Macrophage Derived Galectin-3 Promotes Renal Fibrosis and Diabetic Kidney Disease by Enhancing TGFβ1 Signaling

Abstract

Over 30% of patients with type 2 diabetes develop diabetic kidney disease (DKD), which has emerged as a major contributor to end stage renal disease. Renal fibrosis represents the final pathological outcome of most chronic kidney disease, particularly DKD. This study demonstrates elevated levels of Galectin-3 (Gal3), a lectin associated with inflammatory and fibrotic conditions, in the plasma and kidneys of DKD mice. Positive correlations between Gal3 expression and renal fibrosis are observed in both DKD patients and mice. Macrophage-derived Gal3 is found to promote Transforming growth factor beta 1 (TGFβ1) signaling activation and renal fibrogenesis. Genetic ablation of Gal3 globally or specifically in macrophages, as well as pharmacological inhibition of Gal3, significantly attenuated kidney fibrosis in diabetic mice. Mechanistically, macrophage-derived Gal3 interacted with TGFβ receptor2 (TGFBR2) and Pro-TGFβ1, preventing TGFBR2 proteasomal degradation in fibroblasts and increasing TGFβ1 levels in the diabetic kidney. These events enhances TGFβ1 signaling activation and ultimately facilitated kidney fibrosis. The findings of this study suggest Gal3 as a potential therapeutic target for renal fibrosis and DKD.

Keywords: DKD; TGFβ1 signaling; galectin‐3; protein degradation; renal fibrosis.

Figure 1

Increased serum and kidney Gal3 levels in patients and mice with DKD. a) Serum Gal3 levels in control, diabetic (DM) and DKD patients (n = 15 individuals for control, n = 29 patients for DM, n = 28 patients for DKD). b, c) Human kidney LGALS3 mRNA levels in control, early DKD and advanced DKD patients (b) and the expression correlations of fibrotic genes (TGFB1, ACTA2, FN1, COL1A1, COL3A1) and LGALS3 (c) (from GSE142025 dataset,^{[ 26 ]} n = 9 patients for control, n = 6 patients for early DKD, n = 21 patients for advanced DKD). d) Schematic diagram of experiment. e, f) Kidney weight e) and urine albumin creatinine ratio (UACR) f) of db/db mice at different weeks of age (n = 7–9 mice for each group). g) PAS and Masson staining of db/db mice kidney (n = 6 mice for each group, scale bar 200 µm). h) Protein expression levels of Gal3, FN and aSMA in the kidney of db/db mice (n = 6 mice for each group). i, j) Plasma (i) and kidney (j) Gal3 levels in db/db mice (n = 7–9 mice for each group). k) Immunohistochemistry (IHC) staining of Gal3 in the kidney of db/db mice. n = 6 mice for each group, scale bar 200 µm. l) Plasma Gal3 levels in normal chow (NC) and HFD+STZ mice (n = 6 mice for NC, n = 9 mice for HFD+STZ). m) Kidney Lgals3 mRNA levels in NC and HFD+STZ mice (n = 6 mice for NC, n = 12 mice for HFD+STZ (12 w), n = 10 mice for HFD+STZ (32 w)). Data were analyzed by two‐tailed Student's t test and presented as the mean ± SEM. Pearson correlation coefficient was used to analyze gene expression correlations. * P < 0.05; ** P < 0.01; *** P < 0.001; compared with control mice or indicated groups.

Figure 2

Gal3 knockout ameliorates the kidney fibrosis of DKD and CKD mice. a) Schematic diagram of experiment. b,c) Plasma (b) and kidney (c) Gal3 levels in WT and Gal3‐KO mice with or without DKD (n = 6–11 mice per group). d) UACR in WT and Gal3‐KO mice with or without DKD (n = 8–13 mice per group). e, f) PAS, Masson and Sirius red staining of kidney in WT and Gal3‐KO mice with or without DKD, representative images (e) and statistical analysis (f) (n = 6–9 mice per group, scale bar 100 µm for PAS and 200 µm for Masson and Sirius red). g–i) TEM images (g) and statistical analysis of tight‐slit pore density (h) and thickness of GBM (i) in WT and Gal3‐KO mice kidney with DKD (n = 6 mice per group, scale bar 5 µm or 0.8 µm as indicated, red arrow, ECM deposition). j) Kidney mRNA expression levels of fibrosis related genes (n = 6–11 mice per group). k) Schematic diagram of experiment. l, m) Plasma (l) and kidney (m) Gal3 levels in WT and Gal3‐KO mice with or without CKD (n = 6–10 mice per group). n‐o, UACR n) and BUN o) of WT and Gal3‐KO mice with or without CKD (n = 6–10 mice per group). p, q) HE, Masson and Sirius red staining of kidney in WT and Gal3‐KO mice with or without CKD, representative images (p) and statistical analysis (q) (n = 6–9 mice per group, scale bar 100 µm for HE and 200 µm for Masson and Sirius red, black arrow, inflammatory cell infiltration). Data were analyzed by two‐tailed Student's t test and presented as the mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001; compared with WT CKD mice (o) or indicated groups.

