Single-cell RNA sequencing shows the immune cell landscape in the kidneys of patients with idiopathic membranous nephropathy

Abstract

Idiopathic membranous nephropathy (IMN) is a leading pathological type of the adult primary nephrotic syndrome. Some patients develop end-stage renal disease due to poor response to treatment with steroid and immunosuppressive agents. In order to explore the molecular mechanism of IMN, we collected renal tissue samples from IMN patients and healthy controls and performed analysis by single-cell RNA sequencing (scRNA-seq). A total of 11 kidney cell clusters were identified, including multiple myeloid cell clusters, NK/T cell clusters, and B cell clusters. Most kidney parenchymal and immune cells were enriched in the regulation of immune response, inflammation, fibrosis and endoplasmic reticulum stress. The macrophage population in the IMN group showed a highly activated profile with up-regulated genes related to chemotaxis, inflammation, phagocytosis and fibrosis. CD8+ T cells continued to be cytotoxic in IMN; however, a transition to "inflammageing" GZMK+ CD8+ T cells was observed. The proportion of activated B cells in renal tissues of IMN patients was much higher than that of normal controls, indicating that B cells in IMN might be activated by constant antigenic stimulation. Moreover, the cell-cell interaction analysis revealed the potential communication between renal glomerular cells and immune cells in IMN. Overall, scRNA-seq was applied to IMN to unravel the characteristics of immune cells and elucidate possible underlying mechanisms involved in the pathogenesis of IMN.

Keywords: idiopathic membranous nephropathy; immune cell; pathogenesis; pro-inflammatory; single-cell RNA sequencing.

Figure 1

Single-cell RNA sequencing reveals the kidney cell populations in IMN and control subjects. (A) Overview of single-cell RNA sequencing and data processing. (B) Violin plot of marker genes that identified the eleven distinct cell clusters. (C) Heat map showing marker genes of each cell cluster. (D) Eleven distinct cell clusters identified by UMAP plotting from three IMN patients and four normal controls. (E) Bar plots represent the proportion of cell clusters in each subject. IMN, idiopathic membranous nephropathy; Pod, podocytes; Mes, mesangial cells; Endo, endothelial cells; PEC, parietal epithelial cells; PT, proximal tubule cells; LOH, loop of Henle cells; PC, principal cells; IC, intercalated cells; NKT, natural killer cells and T cells; BC, B cells; Myeloid, myeloid cells.

Figure 2

Myeloid cells have five subgroups. (A) Heatmap of feature genes that clustered each subgroup. (B) Dotplot demonstrates the relative gene expression and percentage of highly expressed genes in each cell group. The color of the dots represents the relative expression levels of genes, while the size of the dots represents the proportion of cells expressing the gene. (C) Barplot denoting the percentages of five myeloid cell subgroups in each sample. (D) Dotplot showing the GO and KEGG pathway enrichment analysis of five myeloid cell subgroups. (E) Violin plots of gene signatures related to MHC-II, inflammation and phagocytosis in different subgroups. (F) Pseudotime cell trajectory analysis on monocytes/macrophages. We correlated the cell trajectory analysis with the three clusters of steady-state kidney monocytes (MC0), inflammatory macrophages (MC1) and alternatively activated (M2-like) macrophages (MC2). Green and red arrowheads indicate clusters mainly composed of controls and IMN patients, respectively.

Figure 3

NK/T cells have ten subgroups. (A) Heatmap of feature genes that clustered each subgroup. (B) Dotplot demonstrates the relative gene expression and percentage of highly expressed genes in each cell group. The color of the dots represents the relative expression levels of genes, while the size of the dots represents the proportion of cells expressing the gene. (C) Barplot denoting the percentages of ten NK/T cell subgroups in each sample. (D) Dotplot showing the GO and KEGG pathway enrichment analysis of ten NK/T cell subgroups. (E) Violin plots of gene signatures related to cytotoxicity and exhaustion in different subgroups. (F) Pseudotime cell trajectory analysis on T cells. We correlated the cell trajectory analysis with the three CD4+ clusters of naive CD4+ T cells (TC0), central memory CD4+ T cells (TC1), and effector memory CD4+ T cells (TC2) (upper), and three CD8+ clusters of resident memory CD8+ T cells (TC4), GZMK+ CD8+ T cells (TC5), and cytotoxic T lymphocyte (TC6) (lower).

Figure 4

B cells have three subgroups. (A) Heatmap of feature genes that clustered each subgroup. (B) Dotplot demonstrates the relative gene expression and percentage of highly expressed genes in each cell group. The color of the dots represents the relative expression levels of genes, while the size of the dots represents the proportion of cells expressing the gene. (C) Barplot denoting the percentages of three B cell subgroups in each sample. (D) Dotplot showing the GO and KEGG pathway enrichment analysis of three B cell subgroups. (E) Pseudotime cell trajectory analysis on B cells. We correlated the cell trajectory analysis with the two clusters of naive (BC0) and memory B cells (BC1). Green and red arrowheads indicate clusters mainly composed of controls and IMN patients, respectively.

Figure 5

Cell-cell interactions of ligands and receptors between different kidney cell types. Heatmap showing ligand-receptor interactions between kidney cell clusters in (A) IMN patients and (B) control subjects. (C) Dot plot showing interactions between kidney parenchymal cell clusters and immune cell clusters. BC, B cells; Endo, endothelial cells; IC, intercalated cells; LOH, loop of Henle; Mes, mesangial cells; Myeloid, myeloid cells; NKT, natural killer cells and T cells; PC, principal cells; PEC: parietal epithelial cells; Pod, podocytes; PT, proximal tubule cells.