Veterinary Immunology and Immunopathology, 24 (1990) 199-209 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands
199
Development of Monoclonal Antibodies Against Canine Glomerular Antigens* DONALD R. KRAWIEC', PETER J. FELSBURG 2, HOWARD B. GELBERG 2 and STEVEN J. DUGAN 3
'Department o[ Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, 1008 West Hazelwood Drive, Urbana, IL 61801 (U.S.A.) 2Department of Veterinary Pathobiology, College of Veterinary Medicine, University of Illinois, Urbana, IL 61801 (U.S.A.) 3Department of Clinical Services, College of Veterinary Medicine and Biomedical Sciences, Ft. Collins, CO 80523 (U.S.A.) (Accepted 16 August 1989)
ABSTRACT
Krawiec, D.R., Felsburg, P.J., Gelberg, H.B. and Dugan, S.J., 1990. Development of monoclonal antibodies against canine glomerular antigens. Vet. Immunol. Immunopathol., 24: 199-209. Monoclonal antibody producing hybridomas were developed by fusing spleen cells from BALB/ c mice immunized against canine glomeruli with SP2 myeloma cells. Monoclonal antibody reactivity was tested using an indirect immunofluorescence assay on various normal canine tissues and canine kidney affected with glomerulonephritis. Two of the hybridomas developed (3H2 and 3A5) reacted with glomeruli and not with renal tubules. Antibody produced by hybridoma 3A5 also reacted with smooth muscle of all other tissues tested and 3H2 with lung tissue. Antigens recognized by monoclonal antibodies were studied by assessing their heat stability and susceptibility to proteolysis and neuraminidase digestion. Antigen and antibody molecular weights were determined by using a western blotting technique. Glomerular proteins that reacted with antibody produced by hybridoma 3H2 had molecular weights ranging from approximately 92 500 daltons to 200 000 daltons. Antigens reacting with both monoclonal antibodies were likely protein antigens. It was concluded that monoclonal antibodies would be useful in the study of glomerular antigens in normal dogs and dogs with glomerulonephritis.
INTRODUCTION
Immunologically mediated renal disease is implicated in a significant number of renal syndromes in dogs and is a common cause of nephrotic syndrome *Supported in part by the University of Illinois Research Board; Biomedical Research Support Grant RR05460 from the National Institute of Health and the American Veterinary Medicine Association Foundation.
0165-2427/90/$03.50
© 1990 Elsevier Science Publishers B.V.
200
D.R. KRAWIEC ET AL.
and chronic renal failure (Cowgill, 1983; Scott, 1983; Lewis and Center, 1984; Robertson, 1984; Dibartola and Chew, 1986). The precise incidence of immunologically mediated renal disease in dogs is unknown. Numerous investigators have identified immune complexes in canine glomeruli (Lewis and Center, 1984; Robertson, 1984; Dibartola and Chew, 1986). Immune complex glomerular disease has been documented in the dog and is primarily idiopathic, but has been associated with pyometra, heartworm disease, neoplasia and systemic inflammatory conditions (Cowgill, 1983; Dibartola and Chew, 1986). Three clinical syndromes currently recognized in the dog and associated with glomerulonephritis are persistent proteninuria, nephrotic syndrome, and chronic renal failure (Lewis and Center, 1984). Morphologic types of canine glomerulonephritis include mesangioproliferative, membranous, membranoproliferative and chronic end-stage glomerulonephritis (Lewis and Center, 1984; Center et al., 1987). There is considerable discussion regarding the importance and occurrence of the various forms of glomerulonephritis and the underlying pathologic mechanisms which allow glomerular disease to occur (Neale and Wilson, 1982; Cowgill, 1983; Scott, 1983; Voyt, 1984). Characterization of the antigens responsible for glomerulonephritis will be useful in clarifying its pathogenesis. Such studies, however, have been hampered by the lack of specific renal cell identification markers. In 1975, Kohler and Milstein (1975) developeda method which results in the production of monoclonal antibodies directed against single antigenic determinants. Monoclonal antibodies have been raised with specificities to human glomerular basement membranes, epithelial cells, endothelial cells, and connective tissue (Falkenberg et al., 1981; Hancock and Atkins, 1983; Michael et al., 1983; Pressey et al., 1983). The purpose of this project was to develop reagents which would allow the study of canine gtomerular antigens in normal dogs and dogs with glomerulonephropathy. More specifically, the objectives were to produce monoclonal antibodies which are directed against glomerular antigens. METHODSANDMATERIALS
Hybridomaproduction Spleen cells from BALB/c mice (Harlan Sprague Dawley, IN) immunized against canine glomeruli were fused with SP2 myeloma cells (supplied by P.J. Felsburg). Glomeruli used for immunization were isolated from normal kidney using graded sieving (Hancock and Atkins, 1983; Allen and Dowling, 1981). Kidney tissue was first minced using scissors, and then forced through a 60mesh (250 mm ) sieve. The material forced through the sieve was collected and passed through additional sieves with sequentially smaller meshes. Each mesh only retained material that is larger than canine glomeruli until the material was passed through a 120-mesh sieve. This sieve retained glomeruli but allowed any smaller material to pass through. Glomeruli were then collected from the
M O N O C L O N A L ANTIBODIES AGAINST G L O M E R U L A R ANTIGENS
201
