Virchows Archiv B Cell Pathol (1992) 62:179-188
V'mehowsArchivB Cell Pathology InclmtinfMolecularPatlwlo~ 9 Springer-Verlag 1992
Intraglomerular fibronectin in rat experimental glomerulonephritis Takashi Yoneyama, Mitsumasa Nagase, Mitsuru Ikeya, Akira Hishida, and Nishio Honda First Department of Medicine,HamamatsuUniversitySchoolof Medicine, Hamamatsu,and First Department of Medicine, Teikyo UniversitySchoolof Medicine,Tokyo,Japan Received November 4, 1991 / Accepted April 28, 1992
Summary. To clarify the mechanisms of glomerular pericapillary fibronectin deposition in human membranous nephropathy and mesangial proliferative glomerulonephritis, intraglomerular fibronectin distribution was examined by light and electron microscopy using the experimental rat models of Heymann and nephrotoxic serum nephritis. As previously demonstrated by immunofluorescence microscopy (Pettersson and Colvin 1978; Ikeya et al. 1985, 1986), fibronectin was distributed in the mesangial areas and occasionally on percicapillary walls of normal glomeruli, while in nephrotoxic serum nephritis and Heymann nephritis, fibronectin was diffusely located along glomerular capillary walls as well as in the mesangium. By immunoelectron microscopy using the immunogold technique, fibronectin was also noted in the mesangial areas and the lamina densa of the glomerular basement membrane (GBM) in normal glomeruli. In nephrotoxic serum nephritis, fibronectin was seen around mesangial cells situated between endothelial cells and the GBM, suggesting that pericapillary fibronectin in nephrotoxic serum nephritis reflects mesangial extension. However, in Heymann nephritis, it was found uniformly in the lamina rara interna, lamina densa and lamina rara externa of the GBM, indicating no specific relation to glomerular cells. When sections of normal and both experimental nephritis kidneys were incubated with fluorescein isothiocyanate conjugated with rat plasma fibronectin, a linear pattern of fluorescein staining along the glomerular capillary walls was observed in Heymann nephritis but not in normal or nephrotoxic serum nephritic rats. The GBM in Heymann nephritis would thus appear to have an affinity for plasma fibronectin. Based on the above findings, fibronectin in the GBM of rats with Heymann nephritis may reasonably be concluded to originate from the plasma.
Correspondence to: T. Yoneyama,First Department of Medicine, Hamamatsu UniversitySchool of Medicine, 3600 Handa-cho, Hamamatsu, 431-31, Japan
Key words: Fibronectin - Heymann nephritis - Nephrotoxic serum nephritis - Immunohistochemistry
Introduction Fibronectin, a large molecular adhesive glycoprotein, is distributed in various tissues, on cellular surfaces, collagenous tissues and in the plasma (Yamada 1978; Hynes and Yamada 1982). By immunofluorescence microscopy, fibronectin has been located predominantly in the mesangial areas in normal glomeruli but is found along capillary walls in diseased glomeruli (Pettersson and Colvin 1978; Weiss et al. 1979; Ikeya et al. 1985, 1986). In our previous study (Ikeya t al. 1985, 1986) using immunofluorescence microscopy, fibronectin was seen to extend from the mesangium to the pericapillary region and along capillary loops with progressive mesangial proliferation. In membranous nephropathy, however, pericapillary extension of fibronectin occurred despite the absence of mesangial proliferation. The present study was conducted to determine the pericapillary distribution of fibronectin at the ultrastructural level in Heymann nephritis and to clarify the mechanisms of deposition of this glycoprotein.
Materials and methods Animals. Female Sprague Dawley rats, weighting 150-200 g, were
used. The animals were allowed free access to a standard pellet diet and water throughout the study. Heymann nephritis. The insoluble rat renal tubular epithelial frac-
tion, F • 1A, was prepared from renal cortices of Sprague Dawley rats as described by Edgington et al. (1976). Heymann nephritis was induced in five rats, each immunized four times at weekly intervals with 20 mg of F x 1A emulsifiedin 0.5 ml of 0.9% saline and conjugated with an equal volume of Freund's complete adjuvant. Followingthe initial immunization,24-h urine samples were collected once a week. The animals were sacrified 20-45 weeks after the initial immunization.
