Bacteriophage-mediated reduction of uropathogenic E. coli from the urogenital epithelium

ABSTRACT

Urinary tract infections (UTIs), primarily caused by uropathogenic Escherichia coli (UPEC), affect millions annually. UPEC gains access to the urinary tract through mucosal reservoirs, including the vaginal tract. With rising antibiotic resistance and frequent recurrence, alternative non-antibiotic strategies like bacteriophage (phage) therapy are gaining attention. We explored the potential of a lytic phage, ΦHP3, as well as a phage cocktail to decolonize UPEC from the urogenital tract using in vitro and in vivo models. Phage demonstrated replication and lytic activity in both bacteriologic medium and simulated vaginal fluid. Pretreatment of human vaginal epithelial cells (VK2/E6E7) and bladder carcinoma cells (HTB-9) with phage reduced adhesion and invasion of UPEC compared with controls. Phage treatment was further able to reduce intracellular UPEC in VK2 cells. Notably, phage pretreatment did not impact phage-resistant UPEC strains, indicating that phage lysis was the primary driver of phenotypes. Live confocal microscopy confirmed the interaction of phage particles with UPEC and with both epithelial cell lines. In vivo, daily intravaginal ΦHP3 administration in humanized microbiota mice significantly reduced vaginal UPEC burden after 4 days. Treatment with a phage cocktail also reduced vaginal and cervical tissue burdens by day 7 post-treatment. UPEC dissemination was observed in uterine and kidney tissues, but burdens were not different between phage and mock-treated groups. In conclusion, we demonstrate that phage and phage cocktails can modestly reduce UPEC urogenital colonization, highlighting the potential of phage therapy as a viable prevention strategy for UTI.

INTRODUCTION

Urinary tract infections (UTIs), at an estimated 400 million cases per year worldwide, are one of the most common bacterial infections (1). UTIs disproportionately afflict women, with more than half of women developing at least one UTI in their lifetime (2). Uropathogenic Escherichia coli (UPEC) is the predominant agent of uncomplicated and complicated UTIs across the human lifespan (2–4). Recurrent UTIs (rUTIs), defined as re-infection within 6 months, occur in about a third of cases (5) and are often caused by the same species or strain (6, 7), suggestive of a host reservoir that reseeds infection. Simultaneous gastrointestinal and bladder detection of the same UPEC strain is observed in UTI (8), rUTI (9, 10), and even asymptomatic bacteriuria (ASB) (11), suggesting gut dissemination as a precursor to UTI. However, rUTI patients show similar patterns of fecal E. coli relative abundance, E. coli blooms, and genetic lineages as controls (12, 13). Similarly, the urinary microbiota, including prevalence of E. coli, is not different between patients with rUTI history compared to controls (14, 15), suggesting the bladder itself is also not a consistent reservoir for UPEC. Conversely, vaginal E. coli prevalence is estimated to be 11%–12% (16–18) but increases up to 35%–40% in women with rUTI (18, 19), up to 80% in women with acute UTI (20), with increased relative abundance also observed in ASB (21). Vaginal introital or peri-urethral E. coli colonization precedes UTI symptom onset (22–24), and isolates are frequently the same strain identified in urine (25). Collectively, these observations support vaginal UPEC reservoirs as a critical driver of UTI and, consequently, identify this as an important site to target for UTI prevention (26).

UTI is the second most common indication of antibiotic prescriptions (27–30), and antibiotic resistance rates among UPEC are on the rise globally (31). Prior antibiotic usage increases the risk for UTI by two- to fivefold (32) and the risk of acquiring a subsequent multidrug-resistant UTI (33). Recurrent cycles of infection and antibiotic treatment are also associated with elevated frequency of antibiotic resistance genes in the urogenital microbiome (14). Along with antibiotic resistance, UPEC undermines antibiotic efficacy by establishing intracellular bacterial communities within the bladder (34, 35) and the vaginal epithelium (36), which are protected from antibiotic therapy and could serve as an additional source of recurrent infection (37). Recalcitrant reservoirs, antibiotic resistance, and detrimental microbiome effects of repeated antibiotic usage necessitate non-antibiotic alternatives to decolonize UPEC from mucosal and intracellular sites.

Bacteriophages (phages), viruses that infect bacteria, are known to impact E. coli mucosal colonization. Expansion of endogenous E. coli phage and concordant decrease in relative E. coli abundance have been observed following fecal microbiome transplant in humans (38). Similarly, diverse populations of endogenous phage populations, predicted to be active against vaginal species, including E. coli, are associated with vaginal bacterial composition (39–41). Phages have been explored as potential therapeutics for UTIs since their discovery more than a century ago, with mixed success (42). Multiple randomized trials have demonstrated the safety of phage to treat UTI (43, 44), but phage was found to be non-superior to placebo controls in a double-blind randomized control trial (43). While phage-mediated E. coli decolonization of the gut has been explored experimentally using phage (45, 46) or phage plus non-pathogenic E. coli competitors (47), therapeutic use of phage or phage products to control vaginal bacterial colonization has only been minimally explored and not in the context of E. coli (48–50).

Hypothesizing that phage can serve as a non-antibiotic alternative to limit UPEC vaginal colonization and intracellular reservoirs, we tested the impact of UPEC targeting phage and phage cocktails on UPEC adherence, invasion, and intracellular persistence in immortalized human urogenital epithelial cells. We further tested the effects of phage and phage cocktails in altering UPEC vaginal colonization and urogenital tissue dissemination in humanized microbiota (^HMbmice) (51, 52). We found that phage and phage cocktails reduced UPEC adherence and invasion in vaginal and bladder epithelial cells and reduced intracellular populations in vaginal cells. Despite promising in vitro activity, vaginal phage treatment displayed muted efficacy in vivo when administered following UPEC vaginal colonization.

