Abstract
Bacteriophages (phages) are being considered as potential alternative therapeutics for the treatment of multidrug resistant bacterial infections. Considering most phages have a narrow host-range, the generally accepted dogma is that therapeutic phages will have a marginal impact on bacterial strains outside of their intended target bacterium. We have discovered that lytic phage infection induces transcription of type VIIb secretion system (T7SS) genes in the pathobiont Enterococcus faecalis. Phage induction of T7SS genes mediates cell contact dependent antagonism of diverse Gram positive bystander bacteria. This phage induced T7SS antagonism is attributed to cell membrane damage. Phage driven T7SS antagonism of neighboring cells is abrogated by deleting essB, a T7SS structural component that is required for secretion of T7SS toxic substrates. Expression of a predicted immunity gene in bystander bacteria confers protection against T7SS mediated inhibition, implicating an upstream LXG domain toxin in intraspecies antagonism. Additionally, phage induction of T7SS gene expression requires IreK, a Serine/Threonine PASTA kinase. Phage induction of T7SS antimicrobial activity signals through a non-canonical IreK stress response pathway. Our findings highlight how phage infection of a target bacterium can unintentionally affect neighboring bystander bacteria. Furthermore, our work indicates that before phages become a standard of care in the clinic, we must clearly understand how bacteria respond to phage infection and what, if any, collateral effects phage therapy may have on non-target bacteria, such as the microbiota.
Introduction
Enterococci constitute a minor component of the healthy human microbiota (1). Enterococci, including Enterococcus faecalis, are also nosocomial pathogens that cause a variety of diseases, including sepsis, endocarditis, surgical-site, urinary tract and mixed bacterial infections (2, 3). Over recent decades, enterococci have acquired extensive antibiotic resistance traits, including resistance to “last-resort” antibiotics such as vancomycin, daptomycin, and linezolid (4-8). Following antibiotic therapy, multi-drug resistant (MDR) enterococci can outgrow to become a dominant member of the intestinal microbiota, resulting in intestinal barrier invasion and blood stream infection (7, 9). The ongoing evolution of MDR enterococci in healthcare settings (4-6, 10, 11) and their ability to transmit antibiotic resistance among diverse bacteria (9, 12-15), emphasize the immediate need for novel therapeutic approaches to control enterococcal infections.
Viruses that infect and kill bacteria (bacteriophages or phages) are receiving attention for their use as antibacterial agents (16). Recent studies have demonstrated the efficacy of anti-enterococcal phages in murine models of bacteremia (17-19) and the administration of phages to reduce E. faecalis burden in the intestine gives rise to phage resistant isolates that are resensitized to antibiotics (20). Considering phages are highly specific for their target bacterium, coupled with the self-limiting nature of their host-dependent replication, suggests that unlike antibiotics which have broad off-target antimicrobial activity, phages are likely to have nominal impact on bacteria outside of their intended target strain (21-23). However, our understanding of how phages interact with bacteria and the bacterial response to phage infection is limited.
While studying the transcriptional response of phage infected E. faecalis cells, we discovered that phage infection induced the expression of genes involved in the biosynthesis of a type VIIb secretion system (T7SS) (24). Firmicutes, including the enterococci, harbor diverse T7SS genes encoding transmembrane and cytoplasmic proteins involved in the secretion of protein substrates (25), and T7SSs promote antagonism of non-kin bacterial cells through contact-dependent secretion of antibacterial effectors and/or toxins (26, 27). The antibacterial activity of T7SSs from Staphylococcus and Streptococcus are well characterized (25) but T7SS-mediated antibacterial antagonism has not been described in Enterococcus. The environmental cues and regulatory pathways that govern T7SS expression and activity are poorly understood, although recent studies indicate that exposure to serum and membrane stresses triggered by pulmonary surfactants, fatty acids and phage infection stimulate T7SS gene expression (24, 28-31). This motivated us to determine if phage induced T7SS gene expression in E. faecalis results in the inhibition of non-kin bacterial cells that are not phage targets (bystanders). We discovered that phage infected E. faecalis produces potent T7SS antibacterial activity against bystander bacteria. Expression of a T7SS antitoxin (immunity factor) gene in bystander cells confers protection against phage mediated T7SS inhibition. Additionally, we discovered that membrane stress during phage infection induces transcription of T7SS genes via a non-canonical IreK signaling pathway. To our knowledge, the enterococcal T7SS is the only example of secretion system induction during phage infection. These data shed light on how phage infection of a cognate bacterial host can influence polymicrobial interactions and raises the possibility that phages may impose unintended compositional shifts among bystander bacteria in the microbiota during phage therapy.
