The Female Vaginal Microbiome in Health and Bacterial Vaginosis

1The human body is a holobiont consisting of the host and multispecies microbes, and the interdependency among them has been progressively enhanced in the approximately half a billion years of human–microbial coevolution. However, the previous knowledge of microbiota in holobionts is shaped by the researches where culture-dependent methods are used to cultivate species. With the advent of new technologies, scientists reveal that biodiversity is far beyond the microbial cells cultivated with culture-dependent methods. In particular, high-throughput sequencing approach provides further understanding of the spectrum of microbial community structure. Scientists have utilized omics approaches, including metabolomics, glycomics, metaproteomics, metatranscriptomics, and metagenomics, to verify that within a habitat, microorganisms exist in dynamic, interactive, and intricate microbial communities. The oral cavity and intestinal tract have been the long-term focus of a large number of researches on the microbial communities of human bodies. Although vaginal health is influentially significant to human reproduction and public health, it has attracted less attention. In recent years, increasing emphasis has been put on the female health, specifically in relation to vaginal microbiome. The vagina harbors a huge microecosystem containing billions of microbes. A systematic detection of the female reproductive tract microbial biomass was conducted by 16S rRNA gene sequencing. Data from 110 persons of reproductive age revealed that the vagina contains 1010–1011 bacteria.

2In the ecosystem, a homeostatic and mutualistic relationship exists between the microbiota and its human host. The host provides a humid, nutritious, and warm habitat for the microbes, whereas the resident microbiota produces antimicrobial and anti-inflammatory factors. Thus, the first line of defense against nonindigenous microorganisms is established. Nevertheless, this balance can be broken by internal and/or external factors. For internal factors, such as hormonal status, age, and immune system, the alteration of the host environment impairs its ability to control opportunistic pathogens contained in the resident microbes that could invade the human body and cause illness. External interferences, such as antibiotics, infections, and environmental microbial exposure, influence the microbiota within the habitat and are potential risk factors for diseases. The variations of these internal and/or external factors lead to the breakdown of a balanced ecosystem, also known as dysbiosis. According to DNA sequencing, reduced diversity in the intestinal ecosystem is associated with disease, whereas high diversity in the vaginal ecosystem is linked to illnesses, such as bacterial vaginosis (BV). Furthermore, BV would trigger numerous health disorders, including adverse pregnancy outcome, human immunodeficiency virus (HIV), human papillomavirus (HPV), and pelvic inflammatory disease (PID).

3This review aims to describe the vaginal microbiome with regards to female health, discuss BV characteristics, present a strong association between BV and diseases, and outline the requirement for comprehensive, accurate, and advanced diagnosis and therapies to lower adverse health outcomes.

4The vaginal microbiome is an intricate and dynamic microecosystem that constantly undergoes fluctuations during the female menstrual cycle and the woman’s entire life. The vaginal mucosa is made up of a stratified squamous nonkeratinized epithelium covered by cervicovaginal secretion. The vaginal mucosa acquires oxygen, glucose, and other nutrients from underlying submucosal tissues through diffusion due to the limited blood supply. This establishes a relatively anaerobic habitat condition. The vagina houses a complex microbial community that subsists in a symbiotic relationship with the host. Thus, the indigenous environment, microorganisms, and their genomes jointly compose the entire habitat, also known as the vaginal microbiome.

