Single-Cell Sequencing Informs That Mesenchymal Stem Cell Alleviates Renal Injury Through Regulating Kidney Regional Immunity in Lupus Nephritis

Abstract

Lupus nephritis (LN) is the common complication of systemic lupus erythematosus. The pathogenesis of LN kidney injury is unclear. In addition to systemic (extrarenal) immune cells, local (intrarenal) immune cells residing in “kidney regional immunity” are momentous in LN. Mesenchymal stem cell (MSC) therapy is effective for LN. However, mechanisms of MSC therapy remains unclear. In this study, we first systematically investigated the effects of MSC on immune cells in kidney regional immunity in LN using single-cell sequencing. We found that MSC reduced proinflammatory central memory CD4⁺ T cells, cytotoxic tissue-resident memory CD8⁺ T cells and exhausted CD8⁺ T cells, increased anti-inflammatory Naive/Effector CD8⁺ T cells and type 1 regulatory T cells; reduced infiltrating proinflammatory Ly6c hi/inter/lo era2⁺ macrophages, increased anti-inflammatory resident macrophage and Ly6c lo ear2⁻ macrophage; and reduced long-lived plasma cells and proinflammatory neutrophils and dendritic cells. This study laid a foundation for clinical applications of MSC.

Introduction

Systemic lupus erythematosus (SLE) is a chronic systemic autoimmune disease affecting multiple organs [1]. Its pathogenesis is not well understood. A variety of factors lead to the dysfunction of immune regulation [2]. A large number of autoantibodies [such as anti-dsDNA antibodies, antinuclear antibodies (ANA)] are generated, forming the deposition of immune complex (IC) in the kidneys, and leading to immune inflammation [3,4]. Lupus nephritis (LN) is the most common and serious complication [5].

The pathogenesis of LN kidney injury is unclear. The activation of systemic (extrarenal) immune cells can cause renal local immune response involved in the occurrence of LN kidney injury. In addition, the activation of local (intrarenal) immune cells also plays a more important role in LN [6,7]. Kidney has unique anatomical structure and microenvironment. During kidney injury, infiltrated immune cells from peripheral blood and local resident immune cells in kidneys together form the “kidney regional immunity” [8]. The latter may have a unique immune regulatory function in the kidneys, maintaining regional immune homeostasis and forming local defenses in the kidneys [9]. Several studies found that inflammation of the kidneys during the course of LN may be associated with regional immune disorders [10,11].

It has been found that there is a significant increase of dendritic cells (DC) in the kidney tissues in LN, which function as antigen presentation to amplify inflammation [12]. In LN, Ly6c⁻ tissue-resident macrophages and Ly6c⁺ monocyte-derived macrophages take on a mixed anti-inflammatory, proinflammatory, and wound healing profile [11]. Tissue-resident memory T cells (Trm), effector memory T cells (Tem), and central memory T cells (Tcm) are also found in renal tissues in LN [13 –15]. A significant increase in the number of CD8⁺CD103⁺Trm cells in the kidney tissues in LN leads to inflammation and injury of the kidneys [16]. Tem and Tcm are also involved in renal regional immunity in LN, but their role is unclear [17]. Treg cells in LN patients are lower than those in healthy people, and increasing the proportion of Treg cells in renal immune cells is also a strategy for LN treatment [18].

In addition, long-lived plasma cells (LLPCs) exist in the kidney tissues of LN, which can produce autoantibodies in the local kidney tissues, form IC in the local kidneys, activate neutrophil and DC, and lead to kidney inflammation [19]. Therefore, effective correction of regional immune dysregulation in LN may be a crucial target for LN.

At present, immunosuppressive agents and hormones are commonly used in the treatment of LN [20]. But, prolonged use of these drugs causes the deficiencies of immune system leading to adverse reactions and even death [21]. Mesenchymal stem cells (MSCs) are nonhematopoietic, multipotent stromal precursor cells resident in most adult tissues. They have reparative potential through their self-renewal and differentiation capabilities and are increasingly recognized as being able to have immunomodulatory activity on a variety of immune cells [22]. MSCs can modulate immune responses during chronic inflammation by regulating the recruitment, function, and fate of cells in the innate and adaptive immune system. These properties make MSC an attractive and promising cell therapy for chronic inflammatory diseases [23].

Animal and clinical studies have shown that MSC therapy is safe and effective for SLE [24,25]. In lupus, MSCs reduced disease severity in the lpr-mutant model of SLE by inhibiting the proliferation of T cell populations and inducing the expansion of Treg cells. MSCs may also ameliorate LN kidney injury by reducing macrophage infiltration and polarizing macrophages into an anti-inflammatory phenotype [26]. Currently, a multicenter clinical study has found that MSC therapy is safe and beneficial for patients with LN who are refractory to current therapies [27].

However, most studies on the treatment of LN by MSCs only focus on observing the changes of a specific immune cell subset, such as Treg, cytotoxic T lymphocyte (CTL), B cells, and macrophages, and there is no systematic and comprehensive study on the effects and mechanism of MSCs on the immune response in the renal region. Therefore, to explore how MSC regulates the infiltrating and resident immune cells in LN renal region immunity is a key issue worth investigating. In this study, we systematically investigated the regulatory effects of MSC therapy on various immune cells in the kidneys of LN using single-cell RNA sequencing. We found some novel immune cell subsets in renal regional immunity. This study provides an initial insight into the local immune mechanisms of MSC in LN, laying the theoretical foundation for future clinical applications of MSC in LN.

Materials and Methods

Animals

The animal experiments in this study were conducted in accordance with the experimental protocol approved by the Medical Ethics Committee of the Chinese PLA General Hospital (Ethic Approval No. SQ2021209). Thirty female MRL/lpr mice (as LN model) and 10 female BALB/C mice (4 weeks of age) were used. These animals were obtained from the Shanghai SLAC Laboratory Animal Company Limited (Shanghai, China). The animals were housed in the Medical Experimental Animal Center of the Chinese PLA General Hospital in a specific pathogen-free environment with a 12-h light–12-h dark cycle, and the temperature was maintained at 24°C–26°C.

Transplantation protocol of MSCs

All umbilical cord MSCs used in this experiment were obtained from Qingdao Hualing Aoyuan Bioengineering Technology Group Co., Ltd. MSC isolation and culture were performed in accordance with the standardized operating procedures of the Stem Cell Research Center. A fourth-generation MSC was used for the treatment. The 16-week-old MRL/lpr mice showed high levels of urinary albumin/creatinine ratio (ACR), indicating the occurrence of LN. Then, the mice were randomized into three groups: (1) the control group, age-matched BALB/C mice (n = 10); (2) the LN model group, MRL/lpr mice were injected with 1 mL normal saline (NS) through the tail vein (n = 14); and (3) the MSC group, MRL/lpr mice were injected with MSCs (1 × 10⁶ cells, 1 mL) through the tail vein (n = 14). Mice in each group were injected with NS or MSCs at 16, 18, 22, and 24 weeks of age, respectively. After treatment, random urine samples were taken every 2 weeks. At 26 weeks of age, 24-h urine samples were taken and mice were anesthetized and sacrificed.

Evaluation of kidney coefficient and spleen coefficient

Body weight of the mice was acquired before euthanasia. The organs were weighed. Organ coefficients were calculated as follows: kidney coefficient = 100 × kidney weight/body weight (g/100 g); spleen coefficient = 100 × spleen weight/body weight (g/100 g) [28].

Analysis of serum and urine samples

The urinary protein concentration was determined by the Albumin Assay Kit (Abcam, USA). The creatinine concentration was measured by the Cayman Urinary Creatinine Kit (Cayman, USA). Urinary ACR was calculated to evaluate mouse urine protein excretion. Concentrations of serum creatinine and urea nitrogen were measured using an automatic biochemical analyzer. Concentrations of serum anti-dsDNA and serum ANA were measured using the ELISA Kit (ADI, USA).

Kidney pathological analysis

Kidney tissues from mice fixed in formaldehyde solution for 2 days was dehydrated, embedded, and sectioned (2 μm). Hematoxylin–Eosin (HE) and periodate Schiff (PAS) staining were used to observe the pathological changes. Images were collected using a light microscope (BX53F; OLYMPUS). Renal histopathology was assessed by an experienced pathologist.