Results
Phage mediated induction of E. faecalis T7SS leads to interspecies antagonism
A hallmark feature of phage therapy is that because phages often have a narrow host range, they will not influence the growth of non-susceptible bacteria occupying the same niche (22). We discovered that infection of E. faecalis OG1RF by phage VPE25, induces the expression of T7SS genes (24). The E. faecalis OG1RF T7SS locus is absent in the commonly studied vancomycin-resistant strain V583, despite conservation of flanking genes (Fig. 1A) (32, 33). Homologs of the E. faecalis T7SS gene exsA are found throughout three of the four Enterococcus species groups (34), including Enterococcus faecium, suggesting a wide distribution of T7SS loci in enterococci (Fig. S1). In silico analyses predict that the E. faecalis T7SS locus encodes multiple WXG100 family effectors and LXG family polymorphic toxins (27, 35). Hence, we hypothesized that induction of T7SS genes during phage infection and consequently the heightened production of T7SS substrates would indirectly influence the growth of non-kin phage-resistant bacterial cells.
To investigate if T7SS factors produced during phage infection of E. faecalis OG1RF interferes with the growth of phage-resistant bystander bacteria, we generated a strain with an in-frame deletion in the T7SS gene essB, encoding a transmembrane protein involved in the transport of T7SS substrates (36). The essB mutant is equally susceptible to phage VPE25 infection compared to wild type E. faecalis OG1RF (Fig. S2). We performed co-culture experiments where phage susceptible wild type E. faecalis OG1RF or ΔessB were mixed with a phage resistant bystander, a strain of E. faecalis V583 deficient in the production the VPE25 receptor (ΔpipV583) (37), at a ratio of 1:1 in the absence and presence of phage VPE25 (multiplicity of infection [MOI] = 0.01) (Fig. 1B). Since the E. faecalis V583 genome does not contain T7SS genes, this strain should lack immunity to T7SS antagonism (33). The viability of E. faecalis ΔpipV583, was reduced nearly 100-fold when co-cultured with E. faecalis OG1RF in the presence of phage VPE25 (Fig. 1C). However, growth inhibition of E. faecalis ΔpipV583 was abrogated during co-culture with phage infected E. faecalis ΔessB. Phage induced T7SS antagonism could be restored by complementation (Fig. 1C), indicating that inhibition of phage resistant E. faecalis ΔpipV583 is T7SS dependent.
T7SS encoded antibacterial toxins secreted by Gram positive bacteria influence intra- and interspecies antagonism (26, 27). While a nuclease toxin produced by Staphylococcus aureus targets a closely related S. aureus strain (26), Streptococcus intermedius exhibits T7SS dependent antagonism against a wide-array of Gram positive bacteria (27). To determine the target range of E. faecalis OG1RF T7SS antibacterial activity, we measured the viability of a panel of VPE25 resistant Gram positive and Gram negative bacteria in our co-culture assay (Fig. 1B). Growth inhibition of the distantly related bacterial species E. faecium and Gram positive bacteria of diverse genera, including S. aureus and Listeria monocytogenes, occurred following co-culture with phage infected wild type E. faecalis OG1RF but not the ΔessB mutant (Fig. 1D). In contrast, Gram positive streptococci were unaffected (Fig. 1D). Similarly, phage induced T7SS activity did not inhibit any Gram negative bacteria tested (Fig. 1D). Collectively, these results show that phage predation of E. faecalis promotes T7SS inhibition of diverse bystander bacteria.
