A recent re-emerging bacterial canker disease incited by Pseudomonas syringae pv. actinidiae (Psa) is causing severe economic losses to Actinidia chinensis and A. deliciosa cultivations in southern Europe, New Zealand, Chile and South Korea. Little is known about the genetic features of this pathovar. We generated genome-wide Illumina sequence data from two Psa strains causing outbreaks of bacterial canker on the A. deliciosa cv. Hayward in Japan (J-Psa, type-strain of the pathovar) and in Italy (I-Psa) in 1984 and 1992, respectively as well as from a Psa strain (I2-Psa) isolated at the beginning of the recent epidemic on A. chinensis cv. Hort16A in Italy. All strains were isolated from typical leaf spot symptoms. The phylogenetic relationships revealed that Psa is more closely related to P. s. pv. theae than to P. avellanae within genomospecies 8. Comparative genomic analyses revealed both relevant intrapathovar variations and putative pathovar-specific genomic regions in Psa. The genomic sequences of J-Psa and I-Psa were very similar. Conversely, the I2-Psa genome encodes four additional effector protein genes, lacks a 50 kb plasmid and the phaseolotoxin gene cluster, argK-tox but has acquired a 160 kb plasmid and putative prophage sequences. Several lines of evidence from the analysis of the genome sequences support the hypothesis that this strain did not evolve from the Psa population that caused the epidemics in 1984–1992 in Japan and Italy but rather is the product of a recent independent evolution of the pathovar actinidiae for infecting Actinidia spp. All Psa strains share the genetic potential for copper resistance, antibiotic detoxification, high affinity iron acquisition and detoxification of nitric oxide of plant origin. Similar to other sequenced phytopathogenic pseudomonads associated with woody plant species, the Psa strains isolated from leaves also display a set of genes involved in the catabolism of plant-derived aromatic compounds.
Citation: Marcelletti S, Ferrante P, Petriccione M, Firrao G, Scortichini M (2011) Pseudomonas syringae pv. actinidiae Draft Genomes Comparison Reveal Strain-Specific Features Involved in Adaptation and Virulence to Actinidia Species. PLoS ONE 6(11): e27297. https://doi.org/10.1371/journal.pone.0027297
Editor: Mark Alexander Webber, University of Birmingham, United Kingdom
Received: July 21, 2011; Accepted: October 13, 2011; Published: November 23, 2011
Copyright: © 2011 Marcelletti et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The work was funded by the Italian Ministry of Agriculture and Forestry through the Agricultural Research Council (C.R.A.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Pseudomonas syringae is a worldwide phytopathogenic microorganism mainly adapted to plant species, both monocotyledon and dicotyledon, and either cultivated or grown in wild habitats. In addition to its well-known dispersal and colonization of cultivated crops by avenues such as seeds, bulbs, bud grafting, rain and wind, there is also evidences that strains of P. syringae strains can be disseminated in various environments through the water cycle ,  and aphids . P. syringae strains have also been isolated from Antarctic areas .
The most common symptoms of P. syringae include leaf spots and necrosis, fruit specks and scabs, flower wilting, twig die-back, branch and trunk cankers and, in particular circumstances, plant death . On the basis of visually assessed symptoms and host range tests and with the aid of biochemical, physiological and nutritional tests and molecular typing, P. syringae (i.e. the P. syringae species complex) is divided into 57 pathovars . To genetically circumscribe 48 P. syringae pathovars and some related species of phytopathogenic pseudomonads, Gardan et al.,  performed DNA-DNA hybridisation and ribotyping analyses and pointed out nine discrete genomospecies. In this study P. s. pv. actinidiae (Psa) was not included. By performing repetitive-sequence PCR, ARDRA and AFLP analyses, this pathovar was subsequently placed into genomospecies 8 together with P. avellanae and P. s. pv. theae , .
Psa is the causal agent of bacterial canker of kiwigreen (Actinidia deliciosa), and was first reported in Japan . It was then subsequently isolated in South Korea  and Italy . In the Asian countries the pathogen caused relevant economic losses , whereas in Italy it has incited occasional leaf spot, twig die-back and bark canker over the past 15 years but never destructive outbreaks . A bacterial canker outbreak on A. deliciosa in central China (Shaanxi province) was observed during 1990–1991 and reported ten years later . Subsequently, another record of this disease on A. deliciosa was reported also in the Anhui province (Southeast China) . Recently, the pathogen has been found in Portugal  and Chile . During 2008–2011, Psa suddenly and very rapidly incited severe epidemics of bacterial canker in central Italy. During these epidemics the kiwigold (A. chinensis) was first affected and, afterwards, A. deliciosa . Psa was first isolated from A. chinensis in southwest China (Sichuan province) in 1989  and later on in southeast China (Anhui province)  and South Korea . Psa was also isolated from wild A. arguta and A. kolomikta plants grown in Japan , . In 2010, the pathogen was also found both on A. chinensis and A. deliciosa in northern Italy  and in France . Psa strains identical to those found in Italy and retained very virulent to both Actinidia species have also been recently identified in New Zealand . In Italy, the kiwigold cultivars (Hort16A, JinTao, Soreli) appear to be very susceptible and as a result, thousands of trees have dead. The main symptom of the disease are leaf spots and necrosis, extensive twig die-back, reddening of the lenticels, bleeding cankers on the trunk and leader with whitish to orange ooze (Figure 1). In Italy molecular typing, which has been performed with repetitive-sequence PCR and MLST, has revealed that there are currently clonal outbreaks of bacterial canker to both A. chinensis and A. deliciosa irrespective of the geographical areas of origin of the isolates and that the strains of the present epidemics are distinct from those causing bacterial canker on A. deliciosa in the past , . As a virulence factor, some Psa strains produce phaseolotoxin , , which is encoded by a mobile gene cluster representing one of the first examples of horizontal gene transfer among phytopathogenic bacteria , .
The sequenced I-Psa and I2-Psa strains from Italy were isolated from the leaves herein showed. A) Leaf symptoms on Actinidia deliciosa cv. Hayward (June, 1992); b) leaf symptom on A. chinensis cv. Hort16A (June, 2008). Note the red-rusty colour of the spots and the chlorotic halo on A. deliciosa and the brownish spot without halo on A. chinensis; c) large canker in deep winter, induced by Psa on the main leader of an adult A. chinensis cv. Hort16A plant, in central Italy (February, 2009). Note the complete destruction of all the external woody tissues of the plant.
These re-emerging, sudden and destructive worldwide cases of bacterial canker on highly-prized crops such as kiwigreen and kiwigold prompted us to an in-depth investigation of the genomic structure of Psa. Comparative genomics can provide insights into the host-pathogen interaction pathways, differential virulence factors and the chronological evolution of pathogens . In recent years, complete or draft genome analyses have been performed for important phytopathogenic pseudomonads such as P. s. pv. tomato , , P. s. pv. phaseolicola , P. s. pv. syringae , P. s. pv. oryzae , P. s. pv. tabaci , P. s. pv. aesculi , P. savastanoi pv. savastanoi  and P. savastanoi pv. glycinea . Sequencing multiple strains of a species or pathovar can provide important information on the possible differential evolution and adaptative mechanisms of phytopathogenic bacteria towards their hosts , , .
For the sequencing, we selected three representative Psa strains: NCPPB3739 ( = KW 11), the type-strain of the pathovar, which was isolated in 1984 in Japan from A. deliciosa, cultivar Hayward ; NCPPB3871, which was isolated in 1992 in Italy from A. deliciosa, cultivar Hayward ; and CRA-FRU 8.43, which was isolated in 2008 in Italy from A. chinensis, cultivar Hort16A , . All strains were isolated from leaf spot symptoms. These strains, which were isolated from the same organ and represent the initial outbreaks of bacterial canker on A. deliciosa in Japan (1984) and Italy (1992) as well as the current severe epidemics on A. chinensis in Italy, are good candidates for elucidating the host-pathogen relationships and the evolutionary adaptation of Psa towards two Actinidia species. The aim of this study was to investigate the biology and evolution of Psa strains that cause bacterial canker to different Actinidia species in many areas of world. We achieved this aim by performing a comparison of the genes found in the draft genomes of the Psa strains with other P. syringae pathovars and by determining the genomic variation among the Psa strains. We demonstrate that Psa shows relevant intrapathovar variations which are probably due to the gain and loss of variable genomic regions. Similar to the other sequenced phytopathogenic pseudomonads associated with woody plant species, the Psa strains isolated from Actinidia spp. leaves also display a set of genes involved in the catabolism of plant-derived aromatic compounds.
Genome-wide sequence data
We generated genome-wide Illumina IIx sequence data from one strain of Psa isolated in Japan, NCPPB3739 ( = KW 11) which is the type-strain of the pathovar and here referred to as J-Psa, and two Psa strains from Italy, NCPPB3871 and CRA-FRU 8.43, which were isolated during two different outbreaks of bacterial canker on Actinidia species and here referred to as I-Psa and I2-Psa, respectively. The Illumina sequencing provided nearly 10 millions of 100 nts reads that passed the quality checking. Sequencing of the J-Psa library provided 1,672,966 reads which were assembled into 833 contigs (N50 = 14,838; largest contig: 67,329) for a total of 5,931,199 nts (a coverage of 27.7 x). Sequencing of the I-Psa library provided 4,083,706 reads which were assembled into 466 contigs (N50 = 27,730; largest contig: 122,209) for a total of 5,938,909 nts (a coverage 67.6 x). Finally, sequencing of the I2-Psa library resulted in 3,823,264 reads which were assembled into 590 contigs (N50 = 22,372; largest contig: 85,982) for a total of 6,144,044 nts (a coverage 61.2 x). Based on the previous sequenced genomes of P. syringae pathovars, the obtained 6 Mb genome of Psa was of the expected size. The G + C content of the three strains ranges from 58.5 and 58.8% (Table 1). The sequences of the assemblies have been deposited in DDBJ/EMBL/GenBank under the following accessions: AFTF00000000 (I-Psa), AFTG00000000 (I2-Psa), AFTH00000000 (J-Psa).
Pairwise alignment between the draft genomes of Psa and the complete genome of Pto DC3000 and occurrence of variable regions
To investigate differences between the genomes of the three Psa strains and Pto DC3000, the closest pathovar of genomospecies 8 according to Gardan et al. , the draft genomes were aligned and compared using MAUVE 2.3.1. software (Figure 2). The percent similarities between the three Psa and Pto DC3000 are of 81.99, 82.03 and 79.48 for J-Psa, I-Psa and I2-Psa, respectively (Table 1). The alignment of the three Psa draft genomes with the complete genome of Pto DC3000 is shown in Figure 2. An ad hoc PERL script was used to establish the similarity of the three Psa genomes. The genomes of J-Psa and I-Psa resulted 99.75% and display only about 14,000 nt of differences, whereas the I2-Psa genome displays a similarity of 88.20% with those of the other two Psa strains. For each Psa strain, the presence of variable regions along the genomes, which are good candidates for horizontal gene transfer, were identified as regions larger than 10 kb in a contig and appeared as a gap in the genome alignment or as regions characterised by a different G+C content with respect to the average content of the three Psa strains. The variable regions found by comparison of the three Psa strains among themselves and with Pto DC3000 are shown in Tables 2, 3 and S1. The highest content of variable regions, namely 12, was found when I2-Psa was compared with J-Psa and I-Psa (Table 4). Variable region 7 was characterised by the presence of a prophage PSPPHO6, which has also been previously found Pph 1448A. Other evidence for the presence of mobile genetic elements has been found in variable region 8 (plasmid-partitioning protein and transposase) and 11 (phage and prophage). Variable region 2 of J-Psa and I-Psa includes the phaseolotoxin cluster (Table 3). An example of the occurrence of variable regions in the Psa genomes is shown in Figure 3.
Colored blocks outline genome sequence that align to part of another genome, and is presumably homologous and internally free from genomic rearrangement (Locally Colinear Blocks or LCBs). Areas that are completely white were not aligned and probably contain sequence elements specific to a particular genome. Blocks below the centre line indicate regions that align in the reverse complement (inverse) orientation. A profile is drawn within each LCB with the height of the color corresponding to the average degree of sequence conservation.
The violet segments (on the right) point out the variable region 3, present in all three Psa strains but not in Pto DC3000; the deep blue (on the left) segments point out the variable region 2, present in J-Psa and I-Psa but absent in I2-Psa and Pto DC3000. The blue segments indicate another variable region present in Pto DC3000 and I2-Psa but not in J-Psa and I2-Psa. The figure shows also some others shorter regions (i.e. light green segments) probable examples of horizontal gene transfer.
