Different mutations in the COL4A5 collagen gene in two patients with different features of Alport syndrome

Kidney International, Vol. 42 (1992), pp. 83—88

Different mutations in the COL4A5 collagen gene in two patients with different features of Alport syndrome HUBERT J.M. SMEETS, Jos J. MELENHORST, HENNY H. LEMMINK, COCK H. SCHRODER, MARCEL R. NELEN, JING ZHOU, SIRKKA L. HOSTIKKA, KARL TRYGGVASON, HANS-HILGER ROPERS, MAAIKE C.E. JANSWEIJER, LEO A.H. MONNENS, HAN G. BRUNNER, and BERNARD A. VAN OOST Department of Human Genetics and Department of Pediatrics, University Hospital Nijmegen, Nijmegen, The Netherlands; Biocenter and Department of Biochemistry, University of Oulu, Oulu, Finland; and Department of Human Genetics, University of Amsterdam, Amsterdam, The Netherlands

cases reported [6], nephritis appeared within nine months after transplantation. Identification of patients predisposed to this anti-GBM disease has not yet proven possible. Recently, the cloning of the basement membrane type IV aS patients. We searched for mutations in this gene in 18 unrelated collagen chain (COL4A5) gene has been reported [7]. AssignDifferent mutations in the COL4A5 collagen gene in two patients with different features of Alport syndrome. Alport syndrome is a hereditary renal disease in which progressive renal failure is often accompanied by sensorineural deafness and ocular abnormalities. Recently, mutations were detected in the type IV collagen a5 chain gene in Alport syndrome

ment of this gene to the region on the long arm of the

patients, and in two patients abnormalities were detected. In the gene of patient BB we identified a complex deletion, which included the exons

X-chromosome, where the AS gene must be located, together with the almost exclusive location of the gene product in the GBM, made the involvement of the COL4A5 gene in AS very likely. This was confirmed by the identification of mutations in the COL4A5 gene in 3 out of 18 AS kindreds [8, 9] and 3 out of 38 in another study [10]. We searched for mutations in the COL4A5 gene in a group of 18 AS patients. First, we have screened Southern blots from the genomic DNAs of our patients for major gene rearrangements. From published data [8—101 it seemed likely that major gene rearrangements were only a minority of the mutations involved. Therefore, we also initiated a search for point mutations. A combination of in vitro amplification of the exons coding for the NC-domain with a search for single stranded-DNA conformation polymorphisms (SSCP) [11, 121 was applied. The conformation of denatured, single stranded DNA is sequence dependent and determines the electrophoretic mobility of the DNA molecule. Most mutations alter the conformation of a DNA fragment and will result in mobility shifts on non-denaturing

encoding the non-collagenous domain and part of the collagenous region. This patient showed early onset nephritis (end-stage renal disease at 17 years) with deafness. Within a year after receiving a kidney from an unrelated donor, he developed an antiglomerular basement membrane nephritis. In patient WJ a point-mutation was detected, changing a tryptophane into a serine in the non-collagenous domain. His clinical features are milder (renal failure at 33 years, no

hearing loss), and a recent renal allograft did not provoke anti-

glomerular basement membrane disease. These initial data suggest that differences in the extent of disruption of the non-collagenous domain

may correlate with the severity and/or heterogeneity of Alport syndrome and with the development of nephritis in renal allografts.

Alport syndrome (AS) is a progressive hereditary glomerulonephritis with a heterogeneous clinical picture. Patients can be classified by their age at end-stage renal disease (ESRD) and the presence of accompanying features, such as hearing loss, ocular malformations, macrothrombocytopenia and leiomyomatosis of

the esophagus [1]. Typical ultrastructural anomalies of the

glomerular basement membrane (GBM) include characteristic polyacrylamide gels. In our group of 18 AS patients, we lesions of the lamina densa. The clinical heterogeneity of AS detected two patients with a mutation in the 3'-end of the may reflect the involvement of different genes, because COL4A5 gene. Both the mutations and the clinical picture differ X-linked dominant, autosomal dominant and autosomal reces- between these patients, which indicates that analysis at the

sive modes of inheritance have been reported for AS [2]. gene level may become an indicator of prognosis for this However, the vast majority of AS pedigrees published thus far disease. show X-linked dominant inheritance [1, 3—51. Some transplanted male patients develop a post-transplant anti-GBM nephritis, leading to irreversible graft failure [6]. In 11 out of 13

