Possible role of autoantibodies against nephrin in an experimental model of chronic graft-versus-host disease

Abstract

Nephrin, a product of the NPHS1 gene, is a component of the slit diaphragms that are found between glomerular foot processes and is a crucial element for glomerular filtration barrier. Recently, nephrin has been focused in a number of studies of proteinuria development including various types of acquired glomerular diseases including minimal change nephrotic syndrome and membranous nephropathy. However, the precise role of nephrin in such acquired glomerular diseases is still unknown. To analyse the role of nephrin further, two kinds of anti-nephrin antibodies were raised in the rabbits and applied to an experimental mouse model of chronic graft-versus-host disease, in which (C57BL/10 × DBA/2) F1 mice developed clinically apparent severe proteinuria with significant glomerular lesions 7 weeks after parental DBA/2 cell transfer. Antibody-sandwich ELISA detected anti-nephrin antibodies during week 2 to week 6, with the peak at week 2 or week 4. Colocalization of nephrin and IgG on week 4, week 6, and week 8 was revealed by confocal microscopic analysis, suggesting that in situ immune complex formation with nephrin in glomerular lesion. Taken together, it seems to be suggested nephrin and its autoantibody have a certain role in the development of glomerular lesion in our model mice.

Introduction

Murine chronic graft-versus-host disease (cGVHD), induced by transfer of parental T lymphocytes to an MHC-incompatible F1 recipient, is characterized by systemic lupus erythematosus (SLE)-like condition including severe immune complex glomerulonephritis associated with polyclonal B cell activation and autoantibodies [1,2]. Investigations done in the last few decades have resulted in the identification of several groups of autoantibody antigens occurring in human or experimental autoimmune related nephritis including DNA, dipeptidyl peptidase IV, glomerular basement membrane [1–4]. Patients with lupus nephritis are immunologically characterized by the production of a wide variety of autoantibodies and the formation and deposition of immune complexes therefore autoantibody formation is a key phenomenon in autoimmunity.

Recent studies in nephrology has revealed that slit diaphragm of glomeruli is a crucial element in the aetiology of proteinuria [5–7]. Blood filtration occurs in the renal glomerulus, which contains a tuft of capillaries located inside the Bowman's capsule. The filtration barrier is composed of a fenestrated endothelial cell layer, glomerular basement membrane, and glomerular epithelial cells (podocytes) facing the urinary space of the glomerulus. Foot processes from adjacent two podocytes form a characteristic interdigitating pattern on the outside of the capillary; slit diaphragm. The slit diaphragm spans 40 nm between foot processes and consists of a zipper-like structure in which there is a central filament with flanking cross-bridges and intervening pores [8]. One of the molecular compositions of the slit diaphragm is a novel highly kidney-specific protein termed nephrin. Nephrin is a 185 kD transmembrane protein that was discovered by positional cloning as a gene linked to congenital nephrotic syndrome of the Finnish type (NPHS1) [5]. Nephrin was subsequently shown to localize in the podocyte filtration slits and has been proposed to have a crucial role in the maintenance of the glomerular filtration barrier, which apparently plays a pivotal role in preventing protein leakage [5,6]. The investigation using nephrin-deficient mice revealed that disruption of slit diaphragm structure results in loss of glomerular permselectivity and severe proteinuria development [9]. In addition, recent further studies of nephrin have also suggested its involvement in most acquired nephrotic syndromes including minimal change nephrotic syndrome and membranous nephropathy [10–13].

We postulated a certain role of nephrin and its antibodies in autoimmune related glomerulonephritis. We employed a murine model of cGVHD using DBA/2 and their F1. (C57BL/10 × DBA/2) F1 donor receiving DBA/2 spleen cells subsequently develops immunostimulant state with severe glomerulonephritis bypassing immunosuppressive acute GVHD. Here, we show the existence of antibodies against nephrin in sera of cGVHD model mice, and colocalization of IgG and nephrin, suggesting in situ immune complex formation. The significance of these findings is discussed.