5In women of reproductive age, physiological changes, such as changes in hormone levels, cause fluctuations in the vaginal microbiome. Marked differences have been reported between non-pregnant and pregnant women in terms of the vaginal microbiome. According to the comparison results, a sharp decline in the diversity and abundance of the vaginal microbiome is observed in pregnant women. Moreover, the predominance of Lactobacillus spp., Actinomycetales, Clostridiales, and Bacteroidales is observed in pregnant women. In non-pregnant women, the predominance of Lactobacillus spp., Actinobacteria, Prevotella, Veillonellaceae, Streptococcus, Proteobacteria, Bifidobacteriaceae, Bacteroides, and Burkholderiales is observed. Thus, the vaginal microbiome would change temporally in a single person. In addition, the vaginal microbiome differs largely among individuals, and the differences are due to variations in sexual activity, douching, chronic stress, regional disparity, race, and other factors. Based on high-throughput sequencing studies, five community state types (CSTs) exist in terms of the vaginal microbiome. Specifically, the research on 396 North American asymptomatic women from four ethnic groups illustrates that the majority of vaginal microbiomes are dominated by single or multiple Lactobacillus species and are classified into five CSTs. CSTs I, II, III, and V are dominated by L. crispatus, L. gasseri, L. iners, and L. jensenii, respectively, whereas the CST IV refers to high diversity of the microbial community characterized by obligate anaerobic bacteria. High Nugent scores are usually linked to CST IV but are also observed in other CSTs. Among the five groups, the CSTs I, II, III, and V exist in 89.7% white women and 80.2% Asian women, whereas these percentages are 61.9% and 59.6% in black and Hispanic women, respectively. A shift in ethnic groups is apparent when CST IV dominated. The differences in vaginal microbiome by race of women might be driven by host genetic factors, such as immune system, ligands on the surface of epithelial cell, and the quantity and components of vaginal discharge. Compared to behavioral and cultural differences, host factors might play a more crucial role in shaping the vaginal microbiome among races. Currently, there are few genotyping studies associated with healthy vaginal microbiome. This is an area of research that would benefit from further investigation.

6Lactobacillus species flourish in the vaginal anaerobic environment and produce various antimicrobial compounds, such as lactic acid, hydrogen peroxide (H2O2), and bacteriocins, thereby contributing to a healthy vaginal microbiome and establishing a defense against the invading pathogens. Lactobacillus species are the main source of l-lactic acid and d-lactic acid that keep the pH value of the habitat lower than 4.5, whereas epithelial cells contribute about 20% l-lactic acid. Nonetheless, the role of H2O2 in the vaginal microbiome remains controversial. Some studies have demonstrated that it has positive effects on the inhibition of the overgrowth of pathogenic microbes. O’Hanlon et al. found that H2O2 at physiological levels displays the undetectable ability of eliminating pathogenic microbes, whereas at high levels, it shows greater antimicrobial ability toward Lactobacillus spp. than pathogenic microbes. This finding indicates that H2O2 is not a vital antimicrobial agent for maintaining the homeostasis of the vaginal microbiome. Lactobacillus also synthesizes bacteriocins, a type of antimicrobial peptides that can permeabilize the microbial cell membrane of nonindigenous microorganisms. Furthermore, they can adhere to vaginal epithelial cells and compete with other microbial cells for binding sites. This finding is important, because the adhesion of pathogen to epithelial cells is the first step and a crucial prerequisite of infection. Notably, the dominant Lactobacillus species determines the extent of vaginal ecosystem protection. For instance, dysbiosis and low stability are usually related to the vaginal microbiota dominated by L. iners. On the contrary, health and high stability of the vaginal community are enhanced by L. crispatus that provides d- and l-lactic acids. Different from other Lactobacillus species, L. iners cannot generate d-lactic acid, which plays a more important role than l-lactic acid.

7BV is a highly prevalent lower genital tract disorder among women of reproductive age globally. It afflicts 23%–29% of women worldwide, and $4.8 billion is spent on symptomatic BV treatment annually. BV is characterized by the loss or sharp decline in the total number of Lactobacillus and a corresponding 100–1000 fold increase in the concentration of facultative or obligate anaerobic microbes, such as Gardnerella, Prevotella, Atopobium, Mobiluncus, Bifidobacterium, Sneathia, Leptotrichia, and some novel bacteria in Clostridiales order referred to as BV-associated bacteria (BVAB) 1–3.

8Gardnerella is the most common microorganism identified from the vaginal samples of women with BV. It was first isolated by Leopold from the cervix swabs of women and the urine of men in 1953. Later, it was found to be related to BV and named Haemophilus vaginalis by Gardner and Dukes in 1955. It is subsequently classified into the genus Corynebacterium. Afterward, due to the results of two taxonomic researches, it was placed into a new genus Gardnerella and renamed G. vaginalis. Until 2019, G. vaginalis was the only identified species in the genus Gardnerella. Recently, a whole genome sequence analysis was performed on 81 Gardnerella strains, and results revealed the existence of 13 disparate species of genus Gardnerella. In the following additional physiological and chemotaxonomic analyses supported by MALDI-TOF and biochemical activity studies, three new Gardnerella species, namely, G. piotii, G. swidsinskii, and G. leopoldii, were described, and the description of G. vaginalis was amended.

