The strongest response was induced by peptide 10–26 followed by peptides 289–306, 117–133/120–133, and 46–70, as determined by high levels of IFN-γ as well as the presence of IL-2 in culture supernatants (Fig. 2). The two peptides 117–133 and 120–133 led to a similar IFN-γ response, although the longer sequence BMN673 induced significantly more IL-2 (p=0.009). In addition, peptide 46–70 stimulated the production of higher amounts of IFN-γ and IL-2 compared to peptide 50–70 (Fig. 2), showing the importance of flanking residues for the induction of an optimal T-cell response. Of note, peptide 305–322, indicated as good binder to DR*0401 by TEPITOPE
(Table 1), did not bind in our assay (Fig. 1) nor did it induce a T-cell response in DR*0401-Tg mice (Fig. 2). Therefore, this peptide was not selected for analysis in RA patients. In conclusion, the four best binders to DR*0401, as determined by binding assays and TEPITOPE program (Table 1), were also the most
immunogenic ones in DR*0401-Tg mice (Fig. 2 and Table 1). We next assessed the potential of the selected peptides to induce production of IFN-γ in PBMC from RA patients and healthy individuals. Freshly isolated PBMC from 33 RA patients and 16 healthy controls were cultured with 13 individual hnRNP-A2 peptides (indicated in bold in Table selleck inhibitor 1) in ELISPOT plates pre-coated with an anti-IFN-γ mAb. In this assay, PBMC from 6 out of 33 (18%) patients showed an IFN-γ response to hnRNP-A2 peptides, five of them (15%) to a main determinant contained in peptide 117–133 (Fig. 3 and Table 2). The mean frequency of IFN-γ-producing cells specific for this dominant epitope was 21±9 out of 106 cells (mean/duplicate for each patient: 25, 15, 11, 20, 39, 15) compared to 2±2 out of 106 cells (3, 0, 0, 4, 5, 0) PAK5 for the medium background. Remarkably, when retesting two of the six reactive patients 3 months after the first evaluation, the T-cell response to the peptide was sustained (Fig. 3 and Table 2). Conversely, PBMC from none of the healthy individuals reacted to hnRNP-A2
peptides (Table 2). Of note, T-cell reactivity to hnRNP-A2 peptides was independent of disease duration, which varied between 3 and 14 years, and immunosuppressive medication (Table 2 and Supporting Information Table 1). Importantly, all six patients with peptide reactivity presented with active disease (DAS28 > 3.2), and four out of five had bone erosions. We next thought to confirm these findings, to show that the responses to peptides 117/120–133 are mediated by CD4+ T cells, and to investigate whether they are selectively found in RA patients. To demonstrate MHC class II restriction, we incubated the cells with an anti-class II Ab together with peptides 117/120–133 and analyzed the proliferative response in 25 additional RA and 28 disease control (DC) patients with osteoarthritis (Supporting Information Table 2).