SummaryGut microbiota typically occupy habitats with definable limits/borders that are comparable to oceanic islands. The gut therefore can be regarded as an 'island' for the assembly of microbial communities within the 'sea' of surrounding environments. This study aims to reveal the ecological mechanisms that govern microbiota in the fish gut 'island' ecosystem. Taxonomic compositions, phylogenetic diversity, and community turnover across host development were analyzed via the high-throughput sequencing of 16S rRNA gene amplicons. The results indicate that the Shannon diversity of gut microbiota in the three examined freshwater fish species all significantly decreased with host development, and the dominant bacterial taxa also changed significantly during host development. Null model and phylogenetic-based mean nearest taxon distance (MNTD) analyses suggest that host gut environmental filtering led to the assembly of microbial communities in the fish gut 'island'. However, the phylogenetic clustering of local communities and deterministic processes that governed community turnover became less distinct as the fish developed. The observed mechanisms that shaped fish gut microbiota seemed to be mainly shaped by the gut environment and by some other selective changes accompanying the host development process. These findings greatly enhance our understanding of stage-specific community assembly patterns in the fish gut ecosystem.

No abstract

Grass carp hemorrhagic disease, caused by the grass carp reovirus (GCRV), is a major disease that hampers the development of grass carp aquaculture in China. The mechanism underlying GCRV infection is still largely unknown. Circular RNAs (circRNAs) are important regulators involved in various biological processes. In the present study, grass carp were infected with GCRV, and spleen samples were collected at 0 (control), 1, 3, 5, and 7 days post-infection (dpi). Samples were used to construct and sequence circRNA libraries, and a total of 5052 circRNAs were identified before and after GCRV infection, of which 41 exhibited differential expression compared with controls. Many parental genes of the differentially expressed circRNAs are involved in metal ion binding, protein ubiquitination, enzyme activity, and nucleotide binding. Moreover, 72 binding miRNAs were predicted from the differentially expressed circRNAs, of which eight targeted genes were predicted to be involved in immune responses, blood coagulation, hemostasis, and complement and coagulation cascades. Upregulation of these genes may lead to endothelial and blood cell damage and hemorrhagic symptoms. Our results indicate that an mRNA–miRNA–circRNA network may be present in grass carp infected with GCRV, providing new insight into the mechanism underlying grass carp reovirus infection.