Intestinal T cell responses to gluten peptides are largely heterogeneous: implications for a peptide-based therapy in celiac disease

Abstract

The identification of gluten peptides eliciting intestinal T cell responses is crucial for the design of a peptide-based immunotherapy in celiac disease (CD). To date, several gluten peptides have been identified to be active in CD. In the present study, we investigated the recognition profile of gluten immunogenic peptides in adult HLA-DQ2(+) celiac patients. Polyclonal, gliadin-reactive T cell lines were generated from jejunal mucosa and assayed for both proliferation and IFN-gamma production in response to 21 peptides from wheat glutenins and alpha-, gamma-, and omega-gliadins. A magnitude analysis of the IFN-gamma responses was performed to assess the hierarchy of peptide potency. Remarkably, 12 of the 14 patients recognized a different array of peptides. All alpha-gliadin stimulatory peptides mapped the 57-89 N-terminal region, thus confirming the relevance of the known polyepitope 33-mer, although it was recognized by only 50% of the patients. By contrast, gamma-gliadin peptides were collectively recognized by the great majority (11 of 14, 78%) of CD volunteers. A 17-mer variant of 33-mer, QLQPFPQPQLPYPQPQP, containing only one copy of DQ2-alpha-I and DQ2-alpha-II epitopes, was as potent as 33-mer in stimulating intestinal T cell responses. A peptide from omega-gliadin, QPQQPFPQPQQPFPWQP, although structurally related to the alpha-gliadin 17-mer, is a distinct epitope and was active in 5 out of 14 patients. In conclusion, these results showed that there is a substantial heterogeneity in intestinal T cell responses to gluten and highlighted the relevance of gamma- and omega-gliadin peptides for CD pathogenesis. Our findings indicated that alpha-gliadin (57-73), gamma-gliadin (139-153), and omega-gliadin (102-118) are the most active gluten peptides in DQ2(+) celiac patients.

FIGURE 1

Recognition profile of peptides from multiple gliadin families. IFN-γ responses of intestinal T cell lines to 21 gluten peptides and to PT-gliadin are illustrated. A, IFN-γ responses of CD patients recognizing peptides derived from all three α-, β-, and γ-gliadins and (B) from two gliadin families (α- and ω-or α- and γ-gliadins) are illustrated. T cells (3 × 10⁴) were stimulated with autologous EBV-B cells (6 × 10⁴) and with TG2-deamidated peptides (30 μg/ml, corresponding to a range of concentration of 8 –19 μ M) or PT-gliadin (50 μg/ml). IFN-γ production was evaluated by ELISA after 48 h of incubation. Results are shown as means ± SD of duplicate wells. Levels of IFN-γ (ng/ml) are referred to 10⁶ T cells/ml. Responses were considered positive (indicated with filled bars) when the peptide-specific IFN-γ level was 2-fold or more than the level of IFN-γ obtained in the absence of peptide. Representative results of three separate experiments performed for each iTCL are illustrated.

FIGURE 2

Recognition profile of peptides from one or none gliadin family. Recognition profile of intestinal T cell lines to 21 gluten peptides and to PT-gliadin are illustrated. A and B, IFN-γ responses of two iTCLs recognizing peptides derived from only one gliadin family are illustrated. C, IFN-γ production and (D) cell proliferation of the iTCLs from patient CD171204 are shown. T cells were stimulated as indicated in Fig. 1. Levels of IFN-γ (ng/ml) are referred to 10⁶ T cells/ml. Positive responses are indicated by the filled bars. Representative results of three separate experiments performed for each iTCL are illustrated.

FIGURE 3

Peptide recognition repertoire and hierarchy. A, Profile of peptide recognition in 14 adult DQ2⁺ CD patients. Filled areas indicate positive IFN-γ responses. B, Peptide hierarchy was assessed by a computational clustering analysis. The intensities of IFN-γ responses elicited by gluten peptides in each iTCL were obtained by calculating the ratio between the optical densities (OD) obtained in response to peptide stimulation and the OD corresponding to medium alone in the ELISA assays. All the peptides were considered except glia-20, α-gliadin 31– 49, glt-156, and glt-19 –39 to which no intestinal T cell response was found in any of the analyzed patients. The scale of IFN-γ intensity is indicated by the different color graduation. Peptides cluster in two groups, as evidenced by branches joining them. The length of branches indicates, in inverse proportion, the similarity of peptide activity.

FIGURE 4

Characterization of an immunogenic ω-gliadin-derived peptide. A, iTCLs from patient CD041051 are assayed for IFN-γ production in response to 10-fold increased concentrations of DQ2-ω-1 peptide. B, Cells from the same patient are stimulated with the DQ2-ω-1 (15 μM) in the presence or absence of anti-HLA-DR and anti-HLA-DQ blocking mAbs (10 μg/ml). C, Alignment of the ω-gliadin peptide DQ2-ω-1 with the α-gliadin 17-mer peptide is illustrated; differences of amino acid residues are delimited by a boxed area. D and E, Analysis of DQ2-ω-1 cross-reactivity on DQ2-α-I/III-specific T cell clones from patient CD230204 is shown. T cells were stimulated with N-terminal α-gliadin peptides and DQ2-ω-1 (each peptide was assayed at 30 μg/ml corresponding to a range of concentrations of 8–19 μM) and assayed as indicated in Fig. 1. Representative IFN-γ and proliferative responses of three separate experiments performed are illustrated.

FIGURE 5

Sequence alignment of 33-mer with its truncated peptides. α-Gliadin peptides were aligned on the basis of their position at the N-terminal region. All 9-mer DQ2-α-I, DQ2-α-II, and DQ2-α-III epitopes contained in each peptide are underlined.

FIGURE 6

Correlation between the stimulatory capacity of N-terminal α-gliadin peptides and the known 9-mer epitopes. Dose-response curves of peptides encompassing the 57–89 region of α-gliadin were performed on both iTCLs and TCCs. Irradiated APCs (autologous B-LCL) were incubated overnight with 10-fold increasing doses of TG2-deamidated peptides before DQ2-α-I/III- or DQ2-α-II-specific T cells were added (3 × 10⁴ for TCLs or 1 × 10⁴ for TCCs). T cell lines and clones were also analyzed in response to DQ2-α-I, -II, and -III 9-mer epitopes (30 μM). A and B, IFN-γ production of DQ2-α-I/III-responsive iTCLs from patient CD230204 is illustrated. C and D, From the same patient, DQ2-α-I/III-specific TCCs were generated and tested for peptide recognition. E and F, Cells from patient CD041051 reacting to DQ2-α-II epitope were stimulated with increasing peptide doses. Values (mean of duplicates) are representative of three separate experiments.

FIGURE 7

The stimulatory activity of 33-mer and of its truncated peptides is processing-independent. A, The proliferative response of CD230204 iTCLs (DQ2-α-I-specific) to 33-mer, 25-mer, 18-mer, 17-mer, and 13-mer peptides of α-gliadins was assessed using both alive or fixed APCs pulsed overnight with peptides (1 μM) or PT-gliadin (50 μg/ml). After a thorough washing, Ag-loaded APCs were incubated with T cells for an additional 48 h. B, Binding to purified DQ2 heterodimers was performed by a competitive inhibitory assay. IC₅₀ indicates the concentration of each analyzed peptide that induced the 50% of inhibition of a probe peptide binding.