Molecular basis of E. faecalis phage–triggered T7SS antagonism
Our data demonstrate that induction of E. faecalis OG1RF T7SS genes during phage infection hinder the growth of select non-kin bacterial species. Antibacterial toxins deployed by Gram negative bacteria via type VI secretion and Gram positive T7SS require physical contact between cells to achieve antagonism (26, 27, 38, 39). Therefore, we investigated if growth inhibition of bystander bacteria is contingent upon direct interaction with phage infected E. faecalis using a trans-well assay (27). We added unfiltered supernatants from wild type E. faecalis OG1RF and ΔessB mutant cultures grown for 24 hrs in the presence and absence of phage VPE25 (MOI = 0.01) to the top of a trans well and deposited phage resistant E. faecalis ΔpipV583 in the bottom of the trans well. The 0.4 µm membrane filter that separates the two wells is permeable to proteins and solutes but prevents bacterial translocation. Supernatant from phage infected wild type E. faecalis OG1RF did not inhibit E. faecalis ΔpipV583 (Fig. 2A) indicating that T7SS mediated growth interference relies on cell to cell contact. To exclude the possibility that T7SS substrates might adhere to the 0.4 µm membrane filter in the trans-well assay, we administered unfiltered culture supernatants directly to E. faecalis ΔpipV583 cells (5×105 CFU/well) at a ratio of 1:10 and monitored growth over a period of 10 hours. Growth kinetics of E. faecalis ΔpipV583 remained similar irrespective of the presence or absence of conditioned supernatant from wild type E. faecalis OG1RF or ΔessB mutant cultures (Fig. 2B), further supporting the requirement of contact-dependent engagement of phage mediated T7SS inhibition.
We discovered that E. faecalis OG1RF inhibits proliferation of non-kin bacterial cells through increased expression of T7SS genes in response to phage infection, but the toxic effectors are unknown. LXG domain containing toxins are widespread in bacteria with a diverse range of predicted antibacterial activities (40, 41). The OG1RF T7SS locus encodes two LXG-domain proteins, expressed from OG1RF_11109 and OG1RF_11121 (Fig. 3A). Polymorphic toxin systems in Gram negative and Gram positive bacteria can encode “orphan” toxins in addition to the primary secreted effectors (42). To identify putative orphan effectors, we aligned OG1RF_11109 and OG1RF_11121 with downstream genes in the T7SS locus. OG1RF_11109 shared conserved sequence with OG1RF_11111 and OG1RF_11113, and OG1RF_11121 had homology to OG1RF_11123 (Fig 3A). Homologs to these putative orphan toxins are found as full-length LXG proteins in E. faecalis and Listeria sp. (Fig. S3), suggesting they are bonafide T7SS effectors. Interestingly, we identified an additional LXG gene product, OG1RF_12414, in a distal locus that is again notably absent from V583 (Fig. S4A and S4B). Although none of the five putative toxins in the T7SS locus have recognizable C-terminal domains, OG1RF_12414 has predicted structural homology to Tne2, a T6SS effector with NADase activity from Pseudomonas protegens (Fig. S4C) (43). Additionally, we identified numerous C-terminal domains in LXG proteins distributed throughout Enterococcus (Fig. S5). These include EndoU and Ntox44 nuclease domains (40, 44, 45), which have been characterized in effectors produced by other polymorphic toxin systems.
Polymorphic toxins are genetically linked to cognate immunity proteins that neutralize antagonistic activity and prevent self-intoxication (40, 45, 46). Each of the five putative toxins in the OG1RF T7SS locus is encoded directly upstream of a small protein that could function in immunity. Whitney et al. demonstrated that the cytoplasmic antagonistic activity of S. intermedius LXG toxins TelA and TelB in E. coli can be rescued by co-expression of the corresponding immunity factors (27). Therefore, we examined if OG1RF_11110, 11112, or 11122, confer immunity to E. faecalis ΔpipV583 during phage infection of E. faecalis OG1RF. Constitutive expression of OG1RF_11122, and not OG1RF_11110 or 11112, partially neutralized phage induced T7SS antagonism (Fig. 3B), confirming an essential role for the OG1RF_11122 gene product in immunity, and suggesting that OG1RF_11121 is at least partly responsible for T7SSb mediated intra-species antagonism.
Sub-lethal antibiotic stress promotes T7SS dependent antagonism
Considering two genetically distinct phages trigger the induction of T7SS genes in E. faecalis (24), we reasoned that T7SS induction could be a result of phage mediated cellular damage and not specifically directed by a phage encoded protein. Antibiotics elicit a range of damage induced stress responses in bacteria (47-49), therefore, we investigated the effects of subinhibitory concentrations of antibiotics on T7SS expression in E. faecalis.
