Abstract Type 1 diabetes (T1D) is a genetic autoimmune disease in children and young adults, with no cure, emphasizing the need for novel genetic-based therapies. Differential analysis using the “Limma” package identified 42 upregulated genes in the peripheral blood of T1D patients. Gene Ontology/Kyoto Encyclopedia of Genes and Genomes analysis suggested that CD38 may contribute to T1D development. LASSO (least absolute shrinkage and selection operator) machine learning successfully screened EMOES, SH2D1B, and TRDV3 from these upregulated genes. We selected SH2D1B, which is included in our natural killer cell–mediated cytotoxicity-related genes gene set, for further protein interaction network analysis, the analysis results revealed a highly correlated network of LY9, SH2D1B, CD244, and SLAMF6. Through immune infiltration analysis, we demonstrated resting natural killer (NK) cells and monocyte migration were upregulated, both were significantly related to the expression levels of LY9, SH2D1B, and CD244. Single-cell RNA sequencing analysis showed that both SH2D1B and CD244 were enriched in NK cells, and SH2D1B has the potential to modulate the function of NK cells. Subsequently, We performed gene set enrichment analysis with SH2D1B grouped into SH2D1Bhigh NK cells, SH2D1Blow NK cells, and SH2D1B- NK cells, which showed that the gene set related to the response to IFNγ is enriched in SH2D1Bhigh NK cells. With the analysis results, we performed differential expression analysis and validated the results by Western blotting. Finally, we identified activation of the IFNγ/JAK1/STAT/CD38 pathway in SH2D1Bhigh NK cells. Our findings suggest that SH2D1Bhigh NK cells may produce adenosine by IFNγ/JAK1/STAT/CD38, to limit autoimmunity and presumably disease development.
Chen et al. (Sun,) studied this question.