Splicing of antigenic peptides by the proteasome

(eng) Antigens presented by MHC class I molecules to CD8+ T lymphocytes consist of peptides of 8-11 amino acids, corresponding to continuous fragments of intracellular proteins. However, some antigenic peptides were found to be created by the splicing of two distinct fragments of the respective parental proteins. Peptide splicing represents a newly described mode of production of antigenic peptides whereby two noncontiguous peptide fragments are joined together after the excision of an intervening segment. Three spliced antigenic peptides have been described. The first example is derived from fibroblast growth factor-5 (FGF-5) and is made up of two fragments of five and four residues. Another is a peptide produced from melanosomal protein gp100 by splicing of two fragments of three and six residues. The third example is a human minor histocompatibility antigenic peptide created by a polymorphism in the SP110 gene. This epitope is composed of two noncontiguous fragments of four and six residues which are spliced together in the reverse order to that in which they appear in the parental protein. In the two latter cases, splicing involves the excision of a short intervening segment of four or six residues and was shown to occur in the proteasome by transpeptidation resulting from the nucleophilic attack of an acyl-enzyme intermediate by the N terminus of a peptide fragment present at the active site. In a first part of this work, we studied the mechanism of production of the FGF-5 spliced peptide, for which the length of the intervening segment (40 aa) is much longer than the two other described examples. We confirmed that the proteasome also produces this spliced peptide by transpeptidation and we showed that reducing the length of the intervening segment increased the production of the spliced peptide. We also tested the possibility of trans-splicing (i.e., splicing of fragments from two distinct substrates) and observed that this event can occur in the cell between two peptide precursors with the same efficiency as cis-splicing. However, trans-splicing between two distinct proteins was far less efficient than splicing of peptide fragments from a single protein substrate and unlikely contributes to the production of the spliced peptides derived from FGF-5 and gp100. The production of the three spliced peptides by the standard proteasome and the immunoproteasome was then compared in details, using a cellular approach and an in vitro approach. We showed that both proteasome types can produce spliced peptides although they differ in their efficiency of production of each peptide. The FGF-5 and gp100 peptides are produced more efficiently by the standard proteasome, while the SP110 peptide is better produced by the immunoproteasome. By showing that splicing depends on the efficiency of production of the splicing partners, these results support the transpeptidation model of peptide splicing. In collaboration with Dr. Paul Robbins from the National Cancer Institute in Bethesda, we have identified a new spliced antigenic peptide, which is derived from the melanosomal protein tyrosinase. This original peptide is composed of two spliced and rearranged peptide fragments, each one containing an aspartate residue resulting from asparagine deamidation. The processing of this antigenic peptide involves retrotranslocation of tyrosinase molecules from the endoplasmic reticulum (ER) to the cytoplasm, deamidation of presumably glycosylated asparagines by peptide-N-glycanase, cleavage and splicing by the standard proteasome, and transport of the resulting peptide into the ER via the TAP transporter. The potential role of an additional, yet unidentified, protease during this process is also discussed in this work.