Nephrotic Plasma Alters Slit Diaphragm–Dependent Signaling and Translocates Nephrin, Podocin, and CD2 Associated Protein in Cultured Human Podocytes

Human Podocyte Cell Culture

Conditionally immortalized WT human podocyte cells were developed by the use of the temperature-sensitive large T antigen-SV40 transgene as described previously (8). These cells have been shown to differentiate fully by 14 d after switching from 33 to 37°C. Experiments to compare normal plasma with disease plasma were performed after day 14. Cell passages between 5 and 20 were used in all experiments. Podocytes were routinely cultured in RPMI-1640 medium with glutamine (R-8758; Sigma, St. Louis, MO) supplemented with 10% FCS (Life Technologies, Grand Island, NY) and insulin transferrin sodium selenite (Sigma I-1184; 1 ml/100 ml). FCS was substituted with human plasma at the same concentration (10%) for the experiments. A second human podocyte cell line, developed in the same way in our laboratory, derived from the normal pole of a kidney with Wilm’s tumor, was also used in some experiments to confirm further immunofluorescence (IF) findings. This was to exclude clonal variation as a potential explanation for differences seen.

Patient Samples

Plasma was obtained from patients who had undergone therapeutic plasma exchange for biopsy-proven FSGS (two boys aged 14 and 12 yr; one girl aged 11 yr) in relapse and the same patients in remission at the end of their plasma exchange course, together with a 14-yr-old girl who required plasma exchange for systemic lupus erythematosus (SLE) and a control patient with no renal disease but on prophylactic plasma exchange for cryoglobulinemia. Normal human plasma (and serum) was examined from an AB Rhesus-negative donor (Sigma) and from healthy adult volunteers. The profiles of the plasma samples are shown in Table 1.

Rescue Experiments

With rescue experiments, day 14 podocytes were initially exposed to 48 h of nephrotic or non-nephrotic plasma, and then the medium was changed with the initial plasma replaced at the same concentration (10%) and 10% of nephrotic or non-nephrotic plasma added for an additional 48 h.

Nephrin Mutant Podocyte Development

A kidney was removed for therapeutic reasons from a Finnish infant (male, 0.9 yr) who had congenital nephrotic syndrome secondary to the homozygous Fin_major mutation (nt121 del 2), which results in a two-nucleotide deletion in exon 2 of chromosome 19 and subsequently a truncation of the protein (1). Another nephrin mutant kidney (female, 1.8 yr) that was removed because of severe early-onset nephrotic syndrome as a result of a homozygous missense exon 11 mutation was also studied. Full informed consent from the family and ethical committee approval were obtained for use of the discarded kidneys. Primary culture, subcloning, and propagation were carried out as described previously (8). These cell lines were transfected with both SV40 large T antigen gene as described previously and a telomerase construct (9), to prevent senescence. This allows the cell line to replicate ad infinitum at 33°C, while the SV40 large T antigen was switched on, but at 37°C, the large T antigen is silenced and the cell line can differentiate. The cell lines when transformed were characterized and expressed WT1, synaptopodin, podocin, and CD2AP on IF and/or Western blotting. No nephrin was detectable on IF of the kidney sections and on IF or Western blotting on the podocyte cell line of the Fin major kidney (Figure 1). Cell passage between 7 and 15 were used for all experiments. Culture conditions were identical to the WT podocytes described above. All experiments on mutant cells were carried out in the Fin major cell line, and the second mutant cell line was used for additional data in the calcium flux experiments.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Characterization of the Fin major human podocyte cell line. Immunofluorescence (IF) staining and Western blotting of differentiated cells, either Fin major nephrin–deficient (−/−) or wild-type (WT) podocytes. (A) Nephrin IF using a polyclonal rabbit antibody K2966. Expression in WT podocyte but not nephrin knockouts. WT podocyte cultured exclusively in FCS giving cytoplasmic distribution of nephrin. (B) Nephrin Western blot. No expression in Fin major cells. Actin loading shown. (C) WT1 expressed in the nuclei of Fin major and WT podocytes. (D) Non-immunized rabbit Ig control. All above antibodies of rabbit origin.

