Identification of autoreactive CD8+ T cell responses targeting chromogranin A in humanized NOD mice and type 1 diabetes patients

Clinical Immunology (2015) 159, 63–71

available at www.sciencedirect.com

Clinical Immunology www.elsevier.com/locate/yclim

Identification of autoreactive CD8 + T cell responses targeting chromogranin A in humanized NOD mice and type 1 diabetes patients Yi Li a,1 , Lina Zhou b,1 , Yashu Li a , Jie Zhang a , Binbin Guo a , Gang Meng c , Xiaoling Chen a , Qian Zheng d , Linlin Zhang a , Mengjun Zhang e,⁎, Li Wang a,⁎ a

Institute of Immunology PLA, Third Military Medical University, Chongqing 400038, China Department of Endocrinology, Southwest Hospital, Third Military Medical University, Chongqing 400038, China c Department of Pathology, Southwest Hospital, Third Military Medical University, Chongqing 400038, China d Department of Phsiology, North Sichuan Medical College, NanChong 637007, China e Department of Pharmaceutical Analysis and Analytical Chemistry, College of Pharmacy, Third Military Medical University, Chongqing 400038, China b

Received 23 August 2014; accepted with revision 26 April 2015 Available online 6 May 2015 KEYWORDS Type 1 diabetes; Epitope; Chronogranin A; Humanized NOD mice

Abstract ChgA has recently been identified as the autoantigen for diabetogenic CD4+ T cells in NOD mice and T1D patients. However, autoreactive CD8+ T-cell responses targeting ChgA haven't been studied yet. Here several HLA-A*0201-restricted peptides derived from mChgA and hChgA were selected by an integrated computational prediction approach, followed by an HLA-A*0201 binding assay. MChgA10–19 and mChgA43–52 peptides, which bound well with HLA-A*0201 molecule, induced significant proliferation and IFN-γ-releasing of splenocytes from diabetic NOD.β2mnull.HHD mice. Notably, flow cytometry analysis found that mChgA10–19 and mChgA43–52 stimulated the production of IFN-γ, perforin, and IL-17 by splenic CD8+ T cells of diabetic NOD.β2mnull.HHD mice. Furthermore, hChgA10–19 and hChgA43–52-induced IFN-γ releasing by specific CD8+ T cells were frequently detected in recent-onset HLA-A*0201-positive T1D patients. Thus, this study demonstrated that autoreactive CD8+ T cells targeting ChgA were present in NOD.β2mnull.HHD mice and T1D patients, and might contribute to pathogenesis of T1D through secreting proinflammatory cytokines and cytotoxic molecules. © 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

⁎ Corresponding authors. Fax: + 86 23 68771928. E-mail addresses: [email protected] (M. Zhang), [email protected] (L. Wang). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.clim.2015.04.017 1521-6616/© 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

64

1. Introduction Type 1 diabetes (T1D) is an autoimmune disease characterized by T cell-mediated destruction of insulin-producing pancreatic islet β cells [1]. It has been commonly accepted that CD4+ T cells play an important role in initiation and progression of T1D [2]. However, autoreactive CD8+ T cells have been proved to play an indispensable key role in the destruction of islet β cells and development of T1D [3]. CD8+ T cells may be the early initiators that mediated the destruction of islet β cell in nonobese diabetic (NOD) mice. MHC class I molecule-deficient NOD mice do not develop insulitis [4,5], and some islet-derived CD8+ T cell clones transfer diabetes in the absence of CD4+ T cells [6,7]. Several observations in humans also inferred the functional importance of CD8+ T cells in the destruction of human islet β cells [8–11]. Thus, the identification of epitopes from islet β cell antigen recognized by CD8+ T cells in humans not only can provide new tools for a more accurate prediction of T1D, but also is important for the elaboration of new immunotherapeutic strategies. Interestingly, most β-cell antigens, such as insulin [12,13], glutamic acid decarboxylase (GAD) [14,15], and islet-specific glucose-6-phosphatase catalytic subunit-related protein (IGRP) [16,17], can be recognized by both islet-specific CD4+ T cells and CD8+ T cells. Chromogranin A (ChgA) has recently been identified as an autoantigen for diabetogenic CD4+ T cells in both NOD mice [18,19] and human T1D subjects [20]. However, it remains unknown whether ChgA can be recognized by autoreactive CD8+ T cells in T1D. Human histocompatibility leukocyte antigen-A*0201 (HLA-A*0201) is the most commonly expressed HLA class I allele in Caucasians and Asians (50%), and contributes to the susceptibility to T1D [21]. T1D onset is significantly accelerated in HLA-A*0201 transgenic NOD mice with HLA-A*0201-restricted CD8 T cells appearing in early, prediabetic insulitic lesions [22]. Therefore, several studies have used humanized NOD.β2mnull.HHD mice to identify β cell autoantigen-derived peptides that are recognized by HLA-A*0201-restricted CD8+ T cells with potential clinical relevance to human T1D [21,23,24]. However limited peptides have been identified as targets recognized by HLA-A2restricted CD8+ T cells in human T1D. In this study, we tested the autoreactivities of CD8+ T cells against several HLAA*0201-restricted candidate epitopes from murine and human ChgA protein in NOD.β2mnull.HHD mice and HLA-A*0201 positive T1D patients, respectively.

