Phage Infection in Vaginal Lactobacilli: An In Vitro Study

Objective: During bacterial vaginosis, an unexplained decrease of vaginal lactobacilli occurs. To identify whether these lactobacilli could be infected by phages, we isolated phages from vaginal lactobacilli and analyzed their potential virulence in attacking vaginal lactobacilli in vitro. Methods: Vaginal samples were obtained from 39 reproductive-aged women. The selective Rogosa SL agar was used to isolate lactobacilli, from which phages induced by mitomycin C or released spontaneoulsy were analyzed by the agar spot method. Results: Of 20 samples from women with vaginal infections, 12 did not have lactobacilli. From the remaining 8 infection samples and the 19 samples from healthy women, 37 Lactobacillus strains were isolated, from which 7 temperate phages were identified. Upon analysis, all 7 phages infected vaginal lactobacilli from the same and/or different women in vitro. Two phages, Φkc005 and Φkc007, had a broad host range, infecting 7 of 8 species tested. A control intestinal Lactobacillus phage also lysed several vaginal strains. One vaginal phage, Φkc039, was apparently lytic against vaginal lactobacilli from 7 other women. This phage was characterized as follows: plaque morphology, small and clear; burst size, 300 phages per cell; spontaneous induction rate, 1 per 106 cells; DNA, double-stranded and linear, 41 kb; and shape, a hexogonal head and a non-contractile tail. Conclusions: Bacteriophages were isolated from vaginal lactobacilli of some women and were shown in vitro to lyse vaginal Lactobacillus strains from the same and/or different women. It was suggested that vaginal lactobacilli might be suppressed by phages.

hominis. 6 Several possible mechanisms by which vaginal lactobacilli decrease have been proposed. These include douching, 7 the use of spermicide such as Nonoxynol-9, 8 and treatment with antibiotics for other infections. However, it is unknown whether vaginal lactobacilli can decrease by natural causes.
Lactobacilli produce lactic acid and hydrogen peroxide (HzOz), which can prevent the overgrowth of other microorganisms in the vagina, z--4 including Escherichia coli, 9,1 G. vaginalis, 11 Chlamydia trachomatis, lz and Neisseria gonorrhoeae. 13 However, during BV, the number of HzOzproducing lactobacilli decreases drastically, while many anaerobic bacteria, such as G. vaginalis, M. hominis, Ureaplasma urealyticum, and Mobiluncus species, overgrow, z-4 It is currently unknown whether vaginal lactobacilli decrease first, allowing the BV-associated anaerobes to overgrow, or the anaerobes overgrow first, suppressing vaginal lactobacilli. Because these anaerobes are sensitive to lactic acid and HzOz produced by vaginal lactobacilli, theoretically they should not outgrow lactobacilli. Therefore, it is possible that lactobacilli decrease first in the vagina before the anaerobes overgrow. If so, other yet unidentified factors might inhibit vaginal lactobacilli.
A natural inhibitor of bacteria is bacteriophage or phage, the virus that infects bacteria. 14 Phages can exist as free viruses in the environment or as parasites in the host bacteria. A bacterium that hosts a latent phage (prophage) is called lysogen, which can release a small number of phages spontaneously or a large number of phages upon induction. Phages can be lyric (virulent) or temperate. The lyric phages can lyse bacteria completely, releasing many new phages into the environment. The temperate phages do not lyse all the infected bacterial cells; they convert some of them into lysogens and coexist with them. However, some temperate phages may become lyric due to mutations of the phage or changes of the bacterial host background.4, s Lactobadllus phages have been isolated from various sources, including dairy products, 6 sausage, 7 human intestines, 18 and sewage. 9 To date, studies on Lactobacillus phages have mainly focused on the food fermentation industry that uses lactobacilli as starter cultures. 6-These foods include yogurts and fermented meats, such as sausages. A study of the production of salami dry sausage showed that phage infection can delay acid production by the lactobadllus starter culture and drastically reduce the number of lactobacilli in the sausage and disrupt the ripening process of the sausage. 7 Likewise, phages may also infect vaginal lactobacilli, delay their acid production, and reduce their numbers. However, whether phages can infect human vaginal lactobacilli has not been reported. Therefore, the purposes of this study were to isolate phages from vaginal lactobacilli, to determine their infectivity and virulence by crossinfecting these phages against a collection of vaginal Lactobacillus strains, and to characterize one of the vaginal Lactobacillus phages.