Kidney immunofluorescent staining

After the paraffin sections were dewaxed, 2 μm sections were placed in water. Antigen recovery was carried out using citrate solution for 10 min at high temperature. The sections were blocked in sheep serum blocking solution (ZSGB-Bio, ZLI-9022) for 1 h and then incubated with PBS diluted primary antibody for 24 h (4°C). The following antibodies were used: IgM (1:50), IgG (1:50), C3(1:50), CD49b (1:100), Lag-3(1:100), and Foxp3(1:100). The sections were then incubated with secondary antibody and DAPI, respectively, and then mounted on slides. Images were acquired using a digital confocal microscope (FV1000-D; OLYMPUS).

Flow cytometry analysis

We prepared and stained kidney single-cell suspensions as described previously in the literature [29]. The antibody was added into 100 μL single-cell suspension tubes of kidney and incubated for 30 min at room temperature in the dark. The erythrocytes were lysed with lysis buffer for 5 min. The following antibodies were purchased from BioLegend: CD45, CD4, CD49B.LAG-3, CD226, CD11B, F4/80, and Ly6c; Dead kidney cells were excluded by staining with the Zombie Aqua™ Fixable Viability Kit (BioLegend). Data were analyzed by FlowJo V. 10.0.7 software.

Antibody array analysis

The antibody microarrays steps were conducted by RayBiotech (Guangzhou, China). The kidneys homogenates (three kidneys from LN and three kidneys from MSC) were analyzed with a glass slide-based antibody cytokine array, including 200 proteins (GSM-CAA-4000-1; RayBiotech). Laser fluorescence scanning (GenePix 4000B scanner) was used to visualize the signals. Signal intensity data after background subtraction and normalization were exported directly into the RayBio Analysis Tool software.

10 × single-cell RNA sequencing

CD45⁺ immune cells were enriched by positive selection from kidneys for single-cell analysis. Before magnetic enrichment, kidneys were minced in Roswell Park Memorial Institute 1640 containing 5% fetal bovine serum, 0.25% Trypsin (Thermo Fisher), and 10 μg/mL DNase I (Sigma), and then incubated at 37°C with gentle rocking for 10 min. CD45⁺ immune cells were magnetically labeled and enriched by positive selection (Miltenyi Biotec). Cells were diluted with Trypan Blue and counted using a hemocytometer. Kidneys from three different mice were pooled per sample, and two samples were prepared according to the MSC transplantation condition. For a target of 5,000 single cells, 8–9,000 live cells were loaded onto the Chromium Controller (10 × genomics).

Single-cell mapping and clustering

The unique molecular identifiers counts for each gene in each cell were imported into R software (version 4.1.2) by Seurat package (version 4.0.5). Quality control was assessed priorly, by filtering out cells with the following criteria: <200 or >6,000 unique genes expressed, or >6% of reads mapping to mitochondria, resulting in a final dataset of 25,945 cells (12,051 from the LN group and 13,894 from the MSC group). Sequencing data from each group was further normalized and scaled, and detected for highly variable features by using Normalized function, ScaleData, and FindVariableFeartures function, respectively, in Seurat package. Thereafter, principal component analysis (PCA) and linear dimensional reduction based on the selected top 3,000 high variable genes on the scaled data were performed. Subsequently, cells were clustered into 20 groups for leukocytes different types of identification using FindNeighbors and FindClusters function in Seurat package.

Differential expression analysis

Marker genes in each cluster were determined using Wilcoxon Rank-Sum test through the FindAllMarkers function in Seurat, with a false discovery rate adjusted P-value cutoff of 0.05. Differential expressed genes (DEGs) in individual cell type between different experimental conditions and between different cell clusters were recognized under limitation of an estimated log fold change >0.25 and detectable expression in >10% of cells in any of the compared cell cluster by FindMarkers function in Seurat. Relevant cell marker gene expression was visualized on Violin Plot to illustrate specific cell genotypes.

Functional enrichment analysis

Gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG) enrichment analysis was performed on Metascape (metascape.org), and gene-set enrichment score was calculated with gene set enrichment analysis (GSEA) software from GSEA Base (version 1.56.0). Marker genes or DEGs with average log2fold change (avg_log2FC) >0.59, adj_P-value <0.05 were used for GO. Gene-sets were downloaded from the Mouse Genome Informatics.

Cell development trajectories construction

Single-cell pseudotime trajectories of monocyte/macrophage populations were reconstructed by Monocle 2 (version 2.22.0) package. Genes used for pseudotime ordering were taken from DEGs identified by function differential GeneTest with full Model Formula Stri set. Dimension reduction and cell ordering along the trajectories were completed with DDRTree method.

Cell–Cell interaction/communication analysis

We conducted cell–cell interaction analysis utilizing cellphonedb function curated by the CellPhoneDB database. The significant cell–cell interactions were selected with P < 0.01.

Statistics

Data are presented as mean ± standard deviation. The Kolmogorov–Smirnov test normality test was conducted for all data in the article using GraphPad Prism 8.0.1 software. The data from analysis of serum and urine samples, kidney and spleen coefficient, kidney pathological, and kidney immunofluorescent staining were compared using one-way analysis of variance followed by post hoc Tukey's multiple comparison tests. Other data from flow cytometry and immunofluorescence costaining were compared using the Mann–Whitney U test. Statistical significance is expressed as follows: No significance; *P < 0.05; **P < 0.01.

Results

MSC transplantation increased survival rate and improved renal function in LN mice

LN mice were injected intravenously with MSCs or NS at 16 weeks of age, and were repeatedly injected after 2, 6, and 8 weeks (Fig. 1A). At 10 weeks after treatment, mortality was 50% in the LN group and 21.4% in the MSC group (Fig. 1B). At 6 weeks after treatment, urinary ACR began to decrease in the MSC group compared with the LN group (Fig. 1C). Compared with the LN group, the ratio of kidney weight/body weight (Kidney Coefficient) and spleen weight/body weight (Spleen Coefficient) in the MSC group were lower (Fig. 1D, E). Compared with the LN group, MSCs also reduced the level of 24-h urine protein (Fig. 1F-a), urinary ACR (Fig. 1F-b), serum creatinine (Fig. 1F-c), and blood urea nitrogen (BUN; Fig. 1F-d).

FIG. 1.

MSC transplantation increased the survival rate and improved renal function in LN mice. (A) schematic diagram of MSC transplantation procedures. Animals were divided into three groups: Control group (Con): the age-matched Blab/c mice (n = 10) were infused with normal saline; LN model group (LN): MRL/lpr mice (n = 14) were infused with normal saline; MSC treatment group (MSC): MRL/lpr mice (n = 14) were infused with MSC. (B) Survival rate in the above three groups at 0–10 weeks after MSC transplantation. (C) Urinary ACR in the above three groups at 0–10 weeks after MSC transplantation. Data are expressed as mean ± SE. *P < 0.05, **P < 0.01. Here asterisk represents significance between LN and MSC transplantation. (D) The size changes of kidney (a) and kidney weight/body weight (Kidney Coefficient) (b) in the above three groups. (E) The size changes of spleen (a) and spleen weight/body weight (Spleen Coefficient) (b) in the above three groups. (F) 24-h urinary albumin (a), ACR (b), serum creatinine (c) and blood urea nitrogen (d) in the above three groups at 10 weeks after MSC transplantation. Data from Con group (n = 6), LN group (n = 6), and MSC group (n = 11) are expressed as mean ± SD. Data obtained from each group were compared using a one-way ANOVA followed by post hoc Tukey's multiple comparison test. *P < 0.05, **P < 0.01. ACR, albumin/creatinine ratio; ANOVA, analysis of variance; LN, lupus nephritis; MSC, mesenchymal stem cell; SD, standard deviation; SE, standard error.

MSC transplantation improved renal pathological injury in LN mice

After treatment of 10 weeks, compared with the LN group, HE and PAS staining showed a significant reduction in mesangial cell proliferation, tubular lesion, and inflammatory cell infiltration in the MSC group (Fig. 2A, B. Supplementary Fig. S1A). Compared with the LN group, the electron-dense deposits under glomerular epithelial cells or endothelial cells, and podocyte effacement were significantly less in the MSC group than the LN group (Fig. 2C, Supplementary Fig. S1B). Compared with the LN group, MSCs reduced deposition of IgM, IgG, and C3 (Fig. 2D, Supplementary Fig. S1C). Compared with the LN group, MSCs also reduced the level of anti-dsDNA and ANA in serum (Fig. 2E, F).

FIG. 2.

