Identification and analysis of multivalent proteolytically resistant peptides from gluten: implications for celiac sprue

Abstract

Dietary gluten proteins from wheat, rye, and barley are the primary triggers for the immuno-pathogenesis of Celiac Sprue, a widespread immune disease of the small intestine. Recent molecular and structural analyses of representative gluten proteins, most notably alpha- and gamma-gliadin proteins from wheat, have improved our understanding of these pathogenic mechanisms. In particular, based on the properties of a 33-mer peptide, generated from alpha-gliadin under physiological conditions, a link between digestive resistance and inflammatory character of gluten has been proposed. Here, we report three lines of investigation in support of this hypothesis. First, biochemical and immunological analysis of deletion mutants of alpha-2 gliadin confirmed that the DQ2 restricted T cell response to the alpha-2 gliadin are directed toward the epitopes clustered within the 33-mer. Second, proteolytic analysis of a representative gamma-gliadin led to the identification of another multivalent 26-mer peptide that was also resistant to further gastric, pancreatic and intestinal brush border degradation, and was a good substrate of human transglutaminase 2 (TG2). Analogous to the 33-mer, the synthetic 26-mer peptide displayed markedly enhanced T cell antigenicity compared to monovalent control peptides. Finally, in silico analysis of the gluten proteome led to the identification of at least 60 putative peptides that share the common characteristics of the 33-mer and the 26-mer peptides. Together, these results highlight the pivotal role of physiologically generated, proteolytically stable, TG2-reactive, multivalent peptides in the immune response to dietary gluten in Celiac Sprue patients. Prolyl endopeptidase treatment was shown to abolish the antigenicity of both the 33-mer and the 26-mer peptides, and was also predicted to have comparable effects on other proline-rich putatively immunotoxic peptides identified from other polypeptides within the gluten proteome.

Figure 1

Protein sequences of the α-2 gliadin deletion mutants. (A) Deletion mutant A1234. (B) Deletion mutant A1278. The peptide sequence in red is the proteolytically resistant 33-mer.

Figure 2

Gastric/pancreatic proteolysis of the wild-type α-2 gliadin (A) and its deletion mutant α1234 (B), illustrated by HPLC with UV detection at 215 nm (black traces) and 280 nm (magenta traces). The 33-mer resulted from the wildtype digestion eluted at a retention time of approximately 16 minutes, which is missing in the deletion mutant digestion.

Figure 3

Stimulation of four HLA-DQ2 restricted T cell lines derived from biopsies of Celiac Sprue patients by the PTCEC (pepsin, trypsin, chymotrypsin, elastase, carboxypeptidase B) and brush-border membranedigestive peptide mixtures from α-2 gliadin deletion mutants α-1234 and α-1278. No T cell stimulation was elicited. Error bars indicate the observed range within the triplicates.

Figure 4

Peptidase catalyzed breakdown of α-2 gliadin. The results of T cell line TCL 422.02.4.2 are shown and similar results were obtained with the T cell lines TCL432.1.4, TCL462.1 and TCL480.1. (A) Brush border membrane (BBM) treatment of α-2 gliadin did not result in efficient antigenic detoxification; (B) BBM and PEP treatment of α-2 gliadin efficiently abrogated T cell responses.

Figure 5

LC-MS analysis of γ-5 gliadin proteolysis. The gliadin protein was treated with pepsin, trypsin, chymotrypsin, elastase, and carboxypeptidase. The blue- and red- colored peptides in the protein sequence are found present in the product mixture.

Figure 6

Known γ-gliadin epitopes are clustered in the 26-mer.

Figure 7

T cell stimulation assay of the 26-mer fragment from γ-5 gliadin. T cell clones were tested with the 26-mer and control shorter peptides. (A) TCC387.3 (reactive to the DQ2-γ-VI epitope) was tested with TG2-treated 26-mer, TG2-treated LQPQQPFPQQPQQPYPQQPQ (“20mer”);, and the reference mono-epitope peptide PEQPFPEQPEQ (“11mer”), (B) TCC430.1.134 (reactive to the DQ2-γ-III epitope) was tested with TG2-treated 26-mer, TG2-treated LQPQQPFPQQPQQPYPQQPQ (“20mer”), and the reference mono-epitope peptide FPEQPEQPYPEQ (“12mer”); (C) TCC387.19 (reactive to DQ2-γ-VII epitope) was tested with the TG2-treated 26-mer and TG2-treated QFPQTQQPQQPFPQPQQTFP (“20mer”); (D) TCC430.1.134 (reactive to DQ2-γ-III) was stimulated with TG2-treated 26-mer, with and without rat intestinal brush border membrane (BBM) and prolyl endopeptidase (PEP) incubation. Numbers in the brackets refer to the positioning of the peptides within the γ-5 gliadin. * This peptide is derived from the 1507333A γ-gliadin (M36999) and identical to sequence 61–80.

Graphic