Drosophila melanogaster'de parkin geniyle etkileşen ve lokomotor davranışı etkileyen genlerin genomik ilişkilendirme modeliyle saptanması
Özet
The aim of this thesis was to locate by genome wide association study (GWAS) the genes which interacts with D. melanogaster parkin, a gene that affects locomotor behavior. The methodology of the thesis was based on release of ‘hidden genetic variation’, which makes detection of the epistatic genes available, by way of phenotypic variation in the expression of a focal gene across different genomes. The gene used this context, parkin, is located in D.melonagaster and is the ortholog of human PARK2 gene. Phenotypic scores were obtained by measuring startle response which is known as an indicative component of locomotor behavior. It is well known that mutations in parkin affect the phenotypic variation in locomotion. In the experiments, parkin mutant and its control were crossed with the 115 lines from the Drosophila Genetic Reference Panel (DGRP), a genomic mapping population with a defined genetic architecture by way of about 5 million SNPs distributed genomewide. With DGRP lines, any biological trait can be easily mapped to its putative genes by using the phenotypic scores obtained. In this study, phenotypes of F1 generations, which were obtained by crossing Drosophila parkin homozygotes (parkin/parkin; w1118/w1118) with the control stock of this gene (parkin+/parkin+; w1118/w1118), were determined separately for 3 different age groups (0, 15 and 30 days). First, it was observed that the mean phenotypic difference between the parkin mutant and its control varied greatly across the genomes. After related ANOVAs, these changes in the magnitude of and the direction of the difference were statistically highly significant confirming the presence of a genotype x genotype interaction. Secondly, a GWAS model based on this genotype x genotype interaction was conducted. After that GWAS, for each age group, genes hit by the SNPs with minor allele frequency of 0.05 were picked out. A second elimination was done by focusing only on the genes which had significant (i.e. p= 10-5) association for at least one of the phenotypic ascertainments (i.e. female effect; male effect; magnitude of the female-male difference). On these criteria, for the three different age groups (0, 15 and 30 days of age) 91, 75 and 64 genetic variants were determined respectively. Those genes were classified with regard to their chromosome locations, relevant biological processes and human orthologs. It is suggested that these genes could be realistic candidates for many biological properties involved in locomotion and various disorders caused by locomotor defects. In conclusion, it is thought that important hints about the interactive genetic architecture of locomotor behavior have been obtained from the results of the studies comprising the thesis. In this repertoire of potentially interacting genes, while there are many new candidate genes to be confirmed for locomotion in general, there seem to be also genes that could enrich the information about the disorders related to locomotion, such as Parkinsonism.
