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 hosts


  • Pablo Rodríguez-Palenzuela,

    1. Centro de Biotecnología y Genómica de Plantas, E.T.S. Ingenieros Agrónomos, Universidad Politécnica de Madrid, Campus de Montegancedo, E-28223 Pozuelo de Alarcón, Madrid, Spain.
    Search for more papers by this author
    • §

      These two authors have contributed equally to this work.

  • Isabel M. Matas,

    1. Instituto de Hortofruticultura Subtropical y Mediterránea ‘La Mayora’, Universidad de Málaga-Consejo Superior de Investigaciones Científicas (IHSM-UMA-CSIC), Área de Genética, Facultad de Ciencias, Universidad de Málaga, Campus Teatinos s/n, E-29010 Málaga, Spain.
    Search for more papers by this author
    • §

      These two authors have contributed equally to this work.

  • Jesús Murillo,

    1. Departamento de Producción Agraria, E.T.S. Ingenieros Agrónomos, Universidad Pública de Navarra, E-31006 Pamplona, Spain.
    Search for more papers by this author
  • Emilia López-Solanilla,

    1. Centro de Biotecnología y Genómica de Plantas, E.T.S. Ingenieros Agrónomos, Universidad Politécnica de Madrid, Campus de Montegancedo, E-28223 Pozuelo de Alarcón, Madrid, Spain.
    Search for more papers by this author
  • Leire Bardaji,

    1. Departamento de Producción Agraria, E.T.S. Ingenieros Agrónomos, Universidad Pública de Navarra, E-31006 Pamplona, Spain.
    Search for more papers by this author
  • Isabel Pérez-Martínez,

    1. Instituto de Hortofruticultura Subtropical y Mediterránea ‘La Mayora’, Universidad de Málaga-Consejo Superior de Investigaciones Científicas (IHSM-UMA-CSIC), Área de Genética, Facultad de Ciencias, Universidad de Málaga, Campus Teatinos s/n, E-29010 Málaga, Spain.
    Search for more papers by this author
    • Present address: Départament de Microbiologie Fondamentale (DMF), Quartier UNIL-Sorge, Bâtiment Biophore, CH-1015 Lausanne, Switzerland;

  • Martín E. Rodríguez-Moskera,

    1. Centro de Biotecnología y Genómica de Plantas, E.T.S. Ingenieros Agrónomos, Universidad Politécnica de Madrid, Campus de Montegancedo, E-28223 Pozuelo de Alarcón, Madrid, Spain.
    Search for more papers by this author
  • Ramón Penyalver,

    1. Centro de Protección Vegetal y Biotecnología, Instituto Valenciano de Invetigaciones Agrarias (IVIA), E-46113 Moncada, Valencia, Spain.
    Search for more papers by this author
  • Maria M. López,

    1. Centro de Protección Vegetal y Biotecnología, Instituto Valenciano de Invetigaciones Agrarias (IVIA), E-46113 Moncada, Valencia, Spain.
    Search for more papers by this author
  • José M. Quesada,

    1. Centro de Protección Vegetal y Biotecnología, Instituto Valenciano de Invetigaciones Agrarias (IVIA), E-46113 Moncada, Valencia, Spain.
    Search for more papers by this author
    • Departamento de Protección Ambiental, Estación Experimental del Zaidín, Consejo Superior de Investigaciones Científicas, c/ Profesor Albareda 1, E-18008 Granada, Spain.

  • Bryan S. Biehl,

    1. University of Wisconsin Biotechnology Center, University of Wisconsin, Madison, WI 53706, USA.
    Search for more papers by this author
  • Nicole T. Perna,

    1. University of Wisconsin Biotechnology Center, University of Wisconsin, Madison, WI 53706, USA.
    Search for more papers by this author
  • Jeremy D. Glasner,

    1. University of Wisconsin Biotechnology Center, University of Wisconsin, Madison, WI 53706, USA.
    Search for more papers by this author
  • Eric L. Cabot,

    1. University of Wisconsin Biotechnology Center, University of Wisconsin, Madison, WI 53706, USA.
    Search for more papers by this author
  • Eric Neeno-Eckwall,

    1. University of Wisconsin Biotechnology Center, University of Wisconsin, Madison, WI 53706, USA.
    Search for more papers by this author
  • Cayo Ramos

    Corresponding author
    1. Instituto de Hortofruticultura Subtropical y Mediterránea ‘La Mayora’, Universidad de Málaga-Consejo Superior de Investigaciones Científicas (IHSM-UMA-CSIC), Área de Genética, Facultad de Ciencias, Universidad de Málaga, Campus Teatinos s/n, E-29010 Málaga, Spain.
    Search for more papers by this author

E-mail; Tel. (+34) 952131955; Fax (+34) 952132001.


Pseudomonas savastanoi pv. savastanoi is a tumour-inducing pathogen of Olea europaea L. causing olive knot disease. Bioinformatic analysis of the draft genome sequence of strain NCPPB 3335, which encodes 5232 predicted coding genes on a total length of 5856 998 bp and a 57.12% G + C, revealed a large degree of conservation with Pseudomonas syringae pv. phaseolicola 1448A and P. syringae pv. tabaci 11528. However, NCPPB 3335 contains twelve variable genomic regions, which are absent in all previously sequenced P. syringae strains. Various features that could contribute to the ability of this strain to survive in a woody host were identified, including broad catabolic and transport capabilities for degrading plant-derived aromatic compounds, the duplication of sequences related to the biosynthesis of the phytohormone indoleacetic acid (iaaM, iaaH) and its amino acid conjugate indoleacetic acid-lysine (iaaL gene), and the repertoire of strain-specific putative type III secretion system effectors. Access to this seventh genome sequence belonging to the ‘P. syringae complex’ allowed us to identify 73 predicted coding genes that are NCPPB 3335-specific. Results shown here provide the basis for detailed functional analysis of a tumour-inducing pathogen of woody hosts and for the study of specific adaptations of a P. savastanoi pathovar.


The Gram-negative plant-pathogenic bacterium Pseudomonas syringae is an economically important pathogen and a relevant model for the study of plant–microbe interactions. An important characteristic of this pathogen is that it infects a wide range of herbaceous and woody plants causing diverse symptoms, such as leaf spots and blights, soft rots of vegetables, wilts, overgrowths, scabs and cankers (Agrios, 2005). Based on pathogenicity and host range, the species is divided into at least 50 pathovars (pvs.), of which the pvs. glycinea, phaseolicola and savastanoi were transferred to the newly created species Pseudomonas savastanoi (Gardan et al., 1992). Nevertheless, this new species and the current nomenclature of P. syringae are in need of a deep revision because taxonomy and phylogeny show a marked disparity. Indeed, DNA–DNA hybridization and multilocus sequence typing allow the division of P. syringae into at least nine different genomospecies (Gardan et al., 1999; Sarkar et al., 2006); in this scheme, P. savastanoi is included in genomospecies 2 and should be renamed according to the taxonomy rules. In order tomaintain the current status quo and avoid introducing more confusion, we will continue using the widespread designation of P. savastanoi pv. savastanoi, and use the trivial name of ‘P. syringae complex’ to refer to the nine genomospecies as a whole.

