pA506, a conjugative plasmid of the plant epiphyte Pseudomonas fluorescens A506

Abstract

Conjugative plasmids are known to facilitate the acquisition and dispersal of genes contributing to the fitness of Pseudomonas spp. Here, we report the characterization of pA506, the 57-kb conjugative plasmid of Pseudomonas fluorescens A506, a plant epiphyte used in the United States for the biological control of fire blight disease of pear and apple. Twenty-nine of the 67 open reading frames (ORFs) of pA506 have putative functions in conjugation, including a type IV secretion system related to that of MOBP6 family plasmids and a gene cluster for type IV pili. We demonstrate that pA506 is self-transmissible via conjugation between A506 and strains of Pseudomonas spp. or the Enterobacteriaceae. The origin of vegetative replication (oriV) of pA506 is typical of those in pPT23A family plasmids, which are present in many pathovars of Pseudomonas syringae, but pA506 lacks repA, a defining locus for pPT23A plasmids, and has a novel partitioning region. We selected a plasmid-cured derivative of A506 and compared it to the wild type to identify plasmid-encoded phenotypes. pA506 conferred UV resistance, presumably due to the plasmid-borne rulAB genes, but did not influence epiphytic fitness of A506 on pear or apple blossoms in the field. pA506 does not appear to confer resistance to antibiotics or other toxic elements. Based on the conjugative nature of pA506 and the large number of its genes that are shared with plasmids from diverse groups of environmental bacteria, the plasmid is likely to serve as a vehicle for genetic exchange between A506 and its coinhabitants on plant surfaces.

Fig 1

Map of pA506. The outer circle shows predicted coding sequences color coded by putative function: dark green, plasmid replication and stability; green, UV resistance; purple, site-specific recombinases; orange, regulation; black, hypothetical proteins; gray, pyocin immunity protein; blue and turquoise, conjugation. Annotated functions or locus tags (i.e., PflA506_p0001 is shown as 1) are shown outside the circle. Intergenic regions with putative functions as the origin of transfer (oriT), origin of vegetative replication (oriV), and partitioning centromere (par) are shown in red. Circles 2 through 8 depict regions of pA506 that are conserved in seven other plasmids as determined by tBLASTn (cutoff of 1e−5). Second circle (lavender), pMPR124 from P. fluorescens; third circle (lavender), PAPI-1 from P. aeruginosa PA14; fourth circle (yellow), pMA4326A from P. syringae pv. maculicola; fifth circle (yellow), pPSR1 from P. syringae pv. syringae; sixth circle (yellow), the small plasmid from P. syringae pv. phaseolicola 1448; seventh circle (gray), pEU30 from E. amylovora; eighth circle (gray), pET49 from E. tasmaniensis. Within each circle, the darkest color indicates nucleotide identity exceeding 40% whereas the lightest color represents identity exceeding 20%. Ninth circle, G+C content. Tenth circle, GC skew. Eleventh circle, coordinates in kilobase pairs. The circular plasmid diagram was generated using BRIG (74).

Fig 2

Identification of oriV of pA506. A map of the region of pA506 extending from coordinate 56556 to 1797 is shown, with coding sequences for PflA506_p0001 (01) and PflA506_p0002 (02) depicted as arrows and intergenic regions depicted as a black line. A 190-bp region conserved with pPSR1 is shown as a gray box, and a sequence that could form a stem-and-loop structure is also shown. Black lines below the map show the sequences that were cloned in pCR2.1-TOPO. The ability of the clones to replicate in A506Δ-pA506 and the sizes of the cloned sequences are shown. The recircularized pCR2.1-TOPO vector did not replicate in A506Δ-pA506.

Fig 3

Evidence for conjugative transfer of pA506 from A506 to Pseudomonas fluorescens SS101 and Pantoea vagans C9-1 and maintenance of pA506 in those bacterial hosts. BamHI digest of plasmids isolated from wild-type strains and recipients with pA506::gfp-km-gm. Lanes: M, 1-kb plus marker; 1, pA506; 2, pA506::gfp-km-gm; 3, P. fluorescens SS101; 4, P. fluorescens SS101 pA506::gfp-km-gm; 5, P. vagans C9-1; 6, P. vagans C9-1 pA506::gfp-km-gm. P. fluorescens SS101 does not have a plasmid. P. vagans C9-1 has 3 megaplasmids that were not isolated with the alkaline lysis method.