180
Fig. 1. Light and electron micrographs of glomeruli from rats with Heymann nephritis rats (A and B) and with nephrotoxic serum nephritis 14 days after duck anti-rat GBM serum injection (C and D). In Heymann nephritis, spikes along the capillary walls are seen by light microscopy. Intraglomerular hypercellularity is not evident (A, periodic acid-silver methenamine stain). By electron miroscopy, subepithelial electron dense deposits and thickening of GBM are
Nephrotoxic serum nephritb. Normal Sprague Dawley rat kidneys were extensively perfused with 0.9% saline. The glomeruli were isolated by sieving the disintegrated kidney cortex through a series of stainless meshes, according to the method of Spiro (1967). From the glomeruli, GBM was obtained by ultransonication. A suspension containing 0.2% of GBM in 0.9% saline was prepared. A 1 ml sample of the suspension, conjugated with an equal volume of Freund's complete adjuvant, was injected intramuscularly into ducks once a week for 4 weeks. At 2 weeks after the last injection, the ducks were bled from the carotid artery. The pooled serum was stored at - 7 0 ~ until used. Five animals were each given a single intravenous injection of 1 ml duck anti-rat GBM serum and sacrified 14 days later. Urine output for 24 h was measured just before sacrifice. Kidney t&sue preparation. Groups of five rats with Heymann nephritis and nephrotoxic serum nephritis and five normal controls were anesthetized with intraperitoneal sodium pentobarbital (0.1 mg/g body weight). After clamping the right renal artery, the left kidneys were perfused with 0.9% saline for 5 min through a catheter inserted into the aorta to wash out the blood, and then with periodate-lysine-paraformaldehyde fixative (McLean and Nakane 1974) for 5 min. Each contralateral kidney was examined by immunofluorescence, conventional light and electron microscopy.
noted (B). In nephrotoxic serum nephritis, mesangial cell proliferation and occasionally double contours are present (C, arrow head, periodic acid-shift stain). Electron microscopic observations showing mesangial expansion to the subendothelial region of capillary loop D. CL, capillary lumen; US, urinary space; M, mesangial cell; En, endothelial cell. A and C, x400; B and D, • 13000
Isolation of rat plasma fibronect&. Gelatin-Sepharose 4B affinity column chromatography was conducted to purify fibroneetin from pooled normal rat plasma (Engvall and Ruoslahti 1977; Gold and Pearlstein 1979). Fibronectin was eluted with 4M urea. The eluate was dialyzed against phosphate-buffered saline. Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate showed two bands, corresponding to molecular weights of 200000 and 220000, by reduction with beta-mercaptoethanol.
Preparation of anti-rat fibronectin antibody. Rabbit antiserum against fibronectin was raised by the intradermal injection of 1 mg of this protein mixed with Freund's complete adjuvant. At 2 weeks after the first injection, two booster injections were given at 2-week intervals. At 2 weeks after the last injection, antiserum was collected by total exsanguination from each animal. The antiserum was absorbed with fibronectin-free plasma. The IgG fraction, precipitated by ammonium sulfate and subsequently applied to a DE-52 column, was examined for monospecificity by the Ouchterlony method. Fluorescein isothiocyanate (FITC) was tagged to the antiserum and the fraction with a F/P molar ratio of 1.15 was used for direct immunofluorescence. A portion of the antiserum was purified by affinity chromatography on fibronectin-Sepharose 4B (Cuatrecacas 1968) and used for the immunoelectron microscopic study.
181
Light and electron microscopy. A portion of a non-perfused kidney was fixed in 20% formalin. Specimens embedded in paraffin wax were sectioned at 2 ~tm thickness and stained with hematoxylin and eosin, periodic acid-shift and periodic acid-silver methenamine. Non-perfused kidney tissue was cut into small pieces and fixed in 2% glutaraldehyde in 0.15 M phosphate buffer (pH 7.4). Following postflxation in 2% osmium tetroxide in 0.15 M phosphate buffer (pH 7.4) for 2 h, the specimens were dehydrated in a series of graded alcohols and embedded in Quetol 812. Ultrathin sections were cut on an ultratome (LKB 8800), and stained with lead citrate and uranyl acetate. Electron micrographs were obtained using a JEOL 100CX (Nihon Denshi, Tokyo, Japan). Immunofluorescence microscopy. Specimens of non-perfused kidney cortex were snap-frozen in n-hexane cooled in dry ice-acetone and stored at - 7 0 ~ C until used. Cryostat (Bright, Huntington, UK) sections 2 ~tm in thickness were cut, placed on glass slides and stained directly with the