JOURNAL OF PATHOLOGY, VOL. 178: 352-358 (1996)

EXPERIMENTAL ANTI-GBM GLOMERULONEPHRITIS INDUCED IN RATS BY IMMUNIZATION WITH

SYNTHETIC PEPTIDES BASED ON SIX a CHAINS OF HUMAN TYPE IV COLLAGEN

KATSUY UKI SUGIHARA*$, YOSHIKAZU S A D O ~ , YOSHIFUMI NINOMIYA~ AND HIROYOSHI WADAS *Shigei Medical Research Hospital, Yamada, Okayama, Jupun; ?Division of Immunology, Shigei Medical Re.rearch Institute,

Yarnadu, Okayunia, .Iupan; $Department of Molecular Biology and Biochemistry, Okuyama University Medicul School, Shikatn-chho, Okriyarnn, Japan; ~ D e p Q r t ~ e n t of’ Pediatrics, Hyogo College of Medicine, Mulcoga wa-cko, N ~ ~ h ~ n i ) t ? ~ i ~ a , Hyogo,

Jupan

SUMMARY The anti-glomerular basement membrane (GBM)-nephritis-inducing activity of six synthetic peptides having an amino acid sequence

consisting of the six a chains of human type IV collagen was examined by injecting the peptides into rats. The peptides consisted of 27 amino acid residues from the non-collagenous domain (NC1) of the a1 to a6 chains and were non-consensus sequences sandwiched between two consensus sequences near the carboxyl terminus. Each peptide was coupled to keyhole limpet haemocyanin and injected with adjuvant into the footpads of 20 female WKYlNCrj rats. The number of rats with proteinuria (over 10.0 mg of urinary proteinll5 h) and haematuria was 2 with the a3 peptide, 8 with the a4 peptide, and 1 with the a5 peptide. Histological changes seen in the glomeruli were characteristic of those in anti-GBM nephritis. Linear deposition of rat IgG along the GBM was observed in five rats injected with the a4 peptide. A nephritogenic monoclonal antibody against the a4 peptide was established using lymph node cells from a rat injected with the a4 peptide. The results indicate that a4(IV)NCl is a potent nephritogenic antigen like a3(1V)NC1, which has already been recognized as a primary target antigen in Goodpasture’s syndrome.

KE\ worn-anti-GBM nephritis; glomerulonephritis; glomerular basement membrane; synthetic peptide; type IV collagen; rat monoclonal antibody

INTRODUCTION

Anti-glomerular basement membrane antibody- induced glomerulonephritis (anti-GBM nephritis) is an autoimmune disease induced by autoantibodies that bind to the glomerular basement membrane (GBM).’ It is frequently accompanied by pulmonary haemorrhage, in which case it is called Goodpasture’s syndrome.2 The antigen responsible for the disease has been studied using sera obtained from patients with Goodpasture’s syndrome. The primary target antigen, termed Goodpasture antigen, is believed to be the non- collagenous domain (NCl) of the a3 chain of type IV collagen, a3(IV)NC 1, because autoantibodies against it have been detected in patients’ sera by Western blotting.? l o

Type IV collagen is one of the major structural components of basement membranes. It is a hetero- trimeric molecule, and each a chain has three distinct domains: the 7s domain at the amino terminus, the central triple-helical domain, and the NCl located at the carboxyl terminus.”.” So far, six distinct but related a chains have been identified: the classical chains al(1V) and n2( IV) and the recently discovered a3(IV) to a6(IV) chains.” The complete primary structure of the six human a chains is known.14 **

Addiessee foi correspondence Katsuyuki Sugihara, Shigei Medical Reaeaich Hospital. 21 17 Ydmdda, Okaydma 701-02, Japan

CCC 0022-341 7/96/030352-07 t ‘ 1996 by John Wiley Bi Sons, Ltd

Research into the nature of the nephritogenic antigen, which is defined as the antigen that induces anti-GBM nephritis in rats, has been conducted to elucidate the mechanism underlying this nephritis and to devise potential forms of therapy.23 27 The nephritogenic anti- gen purified from collagenase-solubilized bovine renal basement membrane resides in the NC 1 fraction of type IV collagen and can be used as an assay antigen to detect Goodpasture antibody in patient serum.26 Although this fraction contains al(1V)NCl to nS(IV)NCl, further isolation of the NC1 of each a chain has not been done.