Methods

Patients Since the age of 10 patient BB was known to suffer from proteinuria and hematuria. He developed sensorineural deafness, Renal biopsies show typical abnormalities of the glomerular basement membrane, consisting of irregular thickening and

Received for publication October 10, 1991 and in revised form February 5, 1992 Accepted for publication February 6, 1992

© 1992 by the International Society of Nephrology 83

84

Smeets et a!: COL4A5 mutations in Alport syndrome NC-domain

Collagenous domain

COL4A5 chain

PL31

HT14 —

JZ4 —

cDNA clones

MD6 PC4b

6kb

Fig. 1. cDNA clones of the COL4A5 gene.

thinning, together with a multi-lamellated aspect with in- exon 2: AS2F: 5'-TTG CGG CAC ATT TTT CCT TGT CTF-3' AS2R: 5'-GGA CCT GAA 'ITA AAG CTA TAA tramembranous granules on electronmicroscopy. The disorder was progressive and at the age of 17 the patient reached GCA-3' end-stage renal disease (ESRD). Transplantation with the kid- exon 3: AS3F: 5'-ATT ATG TTC CTT CTC CTT TTC CTI'-3' AS3R: 5'-ATG ACA AAT GCA AGG AAG AGT ney of an unrelated donor was followed by rapidly progressive antiglomerular basement membrane nephritis, leading to loss of GTA-3' the transplant within a few days of seven months after grafting exon 4: AS4F: 5'-CTT TAC TGT TrF CTC TCC AAA TCT-3' AS4R: 5'-AAA AGT CAC AGC TAA ATC AAT [13]. His affected maternal grandfather died from renal failure at GCC-3' the age of 26 years. Patient WJ had suffered from a progressive loss of renal exon 5: ASSF: 5'-TCT TGT ATA CTG ATT ATT TCG TGG-3' AS5R: 5'-CAG TAG GAA ATT AGA TAT TGA function since the age of 27. Renal disease has been slowly TTA-3' progressive with rapid deterioration of kidney function from the age of 33. A renal biopsy showed similar abnormalities as were observed in the first patient. At age of 35 the patient received a Mutation scanning was performed according to the procedure cadaveric kidney transplant. This transplant has functioned of Orita et al [11, 12]. Single strand DNA was electrophoresed adequately for nine months. Examination by standard audio- under non-denaturing conditions on a 5% polyacrylamide semetry did not show hearing loss. The patient has a similarly quencing gel (with and without 10% glycerol) in the cold room affected brother. The pedigrees of both patients are consistent (4°C) at 5 watts for 20 hours or at 40 watts for three hours. with X-linked inheritance with more severe disease in affected Autoradiographic exposure was overnight. Sequence analysis of double-stranded PCR products was performed using the males than in affected females. sequencing kit of Pharmacia. Template DNA was generated Materials and methods according to the method described by Casanova et al [19]. PCR Chromosomal DNA was isolated from peripheral blood [141. products of COL4A5 exon 4 were digested with BsrI (New DNA (10 g) was digested with EcoRI, PstI and TaqI (Gibco, England Biolabs Inc., Beverly, Massachusetts USA) and anaBRL), resolved by agarose gel electrophoresis and immobilized lyzed on 4% agarose gels, stained with ethidium bromide. on Gene Screen p1USTM (Du Pont, NEN, Boston, MassachuResults setts, USA) by Southern blotting, according to standard proceSouthern blot analysis of DNAs of AS patients was perdures [151. The COL4A5 cDNA probes pJZ4, pHTl4, pPL31, pMD-6 and pPC-4b (Fig. 1; [71, Zhou et a!, unpublished formed to screen for major gene rearrangements. In only one observations), which cover the entire cDNA, were 32P labeled case were structural alterations detected in the COL4A5 gene. by random primer synthesis [16]. Prehybridization and hybrid- No hybridization signals were visible with the probes pMD-6 ization were performed at 65°C in 0.5 M Na,HPO4INaH2PO4 and pPC4-b in the DNA of patient BB (Fig. 2A). Most signifi(pH = 7.2)17% (wtlvol) SDS. Filters were washed three times cant was the absence of the entire NC-domain in the COL4A5 for five minutes and once for 30 minutes at 65°C in 40 mM chain of this patient. His gene was not deleted completely, Na2HPO4INaH2PO4IO. 1% SDS. For autoradiography, filters because probe pPL3 I and pHTl4 showed hybridization with his

