Targeted deletion of Pf prophages from diverse Pseudomonas aeruginosa isolates has differential impacts on quorum sensing and virulence traits

ABSTRACT Pseudomonas aeruginosa is an opportunistic bacterial pathogen that commonly causes medical hardware, wound, and respiratory infections. Temperate filamentous Pf phages that infect P. aeruginosa impact numerous virulence phenotypes. Most work on Pf phages has focused on Pf4 and its host P. aeruginosa PAO1. Expanding from Pf4 and PAO1, this study explores diverse Pf phages infecting P. aeruginosa clinical isolates. We describe a simple technique targeting the Pf lysogeny maintenance gene, pflM (PA0718), that enables the effective elimination of Pf prophages from diverse P. aeruginosa hosts. The pflM gene shows diversity among different Pf phage isolates; however, all examined pflM alleles encode the DUF5447 domain. We demonstrate that pflM deletion results in prophage excision but not replication, leading to total prophage loss, indicating a role for lysis/lysogeny decisions for the DUF5447 domain. This study also assesses the effects different Pf phages have on host quorum sensing, biofilm formation, pigment production, and virulence against the bacterivorous nematode Caenorhabditis elegans. We find that Pf phages have strain-specific impacts on quorum sensing and biofilm formation, but nearly all suppress pigment production and increase C. elegans avoidance behavior. Collectively, this research not only introduces a valuable tool for Pf prophage elimination from diverse P. aeruginosa isolates but also advances our understanding of the complex relationship between P. aeruginosa and filamentous Pf phages. IMPORTANCE Pseudomonas aeruginosa is an opportunistic bacterial pathogen that is frequently infected by filamentous Pf phages (viruses) that integrate into its chromosome, affecting behavior. Although prior work has focused on Pf4 and PAO1, this study investigates diverse Pf in clinical isolates. A simple method targeting the deletion of the Pf lysogeny maintenance gene pflM (PA0718) effectively eliminates Pf prophages from clinical isolates. The research evaluates the impact Pf prophages have on bacterial quorum sensing, biofilm formation, and virulence phenotypes. This work introduces a valuable tool to eliminate Pf prophages from clinical isolates and advances our understanding of P. aeruginosa and filamentous Pf phage interactions.

division.When induced, the Pf prophage is excised from the chromosome, forming a circular double-stranded episome called the replicative form (5). Pf replicative form copy numbers increase in the cytoplasm where they serve as templates for viral transcription.The replicative form is also a template for rolling circle replication, which generates circular single-strand DNA genomes that are packaged into filamentous virions as they are extruded from the cell (3,6).
Filamentous Pf virions are associated with enhanced P. aeruginosa virulence potential by promoting biofilm formation (7) and inhibiting phagocytic uptake by macrophages (8,9).Pf virions also carry a high negative charge density allowing them to sequester cationic antimicrobials such as aminoglycoside antibiotics and antimicrobial peptides (7,10,11).Additionally, Pf phages enhance the virulence potential of P. aeruginosa by modulating the secretion of the quorum-regulated virulence factor pyocyanin (12,13).These properties may explain why the presence of Pf virions at sites of infection is associated with more chronic lung infections and antibiotic resistance in cystic fibrosis patients (8) and why P. aeruginosa strains cured of their Pf infection are less virulent in murine models of pneumonia (14) and wound infection (15).
Most studies to date have focused on interactions between Pf isolate Pf4 and its host P. aeruginosa PAO1 (3,9,11,14).Despite the clear link between Pf4 and the virulence of P. aeruginosa PAO1, the effects of diverse Pf that infect P. aeruginosa clinical isolates have on virulence phenotypes remain unclear.This is in part due to the significant challenge of "curing" clinical isolates of their Pf prophage infections.Prior efforts to delete Pf4 from PAO1 relied on the integration of a selectable marker into the integration site used by Pf4 (14), which precludes complementation studies that re-introduce the Pf4 prophage to the host chromosome.In prior work, we were able to generate a clean Pf4 deletion strain by first deleting the pfiTA toxin-antitoxin module encoded by Pf4, followed by deletion of the rest of the prophage (16).
Here, we find that the Pf4 gene PA0718 maintains Pf4 in a lysogenic state; we therefore refer to PA0718 as the Pf lysogeny maintenance gene pflM.Deletion of PA0718 or homologous alleles from Pf prophages in clinical P. aeruginosa isolates, LESB58, CPA0053, and CPA0087, and the multidrug-resistant strain DDRC3 resulted in the complete loss of Pf prophages from each strain.Furthermore, we observed that some substrains of PAO1 are lysogenized by two Pf phages, Pf4 and Pf6, and we success fully cured PAO1 of both Pf4 and Pf6 prophages.We compare phenotypic differences between wild-type and ∆Pf prophage mutants by assessing Las, Rhl, and PQS quorumsensing activity, biofilm formation, and pyocyanin production.We also examine how Pf prophages impact virulence phenotypes in a C. elegans avoidance model.Overall, we present a new methodology for efficiently curing P. aeruginosa strains of their resident Pf prophages and leverage this tool to gain insight into the diverse impacts Pf phages have on their bacterial hosts.

