Allograft kidney rejection can lead to graft dysfunction and graft loss. Protocol biopsy poses additional risk for recipients with normal renal function. The transcriptome of peripheral blood mononuclear cells (PBMCs) contains tremendous information and has potential application value for non-invasive diagnosis.
From the Gene Expression Omnibus database, we collected three datasets containing 109 rejected samples and 215 normal controls. After data filter and normalization, we performed deconvolution of bulk RNA sequencing data to predict cell type and cell-type specific gene expression. Subsequently, we calculated cell communication analysis by Tensor-cell2cell and conducted the least absolute shrinkage and selection operator (LASSO) logistic regression to screen the robust differentially expressed genes (DEGs). These gene expression levels were validated in mice kidney transplantation acute rejection model. The function of the novel gene ISG15 in monocytes was further confirmed by gene knockdown and lymphocyte-stimulated assay.
The bulk RNA-seq hardly predicted kidney transplant rejection accurately. Seven types of immune cells and transcriptomic characteristics were predicted from the gene expression data. The monocytes showed significant differences in amount and gene expression of rejection. The cell-to-cell communication indicated the enrichment of antigen presentation and T cell activation ligand-receptor pairs. Then 10 robust genes were found by Lasso regression and a novel gene ISG15 remained differential expression in monocytes between rejection samples and normal control both in public data and animal model. Furthermore, ISG15 also showed a critical role in promoting the proliferation of T cells.
This study identified and validated a novel gene ISG15 associated with rejection in peripheral blood after kidney transplantation, which is a significant non-invasive diagnosis and a potential therapeutic target.