were exposed to Kodak X-ray films for one to four days at DNA, though many of the bands, which were visible in the —70°C. DNA of controls, were lacking (Fig. 2B,C). The data are COL4A5 exons 1-5 were amplified in vitro using the PCR suggestive for the presence of two deletions in his gene. The 3' methodology [17]. Exon flanking primers (Pharmacia, Uppsala, part of the gene was completely absent, but pPL3 1 detected Sweden) were synthesized based on the sequence data of Zhou EcoRI fragments of 2.6 kb (Fig. 2B) and 1.4 kb (not visible), et al [18]. The primer sequences are (F indicates the forward indicating that most likely the exons 16 or 17 to 20, which are primer, R the reverse primer): covered by the 5' part of this eDNA probe, were present. However, probe pHTl4, which lay at the 5' side adjacent to exon 1: AS1F: 5'-GAT CTG ATT GTC TTA TTT CTT ATT-3'

AS1R: 5'-TAA AAC ACA AAA GGA ATT CYI' CAA-3'

pPL3 1, also lacked a number of the bands, which were visible in the DNA of control individuals (Fig. 2C). Because there was no map available of the 5' end of the gene, it was unclear which

85

Smeets et al: COL4A5 mutations in Alport syndrome

BB

C

.

V — — 11.5 —-

kb

ee .me * —.

C

kb

2222 2217 2221 2218 2220

ww

C

7.5

____ _____

• — g• !i

S

11.5

_____

9.4

4 -.6.6

J1 i4 2.4

kb

, .#

-

5.1

48

S

.-. 4.5

'I. fm a —c 1.2

5.1

48 415

1*

2.8

A C B Fig. 2. A. Southern blot analysis of EcoR! digested DNA of patient BB and his family with probes pMD-6 and pPC4-b. These probes hybridize with COL4A5 exons 1-11 and detect EcoRI fragments of, respectively, 7.5 kb, 6.6 kb, 2.4 kb, 1.4kb and 2 overlapping fragments of 1.2 kb. The DNA of patient BB (2218) does not show any hybridization. B. Southern blot analysis of EcoRI digested DNA of patient BB and 3 male controls (C) with probe pPL3I. The DNA of the patient displays a normal 2.6kb fragment and ajunction fragment of 6.6 kb. C. Analysis of the same blot as in 2B with probe pHT14. Five of the bands, which are visible in the DNA of controls are missing in the DNA of patient BB. However, four other bands and two junction fragments (6.6 kb and 5.4 kb) are present.

of the exons were missing. Probe pPL3 1 detected one junction fragment of 6.6 kb and probe pHT14 two junction fragments of 6.6 kb and 5.4 kb (Fig. 2B,C). These junction fragments were also present in the DNA of the mother and sister of the patient, who both displayed hematuria, and revealed carriership of the same complex deletion. The junction fragments were absent in the DNA of his healthy grandmother (not shown). Probe pJZ4,

tryptophane codon into a serine codon. The single base muta-

tion also destroyed a BsrI restriction site, which facilitated DNA studies in the family the patient. As shown in Figure 5, the absence of the BsrI site co-segregated with AS. Besides the father of patient, 50 healthy controls (25 males and 25 females)

were tested for the BsrI polymorphism. All normal controls showed presence of the BsrI site. In the other 16 AS patients no

covering the 5' part of the gene, gave a completely normal mutation has been detected so far. picture (data not shown). We screened the five exons coding for the NC-domain for small mutations, using single-strand conformation polymor-

Discussion

In two patients belonging to clinically different types of AS, phism (SSCP) analysis ll, 12]. This method revealed the we have detected mutations in the 3' part of the COL4A5 gene. presence of a mutation in exon 4 (numbered from the 3' end) in For patient BB, two deletions in his COL4A5 gene led to the patient WJ (Fig. 3). Compared to control samples, the mobility absence of the normal NC-domain and part of the collagenous of the amplified fragment was clearly altered. Direct sequencing region. This complex deletion is different from the ones reof the PCR products revealed a single base mutation in the ported before [8, 10], and the gene product might either be a DNA of the patient, which was not present in the DNA of shorter collagen chain or contain a nonsense stretch of amino controls (Fig. 4). His affected brother carried the same mutation acids. Therefore, a relationship to dysfunctioning of the GBM and his mother was heterozygous. The single base alteration led and the development of renal disease is quite obvious. The to a substitution of a guanine for a cytosine, changing a second patient (WJ) shows only a single base alteration in the

86

S,neets et at: COL4A5 mutations in A/port syndrome

SSCP ANALYSIS EXON 4 Male

Male patients

controls

II 1

2

3

4

5

6

7

6

9

10

11

ii

12

Female Female patients

13

14

control

15

Fig. 3. SSCP analysis of cxv,, 4 in AS patients and controls. Exon 4 of patient Wi (lane 0) shows an altered mobility. As expected, no amplification product is visible for the deletion patient BB (lane 6).

