Targeting The Interactions Of Fatty Acid Metabolism With PI3K/mTOR and MAPK As a Novel Therapeutic Strategy In Diffuse Large B-Cell Lymphoma (DLBCL)

Fatty acid metabolism is altered in several cancers, which includes de novo synthesis by fatty acid synthase (FASN) and increased utilization of fatty acids via β-oxidation to meet cellular energy requirements. PI3K and MEK/ERK signal transduction are primary biological pathways that activate tumor-related fatty acid metabolism. Notably, these concepts have emerged primarily from solid tumor studies; there is a comparative paucity of data in lymphoma. We examined the functional roles of fatty acid metabolism in DLBCL cells and their interaction with oncogenic signal transduction pathways including PI3K, MEK/ERK and hypoxia inducible factor-1 alpha (HIF-1a). We also investigated potential therapeutic implications of targeting fatty acid metabolism for the treatment of DLBCL.

DLBCL cell lines OCI-LY3, OCI-LY19, SUDHL4, and SUDHL10 were examined in this study. Cerulenin (FASN inhibitor), Orlistat (FASN and LPL inhibitor), BEZ-235 (Dual PI3K/mTOR inhibitor), AZD6244 (MEK inhibitor) were used to study the effects of enzyme inhibition on lipid metabolism, cell signaling and cell death. We assessed cell viability with the MTT assay and apoptosis by flow cytometric analysis of Annexin-V/propidium iodide (PI). FASN or LPL mRNAs were quantified in DLBCL cell lines by quantitative RT-PCR as well as through gene expression profiling (GEP) analysis (by cell of origin) using the CaBIG dataset. Further, FASN and associated signaling pathways (MEK, ERK, and HIF-1a) were analyzed by Western blot. Finally, for investigation of potential interactions between FASN and HIF-1a or MAPK signaling, we utilized short hairpin RNA interference (shRNA) to knock out (KO) desired pathways.

GEP analysis showed that genes related to fatty acid synthesis and mTOR were more prominently expressed in germinal center (GC) DLBCL (p=0.0006 vs GC control and p=0.0001 vs non-GC DLBCL), whereas genes involved in utilization of fatty acid for energy, PI3K and MAPK were preferentially expressed in non-GC DLBCL (p<0.0001 vs non-GC control and GC DLBCL). Protein level expression of FASN and CPT1 (relevant to β-oxidation) was readily detectable in all DLBCL cell lines, while the expression of LPL was low, except in SUDHL10 cells. We next examined FASN expression following KO of MEK, ERK, and/or HIF-1a using shRNA in OCI-LY3 and SUDHL10 cells. HIF-1a KO in OCI-LY3 cells and MEK, ERK KO in SUDHL10 significantly decreased FASN expression. Similarly, pharmacological inhibition of PI3K/mTOR or MEK using novel small molecule agents BEZ-235 or AZD6244 (selumetinib), we observed down-regulation of both HIF-1a and FASN expression. Treatment with BEZ-235 under fatty acid deprived conditions resulted in synergistic cell death 60% in OCI-LY3, compared to 25% with BEZ-235 alone. Additionally inhibition of MEK/ERK using AZD6244 or U0126 also resulted in down-regulation of CPT1 relevant to β-oxidation. Pharmacologic treatment with cerulenin, however, resulted in inhibition of MEK and ERK phosphorylation and apoptosis in SUDHL4 and OCI-LY3 cells. Furthermore, treatment with AZD6244 or cerulenin both resulted in time- and dose-dependent cell death (Annexin-V/PI) and decreased proliferation (by MTT) in GC and non-GC DLBCL cell lines (P<0.001).

Altogether, we demonstrated that fatty acid metabolism is highly upregulated in DLBCL and that there was preferential gene expression in DLBCL cells in part according to cell of origin. Further, PI3K/mTOR, MAPK and HIF-1a constitutively activated oncogenic pathways in DLBCL appeared here to directly regulate fatty acid metabolism. In addition, we showed that targeting fatty acid metabolism may be harnessed as a potential therapeutic strategy. Continued investigation is warranted to delineate the mechanisms through which MAPK and HIF-1a regulate FASN expression and to determine the in vivo implications of FASN inhibition on DLBCL tumor growth.

No relevant conflicts of interest to declare.