9The close relationship between BV and G. vaginalis seems to indicate that this microorganism is the sole pathogen in BV. However, G. vaginalis has displayed high sensitivity and low specificity when used to determine whether BV is positive or negative. Healthy or asymptomatic women may also carry G. vaginalis, suggesting that the presence of G. vaginalis in the vagina does not always result in BV. Thus, it is important to have a proper understanding of the role of G. vaginalis as a spectator, participant, or causative agent of BV. In numerous studies that aimed to determine the characteristics linked to virulence, various approaches are utilized to analyze the phenotypic, genotypic, and ecotypic diversity of G. vaginalis ( Table 1 ). Based on biochemical tests, 8 and 17 biotypes of G. vaginalis were identified, respectively. Much effort has been devoted to the exploration of the association between its biotypes and BV, but the results remain controversial. For genotyping, four clusters (A1 to A4) of G. vaginalis are defined in the vaginal microbiota based on 60 kDa chaperonin protein (cpn60) universal target region sequencing. Recently, G. vaginalis has been classified into four subgroups (A–D) with the use of cpn60 universal target sequencing. Using whole-genome sequencing, G. vaginalis can reportedly be divided into clades 1, 2, 3, and 4, corresponding to cpn60-based subgroups C, B, D, and A, respectively. The relationship between G. vaginalis genotypes and BV is controversial. In the only study about the ecotypes of G. vaginalis in 2017, three ecotypes were identified as a result of the distinct gene gain/loss of specific functions based on the combination of phylogenetic structure and functional gene analysis. Overall, these efforts have contributed to the identification of the linkage between G. vaginalis and different states (health, asymptomatic BV, and symptomatic BV), thereby ultimately improving the approaches for the accurate diagnosis of BV.

10G. vaginalis harbors a variety of virulence factors associated with pathogenic potential, wherein sialidase and vaginolysin are the most widely investigated factors. Sialidase A gene is associated with BV and the presence of biofilm. With the use of BLAST searches, two more G. vaginalis sialidases (NanH2 and NanH3) were identified. Regarding the substrate specificity of hydrolysis, recombinant NanH2 and NanH3 were active on Siaα2-3/2-6-linked N- and O-glycans containing mucosal substrates, whereas recombinant sialidase A showed a limited ability to hydrolyze mucosal and synthetic substrates. G. vaginalis uses sialidase to hydrolyze sialic acid residue from mucus sialoglycans in the vagina and then catabolizes free carbohydrate, thus contributing to the degradation of vaginal mucus barriers. Notably, some Gardnerella spp., including G. swidsinskii, G. leopoldii, and a certain subgroup of G. vaginalis, possess negative sialidase activity. As for vaginolysin, it is a pore-forming toxic compound belonging to the cholesterol-dependent cytolysin family and facilitates the lysis of target cells, such as vaginal epithelial cells. Other virulence factors, such as prolidase and glycosulfatase, are also reportedly associated with BV.

11Biofilm is a structured community of microbes attached to the abiological or biological surface and inlaid in their own secreted polymeric matrix comprising carbohydrate, protein, and nucleic acid. The formation of biofilms is an intricate, dynamic, and interactive process associated with motile planktonic microbes and microbial aggregates. A variety of bacterial and fungal microbes, such as Gardnerella spp. and Candida spp., can form biofilms. The biofilms produced by pathogenic microorganisms shape an effective protection against host immune responses and antimicrobials. The life cycle of biofilms comprises three phases, namely, attachment to a surface, secretion of polymeric matrix and aggregation of microbes for yielding mature biofilms, and dispersion by detaching from biofilms.