Materials and methods

Polyclonal rabbit anti-mouse nephrin antibodies

For the generation of polyclonal antibodies against mouse nephrin, a recombinant protein containing the intracytoplasmic region corresponding to amino acids 1099–1177 (cNep) of mouse nephrin or extracellular region corresponding to amino acids 768–832 (eNep) was fused to either glutathione-S-transferase (GST) or hexahistidine and produced in E. coli BL21 cells. The DNA insert was amplified by PCR using mouse nephrin cDNA (a gift from Dr Larry Holzman, University of Michigan, Ann Arbor, MI) and verified by automated sequencing. The GST-fused antigen was subsequently used to immunize rabbits (Japan SLC Co., Shizuoka, Japan) and antiserum was then collected and purified using saturated ammonium sulphate, followed by dialysis against phosphate-buffered saline (PBS). This dialysed sample was absorbed to a column containing Affi-Gel 15 (Bio-Rad Laboratories, Hercules, CA, USA) coupled with the same mouse nephrin antigen that had been hexahistidine tagged. The column was washed extensively with PBS, and antibodies were eluted with 100 mM glycine, pH 2·5, and 0·5 ml fractions were collected in tubes containing 50 µl of 1 M Tris, pH 7·4. These fractions were also dialysed against PBS.

Preparation of glomerular lysate

Kidneys were obtained from 8 week old C57BL/10 mice and the glomeruli were isolated by sieving, as described elsewhere [14] and then each samples homogenized in a T-PER buffer (Pierce Biotechnology Inc., Rockford, IL, USA) containing protease inhibitors on ice.

Induction of cGVHD and evaluation of the appearance of proteinuria

Eight week old (C57BL/10 × DBA/2) F1, BDF1 and DBA/2 female mice were purchased from SLC Japan Co. and maintained in our animal facility. Chronic GVHD was induced by an intravenous injection of 5 × 10⁷ viable DBA/2 Spleen cells into BDF1 recipients at day 0 and at day 7, as previously described [15]. A cGVHD group (n = 10) and a non-cGVHD group (n = 6) were prepared for every two week, and both groups were sacrificed at week 0, 2, 4, 6 and 8. Urine and blood samples were collected at the sacrifice. Non-cGVHD controls were injected with the same number of syngenic spleen cells. Proteinuria was measured on an ordered categorical scale by Uristix (Bayer-Sankyo, Tokyo, Japan), with each scale of –, ±, +, 2+, 3+ and 4+ scored as 0, 1, 2, 3, 4, 5 for statistical analysis. Semiquantitative analyses of proteinuria were performed independently by KN and KM.

Light microscopic analysis

Light microscopic studies were performed using standard methods. Briefly, the samples were fixed in 10% formalin in PBS overnight, dehydrated and embedded in paraffin. Sections of 3 µm were made, then rehydrated and stained with periodic-acid Schiff (PAS).

Reagents

Goat anti-mouse IgG antibodies were purchased from Southern Biotechnology Associates (Birmingham, AL, USA). Alexa Fluor 488-conjugated donkey anti-goat antibody and Alexa Fluor 568-conjugated goat anti-rabbit antibody were obtained from Molecular Probes, Inc (Eugene, OR, USA).

Double-staining indirect immunofluorescent microscopy

Cryosections of 5 µm in thickness were placed on Silan-coated slides and dried at room temperature. Sections were then fixed in cold acetone for 5 min at 4 °C and washed in PBS. To block nonspecific binding, the samples were incubated in 5% bovine serum albumin in PBS for 30 min at room temperature. Then, the sections were incubated with rabbit anti-cNep antibody overnight at 4 °C, washed in PBS, and incubated with Alexa Fluor 568-conjugated goat anti-rabbit antibody for 60 min at 37 °C. Following additional blocking with 5% bovine serum albumin and PBS washes, the sections were then incubated with goat anti-mouse IgG and treated with Alexa Fluor 488-conjugated donkey anti-goat antibody. All antibodies were used at dilutions of 1 : 200 in PBS. Qualitative analysis of the distribution of nephrin and IgG deposits was performed by confocal microscopy using a FV1000 confocal laser microscope (Olympus, Tokyo, Japan). Analyses of microscopic findings are performed by two pathologists (KN and IA).

Immunoblotting

For testing the specificity of rabbit anti-nephrin antibodies, the following primary antibodies were used: rabbit anti-cNep antibody and anti-eNep antibody. The horseradish peroxidase-conjugated secondary antibodies were against rabbit (Bio-Rad Laboratories). Glomerular preparation separated by SDS-PAGE was transferred to nitrocellulose filters using a semidry blotting system (ATTO corporation, Tokyo, Japan) and after blocking and incubation with the respective antibodies detected by enhanced chemiluminescence (Amersham Pharmacia, Uppsala, Sweden).

