Post-transcriptional Rac1 activation driven by lipid modification enzymes in atrial fibrillation: a multi-dataset bioinformatics study with experimental validation

Background Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, but the post-translational mechanisms linking mitochondrial dysfunction to arrhythmogenic signaling remain poorly understood. Whether the small GTPase Rac1—central to cardiomyocyte redox signaling—is activated through transcriptional upregulation or post-translational modification (PTM) has not been dissected. Methods We constructed a 77-gene Mitoxyperilysis Gene Set (MGS) and performed differential expression analysis in the discovery cohort GSE115574 (AF = 26, SR = 31; Affymetrix GPL570). ssGSEA scored four mechanistic arms: glutathione (GSH) metabolism, mitochondrial damage, RhoA axis, and mTORC2 axis. A 44-gene PTM enzyme panel was assessed across three datasets (GSE115574, GSE41177, GSE79768; AF = 29, SR = 5 and AF = 11, SR = 9 in the validation cohorts). Rac1 protein activation was examined via immunofluorescence (IF) in human right atrial tissue (AF = 3, SR = 3; TOM20/RAC1 dual staining, eight fields per sample). Results Among 170 differentially expressed genes, 3 MGS genes were significantly altered. The GSH metabolism arm was markedly downregulated in the discovery cohort (P = 1.57 × 10⁻⁴, FDR = 0.0012). The PTM enzyme panel revealed 30/44 genes significantly altered in ≥ 1 dataset, with 5 showing cross-dataset concordant upregulation (GGPS1, ZDHHC4, ZDHHC5, ZDHHC20, AKT1). RAC1 mRNA was not significantly different (log₂FC = 0.12, P = 0.34), and the transcriptional RAC1/RHOA ratio was consistently lower in AF across all three datasets (a negative finding). In human atrial tissue, Rac1 immunofluorescence showed a trend toward higher intensity (8.51 ± 2.95 vs 7.08 ± 2.00, P = 0.53) and membrane-to-cytoplasm ratio (0.93 ± 0.07 vs 0.84 ± 0.13, P = 0.40) in AF (pilot validation, n = 3/group); trimmed-mean and median analyses confirmed that these trends were not driven by bright spots. Conclusions AF atrial tissue exhibits concurrent "transcriptional RhoA relative dominance" and "post-translational Rac1 activation trend," which are not contradictory. Rac1 activation in AF appears to be modification-driven (geranylgeranylation/palmitoylation) rather than transcription-driven, warranting further validation by site-specific PTM assays.