WO1991019010A1

WO1991019010A1 - Method and probes for detection of alport syndrome

Info

Publication number: WO1991019010A1
Application number: PCT/US1991/004051
Authority: WO
Inventors: Karl Tryggvason; Sirkka Liisa Hostikka; Curt L. Atkin; David F. Barker; Arnold Oliphant; Mark H. Skolnick
Original assignee: University Of Utah
Priority date: 1990-06-07
Filing date: 1991-06-07
Publication date: 1991-12-12
Also published as: AU8060591A

Abstract

The present invention relates to a method for identifying an individual possessing a genetic defect associated with Alport syndrome which comprises analyzing the DNA of the individual to be tested by a method capable of detecting variation in a DNA sequence, and observing the presence or absence of a variation in the COL4A5 gene, in comparison with a normal COL4A5 gene.

Description

METHOD AND PROBES FOR DETECTION OF ALPORT SYNDROME

This work has been supported in part by the

National Institutes of Health, Grant No. DK 39497, and No.

RR 00064 Department of Energy Grant DE-F602-88ER60689, and a grant from the Hereditary Nephritis Foundation.

1. FIELD OF INVENTION The present invention relates to a method for early detection of Alport syndrome. In particular, the invention provides DNA probes capable of identifying individuals with Alport syndrome by detection of mutations in the COL4A5 gene.

2. BACKGROUND OF THE INVENTION

2.1. BASEMENT MEMBRANES

Basement membranes are a highly specialized part of the extracellular matrix and they form thin sheets that separate the cells of organs from the fibrillar connective tissues. The basement membranes serve a number of important biological functions in the body such as in the developing embryo where they play a role in cell differenti.ati.on duri.ng the formation of organs. They are also of importance for the correct regeneration of tissues following injuries such as during post-wound reformation of skin and nerves.

Of importance with respect to the present invention is the filtration role of basement membranes in kidneys and blood vessels. This function is exemplified by the renal glomerular basement membrane which makes up the filter between the capillary lumen and the urinary space and hinders the passage of blood cells and large macromolecules (proteins) to the forming urine.

The basement membranes are composed of several proteins, many of which are found only in these structures. Type IV collagen is the major structural component but other specific protein components include laminin, entactin (nidogen) and proteoglycans. Additionally, the basement membranes may contain fibronectin and type VII collagen which are also present in other extracellular matrices. The differences in the molecular composition of basement membranes in different tissues is not well known but a protein known as pemphigold antigen is probably only present in the basement membranes of skin. It is currently believed that there are several other proteins that are specific for basement membranes in certain tissues.

Type IV collagen is the major structural component of basement membranes and it can provide up to 60% of the structure. As is true for collagens in general, type IV collagen molecules contain three α chains that are coiled around each other to form a long triple-helical molecule that is about 1.5 n in diameter and about 400 n in length. At the carboxyl terminal end, the molecule has a large globular noncollagenous domain, called the NC-domain, that has a diameter of about 15 nm. Single type IV collagen molecules are linked with each other into a complex, flexible, network-like structure into which the other basement membrane componentε are bound (Timpl, Eur. J.

Biochem., 180:487-502, 1989).

It was previously thought that the major form of type IV collagen was only composed of two kinds of chains, αl(IV) and α2(IV), with the molecular formula [αl(IV)]_ . . . . 2(IV). The applicants have determined the entire ammo acid sequence of the both chains from man by applying DNA sequencing of cloned cDNA molecules (Soininen, et al. , FEBS

Lett. , 225:188-194, 1987; Hostikka and Tryggvason, J_^ Biol.

Chem., 263:19488-19493, 1988). The results showed that the ' αl(IV) chain contains 1,642 amino acid residues as compared with 1,676 residues in the α2(IV) chain. The carboxyl terminal NC-domains of both chains are very similar with 63% identical amino acid residues. The sequence homology of the two chains in the triple-helical region is 49%. The existence of two other distinct type IV collagen chains, α3(IV) and α4(IV), has recently been reported, indicating that type IV collagen must have several molecular component

(Butkowski et al. , J. Biol. Chem. , 262:7874-7877, 1987; Sau et al^, J^ Biol. Chem. , 263:13374-13380, 1988) .

Of importance with respect to the present invention was the recent discovery of yet another type IV collagen chain, termed α5(IV). cDNA clones that code for about 50% of the human protein sequence from the carboxyl terminal end were isolated and sequenced (Hostikka et al. ,

Proc. Natl. Acad. Sci. USA, 82:1606-1610, 1990) . Amino acid sequence comparison demonstrated that the α5(IV) chain is a distinct gene product which is more closely related to the l(IV) chain than to the α2(IV) chain. In the NC-domain, the sequence identity between the αl(IV) and 5(IV) chains was shown to be as high as 83% whereas it is only 63% identical with the same region from the Q2(IV) chain. For the collagenous domain, the sequence identities were 58% and 46%, respectively. The type IV collagen collagenous domain repeat sequences (Gly-Xaa-Yaa) are characteristically interrupted by noncollagenous sequences. All the interruptions in the collagenous domain of the α5(IV) chain coincide with interruptions in the αl(IV) chain but only partially with interruptions in the α2(IV) chain.