MSC transplantation improved renal pathological injury, reduced deposition of IgM, IgG, and C3 in renal tissues, and decreased concentrations of serum anti-ds-DNA and ANA in LN mice. (A) Representative images of glomerular PAS staining in three groups. 200 × Bar = 50 μm, 400 × Bar = 25 μm. (B) Representative images of glomerular HE staining in the above three groups. 200 × Bar = 50 μm, 400 × Bar = 25 μm. (C) Representative images of electron-dense deposits, foot-process fusion and mesangial cell proliferation in the above three groups by electron microscopy. 5,000 × Bar = 5 μm, 8,000 × Bar = 2 μm. (D) Representative images of deposition of IgM, IgG, and C3 in the kidneys in the above three groups. (E) Serum anti-dsDNA concentration in the above three groups. Bar = 25 μm. (F) Serum ANA concentration in the above three groups. Data from Con group (n = 6), LN group (n = 6), and MSC group (n = 11) are expressed as mean ± SD. Data obtained from each group were compared using a one-way ANOVA followed by post hoc Tukey's multiple comparison test. *P < 0.05, **P < 0.01. ANA, antinuclear antibody; PAS, periodate Schiff.

MSC downregulated the expression of renal inflammation-related proteins in LN mice

RayBiotech antibody microarrays were used to identify renal inflammation-related proteins involved in the renal inflammatory response. PCA analysis showed that the renal inflammatory-related proteins in the MSC group were different from the LN group (Fig. 3A). Differentially expressed proteins between LN and MSC group were defined P < 0.05 and foldchange >1.2 or <0.83(absolute logFC <0.263). Compared with the LN group, MSCs mainly downregulated inflammation-related proteins in the kidneys (16 downregulated and 3 upregulated) (Fig. 3B, C; Supplementary Table S1). GO-BP showed that the differences between the two groups were mainly enriched in immune-related pathways (Fig. 3D). Results suggested that MSCs may alleviate the progression of LN by modulating the renal immune response process.

FIG. 3.

MSCs downregulated the expression of renal inflammatory proteins and the immune activation pathway of immune cells in LN mice. The kidney cytokines were detected by RayBiotech antibody array. The analysis of renal immune cells by Single-Cell RNA Sequencing. (A) PCA results of RayBiotech antibody array. (B) Volcano Plot of differential cytokines between LN model mice (n = 3) and MSC treatment mice (n = 3). Box with down: the downregulated proteins after MSC transplantation, Box with up: the upregulated proteins after MSC transplantation. (C) Clustering heatmap of differential cytokines and corresponding expressions between two groups. (D) Biological process from GO analysis of differentially expressed cytokines. (E) The experimental procedure of 10 × Single-Cell RNA Sequencing. Kidneys were digested into single-cell suspension and live CD45⁺ cells were magnetic beads-sorted and loaded cells for scRNA-seq. (F) Differential genes in renal CD45⁺ immune cells between the two groups. (G) The GSEA hallmark pathways of GO and KEGG terms enriched the two groups. PCA, principal component analysis; GO, gene ontology; GSEA, gene set enrichment analysis; KEGG, kyoto encyclopedia of genes and genomes.

MSC significantly downregulated expression of immune-related genes in renal immune cells in LN mice

Single-cell RNA sequencing for CD45⁺ renal immune cells (Fig. 3E) revealed that MSC affected the activity of various immune-related pathways by altering the expression levels of immune-related genes in renal immune cells. Compared with the LN group, the expression of genes in the immune cells in the MSC group was downregulated, and only a few genes were upregulated (e.g., Mndal, Ifi203, STAT1) (Fig. 3F). GSEA analysis was performed to determine whether immune-related pathways were up- or downregulated between the two groups. GSEA analysis of GO showed that, compared with the LN group, MSCs downregulated the pathways of immune cells activation/proliferation/chemotaxis, while upregulated the pathways of regulatory T cells (Treg cells) and interferon-β (Fig. 3G). GSEA analysis of KEGG showed that immune disease-related pathways and receptor-mediated signaling pathway were downregulated in the MSC group (Fig. 3G).

MSC significantly changed the immune cell landscape in renal regional immunity in LN mice

Using scRNA-Seq, we identified 25945 CD45⁺ immune cells (12,051 cells from the LN group and 13,894 cells from the MSC group). In the merged samples of LN and MSC, based on the expression of typical lineage markers and other genes specifically upregulated in each cluster, we identified them as CD4⁺T cells, CD8⁺T cells, double-positive T cells (DPT), double-negative T cells (DNT), DC, macrophages, B cells, neutrophils, natural killer (NK) cells, and unknown cells (Fig. 4A). CD4⁺T cells expressed CD3E and CD4; CD8⁺T cells expressed CD3E and CD8A; DPT cells expressed CD3E, CD4, and CD8A; DNT cells only expressed CD3E with lack of CD4 and CD8A; DC expressed CD68 and CD11C (Itgax); macrophage expressed CD68; B cell expressed CD79a; neutrophil expressed S100a9; NK lacked CD3E expression (Fig. 4B).

FIG. 4.

Single-cell RNA sequencing of renal CD45⁺ immune cells identified 10 distinct immune cell subpopulations in LN and MSC-treated mice. (A) t-SNE plot of cell clusters in the merged samples. Each point depicts a single cell, colored according to cluster designation. (B) The expression of CD3e, CD4, CD8a, CD68, Itgax (CD11C), CD79a, and S100a9 in different clusters in the merged samples of LN and MSC. (C) The proportion of CD4T, CD8T, DPT, DNT, DC, macrophages, B cells, neutrophils, NK cells in the merged samples of LN and MSC. (D) t-SNE plot of CD45⁺ immune cell populations in different groups. (E) The proportion differences of immune cell populations between the LN mice and MSC transplantation mice. DC, dendritic cells; DNT, double-negative T cells; DPT, double-positive T cells; NK, natural killer; t-SNE, t-distributed stochastic neighbor embedding.

*Correction added on May 22, 2023 after first online publication of May 17, 2023: The panel D of Figure 4 has been corrected.

In the merged samples of LN and MSC, the proportion of T lymphocytes in immune cells was relatively high, followed by neutrophils, DC/macrophages, and the proportion of B cells and NK cells was relatively low (Fig. 4C). There were differences in renal immune cell landscape between LN group and MSC group (Fig. 4D). MSC treatment reduced neutrophils, DC/macrophages, and B cells, while increased NK cells (Fig. 4E; Supplementary Table S2).

MSCs reduced Tcm in CD4⁺T cells in renal regional immunity in LN mice

Once naive T lymphocytes are activated, they proliferate rapidly, producing effector cells that migrate to the site of inflammation [30]. According to the expression level of Ccr7, Effector T cells can be divided into Tcm and Tem [31]. Tem cells respond quickly to the stimulus and produces various effector molecules. Tcm cells have strong proliferation ability and can support and replenish Tem [32]. A recent single-cell sequencing study also found the presence of Tem and Tem in kidney biopsies from LN patients [17].

We identified five clusters of CD4⁺T cells in the merged kidney tissues: Effector memory CD4⁺T cells (Ccl5⁺), Central memory CD4⁺ T cells (Ccr7⁺), Foxp3 CD4⁺T cells (Foxp3⁺), Tr1 CD4⁺T cells (CD226⁺), and ISG-high CD4⁺T cells (ISG15⁺/ISG20⁺) (Fig. 5A–C). Central memory CD⁺4 T cells expressed Ccr7, while Effector memory CD4⁺T cells did not. We found that Central memory CD4⁺T cells in CD4⁺T cells were decreased in the MSC group, while Effector memory CD4⁺T cells did not change significantly (Fig. 5D; Supplementary Table S3). The pathways of activation of immune response, cytokine production in Tem and Tcm were more significantly enriched than Treg, Tr1, and ISG-high CD4⁺T (Fig. 5E). In addition, we also found that Tem was also associated with the pathways of negative regulation of immune activation. Therefore, compared with Tcm, Tem may have both proinflammatory and anti-inflammatory effects.

FIG. 5.

Single-cell RNA sequencing revealed the differences of CD4⁺T lymphocyte in the kidney tissues in LN and MSC-treated mice. (A) t-SNE plot of CD4⁺ T cell clusters in the kidneys. (B) Dot plot displaying the representative maker genes of the five CD4⁺T clusters in the merged samples of LN and MSC. (C) The expression of Ccl5, Ccr7, Fopx3, CD226, ISG20, and ISG15 in the five CD4⁺T clusters in the merged samples of LN and MSC. (D) The proportional differences of Effector memory CD4⁺T cells (Tem), Central Memory CD4⁺T (Tcm), Foxp3 CD4⁺T, Tr1 CD4⁺T, and ISG-high CD4⁺T in the kidneys in LN mice and MSC-treated LN mice. (E) The heatmap of GO terms enriched pathways in the five CD4⁺T clusters in the merged samples of LN and MSC.

