inflammation-induced citrullinated glucose-regulated protein ......glucose-regulated protein 78...

Inflammation-Induced Citrullinated Glucose-RegulatedProtein 78 Elicits Immune Responses in HumanType 1 DiabetesMijke Buitinga,1 Aïsha Callebaut,1 Fernanda Marques Câmara Sodré,1 Inne Crèvecoeur,1

Gabriele Blahnik-Fagan,2 Mei-Ling Yang,3 Marco Bugliani,4 David Arribas-Layton,2 Meghan Marré,5

Dana P. Cook,1 Etienne Waelkens,6,7 Roberto Mallone,8 Jon D. Piganelli,5 Piero Marchetti,4

Mark J. Mamula,3 Rita Derua,6,7 Eddie A. James,2 Chantal Mathieu,1 and Lut Overbergh1

Diabetes 2018;67:2337–2348 | https://doi.org/10.2337/db18-0295

The b-cell has become recognized as a central player inthe pathogenesis of type 1 diabetes with the generationof neoantigens as potential triggers for breaking immunetolerance. We report that posttranslationally modifiedglucose-regulated protein 78 (GRP78) is a novel auto-antigen in human type 1 diabetes. When human isletswere exposed to inflammatory stress induced by inter-leukin-1b, tumor necrosis factor-a, and interferon-g,arginine residue R510 within GRP78 was convertedinto citrulline, as evidenced by liquid chromatography-tandem mass spectrometry. This conversion, knownas citrullination, led to the generation of neoepitopes,which effectively could be presented by HLA-DRB1*04:01molecules. With the use of HLA-DRB1*04:01 tetramersand ELISA techniques, we demonstrate enhanced an-tigenicity of citrullinated GRP78 with significantly in-creased CD4+ T-cell responses and autoantibody titersin patients with type 1 diabetes compared with healthycontrol subjects. Of note, patients with type 1 diabeteshad a predominantly higher percentage of central mem-ory cells and a lower percentage of effector memorycells directed against citrullinated GRP78 comparedwith the native epitope. These results strongly suggestthat citrullination of b-cell proteins, exemplified here bythe citrullination of GRP78, contributes to loss of self-

tolerance toward b-cells in human type 1 diabetes, in-dicating that b-cells actively participate in their owndemise.

Type 1 diabetes is an autoimmune disease characterized byT-cell–mediated destruction of insulin-producing pancre-atic b-cells, leading to insulin deficiency (1). Although it iswell-established that loss of central and peripheral toler-ance toward native self-proteins leads to the escape andaccumulation of autoreactive T-cells, the exact mechanismsthat trigger this break in self-tolerance are not entirelyunderstood (2).

Growing evidence suggests that the b-cell is not a pas-sive target of autoimmunity but actively contributes to itsown destruction by triggering the immune system (2).Many of the triggers associated with type 1 diabetes, suchas inflammatory cytokines, free radicals, and viral infec-tions, result in endoplasmic reticulum (ER) or oxidativestress (3–11). These stresses can lead to modifications ofb-cell proteins, including posttranslational modifications(PTMs) (10,12–17), mRNA alternative splicing, hybridpeptide formation (18,19), and nonconventional trans-lation, potentially generating neoepitopes against which

1Laboratory for Clinical and Experimental Endocrinology, KU Leuven, Leuven,Belgium2Benaroya Research Institute, Seattle, WA3Yale University School of Medicine, New Haven, CT4Pisa University, Pisa, Italy5Division of Pediatric Surgery, University of Pittsburgh, Pittsburgh, PA6Laboratory of Protein Phosphorylation and Proteomics, KU Leuven, Leuven,Belgium7SyBioMa, KU Leuven, Leuven, Belgium8INSERM, U1016, CNRS, UMR8104, Paris Descartes University, Sorbonne ParisCité, Cochin Institute, Paris, France

Corresponding author: Lut Overbergh, [email protected].

Received 9 March 2018 and accepted 31 July 2018.

This article contains Supplementary Data online at http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1.

M.Bui., A.C., and F.M.C.S. contributed equally as first authors.

C.M. and L.O. contributed equally as senior authors.

© 2018 by the American Diabetes Association. Readers may use this article aslong as the work is properly cited, the use is educational and not for profit, and thework is not altered. More information is available at http://www.diabetesjournals.org/content/license.

Diabetes Volume 67, November 2018 2337

IMMUNOLOGY

AND

TRANSPLANTATIO

N

https://doi.org/10.2337/db18-0295http://crossmark.crossref.org/dialog/?doi=10.2337/db18-0295&domain=pdf&date_stamp=2018-10-04mailto:[email protected]://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://www.diabetesjournals.org/content/licensehttp://www.diabetesjournals.org/content/license

no tolerance exists in the immune system (20). In severalautoimmune diseases, such as rheumatoid arthritis, mul-tiple sclerosis, systemic lupus erythematosus, and celiac dis-ease, PTM proteins are established autoantigens (21,22),but their relevance in type 1 diabetes is only starting to beexplored.

One PTM that received recent attention in type 1 di-abetes is citrullination, which involves the deiminationof arginine to citrulline by peptidyl arginine deiminase(PAD) enzymes (15,17,23). Our group previously demon-strated that the ER chaperone glucose-regulated protein78 (GRP78), classically known as an important regulatorof the unfolded protein response (21), is citrullinated inb-cell lines when exposed to inflammatory stress (17). Toour knowledge, we provide here the first direct evidence bytwo-dimensional difference gel electrophoresis (2D-DIGE)and liquid chromatography-tandem mass spectrometry(LC-MS/MS) that GRP78 is citrullinated in human isletsafter cytokine exposure. Together with the identificationof higher CD4+ T-cell frequencies directed against citrulli-nated GRP78 (citGRP78) epitopes and significantly elevatedautoantibody titers in patients with type 1 diabetes, theseresults suggest that citrullination of b-cell proteins contrib-ute to loss of tolerance toward b-cells in human type 1diabetes and identify citGRP78 as a novel islet autoantigen.

RESEARCH DESIGN AND METHODS

Islet Culture and TreatmentHuman islets were obtained from the Alberta DiabetesInstitute Islet Core and Pisa University, with the agree-ment of the local ethics committee from the Universityof Alberta (Edmonton, AB, Canada) (Pro00013094;Pro00001754) and Pisa University (#2615) (24,25) (Sup-plementary Table 1). These were cultured in DMEM sup-plemented with 1% L-glutamine (Gibco), 10% FBS, and100 units/mL penicillin-streptomycin. Islets were exposedto human interleukin-1b (IL-1b) (50 units/mL; R&D Sys-tems), murine tumor necrosis factor-a (TNF-a) (1,000units/mL; R&D Systems), and human interferon-g (IFN-g)(1,000 units/mL; Peprotech) for 24 or 72 h (26), as in-dicated in the figure legends.

Human SubjectsBlood samples for plasma and peripheral blood mono-nuclear cell (PBMC) collection were obtained from indi-viduals with type 1 diabetes and healthy control subjects(Supplementary Tables 2–4). All procedures were approvedby the ethics committee UZ Leuven (S52697; Leuven,Belgium) and conducted in accordance with the principlesof the Declaration of Helsinki, with written informedconsent obtained from all participants before inclusion.