Using cDNA probes and both a somatic cell hybrid and _in s tu hybridization, the gene for the human type IV collagen α5 chain (COL4A5) was localized to the q22 region on the long arm of chromosome X (Hostikka et al. , Proc.

Natl. Acad. Sci. USA, §2:1606-1610, 1990). This is different from the genes (COL4A1) and (COL4A2) for the human αl(IV) and α2(IV) chains, respectively, which have both been assigned to the terminal end of the long arm of chromosome

13 (Boyd et al. , Hum. Genet. , 2i^:121~¹²⁵, 1986; Griffin et al. , Proc. Natl. Acad. Sci. USA, 8^:512-526, 1987). Due to their wide distribution in the body, basement membranes are frequently affected in both local and systemic diseases and, in many instances, the consequent pathological changes lead to severe clinical complications.

These diseases may be genetically inherited due to diseases that are due to gene mutations leading to an abnormal structure and function of the protein or they can be acquired, i.e., complications of diseases that do not primarily involve the basement membrane such as diabetes mellitus. Examples of inherited basement membrane diseases are: (1) the Alport syndrome, a hereditary nephritis caused by abnormal function of the basement membranes resulting in the passage of blood cells into urine (Hematuria) , eye lesions and hearing loss; and, (2) the congenital nephrotic syndrome that is characterized by extensive leakage of proteins through the renal glomerular basement membrane into urine (proteinuria) . Both diseases are fatal but can be treated by renal transplantation.

2.2. COLLAGEN RELATED DISEASES

At present, the occurrence of a number of heritable diseases have been attributed to one or more defects in collagen production (Prockop et al. , New Engl. J. Med. 311:376-386, 1984). For example, osteogenesis imperfecta is known to exist as at least four different . types of disease with defects in Type I procollagen apparently being causally associated with the various symptoms. Type I collagen is the most abundant collagen type found in tissues such as bone, tendon and ligaments where it forms fibrils and fibers which give tension strength to these tissues. Other major fibrillar collagen types are types II and III. Several kinds of mutations have been described in type I and III collagens in osteogenesis imperfecta (brittle bone disease) . For example, deletions of about 500 bases from one allele from the pro-αl(I) chain have been associated with the occurrence of Type I osteogenesis imperfecta. Thi lethal defect resulted in the production of half normal and half shortened chains, which would form disulfide bonds 5 among themselves but which could not form stable helices.

Another type of osteogenesis imperfecta variant exhibited two unrelated mutations, one each in the two alleles for pro-α2(I) chains. Many other different variants have also been observed, all associated with defects in Type I collagen.

Another collagen related disease is Ehlers-Danlos syndrome, which is characterized by joint hypermobility and skin thinness, hyperextensibility and fragility. Many of the types of Ehlers-Danlos syndrome have been associated

15with defects in Type III procollagen. In one case, for example, normal amounts of Type III procollagen can be produced but certain pro-αl(III) chains have an apparent amino acid insertion, and the Type III procollagen with these chains is very sensitive to digestion by proteinases.

20 In another type of variant, the rate of synthesis of pro- αl(III) chains is reduced.

Other types of collagen-related syndromes also exist, often with many phenotypic expressions of the basic syndrome. However, the pattern of these diseases suggests

25 that for each syndrome, a single gene controls the occurrence of the disease. Each kindred exhibiting symptoms of a particular syndrome may have a different mutation, deletion or insertion within that gene, causing a wide variety of expression of the basic disease state.

2.2.1. ALPORT SYNDROME

One such disease for which the basis has not, to date, been elucidated is Alport syndrome. This condition, _a5like those noted above, is a heterogeneous group of diseases; in this case, the diseases are a group of glomerulonephritides, which appear to be primarily X-linked.

(Atkin et al. in Diseases of the Kidney, 4th ed. Schreier and Gottschalk, Eds. Little Brown & Co., Boston, 1988, pp.

617-641) . The phenotypic expression of the disease is varied but certain features tend to predominate: terminal renal failure in males, sensorineural deafness, lenticonus, retinal abnormalities, megathrombocytic cytopenia and

Fechtner inclusions in granulocytes. A number of recent genetic linkage studies (e.g. , Atkin et al. , Am. J. Hum.

Genet. 42:249-255, 1988; Flinter et al. , Genomics 4:335-338,

1989; Szpiro-Tapia et al. , Hum. Genet. 8JL:85-87, 1988) have localized the gene for Alport syndrome as linked to RFLP markers near Xq22, in the middle of the long arm of the X chromosome. The disease has been shown to be associated with progressive ultrastructural abnormalities, such as patchy thickening and thinning of glomerular basement membrane (GBM) , and splitting of the lamina densa; these features have suggested a defect in the major structural component of GBM, Type IV collagen. (Spear, Clin. Nephrol.

1:336-337, 1973).

More recent observations using antibodies to study the noncollagenous (NC) domain of GBM Type IV collagen indicate that there may be some alteration or absence of a

Type IV collagen alpha chain (Kaεhtan et al. , Kidney Int.