PA0718 (PflM) maintains Pf4 lysogeny in P. aeruginosa PAO1
While performing single-gene deletions from the core Pf4 genome (pf4r-intF4) in P. aeruginosa PAO1 (Fig. 1A), we noted that deleting either the Pf4r repressor or the PA0718 gene results in the complete excision of the Pf4 prophage from the P. aeruginosa chromosome (Fig. 1B, upper bands).Prior work demonstrates that deletion of the Pf4r repressor induces Pf4 prophage excision and virion replication (17), but how PA0718 is involved in Pf4 excision is not known.
After excision, Pf4 replicates as a circular episome called the replicative form (5). We used qPCR to measure circular Pf4 replicative form copy numbers in wild-type and ∆PA0718 cells.In wild-type cells, approximately 4,400 replicative form copies were detected for every 10,000 cells; however, the Pf4 replicative form was not detected in ∆PA0718 cells (Fig. 1C), indicating that Pf4 genome replication is not initiated, and the replicative form is lost as cells divide.These results indicate that PA0718 maintains the Pf4 prophage in a lysogenic state and that deleting PA0718 induces Pf4 prophage excision, but not replication, curing PAO1 of its Pf4 infection.Herein, we refer to PA0718 as the Pf lysogeny maintenance gene pflM.
The observation that 4,400 Pf4 replicative form copies are detected for every 10,000 wild-type cells (Fig. 1C) indicates that Pf4 is actively replicating in a subpopulation of cells.We used a multiplex PCR excision assay to measure Pf4 prophage excision and integration in P. aeruginosa populations.In PAO1 populations with no expression vector or those carrying an empty expression vector, both Pf4 prophage integration and excision are observed (Fig. 1D, two bands are present); however, infectious virions were not detected in supernatants by plaque assay, suggesting that Pf4 is replicating at low levels during planktonic growth in LB broth, consistent with prior results (18).The Pf4 excisionase XisF4 regulates Pf4 prophage excision (17), and expressing XisF4 in trans induces complete Pf4 prophage excision (Fig. 1D) and robust virion replication (Fig. 1E).In contrast, expressing PflM in trans maintains the entire population in a lysogenic state (Fig. 1D), and virion replication is not detected (Fig. 1E).When PflM and XisF4 are expressed together, both Pf4 integration and excision products are observed (Fig. 1D), and infectious virions are produced at titers comparable with cells where XisF4 was expressed by itself (Fig. 1E).These results indicate that expressing PflM is not sufficient to inhibit XisF4-mediated Pf4 prophage excision and replication but that PflM can maintain some cells in a lysogenic state during active viral replication.