To answer the question as to whether NCls other than that of a3(IV) are nephritogenic, the peptides corresponding to sequences of human NCl of al(1V) to a6(IV) were synthesized, coupled to a carrier protein, and injected into rats. Even though the immunogenic peptides used were only 27 amino-acid residues long and much smaller than the native molecules, injection of the a3(1V)NCI, u4(TV)NC1, and aS(1V)NCl peptides induced anti-GBM nephritis in rats. In particular, the a4(IV)NC1 peptide induced severe nephritis, and a nephritogenic monoclonal antibody was established from a rat injected with the a4 peptide.

MATERIALS AND METHODS Animals

Female WKY/NCrj inbred rats purchased from Charles River Japan, Inc. (Yokohama, Japan) were

Received 30 November 1994 Accepted 28 June 1995

ANTI-CBM NEPHRlTIS BY SYNTHETIC PEPTIDES

C-terminus

353

Fig. 1-Carboxyl-terminal sequences of the six u chains of human type IV collagen and the peptide sequences used in the present study (boxed). Dashes indicate an amino acid sequence the same as that for the a1 chain; otherwise a substituted residue is indicated. Dots indicate gaps introduced by alignment

used. Rats were housed in polycarbonate plastic cages containing wood shavings as bedding. The animal exper- iments in the present study were conducted according to the Guidelines for Laboratory Animal Experiments of Shigei Medical Research Institute.

Synthetic peptides

Although the amino acid sequences of NCls of the six a chains of human type IV collagen show high homology, the sequences of certain regions are quite different from each other. The sequences selected for this study are shown in Fig. 1. These were 21 amino acids long, near the carboxyl terminus, and sandwiched between consensus sequences.

Sequences were synthesized by the stepwise solid- phase method using an automated peptide synthesizer. A cysteinyl residue was added at the amino terminal of the sequence for coupling to a carrier protein. Peptides were purified to more than 55 per cent purity on a Vydac RP-C18 column (250 mm x 10 mm) (Kurabo Industries Ltd., Osaka, Japan). The peptides were designated A121, A221, A327, A421, A521, and A621 according to the order of the a chains and they were conjugated to keyhole limpet haemocyanin (KLH) by the m- maleimidobenzoyl-N-hydroxysuccinimide-ester method28 for use as an immunogen. The amount of peptide coupled to KLH was estimated by amino acid analysis (Chiron Mimotopes P/L, Victoria, Australia).

Nephritogenicity of the synthetic peptides

Each peptide was tested for nephritogenicity with 20 8-week-old female rats. The rats were injected via hind footpads with an emulsion containing 1OOpg of the peptide conjugated to KLH and an equal amount of Freund's complete adjuvant (FCA). One group of rats injected with an emulsion containing KLH only served as the control group. Fifteen-hour urine (from 6 p.m. to 9 a.m.) was collected at 4, 5, and 6 weeks after injection.

Rats were killed 6 weeks after injection and samples of renal tissue and sera were obtained.

Urinary analysis, histological examination, immunofuorescence study and serum antibody measurement

Urine was assayed for proteinuria by a modified method using sulphosalicylic acid26 and for haematuria by test paper (Hema-Combistix, Miles-Sankyo, Tokyo, Japan). Samples of renal tissue were fixed in Dubosq- Brazil fixative, embedded in paraffin, sectioned at 3 pm, and stained with haematoxylin and eosin, periodic acid- Schiff, and periodic acid-methenamine-silver (PAM). For the immunofluorescence study, 4-pm cryostat sec- tions of the kidney were stained with FITC-conjugated goat antibody against rat IgG and C3 (Cappel Lab., U. S. A.).

The titre in the sera of rats injected with each peptide was assayed by enzyme-linked immunosorbent assay (ELISA) using each peptide as an assay antigen. The titre was expressed as the reciprocal of the maximum dilution yielding a value that was 0.1 absorbance unit higher than the control value.