Isolation of Vaginal Lactobacilli
The clinical research was approved by the Truman Medical Center Institutional Review Board in Kansas City, MO. Vaginal samples were obtained from 39 reproductive-aged, non-pregnant women. All subjects gave written informed consent. First, a sample was obtained to diagnose BV. As previously described, s the diagnosis of BV was based upon the presence of a vaginal discharge with three of the four following characteristics: a homogenous appearance; a pH > 4.5; the presence of a fishy amine odor upon the addition of 10% potassium hydroxide (KOH); and the presence of clue cells. The vaginal discharge of healthy women had no more than one of these four characteristics, and none contained clue cells. Then, a vaginal sample was taken for isolating lactobacilli. A B BL Culturette Collection and Transport System (Becton Dickinson Microbiological Systems, Cockeysville, MD) was used. The Culturette tube contains a rayon swab attached to a cap by a plastic stick and an ampule with 0.5 ml of modified Stuart's transport medium. A swab was inserted into the vagina, rotated a few turns along the vaginal sidewall, and allowed to absorb for a few seconds before withdrawing. The swab was then placed back into the tube containing the transport medium and sent to the laboratory for analysis. The samples were inoculated onto the selective Rogosa SL agar plates (

Phage Induction
Mitomycin C (Sigma, St. Louis, MO) was used to induce phages from vaginal lactobacilli as previously described. 1(' The induction of Lactobacillus prophages was indicated by the lysis of a Lactobacillus culture 4-7 h after the addition of mitomycin C. These lysates were then centrifuged, filtered through a 0.2 pm pore-size filter, and maintained at 4C with a drop of chloroform.

Phage Infectivity Assay
Phage infectivity was determined by the agar spot method as previously described. 16 All positive resuits were verified by single plaque formation, phage DNA isolation, and observation of phages under an electron microscope to rule out possible bacterial inhibitory effects due to bacteriocin, HeOe, or organic acids.

Phage Titration
The temperate (non-virulent or non-lytic) phage, kc039, from a vaginal Lactobadllus isolate (L. delbrueckii subsp, delbruecbii KC039), was found to be lytic (virulent) against another vaginal isolate, L. delbruecbii subsp, delbruecbii KC013. Thus, KC013 was used as an indicator strain to analyze the titer of bkc039. Plaques were counted after 24 h of incubation at 37C on MRS agar with 10 mM CaC1 z.

One-Step Growth Curve
KC013 was grown at 37C in MRS broth until midexponential phase. To ml of this culture, 0.1 ml of bkc039 stock [10 s plaque forming unit (PFU)/ ml; multiplicity of infection, 1:10] was added and mixed. The mixture of cells was added into 4 ml of melted soft 0.6% MRS agar with 10 mM CaC1 e, rapidly mixed, poured onto MRS plates, and incubated at 37C. Plaques were numerated after 24 h of incubation.

Spontaneous Phage Induction
KC039 was grown in 2 ml of MRS broth to midexponential phase. One milliliter of the culture was diluted and plated on MRS agar plates for cell count. Another ml was centrifuged to harvest the supernatant, which was filtered through a sterile 0.2 pm pore-size filter. The supernatant was diluted and used to infect the indicator strain KC013 by the soft agar overlay method as described above. Plaques were numerated after 24 h of incubation at 37C.

Phage Isolation and Cross-Infection
Phage induction was performed with the mitomycin C method in 37 clinically isolated vaginal strains, 2 control vaginal type strains (L. gasseri ATCC 9857 and L. jensenii ATCC 25258), and 2 other control strains (L. gasseri ADH and L. delbruecbii subsp, lactis ATCC 15808). The lysates were used to interact with the 41 Lactobacillus strains to screen for phage-sensitive indicator strains. Seven lysates were confirmed to contain phages because they formed single plaques on the agar plates of sensitive strains. Among the 7 phages, 3 were from 29 Lactobacillus strains of 19 healthy women, 2 from 4 strains of 4 VC patients, and 2 from 4 strains of 4 BV patients. Although the rate of phage isolation was apparently lower from lactobacilli in healthy women than in women with BV or VC, them was no apparent difference in phage sensitivity among vaginal Lactobacillus strains isolated from these women. The sources and host ranges of these phages and the sensitivity of various vaginal Lactobacillus isolates to these phages are shown in Table 1. Strains from all Lactobacillus species studied, except L. fermentum, were infected by phages. The phage bkc005 isolated from a vaginal L. acidophilus strain of a women with VC had a broad host range and was very virulent. Unlike other phages, qbkc005 could continue to spread its infection on a soft agar plate by enlarging its lytic zones with time (Fig. 1). Interestingly, the control phage isolated from the intestinal strain L. gasseri ADH also lysed 6 of these vaginal strains. plates, such as the case of dkc005 (Fig. 1). However, qbkc039 showed clear plaques after the phage infected Lactobacillus strain KC013 on agar plates (data not shown). The complete lysis of KC013 by dkc039 allowed isolation of a large amount of phage particles and gcnomic DNA. This greatly facilitated characterizations of this phage. Moreover, the phage bkc039 and its indicator strain KC013 formed an ideal in vitro model to show that a temperate phage released from a vaginal Lactobacillus stain of one woman could become a lytic phage against a different vaginal Lactobacillus strain from another woman.

