Pharmacological Perturbation of the Immunoproteasome in Hematologic Neoplasias: Therapeutic Implications

Pharmacological inhibition of the canonical proteasome by inhibitors targeting its beta5 (PSMB5) subunit represent a cornerstone of the therapeutic management of plasma cell neoplasias. However, proteasome inhibitors have limited clinical applications in other hematologic malignancies. Notably, a broad spectrum of cell lines and patient-derived samples from lymphoma, leukemia and myeloma tend to express high levels of transcripts for and exhibit substantial activity of the immunoproteasome, an alternative proteasome particle which is enriched in hematopoietic/immune tissues. Established proteasome inhibitors used in clinical practice (e.g. bortezomib, carfilzomib, ixazomib) target both the canonical 20S proteasome and the immunoproteasome. We reasoned though that targeting of the immunoproteasome with more selective small molecule inhibitors that largely spare the canonical proteasome may provide insights into the specific functional role of the immunoproteasome in hematologic neoplasias, and importantly may also uncover underappreciated therapeutic opportunities related to targeting of hematologic neoplasias beyond myeloma, that might be more dependent on the immunoproteasome vs. the canonical proteasome; as well as potential benefits related to the therapeutic index of these agents. To address these questions, we characterized the activity of the previously disclosed selective immunoproteasome inhibitor M3258, which targets the LMP7 (beta5i subunit). The bortezomib-resistant cell line MM1VDR, which has a "gatekeeper" PSMB5 mutation preventing bortezomib binding to the canonical proteasome, remained sensitive to in vitro treatment with M3258, in further support of the functional specificity of this inhibitor. To obtain a comprehensive assessment of the activity of M3258 across hematologic malignancies, we studied a pool of 74 "DNA-barcoded" cell lines (PRISM system) from hematologic malignancies, including multiple myeloma (MM), leukemias and lymphomas, which were treated for 48hr with M3258. MM cell lines (9/16) were strongly represented among the sensitive lines which displayed <50% relative cell viability (compared to DMSO control) after M3258 treatment. Notably, 15 of 58 non-MM lines also exhibited >50% reduction in their cell viability in this analysis; and their responsiveness did not differ significantly from the M3258-responsive MM lines. Importantly, we observed no correlation between the responses of leukemia, lymphoma or MM lines to in vitro treatment with M3258 vs. clinically-relevant pulse of bortezomib. Furthermore, 7 of the 15 lymphoma/leukemia lines that were M3258-responsive had limited, if any, response to a clinically-relevant pulse of bortezomib. Taken together, these observations suggest that immunoproteasome inhibition with M3258 can be active against subsets of hematologic neoplasias with resistance/limited responsiveness to canonical proteasome inhibitors. To further examine the in vivo relevance of these observations, we studied the in vivo activity of M3258 against the MM.1S myeloma cell line, which has only moderate in vitro response to M3258 compared to other blood cancer cell lines tested in our "DNA-barcoded" pooled analysis. M3258 significantly increased the overall survival of mice with diffuse MM-bone lesions established after tail vein injection of MM.1S cells; and of mice harboring MM.1S lesions within a subcutaneously implanted ceramic scaffold-based system engineered to establish a bone marrow-like histological and functional unit with a "humanized" stromal compartment. The anti-tumor activity of M3258 in this scaffold model exhibited a trend for increased efficacy compared to twice weekly subcutaneous treatments with bortezomib. These results highlight that in vivo administration of M3258 exhibits anti-tumor activity against clinically-relevant models of MM lesions, even from cell lines that have modest in vitro responsiveness to this immunoproteasome inhibitor. Furthermore, our studies with a large panel of "DNA-barcoded" cell lines indicates that the anti-tumor effects of M3258 may extend beyond MM and into other classes of hematologic malignancies, including subsets of leukemias and lymphomas, identifying previously underappreciated therapeutic implications for the class of selective immunoproteasome inhibitors.