Journal of Autoimmunity xxx (2014) 1e9

Contents lists available at ScienceDirect

Journal of Autoimmunity journal homepage: www.elsevier.com/locate/jautimm

Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes Tom L. Van Belle a,1, 2, Philippe P. Pagni a,1, Jeanette Liao a, Sowbarnika Sachithanantham a, Amy Dave a, Amira Bel Hani a, Yulia Manenkova a, Natalie Amirian a, Cheng Yang a, Bret Morin a, Haiqing Zhang a, 3, Iain L. Campbell b, Matthias G. von Herrath a, * a b

Type 1 Diabetes Center, La Jolla Institute for Allergy and Immunology, 9420 Athena Circle, La Jolla, CA 92037, USA The School of Molecular Bioscience, University of Sydney, NSW 2006, Australia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 28 June 2013 Received in revised form 23 January 2014 Accepted 3 February 2014

Inflammatory mechanisms play a key role in the pathogenesis of type 1 and type 2 diabetes. IL6, a pleiotropic cytokine with impact on immune and non-immune cell types, has been proposed to be involved in the events causing both forms of diabetes and to play a key role in experimental insulindependent diabetes development. The aim of this study was to investigate how beta-cell specific overexpression of IL-6 influences diabetes development. We developed two lines of rat insulin promoter (RIP)-lymphocytic choriomeningitis virus (LCMV) mice that also co-express IL6 in their beta-cells. Expression of the viral nucleoprotein (NP), which has a predominantly intracellular localization, together with IL6 led to hyperglycemia, which was associated with a loss of GLUT-2 expression in the pancreatic beta-cells and infiltration of CD11bþ cells, but not T cells, in the pancreas. In contrast, overexpression of the LCMV glycoprotein (GP), which can localize to the surface, with IL-6 did not lead to spontaneous diabetes, but accelerated virus-induced diabetes by increasing autoantigen-specific CD8þ T cell responses and reducing the regulatory T cell fraction, leading to increased pancreatic infiltration by CD4þ and CD8þ T cells as well as CD11bþ and CD11cþ cells. The production of IL-6 in beta-cells acts prodiabetic, underscoring the potential benefit of targeting IL6 in diabetes. Ó 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Keywords: Type 1 diabetes Cytokine GLUT-2 Inflammation T cell

1. Introduction Diabetes mellitus is a chronic metabolic disorder that is caused by a failure of the body to produce insulin (type 1 diabetes, T1D) and/or the inability of the body to respond adequately to circulating insulin (type 2 diabetes, T2D). While type 1 and type 2 diabetes

Abbreviations: T1D, type 1 diabetes; IDDM, insulin-dependent diabetes mellitus; NOD, non-obese diabetic; LCMV, lymphocytic choriomeningitis virus; RIP, rat insulin promoter; GP, glycoprotein; NP, nucleoprotein; Tg, transgenic; IL, interleukin; TNF, tumor necrosis factor; IFN-g, interferon-g; Th, T-helper; PFU, plaqueforming units; GLUT-2, glucose transporter 2; XBP-1, X-box binding protein-1; CHOP, C/EBP homologous protein; ER, endoplasmic reticulum. * Corresponding author. Tel.: þ1 858 205 0646; fax: þ1 858 752 6993. E-mail addresses: [email protected] (T.L. Van Belle), matthias@liai. org (M.G. von Herrath). 1 These authors contributed equally to this work. 2 Present address: Clinical and Experimental Endocrinology, KU Leuven, Campus Gasthuisberg, O&N I Herestraat 49 e Box 902, 3000 Leuven, Belgium. 3 Present address: Department of Endocrinology and Metabolism, Shandong Provincial Hospital affiliated to Shandong University, Jinan 250021, China.

have different etiology and pathogenesis, both types of diabetes share inflammation as a common trait. About 5e10% of diabetic patients suffer from T1D, the most common metabolic disease in childhood and adolescence affecting 112,000 children younger than 16 in Europe [1]. In recent years, prevalence has not only risen, but T1D became increasingly diagnosed at younger ages [2]. In type 1 or insulin-dependent diabetes mellitus (IDDM), the pancreatic islets and, specifically, the islet beta-cells become the target of an autoimmune response which destroys their insulin-producing capacity [3]. Human enterovirus (HEV) infections have long been suspected as environmental triggers of human T1D [4]. Cytokines have been implicated in the development of the characteristic islet mononuclear cell infiltrate (termed insulitis) and may also mediate cell toxicity in IDDM [5]. Interleukin (IL)6 is a pleiotropic cytokine with a key impact on both immunoregulation and non-immune events in most cell types and tissues outside the immune system [6,7], including but not limited to fibroblasts, endothelial cells, keratinocytes, monocytes/ macrophages, T cell lines, mast cells, and a variety of tumor cell lines [8]. A vast number of epidemiological, genetic, rodent, and

http://dx.doi.org/10.1016/j.jaut.2014.02.002 0896-8411/Ó 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

