Prevalent CD8(+) T cell response against one peptide/MHC complex in autoimmune diabetes

Abstract

Spontaneous autoimmune diabetes in nonobese diabetic (NOD) mice is the result of a CD4(+) and CD8(+) T cell-dependent autoimmune process directed against the pancreatic beta cells. CD8(+) T cells play a critical role in the initiation and progression of diabetes, but the specificity and diversity of their antigenic repertoire remain unknown. Here, we define the structure of a peptide mimotope that elicits the proliferation, cytokine secretion, differentiation, and cytotoxicity of a diabetogenic H-2K(d)-restricted CD8(+) T cell specificity (NY8.3) that uses a T cell receptor alpha (TCRalpha) rearrangement frequently expressed by CD8(+) T cells propagated from the earliest insulitic lesions of NOD mice (Valpha17-Jalpha42 elements, often joined by the N-region sequence M-R-D/E). Stimulation of splenic CD8(+) T cells from single-chain 8. 3-TCRbeta-transgenic NOD mice with this mimotope leads to preferential expansion of T cells bearing an endogenously derived TCRalpha chain identical to the one used by their islet-associated CD8(+) T cells, which is also identical to the 8.3-TCRalpha sequence. Cytotoxicity assays using islet-derived CD8(+) T cell clones from nontransgenic NOD mice as effectors and peptide-pulsed H-2K(d)-transfected RMA-S cells as targets indicate that nearly half of the CD8(+) T cells recruited to islets in NOD mice specifically recognize the same peptide/H-2K(d) complex. This work demonstrates that beta cell-reactive CD8(+) T cells mount a prevalent response against a single peptide/MHC complex and provides one peptide ligand for CD8(+) T cells in autoimmune diabetes.

Figure 1

Screening of the first-generation peptide libraries. ⁵¹Cr-labeled RMA-SK^d cells were incubated with 60 μg/ml of each of the 72 first-generation libraries at 26°C for 1 h and then challenged with 8.3-CD8⁺ CTL clones at a 1:10 target/effector ratio for 8 h at 37°C. The amino acids/position that consistently elicited 8.3-CTL-induced lysis of peptide-pulsed targets are shown at the top.

Figure 2

Screening of the second-generation peptide libraries. RMA-SK^d cells were pulsed with each of the 12-sec generation libraries or tum and tested in cytotoxicity assays as in Fig. 1. The amino acids that elicited ⁵¹Cr-release values >2 SD above those elicited by tum (horizontal line) are shown at the top. NT, not tested.

Figure 3

Immunological properties of NRP. (A) Proliferation of CTLp from 8.3-NOD and NOD/Lt mice in response to NOD islet-cells (Left) or NOD splenocytes pulsed with NRP or tum (Right) (P < 0.05 for islet cell- or NRP- vs. tum-induced proliferation of 8.3-CTLp, and NRP-induced proliferation of 8.3-CTLp vs. NOD/Lt T cells). (B) Cytokine secretion by 8.3-CTLp in response to NOD islet cells or splenocytes pulsed with 1 μg/ml of NRP or tum (Center and Right) (P < 0.009 for IFN-γ/IL-2 secretion induced by NRP vs. tum). (C) Differentiation of 8.3-CTLp from RAG-2^−/− 8.3-NOD mice into CTL. The panel shows ⁵¹Cr-release assays by using NRP- or tum-pulsed RMA-SK^d cells (Left and Center) or NOD or B6 islet cells as targets, at a 1:10 target/effector ratio (Right) (P < 0.05 for NRP vs. tum reactivity or NOD vs. B6 islet cell reactivity).

Figure 4

Alanine-scanning mutagenesis of NRP. (A) H-2K^d-stabilization assay. RMA-SK^d cells were pulsed with peptides, stained with anti-K^d or anti-D^b mAbs, and analyzed by flow cytometry (P < 0.02 for K^d-binding of all peptides vs. ESP). Data shown correspond to cells pulsed with 1 μg/ml of peptide, but the different peptides displayed comparable differences in their ability to stabilize K^d molecules at lower concentrations. (B) Proliferation of 8.3-CD8⁺ CTLp in response to peptide-pulsed NOD splenocytes [P < 0.05 for analog peptide- (except A4 and A8) vs. tum-induced]. (C and D) Secretion of IFN-γ (C) and IL-2 (D) by 8.3-CTLp in response to peptide-pulsed NOD splenocytes [P < 0.009 for analog peptide- (all except A4 and A8) vs. tum-induced IFN-γ secretion; P < 0.009 for A1, A6 and A7 vs. tum-induced IL-2 secretion]. (E) Cytotoxicity of NRP-differentiated 8.3-CTL against peptide-pulsed RMA-SK^d cells at a 1:10 target/effector ratio (P < 0.05 for A1-, A3-, A6-, and A7- vs. tum-induced).

Figure 5

NRP reactivity of islet-associated CD8⁺ T cells from NOD/Lt mice. (A) NRP and tum reactivity of islet- and spleen-derived CD8⁺ T cell lines from 8.3-NOD and/or NOD/Lt mice. CD3-activated and Vβ8.1-act. are control CD8⁺ CTLs generated by stimulation of NOD splenic CD8⁺ T cells with plate-bound anti-CD28 and anti-CD3 or anti-Vβ8.1/8.2 mAbs. The figure shows results of cytotoxicity assays obtained with one islet line, two splenic lines, and the average values obtained with seven NRP-reactive and one nonreactive islet lines (at a 1:10 target/effector ratio) (P < 0.03). The tum reactivity of the +lines was caused by one line that displayed some cytotoxicity against tum-pulsed targets. (B) NRP reactivity of islet-derived CD8⁺ T cell clones from NOD/Lt mice. Assays were done at a ≈1:1 target/effector ratio. A clone was defined as positive if it triggered ⁵¹Cr-release values from NRP-pulsed targets at least 2 SD above those triggered by tum-pulsed targets. The figure shows results of assays obtained with one clone and average values corresponding to 14 NRP-reactive clones (P < 0.004 for NRP vs. tum reactivity) and 17 non-NRP-reactive clones. (C) Percentage of NRP-reactive CTL lines and clones from NOD/Lt mice (P < 0.0001 for % of NRP- vs. tum-reactive clones).