Molecular and biological characterization of ϕRs551, a filamentous bacteriophage isolated from a race 3 biovar 2 strain of Ralstonia solanacearum

A filamentous bacteriophage, designated ϕRs551, was isolated and purified from the quarantine and select agent phytopathogen Ralstonia solanacearum race 3 biovar 2 strain UW551 (phylotype IIB sequevar 1) grown under normal culture conditions. Electron microscopy suggested that ϕRs551 is a member of the family Inoviridae, and is about 1200 nm long and 7 nm wide. ϕRs551 has a genome of 7929 nucleotides containing 14 open reading frames, and is the first isolated virion that contains a resolvase (ORF13) and putative type-2 phage repressor (ORF14). Unlike other R. solanacearum phages isolated from soil, the genome sequence of ϕRs551 is not only 100% identical to its prophage sequence in the deposited genome of R. solanacearum strain UW551 from which the phage was isolated, but is also surprisingly found with 100% identity in the deposited genomes of 10 other phylotype II sequevar 1 strains of R. solanacearum. Furthermore, it is homologous to genome RS-09-161, resulting in the identification of a new prophage, designated RSM10, in a R. solanacearum strain from India. When ORF13 and a core attP site of ϕRs551 were either deleted individually or in combination, phage integration was not observed, suggesting that similar to other filamentous R. solanacearum ϕRSM phages, ϕRs551 relies on its resolvase and the core att sequence for site-directed integration into its susceptible R. solanacearum strain. The integration occurred four hours after phage infection. Infection of a susceptible R. solanacearum strain RUN302 by ϕRs551 resulted in less fluidal colonies and EPS production, and reduced motilities of the bacterium. Interestingly, infection of RUN302 by ϕRs551 also resulted in reduced virulence, rather than enhanced or loss of virulence caused by other ϕRSM phages. Study of bacteriophages of R. solanacearum would contribute to a better understanding of the phage-bacterium-environment interactions in order to develop integrated management strategies to combat R. solanacearum.

Introduction from a frozen stock onto triphenyltetrazolium chloride plates [21]. Then, a single colony was picked and grown overnight in casamino acid peptone glucose (CPG) broth [22] at 28˚C with shaking. To prepare R. solanacearum inocula, appropriate concentrations of the bacterial suspensions were made in sterile water using OD 600 as an initial measurement of cell density. Final inoculum cell density was confirmed by ten-fold serial dilution plating.

Bacteriophage isolation, purification, and characterization
A bacteriophage designated ϕRs551 was isolated and purified directly from an overnight culture of R. solanacearum strain UW551 grown in CPG broth at 28˚C with shaking, and used for further study. A total of 2 liters of the bacterial culture was grown in order to obtain a sufficient amount of phage particles for morphological characterization and for DNA extraction. The overnight cell culture of UW551 was pelleted at 8000 g for 15 minutes at 18˚C, and the supernatant was filtered through 0.2 μm membrane filters, followed by ultracentrifugation at 109,000 g through a 30% sucrose cushion for 2 hours at 10˚C. The phage pellet was resuspended in SM buffer containing 50 mM Tris/HCl at pH 7.5, 100 mM NaCl, 10 mM MgSO4, and 0.01% gelatin [23], stored at 4˚C in complete darkness, and used for DNA extraction and morphological characterization. To characterize the phage, the purified phage particles were stained with sodium phosphotungstate using the method of Dykstra [24] prior to observation under a Hitachi HT7700 transmission electron microscope. The size of ϕRs551was estimated from 20 phage particles based on scale bars.

