The protective roles of integrin α4β7 and Amphiregulin-expressing innate lymphoid cells in lupus nephritis

Abstract

Type 2 innate lymphoid cells (ILC2s) have emerged as key regulators of the immune response in renal inflammatory diseases such as lupus nephritis. However, the mechanisms underlying ILC2 adhesion and migration in the kidney remain poorly understood. Here, we revealed the critical role of integrin α4β7 in mediating renal ILC2 adhesion and function. We found that integrin α4β7 enables the retention of ILC2s in the kidney by binding to VCAM-1, E-cadherin, or fibronectin on structural cells. Moreover, integrin α4β7 knockdown reduced the production of the reparative cytokine amphiregulin (Areg) by ILC2s. In lupus nephritis, TLR7/9 signaling within the kidney microenvironment downregulates integrin α4β7 expression, leading to decreased Areg production and promoting the egress of ILC2s. Notably, IL-33 treatment upregulated integrin α4β7 and Areg expression in ILC2s, thereby enhancing survival and reducing inflammation in lupus nephritis. Together, these findings highlight the potential of targeting ILC2 adhesion as a therapeutic strategy for autoimmune kidney diseases.

Keywords: Adhesion molecules; Amphiregulin; Innate lymphoid cells; Integrins; Kidney; Lupus nephritis; Tissue residency.

Fig. 1

Reduced kidney ILC2 numbers linked to spontaneous lupus development in MRL-lpr mice. A–C Spleen morphology (A) and weight (B). Representative images of H&E and PAS staining showing the increased glomerular area in old MRL-lpr mice (A, C). Scale bars, 20 μm. D Serum anti-dsDNA IgG, BUN and creatinine (Cr) levels (n = 7–9). E Gene expression levels of Ifng, Tnf, Mcp-1, Il6, and Il1b in renal tissue were quantified using RT‒qPCR (n = 7). F Unbiased immunophenotyping of high-parameter flow cytometry data for CD45⁺ kidney immune cells. G–H Kidney ILCs (lineage⁻CD127⁺) and CD4⁺ T cells (lineage⁺ CD4⁺) (G) and their percentages among CD45⁺ immune cells (H) in 3- to 20-week-old MRL-lpr mice. (n = 4-5) (I) Flow cytometry analysis of Annexin V⁺ apoptotic cells and Ki-67⁺ proliferating cells among ILCs and CD4⁺ T cells (n = 8–10). J Experimental design for the scRNA-seq analysis of ILC-enriched renal cells from pooled samples from young and old MRL-lpr mice. K UMAP plot showing 20 distinct kidney cell types identified by unsupervised clustering (left panel) and their comparison across glomerulonephritis development in MRL-lpr mice (right panel). L Dot plot showing the expression of various inflammatory and regulatory cytokines in immune cell clusters. M Comparison of Areg and TGF-β1 expression assessed using flow cytometry (n = 5–6). All results are shown as the means ± SEMs, and the statistical analysis was performed using the Mann‒Whitney U test or Kruskal‒Wallis test. ns, not significant; *P < 0.05; **P < 0.01; and ***P < 0.001. DC dendritic cell, Mac or Mφ macrophage, Neu neutrophil, B B cell, T T cell, Ery erythrocyte, Epi epithelial cell, Endo endothelial cell, PT proximal tubule cell

Fig. 2

Decreased ILC2 frequencies in kidneys of mice with IMQ-induced lupus. A–C Spleen morphology (A) and weight (B) (n = 12). Representative images of H&E and PAS staining showing the increased glomerular area in the IMQ model (A, C; n = 7–9). Scale bars, 20 μm. D Serum anti-dsDNA IgG, BUN and creatinine (Cr) levels (n = 6–11). E Gene expression levels of Ifng, Tnf, Mcp-1, Il6, and Il1b in renal tissues were quantified using RT‒qPCR (n = 5). Kidney ILCs (lineage⁻CD127⁺) and CD4⁺ T cells (lineage⁺ CD4⁺) (F) and their percentages among CD45⁺ immune cells (G) in the IMQ model (n = 7–8). H Flow cytometry analysis of Annexin V⁺ apoptotic cells and Ki-67⁺ proliferating cells among ILCs and CD4⁺ T cells (n = 7). I Frequencies of AREG- and IL-5-expressing ILC2s in the IMQ model (n = 8). All results are shown as the means ± SEMs, and the statistical analysis was performed using the Mann‒Whitney U test. ns not significant, *P < 0.05; **P < 0.01; ***P < 0.001; and ****P < 0.0001

Fig. 3

Loss of α4β7 expressing ILC2s in lupus nephritis. A Experimental design for intravascular immune cell staining following the tail vein injection of the BV650-CD45 monoclonal antibody. B Comparison of resident (CD69⁺ i.v.CD45⁻) and intravascular (CD69⁻ i.v.CD45⁺) cell compositions between ILC2s and CD4⁺ T cells in the naïve kidney (n = 4). Violin plot showing the expression of integrins (C) and other adhesion- and migration-related molecules (D) across lymphocyte clusters according to the scRNA-seq analysis. E Representative histogram of flow cytometry data for adhesion- and migration-related molecules evaluated in (C, D), excluding S1pr1 and S1pr5, overlaid with the FMO controls. F Heatmap showing the expression of adhesion molecules assessed using flow cytometry in the MRL-lpr and IMQ models, based on the frequency of expressing cells from Fig. 3G, H and Supplementary Fig. 3B, C. G, H Frequency of integrin α4β7-expressing kidney ILC2s in the MRL-lpr (G; n = 4–5) and IMQ lupus models (H; n = 8). I, J Representative flow cytometry plot (I) and frequency of splenic ST2⁺ CD25⁺ ILC2s in the MRL-lpr (n = 5–10) and IMQ lupus models (n = 11–12). All results are shown as the means ± SEMs, and the statistical analysis was performed using the Mann‒Whitney U test or Kruskal‒Wallis test. **P < 0.01, ***P < 0.001, and ****P < 0.0001