Establishment of a monoclonal antibody against the a4(IV) peptide and induction of anti-GBM nephritis by the antibody

A monoclonal antibody against A421 was established by the newly developed rat lymph node method.29 Three female WKY rats were injected with the emulsion con- taining A421 via hind footpads and cell fusion was performed three times at 3, 4, and 5 weeks after injec- tion. A pair of medial iliac lymph nodes were excised and used as the source of B cells for cell fusion to produce hybridomas. Mouse SP2/0-Ag14 myeloma cells were used as fusion partners. A hybridoma cell line was established which produced a monoclonal antibody reacting with both A427 and rat cryostat sections.

The established cells were cultured with a modular minifermenter (Heraeus Instruments, Osterode,

354 K. SUGIHARA ET A L

. . . . . . . . . . a3 8

-cD

4. 0 0 . . 0

: 0. P

Q9 0 0

2%. @O 0.0". *

0 00 0 0 0

7 og 80 0

0

00

KLH A127 A227 A327 A427 A527 A627 only

Fig. 2-Urinary protein and haematuria in rats injected 6 weeks previously with synthetic peptides c', without haematuria; 0 , with haematuria

Germany) to obtain culture supernatant containing a high concentration of the antibody. The culture super- natant was applied to an affinity column where 1 mg of A427 had been coupled to 0.3 g of 2-fluoro-l- methylpyridinium toluene-4-sulphonate activated Cellu- lofine gel (Seikagaku Kogyo Co., Tokyo, Japan). The adsorbed antibody was eluted with 0.1 M glycine-HC1 buffer, pH 2.5, and the pH was adjusted immediately to neutral with 1 M tris(hydroxymethy1)aminomethane. The eluted antibody was dialysed against phosphate- buffered saline (PBS) and concentrated with a collodion bag. The concentration of the antibody was measured by absorbance at 280 nm.

Female WKY rats were injected intraperitoneally with 0, 10, or 50pg of the purified antibody in 1 ml of PBS. Urine was assayed on days 0, 1,2, 3,4, 5,6, 8, 10, 12, 14, 16, 18, and 20. The rats were killed on day 21. Four rats were injected with 5Opg of the antibody and two each were killed on days 7 and 14.

RESULTS Nephvitogenicity of the synthetic peptides

The urinary protein and haematuria of rats at 6 weeks after injection with the synthetic peptides are shown in Fig. 2. Haematuria is a characteristic sign of nephritis and was not detected in the control rats injected with KLH only. Out of 20 animals, the number of rats with haematuria was 2 amongst those injected with A127, 2 with A227, 10 with A327, 13 with A427, 6 with A527, and 2 with A627. Proteinuria was defined as over 10.0 mg of urinary protein/l5 h, because the control rats showed a range of 1.1-7.2mg/15 h. It was detected in two rats injected with A327, in eight rats with A427, and in one rat with A527, and it was always

accompanied by haematuria. Notably, heavy protein- uria (over 100 mg/15 h) was observed in 5 of the 20 rats injected with A427.

Mild pulmonary haemorrhage (below 100 petechiae on the surface of the lung at death) was found in some rats with haematuria. However, it was difficult to judge whether the pulmonary haemorrhage was related to the nephritis, because one rat in the control group also had mild pulmonary haemorrhage.

Histological changes Characteristic changes seen in the glomeruli of rats

having moderate to severe proteinuria were endocapil- lary hypercellularity of mononuclear cells, capsular adhesion, lobulation and sclerosis of capillary tufts, and crescent formation (Fig. 3A). Rats having both mild proteinuria and haematuria also showed mild histologi- cal changes in the golmeruli such as endocapillary hypercellularity and capsular adhesion.

The percentage of glomeruli that exhibited histologi- cal abnormalities depended on the severity of the pro- teinuria. In the rats injected with A327, abnormal glomeruli of the two rats having moderate proteinuria (10-100 mg/l5 h) with haematuria were in the range of 34.548.0 per cent, those of the eight rats having haema- turia only were in the range of 3.6-21.9 per cent and those of the other ten rats having neither haematuria nor proteinuria were in the range of 0-1.3 per cent.