2

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

human in vivo and in vitro studies suggest both protective and pathogenic actions of IL6 in diabetes, both type 1 and type 2 [6]. In human T2D, concentrations of IL6 in the serum are not only predictive of the disease development but are also increased in T2D patients [9,10]. T1D patients also have higher concentrations of IL6 in the serum as well as in the artery wall, compared with nondiabetic subjects [11,12]. Studies in experimental models further show that IL6 is produced by isolated mouse islets after exposure to interferon-g (IFN-g) and tumor necrosis factor-a (TNF-a) [13]. IL6 is also present in infiltrated islets before and during immune infiltration in non-obese diabetes (NOD) mice, a model for type 1 diabetes. The observation of higher IL6 expression in the islets of NOD females than in those of males supports a deleterious effect of IL6 [14,15]. Indeed, beta-cell-specific overexpression of IL6 promotes islet inflammation in NOD mice [16] and in a non-diabetes-prone mouse strain [17] and IL6 is essential for diabetes development in the NOD [18]. Adoptively transferred self-reactive, activated CTL home to the islets and surround the islets of RIP-LCMV transgenic (Tg) mice, but require additional factors, such as upregulation of MHC class II molecules and activation of antigen-presenting cells, to enter the islets (insulitis) and cause hyperglycemia [19]. We tested here the impact of local production of IL6 as additional factor on diabetes development in virus-induced diabetes. Conversely, the observation by Campbell et al. that IL6 expression causes insulitis, but not diabetes, in IL6 Tg mice [17] led us to test whether additional factors such as viral infection and/or co-expression of neo-autoantigens are required to trigger diabetes in IL6 Tg mice. 2. Research design and methods 2.1. Mice, infections and diabetes incidence Mice expressing murine IL6 under the rat insulin promoter (RIP), RIP-IL6, were generated before [17]. RIP-IL6 mice were crossed to RIP-LCMV-nucleoprotein (NP) and RIP-LCMVglycoprotein (GP) (C57Bl/6 background) to generate RIP-NP/IL6 and RIP-GP/IL6, respectively. Where indicated, mice were inoculated with LCMV Armstrong 53b (104 pfu i.p. for diabetes induction in RIP-LCMV, or 106 pfu i.p. for RIP-IL6 in Fig. 6), with Coxsackie Virus B3/GA (CVB3/GA, slower replicator, 5  106 TCID50 i.p.), or with Coxsackie Virus B3/28 (CVB3/28, fast replicator, 5  105 TCID50 i.p.). For spontaneous diabetes, mice were monitored weekly and mice were considered diabetes on the first of 2 consecutive blood glucose readings above 250 mg/dl. Of note, the wild-type littermates of these transgenic mouse strains had relatively high baseline blood glucose concentrations: the ‘NP/IL6’ wildtypes had on average 190 mg/dl (males) and 160 mg/dl (females), and the ‘GP/ IL6’ wildtypes had on average 175 mg/dl (males) and 160 mg/dl (females). For LCMV-induced diabetes, mice were monitored daily and were considered diabetes on the first of 2 consecutive blood glucose readings above 250 mg/dl. 2.2. Immunohistochemistry Pancreata were harvested and snap-frozen in OCT Compound (Tissue-Tek, Sakura Finetek USA, Torrance, CA). Six micron tissue sections were fixed in acetone (10 min RT) before rehydration and biotin/avidin blocking, then incubated with biotinylated anti-mouse CD8, CD4, CD45R/B220, or CD11b in TBS/2%FCS/2%GS (all 1:50, BD Biosciences, San Jose, CA), and guinea pig antiswine insulin (1:500, Dako)(1 h RT). Next, sections were washed before secondary incubation with goat anti-guinea pig AP (1:50, Sigma) and avidin-HRP (1:2000, Vector Labs, Burlingame, CA) in TBS/2%FBS/2%GS. AP or HRP activity was visualized

using Vector Blue AP III (Vector Labs, blue signal) or AEC substrate (red precipitates), respectively. Slides were mounted (Dako Faramount Aqueous Mounting Medium, Dako, Carpinteria, CA) without hematoxylin counterstain. Original JPEG files were adjusted in Photoshop: Image > Auto Tone, then contrast set to 50 and saved as TIFF. 2.3. Immunofluorescence Sections were prepared as for immunohistochemistry. For murine IL6 detection, sections were first incubated in goat anti-mouse IL6 (R&D Systems, Minneapolis, MN; 15 mg/ml in TBS/2%Rabbit Serum (RS); 1 h at RT), washed and incubated with biotinylated rabbit anti-goat (Vector Labs, 1:1000 in TBS/2%RS), washed and incubated in streptavidin-AF488 (Invitrogen, Carlsbad, CA; 1:1000 in TBS/2%RS) (both 45 min at RT). Insulin or GLUT-2 was detected on consecutive sections to circumvent crossreactivity. For insulin, sections were incubated (all 1 h RT) in guinea pig anti-swine insulin antibody (1:140 in TBS/2%GS), washed and incubated with goat polyclonal anti-guinea pig AF488 (Invitrogen; 1:1000 in TBS/2%GS). For GLUT-2 staining, sections were incubated in polyclonal rabbit anti-GLUT-2 (Millipore, Temecula, CA; 1:400 in TBS/2%GS), washed and incubated with goat polyclonal anti-rabbit AF594 antibody (Invitrogen; 1:1000 in TBS/2%GS). Slides were mounted in Prolong Gold anti-fade reagent. Pictures were RGB split to obtain the relevant channel and background subtracted using ImageJ. For full pancreas overview, pictures at 10 magnification were merged using Photomerge in Adobe Photoshop. 2.4. Glucose tolerance tests (GTT) Mice were fasted for 16 h and injected intraperitoneally with glucose (2 g/kg). Blood glucose concentration was monitored as indicated using OneTouch Ultra glucometer via tail vein blood sampling (approximately 5 ml). 2.5. Insulin sensitivity test (IST) Mice were fasted for 4 h and injected intraperitoneally with insulin (Humalog, Eli Lilly, Indianapolis, IN; diluted in 0.9% saline; 0.75 units/kg). Blood glucose concentration was measured at 15, 30, 45 and 60 min after injection. 2.6. Flow cytometry Fluorochrome-labeled mAbs were from BD Biosciences, eBioscience (San Diego, CA), or BioLegend (San Diego, CA). After Fc block (1 mg/ml; 10 min RT), cells were incubated with antibodies to CD3, CD4, CD8a, CD19, CD25 and Foxp3 (20 min 6e8  C). Foxp3 staining was performed using eBioscience reagents. LCMV/GP33-41/H-2Db pentamer staining was done for 30 min at RT (10 ml/sample, ProImmune). LCMV-GP-specific CD8 responses were measured after 5 h ex vivo restimulation with 1 mg/ml gp33-41 peptide and T cell-depleted splenocytes (TdS, only for pLN) in the presence of 10 mg/ml Brefeldin A. TdS were prepared using pan T cell beads (CD90.2, Invitrogen) and irradiated (3000 rad equivalent) using an RS2000 X-ray irradiator (Rad Source, Suwanee, GA). After antigenspecific stimulation, intracellular cytokines IFN-g and TNF-a were detected after fixation and permeabilization using BD Cytofix/ CytoPerm reagents. Exclusion of dead cells was done using the Live Dead dyes (Aqua or Violet, Invitrogen) for 20 min at RT, prior to fixation for ICCS or Foxp3 staining. Stained samples were acquired on LSRII (BD Biosciences). Data were analyzed using FlowJo (Treestar, Ashland, OR).

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

2.7. Q-PCR analysis Pancreas sections from mice were cut and RNA was extracted by placing samples in the RLT buffer and following instructions from Qiagen’s RNEasy extraction kit. Two micrograms of RNA were reverse-transcribed into complementary DNA (cDNA) using the RNA-to-cDNA kit (Applied Biosystems) according to the manufacturer’s instructions. Five microliters of cDNA diluted 1/20 in nuclease-free water was used to run quantitative PCRs in duplicates (total volume of 25 ml). 2.8. Statistical analysis Graphs and statistical analysis were done using Prism 5.0 (GraphPad, La Jolla, CA). For diabetes incidence, significance was calculated by log-rank test (ManteleCox). For cell subset analysis, non-parametric t-test ManneWhitney (two-tailed) was used. Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, ***P < 0.001. 2.9. Ethics All the studies were performed in the La Jolla Institute for Allergy and Immunology upon approval of LIAI’s Animal Care and Use Committee. 3. Results 3.1. Co-expression of viral protein and IL6 can cause spontaneous diabetes development The presence of high numbers of activated autoreactive T cells that recognize an islet self-antigen (e.g. a viral transgene) is not always sufficient for beta-cell destruction [19]. Induction of diabetes in RIP-LCMV mice by inoculation with LCMV causes a rapid upregulation of MHCII and activation of macrophages in the islets [19]. Because macrophages can produce IL6 [20] and IL6 is essential for diabetes onset [18], we examined how beta-cell specific IL6 production impacts immune cell recruitment to the pancreas and diabetes development. We crossed RIP-IL6 Tg mice [17], expressing IL6 specifically in the pancreatic beta-cells (Fig. 1A), to RIP-LCMV-GP or RIP-LCMV-NP C57Bl/6, expressing glycoprotein (GP) or nucleoprotein (NP) of LCMV as a self-antigen in their pancreatic beta-cells and in some lines, in the thymus [21]. Without LCMV inoculation, F1 offspring