^6:669-674, 1989; Savage et al. , Lab. Invest 60:613-618,

1989) . Certain of the chains of Type IV collagen, specifically αl(IV) and Q2(IV) have been previously localized on chromosome 13 (Boyd et al. , Hum. Genet. 74_:121-125, 1986; Griffin et al. , PNAS USA 84,:512-516, 1987) . The discovery of the type IV collagen a chain with gene location at the q22 region on chromosome X (Hostikka et al. , PNAS, 84_:1606-1610, 1990), indicated that this gene might be involved in Alport syndrome. However, preliminary data in this paper, based on Southern blot analysis of DNA fro one male with X-linked Alport syndrome, did not show any abnormalities in restriction enzyme pattern, indicating no rearrangement. Thus, the evidence available did not demonstrate the involvement of the COL4A5 gene in the disease. 5

At present, then, there still exists a need for proven method for conclusive determination of the existence of Alport syndrome in afflicted individuals, as well as in potential carriers. As yet, there has been no proven relationship between the disease and any particular gene. The present invention now provides the means by which individuals carrying this hereditary defect can be readily- identified.

3. SUMMARY OF THE INVENTION is . .

The present invention relates to a method for identifying individuals possessing a genetic defect associated with Alport syndrome. It has now been determine that mutations appearing within the COL4A5 gene are directl associated with the occurrence of the disease. In fact, a 0number of different mutations in the COL4A5 have been identified in association with various phenotypic expressions of Alport syndrome.

Overlapping clones that cover about 50 kb at the

3' end of the human COL4A5 gene have previously been 5 identified (Zhou et al. , Genomics, in press) . Utilizing a pool of insert fragments from these genomic clones and the previously described α5(IV) cDNA clones (Hostikka et al. , PNAS, £J4_:1606-1610, 1990; see copending U.S. Serial No. 377,238, filed July 7, 1989, which is incorporated herein by U reference in its entirety) as a probe, restriction fragments from DNA of normal individuals showed no fragment length variation. On the other hand, analysis of restriction fragments of afflicted individuals routinely showed ₅variation in one or more fragments. Thus, a diagnostic method is provided in which the DNA of an individual to be tested is analyzed using a method capable of detecting variation in a DNA sequence, and verifying the presence or absence of variation in the individual's COL4A5 gene, in comparison with a known normal COL4A5 gene. In one 5 embodiment this is done by RFLP analysis, wherein the analysis is accomplished by hybridizing separated restriction fragments obtained from a restriction enzyme digest of DNA of an individual to be tested with one or more detectable probes comprising a portion of the sequence of the COL4A5 gene, and comparing the restriction fragment pattern so produced with the restriction pattern obtained from an identical restriction enzyme digest of DNA normal individuals. However, any method known in the art to be useful in detecting alterations in a DNA sequence can also

15be employed for this purpose. For example, other useful detection methods include, but are not limited to, allele specific oligonucleotide techniques, single stranded conformation polymorphism analysis, gradient gel electrophoresis, and ligaεe amplification reaction.

20 The invention also relates to probes capable of specifically hybridizing with a particular deletion, insertion or mutation in the COL4A5 sequence. Such probes are capable of detecting specific defects in an individual suspected of possessing an Alport related defect. 25 . . .

The determination of the association between the disease and the COL4A5 gene also provides alternate means of identifying individuals carrying the trait. In addition to detecting the defects directly, it is also possible to detect defects by observation of neutral tightly linked 30 * - marker genes in individuals to be tested.

Finally, the knowledge that the COL4A5 gene causes Alport syndrome provides the basis for development of gene therapy for treatment or prevention of the condition.

_,. It is therefore contemplated, as part of the present invention, that Alport syndrome may be treated or prevented by alteration or substitution for the defective COL4A5 gene, or a product of the defective gene, so as to counteract the effects of the defect or prevent the development of such effects. 5

4. DESCRIPTION OF THE FIGURES Figure 1. Mapping of restriction sites in genomic clones with respect to exons at the 3' end of COL4A5. (A) The stippled boxes indicate exon regions of COL4A5, numbered from the 3' end. Bars labeled PL-31, PL-35 and MD-6 represent previously described cDNA clones which span the indicated exons. The position of EcoRI site in exon 1 is indicated for clones MD-6 and PL-35. Bars labeled

F-7, FM-13, MG-3, and MG-2 represent genomic phage vector

15 clones (Zhou et al. , Geno i.cε, i.n press) wi.th the posi.ti.ons of EcoRI sites marked "E". The lengths, in kilobases, of EcoRI fragments mentioned in the text are shown. The polyadenylation site at the end of the 3' untranslated region was chosen as the origin for the indicated length

20 scale (in kb) for the enti.re regi.on. Exons have been seguenced and intron sizes were also confirmed with heteroduplex analysis. (B) Map of Pstl and EcoRI sites for the region which includes the variable Pstl fragments observed in Alport Kindred P. The 2.2 kb Pstl fragment 25. . indicated is lost and a novel 1.9 kb Pstl fragment is found on mutation-bearing chromosomes in Kindred P. A potential novel Pstl site would result in shortening from the 5' end of this fragment.