Figure 3

Macrophage derived Gal3 induces kidney fibrosis in DKD mice. a) Kidney immunofluorescence co‐staining of Gal3 (red) and other marker proteins (AQP1, Podocalyxin, F4/80, aSMA) (green) in WT and Gal3‐KO mice with DKD, scale bar 100 µm. b–d) Gal3 and JNK signaling pathway protein levels in RAW 264.7 macropahges (b) with statistical analysis (c) and Gal3 levels in cell medium supernatant (d) after LG, HG, PA and LPS treatments. LG, 5.56 mM glucose, HG, 30mM glucose, PA, 200 µM palmitic acid, LPS, 20 ng/mL lipopolysaccharide. e) Schematic diagram of experiment. f, g) Plasma (f) and kidney (g) Gal3 levels in Gal3 f/f and Gal3 MKO mice with DKD (n = 7–8 mice per group). h) UACR in Gal3 f/f and Gal3 MKO mice with DKD (n = 7–10 mice per group). i, j) PAS, Masson and Sirius red staining of kidney in Gal3 f/f and Gal3 MKO mice with DKD, representative images (i) and statistical analysis (j) (n = 6 mice per group, scale bar 100 µm for PAS and 200 µm for Masson and Sirius red). k) Kidney mRNA expression levels of fibrosis related genes in Gal3 f/f and Gal3 MKO mice with DKD (n = 7–10 mice per group). Data were analyzed by two‐tailed Student's t test and presented as the mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001; compared with indicated groups.

Figure 4

Gal3 binds to TGFBR2 and enhances the activation effect of TGFβ1 on renal fibroblast. a) Lgals3 and Tgfb1 mRNA levels after Gal3 overexpression in RAW 264.7 macrophages. b) Protein levels of FN, aSMA, p‐Smad2, Smad2/3 and Gal3 in NRK‐49F fibroblasts after TGFβ1 and Gal3 treatments (n = 3 independent cell samples per group). c) Statistical analysis of (b). d) mRNA levels of Acta2 and Lgals3 in NRK‐49F fibroblasts after TGFβ1 and Gal3 treatments (n = 6 independent cell samples per group). e) Protein levels of FN, aSMA and GFP in NRK‐49F kidney fibroblasts after Gal3‐GFP overexpression and TGFβ1 treatment (n = 3 independent cell samples per group). f) Statistical analysis of (e). g) Co‐immunoprecipitation of Gal3‐GFP and TGFBR1‐Myc and TGFBR2‐Myc in 293T cells. h) Immunofluorescence co‐staining of Gal3 (red) and TGFBR2 (green) in DKD human kidney, scale bar 100 µm. i) Immunofluorescence co‐staining of TGFBR2 (reed) and Gal3‐GFP (green) together with phalloidin in NRK‐49F fibroblast after Gal3‐GFP treatment, Pearson's r value for the colocalization of Gal3‐GFP with TGFBR2, 0.75. The co‐staining was interrupted by Gal3 inhibitor TD139 (1 µM), Pearson's r value for the colocalization of Gal3‐GFP with TGFBR2, 0.51. Scale bar 20 µm. j) Structure diagram of TGFBR2 (transcript variant 1 and variant 2), orange circle, glycosylation site. k) Co‐immunoprecipitation of Gal3‐GFP and TGFBR2‐Myc in 293T cells after PNGase F treatment. l) Co‐immunoprecipitation of Gal3‐GFP and TGFBR2‐Myc with glycosylation site mutation in 293T cells. Data were analyzed by two‐tailed Student's t test and presented as the mean ± SEM. * p < 0.05; ** p < 0.01; *** p < 0.001; compared with indicated groups.

