The Vaginal Microbiome and HIV Transmission Dynamics

Abstract

Purpose of review:

Among women, having a non-optimal, highly diverse vaginal microbiome dominated by bacteria other than optimal Lactobacillus species such as L. crispatus or L. jensenii predicts HIV transmission. Reducing HIV acquisition among women requires a better understanding of the mechanisms through which the vaginal microbiome impacts HIV transmission dynamics and how to more effectively treat and intervene. Technological advancements are improving the ability of researchers to fully characterize interacting host-bacteria mechanisms. Consequently, the purpose of this review was to summarize the most innovative research on the vaginal microbiome and its role in HIV transmission in the past year.

Recent findings:

Studies combining multi-omics, experimental, and translational approaches highlight the associations of a non-optimal microbiome with maladaptive alterations in immune cell functioning, vaginal metabolites, host cell transcription, mucosal immunity, and epithelial barrier integrity. While there are multiple mechanisms proposed to increase HIV acquisition risk, there are virtually zero acceptable and effective treatments to improve vaginal microbiome and immunity.

Summary:

Women-centered solutions to modify the vaginal microbiome and bacterial metabolites should continue to be explored as a mechanism to reduce HIV acquisition.

Introduction

More than half of people living with HIV are women, who mostly acquired HIV through vaginal intercourse, making the vaginal barrier the primary site of HIV acquisition (1). Alterations to microbial or host factors in the vaginal environment can hinder natural immunity and increase the likelihood of HIV acquisition (2–4). It is well established that a more diverse, non-Lactobacillus dominant vaginal microbiome predicts HIV seroconversion (5–10). Across studies, Bacterial Vaginosis (BV), a condition characterized by depleted Lactobacillus and high vaginal microbial diversity, is associated with an estimated 60% increase in risk of HIV acquisition (6). Mechanisms posited to explain the relationship between the vaginal microbiome and HIV acquisition in prior research include the negative impact of BV-associated bacteria on vaginal pH, epithelial barrier function, inflammation, and influxes in HIV-susceptible target cells (4–6,11,12). Vaginal microbicide efficacy may also be altered by vaginal bacteria (13,14). However, prior research efforts have been hindered by the multifactorial nature of host-bacterial interactions, high individual variation, and biobehavioral confounders (2,6,15). Recent technological advancements have supported innovative strategies to study the role of the vaginal microbiome in HIV transmission, including the use of multi-omics, functional assessments, and mixed in vitro/vivo models (16,17). The purpose of this review is to summarize the most innovative research on the vaginal microbiome and HIV transmission dynamics from the past year, also summarized in Figure 1.

Figure 1: Vaginal Microbial Mechanisms Facilitating HIV Acquisition.

An optimal vaginal microbiota is dominated by Lactobacillus spp., which can produce different metabolites such as lactic acid and short-chain fatty acids. This environment is associated with low inflammation, high epithelial integrity, and protection against pathogens (left panel). A non-optimal microbiota (characterized by high diversity, low abundance of Lactobacillus spp., and high abundance of anaerobic bacteria) is associated with increased inflammation, high pH, weak junction and epithelial integrity, and an elevated risk of HIV acquisition (right panel). Created with BioRender.com.

The Vaginal Microbiome and Host Immune Cell Functioning

For HIV acquisition to occur, the virus must penetrate the vaginal mucosal barrier to directly infect epithelial CD4+ T-cells or be taken into antigen presenting cells in the epithelial layer and then presented to HIV-susceptible cells (12). BV is associated with increases in neutrophils and HIV-susceptible target cells in the reproductive tract (11,18). Research over the past year has further demonstrated that a non-optimal vaginal microbiome may disrupt the functioning of two key immune cells - neutrophils and Langerhans cells, thereby promoting epithelial damage and HIV presentation to T-cells (19**,20**).

Costa-Fujishima and colleagues (19**) performed complementary murine and human studies demonstrating alterations in the behavior of vaginal neutrophils in response to BV-associated bacteria. Among mice, there was an influx of neutrophils accumulated in the vaginal epithelium during the diestrus stage of the estrous cycle and after a challenge with BV-associated bacteria (19**). Mice with neutrophil influx in response to BV-associated bacteria (M. mulieris and G. vaginalis) showed damage to the mucosal barrier, including downregulation of Desmoglein-1 expression (a component of cell-cell junctions), thinning/breakdown of the vaginal epithelium, reduced epithelial adherens junction signaling, decreased migration of epithelial cells, and reduced integrin signaling (19**). When neutrophils were depleted after the bacterial challenge, epithelial barrier function improved, supporting the mechanistic role of neutrophils in causing barrier damage (19**). In the human study, women with non-Lactobacillus dominant microbiomes showed upregulated neutrophil-associated proteins and downregulated epithelial cell-cell junction proteins compared to women with Lactobacillus-dominant microbiomes (19**). The results of these studies suggest BV-associated bacteria is associated with altered neutrophil function, and BV-associated bacteria and neutrophils together may impair epithelial barrier integrity (19**). Findings are aligned with prior research showing activation of the innate immune response and HIV-susceptible target cells in response to BV-associated bacteria (11,18).