RESULTS

Phages inhibit UPEC growth in simulated vaginal fluid

Although phage killing of E. coli, including UPEC strains, is widely characterized in bacteriologic media (53–55) and host fluids, including blood (56, 57), urine (53–55), synovial fluid (58), synthetic saliva (59), and cecal fluid (46), phage activity in the context of the vaginal environment has been minimally described, even though phages are readily isolated from vaginal specimens (60–62). For this study, we selected ΦHP3, a well-characterized phage with broad activity against UPEC strains and demonstrated efficacy in multiple infection models (46, 55, 63). We further tested a four-phage cocktail (ΦCocktail) containing ΦHP3, two additional UPEC-targeting phages, ΦE17 and ΦES19, which show superior anti-biofilm efficacy in an in vitro catheter-associated UTI model (55), and ΦHP3.1, a derivative of ΦHP3 that emerged in a directed evolution experiment (54). ΦHP3.1 contains two single nucleotide polymorphism mutations in spike and longtail fiber genes, which likely confer its retained infectivity against parental HP3 phage-resistant E. coli (54), and thus, it has the capacity to counter emerging resistance. This four-phage cocktail has been used to treat extended-spectrum beta-lactamase-producing E. coli UTI in a human case report (64). UPEC cystitis strain UTI89 was mixed with ΦHP3 or the ΦCocktail at a multiplicity of infection (MOI) of 10 and incubated for 12 h in LB broth or a simulated vaginal fluid (SVF) adapted from prior work (65). In LB broth, ΦHP3 and ΦCocktail significantly reduced UPEC growth as measured by optical density for the full 12 h of culture (Fig. 1A), with significantly reduced area under the curve (AUC) values for both phage-exposed conditions (Fig. 1B). We monitored simultaneous CFU and PFU levels over the first 6 h and observed a significant decrease in UTI89 CFU (>10-fold) in the presence of ΦHP3 and ΦCocktail with a concomitant >100-fold increase in phage PFU (Fig. 1C). In SVF, ΦHP3 significantly reduced UPEC growth for the first 7 h, while the ΦCocktail reduced growth for the first 9 h, with no differences in optical density at later time points (Fig. 1D). Even so, AUC values were significantly lower in both phage-exposed conditions (Fig. 1E). Similar to LB, monitoring of CFU and PFU levels in SVF revealed a significant reduction of UTI89 CFU in the presence of ΦHP3 and ΦCocktail (>10,000-fold and >1,000-fold, respectively) and a >100-fold increase in phage PFU over the first 6 h (Fig. 1F). Together, these data support phage replication and lytic activity in both bacteriologic medium and simulated vaginal conditions.

Fig 1.

Phages inhibit UPEC growth and replicate in vitro in bacteriologic medium and simulated vaginal fluid. Overnight UTI89 cultures were diluted in Luria-Bertani (LB) broth or simulated vaginal fluid and infected with ΦHP3 or ΦCocktail at an MOI of 10 or mock-treated as a control. Growth was monitored by measuring OD_600nm at 15-min intervals for 12 h or by measuring CFU and PFU levels for 6 h at 60–120-min intervals. Optical density (A) and area under the curve analyses (B) of UTI89 in LB medium. (C) CFU and PFU quantification in LB medium. Optical density (D) and area under the curve analyses (E) of UTI89 in SVF medium. (F) CFU and PFU quantification in SVF medium. Experiments were performed in technical duplicate across three to four independent replicates. Points represent medians of experimental replicates (A, C, D, and F) or medians of individual experimental replicates (B and E). Lines represent interquartile ranges (A, C, D, and F) or median with interquartile ranges (B and E). Data were analyzed by two-way ANOVA (A and D), one-way ANOVA with Holm-Šídák’s multiple comparisons test (B and E), or mixed-effects ANOVA (C and F), *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Phages decrease UPEC adherence to vaginal and bladder epithelial cells

Phage binding to eukaryotic cells and cellular uptake of phage vary across cell types and phage strains and may promote or inhibit phage efficacy in reducing bacterial pathogens (66–68). To assess whether phages interact with the vaginal epithelium, we performed phage adherence assays using the immortalized human vaginal epithelial cell line VK2/E6E7. ΦHP3 (10⁸ PFU) was incubated with VK2 cell monolayers for 1 h. A subset of monolayers was washed twice with PBS to retain only cell-associated phages (Adherent Φ), whereas unwashed wells retained phages in the supernatant as well as cell-associated phages (Total Φ). Plaque assays of supernatant and cell lysates, with or without washing, revealed that washing significantly reduced cell-free PFU by 170-fold, while cell-associated PFU remained similar between conditions, at an adherence rate of 0.2% of the inoculum (Fig. 2A). Time-lapse confocal imaging further confirmed interaction of ΦHP3, labeled with nucleic acid dye SYBR-gold (69), with VK2 cells (Movie S1). To test the impact of phage on UPEC adherence to the vaginal epithelium, phages were incubated with VK2 cells under the conditions described in Fig. 2A, and then 10⁶ CFU of UTI89 were added to the monolayers and allowed to adhere for 30 min. Both ΦHP3 and ΦCocktail pretreatment reduced UPEC adherence, 10-fold and 150-fold, respectively, in unwashed conditions (Total Φ), whereas only ΦCocktail reduced UPEC adherence 40-fold when non-adherent phages were removed prior to bacterial inoculation (Adherent Φ) (Fig. 2B). These results suggest that both eukaryotic cell-bound and unbound phages contribute to reduced bacterial adhesion, either by directly lysing bacteria or by interfering with their ability to attach to host cells. To test whether lytic activity was required for phage-mediated UPEC reduction, we performed adherence assays using a ΦHP3-resistant UTI89 derivative (Φ^R) with a spontaneous mutation in rfaH resulting in a truncated inner LPS core, the putative ΦHP3 receptor (53). No differences in UTI89 Φ^R adherence to VK2 cells were observed with ΦHP3 treatment in either washed (Adherent Φ) or unwashed (Total Φ) conditions (Fig. 2C). Confocal fluorescence microscopy further corroborated these quantitative findings. SYBR-gold labeled ΦHP3 could be visualized in close proximity with VK2 cells in the presence or absence of RFP-expressing UTI89, frequently co-localizing with bacteria in the latter condition (Fig. 2D).

Fig 2.

Phage pretreatment inhibits UPEC adherence to human vaginal epithelial cells. Human vaginal epithelial VK2 cell monolayers were pretreated with ΦHP3 and ΦCocktail (10⁸ PFU) for 1 h, and non-adherent phages were removed by washing (Adherent Φ) or unwashed to retain all phages (Total Φ). (A) ΦHP3 PFU recovered from filtered supernatant (cell-free) or lysed cells (cell-associated) for Adherent Φ and Total Φ conditions. Monolayers were infected with 10⁶ CFU of UPEC for 30 min, followed by cell lysis and plating to quantify adherent bacteria. (B) Adherence of phage-sensitive WT UTI89 to ΦHP3 and ΦCocktail pretreated cells under Adherent Φ and Total Φ conditions. (C) Adherence of phage-resistant UTI89 Φ^R to ΦHP3 pretreated cells under Adherent Φ and Total Φ conditions. For confocal microscopy, VK2 monolayers were infected with 10⁶ CFU UTI89-RFP (red) for 3 h, prior to treatment with 10⁸ PFU SYBR-Gold labeled ΦHP3 (green), with image acquisition immediately after phage addition. (D) Representative images with cell nuclei stained with Hoechst (blue) and cell membranes visualized with wheat germ agglutinin (magenta). Arrows indicate phage co-localization with host cells (green), UPEC co-localization with host cells (red), or phage co-localization with UPEC (yellow). Experiments were performed in technical duplicate across four to eight independent replicates (A–C) or twice independently (D). Points represent medians of experimental technical replicates, and lines represent medians with interquartile ranges (A–C). Data were analyzed by two-way ANOVA with uncorrected Fisher’s LSD (A), two-way ANOVA with Holm-Šídák’s multiple comparisons test (B), or Kruskal-Wallis with Dunn’s multiple comparisons test (C), **P < 0.01, ****P < 0.0001.

