Background

Patients suffering from focal and segmental glomerulosclerosis (FSGS) and in whom this disease recurs after transplantation are likely to have an active form of the disease and to have a factor(s) (such as, albuminuric factor) present in their blood that alters glomerular permeability for albumin.

Methods

We used a sequential 50 and 70% ammonium sulfate (AS) precipitation of plasma from patients with relapsing FSGS and non-FSGS nephrotic syndrome (NS), in addition to plasma from healthy individuals, to obtain both an immunoglobulin (Ig)-rich fraction (50% AS precipitate) and a non-Ig fraction (70% AS supernatant). These fractions were injected intra-arterially or intravenously/intraperitoneally into Sprague-Dawley rats, and proteinuria (g protein/mmol creatinine) was measured for 24 hours. Ig fractions eluted from immunoadsorption onto protein A were also tested. A biochemical characterization was then carried out on the 70% AS supernatants by ultrafiltration on 30 and 50 kD cut-off membranes and by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Differentially stained bands were sequenced.

Results

The 70% AS supernatants from FSGS patients induced proteinuria when injected intra-arterially into normal rats. This effect was significantly different (P < 0.05) from that observed when similar fractions were prepared from the plasma of patients suffering from non-FSGS NS, but was not different from that observed with fractions from healthy individuals and even with an injection of saline solution. Injections of other plasma fractions did not induce a significant proteinuria in the FSGS group versus the non-FSGS NS group. SDS-PAGE of 70% AS supernatants revealed a protein of 23 kD that was more concentrated in AS supernatants from FSGS plasma than the other plasma samples and that was identified by microsequencing as apolipoprotein A1. After sequential ultrafiltration of 70% AS supernatants on 30 and 50 kD cut-off membranes, a second band of 43 kD was found at a much higher concentration in the FSGS samples than in non-FSGS NS and healthy individuals samples. This band is likely to correspond to a candidate albuminuric factor recently reported by another group¹, and was identified by microsequencing as ₁ acid glycoprotein or orosomucoid. Consequently, purified orosomucoid from the plasma of FSGS, non-FSGS NS patients, or healthy individuals was injected intra-arterially into rats. No differences were found between the proteinuria induced in each group.

Conclusions

These data strongly suggest that in vivo injection of material into the rat is not a reliable model for testing plasma fraction activity and that the 43 kD orosomucoid is not likely to be the albuminuric factor.

METHODS

Animals

Studies were performed on healthy male Sprague-Dawley rats (250 to 275 g; Janvier, Le Genest Saint Isle, France), which were fed standard laboratory food and water for one week prior to injection. Animal experimentation was conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Patient plasma

Plasma samples were obtained from plasmapheresis procedures on patients with FSGS who had presented a relapse of their initial disease. Control plasma was obtained from kidney recipients with NS of an origin other than FSGS, including one patient with nephroangiosclerosis, one with diabetes, and another with IgA nephritis. The first two patients presented an allograft glomerulonephritis related to chronic rejection and the third presented a late relapse of IgA nephritis. All of these control non-FSGS patients were also under immunosuppression with cyclosporine A, and their mean proteinuria and blood albumin levels were not significantly different from the FSGS patients. Table 1 shows the main clinical characteristics of FSGS and control NS patients. FSGS relapse was defined as the appearance of selective proteinuria within the weeks following transplantation and was also diagnosed from renal biopsies of both the patient's own kidney (initial disease) and the graft. Plasma from healthy individuals was obtained from the local blood bank. All plasma samples were handled under sterile conditions and stored at -20°C until used.

Table 1 - Clinical parameters for the initial kidney disease and the nature of graft functional alteration (initial disease relapse or allograft glomerulonephritis) in FSGS and control patients.

Full table

Preparation of 70% ammonium sulfate supernatants and 50% precipitates

A two-step method (50 and 70%) was used to sequentially remove Ig and albumin from plasma. Plasma was thawed, and the quantity of AS (Merck, Nogent s/Marne, France) necessary to obtain 50% saturation was added slowly. The mixture was placed at 4°C overnight under agitation. The mixture was then centrifuged at 3000 g for 30 minutes, and the supernatant was removed. The precipitate was recovered, diluted, with distilled water until complete dissolution and dialyzed (molecular weight cut-off = 3500) against phosphate-buffered saline (PBS; 3 changes) for two days. AS was added to the 50% supernatant to reach 70% saturation and was agitated at 4°C overnight. The precipitate was removed by centrifugation at 3000 g for 30 minutes. The supernatant was dialyzed (Spectra/por dialysis tubing, molecular weight cut-off = 3500; Polylabo, Strasbourg, France) against distilled water (5 changes) and PBS (4 changes) for three days. All supernatants were concentrated tenfold with respect to the initial plasma volume using polyethylene glycol (PEG) (molecular weight 35,000 D; Merck) in order to preserve the initial proportion of the putative "AF" in the solution. The mean final protein concentration in supernatants was 4.9 4.5 g/L, except for one sample (T), in which the protein content was 25.4 g/L. The protein content of each 50% precipitate was measured, and solutions were diluted to obtain a protein concentration in the range of the 70% AS supernatant concentration. Thus, the mean protein concentration in precipitates was 8.62 0.65 g/L.

