“The FSGS Factor”

Chemicals and Supplies

Bovine serum albumin (35% solution), HPLC grade solvents, analytical grade chemicals for buffers and other reagents, and phenylmethylsulfonyl fluoride (PMSF) were obtained from Sigma Chemical Co. (St. Louis, MO). Spectra/Por dialysis tubing was obtained from Spectrum Medical Industries (Laguna Hills, CA). Centricon tubes with 50- and 30-kD molecular weight cutoff (MWCO) ultrafiltration membranes, and stirred cell concentrator and membrane (3.5-kD MWCO) were obtained from Amicon (Beverly, MA). Bovine serum albumin (BSA) and Coomassie Blue Reagent for protein assay were obtained from Bio-Rad (Hercules, CA). Deionized water was purified using NANOpure ultrapure water system from Barnstead (Dubuque, Iowa). Immobilized Protease Sg (Staphylococcus aureus) bound to cross-linked 6% beaded agarose was obtained from Pierce Chemical Co. (Rockford, IL). Precast 4 to 20% gradient minigels (Novell Experimental Technology, San Diego, CA) were used for electrophoresis on equipment from Hoefer Scientific Instruments (San Francisco, CA).

Experimental Animals

Normal male Sprague Dawley rats (120 to 150 g body wt) were maintained on Purina chow with free access to drinking water. Glomeruli were isolated from kidneys removed from rats under Metofane (methoxyflurane; Mallinckrodt Veterinary, Mundelein, IL) anesthesia. Intravenous injection of the 70% supernatant was carried out in rats anesthetized with intraperitoneal Brevital (methohexital sodium, 50 mg/kg; Eli Lilly and Co., Indianapolis, IN).

FSGS Patients and Collection of Plasmapheresis Fluid

Details of the criteria used for the diagnosis of FSGS have been described earlier (13,19). Briefly, diagnosis of FSGS was based on renal biopsies done to evaluate proteinuria or renal insufficiency and on the presence of segmental obliteration of capillaries by increased extracellular matrix in some glomeruli. Mesangial proliferation, segmental deposits of IgM, and/or complement in capillary walls was seen in some biopsy specimens.

Plasma specimens were obtained during the course of therapeutic plasmapheresis from a total of 15 patients. The decision to perform plasmapheresis was made by the primary physician in each case. Eleven patients (nine adults and two children) were being treated for recurrence of FSGS in renal allografts. Of these, seven were male and four were female. An additional three patients were treated with plasmapheresis because of unremitting nephrotic syndrome in their native kidneys. Of these, two were adult males and one was a female child. One plasma sample was obtained from a patient who experienced a fulminant recurrence of FSGS and allograft loss in 1985 and who is treated with three-times-weekly hemodialysis. This patient's serum, plasma, and plasmapheresis fluid have consistently shown high levels of FS activity in the in vitro assay. Plasmapheresis was done using citrate anticoagulation and replacement of removed plasma with 0.9% NaCl and 5% human albumin solutions, and the discarded plasma was shipped to us.

Normal Plasma

Normal plasma was obtained from healthy volunteers or as discarded material from the Blood Center of Southeastern Wisconsin (Milwaukee, WI).

Transport and Storage of Specimens

Plasma was immediately stored at -20°C under sterile conditions or transported to our laboratory on dry ice and stored at -20°C until used. Samples stored at -20°C for up to 3 yr with or without sodium azide and protease inhibitors have been found to retain the FS activity.

Extraction of Total Lipids

In separate experiments, we investigated the possibility that the FSGS factor was a lipid, using lipid extraction from cryoprecipitate-free plasma or after precipitation of chylomicrons and lipoproteins. An aliquot of cryoprecipitate-free plasma from every specimen was extracted twice with 10 vol of a cold chloroform/methanol (2:1) mixture, filtered, and separated into two phases as described by Folch et al. (22). Final upper (aqueous) and lower (chloroform) layers were separated, washed with theoretical lower and upper phases, respectively, and evaporated to dryness under nitrogen. Total lipid residue in the lower (chloroform) phase was reconstituted in the medium containing 5% BSA and used for testing the FS activity (described below). To determine the amount of total lipid removed by dextran sulfate treatment, plasmapheresis fluid was subjected to lipid extraction by chloroform/methanol before and after treatment with dextran sulfate (described below), and the difference in total lipid content was determined.

Enrichment of FSGS Factor

The FSGS factor from patients was concentrated by sequential removal of (1) nonactive cryoprecipitate; (2) lipid-rich lipoproteins and chylomicrons; (3) Ig-rich protein fraction; and (4) albumin-rich protein fraction as described below.

