Detection of selection signatures in farmed coho salmon (Oncorhynchus kisutch) using dense genome-wide information.

Animal domestication and artificial selection give rise to gradual changes at the genomic level in populations. Subsequent footprints of selection, known as selection signatures or selective sweeps, have been traced in the genomes of many animal livestock species by exploiting variation in linkage disequilibrium patterns and/or reduction of genetic diversity. Domestication of most aquatic species is recent in comparison with land animals, and salmonids are one of the most important fish species in aquaculture. Coho salmon (Oncorhynchus kisutch), cultivated primarily in Chile, has been subjected to breeding programs to improve growth, disease resistance traits, and flesh color. This study aimed to identify selection signatures that may be involved in adaptation to culture conditions and traits of productive interest. To do so, individuals of two domestic populations cultured in Chile were genotyped with 200 thousand SNPs, and analyses were conducted using iHS, XP-EHH and CLR. Several signatures of selection on different chromosomal regions were detected across both populations. Some of the identified regions under selection contained genes such anapc2, alad, chp2 and myn, which have been previously associated with body weight in Atlantic salmon, or sec24d and robo1, which have been associated with resistance to Piscirickettsia salmonis in coho salmon. Findings in our study can contribute to an integrated genome-wide map of selection signatures, to help identify the genetic mechanisms of phenotypic diversity in coho salmon.

Carotenoid pigmentation in salmon: variation in expression at BCO2-l locus controls a key fitness trait affecting red coloration.

Carotenoids are primarily responsible for the characteristic red flesh coloration of salmon. Flesh coloration is an economically and evolutionarily significant trait that varies inter- and intra-specifically, yet the underlying genetic mechanism is unknown. Chinook salmon (Oncorhynchus tshawytscha) represents an ideal system to study carotenoid variation as, unlike other salmonids, they exhibit extreme differences in carotenoid utilization due to genetic polymorphisms. Here, we crossed populations of Chinook salmon with fixed differences in flesh coloration (red versus white) for a genome-wide association study to identify loci associated with pigmentation. Here, the beta-carotene oxygenase 2-like (BCO2-l) gene was significantly associated with flesh colour, with the most significant single nucleotide polymorphism explaining 66% of the variation in colour. BCO2 gene disruption is linked to carotenoid accumulation in other taxa, therefore we hypothesize that an ancestral mutation partially disrupting BCO2-l activity (i.e. hypomorphic mutation) allowed the deposition and accumulation of carotenoids within Salmonidae. Indeed, we found elevated transcript levels of BCO2-l in white Chinook salmon relative to red. The long-standing mystery of why salmon are red, while no other fishes are, is thus probably explained by a hypomorphic mutation in the proto-salmonid at the time of divergence of red-fleshed salmonid genera (approx. 30 Ma).

Identification of olfactory receptor genes in Atlantic salmon Salmo salar.

It has been hypothesized that salmonids use olfactory cues to return to their natal rivers and streams. The key components of the molecular pathways involved in imprinting and homing, however, are still unknown. Aquatic chemical cues are received through the nares and into the nasal cavity that contains a single olfactory organ, the olfactory rosette. The olfactory rosette contains sensory neurons, each of which is thought to express only one olfactory receptor. If odorants are involved in salmonid homing migration then olfactory receptors should play a critical role in the dissipation of information from the environment to the fish. Therefore, to understand the molecular basis for imprinting and homing in Atlantic salmon Salmo salar it is important to identify and characterize the repertoire of olfactory receptors in this species. The first public assembly of the S. salar genome was searched for genes encoding three of the superfamilies of fish olfactory receptors: V2R-like (olfc), V1R-like (ora) and main olfactory receptor (mor). A further six ora genes were added to ora1 and ora2, which had been described previously. In addition, 48 putative mors were identified, 24 of which appear to be functional based on their gene structures and predicted amino-acid sequences. Phylogenetic analyses were then used to compare these S. salar olfactory receptor genes with those of zebrafish Danio rerio, two pufferfish species Takifugu rubripes and Tetraodon nigroviridis, medaka Oryzias latipes and three-spined stickleback Gasterosteus aculeatus.

Expression of olfactory receptors in different life stages and life histories of wild Atlantic salmon (Salmo salar).

It has been hypothesized that salmonids use olfactory cues to return to their natal rivers and streams. However, the key components of the molecular pathway involved in imprinting and homing are still unknown. If odorants are involved in salmon homing migration, then olfactory receptors should play a critical role in the dissipation of information from the environment to the fish. Therefore, we examined the expression profiles of a suite of genes encoding olfactory receptors and other olfactory-related genes in the olfactory rosettes of different life stages in two anadromous and one non-anadromous wild Atlantic salmon populations from Newfoundland, Canada. We identified seven differentially expressed OlfC genes in juvenile anadromous salmon compared to returning adults in both populations of anadromous Atlantic salmon. The salmon from the Campbellton River had an additional 10 genes that were differentially expressed in juveniles compared to returning adults. There was no statistically significant difference in gene expression of any of the genes in the non-anadromous population (P < 0.01). The function of the OlfC gene products is not clear, but they are predicted to be amino acid receptors. Other studies have suggested that salmon use amino acids for imprinting and homing. This study, the first to examine the expression of olfactory-related genes in wild North American Atlantic salmon, has identified seven OlfC genes that may be involved in the imprinting and homeward migration of anadromous Atlantic salmon.

