There is considerable evidence that ≥1 circulating plasma factors are involved in the pathogenesis of primary focal segmental glomerulosclerosis (FSGS) without podocyte gene mutations. The soluble identification of the circulating factor has been postulated for many years and has been elusive.1-4 The soluble urokinase plasminogen activator receptor (suPAR) has been suggested as a possible causative factor in FSGS; however, the role of suPAR in FSGS and recurrent FSGS remains controversial.5-8
Urokinase plasminogen activator receptor (uPAR) is a glycolipid-anchored cell surface receptor for urokinase plasminogen activator. The receptor is also involved in nonproteolytic pathways, mainly through its ability to form signaling complexes with other transmembrane proteins such as integrins, caveolin, and G protein–coupled receptors.9 Through these signaling pathways, the versatile uPAR receptor has important roles in inflammation, adhesion, proliferation, and mobilization as well as in severe pathological conditions, such as in malignancies.10-15
In the kidney, Wei et al5 showed that suPAR binds and activates the β3 integrin leading to downstream activation of GTPase that decreases podocyte stress fibers and results in effacement of podocytes and proteinuria. In mice studies, suPAR infusion induced changes of FSGS, which could be prevented by coadministration of anti-uPAR antibodies. However, 2 well-designed separate studies performed in mice failed to show that infusions of suPAR resulted in proteinuria and effacement of podocytes.21,22 A critique of these 2 studies was that suPAR was infused in wild-type (WT) mice, while the experiments performed by Wei et al5 utilize uPAR recombinant mice (uPAR−/−). Therefore, we embarked on experiments to repeat the studies of Wei et al5 and use the uPAR−/− mice to determine whether we could duplicate their observations on the role of suPAR on podocyte injury. We also extended the studies to evaluate the effect of suPAR on podocytes isolated from human glomeruli. Given that recent publications16-19 showing synergistic podocytopathic effects of suPAR and anti-CD40 autoantibody, we also coinjected human suPAR and human CD40 autoantibody isolated from the sera of patients with recurrent FSGS after kidney transplantation, into WT mice, using the same injection regimen as previously used by Wei et al.19
MATERIALS AND METHODS
All chemicals were manufactured by Sigma (St. Louis, MO), and cell culture and fluorescent detection reagents were obtained from Thermo Fisher Scientific (Waltham, MA), unless otherwise stated. Solvents were of analytical grade or higher. uPAR antagonists for blocking human recombinant soluble urokinase plasminogen activator receptor (shuPAR) and stimulation with mouse/human uPAR were performed using recombinant proteins from R&D Systems (Minneapolis, MN). For high-content analysis of podocytes health and morphology, we used rabbit anti-vinculin (Thermo Fisher Scientific, MA, 1:500) coupled to Alexa Fluor 647–conjugated goat anti-rabbit IgG, Alexa 488–conjugated phalloidin, and Hoechst 33342. AP-5 (mouse monoclonal antibody used to detect activated β3 integrin) mouse monoclonal antibody used to detect activated β3 integrin was obtained from the BloodCenter of Wisconsin (Milwaukee, WI).
Animals, Maintenance, and Euthanasia
All animal studies were conducted in accordance with the Principles of Laboratory Animal Care (NIH publication number 85-23, revised 1985) and in accordance with a University of California, San Francisco (San Francisco, CA) Institutional Animal Care and Use Committee protocol. WT C57BL/6 mice (6–9 wk old) were obtained from Harlen Laboratories. uPAR-deficient mice20 were kindly donated by Dr Thomas Bugge, National Institutes of Health. Mice were kept in specific pathogen-free conditions under constant environmental conditions (22°C, 12 h light/dark cycles) and fed with standard laboratory chow and 3% sucrose enriched water. Euthanasia of the sedated mice was performed by cervical dislocation.
