Fatty Acid Oxidation Mediated by Malonyl-CoA Decarboxylase Represses Renal Cell Carcinoma Progression

Abstract

Fatty acid metabolism reprogramming is a prominent feature of clear cell renal cell carcinoma (ccRCC). Increased lipid storage supports ccRCC progression, highlighting the importance of understanding the molecular mechanisms driving altered fatty acid synthesis in tumors. Here, we identified that malonyl-CoA decarboxylase (MLYCD), a key regulator of fatty acid anabolism, was downregulated in ccRCC, and low expression correlated with poor prognosis in patients. Restoring MLYCD expression in ccRCC cells decreased the content of malonyl CoA, which blocked de novo fatty acid synthesis and promoted fatty acid translocation into mitochondria for oxidation. Inhibition of lipid droplet accumulation induced by MLYCD-mediated fatty acid oxidation disrupted endoplasmic reticulum and mitochondrial homeostasis, increased reactive oxygen species levels, and induced ferroptosis. Moreover, overexpressing MLYCD reduced tumor growth and reversed resistance to sunitinib in vitro and in vivo. Mechanistically, HIF2α inhibited MLYCD translation by upregulating expression of eIF4G3 microexons. Together, this study demonstrates that fatty acid catabolism mediated by MLYCD disrupts lipid homeostasis to repress ccRCC progression. Activating MLYCD-mediated fatty acid metabolism could be a promising therapeutic strategy for treating ccRCC.

Significance: MLYCD deficiency facilitates fatty acid synthesis and lipid droplet accumulation to drive progression of renal cell carcinoma, indicating inducing MYLCD as a potential approach to reprogram fatty acid metabolism in kidney cancer.

Figure 1.

MLYCD, as the key gene of FA anabolism, was downregulated and associated with poor prognosis in RCC. A, Violin plots of enrichment scores of FA catabolism and FA synthesis gene signatures in kidney normal versus tumor tissues in the TCGA_KIRC cohort. B, Kaplan–Meier analysis of subgroups (based on the FA catabolism score and FA synthesis score) on overall survival (OS) in the TCGA_KIRC cohort. Q1, low FA catabolism and low FA synthesis; Q2, high FA catabolism and low FA synthesis; Q3, low FA catabolism and high FA synthesis; Q4, high FA catabolism and high FA synthesis. C, Venn diagram highlighting the key gene of FA anabolism. D, Western blot analysis of MLYCD expression in RCC tumor tissues and paired normal kidney tissues (n = 20). Densitometry and statistical analysis. E, Representative images of IHC staining of MLYCD (left), IRS score of MLYCD in RCC tissue microarray (middle), and Kaplan–Meier analysis of MLYCD expression patterns on OS in the tissue microarray (right). F, Western blot analysis of MLYCD expression in RCC cell lines (ACHN, 786-O, 769-P, CAKI-1, A498, OS-RC-2) and the normal renal cell line (HK-2). Densitometry and statistical analysis. G and H, CCK-8 and colony formation assay for assessing the effectiveness of MLYCD restoration in cell proliferation. I and J, Representative images of Transwell assay for assessing the effectiveness of overexpressed MLYCD in cell invasion (I) and migration (J). Student t test; *, P < 0.05; ***, P < 0.001.

Figure 2.

Genetic restoration of MLYCD dramatically altered lipid homeostasis in RCC cells. A, Representative imaging of Oil Red O staining as a visual indicator of intracellular LDs in OS-RC-2 and 786-O cells with/without restoring (left) or knockout (right) MLYCD expression. And statistical analysis of the relative diameters of LDs. B, Heat map of representative lipids species in OS-RC-2 cells with MLYCD overexpressed (n = 6) and the control cells (n = 6) analyzed by the absolute quantitative lipidomics. MG, monoglyceride; SM, sphingomyelins. C–G, The content of representative lipid species in OS-RC-2 cells with MLYCD overexpressed (mean ± SEM, n = 6) and the control cells (mean ± SEM, n = 6). H, Heat map of DEGs in OS-RC-2 cells with stably MLYCD overexpressed (n = 3) compared with the control cells (n = 3), analyzed by RNA sequencing. P < 0.05 and |LogFC| > 1 were considered significant. I, GO analysis was performed to look for biological process associated with MLYCD. J, Combined transcriptomic and lipidomics analysis based on DEGs and differentially lipid species. Student t test; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 3.

