Re-evaluating the generation of a "proteasome-independent" MHC class I-restricted CD8 T cell epitope.

The proteasome is primarily responsible for the generation of MHC class I-restricted CTL epitopes. However, some epitopes, such as NP(147-155) of the influenza nucleoprotein (NP), are presented efficiently in the presence of proteasome inhibitors. The pathways used to generate such apparently "proteasome-independent" epitopes remain poorly defined. We have examined the generation of NP(147-155) and a second proteasome-dependent NP epitope, NP(50-57), using cells adapted to growth in the presence of proteasome inhibitors and also through protease overexpression. We observed that: 1) Ag processing and presentation proceeds in proteasome-inhibitor adapted cells but may become more dependent, at least in part, on nonproteasomal protease(s), 2) tripeptidyl peptidase II does not substitute for the proteasome in the generation of NP(147-155), 3) overexpression of leucine aminopeptidase, thymet oligopeptidase, puromycin-sensitive aminopeptidase, and bleomycin hydrolase, has little impact on the processing and presentation of NP(50-57) or NP(147-155), and 4) proteasome-inhibitor treatment altered the specificity of substrate cleavage by the proteasome using cell-free digests favoring NP(147-155) epitope preservation. Based on these results, we propose a central role for the proteasome in epitope generation even in the presence of proteasome inhibitors, although such inhibitors will likely alter cleavage patterns and may increase the dependence of the processing pathway on postproteasomal enzymes.

The impact of misfolding versus targeted degradation on the efficiency of the MHC class I-restricted antigen processing.

Evidence suggests that most epitopes presented by MHC class I molecules are derived from those newly synthesized proteins that are defective due to errors during manufacture. We examined epitope production from model cytosolic and exocytic proteins modified in various ways. Substrates containing a degradation targeting sequence demonstrated very rapid turnover and enhanced epitope production, as was the case for substrate retargeted from endoplasmic reticulum to cytosol. For less radical alterations, including point mutation and deletion and elimination of glycosylation sites, despite detectable changes in folding, half-life was only moderately decreased and there were no significant increases in epitope production. Puromycin, which causes premature termination of protein synthesis, also had no impact upon epitope production. It appears that most defective proteins are not rapidly dispensed with and the targeting of most nascent proteins for Ag processing is not tied to quality control.