Transcriptome profiles of livers and kidneys from three rainbow trout (Oncorhynchus mykiss) clonal lines distinguish stocks from three allopatric populations

https://doi.org/10.1016/j.cbd.2006.10.001Get rights and content

Abstract

Transcriptome profiling is a powerful means of simultaneously identifying large numbers of genes that respond transcriptionally to stimuli of any sort. Whereas individuality at the level of genomic sequence is readily revealed and can be expected to influence transcriptional responses, knowledge of the global transcriptomic consequences of genomic individuality is in its infancy. Appreciation of the inherent variability of biological systems gives us confidence in predicting that no two individuals in any outbred population will respond identically to a stimulus. More critical for comparative studies, even unstimulated transcriptomes will be distinctive for each individual. To assess the confidence with which inferences may be drawn from transcriptome profiling when genetically identical samples can be assured, we examined the unprovoked transcriptomes of hepatic and pronephric (head kidney) tissues in three clonal lines of Rainbow trout (Oncorhynchus mykiss). Clonal individuals derived from three allopatric populations presented transcriptional profiles for both liver and pronephros that were not statistically significantly different within each clonal line; however each clonal line was distinguished by a subset of genes with constitutively different transcript abundance. Among these, immunologically-relevant genes were over-represented, possibly reflecting evolutionarily recent, pathogen-driven genetic sweeps.

Introduction

Transcriptome profiling is a powerful means to ascertain the relative importance of gene products in any chosen cell, tissue, organism or condition. Consequently, transcriptome databases are expanding rapidly. Several potential sources of unintended variation exist in such datasets, even when experiments are meticulously designed to include relevant controls and protocols are carefully executed. Because variation is the very essence of natural biological systems, even the most carefully executed transcriptome analyses will always reveal differences between samples. In transcriptome studies, differences are the raw material on which informative inferences are based. Uncertainty is introduced if the observed variability is due both to defined variables intentionally built into the experimental design and to inherent but undefined genetic differences between the sources of RNAs in the study population.

A primary aim of this study was to ascertain individual differences between transcriptomes in individuals from three different clonal lines of rainbow trout (Oncorhynchus mykiss). Clonal lines are excellent sources of material for such studies (Purcell et al., 2006) and should facilitate discernment of the technical and the genetic causes of differences. All of the fish in these experiments were raised under the same conditions and not treated in any way that might induce distinctive transcriptional profiles. We posited two hypotheses: If all individuals have experienced identical environmental conditions prior to sampling (1) the same tissue in clonal individuals should reveal very similar transcriptional profiles, and (2) the transcriptome of each clonal individual will more closely resemble transcriptomes of other members of the same clonal line than it resembles those of members of other clonal lines, i.e. an individual can be assigned to a clonal line on the basis of its transcriptome. As biology is inherently variable, we expected to find differences between different allopatric populations of trout and we hypothesized that genes that are differently transcribed in different stocks could reflect geographically isolated, selective events (genetic sweeps) due to environmental factors that could have caused differential mortality. Since epizootics occur commonly among salmonids, we were interested to discern if genes involved in the early acting elements of the innate arm of the immune system would be over-represented among that subset of genes in which clonal line-to-clonal line differences in transcription were constitutive.

Three clonal lines of rainbow trout (O. mykiss) originating from domesticated strains (Young et al., 1996) were produced by androgenesis and gynogenesis. The fish were raised under identical conditions until liver and kidney tissues were collected for the extraction of total RNA and subsequent microarray analysis. Transcriptomes of individual trout were not significantly different within each clonal line; however each clonal line was distinctive in transcript levels for a small number of genes, all of which were immunologically relevant.

Section snippets

Fish

Three clonal lines of rainbow trout (Young et al., 1996) were produced and raised at the Trout Genetics Research Hatchery, School of Biological Sciences, Washington State University.

The OSU line is a homozygous female line which was initially produced by one generation of androgenesis and one generation of gynogenesis. The Arlee and Hot Creek lines are homozygous male lines which were produced by two generations of androgenesis. In order to avoid the possibility that fish tanks might

Relatedness of individuals and precision of technique

As predicted, when data from dye–swaps were clustered hierarchically using Acuity 4.0 algorithms, reciprocally-labeled cDNAs for the three clonal Arlee fish clustered more closely when they were from the same fish than when they were from different fish (Fig. 1). As reflected in lengths (distances) of the horizontal lines to the right of the labels in Fig. 1, Hot Creek and OSU individuals were less distinct than were the Arlee fish; in fact they were similar enough that the algorithm did not

Discussion

DNA microarray analysis has affirmed that unique traits in the transcriptome fingerprints of rainbow trout livers and pronephros distinguish clonal lines of this species. Within each of the two tissues, mRNAs representing most genes scored in this study were present in similar amounts in all individuals regardless of clonal identity. However, cluster analysis showed that the transcriptomes of the individual fish more closely resembled those of its clonemates than those of members of the

Acknowledgements

Funds to support this research were provided by the NIH (ES 03850) and by a U.S. Department of Agriculture Special Research Grant titled “Aquaculture, Idaho and Washington” to G.H.T. The GenePix array scanner and supporting software were purchased with funds provided by the Oregon State University Research Office. Access to facilities and expertise (Caprice Rosato and Scott Givan) in the CGRB Central Services Laboratory at Oregon State University is gratefully acknowledged. The sequences for

References (18)

There are more references available in the full text version of this article.

Cited by (0)

1

Present address: Center for Marine Biotechnology and Biomedicine, Scripps Institute of Oceanography, University of California San Diego, La Jolla CA 92093, U.S.A.

View full text