372/AP male

control

3052

3053

3060 P

T

C

C

T

Leu

T

C

C C

Ser

C

T T C

T T

GATC

GAIC

A T

Leu

C

T A

Ser

Tyr

C

0

Trp A

T T

Ser Ty r

C

T T

GAb

A

1

T

Tyr

Ser Tyr

Fig. 4. Sequence analysis of exon 4 DNA of patient WJ, his affected brother, his mother and a male control. The G to C transversion in the

patients changes a Trp codon (TGG) into an Ser codon (TCG). The mother is heterozygous and displays both a G and a C at that site.

fourth exon of the NC-domain (counted from the 3' end). The stituted tryptophane residue is highly conserved at that position expected effect of the amino acid substitution would be a mild in both halves of the NC-domain in all type IV collagens from alteration of the hydrophobicity of the protein segment, with as Drosophila to humans [20—23]. Evolutionary conservation is a yet unknown consequences for structure and function of the strong indication of functional relevance of amino acids. Secprotein. However, there are two lines of evidence that this base ondly, we tested both the family of patient WJ and a number of substitution is very likely the causal mutation. First, the sub- controls. In the family we found co-segregation of the mutation

Smeets et a!: COL4A5 mutations in Alport syndrome

JJ 3060

3053

LU

372/AP

87

antibodies of this patient was similar to Goodpasture antibodies

and a monoclonal antibody directed against the Goodpasture antigen (MCA-P1) [131. A large deletion is only one of the causes leading to the absence of the entire protein or part of it, a variety of other, small mutations can have a similar effect. For

3054

3052

example, small deletions or insertions, which cause a shift of the reading frame, will result in a nonsense protein, and point mutations, which introduce premature stop codons or prevent normal transcription, will lead to truncation or absence of the protein. Our data suggest that AS patients who have one of those mutations in their COL4A5 gene may be at risk for the development of anti-GBM nephritis upon kidney transplantation. It is obvious that more data on mutations in the COL4AS

3061

gene of patients with anti-GBM nephritis need to be collected to estimate this risk properly. Two mechanisms were hypothesized to explain the development of anti-GBM nephritis in a subgroup of AS patients [241. The first hypothesis assumes that all patients lack the immunological tolerance for the Alport antigen, but that the occurrence 238 bp

128 bp 110 bp

Fig. 5. BsrI digestion of exon 4 of patient WJ and his family. The PCR product is 238 bp long. Digestion with BsrI generates two fragments of 128 and 110 bp in controls (fe., 3054). The mutation has destroyed the BsrI site in the DNA of patients 3053 and 3060. Their mother (3052) and sister (3061) are heterozygous.

with the phenotypic expression of the disease. In contrast, in 50

healthy controls (75 chromosomes) no mutation was found. Therefore, we conclude that this mutation is not a neutral polymorphism, although it still remains possible that other mutations can exist within the remaining unscanned part of the gene.

of anti-GBM nephritis is dependent on the immunological setting of the patient. The second hypothesis postulates a subgroup of patients with specific mutations in their gene that prevent the expression of certain epitopes of the protein. This latter possibility is supported by our data. It is evident that the nature of the mutations in the COL4A5 gene is only one of the

factors involved; other endogenous and exogenous factors contribute as well to the provocation and severity of the immune response, leading to anti-GBM nephritis [13, 24]. A clear demonstration of the involvement of exogenous factors forms a familial case of four male patients, who were all transplanted, but of whom only two lost their allografts because of anti-GBM nephritis [25]. The two mutations described here increase the total number

reported to eight, of which seven have been characterized in detail. These mutations are all different, which helps to explain the clinical heterogeneity of X-linked AS. The two deletion patients (patient BB reported here and patient 197701 from kindred EP [8]) both suffer from severe hearing loss, whereas hearing loss is absent in our patient WJ. This corroborates the suggestion of Barker et al [8] that severity of deafness is directly

correlated with the severity of the defect in COL4A5. A wide The two patients described here have different clinical fea- variety of mutations implies that each AS kindred might carry a