Antibodies

Primary nephrin antibodies used included a rabbit polyclonal antibody (10) that recognizes the intracellular fragment of the molecule and a panel of both rabbit polyclonal and mouse monoclonal antibodies against the extracellular domain of nephrin as described previously (4). These all were gifts from Karl Tryggvason (Karolinska Institute, Stockholm, Sweden). A rabbit polyclonal podocin c-terminal antibody was used for both Western blotting and IF (gift from Corinne Antignac, Paris, France) (8). Other antibodies used included synaptopodin (Progen, Heidelberg, Germany) and polyclonal rabbit antibodies raised against CD2AP and WT1 (Santa Cruz Biotechnology, Santa Cruz, CA). Secondary species-specific horseradish peroxidase–associated antibodies (Sigma) and FITC- and TRITC-labeled antibodies (Jackson Immunoresearch, Philadelphia, PA) were also used. The F-actin cytoskeleton was imaged using a directly conjugated Texas red–Phalloidin stain (Molecular Probes, Eugene OR) and Western blotted with a monoclonal antibody (Sigma).

Immunofluorescence

Immunolabeling was done as described previously (4). Images were obtained by using a Leica photomicroscope attached to a Spot 2 slider digital camera (Diagnostic Instruments, Sterling Heights, MI) and processed with Adobe Photoshop 5.0 software. All images were analyzed by an investigator who was blinded to the identity of the samples and ranked from cortical to cytoplasmic distribution of the molecules on a scale of 0 to 3. At least 20 cells per coverslip were ranked. Each blinded experiment was carried out on three independent occasions. Images were taken at ×400 magnification.

Environmental Scanning Electron Microscopy

Environmental scanning electron microscopy was performed as described previously on these cells (8). The images were categorized by two blinded investigators and ranked according to surface blebbing and fine process formation on a scale from 0 to 4. For each condition, at least five images were obtained and ranked, with four to 15 cells seen per image. Representative images are shown.

Western Blotting and Protein Extraction

The exact techniques as described previously were used for protein extraction and Western blotting (8). Protein was quantified using a modified Bradford assay (11) to ensure equal loading of lanes in comparative studies and confirmed by actin loading.

Lipid Raft Fraction Preparation

For preparation of low-density Triton X-100–insoluble membrane domains, cells were treated for 48 h as before with either normal plasma or FSGS relapse plasma. The protocol was followed exactly as described previously (12). Fractions then were probed with anti-nephrin (48E11) or anti-actin antibody as above.

Intracellular Calcium Studies

Cells were grown on coverslips and then incubated with Fura 2-AM, excited, and assayed using ratiometric fluorescence measurement as described before (13). Undiluted samples of relapse and remission plasma from one patient with FSGS were perfused in random order on the same coverslip with a 20-min wash period of HBSS media (Life Technologies BRL) in between. HBSS was also used as a baseline control. A 10-min preincubation of 70 μM genistein (Sigma) was used to block tyrosine kinase. Control cells of immortalized proximal tubular cells (HK2) (14) and human vein embryonic endothelial cells (HUVEC) (ATCC, Manassas, VA) were treated in exactly the same manner as the podocytes, and their response to plasma was recorded. There was a replicate of at least four in all sets of experiments.

