2184 Diabetes Volume 64, June 2015

Pamela J. Lincez, Iryna Shanina, and Marc S. Horwitz

Reduced Expression of the MDA5 IFIH1 Prevents Autoimmune Diabetes

Diabetes 2015;64:2184–2193 | DOI: 10.2337/db14-1223

Although it is widely accepted that (T1D) further investigation of RNA sensing in T1D patho- is the result of the autoimmune destruction of insulin- genesis (2–4). Understanding how MDA5, a double- producing b-cells in the pancreas, little is known about stranded RNA (dsRNA) virus sensor expressed from the the events leading to islet autoimmunity. Epidemiologi- T1D risk gene IFIH1, controls type I (IFN-I) cal and genetic data have associated virus infections and consequently T1D is critical in understanding the and antiviral type I interferon (IFN-I) response events that lead to autoimmunity and define disease pro- with T1D. Genetic variants in the T1D risk locus inter- gression to develop prophylactic and therapeutic measures. IFIH1) feron induced with C domain 1 ( have In children at risk for T1D, an IFN-I transcriptional fi been identi ed by genome-wide association studies signature precedes islet autoimmunity (5). Recent onset to confer resistance to T1D and result in the reduction of T1D is strongly associated with infection by RNA in expression of the intracellular RNA virus sensor such as enteroviruses including coxsackievirus, known as melanoma differentiation–associated protein further implicating the IFN-I signature (6,7). MDA5 spe- 5 (MDA5). Here, we translate the reduction in IFIH1 gene cifically detects dsRNA intermediates from viruses like expression that results in protection from T1D. Our func- tional studies demonstrate that mice heterozygous at coxsackievirus that are produced in the cytoplasm during the Ifih1 gene express less than half the level of MDA5 RNA virus replication. Upon recognition of its dsRNA fi protein, which leads to a unique antiviral IFN-I signature ligand and proper assembly of a lament structure around and adaptive response after virus infection that protects the dsRNA stem, MDA5 activates and triggers signaling from T1D. IFIH1 heterozygous mice have a regulatory from IFN-b promoter stimulator-1 (IPS-1, also known as rather than effector T-cell response at the site of auto- the mitochondrial antiviral signaling protein MAVS) and

IMMUNOLOGY AND TRANSPLANTATION immunity, supporting IFIH1 expression as an essential IFN regulatory factor-3 and -7 molecules to induce the regulator of the diabetogenic T-cell response and pro- transcription of IFN-b. The specificity and kinetics of viding a potential mechanism for patients carrying IFIH1 MDA5-RNA binding as well as MDA5-induced IFN-I protective polymorphisms. responses that result from viral infection have been well studied (8). Recently, it was demonstrated in mice that a missense mutation, G821S, in the Ifih1 gene disrupts Type 1 diabetes (T1D) is a devastating organ-specific MDA5 responsiveness to dsRNA and allows MDA5 to re- disease resulting from the autoimmune destruction of main constitutively active and induce lupus-like nephritis pancreatic b-cells (1). The events leading to autoimmunity and autoimmunity (9). Funabiki et al. (9) also demon- in T1D are complex and unclear, demanding the design of strated constitutive IFN-I signaling with the common new treatments that consider both the strong genetic in- T1D risk variant A946T. Further, nondiabetogenic mice fluence and environmental stressors linked to the disease. (C57BL/6 mice) heterozygous at Ifih1 develop rapid hy- Identification of protective polymorphisms in the intra- perglycemia after infection with the pancreatropic virus cellular virus receptor melanoma differentiation–associated encephalmyocarditis virus (EMCV)-D owing to virus- protein 5 (MDA5) gene interferon induced with helicase C directed b-cell death (10). Unlike EMCV-D, coxsackievirus domain 1 (IFIH1) that lead to a reduction in MDA5 urges serotype B4 (CB4) infects and, without directly killing

Department of Microbiology and Immunology, The University of British Columbia, This article contains Supplementary Data online at http://diabetes Vancouver, British Columbia, Canada .diabetesjournals.org/lookup/suppl/doi:10.2337/db14-1223/-/DC1. Corresponding author: Marc S. Horwitz, [email protected]. © 2015 by the American Diabetes Association. Readers may use this article as fi Received 8 August 2014 and accepted 7 January 2015. long as the work is properly cited, the use is educational and not for pro t, and the work is not altered. diabetes.diabetesjournals.org Lincez, Shanina, and Horwitz 2185 pancreatic b-cells, induces diabetes in NOD mice (11,12); Cruz Biotech, Santa Cruz, CA) primary antibodies and more intriguingly, it has been strongly associated in recent- IRDye 800CW and IRDye 680 RD secondary antibodies onset T1D in patients (6,7,13). (LI-COR). Membranes were scanned with the LI-COR To better our understanding of the immunological Odyssey Scanner (LI-COR). Protein was quantified using consequences after MDA5 activation and their effects on LI-COR Odyssey 3.0 software. autoimmunity in a susceptible host, we developed a mouse model resonant with IFIH1-protected patients by back- Virus crossing mice deficient in IFIH1 (also known as MDA5) Ten- to 12-week-old mice were infected with sublethal onto the accepted mouse model for T1D, NOD/Ltj mice, doses of 400 plaque-forming units i.p. CB4 Edwards strain and studied mice heterozygote for the deficient MDA5 2 diluted in DMEM. As there is no sex bias in CB4- allele. We demonstrate that these heterozygote mice mediated T1D, equal numbers of male and female mice 2 (MDA5+/ ) express roughly half the level of MDA5 pro- were infected with CB4. Virus stocks were prepared as tein as wild-type mice (MDA5+/+) and when infected with previously described (11). +/2 CB4, a clinically relevant stimulator of T1D, the MDA5 Virus Titer fi mice were protected from T1D, thereby de ning the abil- Free virus particles were detected from tissue homoge- ity of MDA5 to augment autoimmunity and control T1D. nates by plaque assay as previously described (11).