Additionally, we evaluated phage interactions with the immortalized human bladder carcinoma cell line HTB-9. Identical to VK2 cell assays, ΦHP3 binding was measured in unwashed (Total Φ) and washed (Adherent Φ) conditions. Plaque assays of supernatant and cell lysates, with or without washing, revealed that washing significantly reduced cell-free PFU by 250-fold, while cell-associated PFU remained similar between conditions, at an adherence rate of 0.5% of the inoculum (Fig. 3A). Time-lapse confocal imaging further confirmed binding of SYBR Gold-labeled phage particles to HTB-9 cells (Movie S2). Both ΦHP3 and ΦCocktail treatment reduced UTI89 adherence 6.4- and 6.2-fold, respectively, when cell-free phages were retained in the wells, whereas cell-associated only conditions failed to reduce UTI89 adherence to HTB-9 cells (Fig. 3B). As with VK2 cells, reduction in bacterial adherence was dependent on phage lytic activity, as no reduction in UTI89 Φ^R adherence to HTB-9 cells was observed in either washed or unwashed conditions (Fig. 3C). ΦHP3 could be visualized in close proximity to the HTB-9 cell membrane in the presence or absence of RFP-expressing UTI89, often co-localized with bacteria when present (Fig. 3D).

Fig 3.

Phage pretreatment inhibits UPEC adherence to human bladder carcinoma cells. Human bladder carcinoma HTB-9 cell monolayers were pretreated with ΦHP3 and ΦCocktail (10⁸ PFU) for 1 h, and non-adherent phages were removed by washing (Adherent Φ) or unwashed to retain all phages (Total Φ). (A) ΦHP3 PFU recovered from filtered supernatant (cell-free) or lysed cells (cell-associated) for Adherent Φ and Total Φ conditions. Monolayers were infected with 10⁶ CFU of UPEC for 30 min, followed by cell lysis and plating to quantify adherent bacteria. (B) Adherence of phage-sensitive WT UTI89 to ΦHP3 and ΦCocktail pretreated cells under Adherent Φ and Total Φ conditions. (C) Adherence of phage-resistant UTI89 Φ^R to ΦHP3 pretreated cells under Adherent Φ and Total Φ conditions. For confocal microscopy, HTB-9 monolayers were infected with 10⁶ CFU UTI89-RFP (red) for 3 h, prior to treatment with 10⁸ PFU SYBR-Gold labeled ΦHP3 (green), with image acquisition immediately after phage addition. (D) Representative images with cell nuclei stained with Hoechst (blue) and cell membranes visualized with wheat germ agglutinin (magenta). Arrows indicate phage co-localization with host cells (green), UPEC co-localization with host cells (red), or phage co-localization with UPEC (yellow). Experiments were performed in technical duplicate across four to nine independent replicates (A–C) or twice independently (D). Points represent medians of experimental technical replicates, and lines represent medians with interquartile ranges (A–C). Data were analyzed by two-way ANOVA with uncorrected Fisher’s LSD (A), two-way ANOVA with Holm-Šídák’s multiple comparisons test (B), or Kruskal-Wallis with Dunn’s multiple comparisons test (C), **P < 0.01, ****P < 0.0001.

Phages decrease UPEC invasion and intracellular survival in vaginal epithelial cells

UPEC can reside intracellularly within bladder and vaginal epithelial cells (36, 70), establishing a persistent bacterial reservoir that can serve as a nidus for urinary dissemination and recurrent infection. To determine whether phage treatment impacted UPEC invasion, we pretreated monolayers with 10⁸ PFU of ΦHP3 or ΦCocktail for 1 h, washed away non-adherent phages (Adherent Φ), or retained non-adherent phages (Total Φ). Monolayers were then infected with UTI89, and invaded bacteria were quantified using the gentamicin protection method (53). Under these conditions, intracellular bacteria could be visualized within the plasma membrane boundaries of VK2 and HTB-9 cells using confocal microscopy acquired Z-stack images after 3 h of incubation (Fig. S1). Both ΦHP3 and ΦCocktail treatment reduced UTI89 invasion of VK2 cells 5.2- and 7.5-fold, respectively, when cell-free phages were retained in the wells, whereas cell-associated only conditions failed to reduce UTI89 invasion (Fig. 4A). ΦHP3 did not alter UTI89 Φ^R invasion of VK2 cells (Fig. 4B). In HTB-9 cells, only ΦHP3 treatment significantly reduced UTI89 invasion (7.1-fold) and only when non-adherent phages were retained (Fig. 4C). As with VK2 cells, ΦHP3 did not alter UTI89 Φ^R invasion of HTB-9 cells (Fig. 4D).

Fig 4.

Phage modestly inhibits UPEC invasion and intracellular survival in vaginal and bladder cells. For invasion assays, human vaginal epithelial VK2 cell and bladder carcinoma HTB-9 cell monolayers were pretreated with ΦHP3 and ΦCocktail (10⁸ PFU) for 1 h, and non-adherent phages were removed by washing (Adherent Φ) or unwashed to retain all phages (Total Φ). Monolayers were infected with 10⁶ CFU of UPEC for 2 h, gentamicin treatment for 1 h to kill extracellular bacteria, followed by cell lysis and plating to quantify invasive bacteria. VK2 cell invasion of phage-sensitive WT UTI89 to ΦHP3 and ΦCocktail pretreated cells (A), or UTI89 Φ^R to ΦHP3 pretreated cells (B), under Adherent Φ and Total Φ conditions. HTB-9 cell invasion of phage-sensitive WT UTI89 to ΦHP3 and ΦCocktail pretreated cells (C), or UTI89 Φ^R to ΦHP3 pretreated cells (D), under Adherent Φ and Total Φ conditions. For intracellular survival assays, VK2 and HTB-9 monolayers were infected with 10⁶ CFU of UPEC for 2 h and gentamicin treatment for 1 h to kill extracellular bacteria. Monolayers were then treated with 10⁸ PFU phage and gentamicin for 24 h, followed by cell lysis and plating to quantify intracellular bacteria. VK2 intracellular UTI89 or UTI89 Φ^R after treatment with ΦHP3 (E), or UTI89 after treatment with ΦCocktail (F). HTB-9 intracellular UTI89 or UTI89 Φ^R after treatment with ΦHP3 (G), or UTI89 after treatment with ΦCocktail (H). (I) ΦHP3 adherence to VK2 and HTB-9 cells after 30 min of incubation with 10⁸ PFU, followed by washing and cell lysis. (J) Interleukin-8 release in VK2 cell supernatant after 6 h of infection with 10⁶ CFU UTI89 with or without 10⁸ PFU ΦHP3 as measured by ELISA. Experiments were performed in technical duplicate across 3–12 independent replicates. Points represent medians of experimental technical replicates, and lines represent medians with interquartile ranges (A–J). Data were analyzed by two-way ANOVA with Holm-Šídák’s multiple comparisons test (A, C, E, and G), Kruskal-Wallis with Dunn’s multiple comparisons test (B and D), paired t test (F and H), Mann-Whitney U test (I), or two-way ANOVA with uncorrected Fisher’s LSD (J), *P < 0.05, ***P < 0.001.