Psa is phylogenetically closer to P. s. pv. theae than to P. avellanae
To establish phylogenetic relationships between Psa and the other pathovars or species of the P. syringae complex, we used MultiLocus Sequence Type (MLST) analysis. The relationships within the nine genomospecies, which were determined by Gardan et al. , were assessed by considering as many strains of each genomospecies for which the complete or partial sequence of orthologous housekeeping genes was deposited in the NCBI databank as possible. For members of eight out of nine genomospecies, we constructed a concatenated dendrogram based on the neighbour-joining (NJ) algorithm using the gyrB, rpoB and rpoD gene fragments for a total of 1,646 nucleotides. For genomospecies 7 (i.e. P. s. pv. tagetis and P. s. pv. helianthi), there were not enough gene sequences in the databank, and consequently, it was not included in the analysis. The phylogenetic tree is shown in Figure 4. Psa appears closely related to P. s. pv. theae-type strain and slightly distant from P. avellanae-type strain, which are the other two members of genomospecies 8. Genomospecies 3 (P. s. pv. tomato) is the most related to the genomospecies 8, as has already established by Gardan et al., . Furthermore, we analysed the phylogenetic position of the three Psa strains among the P. syringae pathovars using more orthologous genes. In addition, a concatenated tree, based on the maximum likelihood (ML) algorithm and the acnB, fruK, gltA, pgi, rpoB and rpoD gene fragments for a total of 2,926 nucleotides, was also built (Figure S1). Again, the Psa strains are more closely related to P. s. pv. theae than to P. avellanae. In addition, I2-Psa appears to not be identical to the other two Psa strains sequenced here. The high genetic similarity between Psa and P. s. pv. theae observed in the present and previous studies, the records of bacterial canker incited by Psa from both cultivated and wild Actinidia species in eastern Asia countries , , , , , – and the fact that P. s. pv. theae has, so far, been solely reported solely in Japan, led us to postulate that these two closely-related P. syringae pathovars also have their origin in eastern Asia, as has already been hypothesised by Ushiyama et al. . The consideration that the Actinidiaceae and Theaceae families, which are genetically closely related, both originated in eastern Asia would support such an hypothesis.
Phylogenetic relationships were estimated from concatenated sequences from three housekeeping genes, gyrB, rpoB and rpoD (1,646 bp), using the neighbour-joining (NJ) algorithm. Bootstrap values are reported at each branching. Members of all nine genomospecies (Gardan et al., 1999), except genomospecies 7, are included into the analysis. The letter followed by the number reported in brackets indicates the genomospecies sensu Gardan et al. .
Psa harbors putative pathovar-specific mobile genetic elements of potential importance in adaptation to Actinidia spp
With the total protein complement of the Psa strains, we focused on putative proteins encoded by the genome of each strain that showed no significant homologies with proteins encoded by previously sequenced genomes of phytopathogenic pseudomonads. In particular, all the Psa strains display putative phage integrases, integrase family proteins, and transposases. Interestingly, all the Psa strains have putative homologues to the cAMP protein Fic which induces filamentation in the bacterial cells. Such proteins have been found in the Dickeya zeae strain ECH1591, which causes soft rot diseases, but have never been reported in phytopathogenic pseudomonads.
Comparison of the protein complement of Psa
Because figures calculated by MUMMER may be severely overestimated due to the misassemble of the short Illumina reads that may occur in correspondence to repeated sequences, we used the predicted protein complements to estimate the differences among the three Psa strains. An ORF search using GLIMMER predicted 5,670 genes in the J-Psa genome draft which included 20.37% hypothetical proteins and 3.54% conserved hypothetical proteins, 5,557 genes in I-Psa which included 20.54% hypothetical proteins and 3.74% conserved hypothetical proteins and 5,714 genes in I2-Psa which included 22.72% hypothetical proteins and 3.48% conserved hypothetical proteins. For each strain, all predicted proteins were preliminarily annotated by homology search in the RefSeq database and ordered by role categories according to TIGRFAMs. Here we show the protein categorisation of I2-Psa (Table S2). In agreement with the results of the genome-wide comparisons reported above, the resultant putative protein complement of I-Psa and J-Psa were highly similar. As many as 5,076 ORFs were found to be 100% identical in their DNA sequence between the two strains and only seven ORFs were found to be polymorphic, which accounted for a total of 18 SNPs (Table 4), if two cases of ambiguity due to presumptive gene duplication are excluded. The remaining predicted proteins were found to be different from the reciprocal strain due to the differential fragmentation of the draft genomes into contigs or to occasional misassemble, i.e. a small number of predicted ORFs (38 in the J-Psa comparison with I-Psa and 52 in the reciprocal comparison) were not detected in the reciprocal strain as such, but in all cases BLASTn analysis showed that the orthologous sequences were present and had likely been neglected by GLIMMER due to contig fragmentation and consequent loss of signals needed by the program for prediction. Conversely, strain I2-Psa showed significant differences when compared to I-Psa and J-Psa. In the comparison versus J-Psa 2,083 ORFs were identical while 2,140 showed sequence polymorphisms (SNPs). The remaining ORFs were further analysed and, in most cases were found to differ in length. However, whether this difference was due to actual length polymorphism or differential fragmentation of the draft genome into contigs could not be determined. For as many as 398 ORFs, however, no sequence with significant similarity could be found in the I-Psa or in the J-Psa genome draft.
Origin and evolutionary relationships among Psa strains
Further analysis of these I2-Psa specific ORFs was performed to determine their origin. We searched the 35 Pseudomonas spp. and pathovar genomes available as draft or as complete genome sequences from NCBI as well as the draft genomes of P. avellanae BCIP613 and P. syringae pv. theae NCPPB2598, type-strains of these pathogens, whose draft genomes are available (Marcelletti, Firrao, Scortichini, unpublished data in these labs) for comparison using BLASTn. The results of this search showed that among the 398 I2-Psa-specific ORFs, there were 238 ORFs for which no homolog could be found in P. avellanae BCIP 613 or P. s. pv. theae NCPPB2598. It was also found that 49% (i.e. 196) of the 398 proteins that do not have homologous in J-Psa or I-Psa matched sequences in at least one of the genomes of Pto strains assessed, specifically 158, 144, 138, 114 and 116 matches for Pto strains K40, Max13, NCPPB1108, T1 and DC3000, respectively. These evidence as well as the annotation of several of the deduced proteins as phage or prophage proteins strongly suggests that a large part of this I2-Psa-specific DNA has been acquired by horizontal gene transfer from a strain of P. syringae genomospecies 3. However, within the group of 398 I2-Psa specific DNA, we also detected eight ORF that had significant BLASTn hits in the P. avellanae BCIP613 or P. s. pv. theae NCPPB2598 draft genome sequences but no hit in any of the 35 Pseudomonas spp. and pathovars genome sequences available from NCBI (Table 4). The most obvious explanation for this result is that these eight sequences shared by I2-Psa and other strains of genomospecies 8 were lost by I-Psa and J-Psa during their evolution.
To gain access to further information about the origin of I2-Psa, we analysed the DNA sequences of all the genes that were polymorphic in all three Psa strains and compared them with their orthologous in the P. avellanae BCIP613 or P. s. pv. theae NCPPB2598 draft genome sequences. As reported in the Table 4, although the pattern displayed by I2-Psa was more similar to I-Psa than to J-Psa, in most cases I2-Psa displayed the ancestral residue in common with the other strains of genomospecies 8.
Furthermore, we randomly selected 171 ORFs from the list of genes that were found to be orthologues among the three strains and in the draft genomes of P. avellanae BCIP613 or P. s. pv. theae NCPPB2598. The alignment of the concatenation of these genes, which consisted of 166,160 nucleotide positions was used to determine the genealogy with maximum likelihood and perform hypothesis testing. Based on the results, as shown in the tree in Figure 5a, the I2-Psa strains originated from a common ancestor of the I-Psa and J-Psa strains and was not a derivative of either of these strains. To provide statistical support to this presumptive genealogy, we estimated the likelihood of an alternative genealogy with the constraint that I2-Psa and I-Psa were monophyletic (Figure 5b), and performed a test for monophyly to evaluate whether such a null hypothesis could be rejected with statistical significance. The aim of the monophyly test is the evaluation, by means of parametric bootstrap, of the significance of a likelihood ratio calculated comparing a null hypothesis with the unconstrained maximum likelihood tree. The log likelihoods of the null hypothesis and the unconstrained tree were -263874.12 and -263011.44, respectively, and their ratio (Δ = 1725.36) was compared with the delta distribution in a set of alignments of simulated sequences evolved in silico using the unconstrained tree as guidance. The comparison showed that the null hypothesis was to be rejected (P = 0.016). In summary, the analysis of ancestral state residue conservation, the maximum likelihood analysis of genealogies and the evidence of eight ORFs that are genomospecies 8-specific and are present only in I2-Psa strongly suggest that these strains did not evolve from the organisms that caused the epidemic in 1984–1992 but rather from a common ancestor.
Secretion systems associated with pathogenicity and virulence
The three Psa strains sequenced display structural genes involved in the biosynthesis of the type I, II, III, IV and VI secretion systems (TSS). The T1SS encodes for orthologues of the HlyD family secretion protein involved in the transportation of the Escherichia coli α-haemolysin toxin. Both the general secretion Sec-pathway and the Tat-pathway of the T2SS are present in the three Psa strains. These two secretion systems are also present in the other sequenced plant pathogenic pseudomonads, even though the T1SS is not present in Pto DC3000 . In Pto DC3000, the twin-arginine translocation (Tat) system of the T2SS appears to be an important virulence determinant . A complete T3SS, similar to those found in the P. syringae complex  has also been found in the three Psa strains. The T3SS contains the hrp/hrc cluster and the transcriptional regulatory HrpL, HrpR and HrpS proteins. TrbC and VirK protein orthologues for the T4SS are present in the three Psa strains. In addition, I2-Psa also displays other orthologous proteins for the T4SS, namely TraG and VirB8, which are also present in other P. syringae pathovars. Noteworthy the GC contents of these genes are lower than the flanking genomic regions, this indicating the occurrence of possible genomic islands. Finally, the three Psa strains also display two clusters of orthologous proteins of the T6SS: the Vgr family protein, which is identical to that of P. putida GB-1 and the OmpA family protein which is required for the T6SS functionality . ImpA, an inner membrane protein of the T6SS is also present. The precise functions of these secretion systems in Psa have yet to be investigated.
Type III secretion system effectors
A comparison of the effector repertoire of the three Psa strains based on the complete effector repertoire of P. s. pv. tomato DC3000, other P. syringae strains and Psa MAFF302091 reveals a “core” set of 33 hop and 6 avr putative effector genes that are conserved in all strains (Figure 6). In general, in these putative effectors, the amino acid identity is very high (i.e. >90%) to the most similar orthologue of the P. syringae pathovars found in GenBank (Table S3). However, for hopAC1, hopAE1, hopI1, hopM1, and hopZ3, the amino acid identity with respect to their most similar orthologues ranged from 68 to 79%. By contrast, the avrPto1-like effector of the three Psa strains shows an amino acid identity of 44% with the orthologue of P. s. pv. aesculi NCPPB3681. All three Psa strains did not show the presence of hopAB and hopAF which are considered conserved effector genes present in Pto DC3000, Pph 1448A and Psy B728a . Interestingly, I-Psa displays four putative effector genes, namely hopA1, hopAA1-2, hopH1, and hopZ2-like, that are not present in the J-Psa and I-Psa strains. The first three putative effector genes display a relative similarity with the orthologues of Pto DC3000 (91%), Pto DC3000 (99%) and P. s. pv. tomato T1 (97%), whereas hopZ2-like shows a similarity of 40% with the orthologue of P. s. pv. syringae. HopA1 triggers effector-triggered immunity in tobacco, other Nicotiana species  and in many Arabidopsis accessions  and is supposedly involved in host range specificity, even though its virulence functions are still unknown . HopAA1-2 is a paralogue of the hopAA1-1 effector, which is located in the conserved effector locus (CEL) region of the hrp/hrc cluster and is considered to be among the ancient P. syringae effectors that were acquired before the radiation of P. syringae into the current pathovars . HopAA1-2 is present in Pto DC3000 but not in Pto T1, Pph 1448A and Psy B728a. Evidence for a strong virulence function for hopAA1-2 in plants has yet to be fully investigated . Notably, in a wide assessment on the presence/absence of effector genes in 91 strains of P. syringae pathovars, Sarkar et al.  found that hopAA1-2 and hopA1 are among the least distributed effector genes. In fact, these effectors have only been found in eight and 16 out of the 91 strains tested, respectively, and they have only been found present together in three Pto strains. If these two effectors can explain the relevant aggressiveness of I2-Psa towards A. chinensis and A. deliciosa they deserve further in-depth studies. HopH1 is considered to be a variably distributed effector gene in the P. syringae pathovars . The hopZ2-like effector protein showed a low similarity (i.e. 44%) with hopZ2. The possibility that this putative protein represents a different and new effector gene cannot be ruled out. Finally, I2-Psa lacks hopX1, another conserved effector present in other P. syringae pathovars  as well as in J-Psa and I-Psa. These results would indicate that both Actinidia deliciosa and A. chinensis can be infected by different Psa strains displaying different arrays of effector genes and represents a case of convergent evolution of closely-related pseudomonads to the same host genus.