12The polymicrobial biofilms formed on vaginal epithelium play a crucial role in the pathogenesis of BV. G. vaginalis is considered the primary colonizer which can establish a scaffold for the attachment of other BV-associated microbes, thus enabling the development of polymicrobial biofilms. Atopobium vaginae, one of the second colonizers of polymicrobial biofilms, is a strict anaerobic microbe with great predictability for BV. G. vaginalis biofilms have higher tolerance to two common agents of health vaginal discharge, namely, lactic acid and H2O2, than planktonic cells. This may shelter G. vaginalis and other BV-associated microbes from adverse environments. With the use of RNA sequencing, G. vaginalis biofilms are observed to have decreased metabolic activity and down-regulated virulence factor (vaginolysin), which are important for biofilm persistence. These phenotypes are probably linked to the recurrent and chronic BV characteristics. G. vaginalis biofilms are reportedly present in fallopian tube and endometrial samples, indicating that G. vaginalis biofilms could move to the upper genital tract, thus leading to adverse pregnancy outcomes.

13BV is considered a dysbiosis often presenting clinical symptoms which can be caused by a large number of microbes with proinflammatory features, coupled with the host immune response. Vaginal samples from females with BV reportedly harbor high levels of immune mediators, such as interleukin (IL)-8, IL-6, IL-1α, IL-1β, IL-12p70, and TNFα. Different immunological factors might be adopted for various species. For instance, L. crispatus is associated with a marked increase in gamma-induced protein 10 (IP-10), and a significant drop in IL-12 (p70), IL-8, IL-1β, and IL-1α. Nevertheless, according to the analysis of vaginal swabs, G. vaginalis is correlated with the decline in IP-10 and the increase in IL-12 (p70), IL-8, IL-1β, and IL-1α. Similar to G. vaginalis, A. vaginae is also associated with the increase and decrease of the same factors. Greater levels of IL-1β, IL-8, and interferon (IFN)-γ have been observed in females harboring a large quantity of Prevotella spp.. The critical effect of IL-36G on women with BV is verified. Consequently, IL-36G level increases in vaginal samples of women with BV. IL-36G might play a vital role in the immune response to BV and other diseases.

14BV is associated with adverse reproductive health outcomes, such as sexually transmitted infections (STIs) and pelvic inflammatory disease (PID). Moreover, preterm birth (PTB), low birthweight, miscarriage, and other adverse obstetric outcomes are also linked to BV.

15STIs are common acute conditions. Although the majority of STIs are generally not fatal, they lead to a large burden of disease. BV increases the risk of some of the STIs, such as herpes simplex virus type 2 (HSV-2), HPV, HIV, and chlamydial, gonococcal, and trichomonal infection. Women with Nugent scores of 9–10 are at the highest risk, whereas women with Nugent scores of 4–8 are at moderate risk of any bacterial STI (Neisseria gonorrhoeae, Chlamydia trachomatis, and Trichomonas vaginalis). Compared with women with normal vaginal microbiota, women with BV were 4- and 3.4-fold more likely to detect positive for N. gonorrhoeae and C. trachomatis, respectively. In a large longitudinal study (n = 3620), BV is reportedly associated with 1.5–2 times elevated risk for the acquisition of chlamydial, gonococcal, or trichomonal infection. In another cohort study, temporal associations between STI and BV in both directions were assessed. BV and gonorrhea/chlamydia are risk factors for each other, which indicates treating either condition might have a protective effect on the other. Women with BV are at higher risk of acquiring HSV-2 compared with those without BV. More recently, several studies have also found that BV is linked to the elevated risk of STIs. Unlike the abovementioned observational studies, two randomized trials were performed to investigate whether periodic treatment of BV can affect the incidence of STIs. The first study, which involved women with asymptomatic BV, revealed that the incident chlamydia infection is markedly reduced when intravaginal metronidazole gel was used for 6 months. Another study involving women with BV found that the incidence of N. gonorrhoeae, C. trachomatis, or M. genitalium infection is reduced when intravaginal metronidazole and miconazole were used for 12 months.