Nephrin, CD2AP, Podocin, and Actin

After 48 h of exposure, there were differences in the distribution of nephrin, podocin, and CD2AP in differentiated podocytes that were exposed to nephrotic compared with non-nephrotic patient plasma. Nephrotic plasma–exposed podocytes demonstrated a cytoplasmic distribution of all three molecules. Nephrin was observed in a diffuse cytoplasmic and filamentous pattern (average score, 0.4; Figure 2), podocin distribution was predominantly diffusely cytoplasmic (average score, 1.0), and CD2AP was seen in unevenly distributed “spots” in a submembranous and cytoplasmic location (average score, 0.2; Figure 3), as described previously (15). Plasma from all patients with nephrotic syndromes resulted in the same protein distribution, regardless of the underlying nephrotic cause. In contrast, samples from non-nephrotic individuals and healthy volunteers (data not shown) resulted in a highly enriched cortical, plasma membrane, location of nephrin (average score, 2.3), podocin (2.0), and CD2AP (2.6) in the differentiated podocyte (Figures 2 and 3). These enriched areas were not present continuously around the plasma membrane or indeed specifically at cell–cell junctions but seemed prominent at the leading edge of softly rounded cell protrusions, suggestive of a subtle alteration of cell shape to coincide with this distribution. A similar distribution was seen using samples from normal volunteers, patients with no previous renal disease, and patients who were in remission from established FSGS. Intriguing, human plasma (Figure 2) and serum (data not shown) also affected the location of nephrin when compared with FCS in podocytes of the same passage and age, with increased cortically located nephrin in the human compared with bovine samples. Prolonged culture in serum-free medium resulted in podocyte detachment, granulation, and necrosis, so these conditions were not included in the control experiments. Cellular morphology was disrupted in the presence of plasma from nephrotic patients, which occurred as early as 6 h, with retraction of fine process formation (average score, 1.2) and blebbing (average score, 2.9) on the cell surface noted (Figure 4) when viewed by environmental scanning electron microscopy. This appearance was also seen in the FCS-treated cells, although notably process retraction was not as widespread (average score, 2.3). In contrast, in remission and normal plasma–treated samples, there were many long processes (average score, 3.3 and 3.8, respectively) and very few surface blebs (1.9 and 1.4, respectively). Paired t test for either observation, between relapse/FCS and normal/remission, was significant at <0.01. In conjunction with this morphologic change, actin patterning at 48 h was also altered between nephrotic/FCS and non-nephrotic plasma. In cells that were treated with normal or remission plasma, actin filaments in the cytoplasm were weak or absent and cortical actin distribution was strong (average score, 2.1), whereas in cells that were treated with nephrotic plasma or FCS, cytoplasmic actin filaments were strongly expressed throughout, and cortical distribution was relatively weaker (average score, 1.0; paired t test P < 0.05; Figure 4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Nephrin immunolocalization in WT human podocytes incubated with various human plasma samples. Differentiated cells incubated with various plasma samples. Ten percent plasma applied to the cells for 48 h. IF with monoclonal 48E11 antibody. Peripheral nephrin shown by arrows. Representative of three independent, blinded experiments. (A) Normal human plasma. (B) Plasma from patient with focal segmental glomerulosclerosis (FSGS) in remission. (C) Plasma from FSGS patient in relapse. (D) Plasma from patient with systemic lupus erythematosus. (E) FCS alone. (F) Control of nephrin 48E11 antibody applied to proximal tubular HK2 cells. Negative staining.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Effects of nephrotic and non-nephrotic plasma on CD2 associated protein (CD2AP) and podocin in WT human podocytes. Differentiated cells incubated with different plasma samples. Ten percent plasma applied to the cells for 48 h. Representative panels from three independent experiments. Peripheral location of the molecules arrowed. (A) CD2AP from patient in remission from FSGS. (B) CD2AP from normal patient plasma. (C) CD2AP from nephrotic patient (FSGS relapse). (D) Podocin from patient in remission from FSGS. (E) Podocin from normal patient plasma. (F) Podocin from nephrotic patient (FSGS relapse).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Effects on the actin cytoskeleton by human plasma samples and FCS on WT human podocytes. Differentiated cells incubated with different plasma samples. Ten percent plasma applied to the cells for 6 h. (A through D) Environmental scanning electron micrographs of the podocytes, of same passage and age, exposed to the following. (A) Normal human plasma. Process formation illustrated. (B) FSGS patient in remission. Again, process formation demonstrated by arrow. (C) Nephrotic (FSGS relapse). Lack of foot processes and blebs noted. Insets show expanded images of affected areas. (D) FCS. Shows similar blebs to C. (E through H) The actin cytoskeleton by phalloidin staining at 48 h after exposure to the following plasma samples. (E) Normal human plasma. Predominantly cortical distribution. (F) FSGS patient in remission, similar to E. (G) FSGS patient in relapse. Cytoplasmic actin stress fibers more prominent. (H) FCS. Cytoplasmic actin stress fibers more prominent.