During the last few years, research on diseases caused by pseudomonads in herbaceous plants has progressed rapidly and the application of molecular genetics and genomics has provided new insights into the pathogenicity and virulence determinants, and their modes of action. Complete or draft genome sequences are now available for six strains representing four genomospecies and five P. syringae pathovars: pv. syringae (Psy) (Feil et al., 2005), of genomospecies 1; pv. phaseolicola (Pph) (Joardar et al., 2005) and pv. tabaci (Pta) (Studholme et al., 2009), of genomospecies 2; pv. tomato (Pto), of genomospecies 3 (Buell et al., 2003); and pv. oryzae (Por) (Reinhardt et al., 2009), of genomospecies 4. Comparative genomic analysis of these pathovars has not only revealed conserved components of the core genome of the P. syringae complex, but also those components unique to each of these pathogens (Feil et al., 2005; Joardar et al., 2005; Almeida et al., 2009; Studholme et al., 2009).

Bacteria of the P. syringae complex cause disease by shutting down the plant defence responses after injecting specialized proteins, called effectors, into the host cell cytoplasm using a type III secretion system (T3SS), known as the hrp injectisome (Grant et al., 2006). However, and although there is a large variation in the complement of effector and other virulence genes among different strains of the P. syringae complex (Lindeberg et al., 2006; Sarkar et al., 2006), it is not yet clear what other genes besides effectors are needed for disease production, or what are the genes responsible for the definition of host range and pathovar delineation. Also, the variety of symptoms caused on host plants by sequenced strains is limited to blights and leaf spots of herbaceous plants, with knowledge on the virulence and pathogenicity determinants specific for infection of woody plants, including those of tumour-inducing strains lagging far behind.

Tumour induction by Pseudomonas phytopathogens on host plants is restricted to P. savastanoi pvs. nerii (Psn), retacarpa and savastanoi (Psv). Infection of Olea europaea L. by Psv results in hypertrophy formation on the stems and branches, and occasionally on the leaves and fruits. The disease is considered to reduce both olive yield and productivity (Iacobellis, 2001); however, quantitative data on the impact of the disease on crop yield or crop quality are not available. Nevertheless, losses can be caused directly by localized infections that inhibit flowering and affect fruit development as well as taste, and can be caused indirectly by weakening immature main leader branches, resulting in later damage to the tree frame. Currently, the only molecular P. savastanoi determinants known to be involved in knot development are the phytohormones indoleacetic acid (IAA) and cytokinins (CKs) (Surico et al., 1985; Powell and Morris, 1986; Glass and Kosuge, 1988; Rodríguez-Moreno et al., 2008), as well as a functional T3SS (Sisto et al., 2004).

Research on Psv is now ready to move forward because of the development of molecular detection methods for both identifying P. savastanoi pathovars (Penyalver et al., 2000) and typing Psv strains (Quesada et al., 2008; Matas et al., 2009), the selection of virulent and genetically amenable Psv strains such as NCPPB 3335 (Pérez-Martínez et al., 2007), the establishment of the in vitro olive plant as model system for studying the pathogenicity and virulence of P. savastanoi pathovars (Rodríguez-Moreno et al., 2008; 2009), the identification of putative virulence determinants by global genomic analysis of Psv plasmids (Pérez-Martínez et al., 2008) and the description of the endopathogenic lifestyle of Psv NCPPB 3335 in olive knots (Rodríguez-Moreno et al., 2009). Strain NCPPB 3335 was isolated in 1984 from a diseased olive tree in France, and it is being used as a model organism, mostly because it can accept foreign DNA with a high frequency, which is uncommon among Psv strains (Pérez-Martínez et al., 2007). Strain NCPPB 3335 contains two native plasmids (Pérez-Martínez et al., 2008) and is highly virulent in olive trees, both in adult trees and in micropropagated plants (Rodríguez-Moreno et al., 2008).

To identify genetic differences that could account for the differing ability of Psv to infect woody plants in comparison with previously sequenced P. syringae strains, a deep coverage draft genome sequence of Psv strain NCPPB 3335 was obtained using pyrosequencing. The draft genome sequence was annotated and analysed for the presence of genomic regions showing no homology to other previously sequenced P. syringae strains and exhibiting characteristic features of genomic islands. For the purpose of defining the virulence gene complement of this tumour-inducing pathogen of olive, various features were considered in the analysis, including the biosynthesis of phytohormones (IAA, CKs and ethylene) and secretion systems, the repertoire of conserved and potentially novel T3SS effector genes and the catabolism of plant-derived aromatic compounds. Data obtained provide the basis for detailed functional analysis of a tumour-inducing pathogen of woody hosts belonging to the P. syringae complex.

Results and discussion

Sequencing and automatic annotation of the NCPPB 3335 genome

The draft chromosome sequence of P. savastanoi pv. savastanoi NCPPB 3335 yielded 112 supercontigs, between 7 and 270 Kb, with a total length of 5856998 bp and a 57.12% G + C. Most contigs ends consist of repeated sequences, and therefore further assembly was not possible with current data. However, given that the present draft has 15× sequence coverage, it is reasonable to assume that the vast majority of genes important for plant virulence are identifiable from the current draft. There are 5232 predicted coding genes of which 765 (15%) are annotated as hypothetical proteins. The genome sequence and additional information related to each predicted gene, such as gene product annotation, gene ontology, conserved domains or orthologous genes in other Pseudomonas are available in ASAP ( (Glasner et al., 2003).

The maintenance of gene order and location is an important indication of genome conservation; furthermore, genome comparison assists in the identification of lineage-specific regions, which can be involved in adaptation to specific niches and plant hosts. To investigate differences between Psv NCPPB 3335 and the other sequenced P. syringae strains, the genome draft assembly of NCPPB 3335 was compared with the genomes of the closest sequenced relatives Pph 1448A (Joardar et al., 2005) and Psy B728a (Feil et al., 2005). As expected from previous comparisons of P. syringae genomes (Feil et al., 2005; Joardar et al., 2005; Almeida et al., 2009), these three genomes showed extensive regions of conserved sequence and gene order, except for eight major inversions in NCPPB 3335 with respect to 1448A (Fig. 1), of which two (inversions 3 and 5) probably represent inversions in 1448A as the relative orientation is conserved in NCPPB 3335 and B728a.

Figure 1.