Phage susceptibility test
To determine the host specificity of ϕRs551, the purified phage was subject to the spot test and plaque-forming assay using R. solanacearum strains in Table 1 as hosts. This was done by mixing 500 μl of R. solanacearum cells (OD 600 of 0.2) with 3.5 ml of CPG containing 0.45% agar and layering the mixture on top of a CPG plate containing 1.5% (w/v) agar. After the top layer with R. solanacearum was hardened, 3 μl of the purified phage suspension was spotted on the double-layered plate, incubated at 28˚C for 24 h for the presence or absence of plaques. Phage DNA extraction, cloning, sequencing, analysis and characterization Phage DNA was extracted from purified phage (described above) using a phenol-chloroform method [23]. The near-entire full length circular phage genome (7694 out of 7929 bases) was cloned into the Zero Blunt vector (Life Technologies, Inc.) using primers F-ϕRs551 and R-ϕRs551 (Table 2) designed on the prophage region in UW551 contig0570. The genomic DNA sequence of ϕRs551 was determined by primer walking the cloned 7694-bp genome fragment. The F-ϕRs551 and R-ϕRs551 primer sequences were confirmed by sequencing purified phage DNA, and the remaining 235 bp not cloned into the vector were sequenced directly from purified phage DNA to close the genome. Potential ORFs of ϕRs551 were identified using Gene-MarkS [25] or the online program ORFfinder (http://www.ncbi.nlm.nih.gov/orffinder), and were compared with those of other R. solanacearum related phages by BLASTP. To assign possible functions to the ORFs and to identify putative attP region, database searches were performed using FASTA, FASTX, BLASTN and BLASTX programs [26]. To determine the nature of the ϕRs551 genome, approximately 300 to 500 ng of phage genome was subjected to individual enzyme digestions for 30 minutes in a 30-μl reaction volume with 1 unit of DNase I, 3 units of RNase A, 10 units of exonuclease I at 37˚C, or 10 units of S1 nuclease at room temperature (Thermo Fisher Scientific, Waltham, MA.).
Infection of susceptible R. solanacearum strain RUN302 with ϕRs551 To infect susceptible R. solanacearum strain RUN302, which did not contain ϕRs551 sequence in its genome before infection, a single-plaque isolate of ϕRs551 was used. Its titer was determined by mixing 100 μl of the serially diluted phage filtrate with 400 μl of 2 x 10 8 cells of RUN302 for the plaque-forming assay as described above. For propagation and purification, one milliliter of the overnight culture of RUN302 was diluted with 100 ml of fresh CPG in a 500 ml flask. When the culture reached an OD 600 of 0.2, ϕRs551 was added at a multiplicity of infection (moi) of 0.001-1.0. After further growth for 24-48h at 28˚C with shaking, the phage was purified as described above. Extrachromosomal DNA was isolated from phage-infected R. solanacearum cells by the minipreparation method [27].
In vitro growth of R. solanacearum strains To study the effect of phage infection on the in vitro growth of its susceptible R. solanacearum strain, ϕRs551-infected and uninfected R. solanacearum RUN302 strains were grown overnight in 5 ml of CPG broth. One milliliter of the overnight culture (adjusted to OD 600 of 0.003) was then transferred into a well of a 24-well sterile microplate, and was grown at 28˚C with shaking at 200 rpm inside of a Epoch2 microplate reader (BioTek Instruments, Winooski, VT). The absorbance at 600 nm was measured and graphed every three hours over the course of 48 hours. There were three replicates for each time point per strain, and the experiments were repeated three times.

Construction of phage mutants
To determine the role of a core 13-nucleotide attP sequence and ORF13 in the integration of ϕRs551 into R. solanacearum susceptible strain RUN302, three ϕRs551 mutants were constructed: 1) ϕΔattP missing a 395-bp fragment containing the core attP sequence, 2) ϕΔorf13 with a 570-bp deletion including almost all the ORF 13 sequence (190 out of 196 amino acids), and 3) ϕΔattP-Δorf13 missing a 655-bp region containing both the attP and 582 of the 591 nucleotides of the ORF 13 gene sequences. The mutants were generated by PCR amplification of shortened phage regions using primer pairs ( Table 2) designed based on the sequence of ϕRs551. As a positive control, the full length ϕRs551 was also amplified by PCR using primer pair P6-F/P7-R ( Table 2). The PCR product was then purified and circularized with a T4 DNA ligase (Promega Corp., USA). The resulting circularized phage mutant or full length phage DNA was electroporated into competent cells of the wild type strain RUN302 using a Micro-Pulser Electroporator (Bio-Rad, Hercules, CA) with settings of 1.5 KV and 4 milliseconds. This was done by mixing 45 μl of the competent cells with 2μl of the circularized DNA (approximately 40 ng) and placing the mixture in a 0.1 cm-gap electroporation cuvette (Bio-Rad) pre-chilled on ice. The mixture in the cuvette was washed immediately after electroporation with 1 ml of CPG broth and incubated at 28˚C for at least 7 hours. Electro-competent cells were prepared by growing overnight cultures of R. solanacearum to an OD 600 between 0.4 and 0.7, followed by centrifugation at 4182 x g for 10 minutes and four successive washes at 18˚C in 1, 0.5, 0.1 and 0.01 volumes of 10% glycerol [28]. The knocked-out regions in the mutants was confirmed by PCR using primer pairs located within the sequences of the mutated regions that showed a lack of any amplified products.

Determination of the phage integration mechanism
To determine the right and left integration flanking regions (attR and attL) in R. solanacearum Run302 after ϕRs551 integrated into the bacterial genome at a potential attB site, total genomic DNA was extracted from infected RUN302 cells using the DNeasy Blood and Tissue kit (Qiagen, Inc.), and amplified using primer pairs P8/P9 and P10/P11 (Table 2). Since the genome sequence of RUN302 is not available, the P8 and P11 primers targeting R. solanacearum RUN302 sequences were designed based on the potential dif (attB) sequence identified in the fully sequenced R. solanacearum strain P082 [29], a closely related strain which is also a biovar 1, phylotype IIB, sequevar 4 strain like RUN302. The PCR products were purified from agarose gels with the QIAquick gel extraction kit (Qiagen, Inc.) and sequenced commercially.
To determine the time needed for ϕRs551 to integrate into the RUN302 genome, total genomic DNA of ϕRs551-infected RUN302 was extracted every hour for seven hours for PCR determination of the attR and attL regions as described above. As a control, the DNA was also amplified by PCR using primers P12 and P13, designed based on R. solanacearum strain P082, to detect the host bacterial genome.
To determine whether phage integration was affected by mutations in attP and orf13, total genomic DNA from RUN302 infected by each of the three phage mutants constructed above was purified and subjected to PCR using the P8/P9 and P10/P11primers pairs. The primer pair P14/P15 was also used to determine whether circular episomal forms of ϕRs551 were produced. Uninfected wild type RUN302 genomic DNA was used as a control in all the experiments.

