No abstract

Fish genetically selected for growth (GS) and reference (REF) fish were fed with CTRL (15% FM, 5–7% FO) or FUTURE (7.5% FM, 10% poultry meal, 2.2% poultry oil + 2.5% DHA-algae oil) diets during a 12-months production cycle. Samples from initial (t0; November 2019), intermediate (t1; July 2020) and final (t2; November 2020) sampling points were used for Illumina 16S rRNA gene amplicon sequencing of the adherent microbiota of anterior intestine (AI). Samples from the same individuals (t1) were also used for the gene expression profiling of AI by RNA-seq, and subsequent correlation analyses with microbiota abundances. Discriminant analyses indicated the gut bacterial succession along the production cycle with the proliferation of some valuable taxa for facing seasonality and different developmental stages. An effect of genetic background was evidenced along time, decreasing through the progression of the trial, namely the gut microbiota of GS fish was less influenced by changes in diet composition. At the same time, these fish showed wider transcriptomic landmarks in the AI to cope with these changes. Our results highlighted an enhanced intestinal sphingolipid and phospholipid metabolism, epithelial turnover and intestinal motility in GS fish, which would favour their improved performance despite the lack of association with changes in gut microbiota composition. Furthermore, in GS fish, correlation analyses supported the involvement of different taxa with the down-regulated expression of pro-inflammatory markers and the boosting of markers of extracellular remodelling and response to bacterium. Altogether, these findings support the combined action of the gut microbiome and host transcriptionally mediated effects to preserve and improve gut health and function in a scenario of different growth performance and potentiality.

No abstract

Like in other animals, the gut microbiome of fishes contains thousands of microbial species that establish a complex network of relationships among each other and with the host. These interrelationships are shaped by biotic and abiotic factors, but little is known about how they evolved and how they are regulated by the environment, farmers and breeders. Herein, we introduce SAMBA (Structure-Learning of Aquaculture Microbiomes Using a Bayesian-Network Approach), the software implementation of a Bayesian network model for investigating how fish pan-microbiomes and all other variables of a given aquaculture system are related each to other. SAMBA is powered by a Bayesian network trainable model that learns the network structure of an aquaculture system using information from distinct biotic and abiotic variables of importance in fish farming, with special focus on microbial data provided from 16S amplicon sequencing. SAMBA accepts both qualitative and quantitative variables and convincingly deals with the differences in microbial composition derived by the technical or biological variation among microbiomes of distinct specimens. To this end, SAMBA is implemented with a variety of tools to pre-analyze the data and chose a distribution (log-normal, binomial, etc.) to build and train the Bayesian network model. Once the model has been created and validated, the user can interrogate the model and obtain information about the modelled system in two different modes: Report and Prediction. Using the Report mode SAMBA reports how the pan-microbiome and all other variables involved in the modelled aquaculture system influence each other and what are the conditional probabilities of each relation. Under the Prediction mode, the application predicts how the diversity and functional profile of the pan-microbiome would likely change depending on any change made on other variables. Finally, SAMBA implements a comprehensive graphical network editor allowing to navigate, edit and export outcomes. We tested and validated the performance of SAMBA using microbial community standards and gut microbiota communities of farmed gilthead sea bream (Sparus aurata) from different feeding trials.