Mucin 20 (MUC20) Modulates Proteasome Assembly Chaperones through the c-MET Pathway and Is a Biomarker of Proteasome Inhibitor Sensitivity in Myeloma

Background: Modulation of the activity of the ubiquitin-proteasome pathway with the proteasome inhibitor (PI) bortezomib is an established component of therapy for plasma cell disorders. More recently, the second-generation, irreversible proteasome inhibitor carfilzomib has been validated both pre-clinically and clinically, which can overcome bortezomib resistance in some patients. However, resistance emerges to carfilzomib as well, and the emergence of PI resistance in general is incompletely understood. Therefore, there is an urgent need to identify these mechanisms and develop biomarkers that could help guide therapy, and provide novel targets to overcome resistance.

Methods: We generated carfilzomib-resistant (CR) myeloma cell lines by exposing drug-naive ANBL-6, KAS-6/1, U266, and OPM-2 cells to increasing concentrations of carfilzomib. These were then subjected to gene expression profiling (GEP) to identify prominent changes compared to their vehicle-treated counterparts. Molecular and pharmacologic approaches were then pursued to probe the significance of the observed changes to PI resistance in myeloma cell lines, murine models, and clinically annotated GEP databases.

Results: Profiling and pathway analysis of CR myeloma cell lines identified suppressed expression of MUC20 as the most conserved and significant change compared to drug-naïve cells. Decreased levels of MUC20 in CR cells were confirmed by quantitative PCR and Western blotting, and also identified in bortezomib-resistant (BR) cell lines. Moreover, suppression of MUC20 with shRNAs was itself sufficient to confer carfilzomib resistance, which was associated with an increase in the proteasome chymotrypsin-like (ChT-L) activity. Reduced expression of MUC20 led to increased phosphorylation and activation of c-MET, STAT3, and ERK-1/2. In addition, exposure of drug-naïve cells to the c-MET ligand HGF induced activation of this pathway and of the proteasome ChT-L activity, but reduced carfilzomib sensitivity. Conversely, inhibition of c-MET with INCB28060, ARQ-197, or XL-184 restored carfilzomib sensitivity to the CR cell lines in culture, and ARQ-197 overcame resistance in a murine myeloma model. To explore the mechanisms behind the increase in ChT-L activity, we evaluated expression of proteasome components, and while 20S subunits were unchanged, Proteasome maturation protein (POMP), a key proteasome assembly chaperone, was more abundant. The POMP promoter has a consensus ETS-Like Gene 1 (ELK1) binding site, and ELK1 is a target for ERK-1/2. Therefore, we performed chromatin immunoprecipitation and gel mobility shift studies, which documented ELK1 binding, while mutation of this site reduced POMP expression in a reporter assay. Finally, analysis of the Millennium Pharmaceuticals GEP database of patients treated with bortezomib in the relapsed and relapsed/refractory settings showed patients who had higher expression of MUC20 had superior overall and progression-free survival compared to those who had lower expression.

Conclusions: Taken together, our data support a role for signaling through the c-MET/ERK-1/2/ELK1/POMP axis in enhancing proteasome assembly and capacity, thereby reducing sensitivity to proteasome inhibitors like carfilzomib or bortezomib in myeloma. Also, they validate use of drugs suppressing c-MET as a potentially attractive strategy to overcome resistance to carfilzomib in the clinic. Finally, they indicate that MUC20 expression may be a viable biomarker to differentiate patients who have proteasome inhibitor-sensitive or relatively –resistant disease.