Microbial communities define and are defined by their environment. Credit: Marina Corral Spence/Springer Nature

The science of metagenomics has helped researchers characterize microbes as communities, but we are only beginning to understand the complexities of interactions within these communities and with the environment. The rhizosphere of a single plant root includes commensal and pathogenic bacteria that interact with each other as well as with the plant, soil and fungi. Tumorigenesis has been linked to microbiome-induced inflammation in the gut. Complex interactions such as these clearly have important consequences for agriculture as well as disease and health, and methods are needed for deeper exploration.

A key to understanding microbial ecology and host–microbe interactions will be to develop controlled experimental platforms. Gnotobiotic mice provide a blank slate that can be colonized by different gut bacteria and exposed to various diets for comparative studies. For human environments, organ-on-chip technologies that mimic epithelia and accommodate bacterial culture can also help assess microbial interactions. Combinatorial testing of experimental conditions, such as finding which bacteria need to be cocultured in order to grow, can also help to untangle relationships between individual components.

Sequencing and other omic technologies are effective ways to track microbial composition and metabolic activity, and they allow correlations to be made in the context of well-controlled studies. Sequencing RNA from both prokaryotic and eukaryotic cells simultaneously—in the case of intracellular parasites, for example (Nature 529, 496–501, 2016)—can reveal how host and microbe interact at the level of gene expression. Computational modeling and analysis tools need to be developed to tease out environmental correlations and to understand microbial dependencies and coevolution.

Other in situ methods can also help to capture microbial interactions. Techniques for quantitative imaging of labeled bacteria and their surroundings (Cell Host Microbe 18, 478–488, 2015), including fluorescence in situ hybridization labeling of bacteria and noninvasive imaging of extracellular milieu components, add a critical spatial dimension to microbial studies. Metabolic labeling and other methods that can track microbial activity will likewise provide benefits.

New approaches to understanding microbial interactions should help to solve longstanding and emerging questions, such as how microbiomes can protect against pathogenic bacteria.