Novel fatty acid binding protein 5 inhibitors SBFI103 and SBFI1143 impede myeloma cell fitness

Abstract Multiple myeloma (MM) involves clonal expansion of malignant plasma cells in the bone marrow. Obesity is a risk factor for developing MM and is associated with worse outcomes. Despite advances in therapy, MM patients often relapse and suffer from painful therapy-related side effects, highlighting the necessity for additional, less toxic therapeutic targets. Recently, targeting lipid metabolism has shown promise for inhibiting proliferation, migration and drug resistance in cancers including MM, but novel targets and therapeutics are needed. Fatty acid binding protein 5 (FABP5) is an intracellular chaperone of fatty acids that regulates lipid metabolism, cell signaling, and cell growth. High FABP5 expression in MM cells is associated with poor prognosis and pan-FABP family inhibitors suppress myeloma growth. However, the effects of inhibiting FABP5 specifically, and the associated molecular mechanisms, are not known. Thus, in the current study we investigated the role of FABP5 inhibition on MM growth in vitro and in vivo using genetic knockdown of FABP5 and three novel FABP5-specific inhibitors. We first assessed myeloma cell inhibition effects of three FABP5 inhibitors (SBFI102, SBFI103 and SBFI1143) in nine human myeloma cell lines (H929, KMS27, KMS34, MM1S, bone-passaged MM1S-GFPLuc, MM1R, OPM2, RPMI8226, and U266) and two mouse myeloma cell lines (5TGM1-TK and Vk12598). Using Real-time-Glo at 24, 48 and 72 hours, all compounds induced cytotoxicity in all cell lines tested in a dose and time-dependent manner. The IC50 values at 48 and 72 hours respectively ranged from 7.3±3.4 µM and 5.5±2.3 µM for SBFI1143 and 12.365±5.4 µM and 10.121±4.01 µM for SBFI103 across the cell lines. SBFI102 was less effective with IC50 values ~30µM at 72 hours. Although SBFI1143 was the most effective, it is not formulated appropriately for in vivo delivery at this time, and we are therefore exploring SBFI103 for its effects in our syngeneic mouse MM studies currently. Additionally, we genetically silenced FABP5 in a panel of MM cells using stable doxycycline (DOX)- inducible shRNA-mediated knockdown (KD) TRIPZ lentiviral transfection. qRT-PCR analysis showed that FABP5 mRNA was 80% reduced in MM1S KD cells (MM1S shFABP5)compared with non-targeting controls (MM1SshNC, p0.001), and 55% reduced in KMS27shFABP5 versus KMS27shNC (p0.0032) at 48 hours. Lower cell viability was also seen in MM1SshFABP5 (p0.001) and in KMS27shFABP5 (p0.001) at 48 to 96 hours versus controls. Moreover, decreased cell viability was observed in fresh, primary MM patient CD138+ cells treated with shFABP5 versus shNC (p=0.029) at 72 hours. To explore the mechanisms underlying the anti-myeloma effects of the FABP5 inhibitors and KD, we assessed MM cell apoptosis by flow cytometry, protein-level changes by mass spectrometry (MS)-based proteomics, metabolic effects by Seahorse, and transcriptional effects by RNA-sequencing. In KMS34 cells, apoptosis increased 68.8% upon treatment with 12.5 µM SBFI1143 for 48 hours and increased 60.6% when treated with 25 µM SBFI103 at 48 hours. MM1S cells showed similar inductions of apoptosis with these compounds. Seahorse mitochondria function assays showed decreases in both ATP production (p0.05) and basal respiration (p0.01) in MM1SshFABP5 versus MM.1SshNC. MS-based proteomic analysis of MM1SshFABP5 and MM1SshNCcells revealed 43 proteins with |FC| 2 and p0.05. Among these, we found IRF3, IGLL5, and DRA, an HLA class II histocompatibility antigen, downregulated in KMS27shFABP5cells, and SLAMF7, G3BP2, and DOCK7 upregulated in KMS27shFABP5cells versus KMS27shNC, implicating many immunologic and other pathway effects of FABP5. Pathway analyses, western blot confirmations, and comparisons to RNA-sequencing results are ongoing. Finally, to validate FABP5 as a target in vivo, we implanted MM1SshFABP5and MM1SshNCcells into SCID-beige mice using a subcutaneous model that allowed for fluorescent tracking of tumor growth and shRNA expression. Tumors were significantly smaller (p=0.002) in MM1SshFABP5mice than controls and histological analyses of tumors are underway. Combinations with other anti-myeloma agents, verification in syngeneic models, and bone marrow homing myeloma cell models will be used next to validate FABP5 more thoroughly in MM. Combined with prior findings, these new genetic and pharmacological data show that FABP5 specifically supports MM progression and its inhibition holds clinical potential.