on duction and immune cell infiltration. Recent studies on the serine 307 and 310 residues [110, 111]. Proinflammation human islets with T2D demonstrated that the combination molecules activate JNK to promote serine phosphorylation of hyperglycemia and elevated FFAs induced a more effi- of IRS-1 and IRS-2, a process closely related to insulin resis- cient proinflammatory phenotype, mainly via TLR, MyD88, tance. Furthermore, knockout of JNKs ameliorated obesity- and interleukin-1 receptor (IL-1R) I signals [123]. Studies induced insulin resistance [112, 113]. However, there was no in mouse showed that elevated levels of circulating sat- clear evidence showing that HMGB1 was involved in JNK urated FAs activated inflammatory processes within islets activation. On the contrary, the NF-𝜅Bsignalingpathwayis via TLR4/MyD88 pathway. 𝛽 cells produced chemokines closely related to HMGB1. and recruited CD11b(+)Ly-6C(+) M1-type proinflammatory Indeed, NF-𝜅B is a major factor in the regulation of monocytes/macrophages to the islets [124]. IL-1 plays a inflammation [114, 115]. In resting cells, NF-𝜅BandI𝜅B critical role in the islet inflammation. Type 2 diabetic GK complex locates in the cytoplasm and remains inactive. As rat displayed an increased expression of IL-1𝛽,andspe- the inhibitor of the NF-𝜅B, I𝜅BbindstoNF-𝜅Bthroughits cific blockade of IL-1 activity by the IL-1 receptor antag- specific ankyrin repeat motif in C-terminal. This binding onist (IL-1Ra) was associated with reduced inflammatory covers the nuclear localization sequence (NLS) of NF-𝜅Band cytokines/chemokines in GK islets in vitro and in vivo thus prevents its transfer to the nucleus. When the cells are when exposed to metabolic stress [125]. Meanwhile, TLR2 stimulated by extracellular signals, IKK is activated. IKK𝛽, and TLR4 and NLRP3 inflammasome also play a role in the catalytic subunit of IKK, phosphorylates I𝜅B. Following islet inflammation and islet dysfunction [126]. About ten the degradation of I𝜅B, the NLS of NF-𝜅Bisexposedandthe years ago, Steer and colleagues elucidated that IL-1 combined free NF-𝜅B translocates to the nucleus, where it regulates the with HMGB1 promoted islet inflammation and 𝛽-cell death, expression of multiple inflammatory genes including TNF-𝛼, which provided a proof for HMGB1 in the inflammation of IL-1𝛽, IL-2, and IL-6. pancreatic islet cells [127]. However, there was no further Increased NF-𝜅B activity has been observed in an obese study of HMGB1 involvement in islet inflammation later. animal model [116]. Blocking NF-𝜅Bprotectedhighfatdiet Instead, there were studies focusing on the role of HMGB1 mice from insulin resistance [117]. NF-𝜅Bsignalingpathway in islet transplantation [128–133]. HMGB1 inhibition by anti- could be activated by pattern recognition receptors like TLRs IL-6R antibody, HMGB1 A-box, NF-𝜅Binhibitor,orthe and RAGE, both of which have been shown to interact Na(+)/Ca(2+) exchanger inhibitor was able to ameliorate with HMGB1. This suggested that HMGB1 might play a inflammatory responses and islet survival after islet trans- critical role in insulin resistance through NF-𝜅B signaling. plantation by reducing the amount of innate immune cells In a number of human studies with obese state or T2D, and cytokines like TNF-𝛼 and interferon- (IFN-) 𝛾 [129– circulating level of HMGB1 was higher and positively corre- 132]. And blockade of HMGB1 with anti-HMGB1 monoclonal lated with homeostasis model assessment-insulin resistance antibody (mAb, 2g7) inhibited inflammatory response by 6 Journal of Diabetes Research reducing TNF-𝛼 and IL-1𝛽 production and thereby improved represents a promising therapeutic approach for controlling islet viability after islet cells transplantation [134]. inflammation in T2D.

5.4. HMGB1 and 𝛽-Cell Death. Pancreatic 𝛽-cell death is a Competing Interests major feature of the late stage of T2D. It is a consequence of The authors declare that they have no competing interests. insulin resistance and 𝛽-cell overcompensation. In a clinical 𝛽 study of T2D, the reduction of -cell mass was observed in the Authors’ Contributions pancreatic tissue sections [2]. It was reported that decreased 𝛽 𝛽 -cellmassinT2Dwasattributedtopancreatic -cell apop- Yanan Wang and Jixin Zhong contributed equally to this 𝛽 𝛽 tosis and to -cell dedifferentiation [135]. Furthermore, -cell work. dysfunction and apoptosis were associated with islet inflam- mation. Richardson et al. [136] suggested that a persistent, Acknowledgments low-grade enteroviral infection of islet endocrine cells might induce functional changes that contribute to the recruitment This work was supported by grants from the National Natural of macrophages in some patients with T2D. This article gave Science Foundation of China (81271872 and 81670431), NIH evidence that macrophage infiltration was involved in 𝛽- (K01DK105108), and the Natural Science Foundation of cell dysfunction and loss in T2D. In addition, saturated FAs Hubei Province (2015CFA080). activated inflammatory processes within islets via activating TLR4/MyD88 pathway and engendering chemokines [124]. References These results demonstrated that islet inflammation resulted in 𝛽-cell dysfunction and death. Alternately, HMGB1 played [1] IDF, IDF Diabetes Atlas, International Diabetes Federation, 6th an important role in the process of 𝛽-cell apoptosis according edition, 2013. to reports. Li et al. [13] demonstrated that TLR4 was the [2]A.E.Butler,J.Janson,S.Bonner-Weir,R.Ritzel,R.A.Rizza, main receptor of HMGB1 on 𝛽 cells and HMGB1 might and P. C. Butler, “𝛽-cell deficit and increased 𝛽-cell apoptosis in signal through TLR4 to selectively damage 𝛽 cells rather humans with type 2 diabetes,” Diabetes,vol.52,no.1,pp.102– than 𝛼 cells in type 1 diabetes mellitus. Another study in 110, 2003. NOD mice indicated that HMGB1 could be passively released [3] F.M. Ashcroft and P.Rorsman, “Diabetes mellitus and the 𝛽 cell: from damaged pancreatic 𝛽 cells and actively secreted by the last ten years,” Cell, vol. 148, no. 6, pp. 1160–1171, 2012. islet infiltrated immune cells. Blockade of HMGB1 inhibited [4]J.McNelisandJ.Olefsky,“Macrophages,immunity,and diabetes development in younger NOD mice [137]. These metabolic disease,” Immunity,vol.41,no.1,pp.36–48,2014. two studies together suggested that extracellular HMGB1 [5]J.R.Klune,R.Dhupar,J.Cardinal,T.R.Billiar,andA.Tsung, led to islet 𝛽 cells damage via binding to TLR4 on 𝛽 cells. “HMGB1: endogenous danger signaling,” Molecular Medicine, Necrotic islet 𝛽 cells, in turn, released HMGB1 to accelerate vol.14,no.7-8,pp.476–484,2008. cell damage, creating a vicious loop. In addition, HMGB1 [6] L. L. Mantell, W. R. Parrish, and L. Ulloa, “HMGB-1 as a and RAGE interaction induced islet cells apoptosis and therapeutic target for infectious and inflammatory disorders,” progressive 𝛽-cell loss by inducing oxidative stress in type 2 Shock,vol.25,no.1,pp.4–11,2006. diabetic rats [138]. One study showed that high concentration [7] B.-C. Lee and J. Lee, “Cellular and molecular players in adipose of insulin resulted in increased extracelluar HMGB1 and tissue inflammation in the development of obesity-induced induced the apoptosis of rat ovarian granulosa cells in vitro insulin resistance,” Biochimica et Biophysica Acta,vol.1842,no. 3, pp. 446–462, 2014. [139]. It has also been reported that HMGB1 could directly stimulate pancreatic 𝛽 cells to secrete insulin [120]. [8]M.Massaro,E.Scoditti,M.Pellegrinoetal.,“Therapeutic potential of the dual peroxisome proliferator activated receptor (PPAR)𝛼/𝛾 agonist aleglitazar in attenuating TNF-𝛼-mediated 6. Conclusion Remarks inflammation and insulin resistance in human adipocytes,” Pharmacological Research, vol. 107, pp. 125–136, 2016. HMGB1, an ubiquitously expressed and evolutionarily con- [9] P.Dandona, H. Ghanim, A. Bandyopadhyay et al., “Insulin sup- served chromosomal protein, has diverse roles in mediating presses endotoxin-induced oxidative, nitrosative, and inflam- inflammation. Increasing evidence indicates HMGB1 is a vital matory stress in humans,” Diabetes Care,vol.33,no.11,pp.2416– mediator in the onset and progression of diabetes. Elevated 2423, 2010. [10] J. M. Stephens and P. H. Pekala, “Transcriptional repression of level of HMGB1 has been found in serum and islets as 𝛼 well as other tissues like adipose, liver, and muscle from the C/EBP- and GLUT4 genes in 3T3-L1 adipocytes by tumor necrosis factor-𝛼: regulation is coordinate and independent of diabetic patients and animal models. Additionally, increased protein synthesis,” Journal of Biological Chemistry,vol.267,no. expression levels of RAGE and TLRs have also been reported 19, pp. 13580–13584, 1992. in diabetes, and these receptors are critically involved in [11]J.M.Stephens,J.Lee,andP.F.Pilch,“Tumornecrosisfactor-𝛼- the induction of proinflammatory cytokines. The functional induced insulin resistance in 3T3-L1 adipocytes is accompanied association among HMGB1, RAGE, and TLRs augments by a loss of insulin receptor substrate-1 and GLUT4 expression the inflammation in the course of T2D including obesity- without a loss of insulin receptor-mediated signal transduc- induced inflammation, insulin resistance, and islet inflam- tion,” JournalofBiologicalChemistry,vol.272,no.2,pp.971–976, mation. Therefore, blockade of HMGB1 and its receptors 1997. Journal of Diabetes Research 7

[12] G. S. Hotamisligil, N. S. Shargill, and B. M. Spiegelman, formation of polyglutamine aggregates,” Journal of Immunology, “Adipose expression of tumor necrosis factor-𝛼:directrolein vol. 190, no. 4, pp. 1797–1806, 2013. obesity-linked insulin resistance,” Science,vol.259,no.5091,pp. [29] M. E. Bianchi, M. Beltrame, and G. Paonessa, “Specific recogni- 87–91, 1993. tion of cruciform DNA by nuclear protein HMGl,” Science,vol. [13] M.Li,L.Song,X.Gao,W.Chang,andX.Qin,“Toll-likereceptor 243,no.4894,pp.1056–1059,1989. 𝛽 4onislet cells senses expression changes in high-mobility [30] K. Giese, J. Cox, and R. Grosschedl, “The HMG domain of group box 1 and contributes to the initiation of type 1 diabetes,” lymphoid enhancer factor 1 bends DNA and facilitates assembly Experimental & Molecular Medicine,vol.44,no.4,pp.260–267, of functional nucleoprotein structures,” Cell,vol.69,no.1,pp. 2012. 185–195, 1992. [14] D. Tang, R. Kang, H. J. Zeh III, and M. T. Lotze, “High-mobility [31] K. Javaherian, L. F. Liu, and J. C. Wang, “Nonhistone proteins group box 1 and cancer,” Biochimica et Biophysica Acta,vol.1799, HMG1 and HMG2 change the DNA helical structure,” Science, no.1-2,pp.131–140,2010. vol.199,no.4335,pp.1345–1346,1978. [15] G. H. Goodwin and E. W.Johns, “Isolation and characterisation [32]H.Wang,O.Bloom,M.Zhangetal.,“HMG-1asalatemediator of two calf thymus chromatin non histone proteins with high of endotoxin lethality in mice,” Science,vol.285,no.5425,pp. contents of acidic and basic amino acids,” European Journal of 248–251, 1999. Biochemistry,vol.40,no.1,pp.215–219,1973. [33] C. W. Bell, W. Jiang, C. F. Reich III, and D. S. Pisetsky, “The [16] Y. Liu, R. Prasad, and S. H. Wilson, “HMGB1: roles in base extracellular release of HMGB1 during apoptotic cell death,” excision repair and related function,” Biochimica et Biophysica American Journal of Physiology—Cell Physiology,vol.291,no.6, Acta, vol. 1799, no. 1-2, pp. 119–130, 2010. pp. C1318–C1325, 2006. [17] J. O. Thomas and K. Stott, “H1 and HMGB1: modulators of [34] M. T. Lotze and K. J. Tracey, “High-mobility group box 1 protein chromatin structure,” Biochemical Society Transactions,vol.40, (HMGB1): nuclear weapon in the immune arsenal,” Nature no.2,pp.341–346,2012. Reviews Immunology,vol.5,no.4,pp.331–342,2005. [18] R. Kang, R. Chen, Q. Zhang et al., “HMGB1 in health and disease,” Molecular Aspects of Medicine,vol.40,pp.1–116,2014. [35] P. Rovere-Querini, A. Capobianco, P. Scaffidi et al., “HMGB1 is an endogenous immune adjuvant released by necrotic cells,” [19]J.Li,R.Kokkola,S.Tabibzadehetal.,“Structuralbasisforthe EMBO Reports,vol.5,no.8,pp.825–830,2004. proinflammatory cytokine activity of high mobility group box 1,” Molecular Medicine,vol.9,no.1-2,pp.37–45,2003. [36] J. Fucikova, I. Moserova, I. Truxova et al., “High hydrostatic pressure induces immunogenic cell death in human tumor [20] D. Messmer, H. Yang, G. Telusma et al., “High mobility cells,” International Journal of Cancer,vol.135,no.5,pp.1165– group box protein 1: an endogenous signal for dendritic cell 1177, 2014. maturation and Th1 polarization,” Journal of Immunology,vol. 173, no. 1, pp. 307–313, 2004. [37] A. Tsung, R. Sahai, H. Tanaka et al., “The nuclear factor HMGB1 mediates hepatic injury after murine liver ischemia- [21] H. Yang, M. Ochani, J. Li et al., “Reversing established sepsis reperfusion,” Journal of Experimental Medicine,vol.201,no.7, with antagonists of endogenous high-mobility group box 1,” pp. 1135–1143, 2005. Proceedings of the National Academy of Sciences of the United States of America,vol.101,no.1,pp.296–301,2004. [38] U. Andersson and K. J. Tracey, “HMGB1 is a therapeutic [22]Q.Gong,J.-F.Xu,H.Yin,S.-F.Liu,L.-H.Duan,andZ.-L. target for sterile inflammation and infection,” Annual Review of Bian, “Protective effect of antagonist of high-mobility group Immunology,vol.29,pp.139–162,2011. box 1 on lipopolysaccharide-induced acute lung injury in mice,” [39] Q. Gong, H. Zhang, J.-H. Li et al., “High-mobility group box Scandinavian Journal of Immunology,vol.69,no.1,pp.29–35, 1 exacerbates concanavalin A-induced hepatic injury in mice,” 2009. Journal of Molecular Medicine,vol.88,no.12,pp.1289–1298, [23] H. Yang, D. J. Antoine, U. Andersson, and K. J. Tracey, 2010. “The many faces of HMGB1: molecular structure-functional [40] U. Andersson and H. Erlandsson-Harris, “HMGB1 is a potent activity in inflammation, apoptosis, and chemotaxis,” Journal of trigger of arthritis,” Journal of Internal Medicine,vol.255,no.3, Leukocyte Biology,vol.93,no.6,pp.865–873,2013. pp. 344–350, 2004. [24] Y. Tang, X. Zhao, D. Antoine et al., “Regulation of posttrans- [41] N. Taniguchi, K.-I. Kawahara, K. Yone et al., “High mobility lational modifications of HMGB1 during immune responses,” group box chromosomal protein 1 plays a role in the pathogen- Antioxidants & Redox Signaling,vol.24,no.12,pp.620–634, esis of rheumatoid arthritis as a novel cytokine,” Arthritis and 2016. Rheumatism,vol.48,no.4,pp.971–981,2003. [25] I. Ito, J. Fukazawa, and M. Yoshida, “Post-translational methy- [42] J.A.Nogueira-Machado,C.M.D.O.Volpe,C.A.Veloso,andM. lation of high mobility group box 1 (HMGB1) causes its cyto- M. Chaves, “HMGB1, TLR and RAGE: a functional tripod that plasmic localization in neutrophils,” Journal of Biological Chem- leads to diabetic inflammation,” Expert Opinion on Therapeutic istry,vol.282,no.22,pp.16336–16344,2007. Targets, vol. 15, no. 8, pp. 1023–1035, 2011. [26] U. Andersson, H. Wang, K. Palmblad et al., “High mobility [43] J. R. Van Beijnum, W. A. Buurman, and A. W. Griffioen, group 1 protein (HMG-1) stimulates proinflammatory cytokine “Convergence and amplification of toll-like receptor (TLR) and synthesis in human monocytes,” Journal of Experimental receptor for advanced glycation end products (RAGE) signaling Medicine,vol.192,no.4,pp.565–570,2000. pathways via high mobility group B1 (HMGB1),” Angiogenesis, [27] M. Stros,ˇ “HMGB proteins: interactions with DNA and chro- vol.11,no.1,pp.91–99,2008. matin,” Biochimica et Biophysica Acta,vol.1799,no.1-2,pp.101– [44] V. V. Orlova, E. Y. Choi, C. Xie et al., “A novel pathway of 113, 2010. HMGB1-mediated inflammatory cell recruitment that requires [28]H.J.Min,E.A.Ko,J.Wuetal.,“Chaperone-likeactivityof Mac-1-integrin,”The EMBO Journal, vol. 26, no. 4, pp. 1129–1139, high-mobility group box 1 protein and its role in reducing the 2007. 8 Journal of Diabetes Research

[45] M. Schiraldi, A. Raucci, L. M. Mu˜noz et al., “HMGB1 promotes [60]I.E.Dumitriu,P.Baruah,M.E.Bianchi,A.A.Manfredi,and recruitment of inflammatory cells to damaged tissues by form- P. Rovere-Querini, “Requirement of HMGB1 and RAGE for the ing a complex with CXCL12 and signaling via CXCR4,” Journal maturation of human plasmacytoid dendritic cells,” European of Experimental Medicine,vol.209,no.3,pp.551–563,2012. Journal of Immunology,vol.35,no.7,pp.2184–2190,2005. [46] R. R. Kew, M. Penzo, D. M. Habiel, and K. B. Marcu, “The [61]D.Yang,Q.Chen,H.Yang,K.J.Tracey,M.Bustin,andJ.J. IKK𝛼-dependent NF-𝜅B p52/RelB noncanonical pathway is Oppenheim, “High mobility group box-1 protein induces the essential to sustain a CXCL12 autocrine loop in cells migrating migration and activation of human dendritic cells and acts as in response to HMGB1,” Journal of Immunology,vol.188,no.5, an alarmin,” Journal of Leukocyte Biology,vol.81,no.1,pp.59– pp. 2380–2386, 2012. 66, 2007. [47] J. Wu, J. Li, R. Salcedo, N. F. Mivechi, G. Trinchieri, and [62] V.Urbonaviciute, B. G. F¨urnrohr, S. Meister et al., “Induction of A. Horuzsko, “The proinflammatory myeloid cell receptor inflammatory and immune responses by HMGB1-nucleosome TREM-1 controls Kupffer cell activation and development of complexes: implications for the pathogenesis of SLE,” Journal of hepatocellular carcinoma,” Cancer Research,vol.72,no.16,pp. Experimental Medicine, vol. 205, no. 13, pp. 3007–3018, 2008. 3977–3986, 2012. [63] J. S. Park, J. Arcaroli, H.-K. Yum et al., “Activation of gene [48]S.D.Yan,X.Chen,J.Fuetal.,“RAGEandamyloid-𝛽 peptide expression in human neutrophils by high mobility group box neurotoxicity in Alzheimer’s disease,” Nature,vol.382,no.6593, 1protein,”American Journal of Physiology—Cell Physiology,vol. pp. 685–691, 1996. 284,no.4,pp.C870–C879,2003. [49] H. J. Huttunen and H. Rauvala, “Amphoterin as an extracellular [64] K. Li, J. Yang, and X. Han, “Ketamine attenuates sepsis-induced regulator of cell motility: from discovery to disease,” Journal of acute lung injury via regulation of HMGB1-RAGE pathways,” Internal Medicine,vol.255,no.3,pp.351–366,2004. International Immunopharmacology,vol.34,pp.114–128,2016. [50] M. A. Hofmann, S. Drury, C. Fu et al., “RAGE mediates a [65] L. Zhu, Z. Zhang, L. Zhang et al., “HMGB1-RAGE signaling novel proinflammatory axis: a central cell surface receptor for pathway in severe preeclampsia,” Placenta,vol.36,no.10,pp. S100/calgranulin polypeptides,” Cell,vol.97,no.7,pp.889–901, 1148–1152, 2015. 1999. [66] M. B. Manigrasso, J. Juranek, R. Ramasamy, and A. M. Schmidt, [51]O.Hori,J.Brett,T.Slatteryetal.,“Thereceptorforadvanced “Unlocking the biology of RAGE in diabetic microvascular glycation end products (RAGE) is a cellular binding site for complications,” Trends in Endocrinology and Metabolism,vol. amphoterin: mediation of neurite outgrowth and co-expression 25, no. 1, pp. 15–22, 2014. of rage and amphoterin in the developing nervous system,” Journal of Biological Chemistry,vol.270,no.43,pp.25752– [67]G.P.Sims,D.C.Rowe,S.T.Rietdijk,R.Herbst,andA.J.Coyle, 25761, 1995. “HMGB1 and RAGE in inflammation and cancer,” Annual Review of Immunology, vol. 28, pp. 367–388, 2010. [52] A. Taguchi, D. C. Blood, G. Del Toro et al., “Blockade of RAGE-amphoterin signalling suppresses tumour growth and [68]J.Zhang,J.-S.Zhu,Z.Zhou,W.-X.Chen,andN.-W.Chen, metastases,” Nature,vol.405,no.6784,pp.354–360,2000. “Inhibitory effects of ethyl pyruvate administration on human gastric cancer growth via regulation of the HMGB1-RAGE and [53]H.Yang,H.Wang,C.J.Czura,andK.J.Tracey,“Thecytokine Akt pathways in vitro and in vivo,” Oncology Reports,vol.27,no. activity of HMGB1,” Journal of Leukocyte Biology,vol.78,no.1, 5, pp. 1511–1519, 2012. pp.1–8,2005. [69] M. A. Ullah, Z. Loh, W. J. Gan et al., “Receptor for advanced [54]J.Heijmans,N.V.J.A.B¨uller, E. Hoff et al., “Rage signalling glycation end products and its ligand high-mobility group box-1 promotes intestinal tumourigenesis,” Oncogene,vol.32,no.9, mediate allergic airway sensitization and airway inflammation,” pp. 1202–1206, 2013. Journal of Allergy and Clinical Immunology,vol.134,no.2,pp. [55] R. Kang, D. Tang, N. E. Schapiro et al., “The receptor for 440–450.e3, 2014. advanced glycation end products (RAGE) sustains autophagy [70] S. Akira, M. Yamamoto, and K. Takeda, “Role of adapters in and limits apoptosis, promoting pancreatic tumor cell survival,” Toll-like receptor signalling,” Biochemical Society Transactions, Cell Death and Differentiation,vol.17,no.4,pp.666–676,2010. vol.31,no.3,pp.637–642,2003. [56]D.Tang,R.Kang,K.M.Livesey,H.J.Zeh,andM.T.Lotze, “High mobility group box 1 (HMGB1) activates an autophagic [71] S. Akira and K. Takeda, “Toll-like receptor signalling,” Nature response to oxidative stress,” Antioxidants and Redox Signaling, Reviews Immunology,vol.4,no.7,pp.499–511,2004. vol. 15, no. 8, pp. 2185–2195, 2011. [72] M. Yu, H. Wang, A. Ding et al., “HMGB1 signals through toll- [57] H. J. Huttunen, C. Fages, and H. Rauvala, “Receptor for like receptor (TLR) 4 and TLR2,” Shock,vol.26,no.2,pp.174– advanced glycation end products (RAGE)-mediated neurite 179, 2006. outgrowth and activation of NF-𝜅Brequirethecytoplasmic [73]S.Ivanov,A.-M.Dragoi,X.Wangetal.,“Anovelrolefor domain of the receptor but different downstream signaling HMGB1 in TLR9-mediated inflammatory responses to CpG- pathways,” The Journal of Biological Chemistry,vol.274,no.28, DNA,” Blood,vol.110,no.6,pp.1970–1981,2007. pp. 19919–19924, 1999. [74] J. S. Park, D. Svetkauskaite, Q. He et al., “Involvement of toll-like [58] R. Palumbo, B. G. Galvez, T. Pusterla et al., “Cells migrating to receptors 2 and 4 in cellular activation by high mobility group sites of tissue damage in response to the danger signal HMGB1 box 1 protein,” The Journal of Biological Chemistry,vol.279,no. require NF-𝜅Bactivation,”TheJournalofCellBiology,vol.179, 9, pp. 7370–7377,2004. no.1,pp.33–40,2007. [75]H.Yang,H.S.Hreggvidsdottir,K.Palmbladetal.,“Acritical [59] C. J. Treutiger, G. E. Mullins, A.-S. M. Johansson et al., cysteine is required for HMGB1 binding to toll-like receptor 4 “High mobility group 1 B-box mediates activation of human and activation of macrophage cytokine release,” Proceedings of endothelium,” Journal of Internal Medicine,vol.254,no.4,pp. the National Academy of Sciences of the United States of America, 375–385, 2003. vol.107,no.26,pp.11942–11947,2010. Journal of Diabetes Research 9

[76] H. Yang, P.Lundb¨ack, L. Ottosson et al., “Redox modification of [91]K.J.Strissel,Z.Stancheva,H.Miyoshietal.,“Adipocyte cysteine residues regulates the cytokine activity of high mobility death, adipose tissue remodeling, and obesity complications,” group box-1 (HMGB1),” Molecular Medicine,vol.18,no.2,pp. Diabetes,vol.56,no.12,pp.2910–2918,2007. 250–259, 2012. [92] M. A. McArdle, O. M. Finucane, R. M. Connaughton, A. M. [77]J.H.Youn,M.S.Kwak,J.Wuetal.,“Identificationof McMorrow, and H. M. Roche, “Mechanisms of obesity-induced lipopolysaccharide-binding peptide regions within HMGB1 inflammation and insulin resistance: insights into the emerging and their effects on subclinical endotoxemia in a mouse model,” role of nutritional strategies,” Frontiers in Endocrinology,vol.4, European Journal of Immunology,vol.41,no.9,pp.2753–2762, article 52, 2013. 2011. [93] S. Galic, J. S. Oakhill, and G. R. Steinberg, “Adipose tissue as [78] J. H. Youn, Y. J. Oh, E. S. Kim, J. E. Choi, and J.-S. Shin, an endocrine organ,” Molecular and Cellular Endocrinology,vol. “High mobility group box 1 protein binding to lipopolysac- 316, no. 2, pp. 129–139, 2010. charide facilitates transfer of lipopolysaccharide to CD14 and [94] P. Bhargava and C.-H. Lee, “Role and function of macrophages enhances lipopolysaccharide-mediated TNF-𝛼 production in in the metabolic syndrome,” Biochemical Journal,vol.442,no. human monocytes,” The Journal of Immunology,vol.180,no.7, 2, pp. 253–262, 2012. pp. 5067–5074, 2008. [95]C.N.Lumeng,J.B.Delproposto,D.J.Westcott,andA.R. [79]S.Kim,S.YoungKim,J.P.Pribisetal.,“Signalingofhigh Saltiel, “Phenotypic switching of adipose tissue macrophages mobility group box 1 (HMGB1) through toll-like receptor 4 in with obesity is generated by spatiotemporal differences in macrophages requires CD14,” Molecular Medicine,vol.19,no.1, macrophage subtypes,” Diabetes,vol.57,no.12,pp.3239–3246, pp. 88–98, 2013. 2008. [80] H. Yang, H. Wang, Z. Ju et al., “MD-2 is required for disulfide [96]X.Prieur,C.Y.L.Mok,V.R.Velagapudietal.,“Differential HMGB1-dependent TLR4 signaling,” Journal of Experimental lipid partitioning between adipocytes and tissue macrophages Medicine,vol.212,no.1,pp.5–14,2015. modulates macrophage lipotoxicity and M2/M1 polarization in [81] E. Latz, A. Schoenemeyer, A. Visintin et al., “TLR9 signals after obese mice,” Diabetes,vol.60,no.3,pp.797–809,2011. translocating from the ER to CpG DNA in the lysosome,” Nature [97]C.N.Lumeng,S.M.DeYoung,J.L.Bodzin,andA.R.Saltiel, Immunology,vol.5,no.2,pp.190–198,2004. “Increased inflammatory properties of adipose tissue macro- [82] K. Takeda and S. Akira, “Toll-like receptors in innate immunity,” phages recruited during diet-induced obesity,” Diabetes,vol.56, International Immunology,vol.17,no.1,pp.1–14,2005. no. 1, pp. 16–23, 2007. [83] J. Tian, A. M. Avalos, S.-Y. Mao et al., “Toll-like receptor 9- [98] J. M. Olefsky and C. K. Glass, “Macrophages, inflammation, and dependent activation by DNA-containing immune complexes insulin resistance,” Annual Review of Physiology,vol.72,pp.219– is mediated by HMGB1 and RAGE,” Nature Immunology,vol.8, 246, 2010. no.5,pp.487–496,2007. [99] E. Maury, L. No¨el, R. Detry, and S. M. Brichard, “In Vitro [84] S.I.Pachydaki,S.R.Tari,S.E.Leeetal.,“UpregulationofRAGE hyperresponsiveness to tumor necrosis factor-𝛼 contributes and its ligands in proliferative retinal disease,” Experimental Eye to adipokine dysregulation in omental adipocytes of obese Research,vol.82,no.5,pp.807–815,2006. subjects,” Journal of Clinical Endocrinology and Metabolism,vol. [85] Y. Yu, L. Yang, J. Lv et al., “The role of high mobility group 94,no.4,pp.1393–1400,2009. box 1 (HMGB-1) in the diabetic retinopathy inflammation and [100] B. Vandanmagsar, Y.-H. Youm, A. Ravussin et al., “The NLRP3 apoptosis,” International Journal of Clinical and Experimental inflammasome instigates obesity-induced inflammation and Pathology,vol.8,no.6,pp.6807–6813,2015. insulin resistance,” Nature Medicine,vol.17,no.2,pp.179–188, [86] M. R. Dasu, S. Devaraj, S. Park, and I. Jialal, “Increased Toll- 2011. Like Receptor (TLR) activation and TLR ligands in recently [101]R.K.Murumalla,M.K.Gunasekaran,J.K.Padhanetal.,“Fatty diagnosed type 2 diabetic subjects,” Diabetes Care,vol.33,no. acids do not pay the toll: effect of SFA and PUFA on human 4, pp. 861–868, 2010. adipose tissue and mature adipocytes inflammation,” Lipids in [87] J. Skrha,ˇ M. Kalousov´a, J. Svarcov´ˇ a et al., “Relationship of Health and Disease,vol.11,article175,2012. soluble RAGE and RAGE ligands HMGB1 and EN-RAGE to [102]T.Arrigo,V.Chirico,V.Salpietroetal.,“High-mobilitygroup endothelial dysfunction in type 1 and type 2 diabetes mellitus,” protein B1: a new biomarker of metabolic syndrome in obese Experimental and Clinical Endocrinology and Diabetes,vol.120, children,” European Journal of Endocrinology,vol.168,no.4,pp. no. 5, pp. 277–281, 2012. 631–638, 2013. [88]S.Hagiwara,H.Iwasaka,A.Hasegawa,H.Koga,andT. [103] V. N. Montes, S. Subramanian, L. Goodspeed et al., “Anti- Noguchi, “Effects of hyperglycemia and insulin therapy on high HMGB1 antibody reduces weight gain in mice fed a high-fat mobility group box 1 in endotoxin-induced acute lung injury in diet,” Nutrition and Diabetes, vol. 5, article no. e161, 2015. aratmodel,”Critical Care Medicine,vol.36,no.8,pp.2407–2413, [104] P. Kanellakis, A. Agrotis, T. S. Kyaw et al., “High-mobility 2008. group box protein 1 neutralization reduces development of diet- [89] K. Tsoyi, H. J. Jang, I. T. Nizamutdinova et al., “Metformin induced atherosclerosis in apolipoprotein E-deficient mice,” inhibits HMGB1 release in LPS-treated RAW 264.7 cells and Arteriosclerosis, Thrombosis, and Vascular Biology,vol.31,no.2, increases survival rate of endotoxaemic mice,” British Journal pp. 313–319, 2011. of Pharmacology,vol.162,no.7,pp.1498–1508,2011. [105] F. Song, C. H. Del Pozo, R. Rosario et al., “RAGE regulates [90]T.Zhang,X.Hu,Y.Cai,B.Yi,andZ.Wen,“Metformin the metabolic and inflammatory response to high-fat feeding protects against hyperglycemia-induced cardiomyocytes injury in mice,” Diabetes,vol.63,no.6,pp.1948–1965,2014. by inhibiting the expressions of receptor for advanced glycation [106] S. E. Kahn, R. L. Hull, and K. M. Utzschneider, “Mechanisms end products and high mobility group box 1 protein,” Molecular linking obesity to insulin resistance and type 2 diabetes,” Nature, Biology Reports,vol.41,no.3,pp.1335–1340,2014. vol.444,no.7121,pp.840–846,2006. 10 Journal of Diabetes Research

[107] S.B.BiddingerandC.R.Kahn,“Frommicetomen:insightsinto [123] M. B¨oni-Schnetzler, S. Boller, S. Debray et al., “Free fatty acids the insulin resistance syndromes,” Annual Review of Physiology, induce a proinflammatory response in islets via the abundantly vol. 68, pp. 123–158, 2006. expressed interleukin-1 receptor I,” Endocrinology,vol.150,no. [108]H.Xu,G.T.Barnes,Q.Yangetal.,“Chronicinflammationinfat 12, pp. 5218–5229, 2009. plays a crucial role in the development of obesity-related insulin [124] K. Eguchi, I. Manabe, Y. Oishi-Tanaka et al., “Saturated fatty resistance,” The Journal of Clinical Investigation, vol. 112, no. 12, acid and TLR signaling link 𝛽 cell dysfunction and islet inf- pp.1821–1830,2003. lammation,” Cell Metabolism,vol.15,no.4,pp.518–533,2012. [109] S.E.Shoelson,L.Herrero,andA.Naaz,“Obesity,inflammation, [125]J.A.Ehses,G.Lacraz,M.-H.Giroixetal.,“IL-1antagonism and insulin resistance,” Gastroenterology,vol.132,no.6,pp. reduces hyperglycemia and tissue inflammation in the type 2 2169–2180, 2007. diabetic GK rat,” Proceedings of the National Academy of Sciences of the United States of America,vol.106,no.33,pp.13998–14003, [110] S. Schenk, M. Saberi, and J. M. Olefsky, “Insulin sensitivity: 2009. modulation by nutrients and inflammation,” Journal of Clinical Investigation,vol.118,no.9,pp.2992–3002,2008. [126] C. Westwell-Roper, D. Nackiewicz, M. Dan, and J. A. Ehses, “Toll-like receptors and NLRP3 as central regulators of pancre- [111] G. S. Hotamisligil, “Endoplasmic reticulum stress and the atic islet inflammation in type 2 diabetes,” Immunology and Cell inflammatory basis of metabolic disease,” Cell,vol.140,no.6, Biology,vol.92,no.4,pp.314–323,2014. pp. 900–917, 2010. [127]S.A.Steer,A.L.Scarim,K.T.Chambers,andJ.A.Corbett, [112] J. Hirosumi, G. Tuncman, L. Chang et al., “Acentral role for JNK “Interleukin-1 stimulates 𝛽-cell necrosis and release of the in obesity and insulin resistance,” Nature,vol.420,no.6913,pp. immunological adjuvant HMGB1,” PLoS Medicine,vol.3,no.2, 333–336, 2002. article no. e17, 2006. [113]A.Jaeschke,M.P.Czech,andR.J.Davis,“Anessentialroleof [128]G.A.Paredes-Juarez,N.M.Sahasrabudhe,R.S.Tjoelkeret the JIP1 scaffold protein for JNK activation in adipose tissue,” al., “DAMP production by human islets under low oxygen and Genes and Development,vol.18,no.16,pp.1976–1980,2004. nutrients in the presence or absence of an immunoisolating- [114] A.Oeckinghaus,M.S.Hayden,andS.Ghosh,“CrosstalkinNF- capsule and necrostatin-1,” Scientific Reports,vol.5,ArticleID 𝜅Bsignalingpathways,”Nature Immunology,vol.12,no.8,pp. 14623, 2015. 695–708, 2011. [129] T. Itoh, T. Nitta, H. Nishinakamura et al., “HMGB1-mediated early loss of transplanted islets is prevented by anti-IL-6R [115]G.Ghosh,V.Y.-F.Wang,D.-B.Huang,andA.Fusco,“NF- antibody in mice,” Pancreas, vol. 44, no. 1, pp. 166–171, 2015. 𝜅B regulation: lessons from structures,” Immunological Reviews, vol.246,no.1,pp.36–58,2012. [130] E. H. Jo, Y. H. Hwang, and D. Y. Lee, “Encapsulation of pancre- atic islet with HMGB1 fragment for attenuating inflammation,” [116] M. Yuan, N. Konstantopoulos, J. Lee et al., “Reversal of obesity- Biomaterials Research, vol. 19, article 21, 2015. and diet-induced insulin resistance with salicylates or targeted 𝛽 [131] D. Kuraya, M. Watanabe, Y. Koshizuka et al., “Efficacy of disruption of Ikk ,” Science,vol.293,no.5535,pp.1673–1677, 𝜅 2001. DHMEQ, a NF- B inhibitor, in islet transplantation: I. HMGB1 suppression by DHMEQ prevents early islet graft damage,” [117] T. Zeng, J. Zhou, L. He et al., “Blocking nuclear factor-kappa Transplantation,vol.96,no.5,pp.445–453,2013. B protects against diet-induced hepatic steatosis and insulin [132]T.Mera,T.Itoh,S.Kitaetal.,“Pretreatmentofdonorisletswith resistance in mice,” PLoS ONE,vol.11,no.3,ArticleIDe0149677, the Na+/Ca2+ exchanger inhibitor improves the efficiency of 2016. islet transplantation,” American Journal of Transplantation,vol. [118] H. Wang, H. Qu, and H. Deng, “Plasma HMGB-1 levels in 13,no.8,pp.2154–2160,2013. subjects with obesity and type 2 diabetes: a cross-sectional study [133] T. Itoh, S. Iwahashi, M. A. Kanak et al., “Elevation of high- in China,” PLOS ONE,vol.10,no.8,ArticleIDe0136564,2015. mobility group box 1 after clinical autologous islet trans- [119] I.Jialal,U.Rajamani,B.Adams-Huet,andH.Kaur,“Circulating plantation and its inverse correlation with outcomes,” Cell pathogen associated molecular pattern—binding proteins and Transplantation,vol.23,no.2,pp.153–165,2014. High Mobility Group Box protein 1 in nascent metabolic [134]Q.Gao,L.L.Ma,X.Gao,W.Yan,P.Williams,andD.P. syndrome: Implications for cellular Toll-like receptor activity,” Yin, “TLR4 mediates early graft failure after intraportal islet Atherosclerosis,vol.236,no.1,pp.182–187,2014. transplantation,” American Journal of Transplantation,vol.10, [120] R. Guzm´an-Ruiz, F. Ortega, A. Rodr´ıguez et al., “Alarmin high- no. 7, pp. 1588–1596, 2010. mobility group B1 (HMGB1) is regulated in human adipocytes [135] C. Talchai, S. Xuan, H. V. Lin, L. Sussel, and D. Accili, in insulin resistance and influences insulin secretion in 𝛽-cells,” “Pancreatic 𝛽 cell dedifferentiation as a mechanism of diabetic International Journal of Obesity,vol.38,no.12,pp.1545–1554, 𝛽 cell failure,” Cell,vol.150,no.6,pp.1223–1234,2012. 2014. [136]S.J.Richardson,A.Willcox,A.J.Bone,A.K.Foulis,andN. [121] X.-L. Chen, X.-D. Zhang, Y.-Y. Li, X.-M. Chen, D.-R. Tang, G. Morgan, “Islet-associated macrophages in type 2 diabetes,” and R.-J. Ran, “Involvement of HMGB1 mediated signalling Diabetologia,vol.52,no.8,pp.1686–1688,2009. pathway in diabetic retinopathy: evidence from type 2 diabetic [137] J. Han, J. Zhong, W. Wei et al., “Extracellular high-mobility rats and ARPE-19 cells under diabetic condition,” British Journal group box 1 acts as an innate immune mediator to enhance of Ophthalmology,vol.97,no.12,pp.1598–1603,2013. autoimmune progression and diabetes onset in NOD mice,” [122]Y.Chen,F.Qiao,Y.Zhao,Y.Wang,andG.Liu,“HMGB1is Diabetes, vol. 57, no. 8, pp. 2118–2127, 2008. activated in type 2 diabetes mellitus patients and in mesangial [138]B.-W.Lee,H.Y.Chae,S.J.Kwon,S.Y.Park,J.Ihm,andS.-H. cells in response to high glucose,” International Journal of Ihm, “RAGE ligands induce apoptotic cell death of pancreatic Clinical and Experimental Pathology,vol.8,no.6,pp.6683– 𝛽-cells via oxidative stress,” International Journal of Molecular 6691, 2015. Medicine,vol.26,no.6,pp.813–818,2010. Journal of Diabetes Research 11

[139] X.-R. Ni, Z.-J. Sun, G.-H. Hu, and R.-H. Wang, “High concen- tration of insulin promotes apoptosis of primary cultured rat ovarian granulosa cells via its increase in extracellular HMGB1,” Reproductive Sciences, vol. 22, no. 3, pp. 271–277, 2015. Hindawi Publishing Corporation Journal of Diabetes Research Volume 2016, Article ID 7592931, 6 pages http://dx.doi.org/10.1155/2016/7592931

Research Article IL-6 Promotes Islet 𝛽-Cell Dysfunction in Rat Collagen-Induced Arthritis

Huan Jin,1 Yaogui Ning,2 Haotong Zhou,1 and Youlian Wang1

1 Department of Rheumatology, Jiangxi Provincial People’s Hospital, Nanchang 330006, China 2Department of Critical Care Medicine, The First Affiliated Hospital of Xiamen University, Xiamen 361003, China

Correspondence should be addressed to Youlian Wang; [email protected]

Received 21 July 2016; Revised 10 October 2016; Accepted 23 October 2016

Academic Editor: Jixin Zhong

Copyright © 2016 Huan Jin et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The aim of this study was to explore the possible mechanism of rheumatoid arthritis- (RA-) related abnormal glucose metabolism. The model of collagen-induced arthritis (CIA) was established by intradermal injection of type II collagen into Wistar rats; complete Freund’s adjuvant injections were used as the control group. Fasting plasma glucose (FBG) was measured by the glucose oxidase method. Fasting insulin (FIns) and the expressions of IL-6 were detected by ELISA. Islet caspase-3 was examined by immunohistochemistry. On day 17 after immunization, FBG of the CIA group showed an elevated FBG value compared with the control group. Meanwhile, the FIns of group CIA was lower when compared with the control group. Interestingly, the inflammatory cytokine IL-6 expression was significantly increased when compared with the control group. As expected, the abnormal glucose metabolism was accompanied by the increased IL-6 expression. Furthermore, in line with the upregulated IL-6 expression, the apoptosis related enzyme caspase-3 was also markedly increased. These data showed that the elevated FBG in CIA may be associated with the reduced FIns level secondary to the overapoptosis of pancreas islet cells induced by IL-6.

