The Study Of Modifier Genes In SCA2 And Natural Selection On The ATXN2 Gene

A Parkinsonism pedigree was collected by our group. According to clinical characteristics, primary diagnosis showed that it was an autosomal dominant (AD) Parkinson Disease pedigree. The known pathogenic genes of AD Parkinson Disease were excluded in our previously study and a whole genomic scan was performed by gene chip. The pathogenic gene ATXN2 was located on chromosome 12, the candidate region of patient's genome. Thereby, this pedigree was finally confirmed as a Spinocerebellar Ataxia type 2 (SCA2) family.Spinocerebellar ataxia type 2 (SCA2) is a neurodegenerative disease caused by the expansions of CAG trinucleotide repeat in exon 1 of the ATXN2 gene, which has severe influences on sensory and motor controlling functions. Because the patients exhibited different clinical phenotypes in this family with other common type of SCA2 and were sensitive to levodopa treatment, thus, to investigate whether modifier genes exist, log of the odds (LOD) were recomputed according to the settings of this SCA2 family phenotype mode using the preliminary work data of the whole genomic scan. The results showed that two LOD scores were more than 3, one scroe existed on chromosome 5, another was on chromosome 12, where the SCA2 pathogenic gene ATXN2 located at. Subsequently, haplotypes were constructed and analyzed in the location region on chromosome 5 depending on the results above. The same haplotype "3322323" was exhibited among almost all of the patients. Based on these, we presumed that modifier genes existed in this region of chromosome 5, which made the clinical phenotype deviation in favor of Parkinson Disease trails and age of onset ahead of time. Therefore, DNA sequencing was operated on five candidate genes in the location region of chromosome 5 for the purpose of finding whether these modifier genes exist, so as to discuss how modifier genes made SCA2 patients'clinical symptoms changed and to find the reasons for responsing to levodopa. The results showed that a missense mutation ATT/GTT existed on 207 amino acid of the HEXB gene exon 5. From the distributions of this mutation in the pedigree, A allele was carried in all of the patients. Furthermore, A/G mutation in 5'-UTR and two synonymy mutations (serine and lysine, respectively) were detected in ENC1 gene. According to results of 5'-UTR, almost all of the patients carried with A allele. Moreover, no mutation was observed in the other three genes. According to these results above, we considered that the missense mutation of the HEXB gene could be the modifier gene variation in this research which had more possibilities. On account of complex interactional relationships between genes, the mutations of modifier genes could have combined action with other locus. A synonymy mutation also could influence RNA post-transcriptional splicing, processing and the expression of protein, and then affect the phenotype. Perhaps other unidentified complex mechanism and genes exist. These mechanisms need to be clarified.Some researches proposed that the appearance of Parkinsonism in SCA2 patients was associated with frequency of CAA interuptions in CAG repeats of the ATXN2 gene. This association was analyzed by sequencing the CAG repeats. There was only one CAA interruption detected in the CAG repeats of ATXN2 gene in patients of our pedigree, which was discord with the reports that more CAA interruptions made the phenotype in favor of Parkinson Disease. It prompted other factors could have impacts on this family and made them sensitive to levodopa. Therefore, to investigate whether this CAG repeat of the ATXN2 gene is under positive selection, we collected six different Chinese ethnic groups, to evaluate the mechanism of the pathogenic gene and the relation with natural selection.The segments containing CAG repeats in the ATXN2 gene were sequenced for all chromosomes. Significantly negative Tajima's D values were observed in 5 ethnic groups except Hui by performing neutral tests, which suggested positive selection could act on this locus. Hui only showed negative values in these tests, but not significant. Haplotypes were constructed with genotyping data of 8 adjacent SNPs, which also demonstrated that core haplotype reached a high frequency and strong linkage disequilibrium reflected in a large region. These are signatures for positive selection in these 5 ethnic groups. We proposed that positive selection could act on this CAG repeat in these five ethnic groups and make (CAG)8CAA(CAG)4CAA(CAG)8 allele predominant in normal groups, which keeps heredity stability from expanding. Meanwhile, it was suggested that disease causing genes of other polyQ diseases probably have similar mechanism. Whether these related genes are under positive selection or not is worth further investigating.