PCR conditions
PCR using primer pairs P8/P9, P10/P11, P12/P13 or P14/P15 was performed in a 20-μl volume containing 1x GoTaq Green Master Mix (Promega, Madison, WI) and 5 pmol of each primer. Twenty nanograms of DNA was added to the reaction mixture. PCR conditions were 1 cycle of 4 min at 94˚C, 30 cycles of 1 min at 94˚C, 1 min at 60˚C, and 1 min at 72˚C, with a final extension of 10 min at 72˚C. For PCR using primer pairs of P1/P2, P3/P4 or P5/P4, 1x KAPA HiFi HotStart ReadyMix (KAPA Biosystems, Boston, MA) was used and the PCR conditions were 1 cycle of 3 min at 95˚C, 30 cycles of 20 sec at 98˚C, 15 sec at 62˚C, and 2.5 min at 72˚C, with a final extension of 8 min at 72˚C. For PCR using primer pair P6/P7, the PCR conditions were the same as P1/P2, P3/P4 or P5/P4, except that the annealing temperature was 59˚C. Five microliters of each PCR product was mixed and visualized by electrophoresis in 1.0% agarose gels stained with GelRed (Phenix Research Products, Candler, NC).

Extracellular polysaccharide (EPS) assay
EPS in bacterial culture supernatants was determined quantitatively using a modified method of Ahmad et al. [30] and Jeong et al. [31]. Briefly, bacterial strains were grown in CPG medium for 72h at 28˚C with shaking at 200 rpm. The optical density (OD) at 600 nm was measured for each bacterial suspension and adjusted to the same OD 600 with liquid CPG. Ten milliliters of each strain were then centrifuged at 5000 × g for 20 min at 4˚C, and the supernatant was collected and passed through 0.22 um membrane. The filtrate was mixed with 4 volumes of acetone and a final concentration of 20 mM NaCl, and the mixture was kept overnight at 4˚C. To determine the dry weight of EPS, the precipitated EPS was collected by centrifugation and dried at 55˚C prior to measurement. Two replicates were used for each strain and the experiment was repeated three times.

Motility assay
Swimming and swarming motilities of the wild type strain RUN302 of R. solanacearum were examined on CPG containing 0.3% and 0.7% (w/v) agar, respectively, and compared to those of ϕRs551-infected RUN302 strain. Overnight cultures of the RUN302 strains in CPG were centrifuged at 8000 g for 2 min at 4˚C, washed twice with sterile water, and resuspended in water to an OD 600 of 0.1, approximately 10 8 cells per milliliter. Three microliters of the suspension were spotted in the center on an agar plate containing 20 ml of CPG and appropriate amount of agar, and incubated at 28˚C. The plates were photographed 5 days after incubation. For twitching motility assays, 3 μl of the suspension were spotted in the center on a minimal medium plate [19], incubated for 5 days at 28˚C. Twitching motility was then visualized by placing the plate without its lid on the stage of a light microscope (Carl Zeiss Microscope GmbH, Germany) under 40× magnification, photographed using OMAX digital camera (OMAXmicroscope.com), and saved to OMAX ToupView software in gray scale as uncompressed TIFF files. Colonies with twitching motility were characterized by the formation of corrugated trajectories around the colonies.

Virulence assay
Seed grown tomato plants (Lycopersicom esculentum Mill. cv. 'bonnie best') were transplanted approximately 7-10 days after germination, and inoculated two to seven days after transplanting, using the soil drenching method to mimic the natural infection process and maintained as described before [32], except that 50 ml suspensions with 5 x 10 7 cells of R. solanacearum were poured into each pot. Water was used as a negative control. Inoculated plants were rated daily for three weeks using a disease index of 0 to 4 [33]. There were 10 plants per treatment and the experiment was repeated three times.

Statistical analysis
The means of growth rate, EPS dry weight, and disease index between the wild-type and ϕRs551-infected R. solanacearum RUN302 strains were analyzed for significant differences using the t test in Microsoft Excel.

Nucleotide sequence accession numbers
The genome sequence of ϕRs551 has been submitted to GenBank and given accession no. KX179905. The accession numbers for M13 phage and for other R. solanacearum filamentous phage sequences used in this study are: M13, NC_003287; RSM3, AB434711; and RS603, AB937974.