To test whether phage impacts established UPEC intracellular reservoirs, epithelial monolayers were infected with 10⁶ CFU for 2 h, followed by gentamicin treatment for 1 h, after which, 10⁸ PFU were added in fresh media containing gentamicin, and cells were incubated for an additional 24 h. Both ΦHP3 and ΦCocktail treatments modestly but significantly reduced intracellular bacteria by 2.3- and 2.6-fold, respectively (Fig. 4E and F). This reduction was dependent on phage lytic activity, as UTI89 Φ^R intracellular survival was not impacted by phage treatment (Fig. 4E). In contrast, neither ΦHP3 nor ΦCocktail treatment impacted UTI89 or UTI89 Φ^R intracellular survival in HTB-9 cells (Fig. 4G and H). No differences in total phage bound to host cells within 30 min were observed between cell lines (Fig. 4I). UPEC exposure stimulates vaginal epithelial inflammation, including production of chemokine interleukin-8 (IL-8) (71). To test whether phage treatment altered IL-8 production in VK2 cells, monolayers were treated with 10⁸ PFU ΦHP3 for 1 h, followed by infection with 10⁶ CFU of UTI89 for 6 h. IL-8 levels in cell supernatant were significantly increased with UTI89 exposure as measured by ELISA, but ΦHP3 pretreatment did not alter IL-8 production in infected cells or mock-infected controls (Fig. 4J).

Phage treatment reduces murine UPEC vaginal colonization

To test whether phage altered the ability of UPEC to colonize the vaginal mucosa in vivo, we used a murine UPEC vaginal colonization model in humanized microbiota mice (^HMbmice), generated by colonizing germ-free C57BL/6 mice with pooled human fecal microbes (52). We have previously demonstrated that ^HMbmice harbor a more human-like vaginal microbiota that is enriched with Lactobacillus spp. compared to conventional mice and are readily vaginally colonized by UTI89 Strep^R (51). As with prior studies (36, 51), mice were synchronized with β-estradiol (day −2), vaginally inoculated with UTI89 Strep^R (day −1), and were then vaginally administered 10⁸ PFU of phage daily for seven total treatments (Fig. 5A). Vaginal swabs were collected just prior to phage treatment, and urogenital tissues were collected 7 days after initial phage treatment. By day 6, phage treatment significantly promoted UPEC vaginal clearance below the limit of detection in the ΦHP3 group (27% clearance, 10/37 mice) compared to mock-treated controls (11%, 4/37 mice, Fig. 5B). When comparing vaginal UPEC burdens over time, the ΦHP3 group displayed lower UPEC burdens at day 4, with no differences detected at any other time point or in urogenital tissues at day 7 (Fig. 5C and D). We observed significant positive correlations between UPEC CFU across reproductive tract and kidney tissues in mock-treated mice, indicating dissemination to the upper reproductive tract and urinary tract tended to co-occur within the same mice (Fig. 5E). However, in ΦHP3-treated mice, although positive correlations in UPEC burdens across tissues were still observed, the strength of these associations was lower and non-significant between the kidneys and the vagina or cervix (Fig. 5F). While prior work has demonstrated robust vaginal UPEC colonization following transurethral inoculations (36, 72), we observed minimal UPEC colonization of the bladder following vaginal inoculation, similar to previous observations (70). These findings suggest that vaginal inoculation results in transient bladder exposure insufficient to establish colonization, perhaps due to the differences in inoculation dose required to establish persistent colonization of the vagina (10⁴ CFU) (73) compared to the bladder (≥10⁵ CFU) (74).

Fig 5.

ΦHP3 reduces vaginal UPEC colonization in humanized microbiota mice. (A) Experimental timeline for estrus synchronization, UPEC colonization, phage treatment, and sample collection in human microbiota mice (^HMbmice). (B) Percent UTI89 Strep^R vaginal colonization curves of ΦHP3 and mock-treatment mice over 7 days of treatment, with decolonization determined as the lack of detectable UPEC on vaginal swabs for all subsequent samples. (C) UTI89 Strep^R CFU recovered from vaginal swabs over 7 days of treatment. (D) UTI89 Strep^R CFU/g of urinary and reproductive tissues at day 7 post-phage treatment. Correlations of UTI89 Strep^R CFU across day 7 tissues in mock (E) and ΦHP3-treated mice (F). (G) Area under the curve of UTI89 Strep^R recovered from ΦHP3-treated mice vaginal swabs grown in LB for 24 h in the presence or absence of 10⁸ PFU ΦHP3. (H) PFU recovered from vaginal swabs on days 1 through 6 of treatment. (I) Correlations of PFU and UTI89 Strep^R CFU recovered from vaginal swabs on days 1 through 6 of treatment. (J) Vaginal swab CFU of UTI89 Φ^R Strep^R in the presence or absence of ΦHP3 treatment administered on days 0–2. Experiments were performed in at least two independent replicate experiments. Points represent individual mice (C, D, and H–J) or medians of experimental technical replicates (G), and lines represent medians with interquartile ranges. n = 37 (B–F), n = 17 (H and I), and n = 9 (J). Data were analyzed by Mantel-Cox test (B), Mann-Whitney U test with Benjamini, Krieger, and Yekutieli correction for false discovery with a false discovery rate set at 5% (C and J), two-way ANOVA with Holm-Šídák’s multiple comparisons test (D), Pearson correlation (E, F, and I), Mann-Whitney U test (G), or Friedman test with Dunn’s multiple comparisons test to day 1 values (H), *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

Since E. coli is detected in the ^HMbmice vaginal microbiota (51), we tested the susceptibility of endogenous vaginal E. coli isolates to ΦHP3 in vitro. All isolates tested were sensitive to the phage (Fig. 5G). Viable plaques were recovered from most vaginal swabs over the treatment period, with levels decreasing at later time points (Fig. 5H), indicating phages retain capacity for lytic activity in vivo. We performed plaque assays from vaginal swabs from 10 treatment-naive mice, and no plaques were observed, indicating quantified PFU reflected phage treatment and not endogenous phage. Across all time points, vaginal UPEC CFU and ΦHP3 PFU were positively correlated, suggesting that recovery of lytic phage required a viable UPEC host (Fig. 5I). Although E. coli can rapidly develop resistance to phage in vitro and in vivo (53, 54, 75, 76), all tested UPEC CFU recovered from vaginal swabs of phage-treated mice retained susceptibility to ΦHP3 and ΦCocktail (Fig. S2A). To test the impact of phage resistance on vaginal colonization, ^HMbmice were vaginally colonized with UTI89 Φ^R (day −1) in the presence or absence of ΦHP3 treatment administered on days 0, 1, and 2. UTI89 Φ^R poorly colonized the vaginal tract regardless of ΦHP3 treatment, with only one mouse having detectable vaginal CFU at day 2 (Fig. 5J). Together, these data support that daily ΦHP3 treatment modestly reduces UTI89 vaginal colonization, driving clearance in a subset of mice over 7 days. Furthermore, persistent colonization in some animals is not due to the development of phage resistance.

