Diversity among the tailed-bacteriophages that infect the Enterobacteriaceae

Abstract

Complete genome sequences have been determined for 73 tailed-phages that infect members of the bacterial Enterobacteriaceae family. Biological criteria such as genome size, gene organization and gene orientation were used to place these phages into categories. There are 13 such categories, some of which are themselves extremely diverse. The relationships between and within these categories are discussed with an emphasis on the head assembly genes. Although some of them are clearly homologues, suggesting a very ancient origin, there is little evidence for exchange of individual head genes between these phage categories. More recent horizontal exchange of phage tail fiber and early proteins between the categories occurs, but is probably not extremely rapid.

Fig. 1. Genome maps of six tailed-phages that infect enterobacteria

Genomes of the six phages named on the left are shown with gene functions above (one function, like “Heads”, can indicate a cluster of genes that are involved in that function); functions in gray text are poorly documented. The maps are oriented with DNA packaging genes at the left end if the genomes are circularly permuted, and asterisks (*) indicate the virion chromosomes that have specific termini. Transcriptional direction (not RNA polymerase start or stop locations) is indicated below by gray arrows (known or putative early functions) and black arrows (known or putative late functions); NAEnz, nucleic acid or nucleotide metabolism genes; RNPase, RNA polymerase genes; att site, site of integration of prophage into host chromosome; pac site, packaging machinery recognition site where it is known. The accession Nos. for these genome sequences and their annotation are as follows: T1, AY216660; ϕEcoM-GJ1, EF460875; T7, NC_001604; P2, AF063097; 933W, AF125520; Mu, AF083977.

Fig. 2. The tailed-phages of the Enterobacteriaceae with completely sequenced genomes

The seventy-three Enterobacteriaceae phages with completely sequenced genomes as of March 2008 are listed; all are known to be fully functional infectious phages except ϕKO2 which forms virions upon induction, but either they are defective or a susceptible host has not been found [8]. Thin lines separate phage types and subtypes as defined in the text; thick lines separate the lytic and temperate phages; the gray box in the temperate section highlights the “lambdoid” phages with subtypes on different lines, and in the lytic section highlights the phages with T1-like head genes (see text). Phages are listed below the bacterial genus they infect with slashes (/) separating their names (Shigella and Escherichia phages are indicated in the same column since these “genera” are not cleanly separated in evolutionary trees [50]); *, sequence deposited in GenBank but not yet published; †, S. Casjens, R. Hendrix and G. Hatfull, unpublished; ^, R. Hendrix and G. Hatfull, personal communication; @, G. Plunkett III, personal communication; # A. Kropinski, R. Villafane and S. Casjens, unpublished; (E), infects Erwinia amylovora; (K), isolated from Klebsiella oxytoca.

Fig. 3. Head gene maps of the seven lambdoid phage “head types”

Rectangles denote genes. Genes with similar known function or putative function as predicted from protein amino acid similarity are shown in the same color (which does not indicate sequence similarity); white genes have no known or predicted function. The putative protease/coat protein of the Gifsy-2 and BP-4795 genomes is predicted by sequence to contain these two functions, but this has not yet been proven in either case. At the right, the prototypical member of each group is indicated; other members with similar head genes are listed in Fig. 2. Black diamonds below a gene indicates that it is not present in all functional phage members of the subtype.

Fig. 4. Diversity of the phage λ coat protein

A neighbor-joining tree constructed by CLUSTAL X [61] is shown with bootstrap values (out of 1000) on each branch; horizontal branch lengths are proportional to extent of differences. At the right, λ-like coat protein categories I, II and III are indicated. The five fully functional phage coat proteins are indicated by “host species.phage name.gene name” and each prophage coat protein is denoted by its “host species.host strain name.gene number portion of the GenBank locus_tag”, except in the case of the E. coli strain ELD933 prophages [49] where the locus_tag number is replaced by the prophage name 933X or 9933O. Host bacterial species are abbreviated as follows: Ec, E. coli; Se, Salmonella enterica; Sf, Shigella flexneri; Sb, Shigella boydii; Ss, Shigella sonnei; Vh, Vibrio harveyi; Va, Vibrio alginolyticus; Bc, Burkholderia cepacia. The asterisks (*) note that eight additional sequences that are identical to the Gifsy-1 coat protein have been reported in various S. enterica sequences, as have three additional sequences identical to the E. coli prophage 933O coat protein and these identical sequences are not shown in the figure; “trunc” indicates that the gene is truncated in the GenBank sequence.