Variation in plasmid content among Psa strains
Using agarose gel electrophoresis, we compared the number and the size of native plasmids present in the three sequenced Psa strains as well as for comparative purposes, in other representative Psa strains from Japan (i.e. outbreaks of 1984) and Italy (i.e. outbreaks of 2008–2010 in different Italian regions). We found that all the Japanese strains from the 1984 outbreak as well as I-Psa harbour a native plasmid of about 50 kb. This plasmid is absent in I2-Psa and in all the other representative Psa strains isolated in Italy during the recent epidemics of bacterial canker on A. chinensis and A. deliciosa (Figure 7). Remarkably, I2-Psa as well as all the other Psa strains obtained during the recent epidemics of bacterial canker in Italy displayed the presence of a plasmid of about 160 kb, not present in both J-Psa and I-Psa. This represent a substantial acquisition of genetic material for I2-Psa. All the three Psa strains has the the repA gene, essential for plasmid replication, of the P. syringae pPT23A-like plasmid family.
Agarose gel electrophoresis to compare the number and size of native plasmids in the genome of Psa strains. The gels show also other representative Psa strains from the outbreak of bacterial canker in Japan (i.e. 1984) and from the current severe epidemics in Italy. See also Materials and Methods. Note as the ca. 50 kb plasmid present in J-Psa and I-Psa is not contained in all the Psa strains isolated from the current epidemic in Italy. By contrast, I2-Psa and other strains obtained from the recent epidemics of bacterial canker in Italy contain a plasmid of about 160 kb.
Presence and absence of the phaseolotoxin gene cluster and other toxins
The phaseolotoxin gene cluster, argK-tox, is located on the chromosome of Pph and Psa. It comprises three tyrosinase-recombinase-encoding genes: txi1, txi2 and txi3, which are located at the left end of the cluster; the phtE locus, which contains ORFs showing homologies to genes encoding amino acid transferases and ARAC family and fatty acid desaturase; and the argK gene which encodes the phaseolotoxin-resistant ornithine carbamoxyltransferase. The cluster is flanked by two regions, ACT059 and ACT094 . Thirty-eight kilobases of the argK-tox cluster of the Pph MAFF302282 and Psa NCPPB3739 (i.e. J-Psa) strains are identical . In our draft, we found that J-Psa and I-Psa contain the phaseolotoxin gene cluster, which display homologies of 99.94% and 99.99%, respectively, whereas I2-Psa does not contain this 38 kb region of the cluster and the flanking regions, ACT059 and ACT094, are contiguous (Figure S2). Some genetic features indicate that the phaseolotoxin cluster, also referred to as the tox-island, has been acquired by P. syringae pvs. phaseolicola and actinidiae through horizontal gene transfer from an unknown species , . The complete absence of the cluster in many highly virulent Psa strains, as found both in Italy and New Zealand ,  would confirm such a hypothesis and indicate the remarkable aggressiveness of the pathovar even without this virulence factor. These findings also indicate the separate but convergent evolution of genetically different pseudomonads as phytopathogens of Actinidia spp. Notably, different types of leaf spot lesions on Actinidia spp. have been noticed during the two outbreaks of bacterial canker of kiwifruit in Italy in 1992 and 2008-2011 (Figure 1), even though necrotic spots surrounded by a chlorotic halo have also been frequently observed during these severe epidemics on A. chinensis in Italy. The presence of genes coding for coronatine, syringomycin and syringopeptin toxins was also checked but none of these toxins is present in the draft genomes of the Psa strains of the present study.
Copper resistance and antibiotic detoxification
A thorough search for orthologous genes coding for resistance to copper and antibiotics was performed with the draft genomes of the three Psa strains. All strains contain homologues for the copA and copB genes, which are essential for copper resistance . In P. syringae, copA is located in the periplasmic space, whereas copB is in the outer membrane of the bacterial cell. The other two genes, namely copR and copS, which are required for maximum copper resistance were not found. These results confirm a study by Nakajima et al.  who found only copA and copB in the Psa strains, including NCPPB3739, obtained during the initial outbreak of bacterial canker on A. deliciosa in Japan when the regular spraying of copper bactericides was not yet applied to the infected orchards. In addition, all the Psa strains display a vast set of orthologous genes involved in antibiotic resistance. Among the five superfamilies of efflux transporters, Psa has genes belonging to the resistance nodulation division (RND), multi antimicrobial resistance (MAR), multidrug endosomal transporter (MET) and major facilitator superfamily (MFS). Recent studies on the antibiotic resistance mechanism in Gram-negative bacteria, have stressed that resistance greatly depends on the constitutive or inducible expression of active efflux systems . As an example, the disruption of MexB, a gene of the MAR superfamily, that is also present in Psa strains, dramatically increased the susceptibility of P. aeruginosa to beta-lactams, tetracyclines, fluoroquinolones and chloramphenicol . In addition, all the Psa strain genomes include genes involved in the enzymatic inactivation of beta-lactams, tellurium and a macrolide ABC efflux protein, macAB, which confers resistance to cycloheximide. All the Psa strain genomes contain ampD, a gene coding N-acetyl-anhydromuramyl-L-alanine amidase which cleaves the amide bond between N-acetylmuramoyl and L-amino acids in the bacterial cell wall. Finally, I2-Psa also contains a putative lantibiotic dehydratase domain that has never been found in other phytopathogenic pseudomonads.
Iron acquisition, nitric oxide and sucrose metabolism and quorum sensing
The three Psa strain genomes encoded a number of genes involved in iron acquisition, such as the siderophore pyoverdine and enterobactin involved in the isochorismate synthase and yersiniabactin, a siderophore with a very high affinity for iron. In addition, all the genomes contain the TonB protein. This protein spans the periplasm and is anchored to the cytoplasmic membrane interacting with receptors in the outer membrane to facilitate the uptake of iron-siderophore complexes. In human bacterial pathogens, namely Haemophilus influenzae and H. parainfluenzae, the inactivation of TonB decreased the ability to cause disease . The three Psa strains contain two genes involved in the nitric oxide metabolism, namely nitric oxide dioxygenase and anaerobic nitric oxide reductase, with 100% homology to the same genes in P. s. pv. aesculi . These genes might protect Psa from the host defence responses incited by nitric oxide. The three Psa strains do not have the 8 kb cluster coding for sucrose utilization that is present in the phloem infecting P. s. pv. aesculi strain 2250 isolated in Great Britain . The quorum sensing system of Psa apperas to differ from the classic LuxR/LuxI of other P. syringae strains. In fact, the genes of LuxI family are absent from these three strains, which display putative LuxR family genes. The repressor genes rsaM and rsaL are also absent in the three Psa genomes.
Presence of pectolytic enzymes and catabolism of plant-derived aromatic compounds
The three Psa strains contain pectin lyase and polygalacturonase genes that display complete identity to the orthologues of Pto T1. These enzymes are also present in the soft-rot bacterium P. marginalis . Similar to other P. syringae pathovars that infect woody hosts, such as P. s. pv. aesculi and P. savastanoi pv. savastanoi , , Psa also displays genes involved in the catabolism of plant-derived aromatic compounds using both the cathecol branch of the β-chetoadipate and the protocatechuate 3,4-dioxygenase pathways. In fact, all the Psa genomes encode putative proteins involved in the degradation of anthranilate to catechol (i.e. anthranilate dioxygenase reductase and anthranilate phosphoribosyltranferase) as well as proteins involved in the catabolism of cathecol (cathechol 1,2-dioxygenase, muconolactone delta-isomerase, muconate cycloisomerase N-terminal and dienelactone hydrolase. Moreover, the Psa strains have putative pcaG and pcaH genes encoding the two subunits (α and β) of protocatechuate 3,4-dioxygenase, an enzyme involved in the degradation of protocathecuate, which is present in soil-inhabiting bacteria.
Differential multiplication trend of Psa strains in Actinidia spp. leaves and pathogenicity test to tomato
Field evidence from the recent epidemics of bacterial canker in Italy indicate that the current Psa population (I2-Psa) is aggressive to both A. chinensis and A. deliciosa. In the latter species, I2-Psa incited more severe symptoms compared with the Psa population of the past outbreaks in Italy (I-Psa) when solely A. deliciosa was cultivated. Inoculations of A. deliciosa and A. chinensis leaves revealed different multiplication trends between J-Psa, I-Psa and I2-Psa (Figure 8). In fact, I2-Psa performed better than J-Psa and I-Psa in A. chinensis leaves. I2-Psa performed well also on A. deliciosa, although not as well as I-Psa and J-Psa, as three weeks after the inoculation still showed high cell levels (i.e. between 105 and 106 cfu/ml). Conversely J-Psa and I-Psa, that reached a cell concentration of about 1010 cfu/ml 3–4 days after the inoculation with either 103 or 106 cfu/ml when inoculated in A. deliciosa leaves, only grew poorly in A. chinensis leaves where the cell concentration of the strains decreased significantly 21 days after the inoculation. The control leaves did not show any sign of infection. These results indicate that the Psa population currently causing severe damage in Italy and New Zealand is capable to infect both Actinidia species, whereas the Psa population causing outbreaks of bacterial canker to cv. Hayward in Japan and Italy about 20–25 years ago displays more affinity for A. deliciosa. The three Psa strains incited a typical hypersensitivity reaction on tomato leaves, whereas PtoDC3000 caused typical symptoms of bacterial speck.
Multiplication in A. deliciosa cv Hayward (a) and in A. chinensis cv Hort16A (b) leaves. Bacteria were inoculated at 1–2×103 and 1–2×106 cfu/ml. Data represent the mean log of bacterial cell number and standard deviation (SD) as obtained from eight inoculation sites per each sample.
Through Illumina sequencing technology, we have performed a genome-wide survey of genetic variation in Psa strains, the causative agent of bacterial canker of Actinidia spp. worldwide. This study has determined putative variable genomic regions and sets of genes related to the pathogenicity and virulence that could differentially modulate the aggressiveness of pathogen populations towards Actinidia species as well as genes involved in the environmental fitness and adaptation of the bacterium in planta. We confirmed that the re-emerging wave of bacterial canker to A. deliciosa and A. chinensis is being raised by a population of Psa (i.e., I2-Psa) distinct from the population that has led to past outbreaks in Japan and Italy (i.e. J-Psa and I-Psa) , . Moreover, we stress that the current epidemics of bacterial canker are being caused by a Psa population that, most probably, did not originate from that found in Italy about 20 years ago. The origin of this new epidemic wave has rised a large debate worldwide, and it has not been easy to provide convincing answers using conventional approaches. Conversely, the large amount of sequence data obtained in this work has provided unmatched solidity to the reconstructed genealogy of the Psa strains examined.
We refined the taxonomic position of Psa within genomospecies 8 sensu Gardan et al.  using MLST analysis and housekeeping genes, and we found that Psa is phylogenetically closer to P. s. pv. theae than to P. avellanae. Our analysis, performed with all nine genomospecies (except genomospecies 7) currently circumscribes the majority of the P. syringae pathovars and related phytopathogenic pseudomonads and also confirms that the genomospecies 3, including Pto, is the most closely-related cluster. The very high genetic similarity between Psa and P. s. pv. theae, a phytopathogen so far reported solely in Japan, reinforces the assumption that Psa might be of Asian origin. In fact, there are several reports on the occurrence of this pathogen isolated from both A. chinensis and A. deliciosa in China , , , , South Korea , ,  and from A. deliciosa and wild A. arguta and A. kolomikta plants in Japan , , . The possibility of an Asian origin has been already argued .
