Integration of Lipidomics and Transcriptomics Reveals Reprogramming of the Lipid Metabolism and Composition in Clear Cell Renal Cell Carcinoma

Abstract

Clear cell renal cell carcinoma (ccRCC) is fundamentally a metabolic disease. Given the importance of lipids in many cellular processes, in this study we delineated a lipidomic profile of human ccRCC and integrated it with transcriptomic data to connect the variations in cancer lipid metabolism with gene expression changes. Untargeted lipidomic analysis was performed on 20 ccRCC and 20 paired normal tissues, using LC-MS and GC-MS. Different lipid classes were altered in cancer compared to normal tissue. Among the long chain fatty acids (LCFAs), significant accumulations of polyunsaturated fatty acids (PUFAs) were found. Integrated lipidomic and transcriptomic analysis showed that fatty acid desaturation and elongation pathways were enriched in neoplastic tissue. Consistent with these findings, we observed increased expression of stearoyl-CoA desaturase(SCD1) and FA elongase 2 and 5 in ccRCC. Primary renal cancer cells treated with a small molecule SCD1 inhibitor (A939572) proliferated at a slower rate than untreated cancer cells. In addition, after cisplatin treatment, the death rate of tumor cells treated with A939572 was significantly greater than that of untreated cancer cells. In conclusion, our findings delineate a ccRCC lipidomic signature and showed that SCD1 inhibition significantly reduced cancer cell proliferation and increased cisplatin sensitivity, suggesting that this pathway can be involved in ccRCC chemotherapy resistance.

Keywords: SCD1; cholesterol; lipidomics; lipids; renal cell carcinoma.

Figure 1

Volcano plot of the 158 lipids profiled (a). Principal component analysis (PCA) of the global tissue lipidome demonstrated that the two groups (clear cell renal cell carcinoma (ccRCC) vs. normal renal tissue) were clearly distinguishable (b). Hierarchical clustering heatmap analysis of lipids in normal and cancer tissues (c). Metabolic set enrichment analysis (MSEA) showing the most altered biochemical metabolic pathways in ccRCC (d). ChemRICH set enrichment statistical plot. Each node reflects a significantly altered cluster of lipids. Node sizes represent the total number of lipids in each cluster set. The node color scale shows the proportion of increased (red) or decreased (blue) compounds in tumor compared to normal tissue. Purple nodes have both increased and decreased lipids (e).

Figure 2

Lipid classes differentially accumulated between neoplastic (ccRCC) and normal tissue. Y-axis: metabolite relative amount. Small squares indicate outlier in box-and-whisker plots. Solid square indicates extreme outlier.

Figure 3

Among the long chain fatty acids (LCFAs), saturated fatty acids (SFAs), monounsaturated fatty acid (MUFAs), and polyunsaturated fatty acids (PUFAs) were significantly accumulated in cancer (ccRCC) compared to normal tissue (a). Schematic model summarizing the differences in free fatty acids between normal and tumor tissue (b). Exploration of the “Metabologram” Data Portal. The changes in the “Biosynthesis of unsaturated fatty acids” and “Fatty acid elongation” pathways are shown in both transcripts and metabolites when comparing tumors to adjacent normal kidney tissues (c). Y-axis: metabolite relative amount. Small squares indicate outlier in box-and-whisker plots.

Figure 4

The tissue levels of eicosanoids were reduced in tumor samples (ccRCC) compared to normal kidney (a). Exploration of the “Metabologram” Data Portal. The changes in the “Arachidonic acid metabolism” pathway are shown in both transcripts and metabolites when comparing tumors to adjacent normal kidney tissues (b). Y-axis: metabolite relative amount. Small squares indicate outlier in box-and-whisker plots. Solid square indicates extreme outlier.

Figure 5

Cholesterol biosynthesis pathways. A reduced accumulation of main metabolic intermediates in both Kandutsch-Russell and Bloch pathways was observed in neoplastic tissue (ccRCC). Y-axis: metabolite relative amount. Small squares indicate outlier in box-and-whisker plots. Solid square indicates extreme outlier.

Figure 6

Integrated metabolic pathway enrichment analysis. All the matched pathways are displayed as circles. The color and size of each circle are based on the p-value and pathway impact value, respectively (a). Results of pathway analysis according to p-value and false discovery rate (FDR) (b).

Figure 7

Gene set enrichment analysis (GSEA) of the GSE47032 (a) and GSE15641 dataset (b).

Figure 8

Analysis of gene expression by Real time PCR of ATP citrate lyase (ACLY), sterol regulatory element-binding transcription factor 1 (SREBF1), stearoyl-CoA desaturase-1 (SCD1) and fatty acid elongase 2 and 5 (ELOVL2 and ELOVL5) (a). Levels of citrate and acetyl-CoA-to-citrate ratio are increased in ccRCC compared to normal tissue (b). Palmitoleate-to-palmitate ratio and stearate-to-palmitate ratio are increased in ccRCC (c). Small squares indicate outlier in box-and-whisker plots. Solid square indicates extreme outlier.

Figure 9

Gene set enrichment analysis (GSEA) of the GSE41485 dataset (a). SCD1 has a role in RCC resistance to cisplatin (CDDP)-induced cytotoxicity (C). The death rate of treated tumor cells (tumor + A939572 + CDDP) is significantly higher than that of untreated cells (tumor + CDDP) (p < 0.001). No difference is observed in normal cells (p > 0.05) (b). MTT assay reveals significantly decreased cell viability when RCC cells are treated with A939572 before cisplatin incubation (c).

Figure 10

Analysis of gene expression by real-time PCR of prostaglandin-endoperoxide synthase 2 (PTGS2) and prostaglandin E synthase (PTGES) (a), and 3-hydroxy-3-methyl-glutaryl-coenzyme A reductase (HMGCR), mevalonate kinase (MVK), squalene epoxidase (SQLE), and sterol regulatory element-binding transcription factor 2 (SREBF2) (b). Data mining of The Cancer Genome Atlas (TCGA) clear cell renal cell carcinoma patient cohort (KIRC) using GEPIA2 for HMGCR, MVK, and SQLE genes (c).

Figure 11

Analysis of gene expression by real time PCR of CD36, caveolin 1 (CAV1) and low-density lipoprotein receptor (LDLR) (a), and perilipin 2 (PLIN2), hypoxia inducible lipid droplet-associated (HILPDA), and carnitine palmitoyltransferase 1A (CPT1A) (b). Representative images of normal and neoplastic kidney tissue (ccRCC) and normal cortical and ccRCC primary cell cultures captured after Oil Red O (ORO) staining at original magnification of 200×. Scale bars = 100 µm (c).

Figure 12

Median levels of serum cholesterol in ccRCC patients stratified according to pathological stage, Fuhrman grade, lymph node involvement, and visceral metastases (a). Kaplan-Meier cancer-specific survival (CSS) curves, stratified by serum cholesterol levels in the overall population and in a subset of patients with localized disease (b). Small squares indicate outlier in box-and-whisker plots.