WO1995008641A1 - Wilson disease gene - Google Patents

Abstract

Wilson disease (hepatolenticular degeneration) is an autosomal recessive disorder of copper transport, resulting in copper accumulation and toxicity to the liver and brain. The gene (locus WND) has been mapped to chromosome 13 band q14.3. On three overlapping yeast artificial chromosomes (YACs) from this region, a sequence similar to the proposed copper binding motifs of the putative ATPase (MNK) defective in Menkes disease was identified. It was shown that this sequence forms part of a P-type ATPase gene (Wc1) that is very similar to MNK, with at least six putative metal binding domains homologous to those found in prokaryotic heavy metal transporters. This gene lies within a 300 kb region that has been identified as a likely location for WND. The gene is expressed in the liver and kidney. Mutations have been identified within the gene in patients with Wilson disease. This copper-transporting ATPase, designated as ATP7B, is therefore confirmed as the gene defective in Wilson disease.

Description

WILSON DISEASE GENE

BACKGROUND OF THE INVENTION

Copper is an essential trace metal for prokaryotes and eukaryotes, and is a required component for a variety of enzymes, including cytochrome oxidase and other electron transport proteins. Dietary intake of copper generally far exceeds the trace amounts required, and organisms have evolved effective means for the elimination of the excess. Toxicity of copper is believed to act predominantly through the formation of highly reactive hydroxide radicals, which can damage cell membranes, mitochondria, proteins and

DNA (1).

Copper homeostasis requires appropriate mechanisms for copper absorption, cellular transport, incorporation into protein, storage, and excretion. In mammalian systems, various proteins or peptides have been recognized for these functions (2): albumin (and copper histidine) for copper transport in the blood, ceruloplasmin as a possible copper donor to tissues and enzymes (3), and metallothionein for intracellular copper storage (4). The mechanism of copper efflux from tissues has remained an enigma.

Menkes disease and Wilson disease are both caused by a disruption in copper transport (see review (5)). However, these two diseases affect different tissues. In X-linked Menkes disease, copper export is defective in many tissues (5) but is normal in the liver (6). Copper enters into the intestinal cells, but is not transported further, resulting in severe copper deficiency. In contrast, Wilson disease is characterized by failure to incorporate copper into ceruloplasmin in the liver, and failure to excrete copper from the liver into bile. This results in toxic accumulation of copper particularly in the liver, and also in kidney, brain, and cornea. The resulting liver cirrhosis and/or progressive neurological damage has an age of onset from childhood to early adulthood. Consequently, there is a real need to identify the gene responsible for Wilson disease in order to develop new diagnostic and therapeutic strategies useful in the detection and treatment of the disease.

SUMMARY OF THE INVENTION

The Wilson disease gene (WND) has been assigned to a single locus by genetic linkage first to esterase D (7), then to a cluster of polymorphic markers on chromosome 13 band q14.3 (8, 9, 10). Multipoint linkage analysis indicates that WND is flanked proximally by marker D13S31 and distally by D13S59 at distances of 0.4 cM and 1.2 cM respectively (11). Marker D13S31 is sufficiently close to show allelic association with the disease in two different populations (12, 13).

The inventors have isolated new markers between D13S59 and D13S31 and have used them to construct a long range restriction map of the WND region (14, 15). Three CA repeat markers, D13S314, D13S133 and D13S316, have been positioned in a 300 kb region within this map (16). These markers show high allelic association with WND and allowed the identification of specific Wilson disease haplotypes in the region (16). In addition D13S314 was used to define the proximal boundary of the Wilson disease region using a recombination event that is present in one of the Wilson disease families tested.

To isolate the Wilson disease gene, a probe from the proposed copper binding region of Menkes (MNK) was hybridized at low stringency to 19 YACs isolated from a 1-1.5 Mb region immediately distal to marker D13S31. This strategy has lead to the isolation of a gene with all the characteristics required for a copper transporting ATPase, which is predicted to be the gene for Wilson disease.

Accordingly, the present invention provides a DNA sequence containing the gene defective in Wilson disease. In particular, the present invention provides a cDNA sequence containing the gene for a copper transporting ATPase (ATP7B) defective in Wilson disease. More specifically, the present invention provides a nucleotide sequence containing the copper transporting ATPase (ATP7B) which comprises the DNA sequence as illustrated in Figure 10, and the complementary strand or any modified derivative or fragment thereof.

The complete sequence of the metal binding ATPase defective in Wilson disease forms part of the present invention. The DNA sequence includes six copper binding domains, a phosphate domain, a transduction, several potential transmembrane, phosphorylation and ATP binding domains. The sequence of each of these domains forms part of the invention. The 5' and 3' untranslated regions and stated intron sequences are included in the application.

In addition to the cDNA sequence shown in Figure 10, the present application also includes nucleotide sequence adjacent to each exon, shown in Figure 11, which allows the amplification, by PCR, of each exon or a segment of it. Primers have been designed to amplify each exon or segment, however other similar primers can be selected from the sequenced regions, and the patent includes all such

sequences and their complementary strand. This includes the use of the given sequences to obtained further sequence which can then be used for the applications discussed herein.

The present invention also provides a use of the DNA sequence for Wilson disease or any modified derivative or fragment thereof (including the above-mentioned domains) to detect for Wilson disease.

The present invention further includes the use of the DNA sequence for Wilson disease or a modified derivative or any fragment of the given DNA sequence, derived in any way, including amplification by the polymerase chain reaction of RNA or genomic DNA, for the diagnosis of Wilson disease (hepatolenticular degeneration) or the heterozygous state, or for the identification of mutations. This might include such methods as direct sequencing of PCR amplified frag ments, or the examination of differences in small amplified fragments using such methods as single strand confirmation polymorphism (SSCP), denaturing gradient gel electrophoresis, or heteroduplex analysis. Mutations can also be identified by sequencing of the complete gene, in segments, by an automated sequencer. Using the intron primers to amplify each exon, 29 different mutations have been identified. These examples, included in the present application, have been detected by single strand conformation polymorphism analysis (SSCP) and confirmed by sequencing.

Examples of the mutations which can be identified using the sequences described are provided.

The present invention also includes the use of the DNA sequence for Wilson disease or a modified derivative or any fragment of the given DNA sequence, derived in any way, including amplification by the polymerase chain reaction, for use in plasmids or any other vector for therapy of Wilson disease or Menkes disease (another disorder of copper transport). This includes short term use of plasmid containing any part of the sequence in this application as initial therapy for rapid removal of copper, in the early phases of treatment, when patients are at risk from

hemolysis and other complications from rapid release of copper. The use of all or parts of the sequence in

vectors, using lysosomes, is included. The application also includes use for enhancing heavy metal transport in humans or any other organism.

The present invention further includes the use of the DNA sequence for Wilson disease or a modified derivative or any fragment of the given DNA sequence, derived in any way, including amplification by the polymerase chain reaction, to obtain portions of the Wilson disease gene for deriving the homologous ATPase gene from other species. For

example, these sequences have been used to obtain the homologous gene from the rat. The rat sequence, having been derived from the human sequence is included in the present application. This includes the coding region and the 5' and 3' untranslated regions, shown in Figure 17. The cloning of the rat gene has allowed the finding of the basic defect in a mutant rat, the Long Evans Cinnamon (LEC) rat, which is now a model for Wilson disease and can be used to study methods of gene therapy.

Furthermore, the human DNA sequence that is presented here can be used for a variety of other species. In particular the sequence is currently being used to sequence the mouse gene and to determine if a mutation in the toxic mouse lies in this same gene. The mouse can then be used to test gene therapies.

Also included in the present application is the use of the DNA sequence to obtain the sequence ofo the homologous canine gene, which may be defective in Bedlington Terriers and other dog breeds which frequently develop inherited copper toxicosis.

Consequently, the present invention also includes the use of the DNA sequence for Wilson disease or any modified derivative or fragment thereof, derived by any means such as cloning or PCR amplification to obtain the corresponding gene from any other species.

The patent includes the use of the DNA sequence in any way to identify proteins which bind to the DNA sequence and may be involved in conrol of transcription and ultimately in the control of metal related processes.

The present invention further includes the use of the DNA sequence for Wilson disease or any modified derivative or fragment thereof in therapy to remove heavy metal from an organ.

The present invention yet also includes the use of the DNA sequence for Wilson disease, or any modified derivative or fragment thereof in animal breeding, for example to enhance the excretion of copper and other heavy metals.

The invention also includes all of the DNA markers associated with the Wilson disease that the inventors have developed. In particular, the present invention includes DNA markers D13S314, D13S315, D13S316, D13S296 and D13S301 as well as any other markers that detect the same dinucleotide repeat polymorphisms as these five markers.

The invention further includes the use of the above- described DNA makers to detect Wilson disease. The invention further provides a diagnostic kit for detecting Wilson disease comprising at least one marker selected from the group consisting of D13S314, D13S315, D13S316, D13S296 and D13S301.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 Schematic map of the WND candidate region. The positions of the markers relative to the YACs are indicated. D13S314 and D13S315 were derived from cosmids identified by endclones of 27D8 and 235H9 respectively. D13S316 was derived from a cosmid identified by D13S196. The established flanking markers D13S31 and D13S59 are also shown. A physical map of this region has been constructed (Bull and Cox, 1993). YAC isolation and characterization has been described elsewhere (Bull et al. submitted).

Fig. 2 Statistically significant allele distributions. The number of chromosomes carrying a specific allele is shown in white for normal and black for WND chromosomes.

Fig. 3 Isolation of Wc1. a) Restriction maps. At the top is a long range restriction map around the proximal flanking marker for WND, D13S31. Wc1 and markers D13F71S1, D13S196 and D13S31 map to the intervals shown with thick black bars. The location of three microsatellite markers are shown with asterisks. At the bottom are restriction maps of cosmids cosL and cosJ. cDNA clones were mapped to the intervals shown with thick bars. Restriction sites are: M=Mlu I; R=Nru I; N=Not I; H=HindIII; E=EcoRI;

X=XbaI. MluI and NotI sites within YAC 27D8 are shown above and below the line respectively. Three other YACs mentioned in the text (95C3, 53C12 and 68F4) were isolated using PCR primers specific to the right (proximal) end of 27D8. b) Hybridization of probes Me1. a and cosL.d to YACs from the Wilson disease region. YACs 232H4 and 68F4 are overloaded.

Fig. 4 Chromosome mapping of cDNA fragments. a) Each cDNA was hybridized at high stringency to EcoRI and HindIII digests of genomic DNA from Human, hybrid ICD, YAC 95C3 and Hamster. The pattern obtained by hybridization of cDNA clone Wc1.i7 to EcoRI digests of these samples is shown. This clone cross hybridized to a single hamster fragment

(indicated by an asterisk). b) Linkup of clone Wc1.i7 with marker D13S31. Mlul and Nru I digests of the chromosome 13 hybrid ICD were separated with a CHEF PFGE system (Biorad) through 1% agarose with a 300-2000s ramped pulse at 50V for 150h. Sizes (in kb) of detected fragments are indicated on the left. Cross hybridization of Wc(i7) to the hamster component of ICD is indicated with an asterisk.

Fig. 5 Partial DNA sequence of Wc1. Alignment of the amino acid sequences of Wc1 with MNK is indicated, amino acid identities in MNK are indicated with a period. Insertions and deletions in MNK compared to Wc1 are indicated as is the position of a splice site that occurred in one of the cDNA clones. The amino acid numbers of MNK are shown.

Fig. 6 Alignment of cDNA fragments with Me1. The coding region of the Me1 cDNA is represented at the top of the figure by a thick stippled line. The location of functional elements are indicated. The relative positions of probes Me1. a and Mc1.b are shown below with narrow stippled lines. The relative positions of cDNA clones from Wc1 are indicated with thin black lines. The Thick black line shows the part of the Wc1 transcript that has been sequenced. Clone Wc1.f8 contains an unspliced intron.

This is indicated with a dashed line.

Fig. 7 Alignment of nucleotide and derived amino acid sequence of the metal binding domains of Me1 and Wc1. The nucleotide sequence of each domain of Wc1 is shown. A translation of the domains and the corresponding domains of Me1 are shown directly below. Nucleotides and amino acids that are conserved in Me1 are underlined. Residues that are also conserved in bacterial metal binding domains (see Fig. 6) are indicated at the bottom of the figure with an asterisk. Nucleotides that form part of an unspliced intron within the cDNA clone containing the Cu5 domain are shown in italics.

Fig. 8 Alignment of derived amino acid sequence with other genes. Derived amino acid sequences of Wc1 were aligned with Me1 and other proteins from bacteria.

Residues found in both Me1 and Wc1 that are also conserved in one or more of the other proteins are underlined. Those that are invariant in the aligned sequences are shown in bold type. Abbreviations are E. hirae Cu {Enterococcus hirae copper transporting ATPase (26)) E. coli Hg (Escherichia coli mercuric transport protein merP (27)) S. aureus Cd {Staphylococcus aureus cadmium efflux ATPase (28))

R . meliloti Fix (Rhizobium meliloti nitrogen fixation protein FixI (38).

Fig. 9 Northern blot hybridization. Probe Wc1.c8 was hybridized to a Northern blot (Clontech) containing 2 mg of polyA+ RNA from a selection of tissues.

Fig. 10 Complete DNA sequence of the Wilson Disease gene.

Fig. 11 Sequences of exons and flanking intron regions. Exons 9/10/11 and 17/18 have been completely sequenced across intervening introns. Exonic sequence is shown in caps, intronic in lower case. Start ATG and termination TGA are shown in bold. Liver specific sequence is shown in italics. Numbers are for cDNA sequence only and start at the ATG codon. The number refers to the last cDNA nucleotide of that line. Exons 20 and 21 contain alternate 3' exons, expressed in different tissues.

Fig. 12 Exon structure of the ATP7B gene. Exons are shown as numbered, alternating black and white boxes and the location of the various functional regions are

indicated (Cu: copper binding domains, Tm: Transmembrane regions, Td: energy transduction region, Ch: ion channel, Ph: phosphorylation region, ATP: ATP binding). Putative alternate splice sites at exons 5-7, 11 and 20-21 are also indicated.

Fig. 13 Restriction map of genomic DNA in the region of the ATP7B gene. Exon locations are marked with numbered boxes. Exons 9-11 and 17-18 are separated by short introns (see text) and are shown as single blocks. Restriction sites are indicated as follows: B = BamHI, E = EcoRI, H = Hindlll, K = Kpnl, S= Saci. Locations of exons within restriction fragments are arbitrary except exons 1, 2 and 17-18 which contain sites as indicated.

Fig. 14 Diagram of the ATP7B region. The relative locations of the markers in this study are shown.

Positions of the markers have been determined by analysis of recombinant families (12) or overlapping YAC CLONES (46).

Fig. 15 Location of the mutations within the ATP7B gene. Only mutations listed in table 13 are shown. Exons shown as alternating black and white boxes.

Fig. 16 Splice site mutations in ATP7B. Sequences surrounding the alterations are shown and exon/intron boundaries are indicated. The bases affected are incidated with both letters. Translations are shown to indicate the downstream stop codons. Potential splice sites are shown in italics. A: 1615-1G→C, B: 2482+1G→C, C: 3463+1G→A.

Fig. 17 Nucleotide sequence (top) and amino acid sequence (bottom) of the homologous rat gene, including 5' and 3' regions, and coding region.

Fig. 18 Schematic of structure of the rat gene.

Fig. 19 Southern blot showing mapping of the ATP7B gene deletion in the LEC rat.

Fig. 20 RT-PCR analysis of transcripts of the

defective ATP7B gene. DETAILED DESCRIPTION OF THE INVENTION

EXPERIMENT 1 - HAPLOTYPE STUDIES IN WILSON DISEASE

Summary

In 51 families with Wilson disease, we have studied DNA haplotypes of dinucleotide repeat polymorphisms (CA repeats) in the 13ql4.3 region to examine these markers for association with the Wilson disease gene (WND). In addition to a previously described marker (D13S133), we have developed three new highly polymorphic markers (D13S314, D13S315, D13S316) close to the WND locus. We have examined the distribution of marker alleles at the loci studied and have found that D13S314, D13S133, and D13S316 each show non-random distribution on chromosomes carrying the WND mutation. We have studied haplotypes of these three markers and have found that there are highly significant differences between WND and normal haplotypes in Northern European families. These findings have important implications for mutation detection and molecular diagnosis in Wilson disease families.