RESEARCH DESIGN AND METHODS Flow Cytometry Mice Pancreatic lymph node and splenic single-cell suspensions NODmicewerepurchasedfromTheJacksonLaboratory were counted and stained with fluorescently conjugated 2 2 (Bar Harbor, ME). MDA5 knockout (MDA5 / )miceonthe monoclonal antibodies (mAbs) for cell-surface markers C57BL/6 background were a generous gift from Dr. M. CD4 (clone L3T4), CD8 (53-6.7), CD25 (clone PC61), Colonna (The Washington University School of Medicine, CD11b (clone M1/70), CD11c (clone HL3), CD44 (clone Department of Pathology and Immunology). We successfully IM7), and CD62L (clone MEL-14); intracellular transcrip- 2 2 backcrossed MDA5 / mice onto the NOD background and tion factors Foxp3 (clone FJK-16s) and Helios (clone confirmed by single nucleotide polymorphism analysis (per- 22F6); and the inflammatory cytokine IFN-g (XMG1.2). formed by our group and by DartMouse, Lebanon, NH) that All mAbs were purchased from eBiosciences (San Diego, they carry the full complement of NOD idd alleles. More CA) with the exception of Helios from BioLegend (San importantly, the ability to develop spontaneous diabetes is Diego, CA). Stained cells were analyzed by flow cytometry strongly indicative that the required susceptibility loci have with the BD Biosciences LSR II (San Jose, CA) and Flow Jo crossed over and, to this end, littermates to the backcrosses vX.0.6 software (TreeStar, Ashland, OR). 2 2 that were either heterozygote or wild type for the MDA5 / 2 2 alleles developed spontaneous diabetes. MDA5 / mice Intracellular Cytokine Staining 2 were bred with NOD mice and MDA5+/ mice. MDA5+/+ Single-cell suspensions from pancreatic lymph nodes and progeny were bred for further use in experiments. All spleens were restimulated for 4 h at 37°C in Iscove modi- fi miceweremaintainedintheCentreforDiseaseModeling ed DMEM containing 10% FBS with 500 ng/mL phorbol (Life Sciences Centre, Vancouver, BC, Canada) and kept in 12-myristate 13-acetate, 10 ng/mL ionomycin, and Golgi a pathogen-free environment. Diabetes incidence was mon- Plug (BD Biosciences). Cells were stained for surface fi fl itored by nonfasting blood glucose measurements. Disease markers, xed, permeabilized, stained for in ammatory fl onset was determined by two consecutive blood glucose cytokine IFN-g, and analyzed by ow cytometry. . levels 300 mg/dL. Only prediabetic mice were used for In vitro T-Cell Activation experiments.Allanimalworkwasperformedinstrictaccor- Splenic CD4+ T cells were isolated from uninfected dance with the recommendations of the Canadian Council MDA5+/+ mice and both CB4-infected MDA5+/+ and 2 for Animal Care. The protocol was approved by the Animal MDA5+/ mice at day 7 postinfection (pi) using the Easy- Care Committee of The University of British Columbia (cer- Sep Mouse CD4+ T-cell pre-enrichment kit and “The Big fi ti cate numbers A08-0415 and A08-0622). Easy” Silver EasySep Magnet (Stemcell, Vancouver, BC). Western Blotting CD4+ T cells (5 3 107 cells/mL) were then enriched for Mice were stimulated by injection with 100 mg i.p. poly- CD25 activation using the EasySep Mouse CD25 Positive 2 inosinic:polycytidylic acid (P1530; Sigma, St. Louis, MO). Selection kit (Stemcell). CD4+ CD25 T cells from unin- After 24 h of stimulation, spleens were isolated and ho- fected MDA5+/+ mice were mixed in ratios of 2:1, 4:1, 8:1, mogenized by sonication and tissue homogenates were and16:1withCD4+ CD25+ T cells from either CB4-infected 2 lysed with CellLytic MT Mammalian Tissue Lysis Reagent MDA5+/+ or MDA5+/ mice (at a final total concentration (Sigma). Samples were separated on 10% SDS-PAGE, of 3 3 106 cells per 1 mL of RPMI-1640 containing 10% transferred to polyvinylidene fluoride membranes, blocked FCS, 50 mmol/L 2-mercaptoethanol, and penicillin/strepto- with Odyssey Blocking Buffer (LI-COR, Lincoln, NE) and mycin) in a 96-well plate coated for 24 h with anti-CD3e probed with monoclonal rabbit anti-MDA5 (Cell Signaling, mAb (1 mg/mL; BioLegend) and anti-CD28 mAb (1 mg/mL; Danvers, MA) and polyclonal goat anti-tubulin (Santa BD Pharmingen). Mixed cells were also cocultured in 2186 IFIH1 Expression and Autoimmune Diabetes Diabetes Volume 64, June 2015 uncoated wells as a control, and all cells were cultured for AATGAGACTATTG-39, reverse 59-TTCTGAGGCATCAACT 72 h at 37°C. For assessment of T-cell activation, cells were GACAGGTC-39; IFN-a forward 59-TGATGAGCTACTGGT stained with anti-CD4-Pacific Blue and anti-CD25-PE CAGC-39, reverse 59-GATCTCTTAGCACAAGGATGGC-39; (eBioscience). For assessment of T-cell effector function, TLR3 forward 59-GAGAGAGATTCTGGATGCTTGTGTTTG- cells were stimulated with phorbol 12-myristate 13-acetate 39,reverse59-GTCTCATAATGGTTTATCATCTACAAA-39; and ionomycin (Sigma) in the presence of BD Golgi Plug and GAPDH forward 59-AGGTCGGTGTGAACGGATTTG-39, (BD Biosciences) for 4 h at 37°C. The cells were then reverse 59-TGTAGACCATGTAGTTGAGGTCA-39). PCR am- stained with anti-CD4-Pacific Blue and anti-CD25-PE be- plification was performed in 384-well plates with the ABI fore being fixed and permeabilized for intracellular staining 7900HT Fast Real-Time PCR System (Applied Biosystems). with anti-IFN-g-PECy7 (eBioscience). Data were acquired All samples from three independent experiments were using an LSRII flow cytometer and analyzed with FlowJo evaluated in duplicate amplification reactions. mRNA ex- software vX.0.6. pression was normalized to GAPDH. The comparative Ct method was used as previously described (14), and data Immunohistochemical Staining (or Islet Pathology) are shown as 2DC and fold change of 2DC relative to Pancreases were fixed in 70% ethanol for 24 h and t t wild-type (NOD) samples (14). paraffin embedded (Wax-IT, Vancouver, BC, Canada) as previously described (11). Serial tissue sections were Adoptive Transfer Studies stained using standard procedures for hematoxylin-eosin Spleens harvested from donor uninfected mice were used to analyze the anatomical structure and were scored for to generate single-cell suspensions. Cells were stained insulitis according to a three-tiered scale. with fluorescently conjugated mAbs against CD11b and CD11c and sorted with a FACSAria flow cytometer (BD Cytokine Analysis 2 Biosciences). A total of 100,000 CD11b+CD11c and Cytokines interleukin (IL)-2, IL-4, IL-6, IL-10, IL-17, CD11b+CD11c+ cells were sorted (found to be 96% tumor necrosis factor-a, and IFN-g were measured from pure) and, diluted in phosphate-buffered saline with 2% serumatdays0,3,and7post–CB4 infection in a multiplexed FBS, were adoptively transferred intraperitoneally into format using a Cytometric Bead Array (mouse Th1/Th2/ uninfected recipient mice. Recipient mice were infected Th17 cytokine kit; BD Biosciences, Mississauga, ON, intraperitoneally 24 h after adoptive transfer with CB4 Canada). Type 1 IFN-a and -b were measured Edwards strain 2, and spleens, pancreatic lymph nodes, from serum by ELISA using VeriKine Mouse Interferon-a pancreas, and serum were harvested at day 7 pi for anal- and -b ELISA kits (PBL Interferon Source, Piscataway, ysis. Control mice did not receive cells and were infected NJ). with CB4 Edwards strain 2 at the same time as experi- RNA Isolation mental recipient mice. Organs were removed and immediately snap frozen in Statistical Analysis TRIzol reagent (Life Technologies, Burlington, ON, Can- GraphPad Prism 6.0 software (GraphPad, San Diego, CA) ada). Tissues were weighed and organs were homogenized using the Student t test (two-tailed distribution) and a using QIAGEN stainless steel beads and TissueLyser II P value ,0.05 determined statistical significance. Serum benchtop homogenizer at 19/s for 10 min. Total RNA was cytokine concentrations were determined with FCAP prepared with TRIzol reagent according to the manufac- Array Software (BD Biosciences). Data are presented as turer’s protocol (TRIzol; Life Technologies). RNA was means 6 SEM. quantified using a NanoDrop-ND-1000 (Thermo Scien- fi ti c, Wilmington, DE). RESULTS Reverse Transcription and Quantitative Reduction in MDA5 Protects NOD Mice From T1D Real-Time PCR Since patients carrying IFIH1 protective variants are het- cDNA was prepared for 1 mg RNA using a High Capacity erozygous, we followed T1D incidence and pathologies in 2 cDNA Reverse Transcription kit (Applied Biosystems, Fos- NOD mice heterozygous for the MDA5 gene (MDA5+/ ). 2 ter City, CA) according to the manufacturer’s instructions. MDA5+/ mice were generated by successfully backcross- 2 2 RT-PCR was performed with the BioRad T-100 Thermal ing C57BL/6 MDA5 / mice onto the NOD background Cycler. and confirmed by DartMouse. Western blot confirmed cDNA was diluted with UltraPure DNase/RNase-Free that after stimulation with the RNA mimetic, polyinosinic: 2 2 Distilled Water (Life Technologies), and a final RNA polycytidylic acid, MDA5 / mice were unable to gener- 2 concentration equivalent to 10 mg/mL was used for real- ate MDA5 protein, while MDA5+/ mice produced 48% time PCR. for MDA5, IFN-b, IFN-a, less protein than their MDA5+/+ littermates (Fig. 1A). TLR3, and GAPDH was quantified using the iQ SYBR MDA5 expression was also markedly reduced in the Green Supermix (BioRad, Mississauga, ON, Canada) and spleens and in different antigen-presenting cell (APC) specific primers (mouse MDA5 forward 59-GTGATG subsets from the pancreatic lymph nodes (PLNs) of 2 ACGAGGCCAGCAGTTG-39, reverse 59-ATTCATCCGTTT CB4-infected MDA5+/ mice at day 3 pi compared with CGTCCAGTTTCA-39; IFN-b forward 59-GCACTGGGTGG CB4-infected MDA5+/+ mice (Fig. 1D; Supplementary diabetes.diabetesjournals.org Lincez, Shanina, and Horwitz 2187