Quantitatively, Western blotting revealed that nephrin was downregulated by 48 h when podocytes were exposed to nephrotic plasma, but the other SD proteins (podocin and CD2AP) were not affected (Figure 5, A and B). In addition, we used a protocol that isolates lipid rafts on the basis of their insolubility in TX-100 and low buoyant density in sucrose density gradients (16). Nephrin was seen in both the heavy (cytoskeletal bound) fraction and raft fractions in normal plasma–treated cells, which correlates with its distribution in glomerular preparations (16), compared with just being present in the heavy fraction in FSGS relapse plasma–treated cells (Figure 5C). It is interesting that of the molecules studied, the only one that was downregulated early, at 24 h, in response to nephrotic plasma was synaptopodin, an actin associated podocyte protein, suggesting that the podocytes were becoming dedifferentiated (Figure 5A). This observation is consistent with studies of biopsy specimens showing downregulation of synaptopodin in affected podocytes in nephrotic diseases (17–19).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Western blot of podocyte proteins and their modulation by nephrotic plasma. Differentiated cells incubated with various plasma samples. Ten percent plasma applied to the cells for 24 or 48 h. Cells were lysed, and the lysates were probed with the following antibodies. (A) This shows representative blots of synaptopodin (at 24 h), and nephrin, podocin, CD2AP, and actin at 48 h (n = 4). (B) Densitometry on nephrin to actin ratio after 48 h of exposure. Nephrotic groups compared with non-nephrotic groups; four independent experiments. SEM shown. Data analyzed by a two-tailed nonparametric Wilcoxon test. (C) Sucrose density centrifugation of cells that were treated with either FSGS relapse plasma or normal human plasma. Fractions were probed with nephrin monoclonal Ab 48E11. Lighter fractions (1–4) correspond to those that normally contain lipid rafts (16). Control fractions were probed with actin to demonstrate efficiency of the gradient.

By analyzing podocytes from a patient with Fin major congenital nephrotic syndrome, we showed that under control (FCS) conditions, CD2AP was present in its subplasma membrane and cytoplasmic location as before (average score, 0.3), but after 48 h of incubation in normal plasma, there was no translocation to the cell surface (average score, 0.6; P < 0.01; Figure 6). In addition, the cortical concentration of actin seen in response to normal plasma with WT cells was less marked in the mutant cells, with no change to the cytoplasmic filamentous stress fibers (average score, 1.5 versus 1.9; P < 0.1). This demonstrated that CD2AP is present in nephrin-deficient cells but that nephrin needs to be present to allow the plasma membrane translocation of CD2AP and actin reorganization at the cell surface to occur.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Peripheral translocation of CD2AP is dependent on nephrin. WT podocytes and Fin major podocytes were initially cultured in FCS and then exposed for 48 h to normal human plasma. Cells then were immunostained as follows. For wild types: (A) The actin cytoskeleton shows cytoplasmic stress fibers when cultured in FCS, with weak cortical distribution. (B) CD2AP is diffusely distributed with intracytoplasmic spots in FCS culture. (C) On exposure to normal human plasma, the actin cytoskeleton assumes a thick cortical distribution, with almost complete loss of cytoplasmic filaments. (D) CD2AP translocates to the plasma membrane (high magnification view on inset panel). For the Fin major (−/−) podocytes: (E) The actin cytoskeleton is similar to WT podocytes with FCS. (F) CD2AP is diffuse and intracellular in distribution with FCS. (G) With normal human plasma, the actin cytoskeleton does not become cortically distributed but preserves the cytoplasmic stress fibers. (H) CD2AP remains intracellular, diffuse, and in spots. In areas where CD2AP seems to have reached the plasma membrane, the pattern is disrupted and remains submembranous (inset). Magnification, ×400.

Rescue of Nephrin Translocation

We examined whether the nephrotic plasma was responsible for the cytoplasmic relocation of nephrin directly via putative circulating factor(s) or was due to factors that were missing from nephrotic plasma and normally found in human plasma that were important. We found that nephrotic plasma induced a cytoplasmic distribution of nephrin, but this could be rescued and directed to the plasma membrane by subsequently adding an equal amount of non-nephrotic to nephrotic plasma (Figure 7). Importantly, this was not disease-specific, i.e., non-nephrotic plasma co-incubation could rescue FSGS and SLE nephropathic states, which would argue against a specific inhibitory factor present in normal plasma that may be antagonizing a “circulating factor” in FSGS (20).

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Peripheral nephrin and synaptopodin rescue of nephrotic plasma exposed cells by non-nephrotic plasma in WT human podocytes. IF performed using the mouse monoclonal antibody 48E11 (top) or synaptopodin (bottom). Each experiment shows day 14 thermoswitched podocytes exposed to 48 h of initial plasma then fresh co-incubation of original plasma sample with another plasma sample. All samples applied at 10% concentration. The same passage number and culture time of podocyte were used in each experiment. Representative experiment shown. (A) Normal plasma (10%) followed by normal plasma–normal plasma (20%). Peripheral nephrin localization arrowed. (B) Nephrotic plasma for 48 h (10%) followed by nephrotic–nephrotic plasma (20%). Cytoplasmic nephrin distribution. (C) FCS 10% followed by FCS (20%). Cytoplasmic nephrin distribution. (D) Nephrotic plasma (10%) for 48 h followed by nephrotic plasma (10%) and normal plasma (10%) for 48 h. Nephrin again peripherally located (synaptodin up-regulated with normal plasma, Panel A and D). Magnification, ×400.