However, some of the genetic features found in the Psa strains of the past and recent epidemic in Italy and New Zealand indicate that different evolutionary routes have been followed by the ancestor(s) of the two different Psa populations which are represented by J-Psa/I-Psa and by I2-Psa. Our genome-wide analysis indicated that the two strains J-Psa and I-Psa, which were isolated in different years from the same kiwigreen cultivar, in geographically distant areas that were affected by outbreaks of bacterial canker of different severity, are extremely similar. This evidence indicates the major role of climatic conditions on the epidemic of bacterial canker to A. deliciosa. Furthermore, our comparative study of Psa strains shows that despite the apparent prevalence worldwide of a clonal population, a reservoir of diversity of the pathogen has been maintained, which has allowed for a new population, represented by I2-Psa, to emerge about 25 years later from an independent evolutionary line with different genetic characteristics and enhanced epidemic potential. Bacterial canker caused by Psa to both A. chinensis and A. deliciosa has also been recently reported in New Zealand. The molecular typing performed using the MLST analysis of seven housekeeping genes and the detection of 12 effector protein genes allowed to ascertain that the Psa strains of the current epidemics in Italy, here represented by I2-Psa, are identical to those highly virulent strains found in New Zealand, which also lack of the phaseolotoxin gene cluster . Notably, in that country, another Psa population, genetically different and less virulent (i.e. apparently causing only leaf spots) than I2-Psa but capable to infect both kiwigreen and kiwigold has also been identified. Such a population also differs from the two other genetically different Psa populations of the past outbreaks of bacterial canker on A. deliciosa in Asia (i.e. Japan and South Korea) and has been retained as endemically in New Zealand . Whether the highly virulent Psa population currently causing severe economic losses to A. chinensis and A. deliciosa in Italy and New Zealand originated in the area of the origin of the pathogen or from the less virulent endemic Psa population recently isolated in New Zealand still remain to be verified. However, during the last 30 years, four genetically distinct Psa populations have infected, to different extents, different Actinidia species on different continents which is a remarkable case of multiple convergent evolution of phytopathogenic pseudomonad populations of the same pathovar to one single plant genus.
We cannot establish with certainty the origin of this re-emerging wave of bacterial canker in Italy although a likely scenario could be hypothesised. The past outbreaks only caused leaf spot and twig die-back but never the death of thousands of plants. The current I2-Psa population could have been introduced from abroad through latently infected propagative material or infected pollen. Once established in central Italy, it further reached the other Italian regions by means of latently infected propagative material.
We also found, by assessing the Psa growth in the leaves, that A. deliciosa cv. Hayward was more susceptible than A. chinensis cv. Hort16A when inoculated with the Psa strains causing past outbreaks of bacterial canker in Japan and Italy. By contrast, I2-Psa multiplication trend was higher in A. chinensis. However, a relevant inoculum was also found in A. deliciosa three weeks after the inoculation, thus confirming that the re-emerging wave of bacterial canker has been caused by a Psa population that has a high fitness for both Actinidia species. In the Latium region (central Italy), A. chinensis probably largely contributed to the very rapid expansion of such population in the area of kiwigold cultivation, which also acted as reservoir of infection for A. deliciosa.
The question then arises as to how this new highly virulent Psa population originated. The importance of stress factors in promoting bacterial evolution has been recently pointed out. In fact, under stress condition in the host (i.e. nutrient deficiency outside the host, attack by antimicrobial compounds inside the host, and low temperatures), the bacterial competency to uptake DNA is activated and the pathogen can acquire exogenous genetic material that could help it to escape from the stress . In addition, a possible loss of mobile genetic elements carrying avirulence genes can lead to an enhanced virulence . I2-Psa does not have the 50 kb plasmid and the phaseolotoxin gene cluster present in J-Psa and I-Psa but has gained a 160 kb plasmid and a putative prophage that does not occur in the other population. It remains to be verified if any of these stress factors could have promoted the rise of such new Psa population. We did not investigated in details the structure of plasmids but, similarly to other P. syringae pathovars, the three Psa strains have the repA gene that is retained essential for the replication of the pPT23A-like plasmid family .
One of the more striking pieces of evidence of the difference between the two populations is the presence of the phaseolotoxin gene cluster, argK-tox, in J-Psa and I-Psa and its absence in I2-Psa and in all the other strains of the current epidemic assessed so far in Italy and New Zealand , . Phaseolotoxin is considered a major virulence factor for both Psa and Pph , . The argK-tox gene cluster is located on the chromosome and was supposedly acquired through lateral gene transfer from bacteria distantly related to P. syringae or from non pathogenic or avirulent P. syringae strains , . Additionally, one Psy strain, which was isolated from Vicia sativa, displays such a gene cluster , although it showed a nucleotide identity of only 85,3% with that displayed by Psa and Pph . The presence of the argK-tox cluster in one variable region of the J-Psa and I-Psa genomes confirms the acquisition of such genetic trait by lateral gene transfer. Also within Pph have been found many pathogenic strains lacking the phaseolotoxin cluster but very aggressive, similarly to I2-Psa, towards their host plant . Tamura et al. , using an argK-tox- mutant of KW11 (J-Psa) found that the bacterium induced the same type of symptoms in the plant similar to the wild-type, except for the chlorotic halo surrounding the leaf spots. However, the toxin did not promote bacterial growth in planta. It has been postulated that phaseolotoxin, by inhibiting ornitine-carbamoxyltransferase, can reduce or inhibit the growth of other microorganisms . The high identity of the argK-tox clusters found in strains of Psa and Pph suggests a recent acquisition of the cluster by the two pathovars . The worldwide occurrence of many highly virulent strains lacking the argK-tox gene cluster in both Psa and Pph allows us to speculate that the ancestral genomes of these two phytopathogens did not include the phaseolotoxin gene cluster.
Other differences between the two Psa populations were found by the assessment of the effector protein genes. The three strains display an identical core repertoire of 33 hop and 6 avr effector genes. However, the Psa strains also possess some unique effector protein genes. HopX1, hopAR, hopBD2 and hopPmaB are unique to J-Psa and I-Psa, whereas hopA1, hopAA1-2, hopH1, and hopZ2-like are unique to I2-Psa. Such a different effector repertoire found in the three Psa strains might account for the differential aggressiveness showed by the pathogen to Actinidia species. However we still do not know the roles played by these effector genes in terms of pathogenicity and host range for the pathovar actinidiae. In a recent extensive study regarding the evolution of pathogenicity within the P. syringae complex, Baltrus et al.  sequenced also a Psa strain, namely MAFF302091, isolated from A. deliciosa in Japan, in 1984 in the Kanegawa district. Interestingly, this strain differs from J-Psa, isolated in Shizuoka district, for the presence/absence of some effector genes. This would indicate that each single strains might possess a distinct effector repertoire. In a comparative study on the occurrence of effector genes in many P. syringae strains, also Sarkar et al.  analysed Psa MAFF302091. Similar to J-Psa, also Psa MAFF302091 also does not contain hopA1 and hopAA1-2. Noteworthy, these two effectors are rarely found contemporarily present in the 91 P. syringae strains tested and only three Pto strains showed such effectors in their repertoire . However, our pathogenicity test showed that the three Psa strains did not cause infection to tomato. Vanneste et al.  claimed that hopA1 was also present in I-Psa but, according to the genome sequencing, their results appear as an artifact.
Phages, prophages and their morons (i.e. DNA elements inserted between a pair of genes in one phage genome) are known to shape the pathogenicity and virulence of bacterial pathogens and their presence within the bacterial genome can largely contribute to the genetic and phenotypic diversity of bacteria and to the emergence of pathogenic variants . In fact, by carrying various elements contributing to virulence, prophages can also contribute to the individuality of bacteria strains as found in Salmonella, Lactobacillus and Burkholderia , , . In Xylella fastidiosa, the causative agent of infectious diseases of many cultivated crops, prophage-associated chromosomal rearrangements and deletions have been found to be largely responsible for strain-specific differences . Interestingly, in a variable region of I2-Psa, we found the presence of many putative proteins related to the assembly and acquisition of prophage PSPPHO6, which are also present in Pph 1448A. It remains to be ascertained if there is a link between the relevant virulence of I2-Psa to Actinidia spp. and the presence of this prophage which was acquired by horizontal gene transfer and if it could have contributed to the further adaptation of Psa to Actinidia spp.
It is interesting to observe that all three Psa strains sequenced here were isolated from leaf spot symptoms but display a set of genes involved in the degradation of lignin derivatives and other phenolics. In fact, similarly to other P. syringae pathovars associated with woody hosts such as P. syringae pv. aesculi and P. savastanoi pv. savastanoi and to P. putida, a soil-inhabiting species , , the three Psa strains have genes putatively related to the degradation of the anthranilate to the cathecol branch of the β-ketoadipate pathway and to the protocatechuate degradation via the protocathechuate 4,5-dioxygenase pathway. These pathways allow for the utilisation of unsubstituted lignin-related compounds and other plant derived phenolic compounds such as mandalate and phenol . This could explain one of the most striking symptoms induced by Psa on Actinidia spp., the extensive degradation of the woody tissues of the main trunk and leaders mainly occurring during winter. In Italy, I2-Psa incites canker of larger dimensions on A. chinensis cultivars compared to A. deliciosa and sometimes also causes the complete destruction of all external woody tissues, as shown in Figure 1c. It is also worth noting that I2-Psa survived more than 45 days in infected A. chinensis twigs that were pruned and subsequently brought into the lab, without receiving any amendment (i.e. water) (Marcelletti and Scortichini, unpublished data).
The fact that Psa strains can infect both herbaceous (i.e. leaves and young twigs) and woody tissues of the same host plant could mean that differential set of genes are activated when the pathogen is multiplying and infecting different organs of the plant. It would be interesting to further investigate the P. syringae complex to determine whether the capability of infecting herbaceous tissues appeared before or after that of infecting woody tissues.
All Psa strains display also sets of genes that are virulence factors or are important for the survival of the bacterium in planta or for competing with other micro-organisms. In fact, Psa has genes involved in the inhibition of nitric oxide metabolism, namely nitric oxide dioxygenase and anaerobic nitric oxide reduction . Nitric oxide plays a fundamental role in plant disease resistance by acting as a signal-inducing plant gene to synthesise defense-related compounds . The inhibition of nitric oxide synthesis consequently promotes the bacterial growth in planta.
Psa strains contain copA and copB, genes that play a key role in copper resistance . Copper ions are essential for bacterial species but can incite toxic cellular effects if levels of free ions are not controlled. It has been observed that strains of P. syringae with no known history of exposure to copper selection accumulate copper and are resistant. In these bacteria, copper accumulation may have a beneficial role other than in resisting high levels of copper . Interestingly, Nakajima et al.  found that at the beginning of bacterial canker outbreaks in Japan (i.e. 1984–1987) all the Psa strains displayed solely copA and copB. However, after repeated spray treatments with copper-based bactericides, the Psa strains also showed additional genes responsible for the maximal resistance to copper, namely copR and copS.
Psa can counteract the lethal effect of antibiotics by means of multidrug efflux pumps of the multidrug resistance systems encoded by chromosomal genes. These features apparently confer a relevant fitness for the in planta growth of Psa. Notably, I2-Psa also displays a lantibiotic dehydratase protein that might putatively inactivates this class of antibiotics which is produced by Gram-positive bacteria and characterised by its high specific activity against multidrug-resistant bacteria .
The efficient uptake and utilisation of iron through siderophores is regarded as an important virulence factor for phytopathogenic pseudomonads, especially in iron-limited environments . The Psa strains have a set of genes coding for the production of siderophores such as pyoverdine, haemin, enterobactin and yersiniabactin. These last two siderophores are primarily described in the Enterobacteriaceae, and they are characterized by a very high affinity for iron , . Similar to P. s. pv. aesculi, Psa strains also contain hemagglutinin-like proteins. Although they have not been investigated in pseudomonads, in other phytopathogenic bacteria such as Erwinia chrysanthemi, Xylella fastidiosa and Xanthomonas oryzae pv. oryzae these protein have been shown to specifically act in adhesion between the bacterial cell and the plant host cell , .
One laboratory claimed that Psa NCPPB3871 (I-Psa) which was directly received from the National Collection of Plant Pathogenic Bacteria, is not to be a genuine Psa strain . We and other labs  did not find such discrepancy. A putative contamination in the former lab could have occurred.
The comparative genome-wide analysis performed with these three Psa strains representing two different populations of the pathovar, provides important insights into the evolution and adaptation of this pathogen to Actinidia spp. and highlights how a virulence factor like the phaseolotoxin can be lost without decreasing the relative virulence of the bacterium. We also demonstrate how the mobile arsenal of phytopathogenic bacteria (i.e. plasmids and prophages) can be lost and gained by populations of the same pathovar that, consequently, can modulate their virulence accordingly.