Materials and Methods

Families Studied

Peripheral blood was collected from 28 Canadian families, consisting of 22 of Northern European, three of Southern European, two of Oriental and one of Indian origin, and 23 families from the United Kingdom, consisting of nine of Northern European, four of Southern European, four of Indian, three of Sardinian and three of Middle Eastern origin. 21 of these families have been described elsewhere (12, 13). The remaining 30 kindreds consisted of 56 parents, 47 patients and 21 unaffected individuals. The ethnic origin of the parents of each patient was determined where possible. Diagnosis of Wilson disease was originally established by clinical symptomatology, slit lamp examination for Kaiser-Fleischer rings of the cornea, and biochemical tests (plasma copper, ceruloplasmin concentrations and urinary copper excretion, and in most cases, measurement of liver copper levels). In some cases, diagnosis was confirmed by radioactive copper studies (39).

DNA Analysis

DNA was extracted from whole blood collected in EDTA by a salt precipitation method (40).

The markers used in this study are listed in Table 1. All three new markers were derived from cosmid clones isolated from a flow-sorted chromosome 13 library. D13S316 was derived from a cosmid identified by D13S196, an anonymous marker derived by Alu-PCR of a hybrid containing the upper half of chromosome 13 (14). D13S314 and D13S315 were derived from cosmids identified by endclones of YACs 27D8 (identified by D13S196) and 235H9 (identified by D13S31) respectively (14). Endclones were obtained by an inverse PCR method (36). The probes were labelled using a T7

Quickprime random labelling kit (Pharmacia), hybridized to filter lifts of the cosmid library as described previously

(12) and exposed to film. Primary positive signals were picked and replated for secondary and if necessary, tertiary screening. DNA was isolated from the clones, digested with three to five of enzymes, electrophoresed in 1% agarose and transferred to nylon membrane (Hybond N+,

Amersham). The blot was probed with poly-dCdA (Pharmacia) to identify bands containing potentially polymorphic repeat units. Positive bands were subcloned and sequenced to determine the length of the repeat. Primers were designed using the OLIGO software package (Research Genetics).

All markers were typed by amplification of the poly-dCdA tract by the polymerase chain reaction. The reactions were carried out in 10 μl volumes containing 50 mM KCl,

10 mM Tris, pH 8.0, 10 mg/ml BSA, 1.5 mM MgCl₂, 200 μM each of dCTP, dGTP and dTTP, 25μM dATP, 0.2 μCi [a ³⁵S]-dATP and 0.5 units of Amplitaq (Perkin Elmer). Amplification was performed in an MJ research PTC-100-96V Programmable Thermal Controller with 30 cycles of 30 seconds denaturation at 94°C, 30 seconds annealing at the appropriate temperature (Table 1), and 30 seconds extension at 72°C. The samples were then electrophoresed through 6% denaturing polyacrylamide gels, which were dried and exposed to film at room temperature for 1 to 5 days.

Allele sizes were determined by amplification of DNA from the original cosmid clone from which the marker was derived. Consistency in allele size determination was checked by including reactions performed on the DNA of 2 to 3 samples of known genotype on each gel. Gels were read independently by two individuals and ambiguous results were repeated.

Results

All four markers were typed in the 51 Wilson disease families. The data are shown in Tables 3 and 4. The relative positions of the markers used in this study are summarized in Figure 1. Also shown is the location of a candidate for WND gene (Wc1) which we have recently identified as discussed in Experiment 2 and in reference 43.

D13S314 was derived from the endclone of the 27D8 YAC and anchors the proximal end of the candidate region through a recombination event we have described (12). All markers centromeric to this, including D13S315, are also recombinant and the event does not include D13S133 or D13S316. The location of D13S133 relative to the other three markers is based on the pattern of amplification of YACs in the region. Rare recombination events were observed for two markers. One obligate crossover occurred in 68 meioses informative for D13S314 and in 100 meioses informative for D13S315, no crossovers were found in 44 meioses for D13S133 and 124 meioses for D13S316.

Alleles associated with the normal and disease chromosomes were determined. Statistically significant deviations between normal and WND chromosomes were seen in the Northern European families for D13S133, D13S314 and

D13S316. These distributions are shown graphically in

Figure 2. Tests of significance were done by the χ² method for large contingency tables with the following results: D13S133 gave a value of 19.94 (11 d.f., p < 0.05), D13S314 yielded a value of 22.58 (9 d.f., p < 0.01), and D13S316 resulted in values of 26.62 (8 d.f., p « 0.001). Haplotypes of the three markers which gave significant evidence of association were constructed for each family. Table 5 summarizes the haplotypes present on Northern European WND chromosomes and their corresponding frequencies on normal chromosomes. It has been shown that allele differences of single repeat units in microsatellite markers arise more frequently than larger deviations (41). Therefore, because of the variability of CA repeat markers due to single allele slippage and the possibility of misassignment of alleles, haplotypes which differ by no more that two bp at a single marker were grouped. The data are significant at p « 0.0001 (X² = 51.44, 8 d.f.). While D13S315 does not show significant amounts of association, several of the haplotypes can be seen to extend to this marker. Haplotype C is exclusively associated with allele 5 at D13S315 and allele 4 is present on all chromosomes carrying haplotypes D and E. A specific allele of this marker does not segregate exclusively with the other disease haplotypes. There is the possibility that the variant haplotypes are in fact different mutations present on a similar haplotype, and the technique of grouping similar haplotypes should be used with caution. However, in the case of haplotypes A and B, only one variant is present on WND chromosomes and the several normal haplotypes have been grouped, therefore any error in classifying haplotypes results in a more conservative estimate of the relative frequency of these WND chromosomes. In the case of haplotypes C, D, and E, the very low frequency of alleles 5 and 6 on normal chromosomes (0 of 44 haplotypes) as well as the exclusive association with particular alleles of D13S315, make it likely that these groupings are justified. The validity of this technique will be determined once mutations have been identified in the WND gene.

Northern European families were further subdivided into more specific ethnic groups in order to determine if any haplotypes were characteristic of a particular area. Of the 10 chromosomes carrying haplotype A, six are of British, one French, one Dutch, one German and one of unknown origin. Haplotype B chromosomes include seven of British and one of Jewish origin. The seven haplotype C disease chromosomes include two German, two Polish, one

Dutch, one French and one of British origin. Chromosomes with haplotype D consist of three British, three French and one German, and the haplotype 5-11-11, found on a chromosome of French origin is also likely related because it shares an extended haplotype, including D13S315, D13S228 and both RFLPs at D13S31, with other group D disease chromosomes. Haplotype E is found on three chromosomes of German and two of British origin.

Chromosomes from other geographical/ethnic groups show no significant differences from those present on normal chromosomes due to small sample size.

Discussion

We have studied four highly polymorphic dinucleotide repeat polymorphisms near the WND locus and have found that three of them exhibit significant levels of allelic association with the disease. The high degree of association seen between the disease and the markers D13S314 and

D13S316 provides strong support for a candidate gene (Wc1) which we have identified on this YAC as discussed in

Experiment 2 and in reference 43. These three markers also form haplotypes around Wc1 which can be shown to be present on disease chromosomes but not on the normal chromosomes in the same population.

One marker (D13S315) failed to detect significant levels of association with Wilson disease despite its location between Wc1 and two markers (D13S31 and D13S228) which have previously been demonstrated to be in disequilibrium with the WND locus (13) and its specific association with three common Northern European haplotypes (C,D and E). This marker is furthest from the disease gene and is likely too mutable to detect association over this large distance. Another explanation is that the association previously seen is due to the chance association of alleles with WND chromosomes. However, other studies have found similar patterns of high and low disequilibrium across a disease region (42).

The existence of haplotypes found commonly on WND chromosomes also provides clues as to the number of

possible mutations present in the Northern European population. While haplotypes A, B, C, and E are similar, the large number of chromosomes of German, Dutch and Polish origin carrying haplotypes C and E suggest that this is a separate mutation from that present on haplotypes A and B, which are found predominantly on chromosomes of British origin. Haplotypes D and G are very different and almost certainly represent separate mutation events. Evidence of common origins or admixture in the European population can be seen in the existence of the common German haplotypes on three chromosomes of British and one of French origin.

Haplotype D is present on three of six French haplotypes while three British and one German family also carry this haplotype. It is also possible that this represents different mutations that have occurred on the same haplotype in these populations. The additional five haplotypes observed on single WND chromosomes may represent separate mutational events or rearrangements of more common haplotypes by an ancestral recombination event, as is likely the case with haplotype 5-11-11 which is present in a French family and is likely related to haplotype D (5-11-5), and haplotype 6-17-12 which is similar to haplotypes C and E (6-17-10/11). We conclude that there are likely to be at least three common Northern European mutations; one found on haplotypes A and B, another on haplotypes C and E, and a third on haplotype D as well as four or more rare mutations on haplotype G and three of the singly represented haplotypes.

The number of patients in other geographical/ethnic groups are fewer in number it is not possible to get statistically significant data regarding association and haplotypes in most cases. The haplotypes present in our two Sardinian families are interesting because they appear to define three distinctly different haplotypes (4-4-7,

4-7-10, and 7-17-11), unexpected in this island population. This suggests the possibility of at least three different mutations in the Sardinian population. The 7-17-11 haplotype appears on three of six disease chromosomes of Italian origin, in a Jewish family, and is the most common haplotype present on WND chromosomes in the British families. This haplotype may indicate the presence of a common wide-spread mutation instead of different mutations on the same haplotype.

The rarity of observed recombination events make these markers ideal for DNA diagnosis of families in which an affected child is available for testing. The presence of two highly polymorphic loci on either side of the candidate gene ensures that any family will most likely be informative for presymptomatic diagnosis of sibs of patients.

Occasionally, diagnosis of Wilson disease is difficult to establish and haplotype analysis at the present time could provide information in some cases. The presence of haplotypes A or B would not provide information while the presence of haplotypes C through G could provide support for presence of Wilson disease, but would not be definitive. With the identification of a candidate gene, specific mutations may be defined to provide a more definitive diagnosis. EXPERIMENT 2 - ISOLATION OF THE HUMAN WILSON DISEASE GENE Summary

New markers for Wilson Disease have been isolated in the region of the gene on chromosome 13q14.3 as described in Experiment 1. The markers were used to construct a long range restriction map and to obtain 19 YACs in the region. Using the copper-binding motif of the ATPase defective in Menkes disease, a homologous region was identified on three overlapping YACs and on cosmids from a chromosome 13 library. Cosmids were used to isolate cDNA clones by a direct PCR-based cDNA selection strategy.

The sequence of the isolated gene shows considerable homology with the Menkes ATPase throughout all its functional domains, including at least 6 copper-binding

domains, transduction, phosphorylation and ATP-binding domains. The gene is expressed in the liver where there is no expression of the MNK gene. This is compatible with the defect in copper transport in the liver observed in

patients with Wilson disease.

Methodology

General methods. Southern blotting, and PCR were performed as described in (14). Pulsed-field gel electrophoresis

(PFGE) is described previously in (15). DNA Sequencing was done using a T7 sequencing kit (USB).

Markers. Marker D13S31 is an established marker for Wilson disease with alleles that exhibit strong allelic association with WND (12). D13S196 and D13F71S1 were isolated by Alu-PCR and mapped to the region of WND as described in

(14). The three markers were used in the construction of a long range restriction map of the WND region (15). Probe

EHR4 was rescued from the distal end of YAC 235H9 (see Table 6).

Cell lines. ICD is a human-hamster somatic cell hybrid containing the proximal half of chromosome 13 as the only human component (14). YACs. YACs were identified from pooled YAC DNA and then isolated from the CEPH YAC library by D. LePaslier using the primers shown in Table 6. All of them are located between the two established markers for Wilson disease D13S31 and D13S59 (11). Genomic DNA was isolated as described in (35) and sizes were determined by pulsed-field gel electrophoresis (PFGE). YAC 27D8 was characterized in detail. To confirm that it was non-chimeric, probes (27L and 27R) from the left and right ends of the YAC were amplified by inverse PCR (36) and hybridized to Mlu I, Nru I and Not I digests of ICD as described in (15), enabling the YAC to be aligned within the long range restriction map. Further verification was achieved through restriction analysis of the YAC by complete and partial digests with Mlu I and Not I. Partial digestion was achieved using 1-5 min incubation in the presence of 1U restriction enzyme. DNA was separated using a CHEF DR II PFGE system (Biorad) (using a 4-40s pulse for 16h at 200V) and hybridized to probes specific to the right or left arms of the pYAC4 vector. YACs 53C12, 95C3 and 68F4, isolated using primers for probe 27R, have not been fully

characterized. YAC 232 H4, used as a negative control, does not hybridize to any of the probes in the Wilson disease region.

Isolation of Menkes (MNK) cDNA probes. To isolate probes (Me1. a and Mcl.b) for the MNK gene, reverse transcription was performed with a M-MTV reverse transcription kit (BRL) using 0.5 mg total RNA from cultured human myotubes as template and primer Mc4062 (5' GC (A/G) TCATTGAT (T/G)

CC(A/G)TC(C/T)CC 3') corresponding to position 4062 of the MNK cDNA (17) in the presence of 40 U RNase inhibitor

(Pharmacia). Probe Me1. a was amplified from the reverse transcribed template by PCR (14) using primers within the putative copper binding region of MNK: Mc967 (5' CAATGATTC AACAGCCACTT3') and Mc1965(5' TTAATATGTGCTTTGTTGGTTG 3'). Thirty five cycles were performed using an annealing temperature of 60°C. An additional probe Mc1.b was

amplified using primers Mc2942 (5' TTTGCAGACAAACTCAGTGG 3') and Mc3835 (5' GTCTGCAATGGCTATCAAGC 3'). PCR products were directly subcloned into a T-tailed vector (Promega). The relative location of the probes within the MNK cDNA is shown in Fig. 5.

Low stringency hybridization. Probe Me1. a was hybridized (14) at 50°C to HindIII digests of genomic DNA isolated from YACs in the Wilson disease region (Table 6). Filters were washed once in 2XSSC and once in 0.2XSSC at room temperature and exposed overnight. Similar hybridization conditions were used to screen 100,0000 cosmids from a chromosome 13 specific library (Los Alamos).

Isolation of cDNA fragments. cDNA fragments for Wc1 were isolated by a direct selection strategy (22, 23) using purified insert from cosmids cosL and cosJ. The DNA was immobilized on filters and incubated with a combination of primary cDNA pools made from adult and fetal liver.

Following two rounds of hybridization, cDNAs were subcloned into Bluescript vector (Stratagene). Two hundred colonies were picked at random and arrayed. The colonies were screened at low stringency with probes Me1. a and Mc1.b. To check their localization, positively hybridizing clones were hybridized, under normal conditions of stringency, firstly to EcoRI digests of cosL and cosJ and secondly to HindiII and EcoRI digests of YAC 95C3, human and hamster genomic DNA and the human-hamster somatic cell hybrid ICD. Having confirmed that the clones mapped to the correct region of chromosome 13, they were sequenced and placed into contigs based on an overlap region of at least 50bp. Sequencing data was used to align the clones with each another and with the known sequence of MNK. Clone Wc1. f3 was found to be the most 3' fragment. To isolate the 3' end of the cDNA, fragment Wcl.f3 was labelled to a specific activity of 1×10⁸ cpm/μg and used as a probe to screen four human cDNA libraries. 2×10⁶ plaques were screened from an adult liver libraries (Stratagene), 1×106 from a second adult liver library (Clontech) and 1×10⁶ from an human hepatoma library and 1.10⁶ from an adult kidney library (Clontech). From a positively hybridizing clone isolated from the Kidney library, a cDNA fragment (Wc1.bl-1) was amplified using an upper primer (350U: GTG GCT AGC ATT CAC CTT TCC) developed from the 3' end of close Wc1.f3 and a lower primer developed from an arm of the cloning vector. An additional cDNA fragment (Wcl.87-90) was isolated by RT-PCR (described above). The first strand was extended on lμg of poly A+ fetal liver RNA (Clontech) from primer

F3L(5'ATGCGTATCCTTCGGACAGT3'). Forty cycles of PCR were performed using primers 1009U (5'GGCACATGCAGTACCACTCT3' ) and 1662L (5 ' TCTGTCTGGGAGATGTGCTT3 ' ) with an annealing temperature of 67°C.

Cosmid mapping. Cosmids cosJ and cosL were digested to completion with Not I and then partially digested with XbaI , EcoRI or HindIII . Southern blots of the digested DNA were hybridized to probes derived from the arms of the cosmid vector. Fragments detected were used to construct restriction maps of the two cosmids.

Alignment of derived amino acid sequences. Alignment of the cDNA clones to MNK and other proteins was carried out at NCBI using the BLAST network service.