tures. Early onset of ESRD and hearing loss was found in different mutation in COL4A5 and that for each kindred a patient BB, who carries a complex rearrangement in his specific mutation must be identified. For osteogenesis imperCOL4A5 gene, leading to the absence of a large part of the fecta (01) similar findings were reported [26]. More than 70 normal COL4AS chain. The course of the disease is less severe

mutations have been identified in type I collagen genes as the molecular cause of different forms of 01 and insight into the relation between mutation and severity of the disease is growing. Still, the picture is far from complete and it is estimated [26] that the generation of an initial mutation map of type I collagen reason for this is not clear at present. The kidney might be more genes requires the analysis of the phenotypic effects of at least sensitive to alterations of the COL4A5 chain, although tissue- 200 mutations. The size of the COL4AS gene is similar to those specific isoforms of the COL4A5 chain may exist. Further genes and the number of mutations needed for the establishstudies of COL4A5 expression in different tissues will help to ment of a proper correlation between mutations in COL4A5 and solve these questions. AS phenotypes may be similar as well. The complex deletion in the COL4AS gene of patient BB In conclusion, our data suggest that gross rearragements of includes the entire NC-domain and, since the patient lacks the the 3' end of the COL4A5 gene will lead to clinically severe AS normal epitopes of the COL4A5 NC-domain, is probably one of with early onset nephritis and profound deafness. Mild mutathe factors involved in the development of anti-GBM nephritis tions on the other hand, like the missense mutation at the 3' end after kidney transplantation. The reactivity of the anti-GBM of the gene in patient WJ, will lead to a less severe phenotype.

in the second patient and no hearing loss is detectable. No major rearrangements have been found in his COL4A5 gene. The single base mutation affects the NC-domain, but the consequences appear to be limited to the glomerulus. The

88

Smeets et al: COL4A5 mutations in Alport syndrome

Moreover, our finding of a large deletion of the 3' end of the COL4A5 gene in a patient with a de novo anti-GBM nephritis after transplantation implies that a similar correlation may exist between the severity of disruption of the NC-domain and the development of anti-GBM nephritis. Acknowledgments

11. ORITA M, IWAHANA H, KANAZAWA H, HAYASHI K, SEKIYA T:

Detection of polymorphisms of human DNA by gel electrophoresis as single-strand conformation polymorphisms. Proc NatI Acad Sci USA 86:2766—2770, 1989 12. ORITA M, SUZUKI Y, SEKIYA T, HAYASHI K: Rapid and sensitive

detection of point mutations and DNA polymorphisms using the polymerase chain reaction. Genomics 5:874—879, 1989 13. VD HEUVEL LPWJ, SCHRODER CH, SAVAGE COS, MENZEL D, ASSMANN KJM, MONNENS LAH, VEERKAMP JH: The develop-

This work was supported by the Dutch Kidney Foundation (grant C90. 1022) and by grants from the Sigrid Juselius Foundation, the

ment of antiglomerular basement membrane nephritis in two children with Alport's syndrome after renal transplantation: CharacAcademy of Finland and the Finland Cancer Institute. terization of the antibody target. Pediatr Nephrol 3:406—413, 1989 14. MILLER SA, DYKES DD, POLESKY HF: A simple salting out procedure for extracting DNA from human nucleated cells. Nuci Reprint requests to Hubert J.M. Smeets, Ph.D., Department of Acids Res 16:1215, 1988 Human Genetics, University Hospital Nmegen, P.O. Box 9101, NL6500 HB N,jmegen, The Netherlands. 15. SAMBROOK J, FRITSCH EF, MANIATIS T: Molecular Cloning. A

Laboratory Manual (2nd ed). New York, Cold Spring Harbor References 1. HASSTEDT SJ, ATKIN CL, SAN JUAN AC: Genetic heterogeneity among kindreds with Alport Syndrome. Am J Hum Genet 38:940— 953, 1986 2. FEINGOLD J, Bois E, CHOMPRET A, BROYER M, GUEBLER M-C, GRUENFELD JP: Genetic heterogeneity of Alport syndrome. Kidney mt 27:672—677, 1985 3. BRUNNER H, SCHRODER C, VAN BENNEKOM C, LAMBERMON E, TUERLINGS J, MENZEL D, OLBING H, MONNENS L, WIERINGA B,

ROPERS HH: Localization of the gene for X-linked Alport's syndrome. Kidney mt 34:507—510, 1988

