Penicillin is arguably one of the most successful antibiotics, however a subset of patients experience hypersensitivity reactions towards penicillin and other beta-lactam antibiotics, with drug-specific T cells implicated as mediators 1. It is thought that penicillin facilitates the covalent modification of peptide antigens, resulting in the generation of novel haptenated epitopes that trigger drug-specific T cell responses 2. As T cells are activated upon recognition of such peptides presented by human leukocyte antigens (HLAs), much effort has been placed into identifying penicillin-modified ligands that form the epitopes of penicillin-specific T cells. Upon identifying such ligands, we will be able to understand the interactions made between penicillin-specific T cells and the drug-modified HLA ligands. Due to the association noted between HLA-A*02:01 and the onset of drug-induced liver injury 3, we sought to understand how beta-lactam antibiotics might influence the HLA-A*02:01 immunopeptidome using the model drug benzylpenicillin (BP).
Lysine modification was previously reported as the primary site of covalent modification by penicillins, however we found large numbers of spectra containing BP diagnostic ions could not be explained by lysine haptenation. Instead, we observed that penicillin-haptenation occurred primarily on cysteine residues that had previously been modified by cysteinylation (76.4%). Synthetic versions of experimentally identified CysBP-haptenated peptides fragmented identically by CID (PCC 0.85-0.97), confirming sequence assignments and drug modifications. We further demonstrated that modification of the cysteinyl post-translational modification was employed by other beta-lactam antibiotics such as amoxicillin and flucloxacillin. Haptenation of cysteinylated cystines also occurred within proteins in lysates of antigen-presenting cells, providing a source of BP-modified cysteinylated ligands for HLA presentation. Importantly, a patient-derived BP-specific T cell receptor recognised an epitope that was destroyed by reduction with TCEP, consistent with a BP-cysteinylated cysteine modification.
This is the first study to report drug-modified HLA-A*02:01 ligands. It highlights a novel target of penicillin (cysteinylated cysteines) which has been overlooked since haptenation was proposed in 1936 4. This will have major implications for deciphering the epitopes of T cells within the penicillin-induced immunopeptidome, and have the potential to be implemented in screening tools (e.g. peptide/HLA tetramers) to prevent the onset of the wider sphere of allergies induced by beta-lactam antibiotics.