The lipidomic profile of the tumoral periprostatic adipose tissue reveals alterations in tumor cell's metabolic crosstalk

Abstract

Background: Periprostatic adipose tissue (PPAT) plays a role in prostate cancer (PCa) progression. PPAT lipidomic composition study may allow us to understand the tumor metabolic microenvironment and provide new stratification factors.

Methods: We used ultra-high-performance liquid chromatography-mass spectrometry-based non-targeted lipidomics to profile lipids in the PPAT of 40 patients with PCa (n = 20 with low-risk and n = 20 high-risk). Partial least squares-discriminant analysis (PLS-DA) and variable importance in projection (VIP) analysis were used to identify the most relevant features of PPAT between low- and high-risk PCa, and metabolite set enrichment analysis was used to detect disrupted metabolic pathways. Metabolic crosstalk between PPAT and PCa cell lines (PC-3 and LNCaP) was studied using ex vivo experiments. Lipid uptake and lipid accumulation were measured. Lipid metabolic-related genes (SREBP1, FASN, ACACA, LIPE, PPARG, CD36, PNPLA2, FABP4, CPT1A, FATP5, ADIPOQ), inflammatory markers (IL-6, IL-1B, TNFα), and tumor-related markers (ESRRA, MMP-9, TWIST1) were measured by RT-qPCR.

Results: Significant differences in the content of 67 lipid species were identified in PPAT samples between high- and low-risk PCa. PLS-DA and VIP analyses revealed a discriminating lipidomic panel between low- and high-risk PCa, suggesting the occurrence of disordered lipid metabolism in patients related to PCa aggressiveness. Functional analysis revealed that alterations in fatty acid biosynthesis, linoleic acid metabolism, and β-oxidation of very long-chain fatty acids had the greatest impact in the PPAT lipidome. Gene analyses of PPAT samples demonstrated that the expression of genes associated with de novo fatty acid synthesis such as FASN and ACACA were significantly lower in PPAT from high-risk PCa than in low-risk counterparts. This was accompanied by the overexpression of inflammatory markers (IL-6, IL-1B, and TNFα). Co-culture of PPAT explants with PCa cell lines revealed a reduced gene expression of lipid metabolic-related genes (CD36, FASN, PPARG, and CPT1A), contrary to that observed in co-cultured PCa cell lines. This was followed by an increase in lipid uptake and lipid accumulation in PCa cells. Tumor-related genes were increased in co-cultured PCa cell lines.

Conclusions: Disturbances in PPAT lipid metabolism of patients with high-risk PCa are associated with tumor cell metabolic changes.

Keywords: De novo fatty acid synthesis; Lipid metabolism; Lipidomic; Periprostatic adipose tissue; Prostate cancer.

Fig. 1

Identified lipid species in PPAT samples. A Total fatty acids. Error bars represent standard deviation (SD) of the mean. ***p < 0.001. B Other annotated lipids divided by lipid category. Box shows the median, quartiles, and extreme values. *p < 0.05. Abbreviations: SFA, saturated fatty acid; MUFA, monounsaturated fatty acids; ω-3 PUFA, omega 3 polyunsaturated fatty acids; ω-6 PUFA, omega 6 polyunsaturated fatty acids; DG, diglyceride; TG, triglyceride; SM, sphingomyelin; LPI, lysophosphatidylinositol; LPE, lysophosphatidylethanolamine; LPC, lysophosphatidylcholine; PC, phosphatidylcholine (diacylglycerol); PC-O, phosphatidylcholine (alkyl-acyl-glycerol)

Fig. 2

Lipid analysis of PPAT. A Principal component analysis score plot of the 16 selected signatures. B Variance explanation (%) for each principal component is indicated. C Variable Importance of projection (VIP) analysis of the selected lipid signatures with VIP scores ≥ 1. Abbreviations: 13-oxoODE/9oxoODE, 13-Oxo-9,11-octadecadienoic acid/9-oxo-10E,12Z-octadecadienoic acid; 12,13-EpOME, vernolic acid; MG, monoglyceride, TG, triglyceride; LPI, lysophosphatidylinositol; ω-6 PUFA, omega 6 polyunsaturated fatty acids; SFA, saturated fatty acid