Bacterial products in the vagina have the potential to alter the function of epithelial immune cells, causing them to be less effective in inactivating HIV and more likely to transmit HIV to CD4+ T-cells (20**). Van Teijlingen and colleagues (20**) examined the impact of bacterial ligands on the increased susceptibility of vaginal Langerhans cells (LCs) to HIV infection. LCs are macrophages residing in epithelial layers that coordinate immune responses in part by serving as antigen-presenting cells (12,21). LCs have the ability to either deactivate HIV, thereby preventing infection, or become infected with HIV and present the virus to CD4+ T-cells, thereby promoting HIV infection (20**,21). Van Teijlingen et al., (20**) found that vaginal LCs isolated from healthy donor tissue expressed receptors and toll-like receptors that could enable HIV-1 binding and the sensing of bacterial products. When immature LCs were stimulated with bacterial ligands, such as lipopolysaccharide (LPS), they showed increased expression of co-stimulatory molecules that can prime LCs for antigen presentation and activation (20**). After exposure to bacterial ligands, immature LCs became more prone to HIV infection compared to unstimulated immature LCs (20**). Overall, the presence of non-optimal bacteria in the vagina may influence the behavior of host immune cells to increase HIV transmission across the vaginal barrier and presentation to T-cells.

Vaginal Metabolites, Inflammation, and Vaginal Barrier Integrity

Another key component of vaginal immunity are the metabolites, or small molecules in the vagina, produced by bacterial and host cells. The importance of vaginal metabolites for vaginal barrier integrity is supported by the experimental work of Schwecht and colleagues (22*), who treated VK2 vaginal epithelial cells with varying short chain fatty acid (SCFA) and lactic acid combinations designed to mimic Lactobacillus-dominant (“eubiotic”) and non-Lactobacillus-dominant (“dysbiotic”) vaginal microbiome conditions. Treatment with eubiotic metabolites improved epithelial barrier integrity, while a dysbiotic treatment reduced barrier integrity (22*). The eubiotic metabolites protected against barrier function damage after exposure to HIV-1, whereas the dysbiotic metabolites, especially SFCAs, increased HIV-leakage and were associated with poorer epithelial barrier integrity in the context of HIV-1 (22*). Cells treated with eubiotic metabolites had higher expression of tight junction proteins, desmosomal proteins, and adherens junction proteins compared to cells treated with dysbiotic levels of metabolites (22*). Cells treated with the dysbiotic metabolites showed nuclear translocation of NFkB, which has been associated with pro-inflammatory immune responses, and upregulated expression of genes for cytokines involved in barrier function and inflammation (22*). Greater lactic acid was protective at dysbiotic levels of SCFA with respect to suppressing NFkB activation, protecting against damage to epithelial integrity, and decreasing TNF-alpha production (22*). Using hydrochloric acid to lower pH caused more damage to the epithelial cells, highlighting the unique benefits of lactic acid. Findings suggest the need for research on the benefits of lactic acid and SCFAs for promoting epithelial barrier integrity and reducing HIV-1 leakage.

Mucosal Antibodies, Epithelial Barrier Cells, and Tissue-Adherent Microbiome

A key site for HIV prevention is the mucosal and epithelial barrier. Schaefer and colleagues (23*) compared the ability for IgG broadly neutralizing antibodies to trap HIV when colonized with different bacteria in cervicovaginal mucus samples collected from 17 healthy women. HIV trapping was measured as the ability for broadly neutralizing IgG antibodies to limit free diffusion of HIV-like particles in cervicovaginal mucus (23*). Broadly neutralizing IgG antibodies were able to trap HIV effectively in mucus samples from 10 of 17 women (23*). The efficacy of HIV trapping was positively associated with L. crispatus abundance and negatively associated with L. Iners and G. Vaginalis abundance (23*). This work provides evidence that the vaginal microbiome is associated with the effectiveness of IgG antibodies and/or mucin trapping abilities overall (23*). The influence of BV-associated bacteria on antibody effectiveness may have implications for HIV vaccines designed to increase broadly neutralizing antibodies.

