Abstract
(English)
|
Identification of T cell epitopes by combining in vitro proteasome digestion and peptide predictionFrom a given amino acid sequence, one can predict MHC class I binding peptides by computerized algorithms. For this we used the 'syfpeithi' software (http://www.syfpeithi.de) established by our collaborators from Tübingen (Prof. H.-G. Rammensee et al.). However, only some of the peptides predicted to bind to MHC class I molecules are in fact generated by the cellular processing machinery. A major reason for this is that proteasomes digest proteins by specific cleaving, and therefore only some but not all theoretical possible peptides are generated from a given protein. Using proteasomes from human erythrocytes, we performed in vitro digestions of overlapping 22 amino acid long peptides covering the whole sequence of the HOM-Mel 40 / SSX-2 protein. The digestion products were analyzed by mass spectrometry to determine their identity and thus the cleavage sites. These digestion results, combined with the above mentioned peptide prediction approach based on MHC class I binding motifs for HLA-A1, -A2, -A3, -B7, -B27, -B35 were used to define potential MHC-I restricted peptides. Candidate peptides were synthesized and tested in functional T lymphocyte assays, in order to see whether these peptides are indeed relevant T cell epitopes.An additional key element was that tumor infiltrating lymphocytes (TIL) from melanoma patients were used to screen the candidate peptides. The TIL from one patient specifically recognized to the SSX-2 peptide 41-49 (amino acid sequence KASEKIFYV). Specific CD8+ T cells were cloned, allowing to determine that very low concentrations of peptides (in the range of 10 picomolar) mediated target cell recognition. Also, the clones efficiently recognized and killed melanoma cells, which was entirely dependent on their expression of HLA-A*0201 and SSX-2. Thus, the SSX-2 peptide 41-49 is one of the still relatively rare CD8 T cell epitopes that mediates efficient tumor cell lysis. The fact that SSX-2 is absent in all healthy tissues (except on HLA negative cells from the testis) but expressed by a fraction of melanomas, lymphomas, and cancers of colon and head & neck makes this epitope a very attractive candidate to study T cell responses in a wider range of patients, and to envisage new approaches for immunotherapy.Our study is published by Ayyoub et.al. in The Journal of Immunology, 2002, 168: 1717 - 1722.
|