Y. Li et al. 40% female) and 5 HLA-A*0201-negative (10.4 ± 5.0, 5–18, 40%) T1D patients and 8 HLA-A*0201-positive healthy controls (29.6 ± 7.2, 22–41, 50%). All patients with recent-onset T1D (disease duration 5–389 days) were lean and presented at diagnosis with acute onset of symptoms, and have required permanent insulin treatment from the time of diagnosis. These patients were all positive for autoantibodies to islet cell, GAD and/or insulin. HLA-A*0201-positive T1D patients and healthy controls were identified by flow cytometry (BD Bioscience, Franklin Lakes, NJ, USA) using an aliquot of the cells stained with allophycocyanin (APC)-conjugated anti-HLA-A2 mAb BB7.2 (eBiosciences, San Diego, CA, USA). The study protocol was approved by the ethics committee of the Third Military Medical University, and informed consent was obtained from all participating subjects in this study.

2.2. Peptides To screen candidate HLA-A*0201 binding peptides within murine Chromogranin A (mChgA) (GeneBank accession no. P26339) and human ChgA (hChgA) (GeneBank accession no. P10645), an integrated approach combining three online T-cell epitope prediction algorithms including SYFPEITHI (http://www. syfpeithi.de/), IEDB (http://tools.immuneepitope.org/mhci/), and NetMHC (http://www.cbs.dtu.dk/services/NetMHC/) was used. The peptides scored top 10 were collected in SYFPEITHI method, and the cutoff standard of other two methods is IC50 b 500 nm. The predicted candidate m/hChgA peptides and control peptides (HIV Pol476–484 ILKEPVHGV, IGRP206–214 VYLKTNVFL, mInsA2–10 IVDQCCTSI, IGRP265–273 VLFGLGFAI), were synthesized with purity N 95% at Chinese Peptide Company (Hangzhou, China). These peptides were dissolved in DMSO (dimethyl sulfoxide, DMSO) at a concentration of 20 mg/mL, and stored at −80 °C.

2.3. HLA-A*0201 binding assay T2 cells (1 × 106/mL) were incubated with each peptide (10 μg/mL) and 3 μg/mL beta-2-microglobulin (Sigma-Aldrich, St. Louis, MO, USA) for 16 h at 37 °C. Then the cells were washed and stained with anti-HLA-A2 mAb BB7.2 (BD Bioscience), followed by incubation with FITC-conjugated goat antimouse IgG (Beyotime, Jiangsu, China), and analyzed using an FACS Canto cytometer (BD Bioscience).

2.4. Splenocyte proliferation assay

2. Materials and methods 2.1. Mice and T1D subjects NOD.β2mnull.HHD mice were purchased from the Jackson Laboratory (Bar Harbor, Maine, USA). The mice were bred and maintained in specific pathogen-free facilities. All animals were treated according to “Principles of Laboratory Animal Care and Use in Research” (Ministry of Health, Beijing, China). All experimental protocols were approved by the Animal Ethics Committee of the Third Military Medical University. Fresh blood samples were obtained from 10 HLA-A*0201positive (age [mean ± SD] 10.3 ± 3.4 years, range 5–14 years,

Splenocytes freshly isolated from 12- to 16-week-old diabetic female NOD.β2mnull.HHD mice were co-cultured in triplicate with 10 μg/mL indicated peptides in 96-well plates. After incubation at 37 °C for 72 h, [3H] thymidine (1 μCi/well) was added for an additional 16 h of culture, and uptake of [3H] thymidine was determined using a liquid scintillation counter (Beckman Coulter, Brea, CA, USA).

2.5. Mouse or human IFN-γ ELISPOT assays ELISPOT plates were precoated with anti-mouse or antihuman IFN-γ mAb (MabTech, Stockholm, Sweden) overnight at 4 °C, and blocked with RPMI 1640 plus 10% FBS (HyClone

ID of autoreactive CD8+ T cell responses targeting ChgA in humanized NOD mice and T1D patients

65

Corp., Logan, UT, USA). For mouse IFN-γ assays, splenocytes (5 × 105 cells/well) from 12- to 16-week-old diabetic female NOD.β2mnull .HHD mice were incubated with each peptide (50 μg/mL)-pulsed T2 cells (5 × 104 cells/well) for 36 h at 37 °C. For human IFN-γ assays, CD8+ T cells in PBMC were purified from T1D patients or healthy controls by using the EasySep CD8+ T cell enrichment kit (StemCell Technologies, Vancouver, Canada). Purified human CD8+ T cells (purity N90%, 1 × 105 cells/well) were incubated with each peptide (50 μg/mL)-pulsed T2 cells (1 × 104 cells/ well) for 24 h at 37 °C. After incubation, the cells were removed and plates were processed according to the IFN-γ ELISPOT kit (MabTech) manufacturer's instructions. Spots were counted using a spot reader system (Saizhi, Beijing, China).

2.6. Flow cytometry +

Mouse splenic CD8 T cells were purified by negative selection from 12- to 16-week-old diabetic female NOD.β2mnull.HHD mice by using magnetic separation (Miltenyi Biotec, Bergisch Gladbach, Germany). Purified mouse CD8+ T cells (purity N 85%, 1 × 105 cells/200 μL/well) were stimulated with each peptide (50 μg/mL)-pulsed T2 cells (5 × 104 cells/well). One hour later, 0.65 μL/mL GolgiStop™ (BD Bioscience) was added into each culture for additional 5 h-incubation at 37 °C. For intracellular cytokine staining, the cells were firstly surface-stained with PerCP5.5-conjugated anti-mouse CD8 (Sungene, Tianjing, China), and then fixed and stained for APC-conjugated antibodies against IFN-γ and PE-conjugated antibodies against perforin, or PE-conjugated antibodies against mouse IL-17A (eBiosciences) using the Fixation/Permeabilization kit (eBiosciences) according to the manufacturer's instructions. To determine whether the peptide-induced cytokine production is HLA-I-dependent, purified anti-human HLA-ABC antibody (eBioscience) was used at a final concentration of 10 μg/mL. Isotype-matched controls were included in all experiments. Events were collected on the BD Acurri C6 flow cytometer and analyzed with FlowJo software.