MLYCD inhibited de novo FA synthesis by reducing malonyl-CoA content, ultimately induced cell death via increasing ER stress. A, Schematic illustration of the key role of MLYCD in FA anabolism. B and C, Levels of malonyl CoA, FFA, and TGs in OS-RC-2 and 786-O cells with/without overexpression (B) or knockout expression (C) of MLYCD. D and E, Level of de novo FA synthesis in different RCC cells was measured by [¹⁴C] acetate incorporation into lipids. F, Representative fluorescence imaging of LDs stained with BODIPY 493/503 (green) and nuclei stained with DAPI (blue) in OS-RC-2 and 786-O cells with/without MLYCD overexpression and/or malonyl CoA (10 μmol/L) for 24 hours. G, CCK-8 assay for assessing the cell proliferation of 786-O cells with/without MLYCD overexpression and/or malonyl-CoA (10 μmol/L). H, Representative fluorescence imaging of ER tracker (red) staining to assess the ER stress expansion in OS-RC-2 and 786-O cells with/without MLYCD overexpression. Nuclei were stained with DAPI (blue). I, Representative imaging of Western blot analysis for detecting the expression of ER stress markers in OS-RC-2 and 786-O cells with/without MLYCD overexpression. J, CCK-8 assay for assessing the cell proliferation of OS-RC-2 and 786-O cells with/without MLYCD overexpression (B) and/or 4-PBA (10 μmol/L). ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 4.

MLYCD restoration resulted in mitochondrial damage, ROS elevation, and cell activity inhibition via enhancing CPT1A activity. A, CPT1A activity measurement in OS-RC-2 and 786-O cells with/without MLYCD overexpression (left) or knockout (right). B, OCR in OS-RC-2 cells with MLYCD restoration and the control cells. Oligo, oligomycin; FCCP, carbonyl cyanide 4-trifluoromethoxy-phenylhydrazone. C, Representative TEM imaging of the mitochondria in OS-RC-2 and 786-O cells with/without MLYCD restoration. Red arrows, mitochondria. D, Representative fluorescence imaging of mitochondria (red) and nuclei (blue) in OS-RC-2 cells with/without MLYCD overexpression and/or malonyl-CoA (10 μmol/L). E, Total intracellular ROS production was assessed by flow cytometry using DCFH-DA. F, Representative fluorescence imaging of lipid-derived ROS (red) detected with BODIPY 665/676 and mitochondria (green) in OS-RC-2 cells with/without MLYCD restoration and/or CPT1A inhibitor ETO (2 μmol/L). G, CCK-8 assay for assessing the cell proliferation of OS-RC-2 and 786-O cells with/without MLYCD overexpression and/or ETO (2 μmol/L). H, The levels of the unsaturated bonds in lipid based on the absolute quantitative lipidomics (left). The degree of lipid peroxidation was assessed by the malondialdehyde measurement in cells. I and J, Dead cell percentage and cell proliferation was measured after erastin (10 μmol/L) for 48 hours and Ferr-1 (20 μmol/L) treatment by CCK8 assay, respectively. K, Representative imaging of Western blot analysis for detecting the expression of ferroptosis marker genes in OS-RC-2 or 786-O cells with/without MLYCD overexpression (left) or knockout (right). L, Representative fluorescence imaging of lipid-derived ROS (red) detected with BODIPY 665/676 and nuclei (blue) in OS-RC-2 and 786-O cells with/without MLYCD restoration and/or ROS scavenger NAC (10 μmol/L). M, CCK-8 assay for assessing the cell proliferation of OS-RC-2 and 786-O cells with/without MLYCD overexpression and/or NAC (10 μmol/L). Student t test; ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 5.

HIF regulated the expression of MLYCD through eIF4G3 microexons. A and B, Representative imaging of Western blot for detecting the MLYCD expression in OS-RC-2 or 786-O cells with/without knockdown HIF1α/HIF2α. Densitometry and statistical analysis. C, The levels of MLYCD mRNA in OS-RC-2 and 786-O cells with/without knockdown HIF2α. D, Representative imaging of Western blot for detecting the MLYCD expression in OS-RC-2 and 786-O cells with/without HIF2α overexpressing and/or treatment with DMSO or MG132 (20 nmol/L) or chloroquine (50 nmol/L). Densitometry and statistical analysis. E, Venn diagram of the translation-related DEGs according to the RNA sequencing in RCC tissues with paired normal kidney tissues (n = 3) and 786-O or OS-RC-2 cells with knockdown HIF2α (n = 3). DEGs were identified on the basis of P < 0.05 and |LogFC| > 2. F, The correlation of HIF2α and eIF4G3 expression in TCGA_KIRC database. G, eIF4G3 microexon splicing was detected by RT-PCR assay in OS-RC-2 and 786-O cells with/without HIF2α knockdown. Representative imaging of gel map amplifying the spliced products. The GAPDH expression acted as the loading and control. Red rectangle, microexon; orange rectangle, adjacent alternative exon. H, The predicted positions of putative HIF2α-binding motif in −2,000-bp human eIF4G3 promoter. I, ChIP-PCR assays monitoring direct binding of HIF2α to eIF4G3 promoter regions in OS-RC-2 and 786-O cells. J, Luciferase reporter assays were carried out by cotransfecting the WT eIF4G3 promoter or fragment E2-mutant eIF4G3 promoter with HIF2α overexpression vector or the negative vector in OS-RC-2 and 786-O cells. K, Representative imaging of Western blot analysis for detecting the MLYCD and eIF4G3 expression in OS-RC-2 and 786-O cells with/without HIF2α overexpressing and/or eIF4G3 knockout. Densitometry and statistical analysis. Student t test; ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 6.

