The discussion during the question-and-answer period focused on the relevance of using animal models to understand the human microbiome, the relative importance of understanding microbiome composition versus function, and the dietary implications of individual microbiome variation.
Animal Models and the Human Microbiome
Of the mouse model, Nicholson said during his presentation, “This is the first time we have actually had a tool which allows us to measure quantitatively the response of complex organisms to things like probiotic or prebiotic interventions.” During the question-and-answer period, several workshop participants asked questions or commented on the relevance of animal models to understanding the human microbiome. For example, there was a question about the implications of studying the impact of bariatric surgery on bile salt metabolism in rats, given that rats do not have a gall bladder and that human patients that undergo bariatric surgery have an increased incidence of gall bladder disease. Nicholson responded, “The important thing about these rat and mouse models is they help you to develop the tools for studying these complex interactions. We start to get a framework of what sort of pathways are interacting with what sort of bugs.” Nicholson and colleagues are currently finishing a study of 100 bariatric patients, the results of which may indicate whether the rat model is predictive of humans.
Another audience member asked whether there might be a better animal model than mice or rats for studying the human microbiome. Lita Proctor noted that although the HMP was not able to use animal models as a complement to any of the human studies as per NIH Common Fund rules, the door is now open to the development of new animal models. However, it is not clear whether and how the HMP will move in that direction. Meanwhile, the National Institute of General Medical Sciences has been very interested in developing animal models, including some nontraditional animal models (e.g., zebra fish), for microbiome research.
There was also some discussion about the pig model being used to study the human microbiome. Nicholson observed that because of the similarities between pig and human physiology, developing the pig model might be an “important direction” for future research. Regardless of the chosen animal model and regardless of whether the goal is to simulate human disease or a “normal” human microbiome, he emphasized that the metabolic profile generated should approximate a human metabolic profile.