Northern blot hybridization. 10 μg of total RNA isolated from each of the following tissues: brain, lung, spleen, heart, stomach, esophagus, muscle, liver and lymphoblasts, was separated using a 1% agarose gel containing formamide (37). The RNA was transferred onto nylon filters

(HybondN+, Amersham). cDNA Wc1.c8 was labelled to a specific activity of 1×10⁸ cpm/μl and hybridized for 20h to the filter in a solution containing 50% formamide, 6×SSC, 0.1% SDS. The filter was washed once in 2×SSC, and once in 0.2×SSC, 0.1% SDS at 65°C. The filter was exposed to X-Ray film (Kodak) for 6 days. In addition, a Northern blot containing polyA+ RNA from heart, brain, placenta, lung, liver, muscle, kidney and pancreas (Clontech) was probed with Wc1.c8 using conditions recommended by the manufacturer.

Isolation of Wc1

Probe Me1. a from the proposed copper binding region of MNK (nucleotides 965-1965 (17) ) was hybridized at low stringency to the 19 YACs listed in Table 6. Above the background hybridization, shown by all YACs (represented in Fig. 3b by YACs 232H4 and 68F4) additional bands were observed in three overlapping YACs : 27D8, 95C3 and 53C12

(Fig. 3b). Two fragments (2.5 kb and 8.9 kb) were detected only in these YACs. The location of YAC 27D8 with respect to the established marker D13S31 is shown in Fig. 3a. To isolate the cross hybridizing sequence, probe Me1. a was used to screen a chromosome 13 specific cosmid library under the same conditions of low stringency. Two overlapping cosmids (cosJ and cosL) were isolated. From cosL, a non-repetitive probe (cosL.d) was isolated that was also found to cross hybridize to Me1. a. To check the localization of cosL.d, it was hybridized under normal conditions of stringency to YACs 27D8, 53C12 and 95C3 (Fig. 3b). The same 2.5 kb and 8.9 kb fragments were detected.

The cosmids cosL and cosJ were used to isolate expressed sequences from liver using a direct PCR based cDNA selection strategy (22, 23). To isolate clones from regions of Wc1 that were similar to MNK, 200 selected cDNA clones were arrayed and screened at low stringency with probe Me1. a and a probe (Mc1.b) more towards the 3' end of the MNK cDNA in the ATP-binding region (nucleotides

2940-3830 (17)). Thirteen individual cDNA clones of

500-1000 bp in length were isolated, of which eight were characterized in detail. To check that they were all located within the correct region of chromosome 13, each cDNA was hybridized to EcoRI and HindIII digests of cosJ and cosL, YAC 95C3, and a chromosome 13 hybrid, ICD. A representative result is shown in Fig. 4a. All fragments detected by the clones used for further analysis mapped only within cosJ or cosL with no other homologous regions on hybrid ICD. In addition, one of the cDNA fragments (Wc1i7) was hybridized to Mlu I and Not I digests of DNA from hybrid ICD that had been separated by pulsed field gel electrophoresis (PFGE). The probe detected a 2200 kb i\7ru I fragment and 2100 kb and 1200 kb Mlu I fragments that are all common to the established marker D13S31 (15) (Fig. 4b).

In an attempt to isolate larger cDNA fragments, a total of 4×10⁶ colonies from three liver cDNA libraries were screened with clone Wc1. f3. No further cDNAs were obtained.

The order of the cDNA clones was established by mapping each clone on cosmids cosL and cosJ digested with several restriction enzymes (Fig. 3a). The gene covers a region of at least 20 kb.

DNA sequence analysis

DNA sequence was obtained for all the clones shown in Fig. 3a. Sequence analysis of the cDNA clones revealed that Wc1 is very similar to MNK. This enabled the isolated clones to be aligned with the MNK cDNA as shown in Figs. 5, 6 and 10. This alignment agreed with the position of the clones on the cosmid map.

Translation of the nucleotide sequence revealed six putative heavy metal binding domains very similar to the six domains found at the 5' end of MNK (17, 18, 19).

Alignment of these domains with the corresponding domains in MNK is shown in Fig. 7. The six Wc1 copper domains in the figure show a mean amino acid identity of 65 percent with the corresponding copper domains one to six of MNK. One of the clones Wc1.fδ seems to be unspliced message since it contains a splice donor site. The site is also present in genomic DNA (not shown). The cDNA selection method we used occasionally selects such products (22).

Both MNK and Wc1 also contain highly conserved domains characteristic of the P-type family of cation transporting ATPases. This family includes magnesium, calcium,

potassium, sodium and proton pumps from various organisms. Members of the family contain a highly conserved region containing the motif Asp-Lys-Thr-Gly-Thr (DKTGT), that includes an aspartate residue which forms a phosphorylated intermediate during the cation transport cycle. Forty three residues N-terminal to this aspartate is a proline residue thought to be involved in transduction of the energy from ATP hydrolysis to cation transport (24).

C-terminal to the transduction and phosphorylation domains is a highly conserved ATP-binding domain including a Gly-Asp-Gly (GDG) motif. Alignments of MNK and Wc1 around these three domains are shown in Fig. 8. The identity between MNK and Wc1 is 86 percent throughout the transduction/phosphorylation domains and 79 percent throughout the ATP-binding domain.

Also shown in Fig. 7 is the alignment and homology of the functional domains of Wc1 with various heavy metal transporting ATPases from bacteria (for a review see (25)). As has previously been demonstrated for MNK (17, 18, 19), the functional domains of Wc1 are more closely related to these prokaryotic genes than to any characterized

eukaryotic gene, except MNK. The most closely related gene is copA from the gram positive bacteria Enterococcus hirae a gene involved in copper transport (26) Alignments are also shown with a mercuric transporting plasmid encoded protein merP from Escherichia coli (27), a cadmium

exporting ATPase from Staphylococcus aureus (28) and a protein involved in nitrogen fixation (FixI) from the symbiotic bacterium Rhizobium meliloti (26). In addition to the N-terminal metal binding domains characteristic of this sub-group of ATPases, three other conserved residues are present that are not a general feature of P-type

ATPases. These are, two cysteine residues, flanking the invariant proline in the transduction domain and a proline situated 8 residues C-terminal to it. These residues may be involved in conferring metal specificity to the proteins (17). DNA sequence is being submitted to genbank.

Expression

To determine the tissue distribution of the Wc1 message, clone Wc1.C8 was hybridized to Northern blots containing RNA from a variety of tissues. Total RNA was analyzed from brain, lung, spleen, heart, esophagus, muscle, liver and lymphoblasts. Transcript was detected only in the liver, and in relatively low abundance, only a small fraction based on the actin control (data not shown). Poly A+ RNA was analyzed from a number of tissues (Fig. 9). Transcript of 7.5 kb was detected at an almost equal abundance in the liver and kidney. A slight trace of message of a similar size was also detected in heart, brain, lung, muscle, placenta and pancreas. The transcript appeared to be slightly smaller than the MNK transcript which is approximately 8.0-8.5 kb (17, 18, 19).

The placenta appeared to have an additional transcript of about 7 kb.

Discussion

There is strong evidence that the Wc1 gene encodes a copper transporting protein. The gene shows high homology with MNK, which is proposed to be involved in transporting copper from intestinal and other cells. Sequence identity is observed in functionally important regions: the energy transduction, phosphorylation and ATP binding domains are 79% identical or greater. In comparing the metal-binding and transduction domains of Wc1, MNK, and the copper-resistant bacteria E. hirae, there are certain conserved residues that may be specific for copper transport

(Fig. 8).

Wc1 is predicted to be the Wilson disease gene because it lies within a region of chromosome 13 that is known to contain WND. A cluster of three highly polymorphic markers D13S133, D13S314 and D13S316, all located within YAC 27D8 and spanning a region of about 300 kb, show strong allelic association with WND and together define a good candidate region for the gene. Wc1 is flanked proximally by D13S314 and distally by D13S133 and D13S316 (16). No other

homologous copper-binding domains, transduction,

phosphorylation, or ATP-binding domains were found within the Wilson disease region.

The expression patterns of Wc1 and MNK are very different. MNK is expressed in lung, skeletal muscle and heart, but is scarcely detectable in the liver or kidney. In contrast, Wc1 is expressed mainly in the liver and kidney. This tissue expression is appropriate for Wilson disease. A key feature in Wilson disease is accumulation of copper in the liver. The expression in kidney is consistent with the occurrence of kidney damage, believed to be due to copper toxicity, in many Wilson disease patients. Abnormalities of renal tubular function include aminoaciduria, proteinuria, uricosuria, hypercalciuria, defective urinary acidification, renal stones, and occasionally full blown Fanconi syndrome (29, 1).

The two main biochemical characteristics of Wilson disease are the disruption of incorporation of copper into ceruloplasmin in the liver and a severe reduction of copper excretion from the liver into the bile (5). Any candidate gene must have potential for involvement in both processes. Ceruloplasmin deficiency, almost always associated with Wilson disease (30) has been recognized as being very closely related to the basic defect. The localization of the ceruloplasmin locus to chromosome 3 (31) showed that a defect in the ceruloplasmin molecule could not be the basic defect in Wilson disease. However, the deficiency is present in patients in early life, before high levels of copper accumulate in the liver. Ceruloplasmin is a 132 kDa glycoprotein containing six atoms of tightly bound copper per molecule, synthesized in hepatocytes (32), and a possible donor of copper to tissues and enzymes (3).

Copper is incorporated during the biosynthesis of ceruloplasmin which is then secreted from the hepatocytes into the plasma (32). The two processes, copper

incorporation and ceruloplasmin secretion, appear to be independent of one another (32, 33). Incorporation of copper into apoceruloplasmin in vitro can only be achieved under reducing conditions (32). It is therefore

interesting to note that Wc1 contains CXXC motifs in each of its metal binding domains, together with one CXC motif in the transduction domain. Similar motifs are

characteristic of many transition metal binding proteins

(34). The motifs are abundant in metallothionein and bind copper in the reduced (CuI) state (4). Incorporation of copper into ceruloplasmin might require close proximity of the two molecules, and some affinity of ceruloplasmin to the membrane ATPase might be predicted. The pathway involved in copper excretion into bile may be similarly sensitive to the redox state of copper. Wc1 therefore has the potential to play a direct role in copper incorporation into ceruloplasmin, and in copper excretion, by maintaining the metal ion in the correct redox state (Cu I).

Although much is known about the role of copper in many essential enzymes, and about its transport in the blood, the mechanism of copper transport between tissues has remained unclear. The isolation of a second human gene for a putative copper transporting ATPase, with contrasting tissue distribution, helps to reveal exciting new directions in the study of copper transport in health and disease. Wc1 and MNK are the only such metal transporters isolated to date from eukaryotes, but the high degree of homology preserved between the toxic metal binding ATPases of organisms as evolutionarily divergent as bacteria and humans indicates the fundamental importance of this type of molecule.

EXPERIMENT 3 - EXON STRUCTURE OF WILSON DISEASE GENE

In order to do large scale screening of WND patients for mutations it is necessary to know the genomic struc ture, and specifically the exon boundaries, of the ATP7B gene. While sequencing of reverse transcribed, amplified message can be done on patients for which RNA is available, direct amplification of DNA from peripheral blood is a more accessible method for most laboratories.

The ATP7B gene encodes a 7.5 kb transcript, of which 4.2 kb codes for the protein (44). The gene is highly similar to the gene responsible for Menkes disease (ATP7A), which has recently been cloned, and spans 150 kb of genomic sequence on the X chromosome (17-19). In order to determine the number, nature and genomic organization of the exons of the ATP7B coding region, we constructed a cosmid contig across the gene. Primers designed from cDNA

sequence were used directly on cosmid DNA to determine genomic sequence flanking exon/intron boundaries. Primers were then designed to amplify individual exons as fragments of suitable size for SSCP or direct sequencing. The cosmid contig was also restriction mapped and probed with each exon to determine the genomic organization of the WND gene. Materials and Methods

Determination of Exon Boundaries

Exon/intron boundaries within the sequence of the ATP7B cDNA were identified by sequencing of genomic DNA in three cosmids which span the region of the gene. The isolation of two of the cosmids (J and L) has been described elsewhere and a third was isolated from the same chromosome 13 specific library using a cDNA clone from the 3' end of the transcript (Wc1.f3) as a probe (44). Primers were designed from cDNA sequence using the OLIGO program

(Research Genetics) and used in a cycle sequencing protocol (Circumvent, New England BioLabs) to derive genomic

sequence directly from cosmid DNA. These sequences were then compared to the cDNA sequence with the MACDNASIS PRO program (Hitachi) to identify exon boundaries and associated splice sites.

Generation of Genomic Restriction Map A total of 10 μg of DNA from each cosmid was digested to completion with NotI to excise the insert and then digested with 5 units of BamHI, EcoRI, HindIII, KpnI or Sacl for 1,2,5,10,20 and 60 minutes in 10 μl volumes. The partially digested cosmid DNA was electrophoresed through 0.6% agarose at 56 Vhr/cm to resolve large fragments (>6kb) and 1.2% agarose at 40 Vhr/cm to resolve small fragments and blotted on nylon membrane (Hybond N+, Amersham). Each cosmid was probed with primers flanking the insert (45) to generate a restriction map. This map was confirmed by total digestion with the same five enzymes and their double digest combinations and by probing with exonic fragments as described below.

Mapping of Exon Fragments

Primers were designed to amplify each exon and tested on cosmid and genomic DNA. Table 7 lists the primers used and the MgCl₂ concentrations that are optimal for amplification of 10 ng of genomic DNA. Also listed are restriction enzymes which cleave each amplimer into two fragments of appropriate size for SSCP analysis. The use of these primers from SSCP is described elsewhere (46). Individual exons were then used to probe cosmid DNA digested to completion with the enzymes described above. Exons were labelled by amplification of 10 ng of cosmid DNA in 20 μl volumes containing 50 mM KCl, 10 mM Tris, pH 8.0, 10 mg/ml BSA, 1.5 mM or 3 mM MgCl₂ (table 1), 200 μM each of dATP, dGTP and dTTP, 25μM dCTP, 2μCi [α³²P]-dCTP and 0.5 units of Amplitaq (Perkin Elmer). Amplification was performed in an MJ research PTC-100-96V Programmable Thermal Controller with 35 cycles of 30 seconds denaturation at 94°C, 30 seconds annealing at 55°C, and 30 seconds extension at 72°C. Unincorporated label was removed by sephadex G-50 spin column and hybridization was carried out by standard methods.

Exon 21 was mapped by probing cosmid DNA with a primer derived from the cDNA sequence in this region. The primer is located immediately 5' of the stop codon in this exon and its sequence is GGACAGCGGCAGAGCCAGGAAAC.

Results

Determination of Exon Boundaries

Sequencing of the WND cosmids with cDNA primers identified a total of 20 exons in the liver transcript. The sequence of each exon is shown in Figure 11. The first and last nucleotide of each exon, its length, splice sites and domains are listed in table 8. The locations of the functional regions of the gene within the exons are shown in figure 12. The results obtained from sequencing primers in the 5' end of the cDNA sequence include sequence that is further upstream than that published. There is a consensus splice acceptor site located 46 bp upstream of the ATG start codon and this has been defined as the start of exon 1. The 5' untranslated portion of the cDNA is expected to be contained within one or more exons located upstream of this point.

The 3' end of the gene is different in kidney and liver cDNA clones (44, 47). The final coding exon of the liver-derived sequence is contiguous with the genomic sequence and defines a single exon, designated as exon 20 in table 8. The kidney-derived sequence in this region is identical to the liver cDNA for the first 73 base pairs of exon 20 and then diverges. There is a consensus splice donor site at this point, indicating that the difference between the two tissue transcripts is most likely due to the use of alternative splice sites. Therefore, at least some portion of the transcripts in the kidney have an additional exon added at this point. We have designated the kidney specific sequence as exon number 21. Flanking sequence for this exon has not been determined.

Determination of Genomic Structure

The portion of the restriction map of the three WND cosmids containing the coding region of ATP7B and the locations of each of the exons are shown in figure 13. The coding portion of the gene spans approximately 40 kb of genomic DNA. The three overlapping cosmids from which the map was derived span a total region of about 80 kb, with 20 kb on each side of the map shown. Introns 9, 10, and 17 have been completely sequenced and are each less than 200 bp in length (data not shown) thus exons 9-11, and 17-18 are shown as single blocks.

Discussion

The identification of exons in the ATP7B gene is a necessary first step in order to make large scale mutation screening of Wilson disease possible. We have determined the exon/intron structure of the gene in order to design primers flanking each exon for use in SSCP and direct sequencing of mutations, described earlier (46). The exons have also been placed on a genomic map of the region derived from cosmid DNA.

