Fatty Acid Oxidation Mediated by Acyl-CoA Synthetase Long Chain 3 Is Required for Mutant KRAS Lung Tumorigenesis

Abstract

KRAS is one of the most commonly mutated oncogenes in human cancer. Mutant KRAS aberrantly regulates metabolic networks. However, the contribution of cellular metabolism to mutant KRAS tumorigenesis is not completely understood. We report that mutant KRAS regulates intracellular fatty acid metabolism through Acyl-coenzyme A (CoA) synthetase long-chain family member 3 (ACSL3), which converts fatty acids into fatty Acyl-CoA esters, the substrates for lipid synthesis and β-oxidation. ACSL3 suppression is associated with depletion of cellular ATP and causes the death of lung cancer cells. Furthermore, mutant KRAS promotes the cellular uptake, retention, accumulation, and β-oxidation of fatty acids in lung cancer cells in an ACSL3-dependent manner. Finally, ACSL3 is essential for mutant KRAS lung cancer tumorigenesis in vivo and is highly expressed in human lung cancer. Our data demonstrate that mutant KRAS reprograms lipid homeostasis, establishing a metabolic requirement that could be exploited for therapeutic gain.

Keywords: ACSL3; cancer metabolism; fatty acid oxidation; lipid metabolism; lung cancer; mouse cancer models; mutant KRAS.

Figure 1. Mutant KRAS Regulates ACSL3 and ACSL4 in Lung Cancer Cells Both In Vitro and In Vivo

(A) Transgenic (Kras^G12D) and endogenous (Kras^WT) mRNA was quantified in micro-dissected lung tumors of CC10-rtTA/Tet-op-Kras^G12D mice by real-time PCR at the indicated time points after doxy withdrawal. (B) Pan-Ras activity was measured with an ELISA assay in micro-dissected lung tumors at the indicated time points after doxy withdrawal in CC10-rtTA/Tet-o-Kras^G12D mice; ***p < 0.0003. (C) Immunoblot detection of P-Pdk1 in lysates of micro-dissected lung tumors at the indicated time points after Kras^G12D extinction. (D) H&E-stained lung tissue sections demonstrating lung tumor regression at the indicated time points after Kras^G12D extinction; the scale bar represents 1 mm. (E) Heatmaps of the genes found enriched in a genome-wide expression profiling, illustrating the changes in gene expression of the indicated metabolic pathways 24 hr upon doxy withdrawal. Expression level shown is representative of log₂ values of each replicate from micro-dissected lung tumors (n = 4). Red signal denotes higher expression relative to the mean expression level within the group, and blue signal denotes lower expression relative to the mean expression level within the group. (F) GSEA plot of glycolytic processes based on the 24-hr off-doxy versus on-doxy gene expression profiles. NES, normalized enrichment score; p, nominal value. (G) GSEA plot of and lipid metabolic processes based on the 24-hr off-doxy versus on-doxy gene expression profiles. (H) Real-time PCR showing mRNA levels of Hk2 and Ldh-A in lung tumors of Kras^G12D mice before or 24 hr after Kras^G12D extinction. (I) Real-time PCR showing mRNA levels of Acsl family members in lung tumors of Kras^G12D mice before or 24–48 hr after Kras^G12D extinction. *p < 0.02; **p < 0.004. (J) ACSL enzymatic activity in healthy lung parenchyma or in micro-dissected lung tumors after 11 weeks of Kras^G12D induction or after the indicated length of Kras^G12D extinction. **p < 0.003; ***p < 0.001. (K) ACSL enzymatic activity measured in BK fibroblasts treated with doxy for 24 hr. (L) Real-time PCR showing mRNA levels of Acsl family members in BK fibroblasts treated with doxy for 24 hr. (M) ACSL enzymatic activity in A549 NSCLC cells after transfection with the indicated siRNAs. See also Figure S1.

Figure 2. ACSL3 Is Required for the Proliferation of NSCLC Cells Expressing Mutant KRAS

(A) Growth curves of NSCLC cells upon transduction with the indicated shRNAs. KRAS status for each cell line is indicated. Doubling time is mentioned in hours in parentheses. (B) The histogram shows cell viability of the indicated NSCLC cells 96 hr after transfection with ACSL3 or control shRNAs. (C) Schematic representation of the ACSL3 rescue cDNA mutants. shRNA recognition site and sequence of shRNA mutant 1 and 2 are indicated. Red letters indicate mutated nucleotides. (D) The histogram shows H460 (mutant KRAS) cell viability of lung cancer cell line transduced with retroviruses expressing the indicated shRNA and ACSL3 cDNAs. See also Figure S2 and Table S1.

Figure 3. ACSL3 Is Required for the Survival and Anchorage-Independent Growth of NSCLC Cells Expressing Mutant KRAS

(A) Flow cytometry showing the effect of shRNA-mediated ACSL3 knockdown on cell cycle of the indicated NSCLC lines. The percentages of cells in specific phases of the cell cycle are indicated. Note the presence of cells in the sub-G0 phase of the cell cycle in the ACSL3-dependent H460 and H1264 cells, but not in the ACSL3-independent H522 cells. (B) Representative images of tissue culture plates showing soft agar colony formation assay in A549 and H460 (mutant KRAS) cells treated as indicated. (C) Histograms showing quantification of colonies of A549 and H460 cells grown in soft agar upon transduction with the indicated shRNAs; ***p < 0.001. See also Figure S3.

