The biochemical and cell biological basis of GRP94/GP96-mediated tumor rejection: Investigating the peptide binding hypothesis and its alternatives

Vaccination of mice with GRP94/gp96, the endoplasmic reticulum (ER) Hsp90, results in tumor rejection and suppression of metastases; this effect is greatest when GRP94 is purified from the same tumor as that used in subsequent challenge. Two complementary hypotheses have been advanced to account for this tumor-rejection activity. The peptide binding hypothesis suggests that GRP94 binds antigenic peptides in the ER of tumor cells. When GRP94 is released from tumor cells, it is taken up by antigen-presenting cells (APC), and the peptides are represented on MHC class I molecules of the APC. According to the adjuvant hypothesis, GRP94 and other chaperones directly stimulate APC to release cytokines and produce nitric oxide, and initiate dendritic cell maturation.;This thesis examines the basis of GRP94's tumor rejection activity. We critically examine the peptide binding and adjuvant hypotheses, and advance evidence for two alternative mechanisms. Evidence reported here, as well as recent biochemical and immunological findings by others, indicate that GRP94 may not bind and transfer peptides. We observe no detectable peptide binding by GRP94 translated in vitro, and no evidence GRP94-mediated peptide re-presentation in a system of vaccinia virus-driven peptide expression.;We propose the bystander hypothesis, based on the finding that GRP94 copurifies with traces of several enzymologically active proteins. Highly purified preparations of GRP94 contain immunologically significant levels of beta-galactosidase, aminopeptidase and kinase; we suggest that copurifying (poly)peptides ("bystanders"), rather than bound peptides, account for the specificity observed in GRP94-elicited tumor rejection.;The endotoxin hypothesis is advanced based on the finding that many of the adjuvant-like activities attributed to GRP94 are in fact due to lipopolysaccharide (LPS, endotoxin) contamination. Several signaling pathways reported to be stimulated by GRP94, including NF-kappaB activation, nitric oxide production, and p38 phosphorylation, are not activated by low-endotoxin GRP94, but are robustly activated by LPS. However, GRP94 causes rapid phosphorylation of ERK at low concentrations, indicating that it possesses endotoxin-independent cell signaling effects.;These findings imply that GRP94-based immunotherapy may not require custom, tumor-derived GRP94, the method currently under investigation. If GRP94 works via ERK and other pathways, recombinant protein may be equally effective in tumor suppression.