The targeted deletion of pflM cures diverse P. aeruginosa isolates of their Pf prophages
We hypothesized that deleting pflM would provide a convenient way to cure P. aerugi nosa clinical isolates of their Pf prophages.To test this hypothesis, we deleted pflM from the Pf prophages in cystic fibrosis isolate LESB58, two cystic fibrosis isolates from the Stanford Cystic Fibrosis Clinic (CPA0053 and CPA0087), and the multidrug-resistant urine isolate DDRC3 (Table 1).
Pf prophage loss in PAO1 was screened by multiplex PCR (Fig. 2A/C) and confirmed in all strains by long-read whole-genome sequencing.Targeting pflM successfully cured all the above clinical P. aeruginosa isolates of their Pf prophages (Fig. 2B, D through G).Of the Pf prophages we deleted, four were integrated into tRNA genes (three in tRNA-Gly and one in tRNA-Met), and two were integrated into direct repeats (Table 1).Furthermore, Pf prophages fall into two main lineages (I and II, Table 1) (4), and we were successful in deleting representatives from each lineage with an efficiency rate of approximately 17%.These observations indicate neither integration site nor lineage has an influence on pflM-mediated Pf prophage deletion.
To determine if prophage loss impacts bacterial growth, we performed growth curves over 18 hours and compared parental and prophage mutant strains (Fig. S1).In strains PAO1, CPA0053, and DDRC3, deletion of the Pf prophage has no significant impact on bacterial growth (Fig. S1A through C).For strains CPA0087 and LESB58, prophage presence significantly reduces or increases growth rates, respectively (P < 0.0001) (Fig. S1D through E).The decreased growth of strain CPA0087 ∆Pf may be related to the large increase in pyocyanin production, which could negatively impact bacterial growth.The increased growth rate of LESB58 ∆Pf may be explained by relieving the metabolic burden of Pf phage replication.Alternatively, differences in Pf prophage insertion loci in different P. aeruginosa strains could affect bacterial growth dynamics.
Many P. aeruginosa strains are infected by one or more Pf prophages (3).For example, some P. aeruginosa PAO1 sub-isolates are infected by Pf4 and Pf6 (19).Deleting pflM from Pf4 results in the loss of the Pf4 prophage, as does deleting pflM from Pf6 (Fig. 2B and  D).Furthermore, we were able to delete Pf6 from ∆Pf4, producing a PAO1 ∆Pf4/Pf6 double mutant (Fig. 2E).This observation indicates that pflM from one Pf prophage is specific to that prophage and does not compensate for the loss of pflM from another Pf prophage residing in the same host.PflM specificity may be explained by the diversity in the operon encoding pflM or in the pflM allele itself (Fig. 3A).All pflM sequences examined contain a predicted DUF5447 domain (pfam17525), which, as pointed out in prior work, appears to originate from a disrupted gene encoding the Mnt (PHA01513) domain (20).The Mnt protein encoded by Salmonella phage P22 governs lysis-lysogeny decisions by binding phage operator sequences (21,22), suggesting that PflM may regulate Pf lysis-lysogeny decisions by a similar mechanism.In the Pf DDRC3 pflM allele, the Mnt domain is present and is fused to the DUF5447 domain, whereas in Pf4, Pf LESB58, Pf CPA0087, and Pf6, pflM is truncated by a 5' insertion of a gene of unknown function (PA0717 in Pf4) (Fig. 3A).Structural prediction and sequence alignments indicate that despite the different isoforms PflM present within different Pf prophages, residues that are predicted to form an anti-parallel β-sheet characteristic of the DUF5447 domain are conserved (Fig. 3B and C).

Pf phages differentially modulate host quorum sensing
Pf4 is known to suppress PQS and Rhl quorum sensing in P. aeruginosa PAO1 (12,13,23).We hypothesized that Pf phages would likewise modulate quorum sensing in the Pf deletion strains we constructed here.To test this, we used fluorescent transcriptional reporters (12) to measure Las (rsaL), Rhl (rhlA), and PQS (pqsA) transcriptional activity in parental strains and Pf mutants.
Over an 18-hour growth period, we find that differences in Las, Rhl, and PQS quorumsensing activity vary by Pf phage and P. aeruginosa host (Fig. S8).The general trends for each quorum-sensing reporter hold over 18 hours; however, for clarity, we show only the 18-hour time point in Fig. 4.After 18 hours, PAO1 ∆Pf4 Las and PQS signaling are significantly upregulated (P < 0.01 and P < 0.0001, respectively), whereas Rhl transcription is downregulated (Fig. 4A).Pf6 differentially affects PAO1 quorum sensing-Las, Rhl, and PQS signaling are all upregulated in PAO1 ∆Pf6 compared with the parental strain (Fig. 4A).Deleting both Pf4 and Pf6 had no significant impact on Las or Rhl signaling, but PQS signaling was significantly upregulated in PAO1 ∆Pf4/Pf6 compared with the parental strain (Fig. 4A), which is consistent with prior work (12,24).
After 18 hours, PQS transcriptional activity is also significantly (P < 0.001) upregula ted in LESB58 ∆Pf , as are Las and Rhl (Fig. 4B).In strain CPA0053 ∆Pf , Las signaling was significantly upregulated (P < 0.001), whereas Rhl and PQS transcription is reduced when the Pf prophage is deleted (Fig. 4C).PQS and Rhl are upregulated in CPA0087 ∆Pf , whereas Las and signaling are not significantly affected (Fig. 4D).Finally, in DDRC3 ∆Pf , Las and PQS signaling are significantly downregulated (P < 0.001), whereas Rhl signaling is significantly upregulated (P < 0.001) (Fig. 4E).Taken together, these data indicate that Pf phages have diverse and complex relationships with host quorum-sensing pathways that vary significantly by strain.