Quantitative RT-PCRQuantitative RT-PCR was performed as previously de-scribed (17). Primers used were CHOP forward: 59-GAA-CGGCTCAAGCAGGAAATC-39; CHOP reverse: 59-TTCAC-CATTCGGTCAATCAGAG-39; GRP78: Hs.PT.58.22715160

(Integrated DNA Technologies); and ATF3: Hs.PT.56a.41052403.g (Integrated DNA Technologies).

Cell Death AssaysIslets were incubated with propidium iodide (Invitrogen)and Hoechst 33342 (Invitrogen) for 15 min at 37°C.Imaging was performed with a Zeiss Colibri LED micro-scope. At least 15 islets were analyzed by two independentresearchers (one blinded to sample identity).

2D-DIGEAnalysis with 2D-DIGE was performed with protein lysatesfrom 1,000 islets/condition, as previously described (5).

2D Western BlottingThe 2D gels were blotted onto a polyvinylidene fluoridemembrane (GE Healthcare) and probed with anti-GRP78antibody (Santa Cruz) and goat anti-rabbit horseradishperoxidase (Dako) as previously described (17).

Orbitrap LC-MS/MS AnalysisHuman islets were lysed in lysis buffer (7mol/L urea, 2mol/Lthiourea, 4% weight for volume 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate, 40 mmol/L Trisbase, 1% weight for volume dithiothreitol [DTT], anda mixture of protease inhibitors [cOmplete protease in-hibitor; Roche Diagnostics]). Cell lysates (10 mg) werereduced in 5 mmol/L DTT for 30 min at 37°C followed byalkylation with 25 mmol/L iodoacetamide and quenchingof excess iodoacetamide with 25 mmol/L DTT, both byincubating for 30 min at 37°C in the dark. Reduced andalkylated cell lysates were protein precipitated using theWessel-Flügge method (27) and were digested with 1 mgmodified trypsin (Pierce) or 1.5 mg Glu-C (Promega) atpH 8 in the presence of 5% acetonitrile and 0.01%ProteaseMAX (Promega), subjected to desalting with C18ZipTip pipette tips (Millipore) (28), and analyzed by un-biased LC-MS/MS or targeted data-dependent acquisitionwith inclusion list on a hybrid quadrupole-Orbitrap massspectrometer (Q Exactive; Thermo Fisher Scientific). Pep-tides were identified by Mascot (Matrix Science) usingSwissProt (Homo sapiens, 169,779 entries) as a databasethrough Proteome Discoverer 2.2, incorporating Percolatorfor peptide validation. Oxidation (M), deamidation (N/Q),and deamidation (R) (referring to citrullination) were in-cluded as variable modifications, and carbamidomethyla-tion (C) was included as a fixed modification. Two missedcleavages were allowed for both trypsin and Glu-C di-gestion. Peptide tolerance was set at 15 parts per millionand MS/MS tolerance at 20 millimass units. Only peptidesdisplaying high and medium confidence (posterior errorprobabilities # 0.01 and 0.05) were retained.

Because citrullination and deamidation result in a massincrement of 0.98402 Da, whereas a 13C isotope of a nativepeptide results in a mass increment of 1.003355 Da, wemanually checked differences in MH+ of native versusmodified peptide. Moreover, we checked differences inretention time between modified and native peptides,

2338 Immunogenicity of citGRP78 in Type 1 Diabetes Diabetes Volume 67, November 2018

http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1

as previously described (29). To discriminate betweencitrullination and deamidation, we manually checked theMS/MS spectra of the putative citrullinated peptides,inspecting the isotopic distribution of peptide fragmentscontaining N, Q, and R. Synthetic peptides (Synpeptide)IDVNGILRVTAE, IDVDGILRVTAE, and IDVNGILXVTAEwere used to distinguish between citrullinated and dea-midated peptides, using LC-MS/MS fragmentation dataand retention time as discriminating criteria.

Peptide-HLA Binding Prediction and Affinity AssaysNative and citGRP78 epitope candidates with predictedbinding affinity for HLA-DRB1*04:01 were identified usingthe in silico approach outlined in previous work (30) andsynthesized by Synpeptide. Binding affinity to HLA-DRB1*04:01 was measured by incubating increasing con-centrations of peptides in competition with 0.02 mmol/Lbiotinylated influenza hemagglutinin (HA) 306–318 inwells coated with DRB1*04:01 protein. After washing,residual biotin-HA 306–318 was detected using europium-conjugated streptavidin and quantified using a Victor2Dtime-resolved fluorometer (PerkinElmer). IC50 valueswere calculated as the concentration needed to displace50% of the biotin-HA 306–318 peptide.

HLA-DRB1*04:01 Protein and Tetramer ReagentsRecombinant HLA-DRB1*04:01 was purified from insectcell culture supernatants by affinity chromatography anddialyzed against phosphate buffer (pH 6.0). To preparetetramers (Tmrs), DRB1*04:01 was biotinylated in vitroand incubated with 0.2 mg/mL peptide at 37°C in 0.2%n-octyl-b-D-glucopyranoside and 1 mmol/L Pefabloc SC.Monomers were conjugated into Tmrs using R-phycoerythrin(PE) streptavidin at a molar ratio of 8:1.

Isolation of PBMCsPBMCs from HLA-DRB1*04:01+ donors (SupplementaryTables 2 and 3) were isolated from heparinized blood usingLymphoprep and frozen in AIM V Medium (Gibco) con-taining 10% DMSO.

In Vitro Tmr AssaysPBMCs were expanded in vitro as previously described(31). Briefly, 5 3 106 PBMCs were stimulated with6 mmol/L pooled citGRP78 peptides. After 14 days ofin vitro stimulation, PE-labeled Tmrs were used to stainT cells specific for the GRP78 peptides (75 min at 37°C).Analysis was performed by FACSCalibur (BD Biosciences)using FlowJo software (TreeStar). Positivity was defined asthe presence of a distinct population of CD4 bright cells ata percentage more than twofold above background (stain-ing of cells from the unstimulated well set to 0.1% for mostexperiments).

Ex Vivo Tmr AssaysCryopreserved PBMCs (20–30 3 106) were treated with50 nmol/L of dasatinib (Cell Signaling) for 7 min at 37°Cand stained with Tmrs for 2 h at room temperature (32).Cells were labeled with anti-PE and anti-allophycocyanin

magnetic beads for 20 min at 4°C. Next, 1/40th of thecells (preenriched fraction) was reserved to determinethe total cell number used in the assay. The remainingfraction was enriched using Miltenyi Biotec MS magneticcolumns (enriched fraction) according to the manufac-turer’s instructions. Both fractions were stained for15 min at 4°C with CD4-BV421 (BioLegend), CCR7-APC-Cy7 (BioLegend), CD45RA-AF700 (BD Bioscience),SYTOX Green (Thermo Fisher Scientific), and CD20/CD14fluorescein isothiocyanate (eBioscience). Samples were fullycollected on an LSR II flow cytometer (BD Bioscience) andanalyzed using FlowJo version 10.1 software. Data wereexpressed as the frequency of Tmr+ cells per 106 living CD4+

T cells and considered positive when .5 (33).

