Overexpression of Circulating Soluble Nogo-B Improves Diabetic Kidney Disease by Protecting the Vasculature

sNogo-B overexpression in the circulation ameliorates diabetes-mediated GEC proliferation, reduction in glycocalyx lectin content, and podocyte loss. A: GEC proliferation was detected by CD31⁺/Ki67⁺ cells (white arrowhead). Diabetes significantly increases the average number of glomerular CD31⁺/Ki67⁺ cells by sevenfold, which was reduced by sNogo-B overexpression in the circulation (B; n = 8–12/group, average of 30–40 glomeruli/animal). Diabetes led to a 20% reduction in the thickness of the glycocalyx (assessed as lectin content), which was prevented by sNogo-B overexpression in the circulation (C; n = 5/group, with 10–15 capillary loops studied per mouse). Similarly, diabetes-mediated glomerular podocytes loss was ameliorated by sNogo-B overexpression in the circulation (D; n = 7–11/group). Elevated sNogo-B circulating levels had no effect on the diabetes-mediated increase in mesangial volume fraction (Vvmes) (E; n = 9–12/group) or GBM thickness (F; n = 7–11/group) (*P ≤ 0.02, ND-GFP vs. D-GFP; **P = 0.02, ND–sNogo-B vs. D–sNogo-B; #P ≤ 0.01, D-GFP vs. D–sNogo-B). ANOVA with LSD post hoc test (mean ± SD) for all comparisons. ●, AAV-GFP–treated mice; ○, AAV–sNogo-B–treated mice.

sNogo-B overexpression in the circulation ameliorates diabetes-mediated VEGF-A/VEGFR2 signaling and eNOS^Ser1177 phosphorylation in kidney cortex lysate. Diabetes-induced kidney cortex VEGF-A expression (A; n = 11–14/group, in duplicate) and phosphorylated (p-)VEGFR2^Tyr1173 (B; n = 7/group, in duplicate) was blunted in diabetic mice with sNogo-B overexpression in the circulation (*P ≤ 0.003, ND-GFP vs. D-GFP; #P ≤ 0.03, D-GFP vs. D–sNogo-B). Angpt1/Angpt2 ratio (C; n = 7 to 8/group, in duplicate) was reduced in diabetes (*,**P ≤ 0.03, ND-GFP vs. D-GFP and ND–sNogo-B vs. D–sNogo-B) but not altered by elevated sNogo-B circulating levels. The diabetes-mediated increase in ratio of p-eNOS^Ser1177/total eNOS was prevented by sNogo-B overexpression in the circulation (D, n = 8–10/group; *P = 0.017, ND-GFP vs. D-GFP; #P = 0.05, D-GFP vs. D–sNogo-B). E: Representative data showing Western blotting for p-eNOS^Ser1177, total eNOS, and α-tubulin as housekeeping gene. ANOVA with LSD post hoc test (mean ± SD) for all comparisons. ●, AAV-GFP–treated mice; ○, AAV–sNogo-B–treated mice.

sNogo-B overexpression in the circulation modulates diabetes-mediated AKT^Ser473 and GSK3β^Ser9 phosphorylation, while preventing diabetes-mediated β-catenin upregulation. Diabetes was paralleled by a two- to threefold increase in the ratio of phosphorylated (p-)AKT^Ser473/total AKT (A; n = 6–8/group), a significant elevation of the ratio of p-GSK3β^Ser9/total GSK3β (B; n = 8/group) and increased total β-catenin levels (C; n = 8/group) in kidney cortex lysates. Diabetes-mediated AKT and GSK3β phosphorylation and upregulation β-catenin levels were partially or totally prevented by sNogo-B overexpression in the circulation (#P ≤ 0.04, D-GFP vs. D–sNogo-B; *P ≤ 0.007, ND-GFP vs. D-GFP; **P = 0.0001, ND–sNogo-B vs. D–sNogo-B). ANOVA with LSD post hoc test (mean ± SD) for all comparisons. ●, AAV-GFP–treated mice; ○, AAV–sNogo-B–treated mice.

Full-length Nogo-B is expressed in glomeruli, and sNogo-B overexpression in the circulation prevents diabetes-mediated Nogo-B downregulation. A: EM showing the ultrastructure of the kidney glomerulus containing podocytes (POD), ECs, and the GBM. Immunogold labeling showed positive full-length Nogo-B expression in podocytes (black arrowheads) (Ai) and ECs (black arrows) (Aii). Collecting ducts (medullary section) (Aiii) are used as positive control, and negative control (omission of first antibody) is shown in Aiv. Scale bars, 500 nm. B (right): Positive signal for Nogo-B is observed by immunohistochemistry both in glomerular cells (black arrowheads) and in the cortical collecting duct (positive internal control, red arrowheads) (magnification ×40). B (left): Negative control-omission first anti–Nogo-B antiserum. C: Full-length Nogo-B protein expression was downregulated in kidney cortex cell lysate and isolated glomeruli of diabetic mice and was prevented by AAV–sNogo-B overexpression in the circulation (C; n = 10–13/group fort cortex lysate, n = 4 to 5 for isolated glomeruli; *P ≤ 0.045, ND-GFP vs. D-GFP; #P ≤ 0.015, D-GFP vs. D–sNogo-B). D: NgBR was detected in isolated glomeruli; no effect of diabetes or sNogo-B overexpression was noted. E: Nogo-B expression (brown staining) was detected by immunohistochemistry in kidney biopsies from patients with DN and TBMN. Positive staining in the cortical collecting ducts serves as positive internal control (red arrowheads). Scale bars, 100 µm. Quantitative data showing the area of the glomerular tuft containing positive Nogo-B staining (n = 10/group for both TBMN and DN with an average score from 10 glomeruli obtained for each biopsy; *P = 0.0001, TBMN vs. DN). C and D: ANOVA with LSD post hoc test (mean ± SD). E: Unpaired t test (mean ± SD). ●, AAV-GFP–treated mice; ○, AAV–sNogo-B–treated mice.

Incubation of GECs with HG and/or VEGF-A in vitro results in downregulation of full-length Nogo-B and parallel increase in sNogo-B secretion in the supernatant. Fully differentiated GECs were incubated with NG or HG for 72 h in the absence (vehicle [VEH]) or presence of VEGF-A (50 ng/mL). Full-length Nogo-B protein expression was significantly downregulated by HG and/or VEGF-A (A; n = 5 to 6, in duplicate). Conversely, HG and/or VEGF-A was paralleled with an increase in sNogo-B levels, expressed as pg/mL ⋅ µg of cell protein, in the supernatant (B; n = 4 to 5 in duplicate) (NG vs. HG, NG vs. VEGF-A, and NG vs. HG + VEGF-A, *P ≤ 0.03). A: ANOVA with LSD post hoc test (mean ± SD). B: Kruskal-Wallis and Mann-Whitney test (median and interquartile range). ●, NG + VEH; ○, HG + VEH; ■, NG + VEGF-A; □, HG + VEGF-A.