Systems biology identifies sentinel biomarkers bridging ankylosing spondylitis and myocardial infarction

Ankylosing spondylitis (AS) is increasingly recognized as an independent risk factor for premature myocardial infarction (MI), yet the molecular bridge linking chronic axial inflammation to acute coronary events remains poorly mapped. Mining four public microarray cohorts (GSE128470, GSE73754, GSE100927, GSE122897) that profile peripheral blood mononuclear cells from AS patients, MI patients and healthy controls, integrative bioinformatics now delivers a concise pathogenic blueprint. Weighted gene co-expression network analysis identified one AS-related and one MI-related module that significantly overlap; machine-learning (LASSO + SVM-RFE) distilled these to two hub genes—S100A12 and MCEMP1—whose transcript levels rise concordantly across both diseases. ROC curves yield AUCs of 0.92-0.96 for distinguishing AS-MI cases from either disease alone, and a nomogram incorporating age, CRP and the two hubs achieves a net reclassification improvement of 34 %.

Functional dissection places S100A12 and MCEMP1 at the crossroads of damage-associated molecular pattern signalling, neutrophil extracellular trap formation and cholesterol-laden macrophage activation. GSEA shows enrichment for "NOD-like receptor signalling", "IL-17 pathway" and "platelet degranulation"; upstream regulator analysis predicts NF-κB and STAT3 as master transcription factors. Single-cell RNA-seq of coronary plaques from two explanted AS hearts confirms selective expression of both genes in CD14⁺ classical monocytes and CD16⁺ non-classical monocytes, with negligible signal in T cells or fibroblasts. Pseudotime reconstruction suggests a trajectory from resting monocyte → inflammatory macrophage → lipid-associated macrophage under dual AS-MI cues.

Immune infiltration algorithms (CIBERSORT, xCell) reveal that AS-MI peripheral blood harbours elevated neutrophils (35 % vs. 20 % controls), reduced resting CD4⁺ T cells and a compensatory expansion of myeloid-derived suppressor cells. Expression of S100A12 correlates most strongly with neutrophil fraction (r = 0.78), whereas MCEMP1 tracks the non-classical monocyte signature (r = 0.81). Bulk ATAC-seq footprinting further indicates that S100A12 is epigenetically primed by open chromatin at two NF-κB motifs, while MCEMP1 is suppressed by a miR-223-3p binding site that is lost in AS-MI. A TF-miRNA-gene network constructed using ChIP-seq and miRWalk databases couples 24 transcription factors (RUNX1, CEBPB, FOS) and nine miRNAs (miR-155-5p, miR-146a-5p, miR-223-3p) to the two hubs, offering 48 potential intervention nodes.

Drug repositioning screens against the hub-gene signature flagged three clinically approved compounds: enzastaurin (PKCβ inhibitor) down-regulates S100A12 via NF-κB blockade; meglitinide suppresses MCEMP1 transcription by interrupting STAT3 recruitment; and the calcium-channel blocker nifedipine indirectly attenuates both genes through modulation of oxidative stress pathways. Molecular docking shows high affinity (ΔG < −8.5 kcal mol⁻¹) of enzastaurin to the S100A12 promoter pocket and of meglitinide to the MCEMP1 3′-UTR miR-223 target site. In silico validation using LINCS L1000 signatures corroborates reversal of the AS-MI transcriptional program within 24 h of compound exposure.

Collectively, the findings frame S100A12 and MCEMP1 as sentinel biomarkers that integrate chronic systemic inflammation (AS) with acute coronary thrombo-inflammation (MI). Their myeloid-restricted expression, robust diagnostic performance and druggable upstream circuitry nominate them for risk stratification and as molecular targets repurposing existing drugs to prevent MI in AS patients.