Pf phages have contrasting impacts on P. aeruginosa biofilm formation
Pf4 is known to promote P. aeruginosa PAO1 biofilm assembly and function (3,5,7,11,14,25,26).To test if other Pf isolates similarly affect biofilm formation, we used the crystal violet biofilm assay (27) to measure the biofilm formation of lysogenized P. aeruginosa isolates compared with the Pf prophage deletion mutants.In PAO1, deletion of either Pf4 or Pf6 significantly (P < 0.001) reduced biofilm formation by 1.79-fold and 2.33-fold, respectively, whereas deletion of both Pf4 and Pf6 reduced biofilm formation by 7.14-fold (Fig. 5A).This result indicates both Pf4 and Pf6 contribute to PAO1 biofilm formation, which is consistent with prior observations (5,7,11,14,25,26).The clinical isolates in general did not form as robust biofilms as the PAO1 laboratory strain under the in vitro conditions tested.Even so, deleting the Pf prophage from strains CPA0053 and DDRC3 modestly but significantly (P < 0.05) reduced biofilm formation (Fig. 5C and  E).In contrast, biofilm formation was significantly (P < 0.01) increased in strains LESB58 ∆Pf and CPA0087 ∆Pf compared with the parental strains (Fig. 5B and D).The variation in biofilm formation phenotypes is perhaps not surprising, given the variation inherent in the strains examined here, ranging from lab-adapted strains to CF-lung-adapted P. aeruginosa isolates.Although quorum-sensing regulation varies between Pf lysogens and their corresponding Pf prophage mutants, there is no clear trend of modified QS expression and biofilm formation (Fig. 5).

Pf phages suppress P. aeruginosa pyocyanin production in most clinical isolates
Pyocyanin is a redox-active quorum-regulated virulence factor (28). Deleting the Pf4 prophage from PAO1 enhances pyocyanin production (12).We observed increased pyocyanin production in all ∆Pf strains tested except CPA0053, which did not produce much pyocyanin under any condition tested (Fig. 6A through E).We confirmed that neither the pqsABCD operon nor pqsR encoded by CPA0053 is mutated, nor is lasI, lasR, rhlI, or rhlR, indicating that other factors/signaling pathways contribute to reduced pyocyanin production by this strain.These results suggest that some Pf prophages encode gene(s) that inhibit host pyocyanin production.

Pf phages induce avoidance behavior in bacterivorous nematodes
In the environment, bacterivores impose high selective pressures on bacteria (29,30).Pf4 modulation of quorum-regulated virulence factors increases P. aeruginosa fitness against the bacterivorous nematode C. elegans (12).We hypothesized that Pf prophages in P. aeruginosa clinical isolates would similarly protect P. aeruginosa from predation by C. elegans.To test this, we employed C. elegans avoidance assays (31-34) as a metric of bacterial fitness when confronted with nematode predation (Fig. 7A).C. elegans avoided all Pf lysogens, preferring to associate with the ∆Pf strains in every case (Fig. 7B).Note that nematode survival was over 95% over the course of the experiment (8 hours) in all experiments (Fig. 7B, triangles).Collectively, our results suggest that Pf modulates P. aeruginosa virulence phenotypes in ways that repel C. elegans.

