High Frequency of a Novel Filamentous Phage, VCYϕ, within an Environmental Vibrio cholerae Population

Abstract

Environmental Vibrio cholerae strains isolated from a coastal brackish pond (Oyster Pond, Woods Hole, MA) carried a novel filamentous phage, VCYϕ, which can exist as a host genome integrative form (IF) and a plasmid-like replicative form (RF). Outside the cell, the phage displays a morphology typical of Inovirus, with filamentous particles ∼1.8 μm in length and 7 nm in width. Four independent RF isolates had identical genomes, except for 8 single nucleotide polymorphisms clustered in two regions. The overall genome size is 7,103 bp with 11 putative open reading frames organized into three functional modules (replication, structure and assembly, and regulation). VCYϕ shares sequence similarity with other filamentous phages (including cholera disease-associated CTX) in a highly mosaic manner, indicating evolution by horizontal gene transfer and recombination. VCYϕ integrates in the vicinity of the putative translation initiation factor Sui1 in chromosome II of V. cholerae. A screen of 531 closely related host isolates showed that ∼40% harbored phages, with 27% and 13% carrying the IF and RF, respectively. The relative frequencies of the RF and IF differed among strains isolated from the pond or lagoon of Oyster Pond, suggesting that the host habitat influences intracellular phage biology. The overall high prevalence within the host population shows that filamentous phages can be an important component of the environmental biology of V. cholerae.

INTRODUCTION

Filamentous phages of the genus Inovirus are unusual among bacterial viruses in that they do not lyse host cells when new phage particles are produced. Instead, new virions are packaged on the cell surface and extruded (24). These virions contain single-stranded DNA (ssDNA) that typically enters new hosts via a variety of pili positioned on the cell surface (26). Inside the host, inoviruses can persist as a circular, double-stranded replicative form (RF); alternatively, they can integrate into the host chromosome by a variety of mechanisms, including phage-encoded transposases (19) and host-encoded XerC/D (11, 13), which normally resolve chromosome dimers. Production of new phage ssDNA can proceed via rolling-circle replication from the RF. The genomes of inoviruses are composed of modules that encode genome replication, virion structure and assembly, and regulation (3); additionally, like many other phages, inoviruses can undergo extensive recombination, often picking up new genes in the process so that they may act as important mechanisms of gene transfer among hosts (7, 9).

Vibrio cholerae, an environmental bacterium containing strains capable of eliciting the diarrheal disease cholera, has become somewhat of a model for studying Inovirus biology and diversity. This is because an important pathogenicity factor, the cholera toxin (CT), is encoded and transferred by the filamentous phage CTXϕ (21). Infection is mediated by the recognition of a type IV pilus (toxin-coregulated pilus), and the phage genome can irreversibly integrate into the host chromosome at one of two dif sites (dif1 and dif2), which are the targets of XerC/D-mediated recombination with phage att sites (attP) and are present on V. cholerae chromosomes I and II, respectively (22). Different variants of CTXϕ are specific for either dif1 or dif2, where they can integrate as single or tandem copies (6). A number of additional filamentous phages have been described for V. cholerae, including VEJϕ (3), VGJϕ (4), KSF-1ϕ (9), VSKϕ (17), VSKKϕ, fs1ϕ (23), fs2ϕ (8), Vf33ϕ (27), and 493ϕ (16). Importantly, it has recently been shown that several filamentous phages display cooperative interactions and that a process of sequential infection, involving two satellite and three helper phages, may have been important in the evolution of V. cholerae strains associated with the seventh pandemic (11).

Here we characterize a novel filamentous phage, designated VCYϕ, from an environmental V. cholerae population. We also show that VCYϕ had a remarkably widespread distribution in the host population it originated from and that the prevalence of the RF versus the IF in host cells appears to be influenced by the host habitat and lifestyle.