The coding region of the transcript is contained within 21 exons. Most of the exons are less than 300 base pairs in length, with the notable exception of exons 1 and 20. Exon 1 is 1.2 kb in length and includes copper binding domains (Cu) 1-4, while exon 20 includes the 271 bp of coding sequence before the termination codon and an undetermined number of bases beyond this point, and is greater than 400 bp in length. The size of exon 21 is unknown. This makes amplification of individual exons and subsequent SSCP relatively simple and sequencing of the PCR products can be accomplished with one reaction. We have divided exon 1 into six overlapping amplimers in order to cover the entire region and only the coding portion of exon 20 is amplified. All of the sites flanking the exons conform to the consensus for 5' and 3' splice sites (48). With the recent cloning of the homologous gene in the rat discussed in experiment 5 (49), it will be of interest to determine if this exon/intron structure is conserved between ATP7A and ATP7B genes of different species. There is evidence of alternative splicing at several points in the transcript, seen as differences in published liver, brain and kidney cDNA sequences (44, 47). Two segments of the cDNA (transmembrane (Tm) regions 1-4 and transmembrane region 5) are missing in the published brain cDNA sequence (47) but are present in the liver sequence (44). These regions correspond exactly to exons 5,6 and 7 (Tm 1-4) and exon 11 (Tm5) indicating that the difference is due to alternative splicing. Removal of these four exons does not alter the reading frame of the protein product. They also appear to alternate 3' ends in transcripts from liver and kidney, possibly indicating slightly different functional roles in the two tissues.

The restriction mapping of the WND region shows that the ATP7B coding region spans 40 kb of genomic DNA. An additional 20 kb of DNA on either end of cosmid contig presumably contains the 5' and 3' noncoding regions and possibly the promoter and other regulatory portions of the gene. This is in contrast to the reported 150 kb for

Menkes disease (17-19), although the promoter and

polyadenylation sites of the ATP7B gene have not been identified and the WND gene could be of comparable length.

We have identified the exon boundaries and derived primers which can be used for SSCP analysis as well as direct sequencing of genomic DNA of patients, foregoing the need for RNA isolation from patient materials . These primers have been used to screen 50 WND families for mutations and the results of that study are reported elsewhere in Experiment 1 and in Reference 44.

EXPERIMENT 4 - MUTATIONS AND HAPLOTYPES IN WILSON DISEASE GENE

In order to characterize the range of mutations present in the ATP7B, we used primers located in the introns to amplify each exon from genomic DNA. Single strand conformation polymorphism (SSCP) analysis and direct sequencing of exons in patient DNA were used to identify the mutations present in 50 families with Wilson disease. We have also extended our previous haplotype analysis (46) in a total of 65 families by adding two CA repeat markers and have examined the relationship between these haplotypes and the mutations present in the WND gene.

Materials and Methods

Patient Materials

Peripheral blood was collected from 34 Canadian families (consisting of 25 of Northern European, 4 of

Indian, 3 of Southern European, and 2 of Oriental origin), 30 families from the United Kingdom (consisting of 15 of Northern European, 5 of Southern European, 5 of Indian, 3 of Sardinian, and 2 of Middle Eastern origin), and 1 family from Saudi Arabia. 51 of these 68 families have been described in experiment 1 (12, 46, 50). The remaining 14 kindreds consisted of 23 parents, 20 patients and 10 unaffected sibs. The ethnic origin of the parents of each patient was determined where possible. Diagnosis of Wilson disease was originally established as previously described (46). DNA was extracted from whole blood collected in EDTA by a salt precipitation method (40).

Haplotype Analysis

Haplotypes of the markers D13S133, D13S314, D13S315 and D13S316, surrounding the Wilson disease gene, have been described in experiment 1 (46). Two additional markers, D13S296 and D13S301 (50), were added to the haplotypes. The amplification of CA repeats in patient and parent DNA were as described (50, 46). Primers, conditions and size standards used for the new markers are given in table 9 and allele sizes are given in table 10.

Detection of Mutations

Selected patient samples were screened for mutations by the use of single strand conformational polymorphism (SSCP) analysis on individual exons. Patient samples were selected on the basis of the haplotypes derived from markers in the WND region (46), such that all haplotypes were represented at least once. The primers used for each exon, their product length, and MgCl2 concentrations for optimal amplification are given in experiment 3 (51).

Exons were amplified under conditions identical to those used for the CA repeats (46), with an annealing temperature of 55°C, and digested for 2 hours with the appropriate restriction enzyme (51). The samples were then diluted with one volume of SSCP buffer (0.2 M NaOH, 1% SDS) and three volumes of loading buffer (95% formamide, 15 mM EDTA, 0.03% each of xylene cyanol and bromphenol blue) and electrophoresed through 6% non-denaturing polyacrylamide gels at either room temperature with 10% glycerol for 18-24 hours at 10W or at 4°C with no glycerol for 5 hours at 35 W before drying and exposure to film.

Patient samples exhibiting shifts relative to normal samples on SSCP were subjected to direct sequencing to determine the nature of the mutation. Patient and parents (where available) were amplified as above for 35 cycles with 200μM cold dATP and no [α³⁵S]-dATP. Products were purified with a QiaQuick spin column (Qiagen), cycle sequenced (Circumvent, New England Biolabs ) using the PCR primers , and electrophoresed through 6% denaturing polyacrylamide gels.

Results

Haplotype Analysis

Two additional markers (D13S296 and D13S301) have been added to the haplotypes of our WND families. These two markers as well as three that have been previously

described in experiment 1 (46) have been used to examine the haplotypes present on both normal and WND chromosomes. The locations of these markers relative to the disease gene are shown in figure 14. The locations of D13S133, D13S314 and D13S316 have been described in experiment 1. The position of D13S296 telomeric to ATP7B is based on its location near D13S133 (50). D13S301 has been placed centromeric to WND by a recombinant family that we have described previously (12).

The previous study had defined 7 groups haplotypes present more than once on WND chromosomes in our families of Northern European origin. Several of these haplotypes were not found on the normal chromosomes. Table 11 summarizes these haplotypes and includes data from two new markers. The addition of new markers results in a number of different haplotypes within several of the previously defined groups, in particular, groups A and B. One new WND haplotype was found, present on 3 affected chromosomes and not on the normal chromosomes in the same population.

It has also been possible to identify WND haplotypes in populations other than Northern Europeans. These haplotypes are described in table 12. Most of the haplotypes that appear twice in a population are present in a single homozygous patient and would therefore be expected to carry the same mutation due to consanguinity.

Mutation Analysis

A total of 39 sequence changes in the ATP7B gene were detected in our families. Three of these are mutations that have been described previously (47). Eleven of the changes are polymorphic, present in greater than ten percent of chromosomes studied, including some normal

individuals. Four of the changes are conservative amino acid changes present in one or two patients, which are unlikely to represent functionally significant mutations. The remaining 21 mutations include frameshifts, splice site alterations, nonsense mutations, non-conservative amino acid substitutions or changes in invariant residues within functionally important regions of the gene. Table 13 summarizes the mutations and the haplotypes on which they are found and figure 15 shows their locations within the gene. One previously described-mutation, 3457insT (47), was not present in our families. There are a total of 12 small insertion/deletion mutations that result in frameshifts. 750delC was found on two chromosomes of Iranian and Greek origin and results in a stop codon immediately. A single chromosome in a British patient carries 1651delAT, which results in termination of translation 24 amino acids downstream of the deletion.

1652insT was identified on a single chromosome in an Sikh patient and also results in termination at the same point. 2065delA was found on a single affected chromosome in a family of Italian origin, and also results in an immediate stop codon. 2206insC was found on four chromosomes; 2 in British families, and one each in Scottish and Italian kindreds. The frameshift results in a termination codon 27 amino acids downstream in the next exon. The 2881de1C mutation was found to be homozygous in a Sardinian and British patient and heterozygous in another British

patient, resulting in a stop codon 28 amino acids downstream of the deletion. One chromosome in a English

Canadian family was found to carry the 2992delAC mutation which causes a termination 38 amino acids downstream of the deletion. A Sikh patient was found to be homozygous for 3053delC, which causes termination of translation after 67 altered amino acids. 3307delC was found on a single chromosome in a Ukrainian patient and results in a termination codon in the next exon, after 20 amino acids. One chromosome of Swedish origin carries 3533del4, which introduces a stop codon into the sequence after 8 amino acids. 3555del6 does not introduce a frameshift into the protein product but deletes Val₁₁₈₆ and Ile₁₁₈₇ in a single British

patient. One chromosome in a British patient was found to carry 3998delTG which terminates the translation product after 12 amino acids.

Two nonsense mutations were detected in our screen. L905X is caused by a T to A change at the second base of the codon and was found to be present on both chromosomes in a Saudi Arabian family and one chromosome in a Greek kindred. R1288X was found on a single chromosome in a British family and is due to a C to T change at the first base of codon 1288.

Six alterations of splice sites were detected, three of which are polymorphisms. The sequences involved in the three mutations are shown in figure 16. 1615-1G→C alters the invariant G residue at the last position of intron 3 (figure 16A). This would result in the removal of exon 4 from the transcript and loss of the 30 amino acids of the sixth copper binding domain and 24 amino acids downstream. A potential splice acceptor site exists within exon 4 and if used would result in termination of the transcript as shown. A single Pakistani patient was found to be homozygous for this mutation.

The mutation 2482+1G→C alters the invariant G residue at the start of intron 9 which would result in the addition of the intronic DNA to the transcript (figure 16B) . This results in 36 amino acids after the end of exon 9 and then a premature termination codon. There is a near consensus splice donor site 52 bases into the intron which might be used in this transcript but the resulting product would have a frameshift. This mutation was found to be heterozygous in a British patient.

The third splice mutation, 3463+lG→A, was found on a single chromosome in a British family and also alters the invariant first base of the donor site, in intron 15

(figure 16C). This results in 27 amino acids coded by the intron followed by termination of the protein product.

There are no potential splice donor sites located upstream of the stop codon.

The other three base alterations that affect splice sites, 2773-1 3G→C, 3464-6C→T and 3810+6T→C, are not expected to alter splicing of the transcript. The first change alters the splice site toward the consensus and the second does not change consensus. The last base change alters the splice site away from the consensus but this change is found on approximately 40% of affected chromosomes and has been detected on two normal chromosomes by direct sequencing.

Two previously described and six new amino acid changes were identified in our families. One chromosome of Scottish origin was found to carry a T to C change at the first base of codon 501, resulting in a tyrosine to histidine change at this location. This mutation is present on a haplotype that is unique to this patient and not found in the normal population and by Chou-Fasman secondary structure prediction (using the MACDNASIS PRO program) is expected to alter the folding of the region between Cu5 and Cu6. Three chromosomes in two Chinese families carry a C to T change at the first base of codon 747, replacing an arginine with a leucine residue. This substitution occurs in the fourth transmembrane region and replaces a positively charged amino acid with a long chain hydrophobic residue. A single Saudi Arabian patient was homozygous for a G to A transition in codon 912, altering a glycine to a serine in transmembrane region 5. A previously described (47) change, H1038Q, was found to be the most common mutation in our families, occurring in 17 families of German, Dutch, Polish, French, British, Swedish, Ukrainian, and Greek origins. G1070R and I1071T are adjacent amino acid changes that are homozygous in single Indian patients and are present on chromosomes unique to Wilson disease

patients in this population. Six patients of French, British and German origins were found to have an alteration of the invariant Gly₁₂₄₅ in the ATP hinge domain, replacing this residue with a lysine. This glycine residue is highly conserved in all P-type ATPases (44) and is likely vital to their activity. Another previously described mutation, N1239S (47), was found on one chromosome in an Italian patient.

A number of changes within the gene were identified that did not affect the amino acid sequence of the protein product, resulted in conservative changes in apparently non-vital residues within the protein, or were found to be present on normal chromosomes in the population. Table 14 summarizes these changes and listed the restriction sites, if any, altered by each. All of the changes listed are polymorphic on at least 10% of chromosomes with the

exception of the last four. M738V and V964A were found on single chromosomes of British and Sikh origin, respectively, L735V was found on single chromosomes in two British patients and M1138V was found in a British and an Italian patient. Allele frequencies of these polymorphisms within the normal populations were not determined.

Discussion

The identification of mutations in the ATP7B gene is a necessary first step in order to make direct diagnosis of Wilson disease possible. We screened 50 patients with Wilson disease and identified 39 changes in the WND gene, 24 of which appear to be functionally important mutations. The mutational data obtained has been compared to extended haplotypes of Wilson disease patients to assess the usefulness of CA repeat haplotypes in prediction of the disease state.

Extension of the haplotype analysis has resulted in a more complicated picture of the number of mutations present in the populations studied. The two haplotypes previously described (46) as being the most common in the Northern European populations (A and B) have been shown to be a collection of different haplotypes when more markers are added. This is likely due to the fact that the alleles of D13S314, D13S133, and D13S316 present on these original haplotypes are the alleles most commonly found on normal chromosomes in this population, and represent a common background on which a number of mutations may have

occurred. Conversely, haplotypes C, D, and E are seen to remain tightly grouped with only slight variations (no more than 2 bp at the D13S301 locus, 1 bp at D13S316 or one haplotype with an 8 bp change at the D13S296 locus).

Groups C and E can be seen to share a common origin in that the majority of both haplotypes carry the H1038Q mutation. Group D is also associated exclusively with a single mutation, G1235K. Thus, while a few CA repeat markers close to a disease loci can be used in linkage disequilibrium studies and to identify a number of haplotypes which might share a common origin, a greater number of markers will further clarify the data. The ultimate test for common origins of haplotypes is the sharing of a common mutation. The range of alleles at the D 135301 locus on chromosomes which are identical for the other markers suggests that this marker may be more prone to replication slippage and therefore not as useful in haplotype studies.

The large number of subtypes within groups A and B indicate that there are a large number of mutations present in the Wilson disease population. We have described a total of seven mutations on the chromosomes carrying these haplotypes, while another six haplotypes remain. The mutation present on most group C and E haplotypes has been identified as H1038Q. A few chromosomes carry the group C or E haplotype but do not have this mutation, and the changes present on these chromosomes have yet to be identified. These two groups of haplotypes differ by no more that 4 bp at a given locus, most of the variation occurring at the D13S301 marker. The correlation of haplotype group to mutation, in this case, supports our previous system of grouping haplotypes that differ by no more than 2 bp at one locus (46).

The mutation on group D chromosomes has also been identified (G1235K). Group C, D and E chromosomes represent approximately 37% of the Wilson disease chromosomes present in the Northern European population. Careful examination of the haplotypes reveals that the D13S316 marker is diagnostic for these two mutations, in that most chromosomes carrying allele 6 at this marker have the H1038Q mutation and all chromosomes carrying allele 5 have the G1235K mutation. These alleles are not present in the normal chromosomes in our families. This haplotype/ mutation association can be used to rapidly identify chromosomes likely to carry one of these mutations and this could be then confirmed by a single sequencing reaction.

We have revised our estimate of the number of mutations responsible for Wilson disease to include the new marker and haplotype data. We have combined groups C and E due to their sharing of the H1038 mutation, and have split groups A and B into numerous, less common haplotypes as listed in table 11. The two WND haplotypes in group F are now considered separate. This rearrangement results in defining 34 different haplotypes within the Northern

European population, 17 of which have identified mutations (table 13). Haplotypes in the other ethnic groups can be defined to some extent, as shown in table 12. Within the Indian/Pakistani group, at least 10 different haplotypes can be identified, likely each with its own mutations.

Similarly, there are 9 distinct mutations within the

Southern European population. One Southern European haplotype (SE2) is identical to the group C/E haplotypes in the Northern Europeans and indeed carries the same mutation, indicating a common origin for these chromosomes. This may prove true for other haplotypes from all of the ethnic groups. Indeed, the existence of the same mutations in different ethnic groups (ie. 750delC) lends support to this prediction.

A number of mutations appear to be present on very different haplotypes (ie. haplotypes which differ by more than 4 bp at several markers). The 2206insC mutation is present on three very different haplotypes which cannot be explained by recombination. However, this mutation is a insertion of a C into a series of six within the cDNA sequence and thus may represent a spot where DNA polymerase is more likely to make an error during replication. Therefore, these three haplotypes may represent independent origins of the same mutation.

Other cases of mutations present on different haplotypes are more difficult to explain. R747L is present on three different haplotypes and is less likely than 2206insC to represent independent occurrences of the same mutation. This may represent a case of gene conversion that has resulted in the mutation being transferred to another haplotype.

There are also cases of different mutations present on chromosomes with identical haplotypes. Several group C/E haplotypes do not carry the H1038Q mutation, presumably having another lesion. This haplotype is not found within the general population is not likely to be a common background on which two mutations have occurred

Table 14 lists a number of polymorphisms that have been identified within the ATP7B gene, only four of which are rare. These polymorphic bases, some of which alter restriction enzyme sites, could be readily typed in Wilson disease patients in order to extend the haplotype data. The marker loci would not be subject to the polymerase slippage that may cause large variation between related haplotypes consisting solely of CA repeat markers.