FITC-labeled rabbit anti-rat fibronectin antibody. The stained sections were then mounted in buffered glycerol solution for observation under an Olympus Vnox fluorescent microscope provided with an incident illuminator (Olympus, Tokyo, Japan). Immunoelectron microscopy. For immunoelectron microscopy, perfused kidney tissues were cut into small pieces of approximately 0.5 mm 3 and fixed by immersion in periodate-lysin-paraformaldehyde fixative for 6 h at 4~ C. They were then rinsed with 0.05 M glycine-HC1 buffer (pH 7.4) for 16 h and processed at low temperature as follows (Roth et al. 1981 ; Weening et al. 1986). The materials were dehydrated in a graded series of ethanol in water at progressively lower temperature: 30% ethanol for 30 min at 4~ C, 50% for 60 min at - 2 0 ~ C, 70%, 95%, 100% and 100% for 60 min at - 30~ C. The dehydrated specimens were infiltrated with Lowic-
ryl K4M at - 3 0 ~ as follows: one part ethanol and one part Lowicryl K4M for 60 min, one part ethanol and two parts Lowicryl K4M for 60 min, 100% Lowicryl K4M twice for 60 min, and 100% Lowicryl K4M overnight. Polymerization by ultraviolet light (Model B-100A Black-Ray Ultraviolet Lamp UVP, INC) was carfled out at - 3 0 ~ C for 48 h. The polymerized blocks were stored at room temperature. Thin sections (600-800/~) of Lowicryl K4Membedded tissue were mounted on uncoated 300 mesh nickel grids and immunostained. Following incubation in a drop of 0.02 M Tris-buffered saline (TBS) supplemented with 50 mg/ml BSA, 20 nM NAN3, 0.05% Tween 20 and 0.5 M NaCI for 15 min, the specimens on the grids were incubated with affinity-purified rabbit antibody against rat fibronectin, or with rabbit antibody against to rat IgG and albumin (Cappel Laboratories) diluted in 0.1% BSA-TBS (0.02 M Tris-buffered saline supplemented with 1 mg/ml BSA, 20 nM NAN3, 0.05% Tween 20 and 0.5 M NaC1) for 60 min at room temperature. After washing in 0.1% BSA-TBS twice for 5 min, the specimens were incubated with 10 nm gold labeled goat antibody against rabbit IgG (Janssen Pharmaceutica Beers, Belgium) at a dilution of 1 : 10 in 0.1% BSA-TBS for 30 min at room temperature. This was followed by washing twice in 0.1% BSATBS for 5 min, with rinsing in distilled water and drying after each wash. The specimens were stained with uranyl acetate for 5 min and examined in a JEOL 100 CX (Nihon Denshi, Tokyo, Japan). To determine the location of fibronectin in the GBM, the number of gold particles in each layer of GBM was indicated on histograms for controls and rats with Heymann nephritis.
Urinary albumin, IgG and fibronectin excretion. The amount of urinary albumin was determined by single radial immunodiffusion (Becker 1969). Urinary IgG and fibronectin were measured by ELISA as described by Trinick and Laker (1984) and Rennard et al. (1980), respectively.
Fig. 2. Intraglomerular localization of fibronectin in normal rats (A), and those with Heymann nephritis (B), and nephrotoxic serum nephritis (C) (direct immunofluorescence microscopy). In the normal glomerulus (A), fibronectin is localized in the mesangial areas and Bowman's capsule. It is not localized along the capillary loops. In Heymann nephritis (B), a diffuse linear distribution of fibronectin is evident along capillary walls. Mesangial flbronectin is not increased compared with normal controls. In nephrotoxic serum nephritis (C), fibronectin is seen along glomerular capillary loops in a linear pattern. Mesangial fibronectin is expanded in nephrotoxic serum nephritis. A, B and C, x 400
182
Fig. 3 A, B. Electron microscopic localization of fibronectin in normal glomeruli (immuno-goldstaining). Fibronectin is heavily labeled in the mesangial area (A). Mesangial fibronectin has extended continuously to the glomerular capillary walls (A, arrow heads). In peripheral capillary loops, small amounts of fibronectin are distributed in the lamina densa of the GBM (B). CL, capillary lumen; US, urinary space; M, mesangial cell; A, x 13000; B, x 15000
Binding of plasmafibroneetin to glomeruli. Fibronectin at a concentration of 0.7 mg/ml was conjugated with FITC (0.07 mg/ml) in 0.1 M carbonate-bicarbonate buffer (PH 9.3) at room temperature for 2 h with constant stirring. Unreacted dye was removed by get filtration through a Sephadex G-25 column in phosphate buffered saline (PH 7.4) (Chernousov et al. 1985). The ratio of fluorescein to protein was 2.7. The cryostat sections of normal and nephritic rat kidneys were incubated with 0.15 mg/ml fluorescein labeled fibronectin at 37~ C for 2 h. They were then washed twice in phos-
phate-buffered saline at 37~ C, mounted in buffered glycerol solution and observed under an Olympus Vnox fluorescent microscope.