2.7. Statistics Results are expressed as mean ± SD. The unpaired t-test was used to make comparisons between different treatment

Table 1

Figure 1 HLA-A*0201 binding assay of m/hChgA-derived peptides. T2 cells were incubated with each peptide and the stabilization of surface HLA-A2 molecules was detected by flow cytometry. Each bar represents fluorescence index (FI). FI was calculated as follows: FI = (mean fluorescence intensity with the given peptide − mean fluorescence intensity without peptide) / (mean fluorescence intensity without peptide).

groups. P values of b 0.05 were considered statistically significant.

3. Results 3.1. Selection of potential HLA-A*0201-restricted epitopes in mChgA and hChgA To identify the mChgA and hChgA epitopes recognized by CD8+ T cells, we predicted HLA-A*0201-binding candidates through integrating three online algorithms. Peptides m/hChgA 8–16, m/hChgA 10–19, m/hChgA 43–52 were predicted as good HLA-A*0201 binders by all algorithms, whereas peptide m/hChgA75–84 was predicted positively only by SYFTEPITHI (Table 1). These four pairs of m/h ChgA peptides were selected to synthesize because their sequences are identical to or differ from each other only in one single residue (Table 1).

Predicted mChgA and hChgA epitopes binding to HLA-A*0201 molecules.

Source

Position

Sequence

NetMHC score

SYFTEPITHI score

IEDB score

mChgA hChgA mChgA hChgA mChgA hChgA mChgA hChgA

8–16 8–16 10–19 10–19 43–52 43–52 75–84 75–84

ALLLCAGQV ALLLCAGQV LLCAGQVFAL LLCAGQVTAL SLSKPSPMPV TLSKPSPMPV LLKELQDLAL LLKELQDLAL

287 287 37 169 30 56 613 613

24 24 27 27 22 20 25 25

122.21 122.21 82.70 220.54 76.29 104.83 967.12 967.12

Underlined letters indicated amino acid differences between murine and human ChgA peptide sequences.

66

Y. Li et al.

3.2. Binding affinity of candidate peptides for HLAA*0201 molecule To confirm this prediction, the binding affinity of these peptides for HLA-A*0201 molecule was determined by a T2 cell-based binding assay. An HLA-A*0201-restricted epitope HIV Pol476–484 and an H2-Kd restricted epitope IGRP206–214 were used as positive and negative controls, respectively. As indicated in Fig. 1 and Supplementary Fig. 1, m/hChgA10–19 and m/hChgA43–52 which were predicted as strong HLA-A*0201 binders, showed high actual binding affinity for HLA-A*0201; both m/hChgA8–16 predicted as a moderate HLA-A*0201 binder and m/hChgA75–84 predicted as a poor HLA-A*0201 binder displayed undetectable HLA-A*0201-binding capacity. Figure 2 Splenocyte proliferative responses to mChgA peptides. Splenocytes were isolated from seven 12- to 16-week-old diabetic female NOD.β2mnull.HHD mice, and then cells per mice were treated with 10 μg/mL indicated peptides or no peptide for 72 h in 96-well plates (2 × 105 per well) at 37 °C. The cell proliferation was measured by [3H] thymidine incorporation. Data are presented as mean of triplicate cultures per mice. SD are not shown to avoid excessive visual clutter.

3.3. Proliferative and IFN-γ responses of splenocytes from NOD.β2m null .HHD mice to mChgA candidate peptides To assess activation of T cells derived by various peptides, we measured proliferation of splenocytes from 12- to 16-week-old diabetic female NOD.β2mnull.HHD mice upon the stimulation with indicated peptides. Splenocyte proliferation induced by

Figure 3 IFN-γ ELISPOT assays of splenocyte responses to mChgA peptides. A: Representative IFN-γ ELISPOT assay demonstrated splenocyte responses to T2 cell loaded with each indicated peptides or no peptide in 12- to 16-week-old diabetic female NOD.β2mnull. B: The average number of peptide-specific IFN-γ-positive spots per 5 × 105 splenocytes in triplicate cultures was calculated for each indicated peptides or no peptide from twelve separated experiments. SD are not shown to avoid excessive visual clutter.

ID of autoreactive CD8+ T cell responses targeting ChgA in humanized NOD mice and T1D patients mChgA10–19 or mChgA43–52 was obvious, but non-statistically weaker than that induced by the known immunodominant HLA-A*0201-restricted mouse insulin A chain peptide (mInsA2–10) epitope. Whereas no positive splenocyte proliferation induced by either mChgA8–16, mChgA75–84 or medium control was observed (Fig. 2). Furthermore, we performed IFN-γ ELISPOT analysis using each peptide pulsed-T2 cells as target cells. As shown in Figs. 3A and B, compared with peptide-free-T2 cells, the peptide mInsA2–10-pulsed T2 cells stimulated abundant IFN-γ spots forming, and T2 cells loaded with mChgA10–19 and mChgA43–52 could also stimulate obvious IFN-γ secretion by splenocytes from diabetic NOD.β2mnull.HHD mice. No significant specific IFN-γ-producing cell reactivity was detected in splenocytes stimulated with T2 pulsed with mChgA8–16 or mChgA75–84. These results together indicated that HLA-A*0201 well presented peptides mChgA10–19 and mChgA43–52 could effectively stimulate the proliferation and IFN-γ releasing of splenocytes from diabetic NOD.β2mnull.HHD mice.

