Intestinal CXCR6+ ILC3s migrate to the kidney and exacerbate renal fibrosis via IL-23 receptor signaling enhanced by PD-1 expression

Abstract

Group 3 innate lymphoid cells (ILC3s) regulate inflammation and tissue repair at mucosal sites, but whether these functions pertain to other tissues-like the kidneys-remains unclear. Here, we observed that renal fibrosis in humans was associated with increased ILC3s in the kidneys and blood. In mice, we showed that CXCR6⁺ ILC3s rapidly migrated from the intestinal mucosa and accumulated in the kidney via CXCL16 released from the injured tubules. Within the fibrotic kidney, ILC3s increased the expression of programmed cell death-1 (PD-1) and subsequent IL-17A production to directly activate myofibroblasts and fibrotic niche formation. ILC3 expression of PD-1 inhibited IL-23R endocytosis and consequently amplified the JAK2/STAT3/RORγt/IL-17A pathway that was essential for the pro-fibrogenic effect of ILC3s. Thus, we reveal a hitherto unrecognized migration pathway of ILC3s from the intestine to the kidney and the PD-1-dependent function of ILC3s in promoting renal fibrosis.

Keywords: group 3 innate lymphoid cells; gut-kidney axis; programmed death receptor-1; renal fibrosis.

Figure 1.. ILC3s accumulate in human and mouse renal fibrosis

(A-B) CKD patients’ kidney biopsies were divided into different regions according to α-SMA expression. Relationships between α-SMA MFI and immune cells’ density in these areas (n = 30) were analyzed. The Lollipop chart shows the correlation index and p value. The etiologies of CKD patients include IgA Nephropathy (IgAN), Membranous Nephropathy (MN), and Lupus Nephritis (LN). (C-E) Percentages of ILCs within total white blood cells (WBCs) and ILC subsets among ILCs in kidneys of HCs (n = 4) and CKD patients with or without renal fibrosis (n = 15). (F-G) Percentages of ILC subsets among ILCs and absolute number of ILC3s in PBMC of HCs (n = 16) and CKD patients with or without renal fibrosis (n = 51). (H) Correlation between ILC3s percentage among ILCs in PBMC and eGFR (ml/min/1.73m²) of HCs (n = 16) and CKD patients (n = 80). (I) Multiplex immunofluorescence staining of CKD patients’ kidney biopsies with CD3, RORγt, α-SMA and DAPI. Scale bars, 50 μm. The sections were divided into fibrotic areas and non-fibrotic areas according to α-SMA expression, and ILC3s counts in these areas were compared (n = 42). Relationship between α-SMA MFI and ILC3s counts in these areas was analyzed. The etiologies of CKD patients include IgAN, MN, and LN. (J) Absolute number and percentage of renal ILC3s, and concentration of circulating ILC3s in sham and UUO mice were examined by flow cytometry (n = 4–8 per group). (K) Absolute number and percentage of renal ILC3s in vehicle-treated control (n = 4) and FA14d mice were examined by flow cytometry (n = 6). (L-M) Immunofluorescence staining of CD3, RORγt, α-SMA, and DAPI in sham or UUO14d mice kidneys (L). White arrows indicate CD3⁻RORγt⁺ ILC3s.Scale bars, 50 μm. Relationship between α-SMA MFI and renal ILC3s number (n = 19) was analyzed (M). (N) Relationship between serum creatinine level (Scr) and renal ILC3s number in sham and UUO mice (n = 24). (O-P) Immunofluorescence staining of CD3, RORγt, α-SMA, and DAPI in vehicle-treated control and FA14d mice kidneys (O). White arrows indicate CD3⁻RORγt⁺ ILC3s. Scale bars, 50 μm. Relationship between α-SMA MFI and renal ILC3s number (n = 17) was analyzed (P). (Q) Relationships between Scr and blood urea nitrogen (BUN) with renal ILC3s number in vehicle-treated control and FA14d mice (n = 9). Data are pooled from or representative of two to four independent experiments and shown as mean ± SEM. Each symbol in C, E, G, H, J, K, N and Q represents an individual human or mouse. Each symbol in I, M and P represents an individual area or slide. Student’s t test (C, I (left), K), one-way ANOVA test (E, G, J) and Pearson’s correlation test (A, H, I (right), M, N, P, Q) were performed. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05; ns, not significant. See also Figure S1 and Tables S1–S3.