MATERIALS AND METHODS

V. cholerae isolation and propagation.

V. cholerae strains were isolated from surface water of Oyster Pond, Woods Hole, MA, and its lagoon connecting the pond to the coastal ocean on 8 September 2008. The water temperature and salinity were 24.5 and 26°C and 4 and 5 ppt for the pond and the lagoon, respectively. Particle-associated and free-living bacterial populations were collected by sequential filtration of water samples onto filters with different size cutoffs following the protocol in reference 14. For the largest fraction, which is enriched in zooplankton, three replicate water samples of ∼100 liters each were filtered through a 63-μm plankton net (Wildlife Supply Company) and the filtrate was collected for strain isolation in the lab. For the remaining 3 size fractions, 3 replicate 1-liter samples, which had been prefiltered to remove the 63-μm fraction, were collected and transported to the lab for further processing.

In the laboratory, all materials retained on 63-μm filters were homogenized using a tissue grinder (VWR Scientific) and vortexed for 20 min at low speed. The replicate 1-liter water samples from which the >63-μm fraction had been removed were sequentially filtered through 5-, 1-, and 0.2-μm-pore-size filters where the 63- to 5-μm and 5- to 1-μm size fractions were collected using gravity filtration to avoid the breakdown of fragile particles. For these, filtration was repeated with sterile seawater to further remove cells unattached to particles. Subsequently, all filters were placed into 50-ml conical tubes containing 45 ml sterile seawater and vortexed for 20 min at low speed to break up particles and resuspend bacterial cells. Supernatants were used for isolation of V. cholerae by concentrating serial dilutions onto 0.2-μm Supor-200 filters (Pall) using gentle vacuum pressure. These filters were then placed onto agar plates containing Vibrio selective thiosulfate citrate bile salts sucrose medium (BD Difco) with 2% NaCl (marine TCBS). Single colonies were picked and restreaked three times by alternating tryptic soy broth (BD Bacto) with 2% NaCl and marine TCBS medium to obtain pure strains. For all subsequent analyses, the stock cultures were used to avoid unequal treatment of strains. Identification of V. cholerae was done by partial sequencing of the mdh gene as described in reference 1. For routine propagation, strains were grown overnight in Luria-Bertani (LB) broth (Difco) at 25°C in a shaking bath (180 rpm) overnight. Phage was originally detected as a plasmid-like band in genomic DNA preparations analyzed on agarose gels.

DNA isolation and sequencing.

DNA was extracted from V. cholerae for sequencing of the plasmid-like RF of VCYϕ and to determine the insertion site of the integrative form (IF) in the host chromosome. To obtain RF DNA, plasmid-like genomes were isolated from 2 ml samples of overnight cultures of V. cholerae strains 10E09PW02, 10F04PW02, 5G03LW63, and 11H04LW5 using the Qiaprep Spin Miniprep kit (Qiagen Inc.). Subsequently, DNA was electrophoretically separated on 0.8% agarose gels and the bands corresponding to the RF were cut out and purified using gel extraction kits (Qiagen Inc.). RF DNA from strain 10E09PW02 was tagged by barcode A6-B15 (Table 1), while DNA from the remaining three RFs was combined and tagged with barcode A4-B14 for Illumina sequencing.

Table 1.