Injection of Recombinant muPAR Into WT Mice or uPAR−/− Mice
C57BL/6 mice were injected with 0, 20, or 100 μg/mouse of mouse recombinant soluble uPAR-Fc chimera (smuPAR-Fc) chimera (mouse uPAR linked to a human IgG1) i.v. as described before.5,21,22 uPAR-deficient mice were injected with 100 μg/mouse of recombinant mouse smuPAR-Fc protein. As negative controls, mice were injected with Fc and vehicle PBS. Experiment design is illustrated in Figure 1A. Blood and urine were collected at 0, 10, and 24 hours and analyzed for smuPAR, albumin, and creatinine levels (Exocell Inc assays). smuPAR concentrations in the serum and blood were determined using a smuPAR ELISA. A standard curve was generated by adding the indicated amounts of purified recombinant smuPAR-Fc protein to serum or urine from uPAR−/− mice as described before.22 Mice were euthanized 24 hours after injection, and kidneys were retrieved.
Injection of Human CD40 Autoantibody and Human suPAR Into C57BL/6 Mice
As suPAR injection in WT mice did not show proteinuria, we further examined whether “priming” with CD40 autoantibody (isolated from sera of patients with recurrent FSGS after kidney transplantation) could result in proteinuria after suPAR injection. To examine this, we used the injection protocol previously used by Wei et al19 for both agents, and human CD40 autoantibody was injected daily for 6 doses followed by a single dose of suPAR.
Immunohistochemistry (uPAR, IgG1, Integrin β3) and Electron Microscopy on Mouse Kidneys
Immunostains for uPAR were performed on formalin-fixed paraffin-embedded (FFPE) mouse kidneys using standardized immunoperoxidase protocol with goat anti-mouse uPAR antibody (AF534, R&D System) following antigen retrieval at 100°C in pH 8.0, a secondary antibody Rabbit anti-goat HRP, and BOND Polymer detection RTU KIT with mouse HRP Polymer (Abcam) and DAB for detection. The IgG1 immunostain was performed on mouse FFPE kidneys with an anti-human IgG1 antibody (RevMAb Biosciences) using a standardized indirect immunofluorescence technology; human FFPE tonsil tissue was used as positive control. Immunostains against integrin β3 were performed on mouse frozen kidneys with an anti-integrin β3 (AP-5) antibody (Richard H. Aster, Blood Center of Wisconsin) using a standardized immunofluorescence protocol. All immunostains were preformed on a Leica Bond RX Autostainer platform. For electron microscopy, ultrathin (80 nm) sections of the glutaraldehyde-fixed, Epon-embedded mouse kidneys were stained with 2% uranyl acetate. Sections were examined in a Tecnai G212 transmission electron microscope at 80 kV, with images obtained with a Hammamatsu camera model Orca HR.
Immunohistochemistry (uPAR) on Human Kidney Biopsies
Immunostains for uPAR were performed on human frozen kidney biopsies using a mouse monoclonal antibody to uPAR (DAKO, M7294) using standardized indirect immunofluorescence and immunoperoxidase protocols on a Leica Bond RX Autostainer platform. Study groups included native biopsies with FSGS (n = 10) and early posttransplant recurrent FSGS before (n = 15) and after plasmapheresis (n = 7). Native biopsies with membranous nephropathy (MN; n = 10) and minimal change disease (MCD; n = 5) and normal 6-month posttransplant protocol biopsies (n = 10) served as kidney controls. Human lung and endometrial cancer specimens were used as positive controls.
Isolation, Culture, and Automated Microscopy of Primary Human Podocytes
Primary human podocytes were isolated from fresh normal tissues taken from surgically removed kidneys and cultured by a method used previously.23 We used previously described procedures to plate, process, and analyze podocytes using high-content analysis.23
Human Podocyte Injury via Examination of Morphology and Viability
Known triggers of podocyte injury, lipopolysaccharide (LPS) and puromycin aminonucleoside, were used as positive podocyte injury control. Culture medium was used for negative control. Injury was measured using high-content analysis, which compared changes following serial dilutions of shuPAR on podocyte morphology. This was quantified by the adherent cell count and the morphology of the podocyte nucleus (Figure 3A–F).