Ultrafiltration

Seventy percent of the AS supernatants were filtered using an ultrafiltration cell (Amicon, Beverly, MA, US) and a 50 kD Omega membrane (Pall Corporation, St. Germain en Laye, France) at 1.0 bar, under agitation at 4°C. The 50 kD ultrafiltrate was then filtered using a 30 kD membrane (Amicon) at 1.6 bar, under agitation at 4°C. The 30 kD retained fraction constituted the 30 to 50 kD fraction of the 70% AS supernatant.

Preparation of protein A column eluates–anti-Gal antibody depletion

The protein A eluate from FSGS patient plasma was obtained as described previously³. The eluate from NS plasma was obtained by passage through a protein A column: After centrifugation to remove lipids, the plasma was passed through the column. Then the column was washed thoroughly with PBS and eluted with 200 mL of 0.2 mol/L glycine-HCl buffer, pH 2.5. The eluate was recovered, neutralized with 1 mol/L Na₂HPO₄, and dialyzed (Spectra/por dialysis tubing, molecular weight cut-off = 3500) against PBS (3 changes) for two days. The column was successively washed with 2 mol/L urea, 1 mol/L lithium chloride, and 0.2 mol/L glycine-HCl, pH 2.5. The column was stored in PBS 0.05%-NaN₃. The protein concentration in the protein A eluate was 6 g/L for the FSGS sample and 4 g/L for the NS sample.

Synthetic Gal1-3Gal disaccharide-polyacrylamide conjugates covalently linked to Sepharose 6FF were used for the removal of anti-Gal antibodies (Xenotran; Syntesome GmbH, Munich, Germany) by immunoadsorption. Each protein A eluate was passed through the immunoadsorption column and dialyzed against PBS (3 changes) for two days. The column was washed with PBS, eluted with 1% NH₄OH, and stored with PBS-NaN₃. The total protein content, after depletion of anti-Gal Ab, was 4 g/L for the FSGS and 3 g/L for the NS eluate.

SDS-PAGE, Western blot, and protein assays

Proteins (0.6 g) were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), under reducing conditions, on an 12% polyacrylamide gel according to Laemmli¹⁷. Proteins were silver stained, and the gel was stored in a 10% glycerol solution. Alternatively, proteins were transferred onto a nitrocellulose membrane, and Western blots were carried out using the ECL Western Blot kit (Amersham, Arlington Heights, IL, USA) together with peroxidase-conjugated antibodies: anti- (A2290 Sigma; 1/1000), anti- (A7164 Sigma; 1/1000), anti- (A5175 Sigma; 1/1000), and antiorosomucoid (A0011 Dako; 1/5000).

Albumin and Ig (G, A, and M) content was assessed by nephelemetry, using a BN II nephelemeter (Dade Behring S.A, Paris La Défense, France) with Behring assay reagents. In all of the AS supernatants, IgA and IgM were undetectable.

Sequencing of electrophoretic bands

Proteins were separated by SDS-PAGE in a 10% polyacrylamide gel (1.5 mm thick). The gel was fixed twice for 30 minutes in 50% methanol-10% acetic acid, stained overnight with 0.003% amido-black, and destained with water. The protein-containing gel bands were cut out and submitted to protease digestion. The 68 and 43 kD bands were digested using endoprotease Lys C. The 23 kD band was digested using trypsin. Peptides were separated by high-performance liquid chromatography (HPLC) on a 2 mm 25 cm C18 reverse-phase column with a DEAE precolumn. Elution was performed with a gradient of acetonitrile/water 0.1% trifluoacetic acid (TFA). One or two of the best defined peptide peaks of each analysis were submitted to sequencing on a Procise sequencer (Applied Biosystem, Foster City, CA, USA). Digestion, HPLC, and sequencing were performed at the Protein Microsequencing Laboratory at the Pasteur Institute (Paris, France).

Orosomucoid preparation

Isolation was carried out by trichloroacetic acid (TCA) precipitation of all other proteins¹⁸ to obtain an approximate 95% purity. Briefly, cold plasma samples were mixed (vol/vol) with a refrigerated 10% TCA solution under agitation and left on ice overnight. The mixture was centrifuged at 10,000 g for 30 minutes at 4°C. The supernatant was removed and lyophilized. It was dissolved in PBS, and the pH was adjusted to 7.4 with 1N NaOH. Efficient purification was confirmed by SDS-PAGE analysis. The final orosomucoid concentration was between 0.55 and 0.89 g/L.