Primers used in this study

Primer	Sequence	Reference
Illumina adapter A4 up	5′-/5AmMC6/ACACTCTTTCCCTACACGACGCTCTTCCGATCTGCAGG-3′	This study
Illumina adapter A4 down	5′-CCTGCAGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAC/3AmM/-3′	This study
Illumina adapter A6 up	5′-/5AmMC6/ACACTCTTTCCCTACACGACGCTCTTCCGATCTAATTC-3′	This study
Illumina adapter A6 down	5′-GAATTAGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAC/3AmM/-3′	This study
Illumina adapter B14 up	5′-TACTGAGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGC/3AmM/-3′	This study
Illumina adapter B14 down	5′-/5AmMC6/CTCGGCATTCCTGCTGAACCGCTCTTCCGATCTCAGTA-3′	This study
Illumina adapter B15 up	5′-AGCAGAGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGC/3AmM/-3′	This study
Illumina adapter B15 down	5′-/5AmMC6/CTCGGCATTCCTGCTGAACCGCTCTTCCGATCTCTGCT-3′	This study
Illumina _amp_1	5′-AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT-3′	25
Illumina _amp_2	5′-AAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCT-3′	25
VCYϕ_int_F	5′-TTAACATTGTCAAATGATAAATATG-3′	This study
VCYϕ_int_R	5′-ATAATCAACTGATAATGTTGCAAAC-3′	This study
PCRwalking_biotin	5′-biotin-CAACACAGCCCATTATTTTAGCCCC-3′	This study
PCRwalking_biotin-asp	5′-biotin-CATTTCACCATTTTATATTGCGCGT-3′	This study
PCRwalking_biotin_nest	5′-CATTTCACCATTTTATATTGCGCGT-3′	This study
PCRwalking_biotin_nest-asp	5′-TCTGAACTGTTAGACGCCTACAAAA-3′	This study
PCRwalking_anchor-C12	5′-CCACGCGTCGACTAGTAATTCCCCCCCCCCCCDN-3′	This study
PCRwalking_anchor	5′-CCACGCGTCGACTAGTAATT-3′	This study
VCYϕ_Seq_1	5′-TCGATTCATTGTTAAAACTCCCAAAATCG-3′	This study

An Illumina sequencing protocol (25) was modified to allow small plasmid library preparation as follows. DNA libraries were prepared by shearing about 1 μg RF DNA in a volume of 50 μl into fragments with an average length of ∼400 bp. This was done using 14 cycles of alternating 30-s ultrasonic bursts and 30-s pauses in a 4°C water bath in a Bio-Ruptor UCD-200 (Biogenode). The fragment ends were then repaired and phosphorylated using the End-Repair kit (New England BioLabs). The products were subject to a ligation reaction with a 10-fold molecular excess of Illumina adapters (Table 1) using the Quick Ligation kit (New England BioLabs). The ligation product was separated on 1.5% agarose gels, and fragments of 300 to 500 bp were purified with 10 μl elution buffer (EB; Qiagen) using the Qiagen MinElute Reaction Cleanup kit (Qiagen Inc.). The fragments were nick translated with Bst polymerase (New England BioLabs) in a 30-μl final volume. Eight replicate 2-μl reaction products were used without further purification in PCR amplifications using Phusion Hot Start High-Fidelity DNA polymerase (New England BioLabs), and reaction progress was monitored on a Bio-Rad Opticon real-time PCR instrument. The reactions were stopped in the late logarithmic amplification phase, and the DNA from the replicate reaction mixtures was pooled. To generate the ready-to-sequence DNA, libraries were subjected to an additional gel purification step to remove adapter dimers and residual primers. The quality and size distribution of the DNA libraries were checked by Agilent Bioanalyzer DNA-1000 assays (Agilent Technologies, Inc.). The 2 libraries were pooled with 34 other libraries that had different bar codes for deconvolution postsequencing. The samples were loaded onto a cluster of Illumina GAIIx sequencers, and resultant data were analyzed using the Illumina pipeline 1.4.0 to generate fastq files. Sequences were reconstructed and annotated using NextGen 1.9 (Softgenetics Inc.) and DNAmaster software (http://cobamide2.bio.pitt.edu), respectively.

