Understanding how bacterial colonisation antagonizes a pathogenic fungus: metabolites matter
A healthy microbiota can protect us against several diseases. Here, we uncover how colonization by a bacterial member of the microbiota antagonizes the pathogenic yeast Candida albicans.
Candida albicans is probably not the fastest or the strongest of our enemies. This fungus lives in all of us, in several parts of our bodies, mostly as a harmless commensal. However, when there are no defenders to be seen, C. albicans can seize the opportunity and become pathogenic, potentially posing a serious threat to our health. In our group, we want to understand not only what makes this fungus become pathogenic, but also what makes it stay commensal.
We know that this commensal vs. pathogenic dilemma depends on a delicate balance. Several players are crucial: the immune system, the host microbiota, and the epithelial barriers are key to keep C. albicans infections at bay, and our goal is to understand how they do it. We focused on the microbiota: hundreds of studies have already highlighted the importance of a balanced and healthy microbiota for the prevention of diseases. It is well known that some bacterial species that are normal inhabitants of the human body, particularly lactobacilli, can antagonize C. albicans, but how exactly they do it in the human host was still a mystery.
Therefore, we investigated Lactobacillus rhamnosus as a representative of the beneficial microbiota to see how a functional microbiota antagonizes C. albicans pathogenicity. With our first experiments, we found that L. rhamnosus reduced C. albicans growth on epithelial cells, but also made the remaining fungal cells less pathogenic somehow. We also observed that the bacteria need to be alive in order to do their job but, opposite of what we thought before, they do not need to be in direct contact with either the epithelial cells or the fungus. However, the presence of epithelial cells was still somehow required. This suggested that the effect was mediated by the specific environment. Therefore, we took a metabolomics approach to characterize the extracellular environment upon L. rhamnosus colonization.
By analyzing the metabolites in the supernatants of the different conditions (epithelial cells, bacteria, and fungi alone, in pairs, and in the three-way combination) we identified not only which metabolites are changing, but also which of the interaction partners were responsible for the changes. This helped us to start addressing two relevant questions.
Our first question was: how do bacteria grow on epithelial cells? Many bacterial species love cell culture medium and will use any chance to contaminate our flasks, but this is not the case for L. rhamnosus. These bacteria do not grow in cell culture medium, yet they grow perfectly well on epithelial cells. We first used in silico modelling based on the metabolic profiles of the culture supernatants. The models predicted that the metabolic changes that the epithelial cells induced in the medium promoted bacterial biomass production. Based on this, we screened metabolites that were secreted by the epithelial cells but disappeared when L. rhamnosus grew in the model. Using this approach, we confirmed that the epithelial cells provide the bacteria with specific metabolites that they require to grow and to colonize, something essential to protect the host.
Our second question proved to have more answers than we originally thought: how do the bacteria control C. albicans? At first, we took a look at a group of metabolites that were specifically present when the bacteria colonized the host. Several of these metabolites actually suppressed the filamentous growth of C. albicans, a crucial aspect of its virulence. Interestingly, these compounds seemed to synergize with each other. When tested in combination they were far more effective as compared to the individual compounds. In addition, we also noticed that L. rhamnosus were strong competitors for the favoured diet of C. albicans. The bacteria rapidly consumed all the carbon and nitrogen sources that the fungus prefers and only left metabolites not favoured by C. albicans, like lactate (this is probably a familiar scenario for anyone who has siblings). With this, it appeared clear to us: the colonization of epithelial cells by L. rhamnosus gives C. albicans a hard time.
To understand how C. albicans adapted to this stressful environment, we investigated transcriptional alterations, and performed in silico genome scale metabolic modeling. As expected, we observed that the fungus was going through a lot. We observed drastic changes in its transcriptome, particularly the expression of metabolic genes. However, C. albicans seemed to be coping differently at early and late timepoints: at first trying to boost an alternative metabolism to compete, and later almost like giving up. The in silico modeling also confirmed that C. albicans was rewiring its entire metabolism in response to the adverse conditions.
A closer look at the C. albicans gene expression profiles revealed that metabolic changes had serious consequences. Many fungal genes regulate both metabolism and pathogenicity mechanisms, so the confrontation with this adverse environment seems to make C. albicans less dangerous.
We finally performed a screening in which we tested the damaging capacity of C. albicans mutants lacking the genes that were differentially regulated in the presence of bacteria. We identified several mutants that were more damaging than the wild type (hypervirulent) or less damaging (hypovirulent). Interestingly enough, hypovirulent mutants corresponded mostly to genes that were downregulated in the presence of bacteria. On the other hand, the majority of the hypervirulent mutants corresponded to genes upregulated in the presence of bacteria. This means that in response to the presence of L. rhamnosus, C. albicans adaptations are associated with downregulation of fungal virulence genes and upregulation of antivirulence genes, thus making C. albicans less aggressive.
With this, we have come full circle, understanding many things that happen when the host, a member of the microbiota and the pathogenic fungus interact. First, we have observed that the host, usually considered a passive element, is fostering the growth of beneficial members of the microbiota and promoting their colonization. Further, we have seen that, when lactobacilli are in an environment that allows them to thrive and proliferate, they will modify it, and this antagonizes C. albicans in several ways. They not only secrete antivirulence metabolites that reduce fungal pathogenicity, but compete with the fungus by consuming its preferred nutrients. C. albicans reacts by rewiring its metabolism to survive. This metabolic rewiring is, however, quite drastic, and not compatible with virulence. Therefore, C. albicans ends up back in its commensal state that it never would have escaped in the first place if the microbiota would have been balanced.
Understanding the mechanisms by which a healthy human microbiota keeps C. albicans in check as a harmless commensal helps us to better understand antibiotic therapy as a risk factor for Candida infections. Furthermore, the mechanistic insights can help us to reconsider and modify prophylactic probiotic treatments.
Read the paper here:
Alonso-Roman, R., Last, A., Mirhakkak, M.H. et al. Lactobacillus rhamnosus colonisation antagonizes Candida albicans by forcing metabolic adaptations that compromise pathogenicity. Nat Commun 13, 3192 (2022). https://doi.org/10.1038/s41467-022-30661-5