Additionally, the locations of these polymorphisms within the disease gene itself would be expected to result in higher disequilibrium values and better haplotype/mutation correlation, as was shown for the cystic fibrosis gene (53).

An initial screen of our families has identified 24 mutations, three of which have been previously published. These mutations represent 58% of the disease chromosomes present in these families and 50% of the haplotypes.

Therefore, a large number of mutations remain to be

identified. However, the occurrence of a mutation on several different haplotypes and multiple mutations on the same haplotype make it likely that some of the remaining chromosomes actually carry one of the mutations we have identified.

EXPERIMENT 5 - ISOLATION OF HOMOLOGOUS RAT GENE

The Long-Evans Cinnamon (LEC) rat (54) shares many clinical and biochemical features with Wilson disease.

This inbred rat strain was originally established in 1975 from a closed colony of non-inbred Long-Evans (LE) parental rats through successive generations of sibmating. Spontaneous acute hepatitis with severe jaundice occurred in a male rat from the F₂₄ generation at five months of age (54, 55). The mutant allele causing the hepatitis was fixed in subsequent generations through inbreeding, and backcross experiments demonstrated an autosomal recessive pattern of inheritance for this condition (55, 56). LEC rats spontaneously develop acute hepatitis about four months after birth, with clinical features similar to those seen in human fulminant hepatitis, sometimes a feature of Wilson disease. Survivors of this often-fatal attack continue to suffer from chronic hepatitis and usually develop hepatocellular carcinoma at age 12 months or older (55, 57, 58). Copper is abnormally high in the liver of the LEC rats and backcross experiments demonstrated a cosegregation of the hepatitis with the liver copper accumulation and the decreased plasma ceruloplasmin in progeny rats (59, 60). Hepatic copper accumulation occurs prior to the development of the hepatitis and the hepatitis can be prevented by treatment with copper chelating agents such as D-penicillamine, a drug frequently used in treating Wilson disease patients (61). These observations suggest that hepatitis and hepatoma are caused by copper accumulation and that the LEC rat is a model for Wilson disease (62). In contrast to patients with Wilson disease, LEC rats rarely have

prominent neurologic abnormalities (see Table 15).

To test whether the LEC rat is a model for Wilson disease, we initiated studies to determine whether the LEC mutation resides in ATP7B, the rat homologue of the human Wilson disease gene. We report here the cloning of overlapping cDNAs for the rat ATP7B gene and the identification of a partial deletion in the gene, which removes at least 900 basepairs (bp) of the coding region at the 3' end and at least 400 bp of the downstream untranslated region

(UTR).

Methodology

General procedures. The cDNA library was made from the liver of an adult Sprague-Dawley male rat using oligo-dT/random hexamers as primers (Clonetech #RL1023a). Library screening was done according to the standard procedure, using human ATP7B cDNA segments Wc1.g1, Wc1.c8, Wc1.87-90 and Wc1.gb10 as probes (2). Inserts from positive

recombinant bacteriophage λ cDNA clones were generally subcloned into the pBluescript SK(+) phagemid vector

(Stratagene) for restriction mapping and sequence analysis. DNA sequencing was done on double-stranded plasmid DNA using universal and reverse primers and sequence-derived oligomers with the Sequenase kit (United States

Biochemical).

Southern hybridization analysis. LEC rats, two males and two females, were purchased from Charles River Japan, Inc. and LE rats, one male and one female, from Charles River Canada, Inc. Rats were killed at 9 weeks of age. Genomic DNA was extracted from peripheral blood lymphocytes from a control male LE rat and from liver tissues of a male LEC rat. Standard methods were used for restriction enzyme digestion, agarose gel electrophoresis and blotting onto the Hybond-N+ nylon membranes (Amersham). DNA probes were labelled with α-³²P-dCTP using random priming method (25) and hybridization was carried out in 5x SSC, 0.1% SDS, 5x Denhardt, 100 μg/ml sheared and denatured salmon sperm DNA and 10% dextran sulphate at 65°C overnight and final washing conditions were at the same temperature and in 0.1x SSC and 0.1% SDS. Autoradiography was done with Kodak films at -70°C for 1-3 days. RT-PCR analysis. Poly(A+) RNA was extracted from liver tissues of a female LEC and a control female LE rat, using the Fasttrack mRNA isolation kit (Invitrogen). About 2 μg of poly(A+) RNA was reverse- transcribed in a volume of 33 μl, using random hexamers and murine reverse transcriptase and other reagents in the First-strand cDNA synthesis kit (Pharmacia) according to supplier's protocol. PCR was carried out in 20 μl containing 0.5 μl of the reverse-transcribed cDNA templates, 20 pmole of each primer, 100 μM of each dNTPs, 50 mM KCl, 10 mM Tris/pH 8.3,1.5 mM MgCl₂ and 1.0 U Taq polymerase. The step-cycle mode amplification started with one cycle at 95°C for 3 min, 60°C for 2 min and 72°C for 3 min, followed by 30 cycles each at 94°C for 1 min, 60°C for 0.5 min and 72°C for 1 min and terminated with a final extension at 72°C for 15 min. About 10 μl of the reaction products were analysed by electrophoresis on a 1.5% agarose gel. Sequences of the primers (from 5' to 3' ends) used are: 1) 0114F: GACATGGGATTCGAAGCTGC; 2) 0108R: CACTTCTGTGATGCTGTTCC; 3) DF1 : AATGCTCATGGCTCTGTGCTC; 4) DR1: CCACAGCCAGAACCTTCCTG; 5) DR2 : CCAGCATACTTTCCACGTTGC;

6) DR3: CGCACAGCACACCATCAATGG; 7) DR4 : CACTGATGTCACTGG AGATGG; 8) DR5 : CTTCTCCTATGGGAGGGTTGC.

Isolation of cDNAs for ATP7B

Twenty five cDNA clones for ATP7B were isolated from a rat liver cDNA library, using as probes cDNA sequences for the human ATP7B gene (2). A consensus sequence of about 4.7 kilobases (kb) was derived from these overlapping clones and its nucleotide sequence determined. The sequence has a single large open reading frame (ORF) and includes 300 bp of 3' untranslated region. The first in-frame methionine codon starts 17 nucleotides (nt) downstream of the 5' end of the sequence and there is a second in-frame methionine codon located 32 amino acids downstream. The ATP7B cDNA sequence is highly homologous to its human counterpart and, as in humans, is predicted to encode a copper transporting P-type ATPase. There is an amino acid sequence identity of 82% between the rat and human genes. Overall structure as well as each of the individual functional domains are well conserved between the rat and human, with the exception of the lack of metal binding motif 4 in the rat (Figs. 17 & 18). The sequence divergence between the two species in this region was confirmed by sequencing (two clones) and restriction mapping (five clones) independent cDNA clones. The rat sequence of the last 66 amino acids at the

C-terminus differs from that reported by Bull et al. (44) but is similar to that described by Tanzi et al. (47).

This discrepancy is apparently due to alternative splicing of the ATP7B transcripts (Thomas, G.R., unpublished

results).

Characterization of the LEC deletion

The ATP7b cDNA sequences were used to probe Southern blots of genomic DNA digests from an LEC rat and a control LE rat. Using a mixture of probes (p7-5, p7-6 and p7-1) representing the entire 4.7 kb rat cDNA sequence, restriction enzyme digests of the genomic DNA from the two rats gave rise to different hybridization patterns (data not shown). For each of the five restriction enzymes tested, BamHI, HindIII, HincII, PvuII and TaqI, the LEC DNA was found to have, in addition to some hybridization fragments shared with the control LE sample, at least one fragment difference, either missing altogether or altered in size, when compared with the LE pattern. This discrepancy was clearly not due to a polymorphism but could only be

explained by a deletion of part of the gene in the LEC rat, since multiple enzymes were used.

Three probes, p7-5, p 1-8 and p7-1, representing different parts of ATP7B, were used separately to localize the deleted region. Probes p7-5 and pi-8 gave rise to identical hybridization patterns between the LEC and the control rats, whereas probe p7-1 revealed the same hybridization abnormalities noted before in the LEC DNA sample

(data not shown). These results indicated that the deleted region includes the 3' end of the coding region of the gene (Fig. 18).

To map the deletion more precisely, four RsaI fragments from the p7-1 insert were used separately to probe blots containing aliquots of the same genomic DNA digests from the LEC and the control rats. Referring to Fig. 19, the four blots contain aliquots of the same genomic digests of the LEC and the control LE rat DNA. Lanes 1 and 3 in each blot contain digested DNA from the control rat; lanes 2 and 4 contain DNA samples from the LEC rat. The two blots on the left were made from different portions of one gel and the two blots on the right were made from another gel. Sizes of the hybridized fragments are indicated.

Enzymes used for the digestion (Hd, Hindlll; Hc, Hincl) and probes used with each blot are indicated on the top. The relative positions of these probes in the Atp7b cDNA sequence are indicated in Fig. 18. While data are shown only for the two enzymes, results using BamHl, Pstl and Taql gave the same conclusion.

7RsF1 and 7RsF2 detected bands in both the LEC and the control rat DNA samples: 7RsF1 gave rise to identical hybridization patterns in the LEC and the control (an 8.5 kb fragment); 7RsF2 revealed an altered HindIII (5.8 1:b) fragment instead of the normal 11 kb fragment from the LEC sample (Fig. 19). Size alteration was also seen in the BamHI digest tested with 7RsF2 (data not shown). In contrast, 7RsF3 and 7RsF4 hybridized only to the control samples but not to the LEC DNA. Southern hybridization was also carried out with probe 21-3e1, a 297bp fragment from the 3' end of clone 21, a 2.4 kb cDNA clone which extends a further 108 bp 3' of the ATP7B cDNA consensus sequence. Again, hybridization signal was seen only in LE but not in LEC DNA (data not shown). These results indicate that the proximal deletion breakpoint is located between the

sequences represented by the probes 7RsF2 and 7RsF3 and the deletion extends to the most 3' end of the cloned region. Refining the deletion breakpoint

Reverse transcription-polymerase chain reaction (RT-PCR) analysis of poly A⁺ RNA from the rat livers was subsequently done, initially to assess transcriptional status of the defective ATP7B gene in the LEC rat. Figure 20 is an ethidium bromide-stained agarose gel showing RT-PCR products from the LEC and the control LE rat cDNA.

Lanes 1, 5, 7, 9 and 11 contain amplification products from the control LE sample; lanes 2, 6, 8, 10 and 12 contain those from the LEC rat. Primer pairs used are:

0114F/0108R (lanes 1, 2), DF1/DR1 (lanes 5, 6), DF1/DR2 (lanes 7,8), DF1/DR4 (lanes 9, 10), DF1/DR3 (lanes 11, 12). Lanes 3 and 4 show Rsal digests of the 844 bp fragments amplified with the primer pair 0114F/0108R from the LE and LEC cDNA samples, respectively. There were some weak nonspecific bands in lanes 5, 6, 10 and 12. As a control, amplifications with LE rat genomic DNA as template were done for all pairs of primers and no amplification products were detectable on the ethidium bromide-stained get for any primer pairs tested. DNA size markers are indicated on the left.

As shown in Fig. 20, ATP7B cDNA-specific primer pair 0114F and 0108R amplified an 844 bp fragment from both the LEC and the control LE cDNA samples, a size expected from the normal cDNA sequence we determined. The authenticity of these transcripts was further confirmed by digestion with the enzyme Rsal: four fragments of the expected sizes (374 bp, 290 bp, 112 bp and 68 bp) were produced from both the LEC and LE 844 bp fragments. RT-PCR analysis was then applied to refine the proximal deletion breakpoint in the coding region. A sense strand primer, DF1, was chosen in an area that is known to be present in the LEC ATP7B gene (based on the Southern hybridization results) and a series of three complementary strand primers, DR1, DR2 and DR3 were selected adjacent to the inferred deletion breakpoint region (Fig. 18). Amplification with primer pairs DF1/DR1 and DF1/DR2 each gave a fragment of the expected size, 548 bp and 709 bp, respectively, in both the LEC and the control LE samples, whereas DF1/DR3 amplified a fragment of the expected size (922 bp) only in the control sample

(Fig. 20). These results indicate that the deletion breakpoint is located between the sequences represented by DR2 and DR3. Two other primers, DR4 and DR5, located between DR2 and DR3, were designed and used with DF1 in the RT-PCR analysis. A fragment of the predicted size (765 bp) was amplified from both LE and LEC DNA when DF1/DR5 was used (data not shown). With primer pairs DF1/DF4, a fragment of the expected size (853 bp) was produced only from the control LE but not from LEC cDNA (Fig. 20). These results localize the proximal deletion breakpoint within an approximately 90 bp segment between DR4 and DR5. Our data also indicate that transcription still occurs from the defective ATP7B gene in the LEC rat and that the processing of the defective transcripts was apparently carried out in the normal way for the regions examined.

Discussion

We have cloned overlapping cDNAs for ATP7B, the rat homologue of the human Wilson disease gene ATP7B. The coding sequences of the gene show a strong homology between rat and human. Overall structure and particularly individual functional domains are well conserved, confirming the functional importance of these regions. The rat gene has each of the six deletions and insertions (from 3 to 78 amino acids in length) found in the human ATP7B gene in comparison with the Menkes gene, ATP7A (2). It is

interesting to note that the sequences around the region equivalent to the metal binding domain 4 in the human ATP7B sequence are quite dissimilar between rat and human, in both nucleotide and amino acid sequences, as determined for several clones. Apparently, this part of the ATP7B protein has lost its function as a copper binding domain. Results from our Southern hybridization and RT-PCR analyses indicate a deletion in the ATP7B gene in the LEC rat. The deletion removes at least 900 bp of the 3' end coding region and about 400 bp of the 3' UTR. The deletion truncates the ORF of the ATP7B gene and removes the information encoding the conserved ATP binding domain and as such inevitably inactivates the function of the gene as a copper transporter. These results, together with previous

observations that implicated liver copper accumulation causally in the development of hepatitis in the LEC rat

(59, 61), demonstrate that the LEC mutant rat is an animal model for human Wilson disease, in which affected patients have mutations in the homologous ATP7B gene.

The similarities in clinical manifestations of the mutations, mainly those associated with liver disease, between the LEC rats and the Wilson disease patients, make this model an invaluable system for studying liver pathogenesis in Wilson disease and for developing and evaluating new treatment strategies. The chelating agent penicillamine has been used extensively in treatment, but is associated with undesirable side effects in as much as 20% of patients (63). Trientine and the competitive effects of high doses of zinc salts, have also been used, but better assessment of mechanism and protocol for maximum reduction of copper in the liver can now be tested. This disorder could also have potential for gene therapy, using transfected hepatocytes, as first demonstrated to ameliorate hyperlipidemia in the Watanabe rabbit (64).

Another feature of the LEC rats is the extremely high incidence of hepatocellular carcinoma in those rats that survive the initial attack of hepatitis at age 12 months or older. This is in contrast with Wilson disease, in which patients rarely develop liver cancer. This difference might be due, at least partially, to many biological differences between the two species. In accord with this hypothesis, transgenic mice with excessive storage of abnormal Z α₁-antitrypsin, as in human α₁-antitrypsin deficiency, also develop hepatocellular carcinoma (65), while this rarely occurs in the human patients. Another possibility is that patients with untreated Wilson disease may not survive long enough to develop cancers. The mechanisms leading to carcinogenesis in rats are still poorly understood, however, it is likely that the

abnormally high level of copper accumulation plays an important role in the process possible actions of the toxic level of copper include DNA-damaging effects from copper ions directly, or indirectly such as through the generation of free radicals (66, 67) and disturbance of expression of those genes controlled by certain zinc-finger transcription factors through replacement of zinc ion by copper.

This rat model can also be used to gain information on normal copper transport. Copper transport in the plasma has been well studied, but the mechanism of efflux from the liver is not well understood. The function of the Wilson disease gene must be closely associated with incorporation of copper into ceruloplasmin, as reflected by a very low ceruloplasmin concentration in most patients and in LEC rats. Questions of how the transport takes place and whether other proteins are involved can now be addressed. Also, because of the high degree of similarity between the cadmium and copper transporting ATPases in bacteria and the Wilson and Menkes disease genes (68), other heavy metals such as cadmium, may share or interact with the same transport system (69).

EXAMPLES OF APPLICATIONS OF THE INVENTION

The identification of the gene responsible for Wilson disease as well as markers associated with the disease has important implications for the development of new diagnostic and therapeutic strategies for the disease. Below are some examples of the use of the present invention in the diagnosis and treatment of Wilson disease.