β3 Integrin Activation of Human Podocyte Primary Culture
Human podocytes were pretreated with suPAR 0.0039-1 μg/mL for 24 hours or with 50 mmol/L EDTA for 1 hour as positive control. The activated β3 integrin was detected by the AP-5 antibody. The AP-5 antibody was diluted in PBS without Ca2+ or Mg2+; cells were independently labeled using the AP-5 mouse monoclonal antibody at a 1:100 dilution of the stock concentration followed by an Alexa 488–conjugated goat anti-mouse antibody. The Spot Detector BioApplication was used to measure total AP-5 Ab signal/cell in peripheral adhesions as integrated brightness of all AP-5 foci per cell.
Quantitative variables were expressed as the mean ± SEM, and the qualitative variables were expressed as a proportion. Groups’ distributions were compared utilizing an exact version of the Wilcoxon rank-sum test. Statistical significance was defined as P <0.05.
Recombinant smuPAR Did Not Induce Proteinuria in WT Mice and Did Not Cause Proteinuria in uPAR-deficient Mice
We administered (i.v.) a commercially available uPAR-Fc chimera (smuPAR linked to a human IgG1 Fc; R&D Systems) to both WT C57BL/6 and uPAR−/− mice.20 WT mice were injected with 2 different doses of smuPAR-Fc chimera, 20 or 100 μg/mouse, and the uPAR-deficient mice were injected with 100 μg/mouse. As controls, mice were injected with Fc (human IgG1 Fc; R&D Systems) and vehicle PBS. Experimental design is illustrated in Figure 1A. The functionality of the injected smuPAR-Fc was verified by the ability of the receptor recognition of its ligand, urokinase plasminogen activator (uPA), and the functionality of the bound ligand by enzyme assay. The enzymatic activity of the bound uPA was measured by fluorogenic assays using serial dilutions (0–625 ng/mL) of mouse uPA (Active Mouse Urokinase, Molecular Innovations, Inc.; Figure 1B). No increase in urine protein excretion at 10 or 24 hours was detected by urine albumin/creatinine ratio analysis of WT smuPAR-Fc–injected mice relative to controls (Figure 1C). Injection of 100 μg smuPAR-Fc chimera to uPAR−/− mice leads to proteinuria; however, injection of Fc by itself was sufficient to induce proteinuria in those mice (Figure 1C). Circulating mouse uPAR concentrations in serum and urine were measured by ELISA (Figure 1D and E) against a recombinant smuPAR-Fc standard curve in serum or urine, respectively (Figure 1C), as described by Spinale et al.22 Priming with human anti-CD40 autoantibody followed by an injection of human suPAR did result in a modest increase in proteinuria (Figure 1F).
Infused smuPAR-FC Did Not Localize to Podocytes
Immunohistochemistry (IHC) revealed weak baseline endothelial expression of uPAR in control WT mice. Stronger uPAR signal in the glomerular endothelial cells was observed following injection of smuPAR-FC to WT mice compared with controls (Figure 2A). The strength of the IHC signal correlated with the smuPAR-FC dose with stronger uPAR signal in the mice injected with 100 µg compared with those injected with 20 μg. uPAR IHC was negative in the kidneys of uPAR knockout mice and remained negative following injection of 100 μg of suPAR or Fc control (Figure 2A). The anti-IgG1 stains in WT FFPE mouse kidneys from animals injected with 100 μg msuPAR-Fc or with Fc alone were negative.
The results of the AP-5 stains showed strong glomerular positivity with 100 μg dose and weak positivity with the 20 μg dose. However, the precise cellular localization of AP-5 could not be determined due to freezing artifacts, and therefore it is not clear if a super high dose of uPAR-Fc is required to demonstrate AP-5, a marker of β3 integrin activation, or it reflects nonspecific staining. Electron microscopy analysis of the smuPAR-FC–injected WT mouse kidney revealed minimal focal foot process effacement. No damage was detected in the uPAR−/− kidneys (Figure 2B).