3

expressing either a neo-autoantigen or IL6 alone did not develop diabetes (Fig. 1B and C) [17,21], indicating the produced levels of IL6 are not toxic to the beta-cells. Remarkably, simultaneous expression of both IL6 and the neo-autoantigen NP caused hyperglycemia in all mice, even without LCMV inoculation (Fig. 1B). In male RIPNPþ/IL6þ double Tg mice, incidence of hyperglycemia started by week 5 of age and afflicted all mice by week 9. In female RIP-NPþ/ IL6þ double Tg mice, hyperglycemia was observed starting week 6 and reaching 100% penetration by week 16. In contrast, simultaneous expression of IL6 and the neo-autoantigen GP did not raise blood glycemia in female mice and only in 25% of the male RIP-GPþ/ IL6þ double Tg mice (Fig. 1C). We next tested whether the differences in the magnitude of expression of IL6 explained the very different pattern of diabetes development displayed by RIP-GP and RIP-NP mice when crossed with RIP-IL6 mice. This revealed that the amounts of IL6 transcript in the pancreas did not differ between the RIP-NP/IL6 and the RIP-GP/IL6 line (Fig. 1D), even though IL-6 Tg mice contained clearly higher amounts of IL-6 transcripts in the pancreas than IL-6 non-Tg mice. 3.2. CD11bþ cells, but not T cells, infiltrate the pancreas in RIP-NPþ/ IL6þ double Tg mice Previous histological examination revealed that islets often displayed irregular boundaries, numerous ducts surrounding the islets and islet inflammation [17]. We confirm this in RIP-NP/IL6þ and RIP-NPþ/IL6þ mice (data not shown) and further determined which immune cell types infiltrated the pancreas of hyperglycemic RIP-NPþ/IL6þ mice. As expected, no CD11bþ, CD4þ or CD8þ T cells were detectable in non-Tg littermates. In line with our previous data on mice at older age [17], CD11bþ cells infiltrated the pancreas in RIP-NP/IL6þ mice already at the age of 5 weeks (Fig. 2A). However, CD4þ and CD8þ T cells were absent from immune infiltrates in islets of RIP-NP/IL6þ mice. We then examined the immune infiltrate in pancreata of RIP-NPþ/IL6þ double Tg mice at 5 weeks of age, i.e. prior to the development of hyperglycemia, and found that co-expression of NP and IL6 increased the amount of CD11bþ cells in the immune infiltrate, but did not cause infiltration by T cells at this age. To determine whether the increased influx of CD11b also reflects a qualitative change in the nature of the CD11bþ cells, we assessed if IL6 expression in the pancreas shifts the balance of macrophages from an M2-phenotype to a more proinflammatory M1-phenotype. We measured (in the pancreas) the expression inducible nitric oxide synthase (Inos, pro-M1), arginase-1 (Arg1, pro-M2) as well as

Fig. 1. Simultaneous production of autoantigen and IL6 in the beta-cells causes diabetes. A: Immunofluorescence staining of IL6 in the pancreas of RIP-IL6 Tg mice. Green: IL6; Blue: DAPI; dotted lines delineate pancreatic islets. Pancreatic islets shown are representative for at least 200 islets from 4 mice. B, C: Percentage of hyperglycemic mice in male (left) and female (right) RIP-NP/IL6 double Tg mice (B) and RIP-GP/IL6 double Tg mice (C), as indicated. Statistical significance was determined by log-rank (ManteleCox) test. ***P < 0.001. D: IL6 gene expression in pancreas sections of RIP-NP/IL6 (top) and RIP-GP/IL6 (bottom) transgenic mice. Data are cumulated from two independent experiments are shown (n ¼ 5e7). **P < 0.01. Lines represent means.

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

4

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

Fig. 2. CD11bþ cells, but not T cells, infiltrate the pancreas in RIP-NPþ/IL6þ double Tg mice. Representative cryosections of pancreata from 5-week old mice, stained for CD11b, CD4, or CD8 (red-brown) and insulin (blue). Islets shown are representative for the predominant infiltration type (40e130 islets per group, 3e6 mice per group). B, C: mRNA expression of inducible nitric oxide synthase (Inos; left), arginase-1 (Arg1, middle) and the ratio of Inos/Arg1 mRNA expression (right) in pancreases of RIP-NP/IL6 (B; n ¼ 5e6) and RIP-GP/IL6 (C; n ¼ 5e7) transgenic mice, analyzed by RT-qPCR and normalized to GAPDH using the DDCt method. Data cumulated from two independent experiments are shown. Lines represent means.

Inos/Arg1 expression ratios [22] and found no significant change in the expression of these genes, arguing against a shift toward proinflammatory M1 macrophages in the pancreas of RIP-NP/IL6 mice (Fig. 2B). We obtained similar results in the RIP-GP/IL6 strain (Fig. 2C). 3.3. Deficient GLUT-2 expression in RIP-NPþ/IL6þ and RIP-GPþ/IL6þ double Tg mice The absence of immune infiltrate suggests that the hyperglycemia in RIP-NPþ/IL6þ mice is not driven by autoimmunity, but perhaps by perturbed insulin production and/or response. Immunohistochemistry (data not shown) and immunofluorescence microscopy revealed that beta-cells from RIP-NPþ/IL6þ mice still contain insulin, but some islets showed an irregular insulin staining pattern (Fig. 3A). This is apparent in all islets of these mice (Fig. 3A, left column showing whole pancreas section overview). Gene expression analysis nevertheless confirms that Ins1 and Ins2 transcripts are not present at lower amounts in the islets of RIP-NP/IL6 transgenic mice (data not shown). Restoring euglycemia using an insulin pellet normalized this pattern, indicating that these mice

can produce normal amounts of insulin and suggesting that the irregular pattern of the insulin staining reflects the partial release of insulin-containing secretory granules in hyperglycemic RIP-NPþ/ IL6þ mice. Besides insulin, also GLUT-2 (solute carrier family 2 member 2, SLC2A2) is required for glucose homeostasis and normal function and development of the endocrine pancreas. Absence of GLUT-2 prevents glucose-stimulated insulin secretion by beta-cells. It is therefore interesting that we observed deficient GLUT-2 expression in the pancreas of RIP-NPþ/IL6þ mice, regardless of whether they were hyperglycemic or normoglycemic by insulin pellet (Fig. 3A). Time-kinetic analysis showed that the disappearance of GLUT-2 expression preceded the hyperglycemia and that insulin expression in the islets remained intact before, during and after hyperglycemia (Fig. 3B). To test the functional complications of these observations, we performed an insulin tolerance test showing a comparable outcome in RIP-NPþ/IL6þ double Tg mice and single Tg or non-Tg littermates, indicating the response to insulin was not impaired (Fig. 3C). Of note, the response to insulin was not affected by the starting glycemia, as blood glycemia declined proportionally equal in both normoglycemic and hyperglycemic double Tg mice.