sieve for use in immunizing mice. Mice were immunized with 2 × 103 glomeruli by intraperitoneal injection. After 3 weeks, the mice received a secondary intraperitoneal immunization with the same dose of glomeruli. Three days after the second immunization, immunized spleen cells were collected and fused with myeloma cells using standard techniques (Krawiec and Muscoplat, 1984a,b). Fused cells were evenly distributed in two 24-well flat-bottomed culture plates. Media from wells were tested for anti-canine glomerular antibody in 10 to 14 days. Media from wells containing hybridomas were initiallyassayed for antibody, using an enzyme immunofiltration assay, with isolated glomeruli as antigen (Handley et al.,1982). Whole glomeruli were isolated for this procedure, and an aliquot was placed in each well of a 99-well disposable microfold (Isolab, OH). Glomeruli diluted in 0.01 M phosphate-buffered saline containing 1% bovine serum albumin (Sigma, M O ) (BPBS) were added to each well and were washed three times with 0.01 M phosphate-buffered saline containing 0.3% swine skin type I gelatin (Sigma, M O ) (GPBS). Fifty/A of media from wells with growing hybridomas were added to each well containing glomeruli and incubated for 30 rain. The media was suctioned from the wells, and the glomeruli were washed three times with GPBS. A second antibody was then added to each well. This second antibody was horseradish peroxidase labeled rabbit anti-mouse IgG (Cooper, PA). This was left in the wells for 30 rain at room temperature and then suctioned from the wells. The glomeruli were again washed three times with GPBS. Positive wells were identifed using O-phenylenediamine (Sigma, M O ) as substrate. Media from wells without growing hybridomas were used as negative controls,and sera from mice inoculated with canine glomeruli were used as positive controls. Hybridoma cell lines producing antiglomerular antibody as determined by the enzyme immunofiltration assay were cloned two times by limiting dilution and frozen in D M E M + 9 5 % fetal bovine serum (Krawiec and Muscoplat, 1984a,b). Antigen distribution on canine frozen tissue sections Antibody reactivity was tested on the following canine tissues using an indirect immunofluorescent (IFA) assay: normal kidney, kidney tissues affected with glomerulonephritis, liver, spleen, skin, lung, brain, skeletal muscle, and bladder (Allen and Dowling, 1981; Krawiec and Muscoplat, 1984a,b). Tissues were collected fresh, placed in a vial and frozen in liquid nitrogen. The tissue was processed and 6-8/~m sections were made. The IFA were performed using standard techniques (Allen and Dowling, 1981; Krawiec and Muscoplat, 1984a,b). Tissue sections were fixed in acetone and washed with BPBS. Sections were incubated for 30 min at room temperature with 50 #1 of medium in which cloned hybridoma cells were growing. This anti-glomerular antibody was washed from the tissue using BPBS. After washing, the sections were incubated for 30 rain at room temperature with 50/~l of a 1:10 dilution of fluo-
202
D.R. KRAWIEC ET AL.
rescein-conjugated rabbit anti-mouse IgG (F (ab') 2 fragment) (Cooper, PA ). Slides were washed with BPBS, mounted and were examined, using a fluorescent microscope. Media from wells that did not contain growing hybridoma antibody were substituted for anti-glomerular hybridoma antibody for use as negative controls, and serum from mice immunized with canine glomeruli were used as positive controls.
Characterization of antigen Antigens recognized by monoclonal antibodies were studied by assessing their heat stability and susceptibility to proteolysis and neuraminidase digestion (Hancock and Atkins, 1983 ). These procedures were performed on tissue sections followed by immunofluorescent labelling. Heat stability was assessed by boiling kidney tissue sections for 20 min. Susceptibility to proteolysis was determined by incubating kidney tissue sections for 30 min with 0.01% trypsin (Sigma, MO) in 0.1% CaCl2 pH 7.8 and for 10 min with 0.01% pronase ( Sigma, MO) in 0.05 M Tris-HC1 (Sigma, MO) pH 8.0. Finally, tissue sections were incubated with 0.01% and 0.1% neuraminidase (Sigma, MO) in 0.1 M phosphate buffer pH 6.0 at 37°C for 16 h.
Western blotting Antigen and antibody molecular weight determination was performed by using a western blotting technique (Towbin et al., 1979; Burnette, 1981 ). Antigen molecular weight was determined by washing and suspending glomerular cells in a 1% Triton X-100 (Bio-Rad, CA) solution with 50 mMTris-HC1 (BioRad, CA), pH 8.0, which dissolves and solubilizes cellular membranes. Cellular debris was centrifuged out at 45 000 Xg and supernatant with canine glomerular antigens was concentrated and subjected to sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) electrophoresis (Haas and Kennett, 1980). The glomerular antigen was diluted 1 : 1 in sample buffer containing 33% 2mercaptoethanol (Bio-Rad, CA) and boiled for 60 sec (Krawiec and Muscoplat, 1984a,b). A 10% SDS-PAGE was prepared, and the boiled antigen was added to sample wells. The gel was run at a constant current of 30 mA for about 2.5 h. The gel was then used in a western blotting procedure. Proteins in the gel were electrophoretically transferred to nitrocellulose (BioRad, CA) using a western blotting technique (Burnette, 1981). Nonreacting protein binding sites were blocked using nonfat dry milk (Johnson et al., 1984). Glomerular antigens were identified by reacting the nitrocellulose with hybridoma antibody for 3 h. The paper was then washed and reacted with peroxidase conjugated goat anti-mouse IgG (Cooper, PA) for 30 rain. The paper was washed again and developed in O-phenylenediamine for 20-30 min. Molecular weight standards (Bio-Rad, CA) were also electrophoretically transferred to nitrocellulose paper. The area on the nitrocellulose paper containing the molecular weight standards was separated from the rest of glomerular antigen containing paper immediately after the blotting procedure and was stained separately for
MONOCI.DNAL ANTIBODIES AGAINST GLOMERULAR ANTIGENS
203
protein. This was then used to determine the molecular weights of the reacting glomerular antigens.