To investigate the influence of sublethal antibiotic concentrations on E. faecalis OG1RF T7SS transcription, we determined the minimum inhibitory concentrations (MIC) of ampicillin, vancomycin, and daptomycin (Fig.S6A-S6C) and monitored T7SS gene expression in E. faecalis OG1RF cells treated with a sub-lethal dose of antibiotic (50% of the MIC). We found that bacterial T7SS genes were significantly upregulated in the presence of the cell membrane targeting antibiotic, daptomycin, relative to the untreated control (Fig. 4A). In contrast, the cell wall biosynthesis inhibitors ampicillin and vancomycin either did not induce or had a minor impact on T7SS mRNA levels, respectively (Fig. 4A). Additionally, induction of T7SS transcription occurred when bacteria are challenged with a sub-inhibitory concentration of the DNA targeting antibiotics ciprofloxacin and mitomycin C (Fig. 4B, Fig. S6D–S6E). Collectively, these data show that T7SS induction in E. faecalis occurs in response to cell envelope and DNA stress.
We next assessed the influence of daptomycin driven T7SS induction on inter-enterococal antagonism. Since T7SS expression is less robust in the presence of daptomycin compared to phage infection (24), a 10:1 ratio of daptomycin treated E. faecalis OG1RF was required for growth inhibition of E. faecalis ΔpipV583 during co-culture (Fig. 4C). Consistent with our previous results, daptomycin induced T7SS inhibition of E. faecalis ΔpipV583 was contact dependent (Fig. 4D). To facilitate T7SS mediated contact-dependent killing of the target strain during daptomycin exposure, we performed the inhibition assay on nutrient agar plates. The sub-inhibitory concentration of daptomycin (2.5 µg/ml) used in liquid culture was toxic to the cells on agar plates, so we lowered the daptomycin concentration to 0.5 µg/ml to prevent drug toxicity in the agar-based antagonism assay. Plating T7SS producing E. faecalis OG1RF cells and E. faecalis ΔpipV583 bystander cells at a ratio of 10:1 resulted in ∼10–fold inhibition of bystander growth (Fig. 4E). These data show that in addition to phages, antibiotics can be sensed by E. faecalis thereby inducing T7SS antagonism of non-kin bacterial cells. These data also suggest that the magnitude of T7SS gene expression is directly related to the potency of T7SS inhibition.
IreK facilitates T7SS expression in phage infected E. faecalis OG1RF via a non-canonical signaling pathway
Having established that phage and daptomycin mediated membrane damage stimulates heightened E. faecalis OG1RF T7SS gene expression and antagonistic activity, we next sought to identify the genetic determinants that sense this damage and promote T7SS transcription. Two-component systems, LiaR/S and CroS/R, and the PASTA kinase family protein IreK are well-characterized modulators of enterococcal cell envelope homeostasis and antimicrobial tolerance (50-52). Aberrant cardiolipin microdomain remodeling in the bacterial cell membrane in the absence of the LiaR response regulator results in daptomycin hypersensitivity and virulence attenuation (53). CroS/R signaling and subsequent modulation of gene expression govern cell wall integrity and promote resistance to cephalosporins, glycopeptides and beta—lactam antibiotics (54-56). The ireK encoded transmembrane Ser/Thr kinase regulates cell wall homeostasis, antimicrobial resistance, and contributes to bacterial fitness during long-term colonization of the intestinal tract (51, 57, 58). Recently it has been shown that direct cross-talk between IreK and the CroS/R system positively impacts enterococcal cephalosporin resistance (59).