HIF2α regulated FA anabolism by inhibiting MLYCD expression. A, Representative images of Western blotting for detecting MLYCD and HIF2α expression in OS-RC-2 and 786-O cells with/without HIF2α knockdown (shHIF2α) and/or MLYCD knockout (sgMLYCD). B, Levels of malonyl-CoA, de novo FA synthesis, and CPT1A activity in cells with/without HIF2α knockdown and/or MLYCD knockout. C, Levels of FFA and TG in cells with or without HIF2α knockdown and/or MLYCD knockout. D, Representative fluorescence images of LDs stained with BODIPY 493/503 (green) and nuclei stained with DAPI (blue) in cells with or without HIF2α knockdown and/or MLYCD knockout. E and F, CCK-8 and Transwell assays to assess the effect of HIF2α knockdown and/or MLYCD knockout on cell proliferation and invasion. G, Representative TEM images of mitochondria in OS-RC-2 cells with or without HIF2α knockdown and/or MLYCD knockout. Red arrows, mitochondria. Green arrows, endoplasmic reticulum. H, OCR measurement in OS-RC-2 cells with or without HIF2α knockdown and/or MLYCD knockout. I, Representative fluorescence imaging of lipid-derived ROS (red) and mitochondria (green) in OS-RC-2 cells with or without HIF2α knockdown and/or MLYCD knockout. J, The percentage of dead cells was detected by the CCK8 assay after treatment with erastin (10 μmol/L) for 48 hours. K, Representative images of Western blotting for detecting the expression of ferroptosis marker genes and PLIN2 in cells with/without HIF2α knockdown and/or MLYCD knockout. Student t test; ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 7.

MLYCD reversed RCC resistance to sunitinib in vitro and in vivo. A, Representative fluorescence images and quantification of the corrected total cell fluorescence (CTCF) of Bodipy (green) and CD31 (red) in different stages of OS-RC-2 tumor xenograft with sunitinib treatment. Su1rd, first generation tumor; Su2rd, second generation tumor; Su3rd, third generation tumor. B and C, Levels of malonyl-CoA content, de novo FA synthesis and CPT1A activity, and representative images of Western blot for detecting the expression of MLYCD in cells from different stages of OS-RC-2 and 786-O tumor xenograft with sunitinib treatment. D and E, The sunitinib sensitivity and cell proliferation of the sunitinib resistant cells (OS-RC-2-Su3rd and 786-O-Su3rd) with/without MLYCD restoration was detected by CCK8 assay. F, The sunitinib sensitivity and cell proliferation of the cells (OS-RC-2 and 786-O) with/without MLYCD restoration was detected by CCK8 assay. G–I, BALB/c nude mice bearing OS-RC-2 xenografts with MLYCD overexpression were treated with sunitinib (40 mg/kg/day) or PBS according to standard therapy (4 weeks on and 2 weeks off treatment). Measured the tumor volumes every 3 days. Weighed the tumors after resection. J, Representative imaging of IHC of MLYCD, Ki67, PERK, PTGS2, and caspase-3 or IF for LDs via staining Bodipy in tumor tissues from A. Student t test; ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 8.

Conditional Mlycd-knockout mice confirmed the function of endogenous Mlycd. A and B, Representative imaging of IHC of Mlycd and hematoxylin and eosin (H&E) staining in kidney tissues from the kidney proximal tubules–specific Mlycd-knockout mice (Mlycd-CKO) and the control mice (WT). Statistical analysis based on immunoreactivity score (IRS). C, Representative imaging of IHC of RCC biomarkers (Ki67, vimentin, Ca-ix, Rcc, Ck-7) in kidney tissues from Mlycd-CKO and WT mice. D, Volcano plot demonstrating DEGs in kidney tissues from Mlycd-CKO versus WT mice. E, Bar graph showing that the Top 15 GSEA and GO of hallmark gene sets based on the P value. F, Representative imaging of LDs staining with Bodipy in kidney tissues from Mlycd-CKO and WT mice. G, Levels of malonyl-CoA, FFA and TG in kidney tissues from Mlycd-CKO and WT mice. H, Representative imaging of IHC of Plin2, Perk, Ptgs2, Scd, caspase-3, and Tunel in kidney tissues from the kidney proximal tubules–specific Mlycd-knockout mice (Mlycd-CKO) and the control mice (WT). Statistical analysis based on immunoreactivity score (IRS) or percentage of positive cells. Student t test; ns, not significant; **, P < 0.01; ***, P < 0.001.