16HIV infection is diagnosed in over 1 million females yearly, and BV is a major risk factor of HIV infection. For instance, BV is linked to a greater risk of HIV infection, and the level of HIV is greater in the vaginal discharge of women who are infected with HIV and BV than in those of women who are infected with HIV but have no BV. In Uganda, compared with the frequency of HIV-1 in women with normal vaginal flora (14.2%), a markedly elevated frequency (26.7%) was observed in women with severe BV (Nugent scores 9–10). Similarly, a study on HIV-infected Indian women showed the elevated risk of BV in HIV-infected and HPV-positive women and the high prevalence of BV among women infected with HIV. Women with HIV-BV+, HIV-BV-, HIV+BV-, and HIV+BV+ were compared in terms of microbiota of the lower genital tract, and results showed that HIV infection was linked to increased diversity. Furthermore, vaginal microbial diversity is important in HIV prevention. According to the results of the clinical trial, the efficacy of tenofovir gel microbicide, which prevents HIV transmission, is dependent on vaginal microbiota. Tenofovir decreases the incidence of HIV by 61% in women with Lactobacillus-dominant microbiota but merely by 18% in women without Lactobacillus. Further, the concentrations of tenofovir are lower in women without Lactobacillus and are negatively correlated with BV-associated microbes that could degrade tenofovir by metabolism. In addition, compared with women with lower levels of P. bivia, women with higher levels (>1%) are linked to 13-fold increased likelihood to become infected with HIV. Women with microbiota dominated by BV-associated microbes have increased amount of HIV target cells, namely CD4 T cells, in the genital mucosa. This may contribute to the elevated HIV acquisition risk.

17Among young women, the most common sexually transmitted infection is HPV, which contributes greatly to cervical cancer. HPV infection is affected significantly by BV. CST IV-BV is a risk factor for persistent HPV, and the biomarkers of persistent HPV are Atopobium spp. and the sialidase gene of G. vaginalis. The BV prevalence in the high-risk (HR)-HPV clearing group was 5.0%, which was lower compared with the increased BV prevalence of 11.2% in the HR-HPV persistent group. Further, women who currently have BV reported a lower clearance of HPV than women without BV. Similarly, BV is reportedly associated with delayed HPV clearance and elevated risk of HPV incident and prevalence. In Korea, increased bacterial diversity with decreased Lactobacillus spp. was observed in women with HPV. Particularly, BV-associated bacteria, namely Sneathia spp., is considered a possible biomarker linked to HPV. In a study on Nigerian women, a similar linkage was found between the increased proportion of BV-associated bacteria, such as Leptotrichia and Prevotella, with the decreased in Lactobacillus spp. and HR-HPV. Among Swedish HPV-infected women, a large proportion of Megasphaera, Prevotella, Sneathia, BVAB1, and BVAB2 was identified in the vaginal microbiota. Women with a total loss of Lactobacillus in their vaginal microbiota are 2-fold more likely to have oncogenic HPV.

18PID is an infection-caused inflammation of the upper genital tract. BV is considered a risk factor for PID that can cause adverse reproductive sequelae, such as infertility, chronic pelvic pain, and ectopic pregnancy. BV-associated microbes were reportedly linked to the elevated risk of PID development, whereas non-BV-associated microbes showed no effect on the risk of PID. Patients with acute endometritis are more likely to possess BV and less likely to carry lactobacilli. Compared with women with normal vaginal microbiota, subclinical PID is 2.7 times more frequently identified in women with BV. The presence of A. vaginae, S. amnionii, BVAB1, or S. sanguinegens is associated with PID and its sequelae, including recurrent PID and infertility. In a recent study of women at high risk of STI, the presence of BV-associated microbes, such as A. vaginae, Megasphaera spp., Sneathia spp., Prevotella amnii, and Eggerthella-like bacterium, in the vagina can increase the likelihood of PID development. Further, a larger bacterial load of BV-associated microbes predicted PID. The identification of BV-associated microbes in PID indicated lower to upper reproductive tract ascension. This finding may be due to the enzymes produced by BV-associated microbes. These enzymes, such as mucinase and sialidase, may degrade mucin barriers and facilitate ascending infection, thereby leading to PID..