MSCs increased Tr1 Treg in CD4⁺T cells, not Foxp3 Treg, in renal regional immunity in LN mice

Treg cells can be recruited to the kidneys to play an immunosuppressive role during inflammation [18]. Treg cells may reside in the tissues for a long time and do not enter the circulation [31]. Treg cells can be divided into naturally generated natural Treg cells (nTregs) and induced adaptive Treg cells (aTregs or iTregs), such as Tr1 (type 1 Treg cell) and Th3 [33]. nTregs express the transcription factor Foxp3. Tr1 cells do not express Foxp3. Tissue-resident Treg cells may originate in the thymus (nTREG) or may be induced in peripheral tissues (iTREG). Both typically Treg cells and aTreg cells can inhibit immune inflammation [34].

We have identified two types of Treg cells in renal tissue by single-cell sequencing: Foxp3⁺ Treg and Tr1. Tr1 cells are marked by CD226, CD49b, and Lag3 [35,36]. Our results showed that Tr1 did not express Foxp3 but expressed CD226 (Fig. 5C). Tr1 and Foxp3 Treg cells showed similar gene expression (e.g., Tnfrsf4, CTLA-4, IL2ra, Tnfrsf18, Pdcd1, and Ikzf2) (Supplementary Fig. S2A), and these genes are also commonly expressed in Treg cells [37]. We found that the proportion of Tr1 cells in CD4⁺T increased after MSC transplantation, but there was no change in Foxp3 Treg (Fig. 5D). Functional enrichment of CD4⁺T cell subsets also showed that Foxp3 Treg and Tr1 may have similar functional characteristics. Foxp3 Treg and Tr1 CD4⁺T cells were associated with the pathways of negative immune cell activation and cytokine production. (Fig. 5E).

Therefore, we speculated that MSCs may exert its therapeutic effect on LN by Tr1 rather than Foxp3 Treg. To test this hypothesis, we analyzed the proportional changes in Tr1 and Foxp3 Tregs in renal tissues after MSC transplantation by flow cytometry and immunofluorescence. Renal tissue flow cytometry showed that the mean proportion of CD49b⁺ Lag3⁺CD4⁺ Tr1 cells in CD4⁺T in the LN group was 3.35%, while the mean proportion in the MSC group was 24.6%. Compared with the LN group, the proportion of Tr1 was significantly higher (P < 0.01), but there was no difference in the proportion of Foxp3 Treg between the two groups (P > 0.05; Fig. 6A–C). Kidney immunofluorescence showed similar results. Compared with the LN group, the counts of CD49b⁺ Lag3⁺ Tr1 cells was higher (P < 0.01; Fig. 6D, E). Thus, MSC seemed to have a greater regulatory effect on Tr1 cells in the kidney than Foxp3 Treg.

FIG. 6.

Flow cytometry and immunofluorescence revealed proportional differences of Foxp3 Treg, Tr1 in kidneys between the LN mice and the MSC-treated mice. (A) The gating strategies of Foxp3 Treg and Tr1. (B) The representative flow cytometric dot plots of Foxp3 Treg and Tr1 between two groups. (C) The flow cytometric graphical analysis of Foxp3 Treg and Tr1 between two groups. Data from LN group (n = 5) and MSC group (n = 5) are expressed as mean ± SD. (D) The representative immunofluorescent images of Tr1 in the kidneys. (E) The counts of Tr1 in the kidneys by immunofluorescence costaining. Data from LN group (n = 6) and MSC group (n = 11) are expressed as mean ± SD. Data obtained from each group were compared using the Mann–Whitney U test. *P < 0.05, **P < 0.01.

In addition, we also found that ISG-high CD4⁺T cells were increased and related to the pathways of positive regulation of interferon-beta production and interleukin-2 production (Fig. 5E), suggesting that this subset of cells may also have a renal protective effect.

MSCs reduced resident memory cytotoxic CD8⁺T cells (Trm CD8⁺T) in CD8⁺T cells in renal regional immunity in LN mice

In LN, CD8⁺T cells are heterogeneous after infiltrating into tissues [38]. In LN single-cell sequencing studies, CD8⁺T cells present different phenotypes in the kidneys: exhaustion T (Tex), Tem, Trm, CTL, etc. [17]. Trm is a subset of memory T cells residing in organs with high expression of CD69, CD49a, and CD103, and does not participate in systemic circulation [32,39]. Studies have shown that CD8⁺Trm cells are amplified in the kidneys of LN patients or MRL/lpr mice [40].

We identified three subsets of CD8⁺ T cells: cytotoxic CD8⁺ T cells (Gzmb⁺), exhausted CD8⁺ T cells (Pdcd1⁺), and Naive/Effector CD8⁺ T cells (Sell⁺/IL7r⁺) (Fig. 7A–C). We found that cytotoxic CD8⁺ T not only expressed CTL marker genes (Gzmb), but also specifically expressed tissue-resident memory markers, CD103(Itgae) and CD49a(Itga1) (Fig. 7C). In other words, Trm cells and cytotoxic CD8⁺T cells had the overlapped gene expression in LN, and both may be the same kind of CD8⁺T cells, resident memory cytotoxic CD⁺8 T cells (Trm CD8⁺T). MSCs reduced Trm CD8⁺T and exhausted CD8⁺T in CD8⁺T, while increased Naive/Effector CD8⁺T cells (Fig. 7D; Supplementary Table S4). Functional enrichment analysis for these CD8⁺T cell subsets (Fig. 7E) showed that cytotoxic CD8⁺ T cells and exhausted CD8⁺ T cells were associated with pathways of lymphocyte activation and cytokine production. Besides, cytotoxic CD8⁺ T cells were more related to the pathways of immune activation than exhausted CD8⁺ T cells.

FIG. 7.

Single-cell RNA sequencing revealed CD8⁺T lymphocyte differences in the LN mice and the MSC-treated mice. (A) t-SNE plot of CD8⁺ T cell clusters in the kidneys. (B) Dot plot displaying the representative maker genes of the three CD8⁺T clusters in the merged samples of LN and MSC. (C) The expression of Sell, IL7r, Pdcd1, Gzmb, CD103, and CD49a in the three CD8⁺T clusters in the merged samples of LN and MSC. (D) The proportion of proportion cytotoxic CD8⁺ T cells, exhausted CD8⁺ T cells, and Naive/Effector CD8⁺ T cell in the kidneys. (E) The heatmap of GO terms enriched pathways in the three CD8⁺T clusters in the merged samples of LN and MSC.

MSCs increased anti-inflammatory resident macrophages, and reduced proinflammatory infiltrating macrophages in renal regional immunity in LN mice

Macrophages can be divided into two categories: Monocyte-derived macrophages and Tissue-resident macrophages [41]. Renal macrophages in LN consist of infiltrating macrophages and resident macrophages [11]. Ly6c is a cell surface glycoprotein that is used to identify the type of macrophages. Ly6c⁺ macrophages are considered to be from the peripheral blood and have proinflammatory function, while Ly6c⁻ macrophages, also known as tissue-resident macrophages, have anti-inflammatory function [42].

First, macrophages were defined as two groups: Ly6c⁺ macrophages and Ly6c⁻ macrophages, based on the expression level of Ly6c. Ly6c⁺ macrophages were further divided into Ly6c hi macrophage, Ly6c inter macrophage, and Ly6c lo macrophage. All three groups of macrophages could be regarded as peripheral infiltrating macrophages. While Ly6c− macrophage could be considered as a resident macrophage. Then Ly6c lo macrophages could also be divided into two different groups of cells, according to ear2 gene expression. So, macrophages in our single-cell sequencing data were divided into five distinct subgroups: Ly6c hi macrophage (Ly6c hi Mφ), Ly6c inter macrophage (Ly6c inter Mφ), Ly6c lo ear2⁺ macrophage (Ly6c lo ear2⁺ Mφ), Ly6c lo ear2⁻ macrophage (Ly6c lo ear2⁻ Mφ), and resident macrophage (resident Mφ) (Fig. 8A–C). Besides, CD14 and CD16 are also common markers used to distinguish peripheral origin mononuclear macrophages from tissue-resident macrophages. CD14⁺CD16⁺ macrophages are commonly referred to as proinflammatory monocyte-derived macrophages [43].