1. Introduction Previous studies have defined the relationship between inflammatory factors and insulin resistance [10–12]. In- Rheumatoid arthritis (RA) is a chronic disease characterized creasedlevelsofseruminflammatorymarkerIL-6havebeen by synovitis and autoantibody such as rheumatoid factor (RF) found in patients with type 2 diabetes [12]. In addition, IL- and anti-cyclic citrullinated peptide antibody (anti-CCP) [1, 6 treatment induces beta-cell apoptosis via STAT-3-mediated 2]. Besides the destruction of cartilage and bone, RA also nitric oxide production [13–15]. Recent reports have shown leads to system damage including cardiovascular, pulmonary, that the prevalence of diabetes mellitus (DM) in patients and psychological systems [3, 4]. Blockage of proinflamma- with RA is about 15%–19%, which is significantly higher than tory cytokines activity involved in the chronic inflammation that of the global DM [16, 17]. The fasting plasma glucose of RA, such as tumor necrosis factor-alpha (TNF-𝛼)and (FBG) in RA patients with type 2 diabetes mellitus (T2DM) interleukin-6 (IL-6), was applied in clinical therapy [5]. is higher than that of T2DM patients, suggesting that the IL-6 is a pleiotropic proinflammatory cytokine produced islet 𝛽-cell damage in RA patients complicated with T2DM by synovial cells and endothelial cells [6, 7]. IL-6 favors is more serious [11], and these may be related to their long- T-cell activation, B-cell proliferation, and immunoglobulin term systemic inflammatory status [18, 19]. However, the secretion [8]. Moreover, IL-6 participates in the synthesis of mechanismofRAwith𝛽-cell damage is still uncertain. acute phase proteins in the liver and proliferation and differ- In this study, the model of CIA was established to observe entiation of hematopoietic precursor cells. Its overexpression the expressions of FBG, FIns, IL-6, and islet caspase-3. As can promote the occurrence of RA and other autoimmune expected, the abnormal glucose metabolism was accom- diseases [9]. panied by the increased IL-6 expression. Furthermore, in 2 Journal of Diabetes Research line with increased IL-6 expression, the apoptosis related 2.5. Statistical Analysis. All statistical analyses were per- enzyme caspase-3 was also markedly increased. These data formed using Prism 5 software. Data are presented as the showed that the possible mechanism of RA complicated with mean ± standard error (SD). Group comparisons were abnormal glucose metabolism might result from the 𝛽-cell performed using Student’s 𝑡-test. Paired Student’s 𝑡-test was apoptosis induced by IL-6. used for paired data. A value of 𝑝 < 0.05 was considered as statistically significant. 2. Materials and Methods 3. Results 2.1. Experimental Animals and Reagents. Female Wistar rats, weighing 150–180 g, were purchased from Kaifu district, 3.1. Abnormal Glucose Metabolism in Rat Collagen-Induced Changsha city, East Laboratory Animal Service Department, Arthritis. In recent studies, there has been increased evi- andwerebredinaspecificpathogen-free(SPF)environment. dence of the relationship between glucose metabolism dys- The rats were adaptively bred one week before experiment. function and human autoimmune disease. To explore the ∘ All rats were maintained (20–24 C and 55–60% humid- mechanism of this observation in rheumatoid arthritis, rat ity) under a standard 12-hour light/dark cycle. IL-6 ELISA collagen-induced arthritis model was established. In this kits were purchased from RayBiotech Company (Norcross, study, on day 17 after the primary immunization, the CIA rats USA); insulin ELISA kits were purchased from Wuhan showed mild edema in one or both sides of the hind limbs. Elabscience Biotechnology Co., Ltd. (Wuhan, China); bovine Furthermore, histological analysis showed hyperplasia of the type II collagen (C7806) and complete Freund’s adjuvant synovial membrane and the infiltration of inflammatory cells were purchased from Sigma Company (Shanghai, China); in CIA rats (Figures 1(a) and 1(b)). As expected, a markedly rabbit anti-rat caspase-3 was provided by Abcam Company increased level of FBG was observed in CIA rats when (Cambridge, USA). Two-step immunohistochemical kits and compared with the control group. Meanwhile, a decreased diaminobenzidine (DAB) color liquid were purchased from level of FIns was observed in CIA rats when compared with Beijing Zhong Shan-Golden Bridge Biological Technology the control group (Figure 2). These data suggest abnormal Co., Ltd. (Beijing, China). glucosemetabolisminratcollagen-inducedarthritis.

2.2. Animal Models. Bovine type II collagen (10 mg) was 3.2. Effect of Increased IL-6 Expression on FBG and FIns. dissolved in 5 mL 0.1 mol/L acetic acid solution, maintained Previous studies have defined the relationship between ∘ at 4 C overnight, and set aside. An equal dose of collagen inflammatory cytokines and insulin resistance [15]. IL-6 is solution and complete Freund’s adjuvant was thoroughly a critical inflammatory cytokine in the pathogenesis of RA, mixed and completely emulsified; then, a CII/IFA emulsion and blockage of its activity by IL-6R antibody was applied (1mg/mL)wasfreshlymanufactured.Ratswereinjectedwith in clinical treatment. Here, we also observed an increased emulsionatthebaseofthetailandbackwithinthreepoints expression of IL-6 in CIA rats when compared with control (day 1); animals were boosted with the dose of 0.2 mL CII- (Figure 3). An increased IL-6 level was also observed in IFAemulsionatday7bythesamemethod.Thecontrolgroup patients with type 2 diabetes. Next, to determine the role were treated with complete Freund’s adjuvant according to of IL-6 in the regulation of glucose metabolism in CIA the same protocol. After initial immunization [20], the rats’ rats, the correlation between IL-6 and FBG or FIns was arthritis index (AI) was scored as in a previous report. Blood analyzed. The result met our expectations; a significantly samples were collected from the intraocular adjoined veins of the rats on the 17th day after initial immunization. positive correlation was observed between IL-6 and FBG (Figure 4(a)). In addition, a negative correlation was depicted in IL-6 and FIns (Figure 4(b)). 2.3. Immunohistochemistry. The pancreatic tissue was fixed in 4% paraformaldehyde (pH 7.4) for histological analysis. To detect the expression of caspase-3, the slides were incubated 3.3. IL-6 Promotes Caspase-3 Expression in Pancreas Islet. A ∘ 𝛽 with antibody diluted at 1 : 80 overnight at 4 C, followed critical role of -cell apoptosis has been demonstrated in the by DAB coloration, and stained with hematoxylin dye and development of diabetes. In line with an increased expression dehydrated (transparent). Image-Pro Plus 6.0 was used to of IL-6 (Figure 3), the apoptosis related enzyme caspase- analyze and carefully delineate the scope of each field in the 3 was also markedly increased in CIA rats (Figures 5(a)– islets, to measure the mean absorbance of the islets, as an 5(c)). Furthermore, the cleaved caspase-3 was also detected, index for expression of islets caspase-3. as expected, and the CIA rats showed an increased level when compared with control rats. Previous studies have shown 2.4. Immunoblots. The protein extracted from the pancreatic that IL-6 treatment induces beta-cell apoptosis via STAT- tissue was assessed using an immunoblot assay. Antibodies 3-mediated nitric oxide production. Accordingly, a positive against cleaved caspase-3 were obtained from Cell Signaling correlation between caspase-3 and IL-6 was observed in our Technology (MA, USA). After incubating with secondary study. Taken together, these results suggest that the abnormal antibody conjugated horseradish peroxidase, the immunore- glucosemetabolismmightbeduetothe𝛽-cell apoptosis activity was developed by an ECL system (Pierce). induced by inflammatory cytokine IL-6 (Figure 5(d)). Journal of Diabetes Research 3 Control

(a) CIA

(b)

Figure 1: Establishment of rat collagen-induced arthritis. After day 17 of immunization, the macroscopic change of hind limbs and the histology of joint tissue sections from CIA rats and control rats were analyzed. All data are representative of one out of three independent experiments.

10 FBG 20 FIns

8 ∗ 15 ∗ 6 10

4 (ng/mL) (mmol/L) 5 2

0 0 Control CIA Control CIA Figure 2: Glucose metabolism dysfunction in CIA rats. Blood samples were collected from the intraocular adjoined veins of the rats on day 17 after initial immunization from CIA rats and control rats. FBG and FIns in sera of both groups were determined as described in the Materials ∗ and Methods. The data are presented as mean ± SD (𝑛=6–8/group) and are representative of three independent experiments. 𝑝 < 0.05.

4. Discussion overexpressed proinflammatory cytokines [21]. In this exper- iment, the expression of IL-6 in group CIA was significantly Recent studies have shown that the prevalence of RA higher than that of control group, and the abnormal glucose patients with abnormal glucose metabolism was significantly metabolism was accompanied by the increased IL-6 expres- increased [11, 16], and the decreased insulin sensitivity and sion. Furthermore, in line with increased IL-6 expression, islet 𝛽-cell function due to islet 𝛽-cell excessive apoptosis theapoptosisrelatedenzymecaspase-3wasalsomarkedly in these patients and the mechanism might be related to increased. These data showed that the abnormal glucose 4 Journal of Diabetes Research

IL-6 600 ∗

400 (pg/mL) 200

0 Control CIA Figure 3: Increased IL-6 expression in rat collagen-induced arthritis. Blood samples were collected from the intraocular adjoined veins of the rats on day 17 after initial immunization from CIA rats and control rats. IL-6 in sera was determined by ELISA kit. The data are presented ∗ as mean ± SD (𝑛=6–8/group) and are representative of three independent experiments. 𝑝 < 0.05.

8.0000 20.0000

18.0000 7.0000 16.0000

6.0000 14.0000 FIns (ng/mL) FIns FBG (mmol/L) 12.0000 5.0000 10.0000

4.0000 8.0000

200.0000300.0000 400.0000 500.0000 600.0000 700.0000 200.0000 300.0000 400.0000 500.0000 600.0000 700.0000 IL-6 (pg/mL) IL-6 (pg/mL) (a) (b)

Figure 4: Correlation between IL-6 and FBG and FIns. The sera were harvested after day 17 of immunization. Levels of IL-6, FBG, andFIns were determined. The correlations between IL-6 and these factors in all the rats were analyzed by Spearman correlation test. 𝑝 value < 0.05 was considered as statistically significant. Data are representative of three independent experiments.

metabolisminRAmightresultfromthe𝛽-cell apoptosis IL-6 can also induce the production of IL-1𝛽,IL-2,and induced by IL-6. TNF-𝛼 and enhance the biological effects when combined IL-6, an important proinflammatory cytokine, could with these cytokines [9]. The previous studies have also promote insulin secretion at low concentration of IL-6 [22]. found that TNF-𝛼 produced by adipose tissue can induce Interestingly, the high concentration of IL-6 would damage 𝛽- insulin resistance (IR); these data showed the relationship cells and promote apoptosis [13]. The cytotoxicity of high IL- between inflammatory factors and insulin resistance. Not 6 concentrations could upregulate the expression of cleaved only were IL-6 levels in CIA rats significantly increased, but caspase-3 and phosphorylated p38 mitogen-activated protein also other proinflammatory cytokines like TNF-𝛼 and IFN-𝛾 kinase and phosphorylated nuclear factor-𝜅B, in part via were overexpressed [5]. IL-6 can also be combined with TNF- signal transducers and activators of transcription-3-mediated 𝛼,IL-1𝛽, and induced classical caspase-dependent apoptosis NO production [8]. In addition, IL-6 can increase lipolysis pathway, thereby increasing the apoptosis of 𝛽-cells and in adipocytes as well as the release of free fatty acids, causing a reduction of insulin secretion [14]. These might be thereby damaging the mitochondria and glucose transporter the reason for the high prevalence of diabetes mellitus (DM) 2 (GLUT2) function and insulin sensitivity [23]. Here, in in patients with RA when compared with healthy controls. ourstudy,wealsoobservedelevatedIL-6expressionin Insulin is one of the most important hormones in an CIA, which positively correlated with glucose. Furthermore, organism that regulates blood sugar levels, which is secreted Journal of Diabetes Research 5 CIA Control

(a) 0.6 Caspase-3 ∗ 0.4

0.2 Mean OD

0 Control CIA Control CIA (b) (c) 0.8 1.0

0.8 0.6

0.6 0.4 0.4 IL-6 (ng/mL) IL-6 0.2 0.2 Caspase-3 OD) (mean

0.0 0.0 IL-6 Caspase 3 (d)

Figure 5: Potential of IL-6 in the apoptosis of 𝛽-cells. The pancreas islets were harvested after day 17 of immunization. (a) The apoptosis related gene caspase-3 was analyzed by immunohistochemistry. (b) The expression of caspase-3 was analyzed by OD value. Image-Pro Plus 6.0 was used to delineate the scope of each field in the islets and measure the mean absorbance of the islets. (c) The proteins of pancreas islets were extracted for analyzing cleaved caspase-3 expression by immunoblots assays. Data are representative of three independent experiments (d) Paired Student’s 𝑡-test was used for analyzing the paired data of IL-6 and caspase-3. Data are shown as the mean ± SD (𝑛=6–8/group) ∗ and are representative of three independent experiments. 𝑝 < 0.05.

by pancreatic 𝛽-cells; insulin deficiency can lead to elevated This study suggests the significantly increased FBG and blood glucose levels [21]. As expected, we observed that decreasedFInsinCIArats,whichmightbearesultfrom fasting serum insulin levels were significantly decreased in autoimmune inflammatory mediators such as IL-6. IL-6 CIA. Furthermore, the increased expression of islet caspase- induced abnormal glucose metabolism might be through 3 was also detected. Caspase-3 is an important proapoptotic upregulating caspase-3 expression, because of the positive protein and is the performer of cell apoptosis. When the correlation between IL-6 and apoptosis related enzyme cells are stimulated by the proapoptotic factors, caspase-3 caspase-3. Taken together, these data showed that the elevated canbeactivatedandtheninducetheirapoptosis[24,25]. FBGandreducedFInslevelinCIAmaybeinducedby In addition, IL-6 was positively correlated with caspase-3 the inflammatory cytokines generated by arthritis, which production which leads 𝛽-cell to apoptosis in our study. The is significantly different from the traditional T2DM. The elevated levels of IL-6 inhibiting the insulin secretion of islet treatment of the primary disease in RA combined with DM 𝛽-cells might be through promoting 𝛽-cells apoptosis. can improve glucose metabolism. 6 Journal of Diabetes Research

Competing Interests nitric oxide production,” Diabetes/Metabolism Research and Reviews,vol.27,no.8,pp.813–819,2011. The authors declare no competing interests regarding the [14] S. Kim, K. Kim, K. Suk, Y.Kim, S. H. Oh, and M. Lee, “Apoptosis publication of this paper. of human islet cells by cytokines,” Immune Network,vol.12,no. 3, pp. 113–117, 2012. Authors’ Contributions [15] C. A. Delaney, D. Pavlovic, A. Hoorens, D. G. Pipeleers, and D. L. Eizirik, “Cytokines induce deoxyribonucleic acid strand Huan Jin and Yaogui Ning contributed equally to this work. breaks and apoptosis in human pancreatic islet cells,” Endo- crinology,vol.138,no.6,pp.2610–2614,1997. [16] J. F. Simard and M. A. Mittleman, “Prevalent rheumatoid Acknowledgments arthritis and diabetes among NHANES III participants aged 60 and older,” The Journal of Rheumatology,vol.34,no.3,pp.469– This project was supported by the Jiangxi Provincial Science 473, 2007. and Technology Department (20112BBG70040). [17] O.¨ N. Pamuk, E. Unl¨ u,¨ and N. C¸akir,“Roleofinsulinresistance in increased frequency of atherosclerosis detected by carotid References ultrasonography in rheumatoid arthritis,” Journal of Rheuma- tology, vol. 33, no. 12, pp. 2447–2452, 2006. [1] M. Kourilovitch, C. Galarza-Maldonado, and E. Ortiz-Prado, [18] S. Mavrogeni, T. Dimitroulas, C. Bucciarelli-Ducci et al., “Diagnosis and classification of rheumatoid arthritis,” Journal “Rheumatoid arthritis: an autoimmune disease with female of Autoimmunity,vol.48-49,pp.26–30,2014. preponderance and cardiovascular risk equivalent to diabetes [2] J. T. Cassidy, J. E. Levinson, J. C. Bass et al., “A study of clas- mellitus: Role of cardiovascular magnetic resonance,” Inflam- sification criteria for a diagnosis of juvenile rheumatoid arthri- mation and Allergy-Drug Targets,vol.13,no.2,pp.81–93,2014. tis,” Arthritis and Rheumatism,vol.29,no.2,pp.274–281,1986. [19] C. M. Bartels, J. M. Saucier, C. T. Thorpe et al., “Monitoring [3]H.Maradit-Kremers,P.J.Nicola,C.S.Crowson,K.V.Ballman, diabetes in patients with and without rheumatoid arthritis: A and S. E. Gabriel, “Cardiovascular death in rheumatoid arthritis: Medicare Study,” Arthritis Research and Therapy,vol.14,no.4, a population-based study,” Arthritis & Rheumatism,vol.52,no. article R166, 2012. 3, pp. 722–732, 2005. [20] D. Salvemini, E. Mazzon, L. Dugo et al., “Amelioration of joint [4]A.A.Wagan,T.H.Mahmud,A.Rasheed,Z.A.Zafar,A.U. disease in a rat model of collagen-induced arthritis by M40403, Rehman, and A. Ali, “Cardiovascular risk score in Rheumatoid a superoxide dismutase mimetic,” Arthritis & Rheumatism,vol. Arthritis,” Pakistan Journal of Medical Sciences,vol.32,no.3,pp. 44,no.12,pp.2909–2921,2001. 534–538, 2016. [21] H. Khodabandehloo, S. Gorgani-Firuzjaee, G. Panahi, and [5] I. B. McInnes, C. D. Buckley, and J. D. Isaacs, “Cytokines in R. Meshkani, “Molecular and cellular mechanisms linking rheumatoid arthritis-shaping the immunological landscape,” inflammation to insulin resistance and 𝛽-cell dysfunction,” Nature Reviews Rheumatology,vol.12,no.1,pp.63–68,2016. Translational Research,vol.167,no.1,pp.228–256,2016. [6]T.Koga,S.Yamasaki,K.Migitaetal.,“Post-transcriptional [22] J. Osorio,´ “Diabetes: IL-6 mediates the protective effects of regulation of IL-6 production by Zc3h12a in fibroblast-like exercise on 𝛽 cells,” Nature Reviews Endocrinology, vol. 11, no. synovial cells,” Clinical and Experimental Rheumatology,vol.29, 4, p. 193, 2015. no. 6, pp. 906–912, 2011. [23] P. H. Dessein, B. I. Joffe, and A. E. Stanwix, “Effects of disease [7]M.Mihara,Y.Moriya,T.Kishimoto,andY.Ohsugi,“Inter- modifying agents and dietary intervention on insulin resistance leukin-6 (IL-6) induces the proliferation of synovial fibroblastic and dyslipidemia in inflammatory arthritis: a pilot study,” cells in the presence of soluble IL-6 receptor,” British Journal of Arthritis Research,vol.4,no.6,articleR12,2002. Rheumatology,vol.34,no.4,pp.321–325,1995. [24] P. Augstein, J. Bahr, G. Wachlin et al., “Cytokines activate [8] M. F. Neurath and S. Finotto, “IL-6 signaling in autoimmunity, caspase-3 in insulinoma cells of diabetes-prone NOD mice chronic inflammation and inflammation-associated cancer,” directly and via upregulation of Fas,” Journal of Autoimmunity, CytokineandGrowthFactorReviews,vol.22,no.2,pp.83–89, vol.23,no.4,pp.301–309,2004. 2011. [25] S. Reddy, J. Bradley, and J. M. Ross, “Immunolocalization of [9] L. H. Calabrese and S. Rose-John, “IL-6 biology: implications caspase-3 in pancreatic islets of NOD mice during cyclophos- for clinical targeting in rheumatic disease,” Nature Reviews phamide-accelerated diabetes,” Annals of the New York Academy Rheumatology,vol.10,no.12,pp.720–727,2014. of Sciences,vol.1005,pp.192–195,2003. [10] S. Manrique-Arija, I. Urena,˜ P. Valdivielso et al., “Insulin resistance and levels of adipokines in patients with untreated early rheumatoid arthritis,” Clinical Rheumatology,vol.35,no. 1,pp.43–53,2016. [11] D. Shahin, E. Eltoraby, A. Mesbah, and M. Houssen, “Insulin resistance in early untreated rheumatoid arthritis patients,” Clinical Biochemistry,vol.43,no.7-8,pp.661–665,2010. [12] B. Feve` and J.-P. Bastard, “The role of interleukins in insulin resistance and type 2 diabetes mellitus,” Nature Reviews Endocrinology,vol.5,no.6,pp.305–311,2009. [13]Y.S.Oh,Y.-J.Lee,E.Y.Park,andH.-S.Jun,“Interleukin- 6 treatment induces beta-cell apoptosis via STAT-3-mediated Hindawi Publishing Corporation Journal of Diabetes Research Volume 2016, Article ID 2846570, 9 pages http://dx.doi.org/10.1155/2016/2846570

Review Article Considerations for Defining Cytokine Dose, Duration, and Milieu That Are Appropriate for Modeling Chronic Low-Grade Inflammation in Type 2 Diabetes

Craig S. Nunemaker1,2

1 Diabetes Institute, Heritage College of Osteopathic Medicine, Ohio University, Athens, OH, USA 2Department of Biomedical Sciences, Heritage College of Osteopathic Medicine, Ohio University, Athens, OH, USA

Correspondence should be addressed to Craig S. Nunemaker; [email protected]

Received 22 July 2016; Accepted 25 September 2016

Academic Editor: Akira Mima

Copyright © 2016 Craig S. Nunemaker. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Proinflammatory cytokines have been implicated in the pathophysiology of both type 1 diabetes (T1D) and type 2 diabetes (T2D). T1D is an autoimmune disease involving the adaptive immune system responding to pancreatic beta-cells as antigen-presenting cells. This attracts immune cells that surround pancreatic islets (insulitis) and secrete cytokines, such as IL-1beta, IFN-gamma, and TNF-alpha, in close proximity to pancreatic beta-cells. In contrast, there is little evidence for such a focused autoimmune response in T2D. Instead, the innate immune system, which responds to cellular damage and pathogens, appears to play a key role. There are three major sources of proinflammatory cytokines thatmay impact islet/beta-cell function in T2D: (1) from islet cells, (2) from increased numbers of intraislet macrophages/immune cells, and (3) from increased circulating levels of proinflammatory cytokines due to obesity, presumably coming from inflamed adipose tissue. These differences between T1D and T2D are reflected by significant differences in the cytokine concentration, duration, and milieu. This review focuses on chronic versus acute cytokine action, cytokine concentrations, and cytokine milieu from the perspective of the pancreatic islet in T2D. We conclude that new cytokine models may be needed to reflect the pathophysiology of T2D more effectively than what are currently employed.

1. Introduction exposure that may more accurately reflect the pancreatic environment in T2D, with particular emphasis on the islet. Nearly a century has passed since insulin codiscoverers Frederick Banting and Charles Best observed the first patient to receive insulin therapy in 1922, instantly transforming the 2. Cytokines in T1D versus T2D prognosis of type 1 diabetes (T1D) from a death sentence to a manageable disease. However, the root cause of T1D is still There are notable similarities and differences in the action not known, but the processes that lead to the destruction of of cytokines in the development of T1D versus T2D [9, 10], insulin-producing pancreatic beta-cells have been fairly well as shown in Figure 1 and described below. In T1D, beta-cells elucidated. T1D is now considered an autoimmune disease are the direct target of an autoimmune invasion beginning that causes the gradual, systematic destruction of pancreatic with peri-insulitis and ending in beta-cell death [10–12]. beta-cells within the islets of Langerhans. Proinflammatory In T2D, metabolic stress is thought to activate the innate cytokines play a prominent role in the pathophysiology of immune system, resulting in a chronic inflammatory state T1D [1–6], but increasing evidence also suggests a significant marked by increased cytokines, increased islet-associated role for cytokines in islet dysfunction in T2D as well [6– macrophages, and beta-cell apoptosis [10, 13, 14]. In T1D, 8]. In this review, I address the substantial differences in autoimmunity is the prime effector of beta-cell destruction. the inflammatory environment of the pancreatic islet inT1D In T2D, a number of metabolic stress factors related to excess versus T2D and then consider alternative models of cytokine nutrients are thought to contribute to beta-cell decline and 2 Journal of Diabetes Research

Type 1 diabetes Type 2 diabetes 3. Considerations for a Model of Cytokine Antigen Metabolic Action Consistent with T2D Trigger presentation stress A number of proinflammatory cytokines are elevated in the Autoimmune Islet-derived and/or general circulation of obese individuals compared to lean Process response circulating cytokines individuals, and these increased levels are associated with an increased risk of developing T2D [45–53]. This low- grade systemic inflammation may also play a direct role in triggering beta-cell dysfunction, particularly in T2D. We have Islet infiltration Increased previously reported that cytokines can directly affect aspects and focal numbers of of islet function in rodent islets at circulating concentrations cytokine macrophages in vitro [54–56]. However, these effects, which are primarily release in islets related to intracellular calcium handling, do not appear to impact cell death rate or insulin secretion in normal islets [55]. As shown in Figure 2, the “T2D-like” cytokines, in the Beta-cell Islet dysfunction, Result destruction beta-cell death(?) range of circulating levels in the blood, do not increase cell death or decrease insulin secretion significantly. The “T1D- Figure 1: A flowchart of very basic differences in the patho- like” concentrations of cytokines necessary to induce cell physiology of T1D and T2D. Each disease is described by the death are at least 50-foldhigherthan what is observed in low- immune-mediated triggers, processes, and results with respect to the grade systemic inflammation in T2D (see also similar pub- pancreatic beta-cell/islet. lished data in [55]). The 5 ng/mL concentration of IL-1beta commonly found in the literature is 10 times greater than the T1D-like concentration and 500 times greater than the T2D- like concentration shown in Figure 2. This suggests that the destruction including glucotoxicity [15–17], lipotoxicity [15– use of high concentrations of death-inducing cytokines is not 17], oxidative stress [16, 18], endoplasmic reticulum (ER) necessarily appropriate for recreating an islet environment stress [19, 20], amyloid deposition [21], and inflammation relatedtoT2D.Tobuildanappropriateinvitromodelof [8,13].Someevidencesuggeststhatallofthesefactorsmay chronic low-grade inflammation that can reasonably mimic even be connected through inflammasome activation [22– the islet environment in T2D in vivo, several questions must 24]. To bring this discussion full circle, emerging research be addressed: suggests that some of the putative factors causing beta-cell dysfunction in T2D, most notably ER stress and/or oxidative (1) What cytokines and chemokines are elevated in the stress, may play a role in antigen presentation to trigger the systemic circulation in obesity and T2D that can autoimmune response in T1D [25–27]. negatively impact islet function? The key proinflammatory cytokines in T1D are interleu- (2) What immune cells are found in or near islets that can kin- (IL-) 1beta, tumor necrosis factor alpha (TNF-alpha), secrete cytokines in close proximity? and interferon-gamma (IFN-gamma). IL-1beta is produced (3) What cytokines are secreted by islet cells themselves by monocytes (leukocytes), macrophages, and other immune in response to damage or stress? cells. TNF-alpha is produced by macrophages, lymphoid, stromal cells, and many other cell types and can exist in (4) What concentration and duration of exposure mimic membrane-bound as well as in soluble forms [28, 29]. IFN- chronic low-grade inflammation? gamma is primarily produced by natural killer cells and certain types of T cells [30], although macrophages can Each of these considerations is discussed in Sections 3.1–3.4. also produce IFN-gamma under certain conditions, such as exposure to IL-12 and IL-18 [31, 32]. Infiltrating immune 3.1. Circulating Cytokines and Chemokines. Circulating levels cells are thought to secrete these three cytokines in close of several cytokines and chemokines have been identified as proximity to the beta-cell at high concentrations in T1D, potential risk factors for developing T2D, including TNF- and these cytokines interact synergistically to inflict cytotoxic alpha, C-reactive protein (CRP), monocyte chemoattractant effects [3, 33–36]. Although rodent islets may succumb to protein-1 (MCP-1), IL-1beta, and IL-6 [45–53]. Chief among high concentrations of any of these cytokines, classic studies these risk factors are IL-1beta and IL-6 [57]. IL-1beta at demonstrated that combinations of these cytokines produced high enough concentrations can destroy beta-cells [58, 59]. more consistent cytotoxic effects on human islets [3, 4, 34, Antagonizing the actions of IL-1beta has been shown to 37, 38]. In T2D, the metabolic stress of obesity is thought have clinical value in partially mitigating the symptoms of to elevate cytokine production in adipose tissue [39–43] T2D[60–63],suggestingthatIL-1betaplaysaroleinthe and other organs to a lesser extent [44], resulting in an pathophysiology of the disease. Numerous studies have also overall chronic low-grade inflammatory state. In the next associated increased levels of IL-6 with the development section, key parameters will be examined to define a model of of type 2 diabetes [64–67]. However, IL-6 has also been islet exposure to cytokines that may resemble the pancreatic associated with beneficial effects of exercise (reviewed in environment more closely during the development of T2D. [68]) and with improving islet function [69, 70], resulting Journal of Diabetes Research 3

Cytokine-induced cell death 1050 control mice, suggesting that low-grade systemic inflamma- tion develops early in the disease process [74]. In addition, many other cytokines and chemokines are 850 ∗ increased in obesity including leptin, resistin, IL-7, IL-8, retinol binding protein- (RBP-) 4, plasminogen activator inhibitor- (PAI-) 1, chemokine- (C-X-C motif-) ligand 5 650 (CXCL5), visfatin, chemerin, and vaspin [75, 76]. Other cytokines are decreased with increased obesity including adiponectin, IL-10, and omentin [75, 76]. We conducted a