4. ATKIN CL, HASSTEDT SJ, MENLOVE L, CANNON L, KIRSCHNERN,

SCHWARTZ C, NGUYEN K, SKOLNICK M: Mapping of Alport syndrome to the long arm of the X chromosome. Am J Hum Genet 42:249—255, 1988

5. FLINTER FA, ABBs S, BoRRow M: Localization of the gene for classic Alport syndrome. Genomics 4:335—338, 1989

6. CAMERON JS: Recurrent primary disease and de novo nephritis following renal transplantation. Pedialr Nephrol 5:412—421, 1991 7. HOSTIKKA SL, EDDY RL, BYERS MG, HOYHTYA M, SHOWS TB,

Laboratory Press, 1989 16. FEINBERG AP, VOGELSTEIN B: A technique for radiolabeling DNA

restriction endonuclease fragments to high specific activity. Anal Biochem 132:6—13, 1983

17. SAlK! RK, GELFAND DH, STOFFEL S, SCHARF SJ, HIGUCHI R, HORN GT, MULLIS KB, ERLICH HA: Primer-directed enzymatic

amplification of DNA with a thermostable polymerase. Science 239:487—491, 1988 18. ZHOU J, HOSTIKKA SL, CHOW LT, TRYGGVASON K: Characteriza-

tion of the 3' half of the human type IV collagen aS gene that is affected in the Alport syndrome. Genomics 9:1—9, 1991 19. CASANOVA JL, PANNETIER C, JAULIN C, KOURILSKY P: Optimal

conditions for directly sequencing double-stranded PCR products with sequenase. Nucl Acids Res 18:4028, 1990 20. BLUMBERG B, MACKRELL AJ, OLSON PF, KURKINEN M, MONSON

JM, NATZLE JE, FESSLER JH: Basement membrane procollagen IV

and its specialized carboxyl domain are conserved in Drosphila, mouse and human. J Biol Chem 262:5947—5950, 1987 21. GUO X, KRAMER JM: The two Caenorhabditis elegans basement membrane (type IV) collagen genes are located on separate chromosomes. J Biol Chem 264:17,574—17,582, 1989

TRYGGVASON K: Identification of a distinct type IV collagen a chain with restricted kidney distribution and assignment of its gene to the locus of X chromosome-linked Alport syndrome. Proc NatI

22. HOSTIKKA SL, TRYGGVASON K: The complete primary structure of

AcadSci USA 87:1606—1610, 1990 8. BARKER DF, HOSTIKKA SL, ZHOU J, CHOW LT, OLIPHANT AR,

23. MORRISON KE, GERMINO GG, REEDERS ST: Use of the polymerase

GERKEN SC, GREGORY MC, SKOLNiCK MH, ATKIN CL, TRYGGVA-

SON K: Identification of mutations in the COL4A5 collagen gene in Alport syndrome, Science 248:1224—1227, 1990 9. ZHOU J, BARKER DF, HOSTIKKA SL, GREGORY MC, ATKIN CL, TRYGGVASON K: Single base mutation in a5(IV) collagen chain gene converting a conserved cysteine to serine in Alport syndrome. Genomics 9:10—18, 1991

10. BOYE E, VETRIE D, FLINTER F, BUCKLE B, PIHLAiANIEMI T, HAMALAINEN E-R, MYERS JC, BOBROW M, HARRIS A: Major

rearrangements in the aS(IV) collagen gene in three patients with Alport syndrome. Genomics 11:1125—1132, 1991

the a2 chain of human type IV collagen and comparison with the al(IV) chain. J Biol Chem 263:19,488—19,493, 1988

chain reaction to clone and sequence a cDNA encoding the bovine a3 chain of type IV collagen. J Biol Chem 266:35—39, 1991 24. KASHTAN CE, KLEPPEL MM, BUTKOWSKI RJ, MICHAEL AF, FISH

AJ: Alport syndrome, basement membranes and collagen. Pediatr Nephrol 4:523—532, 1990

25. KASHTAN CE, BUTKOWSKI Ri, KLEPPEL MM, FIRST MR, MICHAEL AF: Posttransplant anti-glomerular basement membrane nephritis in related males with Alport syndrome. J Lab Clin Med 116:508—515, 1990

26. BYER5 PH, WALLIS GA, WILLING MC: Osteogenesis imperfecta: Translation of mutation to phenotype. J Med Genet 28:433—442, 1991