• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Mar 22, 2005; 102(12): 4419–4424.
Published online Mar 8, 2005. doi:  10.1073/pnas.0406620102
PMCID: PMC555480

Integrons in Xanthomonas: A source of species genome diversity


Integrons are best known for assembling antibiotic resistance genes in clinical bacteria. They capture genes by using integrase-mediated site-specific recombination of mobile gene cassettes. Integrons also occur in the chromosomes of many bacteria, notably β- and γ-Proteobacteria. In a survey of Xanthomonas, integrons were found in all 32 strains representing 12 pathovars of two species. Their chromosomal location was downstream from the acid dehydratase gene, ilvD, suggesting that an integron was present at this site in the ancestral xanthomonad. There was considerable sequence and structural diversity among the extant integrons. The majority of integrase genes were predicted to be inactivated by frameshifts, stop codons, or large deletions, suggesting that the associated gene cassettes can no longer be mobilized. In support, groups of strains with the same deletions or stop codons/frameshifts in their integrase gene usually contained identical arrays of gene cassettes. In general, strains within individual pathovars had identical cassettes, and these exhibited no similarity to cassettes detected in other pathovars. The variety and characteristics of contemporary gene cassettes suggests that the ancestral integron had access to a diverse pool of these mobile elements, and that their genes originated outside the Xanthomonas genome. Subsequent inactivation of the integrase gene in particular lineages has largely fixed the gene cassette arrays in particular pathovars during their differentiation and specialization into ecological niches. The acquisition of diverse gene cassettes by different lineages within Xanthomonas has contributed to the species-genome diversity of the genus. The role of gene cassettes in survival on plant surfaces is currently unknown.

Keywords: genome evolution, lateral gene transfer, mobile DNA, pathovar

Genome sequencing of an individual eukaryote recovers most genes in a species. The same cannot be said of prokaryotes, where strains of bacterial species can differ by ≤20% in their gene content. The characterization of such intraspecific variation is central to assembling the species genome, the individual elements of which do not reside in any one cell but are dispersed among clonal lineages within species (1, 2). The diversity inherent in bacterial species genomes emerges through the interactions of various genetic elements, environmental selection, and genetic isolation through niche specialization.

Intraspecific variation can be characterized by using restriction mapping, subtractive hybridization, or a comparison of genome sequences (1, 2). These methods can identify large DNA insertions, which are often acquired by lateral gene transfer (3). However, the acquisition of one, or a few, key genes can lead to rapid selection and fixation in particular niches. The acquisition of antibiotic resistance in clinical bacteria is a clear example of how bacterial populations can employ mosaic structures of mobile genetic elements to rapidly respond to environmental change (see ref. 4).

Plasmid- or transposon-borne integrons are a key player in this process, being able to acquire, rearrange, and express genes, in this case, those conferring antibiotic resistance (57). Whether integrons are located on a plasmid or chromosome, their structure and function are similar. They contain a gene for a DNA integrase (intI), which catalyzes the site-specific recombination of gene cassettes at the integron-associated recombination site (attI). Acquisition of multiple cassettes results in a contiguous array of genes. Each gene cassette typically consists of a single ORF and a further recombination site known as a 59-base element (be). Transcription of integrated gene cassettes is driven by a promoter, Pc (refs. 5, 8, and 9 and Fig. 1).

Fig. 1.
Schematic illustration of the integron in Xanthomonas campestris pv. campestris ATCC33913 (15), showing binding sites for some of the primers used in this study. Only a portion of the cassette array in ATCC33913 is shown. Integrase activity can also catalyze ...

Integrons are not restricted to clinical settings, because they can be amplified from soil DNA and are found in many genome sequences (1012). Integrons and gene cassette arrays have been found in the chromosomes of Vibrio, Pseudomonas, Xanthomonas, Microbulbifer, Treponema, Geobacter, Dechloromonas, Methylobacillus, and Shewanella species (1316). Generally, identification of chromosomal integrons is based on conservation of structure and comparative sequence analysis rather than demonstration of function. However, the presence of an intI homolog, a plausible attI recombination site, and a gene cassette array bounded by 59-be sites (17), argues that they have properties similar to those of the extensively characterized class 1 and 3 integrons (6, 18, 19).

