Full-length soluble urokinase plasminogen activator receptor down-modulates nephrin expression in podocytes

Increased plasma level of soluble urokinase-type plasminogen activator receptor (suPAR) was associated recently with focal segmental glomerulosclerosis (FSGS). In addition, different clinical studies observed increased concentration of suPAR in various glomerular diseases and in other human pathologies with nephrotic syndromes such as HIV and Hantavirus infection, diabetes and cardiovascular disorders. Here, we show that suPAR induces nephrin down-modulation in human podocytes. This phenomenon is mediated only by full-length suPAR, is time-and dose-dependent and is associated with the suppression of Wilms’ tumor 1 (WT-1) transcription factor expression. Moreover, an antagonist of αvβ3 integrin RGDfv blocked suPAR-induced suppression of nephrin. These in vitro data were confirmed in an in vivo uPAR knock out Plaur−/− mice model by demonstrating that the infusion of suPAR inhibits expression of nephrin and WT-1 in podocytes and induces proteinuria. This study unveiled that interaction of full-length suPAR with αvβ3 integrin expressed on podocytes results in down-modulation of nephrin that may affect kidney functionality in different human pathologies characterized by increased concentration of suPAR.


Results
SuPAR induces down-modulation of nephrin in human primary podocytes. Various, genetic and functional, nephrin defects lead to nephritic syndromes and proteinuria [22][23][24][25][26][27] . To verify the effect of suPAR stimulation on nephrin expression, we used human primary podocytes obtained from renal tissue of patients affected by renal adenocarcinoma who underwent a radical nephrectomy. All tissue specimens used for this study were collected from the distal part of the pathological tissue and were free from any disease as was verified by haematoxylin-eosin-staining (Fig. 1a). We then cultured freshly isolated glomerula in order to obtain human podocytes (Fig. 1a), whose phenotype was verified by qPCR analysis detecting gene expression of the following specific podocytes markers: Wilms' tumor 1 (WT-1), synaptopodin, nephrin and podocin (data not shown). Mean serum and plasma concentration of suPAR in healthy adults have been reported to be 2 ng/mL 18 , whereas it reaches 20 ng/mL in patients with various diseases, such as cancer, sepsis, liver disease, HIV infection and FSGS 8,14,[18][19][20][21] . We then proceeded to incubate human primary podocytes with 20 ng/mL of human recombinant fl-suPAR. After 24 hours of stimulation we detected by fluorescence microscopy a significant down-modulation of nephrin at the protein level (Fig. 1b). Immunoflourescence staining of frozen tissue from human normal kidney served to confirm the specificity of anti-nephrin rabbit clonal antibody (Ab) used for this study (Fig. 1c). The decreased amount of nephrin following incubation of primary podocytes with suPAR was also confirmed by qPCR, thus suggesting a transcriptional control of nephrin expression by suPAR (Fig. 1d). To both confirm the modulation of nephrin exerted by suPAR and to disclose the related mechanism in a post-mitotic podocyte, we repeated the same experimental approach in conditional immortalized human podocytes (CIHPs) in vitro 34 . By performing assessments of protein expression with immune-flourescence in CIHPs incubated with different concentrations of suPAR for 24 hours, we detected a significant reduction of nephrin flourescence intensity and thus a lower level of nephrin protein in suPAR stimulated CIHPs comparing to control experiments, with the maximum inhibition between 10-20 ng/mL of suPAR (Fig. 2a). To determine whether suPAR stimulation affects nephrin expression at the transcriptional level, we performed qPCR experiments. As shown in Fig. 2b, quantification of nephrin mRNA in CIHPs incubated with suPAR showed a rapid and progressive decrease of Nephrin gene expression (Fig. 2a-c). The reduction of nephrin transcripts was already detectable after 3 hours and became significant after 6 hours of treatment, when it also reached a plateau that was maintained even after 24 hours of stimulation. However, we did not observe any significant suPAR-mediated down-modulation of the expression of synaptopodin, another specific podocyte marker, thus indicating the specificity of suPAR in suppressing nephrin expression in CIHPs (Fig. 2c).