Figure 2. Genomic Southern blots demonstrating apparent mutations in the COL4A5 gene in Alport-affected males. Lanes marked "M" contain molecular size standards which are, reading from the smallest visible band, 1.1, 1.5,

2.65, 4.3, 5.3, 6.0, 7.95, 10.6, 13.3 and 15.9 kb followed ₅ by a ladder of bands at 2.65 kb intervals. All other lanes are DNAs from individual Alport-affected males. DNAs were digested with the indicated restriction enzymes, transferred to Zetabind membranes and probed with 32P-labeled mixture of insert DNA fragments corresponding to cDNA clones Pl-31, PL-35 and MD-6 (see Fig. 1) as described (Barker et al. , Am. J. Hum. Genet. 3_6_:1159-1171, 1984). Lanes numbered "1" and "2" correspond to individuals 500081 and 500242 who show an altered Pstl fragment pattern (indicated by arrow to left of panel C) , but apparently normal Taql and EcoRI patterns. Lanes marked "3" correspond to individual 198002 who is missing a faintly hybridizing Taql fragment at the position of the arrow to the left of panel A, but has a normal Pstl fragment pattern. Lanes marked "4" represent individual 198801, who shows missing or significantly reduced intensity of bands with all three enzymes, at the positions of the ¹⁶arrows to th3- right of each panel.

Figure 3. Inheritance of the COL4A5 Pstl site variant in two portions of Kindred P. Southern transfers were prepared and probed as described in the legend to

Figure 2. Only the region of the pattern containing the

20 relevant 2.2 and 1.9 kb bands (arrows) i.s shown. A faintly hybridizing constant fragment is present at a position corresponding to approximately 1.95 kb. Obligate carrier mothers or females with diagnosed renal phenotype are shown as half-darkened circles. Unaffected females are clear

25 circles. Unaffected males are open boxes and affected males are darkened boxes.

5. DETAILED DESCRIPTION OF THE INVENTION

The present invention resulted from a study of a 30 number of Alport Kindreds; DNAs from one or two affected male members from each of 18 independently ascertained

Alport Kindreds were examined by conventional Southern blot analysis (Barker et al. , Am. J. Hum. Genet. 36:1159-1171,

₃₅1984; L.-C. Tsui et al. , Science 230:1054-1057, 1985) with a pool of cDNA probes comprising fragments encompassing the 3' end of the COL4A5 gene, to determine the presence of structural alterations in the COL4A5 gene. No fragment length variation was found in normal controls for EcoRI (63 chromosomes) , Taql (91 chromosomes) or Pstl (102 chromosomes) . Based on this observation, the further observation of any fragment pattern variation in Alport Kindreds is therefore a strong indication of a disease causing mutation. In three of 18 Kindreds tested, variant fragment patterns (see Figure 2) which appear to reflect 10_Sp_ecifi_{c anc}j distinct mutations in the COL4A5 gene were detected.

5.1. RFLP ANALYSIS The original discovery of the COL4A5 gene-disease

^association was made by RFLP analysis. The probe utilized in determining fragment length variation comprised a pool of cDNAs previously described (Hostikka et al. , PNAS USA 87: 1606-1610, 1990) . The insert fragments were obtained from cDNA clones MD-6, PL-35 and PL-31. On June 6, 1^'990, pooled

^samples of the plasmids containing the relevant inserts were deposited with the American Type Culture Collection, Rockville, Maryland, under accession number ATCC 40827. Separation of the inserts can be achieved by transformation of an appropriate host bacterium, such as E_^ coli strain

2°JM83, or other suitable strain, with the plasmids, isolation of the DNA from the transformants and identification of one member of each class by agarose gel electrophoresis. Collectively, these clones cover approximately 50 kb at the

3^# end of the human COL4A5 gene. A map of the restriction

30sites in these clones relative to the exons of the 3' end of the gene is shown in Figure 1. These probes permit the identification of variation in the 3' end of the COL4A5 gene. Practice of the method is not limited, however, to the use of inserts from these specific clones from these 5 probes, but may also be performed with nucleic acid fragments having substantial homology with these fragments. This can be readily determined by hybridization of the fragments to be used with relevant fragments from the disclosed clones. Alternately, oligonucleotide probes can **be prepared based on the cDNA sequence disclosed in the Hostikka et al. reference (supra) for COL4A5; these probes can in turn be used to screen other cDNA libraries to obtain other useful clones.

The mutations reported herein were found as the

10result of restriction digests using EcoRI, Taql and Pstl. However, any one of the numerous restriction enzymes which is capable of cutting mammalian DNA may also be used. Fragment separation is most commonly carried out by gel electrophoresis in agarose, any sizing method, including

'^methods based on charge and size, is sufficient and is encompassed in the invention, since sizing methods for DNA fragments are well known in the art (e.g. Southern, Meth. Enzymol. 68:152-176, 1979).