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

5

Fig. 3. Normal insulin sensitivity and impaired GLUT-2 production in RIP-NPþ/IL6þ double Tg mice. A: Expression of insulin (green) and GLUT-2 (red) revealed by immunofluorescence. Shown are whole pancreas cryosections from a representative mouse (1st and 3rd column) and pictures of individual islets (2nd and 4th column, representative for approximately 120e180 islets from 2 to 3 mice per group). Data were confirmed by IHC (data not shown). B: Time-kinetic analysis of insulin (green) and GLUT-2 (red) expression by immunofluorescence of pancreatic islets in female RIP-NPþ/IL6þ double Tg mice before, at or after diabetes onset. Dotted line delineates boundary of pancreatic islet. C: Insulin tolerance test. Displayed is the blood glycemia normalized to the fasting glycemia in RIP-LCMV-NP/IL6 single and double Tg mice, as indicated.

We reasoned that local expression IL-6 might have resulted in the production of inflammatory factors (e.g., prostaglandins) that in turn influence the expression of insulin or Glut2 in RIP-IL6 Tg mice. Because cyclooxygenase-2 (COX-2), an enzyme encoded by the Prostaglandin-endoperoxide synthase 2 (Ptgs2) gene and is responsible for the synthesis of prostaglandin H2 (PGH2) [23], is a very important source of prostaglandins, we measured the mRNA expression levels of COX-2 in the pancreas of RIP-NP/IL6 and RIPGP/IL6 mice. We could not find statistically significant differences

in RIP/IL6þ Tg mice compared to RIP/IL6 counterparts (data not shown). If anything, we observed a non-significant trend for lower COX-2 expression in RIP-NP þ IL6þ double Tg mice, ruling out an important role played by this particular mediators in the phenotypes we had observed. The RIP-GP/IL6 line showed a phenotypic pattern that was largely similar to the RIP-NP/IL6 line, namely pancreatic beta-cells from RIP-GPþ/IL6þ double Tg mice also expressed insulin (Fig. 4A) and gene expression analysis confirms that Ins1 and Ins2 transcripts

Fig. 4. Normal glucose sensitivity, but impaired GLUT-2 expression in RIP-GPþ/IL6þ mice. A: Expression of insulin (columns 1 and 2, green) and GLUT-2 (columns 3 and 4, red) revealed by immunofluorescence. Shown are pictures of individual islets of male or female mice as indicated (representative for approximately 120e180 islets from 2 to 3 mice per group). Dotted line delineates boundary of pancreatic islet. N/A: Not available because RIP-GPþ/IL6þ double Tg female mice do not become diabetic spontaneously. B: Glucose tolerance test in RIP-GP/IL6 Tg mice. Shown are the mean  SEM of blood glycemia normalized to the fasting glycemia, per group as indicated.

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

6

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

are not present at lower amounts in the islets of RIP-GP/IL6 transgenic mice (data not shown). However, similar to our observations in the RIP-NPþ/IL6þ mice, GLUT-2 expression was not detectable (Fig. 4A). To test the functional complications of these observations, we performed a glucose tolerance test. The blood glucose concentration after the glucose injection followed a comparable curve in RIP-GPþ/IL6þ double Tg mice and single Tg or non-Tg littermates, indicating the RIP-GPþ/IL6þ double Tg mice were capable of controlling a glucose bolus (Fig. 4B). 3.4. Beta-cell specific production of IL6 accelerates virus-induced diabetes Beta-cell specific production of IL6 alone does not induce diabetes [17], nor precipitates diabetes by LCMV virus on a diabetes resistant background (Fig. 1). We examined whether IL6 increased diabetogenic responses in the virus-induced model of autoimmune diabetes [21,24,25]. Inoculation of RIP-LCMV-GPþ/IL6 mice induced diabetes starting at day 10, and reaching 100% disease penetrance by day 12 (Fig. 5A). Strikingly, beta-cell specific coexpression of IL6 and GP significantly accelerated diabetes onset in the RIP-GPþ/IL6þ mice, starting as early as day 6 and reaching 100% disease penetrance by day 7 (Fig. 5A). Immunohistochemistry analysis revealed more CD4þ and CD8þ cells around and/or in the pancreatic islets and a strong influx of CD11bþ and CD11cþ cells in LCMV-inoculated GPþ/IL6þ versus GPþ/ IL6 mice (Fig. 5B). This suggests that IL6 creates a proinflammatory milieu supporting migration of immune cells and diabetogenicity. Of note, without LCMV inoculation, GPþ/IL6þ or GPþ/IL6 did not show infiltration of T cells (data not shown), similar to NPþ/IL6þ or NPþ/IL6 without LCMV inoculation (Fig. 2). IL6 did not increase autoantigen-specific CD8þ T cells in spleen or pancreatic lymph nodes, as determined by pentamer staining

(Fig. 5C), but a higher fraction of CD8þ T cells produced IFN-g (Fig. 5D) or both TNF-a and IFN-g (data not shown) upon ex vivo restimulation with GP33-41 peptide. This was because more of the GP33-41/H-2Db-pentamerþ CD8þ T cells actually produced IFN-g when restimulated ex vivo (Fig. 5E). We could not detect IL-17 production by T cells from RIP-LCMV-GPþ/IL6þ mice (data not shown) [26], despite IL6 being an important driver for IL-17 production [27]. On the other hand, we detected less CD4þ T cells with a CD4þCD25þFoxp3þ Treg phenotype in the spleen and pancreatic lymph node (Fig. 5F). Thus, beta-cell specific IL6 expression increased autoantigenspecific CD8þ T cell responses, while reducing the regulatory T cell fraction. 3.5. Inoculation with Coxsackie virus B (CVB) or LCMV does not cause diabetes RIP-IL6 mice Viral infections, especially with human enterovirus (HEV), have received much attention as potential environmental trigger of human T1D [4]. CVB has been shown to bear both enhancing and protective capacity in the NOD mouse model of T1D, depending on the context. CVB4 promotes diabetes in NOD mice [28], though possibly depending on the time point at which infection occurs [29]. CVB3 effectively suppresses diabetes incidence when inoculated into healthy young NOD mice [30,31], but can rapidly trigger diabetes in older, prediabetic NOD mice [32]. LCMV strain Armstrong is used as a means to induce diabetes in mice that transgenically express LCMV antigens in their beta-cells [24,25,33], but otherwise LCMV infection efficiently prevents diabetes in NOD mice [34,35]. We hypothesized that the local production of IL6 creates a proinflammatory milieu in the pancreas that suffices to confer diabetogenicity on infection with a pancreatropic virus. We inoculated RIP-IL6 Tg mice with CVB3, a pancreatropic enterovirus