shuPAR Did Not Cause Injury to Primary Human Podocytes
It has been reported that suPAR alters podocyte morphology and function in cell culture and animals. To evaluate whether suPAR directly mediates podocyte injury, we compared the effect of shuPAR to other known triggers of podocyte injury such as LPS and puromycin aminonucleoside or culture medium as the negative control (Figure 3). No difference was observed in cell viability between shuPAR-treated and shuPAR-untreated control podocytes by an MTS (tetrazolium compound [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) assay (Figure 3A). Injury was also measured using high-content analysis comparing the changes following serial dilutions of shuPAR on podocyte morphology as measured by high-content analysis of adherent cell count and nuclear morphology (Figure 3B and C). The positive controls, LPS and PAN, did affect podocyte viability, morphology, and adhesion as expected. F-actin rearrangement and AP-5 activation were not observed in the shuPAR-treated podocytes (Figure 3D–F).
uPAR Immunostaining of Human Kidneys
The uPAR immunostain in kidneys from patients with native kidney FSGS as well as in patients with early recurrence of FSGS (with demonstrated effacement of podocytes by EM) was negative. uPAR staining was also negative in patients with MCD and MN.
The initial excitement associated with the publication of the study by Wei et al5 that suPAR was the circulating factor associated with recurrent FSGS has been replaced by skepticism and frustration. The clinical findings of selective elevation of suPAR in FSGS patients and their ability to predict recurrence after transplantation have not been confirmed in several studies.22 However, several factors could potentially account for the discordant results, including the degree of renal failure of the patients because the glomerular filtration rate is an important determinant of suPAR levels, ethnicity, the heterogenicity of the disease itself, and methodology for measuring suPAR levels (although most, if not all, studies have used the commercially available ELISA assay). In fact, a critique of the suPAR studies is that the commercially available ELISA assay does not differentiate the different forms of circulating suPAR.8 In a recently published study from our laboratory, we used a time-resolved fluorescent immunoassay (TR-FIA) to measure the different fragments of circulating suPAR in patients with recurrent FSGS to determine whether a pathological fragment was present and could be associated or predictive of recurrence.8 Similar to the results obtained with the ELISA assay, the TR-FIA of the free fragment of suPAR was not found to be predictive of posttransplant recurrence.8 Furthermore, concerns have been raised on results of the elegantly performed experiments in mice by Wei et al5 that showed a robust causal relationship between suPAR and the development of histological and functional injury to the podocytes in vivo. Two independent studies in mice failed to duplicate the findings of Wei et al5 that infusion of suPAR induces podocyte injury and proteinuria.
Cathelin et al21 used 2 well-characterized recombinant forms of mouse suPAR produced by eukaryotic cells that were administered over the short and long term to WT mice. suPAR was deposited in the glomeruli of mice but did not alter the podocytes histologically (ie, foot effacement) or functionally (ie, proteinuria). Spinale et al22 used the commercially available Fc-mouse suPAR used by Wei et al5 at the same concentration of 20 μg in WT mice. The injected mice had a 6- to 12-fold increase in serum suPAR levels over 4–24 hours but did not develop proteinuria. An inducible transgenic mouse model that maintained elevated serum suPAR levels for 6 weeks did not also injure the glomeruli or induce proteinuria. Were the results of these 2 studies negative because both groups of investigators used WT rather than uPAR−/− mice? The investigators reasoned that unlike the uPAR−/− mice utilized in all the experiments by Wei et al,5 the WT mice are more physiological and represent a better experimental model.
In a different set of experiments, Alfano et al24 showed that infusion of 20 μg of recombinant murine suPAR using uPAR−/− mice induced deposits of suPAR in the glomeruli and increased proteinuria. In addition, Alfano et al24 observed downmodulation of nephrin. The investigators also reported similar findings in vitro using immortalized podocytes: suPAR downregulated nephrin expression and this effect was blocked by adding an antagonist to the αVβ3 integrin, supportive evidence that downstream signaling requires suPAR binding to the αVβ3 integrin. These studies add more complexities to understanding the suPAR effect. Is the suPAR mediated injury then only reproducible in uPAR−/− mice and WT mice are resistant to the effect of high circulating levels of suPAR?