Antibody subclass Monoclonal antibody subclass was determined by reacting normal kidney tissue frozen sections firstwith hybridoma-derived antiglomerular antibody, and with goat anti-mouse IgGl, IgG2a, IgG2b, IgG3, IgM, IgA, and kappa and lambda light chains (Cooper, PA). The tissue sections were lastlyreacted with fluorescein-conjugated rabbit anti-goat IgG (Cooper, PA). These sections were washed, mounted, and observed for fluorescence using a fluorescent microscope. RESULTS
Hybridoma production Six fusions were performed. Hybridomas developed in approximately 300 wells. Thirty-two of the tested wells were positive for canine renal tissue using the immunofiltration assay. Media from these 32 wells were tested for anticanine renal antibody using an indirect fluorescent antibody test. Six of the hybridomas tested reacted only with canine glomeruli. Two of the hybridomas,
Fig. 1. Section of canine kidney tissue with only a glomerulus staining after incubation with tissue culture media containing antibody 3H2, followed by incubation with fluorescein-conjugated rabbit anti-mouse IgG, X 250.
204
D.R. KRAWIECET AL.
%
Fig. 2. Section of canine kidney tissue with a glomerulus staining after incubation with tissue culture media containing antibody 3A5, followed by incubation with fluorescein-conjugated rabbit anti-mouse IgG, X 250.
which reacted with glomeruli (3H2, 3A5), were chosen for further study because they appeared to be stable in culture and had differing patterns of glomerular fluorescence (Figs. 1 and 2). 3A5 and 3H2 antigen distribution in normal canine tissue Antibody 3A5 reacted with components of renal glomeruli and smooth muscle in all tissues tested (Figs. 2, 3a, and 3b). It stained normal glomeruli in a linear pattern (Fig. 2). In glomeruli affected with membranoproliferative glomerulonephritis it also stained Bowmans capsule (Fig. 4). Antibody 3H2 reacted only with kidney and lung. It reacted diffusely with the entire normal glomerulus and its staining characteristics did not change in dogs affected with membranoproliferative glomerulonephritis. It reacted with lung tissue alveoli in a linear pattern and did not react with bronchioles or blood vessels. The intensity of fluorescence of lung was less than in the glomerulus. Neither antibody reacted with renal tubules. Characterization of antigen Boiling tissue sections completely abolished antibody 3A5 labelling and staining but did not affect antibody 3H2. Trypsin digestion had no effect on
MONOCLONAL ANTIBODIES AGAINST GLOMERULAR ANTIGENS
205
Fig. 3a. Section of canine bladder with smooth muscle staining afterincubation with tissueculture media containing antibody 3A5, followed by incubation with fluorescein-conjugatedrabbit antimouse IgG, × 250.
Fig. 3b. T h e same area of canine bladder as 3. Stained with H & E.
206
D.R. KRAWIEC ET AL.
Fig. 4. Section of canine kidney tissue affected with membranoproliferative glomerulonephritis with Bowmans capsule staining after incubation with tissue culture media containing antibody 3A5, followed by incubation with fluorescein-conjugated rabbit anti-mouse IgG, × 400.
la.
lb,
a ~
I
f
Fig. 5. Photograph of western blot with glomerular antigen identified by antibody 3H2. la=molecular weight standards; a=myosin (200000 D); b=fl-galactosidase (116250 D); c = phosphorylase fl (92 500 D ); d = bovine serum albumin ( 66 200 D ); e = ovalbumin (45 000 D ); lb = glomerular antigen blot; f= glomerular antigen.
MONOCLONAL ANTIBODIES AGAINST GLOMERULAR ANTIGENS
207
antibody 3A5 staining but completely abolished antibody 3H2 staining. Pronase digestion resulted in only a slight decrease in intensity of labelling with both antibodies, and neuraminidase treatment had no effect.
Western blotting 3H2 antigen appeared to have four major peptide components. The glomerular proteins that reacted with this antibody had molecular weights between 92 500 and 200 000 daltons (Fig. 5). 3A5 antigen was not identifiable using this technique.
Hybridoma antibody class and subclass Both antibodies were identified as immunoglobulin subclass IgG1 with kappa light chains. DISCUSSION In this study, whole canine glomeruli were used to develop monoclonal antibodies against glomerular antigens. Antibodies 3A5 and 3H2 developed by this technique reacted specifically with glomerular components but not with renal tubules. 3A5 reacted in a linear pattern within the glomerulus. It also reacted with smooth muscle of other tissues tested. This would lead to the speculation that 3A5 is reacting to the smooth muscle of arterioles within the glomerulus or the basement membrane of both. 3H2 reacted in a more diffuse pattern within the giomerulus. Further studies using immunoelectronmicroscopy or thin sections will be necessary to determine the precise component within the glomerulus these antibodies are identifying. 3H2 did not react in a linear pattern. It is, therefore, unlikely reacting with the glomerular basement membrane. It could possibly be reacting to a component in the glomerular mesangium. It also reacted with lung tissue. The staining was much less intense with the lung, but it stained more than controls. The exact location of the antigen with the alveoli or the significance of this cross reactivity is unknown. The peptide antigens identified by 3H2 are high molecular weight ranging from 200 000 daltons to 92 500 daltons. Antibody 3H2 may be reacting to more than one protein, or the glomerular protein may have been broken into subunits as a result of the processing for use in the SDS-PAGE electrophoresis. The antigen reacting with antibody 3A5 was not identified by the technique used in this study. Preparation of antigens for SDS-PAGE requires boiling. This likely denatured the antigen making it unable to react with antibody 3A5. Further studies using nondenaturing conditions will be required to identify the molecular weight of this antigen. The fact that 3A5 antigen is susceptible to boiling and 3H2 antigen is sus-
208
D.R. KRAWIEC ET AL.