In vivo injections of plasma fractions into rats

Intra-arterial injections were performed first. Animals were placed in metabolic cages for 24 hours before injection to determine the preinjection level of proteinuria. Rats were anesthetized by a mixture of isofluoran/oxygen, and the right kidney was removed. The aorta was clamped above and below the left kidney artery, and the plasma fractions (1 mL) were slowly (5 to 8 minutes) injected directly into the aorta under the junction with the left renal artery (the total time of ischemia was about 10 minutes). Rats were allowed to recover from anesthesia and were transferred to metabolic cages. Urine samples were collected for 24 hours following the injection, and diuresis was measured.

This procedure was chosen in order to increase the exposure of the recipient's kidney to the putative albuminuric factor. As it could also have induced aspecific lesions caused by local ischemic damage, we separately collected the urine produced within the six hours following injection and that produced during the following 18 hours (6 to 24 hours).

Finally, for the same reasons, the effect of IV/IP injections of plasma fractions was also investigated. Animals whose proteinuria level had previously been measured were anesthetized, and the plasma fractions (1 mL) were injected into the penis dorsal vein. One milliliter of the same fraction was then injected intraperitoneally. The same two (IV-IP) injections were repeated 48 and 72 hours later. Urine was collected in metabolic cages for 24 hours after the first and the third injection.

Each rat received one sample of individual fraction (from one patient), and this sample was injected into at least three rats (from 3 to 7 rats for one individual sample in one type of injection).

Measurement of proteinuria and urine electrophoreses

Rats were placed in metabolic cages with free access to water but without food. The total protein concentration (g/L) of urine was measured by a colorimetric test using an autoanalyzer Hitachi 717 (Boehringer, Mannheim, Germany). Urinary creatinine was measured by the Jaffe method and expressed in mmol/L. Proteinuria was expressed as the ratio (g/mmol) of total protein (g/L) to urinary creatinine (mmol/L).

Agarose electrophoresis of urinary proteins was performed using the Protur Hisi kit (Beckman, Gagny, France). Proteins were stained with amido-black. The scanning of nondenaturating electrophoresis by a densitometer (Hysys, Sebia, Issy Leo Noulineaux, France) allowed an integration of the electrophoresis pattern, which determined the percentage of albumin contained within the urine samples.

Statistical analysis

Results are expressed as mean SD. The nonparametric Wilcoxon test was used to compare the effects of the various plasma fractions on rat proteinuria before and 24 hours following the injection. P < 0.05 was accepted as significant.

RESULTS

Protein content of different plasma fractions

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of 70% AS supernatants showed one major and multiple minor protein bands Figure 1. The major band (molecular weight of 68 kD) was confirmed to be albumin Table 2. The minor band of approximately 23 kD was more concentrated in FSGS than in non-FSGS NS and normal plasma fractions. Upon sequencing, it was identified as apoliprotein A1 Table 2. Other proteins appeared to be equally distributed among the different supernatants. Heavy and light Ig chains appeared as minor components, although they could be detected by Western blotting (data not shown). After PEG treatment, the albumin concentration in the 70% AS supernatants was below 2.5 g/L, except for the supernatant from the single FSGS patient T, where it was 11.9 g/L. The Ig concentration (exclusively IgG) was always below 0.5 g/L.

Figure 1.

SDS-PAGE of 70% ammonium sulfate (AS) supernatants of different plasma. Proteins (0.6 g) were analyzed in a 12% polyacrylamide gel under reducing conditions with Laemmli buffers. Proteins were silver stained. Lanes are: (1) healthy individuals, (2–4) focal and segmental glomerulosclerosis (FSGS), (5 and 6) non-FSGS nephrotic syndrome (NS), and (7) protein A column eluate of plasma from a patient with hemochromatosis, enabling the position of albumin, heavy and light Ig chains to be located (arrows).

Full figure and legend (154K)

Table 2 - Sequence of peptides separated by HPLC after protease digestion of proteins isolated by SDS-PAGE.

Full table

After ultrafiltration on 50 and 30 kD membranes, the 30 to 50 kD fraction of the 70% AS supernatant from an FSGS plasma sample contained three major proteins: albumin, a 23 kD protein confirmed as apolipoprotein AI, and a 43 kD protein, which was not detected in the non-FSGS NS plasma Figure 2. The 43 kD protein was isolated by preparative gel electrophoresis and sequenced Table 2. It was identified as the ₁ acid glycoprotein, also known as orosomucoid. Orosomucoid was present in the fractions prepared from other FSGS patients; although not silver stained, it was also detected by Western blot in preparations from both normal individuals and non-FSGS NS patients Figure 3.

Figure 2.

SDS-PAGE of the 30 to 50 kD fraction of 70% ammonium sulfate supernatants. Lane 1, FSGS plasma; lane 2, non-FSGS nephrotic syndrome plasma.

