Dynamic host–bacteria interactions during an acidotic state induction

Abstract

Mainly due to the lack of time-series data, we know very little about the underlying interactions leading to adverse states in the gut. I therefore investigate the host-bacteria dynamic interactions in a recently published acidotic state induction time-series experiment. Too high levels of fermentation in the hindgut may lead to elevated serum lactate levels, which is a characteristic of the acidotic state. The acidotic state was induced through injection of oligofructose directly into the caecum of five horses, with subsequent temporal measurements of host serum lactate, and a range of caecum bacteria and metabolites. Principal component analysis (PCA) showed that equine hindgut streptococcal species (EHSS) showed the main positive correlation to caecum lactate, and negative correlation to acetate. By dynamic modelling I found that serum lactate interacted with several caecum components, with the main interactions being with caecum lactate and Enterobacteriaceae. Enterobacteriaceae showed a positive interaction with serum lactate, while the interaction between serum and caecum lactate was more complex. The main individual differences in horse serum lactate could be explained by the levels of Enterobacteriaceae and caecum lactate. In conclusion, the dynamic models revealed simple host-bacteria interactions that can explain changes and individual differences in serum lactate.