Figure 1—A reduction in MDA5 protects NOD mice from spontaneous and virus-mediated T1D. A: MDA5+/+, MDA5+/2, and MDA52/2 mice were stimulated with polyinosinic:polycytidylic acid as described in RESEARCH DESIGN AND METHODS. After 24 h stimulation, a 48% reduction in MDA5 protein from MDA5+/2 spleens compared with MDA5+/+ was confirmed by Western blot. B: Diabetes incidence was monitored in MDA5+/+ (n = 41), MDA5+/2 (n = 23), and MDA52/2 (n = 20) mice for 30 weeks. Two consecutive blood glucose levels >300 mg/dL determined diabetes incidence. C: Serial sections of pancreas tissue from MDA5+/+ and MDA5+/2 at days 3 and 7 post–CB4 infection were stained for hematoxylin-eosin and scored for insulitis according to a three-tiered scale (no insulitis, peri-insulitis, and insulitis). Infected 10- to 12-week-old MDA5+/+ (n = 10) and MDA5+/2 (n = 15) and uninfected MDA5+/+ (n = 12) and MDA5+/2 (n = 8) mice were monitored for cumulative diabetes incidence up to 14 days pi (D), and viral titers were quantified by standard plaque assay from pancreas and spleens at 3, 7, and 14 days pi (E). Results are shown as means 6 SEM, and statistical significance was determined by Student t test. *P < 0.05.