In the rats injected with A427, abnormal glomeruli of the five rats having severe proteinuria (over 100 mg/l5 h) with haematuria were in the range of 84.0-97.6 per cent, those of the three rats having moderate proteinuria with haematuria were in the range of 5.4-29.0 per cent, those of the four rats having haematuria but no pro- teinuria were in the range of 2-4-6.3 per cent, and those

ANTI-GBM NEPHRITIS BY SYNTHETIC PEPTIDES 355

Fig. 3-(A) Histology of the kidney of a heavily proteinuric rat 6 weeks after injection with peptide corresponding to a4(IV)NCI. Severe damage such as lobulation and sclerosis of the capillary tufts and crescent formation is found. PAM staining. x 160. (B) Direct immunofluorescence of the kidney of a rat injected with peptide corresponding to a4(IV)NCl. The rat had heavy proteinuria and was killed 6 weeks after injection. Fluorescence is observed in a linear pattern along the GBM. x 280

of the other seven rats having neither haematuria nor proteinuria were in the range of 0-1.6 per cent.

In the rats injected with A527, abnormal glomeruli of the one rat having moderate proteinuria with haema- turia were 16.1 per cent, those of the five rats having haematuria but no proteinuria were in the range of 1.4-23-6 per cent, and those of the other 14 rats having neither haematuria nor proteinuria were in the range of 0-1.5 per cent.

In the rats injected with A127, A227, or A627, abnor- mal glomeruli of the six rats having haematuria but no proteinuria were in the range of 1.4-7.1 per cent and those of the other 54 rats having neither haematuria nor proteinuria were in the range of 0-1.5 per cent. Those of the 20 control rats injected with KLH only were in the range of 0-1.3 per cent.

Immunofluorescence study

An immunofluorescence study using FITC- conjugated antibody against rat IgG revealed intense linear deposition of IgG along the GBM of 5 of the 20 rats injected with A427 (Fig. 3B), but no deposition in the other rats used in this study. It was impossible to discriminate faint to moderate staining of immunized rats from the control staining using age-matched control rats, because of the presence of serum IgG along the GBM of the control rats even after the kidney had been perfused with phosphate-buffered saline. No C3 deposition was detected along the GBM in any of the rats.

Serum titre against the injected peptides Serum titres against the injected peptides in the case of

A127 were in the range of 1000-28 000 (median 7800); in the case of A227, 8900-80 000 (median 25 000); in the case of A327, 3900-1 10 000 (median 22 000); in the case

of A427, 24 000-130 000 (median 55 000); in the case of A527, 900042 000 (median 17 500); and in the case of A627, 21 000-150 000 (median 58 000). These high values indicated there was no trouble with the immunization process.

Establishment of a nephritogenic monoclonal antibody against the a4(IV) peptide

A hybridoma cell line was established which produced a monoclonal antibody reacting with both A427 and rat cryostat sections. The monoclonal antibody was named H43. Its subclass was rat IgG2b. It reacted with A427, but not with A127, A227, A327, A527, or A627 by ELISA. Its epitope was examined by multipin-peptide scanning with overlapping octapeptides homologous to A421 and by a bovine sequence corresponding to A427. It reacted with the human peptide sequence PAPDTLKE and the bovine sequence PLPDTLKE.

Rat cryostat sections were stained with H43 culture supernatant for indirect immunofluorescence (Fig. 4). The GBM, tubular basement membrane (TBM), and Bowman’s capsular basement membrane were faintly stained. In human renal sections, the GBM and part of the TBM were intensively stained and Bowman’s capsular basement membrane was faintly stained.

Induction of anti-GBM nephritis by monoclonal antibody H43

Purified H43 (10 or 50pg) was injected into female WKY rats intraperitoneally. The time course of pro- teinuria is shown in Fig. 5. Proteinuria with haematuria began on day 2-3 and became severe as time progressed. The characteristic histological change seen in the glomeruli of rats injected with 50pg of H43 was endo- capillary hypercellularity of the mononuclear cells (Fig. 6A). Mild capsular adhesion, lobulation and sclerosis of

356 K. SUGIHARA ET A L

Fig. &Indirect immunofluorescence of a normal rat kidney incubated with H43. The cryostat section was first incubated for 1 h with H43 (culture supernatant) and then stained with FITC-conjugated anti-rat IgG antibody. Fluorescence is observed along the glomerular and tubular basement membranes x 280

the capillary tufts, and crescent formation was also seen. The rats given 10pg of H43 exhibited proteinuria, haematuria, and histological changes in the glomeruli, but with lesser severity than rats given 5 0 p g of H43. The monoclonal antibody was detected in a linear pattern along the GBM by direct immunofluorescence (Fig. 6B). Control rats injected with PBS had neither proteinuria, haematuria, nor linear immunofluorescence on the GBM.