FIG. 8.

Single-cell RNA sequencing revealed macrophage differences in the LN mice and the MSC-treated mice. (A) t-SNE plot of macrophage clusters in the kidneys. (B) Dot plot displaying the representative marker genes of the five macrophage clusters in the merged samples of LN and MSC. (C) The expression of Ly6c, Ear2, and C1qc in the five macrophage clusters in the merged samples of LN and MSC. (D) The proportion of Ly6c hi macrophage, Ly6c inter macrophage, Ly6c lo ear2⁺ macrophage, Ly6c lo ear2⁻ macrophage, Resident macrophage in the kidneys. (E) The heatmap of GO terms enriched pathways in the five macrophage clusters in the merged samples of LN and MSC.

In our data, Ly6c hi, Ly6c inter macrophage, and Ly6c lo ear2⁺ macrophage also highly expressed CD14 and CD16, while Ly6c lo ear2⁻ macrophage and resident macrophage did not (Supplementary Fig. S2B, C).

MSCs prominently decreased the proportion of Ly6c hi, Inter, and lo ear2⁺ macrophages, and increased the proportion of Ly6c⁻ resident macrophage and Ly6c lo ear2⁻ macrophage (Fig. 8D; Supplementary Table S5). Next, we performed functional enrichment analysis for these macrophage subsets (Fig. 8E). The results showed that Ly6c hi, inter, and lo ear2⁺ macrophages were primarily involved in the function of immune cells activation and proinflammation. Resident macrophages were more associated with anti-inflammatory effect. Except for positive regulation for interleukin-2 production pathway, Ly6c lo ear2⁻ macrophages were not as enriched in other immune-related pathways as the other four macrophage clusters.

Then, renal flow cytometry was used to further verify the changes of renal Ly6c⁺ and Ly6c⁻ macrophages after MSC transplantation. The results showed that MSCs decreased infiltrating Ly6c⁺ macrophages and increased Ly6c⁻ tissue-resident macrophages (Fig. 9A–C). From the results of single-cell sequencing and flow cytometry, we could see that MSCs could increase anti-inflammatory renal resident macrophages and decrease proinflammatory infiltrating macrophages.

FIG. 9.

Flow cytometry revealed proportion differences of Ly6c⁺/Ly6c⁻ macrophage in kidneys between the LN and the MSC-treated mice. Single-cell RNA sequencing revealed the differentiation trajectories of macrophage in the LN and the MSC-treated mice. (A) The gating strategies of Ly6c⁺ and Ly6c⁻ macrophage. (B) The representative flow cytometric dot plots of Ly6c⁺ and Ly6c⁻ macrophage between two groups. (C) The flow cytometric graphical analysis of Ly6c⁺ and Ly6c⁻ macrophage between the two groups. Data from LN group (n = 5) and MSC group (n = 5) are expressed as mean ± SD. Data obtained from each group were compared using the Mann–Whitney U test. *P < 0.05, **P < 0.01. (D) Visualization of cell fate in trajectories of macrophage. (E) Pseudotemporal expression pattern of clustering genes heatmap among macrophage clusters. (F) The differentiation trajectories of five macrophage clusters. (G) The differentiation trajectories of macrophage between two groups.

MSCs facilitated the conversion of macrophages into renal resident macrophages and Ly6c lo ear2⁻ macrophages in LN mice

It is generally believed that early fetal yolk sac macrophages are the main source of renal resident macrophages. However, many studies have shown that monocyte-derived macrophages can also be the complement of classical tissue-resident macrophages, which are recruited from the peripheral and then colonized in tissues [44]. To further reveal the dynamic differentiation changes between infiltrating macrophages (Ly6c hi/inter/lo macrophage) and resident macrophages (Ly6c⁻ resident macrophage), we constructed the differentiation trajectories of macrophages based on scRNA-seq data.

Using pseudotemporal analysis, we divided macrophages into three distinct fate stages: fate1, fate2, and fate3 (Fig. 9D, E). Ly6c hi/inter/lo ear2⁺ macrophages were located in the trajectories of fate1; Ly6c⁻ resident macrophages were located in the trajectories of fate2, and Ly6c lo ear2⁻ macrophages were located in the trajectories of fate3 (Fig. 9F). Therefore, we identified proinflammatory infiltrating Ly6c hi/inter/lo ear2⁺ macrophage as fate1 macrophage; anti-inflammatory resident macrophage as fate2 macrophage; Ly6c lo ear2⁻ macrophage with hypoimmunofunction as fate3 macrophage. Based on macrophage differentiation studies and pseudotemporal analysis data, we suggested that peripherally recruited proinflammatory infiltrating macrophages in fate1 phase could differentiate to resident macrophages in fate2 phase and Ly6c lo esr2⁻ macrophage in fate3 phase.

We also found that macrophages in the LN group were distributed at the trajectories of fate1, while macrophages in the MSC group were distributed at the trajectories of fate2 and fate3 (Fig. 9G). In summary, we speculated that MSCs may promote differentiation of proinflammatory mononuclear macrophages into anti-inflammatory resident macrophages and immunocompromised macrophages, thus reducing renal inflammation.

MSCs reduced LLPCs in renal regional immunity in LN mice

Naive B cells can be activated to differentiate into plasma cells (PC). Autoantibodies are generally produced by short- or long-lived plasma cells. PC migrates to the organ and further differentiated into LLPCs [45]. In LN, the kidneys are the primary source of PCs that produce autoantibodies, and the proportion of self-reactive PCs in the kidneys appears to be higher than in other organs (about 50% of PC is in kidneys, 20% in spleen, and 30% in bone marrow) [19]. In our data, we detected a B cell subset characterized by CD79a⁺, CD79b⁺, and Ly6d⁺ (Fig. 10A, B). This subset also expressed the marker genes of PCs, Jchain, Eaf2, and CD138(Sdc1) (Supplementary Fig. S1B). CD138 is often used to distinguish LLPCs from other B cells [46]. We found that MSCs reduced LLPCs in the kidneys (416 cells from LN group and 85 cells from MSC group; Supplementary Table S1). In addition, this subset cells were mainly related to immune activation and cytokine production (Fig. 10C). In addition, GSEA functional enrichment analysis showed that MSCs also downregulated pathways related to immune activation (Fig. 10D).

FIG. 10.

Single-cell RNA sequencing revealed B lymphocyte differences, cell–cell interactions, and renal regional immunity in the LN and the MSC-treated mice. (A) t-SNE plot of B cell cluster in the kidneys. (B) The expression of CD79a, CD79b, Ly6d, Jchain, Eaf2, and CD138(Sdc1) in B cell cluster in the merged samples of LN and MSC. (C) The GO terms enriched pathways in B cell cluster in the merged samples of LN and MSC. (D) The GSEA pathways of GO terms enriched in B cell cluster between the two groups. (D) Heatmaps showing capacity for intercellular communication between the two groups. Color from blue to red indicates the number of ligand–receptor pairs from low to high. (E) An overview of the major immune cell types implicated in kidney regional immunity of lupus nephritis.

MSCs reduced neutrophils and DC in renal regional immunity in LN mice

Neutrophils can induce autoimmunity to LN through a unique form of cell death called NETosis (neutrophils extracellular trap) [7]. In LN, DC infiltrate the kidneys and perform the function of presenting antigens, amplifying inflammation. DC and monocyte-derived macrophages originate from myeloid progenitor cells in the bone marrow [11]. In contrast to macrophages, CD11C expression is high in DC [47]. In our RNA-Seq data, 2,891 neutrophils (2,677 cells from the LN group and 214 cells from the MSC group) were found (Supplementary Fig. S3A, B; Supplementary Table S2); 469 DCs specifically expressing CD11C and Ciita were detected (662 from the LN group and 253 from the MSC group; Supplementary Fig. S3E, F; Supplementary Table S2). The number of proinflammatory neutrophils and DC were decreased after MSC transplantation. MSCs also downregulated the pathways of immune cell activation and cytokine production in neutrophils and DC (Supplementary Fig. S3C, D, G, H).