Fig. 1). Importantly, we observed that MDA5 deficiency population of autoreactive T cells to elicit autoimmunity 2 2 (MDA5 / ) protected mice from spontaneous disease, (11,12). Resident APCs engulf infected b-cells, present 2 while loss of a single MDA5 allele (MDA5+/ ) reduced sequestered islet and viral antigens to resting autoreactive the incidence of spontaneous diabetes compared with T cells, and induce a population of T cells directed at the virus MDA5+/+ littermates (Fig. 1B). These results implicate and specific to the pancreatic islets. The b-cell–directed MDA5 signaling in the development of T1D in NOD T cells then destroy the islets and accelerate insulin loss 2 mice and justify the study of MDA5+/ as a model for and T1D disease pathology (11). Though strongly linked the protective IFIH1 variants found in patients. to T1D in patients, CB4 may be responsible for only a sub- CB4 has been isolated from patients with T1D, set of new-onset cases, as other viruses such as CB1 and associated with disease onset, and found to accelerate rotavirus have been more recently implicated (13). As CB4 NOD diabetes. CB4 infects, but does not directly kill, is a clinically relevant inducer of T1D that stimulates pancreatic b-cells and instead activates a preexisting MDA5 and accelerates NOD diabetes, it clearly models 2188 IFIH1 Expression and Autoimmune Diabetes Diabetes Volume 64, June 2015 the events leading to T1D in mice by mimicking environ- Toll-like receptor 3 (TLR3het mice) were infected and mental influences that lead to T1D. Cumulative diabetes measured for serum IFN-I. TLR3het mice show an in- 2 incidence was monitored after infection with CB4 in 10- crease in IFN-b earlier than MDA5+/ by day 2 pi and 2 2 to 12-week-old MDA5+/ mice. We observed that MDA5+/ asignificant rise in IFN-a at day 3 compared with mice were completely protected from the development of MDA5+/+ mice (Fig. 2C). T1D compared with infected age-matched MDA5+/+ litter- Further, the variance in IFN-I production in the serum 2 mates that developed disease incidence at 50% by 14 of our MDA5+/ mice at day 3 pi was reflected in the level days pi (Fig. 1D). As expected of mice 10–12 weeks old, of mRNA expression, where pancreas and spleen tissue 2 2 uninfected MDA5+/ and MDA5+/+ littermates did not from infected MDA5+/ mice also showed a unique profile develop disease (Fig. 1D). It is astonishing, however, of IFN-I expression (Fig. 2D). IFN-b expression was ele- 2 2 that MDA5+/ mice demonstrated sufficient antiviral vated in MDA5+/ pancreas, whereas IFN-a was greater 2 responses to clear the virus similarly to MDA5+/+ mice in MDA5+/ spleen at day 3 pi compared with infected 2 (Fig. 1E), despite the reduction in MDA5 expression in MDA5+/+. The increase in IFN-b expression in MDA5+/ the heterozygote mice (Fig. 1A and D; Supplementary Fig. pancreas also correlated with higher levels of MDA5 ex- 1). Overall, virus was cleared in both MDA5+/+ and pression compared with infected MDA5+/+ pancreas. In 2 2 MDA5+/ mice by day 14 pi, with an equal rate of clear- MDA5+/ spleen, where MDA5 expression was lower ance in the pancreas, and a slightly slower rate in the than MDA5+/+, we observed elevated TLR3 and IFN-a 2 spleen for the MDA5+/ mice (Fig. 1E). Despite reduced expression (Fig. 2D). It is possible that the unique IFN-I 2 expression and function of a critical innate immune sen- profile observed in MDA5+/ mice acts in concert with 2 sor of CB4, the MDA5+/ mice did not exhibit immuno- TLR3-induced IFN-I responses to control virus infection, suppression or a reduced ability to handle virus infection. yet sustain a level of IFN-I production that does not abro- As expected, after infection, diabetic MDA5+/+ mice gate further inflammation and the onset of autoimmunity. exhibited a high degree of pancreatic insulitis, reflecting For establishment of whether other cytokines were 2 the observed hyperglycemia (11,12), while MDA5+/ mice equally affected after infection under conditions of at the same time pi lacked the same level of insulitis, reduced MDA5 signaling, serum was sampled from both reflecting their protected status (Fig. 1C). While mice mice after infection and measured for inflammatory with a full complement of MDA5 show dramatic increases cytokines, and the levels of IL-6, IL-10, tumor necrosis in pancreatic islet insulitis from the peak of infection (day factor-a, and IFN-g were similar from both pre– and 2 2 3 pi) until the initiation of diabetes (day 7 pi), MDA5+/ post–CB4 infection when compared between MDA5+/ mice show no increase in insulitis. This suggested that mice and MDA5+/+ mice (Fig. 2A). While the differences reduced MDA5 expression and function altered the ability in MDA5 innate signaling altered the IFN-I signature, no of effector T cells to home to the pancreatic islets and overall change was observed in other major inflammatory destroy the pancreatic b-cells. mediators.