Pairwise alignments between the draft genome of P. savastanoi pv. savastanoi NCPPB 3335 (Psv) and the complete genomes of P. syringae pv. syringae B728a (Psy) and P. syringae pv. phaseolicola 1448A (Pph). BLASTN analysis was performed using WebACT and displayed with the ACT software; red and blue bars indicate collinear and inverted regions of identity respectively. Only matches larger than 1 kb are shown. Numbers indicate major inversions in the genome of Psv NCPPB 3335 with respect to Pph 1448A.

Phylogenetic relationship of Psv NCPPB 3335 with other Pseudomonas sequenced strains

To establish the phylogenetic relationship of Psv with respect to other selected members of the genus Pseudomonas with complete sequenced genomes, we compared a set of eight protein-coding house-keeping genes, namely: gyrB, rpoB, rpoA, recA, gyrA, rpoD, gltA and gapA. These genes are well conserved among the 10 taxa examined. We created an alignment of the proteins and reconstructed the phylogenetic tree showed in Fig. 2, using neighbour-joining methods. Maximum parsimony and minimal evolution methods were also used rendering phylogenetic trees with the same topology as the one shown in Fig. 2 (data not shown). As expected, because they are all included in genomospecies 2, Psv, Pph and Pta belong to the same cluster of closely related bacteria. Moreover, and in agreement with previous results using PCR fingerprinting (Stead et al., 2003), Psv appears to be separated from the other two related species included in this cluster, although the bootstrap value associated with this node is only 73. The observed divergence could reflect the specific adaptation of this pathogen to woody hosts.

Figure 2.

Evolutionary relationships of P. savastanoi and selected Pseudomonas. The evolutionary history was inferred using the neighbour-joining method (Saitou and Nei, 1987); the values of the bootstrap test (100 000 replicates) are shown in the branches. The evolutionary distances are in the units of the number of amino acid substitutions per site. Phylogenetic analyses were conducted in MEGA4 (Tamura et al., 2007). Pseudomonas strains were designateded as follows: Pph, P. syringae pv. phaseolicola 1448A; Pta, P. syringae pv. tabaci ATCC 11528; Psy, P. syringae pv. syringae B728a; Pto, P. syringae pv. tomato DC3000; Psv, P. savastanoi pv. savastanoi NCPPB 3335; Pfl, Pseudomonas fluorescens Pf5; Ppu, P. putida KT2440; Pen, Pseudomonas entomophila L48; Pae, Pseudomonas aeruginosa PAO1; Pst, Pseudomonas stutzeri A1501.

Mobile genetic elements

The genome of Psv NCPPB 3335 includes sequences with high similarity to the majority of the transposases found in Pph 1448A (Joardar et al., 2005) including IS801, ISPsy2, ISPsy3, ISPsy4, ISPsy16, ISPsy17, ISPsy19, ISPsy20 and ISPsy24, as well as the putative transposases PSPPH_3494, PSPPH_4464 and PSPPH_5043, indicating that they were acquired before the divergence of these two pathovars. ISPsy17 is present in nearly fifty intact copies in 1448A (Joardar et al., 2005) and is responsible for the emergence of race 6 and other virulent races of P. syringae pv. phaseolicola (Rivas et al., 2005). Unfortunately, the limited length of the 454 reads, of less than 300 nucleotides, do not allowed for the correct resolution and assembly of repeated sequences, and it is therefore not possible to accurately estimate the number of the different insertion sequences that are in more than one copy. NCPPB 3335 also contains homologous of ISPssy, ISPsy8 and ISPsy10, which are not present in 1448A. Remarkably, NCPPB 3335 contains at least 10 new transposase sequences that are similar to transposases from very diverse bacterial genera and at least 27 genes coding for integrases and resolvases, suggesting a high rate of horizontal exchange in this Psv strain.

Comparison of the protein complement of Psv NCPPB 3355 versus Pph 1448A and other pseudomonads

The orthologous genes with respect to Pph were identified by ‘reciprocal best hits’ and filtered to meet the criteria of showing more than 65% of identity in more than 65% of the length. The resulting 4132 out of 5232 (79%) genes could be considered the core genome of this bacterium. The remaining 1100 predicted genes were further investigated; 146 were identified as putative plasmid-associated genes; another group of 881 showed sequence similarities with other P. syringae genes, and these likely include both paralogues and xenologs in addition to orthologues that missed the threshold in the filtering step; the remaining group of 73 genes, which are not present in the known genomes of P. syringae species, were considered as Psv-specific genes (Table 1). Among this subset, 17 genes are related to DNA translocation and phage functions, 6 genes with DNA replication, 6 regulatory genes, 4 genes encoding transporter proteins, 4 genes encoding putative membrane proteins and 7 hypothetical proteins. In addition, there is a cluster of 8 genes related to degradation of aromatic compounds (see below) and 21 genes encoding other functions.

Table 1.  Proteins encoded by the draft Psv NCPPB 3335 genome showing no detectable homologues on the previously sequenced P. syringae genomes.
Psv IDaProductHighest identity
% aabOrganism
  • a. 

    ASAP ID number for Psv strain NCPPB 3335.

  • b. 

    % amino acid identity.