ELISAImmunoreactivity of human plasma to native and citGRP78was performed by ELISA (Supplementary Table 4). Briefly,0.125 mg of recombinant human native or citGRP78 pro-tein (constructed and produced as previously described[17]) in 0.05 mol/L carbonate-bicarbonate buffer (pH 9.6)(Sigma) was coated onto Immulon 2 HB plates (ThermoFisher Scientific) overnight at 4°C. The plate was blockedwith 1% BSA in PBS containing 0.05% Tween 20 for 2 h atroom temperature. Plasma was diluted 1:100 in PBS-Tween 20 containing 0.3% BSA and incubated for 1 h.Mouse anti-GRP78 monoclonal antibody (Santa Cruz)served as a positive control. Species-specific goat anti-human IgG (Southern Biotech) or rat anti-mouse IgMalkaline phosphatase (Santa Cruz) was added as a second-ary reagent. Color was developed with para-nitrophenyl-phosphate substrate (Sigma). All readings were normalizedto nonspecific plasma binding from blank wells (no antigenadded). In parallel assays, wells coated with PAD enzymesonly confirmed specific binding to citGRP78.

Statistical AnalysesData are shown as mean 6 SD. Significant differencesbetween experimental conditions were assessed by pairedStudent t test, Wilcoxon rank sum test, Kruskal-Wallis testwith Mann-Whitney U comparison, or MANOVA as indi-cated (GraphPad Prism version 7.02 and SPSS version 22software). P , 0.05 was considered statistically significant.

RESULTS

GRP78 Is Citrullinated in Human Islets of LangerhansUpon Inflammatory StressTo study the effect of inflammation on the modification ofGRP78, islets were exposed in vitro to IL-1b, IFN-g, andTNF-a for 24 or 72 h followed by 2D-DIGE analyses.Cytokine exposure significantly induced cell death after72 h (11.39 vs. 4.41% in control islets; P , 0.05, n = 5)(Fig. 1A). Whereas after 24 h no increased cell deathwas observed (Fig. 1A), upregulation of CHOP, GRP78,and ATF3 mRNA (P , 0.05, n = 3) suggested thatincreased ER stress preceded the induction of cell death(Fig. 1B–D).

diabetes.diabetesjournals.org Buitinga and Associates 2339

http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1

Analysis of the islet 2D-DIGE pattern in control isletsrevealed the presence of five GRP78 spots (I1–I5) differ-ing in isoelectric point but not in molecular weight (Fig.1E). This pointed to the presence of at least five GRP78isoforms of which the identity was confirmed by 2DWestern blotting (Fig. 1F). Quantitative 2D-DIGE anal-ysis of cytokine-exposed versus control islets (n = 5;donors D1–D5) (Supplementary Table 1) revealed a sig-nificant upregulation of the two most abundant GRP78isoforms (isoform I2: 1.5-fold upregulated [P , 0.05];isoform I3: 1.64-fold upregulated [P, 0.05]). In three offive islet donors (D1–D3), this upregulation was paral-leled by a shift in relative abundance toward the isoformwith the lower isoelectric point (isoform I3), which is

indicative of an increased acidic nature consistent withcitrullination or other PTMs, such as phosphorylation,deamidation, and ribosylation, among others (Fig. 1Gand H).

To identify which GRP78 amino acid residues weremodified, we investigated the human islet proteome byunbiased LC-MS/MS (n = 5; donors D6–D10) (Supplemen-tary Table 1). On the basis of our earlier observation thatGRP78 is citrullinated in INS-1E cells (17), we specificallyfocused on the identification of citrullinated arginine (R)residues. GRP78 contains 28 arginines, all potential sitesfor citrullination. Combined LC-MS/MS analysis of trypsinand Glu-C–digested islets from five islet donors, control orcytokine-exposed, resulted in 67.58% and 71.71% coverage of

Figure 1—GRP78 is PTM in human islets of Langerhans upon cytokine exposure. A: Apoptosis in human control islets and after exposure toIL-1b (50 units/mL), IFN-g (1,000 units/mL), and TNF-a (1,000 units/mL) for 24 or 72 h (n = 5). B–D: Changes in mRNA levels of ERstress markers CHOP (B), GRP78 (C ), and ATF3 (D) in human control islets and after exposure to IL-1b (50 units/mL), IFN-g (1,000 units/mL),and TNF-a (1,000 units/mL) for 24 h (n = 3). E and F: Human islets express five isoforms (I1–I5) of GRP78 that differ in isoelectric point butnot in molecular weight, as indicated by 2D-DIGE (pH 4–7) (E) and 2D Western blot (F ). G: Three-dimensional graph view of the DeCyderanalysis of the 2D-DIGE of human control islets and cytokine-exposed islets (n = 5). I1–I5 represent the five different isoforms of GRP78. H:The ratio of GRP78 isoforms I3/I2 is increased in cytokine-exposed islets from donors D1, D2, and D3, whereas they were decreased in isletsfrom donors D4 and D5. Data in A–D are mean 6 SD and analyzed by a paired Student t test. *P , 0.05. Ctr, control islets; Cyt, cytokine-exposed islets.


http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1

GRP78, respectively, including 22 of the 28 arginines (Fig.2A). However, a conversion to citrulline could be unambig-uously identified for none of these covered arginine residues.

We hypothesized that the inability to detect citrulli-nation of GRP78 could indicate that either citrullinationdoes not occur, the citrullinated arginine residues arenot covered in the current approach, or the citrullinatedarginine residues are not abundant enough to be detectedby unbiased LC-MS/MS. We therefore applied a targetedMS/MS strategy (data-dependent acquisition) specificallyfocusing on R510, the arginine residue we have previouslyidentified as being citrullinated in INS-1E upon cytokine-exposure and against which diabetic NOD mice haveelevated levels of circulating autoantibodies and autor-eactive T cells (17). Because citrullination often is difficultto distinguish from deamidation, we first analyzed a na-tive, citrullinated, and deamidated version of a syntheticGlu-C peptide harboring R510 by LC-MS/MS. We mea-sured a clear difference in retention time of the variouspeptides: 182.53 6 1.87, 209.92 6 1.91, and 189.18 6

0.66 min, respectively (Fig. 2B). Unbiased LC-MS/MSanalysis of five different human islet samples, controlor cytokine-exposed, revealed the presence of precursorions with a mass, charge, and retention time correspond-ing to those of the synthetic native and citrullinatedpeptides. By subsequent targeted LC-MS/MS, fragmen-tation data were obtained that enabled us to unambig-uously assign these precursor ions to the native andcitrullinated R510-harboring peptides and to distinguishthem from the deamidated peptide, the latter on the basisof the diagnostic b+-ion of charge/mass ratio (m/z) 612.34(Fig. 2C). A combination of correct m/z, retention times,and diagnostic fragmentation data led us to conclude thatR510 is citrullinated in human islets.