Could the activity of chromosomal integrons contribute to genetic diversity by assembling different cassette arrays as lineages diverge? Could the capacity of integrons to assemble functionally complementary gene combinations result in ecological differentiation and, ultimately, speciation? We can address these questions by examining cases where a bacterial lineage has had a stable association with a single integron lineage throughout its evolutionary and ecological radiation. The genus Xanthomonas is such a model system. Xanthomonads are closely associated with plants, either as phytopathogens or as nonpathogenic inhabitants of the phylloplane (20, 21). The genus is remarkably uniform in all but one phenotypic trait; the ability to cause disease on particular plants (22, 23). The economic importance of this property led to a special-purpose classification, the pathovar (24). A pathovar is defined as “a strain or set of strains with the same or similar characteristics, differentiated at the infrasubspecific level from other strains of the same species or subspecies on the basis of distinctive pathogenicity to one or more plant hosts” (25). Despite the wide acceptance of this concept, practical issues mean that comprehensive testing of host range limits has not been performed for most isolates, and it is difficult to integrate pathovar classifications with classifications of xanthomonads that are isolated as benign commensals (26). Nevertheless, despite debate about the nomenclatural status of pathovars (27, 28), the broad concept that strains isolated from a particular diseased plant species constitute a genetically related population adapted to causing disease on that host is supported by genetic and pathogenicity data (26, 2931). Consequently, Xanthomonas strains isolated from the same host species are likely to be members of an ecologically defined population, whereas strains isolated from other host species are likely to be ecologically distinct populations.

The genomes of single strains of two Xanthomonas species (15) each contain an intI homolog at the same location. We postulated that these represent an integron lineage that is an ancestral feature of the genus and used this to observe the consequences of integron activity during an evolutionary radiation. In this study, we tested a collection of Xanthomonas pathovars for integrons and partially mapped their gene cassette arrays. Here, we show that the integron locus is the site of fixed genetic differences in strains that were recovered from distinct habitats (i.e., plant hosts).

Materials and Methods

Source and Characterization of Isolates. Xanthomonas strains were obtained from the Australian National Collection of Plant Pathogenic Bacteria (DAR prefix to strain numbers). DNA was isolated by using an SDS/phenol method (32). The strains and their characteristics are listed in Table 1. Strains were collected over a 50-year period and stored in a single repository. To confirm identity as xanthomonads, all cultures were subjected to amplified ribosomal DNA restriction analysis (ARDRA) with RsaI and HinfI by using standard protocols (33). Restriction patterns were compared with those predicted from 16S ribosomal RNA-encoding DNA (rDNA) sequences of xanthomonads available through the National Center for Biotechnology Information database. To confirm the independence of each strain, DNA fingerprinting with repetitive-element PCR (rep-PCR) was used (34). DNA fingerprints were generated from replicate DNA extractions of each strain by using both the BOXA1R primer and the ERIC1R and ERIC2 primer sets (35) and following published protocols (36).

Table 1.
List of Xanthomonas isolates used in this study

Sequencing of Integron Components and Flanking DNA. Genome sequences of Xanthomonas campestris pv. campestris and Xanthomonas axonopodis pv. citri (15) revealed integron-related sequences that were conserved in the two strains. We used these data to develop PCR protocols to screen our strains for the presence of attI, intI, and gene cassette arrays and to test whether integrons were located adjacent to ilvD. The presence of a core integron was determined by using primers targeting conserved motifs in attI (MRG18) and intI (AJH72). This PCR is predicted to generate a 300-bp product. Integron location was confirmed by using primers targeting intI (AJH73) or attI (MRG18) and ilvD (AJH95). This PCR is predicted to give a product of 1,700 bp. Cassette arrays were detected by using primers targeting 59-be (AJH60) and intI (AJH72). This PCR is predicted to give multiple products of various sizes. Details of primer sequences, binding locations, and PCR conditions are given in Fig. 1 and Table 2, which is published as supporting information on the PNAS web site.

Amplified fragments were purified by using PCR clean-up columns (Promega), and DNA was sequenced directly by using the amplification primers or cloned into pGEM-T (Promega) according to the manufacturer's instructions. Plasmids were sequenced with primers pGEMF and pGEMR and by using internal primers where necessary for large clones. DNA sequencing was performed at the Macquarie University Sequencing Facility by using an ABI Prism 377 (PE Biosystems, Foster City, CA). Contigs spanning the regions ilvD to attI and intI to the most distal gene cassette recovered were edited and connected by using dnasis software (Hitachi, Brisbane, CA).