MSCs increased NK, DNT, and DPT cells in renal regional immunity in LN mice

In our scRNA-seq data, we identified a cluster of cells that do not express the marker of T, B, DC, neutrophils, and macrophages (Supplementary Fig. S4A, B). These cells also did not express NK marker genes, KLRB1 and TYROBP. However, this subset of cells expressed Klrc2 and Ikzf2, which also are common genes in NK cells (Supplementary Fig. S4C) [48,49]. This subset of cells was mainly distributed in the kidneys of MSC group (21 from the LN group and 671 from the MSC group; Supplementary Table S2). According to the GSEA analysis, we inferred that these NK cells were different from the traditional NK cells, mainly manifested as the silencing and loss of immune cell markers and cell functions (Supplementary Fig. S4D).

DNT cells are a unique subset of T lymphocytes, which do not express CD4, CD8, and NK cell markers [50]. DPT cells express both CD4 and CD8 [51]. Two hundred forty-three DNT cells were detected (90 from the LN group and 153 from the MSC group) and 368 DPT cells (140 from the LN group and 228 from the MSC group) were detected (Supplementary Table S2). DNT and DPT cells accounted for a low proportion of renal immune cells (1.5% and 1.0%). To get a better understanding of these two clusters, we need to enrich them.

MSCs reduced immune cell interaction in renal regional immunity in LN mice

The pathogenesis of LN is complex, involving activation and infiltration of various immune cells in renal regional immunity. CellPhoneDB was used to identify ligands/receptors and molecular interactions among the major cell types. Among them, myeloid cells (macrophage/DC/Neutrophil) harbored the highest pair numbers (Fig. 10D). The results indicated that myeloid cells could be the key effectors and amplifiers of the proinflammatory response in the LN. MSCs reduced ligand/receptor pairs between myeloid cells and other cells. Thus, MSC transplantation suppressed immune cell interactions and attenuated the proinflammatory response in the kidneys.

Discussion

LN is the most common and serious complication of SLE. Long-term usage of glucocorticoids and other immunosuppressive agents often induces diverse side effects. MSCs have attracted attention because of their powerful immunoregulatory ability. In this study, we found that MSC therapy is safe and effective for LN. Even though the application of intravascular MSC administration appears to be safe, the majority of the MSCs get trapped in the lungs with only a small fraction reaching their target tissue. The cause of this problem may be the formation of microthrombi in the lungs. It has been shown that a targeted MSC administration through the renal artery immediately reduces renal perfusion, coinciding with the presence of microthrombi in the glomeruli and signs of an instant blood-mediated inflammatory reaction [52]. MSCs can have podoplanin on them that binds and activates platelets causing microthrombi, which is a key issue worth investigating [53].

There is growing evidence that some subsets of nonrecirculating cells, such as macrophages, NK, intrinsic lymphocytes, memory T cells, and Tregs, exist in multiple nonlymphoid tissues and promote regional immune responses according to the characteristics of each regional tissue [31]. These infiltrating and resident immune cells in the kidneys form specific “renal regional immunity”. In this study, we found that CD4⁺T cells (Tem, Tcm, Treg, Tr1, ISG-high CD4⁺T), CD8⁺T cells (Trm, Tex, Naive/Effector CD8⁺T), macrophages (Ly6c hi Mφ, Ly6c inter Mφ, Ly6c lo ear2⁺ Mφ, Ly6c lo ear2⁻Mφ, resident Mφ), DC, B cells, neutrophils, NK cells, DPT, and DNT constituted the kidney regional immunity (Fig. 10E). MSCs not only changed the proportion of infiltrating immune cells in the kidneys (such as Tcm, Tem, Ly6c hi/inter/lo Mφ), but also changed the proportion of resident immune cells (such as Trm, resident Mφ, LLPC).

Treg cells are necessary to control the immune response. There are two major subsets, which show a strong capacity to suppress effector T cells: Foxp3⁺ Treg cells and Tr1 cells. In both types of Treg cells, IL-10 plays a major role in cell maintenance and, thereby, in sustaining homeostasis. They have the potential to suppress Th17 and Th1 cells directly through IL-10 [54]. Tr1 cells have been less studied than classical Tregs and even less in kidney diseases. A 2022 clinical study of LN found that enhancing the expression of functional CD4⁺ CD46⁺ Tr1 cells may be an effective treatment for LN [55]. Besides, a SLE study found that the proportion of Foxp3⁺Treg did not change significantly after MSC treatment, which was consistent with our results [56]. It was found that function of classic Treg was deficient in SLE patients compared with normal people [57]. Therefore, the functional role of Tr1 cells in LN is a key issue worth investigating. The limitation of this study is the lack of in-depth studies to verify the function of Tr1 cells and elucidate how MSC induces Tr1 production. In the following studies, we will focus on further research on Tr1 cells.

Based on the expression level of Ly6c, macrophages are defined as Ly6c⁺ peripheral mononuclear macrophages and Ly6c⁻ resident macrophages. Intrarenal monocyte recruitment and Ly6c hi macrophage expansion can induce renal parenchymal cell apoptosis during LN [11]. Some resident macrophage populations are maintained by monocytes through longevity, proliferation, and recruitment. A recent study in 2022 showed that whereas established resident peritoneal macrophages largely fail to migrate to the omentum, monocyte-derived resident cells readily migrate and form a substantial component of omental CD102⁺ macrophages in the months following resolution of peritoneal inflammation [58]. The resident macrophages differentiated by peripheral infiltrating macrophages had higher migration capacity. A full appreciation of the complexity and functions of renal macrophages in LN will require further exploration of their distinct molecular and functional characteristics. In our single-cell data, we also found some complexity in identification and function of macrophage subsets.

What is more, recent renal single-cell transcriptional sequencing studies have shown that ear2 is expressed in Ly6c lo macrophages [59]. However, there is relatively little research on this subset of macrophages. Our single-cell data suggested that Ly6c lo ear2⁺ macrophages and Ly6c lo ear2⁻macrophages had different immune functions. In our next study, genes and pathways of interest can be studied in more detail in LN so as to determine which cellular functions are pathogenic and which are protective. The ultimate goal is to make the MSC more targeted at preventing kidney inflammation by preventing either the influx of pathogenic macrophages or the activation of pathogenic pathways.

From this study, we could see that neutrophils accumulate in LN kidneys. MSCs could reduce the recruitment of neutrophils in the kidney. An increased number of neutrophils is common in patients with lupus. Neutrophils are important contributors to autoimmunity through a unique form of cell death called NETosis (neutrophil extracellular trap) [60]. Endothelial cell injury plays an important role in the inflammatory processes associated with SLE [61]. Meanwhile, a study has shown that there is a crosstalk between endothelial cells and MSC. MSCs can inhibit neutrophil recruitment through endothelial cells [62]. In LN, MSCs may inhibit neutrophil recruitment by regulating endothelial cells, which needs to be further verified in our next study.

So far, only a limited number of articles on the use of scRNA-seq in LN have been published. In this study, we used scRNA-seq to perform a comprehensive classification and description of these immune cell subtypes, including their specific marker genes and functional characteristics. Taken together, the results in this study may provide us with a more comprehensive insight into the immunological characteristics of the renal regional immunity in LN in mice after MSC transplantation, as well as valuable information for future clinical applications of MSC.

Ethics Approval and Consent to Participate

The animal experiments in this study were conducted in accordance with the experimental protocol approved by the Medical Ethics Committee of the Chinese PLA General Hospital (Ethics Approval No. SQ2021209). Procedures for the MSC culture and consent of patients followed the standards approved by the Chinese PLA General Hospital Ethics Committee (No. S2019-059-01).

Availability of Data and Materials

The data used to support the findings of this study are available from the corresponding author upon request.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the National Key Research and Development Program of China (Grant No. 2020YFA0113004) and the National Natural Science Foundation of China (Grant No. 81830060).

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Figure S3

Supplementary Figure S4

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4

Supplementary Table S5

References

Kiriakidou

and Ching

. (2020). Systemic lupus erythematosus. Ann Intern Med, 172:ITC81–ITC96.

Zucchi

, Elefante

, Schilirò

, Signorini

, Trentin

, Bortoluzzi

and Tani

. (2022). One year in review 2022: systemic lupus erythematosus. Clin Exp Rheumatol, 40:4–14.

Wang

and Xia

. (2019). Anti-double stranded DNA antibodies: origin, pathogenicity, and targeted therapies. Front Immunol, 10:1667.

Pisetsky

, Bossuyt

and Meroni

. (2019). ANA as an entry criterion for the classification of SLE. Autoimmun Rev, 18:102400.

