Abstract
Proteasome inhibitors (PIs) represent the gold standard in the treatment of multiple myeloma. Among PIs, Bortezomib (BTZ) is frequently used as first line therapy, but peripheral neuropathy (PN), occurring approximately in 50% of patients, impairs their life, representing a dose-limiting toxicity. Carfilzomib (CFZ), a second-generation PI, induces a significantly less severe PN. We investigated possible BTZ and CFZ off-targets able to explain their different neurotoxicity profiles. In order to identify the possible PIs off-targets we used the SPILLO-PBSS software that performs a structure-based in silico screening on a proteome-wide scale. Among the top-ranked off-targets of BTZ identified by SPILLO-PBSS we focused on tubulin which, by contrast, did not turn out to be an off-target of CFZ. We tested the hypothesis that the direct interaction between BTZ and microtubules would inhibit the tubulin alfa GTPase activity, thus reducing the microtubule catastrophe and consequently furthering the microtubules polymerization. This hypothesis was validated in a cell-free model, since BTZ (but not CFZ) reduces the concentration of the free phosphate released during GTP hydrolysis. Moreover, NMR binding studies clearly demonstrated that BTZ, unlike CFZ, is able to interact with both tubulin dimers and polymerized form. Our data suggest that different BTZ and CFZ neurotoxicity profiles are independent from their proteasome inhibition, as demonstrated in adult mice dorsal root ganglia primary sensory neurons, and, first, we demonstrate, in a cell free model, that BTZ is able to directly bind and perturb microtubules.