Mechanistic and therapeutic insights in autoimmune diabetes would benefit from a more complete identification of relevant autoantigens. BDC2.5 TCR transgenic NOD mice express transgenes for TCR Vα1 and Vβ4 chains from the highly diabetogenic BDC2.5 CD4+ T cell clone, which recognizes pancreatic β cell membrane Ags presented by NOD I-Ag7 MHC class II molecules. The antigenic epitope of BDC2.5 TCR is absent in β cells that do not express chromogranin A (ChgA) protein. However, characterization of the BDC2.5 epitope in ChgA has given inconclusive results. We have now identified a ChgA29–42 peptide within vasostatin-1, an N-terminal natural derivative of ChgA as the BDC2.5 TCR epitope. Having the necessary motif for binding to I-Ag7, it activates BDC2.5 T cells and induces an IFN-γ response. More importantly, adoptive transfer of naive BDC2.5 splenocytes activated with ChgA29–42 peptide transferred diabetes into NOD/SCID mice.
A restricted T cell repertoire targeted to a major autoantigen may represent an important early event in the onset of type 1 diabetes (1). BDC2.5 TCR transgenic NOD mice express genes for Vα1 and Vβ4 chains of TCR for BDC2.5 T cells clone (2, 3) isolated from the spleen and lymph nodes of diabetic NOD mice (3–6). The BDC2.5 T cell clone has a Th1 phenotype and reacts with β cell membrane Ags in the context of I-Ag7 MHC class II (2). It has been shown that BDC2.5 CD4+ T cells of BDC2.5 NOD mice can recognize glutamic acid decarboxylase (GAD)524–543 epitope (7) and a variety of related mimotopes (8).
Recent studies have identified chromogranin A (ChgA), a 463 aa secretory protein of granin family, as the autoantigen for BDC2.5 cells (9). ChgA is a precursor to several functional peptides including vasostatin-1 (VS-1; ChgA1–76), VS-2 (ChgA1–113), and WE-14 (ChgA358–371) (10). A similarity between a common binding motif (WSRMD) within mimotopes and ChgA sequence led to the assignment of naturally processed WE14 peptide as the nominal antigenic epitope of BDC2.5 T cell clone (9). However, WE-14 lacks a full binding motif for I-Ag7, the key class II restriction molecule in NOD mice (9). As T cells in BDC2.5 transgenic NOD mice reflect much more physiological relevance to an in vitro-differentiated BDC2.5 clone (2), we investigated other potential antigenic determinants in ChgA protein that may stimulate naive CD4+ cells from BDC2.5 TCR transgenic NOD mice. We scanned peptides from ChgA protein that are able to bind to I-Ag7 class II MHC molecule based on predictions for binding motif of peptides to the I-Ag7 molecule (11, 12) and identified a new epitope that is responsible for the activation of BDC2.5 TCR CD4+ cells. In this study, we report for the first time, to our knowledge, that ChgA29–42 peptide, which is part of VS-1, is the natural epitope of BDC2.5 T cells.
Materials and Methods
BDC2.5 TCR transgenic NOD, NOD/Ltj, and NOD/SCID mice were bred and housed under specific pathogen-free animal facility at the University of Western Ontario. BDC2.5 TCR transgenic NOD mice were obtained from The Jackson Laboratory (Bar Harbor, ME). Mice were used in accordance with institutional guidelines.
NOD MHC class II (I-Ag7) peptide-binding motifs previously described (11, 12) were used to identify potential I-Ag7 binding sites within the murine ChgA amino acid sequence. ChgA29–42 (DTKVMKCVLEVISD), ChgA34–44 (KCVLEVISDSL), ChgA36–44 (VLEVISDSL), WE14: ChgA358–371 (WSRMDQLAKELTAE), PS3 mimotope (SRLGLWVRME), GAD524–538 (SRLSKVAPVIKARMM), and negative control nonstimulatory peptide LSIALHVGFDH (NEG) were synthesized and characterized in our laboratory as previously described (12). Characterization data for ChgA29–42 and WE-14 peptides are presented in Supplemental Fig. 1. Peptides were dissolved in water, pH adjusted with 0.1 M acetic acid or 0.1 M HCl, and sterilized by filtration through a 0.22-μm filter.