DISCUSSION
This study describes a convenient method to cure P. aeruginosa isolates of their Pf prophage infections and explores relationships between diverse Pf phages and their P. aeruginosa hosts.Overall, different Pf isolates exhibit varying effects on host quorum sensing and biofilm formation.One commonality between all Pf isolates examined is their ability to suppress pyocyanin production and repel C. elegans away from P. aeruginosa, protecting their host from predation.
Our results indicate that PflM maintains Pf in a lysogenic state and that deleting the pflM gene induces Pf prophage excision, but not replication.In Pf4, the site-specific tyrosine recombinase IntF4 catalyzes Pf4 prophage integration into and excision from the chromosome, whereas the Pf4 excisionase XisF4 regulates Pf4 prophage excision by promoting interactions between IntF4 and Pf4 attachment sites as well as inducing the expression of the replication initiation protein PA0727 (4,17).In response to stimuli such as oxidative stress (35), these coordinated events induce Pf4 prophage excision and initiate episomal replication, allowing Pf4 to complete its lifecycle.
Although it is presently not known how PflM maintains Pf lysogeny, it is possible that PflM regulates the IntF recombinase or inhibits the XisF excisionase, causing the Pf prophage to excise from the chromosome without concurrently inducing the replication initiation protein PA0727 ( 6), thus resulting in Pf prophage excision without initiating episomal Pf replication.Additionally, the DUF5447 domain (pfam17525) encoded by PflM is found only in Pseudomonads, indicating that the mechanism of PflM in maintaining Pf lysogeny is unique, compared with other lysogeny maintenance mechanisms employed by other phages.
Our study highlights the role of Pf phages in manipulating P. aeruginosa quorum sensing.Pf phages have varying effects on host quorum sensing; broadly, we determine that Pf phage modulates quorum sensing activity and quorum-regulated phenotypes in all bacterial strains tested.These findings imply that different Pf isolates interact with host quorum-sensing networks in diverse ways, indicating a complex interplay between Pf phages and host regulatory systems.
Quorum sensing regulates P. aeruginosa biofilm formation and Pf4 contributes to biofilm formation in PAO1 (5,7,14,18,36).Consistently, we find that both Pf4 and Pf6 contribute to PAO1 biofilm formation.Interestingly, the impact of Pf prophage deletion on biofilm formation varies among clinical isolates, which may be related to different quorum-sensing hierarchies present in clinical P. aeruginosa isolates (37).Alternatively, Pf prophage encodes accessory genes that may modulate host quorum-sensing systems.
Despite differences in interactions between Pf phages and host quorum sensing, deleting Pf prophages from the host chromosome enhances pyocyanin production in all strains tested except for strain CPA0053, which produces low levels of pyocyanin compared with all other strains tested.It is possible that the loss of PfsE in the ∆Pf strains in this study is responsible for the observed increase in pyocyanin production in most strains.As pyocyanin is the terminal signaling molecule in P. aeruginosa quorum-sensing networks (28), these results suggest this inhibition is the ultimate goal of Pf phages and may be beneficial to Pf phages during active replication.Pyocyanin and other redox-active phenazines are toxic to bacteria; it is possible that stress responses that are induced by pyocyanin-producing P. aeruginosa are detrimental to Pf replication.
We recently discovered that Pf phages encode a protein called PfsE (PA0721), which inhibits PQS signaling by binding to the anthranilate-coenzyme A ligase PqsA, and that this results in enhanced Pf replication (13).One possibility to explain the increased growth rate of LESB58 ∆Pf could be related to the dramatic increase in pyocya nin production by this strain, which could negatively impact bacterial growth.Further investigation is required to determine if this is true.
Pf lysogens induce avoidance behavior by C. elegans, which prefers to associate with the ∆Pf strains.Strikingly, although strains lacking Pf prophages are less virulent in a nematode infection model and attract C. elegans in the model described here, the reduced virulence of ∆Pf strains contrasts with their high pyocyanin virulence factor production.This discrepancy may be partly explained by our prior findings that Pf4 suppresses pyocyanin and other bacterial pigment production as a means to avoid detection by innate host immune responses (12) that are regulated by the aryl hydrocar bon receptor (37,38).
In summary, this research reveals the crucial role of the PflM gene in maintaining Pf lysogeny, demonstrates strain-specific effects on quorum sensing and biofilm formation, reveals the consistent inhibition of pyocyanin production by Pf phages, and suggests a role for Pf phages in protecting P. aeruginosa against nematode predation.

Strains, plasmids, primers, and growth conditions
Strains, plasmids, and their sources are listed in Table 2. Unless otherwise indicated, bacteria were grown in lysogeny broth (LB) at 37°C with 230 rpm shaking and supple mented with gentamicin (Sigma) where appropriate, at either 10 or 30 µg mL -1 .