Fig. 5. Beta-cell specific production of IL6 accelerates virus-induced diabetes. RIP-LCMV-GPþ/IL6þ mice or RIP-LCMV-GPþ/IL6 littermates were inoculated with LCMV to induce diabetes. A: Blood glycemia was monitored and percentage of hyperglycemic mice was plotted. n ¼ 4e6 per group. **P ¼ 0.0013 by Log-rank (ManteleCox) test. B: Pancreatic infiltration of CD4þ, CD8þ, CD11bþ and CD11cþ cells on day 7, revealed by immunohistochemistry (blue: insulin, red: cell subset marker). Shown are pictures of representative islets. CeF: On day 7, spleen (top) and pancreatic lymph nodes (bottom) were harvested to measure autoantigen-specific CD8þ T cells by pentamer staining (C), by ICCS upon stimulation with GP33-41 peptide in the presence of Brefeldin A (D), or via IFN-g production by GP33-41/Db pentamer positive CD8 T cells (E), as well as regulatory T cells (F; shown as % Foxp3þCD25þ in the live CD4 gate). Each dot represents one mouse. *P < 0.05 by two-tailed ManneWhitney test.

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

that can infect beta-cells [36] and that causes a systemic, acute infection that is nonpersisting but leads to immune-mediated myocarditis, or with LCMV Armstrong, an arenavirus that causes a systemic, acute infection that is asymptomatic and also nonpersisting. For CVB3 infection we used either the wellcharacterized, poorly pathogenic CVB3 strain CVB3/GA as a slowreplicator versus the fast-replicating CVB3/28 [30]. Infection with any of these three viruses did not increase the blood glycemia (Fig. 6A) and did not cause insulitis in the RIP-IL6 Tg mice (Fig. 6B). This suggests that IL6-induced local inflammation is insufficient to allow viral infections to breach tolerogenic mechanisms. 4. Discussion Studies looking into the role of IL6 in the pathogeneses underlying obesity, insulin resistance, beta-cell destruction, type 1 diabetes, and type 2 diabetes suggest both protective and pathogenic actions of IL6 in diabetes [6]. We have shown here that production of IL6 and antigen by b-cells can cause spontaneous hyperglycemia and interferes with GLUT-2 expression. In the virus-induced mouse model of autoimmune diabetes, IL6 expression accelerates diabetes that is associated with increasing autoantigen-specific CD8 T cell responses. IL-6 is secreted by antigen-presenting cells in response to external stimuli such as TNF-a, IL-1b but also bacterial and fungal components. IL-6 is antagonistic for regulatory T cells and

Fig. 6. Infection with CVB3 or LCMV does not cause diabetes in Tg mice IL6 production in the beta-cells. RIP-IL6 Tg mice were infected with Coxsackie Virus B3/28 (fast replicator, 5  105 TCID50 i.p.), Coxsackie Virus B3/GA (slower replicator, 5  106 TCID50 i.p.) or LCMV Armstrong 53b (106 pfu i.p.). A: Average of blood glucose concentration, and B: H&E staining of cryosections of pancreas, harvested at week 6 after inoculation (3 mice per group).

7

important for Th17 differentiation [37]. IL-6 is also sufficient and necessary for CD4þ T cells to induce production of IL-21 [38], a cytokine critically required for autoimmune diabetes in NOD mice [39,40]. Given such immunoregulatory actions of IL6, we wanted to study the influence of local production of IL6 during virus-induced diabetes using the RIPeLCMV system, a mouse model in which diabetes is initiated by viral infection [21,24,25]. In LCMVinoculated RIP-LCMV-GP/IL6 line, we found that IL6 increased the pancreatic infiltration by several immune cells, including T cells, macrophages, dendritic cells and B cells (data not shown), thus causing accelerated diabetes onset. This was associated with increased LCMV-GP-specific effector CD8þ T cell responses and reduced regulatory T cell fractions in the periphery, in line with a proinflammatory role for IL6. On the other hand however, local expression of IL6 is insufficient to confer diabetogenicity on viruses that can infect pancreatic betacells. We tested LCMV as well as a slow and a fast-replicating CVB3. Our data suggest that the viruses that could be involved in type 1 diabetes etiology still need a diabetes-prone genetic background and that local IL6 expression is insufficient to overcome these diabetes-resistance mechanisms. Conversely, a number of viruses that paradoxically induce strong inflammation and immune activity locally in the pancreas (while sparing b cells) during the prediabetic phase are known to prevent T1D in NOD mice [41]. Indeed, infection of prediabetic NOD mice with Coxsackie virus B3 or LCMV delayed diabetes onset and reduced disease incidence [31]. We do not yet know whether beta-cell specific IL6 expression can negate such protection by viruses. Surprisingly, the RIP-NPþ/IL6þ double Tg mice became hyperglycemic spontaneously, but the underlying mechanisms remain elusive. The absence of pancreatic infiltration by T cells suggested that the hyperglycemia is not associated with T cell autoimmunity. On the other hand, our data tend to indicate that the significant increase in CD11b-positive cells in the pancreas might be enough to cause impaired expression of Glut2, but that this goes without an obvious shift to the proinflammatory M1-phenotype. Further support that the reduction in GLUT-2 is not the result of proinflammatory factors comes from the lack of increased expression in IL-6 Tg mice of COX-2 (data not shown), an enzyme that is unexpressed under normal conditions in most cells, but becomes upregulated only in cells where prostaglandins are upregulated, i.e. during inflammation. Further studies are needed to assess the exact nature and contribution of the infiltrating CD11bþ cells. Possibly, the hyperglycemia could thus simply be the consequence of disturbed glucose homeostasis, or insulin resistance, as can occur in type 2 diabetes. Indeed, IL6 induces insulin resistance in cell culture systems [42,43], underscoring its role in the development of diabetes in man. Nevertheless, both insulin levels in the pancreatic beta-cells and response to insulin and glucose seemed to be normal in the RIP-LCMV/IL6 double Tg mice. Interestingly, pancreatic betacells of double Tg mice did not express detectable levels of GLUT-2, a transmembrane protein that is normally highly expressed in pancreatic beta-cells and is important in the up-take and metabolism of glucose at the cellular level [44,45]. In streptozotocintreated rats, a decrease in the level of GLUT-2 correlated with the diabetic state and with the glucose unresponsiveness of their islet beta-cells [46]. However, it is puzzling why the RIP-GPþ/IL6þ double transgenic mice are normoglycemic under steady state and respond normal to GTT, yet have reduced expression of GLUT-2 in the pancreatic islets. It is possible that the reduction of Glut2 expression in the RIP-GP/IL6 line is either compensated by other GLUTs, or is not sufficiently low to impair the glucose sensitivity. While we do not rule out compensation by other GLUTs, we deem it less likely, because Glut2 is expressed at a very high level in pancreatic beta-cells (and in the basolateral membranes of