Materials and Methods
Psa NCPPB3739 is the type strain of the pathovar and was isolated in 1984, in Japan (Shizuoka district) from a leaf spot lesion of A. deliciosa cv. Hayward . Psa NCPPB3871 was isolated in 1992 in the Roma province (central Italy) from a leaf spot lesion on A. deliciosa cv. Hayward . Psa CRA-FRU 8.43 was isolated in the province of Latina (central Italy) from a leaf spot lesion of A. chinensis cv. Hort16A , , and is the first isolate from the epidemic of bacterial canker currently causing severe damage to the cultivation of A. chinensis and A. deliciosa in Italy . In this paper, the Psa strains are referred to as J-Psa (NCPPB3739), I-Psa (NCPPB3871) and I2-Psa (CRA-FRU 8.43). Figure 1 shows the symptomatic leaves of A. deliciosa and A. chinensis in Italy from where isolations were performed. For this study, the Psa strains were maintained on nutrient agar amended with 5% w/v sucrose (NSA) and incubated at 25±1°C:
Library preparation and Illumina sequencing
Bacterial genomic DNA was extracted from 1 ml of overnight J-Psa, I-Psa and I2-Psa cultures grown in KB broth DNA using a Wizard DNA purification kit (Promega Italia, Padova, Italy) following the manufacturer's instructions. DNA was measured and checked for quality using a NanoDrop (NanoDrop products, Wilmington, DE, USA). A total of 10 µg of DNA from each sample was fragmented by incubation for 70 min with 5 µl of dsDNA Fragmentase (New England Biolabs MA, USA). The reaction was stopped with EDTA and purified using a QIAquick PCR purification kit (QIAGEN, Hilden, Germany). The eluate was end repaired using an End Repair kit (New England Biolabs, MA, USA) for 30 min at 20°C. The end-repaired DNA was A-tailed for 30 min at 37°C using a d-A Tailing kit (New England Biolabs, MA, USA). After purification using the MinElute purification kit (QIAGEN), the DNA was ligated using Quick T4 DNA ligase (New England Biolabs) to 500 pmol of Illumina adaptors that had been previously annealed by heating at 98°C for 3 min and then slowly cooling to 16°C in a thermocycler. After further purification using the MinElute purification kit (QIAGEN), 1 µl of each reaction was quantified by labelling with biotin, spotted on nitrocellulose after a serial dilution, and detected using an anti-biotin-AP conjugate (Roche Diagnostics, Monza, Italy) following manufacturer's instructions. Equal amounts of DNA from samples were pooled together and size fractionated by 2% MS-6 agarose (Conda, Madrid, Spain) gel electrophoresis in TAE buffer at 120 V for 60 min. Gel slices containing DNA in the 400 to 600 bp estimated range were cut and purified using QIAquick gel extraction kit (QIAGEN) and used for sample preparation according to the protocol for genomic DNA sequencing using the Illumina Genome Analyser IIx (Illumina, USA). The samples were run at the Istituto di Genomica Applicata (Udine, Italy).
The Illumina sequencing provided complessively nearly 10 millions 100 nt reads from the genomic DNA of the three Psa strains that passed the quality check. This amount of sequence represents approximately 27.7, 67.6 and 61.2 X coverage for J-Psa, I-Psa and I2-Psa, respectively, i.e. the expected genomic size of these strains, based on the previously sequenced genomes of the P. syringae pathovars is 6 Mb.
Whole-genome assembly and alignment of Illumina genomes
Paired reads of 100 nts were assembled into contigs using the de novo (i.e. without using a reference genome) assembly option of the CLC genomic workbench (CLC-bio, Aarhus, Denmark). Contigs sequences were scanned for ORFs by GLIMMER, version 3.02.  which had been previously trained on the complete genome sequences of Pseudomonas syringae pv. tomato strain DC3000 (NC_004578.1, i.e. Pto DC3000), P. s. pv. phaseolicola strain 1448A (NC_005773.3, i.e. Pph 1448A), and P. s. pv. syringae strain B728a (NC_007005.1, i.e. Psy B728a). The putative proteins were annotated against the RefSeq database using a perl script for recursive blastx searches. Additional genome sequence analyses was performed with the aid of the software packages mummer  and mauve . Several ad hoc PERL scripts were developed to assist the comparison of genome sequence draft and its putative protein complement with respect to each one of the three Psa strains and Pto DC3000, Pph 1449A and Psy B728a. Additional assessment of the effector genes repertoire was performed upon the paper of Baltrus et al. 
Protein categorization and searching for variable regions in Psa genomes
For each strain, all predicted proteins were ordered by role categories according to TIGRFAMs. In addition, for each Psa strain the presence along the genomes of variable regions was also analysed. Such genomic regions were identified by the following criteria using MAUVE and an ad hoc script PERL “island finder” a) DNA region larger than 10 kb in a contig that appeared as a gap in the genome alignment and b) different G+C content with respect to the average content of the three Psa strains, which was performed using the tool: http://www.sciencebuddies.org/science-fair-projects/projectideas/GenomGCcalculator. These regions were retained as good candidates to be the results of horizontal gene transfer .
All the sequences used in the MLST analysis to build the phylogenetic trees were edited and aligned using ClustalW 1.83 (http://www.ebi.ac.uk/tools/clustalw2/) and concatenated using Geneious 5.1.4 (http://www.geneious.com/). The phylogenetic tree built using the neighbor-joining (NJ) algorithm was obtained using the Splits-Tree software  and the Hamming distances method using three housekeeping gene fragments, namely gyrB, rpoB and rpo, for a total of 1,646 bp. Bootstrap analysis was performed using the same software. The program TOPALi version 2.5, available at (http://www.topali.org/) , was used to determine the best model of evolution for the second phylogenetic tree based on maximum likelihood (ML) algorithm using six housekeeping fragment genes, acnB, fruK, gltA, pgi, rpoB and rpoD, for a total of 2,926 bp. The PhyML  method was used to determine the best model of evolution for the ML analysis. Both the hierarchical likelihood ratio test (hLRT) and the standard Akaike Information Criterion (AIC) were used to evaluate the model scores. Bootstrap analysis was performed using the same software. The hypotheses about the genealogy of the Psa strains were tested using the likelihood-ratio test for monophyly which was developed by Huelsenbeck et al. . Likelihoods were estimated using the Phangorn module  of the statistical package R (R development core team, 2007). The significance of the likelihood ratio was estimated by parametric bootstrap according to Huelsenbeck et al.  by simulation of 1,000 replicated datasets generated with Indel-Seq-Gen .
Analysis of plasmid content
Plasmid isolation was performed using the PureYieldTM Plasmid Miniprep System (Promega, Madison, WI, U.S.A.) protocol. The strains submitted for plasmid detection were the object of this study as well as Psa KW30 and Psa KW31, which were isolated in 1984 in Japan from leaf spot lesions of A. deliciosa cv. Hayward  and additional 11 Psa strains isolated in different regions of Italy during current epidemics , , . Plasmids from Escherichia coli strain 39R861 were used as molecular weight marker .
Multiplication of Psa in Actinidia spp. leaves and pathogencity test on tomato
To compare the capability of infection in different Actinidia spp., we artificially inoculated both A. chinensis and A. deliciosa leaves with all three of the sequenced Psa strains. For inoculation, bacteria were grown for 48 h on NSA, at 25±1°C, and the plants were covered 24 h before the inoculation with a plastic bag. Leaf areas of approximately 1 cm in diameter on one-year-old potted A. deliciosa cv Hayward and A. chinensis cv Hort16A plants were inoculated using a needleless sterile syringe with a bacterial suspension in sterile saline (0,85% NaCl in distilled water) at the concentrations of 1–2×103 and 1–2×106 cfu/mL. For each thesis, 10 leaves were inoculated in four sites. Control plants were treated using solely sterile saline. To determine bacterial growth in planta, leaf disks of about 0,5 cm of diameter were sampled from each species and inoculation site at regular intervals and ground in 1 mL of sterile saline, and serial ten-fold dilutions were spotted onto NSA medium. Colonies were counted two days after incubation at 25±1°C. According to a comparative study on host-specific virulence factors and effector genes , it has been observed that the contemporary presence of the effector genes hopAA1-2 and hopA1 in single strains is quite rare. Only three Pto strains displayed such a combination. To verify that Psa, in particular I2-Psa showing such effector genes, can multiply and also infect tomato plants, we performed a pathogenicity test by inoculating potted-seedlings of Lycopersicon esculentum (tomato) cv Lancelot. The inoculation was performed as described above and Pto DC3000 was used as positive control. Symptoms caused by inoculation of the three Psa strains were observed and compared with those caused by Pto DC3000.
Evolutionary relationships of Psa strains to other phytopathogenic pseudomonads. Phylogenetic relationships were estimated from concatenated sequences from six housekeeping genes, acnB, fruK, gltA, pgi, rpoB and rpoD (2,926 bp), using the maximum likelihood (ML) algorithm. Bootstrap values are reported at each branching. P. putida was used as outgroup.
Presence and absence of phaseolotoxin in Psa strains. Diagrammatic representation of the phaseolotoxin gene cluster, argK-tox, and the flanking regions. The phaseolotoxin cluster is present in J-Psa and I-Psa (upper part) but not in I2-Psa (lower part).
Variable regions (VR) found in the draft genome of I2-Psa compared with J-Psa and I-Psa draft genomes.
Categorization according to TIGRFAMs of protein complement displayed by the draft genome of I2-Psa.
We thank Robert W. Jackson of School of Biological Sciences, Reading, United Kingdom for having supplied the E. coli strain 39R861.
Conceived and designed the experiments: MS GF. Performed the experiments: SM GF PF MP MS. Analyzed the data: GF SM MP PF MS. Contributed reagents/materials/analysis tools: GF SM MP PF MS. Wrote the paper: MS GF.
- 1. Amato P, Parazols M, Sancelme M, Laj P, Mailhot G, et al. (2007) Microorganisms isolated from the water phase of trophospheric clouds at the Puy de Dome: major groups and growth abilities at low temperatures. FEMS Microbiol Ecol 59: 242–254.P. AmatoM. ParazolsM. SancelmeP. LajG. Mailhot2007Microorganisms isolated from the water phase of trophospheric clouds at the Puy de Dome: major groups and growth abilities at low temperatures.FEMS Microbiol Ecol59242254
- 2. Morris CE, Sands DC, Vinatzer BA, Glaux C, Guilbaud C, et al. (2008) The life history of the plant pathogen Pseudomonas syringae is linked to the water cycle. ISME J 2: 321–334.CE MorrisDC SandsBA VinatzerC. GlauxC. Guilbaud2008The life history of the plant pathogen Pseudomonas syringae is linked to the water cycle.ISME J2321334
- 3. Stavrinides J, Mc Closkey JK, Ochman H (2009) Pea aphid as both host and vector for the phytopathogenic bacterium Pseudomonas syringae. Appl Env Microbiol 75: 2230–2235.J. StavrinidesJK Mc CloskeyH. Ochman2009Pea aphid as both host and vector for the phytopathogenic bacterium Pseudomonas syringae.Appl Env Microbiol7522302235
- 4. Seshu Kumar G, Jagannadham MV, Ray MK (2002) Low-temperature-induced changes in composition and fluidity of lipopolysaccharides in the Antarctic psychrotrophic bacterium Pseudomonas syringae. J Bacteriol 184: 6746–6749.G. Seshu KumarMV JagannadhamMK Ray2002Low-temperature-induced changes in composition and fluidity of lipopolysaccharides in the Antarctic psychrotrophic bacterium Pseudomonas syringae.J Bacteriol18467466749
- 5. Agrios GN (2005) Plant Pathology 5th edition, Academic Press. 952 p.GN Agrios2005Plant Pathology 5th edition, Academic Press952
- 6. Bull CT, De Boer SH, Denny TP, Firrao G, Fischer-Le Saux M, et al. (2010) Comprehensive list of names of plant pathogenic bacteria, 1980-2007. J Plant Pathol 92: 551–592.CT BullSH De BoerTP DennyG. FirraoM. Fischer-Le Saux2010Comprehensive list of names of plant pathogenic bacteria, 1980-2007.J Plant Pathol92551592
- 7. Gardan L, Shafik H, Belouin S, Broch R, Grimont F, et al. (1999) DNA relatedness among the pathovars of Pseudomonas syringae and description of Pseudomonas tremae sp. nov. and Pseudomonas cannabina sp. nov. (ex Sutic and Dowson 1959). Int J Syst Bacteriol 49: 469–478.L. GardanH. ShafikS. BelouinR. BrochF. Grimont1999DNA relatedness among the pathovars of Pseudomonas syringae and description of Pseudomonas tremae sp. nov. and Pseudomonas cannabina sp. nov. (ex Sutic and Dowson 1959).Int J Syst Bacteriol49469478
- 8. Scortichini M, Marchesi U, Di Prospero P (2002) Genetic relatedness among Pseudomonas avellanae, P. syringae pv. theae and P. s. pv. actinidiae, and their identification. Eur J Plant Pathol 108: 269–278.M. ScortichiniU. MarchesiP. Di Prospero2002Genetic relatedness among Pseudomonas avellanae, P. syringae pv. theae and P. s. pv. actinidiae, and their identification.Eur J Plant Pathol108269278
- 9. Manceau C, Brin C (2003) Pathovars of Pseudomonas syringae are structured in genetic populations allowing the selection of specific markers for their detection in plant samples. In: Iacobellis NS, Collmer A, Hutcheson SW, Mansfield JW, Morris CE, Murillo J, Schaad NW, Stead DE, Surico G, Ulrich MS, editors. Pseudomonas syringae and related pathogens. Kluwer Academic Publishers. pp. 503–512.C. ManceauC. Brin2003Pathovars of Pseudomonas syringae are structured in genetic populations allowing the selection of specific markers for their detection in plant samples.NS IacobellisA. CollmerSW HutchesonJW MansfieldCE MorrisJ. MurilloNW SchaadDE SteadG. SuricoMS UlrichPseudomonas syringae and related pathogensKluwer Academic Publishers503512
- 10. Takikawa Y, Serizawa S, Ichikawa T, Tsuyumu S, Goto M (1989) Pseudomonas syringae pv. actinidiae pv. nov.: the causal bacterium of canker of kiwifruit in Japan. Ann Phytopathol Soc Japan 55: 437–444.Y. TakikawaS. SerizawaT. IchikawaS. TsuyumuM. Goto1989Pseudomonas syringae pv. actinidiae pv. nov.: the causal bacterium of canker of kiwifruit in Japan.Ann Phytopathol Soc Japan55437444
- 11. Koh JK, Cha BJ, Chung HJ, Lee DH (1994) Outbreak and spread of bacterial canker in kiwifruit. Korean J Plant Pathol 10: 68–72.JK KohBJ ChaHJ ChungDH Lee1994Outbreak and spread of bacterial canker in kiwifruit.Korean J Plant Pathol106872
- 12. Scortichini M (1994) Occurrence of Pseudomonas syringae pv. actinidiae on kiwifruit in Italy. Plant Pathol 43: 1035–1038.M. Scortichini1994Occurrence of Pseudomonas syringae pv. actinidiae on kiwifruit in Italy.Plant Pathol4310351038
- 13. Koh JK, Lung JS, Hur JS (2003) Current status of occurrence of major diseases on kiwifruit and their control in Korea. Acta Horticulturae 610: 437–443.JK KohJS LungJS Hur2003Current status of occurrence of major diseases on kiwifruit and their control in Korea.Acta Horticulturae610437443
- 14. Ferrante P, Scortichini M (2009) Identification of Pseudomonas syringae pv. actinidiae as causal agent of bacterial canker of yellow kiwifruit (Actinidia chinensis Planchon) in central Italy. J Phytopathol 157: 768–770.P. FerranteM. Scortichini2009Identification of Pseudomonas syringae pv. actinidiae as causal agent of bacterial canker of yellow kiwifruit (Actinidia chinensis Planchon) in central Italy.J Phytopathol157768770
- 15. Liang Y, Zhang X, Tian C, Gao A, Wang P (2000) Pathogenic identification of kiwifruit bacterial canker in Shaanxi. J Northwest Forestry College. Y. LiangX. ZhangC. TianA. GaoP. Wang2000Pathogenic identification of kiwifruit bacterial canker in Shaanxi.J Northwest Forestry Collegeunpaginated. http://en.cnki.com.cn/Article_en/CJFDTOTAL-XBLX200001006.htm. unpaginated. http://en.cnki.com.cn/Article_en/CJFDTOTAL-XBLX200001006.htm.