To determine the host chromosomal region of phage insertion in strains 4A03LW1 and 4B03LW1, a walking PCR protocol (18, 20) was used taking advantage of the fact that the attP site is split during phage insertion into the host chromosome. Biotinylated primers, PCRwalking_biotin and PCRwalking_biotin-asp (Table 1), facing outward from the predicted attP site were designed and used to obtain single-stranded PCR products. In a typical reaction mixture, 20 ng host DNA containing integrated VCYϕ was mixed with 0.5 μmol biotinylated primer and 0.5 U Platinum Taq Hi-Fidelity (Invitrogen). Amplification used a three-step cycling program (94°C for 30 s, 45°C for 30 s, and 68°C for 5 min) for 35 cycles. The extension products were captured on streptavidin beads (Promega), purified, and stored in 1× terminal deoxynucleotidyl transferase buffer. A poly(G) tail was added to the purified extended products by incubation with 4 mM dGTP and 4 U terminal deoxynucleotidyl transferase enzyme (Promega) at 37°C in a shaking bath (200 rpm) for 2 h. The poly(G)-tailed products were made double stranded by using the PCRwalking_anchor-C12 and PCRwalking_nest-asp primers (Table 1). The PCR products were separated on a 1% agarose gel, and fragments of 2 to 4 kb were purified using the QIAquick gel extraction kit (Qiagen). The purified DNA fragments were reamplified with primers PCRwalking_nest and PCRwalking_anchor (same as PCRwalking_anchor-C12 but lacks the run of 12 C residues) (Table 1). The DNA was purified again by agarose gel and the QIAquick gel extraction kit and sequenced using the Sanger method with either primer VCYϕ_seq_1 or VCYϕ_seq_2 (Table 1).

To test whether the phage DNA was in single-stranded form, DNA was isolated from phage particles as described by Faruque et al. (9) and digested with DNase I.

PCR-based phage identification and screening for the RF or IF in host cells.

Because different strains were used for sequencing and electron microscopy (EM) of phage and to ensure that the RF and IF are similar phages, we devised specific PCR primers targeting a gene (open reading frame 9 [ORF9]) currently unique to VCYϕ (Table 1).

To identify host isolates containing the RF and/or IF, a PCR protocol was devised that can differentiate either form. This was achieved by designing one set of primers flanking the attB site in the bacterial chromosome (primers VCYϕ_int_F and VCYϕ_int_R; Table 1), which is split during integration, so that these primers yield only a product for strains not carrying the IF of VCYϕ. Similarly, a second set of primers flanking the attP site of the phage (primers VCYϕ_Seq_1 and VCYϕ_Seq_2; Table 1) was used to confirm the presence of the RF of VCYϕ. For identification of the IF, we used primers VCYϕ_int_F and VCYϕ_Seq_1 or VCYϕ_Seq_2, which can produce a 190-bp or 150-bp PCR product if VCYϕ is integrated into the attB site.

In a typical reaction mixture, 20 ng genomic DNA or 2 μl 1:10-diluted cultured strains were used as the template and mixed with 0.4 μM RF-specific or IF-specific primers and a polymerase mixture from the Qiagen HotStarTaq Master Mix Kit using a three-step cycling program consisting of initial denaturation at 95°C for 15 min; 30 cycles of a three-step procedure including denaturation at 94°C for 30s, annealing at 52°C for 30s, and extension at 72°C for 30s; and a final extension at 72°C for 5 min. The PCR products were separated on a 1.8% agarose gel prepared with 0.5× Tris-borate-EDTA buffer.

EM.

To prepare phages for EM, V. cholerae strain 7D07PW5, which carries the RF of VCYϕ, was grown overnight in 100 ml LB medium at room temperature in a shaking water bath (180 rpm). The supernatant containing phages was collected by centrifuging the culture at 8,000 × g for 15 min and subsequently filtering it through a 0.22-μm-pore-size filter. A 100-μl aliquot of the filtered supernatant was spread on an LB agarose plate for sterility assurance. To precipitate the phage particles, NaCl and polyethylene glycol 6000 were added to the filtrate to final concentrations of 2.5 and 5%, respectively. The mixture was incubated on ice for 30 min and then centrifuged at 13,000 × g for 30 min. The phage-containing pellet was collected and resuspended in 500 μl phosphate-buffered saline.

For EM, purified phage particles were negatively stained with 4% (wt/vol) uranyl acetate and mounted on freshly prepared Formvar grids. Phage samples were photographed under a FEI Technai Spirit transmission electron microscope. The average length and width of the phages were determined from six individual particles.

Nucleotide sequence accession number.