For detecting immunoprecipitates, diluted (×100) pooled mouse sera were purified with saturated ammonium sulphate then used as primary antibodies. The horseradish peroxidase-conjugated secondary antibodies were against mouse (Bio-Rad Laboratories), and immunoprecipitates were revealed by enhanced chemiluminescence (Amersham Pharmacia).

Enzyme-linked immunosorbent assay (ELISA)

To detect anti-DNA antibodies, calf thymus double-stranded DNA (Sigma-Aldrich, St. Louis, MO, USA) was denatured at 100 °C to generate single-stranded DNA. Ninety-six-microwell plates (Corning, Corning, NY, USA) were precoated overnight at 4 °C with methylated BSA (Sigma-Aldrich), then washed with PBS and 0·02% Tween 20 and coated with the single-stranded calf thymus DNA. The plates were washed again and blocked with 2% BSA in PBS. Diluted (×200) sera were then added for 2 h at 37 °C. After additional washing with PBS-Tween, the plates were overlaid with phosphatase-labelled anti-mouse IgG for 2 h at 37 °C. The concentration of specific bound antibodies was determined by the addition of p-nitrophenylphosphate (Kirkegaard and Perry Laboratories, Gaithersburg, MD, USA) and optical density (OD) at 405 nm was measured using a microplate reader (Benchmark, Bio-Rad Laboratories) with high-titre antisera.

Antibodies against mouse nephrin were detected using antibody-sandwich ELISA. 96-well plates were precoated overnight at 4°C with 5 µg/ml of rabbit polyclonal anti-nephrin antibodies (anti-cNep or anti-eNep) in PBS, then washed with PBS-Tween and blocked with 2% BSA in PBS. 20 µg/ml of mouse glomerular preparation was then added for 2 h at 37°C, and washed with PBS-Tween and blocked with 2% BSA-PBS. Diluted (×10) sera were added and incubate for 2 h at 37°C. After additional washing with PBS-Tween, the plates were overlaid with phosphatase-labelled anti-mouse IgG for 2 h at 37°C. The concentration of bound antibodies was determined by the same method as that of anti-DNA antibodies.

Immunoprecipitation of nephrin

Anti-nephrin antibody (5 µg) was mixed with 50 µl of protein G-Sepharose (Amersham Pharmacia). After 2-h incubation at 4°C, glomerular lysate was added and incubated for 1 h at 4°C. Then, the unbound proteins were removed by washing the beads three times with wash buffer (50 mM Tris-HCl, pH 7·5, 150 mM NaCl, 5 mM EDTA, 0·05% Triton X-100). The immunoprecipitates were resuspended in 50 µl of sample buffer boiled for 5 min and analysed by immunoblotting.

Statistical analyses

Data are expressed as mean ± SEM. Statistical significance of differences in proteinuria was determined by the Wilcoxin-Mann–Whitney rank sums test with an adjustment for ties. Values of antibodies were log-transformed to achieve a normal distribution, and then statistical significance of differences in them was determined by the Student t-test. All statistical calculations were performed by KaleidaGraph (Synergy Software, Reading, PA, USA). P-value of less than 0·05 was considered to be significant.

Results

Production and characterization of rabbit anti-nephrin antibodies

In the initial studies, we conducted to produce anti-nephrin antibody in the rabbit as described in materials and methods. Rabbits were immunized with cytoplasmic and extracellular region of nephrin, respectively. Both of raised nephrin antibodies, anti-cNep and anti-eNep antibodies, were tested for their specificity by immunoblotting and immunofluorescence studies using glomerular lysate from 8 week old C57BL/10 mice. A clear 185 kD bands were detected in the immunoblotting experiments, consistent with the molecular size of the nephrin protein (Fig. 1a). The specific immunoreactivity of the antibodies was further identified by the glomerular-specific patterns of expression that were identical to those previously described for nephrin in naive mice [16], which express nephrin in fine granular patterns along the glomerular capillary walls (Fig. 1b).

Fig. 1.