Morphology and host specificity of ϕRs551
When the supernatant of the overnight culture of R. solanacearum strain UW551 grown under normal growth conditions was purified, a flexible filamentous particle, resembling the members of Inoviridae, was observed under the electron microscope, and designated ϕRs551 (Fig  1). This bacteriophage has an average size of~1200 nm in length and~7 nm in width, similar in morphology to other bacteriophages of R. solanacearum isolated from soil, such as ϕRSM1 (1500±300 nm in length, 6±0.7 nm in width) [8] and ϕRS603 (1120 nm in length, 8 nm in width) [9].
To determine the host specificity of ϕRs551, 12 R. solanacearum strains including nine r3b2 and three non-r3b2 strains originally isolated from different parts of the world were tested for their susceptibility to the phage (Table 1). ϕRs551 infected two of the nine r3b2 strains, RUN628 and 4155, and one of the three non-r3b2 strains, RUN302 (Table 1), as evidenced by the formation of plaques on CPG plates layered first with the tested R. solanacearum strains and spotted later with ϕRs551.

Genome characterization of ϕRs551
The genome of ϕRs551 was degraded by DNase I, but not by RNase A. It was also degraded by S1 nuclease, specific for circular single stranded nucleic acids, but not by Exonuclease I, which cleaves nucleotides from the end of a polynucleotide chains. The complete nucleotide sequence of ϕRs551 was determined (S1 Fig) and submitted to GenBank (accession no. KX179905). The genome consists of 7929 nucleotides and has a G+C content of 61%. Fourteen putative open reading frames (ORFs) were identified in ϕRs551 and compared by BLASTP to genome sequences of other R. solanacearum filamentous phages including ϕRSM1, ϕRSM3, and ϕRS603 ( Table 3). Each of the ORFs is capable of encoding a protein with a molecular mass of greater than 7.5 kDa. The deduced ORFs, their positions, putative functions, and identities to the closest R. solanacearum phage homologues are summarized in Table 3. The predicted ORFs could be arranged in a similar modular structure to that of previously characterized R. solanacearum filamentous phages of the Ff group ϕRSM3 and ϕRS603 [9,20], with structural differences in two variable regions; one region located between the replication module and the structural module and the second within the replication module (Fig 2). Between the replication module and structural module, ϕRs551 has only one predicted ORF, similar to ϕRS603, but different from ϕRSM3 which has two ORFs (Fig 2). Within the replication module, Filamentous phage ϕRs551 isolated from a r3b2 strain of Ralstonia solanacearum ϕRs551 has three reverse-oriented ORFs, 12, 13 and 14, as in ϕRSM3 (ORFs 13, 14 and 15). This is different from ϕRS603, which only has one reversed ORF, ORF13, corresponding to the ORF14 of ϕRs551 and the ORF 15 of ϕRSM3. ORF13 of ϕRs551 was annotated as a putative resolvase (ORF14 in ϕRSM3) that ϕRS603 does not have. ORF14 in ϕRs551 was annotated as a putative phage repressor, similar to the annotation for the ORF13 in ϕRS603 and the ORF15 in ϕRSM3, but the ORF14 in ϕRs551 shared 81% homology to ORF13 in ϕRS603, and no significant homology to ORF15 in ϕRSM3.
Presence of ϕRs551 genome in R. solanacearum strains and Phylogenetic analysis with well-identified R. solanacearum filamentous phages The ϕRs551 nucleotide (nt) sequence was found with 100% identity in 11 deposited genomes of R. solanacearum strains: UW551, UY031, IPO1609, NCPPB909, CFIA906, UW24, UW365, RS2, Geo_56, Geo_96, and Geo_99. The first eight strains were deposited as r3b2 strains, although the race and biovar nature of the last three unpublished strains was unknown. Such phage sequence, however, was not found in two other R. solanacearum strains deposited as r3b2: NCPPB282 and 23-10BR. Askora et al. [15] previously identified a prophage sequence in UW551, putatively naming it ϕRSM4, and similarly predicated the ORFs [15,20], but no research was conducted beyond its annotation. We named our ORFs in ϕRs551 consecutively from ORF1 to 14 and annotated ORF2, ORF3 (ORF4 in ϕRSM4), and ORF13 (ORF14 in RSM4) to have 81, 92, and 196 instead of 59, 98, and 237 amino acids, respectively. ϕRs551 is most closely related to the prophage RSM6 identified in the phylotype III R. solanacearum strain CMR15 [20] and the phage ϕRS603 [9], matching 89% and 83% of the genome with 88% and 82% identity at the nt level, respectively. ϕRs551 differs from the prophage RSM6 in the middle of ORF 8, which could have an impact on host range [15], as well as in ORFs 2, 3, and part of the intergenic region. The difference between ϕRs551 and ϕRS603 is that ϕRs551 has an annotated resolvase that is not found in ϕRS603 (ORF13), but is missing a second ORF in the A-S module (Fig 2) that ϕRS603 has in common with ϕRSS phages [9]. Interestingly, the nt sequence of ϕRs551 also matched genome RS-09-161, a R. solanacearum strain isolated from India [34], with 84% identity to 74% coverage of ϕRs551. Inspection of this homologous RS-09-161 sequence revealed a previously unidentified prophage of 9947 nt (position 3189 to 13135 in GenBank accession number JHBO01000069) with 17 predicted ORFs, which is two to four more ORFs than the other R. solanacearum Ff-type phages including ϕRSM, ϕRs551 and RS603 (Fig 2). We designated the newly identified prophage ϕRSM10 to continue the naming and numbering system used by Yamada's group for prophages identified previously in Ralstonia species [20]. The ϕRSM10 sequence has an insertion in the replication module that is similar in structure to ϕRSM3 and ϕRs551 (Fig 2). The structure of its A-S region is similar to ϕRS603 and ϕRSS phages [9]. Also, there is an additional ORF (ORF 16) (position 4188 to 4808 in JHBO01000069) (Fig 2) between the replication module and structural module that is not found in any other ϕRSM or ϕRSS phage and has 85% identity to a Filamentous phage ϕRs551 isolated from a r3b2 strain of Ralstonia solanacearum hypothetical protein in Methylomicrobium album BG8 (Accession number EIC30310). ϕRSM10 also differs from ϕRs551 in the repressor region, where ORF15L in ϕRSM10 (position 3240 to 3554 in JHBO01000069) is 100% identical to ORF15 in ϕRSM3 (Fig 2).