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

8

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

intestinal and kidney epithelial cells and of hepatocytes), whereas other Gluts mainly have a different cellular or anatomical distribution [44]. At the moment, we therefore speculate that the remaining expression of Glut2, although no longer detected using standard immunofluorescence microscopy, suffices to maintain normoglycemia in the RIP-GP/IL6 line. This is based on the uniquely high Km of Glut2 for glucose (w17 mM) [47], which ensures fast equilibration of glucose between the extracellular space and the cell cytosol at all physiological or diabetes-associated glycemic levels. This process is so efficient that the rate of glucose metabolism in the beta-cells is controlled at the glucose phosphorylation step and thus modulation of Glut2 surface expression usually does not regulate metabolism except if its reduction is sufficient to limit glucose access to the hexokinases, as can happen in diabetic conditions in beta-cells, or as can be seen in Glut2 gene-knockouts were absence of Glut2 prevents glucose-stimulated insulin secretion by beta-cells (and the regulation of glucose-sensitive gene expression in hepatocytes). Because complete absence of GLUT-2 prevents glucose-stimulated insulin secretion by beta-cells, it is possible that the lack of GLUT-2 in the RIP-NP/IL6 line impedes rapid equilibration between the beta-cell cytosol and the extracellular thereby impairing adequate release of insulin, resulting in a hyperglycemic organism with beta-cells still ‘full’ of insulin. Further studies should unravel these puzzling observations. The RIP-LCMV-GP strain does not express the GP in the thymus, whereas the RIP-LCMV-NP strain expresses the NP in the thymus and deletes the majority of the autoreactive T cells. Nevertheless, in both cases, naive autoreactive CD8þ CTL are present in the periphery [48] and it is therefore unclear how, once crossed with RIPIL6 mice, spontaneous hyperglycemia occurs in one line and less in the other. The very different pattern of diabetes development displayed by RIP-GP and RIP-NP mice when crossed with RIP-IL6 mice was not the consequence of differences in the expression level of IL6 in the pancreas of RIP-NPþ/IL6þ and RIP-GPþ/IL6þ mice. We thus reasoned that the localization patterns of the NP or GP protein can explain the different phenotypes. Based on work by the Buchmeier lab [49], we conclude that the LCMV-GP is detectable in the cytoplasm and on the surface of LCMV-infected cells. Work by Zeller et al. [50] showed that most of the NP protein remains intracellular, although a small part/epitope does appear on the extracellular side. NP does not have a classical signal sequence (to target it to the plasma membrane) e it has 8 hydrophobic regions with a long enough sequence to be transmembrane. In the context of infection it definitely makes it to the plasma membrane e and is transmembrane e but we do not know if it requires accessory proteins (viral GP) to localize effectively or not. It is possible, although speculative, that the lack of a specific signal peptide causes the NP protein to remain inside the cell in the transgenic non-infected setting. However, not much is known about the subcellular localization of NP and GP in the transgenic non-infected setting. One of the few things we know, comes from one of our earlier publications, in which we have shown reactivity to an antiLCMV serum in pancreatic sections of RIP-LCMV-GP strains, and that expression level of GP in beta-cells corresponded to the degree of diabetes pathogenesis [51]. However, the staining did not address the localization of the GP in this transgenic setting. We do know that MHC:peptide complexes must be on the surface of betacells in the transgenic non-infected setting, because when LCMVspecific cytotoxic T lymphocyte (CTL) clones were adoptively transferred into RIP-LCMV-GP or RIP-LCMV-NP mice, they migrated specifically to the islets of Langerhans [24,25]. That said, a difference in surface expression of MHC:peptide complexes is unlikely to explain the difference between the RIP-GP/IL and the RIP-NP/IL6 lines, given the absence of CD4þ or CD8þ T cell infiltration in RIPNPþ/IL6þ double transgenic diabetic mice.

Contrary to surface-bound GP expression, simultaneous expression of predominantly intracellular NP together with IL6 may lead to ER stress resulting in increased beta-cell death and antigen release, thereby fueling the autoimmune response without the need for prior LCMV infection. To address this, we determined the expression of molecules upregulated during the course of the unfolded protein response (UPR), namely C/EBP homologous Protein (CHOP) and spliced Xbox binding protein-1 (sXBP-1) (19; 28). However, experiments on samples from RIP-GP/IL6 transgenic mice revealed no significant differences in expression of CHOP or sXBP-1 between single, double or non-transgenic littermates (data not shown). Likewise, comparing RIP-NPþ/IL6þ versus RIP-GPþ/IL6þ double transgenic mice revealed no differences in expression of the tested ER-stress-related molecules (data not shown). Thus, the mechanisms underlying the different clinical outcomes in RIP-NPþ/ IL6þ and RIP-GPþ/IL6þ mice still warrant further investigation. IL6 plays a crucial role in the pathogenesis of many chronic inflammatory and autoimmune diseases ranging from rheumatoid arthritis and multiple sclerosis to uveoretinitis and asthma [52]. Our data confirm and extend reports on an important pathogenic role for IL6 in diabetes [17,18], consistent with the idea that IL6, like TNF-a and Th1 cytokines, plays a key role in experimental IDDM development [53] and increased levels of IL6 in patients with diabetes [9e12]. Continued interest in the development of anti-IL6 agents might also have a beneficial outcome in type 1 diabetes therapy. 5. Conclusions We have attempted here to study the effect of b-cell specific IL-6 expression on spontaneous and virus-induced diabetes. We have shown here that transgenic expression of IL-6 and LCMV viral proteins in b-cells leads to disappearance of the glucose transporter GLUT-2 in b-cells of both NPþ/IL6þ and GPþ/IL6þ Tg mice. However, diabetes development is dramatically differently affected, depending on whether the viral protein is a surface antigen (glycoprotein, GP) or a nuclear antigen (nucleoprotein, NP). Mice expressing IL-6 and nuclear protein (NP) in the b-cells became spontaneously diabetic. In contrast, mice expressing IL-6 and surface protein (GP) did not become spontaneously diabetic, but showed accelerated LCMV-induced diabetes. Although further research will be needed to unravel the mechanisms underlying these observations, our data underscore the prodiabetic role for IL6, driving spontaneous and virus-induced diabetes, and highlight the need for IL6 antagonists in type 1 diabetes therapy. Conflict of interest statement The authors have no conflict of interest. Acknowledgments We thank Malina McClure and the staff at the LIAI Animal Facility for mouse colony maintenance, Priscilla Colby for administrative support, Malina McClure and Therese Juntti for technical assistance. TLVB and this work were funded by grant 24-2007-439 from the Juvenile Diabetes Research Foundation. PPP received funding from the American Diabetes Association (Mentor-Based Postdoctoral Fellowship 7-09-MN-54) and The Philippe Foundation, Inc. This work was funded by U prevention Center grant (U01 DK078013-05S1). We thank the Brehm Coalition for generous support. The funding sources had no involvement in the study design, interpretation of data, or decision to submit the article for publication.