ceptible to endopeptidase (trypsin) digestion would suggest the antibodies are reacting to protein antigens. Their resistance to neuraminidase would indicate t h a t they are not sialoproteins. Sialoproteins have been reported to cover podocytes within the glomerulus and apparently play a role in glomerular filtration. Studies using monoclonal antibodies reactive with canine glomerular antigens may provide significant information on the pathogenesis of immune complex glomerulonephritis. Monoclonal antibodies can, for example, be used to identify the proliferating cells in proliferative glomerulonephritis. These cells are often difficult to identify histologically. Other organs can be evaluated for the presence of these antigens thus aiding the search for the source of offending antigens in idiopathic glomerulonephritis. Also, defining changes in the distribution, amount, and characteristics of normal renal antigens in glomerulonephritis may aid in defining its pathogenesis. Only by understanding the type, amount, and distribution of normal renal antigens and cells can we begin to understand and study changes associated with renal diseases. Characterizing glomerulonephritis in this way may also prove to be important in clarifying pathologic mechanisms, assessing prognosis, and instituting therapy.
REFERENCES Allen, D.E. and Dowling,J.P., 1981. Immunofluorescenceand immunoperoxidaseprocedures. In: D.E. Allen and J.P. Dowling (Editors), Techniques for Nephropathology.CRC Press, Boca Raton, FL, pp. 73-78. Burnette, W.N., 1981. Western blotting electrophoretictransfer of proteins from sodium dodecyl sulfate-plyacrylamidegels to unmodifiednitrocelluloseand radiographic detection with antibody and radioiodinatedprotein A. Anal. Chem., 112: 195-203. Center, S.A., Smith, C.A., Wilkinson, E., Erb, H.N. and Lewis, R.M., 1987. Clinicopathologic, renal immunofluorescent,and light microscopicfeatures of glomerulonephritisin the dog: 41 cases (1975-1985). J. Am. Vet. Med. Assoc., 190: 81-90. Cowgill, L.D., 1983. Diseases of the kidney. In: S.J. Ettinger {Editor), Textbook of Veterinary Internal Medicine,W.B. Saunders, Philadelphia, PA, pp. 1793-1878. Dibartola, S.P. and Chew, D.J., 1986. Glomerulardisease in the dog and cat. In: R.W. Kirk (Editor), Current VeterinaryTherapy IX, W.B. Saunders, Philadelphia, PA, pp. 1132-1138. Falkenberg, F.W., Muller, E. and Riffiemann,H., 1981. The production of monoclonalantibodies against glomerular and other antigens of the human nephron. Renal Physiol. Base., 4: 150156. Haas, J.B. and Kennett, R.H., 1980. Characterization of hybridoma immunoglobulinby sodium dodecylsulfate-polyacrylamidegel electrophoresis.In: R.H. Kennett, T.J. McKearn,and K.B. Bechtol (Editors), MonoclonalAntibodies. Plenum Press, New York, NY, pp. 406-411. Handley, H.H., Glassy, M.C., Cleveland, P.H. and Royston, I., 1982. Development of a rapid microELISAassay for screening hybridoma supernatants for murine monoclonalantibodies. J. Immunol. Methods,54: 291-296. Hancock, W.W. and Atkins, R.C., 1983. Monoclonalantibodies to human glomerulonephritis:A marker for glomerularepithelial cells. Nephron, 33: 83-90. Johnson, D.A., Gautsch, J.W., Sportsman, J.R. and Elder, J.H., 1980. Improvedtechnique utiliz-
MONOCLONALANTIBODIESAGAINSTGLOMERULARANTIGENS
209
ing nonfat dry milk for analysis of proteins and nucleic acids transferred to nitrocellulose. Gene Anal. Techn., 1: 3-8. Kohler, G. and Milstein, C., 1975. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature, 256: 495-497. Krawiec, D.R. and Muscoplat, C.C., 1984a. Development and characterization of a hybridomaderived antibody (Aby 1A1 ) with specificity to canine thymocytes and peripheral T lymphocytes. Am. J. Vet. Res., 45: 491-498. Krawiec, D.R. and Muscoplat, C.C., 1984b. Development and characterization of a monoclonal antibody (Aby 6C6) that distinguishes medullary from cortical thymocytes. Am. J. Vet. Res., 45: 499-505. Lewis, R.M. and Center, S.A., 1984. Primary diseases affecting glomerulus. In: K.C. Bovee {Editor), Canine Nephrology. Harwell Publishing Co., Media, PA, pp. 461-480. Michael, A.F., Yang, J.W., Falk, R.J., Bennington, M.J., Scheinman, J.I., Vernier, R.L. and Fish, A.S., 1983. Monoclonal antibodies to human renal basement membranes: Heterogenic and ontogenic changes. Kidney Intern., 24: 77-88. Neale, T.J. and Wilson, C.B., 1982. Glomerular antigens in glomerulonephritis. Springer Semin. Immunopathol., 5: 221-249. Pressey, A., Pusey, C.D., Dash, A., Peters, D.K. and Lockwood, C.M., 1983. Production ofa monoclonal antibody to autoantigenic components of human glomerular basement membrane. Clin. Exp. Immunol., 54: 178-194. Robertson, J.L., 1984. Immunologic injury to the kidney and the renal response. In: K.C. Bovee (Editor), Canine Nephrology. Harwell Publishing, Media, PA, pp. 439-460. Scott, R.C., 1983. Immune-mediated renal disease. In: R.W. Kirk (Editor), Current Veterinary Therapy VIII, W.B. Saunders, Philadelphia, PA, pp. 966-970. Towbin, H., Staehelin, T. and Gordon, J., 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci., 4350-4354. Voyt, A., 1984. New aspects of the pathogenesis of immune complex glomerulonephritis: Formation of subepithelial deposits. Clin. Nephrol., 21: 15-20.
