1. Antimicrobial Effects of H2O2: Epidemiological Studies on Bacterial Vaginosis
The first studies focusing on this topic were in vivo observational studies beginning from the late 1980s. Their provide the most supporting data about the positive correlation between the vaginal presence and dominance of H
2O
2-producing
Lactobacillus species and healthy vaginal microbiota. One group of these studies focused on the association between H
2O
2 lactobacilli and bacterial vaginosis. BV is thought to be a vaginal dysbacteriosis with anaerobic overgrowth (e.g.,
Gardnerella,
Atopobium) with or without symptoms
[1][2]. Women with bacterial vaginosis have higher risks for preterm birth, late miscarriage, and HIV infection
[3][4][5].
The epidemiological studies on bacterial vaginosis demonstrated a remarkable difference in the prevalence of isolated H
2O
2-producing lactobacilli in healthy, nonpregnant women and women with bacterial vaginosis. Hydrogen peroxide-generating lactobacilli were detected in the large majority of healthy women whereas only in a small part of women with BV
[6][7]. Moreover, differences in vaginal colonization by lactobacilli in healthy women and those with BV were only observed in the H
2O
2-producing group of
Lactobacillus [7][8]. Since the absence of these bacteria in women with bacterial vaginosis was more prevalent than the increased colonization rates of anaerobic bacteria (
Gardnerella,
Mobiluncus, Mycoplasma), the theory of the presence of H
2O
2 producing lactobacilli as a critical protective factor in the healthy vaginal microbiota was reasonable
[6].
These findings also raised an important question regarding the pathogenesis of BV and the chronology of microbial events resulting in the disease. Do H
2O
2-positive
Lactobacillus species first disappear, and do obligate anaerobic bacteria take over their place or the other way round? A possible answer was provided by longitudinal studies of healthy, nonpregnant women with follow-up visits. They confirmed the lack of H
2O
2-forming lactobacilli as a primary risk factor for bacterial vaginosis
[9]. BV development was four times higher in women without Lactobacillus species producing H
2O
2 than in women colonized by these bacteria. Harboring any lactobacilli reduced the risk of BV twofold
[9]. The acquisition of bacterial vaginosis was significantly higher among women initially colonized with H
2O
2-producing strains and lost colonization of these species later, compared to persistently colonized women
[10]. H
2O
2-positive strains of
L. crispatus and
L. jensenii were the most likely to maintain persistent vaginal colonization over the period of the study, suggesting optimized host-microbiota interactions.
Interestingly, vaginal and rectal co-colonization by H
2O
2-producing
L. crispatus species is suggested as another protective factor against BV development. In a cross-sectional study, co-colonization by H
2O
2-positive lactobacilli was shown to occur very often and was associated with a reduced risk of BV 4-fold compared with vaginal colonization only
[11]. According to this observation, H
2O
2-producing lactobacilli in the distal gastrointestinal tract could contribute to maintaining the dominance and supplying eventual shortages of vaginal lactobacilli.
Epidemiological studies also helped identify several demographic and behavioral factors correlated either positively or negatively with vaginal colonization of H
2O
2-producing
Lactobacillus species. Certainly, several of these factors are also correlated to the development of bacterial vaginosis. Vaginal colonization with H
2O
2-forming
Lactobacillus strains was associated with white race, higher education, the use of barrier contraception and less smoking
[9]. Older age, parity, alcohol use, having ≥1 act of vaginal intercourse per week, vaginal cleansing, current BV, and recent use of antibiotics were associated with decreased H
2O
2+
Lactobacillus isolation. women having ≥1 act of vaginal intercourse per week (no information about condom use) or antibiotic treatment were more likely to lose colonization
[10][12]. The possible association with host-specific health issues (e.g., hormone and immunologic status) has not been investigated.
When determining the levels of H
2O
2 in vaginal secretions, women with BV had lower levels of H
2O
2 than healthy women’s concentrations (0.04 μg/mL vs. 0.17 μg/mL)
[13].