Anti–I-Ag7 mAb 10-2.16 (American Type Culture Collection) was purified from hybridoma cultures (12). allophycocyanin-conjugated anti-mouse CD4 (L3T4), mouse IgG2b isotype control (eBioscience, San Diego, CA), and PE-conjugated anti-mouse Vβ4 TCR (KT4) was obtained from BD Pharmingen (San Jose, CA).
T cell proliferation assay
Spleen cells from 4-wk-old BDC2.5 TCR transgenic NOD mice were cultured in 96-well round-bottom plates (BD Biosciences) at 2 × 105 cells/well in the presence of ChgA29–42 or other peptides indicated at a concentration of 200, 100, 50, 20, 10, and 5 μg/ml in RPMI 1640 media (Invitrogen) supplemented with 10% FCS (HyClone), 10 mM HEPES, 2 mM l-glutamine, 5 × 10−5 M 2-ME, and 1 U/ml penicillin/streptomycin) at 37°C for 72 h. After 3 d, cultures were pulsed with 1 μCi/well [3H]thymidine (PerkinElmer) for 18 h, and uptake of [3H]thymidine was measured using a liquid scintillation counter (LKB Instruments).
I-Ag7 dimer peptide binding assay
The I-Ag7 dimer used in this study contains two empty truncated I-Ag7 molecules fused to the Fc portion of mouse IgG2a Ab (13). For ELISA-based assay (14), goat anti-mouse IgG2a Ab (Caltag Laboratories) was coated on an ELISA plate overnight. Biotinylated HEL peptide (200 nM), competitor peptides at various concentrations, and I-Ag7 dimer (200 ng/well) were incubated for 24 h at room temperature in binding buffer (6.7 mM citric phosphate buffer [pH 7], 0.15 M NaCl, 2% Nonidet P-40, 2 mM EDTA, and protease inhibitors leupeptin and aprotinin). The mixtures were then transferred to the washed ELISA plate and incubated for 2 h at room temperature. I-Ag7 dimer/biotinylated HEL complexes that bound to the ELISA plate were detected by adding streptavidin-AKP and NPP substrate, followed by reading the plate at 405 nm.
Spleen cells were harvested from 3-wk-old BDC2.5 TCR transgenic NOD mice and cultured in 96-well round-bottom plates at 2 × 105 cells/well in the presence of 100 μg/ml ChgA29–42, WE14, PS3 mimotope, or NEG control peptides in RPMI-10 culture media at 37°C. After 4 d, peptide-primed cells (10 × 106) were injected i.v. into NOD/SCID recipients. Mice were monitored for the development of diabetes by measuring urine glucose with Diastix strips (Bayer) twice a week.
For CFSE dye dilution assay, splenocytes from 2- to 3-wk old BDC2.5 NOD mice were stimulated with the indicated peptides and labeled with 5 μM CFSE and injected i.p. into 10-d-old NOD mice. After 1 wk, spleen, pancreatic lymph nodes (PLN) and mesenteric lymph nodes (MLN) were analyzed for recruitment and proliferation of CFSE-labeled Vβ4+ CD4+ donor cells.
All experiments were repeated at least three times with reproducible results. Figures show mean ± SD of data from representative experiments. The Fisher exact probability test was used to analyze adoptive transfer data for diabetes development, and a p value <0.05 was considered significant.
Results and Discussion
ChgA29–42 is an epitope of BDC2.5 TCR CD4+ T cells in BDC2.5 TCR transgenic NOD mice
Alignment of ChgA29–42 aa sequence with putative I-Ag7 binding motif shows this peptide has three critical amino acid positions (L, V, D for P4, P6, and P9 amino acid position, respectively) that fit well with the I-Ag7 binding motif (Fig. 1A). This peptide is part of VS-1, which is an N-terminal fragment of ChgA. To determine the T cell proliferative response of ChgA29–42 peptide, splenocytes from 6-wk-old BDC2.5 NOD mice were stimulated in the presence or absence of the peptide at various concentrations. After 72 h, proliferation was measured by [3H]thymidine incorporation (Fig. 1B). The PS3 mimotope peptide (9) that is known to strongly stimulate BDC2.5 TCR CD4+ T cells was used as a positive control. As shown in Fig. 1B, ChgA29–42 and PS3 peptides were able to induce proliferation of BDC2.5 splenocytes at various concentration ranges. In contrast, WE14 epitope (9) of ChgA protein and the GAD65 epitope GAD524–538 only weakly stimulated BDC2.5 cells. The GAD524–538 proliferative response was much lower than ChgA29–42 epitope, but it was significantly higher than WE14 peptide. We therefore conclude that ChgA29–42 is the ChgA determinant recognized by BDC2.5 T cells. We also analyzed the cytokines in supernatants of BDC2.5 splenocytes after stimulation with ChgA29–42 peptide at 24, 48, and 72 h. As shown in Fig. 1C, compared with controls, there was a gradual increase in the level of IFN-γ and IL-2 cytokines in cultures of cells stimulated with ChgA29–42 at all time points. Stimulation with PS3 mimotope caused secretion of significantly higher levels of IFN-γ after 24 h that lasted for 72 h. There was a slight buildup of the IL-2 cytokine secretion at 48 and 72 h after stimulation with ChgA29–42 peptide. PS3 mimotope caused BDC2.5 splenocytes to secrete the highest level of IL-2 at 24 h, but the cytokine response tapered off over time.