Removal of the Cryoprecipitate. Frozen plasma was allowed to thaw at 4°C, and sodium azide (0.05%) and PMSF (1 mM) were added immediately followed by centrifugation of the plasma at 4°C (5000 × g for 30 min) to remove the cryoprecipitate formed during storage.

Removal of Lipoproteins and Chylomicrons. To remove protein-bound lipids in an aqueous medium, we precipitated lipoproteins and chylomicrons (23,24). An equal volume of 0.1% sodium chloride was added to the cryoprecipitate-free plasma specimen. After mixing, 20 μl/ml of 10% dextran sulfate (molecular weight, 0.5 × 10⁶) was added followed by 100 μl/ml of 1 M calcium chloride. The precipitate was removed by centrifugation at 4°C, and the supernatant was dialyzed against Tris-buffered saline (TBS; 1% sodium chloride in 10 mM Tris-HCl, pH 7.8). The precipitate could not be redissolved for assay of FS activity. Total lipids, protein, and FS activity were determined in the supernatant as described.

Removal of Ig-Rich Protein Fraction by Ammonium Sulfate. The plasma specimen, free of the majority of lipoproteins and chylomicrons and the lipids associated with them, was subjected to sequential ammonium sulfate precipitation. First, a 50% saturation of ammonium sulfate was obtained by gradual addition of dry ammonium sulfate with continuous stirring at 4°C. The precipitate was separated by centrifugation at 10,000 × g for 15 min (4°C). The precipitate obtained at 50% ammonium sulfate saturation could not be redissolved completely and was discarded.

Removal of Albumin-Rich Protein Fraction by Ammonium Sulfate. Ammonium sulfate concentration of the supernatant was then increased to 70% saturation, and the resulting precipitate was removed by centrifugation at 20,000 × g for 15 min at 4°C. The precipitate at 70% ammonium sulfate saturation was readily soluble in TBS. Residual protein in the supernatant at 70% ammonium sulfate saturation (70% supernatant) could be precipitated by increasing ammonium sulfate to 80% saturation (70 to 80% precipitate).

Characteristics of the FSGS Factor

The 70% supernatant fraction of the plasma contained very little protein. However, this fraction retained the FS activity and was used to study some of its characteristics, including (1) effect of heat; (2) effect of incubation with a nonspecific proteolytic enzyme; and (3) size.

Heat Inactivation. The 70% supernatant (5 mg protein/ml) was incubated in boiling water for 10 to 40 min, cooled, centrifuged, and the supernatant was tested for FS activity.

Protease Digestion. The 70% supernatant fraction (5 mg protein/ml) was mixed with 1 ml slurry of Protease-Sg immobilized on agarose (2.4 benzoyl arginine ethyl ester U/ml gel; Pierce Chemical Co.) in TBS containing 5 mM calcium chloride. The mixture was incubated overnight at room temperature with gentle shaking. The slurry was washed with TBS and centrifuged. The supernatant was dialyzed against TBS, concentrated under pressure (20 to 25 psi) on an ultrafiltration membrane (3.5 kD MWCO) using an Amicon concentrator (Amicon, Beverly, MA), and tested for FS activity.

Size Determination. Dialysis. Routine dialysis was carried out using 3.5-kD MWCO dialysis membranes with at least three changes of 10 vol of TBS. To determine the molecular size of the active component, the 70% supernatant fraction was dialyzed using 50-kD or 12- to 14-kD MWCO dialysis membranes with three to four changes of 10 vol of TBS. The retentate in the dialysis tube was used to test the FS activity in vitro.

Centrifugation-Based Membrane Ultrafiltration. Centrifuge tubes fitted with 50- or 30-kD MWCO ultrafiltration membranes (Amicon, Beverly, MA) were used for determining the size of the FSGS factor in the 70% supernatant. The 50-kD MWCO membrane was washed thoroughly with TBS, and a 5-ml aliquot of the sample was added to the reservoir held in the filtrate cup. The device was centrifuged on a fixed angle rotor at 5000 × g for 15 min. An aliquot of the filtrate was saved, and the rest of the filtrate was transferred to a thoroughly washed 30-kD MWCO membrane reservoir and spun at 5000 × g for 30 min. The retentate and the filtrate were dialyzed in TBS and used to test the FS activity.