Figure 4. Suppression of Fatty Acids Uptake and β-Oxidation upon Mutant KRAS Suppression

(A) Heatmaps illustrating the changes in expression of FA metabolism/β-oxidation genes 24 hr upon doxy withdrawal in mutant Kras lung cancer tumors in transgenic mice. Expression level shown is representative of log₂ values of each replicate from micro-dissected lung tumors. Red signal, higher expression; blue signal, lower expression relative to the mean expression level within the group. (B) GSEA plot of FA metabolism/β-oxidation genes based on the 24-hr off-doxy versus on-doxy gene expression profiles. (C) Real-time PCR showing the mRNA level of the indicated genes upon Kras^G12D extinction in vivo; genes responsible for long-chain FA transport carnitine O-octanoyltransferase (Crot) and fatty acid transporter-member 2 (Fatp2/Slc27a2) and 4 (Fatp4/Slc27a4); and genes that control FA β-oxidation, such as enoyl coenzyme A hydratase (Echs1) and carnitine O-acetyltransferase (Crat); ns, statistically non-significant; *p < 0.01; **p < 0.003; ***p < 0.001. (D) BODIPY-FA uptake in BK cells treated with doxy as indicated. Red and white arrows indicate positive and negative cells, respectively (cells with both cytoplasmic and membrane staining were considered positive); the scale bar represents 50 μm. Note that mutant KRAS induces the uptake/retention of BODIPY-FA. (E) Fluorescence-activated cell sorting analysis of BK cells after a 30-min pulse labeling with BODIPY-FA. Note uptake/retention of BODIPY-FA in BK cells expressing mutant KRAS. (F) BODIPY-FA retention in mutant KRAS HCC44 NSCLC cells expressing the indicated shRNAs. Red and white arrows indicate representative positive and negative cells, respectively; the scale bar represents 50 μm. (G) The histogram shows the percentage of BODIPY-FA-positive HCC44 and H460 cells transduced with indicated ACSL3 shRNA (100 cells were counted in three separate areas). (H) Composition of lysophosphatidylcholine (LPC) and triacylglycerides (TAGs) in BK cells with and without doxy treatment (*p < 0.05; **p < 0.001). (I) Composition of LPC and TAGs in H460 NSCLC cells treated with the indicated ACSL3 shRNA (*p < 0.05). (J) β-oxidation assay in H460 NSCLC cells treated with the indicated ACSL3 shRNAs or etomoxir; ***p < 0.001. (K) Measurement of ATP in H460 NSCLC cells treated with the indicated ACSL3 shRNAs or etomoxir. (L) The histogram shows the viability of the indicated mutant KRAS NSCLC cells 72 hr after the indicated treatments; *p < 0.01; **p < 0.004. (M) The histogram shows the viability of the indicated mutant KRAS NSCLC cells 72 hr after the treatment with 4-bromocrotonic acid; **p < 0.004. See also Figure S4.

Figure 5. ACSL3 Suppression Impairs Tumorigenesis In Vivo

(A and B) Tumor volume of xenografts of A549 and H460 NSCLC cells upon transduction with the indicated shRNAs. (C and D) Kaplan-Meier curves of nude mice carrying xenografts of A549 and H460 NSCLC cells expressing the indicated shRNA upon transduction with lentiviruses (n = 6/group).

Figure 6. ACSL3 Is Required for the Initiation of KRAS-Driven NSCLC in Mice and Is Expressed in Patients

(A) Images of Kras^G12D lung cancers stained with H&E after 8 weeks of adeno-Cre infection. Lung nodules are indicated as A: adenoma, AAH: atypical adenomatous hyperplasia, and BH: bronchial hyperplasia (magnification: upper panels 12.5×; bottom panels 100×). (B) Histogram of tumor type distribution in mice with the indicated genotypes (*p < 0.05; ***p < 0.001; n, number of mice samples per genotype). Tumor types are indicated. (C) Images of murine lung tumors stained with SP-C antibody. Positive staining confirms their origin from type II alveolar cells (magnification 400×). Arrows indicate the location of the inlets. (D) Images of murine lung tumors stained with indicated antibodies. Antibody staining is in brown; nuclear counter staining is in blue (magnification 400×). Arrows indicate the location of inlets. (E) The histogram shows the percentage of P-H3-positive cells in lung tumors of Kras^G12D; Acsl3^+/+ and Kras^G12D; Acsl^−/− mice (100 cells were counted in three separate areas). (F) Representative microphotographs of ACSL3 in the indicated representative human samples (respective H-score for tumor samples are indicated). Black arrowhead indicates ACSL3 in type II pneumocytes. Arrows indicate the location of inlet (the scale bar represents 200 μm). (G) Histogram showing H-score for ACSL3 staining intensity in tumor samples according to their pathologic stages (*p < 0.05; ns, statistically insignificant; n, number of patient samples). See also Figure S5.