We performed our mice studies to specifically determine if any differences exist in response to suPAR infusions between WT and uPAR−/− mice. We administered smuPAR-Fc to mice with the same dose used in previous studies, 20 μg, as well as the higher dose of 100 μg. Neither dose produced effacement of podocytes nor proteinuria in WT or uPAR−/− mice. By IHC, the infused suPAR deposited within the glomeruli but only in the endothelium and not in podocytes and only in WT mice. We also could not demonstrate a difference in viability or morphology of the cytoskeleton of human podocytes treated with suPAR in contrast to the podocyte cytoskeleton disruption observed with LPS and puromycin. In contrast to previous reports, activation of β3 integrin signaling has not observed when shuPAR was added to human podocytes as AP-5 antibody labeling did not show increased signals in focal adhesions (Figure 3F). Interestingly, a small surge in proteinuria was seen after suPAR was coinjected with anti-CD40 autoantibody, isolated from the plasma of patients with recurrence FSGS, following a short course of prior priming with this antibody, which supports prior observations of multiple pathways possibly acting in synergy to drive podocyte injury. The exact role and pathogenetic mechanism of podocyte injury of CD40 antibodies require additional investigation. Delville et al16 showed that the presence of anti-CD40 antibodies predicted recurrence of FSGS post-transplant and in vitro podocyte injury. In fact, a clinical trial with Bleselumab, a humanized anti-CD40 monoclonal antibody which presumably neutralizes the pathogenic CD40 autoantibodies, is underway in FSGS patients to prevent recurrence of the disease (Clinicaltrials.gov; NCT02921789).
Our observation of a lack of immunostaining for uPAR in kidney diseases characterized by podocyte effacement further complicates our understanding of the role of suPAR on podocyte injury in FSGS.
Our studies in mice and patients do not necessarily relegate suPAR to irrelevancy in FSGS. A follow-up study in WT mice performed by Wei et al19 showed that podocyte effacement and proteinuria occurred when suPAR was coadministered patients with anti-CD40 antibodies isolated from plasma of patient with recurrent FSGS but not when only suPAR was infused. In fact, neither the esuPAR nor the anti-CD40 alone could produce podocyte injury.19 In summary, suPAR role in FSGS and recurrent FSGS as well as in experimental models is more complex than initially suggested. The synergistic dual effect of suPAR and anti-CD40 antibodies requires additional studies. Before therapeutic interventions are initiated, a better understanding of the role of suPAR in FSGS and recurrent FSGS will be required.
We would like to thank Dr Thomas Bugge from the National Institutes of Health for donating urokinase plasminogen activator receptor–deficient mice. We also thank Murali Gopalakrishnan for project guidance and Stella Markosyan for management of logistics. We thank the AbbVie-University of California, San Francisco (UCSF) the Joint Scientific Committee and Michael Ploug for useful discussions on the project. We thank Donghui Wang (UCSF, USA) for help with the sample collection in mouse models.