MDA5+/2 Mice Have a Unique IFN-I Signature After Reduction in MDA5 Induces Regulatory Rather Than CB4 Infection Effector Immune Responses After the sensing of dsRNA, MDA5 triggers a signaling Type 1 IFNs have diverse immunomodulatory functions pathway leading to the induction of an antiviral response that play an important role in many autoimmune driven by type 1 IFNs. Analysis of serum levels of IFN-a diseases including T1D (15). Dendritic cell activation and -b showed a significantly different IFN-I pattern of and the presentation of sequestered self-antigens to 2 expression in the MDA5+/ compared with MDA5+/+ preexisting autoreactive T cells can be directly affected 2 mice. At 3 days post–CB4 infection, MDA5+/ showed by IFN-a and -b (16). To determine whether the changes a significantly greater amount of IFN-b in the serum in MDA5 and IFN-I expression within the infected 2 compared with MDA5+/+ (Fig. 2C). By day 7, the MDA5+/ mice altered APC activation and subsequent 2 MDA5+/ mice had returned to preinfection levels of T-cell polarization leading to protection, we analyzed IFN-b, where IFN-b levels in MDA5+/+ mice steadily the expression of MHC and costimulatory molecules rose over the 7-day time course, with greater IFN-b at (CD40, CD80, CD86, and F4/80) on APCs (CD11b+CD11c+ 2 day 7 than MDA5+/ mice (Fig. 2C). In both mice, IFN-a or CD11b+CD11c- cells) from the spleens and PLNs of 2 2 levels steadily rose pi, with greater levels in the MDA5+/ MDA5+/ mice and wild-type mice. At 48 h pi, no differ- mice. The differences in IFN-I signature pi are best repre- encesinactivationofAPCsfrombothspleenandPLNs 2 sented as a ratio of IFN-b to IFN-a,wheretheMDA5+/ were observed between the infected mice (Supplemen- mice show a rise at day 3 and return to balanced levels tary Fig. 2). by day 7, while the MDA5+/+ mice show a steady increase Development of spontaneous diabetes in NOD mice is in the ratio to maximum at day 7 (Fig. 2B). For estab- critically linked to the balance and polarization of effector lishment of whether a concomitant increase was reflected and regulatory T cells (17). A growing pancreatic insulitis in mice deficient for the other major sensor of coxsackie- composed of diabetogenic effector T cells occurs over time virus dsRNA, NOD mice heterozygously deficient for and results in the loss and destruction of insulin- diabetes.diabetesjournals.org Lincez, Shanina, and Horwitz 2189

Figure 2—MDA5 expression alters the levels of type 1 IFN and not inflammatory cytokines that respond to CB4 infection. Cytokines were measured by FACS bead array (A) and ELISA (B and C) in sera harvested from MDA5+/+ (n =8–15) and MDA5+/2 (n =8–15) mice before infection (day 0) and days 2, 3, and 7 post–CB4 infection. Statistical significance was determined by Student t test. *P < 0.05, **P < 0.001, and ****P < 0.0001. D: Relative mRNA expression levels of MDA5, TLR3, and IFN-I from the spleen and pancreas from MDA5+/+ (n =8–10) and MDA5+/2 (n =8–10) mice at days 0 (not shown), 3, and 7 (not shown) post–CB4 infection were quantified by quantitative real-time +/+ PCR and normalized to GAPDH. The comparative Ct method was used to calculate mean relative expression 6 SEM against MDA5 mice as described in RESEARCH DESIGN AND METHODS. Data shown are from duplicate samples from two independent experiments.