The probes utilized in this method may be

^labelled by any means" known in the art. Although a common method is radioisotope labelling, many nonradioactive labels are also available. For example, the method can also be performed with chemiluminescent molecules, bioluminescent molecules or detectably labeled ligands such as biotin.

25 Labelling techniques are well known in the art and several are described (see for example, in Keller and Manak, DNA Probes, Macmillan Publishers, Ltd., 1989).

5.2. ALTERNATE METHODS OF DETECTION

30 . . .

The previous section describes the use of DNA probes encompassing all or part of the normal COL4A5 gene to detect variations within the gene indicative of Alport syndrome. However, this is not the only method by which the

Alport-related defects can be determined. 35 Given the knowledge of a particular mutation being associated with Alport syndrome, it is also possible to devise probes which will hybridize directly with the defective portion of the gene, or alternately, probes which will only hybridize with normal COL4A5 genes. The availability of such probes thus provides a means of direct detection of defective DNA, by contacting a probe capable o selectively hybridizing with a region of DNA comprising a specific mutation, deletion or insertion in the COL4A5 gene, with DNA of an individual to be tested, and observing the '"presence or absence of a hybridization reaction. For example, allele specific oligonucleotide (ASO) technology has been used to detect single base mismatches in DNA (Wallace et al. , Nucl. Acids Res. 6_:3543-3557, 1979; Conner et a , PNAS USA 80:278-282, 1980).

15 Methods known m the art for DNA polymorphism analysis also include ligase amplification reaction

(EP 336 731) , single stranded conformation polymorphisms .

(SSCP) (Orita et al^ Genomics 5_^:874-879; PNAS USA 86_:2766-

2770, 1989) , and gradient gel electrophoresis. Alternate n methods will be recognized by those skilled in the art.

It is also possible to identify the presence of a defective Alport gene by observation of a tightly linked marker encoding a neutral polymorphism not directly involved in the disease itself. The speed, efficiency and 5reliability of the genetic linkage studies can be improved by the identification of more highly informative markers which lie near or within the COL4A5 gene and which may be assayed by PCR and other methods. For example, simple repeats have been shown to be frequently associated with 0 length heterogeneity (Weber and May, Am. J. Hum. Genet.

44_^:388-396, 1989). Additionally, genomic clones can be restricted, blotted and hybridized with poly (dC-dA)/poly (dG-dT) and with an Alu-repeat probe to identify fragments containing CA tracts on simple repeats associated with Alu D elements (Zuliani and Hobbs, Am. J. Hum. Genet. 46:963-969,

1990) . Sequencing and PCR amplification of regions containing repeats and PCR amplification of samples of DNA from unrelated individuals can be used to determine the presence of useful variation for marker purposes. 5

5.3. EVALUATION OF ALPORT KINDREDS Studies on inheritance of Alport Syndrome and family pedigrees of a number of affected Utah kindreds have •previously been described by Hasstedt et al. (Am. J. Hum. ¹⁰Genet. 3J3:940-953, 1986 and O'Neill et al^, Ann. Intern Med. :176-182, 1978). The observations of the present study were also based on members of some of the same families.

5.3.1. Utah Kindred EP

With the pooled cDNA probe, DNA from affected male 197701 (lanes marked "4" in Fig. 2) from Utah Kindred EP (Hasstedt et al. , supra) , showed the absence of 7.5, 2.4 and 1.1 kb EcoRI fragments. He also showed missing 16, 10 and 5.5 kb Pstl fragments and 3.5, 3.2 and 2.7 kb Taql * fragments. The concordant absence of fragments produced by all three restriction enzymes indicates the deletion of a portion of the 3' end of the gene. It was shown that the deletion is internal to the COL4A5 gene by hybridization of the genomic probes F7, FM-13 and MG-2 (Fig. 1) to the

25patient DNA. The MG-2 probe, which spans the 3' end of the gene, hybridized with a 6.6 kb EcoRI fragment which is known to contain a 5' portion of exon 1 and the complete exons 2, 3 and 4 as counted from the 3' end (Fig. 1A) . In contrast, the FM13 probe, which contains exons 5 through 10,

30 hybridized to several fragments in the control samples which were not present in 197701. The fragments affected include the 2.4 kb fragment containing exons 5 and 6, the 7.5 kb fragment containing exons 7, 8 and 9 and a 2.6 kb genomic fragment which is 5' to exon 10. The presence of the 5.5 kb EcoRI fragment containing exons 12, 13 and 14, and the 3.6 kb fragment which is 3' to exon 11 was confirmed by hybridization of the patient sample with the genomic probe F7. Thus, exons 1-4 and exons 11 and upstream are intact in 197701 and the deletion affects the region containing exons ⁵5 through 10.

The deletion in Kindred EP is a mutation that would result in an alpha 5 (IV) chain with severely altered structure. The sequence of the exons in this region (Hostikka et al. , supra;) demonstrates that if the deletion O includes all of exons 5 through 10, the reading frame would not be changed. However, the variant chain would lack 193 residues from the carboxyl terminal end of the collagenous domain and 47 residues from the non-collagenous (NC) domain. The NC domain is essential for triple helix formation and cross-linking of type IV collagen molecules. Consequently, the abnormally short variant chain would not be properly incorporated into homo- or heterotrimers to form a normal type IV collagen network. If a deletion endpoint occurs within exon 5 or 10, then it is likely that the effects on RNA splicing or translation frame would result in more extreme changes in the gene product.

