Mutational analysis of ATP 7 B gene and the genotype-phenotype correlation in patients with Wilson ' s disease in Serbia

Background/Aim. Wilson’s disease (WD) is an autosomal-recessive disorder which is characterized with a marked clinical heterogeneity. The gene responsible for WD is located in 13q14.3 chromosome, contains 21 exons and codes for copper specific transporting P-type adenosinetriphosphatase (ATPase) (ATP7B). Mutations in ATP7B gene change biosynthetic and transporting role of ATPase in cell leading to damaged billiary excretion of copper and its accumulation in the liver, brain, cornea and other tissues. Until now, it has been described more than 400 mutations in ATP7B gene with characteristic geographic distribution. The aim of this study was to assess the spectrum of mutations of ATP7B gene on a large number of patients in Serbian population and to make a correlation between particular genotypes and specific phenotypes. Methods. Eighty-six consecutive patients with WD from WD Clinical Research programme were included in this study. Genetic analysis was performed by direct gene sequencing method. Results. Mutations in ATP7B gene were found in 93% analyzed patients (81.4% of all alleles analyzed). Thirteen mutations were identified, one of which (G998E) was the novel one, so far undescribed in the literature. The most frequent mutation in our population was H1069Q, which was present in 38.4% patients, and the second most frequent mutation was 2304-2305insC (11.6%). Also, a great number of gene polymorphisms of DNA sequences, which do not disturb the ATP7B gene function, was identified. Although neurological form of the disease was more frequent in the group of homozygous for H1069Q and the group of nonH1069Q carriers, there was no statistically significant difference between the groups. Conclusion. Our research showed that genetic diagnosis of WD can be done in 80% of cases by analysis of 5 most common mutations in our population, which facilitate diagnosis significantly. There was no correlation between different genotypes and specific phenotypic features of WD, the presence of psychiatric disturbances and cognitive deterioration.


Introduction
Wilson's disease (WD) is an autosomal-recessive disorder characterized with marked clinical heterogeneity.Gene responsible for WD is located on chromosome 13q14.3 1,2 It consists of 21 exons and encodes for a copper-transporting P-type adenosinetriphosphatase(ATPase) (ATP7B) 3 .Mutations in ATP7B gene change biosynthetic and transporting role of ATPase in cell, resulting in impaired billiary excretion of copper and its accumulation in liver, brain, cornea and other tissues.The disease incidence is estimated to be 1 in 30,000 to 1 in 50,000 4,5 , and it is supposed that the prevalence is 30 per million, with frequency of heterozygous mutations carriers of about 1 in 90 to 1 in 150 6,7 .Until now, it has been identified more than 400 mutations in ATP7B gene with characteristic geographic distribution 8,9 .Most mutations are present in particular families, in particular cases. I previous study of our population, molecular gene defect was identified in 80% of alleles of WD patients 10 , with 11 different mutations (3 of which were new ones).The most frequent mutations were H1069Q (48.9% of analyzed alleles), 2304-2305insC (11.4%), and A1003T (5.7%) 10 .
The disease can be presented through 3 forms: hepatic, neurological and psychiatric 11 .There is no clear evidence of associations between specific genotypes and some clinical features of the disease.Therefore, the aim of this study was to assess spectrum of mutations in ATP7B gene on a larger number of patients in Serbian population, and to make a correlation between particular genotypes and specific phenotypes.