As shown in Supplemental Fig. 2, we monitored BDC2.5 CD4+ T cells for the upregulation of the activation markers CD69, CD44, and CD25 after stimulation with ChgA29–42, PS3 mimotope, WE14, or NEG control peptide. We found large phenotypic changes with ChgA29–42 peptide and PS3 mimotope, but WE14 induced only marginal changes.
We next examined the amino acid residues of ChgA29–42 peptide that are critical for BDC2.5 T cell stimulation. We measured the proliferative response of splenocytes from BDC2.5 TCR NOD mice to several N-terminal truncated peptides (Fig. 2A). As shown in Fig. 2B, the ChgA36–44 peptide over a wide range of doses had minimal activity, and ChgA34–44 peptide weakly stimulated BDC2.5 T cells. This showed that although amino acids located at the C-terminal part of ChgA29–42 molecule are necessary for binding to I-Ag7 and BDC2.5 TCR stimulation, amino acids at the N-terminal of the peptide increase further the magnitude of BDC2.5 T cell stimulation. In the competitive binding experiments for the above peptides (Fig. 2C), we found that ChgA29–33 sequence of the ChgA29–42 epitope greatly enhanced its binding to I-Ag7 compared with the shorter ChgA34–44 peptide. We also found that the binding of ChgA29–42 peptide to I-Ag7 is much greater than that of WE14 peptide. There was a hierarchy in binding affinity of peptides to I-Ag7 in the order of PS3 > ChgA29–42 > ChgA34–44 > ChgA36–44 = WE14.
Presentation of ChgA29–42 epitope by I-Ag7 molecule was confirmed by blocking studies with anti–I-Ag7 mAb. Single-cell suspension of splenocytes from 6-wk-old BDC2.5 NOD were cultured with ChgA29–42 peptide (100 μg/ml) in the presence or absence of anti-MHC class II I-Ag7 Ab (10-2.16) at concentrations of 10, 1, 0.1, and 0.01 μg/ml. Anti–I-Ag7 mAb (10 μg/ml) inhibited proliferation of BDC2.5 cells stimulated with ChgA29–42 peptide (Fig. 3), and a decrease in anti–I-Ag7 concentrations released proliferating BDC2.5 cells from blocking effect of the Ab. This suggests ChgA29–42 activates BDC2.5 T cells specifically in the context of I-Ag7 in NOD mice.
ChgA29–42 peptide primed BDC2.5 splenocytes proliferate in PLN of NOD mice
We then investigated the in vivo proliferative capacity of BDC2.5 T cells activated with ChgA29–42 in the PLN of NOD mice. Splenocytes from young BDC2.5 NOD mice, representing mostly naive T cells, were stimulated with ChgA29–42, PS3 mimotope, or NEG-negative control peptides. After 3 d, cells were labeled with CFSE dye and adoptively transferred (i.p. injection) into 10-d-old NOD mice. Analysis of cells gated for CD4+ lymphocytes expressing Vβ4+ TCR in PLN showed more proliferation and expansion of the BDC2.5 CD4+ T cells activated with ChgA29–42 or PS3 compared with cells treated with negative control peptide (Fig. 4). BDC2.5 cells stimulated with ChgA29–42 epitope and PS3 mimotope showed dilution of CFSE dye (33.4 and 44%, respectively), whereas BDC2.5 treated with negative control peptide failed to dilute the dye in the PLNs. Adoptively transferred cells failed to proliferate in the control MLN. As Ag-activated BDC2.5 T cells had already experienced Ag in vitro and showed more memory phenotype, it is reasonable that these cells expand upon re-encounter with Ag in vivo in PLN due to lower threshold for activation.