Results
Conventional light and electron microscopic observation T h e m o r p h o l o g i c findings in b o t h e x p e r i m e n t a l m o d e l s were essentially the s a m e as t h o s e r e p o r t e d b y o t h e r in-
183
Fig. 4. Localization of fibronectin in nephrotoxic serum nephritis, as observed by immunoelectron microscopy using immunogold staining. Mesangial interposition in the subendothelial layers and fibronectin distributed about mesangial cells between endothelial cells and GBM are apparent, this being particularly evident be-
tween endothelial and mesangial cells. Increase in the number of gold particles, indicating increase in fibronectin, can be seen in the mesangial matrix (arrow heads). CL, capillary lumen; US, urinary space; M, mesangial cell. • 13000
vestigators (Alousi et al. 1969; Fujimoto et al. 1964). In Heymann nephritis, spike formation (Fig. 1 A). The presence of subepithelial electron-dense deposits and thickening of the GBM (Fig. 1 B) were noted as characteristic features. In nephrotoxic serum nephritis, mesangial expansion with moderate hypercellularity was detected by light microscopy (Fig. 1 C). Electron microscopic examinations showed mesangial expansion to the subendothelial region of capillary loops (moderate mesangial interposition) (Fig. 1 D).
of experimental nephritis (Fig. 2 B, 2 C). Mesangial fibronectin was more densely labeled in nephrotoxic serum nephritis than in normal controls, but not in Heymann nephritis.
Immunofluorescence microscopic examination In normal rats, the glomerular mesangial areas were labeled with fluorescein conjugated antibody against fibronectin. No peripheral capillary loops were labeled (Fig. 2A). Diffuse linear fibronectin distribution along the glomerular capillary wails was noted in both types
Immunoelectron microscopic examination To demonstrate the ultrastructural localization of fibronectin, immunoelectron microscopic examination using immuno-gold staining was conducted on perfused kidneys. In perfused kidneys from normal rats, IgG and albumin was not detected in the glomeruli. Fibronectin was heavily labeled in the mesangial areas (Fig. 3A). On the capillary walls, small numbers of gold particles were seen in the GBM (Fig. 3 B). At 14 days after the anti-rat GBM serum injection in nephrotoxic serum nephritis, fibronectin was seen around mesangial cells situated between the endothelial
184
Fig. 5A-C. Localization of fibronectin in Heymann nephritis as observed by immunoelectron microscopy with immuno-gold staining. Fibronectin is located uniformly in the lamina rara interna, lamina densa and lamina rara externa of the thickened GBM (A, arrow heads and B). Pericapillary fibronectin distribution shows
no specific relation to mesangial (A), endothelial or epithelial cells (B). Amount of fibronectin in the thickened GBM (B) is greater than in the unthickened GBM (C). CL, capillary lumen; US, urinary space; M, mesangial cell; GBM, glomerular basement membrane. A x 15000; B and C, x 17000
185 Control rats
n=5 50,
T
~, 40. *9
~
T
r
~ 20.
iZ 1 0 . G
Endothelium
B
M
Epithelium
Heymann nephritis
,,0 70. o
'~60,
I1
-~ o 50, 40
i
TTTflI T
30
Urinary excretion of album&, IgG andfibronectin
z 20 10 Endothelium
normal controls and rats with Heymann nephritis were also apparent on morphometrical analysis (Fig. 6). As shown by the histograms, the particles were concentrated in the inner half of GBM in normal rats, while a denser distribution was evident through out the thickened GBM in Heymann nephritis. Gold labeling of the GBM was significantly denser in Heymann nephritis than in normal rats (18.0+ 3.8 particles/pm 2 in normal controls compared with 52.0+ll.0particles/pm 2 in Heymann nephritis, P < 0.001). By immunofluorescence microscopy, diffuse linear fibronectin staining was evident along the glomerular capillary walls in Heymann nephritis. Fibronectin was present in small amounts in the endothelial and epithelial cell cytoplasm and IgG was found in glomerular subepithelial electron dense deposits in Heymann nephritis, while IgG was distributed continuously along the GBM in nephrotoxic nephritis (Fig. 7 A, B). No albumin was detected in the glomeruli of rats with either type of experimental nephritis.