Intracellular Calcium Signaling

We studied the effect of plasma from a nephrotic patient in relapse and in remission on both WT and on two different nephrin mutant podocyte cell lines. We found in WT cells, with remission plasma, that there was no [Ca²⁺] response (peak/baseline ratio = 1), whereas relapse plasma caused a statistically significant increase in peak [Ca²⁺], which was inhibited by the tyrosine kinase inhibitor genistein. The effect on nephrin mutant cells was revealing: With remission plasma, there was a considerable increase in peak [Ca²⁺] in both mutant cell lines, compared with WT cells, suggesting that the presence of nephrin in the correct conformation and cellular location prevents an increase in [Ca²⁺] brought about by this plasma. With relapse plasma, there was a constant peak [Ca²⁺] response that was consistent between podocyte cell types (Figure 8, A and B). It is interesting that genistein had no effect on the nephrin mutant cells’ response to remission plasma (Figure 8C). This shows that the effect of remission plasma on intracellular calcium is independent of tyrosine kinase activation in the absence of nephrin expression. This therefore supports the suggestion that nephrin plays a role in the calcium response in WT cells and that the response seen in nephrin knockout cells is a result of perturbation of this normal pathway. The effect of plasma seemed to be podocyte specific in that neither relapse nor remission plasma caused any increase in [Ca²⁺] in either proximal tubular or human endothelial cells (Figure 8D).

• Download figure
• Open in new tab
• Download powerpoint

Figure 8.

Podocyte calcium flux data. Cells were grown on coverslips and then incubated with Fura 2-AM, excited, and assayed using ratiometric fluorescence measurement. Undiluted samples of relapse and remission plasma from one patient with FSGS were perfused in random order on the same coverslip with a 20-min wash period of HBSS medium in between. HBSS was also used as a baseline control. Data expressed as ratio of peak/baseline of [Ca²⁺]_Intensity. (A) Relapse plasma significantly increases [Ca²⁺]_i levels in differentiated WT podocytes (1.67 ± 0.19; paired t test P < 0.001; n = 9), exon 11 nephrin mutant podocytes (relapse, 1.5 ± 0.15; paired t test P < 0.001; n = 7) and Fin major podocytes (relapse, 1.6 ± 0.12; paired t test P < 0.001; n = 7) compared with baseline. Remission plasma does not significantly increase [Ca²⁺]_I levels in WT podocytes (0.9 ± 0.146; t test P < 0.001; n = 9); however, it does in both exon 11 mutant podocytes (remission, 2.6 ± 0.3; paired t test P < 0.001; n = 7) and Fin major podocytes (remission, 2.2 ± 0.37; paired t test P < 0.005; n = 6) compared with baseline. (B) Preincubation of WT podocytes with 70 mM genistein attenuates the [Ca²⁺]_i response stimulated by relapse plasma (relapse, 121.5 ± 34.7 nM [Ca²⁺] versus 54.8 ± 4.9 [Ca²⁺]; ANOVA Bonferroni multiple comparison test, P < 0.05, n = 5; genistein/relapse, 59.8 ± 12.9 nM [Ca²⁺] versus 44.3 ± 7.9 nM [Ca²⁺], NS). *P < 0.05. (C) Preincubation of Fin major podocytes with 70 mM genistein shows no significant abrogation of the increased [Ca²⁺] response to remission plasma (seen in A; n = 4; paired t test P < 0.004, remission versus remission/genistein). (D) Effect of the plasma is podocyte specific. The data are expressed as a ratio of peak/baseline [Ca²⁺]. In HK2 cells (a proximal tubular epithelial cell line) and human vein embryonic endothelial cells (HUVEC), neither remission nor relapse plasma has a significant effect on [Ca²⁺] flux (HK2 cell: remission 0.77 ± 0.14, relapse: 0.96 ± 0.11; HUVEC: remission 1 ± 0.11, relapse 0.8 ± 0.1 versus baseline; ANOVA Bonferroni multiple comparison test; n = 7).