A. Diagnosis of presymptomatic sibs After the first individual in a family is diagnosed with Wilson disease, there is frequently difficulty in determining whether other sibs, who have a one in four chance of being affected, are actually patients. We have demonstrated in some families (Cox, D.W. and Billingsley, G.D., The application of DNA markers to the diagnosis of presymptomatic Wilson disease. Proceedings of: Genetics of Psychiatric Diseases Wenner-Gren International Symposium, et. L. Wetterberg, Stockholm, pp. 167,988; Houwen et al., H. Hepatol. 17:269, 1993) that an incorrect diagnosis can be made, even when all possible biochemical and radioactive studies are carried out. Reliable diagnosis can be made with the markers we have developed. While other markers have been developed by others in this region, ours are particularly useful in that they are within about 200 kb of the Wilson disease gene, are very highly polymorphic, and the combination of these alleles, or haplotypes, have been studied both in our patients and in normal individuals (see Experiment 1). The markers we have found particularly useful are as follows:

D13S314 - 12 alleles

D13S315 - 9 alleles

D13S316 - 9 alleles

In addition, we have used, in our haplotypes, D13S133, a marker which was developed by others, which we have

identified as being very close to the Wilson disease locus.

Our own markers can be used to reliably diagnose

Wilson disease in sibs, and because of the high variability are most likely to be informative in all families. We have already successfully carried out presymptomatic diagnosis in at least six families.

B. Diagnosis of patients

Diagnosis of Wilson disease is particularly difficult for those with liver disease, since copper accumulation, characteristic of Wilson disease, also occurs in other liver diseases which have a biliary obstructive component. Every abnormal biochemical test in Wilson disease can be found to be abnormal in some other type of liver disease. For example, in addition to high liver copper, ceruloplasmin typically decreased in Wilson disease, may be elevated into the normal range.

1) Determination of haplotypes

In some cases, the haplotypes we have developed with our DNA markers, along with D13S133, can be used to increase the certainty of a diagnosis of WND that a patient has Wilson disease. This is because some of the haplotypes which occur in patients are rare in the general population. If a patient has one of these haplotypes, the chances of having a Wilson disease mutation are high. In combination with biochemical data, positive support for a diagnosis of Wilson disease could be obtained and treatment initiated immediately. Examples of haplotypes which are considerably more common in Wilson disease, and have not been found in the normal population are as follows: (refer to Experiment 1 for further description of haplotypes). These haplotypes are comprised of the following markers:

D13S316 - D16S133 - D13S314 - D13S315

Haplotype C: 6 - 17 - 10 - 5 (particularly in German patients)

Haplotype D: 5 - 11 - (5 or 4) - 5 (particularly in French patients)

Haplotype E: 6 - 17 - 11 - 4 (particularly in German and

British patients)

Among our patients of Northern European origin, these haplotypes represent 40% of a series of 47 random patients. This suggests that the haplotype approach could be useful in a relatively large proportion of cases.

This approach is useful even when the mutation is not known. However, direct analysis of the mutation will of course be more reliable. Typically, detection of all mutations for disease takes a considerable length of time, and may not be complete for years. 2) Mutation analysis

The proposed sequence can be used for the analysis of specific mutations in patients with Wilson disease. The direct analysis of such mutations has important implications for diagnosis. All regions of the sequence can be analyzed by methods such as the polymerase chain reaction, with primer sequences from within the cDNA region as given, or from intron sequences not presented as part of the present sequence. Any of the sequence which is amplified is included in the invention, whether amplified from sequences given or from sequences lying immediately

adjacent (in introns). The amplified portions of the sequence also include similar sequences which may have one or a few nucleotides altered, with the end result being amplilfication of the sequence given. Regions of 250 to

300 base pairs can be analyzed through mutation analysis by direct sequencing. Another method for detecting mutations is through the examination of fragments of 200 to 300 base pairs, which are then analyzed by single strand polymorphism confirmation (SSCP) analysis or by heteroduplex analysis. Either of these meethods can detect differences from the normal sequence. The exact mutation can then be confirmed by sequencing. However, once mutations are established, such a survey will be useful for direct mutation detection.

We have used specific primers, as shown below, to amplify a 275 base pair portion of the WND gene, followed by single strand conformation polymorphism (SSCP) and heteroduplex analysis: Two patients have been identified to date with this specific mutation.

B8.3a, 21-mer,

5' TGT AAT CCA GGT GAC AAG CAG 3'

B8.3b, 19-mer,

5' CAC AGC ATG GAA GGG AGA G

The same approach can be used to identify other mutations throughout the 4120 base pair sequence of the gene. a) Detection of point mutations

The sequence we have obtained is useful for the direct detection of mutations. Based on this sequence, we have developed PCR primers to amplify the functional motifs of the protein: copper binding, energy transduction, phosphorylation, and ATP binding. From our sequence, we have developed sequencing primers to sequence PCR products to identify mutations.

b) Detection of deletions or duplications

We have also developed primers which will be useful for the detection of deletions. We expect that a large proportion of the mutations in the Wilson disease gene will involve deletion (or duplications), particularly of the copper binding regions. Because there are six very similar motifs in the copper binding region, as we know from studies of the immunoglobulin heavy chain region carried out in our laboratory, deletions and duplications tend to occur frequently in the present of repeated sequences. In fact this has been demonstrated for Menkes disease (17). There are about 16% of patients with Menkes disease who have deletions in the copper binding region. The PCR primers we have developed can be used directly to identify such deletions. All of these primer sequences lie within the region we have submitted in this application.

For additional mutations, the intron exon boundaries we have sequenced will provide a useful source for PCR primers to amplify exons of the gene for the further search for mutations.

Therapy

Therapy for Wilson disease at the present time involves chelation of excess copper through the use of a chelating agent such as penicillamine, a potent agent which binds copper through its cysteine residues. But there are problems with the use of this agent, and the neurological symptoms can be worsened on initial treatment as copper is released from to the liver and transfers to other tissues, for example the brain. In addition, about 15% of the patients experience side effects from the therapy, including depression of the immune system, and reduction in the number of white and red blood cells. Zinc therapy is being used in some cases, but tends to cause intestinal irritation and is not tolerated well by some patients.

The new basis of therapy would involve introduction of the Wilson disease gene in a plasmid. We have discussed the copper and mercury containing plasmids in our publication (Thomas et al., manuscript). Copper is used

extensively in agriculture as a fungicide and bactericide. Certain bacteria have adapted to survive high copper conditions by replicating a high copy number of a plasmid which contains a sequence to encode an ATPase with a copper binding domain, very similar to the Wilson disease gene.

Creation of a construct similar to that found in the copper resistant bacteria therefore appears to be a possible approach.

Such a construct would then have to enter into the liver. This experiment has already been shown to be successful in the Watanabe rabbit, which has heritable hyperlipidemia, and demonstrated that allogenic hepatocytes can be transplanted into affected rabbits to ameliorate hypercholesterolemia (Wilson et al. PNAS 85:4421, 1988). In this rabbit, the gene was introduced in a plasmid construct attached to an asialoglycoprotein receptor, which targets to the liver cell. A similar approach is therefor feasible for a plasmid containing the Wilson disease gene. Human hepatocytes, cultured in vitro, can be transfected with an adenovirus containing the Wilson disease gene, and returned to the affected donor into the peripheral circulation. This model has already been tested in rats with the alpha₁-antitrypsin gene (Jaffe et al. Nature

Genetics 1, 1992). Partial hepatectomy can improve the stability of targeted DNA (Wilson et al. J. Biochem.

267:963, 1992). It is of interest that in these studies, DNA in the hepatocytes was present in stabilized plasmids, which do not self-replicate. The Wilson disease therefore could be used directly as a plasmid to be added to cultured hepatocytes, or perhaps to be administered directly through the portal system. Episomes which contained alpha₁-antitrypsin were found to remain relatively stable and produced the product (alpha₁-antitrypsin) for at least four months. Since even a low production of alpha₁-antitrypsin product should avoid the copper accumulation which takes place, this approach is technically feasible.

Other potential therapies

We have outlined in Example 2 that the Wilson disease gene is similar to genes on cadmium resistance and mercury resistance plasmids in bacteria. The similarity exists through all of the functional domains; metal binding, transduction phosphorylation and ATP binding. The Wilson disease gene could therefore be used, if incorporated into a plasmid construct, to remove excess cadmium or mercury from tissues. As expressed above, this is feasible for removal from the liver. Cadmium is particularly carcinogenic in the kidney, and it is of interest that the Wilson disease gene is expressed in kidney (Experiment 2).

Targeting of the gene to the kidney could alleviate cadmium toxicity in those who have been inadvertently exposed. The metal binding regions are very similar for the Wilson disease gene, and for the mercury and cadmium resistance plasmids. It is very likely that this sequence will be found to bind the other heavy metals. The differences outlined in Experiment 2, Figure 8, may suggest that slight alteration in the copper binding region could increase the specific binding for mercury and cadmium.

A construct containing the Wilson disease gene could potentially be used to overcome the defect in Menkes disease, since the copper binding region is very similar. A new process of targeting tissues with DNA-coated gold pellets (Yang et al. PNAS 87:9568, 1993) suggest that the intestinal cells, the site of the defect in Menkes disease, could be induced to incorporate Wilson disease DNA to allow transport of copper out of that tissue. Introduction of the plasmid into the intestinal epithelial cells seems also to be feasible.

Another approach, for both Wilson and Menkes disease would be to induce overexpression of the defective gene, which may be possible if there is residual activity of the gene product. We have found from our haplotype studies that most patients with Wilson disease appear to be genetic compounds, that is they probably carry two different mutations. At least one of these may have residual activity. Non-human applications

The Wilson disease gene could be targeted into the germ line of organisms for which the accumulation of toxic metals is a problem. For example, the targeting of the Wilson disease or of similar sequence into a plasmid into the germ line of fish stocks could increase the ability of such stocks to eliminate heavy metals, in regions which have naturally-occurring or pollution induced metal contamination.

Copper toxicity has been noted as a problem in sheep, as may also be a problem in other domestic species. It is possible that this toxicity in sheep is due to particularly low levels of expression of the homologous gene to the P-type ATPase described in this application for WND. The sequence presented may therefore have some application in therapy for toxicity in sheep, or in other animal species, or could be used in breeding to produce sheep, or other species which are more copper resistant. The sheep is given as only one example of an animal sensitive to copper toxicity. Other uses are also envisioned for the removal of copper or other toxic metals not only from sheep, but a variety of other organisms, including the removal of mercury from fish or any other species. The DNA sequence of the present invention can be used to obtain the equivalent gene from the mouse, to study the homologous gene. The human sequence in this application could be used to facilitate obtaining the sequence for the homologous gene in the toxic milk mouse, an inbred strain of mutant mouse, the defect in copper metabolism which may be identical to that of Wilson disease. Any use of the human sequence or a portion of it to be used for study of the toxic milk mouse and its normal counterpart are ineluded in this application.

While the above refer to specific applications of the present invention, it is to be appreciated that other uses, that are conceivable by one skilled in the art, are also within the scope of the present invention.

Table 1

Markers used in this study

Annealing No.

Locus Primers Temp Alleles PIC 133201^a 133202^a

D13S314 GAGTGGAGGAGGAGAAAAGA -62° 12 0.76 141/141 145/141

GTGTGACTGGATGGATGTGA

D13S315 GCCATCCAGAGTTAAACCA 58° 8 0.45 164/162 164/162

TTATAGCTTTTCTCATGCATTC

D13S316 GCAGCAATGCTTTGTGCATAA 62° 9 0.73 140/140 140/140

TGTTTCCCACCAATCTTACCG

D13S133 See Ref. (Petrukhin et al. 1993) ^aReference genotypes from CEPH family 1332. Numbers are allele sizes in base pairs.

Table 3

Marker distributions on Wilson disease family chromosomes

NE^a SE^a Sard^a ME^a Or^a Ip^a

Marker Allele N W N W N W N W N W N W

D13S316 1 2 0 0 0 0 0 0 0 0 0 0 0

2 1 1 2 0 0 0 0 0 0 0 0 0

3 2 3 2 1 1 0 1 0 0 0 1 0

4 19 3 2 1 3 3 2 2 0 0 2 10

5 3 9 2 3 0 0 1 1 1 1 2 0

6 4 16 1 0 1 0 0 1 0 0 0 0

7 19 26 2 8 0 3 1 2 3 3 2 1

8 4 2 1 1 0 0 0 0 0 0 3 0

9 1 0 2 0 0 0 0 0 0 0 0 0 ...... ....... ...... ....... ....... ....... ........

Total 56 60 14 14 5 6 5 6 4 4 10 11

D13S314 1 0 0 0 0 0 0 1 0 0 0 0 0

2 0 0 0 0 1 0 1 0 0 0 0 0

3 1 0 0 0 0 0 0 0 0 0 0 0

4 3 3 0 1 0 0 0 0 0 0 0 0

5 5 7 1 0 0 0 1 0 0 0 1 3

6 1 0 0 0 0 0 0 0 0 0 2 2

7 5 5 0 2 1 1 0 1 0 2 0 2

8 5 0 4 0 0 0 0 0 0 0 0 0

9 3 0 1 0 0 0 0 0 0 0 0 0

10 7 22 2 5 0 1 2 0 0 0 2 2

11 15 15 5 6 2 2 0 5 1 0 4 2

12 7 1 0 0 0 0 0 0 0 0 0 0 ....... ....... ...... ....... ....... ....... ........

Total 52 53 13 14 4 4 14 17 1 2 10 11 ^aFamilies were grouped according to geographical origin. NE=Northern European,

SE=Southern European, Sard=Sardinian, ME=Middle Eastern, Or=Oriental, IP=Indian/Pakistani Table 4

Marker distributions on Wilson disease family chromosomes

NE^a SE^a Sard³ ME^a Or^a IP^a

Marker Allele N W N W N W N W N W N W

D13S133 1 0 0 0 0 0 0 1 0 0 0 0 0

2 0 1 1 0 0 0 1 0 0 0 0 0

3 2 2 0 0 1 0 0 0 0 0 0 0

4 3 0 0 2 0 1 0 0 0 0 0 0

5 0 1 2 0 1 0 0 0 0 0 2 5

6 3 3 2 1 0 0 0 0 0 0 1 2

7 6 0 0 0 0 1 0 1 1 2 0 0

8 3 3 0 0 0 0 1 0 0 0 0 0

9 2 0 0 1 0 0 0 1 0 0 1 1

10 1 1 0 0 0 0 1 0 0 0 0 0

11 5 7 0 0 0 0 0 0 0 0 1 0

12 0 0 0 0 0 0 0 1 0 0 0 0

13 0 0 1 0 0 0 0 0 0 0 0 0

14 0 0 1 0 0 0 0 0 0 0 0 0

15 0 0 0 0 0 0 1 0 0 0 0 0

16 2 0 0 0 1 0 0 0 0 0 1 0

17 22 40 5 8 0 2 0 3 3 2 2 1 ...... ....... ...... ....... ....... ....... ........

Total 49 58 12 12 4 4 5 6 4 4 10 11

D13S315 1 3 0 0 0 1 1 1 0 0 0 0 0

2 5 4 3 0 1 0 1 0 0 1 2 1

3 4 9 1 1 1 0 1 2 0 2 3 2

4 35 27 8 9 3 4 2 1 3 1 2 3

5 9 17 1 1 0 0 0 0 0 0 3 2

6 0 0 1 2 0 0 0 1 0 0 0 0

7 1 1 0 0 0 1 0 2 1 0 0 3

8 0 1 0 1 0 0 0 0 0 0 0 0 ...... ....... ...... ....... ....... ....... ........

Total 57 58 14 14 6 6 5 6 4 4 10 11 ^aSee notes for table 2 Table 5

Haplotype distribution on chromosomes of Northern European origin

WND Normal

Group Haplotypes³ No. Freq No. Freq

A 7-17-10 9 3

8-17-10 1 1

10 0.21 4 0.09

B 7-17-11 8 8

7-16-11 0 1

8-17-11 0 3

8 0.17 12 0.27

C 6-17-10 7 0.15 0 0.00

D 5-11-5 4 0

5-11-4 2 0

5-10-5 1 0

7 0.15 0 0.00

E 6-17-11 5 0.11 0 0.00

F 7-17-5 1 0

7-17-4 1 0

8-17-5 1 0

3 0.06 0 0.00

G 2-6-7 1 0

3-6-7 1 0

2 0.04 0 0.00

5 others 5 0.11 4 0.09

18 others 0 0.00 24 0.55 ......... ......... ....... ....... ....... .......