- 16. Li Y, Cheng H, Fang S, Qian Z (2001) Y. LiH. ChengS. FangZ. Qian2001Ecological factors affecting prevalence of kiwifruit bacterial canker and bacteriostatic action of bacteriocides on Pseudomonas syringae pv. actinidiae. Chinese J. Appl. Ecology, unpaginated. http://en.cnki.com.cn/Article_en/CJFDTOTAL-YYSB200103012.htm. Ecological factors affecting prevalence of kiwifruit bacterial canker and bacteriostatic action of bacteriocides on Pseudomonas syringae pv. actinidiae. Chinese J. Appl. Ecology, unpaginated. http://en.cnki.com.cn/Article_en/CJFDTOTAL-YYSB200103012.htm.
- 17. Balestra GM, Renzi M, Mazzaglia A (2010) First report of bacterial canker of Actinidia deliciosa caused by Pseudomonas syringae pv. actinidiae in Portugal. New Disease Reports 22: 10 [doi. GM BalestraM. RenziA. Mazzaglia2010First report of bacterial canker of Actinidia deliciosa caused by Pseudomonas syringae pv.actinidiae in Portugal. New Disease Reports 2210 [doi10.5197/j.2044-0588.2010.022.010]. 10.5197/j.2044-0588.2010.022.010].
- 18. European Plant Protection Organization (2011) First report of Pseudomonas syringae pv. actinidiae in Chile. EPPO Reporting Service, n°3, 2011/055. European Plant Protection Organization2011First report of Pseudomonas syringae pv. actinidiae in Chile.EPPO Reporting Service, n°3, 2011/055
- 19. Ferrante P, Scortichini M (2010) Molecular and phenotypic features of Pseudomonas syringae pv. actinidiae isolated during recent epidemics of bacterial canker on yellow kiwifruit (Actinidia chinensis) in central Italy. Plant Pathol 69: 954–962.P. FerranteM. Scortichini2010Molecular and phenotypic features of Pseudomonas syringae pv. actinidiae isolated during recent epidemics of bacterial canker on yellow kiwifruit (Actinidia chinensis) in central Italy.Plant Pathol69954962
- 20. Wang Z, Tang X, Liu S (1992) Z. WangX. TangS. Liu1992Identification of the pathogenic bacterium for bacterial canker on Actinidia in Sichuan. J. Southwest Agricultural University, unpaginated. http://en.cnki.com.cn/Article_en/CJFDTOTAL-XNND199206007.htm. Identification of the pathogenic bacterium for bacterial canker on Actinidia in Sichuan. J. Southwest Agricultural University, unpaginated. http://en.cnki.com.cn/Article_en/CJFDTOTAL-XNND199206007.htm.
- 21. Li M, Tan G, Li Y, Cheng H, Han X, et al. (2004) Resistance of different Chinese gooseberry cultivars to Chinese gooseberry bacterial canker caused by Pseudomonas syringae pv. actinidiae and their cluster analysis. Plant Protection 30: 51–54.M. LiG. TanY. LiH. ChengX. Han2004Resistance of different Chinese gooseberry cultivars to Chinese gooseberry bacterial canker caused by Pseudomonas syringae pv. actinidiae and their cluster analysis.Plant Protection305154
- 22. Koh YJ, Kim GH, Jung JS, Lee YS, Hur JS (2010) Outbreak of bacterial canker on Hort16A (Actinidia chinensis Planchon) caused by Pseudomonas syringae pv. actinidiae in Korea. NZ J Crop Hort Sci 38: 275–282.YJ KohGH KimJS JungYS LeeJS Hur2010Outbreak of bacterial canker on Hort16A (Actinidia chinensis Planchon) caused by Pseudomonas syringae pv. actinidiae in Korea.NZ J Crop Hort Sci38275282
- 23. Ushiyama K, Kita N, Suyama K, Aono N, Ogawa J, et al. (1992) Bacterial canker disease of wild Actinidia plants as the infection source of outbreak of bacterial canker of kiwifruit caused by Pseudomonas syringae pv. actinidiae. Ann Phytopathol Soc Japan 58: 426–430.K. UshiyamaN. KitaK. SuyamaN. AonoJ. Ogawa1992Bacterial canker disease of wild Actinidia plants as the infection source of outbreak of bacterial canker of kiwifruit caused by Pseudomonas syringae pv. actinidiae.Ann Phytopathol Soc Japan58426430
- 24. Ushijama K, Suyama K, Kita N, Aono N, Fujii H (1992) Isolation of kiwifruit canker pathogen, Pseudomonas syringae pv. actinidiae from leaf spot of tara vine (Actinidia arguta Planch). Ann Phytopathol Soc Japan 58: 476–479.K. UshijamaK. SuyamaN. KitaN. AonoH. Fujii1992Isolation of kiwifruit canker pathogen, Pseudomonas syringae pv. actinidiae from leaf spot of tara vine (Actinidia arguta Planch).Ann Phytopathol Soc Japan58476479
- 25. Marcelleti S, Scortichini M (2011) Clonal outbreaks of bacterial canker caused by Pseudomonas syringae pv. actinidiae on Actinidia chinensis and A. deliciosa in Italy. J Plant Pathol 93: 479–483.S. MarcelletiM. Scortichini2011Clonal outbreaks of bacterial canker caused by Pseudomonas syringae pv. actinidiae on Actinidia chinensis and A. deliciosa in Italy.J Plant Pathol93479483
- 26. European Plant Protection Organization (2010) First record of Pseudomonas syringae pv. actinidiae in France. EPPO Reporting Service 2010/188. European Plant Protection Organization2010First record of Pseudomonas syringae pv. actinidiae in France.EPPO Reporting Service 2010/188
- 27. Chapman J, Taylor R, Alexander B (2011) Second report on characterisation of Pseudomonas syringae pv. actinidiae (Psa) isolates in New Zealand. Ministry of Agriculture and Forestry 10: J. ChapmanR. TaylorB. Alexander2011Second report on characterisation of Pseudomonas syringae pv. actinidiae (Psa) isolates in New Zealand.Ministry of Agriculture and Forestry10
- 28. Tamura K, Takikawa Y, Tsuyumu S, Goto M (1989) Characterization of the toxin produced by Pseudomonas syringae pv. actinidiae, the causal bacterium of kiwifruit canker. Ann Phytopathol Soc Japan 55: 512.K. TamuraY. TakikawaS. TsuyumuM. Goto1989Characterization of the toxin produced by Pseudomonas syringae pv. actinidiae, the causal bacterium of kiwifruit canker.Ann Phytopathol Soc Japan55512
- 29. Tamura K, Imamura M, Yoneyama K, Kohno Y, Takikawa Y, et al. (2002) Role of phaseolotoxin production by Pseudomonas syringae pv. actinidiae in the formation of halo lesions of kiwifruit canker disease. Physiol Mol Plant Pathol 60: 207–214.K. TamuraM. ImamuraK. YoneyamaY. KohnoY. Takikawa2002Role of phaseolotoxin production by Pseudomonas syringae pv. actinidiae in the formation of halo lesions of kiwifruit canker disease.Physiol Mol Plant Pathol60207214
- 30. Sawada H, Takeuchi T, Matsuda I (1997) Comparative analysis of Pseudomonas syringae pv. actinidiae and pv. phaseolicola based on phaseolotoxin-resistant ornithine carbamoyltransferase gene (argK) and 16S-23S rRNA intergenic spacer sequences. Appl Environ Microbiol 63: 282–288.H. SawadaT. TakeuchiI. Matsuda1997Comparative analysis of Pseudomonas syringae pv. actinidiae and pv. phaseolicola based on phaseolotoxin-resistant ornithine carbamoyltransferase gene (argK) and 16S-23S rRNA intergenic spacer sequences.Appl Environ Microbiol63282288
- 31. Sawada H, Suzuki F, Matsuda I, Saitou N (1999) Phylogenetic analysis of Pseudomonas syringae pathovars suggests the horizontal gene transfer of argK and the evolutionary stability of hrp gene cluster. J Mol Evol 49: 627–644.H. SawadaF. SuzukiI. MatsudaN. Saitou1999Phylogenetic analysis of Pseudomonas syringae pathovars suggests the horizontal gene transfer of argK and the evolutionary stability of hrp gene cluster.J Mol Evol49627644
- 32. Ahmed N (2009) A flood of microbial genomics-do we need for more ? PLoS One 4: e5831.N. Ahmed2009A flood of microbial genomics-do we need for more ?PLoS One4e5831
- 33. Buell CR, Joardar V, Lindeberg M, Selengut J, Paulsen IT, et al. (2003) The complete genome sequence of the Arabidopsis and tomato pathogen Pseudomonas syringae pv. tomato DC3000. Proc Natl Acad Sci U S A 100: 10181–10186.CR BuellV. JoardarM. LindebergJ. SelengutIT Paulsen2003The complete genome sequence of the Arabidopsis and tomato pathogen Pseudomonas syringae pv. tomato DC3000.Proc Natl Acad Sci U S A1001018110186
- 34. Almeida NF, Yan S, Lindeberg M, Studholme DJ, Schneider DJ, et al. (2009) A draft genome sequences of Pseudomonas syringae pv. tomato T1reveals a type III effector repertoire significantly divergent from that of Pseudomonas syringae pv. tomato DC3000. Mol Plant Microbe Interact 22: 52–62.NF AlmeidaS. YanM. LindebergDJ StudholmeDJ Schneider2009A draft genome sequences of Pseudomonas syringae pv. tomato T1reveals a type III effector repertoire significantly divergent from that of Pseudomonas syringae pv. tomato DC3000.Mol Plant Microbe Interact225262
- 35. Joardar V, Lindeberg M, Jackson RW, Selengut J, Dodson R, et al. (2005) Whole-genome sequence analysis of Pseudomonas syringae pv. phaseolicola 1448A reveals divergence among pathovars in genes involved in virulence and transposition. J Bacteriol 187: 6488–6498.V. JoardarM. LindebergRW JacksonJ. SelengutR. Dodson2005Whole-genome sequence analysis of Pseudomonas syringae pv. phaseolicola 1448A reveals divergence among pathovars in genes involved in virulence and transposition.J Bacteriol18764886498
- 36. Feil H, Feil WS, Chain P, Larimer F, Di Bartolo G, et al. (2005) Comparison of the complete genome sequences of Pseudomonas syringae pv. syringae B728a and pv. tomato DC3000. Proc Natl Acad Sci U S A 102: 11064–11069.H. FeilWS FeilP. ChainF. LarimerG. Di Bartolo2005Comparison of the complete genome sequences of Pseudomonas syringae pv. syringae B728a and pv. tomato DC3000.Proc Natl Acad Sci U S A1021106411069
- 37. Reinhardt JA, Baltrus DA, Nishimura MT, Jeck WR, Jones CD, et al. (2009) De novo assembly using low-coverage short read sequence data from the rice pathogen Pseudomonas syringae pv. oryzae. Genome Res 19: 294–305.JA ReinhardtDA BaltrusMT NishimuraWR JeckCD Jones2009De novo assembly using low-coverage short read sequence data from the rice pathogen Pseudomonas syringae pv. oryzae.Genome Res19294305
- 38. Studholme DJ, Gimenez Ibanez S, Maclean D, Dangl JL, Chang JH, et al. (2009) A draft genome sequence and functional screen reveals the repertoire of type III secreted proteins of Pseudomonas syringae pathovar tabaci 11528. BMC Genomics 10: 395.DJ StudholmeS. Gimenez IbanezD. MacleanJL DanglJH Chang2009A draft genome sequence and functional screen reveals the repertoire of type III secreted proteins of Pseudomonas syringae pathovar tabaci 11528.BMC Genomics10395
- 39. Green S, Studholme DJ, Laue BE, Dorati F, Lovell H, et al. (2010) Comparative genome analysis provides insights into the evolution and adaptation of Pseudomonas syringae pv. aesculi on Aesculus hippocastanum. PLoS One 5: e10224.S. GreenDJ StudholmeBE LaueF. DoratiH. Lovell2010Comparative genome analysis provides insights into the evolution and adaptation of Pseudomonas syringae pv. aesculi on Aesculus hippocastanum.PLoS One5e10224