Specificity of nephrin polyclonal antibody. (a) Western blot analysis of mouse glomerular preparation. A nitrocellulose strip was incubated with preimmune rabbit serum (lane 1), anti-mouse nephrin antibody raised in rabbits against a recombinant mouse nephrin intracytoplasmic region (aa1099–1177) (lane 2), antibody to extracellular region (aa768–832) (lane 3), or with antibody against intracytoplasmic region preincubated with recombinant antigen (lane 4). Western blot analysis using antibody against extracellular region preincubated with its antigen also showed no band (data not shown). (b) Immunostaining with the polyclonal nephrin antibody to intracellular region. A fresh frozen adult mouse kidney section was fixed with acetone and incubated with the anti-nephrin antibody recognizing its intracellular region. Nephrin staining is observed as a fine granular pattern along the glomerular capillary walls. (original magnification ×400) Immunostaining of nephrin antibody to extracellular region was the same pattern (data not shown).

Proteinuria levels and DNA antibody immunoreactivity in a mouse model of lupus nephritis

As previously noted, cGVHD mice developed proteinuria and anti-DNA antibody. Proteinuria extraction from cGVHD model mice began to increase significantly at 7 weeks after donor cell transfer (Fig. 2a). These mice developed autoantibody titres against ssDNA significantly on week 2 to week 6 (Fig. 2b).

Fig. 2.

Proteinuria (a) and serum DNA autoantibody (b). Values are represented as the mean ± SEM. Results showing statistically significant differences from the non-cGVHD group calculated by the Student t-test; *P < 0·05, **P < 0·0001. Statistical significance was calculated by the Wilcoxin-Mann–Whitney rank sums test (a) and by the Student t-test (b). OD, optical density.

Autoantibodies against nephrin

To test our hypothesis, the sera from cGVHD mice were analysed by nephrin-specific antibody-sandwich ELISA using anti-cNep and anti-eNep. Double antibody-sandwich ELISA revealed antibodies against nephrin in cGVHD model mice (Fig. 3). Both data from precoated plates with anti-cNep and anti-eNep antibodies were very similar; antibodies against nephrin raised significantly from week 2 to week 6, peaking at week 2 or week 4. These results strongly suggest that autoantibodies against nephrin were raised in cGVHD model mice before and during proteinuria development.

Fig. 3.

Serum antibody titres against nephrin determined by antibody-sandwich ELISA using precoated wells with antibodies against (a) intracytoplasmic region and (b) extracellular region of nephrin. Values are represented as the mean ± SEM. Results showing statistically significant differences from the noncGVHD group calculated by the Student t-test; *P < 0·05, **P < 0·0001. OD, optical density.

To confirm the existence of autoantibodies against nephrin, glomerular lysate was immunoprecipitated with anti-cNep antibody, then subjected to SDS-PAGE, transferred to nitrocellulose membrane, and immunoblotted with pooled mouse sera at week 2 and week 4 after purified with ammonium sulphate (Fig. 4). The results showed that the pooled sera from cGVHD mice, not non-cGVHD mice, successfully revealed the band around 185 kD, suggesting nephrin (lane 2 and 3).

Fig. 4.

Immunoblotting analysis of immunoprecipitated nephrin. Incubated with rabbit anti-nephrin antibody (lane a), pooled mouse sera from cGVHD group of week 2 (lane b) and week 4 (lane c), and pooled mouse sera from noncGVHD group of week 2 (lane d).

Morphological studies of kidney specimens

Light microscopy analysis was performed using kidney specimens from both cGVHD and non-cGVHD mice. Our observations using light microscopy showed that mesangial proliferative glomerulonephritis with focal crescent formation was evident up to week 8 (Fig. 5).

Fig. 5.

Representative micrographs of glomeruli from cGVHD model mice from week 0 to week 8, stained with periodic acid-schiff (PAS). No obvious changes were observed until week 6. At week 8, mesangial proliferative glomerulonephritis with focal crescent formation was detectable.

Because antibody-sandwich ELISA suggested the existence of autoantibodies against nephrin, we tested for possible colocalization using double immunofluorescence staining for IgG and nephrin. Confocal microscopic analysis suggested that there were several foci of colocalization of IgG and nephrin along the glomerular capillary walls in cGVHD mice at week 4 and the colocalization became conspicuous at week 6 and week 8 (Fig. 6). These results indicated that immune complexes comprised of IgG and anti-nephrin antibodies were formed along the glomerular basement membrane. No IgG deposition was noted in glomeruli of non-cGVHD group (data not shown).

Fig. 6.