67

3.4. HLA-A*0201-restricted CD8+ T-cell responses to mChgA10–19 and mChgA43–52 in NOD.β2mnull.HHD mice To establish whether the mChgA peptides could induce HLAA*0201-restricted CD8+ T-cell responses in NOD.β2mnull.HHD mice, splenic CD8+ T cells were purified and their capacity to produce IFN-γ, perforin, and IL-17 were analyzed by flow cytometry. As shown in Figs. 4A and B, T2 cells pulsed with mChgA10–19, mChgA43–52 as well as mInsA2–10 stimulated the production of IFN-γ, perforin, and IL-17 by splenic CD8+ T cells of diabetic NOD.β2mnull.HHD mice. The IFN-γ-producing CD8+ T cells simultaneously secreted perforin, but no IL-17, whereas the IL-17-producing CD8+ T cells expressed high amounts of IL-17, but no IFN-γ (data not shown). Moreover, the production of these effectors molecules by CD8+ T cells was almost completely blocked by adding the anti-HLA-ABC antibody (Fig. 4A). These results demonstrated that mChgA10–19 and mChgA43–52 peptides are novel targets of HLA-A*0201-

Figure 4 Intracellular cytokine staining of peptide-specific CD8+ T cells ex vivo. The expression of IFN-γ, perforin and IL-17A by purified CD8+ T cells stimulated by T2 cells pulsed with mChgA10–19, mChgA43–52 or mInsA2–10 was detected by flow cytometry. The T2 alone (no peptide) was used as blank control. Anti-HLA-ABC (10 μg/mL) (+) or isotypic control (10 μg/mL) (−) was added for analyzing the HLA class I restriction of CD8+ T-cell responses to each peptide. The representative FACS density plots (A) and statistical analysis (B) of three independent experiments are shown.

68 Table 2

Y. Li et al. IFN-γ ELISPOT analysis of human CD8+ T cell responses to hChgA peptides. HLA-A*0201+ T1D patients (n = 10)

% positive

hChgA10–19 hChgA43–52 IGRP265–273 Positive control Basal + 3SD Basal

hChgA10–19 hChgA43–52 IGRP265–273 Positive control Basal + 3SD Basal

20 (2/10) 30 (3/10) 40 (4/10) 100 (10/10)

P01

P02

P03

P04

P05

P06

P07

P08

P09

P10

54 51 77 349 62 47

58 63 150 524 133 80

72 88 67 408 77 50

85 81 112 636 93 76

160 171 177 681 151 117

47 59 53 312 70 52

65 66 93 279 79 63

106 83 91 445 99 85

102 127 105 530 115 91

105 98 102 550 135 107

HLA-A*0201− T1D patients (n = 5)

HLA-A*0201+ healthy controls (n = 8)

% positive

P11

P12

P13

P14

P15

C1

C2

C3

C4

C5

C6

C7

C8

0 0 0 100

156 123 151 612 191 159

33 28 50 357 55 32

79 75 83 572 98 81

80 89 95 553 110 92

61 56 64 448 80 62

141 135 156 841 170 128

56 52 99 535 112 65

99 93 101 468 118 97

111 117 106 589 137 113

59 57 66 368 78 63

82 75 86 412 101 77

127 135 143 639 161 132

96 89 87 558 109 94

(0/13) (0/13) (0/13) (13/13)

Positive control: a mix of 20 ng/mL Phorbol-12-myristate-13-acetate (PMA) and 400 ng/mL ionophore (Ion). Readouts are expressed as spot-forming cells/105 CD8+ T cells. Data are presented as mean of triplicate cultures. The positive cut-off was set at 3SD above the average background responses against T2 pulsed with no peptide. Positive reactivities corresponding to values above basal + 3SD are underlined.

restricted autoreactive CD8+ T cells with the proinflammatory and cytolytic capacity in diabetic NOD.β2mnull.HHD mice.

3.5. CD8+ T-cell reactivities to hChgA10–19 and hChgA43–52 in T1D patients We then performed human IFN-γ ELISPOT directly to investigate whether CD8+ T-cell responses to hChgA10–19 and hChgA43–52 exist in T1D patients. As shown in Table 2 and Fig. 5, hChgA10–19 and hChgA43–52 were recognized by CD8+ T cells in about 20% and 30% of HLA-A*0201+ T1D patients, respectively. The CD8+ T cell response to a known immunodominant HLA-A*0201-restricted epitope IGRP265–273 was detected in almost 40% of HLA-A*0201+ T1D patients. And the two tested hChgA epitopes weren't recognized by CD8+ T cells from either HLA-A*0201+ healthy controls or HLA-A*0201− T1D patients. These results indicated that hChgA10–19 and hChgA43–52 were novel targets of HLA-A*0201-restricted autoreactive CD8+ T cells in T1D patients.