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

T.L. Van Belle et al. / Journal of Autoimmunity xxx (2014) 1e9

References [1] Ehehalt S, Dietz K, Willasch AM, Neu A, for the D-GB-W. Epidemiological perspectives on type 1 diabetes in childhood and adolescence in Germany: 20 years of the DIARY registry. Diabetes Care 2010;3:338e40. [2] Patterson CC, Dahlquist GG, Gyurus E, Green A, Soltesz G, Group ES. Incidence trends for childhood type 1 diabetes in Europe during 1989e2003 and predicted new cases 2005e20: a multicentre prospective registration study. Lancet 2009;373:2027e33. [3] Van Belle TL, Coppieters KT, von Herrath MG. Type 1 diabetes: etiology, immunology, and therapeutic strategies. Physiol Rev 2011;91:79e118. [4] Hyoty H, Hiltunen M, Knip M, Laakkonen M, Vahasalo P, Karjalainen J, et al. A prospective study of the role of coxsackie B and other enterovirus infections in the pathogenesis of IDDM. Childhood Diabetes in Finland (DiMe) Study Group. Diabetes 1995;44:652e7. [5] Eizirik DL, Colli ML, Ortis F. The role of inflammation in insulitis and beta-cell loss in type 1 diabetes. Nat Rev Endocrinol 2009;5:219e26. [6] Kristiansen OP, Mandrup-Poulsen T. Interleukin-6 and diabetes: the good, the bad, or the indifferent? Diabetes 2005;54(Suppl. 2):S114e24. [7] Kamimura D, Ishihara K, Hirano T. IL-6 signal transduction and its physiological roles: the signal orchestration model. Rev Physiol Biochem Pharmacol 2003;149:1e38. [8] Camporeale A, Poli V. IL-6, IL-17 and STAT3: a holy trinity in auto-immunity? Front Biosci (Landmark Ed) 2012;17:2306e26. [9] Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. J Am Med Assoc 2001;286:327e34. [10] Wedrychowicz A, Dziatkowiak H, Sztefko K. Interleukin-6 (IL-6) and IGFIGFBP system in children and adolescents with type 1 diabetes mellitus. Exp Clin Endocrinol Diabetes 2004;112:435e9. [11] Trinanes J, Salido E, Fernandez J, Rufino M, Gonzalez-Posada JM, Torres A, et al. Type 1 diabetes increases the expression of proinflammatory cytokines and adhesion molecules in the artery wall of candidate patients for kidney transplantation. Diabetes Care 2012;35:427e33. [12] Gabbay MA, Sato MN, Duarte AJ, Dib SA. Serum titres of anti-glutamic acid decarboxylase-65 and anti-IA-2 autoantibodies are associated with different immunoregulatory milieu in newly diagnosed type 1 diabetes patients. Clin Exp Immunol 2012;168:60e7. [13] Campbell IL, Cutri A, Wilson A, Harrison LC. Evidence for IL-6 production by and effects on the pancreatic beta-cell. J Immunol 1989;143:1188e91. [14] Anderson JT, Cornelius JG, Jarpe AJ, Winter WE, Peck AB. Insulin-dependent diabetes in the NOD mouse model. II. Beta cell destruction in autoimmune diabetes is a TH2 and not a TH1 mediated event. Autoimmunity 1993;15: 113e22. [15] Rabinovitch A, Suarez-Pinzon WL. Role of cytokines in the pathogenesis of autoimmune diabetes mellitus. Rev Endocr Metabolic Disord 2003;4:291e9. [16] DiCosmo BF, Picarella D, Flavell RA. Local production of human IL-6 promotes insulitis but retards the onset of insulin-dependent diabetes mellitus in nonobese diabetic mice. Int Immunol 1994;6:1829e37. [17] Campbell IL, Hobbs MV, Dockter J, Oldstone MB, Allison J. Islet inflammation and hyperplasia induced by the pancreatic islet-specific overexpression of interleukin-6 in transgenic mice. Am J Pathol 1994;145:157e66. [18] Campbell IL, Kay TW, Oxbrow L, Harrison LC. Essential role for interferongamma and interleukin-6 in autoimmune insulin-dependent diabetes in NOD/Wehi mice. J Clin Invest 1991;87:739e42. [19] von Herrath M, Holz A. Pathological changes in the islet milieu precede infiltration of islets and destruction of beta-cells by autoreactive lymphocytes in a transgenic model of virus-induced IDDM. J Autoimmun 1997;10:231e8. [20] Mosser DM, Edwards JP. Exploring the full spectrum of macrophage activation. Nat Rev Immunol 2008;8:958e69. [21] von Herrath MG, Dockter J, Oldstone MB. How virus induces a rapid or slow onset insulin-dependent diabetes mellitus in a transgenic model. Immunity 1994;1:231e42. [22] Olefsky JM, Glass CK. Macrophages, inflammation, and insulin resistance. Annu Rev Physiol 2010;72:219e46. [23] Xie WL, Chipman JG, Robertson DL, Erikson RL, Simmons DL. Expression of a mitogen-responsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Natl Acad Sci U S A 1991;88:2692e6. [24] Ohashi PS, Oehen S, Buerki K, Pircher H, Ohashi CT, Odermatt B, et al. Ablation of "tolerance" and induction of diabetes by virus infection in viral antigen transgenic mice. Cell 1991;65:305e17. [25] Oldstone MB, Nerenberg M, Southern P, Price J, Lewicki H. Virus infection triggers insulin-dependent diabetes mellitus in a transgenic model: role of anti-self (virus) immune response. Cell 1991;65:319e31. [26] Van Belle TL, Esplugues E, Liao J, Juntti T, Flavell RA, von Herrath MG. Development of autoimmune diabetes in the absence of detectable IL-17A in a CD8-driven virally induced model. J Immunol 2011;187:2915e22. [27] Zhou L, Ivanov II, Spolski R, Min R, Shenderov K, Egawa T, et al. IL-6 programs T(H)-17 cell differentiation by promoting sequential engagement of the IL-21 and IL-23 pathways. Nat Immunol 2007;8:967e74.