Electrophoresis

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed using 4 to 20% gradient gels in Tris-glycine buffer to detect the protein profile of the fractions obtained during purification. Gels were stained for proteins using Coomassie blue or silver staining method. Electrophoretic pattern of the 70% supernatants from normal and FSGS plasma were also compared under nondenaturing and nonreducing conditions (native conditions).

Protein Determination

Total protein concentration was determined by the Bradford method (25). The Coomassie Blue Reagent (Bio-Rad, Hercules, CA) was added to protein solutions, and the intensity of the color was measured spectrophotometrically at 595 nm.

Measurement of FS Activity

Fractions obtained during purification were tested for their capacity to increase glomerular albumin permeability (P_alb) by the method established in our laboratory. Details of the rationale and methodology for using isolated glomeruli, measurement of reflection coefficient, and albumin permeability have been described previously (18,26). Briefly, glomeruli were isolated by sequential sieving in Krebs-Ringer's buffer containing 5 g/dl BSA. Control plasma, FSGS plasma, or its fractions (test samples) were dialyzed (3 × 10 vol) and diluted in TBS to obtain a protein concentration of 4 to 5 mg/ml. For routine bioassay of the activity, 20-μl aliquots of test samples containing 80 to 100 μg of protein were incubated with glomeruli in a final volume of 1 ml at 37°C for 10 min, and the initial image was recorded by videomicroscopy. The medium was changed from 5 g/dl to 1 g/dl BSA to produce an oncotic gradient across the glomerular capillary wall causing movement of fluid into the capillaries and an increase in glomerular size. A second video image was recorded 2 to 3 min after adding 1% BSA. The average diameter of each video image was measured, and the relative volume increase of each glomerulus was calculated using change in volume (ΔV) due to oncotic gradient. Albumin reflection coefficient was calculated as: σ_alb = ΔV_experimental/ΔV_control. Convectional permeability (P_alb) was calculated as: P_alb = 1 - σ_alb. When σ_alb is zero, albumin moves across the membrane at same rate as water and P_alb equals 1.0. When σ_alb is one, albumin cannot cross the membrane with water and P_alb equals zero. A P_alb value of 0.5 or higher is accepted as positive.

Incubation of glomeruli with medium containing pooled normal serum (1:50 dilution) served as a negative control, and incubation with known active FSGS serum (1:50 dilution) served as a positive control. To determine the lowest concentration of protein required to cause an increase in permeability (P_alb ≥ 0.5), test samples were serially diluted with TBS and activity was measured in duplicate experiments.

Injection of 70% Supernatant, Collection of Pre- and Postinjection Urine, and Measurement of Urinary Protein and Creatinine

Seventy percent supernatant from plasma of four patients with recurrent FSGS in renal allografts or from normal pooled plasma were used for intravenous injection into rats. Urine samples were collected on sodium azide (50 μl of 5%) to avoid bacterial contamination. Rats were housed in metabolic cages overnight (about 18 h) before injection to collect urine for preinjection urinary protein determination. The 70% supernatant from normal or FSGS plasma was injected intravenously in anesthetized rats. Rats were allowed to recover from anesthesia and transferred to metabolic cages. Preliminary data showed no difference in urine protein up to 6 h after injection of FSGS plasma preparation, and this duration allowed the animals to recover from anesthesia. Urine samples were collected for 18 h, after the 6-h recovery period; volumes were recorded, centrifuged at 4°C, and supernatants were stored at -20°C.

Creatinine in urine samples was measured using a creatinine/PAP kit (Boehringer Mannheim, Indianapolis, IN). This assay is based on an enzymatic conversion of creatinine to creatine to sarcosine. The latter is oxidized to generate hydrogen peroxide for use in producing a red-colored benzoquinone-imine dye, which is measured spectro-photometrically at 510 nm. Urine protein was measured by the Coomassie Blue dye reagent as described above.

Statistical Analyses

Results are expressed as mean ± SD in Tables 1 and 2. P_alb values (mean ± SEM) for FSGS plasma fractions were compared with normal plasma using t test, and P < 0.01 was accepted as significant (see Figure 2, A and B). Values for urinary proteins and creatinine are expressed as mean ± SEM. n represents the number of rats in each group. Values among various groups were compared using one-way ANOVA, and significance was defined as P < 0.01 (see Figure 5).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