The genome sequence of VCYϕ from strain 10E09PW02 has been deposited in GenBank under accession number JN848801.

RESULTS AND DISCUSSION

Characterization of the RF of filamentous phage VCYϕ.

Among a collection of 531 environmental V. cholerae isolates from Oyster Pond, 77 contained putative episomal elements when screened by gel electrophoresis. Seven strains contained episomal elements of various sizes; however, 70 appeared similar in size. Restriction endonuclease analysis for a subset of 10 of these elements using BamHI, EcoRI, and PstI revealed identical patterns, suggesting a closely related, double-stranded plasmid-like element with a size of approximately 7 kbp (data not shown). Subsequent genome sequencing of 4 of these 7-kbp plasmids suggested them to be the RF of a new filamentous phage, which we call VCYϕ (Fig. 1).

Fig 1.

Genome organization of phage VCYϕ. Linear ORF maps of phages VGJϕ, KSF-1ϕ, and VCYϕ were aligned based on their modular structures. ORFs or genes are represented by arrows oriented in the direction of transcription. Black, white, and light gray arrows represent replication, structure-and-assembly, and regulation modules, respectively. Dark gray arrows represent unknown ORFs. The attP sequences of VGJϕ and two SNP regions (SNP-A and SNP-B) are also indicated.

The whole genome of phage VCYϕ consists of 7,103 nucleotides (Fig. 1) with a G+C content of 41.8 mol%. Among the 4 sequenced genomes, only 8 single nucleotide polymorphisms (SNPs), clustered in two regions, were evident: 3 SNPs in a 7-bp stretch (SNP-A in Fig. 1) and 5 SNPs in a 14-bp stretch (SNP-B in Fig. 1). Overall, the phage contains 11 ORFs (ORF1 to ORF11) predicted by a BLAST search. Of these ORFs, 9 are homologous to protein-coding genes previously reported from other filamentous phages, including KSF-1ϕ (9) and VGJϕ (4). Based on similarity in sequence and organization to these other phages, the ORFs of VCYϕ can be classified into functional modules for replication, structure and assembly, and regulation (Fig. 1).

The putative replication module, composed of ORF1 to ORF3 (Fig. 1), maps to the same position as rstA and rstB in CTXϕ (8) and gII and gV in phage M13 (2). ORF2 and ORF3 share amino acid sequence similarity with the protein of potential phage replication genes and genes of potential ssDNA-binding proteins (26). We therefore suggest that ORF2 and ORF3 play similar roles in VCYϕ. A hypothetical gene, ORF1 is associated with the replication module based on its map position and the overlap of its stop codon with ORF2; however, its function remains unknown.

The putative structural and assembly module consists of ORF4 to ORF8 (Fig. 1), each sharing similarity in size, genome position, and sequence with the corresponding capsid proteins of other filamentous phages (3, 4, 8, 9). For instance, ORF6 exhibits similar size and genome position to gIII of CTXϕ, which encodes a minor capsid protein, pIII, that recognizes and interacts with receptors and coreceptors. The protein encoded by ORF8 is similar to the pI protein of phage ϕLF of Xanthomonas campestris (5) and the Zot protein of CTXϕ, which are both required for viral particle packaging and secretion.

ORF10 and ORF11 likely encode regulatory proteins constituting the third module. Both ORFs are oriented in opposite direction to the rest of the ORFs and exhibit homology to ORF136 and ORF154 of VGJϕ (4) that encode a potential regulatory and repressor protein, respectively. Finally, ORF9 is a conserved hypothetical protein whose function has not been established. Its location between the structure-and-assembly and regulation modules is the same as ctxA and ctxB of CTXϕ. However, ORF9 does not share homology with these genes, which code for CT, an important pathogenicity determinant in the diarrheal disease cholera. Based on these comparisons, it seems likely that ORF9 is not associated with any of the three modules and may provide additional but currently unknown function to the phage.