DISCUSSION

The present study clearly indicates for the first time that synthetic peptides can induce anti-GBM nephritis in rats. Compared with the cases in which purified bovine23 or rat24 native antigen was injected, the nephritis

1

3

1 0 2 4 6 8 10 12 14 16 18 20

Days after injection Fig. 5-Time course and dose-effect of proteinuria induced by injec- tion with monoclonal antibody H43. WKY rats were injected with 1Opg ( A ) or 50pg (0) of H43, or PBS (0). Values are expressed as an average of the urinary protein of four rats. Open and closed symbols indicate without and with haematuria, respectively. The severity of the histological changes is well correlated with the severity of urinary protein loss.

induced in the present study was mild on the whole and the individual differences were very large. This was due to both fewer antigenic determinants on the immunogen and fewer conformational antigenic determinants which exist on the native antigens. The importance of confor- mational antigenic determinants in glomerulonephritis has been demonstrated by Kalluri et dX The reactivity of Goodpasture sera with solubilized GBM antigen was markedly reduced when conformational determinants of the assay antigen were destroyed by reducing the disulphide bonds.

The amino acid sequences of human type IV collagen were used in the present study as peptide sequences, because those of rat type IV collagen have not yet been reported. Rat sequences are probably very similar to

Fig. &(A) Histology of the kidney of a rat injected with monoclonal antibody H43 on day 21. Mild damage is found in glomeruli, such as endocapillary hypercellularity and lobulation of the capillary tufts. PAM staining. x 160. (B) Direct immunofluorescence of the kidney of a rat injected with 50pg of H43 on day 21. The fluorescence pattern is linear along the GBM. x 280

ANTI-GBM NEPHRITIS BY SYNTHETIC PEPTIDES 357

those of human type IV collagen, because the sequences of the NCI of mouse a l , ~ 2 , ~ ~ a3, a4, and a5 chains;31 bovine and a4 chains;33 rabbit al, and a4 chains;35 and dog a5 chain36 have all been reported and comparisons indicate that mammalian sequences are very similar among the same a chains. Active-type anti-GBM nephritis can be induced in WKY rats by the injection of both heterologous (bovine)23 and homolo- gous (rat)24 NCI fractions of type IV collagen from renal basement membranes. The use of the human sequences in place of the rat sequences in the present study is based on these facts.

Although several studies have reported that exper- imental anti-GBM nephritis can be induced by the injection of solubilized GBM into rats, none has clearly analysed the nephritogenic component^.^^-^^ Because the synthetic peptides used in this study were not con- taminated with other a chains, the results obtained were very reliable.

The synthetic peptides A127 to A627, correspond- ing to al(1V)NCl to a6(IV)NCl, respectively, had sequences near the carboxyl-terminal regions of human a(1V)NCl. These were non-consensus sequences sandwiched between two consensus sequences, although they shared short common sequences. The synthetic peptides A327, A427, and A527 exhibited nephrito- genic activity and A427 induced particularly severe nephritis. On the other hand, the synthetic peptides A127, A227, and A627 exhibited little nephritogenic activity. This result clearly indicates that a3(IV)NC 1, a4(IV)NCl, and aS(1V)NCl are nephritogenic antigens with regard to experimental anti-GBM nephritis in rats.