PI fluorescence (a.u.) PI fluorescence 450 32-plex cytokine detection array of mouse blood serum from leptin-receptor-deficient male BKS.Cg-Dock7m+/+ Leprdb/J (db/db) mice and heterozygous controls at different ages to 250 identify additional cytokines and chemokines [77]. Many of Untreated T2D-like cytokines T1D-like cytokines the 32 cytokines tested were above or below the sensitivity Cytokine dose for appropriate detection or were highly variable. However, CXCL1 and CXCL5 were increased significantly in serum (a) at the onset of obesity and T2D. We further showed that 70 Cytokine-induced reduction in GSIS exposure to circulating concentrations of these chemokines synergistically produced mild inhibition of islet function and ∗ that expression of CXCL1 and CXCL5 increased markedly in 60 ∗ islets in response to low-grade inflammation [77]. 50 Regarding the classic trio of T1D cytokines, IL-1beta appears to also play a role in T2D, but TNF-alpha and IFN- 40 gamma may or may not be as prominent. In our studies, we did not observe elevated serum levels of TNF-alpha in prediabetic mice (IFN-gamma was not examined), and TNF- 30 alpha does not appear to have the same degree of impact as Insulin (ng/mL) Insulin IL-1beta and IL-6 on the development of T2D in humans [47]. 20 IFN-gamma appears to be elevated in obesity and may play a role in islet dysfunction in T2D [78, 79]. However, at present 10 there is not sufficient evidence that circulating levels of IFN- gamma impact islet function or that IFN-producing immune 0 cells are localized within islets in models of T2D. Additional Untreated T2D-like cytokines T1D-like cytokines studies are needed to determine the full milieu of obesity- Cytokine dose associated cytokines and the extent to which this milieu of 3 G circulating cytokines can impact islet function. 28 G (b) 3.2. Cytokines Produced by Macrophages and Other Immune Figure 2: Islet viability and function following overnight treatment Cells within Islets. In addition to the chronic inflammatory with different cytokine concentrations. (a) Cell death measurements state marked by increased circulating cytokines, signs of by propidium iodide (PI) were made on islets following treatment increased inflammation in T2D have also been observed with Roswell Park Memorial Institute 1640 media (untreated), within the pancreas itself. Specifically, increased numbers of 10 pg/mL IL-1beta + 20 pg/mL TNF-alpha + 200 pg/mL IFN- islet-associated macrophages have been reported in rodent gamma (T2D-like), or 500 pg/mL IL-1beta + 1000 pg/mL TNF-alpha models of T2D and in humans with T2D [80, 81]. In the + 10,000 pg/mL IFN-gamma (T1D-like). (b) Glucose-stimulated db/db mouse model of diabetes, a 4-fold increase in M1 type insulin secretion (GSIS) in 3 mM glucose (3G, open bars) and macrophages was reported as early as 8 weeks of age [82]. 28 mM glucose (28G, filled bars) following cytokine treatments Macrophages are involved in adaptation and cellular repair ∗𝑃 < listed in (a). value 0.05. as well as in the discrete removal of dead and dying cells. Thus, increased islet presence of macrophages could merely be an indicator of more dying cells that must be cleared away. in a complex and evolving view over the role(s) of IL-6 in However, macrophages could also be active instigators of T2D [71–73]. A prospective study focused primarily on links destruction by secreting various proinflammatory cytokines between nutrition and cancer generated an intriguing report in close proximity to islet cells. The potential of macrophage that identified the combination of both IL-1beta and IL-6 instigation of islet dysfunction has been confirmed by stud- as key cocontributors to the development of T2D [47]. We ies showing that palmitate causes macrophage accumula- reported that serum levels of IL-1beta and IL-6 in diabetes- tion, increased cytokine/chemokine production within islets, prone mice at an age before hyperglycemia developed and islet dysfunction, but islet dysfunction is prevented were 2-3 times higher than for age-matched heterozygous when macrophage accumulation is blocked [83, 84]. These 4 Journal of Diabetes Research findings suggest that macrophages play an important role or negative effects on the beta-cell. In addition, synergistic in inflammation caused by metabolic stress. In addition activity among multiple cytokines can alter or amplify sig- to macrophage involvement, recent work also suggests that naling pathways [3, 4, 107, 108], adding an additional layer clusters of differentiation- (CD-) 45-positive leukocytes of complexity to cytokine action. are also increased in islets from T2D donors, suggesting a possible role of the adaptive immune system in T2D [85, 86]. 4. Cytokine Susceptibility and Signaling in T2D-Prone Islets 3.3. Islet-Derived Cytokines. There is also evidence that cells Our recent work has focused on elucidating possible dif- within islets are capable of producing proinflammatory ferences in response to chronic low-grade inflammation cytokines. Chronic high levels of glucose, for example, can between islets from diabetes-prone and normal (nondiabe- stimulate IL-1beta production in beta-cells [87]. Exposure tes-prone) environments. Male db/db mice and heterozygous to certain cytokines can also stimulate islet expression of controls were used at 4-5 weeks of age, an age at which the cytokines and chemokines. For example, treatment with capacity for glucose-stimulated insulin secretion and intra- IFN-gamma + TNF-alpha increases islet expression of IL- cellular calcium responses are similar for both mouse strains 15, interferon-gamma-induced protein- (IP-) 10, CXCL9, [74]. The combination of IL-1beta and IL-6 significantly CXCL11, and chemokine (C-C motif) ligand- (CCL-) 20 [88]. impaired islet function in ways that either alone could not IL-1beta exposure induces expression of several neutrophil- [74]. Further, islets isolated from prediabetic db/db mice and attracting proteins including MCP-1 [89]. Exposure to free exposed in vitro to IL-1beta + IL-6 overnight at approximately fatty acids, palmitate in particular, can induce islet expression circulating levels (in mice) caused a significant decrease in of numerous cytokines and chemokines, including IL-1beta, glucose-stimulated insulin secretion, a significant increase in TNF-alpha, IL-6, IL-8, CXCL1, and MCP-1 [83, 84, 90]. ER stress markers (nitric oxide synthase 2 (NOS2), 78-kDa The importance of intraislet cytokine production is that glucose-regulated protein (GRP78), activating transcription cytokine concentrations within islets may be orders of magni- factor 4 (ATF4), and DNA Damage Inducible Transcript tude higher when cytokines are produced by islet cells and/or 3 (DDIT3), also known as C/EBP homologous protein resident immune cells within the islet in comparison to cir- (CHOP)), and increased cell death; these cytokines produced culating levels. It should be noted that the increased cytokine no such effect in heterozygous control islets [74]. Further, expression in many of these studies could not be attributed to when nondiabetic control mice were implanted with subcu- any specific cell type(s) within the islet. Thus, it is possible that taneous osmotic mini-pumps containing IL-1beta + IL-6 to immune cells within the islet could be responsible for most mimic the serum increases found in prediabetic db/db mice, or all of the increased expression of these cytokines, rather we were able to produce 2-3-fold increases in circulating than insulin-producing beta-cells, glucagon-producing alpha cytokines, thus reproducing cytokine levels observed in low- cell, or other endocrine cell types in the islet. Thus, intraislet grade inflammation in obesity. The increased circulating secretion of cytokines, regardless of the source, may markedly levels of IL-1beta + IL-6 were insufficient to impact physi- increase the potential concentration of cytokine exposure in ology in these normal mice. When compared with saline- islet cells. implanted controls, however, isolated islets from cytokine- pump mice showed deficiencies in calcium handling and 3.4. Cytokine Dose, Duration, and Milieu. The net effect of insulin secretion that were similar to cytokine effects on islets cytokine exposure on tissue depends markedly on factors in vitro [74]. Finally, these low level cytokine exposures could such as cytokine concentration, duration of cytokine expo- also impair human islet function [74]. These results in mice sure, and the combination of cytokines involved (also called suggest that mild increases in circulating cytokines may be the cytokine milieu [91, 92]). IL-1beta, for example, causes sufficient to trigger islet dysfunction leading to islet failure in islet dysfunction or cell death in a number of studies of genetically susceptible individuals [74]. chronic exposure [37, 58], and blockade of the IL-1 receptor The cytokine signaling pathways may also differ in con- mitigates many of these effects [60, 93, 94]. However, IL- ditions of low-grade inflammation when compared to classic 1beta can also enhance insulin secretion [95–97], and recent cytokine cocktails associated with T1D. A microarray study work even suggests IL-1beta may play a key role in islet of islet gene expression in response to low concentrations compensation for nutrient overload in the early stages of 10 pg/mL IL-1beta + 20 pg/mL IL-6 produced a large number metabolic disorders [98]. In one early series of studies, both of gene hits that were apparently not associated with canon- positive and negative impacts of IL-1beta were described ical signaling pathways for IL-1beta and IL-6 [109]. Several based on differences and dose and duration of exposure highly cytokine-induced genes related to proteins involved [99–101]. In the context of T2D, the duration of low-grade with iron regulation, including Steap4 (six-transmembrane inflammation may be a key factor, as cytokines could have epithelial antigen of prostate 4), lipocalin-2 (Lcn-2), and compensatory/protective effects initially that become delete- hepcidin antimicrobial peptide (Hamp). Moreover, seven rious under chronic conditions. Thus, the dose and duration cytokine-sensitive genes were identified for single nucleotide of cytokine exposure [98, 101, 102], species being tested [103– polymorphisms related to the acute insulin response to 105], and milieu of interacting cytokines [3, 4, 95, 105, 106] glucose (AIRg, a test of islet function) in a genome-wide are all important factors that contribute to both positive association scan of a population with a high prevalence Journal of Diabetes Research 5

Putative sources of cytokines affecting islets in T2D IL-1beta and 10–20 pg/mL IL-6 as being sufficient to promote Circulating islet dysfunction in T2D. Immune-cell-derived cytokines Islet-cell-derived To produce these levels of low-grade systemic inflam- cytokines cytokines mation in vivo is quite possible. Osmotic mini-pumps are ideally designed to produce the small and precise changes in the circulating levels of cytokines that define chronic inflammation. We induced low-grade inflammation in vivo Beta-cell in normal healthy mice loading ALZET osmotic mini-pumps (model 1007d, rate 0.5 𝜇L/h for 7 days) with 32 𝜇g/mL for IL- 1B plus 4 𝜇g/mLforIL-6insalineorsalineonlycontrols. We inserted each pump subcutaneously into an incision Figure 3: A basic depiction of the putative sources of cytokines affecting islets in T2D. at the nape of the neck to cause low-grade inflammation in mice without significantly impacting blood glucose or insulin levels. ELISA assays confirmed that serum levels of IL-1B and IL-6 were approximately doubled in mice for T2D [109]. The seven genes were Arap3 (ArfGAP with treated with cytokine mini-pumps versus saline control RhoGAP domain, ankyrin repeat, and PH domain 3), F13a1 pumps. Islets isolated from cytokine-pump-treated mice (coagulation factor XIII, A1 polypeptide), Klhl6 (kelch-like showed impaired function compared to saline-pump con- protein 6), Nid1 (nidogen 1), Pamr1 (peptidase domain- trols. These findings are detailed in [74] and provide a containing protein associated with muscle regeneration 1), method for elevating specific circulating cytokines related Ripk2 (receptor-interacting protein kinase 2), and Steap4 to chronic low-grade inflammation in T2D, independent of [109]. These findings suggest novel elements of cytokine obesity. signaling that require further study. 6. Final Thoughts 5. Modeling Inflammation in T2D Differences in susceptibility to cytokine-induced damage The microenvironment of the pancreatic islet during the between diabetes-prone and nondiabetes-prone islets sug- development of T2D has yet to be fully elucidated. With gest that certain genetic predispositions may render some regard to inflammatory-mediated processes, several sources individuals much more susceptible to the negative impact of proinflammatory cytokines have been suggested through- of proinflammatory cytokines than others. How many other out this review: (1) immune-cell-derived cytokines from chronic diseases could play out in this way? The rela- macrophages and lymphocytes, (2) cytokines derived from tively small increases in circulating cytokines caused by inflamed fat tissues or other distal sources that increase obesity or other chronic inflammatory conditions could cytokine concentration in the general circulation, and (3) conceivably contribute to a plethora of conditions including cytokines/chemokines derived from peptide-producing islet asthma [110, 111], rheumatoid arthritis [112], heart disease cells such as alpha-cells or beta-cells (see Figure 3). Creating [113], various cancers [114–116], polycystic ovarian syn- accurate in vitro models of the concentration, duration, and drome [117, 118], and even mood disorders [119]. Thus, combination of key cytokines involved with each of these the combination of the genetic predisposition of the indi- sources will contribute greatly toward a better understanding vidual and the dose, duration, and milieu of cytokine of how islets and other organs such as the liver, muscle, fat, exposure may significantly contribute to numerous chronic and kidney respond to low-grade inflammation. diseases. Wehavemadeafirstattemptatamodelofthecirculating levels of proinflammatory cytokines utilizing IL-1beta and IL-6. Although the inflammatory environment is dynamic, Competing Interests we focused on cytokines involved prior to the onset of hyperglycemia (the prediabetic stage). Several observations The author declares that there is no conflict of interests served as our rationale for developing this model of cir- regarding the publication of this paper. culating cytokines in T2D. First, the combination of these two cytokines is sufficient to significantly increase the risk of developing T2D in humans [47]. Second, increased levels Acknowledgments of IL-1beta and IL-6 were observed in db/db mice prior to hyperglycemia or substantial differences in weight (<10%) This work was supported by NIDDK R01 089182 and the Dia- compared to control mice [74]. Differences in several other betes Institute and Heritage College of Osteopathic Medicine cytokines, including TNF-alpha [74], were not observed at Ohio University. Special thanks are due to Catherine Hajm- [77]. Third, exposure to combinations of these cytokines at rle and Patrick E. MacDonald for stimulating the genesis circulating levels affected islets in unique ways that treatment of this review and to Karie Cook for editing. Thanks are with individual cytokines could not replicate [55, 74]. These due to Stacey Dula and Kathryn Corbin for data related to studies collectively point to the combination of 5–10 pg/mL Figure 2. 6 Journal of Diabetes Research

References [17] V. Poitout, “Glucolipotoxicity of the pancreatic 𝛽-cell: myth or reality?” Biochemical Society Transactions,vol.36,no.5,pp.901– [1] T. Mandrup-Poulsen, K. Bendtzen, J. H. Nielsen, G. Bendixen, 904, 2008. and J. Nerup, “Cytokines cause functional and structural dam- [18] M. R. Duchen, “Roles of mitochondria in health and disease,” age to isolated islets of Langerhans,” Allergy,vol.40,no.6,pp. Diabetes,vol.53,no.1,pp.S96–S102,2004. 424–429, 1985. [19] E. Araki, S. Oyadomari, and M. Mori, “Endoplasmic reticulum [2] K. Bendtzen, T. Mandrup-Poulsen, J. Nerup, J. H. Nielsen, C. stress and diabetes mellitus,” Internal Medicine,vol.42,no.1,pp. A. Dinarello, and M. Svenson, “Cytotoxicity of human pI 7 7–14, 2003. Interleukin-1 for pancreatic islets of Langerhans,” Science,vol. [20] D. L. Eizirik, A. K. Cardozo, and M. Cnop, “The role for 232,no.4757,pp.1545–1547,1986. endoplasmic reticulum stress in diabetes mellitus,” Endocrine [3] C. Pukel, H. Baquerizo, and A. Rabinovitch, “Destruction of Reviews,vol.29,no.1,pp.42–61,2008. rat islet cell monolayers by cytokines. Synergistic interactions [21] L. Haataja, T. Gurlo, C. J. Huang, and P. C. Butler, “Islet amyloid of interferon-gamma, tumor necrosis factor, lymphotoxin, and in type 2 diabetes, and the toxic oligomer hypothesis,” Endocrine interleukin 1,” Diabetes,vol.37,no.1,pp.133–136,1988. Reviews,vol.29,no.3,pp.303–316,2008. [4]J.A.Corbett,M.A.Sweetland,J.L.Wang,J.R.LancasterJr., [22] K. Schroder, R. Zhou, and J. Tschopp, “The NLRP3 inflamma- and M. L. McDaniel, “Nitric oxide mediates cytokine-induced some: a sensor for metabolic danger?” Science,vol.327,no.5963, inhibition of insulin secretion by human islets of Langerhans,” pp. 296–300, 2010. Proceedings of the National Academy of Sciences of the United [23] S. L. Masters, A. Dunne, S. L. Subramanian et al., “Activation States of America,vol.90,no.5,pp.1731–1735,1993. of the NLRP3 inflammasome by islet amyloid polypeptide [5]A.RabinovitchandW.L.Suarez-Pinzon,“Cytokinesand provides a mechanism for enhanced IL-1𝛽 2 in type 2 diabetes,” their roles in pancreatic islet 𝛽-cell destruction and insulin- Nature Immunology, vol. 11, no. 10, pp. 897–904, 2010. dependent diabetes mellitus,” Biochemical Pharmacology,vol. [24] B. Vandanmagsar, Y.-H. Youm, A. Ravussin et al., “The NLRP3 55, no. 8, pp. 1139–1149, 1998. inflammasome instigates obesity-induced inflammation and [6]J.W.Ramadan,S.R.Steiner,C.M.O’Neill,andC.S.Nune- insulin resistance,” Nature Medicine,vol.17,no.2,pp.179–188, maker, “The central role of calcium in the effects of cytokines on 2011. beta-cell function: implications for type 1 and type 2 diabetes,” [25]S.A.Tersey,Y.Nishiki,A.T.Templinetal.,“Islet𝛽-cell Cell Calcium,vol.50,no.6,pp.481–490,2011. endoplasmic reticulum stress precedes the onset of type 1 [7]A.S.GreenbergandM.L.McDaniel,“Identifyingthelinks diabetes in the nonobese diabetic mouse model,” Diabetes,vol. between obesity, insulin resistance and 𝛽-cell function: poten- 61,no.4,pp.818–827,2012. tial role of adipocyte-derived cytokines in the pathogenesis of [26]Y.Imai,A.D.Dobrian,M.A.Morris,andJ.L.Nadler,“Islet type 2 diabetes,” European Journal of Clinical Investigation,vol. inflammation: a unifying target for diabetes treatment?” Trends 32,no.3,pp.24–34,2002. in Endocrinology and Metabolism,vol.24,no.7,pp.351–360, [8] G. S. Hotamisligil, “Inflammation and metabolic disorders,” 2013. Nature,vol.444,no.7121,pp.860–867,2006. [27] A. Maganti, C. Evans-Molina, and R. Mirmira, “From immuno- 𝛽 [9]M.Y.Donath,J.Størling,K.Maedler,andT.Mandrup-Poulsen, biology to -cell biology: the changing perspective on type 1 “Inflammatory mediators and islet 𝛽-cell failure: a link between diabetes,” Islets,vol.6,no.2,ArticleIDe28778,2014. type 1 and type 2 diabetes,” Journal of Molecular Medicine,vol. [28] K. Pfeffer, “Biological functions of tumor necrosis factor 81, no. 8, pp. 455–470, 2003. cytokines and their receptors,” Cytokine and Growth Factor [10]M.Cnop,N.Welsh,J.-C.Jonas,A.J¨orns,S.Lenzen,andD.L. Reviews,vol.14,no.3-4,pp.185–191,2003. Eizirik, “Mechanisms of pancreatic 𝛽-cell death in type 1 and [29] A. V. Tumanov, S. I. Grivennikov, A. A. Kruglov et al., “Cellular type 2 diabetes: many differences, few similarities,” Diabetes,vol. source and molecular form of TNF specify its distinct functions 54, no. 2, pp. S97–S107, 2005. in organization of secondary lymphoid organs,” Blood,vol.116, no.18,pp.3456–3464,2010. [11] D. L. Eizirik, M. L. Colli, and F. Ortis, “The role of inflammation 𝛽 [30] J. R. Schoenborn and C. B. Wilson, “Regulation of interferon- in insulitis and -cell loss in type 1 diabetes,” Nature Reviews 𝛾 Endocrinology,vol.5,no.4,pp.219–226,2009. during innate and adaptive immune responses,” Advances in Immunology,vol.96,pp.41–101,2007. [12] R. Bergholdt, P. Heding, K. Nielsen et al., “Type 1 database [31] M. Munder, M. Mallo, K. Eichmann, and M. Modolell, “Murine mellitus: an inflammatory disease of the islet,” Advances in macrophages secrete interferon 𝛾 upon combined stimulation Experimental Medicine and Biology,vol.552,pp.129–153,2004. with interleukin (IL)-12 and IL-18: a novel pathway of autocrine [13] M. Y. Donath, M. B¨oni-Schnetzler, H. Ellingsgaard, and J. A. macrophage activation,” Journal of Experimental Medicine,vol. 𝛽 Ehses, “Islet inflammation impairs the pancreatic -cell in type 187, no. 12, pp. 2103–2108, 1998. 2diabetes,”Physiology,vol.24,no.6,pp.325–331,2009. [32]L.Darwich,G.Coma,R.Pe˜na et al., “Secretion of interferon- [14] J. A. Ehses, M. B¨oni-Schnetzler, M. Faulenbach, and M. Y. 𝛾 by human macrophages demonstrated at the single-cell Donath, “Macrophages, cytokines and 𝛽-cell death in Type 2 level after costimulation with interleukin (IL)-12 plus IL-18,” diabetes,” Biochemical Society Transactions,vol.36,no.3,pp. Immunology,vol.126,no.3,pp.386–393,2009. 340–342, 2008. [33] A. Rabinovitch, “An update on cytokines in the pathogenesis [15] V. Poitout and R. P. Robertson, “Minireview: secondary 𝛽-cell of insulin-dependent diabetes mellitus,” Diabetes/Metabolism failure in type 2 diabetes—a convergence of glucotoxicity and Reviews,vol.14,no.2,pp.129–151,1998. lipotoxicity,” Endocrinology,vol.143,no.2,pp.339–342,2002. [34] T. Mandrup-Poulsen, K. Bendtzen, C. A. Dinarello, and J. [16]R.P.Robertson,J.Harmon,P.O.T.Tran,andV.Poitout,“𝛽- Nerup, “Human tumor necrosis factor potentiates human inter- cell glucose toxicity, lipotoxicity, and chronic oxidative stress in leukin 1-mediated rat pancreatic 𝛽-cell cytotoxicity,” The Journal type 2 diabetes,” Diabetes,vol.53,no.1,pp.S119–S124,2004. of Immunology,vol.139,no.12,pp.4077–4082,1987. Journal of Diabetes Research 7

[35]I.L.Campbell,A.Iscaro,andL.C.Harrison,“IFN-𝛾 and tumor [51]K.I.Alexandraki,C.Piperi,P.D.Ziakasetal.,“Cytokine necrosis factor-𝛼: cytotoxicity to murine islets of Langerhans,” secretion in long-standing diabetes mellitus type 1 and 2: Journal of Immunology,vol.141,no.7,pp.2325–2329,1988. associations with low-grade systemic inflammation,” Journal of [36] M. R. Heitmeier, A. L. Scarim, and J. A. Corbett, “Interferon- Clinical Immunology,vol.28,no.4,pp.314–321,2008. 𝛾 increases the sensitivity of islets of langerhans for inducible [52] K. Alexandraki, C. Piperi, C. Kalofoutis, J. Singh, A. Alaveras, nitric-oxide synthase expression induced by interleukin 1,” The and A. Kalofoutis, “Inflammatory process in type 2 diabetes: the JournalofBiologicalChemistry,vol.272,no.21,pp.13697–13704, role of cytokines,” Annals of the New York Academy of Sciences, 1997. vol. 1084, pp. 89–117, 2006. [37] A. Rabinovitch, W. Sumoski, R. V. Rajotte, and G. L. Warnock, [53] H. Kolb and T. Mandrup-Poulsen, “The global diabetes epi- “Cytotoxic effects of cytokines on human pancreatic islet cells demic as a consequence of lifestyle-induced low-grade inflam- in monolayer culture,” The Journal of Clinical Endocrinology & mation,” Diabetologia,vol.53,no.1,pp.10–20,2010. Metabolism, vol. 71, no. 1, pp. 152–156, 1990. [54] P. Jahanshahi, R. Wu, J. D. Carter, and C. S. Nunemaker, [38] G. Soldevila, M. Buscema, M. Doshi, R. F. L. James, G. F. “Evidence of diminished glucose stimulation and endoplas- Bottazzo, and R. Pujol-Borrell, “Cytotoxic effect of IFN-𝛾 plus mic reticulum function in nonoscillatory pancreatic islets,” TNF-𝛼 on human islet cells,” Journal of Autoimmunity,vol.4, Endocrinology,vol.150,no.2,pp.607–615,2009. no.2,pp.291–306,1991. [55]S.B.Dula,M.Jecmenica,R.Wuetal.,“Evidencethatlow- [39]Y.-H.LeeandR.E.Pratley,“Theevolvingroleofinflammation grade systemic inflammation can induce islet dysfunction as in obesity and the metabolic syndrome,” Current Diabetes measured by impaired calcium handling,” Cell Calcium,vol.48, Reports,vol.5,no.1,pp.70–75,2005. no.2-3,pp.133–142,2010. [40] G. Fantuzzi, “Adipose tissue, adipokines, and inflammation,” [56] K. L. Corbin, T. E. Hall, R. Haile, and C. S. Nunemaker, “Anovel Journal of Allergy and Clinical Immunology, vol. 115, no. 5, pp. fluorescence imaging approach for comparative measurements 911–920, 2005. of pancreatic islet function in vitro,” Islets,vol.3,no.1,pp.14–20, 2011. [41] A.S.GreenbergandM.S.Obin,“Obesityandtheroleofadipose [57] V. Wieser, A. R. Moschen, and H. Tilg, “Inflammation, tissue in inflammation and metabolism,” American Journal of cytokines and insulin resistance: a clinical perspective,” Clinical Nutrition,vol.83,no.2,pp.461S–465S,2006. Archivum Immunologiae et Therapiae Experimentalis,vol.61, [42] B. Gustafson, “Adipose tissue, inflammation and atherosclero- no. 2, pp. 119–125, 2013. sis,” Journal of Atherosclerosis and Thrombosis,vol.17,no.4,pp. [58]T.Mandrup-Poulsen,G.A.Spinas,S.J.Prowseetal.,“Islet 332–341, 2010. cytotoxicity of interleukin 1. Influence of culture conditions and [43] K. A. Harford, C. M. Reynolds, F. C. McGillicuddy, and H. islet donor characteristics,” Diabetes,vol.36,no.5,pp.641–647, M. Roche, “Fats, inflammation and insulin resistance: insights 1987. to the role of macrophage and T-cell accumulation in adipose [59] C. A. Dinarello, M. Y. Donath, and T. Mandrup-Poulsen, “Role tissue,” Proceedings of the Nutrition Society,vol.70,no.4,pp. of IL-1𝛽 in type 2 diabetes,” Current Opinion in Endocrinology, 408–417, 2011. Diabetes and Obesity,vol.17,no.4,pp.314–321,2010. [44]M.Arnush,M.R.Heitmeier,A.L.Scarim,M.H.Marino,P. [60] C. M. Larsen, M. Faulenbach, A. Vaag et al., “Interleukin- T. Manning, and J. A. Corbett, “IL-1 produced and released 1-receptor antagonist in type 2 diabetes mellitus,” The New 𝛽 endogenously within human islets inhibits cell function,” England Journal of Medicine,vol.356,no.15,pp.1517–1526,2007. Journal of Clinical Investigation,vol.102,no.3,pp.516–526,1998. [61] A. B. Goldfine, R. Silver, W. Aldhahi et al., “Use of salsalate to [45]A.D.Pradhan,J.E.Manson,N.Rifai,J.E.Buring,and target inflammation in the treatment of insulin resistance and P. M. Ridker, “C-reactive protein, interleukin 6, and risk of type 2 diabetes,” Clinical and Translational Science,vol.1,no.1, developing type 2 diabetes mellitus,” The Journal of the American pp. 36–43, 2008. Medical Association,vol.286,no.3,pp.327–334,2001. [62] A. Schwarznau, M. S. Hanson, J. M. Sperger et al., “IL- [46]F.B.Hu,J.B.Meigs,T.Y.Li,N.Rifai,andJ.E.Manson, 1𝛽 receptor blockade protects islets against pro-inflammatory “Inflammatory markers and risk of developing type 2 diabetes cytokine induced necrosis and apoptosis,” Journal of Cellular in women,” Diabetes,vol.53,no.3,pp.693–700,2004. Physiology,vol.220,no.2,pp.341–347,2009. [47] J.Spranger,A.Kroke,M.M¨ohlig et al., “Inflammatory cytokines [63] M. Y. Donath, “Targeting inflammation in the treatment of type and the risk to develop type 2 diabetes: Results of the prospec- 2diabetes,”Diabetes, Obesity and Metabolism,vol.15,no.3,pp. tive population-based European Prospective Investigation into 193–196, 2013. Cancer and Nutrition (EPIC)-Potsdam study,” Diabetes,vol.52, [64]B.B.Duncan,M.I.Schmidt,J.S.Pankowetal.,“Low-grade no.3,pp.812–817,2003. systemic inflammation and the development of type 2 diabetes: [48]B.Vozarova,C.Weyer,R.S.Lindsay,R.E.Pratley,C.Bogardus, the atherosclerosis risk in communities study,” Diabetes,vol.52, and P. A. Tataranni, “High white blood cell count is associated no. 7, pp. 1799–1805, 2003. with a worsening of insulin sensitivity and predicts the devel- [65] T. Illig, F. Bongardt, A. Sch¨opfer et al., “Significant association opment of type 2 diabetes,” Diabetes,vol.51,no.2,pp.455–461, of the interleukin-6 gene polymorphisms C-174G and A-598G 2002. with type 2 diabetes,” Journal of Clinical Endocrinology and [49] K. E. Wellen and G. S. Hotamisligil, “Inflammation, stress, and Metabolism,vol.89,no.10,pp.5053–5058,2004. diabetes,” The Journal of Clinical Investigation, vol. 115, no. 5, pp. [66] J.-H. Kim, R. A. Bachmann, and J. Chen, “Chapter 21 1111–1119, 2005. interleukin−6 and insulin resistance,” Vitamins & Hormones, [50] J.-M. Fern´andez-Real, “Genetic predispositions to low-grade vol. 80, pp. 613–633, 2009. inflammation and type 2 diabetes,” Diabetes Technology and [67] E. M. Dahl´en, A. Tengblad, T. L¨anne et al., “Abdominal obesity Therapeutics,vol.8,no.1,pp.55–66,2006. and low-grade systemic inflammation as markers of subclinical 8 Journal of Diabetes Research

organdamageintype2diabetes,”Diabetes and Metabolism,vol. [84] K. Eguchi and I. Manabe, “Macrophages and islet inflammation 40,no.1,pp.76–81,2014. in type 2 diabetes,” Diabetes, Obesity and Metabolism,vol.15,no. [68] C. P.Fischer, “Interleukin-6 in acute exercise and training: what 3, pp. 152–158, 2013. is the biological relevance?” Exercise Immunology Review,vol. [85] M. J. Butcher, D. Hallinger, E. Garcia et al., “Association of 12, pp. 6–33, 2006. proinflammatory cytokines and islet resident leucocytes with [69] S.-E. Choi, K.-M. Choi, I.-H. Yoon et al., “IL-6 protects islet dysfunction in type 2 diabetes,” Diabetologia,vol.57,no. pancreatic islet beta cells from pro-inflammatory cytokines- 3, pp. 491–501, 2014. induced cell death and functional impairment in vitro and in [86] C. Westwell-Roper and J. A. Ehses, “Is there a role for the vivo,” Transplant Immunology,vol.13,no.1,pp.43–53,2004. adaptive immune system in pancreatic beta cell failure in type 2 diabetes?” Diabetologia,vol.57,no.3,pp.447–450,2014. [70]H.Ellingsgaard,J.A.Ehses,E.B.Hammaretal.,“Interleukin-6 𝛽 regulates pancreatic 𝛼-cell mass expansion,” Proceedings of the [87] K. Maedler, P. Sergeev, F. Ris et al., “Glucose-induced cell 𝛽 National Academy of Sciences of the United States of America, production of IL-1 contributes to glucotoxicity in human vol.105,no.35,pp.13163–13168,2008. pancreatic islets,” The Journal of Clinical Investigation, vol. 110, no. 6, pp. 851–860, 2002. [71] A. L. Carey and M. A. Febbraio, “Interleukin-6 and insulin sensitivity: friend or foe?” Diabetologia,vol.47,no.7,pp.1135– [88] A.K.Cardozo,P.Proost,C.Gysemans,M.-C.Chen,C.Mathieu, and D. L. Eizirik, “IL-1𝛽 and IFN-𝛾 induce the expression of 1142, 2004. diverse chemokines and IL-15 in human and rat pancreatic islet [72] O. P. Kristiansen and T. Mandrup-Poulsen, “Interleukin-6 and cells, and in islets from pre-diabetic NOD mice,” Diabetologia, diabetes: the good, the bad, or the indifferent?” Diabetes,vol.54, vol. 46, no. 2, pp. 255–266, 2003. supplement 2, pp. S114–S124, 2005. [89]M.-C.Chen,E.Schuit,andD.L.Eizirik,“IdentificationofIL- [73] M. Pal, M. A. Febbraio, and M. Whitham, “From cytokine 1𝛽-induced messenger RNAs in rat pancreatic beta cells by to myokine: the emerging role of interleukin-6 in metabolic differential display of messenger RNA,” Diabetologia,vol.42,no. regulation,” Immunology and Cell Biology,vol.92,no.4,pp.331– 10, pp. 1199–1203, 1999. 339, 2014. [90] M. Igoillo-Esteve, L. Marselli, D. A. Cunha et al., “Palmitate [74]C.M.O’Neill,C.Lu,K.L.Corbinetal.,“Circulatinglevels induces a pro-inflammatory response in human pancreatic of IL-1B+IL-6 cause ER stress and dysfunction in islets from islets that mimics CCL2 expression by beta cells in type 2 prediabetic male mice,” Endocrinology,vol.154,no.9,pp.3077– diabetes,” Diabetologia,vol.53,no.7,pp.1395–1405,2010. 3088, 2013. [91] K. Esposito, F.Nappo, F.Giugliano et al., “Cytokine milieu tends [75] K. Makki, P. Froguel, and I. Wolowczuk, “Adipose tissue in toward inflammation in type 2 diabetes,” Diabetes Care,vol.26, obesity-related inflammation and insulin resistance: cells, no.5,p.1647,2003. cytokines, and chemokines,” ISRN Inflammation,vol.2013, [92] N. Welsh, M. Cnop, I. Kharroubi et al., “Is there a role for Article ID 139239, 12 pages, 2013. locally produced interleukin-1 in the deleterious effects of high [76] M. K. Piya, P. G. McTernan, and S. Kumar, “Adipokine inflam- glucose or the type 2 diabetes milieu to human pancreatic mation and insulin resistance: the role of glucose, lipids and islets?” Diabetes,vol.54,no.11,pp.3238–3244,2005. endotoxin,” Journal of Endocrinology,vol.216,no.1,pp.T1–T15, [93] N. S. Sauter, F. T. Schulthess, R. Galasso, L. W. Castellani, 2013. and K. Maedler, “The antiinflammatory cytokine interleukin-1 [77] C. S. Nunemaker, H. G. Chung, G. M. Verrilli, K. L. Corbin, receptor antagonist protects from high-fat diet-induced hyper- A. Upadhye, and P. R. Sharma, “Increased serum CXCL1 and glycemia,” Endocrinology, vol. 149, no. 5, pp. 2208–2218, 2008. CXCL5 are linked to obesity, hyperglycemia, and impaired islet [94] C. Westwell-Roper, D. L. Dai, G. Soukhatcheva et al., “IL-1 function,” Journal of Endocrinology, vol. 222, no. 2, pp. 267–276, blockade attenuates islet amyloid polypeptide-induced proin- 2014. flammatory cytokine release and pancreatic islet graft dysfunc- [78]R.W.O’Rourke,A.E.White,M.D.Metcalfetal.,“Systemic tion,” The Journal of Immunology,vol.187,no.5,pp.2755–2765, inflammation and insulin sensitivity in obese IFN-𝛾 knockout 2011. mice,” Metabolism, vol. 61, no. 8, pp. 1152–1161, 2012. [95] V. K. Mehta, W. Hao, B. M. Brooks-Worrell, and J. P. Palmer, [79] F. M. Wensveen, S. Valenti´c, M. Sestan,ˇ T. Turk Wensveen, and “Low-dose interleukin 1 and tumor necrosis factor individually B. Poli´c, “The ‘Big Bang’ in obese fat: events initiating obesity- stimulate insulin release but in combination cause suppression,” induced adipose tissue inflammation,” European Journal of European Journal of Endocrinology,vol.130,no.2,pp.208–214, Immunology,vol.45,no.9,pp.2446–2456,2015. 1994. [96] D. L. Eizirik, S. Sandler, N. Welsh, L. Juntti-Berggren, and P.- [80] J. A. Ehses, A. Perren, E. Eppler et al., “Increased number of O. Berggren, “Interleukin-1𝛽-induced stimulation of insulin islet-associated macrophages in type 2 diabetes,” Diabetes,vol. release in mouse pancreatic islets is related to diacylglycerol 56,no.9,pp.2356–2370,2007. production and protein kinase C activation,” Molecular and [81] M. B¨oni-Schnetzler, J. A. Ehses, M. Faulenbach, and M. Y. Cellular Endocrinology, vol. 111, no. 2, pp. 159–165, 1995. Donath, “Insulitis in type 2 diabetes,” Diabetes, Obesity and [97] C. Arous, P. G. Ferreira, E. T. Dermitzakis, and P. A. Halban, Metabolism, vol. 10, no. 4, pp. 201–204, 2008. “Short term exposure of beta cells to low concentrations of [82] H. Cucak, L. G. Grunnet, and A. Rosendahl, “Accumulation of interleukin-1𝛽 improves insulin secretion through focal adhe- M1-like macrophages in type 2 diabetic islets is followed by a sion and actin remodeling and regulation of gene expression,” systemic shift in macrophage polarization,” Journal of Leukocyte Journal of Biological Chemistry,vol.290,no.10,pp.6653–6669, Biology,vol.95,no.1,pp.149–160,2014. 2015. [83] K. Eguchi, I. Manabe, Y. Oishi-Tanaka et al., “Saturated fatty [98] C. Hajmrle, N. Smith, A. F. Spigelman et al., “Interleukin-1 acid and TLR signaling link 𝛽 cell dysfunction and islet signaling contributes to acute islet compensation,” JCI Insight, inflammation,” Cell Metabolism,vol.15,no.4,pp.518–533,2012. vol. 1, no. 4, Article ID e86055, 2016. Journal of Diabetes Research 9

[99]G.A.Spinas,T.Mandrup-Poulsen,J.Mølvigetal.,“Lowcon- the American College of Cardiology,vol.54,no.23,pp.2129– centrations of interleukin-1 stimulate and high concentrations 2138, 2009. inhibit insulin release from isolated rat islets of Langerhans,” [114] E.K.Lund,N.J.Belshaw,G.O.Elliott,andI.T.Johnson,“Recent Acta Endocrinologica, vol. 113, no. 4, pp. 551–558, 1986. advances in understanding the role of diet and obesity in the [100] G. A. Spinas, B. S. Hansen, S. Linde et al., “Interleukin 1 dose- development of colorectal cancer,” Proceedings of the Nutrition dependently affects the biosynthesis of (pro)insulin in isolated Society,vol.70,no.2,pp.194–204,2011. rat islets of Langerhans,” Diabetologia,vol.30,no.7,pp.474– [115]A.M.Ryan,M.Duong,L.Healyetal.,“Obesity,metabolicsyn- 480, 1987. drome and esophageal adenocarcinoma: epidemiology, etiology [101]G.A.Spinas,J.P.Palmer,T.Mandrup-Poulsen,H.Andersen,J. and new targets,” Cancer Epidemiology,vol.35,no.4,pp.309– H. Nielsen, and J. Nerup, “The bimodal effect of interleukin 1 on 319, 2011. rat pancreatic beta-cells—stimulation followed by inhibition— [116]G.Sethi,M.K.Shanmugam,L.Ramachandran,A.P.Kumar, depends upon dose, duration of exposure, and ambient glucose and V. Tergaonkar, “Multifaceted link between cancer and concentration,” Acta Endocrinologica,vol.119,no.2,pp.307–311, inflammation,” Bioscience Reports,vol.32,no.1,pp.1–15,2012. 1988. [117] T. Sathyapalan and S. L. Atkin, “Mediators of inflammation in [102]A.L.Scarim,M.R.Heitmeier,andJ.A.Corbett,“Irreversible polycystic ovary syndrome in relation to adiposity,” Mediators inhibition of metabolic function and islet destruction after a 36- of Inflammation, vol. 2010, Article ID 758656, 5 pages, 2010. 𝛽 hour exposure to interleukin-1 ,” Endocrinology,vol.138,no.12, [118] A. Repaci, A. Gambineri, and R. Pasquali, “The role of low- pp. 5301–5307, 1997. grade inflammation in the polycystic ovary syndrome,” Molec- [103] D. J. Kawahara and J. S. Kenney, “Species differences in human ular and Cellular Endocrinology,vol.335,no.1,pp.30–41,2011. and rat islet sensitivity to human cytokines. Monoclonal anti- [119]J.K.Soczynska,S.H.Kennedy,H.O.Woldeyohannesetal., interleukin-1 (IL-1) influences on direct and indirect IL-1- “Mood disorders and obesity: understanding inflammation as mediated islet effects,” Cytokine,vol.3,no.2,pp.117–124,1991. a pathophysiological nexus,” NeuroMolecular Medicine,vol.13, [104] D. L. Eizirik, S. Sandler, N. Welsh et al., “Cytokines suppress no. 2, pp. 93–116, 2011. human islet function irrespective of their effects on nitric oxide generation,” The Journal of Clinical Investigation,vol.93,no.5, pp. 1968–1974, 1994. [105]D.L.Eizirik,D.G.Pipeleers,Z.Ling,N.Welsh,C.Hellerstr¨om, and A. Andersson, “Major species differences between humans and rodents in the susceptibility to pancreatic 𝛽-cell injury,” Proceedings of the National Academy of Sciences of the United States of America,vol.91,no.20,pp.9253–9256,1994. [106] J. A. Corbett, G. Kwon, J. Turk, and M. L. McDaniel, “IL-1 beta induces the coexpression of both nitric oxide synthase and cyclooxygenase by islets of Langerhans: activation of cyclooxygenase by nitric oxide,” Biochemistry,vol.32,no.50, pp. 13767–13770, 1993. [107]V.Ablamunits,F.Baranova,T.Mandrup-Poulsen,andJ.Nerup, “In vitro inhibition of insulin release by blood mononuclear cells from insulin-dependent diabetic and healthy subjects: synergistic action of IL-1 and TNF,” Cell Transplantation,vol.3, no.1,pp.55–60,1994. [108] M. Cetkovic-Cvrlje and D. L. Eizirik, “TNF-𝛼 and IFN-𝛾 poten- tiate the deleterious effects of IL-1𝛽 on mouse pancreatic islets mainly via generation of nitric oxide,” Cytokine,vol.6,no.4,pp. 399–406, 1994. [109]P.R.Sharma,A.J.Mackey,E.A.Dejeneetal.,“Anislet- targeted genome-wide association scan identifies novel genes implicated in cytokine-mediated islet stress in type 2 diabetes,” Endocrinology,vol.156,no.9,pp.3147–3156,2015. [110]D.Rastogi,S.M.Canfield,A.Andradeetal.,“Obesity-associ- ated asthma in children: a distinct entity,” Chest, vol. 141, no. 4, pp. 895–905, 2012. [111] C. S. Farah and C. M. Salome, “Asthma and obesity: a known associationbutunknownmechanism,”Respirology,vol.17,no. 3, pp. 412–421, 2012. [112]C.S.Derdemezis,P.V.Voulgari,A.A.Drosos,andD.N. Kiortsis, “Obesity, adipose tissue and rheumatoid arthritis: coincidence or more complex relationship?” Clinical and Exper- imental Rheumatology,vol.29,no.4,pp.712–727,2011. [113]P.Libby,P.M.Ridker,andG.K.Hansson,“Inflammationin atherosclerosis: from pathophysiology to practice,” Journal of Hindawi Publishing Corporation Journal of Diabetes Research Volume 2016, Article ID 7053963, 19 pages http://dx.doi.org/10.1155/2016/7053963

Research Article Altered Macrophage and Dendritic Cell Response in Mif −/− Mice Reveals a Role of Mif for Inflammatory-Th1 Response in Type 1 Diabetes

Yuriko Itzel Sánchez-Zamora,1 Imelda Juarez-Avelar,1 Alicia Vazquez-Mendoza,2 Marcia Hiriart,3 and Miriam Rodriguez-Sosa1

1 Unidad de Biomedicina, Facultad de Estudios Superiores (FES) Iztacala, Universidad Nacional Autonoma´ de Mexico´ (UNAM), 54090 Tlalnepantla, MEX, Mexico 2Carrera de Optometr´ıa, FES-Iztacala, UNAM, 54090 Tlalnepantla, MEX, Mexico 3Departamento de Neurodesarrollo y Fisiolog´ıa, Instituto de Fisiolog´ıa Celular, UNAM, 04510 Coyoacan,MEX,Mexico´

Correspondence should be addressed to Miriam Rodriguez-Sosa; [email protected]

Received 31 May 2016; Accepted 10 July 2016

Academic Editor: Jixin Zhong

Copyright © 2016 Yuriko Itzel Sanchez-Zamora´ et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Macrophage migration inhibitory factor (Mif) is highly expressed in type 1 diabetes mellitus (T1DM). However, there is limited information about how Mif influences the activation of macrophages (M𝜑) and dendritic cells (DC) in T1DM. To address this issue, we induced T1DM by administering multiple low doses of streptozotocin (STZ) to Mif−/− or wild-type (Wt) BALB/c mice. We found that Mif−/− mice treated with STZ (Mif−/−STZ) developed lower levels of hyperglycemia, inflammatory cytokines, and specific pancreatic islet antigen- (PIAg-) IgG and displayed reduced cellular infiltration into the pancreatic islets compared toWt mice treated with STZ (WtSTZ). Moreover, M𝜑 and DC from Mif−/−STZ displayed lower expression of MHC-II, costimulatory molecules CD80, CD86, and CD40, Toll-like receptor- (TLR-) 2, and TLR-4 than WtSTZ. These changes were associated with a reduced capacity of M𝜑 and DC from Mif−/−STZ to induce proliferation in ovalbumin-specific T cells. All the deficiencies observed in Mif−/−STZ were recovered by exogenous administration of recombinant Mif. These findings suggest that Mif plays a role in the molecularmechanismsofM𝜑 and DC activation and drives T cell responses involved in the pathology of T1DM. Therefore, Mif is a potential therapeutic target to reduce the pathology of T1DM.