citGRP78 Peptides Can Be Presented in the Context ofHLA-DRB1*04:01On the basis of the finding that GRP78 is citrullinated incytokine-exposed human islets, we investigated whetherGRP78 citrullination could result in the generation of

Figure 2—Identification of citGRP78 residues in control and cytokine-exposed human islets using LC-MS/MS.A: Protein sequence coverageof GRP78 in control and cytokine-exposed (IL-1b [50 units/mL], IFN-g [1,000 units/mL], and TNF-a [1,000 units/mL] for 72 h) human islets.Coverage is indicated in red, arginine residues are indicated in bold, the signal peptide is indicated in italics, and the synthetic peptide used forLC-MS/MS is underlined. B: Fragmentation data, m/z values, and retention times (RTs) of synthetic IDVNGILRVTAE (native), IDVNGILXVTAE(citrullinated), and IDVDGILRVTAE (deamidated) peptides using LC-MS/MS. C: Fragmentation data, m/z values, and RTs of the native andcitrullinated peptides found in human islets using targeted LC-MS/MS (data-dependent acquisition). Detected ions are indicated in bold, andthe diagnostic ion to distinguish citrullination from deamidation is underlined.


neoepitopes recognized in patients with type 1 diabetes.To this end, we first identified GRP78 peptide sequencescontaining possible HLA-binding motifs. Because we couldnot cover the full-length GRP78 protein sequence by MSanalysis, thus possibly missing other citrullinated arginineresidues, we screened the entire GRP78 sequence forpotential HLA-DRB1*04:01 binders. We specifically fo-cused on HLA-DRB1*04:01 because binding to this allo-type has been shown to be enhanced by citrullinationat certain amino acid positions (34,35). A previously de-veloped algorithm was used to predict in silico peptidesequences with a high affinity for HLA-DRB1*04:01 intheir native or citrullinated forms. Thirteen peptides wereidentified as potential binders (Table 1). Of note, amongthese peptides were 498–512X and 500–514X citrullinatedat R510, the residue identified here by LC-MS/MS ascitrullinated in cytokine-exposed islets and previouslyidentified as an autoantigenic epitope in NOD mice (17).In vitro binding affinity assays demonstrated that 3 of the13 citrullinated peptides bound to DRB1*04:01, namely195–209X (citrullinated on R197, covered by LC-MS/MS,but not modified), 498–512X (citrullinated on R510), and500–514X (citrullinated on R510) (Table 1). Two otherpeptides (58–72R, 65–79R) bound DRB1*04:01 only intheir native form but with much lower affinity.

Circulating CD4+ T Cells Directed Against citGRP78498–512X Are Present at Higher Frequencies in PatientsWith Type 1 Diabetes Compared With Healthy SubjectsWe assessed the ability of the citrullinated peptides 195–209X, 498–512X, and 500–514X to elicit CD4+ T-cellresponses in PBMCs from patients with type 1 diabetesand healthy subjects (Supplementary Table 2) stimulatedin vitro for 14 days followed by detection with DRB1*04:01

Tmrs. Despite the low binding affinity of 498–512X and500–514X (Table 1), stable Tmrs could be generated. Allpeptides elicited positive CD4+ T-cell responses in multiplesubjects (Fig. 3). One epitope, 498–512X, was exclusivelyrecognized by CD4+ T cells from patients with type 1 di-abetes. Peptide 195–209X elicited T-cell responses in oneof six healthy subjects and in one patient with type 1 di-abetes. In two of six healthy subjects, a response againstpeptide 500–514X was observed, whereas only one patientwith type 1 diabetes was responsive (Fig. 3).

On the basis of this first evidence that citGRP78epitopes are able to elicit T-cell responses, we assessedthe frequencies of CD4+ T cells that recognize citGRP78epitopes in PBMCs from patients with new-onset type 1diabetes (,1 year, n = 4), patients with long-standing type1 diabetes ($1 year, n = 11), and healthy subjects (n = 8)by direct ex vivo Tmr assays (Fig. 4A–H) (donor char-acteristics listed in Supplementary Table 3). We focused onpeptides that elicited a similar or higher T-cell responsein patients with type 1 diabetes compared with healthysubjects on the basis of the in vitro Tmr assay results (Fig.3). The 498–512X reactive CD4+ T cells were more fre-quent in patients with type 1 diabetes than in healthysubjects, with a positive response in 7 of 15 patients (Fig.4J). Although the native variant 498–512R also eliciteda positive response in 5 of 15 patients, the frequency of498–512R-reactive CD4+ T cells was not significantlyhigher in patients with type 1 diabetes than in healthysubjects (Fig. 4I). For peptide 195–209X (Fig. 4L) and itscorresponding unmodified sequence (Fig. 4K), a positiveresponse was observed in a subset of patients, but thefrequency of Tmr+ cells was not significantly higher than inhealthy subjects. Although no significant difference wasobserved when comparing T-cell frequencies recognizing

Table 1—Motif analysis for native and citrullinated DRB1*0401 GRP78 sequences

Native GRP78 peptides citGRP78 peptides

Peptide Sequence IC50 (mmol/L)a Peptide Cit site Sequenceb IC50 (mmol/L)

a

7–20R AAMLLLLSAARAEE ND 7–20X R17 AAMLLLLSAAXAEE ND

39–53R YSCVGVFKNGRVEII ND 39–53X R49 YSCVGVFKNGXVEII ND

58–72R GNRITPSYVAFTPEG 63 58–72X R60 GNXITPSYVAFTPEG ND

65–79R YVAFTPEGERLIGDA 97 65–79X R74 YVAFTPEGEXLIGDA ND

97–110R RLIGRTWNDPSVQQ ND 97–110X R97; R101 RLIGXTWNDPSVQQ ND

172–186R VPAYFNDAQRQATKD ND 172–186X R181 VPAYFNDAQXQATKD ND

195–209R VMRIINEPTAAAIAY 0.04 195–209X R197 VMXIINEPTAAAIAY 0.20

292–305R VEKAKRALSSQHQA ND 292–305X R297 VEKAKXALSSQHQA ND

434–448R TKLIPRNTVVPTKKS ND 434–448X R439 TKLIPXNTVVPTKKS ND

498–512R EVTFEIDVNGILRVT ND 498–512X R510 EVTFEIDVNGILXVT 2.71

500–514R TFEIDVNGILRVTAE ND 500–514X R510 TFEIDVNGILXVTAE 4.89

523–536R KITITNDQNRLTPE ND 523–536X R532 KITITNDQNXLTPE ND

554–566R KLKERIDTRNELE ND 554–566X R558; R562 KLKERIDTXNELE ND

Cit, citrullinated; ND, not detectable. aIC50: the peptide concentration that displaces one-half of the reference peptide (biotinylatedHA306–318 peptide). A lower IC50 indicates better binding.

bThe residue that is modulated is shown in boldface in each sequence.X indicates citrulline.


http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1

the native and citrullinated form of each peptide in eachindividual (Fig. 4M–P), a trend toward higher frequenciesof Tmr+ T-cells directed against 498–512X was observed(P = 0.058) (Fig. 4N).