Figure 2.. Intestine-derived ILC3s rapidly accumulate in fibrotic kidney

(A) Representative immunofluorescence staining of Ki67 or CD69, RORγt, CD127, and DAPI in UUO3d kidneys. ILC3s are indicated as CD127⁺ RORγt⁺. White stars indicate Ki67⁻ or CD69⁻ ILC3s. Open white triangles indicate Ki67⁺ or CD69⁺ ILC3s. Scale bars, 50 μm. ILC3 expression of Ki67and CD69 in UUO3d and UUO14d kidneys was quantified (n = 5 per group). (B) Expression of Ki67, CD103, and CD49a on ILC3s in obstructed kidneys (n = 5 per group). (C) Immunofluorescence staining of UUO3d kidneys with CD3, RORγt, CD127 and DAPI. White stars indicate CD3⁻CD127⁺RORγt⁺ ILC3s. “V” indicates blood vessels. Scale bars, 50 μm. (D) ILC3s in UUO mice kidneys, PBMC, small intestines (SI), lungs, spleens and bone marrow (n = 4–8 per group). (E) SI-ILC3s in vehicle-treated control (n = 4) or FA14d mice (n = 5). (F) Kaede transgenic mice SI were photoconverted (green to red) immediately before UUO, and the migration of cells to the kidney was examined at UUO3d. (G) Confocal microscopy of Kaede red⁺ cells in Kaede-UUO mice kidneys. Cell nuclei were counterstained with DRAQ5. Scale bars, 20 μm. (H) Kaede-red expression in renal ILC3s of Kaede-UUO mice with or without intestinal photoconversion (n = 3 per group). (I) Kaede-red expression in renal ILCs, T cells, B cells, monocytes and neutrophils of Kaede-UUO mice with intestinal photoconversion (n = 3). (J) Kaede red⁺ immune cell proportions among total Kaede red⁺ cells in intestine-photoconverted Kaede-UUO mice kidneys (n = 3). Data are pooled from or representative of two to four independent experiments and shown as mean ± SEM. Each symbol in E represents an individual mouse. One-way ANOVA test (D) and Student’s t test (E), were performed. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05; ns, not significant. See also Figure S2.

Figure 3.. Intestinal ILC3s migrate into injured kidney via CXCR6-CXCL16 axis

(A) Strategy for scRNA-seq. (B) ILC subsets in kidney (left) and ILC3s in intestine and kidney (right) were analyzed by UMAP. (C) Unsupervised transcriptional trajectory of intestinal and renal ILC3s, colored by cell pseudo-time and samples. (D) Heatmap showing chemokine receptors in renal and intestinal ILC3s. (E-F) Expression of CCR2, CXCR6 and CXCR4 on ILC3s in UUO3d mice intestines or kidneys (n = 5 per group) were measured by flow cytometry. (G) Confocal images of mice kidneys with CXCL16. Scale bars, 50 μm. (H) Confocal images of human kidneys with CXCL16, CXCL16 expression in the kidneys of HCs (n = 3) and CKD patients (n = 5) were compared. Scale bars, 50 μm. (I-J) Serum sCXCL16 in mice (n = 5–14 per group) (I) and human (HCs, n = 15; non-fibrosis, n = 7; fibrosis, n = 43) (J) were detected with ELISA. (K-L) Relationships between serum sCXCL16 with ILC3s counts in UUO mice kidneys (n = 19) (K), and frequency of ILC3s in patients’ PBMC (n = 65) (L). (M) Experimental design for (N-P). (N-P) ILC3s in kidneys (N) and intestines (O) of WT-UUO mice (n = 4–6 per group), and Kaede red⁺ ILC3s in Kaede-UUO mice (n = 3 per group) (P) were measured by flow cytometry. (Q) SI-ILC3s in Cxcr6^gfp/gfp-sham and Cxcr6^gfp/gfp-UUO3d mice were analyzed by flow cytometry (n = 5 per group). (R) Experimental design for (S). (S) Kidney ILC3s were measured by flow cytometry (n = 5 per group). Data are pooled from or representative of two to four independent experiments and shown as mean ± SEM. Each symbol in E, I, J-L, N-Q and S represents an individual human or mouse. Student’s t test (H, Q), one-way ANOVA test (I, J, N-P, S) and Pearson’s correlation test (K, L) were performed. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05; ns, not significant. See also Figure S3.