The hypothesis of the protective role of H
2O
2-producing lactobacilli in the vagina was questioned by some studies suggesting that bacterial vaginosis may develop despite the presence of lactobacilli with H
2O
2 formation
[14][15]. For example, in most of the investigated BV cases with large numbers of BV-associated species, simultaneous colonization of vaginal lactobacilli in large numbers (10
5–10
6 colony forming units (CFU)/mL) was demonstrated. Moreover, as shown in vitro, strong H
2O
2 producers were identified in BV cases as well
[14][15]. However, due to the more significant number of lactobacilli observed in healthy women, it might be the case that overall higher amounts of vaginal H
2O
2 would be generated in them than in women with bacterial vaginosis
[15][16].
2. Antimicrobial Effects of H2O2: Epidemiological Studies on Vulvovaginal Candidiasis (VVC)
One of the most frequent vaginal disorders is vulvovaginal candidiasis, caused by several species of the yeast
Candida, predominantly by
C. albicans.
Candidal vulvovaginitis occurs when
Candida species members of vaginal microbiota superficially penetrate the mucosal lining of the vagina leading to a secondary inflammatory response
[17]. An association between
Candida overgrowth and levels of lactobacilli overall, neither a deficiency nor colonization with unusual
Lactobacillus species could be observed
[18]. Most women with candidiasis had the highest lactobacilli counts; even previous antibiotic treatment did not affect lactobacilli density
[19]. Similar results were shown later: hydrogen peroxide-producing
Lactobacillus species were almost equally isolated in women with normal microbiota and women with VVC
[19]. Furthermore, a longitudinal study showed no correlation between initial H
2O
2−/H
2O
2+
Lactobacillus colonization with the development of symptomatic candidiasis. The demonstration of the protective role of H
2O
2-positive lactobacilli against the acquisition of VVC failed and therefore a possible correlation was questioned
[20]. Vaginal
Candida propagation may be facilitated by fungal intrinsic and/or local extrinsic factors more powerful than
Lactobacillus-related defense.
3. Antimicrobial Effects of H2O2: Epidemiological Studies on STI Pathogens
Hydrogen peroxide produced by lactobacilli in the vagina may not only maintain their dominance and control the physiological composition of the vaginal microbiota, but they may protect against colonization of pathogens. Preventing sexually transmitted diseases and ascending infection of the chorioamniotic membranes and uterine cavity in pregnant women is of great medical importance. Compared to in vitro studies, there are just a few investigations dealing with epidemiologic correlation between H
2O
2-producing lactobacilli and vaginal infection
[9][19][20][21][22]. Most of these epidemiological studies are complex, analyzing normal microbiota, bacterial vaginosis, and the most frequent infections simultaneously.
Infections by the protozoon
Trichomonas vaginalis belong to the group of sexually transmitted diseases, with having a new sex partner as the most important risk factor. Since
T. vaginalis is able to the phagocytosis and indirect killing of vaginal lactobacilli with its toxic metabolic products, abnormal vaginal flora and/or reduction of lactobacilli are thought to be additional risk factors for the infection
[19][23][24]. Most epidemiological studies investigating the possible role of H
2O
2-producing
Lactobacillus species in normal vaginal flora, bacterial vaginosis and VVC failed to show any correlation of the bacteria with trichomoniasis
[9][19][20].
In the case of the STI caused by
Neisseria gonorrhoeae,
Lactobacillus-dominant vaginal microbial community was shown to protect individuals from lower genital tract infection with
N. gonorrhoeae [25]. Furthermore, women colonized by H
2O
2-generating
Lactobacillus species were less frequently infected by gonococci than women lacking H
2O
2+ lactobacilli
[9].
The association between HIV infections and vaginal colonization by H
2O
2-forming lactobacilli was investigated in HIV seronegative and seropositive women. Compared with African female sexual workers carrying H
2O
2+
Lactobacillus, women without lactobacilli had a 2.5-fold higher risk of HIV-1 infection. Women with only H
2O
2-negative strains were at intermediate risk. The abundance of H
2O
2-generating lactobacilli in HIV-positive women was significantly reduced than in HIV-negative women
[21]. Analyzing the occurrence of different hydrogen peroxide-producing
Lactobacillus species in HIV seropositive women, it was demonstrated that H
2O
2-producing
L. gasseri is more prevalent in the population of HIV-1 infected women
[22]. Moreover, it was the predominant species detected among women who had high quantities of H
2O
2-producing
Lactobacillus but were negative for both
L. crispatus and
L. jensenii, suggesting an alteration of
Lactobacillus species in the vaginal flora of HIV-positive women
[22].