DNA translocation and phage functions   
 AER-0000385Phage-related lysozyme (muraminidase)76P. putida GB-1
 AER-0000612Tn5045 transposase86Xanthomonas axonopodis pv. citri 306
 AER-0000684Putative phage repressor53P. entomophila L48
 AER-0001423Putative integrase77P. putida W619
 AER-0001425Phage integrase82P. putida F1
 AER-0003446Prophage CP4-57 regulatory63Pseudomonas viridiflava
 AER-0003448Integrase72P. fluorescens SBW25
 AER-0003640FOG: transposase40Legionella drancourtii LLAP12
 AER-0003729Transposase IS3/IS911 family protein81P. putida W619
 AER-0003761ISPsy14, transposase88P. resinovorans
 AER-0003843ISSod6 transposase, IS130175Comamonas testosteroni CNB-1
 AER-0003846Putative site-specific recombinase87P. putida GB-1
 AER-0004165Filamentation induced by cAMP protein Fic53Pseudoalteromonas atlantica T6c
 AER-0004372Putative site-specific recombinase43Thiomicrospira crunogena XCL-2
 AER-0004373Putative enzyme; integration, recombination37T. crunogena XCL-2
 AER-0004384FOG: HEAT repeat46Xanthobacter autotrophicus Py2
 AER-0004920Phage integrase91P. putida F3
DNA replication   
 AER-0000348Putative DNA/RNA helicase48Clostridium perfringens C str. JGS1495
 AER-0000368DNA polymerase III α-subunit68P. putida GB-1
 AER-0001119Putative DNA helicase81P. fluorescens SBW25
 AER-0001120DNA helicase84P. fluorescens SBW25
 AER-0001534DNA replication protein dnaC59P. putida GB-1
 AER-0005045DEAD/DEAH box helicase domain-containing protein60Pseudomonas mendocina ymp
Regulatory genes   
 AER-0000170Transcriptional regulator66Escherichia coli O127
 AER-0000186Transcriptional regulator70Ralstonia solanacearum MolK2
 AER-0000242Diguanylate cyclase/phosphodiesterase with GAF sensor28Erythrobacter sp. SD-21
 AER-0001346Transcriptional regulator, RpiR family73P. fluorescens Pf-5
 AER-0001899Transcriptional activator BenR52Lutiella nitroferrum 2002
 AER-0001907Transcriptional activator BenR50Acinetobacter baumannii ATCC 19606
Degradation of aromatic compounds   
 AER-0001892Catechol 1,2-dioxygenase71P. fluorescens Pf-5
 AER-0001893Muconolactone delta-isomerase62Escherichia fergusonii ATCC 35469
 AER-0001894Cis-muconate cycloisomerase CatB70Pseudomonas sp. CA10
 AER-0001895Anthranilate dioxygenase reductase69P. resinovorans
 AER-0001896Benzoate 1,2-dioxygenase β-subunit92P. resinovorans
 AER-0001898Toluate 1,2-dioxygenase α-subunit89P. resinovorans
 AER-0001903Putative oxygenase subunit66Acinetobacter sp. ADP1
 AER-0001905Carboxymethylenebutenolidase69Burkholderia sp. H160
Genes encoding transporter proteins   
 AER-0000173Amino acid transporters58P. stutzeri A1501
 AER-0003953Transporter, LysE family62X. oryzae pv. oryzae MAFF
 AER-0004697Major facilitator superfamily MFS_162Dickeya dadantii Ech703
 AER-0005083Major facilitator superfamily MFS_148Rhodopseudomonas palustris BisB18
Putative membrane proteins   
 AER-0003943Probable membrane-fusion protein75P. fluorescens PfO-1
 AER-0004085Membrane protein, putative29P. aeruginosa UCBPP-PA14
 AER-0004086Membrane protein, putative37P. aeruginosa UCBPP-PA14
 AER-0004919Membrane protein, putative90P. putida F1
Hypothetical proteins   
 AER-0001536Hypothetical protein45Xylella fastidiosa 9a5c
 AER-0002231Hypothetical protein45Vibrio fischeri ES114
 AER-0004230Hypothetical protein93P. putida F1
 AER-0004374Hypothetical protein45T. crunogena XCL-2
 AER-0004726Hypothetical protein31Idiomarina baltica OS145
 AER-0004727Hypothetical protein32Haemophilus influenzae 3655
 AER-0004918Hypothetical protein90P. putida F1
Other genes   
 AER-0000190Reductase SDR62Methylobacillus flagellatus KT
 AER-0000328Phosphotyrosine protein phosphatase57Xanthomonas campestris pv. vesicatoria 85–10
 AER-0000369SMC domain-containing protein48P. putida GB-1
 AER-0000379Putative DNA binding protein55Klebsiella pneumoniae 342
 AER-0000381Endodeoxyribonuclease RusA54P. aeruginosa PA7
 AER-0000504Aminotransferase class IV64Variovorax paradoxus S110
 AER-0000505Cyclohexadienyl dehydratase60V. paradoxus S110
 AER-0000506Menaquinone biosynthesis-related protein47Burkholderia phytofirmans PsJN
 AER-0001421ATP/GTP binding protein95P. putida W619
 AER-0001728Aspartyl-tRNA(Asn) amidotransferase30Rhodococcus opacus B4
 AER-0003442Conserved domain protein76P. fluorescens SBW25
 AER-0003466Modification methylase Eco57IB48K. pneumoniae subsp. rhinoscleromatis ATCC 13884
 AER-0003467Type II restriction enzyme69Flavobacterium psychrophilum JIP02/86
 AER-0003468Phosphoglycolate phosphatase37Salmonella enterica subsp. enterica SPB7
 AER-0003951Glutamine synthetase family protein72Burkholderia phymatum STM815
 AER-0003952Homoserine O-succinyltransferase69B. phymatum STM815
 AER-0003955Fatty acid desaturase72B. phymatum STM815
 AER-0003956O-acetylhomoserine sulfhydrylase72B. phymatum STM815
 AER-0003957Pyruvate aminotransferase63B. phymatum STM815
 AER-0004933Sulfatase family protein83P. fluorescens Pf0-1
 AER-0004941Surface antigen gene44P. fluorescens Pf-5

We have also analysed the presence of variable regions (VR) in the Psv genome. These regions, exhibiting either high level of variation in content, alignment, or both relative to related genomes (Hacker and Kaper, 2000; Welch et al., 2002), are shaped by a range of processes including excision, recombination, duplication and horizontal transfer. In addition, when components are implicated in virulence functions, VRs are described as pathogenicity islands (Gal-Mor and Finlay, 2006). Variable regions were defined as DNA regions larger than 10 kb that met both of the following criteria: (i) they appeared as gaps in genome alignments with respect to the genome of Pph 1448A using the MAUVE software (Darling et al., 2004) and (ii) their component genes were not found in other P. syringae strains in Megablast comparisons. The borders of the identified VRs were analysed using ACT and checking for lack of synteny with Psy B728A and Pph 1448A. Table 2 shows the location and gene content of the 12 VRs identified. These regions are good candidates to be the result of horizontal transfer. In all cases except in VRs 3, 7 and 8, the presence of mobile genetic elements, tRNAs and phage-related proteins was observed. Also it must be pointed out that the VR regions are enriched in hypothetical proteins, constituting 33.6% of the total number of genes in these regions, compared with the overall 15% of hypothetical proteins. Only a small number of genes found in the VRs are likely to code for virulence factors, such as the genes for haemagglutinin and haemolysin (VR1), two MCPs (methyl accepting chemotaxis proteins) (VRs 2 and 5), one protease, one putative efflux pump (VR3) and two genes encoding for Fe-chelating functions (VR7). The most striking feature found in the VRs is a cluster of genes (VR8, discussed below) involved in degradation of phenolic compounds. This VR does not show any landmark of horizontal transfer, such as the presence of mobile genetic elements.

