Anti-Tumor Activity of Immunoproteasome Selective Inhibitors.

Background: The immunoproteasome is a form of the proteasome that is distinct from the broadly expressed constitutive proteasome. The immunoproteasome, which is predominant in hematopoetic cells, contains the unique active site subunits, LMP7, LMP2 and MECL1, accounting for the chymotrypsin-like, caspase-like and trypsin-like activity respectively. The proteasome inhibitors bortezomib and carfilzomib have validated the proteasome as a therapeutic target in hematologic malignancies; however these inhibitors target both the constitutive and immunoproteasome. We hypothesized that selective inhibition of the immunoproteasome would have anti-tumor activity on hematologic tumors and may avoid toxicities associated with constitutive proteasome inhibition.

Methods: The level of immunoproteasome expression in hematologic malignant cells was not known. Hematologic tumor cell lines and primary patient samples were assessed for relative levels of immunoproteasome versus constitutive proteasome using an ELISA that quantitates proteasomal subunits. Epoxyketone-based inhibitors were generated that are >20-fold selective for the immunoproteasome subunit, LMP7, over the corresponding constitutive subunit, Beta5. Anti-tumor activity of LMP7 selective inhibitors was determined by treating tumor cell lines and primary patient cells that are predominately LMP7 expressing. Furthermore, anti-tumor effects were assessed by combining LMP7 and Beta5 inhibition using genetic and chemical methods Results: Immunoproteasome expression was variable in primary patient samples and tumor cell lines. In the majority of acute myeloid leukemia, acute lymphocytic leukemia, and chronic lymphocytic leukemia patient samples, the immunoproteasome accounted for >95% of the total proteasome. In HS-Sultan, Molt4 and RL cell lines, the immunoproteasome accounted for 70%, 50% and 50% of the total proteasome, respectively. In these cell lines, complete inhibition of LMP7 was not sufficient to induce to cell death. LMP7 selective inhibitors were only cytotoxic in these cell lines under conditions where Beta5 was depleted either genetically or chemically. Genetic depletion was accomplished using Beta5-specific shRNA expression in HS-Sultan, Molt4 and RL cells which increased the relative LMP7 levels to 95%, 80% and 70% respectively. These cells were significantly more sensitive to LMP7 inhibition than their parental cells, suggesting that cell kill requires suppression of LMP7 and Beta5. Chemically, combinations of Beta5 and LMP7 selective inhibitors showed similar results, i.e. greater cytotoxicity in combination than either inhibitor alone. Acute lymphocytic leukemia patient samples were screened for LMP7 expression and anti-tumor activity of the LMP7 selective inhibitors correlated with the relative levels of LMP7 expression.

Conclusion: This work suggests that tumor cells with high relative levels of LMP7 will be sensitive to LMP7 selective inhibition. These results support the development of immunoproteasome-selective inhibitors for the treatment of hematologic malignancies where high levels of immunoproteasome expression are observed. An LMP7 selective inhibitor has been identified as a clinical candidate for development in leukemia.