Integration mechanism of ϕRs551
The discovery that the full sequence of ϕRs551 was present in 11 deposited genomes of R. solanacearum strains in GenBank suggests that ϕRs551 has a lysogenic ability and is able to integrate into the host genome. By BLAST search, we identified a possible dif-like sequence for attP in ϕRs551 that contains a 13-bp core sequence, 5 0 TGGCGGAGAGGGT 3 0 , corresponding to nucleotides 7410 to 7422 in ϕRs551, as well as a putative resolvase encoded by ORF13. This core sequence showed 100% identity to the attP sequences of other Ralstonia RSM phages including ϕRSM3 and our newly identified prophage ϕRSM10 (Fig 3A), as well as the attB sequence of R. solanacearum strains such as RUN302 (Fig 3A). To confirm that this core sequence is involved in the integration of ϕRs551 into its susceptible host strain RUN302, we amplified and determined the left and right integration flanking regions, attL and attR, by PCR in ϕRs551-infected Run302 (Fig 3A and 3B). Both the attL and attR fragments contained the 13-nt core att sequence that is in both the bacterial RUN302 and the viral ϕRs551 genomes (Fig 3A). The attL site had RUN302 genome sequence upstream and ϕRs551 sequence downstream of this core sequence. Reversely, the attR fragment contained ϕRs551 sequence upstream and Run302 sequence downstream of the core sequence, suggesting that the core sequence is where the phage integration happened.
The time needed for ϕRs551 to integrate into the RUN302 genome was determined to be between four to five hours after phage infection, since the integration flanking regions attL and attR were not amplified until five hours after RUN302 was infected by ϕRs551 (Fig 3C). A 147-bp bacterial genome target, however, was detected in the wt and ϕRs551-infected RUN302 genome regardless of phage integration (Fig 3C).
To determine the role of ORF13 and to further investigate the importance of the core att sequence (attP) in the integration of ϕRs551 into its susceptible R. solanacearum host strain RUN302, two single phage mutants, ϕΔattP missing the attP and ϕΔorf13 missing the ORF13 gene, and one double mutant, ϕΔorf13&ΔϕattP missing both the attP and the ORF13 gene, were generated (Fig 3D). None of the phage mutants integrated into the genome of RUN302 after infection, since no attL and attR fragments were detected by PCR using primer pairs P10/P11 and P8/P9 (Fig 3E). Instead, they replicated as circular episomal forms, since they were amplified by the primer pair P14/P15 that target ϕRs551 regions away from the mutated regions, and yield PCR products only if the phages were circular and outside of the bacterial genome (Fig 3E). Both the wild type ϕRs551 and the ϕRs551 generated by PCR existed as both lysogenic and episomal forms, as evidenced by amplification of attL, attR and the partial episomal ϕRs551 by primer pairs P10/P11, P8/P9 and P14/P15 (Fig 3E).