producing b-cells in the islets. CB4 infection acts to ac- compared with infected MDA5+/+ mice that have greater celerate this insulitis by exposing infected b-cells to the CD4+ IFN-g T cells in the PLNs (Fig. 3D). autoreactive T cells causing reactivation (11). The induc- To demonstrate regulatory T-cell function, we per- tion of diabetes is the result of an alteration in the bal- formed Treg suppression assays by stimulating CD4+ 2 ance between effector autoreactive T cells and regulatory CD25 T cells from uninfected MDA5+/+ mice and 2 T cells forming the pancreatic insulitis. In MDA5+/ mice testing whether CD4+ CD25+ Tregs from either infected 2 that show protection from diabetes, an expansion of MDA5+/ or MDA5+/+ mice limited T-cell proliferation. Tregs (CD4+, CD25+, and Foxp3+) is observed by day 7 Cells were mixed at effector T cell–to–Treg ratios ranging pi in the pancreas with a concomitant decrease in effector from 2:1 to 16:1 with titratable suppression and optimal CD4+ T cells (CD4+CD44lowCD62high) compared with results at 2:1 presented in Fig. 3E. The Tregs isolated 2 infected MDA5+/+ mice (Fig. 3A). A decrease in effector from MDA5+/ mice demonstrated a significantly greater CD4+ T cells and a significant decrease in effector CD8+ T ability to suppress IFN-g–producing CD4+ T cells than 2 cells were observed in MDA5+/ spleens compared with Tregs from infected MDA5+/+ mice (Fig. 3E). An overall infected MDA5+/+ mice (Fig. 3B and C). Further, at day 7 change in the balance of the pancreatic infiltrating effec- 2 2 pi, CB4-challenged MDA5+/ mice harbor greater num- tor and regulatory T cells was observed in MDA5+/ mice, bers of CD4+ T cells secreting IFN-g in their spleens and the increased suppressive function of regulatory 2190 IFIH1 Expression and Autoimmune Diabetes Diabetes Volume 64, June 2015

Figure 3—Reduced MDA5 expression polarizes a regulatory T-cell response. MDA5+/2 mice have increased levels of regulatory T cells (Foxp3+ of CD4+ T cells) in the PLNs (A), decreased effector CD4+ and CD8+ (CD44highCD62low) T cells in both PLNs and spleen (B and C), and increased IFN-g–producing CD4+ T cells in the spleen compared with infected MDA5+/+ mice at day 7 post–CB4 infection (D). T cells were isolated from MDA5+/+ (n = 5) and MDA5+/2 (n = 5) mice at day 7 pi and with classical activation and maturation marker antibodies for FACS analysis. Results are shown as means 6 SEM of a representative from three independent experiments. P values were determined using Student two-tailed paired t test. **P < 0.01, *P < 0.05. E: Tregs from MDA5+/2 mice suppress IFN-g production from MDA5+/+ CD4+ T cells in vitro. CD4+ CD252 T cells were isolated from spleens of uninfected MDA5+/+ mice and mixed at a ratio of 2:1 (CD4+ CD252 T cells: +/2 +/+ + + Tregs) with MDA5 or MDA5 CD4 CD25 Tregs isolated from CB4-infected mice at day 7 pi as described in RESEARCH DESIGN AND + 2 METHODS. Cells were mixed at CD4 CD25 T cells:Treg ratios ranging from 2:1 to 16:1 with titratable suppression, and optimal results at the 2:1 ratio are presented here. After 72 h, cells were restimulated, and intracellular staining determined the percent of IFN-g–producing CD4+ T cells. One of two independent experiments that yielded similar results is shown. hi, high.

2 2 2 T cells generated by day 7 pi in these mice likely leads to MDA5+/ (data not shown), and MDA5 / (data not the observed diabetes resistance. shown) recipients 24 h prior to CB4 challenge. After 7 days pi, we observed a significant increase in the percent- CD11b+ CD11c+ Cells From MDA5+/2 Mice Induce age of Tregs in the PLNs of infected MDA5+/+ recipient 2 Regulatory T Cells mice that received MDA5+/ CD11b+ CD11c+ cells com- Polarization of T cells suggested that MDA5 was acting pared with infected MDA5+/+ recipients that did not re- within APCs, and for confirmation of this, APCs were ceive cells prior to infection (no cells [Fig. 4A]). APCs isolated, subdivided (CD11b+ CD11c+ and CD11b+ transferred from MDA5+/+ donors did not induce changes 2 2 CD11c ) from either MDA5+/ or MDA5+/+ donor in Treg levels in recipient mice but, rather, boosted CD4+ spleens, and adoptively transferred into MDA5+/+ (Fig. 4), effector (CD44high CD62Llow) T cells in the PLN and diabetes.diabetesjournals.org Lincez, Shanina, and Horwitz 2191

Figure 4—CD11b+CD11c+ cells from MDA5+/2 mice polarize a regulatory T-cell response. APCs (CD11b+ CD11c+ or CD11b+ CD11c2) isolated from MDA5+/2 or MDA5+/+ spleens were adoptively transferred to MDA5+/+ (n =10–15), MDA5+/2 (not shown, n =10–15), and MDA52/2 (not shown, n =10–15) recipients. After 24 h, recipient mice including age-matched controls that did not receive cells (no cells) were infected with 400 pfu CB4. A: After 7 days pi, lymphocytes from spleen and PLN of infected mice were stained for FACS analysis. B: IFN-I concentrations in serum from recipient mice were measured by ELISA as described in RESEARCH DESIGN AND METHODS. Data shown are means 6 SEM of pooled samples from three independent experiments. P values were determined using Student two-tailed paired t test. *P < 0.05.