G
B
M
Epithelium
Fig. 6. Distribution of immunolabeled fibronectin in the GBM of normal and Heymann nephritis rats. Histogram showing gold particles to be concentrated most in inner half of GBM in normal rats. More diffuse distribution of gold particles is evident throughout the thickened GBM in Heymann nephritis. GBM, glomerular basement membrane
In both glomerulonephritis models, the urinary excretion of albumin, IgG and fibronectin was significantly greater than that in normal controls (P
V'mehowsArchivB Cell Pathology InclmtinfMolecularPatlwlo~ 9 Springer-Verlag 1992
Intraglomerular fibronectin in rat experimental glomerulonephritis Takashi Yoneyama, Mitsumasa Nagase, Mitsuru Ikeya, Akira Hishida, and Nishio Honda First Department of Medicine,HamamatsuUniversitySchoolof Medicine, Hamamatsu,and First Department of Medicine, Teikyo UniversitySchoolof Medicine,Tokyo,Japan Received November 4, 1991 / Accepted April 28, 1992
Summary. To clarify the mechanisms of glomerular pericapillary fibronectin deposition in human membranous nephropathy and mesangial proliferative glomerulonephritis, intraglomerular fibronectin distribution was examined by light and electron microscopy using the experimental rat models of Heymann and nephrotoxic serum nephritis. As previously demonstrated by immunofluorescence microscopy (Pettersson and Colvin 1978; Ikeya et al. 1985, 1986), fibronectin was distributed in the mesangial areas and occasionally on percicapillary walls of normal glomeruli, while in nephrotoxic serum nephritis and Heymann nephritis, fibronectin was diffusely located along glomerular capillary walls as well as in the mesangium. By immunoelectron microscopy using the immunogold technique, fibronectin was also noted in the mesangial areas and the lamina densa of the glomerular basement membrane (GBM) in normal glomeruli. In nephrotoxic serum nephritis, fibronectin was seen around mesangial cells situated between endothelial cells and the GBM, suggesting that pericapillary fibronectin in nephrotoxic serum nephritis reflects mesangial extension. However, in Heymann nephritis, it was found uniformly in the lamina rara interna, lamina densa and lamina rara externa of the GBM, indicating no specific relation to glomerular cells. When sections of normal and both experimental nephritis kidneys were incubated with fluorescein isothiocyanate conjugated with rat plasma fibronectin, a linear pattern of fluorescein staining along the glomerular capillary walls was observed in Heymann nephritis but not in normal or nephrotoxic serum nephritic rats. The GBM in Heymann nephritis would thus appear to have an affinity for plasma fibronectin. Based on the above findings, fibronectin in the GBM of rats with Heymann nephritis may reasonably be concluded to originate from the plasma.
Correspondence to: T. Yoneyama,First Department of Medicine, Hamamatsu UniversitySchool of Medicine, 3600 Handa-cho, Hamamatsu, 431-31, Japan
Key words: Fibronectin - Heymann nephritis - Nephrotoxic serum nephritis - Immunohistochemistry
Introduction Fibronectin, a large molecular adhesive glycoprotein, is distributed in various tissues, on cellular surfaces, collagenous tissues and in the plasma (Yamada 1978; Hynes and Yamada 1982). By immunofluorescence microscopy, fibronectin has been located predominantly in the mesangial areas in normal glomeruli but is found along capillary walls in diseased glomeruli (Pettersson and Colvin 1978; Weiss et al. 1979; Ikeya et al. 1985, 1986). In our previous study (Ikeya t al. 1985, 1986) using immunofluorescence microscopy, fibronectin was seen to extend from the mesangium to the pericapillary region and along capillary loops with progressive mesangial proliferation. In membranous nephropathy, however, pericapillary extension of fibronectin occurred despite the absence of mesangial proliferation. The present study was conducted to determine the pericapillary distribution of fibronectin at the ultrastructural level in Heymann nephritis and to clarify the mechanisms of deposition of this glycoprotein.
Materials and methods Animals. Female Sprague Dawley rats, weighting 150-200 g, were
used. The animals were allowed free access to a standard pellet diet and water throughout the study. Heymann nephritis. The insoluble rat renal tubular epithelial frac-
tion, F • 1A, was prepared from renal cortices of Sprague Dawley rats as described by Edgington et al. (1976). Heymann nephritis was induced in five rats, each immunized four times at weekly intervals with 20 mg of F x 1A emulsifiedin 0.5 ml of 0.9% saline and conjugated with an equal volume of Freund's complete adjuvant. Followingthe initial immunization,24-h urine samples were collected once a week. The animals were sacrified 20-45 weeks after the initial immunization.