19PTB, which refers to gestation within 37 weeks, has become a serious health challenge worldwide. PTB occurs in about 15 million pregnancies each year and is a primary risk factor of neonatal death. PTB and other adverse obstetric outcomes have reportedly been associated with BV. BV is verified in a meta-analysis to be powerful risk factor for late miscarriage and associated with maternity infectious incidence rate and preterm delivery. The BV in early pregnancy is linked to preterm delivery or delivery of an infant with a low birthweight. High levels of BV-associated microbes, such as A. vaginae and G. vaginalis, are strongly related to the risk of PTB (hazard ratio 3.3). Other BV-associated microbes, such as Sneathia sanguinegens, Atopobium, and Mobiluncus curtsii/mulieris, are risk factors for spontaneous PTB. A recent multi-omic study of large samples showed that increased level of BV-associated microbes and a sharp decline in the level of L. crispatus are observed in women with PTB. As a method of preventing PTB, BV treatment during pregnancy has been evaluated in many studies. However, PTB incidence is not always decreased by antibiotic therapy of BV. BV-associated microbes might cause infection during gestation, because they could move into the uterus before gestation.

20These abovementioned studies highlight the association between BV and many diseases, such as STI, PID, and PTB. BV is a highly prevalent disorder. Thus, the interventions that decrease BV incidence could reduce the incidence of BV-associated diseases. The accurate and efficient diagnosis and treatment of BV may be the key to preventing such diseases.

21BV can be diagnosed by performing two standardized tests, namely, Amsel criteria and Nugent score, by using vaginal swabs. Amsel criteria is used for clinical diagnosis, and at least 3 of the following 4 characteristics are required to determine BV-positive patients: homogeneous and milk–like vaginal fluid ( Figures 1A, B ), increased vaginal pH, fishy odor, and the observation of clue cells (vaginal epithelial cells covered with bacteria) through microscopy. Nevertheless, the characteristics above are often absent in some cases, and the corresponding diagnosis is somewhat subjective. Nugent score relies on the quantitative analysis of the morphotypes of different microorganisms, such as Lactobacillus and Gardnerella, in Gram-stained vaginal discharge ( Figures 1C, D ). This laboratory-based approach uses a score system, in which the scores of 0–3, 4–6, and 7–10 are considered normal, intermediate, and BV, respectively. However, 27% of asymptomatic women possess a high diversity of microbial community dominated by obligate anaerobic bacteria instead of Lactobacillus. Thus, the combination of the two approaches above with clinical and microbiological morphology findings may result in a more accurate and reliable diagnosis.

22There are many point-of-care (POC) methods available for the diagnosis of BV, such as FemExam card, OSOM BVBlue, VGTest ion motility spectrometry (IMS), and wet mount. FemExam card measures amines, proline aminopeptidase, and pH of vaginal discharge. The OSOM BV Blue approach is a chromogenic test that detects the level of sialidase in vaginal discharge. VGTest-IMS can be used as an electronic molecular sensor of biogenic amines in vaginal discharge, particularly trimethylamine of BV. Given the limitations of the aforementioned methods applicable to the diagnosis of BV, attentions have been paid to molecular diagnostic approaches which have the capability to detect and quantify fastidious microbes. Commercially available molecular tests are direct probe tests and nucleic acid amplification tests. Direct probe tests, such as Affirm VP test and bacterial vaginosis/vaginitis panel, apply DNA probe to identify the specific sequences of targeted microbes from vaginal discharge. These tests are most useful for symptomatic women since the targeted microbe for BV, namely G. vaginalis, may exist in women with healthy vaginal microbiome. Nucleic acid amplification tests include NuSwab, SureSwab BV DNA quantitative real-time PCR, BD Max vaginal panel, and BV panel. NuSwab meartures three positive indicators of BV diagnosis, namely Megasphaera type 1, BVAB2, and A. vaginae, and one negative indicator of BV diagnosis, namely L. crispatus. SureSwab BV real-time PCR is used in the detection of three H2O2-producing Lactobacillus species (Lactobacillus acidophilus, L. jensenii, and L. crispatus) and three BV-associated microbes (Megasphaera spp, A. vaginae, and G. vaginalis). BD Max vaginal panel can detect two Lactobacillus species (L. jensenii and L. crispatus) and four BV-associated microbes (A. vaginae, G. vaginalis, BVAB2, and Megasphaera type 1). BV panel identifies A. vaginae, G. vaginalis, and Megasphaera type 1 and 2 as accurate predictors for diagnosis of symptomatic BV.