199
Development of Monoclonal Antibodies Against Canine Glomerular Antigens* DONALD R. KRAWIEC', PETER J. FELSBURG 2, HOWARD B. GELBERG 2 and STEVEN J. DUGAN 3
'Department o[ Veterinary Clinical Medicine, College of Veterinary Medicine, University of Illinois, 1008 West Hazelwood Drive, Urbana, IL 61801 (U.S.A.) 2Department of Veterinary Pathobiology, College of Veterinary Medicine, University of Illinois, Urbana, IL 61801 (U.S.A.) 3Department of Clinical Services, College of Veterinary Medicine and Biomedical Sciences, Ft. Collins, CO 80523 (U.S.A.) (Accepted 16 August 1989)
ABSTRACT
Krawiec, D.R., Felsburg, P.J., Gelberg, H.B. and Dugan, S.J., 1990. Development of monoclonal antibodies against canine glomerular antigens. Vet. Immunol. Immunopathol., 24: 199-209. Monoclonal antibody producing hybridomas were developed by fusing spleen cells from BALB/ c mice immunized against canine glomeruli with SP2 myeloma cells. Monoclonal antibody reactivity was tested using an indirect immunofluorescence assay on various normal canine tissues and canine kidney affected with glomerulonephritis. Two of the hybridomas developed (3H2 and 3A5) reacted with glomeruli and not with renal tubules. Antibody produced by hybridoma 3A5 also reacted with smooth muscle of all other tissues tested and 3H2 with lung tissue. Antigens recognized by monoclonal antibodies were studied by assessing their heat stability and susceptibility to proteolysis and neuraminidase digestion. Antigen and antibody molecular weights were determined by using a western blotting technique. Glomerular proteins that reacted with antibody produced by hybridoma 3H2 had molecular weights ranging from approximately 92 500 daltons to 200 000 daltons. Antigens reacting with both monoclonal antibodies were likely protein antigens. It was concluded that monoclonal antibodies would be useful in the study of glomerular antigens in normal dogs and dogs with glomerulonephritis.
INTRODUCTION
Immunologically mediated renal disease is implicated in a significant number of renal syndromes in dogs and is a common cause of nephrotic syndrome *Supported in part by the University of Illinois Research Board; Biomedical Research Support Grant RR05460 from the National Institute of Health and the American Veterinary Medicine Association Foundation.
0165-2427/90/$03.50
© 1990 Elsevier Science Publishers B.V.
200
D.R. KRAWIEC ET AL.
and chronic renal failure (Cowgill, 1983; Scott, 1983; Lewis and Center, 1984; Robertson, 1984; Dibartola and Chew, 1986). The precise incidence of immunologically mediated renal disease in dogs is unknown. Numerous investigators have identified immune complexes in canine glomeruli (Lewis and Center, 1984; Robertson, 1984; Dibartola and Chew, 1986). Immune complex glomerular disease has been documented in the dog and is primarily idiopathic, but has been associated with pyometra, heartworm disease, neoplasia and systemic inflammatory conditions (Cowgill, 1983; Dibartola and Chew, 1986). Three clinical syndromes currently recognized in the dog and associated with glomerulonephritis are persistent proteninuria, nephrotic syndrome, and chronic renal failure (Lewis and Center, 1984). Morphologic types of canine glomerulonephritis include mesangioproliferative, membranous, membranoproliferative and chronic end-stage glomerulonephritis (Lewis and Center, 1984; Center et al., 1987). There is considerable discussion regarding the importance and occurrence of the various forms of glomerulonephritis and the underlying pathologic mechanisms which allow glomerular disease to occur (Neale and Wilson, 1982; Cowgill, 1983; Scott, 1983; Voyt, 1984). Characterization of the antigens responsible for glomerulonephritis will be useful in clarifying its pathogenesis. Such studies, however, have been hampered by the lack of specific renal cell identification markers. In 1975, Kohler and Milstein (1975) developeda method which results in the production of monoclonal antibodies directed against single antigenic determinants. Monoclonal antibodies have been raised with specificities to human glomerular basement membranes, epithelial cells, endothelial cells, and connective tissue (Falkenberg et al., 1981; Hancock and Atkins, 1983; Michael et al., 1983; Pressey et al., 1983). The purpose of this project was to develop reagents which would allow the study of canine gtomerular antigens in normal dogs and dogs with glomerulonephropathy. More specifically, the objectives were to produce monoclonal antibodies which are directed against glomerular antigens. METHODSANDMATERIALS
Hybridomaproduction Spleen cells from BALB/c mice (Harlan Sprague Dawley, IN) immunized against canine glomeruli were fused with SP2 myeloma cells (supplied by P.J. Felsburg). Glomeruli used for immunization were isolated from normal kidney using graded sieving (Hancock and Atkins, 1983; Allen and Dowling, 1981). Kidney tissue was first minced using scissors, and then forced through a 60mesh (250 mm ) sieve. The material forced through the sieve was collected and passed through additional sieves with sequentially smaller meshes. Each mesh only retained material that is larger than canine glomeruli until the material was passed through a 120-mesh sieve. This sieve retained glomeruli but allowed any smaller material to pass through. Glomeruli were then collected from the
M O N O C L O N A L ANTIBODIES AGAINST G L O M E R U L A R ANTIGENS
201