We further tested whether the four-phage cocktail altered UPEC vaginal colonization. Mice were inoculated with UTI89 and treated daily with phage as in Fig. 5A. Although 28% (8/29) of ΦCocktail-treated mice cleared vaginal UPEC by day 6, this was not statistically different from mock-treated mice with 8% (2/25) UPEC clearance (Fig. 6A). No differences were observed in vaginal UPEC CFU from days 1–6 of treatment; however, day 7 vaginal and cervical burdens were significantly lower in ΦCocktail-treated mice (Fig. 6B and C). Similar to ΦHP3 kinetics, ΦCocktail vaginal PFUs were reduced at later time points compared to day 1, and UPEC CFU and phage PFU were positively correlated (Fig. 6D and E). Because phage treatment showed only modest reductions in UPEC burden in vivo, despite potent in vitro activity, we tested several strategies to improve in vivo efficacy. Inoculation with a 100-fold lower UPEC inoculum failed to further separate UPEC CFU colonization and dissemination over the 7-day treatment (Fig. S2B and C). Additionally, administering phage intravaginally just 30 min post-infection with UPEC did not further separate bacterial burdens between phage-treated groups and mock controls. In fact, this treatment regimen resulted in higher day 5 vaginal CFU and day 7 cervical burdens in ΦCocktail-treated mice compared to mock controls (Fig. 6F and G). Given the established impact of the estrous stage on UPEC vaginal colonization (70), we further evaluated whether the estrous cycle modified the capacity for phage to persist in the vaginal environment. After estrous staging via vaginal lavage fluid (Fig. S2D) (77), ^HMbmice were inoculated with a single dose of 10⁸ PFU of ΦHP3 or ΦCocktail in the absence of UTI89 to limit confounding effects of phage replication within the bacterial host. Phage persistence was measured by collecting vaginal swabs at 4 and 24 h post-inoculation. Mice that were inoculated with ΦHP3 during estrogen-dominant stages (proestrus and estrus) had significantly higher phage titers recovered at 4 h compared to mice that were in progesterone-dominant stages (metestrus and diestrus) at the time of inoculation (Fig. S2E). Similar patterns were observed with ΦCocktail-treated mice, although no significant differences were detected (Fig. S2E). In contrast to Fig. 5H and 6D, no PFUs were detected in any samples at 24 h post-inoculation, possibly due to the absence of bacterial hosts. These results suggest that the estrous stage may influence phage-bacteria and phage-epithelial dynamics at the vaginal mucosa.

Fig 6.

ΦCocktail modestly affects vaginal UPEC colonization in humanized microbiota mice. ^HMbmice were vaginally inoculated with UPEC and treated daily with phage as described in Fig. 5A. (A) Percent UTI89 Strep^R vaginal colonization curves of ΦCocktail and mock-treated mice over 7 days of treatment, with decolonization determined as the lack of detectable UPEC on vaginal swabs for all subsequent samples. (B) UTI89 Strep^R CFU recovered from vaginal swabs over 7 days of treatment. (C) UTI89 Strep^R CFU/g of urinary and reproductive tissues at day 7 post-phage treatment. (D) PFU recovered from vaginal swabs on days 1 through 6 of treatment. (E) Correlations of PFU and UTI89 Strep^R CFU recovered from vaginal swabs on days 1 through 6 of treatment. In a subset of mice, ΦHP3 or ΦCocktail was administered 30 min after UPEC inoculation and daily thereafter for an additional 6 days. (F) UTI89 Strep^R CFU recovered from vaginal swabs over 7 days of treatment. (G) UTI89 Strep^R CFU/g of urinary and reproductive tissues at day 7 post-phage treatment. Experiments were performed in at least two independent replicate experiments. Points represent individual mice, and lines represent medians with interquartile ranges (B–G). n = 25–28 (A–C), n = 28 (D and E), and n = 7–10 (F and G). Data were analyzed by Mantel-Cox test (A), Mann-Whitney U test with Benjamini, Krieger, and Yekutieli correction for false discovery with a false discovery rate set at 5% (B), two-way ANOVA with Holm-Šídák’s multiple comparisons test (C), Friedman test with Dunn’s multiple comparisons test to day 1 values (D), Pearson correlation (E), or mixed-effects model with Dunnett’s multiple comparisons test (F and G), *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.

DISCUSSION

While clinical evidence highlights the importance of UPEC vaginal colonization as a predictor of UTI risk (20, 22–24), with genetically identical vaginal and UTI isolates reported in some cases (78), there are limited options to control UPEC vaginal colonization. In this study, we explored the use of lytic bacteriophage as a UPEC urogenital decolonization strategy. We found that phage readily bound to human vaginal and bladder epithelial cells, reduced UPEC adherence in both cell types, and limited UPEC intracellular survival in vaginal cells in vitro. In a murine UPEC vaginal colonization model, although a lytic phage was recovered and the development of phage resistance was not observed, phage treatment only modestly reduced UPEC vaginal colonization, suggesting both a potential clinical benefit and a need for further optimization.

In human vaginal samples, phages represent the largest portion (83%) of total viral sequences, are detected in up to 90% of human vaginal samples, and are predominantly predicted to be lysogenic (temperate) phages that incorporate into the bacterial genome rather than causing cell lysis (39, 40, 79). Endogenous phages predicted to infect E. coli are detected in about 10% of women (79) and are enriched in women with bacterial vaginosis (40), suggesting these lysogenic phages may confer a fitness advantage through enhanced virulence, stress response, or antibiotic resistance. While the therapeutic application of lytic phage in the vaginal tract is still in early stages, multiple studies have evaluated the potential of phage-derived endolysins to reduce vaginal pathobionts, including group B Streptococcus (50) and Gardnerella spp. (80). Even so, there is substantial experimental work testing phage-mediated decolonization of other tissues (81) and across a range of model systems, including invertebrate animals (82). A recent systematic review revealed that 80% of experimental studies report phage-mediated reduction of the target organism at sites, including the gastrointestinal tract, skin, lung, and urinary tract (81).

Building upon prior work, we selected ΦHP3, a Myoviridae family phage with broad efficacy against UPEC strains and established efficacy in human blood and urine (46, 53, 55, 63), to assess UPEC decolonization of the urogenital epithelium. Our four-phage cocktail included ΦHP3, as well as ΦHP3.1, to counter emerging bacterial resistors, and two phages, E17 and ES19, with superior anti-biofilm activity (54, 55). In contrast with other studies (75, 83), we did not consistently identify increased activity of the phage cocktail compared to single phage treatment either in vitro or in vivo. This could be explained by overlapping bacterial targets, believed to be primarily LPS for both HP3 and ES17 (53, 54), or that the phage cocktail activity was primarily driven by ΦHP3, and that bacterial resistance and biofilm production were not key persistence mechanisms in our models.