5.3.2. Utah Kindred P

The extensively studied Utah Kindred P (e.g., 5 Atkin et al. , supra, O'Neill et al. , supra; O'Neill et al. ,

Path. Res. Pract. 168:146-162, 1980; Hasstedt & Atkin, Am.

J. Hum. Genet. 3_5:1241-1251, 1983; Hasstedt & Atkin, supra,

Kashtan et al. , supra) was represented by affected males

500081 and 500242. Their DNA Pstl fragment patterns (lanes 0 marked "1" and "2" in Fig. 2) show the absence of a 2.2 kb fragment and the appearance of a 1.9 kb fragment not found in normals or affected individuals from other kindreds.

Both the 2.2 kb and 1.9 kb fragments hybridize with the cDNA inserts of clones PL-35 and MD-6. PL-35 and MD-6 contain 5 coding sequences extending in opposite directions from the EcoRI site in exon 1 (Fig. IB) and therefore, the affected fragment must be the 2.2 kb Pstl fragment that contains the 5' end of exon 1 as well as exons 2 and 3 (Figs. 1A and IB) .

The association of the Pstl variation with the ° lport phenotype was confirmed in 122 individual members of Kindred P by probing Southern blots of Pstl-digested DNA samples with the pooled cDNA probe. Fragment patterns, exemplified in Fig. 3, showed full concordance of clinically diagnosed Alport syndrome with the 1.9 kb Pstl variant Ofragment. In particular, all 23 gene-carrier mothers were found to carry one copy of each of the 1.9 kb and 2.2 kb fragments which segregated to a total of 45 sons in complete linkage with the presence or absence, respectively, of the Alport phenotype. Within Kindred P all 30 affected males ' carried only the 1.9 kb fragment while all 32 unaffected males had only the 2.2 kb fragment.

The molecular basis for the Pstl site variation appears to be a point mutation of a guanine to a cytosine at the 5' end of the third exon (from the 3' end of the gene).

20 Specifically, with reference to the cDNA sequence published in Hostikka et al. , (supra) , the mutation is in G-1952. This converts the second cysteine (Cys-651) in the NCI region to a serine. This change from guanine to cytosine generates in the gene two new restriction sites, a Bglll and

25a Pstl. The absence of the fifth cysteine would be enough to cause an abnormal folding of the NC-domain which is essential both for triple helix formation and intermolecular crosslinking.

30

5.3.3. Utah Kindred 1970

Initial examination of affected male 197002 from Utah Kindred 1970 shows the absence of a faintly hybridizing 6.1 kb Taql fragment (lanes marked "3" in Fig. 2). However,

35 subsequent studies have shown the presence of a polymorphis indicating the presence of an as yet uncharacterized mutation.

5.3.4. Discussion ° The present results demonstrate three different mutations in the COL4A5 gene in three kindreds with Alport syndrome, supporting the conclusion that mutations in this gene account for at least a portion and possibly all of X- linked Alport. The 5' portion of the α5(IV) gene has not

Oyet been characterized and it is possible that mutations in that region could be responsible for the disease in kindreds that did not show an abnormal Southern pattern with the 3' cDNA probes used here. Furthermore, based on the observed variation in mutation patterns, it is reasonable to expect

'°that other mutations remain to be detected in the 3' end of the gene.

The present results also constitute the first example of a human genetic disease affecting a basement membrane component and characterization of the different *"mutations may shed light on the role of the type IV collagen chains in basement membrane structure. The present data suggest that a wide variety of mutations of the COL4A5 gene will be found in Alport syndrome, as in osteogenesis imperfecta, where mutations of varying phenotypic severity

25occur in COL1A1 and COL1A2 (Prockop et al. , supra; Sykes et al. , Am. J. Hum. Genet 4_6:293-307 1990). Basement membranes in different tissues must have differences in structural composition reflecting biological function; monoclonal antibodies to NC fragments of GBM type IV collagen indeed 0 showed organ-specific distribution in basement membranes.

It has generally been assumed that the major chains, αl(IV) and α2(IV), are ubiquitous and generally essentially structural components and, at present, there is no evidence for their tissue specific distribution. In contrast, region 5 specificity has been shown for the α3(IV) and 5(lV) chains, at least in the kidney where the α5(IV) chain is almost exclusively located in the GBM (Hostikka et al. , supra) . Although it has been shown that the α5(IV) chain is expressed in many other tissues such as in lung and spleen, **Alport patients have not been reported to have complications in these organs. The basis for the organ specificity and a molecular understanding of how a structurally defective α5(IV) chain results in the loss of GBM integrity and hematuria in Alport patients remain to be established. '0 The fact that Alport patients usually have moderate to severe hearing loss suggests that the α5(IV) chain is also an important component in the basement membrane of a sensory portion of the inner ear. The finding of mutations which lie entirely within the boundaries of COL4A5 in two kindreds cosegregating for nephritis and deafness demonstrates that mutations in a single gene can cause both phenotypes. Furthermore, the deafness of the individual with the large deletion is among the most profound that has been observed in Alport kindreds, implying

2 *Ω^υthat the severity of the deafness is directly correlated with the severity of the defect in COL4A5.