Figure 5

Gal3 inhibits the ubiquitin‐proteasome degradation pathway of TGFBR2. a) Protein levels of FN, aSMA and TGFBR2 in the kidney of WT and Gal3‐KO mice with or without DKD (n = 6 mice for NC‐fed WT, Gal3‐KO mice and DKD WT mice, n = 7 mice for DKD Gal3‐KO mice). b) statistical analysis of (a). c) mRNA levels of TGFβ1 receptors in WT mice with or without DKD (n = 6–10 mice per group). d) Protein levels of TGFBR2 in NRK‐49F fibroblasts after TGFβ1 and Gal3 treatments (n = 3 independent cell samples per group). e) mRNA levels of TGFβ1 receptors after TGFβ1 and Gal3 treatments (n = 6 independent cell samples per group). f) Human kidney mRNA levels of TGFBR2 in control, early DKD and advanced DKD patients (from GSE142025 dataset,^{[ 26 ]} n = 9 patients for control, n = 6 patients for early DKD, n = 21 patients for advanced DKD). g) Protein levels of TGFBR2 after Gal3 and CHX treatments (n = 3 independent biological experiments) and Gal3 levels in cell medium of NRK‐49F fibroblasts. h) Protein levels of TGFBR2‐Myc after Gal3 and CHX treatments (n = 3 independent biological experiments) and Gal3 levels in cell medium of 293T cells. i) Protein levels of TGFBR2 and Gal3 in NRK‐49F fibroblasts after Gal3, proteasome inhibitor (MG132 and Epoxomicin) and lysosome inhibitor (NH4Cl and Bafilomycin) treatments (n = 4 independent cell samples per group). j) Protein levels of TGFBR2 in NRK‐49F fibroblasts after MG132 and NH4Cl treatments with or without Gal3 (n = 4 independent cell samples per group). k) Gal3 inhibited the ubiquitination of TGFBR2 in NRK‐49F fibroblasts treated with CHX. l) Gal3 inhibits K48‐linked TGFBR2 ubiquitination. Data were analyzed by two‐tailed Student's t test and presented as the mean ± SEM. * p < 0.05; ** p < 0.01; compared with indicated groups.

Figure 6

Gal3 binds with Pro‐TGFβ1 and inhibits its degradation. a) Co‐immunoprecipitation of Gal3‐GFP and Pro‐TGFβ1‐His in 293T cells. b) Gal3 and Pro‐TGFβ1 binding affinity assay using a BLI system. c) Structure diagram of Pro‐TGFβ1, red circle, glycosylation site. d) Co‐immunoprecipitation of Gal3‐GFP and Pro‐TGFβ1‐His in 293T cells after PNGase F treatment. e) Co‐immunoprecipitation of Gal3‐GFP and Pro‐TGFβ1‐His with glycosylation site mutation in 293T cells. f) Protein levels of Pro‐TGFβ1 and LAP after Gal3 and CHX treatments (n = 3 independent biological experiments) and Gal3 levels in cell medium of 293T cells. g) Mature TGFβ1 levels in the kidney of WT and Gal3‐KO mice with or without DKD (n = 6–9 mice per group). h) Protein levels of FN and aSMA in NRK‐49F fibroblasts after TGFBR2 inhibitor LY2109761 and TGFβ1 signaling inhibitor PD169316 treatments with or without Gal3 and TGFβ1 (n = 4 independent cell samples per group). i) Statistical analysis of (h). Data were analyzed by two‐tailed Student's t test and presented as the mean ± SEM. * p < 0.05; ** p < 0.01; compared with CHX group or indicated groups.

Figure 7

Pharmacological inhibition of Gal3 ameliorates kidney fibrosis in DKD mice. a) Co‐immunoprecipitation of Gal3‐GFP and TGFβ1‐Myc in 293T cells after Gal3 inhibitors TD139 and GB1107 treatments. b) Gal3 inhibitor TD139 restored the ubiquitination of TGFBR2 in NRK‐49F fibroblasts after Gal3 and CHX treatments. c) Protein levels of TGFBR2 in NRK‐49F fibroblasts after Gal3, TD139 and CHX treatments (n = 3 independent cell samples per group). d) Protein levels of FN, aSMA, p‐Smad2 and Smad2/3 in NRK‐49F fibroblasts after Gal3, TD139 and TGFβ1 treatments (n = 3 independent cell samples per group). e) Statistical analysis of (d). f) Schematic diagram of experiment. g–i) UACR (g), representative kidney images (h) and kidney index (left kidney weight / body weight) (i) of DKD mice after GB1107 treatment. j) Volume of 24 hours urine in DKD mice after 8 weeks GB1107 treatment. k, l) Plasma (k) and kidney (l) Gal3 levels in DKD mice after GB1107 treatment. m, n) PAS, Masson and Sirius red staining of kidney in NC, DKD mice with vehicle or GB1107 treatments, representative images (m) and statistical analysis (n) (scale bar 100 µm for PAS and 200 µm for Masson and Sirius red). o–q) TEM images (o) and statistical analysis of tight‐slit pore density (p) and thickness of GBM (q) in NC, DKD mice with vehicle or GB1107 treatments (scale bar 5 µm or 0.8 µm as indicated, red arrow, ECM deposition). r) Kidney mRNA expression levels of fibrosis related genes in NC, DKD mice with vehicle or GB1107 treatments. n = 6–10 mice per group in g‐r. Data were analyzed by two‐tailed Student's t test and presented as the mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001; compared with indicated groups.