180
Fig. 1. Light and electron micrographs of glomeruli from rats with Heymann nephritis rats (A and B) and with nephrotoxic serum nephritis 14 days after duck anti-rat GBM serum injection (C and D). In Heymann nephritis, spikes along the capillary walls are seen by light microscopy. Intraglomerular hypercellularity is not evident (A, periodic acid-silver methenamine stain). By electron miroscopy, subepithelial electron dense deposits and thickening of GBM are
Nephrotoxic serum nephritb. Normal Sprague Dawley rat kidneys were extensively perfused with 0.9% saline. The glomeruli were isolated by sieving the disintegrated kidney cortex through a series of stainless meshes, according to the method of Spiro (1967). From the glomeruli, GBM was obtained by ultransonication. A suspension containing 0.2% of GBM in 0.9% saline was prepared. A 1 ml sample of the suspension, conjugated with an equal volume of Freund's complete adjuvant, was injected intramuscularly into ducks once a week for 4 weeks. At 2 weeks after the last injection, the ducks were bled from the carotid artery. The pooled serum was stored at - 7 0 ~ until used. Five animals were each given a single intravenous injection of 1 ml duck anti-rat GBM serum and sacrified 14 days later. Urine output for 24 h was measured just before sacrifice. Kidney t&sue preparation. Groups of five rats with Heymann nephritis and nephrotoxic serum nephritis and five normal controls were anesthetized with intraperitoneal sodium pentobarbital (0.1 mg/g body weight). After clamping the right renal artery, the left kidneys were perfused with 0.9% saline for 5 min through a catheter inserted into the aorta to wash out the blood, and then with periodate-lysine-paraformaldehyde fixative (McLean and Nakane 1974) for 5 min. Each contralateral kidney was examined by immunofluorescence, conventional light and electron microscopy.
noted (B). In nephrotoxic serum nephritis, mesangial cell proliferation and occasionally double contours are present (C, arrow head, periodic acid-shift stain). Electron microscopic observations showing mesangial expansion to the subendothelial region of capillary loop D. CL, capillary lumen; US, urinary space; M, mesangial cell; En, endothelial cell. A and C, x400; B and D, • 13000
Isolation of rat plasma fibronect&. Gelatin-Sepharose 4B affinity column chromatography was conducted to purify fibroneetin from pooled normal rat plasma (Engvall and Ruoslahti 1977; Gold and Pearlstein 1979). Fibronectin was eluted with 4M urea. The eluate was dialyzed against phosphate-buffered saline. Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate showed two bands, corresponding to molecular weights of 200000 and 220000, by reduction with beta-mercaptoethanol.
Preparation of anti-rat fibronectin antibody. Rabbit antiserum against fibronectin was raised by the intradermal injection of 1 mg of this protein mixed with Freund's complete adjuvant. At 2 weeks after the first injection, two booster injections were given at 2-week intervals. At 2 weeks after the last injection, antiserum was collected by total exsanguination from each animal. The antiserum was absorbed with fibronectin-free plasma. The IgG fraction, precipitated by ammonium sulfate and subsequently applied to a DE-52 column, was examined for monospecificity by the Ouchterlony method. Fluorescein isothiocyanate (FITC) was tagged to the antiserum and the fraction with a F/P molar ratio of 1.15 was used for direct immunofluorescence. A portion of the antiserum was purified by affinity chromatography on fibronectin-Sepharose 4B (Cuatrecacas 1968) and used for the immunoelectron microscopic study.
181
Light and electron microscopy. A portion of a non-perfused kidney was fixed in 20% formalin. Specimens embedded in paraffin wax were sectioned at 2 ~tm thickness and stained with hematoxylin and eosin, periodic acid-shift and periodic acid-silver methenamine. Non-perfused kidney tissue was cut into small pieces and fixed in 2% glutaraldehyde in 0.15 M phosphate buffer (pH 7.4). Following postflxation in 2% osmium tetroxide in 0.15 M phosphate buffer (pH 7.4) for 2 h, the specimens were dehydrated in a series of graded alcohols and embedded in Quetol 812. Ultrathin sections were cut on an ultratome (LKB 8800), and stained with lead citrate and uranyl acetate. Electron micrographs were obtained using a JEOL 100CX (Nihon Denshi, Tokyo, Japan). Immunofluorescence microscopy. Specimens of non-perfused kidney cortex were snap-frozen in n-hexane cooled in dry ice-acetone and stored at - 7 0 ~ C until used. Cryostat (Bright, Huntington, UK) sections 2 ~tm in thickness were cut, placed on glass slides and stained directly with the