In samples from patients with type 1 diabetes with.10Tmr+ cells, we further examined the activation phenotypeof 498–512R- and 498–512X-experienced Tmr+ T cells onthe basis of CCR7 and CD45RA (Fig. 5A–C). In patientswith long-standing type 1 diabetes, the 498–512X Tmr+

fraction displayed a higher percentage of central memoryand a lower percentage of effector memory cells comparedwith the 498–512R Tmr+ fraction (Fig. 5D and E). Nodifferences were observed in the frequency of naive T cellsspecific for 498–512R and 498–512X (Fig. 5F).

Native and citGRP78 Are Targets for Autoantibodies inPatients With Type 1 DiabetesNext, we assessed autoantibody reactivity againstcitGRP78 and its native counterpart in plasma of patientswith new-onset type 1 diabetes (n = 47), long-standingtype 1 diabetes (n = 61), and of healthy subjects (n = 89)(Fig. 6A and Supplementary Table 4). With 33% of thepatients with long-standing type 1 diabetes being autoan-tibody positive, these patients had significantly higheranti-citGRP78 autoantibody titers than healthy subjects(Fig. 6A and B). When comparing the reactivity againstnative and citGRP78 of the same subjects, patients withnew-onset and long-standing type 1 diabetes had signifi-cantly higher titers of anti-citGRP78 autoantibodies. Of note,4 of 5 citGRP78 autoantibody-positive patients with new-onset type 1 diabetes and 15 of 20 citGRP78 autoantibody-positive patients with long-standing type 1 diabetes alsoshowed reactivity against the native epitope (Fig. 6C–E).

Autoantibodies against native and citGRP78 primarilywere present in patients with multiple autoantibodies against

insulin, GADA, and/or IA-2 antigen (Fig. 6F and G). Whencomparing autoantibody positivity for native and/orcitGRP78 in relation to diabetes duration, autoantibodiesagainst native and/or citGRP78 most often were detectedin patients diagnosed within 1–5 years (Fig. 6H).

DISCUSSION

We report on the citrullinated ER chaperone GRP78 as a novelautoantigen in human type 1 diabetes.We demonstrate thatcitrullination of GRP78 occurs in human islets exposed toinflammatory stress. With some citGRP78 peptides display-ing strong binding to HLA DRB1*04:01, we identified thepresence of CD4+ T cells and circulating autoantibodies reactiveagainst citGRP78 in a subset of patients with type 1 diabetes.We thus provide the first direct evidence in our knowledge thatb-cell proteins can be citrullinated and that islet cell–derivedGRP78 epitopes may be involved in the pathogenesis ofhuman type 1 diabetes through the formation of citrullinatedneoantigens that evoke CD4+ T-cell and B-cell responses.

In recent years, PTMs of b-cell proteins have emergedas a potential source for neoantigens in type 1 diabetes,with many studies focusing on their role in inducing T-cellresponses (12–16,18,19,36). Here, we provide a morecomprehensive insight by covering not only the interactionbetween the modified proteins and the immune systembut also the role of inflammation in inducing these mod-ifications. Given that the PTM citrullination is closelyassociated with inflammatory conditions and that inflam-mation plays a pivotal role in the induction, amplification,and maintenance of type 1 diabetes (2), this modificationalso may be important in this disease (37). The observa-tion that T-cell responses are detected against citrulli-nated peptides derived from GAD65 (15), islet amyloidpolypeptide (14), and now GRP78 is an indication that

Figure 3—Analysis of CD4+ T-cell responses to citGRP78 peptides using in vitro Tmr assays. To identify immunogenic citrullinated peptidesderived fromGRP78, PBMCs from six patients with type 1 diabetes (T1D) and six healthy control (HC) subjects were expanded for 2 weeks bystimulating with peptide and then stained with the corresponding DRB1*04:01 Tmr. The representative FACS plots depict Tmr+ CD4+ T cells(percentages shown in each upper-right quadrant) directed against the citGRP78 peptide 195–209X (A and B), 498–512X (C and D), or 500–514X (E and F ) in HC subjects (A, C, and E ) and in patients with T1D (B, D, and F ). With use of a previously established cutoff, Tmr staining oftwice the mean value of the negative (Neg.) control (Tmr staining of unstimulated T cells) (G) was considered to be a positive response (H).


http://diabetes.diabetesjournals.org/lookup/suppl/doi:10.2337/db18-0295/-/DC1

citrullination indeed plays a role in human type 1 diabetes.Of note, we also show for the first time in our knowledgethat b-cell proteins can be citrullinated when human isletsare exposed to inflammatory cytokines (IL-1b, IFN-g, andTNF-a), as exemplified by the citrullination of GRP78. Thisfinding reinforces the concept that inflammation modu-lates the cross-talk between the b-cell and the immunesystem not only by mediating b-cell dysfunction and death(3) but also by inducing b-cell neoantigen formation.Recent data on islet-reactive CD8+ T cells have suggestedthat the autoimmune repertoire between patients withtype 1 diabetes and healthy control subjects is largelyoverlapping (38) and that the critical ingredient leadingto the priming of this repertoire may be the availability ofits target islet antigens in the inflammatory milieu ofinsulitis. In light of the evidence that GRP78 citrullinationis induced by proinflammatory cytokines, we hypothesizethat this paradigm also may apply to citGRP78.

We have used a combination of IL-1b, IFN-g, and TNF-a,where TNF-a may be crucial to evoke cytokine-inducedPTMs in human islets, because the combination of IL-1b

and IFN-g alone does not induce GRP78 modifications inhuman islets (39). Of importance, PTMs of GRP78 were notinduced in all islet donors investigated; in two islet prep-arations, the levels of modified GRP78 were already elevatedin control islets and could not be enhanced further bycytokine exposure. This may be explained by higher basallevels of stress that are intrinsically associated with theislet isolation procedure, such as a more harsh enzymaticdigestion (40) or a longer cold ischemia time (41). On theother hand, it may also reflect the inherent heterogeneity inhumans, with not every individual being prone to citrulli-nation even when exposed to inflammation. Whether thedifference in induction of GRP78 PTMs between isletdonors is due to variations in genetic background, possiblyleading to abnormal expression levels of PAD enzymes, orto differences in susceptibility to environmental triggers,which may result in increased Ca2+ fluxes necessary for theactivation of PAD enzymes, remains to be elucidated.