4. Discussion Although the major role of autoreactive CD8+ T cells in T1D has been well established, knowledge of the target autoantigens for CD8+ T cells especially in human T1D patients is relatively limited. ChgA, present in secretory granules of islet β cells and other neuroendocrine tissues, is identified to be a new autoantigen for autoreactive CD4+ T cells in NOD mice and human T1D [18,20]. However, autoreactive CD8+ T-cell responses targeting ChgA have not been studied yet. Herein, we firstly reported that autoreactive CD8+ T cells targeting ChgA existed in both NOD.β2mnull.HHD mice and T1D patients. Peptides ChgA10–19

and ChgA43–52 were identified as novel HLA-A*0201-restricted epitopes relevant to T1D. So far, several β-cell antigens, such as insulin [25], GAD65 [26], zinc transporter 8 [27], as well as ChgA, have been found to be targeted by both diabetogenic CD8+ and CD4+ T cells in NOD mice and T1D patients, suggesting that the recognition of the same autoantigen by autoreactive CD4+ and CD8+ T cells may benefit the formation of a proximal relationship between them and greatly contribute to the pathogenesis of autoimmune diseases. It is indicated that the peptide WE14, a natural proteolytic cleavage product from ChgA, can be converted into a highly antigenic epitope for BDC-2.5 CD4+ T cell clone through the enzyme transglutaminase-dependent modification [28]. Recent study shows that WE14 is also recognized by T cells from diabetic patients and treatment with transglutaminase increased reactivity to WE14 peptide in some patients [20]. The two HLA-A*0201-restricted ChgA epitopes identified here are not included in WE14, but ChgA43–52 is adjacent to ChgA 29–42 peptide, which is shown to be able to induce cellular and humoral immune responses in NOD mice [29]. HLA-A*0201 is one important HLA class I allele relevant to T1D [21]. Some autoantigenic peptides that were identified to be recognized by CD8+ T cells in NOD mice expressing HLA-A*0201 molecules have been shown to be targets for HLA-A*0201-restricted CD8+ T cells in T1D patients [30,31]. Through an integrated computational prediction approach, we selected four HLA-A*0201-binding candidate peptide pairs derived from murine and human ChgA protein. A positive correlation between the predicted binding affinity and the magnitude of HLA-A*0201 stabilization was observed, and two peptides mChgA10–19 and mChgA43–52 with an actual positive binding affinity for HLA-A*0201 could obviously stimulate the proliferation of splenocytes from HLA-A*0201-transgenic NOD mice, again suggesting the high

ID of autoreactive CD8+ T cell responses targeting ChgA in humanized NOD mice and T1D patients

69

Figure 5 A representative image of an IFN-γ ELISPOT assay for CD8+ T cells from HLA-A*0201+ T1D patients, HLA-A*0201− T1D patients, and HLA-A*0201+ healthy controls. Spot number counts are expressed as spot-forming cells/105 CD8+ T cells for each well.

accuracy of integrated approach for the prediction of CD8+ T cell epitopes [32]. IFN-γ, which is mainly produced by autoreactive T cells, is generally believed to be a key pathogenic factor in T1D. IFN-γ ELISPOT assay has been always used for detection and enumeration of autoreactive T cells in T1D mouse models and patients. We showed that peptides mChgA10–19 and mChgA43–52 could induce specific T cell response in NOD.β2mnull.HHD mice as determined by IFN-γ ELISPOT assay. However, this assay cannot identify individual cytokine-producing cells. By combining surface staining and intracellular cytokine staining, we further demonstrated that mChgA peptide-induced cytokine production was absolutely medicated by CD8+ T cells and this process was completely blocked by a antibody against HLA class I molecules, again confirming the HLA-A*0201 restriction. Very interestingly, we observed that some of mChgA peptidereactive CD8+ T cell clones displayed a marked cytotoxic activity by producing IFN-γ and perforin, and some other CD8+ T cell clones released high amounts of IL-17 but little IFN-γ, just similar to mInsA2–10-induced CD8+ T cell response profile. Although the autoreactive cytotoxic CD8+ T cells have been shown to directly mediate the destruction of islet β cells, several recent findings highlight the role of proinflammatory

IL-17 immunity in the pathogenesis of autoimmune diabetes mice model [33] and human T1D [34]. An increase in both CD4+ and CD8+ T cells that secreted IL-17 has been reported in the peripheral blood of children with recent onset T1D [35]. Blocking the generation of Th17 cells with the administration of B7-H4.Ig effectively prevented T1D development in NOD mice [36]. A recent study demonstrated that the autoantigenspecific IL-17-producing CD8+ T cells (Tc17) could be home to the pancreatic lymph nodes without causing any pancreatic infiltration or tissue destruction when transferred alone, but could potentiate the Th1-mediated disease progression [37]. Our results indicated β-cell autoantigens such as insulin and ChgA could generate HLA-A*0201-restricted autoreactive CD8+ T cells with the proinflammatory and cytolytic capacity, which might both participate in the pathogenesis of T1D. It has been commonly accepted that the interaction between epitope peptide and HLA-A*0201 molecule mainly depends on anchor residues at position 2 and 9 [38]. Our results indicated that the mouse ChgA-derived peptides mChgA10–19/mChgA43–52 and their human counterparts, which differ from each other by only a single residue at position 8 or 1, have similar binding affinity with HLA-A*0201, intensively implying that hChgA10–19 and hChgA43–52 peptides would be