Full figure and legend (22K)

Figure 3.

Western blot with anti-orosomucoid performed on 70% AS supernatants of different plasma. Lanes are: (1) healthy individuals, (2–4) FSGS, (5 and 6) non-FSGS NS, and (7) orosomucoid isolated from plasma by TCA precipitation.

Full figure and legend (36K)

Three bands were detected following SDS-PAGE of 50% AS precipitates; these were identified as albumin, and Ig light and heavy chains (data not shown). The electrophoresis of different protein A column eluates revealed the two major bands corresponding to heavy and light Ig chains in addition to albumin and other minor components (Figure 1, lane 7).

Effect of injection of 70% AS supernatants into rats

The effect of the intra-arterial injections of 70% AS supernatants and of an isotonic saline solution on proteinuria was compared first. Figure 4a shows a borderline significant increase in urinary protein excretion in the FSGS group 24 hours following injection, as compared with preinjection values (0.178 0.019 vs. 0.29 0.075 g/mmol, P < 0.05). Proteinuria reverted to the baseline value 48 hours following injection (data not shown). Preparations from non-FSGS NS patients only moderately influenced basal proteinuria values (0.167 0.034 vs. 0.23 0.014 g/mmol, P = NS). However, an injection of AS supernatants from healthy individuals was also associated with a rise in protein excretion of approximately the same magnitude as that seen after injection of AS supernatants from FSGS patients (0.169 0.022 vs. 0.29 0.059 g/mmol). Furthermore, a similar volume of isotonic saline also had a significant effect of roughly the same magnitude (0.171 0.025 vs. 0.279 0.059 g/mmol). Separate analysis of urine collected from 0 to 6 hours and from 6 to 24 hours postinjection¹ gave similar results (data not shown).

Figure 4.

Proteinuria after injection of plasma fractions into the rat vasculature. Urine was collected for 24 hours. Data are expressed as mean protein (g/L)/creatinine (mmol/L) SD (error bars). The Wilcoxon statistical test was used to compare preinjection data and postinjection data for each group. A significant P was indicated by an asterisk. (A) Effect of intra-arterial injection of 70% AS supernatants. Four groups were tested: the non-FSGS NS (3 patient fractions), the FSGS (5 patient fractions), the healthy individuals group (19 rats, 4 patient fractions, not shown), and the saline group (10 rats). (B) Effect of intravenous and IP injection of 70% AS supernatants. Three groups were tested: the non-FSGS NS (3 patient fractions), the FSGS (5 patient fractions), and the healthy individuals group (12 rats, 3 patient fractions, not shown). (C) Effect of intra-arterial injection of 50% AS precipitates. Three groups were tested: the non-FSGS NS (2 patient fractions), the FSGS (3 patient fractions), and the healthy individuals group (7 rats, 2 patient fractions, not shown). (D) Effect of IV-IP injection of 50% AS precipitates. Three groups were tested: the non-FSGS NS (2 patient fractions), the FSGS (3 patient fractions), and the healthy individuals group (12 rats, 2 patient fractions, not shown). N = total number of rats in each group. Symbols in A and C are: () pre-injection; () post-injection. Symbols in B and D are: () pre-injection; () post-injection; () third injection.

Full figure and legend (60K)

Before the injection, albumin was found to account for 76 20% of the protein in rat urine. The percentage of albumin present in the urine did not increase significantly during the 24 hours following injection of AS supernatants from FSGS (79.7 8.5%), non-FSGS NS patients (78.5 9.6%), healthy individuals (83.8 8%), or saline (84 0.7%).

No significant difference was found between the pre- and the post-IV-IP injection proteinuria level in the three groups Figure 4b: the non-FSGS NS group (0.123 0.03 vs. 0.14 0.032 g/mmol), the FSGS group (0.136 0.03 vs. 0.15 0.026 g/mmol), and the healthy group (0.139 0.05 vs. 0.17 0.068 g/mmol).

Effect of injection of 50% AS precipitates into rats

After the intra-arterial injection of 50% AS precipitates (2.2 0.65 g/L albumin concentration), a nonspecific increase in proteinuria occurred in the three groups: non-FSGS NS (0.111 0.032 vs. 0.315 0.122 g/mmol), FSGS (0.112 0.021 vs. 0.278 0.086 g/mmol; Figure 4c), and the healthy group (0.112 0.021 vs. 0.311 0.15 g/mmol). Similarly, IV-IP injections of 50% AS precipitates from the various groups (non-FSGS NS, FSGS, and healthy individuals) had no effect on proteinuria Figure 4d.