Table 2.  Variable regions found in the genome of Psv NCPPB 3335.
VRContigscoordinate% G + CVR-encoded genes
 1Contig00620 636–64 06053.19 Hypothetical proteins
4 Transposases
2 Recombinases
Outer membrane porin, OprD family
Probable haemagglutinin
Haemolysin activator protein precursor
GCN5-related N-acetyltransferase
 2Contig02745–59 3825210 Hypothetical proteins
2 Integrases
7 Transferases
Prophage CP4-57 regulatory
DnaJ domain protein
Methyl-accepting chemotaxis protein
ATP binding component of ABC-transporter
O-antigen export system permease protein RfbD
dTDP-4-dehydrorhamnose 3,5-epimerase
dTDP-4-dehydrorhamnose reductase
dTDP-glucose 4,6-dehydratase
Modification methylase Eco57IB
Type II restriction enzyme, methylase subunit
Phosphoglycolate phosphatase
MobB protein
Integral membrane protein
 3Contig02815 794–31 23760.23 Hypothetical proteins
Outer membrane efflux protein superfamily
Flagellar hook-length control protein fliK
Microcystin-dependent protein
Histone acetyltransferase HPA2 and related acetyltransferases
Zn-dependent proteases
 4Contig032790–37 38658.523 Hypothetical proteins
9 Phage-related proteins
2 Holliday junction resolvase
Nin protein
DNA polymerase III α-subunit
Chromosome partition protein smc
Transcriptional regulator, Cro/CI family
putative DNA binding protein
 5Contig04813 417–60 03455.423 Hypothetical proteins
17 Phage-related proteins
2 Terminases
DNA methyltransferase
Anti-termination protein Q, putative
Methyl-accepting chemotaxis protein
Tail fibre domain protein
SSU ribosomal protein S2p (SAe)
 6Contig04932 149–46 92751.810 Hypothetical proteins
2 Prophage lambda integrases
C-5 cytosine-specific DNA methylase family protein
Repressor protein c2
DNA replication protein dnaC
 7Contig05316 426–26 93756.2Sulfatase family protein
Ais protein, putative
PAP2 superfamily protein
Two-component system sensor protein
Hypothetical protein
    RNA polymerase sigma-70 factor, ECF subfamily
Fe2+-dicitrate sensor, membrane component
Ferrichrome-iron receptor
Surface antigen gene
 8Contig065188 176–202 89060.1Catechol degradation (Table 4)
2 277–7 348
54.42 Hypothetical proteins
Phage integrase : Phage integrase, N-terminal SAM-like
ISPpu14, transposase Orf3
N-succinyl-L,L-diaminopimelate desuccinylase
N-succinyl-L,L-diaminopimelate aminotransferase alternative
Acetylornithine deacetylase/succinyl-diaminopimelate desuccinylase
Putative transport protein
Enoyl-[acyl-carrier-protein] reductase [FMN]
Phage tail sheath protein FI
FOG : TPR repeat
10Contig07576 444–91 33452.92 Phage integrases
FOG : GGDEF domain
Ethanolamine operon regulatory protein
Protein involved in meta-pathway of phenol degradation
Putative amidase
ATP/GTP binding protein
MazG nucleotide pyrophosphohydrolase domain protein
Hypothetical protein
11Contig08115 625–26 784536 Hypothetical proteins
12Contig83112 202–143 759564 Aldo-keto reductases
5 Transcriptional regulators
2 Transposases
Putative tRNA-m1A22 methylase
Putative translation initiation inhibitor, yjgF family
Glutathione S-transferase
DnaJ-class molecular chaperone
ATP-dependent helicase HrpB
putative phage-related hypothetical protein
NAD-dependent protein deacetylase of SIR2 family
Amino acid transporters
hypothetical protein
4-Carboxymuconolactone decarboxylase
Xanthine dehydrogenase iron-sulfur subunit
Putative xanthine dehydrogenase yagS, FAD binding subunit
Xanthine dehydrogenase, molybdenum binding subunit
Xanthine and CO dehydrogenases maturation factor
4-Carboxymuconolactone decarboxylase
Probable transport transmembrane protein

Metabolic potential of the Psv NCPPB 3335 genome

Through the past century, Pseudomonas species have figured prominently in efforts to unravel how microbes recycle disparate organic molecules in the environment, including aromatic hydrocarbons (Wackett, 2003). However, and in spite of the availability of the genome sequences of five P. syringae pathovars, little attention has been paid to date to the catabolic versatility of this bacterial complex.

Oxygenases play key roles in the chemical transformation of recalcitrant compounds (Resnick and Gibson, 1996), and several dioxygenase-coding genes are usually found in the genome of metabolically versatile bacteria such as Pseudomonas putida (Ppu) KT2440 (Nelson et al., 2002). Analysis of the Psv NCPPB 3335 genome reveals the existence of 46 putative oxygenase-related genes, of which 21 were predicted to be dioxygenases, including the α- and β-subunits of protocatechuate 3,4-dioxygenase (AER-0000733 and AER-0000732, respectively) and catechol 1,2-dioxygenase (catA gene, AER-0001892) structurally and functionally characterized in Ppu KT2440. Besides, three homologues of genes coding for taurine family dioxygenases were also found (AER-0002870, AER-0002989 and AER-0004755). Out of these three loci, only AER-0004755 showed an orthologue gene in the genome of Ppu KT2440 (PP_0169). Interestingly, four of the 21 Psv genes encoding putative dioxygenases (AER-0001892, catA; AER-0001898, antA; AER-0001897/ AER-0001896, antB and AER-0001895, antC) were found in a VR of the Psv NCPPB 3335 genome (VR8, Table 2), which showed high identity with the carbazole-degradative plasmid pCAR1 from Pseudomonas resinovorans CA10 (Nojiri et al., 2002; Urata et al., 2004). Besides the antABC gene cluster, encoding two-component anthranilate 1,2-dioxygenase, involved in the degradation of anthranilate to catechol in P. resinovorans CA10, VR8 also encodes antR (Table 3), the positive regulator of the antABC operon (Urata et al., 2004). Furthermore, VR8 also encodes homologues to catB and catC genes, organized together with the catA gene in identical orientation to that shown by the catechol-degradative gene cluster catBCA of plasmid pCAR1 (Nojiri et al., 2002) and other pseudomonads (Table 3). Conjugative transfer of pCAR1 from P. resinovorans CA10 enabled Ppu KT2440 to utilize either carbazole or anthranilate as its sole source of carbon and nitrogen (Miyakoshi et al., 2007). Thus, all these results suggest that Psv NCPPB 3335 could utilize anthranilate as sole carbon and nitrogen source. Incubation of NCPPB 3335 in minimal succinate medium (SSM) (Meyer and Abdallah, 1978) containing anthranilate as the sole source of nitrogen resulted in browning of the medium, whereas P. syringae strains not encoding the antABC operon (Pph 1448A and Pto DC3000) did not accumulate the brown compound (data not shown). This phenotype, attributed to the possible accumulation of a substituted cathecol, has been previously observed in P. putida strains affected in their ability to degrade toluene (Mars et al., 1998). Also, it is tempting to speculate that catabolic pathways encoded within VR8 from Psv NCPPB 3335, a genomic region absent in all sequenced P. syringae strains infecting herbaceous plants but shared with P. resinovorans, commonly found in the lubricating oils of wood mills, could offer a selective advantage for growth of P. savastanoi strains in woody hosts. It is well known that phenolic compounds provide a natural defence against pathogen attack (Agrios, 2005) and also that olive tree tissues are particularly abundant in these compounds (Oi-kano et al., 2008). Moreover, it has been reported that the production of phenolic compounds is greatly increased in olive tree knots, upon Psv attack (Cayuela et al., 2006), which strongly suggests that bacterial resistance to phenols could be of paramount importance in the pathogenicity of this bacterium. However, all these hypotheses remain to be tested.