Construction of deletion mutants
We used allelic exchange to delete alleles from P. aeruginosa (43).Briefly, to delete pflM (PA0718), upstream and downstream homologous sequences (~500 bp) were amplified through PCR from PAO1 genomic DNA using the UP and DOWN primers listed in Table 3.These amplicons were then ligated through splicing-by-overlap extension (SOE)-PCR to construct a contiguous deletion allele.This amplicon was then run on a 0.5% agarose gel, gel extracted (New England Biolabs #T3010L), and cloned (Gateway, Invitrogen) into a pENTRpEX18-Gm backbone to produce the deletion construct.The deletion construct was then transformed into DH5α, mini-prepped (New England Biolabs #T1010L), and sequenced (Plasmidsaurus.com).Sequencing-confirmed vectors were then transformed into Escherichia coli S17 Donor cells for biparental mating with the recipient P. aeruginosa strain.Single crossovers were isolated on VBMM (Vogel-Bonner minimal medium) agar supplemented with 30 µg/mL gentamicin, followed by the selection of double crossovers on no salt sucrose.The final obtained mutants were confirmed by excision assay (see below), Sanger sequencing of excision assay products, and whole genome sequencing.

Excision assays
Excision assays were designed as described previously (45).Briefly, a multiplex PCR assay was designed to produce amplicons of distinct sizes if the Pf prophage was integrated (primers Fwd_1 and Rev produce a smaller band) or excised (primers Fwd_2 and Rev produce a larger band) using Phusion Plus PCR Mastermix (Thermo Scientific # F631L).Primers were used at a final concentration of 0.5 µM and are listed in (Table 3).

Plaque assays
Plaque assays were performed using ΔPf4 as the indicator strain grown on LB plates.Phage in filtered supernatants was serially diluted 10× in phosphate buffered saline (PBS) and spotted onto lawns of PAO1 ΔPf4 .Plaques were imaged after 18 hours of growth at 37°C.PFUs/mL were then calculated.

Quantitative PCR
Cultures were grown overnight in LB broth with shaking at 37°C.Following 18 hours of incubation, cultures were pelleted at 16,000 × g for 5 minutes, washed 3× in 1× PBS, and treated with DNase at a final concentration of 0.1 mg/mL.qPCR was performed using SsoAdvanced Universal SYBR Green Supermix (BioRad #1725270) on the BioRad CFX Duet.For the standard curves, the sequence targeted by the primers was inserted into vectors pLM61 and pUC57-rplU, respectively, and 10-fold serial dilutions of the standard were used in the qPCR reactions with the appropriate primers (Table 3).Normalization to chromosomal copy number was performed as previously described (44) using 50S ribosomal protein gene rpIU.

Pyocyanin extraction and measurement
Pyocyanin was measured as previously described (20,46).Briefly, 18-hour cultures were treated with chloroform at 50% vol/vol.Samples were vortexed vigorously, and the organic phase was separated by centrifuging samples at 6,000 × g for 5 minutes.The chloroform layer was removed to a fresh tube, and 20% of the volume of 0.1 N HCl was added; then, the mixture vortexed vigorously.Once separated, the aqueous fraction was aliquoted to a 96-well plate, and the absorbance was measured at 520 nm.The concentration of pyocyanin, expressed as μg/mL, was obtained by multiplying the OD 520 nm by 17.072, as described previously (46).

Quorum-sensing reporters
Competent P. aeruginosa cells were prepared by washing overnight cultures in 300 mM sucrose, followed by transformation by electroporation (47) with the plasmids CP1 PBBR-MCS5 Empty, CP53 PBBR1-MCS5 pqsA-gfp, CP57 PBBR1-MCS5 rhlA-gfp, and CP59 PBBR1-MCS5 rsaL-gfp listed in Table 2. Transformants were selected by plating on the appropriate antibiotic selection media.The indicated strains were grown in buffered LB containing 50 mM MOPS (3-morpholinopropane-1-sulfonic acid) buffer and 100 µg mL -1 gentamicin for 18 hours.Cultures were then sub-cultured 1:100 into fresh LB MOPS buffer and grown to an OD 600 of 0.3.To measure reporter fluorescence, each strain was added to a 96-well plate containing 200 µL LB MOPS with a final bacterial density of OD 600 0.1 and incubated at 37°C in a CLARIOstar BMG LABTECH plate reader.Prior to each measurement, plates were shaken at 230 rpm for a duration of 2 minutes.A measurement was taken every 15 minutes for both growth (OD 600 ) and fluorescence (excitation at 485-15 nm and emission at 535-15 nm).End-point measurements at 18 hours were normalized to cell density.

C. elegans growth conditions
Synchronized adult N2 C. elegans were propagated on normal nematode growth medium (NNGM) agar plates with E. coli OP50 as a food source.

