Responses of rumen bacterial microflora to aflatoxin B1 challenge: potential roles of AHLs/AI-2 mediated quorum sensing

Background
As a universal cell-to-cell communication mode in the diverse microbial communities, quorum sensing (QS) has been discovered in the ruminal microbiome and supposed to play a significant role in feed digestion and microorganism adaptation during rumen fermentation. Besides, the mycotoxin aflatoxin B1 (AFB1) has been reported to cause rumen dysbacteriosis and alter the abundance of some specific ruminal bacteria that possess the QS-related genes. Threfore, we hypothesized that QS is involved in the responses of rumen bacterial microbiota to AFB1 exposure, and employed the rumen simulation technique (RUSITEC) to investigate the potential roles of QS in the AFB1-induced shifts in ruminal microbiota and relevant metabolisms.

Results
AFB1 raised ruminal pH and NH3-N concentration, but decreased total volatile fatty acid (TVFA) production, the molar proportions of acetate, propionate, butyrate, isobutyrate, valerate, isovalerate, and the disappearance of acid detergent fiber (ADF). The concentrations of QS signaling molecules including AI-2, C4-HSL, C6-HSL, and 3-oxo-C6-HSL were reduced by AFB1, whereas the microbial protein and lipopolysaccharide (LPS) were unchanged. The copy numbers of Ruminococcus flavefaciens, methanogens, and fungi dropped, while the density of Selenomonas ruminantium rose in response to AFB1. LEfSe analysis showed that the species of unclassified_Prevotella, UCG_004_bacterium_YGB2007, Cenchrus americanus, uncultured_rumen_bacterium affiliated to the class Kiritimatiella, Butyrivibrio fibrisolvens, and Anaerovibrio_sp_JC8 were differentially enriched in the rumen liquid of different treatments, whilst another uncultured_rumen_bacterium was the only distinctively enriched species in the rumen solid. Further, the metabolomics analysis indicated that ascorbate and aldarate metabolism, histidine metabolism, sphingolipid metabolism, pyruvate metabolism, glycine, serine and threonine metabolism, biosynthesis of unsaturated fatty acids, and arginine and proline metabolism were the commonly differential metabolic pathways. As illustrated by the Spearman analysis, the AI-2 and C6-HSL were positively correlated with the volatile fatty acids (VFAs), and C6-HSL was positively correlated with the copy numbers of Selenomonas ruminantium and Ruminococcus flavefaciens. Moreover, the positive or negative interconnections between the QS signal molecules and the relative abundances of differential bacterial taxa were also exhibited.

Conclusions
The present RUSITEC fermentation study revealed the potential involvements of AHLs/AI-2 mediated QS in the fluctuations of ruminal microbes and relevant rumen metabolisms under the AFB1 challenge, which could lead to the overall impairment of ruminal microbial fermentation. Future studies are necessitated to elucidate and testify the mechanisms of those correlations discovered in the current trial.