Effect of injection of protein A column eluates

The intra-arterial injection of protein A column eluates that were not depleted of anti-Gal Ab Figure 5a resulted in a roughly similar proteinuria in the two groups (0.125 0.022 vs. 0.187 0.079 g/mmol for the non–FSGS NS eluates and 0.125 0.022 vs. 0.216 0.82 g/mmol for the FSGS eluates, P = NS). Human serum contains high levels (estimated to be 1% of total Ig) of xenoantibodies directed against most mammalian endothelial cells¹⁹. As the injected fractions contain xenoantibodies, which are able to bind strongly and activate the renal endothelium and possibly result in nonspecific proteinuria, thus masking the effect of the putative albuminuric factor, we decided to remove the anti-Gal antibodies by immunoadsorption. Depletion of anti-Gal Ab (Figure 5 c, d) did not reveal an effect of FSGS versus non-FSGS NS fractions.

Figure 5.

Proteinuria after injection of plasma fractions into the rat vasculature. Urine was collected for 24 hours. Data are expressed as mean protein (g/L)/creatinine (mmol/L) SD (error bars). The Wilcoxon statistical test was used to compare pre-injection data and post-injection data for each group. A significant P was indicated by an asterisk. One patient fraction was used in each group. (A) Effect of intra-arterial injection of protein A column eluates. Two groups were tested: the non-FSGS NS and the FSGS group. (B) Effect of IV-IP injection of protein A column eluates. Two groups were tested: the non-FSGS NS and the FSGS group. (C) Effect of intra-arterial injection of protein A column eluates depleted of anti-Gal antibodies. Two groups were tested: the non-FSGS NS and the FSGS group. (D) Effect of IV-IP injection of protein A column eluates depleted of anti-Gal antibodies. Two groups were tested: the non-FSGS NS and the FSGS group. N = total number of rats in each group. Symbols in A and C are: () pre-injection; () post-injection. Symbols in B and D are: () pre-injection; () post-injection; () third injection.

Full figure and legend (47K)

Correlation between albumin content in AS supernatants and proteinuria

Since our aim was to determine whether plasma from relapsing patients contained an uncharacterized AF, we injected a similar volume of the material from FSGS, non-FSGS NS, and normal plasma into the rat. However, as the injected protein content varied, particularly with respect to albumin, and since in some conditions proteinuria has been related to protein overload^20,21,22, we checked whether the quantity of albumin injected and the level of proteinuria were correlated. No significant relationship (r = 0.45) was detected (data not shown), thus suggesting that the difference observed was not caused by protein overload. A similar conclusion was obtained when the total amount of injected protein was plotted against proteinuria.

Effect of intra-arterial injection of patient-derived orosomucoid into rats

Since orosomucoid from FSGS patients could have undergone discrete modifications post-transplant that could have altered its capacity to interfere with glomerular permselectivity, orosomucoid preparations from both FSGS and non-FSGS NS plasma were injected into the rat renal artery.

As it was reported that significant differences in proteinuria following AS supernatant injection were observed in urine collected between 6- and 24-hours postinjection¹, we analyzed the proteinuria during the first 6 hours and during the following 18 hours. Proteinuria during the first six hours was increased to the same extent in each group (data not shown). Figure 6 shows that there was no difference in 24-hour urinary protein excretion between rats injected intra-arterially with the orosomucoid fraction prepared from normal plasma, plasma from FSGS patients, and plasma from non-FSGS NS patients (P = 0.018, P = 0.028, and P = 0.018, respectively).

Figure 6.

Proteinuria after injection of orosomucoid purified from plasma. Urine was collected for 24 hours. Data are expressed as protein (g/L)/creatinine (mmol/L) SD. Three groups of seven rats were tested with orosomucoid derived from FSGS (3 patient fractions), non-FSGS NS (one patient fraction), and healthy individuals plasma (one patient fraction). P was significant in the three groups. Symbols are: () pre-injection; () post-injection.

Full figure and legend (44K)

