Abstract:Metabolic alterations in cancer are not only functional to ensure malignant cell growth but also shape a pro-tumor behavior of normal cell populations in the tumor microenvironment. Multiple myeloma (MM) is the only human cancer that is both glutamine-addicted and glutamine-auxotroph, a feature that renders MM growth completely dependent upon extracellular glutamine availability. Indeed, plasma cells from most MM patients do not express Glutamine Synthetase (GS), the only enzyme able to synthetize glutamine from glutamate and ammonium, while express high levels of Glutaminase (GLS), which catalyze the first step of glutaminolysis by deamidating glutamine into glutamate and ammonium. As a consequence, the bone marrow (BM) plasma of MM patients has low-glutamine/high glutamate levels compared to patients with smoldering MM (SMM) and Monoclonal Gammopathy of Uncertain Significance (MGUS), as firstly demonstrated by our group. This particular metabolic microenvironment induces GS expression in mesenchymal stromal cells (MSCs) impairing osteoblast (OB) differentiation of MSCs, thus favoring MM bone lesions typical of active MM. The impact of these metabolic alterations on other BM cell populations is much less defined and it has been investigated in this study. For instance, although MM BM displays an oversized adipocyte population, which sustain MM growth by providing free fatty acids, the effects of MM metabolism on MSCs and their adipocyte differentiation remain to be characterized. First we show that 13C 5-Gln isotopomer distribution revealed that almost 50% of total glutamate in human myeloma cell lines (HMLCs) directly derives from glutamine deamidation by GLS. Furthermore, in standard culture conditions, HMLCs secreted glutamate likely through the SLC7A11 transporter, whose activity was higher in HMLCs than in MSCs. On the other hand, HMCLs were not able to take up glutamate from the extracellular space, while human primary MSCs isolated from healthy donors displayed high activity of the sodium-dependent glutamate transporters of the EAAT family. Moreover, glutamate uptake was higher in undifferentiated MSCs than in MSC incubated for 14 days in osteogenic medium (10 -8 M dexamethasone and 50 μg/ml ascorbic acid). Consistently, a public transcriptional profile of MSCs and OBs obtained from bone biopsies of healthy donors (n=7) or MM patients (n=16) revealed that the expression of the inward Glu transporter EAAT3 is higher in MSCs compared to OBs. In glutamine-free conditions, MSCs produced and secreted glutamine, a process boosted by extracellular glutamate in a dose-dependent manner. Glutamine secretion was hindered by either l-methionine sulfoximine (MSO), an irreversible inhibitor of GS, or D-aspartate (D-Asp), a high-affinity inhibitor of EAAT3 that hinders glutamate uptake. In a co-culture system, HMCLs induced the expression of SNAT5 glutamine efflux transporter in MSCs. Consistently, upon glutamine withdrawal, primary undifferentiated MSCs from healthy donors sustained HMCLs growth, while this nutritional support was markedly impaired by either inhibition or silencing of GS and EAAT3, with a substantial decrease in MM cell viability. Lastly, primary human MSCs were incubated under adipogenic conditions (0.5 mM 3-Isobutyl-1-methylxanthine, 5 μM indomethacin, 50 μM dexamethasone and 10 mg/ml human insulin) in the presence or in the absence of glutamine to mimic, respectively, normal or MM BM microenvironment. Glutamine deprivation increased lipogenesis, assessed with Oil Red O staining, as well as the expression of the adipocyte markers PPARG, LEP, and ADIPOQ. These data point to a MM-driven metabolic pro-tumor BM niche in which MSCs feed glutamine-addicted MM cells, and MSC differentiation is skewed from osteogenesis to adipogenesis. Several steps of these deranged pathways are amenable to pharmacological inhibition, pointing to possible novel therapeutic approaches to counteract MM growth and its effects on the BM niche.

Glutamine Addiction of Multiple Myeloma Cells Reprograms Mesenchymal Stromal Cell Enzymes and Transporters Towards a Pro-Tumor Phenotype