Characterization of rainbow trout (Oncorhynchus mykiss) intestinal microbiota and inflammatory marker gene expression in a recirculating aquaculture system

Abstract

Intestinal microbiota and host inflammatory marker gene expression were characterized in rainbow trout (Oncorhynchus mykiss) during a feeding trial designed to determine the digestibility of soybean meal and fish meal. The trial was conducted in a freshwater, biofiltered recirculating aquaculture system in Saskatoon, Canada. Intestinal contents and tissue were collected from nine fish on each of the two experimental diets and a synthetic, casein-based reference diet. Total DNA from contents of each group was pooled to create template for PCR and construction of libraries of cloned cpn60 universal target sequences. A total of 3357 sequences were produced. The most frequently observed sequences were similar to Firmicutes (particularly Carnobacterium maltaromaticum) and Gamma Proteobacteria, including enterobacteria genera Hafnia, Pseudomonas and Aeromonas. The fish meal diet associated microbiota was balanced between Gamma Proteobacteria and Firmicutes while the soybean meal diet associated microbiota was less diverse and dominated by C. maltaromaticum. Species-specific qPCR of several bacterial species revealed a large amount of variation in target abundance between individual fish. Analysis of inflammatory marker genes including proliferating cell nuclear antigen (PCNA), immunoglobulin M (IgM), and interleukin-1 beta (IL-1β) showed a significantly higher level of PCNA associated with the reference diet and a non-significant trend of elevated IL-1β, suggesting that the semi-purified reference diet may be associated with sub-acute intestinal damage. A high level of variation between individual animals was also observed in the gene expression analysis. An examination of correlations within the combined results support the model that diet effects on the fish intestine include indirect effects mediated through modifications of the intestinal microbiota. These results illustrate the benefits of collecting experimental data from individual fish to increase the analytical power of results.

Introduction

The complex microbial community of the intestine (intestinal microbiota) of terrestrial animals plays a critical role in the digestion of food, pathogen exclusion and the development and maturation of the immune system. In fish, the intestinal microbiota is not as dense (viable counts approximately 10⁸ cfu g^− 1 (Kim et al., 2007, Navarrete et al., 2010)) but is presumed to play a similar role in the health and growth of the host. Recent research focused on the potential replacement of fish meal in aqua feeds with plant-derived proteins has led to the observation of pathological changes and inflammation in the intestinal mucosa which are thought to be either a direct effect of anti-nutritional factors on the gut epithelium and/or the result of diet-induced changes in the microbiota structure and function (Bakke-McKellep et al., 2000, Bakke-McKellep et al., 2007, Krogdahl et al., 2003, Sanden et al., 2005). Given the interaction between host, microbiota and diet, understanding these relationships is critical to development and evaluation of novel diets and maximizing fish health and welfare.

Historically, our understanding of the composition of the intestinal microbiota was limited by reliance on culture-based methods where only those organisms that could be readily cultivated in the laboratory could be identified. More recently, culture-independent, molecular methods have been applied to the study of intestinal microbiology, complementing culture-based studies and resulting in an increased appreciation of the complexity of the community. The predominant molecular methods employed in studying fish intestinal microbiota have been fingerprint-style methods, particularly denaturing gradient gel electrophoresis (DGGE) on microbial community samples or restriction fragment length polymorphism (RFLP) analysis for identification of isolates. Although DNA sequencing has been widely employed in the identification of cultured bacterial isolates from the intestinal environment, very few sequence-based studies of the entire intestinal microbiota have been conducted and this work has been done almost exclusively based on sequencing of cloned 16S rRNA genes (Holben et al., 2002, Kim et al., 2007).

One particularly important outcome of work to date has been the description of variation in microbial composition and population density between fish species and among individuals of the same species, as well as variation associated with environmental factors such as management style, temperature and salinity (reviewed by Cahill, 1990, Ringø et al., 1995). This is consistent with studies of other animal species where the microbiota is observed to change in structure with development and environmental changes. These observations emphasize the necessity of having a thorough understanding of the particular experimental system in which nutrition and growth studies take place, including quantifying normal variation in microbiota within the system (within and between animals), making it possible to recognize biologically significant deviations. Currently employed molecular methods have been of limited use since most do not provide sufficient detail or resolution to adequately describe differences or recognize shifts in microbiota composition.

The cpn60 gene (encoding the universal 60 kDa chaperonin) has been established as a target for sequence-based microbial ecology studies and has been demonstrated to be generally more informative than 16S rRNA, providing better discrimination between closely related organisms (Hill et al., 2004). A 552–567 bp region of cpn60, corresponding to nucleotides 274–828 of the E. coli cpn60 gene, can be amplified with universal, degenerate PCR primers (Hill et al., 2006). To date, cpn60 sequence-based methods have been applied to the study of intestinal and urogenital microbiota of a variety of terrestrial animals, including humans (Desai et al., 2009, Dumonceaux et al., 2006a, Hill et al., 2005a, Hill et al., 2005b, Lukwinski et al., 2006). The discriminating power of the cpn60 universal target sequence also makes it an ideal target for the development of species-specific PCR assays for the detection and quantification of bacteria in complex samples (Dumonceaux et al., 2006b).

The challenges in characterization of the intestinal microbiota also apply to investigations of host response, where natural variation in marker gene expression must be understood before conclusions can be drawn regarding the effects of experimental changes in diet or environment. Variation in host gene expression, even for so-called housekeeping genes, is widely experienced but rarely reported (Juul-Madsen et al., 1992, Nath et al., 2006). As a result, it is necessary to examine a number of gene markers associated with any particular physiological pathway and/or systematically identify the markers that are most informative.

The Prairie Aquaculture Research Centre in Saskatoon, Canada is a biologically filtered, recirculating, freshwater system using a de-chlorinated municipal water supply and housing rainbow trout. The objective of our study was to take advantage of an ongoing digestibility trial to characterize the intestinal contents microbiota of rainbow trout on three different diets and to measure host inflammatory marker gene expression in the intestines of fish in this research facility. In addition, we applied species-specific quantitative PCR to intestinal contents samples from individual fish in order to investigate the level of animal to animal variation within and between diet groups. The results of this study add to our understanding of the intestinal microbiota of rainbow trout and further describe the extent of individual to individual variation in fish that should be an important consideration in experimental design. These results form the foundation for future trials and experiments aimed at understanding the interactions between host, microbiota and diet.