In mammals, epithelial cells are coated with a glycan coat (i.e., glycocalyx), which provides a layer of protection from pathogens. Agarwal and colleagues (24*) sought to compare the vaginal epithelial glycocalyx among women with and without BV and examine the impact of Gardnerella-produced sialidases (which removes sialic acids from glycoproteins) on epithelial glycans and cell gene expression. The function of sialic acid is to cap the terminal ends of galactose residues and protect the underlying carbohydrates from being degraded. Women without BV typically had a visible glycan coat on vaginal epithelial cells, with sialic acids present (24*). In comparison, women with BV had a diminished visible glycocalyx, higher endogenous sialidase activity, lower sialic acid in the glycocalyx, lower levels of sialylated epithelial N-glycans and O-glycans, a greater portion of exposed galactose residues on epithelial cells, and large numbers of bacteria on cell surfaces (24*). When normal N- and O-glycans were exposed to a commercially prepared exogenous sialidase, they became more similar to N- and O-glycans seen in BV, with fewer sialylated peaks (24*). Treatment with Gardnerella-produced sialidases led to degraded vaginal epithelial glycans and a diminished glycocalyx (24). Treatment of vaginal cells with Gardnerella sialidase was associated with altered gene transcription, including enrichment of genes in pathways associated with cell death and immune responses (24*). This study provides evidence that BV-associated bacteria and their metabolites have the potential to deplete the epithelial glycan coat and alter related cellular and immune processes (24*).

Edfeltdt and colleagues (25*) compared the microbiota adhered to ectocervical tissue, collected via tissue biopsies, to the luminal microbiota, collected via cervicovaginal lavage, among sex workers in Kenya. The luminal and cervicovaginal tissue-adherent microbiomes showed multiple distinctions, with the most common genus in the luminal microbiome being Lactobacillus and the most common genus in the tissue-adherent microbiome dominant being Gardnerella, which was present in tissue samples of nearly all participants (25*). About 65% of the sample had similar dominant genera in both samples. However, with the exception of participants with a L. crispatus dominant luminal microbiome, all other groups exhibited higher alpha diversity in tissue compared to luminal samples (25*). The authors suggest the presence of Gardnerella and other BV-associated bacteria in tissue may contribute to the reseeding of the luminal microbiome with BV-associated bacteria after treatment, partially explaining high BV recurrence rates; however, research is needed to confirm this conclusion (25*).

Epigenetics

The genetic expression of host cells can be altered due to host-microbe interactions, which could potentially impact processes related to HIV immunity. For example, host cells exchange non-coding micro-RNAs (miRNAs) that downregulate post-transcriptional gene expression by encasing miRNAs in extracellular vesicles (26*). This allows for the measurement of extracellular vesicle-contained miRNAs (EV-miRNAs) to assess the association of the microbiome with the miRNA transcriptome (26*). Cezar-de-Mello and colleagues (26*) used an in vitro human vaginal colonization model to assess EV-miRNAs in response to colonization with L. crispatus, the parasitic infection Trichomonas vaginalis, the BV-associated bacteria Prevotella bivia, or a non-colonized control sample. Colonization with Trichomonas vaginalis and P. bivia led to substantial alterations in EV-miRNA cargo, whereas L. crispatus maintained miRNA homeostasis (26*). There were overlaps in the transcriptome alterations between Trichomonas vaginalis and P. bivia, including among multiple pathways affiliated with steroid hormone receptor functioning in the reproductive tract (26*). Theoretically, decreases in estrogen receptor 1 transcription could explain elevated inflammation in the vagina in response to colonization by pathogens, but this needs to be confirmed (26*). This study provides evidence that pathogenic parasites and bacteria in the vagina could impact host epigenetic mechanisms (26*).

Edfeldt and colleague’s (25*) study comparing the tissue-adherent versus luminal vaginal microbiome provides further evidence for an association between the microbiome and host gene transcription. Women with a highly diverse luminal microbiota showed increased transcription of genes associated with innate immunity, epithelial remodeling, glucose catabolism, estrogen transcription factors, and membrane budding compared to Lactobacillus-dominant microbiomes (25*). BV-associated bacteria were associated with alterations in host cell arachidonic and linoleic acid metabolism, while other pathogenic bacteria were associated with immune activation pathways (25*).

Multi-Omics Interactions

Multi-omics approaches can connect the multiple interacting mechanisms between bacterial communities, vaginal metabolites, immune cell function, and epithelial barrier integrity examined in experimental research. Using a multi-omics approach, Farr Zuend and colleagues (27*) examined the relationship of cervicovaginal inflammation with the vaginal microbiome, metabolites, immune cell phenotypes, the mucosal proteome, and functional proteomic and microbiome pathways among 42 healthy women in Canada. Higher vaginal inflammation was associated with lower abundance of L. crispatus and a reduction in microbiome functions associated with carbohydrate metabolism and lactic acid, as well as elevated vaginal pH, increases in the metabolite xanthine, and higher levels of endocervical antigen-presenting cells (27*). Higher vaginal inflammation was also associated with upregulation of mucosal protein pathways affiliated with innate immunity, neutrophil degranulation, integrin signaling, complement/coagulation cascades, and leukocyte migration, as well as downregulation of pathways associated with the formation of the cornified envelope, cell-cell adherens junctions, and keratinization (27*). A Bayesian network analysis showed multiple interactions between the microbiome, bacterial functions, antigen-presenting cells, epithelial barrier dysfunction, neutrophil activation, and metabolites succinate and xanthine. The authors suggest succinate may link the vaginal microbiome to antigen-presenting cell activity and inflammation (27*). Future research would benefit from examining whether metabolites such as succinate are a modifiable area for intervention to reduce loss of epithelial barrier integrity and immune activation.

