Profiling Shared Cytotoxic Immune Signatures in SLE-Associated Coronary Injury Through Transcriptomics and Machine...

This study investigated shared molecular pathways linking systemic lupus erythematosus (SLE) and coronary artery disease (CAD) to uncover mechanisms of coronary injury in SLE.

Bulk transcriptomic datasets (GSE45291 for SLE, GSE61145 for CAD) were analyzed to identify differentially expressed genes (DEGs), immune cell infiltration patterns, and co-expression networks. A diagnostic model was constructed and validated using external cohorts (GSE49454 for SLE, GSE179789 for CAD). Machine learning prioritized core genes, validated in both external cohorts and in SLE patients with coronary injury (GSE264125). Cellular localization and intercellular communication were explored by analyzing single-cell RNA-seq data (GSE135779). qPCR was used to validate the gene expression in peripheral blood mononuclear cells (PBMCs) from patients.

We identified 146 common DEGs enriched in immune pathways related to cell toxicity, and found shared dysregulation in cytotoxic lymphocytes such as natural killer (NK) cells and CD8⁺ T cells. Through co-expression analysis and DEG intersection, we pinpointed 11 hub genes (eg, GZMK, KLRK1, GNLY). A diagnostic model based on these genes showed strong performance (SLE: AUC 0.881 training, 0.666 validation; CAD: AUC 0.897 training, 0.781 validation). Machine learning highlighted GZMK and KLRK1 as core genes, which were further validated for their combined diagnostic utility (AUC: 0.782-1.000) in SLE-related coronary injury. Single-cell analysis revealed that these genes are primarily active in cytotoxic CD8⁺ T cells and NK cells, with GZMK linked to CLEC-mediated signaling and KLRK1 to HLA activation. Finally, we confirmed higher expression of these genes in blood cells from SLE patients with coronary artery disease using qPCR.

SLE and CAD share a cytotoxic lymphocyte-driven molecular axis, with GZMK/KLRK1-mediated immune dysregulation as a key contributor to coronary injury in SLE. GZMK and KLRK1 may represent promising biomarkers for early detection and risk stratification of SLE-associated coronary complications. Notably, the discrimination (AUC=1.000) was observed in a limited subgroup of SLE patients with coronary microvascular dysfunction (n=4), warranting further validation in expanded cohorts.