sieve for use in immunizing mice. Mice were immunized with 2 × 103 glomeruli by intraperitoneal injection. After 3 weeks, the mice received a secondary intraperitoneal immunization with the same dose of glomeruli. Three days after the second immunization, immunized spleen cells were collected and fused with myeloma cells using standard techniques (Krawiec and Muscoplat, 1984a,b). Fused cells were evenly distributed in two 24-well flat-bottomed culture plates. Media from wells were tested for anti-canine glomerular antibody in 10 to 14 days. Media from wells containing hybridomas were initiallyassayed for antibody, using an enzyme immunofiltration assay, with isolated glomeruli as antigen (Handley et al.,1982). Whole glomeruli were isolated for this procedure, and an aliquot was placed in each well of a 99-well disposable microfold (Isolab, OH). Glomeruli diluted in 0.01 M phosphate-buffered saline containing 1% bovine serum albumin (Sigma, M O ) (BPBS) were added to each well and were washed three times with 0.01 M phosphate-buffered saline containing 0.3% swine skin type I gelatin (Sigma, M O ) (GPBS). Fifty/A of media from wells with growing hybridomas were added to each well containing glomeruli and incubated for 30 rain. The media was suctioned from the wells, and the glomeruli were washed three times with GPBS. A second antibody was then added to each well. This second antibody was horseradish peroxidase labeled rabbit anti-mouse IgG (Cooper, PA). This was left in the wells for 30 rain at room temperature and then suctioned from the wells. The glomeruli were again washed three times with GPBS. Positive wells were identifed using O-phenylenediamine (Sigma, M O ) as substrate. Media from wells without growing hybridomas were used as negative controls,and sera from mice inoculated with canine glomeruli were used as positive controls. Hybridoma cell lines producing antiglomerular antibody as determined by the enzyme immunofiltration assay were cloned two times by limiting dilution and frozen in D M E M + 9 5 % fetal bovine serum (Krawiec and Muscoplat, 1984a,b). Antigen distribution on canine frozen tissue sections Antibody reactivity was tested on the following canine tissues using an indirect immunofluorescent (IFA) assay: normal kidney, kidney tissues affected with glomerulonephritis, liver, spleen, skin, lung, brain, skeletal muscle, and bladder (Allen and Dowling, 1981; Krawiec and Muscoplat, 1984a,b). Tissues were collected fresh, placed in a vial and frozen in liquid nitrogen. The tissue was processed and 6-8/~m sections were made. The IFA were performed using standard techniques (Allen and Dowling, 1981; Krawiec and Muscoplat, 1984a,b). Tissue sections were fixed in acetone and washed with BPBS. Sections were incubated for 30 min at room temperature with 50 #1 of medium in which cloned hybridoma cells were growing. This anti-glomerular antibody was washed from the tissue using BPBS. After washing, the sections were incubated for 30 rain at room temperature with 50/~l of a 1:10 dilution of fluo-
202
D.R. KRAWIEC ET AL.
rescein-conjugated rabbit anti-mouse IgG (F (ab') 2 fragment) (Cooper, PA ). Slides were washed with BPBS, mounted and were examined, using a fluorescent microscope. Media from wells that did not contain growing hybridoma antibody were substituted for anti-glomerular hybridoma antibody for use as negative controls, and serum from mice immunized with canine glomeruli were used as positive controls.
Characterization of antigen Antigens recognized by monoclonal antibodies were studied by assessing their heat stability and susceptibility to proteolysis and neuraminidase digestion (Hancock and Atkins, 1983 ). These procedures were performed on tissue sections followed by immunofluorescent labelling. Heat stability was assessed by boiling kidney tissue sections for 20 min. Susceptibility to proteolysis was determined by incubating kidney tissue sections for 30 min with 0.01% trypsin (Sigma, MO) in 0.1% CaCl2 pH 7.8 and for 10 min with 0.01% pronase ( Sigma, MO) in 0.05 M Tris-HC1 (Sigma, MO) pH 8.0. Finally, tissue sections were incubated with 0.01% and 0.1% neuraminidase (Sigma, MO) in 0.1 M phosphate buffer pH 6.0 at 37°C for 16 h.
Western blotting Antigen and antibody molecular weight determination was performed by using a western blotting technique (Towbin et al., 1979; Burnette, 1981 ). Antigen molecular weight was determined by washing and suspending glomerular cells in a 1% Triton X-100 (Bio-Rad, CA) solution with 50 mMTris-HC1 (BioRad, CA), pH 8.0, which dissolves and solubilizes cellular membranes. Cellular debris was centrifuged out at 45 000 Xg and supernatant with canine glomerular antigens was concentrated and subjected to sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) electrophoresis (Haas and Kennett, 1980). The glomerular antigen was diluted 1 : 1 in sample buffer containing 33% 2mercaptoethanol (Bio-Rad, CA) and boiled for 60 sec (Krawiec and Muscoplat, 1984a,b). A 10% SDS-PAGE was prepared, and the boiled antigen was added to sample wells. The gel was run at a constant current of 30 mA for about 2.5 h. The gel was then used in a western blotting procedure. Proteins in the gel were electrophoretically transferred to nitrocellulose (BioRad, CA) using a western blotting technique (Burnette, 1981). Nonreacting protein binding sites were blocked using nonfat dry milk (Johnson et al., 1984). Glomerular antigens were identified by reacting the nitrocellulose with hybridoma antibody for 3 h. The paper was then washed and reacted with peroxidase conjugated goat anti-mouse IgG (Cooper, PA) for 30 rain. The paper was washed again and developed in O-phenylenediamine for 20-30 min. Molecular weight standards (Bio-Rad, CA) were also electrophoretically transferred to nitrocellulose paper. The area on the nitrocellulose paper containing the molecular weight standards was separated from the rest of glomerular antigen containing paper immediately after the blotting procedure and was stained separately for
MONOCI.DNAL ANTIBODIES AGAINST GLOMERULAR ANTIGENS
203
protein. This was then used to determine the molecular weights of the reacting glomerular antigens.