9

[28] Horwitz MS, Bradley LM, Harbertson J, Krahl T, Lee J, Sarvetnick N. Diabetes induced by Coxsackie virus: initiation by bystander damage and not molecular mimicry. Nat Med 1998;4:781e5. [29] Serreze DV, Ottendorfer EW, Ellis TM, Gauntt CJ, Atkinson MA. Acceleration of type 1 diabetes by a coxsackievirus infection requires a preexisting critical mass of autoreactive T-cells in pancreatic islets. Diabetes 2000;49: 708e11. [30] Tracy S, Drescher KM, Chapman NM, Kim KS, Carson SD, Pirruccello S, et al. Toward testing the hypothesis that group B coxsackieviruses (CVB) trigger insulin-dependent diabetes: inoculating nonobese diabetic mice with CVB markedly lowers diabetes incidence. J Virol 2002;76:12097e111. [31] Filippi CM, Estes EA, Oldham JE, von Herrath MG. Immunoregulatory mechanisms triggered by viral infections protect from type 1 diabetes in mice. J Clin Invest 2009;119:1515e23. [32] Drescher KM, Kono K, Bopegamage S, Carson SD, Tracy S. Coxsackievirus B3 infection and type 1 diabetes development in NOD mice: insulitis determines susceptibility of pancreatic islets to virus infection. Virology 2004;329:381e 94. [33] Martinic MM, Juedes AE, Bresson D, Homann D, Skak K, Huber C, et al. Minimal impact of a de novo-expressed beta-cell autoantigen on spontaneous diabetes development in NOD mice. Diabetes 2007;56:1059e68. [34] Oldstone MB. Prevention of type I diabetes in nonobese diabetic mice by virus infection. Science 1988;239:500e2. [35] Christen U, Benke D, Wolfe T, Rodrigo E, Rhode A, Hughes AC, et al. Cure of prediabetic mice by viral infections involves lymphocyte recruitment along an IP-10 gradient. J Clin Invest 2004;113:74e84. [36] Kanno T, Kim K, Kono K, Drescher KM, Chapman NM, Tracy S. Group B coxsackievirus diabetogenic phenotype correlates with replication efficiency. J Virol 2006;80:5637e43. [37] Bettelli E, Oukka M, Kuchroo VK. T(H)-17 cells in the circle of immunity and autoimmunity. Nat Immunol 2007;8:345e50. [38] Dienz O, Eaton SM, Bond JP, Neveu W, Moquin D, Noubade R, et al. The induction of antibody production by IL-6 is indirectly mediated by IL-21 produced by CD4þ T cells. J Exp Med 2009;206:69e78. [39] Sutherland APR, Van Belle T, Wurster AL, Suto A, Michaud M, Zhang D, et al. Interleukin-21 is required for the development of type 1 diabetes in NOD mice. Diabetes 2009;58:1144e55. [40] Van Belle TL, Nierkens S, Arens R, von Herrath MG. Interleukin-21 receptormediated signals control autoreactive T cell infiltration in pancreatic islets. Immunity 2012;36:1060e72. [41] Shoda LK, Young DL, Ramanujan S, Whiting CC, Atkinson MA, Bluestone JA, et al. A comprehensive review of interventions in the NOD mouse and implications for translation. Immunity 2005;23:115e26. [42] Lagathu C, Bastard JP, Auclair M, Maachi M, Capeau J, Caron M. Chronic interleukin-6 (IL-6) treatment increased IL-6 secretion and induced insulin resistance in adipocyte: prevention by rosiglitazone. Biochem Biophys Res Commun 2003;311:372e9. [43] Rotter V, Nagaev I, Smith U. Interleukin-6 (IL-6) induces insulin resistance in 3T3-L1 adipocytes and is, like IL-8 and tumor necrosis factor-alpha, overexpressed in human fat cells from insulin-resistant subjects. J Biol Chem 2003;278:45777e84. [44] Thorens B, Mueckler M. Glucose transporters in the 21st century. Am J Physiol Endocrinol Metab 2010;298:E141e5. [45] Orci L, Thorens B, Ravazzola M, Lodish HF. Localization of the pancreatic beta cell glucose transporter to specific plasma membrane domains. Science 1989;245:295e7. [46] Thorens B, Weir GC, Leahy JL, Lodish HF, Bonner-Weir S. Reduced expression of the liver/beta-cell glucose transporter isoform in glucose-insensitive pancreatic beta cells of diabetic rats. Proc Natl Acad Sci U S A 1990;87: 6492e6. [47] Thorens B. Molecular and cellular physiology of GLUT-2, a high-Km facilitated diffusion glucose transporter. Int Rev Cytol 1992;137:209e38. [48] von Herrath MG, Guerder S, Lewicki H, Flavell RA, Oldstone MB. Coexpression of B7-1 and viral (“self”) transgenes in pancreatic beta cells can break peripheral ignorance and lead to spontaneous autoimmune diabetes. Immunity 1995;3:727e38. [49] Wright KE, Spiro RC, Burns JW, Buchmeier MJ. Post-translational processing of the glycoproteins of lymphocytic choriomeningitis virus. Virology 1990;177: 175e83. [50] Zeller W, Bruns M, Lehmann-Grube F. Lymphocytic choriomeningitis virus. X. Demonstration of nucleoprotein on the surface of infected cells. Virology 1988;162:90e7. [51] Martinic MM, Huber C, Coppieters K, Oldham JE, Gavin AL, von Herrath MG. Expression level of a pancreatic neo-antigen in beta cells determines degree of diabetes pathogenesis. J Autoimmun 2010;35:404e13. [52] Neurath MF, Finotto S. IL-6 signaling in autoimmunity, chronic inflammation and inflammation-associated cancer. Cytokine Growth Factor Rev 2011;22: 83e9. [53] Rabinovitch A. An update on cytokines in the pathogenesis of insulindependent diabetes mellitus. Diabetes Metab Rev 1998;14:129e51.

Please cite this article in press as: Van Belle TL, et al., Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes, Journal of Autoimmunity (2014), http://dx.doi.org/10.1016/j.jaut.2014.02.002

Beta-cell specific production of IL6 in conjunction with a mainly intracellular but not mainly surface viral protein causes diabetes.

Inflammatory mechanisms play a key role in the pathogenesis of type 1 and type 2 diabetes. IL6, a pleiotropic cytokine with impact on immune and non-i...
3MB Sizes 1 Downloads 3 Views