Table 3.  Putative genes related with the metabolism of aromatic compound substrates identified by in silico whole genome analysis of P. savastanoi pv. savastanoi NCPPB 3335. Thumbnail image of

Vanillin and many other plant products such as lignin-related aromatic compounds are metabolized through protocatechuate. Genes homologous to the loci vdh, vanA and vanB, which are involved in the bioconversion of vanillin to the central intermediate protocatechuate by Pseudomonas sp. strain HR199 (Priefert et al., 1997), were identified in the genome of Psv NCPPB 3335 and other sequenced P. syringae strains (Table 3). In addition, the amino acid sequences deduced from two corresponding loci identified in the genome of Psv NCPPB 3335 revealed high degrees of identity to pcaG and pcaH genes, encoding the two subunits of protocatechuate 3,4-dioxygenase, a key ring-cleaving dioxygenase of the β-ketoadipate pathway involved in the degradation of protocatechuate in soil (Overhage et al., 1999) bacteria. Moreover, Psv NCPPB 3335 also contains homologues of the pobA (4-hydroxybenzoate hydroxylase gene) and pobR (a regulator that activates pobA expression) genes, involved in the catabolism of 4-hydroxybenzoate to protocatechuate in P. putida (Bertani et al., 2001) (Table 3).

Table 3 shows the identification of other Psv NCPPB 3335 genes potentially involved in the biosynthesis of ubiquinone from 4-hydroxybenzoate, the meta-pathway of phenol degradation or the transport and tolerance to toluene.

Phytohormones biosynthetic genes

The knot forming P. savastanoi pathogens differ from almost all other bacterial plant pathogens in the production of substantial quantities of the phytohormones IAA and CKs. Whereas knot development in plants infected with P. savastanoi is dependent on bacterial production of IAA, CKs production is usually regarded as contributing to virulence in P. savastanoi and all other gall forming bacteria (Surico et al., 1985; Glass and Kosuge, 1988; Kennelly et al., 2007; Rodríguez-Moreno et al., 2008). In P. savastanoi, IAA is synthesized from tryptophan in two steps catalysed by the products of the iaaM (tryptophan monooxygenase) and iaaH (indoleacetamide hydrolase) genes. Pseudomonas savastanoi also converts IAA to IAA-lysine through the action of the iaaL gene. Recently, it has been reported that Psv strains contains two iaaL paralogues, which are both chromosomally located in most strains (Matas et al., 2009). In agreement with these results, two iaaL paralogues were found in the genome of Psv NCPPB 3335, with nucleotide sequences that are 90% identical. Furthermore, sequence analysis showed that this strain also contains two iaaM and iaaH paralogues sequences, which are probably organized in two different iaaMH operons, located in contigs 26 and 34, with the iaaM locus first in the gene cluster, as suggested by the contiguous ASAP numbers automatically assigned to both iaaM-iaaH genes partners (Table 4). Interestingly, Southern hybridization analysis of total DNA isolated from 10 different Psv strains with iaaM and iaaH probes revealed that all of them harboured two copies of those two genes (data not shown). However, the nucleotide sequences of AER-0003958 (iaaM-2) and AER-0003642 (iaaH-1) show a 22-nucleotide insertion causing a frameshift and 158 ambiguous nucleotides respectively. Thus, to determine whether these sequences encode functional genes or pseudogenes the nucleotide sequences of these two loci should first be confirmed.

Table 4.  Indole-3-acetic acid biosynthetic genes identified by in silico whole genome analysis of P. savastanoi pv. savastanoi NCPPB 3335.
Psv IDaContigGenebProduct/commentHighest identityOrthologuesc
OrganismcAccession number% ndPpuPphPto
  • a. 

    ASAP ID number for Psv strain NCPPB 3335.

  • b. 

    Asterisks indicate putative pseudogenes.

  • c. 

    Ppu, P. putida KT2440; Pph, P. syringae pv. phaseolicola 1448A; Pto, P. syringae pv. tomato DC3000; Psn, P. savastanoi pv. nerii (plasmid pIAA1); Psy, P. syringae pv. syringae.

  • d. 

    % Nucleotide identity. Ambiguous nucleotides were not considered for determination of the identity.

  • e. 

    AER-0003637 and AER-0004688 correspond to alleles iaaLPsn and iaaLPsv (Matas et al., 2009) respectively.

AER-0003641034iaaM-1Tryptophan 2-monooxygenasePsn plasmid pIAA1M11035100
Psy 3023AY53053694
AER-0003958026iaaM-2*Truncated (premature termination)Psy B728aCP00007597
Psy Y30U0435897
AER-0003642034iaaH-1*Sequence containing 158 ambiguous nucleotidesPsn plasmid pIAA1M11035100_
Psy B728aCP00007593
AER-0003959026iaaH-2Indoleacetamide hydrolasePsy B728aCP00007599
Psy Y30U0435899
AER-0003637e034iaaL-1Indoleacetate-lysine synthetasePsn plasmid pIAA1M3537398+
AER-0004688e092iaaL-2Indoleacetate-lysine synthetasePtoAE01685391+
Psn plasmid pIAA1M3537390

A sequence homologous to the trans-zeatin producing gene (ptz) from plasmid pCK1 of Psn (GenBank X03679.1) was found in the genome of Psv NCPPB 3335; additionally, plasmid pPsv48A (approximately 73 Kb), isolated from NCPPB 3335, was previously shown to hybridize with a ptz probe (Pérez-Martínez et al., 2008), indicating that CKs biosynthesis is specified, at least in part, by plasmid-borne genes in this strain.

blast searches using the amino acid sequences of ethylene forming enzymes from Pph, P. syringae pv. glycinea or P. syringae pv. pisi (GenBank accession numbers AAD16440.1, AAD16439.1 and AAD16443.1, respectively) yielded no significant hits against the genome of Psv NCPPB 3335. Previously, it was reported that homology to the efe gene was neither found by hybridization analysis of 32 different Psv plasmids with an efe probe (Pérez-Martínez et al., 2008). Together, all these results suggest that Psv isolates do not belong to the group of 2-oxoglutarate-dependent ethylene producers, a pathway dependent on the efe gene in several P. syringae pathovars (Weingart et al., 1999).