(A) Albumin permeability (P_alb) of isolated rat glomeruli after incubation with fractions of 70% FSGS supernatant obtained by dialysis. Size determination by dialysis through membranes of different molecular weight cutoff (MWCO) was carried out as described. NPP, normal pooled plasma. Values are mean ± SEM (n = 10). ^*P < 0.01 compared with NPP. 70% FSGS is the supernatant at 70% ammonium sulfate saturation obtained from FSGS plasma. The 12- to 14-kD Retentate and 50-kD Retentate are the residual fractions retained after dialysis in 12- to 14-kD or 50-kD MWCO tube. (B) P_alb of isolated rat glomeruli after incubation with fractions of 70% FSGS supernatant obtained by ultrafiltration through membranes. Size determination by ultrafiltration through membranes of different MWCO was carried out as described. Values are mean ± SEM (n = 10 patients). ^*P < 0.01, significantly different values from NPP. 70% represents the untreated supernatant at 70% ammonium sulfate saturation. 50-kD Retentate and 30-kD Retentate are the residual protein that did not pass through membrane of 50-kD and 30-kD MWCO, respectively. 50-kD Filtrate and 30-kD Filtrate are the protein that passed through 50-kD and 30-kD membrane, respectively.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Proteinuria in rats after intravenous injection of 70% supernatant from FSGS plasma. Effects of injecting 70% supernatant from normal pooled plasma or FSGS plasma on rat urinary protein are shown. Preinjection urine protein and creatinine were determined in overnight (18 h) collection for each rat. Rats were anesthetized and injected intravenously with 1 ml of 70% supernatant containing 12 mg of protein from normal or FSGS plasma. Postinjection urine samples were collected from 6 to 24 h after injection, constituting an 18-h collection period. Seventy percent supernatants from four FSGS plasma samples were used in separate experiments. Pre- and postinjection urinary protein (mg protein/mg creatinine) were compared. Values represent mean ± SEM. n represents the number of animals studied, and P < 0.01 was accepted as significant.

Documentation of FS Activity in Plasma of FSGS Patients

Demographic characteristics of the patients studied and the initial P_alb of the plasmapheresis samples are shown in Table 1. Plasma from each patient with FSGS caused a significant increase in P_alb. The volume of plasma obtained from plasma-pheresis ranged from 960 to 4450 ml (typically about 4000 ml). It is important to note that plasmapheresis fluid contains both patient plasma and the human albumin that was used as a replacement solution. Total protein in the samples ranged from 30 to 170 g.

Effect of Removal of Lipids by Organic Solvents

To determine whether the FSGS factor is a lipid, we used an aliquot of cryoprecipitate-free plasma and extracted total lipids by chloroform/methanol (2:1) as described. The amount of total lipids in the plasma specimens averaged 570 mg/dl (Table 2). The lower chloroform phase contained 570 ± 210 mg/dl lipids and no detectable protein. The lipids extracted by this method did not increase P_alb of isolated rat glomeruli.

Enrichment of FSGS Factor

Precipitation of Lipoproteins and Chylomicrons. Initial treatment of plasma with high molecular weight dextran sulfate and calcium chloride removed 83.5% of total lipids and 15.5% of the total protein without affecting the FS activity (Table 2).

Precipitation of Proteins by Ammonium Sulfate. The lipoprotein and chylomicron-free supernatant was further enriched by a two-stage precipitation of plasma proteins with ammonium sulfate. In the first step, ammonium sulfate was added to 50% saturation, which precipitated 38% of the proteins. This precipitate did not redissolve completely and was discarded. The supernatant obtained after centrifugation contained about 62% of the initial plasma protein and tested positive for the FS activity.

Increasing the concentration of ammonium sulfate in the supernatant from 50 to 70% resulted into precipitation of the majority of plasma proteins. After centrifugation, a supernatant containing 1.8% of the initial plasma protein (Figure 1) was obtained. The precipitate at 70% ammonium sulfate saturation could be redissolved and did not show FS activity. The residual protein in the supernatant at 70% ammonium sulfate saturation (70% supernatant) could be precipitated by increasing ammonium sulfate to 80% saturation (70 to 80% precipitate). The 70 to 80% precipitate had FS activity in that it increased glomerular albumin permeability in the in vitro bioassay. The supernatant at 80% ammonium sulfate saturation did not contain any detectable protein (data not shown) and tested negative for FS activity. As shown in Table 3, we achieved a 100-fold concentration of the FSGS factor by sequential removal of the cryoprecipitate, lipoproteins, and most immunoglobulins and albumin from the plasma of FSGS patients by following the protocol described above. No further enrichment was afforded by precipitation at 80% ammonium sulfate saturation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Protein removal by sequential precipitation. Plasmapheresis fluid from individual focal segmental glomerulosclerosis (FSGS) patients was processed separately according to the methods described in the text. Average values of the percent protein removed from the plasma are presented as mean ± SD (n = 15 patients). Values for protein in supernatants after lipoprotein removal (Post Lipo) at 50 and 70% ammonium sulfate saturation are shown. The amount of protein in the supernatant at 80% saturation of ammonium sulfate was calculated as: Protein in the 70% supernatant — The amount of protein in the 80% precipitate. The number above each bar represents the percent of protein removed from the plasma specimen.