Fig. 3

Heat map of the clustering analysis. Samples of PPAT categorized by ISUP group naturally clustered into experimental studied groups: low-risk (group I) and high-risk (groups III, IV, and V). Metabolite intensities are displayed as colors ranging from red grading (more abundant metabolites) to blue grading (less abundant metabolites). Classification of lipid according to extraction method: FAME: ω-6 PUFA, SFA, total PUFA, and behenic acid. LIP-I: 13-OxoODE/9oxoODE; 12,13-EpOME; LPI, lauric acid and linoleic acid. LIP-II: MG, and TG

Fig. 4

Biochemical metabolic analysis. A Metabolic set enrichment analysis showing altered lipid signatures analyzed by SMPDB metabolite set library. B, C Significantly altered lipid metabolite box plots allocated to related pathways. Patients were stratified regarding ISUP group. Statistic comparation was performed between high-risk (groups III, IV, and V) and low-risk (group I). Each box shows the median, quartiles, and extreme values. *p < 0.05, **p < 0.01

Fig. 5

Analysis of gene expression of PPAT samples. A–D PPAT displays an altered expression profile of lipid metabolism-related genes and E inflammatory-related genes. Each box shows the median, quartiles, and extreme values of n = 20 samples

Fig. 6

A bidirectional crosstalk occurs between PPAT explants and PCa cell lines. PCa cell lines (PC-3 or LNCaP) were co-cultivated with PPAT explants for 48 h and expression of lipid-, inflammatory-, and tumor-related genes were evaluated by RT-qPCR. A Co-cultured explants with LNCaP or PC-3 cells. B Co-cultured PC-3 cells. C Co-cultured LNCaP cells. D Uptake of the fatty acid analog TF2-C12 at 30 and 60 min in PCa cell lines co-cultured with PPAT explants. E Lipid accumulation measured with Nile Red dye after 15-min incubation in PCa cell lines co-cultured with PPAT explants. Different lettering over the boxes indicates statistical differences. Significant differences are established at p < 0.05. Data are expressed as mean ± SEM (n = 6 experiments)

Fig. 7

Scheme illustrating lipid metabolite alterations, related pathways, and implicated genes in human PPAT and PCa cell lines. In “aggressive” PPAT adipocytes, lower concentrations of 13-oxoODE, an endogenous ligand for PPARG, may provoke transcriptional upregulation of the proinflammatory cytokines IL-6, IL-1B, and TNFα. De novo lipogenesis is also reduced, as the expression levels of ACACA and FASN are downregulated. The resulting process translated to reduced concentrations of palmitic acid and its saturated FA intermediates (produced by elongation process): stearic acid, arachidic acid, behenic acid, and lignoceric acid. FAs from TG lipolysis can also be contemplated as another FA source. Reduced expression levels of lipase LIPE and the mitochondrial transporter CPT1A may indicate suppressed FA lipolysis. An increase in the expression of the FABP4 lipid transporter gene points to active FA mobilization to PCa cells. PCa cells can take-up the FA released from PPAT adipocytes followed by an increased expression of lipid transporter genes (CD36, FATP5), fuelling cellular de novo lipogenesis and lipolysis. An active proinflammatory state is also observed in PCa cells. Abbreviations: 13-HODE, 13 hydroxy-octadecadienoic acid; 13-oxoODE, 13-oxo-octadecadienoic; PPARG, peroxisome proliferator-activated receptor gamma; IL-6, interleukin 6; IL-1B, interleukin 1 B; TNFα, tumor necrosis factor alpha; FASN, fatty acid synthase; SREBP-1, sterol regulatory element binding transcription factor 1; ACACA, acetyl-CoA; CPT1A, carboxylase alpha; Carnitine Palmitoyl transferase 1A; LIPE, hormone-sensitive lipase; MG, monoglyceride; DG, diglyceride; TG, triglyceride; FABP4, fatty acid binding protein 4; FA, fatty acid; PNPLA2, patatin-like phospholipase domain containing 2; CD36, CD36 molecule; LPL, lipoprotein lipase