FITC-labeled rabbit anti-rat fibronectin antibody. The stained sections were then mounted in buffered glycerol solution for observation under an Olympus Vnox fluorescent microscope provided with an incident illuminator (Olympus, Tokyo, Japan). Immunoelectron microscopy. For immunoelectron microscopy, perfused kidney tissues were cut into small pieces of approximately 0.5 mm 3 and fixed by immersion in periodate-lysin-paraformaldehyde fixative for 6 h at 4~ C. They were then rinsed with 0.05 M glycine-HC1 buffer (pH 7.4) for 16 h and processed at low temperature as follows (Roth et al. 1981 ; Weening et al. 1986). The materials were dehydrated in a graded series of ethanol in water at progressively lower temperature: 30% ethanol for 30 min at 4~ C, 50% for 60 min at - 2 0 ~ C, 70%, 95%, 100% and 100% for 60 min at - 30~ C. The dehydrated specimens were infiltrated with Lowic-
ryl K4M at - 3 0 ~ as follows: one part ethanol and one part Lowicryl K4M for 60 min, one part ethanol and two parts Lowicryl K4M for 60 min, 100% Lowicryl K4M twice for 60 min, and 100% Lowicryl K4M overnight. Polymerization by ultraviolet light (Model B-100A Black-Ray Ultraviolet Lamp UVP, INC) was carfled out at - 3 0 ~ C for 48 h. The polymerized blocks were stored at room temperature. Thin sections (600-800/~) of Lowicryl K4Membedded tissue were mounted on uncoated 300 mesh nickel grids and immunostained. Following incubation in a drop of 0.02 M Tris-buffered saline (TBS) supplemented with 50 mg/ml BSA, 20 nM NAN3, 0.05% Tween 20 and 0.5 M NaCI for 15 min, the specimens on the grids were incubated with affinity-purified rabbit antibody against rat fibronectin, or with rabbit antibody against to rat IgG and albumin (Cappel Laboratories) diluted in 0.1% BSA-TBS (0.02 M Tris-buffered saline supplemented with 1 mg/ml BSA, 20 nM NAN3, 0.05% Tween 20 and 0.5 M NaC1) for 60 min at room temperature. After washing in 0.1% BSA-TBS twice for 5 min, the specimens were incubated with 10 nm gold labeled goat antibody against rabbit IgG (Janssen Pharmaceutica Beers, Belgium) at a dilution of 1 : 10 in 0.1% BSA-TBS for 30 min at room temperature. This was followed by washing twice in 0.1% BSATBS for 5 min, with rinsing in distilled water and drying after each wash. The specimens were stained with uranyl acetate for 5 min and examined in a JEOL 100 CX (Nihon Denshi, Tokyo, Japan). To determine the location of fibronectin in the GBM, the number of gold particles in each layer of GBM was indicated on histograms for controls and rats with Heymann nephritis.
Urinary albumin, IgG and fibronectin excretion. The amount of urinary albumin was determined by single radial immunodiffusion (Becker 1969). Urinary IgG and fibronectin were measured by ELISA as described by Trinick and Laker (1984) and Rennard et al. (1980), respectively.
Fig. 2. Intraglomerular localization of fibronectin in normal rats (A), and those with Heymann nephritis (B), and nephrotoxic serum nephritis (C) (direct immunofluorescence microscopy). In the normal glomerulus (A), fibronectin is localized in the mesangial areas and Bowman's capsule. It is not localized along the capillary loops. In Heymann nephritis (B), a diffuse linear distribution of fibronectin is evident along capillary walls. Mesangial flbronectin is not increased compared with normal controls. In nephrotoxic serum nephritis (C), fibronectin is seen along glomerular capillary loops in a linear pattern. Mesangial fibronectin is expanded in nephrotoxic serum nephritis. A, B and C, x 400
182
Fig. 3 A, B. Electron microscopic localization of fibronectin in normal glomeruli (immuno-goldstaining). Fibronectin is heavily labeled in the mesangial area (A). Mesangial fibronectin has extended continuously to the glomerular capillary walls (A, arrow heads). In peripheral capillary loops, small amounts of fibronectin are distributed in the lamina densa of the GBM (B). CL, capillary lumen; US, urinary space; M, mesangial cell; A, x 13000; B, x 15000
Binding of plasmafibroneetin to glomeruli. Fibronectin at a concentration of 0.7 mg/ml was conjugated with FITC (0.07 mg/ml) in 0.1 M carbonate-bicarbonate buffer (PH 9.3) at room temperature for 2 h with constant stirring. Unreacted dye was removed by get filtration through a Sephadex G-25 column in phosphate buffered saline (PH 7.4) (Chernousov et al. 1985). The ratio of fluorescein to protein was 2.7. The cryostat sections of normal and nephritic rat kidneys were incubated with 0.15 mg/ml fluorescein labeled fibronectin at 37~ C for 2 h. They were then washed twice in phos-
phate-buffered saline at 37~ C, mounted in buffered glycerol solution and observed under an Olympus Vnox fluorescent microscope.
