A multi-omics exploration of PPARG activation in colon cancer: kinases featuring a PPRE sequence within regulatory regions

Abstract

Background: As members of the nuclear receptor (NR) family of transcription factors, peroxisome proliferator-activated receptors (PPARs) regulate essential cellular processes, including lipid metabolism, glucose uptake, cell proliferation, and programmed cell death through ligand-mediated activation. Within the PPAR subfamilies, PPAR-γ (PPARG) is crucial to the development of fat cells, sensitivity to insulin, apoptosis, and metastasis. Furthermore, it demonstrates properties that counteract fibrosis and inflammation, thus establishing itself as a notable target for therapeutic interventions against conditions such as type 2 diabetes and cancer. PPARG is reported to be a promising target for patients diagnosed with colorectal cancer (CRC). Globally, colorectal cancer ranks as the third most prevalent malignancy and is responsible for approximately 10% of all cancer mortalities, and PPARG is significantly expressed in 70% of the sporadic CRC. In individuals with CRC, the precise function of PPARG remains not entirely comprehended and elucidation of the PPARG transcriptional regulation in CRC seems promising.

Results: This study integrates RNA-seq and ChIP-seq reads to analyze the effects of Rosiglitazone on HT-29 colon cancer cells. Peak calling analysis from ChIP-seq data identified 14,000 to 34,000 binding sites for PPARG across different experimental conditions. RNA-seq analysis highlighted significant differential gene expression in Rosiglitazone-treated cells, with 4362 and 6780 genes significantly regulated at 24 and 48 h, respectively. The correlation of these datasets with PPRE-associated kinases resulted in the identification of 18 differentially expressed genes (DEGs), followed by subsequent analysis of gene ontology, pathway enrichment, and protein-protein interactions, culminating in the elucidation of seven hub genes (PTK2, HGS, CDK8, PRPF6, PRKDC, PRKCZ, MET). Further these hub genes correlated with CRC progression and patient survival. Validation using independent GEO datasets (GSE113513 and GSE210693) and gene effect scores derived from CRISPR knockout screens further supported the functional impact of these hub genes. Disease ontology and mutational analyses implicated the hub genes in various cancers, including CRC. Moreover, miRNA analysis identified 37 experimentally validated miRNAs potentially modulating hub gene expression.

Conclusions: These findings advance our understanding of PPARG's regulatory network and underscore its potential as a therapeutic target, establishing a robust framework for future research in PPARG-related pathways.

Keywords: ChIP-seq; Gene enrichment; Kinase; Mi-RNA; Network analysis; Peroxisome Proliferator-Activated Receptors-γ; R studio; RNA-seq.

Fig. 1

Peak distribution of all four ChIP-seq samples. 13.59–16.65% of peaks were located near the gene promoter (≤ 1 Kb), 21.54–22.24% of the peaks were located in the distal intergenic region, and 21.51–23.36% of the peaks were located in the intronic region

Fig. 2

Volcano plot of the differentially expressed genes (DEGs) A at 24 h of Rosiglitazone-treated HT-29 cells. B 48 h of Rosiglitazone-treated HT-29 cells. (NS- non-significant)

Fig. 3

Top 50 differentially expressed genes (DEGs) (based on padj values) both in A 24 h and B 48 h Rosiglitazone treated cells

Fig. 4

Bar graph for differential expression of the kinases, overlapping between PPRE-associated genes and the DEGs (P_adj ≤ 0.5) in the RNA-seq dataset

Fig. 5

Gene Ontology (GO) enrichment of the DEGs at both 24 and 48 h, and 48 h only

Fig. 6

Pathway analysis of the DEGs at 24 and 48 h: The genes are substantially involved in the m-TOR signaling pathway, PDGF, GMCSF, IFN-gamma, IL-5, and IL-3 mediated signaling pathways

Fig. 7

The cytoHubba interface integrated into the Cytoscape platform was utilized and the foremost 10 hub genes were outlined based on closeness, degree, and betweenness

Fig. 8

Box plots of relative expression level of the hub genes in normal and different cancer stage of colon adenocarcinoma (COAD)

Fig. 9

The impact of identified hub gene expression level on the overall survival of CRC patients: Low expression is indicated in black colour while red colour represents the high gene expression

Fig. 10

(A) The hub genes expression validated with microarray data containing samples from non-cancerous and CRC patients. All the hub genes expressed differential regulation in cancerous samples as compared to non-cancerous controls (*** represents padj ≤ 0.05). (B) On HT-1197 cells, T0070907, a PPARG antagonist, exhibits a reversal in the expression of the hub genes compared to the rosiglitazone treatment

Fig. 11

Heatmap of the hub genes and disease associations (p ≤ 0.05) was constructed using the DOSE package in R Studio. It was observed that the genes are involved in stomach, breast, lung, pancreas, prostate carcinoma, adenoma, glioma, severe combined immunodeficiency, Wiskott-Aldrich syndrome, and ACTH-secreting pituitary adenoma