Background: Exposure to the antibiotic amoxicillin is associated with the development of idiosyncratic adverse drug reactions. Amoxicillin-specific T-cells have been detected in patients with severe liver injury, suggestive of an immune disease aetiology. This theory has gained credence with the discovery of a number genetic associations including the HLA DRB1*15:01-DQB1*06:02 haplotype. β-lactam antibiotics form drug-protein adducts by conjugating with lysine residues that are postulated to activate T-cells, following protein processing and the liberation of peptide epitopes. In vitro studies have characterised drug-modified proteins in patient sera, however, the exact epitopes being presented on T-cells have not yet been identified. This study involves a parallel approach to identify the nature of class II epitopes including a designer drug-modified peptide methodology and an interrogation of the naturally eluted immunopeptidome. T-cell models were employed to assess the immunogenicity of candidate drug-associated antigens.
Methods & Results: Designer peptides were synthesised containing (1) anchors for the HLA DRB1*15:01-DQB1*06:02 haplotype and (2) amoxicillin-bound lysine residues in several locations representative of possible drug-TCR contact sites. The immunoaffinity capture of MHC class II antigens was carried out on B-cell lines derived from healthy volunteers homozygous for the risk alleles. Both designer peptides and MHC eluted peptides were purified using HPLC and characterised using proteomics/ immunopeptidomics methodologies on a TripleTOF 6600 mass spectrometer (AB Sciex). Candidate peptides were incubated with PBMCs from hypersensitive patients positive for the risk haplotype to generate antigen-responsive T-cell lines. Amoxicillin-modified peptide-specific T-cells proliferated and secreted cytokines such as IFNγ in a dose dependent manner with high specificity to the antigen showing no cross reactivity with unmodified peptides or with positional derivatives at different TCR contact sites. T-cell responses were restricted to the HLA DRB1*15:01-DQB1*06:02 risk alleles.
Conclusion: Here we demonstrate an approach to elucidate and validate drug-associated-antigens which may be implicated in drug hypersensitivity reactions.