Full-length suPAR, but not the truncated variant, down-regulates nephrin expression in human podocytes. Structurally, uPAR is a GPI-anchored membrane glycoprotein consisting of three homologous domains (D I D II D III ) 35 . Cleavage of uPAR from the cell membrane is catalyzed by various enzymes and can occur both at the GPI-anchor and the linker region between D I and D II . The suPAR resulting from cleavage may thus consist of domain D I D II D III , D II D III or D I . Since the increased level of plasma-associated suPAR in human disorders reflects the increased concentration of both D I D II D III (fl-suPAR) and D II D III (c-suPAR) variants, we evaluated the contribution of suPAR variants in down-modulation of nephrin in CIHPs 4,6 . To compare the dose-dependent effects of fl-suPAR and c-suPAR variants on nephrin inhibition, two suPAR variants were tested in the range of concentration between 0.4-0.02 nM that resulted with the previously observed maximum inhibitory effect of fl-suPAR (Fig. 2b). Interestingly, only fl-suPAR protein was able to significantly reduce nephrin expression at transcription level (Fig. 3a). In addition immunoflourescence analysis showed a statistically significant reduction of nephrin also at protein level only after stimulation with fl-suPAR and not with its short

SuPAR mediated down-regulation of nephrin in human podocytes occurs through interaction with αvβ3 integrin and is associated with suppression of the WT-1 transcription factor. It
has been demonstrated that suPAR binds and activates α vβ 3 integrin in human podocytes 8 . In order to understand whether the α vβ 3 integrin molecule is involved in suPAR dependent down-regulation of nephrin we used the α vβ 3 small molecule-inhibitor cycloRGDfv 8 . Pre-treatment of CIHPs constitutively expressing α vβ 3 integrin with cycloRGDfv prior stimulation with fl-suPAR resulted in a significant inhibition of suPAR-dependent downmodulation of nephrin at the transcription level (Fig. 4a). Moreover, qPCR analysis revealed that the effect of cycloRGDfv on nephrin expression on CIHPs incubated with suPAR is dose dependent, since incubation with 5 μ M and 10 μ M of this α vβ 3 inhibitor, respectively, partially or fully restored the amount of nephrin transcripts compared to control experiments. These findings suggest that suPAR is able to induce the down-modulation of nephrin in CIHPs via α vβ 3 interaction Different activating and suppressing transcription factors have been identified as being involved in the transcriptional regulation of Nephrin gene expression [36][37][38] . The most documented transcription factor involved in the regulation of Nphs1 gene expression is WT-1. Knockout, transgenic, and siRNA analyses have demonstrated the importance of WT-1 at several stages of kidney development [39][40][41] . Moreover, expression of WT-1 continues in podocytes of adult kidneys and is required for physiological levels of nephrin expression. Consistent with this observation, it was demonstrated that nephrin expression in podocytes is lower in the kidneys of mice with reduced expression of WT-1 37,42 . In addition, both in vitro and in vivo functional approaches have shown that WT-1 can bind and activate the nephrin promoter and that this binding is essential for nephrin-specific expression in vivo 37,43 . Since our results showed suPAR-dependent down-regulation of nephrin at transcription level, we assessed whether WT-1 is involved in this process. QPCR analysis of WT-1 transcription showed a statistically significant decrease of WT-1 expression in CIHPs after treatment with fl-suPAR (Fig. 4b), thus indicating that WT-1 is a possible target of an activated suPAR-α vβ 3 signaling down-stream pathway. In line with the kinetic observed for the down-modulation of nephrin (Fig. 4a), experiments performed with different concentrations of cycloRGDfv inhibitor revealed a full restoration of WT-1 transcripts after pre-treatment with 10 μ M of cycloRGDfv (Fig. 4b), while we still could observe a lower but significant inhibition of nephrin incubating CIHPs with 5 μ M of the α vβ 3 inhibitor. To assess the specific suPAR-dependent attenuating role of WT-1 transcription factor in nephrin gene expression we evaluated the binding of WT-1 in the promoter region of Nphs1 by chromatin immunoprecipitation assay (ChIP). We detected a significantly lower binding of WT-1 protein in the regulatory region of Nphs1 gene in fl-suPAR treated cells compared to the non-stimulated podocytes (Fig. 5a). Amplification of GAPDH promoter was used as a positive control in CIHPs by immunoprecipitation of chromatin with anti-RNA polymerase II antibody (Fig. 5b). GAPDH promoter is lacking any WT-1 site thus as a negative control we amplified GAPDH promoter after chromatin immunoprecipitation with IgG or anti-WT-1 antibody in Mock and suPAR treated samples, however, we did not observe any significant changes (Fig. 5b). Specificity of the anti-WT-1 antibody used in ChIP assay was tested in the Jurkat and in K562 cell lines, known to be respectively negative and positive cells for WT-1 expression (Fig. 5c-d) 44,45 . These data strongly indicate that the suPAR-dependent down regulation of nephrin might occur through decreased transcription levels of WT-1 factor resulting in the attenuated binding to Nphs1 gene promoter and thus lower transcription of nephrin. In addition, we analyzed the expression of the transcription regulator Snail that has recently been proposed as a repressor of Nephrin gene expression 38 . However, although we found detectable levels of Snail in CIHPs, we did not measure any increased amount of Snail after suPAR treatment (data not shown).

