The vaginal microbiota through the lens of systems biology

The human body is a complex ecosystem of co-existing microbes, including those in the gut, skin and vagina of women. They play an important role in health and disease. However, much remains to be learned about them.

A new paper recently published online in Trends in microbiology Journal reviews the systems biology approach to studying the vaginal microbiome (VMB), which helps to understand its composition and function and how it interacts with the host.

Review: New perspectives on the vaginal microbiota through systems biology. Image credit: Design_Cells / Shutterstock

Introduction

The VMB is important for female fertility, and disruptions can be associated with pregnancy disorders, gynecological conditions such as pelvic inflammatory disease (PID), and a number of infections related to the female reproductive system. In addition, VMB can affect the effectiveness of drugs in women.

However, VMB is little understood beyond a vague concept as an overarching one Lactobacillus is associated with a “good” state with a homogenous social structure. Conversely, the undesirable state of VMB exists when more diverse species are identified in higher abundance.

This latter undesirable condition is often associated with bacterial vaginosis (BV), which occurs in one in three women during their reproductive years, which can have serious consequences on their fertility. As such, research in this area is necessary to understand the direction and extent of such associations.

The problem

Although many studies have been conducted in this area, it is difficult to understand what the optimal VMB looks like due to the complex interactions between microbes and other host factors. This means that a healthy VMB can vary considerably between women and at different points in the same person’s life cycle.

Such changes occur within days, in contrast to the much slower change seen with gut, skin and oral microbes, which can change over months or even years. Unfortunately, this makes cross-sectional data rather inconclusive when it comes to investigating the relationship between VMB composition, function, and disease—thus making most of this data less useful than it could be.

Again, human VMB differs significantly from animal, as well as culture-based models. In the former, even non-human primates manage to exhibit characteristic human vaginal conditions, including an acidic pH and Lactobacillus control

In the latter, some microorganisms are remarkably resistant to culture in vitro, but various culture conditions are used in different laboratories, depending on the media. This could make the growth environment quite different from that of the human cervix and vagina, invalidating the results of such experiments.

As such, clinical specimens in which vaginal microflora are cultured, identified and quantified are the primary source of information on human VMB. These data are colored by experimental and host variables, which require sophisticated statistical adjustments to achieve a valid result.

Although it is appropriate for all microbial sites, [this] is particularly relevant for VMB due to the lack of experimental models that allow studying the vaginal microbiota under controlled conditions.”

The solution

Such conundrums can be resolved by a systems biology approach, where quantitative analyzes are used to extract important factors that influence the behavior and function of a microbial community. As such,Utilizing systems biology techniques applied to other microbes, as well as developing new strategies and applying these strategies to VMB, will have a significant impact on improving women’s health.”

The application of systems biology can overcome the challenges of such complex and multiple external and internal interacting networks. Furthermore, many methods can be used, depending on the type of information available and the objective of the study.

Thus, statistical or data-driven methods are ideal when high throughput is available in a relatively new field of study. This can help indicate which microbial profiles are associated with disease or health. As little is known about the VMB, data-driven models have dominated so far.

In contrast, hypothesis-driven mechanistic approaches are better when much is already known about a system, or at least the basic data are available, and the need is to understand the mechanisms of cause-effect relationships that underlie biological activity. In addition, they help set the stage in which microbial assemblages and interactions can occur under normal and abnormal conditions.

Some mechanistic approaches include mass kinetic or population dynamics models (based on differential equations), genome-scale metabolic models (GEMs), and biochemical models (ABMs).

What has been achieved?

The systems biology approach has already helped to identify and classify community state types (CSTs) associated with health, disease or transitions between the two. They were first defined by microbial abundance and incorporated patient demographic and health data to form hierarchical groups. In addition, other methods such as clustering of nearest centroids have been developed to overcome the inherent variation in the dataset with the previous approach.

CST groups help to simplify VMB composition, thereby pointing out relationships with community composition and functioning. But this comes at the cost of overlooking community-specific factors specific to different categories.

Multi-omics approaches could be integrated with systems biology approaches to identify relationships with different types of communities and specific metabolomics, transcriptomics and metagenomics profiles, for example. In addition, random forest models and other advanced machine learning models are being pushed to help identify VMBs dominated by different microbes, such as L. crispatus against. L. iners or Bifidobacteriaceae.

Interestingly, neural network modeling has demonstrated the superiority of metabolomics in accurately describing the cervical environment compared to either VMB composition or immunoproteomics. Integrated use of these methods could help select important drivers of VMB states in health and disease.

Of particular importance may be the insights gained regarding the risk of sexually transmitted infections (STIs) with increased levels of “bad” microbes. For example, an increase in L. iners appears to be associated with a higher risk of STDs, meanwhile L. gasseri related to health. On the other hand, Gardnerella vaginalis and Prevotella species are associated with chlamydial infection.

Mechanistic models include the technique called MIMOSA (Model-based Integration of Metabolite Observations and Species Abundances) which uses metabolic system models to understand the dynamics of the community through its gene content. This helped identify Prevotella species and Atopobium vagina as key regulators of VMB, using a calculated community-based metabolite capacity (CMP) score. The CMP shows the turnover of each metabolite in each community.

Similarly, genome-scale reconstructions (GENREs) could help to understand the role of challenging microbes in VMB. Ordinary differential equation (ODE) modeling is used to explore how drugs can affect the VMB and the ecology of this system, showing how the composition fluctuates with exposure to different factors.

What lies in the future?

Much research has focused on the gut microbiome, with nearly $150 million poured into developing and standardizing new tools for its exploration. VMB researchers could potentially use this to serve their goals. This includes BURRITO, a web tool that helps visualize the microbial community with relative abundance. This could be extended to look at VMB metagenomics, showing how patient characteristics relate to CSTs.

Supervised machine learning approaches to better understand VMB include Data Integration Analysis for Biomarker Discovery using Latent Components (DIABLO), where omics datasets are integrated by correlation, and Sparse regularized generalized canonical correlation analysis (SRGCCA), used in Crohn’s disease.

To overcome the limitations imposed by the lack of knowledge of the functional classification of VMB, unsupervised learning methods may be useful, such as multi-omic factor analysis (MOFA).

Multiple ODE models based on the Generalized Lotka–Volterra (gLV) models can also be used. These include web-gLV, the Microbial System Inference Engine for Microbial Time Series Analysis (MDSINE), and the Learning Interactions from Microbial Time Series (LIMITS) method, as well as new adaptations such as Lotka–Volterra combinations (cLV) and the “Biomass Estimation and Inference Model with Expectation Maximization” algorithm ( BEEM), which do not depend on community cultivability or the availability of extensive longitudinal datasets.

Newer methods include algorithms such as the Constant yield expectation framework (conYE) and MMinte, which simulate the conditions for metabolism and community growth based on the dense interactions of the species. Such novel adaptations and approaches could help to understand the factors that shape the dynamic VMB in health and disease in different populations.

Leave a Comment

Your email address will not be published. Required fields are marked *