Characteristics of the FSGS Factor

Every plasma specimen caused an increase in P_alb of isolated rat glomeruli. The activity was not found to alter after storage up to 3 yr at -20°C. Addition of the protease inhibitor PMSF or a cocktail of inhibitors also did not make any discernible difference in activity (data not shown). Thus, we have not found any obvious difference in FS activity due to prolonged storage of samples at -20°C or by addition of protease inhibitors.

Some biochemical characteristics of the enriched FS activity in the 70% supernatant fraction are shown in Table 4 and Figure 2, A and B. Heating the 70% supernatant for 10 min or longer at 100°C abolished its effect on glomerular albumin permeability. The addition of protease also eliminated the FS activity (Table 4).

As shown in Figure 2A, FS activity in the 70% supernatant was retained in the 12- to 14-kD MWCO dialysis tube but not in the 50-kD MWCO membrane. However, FS activity was recovered from the retentate when untreated FSGS plasma was dialyzed in a 50-kD membrane.

Centrifugation-based membrane ultrafiltration of the 70% supernatant from FSGS plasma showed that the FS activity passed through a 50-kD membrane and was recovered in the 50-kD filtrate. However, the FS activity did not pass through 30-kD MWCO membrane and was recovered in the 30-kD retentate (Figure 2B). These results suggest that the FS activity in the 70% supernatant can be recovered from a 30- to 50-kD fraction.

Electrophoresis

SDS-PAGE of plasma and its fractions showed several bands corresponding to proteins of a wide molecular size range. Figure 3 represents a typical SDS-PAGE separation of proteins in different fractions obtained from FSGS plasma. A comparison of initial plasma protein and the 70% ammonium sulfate fractions showed a decrease in high molecular weight proteins consistent with the removal of proteins such as immunoglobulins and lipoproteins.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of proteins at different stages of purification of the FSGS factor. An equal amount of total protein from each stage of purification was analyzed by SDS-PAGE on a 4 to 20% gradient Tris-glycine gel and developed by Coomassie-G-250 staining. Lanes on the gel were designated as follows: STD, wide range standard protein mixture was used as a reference; lane 1, serum from FSGS patient; lane 2, FSGS plasmapheresis fluid; lane 3, supernatant after precipitation with dextran sulfate; lane 4, supernatant at 50% ammonium sulfate saturation; lane 5, supernatant at 70% ammonium sulfate saturation.

Figure 4 illustrates electrophoretic separation of the proteins in 70% supernatants from normal (lane A) and plasma from two FSGS patients who experienced recurrence of FSGS in renal allografts (lanes B and C) under nondenaturing conditions. Lane B is from an anuric patient on hemodialysis, and lane C is from a patient with nephrotic syndrome due to recurrent FSGS. An increase in the low molecular weight protein fraction is evident in both lanes B and C compared with lane A.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Polyacrylamide gel electrophoresis of 70% supernatants from normal and FSGS plasma under nondenaturing conditions. Electrophoretic analysis of total protein in the 70% supernatants using a 4 to 20% gradient Tris-glycine gel under nondenaturing conditions followed by silver staining. Supernatants at 70% ammonium sulfate saturation containing equal amounts of protein from three patients were used for electrophoresis in separate lanes. Lanes on the gel were designated as follows: Lane A, normal pooled plasma; lane B, plasma from a patient currently on hemodialysis with end-stage renal disease after recurrent FSGS; lane C, plasma from recurrent FSGS patient.

In Vivo Effect of the 70% Supernatant

Amount of urinary protein (mg protein/mg of creatinine) before and after intravenous injection of 70% supernatant from normal or FSGS plasma into rats is shown in Figure 5. Injection of FSGS 70% supernatant caused a significant increase in urinary protein from 3.1 ± 1.8 (preinjection) to 9.4 ± 2.4 mg (6 to 24 h postinjection). There was no increase in urinary protein (mg protein/mg creatinine) in rats injected with normal 70% supernatant. As shown, the increase in urinary protein caused by the injection of FSGS 70% supernatant from four patients with recurrent FSGS was consistently present, suggesting that the increase in urine protein by FSGS 70% supernatant is not a patient-specific phenomenon.