1. Introduction [5–7]. Consequently, molecules such as preinsulin, insulin, glutamate decarboxylase-65 (GAD-65), and islet antigen 2𝛽 T1DM is an autoimmune-mediated disease characterized become recognized as self-antigens [8]. 𝛽 by selective destruction of insulin-producing pancreatic - Different types of innate immune cells such as T cells, cells, resulting in the need for lifelong administration of eosinophils, macrophages (M𝜑), and dendritic cells (DC) exogenous insulin for patient survival [1], and represents 5– as well as proinflammatory cytokines and chemokines are 10%ofallcasesofdiabetes[2,3].Thepathologicalcondition present during insulitis [4, 9]. The relationship between begins with an autoimmune inflammatory process known T1DM and high levels of inflammatory cytokines such as as insulitis, which leads to leucocyte infiltration into the interleukin- (IL-) 1𝛽 [10–14], interferon- (IFN-) 𝛾 [15], tumor pancreatic islets of Langerhans, resulting in the autoimmune necrosis factor- (TNF-) 𝛼 [16], IL-12 [17, 18], and macrophage destruction of pancreatic 𝛽-cells [4]. Thus, the onset of migration inhibitory factor (Mif) [19–24] has been widely T1DMistheresultoftherecognitionofself-antigensdue recognized. to molecular mimicry against infection with various viruses Mif is a pleiotropic cytokine produced during the im- such as Coxsackie virus, 𝛽-cell cytotoxicity, or 𝛽-cell cytolysis mune response by activated T cells, M𝜑,DC,andavariety 2 Journal of Diabetes Research of nonimmune cells and plays a pivotal role in the systemic were approved by the local Institutional Animal Care and inflammatory immune response by promoting the produc- Use Committee. All efforts were made to minimize animal tion of proinflammatory cytokines including TNF-𝛼 and IL- sufferingoverthecourseofthesestudies. 6, which are involved in inflammatory and autoimmune diseases such as septic shock [25], cancer [26], inflammatory 2.2. Animals. Six- to 8-week-old male BALB/c mice were bowel disease [27, 28], rheumatoid arthritis [29, 30], obesity purchased from Harlan Laboratories (Mexico City, Mex.) [31, 32], and diabetes [33–35]. Moreover, Mif has recently and were maintained as a breeding colony in a pathogen- been proposed as a diagnostic biomarker for autoimmune free environment at our animal facilities in accordance with diseases [36] such as arthritis [37, 38], type 2 diabetes [35], institutional guidelines. Mif−/− mice were kindly provided and ulcerative colitis [39] in both animals and humans. by Dr. Abhay R. Satoskar (The Ohio State University, USA) The pathogenic contribution of Mif to T1DM was demon- and were maintained as breeding colonies for more than strated by showing that Mif mRNA expression was upregu- 10 generations in the BALB/c genetic background on the lated in splenic lymphocytes during the development of spon- Transgenic Mouse Core Facility at our institution. Geno- taneouslydiabeticnonobesediabetic(NOD)mice,aswell typing of Mif−/− mice was routinely performed on DNA as cyclophosphamide-treated NOD mice. Diabetes incidence isolated from tail snips using a PCR procedure [40]. The PCR were performed using the following primers: Mif: for- was increased to 86% in NOD mice treated with recombi- 󸀠 󸀠 󸀠 ward 5 -AGACCACGTGCTTAGCTGAG-3 and reverse 5 - nant Mif (rMif) protein, compared with the 55% incidence 󸀠 GCATCGCTACCGGTGGATAA-3 ;Neomycin(Neo):for- observed in untreated control NOD mice [20]. Furthermore, 󸀠 󸀠 − − ward 5 -ATTGAACAAGATGGATTGCAC-3 and reverse studies performed using Mif / mice rendered diabetic by 󸀠 󸀠 administering multiple low doses of streptozotocin (STZ) 5 -CGTCCAGATCATCCTGATC-3 .PCRfortheamplifi- have shown that the absence of Mif affected several aspects of cation of Mif and NEO was performed by adding 100 ng experimental T1DM, including initial immunopathological of the extracted DNA to 25 𝜇L of a reaction mixture that events and the production of proinflammatory and cytotoxic contained 18.4 𝜇Lofdistilledwater,2.5𝜇Lof10xPCRbuffer, mediators, thereby interfering with both inflammation and 0.4 𝜇L of dNTPs (10 mM), 1.5 𝜇L of MgCl2 (25 mM), 1 𝜇Lof tissue destruction [22]. theforwardandreverseprimers,and0.2𝜇L(2.5units)of All the results described above provide evidence that Mif Taq DNA polymerase (Ampliqon, Bioreagents and Molecular Diagnostics). The amplification protocol consisted of 5 min plays a critical role in the pathogenesis of T1DM. However, ∘ ∘ ∘ at 95 C, 35 cycles of 95 Cfor30sec,58Cfor40sec,and the precise mechanism by which Mif promotes insulitis and ∘ ∘ the mechanism underlying its proinflammatory role remain 72 C for 30 sec and a final extension at 72 Cfor5min.All unclear. The activities of Mif may reside at the levels of PCR experiments were conducted with positive and negative both the inductive and effector phases of the inflammatory controls. A PCR fragment of 200 bp, corresponding to Mif, response attributed to antigen-presenting cells. or 500 bp, corresponding to NEO, was visualized to identify Wt or Mif−/− mice, respectively. The PCR products were Here, we analyzed the influence of Mif on M𝜑 and DC analyzed by electrophoresis on a 1.5% agarose gel and were activation using an autoimmune diabetes model in which viewed under UV light (Bio-Rad, USA). multiple low doses of STZ were administered to Mif−/− and wild-type (Wt) mice (Mif−/−STZ and WtSTZ, resp.) in 2.3. Induction of T1DM. Mif−/− and Wt mice were deprived the BALB/c genetic background. As previously demonstrated of food for 8 h before induction of diabetes via intraperitoneal Mif−/−STZ developed less severe hyperglycemia, reduced (i.p.) injection of STZ at doses of 40 mg/kg of body weight, levels of IFN-𝛾 and TNF-𝛼,asmalleramountofpancre- daily for five consecutive days (days 0–4) (Sigma-Aldrich, atic islet antigen- (PIAg-) specific IgG, and decreased cell St.Louis,MO,USA).STZwasdilutedincold0.01Mcitrate infiltration into the pancreatic islets compared to WtSTZ. buffer (pH 4.5) and was used within 5 min of preparation, in Interestingly,wefoundforthefirsttimethatM𝜑 and DC from accordance with a previously reported protocol [41]. Healthy Mif−/−STZ displayed decreased expression of costimulatory mice from each group received i.p. injections of an equivalent molecules CD80, CD86, and CD40, as well as Toll-like volume of vehicle (citrate buffer) as negative controls. receptor- (TLR-) 2, TLR-4, and major histocompatibility complex- (MHC-) II. Importantly, we demonstrated that 2.4.AnalysisofBloodGlucose,SerumInsulin,andCytokine due to diminished upregulation of costimulatory molecules, Levels. Blood samples were collected by tail snipping from these cells exhibited a reduced capacity to induce prolifer- Wt and Mif−/− mice that had been fasting for 6 h. Samples ation and cytokine expression in cocultures with allogeneic were obtained once before STZ injection and 2, 4, and 8 weeks ovalbumin- (OVA-) specific T cells. All deficiencies observed afterSTZinjection.Bloodglucoselevelsweremeasuredwith in Mif−/−STZ were reversed by exogenous rMif protein aportableglucometer(Accu-ChekSensorglucometer;Roche administration. Diagnostics, Indianapolis, IN, USA). Mice with a glucose concentration exceeding 300 mg/dL were considered to have 2. Materials and Methods T1DM. Blood was collected and centrifuged at 1300 ×g, and the serum levels of Mif (Neobiolab, USA), IL-12, IFN-𝛾,IL- 2.1. Ethics Statement. All experiments in this study were 17, IL-4, IL-13 (PeproTech, Mex), and insulin (Lincoln, St. performed according to the guidelines in the Mexican Reg- Charles, MO, USA) were determined via ELISA according to ulations on Animal Care (NOM-062-ZOO-1999, 2001) and the manufacturer’s instructions. Journal of Diabetes Research 3

2.5. Pancreatic Islets Antigen (PIAg) Isolation. PIAg islets were GIBCO, BRL, Grand Island, NY, USA). Spleen cells were 6 obtained from healthy Wt mice as described below. After suspended at 5 × 10 cells/mL in the same medium. The isolation, the islets were lysed by five freeze-thaw cycles pancreas was also removed under sterile conditions, and pan- followed by sonication (six 60-Hz cycles for 1 min each) on creaticisletswereisolatedusingthecollagenasemethodas ∘ ice. After centrifugation (10 000 g, 15 min, 4 C), supernatants previously described [44]. Briefly, the pancreas was removed were collected and filtered through a 0.2 𝜇mmembrane and cut into small pieces (approximately 3 mm in size). The (Corning, Cambridge, MA, USA). The protein concentration tissue was subsequently incubated in collagenase (0.3 mg/mL; was determined using the Lowry method [42], and PIAg Roche Diagnostics Corp., Indianapolis, IN, USA) for 10 ∘ ∘ aliquots were stored at −70 Cuntiluse. minutes at 37 Cinatotalof1mLofdigestionsolution under constant shaking and intermittent vortexing. Islets 2.6. Analysis of IgG Antibody Production. Serum samples were subsequently washed several times in HBSS containing were analyzed for the levels of pancreatic islet-specific Th1- BSA (5 mg/mL) and were hand-picked under a dissecting associated IgG2a and Th2-associated IgG1 antibodies by microscope. Islets were dispersed into single cells by suspen- ELISA. Briefly, 96-well ELISA plates (Costar) were coated sion in trypsin-EDTA (GIBCO, BRL) and passage through a with 100 𝜇L/wellsolublePIAginTrisbuffer,pH7.8.The siliconized Pasteur pipette. Then, the cells were incubated at ∘ ∘ plates were incubated overnight at 4 C. Then, the wells were 37 C. washed thoroughly with phosphate-buffered saline (PBS) containing 0.05% Tween-20 (PBS-Tween; Merck, France) 2.9. Flow Cytometry Analysis. Cells from the spleen or and were blocked with PBS supplemented with 1% bovine pancreas obtained as described above were then used for serum albumin (PBS-BSA; Sigma-Aldrich) for one hour at flow cytometry analysis. In brief, cells were washed in room temperature (RT). The serum samples were diluted flow cytometry wash solution (Dulbecco’s PBS containing 1 : 100, followed by serial dilution of each sample (in PBS- 1% FCS and 0.05% sodium azide), followed by incubation BSA) from healthy or STZ-treated Wt or Mif−/− mice. The with allophycocyanin- (APC-) conjugated anti-F4/80 and ∘ plates were then incubated at 4 C overnight. After extensive anti-CD11c antibodies for differentiation of𝜑 M and DC, washing with PBS-Tween, the samples were incubated for respectively. Then, the selected cells were incubated in 3% 45 min at RT with isotype-specific peroxidase-labeled goat BSA-PBS containing a phycoerythrin- (PE-) labeled anti- anti-mouse antibodies (anti-IgG1 and anti-IgG2a at 1/1000 CD80, anti-CCR5, or anti-TLR-4 antibody or a fluores- dilutions; Zymed, San Francisco, CA, USA). Then, the plates cein isothiocyanate- (FITC-) labeled anti-CD86, anti-CD40, anti-MHC-II, or anti-TLR-2 antibody (all antibodies from were washed, and immunoreactivity was detected with ABTS ∘ solution (Zymed). The results were expressed as endpoint Biolegend, San Diego, CA, USA) at 4 Cfor30min.After titers based on optical density. incubation, the cells were washed several times in buffer, fixed in 1% paraformaldehyde (Sigma-Aldrich), and stored at ∘ 2.7. Histopathology. Pancreatic tissues from Wt and Mif−/− 4 C in the dark, followed by analysis using a FACSCalibur mice healthy or injected with STZ were removed, fixed flow cytometer and CellQuest software (Becton Dickinson, overnight in 4% formaldehyde, and embedded in paraffin Franklin Lakes, NJ, USA). blocks. Afterwards, 5 to 7 𝜇m transverse sections of pancre- atic tissue were sliced from the paraffinized tissue blocks, 2.10. Coculture of Macrophages with Spleen Cells. Coculture mounted on slides, and subsequently stained with eosin- of M𝜑 with naive spleen cells was performed as follows. hematoxylin (E&H; Sigma-Aldrich). For each mouse, one Adherent M𝜑 among peritoneal exudate cells (PECs) from histological containing 1–3 nonsuccessive slices was scored healthy or 8 weeks after STZ-treatment Wt or Mif−/− mice were obtained. Briefly, the𝜑 M density was adjusted to 5 × for infiltration as previously described [43] according to the 6 following scale: grade 0 (no insulitis) = 0% infiltration within 10 cells/mL, and the cells were plated (100 𝜇L) in 96-well the islets; grade 1 (peri-insulitis) = 1–10% infiltration; grade flat-bottom plates (Costar, Cambridge, MA, USA). Three 2 (moderate insulitis) = 11–<50% infiltration; grade 3 (severe hours later, the PECs were washed three times with warm insulitis) = >50% infiltration; or grade 4 (complete insulitis) sterile PBS to remove nonadherent cells, and 10 𝜇gofOVA = extensive infiltration with few or no detectable pancreatic (Worthington, USA) in 100 𝜇L of DMEM supplemented islet cells. Using an Olympus BX51 microscope (Olympus media was added. Three hours later, adherent M𝜑 were America, Melville, NY, USA) equipped with a digital video washed three times with warm sterile DMEM to remove camera, 30 islets of Langerhans were evaluated per mouse. excess OVA that had not been phagocytosed. Spleen cells from OVA-transgenic mice were obtained as previously 6 2.8. Cell and Pancreatic Islet Isolation. Spleen and pancreatic described [45], suspended at 1 × 10 cells/mL, and added islet cells from WtSTZ and Mif−/−STZ were collected after (100 𝜇L) to PECs at a ratio of 1 M𝜑 : 5 spleen cells. The ∘ 0, 2, 4, and 8 weeks of injection with STZ and stained for cocultures were maintained at 37 Cin5%CO2 for 5 days. flow cytometry analysis. Briefly, the spleen was removed Then, [3H]-thymidine (185 GBb/mmol activity, Amersham, under sterile conditions, and spleen cells were obtained by UK) was added at 0.5 𝜇Ci/well, and the cells were incubated mincing and filtering the tissue, followed by washing and for a further 18 h. The cells were harvested using a 96-well suspension in DMEM culture medium supplemented with harvester (Tomtec, Toku, Finland) and then counted using 10% fetal bovine serum, 100 units of penicillin/streptomycin, a microplate counter (Trilux, Toku, Finland). The values are 2 mM glutamine, and 1% nonessential amino acids (all from presented as counts per minute (CPM) from triplicate wells. 4 Journal of Diabetes Research

600 0.5 ∗∗∗ ∗∗∗ ∗∗∗ 0.4 ∗∗∗ 400 0.3 ∗

0.2 200 Blood glucose (mg/dL) Serum (pg/mL) insulin 0.1 ∗

0 0.0 024680 2468 Weeks after induction Weeks after induction WtSTZ Healthy Wt WtSTZ Healthy Wt Mif −/−STZ Healthy Mif −/− Mif −/−STZ Healthy Mif −/− (a) (b)

Figure 1: Mif−/− mice developed less severe hyperglycemia than Wt mice after type 1 diabetes mellitus induction. Blood glucose (every 2 weeks) (a) and serum insulin (0, 2,, and 8 weeks) (b) levels were monitored as described in Section 2 in Wt (◼)andMif−/− (󳶃) mice after STZ administration; as controls, healthy Wt (I)orhealthyMif −/− (Δ) mice were plotted as well. All data are representative of three independent ∗ ∗∗∗ experiments and are expressed as the means ± SE (n = at least 5–7 animals per group/time point). 𝑝 < 0.05, 𝑝 < 0.001, GraphPad Prism software 6.

2.11. Mif Reconstitution. To establish the Mif levels under Mif−/−STZ. The blood glucose level of Mif−/−STZpeakedat conditions of T1DM, the serum Mif levels were deter- 219±23 mg/dLonweek4,andthebloodglucoselevelinsome mined weekly until 8 weeks after STZ treatment in WtSTZ. mice decreased between weeks 6 and 8 (226 ± 15 and 182 ± Mif−/−STZ received similar concentrations of rMif (R&D 16 mg/dL, resp.) (Figure 1(a), Mif−/−STZ: inverted triangles). Systems, USA) to the Mif levels observed in WtSTZ to This observation suggested that a slight recovery of the emulate physiological conditions in WtSTZ. Briefly, Mif−/− glucoseresponsemayoccurinMif−/−STZ at approximately mice received i.p. injection of STZ together with 500 pg of week 8 of STZ treatment. rMif in 100 𝜇L of saline solution as a vehicle. Upon STZ WtSTZshowedapeakbloodinsulinlevelat2weeks treatment, Mif−/− mice received an i.p. injection of rMif after STZ treatment, and after the sixth and eighth weeks of every three days at the following doses: 850 pg on week 1; treatment, their blood insulin levels were significantly lower 1280 pg on week 2; 1292 pg on week 3; 1305 pg on week 4; than those of healthy Wt mice (Figure 1(b), WtSTZ: squares 4300 pg on week 5; 7300 pg on week 6; and, finally, 13200 pg with solid line; Wt mice: white circles with dotted line). on week 7. Other Mif−/−STZmicewereinjectedwithan Interestingly, Mif−/−STZshowednosignificantchangesin equivalent volume of saline solution (100 𝜇L) as controls. insulinlevelscomparedtohealthyMif−/− or Wt mice over the course of the experiment (Figure 1(b), Mif −/−STZ: inverted 2.12. Statistical Analysis. Comparisons between Wt and triangles with solid line; healthy Mif−/− mice: white triangles Mif−/− mice that were healthy or treated with STZ were with dotted line). performed using either Student’s unpaired 𝑡-test or ANOVA followed by Turkey’s multiple comparisons test for data that 3.2. Mif−/− Mice Produced Lower Proinflammatory Cytokine displayed a normal distribution. 𝑝 values less than 0.05 were Levels Than Wt Mice after T1DM Induction. We measured ∗ considered significant and were designated as 𝑝 < 0.05, the serum levels of proinflammatory and anti-inflammatory ∗∗ ∗∗∗ 𝑝 < 0.01,or 𝑝 < 0.001.Alldatawereanalyzedusing cytokines in Wt and Mif−/− mice after STZ administra- GraphPad Prism 6 software (San Diego, CA, USA). tion. As expected, the WtSTZ displayed gradually increasing serum levels of inflammatory cytokines such as Mif, IL- 3. Results 12, and IFN-𝛾 (Figures 2(a), 2(b), and 2(c), resp., WtSTZ: squares) between 2 and 8 weeks after T1DM induction. The 3.1. Mif−/− Mice Developed Less Severe Hyperglycemia Than level of the proinflammatory cytokine IL-17 was significantly Wt Mice after STZ Administration. We first investigated the increased at 8 weeks after T1DM induction in WtSTZ clinical effects of STZ on Mif−/− and Wt mice. Our data (Figure 2(d), squares). In contrast, Mif−/−STZ displayed demonstrated that Wt mice rapidly developed hyperglycemia significantly lower serum levels of Mif, IL-12, IFN-𝛾,and after STZ administration. These mice sustained high blood IL-17 than WtSTZ (Figures 2(a), 2(b), 2(c), and 2(d), resp.; glucose levels from week 2 (281.7 ± 16 mg/dL) until week 8 of Mif−/−STZ: inverted triangles). STZ administration (470.5 ± 24 mg/dL) (Figure 1(a), WtSTZ: The serum levels of the anti-inflammatory cytokine IL- squares). In contrast, hyperglycemia developed gradually in 4 were significantly lower in WtSTZ (squares) than in Journal of Diabetes Research 5

16000 MIF ∗∗∗ 1200 IL-12 ∗∗∗ 14000

12000 800 ∗∗∗

10000 3000 ∗ (pg/mL) (pg/mL) 400 2000 ∗ ∗ 1000

0 0 0246 8 0 2468 Weeks after induction Weeks after induction WtSTZ WtSTZ Mif −/−STZ Mif −/−STZ (a) (b)

1500 IFN-𝛾 300 IL-17 ∗∗∗ ∗∗∗ ∗∗∗ 1000 200 (pg/mL) (pg/mL) 500 100

0 0 0 2468 0 2468 Weeks after induction Weeks after induction WtSTZ WtSTZ Mif −/−STZ Mif −/−STZ (c) (d) 6000 200 IL-4 IL-13

5000 ∗∗∗ 180 4000 ∗∗

3000 160 (pg/mL) (pg/mL) 2000 140 1000

0 120 0 2468 0 2468 Weeks after induction Weeks after induction WtSTZ WtSTZ Mif −/−STZ Mif −/−STZ (e) (f)

Figure 2: Mif deficiency prevents the elevation of proinflammatory cytokine production. The levels of Mif (a), IL-12 (b), IFN-𝛾 (c), IL-17 (d), IL-4 (e), and IL-13 (f) in the sera from Wt (◼)andMif −/− (󳶃) mice at 0, 2, 4, and 8 weeks after STZ administration were measured via enzyme-linked immunosorbent assay (ELISA) in triplicate, as indicated in Section 2. Data are expressed as the means ± SE (n =atleast5–7 ∗ ∗∗ ∗∗∗ mice per group/time point). 𝑝 < 0.05, 𝑝 < 0.01,or 𝑝 < 0.001, GraphPad Prism software 6. 6 Journal of Diabetes Research

Table 1: Assessment of the extent of leukocyte infiltration into 3.4. Mif−/− Mice Showed Lower Levels of Specific Pancreatic pancreatic islets of WT and Mif −/− mice at 8 weeks after STZ Islet Antibodies Than Wt Mice after T1DM Induction. In administration. T1DM, the immune system targets self-antigens within pan- creatic islets and destroys the inhabiting insulin-secreting 𝛽- Groups Number of Infiltrated Islets that lost circular islets islets morphology cells. Therefore, autoantibody detection serves as a predictive Wt 100 ND ND factor for the onset of diabetes in both humans and mice [46, 47]. WtSTZ 100 100 65 To investigate how Mif contributes to autoantibody pro- −/− Mif 100 ND ND duction in T1DM, the serum levels of pancreatic islet-specific Mif −/−STZ 100 33 25 IgG2a and IgG1 antibodies were determined in Wt and ND: none detected; Wt: wild-type. Mif−/− mice after 2, 4, and 8 weeks after STZ administration. WtSTZproducedhighlevelsofIgG2aatalltimepoints analyzed (Figure 4(a)) and IgG1 only at week 8 (Figure 4(b)). Mif−/−STZ (inverted triangles) (Figure 2(e)). No differences Importantly, Mif−/−STZmicedisplayedsignificantlylower in the serum IL-13 levels were found between WtSTZ and levels of both IgG2a and IgG1 than WtSTZ (Figures 4(a) Mif−/−STZ (Figure 2(f)). and 4(b), resp.), suggesting suppressed development of an adaptive immune response. These findings suggest that Mif 3.3. Mif−/− Mice Showed Reduced Pancreatic Islet Damage modulates the incidence and severity of diabetes by favoring and Cellular Infiltration Compared to Wt Mice after T1DM the development of autoantibodies in T1DM. Induction. To confirm that STZ reached its target, Wt and Mif−/− mice were treated with a single high dose of 3.5. Splenic M𝜑 and DC from Mif−/−STZ Mice Expressed STZ (150 mg/kg). The toxic effect of high-dose STZ was Lower Levels of Costimulatory Molecules Than Those from similar between Wt and Mif−/− mice (Supplementary WtSTZ Mice. We showed that the reduced severity of T1DM Figure 1 in Supplementary Material available online at in Mif −/−STZ is associated with decreased pancreatic islet http://dx.doi.org/10.1155/2016/7053963). In both experimen- damage and diminished adaptive immune responses. To tal groups, the toxic effect of STZ was present, and the high determine whether this phenotype could be related to the dose of STZ destroys the insulin-producers beta cells, leading degree of activation of antigen-presenting cells, we charac- to acute nonimmune-mediated diabetes; in contrast the mul- terized the expression of the costimulatory molecules CD80, tiple low doses of STZ at 40 mg/Kg, daily for five consecutive CD86, and MHC-II and of the receptors TLR-2 and TLR-4 in days, require participation of immune-inflammatory events M𝜑 and DC from the spleen and pancreas of Mif−/− and Wt for T1DM development [41]. mice after 0, 2, 4, and 8 weeks of STZ administration. To assess the damage to 𝛽 cells in a model of T1DM In cells isolated from the spleen, the percentages of induced by multiple low doses of STZ, we sacrificed the mice CD80-, CD86-, MHC-II-, TLR2-, and TLR4-expressing M𝜑 + at 8 weeks after T1DM induction and removed the pancreas (F4/80 )weresimilarbetweenhealthyMif−/− and Wt mice for H&E staining and histological analysis. We assessed the (Figures 5(a), 5(b), 5(c), 5(d), and 5(e), resp., at time 0 number and size of normal and infiltrated pancreatic islets post-STZ administration, right panel). Interestingly, upon in each slide for the experimental and healthy mice. Thirty STZ administration, Mif −/− mouse M𝜑 showed impaired pancreatic islets were quantified per experimental group. activation, characterized by reduced expression of CD80, Islets from healthy Wt and Mif−/− mice had a round CD86, MHC-II, TLR-2, and TLR-4, compared with M𝜑 from morphology and well-defined borders without cellular infil- WtSTZ mice (Figures 5(a), 5(b), 5(c), 5(d), and 5(e), resp.). + tration (Figures 3(a) and 3(c)). As expected, the pancreas Similarly, DCs (CD11c )isolatedfromthespleenof from WtSTZ showed fewer and smaller islets than those from Mif−/−STZ mice expressed lower levels of CD80, CD86, healthy Wt mice. Additionally, evident cellular infiltration and MHC-II at 4 through 8 weeks after STZ administration ledtothebreakdownofisletmorphologyinthepancreas (Figures 6(a), 6(b), and 6(c), resp.), whereas the expression of WtSTZ (Figure 3(b)). In contrast, partial islet damage of TLR-2 and TLR-4 in DCs was reduced at 4 weeks in (Figure 3(d)) and no significant reduction in islet number Mif−/−STZ compared to WtSTZ (Figures 6(d) and 6(e), or size were observed in Mif−/−STZ compared to WtSTZ resp.). (Table 1). As shown in Figure 3(e), Mif−/− mice at 8 weeks after 3.6. Pancreatic M𝜑 and DC from Mif−/−STZ Mice Express STZ administration displayed marked reductions in both Lower Levels of Costimulatory Molecules Than Those from invasive insulitis (insulitis grades 3 and 4) and mild peri- WtSTZ Mice. To investigate the role of Mif in the acti- insulitis (insulitis grade 2) in pancreatic islets compared to vation of APCs in the pancreas, we determined MHC-II WtSTZ. Thus, pancreatic islets from Mif−/−STZ had reduced and costimulatory molecule expression in M𝜑 and DC in damage associated with reduced leucocyte infiltration com- the pancreas of Mif−/−STZ and WtSTZ mice. M𝜑 from pared to pancreatic islets from WtSTZ. These results demon- Mif−/−STZ expressed lower levels of CD80 and CD86 from strated that STZ administration triggers 𝛽-cell destruction 4 through 8 weeks after STZ administration (Figures 7(a) followed by insulin deficiency and hyperglycemia in Wt mice. and 7(b), resp.). Alternatively, reduced expression of MHC-II, Although STZ reached the pancreatic islets in Mif−/− mice, TLR-2, and TLR-4 was detected earlier, from 2 weeks until 8 Mif−/−STZexhibitedminordamagecomparedtoWtSTZ. weeks after STZ treatment (Figures 7(c), 7(d), and 7(e), resp.). Journal of Diabetes Research 7

Healthy STZ Wt Wt

40x 40x (a) (b) Healthy STZ −/− −/− Mif Mif

40x 40x (c) (d) 50

∗∗ 40

30

Islets damage Islets 20

10

0 WtSTZ Mif −/−STZ

4 1 3 0 2 (e)

Figure 3: Mif−/− mice conserve their healthy anatomy and exhibit limited cellular infiltration into pancreatic islets after STZ administration. Pancreases isolated from Wt or Mif −/− mice were examined via histological analysis of eosin-hematoxylin (E&H) staining to establish the number of pancreatic islets and to determine lymphocyte infiltration (arrowhead in (b)) at 8 weeks after STZ administration. Pancreatic islets from healthy Wt mice (a); Wt mice treated with STZ, WtSTZ (b); healthy Mif −/− mice (c), and Mif −/− mice treated with STZ, Mif −/−STZ (d). Compilation of infiltration stages in the pancreas of Wt and Mif −/− mice after STZ administration (e). Pancreatic islets were scored using the following scale: grade 0 (no insulitis) = 0% infiltration; grade 1 (peri-insulitis) = 1–10% infiltration; grade 2 (moderate insulitis) = 11–<50% infiltration; grade 3 (severe insulitis) = >50% infiltration; or grade 4 (complete insulitis) = complete infiltration. We counted 30–40 islets per experiment using six mice per experimental group, depending on the number of islets that were present in the sections. All data are ∗∗ representative of two independent experiments, 𝑛=12from two experiments. 𝑝 < 0.01, GraphPad Prism software 6. 8 Journal of Diabetes Research

60000 10000 IgG2a IgG1

∗∗∗

8000 ∗∗∗

40000 6000

4000 titer Endpoint Endpoint titer Endpoint ∗∗∗ 20000 ∗∗ 2000

0 0 0 2468 0 2468 Weeks after induction Weeks after induction WtSTZ WtSTZ Mif −/−STZ Mif −/−STZ (a) (b)

Figure 4: Kinetics of antibody responses in type 1 diabetes mellitus models induced by low-dose STZ. Specific production of IgG2a (a) and IgG1 (b) antipancreatic islet antigen antibodies. Sera from Wt (◼)andMif−/− mice (󳶃) at weeks 0, 2, 4, and 8 after STZ administration were processed by ELISA, using pancreatic islets antigen (PIAg) as the source of specific antigen as indicated in Section 2. Data are expressed as the mean reciprocal endpoint titer ± SEM. Five to six animals were analyzed in each group. The data shown are representative of one out of ∗∗ ∗∗∗ three identical experiments with similar results. Mann-Whitney U test. 𝑝 < 0.01, 𝑝 < 0.001, GraphPad Prism software 6.

Moreover, DC from Mif−/−STZ mice displayed reduced levels similar to those in WtSTZ during the first six weeks CD80 at 4 to 8 weeks (Figure 8(a)), CD86 at all time after STZ treatment. However, at 8 weeks after STZ treatment, points analyzed (Figure 8(b)), and reduced TLR-2 and TLR- the glucose levels were not increased in Mif−/−STZ+rMif 4 expression at 4 weeks and 2 weeks, respectively, compared compared to WtSTZ (Figure 10(a)). to DCs from WtSTZ (Figures 8(d) and 8(e), resp.). No ComparableserumlevelsofthecytokinesIL-6andIL- significant differences in MHC-II expression were detected 12 were observed between Mif−/−STZ+rMif and WtSTZ (Figure 8(c)). during the first 6 weeks after STZ treatment. However, the

+ + levels of these inflammatory cytokines significantly decreased 3.7. F4/80 Macrophages and CD11b Monocytes Obtained in Mif −/−STZ+rMif mice at 8 weeks after STZ treatment from Mif−/−STZShowImpairedCapacitytoInduceSpleen compared to WtSTZ (Figures 10(b) and 10(d)). Interestingly, Cell Proliferation. The ability of M𝜑 to activate T cells was the serum levels of TNF-𝛼 from Mif−/−STZ+rMif were investigated using OVA-transgenic T cell cocultures. M𝜑 higher than those of WtSTZ at all time points analyzed were collected from Wt or Mif−/− mice treated or untreated (Figure 10(c)). These results confirm that Mif acts asa for 8 weeks with STZ. As shown in Figure 9(a), after priming powerful inducer of proinflammatory cytokines involved in the cells with OVA in vitro,M𝜑 from Mif−/−STZ mice the development of experimental T1DM. induced less T cell proliferation in response to OVA than M𝜑 from WtSTZ. A similar trend was observed in cocultures + of CD11b cells from Mif −/−STZ and OVA-transgenic T 4. Discussion cells (Figure 9(b)). These data suggest a role of Mif in promoting M𝜑 activation, which in turn induces specific Currently, there is no doubt that Mif is a key molecule T cell proliferation, particularly in this experimental T1DM that promotes proinflammatory immune responses [48]. This model. proinflammatory property of Mif contributes to developing protective inflammatory-Th1 immune response in different 3.8.MifReconstitutioninSTZ-TreatedMif−/− Mice Promotes models of parasitic diseases [49]. In contrast the same proin- Hyperglycemia and Reestablishes the Production of Proinflam- flammatory property of Mif participates in the pathogenesis matory Cytokines in This T1DM Model. The systemic levels of many inflammatory diseases [50]. In line with the last one, of Mif were reconstituted in Mif−/−STZmiceduringthe recently it has been established that high blood levels of Mif course of T1DM, as described in Section 2. Mif−/−STZ that are associated with human T1DM, similar to the findings in received rMif (Mif−/−STZ+rMif) displayed blood glucose experimental mouse models of T1DM [20, 21, 24, 51]. Journal of Diabetes Research 9

Spleen macrophages 200

∗∗∗ ∗∗∗ 150 cells +

2 weeks 4 weeks 8 weeks 100 CD80 + MFI MFI MFI

F4/80 50

CD80 0 02468 Weeks after induction (a) Spleen macrophages 250

∗ 200 ∗ cells

+ 150 2 weeks 4 weeks 8 weeks ∗∗ CD86

+ 100 MFI MFI MFIMFI F4/80 50

CD86 0 0 2468 Weeks after induction (b) Spleen macrophages 250

∗∗∗

cells ∗∗∗ + 200

2 weeks 4 weeks 8 weeks II MFI

MFI MFI MHC- + 150 F4/80

MHC-II 100 0 2468 Weeks after induction (c) Spleen macrophages 400

∗∗∗ 300 ∗∗∗ cells

2 weeks 4 weeks 8 weeks + ∗∗∗ 200 TLR2

MFI MFI MFI +

F4/80 100

TLR-2 0 0 2468 Weeks after induction (d)

Figure 5: Continued. 10 Journal of Diabetes Research

Spleen macrophages 400

∗∗∗ 2 weeks 4 weeks 300 8 weeks cells ∗∗∗ ∗∗∗ +

MFI MFI 200

MFI TLR4 +

100 F4/80

TLR-4 0 0 2468 Weeks after induction (e)

Figure 5: Mif promotes costimulatory molecule expression on spleen M𝜑.Thetimecourseofcostimulatorymolecule:CD80(a),CD86(b), + MHC-II (c), TLR-2 (d), and TLR-4 (e) expression on spleen F4/80 M𝜑.M𝜑 from spleen of Wt and Mif−/− mice were harvested at 0, 2, 4, 7 and 8 weeks after STZ administration and 1 × 10 cells/mL were processed by flow cytometric analysis as indicated in Section 2. Analyses of expression are shown in the right panels: WtSTZ (◼); Mif−/−STZ (󳶃). And representative histograms of expression based on fluorescence are shown in the left panels; isotype controls are indicated in gray shadow, the black line represents expression by Mif−/−STZ cells, and the blue line shows expression by WtSTZ cells. The data are presented as the mean fluorescence intensities (MFI) from one representative experiment. ∗ ∗∗ ∗∗∗ Each experiment was repeated three times (𝑛=3) and individually analyzed. 𝑝 < 0.05, 𝑝 < 0.01,or 𝑝 < 0.001, GraphPad Prism software 6.

