Identification of fatty acid oxidation-related subtypes by integrated analysis of bulk- and single-cell transcriptome profiling in colorectal cancer.

Abstract

As one of the major metabolic reprogramming pathways, fatty acid oxidation (FAO) contributes to rapid progression in tumor cells. Nevertheless, the genomic patterns of patients' FAO levels in colorectal cancer (CRC) remain unknown. Hence, it is crucial to identify the interplay mechanisms of molecular biochemical features of FAO in CRC. Data of patients with CRC were accessed from The Cancer Genome Atlas (TCGA). Unsupervised consensus clustering related to FAO sores was conducted. The differentially expressed genes (DEGs) were screened by clustering according to FAO status polarized in TCGA, followed by the construction of the scores of genes related to FAO (GFAO_Score). Enrichment of FAO and carcinogenesis at the cell level were calculated based on the single-cell RNA (scRNA) sequencing analysis. The clinical values and drug analysis of GFAO_Score were evaluated by external validation cohorts from Gene Expression Omnibus (GEO) datasets. We classified patients into two distinct FAO clusters which indicated those with lower FAO levels had poor prognosis and high enrichment of carcinogenic-gene pathways. Further, the high FAO-enriched subtypes in epithelial cells revealed carcinogenesis. Three FAO-related genes (ZFHX4, AQP8, and AKR1B10) were screened to construct the GFAO_Score. The high GFAO_Score group leaned toward advanced CRC and unfavorable survival outcomes in the validation cohort. The low GFAO_Score group possessed a better response to immunotherapy and exhibited lower IC50 (50% inhibition concentration) values for certain chemotherapy drugs, such as 5-fluorouracil, irinotecan, oxaliplatin, paclitaxel, and camptothecin. FAO patterns vary in patients with CRC. The GFAO_Score might contribute to the precise screening of patients according to metabolism reprogramming and optimization of strategies in clinical practice.