Integration of Multi-Omics and Machine Learning Identifies TGFB1 and SERPINE1 as Biomarkers of Vascular Smooth Muscle Cell Senescence in Intracranial Aneurysms.

Single-cell RNA sequencing analysis of a mouse elastase-induced IA model revealed 26 clusters across 10 cell types, with IA samples showing VSMC depletion, immune cell enrichment, and fibroblast expansion.

• Data were derived from GSE193533, a mouse elastase-induced IA model
• Cell clustering identified 26 distinct clusters spanning 10 cell types
• IA samples specifically demonstrated VSMC depletion compared to controls
• Immune cell enrichment and fibroblast expansion were observed in IA samples relative to controls

VSMCs were categorized into five subsets (VSMC1-5), with the extracellular matrix remodeling and synthetic-inflammatory subtype (VSMC1) significantly increased in IA.

• Five distinct VSMC subsets were identified through subclustering analysis
• VSMC1 was characterized as an extracellular matrix remodeling and synthetic-inflammatory subtype
• VSMC1 proportion was significantly increased in IA samples compared to controls
• Pseudotime trajectory inference was used to analyze VSMC subset relationships

Differential expression analysis of VSMCs intersecting with senescence-associated genes from the Aging Atlas database identified 159 senescence-associated genes (SAGs) enriched in inflammatory and apoptotic pathways.

• VSMC differential expression data were cross-referenced with the Aging Atlas database to identify SAGs
• 159 SAGs were identified at the intersection of VSMC differential expression and known senescence-associated genes
• These 159 SAGs were functionally enriched in inflammatory and apoptotic pathways
• Functional enrichment analysis was performed as part of the analytical pipeline

Integration of hdWGCNA and bulk transcriptomics narrowed the candidate senescence-associated genes to 45 key SAGs.

• High-dimensional weighted gene co-expression network analysis (hdWGCNA) was applied to single-cell data
• Results were integrated with human bulk RNA sequencing and microarray datasets
• The integration reduced the 159 candidate SAGs to 45 key SAGs
• Both bulk RNA sequencing and microarray datasets from human samples were incorporated in the analysis

Machine learning identified TGFB1 and SERPINE1 as robust biomarkers of senescence-associated VSMCs in intracranial aneurysms.

• Machine learning algorithms were applied to the 45 key SAGs to identify candidate biomarkers
• TGFB1 and SERPINE1 emerged as the top candidate biomarkers from the machine learning analysis
• Both genes showed high diagnostic accuracy with AUC > 0.75
• Validation was performed using external datasets and immunohistochemical analysis of human IA tissues
• Both TGFB1 and SERPINE1 were upregulated in IA tissues compared to controls

VSMC senescence is implicated in intracranial aneurysm pathogenesis, with senescence-associated processes including cell-cell communication changes observed in IA.

• Cell-cell communication analysis was conducted as part of the multi-omics integration pipeline
• VSMC senescence had not been previously explored in IA pathogenesis according to the authors
• Senescence-associated genes were enriched in inflammatory and apoptotic pathways relevant to IA progression
• The study provides evidence that VSMC senescence may contribute to IA progression