Adoptive transfer of ChgA29–42-activated BDC2.5 splenocytes into NOD mice induces diabetes
BDC2.5 T cells are mostly naive, and nonspecific activation with anti-CD3 and anti-CD28 leads to diabetes development upon transfer to the recipient NOD/SCID mice. We sought to investigate the potential of ChgA29–42-activated BDC2.5 lymphocytes to induce diabetes. Splenocytes from young, 3-wk-old BDC2.5 NOD mice were stimulated with 100 μg/ml ChgA29–42 peptide for 4 d. Cells activated nonspecifically with anti-CD3 plus anti-CD28 were used as positive control. BDC2.5 splenocytes were also treated with PS3 mimotope and BDC2.5-negative control peptide. After 4 d of in vitro stimulation, splenocytes (10 × 106 for each mouse) were injected into the tail veins of NOD/SCID mice (Fig. 5). Mice that received splenocytes treated with control NEG peptide (n = 6) remained diabetes free. In contrast, 80% of NOD mice that received activated BDC2.5 cells with anti-CD3 and anti-CD28 (n = 6) developed diabetes at day 9 posttransfer with all becoming diabetic by day 14 following adoptive transfer. Similarly, NOD/SCID mice (n = 10) that received BDC2.5 splenocytes activated with ChgA29–42 epitope developed diabetes (70%) at day 7 of adoptive transfer, and all of them were diabetic after 14 d of adoptive transfer. All of the NOD/SCID mice receiving BDC2.5 splenocytes stimulated with mimotope PS3 peptide (n = 8) were diabetic by day 6. In the WE14 group, only 3 out of 11 (27%) mice became diabetic. Therefore, ChgA29–42 peptide activated and induced functional immune responses in T cells derived from BDC2.5 TCR transgenic NOD cells.
In conclusion, we have identified a new epitope from ChgA protein that is able to bind specifically to I-Ag7 MHC class II molecule to induce proliferation of BDC2.5 CD4+ T cell in NOD mice. In previous studies, Stadinski et al. (9, 15) identified a naturally processed peptide of ChgA protein, WE14, as an antigenic target of BDC2.5 T cells. The I-Ag7 binding motif in WE14 peptide (WSRMD) is very similar to the binding motif of various mimotope peptides (WVRME) and, theoretically, makes it an ideal candidate as an antigenic epitope for BDC2.5 T cells. However, the WEDKRWSRMD amino acid sequence of WE14 peptide was not able to raise an immune response in BDC2.5 T cell clones (9). Stadinski et al. (9, 15) has suggested that ChgA-derived WE-14 peptide has an unusual binding to IAg7 as it only partially filled the peptide-binding groove, leaving positions P1–P4 empty, and it extended far beyond the peptide-binding groove at its C terminus. Although the presence of ChgA in β cell membrane was required for activation of the BDC2.5 clone, WE14 peptide of ChgA protein was weaker than β cell crude extract (9). We suggest that WE14 is not an epitope but a weak mimotope for BDC2.5 T cells. Our results show that ChgA29–42 epitope is able to induce and mediate a significant immune response by BDC2.5 CD4+ T cells in NOD mice. Moreover, compared with WE14 peptide, the ChgA29–42 sequence has the necessary motif for the I-Ag7 binding cleft without a need to bind to an outer part of the MHC molecule. We therefore propose ChgA29–42 as the antigenic epitope of BDC2.5 T cells that will resolve the above-mentioned problems inherent in WE14 as an epitope.
The authors have no financial conflicts of interest.
We thank Dr. Joaquin Madreans and Dr. Anthony M. Jevnikar for helpful comments and suggestions.
This work was supported by grants from the Canadian Institutes of Health Research.
The online version of this article contains supplemental material.
Abbreviations used in this article:
- chromogranin A
- glutamic acid decarboxylase
- mesenteric lymph node
- nonstimulatory peptide LSIALHVGFDH
- pancreatic lymph node
- Received November 8, 2010.
- Accepted February 1, 2011.
- Copyright © 2011 by The American Association of Immunologists, Inc.