High-dose of suPAR in knock out Plaur −/− mice inhibits nephrin and WT-1 expression in podocytes and induces proteinuria.
We next examined whether exogenous circulating full-length suPAR could cause nephrin down-modulation in uPAR-knockout (Plaur −/− ) mice in which we injected i.v. 20 μ g (1 mg/Kg) of murine full-length recombinant mouse suPAR. After 24 hours, we observed an increased level of proteinuria and higher amount of suPAR deposit in the glomeruli of Plaur −/− mice infused with full-length recombinant murine suPAR compared to control experiments (Fig. 6a) 8 . Moreover, we performed experiments using confocal microscopy in the same experimental setting and we observed a significant down-modulation of nephrin expression with no changes in synaptopodin levels in Plaur −/− mice infused with full-length recombinant murine suPAR compared to control experiments (Fig. 6b). Finally, we confirmed our results obtained in vitro in the in vivo model by showing that the decreased expression of nephrin is associated with the down-modulation of WT-1 in suPAR treated Plaur −/− mice  (Fig. 6c). These data demonstrate that suPAR is able to activate a specific repressor-signaling pathway that leads to suppression of WT-1 and Nephrin genes.

Discussion
Recently scientific opinion hailed the discovery of suPAR as a possible pathogenic factor as well as a mere biomarker of FSGS 8 . In addition different clinical studies observed increased suPAR concentration in various glomerular diseases thus implying on one hand the non specific role of suPAR in FSGS and on the other hand its emerging active pathological role in different glomerular and proteinuric unrelated to FSGS, disorders [8][9][10][11][12][13] . Indeed, in all studied renal disorders, increased suPAR was inversely associated to estimate glomerular filtration rate (eGFR) and in some reports to proteinuria. Nephritic syndromes represent characteristic clinical features also of other human diseases such as HIV and Hantavirus infection, diabetes and cardiovascular disorders that have been associated with increased level of suPAR 14,[46][47][48][49][50][51][52] . Experimental studies both in vitro and in vivo clearly demonstrated the effect of suPAR on α vβ 3 integrin activation in podocytes 8,33 . In addition, studies using the Plaur −/− mice model confirmed the ability of high dose of suPAR to induce proteinuria 8 . Our study demonstrated that full length suPAR induced selective down-modulation of nephrin expression in human podocytes via interaction with α Vβ 3 integrin. This negative regulation of nephrin was observed both at the protein and the transcriptional levels, and was associated with a reduced level of the transcription factor WT-1. Furthermore, in the in vivo suPAR knock out Plaur −/− mice model, the infusion of a high dose of suPAR correlates with lower expressions of nephrin and WT-1 in podocytes and glomerular permeability. Controversial results were obtained in wild type mice infused with high dose of suPAR 53 . These observations suggest that different molecular mechanism(s) may be involved in the detrimental action of suPAR in kidney physiopathology and various factors may control, inhibit or emphasize the toxic action of suPAR in pathological conditions. In this context, expression of α vβ 3 integrin, which is expressed at low levels in podocytes under physiological conditions and increases in some pathologies, such as diabetic nephropathy, could play an important role 54,55 . Activation of α vβ 3 integrin in podocytes can be also inhibited by other integrins such as α 3β 1 that represents the principal integrin expressed in podocytes and interacts with glomerular basement membrane 30 .On the other hand, an α vβ 3 genetic polymorphism or other circulating factors such as TNF-α may affect suPAR activity 56,57 . Finally, the heterogeneity of circulating suPAR isoforms might explain why this biomarker, although being elevated in a variety of diseases, lacks disease-specificity and shows heterogeneous pathogenic action 2 . Besides the full length and cleaved form of suPAR various glycosylated variants of suPAR among different cell types have been reported 2 .Here, we show that only the full-length suPAR, but not c-suPAR, induces the down-modulation of nephrin, providing a conceptual framework for its pathogenetic action on podocytes in different human pathologies characterized by elevated suPAR and opening new therapeutic perspectives in the field. As an example, it might be interesting to evaluate the possible pathogenic role of suPAR in HIV infection since one of the clinical manifestation of kidney disorders in HIV pathogenesis is FSGS and increased levels of plasma suPAR in HIV patients have been correlated with disease progression 14,58,59 . Lymphoid organs of HIV infected individuals showed as an important site of production and release of suPAR and in particular full-length suPAR was found to be increased and contributes to prevent the anti-HIV activity of uPA 6,14,60,61 . In addition, we observed that plasma from HIV-infected individuals with increased levels of plasma suPAR have potential to induce downmodulation of nephrin (unpublished observation) and thus implicate suPAR as a possible renal risk factor in HIV pathogenesis.