Physiological effects of ϕRs551 on R. solanacearum strain RUN302
To determine the effects of ϕRs551 infection on R. solanacearum strain RUN302, we first compared the in vitro growth of the wild type (wt) with that of the ϕRs551-infected strain RUN302. The growth curve of the two strains was measured every hour for 48 hours, and graphed every three hours (Fig 4A). Although the wt RUN302 grew faster during the exponential growth period (9 to 27 hours) than the phage-infected strain, the difference was not significant at all but one (12 h) time point. Overall, the two strains grew at a similar rate during the entire 48 hour period (Fig 4A). When the two strains grew on regular TZC medium plates, the colonies of the wt strain grew irregular and fluidal, while the phage-infected strain appeared circular and dry (Fig 4B), suggesting low production of EPS. This was confirmed by an EPS quantitative assay that showed ϕRs551-infected cells produced significantly lower amounts of EPS (71.5±3 mg/10 ml) than the wt cells (107±5 mg/10 ml). The two strains also displayed distinct difference in motilities when compared using the twitching, swimming and swarming motility assays (Fig 4C, 4D and 4E). For the wt strain, under a microscope, we observed the formation of corrugated trajectories around the edge of its colonies, indicating twitching motility (Fig 4C, left). Such trajectories, however, were largely missing in the phage-infected RUN302 strain and the edge of its colonies was smooth (Fig 4C, right). Our assay for swimming activity revealed that the wt RUN302 expanded almost to the entire plate five days after The attP sequences identified in ϕRs551 and the prophage RSM10 are compared with that in ϕRSM3 of R. solanacearum. The 13-nucleotide core att sequence is shown in red. For prophage RSM10 and ϕRSM3, only nucleotides different from the ones in ϕRs551 are shown. The att sequences before (attB) and after ϕRs551 integration into R. solanacearum strain RUN302 (attL and attR) were shown. Sequences from the bacterial strain RUN302 and the phage ϕRs551 are shown in green and black, respectively. (B) After integration, the presence of the left (attL) and right (attR) integration flanking regions in ϕRs551-infected RUN302 were amplified by PCR with primer pairs P8/P9 and P6/P7, respectively, and separated by gel electrophoresis. (C) The attL and attR regions were amplified by PCR every hour for seven hours after infection of R. solanacearum RUN302 with ϕRs551 to determine the time needed for phage integration to occur. Note that no attL and attR fragments were detected in the uninfected wild type RUN302 strain, although the R. solanacearum (Rs) fragment was detected in all strains/time points by PCR with primer pair P10/P11. (D) PCR amplification generated ϕRs551 mutants consisting of shortened phage regions using primer pairs: P1-F/P2-R to obtain ϕΔattP missing a 395-bp fragment containing the core attP sequence, P3-F/P4-R to obtain ϕΔorf13 missing 570 of the 591 nucleotides of the orf13 gene sequence, and P5-F/P4-R to obtain ΔattP-Δorf13 missing a 655-bp region containing both the core attP and 582 nucleotides of the orf13 gene sequences. The primer sequences are in bold, and their directions are indicated by arrows. The orf13 gene and the 13-nucleotide core attP sequences are boxed. Five hundred and sixty nucleotides between P4-R and P3-F, and 336 between P5-F and P1-F are not displayed. (E) Deletion of ϕRs551's attP or/and ORF13 (the integrase) abolished its ability to integrate in the genome of R. solanacearum strain RUN302. Lanes: RUN302 infected by the wild type ϕRs551 (1), the single mutant phage ϕΔattP (2), ϕΔORF13 (3), the double mutant phage ϕΔattP&ΔORF3 (4), and the full-length phage ϕRs551 generated by PCR (6), as well as uninfected wild type RUN302 (5). M, 1 kb DNA ladder. Arrows indicate the fragments for attL, attR, Rs and the partial episomal phages. https://doi.org/10.1371/journal.pone.0185034.g003 Filamentous phage ϕRs551 isolated from a r3b2 strain of Ralstonia solanacearum inoculation of 3 μl of the bacterium in the center of the plate (Fig 4D, left), while the phageinfected RUN302 strain did not expand to even half of the plate radius (Fig 4D, right). Similarly, in our swarming assay, the wt RUN302 strain was observed to swarm over half of the plate radius (Fig 4E, left), while the phage-infected strain only moved approximately one third of the distance as compared to its wt (Fig 4E, right).

Effects of ϕRs551 on the virulence of R. solanacearum strain RUN302
To study the effect of ϕRs551 on the virulence of its susceptible strain RUN302 of R. solanacearum, we compared the virulence of the wild type to that of the ϕRs551-infected RUN302 of R. solanacearum. The wild type strain RUN302 started to cause disease symptoms four days after soil drenching inoculation (disease index (DI) > 0), reached a DI of 2 (50% of leaf area wilted) in fewer than six days, and completely wilted all inoculated plants (DI = 4) by day 11 (Fig 5). On the contrary, the virulence level caused by ϕRs551-infected RUN302 strain of R. solanacearum was significantly lower (Fig 5). It did not cause any disease symptoms until five days after inoculation, took more than 14 days to cause a DI of 2, and was unable to cause any disease symptoms on some inoculated plants even 21 days after soil drenching inoculation (Fig 5). Filamentous phage ϕRs551 isolated from a r3b2 strain of Ralstonia solanacearum