70 recognized by CD8+ T cells in HLA-A*0201+ T1D patients. Expectedly, we demonstrated that the two peptides hChgA10–19 and hChgA43–52 induced IFN-γ response in a fraction of HLA-A*0201+ T1D patients, but neither in HLA-A*0201+ healthy controls nor in HLA-A*0201− T1D patients. Our human IFN-γ ELISPOT assay provided direct evidence that CD8+ T cells, purified from patients' blood samples, are responsible for the production of IFN-γ. Autologous or allogeneic HLA-matched APC can be used for ELISPOT assays on purified CD4+ and CD8+ T cells. As the finite amount of blood available from patients, autologous APC is rarely obtained in sufficient numbers. Therefore, the use of a common allogeneic peptide-presenting cell line appears advantageous. The human mutant lymphoid cell line T2 was widely used as allo-APC for HLA-A*0201-restricted CD8+ T cells. However, we and others have observed that significant background spot formation can be induce by T2-reactive T cells [39]. Under the high ‘background’, the mean + 3SD of the negative control triplicates was often selected as cutoff for positive response base on receiver-operator characteristic analysis for human or IFN-γ ELISPOT [23,40]. However, we observed that a CD8+ T cell response targeting human IGRP peptide IGRP265–273 was present in about 40% HLA-A*0201+ T1D patients. IGRP265–273 peptide, which is completely conserved between mouse and human IGRP, was originally identified in HLA-A*0201 transgenic NOD mice [41] and subsequently confirmed to be recognized by CD8+ T cells in some HLA-A2+ T1D patients [31]. Another study showed that 90% candidate epitopes from GAD65 and insulinoma-associated protein identified by immunization of HLA-A*0201 transgenic mice were specifically recognized by CD8+ T cells from newly diagnosed T1D patients with the different magnitude of responses [23]. As the finite amount of blood available from T1D patients, the HLA-A*0201-transgenic NOD mice model is efficient to identify autoantigen-derived epitopes relevant to human T1D. Future studies will be needed to more fully characterize the ChgA-specific CD8+ T cell responses in terms of human T1D pathogenesis. In conclusion, we firstly found that ChgA-specific CD8+ T cells were present in both HLA-A*0201 transgenic NOD mice and T1D patients, and peptides ChgA10–19 and ChgA43–52 of murine and human were identified as novel HLA-A*0201-restricted epitopes that induce the proinflammatory and cytolytic response by CD8+ T cells. The identification of the ChgA peptides is an important first step towards understanding their potential role in the pathogenesis of T1D and their possible value in the future research in human T1D patients. Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.clim.2015.04.017.

Conflict of interest statement The authors declare that there is no duality of interest associated with this manuscript.

Acknowledgment The authors acknowledge the support of the National Natural Science Foundation of China (No. 31100653 and No. 81273213) and the Fundamental and Advanced Research Projects of Chongqing (No. cstc2013jcyjA10098).

Y. Li et al.

References [1] M. Knip, H. Siljander, Autoimmune mechanisms in type 1 diabetes, Autoimmun. Rev. 7 (2008) 550–557. [2] K. Haskins, A. Cooke, CD4 T cells and their antigens in the pathogenesis of autoimmune diabetes, Curr. Opin. Immunol. 23 (2011) 739–745. [3] S. Tsai, A. Shameli, P. Santamaria, CD8+ T cells in type 1 diabetes, Adv. Immunol. 100 (2008) 79–124. [4] D.V. Serreze, E.H. Leiter, G.J. Christianson, D. Greiner, D.C. Roopenian, Major histocompatibility complex class I-deficient NOD-B2m null mice are diabetes and insulitis resistant, Diabetes 43 (1994) 505–509. [5] L.S. Wicker, E.H. Leiter, J.A. Todd, R.J. Renjilian, E. Peterson, P.A. Fischer, P.L. Podolin, M. Zijlstra, R. Jaenisch, L.B. Peterson, Beta 2-microglobulin-deficient NOD mice do not develop insulitis or diabetes, Diabetes 43 (1994) 500–504. [6] F.S. Wong, I. Visintin, L. Wen, R.A. Flavell, C.J. Janeway, CD8 T cell clones from young nonobese diabetic (NOD) islets can transfer rapid onset of diabetes in NOD mice in the absence of CD4 cells, J. Exp. Med. 183 (1996) 67–76. [7] R.T. Graser, T.P. DiLorenzo, F. Wang, G.J. Christianson, H.D. Chapman, D.C. Roopenian, S.G. Nathenson, D.V. Serreze, Identification of a CD8 T cell that can independently mediate autoimmune diabetes development in the complete absence of CD4 T cell helper functions, J. Immunol. 164 (2000) 3913–3918. [8] A. Imagawa, T. Hanafusa, S. Tamura, M. Moriwaki, N. Itoh, K. Yamamoto, H. Iwahashi, K. Yamagata, M. Waguri, T. Nanmo, S. Uno, H. Nakajima, M. Namba, S. Kawata, J.I. Miyagawa, Y. Matsuzawa, Pancreatic biopsy as a procedure for detecting in situ autoimmune phenomena in type 1 diabetes: close correlation between serological markers and histological evidence of cellular autoimmunity, Diabetes 50 (2001) 1269–1273. [9] A.M. Bulek, D.K. Cole, A. Skowera, G. Dolton, S. Gras, F. Madura, A. Fuller, J.J. Miles, E. Gostick, D.A. Price, J.W. Drijfhout, R.R. Knight, G.C. Huang, N. Lissin, P.E. Molloy, L. Wooldridge, B.K. Jakobsen, J. Rossjohn, M. Peakman, P.J. Rizkallah, A.K. Sewell, Structural basis for the killing of human beta cells by CD8(+) T cells in type 1 diabetes, Nat. Immunol. 13 (2012) 283–289. [10] W.W. Unger, T. Pearson, J.R. Abreu, S. Laban, A.R. van der Slik, S.M. der Kracht, M.G. Kester, D.V. Serreze, L.D. Shultz, M. Griffioen, J.W. Drijfhout, D.L. Greiner, B.O. Roep, Isletspecific CTL cloned from a type 1 diabetes patient cause beta-cell destruction after engraftment into HLA-A2 transgenic NOD/scid/IL2RG null mice, PLoS One 7 (2012) e49213. [11] A. Skowera, R.J. Ellis, R. Varela-Calvino, S. Arif, G.C. Huang, C. Van-Krinks, A. Zaremba, C. Rackham, J.S. Allen, T.I. Tree, M. Zhao, C.M. Dayan, A.K. Sewell, W.W. Unger, J.W. Drijfhout, F. Ossendorp, B.O. Roep, M. Peakman, CTLs are targeted to kill beta cells in patients with type 1 diabetes through recognition of a glucose-regulated preproinsulin epitope, J. Clin. Invest. 118 (2008) 3390–3402. [12] G.G. Pinkse, O.H. Tysma, C.A. Bergen, M.G. Kester, F. Ossendorp, P.A. van Veelen, B. Keymeulen, D. Pipeleers, J.W. Drijfhout, B.O. Roep, Autoreactive CD8 T cells associated with beta cell destruction in type 1 diabetes, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 18425–18430. [13] S.C. Kent, Y. Chen, L. Bregoli, S.M. Clemmings, N.S. Kenyon, C. Ricordi, B.J. Hering, D.A. Hafler, Expanded T cells from pancreatic lymph nodes of type 1 diabetic subjects recognize an insulin epitope, Nature 435 (2005) 224–228. [14] P. Panina-Bordignon, R. Lang, P.M. van Endert, E. Benazzi, A.M. Felix, R.M. Pastore, G.A. Spinas, F. Sinigaglia, Cytotoxic T cells specific for glutamic acid decarboxylase in autoimmune diabetes, J. Exp. Med. 181 (1995) 1923–1927. [15] G.T. Nepom, J.D. Lippolis, F.M. White, S. Masewicz, J.A. Marto, A. Herman, C.J. Luckey, B. Falk, J. Shabanowitz, D.F.