DNA Analysis. DNA sequences were used to search databases by using the National Center for Biotechnology Information blast web site (www.ncbi.nlm.nih.gov:80/blast). blastn and blastx functions were used to identify key features, including ilvD, intI, and the presence of transposons. Identification of 59-be was based on sequence similarity detected in blastn searches or by inspection of intergenic regions. Putative integron-associated recombination sites (attI) were identified by using alignments between integron sequences of different pathovars. Hypothetical genes in cassettes were identified on the basis of blastn searches in a small number of cases, and in other cases, according to the following criteria: a reading frame in the opposite orientation to intI; a start codon within 30 bp of attI or the 59-be; a stop codon in or adjacent to the next 59-be; and being the largest ORF bounded by two 59-be. Sequences were annotated and submitted to the GenBank database through the National Center for Biotechnology Information by using sequin Version 5 software (www.ncbi.nlm.nih.gov/sequin).


Characterization of Isolates. All strains used in this study conformed to the expected ARDRA patterns (33) generated by digestion of amplified 16S rDNA from Xanthomonas species (data not shown). To establish the independence of strains in the collection and absence of cross-contamination, each strain was subjected to rep-PCR by using BOX and ERIC primers (36). This procedure generated a complex fingerprint upon electrophoresis, with DNA fragments ranging from 100 to 2,500 bp. No two strains generated identical fingerprints when this method was used (Fig. 2 Top and Middle). Although detailed comparisons were not made, it was noted that within-pathovar similarity was greater than between-pathovar similarity. It was concluded that each strain represented an independent isolate of the nominal pathovar. The fingerprints suggested that some strains had been misidentified or mislabeled during curation or handling, because [DAR61714 oryzae and DAR73878 vitians (= oryzae?)] and [DAR34895 vesicatoria, DAR73877 vesicatoria, and DAR61713 oryzae (= vesicatoria?)] each formed coherent groups upon rep-PCR, gene cassette PCR, and integron-mapping (Figs. (Figs.22 and and3).3). This suggestion was borne out by DNA sequencing of the rDNA intergenic spacer region (Supporting Text, which is published as supporting information on the PNAS web site), because DAR73878 exhibited 100% similarity to DAR61714 over the sequenced region (539 bp) but had lower matches to the two other vitians isolates, which were identical. DAR61713 had 99.8% and 100% homology, respectively, to DAR34895 and DAR73877 and 100% homology to vesicatoria isolates available in GenBank, confirming the PCR fingerprinting results.

Fig. 2.
Molecular analysis of Xanthomonas strains. Total genomic DNA of different Xanthomonas pathovars was amplified by using rep-PCR (30) or integron gene cassette PCR. (Top) BOX-PCR fingerprinting. (Middle) ERIC-PCR fingerprinting. (Bottom) Amplification of ...
Fig. 3.
Maps of X. campestris and X. axonopodis integrons characterized in this paper, aligned below the map for ATCC33913 (see Fig. 1 and ref. 15). From left to right, each map consists of pathovar and isolate numbers; the acid dehydratase gene (ilvD), the integrase ...

Screening for Integron Content. A functional integron has two essential components, intI and an associated recombination site, attI. To confirm the presence of an attI site adjacent to intI, all strains were screened with PCR by using primers designed to the proximal part of intI (AJH72) and to the attI region (MRG18) (Fig. 1 and Table 2). Every strain generated a single product of the expected size (≈300 bp), confirmed by sequencing to be related to the intI/attI components of X. axonopodis pv. citri str. 306 and X. campestris pv. campestris ATCC33913. These data suggest that an integron was present in all 32 strains in the collection (Table 1).

Determining Integron Location. To test whether this integron family had been present throughout the radiation of Xanthomonas, PCR was used to confirm its presence adjacent to the dihydroxyacid dehydratase gene, ilvD (ref. 15 and Fig. 1). All strains gave a product, but the expected size (1,700 bp) was seen in only 9 of 32 strains. For 23 selected strains, this PCR product was DNA-sequenced and shown to include homologs of ilvD and intI. The length discrepancies were caused by indels in the intergenic region between ilvD and intI or within intI (see below). These data confirmed that an integron had been present at ilvD throughout the radiation of the strains and showed that independent loss of function mutations had occurred in different sublineages.

