Proteasome Inhibitors Do Not Inhibit the Serine Protease HtrA2/Omi

Abstract 5023

Based on unprecedented efficacy, the proteasome inhibitor (PI) bortezomib has become the cornerstone of multiple myeloma treatment. Nevertheless, in a subset of patients bortezomib causes painful peripheral neuropathy and this side effect can limit its potential benefit for those patients. Although the mechanism of bortezomib associated neuropathy is unknown, we have previously suggested that it is related to the mechanism of action (Csizmadia at. al. Vet Pathol 2010; 47:358–367.). Recently Arastu-Kapur et al. (Clin Cancer Res 2011;17:2734–2743.) have reported that the serine protease HtrA2/Omi was inhibited by bortezomib (a peptide boronate proteasome inhibitor) and not by carfilzomib (an epoxyketone proteasome inhibitor). Further, since HtrA2/Omi is involved in neuronal survival (Martins et al. Mol Cell Biol 2004; 24: 9848–9862.) they suggested that this off target inhibition by bortezomib could be the mechanism underlying bortezomib associated peripheral neuropathy. To confirm and extend these published results, we investigated the effects of these two PIs on HtrA2/Omi activity in recombinant enzyme assays, in SH-SY5Y neuroblastoma-, and wild type and HtrA2/Omi double negative mouse embryonic fibroblast cells (MEF). In contrast to the results of Arastu-Kapur et al., our results clearly demonstrated that neither bortezomib nor carfilzomib inhibits HtrA2/Omi in recombinant enzyme assays at concentrations up to 100μM. As a positive control we used Ucf-101 an HtrA2/Omi specific inhibitor (Cilenti et al. J Biol Chem 2003; 278:11489–11494.) which in our assay behaved in a manner consistent with the published literature. Similarly, in MEF cells, only Ucf-101 prevented the degradation of validated HtrA2/Omi substrates eIF4G1 and UCH-L1, while neither bortezomib nor carfilzomib prevented the degradation of these two substrates. In conclusion, we have assessed the protease activity of HtrA2/Omi both in vitro with purified enzyme and in cultured cells and we find that neither PI inhibits this protease. Therefore we think it is unlikely that PI associated peripheral neuropathy is caused by off target inhibition of HtrA2/Omi. Further research is needed to understand the side effects of PIs.