In view of the aforenoted heterogeneity in mutation, the present method is not limited to detection only of the three mutations already observed, but also

25encompasses detection of any mutation in the COL4A5 gene.

Detection of other mutations is well within the skill in the art, utilizing probes encompassing significant portions of the COL4A5 gene, and commercially available restriction enzymes. Although the mutations described herein have been

30 detected using Eco RI, Pstl, and Taql, the use of any restriction enzyme which will cut mammalian DNA is suitable to digest DNA for probing. The detection of additional mutations may be conveniently achieved by conventional Southern hybridization. Alternately, single stranded conformation polymorphisms (Orita et al. , Geno ics 5_:874- 879, 1989) may be used to detect small insertions, deletion and single base changes not found by conventional Southerns Moreover, any of the additional detection methods described above will be useful in detecting other mutations in the ⁵COL4A5 gene.

Moreover, although Alport syndrome is primarily human condition, the method can be applied to other animals For example, an Alport-like condition has also been reporte . in dogs (Thorner et al. , Kidney Int. 3_5:843-850, 1989). 0 The discovery of the association of the gene wit the Alport phenotype allows the use of the human COL4A5 gen to identify, by homology, the respective animal COL4A5 gene and the Alport-like phenotype in animals. Animal models ca be created based either on naturally occurring Alport-like

' defects in certain animals, or by insertion of a cloned defective gene into a convenient animal, such as a mouse or rat. Such animal models which exhibit Alport-like phenotyp will be useful for screening pharmaceuticals for human use.

⁰ 5.4. GENE THERAPY

Determination of the association of the COL4A5 gene with Alport syndrome provides the basic information necessary to develop programs of gene therapy. Given that the disease has been shown to be associated with deletions 5or mutations within the gene, a number of approaches can be taken to prevent the damaging effects in individuals so afflicted, or to prevent expression of the trait in the child of one or more parents carrying the relevant mutation.

For example, diagnosis of individuals having an 0 . . -

Alport-related variation permits manipulation of the germ cells or embryos of such individuals so as to negate the expression of the defective gene. This may be done by transforming the relevant cells, at an early stage of development, with vectors carrying the normal COL4A5 gene. 5 This permits expression in the transformed cells of normal type IV collagen, and assuming adequate passage of the normal gene through the cell lines derived from transformed cells, any or all of the ill effects of the expression of the abnormal gene may be avoided.

5 Alternately, grafts of functioning cells into the appropriate region of the body, depending on the nature of expression of the disease in the individual, may also be performed. Reverse transplants of fibroblasts producing functional Type IV collagen could also be made. Finally, O infusion of the appropriate type of collagen to affected body regions may also serve to alleviate or reverse the symptoms of Alport syndrome.

The following paragraphs provide one non-limiting example of the practice of the present method. However,

'⁵those skilled in the art will readily recognize possible modifications of the disclosed method which will fall within the scope of the claimed invention.

6. EXAMPLE * The techniques used m the following Example were performed in accordance with methods described in Sambrook et al. (Molecular Cloning, a Laboratory Manual, Cold Spring Harbor Laboratory Press, 1989) , unless otherwise stated.

The example describes the conduct of a Southern

25hybri.di.zati.on useful m determining the presence or absence of an Alport related mutation.

6.1. PREPARATION OF COL4A5-SPECIFIC PROBE

The probe which has been used to detect those

30 mutations found to date may be prepared as follows.

The bacterial strains which carry the three plasmids containing the COL4A5-specific segments known as

PL-31, PL-35 and MD-6 are grown in LB medium containing

50 μl/ml ampicillin. The plas id DNAs are purified from the 35 cell culture by any standard method which yields DNA of sufficient purity to cut with a restriction enzyme. Following the plasmid purification, 20 μg of each DNA sampl is digested with EcoRI under conditions suggested by the supplier. The digested DNAs are separated by size on an *>agarose gel. The fragments which correspond to the COL4A5 specific segments are removed from the gel, purified free o any contaminating agarose and pooled. About 100-250 ng of the pooled mixture may be used to produce a radioactive or other suitably labeled probe. A COL4A5-specific probe has Oalso been made that can be detected by chemiluminescence. The fragments may also be used separately for the purpose o visualizing only the part of the COL4A5 gene to which they correspond. The PL-31 segment alone appears to be better for visualizing the Taql variant that is difficult to see, than is the pooled probe.

6.2. PREPARATION OF HUMAN DNA SAMPLES AND SOUTHERN BLOTS A sample of blood is taken into an anti-coagulan tube from an individual to be tested. The nuclei from the onwhite blood cells are puri.fi.ed by lysi.s of the red blood cells and the exterior membranes of the white cells. DNA i purified by proteinaεe K digeεtion followed by extraction with phenol and precipitation with ethanol.