1. Artero ML, Sharma R, Savin VJ, et al. Plasmapheresis reduces proteinuria and serum capacity to injure glomeruli in patients with recurrent focal glomerulosclerosis. Am J Kidney Dis. 1994; 23:574–581
2. Cattran D, Neogi T, Sharma R, et al. Serial estimates of serum permeability activity and clinical correlates in patients with native kidney focal segmental glomerulosclerosis. J Am Soc Nephrol. 2003; 14:448–453
3. Savin VJ, Sharma R, Sharma M, et al. Circulating factor associated with increased glomerular permeability to albumin in recurrent focal segmental glomerulosclerosis. N Engl J Med. 1996; 334:878–883
4. Vinai M, Waber P, Seikaly MG. Recurrence of focal segmental glomerulosclerosis in renal allograft: an in-depth review. Pediatr Transplant. 2010; 14:314–325
5. Wei C, El Hindi S, Li J, et al. Circulating urokinase receptor as a cause of focal segmental glomerulosclerosis. Nat Med. 2011; 17:952–960
6. Königshausen E, Sellin L. Circulating permeability factors in primary focal segmental glomerulosclerosis: a review of proposed candidates. Biomed Res Int. 2016; 2016:3765608
7. Wei C, Trachtman H, Li J, et al; PodoNet and FSGS CT Study Consortia. Circulating suPAR in two cohorts of primary FSGS. J Am Soc Nephrol. 2012; 23:2051–2059
8. Harel E, Shoji J, Abraham V, et al. Identifying a potential biomarker for primary focal segmental glomerulosclerosis and its association with recurrence after transplantation. Clin Transplant. 2019; 33:e13487
9. Ploug M, Rønne E, Behrendt N, et al. Cellular receptor for urokinase plasminogen activator. Carboxyl-terminal processing and membrane anchoring by glycosyl-phosphatidylinositol. J Biol Chem. 1991; 266:1926–1933
10. Blasi F, Carmeliet P. uPAR: a versatile signalling orchestrator. Nat Rev Mol Cell Biol. 2002; 3:932–943
11. Høyer-Hansen G, Lund IK. Urokinase receptor variants in tissue and body fluids. Adv Clin Chem. 2007; 44:65–102
12. Høyer-Hansen G, Pessara U, Holm A, et al. Urokinase-catalysed cleavage of the urokinase receptor requires an intact glycolipid anchor. Biochem J. 2001; 358Pt 3673–679
13. Smith HW, Marshall CJ. Regulation of cell signalling by uPAR. Nat Rev Mol Cell Biol. 2010; 11:23–36
14. Thunø M, Macho B, Eugen-Olsen J. SuPAR: the molecular crystal ball. Dis Markers. 2009; 27:157–172
15. Thurison T, Lomholt AF, Rasch MG, et al. A new assay for measurement of the liberated domain I of the urokinase receptor in plasma improves the prediction of survival in colorectal cancer. Clin Chem. 2010; 56:1636–1640
16. Delville M, Sigdel TK, Wei C, et al. A circulating antibody panel for pretransplant prediction of FSGS recurrence after kidney transplantation. Sci Transl Med. 2014; 6:256ra136
17. Donner AJ, Yeh ST, Hung G, et al. CD40 generation 2.5 antisense oligonucleotide treatment attenuates doxorubicin-induced nephropathy and kidney inflammation. Mol Ther Nucleic Acids. 2015; 4:e265
18. Beaudreuil S, Lorenzo HK, Durrbach A. The anti-CD40 auto-antibody: a biomarker or a factor for the permeability of recurrent focal segmental glomerulosclerosis? Ann Transl Med. 2015; 3:115
19. Wei C, Sigdel TK, Sarwal MM, et al. Circulating CD40 autoantibody and suPAR synergy drives glomerular injury. Ann Transl Med. 2015; 3:300
20. Bugge TH, Flick MJ, Danton MJ, et al. Urokinase-type plasminogen activator is effective in fibrin clearance in the absence of its receptor or tissue-type plasminogen activator. Proc Natl Acad Sci U S A. 1996; 93:5899–5904
21. Cathelin D, Placier S, Ploug M, et al. Administration of recombinant soluble urokinase receptor per se is not sufficient to induce podocyte alterations and proteinuria in mice. J Am Soc Nephrol. 2014; 25:1662–1668
22. Spinale JM, Mariani LH, Kapoor S, et al; Nephrotic Syndrome Study Network. A reassessment of soluble urokinase-type plasminogen activator receptor in glomerular disease. Kidney Int. 2015; 87:564–574
23. Abraham VC, Miller LN, Pratt SD, et al. Implementation of a human podocyte injury model of chronic kidney disease for profiling of renoprotective compounds. Eur J Pharmacol. 2017; 815:219–232
24. Alfano M, Cinque P, Giusti G, et al. Full-length soluble urokinase plasminogen activator receptor down-modulates nephrin expression in podocytes. Sci Rep. 2015; 5:13647