Amplification of Cassette Arrays. To detect cassette arrays and test for heterogeneity in those arrays, a PCR anchored in intI (AJH72) was used in conjunction with a primer targeting 59-be sites (AJH60) (Fig. 1). Because AJH60 has multiple binding sites in a cassette array, this assay resulted in multiple PCR products from each strain, with stepwise increases in size corresponding to amplification of one, two, or more cassettes adjacent to attI (Fig. 2 Bottom). All 32 strains generated complex banding patterns, corresponding to 12 distinct cassette arrays. Groups of strains that shared a cassette array fingerprint corresponded precisely to groups identified by using rep-PCR. Cassette array PCRs were identical within each of the pathovars begoniae, oryzae, pruni, pelargoni, vitians, citri, and vesicatoria (Fig. 2 Bottom). DNA sequencing was performed on the largest cassette PCR product cloned from each of the 23 strains whose ilvD/attI had already been sequenced. This procedure revealed at least one and up to five putative cassettes in a typical integron-like structure. Strains with identical cassette PCR patterns shared the same gene cassettes in the same order over the sequenced region. Strains with nonidentical cassette PCR patterns had no cassettes in common. Consequently, gene cassette order and content tend to be conserved within, but not between, cassette arrays in Xanthomonas pathovars.

Assembly of Xanthomonas Integron Sequences. The DNA sequence of each integron was assembled from two PCR-derived sequences: the intI/attI fragment and the largest fragment recovered from the cassette PCR. Validity of the connected sequences was confirmed by restriction mapping of long-range PCR products (see Supporting Text). Scale maps of the sequenced regions of each strain are shown in Fig. 3. The salient features of these assemblies are as follows:

Integrase gene. Each strain contained an integron integrase homolog, but the gene was predicted to be nonfunctional in the majority of strains. Only four strains examined potentially encoded a functional integrase (pathovars campestris DAR30538, begoniae DAR69819 and DAR54703, and vesicatoria DAR26930). A total of seven strains representing pathovars translucens, orzyae, vitians, vesicatoria, and axonopodis contained frameshifts or stop codons in intI that are predicted to lead to truncated or nonfunctional proteins. The remaining strains had various large indels in intI and represented three pathovars of X. campestris (pruni, pelargoni, and vitians) and three pathovars of X. axonopodis (citri, citrumelo, and unknown). Deletion end points were identical in pathovars pelargoni and vitians and in three pathovars of X. axonopodis. In one strain (pv. vitians DAR30526), intI was disrupted by the insertion of tranposon-related sequences.

Integron-associated recombination site, attI. Each strain tested contained a potential attI site, identified by alignment of the region between intI and the first cassette in the associated array. All potential attI sites included a sequence consistent with an attI-like simple site (37). The attI sequences were moderately conserved within Xanthomonas but were highly divergent, compared with attI in other integron classes. Although we have not confirmed their activity in recombination assays, the sequences reported here have almost certainly descended from a functional attI site.

Cassette arrays. A total of 27 complete or partial gene cassettes were identified, adding to the ≈30 gene cassettes reported from Xanthomonas and available in public databases. In two instances, our cassettes were interrupted by the insertion of putative transposons (pathovars citri and translucens). Our sequence data are unlikely to represent the entire cassette array in any of these strains. Nevertheless, a number of significant observations can be made.

First, particular cassette arrays were directly correlated with groups defined on the basis of rep-PCR analysis. X. campestris pathovars campestris, begoniae, pruni, pelargoni, and vitians each had arrays composed of gene cassettes found in no other pathovar. X. axonopodis pathovars citri and citrumelo had an identical cassette array, whereas the remaining X. axonopodis pathovars (vesicatoria, axonopodis, and unknown) each had unique cassette arrays. Strains of X. campestris pv. campestris and X. axonopodis pv. citri mapped in this study had cassette arrays identical to the genome-sequenced strains of these pathovars (15).

Second, individual gene cassettes did not occur in multiple contexts. Either the entire integrase/attI/cassette array was identical across the sequenced region, or arrays had no cassettes in common. One subset of cassettes constituted a family sharing >84% DNA sequence identity. These cassettes were homologous to pigH, a gene of unknown function associated with the xanthomonadin biosynthesis gene cluster elsewhere in the Xanthomonas genome. Members of this group were found in three different contexts, which could reflect ancient gene transfer and subsequent divergence. None of the other cassette genes showed relationship to genes of known function, based on blastn or blastx searches.