Fig 4

The putative partitioning region of pA506. (A) A map of the region of pA506 extending from coordinates 12602 to 14265 is shown, with coding sequences depicted as open arrows and the intergenic regions depicted as a black line. The enlargement shows 313 bp of the intergenic region between PflA506_p0020 and PflA506_p0021. Each pentagon represents one of the repeated six-nucleotide sequences (GAATTC) present in the region, which are separated by 4- to 7-bp spacer sequences with low G+C content. Three longer 18-bp spacer sequences with low G+C content (35 to 39%) are depicted with shaded boxes. (B) The sequence motif logo was generated from the conserved six-nucleotide repeat regions depicted as pentagons in panel A (http://weblogo.berkeley.edu/).

Fig 5

Relaxase-based phylogeny of pA506 and other plasmids in the MOB_P6 family. The N-terminal 350 aa from PflA506_p0066 were used to identify novel MOB_P6 family plasmids, as previously described (35). Relaxase sequences from known MOB_P5, MOB_P6, and MOB_P7 family plasmids were also used in generation of this tree. Sequences were aligned by MAFFT L-INS-I, and the LG+G substitution model was identified as the best model using the BIC selection in ProtTest 3. The maximum-likelihood tree was generated using RAxML with the above data as input, and the interior node values indicate the percentage of bootstraps out of 1,000. Bootstrap values below 50 are not shown. Sequences that were significantly shorter than the 350 aa used in the BLASTP search were removed prior to alignment and tree generation. MOB_P6 family plasmids are in Pseudomonas spp. (circles) or species of the Enterobacteriaceae (diamonds), other gammaproteobacteria (squares), or betaproteobacteria (pentagons).

Fig 6

Influence of pA506 on twitching motility of A506. Twitching motility of Pseudomonas fluorescens strain A506 (■) and the plasmidless derivative A506-ΔpA506 (□) was assessed on King's medium B solidified with 1% agar. The motility zone was measured over 24 h at the interface of the petri dish and the bottom of the medium. The mean diameter of motility was calculated from five replicate plates per strain; the vertical lines indicate 1 standard error of the mean. Asterisks show time points when the motility of the two strains differed significantly according to Fisher's protected least significant difference at P = 0.05.

Fig 7

pA506 increased tolerance of A506 to UV irradiation. Tolerance of Pseudomonas fluorescens strain A506 (■) and the plasmidless derivative A506-ΔpA506 (□) to UV irradiation was assessed by exposing cells spread on medium to various doses of UVC. Colonies were counted after a 3-day incubation period in the dark. The survival ratio is the population size after exposure to UV divided by the population that was not exposed. The results were averaged from 5 replicate experiments. Vertical bars represent 1 standard error of the mean. Asterisks indicate a significant difference (Fisher's protected least significant difference, P = 0.05) between the mean log (survival ratio) of A506 and its plasmidless derivative at various UVC doses.

Fig 8

pA506 did not have a detectable influence on epiphytic fitness of Pseudomonas fluorescens A506 on pear or apple flowers. Bacterial populations were estimated from flowers on Bartlett pear (A), Gala apple (B), and Red Delicious apple (C) trees at the Botany and Plant Pathology Field Laboratory of Oregon State University near Corvallis, OR. Open flowers were sprayed with 1 × 10⁸ CFU/ml suspensions of A506 (■) or the plasmidless derivative A506-ΔpA506 (□). At 2 to 9 days following inoculation, culturable populations of bacterial strains on the pistillate structures of flowers were estimated by spreading dilutions of tissue washes on selective media. Vertical bars represent 1 standard error of the mean. No significant differences (Fisher's protected least significant difference, P = 0.05) between mean population sizes of A506 and its plasmidless derivative were observed.