Studies in NOD mice or mice treated with multiple low for these two cytokines. These observations confirm that Mif doses of STZ (despite the pathogenic differences between exogenous acts as a powerful inducer of IL-6 and IL-12, but these models) have shown that pancreatic 𝛽 cell destruction only if it is present in steady high concentration. − results from the toxic effect of free radicals (O2 ,H2O2, In addition, we found that Mif−/−STZ did not produce and nitric oxide) and inflammatory cytokines released by detectable levels of islet autoantibodies, in contrast to WtSTZ, activated M𝜑 and T cells [15, 41, 52]. Therefore, both models which produced high levels of islet autoantibodies. It is have been widely used to dissect the role of Mif in the well known that the early presence of islet autoantibodies pathogenesis of T1DM using anti-Mif monoclonal antibody is decisive in the development of diabetes by NOD mice as treatment or using Mif −/− mice. In all cases, the lack of wellashumans[46,47,53].OurresultsconfirmthatMifis Mif resulted in diminished manifestation of the disease, essential for the development of hyperglycemia and suggest a decreased glucose blood levels, and reduced production role for Mif not only in the innate immune response but also inflammatory cytokines associated with the development of intheadaptiveimmuneresponseinT1DM. T1DM, including TNF-𝛼,IL-1𝛽,IFN-𝛾,IL-12,andIL-23[21, Recently, it has been reported that Mif is produced by 22, 24]. Our study validates and extends these findings by pancreatic 𝛽-cells and that Mif is released by insulin granules demonstrating an important role for Mif in promoting cos- in an autocrine fashion [54, 55]. The chemical destruction of timulatory molecule expression in M𝜑 and DC during T1DM 𝛽-pancreatic islets by STZ in Wt mice damaged 𝛽-pancreatic development. We further demonstrate that, in addition to islets;thisconditioncouldreduceonemajorsourceofMifin regulating M𝜑 and DC activation in T1DM, M𝜑 isolated from this experimental T1DM model. However, we did not observe T1DM Mif−/− mice exhibit reduced T cell activation. a reduction in Mif levels in this model; in contrast, Wt mice Here,weobservedthatWtmicetreatedwithSTZ produced high serum levels of Mif after STZ administration. exhibited high blood glucose levels greater than 400 mg/dL In line with this finding, it is known that the pancreatic but Mif−/−STZmicedevelopedlowerglucoselevelsof islets remaining after treatment with STZ produce high levels approximately 200 mg/dL, in association with lower serum of Mif [21] and that an elevation of Mif secretion precedes levels of proinflammatory cytokines. After reconstituting Mif pancreatic islet death induced by IFN-𝛾,TNF-𝛼,andIL-1𝛽 using exogenous rMif, Mif−/−STZshowedbloodglucose [24]. This evidence establishes that STZ did not influence levels and IL-6 and IL-12 levels similar to those in WtSTZ Mif production/release by pancreatic islets or other cellular from 2 to 6 weeks after STZ treatment. Interestingly, TNF-𝛼 sources, such as T cells, DC, and M𝜑 infiltrating the pancreas. serum levels from Mif−/−STZ+rMif were higher than those The loss of insulin production in T1DM is related to pan- of WtSTZ at all time points analyzed. The reduction levels creatic 𝛽-celldestructionduetoinsulitis[56].Weobserved of blood glucose, IL-6, and IL-12 observed in Mif−/−STZ that WtSTZ developed high serum levels of Mif and low mice,onweek8,probablywerebecausetheanimalsdidnot insulin levels compared to Mif−/−STZ, which expressed receive rMif injections on week 8, as mentioned in Section 2, insulin levels comparable to those in healthy mice. The so the residual effect of rMif from week 1 to week 7 was histological analysis of pancreatic islets showed that WtSTZ insufficient to induce IL-6 and IL-12 levels at 8 week similar displayed100%insulitis,comparedtothe33%insulitis to that observed in the WtSTZ mice in this point, at least observed in Mif −/−STZ mice. These observations confirm Journal of Diabetes Research 11

Spleen dendritic cells 200

∗∗∗ 150

cells ∗∗ + 2 weeks 4 weeks 8 weeks 100 CD80 +

MFI MFI MFI

CD11c 50

CD80 0 0 2468 Weeks after induction (a) Spleen dendritic cells 150 ∗∗ ∗∗ ∗∗∗

cells 100 + 2 weeks 4 weeks 8 weeks CD86 + 50 MFI MFI

MFI CD11c

CD86 0 02468 Weeks after induction (b) Spleen dendritic cells 400 ∗∗∗

300 cells + II ∗∗∗ 2 weeks 4 weeks 8 weeks 200 MHC-

MFI MFI MFI + 100 CD11c

MHC-II 0 0 2468 Weeks after induction (c) Spleen dendritic cells 250 ∗ 200 ∗∗∗ ∗ cells

+ 150 2 weeks 4 weeks 8 weeks TLR2

+ 100

MFI MFI MFI

CD11c 50

TLR-2 0 0 2468 Weeks after induction (d) Figure 6: Continued. 12 Journal of Diabetes Research

Spleen dendritic cells 250 ∗∗∗ ∗ 200 cells

+ 150 2 weeks 4 weeks 8 weeks TLR4

+ 100 MFI MFI MFI

CD11c 50

TLR-4 0 0 2468 Weeks after induction (e)

Figure 6: Mif promotes costimulatory molecules expression on spleen DC. The time course of costimulatory molecule: CD80 (a), CD86 (b), + MHC-II (c), TLR-2 (d), and TLR-4 (e) expression on spleen CD11c DC.DCfromspleenofWtandMif−/− micewereharvestedat0,2,4, 7 and 8 weeks after STZ administration and 1 × 10 cells/mL were processed by flow cytometric analysis as indicated in Section 2. Analyses of expression are shown in the right panels: WtSTZ (◼); Mif−/−STZ (󳶃). And representative histograms of expression based on fluorescence are shown in the left panels. Isotype controls are indicated in gray shadow, the black line represents expression by Mif−/−STZ cells, and the blue line shows expression by WtSTZ cells. The data are presented as the mean fluorescence intensities (MFI) from one representative experiment. ∗ ∗∗ ∗∗∗ Each experiment was repeated three times (𝑛=3) and individually analyzed. 𝑝 < 0.05, 𝑝 < 0.01,or 𝑝 < 0.001, GraphPad Prism software 6.

Pancreatic macrophages 200 ∗ ∗∗∗ 150 cells 2 weeks 4 weeks 8 weeks + 100 CD80 +

MFI MFI MFI F4/80 50

CD80 0 0 2468 Weeks after induction (a) Pancreatic macrophages 200

∗ 150 ∗ cells + 2 weeks 4 weeks 8 weeks 100 CD86 +

MFI MFI MFI F4/80 50

CD86 0 0 2468 Weeks after induction (b)

Figure 7: Continued. Journal of Diabetes Research 13

Pancreatic macrophages 1000 ∗∗∗

∗ 800 cells + II 600 ∗ 2 weeks 4 weeks 8 weeks

MHC- 400 +

MFI MFI MFI F4/80 200

MHC-II 0 0 2468 Weeks after induction (c) Pancreatic macrophages 1500

∗∗

cells 1000 + 2 weeks 4 weeks 8 weeks

TLR2 ∗ + ∗ 500 MFI MFI MFI F4/80

TLR-2 0 0 2468 Weeks after induction (d) Pancreatic macrophages 1000 ∗ 800 cells

2 weeks 4 weeks + 8 weeks 600 ∗∗∗ TLR4

+ 400 MFI MFI MFI

F4/80 200

TLR-4 0 0 2468 Weeks after induction (e)

Figure 7: Mif promotes costimulatory molecule expression on pancreatic M𝜑. The time course of costimulatory molecule: CD80 (a), CD86 + (b), MHC-II (c), TLR-2 (d), and TLR-4 (e) expression on pancreatic F4/80 M𝜑.M𝜑 from pancreas of Wt and Mif −/− micewereharvestedat 7 0, 2, 4, and 8 weeks after STZ administration and 1 × 10 cells/mL were processed by flow cytometric analysis as indicated in Section 2. Analyses of expression are shown in the right panels: WtSTZ (◼); Mif −/−STZ (󳶃). And representative histograms of expression based on fluorescence are shown in the left panels. Isotype controls are indicated in gray shadow, the black line represents expression by Mif−/−STZ cells, and the blue line shows expression by WtSTZ cells. The data are presented as the mean fluorescence intensities (MFI) from one representative ∗ ∗∗ ∗∗∗ experiment. Each experiment was repeated three times (𝑛=3) and individually analyzed. 𝑝 < 0.05, 𝑝 < 0.01,or 𝑝 < 0.001,GraphPad Prism software 6.

that Mif deficiency resulted in pancreatic islet protection, as a chemokine. For example, Mif plays a crucial role in leuko- probably by controlling the functional activity and modu- cyte recruitment and arrest during atherosclerosis develop- lating the secretory capacity of proinflammatory cytokines ment [57]. Therefore, Mif could participate in the process produced by T cells, DC, and M𝜑 that reach the pancreatic of insulitis to promote the production of proinflammatory cells. cytokines, but Mif could also promote leukocyte recruitment Mif has been recognized as a molecule that not only to pancreatic 𝛽-cells. For this reason, Mif −/−STZ exhibited promotes proinflammatory cytokine production but also acts reduced insulitis. 14 Journal of Diabetes Research

Pancreatic dendritic cells

200 ∗∗ ∗∗∗ 150 cells

2 weeks 4 weeks 8 weeks +

CD80 100 +

MFI MFI MFI CD11c 50

CD80 0 0 2468 Weeks after induction (a) 200 ∗ 150

cells ∗ 2 weeks 4 weeks 8 weeks + ∗ 100 CD86 + MFI MFI MFI 50 CD11c

CD86 0 02468 Weeks after induction (b) 600 cells

+ 400 II

2 weeks 4 weeks 8 weeks MHC- + 200 MFI MFI MFI CD11c

MHC-II 0 0 2468 Weeks after induction (c) 1500 ∗ cells

2 weeks 4 weeks 8 weeks + 1000 TLR2 + MFI MFI MFI 500 CD11c TLR-2 0 0 2468 Weeks after induction (d)

Figure 8: Continued. Journal of Diabetes Research 15

800

600 ∗ cells +

400 2 weeks 4 weeks 8 weeks TLR4 +

CD11c 200 MFI MFI MFI 0 TLR-4 0 2468 Weeks after induction (e)

Figure 8: Mif promotes costimulatory molecules expression on pancreatic DC. The time course of costimulatory molecule: CD80 (a), CD86 (b), MHC-II (c), TLR-2 (d), and TLR-4 (e) expression on pancreatic CD11c DC. DC from pancreas of Wt and Mif −/− micewereharvestedat0, 7 2, 4, and 8 weeks after STZ administration and 1 × 10 cells/mL were processed by flow cytometric analysis as indicated in Section 2. Analyses of expression are shown in the right panels: WtSTZ (◼); Mif−/−STZ (󳶃). And representative histograms of expression based on fluorescence are shown in the left panels. Isotype controls are indicated in gray shadow, the black line represents expression by Mif−/−STZ cells, and the blue line shows expression by WtSTZ cells. The data are presented as the mean fluorescence intensities (MFI) from one representative ∗ ∗∗ ∗∗∗ experiment. Each experiment was repeated three times (𝑛=3) and individually analyzed. 𝑝 < 0.05, 𝑝 < 0.01,or 𝑝 < 0.001,GraphPad Prism software 6.

8000 40000 ∗

6000 30000

∗∗ 4000 20000 3H-TdR up-take 3H-TdR 3H-TdR up-take 3H-TdR 2000 10000 in response to OVA (CPM) OVA to in response in response to OVA (CPM) OVA to in response

0 0

Healthy STZ Healthy STZ Wt Wt Mif −/− Mif −/−

(a) (b)

Figure 9: Macrophages and monocytes from Mif−/− mice have impaired ability to induce lymphocyte proliferation. At 8 weeks after STZ + + treatment, Wt (◼)orMif−/− (◻)M𝜑 (F4/80 ) and monocytes (CD11b )primedwith10𝜇g/mL OVA were cocultured with OVA-transgenic T cells for 5 days. Subsequently, [3H]-thymidine was added for 18 h, and [3H]-thymidine incorporation was measured. The values are presented ∗ ∗∗ as means ± SEM counts per minute (CPM) from triplicate wells of three independent experiments (𝑛=8). 𝑝 < 0.05 or 𝑝 < 0.01.GraphPad Prism software 6, GraphPad Prism software 6.

Antigen-presenting cells, M𝜑 and DC, are key mediators Our results demonstrated that both M𝜑 and DC from of the development of T1DM [4]. Moreover, it has been the spleen and pancreas of Mif−/−STZ mice expressed lower proposed that DC orchestrate the autoimmune response in levels of CD80, CD86, MHC-II, TLR-2, and TLR-4 than those T1DM via TLR-2 and TLR-4 [58]. Previous studies by us of WtSTZ. These results demonstrate a role of Mif in the and others have shown that Mif induces the expression of activation of M𝜑 and DC to promote costimulatory molecule costimulatory molecules on M𝜑 andDCinsomepathological expression, which might drive pancreas-specific T cell acti- infections [59–61]. Here, we identified the expression of vation and effector Th1 subset differentiation, processes that costimulatory molecules and TLR-2 and TLR-4 on M𝜑 and have been associated with subsequent pancreatic injury in DC, as well as the ability of M𝜑 to activate T lymphocytes. T1DM. 16 Journal of Diabetes Research

600 Blood glucose 800 IL-6

∗∗∗ ∗∗∗ 600 400

400 (mg/dL) (pg/mL)

200

200

0 0 0 24680 2468 Weeks after STZ administration Weeks after STZ administration WtSTZ WtSTZ Mif −/−STZ Mif −/−STZ (a) (b) 1200 IL-12 10000 TNF-𝛼 ∗

8000 900 ∗

6000 600 (pg/mL)

(pg/mL) ∗∗ 4000

∗∗∗ 300 2000

0 0 0 2468 0 2468 Weeks after STZ administration Weeks after STZ administration WtSTZ WtSTZ Mif −/−STZ Mif −/−STZ (c) (d)

Figure 10: After Mif reconstitution, proinflammatory cytokine production in Mif−/− mice was increased.Mif reconstitution was initiated at the same time of STZ administration to Wt (◼)orMif−/− mice (󳶃) until 7 weeks after STZ treatment. The glucose levels (a) and the levels of the inflammatory cytokines IL-6 (b), TNF-a (c), and IL-12 (d) were measured. Control mice receiving vehicle injection did not display significant changes on cytokine production compared to week 0 (data do not shown). The values are expressed as the means ± SEM of three ∗ ∗∗ ∗∗∗ independent experiments (𝑛=8). 𝑝 < 0.05, 𝑝 < 0.01,or 𝑝 < 0.001, GraphPad Prism software.

In line with the results described above, we observed that Mif−/−STZ mice. This finding is consistent with the evidence + both M𝜑 (F4/80) and monocytes (CD11b ) from WtSTZ were that proliferation of antigen-specific T cells from TCR- more reactive and had greater ability to induce T lymphocyte- transgenic mice is highly dependent on CD28/CD86 costim- specific proliferation in response to OVA than those from ulation. Disrupting this interaction dramatically reduces cell Journal of Diabetes Research 17 proliferation [62]. This conclusion agrees with the result that was supported by Grant IN212215 from PAPIIT (DGAPA, blocking CD86 prevents the development of diabetes in NOD UNAM, Mexico)´ and by Grant 152224 from CONACYT to mice [63]. Moreover, mutations in the MHC-II molecule lead Miriam Rodriguez-Sosa. to development of autoimmune diabetes [64]. By another hand, the differentiation state of M𝜑 is an References important determinant for T cell response in T1DM. Two major populations have been defined, the classically acti- [1] F. S. Malik and C. E. Taplin, “Insulin therapy in children and vated (CA) M𝜑 secreting proinflammatory cytokines such adolescents with type 1 diabetes,” Pediatric Drugs,vol.16,no.2, as TNF-𝛼, IL-6, and IL-1𝛽 and reactive oxygen species; pp. 141–150, 2014. and alternatively activated (AA) M𝜑 which secrete anti- [2] American Diabetes Association, “Diagnosis and classification inflammatory factors including TGF-𝛽 and IL-10. In T1DM of diabetes mellitus,” Diabetes Care, vol. 37, supplement 1, pp. S81–S90, 2013. it has been established that CAM𝜑 trigger inflammatory [3] S. Canivell and R. Gomis, “Diagnosis and classification of responses which initiates insulitis and pancreatic 𝛽 cell death, 𝜑 autoimmune diabetes mellitus,” Autoimmunity Reviews,vol.13, whereas the AAM decreases hyperglycemia, insulitis, and no. 4-5, pp. 403–407, 2014. inflammation in the pancreas [65]. [4]A.Willcox,S.J.Richardson,A.J.Bone,A.K.Foulis,andN. AsMifisaregulatorofmanyproinflammatorycytokines G. Morgan, “Analysis of islet inflammation in human type 1 𝜑 that are characteristic for the CAM ,Mifhasbeenproposed diabetes,” Clinical&ExperimentalImmunology,vol.155,no.2, as CAM𝜑 macrophage-polarizing factor [66]. However, there pp. 173–181, 2009. are few and contrary experimental evidences about how Mif [5] M.E.Craig,S.Nair,H.Stein,andW.D.Rawlinson,“Virusesand might participate in the polarization to AAM𝜑 or CAM𝜑. type 1 diabetes: a new look at an old story,” Pediatric Diabetes, In a mouse model of nonalcoholic fatty liver disease M𝜑 in vol.14,no.3,pp.149–158,2013. liver from Mif−/− mice were skewed toward AAM𝜑 [67]. By [6]A.Salvatoni,A.Baj,G.Bianchi,G.Federico,M.Colombo,and contrary, in the melanoma mouse model M𝜑-derived Mif A. Toniolo, “Intrafamilial spread of enterovirus infections at the participates in AAM𝜑 polarization [68]. Moreover, murine clinical onset of type 1 diabetes,” Pediatric Diabetes,vol.14,no. Mif and filarial nematode parasite (Brugia) Mif protein 6, pp. 407–416, 2013. induced proinflammatory cytokines release. However, Mif [7]G.AfonsoandR.Mallone,“Infectioustriggersintype1 also induce upregulation of IL-4 on bone marrow-derived diabetes: is there a case for epitope mimicry?” Diabetes, Obesity mouse M𝜑,whichwhentreatedin vitro with Mif and IL-4 and Metabolism, vol. 15, supplement 3, pp. 82–88, 2013. induce AAM𝜑 [69]. Here, we show that M𝜑 from Mif−/−STZ [8] A.˚ Lernmark and C.-D. Agardh, “Immunomodulation with mice display reduced proinflammatory cytokine production human recombinant autoantigens,” Trends in Immunology,vol. and exhibit reduced ability to induce T lymphocyte prolifer- 26,no.11,pp.608–612,2005. ation in response to OVA. It is possible that Mif deficiency [9] M. Pearl-Yafe, A. Kaminitz, E. S. Yolcu, I. Yaniv, J. Stein, and N. 𝜑 Askenasy, “Pancreatic islets under attack: cellular and molecular influences on AAM polarization in this model; however, effectors,” Current Pharmaceutical Design,vol.13,no.7,pp.749– more experiments are necessary to establish this. 760, 2007. [10] Y. Dogan, S. Akarsu, B. Ustundag, E. Yilmaz, and M. K. 5. Conclusions Gurgoze, “Serum IL-1𝛽, IL-2, and IL-6 in insulin-dependent diabetic children,” Mediators of Inflammation,vol.2006,Article We show for the first time a role of Mif in promoting ID 59206, 6 pages, 2006. costimulatory molecule expression in M𝜑 and DC in T1DM. [11] B. Baumann, H. H. Salem, and B. O. Boehm, “Anti-inflam- These results reveal Mif as a key regulator of proinflammatory matory therapy in type 1 diabetes,” Current Diabetes Reports,vol. function and M𝜑 and DC activation in T1DM. Although 12, no. 5, pp. 499–509, 2012. more specific experiments are required, there is no doubt that [12] C. Y. Westwell-Roper, J. A. Ehses, and C. B. Verchere, “Resident Mif represents a potential target for anti-Mif therapy, which macrophages mediate islet amyloid polypeptide-induced islet might attenuate the autoimmune process in T1DM. IL-1𝛽 production and 𝛽-cell dysfunction,” Diabetes,vol.63,no. 5, pp. 1698–1711, 2014. Competing Interests [13] T. Mandrup-Poulsen, L. Pickersgill, and M. Y. Donath, “Block- ade of interleukin 1 in type 1 diabetes mellitus,” Nature Reviews The authors have no conflicting financial or commercial Endocrinology,vol.6,no.3,pp.158–166,2010. interests. [14] T. Mandrup-Poulsen, “Interleukin-1 antagonists and other cytokine blockade strategies for type 1 diabetes,” Review of Diabetic Studies,vol.9,no.4,pp.338–347,2012. Acknowledgments [15] X.-L. Zou, Z.-Y. Zhao, Y.-Y. Wan, Z.-Q. Su, and M. Xiang, “Diabetogenic T cells induce autoimmune diabetes in BALB/c This study is part of the work of Yuriko Itzel Sanchez-Zamora´ mice,” Chinese Medical Sciences Journal,vol.23,no.2,pp.88–94, to obtain a Ph.D. degree in the Programa de Doctorado en 2008. Ciencias Biomedicas,´ Universidad Nacional Autonoma´ de [16] J. My´sliwska, M. Smardzewski, N. Marek-Trzonkowska, M. + + Mexico´ (UNAM), and she received a Fellowship (no. 349680) My´sliwiec, and K. Raczynska,´ “Expansion of CD14 CD16 from CONACYT. The authors thank Dr. J. de D. Ruiz- monocytes producing TNF-𝛼 in complication-free diabetes Rosado and Nikhil Reddy for critically reading this paper type 1 juvenile onset patients,” Cytokine,vol.60,no.1,pp.309– (Nationwide Children’s Hospital, OSU, USA). This work 317, 2012. 18 Journal of Diabetes Research

[17] M. Rachmiel, O. Bloch, A. A. Shaul et al., “Young patients with comorbidities,” Mediators of Inflammation,vol.2010,ArticleID both type 1 diabetes mellitus and asthma have a unique IL-12 610479, 7 pages, 2010. and IL-18 secretory pattern,” Pediatric Diabetes,vol.12,no.7,pp. [33]C.Toso,J.A.Emamaullee,S.Merani,andA.M.J.Shapiro, 596–603, 2011. “The role of macrophage migration inhibitory factor on glucose [18] A. E. Altinova, D. Engin, E. Akbay et al., “Association of poly- metabolism and diabetes,” Diabetologia, vol. 51, no. 11, pp. 1937– morphisms in the IL-18 and IL-12 genes with susceptibility to 1946, 2008. type 1 diabetes in Turkish patients,” Journal of Endocrinological [34] J. Nishihira, “Novel pathophysiological aspects of macrophage Investigation,vol.33,no.7,pp.451–454,2010. migration inhibitory factor (review),” International Journal of [19] P. Hanifi-Moghaddam, N. C. Schloot, S. Kappler, J. Seißler, Molecular Medicine,vol.2,no.1,pp.17–28,1998. and H. Kolb, “An association of autoantibody status and serum [35] Y. Sanchez-Zamora, L. I. Terrazas, A. Vilches-Flores et al., cytokine levels in type 1 diabetes,” Diabetes,vol.52,no.5,pp. “Macrophage migration inhibitory factor is a therapeutic target 1137–1142, 2003. in treatment of non-insulin-dependent diabetes mellitus,” The [20] J. Bojunga, K. Kusterer, M. Bacher, R. Kurek, K.-H. Usadel, and FASEB Journal,vol.24,no.7,pp.2583–2590,2010. H. Renneberg, “Macrophage migration inhibitory factor and [36] G. Grieb, M. Merk, J. Bernhagen, and R. Bucala, “Macrophage development of type-1 diabetes in non-obese diabetic mice,” Migration Inhibitory Factor (MIF): a promising biomarker,” Cytokine,vol.21,no.4,pp.179–186,2003. Drug News and Perspectives,vol.23,no.4,pp.257–264,2010. [21] I. Cvetkovic, Y. Al-Abed, D. Miljkovic et al., “Critical role of [37] M. Leech, C. Metz, P. Hall et al., “Macrophage migration macrophage migration inhibitory factor activity in experimen- inhibitory factor in rheumatoid arthritis: evidence of proin- tal autoimmune diabetes,” Endocrinology,vol.146,no.7,pp. flammatory function and regulation by glucocorticoids,” Arthri- 2942–2951, 2005. tis and Rheumatism,vol.42,no.8,pp.1601–1608,1999. [22] S. Stosic-Grujicic, I. Stojanovic, D. Maksimovic-Ivanic et al., [38] S. Onodera, H. Tanji, K. Suzuki et al., “High expression of “Macrophage migration inhibitory factor (MIF) is necessary macrophage migration inhibitory factor in the synovial tissues for progression of autoimmune diabetes mellitus,” Journal of of rheumatoid joints,” Cytokine,vol.11,no.2,pp.163–167,1999. Cellular Physiology,vol.215,no.3,pp.665–675,2008. [39] J. Yun, C.-T. Xu, and B.-R. Pan, “Epidemiology and gene [23] Y. I. Sanchez-Zamora´ and M. Rodriguez-Sosa, “The role of markers of ulcerative colitis in the Chinese,” World Journal of MIF in type 1 and type 2 diabetes mellitus,” Journal of Diabetes Gastroenterology,vol.15,no.7,pp.788–803,2009. Research,vol.2014,ArticleID804519,6pages,2014. [40] P. W. Laird, A. Zijderveld, K. Linders, M. A. Rudnicki, R. [24] I. Stojanovic, T. Saksida, I. Nikolic, F. Nicoletti, and S. Stosic- Jaenisch, and A. Berns, “Simplified mammalian DNA isolation Grujicic, “Macrophage migration inhibitory factor deficiency procedure,” Nucleic Acids Research, vol. 19, no. 15, article 4293, protects pancreatic islets from cytokine-induced apoptosis in 1991. vitro,” Clinical and Experimental Immunology,vol.169,no.2,pp. [41] M. L. Lukic, S. Stoˇsic-Grujicic, and A. Shahin, “Effector mech- 156–163, 2012. anisms in low-dose streptozotocin-induced diabetes,” Develop- mental Immunology,vol.6,no.1-2,pp.119–128,1998. [25] T. Roger, M. P. Glauser, and T. Calandra, “Macrophage migration inhibitory factor (MIF) modulates innate immune [42] M. A. K. Markwell, S. M. Haas, L. L. Bieber, and N. E. responses induced by endotoxin and Gram-negative bacteria,” Tolbert, “A modification of the Lowry procedure to simplify Journal of Endotoxin Research,vol.7,no.6,pp.456–460,2001. protein determination in membrane and lipoprotein samples,” Analytical Biochemistry,vol.87,no.1,pp.206–210,1978. [26]S.N.Babu,G.Chetal,andS.Kumar,“Macrophagemigration inhibitory factor: a potential marker for cancer diagnosis and [43] E. E. Hillhouse, R. Collin, G. Chabot-Roy et al., “Nearby therapy,” Asian Pacific Journal of Cancer Prevention,vol.13,no. construction impedes the progression to overt autoimmune 5, pp. 1737–1744, 2012. diabetes in NOD mice,” Journal of Diabetes Research,vol.2013, ArticleID620313,7pages,2013. [27] J. Nishihira, “Molecular function of macrophage migration [44] D.S.Li,Y.Yuan,H.Tu,Q.Liang,andL.Dai,“Aprotocolforislet inhibitory factor and a novel therapy for inflammatory bowel isolationfrommousepancreas,”Nature Protocols,vol.4,no.11, disease,” Annals of the New York Academy of Sciences,vol.1271, pp.1649–1652,2009. no. 1, pp. 53–57, 2012. [45] A. M. Ortiz-Flores, Y. Ledesma-Soto, E. A. Calleja, M. [28] J. Nishihira and K. Mitsuyama, “Overview of the role of Rodr´ıguez-Sosa, I. Juarez,´ and L. I. Terrazas, “Taenia crassiceps macrophage migration inhibitory factor (MIF) in inflammatory infection does not influence the development of experimental bowel disease,” Current Pharmaceutical Design,vol.15,no.18, rheumatoid arthritis,” BioMed Research International,vol.2013, pp.2104–2109,2009. Article ID 316980, 9 pages, 2013. [29] L. L. Santos and E. F. Morand, “Macrophage migration [46] L. Yu, D. T. Robles, N. Abiru et al., “Early expression of antiin- inhibitory factor: a key cytokine in RA, SLE and atherosclero- sulin autoantibodies of humans and the NOD mouse: evidence sis,” Clinica Chimica Acta,vol.399,no.1-2,pp.1–7,2009. for early determination of subsequent diabetes,” Proceedings of [30] Z. Szekanecz, T. Besenyei, A.´ Szentpetery,´ and A. E. Koch, the National Academy of Sciences of the United States of America, “Angiogenesis and vasculogenesis in rheumatoid arthritis,” vol.97,no.4,pp.1701–1706,2000. Current Opinion in Rheumatology,vol.22,no.3,pp.299–306, [47] A. K. Steck, K. Johnson, K. J. Barriga et al., “Age of islet 2010. autoantibody appearance and mean levels of insulin, but not [31] O. M. Finucane, C. M. Reynolds, F. C. McGillicuddy, and H. GAD or IA-2 autoantibodies, predict age of diagnosis of type M. Roche, “Insights into the role of macrophage migration 1 diabetes: diabetes autoimmunity study in the young,” Diabetes inhibitory factor in obesity and insulin resistance,” Proceedings Care, vol. 34, no. 6, pp. 1397–1399, 2011. of the Nutrition Society,vol.71,no.4,pp.622–633,2012. [48] S. Rex, S. Kraemer, G. Grieb et al., “The role of macrophage [32] R. Kleemann and R. Bucala, “Macrophage migration inhibitory migration inhibitory factor in critical illness,” Mini-Reviews in factor:criticalroleinobesity,insulinresistance,andassociated Medicinal Chemistry, vol. 14, no. 14, pp. 1116–1124, 2014. Journal of Diabetes Research 19

[49] J. D. D. Rosado and M. Rodriguez-Sosa, “Macrophage migra- [64]J.J.Racine,M.Zhang,M.Wang,W.Morales,C.Shen,andD. tion inhibitory factor (MIF): a key player in protozoan infec- Zeng, “MHC-mismatched mixed chimerism mediates thymic tions,” International Journal of Biological Sciences,vol.7,no.9, deletion of cross-reactive autoreactive T cells and prevents pp.1239–1256,2011. insulitis in nonobese diabetic mice,” The Journal of Immunology, [50] K. Kim and H. Kim, “Macrophage migration inhibitory factor: vol.194,no.1,pp.407–417,2015. a potential therapeutic target for rheumatoid arthritis,” The [65] A. Espinoza-Jimenez,A.N.Pe´ on,´ and L. I. Terrazas, “Alter- Korean Journal of Internal Medicine,vol.31,no.4,pp.634–642, natively activated macrophages in types 1 and 2 diabetes,” 2016. Mediators of Inflammation, vol. 2012, Article ID 815953, 10 [51]C.Pfleger,N.C.Schloot,M.D.Brendeletal.,“Circulating pages, 2012. cytokines are associated with human islet graft function in type [66]B.-S.Kim,N.Pallua,J.Bernhagen,andR.Bucala,“The 1diabetes,”Clinical Immunology,vol.138,no.2,pp.154–161,2011. macrophage migration inhibitory factor protein superfamily [52] H. Kolb, “Mouse models of insulin dependent diabetes: low- in obesity and wound repair,” Experimental and Molecular dose streptozocin-induced diabetes and nonobese diabetic Medicine, vol. 47, article e161, 2015. (NOD) mice,” Diabetes/Metabolism Reviews,vol.3,no.3,pp. [67] D. Heinrichs, M.-L. Berres, M. Coeuru et al., “Protective role 751–778, 1987. of macrophage migration inhibitory factor in nonalcoholic [53] A. K. Steck, K. Vehik, E. Bonifacio et al., “Predictors of steatohepatitis,” The FASEB Journal,vol.28,no.12,pp.5136– progression from the appearance of islet autoantibodies to 5147, 2014. early childhood diabetes: the Environmental Determinants of [68] K. Yaddanapudi, K. Putty, B. E. Rendon et al., “Con- Diabetes in the Young (TEDDY),” Diabetes Care,vol.38,no.5, trol of tumor-associated macrophage alternative activation pp.808–813,2015. by macrophage migration inhibitory factor,” The Journal of [54] G. Waeber, T. Calandra, R. Roduit et al., “Insulin secretion Immunology, vol. 190, no. 6, pp. 2984–2993, 2013. 𝛽 is regulated by the glucose-dependent production of islet [69] L. Prieto-Lafuente, W.F. Gregory, J. E. Allen, and R. M. Maizels, cell macrophage migration inhibitory factor,” Proceedings of the “MIF homologues from a filarial nematode parasite synergize National Academy of Sciences of the United States of America, with IL-4 to induce alternative activation of host macrophages,” vol. 94, no. 9, pp. 4782–4787, 1997. Journal of Leukocyte Biology,vol.85,no.5,pp.844–854,2009. [55] G. Waeber, T. Calandra, C. Bonny, and R. Bucala, “A role for the endocrine and pro-inflammatory mediator MIF in the control of insulin secretion during stress,” Diabetes/Metabolism Research and Reviews,vol.15,no.1,pp.47–54,1999. [56]N.G.Morgan,P.Leete,A.K.Foulis,andS.J.Richardson,“Islet inflammation in human type 1 diabetes mellitus,” IUBMB Life, vol. 66, no. 11, pp. 723–734, 2014. [57] E. P. C. van der Vorst, Y. Doring,¨ and C. Weber, “Chemokines and their receptors in Atherosclerosis,” Journal of Molecular Medicine,vol.93,no.9,pp.963–971,2015. [58] D. L. Eizirik, M. L. Colli, and F. Ortis, “The role of inflammation in insulitis and 𝛽-cell loss in type 1 diabetes,” Nature Reviews Endocrinology,vol.5,no.4,pp.219–226,2009. [59] A. B. Stavitsky and J. Xianli, “In vitro and in vivo regulation by macrophage migration inhibitory factor (MIF) of expression of MHC-II, costimulatory, adhesion, receptor, and cytokine molecules,” Cellular Immunology,vol.217,no.1-2,pp.95–104, 2002. [60] C. A. Terrazas, E. Huitron, A. Vazquez et al., “MIF synergizes with Trypanosoma cruzi antigens to promote efficient dendritic cell maturation and IL-12 production via p38 MAPK,” Interna- tional Journal of Biological Sciences,vol.7,no.9,pp.1298–1310, 2011. [61] J. de Dios Ruiz-Rosado, J. E. Olgu´ın, I. Juarez-Avelar´ et al., “MIF promotes classical activation and conversion of inflammatory ℎ𝑖𝑔ℎ Ly6C monocytes into tipdcs during murine toxoplasmosis,” Mediators of Inflammation,vol.2016,ArticleID9101762,18 pages, 2016. [62] B. Salomon and J. A. Bluestone, “Complexities of CD28/B7: CTLA-4 costimulatory pathways in autoimmunity and trans- plantation,” Annual Review of Immunology,vol.19,pp.225–252, 2001. [63] D. J. Lenschow, S. C. Ho, H. Sattar et al., “Differential effects of anti-B7-1 and anti-B7-2 monoclonal antibody treatment on the development of diabetes in the nonobese diabetic mouse,” Journal of Experimental Medicine, vol. 181, no. 3, pp. 1145–1155, 1995. Hindawi Publishing Corporation Journal of Diabetes Research Volume 2016, Article ID 1028401, 7 pages http://dx.doi.org/10.1155/2016/1028401

Research Article Potential of IL-33 for Preventing the Kidney Injury via Regulating the Lipid Metabolism in Gout Patients

Lihua Duan,1 Yan Huang,1 Qun Su,1 Qingyan Lin,1 Wen Liu,1 Jiao Luo,1 Bing Yu,1 Yan He,1 Hongyan Qian,1 Yuan Liu,1 Jie Chen,2 and Guixiu Shi1

1 Department of Rheumatology and Clinical Immunology, The First Hospital of Xiamen University, Xiamen 361003, China 2Department of Immunology, Basic Medicine Science, Medical College, Xiamen University, Xiamen 361005, China

Correspondence should be addressed to Jie Chen; [email protected] and Guixiu Shi; [email protected]

Received 5 May 2016; Accepted 5 July 2016

Academic Editor: Jixin Zhong

Copyright © 2016 Lihua Duan et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Interleukin-33 (IL-33), the most recently discovered member of the IL-1 superfamily, has been linked to several human pathologies including autoimmune diseases, sepsis, and allergy through its specific IL-1 receptor ST2. However, there is little information regarding the role of IL-33 in gout. In this study, we investigated the potential role of IL-33 in gout patients. The serum level of IL-33 was measured by ELISA, and the clinical and laboratory parameters, serum creatinine, urea, and lipid, were extracted from medical record system. The serum IL-33 expression was predominantly increased in gout patients compared to healthy controls, and the IL-33 levels were higher in patients without kidney injury. Furthermore, IL-33 showed a negative correlation with biomarkers of kidney injury, such as CRE and urea. The lipid metabolism dysfunction, tophi, and hypertension are the common reasons for kidney injury in gout. Interestingly, inverse and positive correlation of IL-33 expression was observed in LDL and HDL, respectively. However, there was no significant alteration in the gout patients with hypertension and tophi. These data suggested that IL-33 might act as a protective role in kidney injury through regulating the lipid metabolism in gout.

