Multi-Omics Analysis Combined with Machine Learning Identified FABP4 in Smooth Muscle Cells as a Pathogenic Factor in Atherosclerosis

Abstract

Yinyu Wang,1 Tingting Shang,1 Rui Cai,2 Cuiping Wang1 1Department of Cardiology, Affiliated Hospital of Jiangsu University, Zhenjiang, People’s Republic of China; 2Department of Cardiology, The Affiliated Yancheng Maternity & Child Health Hospital of Yangzhou University, Yancheng, People’s Republic of ChinaCorrespondence: Cuiping Wang, Department of Cardiology, Affiliated Hospital of Jiangsu University, No. 438 Jiefang Road, Jingkou District, Zhenjiang City, Jiangsu Province, People’s Republic of China, Email [email protected]: Atherosclerosis is the pathological basis of coronary heart disease, stroke, and peripheral arterial disease. Smooth muscle cells (SMCs) play a crucial role in atherosclerotic pathogenesis. However, effective drugs and therapy targeting SMCs for treating atherosclerosis are still lacking.Methods: We utilized single-cell RNA sequencing (scRNA-seq) (GSE155512 and GSE159677) and array data (GSE43292 and GSE125771) to identify Scissor+ SMCs (SMCs positively associated with atherosclerosis) and Scissor- SMCs (SMCs negatively associated with atherosclerosis) by using Scissor package. We analyzed their functional changes, cell-cell communication, and differentiation potential. Machine learning techniques were employed to analyze the marker in SMCs of atherosclerosis. qRT-PCR was used to examine the expression of these genes in MOVAS stimulated by ox-LDL. Potential inhibitors of the identified proteins were predicted, and their binding sites were analyzed.Results: We identified 475 Scissor+ SMCs and 1363 Scissor- SMCs. Functional enrichment analysis revealed that Scissor+ SMCs exhibited downregulation of Rho-related pathways, while pro-inflammatory pathways were upregulated. Cell-cell communication analysis indicated tighter interactions between SMCs and endothelial cells. Differential expression analysis identified 20 genes highly expressed in both scRNA-seq and array data. The LASSO regression, random forest, support vector machine and receiver operating characteristic curve suggested a strong correlation between fatty acid-binding protein 4 (FABP4) and atherosclerosis. The qRT-PCR results showed that FABP4 was highly expressed in MOVAS stimulated by ox-LDL. Drug prediction revealed that (S)-RP-6306 acted as an inhibitor, via forming a polar bond with Arg-126. In vitro experiments confirmed that (S)-RP-6306 significantly reduced the expression of FABP4.Conclusion: Scissor+ SMCs differed significantly from Scissor- SMCs in cellular function, cell-cell communication, and differentiation potential. The high expression of FABP4 in this subgroup of SMCs presented a promising therapeutic target for atherosclerosis, with (S)-RP-6306 showing potential as a drug targeting FABP4. Keywords: atherosclerosis, smooth muscle cell, machine learning, molecular docking

Keywords

• atherosclerosis
• smooth muscle cell
• machine learning
• molecular docking