With the use of MS, we identified R510 as a citrullinationsite in human GRP78. The 2D-DIGE profile showed, how-ever, at least five different isoforms, suggesting modifications

Figure 4—Identification of CD4+ T-cell frequencies directed against native and citGRP78 peptides in healthy control (HC) subjects andpatients with type 1 diabetes (T1D) analyzed by ex vivo Tmr assays. A–H: In HC subjects and patients with T1D, respectively, representativeFACS plots depict CD4+ T cells in the enriched fraction directed against the native GRP78 peptide 498–512R (A andB), against the citGRP78peptide 498–512X (C andD), against the native GRP78 peptide 195–209R (E and F ), and against the citGRP78 peptide 195–209X (G andH).I–L: Total CD4+ T-cell frequencies in the enriched fraction directed against 498–512R (I), 498–512X (J), 195–209R (K ), and 195–209X (L) inHC subjects (n = 8;■) and in patients with T1D (n = 15) subdivided by new-onset T1D (,1 year; n = 4;△) and long-standing T1D ($1 year;n = 11; ●). CD4+ T-cell frequencies .5 Tmr+ cells/106 (dashed line) are considered positive (indicated in red). M–P: Total CD4+ T-cellfrequencies in the enriched fraction comparing the native and citrullinated forms of peptide 498–512 in HC subjects (M ) and patients in T1D(N ) and of peptide 195–209 in HC subjects (O) and patients with T1D (P). Data in I–L are mean6 SD and were analyzed by Mann-WhitneyU test. *P , 0.05. M–P data were analyzed using the Wilcoxon rank sum test. APC, allophycocyanin.


on several residues. With a sequence coverage of 68–72%identifying 22 of the 28 arginines within GRP78, wepotentially missed other citrullination sites. The use ofalternative digestion enzymes, such as Lys-C or elastase, orthe use of a more sensitive targeted approach as performedhere for the R510 region, may result in a more completepicture of GRP78 citrullination. Next to this, other PTMs,such as deamidation, also may result in the multiple iso-form spots observed in the 2D-DIGE (42).

Alterations in peptide epitopes are known to affectimmune recognition by modulating HLA binding or influ-encing T-cell recognition (34,35,43). Of the 13 citGRP78peptides predicted to bind HLA-DRB1*04:01, 2 witha R510 citrullination site (498–512X and 500–514X) wereconfirmed to be preferential binders compared with theirnative counterparts. This observation supports the notionthat some HLA allotypes, such as HLA-DRB1*04:01, prefercitrulline over arginine at key peptide-binding positions(34,35). This may reflect either a direct interaction ofcitrulline with HLA anchor positions or an indirect effectmediated by a change in peptide orientation within thebinding groove. The binding motif of HLA-DRB1*04:01displays a strong preference for aliphatic and aromatic res-idues in the first binding pocket. According to this motif,

the canonical nonamer motif of epitope 498–512X mostlikely to bind to HLA-DRB1*04:01 would be FEIDVNGILbecause it contains an optimal F residue at P1, a highlyfavorable D at P4, and a preferred N at P6 (44–46). Thisplaces citrulline in position 10, which according to Yassaiet al. (47) can modulate peptide-binding affinity and in-duce an alternative downward configuration of the peptidein the binding groove, thus shifting the T-cell receptorcontact positions.

Although the binding affinity of peptide 195–209 wasimproved upon citrullination, the frequency of Tmr+ T cellsdirected against this epitope was not significantly in-creased in patients with type 1 diabetes compared withhealthy subjects. Together with our observation that R197is not citrullinated in cytokine-treated human islets (Fig.2A), these results may indicate that citrullination of R197does not readily occur in human type 1 diabetes.

R510 is exactly the same arginine residue as previouslyfound to be citrullinated in INS-1E cells upon cytokineexposure and against which diabetic NOD mice have ele-vated levels of circulating autoantibodies and autoreactiveT cells (17). Using ex vivo Tmr assays and ELISAs, we nowdemonstrate elevated T-cell and autoantibody responsesagainst 498–512X in patients with type 1 diabetes, which

Figure 5—Phenotypic characterization of 498–512R- and 498–512X-experienced CD4+ T cells in patients with type 1 diabetes. A–C: CCR7and CD45RA were used to discriminate between antigen-experienced central memory CD4+ T cells (TCM) (CD45RA2CCR7+), effectormemory CD4+ T cells (TEM) (CD45RA2CCR72), and naive CD4+ T cells (CD45RA+CCR7+). Gating was performed on the CD4+ T cells of thepreenriched fraction (A) and imposed on the CD4+ Tmr+ cells against 498–512R (B) and 498–512X (C). Frequency of TCM (D), TEM (E), andnaive (F) CD4+ T-cells in patientswith long-standing type 1diabetes (n=6). Data inD–Fwere analyzed using theWilcoxon rank sum test. *P,0.05.


Figure 6—Autoantibody responses directed against native and citGRP78 in healthy control (HC) subjects and patients with type 1 diabetes(T1D). A and B: Titers of anti-GRP78 and anti-citGRP78 autoantibodies (A) and frequencies of anti-GRP78 and anti-citGRP78 autoantibodypositivity (B) in plasma of HC subjects (n = 89;■), patients with new-onset T1D (nT1D) (,1 year; n = 47;△), and patients with long-standingT1D ($1 year; n = 61;●). Titers are considered positive (indicated in red) when the optical density (OD) is higher than the mean + 2 SDs of thecontrol group.C–E: Comparison of anti-GRP78 and anti-citGRP78 autoantibody titers in HC subjects (C ), patients with nT1D (D), and patientswith T1D (E). F and G: Anti-GRP78 and anti-citGRP78 autoantibody positivity in patients with nT1D (F ) and long-standing T1D (G) who arepositive for none, one, or multiple autoantibodies (autoAbs) for insulin, GADA, and/or IA-2 antigen. H: Anti-GRP78 and anti-citGRP78autoantibody positivity plotted against diabetes duration. Data were analyzed using MANOVA with Bonferroni-corrected post hoccomparisons within (native vs. citrullinated) and between (HC vs. nT1D vs. T1D) subjects. *P , 0.05, **P , 0.01, ***P , 0.001.


strongly indicates that the same epitope as in NOD mice isrecognized in human type 1 diabetes. The observation thatthere are patients with CD4+ T-cells and autoantibodyresponses to the native, the citrullinated, or both epitopesmay suggest that epitope spreading occurs (48). There isgreat interest in knowing to which epitope the immunesystem initially responds, following the hypothesis that thisepitope may behave as a disease driver (49) and may bea potential target for antigen-specific treatments (48).

That B- and T-cell responses were more readily observedin patients with long-standing ($1 year) type 1 diabetesthan in patients with new-onset type 1 diabetes under-scores our hypothesis that citGRP78, rather than beinga primary autoantigen, may be involved in the amplifica-tion of islet autoimmunity once inflammation is alreadyongoing. The activation of PAD enzymes requires high levelsof cytosolic Ca2+ (11), which is favored by ER dysfunction incombinationwith excessive insulin demand. A similar processhas been described for the Ca2+-dependent PTM enzymetissue transglutaminase 2 in which activity is increased byER stress, leading to b-cell stress-dependent immunogenicity(10). In this scenario, our observation that GRP78 is trans-located to the plasma membrane upon cytokine exposure inrodentb-cells (17) and human islets (S. Vig, L.O., unpublishedobservations) is noteworthy. Although the mechanisms havenot yet been elucidated, the combination of membrane trans-location and increased cytosolic Ca2+ concentrations may bea prerequisite for GRP78 citrullination to occur.

Similar to other autoantigens (12,15), we show thatboth B- and T-cell responses to citGRP78 are not homo-geneous across the patient population, which may reflectthe heterogeneity of the disease or the sequestration ofautoimmune T cells in diseased organs (38). This hetero-geneity suggests that it is unlikely that a single marker canpredict disease onset or progression, emphasizing the im-portance of identifying other autoantigens and to mapT-cell responses and autoantibody generation in at-riskindividuals and patients with type 1 diabetes during dis-ease initiation and progression.