1. Introduction gout was often accompanied by lipid metabolic dysregula- tion. Previous study has shown that excess triglyceride-rich Gout is the most common metabolic disease in human, which lipoproteins are oxidized by oxidative stress in chronic kidney results from the purine metabolism disorder [1]. Due to loss disease. The prevention of hypertriglyceridemia is one of the of uricase in human, purine metabolism in human and apes most important steps for the treatment of chronic kidney is completely different form that in other mammalian species. disease [7]. Therefore, lipoprotein metabolic disorder might Monosodium urate (MSU) crystals formation results from contribute to the kidney injury in gout patients. serum uric acid concentrations rise above the physiological IL-33 is a member of the IL-1 cytokine family, and the saturation, which deposit in joints and soft tissues leading intracellular pathway of IL-33 signaling is similar to that of to arthritis, soft tissue masses, nephrolithiasis, and urate IL-1 signaling [8]. The IL-1 family of cytokines includes IL- nephropathy [2]. It is reported that 75%–80% of gout is 1𝛼,IL-1𝛽, and IL-18 [9]. IL-33 signaling occurs by binding to associated with metabolism disorder. More than 80% of the ST2, IL-1RAcP, subsequent intracellular signaling pathways patients with hyperlipoproteinemia are accompanied with including MyD88, IRAK (IRAK1, IRAK4), and TRAF6, lead- hyperuricaemia [3]. The finding of hyperlipidaemia in gout ing to the activation of NF-𝜅B and MAPKs (p38, JNK, and patients is common, including an increase in small dense ERK) pathways [10, 11]. IL-33 production is upregulated in LDL cholesterol and a reduction in HDL cholesterol [4]. It inflamed tissues and acts as an early inducer of inflammation was shown that metabolic syndrome including high levels contributing to the further amplification of inflammatory of LDL and TG might be an important factor in the cause responses [12]. Furthermore, recent studies have demon- of chronic kidney disease (CKD) [5, 6]. The pathogenesis of stratedthatIL-33playedacriticalroleinobesity-associated 2 Journal of Diabetes Research

Table 1: Demographic data and clinical characteristics of subjects in the study.

Characteristics Gout patients Healthy control Total 41 44 Male sex 39 (95%) 41 (93%) Age at study mean (SD) years 58.3 (15.55) 56.8 (17.91) Course of disease, mean (range) years 10.6 (0.3–40) — Clinical manifestation, 𝑛 (%) Hyperuricemia 27 (65.8%) — Acute phase 36 (87%) — Tophi 15 (36.5%) — Gouty kidney damage 13 (31.7%) — Hypercholesteremia 18 (43.9%) — Hypertension 21 (51.2%) — Osteoporosis 19 (46.3%) — Fever 4 (9.7%) —

inflammation, atherosclerosis, and metabolic abnormalities gathered. The number and clinical characteristics of healthy through promoting the production of T helper type 2 (Th2) controls and patients with gout were summarized in Table 1. cytokines and polarizing macrophages towards a protective alternatively activated phenotype. 2.2. Detection of Cytokines by ELISA. Four milliliters of Although the detrimental role for IL-33 in the RA and bloodwascollectedinsterilecoagulanttubesandwasthen SpA has been reported, the role of IL-33 in gout arthritis is centrifuged at 3,500 rpm for 5 min at ambient temperature to still unknown. In the present study, we examine serum levels obtain serum, which was immediately frozen and stored at ∘ of IL-33 by ELISA from gout patients, and correlation with −80 C until batch analysis. Concentrations of IL-33 (Pepro- clinical markers in patients with gout was analyzed. We found Tech, USA) were determined by ELISA kits according to the that serum levels of IL-33 in gout patients were significantly manufacturers’ protocols. higher than healthy subjects. Particularly, the IL-33 levels were higher in patients without kidney injury. Previous 2.3. Statistical Analysis. AlldatawereanalyzedinGraphPad studies showed that IL-33 exacerbates acute kidney injury, Prism5. Results are presented as mean ± SEM. The Mann- whereas IL-33 showed a negative correlation with biomarker Whitney 𝑈 test and Spearman’s correlation analysis were used of kidney injury, such as CRE and urea in gout patients. There to calculate significance. Statistical significance was accepted were no significant differences in the gout patients with tophi for 𝑝 values < 0.05. or hypertension which can cause kidney damage. However, an inverse correlation of IL-33 expression was observed in LDL, and a significantly positive correlation was observed 3. Results in HDL. These data suggested that IL-33 might prevent the kidney injury through regulating the lipid metabolism in gout 3.1. Clinical Characteristics of Gout Patients. The clinical patients. characteristics of gout patients (see Table 1) were summa- rized for this study. Forty-one patients with gout and forty- four healthy controls of Southern Chinese population were 2. Materials and Methods enrolled.Themeanageforgoutpatientswas58.3years with range (26–91); there were 39 males and 2 females. The 2.1. Patients and Controls. A total of 41 patients diagnosed mean course of the disease was for 10.6 years with range with gout (2 women and 39 men, age 26–91, mean 58.3 of 0.3–40 years. Among these 41 patients, 36 patients (87%) years) were recruited from the outpatient clinic and ward of had acute phase, 15 patients (36.5%) had tophi in Gout, 27 the Department of Rheumatology and Clinical Immunology, patients (65.8%) had hyperuricemia, 18 patients (43.9%) had the First Hospital of Xiamen University. All patients were hypercholesteremia, 21 patients (51.2%) had hypertension, diagnosedbasedontheAmericanCollegeofRheumatology 13 patients (31.7%) had gouty kidney damage, 19 patients classification criteria [13]. The results were compared with (46.3%) had osteoporosis, and 4 (9.7%) had fever. a population of 44 healthy volunteers (healthy controls) matchedforsexandage.Localethicscommitteeapprovedthe 3.2. Serum IL-33 Levels Are Increased in Gout Patients. It study and informed consent was obtained from patients and has been demonstrated that IL-33 played a critical role in control subjects. Clinical data and several laboratory param- many diseases. However, the role of IL-33 in gout patients eters, such as CRE, urea, HDL, LDL, and UA, of patients were has not been described until now. Currently, compared with Journal of Diabetes Research 3

800 p = 0.027 800 p < 0.0001

600 600

400 400

Serum (pg/mL) IL-33 200 Serum (pg/mL) IL-33 200

0 0 Gout HC Normal Injury (a) (b)

Figure 1: Increased IL-33 expression in gout patients with renal damage. (a) The serum levels of IL-33 in gout patients (𝑛=41) and healthy controls (HC) (𝑛=44) were detected by ELISA; each symbol represented an individual sample and horizontal lines showed median values. (b) The gout patients were divided into two groups, one group with kidney function injury𝑛=13 ( ) and one without (𝑛=28). Mann-Whitney 𝑈 test was conducted to compare the data between two groups.

healthy controls, the serum IL-33 levels in gout patients were 3.5. Potential of IL-33 in the Modulation of Lipid Metabolism. significantly increased𝑝 ( < 0.0001). As shown in Figure 1(a), Furthermore, lipid metabolism disorder often occurred in themedianlevelofIL-33ingoutpatientswas184.6pg/mL gout patients, which caused kidney damage (Figure 3(c)) in with range of 31.3–741.3 pg/mL, while healthy control was gout patients, and recent study has shown that IL-33 plays 65.6 pg/mL (2.16–257.8 pg/mL). Furthermore, we observed aprotectiveroleinatherosclerosis.Thus,thecorrelationof that the gout patients with renal injury showed lower IL- IL-33 with lipid was analyzed. We found that a positive 33 levels (median 114.9 pg/mL: 31.3–447.8) when compared correlation between IL-33 and HDL (𝑟 = 0.41, 𝑝 = 0.007) with gout patients without renal injury (median 299.5 pg/mL: was observed (Figure 4(a)). Consistently, LDL, a critical 57.9–741.3) (𝑝 = 0.027) (Figure 1(b)). lipid for the development of atherosclerosis, was in negative correlation with IL-33 (𝑟 = −0.33, 𝑝 = 0.03) (Figure 4(b)). A negative correlation between IL-33 and TG was observed, 3.3. Association of IL-33 Expression in Gout Patients with although there is no significant difference𝑟=−0.12 ( , 𝑝= Tophi, Hypertension, and HDL. The renal function in gout 0.42) (Figure 4(c)). patients was often damaged by tophi, hyperlipoidemia, and hypertension. Gout patients were grouped according to tophi, hypertension, and HDL. The IL-33 expression of 4. Discussion gout patients with tophi (median 154.0 pg/mL: 57.9–589.1) or not (median 185.7 pg/mL: 31.3–741.3) (𝑝 = 0.635)and In the present study, we demonstrated the IL-33 expression with hypertension (median 150.7 pg/mL: 31.3–741.3) or not anditspotentialroleingoutpatients.SerumIL-33levels (median 196.7 pg/mL: 71.7–614.0) (𝑝 = 0.804)didnotshow were significantly elevated in patients with gout patients a significant difference (Figures 2(a) and 2(b)). Interestingly, when compared with healthy control subjects. Furthermore, when compared with low serum HDL levels, the gout patients theelevatedIL-33levelswereconsiderablyreducedinrenal with high-HDL levels showed an increased IL-33 expression impairment when compared with normal renal function in (328.9 ± 231.5 versus 154.2 ± 97.35 pg/mL, 𝑝 = 0.01). gout patients. There were no significant differences whether thegoutpatientshadtophiorhypertensionwhichcancause kidney damage. However, an inverse correlation of IL-33 3.4. Correlation of Serum IL-33 Levels with Kidney Function expression was observed in LDL, and a significantly positive Indictors in Gout Patients. In addition, the role of IL-33 in correlation was observed in HDL. These data suggested that thegoutpatientswithkidneyinjurywasanalyzed.Spearman’s IL-33 might prevent the kidney injury through regulating the correlation coefficient was used for the assessment of correla- lipidmetabolismingoutpatients. tion between IL-33 levels and CRE or urea. As expected, the It was first demonstrated that IL-33 activates T helper IL-33 levels were negatively correlated with CRE (𝑟 = −0.57, type 2 (Th2) cells and mast cells to secrete Th2 cell- 𝑝 < 0.0001)andurea(𝑟 = −0.49, 𝑝 = 0.0008) (Figures associated proinflammatory cytokines and chemokines [8]. 3(a) and 3(b)). The above results showed that HDL played a Additionally, IL-33 can also induce proinflammatory effects protect role in the development of the kidney injury in gout depending on Th1/Th17 immune response [14, 15]. Previous patients. Here, a negative correlation between HDL and CRE studies have reported that IL-33 acts as an endogenous was depicted in Figure 3(c). “danger signal” or “alarmin” that may alert the immune 4 Journal of Diabetes Research

800 p > 0.05 800 p > 0.05

600 600

400 400 Serum (pg/mL) IL-33

200 Serum (pg/mL) IL-33 200

0 0 No-tophi Tophi No-hypertension Hypertension (a) (b)

800 p = 0.01

600

400

Serum (pg/mL) IL-33 200

0 High-HDL Low-HDL (c)

Figure 2: Difference of IL-33 expression in gout patients with tophi, hypertension, and HDL. The renal function was affected by tophi, hyperlipoidemia, and hypertension in gout patients. The serum levels of IL-33 in patients were grouped according to tophi, hypertension, and HDL. These groups were evaluated by Mann-Whitney 𝑈 test.

system in response to inflammatory diseases [16], such AcLDL/OxLDL uptake and storage of cholesteryl esters and as hypersensitive diseases like asthma [17], autoimmune triglycerides and cholesterol efflux/transport and by inducing diseases like rheumatoid arthritis [18], allergic rhinitis [19], a phenotypical Th1-to-Th2 switch [27]. IL-33 may play a and autoimmune conjunctivitis [20]. IL-33 from necrotic protective role in the development of atherosclerosis via the cells might induce autophagy, which can further balance the induction of IL-5 and ox-LDL antibodies [28]. Low serum effects of increased apoptosis secondary to contrast-induced high-density lipoprotein cholesterol level was an independent nephropathy in diabetic kidney disease [21]. In addition, predictor for gouty flares [29], while gout patients who had IL-33 promotes acute kidney injury through CD4 T cell- higher high density lipoprotein (HDL) levels with metabolic mediated production of CXCL1 [22]. Although the cytokine syndrome were at the lowest risk [30]. In line with above secretion by macrophage can be enhanced by IL-33 [23–25], results, we also observed that serum levels of IL-33 were in our study IL-33 was found increased in gout patients with positively correlated with HDL from gout patients, and the nokidneyinjury.ThereasonforthedualfunctionofIL-33 expressya ion of IL-33 was markedly high in the patient might be the various inflammatory environments in different without kidney damage. As expected, serum IL-33 levels diseases. showed an inverse correlation with CRE, urea. and UA from Itiswellknownthatgoutisassociatedwithcardiovas- gout patients. cular and metabolic diseases, such as hypertension, hyper- MSU has been identified as an endogenous factor for lipidaemia, and diabetes mellitus. 75%–80% of gout patients kidney damage [11]. NLRP3 activation induced by MSU crys- had combined high blood lipoprotein [4, 26]. Interestingly, tal triggers caspase-1-dependent IL-1𝛽 and IL-18 secretion, it has been demonstrated that IL-33 is a potent inhibitor which induces a general inflammatory response including of macrophage foam cell formation in vivo and in vitro recruitment of neutrophils and macrophages to the site of and attenuates atherosclerosis. IL-33 blocks foam cell for- crystal formation. These cytokines and infiltrated immune mation by directly regulating the expression of genes for cells in the kidney interstitial tissue result in the renal Journal of Diabetes Research 5

200 30 r=− p = r = −0.57, p < 0.0001 0.50, 0.0008

150 20 mol/L) mol/L)

100 𝜇 𝜇 Urea ( CRE ( 10 50

0 0 0 200 400 600 800 0 200 400 600 800 Serum IL-33 (pg/mL) Serum IL-33 (pg/mL) (a) (b)

200 r=−0.37, p = 0.01

150

mol/L) 100 𝜇 CRE ( 50

0 0.0 0.5 1.0 1.5 2.0 HDL (𝜇mol/L) (c)

Figure 3: Negative correlation of IL-33 with CRE and urea in gout patients. (a, b) The determination of linear relationships between IL-33 expression and CRE and urea in gout patients was performed by Spearman correlation coefficient. (c) Serum HDL levels showed a negative correlation with Cre. Spearman’s correlation analysis was used to calculate significance.

tubular epithelial cell damage and interstitial fibrosis, and protective role in the kidney injury of gout through regulating the inflammation condition is easy to cause the glomerulitis lipid metabolism. andrenaldysfunction.Althoughthecytokinesecretionby In conclusion, the results presented here suggest that the macrophage can be enhanced by IL-33 [23–25], in our study serum IL-33 could be a sensitive marker for kidney function gout patients with no kidney injury showed a high IL- in gout patients. Interestingly, it is noted that IL-33 was 33 expression. A number of epidemiological studies have associated with HDL and CRE in gout, suggesting that the reported that serum uric acid levels are the high-risk factor IL-33 may play a beneficial role in the pathogenesis of gout. in the pathogenesis of hypertension, and high plasma uric Of course, further studies are required to explore the specific acidlevelsarecommoninpatientswitharterialhyper- regulation mechanism between IL-33 and renal function. These data maybe further suggest a novel approach in treating tension. Hypertension is commonly associated with renal gouty arthritis. vasoconstriction, which also leads to kidney injury, but no correlation between hypertension and IL-33 expression was observed in this study. Furthermore, the spectrum of renal Competing Interests diseases in gout patients includes urate stones, acute uric acid The authors declare no conflict of interests regarding the nephropathy, and chronic urate nephropathy. Chronic urate publication of this paper. nephropathy may be associated with a series of complicating factors, rather than simply with excessive depositions of uric acid within renal tissues, such as microvascular damage from Authors’ Contributions untreated hypertension. There was also no difference caused Lihua Duan, Yan Huang, and Qun Su contributed equally to by the depositions of uric acid. Thus, IL-33 might play a this work. 6 Journal of Diabetes Research

6 2.0 r=0.41, p = 0.007 r=−0.33, p = 0.03

1.5 4 mol/L) mol/L) 𝜇

𝜇 1.0 ( LDL

HDL ( HDL 2 0.5

0.0 0 0 200 400 600 800 0 200 400 600 800 Serum IL-33 (pg/mL) Serum IL-33 (pg/mL) (a) (b)

6 r=−0.12, p = 0.42

4 mol/L) 𝜇

TG ( 2

0 0 200 400 600 800 Serum IL-33 (pg/mL) (c)

Figure 4: The correlation of IL-33 with lipoprotein in gout patients. (a) Serum IL-33 was positively correlated with HDL in gout patients (𝑟 = 0.41, 𝑝 = 0.007). (b, c) Serum IL-33 level was negatively correlated with LDL (𝑟 = −0.33, 𝑝 = 0.03)andTG(𝑟 = −0.12, 𝑝 = 0.42)ingout patients. Serum IL-33 level was not detected to correlate with TG from gout patients. Spearman correlation analysis was conducted.

Acknowledgments [5] J. Chen, P. Muntner, L. L. Hamm et al., “The metabolic syndrome and chronic kidney disease in U.S. adults,” Annals of This work was supported by the National Natural Science Internal Medicine,vol.140,no.3,pp.167–I39,2004. Foundation of China (NSFC 81302564 to LH Duan, NFSC [6]M.Kurella,J.C.Lo,andG.M.Chertow,“Metabolicsyndrome 81471534 to GX Shi, and NSFC 81301786 to J Chen). and the risk for chronic kidney disease among nondiabetic adults,” JournaloftheAmericanSocietyofNephrology,vol.16, no.7,pp.2134–2140,2005. References [7] T. Saito, “Features of hypertriglyceridemia in chronic kidney [1] P. Richette and T. Bardin, “Gout,” The Lancet, vol. 375, no. 9711, disease (CKD),” Nihon Rinsho,vol.71,no.9,pp.1618–1622,2013. pp.318–328,2010. [8] J. Schmitz, A. Owyang, E. Oldham et al., “IL-33, an interleukin- 1-like cytokine that signals via the IL-1 receptor-related protein [2]Y.-F.Qing,Q.-B.Zhang,J.-G.Zhou,andL.Jiang,“Changes ST2 and induces T helper type 2-associated cytokines,” Immu- 𝜅 𝛽 in toll-like receptor (TLR)4-NF B-IL1 signaling in male gout nity,vol.23,no.5,pp.479–490,2005. patients might be involved in the pathogenesis of primary gouty [9] T. Nabe, “Interleukin (IL)-33: new therapeutic target for atopic arthritis,” Rheumatology International,vol.34,no.2,pp.213– diseases,” Journal of Pharmacological Sciences, vol. 126, no. 2, pp. 220, 2014. 85–91, 2014. [3] S. Fujimori, “Gout and hyperuricemia,” Japanese Journal of [10] W.-D. Xu, M. Zhang, Y.-J. Zhang, and D.-Q. Ye, “IL-33 in Clinical Medicine, vol. 59, supplement 8, pp. 341–348, 2001. rheumatoid arthritis: potential role in pathogenesis and ther- [4]S.Takahashi,T.Yamamoto,Y.Moriwaki,Z.Tsutsumi,andK. apy,” Human Immunology,vol.74,no.9,pp.1057–1060,2013. Higashino, “Impaired lipoprotein metabolism in patients with [11] F. Ghaemi-Oskouie and Y. Shi, “The role of uric acid as an primary gout - Influence of alcohol intake and body weight,” endogenous danger signal in immunity and inflammation,” British Journal of Rheumatology,vol.33,no.8,pp.731–734,1994. Current Rheumatology Reports,vol.13,no.2,pp.160–166,2011. Journal of Diabetes Research 7

[12] G. Palmer and C. Gabay, “Interleukin-33 biology with potential [29] A. Mak, R. C.-M. Ho, J. Y.-S.Tan et al., “Atherogenic serum lipid insights into human diseases,” Nature Reviews Rheumatology, profile is an independent predictor for gouty flares in patients vol. 7, no. 6, pp. 321–329, 2011. with gouty arthropathy,” Rheumatology,vol.48,no.3,pp.262– [13]S.L.Wallace,H.Robinson,A.T.Masi,J.L.Decker,D.J. 265, 2009. McCarty,andT.F.Yu,¨ “Preliminary criteria for the classification [30] Y. H. Rho, S. J. Choi, Y. H. Lee et al., “The prevalence of of the acute arthritis of primary gout,” Arthritis and Rheuma- metabolic syndrome in patients with gout: a multicenter study,” tism,vol.20,no.3,pp.895–900,1977. Journal of Korean Medical Science,vol.20,pp.1029–1033,2005. [14]B.P.Leung,D.Xu,S.Culshaw,I.B.McInnes,andF.Y.Liew,“A novel therapy of murine collagen-induced arthritis with soluble T1/ST2,” The Journal of Immunology,vol.173,no.1,pp.145–150, 2004. [15] Y.-S. Hong, S.-J. Moon, Y.-B. Joo et al., “Measurement of interleukin-33 (IL-33) and IL-33 receptors (sST2 and ST2L) in patients with rheumatoid arthritis,” Journal of Korean Medical Science, vol. 26, no. 9, pp. 1132–1139, 2011. [16] C. J. Nile, E. Barksby, P. Jitprasertwong, P. M. Preshaw, and J. J. Taylor, “Expression and regulation of interleukin-33 in human monocytes,” Immunology,vol.130,no.2,pp.172–180,2010. [17] D. Jackson, H. Makrinioti, B. M. Rana et al., “IL-33-dependent Type 2 inflammation during rhinovirus-induced asthma exac- erbations in vivo,” American Journal of Respiratory and Critical Care Medicine,vol.12,pp.1373–1382,1990. [18]C.Li,R.Mu,J.Guoetal.,“GeneticvariantinIL33isassociated with susceptibility to rheumatoid arthritis,” Arthritis Research & Therapy,vol.16,no.2,articleR105,2014. [19] Y. Haenuki, K. Matsushita, S. Futatsugi-Yumikura et al., “A critical role of IL-33 in experimental allergic rhinitis,” The Journal of Allergy and Clinical Immunology,vol.130,no.1,pp. 184–194.e11, 2012. [20] A. Matsuda, Y.Okayama, N. Terai et al., “The role of interleukin- 33 in chronic allergic conjunctivitis,” Investigative Ophthalmol- ogy & Visual Science, vol. 50, no. 10, pp. 4646–4652, 2009. [21] L. Demirtas, K. Turkmen, F. M. Kandemir et al., “The possible role of interleukin-33 as a new player in the pathogenesis of contrast-induced nephropathy in diabetic rats,” Renal Failure, vol. 38, no. 6, pp. 952–960, 2016. [22] A. Akcay, Q. Nguyen, Z. He et al., “IL-33 exacerbates acute kidney injury,” JournaloftheAmericanSocietyofNephrology, vol. 22, no. 11, pp. 2057–2067, 2011. [23] Q. Espinassous, E. Garcia-de-Paco, I. Garcia-Verdugo et al., “IL-33 enhances lipopolysaccharide-induced inflammatory cytokine production from mouse macrophages by regulating lipopolysaccharide receptor complex,” TheJournalofImmunol- ogy,vol.183,no.2,pp.1446–1455,2009. [24]T.Ohno,K.Oboki,N.Kajiwaraetal.,“Caspase-1,caspase-8, and calpain are dispensable for IL-33 release by macrophages,” Journal of Immunology,vol.183,no.12,pp.7890–7897,2009. [25] H. Hayakawa, M. Hayakawa, A. Kume, and S.-I. Tominaga, “Soluble ST2 blocks interleukin-33 signaling in allergic airway inflammation,” The Journal of Biological Chemistry,vol.282,no. 36, pp. 26369–26380, 2007. [26] J.-Y. Choe, S.-H. Park, and S.-K. Kim, “Serum cystatin C is a potential endogenous marker for the estimation of renal function in male gout patients with renal impairment,” Journal of Korean Medical Science,vol.25,no.1,pp.42–48,2010. [27] J. E. McLaren, D. R. Michael, R. C. Salter et al., “IL-33 reduces macrophage foam cell formation,” Journal of Immunology,vol. 185, no. 2, pp. 1222–1229, 2010. [28] A. M. Miller, D. Xu, D. L. Asquith et al., “IL-33 reduces the development of atherosclerosis,” The Journal of Experimental Medicine,vol.205,no.2,pp.339–346,2008. Hindawi Publishing Corporation Journal of Diabetes Research Volume 2016, Article ID 8017571, 6 pages http://dx.doi.org/10.1155/2016/8017571

Review Article Does Hsp60 Provide a Link between Mitochondrial Stress and Inflammation in Diabetes Mellitus?

Joshua Juwono and Ryan D. Martinus

School of Science, University of Waikato, Private Bag 3105, Hamilton 3240, New Zealand

Correspondence should be addressed to Ryan D. Martinus; [email protected]

Received 11 April 2016; Revised 10 June 2016; Accepted 13 June 2016

Academic Editor: Akira Mima

Copyright © 2016 J. Juwono and R. D. Martinus. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The focus of this review is to summarise the known relationships between the expression of heat shock protein 60 (Hsp60) andits association with the pathogenesis of Type 1 and Type 2 diabetes mellitus. Hsp60 is a mitochondrial stress protein that is induced by mitochondrial impairment. It is known to be secreted from a number of cell types and circulating levels have been documented in both Types 1 and 2 diabetes mellitus patients. The biological significance of extracellular Hsp60, however, remains to be established. We will examine the links between Hsp60 and cellular anti- and proinflammatory processes and specifically address how Hsp60 appears to affect immune inflammation by at least two different mechanisms: as a ligand for innate immune receptors andas an antigen recognised by adaptive immune receptors. We will also look at the role of Hsp60 during immune cell activation in atherosclerosis, a significant risk factor during the pathogenesis of diabetes mellitus.

1. Introduction 2. Role of Mitochondrial Hsp60 in Types 1 and 2 Diabetes Mellitus Diabetes mellitus is a spectrum of metabolic disorders characterised by chronic hyperglycaemia and abnormalities Heat shock protein 60 (Hsp60) is a molecular stress protein within the metabolism of proteins, fats, and carbohydrates predominantly localised to the mitochondrion where it is [1]. The two most common forms of diabetes are classified as knowntoplayaroleinthefoldingofproteinsinthe Type 1 and 2. Type 1 also known as insulin-dependent diabetes mitochondrial matrix. Hsp60 is upregulated in response to (IDDM) is characterised by the autoimmune destruction of mitochondrial impairment [6] and is considered to be an the 𝛽-cells of the pancreatic islets which ultimately results indicator of mitochondrial stress. Interestingly, Hsp60 has in loss of insulin production leading to hyperglycaemia [2]. alsobeenshowntobesecretedfromavarietyofmammalian On the other hand, Type 2 diabetes is linked to disorders of celltypes[7–9]andisknowntobefoundatelevated bothinsulinsecretionandinsulinaction[3].Itisbecoming levels in both Types 1 and 2 diabetes mellitus [10]. The increasingly clear that Type 2 diabetes is also associated physiological/pathological consequence of having elevated with a progressive destruction of 𝛽-islet cells by autoimmune levels of Hsp60 in systemic circulation and the cell types processes linked to inflammation [4]. However, the key responsible for secretion of Hsp60 into circulation in diabetes triggers and molecular mechanisms responsible for the loss mellitus is not yet known. of 𝛽-cells have not yet been elucidated. Since mitochondria For many years Hsp60 has been observed in nonobese play a key role in the secretion of insulin from 𝛽-islet cells [5], (NOD) mouse model of diabetes and has been linked to in this review we will examine the role of the mitochondrial playaroleinthedestructionofpancreatic𝛽-islet cells molecular stress protein Hsp60 in the pathogenesis of both caused by the spontaneous development of autoimmune Types 1 and 2 diabetes and the potential links between Hsp60 T-lymphocytes [11]. Several epitopes have been identified expression and inflammation. to have significant anti-Hsp60 T-cells responses, but one 2 Journal of Diabetes Research in particular, p277 peptide, has shown the greatest anti- Even though T-cells targets non-tissue-specific molecules, Hsp60 T-cells responses [12]. The p277 peptide of Hsp60 it may have a greater preference over vulnerable, damaged has been widely studied over the years; it is 24-amino 𝛽-cells, ultimately causing distress to 𝛽-cells, and allowing acid peptide within amino acid residues 427–460 derived the release of tissue-specific antigens [20]. Expression of from monocytes human Hsp60 [13]. Hyperglycaemia and Hsp60 in 𝛽-cellmayalsobeaugmentedbyenvironmental insulitis developed when standard strains of mice, not prone tissue-specific trigger such as viral infection via 𝛾-interferon, to spontaneous diabetes, were immunized with p277 cova- thus causing 𝛽-cells to be targets for anti-Hsp60 T-cells lently conjugated to a foreign immunogenic carrier molecule [20]. [14]. Thus what is the purpose and what can the immune Interestingly, Hsp60 and p277 peptides can also lead to system gain from recognising Hsp60? Hsp60 is a highly protection of 𝛽-cell functions [15, 16]. This protection is conserved protein; therefore both bacterial (foreign) and thought to be due to modulation of the autoimmune process endogenous Hsp60 (self) can act as an antigen for 𝛽-cells [23]. responsible for 𝛽-cell destruction. Therapeutic vaccination At first, production of antibodies against Hsp60 was thought with p277 has been documented to slow and inhibit the to be a mechanism to fight bacterial infection or vaccination destruction of 𝛽-cells both in NOD mice and in humans. [23]; however it was subsequently found that autoantibod- Administration of p277 has been shown to downregulate ies against self-Hsp60 were also found to be associated T-cell reactivity to 𝛽-cell antigens. This process is thought with various autoimmune diseases such as Type 1 diabetes to be associated with a shift in cytokine profile from the [24]. proinflammatory T-helper 1 (Th1) phenotype to the anti- Various studies with human patients have shown that inflammatory T-helper-2 (Th2) phenotype. Thus, p277 has treatment of Type 1 diabetes with p277 epitope can preserve been shown to increase IL-4 and IL-10 secretion and decrease part of the endogenous insulin production by arresting the 𝛾-IF secretion and is thought to be mediated by p277 binding destruction of the 𝛽-cell mass in the pancreas [25, 26]. to TLR2 receptors [17, 18]. ThepaperbyStuartetal.,2012,statesthatanumberof Therefore, it is evident that Hsp60 is able to influence T- Hsp60 peptide epitopes that can bind multiple allelic variants cell responses in two ways: as a ligand of toll-like receptor of the human major histocompatibility complex molecule 2 signalling and as an antigen. But how can T-lymphocytes HLA-DRarecalledpan-DRepitopeswhichcaninducelow target Hsp60 (which is ubiquitous) and p227 epitope be peptide-specific proliferative responses and peptide-specific involved with the destruction and protection of pancreatic 𝛽- production of intracellular cytokines such as interleukin- cells? One explanation for this is molecular mimicry. Hsp60 10 (IL-10), an anti-inflammatory cytokine, and interferon-𝛾 and p277 could possibly mimic a tissue-specific antigen of the (IFN-𝛾) in Type 1 diabetes [26]. This suggests that Hsp60 and 𝛽-islet cells of the pancreas that has a p277-like epitope [19]. A peptides derived from the full-length molecule can induce study by [20] reported that mouse Hsp60 molecules in 𝛽-cells both proinflammatory and anti-inflammatory cytokines. This of NOD mouse are the target of anti-H-p277. However this confirms Hsp60 as an important modulator of inflamma- finding poses the question of how a ubiquitous molecule such tion in Type 1 diabetes mellitus. It should be noted that as Hsp60 can be the target of a tissue-specific autoimmune Hsp60 is also thought to downregulate inflammation via disease? There is no evidence to suggest that tissue-specific activated effector T-cells upregulating Hsp60 and presenting Hsp60 exists, given that there is a fibroblast homologue which their own Hsp60 epitopes to antiergotypic regulatory T-cells contains identical sequence to the pancreatic 𝛽-islet Hsp60 [27]. [20]. However, several suggestions have been proposed in Evidence is also accumulating which seem to suggest order to address this question; Hsp60 is present in a unique that Hsp60 may also be involved in the pathogenesis of way in secretory vesicles of the 𝛽-cells, and in the event of Type 2 diabetes mellitus. A number of studies have shown insulin secretion, these vesicles fuse with the 𝛽-cell mem- elevated levels of Hsp60 protein in systemic circulation in brane causing Hsp60 to be presented on the 𝛽-cell membrane Type 2 diabetes patients. Yuan et al. [10] reported the elevated and/orevensecretedintheabsenceofmitochondrialstress, presence of Hsp60 in both serum and saliva of Type 2 causing differences in secretion characteristic from other diabetics compared to nondiabetic control subjects. Salivary cells [20, 21]. When C9 cells (a diabetogenic NOD T-cell Hsp60wasfoundtobefourfoldhigherintype2diabetics clonethatrespondstop277)areinjectedintoNODmice, compared to nondiabetics, and serum Hsp60 was found they migrate to the pancreas and kidney [20]. Another tobe16-foldhigherthanthesalivaryHsp60inType2 experiment showed that there are 8–12 genes for Hsp60 in diabetics. The presence of Hsp60 as a molecular marker that the vertebrate genome; however, when sequenced, all but represents mitochondrial stress opens up the opportunity one were pseudogenes [22]. All of these experiments suggest for a noninvasive diagnostic route to further investigate the that although 𝛽-cell Hsp60 is not tissue-specific, there may relationship of Hsp60 and diabetes [10]. Evidence has been be tissue-specific posttranslational modifications of Hsp60 documented which shows that when human HeLa cells are [20]. Experiments have also shown how T-cells are still grown in the presence of mitochondrial inhibitors (such as able to proliferate in the presence of 𝛽-cell and absence of sodium azide, hydrogen peroxide, and high glucose) there other antigen-presenting cells suggesting that Hsp60 in 𝛽- is a significant upregulation of Hsp60 at the protein level cell may be processed specifically to that tissue, enabling [28]. This suggests that the increased level of serum Hsp60 the 𝛽-celltopresentitsownHsp60totheT-cells,andulti- detected in Type 2 diabetes mellitus patients might also be mately present specific immunogenic peptide like p277 [20]. duetomitochondrialstress. Journal of Diabetes Research 3

3. Hsp60 and Inflammation develop RNAi based therapeutic strategies to treat Type 2 diabetes clinically. A number of studies suggest that extracellular Hsp60 plays Many studies have also shown that people suffering from a role as a cellular “danger” signal for cellular and humoral Type 1 and Type 2 diabetes have accelerated atherosclerosis immune reactions [29, 30]. In 1997, a hypothesis was pro- and are in greater risk of developing atherosclerosis [47]. posed suggesting that elevated acute-phase/stress reactants Atherosclerosis is a disease where plaque builds up inside (such as Hsp60) and their major cytokine are associated with the arteries and is the cause of a majority of cardiovascular Type 2 diabetes [31]. Since then, many studies have been diseases [48]. Early atherosclerosis is characterised by the conducted looking at circulating markers of inflammation, penetration of agranulocyte or mononuclear cells, in partic- and their association with Type 2 diabetes mellitus [32]. ular monocytes, macrophages, and T-lymphocytes [49]. In Inflammationhasbeenfoundtobeanimportantcausative the late atherosclerosis lesions, T-lymphocytes were seen to factor in the pathogenesis of Type 2 diabetes and insulin be activated, and a substantial proportion of the cells are resistance[33].Thereisalsoanobservableassociation thought to be reacting against Hsp60 [50, 51]. A study done between the pathogenesis of insulin resistance, diabetes and using rabbits immunized with mycobacterial Hsp60 have atherosclerosis, and the activation of innate immune system shown that atherosclerotic lesions can be prevented when the by toll-like receptors (TLRs) [34–36]. Interestingly, a link rabbit’s T-lymphocytes are depleted [52, 53]. On the other between TLR2 and TLR4 polymorphisms and Type 2 diabetes hand, when LDL-receptor deficient mice are introduced to has been documented, suggesting that TLRs may play a theHsp60reactiveT-lymphocytes,themicewereableto causativeroleindiabetes[37,38].Otherstudieshaveshown induce pronounced atherosclerotic vessel wall changes [54]. the increased expression of TLR2 and TLR4 in conventional A study done in 2007 found a correlation between insulin resistance target tissues like skeletal muscle and atherosclerosis and T-cell reactivity to Hsp60 in young males adiposetissueofType2diabetics[39,40]. but not in men aged 50 and above. This suggests that the TLRs are a family of protein that senses the invasion T-cell reactivity to Hsp60 is more prominent in young and of microorganism. This in turn stimulates the TLRs and very early stages of arteriosclerosis [55]. It is thought that T- initiates a range of host defence mechanisms [41]. Each cell reactivity to Hsp60 is less prominent in men age 50 and member of the TLR family is set to recognise a specific over because the majority of the T-cells have already formed pathogen component, and when activated it will create a frombloodtothesiteofinflammationinatherosclerotic signalling cascade that ultimately leads to the production plaques and lymphocytes from peripheral blood may no 𝛽 of cytokines (such as IL-1 ,IL-6,IL-8,monocytechemoat- longer present the specific antigen repertoire of cells in 𝛼 tractant protein-1 (MCP-1), and tumour necrosis factor- vessel walls [55]. This T-cell reactivity to Hsp60 is capable 𝛼 (TNF- )) and adaptive immune response [41]. Ligands for of triggering both innate and adaptive immune responses TLR2 and TLR4 include Hsp60, Hsp70, high mobility group that initiate the earliest inflammatory stage of atherosclerosis, B1 protein, endotoxin, hyaluronan, advanced glycation end and mitochondrial Hsp60 is increasingly being recognised products, and extracellular matrix components [42]. In 2009, as a key autoantigen at the sites of endothelial inflammation it was postulated that the effects of Hsp60 on the innate [56, 57]. However, the mechanisms leading to expression of immune system may be due to the presence of bacterial Hsp60 during the initiation of arteriosclerosis due to T-cell contaminants (LPS, lipopolysaccharide, a major cell wall reactivity to Hsp60 are still not well understood. component of gram negative bacteria) in preparations of the recombinant mammalian Hsp60 preparations [43]; however 4. Conclusion it is now clear that activation of innate immune receptors canbecausedbyHsp60onitsownandnotbyassociated There is a clear association between Hsp60 and Type 1and contaminants [44]. A study done in 2010 by Dasu et al. Type 2 diabetes. In Type 1 diabetes, Hsp60 protein is able to confirmed that expression of TLR2 and TLR4 in Type 2 induce the production of anti-Hsp60 antibodies as a defence diabetics is greatly increased when compared to nondiabetics. mechanism against pathogens; anti-Hsp60 antibodies also Furthermore, due to the increase of TLR2 and TLR4 expres- target endogenous Hsp60 (p277 epitope) and result in the sion, there was a subsequent increase of inflammation, which destruction of 𝛽-islet cells. However, both Hsp60 and p277 𝜅 wasmediatedbynuclearfactorKappaBeta(NF- B) p65 [45]. peptides can also prevent 𝛽-cell destruction by upregulation 𝛽 The concentrations of proinflammatory mediators IL-1 ,IL- of the anti-inflammatory Th2 cytokine pathway. Since the 𝛼 6, IL-8, MCP-1, and TNF- in the serum were also found to loss of 𝛽-islet cells is primarily thought to be driven by a be significantly increased in Type 2 diabetics, compared to proinflammatory Th1 cytokine response, the shift of Th1 to nondiabetics. This novel finding suggests that Hsp60 could be Th2 by Hsp60 and p277 may be involved in attenuation playing modulatory responses in inflammation, a metabolic of Type 1 diabetes mellitus (Figure 1). The high levels of characteristic of Type 2 diabetes, through the activation of Hsp60 found in the serum in Type 2 diabetic may also TLRs. lead to the initiation of proinflammatory cytokines in target Interestingly, antihuman Hsp60 small-hairpin RNAs cells (such as vascular endothelial cells) by interacting with (shRNAs) have been documented to downregulate the TLR2 and TLR4 receptors (Figure 2). Thus, Hsp60 acting as expression of endogenous Hsp60 mRNA 48 hours after a proinflammatory signalling molecule may play a role in transfection in human cells [46]. The study proves that Hsp60 the nonresolved vascular inflammation, which is increasingly can be regulated using RNAi and opens the possibility to being recognised as a feature of Type 2 diabetes. It is 4 Journal of Diabetes Research

T-cell

Nucleus

𝛽-islet cell

Molecular mimicry Destruction of T-cell recognition target cell Cease insulin production Hyperglycemia Nucleus (𝛽-islet cells)

Increase of IL-4 and IL-10 and decrease of 𝛾-IF can reverse 𝛽-islet cell destruction initiated by Mitochondria ↑ Hsp60 proinflammatory Administration of p277 ↑ p277 epitope cytokines

Downregulate T-cell reactivity 𝛽 to -islet cells Nucleus

Stress TLR2 ↑ IL-4 ↓ 𝛾-IF ↑ IL-10

Figure 1: Hsp60 and type 1 diabetes.