Wild type E. faecalis OG1RF, an ireK in-frame deletion mutant (51) and transposon (Tn) insertion mutants of liaR, liaS, croR, and croS (60) all display similar growth kinetics in the absence of phage VPE25 infection (Fig. S7A). Although croR-Tn and croS-Tn exhibit reductions in the plaquing efficiency of VPE25 particles, none of these genetic elements of enterococcal cell wall homeostasis and antibiotic resistance were required for VPE25 infection (Fig. S7B). We queried the expression levels of T7SS genes in these isogenic mutants during phage VPE25 infection (MOI = 1). T7SS gene expression was not enhanced in the ΔireK mutant during phage infection (Fig. 5A), whereas liaR-Tn, liaS-Tn, croR-Tn, and croS-Tn produced heightened levels of T7SS transcripts similar to the wild type E. faecalis OG1RF compared to the uninfected controls (Fig. S8A-S8F). A sub-lethal concentration of the cephalosporin ceftriaxone did not induce T7SS gene expression (Fig. S9), indicating that expression of T7SS genes following phage mediated membrane damage signals through a pathway that is distinct from the IreK response to cephalosporin stress. Additionally, the ΔireK mutant phenocopies the ΔessB mutant strain in the interbacterial antagonism co-culture assay, wherein the ΔireK mutant is unable to mediate phage induced T7SS dependent killing of the phage resistant E. faecalis ΔpipV583 non-kin cells (Fig. 5B). Collectively, these results indicate that IreK senses phage mediated membrane damage promoting T7SS transcription independent of the CroS/R pathway.
Discussion
Despite the fact that bacteria exist in complex microbial communities and engage in social interactions (61, 62), phage predation studies have primarily been performed in monoculture (24, 63-65). Studies report phage-mediated effects on non-target bacteria linked to interbacterial interactions and evolved phage tropism for non-cognate bacteria (66-68), whereas other studies have identified minimal changes in microbiota diversity during phage therapy (66, 69).
Our results extend previous work that observed the induction of E. faecalis OG1RF T7SS gene expression in response to phage infection (24). By using an in vitro antibacterial antagonism assay, we discovered that phage predation of E. faecalis OG1RF has an inhibitory effect on non-phage targeted bacterial species during co-culture. Our work shows that phage mediated inhibition of Gram positive bystander bacteria relies on the expression and activity of T7SS genes. This work emphasizes how phage infection of target bacteria can extend beyond intended phage targets and inhibit other members of a microbial community. This discovery could have profound ramifications on how microbial communities like the microbiota respond to phage therapy.
Our data suggest that membrane stress associated with phage infection or sub-lethal daptomycin treatment stimulates T7SS mediated antibacterial antagonism of E. faecalis OG1RF. Given that daptomycin is used to target vancomycin-resistant enterococcal infections, this finding provides a model by which antibiotic-resistant enterococci may overgrow and dominate the microbiota after antibiotic treatment. Although further investigation is required to understand how T7SS induction might contribute to enterococcal fitness in polymicrobial environments, disruption of T7SS loci in other bacteria compromises bacterial membrane integrity and attenuates virulence (31, 70). It is possible that environmental conditions encountered in the intestinal tract, including bile salts, antimicrobial proteins, and competition for nutrient resources could influence T7SS activity in E. faecalis to facilitate niche establishment and/or persistence within a complex microbial community. Indeed, E. faecalis T7SS mutants have a defect in their ability to colonize the murine reproductive tract (manuscript in preparation). Further, we discovered that transcriptional activation of the T7SS during phage infection relies on IreK. Previously characterized IreK–mediated stress response pathways, including cephalosporin stress or the CroS/R signaling, did not contribute to T7SS expression. We hypothesize that IreK senses diverse environmental stressors and coordinates distinct outputs in response to specific stimuli. Considering that IreK signaling is important for E. faecalis intestinal colonization (58), it is possible that IreK–dependent T7SS expression in response to intestinal cues modulate interbacterial interactions and enterococcal persistence in the intestine. However, the molecular mechanism by which IreK facilitates T7SS transcription remains unanswered.
Antibacterial properties of T7SS substrates have been demonstrated (26, 27). Here we show that the expression of an immunity gene OG1RF_11122 in T7SS targeted E. faecalis cells likely confers protection from inhibition by the upstream LXG toxin encoded by OG1RF_11121. Aside from its LXG domain, OG1RF_11121 does not harbor any other recognizable protein domains, hence the mechanism underlying its toxicity is unclear. Whitney et al. demonstrated that LXG toxin antagonism is contact–dependent, having minimal to no impact on target cells in liquid media (27). Although we found that physical engagement is crucial for E. faecalis T7SS mediated antagonism, we observed a significant reduction in target cell growth in liquid media both during phage and daptomycin treatment of T7SS proficient E. faecalis. Together, these data suggest that heightened T7SS transcription upon phage exposure compared to daptomycin exposure may account for robust E. faecalis OG1RF inhibition of other bacteria in liquid culture.