Total 47 44

^aHaplotypes are given in the order: D13S316 - D13S133 - D13S314 Table 6 YACs in the Wilson disease region

D Number Probe Primers YACs

D13F71S1/2 pB32.3 CCGGGTATCTTAATTGGTGT 11G2;102F4;296G5

CTGGGGCCAACAATGTATTA 355C2

D13S196 pB40.3 GCAAAGTTCATAGGAAACCAGG 27D8;86A3;90H11;

ACATTTTGGTCAGACACTGGC 220A9£98H2

27R ATTGGGCATCTCTTGCTGTT 9B2;53C12;68F3

TGCAGGAATTCACTGTGTGA 95C3;407F11;117E9

EHR4 GGCCAGAATGACAAAATTCA 378B12;215B5;407F3

GGCTTCATGAGTGTGGTCCT

D13S31 235H9

Table 7

Primers and Conditions for SSCP and Sequencing

Product

Exon Primers length MgCl₂ Enzyme Fragments

1a gtttcaaggttaaaaaatgt 298 bp 1.5mM Banll 144 bp / 154 bp ggcacatatttcacagtgg

1b ggccaccagcacagtc 253 bp 1.5mM EcoNI 158 bp / 95 bp ctgggcaggcaaggac

1c gaggccagcattgcaga 282 bp 1.5mM Hintl 155 bp / 127 bp agccactttgctcttgatg

1d atgacatgggatttgaag 386 bp 1.5mM Hphl 158 bp / 228 bp tccgacaggaagagaaac

1e gcccaagtaaagtatgaccc 293 bp 1.5mM Banll 155 bp / 138 bp gacaccgatattrgctgcac

1f ggcacatgcagtaccactct 305 bp 1.5mM Haelll 118 bp / 187 bp agggctacctatacaccaiec

2 gatatttctgacattttatcc 320 bp 3.0 mM Mspl 152 bp / 168 bp gcagcattcctaagttca

3 ccacccagagtgttacagcc 229 bp 1.5mM Ddel 123 bp / 106 bp accccctaacgcaccca

4 cctgggtctgtgggattct 232 bp 1.5mM Mspl 158 bp / 74 bp aaaggtgactacaatttttaatga

5 ctgccaatgcaiaimaac 200 bp 1.5mM Haelll 93 bp / 107 bp ggtagaggaagggacttaga

6 tgtaatccaggtgacaagcag 276 bp 1.5mM Ncol 154 bp / 122 bp cacagcatggaagggagag

7 aacccttcactgrccttgtc 296 bp 1.5mM Banll 182 bp / 114 bp aggcagctctmctgaac

8 tttcgatagctctcaπrcaca 241 bp 1.5mM Styl 105 bp /136 bp tgcccacactcacaaggtc

9 agtcgccatgtaagtgataa 193 bp 1.5mM Avall 88 bp / 105 bp ctgagggaacaigaaacaa

10 ctgtcaggtcacatagtgct 278 bp 1.5mM Alul 149 bp / 129 bp tttcccagaactcttcaca

11 cttgtggtgttttatttcttc 229 bp 1.5mM Hincll 128 bp / 101 bp accaccaαtagcccaag Table 7 σont.

12 tgaactctcaacctgcct 266 bp 1.5mM Nlalll 162bp/104bp tctcagatgggaaagccg

13 tccatctgtattgtggtcag 301 bp 1.5mM Ddel 140 bp / 161 bp cagctaggagagaaggacat

14 ctttcacttcacccctct 255 bp 1.5mM Fokl 134 bp / 121 bp agctgaragagacaaaagc

15 203 bp 1.5mM Cfol 89bp/114bp aggaaggcagaagcaga

16 caagtgtggtatcttggtg 279 bp 1.5mM Nlalll 147bp/132bp ctggtgcttacttttgtctc

17 ttttgccaacactagcattcc 275 bp 1.5mM Ncol 125 bp / 150 bp tcccagcacccacagcc

18 ggcagaccccttcctcac 214 bp 1.5mM Avail 110 bp / 104 bp cctgggagagagaagccttt

19 ctaggtgtgagtgcgagtt 256 bp 1.5mM AM 127bp/129bp cagcatttgtcccaggt

20 aatggctcagatgctgtt 361 bp 1.5mM Ncol 163 bp / 198 bp gcttgtggtgagtggagg

Table ⁸

ATP7B Exons

Exon Start End Length Splice Acceptor& Donor Sites Domains

1 -44 1192 1237 tcttttctttttagATCTT...TTCTGgtacgt Cul, 2, 3 & 4

2 1193 1450 258 acattttatcctagAAAGC...AGCTGgtaaga Cu5

3 1451 1614 164 cctgatggttccagGTGTT...TGACAgtaagt Cu5

4 1615 1776 162 atcctgtgttgcagATCAC.TTGAGgtaagt Cu6

5 1777 1853 77 tttttaatgacaagGAAAT...AAGCAgtaggt

6 1854 2028 175 catttgctttccagTTGGA...TCCAGgtatat Tml & 2

7 2029 2262 234 ctccttgtctttagCTCCT...CAAAGgtaaca Tm2, 3 & 4

8 2263 2354 92 tggttatttcctagAGCAA...ATCAGgtgagt

9 2355 2482 128 ggcgtttgttgcagGGAGG...CACAGgtgaga Td

10 2483 2637 155 ttcctacgtcctagGAGAA...CAAAGgtaatg

11 2638 2772 135 tmattttcatagGCACC-TTCCTgtaagc Tm5

12 2773 2967 195 tgtcctgtmcagAACCC...ACAAGgtcagc Ch/Tm6, Ph

13 2968 3150 183 gttgtttttggcagATAAA...AAGAGgtacgt Ph, Tm7

14 3151 3319 169 ttccaccttcccagGAACT...AAAAGgtattg

15 3320 3463 144 cctctmgaatagATGCA...TGACGgtatct Tm8

16 3464 3606 143 ctggttcgctccagGTGTG...CCCAGgtacag Tm8

17 3607 3810 204 ttccttttgtctagGTTGG...TCAGAgtgagt ATP

18 3811 3928 118 ctcctctccatcagAATGA...AGCAGgtagga Tm9

19 3929 4031 103 tctttcttccccagGTGTC.AAGTGgtgagt Tm9

20 4032 ? ^b >400 gcctcctcttccagCTATA...CCCAGgtcagt° Table 8 ( cont . )

21 4205 ^{d a} Intronic sequences are shown as lowercase letters, exonic sequences are uppercase. The invariant residues at start and end of each intron are shown in bold.

b The 3' end of exons 20 has not been determined. The stop codon of the liver transcript is located within this exon.

o The given splice donor site for exon 20 is the site located 73 bp into die exon and appears to be used in kidney transcripts.

d The 3' end and flanking sequence of exon 21 have not been determined.

Table 9

Additional markers used in this study

Annealing No.

Locus Primers Temp Alleles PIC 133201^a 133202

D13S296 CAAACTTTTAGTATGAGTCTATCTC 62* 12 0.76 141/141 145/141

AAGTGAGGAGTGAGGTAAATG

D13S301 ATCATACCTGGTTGTGCAA 58* 8 0.45 164/162 164/162

TGATGCTTCTTTCTAAACACA

Reference genotypes from CEPH family 1332. Numbers are allele sizes in base pairs.

Table 10

Allele sizes for D13S296 and D13S301

Allele D13S296 D13S301

1 133 bp 152 bp

2 131 bp 150 bp

3 148 bp

4 127 bp 146 bp

5 144 bp

6 123 bp 142 bp

7 140 bp

8 119 bp 138 bp

9

10 115 bp 134 bp

11 113 bp

12 111 bp

13 109 bp

15 105 bp

17 101 bp 124 bp

Table 11

Haplotype distribution on chromosomes of Northern European origin

Original Extended WND Normal

Group Haplotype^a Haplotype^b No. Freq No. Freq

A 10-17-7 10-4-13-17-7 3 1

10-5-13-17-7 3 1 10-2-13-17-7 2 0

10-6-13-17-7 2 0

10-3-13-17-7 1 1

4 others 4 0

10-17-8 10-3-15-17-8 1 0

10-5-13-17-8 0 1

Total A 16 0.211 4 0.065

B 11-17-7 11-4-13-17-7 4 2

11-6-13-17-7 2 3

11-7-13-17-7 1 0

7 others 0 10

11-16-7 11-5-13-16-7 0 1

11-17-8 4 others 0 4

Total B 7 0.092 20 0.323

C 10-17-6 10-4-11-17-6 13 0

10-5-11-17-6 2 0

10-4- 11- 17-5 ά 2 0 10-4-15-17-6 1 0

10-5-11-17-5.5 1 0

Total C 19 0.250 0 0.000

D 3-11-5 3-3-8-11-5 6 0

3-10-5 3-3-8-10-5 1 0

Total D 7 0.092 0 0.000 Table 11 ( cont . )

E 11-17-6 11-4-11-17-6 5 0.066 0 0.000

F 3-17-7 3-4-15-17-7 1 0

4-17-7 4-5-15-17-7 0 1

4-17-8 4-4-15-17-8 1 0

Total F 2 0.026 1 0.016

G 7-6-2 7-7-1-6-2 1 0.013 0 0.000

H ND 7-8-6-4 3 0

7-7-64 1 1

Total H 4 0.053 1 0.016

15 others 15 0.197 5 0.081

28 others 0 0.000 31 0.500

Total Haplotypes 76 62 ^a Original haplotypes as defined previously [11] given in the order: D13S314 - D13S133 -

D13S316. ND = Not previously defined.

b Extended versions of original haplotypes are given in the order: D13S314 - D13S301 - D13S296

- D13S133 - D13S316. New haplotypes (ie. H) do not include D13S133.

Table 12

Haplotypes on non-Northern European chromosomes

Group^a Haplotypep^b WND Normal

IP 1 7-6-6-4 2 0

7-7-6-4 1 0

Total IP 1 3 0

IP 2 5-5-6-4 2 0

5-6-6-4 1 1

Total IP 2 3 1

IP 3 3-7-6-4 2 0

IP 4 10-17-6-4 2 0

IP 5 12-5-13-7 2 0

5 others 5 1

9 others 0 9

S I 10-4-6-4 2 0

S 2 11-4-13-7 2 0

2 others 2 0

3 others 0 3

SE 1 11-4-13-7 3 1

SE 2 10-4-11-6 2 0

7 others 7 1

8 others 0 8 ^a IP = Indian/Pakistani, S = Sardinian, SE = Southern European

b Haplotypes are given in the order D13S314 - D13S301 - D13S296 - D13S316 Table 13

Mutations in ATP7B

Changes Domain^a Ref Ethnic Groups Haplotypes^b No.

750delC Cu3 * Iranian 11-3-6-4 (†) 1

Greek 10-3-6-5 (†) 1

Y501H btw Cu5 & Cu6 * Scottish 7-6-6-4 (H) 1

1615-1G→C Cu6 * Indian/Pakistani 12-5-13-7 (IP5) 2

1651delAT Cu6 * British 10-10-15-7 (A) 1

1652insT Cu6 * Indian/Pakistani 3-3-6-4 (†) 1

2065dclA Tm3 * Italian 1-2-15-8 (†) 1

2206insC Tm4 * Scottish, Italian 10-11-8-5 (†) 2

British 12-3-13-7 (†) 1

British 10-2-15-8 (A) 1

R747L Tm4 * Chinese 6-7-10-5 (†) 1

Chinese 5-7-10-7 (†) 1

Chinese 6-3-13-7 (†) 1

2482+1G→C Td * British 10-3-13-7 (A) 1 L905X Tm5 * Saudi 11-7-6-4 (†) 1

Saudi, Greek 11-7-8-5 (†) 2

G912S Tm5 * Saudi 11-3-13-7 (†) 1

Saudi 11-4-13-7 (†) 1

2881delC Ch/Tm6 * British 11-5-13-7 (A) 1

British 11-6-13-7 (A) 1 Table 13 (cent. )

British 134-15-7 (†) 1

Sardinian 11-3-13-7 (S2) 1

Sardinian 10-3-64 (S1) 1

2992delAC Ph * British 7-6-1-2 (†) 1 3053delC Tm7 * Indian/Pakistani 3-6-64 (IP3) 2

H1038Q Tm7 [5] Eastern European 10-3-11-6 (C) 15

German, British 11-3-11-6 (E) 3

Eastern European 10-3-11-5.5 (C) 2

British 104-11-5.5 (C) 1

German 114-11-6 (E) 1

French 10-3-15-6 (E) 1

German 10-3-11-7 (A) 1

G1070R btw Tm7 & Tm8 * Indian/Pakistani 10-10-64 (IP4) 1

I1071T btw Tm7 & Tm 8 * Indian/Pakistani 7-5-64 (IP1) 2

3307delC btw Tm7 & Tm8 [5] Ukrainian 3-2-64 (†) 1

3463+1G→A Tm8 * British 3-3-15-7 (†) 1

3533del4 Tm8 * Swedish 11-6-64 (†) 1

3555del6 Tm8 * British 10-6-13-7 (A) 1

G1235K ATP hinge * French, British 3-2-8-5 (D) 7

N1239S ATP Hinge [5] Italian 11-10-13-7 (†) 1

R1288X btw ATP & Tm9 * British unknown 1

3998delTG after Tm9 * British 12-5-11-7 (†) 1 * : decribed in diis study Table 13 (cont. ) ^a Cu: copper binding domains, Tm: transmembrane regions, Td: energy transduction region, Ch: ion channel, Ph: phosphorylation region, ATP: ATP binding.

b Haplotype group (tables 3 and 4) is given in parentheses after the haplotype.

† : unique haplotype

The numbering of bases begins at the ATG initiator codon.

Table 14

Polymorphisms in ATP7B

Enzyme Sites

Changes Domain Ethnic Groups Creates Destroys

A375S Cu4 All Mnll NspBII

L435V btw Cu4 & Cu5 All - Pstl

K921R Tm5 All - -

2773-13G→C Ch/Tm6 All - Tthllll

T946M Ch/Tm6 All Nlalll -

2880G→A Ch/Tm6 All Taqll -

2916G→A Ch/Tm6 All - Cfol

A1109V Tm8 All - BbvI

3464-6C→T Tm8 British - Gsul

3606+27T→C Tm8 British - -

3810+6T→C ATP All - -

M738V Tm4 British HgiAI Nlalll

L745V Tm4 British Dsal EcoRII

V964A Ch/Tm6 Sikh Ncol -

M1138V Tm8 British, Italian Maelll -

The numbering of bases begins at the ATG initiator codon. Table 15 Comparison of LEC and Wilson disease

mutations and their associated features^a

LEC Wilson

Inheritance Autosomal recessive Autosomal recessive

Defective gene ATP7B^b ATP7B

Clinical

Liver disease Major clinical feature Major clinical feature

Hepatocellular carcinoma Yes Rare

Neurological abnormalities Rare Major clinical feature

Kayser-Fleischer ring No Yes

Laboratory findings

Liver copper Increased Increased

Ceruloplasmin Decreased Decreased

Serum copper Decreased Decreased

Treatment with chelating agents Effective Effective ^a Data are summarized from the references cited in the text.

b This study.

References

1. Brewer, G. J. & Yuzbasiyan-Gurkan, V. Wilson Disease.

Medicine 71, 139-164 (1992).

2. Sarkar, B. in Metal ions in biological systems (ed

Sigel, H.) 233-281 (Marcel Dekker, Inc., New York, 1981).

3. Orena, S. J., Goode, C. A. & Linder, M. C. Binding and uptake of copper from ceruloplasmin. Biochem. Biophys . Res . Comm . 139, 822-829 (1986).

4. Kagi, J. H. R. & Schaffer, A. Biochemistry of

metallothionein. Biochem 27, 8509-8515 (1988).

5. Danks, D. M. in Metaboli c Basis of Inheri ted Disease

(eds Beaudet, A.L., Sly, W.S. & Valle, D.) Vol. 6, 1411-1431 (McGraw-Hill, New York, 1989).

6. Darwish, H. M., Hoke, J. E. & Ettinger, M. J. Kinetics of Cu(II) transport and accumulation by hepatocytes from copper-deficient mice and the brindled mouse model of Menkes disease. J. Biol . Chem. 258,

13621-13626 (1983).

7. Frydman, M., Bonne-Tamir, B., Farrer, L. A., et al.

Assignment of the gene for Wilson disease to

chromosome 13. Proc Natl Acad Sci USA 82, 1819-1821 (1985).

8. Bowcock, A. M., Farrer, L. A., Cavalli-Sforza, L. L., et al . Mapping the Wilson disease locus to a cluster of linked polymorphic markers on chromosome 13. Am J Hum Genet 41, 27-35 (1987).

