High-density transcriptional initiation signals underline genomic islands in bacteria

• Published in: PloS one Vol. 7; no. 3; p. e33759
• Main Authors: Huang, Qianli, Cheng, Xuanjin, Cheung, Man Kit, Kiselev, Sergey S, Ozoline, Olga N, Kwan, Hoi Shan
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 20-03-2012; Public Library of Science (PLoS)
• Subjects: Acids; Antibiotics; Bacteria; Bacteria - genetics; Bias; Biofilms; Biology; Biophysics; Codons; Computational Biology; E coli; Escherichia coli; Gene expression; Gene regulation; Genes; Genetic aspects; Genome, Bacterial; Genomes; Genomic Islands; Genomics; Life sciences; Localization; Metabolism; Multigene Family; Outbreaks; Pathogenicity; Pathogens; RNA polymerase; Salmonella; Salmonella Typhimurium; Subcellular Fractions; Transcription (Genetics); Transcription initiation; Transcription Initiation Site; Transcription, Genetic; China
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0033759

Genomic islands (GIs), frequently associated with the pathogenicity of bacteria and having a substantial influence on bacterial evolution, are groups of "alien" elements which probably undergo special temporal-spatial regulation in the host genome. Are there particular hallmark transcriptional signals for these "exotic" regions? We here explore the potential transcriptional signals that underline the GIs beyond the conventional views on basic sequence composition, such as codon usage and GC property bias. It showed that there is a significant enrichment of the transcription start positions (TSPs) in the GI regions compared to the whole genome of Salmonella enterica and Escherichia coli. There was up to a four-fold increase for the 70% GIs, implying high-density TSPs profile can potentially differentiate the GI regions. Based on this feature, we developed a new sliding window method GIST, Genomic-island Identification by Signals of Transcription, to identify these regions. Subsequently, we compared the known GI-associated features of the GIs detected by GIST and by the existing method Islandviewer to those of the whole genome. Our method demonstrates high sensitivity in detecting GIs harboring genes with biased GI-like function, preferred subcellular localization, skewed GC property, shorter gene length and biased "non-optimal" codon usage. The special transcriptional signals discovered here may contribute to the coordinate expression regulation of foreign genes. Finally, by using GIST, we detected many interesting GIs in the 2011 German E. coli O104:H4 outbreak strain TY-2482, including the microcin H47 system and gene cluster ycgXEFZ-ymgABC that activates the production of biofilm matrix. The aforesaid findings highlight the power of GIST to predict GIs with distinct intrinsic features to the genome. The heterogeneity of cumulative TSPs profiles may not only be a better identity for "alien" regions, but also provide hints to the special evolutionary course and transcriptional regulation of GI regions.