Methods
Isolation and culture of human podocytes. Human renal tissue was obtained at the Department of Urology, Istituto Clinico Humanitas (Milan, Italy) from patients that underwent to both laparoscopic or open radical nephrectomy due to the renal cell carcinoma. All patients participated in this study provided written informed consent. All experimental protocols were approved by IRB (Authorization nr. 794/2011 Ethic Committee, Humanitas Clinical and Research Center, Milan). All methods were carried out in accordance with the approved guidelines. All tissues were collected from the distal area of the pathological tissue and macroscopically free from any disease as verified by haematoxylin-eosin-staining. Under aseptic conditions kidneys were minced into small pieces and then pressed through a series of stainless steel sieves (sieving method) with decreasing pore size of 200-μ m, 100-μ m and 75-μ m. As a final step the glomeruli were collected on 75-μ m sieve, washed twice and cultured in collagen IV (Sigma-Aldrich), coated plates in F-12 medium (Sigma-Aldrich) with 10% FBS, 2 mM ultra-glutamine, 100 U/mL penicillin, 100 μ g/mL streptomycin, nonessential aminoacids (all purchased from Lonza Verviers Sprl) and supplemented with insulin-transferrin-sodium-selenite media supplement (100X, Sigma-Aldrich) and 0.35 ng/mL hydrocortisone (Sigma-Aldrich). After 3-4 days non attached glomeruli were washed and cultured for another 5-10 days. Conditionally immortalized human podocytes (CIHPs, kind gift from Dr. M. A. Saleem) were developed from primary human podocytes by transfection with the temperature-sensitive SV40-T and cultured as described in Saleem M.A. et al. 34 . Reagents. Human full length suPAR (composed of the three domains DI, DII and DIII) and truncated suPAR (c-suPAR, composed of the two domains DII and DIII) were purified as previously described 62

QPCR.
Total RNA was extracted using RNeasy mini columns (Qiagen, Valencia, CA), following manufacturer's instructions, and its concentration was determined by spectrophotometry. One μ g of total RNA was used to generate cDNA templates for RT-PCR, using random primers, RNase inhibitor and High-Capacity cDNA Reverse Transcription Kit from Applied Biosystem (Foster City, CA). All mouse and human gene expression were analyzed by the Taqman ® mRNA specific assays for: nephrin, podocin, synaptopodin, WT-1, Snail and GAPDH (Applied Biosystem).
Immunofluorescence. Human kidney tissue embedded and frozen in OCT or cells grown on coverslips coated with human collagen IV were fixed with 4% paraformaldehyde (PFA) then washed, and immunolabeled over night at 4 °C with rabbit polyclonal anti-nephrin (clone Y17-R, Acris Antibodies, San Diego, CA, USA) or rabbit polyclonal anti-α vβ 3-integrin (clone 23C6, Santa Cruz) Ab. The bound antibody was stained with FITC/Alexa Fluor 594-conjugated goat anti-rabbit Ab respectively. Mouse kidney tissue embedded and frozen in OCT were fixed in acetone and stained for anti-nephrin (clone Y17-R) and subsequently with anti-synaptopodin (clone G1D4) followed by staining with FITC/Alexa Fluor 594-conjugated goat anti-rabbit and anti-mouse Ab respectively. SuPAR deposits in glomeruli were detected in frozen kidney tissue, fixed with 4% PFA and stained with antibody against murine uPAR (si420), kindly provided by Dr. Nicolai Sidenius (IFOM-IEO, Milan, Italy). Nuclei were stained with DAPI.