Secretion systems associated with virulence

Structural genes involved in the biosynthesis of secretion systems classified as I, II, III, IV and VI were identified in the genome of Psv NCPPB 3335. Genes related with type II, III, IV and VI secretion systems are shown in Table S1. Like Pph 1448A (Joardar et al., 2005), Psv NCPPB 3335 also encodes multiple type I secretion pathway components and two distinct type II secretion pathways: the Sec and the Tat systems. These major secretion pathways are found in many pathogenic bacteria including other Pseudomonas (Duong et al., 2001). NCPPB 3335 also contained a complete T3SS similar to those found in pathogenic strains of the P. syringae complex (Grant et al., 2006); this was not unexpected, because it was demonstrated that Psv needs a functional T3SS to cause tumours in olive plants (Sisto et al., 2004). Additionally, Psv NCPPB 3335 contains a cluster of 22 genes (AER-0001650 to AER-0001671; Table S1) encoding components of a putative T3SS; this cluster is also present in Pph 1448A, Pta 11528 and Por 1–6, but not in Pto DC3000, Psy B728a or other sequenced Pseudomonas. Remarkably, the genes in this second T3SS cluster are not preceded by regulatory sequences typical of the HrpL regulon and, as deduced from the analysis of specific mutants in the canonical T3SS (Sisto et al., 2004; Joardar et al., 2005), it appears that it does not transfer effectors to plant cells and that is not essential for pathogenicity.

Genes related with type IVA and IVB secretion systems were also found in the genome of Psv NCPPB 3335. With the exception of virB3 and virB7, homologues of all other type IVA secretion system genes were found in the genome of this strain; however, these two genes were detected by hybridization with specific probes in a plasmid of Psv NCPPB 3335 (Pérez-Martínez et al., 2008). In agreement with these authors, sequence analysis of this strain also identified a truncated type IVB secretion system (Table S1).

There are two clusters of genes encoding a T6SS. The first one (AER-0002618 to AER-0002633) is highly similar to that found in Pph 1448A. However, and although none of the genes related with the second T6SS (AER-0002971 to AER-0002983) are found in Pph 1448A, both gene order and sequence are conserved in Pto DC3000 (Table S1). At present, there is not enough information in the scientific literature to evaluate the possible completeness of this second set of T6SS genes.

Type III secretion system effectors

A prediction of putative hop genes encoded in the genome of Psv NCPPB 3335 was performed based in three different criteria: (i) sequence homology with the catalogue of confirmed and predicted hop genes available in the Hop database (HopDB, (Lindeberg et al., 2006); (ii) N-terminal sequence features, using the programme EffectiveT3 ( (Arnold et al., 2009) and (iii) identification of potential HrpL boxes in the promoter sequences (220 nucleotides upstream of the start codon) following the criteria described by Fouts and colleagues (2002). Table S2 shows 16 putative effectors with amino acid identity higher than 80% to previously described effectors and three candidate genes that showed identity values between 65% and 70% but met at least one of the two other criteria. Moreover, 11 candidate genes were found, which do not share sequence similarity with known effectors, but they met the two other criteria. A comparison of the effector repertoire of Psv NCPPB 3335, Pph 1448A and Pta 11528 allowed definition of a ‘core’ set of 7 hop genes that are conserved in all three pathovars. Three of these seven genes were also found in the genomes of Pto DC3000 and Psy B728a; however, hopR1 and hopV1 are only present in Pto DC3000 and hopAE1 in Psy B728a. In contrast, hopAS1, plasmid-encoded in Psv NCPPB 3335 (Fig. 3), was not found either in Pto DC3000 or Psy B728a. The subset of effectors in Psv NCPPB 3335 not shared with Pph 1448A or Pta 11528 comprises another five genes, three of which appeared to be specific to Psv NCPPB 3335; thus, it is possible that they play specific roles in the interaction of this bacterium with woody plants. Also, among the 11 candidate effectors with no homology with known hop genes, there are three that are specific to Psv NCPPB 3335. Experimental data showing the in planta T3SS-dependent translocation are needed to ascertain their identity as Hop proteins.

Figure 3.

Comparison of the type III effector gene complements of P. savastanoi pv. savastanoi (Psv) NCPPB 3335 and other sequenced plant-pathogenic pseudomonads. Those genes that are conserved in P. syringae pv. syringae (Psy) B728a and P. syringae pv. tomato (Pto) DC3000 are shown in bold face and underlined respectively. Gene hopAF1–2, plasmid encoded in Psv NCPPB 3335, shows 73%–74% amino acid identity with hopAF1 from Psy B728a, Pph 1448A and Pto DC3000 (Table S2); however, these three strains contains only a hopAF1 gene encoded in their chromosomes. Pph, P. syringae pv. phaseolicola; Pta, P, syringae pv. tabaci. #Plasmid-encoded gene; asterisks indicate putative pseudogenes; hop genes that are truncated by a frameshift or a premature stop codon are indicated by the addition of a single quotation mark to the gene name (Lindeberg et al., 2005).

Other virulence-related genes

Known virulence determinants in plant-pathogenic Pseudomonas includes phytotoxins, cell wall-degrading hydrolytic enzymes, extracellular polysaccharides, iron-uptake systems, resistance mechanisms to plant-derived antimicrobials, adhesion, as well as the general processes of motility and chemotaxis (Agrios, 2005). Based on the annotation of the Psv NCPPB 3335 draft genome, we have found 551 genes potentially involved in several aspects of the virulence of this bacterium, which are shown in Table S1. Most of these genes are conserved in Pph 1448A; consequently, the subset of Psv NCPPB 3335-specific genes (Table 1) merits special attention as they could be involved in the ability of this pathogen to infect woody plants.

The genome of Psv NCPPB 3335 encodes several putative cell wall-degrading enzymes: two cellulases, two pectate lyases, two pectin lyases and a polygalacturonase (Table S1). Interestingly, out of these seven candidate enzymes, a cellulase (AER-0003969) and a pectate lyase (AER-0003025) are not found in Pph 1448A, although they are present in other plant-pathogenic Pseudomonas.

Exopolysaccharides, which are components of the bacterial capsule, play a role in both adhesion and protection of bacterial cells from external stresses (Gacesa, 1998). In phytopathogenic Pseudomonas, the exopolysaccharide alginate has been reported as an epiphytic fitness and virulence factor in relation to the production of water soaked lesions (Peñaloza-Vazquez et al., 2004). In addition, levansucrase, an enzyme required for biosynthesis of the polysaccharide levan, seems to have a role in the early phase of infection by creating a separating layer between bacteria and plant cell wall to prevent pathogen recognition (Hettwer et al., 1995). A set of 18 genes involved in alginate biosynthesis and present in Pph 1448A was found in Psv NCPPB 3335. By contrast, while only a single levansucrase-coding gene (AER-000661) was identified in Psv NCPPB 3335 (Table S1), Pph 1448A encodes three different levansucrase genes. These results might be related to the fact that Psv strains are in general levan-negative whereas the P. syringae pathovars of subgroup 1a are all levan-positive (Lelliott and Stead, 1987).