In conclusion, this study reinforces the notion thatb-cells are actively involved in their own destruction bygenerating citrullinated b-cell protein epitopes when ex-posed to inflammatory stress, as shown here for GRP78. Wedemonstrate that one of these citGRP78 epitopes is able toelicit CD4+ T-cell and autoantibody responses primarily inpatients with long-standing type 1 diabetes. These resultsmay suggest that immune responses against citGRP78contribute to the acceleration of disease rather than tothe precipitation of its development. On the other hand, itmay also be possible that GRP78-specific immune cells weremainly sequestered in the pancreas and pancreatic lymphnodes until after disease onset, after which they emergedin the periphery. Our findings further suggest that epitopespreading occurs, with detectable T- and B-cell responsesdirected against both the native and citrullinated epitope.Given the heterogeneity of type 1 diabetes, the identif-ication of PTM b-cell proteins as potential neoantigens is

critical for a better understanding of disease pathology,the discovery of new markers for earlier diagnosis andpatient stratification, and the development of novel diseaseinterventions, especially in the view of more personalizedtherapies. Determination of the stage of disease duringwhich T- and B-cell reactivity toward modified and nativeself-antigens occurs will shed more light on the questionof whether responses to modified epitopes are primaryresponses and thus drivers of the disease process. Thiswill be of crucial importance for disease prediction, pre-vention, and development of antigen-specific therapies.

Acknowledgments. The authors thank Frea Coun and Dries Swinnen(KU Leuven, Leuven, Belgium) for technical assistance, Patrick MacDonald (AlbertaDiabetes Institute, University of Alberta, Edmonton, AB, Canada) for the supply ofhuman islets, Dr. Sebastien Carpentier (SyBioMa Mass Spectrometry, KU Leuven)for helpful advice, and Hilde Morobé (UZ Leuven) for patient recruitment andcollection of blood samples. The authors also thank all the patients and healthyvolunteers for participating in this study.Funding. This work was supported by the Innovative Medicines Initiative JointUndertaking under grant agreement 115797 (INNODIA). This joint undertakingreceives support from the European Union’s Horizon 2020 research and innovationprogram and the European Federation of Pharmaceutical Industries and Associ-ations, JDRF, and The Leona M. and Harry B. Helmsley Charitable Trust; KU Leuven(GOA14/010); JDRF (2-SRA-2015-52-Q-R, 1-INO-2018-638-A-N); a JDRF post-doctoral fellowship 3-PDF-2018-593-A-N (to M.Bui.); and Fonds WetenschappelijkOnderzoek (for a PhD fellowship for A.C. [1189518N] and D.P.C. [11Y6716N]).Duality of Interest. No potential conflicts of interest relevant to this articlewere reported.Author Contributions. M.Bui., A.C., F.M.C.S., and I.C. contributed to thedesign, conduct, analysis, and interpretation of the data and wrote and edited themanuscript. M.Bui., A.C., F.M.C.S., I.C., G.B.-F., M.-L.Y., M.Bug., D.A.-L., M.M.,D.P.C., E.W., R.M., J.D.P., P.M., M.J.M., R.D., E.A.J., C.M., and L.O. revised thearticle and gave their final approval of the version to be published. A.C., E.W., andR.D. conducted, analyzed, and interpreted the proteomics data. G.B.-F., D.A.-L.,D.P.C., and E.A.J. contributed to the design, analysis, and/or interpretation ofthe peptide-binding affinity and Tmr data. M.-L.Y. and M.J.M. designed andperformed the ELISA experiments as well as contributed to the analysis andinterpretation. M.Bug. and P.M. provided human islets. M.M., R.M., and J.D.P.provided intellectual input. C.M. and L.O. designed research, interpreted data, andwrote and edited the manuscript. C.M. and L.O. are the guarantors of this workand, as such, had full access to all the data in the study and take responsibilityfor the integrity of the data and the accuracy of the data analysis.Prior Presentation. Parts of this study were presented in abstract format the Immunology of Diabetes Society Congress, London, U.K., 25–29 October2018, and the Human Proteome Organization World Congress, Orlando, FL,30 September–3 October 2018.

References1. Cnop M, Welsh N, Jonas JC, Jörns A, Lenzen S, Eizirik DL. Mechanisms ofpancreatic beta-cell death in type 1 and type 2 diabetes: many differences, fewsimilarities. Diabetes 2005;54(Suppl. 2):S97–S1072. Eizirik DL, Colli ML, Ortis F. The role of inflammation in insulitis and beta-cellloss in type 1 diabetes. Nat Rev Endocrinol 2009;5:219–2263. Eizirik DL, Miani M, Cardozo AK. Signalling danger: endoplasmic reticulumstress and the unfolded protein response in pancreatic islet inflammation. Dia-betologia 2013;56:234–2414. Gurzov EN, Eizirik DL. Bcl-2 proteins in diabetes: mitochondrial pathways ofb-cell death and dysfunction. Trends Cell Biol 2011;21:424–431


