IS4 family goes genomic

Abstract

Background: Insertion sequences (ISs) are small, mobile DNA entities able to expand in prokaryotic genomes and trigger important rearrangements. To understand their role in evolution, accurate IS taxonomy is essential. The IS4 family is composed of approximately 70 elements and, like some other families, displays extremely elevated levels of internal divergence impeding its classification. The increasing availability of complete genome sequences provides a valuable source for the discovery of additional IS4 elements. In this study, this genomic database was used to update the structural and functional definition of the IS4 family.

Results: A total of 227 IS4-related sequences were collected among more than 500 sequenced bacterial and archaeal genomes, representing more than a three fold increase of the initial inventory. A clear division into seven coherent subgroups was discovered as well as three emerging families, which displayed distinct structural and functional properties. The IS4 family was sporadically present in 17 % of analyzed genomes, with most of them displaying single or a small number of IS4 elements. Significant expansions were detected only in some pathogens as well as among certain extremophiles, suggesting the probable involvement of some elements in bacterial and archaeal adaptation and/or evolution. Finally, it should be noted that some IS4 subgroups and two emerging families occurred preferentially in specific phyla or exclusively inside a specific genus.

Conclusion: The present taxonomic update of IS4 and emerging families will facilitate the classification of future elements as they arise from ongoing genome sequencing. Their narrow genomic impact and the existence of both IS-poor and IS-rich thriving prokaryotes suggested that these families, and probably ISs in general, are occasionally used as a tool for genome flexibility and evolution, rather than just representing self sustaining DNA entities.

Figure 1

Transposases vs. IRs and DRs. Comparative overview of IS sequence features. A. Dendrogram representing an alignment of 227 transposases. The first common node of each family was pointed out on the left tree side. Subgroups of family IS4 are mentioned on the right tree side. Individual IS names were removed for clarity. B. Consensus sequences (5' to 3') of left (upper line) and right (lower line) DNA extremities of distinct subgroups/families. '+' symbols are used when the highest conservation level is shared by more than one residue. Percentages of nucleotide conservation at all positions are indicated by black bars. Decimal nucleotide numbering is marked by dotted vertical lines. Full alignments that generated each consensus can be found in Additional files 1 to 10. Note that in the case of family IS701, the exposed consensus does not represent the entire family. For further details, see Additional file 8. IRL, left TIR; IRR, right TIR. C. Target duplication length range in bps.

Figure 2

DDE and YREK motifs of IS4 subgroups and emerging families. Comparative overview of conserved transposase regions among IS4 subgroups and emerging families. Each line represents a part of the amino acid consensus obtained from multiple alignments of members belonging to the displayed IS groups. Numbers in brackets correspond to the mean amino acid spacer (accompanied by standard deviation) between the two aspartate residues or the aspartate and glutamate residues of the DDE motifs among transposases form a given group. The conserved transposase regions N2, N3 and C1 are mentioned on top of alignment. Symbols and colors are used as depicted in keys.

Figure 3

Target sites : duplications and conservations. Consensus sequences of in silico reconstructed and aligned target sites (5' to 3') typical for each subgroup or family. Only ISs found at least at five different genomic locations and flanked by DRs were considered, as well as elements described previously in literature. Each box represents one base position. Target sites are divided in three parts. The central sequence, which is duplicated upon insertion, is flanked by one upstream (left) and one downstream (right) target arm. Ten bps of each arm are shown. The number of target sites considered for each element is indicated. If more than one element displayed similar insertion specificity, their insertion sites were combined into a single line and their names listed above it. Gray boxes inside a duplication consensus indicate that DRs of variable length can be found for the given element. W, A or T; S, G or C; R, A or G; Y, T or C. * 'IS231' stands for following elements: IS231A, C, F, M, S, T, U, Y, ISBce4, 5, 6, 11, 12 and ISBth5. ** For the 6000 sites, see references [75–77].

Figure 4

Prokaryote distribution of IS4 subgroups and emerging families. A. Three dimensional histogram of the number of distinct elements belonging to IS4 subgroups and emerging families, as they can be found among major prokaryotic clades. Each IS subgroup/family is represented by a different color. Iso-forms (which we defined as elements that show a divergence of less than 2% in the amino acid sequence of their potential proteins) were not included. The line 'Others' stands for Aquificae, Chlamydiae/Verrucomicrobia, Chloroflexi, Chrysiogenetes, Dictyoglomi, Fusobacteria, Nitrospirae, Spirochetes, Thermodesulfobacteria and Thermotogae. For interpretation, see main text. B. To avoid misinterpretation triggered by unequal sequencing efforts between different microbial groups, the number of genome projects, as of 1^st April 2007, is depicted by histogram.

Figure 5

Genomic IS copy number. Genomic copy numbers of four families as they were found on chromosomes and plasmids. Families IS4, IS1634, ISH3 and IS701 are represented in red, blue, green and purple, respectively. The number of copies located on chromosomes is represented by cylinders while those located on plasmids is shown as cones. The height of each cylinder/cone indicates the number of chromosome/plasmid displaying the corresponding number of IS family members. No distinction was made when different elements of the same IS family occurred in the same genome. The histograms include intact elements, disrupted full length copies and large partial fragments displaying more than 95 % DNA sequence identity. Genomes without any copy of the aforementioned families were not included.