Enterococci occupy polymicrobial infections often interacting with other bacteria (71-74). Although commensal E. faecalis antagonize virulent S. aureus through the production of superoxide (75), the two species also exhibit growth synergy via exchange of critical nutrients (76). Here, we show that phage treatment of E. faecalis OG1RF can indirectly impact the growth of neighboring phage-resistant bacteria, including S. aureus, in a T7SS–dependent manner, suggesting that phage therapy directed against enterococci driving T7SS activity could be useful for the treatment of polymicrobial infections. However, the counter argument is that phage therapy directed against enterococci could push a bacterial community toward dysbiosis, as phage induced T7SS activity could directly inhibit beneficial bystander bacteria. This raises questions about the consequences of phage mediated off-target effects on bacteria. Could phage induced T7SS activity be used to reduce phage expansion into other closely related strains as a means to dilute phages out of a population, or is it simply that phage induction of the T7SS serves as a mechanism that benefits a select few within a population to aid in their reoccupation of a niche upon overcoming phage infection? Future studies aimed at exploring enterococcal T7SS antagonism in polymicrobial communities should help elucidate the impact of phages on microbial community composition.
Materials and Methods
Bacteria and bacteriophage
Bacteria and phages used in this study are listed in Table S1. Bacteria were grown with aeration in Todd-Hewitt broth (THB) or on THB agar supplemented with 10mM MgSO4 at 37°C. The following antibiotic concentrations were added to media for the selection of specific bacterial strains or species: E. faecalis OG1RF (25 μg/ml fusidic acid, 50 μg/ml rifampin), E. faecalis V583 ΔpipV583 (100 μg/ml gentamycin), S. aureus AH2146 LAC Φ11:LL29 (1 μg/ml tetracyclin), L. monocytogenes 10403S (100 μg/ml streptomycin), V. cholerae C6706 int I4::TnFL63 and S. enterica serovar Typhimurium 140285 put::Kan (50 μg/ml kanamycin). S. agalactiae COH1 was distinguished from E. faecalis on Chrome indicator Agar (CHROMagar StrepB SB282). We were unable to differentially select E. coli, S. pyogenes and S. mitis from E. faecalis based on antibiotic sensitivity. Therefore, colony counts of these bacteria in co-culture experiments were acquired by subtracting the E. faecalis colony numbers on selective media from total number of colonies on non-selective media. Strains harboring pLZ12A and its derivatives were grown in the presence of 20 μg/ml chloramphenicol.
Bioinformatic analyses
Genome sequences of E. faecalis V583 (NC_004668.1) and OG1RF (NC_017316.1) were obtained from NCBI. Alignments were generated and visualized using EasyFig (77). OG1RF protein domains were identified using KEGG (78) and ExPASy PROSITE (79). Structure modeling of OG1RF_12414 was done with Phyre2 (80). The EsxA phylogenetic tree was constructed in MEGA version X (81) using non-redundant protein sequences obtained from NCBI BLAST (82) with OG1RF_11100 as input and was edited using the Interactive Tree Of Life browser (83). OG1RF_11109 was used as an input for the NCBI Conserved Domain Architecture Retrieval Tool (84) to identify protein domains that co-occur with LXG domains in Enterococcus.
Antibiotic sensitivity profiles
Antibiotic susceptibility profiles for ampicillin, vancomycin, and daptomycin were determined using a broth microdilution assay. Overnight (O/N) E. faecalis OG1RF cultures were diluted to 5 × 106 CFU/ml and 100 µl was added to each well of a 96-well plate to give a final cell density of 5 × 105 CFU/ml. Antibiotic stocks were added to the first column of each row, mixed thoroughly, and were serially diluted 2-fold across the rows. The last column was used as a no drug control. Cultures containing daptomycin were supplemented with 50 µg/ml CaCl2. Bacterial growth was monitored by measuring absorbance (OD600) using a Synergy H1 microplate reader set to 37°C with continuous shaking O/N. Growth curves are presented as the average of three biological replicates. A concentration of antibiotic just below the drug amount that inhibits bacterial growth was deemed sub-lethal and used to examine T7SS genes expression.