Antibody subclass Monoclonal antibody subclass was determined by reacting normal kidney tissue frozen sections firstwith hybridoma-derived antiglomerular antibody, and with goat anti-mouse IgGl, IgG2a, IgG2b, IgG3, IgM, IgA, and kappa and lambda light chains (Cooper, PA). The tissue sections were lastlyreacted with fluorescein-conjugated rabbit anti-goat IgG (Cooper, PA). These sections were washed, mounted, and observed for fluorescence using a fluorescent microscope. RESULTS
Hybridoma production Six fusions were performed. Hybridomas developed in approximately 300 wells. Thirty-two of the tested wells were positive for canine renal tissue using the immunofiltration assay. Media from these 32 wells were tested for anticanine renal antibody using an indirect fluorescent antibody test. Six of the hybridomas tested reacted only with canine glomeruli. Two of the hybridomas,
Fig. 1. Section of canine kidney tissue with only a glomerulus staining after incubation with tissue culture media containing antibody 3H2, followed by incubation with fluorescein-conjugated rabbit anti-mouse IgG, X 250.
204
D.R. KRAWIECET AL.
%
Fig. 2. Section of canine kidney tissue with a glomerulus staining after incubation with tissue culture media containing antibody 3A5, followed by incubation with fluorescein-conjugated rabbit anti-mouse IgG, X 250.
which reacted with glomeruli (3H2, 3A5), were chosen for further study because they appeared to be stable in culture and had differing patterns of glomerular fluorescence (Figs. 1 and 2). 3A5 and 3H2 antigen distribution in normal canine tissue Antibody 3A5 reacted with components of renal glomeruli and smooth muscle in all tissues tested (Figs. 2, 3a, and 3b). It stained normal glomeruli in a linear pattern (Fig. 2). In glomeruli affected with membranoproliferative glomerulonephritis it also stained Bowmans capsule (Fig. 4). Antibody 3H2 reacted only with kidney and lung. It reacted diffusely with the entire normal glomerulus and its staining characteristics did not change in dogs affected with membranoproliferative glomerulonephritis. It reacted with lung tissue alveoli in a linear pattern and did not react with bronchioles or blood vessels. The intensity of fluorescence of lung was less than in the glomerulus. Neither antibody reacted with renal tubules. Characterization of antigen Boiling tissue sections completely abolished antibody 3A5 labelling and staining but did not affect antibody 3H2. Trypsin digestion had no effect on
MONOCLONAL ANTIBODIES AGAINST GLOMERULAR ANTIGENS
205
Fig. 3a. Section of canine bladder with smooth muscle staining afterincubation with tissueculture media containing antibody 3A5, followed by incubation with fluorescein-conjugatedrabbit antimouse IgG, × 250.
Fig. 3b. T h e same area of canine bladder as 3. Stained with H & E.
206
D.R. KRAWIEC ET AL.
Fig. 4. Section of canine kidney tissue affected with membranoproliferative glomerulonephritis with Bowmans capsule staining after incubation with tissue culture media containing antibody 3A5, followed by incubation with fluorescein-conjugated rabbit anti-mouse IgG, × 400.
la.
lb,
a ~
I
f
Fig. 5. Photograph of western blot with glomerular antigen identified by antibody 3H2. la=molecular weight standards; a=myosin (200000 D); b=fl-galactosidase (116250 D); c = phosphorylase fl (92 500 D ); d = bovine serum albumin ( 66 200 D ); e = ovalbumin (45 000 D ); lb = glomerular antigen blot; f= glomerular antigen.
MONOCLONAL ANTIBODIES AGAINST GLOMERULAR ANTIGENS
207
antibody 3A5 staining but completely abolished antibody 3H2 staining. Pronase digestion resulted in only a slight decrease in intensity of labelling with both antibodies, and neuraminidase treatment had no effect.
Western blotting 3H2 antigen appeared to have four major peptide components. The glomerular proteins that reacted with this antibody had molecular weights between 92 500 and 200 000 daltons (Fig. 5). 3A5 antigen was not identifiable using this technique.
Hybridoma antibody class and subclass Both antibodies were identified as immunoglobulin subclass IgG1 with kappa light chains. DISCUSSION In this study, whole canine glomeruli were used to develop monoclonal antibodies against glomerular antigens. Antibodies 3A5 and 3H2 developed by this technique reacted specifically with glomerular components but not with renal tubules. 3A5 reacted in a linear pattern within the glomerulus. It also reacted with smooth muscle of other tissues tested. This would lead to the speculation that 3A5 is reacting to the smooth muscle of arterioles within the glomerulus or the basement membrane of both. 3H2 reacted in a more diffuse pattern within the giomerulus. Further studies using immunoelectronmicroscopy or thin sections will be necessary to determine the precise component within the glomerulus these antibodies are identifying. 3H2 did not react in a linear pattern. It is, therefore, unlikely reacting with the glomerular basement membrane. It could possibly be reacting to a component in the glomerular mesangium. It also reacted with lung tissue. The staining was much less intense with the lung, but it stained more than controls. The exact location of the antigen with the alveoli or the significance of this cross reactivity is unknown. The peptide antigens identified by 3H2 are high molecular weight ranging from 200 000 daltons to 92 500 daltons. Antibody 3H2 may be reacting to more than one protein, or the glomerular protein may have been broken into subunits as a result of the processing for use in the SDS-PAGE electrophoresis. The antigen reacting with antibody 3A5 was not identified by the technique used in this study. Preparation of antigens for SDS-PAGE requires boiling. This likely denatured the antigen making it unable to react with antibody 3A5. Further studies using nondenaturing conditions will be required to identify the molecular weight of this antigen. The fact that 3A5 antigen is susceptible to boiling and 3H2 antigen is sus-
208
D.R. KRAWIEC ET AL.
