Defining the Most Effective Proteasome Inhibition in Multiple Myeloma: Head-to-Head Comparison of Currently Available Proteasome Inhibitors

Introduction: Proteasome inhibitors (PIs) are a backbone of multiple myeloma (MM) therapy. The proteasome harbors six proteolytically active subunits (β1, β2 and β5 of the constitutive (c) and the immunoproteasome (i), respectively). They differ in substrate preferences and affinity for PIs. β5c was identified as the rate-limiting proteolytic proteasome activity, while β2 and β1 activity received little attention. Consecutively, all proteasome-inhibiting drug candidates (bortezomib, carfilzomib, ixazomib, oprozomib, delanzomib, marizomib) by design target β5c, but differ in co-inhibition of non-β5 activities. Current chemical biology tools allow highly selective inhibition and visualization of individual proteasome subunits in living cells. We lack data that identify the most effective pattern of β5 and non-β5 type of proteasome inhibition in MM cells, as well as a head to head comparison of PIs and drug candidates, and their proteasome subunit inhibition profile in relation to pharmacodynamic activity. Such data shall allow rational selection and positioning of proteasome-inhibiting drugs.

Methods: Head to head comparison of bortezomib, carfilzomib, ixazomib, oprozomib, delanzomib and marizomib was performed in MM cells. Non-commercial, highly subunit-selective PIs and activity-based fluorescent chemical probes (ABP) were used to assess the most effective pattern of inhibition of proteasome subunits, in conjunction with a fluorescent ubiquitin Ub-^G67V-GFP construct, and related to the viability of MM.

Results: All commercial PI tested target primarily β5 activity and in higher concentrations co-targeted β2 and/or β1 (Table 1). In equimolar comparison, most effective PI was marizomib, followed by bortezomib, delanzomib, carfilzomib, ixazomib and oprozomib. MM cell viability inversely correlated with functional proteasome inhibition. Surprisingly, during short term, 1h pulse exposure mimicking the pharmacokinetics of i.v. proteasome inhibiting drugs, β5-selective proteasome inhibition was not sufficient to functionally inhibit protein degradation or induce cytotoxicity, even in PI-sensitive MM cells. Only co-inhibition β5 with either β1 or β2 resulted in accumulation of polyubiquitinated protein and decreased MM cell viability. Long time (48h) continuous β5-selective inhibition likewise induced cell death in PI-sensitive MM, however, only co-inhibition of β5 and β2 activity induced cytotoxicity in PI-resistant MM.

Conclusion: All commercial PIs by design target β5c activity and differ in their co-targeting of non-β5 activities. Co-inhibition of β1 and β2 determines cytotoxicity. Selective inhibition of β5 by pulse exposure is not sufficient, but co-inhibition of β5 and β2 is the most effective combination to induce cytotoxicity, in particularly in PI resistant cells. This inhibition pattern is provided only by marizomib and high dose carfilzomib. By contrast, low dose of carfilzomib does not deliver co-inhibition of β2 and therefore has no advantage over bortezomib. Together, these results give a rational basis to select individual proteasome inhibitors for the treatment of MM: cytotoxicity in PI-sensitive MM is achieved by co-targeting of β5 together with β1 or β2 with a pulse treatment, or by prolonged β5 inhibition (s.c. bortezomib, orally administered ixazomib). In PI-resistant MM only dual β2/β5 inhibition is able to achieve meaningful cytotoxicity, which is provided by carfilzomib or marizomib.

Table 1

View largeDownload slide

Table 1

View largeDownload slide

Close modal