Third, the G+C content of the ilvD/intI/attI region of the strain collection ranged from 61.1% to 64.9% and is similar to the G + C content of the Xanthomonas genome, whereas the G+C content of the cassette arrays was much lower, ranging from 47.7% to 57.8%. This observation suggests that gene cassettes were acquired by lateral gene transfer of foreign DNA, a conclusion also drawn from an analysis of the X. campestris ATCC33913 genome sequence (38).


Integrons and gene cassettes are a frequent component of bacterial genomes, being readily recovered from environmental samples and having been observed in ≈5% of genome sequencing projects (1115, 17, 39). The best-characterized integrons (class 1) are borne on mobile elements. Their activity facilitated a community level response to intensive antibiotic use, resulting in the emergence of integron-encoded, multiple antibiotic resistance in disparate bacterial species. Available data suggest that the general properties of class 1 integrons are shared by other integrons (12, 17, 19, 40, 41). More recently discovered integrons are apparently nonmobile components of bacterial chromosomes, raising the question of whether they could also facilitate within-species response to environmental change. One way to investigate the long-term evolutionary significance of chromosomal integrons is to examine their associated cassette arrays in bacterial populations known to share common ancestry.

Our dataset was based on 32 Xanthomonas strains encompassing two species and 12 pathovars that were chosen to reflect a diversity of collection dates, locations, and plant hosts. Integrons were detected in all strains, and partial integron maps consisting of flanking sequence, the integrase gene, and a partial cassette array were generated for 23 of these strains. Phylogenetic analyses of intI/attI and ilvD strongly support the Xanthomonas integrons as a monophyletic group (Supporting Text and Figs. 4 and 5, which are published as supporting information on the PNAS web site). Significantly, the location of the integron was the same in all cases, downstream from the gene for dihydroxyacid dehydratase, ilvD (15). Other organisms for which genome sequences are available do not carry integrons at this locus. The simplest interpretation of these data is that an integron was acquired at this locus by the common ancestor of Xanthomonas and has been maintained throughout the radiation of the genus. Our primary question was then: Has integron activity contributed to genetic diversity? Consequently, we examined the proximal portion of the gene cassette array from all pathovars. There was no evidence of integrons generating diversity within pathovars but strong evidence for diversity between pathovars.

With some minor exceptions, individual pathovars had distinct proximal gene cassette arrays, and every cassette identified was found only in one pathovar. It is uncertain how many cassettes might constitute the arrays in each pathovar, because this determination will have to await complete sequencing of each array. However, the genome sequenced arrays in X. campestris pv. campestris ATCC33913 and X. axonopodis pv. citri str. 306 have 23 and 4 cassettes, respectively (15). Two different chromosomal integrons in Pseudomonas stutzeri have 10 and 14 cassettes (ref. 12; M.R.G. and A.J.H., unpublished data), whereas Pseudomonas alcaligenes ATCC55044 has 33 cassettes in its array (14). Again, none of the known cassettes in these Pseudomonas species are found in more than one array. In the genus Vibrio, integrons can contain hundreds of unique gene cassettes (17, 42). Hence, it seems likely that the arrays in Xanthomonas pathovars will also consist of multiple cassettes, and that many of these cassettes will be found in only one pathovar. It also seems likely that the ancestral Xanthomonas integron, and indeed all extant integrons, had access to a very large pool of gene cassettes, and that considerable genomic diversity can be generated through integron activity. This conclusion also has been drawn from analysis of gene cassettes recovered directly from environmental DNA (16) and from spatial analysis of the abundance of gene cassettes in natural systems (43).

The Xanthomonas cassette arrays are phenotypically cryptic, because their products have not yet been associated with properties relevant to epiphytic survival or pathogenicity. However, the following points may be argued: the strain collection shares a recent common ancestor (being monophyletic by 16S rDNA and ilvD); an integron was present in this ancestor (congruence of intI, ilvD, and 16S rDNA phylogenies); the lineage has been subjected to evolutionary pressures that resulted in multiple, distinct genomic lineages (well supported subgroups are seen by various genetic typing methods); and over the evolutionary radiation of Xanthomonas, integron activity has assembled diverse gene cassettes in the proximal region (12 different cassette arrays detected). The tight correspondence among various genetic typing methods (rep-PCR, ilvD, and intI phylogenies), cassette arrays, and pathovar status is strongly suggestive of linkage between the genetic events that result in fixation of cassette arrays and the events that lead to host-specific pathogenesis.

