SuPAR mediates viral response proteinuria by rapidly changing podocyte function

Abstract

Elevation in soluble urokinase receptor (suPAR) and proteinuria are common signs in patients with moderate to severe coronavirus disease 2019 (COVID-19). Here we characterize a new type of proteinuria originating as part of a viral response. Inoculation of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes increased suPAR levels and glomerulopathy in African green monkeys. Using an engineered mouse model with high suPAR expression, inhaled variants of SARS-CoV-2 spike S1 protein elicite proteinuria that could be blocked by either suPAR antibody or SARS-CoV-2 vaccination. In a cohort of 1991 COVID-19 patients, suPAR levels exhibit a stepwise association with proteinuria in non-Omicron, but not in Omicron infections, supporting our findings of biophysical and functional differences between variants of SARS-CoV-2 spike S1 protein and their binding to podocyte integrins. These insights are not limited to SARS-CoV-2 and define viral response proteinuria (VRP) as an innate immune mechanism and co-activation of podocyte integrins.

Fig. 1. Induction of glomerular disease in AGMs by SARS-CoV-2 2019-nCov inoculation.

a Experimental scheme for SARS-Cov-2 2019-nCov inoculation (n = 4 per group, 2 male, 2 female). Age and gender-matched control African green monkeys (AGMs) were either mock-infected with culture media used for virus propagation (n = 1) or untreated (n = 3 biological replicates). The illustration of AGM was created with BioRender.com. b Plasma soluble urokinase receptor (suPAR) levels in 2019-nCov inoculated AGMs upon nephropsy (n = 4 per group). P = 0.0034, two-tailed t test. c Proteinuria in AGMs upon nephropsy (n = 4 per group). Alb, albumin; Creat, creatinine. P = 0.011, two-tailed t test. d Renal histological examination, Periodic acid-Schiff (PAS) staining. Black scale bar, 100 μm. e Renal histological examination, Sirius red staining. Black scale bar, 150 μm. d, e The experiments were repeated four times, generating similar results. f Maximum intensity projection images obtained with podocyte exact morphology measurement procedure (PEMP), indicating the disorganized slit diaphgram in “2019-nCov”. White scale bar, 2 μm. g Calculated filtration slit density (FSD) that reflects the extent of foot process effacement (n = 30 replicates for control, n = 60 replicates for 2019-nCov). The analysis was conducted with two-sided Mann–Whitney test. b, c, g Data were presented as Mean ± SD. *P < 0.05, **P < 0.01, ****P < 0.0001.

Fig. 2. Induction of glomerular disease in male mice with high levels of suPAR via intranasal inhalation of 2019-nCov spike S1 protein.

a Experimental scheme. I/N, intranasal (n = 8 mice per group, all male). The illustrations of SARS-Cov-2 and mouse were created with BioRender.com. b Proteinuria. 2019-nCov spike S1 protein inoculation induced proteinuria in mouse soluble urokinase receptor transgenic (suPAR-Tg) mice, but not in wild-type B6, or mouse urokinase receptor (uPAR) deficiency (Plaur^−/−) mice (n = 8 per group, male). Alb albumin. Creat creatinine. c Transmission electron microscopy (TEM) analyses. TEM images showing glomerular filtration barrier and foot process effacement (red arrow). Representative of ten images per group. Blue scale bar, 1 μm. d Foot process count per μm GBM. FP foot process. GBM glomerular basement membrane (n = 10 replicates per group). e Blocking suPAR with uPAR monoclonal antibody (Mab) as shown in treatment scheme reduced proteinuria in male suPAR-Tg mice upon 2019-nCov spike S1 protein inoculation (n = 6 per group). IgG isotype IgG control. D day. f Vaccination with single dose BNT162b2 as shown in the treatment scheme prevented 2019-nCov spike S1 protein induced proteinuria in male suPAR-Tg mice (n = 6 per group). Saline, phosphate-buffered saline (PBS). q.d. once a day, I.M, intramuscular. I/N, intranasal. In (b), (e) and (f), two-way analysis of variance with Tukey’s multiple comparisons test was conducted. d One-way analysis of variance with Tukey’s multiple comparisons test was performed. Data are presented as mean ± SD. ***P < 0.001, ****P < 0.0001.

Fig. 3. Viral protein-caused proteinuria in suPAR-Tg mice was SARS-CoV-2 variant-dependent and virus-specific.

a Inoculation of SARS-CoV-2 2019-nCov nucleocapsid protein in the same scheme as with 2019-nCov S1 protein did not cause proteinuria in male mice with high levels of suPAR (n = 6). P = 0.474, non-significant, two-sided Mann–Whitney test. b Inoculation of SARS-CoV-2 Delta variant (B.1.617.2) S1 protein induced proteinuria in male mice with high levels of suPAR (n = 5). P = 0.0079, two-sided Mann–Whitney test. c Inoculation of SARS-CoV-2 Omicron variant (B.1.1.529) S1 protein in the same scheme as with 2019-nCov S1 failed to elicit proteinuria in male suPAR-Tg mice (n = 6 biological replicates). P = 0.5714, non-significant, two-sided Mann–Whitney test. d Injection of human immunodeficiency virus type 1 (HIV-1) envelop gp120 protein caused proteinuria in suPAR-Tg mice (n = 6, 3 male, 3 female). P = 0.0087, two-sided Mann–Whitney test. e Injection of hepatitis C virus (HCV) E2 protein did not cause proteinuria in suPAR-Tg mice (n = 4, 2 male, 2 female). P = 0.8286, non-significant. f suPAR-Tg mice inoculated with influenza B HA protein did not develop proteinuria (n = 4, male). HA hemagglutinin A. P = 0.6571, non-significant, two-sided Mann–Whitney test. I/N intranasal. I.P intraperitoneal. Data are presented as mean ± SD in all graphs. **P < 0.01.

Fig. 4. Omicron S1 protein distinguished from its 2019-nCov counterpart both biophysically and functionally.

a–d S1 protein binding affinity as indicated by surface plasmon resonance assays. As shown, S1 protein was immobilized onto a CM5 sensor chip, while αvβ3 integrin or ACE2 was applied as an analyte in a series of increasing concentrations. Calculation of K_D value indicates that 2019-nCov S1 protein bound more tightly to ACE2 (a) and αvβ3 integrin (c), as compared to Omicron S1 protein (b, d). e, f The effects of S1 protein on cultured human podocytes. Fully differentiated human podocytes were treated for 16 h before harvest for immunofluorescence staining and imaging (n = 5 biological replicates). 2019-nCov but not Omicron S1 protein enhanced ACE2 expression in human podocytes (e). 2019-nCov but not Omicron S1 protein induced αvβ3 integrin activity in human podocytes, as indicated by AP5 (an antibody specific for αvβ3 integrin activation) immunostaining (f). Notably, P < 0.01 for 2019-nCov S1 compared to Omicron S1 in both (e, f). MFI mean fluorescence intensity. Data are presented as mean ± SD. The analyses were conducted with two-way analysis of variance with Tukey’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 5. The association between suPAR and proteinuria in patients with COVID-19 was dependent on the underlying variant.

Violin plot depicting the median suPAR level across four different proteinuria classes [“none or trace”, 1+ (30–99 mg/dL), 2+ (100–300 mg/dL), or 3+ (>300 mg/dL) respectively], stratified by non-Omicron and Omicron SARS-Cov-2 infections. The central line in box represents median, box represents the interquartile, and whiskers represent the minimum and maximum, respectively. Plot was generated using R with ggplot2 package. The sample size (n) of each class is shown below the graphs.