Curative responses in the treatment of multiple myeloma (MM) are limited by the emergence of therapeutic resistance. To address this problem, we set out to identify druggable mechanisms that convey resistance to proteasome inhibitors (PIs; e.g., bortezomib), which are cornerstone agents in the treatment of MM. In isogenic pairs of PI sensitive and resistant cells, we observed stark differences in cellular bioenergetics between the divergent phenotypes. PI resistant cells exhibited increased mitochondrial respiration driven by glutamine as the principle fuel source. To target glutamine-induced respiration in PI resistant cells, we utilized the glutaminase-1 inhibitor, CB-839. CB-839 inhibited mitochondrial respiration and was more cytotoxic in PI resistant cells as a single agent. Furthermore, we found that CB-839 synergistically enhanced the activity of multiple PIs with the most dramatic synergy being observed with carfilzomib (Crflz), which was confirmed in a panel of genetically diverse PI sensitive and resistant MM cells. Mechanistically, CB-839 enhanced Crflz-induced ER stress and apoptosis, characterized by a robust induction of ATF4 and CHOP and the activation of caspases. Our findings suggest that the acquisition of PI resistance involves adaptations in cellular bioenergetics, supporting the combination of CB-839 with Crflz for the treatment of refractory MM.

Proteasome inhibitors (PIs) are cornerstone agents in the treatment of Multiple Myeloma (MM). Although initially effective, resistance to PIs inevitably emerges, presenting an obstacle to sustained and durable treatment responses in the clinic. To address this limitation, we set out to discover new small molecules that are able to restore PI sensitivity in resistant MM cells. We screened multiple chemical libraries using a cell-based screening method that identifies synergistic combinations with the PI bortezomib (Btz). This method uncovered compound E61 that, in subsequent rounds of screening, demonstrated potent PI re-sensitizing activity. E61 synergistically enhanced the activity of multiple PIs, including Btz, carfilzomib, ixazomib, and oprozomib by 3-15 fold in a genetically diverse panel of PI sensitive and resistant MM cells. In addition, E61 exhibited strong anti-MM activity as a single agent after extended treatment times (48 hours) and demonstrated >10-fold selectivity for MM cell lines over normal peripheral blood mononuclear cells (PBMCs), normal lymphocytes, and a panel of normal fibroblast cell lines. Importantly, the PI sensitizing activity of E61 was also limited to MM cells, as the drug failed to enhance the activity of PIs in normal cells. For a hit stage molecule, E61 showed exceptional tolerability and activity in vivo, significantly improving animal survival and reducing the number of CD138+ MM plasma cells in the bone marrow of mice. We used a xenotransplant model where NOD-SCID IL2Rgamma-/- (NSG) mice were injected via the lateral tail vein with PI resistant MM cells. Using this model, mice reliably reached the survival endpoint between weeks six and seven, with death being caused by the infiltration of mouse bone marrow by MM plasma cells and the development of bone lesions that closely resemble the human MM pathology. Continuous dosing with E61 (50 mg/kg, i.p., daily) was able to cure 38% of mice, with surviving mice showing only minimal residual disease (i.e., 1.0-1.5% CD138+ MM cells in the bone marrow) at the termination of the experiment. The molecular effects of E61 which lead to its anti-MM activity are characterized primarily by oxidative and endoplasmic stress responses. E61 induces reactive oxygen species (ROS) formation and oxidative damage to proteins, effects that are synergistically potentiated by the addition of PIs. This oxidative burst is critical to the anti-MM activity of E61, as the neutralizing of ROS with various molecular scavengers blocks the pro-apoptotic effects of E61. E61 triggers a robust induction of canonical ER stress markers including phospho-eIF2a, ATF4, XBP-1s, and CHOP. In order to identify the direct molecular target of E61, we chemically modified the molecule to enable copper-catalyzed azide-alkyne click chemistry coupling to fluorescent dyes and immobilizing agents. This strategy was used in tandem with peptide mass fingerprinting, which identified a small set of E61 protein binding partners that are principally involved in the proper folding of nascent polypeptide chains. Additionally, we have synthesized a small library of derivatives which retain the PI re-sensitization phenotype, that have increased metabolic stability against human liver microsomes by at least 5-fold. Ongoing efforts by our group are aimed at further validating and confirming this mechanism of action of E61 as well as further optimizing the chemical structure of E61 for enhanced potency and pharmaceutical properties toward the goal of clinical development. Overall, this work demonstrates the potential of developmental compound E61, a new class of small molecule with an apparent novel mechanism of action, as a new drug candidate for the treatment of refractory MM. Disclosures No relevant conflicts of interest to declare.

Multiple myeloma (MM) remains largely incurable due to the emergence of therapeutic resistance. We therefore set out in this study to identify druggable molecular mechanisms that convey resistance to proteasome inhibitors (PIs; e.g., bortezomib/VELCADE, carfilzomib/KYPROLIS), which are cornerstone agents in the treatment of MM. In comparing isogenic pairs of PI sensitive and resistant cells, we observed stark differences in cellular bioenergetics between the divergent phenotypes. While glycolysis rates between cell lines were similar, PI resistant cells exhibited increased mitochondrial respiration characterized by higher basal oxygen consumption rates (OCR) and overall respiratory capacity. Additionally, PI resistant cells were found to have lower activation of the AMP kinase (AMPK) energy stress pathway and increased levels of NAD(P)H, which serve as electron carriers in the mitochondrial process of oxidative phosphorylation. We determined that glutamine was the principle source of fuel driving mitochondrial respiration as removal of glutamine completely inhibited OCR as well as cell proliferation, whereas glucose and pyruvate were dispensable. Given the propensity for mitochondrial respiration in PI resistant cells and the dependence on glutamine for this process, we hypothesized that targeting glutamine utilization by PI resistant cells would restore their sensitivity to the cytotoxic effects of PIs. To test this possibility, we inhibited glutamine metabolism using the small molecule GLS1 inhibitor CB-839. CB-839 repressed basal OCR and total respiratory capacity and reduced cell viability to varying degrees in a panel of PI sensitive and resistant MM cell lines. Most notably, we found that CB-839 synergistically enhanced the cytotoxic activity of multiple PIs, including bortezomib, carfilzomib, ixazomib, and oprozomib, in a genetically diverse panel of 15 PI sensitive and resistant MM cell lines. The effects of CB-839 were the most apparent in combination with carfilzomib (Crflz), where it enhanced Crflz-induced death by >4-fold. CB-839 enhanced Crflz-induced apoptosis as measured by the activation of caspase 3, 7, 8 and the cleavage of the caspase-3 substrate PARP. Mechanistically, the combination of CB-839 and Crflz induces a strong and synergistic ER stress response, characterized by the induction of ATF4 and CHOP. Our findings suggest that the acquisition of PI resistance may involve adaptations in cellular bioenergetics that may be exploited therapeutically by targeting glutamine metabolism. Furthermore, our results support the combination of clinical stage compound CB-839 with PIs, particularly Crflz, for the treatment of refractory MM. Disclosures No relevant conflicts of interest to declare.