ID of autoreactive CD8+ T cell responses targeting ChgA in humanized NOD mice and T1D patients

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

Hunt, V.H. Engelhard, B.S. Nepom, Identification and modulation of a naturally processed T cell epitope from the diabetesassociated autoantigen human glutamic acid decarboxylase 65 (hGAD65), Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 1763–1768. S.M. Lieberman, A.M. Evans, B. Han, T. Takaki, Y. Vinnitskaya, J.A. Caldwell, D.V. Serreze, J. Shabanowitz, D.F. Hunt, S.G. Nathenson, P. Santamaria, T.P. DiLorenzo, Identification of the beta cell antigen targeted by a prevalent population of pathogenic CD8+ T cells in autoimmune diabetes, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 8384–8388. J. Yang, N.A. Danke, D. Berger, S. Reichstetter, H. Reijonen, C. Greenbaum, C. Pihoker, E.A. James, W.W. Kwok, Islet-specific glucose-6-phosphatase catalytic subunit-related protein-reactive CD4+ T cells in human subjects, J. Immunol. 176 (2006) 2781–2789. B.D. Stadinski, T. Delong, N. Reisdorph, R. Reisdorph, R.L. Powell, M. Armstrong, J.D. Piganelli, G. Barbour, B. Bradley, F. Crawford, P. Marrack, S.K. Mahata, J.W. Kappler, K. Haskins, Chromogranin A is an autoantigen in type 1 diabetes, Nat. Immunol. 11 (2010) 225–231. K. Yoshida, T. Martin, K. Yamamoto, C. Dobbs, C. Munz, N. Kamikawaji, N. Nakano, H.G. Rammensee, T. Sasazuki, K. Haskins, H. Kikutani, Evidence for shared recognition of a peptide ligand by a diverse panel of non-obese diabetic mice-derived, islet-specific, diabetogenic T cell clones, Int. Immunol. 14 (2002) 1439–1447. P.A. Gottlieb, T. Delong, R.L. Baker, L. Fitzgerald-Miller, R. Wagner, G. Cook, M.R. Rewers, A. Michels, K. Haskins, Chromogranin A is a T cell antigen in human type 1 diabetes, J. Autoimmun. 50 (2014) 38–41. T. Takaki, M.P. Marron, C.E. Mathews, S.T. Guttmann, R. Bottino, M. Trucco, T.P. DiLorenzo, D.V. Serreze, HLA-A*0201restricted T cells from humanized NOD mice recognize autoantigens of potential clinical relevance to type 1 diabetes, J. Immunol. 176 (2006) 3257–3265. M.P. Marron, R.T. Graser, H.D. Chapman, D.V. Serreze, Functional evidence for the mediation of diabetogenic T cell responses by HLA-A2.1 MHC class I molecules through transgenic expression in NOD mice, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 13753–13758. P. Blancou, R. Mallone, E. Martinuzzi, S. Severe, S. Pogu, G. Novelli, G. Bruno, B. Charbonnel, M. Dolz, L. Chaillous, P. van Endert, J.M. Bach, Immunization of HLA class I transgenic mice identifies autoantigenic epitopes eliciting dominant responses in type 1 diabetes patients, J. Immunol. 178 (2007) 7458–7466. D.V. Serreze, M.P. Marron, T.P. Dilorenzo, “Humanized” HLA transgenic NOD mice to identify pancreatic beta cell autoantigens of potential clinical relevance to type 1 diabetes, Ann. N. Y. Acad. Sci. 1103 (2007) 103–111. F.S. Wong, J. Karttunen, C. Dumont, L. Wen, I. Visintin, I.M. Pilip, N. Shastri, E.G. Pamer, C.J. Janeway, Identification of an MHC class I-restricted autoantigen in type 1 diabetes by screening an organ-specific cDNA library, Nat. Med. 5 (1999) 1026–1031. A. Quinn, M.F. McInerney, E.E. Sercarz, MHC class I-restricted determinants on the glutamic acid decarboxylase 65 molecule induce spontaneous CTL activity, J. Immunol. 167 (2001) 1748–1757. E. Enee, R. Kratzer, J.B. Arnoux, E. Barilleau, Y. Hamel, C. Marchi, J. Beltrand, B. Michaud, L. Chatenoud, J.J. Robert, P.