suPAR knock out Plaur −/− mice model. UPAR knock-out (Plaur −/− ) mice 63 were kindly provided by Dr. Nicolai Sidenius (IFOM-IEO, Milan, Italy) and maintained on C57BL6/N genetic background under specific pathogen-free conditions. Eight to ten week-old Plaur −/− mice were intravenously injected with 20 μ g (1 mg/Kg) of murine recombinant full length suPAR (R&D system). Twenty-four hours after injections, urine were collected and analyzed for creatinine and total protein content. Animals were then sacrificed and kidneys were collected and stored in OCT for immune fluorescence analyses and RNA extraction. Animal experiments adhered to the requirements of the Commission Directive 86/609/EEC and to the Italian legislation (Decreto Legislativo 116; 27 January 1992). All experimental protocols were approved by the Animal Care and Use Committee (Authorization nr. 192/2012-B, Humanitas Clinical and Research Center, Milan, Italy). All methods were carried out in accordance with the approved guidelines.
Chromatin Immunoprecipitation. Chromatin immunoprecipitation (ChIP) assays were performed with the use of EZ-Magna ChIP Chromatin Immunoprecipitation Kit (Millipore) following manufacturer's instructions. CIHPs were growth at 80-90% of confluency. Jurkat and K562 cell lines (ATCC) were cultured in RPMI 1640 medium supplemented with 10% FBS, 2 mM ultra-glutamine, 100 U/mL penicillin, 100 μ g/mL streptomycin at concentration of 2 × 10 6 cells/mL. Protein-DNA complexes were cross linked with 1% formaldehyde followed by glycine 0.125 M treatment, then cells were harvested and nuclear extraction was performed. Nuclei were collected by centrifugation at 12000 g and were suspended in sonication buffer with protease inhibitor cocktail and sheared to an average length of 750 bp by using Bioruptor Plus UCD-300 on high power for 36 cycles (30'' ON and 30'' OFF). Aliquots of cross-linked chromatin (50 uL) were diluted with 450 uL of ChIP dilution buffer and incubated overnight at 4 °C with 20 uL protein A/G magnetic beads and 5.0 μ g/mL of rabbit polyclonal ChIP grade anti-WT-1 antibody (clone-C19, Santa Cruz Biotechnology). Mouse monoclonal anti-RNA polymerase II (CTD4H8) as the positive control was used and mouse/rabbit normal IgG were used as negative controls. 1% of non immunoprecipitated chromatin was saved as input sample. Cross-links between proteins and DNA were reversed by addition of ChIP elution buffer with proteinase K and incubation at 65 °C. DNA was purified using spin columns. Quantitative amplification of precipitated DNA fragments was performed on a 7900HT Fast Real-Time System (Applied Biosystem) using SYBR Green assay. The following primer pairs were used Nphs1 promoter: 5′ CGCCCAGTCTCTTTATCTTTC-3′ , 5′ -GACAAGGAGCAGGAGTGAG-3′ ; GAPDH promoter: 5′ -TACTAGCGGTTTTACGGGCG-3′ , 5′ -CGAACAGGAGGAGCAGAGAGCGA-3′ . The specificity of anti-WT-1 antibody used for ChIP assay was tested in Western blotting assay. All Inputs and chromatin immunoprecipitated samples with IgG or with anti-WT-1 rabbit polyclonal antibody (clone-C19, Santa Cruz Biotechnology) were separated by SDS-PAGE and transferred to nitrocellulose membrane. The membrane was blocked in 5% milk, incubated overnight at 4 °C with the goat polyclonal anti-WT-1 primary antibodies (Abcam, ab96792) or goat polyclonal anti-β -actin (clone C-11, Santa-Cruz Biotechnology, Santa Cruz, CA, USA), washed and incubated with secondary Ab -conjugated with horseradish peroxidase. Western blot analysis was conducted according to standard procedures using Immun-Star TM WesternC TM chemiluminescence detection substrate kit (Bio-Rad, Hercules, CA, USA).