Similar to prior work with an Autographiviridae phage (84), we observed phage-mediated reduction of UPEC interactions with HTB-9 cells and extended these studies to vaginal epithelial (VK2) cells. In both cell lines, reduction of UPEC was dependent on bacterial susceptibility to phage lysis since no effects of phage were observed using phage-resistant UTI89. Both ΦHP3 and ΦCocktail were somewhat more active on vaginal-derived cells compared to bladder-derived cells across adherence, invasion, and intracellular survival assays, even though lytic phage was recovered at similar levels from both lines, suggesting discordant phenotypes may be due to differences in extra- and intracellular localization of phage and bacteria between cell lines or impacts of cell culture media on phage activity. Similar to our findings, in a human bladder organoid model of non-transformed uroepithelial cells, ΦHP3 significantly reduced UTI89 adherence but did not impact established intracellular reservoirs (85). While eukaryotic cell uptake of phage may limit the therapeutic efficacy of phage against extracellular bacteria (67), this uptake may allow phage to access intracellular bacterial populations. Phage reduction of intracellular bacterial reservoirs has been reported in murine macrophages (86) and human T24 bladder epithelial cells (87), and in this study, we report phage-mediated reduction of intracellular UPEC in vaginal epithelial cells, but not HTB-9 bladder cells. Although our data suggest that phage gains access to intracellular UPEC, we were unable to visualize intracellular phage using SYBR gold due to the loss of labeling within a short time frame. This could be due to loss of labeling during phage replication or dye destabilization due to changes in intracellular pH. Generation of phage-specific antibodies could provide more definitive evidence of surface versus intracellular localization in future work. Another limitation is that we did not identify the mechanism of phage entry in vaginal or bladder epithelial cells, although other phages have been described to undergo micropinocytosis and clathrin- or caveolae-mediated endocytosis (88). Furthermore, we did not determine the eventual fate of the phage or phage DNA, although previous studies have shown that internalized phages are trafficked through the endolysosome and degraded (89). Furthermore, unlike prior studies with UTI89 and uroepithelial cells (90), we observed no differences in vaginal epithelial cell production of IL-8 in the presence of phage, suggesting that augmentation of cellular immune responses did not contribute to our phenotypes.

Despite potent activity in vitro, phages have demonstrated mixed effects in animal models of pathogen colonization. Similar to our current study focused on UPEC urogenital colonization, prior studies have demonstrated that phages have limited efficacy in reducing exogenous or endogenous gut E. coli, even when active phages were recovered from the mucosal site (47, 91, 92). There are several plausible reasons for this. First, it is possible that the relative proportion, or MOI, of phage to bacterial host must be optimized to reduce bacterial colonization. Two phages used in our study, HP3 and ES17 (included in the cocktail), were effective in reducing extra-intestinal pathogenic E. coli murine gut colonization at a 10¹⁰ PFU dose, but not at a 10⁹ PFU dose (46). In contrast, a three-phage cocktail reduced UPEC gut colonization at both 10⁵ and 10⁷ doses (45), highlighting the range in activity of different phage and bacteria pairings. A second possibility is that the in vivo environment directly or indirectly impacts phage efficacy, with phage showing efficacy in some tissues but not others (93). Phage binding to host mucins may impact in vivo efficacy by bringing them in closer proximity to the bacterial host, as seen in the case of ΦES17 in the ΦCocktail (46, 94) or by partially limiting the ability to bind bacterial hosts, as is the case for ΦHP3 (46). However, the similar activity between ΦHP3 and ΦCocktail suggests that mucin binding is not a strong driver of phenotypes in our models. Moreover, the in vivo environment may provide the bacterial host a protected niche or low metabolic state that limits phage activity and replication. Similar to a study with E. coli mono-colonized mice (76), we observed positive correlations between recovered CFU and PFU with recovered bacteria post-phage treatment retaining phage susceptibility. Although we observed that endogenous murine E. coli vaginal isolates, a common member of the ^HMbmice vaginal microbiota (51), were susceptible to the phage used in our experiments, we did not track endogenous E. coli levels before or after phage treatment. Thus, it is possible that endogenous E. coli was an additional variable influencing phage treatment success. Finally, estrous cycling may further impact phage-UPEC and phage-epithelial interactions. Synchronization into estrus via estradiol promotes murine vaginal colonization of multiple pathogens, including UPEC (70, 73, 95); however, the impact of estrous stage on phage interactions with host cells is less well-characterized. In this study, we found that phage persisted at higher titers when mice were inoculated in estrogen-dominant (proestrus or estrus) stages compared to progesterone-dominant stages. Thus, it is possible that phages bind cornified and non-cornified epithelial cells, which are more abundant in the vaginal lumen when estrogen levels are high, or conversely, phages are removed or inhibited by infiltrating neutrophils during progesterone-dominant stages (77). If so, these dynamics may explain why phage activity improved at 4 days post-synchronization if synchronized animals were returning to proestrus/estrus stages of the next cycle. As additional limitations, we did not monitor the estrous stage mice during UPEC colonization and phage treatment, nor did we visualize the location of UPEC and phage in vivo. While all these possibilities may be at play in our model, neither reducing the UPEC inoculum nor altering the timing of phage treatment further augmented phage efficacy in reducing UPEC vaginal levels, suggesting further optimization is needed.

Although beyond the scope of this work, it is possible that phage treatment may synergize with bacteria that outcompete UPEC in the urogenital environment. Combining phage with non-pathogenic E. coli reduces pathogenic E. coli colonization of the gastrointestinal tract in vitro (96) and in vivo (47); however, since E. coli is a lower abundance vaginal organism associated with a non-optimal vaginal microbiota (21), this is likely not a viable strategy for UPEC vaginal decolonization. Alternatively, inclusion of a vaginal Lactobacillus spp., associated with an optimal vaginal microbiota (97), that also displays anti-UPEC activity may serve as a promising approach to control UPEC vaginal colonization.

In summary, our study shows that phage can limit UPEC urogenital colonization across multiple experimental models; however, the biological factors dictating phage activity in this environment remain to be identified. Phage effectiveness required bacterial susceptibility to lysis. While spontaneous development of phage resistance was not readily detected in our models, experimentally acquired phage-resistant strains displayed reduced colonization fitness in vivo. These findings deepen our knowledge of phage activity toward UPEC in the urogenital environment and support their continued development as a strategy to prevent UTIs.

