Abstract
Advancing microbiome-gut-brain axis science requires systematic, rational and translational approaches to bridge the critical knowledge gaps currently preventing full exploitation of the gut microbiome as a tractable therapeutic target for gastrointestinal, mental and brain health. Current research is still marked by many open questions that undermine widespread application to humans. For example, the lack of mechanistic understanding of probiotic effects means it remains unclear why even apparently closely related strains exhibit different effects in vivo. For the therapeutic application of live microbial psychobiotics, consensus on their application as adjunct treatments to conventional neuromodulators, use in unmedicated populations or in at-risk cohorts with sub-clinical symptomatology is warranted. This missing information on both sides of the therapeutic equation when treating central nervous system (CNS) conditions makes psychobiotic research challenging, especially when compared to other pharmaceutical or functional food approaches. Expediting the transition from positive preclinical data to proven benefits in humans includes interpreting the promises and pitfalls of animal behavioural assays, as well as navigating mechanism-informed decision making to select the right microbe(s) for the job. In this review, we consider how these decisions can be supported in light of information accrued from a range of clinical studies across healthy, at-risk and pathological study populations, where specific strains have been evaluated in the context of gastrointestinal physiology, brain function and behaviour. Examples of successful, partial and unsuccessful translation from bench to bedside are considered. We also discuss the developments in in silico analyses that have enhanced our understanding of the gut microbiome and that have moved research towards pinpointing the host-microbe interactions most important for optimal gut-brain axis function. Combining this information with knowledge from functional assays across in vitro and ex vivo domains and incorporating model organisms can prime the discovery pipelines with the most promising and rationally selected psychobiotic candidates.