4. Antimicrobial Effects of H2O2: Experimental Studies
Alongside the population-based studies, the hypothesized direct antimicrobial role of vaginal lactobacilli-derived hydrogen peroxide was investigated in a series of experiments. In these microbiological in vitro studies, the H2O2-mediated killing/inhibition of target pathogens was observed by co-culturing them either with different types of lactobacilli or with its supernatants or with cervicovaginal fluid.
The most convincing epidemiological correlation was demonstrated in women with bacterial vaginosis, where vaginal colonization by H2O2-producing Lactobacillus strains was reduced. The key bacterium of BV is thought to be Gardnerella vaginalis, with its overgrowth potential beside certain other anaerobes. Initial findings of in vivo studies assumed the control of H2O2-forming Lactobacillus species over other members of the vaginal microbiota, especially over G. vaginalis with the, at least partially, direct bactericidal effect of H2O2.
In a liquid co-culture assay with lactobacilli and BV-associated organisms,
G. vaginalis and
P. bivia at pH 5 resembling vaginal acidic conditions, killing of the pathogens could be observed only when H
2O
2-producing lactobacilli were added to the system, H
2O
2-negative lactobacilli showed no effect
[26]. The H
2O
2-dependent antibacterial mechanism was demonstrated by adding catalase to the assay, which abandoned the reduction of
G. vaginalis and
P. bivia. Moreover, toxic effects on
G. vaginalis could be augmented by adding peroxidase and a halide to the co-culture
[26]. However, the study has some limitations in its interpretation in vivo. First, as shown in studies later, reduced pH alone can inhibit the growth of several vaginal microbiota members, including
G. vaginalis [27]. Therefore, bacterial depletion should have also occurred in the test system with H
2O
2-nonproducer lactobacilli. Furthermore, experiments were carried out under fully aerobic conditions, which is questionable in the vaginal mucosa. At toxic concentrations of H
2O
2 complemented with peroxidase and a halide in the experiments, there was a fall in the viable cell count of the H
2O
2-producing lactobacilli, suggesting autoinhibitory effects. Since lactobacilli dominate the vaginal microbiota, this suggests more reduced H
2O
2 concentrations in vivo and a rather subordinate role of oxidative stress caused by H
2O
2 in controlling the growth of BV-associated bacteria.
Another study of the same year with co-culture test systems, although on solid agar media, supported these concerns regarding in vitro experiments
[27]. 5 of 20 H
2O
2-positive lactobacilli isolated from healthy women or women with bacterial vaginosis exerted some inhibitory effects against a few
Mobiluncus and
Peptostreptococcus strains. In contrast, others failed to reduce the growth of
Gardnerella vaginalis,
Bacteroides spp. and other strains of
Mobiluncus and
Peptostreptococcus. Similar results were observed with H
2O
2-negative lactobacilli. The observed antimicrobial effect was hardly influenced by the pH of the medium, suggesting rather pH-dependent growth inhibition. A growth inhibitory activity of H
2O
2 alone at different concentrations (0.0003–0.3%) was not observed either. It should be noted that these experiments were carried out under anaerobic conditions, and concentrations of target bacteria were chosen arbitrarily and, therefore, probably too high for inhibition detection (10
6 CFU/mL)
[1].