Parikh

, Almaani

, Brodsky

and Rovin

. (2020). Update on lupus nephritis: core curriculum 2020. Am J Kidney Dis, 76:265–281.

Lech

and Anders

. (2013). The pathogenesis of lupus nephritis. J Am Soc Nephrol, 24:1357–1366.

Frangou

, Georgakis

and Bertsias

. (2020). Update on the cellular and molecular aspects of lupus nephritis. Clin Immunol, 216:108445.

Tian

, Cao

, Chen

and Lyu

. (2016). Regional immunity in tissue homeostasis and diseases. Sci China Life Sci, 59:1205–1209.

Zhou

, Sun

, Zhao

and Zhang

. (2017). Advancement in regional immunity and its clinical implication. Sci China Life Sci, 60:1178–1190.

10.

Kurts

, Panzer

, Anders

and Rees

. (2013). The immune system and kidney disease: basic concepts and clinical implications. Nat Rev Immunol, 13:738–753.

11.

Maria

and Davidson

. (2017). Renal macrophages and dendritic cells in SLE nephritis. Curr Rheumatol Rep, 19:81.

12.

Sahu

, Bethunaickan

, Singh

and Davidson

. (2014). Structure and function of renal macrophages and dendritic cells from lupus-prone mice. Arthritis Rheumatol, 66:1596–1607.

13.

Dolff

, Abdulahad

, van Dijk

, Limburg

, Kallenberg

and Bijl

. (2010). Urinary T cells in active lupus nephritis show an effector memory phenotype. Ann Rheumat Dis, 69:2034–2041.

14.

Gunawan

, Her

, Liu

, Tan

, Chan

, Tan

WWS

, Dharmaraaja

, Fan

, Ong

, Loh

, Chang

KTE

, Tan

, Chan

JKY

and Chen

. (2017). A novel human systemic lupus erythematosus model in humanised mice. Sci Rep, 7:16642.

15.

, Tang

, Zhang

, Jia

, Wang

, Li

, Han

, Peng

, Xie

, Wu

, Wang

, Zhen

, Wang

, Liu

, Sun

, Zhang

and Yi

. (2022). Targeting tissue-resident memory CD8+ T cells in the kidney is a potential therapeutic strategy to ameliorate podocyte injury and glomerulosclerosis. Mol Ther, 30:2746–2759.

16.

Zhou

, Guo

, Li

, Huang

, Li

, Zhang

, Zhao

, Wang

, Zhang

and Yang

. (2020). JAK/STAT signaling controls the fate of CD8+CD103+ tissue-resident memory T cell in lupus nephritis. J Autoimmun, 109:102424.

17.

Arazi

, Rao

, Berthier

, Davidson

, Liu

, Hoover

, Chicoine

, Eisenhaure

, Jonsson

, Li

, Lieb

, Zhang

, Slowikowski

, Browne

, Noma

, Sutherby

, Steelman

, Smilek

, Tosta

, Apruzzese

, … Accelerating Medicines Partnership in Network

SLE

. (2019). The immune cell landscape in kidneys of patients with lupus nephritis. Nat Immunol, 20:902–914.

18.

, Tang

, Yin

and Dai

. (2022). New insights for regulatory T cell in lupus nephritis. Autoimmun Rev, 21:103134.

19.

Espeli

, Bökers

, Giannico

, Dickinson

, Bardsley

, Fogo

and Smith

. (2011). Local renal autoantibody production in lupus nephritis. J Am Soc Nephrol, 22:296–305.

20.

Anders

, Saxena

, Zhao

, Parodis

, Salmon

and Mohan

. (2020). Lupus nephritis. Nat Rev Dis Primers, 6:7.

21.

and Tsokos

. (2012). Treatment of systemic lupus erythematosus: new advances in targeted therapy. Ann N Y Acad Sci, 1247:138–152.

22.

Castro-Manrreza

and Montesinos

. (2015). Immunoregulation by mesenchymal stem cells: biological aspects and clinical applications. J Immunol Res, 2015:394917.

23.

Munir

and McGettrick

. (2015). Mesenchymal stem cell therapy for autoimmune disease: risks and rewards. Stem Cells Dev, 24:2091–2100.

24.

Chun

, Choi

, Kim

, Nam

, Lee

, Kim

, Han

and Bae

. (2022). Safety and tolerability of bone marrow-derived mesenchymal stem cells in lupus animal models and a phase I clinical trial in humans. Lupus, 31:1245–1253.

25.

, Guo

, Pan

, Chen

, Liu

, and Pan

. (2021). Mesenchymal stem cell therapy: hope for patients with systemic lupus erythematosus. Front Immunol, 12:728190.

26.

Zhang

, Niu

, Tang

, Feng

, Yao

, Chen

, Li

, Feng

, Chen

, and Sun

. (2019). Mesenchymal stem cells prevent podocyte injury in lupus-prone B6.MRL-Faslpr mice via polarizing macrophage into an anti-inflammatory phenotype. Nephrol Dialysis Transpl, 34:597–605.

27.

Wang

, Li

, Zhang

, Chen

, Li

, Hu

, Jiang

, Shi

, and Sun

. (2014). Umbilical cord mesenchymal stem cell transplantation in active and refractory systemic lupus erythematosus: a multicenter clinical study. Arthritis Res Ther, 16:R79.

28.

Liu

, Lu

, Lou

, Cai

, Cui

, Wang

, Nie

, Chen

, Li

and Luo

. (2019). Xenogeneic transplantation of human placenta-derived mesenchymal stem cells alleviates renal injury and reduces inflammation in a mouse model of lupus nephritis. BioMed Res Int, 2019:9370919.

29.

Deng

, Zhou

, Liao

, Guo

, Wang

, Li

, Wu

, Xu

, Hu

, Pei

, Liao

, Yao

, Yang

, Zeng

and Xu

. (2021). Separated parabiont reveals the fate and lifespan of peripheral-derived immune cells in normal and ischaemia-induced injured kidneys. Open Biol, 11:200340.

30.

Mahnke

, Brodie

, Sallusto

, Roederer

and Lugli

. (2013). The who's who of T-cell differentiation: human memory T-cell subsets. Eur J Immunol, 43:2797–2809.

31.

Mueller

and Mackay

. (2016). Tissue-resident memory T cells: local specialists in immune defence. Nat Rev Immunol, 16:79–89.

32.

Asada

, Ginsberg

, Gagliani

, Mittrücker

and Panzer

. (2022). Tissue-resident memory T cells in the kidney. Semin Immunopathol, 44:801–811.

33.

Workman

, Szymczak-Workman

, Collison

, Pillai

and Vignali

. (2009). The development and function of regulatory T cells. Cell Mol Life Sci, 66:2603–2622.

34.

Alikhan

, Huynh

, Kitching

and Ooi

. (2018). Regulatory T cells in renal disease. Clin Transl Immunol, 7:e1004.

35.

Meng

, Zheng

, Zhou

, Gao

, Ni

, Liang

and Chen

. (2021). FoxP3- Tr1 cell in generalized myasthenia gravis and its relationship with the anti-AChR antibody and immunomodulatory cytokines. Front Neurol, 12:755356.

36.

Wang

, Chen

, Song

, Shen

, Li

, Fang

and Chen

. (2021). CD226 deficiency attenuates the homeostasis and suppressive capacity of Tr1 cells. Mol Immunol, 132:192–198.

37.

Scheinecker

, Göschl

and Bonelli

. (2020). Treg cells in health and autoimmune diseases: New insights from single cell analysis. J Autoimmun, 110:102376.

38.

D'Agati

, Appel

, Estes

, Knowles

, 2nd and Pirani

. (1986). Monoclonal antibody identification of infiltrating mononuclear leukocytes in lupus nephritis. Kidney Int, 30:573–581.

39.

Mei

, Li

, Zhou

, Cheng

and Cui

. (2022). The emerging role of tissue-resident memory CD8+ T lymphocytes in human digestive tract cancers. Front Oncol, 11:819505.

40.

Winchester

, Wiesendanger

, Zhang

, Steshenko

, Peterson

, Geraldino-Pardilla

, Ruiz-Vazquez

and D'Agati

. (2012). Immunologic characteristics of intrarenal T cells: trafficking of expanded CD8+ T cell β-chain clonotypes in progressive lupus nephritis. Arthritis Rheumat, 64:1589–1600.

41.