MATERIALS AND METHODS

Bacterial strains and mammalian cell lines

Uropathogenic E. coli strain UTI89 (98), an HP3 phage-resistant UTI89 derivative (UTI89-2) (53), here designated Φ^R, a spontaneous streptomycin-resistant mutant of UTI89 (UTI89 Strep^R) (51), UTI89 Φ^R Strep^R (this study using the same method), and UTI89-RFP Kan^R were used in this study (99). Strains were grown in Luria-Bertani (LB) broth overnight at 37°C with shaking and supplemented with streptomycin (1,000 µg/mL) or kanamycin (50 µg/mL) where applicable. E. coli DH5α (ATCC BAA-3219) was grown in LB broth to the stationary phase and used as the bacterial host for phage plaque assays. Bacterial stocks were maintained at −80°C in 20% glycerol. Immortalized human vaginal epithelial cells (VK2/E6E7, ATCC CRL-2616) were grown in keratinocyte serum-free medium (KSFM) containing 50 mg/mL bovine pituitary extract and 0.1 ng/mL human recombinant epithelial growth factor. Human bladder epithelium carcinoma cells (5637, ATCC HTB-9) were grown in RPMI-1640 (Corning) containing 10% heat-inactivated fetal bovine serum. The cells were incubated at 37°C in 5% CO₂ and 100% humidity and passaged every 3–5 days until achieving 80%–90% confluence prior to seeding into 24-well tissue culture-treated plates.

Bacteriophage strains, titering, and cocktail preparation

Bacteriophage HP3 (accession no. KY608967) (63), ES17 (accession no. MN508615), and ES19 (accession no. MN508616) were isolated from environmental sources (HP3) and wastewater (ES17 and ES19). HP3.1 is a derivative of HP3 (54). Individual phage stocks were prepared and purified by cesium chloride gradient centrifugation as described previously (63). Endotrap HD Columns (LIONEX GmbH) were used to deplete endotoxin per the manufacturer’s instructions. Phage titers were determined by culture-based, double agar overlay plaque assays. Briefly, LB agar plates were overlaid with molten soft agar (LB containing 0.5% agarose), cooled to 4°C, and inoculated with 10⁷ CFU of E. coli DH5α as prey. After solidification, serially diluted phage lysates were spotted on top agar, incubated overnight at 37°C, and plaques were quantified to calculate phage titers in PFU/mL. Phage preparations were stored in phage buffer (54) at 4°C until use.

UPEC growth and CFU and PFU quantification

Overnight UPEC cultures were pelleted, resuspended in fresh LB or SVF medium at a 1:10 dilution, and transferred to 96-well microtiter plates. SVF, a mimetic of human vaginal fluid, was prepared as described previously (65), with the exception that the final pH (5.0) was not further reduced to 4.5 with lactic acid. For the phage challenge, 90 µL of adjusted bacterial inoculum in LB was mixed with 10 µL of ΦHP3 or ΦCocktail (10⁹ PFU/mL) in 96-well plates, performed in technical duplicate. Control wells were treated with LB or SVF media alone. Plates were incubated at 37°C with orbital shaking in a Tecan Infinite 200 plate reader, and growth was monitored by measuring OD_600nm at 15-min intervals for 12 h. Area under the curve was calculated using GraphPad Prism version 10.5.0. To evaluate the bactericidal activity of ΦHP3 or ΦCocktail in LB and SVF medium, overnight UPEC cultures were diluted and infected with ΦHP3 or ΦCocktail as in growth assays, or phage buffer as a control, and incubated for 6 h under shaking conditions. Samples were collected at the indicated time points, and bacterial viability was determined by enumerating CFUs on LB agar plates for both mock-treated and phage-treated groups. PFUs were determined via plaque assays as described below.

Phage plaque assays

To determine phage titers, 100 µL of culture, cell lysates, or murine samples was centrifuged to pellet bacterial cells and debris. Subsequently, 20 µL of the cell-free supernatant was mixed with an equal volume of chloroform. Phage plaques were determined using culture-based double agar overlay. Briefly, supernatants containing phage were filtered (0.2 μm) and stored at 4°C until use. LB plates were overlaid with 3.5 mL of molten soft agarose (cooled to 45°C) containing 100 µL of stationary phase E. coli DH5α cultures. After solidification, samples were serially diluted and spotted (5 µL) on the “top agar” surface. Plates were incubated overnight at 37°C, and plaques were counted to calculate phage titers as PFU/mL.

Adherence and invasion assays

UPEC adhesion and invasion assays were performed as previously described (53) with some minor modifications. Confluent monolayers grown in 24-well tissue culture plates were replenished with fresh KSFM or RPMI 1640 + 10% FBS medium for VK2 and HTB-9 cells, respectively. For adherence assays, cells were pretreated with ՓHP3 or ΦCocktail (10⁸ PFU/well) for 1 h. Following phage treatment, selected wells were washed with PBS twice to retain only cell-adherent phage, while others remained unwashed to retain the total phage population within the well. Bacteria (UTI89 or UTI89 Φ^R) were added at 10⁶ CFU/well, centrifuged for 2 min at 200 × g, and incubated for 30 min at 37°C in 5% CO₂. After 30 min of incubation, media were removed, and wells were washed at least three times with PBS to remove non-adherent extracellular bacteria. Cells were lifted by incubation with 100 µL of 0.025% Trypsin-EDTA for 8 min at 37°C. Thereafter, 400 µL of 0.025% Triton X-100 was added to each well. Cells were lysed by vigorous pipetting (30×), and 10-fold serial dilutions of cell lysates were plated on LB agar to quantify CFU/well and percent adherence of the inoculum. For invasion assays, UPEC strains were incubated with host cell monolayers for 2 h, cells were washed with PBS three times, and then incubated with medium containing gentamicin (100 µg/mL) for 1 h to kill extracellular bacteria. Host cells were then trypsinized and permeabilized, serially diluted, and plated as described for adhesion assays to quantify invaded bacteria.

Intracellular survival assays

As with adherence and invasion assays, VK2 and HTB-9 cell monolayers were seeded into 24-well tissue culture plates and replaced with fresh medium. Cells were infected with 10⁶ CFU/well UTI89 or UTI89 Φ^R (MOI of 10), plates were centrifuged at 200 × g for 2 min, and incubated at 37°C with 5% CO₂ for UPEC internalization. After 2 h of incubation, cells were washed with 1× PBS, and gentamicin was added at a concentration of 100 μg/mL for 1 h to kill extracellular bacteria. Infected cells were washed with PBS three times and treated with 100 μL of 10⁸ PFU/well of ΦHP3 or ΦCocktail in fresh medium containing gentamicin (10 μg/mL) and incubated at 37°C. After 24 h of incubation with phage, cells were trypsinized and permeabilized, serially diluted, and plated as described above. Intracellular CFUs were quantified the following day and expressed as CFU/well.

Cytokine assays

Confluent monolayers of VK2 cells were pretreated with ΦHP3 (10⁸ PFU/well) and incubated for 1 h. Cells were challenged with 10⁶ CFU UTI89 (MOI of 10) and further incubated for 6 h at 37°C with 5% CO₂. After incubation, cell supernatants were collected and stored at –20°C until analysis. Interleukin-8 in undiluted VK2 supernatant was quantified via ELISA per the manufacturers’ instructions (R&D Systems, DY208-05).