A more complex study provided detailed data about the circumstances of H
2O
2 production and H
2O
2-mediated toxicity by vaginal lactobacilli
[28]. In these experiments, the inhibitory effects of 22 isolated H
2O
2-producing vaginal lactobacilli were determined on different
G. vaginalis strains. Agar well diffusion assay measured the cell growth reduction induced by
Lactobacillus culture filtrates. The influence of several culture parameters (pH, H
2O
2 presence, anaerobic/aerobic conditions) was tested independently. A low pH of around four and lactic acid accounted for 60 to 95%
Lactobacillus-derived inhibitory activity, and H
2O
2 accounted for only 0 to 30% after its denaturation with catalase treatment. H
2O
2 production was not detectable under anaerobic or static aerobic conditions.
Comparison of the antimicrobial effect of pure H
2O
2 and culture supernatants of H
2O
2-producing vaginal lactobacilli against BV-associated microorganisms demonstrated significant H
2O
2 sensitivity of
Gardnerella and
Prevotella, however, experiment conditions are unlikely to occur in vivo
[29]. Catalase treatment neutralized growth inhibition of pure H
2O
2 but did not affect the toxicity of culture supernatants proposing other toxic mechanisms than H
2O
2. Cultured media of
Lactobacillus species with moderate or low H
2O
2 production appeared to be less effective or ineffective on the growth of G. vaginalis, indicating dose-dependent toxicity and varying levels of H
2O
2 and at least in vitro
[30].
In the case of candidiasis, as seen in the epidemiological studies, the protective role of lactobacilli-derived hydrogen peroxide is rather doubtful since different
Candida species were found to be resistant to relatively high concentrations of H
2O
2 [31]. 30 g/L H
2O
2 was necessary to kill all
Candida yeast cells, and 3 g/L was inhibitory for only some
Candida cells
[31]. No
Lactobacillus strain was found to produce H
2O
2 in this high concentration. Moreover, this concentration of H
2O
2 would also act as an autoinhibitory
[31]. In contrast to these results, using
Lactobacillus culture supernatants, in which H
2O
2 reached concentrations from 0.05 to 1.0 mM, they effectively could inhibit
Candida growth. Still, it could not be neutralized with catalase, suggesting other toxic mechanisms
[29]. Furthermore,
Candida species can produce their catalase for H
2O
2 degradation, which, in turn, could be further stimulated by
Lactobacillus-derived H
2O
2 [29]. In another study of the same year, the minimal bactericidal concentration of H
2O
2 on a single
C. albicans strain was 2.52 mmol/L;
Candida albicans appeared to be approximately six times more tolerant than
G. vaginalis to H
2O
2-mediated inhibition
[30]. Supernatants of vaginal H
2O
2-producing lactobacilli, treated with proteinase K to neutralize antimicrobial peptides but with maintained H
2O
2 activity, showed insufficient eradication of
C. albicans [30].
Liquid co-culture experiments demonstrated pH-dependent growth inhibitory effects of vaginal lactobacilli on the pathogen
Neisseria gonorrhoeae with the enhancement of toxicity under acidic conditions
[32]. Gonococcal growth was significantly more inhibited by H
2O
2-producing lactobacilli. Moreover, H
2O
2-positive lactobacilli could increase catalase production by
N. gonorrhoeae at least at neutral pH, and catalase activity parallel decreased with pH, probably due to the bactericide effect of low pH on gonococci
[32]. Co-culture experiments obtained similar results based on the sandwich method with agar plates investigating the inhibitory effect of four isolated
Lactobacillus strains (
L. crispatus,
L. jensenii,
L. gasseri,
L. acidophilus) on two Gonococcal laboratory strains
[33]. All four
Lactobacillus strains inhibited the growth of all Gonococcal strains tested at low pH. Since adding catalase could effectively neutralize
Lactobacillus-mediated Gonococcal killing, H
2O
2 was suggested as the primary mediator of inhibition
[33].
In one study, supernatants of H
2O
2-producing
Lactobacillus species were shown to inactivate elementary bodies of
Chlamydia trachomatis mainly through a lactate acid-dependent mechanism since catalase-treatment could not reverse the inhibition by neutralizing H
2O
2 [34]. It was hypothesized that the rigid outer membrane of Chlamydia could prevent H
2O
2 from entering the cell
[34].