DISCUSSION

The immediate recurrence of idiopathic NS after renal transplantation strongly suggests the involvement of circulating factor(s) in the pathogenesis of this disease. Thus, characterization of this factor would represent a major breakthrough in understanding its pathology. Some authors have suspected a T-cell product to be involved^23,24,25. Others have reported vascular permeability activity^9,26,27 or implied that cationic agents allowed a leakage of albumin through the glomerular urinary space^28,29,30. However, the nature of the factor(s) responsible for this disease remains elusive. Recently, we showed the dramatic effect of ex vivo protein A³ and anti-human Ig immunoaffinity columns⁴ on proteinuria in relapsing patients, suggesting the involvement of immunoglobulin in the recurrence of FSGS following transplantation. Although in our experience, membranous glomerulonephritis and primary FSGS patients were not always sensitive to protein A adsorption^3,31, this treatment may decrease albuminuria in other forms of NS³². Since fractions from protein A column eluates with a molecular weight below 100 kD that were from two of the four patients originally tested have induced albuminuria in rats³, we hypothesized that the factor could be either an immunoglobulin fragment or an immunoglobulin binding protein. We have now reinvestigated the in vivo effect of Ig-depleted fractions obtained from plasmapheresis fluids when injected into rats. Thus, using a simple technique of sequential protein precipitation with AS, similar to that recently utilized by Sharma et al¹, we removed almost all immunoglobulin and the majority of albumin from plasma. Finally, the quantity of injected proteins was lower than that able to give a nonspecific protein-loading effect^20,21,22, as also shown by the lack of correlation between the amount of human albumin injected into rats and the induction of proteinuria. The FSGS-derived sample was the only fraction in the 70% AS supernatants to induce a significant increase in urinary protein excretion at 24 hours as compared with preinjection values. Surprisingly, an injection of AS supernatants from healthy individuals as well as the injection of saline also resulted in a large increase in proteinuria as compared with the 24-hour preinjection values. This increase in proteinuria following the injection of a similar volume of preparations from healthy individuals or of saline suggests that this effect was due to an increase in the plasma volume and in the glomerular filtration rate following injection. However, the minor increase in proteinuria in the non-FSGS NS group suggests that more complex interactions can operate when NS samples are used. Therefore, this highlights the importance of comparing the effect of FSGS-derived supernatants with appropriate controls such as non-FSGS nephrotic sera and of mimicking as closely as possible the volume of material injected. Furthermore, the absence of any increase in the albumin content of urine suggests the occurrence of a nonspecific effect. Finally, because direct intra-arterial injection of the various fractions tested could have induced nonspecific ischemic lesions, which may have hidden an effect of the material injected and that anti-Gal antibodies were not involved, all of the experiments were repeated with IV-IP injection, without any further effect. These series of experiments, which have been performed on a much larger scale both in the number of sera tested individually (non-FSGS NS = 3 different sera, FSGS = 5 different sera) and in the number of rats injected (7 to 20 per group for intra-arterial injections and 12 per group for IV-IP injections, a total of 237 rats were injected), do not confirm our previous data obtained for two patients (from the first 4 who underwent the initial protein A procedure reported³).

A recent report stated that similar AS preparations from plasma from four relapsing FSGS patients significantly induced proteinuria in rats when compared with AS preparations obtained from normal individuals¹. However, the authors did not show the effect of preparations from patients with non-FSGS NS, thus making any comparison between different experimental series difficult. Interestingly, when urinary protein levels for the first six hours following injection were not taken into account (as in their report¹), the general proteinuria pattern between groups in our experiments was not modified.

Several factors could explain the transient or lack of effect on proteinuria following the transfer of FSGS biological materials: (1) The factor may be delivered at a much lower concentration than in the pathological state (even when injected directly into the kidney vasculature) and could also be rapidly eliminated with no continuous synthesis occurring. (2) The factor could be neutralized in the normal rat plasma by an antifactor, as suggested by the inhibitory effect of normal rat sera on the activity of FSGS plasma on albumin permselectivity in isolated rat glomeruli in vitro (abstract; Savin, Nephrol Dial Transplant 9:947, 1994). However, the extensive series of transfer experiments performed in this study, including intra-arterial or IV-IP procedures, indicates that this approach, which has also been used recently¹, is unlikely to yield convincing evidence concerning the presence of a putative factor. Moreover, although we injected a greater volume (IV and IP), the total quantity of protein injected in our procedure was similar to that injected by Sharma et al (9.8 9 mg/mL vs. 12 mg/mL, respectively)¹, suggesting that the discrepancy between the two studies cannot be explained by a difference in the quantity of protein injected.

Electrophoresis of injected AS supernatants revealed only a few differences between samples prepared from FSGS and non-FSGS NS patients. After sequencing of a protein (23 kD), which was found to be present in a significantly higher proportion in the FSGS than in non-FSGS NS, plasma fractions revealed apolipoprotein AI (Apo AI). Despite the fact that its molecular weight is consistent with the molecular weight estimation of the factor suggested in prior studies^33,34,35, its presence in low levels in preparations from normal (Figure 1, line 1) and non-FSGS NS sera (Figure 1, lines 5 and 6), as well as its anionic properties, suggests that Apo AI has no role in FSGS. Furthermore, a 43 kD band only detected in FSGS-derived preparations was identified as orosomucoid. A band with similar characteristics (but which was not sequenced) was recently proposed as a putative albuminuric factor by Sharma et al¹. Interestingly, orosomucoid was shown to restore glomerular permselectivity for albumin during ex vivo perfusion of isolated rat kidneys¹⁶. Since discrete alterations in the biochemical properties of this highly anionic molecule in FSGS patients could form a basis for its potential role in FSGS, we purified orosomucoid from the plasma of FSGS, non-FSGS NS, and healthy individuals, and injected it into the rat kidney vasculature. No evidence for a specific induction of proteinuria was found.