Berard and colleagues (28*) conducted a multi-omics study combining in vivo and in vitro methods to assess the host and bacterial proteomic, metabolomic, transcriptomic, and immune mechanisms driving vaginal epithelial barrier dysfunction among 405 HIV-negative women Uganda and Kenya in a pre-exposure prophylaxis (PrEP) clinical trial. Multi-omics pathways and meta-models were constructed to describe functional relationships between host and bacterial factors. Gardnerella, M. mulieris, and their metabolites (e.g., imidazole propionate) were associated with epithelial barrier dysfunction and mucosal inflammation, which was likely driven in part by the mammalian target of rapamycin (mTOR) pathway (28*). mTOR is a human enzyme that regulates widespread physiological activities, including those related to immunity and cellular proliferation, growth, motility, and survival (28*). Variance was found between individuals, which suggests there are host factors (e.g., frequency of unprotected intercourse) that modify the effects of bacteria on vaginal barrier disruption (28*). These host factors require further investigation to identify characteristics that place individuals at higher risk for barrier dysfunction.

A Lack of Solutions

Women-centric methods to improve vaginal immunity are critical for ending the HIV epidemic; yet, solutions to limit vaginal HIV transmission are lacking (1,2,29). The only two forms of HIV PrEP approved for cisgender women -- Truvada and Apretude -- show insufficient uptake (30). The efficacy of oral PrEP among women is dependent on a higher level of medication adherence compared to men, although injectable PrEP may reduce daily barriers to adherence (31). Furthermore, trials of intravaginal HIV PrEP solutions have extremely low efficacy, and studies have demonstrated that this is likely, at least in part, due to direct metabolism by BV-associated bacteria in women with BV (13,14).

There are four US Federal Drug Administration (FDA)-approved antibiotics for BV: vaginal Clindamycin Phosphate, vaginal Metronidazole, oral Secnidazole, and oral Tinidazole (32). Oral Clindamycin hydrochloride is FDA-approved for anaerobic vaginal infections, and the CDC recommends oral Metronidazole for symptomatic BV (33). Although these treatments are superior to placebo, substantial BV persistence and reoccurrence demonstrate their limited benefits (32,34–36). In a meta-analysis of BV treatment studies, clinical cure rates ranged from 47% to 96%, with a pooled cure rate of 76% (36). Similar cure rates of 48% to 85% have been reported in studies of pregnant women (35). There is limited research on treating asymptomatic BV to reduce HIV acquisition.

HIV-prevention interventions should be woman-controlled, culturally acceptable, and effective for women with diverse vaginal microbiomes (14,29). Promising areas for future research include novel strategies for vaginal product delivery, such as hydrogels that avoid harming vaginal cells or Lactobacillus (37); multi-purpose treatments for HIV, BV, and pregnancy prevention (38); non-invasive biomarkers for epithelial disruption, such as soluble E-cadherin (39); and the use of community-based research to collect data from large, global samples of women (40). Research is needed to understand how social and environmental conditions, mental health, sexual behaviors, and intravaginal practices influence the vaginal microbiome and its’ association with HIV transmission (41).

Summary/Conclusion

A non-optimal vaginal microbiome characterized by less Lactobacillus dominance, fewer lactic-acid producing bacteria, and greater diversity and abundance of BV-associated bacteria may elevate the risk of HIV acquisition by increasing vaginal pH, damaging the epithelial barrier, promoting HIV-leakage across epithelial cells, reducing the efficacy of mucosal antibodies in trapping HIV, and increasing immune cell presentation of HIV to CD4+ T-Cells. More effective woman-controlled solutions are crucial to support vaginal health by promoting the vaginal microbiome and inducing mucosal immune mechanisms that are more resilient against HIV infection.

Key Points.

• A non-optimal vaginal microbiome can increase risk of HIV acquisition, yet few efficacious treatments currently exist and are critically needed.

• The impact of vaginal bacteria and metabolites on immune cell function, epithelial barrier integrity, and mucosal immunology presents areas for potential intervention.

• Research is still needed to develop women-controlled products to improve women’s health and wellbeing, and enhance vaginal defenses against HIV.