The role of the microbiome in the neurobiology of social behaviour.

Microbes colonise all multicellular life, and the gut microbiome has been shown to influence a range of host physiological and behavioural phenotypes. One of the most intriguing and least understood of these influences lies in the domain of the microbiome's interactions with host social behaviour, with new evidence revealing that the gut microbiome makes important contributions to animal sociality. However, little is known about the biological processes through which the microbiome might influence host social behaviour. Here, we synthesise evidence of the gut microbiome's interactions with various aspects of host sociality, including sociability, social cognition, social stress, and autism. We discuss evidence of microbial associations with the most likely physiological mediators of animal social interaction. These include the structure and function of regions of the 'social' brain (the amygdala, the prefrontal cortex, and the hippocampus) and the regulation of 'social' signalling molecules (glucocorticoids including corticosterone and cortisol, sex hormones including testosterone, oestrogens, and progestogens, neuropeptide hormones such as oxytocin and arginine vasopressin, and monoamine neurotransmitters such as serotonin and dopamine). We also discuss microbiome-associated host genetic and epigenetic processes relevant to social behaviour. We then review research on microbial interactions with olfaction in insects and mammals, which contribute to social signalling and communication. Following these discussions, we examine evidence of microbial associations with emotion and social behaviour in humans, focussing on psychobiotic studies, microbe-depression correlations, early human development, autism, and issues of statistical power, replication, and causality. We analyse how the putative physiological mediators of the microbiome-sociality connection may be investigated, and discuss issues relating to the interpretation of results. We also suggest that other candidate molecules should be studied, insofar as they exert effects on social behaviour and are known to interact with the microbiome. Finally, we consider different models of the sequence of microbial effects on host physiological development, and how these may contribute to host social behaviour.