Characterization of bkc039
The infection cycle of bkc039 against its new host strain KC013 in MRS broth with 10 mM CaC1 e was characterized by its one-step growth kinetics.  by an average of the sums of the molecular sizes of these restriction fragments. The agarose gel electrophoresis, which allows discrimination among covalently closed supercoiled circular, nick-relaxed circular, and linear DNA molecules showed only one distinct band of undigested phage qbkc039 DNA (data not shown). This indicated that the phage genome was a double-stranded linear DNA. The electron micrograph illustrates that the phage bkc039 has a hexagonal head, about 67 nm in diameter, and a flexible but non-contractile tail about 250 nm in length with about 60 discs (Fig. 4). This phage morphology apparently belongs to Bradley type B phages, zz DISCUSSION BV is a common vaginal infection that may result in serious sequelae, such as increased risk of preterm delivery, low birth weight, and mid-trimester loss. 6 Although the exact cause of BV is unknown, it has been well documented that during BV, the normal,  However, an increased number of cases will be needed to confirm these observations. Since it is difficult to test phage infections in vaginal lactobacilli in vivo in humans, in vitro studies were performed to show that phages released from some women's vaginal lactobacilli could lyse vaginal lactobacilli isolated from the same and/or different women (Table 1). This observation suggested that vaginal lactobacilli could be suppressed by phages in vivo in humans.
The analysis of these phages revealed some un-  (Table 1). The phage qbkc005 could also continue its infection after the host cell had stopped active growth (Fig. 1). Second, a bacterium that carries a prophagc, namely a lysogcn, is normally immune from infection by other phages of the same type. This characteristic is called "super-infection immunity," which is controlled by a phage repressor protein encoded by the prophagc genes already in the host bacterial chromosome or by other mechanisms. 14 This may explain why the phages released by a vaginal Lactobacillus strain normally would not lysc its own bacterial host. However, the "superinfection immunity" may not protect vaginal lactobacilli from infections by different types of phages. The observation that phages released from one Lactobacillus strain can lyse other lysogenic strains suggests that different types of phages may exist in these vaginal lactobacilli. But exceptions exist. Two phages, qbkc007 and qbkc012, lysed their own host strains, suggesting that these phages might be defective in their super-infection immunity.
Third, truly lytic or virulent phages are usually short-lived. Once they appear, the virulent phages can rapidly eliminate their host bacteria; as a result, they lose their shelter that enables them to live and to reproduce themselves. Therefore, phages that are temperate to some bacteria but lytic to others are of concern. All 7 vaginal Lactobacillus phages isolated in this study were temperate phages from lysogenic vaginal Lactobacillus strains. It is well known that some temperate phages can become 42 INFt,;C'I'IOUS DISEASIS IN OBS'IT{7'RICS AN/) GYNI,JCOLOGY virulent due to genetic mutations, is However, it is unknown why some temperate phages, such as +kc039, can become lyric against other vaginal Lactobacillus strains. Probably certain differences in bacterial host backgrounds prohibited these phages from integrating their DNA into the chromosome of their new hosts to form lysogens, z4 Among many characteristics of these phages, the spontaneous release of a small number of phages by lysogenic lactobacilli is noteworthy. This suggests that lysogcnic vaginal lactobacilli may be a source of potentially infectious phages.
In summary, phages were isolated from vaginal lactobacilli. In vitro studies showed that phages isolated from vaginal lactobacilli of some women could infect vaginal lactobacilli of the same and/or different women. Also, a phage isolated from a human intestinal Lactobacil/us strain lysed some of these vaginal lactobacilli, suggesting that infectious phages in the vagina may come from the fecalurogenital route. Although the phage infection in vaginal lactobacilli observed in vitro may not necessarily represent that a similar situation could happen in vivo, the results imply that vaginal lactobacilli may be inhibited by phages. This implication may be important for studying the etiology of BV that is associated with the decrease of vaginal lactobacilli. However, further studies with an increased number of clinical cases will be needed to associate phage infections in vaginal lactobacilli with women's vaginal health status.