Target cells: 𝛽-islet cells

Excretion of Hsp60 into the Interaction of Nucleus serum Hsp60 with Nucleus TLR2 and TLR4

↑ 1𝛽 6 8 1 Mitochondria High levels of IL- , IL- , IL- , MCP- , ↑ Hsp60 Hsp60 in the and TNF-𝛼 serum

Proinflammatory cytokines released into the serum Inflammation

Stress Altered insulin production

Hyperglycemia

Figure 2: Hsp60 and type 2 diabetes. Journal of Diabetes Research 5

suggesting that Hsp60 does indeed play a key regulatory role with a T-cell epitope of the human 65-kDa heat shock protein,” in modulating inflammatory processes in diabetes mellitus Proceedings of the National Academy of Sciences of the United and could also provide a key link between mitochondrial States of America, vol. 88, no. 8, pp. 3088–3091, 1991. stress and inflammation in diabetes mellitus. [14] D. Elias, H. Marcus, T. Reshef, V. Ablamunits, and I. R. Cohen, “Induction of diabetes in standard mice by immunization with Competing Interests the p277 peptide of a 60-kDa heat shock protein,” European Journal of Immunology, vol. 25, no. 10, pp. 2851–2857, 1995. The authors declare that they have no competing interests. [15] I. Raz, D. Elias, A. Avron, M. Tamir, M. Metzger, and I. R. Cohen, “Beta-cell function in new-onset type 1 diabetes Authors’ Contributions and immunomodulation with a heat-shock protein peptide (DiaPep277): a randomised, double-blind, phase II trial,” The Joshua Juwono was responsible for researching the areas Lancet,vol.358,no.9295,pp.1749–1753,2001. coveredinthereviewunderthesupervisionofRyanD. [16] G. Sarikonda, S. Sachithanantham, J. F. Miller, P. P. Pagni, K. Martinus.ThereviewpaperwaswrittenbyJoshuaJuwono T. Coppieters, and M. von Herrath, “The Hsp60 peptide p277 andeditedbyRyanD.Martinus. enhances anti-CD3 mediated diabetes remission in non-obese diabetic mice,” Journal of Autoimmunity,vol.59,pp.61–66,2015. [17] D.-H. Kim, J.-C. Lee, M.-K. Lee, K.-W. Kim, and M.-S. Lee, References “Treatment of autoimmune diabetes in NOD mice by Toll-like receptor 2 tolerance in conjunction with dipeptidyl peptidase 4 [1] K. S. Polonsky, “The past 200 years in diabetes,” The New inhibition,” Diabetologia,vol.55,no.12,pp.3308–3317,2012. England Journal of Medicine,vol.367,no.14,pp.1332–1340,2012. [2] B. O. Roep and T. I. M. Tree, “Immune modulation in humans: [18] G. Nussbaum, A. Zanin-Zhorov, F. Quintana, O. Lider, and I. implications for type 1 diabetes mellitus,” Nature Reviews R. Cohen, “Peptide p277 of HSP60 signals T cells: inhibition of Endocrinology,vol.10,no.4,pp.229–242,2014. inflammatory chemotaxis,” International Immunology,vol.18, no. 10, pp. 1413–1419, 2006. [3]A.B.Olokoba,O.A.Obateru,andL.B.Olokoba,“Type2 diabetes mellitus: a review of current trends,” Oman Medical [19] D. B. Jones, A. F. W.Coulson, and G. W.Duff, “Sequence homo- Journal,vol.27,no.4,pp.269–273,2012. logies between hsp60 and autoantigens,” Immunology Today, [4] B. K. Itariu and T. M. Stulnig, “Autoimmune aspects of type 2 vol.14,no.3,pp.115–118,1993. diabetes mellitus—a mini-review,” Gerontology,vol.60,no.3, [20]O.S.Birk,D.Elias,A.S.Weissetal.,“NODmousediabetes: pp. 189–196, 2014. the ubiquitous mouse hsp60 is a 𝛽-cell target antigen of auto- [5] P. Maechler, “Mitochondrial function and insulin secretion,” immune T cells,” Journal of Autoimmunity,vol.9,no.2,pp.159– Molecular and Cellular Endocrinology,vol.379,no.1-2,pp.12– 166, 1996. 18, 2013. [21] K. Brudzynski, V.Martinez, and R. S. Gupta, “Secretory granule [6] M. W.Pellegrino, A. M. Nargund, and C. M. Haynes, “Signaling autoantigen in insulin-dependent diabetes mellitus is related to the mitochondrial unfolded protein response,” Biochimica et 62 kDa heat-shock protein (hsp60),” Journal of Autoimmunity, Biophysica Acta—Molecular Cell Research,vol.1833,no.2,pp. vol. 5, no. 4, pp. 453–463, 1992. 410–416, 2013. [22] T. J. Venner, B. Singh, and R. S. Gupta, “Nucleotide sequences [7]S.GuptaandA.A.Knowlton,“ExosomedependentHSP60 and novel structural features of human and Chinese hamster secretion in rat cardiac myocytes,” JournalofMolecularand hsp60 (chaperonin) gene families,” DNA and Cell Biology,vol. Cellular Cardiology,vol.40,no.6,pp.875–876,2006. 9, no. 8, pp. 545–552, 1990. [8] A. M. Merendino, F. Bucchieri, C. Campanella et al., “Hsp60 is [23] S. H. Kaufmann, “Immunity to bacteria,” Current Opinion in actively secreted by human tumor cells,” PLoS ONE,vol.5,no. Immunology,vol.2,no.3,pp.353–359,1989. 2, article e9247, 2010. [24] F. J. Quintana and I. R. Cohen, “The HSP60 immune system [9] S. Swaroop, N. Sengupta, A. R. Suryawanshi, Y. K. Adlakha, network,” Trends in Immunology,vol.32,no.2,pp.89–95,2011. and A. Basu, “HSP60 plays a regulatory role in IL-1𝛽-induced [25] V. A. L. Huurman, P. E. Van Der Meide, G. Duinkerken et al., microglial inflammation via TLR4-p38 MAPK axis,” Journal of “Immunological efficacy of heat shock protein 60 peptide Dia- Neuroinflammation,vol.13,no.1,article27,2016. Pep2776 therapy in clinical type I diabetes,” Clinical and Experi- [10] J. Yuan, P. Dunn, and R. D. Martinus, “Detection of Hsp60 in mental Immunology,vol.152,no.3,pp.488–497,2008. saliva and serum from type 2 diabetic and non-diabetic control subjects,” Cell Stress & Chaperones,vol.16,no.6,pp.689–693, [26] A. A. V. Stuart, W. de Jager, M. R. Klein et al., “Recognition of 2011. heat shock protein 60 epitopes in children with type 1 diabetes,” Diabetes/Metabolism Research and Reviews,vol.28,no.6,pp. [11] O. S. Birk, D. C. Douek, D. Elias et al., “A role of Hsp60 in 527–534, 2012. autoimmune diabetes: analysis in a transgenic model,” Proceed- ings of the National Academy of Sciences of the United States of [27] F. J. Quintana, A. Mimran, P. Carmi, F. Mor, and I. R. Cohen, America,vol.93,no.3,pp.1032–1037,1996. “HSP60 as a target of anti-ergotypic regulatory T cells,” PLoS [12] R. Abulafia-Lapid, D. Elias, I. Raz, Y. Keren-Zur, H. Atlan, ONE,vol.3,no.12,ArticleIDe4026,2008. andI.R.Cohen,“Tcellproliferativeresponsesoftype1dia- [28] L. Hall and R. D. Martinus, “Hyperglycaemia and oxidative betes patients and healthy individuals to human hsp60 and its stress upregulate HSP60 & HSP70 expression in HeLa cells,” peptides,” Journal of Autoimmunity,vol.12,no.2,pp.121–129, SpringerPlus,vol.2,no.1,pp.1–10,2013. 1999. [29]S.K.Calderwood,S.S.Mambula,P.J.GrayJr.,andJ.R. [13]D.Elias,T.Reshef,O.S.Birk,R.vanderZee,M.D.Walker,and Theriault, “Extracellular heat shock proteins in cell signaling,” I. R. Cohen, “Vaccination against autoimmune mouse diabetes FEBS Letters,vol.581,no.19,pp.3689–3694,2007. 6 Journal of Diabetes Research

[30] J. J. De Haan, M. B. Smeets, G. Pasterkamp, and F. Arslan, [47]A.S.D.M.Matheus,L.R.M.Tannus,R.A.Cobas,C.C.S. “Danger signals in the initiation of the inflammatory response Palma, C. A. Negrato, and M. D. B. Gomes, “Impact of diabetes after myocardial infarction,” Mediators of Inflammation,vol. on cardiovascular disease: an update,” International Journal of 2013, Article ID 206039, 13 pages, 2013. Hypertension, vol. 2013, Article ID 653789, 15 pages, 2013. [31]J.C.Pickup,M.B.Mattock,G.D.Chusney,andD.Burt, [48] I. Tabas, G. Garc´ıa-Cardena,˜ and G. K. Owens, “Recent insights “NIDDM as a disease of the innate immune system: association into the cellular biology of atherosclerosis,” TheJournalofCell of acute-phase reactants and interleukin-6 with metabolic syn- Biology,vol.209,no.1,pp.13–22,2015. drome X,” Diabetologia,vol.40,no.11,pp.1286–1292,1997. [49]Q.B.Xu,G.Oberhuber,M.Gruschwitz,andG.Wick,“Immu- [32] J. C. Pickup, “Inflammation and activated innate immunity in nology of atherosclerosis: cellular composition and major histo- the pathogenesis of type 2 diabetes,” Diabetes Care,vol.27,no. compatibility complex class II antigen expression in aortic 3, pp. 813–823, 2004. intima, fatty streaks, and atherosclerotic plaques in young [33] K. E. Wellen and G. S. Hotamisligil, “Inflammation, stress, and and aged human specimens,” Clinical Immunology and Immu- diabetes,” The Journal of Clinical Investigation,vol.115,no.5,pp. nopathology,vol.56,no.3,pp.344–359,1990. 1111–1119, 2005. [50] M. Benagiano, M. M. D’Elios, A. Amedei et al., “Human 60- kDa heat shock protein is a target autoantigen of T cells derived [34] L. K. Curtiss and P. S. Tobias, “Emerging role of Toll-like from atherosclerotic plaques,” The Journal of Immunology,vol. receptors in atherosclerosis,” JournalofLipidResearch,vol.50, 174, no. 10, pp. 6509–6517, 2005. pp. S340–S345, 2009. [51]A.J.Curry,I.Portig,J.C.Goodall,P.J.Kirkpatrick,andJ.S.H. [35] H. Shi, M. V. Kokoeva, K. Inouye, I. Tzameli, H. Yin, and J. Gaston, “T lymphocyte lines isolated from atheromatous plaque S. Flier, “TLR4 links innate immunity and fatty acid-induced contain cells capable of responding to Chlamydia antigens,” insulin resistance,” The Journal of Clinical Investigation,vol.116, Clinical and Experimental Immunology,vol.121,no.2,pp.261– no. 11, pp. 3015–3025, 2006. 269, 2000. [36] F. S. Wong and L. Wen, “Toll-like receptors and diabetes,” [52] B. Metzler, M. Mayr, H. Dietrich et al., “Inhibition of arte- Annals of the New York Academy of Sciences, vol. 1150, pp. 123– riosclerosis by T-cell depletion in normocholesterolemic rabbits 132, 2008. immunized with heat shock protein 65,” Arteriosclerosis, Throm- [37] R. A. Bagarolli, M. J. A. Saad, and S. T. O. Saad, “Toll-like recep- bosis, and Vascular Biology,vol.19,no.8,pp.1905–1911,1999. tor 4 and inducible nitric oxide synthase gene polymorphisms [53] Q. Xu, H. Dietrich, H. J. Steiner et al., “Induction of arte- are associated with Type 2 diabetes,” Journal of Diabetes and Its riosclerosis in normocholesterolemic rabbits by immunization Complications, vol. 24, no. 3, pp. 192–198, 2010. with heat shock protein 65,” Arteriosclerosis, Thrombosis, and [38] E. Gulden¨ and L. Wen, “Toll-like receptor activation in Vascular Biology,vol.12,no.7,pp.789–799,1992. immunity vs. Tolerance in autoimmune diabetes,” Frontiers in [54] J. George, A. Afek, B. Gilburd, Y. Shoenfeld, and D. Harats, Immunology, vol. 5, article 119, 2014. “Cellular and humoral immune responses to heat shock protein [39] S. J. Creely, P. G. McTernan, C. M. Kusminski et al., “Lipopoly- 65 are both involved in promoting fatty-streak formation in saccharide activates an innate immune system response in LDL-receptor deficient mice,” Journal of the American College humanadiposetissueinobesityandtype2diabetes,”American of Cardiology,vol.38,no.3,pp.900–905,2001. JournalofPhysiology—EndocrinologyandMetabolism,vol.292, [55] M. Knoflach, S. Kiechl, B. Mayrl et al., “T-cell reactivity no. 3, pp. E740–E747, 2007. against HSP60 relates to early but not advanced atherosclerosis,” [40] S. M. Reyna, S. Ghosh, P. Tantiwong et al., “Elevated toll-like Atherosclerosis,vol.195,no.2,pp.333–338,2007. receptor 4 expression and signaling in muscle from insulin- [56] C. Grundtman, S. B. Kreutmayer, G. Almanzar, M. C. Wick, and resistant subjects,” Diabetes,vol.57,no.10,pp.2595–2602,2008. G. Wick, “Heat shock protein 60 and immune inflammatory [41] K. Takeda, T. Kaisho, and S. Akira, “Toll-like receptors,” Annual responses in atherosclerosis,” Arteriosclerosis, Thrombosis, and Review of Immunology,vol.21,pp.335–376,2003. Vascular Biology,vol.31,no.5,pp.960–968,2011. [42] M.-F. Tsan and B. Gao, “Endogenous ligands of Toll-like recep- [57] C. Wick, “Tolerization against atherosclerosis using heat shock tors,” Journal of Leukocyte Biology,vol.76,no.3,pp.514–519, protein 60,” Cell Stress and Chaperones, vol. 21, no. 2, pp. 201–211, 2004. 2016. [43] M.-F. Tsan and B. Gao, “Heat shock proteins and immune sys- tem,” JournalofLeukocyteBiology,vol.85,no.6,pp.905–910, 2009. [44] B.Henderson,S.K.Calderwood,A.R.M.Coatesetal.,“Caught with their PAMPs down? The extracellular signalling actions of molecular chaperones are not due to microbial contaminants,” Cell Stress & Chaperones,vol.15,no.2,pp.123–141,2010. [45]M.R.Dasu,S.Devaraj,S.Park,andI.Jialal,“IncreasedToll- Like Receptor (TLR) activation and TLR ligands in recently diagnosed type 2 diabetic subjects,” Diabetes Care,vol.33,no. 4, pp. 861–868, 2010. [46] T. J. Corydon, J. Hansen, P. Bross, and T. G. Jensen, “Down- regulation of Hsp60 expression by RNAi impairs folding of medium-chain acyl-CoA dehydrogenase wild-type and dis- ease-associated proteins,” Molecular Genetics and Metabolism, vol. 85, no. 4, pp. 260–270, 2005. Hindawi Publishing Corporation Journal of Diabetes Research Volume 2016, Article ID 7965317, 10 pages http://dx.doi.org/10.1155/2016/7965317

Research Article Analysis of Inflammatory Mediators in Prediabetes and Newly Diagnosed Type 2 Diabetes Patients

Zhen Wang,1 Xu-Hui Shen,2 Wen-Ming Feng,3 Guo-fen Ye,4 Wei Qiu,5 and Bo Li6

1 Department of Clinical Medicine, School of Nursing & Medicine, Huzhou University, Huzhou, Zhejiang 313000, China 2Department of Nursing Medicine, School of Nursing & Medicine, Huzhou University, Huzhou, Zhejiang 313000, China 3Surgery Department, Huzhou Wu Xing People’s Hospital, Huzhou, Zhejiang 313008, China 4Physical Examination Center, Zhebei Clinical Medicine Hospital, Huzhou University, Huzhou, Zhejiang 313000, China 5Endocrinology Department, Zhebei Clinical Medicine Hospital, Huzhou University, Huzhou, Zhejiang 313000, China 6Huanzhu Street Community Health Center, Huzhou, Zhejiang 313000, China

Correspondence should be addressed to Xu-Hui Shen; [email protected]

Received 12 April 2016; Accepted 5 June 2016

Academic Editor: Jixin Zhong

Copyright © 2016 Zhen Wang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This study evaluated the inflammatory markers in prediabetes and newly diagnosed type 2 diabetes mellitus (T2DM). Inflammatory markers levels were analyzed using one-way analysis of covariance and the association with prediabetes or T2DM risks was examined by logistic regression models. Our data showed increased levels of hypersensitivity C-reactive protein (hs-CRP), interleukin (IL-4), IL-10, and tryptase in prediabetes subjects and hs-CRP, immunoglobulin E (IgE), IL-4, and IL-10 in T2DM subjects. We concluded that Hs-CRP, IgE, IL-4, IL-10, and tryptase were positively associated with prediabetes or T2DM. Further large prospective studies are warranted to assess a temporal relation between inflammatory biomarkers and incidence of prediabetes or T2DM and its associated chronic diseases.

1. Introduction Many single nucleotide polymorphisms (SNPs) in var- ious genes including those of pro- and anti-inflammatory Type 2 diabetes mellitus (T2DM) is a complex disease in cytokines have been reported as a risk for T2DM [1–3]. But which both genetic and environmental factors interact in not all SNPs have been confirmed by unifying results in 𝛽 determining impaired -cell insulin secretion and peripheral different studies and wide variations have been reported in insulin resistance [1–3]. T2DM is also a metabolic disorder variousethnicgroups[2–4].GeneticpolymorphismsofC- between pro- and anti-inflammatory characterized by reactive protein (CRP) and their association with prediabetes chronic hyperglycemia and increased or decreased levels of and T2DM have been widely studied [5]. IL-6 was regarded circulating cytokines [4]. The rise in the proinflammatory as a kind of pro- and anti-inflammatory factor; one of the cytokines (e.g., interleukin- (IL-) 1, IL-6, tumor necrosis common polymorphisms in the IL-6 gene promoter (C-174G) 𝛼 factor- (TNF-) , C-reactive protein (CRP), transforming was considered as risk factors for T2DM development [2]. IL- 𝛽 growth factor- (TGF-) , and leptin) or the fall in anti- 10 is an anti-inflammatory cytokine protecting against T2DM inflammatory cytokines (e.g., interleukin-1 receptor and inflammation; several variants in the IL-10 gene promoter antagonist (IL-1Ra), IL-4, IL-10, IL-13, and adiponectin) region have been identified and showed the association with is the essential step in glucotoxicity and lipotoxicity induced the development of T2DM [3, 4]. It is reported that TNF-𝛼 mitochondrial injury, oxidative stress, and beta cell apoptosis is a possible mediator of insulin resistance and diabetes since in T2DM [2–4]. These pro- and anti-inflammatory cytokines it inhibits insulin signaling and impairs its secretion [2–4]. canenhanceinsulinresistancedirectlyinadipocytes,muscle, One of the SNPs in TNF-𝛼 gene showed a twofold increase and hepatic cells, leading to systemic disruption of insulin in transcriptional activity and an association of TNF-𝛼 SNPs sensitivity and impaired glucose homeostasis [5]. with T2DM [2]. Genetic variants in some inflammatory 2 Journal of Diabetes Research factors associated with T2DM may provide a rationale for 2.2. Data Collection. Trained staff interviewed participants further studying their roles as biomarkers for disease early using a self-designed questionnaire to obtain information risk prediction and therapeutic targets for T2DM and related on demographic characteristics and anthropometric and complications. lifestyle variables. Physical activity assessments were per- Although accumulating evidences support the patholog- formed using self-reported Total Energy Expenditure Ques- ical role of inflammatory cytokines in T2DM, most stud- tionnaire (TEEQ), a nine-step scale where every step was ies only focused on a few specific inflammatory factors assignedafixedvalueintermsofmultipleofMetabolic and were done in relatively small population groups in a Energy Turnover (MET) [8]. Anthropometric measurements particular ethnic group. In our previous studies, we have (body height and weight, waist circumference (WC), hip explored the associations between few several inflammatory circumference, and blood pressure) were attained at the initial cytokines (e.g., CRP, immunoglobulin E (IgE), chymase, and screening visit. Two sitting blood pressure measurements tryptase) with prediabetes and T2DM and concluded that were taken for each participant using a mercury sphygmo- IgE and CRP are risk factors of prediabetes and T2DM only manometer according to a standard protocol. The mean of basedonasmallsampleofcross-sectionalstudydesign[6, these two blood pressure measurements was used in the data 7]. analysis. In this study, we extend the analysis of the role of pro- The biochemical parameters, including FPG, 2 h OGTT, and anti-inflammatory cytokines in prediabetes and newly HbA1c, fasting insulin, plasma total cholesterol (TC), triglyc- diagnosed T2DM based on a larger sample of cross-sectional eride (TG), low-density lipoprotein cholesterol (LDL-c), study design, by measuring the levels of several proin- high-density lipoprotein cholesterol (HDL-c), hs-CRP, and flammatory cytokines (e.g., IgE, hs-CRP, IL-6, TNF-𝛼,and IgE, were measured in the Clinical Biochemistry Unit of tryptase) and anti-inflammatory cytokines (e.g., IL-4, IL-10, Huzhou First Hospital, a teaching hospital of Huzhou Uni- and forkhead/winged helix transcription factor 3+ (Foxp3+)) versity. The cytokine concentrations, including IL-6,𝛼 TNF- , in prediabetes and T2DM, and correlate them with other tryptase, IL-4, IL-10, and Foxp3+, were measured in patients laboratory and clinical biochemical indicators. Our find- sera using commercially available ELISA double antibody ings will lend support to the hypothesis that some specific sandwich method assays (R&D Systems), performed accord- inflammatory factors may play an etiological role in the ing to the manufacturer’s instructions. Detection kits were pathogenesis of T2DM and also will offer new insights into produced by Wuhan Gene Biotech Co., Ltd. the potential clinical value of these inflammatory factors as biomarkers for disease early risk prediction and therapeutic 2.3. Clinical Criteria. Diabetes and prediabetes were grouped targets for T2DM and related complications. accordingtoAmericanDiabetesAssociation2010(ADA 2010) criteria or HbA1c-based diagnosed criteria [9, 10]. Diabetes was classified with a fasting plasma glucose (FPG) ≥ 2. Materials and Methods 7.0 mmol/L or 2 h OGTT ≥ 11.1 mmol/L or HbA1c ≥ 6.5%, whereas prediabetes was defined as FPG ≥ 5.6 and 2.1. Study Population. The study is part of the Diabetes <7.0 mmol/L or 2 h OGTT ≥ 7. 8 a n d <11.1 mmol/L or HbA1c ≥ Intervention Project (DIP) started in 2012 from School of 6.0% and HbA1c < 6.5%. Subjects were classified as having a Nursing and Medicine, Huzhou University, Zhejiang, China. normal glucose profile if FPG < 5.6 mmol/L and 2 h OGTT < From March to December 2012, a total of 7054 rural residents 7. 8 m m o l / L o r H b A 1 c < 6.0%. aged 50–75 years from eight rural communities in the city of Huzhou participated in physical examination. Based on < 2.4. Statistical Analysis. The mean and standard deviation the criteria of fasting plasma glucose (FPG) 5.6 mmol/L, ± 898 (12.73%) were classified into prediabetes. During July (mean SD) of continuous and normal distributional vari- to August 2013, after excluding subjects with known DM or ables and median and quartile range of continuous but receiving hypoglycaemic medications or with cardiovascular skewed distributional variables were used. Body mass index (BMI) was calculated as weight in kilo- disease, cerebrovascular disease, malignant disease, chronic 2 liverdisease,orkidneyfailure,orthoseundermedications, grams divided by the square of the height in meters (kg/m ). 825 cases of 898 prediabetes subjects and another 300 Waist-to-hip ratio (WHR) was calculated as waist divided by randomized sampling normal glucose subjects were invited hip circumference. Homeostasis model assessment-insulin forFPG,2-houroralglucosetolerancetest(2hOGTT),and resistance (HOMA-IR) was calculated as value of FPG × value haemoglobin A1c (HbA1c) tests as part of the prediabetes and of fasting insulin/22.5 and homeostasis model assessment- T2DM screening. But only 560 subjects would participate 𝛽 cell function (HOMA-𝛽) was calculated as 20 × value of in this study, according to the following clinical criteria fasting insulin/(FPG-3.5). based on FPG, 2 h OGTT, and HbA1c test: 219 (39.11%) BasedonChina2006BloodPressureControlCriteria hadnormalglucosetolerance(NGT),215(38.39%)were and China Prevention and Treatment Classification Recom- diagnosed as prediabetes subjects, and 126 (22.5%) were mendation on Dyslipidemia [11], hypertension was defined diagnosedasT2DMsubjects.Thisstudywasapprovedby as systolic blood pressure (SBP) and/or diastolic blood pres- the Huzhou City Ethics Committee and all subjects gave sure (DBP) ≥ 140/90 mmHg or as receiving blood pressure written, informed consent prior to participating in the lowering medications; high TG was defined as a fasting study. plasma TG ≥ 1.70 mmol/L, low HDL-C as a fasting HDL-C Journal of Diabetes Research 3

≤ 0.9 mmol/L, high TC as TC ≥ 5.72 mmol/L, and low LDL-C subjects (0.5 mg/L) (𝑃 < 0.05). Mean levels of tryptase as a fasting LDL-C ≤ 3.64mmol/L.BasedonChinaObesity in PDG subjects (mean ± SD: 1545.36 ± 291.45 ng/mL) or Task Group Recommendation [12], overweight was classified T2DMG subjects (mean ± SD: 1524.96 ± 286.65 ng/mL) were 2 when a body mass index (BMI) ≥ 24 kg/m and obesity was significantly higher compared with NGG subjects (mean ± 2 classified when a body mass index (BMI) ≥ 29 kg/m . SD: 1481.21 ± 271.44 ng/mL) (𝑃 < 0.05). Mean levels of IL-4 In analyzing the relationships of inflammatory cytokines in PDG subjects (mean ± SD: 303.70±56.84 ng/L) or T2DMG to prediabetes or T2DM, IgE was classified as normal/abnor- subjects (mean ± SD: 304.45 ± 55.87 ng/L) were significantly mal according to upper quartiles (P75 = 34.0 IU/L); CRP was higher compared with NGG subjects (mean ± SD: 283.45 ± classified as normal/abnormal according to upper quartiles 52.50 ng/L) (𝑃 < 0.05). Mean levels of IL-10 in PDG subjects (P75 = 1.0 mg/L); tryptase was classified as normal/abnormal (mean ± SD: 244.03 ± 53.34 ng/L) or T2DMG subjects (mean according to upper quartiles (P75 = 1699.94 ng/mL) and ± SD: 265.04±40.42 ng/L) were significantly higher compared TNF-𝛼 was classified as normal/abnormal according to with NGG subjects (mean ± SD: 229.23 ± 46.54 ng/L) (𝑃< upper quartiles (P75 = 325.14 pg/mL); IL-4 was classified 0.05). But there were no significant differences in IL-6, TNF- as normal/abnormal according to upper quartiles (P75 = 𝛼, and Foxp3+ (Table 1). 323.91 ng/L); IL-10 was classified as normal/abnormal accord- Spearman correlation analysis between inflammatory ingtoupperquartiles(P75=256.22ng/L);FOXP3+was cytokines and the traditional cardiovascular factors indicated classified as normal/abnormal according to upper quartiles that whether, in NGG, the PDG, or the T2DMG, there (P75 = 411.17 pg/mL). were significant relationships between most of the tradi- One-way analysis of covariance was used to test for dif- tional cardiovascular factors and between most of inflam- ferences in continuous distributional variables between three matory cytokines. However, there were seldom significant groups. The nonparametric Kruskal-Wallis test was used to relationships between the traditional cardiovascular factors test for differences in continuous but skewed distributional and inflammatory cytokines (Tables 2–4; Tables S1–S6) (in 2 variables between two groups, and 𝜒 testwasusedtotest Supplementary Material available online at http://dx.doi.org/ differences in proportions between three groups. 10.1155/2016/7965317). Spearman correlation coefficients were calculated to eval- After adjusting confounding factors by multivariable uate associations between inflammatory cytokines and the combined with inflammatory biomarkers mutually adjusted traditional cardiovascular factors. model, plasma IgE was associated with T2DM compared with NGG (OR (odds ratio): 3.46 (1.92–6.23, 95% CI), 𝑃< Binary logistic regression model was used to estimate 0.001 theoddsratios(ORs)and95%CIsforprediabetesand ) or compared with PDG (OR: 2.57 (1.54–4.30, 95% CI), 𝑃 < 0.001); plasma CRP was associated with PDG (OR: T2DM according to dichotomy of inflammatory biomarkers 𝑃 < 0.001 concentration, using the lower values as the referent category. 2.30 (1.46–3.62, 95% CI), )orT2DM(OR:16.24 (7.99–33.02, 95% CI), 𝑃 < 0.001) compared with NGG or Considering the potential confounding factors, we applied 𝑃 < 0.001 10 models to assess the association between biomarkers and T2DM (OR: 5.93 (3.29–10.72, 95% CI), )compared prediabetes or T2DM, including unadjusted mode, age and with PDG; serum IL-4 was associated with PDG (OR: 1.68 (1.05–2.69, 95% CI), 𝑃 = 0.031)orT2DM(OR:1.94(1.10– sex-adjusted model, multivariable (age, sex, BMI, WHR, SBP, 𝑃 = 0.023 DBP,TC, TG, level of physical activity, dietary intake, alcohol 3.53, 95% CI), ) compared with NGG; serum IL- 10 was associated with PDG (OR: 2.13 (1.33–3.41, 95% CI), intake,smokingstatus,presenceorabsenceoffamilyhistory 𝑃 = 0.002 𝑃< of diabetes, hypertension, heart disease, stroke, and hyperc- ) or T2DM (OR: 8.62 (4.50–16.49, 95% CI), 0.001) compared with NGG or T2DM (OR: 3.04 (1.79–5.13, holesterolemia) adjusted model, and multivariable combined 𝑃 < 0.001 with inflammatory biomarkers mutually adjusted model. 95% CI), ) compared with PDG (Table 5). Serum tryptase was associated with prediabetes (OR: 1.53 (1.01–2.32, All statistical analysis was conducted using SPSS statisti- 95% CI), 𝑃 = 0.044) before adjustment and (OR: 1.63 (1.02– cal software (version 19.0). 2.61, 95% CI), 𝑃 = 0.041) after adjustment by multivariable + IL-4, compared with NGG. Serum IL-6, TNF-𝛼,andFoxp3+ 3. Results were not associated with PDG or T2DM (Table 5). The descriptive characteristics of 560 study participants were presented separately for participants with normal glucose 4. Discussion tolerance, prediabetes, and T2DM (Table 1). A total of 215 were prediabetes, 126 were T2DM, and 219 were normal OurdatashowedincreasedlevelsofCRP,IL-4,IL-10,and glucose subjects. Overall, except smoking status, alcohol use, tryptase in prediabetes subjects and increased levels of CRP, disease family history, physical activity level, and dietary IL-4, and IL-10 in T2DM subjects compared with normal intake, all traditional vascular risk factors were worse in the glucose subjects. Not all inflammatory cytokines in our study PDGortheT2DMGpatientsthanNGGsubjects.Median were in agreement with previous findings. levels of IgE in T2DMG subjects (40 IU/L) were significantly With the associations between inflammatory cytokines higher compared with NGG subjects (18 IU/L) (𝑃 < 0.05) (e.g., IL-6, TNF-𝛼, and hs-CRP) and insulin resistance, and with PDG subjects (20 IU/L) (𝑃 < 0.05). Median levels prediabetes or T2DM had been widely researched [13– of CRP in PDG subjects (0.8 mg/L) or T2DMG subjects 21]. Most studies found that subjects with insulin resis- (1.5 mg/L) were significantly higher compared with NGG tance, prediabetes, or T2DM had increased levels of IL-6, 4 Journal of Diabetes Research

Table 1: Baseline characteristics of 560 subjects with different glucose status grouped according to fasting, postload glucose levels, and HbA1c.

Variable NGG (𝑛 = 219)PDG(𝑛 = 215)T2DMG(𝑛 = 126) 𝑃 value ∗ Age (years) 58.13 ± 6.37 59.31 ± 6.86 59.37 ± 7.07 0.122 Sex (%, male) 47.0 40.5 40.5 0.310 2 ∗ † ‡ BMI (kg/m ) 23.58 ± 3.17 25.11 ± 3.51 24.55 ± 3.15 <0.001 ∗ † ‡ WC (cm) 79.92 ± 9.19 85.00 ± 9.89 83.85 ± 8.90 <0.001 ∗ † ‡ WHR 0.84 ± 0.06 0.87 ± 0.05 0.87 ± 0.06 <0.001 Smoking status (%) 0.071 Current 63.9 28.8 7.3 — Never 73.0 19.1 7.9 — Past 63.0 31.7 7.9 — Alcohol use (in the past 12 months) (%, drinker) 25.1 17.2 22.2 0.130 History of diabetes (%, yes) 6.8 12.6 29.4 <0.001 History of myocardial infarction (%, yes) 0.9 0.9 2.4 0.493 History of high blood pressure (%, yes) 15.1 14.0 18.3 0.689 History of stroke (%, yes) 4.6 5.6 4.8 0.731 History of dyslipidemia (%, yes) 8.7 6.5 13.5 0.360 ∗ Physical activity (MET/week) 440.47 ± 146.79 447.92 ± 151.39 427.49 ± 152.15 0.480 Dietary intake (%, excess energy intake) 74.9 72.1 73.0 0.800 Carbohydrate intake (%) 0.740 <55% of total energy intake 56.6 54.9 50.8 — 55–65% of total energy intake 24.2 24.2 23.8 — >65% of total energy intake 19.2 20.9 25.4 — Fat intake (%) 0.005 <20% of total energy intake 11.9 20.0 23.8 — 20–25% of total energy intake 41.1 38.1 46.8 — >25% of total energy intake 47.0 49.1 29.4 — Protein intake (%) 0.529 <60% of total energy intake 16.9 11.2 12.7 60–65% of total energy intake 29.7 31.6 31.7 >65% of total energy intake 53.4 57.2 55.6 ∗∗ † ‡, FPG (mmol/L) 5.17 (4.94–5.34) 5.97 (5.74–6.20) 7.72 (7.35–8.89) § <0.001 ∗ † ‡, 2hOGTT(mmol/L) 5.35 ± 1.08 7.64 ± 1.48 12.16 ± 1.55 § <0.001 ∗ ‡, HbA1c (%) 5.06 ± 0.80 5.20 ± 1.03 6.64 ± 1.64 § <0.001 ∗∗ † ‡, Fasting insulin (mU/L) 7.0 (6.55–7.36) 7.36 (6.89–8.70) 10.10 (8.54–13.7) § <0.001 ∗∗ † ‡, HOMA-IR 1.60 (1.42–1.73) 2.04 (1.81–2.32) 3.64 (2.93–4.58) § <0.001 ∗∗ † ‡, HOMA-𝛽 84.14 (72.72–99.79) 60.27 (52.46–70.64) 48.67 (32.96–66.35) § <0.001 ∗ † ‡ SBP (mmHg) 121.74 ± 20.01 130.94 ± 23.61 130.49 ± 18.90 <0.001 ∗ † DBP (mmHg) 76.46 ± 9.97 80.53 ± 11.29 77.83 ± 8.31§ <0.001 ∗ † ‡ TC (mmol/L) 4.85 ± 0.84 5.00 ± 0.90 5.24 ± 1.12 0.001 ∗∗ † ‡ TG (mmol/L) 1.33 (1.04–1.94) 1.72 (1.30–2.44) 1.95 (1.31–3.17) <0.001 ∗ † ‡ HDL-c (mmol/L) 1.29 ± 0.32 1.20 ± 0.30 1.16 ± 0.28 <0.001 ∗ LDL-c (mmol/L) 2.85 ± 0.75 2.80 ± 0.92 2.90 ± 1.04 0.613 ∗∗ ‡, IgE (IU/L) 18 (7–34) 20 (7–42.5) 40 (16–64) § <0.001 ∗∗ † ‡, hs-CRP (mg/L) 0.5 (0.3–1.0) 0.8 (0.4–2.0) 1.50 (1.00–2.40) § <0.001 ∗ IL-6 (ng/L) 58.56 ± 12.66 60.92 ± 12.82 60.39 ± 10.28 0.117 Journal of Diabetes Research 5

Table 1: Continued.