The antiviral potential of H
2O
2 was studied for Herpes Simplex Virus Type 2 (HSV-2) and Human Immunodeficiency Virus Type 1 (HIV-1)
[35][36]. Inhibition of HSV-2 multiplication by lactobacilli was demonstrated on HSV-2-infected Vero cells incubated with bacteria. However, culture supernatants of H
2O
2-forming lactobacilli could neither modify the infectivity of HSV-2 virions nor affect intracellular events of virus multiplication. Investigation of the effects of pure H
2O
2 in cell culture experiments was hindered by the fact that H
2O
2 was metabolized promptly in the cell culture, and maintaining constant levels of H
2O
2 was not possible. Incubating HSV-2 virions with H
2O
2, hydrogen peroxide showed a dose-dependent reduction of the infection capacity with a 50% inhibition at 184 μM after one h incubation. Still, that high activity level in the vagina is rather unlikely
[34][35][37]. In a simple, early study on HIV-1, H
2O
2-producing
L. acidophilus at a concentration of 10
7 CFU/mL was viricidal to HIV-1 virions in lactate buffer with pH 5.0
[36]. The role of H
2O
2 was demonstrated with the addition of myeloperoxidase and chloride to lower concentrations of the
Lactobacillus with ineffective viricidal activity alone, showing enhancement of viral reduction in the in vitro test system
[36].
The correlation of these studies with physiological conditions in the vaginal mucosa is challenging to estimate since in vitro studies contain several limiting factors for their proper interpretation. Most of the studies worked with isolated
Lactobacillus strains and not with bacterial communities; concentrations of lactobacilli and target bacteria varied in the experiments, as was the case of O
2 tension and sometimes pH values during incubation periods. Using pure H
2O
2 in inhibitory experiments, it should always be remembered that lactobacilli are H
2O
2-sensitive as well. Furthermore, cervical mucosa may harbor additional molecules able to react with and inactivate hydrogen peroxide. Using cervicovaginal fluid (CVF) from healthy women as the natural source of lactobacilli-derived compounds in in vitro experiments could eliminate some significant issues regarding microbiological tests
[38][39]. Under hypoxic conditions, CVF lost its H
2O
2 content within one hour, significantly suggesting inactivating mechanism present in CVF
[38]. The mean hydrogen peroxide content in CVF samples after aeration was only 23 ± 5 μM, one hundred times lower than maximal aerobic in vitro production (~2 mM). But even 50 μM hydrogen peroxide could not contain inactive pathogens like HSV-2,
N. gonorrhoeae,
H. ducreyii and several BV-associated bacteria. Additionally, adding 1% CVF reversed the in vitro inactivation of
G. vaginalis and
P. bivia by H
2O
2-producing
L. crispatus, suggesting vigorous H
2O
2-blocking activity of CVF and questioning the dominant role of hydrogen peroxide in the maintenance of healthy vaginal microenvironment
[38].
5. Impact of H2O2-Producing Lactobacilli on Fertility and Pregnancy Outcome
In healthy pregnancy, the abundance of vaginal H
2O
2-producing lactobacilli was lower than that in nonpregnant females
[40]. The proportion of H
2O
2+ strains decreased with the gravidae’s age and increased with pregnancy trimesters
[40].
Several studies demonstrated the association of vaginal dysbiosis with the negative outcome of fertility treatments suggesting altered vaginal microbiota and bacterial vaginosis as independent risk factors with a predictive value
[41][42][43][44][45]. Moreover, investigating 135 vaginal
Lactobacillus strains belonging to the species
L. crispatus,
L. jensenii and
L. gasseri, lactobacilli strains from healthy women generated significantly higher amounts of H
2O
2 than strains of infertile women
[46]. Colonization with
Lactobacillus species that produce hydrogen peroxide (H
2O
2) and bacterial vaginosis have been associated with lower rates of preterm birth as well, suggesting the protective role of these strains against ascending infections
[47]. Furthermore, the presence of H
2O
2-producing species (
L. jensenii and/or
L. vaginalis) during pregnancy was associated with significantly reduced rates of preterm birth and/or chorioamnionitis
[48].
This entry is adapted from the peer-reviewed paper 10.3390/antiox12051055