Representative micrographs following confocal microscopic analysis of glomeruli from cGVHD model mice. Green fluorescent signals indicate IgG deposition and red fluorescence represents nephrin staining. Apparent colocalization of both fluorescence along the glomerular capillary walls began on week 4, then foci of colocalization became conspicuous on week 6 and week 8.

Discussion

The antibodies we produced in rabbit showed remarkable specificity with minimal nonspecific background, which was indicated by immunoblotting and immunofluorescent microscopy (Fig. 1). Then we applied antibody-sandwich ELISA and confocal microscopic analysis to detect autoantibodies against nephrin. Initially, we tried to establish conventional anti-nephrin ELISA system. However, we failed to obtain enough amount of purified recombinant nephrin for coating antigen in ELISA, thus we developed sandwich ELISA using rabbit anti-cNep and e-Nep antibodies which recognize different epitopes of nephrin, intracytoplasmic domain and extracellular region, respectively. Applied sera from cGVHD mice successfully developed the positive reaction (Fig. 3), and sera from cGVHD mice at week 2 and week 4 recognized immunoprecipitated nephrin (Fig. 4). Furthermore, colocalization of IgG and nephrin were suggested to begin week 4 and increase toward week 8 with double-staining confocal microscopy (Fig. 6). Positive detection of nephrin and IgG at the same site indicates that IgG revealed by confocal microscopy might direct to the nephrin and form in situ immune complex. This finding was observed from week 4, prior to the development of significant proteinuria at week 7. Increasing in the number of colocalization spots in parallel with light microscopic changes seems to suggest that IgG-immune complex with nephrin in situ has a certain role in the disruption of slit diaphragm to develop proteinuria in cGVHD model mice.

In human, antibodies against nephrin have been already observed. Patrakka et al. [17] detected autoantibodies against nephrin in NPHS1 patients who developed recurrent nephrotic syndrome in renal grafts after transplantation. They found that antibodies reacting against glomeruli in eight of nine patients with recurrence and high anti-nephrin antibody levels in four patients. All of the patients who developed recurrent nephrotic syndrome had a Fin-major/Fin-major genotype, which leads to complete absence of nephrin in the native kidney, thus lack of immunological tolerance to nephrin molecule and immune reaction against nephrin antigen seems to be involved in the pathogenesis of the recurrence. Moreover, injection of monoclonal antibody 5-1-6, directed against the extracellular domain of rat nephrin, causes severe proteinuria in rats within days after injection [18,19]. In this rat model, antibodies obviously penetrate the GBM and reach subepithelial spaces in which nephrin exists. Our present results are consistent with these observations and strongly suggest the pathogenic role of anti-nephrin antibodies in proteinuric state.

The precise mechanism under the development of autoantibodies against nephrin in our model mouse still remains unclear. In murine cGHVD model, polyclonal B cell activation has been demonstrated and are considered to be important for autoantibody formation [20] and additionally IgG class switch is critical for the development of immune complex nephritis [21], thus it would be possible that polyclonal B cell activation might be involved in the anti-nephrin antibody formation and the development of IgG class antibody against nephrin might be critical for the development of glomerulonephritis. Antigen spreading has been suggested to be another mechanism [22]. That is to say, deposited immune complexes or anti-GBM antibodies along the GBM of glomeruli would cause inflammatory injury on podocyte foot processes and slit diaphragm, originally attached urinary space side of the GBM, then lead to spreading of the autoimmune response to components of foot processes by helper T cells specific for GBM components like laminin. This mechanism also might explain further antigenic stimulation to promote IgG-class antibody, which is consistent with our finding of IgG deposition along with nephrin.

The present study provides a basis for understanding the significance of slit diaphragm and their antibodies in autoimmune glomerulonephritis. However, it is still not clear from our experiments whether the nephrin antibodies directly cause disturbance of slit diaphragms to develop lupus-like glomerulonephritis in cGVHD model mice. Actually, our experiments did not show close temporal relationship between anti-nephrin antibody development (week 2) and severe proteinuria development (week 7). It would be speculated that certain amount of antibody deposition is required to develop severe glomerulonephritis. Further studies are required to be aimed at elucidating the pathogenic role of autoantibodies against nephrin and confirming antibody–nephrin interaction with electron microscopy. Transfer of the autoantibodies eluted from sera and glomeruli from model mice into naïve recipients might provide novel insights for the pathogenic role of nephrin. In addition, it remains to be determined specificity and precise epitopes of autoantibodies against nephrin to understand precise mechanism of autoantibody production.