The recombination mechanisms by which integrons acquire diverse genes also can result in the loss and rearrangement of those genes, and, consequently, the cassette arrays in active integrons might not be stable over long periods. One way of stabilizing arrays in particular lineages is to inactivate the integron machinery by loss of integrase activity or mutations in the recombination sites attI and/or 59-be. Of the 25 integrase gene fragments examined in this study, only 4 were intact. A number of pathovars contained point mutations in the integrase gene that resulted in frameshifts or stop codons that would cause premature truncation or nonsense translation, thus affecting the critical tyrosine residue that is essential for site-specific recombinase activity (44). Strains with identical frameshift or stop mutations also had identical cassette arrays as determined by mapping or amplification of gene cassettes, suggesting that fixation of cassette arrays has taken place. The conservation of cassette arrays in pathovars begoniae and campestris raises the possibility that these integrons also are nonfunctional for recombination activity, despite having an apparently intact intI gene.

In a further 14 strains, large deletions of intI were noted, all of which would inactivate the gene. Some pairs of pathovars (pelargoni/vitians and citri/DAR69855, pv. unknown) shared identical deletion end points, suggesting that the deletion events predated differentiation into the current pathovars. However, each of these pathovars had a distinct cassette array. The most likely explanation for this circumstance is homologous recombination between arrays, and indeed, sequence alignment of the attI regions of pathovars pelargoni and vitians suggests that one crossover junction may be located within this region (data not shown).

Comparison between integrons in different pathovars of Xanthomonas allows us to reconstruct some of the genomic evolution in this group and to suggest how rearrangements of gene cassettes have accompanied niche specialization of pathovars. In this model, an ancestral integron containing at least a functional integrase gene and attendant attI site was acquired adjacent to ilvD. Subsequently, integron activity accessed the environmental pool of gene cassettes to generate large numbers of diverse lineages, each containing a multicassette array. Inactivation of intI then fixed particular cassette arrays concomitant with their specialization or selection as pathovars of different plant hosts. Subsequent rearrangements of arrays may have been driven by in trans integrase activity or homologous recombination, again accompanied by niche specialization. It has already been suggested that the phylloplane is a hot spot for lateral gene transfer driven by conjugation and transduction, and that, consequently, it is an important niche for generating microbial genetic diversity (see ref. 45). Integrons may be an important mechanism for incorporating laterally transferred genes into the chromosomes of phylloplane organisms.

Recent data suggest that the genus Xanthomonas is composed of two types of strains: (i) those commonly isolated as plant pathogens (pathovars), and (ii) a genetically diverse, more rarely isolated group of apparently nonpathogenic strains. It further appears that classification based solely on phytopathogenicity does not accurately reflect the phylogeny or inherent diversity of the genus (26). The pathovars that we currently recognize could be the result of agricultural monocultures favoring particular strains that have the ability to invade and proliferate in the vast biomass of agronomic crops. Consequently, the appearance of pathovars may be the result of the selection pressures brought about by human activity, in a manner analogous to the appearance of integron-driven antibiotic resistance in clinical settings. Whether integrons and their associated gene cassettes are responsible for adaptation as phytopathogens of particular plant hosts or are fixed in lineages by simple linkage will not be known until the diversity and functions of gene cassettes in Xanthomonas are more fully understood.

Supplementary Material

Supporting Information:


We thank J. Brown, D. Wood, and P. Worden for technical assistance. This work was supported by the Australian Research Council and the Macquarie University Research and Development Scheme.


Author contributions: M.R.G., H.W.S., and A.J.H. designed research; M.R.G., M.P.H., and A.J.H. performed research; M.R.G. and A.J.H. contributed new reagents/analytic tools; M.R.G., M.P.H., H.W.S., and A.J.H. analyzed data; and M.R.G., H.W.S., and A.J.H. wrote the paper.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviations: be, base element; rDNA, ribosomal RNA-encoding DNA; rep-PCR, repetitive-element PCR.

Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. AY928773–AY928797).

