Phage as signatures of healthy microbiomes

Abstract

Parasites are foundational to ecosystem health both as indicator species of community productivity but also as drivers of diversity. In bacterial communities, bacteriophage viruses can play such a role as they track the dynamic composition of bacterial hosts, and in the case of lytic phages, confer a growth advantage to lower abundance bacteria while adapting to more common ones. We set out to test whether viromes can be used as signatures of microbiome health using previously published results across systems. By comparing observed phage and bacterial diversity between microbiomes characterized by disturbance (so-called dysbiosis) and those considered control populations, we were able to identify some key commonalities. While just under half of studies report significant changes in viral species richness in dysbiosis, just under two thirds of studies find the viral composition to shift in dysbiosis, with specific viral taxa enrichment acting as a common signature of dysbiosis. Our analyses also suggest that the positive relationship between bacteriome and virome alpha diversity observed in health breaks down under microbiome disturbance. Overall, while specific viral signatures of dysbiosis are likely to be highly disease- and condition-specific, existing ecological theory shows clear promise in predicting and explaining microbiome health. Future data on bacteria-phage diversity relationships may provide us with much needed opportunity to diagnose, treat, and better understand the causes of dysbiosis. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSBeing able to identify signatures of microbiome health (or lack thereof) has the potential to improve the way we diagnose and treat disease. To do this, the bacterial microbiome is traditionally characterised at the 16S taxonomic level, and changes in composition are linked to changes in disease status. More recently, the field of viromics has gained attention, and studies have begun to probe the relationship between the virome and health or disturbance ( dysbiosis). This work has focused to date on finding single phages that indicate presence of known pathogens, or in a few cases the relationship between viral diversity and disease. To our knowledge, no work has yet sought to identify a common signature of dysbiosis or find commonalities across systems that suggest a role for phages in dysbiosis. Decades of ecological theory has shown how parasites can shape the ecology and evolution of their hosts, and here we argue that bacteriophage viruses have the potential to shape these same processes within microbial communities. The motivation for the current work was thus to ask whether existing ecological theory could help us identify viral signatures of dysbiosis in the microbiome. Added value of this studyThis study employed a systematic review and meta-analysis to test whether and when phage communities can be used as signatures of microbiome health. To do this, we synthesized previously published results that measure composition of the virome between bacterial microbiomes characterised by health or dysbiosis. We found a total of 39 studies across human, mouse, pig and cow hosts that spanned a diverse spectrum of dysbioses, including bacterial infections, viral infections, and varied diseases such as cancer, cirrhosis, and inflammatory bowel disease, and identified a number of commonalities. Just under half of these studies reported a significant change in viral species richness in dysbiosis, and just under two thirds reported the viral composition to shift in dysbiosis. While the vast majority of studies report an enrichment of specific viral taxa associated with dysbiosis, there is little overlap among studies regarding the identity of these enriched taxa. Finally, our analysis provides evidence that the positive relationship between bacteriome and virome alpha diversity breaks down in dysbiosis. Implications of all available evidenceSynthesis of the available evidence suggests that while looking for specific viral taxa as signatures may be limited to associations that are highly disease or condition specific, there is promise for the use of existing ecological theory in predicting and explaining microbiome health when considering compositional changes in the virome. Prospective studies should look to expand the data we have on bacteria-phage relationships at the level of species richness and community compositions, and we argue that more routinely investigating the virome or phageome, in addition to collecting 16S taxonomic descriptions of the microbial community, would help improve our ability to identify signatures of microbiome health. These viral signatures may offer early warning signs of microbiome disturbance and disease. This has clear relevance to our ability to diagnose, treat, and understand the underlying causes of disease.