Figure 4.. ILC3s mediate fibrotic niche formation and contribute to renal fibrosis

(A) Experimental design for (B-D). (B-C) Immunofluorescence staining of α-SMA and DAPI in kidneys (B). Scale bars, 100 μm. Fibrotic niche size (α-SMA) and α-SMA MFI were compared (n = 5 per group) (C). (D) Immunofluorescence staining of CD3, RORγt, α-SMA and DAPI in kidneys. Scale bars, 50 μm. White arrows indicate CD3⁻RORγt⁺ ILC3s. Relationship between fibrotic niche sizes (α-SMA) and ILC3 numbers in fibrotic niches was analyzed. (E) Experimental design for (F-G). (F) Immunofluorescence staining of α-SMA and DAPI in kidneys. Scale bars, 50 μm. Fibrotic niche sizes (α-SMA) and α-SMA MFI were compared (n = 5 per group). (G) Kidney expression of α-SMA, collagen I (COL I) and fibronectin was determined by Western blot (n = 3 per group). (H) Experimental design for (I-J). (I) Immunofluorescence staining of α-SMA and DAPI in kidneys. Scale bars, 50 μm. Fibrotic niche sizes (α-SMA) and α-SMA MFI were compared (n = 3 per group). (J) Kidney expression of α-SMA, COL I and fibronectin was determined by Western blot (n = 3 per group). (K) Experimental design for (L-M). (L) Immunofluorescence staining of α-SMA and DAPI in mice kidneys. Scale bars, 50 μm. Fibrotic niche sizes (α-SMA) and α-SMA MFI were compared (n = 4 per group). (M) Kidney expression of α-SMA, COL I and fibronectin was determined by Western blot (n = 4 per group). Data are pooled from or representative of two to four independent experiments and shown as mean ± SEM. Each symbol in D represents an individual mouse. Student’s t test (F, G, I, J, L, M), one-way ANOVA test (C) and Pearson’s correlation test (D) were performed. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05. See also Figure S4.

Figure 5.. Elevated PD-1 expression enhanced the pro-fibrotic role of kidney-infiltrating ILC3s

(A) Heatmap showing immune checkpoints in intestinal and renal ILC3s. (B) scRNA-seq data showing Cxcr6 expression on Pdcd1-positive ILC3s. (C) PD-1 and CXCR6 expression on renal ILC3s in UUO14d mice (n = 5). (D) PD-1 MFI on renal ILC3s and PD-1⁺ ILC3s percentage among ILC3s between sham and UUO mice (n = 4–8 per group). (E) PD-1 MFI on renal ILC3s and PD-1⁺ ILC3s percentage among ILC3s in vehicle-treated control (n = 4) and FA14d mice (n = 6). (F-G) Immunofluorescence staining of α-SMA, CD127, PD-1 and DAPI in UUO14d kidneys (F). ILCs are indicated as CD127⁺. Scale bars, 50 μm. White stars indicate PD-1⁺ ILCs. Open white triangles indicate PD-1⁻ ILCs. Area (i) and (ii) represent areas with mild or severe fibrosis respectively. PD-1⁺ ILCs frequency among ILCs in areas with mild fibrosis (n = 3) or severe fibrosis (n = 4) was compared (G). (H) Experimental design for (I-J). (I) Percentage of myofibroblasts (α-SMA⁺) in kidneys was measured by flow cytometry (n = 4 per group). (J) Kidney expression of α-SMA, COL I and fibronectin was determined by Western blot (n = 3 per group). Data are pooled from or representative of two to four independent experiments and shown as mean ± SEM. Each symbol in D, E and I represents an individual mouse. One-way ANOVA test (D, I, J) and Student’s t test (E, G) were performed. ****p < 0.0001; **p < 0.01; *p < 0.05; ns, not significant. See also Figure S5.

