Developing a novel immune infiltration-associated mitophagy prediction model for amyotrophic lateral sclerosis using bioinformatics strategies

Abstract

Background: Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease, which leads to muscle weakness and eventual paralysis. Numerous studies have indicated that mitophagy and immune inflammation have a significant impact on the onset and advancement of ALS. Nevertheless, the possible diagnostic and prognostic significance of mitophagy-related genes associated with immune infiltration in ALS is uncertain. The purpose of this study is to create a predictive model for ALS using genes linked with mitophagy-associated immune infiltration. Methods: ALS gene expression profiles were downloaded from the Gene Expression Omnibus (GEO) database. Univariate Cox analysis and machine learning methods were applied to analyze mitophagy-associated genes and develop a prognostic risk score model. Subsequently, functional and immune infiltration analyses were conducted to study the biological attributes and immune cell enrichment in individuals with ALS. Additionally, validation of identified feature genes in the prediction model was performed using ALS mouse models and ALS patients. Results: In this study, a comprehensive analysis revealed the identification of 22 mitophagy-related differential expression genes and 40 prognostic genes. Additionally, an 18-gene prognostic signature was identified with machine learning, which was utilized to construct a prognostic risk score model. Functional enrichment analysis demonstrated the enrichment of various pathways, including oxidative phosphorylation, unfolded proteins, KRAS, and mTOR signaling pathways, as well as other immune-related pathways. The analysis of immune infiltration revealed notable distinctions in certain congenital immune cells and adaptive immune cells between the low-risk and high-risk groups, particularly concerning the T lymphocyte subgroup. ALS mouse models and ALS clinical samples demonstrated consistent expression levels of four mitophagy-related immune infiltration genes (BCKDHA, JTB, KYNU, and GTF2H5) with the results of bioinformatics analysis. Conclusion: This study has successfully devised and verified a pioneering prognostic predictive risk score for ALS, utilizing eighteen mitophagy-related genes. Furthermore, the findings indicate that four of these genes exhibit promising roles in the context of ALS prognostic.