9. Bowcock, A. M., Farrer, L. A., Hebert, J. M., et al.

Eight closely linked loci place the Wilson disease locus within 13q14-q21. Am J Hum Genet 43, 664-674 (1988).

10. Yuzbasiyan-Gurkan, V., Brewer, G. J., Boerwinkle, E. & Venta, P. J. Linkage of the Wilson disease gene to chromosome 13 in North-American pedigrees. Am J Hum Genet 42, 825-829 (1988). 11. Farrer, L. A., Bowcock, A. M., Hebert, J. M., et al. Predictive testing for Wilson's disease using tightly linked and flanking DNA markers. Neurology 41, 992-999 (1991).

12. Thomas, G. R., Roberts, E. A., Rosales, T. O., et al.

Allelic association and linkage studies in Wilson disease. Hum. Mol . Genet . (1993) 2:1401-1405.

13. Houwen, R. H. J., Berger, R., Cox, D. W. & Buys, C. H.

C. M. Allelic association for Wilson disease-D13S31. J Hepatol 16, S15 (1992).

14. Bull, P. C, Barwell, J. A., Hannah, H. T-L., et al.

Isolation of new probes in the region of the Wilson disease locus, 13gl4.2-14.3. Cytogenet Cell Genet 64, 12-17 (1993).

15. Bull, P. C. & Cox, D. W. Long range restriction

mapping of 13q14.3 focused on the Wilson disease region. Genomics 16, 593-598 (1993).

16. Thomas, G. R., Bull, P. C, Roberts, E. A., Walshe, J.

R. & Cox, D. W. Haplotype studies in wilson disease. Am J Hum Genet (1993). (in press)

17. Vulpe, C, Levinson, B., Whitney, S., Packman, S. & Gitschier, J. Isolation of a candidate gene for Menkes disease and evidence that it encodes a

copper-transporting ATPase. Nature Genet . 3, 7-13 (1993).

18. Mercer, J. F. B., Livingstone, J., Hall, B., et al.

Isolation of a partial candidate gene for Menkes disease by positional cloning. Nature Genet . 3, 20-25 (1993).

19. Chelly, J., Turner, Z., Tonnesen, T., et al . Isolation of a candidate gene for Menkes disease that encodes a potential heavy metal binding protein. Nature Genet . 3, 14-19 (1993).

20. Sass-Kortsak, A. Copper metabolism. Adv. Clin . Chem.

8, 1-67 (1965). 21. Owen, C. A. Jr. Metabolism of radiocopper (⁶⁴Cu) in the rat. Am. J. Physiol . 209, 900-904 (1965).

22. Rommens, J. M., Lin, B., Hutchinson, G. B., et al. A transcription map of the region containing the

Huntington disease gene. Hum. Mol . Genet . 2 , 901-907

(1993).

23. Goldberg, Y. P., Lin, B. -Y., Andrew, S. E., et al.

Cloning and mapping of the a-adducin gene close to D4S95 and assessment of its relationship to Huntington disease. Hum. Mol . Genet . 1, 669-675 (1992).

24. Vilsen, B., Andersen, J. P., Clarke, D. M. &

MacLennan, D. H. Functional consequences of proline mutations in the cytoplasmic and transmembrane sectors of Ca⁺⁺-ATPase of sarcoplasmic reticulum. J. Biol . Chem. 264, 21024-21030 (1989).

25. Silver, S., Nucifora, G., Chu, L. & Misra, T. K.

Bacterial resistance ATPases: primary pumps for exporting toxic cations and anions. Trends Biochem Sci 14, 76-80 (1989).

26. Odermatt, A., Suter, H., Krapf, R. & Solioz, M.

Primary structure of two P-type ATPases involved in copper homeostasis in Enterococcus hirae . J. Biol . Chem. 268, 12775-12779 (1993).

27. Griffin, H. G., Foster, T. J., Silver, S. & Misra, T. K. Cloning and DNA sequence of the mercuric and organomercurial resistance determinants of plasmid pDU1358. Proc Natl Acad Sci USA 84, 3112-3116 (1987).

28. Nucifora, G., Chu, L., Misra, T. K. & Silver, S.

Cadmium resistance from Staphylococcus aureus plasmid pI258 cadA gene results from a cadmium-efflux ATPase. Proc Natl Acad Sci USA 86, 3544-3548 (1989).

29. Yarze, J. C, Martin, P., Munoz, S. J. & Friedman, L.

S. Wilson Disease: current status. Am . J. Med. 92, 643-654 (1992).

30. Sternlieb, I. Perspectives on Wilson Disease.

Hepatology 12, 1234-1239 (1990). 31. Yang, F., Naylor, S. L., Lum, J. B., et al.

Characterization, mapping, and expression of the human ceruloplasmin gene. Proc Natl Acad Sci USA 83,

3257-3261 (1986).

32. Sato, M. & Gitlin, J. D. Mechanisms of copper

incorporation during the biosythesis of human

ceruloplasmin. J. Biol . Chem . 266, 5128-5134 (1991).

33. Gitlin, J. D., Schroeder, J. J., Lee-Ambrose, L. M. & Cousins, R. J. Mechanisms of ceruloplasmin

biosynthesis in normal and copper-deficient rats.

Biochem. J. 282, 835-839 (1992).

34. O'Halloran, T. V. Transition metals in control of gene expression. Science 261, 715-725 (1993).

35. Scherer, S. & Tsui, L-T. in Advanced techniques in

chromosome research (ed Adolph, K.W.) 33-72 (Marcel Dekker, Inc., New York, Basel, Hong kong, 1991).

36. Triglia, T., Peterson, M. G. & Kemp, D. J. A procedure for in vitro amplification of DNA segments that lie outside the boundaries of known sequences. Nucl Acids Res 16, 8186-8180 (1988).

37. Sambrook, J. Fritsch, E. F. & Maniatis, T. Molecular cloning, a laboratory manual (Cold Spring Harbor

Laboratory Press, Cold Spring Harbor, 1989).

38. Kahn, D., David, M., Domergue, O., et al. Rhizobium melitoli fixGHI sequence predicts involvement of a specific cation pump in symbiotic nitrogen fixation. J bacteriol 171, 929-939 (1989).

39. Cox DW, Fraser FC and Sass-Kortsak A (1972) A genetic study of Wilson's disease: evidence for heterogeneity. Am J Hum Genet 24:646-666

40. Miller SA, Dykes DD and Polesky HF (1988) A simple

salting out procedure for extracting DNA from human nucleated cells. Nucl Acids Res 16:1215

41. Oudet C, Mornet E, Serre JL, Thomas F,

Lentes-Zengerling S, Kretz C, Deluchat C, et al (1993) Linkage disequilibrium between the fragile X mutation and two closely linked CA repeats suggests that fragile X chromosomes are derived from a small number of founder chromosomes. Am J Hum Genet 52:297-304

42. MacDonald ME, Lin C, Srinidhi L, Bates G, Altherr M, Whaley WL, Lehrach H, et al (1991) Complex pattern of linkage disequilibrium in the Huntington disease region. Am J Hum Genet 49:723-734

43. Bull PC, Thomas GR, Rommens JM, Forbes JR and Cox DW

(1993) Identification of a putative copper binding P-type ATPase similar to the Menkes (MNK) gene: a candidate for the Wilson disease gene. (submitted)

44. Bull PC, Thomas GR, Rommens JM, Forbes JR, Cox DW:

The Wilson disease gene is a putative copper

transporting P-type ATPase similar to the Menkes gene. Nat Genet 1993; 5: 327-337.

45. Byth BC, Thomas GR, Hofland N, Cox DW: Application of inverse PCR to isolation of end probes from cosmids. Nucl Acids Res 1994; 22: 1766-1767.

46. Thomas GR, Roberts EA, Walshe JM, Cox DW: The Wilson disease gene: mutations and haplotypes. Am J Hum Genet 1994; 54: 71-78.

47. Tanzi RE, Petrukhin KE, Chernov I, et al: The Wilson disease gene is a copper transporting ATPase with homology to the Menkes disease gene. Nat Genet 1993; 5: 344-350.

48. Breathnach R, Chambon P: Organization and expression of eucaryotic split genes coding for proteins. Ann Rev Biochem 1981; 50: 349-383.

49. Wu J, Forbes JR, Shiene Chen H, Cox DW: The LEC rat has a deletion in the copper transporting ATPase gene homologous to the Wilson disease gene. Nat Genet 1994; 7: 541-545.

50. Petrukhin KE, Fischer SG, Pirastu M, et al: Mapping, cloning and genetic characterization of the region containing the wilson disease gene. Nat Genet 1993; 5: 338-343. 51. Thomas GR, Forbes JR, Cox DW: The Wilson disease gene: exon structure. Genomics 1994; (In Press).

52. Houwen RHJ, Thomas GR, Roberts EA, Cox DW: DNA

markers for the diagnosis of Wilson disease.

J Hepatol 1993; 17: 269-276.

53. Kerem B, Rommens JM, Buchanan JA, et al: Identification of the cystic fibrosis gene: genetic

analysis. Science 1989; 245: 1073-1080.

54. Sasaki, M. et al. Spontaneous hepatitis in an inbred strain of Long-Evans rats. Rat News Lett. 14, 4-6 (1985).

55. Yoshida, M.C. et al. New mutation causing hereditary hepatitis in the laboratory rate. J. Hered. 78, 361- 365 (1987).

56. Masuda, R., Yoshida, M.C, Sasaki, M., Dempo, K. &

Mori, M. Hereditary hepatitis of LEC rats is

controlled by a single autosomal recessive gene. Lab. Animal 22, 166-169 (1988).

57. Masuda, R., Yoshida, M.C, Dempo, K. & Mori, M. High susceptibility to hepatocellular carcinoma development in LEC rats with hereditary hepatitis. Jpn. J. cancer Res. 79, 925-835 (1988).

58. Sawaki, M., Enomoto, K., Takahashi, H., Nakajima, Y. & Mori, M. Phenotype of preneoplastic and neoplastic liver lesions during spontaneous liver carcinogenesis of LEC rats. Carcinogenesis 11, 1875-1861 (1990).

59. Sone, H. et al. Genetic linkage between copper

accumulation and hepatitis/hepatoma development in LEC rats. Molec. Carcinog. 5, 199-204 (1992).

60. Ono, T., Abe, S. & Yoshida, M.C. Hereditary low level of plasma ceruloplasmin in LEC rats associated with spontaneous development of hepatitis and liver cancer. Jpn. J. cancer Res. 82, 486-489 (1991).

61. Li, Y., Togashi, Y. & Takeichi, N. Abnormal copper accumulation in the liver of LEC rats: a rat form of Wilson's disease. In The LEC rat, a New Model for Hepatitis and Liver Cancer (eds Mori, M., Yoshida, M.C, Takeichi, N. & Taniguchi, N.) 122-132 (SpringerVerlag, Tokyo, 1991).

62. Li, Y. et al. Spontaneous hepatic copper accumulation in Long-Evans cinnamon rats with hereditary hepatitis: A model of Wilson's disease. J. clin. Invest. 87, 1858-1861 (1991).

63. Walshe, J.M. Wilson's disease presenting with

features of hepatic dysfunction: a clinical analysis of eight-seven patients. Q.J. Med. New. Ser. 70, 253- 263 (1989).

64. Wilson, J.M., Johnston, D.E., Jefferson, D.M. &

Mulligan, R.C Correction of the genetic defect in hepatocytes from the Watanabe heritable hyperlipidemic rabbit. Proc. natn. Acad. Sci. U.S.A. 85, 4421-4425 (1988).

65. Geller, S.A. et al. Hepatocarcinogenesis is the

sequel to hepatitis in Z#2 eel-antitrypsin transgenic mice: histopathological and DNA ploidy studies.

Hepatology 19, 389-397, 1994.

66. Tkeshelashvili, L.K., McBride, T., Spence, K. & Loeb, L.A. Mutation spectrum of copper-induced DNA damage. J. biol. Chem. 266, 6401-6406 (1992).

67. Loeb, L.A., James, E.Q., Waltersdorph, A.M. &

Klebanoff, S.J. Mutagenesis by the autoxidation of iron with isolated DNA. Proc. natn. Acad. Sci. u.s.a. 85, 3918-3922 (1988).

68. Bull, P.C & Cox, D.W. Wilson disease and Menkes

disease: new handles on heavy metal transport.

Trends Genet, (in the press).

69. Dancis, A. et al. Molecular characterization of a

copper transport protein in S. cerevisiae: An

unexpected role for copper in iron transport. Cell 76, 393-402 (1994).

Claims

CLAIMS :

1. A nucleotide sequence containing the gene for Wilson disease.

2. The nucleotide sequence according to claim 1 which comprises the DNA sequence as illustrated in Figure 5, the complementary strand or any modified derivative or fragment thereof.

3. A nucleotide sequence containing the gene for a copper transporting ATPase defective in Wilson disease.

4. The nucleotide sequence according to claim 1 or 3 which comprises the DNA sequence as illustrated in

Figure 10 the complementary strand or any modified

derivative or fragment thereof.

5. A copper binding domain of the gene for Wilson disease which has the DNA sequence designated as Cu1 in Figure 10.

6. A copper binding domain of the gene for Wilson disease which has the DNA sequence designated as Cu2 in Figure 10.

7. A copper binding domain of the gene for Wilson disease which has the DNA sequence designated as Cu3 in Figure 10.

8. A copper binding domain of the gene for Wilson disease which has the DNA sequence designated as Cu4 in Figure 10.

9. A copper binding domain of the gene for Wilson disease which has the DNA sequence designated as Cu5 in Figure 10.

10. A copper binding domain of the gene for Wilson disease which has the DNA sequence designated as Cu6 in Figure 10.

11. A phosphatase-transduction domain of the gene for Wilson disease which has the DNA sequence designated as Pt/T in Figure 10.

12. A transmembrane domain of the gene for Wilson disease which has the DNA sequence designated as Tm in Figure 10.

13. A phosphorylation domain of the gene for Wilson disease which has the DNA sequence designated as Ph in Figure 10.

14. A ATP binding domain of the gene for Wilson disease which has the DNA sequence designated as ATP-hinge in

Figure 10.

15. A use of a nucleotide sequence according to any one of claims 1 to 14, or a derivative or fragment thereof, to detect Wilson disease.

16. A DNA marker associated with the gene for Wilson disease characterized in that it detects the same dinucleotide repeat polymorphism as DNA marker D13S314.

17. The DNA marker according to claim 16 wherein said marker is D13S314 as defined in Table 1.

18. A DNA marker associated with the gene for Wilson disease characterized in that it detects the same

dinucleotide repeat polymorphism as DNA marker D13S315.

19. The DNA marker according to claim 18 wherein said marker is D13S315 as defined in Table 1.

20. A DNA marker associated with the gene for Wilson disease characterized in that it detects the same

dinucleotide repeat polymorphism as DNA marker D13S316.

21. The DNA marker according to claim 20 wherein said marker is D13S316 as defined in Table 1.

22. A DNA marker associated with the gene for Wilson disease characterized in that it detects the same

dinuclotide repeat polymorphism as DNA marker D13S301.

23. The DNA marker according to claim 22 whereins aid marker is D13S296 as defined in Table 9.

24. A DNA marker associated with the gene for Wilson disease characterized in that it detects the same

dinuclotide repeat polymorphism as DNA marker D13S301.

25. The DNA marker according to claim 24 wherein said marker is D13S301 as defined in Table 9.

26. A use of a DNA marker according to any one of

claims 16 to 25 to detect Wilson disease.

27. A kit for detecting Wilson disease comprising at least one DNA marker selected from the group consisting of

D13S314, D13S315, D13S316 D13S296 and D13S301.

28. A use of a nucleotide sequence according to any one of claims 1 to 14, or a derivative or fragment thereof, to treat Wilson disease.

29. A use of a nucleotide sequence according to any one of claims 1 to 14, or a derivative or fragment thereof, to isolate the Wilson disease gene, or fragment thereof, from a mammal.

30. A use of a primer derived from a nucleotide sequence according to any one of claims 1 to 14 to detect a mutation in the Wilson disease gene in a Wilson disease patient.

31. A use according to claim 30 wherein said primer has the following DNA sequence:

5' TGT AAT CCA GGT GAC AAG CG 3'.

32. A use according to claim 30 wherein said primer has the following DNA sequence:

5' CAC AGC ATG GAA GGG AGA G 3' .

33. A use of a nucleotide sequence according to any one of claims 1 to 14 or a modified derivative or fragment thereof to reduce metal toxicity in an animal.

34. A nucleotide sequence containing the gene for a copper transporting ATPase defective in teh long Evans Cinnamon rat.

35. The nucleotide sequence according to claim 34 which comprises the DNA sequence as illustrated in Figure 17 the complementary strand or any modified derivative or fragment thereof.