Variable NGG (𝑛 = 219)PDG(𝑛 = 215)T2DMG(𝑛 = 126) 𝑃 value ∗ TNF-𝛼 (pg/mL) 284.94 ± 61.17 290.13 ± 68.22 295.37 ± 57.25 0.326 ∗ † ‡ Tryptase (ng/mL) 1481.21 ± 271.44 1545.36 ± 291.45 1524.96 ± 286.65 0.036 ∗ † ‡ IL-4 (ng/L) 283.45 ± 52.50 303.70 ± 56.84 304.45 ± 55.87 <0.001 ∗ † ‡, IL-10 (ng/L) 229.23 ± 46.54 244.03 ± 53.34 265.04 ± 40.42 § <0.001 ∗ Foxp3+ (pg/mL) 364.22 ± 72.27 367.26 ± 24.46 356.18 ± 83.12 0.418 T2DM: type 2 diabetes mellitus; NGG: normal glucose group; PDG: prediabetes group; T2DMG: T2DM group; BMI: body mass index; WC: waist circumference; WHR: waist hip ratio; HOMA-IR: homeostasis model assessment-insulin resistance; HOMA-𝛽: homeostasis model assessment-𝛽 cell function; FPG: fasting plasma glucose; 2 h OGTT: 2-hour oral glucose tolerance test; HbA1c: haemoglobin A1c; SBP: systolic blood pressure; DBP: diastolic blood pressure; TC: total cholesterol; TG: triglyceride; LDL-c: low-density lipoprotein cholesterol; HDL-c: high-density lipoprotein cholesterol; IgE: immunoglobulin E; hs- CRP: hypersensitivity C-reactive protein; TNF-𝛼: tumor necrosis factor; IL-6: interleukin-6; IL-4: interleukin-4; IL-10: interleukin-10; Foxp3+: forkhead/winged helix transcription factor 3+. ∗ ∗∗ † Variable is described using mean and standard deviation. Variable is described using median and interquartile range. 𝑃 <0.05, PDG compared with the ‡ NGG group; 𝑃 <0.05, DMG compared with the NGG group. §𝑃 <0.05, T2DMG compared with the PDG group.

Table 2: Spearman correlation coefficients between cardiovascular factors in normal glucose group.

TC TG HDL CLDLc HOMA-IR HOMA-𝛽 BMI WHR SBP TC 1 ∗∗ TG 0.321 1 ∗∗ ∗∗ HDL C −0.298 −.0363 1 ∗∗ LDL 0.878 0.098 0.107 1 ⋅ HOMA-IR 0.062 0.042 −0.120 0.036 1 ⋅⋅ ∗ ∗ HOMA-𝛽−0.094 −0.137 0.096 −0.102 −0.140 1 ∗∗ ∗∗ ∗∗ ∗∗ ∗∗ BMI 0.203 0.374 −0.279 0.209 0.062 −0.213 1 ∗∗ ∗∗ ∗∗ WHR 0.029 0.363 −0.386 0.054 0.069 −0.112 0.552 1 ⋅ ∗∗ ∗ ∗ ∗∗ ∗∗ SBP 0.184 0.144 −0.014 0.161 0.053 −0.124 0.284 0.252 1 ∗ ∗∗ ∗ ∗∗ ∗∗ ∗∗ DBP 0.166 0.220 −0.105 0.148 0.057 −0.074 0.406 0.335 0.716 TC: total cholesterol; TG: triglyceride; LDL-c: low-density lipoprotein cholesterol; HDL-c: high-density lipoprotein cholesterol; HOMA-IR: homeostasis model assessment-insulin resistance; HOMA-𝛽: homeostasis model assessment-𝛽 cell function; BMI: body mass index; WHR: waist hip ratio; SBP: systolic blood pressure; DBP: diastolic blood pressure. ∗ ∗ ∗∗ Data on all subjects without missing values for all of these variables. 𝑃 <0.05. 𝑃 <0.01.

TNF-𝛼, and hs-CRP [13–16]. A study reported that high study demonstrated that plasma IgE levels strongly corre- levels of inflammatory cytokines appeared in early stage of lated with T2DM (Table 5), which was in agreement with T2DM and were capable of predicting the development of our previous studies [6, 7]. Further large population-based type 2 diabetes through diminishing insulin sensitivity [14]. prospectivestudiesarewarrantedtoassesstheroleofIgEin But a recent study showed decreased levels of IL-6 and T2DM. TNF-𝛼 in T2DM compared to healthy controls [17]. The IL-4 was major anti-inflammatory cytokine and had been author in this paper interpreted that this discrepancy could proposed to play a crucial role in the pathophysiology of be attributed to duration of the diseases, small sample size, T2DM, and several candidate genes have been identified [22]. andthedifferencesinageandsexofthestudiedgroups[17]. Our data showed that serum IL-4 was associated with PDG Elevated plasma levels of hs-CRP in prediabetes and T2DM or T2DM (Tables 1 and 5). subjects have been demonstrated in this paper (Tables 1 and IL-10 was anti-inflammatory cytokine and has been 5), our previous studies, and other studies [6, 7, 14–21]. Our shown to improve impaired insulin signaling [23], prevent data showed that mean levels of TNF-𝛼 or IL-6 in PDG the development of IL-6 [24], inhibit NADPH oxidase [25], or T2DMG subjects were higher but insignificant compared prevent pancreatic beta cell destruction [26], and play an with NGG subjects (Tables 1 and 5), which could be explained important role in modulation of cardiovascular insulin resis- by the differences of some incompletely measured residual tance [23–26]. Two studies indicated that low IL-10 level was confounding factors or inflammation-related diseases (e.g., associated with high HbA1c and serum IL-10 level may be rheumatoid arthritis (RA) or osteoarthritis (OA)) or allergies one of the predictors of glycemia [27, 28]. Increased level of in the studied groups. IL-10 in prediabetes and T2DM subjects in our study was in IgE stimulation during allergic reactions and infections disagreement with these findings [2, 26–28], which could be is the natural defense mechanism. It also plays a crucial explained by high levels of inflammatory cytokines appearing role in the pathophysiology of T2DM [6, 7]. This current in early stage of T2DM [14, 17]. 6 Journal of Diabetes Research

Table 3: Spearman correlation coefficients between inflammatory markers in normal glucose group.

IgE CRP IL-4 IL-6 IL-10 Foxp3+ Tryptase IgE 1 CRP 0.117 1 ∗∗ IL-4 0.059 −0.174 1 ⋅⋅ ∗∗ IL-6 0.026 0.06 0.303 1 ∗∗ ∗∗ IL-10 0.087 −0.012 0.272 0.259 1 ∗∗ ∗∗ ∗∗ Foxp3+ 0.071 −0.066 0.211 0.245 0.430 1 ⋅ ∗∗ ∗∗ ∗∗ ∗∗ Tryptase −0.009 −0.106 0.176 0.188 0.302 0.458 1 ∗ ∗∗ TNF-𝛼−0.052 −0.173 0.057 0.049 0.08 0.085 0.212 IgE: immunoglobulin E; hs-CRP: hypersensitivity C-reactive protein; IL-4: interleukin-4; IL-6: interleukin-6; IL-10: interleukin-10; Foxp3+: forkhead/winged helix transcription factor 3+; TNF-𝛼: tumor necrosis factor. ∗ ∗ ∗∗ Data on all subjects without missing values for all of these variables. 𝑃 < 0.05. 𝑃 < 0.01.

Table 4: Spearman correlation coefficients between inflammatory markers and cardiovascular factors in normal glucose group.

IgE CRP IL-4 IL-6 IL-10 Foxp3 Tryptase TNF-𝛼 ∗ ∗ TC −0.087 0.007 0.038 −0.083 −0.031 −0.016 0.138 0.163 TG 0.020 0.122 0.048 0.023 0.073 0.052 0.077 0.111 ∗ ∗ HDL-c 0.017 −0.146 −0.085 −0.136 −0.031 0.050 0.065 0.082 LDL-c −0.043 −0.001 0.089 −0.001 −0.026 −0.027 0.115 0.111 HOMA-IR 0.012 0.102 0.093 −0.003 0.007 −0.114 −0.085 −0.068 HOMA-𝛽 0.030 0.092 −0.006 0.053 −0.055 −0.047 0.014 −0.117 BMI 0.037 0.001 −0.006 −0.040 0.026 0.057 0.028 0.114 ∗ WHR −0.038 −0.042 0.146 0.106 0.108 0.126 0.095 0.085 SBP 0.091 0.024 0.000 −0.071 −0.003 0.041 0.072 0.054 DBP 0.132 0.043 0.045 −0.024 −0.045 0.026 −0.028 0.020 TC: total cholesterol; TG: triglyceride; LDL-c: low-density lipoprotein cholesterol; HDL-c: high-density lipoprotein cholesterol; HOMA-IR: homeostasis model assessment-insulin resistance; HOMA-𝛽: homeostasis model assessment-𝛽 cell function; BMI: body mass index; WHR: waist hip ratio; SBP: systolic blood pressure; DBP: diastolic blood pressure; IgE: immunoglobulin E; hs-CRP: hypersensitivity C-reactive protein; IL-4: interleukin-4; IL-6: interleukin-6; IL-10: interleukin-10; Foxp3+: forkhead/winged helix transcription factor 3+; TNF-𝛼: tumor necrosis factor. ∗ ∗ Data on all subjects without missing values for all of these variables. 𝑃 < 0.05.

T regulatory (Treg) cells are a unique population of T- could exist between these inflammatory cytokines or between cells which play a crucial role in the maintenance of self- cardiovascular factors [1–3, 20, 21]. tolerance and suppression of potentially inflammatory T- Results of our study extend other evidences linking cells [29]. Chronic low-grade inflammation in obesity and inflammatory cytokines to insulin resistance and risk of dia- impaired insulin sensitivity has been associated with fewer betes based on a larger sample of cross-sectional study design. Tregs in adipose tissue [29–31]. Our study did not prove the Ourfindingsfurtherlendsupporttothehypothesisthatsome associations between FOXP3+ with the PDS or T2DM (Tables specific inflammatory factors may play an important role in 1and5). the pathogenesis of T2DM and also offer new insights into Mast cell tryptase is an important protease that has been the potential clinical value of these inflammatory factors as implicated in cardiovascular diseases [32–34]. Our previous biomarkers for T2DM early prediction and treatment [13–17]. study did not demonstrate the association between plasma Our study has several limitations. We assessed the associ- tryptase with prediabetes or T2DM [6, 7]. But in this paper, ations between few inflammatory cytokines and prediabetes our data indicated that tryptase was mildly associated with and T2DM by means of plasma or serum biomarker only prediabetes (OR: 1.63 (1.02–2.61, 95% CI), 𝑃 = 0.041)after based on cross-sectional study design and cannot demon- adjustment by multivariable + IL-4, compared with NGG, strate that altered plasma or serum levels of inflammatory which could be explained for the synergistic effect of IL-4. biomarkers are predictors of incident diabetes by large Our data also showed that there were significant relation- prospective cohort study. Our comprehensive assessment of ships between most of the traditional cardiovascular factors diabetes risk factors allowed statistical control for important and between most of inflammatory cytokines. However, there confounding factors in the pathogenesis of diabetes, but were seldom significant relationships between the traditional residual confounding could remain in the analysis. In par- cardiovascular factors and inflammatory cytokines (Tables 2– ticular, we did not consider potential inflammation-related 4; Tables S1–S6), which indicated that complex interactions diseases, such as RA or OA, which widely exist in the senior, Journal of Diabetes Research 7

Table 5: ORs of prediabetes and T2DM according to the inflammatory markers.

OR (95% CI) PDG versus NGG 𝑃 DMG versus NGG 𝑃 DMG versus PDG 𝑃 IgE † Model 1 (unadjusted) 1.34 (0.89–2.03) 0.166 3.05 (1.93–4.84) <0.001 2.28 (1.45–3.28) <0.001 ‡ Model2(ageandsex) 1.35 (0.89–2.65) 0.156 3.11 (1.95–4.95) <0.001 2.29 (1.46–3.59) <0.001 Model 3 (multivariable)§ 1.41 (0.89–2.24) 0.147 3.51 (1.96–6.29) <0.001 2.55 (1.52–4.25) <0.001 Model 4 (multivariable + CRP) 1.36 (0.85–2.19) 0.198 3.79 (1.89–7.73) <0.001 2.58 (1.49–4.46) 0.001 Model 5 (multivariable + tryptase) 1.40 (0.88–2.23) 0.156 3.48 (1.94–6.25) <0.001 2.52 (1.51–4.23) <0.001 Model 6 (multivariable + IL-6) 1.40 (0.88–2.22) 0.158 3.51 (1.96–6.29) <0.001 2.54 (1.52–4.25) <0.001 Model 7 (multivariable + TNF-𝛼) 1.42 (0.89–2.89) 0.146 3.49 (1.95–6.26) <0.001 2.55 (1.52–4.25) <0.001 Model 8 (multivariable + IL-4) 1.32 (0.83–2.11) 0.244 3.44 (1.91–6.18) <0.001 2.55 (1.53–4.26) <0.001 Model 9 (multivariable + IL-10) 1.30 (1.81–2.07) 0.283 2.72 (1.45–5.10) 0.002 2.33 (1.38–3.93) 0.002 Model 10 (multivariable + Foxp3) 1.39 (0.87–2.21) 0.167 3.46 (1.92–6.23) <0.001 2.57 (1.54–4.30) <0.001 CRP † Model 1 (unadjusted) 2.35 (1.57–3.52) <0.001 11.76 (6.90–20.03) <0.001 5.01 (3.00–8.37) <0.001 ‡ Model2(ageandsex) 2.32 (1.55–3.49) <0.001 11.70 (6.84–20.02) <0.001 5.11 (3.04–8.57) <0.001 Model 3 (multivariable)§ 2.34 (1.49–3.67) <0.001 16.14 (7.99–32.59) <0.001 5.73 (3.19–10.29) <0.001 Model 4 (multivariable + IgE) 2.31 (1.47–3.64) <0.001 17.23 (8.17–36.33) <0.001 5.81 (3.19–10.59) <0.001 Model 5 (multivariable + tryptase) 2.31 (1.47–3.63) <0.001 16.17 (7.99–32.70) <0.001 5.81 (3.23–10.48) <0.001 Model 6 (multivariable + IL-6) 2.34 (1.49–3.68) <0.001 16.30 (7.99–32.59) <0.001 5.79 (3.21–10.43) <0.001 Model 7 (multivariable + TNF-𝛼) 2.36 (1.50–3.72) <0.001 17.30 (8.43–35.52) <0.001 5.75 (3.20–10.34) <0.001 Model 8 (multivariable + IL-4) 2.23 (1.41–3.51) 0.001 17.34 (8.39–35.86) <0.001 5.87 (3.24–10.61) <0.001 Model 9 (multivariable + IL-10) 2.25 (1.42–3.54) 0.001 17.35 (8.24–35.74) <0.001 5.79 (3.16–10.59) <0.001 Model 10 (multivariable + Foxp3) 2.30 (1.46–3.62) <0.001 16.24 (7.99–33.02) <0.001 5.93 (3.29–10.72) <0.001 Tryptase † Model 1 (unadjusted) 1.53 (1.01–2.32) 0.044 1.15 (0.70–1.88) 0.588 0.75 (0.46–1.21) 0.237 ‡ Model2(ageandsex) 1.55 (0.99–2.29) 0.055 1.21 (0.68–1.85) 0.653 0.75 (0.46–1.21) 0.236 Model 3 (multivariable)§ 1.57 (0.99–2.50) 0.057 1.35 (0.73–2.47) 0.336 0.75 (0.44–1.29) 0.305 Model 4 (multivariable + IgE) 1.56 (0.98–2.45) 0.06 1.32 (0.66–2.36) 0224 0.78 (0.45–1.36) 0.485 Model 5 (multivariable + CRP) 1.52 (0.95–2.44) 0.082 1.34 (0.66–2.73) 0.417 0.71 (0.40–1.25) 0.235 Model 6 (multivariable + IL-6) 1.22 (0.74–2.00) 0.436 1.39 (0.74–2.61) 0.306 0.83 (0.47–1.47) 0.527 Model 7 (multivariable + TNF-𝛼) 1.47 (0.90–2.38) 0.126 1.28 (0.68–2.41) 0.440 0.71 (0.40–1.27) 0.246 Model 8 (multivariable + IL-4) 1.63 (1.02–2.61) 0.041 1.26 (0.67–2.30) 0.484 0.73 (0.42–1.26) 0.263 Model 9 (multivariable + IL-10) 1.36 (0.84–2.19) 0.213 0.93 (0.47–1.84) 0.841 0.64 (0.37–1.12) 0.117 Model 10 (multivariable + Foxp3) 1.50 (0.92–2.44) 0.107 1.21 (0.58–2.51) 0.612 0.77 (0.41–1.43) 0.399 IL-6 † Model 1 (unadjusted) 1.37 (0.91–2.08) 0.136 0.85 (0.51–1.42) 0.534 0.62 (0.37–1.03) 0.062 ‡ Model2(ageandsex) 1.36 (0.90–2.07) 0.146 0.86 (0.51–1.44) 0.560 0.62 (0.37–1.03) 0.062 Model 3 (multivariable)§ 1.37 (0.86–2.19) 0.188 0.97 (0.51–1.85) 0.922 0.68 (0.39–1.18) 0.169 Model 4 (multivariable + IgE) 1.36 (0.85–2.18) 0.202 1.01 (0.52–1.96) 0.985 0.68 (0.40–1.20) 0.188 Model 5 (multivariable + CRP) 1.37 (0.85–2.22) 0.193 0.99 (0.46–2.11) 0.977 0.64 (0.36–1.18) 0.159 Model 6 (multivariable + tryptase) 1.22 (0.74–2.00) 0.436 0.88 (0.45–1.73) 0.720 0.72 (0.40–1.29) 0.264 Model 7 (multivariable + TNF-𝛼) 1.41 (0.88–2.27) 0.157 0.95 (0.50–1.82) 0.876 0.67 (0.38–1.17) 0.161 Model 8 (multivariable + IL-4) 1.25 (0.77–2.02) 0.374 0.85 (0.43–1.65) 0.624 0.62 (0.35–1.12) 0.111 Model 9 (multivariable + IL-10) 1.19 (0.73–1.93) 0.483 0.68 (0.33–1.39) 0.287 0.56 (0.31–1.00) 0.05 Model 10 (multivariable + Foxp3) 0.68 (0.39–1.20) 0.183 0.91 (0.47–1.75) 0.769 0.68 (0.39–1.20) 0.183 TNF-𝛼 † Model 1 (unadjusted) 1.29 (0.85–2.97) 0.239 1.24 (0.76–2.02) 0.390 0.96 (0.59–1.53) 0.866 ‡ Model2(ageandsex) 1.31 (0.85–2.00) 0.218 1.26 (0.77–2.07) 0.358 0.96 (0.59–1.55) 0.867 Model 3 (multivariable)§ 1.43 (0.89–2.29) 0.140 1.29 (0.70–2.38) 0.420 1.04 (0.60–1.81) 0.883 Model 4 (multivariable + IgE) 1.42 (0.89–2.89) 0.146 1.26 (0.67–2.38) 0.480 1.04 (0.59–1.83) 0.885 Model 5 (multivariable + CRP) 1.46 (0.90–2.37) 0.127 1.83 (0.87–3.85) 0.111 0.94 (0.52–1.68) 0.829 8 Journal of Diabetes Research

Table 5: Continued. OR (95% CI) PDG versus NGG 𝑃 DMG versus NGG 𝑃 DMG versus PDG 𝑃 Model 6 (multivariable + tryptase) 1.27 (0.77–2.09) 0.347 1.20 (0.63–2.28) 0.572 1.19 (0.66–2.15) 0.570 Model 7 (multivariable + IL-6) 1.41 (0.88–2.27) 0.157 1.29 (0.70–2.40) 0.414 1.10 (0.63–1.91) 0.751 Model 8 (multivariable + IL-4) 1.35 (0.84–2.18) 0.218 1.20 (0.64–2.25) 0.568 1.02 (0.58–1.78) 0.944 Model 9 (multivariable + IL-10) 1.27 (0.77–2.05) 0.352 1.24 (0.56–2.14) 0.652 0.89 (0.51–1.58) 0.694 Model 10 (multivariable + Foxp3) 1.06 (0.61–1.85) 0.840 1.24 (0.66–2.30) 0.504 1.06 (0.61–1.85) 0.840 IL-4 † Model 1 (unadjusted) 1.90 (1.27–2.86) 0.002 2.26 (1.42–3.59) 0.001 1.19 (0.76–1.85) 0.456 ‡ Model2(ageandsex) 1.93 (1.28–2.92) 0.002 2.27 (1.41–3.63) 0.001 1.19 (0.76–1.86) 0.457 Model 3 (multivariable)§ 1.73 (1.09–2.75) 0.020 2.06 (1.15–3.68) 0.015 1.13 (0.68–1.87) 0.649 Model 4 (multivariable + IgE) 1.67 (1.05–2.66) 0.030 1.98 (1.10–3.58) 0.023 1.14 (0.68–1.90) 0.630 Model 5 (multivariable + CRP) 1.58 (1.00–2.52) 0.057 2.47 (1.22–4.98) 0.012 0.86 (0.50–1.50) 0.601 Model 6 (multivariable + tryptase) 1.53 (1.02–2.61) 0.041 2.01 (1.02–3.60) 0.019 1.18 (0.71–1.99) 0.522 Model 7 (multivariable + IL-6) 1.67 (1.04–2.66) 0.033 2.11 (1.17–3.80) 0.013 1.29 (0.76–2.21) 0.350 Model 8 (multivariable + TNF-𝛼) 1.68 (1.06–2.67) 0.026 2.03 (1.13–3.63) 0.017 1.12 (0.67–1.87) 0.663 Model 9 (multivariable + IL-10) 1.46 (0.90–2.36) 0.125 1.58 (0.84–2.97) 0.158 0.88 (0.52–1.51) 0.645 Model 10 (multivariable + Foxp3) 1.68 (1.05–2.69) 0.031 1.94 (1.10–3.53) 0.023 1.15 (0.69–1.92) 0.601 IL-10 † Model 1 (unadjusted) 1.99 (1.32–2.99) 0.001 5.76 (3.57–9.29) <0.001 2.90 (1.84–4.58) <0.001 ‡ Model2(ageandsex) 1.03 (1.00–1.06) 0.060 6.07 (3.72–9.91) <0.001 2.92 (1.84–4.62) <0.001 Model 3 (multivariable)§ 2.17 (1.37–3.46) 0.001 8.61 (4.51–16.41) <0.001 3.03 (1.79–5.12) <0.001 Model 4 (multivariable + IgE) 2.11 (1.32–3.37) 0.002 7.56 (3.92–14.57) <0.001 2.82 (1.65–4.81) <0.001 Model 5 (multivariable + CRP) 2.07 (1.29–3.31) 0.002 7.33 (3.27–13.68) <0.001 3.06 (1.73–5.39) <0.001 Model 6 (multivariable + tryptase) 2.05 (1.28–3.29) 0.003 8.70 (4.52–16.73) <0.001 3.21 (1.18–5.48) <0.001 Model 7 (multivariable + IL-6) 2.11 (1.32–3.38) 0.002 8.79 (4.12–16.04) <0.001 3.28 (1.92–5.60) <0.001 Model 8 (multivariable + TNF-𝛼) 2.10 (1.31–3.36) 0.002 8.66 (3.99–15.86) <0.001 3.07 (1.81–5.22) <0.001 Model 9 (multivariable + IL-4) 1.98 (1.23–3.20) 0.005 8.86 (4.11–15.06) <0.001 3.11 (1.81–5.33) <0.001 Model 10 (multivariable + Foxp3) 2.13 (1.33–3.41) 0.002 8.62 (4.50–16.49) <0.001 3.04 (1.79–5.13) <0.001 Foxp3+ † Model 1 (unadjusted) 1.35 (0.89–2.06) 0.162 1.10 (0.67–1.81) 0.702 0.82 (0.50–1.33) 0.415 ‡ Model2(ageandsex) 1.33 (0.87–2.03) 0.190 1.07 (0.65–1.77) 0.788 0.81 (0.50–1.33) 0.410 Model 3 (multivariable)§ 1.31 (0.81–2.17) 0.270 1.69 (0.91–3.12) 0.097 0.90 (0.52–1.57) 0.720 Model 4 (multivariable + IgE) 1.35 (0.84–2.17) 0.211 1.61 (0.85–3.04) 0.148 0.84 (0.48–1.48) 0.551 Model 5 (multivariable + CRP) 1.19 (0.73–1.95) 0.478 150 (0.65–3.24) 0.241 0.73 (0.40–1.31) 0.293 Model 6 (multivariable + tryptase) 1.17 (0.71–1.93) 0.544 1.61 (0.85–3.04) 0.144 1.01 (0.56–1.83) 0.977 Model 7 (multivariable + IL-6) 1.24 (0.76–2.02) 0.395 1.70 (0.92–3.17) 0.093 0.98 (0.56–1.73) 0.947 Model 8 (multivariable + TNF-𝛼) 1.26 (0.78–2.04) 0.349 1.66 (0.89–3.07) 0.111 0.90 (0.51–1.56) 0.701 Model 9 (multivariable + IL-4) 1.17 (0.72–1.92) 0.525 1.57 (0.84–2.93) 0.161 0.88 (0.50–1.54) 0.659 Model 10 (multivariable + IL-10) 1.13 (0.69–1.86) 0.624 1.68 (0.86–3.28) 0.132 0.88 (0.50–1.55) 0.651 ORs: odd ratios; T2DM: type 2 diabetes mellitus; TC: total cholesterol; TG: triglyceride; BMI: body mass index; WHR: waist hip ratio; SBP: systolic blood pressure; DBP: diastolic blood pressure; IgE: immunoglobulin E; hs-CRP: hypersensitivity C-reactive protein; IL-4: interleukin-4; IL-6: interleukin-6; IL-10: interleukin-10; Foxp3+: forkhead/winged helix transcription factor 3+; TNF-𝛼: tumor necrosis factor. Data are OR (95% CI) unless otherwise indicated. † Model 1 was not adjusted for any variable. ‡ Model 2 was adjusted for age and sex. § Model 3 was adjusted for age, sex, BMI, WHR, SBP, DBP, TC, TG, level of physical activity, dietary intake, alcohol intake, smoking status, presence or absence of family history of diabetes, hypertension, heart disease, stroke, and hypercholesterolemia. Journal of Diabetes Research 9 or allergies which are also very common among individuals, [8] Y. T. Lagerros, L. A. Mucci, R. Bellocco, O. Nyren,´ O. Balter,¨ and these unmeasured residual confounding factors might andK.A.Balter,¨ “Validity and reliability of self-reported total result in discrepancies. energy expenditure using a novel instrument,” European Journal In conclusion, circulating inflammatory mediators, hs- of Epidemiology,vol.21,no.3,pp.227–236,2006. CRP, IgE, IL-4, IL-10, and tryptase, were positively associated [9]N.Alqahtani,W.A.G.Khan,M.H.Alhumaidi,andY.A.A. withprediabetesorT2DM.Furtherlargeprospectivestudies R. Ahmed, “Use of glycated hemoglobin in the diagnosis of are warranted to assess a temporal relation between baseline diabetes mellitus and pre-diabetes and role of fasting plasma levels of inflammatory biomarkers and incidence of predia- glucose, oral glucose tolerance test,” International Journal of Preventive Medicine, vol. 4, no. 9, pp. 1025–1029, 2013. betesorT2DManditsassociatedchronicdiseases. [10]D.E.Olson,M.K.Rhee,K.Herrick,D.C.Ziemer,J.G. Twombly, and L. S. Phillips, “Screening for diabetes and pre- Competing Interests diabetes with proposed A1C-based diagnostic criteria,” Diabetes Care,vol.33,no.10,pp.2184–2189,2010. The authors declare that there is no conflict of interests [11] Y. Wu, R. Huxley, L. Li et al., “Prevalence, awareness, treatment, regarding the publication of this paper. The authors confirm and control of hypertension in China: data from the China that the mentioned received funding in the Acknowledg- National Nutrition and Health Survey 2002,” Circulation,vol. ments did not lead to any conflict of interests regarding the 118, no. 25, pp. 2679–2686, 2008. publication of this paper. [12] P.-H. Zhang, S.-F. Jiao, Y. Zhou et al., “Study on chronic disease related behavior and lifestyle in adults in Beijing, 2005,” Chinese Journal of Epidemiology,vol.28,no.12,pp.1162–1166,2007. Acknowledgments [13]P.Marques-Vidal,F.Bastardot,R.VonKanel¨ et al., “Association between circulating cytokine levels, diabetes and insulin resis- The authors thank all the volunteers who participated in this tance in a population-based sample (CoLaus study),” Clinical study and the clinic staff members from the Huzhou First Endocrinology,vol.78,no.2,pp.232–241,2013. Hospital, Zhebei Clinical Medicine Hospital, and Huanzhu [14] M. I. Schmidt, B. B. Duncan, A. R. Sharrett et al., “Markers Street Community Health Service Center for their assistance of inflammation and prediction of diabetes mellitus in adults in clinical data collections. This work was supported by (Atherosclerosis Risk in Communities study): a cohort study,” grants from the Huzhou Municipal Science and Technology The Lancet,vol.353,no.9165,pp.1649–1652,1999. Agency (2012GS02) (Zhen Wang), the Zhejiang Province [15] J. M. Guzman-Flores,´ J. F. Munoz-Valle,˜ J. Sanchez-Corona´ Natural science fund (LY13H070004) (Xu-Hui Shen), and et al., “Tumor necrosis factor-alpha gene promoter −308G/A the Zhejiang Province Science Technology Department Plan and −238G/A polymorphisms in Mexican patients with type 2 Projects (2015C33283) (Zhen Wang). diabetes mellitus,” Disease Markers,vol.30,no.1,pp.19–24,2011. [16] S. Tangvarasittichai, S. Pongthaisong, and O. Tangvarasittichai, “Tumor necrosis factor-Α, interleukin-6, C-reactive protein References levels and insulin resistance associated with type 2 diabetes in [1] A. Nadeem, S. Mumtaz, A. K. Naveed et al., “Gene-gene, gene- abdominal obesity women,” Indian Journal of Clinical Biochem- environment, gene-nutrient interactions and single nucleotide istry,vol.31,no.1,pp.68–74,2016. polymorphisms of inflammatory cytokines,” World Journal of [17] A. Al-Shukaili, S. AL-Ghafri, S. Al-Marhoobi, S. Al-Abri, Diabetes,vol.6,no.4,pp.642–647,2015. J. Al-Lawati, and M. Al-Maskari, “Analysis of inflammatory mediators in type 2 diabetes patients,” International Journal of [2] M. Banerjee and M. Saxena, “Genetic polymorphisms of Endocrinology, vol. 2013, Article ID 976810, 7 pages, 2013. cytokine genes in type 2 diabetes mellitus,” World Journal of Diabetes,vol.5,no.4,pp.493–504,2014. [18] G. Mavridis, E. Souliou, E. Diza et al., “Inflammatory cytokines in insulin-treated patients with type 2 diabetes,” Nutrition, [3]A.Brunetti,E.Chiefari,andD.Foti,“Recentadvancesinthe Metabolism and Cardiovascular Diseases,vol.18,no.7,pp.471– molecular genetics of type 2 diabetes mellitus,” World Journal of 476, 2008. Diabetes,vol.5,no.2,pp.128–140,2014. [19] C. Sabanayagam, A. Shankar, S. C. Lim, J. Lee, E. S. Tai, and [4] M. Saxena, N. Srivastava, and M. Banerjee, “Association of IL- T. Y. Wong, “Serum C-reactive protein level and prediabetes in 𝛼 6, TNF- and IL-10 gene polymorphisms with type 2 diabetes two Asian populations,” Diabetologia,vol.54,no.4,pp.767–775, mellitus,” Molecular Biology Reports,vol.40,no.11,pp.6271– 2011. 6279, 2013. [20] X. Ye, G. Zong, X. Liu et al., “Development of a new risk score [5]P.Arora,B.Garcia-Bailo,Z.Dastanietal.,“Geneticpolymor- for incident type 2 diabetes using updated diagnostic criteria in phisms of innate immunity-related inflammatory pathways and middle-aged and older Chinese,” PLoS ONE,vol.9,no.5,Article their association with factors related to type 2 diabetes,” BMC ID e97042, 2014. Medical Genetics, vol. 12, article 95, 2011. [21] I. Isordia-Salas, M. E. Galvan-Plata,A.Lea´ nos-Miranda˜ et al., [6]Z.Wang,H.Zhang,X.-H.Shenetal.,“Immunoglobulineand “Proinflammatory and prothrombotic state in subjects with mast cell proteases are potential risk factors of human pre- different glucose tolerance status before cardiovascular disease,” diabetes and diabetes mellitus,” PLoS ONE,vol.6,no.12,Article Journal of Diabetes Research,vol.2014,ArticleID631902,9 ID e28962, 2011. pages, 2014. [7] Z. Wang, H. Zhang, X.-H. Shen et al., “Immunoglobulin E and [22] M. Hulsmeyer,¨ C. Scheufler, and M. K. Dreyer, “Structure of mast cell proteases are potential risk factors of impaired fasting interleukin 4 mutant E9A suggests polar steering in receptor- glucose and impaired glucose tolerance in humans,” Annals of complex formation,” Acta Crystallographica Section D Biological Medicine, vol. 45, no. 3, pp. 220–229, 2013. Crystallography,vol.57,no.9,pp.1334–1336,2001. 10 Journal of Diabetes Research

[23] A. R. Aroor, S. McKarns, V. G. Demarco, G. Jia, and J. R. Sowers, “Maladaptive immune and inflammatory pathways lead to cardiovascular insulin resistance,” Metabolism: Clinical and Experimental, vol. 62, no. 11, pp. 1543–1552, 2013. [24] M.Kassan,M.Galan,M.Partyka,M.Trebak,andK.Matrougui, + + “Interleukin-10 released by CD4 CD25 natural regulatory T cells improves microvascular endothelial function through inhibition of NADPH oxidase activity in hypertensive mice,” Arteriosclerosis, Thrombosis, and Vascular Biology, vol. 31, no. 11, pp. 2534–2542, 2011. [25] K. Ohshima, M. Mogi, F. Jing et al., “Roles of interleukin 17 in angiotensin II type 1 receptor-mediated insulin resistance,” Hypertension,vol.59,no.2,pp.493–499,2012. [26]M.Saxena,C.C.Agrawal,H.K.Bid,andM.Banerjee,“An interleukin-10 gene promoter polymorphism (−592A/C) asso- ciated with type 2 diabetes: a North Indian study,” Biochemical Genetics,vol.50,pp.549–559,2012. [27] A. B. Acharya, S. Thakur, and M. V. Muddapur, “Evaluation of serum interleukin-10 levels as a predictor of glycemic alteration in chronic periodontitis and type 2 diabetes mellitus,” Journal of Indian Society of Periodontology,vol.19,no.4,pp.388–392, 2015. [28] A. B. Acharya, S. Thakur, and M. V. Muddapur, “J Indian effect of scaling and root planing on serum interleukin-10 levels and glycemic control in chronic periodontitis and type 2 diabetes mellitus,” Journal of Indian Society of Periodontology,vol.19,no. 2, pp. 188–193, 2015. [29] G. Liu, H. Ma, L. Qiu et al., “Phenotypic and functional switch of macrophages induced by regulatory CD4 CD25 T cells in mice,” Immunology and Cell Biology,vol.89,no.1,pp.130–142,2011. [30] S.Sun,Y.Ji,S.Kersten,andL.Qi,“Mechanismsofinflammatory responses in obese adipose tissue,” Annual Review of Nutrition, vol. 32, pp. 261–286, 2012. [31] M. Feuerer, L. Herrero, D. Cipolletta et al., “Lean, but not obese, fat is enriched for a unique population of regulatory T cells that affect metabolic parameters,” Nature Medicine,vol.15,no.8,pp. 930–939, 2009. [32] J. Sun, J. Zhang, J. S. Lindholt et al., “Critical role of mast cell chymase in mouse abdominal aortic aneurysm formation,” Circulation,vol.120,no.11,pp.973–982,2009. [33] J. Zhang, J. Sun, J. S. Lindholt et al., “Mast cell tryptase deficiency attenuates mouse abdominal aortic aneurysm formation,” Cir- culation Research, vol. 108, no. 11, pp. 1316–1327, 2011. [34] M. Xiang, J. Sun, Y.Lin et al., “Usefulness of serum tryptase level as an independent biomarker for coronary plaque instability in aChinesepopulation,”Atherosclerosis,vol.215,no.2,pp.494– 499, 2011.