Wynn

and Vannella

. (2016). Macrophages in tissue repair, regeneration, and fibrosis. Immunity, 44:450–462.

42.

Wen

, Yan

, Wang

and Liu

. (2021). Macrophage heterogeneity in kidney injury and fibrosis. Front Immunol, 12:681748.

43.

Rogacev

, Seiler

, Zawada

, Reichart

, Herath

, Roth

, Ulrich

, Fliser

and Heine

. (2011). CD14++CD16+ monocytes and cardiovascular outcome in patients with chronic kidney disease. Eur Heart J, 32:84–92.

44.

Perdiguero

and Geissmann

. (2016). The development and maintenance of resident macrophages. Nat Immunol, 17:2–8.

45.

Cyster

and Allen

CDC

. (2019). B cell responses: cell interaction dynamics and decisions. Cell, 177:524–540.

46.

Halliley

, Tipton

, Liesveld

, Rosenberg

, Darce

, Gregoretti

, Popova

, Kaminiski

, Fucile

, Albizua

, Kyu

, Chiang

, Bradley

, Burack

, Slifka

, Hammarlund

, Wu

, Zhao

, Walsh

, Falsey

, … Lee

. (2015). Long-lived plasma cells are contained within the CD19(-)CD38(hi)CD138(+) subset in human bone marrow. Immunity, 43:132–145.

47.

Wang

, Kwak

, Fassett

, Liu

, Yao

, Weng

, Xu

, Bache

, Mueller

and Chen

. (2017). Role of bone marrow-derived CD11c+ dendritic cells in systolic overload-induced left ventricular inflammation, fibrosis and hypertrophy. Basic Res Cardiol, 112:25.

48.

Mazzurana

, Forkel

, Rao

, Van Acker

, Kokkinou

, Ichiya

, Almer

, Höög

, Friberg

and Mjösberg

. (2019). Suppression of Aiolos and Ikaros expression by lenalidomide reduces human ILC3-ILC1/NK cell transdifferentiation. Eur J Immunol, 49:1344–1355.

49.

Goncalves

, Makalo

, Joof

, Burr

, Ramadhani

, Massae

, Malisa

, Mtuy

, Derrick

, Last

, Nabicassa

, Cassama

, Houghton

, Palmer

, Pickering

, Burton

, Mabey

, Bailey

, Goodier

, Holland

, … Roberts

. (2016). Differential frequency of NKG2C/KLRC2 deletion in distinct African populations and susceptibility to Trachoma: a new method for imputation of KLRC2 genotypes from SNP genotyping data. Human Genet, 135:939–951.

50.

Yang

, Zhu

, Tian

, Wang

, Guo

, Sun

, Jin

, Zhang

, Shi

, Gershwin

, Zhang

, Zhao

and Zhang

. (2021). Transcriptome landscape of double negative T cells by single-cell RNA sequencing. J Autoimmun, 121:102653.

51.

Choi

, Park

, Choi

, Jung

and Lee

. (2021). Cellular heterogeneity of circulating CD4+CD8+ double-positive T cells characterized by single-cell RNA sequencing. Sci Rep, 11:23607.

52.

Munk

, Duvald

, Pedersen

, Lohmann

, Keller

, Møller

, Ringgaard

, Buus

, Jespersen

and Eijken

. (2022). Dosing limitation for intra-renal arterial infusion of mesenchymal stromal cells. Int J Mol Sci, 23:8268.

53.

Ward

LSC

, Sheriff

, Marshall

, Manning

, Brill

, Nash

and McGettrick

. (2019). Podoplanin regulates the migration of mesenchymal stromal cells and their interaction with platelets. J Cell Sci, 132:jcs222067.

54.

Soukou

, Huber

and Krebs

. (2021). T cell plasticity in renal autoimmune disease. Cell Tissue Res, 385:323–333.

55.

Tsai

, Chien

, Chiu

, Su

, Chiu

, Hsiao

and Lin

. (2022). Lupus nephritis with corticosteroid responsiveness: molecular changes of CD46-mediated type 1 regulatory T cells. Pediatr Res, 92:1099–1107.

56.

Jang

, Jeong

, Kim

, Jang

, Kang

, Lee

, Bae

, Kim

and Youn

. (2016). Infusion of human bone marrow-derived mesenchymal stem cells alleviates autoimmune nephritis in a lupus model by suppressing follicular helper T-cell development. Cell Transpl, 25:1–15.

57.

Kato

and Perl

. (2018). Blockade of Treg cell differentiation and function by the interleukin-21-mechanistic target of rapamycin axis via suppression of autophagy in patients with systemic lupus erythematosus. Arthritis Rheumatol, 70:427–438.

58.

Louwe

, Forbes

, Bénézech

, Pridans

and Jenkins

. (2022). Cell origin and niche availability dictate the capacity of peritoneal macrophages to colonize the cavity and omentum. Immunol, 166:458–474.

59.

Chen

, Huang

, Yang

, Yiu

, Yu

, Tang

SCW

and Lan

. (2022). Single-cell RNA sequencing identified novel Nr4a1+ Ear2+ anti-inflammatory macrophage phenotype under myeloid-TLR4 dependent regulation in anti-glomerular basement membrane (GBM) crescentic glomerulonephritis (cGN). Adv Sci, 9:e2200668.

60.

Obrişcă

, Sorohan

, Tuţă

and Ismail

. (2021). Advances in lupus nephritis pathogenesis: from bench to bedside. Int Mol Sci, 22:3766.

61.

Liu

, Deng

, Meng

, Qiu

, Sun

, and Dai

. (2018). CD8+iTregs attenuate glomerular endothelial cell injury in lupus-prone mice through blocking the activation of p38 MAPK and NF-κB. Mol Immunol, 103:133–143.

62.

Munir

, Luu

, Clarke

, Nash

and McGettrick

. (2016). Comparative ability of mesenchymal stromal cells from different tissues to limit neutrophil recruitment to inflamed endothelium. PLoS One, 11:e0155161.

Single-Cell Sequencing Informs That Mesenchymal Stem Cell Alleviates Renal Injury Through Regulating Kidney Regional Immunity in Lupus Nephritis

Abstract

Introduction

Materials and Methods

Animals

Transplantation protocol of MSCs

Evaluation of kidney coefficient and spleen coefficient

Analysis of serum and urine samples

Kidney pathological analysis

Kidney immunofluorescent staining

Flow cytometry analysis

Antibody array analysis

10 × single-cell RNA sequencing

Single-cell mapping and clustering

Differential expression analysis

Functional enrichment analysis

Cell development trajectories construction

Cell–Cell interaction/communication analysis

Statistics

Results

MSC transplantation increased survival rate and improved renal function in LN mice

MSC transplantation improved renal pathological injury in LN mice

MSC downregulated the expression of renal inflammation-related proteins in LN mice

MSC significantly downregulated expression of immune-related genes in renal immune cells in LN mice

MSC significantly changed the immune cell landscape in renal regional immunity in LN mice

MSCs reduced Tcm in CD4+T cells in renal regional immunity in LN mice

MSCs increased Tr1 Treg in CD4+T cells, not Foxp3 Treg, in renal regional immunity in LN mice

MSCs reduced resident memory cytotoxic CD8+T cells (Trm CD8+T) in CD8+T cells in renal regional immunity in LN mice

MSCs increased anti-inflammatory resident macrophages, and reduced proinflammatory infiltrating macrophages in renal regional immunity in LN mice

MSCs facilitated the conversion of macrophages into renal resident macrophages and Ly6c lo ear2− macrophages in LN mice

MSCs reduced LLPCs in renal regional immunity in LN mice

MSCs reduced neutrophils and DC in renal regional immunity in LN mice

MSCs increased NK, DNT, and DPT cells in renal regional immunity in LN mice

MSCs reduced immune cell interaction in renal regional immunity in LN mice

Discussion

Ethics Approval and Consent to Participate

Availability of Data and Materials

Footnotes

Author Disclosure Statement

Funding Information

Supplementary Material

References

MSCs reduced Tcm in CD4⁺T cells in renal regional immunity in LN mice

MSCs increased Tr1 Treg in CD4⁺T cells, not Foxp3 Treg, in renal regional immunity in LN mice

MSCs reduced resident memory cytotoxic CD8⁺T cells (Trm CD8⁺T) in CD8⁺T cells in renal regional immunity in LN mice

MSCs facilitated the conversion of macrophages into renal resident macrophages and Ly6c lo ear2⁻ macrophages in LN mice