Discussion
We purified an Ff-like filamentous phage ϕRs551 from the supernatant of R. solanacearum strain UW551 grown under normal growth conditions. We also characterized ϕRs551's morphology and genome, and determined its integration mechanism and effect on the morphology, EPS production, in vitro growth, motility and virulence of its host bacterial cells. Different from previously isolated filamentous R. solanacearum phages, ϕRs551 is the first active virion identified and purified directly from the supernatant of a normally grown R. solanacearum strain without any form of induction [35], rather than from the soil of a crop field like ϕRSM1 and ϕRSS1 [8] or regenerated like ϕRSM3 from a prophage in a lysogenic R. solanacearum strain [15]. In addition, the genome sequence of ϕRs551 was found with 100% identity to its prophage sequence in the deposited genome of R. solanacearum strain UW551 from which the phage was purified. Surprisingly, when the genome sequence of ϕRs551 was used as a query sequence to blast GenBank's nucleotide collection, NCBI genomes and whole genome Filamentous phage ϕRs551 isolated from a r3b2 strain of Ralstonia solanacearum shotgun contigs databases, the sequence was also found with 100% identity in the genomes of ten other R. solanacearum strains sequenced and deposited independently by different research groups: Geo_57, Geo_96, Geo_99, UY031, IPO1609, NCPPB909, CFIA906, UW24, UW365, and RS2. The race and biovar nature of strains Geo_57, _96 and _99 were not given in GenBank, but the other eight strains including UW551 are all deposited as r3b2 strains. All 11 strains, however, belong to phylotype IIB, sequevar 1 as confirmed by Stulberg et al. [32] for UW551 and IPO1609, reported by Clarke et al. [36] for NCPPB909, POPS2 (alternative ID of UW365) and RS2, and determined in this study for Geo_57, Geo_96, Geo_99, UY031, CFIA906 and UW24 using a R. solanacearum typing computer program [37]. Such phage sequence, however, was not found in two other genomes deposited as r3b2 strains: NCPPB282 and 23-10BR, although the former belongs to phylotype IIB sequevar 2 [36] and the latter with an alternate ID of UW349 was determined previously by Stulberg et al. [32] to be an atypical biovar 2 strain, belong to phylotype IIB sequevar 27 and not as cold virulent as UW551. The  [36], suggesting that the phage infection might have occurred in a common ancestor of the R. solanacearum strains and been stably maintained over the years after the strains were disseminated in Asia, Africa, Europe and South America. It is also possible that ϕRs551 released from the ancestral lysogenic R. solanacearum strain infected local susceptible R. solanacearum strains when the ancestral strain was disseminated into different geographic locations. This is different from other R. solanacearum filamentous phages, where their prophage sequences in R. solanacearum were similar or closely related to the phages, but not exactly the same. For example, the prophage ϕRSM3's sequence was identified in R. solanacearum strain MAFF730139, and is 75 bp shorter than and had 93% nucleotide identity to the isolated phage ϕRSM1 [15]. Currently, a leading hypothesis of the origins of r3b2 strains posits that they disseminated out of South America and are largely clonal [36,38]. Since ϕRs551 has been identified in the genomes of sequevar 1 strains, its presence or absence might be useful for tracing the roots of the r3b2 strains of R. solanacearum.
Askora et al. [16] provided the first evidence that the ORF14 of the R. solanacearum filamentous phage ϕRSM1 functioned as a small site-specific serine recombinase from the resolvase/invertase subfamily for integrative and excisive recombination of the phage in R. solanacearum. They used an indirect approach for the study by cloning ϕRSM1's ORF14 and attP sequences into an E. coli plasmid and observed integration of the resulting plasmid pT-orf14-attP into the attB sequence of R. solanacearum [16]. They also observed the occurrence of an intermolecular recombination between the E. coli plasmid pT-orf14-attP and a R. solanacearum plasmid containing attB in R. solanacearum cells [16]. In our study, we used a direct approach by constructing ϕRs551 mutants missing orf13 and attP either alone or in combination, and found that all three phage mutants ϕΔorf13, ϕΔattP and ϕΔattP&Δorf3 were unable to integrate into RUN302 (Fig 3E). Our study, therefore, confirms that, similar to ϕRSM phages, ϕRs551 relies on its own ORF13 (ORF14 in ϕRSM1 and ϕRSM3), a resolvase, and a core att sequence for site-specific integration into its susceptible R. solanacearum host genome. It is worth mentioning that the filamentous R. solanacearum phage ϕRS603 has not been found in an integrated state [9], and a counterpart resolvase of the ORF13 of ϕRs551 or ORF14 of the ϕRSM has not been identified in ϕRS603, further suggesting the involvement of a resolvase in the integration of R. solanacearum filamentous phages.