As to nephritogenic antigens from the renal basement membrane, we earlier purified a type IV collagen NCI fraction, P3, having intense nephritogenicity and another type IV collagen NCI fraction, P2, having very weak nephritogenicity, from bovine renal basement membrane.26 We have already reported that the P3 fraction has the properties of the Goodpasture anti- gen.26 The bovine P2 fraction contained al(1V)NCl and a2(IV)NC1 as the major constituents, whereas the P3 fraction contained al(1V)NCl to aS(1V)NCI. The nephritogenicity of al(1V)NCl and a2(IV)NCl was thus concluded to be negligible, because the P2 fraction was very weakly nephritogenic. As a consequence, either one or more than one of a3(IV)NCI, a4(IV)NC1 and aS(1V)NCl was considered to have nephritogenicity.

Induction of passive-type anti-GBM nephritis in rats by a homologous monoclonal antibody has already been reported.37 The monoclonal antibodies were against the NC1 fraction of type IV collagen but their a-chain specificity was not determined. In contrast, H43 pro- duced in the present study was well characterized. Its epitope was mapped by multipin-peptide scanning and was specific against the a4(IV) chain. Because the origin cell of the hybridoma producing H43 was obtained from a rat injected with A427, it is obvious that the active model of glomerulonephritis induced by the injection of A427 was anti-GBM nephritis. It is likely that the nephritis induced by the injection of A327 and A527 was also anti-GBM nephritis.

Several studies have reported that a3(IV)NC1 is the primary target antigen, termed Goodpasture antigen.”O However, the nephritogenic activity of a3(IV)NC 1 has not yet been demonstrated. The results of the present study show that a3(IV)NC1 is a nephritogenic antigen and strongly support the view that it is the primary target antigen in human Goodpasture’s syndrome.

The results of the present study suggest that a4(IV)NC1 is also a primary target antigen in human Goodpasture’s syndrome for the three reasons: because the human a4(IV) chain has a similar structure at the amino-acid sequence level to that of the a3(IV) chain; because its gene exists in human chromosome 2 in a head-to-head fashion with the gene of u ~ ( I V ) ‘ ~ ; and because a4(IV) molecules are distributed in a pattern identical to that of the human a3(IV) molecules as seen by immunostaining of t i s ~ u e s . ~ * J ~ Another piece of supporting evidence is that the autoantibody against a4(IV)NC1 has been detected in Goodpasture sera by Western blotting.40

This model of anti-GBM nephritis induced by syn- thetic peptides is simple, but knowledge obtained with the defined antigen is of fundamental importance in delineating the aetiology of anti-GBM nephritis. Epitope-defined nephritogenic monoclonal antibodies are also valuable in research on the development of anti-GBM nephritis.

ACKNOWLEDGEMENTS

We wish to thank Drs Tohru Okigaki and Ichiro Naito (Shigei Medical Research Institute) and Drs Masafumi Taki and Hiroyuki Ohmori (Shigei Medical Research Hospital) for their valuable advice; Ms Megumi Kagawa and Ms Yumiko Kishiro for technical assistance and advice; Mrs Fusae Ueki for preparation of the histological sections; and Mrs Chieko Takahashi for vivarium care. Special thanks are due to Dr Jerome M. Seyer (University of Tennessee, Memphis) for synthesizing the peptides used in this study.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

Wilson CB, Dixon FJ. Anti-glomerular basement membrane antibody- induced glomerulonephritis. Kidney In1 1973; 3 7489. Benoit FL, Rulon DB, Theil GB, Doolan PD, Watten RH. Goodpasture’s syndrome. A clinicopathologic entity. Am J Med 1964; 37: 424444. Wieslander J, Barr JF, Butkowski RJ, ef ul. Goodpasture antigen of the glomerular basement membrane: localization to noncollagenous regions of type IV collagen. Proc Nut1 Acud Sci USA 1987; 262: 7874-7871. Butkowski RJ, Langeveld JPM, Wieslander J, Hamilton J, Hudson BG. Localization of the Goodpasture epitope to a novel chain of basement membrane collagen. J Bid Chenz 1987; 262 7874-7877. Saus J, Wieslander J, Langeveld JPM, Quinones S, Hudson BG. Identifica- tion of the Goodpasture antigen as the u3(IV) chain of collagen IV. J Biol Chem 1988; 263: 13374-13380. Gonwar S, Ballester F, Kalluri R, et u/. Glomerular basement membrane. ldentification of dimeric subunits of the noncollagenous domain (hexamer) of collagen IV and the Goodpasture antigen. J Biol Chrm 1991; 266 15318-15324. Hudson BG, Wieslander J, Wisdom BJ Jr, Noelken ME. Goodpasture syndrome: molecular architecture and function of basement membrane antigen. Lab Invest; 1989; 61: 256-269. Kalluri R, Gunwar S, Reeders ST, e l ul. Goodpasture syndrome. Localira- tion of the epitope for the autoantibodies to the carboxyl-terminal region of the u3(IV) chain of basement membrane collagen. J Bid Chem 1991; 266 24018-24024.