spleen of MDA5+/+ recipients by day 7 pi (Supplementary an essential regulator of the diabetogenic T-cell response. Fig. 3). Both IFN-a and -b were at significantly higher We have observed that partial loss of MDA5 expression detectable levels in the serum of recipient mice (MDA5+/+) creates a unique IFN-I signature, in which a burst of IFN-I that received CD11b+ CD11c+ cells from MDA5+/+ com- is induced early pi, likely to help clear viral infection, 2 pared with MDA5+/ donors (Fig. 4B). The induction of followed by a drop in IFN-I as the infection is cleared and Tregs and decreased expression of IFN-I in recipients likely prevents the triggering of autoimmunity. MDA5 (MDA5+/+) after transfer of CD11b+ CD11c+ cells from signaling of IFN-I acts in partnership over the course of 2 MDA5+/ donors versus MDA5+/+ donors strongly suggest infection with another RNA sensor, TLR3, to develop that the cells responsible for the sustained expression of auniquepathogen-specific signature. In the case of the 2 this unique IFN-I signature in MDA5+/ mice are CD11b+ pancreatropic virus EMCV, the loss of one MDA5 allele, CD11c+ cells. though on the T1D resistant C57BL/6 mouse background, resulted in transient hyperglycemia due to direct killing of DISCUSSION pancreatic b-cells and did not protect from EMCV-induced Our findings demonstrate that control of the interferon diabetes (10). Herein, our work demonstrates the impor- signature pathway after an environmental insult regulates tance of MDA5 signaling kinetics pi, as the partial loss of a critical balance between effector and regulatory T cells, MDA5 expression in diabetes-susceptible NOD mice pro- thereby influencing disease. In T1D, MDA5 sensing acts as tects from the establishment of T1D through a change in 2192 IFIH1 Expression and Autoimmune Diabetes Diabetes Volume 64, June 2015 the polarization of the T-cell response to increase regula- tion of the autoimmune response. The host response is Acknowledgments. The authors are grateful to Dr. M. Colonna, of The not significantly diminished in its ability to clear virus Washington University School of Medicine, Department of Pathology and Im- 2 2 infection, as CB4 does not replicate out of control and munology, for the generous gift of C57BL/6 MDA5 / mice that the authors directly kill the pancreatic b-cells. Rather, CB4 mimics backcrossed onto the NOD background. The authors thank The University of clinical diabetes onset with the presentation of self- British Columbia flow cytometry facility, especially Justin Wong and Andy Johnson, antigen through resident CB4-infected cells in the pan- for sorting and Alissa Cait and Adam Plumb for their advice and technical expertise. creas (11), and altering MDA5 sensing simply regulates Duality of Interest. No potential conflicts of interest relevant to this article the autoreactive component of the host response to were reported. infection. Author Contributions. P.J.L. designed and performed the experiments Specifically, we observed that a reduction in MDA5 fi and wrote the manuscript. I.S. performed the experiments. M.S.H. designed the alters IFN-I signaling in a tissue- and cell-speci c manner experiments and wrote the manuscript. M.S.H. is the guarantor of this work and, that allows for Tregs in the PLN at the site of auto- as such, had full access to all the data in the study and takes responsibility for + immunity. Further, we have observed that CD11b the integrity of the data and the accuracy of the data analysis. 2 CD11c+ cells from MDA5+/ mice (and not from +/+ MDA5 donors) induce Tregs and maintain lower levels References + + of IFN-I pi, suggesting that CD11b CD11c cells are re- 1. Eisenbarth GS. Type I diabetes mellitus. A chronic autoimmune disease. N sponsible for driving the unique IFN-I signature observed Engl J Med 1986;314:1360–1368 +/2 in MDA5 mice that protects from T1D. Early interven- 2. Nejentsev S, Walker N, Riches D, Egholm M, Todd JA. Rare variants of tion of plasmacytoid dendritic cell (pDC) function has IFIH1, a gene implicated in antiviral responses, protect against type 1 diabetes. recently been reported to prevent autoimmune patholo- Science 2009;324:387–389 gies in an autoimmune lupus model (18). Further, chil- 3. Downes K, Pekalski M, Angus KL, et al. Reduced expression of IFIH1 is dren at risk for T1D have shown a unique IFN-I protective for type 1 diabetes. PLoS ONE 2010;5:e12646 transcriptional signature that precedes islet autoimmu- 4. Smyth DJ, Cooper JD, Bailey R, et al. A genome-wide association study of nity, recent onset of T1D has been strongly associated nonsynonymous SNPs identifies a type 1 diabetes locus in the interferon-induced – with infection by RNA viruses like coxsackievirus, and helicase (IFIH1) region. Nat Genet 2006;38:617 619 an increase in pDCs and IFN-a production, especially af- 5. Ferreira RC, Guo H, Coulson RM, et al. A type I interferon transcriptional signature precedes autoimmunity in children genetically at risk for type 1 di- ter CB4 infection, has been observed in recent-onset T1D abetes. Diabetes 2014;63:2538–2550 patients (5,6,19–21). These studies and our work suggest – 6. Dotta F, Sebastiani G. Enteroviral infections and development of type 1 an important role for natural IFN-I producing cells like diabetes: The Brothers Karamazov within the CVBs. Diabetes 2014;63:384– pDCs and their contribution, in response to environmen- 386 tal T1D-associated IFN-I stimulators like CB4, to an early 7. Richardson SJ, Horwitz MS. Is type 1 diabetes “going viral”? Diabetes IFN-I signature in the development of autoimmunity. We 2014;63:2203–2205 have shown here that a reduction in MDA5 creates 8. Reikine S, Nguyen JB, Modis Y. Pattern Recognition and Signaling Mech- a unique IFN-I signature and leads to development of anisms of RIG-I and MDA5. Front Immunol 2014;5:342 a regulatory rather than an effector T-cell response 9. Funabiki M, Kato H, Miyachi Y, et al. Autoimmune disorders associ- after infection with a T1D-inducing RNA virus in a ated with gain of function of the intracellular sensor MDA5. Immunity – T1D-susceptible model. 2014;40:199 212 This work identifies MDA5 as an important target 10. McCartney SA, Vermi W, Lonardi S, et al. RNA sensor-induced type I IFN prevents diabetes caused by a b cell-tropic virus in mice. J Clin Invest 2011;121: for preventative and therapeutic strategies to halt T1D. 1497–1507 Our demonstration that a reduction in MDA5 signaling 11. Horwitz MS, Bradley LM, Harbertson J, Krahl T, Lee J, Sarvetnick N. Di- induces regulatory T-cell levels at the site of autoimmunity abetes induced by Coxsackie virus: initiation by bystander damage and not suggests a potential protective mechanism. Ex vivo expan- molecular mimicry. Nat Med 1998;4:781–785 sion and transplantation of regulatory T cells to protect 12. Serreze DV, Ottendorfer EW, Ellis TM, Gauntt CJ, Atkinson MA. Accel- b-cells from destruction have successfully prevented and eration of type 1 diabetes by a coxsackievirus infection requires a preex- reversed T1D in NOD mice and have shown great promise isting critical mass of autoreactive T-cells in pancreatic islets. Diabetes as a clinical intervention in children with recent onset of 2000;49:708–711 T1D (22–24), though current therapies aimed at expanding 13. Kondrashova A, Hyöty H. Role of viruses and other microbes in the path- regulatory T cells to treat autoimmunity have been unsuc- ogenesis of type 1 diabetes. Int Rev Immunol 2014;33:284–295 cessful in maintaining protection over time (22,25). Here, 14. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc 2008;3:1101–1108 by reducing MDA5 gene expression we have uncovered 15. González-Navajas JM, Lee J, David M, Raz E. Immunomodulatory functions a unique IFN-I signature in NOD mice that has led to the of type I interferons. Nat Rev Immunol 2012;12:125–135 expansion of regulatory T cells at the site of autoimmu- 16. Guerder S, Joncker N, Mahiddine K, Serre L. Dendritic cells in tolerance and nity and to the protection from T1D, suggesting a new autoimmune diabetes. Curr Opin Immunol 2013;25:670–675 avenue for T1D therapy, in targeting MDA5, that con- 17. Jeker LT, Bour-Jordan H, Bluestone JA. Breakdown in peripheral tolerance siders both genetic and environmental factors known to in type 1 diabetes in mice and humans. Cold Spring Harb Perspect Med 2012;2: alter disease pathogenesis. a007807 diabetes.diabetesjournals.org Lincez, Shanina, and Horwitz 2193