Fluorescence confocal microscopy

For confocal microscopy, VK2 and HTB-9 cells were seeded (80,000 cells/well) in 24-well plates (Cellvis, P-1.5H-N). The following day, cells were washed with PBS and replaced with 400 μL of fresh RPMI 1640 medium without phenol red. Monolayers were infected with 10⁶ CFU of UTI89-RFP and incubated for 3 h at 37°C with 5% CO₂. Unadhered bacteria were washed with PBS and replaced with fresh RPMI 1640 media without phenol red. Purified phages were labeled with SYBR-Gold nucleic acid gel stain (2.5× concentration, Invitrogen, S11494) for 1 h in the dark at 4°C, followed by three washes with phage buffer on an Amicon-Ultra centrifugal unit 100-kDa membrane to remove excess stain. Washed phages were resuspended in a final volume of 1 mL in phage buffer at 10⁹ PFU/mL. Labeled phages (10⁹ PFU/mL in 100 μL) were added to the preincubated wells containing cells and bacteria for 1 h. Following incubation, these wells were fixed with 4% PFA for 15 min and washed with PBS. Cells were stained with Wheat Germ Agglutinin Alexa Fluor 647 Conjugate (5 µg/mL) and Hoechst 33342 nucleic acid stain (1 µg/mL) and incubated for 20 min. Cells were washed with PBS and replaced with fresh RPMI 1640 media. Both 2D and 3D images were captured using a Nikon Ti2 ECLIPSE confocal microscope with a 60× oil-immersion objective. Z-stacks were acquired at a step size interval of 0.3 µm along the Z-axis. Time-lapse imaging was performed on VK2 and HTB-9 cells after adding labeled phage (10⁹ PFU/mL in 100 μL). The image acquisition was initiated immediately after phage addition with time lapse settings of 30-s intervals for a total of 10 min using a Nikon Ti2 ECLIPSE confocal microscope with a 60× oil-immersion objective. All the images were processed in Fiji, and representative sequential frames (Substack) were extracted in ImageJ to visualize phage, UPEC, and host cell interactions.

Animals

Mice were given food and water ad libitum. Humanized microbiota mice (^HMbmice) were bred and maintained as described previously (52). Mice were randomly distributed so that each treatment group contained a similar age range (6–8 weeks). Mice were acclimatized for 1 week in the biohazard room prior to experiments.

Murine UPEC vaginal colonization

Vaginal colonization studies were adapted from our previous studies (51, 100). Briefly, mice were synchronized with 0.5 mg β-estradiol administered intraperitoneally 24 h prior to vaginal inoculation of 10⁷ CFU of UTI89 Strep^R or UTI89 Φ^R Strep^R in a 10 μL volume. Beginning 24 h post-infection, mice received daily doses of purified UPEC-targeted ΦHP3 or ΦCocktail (10⁸ PFU in 10 μL), or phage buffer only as a mock control, for seven consecutive days. Vaginal swabs were collected daily just prior to phage treatment as described previously (51), and swabs were suspended in 100 μL of PBS. To quantify CFU, the swab sample (10 μL) was serially diluted and plated on LB agar containing streptomycin (1,000 μg/mL). To quantify PFU, 20 µL of the remaining swab sample was mixed with an equal volume of chloroform, incubated for 15 min to kill bacteria, serially diluted in phage buffer, and spotted onto top agar plates (5 µL). After 1 week of phage treatment, urinary and reproductive organs of mice were harvested and homogenized as described previously (101) and plated on LB agar with streptomycin to assess UPEC dissemination. Several modifications were made to the protocol to test the impact on phage efficacy. In a subset of experiments, mice received 10⁸ PFU of phage, administered intravaginally 30 min after UPEC infection (Fig. 6F and G). Additionally, mice were vaginally inoculated with a lower dose of UTI89 Strep^R (10⁵ CFU) followed by administration of 10⁸ PFU of phage 24 h post-infection. To assess whether phage recovery in ^HMbmice is influenced by the estrous stage, estrous staging was performed on vaginal lavage (50 µL) adapted from a previously described protocol (95), using brightfield microscopy at 100× magnification on an ECHO Revolve microscope. Mice were then intravaginally administered 10 µL containing 10⁸ PFU of ΦHP3 or ΦCocktail. Vaginal swabs were collected at 4 and 24 h post-inoculation to quantify recoverable phage (PFU) using double agar overlay plaque assays as described above.

Statistics

In vitro experiments were performed at least three times independently with at least two technical duplicates. Mean values of independent experiments were used to represent experimental replicates for statistical analyses. In vivo experiments were conducted at least twice independently, with individual mice serving as biological replicates. Experimental data were combined prior to statistical analyses. Two-way ANOVAs were used to compare bacterial growth curves, UTI89 adherence and invasion, intracellular survival (ΦHP3 treatment), VK2 IL-8 secretion, and murine tissue CFUs and PFUs with multiple comparisons tests, including uncorrected Fisher’s LSD or Holm-Šídák’s test as indicated in figure legends. A mixed-effects model with Dunnett’s multiple comparisons test was used for in vitro CFU and PFU recovery and murine tissues with three group comparisons. One-way ANOVA with Holm-Šídák’s multiple comparisons test or Mann-Whitney U tests were used to compare areas under the curves, as indicated in the figure legends. Kruskal-Wallis tests with Dunn’s multiple comparisons test were used to compare UTI89 Φ^R adherence and invasion. Paired t tests were used to compare intracellular survival (ΦCocktail treatment). Mann-Whitney U tests were used to compare adherent PFU and vaginal swab CFUs with the addition of Benjamini, Krieger, and Yekutieli correction for false discovery with a false discovery rate set at 5%. Mantel-Cox (log-rank) tests were used to compare percent colonization over time. Pearson correlations were used to correlate CFU across tissues and CFU and PFU within the same tissue. Friedman test with Dunn’s multiple comparisons test was used to compare vaginal PFU to day 1 values to subsequent time points. Statistical analyses were performed using Prism, version 10.5.0 (GraphPad Software Inc., La Jolla, CA, USA). P values of <0.05 were considered statistically significant.

ETHICS APPROVAL

Animal experiments were approved by the Baylor College of Medicine (BCM) Institutional Animal Care and Use Committee (protocol AN-8233) and were performed under accepted veterinary standards.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/iai.00543-25.

Supplemental material. iai.00543-25-s0001.pdf.

Fig. S1 and S2; Supplemental movie captions.

Movie S1. iai.00543-25-s0002.avi.

ΦHP3 phage adherence to human vaginal epithelial VK2 cells, related to Fig. 2.

Movie S2. iai.00543-25-s0003.avi.

ΦHP3 phage adherence to human bladder carcinoma HTB-9 cells, related to Fig. 3.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.