Fig. 4

α4β7 integrin mediates kidney ILC2 retention. A Expression of Madcam1, Vcam1, Cdh1, and Fn1 in various clusters of kidney cells from public mouse kidney scRNA-seq data assessed using KidneyCellExplorer (http://cello.shinyapps.io/kidneycellexplorer/). Representative flow cytometry plot (B) and frequencies (C) of MAdCAM-1, VCAM-1, E-cadherin, and fibronectin expression in endothelial (CD31⁺), epithelial (EpCAM⁺), and other nonimmune cells (CD45⁻ CD31⁻ EpCAM⁻) from the naïve mouse kidney (n = 9). D Expression of key adhesion molecules in the renal intrinsic cells from the IMQ model (n = 9). E Receptor‒ligand interactions between ILC2s and renal intrinsic cell clusters through Itga4 (up) and Itgb7 (down) expressed on ILC2s, as assessed by calculating the ligand‒receptor score (L-R score) from the analysis using the SingleCellSignalR algorithm. F Graphical summary showing the predicted interactions between ILC2s and renal intrinsic cells. All results are shown as the means ± SEMs, and the statistical analysis was performed using the Mann‒Whitney U test. ns not significant; ***P < 0.001. E-cad E-cadherin, FN fibronectin, Epi epithelial cells, Endo endothelial cells, PT proximal tubule cells

Fig. 5

Integrin α4β7 guides renal ILC2 adhesion and migration. A Experimental design for evaluating the adhesion of kidney ILC2s to prospective ligands expressed on renal tissues using live cell imaging. B Relative adhesion of ILC2s to the ligands compared to the control group (control, n = 10; MAdCAM-1, VCAM-1, and E-cadherin, n = 7; fibronectin, n = 3). C Images of ILC2s adhered to ligand-coated plates. Scale bars, 10 μm. D Trajectories of the ILC2s tracked over time on ligand-coated plates. E Velocity, accumulated distance, and Euclidian distance of the ILC2 trajectory on ligand-coated plates. F Expression of integrin α4β7 and β1 in human kidney ILCs (n = 4). G Images of human blood ILCs adhered to ligand-coated wells. Scale bars, 10 μm. All results are shown as the means ± SEMs, and the statistical analysis was performed using the Kruskal‒Wallis test. ns not significant, *P < 0.05; ***P < 0.001; and ****P < 0.0001

Fig. 6

Reduced integrin expression in kidney ILC2s via TLR7/9. A Experimental protocol for in vivo TLR7 agonist treatment of IL-33-treated mice. Representative histograms (B) and frequencies (C) of integrin α4-, β7-, αv-, and β1-expressing ILC2s in IL-33-treated mice exposed to IMQ (n = 8). D Frequency of ILC2s in recombinant IL-33-treated mice exposed to IMQ (n = 8). E Experimental design for blocking integrin β7 in recombinant IL-33-treated old MRL-lpr (17 weeks old) mice. F Representative flow cytometry plot of ILCs and their frequencies after blocking integrin β7 in recombinant IL-33-treated old lpr mice (n = 6–9). G Graphical scheme of Itga4 MO-mediated downregulation of gene expression. H Expression (MFI) of integrin α4 and β7 after Itga4 MO transfection of kidney ILC2s (n = 5). I Expression (MFI) of CD127, ST2, and CD25 after Itga4 MO transfection of kidney ILC2s (n = 5). J Frequencies of AREG and TGF-β1 induced by Itga4 MO in kidney ILC2s (n = 5). All results are shown as the means ± SEMs, and the statistical analysis was performed using the Mann‒Whitney U test or Kruskal‒Wallis test. ns not significant; **P < 0.01; and ***P < 0.001

Fig. 7

IL-33 enhances kidney ILC2s and renal function in lupus nephritis. A Experimental design of in vivo ILC2 expansion in the IMQ model. B Survival kinetics of the IL-33-treated IMQ model mice (n = 14). C Glomerular size in the IL-33-treated IMQ model mice was assessed using H&E staining (n = 7–10). D Gene expression levels of Ifng, Il17a, Il1b, Il6, and Ccl2 in renal tissue were quantified using RT‒qPCR (n = 7‒9). E Frequencies of ILCs in the IMQ model mice after IL-33 treatment (n = 10–13). F, G Flow cytometry analysis of the expression of integrin α4 (F) and β7 (G) in kidney ILC2s (n = 3–5). H Glomerular size in the ILC2 adoptive transfer (ivILC2) or rAREG treated IMQ model mice was assessed using H&E staining (n = 6–11). I Experimental design of the in vitro assay in which splenic CD4⁺ T cells from old MRL-lpr mice and IL-33-treated kidney ILC2s from young MRL-lpr mice were cocultured. J Frequencies of IFN-γ and IL-17A in splenic CD4⁺ T cells from old MRL-lpr mice cocultured with IL-33-treated kidney ILC2s from young MRL-lpr mice. K Comparison of Areg expression between ILC2s and CD4⁺ T cells from cocultures (n = 5). All results are shown as the means ± SEMs, and the statistical analysis was performed using the Mann‒Whitney U test or Kruskal‒Wallis test. ns not significant; *P < 0.05; **P < 0.01; and ****P < 0.0001