Figure 6.. PD-1 on ILC3s increases IL-17A production to enhance fibroblasts activation

(A) Differentially regulated genes between renal ILC3s of WT-UUO3d and Pdcd1^−/−-UUO3d mice. (B) Differentially expressed cytokine genes in Pdcd1^−/−- UUO3d-ILC3s compared to WT- UUO3d-ILC3s. The x and y axes show the Normalization Gene Expression (NGE) for each group, and Genes above the diagonal are increased and those below are decreased in Pdcd1^−/−-UUO3d-ILC3s. Genes outside the gray area are differentially expressed. (C-D) The renal cell suspension was stimulated with cell stimulation cocktails (eBioscience) and IL-17A expression in renal ILC3s of HCs and CKD patients (n = 5 per group) (C), and WT-UUO3d and Pdcd1^−/−-UUO3d mice were detected by flow cytometry and compared (D) (n = 3 per group). (E) Schematic of Pdcd1 targeting strategy. (F) Schematic of generation PD-1 conditional deletion mice and experimental design. (G) The renal cell suspension was stimulated with cell stimulation cocktails (eBioscience) and IL-17A expression in renal ILC3s of Pdcd1^fl/fl-UUO3d and Pdcd1^fl/flRorc-cre-UUO3d mice were detected by flow cytometry and compared (n = 3–4 per group). (H) Expression of α-SMA, COL I and fibronectin in Pdcd1^fl/fl-UUO3d and Pdcd1^fl/flRorc-cre-UUO3d mice kidneys were determined by Western blot (n = 3 per group). (I-K) Primary renal fibroblasts were cultured alone, stimulated with rmIL-17A or cocultured with WT-ILC3, Pdcd1^−/− -ILC3s for 24h or 48h. The neutralizing anti-IL-17A antibody was added in some systems (n = 4–6 per group). The expression of α-SMA, COL I and fibronectin by fibroblasts was determined by real-time PCR (GAPDH was the standard) or Western blot. Data are pooled from or representative of two to four independent experiments and shown as mean ± SEM. One-way ANOVA test (D, J) and Student’s t test (C, G, H, I) were performed. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05; ns, not significant. See also Figure S6 and Table S4.

Figure 7.. PD-1 amplifies IL-23-induced JAK2/STAT3/RORγt/IL-17A signaling in ILC3s by competitively binding IL-23R to inhibit endocytosis.

(A) RORγt expression of renal ILC3s from WT-UUO3d and Pdcd1^−/−-UUO3d mice were compared by flow cytometry (n = 3 per group). (B) WT-ILC3s and Pdcd1^−/−-ILC3s were cultured alone or stimulated with IL-23/IL-1β for 20 min. p-STAT3 expression was measured by flow cytometry (n = 2–6 per group). (C-D) WT-ILC3s and Pdcd1^−/−-ILC3s were activated by IL-23/IL-1β with or without Stattic for 48h. IL-17A (C) and RORγt expression (D) on ILC3s were measured by flow cytometry (n = 2–11 per group). (E-G) WT-ILC3s and Pdcd1^−/−-ILC3s were activated by IL-23/IL-1β for 2h or 48h. Surface IL-23R (E), p-JAK2 (F) and total IL-23R (G) were detected with flow cytometry or immunofluorescence (n = 3–9 per group). (H-J) WT-ILC3s and Pdcd1^−/−-ILC3s were preincubated for 30 min in the medium with or without chlorpromazine (CPZ), and then activated by IL-23 for 2h or 48h. Surface IL-23R (H), RORγt (I) and IL-17A (J) were detected by flow cytometry (n = 3 per group). (K) Structural prediction of PD-1 and IL-23R protein combination with AlphaFold2. The interfaces were colored by hydrophobicity. Yellow means hydrophobic and blue means hydrophilic. The sites of hydrophobic interactions were outlined by red dotted lines. (L-M) Co-immunoprecipitation of PD-1, IL-23R-3×FLAG, and AP2 in HEK-293T. (N) Schematic of PD-1 mechanism in ILC3s. Data are pooled from or representative of two to four independent experiments and shown as mean ± SEM. Student’s t test (A, F-J) and One-way ANOVA test (B-E) were performed. ****p < 0.0001; ***p < 0.001; **p < 0.01; *p < 0.05; ns, not significant. See also Figure S7.