5. D’Hertog W, Overbergh L, Lage K, et al. Proteomics analysis of cytokine-induced dysfunction and death in insulin-producing INS-1E cells: new insightsinto the pathways involved. Mol Cell Proteomics 2007;6:2180–21996. Lenzen S. Chemistry and biology of reactive species with special referenceto the antioxidative defence status in pancreatic b-cells. Biochim Biophys Acta2017;1861:1929–19427. Yang C, Diiorio P, Jurczyk A, O’Sullivan-Murphy B, Urano F, Bortell R.Pathological endoplasmic reticulum stress mediated by the IRE1 pathway con-tributes to pre-insulitic beta cell apoptosis in a virus-induced rat model of type 1diabetes. Diabetologia 2013;56:2638–26468. Op de Beeck A, Eizirik DL. Viral infections in type 1 diabetes mellitus–why theb cells? Nat Rev Endocrinol 2016;12:263–2739. Marré ML, James EA, Piganelli JD. b cell ER stress and the implications forimmunogenicity in type 1 diabetes. Front Cell Dev Biol 2015;3:6710. Marré ML, Profozich JL, Coneybeer JT, et al. Inherent ER stress in pancreaticislet b cells causes self-recognition by autoreactive T cells in type 1 diabetes. JAutoimmun 2016;72:33–4611. Marré ML, Piganelli JD. Environmental factors contribute to b cell endo-plasmic reticulum stress and neo-antigen formation in type 1 diabetes. FrontEndocrinol (Lausanne) 2017;8:26212. van Lummel M, Duinkerken G, van Veelen PA, et al. Posttranslationalmodification of HLA-DQ binding islet autoantigens in type 1 diabetes. Diabetes2014;63:237–24713. Delong T, Baker RL, He J, Barbour G, Bradley B, Haskins K. DiabetogenicT-cell clones recognize an altered peptide of chromogranin A. Diabetes 2012;61:3239–324614. Babon JA, DeNicola ME, Blodgett DM, et al. Corrigendum: analysis of self-antigen specificity of islet-infiltrating T cells from human donors with type 1 di-abetes [published corrections appear in Nat Med 2016;22:1482–1487 and NatMed 2017;23:1004]. Nat Med 2017;23:26415. McGinty JW, Chow IT, Greenbaum C, Odegard J, Kwok WW, James EA.Recognition of posttranslationally modified GAD65 epitopes in subjects with type1 diabetes. Diabetes 2014;63:3033–304016. Mannering SI, Harrison LC, Williamson NA, et al. The insulin A-chain epitoperecognized by human T cells is posttranslationally modified. J Exp Med 2005;202:1191–119717. Rondas D, Crèvecoeur I, D’Hertog W, et al. Citrullinated glucose-regulatedprotein 78 is an autoantigen in type 1 diabetes. Diabetes 2015;64:573–58618. Delong T, Wiles TA, Baker RL, et al. Pathogenic CD4 T cells in type 1 diabetesrecognize epitopes formed by peptide fusion. Science 2016;351:711–71419. Wiles TA, Delong T, Baker RL, et al. An insulin-IAPP hybrid peptide is anendogenous antigen for CD4 T cells in the non-obese diabetic mouse. J Auto-immun 2017;78:11–1820. Eizirik DL, Cardozo AK, Cnop M. The role for endoplasmic reticulum stress indiabetes mellitus. Endocr Rev 2008;29:42–6121. Sollid LM, Jabri B. Celiac disease and transglutaminase 2: a model forposttranslational modification of antigens and HLA association in the pathogenesisof autoimmune disorders. Curr Opin Immunol 2011;23:732–73822. Bicker KL, Thompson PR. The protein arginine deiminases: structure,function, inhibition, and disease. Biopolymers 2013;99:155–16323. Muller S, Radic M. Citrullinated autoantigens: from diagnostic markers topathogenetic mechanisms. Clin Rev Allergy Immunol 2015;49:232–23924. Marchetti P, Suleiman M, Marselli L. Organ donor pancreases for the study ofhuman islet cell histology and pathophysiology: a precious and valuable resource.Diabetologia 2018;61:770–77425. Solimena M, Schulte AM, Marselli L, et al. Systems biology of the IMIDIAbiobank from organ donors and pancreatectomised patients defines a noveltranscriptomic signature of islets from individuals with type 2 diabetes. Dia-betologia 2018;61:641–65726. Brozzi F, Nardelli TR, Lopes M, et al. Cytokines induce endoplasmic reticulumstress in human, rat and mouse beta cells via different mechanisms. Diabetologia2015;58:2307–2316

27. Wessel D, Flügge UI. A method for the quantitative recovery of protein indilute solution in the presence of detergents and lipids. Anal Biochem 1984;138:141–14328. Yeung YG, Stanley ER. Rapid detergent removal from peptide samples withethyl acetate for mass spectrometry analysis. Curr Protoc Protein Sci 2010;Chapter 16:Unit 16.1229. Raijmakers R, van Beers JJBC, El-Azzouny M, et al. Elevated levels offibrinogen-derived endogenous citrullinated peptides in synovial fluid of rheu-matoid arthritis patients. Arthritis Res Ther 2012;14:R11430. James EA, Moustakas AK, Bui J, et al. HLA-DR1001 presents “altered-self”peptides derived from joint-associated proteins by accepting citrulline in threeof its binding pockets. Arthritis Rheum 2010;62:2909–291831. Yang J, Danke N, Roti M, et al. CD4+ T cells from type 1 diabetic and healthysubjects exhibit different thresholds of activation to a naturally processed pro-insulin epitope. J Autoimmun 2008;31:30–4132. Yang J, James E, Gates TJ, et al. CD4+ T cells recognize unique and con-served 2009 H1N1 influenza hemagglutinin epitopes after natural infection andvaccination. Int Immunol 2013;25:447–45733. Kwok WW, Tan V, Gillette L, et al. Frequency of epitope-specific naive CD4(+)T cells correlates with immunodominance in the human memory repertoire. JImmunol 2012;188:2537–254434. Nguyen H, James EA. Immune recognition of citrullinated epitopes. Im-munology 2016;149:131–13835. McGinty JW, Marré ML, Bajzik V, Piganelli JD, James EA. T cell epitopesand post-translationally modified epitopes in type 1 diabetes. Curr Diab Rep2015;15:9036. Kracht MJL, van Lummel M, Nikolic T, et al. Autoimmunity against a defectiveribosomal insulin gene product in type 1 diabetes. Nat Med 2017;23:501–50737. Dudek NL, Purcell AW. The Beta Cell Immunopeptidome. In Vitamins andHormones. Vol. 95, Litwack G, Ed., Burlington, VT, Academic Press, 2014, p. 115–14438. Culina S, Lalanne AI, Afonso G, et al.; ImMaDiab Study Group. Islet-reactiveCD8+ T cell frequencies in the pancreas, but not in blood, distinguish type 1diabetic patients from healthy donors. Sci Immunol 2018;3:1–1639. Rondas D, Bugliani M, D’Hertog W, et al. Glucagon-like peptide-1 protectshuman islets against cytokine-mediated b-cell dysfunction and death: a proteomicstudy of the pathways involved. J Proteome Res 2013;12:4193–420640. Benhamou PY, Watt PC, Mullen Y, et al. Human islet isolation in 104 con-secutive cases. Factors affecting isolation success. Transplantation 1994;57:1804–181041. Pileggi A, Ribeiro MM, Hogan AR, et al. Effects of pancreas cold ischemia onislet function and quality. Transplant Proc 2009;41:1808–180942. Tsiatsiani L, Heck AJR. Proteomics beyond trypsin. FEBS J 2015;282:2612–262643. Chow I-T, James EA, Gates TJ, et al. Differential binding of pyruvate de-hydrogenase complex-E2 epitopes by DRB1*08:01 and DRB1*11:01 is predictedby their structural motifs and correlates with disease risk. J Immunol 2013;190:4516–452444. James EA, Rieck M, Pieper J, et al. Citrulline-specific Th1 cells are in-creased in rheumatoid arthritis and their frequency is influenced by diseaseduration and therapy. Arthritis Rheumatol 2014;66:1712–172245. Southwood S, Sidney J, Kondo A, et al. Several common HLA-DR types sharelargely overlapping peptide binding repertoires. J Immunol 1998;160:3363–337346. Scally SW, Petersen J, Law SC, et al. A molecular basis for the association ofthe HLA-DRB1 locus, citrullination, and rheumatoid arthritis. J Exp Med 2013;210:2569–258247. Yassai M, Afsari A, Garlie J, Gorski J. C-terminal anchoring of a peptide toclass II MHC via the P10 residue is compatible with a peptide bulge. J Immunol2002;168:1281–128548. Vanderlugt CL, Miller SD. Epitope spreading in immune-mediated diseases:implications for immunotherapy. Nat Rev Immunol 2002;2:85–9549. Pugliese A. Autoreactive T cells in type 1 diabetes. J Clin Invest 2017;127:2881–2891


inflammation-induced citrullinated glucose-regulated protein ......glucose-regulated protein 78...

Documents