Xuanlin Wu, Tao Pan, Zhihao Fang, Titi Hui, Xiaoxiao Yu, Changxu Liu, Zihao Guo, Chang Liu

Department of General Surgery, Fourth Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang, People’s Republic of China

Correspondence: Chang Liu, Department of General Surgery, Fourth Affiliated Hospital of Harbin Medical University, 37 Yiyuan Street, Nangang District, Harbin, Heilongjiang, 150001, People’s Republic of China, Tel +86-13313699697, Email lc19726666@163.com

Background: Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD), as a common chronic liver condition globally, is experiencing an increasing incidence rate which poses significant health risks. Despite this, the detailed mechanisms underlying the disease’s onset and progression remain poorly understood. In this study, we aim to identify effective diagnostic biomarkers for MASLD using microarray data combined with machine learning techniques, which will aid in further understanding the pathogenesis of MASLD.
Methods: We collected six datasets from the Gene Expression Omnibus (GEO) database, using five of them as training sets and one as a validation set. We employed three machine learning methods—LASSO, SVM, and Random Forest (RF)—to identify hub genes associated with MASLD. These genes were further validated using the external dataset GSE164760. Additionally, functional enrichment analysis, immune infiltration analysis, and immune function analysis were conducted. A TF-miRNA-mRNA network was constructed, and single-cell RNA sequencing was used to determine the distribution of key genes within key cell clusters. Finally, the expression of the key genes was further validated using the palmitic acid-induced AML-12 cell line and the MCD mouse model.
Results: In this study, through differential gene expression (DEGs) analysis and machine learning techniques, we successfully identified 10 hub genes. Among these, the key gene EGR1 was validated and screened using an external dataset, with an area under the curve (AUC) of 0.882. Enrichment analyses and immune infiltration assessments revealed multiple pathways involving EGR1 in the pathogenesis and progression of MASLD, showing significant correlations with various immune cells. Furthermore, additional cellular experiments and animal model validations confirmed that the expression trends of EGR1 are highly consistent with our analytical findings.
Conclusion: Our research has confirmed EGR1 as a key gene in MASLD, providing novel insights into the disease’s pathogenesis and identifying new therapeutic targets for its treatment.

Keywords: metabolic dysfunction-associated steatotic liver disease, immune infiltration, machine learning, TF-miRNA-mRNA network, EGR1