C. elegans avoidance assays
C. elegans avoidance assays were performed as previously described (33).Briefly, synchronized adult N2 worms were propagated at 24°C on 3.5 cm NNGM agar plates with E. coli OP50 for 48 hours, collected, and washed 4× to remove residual OP50.NNGM agar was spotted with 20 µL of P. aeruginosa (Pf lysogens and their isogenic ∆Pf mutant) overnight cultures (LB broth) as shown in Fig. 7A and grown for 18 hours at 37°C.Worms

Statistical analyses
Differences between data sets were evaluated with a Student's t-test (unpaired, two-tailed) or two-way ANOVA using the Šidák correction (95% confidence interval threshold) where appropriate.P values of < 0.05 were considered statistically significant.GraphPad Prism version 9.4.1 (GraphPad Software, San Diego, CA) was used for all analyses.

FIG 1
FIG 1 The Pf4 phage gene PA0718 (pflM) maintains lysogeny.(A) The Pf4 prophage is shown.(B) Multiplex PCR was used to measure Pf4 prophage integration and excision from the PAO1 chromosome in Pf4 single-gene mutants, which were generated through allelic exchange.Deletion of the Pf4 repressor (Pf4r) and PA0718 results in prophage excision.Despite our best efforts, we were unable to generate a single-gene mutant of Pf4 gene pfsE, marked with an asterisk.(C) Quantitative PCR (qPCR) was used to measure episomal Pf4 replicative form in cells after 18 hours of growth in lysogeny broth (LB) broth.Data are the mean ± SEM of three replicate experiments.The lower limit of detection for the assay is 37.85 copy numbers per 10,000 cells.**P < 0.01, Student's t-test.(D) PA0718and/or XisF4 were expressed from an inducible plasmid in P. aeruginosa PAO1.After 18 hours, Pf4 integration and excision were measured by excision assay.(E)Pf4 virions in filtered supernatants collected from the indicated strains were titered on lawns of P. aeruginosa ∆Pf4.A representative image is shown.

FIG 2
FIG 2 Targeted deletion of pflM cures diverse P. aeruginosa isolates of their Pf prophage infections.A multiplex PCR assay and long-read whole genome sequencing were used to confirm the loss of (A and B) the Pf4 prophage or (C and D) the Pf6 prophage from the PAO1 chromosome.(E-I) Long-read whole genome sequencing was used to confirm the successful deletion of the indicated Pf prophages.Reads were aligned to 50 kb sequences flanking the Pf prophage insertion sites in the parental chromosome.The genomic coordinates for each Pf prophage are shown above each bracket.

FIG 3
FIG 3 Sequence and structural analysis of PflM.(A) Pf prophage annotation was performed with Rapid Annotation using Subsystem Technology (RAST).Inset: Clinker was used to generate alignments of the pflM loci spanning from xisF to PA0719 in the Pf4 reference sequence.DUF5447 and mnt domains in pflM are indicated by orange or green, respectively, in the inset.(B) PflM structures were predicted using AlphaFold2and aligned using ChimeraX.The highlighted region indicates the DUF5447 domain and the Mnt domain is indicated for the PflM sequence from strain DDRC3.(C) A protein sequence logo for PflM was generated with WebLogo3 (v.2.8.2); dashes represent variable residues.The DUF5447 domain is highlighted.

FIG 7 C
FIG 7 C. elegans actively avoids P. aeruginosa Pf lysogens.(A) Experimental design: P. aeruginosa and isogenic ∆Pf mutants were spotted on normal nematode growth medium (NNGM) plates with wild-type N2 C. elegans at the indicated locations.C. elegans localization to the indicated quadrants was measured hourly.(B) C. elegans association with P. aeruginosa (circles) or isogenic ∆Pf mutants (squares) in the indicated strain backgrounds was measured hourly over 8 hours (three experiments with N = 30 per replicate [90 animals total]).P values were calculated by two-way ANOVA (analysis of variance) comparing ∆Pf strains with the parental strains using the Šidák correction (95% confidence interval threshold), **P < 0.01, ***P < 0.001, ****P < 0.0001.

TABLE 1 P
. aeruginosa isolates and Pf prophage characteristics

TABLE 2
Strains and plasmids used in this study

TABLE 3
Primers used in this study plated in triplicate and incubated at 24°C.C. elegans migration was monitored hourly for 8 hours. were