Note Added in Proof. The genome sequence of a third species, Xanthomonas oryzae pv. oryzae, has now been published (ref. 46; GenBank accession no. AF013598). An integron integrase gene also is located downstream from ilvD in this sequence, although the cassette array appears to be heavily disrupted by transposons.


1. Lan, R. & Reeves, P. R. (2000) Trends Microbiol. 8, 396–401. [PubMed]
2. Boucher, Y., Nesbo, C. L. & Doolittle, W. F. (2001) Curr. Opin. Microbiol. 4, 285–289. [PubMed]
3. Ochman, H., Lawrence, J. G. & Groisman, E. A. (2000) Nature 405, 299–304. [PubMed]
4. Toussaint, A. & Merlin, C. (2002) Plasmid 47, 26–25. [PubMed]
5. Stokes, H. W. & Hall, R. M. (1989) Mol. Microbiol. 3, 1669–1683. [PubMed]
6. Hall, R. M., Collis, C. M., Kim, M.-J., Partridge, S. R., Recchia, G. D. & Stokes, H. W. (1999) Ann. N.Y. Acad. Sci. 870, 68–80. [PubMed]
7. Collis, C. M., Recchia, G. D., Kim, M.-J., Stokes, H. W. & Hall, R. M. (2001) J. Bacteriol. 183, 2535–2542. [PMC free article] [PubMed]
8. Levesque, C., Brassard, S., Lapointe, J. & Roy, P. H. (1994) Gene 142, 49–54. [PubMed]
9. Collis, C. M. & Hall, R. M. (1995) Antimicrob. Agents Chemother. 39, 155–162. [PMC free article] [PubMed]
10. Nield, B. S., Holmes, A. J., Gillings, M. R., Recchia, G. D., Mabbutt, B. C., Nevalainen, K. M. H. & Stokes, H. W. (2001) FEMS Microbiol. Lett. 195, 59–65. [PubMed]
11. Rowe-Magnus, D. A., Guerot, A. M., Ploncard, P., Dychinco, B., Davies, J. & Mazel, D. (2001) Proc. Natl. Acad. Sci. USA 98, 652–657. [PMC free article] [PubMed]
12. Holmes, A. J., Holley, M. P., Mahon, A., Nield, B. S., Gillings, M. R. & Stokes, H. W. (2003) J. Bacteriol. 185, 918–928. [PMC free article] [PubMed]
13. Heidelberg, J. F., Eisen, J. A., Nelson, W. C., Clayton, R. A., Gwinn, M. L., Doson, R. J., Haft, D. H., Hickey, E. K., Peterson, J. D., Umayam, L., et al. (2000) Nature 406, 477–483. [PubMed]
14. Vaisvila, R., Morgan, R. D., Posfai, J. & Raleigh, E. A. (2001) Mol. Microbiol. 42, 587–601. [PubMed]
15. da Silva, A. C. R., Ferro, J. A., Reinach, F. C., Farah, C. S., Furlan, L. R., Quaggio, R. B., Monteiro-Vitorello, C. B., Van Sluys, M. A., Almeida, N. F., Alves, L. M. C., et al. (2002) Nature 417, 459–463. [PubMed]
16. Holmes, A. J., Gillings, M. R., Nield, B. S., Mabbutt, B. C., Nevalainen, K. M. H. & Stokes, H. W. (2003) Env. Microbiol. 5, 383–394. [PubMed]
17. Rowe-Magnus, D. A., Guerot, A.-M., Biskri, L., Bouige, P. & Mazel, D. (2003) Genome Res. 13, 428–442. [PMC free article] [PubMed]
18. Stokes, H. W., O'Gorman, D. B., Recchia, G. D., Parsekhian, M. & Hall, R. M. (1997) Mol. Microbiol. 4, 731–745. [PubMed]
19. Collis, C. M., Kim, M., Partridge, S. R., Stokes, H. W. & Hall, R. M. (2002) J. Bacteriol. 184, 3017–3026. [PMC free article] [PubMed]
20. Hayward, A. C. (1993) in Xanthomonas, eds. Swings, J. G. & Civerolo, E. L. (Chapman & Hall, London), pp. 1–119.
21. Vauterin, L., Yang, P., Alvarez, A., Takikawa, Y., Roth, D. A., Vidaver, A. K., Stall, R. E., Kersters, K. & Swings, J. (1996) Syst. Appl. Microbiol. 19, 96–105.