- 40. Rodriguez-Palenzuela P, Matas IM, Murillo J, Lopez-Solanilla E, Bardaji L, et al. (2010) Annotation and overview of the Pseudomonas savastanoi pv. savastanoi NCPPB 3335 draft genome reveals the virulence gene complement of a tumour-inducing pathogen of woody host. Environ Microbiol 12: 1604–1620.P. Rodriguez-PalenzuelaIM MatasJ. MurilloE. Lopez-SolanillaL. Bardaji2010Annotation and overview of the Pseudomonas savastanoi pv. savastanoi NCPPB 3335 draft genome reveals the virulence gene complement of a tumour-inducing pathogen of woody host.Environ Microbiol1216041620
- 41. Qi M, Wang D, Bradley CA, Zhao Y (2011) Genome sequence analysis of Pseudomonas savastanoi pv. glycinea and subtractive hybridization-based comparative genomics with nine pseudomonads. PLoS One 6: 16451.M. QiD. WangCA BradleyY. Zhao2011Genome sequence analysis of Pseudomonas savastanoi pv. glycinea and subtractive hybridization-based comparative genomics with nine pseudomonads.PLoS One616451
- 42. Cunnac S, Lindeberg M, Collmer A (2009) Pseudomonas syringae type III secretion system effectors: repertoires in search of functions. Curr Opin Microbiol 12: 53–60.S. CunnacM. LindebergA. Collmer2009Pseudomonas syringae type III secretion system effectors: repertoires in search of functions.Curr Opin Microbiol125360
- 43. Bronstein PA, Marrichi M, Cartinhour S, Schneider DJ, De Lisa MP (2005) Identification of a twin-arginine translocation system in Pseudomonas syringae pv. tomato DC3000 and its contribution to pathogenicity and fitness. J Bacteriol 187: 8450–8461.PA BronsteinM. MarrichiS. CartinhourDJ SchneiderMP De Lisa2005Identification of a twin-arginine translocation system in Pseudomonas syringae pv. tomato DC3000 and its contribution to pathogenicity and fitness.J Bacteriol18784508461
- 44. Grant SR, Fisher EJ, Chang JH, Mole BM, Dangl JL (2006) Subterfuge and manipulation: type III effector proteins of phytopathogenic bacteria. Ann Rev. Microbiol 60: 425–449.SR GrantEJ FisherJH ChangBM MoleJL Dangl2006Subterfuge and manipulation: type III effector proteins of phytopathogenic bacteria. Ann Rev.Microbiol60425449
- 45. Aschtgen M-S, Gavioli M, Dessen A, Lloubes L, Cascalés E (2010) The SciZ protein anchors the enteroaggregative Escherichia coli type VI secretion system to the cell wall. Mol Microbiol 75: 886–899.M-S AschtgenM. GavioliA. DessenL. LloubesE. Cascalés2010The SciZ protein anchors the enteroaggregative Escherichia coli type VI secretion system to the cell wall.Mol Microbiol75886899
- 46. Lindeberg M, Cartinhour S, Myers CR, Schechter LM, Schneider DJ, et al. (2006) Closing the circle on the discovery of genes encoding Hrp regulon members and type III secretion system effectors in the genomes of three model Pseudomonas syringae strains. Mol Plant Microbe Interact 19: 1151–1158.M. LindebergS. CartinhourCR MyersLM SchechterDJ Schneider2006Closing the circle on the discovery of genes encoding Hrp regulon members and type III secretion system effectors in the genomes of three model Pseudomonas syringae strains.Mol Plant Microbe Interact1911511158
- 47. Alfano JR, Kim HS, Delaney TP, Collmer A (1997) Evidence that the Pseudomonas syringae pv. syringae hrp-linked hrmA gene encodes an Avr-like protein that acts in an hrp-dependent manner within tobacco cells. Mol Plant Microbe Interact 10: 580–588.JR AlfanoHS KimTP DelaneyA. Collmer1997Evidence that the Pseudomonas syringae pv. syringae hrp-linked hrmA gene encodes an Avr-like protein that acts in an hrp-dependent manner within tobacco cells.Mol Plant Microbe Interact10580588
- 48. Gassmann W (2005) Natural variation in the Arabidopsis response to the avirulence gene hopPsyA uncouples the hypersensitive response from disease resistance. Mol Plant Microbe Interact 18: 1054–1060.W. Gassmann2005Natural variation in the Arabidopsis response to the avirulence gene hopPsyA uncouples the hypersensitive response from disease resistance.Mol Plant Microbe Interact1810541060
- 49. Kim MG, Geng X, Lee SY, Mackey D (2009) The Pseudomonas syringae type III effector AvrRpm1 induces significant defenses by activating the Arabidopsis nucleotide-binding leucine-rich repeat protein RPS2. Plant J 57: 645–653.MG KimX. GengSY LeeD. Mackey2009The Pseudomonas syringae type III effector AvrRpm1 induces significant defenses by activating the Arabidopsis nucleotide-binding leucine-rich repeat protein RPS2.Plant J57645653
- 50. Rohmer L, Guttman DS, Dangl JL (2004) Diverse evolutionary mechanisms shape the type III effector virulence factor repertoire in the plant pathogen Pseudomonas syringae. Genetics 167: 1341–1360.L. RohmerDS GuttmanJL Dangl2004Diverse evolutionary mechanisms shape the type III effector virulence factor repertoire in the plant pathogen Pseudomonas syringae.Genetics16713411360
- 51. Munkvold KR, Russell AB, Kvitko BH, Collmer A (2009) Pseudomonas syringae pv. tomato DC3000 type III effector HopAA1-1 functions redundantly with chlorosis-promoting factor PSPTO4223 to produce bacterial speck lesions in host tomato. Mol Plant Microbe Interact 22: 1341–1355.KR MunkvoldAB RussellBH KvitkoA. Collmer2009Pseudomonas syringae pv. tomato DC3000 type III effector HopAA1-1 functions redundantly with chlorosis-promoting factor PSPTO4223 to produce bacterial speck lesions in host tomato.Mol Plant Microbe Interact2213411355
- 52. Sarkar S, Gordon J, Martin G, Guttman D (2006) Comparative genomics of host-specific virulence in Pseudomonas syringae. Genetics 174: 1041–1056.S. SarkarJ. GordonG. MartinD. Guttman2006Comparative genomics of host-specific virulence in Pseudomonas syringae.Genetics17410411056
- 53. Wei CF, Kvitko BH, Shimizu R, Crabill E, Alfano JR, et al. (2007) A Pseudomonas syringae pv. tomato DC3000 mutant lacking the type III effector HopQ1-1 is able to cause disease in the model plant Nicotiana benthamiana. Plant J 51: 32–46.CF WeiBH KvitkoR. ShimizuE. CrabillJR Alfano2007A Pseudomonas syringae pv. tomato DC3000 mutant lacking the type III effector HopQ1-1 is able to cause disease in the model plant Nicotiana benthamiana.Plant J513246
- 54. Genka H, Baba T, Tsuda M, Kanaya S, Mori H, et al. (2006) Comparative analysis of argK-tox clusters and their flanking regions in phaseolotoxin-producing Pseudomonas syringae pathovars. J Mol Evol 63: 401–414.H. GenkaT. BabaM. TsudaS. KanayaH. Mori2006Comparative analysis of argK-tox clusters and their flanking regions in phaseolotoxin-producing Pseudomonas syringae pathovars.J Mol Evol63401414
- 55. Sawada H, Kanaya S, Tsuda M, Suzuki F, Azegami K, et al. (2002) A phylogenetic study of the OCTase genes in Pseudomonas syringae pathovars suggests the horizontal gene transfer of argK and the evolutionary stability of hrp gene cluster. J Mol Evol 49: 627–644.H. SawadaS. KanayaM. TsudaF. SuzukiK. Azegami2002A phylogenetic study of the OCTase genes in Pseudomonas syringae pathovars suggests the horizontal gene transfer of argK and the evolutionary stability of hrp gene cluster.J Mol Evol49627644
- 56. Cooksey DA (1993) Molecular mechanism of copper resistance and accumulation in bacteria. FEMS Microbiol Rev 14: 381–386.DA Cooksey1993Molecular mechanism of copper resistance and accumulation in bacteria.FEMS Microbiol Rev14381386
- 57. Nakajima M, Goto M, Hibi T (2002) Similarity between copper resistance genes from Pseudomonas syringae pv. actinidiae and P. syringae pv. tomato. J Gen Plant Pathol 68: 68–74.M. NakajimaM. GotoT. Hibi2002Similarity between copper resistance genes from Pseudomonas syringae pv. actinidiae and P. syringae pv. tomato.J Gen Plant Pathol686874
- 58. Nikaido H (2009) Multidrug resistance in bacteria. Ann Rev Biochem 78: 119–146.H. Nikaido2009Multidrug resistance in bacteria.Ann Rev Biochem78119146
- 59. Pollard AM, Scherbinske SE, Nichols WA (2007) The tonB gene of Haemophilus parainfluenzae demonstrates strong sequence identity with that of Haemophilus influenzae. J Appl Res 7: 32–38.AM PollardSE ScherbinskeWA Nichols2007The tonB gene of Haemophilus parainfluenzae demonstrates strong sequence identity with that of Haemophilus influenzae.J Appl Res73238
- 60. Papi R, Kyriakidis D (2003) Overexpression of the pectin lyase gene of Pseudomonas marginalis in Escherichia coli and purification of the active enzyme. Biotech Appl Biochem 37: 187–194.R, Kyriakidis D. Papi2003Overexpression of the pectin lyase gene of Pseudomonas marginalis in Escherichia coli and purification of the active enzyme.Biotech Appl Biochem37187194
- 61. Arnold DL, Jackson RW, Waterfield NR, Mansfield JW (2007) Evolution of microbial virulence: the benefits of stress. Trends Genet 23: 293–300.DL ArnoldRW JacksonNR WaterfieldJW Mansfield2007Evolution of microbial virulence: the benefits of stress.Trends Genet23293300
- 62. Jackson RW, Johnson LJ, Clarke SR, Arnold DL (2011) Bacterial pathogen evolution: breaking news. Trends Genet 27: 32–40.RW JacksonLJ JohnsonSR ClarkeDL Arnold2011Bacterial pathogen evolution: breaking news.Trends Genet273240
- 63. Sesma A, Sundin GW, Murillo J (2000) Phylogeny of the replication regions of pPT23A-like plasmids from Pseudomonas syringae. Microbiol 146: 2375–2384.A. SesmaGW SundinJ. Murillo2000Phylogeny of the replication regions of pPT23A-like plasmids from Pseudomonas syringae.Microbiol14623752384
- 64. Bender CL, Alarcon-Chaidez F, Gross DC (1999) Pseudomonas syringae phytotoxins: mode of action, regulation and byosinthesis by peptide and polyketide synthetase. Microbiol Mol Biol Rev 63: 266–292.CL BenderF. Alarcon-ChaidezDC Gross1999Pseudomonas syringae phytotoxins: mode of action, regulation and byosinthesis by peptide and polyketide synthetase.Microbiol Mol Biol Rev63266292
- 65. Tourte C, Manceau C (1995) A strain of Pseudomonas syringae which does not belong to pathovar phaseolicola produces phaseolotoxin. Eur J Plant Pathol 101: 483–490.C. TourteC. Manceau1995A strain of Pseudomonas syringae which does not belong to pathovar phaseolicola produces phaseolotoxin.Eur J Plant Pathol101483490