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

71

van Endert, ZnT8 is a major CD8+ T cell-recognized autoantigen in pediatric type 1 diabetes, Diabetes 61 (2012) 1779–1784. T. Delong, R.L. Baker, J. He, G. Barbour, B. Bradley, K. Haskins, Diabetogenic T-cell clones recognize an altered peptide of chromogranin A, Diabetes 61 (2012) 3239–3246. E. Nikoopour, R. Cheung, S. Bellemore, O. Krougly, E. LeeChan, M. Stridsberg, B. Singh, Vasostatin-1 antigenic epitope mapping for induction of cellular and humoral immune responses to chromogranin A autoantigen in NOD mice, Eur. J. Immunol. 44 (2014) 1170–1180. I. Jarchum, J.C. Baker, T. Yamada, T. Takaki, M.P. Marron, D.V. Serreze, T.P. DiLorenzo, In vivo cytotoxicity of insulinspecific CD8+ T-cells in HLA-A*0201 transgenic NOD mice, Diabetes 56 (2007) 2551–2560. I. Jarchum, L. Nichol, M. Trucco, P. Santamaria, T.P. DiLorenzo, Identification of novel IGRP epitopes targeted in type 1 diabetes patients, Clin. Immunol. 127 (2008) 359–365. B. Trost, M. Bickis, A. Kusalik, Strength in numbers: achieving greater accuracy in MHC-I binding prediction by combining the results from multiple prediction tools, Immunome Res. 3 (2007) 5. J. Honkanen, J.K. Nieminen, R. Gao, K. Luopajarvi, H.M. Salo, J. Ilonen, M. Knip, T. Otonkoski, O. Vaarala, IL-17 immunity in human type 1 diabetes, J. Immunol. 185 (2010) 1959–1967. V.S. Costa, T.C. Mattana, S.M. Da, Unregulated IL-23/IL-17 immune response in autoimmune diseases, Diabetes Res. Clin. Pract. 88 (2010) 222–226. A.K. Marwaha, S.Q. Crome, C. Panagiotopoulos, K.B. Berg, H. Qin, Q. Ouyang, L. Xu, J.J. Priatel, M.K. Levings, R. Tan, Cutting edge: increased IL-17-secreting T cells in children with new-onset type 1 diabetes, J. Immunol. 185 (2010) 3814–3818. I.F. Lee, X. Wang, J. Hao, N. Akhoundsadegh, L. Chen, L. Liu, S. Langermann, D. Ou, G.L. Warnock, B7-H4.Ig inhibits the development of type 1 diabetes by regulating Th17 cells in NOD mice, Cell. Immunol. 282 (2013) 1–8. A. Saxena, S. Desbois, N. Carrie, M. Lawand, L.T. Mars, R.S. Liblau, Tc17 CD8+ T cells potentiate Th1-mediated autoimmune diabetes in a mouse model, J. Immunol. 189 (2012) 3140–3149. K.C. Parker, M.A. Bednarek, L.K. Hull, U. Utz, B. Cunningham, H.J. Zweerink, W.E. Biddison, J.E. Coligan, Sequence motifs important for peptide binding to the human MHC class I molecule, HLA-A2, J. Immunol. 149 (1992) 3580–3587. C.M. Britten, R.G. Meyer, T. Kreer, I. Drexler, T. Wolfel, W. Herr, The use of HLA-A*0201-transfected K562 as standard antigen-presenting cells for CD8(+) T lymphocytes in IFNgamma ELISPOT assays, J. Immunol. Methods 259 (2002) 95–110. E. Enee, E. Martinuzzi, P. Blancou, J.M. Bach, R. Mallone, P. van Endert, Equivalent specificity of peripheral blood and isletinfiltrating CD8+ T lymphocytes in spontaneously diabetic HLAA2 transgenic NOD mice, J. Immunol. 180 (2008) 5430–5438. T. Takaki, M.P. Marron, C.E. Mathews, S.T. Guttmann, R. Bottino, M. Trucco, T.P. DiLorenzo, D.V. Serreze, HLA-A*0201restricted T cells from humanized NOD mice recognize autoantigens of potential clinical relevance to type 1 diabetes, J. Immunol. 176 (2006) 3257–3265.