Circulating urokinase receptor as a cause of focal segmental glomerulosclerosis

Abstract

Focal segmental glomerulosclerosis (FSGS) is a cause of proteinuric kidney disease, compromising both native and transplanted kidneys. Treatment is limited because of a complex pathogenesis, including unknown serum factors. Here we report that serum soluble urokinase receptor (suPAR) is elevated in two-thirds of subjects with primary FSGS, but not in people with other glomerular diseases. We further find that a higher concentration of suPAR before transplantation underlies an increased risk for recurrence of FSGS after transplantation. Using three mouse models, we explore the effects of suPAR on kidney function and morphology. We show that circulating suPAR activates podocyte β(3) integrin in both native and grafted kidneys, causing foot process effacement, proteinuria and FSGS-like glomerulopathy. Our findings suggest that the renal disease only develops when suPAR sufficiently activates podocyte β(3) integrin. Thus, the disease can be abrogated by lowering serum suPAR concentrations through plasmapheresis, or by interfering with the suPAR-β(3) integrin interaction through antibodies and small molecules targeting either uPAR or β(3) integrin. Our study identifies serum suPAR as a circulating factor that may cause FSGS.

Figure 1. suPAR measurement in the serum of subjects with glomerular disease

For transplant subjects, the suPAR values were measured from pretransplantation serum unless otherwise indicated. Data is presented as means ± s.e.m. (a) Serum suPAR concentration in subjects with glomerular disease and healthy human subjects. MN, membranous nephropathy. *P < 0.05 for FSGS versus MN and preeclampsia; ^#P < 0.001 for FSGS versus healthy, MCD relapse, and MCD remission. Note that the values highlighted with red or green dots in the healthy subject and FSGS columns are identical twin pairs; in each case, one is healthy and has a twin brother with FSGS. (b) Serum suPAR in different population of subjects with primary FSGS. **P < 0.01 for recurrent FSGS versus nonrecurrent FSGS and nontransplant primary FSGS, respectively. (c) Serum suPAR concentrations after transplantation. ^#P < 0.001. (d) Correlation analysis of pretransplantation suPAR with proteinuria after transplantation. Pearson r = 0.16, P = 0.50. (e) Correlation analysis of pretransplantation suPAR with eGFR. Pearson r = 0.36, P = 0.16. (f) Correlation analysis of suPAR after transplantation with eGFR. Pearson r = 0.10, P = 0.58.

Figure 2. suPAR binds to and activates β₃ integrin on podocytes

(a) Western blot showing that suPAR binds β₃ integrin (representative of three experiments). EIF1B–GFP, encoding a translation initiation factor, and Raver-Flag encoding a ribonucleoprotein served as negative binding controls. β₃ integrin is encoded by Itgb3. S, sPlaur_WT (encoding suPAR); M, Plaur_WT (encoding membrane-bound uPAR); IP, immunoprecipitation. (b) AP5 immunostaining of differentiated human podocytes incubated with suPAR-rich recurrent FSGS serum (rec-FSGS serum), co-treated with the monoclonal antibody to human uPAR (uPAR mAb), and with cycloRGDfv, a small molecule that blocks β₃ integrin activity. AP5-specific antibody detects the active form of β₃ integrin. Bovine serum, negative control; suPAR, recombinant human suPAR protein. (c) Immunohistochemistry of AP5 on kidney biopsies from patients with glomerular disease. Top, representative AP5 staining in the glomerulus of subjects with FSGS. Bottom, the percentage (mean ± s.e.m.) of AP5-positive glomeruli. *P < 0.05 for primary FSGS versus control; **P < 0.01 for recurrent FSGS versus control. (d) Double immunofluorescent staining in glomeruli of kidney grafts for AP5 (green) and the podocyte marker synaptopodin (red). Top and bottom left, AP5 in the graft glomerulus 2 h after reperfusion in recurrent and nonrecurrent transplant biopsies (n = 2 per group). Top right, AP5 signal in recurrent transplant biopsies (n = 3) and nonrecurrent grafts (n = 5). Bottom right, normal kidney sections (n = 2) and biopsies from acute T cell-mediated rejections (n = 3) served as controls. Scale bars, 30 µm.

Figure 3. suPAR serum concentrations and podocyte β₃ integrin activity determine treatment response to plasmapheresis in recurrent FSGS