As in other sequenced filamentous phages from Vibrio strains (4, 9, 10), the three modules of phage VCYϕ appear to be evolutionary mosaics assembled by horizontal gene transfer. At the nucleic acid level, the structure-and-assembly module of phage VCYϕ and filamentous phage KSF-1ϕ from V. cholerae share 76% identity (9) while the regulatory module of phage VCYϕ displays only low similarity to that of KSF-1ϕ; instead, it is 80% similar to the corresponding module of phage VGJϕ (4). The hybrid genome of phage VCYϕ thus confirms that horizontal gene transfer, possibly by coinfection, is a significant driving force in the evolution of filamentous vibriophages.

Characterization of the viral particle.

A filamentous phage structure was detected by EM in precipitates obtained from filtered supernatant of strain 7D07PW5, which had been shown to contain the 7-kbp plasmid-like structure in the gel assay. These phage-like particles were 1.762 ± 0.016 μm in length and 7 nm in width (n = 6) (Fig. 2). The size of VCYϕ is similar to those typically found in the genus Inovirus (28) (0.8 to 2 μm in length and 6 to 7 nm in width), including fs2ϕ (15) and KSF-1ϕ (9).

Fig 2.

Electron micrograph of phage VCYϕ particles. Phage particles were isolated from the culture supernatant of strain 7D07PW5. Bars, 100 nm.

Comparison of RF and IF by PCR screening.

To gain further confidence that the RF and IF of the phage detected in host cells represent the same phage, we used a specific PCR assay targeting ORF9 (Table 1), which is currently unique to VCYϕ. This gave positive results for all 10 strains assayed, including those used for sequencing and EM. Together with the PCR assays differentiating IF and RF, this suggests that the phages present in the Oyster Pond isolates were of a highly similar nature.

Characterization of the integration site of VCYϕ.

Because several filamentous phages have been shown to integrate into the V. cholerae chromosome, we investigated this ability in VCYϕ. We first identified a putative 28-nucleotide-long attP site by comparison with VGJϕ (4) since this phage has a regulation module distinct from that of VCYϕ by only 20% nucleotide differences (Fig. 3). This comparison also revealed potential binding sites for XerC and XerD, which mediate phage-host recombination (22). The XerC site of phage VCYϕ differs from that of phage VGJϕ in four nucleotides, whereas the XerD sites are identical. Similar to those of phage VGJϕ, the XerC and XerD sites of phage VCYϕ are also separated by 7 nucleotides. The attP sequences of some filamentous phages from Vibrio are homologous, suggesting that the attP structure is important for recognition and recombination by XerC and XerD.

Fig 3.

attP site of VCYϕ and attB site of integration of VCYϕ into chromosome II of strain 4A01LW1. (A) Sequence alignment of the attP regions of VCYϕ and VGJϕ (accession no. AY242528) and of the attB regions of strain 4A01LW1 and dif1 of V. cholerae N16961. (B) Schematic representation of the integration region of chromosome II of strain 4A01LW1. The attB sequences region is also indicated.

To characterize the phage integration site of the host chromosome (attB), we first developed a PCR screen to distinguish V. cholerae strains containing either the IF or RF of phage VCYϕ. Strain 4A01LW1, for which this analysis detected an integrated phage, was chosen for further characterization of the chromosomal location of the attB site using walking PCR. Sequence analysis of PCR products and comparison with the genome sequence of V. cholerae O1 N16961 (12) identified a putative attB site 28 bp long and identical to the site of dif1 of V. cholerae O1 N16961 except for a single nucleotide position (Fig. 3A). However, unlike the dif1 site of V. cholerae O1 N16961, which is located on chromosome I, the 1,054-bp region flanking the attB site in strain 4A01LW1 shared sequence similarity (84%) with chromosome II. In this strain, the attB site is located between a putative transposase and sui1, which encodes a translation initiation factor (Fig. 3B). This suggests that phage VCYϕ, like other filamentous phages in Vibrio strains, uses the XerC and XerD recombination system to integrate into chromosomal dif-like sequences.

Distribution of VCYϕ across the environmental V. cholerae population.