- 66. Murillo J, Bardaji L, Navarro de la Fuente L, Fuhrer ME, Aguilera S, et al. (2011) Variation in conservation of the cluster for biosynthesis of the phytotoxin phaseolotoxin in Pseudomonas syringae suggests at least two events of horizontal acquisition. Res Microbiol 162: 253–261.J. MurilloL. BardajiL. Navarro de la FuenteME FuhrerS. Aguilera2011Variation in conservation of the cluster for biosynthesis of the phytotoxin phaseolotoxin in Pseudomonas syringae suggests at least two events of horizontal acquisition.Res Microbiol162253261
- 67. Rico A, Lopez R, Asensio C, Aizpun M, Asensio-S Manzanera C, et al. (2003) Nontoxigenic strains of Pseudomonas syringae pv. phaseolicola are a main cause of halo blight of bean in Spain and escape current detection methods. Phytopathol 93: 1553–1559.A. RicoR. LopezC. AsensioM. AizpunC. Asensio-S Manzanera2003Nontoxigenic strains of Pseudomonas syringae pv. phaseolicola are a main cause of halo blight of bean in Spain and escape current detection methods.Phytopathol9315531559
- 68. Baltrus DA, Nishimura MT, Romanchuk A, Chang JH, Mukhtar MS, et al. (2011) Dynamic evolution of pathogenicity revealed by sequencing and comparative genomics of 19 Pseudomonas syringae isolates. PLoS Pathogens 7: e1002132.DA BaltrusMT NishimuraA. RomanchukJH ChangMS Mukhtar2011Dynamic evolution of pathogenicity revealed by sequencing and comparative genomics of 19 Pseudomonas syringae isolates.PLoS Pathogens7e1002132
- 69. Sarkar S, Gordon J, Martin G, Guttman D (2006) Comparative genomics of host-specific virulence in Pseudomonas syringae. Genetics 174: 1041–1056.S. SarkarJ. GordonG. MartinD. Guttman2006Comparative genomics of host-specific virulence in Pseudomonas syringae.Genetics17410411056
- 70. Vanneste JL, Cornish DA, Yu J, Audusseau C, Paillard S, et al. (2011) Presence of the effector gene hopA1 in strains of Pseudmonas syringae pv. actinidiae isolated from France and Italy. N Z Plant Prot 64: 252–258.JL VannesteDA CornishJ. YuC. AudusseauS. Paillard2011Presence of the effector gene hopA1 in strains of Pseudmonas syringae pv. actinidiae isolated from France and Italy.N Z Plant Prot64252258
- 71. Ronning CM, Losada L, Brinkac L, Inman J, Ulrich RL, et al. (2010) Genetic and phenotypic diversity in Burkholderia: contributions by prophage and phage-like elements. BMC Microbiol 10: 202.CM RonningL. LosadaL. BrinkacJ. InmanRL Ulrich2010Genetic and phenotypic diversity in Burkholderia: contributions by prophage and phage-like elements.BMC Microbiol10202
- 72. Figueroa-Bossi N, Uzzau S, Maloriol D, Bossi L (2001) Variable assortment of prophages provides a transferable repertoire of pathogenic determinants in Salmonella. Mol Microbiol 39: 260–271.N. Figueroa-BossiS. UzzauD. MaloriolL. Bossi2001Variable assortment of prophages provides a transferable repertoire of pathogenic determinants in Salmonella.Mol Microbiol39260271
- 73. Ventura M, Canchaya C, Pridmore D, Berger B, Brussow H (2003) Integration and distribution of Lactobacillus johnsonii prophages. J Bacteriol 185: 4603–4608.M. VenturaC. CanchayaD. PridmoreB. BergerH. Brussow2003Integration and distribution of Lactobacillus johnsonii prophages.J Bacteriol18546034608
- 74. Van Sluys MA, De Oliveira MC, Monteiro-Vitorello CB, Miyaki CY, Furlan RL, et al. (2003) Comparative analysis of the complete genome sequences of Pierce's disease and citrus variegated chlorosis strains of Xylella fastidiosa. J Bacteriol 185: 1018–1026.MA Van SluysMC De OliveiraCB Monteiro-VitorelloCY MiyakiRL Furlan2003Comparative analysis of the complete genome sequences of Pierce's disease and citrus variegated chlorosis strains of Xylella fastidiosa.J Bacteriol18510181026
- 75. Harwood CS, Parales RE (1996) The β-ketoadipate pathway and the biology of self-identity. Annu Rev Microbiol 50: 553–590.CS HarwoodRE Parales1996The β-ketoadipate pathway and the biology of self-identity.Annu Rev Microbiol50553590
- 76. Helmick RA, Fletcher AE, Gardner AM, Gessner CR, Hvitved AN, et al. (2005) Imidazole antibiotics inhibit the nitric oxide dioxygenase function of microbial flavohemoglobin. Antimicrob Agents Chemoter 49: 1837–1843.RA HelmickAE FletcherAM GardnerCR GessnerAN Hvitved2005Imidazole antibiotics inhibit the nitric oxide dioxygenase function of microbial flavohemoglobin.Antimicrob Agents Chemoter4918371843
- 77. Delledonne M, Xia Y, Dixon RA, Lamb C (1998) Nitric oxide functions as a signal in plant disease resistance. Nature 394: 585–588.M. DelledonneY. XiaRA DixonC. Lamb1998Nitric oxide functions as a signal in plant disease resistance.Nature394585588
- 78. Brötz H, Sahl H-G (2000) New insights into the mechanism of action of lantibiotics—diverse biological effects by binding to the same molecular target. J Antimicrob Chem 46: 1–6.H. BrötzH-G Sahl2000New insights into the mechanism of action of lantibiotics—diverse biological effects by binding to the same molecular target.J Antimicrob Chem4616
- 79. Neilands JB (1995) Siderophores-structure and functions of microbial iron transport compounds. J Biol Chem 270: 26723–26726.JB Neilands1995Siderophores-structure and functions of microbial iron transport compounds.J Biol Chem2702672326726
- 80. Carniel E (2001) The Yersinia high-pathogenicity island: an iron-uptake island. Microb Infect 3: 561–69.E. Carniel2001The Yersinia high-pathogenicity island: an iron-uptake island.Microb Infect356169
- 81. Raymond KN, Dertz EA, Kim SS (2003) Enterobactin: an archetype for microbial iron transport. Proc Natl Acad Sci U S A 100: 3584–3588.KN RaymondEA DertzSS Kim2003Enterobactin: an archetype for microbial iron transport.Proc Natl Acad Sci U S A10035843588
- 82. Guilhabert MR, Kirkpatrick BC (2005) Identification of Xylella fastidiosa antivirulence genes: hemagglutinin adhesins contribute a biofilm maturation to X. fastidiosa and colonization and attenuate virulence. Mol Plant Microbe Int 18: 856–868.MR GuilhabertBC Kirkpatrick2005Identification of Xylella fastidiosa antivirulence genes: hemagglutinin adhesins contribute a biofilm maturation to X. fastidiosa and colonization and attenuate virulence.Mol Plant Microbe Int18856868
- 83. Das A, Rangaraj N, Sonti RV (2009) Multiple adhesin-like functions of Xanthomonas oryzae pv. oryzae are involved in promoting leaf attachment, entry, and virulence on rice. Mol Plant Microbe In 22: 73–85.A. DasN. RangarajRV Sonti2009Multiple adhesin-like functions of Xanthomonas oryzae pv. oryzae are involved in promoting leaf attachment, entry, and virulence on rice.Mol Plant Microbe In227385
- 84. Rees-George J, Vanneste JL, Cornish DA, Pushparajah IPS, Yu J, et al. (2010) Detection of Pseudomonas syringae pv. actinidiae using polymerase chain reaction (PCR) primers based on the 16S–23S rDNA intertranscribed spacer region and comparison with PCR primers based on other gene regions. Plant Pathol 49: 453–464.J. Rees-GeorgeJL VannesteDA CornishIPS PushparajahJ. Yu2010Detection of Pseudomonas syringae pv. actinidiae using polymerase chain reaction (PCR) primers based on the 16S–23S rDNA intertranscribed spacer region and comparison with PCR primers based on other gene regions.Plant Pathol49453464
- 85. Delcher AL, Bratke KA, Powers EC, Salzberg SL (2007) Identifying bacterial genes and endosymbiont DNA with Glimmer. Bioinformatics 23: 673–679.AL DelcherKA BratkeEC PowersSL Salzberg2007Identifying bacterial genes and endosymbiont DNA with Glimmer.Bioinformatics23673679
- 86. Kurtz S, Phillippy A, Delcher AL, Smoot M, Shumway M, et al. (2004) Versatile and open software for comparing large genomes. Genome Biol 5: R12.S. KurtzA. PhillippyAL DelcherM. SmootM. Shumway2004Versatile and open software for comparing large genomes.Genome Biol5R12
- 87. Darling AC, Mau B, Blattner FR, Perna NT (2004) MAUVE: multiple alignment of conserved genomic sequence with arrangements. Genome Res 14: 1394–1403.AC DarlingB. MauFR BlattnerNT Perna2004MAUVE: multiple alignment of conserved genomic sequence with arrangements.Genome Res1413941403
- 88. Gal-Mor O, Finlay BB (2006) Pathogenicity islands: a molecular toolbox for bacterial virulence. Cell Microbiol 8: 101–113.O. Gal-MorBB Finlay2006Pathogenicity islands: a molecular toolbox for bacterial virulence.Cell Microbiol8101113
- 89. Huson DH, Bryant D (2006) Application of phylogenetic networks in evolutionary studies. Mol Biol Evol 23: 254–267.DH HusonD. Bryant2006Application of phylogenetic networks in evolutionary studies.Mol Biol Evol23254267
- 90. Milne I, Lindner D, Bayer M, Husmeier D, McGuire G, et al. (2008) TOPALi v2: a rich graphical interface for evolutionary analyses of multiple alignments on HPC clusters and multi-core desktops. Bioinformatics 24: 126–127.I. MilneD. LindnerM. BayerD. HusmeierG. McGuire2008TOPALi v2: a rich graphical interface for evolutionary analyses of multiple alignments on HPC clusters and multi-core desktops.Bioinformatics24126127
- 91. Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate phylogenies by maximum likelihood. Syst Biol 52: 696–704.S. GuindonO. Gascuel2003A simple, fast, and accurate algorithm to estimate phylogenies by maximum likelihood.Syst Biol52696704
- 92. Huelsenbeck JP, Hillis DM, Nielsen R (1996) A likelihood-ratio test of monophyly. Syst Biol 45: 546–558.JP HuelsenbeckDM HillisR. Nielsen1996A likelihood-ratio test of monophyly.Syst Biol45546558
- 93. Schliep KP (2011) Phangorn: phylogenetic analysis in R. Bioinformatics 27: 592–593.KP Schliep2011Phangorn: phylogenetic analysis in R. Bioinformatics27592593
- 94. Strope CL, Abel K, Scott SD, Moriyama E N (2009) Biological sequence simulation for testing complex evolutionary hypotheses: indel-Seq-Gen version 2.0. Mol Biol Evol 26: 2581–2593.CL StropeK. AbelSD ScottN. Moriyama E2009Biological sequence simulation for testing complex evolutionary hypotheses: indel-Seq-Gen version 2.0.Mol Biol Evol2625812593
- 95. Jackson RW, Athanassopoulos E, Tsiamis G, Mansfield JW, Sesma A (1999) Identification of a pathogenicity island, which contains genes for virulence and avirulence, on a large native plasmid in the bean pathogen Pseudomonas syringae pathovar phaseolicola. Proc Natl Acad Sci U S A 96: 10875–10880.RW JacksonE. AthanassopoulosG. TsiamisJW MansfieldA. Sesma1999Identification of a pathogenicity island, which contains genes for virulence and avirulence, on a large native plasmid in the bean pathogen Pseudomonas syringae pathovar phaseolicola.Proc Natl Acad Sci U S A961087510880