Molecular and culture-based assessment of the microbiome in a zebrafish (Danio rerio) housing system during set-up and equilibration.

Background: Zebrafish used in research settings are often housed in recirculating aquaculture systems (RAS) which rely on the system microbiome, typically enriched in a biofiltration substrate, to remove the harmful ammonia generated by fish via oxidation. Commercial RAS must be allowed to equilibrate following installation, before fish can be introduced. There is little information available regarding the bacterial community structure in commercial zebrafish housing systems, or the time-point at which the system or biofilter reaches a microbiological equilibrium in RAS in general.

Methods: A zebrafish housing system was monitored at multiple different system sites including tank water in six different tanks, pre- and post-particulate filter water, the fluidized bed biofilter substrate, post-carbon filter water, and water leaving the ultra-violet (UV) disinfection unit and entering the tanks. All of these samples were collected in quadruplicate, from prior to population of the system with zebrafish through 18 weeks post-population, and analyzed using both 16S rRNA amplicon sequencing and culture using multiple agars and annotation of isolates via matrix-assisted laser desorption/ionization-time-of-flight (MALDI-TOF) mass spectrometry. Sequencing data were analyzed using traditional methods, network analyses of longitudinal data, and integration of culture and sequence data.

Results: The water microbiome, dominated by Cutibacterium and Staphylococcus spp., reached a relatively stable richness and composition by approximately three to four weeks post-population, but continued to evolve in composition throughout the study duration. The microbiomes of the fluidized bed biofilter and water leaving the UV disinfection unit were distinct from water at all other sites. Core taxa detected using molecular methods comprised 36 amplicon sequence variants, 15 of which represented Proteobacteria including multiple members of the families Burkholderiaceae and Sphingomonadaceae. Culture-based screening yielded 36 distinct isolates, and showed moderate agreement with sequencing data.

Conclusions: The microbiome of commercial RAS used for research zebrafish reaches a relatively stable state by four weeks post-population and would be expected to be suitable for experimental use following that time-point.