Clinical Relevance and Drug Modulation of PPAR Signaling Pathway in Triple-Negative Breast Cancer: A Comprehensive Analysis

Abstract

Triple-negative breast cancer (TNBC) is highly heterogeneous and poses a significant medical challenge due to limited treatment options and poor outcomes. Peroxisome proliferator-activated receptors (PPARs) play a crucial role in regulating metabolism and cell fate. While the association between PPAR signal and human cancers has been a topic of concern, its specific relationship with TNBC remains unclear. Integrated analysis of large published datasets from clinical cohorts and cell lines through databases has proven to be a powerful and essential approach for understanding cancer and uncovering new molecular targets. Here, we conducted a comprehensive study investigating the clinical relevance and drug modulation of the PPAR signaling pathway in TNBC, using data from The Cancer Genome Atlas (TCGA) for TNBC patients and Genomics of Drug Sensitivity in Cancer (GDSC) for TNBC cell lines, along with drug perturbation information from Connectivity Map (CMap). In the TCGA-TNBC cohort, higher PPAR signaling activity was not associated with clinical stage, prognosis, tumor mutational burden, microsatellite instability, homologous recombination deficiency, stemness, or proliferation status. However, it was linked to older age; an elevated rate of piccolo presynaptic cytomatrix protein (PCLO) mutations; and oncogenic signal transduction involving MAPK, Ras, and PI3K-Akt pathways. Additionally, it influenced biological pathways including fatty acid metabolism, AMPK signaling, and ferroptosis. Strikingly, higher PPAR activity appeared to promote the formation of an antitumor immune and microbial microenvironment. In the GDSC-TNBC cells, nevertheless, it seemed to incur chemoresistance. Furthermore, we identified a batch of potential compounds that can regulate the PPAR signaling pathway. Lastly, our experimental validation demonstrated the ability of the histone deacetylase (HDAC) inhibitor chidamide to activate the PPAR signal in TNBC cells. In conclusion, the PPAR signaling pathway likely has pleiotropic biological effects in TNBC. These preliminary but interesting findings enhance our understanding of the role played by PPAR signal and provide new insights into the heterogeneity driven by it in TNBC.

Figure 1

Genetic landscapes and clinical significance of altered PPAR signaling in TCGA-TNBC cohort. (a) Expression pattern of genes in the PPAR signaling pathway and overall PPAR activity (GSVA score) in 190 primary tumors and 113 normal tissues. (b) Association between PPAR activity and age or clinical stage, with significance determined by the Wilcoxon rank-sum test. (c) Relationship between PPAR activity and overall survival, with significance determined by the Kaplan–Meier analysis and log-rank test. (d) Waterfall plots showing the top 20 genes with the highest mutation frequencies in the high and low PPAR activity groups. (e) Difference in mutation frequency of PCLO between the high and low PPAR activity groups, with significance determined by Fisher's exact test. (f) Pearson's correlations between PPAR activity and TMB, MSI, HRD, stemness indices mDNAsi and mRNAsi, or the expression of MKI67, PCNA, and MCM2. ⁣^∗p < 0.05, ⁣^∗∗p < 0.01, ns: not significant.

Figure 2

Connections between PPAR activity and oncogene expression or KEGG pathway activity in TCGA-TNBC cohort. (a) Volcano plot displaying the transcriptomic changes between the high and low PPAR activity groups with the differentially expressed oncogenes highlighted. (b) Difference in expression of oncogenes between the high and low PPAR activity groups, with significance determined by the Wilcoxon rank-sum test. ⁣^∗∗p < 0.01, ⁣^∗∗∗p < 0.001. ORA-based KEGG pathway enrichment analysis for (c) differential oncogenes and (d) up- or downregulated genes between high and low PPAR activity groups.

Figure 3

Relationships between PPAR activity and tumor immune microenvironment in TCGA-TNBC cohort. (a) Pearson's correlations between PPAR activity and StromalScore, ImmuneScore, ESTIMATEScore, or tumor purity. (b) Immune cell composition assessed by TIMER and CIBERSORT algorithms. (c) Pearson's correlations between PPAR activity and infiltration scores of dendritic cells, neutrophils, macrophages, M1 macrophages, activated CD4+ memory T cells, or naive B cells. (d) Pearson's correlations between PPAR activity and immune-related gene expression. (e) GSEA plots of nine immune-related pathways significantly enriched in the high compared to low PPAR activity group. (f) Pearson's correlations between PPAR activity and scores of TCR Shannon or CYT. ⁣^∗p < 0.05, ⁣^∗∗p < 0.01, ⁣^∗∗∗p < 0.001.

Figure 4

Associations between PPAR activity and TME-resident microorganisms in TCGA-TNBC cohort. (a) Distribution pattern of 1406 microorganisms in 190 TNBC samples. (b) Pearson's correlations between PPAR activity and scores of Chao1, ACE, Shannon, or Simpson. ns: not significant. (c) Pearson's correlations between PPAR activity and specific microorganism content.

Figure 5

Drug sensitivity correlation of PPAR activity and prediction of drugs modulating it. (a) Pearson's correlations between PPAR activity and IC50 values of specific drugs in GDSC-TNBC cells. ⁣^∗p < 0.05, ⁣^∗∗p < 0.01. (b) Volcano plot displaying the transcriptomic changes of 938 chemoresistance-related genes between the high and low PPAR activity groups in TCGA-TNBC cohort. (c) Pearson's correlations among the gene expression profiles of 6 CMap-derived non-TNBC and 20 GDSC-derived TNBC cell lines. (d) Identified potential compounds that may regulate the PPAR signaling pathway in TNBC by CMap analysis.

Figure 6

Upregulation of the PPAR signaling pathway by HDAC inhibitor chidamide in human TNBC MDA-MB-231 cells. (a) Volcano plot displaying the transcriptomic changes in MDA-MB-231 cells following chidamide treatment with the genes involved in the PPAR signaling pathway highlighted. Dots representing 25 upregulated, 9 downregulated, and 25 nondifferential genes were colored in red, blue, and gray, respectively. (b) Heat map displaying the expression patterns of 25 upregulated and 9 downregulated genes involved in the PPAR signaling pathway from (a). (c) Pathview mapping result of genes involved in the PPAR signaling pathway from (a), with rectangles indicating upregulated, downregulated, and nondifferential genes filled in red, blue, and gray, respectively. (d) qRT-PCR analysis for 10 representative genes, with significance determined by Student's t-test. ⁣^∗∗p < 0.01 and ⁣^∗∗∗p < 0.001. (e) Western blot analysis for PPARα and CPT1α.