358 K. SUGIHARA ET AL.

9

10

I I

12

13

14

15

16

17

18

19.

20.

21.

22

23.

24.

Neilson EG, Kalluri R, Sun MJ, ef a/. Specificity of Goodpasture auto- antibodies for the recombinant noncollagenous domains of human type IV collagen. J Biol Chem 1993; 268: 8402-8405. Hudson BG, Kalluri R, Gunwar S, Noelken ME, Mariyama M, Reeders ST. Molecular characteristics of the Goodpasture autoantigen. Kidney Int 1993; 43: 135-139. Martin GR, Timpl R, Kuhn K. Basement membrane proteins: molecular structure and function. Adv Protein Chem 1988; 3 9 1-50. Timpl R, Wiedemann H, Van Delden V, Furthmayr H, Kuhn K. A network model for the organization of type IV collagen molecules in basement membranes. Eur J Biochem 1981; 120: 203-211. Hudson BG, Reeders ST, Tryggvason K. Type 1V collagen: structure, gene organization, and role in human disease. J Bid Chem 1993; 268 26033-26036. Brazel D, Oberbaumer I, Dieringer H, et a/. Completion of the amino acid sequence of the nl chain of human basement membrane collagen (type IV) reveals 2 I non-triplet interruptions located within the collagenous domain. Eur J Biodiern 1987; 168: 529-536. Soininen R, Haka-Risku T, Prockop DJ, Tryggvason K. Complete primary structure of the al-chain of human basement membrane (type IV) collagen. FEBS Left 1987; 225: 188-194. Brazel D, Pollner R, Oberbaumer 1, Kiihn K. Human basement membrane collagen (type IV). The amino acid sequence of the a2(IV) chain and its comparison with the ul(lV) chain reveals deletions in the nl(lV) chain. Eur J Biochem 1988; 172: 3542. Hostikka SL, TryggVdSOn K. The complete primary structure of the u2 chain of human type IV collagen and comparison with the al(IV) chain. J Biol Chem 1988; 263: 19488-19493. Mariyama M, Leinonen A, Mochizuki T, Tryggvason K, Reeders ST. Complete primary structure of the human a3(IV) collagen chain: coexpres- sion of the n3(1V) and n4(IV) collagen chains in human tissues. J B i d Chem 1994; 269 23013-23017. Leinonen A, Mariyama M. Mochizuki T, Tryggvason K, Reeders ST. Complete primary structure of the human type IV collagen a4(IV) chain. J Biol Chem 1994; 269 26172-26177. Zhou J, Hertz JM, Leinonen A, Tryggvason K. Complete amino acid sequence of the human a5(1V) collagen chain and identification of a single-base mutation in exon 23 converting glycine 521 in the collagenous domain to cysteine in an Alport syndrome patient. J Biol Chem 1992; 267: 12475--1248 1. Oohashi T, Sugimoto M, Matei M, Ninomiya Y. Identification of a new collagen IV chain, a6(IV), by cDNA isolation and assignment of the gene to chromosome Xq22, which is the same locus for COL4A5. .I Biol Chem 1994; 269 7520-7526. Zhou J. Ding M, Zhao Z, Reeders ST. Complete primary structure of the sixth chain of human basement membrane collagen, u6(IV). J Biol Chem 1994; 269 13193-13199. Sado Y, Okigaki T, Takamiya H, Seno S. Experimental autoimmune glomerulonephriti5 with pulmonary hemorrhage in rats. The dose-effect relationship of the nephritogenic antigen from bovine glomerular basement membrane. J Clin Lab Inzmuwol 1984; 1 5 199-204. Sado Y , Naito 1. Experimental autoimmune glomerulonephritis in rats by soluble isologous or homologous antigens from glomerular and tubular basement membranes. Br J Exp Pufhol 1987; 6 8 695-704.