Co-culture bacterial antagonism assays
For inter- and intraspecies antagonism assays in liquid media, O/N cultures of different bacteria were diluted in THB containing 10mM MgSO4 to an OD600 of 0.2 and mixed together in a 1:1 or 10:1 ratio. The mixed cell suspensions were either left untreated or treated with phage VPE25 (MOI 0.01) / daptomycin (2.5 μg/ml) and grown at 37°C with aeration. For antagonism experiments on agar plates, O/N cultures of different strains were diluted to an OD600 of 0.2 and mixed together in a 1:1 or 10:1 ratio. A total of 107 cells from mixed culture suspension was added to 5 ml THB + 0.35% agar at 55°C and were poured over the surface of a THB agar plate in the absence or presence of daptomycin (0.5 µg/ml). The plates were incubated at 37°C under static conditions for 24 hours. Cells were harvested by scraping off the top agar, resuspending in 5 ml of PBS, and the cfus were obtained by plating serially diluted cell suspension on appropriate selective or differentiating agar plates. Relative viability was calculated from the ratio of target strain cfu in the treated versus the untreated co-culture. The assays were performed in biological triplicates.
RNA extraction and quantitative PCR
RNA was extracted from phage treated or untreated E. faecalis OG1RF cells by using an RNeasy Mini Kit (Qiagen) with the following published modifications (24). cDNA was generated from 1 µg of RNA using qScript cDNA SuperMix (QuantaBio) and transcript levels were analyzed by qPCR using PowerUp™SYBR Green Master Mix (Applied Biosystems). Transcript abundances were normalized to the 16S rRNA gene transcripts and fold–change was calculated by comparing to untreated controls. All data are represented as the average of three biological replicates.
Bacterial growth curves
25 ml of 10mM MgSO4 supplemented THB was inoculated with O/N cultures of E. faecalis diluted to an OD600 of 0.025 and distributed to a 96-well plate in 0.1 ml volumes. Cultures were incubated at 37° C with aeration. OD600 was measured periodically for 18 hours in a Synergy H1 microplate reader.
Efficiency of plating (EOP) assays
To investigate if phage VPE25 can infect and lyse E. faecalis mutants and various other bacterial species, 107 PFU/ml of phage was serially diluted and the phage was titered on each strain using a THB agar overlay plaque assay. EOP is expressed as the percentage of phage titer from each strain relative to the wild type E. faecalis OG1RF control. Data are presented as the average of three biological replicates.
Construction of E. faecalis mutants and complementation
Isolation of E. faecalis genomic DNA was performed using a ZymoBIOMICS™ DNA Miniprep Kit (Zymo Research). All PCR reactions used for cloning were performed with high fidelity KOD Hot Start DNA Polymerase (EMD Millipore). E. faecalis ΔessB was generated by allelic replacement by cloning an in frame essB deletion product into pLT06 using Gibson Assembly® Master Mix (New England Biolabs), integrating this construct into the chromosome, and resolving the deletion mutant by homologous recombination (85-87). For ectopic expression of putative immunity proteins, coding regions of OG1RF_11110, OG1RF_11112, and OG1RF_11122 were cloned downstream of the bacA promoter (PbacA) by restriction digestion and ligation into the shuttle vector pLZ12A (20). Primer sequences and restriction enzymes used for cloning are listed in Table S1. Plasmids were introduced into electrocompetent E. faecalis cells as previously described (20).
Statistical analysis
Statistical tests were performed using GraphPad – Prism version 8.2.1. For qPCR and bacterial competition assays, unpaired Student’s t-tests were used. P values are indicated in the figure legends.
Figure Legends
Acknowledgments
This work was supported by National Institutes of Health grants R01AI141479 (B.A.D.) and R01AI122742 (G.M.D.). J.L.E.W. was supported by American Heart Association Grant 19POST34450124 / Julia Willett / 2018. We would like to thank Andrés Vázquez-Torres, Laurel Lenz, Alex Horswill, Stefan Pukatzki, Kelly Doran, and their lab members for sharing bacterial strains used in this study.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.
- 6.
- 7.↵
- 8.↵
- 9.↵
- 10.
- 11.
- 12.↵
- 13.
- 14.
- 15.↵
- 16.↵
- 17.↵
- 18.
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.
- 30.
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.
- 65.↵
- 66.↵
- 67.
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.
- 73.
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
- 86.
- 87.↵