Fatty acid synthesis is a therapeutic target in human liposarcoma

Abstract

Liposarcomas (LS) are mesenchymal tumors that can recur after surgical resection and often do not respond to presently available medical therapies. This study demonstrates the dependence of LS on de novo long-chain fatty acid synthesis for growth. Lipogenesis can be impaired by inhibiting the activities of lipogenic enzymes, including acetyl CoA-carboxylase (ACC) and fatty acid synthase (FASN), or by suppressing the expression of key genes involved in the pathway and its regulation. The FASN inhibitors cerulenin and orlistat reduced the growth of two LS cell lines (LiSa2, SW872), as did inhibition of ACC with soraphen A. CDDO-Me, a synthetic triterpenoid, suppressed expression of Spot 14 and FASN genes and likewise inhibited LS cell growth. Importantly, the anti-proliferative effect of each agent was prevented by the co-administration of palmitate, the major product of cellular long-chain fatty acid synthesis. In stark contrast to LS cells, these compounds had no effect on the growth of fibroblasts. Four biochemically distinct agents that target critical points in the fatty acid synthetic pathway exert anti-proliferative effects on LS cells, and rescue of cell growth by palmitic acid suggests that reduced tumor cell lipogenesis mediates the growth inhibition. These findings warrant further studies aimed at the clinical exploitation of the dependence of LS cell growth on fatty acids.

Figure 1

Lipogenesis-related RT-PCR gene profile for liposarcoma cell lines. An ethidium-stained agarose gel electrophoresis of RT-PCR products after 25 cycles is shown. Primers for fatty acid synthase (FASN), Spot 14 (S14) and cyclophilin (CYC) were used to analyze two different liposarcoma cell lines (LiSa-2, SW872) and fibroblasts (Fibro). Template mRNA was harvested after induction of differentiation in the liposarcoma cells.

Figure 2

Anti-proliferative effect of cerulenin. (A) LiSa-2 cells were grown for 5 days in the indicated concentrations of cerulenin with or without 32 μM palmitate. DNA content/well was assessed using the Hoechst assay. (B and C) Replicated experiment in SW872 and fibroblast cells. Data are mean ± SEM normalized to the control group; n=8/group. ^*p<0.0001 comparing cerulenin treatment to control; ^#p<0.03 comparing palmitate with cerulenin to corresponding cerulenin dose alone.

Figure 3

Anti-proliferative effect of orlistat. (A) LiSa-2 cells were grown for 5 days in vehicle or 20 μM orlistat in the presence or absence of 32 μM palmitate. DNA content/well was assessed using the Hoechst assay. (B) Replicated experiment in fibroblasts. DNA content/well was assessed using the Hoechst assay. Data are mean ± SEM normalized to the control group; n=8/group. ^*p<0.0001 comparing orlistat treatment to control; ^#p<0.000001 comparing palmitate with orlistat to the corresponding orlistat dose alone.

Figure 4

Effects of soraphen A. (A) LiSa-2 cells were grown for 24 h in vehicle or 75 nM soraphen A with or without 32 μM palmitate, at which time lipid synthesis (cpm incorporated/mg protein) was assessed over a 3-h labeling period with 1-[14-C]-actetate. (B) LiSa-2 cells were grown for 5 days in 75 nM soraphen A in the presence or absence of 32 μM palmitate, and DNA content/well was assessed using the Hoechst assay. Data are mean ± SEM with n=8/group. ^*p<0.01 compared to vehicle control.

Figure 5

Anti-proliferative effects of CDDO-Me. (A) LiSa-2 cells were grown for 12 days in 100 nM CDDO-Me in the presence or absence of 32 μM palmitate. DNA content/well was assessed using the Hoechst assay. (B) Replicated experiment in fibroblasts. Data are expressed as mean ± SEM with n=8/group normalized to the control groups. ^*p<0.001 comparing CDDO-Me treatment to control; ^#p<0.05 comparing palmitate with CDDO-Me to CDDO-Me alone.

Figure 6

CDDO-Me downregulates FASN and S14 expression in LiSa-2 cells. Quantitative real-time PCR (45 cycles) results for fatty acid synthase (FASN) and S14 mRNAs in LiSa-2 (A and C) and fibroblast (B and D) cells after 12-day exposure to 100 nM CDDO-Me or vehicle alone. Data (mean ± SEM, n=6/group) are normalized to the cyclophilin signal and are expressed as a percent of control. ^*p<0.01 compared to vehicle treatment.