(a) Human podocytes incubated with different pooled serum samples and assayed for β₃ integrin activity. MFI, mean fluorescence intensity. *P < 0.05 for nonrecurrent FSGS versus normal subjects, ***P < 0.001 for recurrent versus nonrecurrent FSGS or versus healthy subjects. The respective suPAR concentration of the pooled sera is marked in red. NS, normal (healthy) subject; NR, nonrecurrent FSGS; REC, recurrent FSGS (representative of three experiments). (b) Pharmacological modulation of β₃ integrin activity in podocytes. **P < 0.01 for cylcoRGDfv co-treated cells versus recurrent FSGS serum alone; ***P < 0.001 for uPAR-specific mAb co-treated cells versus recurrent FSGS serum alone. (c) suPAR in serum from subjects with recurrent FSGS (n = 4) before and after a course of plasmapheresis. **P < 0.01. (d) Effect of plasmapheresis on β₃ integrin activity in podocytes incubated with recurrent FSGS serum (n = 6), collected before and after serial treatment with plasmapheresis. ***P < 0.001. (e–h) Clinical cases of recurrent FSGS. Top graphs show serum suPAR, urine protein/creatinine ratio (g/g) and individual plasmapheresis treatment as indicated by arrows and plotted over time (d) from before (−1) to after transplantation. Bottom graphs and images show podocyte β₃ integrin activity measured by FACS (left) and immunofluorescence (right) as a result of incubation with pretransplantation serum, or with the after-transplantation serum collected after repetitive plasmapheresis treatments. As a reference, the mean concentration of AP5 from a is marked as a dashed line. (e,f) Patients who obtained full remission after pheresis. (g,h) Patients who did not achieve remission after pheresis. Scale bars, 30 µm. Whiskers in plots of AP5 activity and serum suPAR show minimum to maximum.

Figure 4. suPAR activates β₃ integrin and causes foot process effacement in Plaur^−/− mouse kidneys and albuminuria in Plaur^−/− mice

(a) Injection (i.v.) of high doses of recombinant mouse suPAR into Plaur^−/− mice (n = 4 per group) induces proteinuria. **P < 0.01 for mice injected with 20 µg of suPAR at 24 h versus mice injected with other doses or versus other time points. (b) Injection (i.v.) of high doses of recombinant suPAR deposits into podocytes. Green, uPAR; red, synaptopodin (Synpo). (c) AP5 activity induced in the podocytes of high-dosage suPAR-injected Plaur^−/− mice (n = 4). Green, AP5; red, Synpo. (d,e) LPS induced endogenous suPAR in wild-type mice (n = 6). (d) Serum suPAR concentrations in LPS-treated mice. ***P < 0.001 for LPS-injected mice at 24 h versus PBS control, and versus LPS-injected mice at 0 h. **P < 0.01 for LPS-injected mice at 48 h versus at 0 h. (e) Urinary suPAR concentrations. ***P < 0.001 for LPS-injected mice at 48 h versus 0 h, and versus PBS control at any time point. **P < 0.01 for LPS-injected mice at 24 h versus 0 h. (f) Generation of a hybrid-kidney mouse model. (g) Electron microscope analysis of the PBS (n = 3) or LPS (n = 5) treated hybrid kidney. (h) uPAR expression in the native or Plaur^−/− kidneys from the hybrid-kidney mice with or without LPS treatment. Scale bars, 30 µm in b,c and h; 250 nm in g. Error bars, means ± s.e.m. in a; means ± s.d. in d,e.

Figure 5. Sustained overexpression of suPAR in the blood of wild-type mice leads to an FSGS-like glomerulopathy

(a) Generation of β₃ integrin binding–deficient suPAR mutants. (b) Serum suPAR concentrations in the sPlaur_WT engineered mice. *P < 0.05 at day 7 versus day 0 (before initial electroporation) (c) Urinary suPAR in sPlaur_WT engineered mice. ***P < 0.001 for days 7, 14 and 28 versus day 0; *P < 0.05 for day 28 versus day 7. (n = 4 in each group). (d) Albuminuria in sPlaur_WT and sPlaur_E134A mice. *P < 0.05 for sPlaur_WT mice at day 7 versus before treatment or versus sPlaur_E134A mice at day 7. **P < 0.01 for sPlaur_WT engineered mice at day 14 versus before treatment or versus sPlaur_E134A treated mice at day 7 or 14. (e) Kidney EM analysis of sPlaur engineered mice. Podocyte damage is reflected by relating the length of effaced foot process (FP) to the total length of the glomerular basement membrane (GBM) analyzed. Scale bars, 1 µm for upper image, 250 nm for lower image. **P < 0.01. (f) Histochemistry and light microscopy of the kidney from sPlaur engineered mice. PAS, periodic acid–Schiff. Scale bars, 30 µm. (g) Histopathological alteration of the kidneys was semiquantitatively scored. *P < 0.05. Error bars, means ± s.e.m. in d; means ± s.d. in b,c,e and g.

Figure 6. Administration of blocking antibody to uPAR ameliorates suPAR-caused kidney damage. (n = 4 in each group)

(a) Proteinuria in the antibody treated sPlaur_WT mice. *P < 0.05 for sPlaur_WT mice receiving isotype control at day 7 versus before initial electroporation at day 0 or versus mice treated with antibody to uPAR at day 7; ***P < 0.01 for sPlaur_WT mice receiving isotype control at day 21 versus at day 0 or versus antibody to uPAR-treated mice at day 21. (b) Morphological examination of the antibody treated sPlaur_WT kidney. Scale bars, 30 µm. (c) Pathology score. *P < 0.05. (d) Electron microscopic analysis of the antibody treated kidney from sPlaur_WT engineered mice. **P < 0.01 for IgG isotype control versus uPAR-specific antibody–treated sPlaur_WT engineered mice with respect to the ratio of effaced foot process (FP) to total GBM length measured. Scale bar, 360 nm. Error bars, means ± s.e.m.