Gene expression signatures from single-cell transcriptomics predict Sjögren's syndrome

Sjögren's syndrome (SjS) is a systemic autoimmune disorder characterized by lymphocytic infiltration of exocrine glands, leading to dry eyes and mouth. While previous genome-wide association studies (GWAS) and transcriptomic analyses have identified genes associated with SjS, predictive models based on single-cell resolution are limited. In this study, single-cell RNA sequencing (scRNA-seq) data from peripheral blood mononuclear cells (PBMCs) of SjS patients were analyzed to map immune cell alterations linked to the disease. Compared with healthy controls, SjS patients displayed decreased proportions of naïve CD8⁺ T cells and Helios FOXP3^lo CD4⁺ Tregs, alongside increased frequencies of CTLA−4⁺ CD8⁺ inhibitory T cells and TRDC γδ T cells. Using machine learning, a predictive model for SjS diagnosis was developed based on a 12-gene signature (SjS. Sig: GIMAP7, PSMB8, CD27, CCR7, TAGAP, UQCR10, HCLS1, LCK, TNFAIP3, ISG15, GIMAP4, and HLA-DRB1), which effectively differentiated patients from healthy individuals. Key genes such as CD27, PSMB8, HCLS1, LCK, UQCR10, and GIMAP4 were validated in clinical samples through flow cytometry and real-time quantitative PCR. These findings provide insights into the immune landscape of SjS at a single-cell resolution and propose a reliable molecular signature for diagnosis and immune monitoring.