Agriculture Reference
In-Depth Information
taxa at different phylogenetic ranks. The utilization of NGS to investigate the diversity of the
bacterial populations in the GI tract of fish is becoming common practice; examples include
studies in zebrafish (Roeselers et al. 2011; Rodiles et al . 2014), common carp (van Kessel
et al. 2011; Li et al . 2013), grass carp (Wu et al . 2012), European sea bass Dicentrarchus
labrax (Carda-DiƩguez et al . 2013; Peggs et al . 2014) and rainbow trout (Desai et al . 2012).
While the study of common carp GI microbiota by van Kessel et al . (2011) used multiple
primer sets and classified 17,641 sequence reads, there was no mention of coverage or rar-
efaction curves. In the study of grass carp, the 16S rRNA V1-V3 region was targeted, and
259 to 2773 OTUs were identified from 6990 to 18,993 sequence reads (Wu et al . 2012). This
sequence coverage resulted in rarefaction curves that approached the saturation plateau, and
Good's coverage estimations revealed that 85% to 98% of the phylotypes were obtained. The
study of Roeselers et al. (2011) revealed that rarefaction curves generated from 16S V1-V2
sequences in this study did not reach a plateau even after more than 9000 sequence reads. These
studies reveal a far greater level of diversity of the gut microbiota of omnivorous cyprinids than
previous studies that lacked an NGS approach have indicated. However, in the study on the
carnivorous rainbow trout by Desai et al . (2012), which targeted the cpn60 gene, estimates of
Good's coverage ranged from 94.5% to 97.8% when using an average of only 1660 reads per
sample. Although limited studies are available, this comparison between cyprinids and trout is
suggestive of potentially large differences in the level of diversity. Thus it is important to assess
the degree of sampling redundancy by generating rarefaction curves that plot the number of
unique sequences at some percentage identity (e.g. 97%) versus the total number of sequences
generated, which may be readily accomplished using the open-source software programs such
as DOTUR as part of the MOTHUR project (Schloss et al. 2009).
In addition to surveys of bacterial communities, NGS permits the more facile de novo
sequencing of bacterial genomes and community genomes. In particular for probiotic bacterial
strains, de novo sequencing enables the determination of whether strains contain any potential
virulence factors or pathways for secondary metabolite synthesis, and enables design of unique
sequences for primer or probe design that are specific to the probiotic strain (Altermann et al.
2005; Callanan 2005; O'Flaherty and Klaenhammer 2011). The gene sequences from probi-
otic strains can be used to determine gene expression in vivo and investigate the impact of host
microenvironments or different diet formulations on the activity of probiotic strains within
the fish intestine. Entire microbial communities may also be subjected to community genomic
(a.k.a. 'metagenomic') analysis, by direct sequencing of bacterial genomes or gene transcripts,
or by cloning of the DNA or RNA extracted from an environmental sample (Rondon et al.
2000). Metagenomic analysis of human gut samples has revealed important contributions of
gut microorganisms to obesity and identified bacterial phylotypes common among the gut
microbiota of different individuals (Turnbaugh et al. 2009; Qin et al. 2010). Future work in
understanding the contributions of probiotics to fish intestinal health and disease may investi-
gate the functional genomics of probiotic strains and how these specific strains interact with
their host and its diverse intestinal microbiota. There are many other bioinformatics analyses
that may be undertaken to investigate (meta)genome sequences, which are beyond the realm
of this chapter.
Clearly the field of probiotics as it relates to fish growth performance, immune com-
petence, and disease will be better informed in the future through application of these
culture-independent technologies.
Search WWH ::




Custom Search