Transcriptomic exploration of key genes related to mitochondria and ferroptosis in inflammatory bowel disease and experimental validation

Inflammatory bowel disease (IBD) is a systemic condition with multifactorial origins. Mitochondria and ferroptosis have been implicated in the disease, but their potential interaction in IBD remains unclear. Therefore, investigating the relationship between IBD and these factors, along with elucidating the underlying mechanisms, is essential. Differentially expressed genes (DEGs) from the GSE75214 dataset were intersected with mitochondria-related genes (MRGs) and ferroptosis-related genes (FRGs) retrieved from literature to identify candidate genes. Key genes were finalized through machine learning, receiver operating characteristic (ROC) analysis, and expression validation. Subsequently, an artificial neural network (ANN) disease prediction model based on these key genes was constructed and evaluated. Further analyses included enrichment analysis of key genes, immune infiltration analysis, chromosomal localization, and drug prediction. Finally, the expression levels of key genes were validated using reverse transcription quantitative real-time PCR (RT-qPCR). Four genes (AQP8, ACSF2, ACSL4, and IL1B) were identified as key genes. The ANN model demonstrated a strong association between these key genes and IBD, with reliable performance (GSE75214: area under the curve (AUC) = 0.855; GSE59071: AUC = 0.859). Additionally, adaptive immune response, tricarboxylic acid (TCA) cycle, and inositol phosphate metabolism were found to be associated with these key genes in IBD. Specifically, ACSL4 showed the strongest positive correlation with immature dendritic cells (correlation coefficient (cor) = 0.93, p < 0.01), while ACSF2 exhibited the strongest negative correlation with effector memory CD8⁺ T cells (cor = -0.81, p < 0.01). The key genes were localized to different chromosomes. Potential therapeutic drugs for IBD, such as 1-methyl-3-isobutylxanthine, were identified through target drug prediction. RT-qPCR experiments confirmed that AQP8 and ACSF2 expression levels were significantly reduced in IBD (p < 0.01), whereas ACSL4 and IL1B expression levels were significantly increased (p < 0.05). AQP8, ACSF2, ACSL4, and IL1B were identified as key genes in IBD, providing a foundation for future therapeutic research on this disease.