Abstract:
This thesis describes the development and evaluation of novel T cell-independent strategies for the identification of disease-associated MHC-ligands by mass spectrometry. The 'Predict, Calibrate and Detect' (PCD)-approach, a combination of epitope prediction and HPLC-mass spectrometry, allowed the identification of new HLA-A2-ligands from the tumor-associated proteins CEA, p53 and MAGE-A1 with a sensitivity of <10 copies/cell. Moreover, for CEA694-702 presentation on a solid gastrointestinal carcinoma could be shown. MAGE-A1278-286 could be verified as a CTL-epitope in transgenic mice, while the p53 and CEA-derived ligands are still under investigation. These peptides are candidates for the peptide-specific immunotherapy of tumors.
Allel-specific peptide motifs are the main prerequisite for the PCD-approach as well as for other strategies that employ epitope prediction for the identification of CTL-epitopes from defined antigens. The determination of the peptide motifs of HLA-B*1510 and HLA-DQ(a1*0201/b1*0202) has paved the way for epitope prediction for these alleles.
Moreover, the first Borna Disease Virus-derived CTL-epitope could be characterized. which also represents the first known virus-derived MHC-ligands from the rat.
Furthermore, MHC class II-ligands from the diabetes-relevant autoantigen GAD65 could be identified by nanoelectrospray MSMS analysis, using a novel approach for the specific detection of MHC-ligands from a defined antigen by radioactive labeling of the protein.
Finally, nanoelectrospray MSMS analysis allowed the identification of the cytosolic proteases puromycin-sensitive aminopeptidase and bleomycin hydrolase, that are probably involved in N-terminal trimming of proteasomal digestion products and might play an important role in the generation of MHC class I-ligands.