Using data from gel analysis and the PCR-based screening for either the RF or the IF, we further investigated the frequencies of the two forms of phage VCYϕ across a large collection of V. cholerae isolates from Oyster Pond. This showed that 220 (41.4%) of a total of 531 isolates contained either the RF or IF of phage VCYϕ (Table 2), and 70 (13.2%) of 531 strains carried only the 7-kbp RF of phage VCYϕ, as suggested by gel electrophoresis (Table 2). Because the IF was detectable by IF-specific PCR assay in none of these strains, the RF of phage VCYϕ appears to be able to replicate without integration into the chromosome. Overall, this suggests a remarkably high prevalence of this phage within this environmental V. cholerae population.

Table 2.

Frequencies of IF and RF of phage VCYϕ in a collection of 531 V. cholerae isolates from coastal Oyster Pond, MA, and its lagoon

Source	Total no. of isolates	No. (%) of isolates with:		Total no. (%) of strains containing phage VCYϕ
		RFa	IFb
Pond	360	59 (16.4)	79 (21.9)	138 (38.3)
Lagoon	171	11 (6.4)	71 (41.5)	82 (48.0)
Both	531	70 (13.2)	150 (28.2)	220 (41.4)

^aThe RF was detected as a 7-kbp, plasmid-like band by agarose gel electrophoresis and by an additional RF-specific PCR assays followed the IF-specific PCR assays.

^bThe IF was identified by IF-specific PCR assays.

Although it is impossible to know exactly how initial isolation has affected the transition between the RF and the IF or loss of the phage, we suggest that the values provided represent the lower boundary of the frequency of the phage in this environmental V. cholerae population. We found that after regrowth of strains from liquid stock cultures, 14% of the strains had lost the RF but in only one strain had the RF transitioned to the IF. Moreover, only a single loss of the IF was observed when 2 strains (4A01 and 4B03) were streaked from liquid medium and a total of 39 single colonies were assayed. This suggests that the RF and IF are moderately stable when strains are propagated but also indicates that an even higher portion of strains in the environmental population may have been harboring phages.

Another 150 (28.2%) of the 531 host isolates carried the IF of phage VCYϕ detectable by IF-specific PCR screen, suggesting a high prevalence of the integrated phage in the V. cholerae population. Although none of these strains showed a visible 7-kbp band by agarose gel electrophoresis, 113 also gave positive results with the RF-specific PCR screening (Table 2). Because lysogen repression is never absolute, it is possible that a small subpopulation in each culture tube had transitioned from the IF to the RF. Alternatively, a subpopulation in each culture tube had transitioned from the IF to the RF. These host strains were therefore scored as containing the IF only, but the presence of small amounts of the RF underscores that the exact proportions of strains containing the RF and IF may have shifted during culturing. We therefore only stress trends in populations from different habitats based on the assumption that these should be unaffected by transitions between the two forms during regrowth of strains, which was highly standardized (three streaks postisolation with subsequent analyses carried out with freezer stock cultures).

The overall frequencies of the phage (RF and IF) were similar among host isolates from the lagoon and the pond (Table 2); however, they displayed different trends in the presence of the IF and RF. While the frequencies of the IF (22%) and RF (16%) were roughly equal in the pond, in the lagoon, the vast majority was in the IF (42%) rather than the RF (6%) (Table 2). As detailed above, we deem it unlikely that such a difference might arise postisolation of host strains, so that environmental factors may play a role in the transition between the RF and IF. The population in the lagoon may thus have been producing fewer phage particles than its equivalent in the pond; however, we emphasize that this observation will have to be verified by culture-independent methods in the future.

In summary, we have described a novel filamentous phage infecting V. cholerae, adding to the already considerable number of such phages in this species. We show that this phage had a surprisingly high prevalence in environmental host populations when sampled in late summer and that the transition between the IF and RF may be influenced by environmental factors. Overall, filamentous phages appear to be an important factor in the environmental biology of V. cholerae and can affect a large fraction of the cells within a population.