18. Rowland SL, Riggs JM, Gilfillan S, et al. Early, transient depletion of plas- pDCs augment Th1 responses through IFN-a production. J Immunol 2014;193: macytoid dendritic cells ameliorates autoimmunity in a lupus model. J Exp Med 1024–1034 2014;211:1977–1991 22. Marek-Trzonkowska N, Myśliwiec M, Dobyszuk A, et al. Therapy of type 1 19. Dotta F, Censini S, van Halteren AG, et al. Coxsackie B4 virus infection of diabetes with CD4(+)CD25(high)CD127-regulatory T cells prolongs survival of beta cells and natural killer cell insulitis in recent-onset type 1 diabetic patients. pancreatic islets - results of one year follow-up. Clin Immunol 2014;153:23–30 Proc Natl Acad Sci USA 2007;104:5115–5120 23. Bluestone JA, Tang Q. Therapeutic vaccination using CD4+CD25+ antigen- 20. Yeung WC, Rawlinson WD, Craig ME. Enterovirus infection and type 1 di- specific regulatory T cells. Proc Natl Acad Sci USA 2004;101(Suppl. 2):14622–14626 abetes mellitus: systematic review and meta-analysis of observational molecular 24. Tang Q, Henriksen KJ, Bi M, et al. In vitro-expanded antigen-specific reg- studies. BMJ 2011;342:d35 ulatory T cells suppress autoimmune diabetes. J Exp Med 2004;199:1455–1465 21. Xia CQ, Peng R, Chernatynskaya AV, et al. Increased IFN-a-producing 25. Smilek DE, Ehlers MR, Nepom GT. Restoring the balance: immunothera- plasmacytoid dendritic cells (pDCs) in human Th1-mediated type 1 diabetes: peutic combinations for autoimmune disease. Dis Model Mech 2014;7:503–513