A growing body of evidence has highlighted the influence of the gut microbial community on host health, with dysbiosis being implicated in diseases as diverse as autism, obesity and colitis. Characterisation of the mechanisms through which the gut microbiota and the hot interact is required to build an understanding of how to effectively manage and manipulate host health. This work employs computational simulation to elucidate the influence of diet composition and feeding regimen on gut bacterial community composition.
This modelling work builds on a real mouse study in which 250 mice were administered one of 30 diets. Each diet consists of different ratios of carbohydrate, protein and fat, and in addition, varying quantities of cellulose that adjust the host-accessible nutrient density of the diet. We simulate each mouse in this study, and assess how the diet composition and varying lengths and regularities of feeding versus diet periods drive bacterial community composition. Our simulation explicitly represents individual bacteria as belonging to one of five functional guilds, defined by their preferential access to carbon and nitrogen from a subset of feed-derived protein, cornstarch, dyetrose, or intestinal mucin. We simulate heterogeneous bacterial communities, where individual bacteria compete for limited nutrient resources, have varying growth and death rates, and undergo a stress response in response to periods of nutrient scarcity.
Through simulation we can predict how different diets, the periods at which food is eaten, and periods of fasting can influence bacterial community composition. We find that periodic caloric restriction drives a greater relative abundance of bacterial guilds that preferentially access carbon and nitrogen from mucin, and that the benefit to these guilds varies with host diet composition. We can assess the quantities and times at which fibre supplements can be administered to achieve optimal effect in manipulating bacterial community composition.