Following resuspension, the DNA is digested with

25 . . . . the restriction endoculease of choice, using the conditions provided by the suppliers. The fragments are separated on an agarose gel, denatured by exposure to 0.2 N NaOH and transferred to a charged nylon membrane using standard procedures (Sambrook et al.) . 30

6.3. PREPARATION OF RADIOLABELED PROBE AND HYBRIDIZATION

The COL4A5-specific probe is radiolabeled by the

"nick-translation" procedure. The DNA is briefly treated

„_with DNAse I to introduce a limited number of nicks. DNA 35 polymerase I and a mixture of nucleotides including one radioactive nucleotide is then added and this second enzyme causes these nucleotides to be incorporated in stretches which begin at the positions of the introduced nicks. Following the reaction, the labeled DNA iε separated from °other reaction components by spermine precipitation.

The labeled probe is denatured by boiling and mixed with a standard hybridization solution (Sambrook et al.) and then added to the membrane which carries the array of separated fragments. Following incubation for 16-24 0hours, the nylon membrane is removed from the solution and washed to remove non-specifically bound labeled probe. Following the wash, the membrane is exposed to X-ray film backed by a Dupont-Cronex intensifier screen for 1-3 days at -80 ° C. The film iε developed in an automatic X-ray film processor.

0

5

0

5

Claims

WHAT WE CLAIM IS :

1. A method for identifying an individual hav a genetic defect associated with Alport syndrome comprisi subjecting the DNA of the individual to be tested to

°analysis by a technique capable of detecting variation in DNA sequence, and verifying the presence or absence of variation in the individual's COL4A5 gene, in comparison with a normal COL4A5 gene.

2. The method of Claim 1 in which the techniq ⁰ is RFLP analysis.

3. The method of Claim 1 in which the techniq is SSCP.

4. The method of Claim 1 in which the techniq is allele specific oligonucleotides.

5. The method of Claim 1 in which the techniq is gradient gel electrophoresis.

6. A method for identifying the presence or absence of a genetic defect associated with Alport syndro which comprises: 0 (a) contacti.ng separated restri.ction fragment obtained from a restriction enzyme digest of DNA of an individual to be tested with one or more detectable probe , comprising a portion of the sequence of the COL4A5 gene;

(b) comparing the restriction fragment patter 5εo produced with a restriction fragment pattern obtained from an identical restriction enzyme digest of DNA of nor individuals; and

(c) identifying the presence or absence of variation in fragment length in the individual to be teste

7. The method of Claim 1 in which the probe comprises one or more restriction fragments of at least on clone selected from the group consisting of MD-6, PL-3 and PL-35, or fragments homologous therewith.

8. The method of Claim 7 in which the probe iε obtained from clones deposited at the American Type Culture Collection under Accession No. 40827.

9. The method of Claim 1 in which the

-*restriction enzyme comprises at least one of EcoRI, Taql, or Pstl.

10. The method of Claim 1 in which the probe is radiolabelled. 0

11. The method of Claim 9 in which the variation comprises the absence of at least one EcoRI fragment.

12. The method of Claim 9 in which the variation 5comprises the absence of at least one Pstl fragment.

13. The method of Claim 9 in which the variation comprises the absence of at least one Taql fragment.

0 14. The method of Clai.m 9 i.n whi.ch the vari.ati.on comprises an additional Pstl fragment.

15. The method of Claim 14 in which the variation results from a point mutation of guanine to 5cytosine at position 1352 of the C0L4A5 sequence.

16. A method for identifying the presence or absence of a genetic defect associated with Alport syndrome which comprises: 0 . . .

(a) contacting DNA of an individual to be tested with a detectable probe capable of specifically hybridizing with a segment of DNA

5 comprising a deletion, insertion or mutation in the COL4A5 gene sequence or specifically hybridizing with a normal COL4A5 sequence; and (b) observing the presence or absence of- hybridization between the probe and the DNA bei ⁵ tested.

17. The method of Claim 16 in which the probe comprises a deletion in the exon 3 of the COL4A5 gene.

'0 is. The method of Claim 16 in which the probe comprises a mutation in guanine at position 1352 in the COL4A5 sequence.

19. The method of Claim 16 in which the probe comprises a deletion in all or part of exons 5-10 of the COL4A5 gene.

20. A method for identifying individuals possessing a genetic defect associated with Alport syndrome 0which comprises detecti.ng i.n the indi.vi.dual to be tested th presence or absence of a neutral marker which is tightly linked to a defective COL4A5 gene, whereby the presence of the marker is indicative of the occurence of the defect.

5 21. A method for treatment or prevention of

Alport syndrome in an individual possessing a defect in the

COL4A5 gene which comprises providing to the individual normal copies of the normal COL4A5 gene, or products of the expression of the COL4A5 gene, so as to counteract the 0 effects of the defect or to prevent development of the effects of the defect.

22. A nonhuman animal which possesses a genetic defect in a gene homologous to the human COL4A5 gene. 5

23. A nonhuman transgenic animal which possesses a defective COL4A5 gene and which exhibit an Alport-like phenotype.