ceptible to endopeptidase (trypsin) digestion would suggest the antibodies are reacting to protein antigens. Their resistance to neuraminidase would indicate t h a t they are not sialoproteins. Sialoproteins have been reported to cover podocytes within the glomerulus and apparently play a role in glomerular filtration. Studies using monoclonal antibodies reactive with canine glomerular antigens may provide significant information on the pathogenesis of immune complex glomerulonephritis. Monoclonal antibodies can, for example, be used to identify the proliferating cells in proliferative glomerulonephritis. These cells are often difficult to identify histologically. Other organs can be evaluated for the presence of these antigens thus aiding the search for the source of offending antigens in idiopathic glomerulonephritis. Also, defining changes in the distribution, amount, and characteristics of normal renal antigens in glomerulonephritis may aid in defining its pathogenesis. Only by understanding the type, amount, and distribution of normal renal antigens and cells can we begin to understand and study changes associated with renal diseases. Characterizing glomerulonephritis in this way may also prove to be important in clarifying pathologic mechanisms, assessing prognosis, and instituting therapy.
REFERENCES Allen, D.E. and Dowling,J.P., 1981. Immunofluorescenceand immunoperoxidaseprocedures. In: D.E. Allen and J.P. Dowling (Editors), Techniques for Nephropathology.CRC Press, Boca Raton, FL, pp. 73-78. Burnette, W.N., 1981. Western blotting electrophoretictransfer of proteins from sodium dodecyl sulfate-plyacrylamidegels to unmodifiednitrocelluloseand radiographic detection with antibody and radioiodinatedprotein A. Anal. Chem., 112: 195-203. Center, S.A., Smith, C.A., Wilkinson, E., Erb, H.N. and Lewis, R.M., 1987. Clinicopathologic, renal immunofluorescent,and light microscopicfeatures of glomerulonephritisin the dog: 41 cases (1975-1985). J. Am. Vet. Med. Assoc., 190: 81-90. Cowgill, L.D., 1983. Diseases of the kidney. In: S.J. Ettinger {Editor), Textbook of Veterinary Internal Medicine,W.B. Saunders, Philadelphia, PA, pp. 1793-1878. Dibartola, S.P. and Chew, D.J., 1986. Glomerulardisease in the dog and cat. In: R.W. Kirk (Editor), Current VeterinaryTherapy IX, W.B. Saunders, Philadelphia, PA, pp. 1132-1138. Falkenberg, F.W., Muller, E. and Riffiemann,H., 1981. The production of monoclonalantibodies against glomerular and other antigens of the human nephron. Renal Physiol. Base., 4: 150156. Haas, J.B. and Kennett, R.H., 1980. Characterization of hybridoma immunoglobulinby sodium dodecylsulfate-polyacrylamidegel electrophoresis.In: R.H. Kennett, T.J. McKearn,and K.B. Bechtol (Editors), MonoclonalAntibodies. Plenum Press, New York, NY, pp. 406-411. Handley, H.H., Glassy, M.C., Cleveland, P.H. and Royston, I., 1982. Development of a rapid microELISAassay for screening hybridoma supernatants for murine monoclonalantibodies. J. Immunol. Methods,54: 291-296. Hancock, W.W. and Atkins, R.C., 1983. Monoclonalantibodies to human glomerulonephritis:A marker for glomerularepithelial cells. Nephron, 33: 83-90. Johnson, D.A., Gautsch, J.W., Sportsman, J.R. and Elder, J.H., 1980. Improvedtechnique utiliz-
MONOCLONALANTIBODIESAGAINSTGLOMERULARANTIGENS
209
ing nonfat dry milk for analysis of proteins and nucleic acids transferred to nitrocellulose. Gene Anal. Techn., 1: 3-8. Kohler, G. and Milstein, C., 1975. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature, 256: 495-497. Krawiec, D.R. and Muscoplat, C.C., 1984a. Development and characterization of a hybridomaderived antibody (Aby 1A1 ) with specificity to canine thymocytes and peripheral T lymphocytes. Am. J. Vet. Res., 45: 491-498. Krawiec, D.R. and Muscoplat, C.C., 1984b. Development and characterization of a monoclonal antibody (Aby 6C6) that distinguishes medullary from cortical thymocytes. Am. J. Vet. Res., 45: 499-505. Lewis, R.M. and Center, S.A., 1984. Primary diseases affecting glomerulus. In: K.C. Bovee {Editor), Canine Nephrology. Harwell Publishing Co., Media, PA, pp. 461-480. Michael, A.F., Yang, J.W., Falk, R.J., Bennington, M.J., Scheinman, J.I., Vernier, R.L. and Fish, A.S., 1983. Monoclonal antibodies to human renal basement membranes: Heterogenic and ontogenic changes. Kidney Intern., 24: 77-88. Neale, T.J. and Wilson, C.B., 1982. Glomerular antigens in glomerulonephritis. Springer Semin. Immunopathol., 5: 221-249. Pressey, A., Pusey, C.D., Dash, A., Peters, D.K. and Lockwood, C.M., 1983. Production ofa monoclonal antibody to autoantigenic components of human glomerular basement membrane. Clin. Exp. Immunol., 54: 178-194. Robertson, J.L., 1984. Immunologic injury to the kidney and the renal response. In: K.C. Bovee (Editor), Canine Nephrology. Harwell Publishing, Media, PA, pp. 439-460. Scott, R.C., 1983. Immune-mediated renal disease. In: R.W. Kirk (Editor), Current Veterinary Therapy VIII, W.B. Saunders, Philadelphia, PA, pp. 966-970. Towbin, H., Staehelin, T. and Gordon, J., 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci., 4350-4354. Voyt, A., 1984. New aspects of the pathogenesis of immune complex glomerulonephritis: Formation of subepithelial deposits. Clin. Nephrol., 21: 15-20.