23The etiology of BV remains a persistent conundrum, and consequently, developing and applying more comprehensive, accurate, and advanced approaches for its diagnosis would be necessary. The use of novel approaches, such as deep sequencing of the 16S rRNA gene, lipidomics, glycomics, metabolomics, and proteomics, provides further insight into the features of BV. Thus, the improved methods will fuel the application of accurate diagnostic strategies and will contribute critically to the development and application of definitive diagnostic methods. For example, based on high-throughput sequencing, women with BV possess heterogeneous vaginal microbiota with high species diversity and richness is observed. Afterward, with the use of the multi-omic approach (16S rRNA gene sequencing and metabolomics) approach, useful biomarkers (putrescine, cadaverine, 2-methyl-2-hydroxybutanoic acid, and diethylene glycol) are identified for the diagnosis of BV. In addition, artificial intelligence, which refers to machine learning techniques, is performed on the diagnosis of BV and is found to have high interrater reliability and automaticity.

24In short, various diagnostic approaches can be used to diagnose BV. Clinicians should choose a feasible method for BV diagnosis in different cases based on the evaluation of necessary time, cost, and accuracy. For instance, the high-throughput sequencing method might be more suitable for diagnosis among women with recurrent and intractable BV, whereas the wet mount method is widely used for its rapid results and low cost.

25Currently, according to Centers for Disease Control and Prevention, the first-line therapeutic strategies are oral metronidazole (500 mg twice a day for 7 days), intravaginal metronidazole gel (5 g once a day for 5 days), and intravaginal clindamycin cream (5 g once a day for 7 days). Metronidazole is a 5-nitroimidazole originally used for the treatment of trichomoniasis and is found to be useful against anaerobic infection. Clindamycin is a lincosamide antibiotic that does not differ greatly in terms of effectiveness against BV when compared with metronidazole. Vaginal dequalinium is an alternative treatment regimen for BV according to European International Union Against Sexually Transmitted Infections World Health Organization. Dequalinium chloride is a quaternary ammonium compound that possesses cure rates similar to those of clindamycin. Being an agent in the second-line therapeutic strategy, tinidazole has few side effects and good pharmacokinetic profile, and it is also an agent in 5-nitroimidazole class. The food and drug administration has recently approved the single-dose of 2 g secnidazole in granule formation for the treatment of BV. A research has revealed that 2 g secnidazole and 500 mg metronidazole twice a day for 5 days have the same effect. In addition, the single-dose regimen of secnidazole may improve patient compliance. Thus, this next-generation 5-nitroimidazole may serve as a substitute for the 5–7 days therapeutic strategy against BV.

26Although these antibiotics are effective against BV-associated bacteria and somewhat relieve symptoms, the remission is usually temporary, and many patients relapse after treatment. The high recurrence rate (50%–67%) may be the result of the inability of antibiotics to eliminate the biofilm-associated bacteria of BV in vagina ( Figure 2 ). For example, Ahrens et al. reported that G. vaginalis and other BV-associated bacteria are eradicated or are largely decreased in 58% of the patients treated with metronidazole. Meanwhile, none of these bacteria are eliminated in the remaining half of the patients. This phenomenon is attributed to the sheltering of G. vaginalis and other bacteria by biofilms. In addition, Swidsinski et al. observed the persistence of G. vaginalis biofilms after oral metronidazole therapy. Therefore, a promising therapeutic strategy against BV, i.e., adjuvant-based therapy with the ability to disrupt biofilms, has been developed ( Figure 2 ). Extracellular DNA is an important integrant of G. vaginalis biofilms. The enzymatic activity of DNase can be used to disrupt newly formed and mature biofilms. DNase has shown its ability to release microorganisms from biofilms to supernatant fractions and to enhance the activity of metronidazole. Similarly, lysozyme can reportedly disrupt biofilms and prevent their formation. The combination of lysozyme and antibiotic (clindamycin or metronidazole) can potentiate both the ability of antibiotics and biofilm disruption. In addition, the use of amphoteric tenside sodium cocoamphoacetate causes degradation of the established biofilms and enhances the effect of metronidazole. Lauramide arginine ethyl ester and subtilosin reportedly have great bactericidal effect on G. vaginalis biofilms. These two compounds have displayed a synergistic effect with antibiotics (clindamycin and metronidazole) against G. vaginalis biofilms. Besides the abovementioned adjuvants investigated with antibiotics, many others have been studied individually. For instance, L. reuteri RC-14 can degrade surface density, depth, and area of G. vaginalis biofilms. Thymol is effective in the inhibition of mature and native G. vaginalis biofilms. Endolysin (PM-477) can selectively and effectively eliminate Gardnerella in native polymicrobial biofilm and in cultures of isolated strains. Other strategies involving the utilization of boric acid, octenidine, cationic amphiphiles, and amphoteric tenside (WO3191) also reportedly have a positive effect on BV-associated biofilms.