Pathogenic bacteria rely on several cell surface associated factors to increase adhesion to the host surface (Wall and Kaiser, 1999), and filamentous haemagglutinin has been shown to be an adhesion and a virulence factor in the plant pathogen Dickeya dadantii (Rojas et al., 2002). Psv NCPPB 3335 contains one putative filamentous haemagglutinin (AER-0004375), which is also present in Pph 1448A, although truncated by the insertion of a mobile element. In addition, 31 genes with annotations related to the biosynthesis of type IV pili, a major group of adhesion factors usually found in pseudomonads (Wall and Kaiser, 1999), were also found in both Psv NCPPB 3335 (Table S1) and Pph 1448A.

Chemotaxis and motility phenomena have been associated with virulence in phytopathogenic bacteria, particularly during the first stages of the infection process (Tans-Kersten et al., 2001; Antúnez-Lamas et al., 2009). The genome of Psv NCPPB 3335 shows a nearly duplicate set of genes involved in the transduction pathway of the chemotaxis process (che genes, Table S1). Moreover, 50 genes with Pfam signature domain corresponding to MCPs were identified in the genome of this strain. This high number of MCPs is common in other phytopathogenic pseudomonads and probably reflects the complex navigation skills required for this particular bacterial lifestyle.

Table S1 also shows the identification in the genome of Psv NCPPB 3335 of other virulence-related factors, such as toxins, siderophores, quorum sensing regulation and multidrug transporters. The relevance of these factors in P. savastanoi has not been reported to date.


Bioinformatic analysis of the draft genome of Psv NCPPB 3335, a pathogen of olive, revealed a large degree of conservation with other pseudomonads belonging to the P. syringae complex, principally with Pph 1448A and Pta 11528. However, Psv NCPPB 3335 genome showed eight major inversions with respect to Pph 1448A and 12 VRs not present in any other sequenced P. syringae strain. The existence of 21 putative dioxygenase-coding genes in the genome of this strain, several of which were predicted to be involved in the degradation of aromatic hydrocarbons and are located in a variable genome region, revealed the potential metabolic versatility of this strain in comparison with other P. syringae strains. From this sequence, combined with a Southern hybridization screen, it was also found that Psv strains encode two paralogues sequences of iaaM and iaaH, further supporting the role of IAA in pathogenicity. Our study has also highlighted particular classes of virulence-related genes, many of which are specific for Psv NCPPB 3335. Detailed analysis of the repertoire of genes encoding putative T3SS effector proteins in this strain, and comparison in other sequenced P. syringae, revealed both the core arsenal of P. syringaeP. savastanoi effectors and those strain-specific. In summary, results shown here provide, for the first time, the basis for detailed functional analysis of a tumour-inducing pathogen of woody hosts and for the study of specific adaptations of a P. savastanoi pathovar.

Experimental procedures

Growth of P. savastanoi pv. savastanoi NCPPB 3335 and DNA isolation

Pseudomonas savastanoi pv. savastanoi NCPPB 3335 (Pérez-Martínez et al., 2007) was grown on LB medium at 28°C. DNA was extracted using Jet-Flex genomic DNA purification kit (Genomed GmbH, Löhne, Germany) following the manufacturer's instructions. The DNA sample was further purified by extraction with phenol : chloroform and precipitation with ethanol and resuspended in MilliQ water. NanoDrop measurements gave a concentration of 1.7 µg µl−1 (in total 550 µg of DNA) with A260/A280 of 1.95.

Genome sequencing

454 pyrosequencing was carried out by Eurofins Medigenomix GmbH (Planegg-Martinsried, Germany) with GS20/FLX technology (Roche, Basel, Switzerland) using 20 µg of DNA. Shotgun sequencing was performed to a redundancy of 15× followed by a pair-end library analysis to determine the orientation and relative position of contigs produced by de novo shotgun sequencing. The sequence read length was 200–300 bp. Assembly was performed by using the Newbler Assembler software provided by 454 Life Sciences.

Annotation and genome analysis

Scaffolds were aligned against the complete genome sequence of P. syringae pv. phaseolicola 1448A using MAUVE (Darling et al., 2004) and automatically annotated using the RAST server (Meyer et al., 2008). Genomes were also aligned using a genome-wide blast comparison and visualized through ACT (Carver et al., 2005), and manually inspected. The genome sequence is available in and can be visualized using the ASAP software (Glasner et al., 2003); it was also deposited at DDBJ/EMBL/GenBank under the accession ADMI00000000 and

Phylogeny of P. savastanoi pv. savastanoi NCPPB 3335 and other strains belonging to the genus Pseudomonas was performed by multilocus sequence analysis using a concatenated data set of gyrA, gyrB, rpoA, rpoA, rpoB, rpoD, recA, gltA and gapA genes. Multiple alignments were performed with Clustal W (Larkin et al., 2007), and a phylogenetic tree was obtained using the neighbour-joining method (Saitou and Nei, 1987). The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (100000 replicates) were shown next to the branches in Fig. 2 (Felsenstein, 1985). The evolutionary distances were computed using the Poisson correction method (Zuckerkandl and Pauling, 1965) and are in the units of the number of amino acid substitutions per site. All positions containing gaps and missing data were eliminated from the dataset (Complete deletion option). There were a total of 4895 positions in the final dataset. Phylogenetic analyses were conducted in MEGA4 (Tamura et al., 2007).

Orthologous genes to Pph 1448A (core genome) were determined by reciprocal best hits using the bioinformatics facilities of ASAP (Glasner et al., 2003). Further annotation was performed with Blast2GO (Conesa et al., 2005; Götz et al., 2008) and Pfam profiles with the aid of ad hoc Perl scripts.

Effectors prediction was performed with the programme EffectiveT3 ( (Arnold et al., 2009) with ad hoc Perl scripts. The promoter regions of AER-0000625 (hopAU1), AER-0000968 (hopAF1-1) and AER-0002657 (hopA1), whose ORFs are located at the end of a contig, were amplified by arbitrary PCR (Espinosa-Urgel et al., 2000) and sequenced prior to effector prediction.


This research has been supported by the Spanish Plan Nacional I+D+I grants AGL2008-55311-CO2-01, AGL2008-55311-CO2-02 and AGL-2009-12757, as well as grant refs. P08-CVI-03475 from the Junta de Andalucía, 57/2007 from the ‘Departamento de Educación y Cultura, Gobierno de Navarra’ (Spain) and CAO00-007 from the INIA (Spain). I.M.M. was supported by the Ramón Areces Foundation (Spain). Bioinformatic support for annotation and release of the genome was provided in part by funds (2009-65109-05719) awarded to C.R. Buell and N.T. Perna from the US Department of Agriculture National Institute of Food and Agriculture. M. Lindeberg (Cornell University, USA) is thanked for technical advice with the nomenclature of Hop genes. The authors thank Theresa H. Osinga for critical reading of the manuscript and advice on English usage.