25. Pusey CD, Holland MJ, Cashman SJ, e f ul. Experimental autoimmune glomerulonephritis induced by homologous and isologous glomerular base- ment membrane in Brown-Norway rats. Nephrol D i d Trcmsplunf 199 I ; 6 457465.

26. Sado Y, Kagawa M, Naito 1, Okigaki T. Properties of bovine nephritogenic antigen that induce anti-GBM nephritis in rats and its similarity to the Goodpasture antigen. VirchoMis Arch B [Cell Pathol] 1991: 60: 345-351

27. Kalluri R. Gattone VH 11. Noelken ME. Hudson BG. The a3 chain of tvoe ,>

1V collagen induces autoimmune Goodpasture syndrome. Proc Nut/ Acucl Sci USA 1994; 91: 6201-6205.

28. Maloy WL, Coligan JE, Paterson Y . Production o f antipeptide antisera. In: Coligan JE, et al., eds. Current Protocols in Immunology, 9.4.1 Peptides. New York: John Wiley, 1991.

29. Kishiro Y , Kagawa M, Naito I , Sado Y. A novel method of preparing rat-monoclonal antibody-producing hybridomas by using rat medial iliac lymph node cells. Cell Strucf Funcf 1995: 2 0 151-156.

30. Muthukumaran G, Blumberg B, Kurkinen M. The complete primary structure for the al-chain of mouse collagen IV. Differential evolution of collagen IV domains. J B i d Chem 1989; 264 6310-6317.

31. Miner JH, Sanes JR. Collagen IV n3, n4, and n5 chains in rodent basal laminae: sequence, distribution, association with laminins, and developmen- tal switches. J Cell Biol 1994; 127: 879-891.

32. Morrison KE, Germino GG, Reeders ST. Use of the polymerase chain reaction to clone and sequence cDNA encoding the bovine n3 chain of type IV collagen. J BioI Chem 1991; 266 3439.

33. Mariyama M, Kalluri R, Hudson BG, Reeders ST. The a4(IV) chain of basement membrane collagen: isolation of cDNA encoding bovine a4(IV) and comparison with other type IV collagens. J Bid Chem 1992: 267: 1253-1258.

34. Yamaguchi N. Sato N, KO JS, Ninomiya Y. Cloning of al(1V) and n2(IV) collagen cDNAs from rabbit corneal endothelial cell RNA. Invest Ophthul- mol Vis Sci 1991; 3 2 29242930,

35. Kamagata Y, Mattei M-G, Ninomiya Y. Isolation and sequencing of cDNAs and genomic DNAs encoding the a4 chain o f basement membrane collagen type IV and assignment of the gene to the distal long arm of human chromosome 2. J Biol Chem 1992; 267: 23753-23758.

36. Zheng K, Thorner PS, Marrano P, Baumal R, Mclnnes RR. Canine X chromosome-linked hereditary nephritis: a genetic model for human X-linked hereditary nephritis resulting from a single base mutation in the gene encoding the a5 chain of collagen type IV. Proc Nut1 Arud Sci USA 1994; 91: 3989-3993.

37. Sado Y . Kagawa M. Rauf S. Naito 1. Moritoh C. Okieaki T. lsoloeous monoclonal antibodies can induce anti-GBM glomerulon-ephritis in rays J Pathol 1992, 168 221-227

38. Kleppel MM, Santi PA, Cameron JD, Wieslander J, Michael AF. Human tissue distribution of novel basement membrane collagen. Am J Puthol 1989, 134 813-825.

39. Nakanishi K, Yoshikawa N, Iijima K, et a/. lmmunohistochemical study of al-5 chains of type TV collagen in hereditary nephritis. Kidney Int 1994; 4 6 1413-1421.

40. Derry CJ, Pusey CD. Tissue-specific distribution of the Goodpasture antigen demonstrated by 2-D electrophoresis and Western blotting. Nephrol Dial Transplant 1994; 9 355-361