Résumé
Proteasome inhibitors (PI) proved to be extremely effective against different types of cancer, particularly against Multiple Myeloma (MM), a frequent and still incurable plasma cell malignancy. Phase II clinical trials showed that more than 50% of MM patients fail to respond to bortezomib, the only PI currently approved for clinical use. However, the mechanisms of action and bases of individual susceptibility to PI remain largely unclear, with no reliable predictor of response identified so far. Recent evidences linking proteasome activity and Ig synthesis to susceptibility to PI suggest that the exquisite sensitivity of MM cells (MMC) to PI might be explained by an imbalance between the efficiency of the ubiquitin (Ub)-proteasome pathway and the demand for proteasome-mediated degradation. We set out to explore this hypothesis both in vitro and ex vivo. To achieve this aim, we employed human MM cell lines characterized by differential apoptotic sensitivity to PI (U266 and RPMI8226, fairly resistant cell lines, versus MM.1S, an extremely sensitive one) and primary, patient derived MMC. In MM cell lines, we found that high apoptotic sensitivity to PI is associated with lower expression of active proteasomes (as assessed by decreased expression of cleaved catalytic subunits and enzymatic assays with fluorogenic substrates in cell extracts), together with higher proteasomal workload (demonstrated by higher proteasome-dependent loss of TCA-insoluble radioactivity in pulse-chase assays). Indeed, MM.1S cells displayed 2–3 times lower proteasomal activity as compared to the more resistant U266 and RPMI8226 cells, both on a per cell basis and upon normalization by protein content. Together with the reduced proteasome capacity, MM.1S cells showed a consistently higher production of client proteins for the Ub-proteasome pathway. Such an increased load appears to be the consequence of a higher production of Rapidly Degraded Polypeptides (RDP). These are newly synthesized proteins which are quickly redirected to proteasome-mediated degradation. The imbalance between proteasomal load and capacity results in remarkable accumulation of poly-Ub proteins at the expense of free Ub (as established by both western blotting and immunofluorescence), unveiling basal proteasome stress in PI-sensitive MMC. In order to establish a causal link between proteasome stress and sensitivity to PI, we pharmacologically modulated either proteasome expression or workload and successfully altered PI-induced apoptosis. As predicted, increasing proteasome workload by means of ER stressors (e.g. tunicamycin, thapsigargin, brefeldin A) dramatically enhances susceptibility to PI, while a raise in proteasomal activity (achieved by exploiting the proteasome stress response, an adaptive mechanism by which mammalian cells induce proteasome biogenesis in response to either decreased proteasome function or increased proteasomal demand), confers marked resistance to PI-induced apoptosis. Having established cause-effect relationships between determinants of proteasome stress and vulnerability to PI in vitro, we then asked if our model could be used to predict responsiveness to PI in MM patients. In keeping with this hypothesis, intracellular immunostaining in primary, patient-derived MMC reveals that accumulation of poly-Ub proteins specifically hallmarks neoplastic plasma cells, indicating that the cancer compartment in MM patients suffers from proteasome stress. Moreover, poly-Ub levels positively correlate with Ig content, both intra- and inter-patient, suggesting a direct effect of Ig synthesis and/or retention on proteasome functional load. Finally, overall proteasome activity of primary MMC inversely correlates with the intrinsic apoptotic sensitivity to PI as assessed ex vivo, providing a rationale for the assessment of this parameter as a potential predictor of the in vivo response to bortezomib or other PI. Altogether, our data indicate that the balance between proteasome workload and degradative capacity represents a critical determinant of apoptotic sensitivity of MMC to PI, providing both a novel predictive tool of potential prognostic value and the framework for novel combination therapies aimed at exacerbating proteasome stress in MM.