27Microbial-based therapeutics have recently attracted an increasing amount of interest owing to the beneficial effects to the host health. As pivotal bacteria in the healthy vaginal microbiome, Lactobacillus species can act as antimicrobial adjuvants due to their ability to potentiate the effect of antibiotics. In 2020, a phase 2b trial (NCT02766023) was conducted on 228 women to assess the efficiency of L. crispatus CTV-05 in preventing BV relapse. When L. crispatus CTV-05 was used after antibiotic treatment, a remarkably lower rate of BV recurrence was observed at 3 months compared with placebo. Vaginal microbiota transplantation (VMT), in which optimum vaginal microbiota is transplanted to patients, is another microbial-based therapeutic strategy. In 2019, the first exploratory research has reported the viability of utilizing VMT as a long-term therapeutic regimen for women with intractable BV. In this study, five patients with intractable BV were treated with VMT after antibiotic treatment, and long-lasting relief was shown in four of these patients in the follow-up period of 5–21 months after VMT. This finding suggested the remarkable alleviation of symptoms and the rebuilding of the vaginal microbiota with Lactobacillus spp. dominance. The risks of this treatment include the acquisition of antimicrobial-resistant microbes, sperms, undetected pathogens, and other clinically silent phenotypes from the donors. Overall, antibiotics, biofilm-disrupting agents, probiotic Lactobacillus, and VMT can be utilized separately or in combination to regulate the microbiome through the reestablishment of vaginal eubiosis.

28The intricate and dynamic vaginal microbiome brings the challenge for diagnosis. The challenge for treatment is attributed to inaccurate diagnostics, biofilm, antibiotic resistance, and the simultaneous elimination of both pathogenic bacteria and probiotics, such as Lactobacillus. It is expected that future research should be conducted to target specific microbes, thus eliminating more pathogenic bacteria, but without affecting probiotic microorganisms. In addition, it is necessary to develop personalized diagnosis and treatment by taking individual differences in vaginal microbiome into consideration.

29Further, with increasing knowledge of BV-associated epidemiology and pathogenesis, more improved novel therapeutic strategies will be developed in the future.

30The vaginal microbiome forms a homeostatic and mutualistic relationship with human host and plays an important role in vaginal health and disease. The variations of internal and/or external factors lead to the breakdown of a balanced ecosystem, which is also known as dysbiosis. Although the increasing scientific knowledge has already provided insights into the characteristics of vaginal microbiome and its correlation with diseases, such as STIs, PID, PTB, there is still inadequate understanding of interactions between microbiota and the host. Efforts should be made to reveal the mechanism of interactions between species and their impact on vaginal microbiome.

31As a highly prevalent dysbiosis, BV triggers numerous adverse health outcomes and becomes a burden to individuals and public health. G. vaginalis is the most common microorganism detected in BV, but its presence in the vagina does not always lead to BV. Other microorganisms, such as Atopobium and Prevotella, also have a strong relationship with BV. These BV-associated microbes can affect immune mediators, which may serve as predictive biomarkers for dysbiosis, in the vaginal environment. Although the epidemiology and pathogenesis of BV is not fully understood, studies based on genomic, lipidomic, glycomic, metabolomic, and proteomic techniques may provide further insights. More improved novel and accurate diagnosis and therapeutic strategies will be developed based on the accumulated information on BV.

32XC drafted the manuscript for publication. YL participated in writing chapter diagnosis and treatment of BV. TC prepared the first draft of Figure 1. RL reviewed and revised the manuscript. All authors contributed to the article and approved the submitted version.

33This work was supported by Science and Technology Project of Jinan Municipal Health Commission (2019-1-25 and 2020-4-71).

34The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.