Methods
Eighty six consecutive patients with WD from the WD Clinical Research programme were included in this study.The diagnosis of WD was made based on standard criteria: anamnesis, clinical examination, ceruloplasmin and copper serum level, copper level in 24-hour urine, liver biopsy, "slit"-lamp examination for the presence of Kayser-Fleischer's ring.
Study was approved by the Ethic Committee, Clinical Center of Serbia, Belgrade.Upon signing informed consent, the patients were interviewed in order to obtain demographic and clinical data on disease course, therapy, etc.All the patients passed through neurological and hepatic examination, as well as brain nuclear magnetic resonance (NMR).
Disability level and grading of WD and its multisystem manifestations were measured by the Global Assessment Scale (GAS) for WD 12 .This scale has 2 tiers: tier 1 assesses global disability in 4 domains: liver, cognition and behaviour, motor and osseomuscular functions (each domain is scored independently on a 6-point ascendant scale (0-5) and results are not added), while tier 2 is multidimensional scale for a fine grained evaluation of the neurological dysfunction 14 items are scored on a 5-point ascendant scale (0-4) and their sum (0-56) presents a total score of tier 2].Psychiatric comorbidity was examined through the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID) 13 .The SCID for every patient was performed by the trained psychiatrist.Cognition was estimated by the Mini Mental State Examination (MMSE) scale 14 .Genotype-phenotype analysis was made upon association of specific molecular defect and disease form, age at onset, the presence of psychiatric disturbances, and GAS for WD scores.
DNA extraction and polymerase chain reaction (PCR) analysis were made by standard methods.Mutation detection was performed by the direct sequencing method.Exons 5, 8, 13-15 of ATP7B gene were amplified using primers complementary to DNA sequences flanking the exon-intron boundaries.PCR products were purified using a standard purifying protocol and then sequenced on an automatic sequencer (ABI 310 Genetic Analyzer).Point mutations were detected reliably by means of manual inspection of characteristic double peaks.Mutation analysis was carried out according to following strategy: firstly, exon 14 was analyzed for the most frequent mutation in Serbian WD patients (H1069Q), secondly, if no or only one mutation was found, exons 8, 5, 15 and 13 were analyzed, as locations of other most common mutations.
Data were analyzed using methods of descriptive statistic, analysis of variance (ANOVA) and 2 -test.

Results
The study consisted of 86 patients with demographic and clinical characteristics shown in Table 1.A predominant disease form was neurological (54.7%), while hepatic form was present in 41.9% of the patients.Therapy was taken continuously in 71.8% of the patients, mostly dpenicillamine (84.4%).
Molecular defects in ATP7B gene were found in 93% patients (81.4% of all the analyzed alleles).Thirteen mutations were identified: 2 frameshifts, 1splicing and 10 mis-Strana 459 sense mutations, one of which (G998E) was the novel one, described for the first time in the literature.Six patients had no known mutations, while three of them had gene polymorphisms.The most frequent mutation in our population, H1069Q, was present in 38.4% of the analyzed alleles, followed by insertion 2304-2305insC in exon 8 (11.63% of analyzed alleles).Two missense mutations, A1003T in exon 13 and R616Q in exon 5 occurred with the frequencies of 9.3% and 8.1%, respectively.Relatively frequently found mutations were also 3402delC (2.32%) and R969Q (1.74%), while G591D and I1102T were individually observed in two heterozygous patients (1.16%).Each of the remaining mutations was present only on one allele of the single patient (Table 2).Furthermore, some DNA polymorphisms that do not disrupt ATB7B gene functions were identified (Table 3).
The patients were divided into three groups according to the found mutations: the group of homozygous for H1069Q, the group of heterozygous for H1069Q, the non-H1069Q patients.H1069Q mutation was identified in 53 patients (13 homozygous and 40 heterozygous).Twenty-seven patients had no H1069Q mutation, but some other ATP7B gene mutation, while 6 patients did not have any of the known mutations.In the group of homozygous for H1069Q more patients had neurological (10) then hepatic (2) disease form.Both clinical presentations had the same frequency (20) in patients heterozygous for H1069Q.There was no significant difference between neurological and hepatic form (13 and 12) in the group of non-H1069Q patients.Although neurological form of disease was more frequent in the groups of homozygous for H1069Q and non-H1069Q patients, there was no statistically significant difference between the examined groups ( 2 = 4.664; p = 0.198).Also, there was no statistically significant difference in frequency of specific mutations in different neurological forms ( 2 = 12.78; p = 0.172).
Age at onset was not statistically significantly different in the examined groups with different genotypes (ANOVA; F = 1.236; p = 0.302).In group of homozygous for H1069Q dis- ease onset was at the age of 22.6 ± 7.6 years, in heterozygous for H1069Q at 25.6 ± 9.0 years, while in non-H1069Q patients disease started at the age of 23.0 ± 8.9 years (Table 4).The influence of genotype on severity of clinical presentation was examined through correlations of specific mutations and GAS for WD scores.There was no statistically significant difference among the examined genotype groups and subscores for the assessment of hepatic (ANOVA, F = 0.131; p = 0.941), cognitive and behavioural (ANOVA; F = 1.476; p = 0.231), motor (ANOVA; F = 0.550; p = 0.650) and osseomuscular impairment (ANOVA; F = 0.681; p = 0.567).Neither cognitive impairment, estimated by MMSE, (ANOVA; F = 0.428; p = 0.734) (Table 5), nor the frequency of psychiatric manifestations (Table 4) ( 2 = 9.109; p = 0.168) differed among various genotype groups.