In summary, we have shown that a single intra-arterial injection of low-protein-containing plasma fractions from patients with an FSGS recurrence following transplantation induces a mild and transient proteinuria. However, although preparations from non-FSGS NS could not reproduce this effect, normal sera or even saline resulted in a signal of a similar magnitude. Furthermore, we show that the 43 kD fraction proposed as a putative AF¹ is orosomucoid, and that injection of this material derived from FSGS plasma into the rat kidney vasculature induced the same level of proteinuria as compared with the 43 kD fraction prepared from the control plasma.

References

References

1.	Sharma M, Sharma R, McCarthy E & Savin V. "The FSGS factor": Enrichment and in vivo effect of activity from focal segmental glomerulosclerosis plasma. J Am Soc Nephrol 1999; 10: 552−561. | PubMed | ISI | ChemPort |
2.	Dantal J, Giral M, Hourmant M & Soulillou J-P. Glomerulonephritis recurrences after kidney transplantation. Curr Opin Nephrol Hypertens 1995; 4: 146−154. | PubMed | ChemPort |
3.	Dantal J, Bigot E, Bogers W, Testa A, Kriaa F, Jacques Y, Hurault D, Ligny B, Niaudet P, Charpentier B & Soulillou J-P. Effect of plasma protein adsorption on protein excretion in kidney-transplant recipients with recurrent nephrotic syndrome. N Engl J Med 1994; 330: 7−14. | Article | PubMed | ISI | ChemPort |
4.	Dantal J, Godfrin Y, Koll R, Perretto S, Naulet J, Bouhours J-F & Soulillou J-P. Antihuman immunoglobulin affinity immunoadsorption strongly decreases proteinuria in patients with relapsing nephrotic syndrome. J Am Soc Nephrol 1998; 9: 1709−1715. | PubMed | ISI | ChemPort |
5.	Godfrin Y, Dantal J, Bouhours J-F, Heslan J-M & Soulillou J-P. A new method of measuring albumin permeability in isolated glomeruli. Kidney Int 1996; 50: 1352−1357. | PubMed | ISI | ChemPort |
6.	Godfrin Y, Dantal J, Perretto S, Hristea D, Legendre C, Kreis H & Soulillou J-P. Study of the in vitro effect on glomerular albumin permselectivity of serum before and after transplantation in focal segmental glomerulosclerosis. Transplantation 1997; 64: 1711−1715. | PubMed | ISI | ChemPort |
7.	Savin V, Sharma R, Lovell H & Welling D. Measurement of albumin reflection coefficient with isolated rat glomeruli. J Am Soc Nephrol 1992; 3: 1−10. | PubMed | ISI | ChemPort |
8.	Savin V, Sharma R, Sharma M, McCarthy E, Swan S, Ellis E, Lovell H, Warady B, Gunwar S, Chonko A, Artero M & Vincenti F. Circulating factor associated with increased glomerular permeability to albumin in recurrent focal segmental glomerulosclerosis. N Engl J Med 1996; 334: 878−883. | Article | PubMed | ISI | ChemPort |
9.	Zimmermann S. Increased urinary protein excretion in the rat produced by serum from a patient with recurrent focal glomerular sclerosis after renal transplantation. Clin Nephrol 1984; 22: 32−38. | PubMed | ISI | ChemPort |
10.	Wilkinson A, Gillespie C, Hartley B & Williams D. Increase in proteinuria and reduction in number of anionic sites on the glomerular basement membrane in rabbits by infusion of human nephrotic plasma in vivo. Clin Sci 1989; 77: 43−48. | PubMed | ISI | ChemPort |
11.	Bakker WW & van Luijk WH. Do circulating factors play a role in the pathogenesis of minimal change nephrotic syndrome? Pediatr Nephrol 1989; 3: 341−349. | Article | PubMed | ISI | ChemPort |
12.	Boulton Jones J, Tulloch I, Dore B & McLay A. Changes in the glomerular capillary wall induced by lymphocyte products and serum of nephrotic patients. Clin Nephrol 1983; 20: 72−77. | PubMed | ChemPort |
13.	Cheung PA, Stulp B, Immenschuh S, Borghuis T, Baller JF & Bakker WW. Is 100KF an isoform of hemopexin? Immunochemical characterisation of the vasoactive plasma factor 100KF. J Am Soc Nephrol 1999; 10: 1700−1708. | PubMed | ISI | ChemPort |
14.	Bakker WW, Baller JF & Luijk WH. A Kallikrein-like molecule and plasma vasoactivity in minimal change disease. Contrib Nephrol 1988; 67: 31−36. | PubMed | ChemPort |
15.	Haraldsson BS, Johnsson EK & Pippe B. Glomerular permselectivity is dependent on adequate serum concentrations of orosomucoid. Kidney Int 1992; 41: 310−316. | PubMed | ISI | ChemPort |