ϕRs551 maintained both prophage and free particle states, unlike R. solanacearum RSM and RSS phages in which episomal DNA and virions were rarely identified once a prophage state was established in the lysogenic strains [39]. The phage ϕRs551 appeared spontaneously in the supernatant of the R. solanacearum lysogenic cells and was able to replicate to a rate of 10 11 pfu/ml even after the prophage state was established in RUN302. Yamada [39] hypothesized that the prophage state and phage immunity in the RSS and RSM phages were possibly maintained by the regulatory function of the putative phage repressor encoded by the orf13 of ϕRSS and the orf15 of ϕRSM. Interestingly, the putative repressor gene identified in ϕRs551 and encoded by the ORF14 showed no homology to the putative repressor of ϕRSM3 which we named a type 1 repressor, but shared a high homology (81%) with the putative repressor of ϕRS603 which we named a type 2 repressor, suggesting that the type 2 repressor may not suppress phage replication as strongly as the type 1 repressor. Our naming of the two types of putative repressor proteins is consistent with previous findings that ORFs encoding for the putative repressors in phages ϕRSM1 and ϕRSM3 and the prophage RSM7 shared weak homology with the ones in the prophages RSM4, RSM5 and RSM6, as well as in the phage ϕRS603, although the ORFs within each of the two groups shared high amino acid similarity [9,20]. It is worth mentioning that ϕRs551 is the first isolated virion that contains an integrase (ORF13) and a putative type-2 phage repressor (ORF14), since the RSS and RSM phages contain an integrase but a type 1 repressor and the RS603 phage contained a type 2 repressor but did not encode any integrase. The significance of such an integrase and repressor combination is unknown, and the biological functions and competiveness advantages, if any, such a combination offers to the phage and its host bacterium will be the focus of further study.
Infection of a R. solanacearum host strain RUN302 by ϕRs551 resulted in less fluidal bacterial colonies with significantly reduced production of the extracellular polysaccharide, a major virulence factor [40]. The overall in vitro growth of the bacterium in liquid CPG, however, was not significantly affected. This is similar to RSM3-infected R. solanacearum cells that had similar growth curve to their uninfected wild type cells [19]. The phage infection also reduced the bacterium's twitching, swimming and swarming motilities, which are important virulence factors, especially during root invasion and early colonization [41,42]. As a result, at least partially, it is not surprising that the virulence of the ϕRs551-infected R. solanacearum was significantly reduced, although the reduced virulence is different from the loss of virulence caused by infection with the filamentous RSM-type phage ϕRSM3 [19]. Addy et al. [19] demonstrated that the virulence can be restored when the putative phage repressor encoded by the ORF15 of ϕRSM3 was deleted, suggesting negative regulation of virulence by the ϕRSM3-encoded repressor protein. Since the putative phage repressor in ϕRs551 (a type 2) is different from the one in the RSM phages (a type 1), the putative repressor in ϕRs551 may not regulate virulence as tightly as the one in ϕRSM3, resulting in a significantly reduced but not loss of virulence in ϕRs551-infected R. solanacearum cells in tomato plants. Our host specificity study revealed that ϕRs551 infected a few but not all tested r3b2 strains, as well as a closely related non-r3b2 sequevar 4 strain of R. solanacearum ( Table 1), suggesting that ϕRs551 is a more selective phage such as ϕRSM1 [15]. This combined with the reduced, not complete loss of virulence caused by ϕRs551 argues against its potential as a biocontrol agent by itself. Future research, however, is needed to study the roles, if any, played by the phage in other aspects of its host bacterium. For example, will producing and releasing ϕRs551 virions offer any fitness or competitive advantage to the R. solanacearum lysogenic strains by infecting susceptible R. solanacearum strains occupying the same environment, competing for the same ecological niche for plant infection and survival? Better understanding of the phage-bacterium-environment interactions will facilitate the development of integrated management strategies to combat R. solanacearum. Future research is also needed to determine the regulatory role played by the putative repressor of ϕRs551 in integration and excision of the phage, as well as in virulence and competiveness fitness of its bacterial hosts.
In this study, a new prophage RSM10 was identified from the deposited genome sequence of R. solanacearum RS-09-161. Although its entire sequence is homologous to a combination of ϕRS603, ϕRs551and the RSM phages, its structure in the A-S region is similar to ϕRS603 and ϕRSS phages, suggesting that the RSM10 prophage might be an evolutionary intermediate between ϕRSM and ϕRSS phages. The discovery that the RSM10 prophage also contains an ORF that is homologous to a hypothetical protein in Methylomicrobium album suggests that genetic material has been exchanged; although it is unknown how this horizontal gene transfer has occurred. It is worth mentioning that the putative repressor (ORF15L, Fig 2) annotated for the new prophage RSM10 is 100% identical at the amino acid level to the putative repressor protein (ORF15) of the prophage RSM3 (Fig 2), suggesting that different from ϕRs551, ϕRSM10 contains a type 1 repressor (Fig 2). Whether an active virion of RSM10 can be isolated from R. solanacearum strain RS-09-161, what roles played by the type 1 repressor in ϕRSM10, and what are the effects of the prophage on the evolution and host virulence in R. solanacearum remain to be determined.