Results
Conventional light and electron microscopic observation T h e m o r p h o l o g i c findings in b o t h e x p e r i m e n t a l m o d e l s were essentially the s a m e as t h o s e r e p o r t e d b y o t h e r in-
183
Fig. 4. Localization of fibronectin in nephrotoxic serum nephritis, as observed by immunoelectron microscopy using immunogold staining. Mesangial interposition in the subendothelial layers and fibronectin distributed about mesangial cells between endothelial cells and GBM are apparent, this being particularly evident be-
tween endothelial and mesangial cells. Increase in the number of gold particles, indicating increase in fibronectin, can be seen in the mesangial matrix (arrow heads). CL, capillary lumen; US, urinary space; M, mesangial cell. • 13000
vestigators (Alousi et al. 1969; Fujimoto et al. 1964). In Heymann nephritis, spike formation (Fig. 1 A). The presence of subepithelial electron-dense deposits and thickening of the GBM (Fig. 1 B) were noted as characteristic features. In nephrotoxic serum nephritis, mesangial expansion with moderate hypercellularity was detected by light microscopy (Fig. 1 C). Electron microscopic examinations showed mesangial expansion to the subendothelial region of capillary loops (moderate mesangial interposition) (Fig. 1 D).
of experimental nephritis (Fig. 2 B, 2 C). Mesangial fibronectin was more densely labeled in nephrotoxic serum nephritis than in normal controls, but not in Heymann nephritis.
Immunofluorescence microscopic examination In normal rats, the glomerular mesangial areas were labeled with fluorescein conjugated antibody against fibronectin. No peripheral capillary loops were labeled (Fig. 2A). Diffuse linear fibronectin distribution along the glomerular capillary wails was noted in both types
Immunoelectron microscopic examination To demonstrate the ultrastructural localization of fibronectin, immunoelectron microscopic examination using immuno-gold staining was conducted on perfused kidneys. In perfused kidneys from normal rats, IgG and albumin was not detected in the glomeruli. Fibronectin was heavily labeled in the mesangial areas (Fig. 3A). On the capillary walls, small numbers of gold particles were seen in the GBM (Fig. 3 B). At 14 days after the anti-rat GBM serum injection in nephrotoxic serum nephritis, fibronectin was seen around mesangial cells situated between the endothelial
184
Fig. 5A-C. Localization of fibronectin in Heymann nephritis as observed by immunoelectron microscopy with immuno-gold staining. Fibronectin is located uniformly in the lamina rara interna, lamina densa and lamina rara externa of the thickened GBM (A, arrow heads and B). Pericapillary fibronectin distribution shows
no specific relation to mesangial (A), endothelial or epithelial cells (B). Amount of fibronectin in the thickened GBM (B) is greater than in the unthickened GBM (C). CL, capillary lumen; US, urinary space; M, mesangial cell; GBM, glomerular basement membrane. A x 15000; B and C, x 17000
185 Control rats
n=5 50,
T
~, 40. *9
~
T
r
~ 20.
iZ 1 0 . G
Endothelium
B
M
Epithelium
Heymann nephritis
,,0 70. o
'~60,
I1
-~ o 50, 40
i
TTTflI T
30
Urinary excretion of album&, IgG andfibronectin
z 20 10 Endothelium
normal controls and rats with Heymann nephritis were also apparent on morphometrical analysis (Fig. 6). As shown by the histograms, the particles were concentrated in the inner half of GBM in normal rats, while a denser distribution was evident through out the thickened GBM in Heymann nephritis. Gold labeling of the GBM was significantly denser in Heymann nephritis than in normal rats (18.0+ 3.8 particles/pm 2 in normal controls compared with 52.0+ll.0particles/pm 2 in Heymann nephritis, P < 0.001). By immunofluorescence microscopy, diffuse linear fibronectin staining was evident along the glomerular capillary walls in Heymann nephritis. Fibronectin was present in small amounts in the endothelial and epithelial cell cytoplasm and IgG was found in glomerular subepithelial electron dense deposits in Heymann nephritis, while IgG was distributed continuously along the GBM in nephrotoxic nephritis (Fig. 7 A, B). No albumin was detected in the glomeruli of rats with either type of experimental nephritis.
G
B
M
Epithelium
Fig. 6. Distribution of immunolabeled fibronectin in the GBM of normal and Heymann nephritis rats. Histogram showing gold particles to be concentrated most in inner half of GBM in normal rats. More diffuse distribution of gold particles is evident throughout the thickened GBM in Heymann nephritis. GBM, glomerular basement membrane
In both glomerulonephritis models, the urinary excretion of albumin, IgG and fibronectin was significantly greater than that in normal controls (P