16.	Johnsson E & Haraldsson B. Addition of purified orosomucoid preserves the glomerular permability for albumin in isolated perfused rat kidneys. Acta Physiol Scand 1993; 147: 1−8. | PubMed | ISI | ChemPort |
17.	Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227: 680−685. | Article | PubMed | ISI | ChemPort |
18.	Popenoe EA. Characterization of a glycoprotein in the urine of patients with proteinuria. J Biol Chem 1955; 217: 61−66. | PubMed | ISI | ChemPort |
19.	Galilli U. Interaction of the natural anti-Gal antibody with -galactosyl epitopes: A major obstacle for xenotransplantation in humans. Immunol Today 1993; 14: 480−482. | Article | PubMed | ISI | ChemPort |
20.	Eddy AA. Interstitial nephritis induced by protein-overload proteinuria. Am J Pathol 1989; 135: 719−733. | PubMed | ISI | ChemPort |
21.	Roy L, Vernier R & Michael A. Effect of protein-load proteinuria on glomerular polyanion. Proc Soc Exp Biol Med 1972; 141: 870−874. | PubMed | ISI | ChemPort |
22.	Chen A, Wei CH, Sheu LF, Ding SL & Lee WH. Induction of proteinuria by adriamycin or bovine serum albumin in the mouse. Nephron 1995; 69: 293−300. | PubMed | ISI | ChemPort |
23.	Yoshizawa N, Kusumi Y, Matsumoto K, Oshima S, Takeuchi A, Kawamura O, Kubota T, Kondo S & Niwa H. Studies of a glomerular permeability factor in patients with minimal−change nephrotic syndrome. Nephron 1989; 51: 370−376. | PubMed | ISI | ChemPort |
24.	Koyama A, Fujisaki M, Kobayashi M, Igarashi M & Narita M. A glomerular permeability factor produced by human T cell hybridomas. Kidney Int 1991; 40: 453−460. | PubMed | ISI | ChemPort |
25.	Lagrue G, Xheneumont S, Branellec A & Weil B. Lymphokines and nephrotic syndrome (letter). Lancet 1975; 1: 271−272. | PubMed | ISI | ChemPort |
26.	Lagrue G, Xheneumont S, Branellec A, Hirbec G & Weil B. A vascular permeability factor elaborated from lymphocytes. I. Demonstration in patients with nephrotic syndrome. Biomedicine 1975; 23: 37−40. | PubMed | ISI | ChemPort |
27.	Sobel A, Branellec A, Blanc C & Lagrue G. Physicochemical characterization of vascular permeability factor produced by CON A-stimulated human lymphocytes. J Immunol 1977; 119: 1230−1234. | PubMed | ISI | ChemPort |
28.	Vehaskari M, Root E, Germuth F & Robson A. Glomerular charge and urinary protein excretion: Effects of systemic and intrarenal polycation infusion in the rat. Kidney Int 1982; 22: 127−135. | PubMed | ISI |
29.	Hunsicker L, Shearer T & Shaffer S. Acute reversible proteinuria induced by infusion of the polycation hexadimethrine. Kidney Int 1981; 20: 7−17. | PubMed | ISI | ChemPort |
30.	Barnes JL, Radnik RA, Gilchrist EP & Venkatachalam MA. Size and charge selective permeability defects induced in glomerular basement membrane by a polycation. Kidney Int 1984; 25: 11−19. | PubMed | ISI | ChemPort |
31.	Haas M, Godfrin Y, Oberbauer R, Yilmaz N, Borchhardt K, Regele H, Drumi W, Derfler K & Mayer G. Plasma immunoadsorption treatment in patients with primary focal and segmental glomerulosclerosis. Nephrol Dial Transplant 1998; 13: 2013−2016 10.1093/ndt/13.8.2013. | PubMed | ISI | ChemPort |
32.	Esnault V, Besnier D, Testa A, Covill P, Simon P, Subra J-F & Audrain M. Effect of protein A immunoadsorption in nephrotic syndrome of various etiologies. J Am Soc Nephrol 1999; 10: 2014−2017. | PubMed | ISI | ChemPort |
33.	Garin H & Corontzes N. Effect of lymphokine from nephrotic peripheral blood mononuclear cells on catabolism of rat glomerular basement membrane sulfated compounds. Nephron 1992; 62: 416−421. | PubMed | ISI |
34.	Schnapper HW & Aune TM. Identification of the lymphokine soluble immune response suppressor in urine of nephrotic children. J Clin Invest 1985; 76: 341−349. | PubMed |
35.	Cheng IKP, Jones BM & Chan MK. The role of soluble immune response suppressor lymphokine in the prediction of steroid responsiveness in idiopathic nephrotic syndrome. Clin Nephrol 1989; 4: 168−172.

Acknowledgments

This work was supported in part by PHRC program number BRD 99/3-E and by the Foundation Transvie. We thank Ms. Joanna Ashton for her help in editing the manuscript.