22. Van den Mooter, M. & Swings, J. (1990) Int. J. Syst. Bacteriol. 40, 472–489.
23. Hauben, L., Vauterin, L., Swings, J. & Moore, E. R. B. (1997) Int. J. Syst. Bacteriol. 47, 328–335. [PubMed]
24. Young, J. M., Dye, D. W., Bradbury, J. F., Panagopoulos, C. G. & Robbs, C. F. (1978) N. Z. J. Agr. Res. 21, 153–177.
25. Dye, D., Bradbury, J., Goto, M., Hayward, A. C., Lelliot, R. A. & Schroth, M. N. (1980) Rev. Plant Pathol. 59, 153–168.
26. Vauterin, L. & Swings, J. (1997) J. Ind. Microbiol. Biotechnol. 19, 77–82.
27. Schaad, N. W., Vidaver, A. K., Lacy, G. H., Rudolph, K. & Jones, J. B. (2000) Phytopathology 90, 208–213. [PubMed]
28. Young, J. M., Bull, C. T., De Boer, S. H., Firrao, G., Gardan, L., Saddler, G. E., Stead, D. E. & Takikawa, Y. (2001) Phytopathology 91, 617–620. [PubMed]
29. Bouzar, H., Jones, J. B., Stall, R. E., Louws, F. J., Schneider, M., Rademaker, J. L. W., de Bruijn, F. J. & Jackson, L. E. (1999) Phytopathology 89, 328–335. [PubMed]
30. Vicente, J. G., Conway, J., Roberts, S. J. & Taylor, J. D. (2001) Phytopathology 91, 492–499. [PubMed]
31. Barak, J. D. & Gilbertson, R. L. (2003) Phytopathology 93, 596–603. [PubMed]
32. Gillings, M. & Fahy, P. C. (1993) Plant Pathol. 42, 744–753.
33. Heyndrickx, M., Vauterin, L., Vandamme, P., Kersters, K. & De Vos, P. (1996) J. Microbiol. Methods 26, 247–259.
34. Louws, F. J., Fulbright, D. W., Stephens, C. T. & de Bruijn, F. J. (1994) Appl. Env. Microbiol. 60, 2286–2295. [PMC free article] [PubMed]
35. Koeuth, T., Versalovic, J. & Lupski, J. R. (1995) Genome Res. 5, 408–418. [PubMed]
36. Gillings, M., Fahy, P. C. & Bradley, J. (1998) Australasian Plant Pathol. 27, 97–103.
37. Partridge, S. R., Recchia, G. D., Scaramuzzi, C., Collis, C. M., Stokes, H. W. & Hall, R. M. (2000) Microbiology 146, 2855–2864. [PubMed]
38. Garcia-Vallve, S., Guzman, E., Montero, M. A. & Romeu, A. (2003) Nucleic Acids Res. 31, 187–189. [PMC free article] [PubMed]
39. Nemergut, D. R., Martin, A. P. & Schmidt, S. K. (2004) Appl. Environ. Microbiol. 70, 1160–1168. [PMC free article] [PubMed]
40. Drouin, F., Mélançon, J. & Roy, P. H. (2002) J. Bacteriol. 184, 1811–1815. [PMC free article] [PubMed]
41. Léon, G. & Roy, P. H. (2003) J. Bacteriol. 185, 2036–2041. [PMC free article] [PubMed]
42. Clark, C. A., Purins, L., Kaewrakon, P., Focareta, T. & Manning, P. A. (2000) Microbiology 146, 2605–2612. [PubMed]
43. Michael, C. A., Gillings, M. R., Holmes, A. J., Hughes, L., Andrew, N., Holley, M. P. & Stokes, H. W. (2004) Am. Nat. 164, 1–12. [PubMed]
44. Grainge, I. & Jayaram, M. (1999) Mol. Microbiol. 33, 449–456. [PubMed]
45. Lindow, S. E. & Leveau, J. H. J. (2002) Curr. Opin. Biotechnol. 13, 238–243. [PubMed]
46. Lee, B. M., Park, Y. J., Park, D. S., Kang, H. W., Kim, J. G., Song, E. S., Park, I. C., Yoon, U. H., Hahn, J. H., Koo, B. S., et al. (2005) Nucleic Acids Res. 33, 577–586. [PMC free article] [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

    Your browsing activity is empty.

    Activity recording is turned off.

    Turn recording back on

    See more...