Association between Dna Methylation and Pathogenicity in Uropathogenic Escherichia Coli

Abstract Epigenetic study is important in eukaryote, because it regulates transcription. In bacteria, on the other hand, DNA modifications do not seem to be involved in transcription in general. However, publications reporting involvement of DNA methylation to transcription in bacteria are increasing year by year. In addition, reports of association between DNA modification and pathogenicity are also increasing, while the mechanisms are not uncovered in the most cases. Therefore, it is important to investigate how they are involved in each other. We have found that distribution of some genes encoding DNA methyltransferases (DNMTase), which conflict with a modification-specific restriction enzyme, McrBC, is significantly higher in urinary tract infections associated E. coli (UPEC) than in the fecal isolates. This result suggests that the DNA methylation associates with pathogenicity in UPEC. Here we propose three projects to investigate mechanisms of the association in UPEC as a model. Determinant genes encoding the DNA methyltransferase, which show higher distribution in UPEC, have been identified. Then, we will construct mutants of them in UPEC strains, and carry out transcriptome analysis to see whether the DNA modification is involved in transcription or not, and animal experiments to see whether the DNA modification affects virulence or not. If expression of a virulence gene(s) is changed in the mutants, we could reveal that the DNA modification directly involved in virulence, and then, the result of animal experiments would be also changed by the mutation defective in the methylation. We will screen another conflict between DNMTase and modification-specific restriction enzyme in urinary tract infectious E. coli. Such conflict would affect to horizontal gene transfer of genomic islands and evolution to be a pathogenic strain. Therefore, we could identify the restriction-modification enzymes to be involved in directional evolution arising pathogenic strains. We are going to develop two of bioinfomatic methods for whole genome epidemiological analysis, and prediction of virulence factor by self organizing map. These developed methods themselves would have strong potential to help the clinical examination and infectious disease study. Furthermore, genomes of clinically isolated E. coli would be analyzed using the developed methods to investigate the association between the conflict and pathogenicity. Consequently, a novel therapy targeting the restriction-modification enzymes would be developed to repress the expression of virulence factor or avoid horizontal gene transfer to become a pathogenic strain.