Discussion
Our study identified 13 mutations (12 formerly known, 1 new) in ATP7B gene in 81.3% of the analyzed alleles.Our results confirm that the spectum of mutations in WD consists of a small number of relatively frequent mutations and a large number of rare mutatons.
The most frequent mutatation in our population is substitution of histidine with glutamine on the position 1069 (H1069Q) in exon 14, which is found in 38.4% of the analyzed alleles.Also, this is the most common mutation responsible for WD worldwide, present only in Caucasions 10 , and seems to originate from the Central or Eastern Europe 15 .Its frequency is highest in Poland and Eastern Germany (50-80%) 4 .In Europe, south of the Alps, this mutation becomes infrequent and it is totally absent in Sardinia and Sicily, which can bee explained by teritorial isolation and high rate of consanguinity (incidence of WD in these islands is approximately one per 7,000) 16 .This mutation is always found in the same haplotype, haplotype VIII defined by STR loci D13S301, D13S296, D13S297, D13S298 17 (allele combination 5, 9, 4, 3, respectively) in every analyzed population.Regarding this, it can be concluded that H1069Q is a very old mutation 16,17 .
The second most frequent mutation in our population, insertion 2304-2305insC in exon 8 (11.6% of examined alleles) is present in 6% of WD patients of the continental part of Italy and in 2.6% of population of Russia and North America 16 .However, the third most common mutation in our group, substitution of alanine with threonine (A1003T) in exon 13 (9.3% of the analyzed alleles) shows no important frequency in other European population.Substitution of arginine with glutamine in the position 616 (R616Q) in exon 5 is the ultimate mutation with significant frequency in our research that was previously described only in one WD patient in Great Britain 18 and Italy 19 .Deguti et al. 20 described deletion 3402delC in exon 15 (2.3% of the analyzed alleles in our population) as the most common gene defect in Brazil with the frequency of 30.8%, as well as in Russian population (19%) 4,10 .It is interesting that this mutation was not found in Mediterranean region, although the dominant immigrant group in the South America was from this region 19,21 .Substitution R969Q is frequent in Greece (18%) 22 , while in our study it was identified only in 1.74% of the analyzed alleles.
Strategy for search of pathological mutations in WD by the direct sequencing method was initially directed towards exons 5, 8, 13, 14 and 15, as loci with majority of the known mutations in our population, as well as in ethnical groups of Central and Eastern Europe.
The main reason for great clinical heterogeneity in WD has not been explained yet, but it is believed that this is caused by different types of mutations in ATP7B gene and their effect on ATP7B function (various mutations cause different impairments of ATP7B protein func-  tions) 23 .Studies that tried to explore influence of ATP7B genotype on clinical presentation and disease course showed no consistent results.Estimation of genotypephenotype correlation is difficult because of large number of mutations (more than 400) with a marked genetical heterogeneity, small number of patients in the majority of studies, and significant hetererozigocity 23 .The frequency of H1069Q mutations in homozygous state was the only one that met criteria for gentype-phenotype analysis.Some studies showed that homozygous for H1069Q presented first disease symptoms up to 12 years later than compound heterozygotes and non-H1069Q patients, as well as less severe disturbance in copper metabolism 24 .This clinical expression can be explained by partially perserved ATP7B protein function, with residual billiary copper excretion present, resulting in slower accumulation in liver.Our study did not confirm prevously described clinical characteristic of patients with homozygous H1069Q mutation.It has been suggested that nonsense, splice-site and frameshift mutations are associated with severe phenotypic expression due to the production of non-functional protein product, while missense mutations, that do not disturb pro-tein function totally, are marked as "mild" 24,25 .Gromadzka et al. 24 demonstrated on 142 WD patients that carry of one or two "severe" mutation have dose-dependent effect on lowering of ceruloplasmin level and younger disease onset.We tried to asses a correlation between genotype and severity of clinical expression using GAS for WD, but no significant associations were observed.

Conclusion
Our study showed that the genetic diagnosis of WD can be made by analysis of the 5 most frequent mutations in our population in more than 80% of cases, which can significantly facilitate diagnosis.Unfortunately, there was no correlation made between the examined genotypes and the specific phenotypic features of WD (disease form and severity of clinical expression), the presence of psychiatric disturbances and cognitive deterioration.

Table 4 Influence of genotype of the patients with Wilson's disease on the disease onset
N -neurological form; H -

Table 5 Influence of genotype of the patients with Wilson's disease (WD) on severity of the disease presentation
GAS -Global Assessment scale; MMSE -Mini-Mental State Examination scale.