Nicotinamide phosphoribosyltransferase (NAMPT) inhibitor-resistant rhabdomyosarcoma (RMS) models exhibit alterations in metabolic and genomic profiles

RMS is a common pediatric soft tissue sarcoma for which new therapies are critically needed. We previously demonstrated that RMS is highly sensitive to inhibitors of NAMPT, which catalyzes the rate-limiting step of the NAD + salvage pathway and is the only pharmacologically targetable NAD + production enzyme. Treatment with the NAMPT inhibitor OT-82 results in complete tumor regressions in vivo, however, upon intermittent treatment, acquired resistance develops in some models. As acquired drug resistance is a known impediment to the clinical efficacy of targeted agents, we sought to elucidate potential mechanisms of OT-82 resistance in RMS. Mice with orthotopic fusion-positive (FP) and fusion-negative (FN) RMS xenograft tumors were treated with OT-82 for 8 weeks on the clinical dosing schedule. After stopping treatment, mice were observed for recurrence and retreated when tumors regrew to >900 mm 3 . Tumors that progressed on treatment were harvested and converted to cell lines. Two resistant cell lines (1 FP - Rh30-mRes and 1 FN - RD-mRes) were selected for further study. Incucyte live cell analysis confirmed retention of OT-82 resistance in vitro with resistant cells maintaining proliferation at doses of OT-82 up to 30X above the IC 50 of parental cells. Resistant cells exposed to OT-82 maintained ATP levels consistent with that of untreated controls. After 24h of OT-82 treatment, NAD + loss was observed in both parental and resistant cells, however resistant cells recovered NAD + levels within 48-96h. Effects on glucose metabolism, measured using extracellular flux and metabolomic analyses demonstrated that in the presence of OT-82, only resistant cells maintained glycolytic function. Specifically, metabolites downstream of the NAD + -dependent enzyme glyceraldehyde-3-phosphase dehydrogenase were reduced in parental cells but maintained in resistant cells. Analysis of protein expression of NAD + synthesis enzymes NAMPT, NAPRT, and QPRT revealed that Rh30-mRes expresses more QPRT, however, genetic silencing of QPRT did not reverse resistance, suggesting upregulation of compensatory NAD production enzymes is not a primary mechanism of resistance. Whole exome sequencing revealed that each resistant cell line has a distinct, previously unreported mutation in NAMPT . Protein modeling suggests each mutation affects the drug binding pocket of NAMPT, with the S17F variant in Rh30-mRes resulting in a collapse of the pocket and the S241C variant in RD-mRes reducing binding affinity of OT-82. These findings are consistent with functional studies demonstrating that Rh30-mRes is resistant to multiple other NAMPT inhibitors whereas RD-mRes is resistant only to OT-82. Together, these data suggest that acquired resistance to NAMPT inhibitors in RMS models involves the development of mutations in the target protein affecting drug binding and affinity.