General and family-specific gene expression responses to viral hemorrhagic septicaemia virus infection in rainbow trout (Oncorhynchus mykiss).

The ability of rainbow trout (Oncorhynchus mykiss) to respond successfully to infection by viral hemorrhagic septicaemia virus (VHSV) is expected to involve a large number of biochemical processes. We hypothesized that this would be reflected at the gene expression level in infected fish, and we tested it by examining gene expression levels in the head kidney of trout at a genome-wide scale with a 16K cDNA microarray for salmonids. Expression levels were recorded during 16 days following bath challenge. The challenge experiment included a relatively low susceptibility (32% survival following challenge) and a relatively high susceptibility (18% survival following challenge) trout family that were both split into a group exposed to virus and a non-exposed control group. In total, 939 genes were differentially expressed between infected and non-infected fish (FDR p=0.05). Five groups of Gene Ontology categories were involved in immune-related processes and over-represented in infected fish: (i) stress and defense response, (ii) NFkappaB signal transduction, (iii) response to non-self, (iv) antigen processing and presentation, and (v) proteasome complexes. The first four categories were also over-represented among the 642 differentially expressed genes in the low-susceptibility trout family but not among the 556 differentially expressed genes in the high-susceptibility trout family. Expression profiles for most immune genes discussed showed increased transcription from day 3 post-challenge. The results suggest that the innate immune system may play an important role in the successful response to VHSV in rainbow trout. In addition, the results indicate that a superior regulation of the transcription of several key innate immune-related genes contribute to the increased survival in resistant fish.

Identification of the sex chromosomes of brown trout (Salmo trutta) and their comparison with the corresponding chromosomes in Atlantic salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss).

Males are the heterogametic sex in salmonid fishes. In brown trout (Salmo trutta) the sex-determining locus, SEX, has been mapped to the end of linkage group BT-28, which corresponds to linkage group AS-8 and chromosome SSA15 in Atlantic salmon (Salmo salar). We set out to identify the sex chromosomes in brown trout. We isolated Atlantic salmon BAC clones containing microsatellite markers that are on BT-28 and also on AS-8, and used these BACs as probes for fluorescent in situ hybridization (FISH) analysis. SEX is located on the short arm of a small subtelocentric/acrocentric chromosome in brown trout, which is consistent with linkage analysis. The acrocentric chromosome SSA15 in Atlantic salmon appears to have arisen by a centric fusion of 2 small acrocentric chromosomes in the common ancestor of Salmo sp. We speculate that the fusion process that produced Atlantic salmon chromosome SSA15 disrupted the ancestral sex-determining locus in the Atlantic salmon lineage, providing the impetus either for the relocation of SEX or selection pressure for a novel sex-determining gene to arise in this species. Thus, the sex-determining genes may differ in Atlantic salmon and brown trout.

Grayling (Thymallinae) phylogeny within salmonids: complete mitochondrial DNA sequences of Thymallus arcticus and Thymallus thymallus.

The phylogenetic relationships among the three subfamilies (Salmoninae, Coregoninae and Thymallinae) in the Salmonidae have not been addressed extensively at the molecular level. In this study, the whole mitochondrial genomes of two Thymallinae species, Thymallus arcticus and Thymallus thymallus were sequenced, and the published mitochondrial genome sequences of other salmonids were used for Bayesian and maximum-likelihood phylogenetic analyses. These results support an ancestral Coregoninae, branching within the Salmonidae, with Thymallinae as the sister group to Salmoninae.

The sex determining loci and sex chromosomes in the family salmonidae.

Salmonids are descended from a common ancestor that underwent an autotetraploidization event. After a whole genome duplication species could deal with sex determination by deleting one copy of SEX, the sex determining locus, or by recruiting a duplicated transcription factor to become a novel sex determining gene. It is not known which if any of these strategies salmonids adopted, but it appears that they all have primarily a genetic mechanism of sex determination with male heterogamety. The sharing of sex-linked markers on the X and Y chromosomes and the difficulty in identifying Y-specific markers indicate that X and Y chromosomes in salmonids have a large pseudoautosomal region and a small sex determining region. Linkage analyses suggest that either SEX differs in different lineages or else has remained the same and moved by transposition to different chromosomes. The identification of the sex chromosomes in salmonid species has not resolved this issue. It is clear that salmonids are at an early stage in sex chromosome differentiation and therefore provide a wonderful opportunity to study the evolution of sex determination. The availability of a reference salmonid genome sequence would provide an important resource for research in this area.

Rainbow smelt (Osmerus mordax) genomic library and EST resources.

Genomic resources in rainbow smelt (Osmerus mordax) enable us to examine the genome duplication process in salmonids and test hypotheses relating to the fate of duplicated genes. They further enable us to pursue physiological and ecological studies in smelt. A bacterial artificial chromosome library containing 52,410 clones with an average insert size of 146 kb was constructed. This library represents an 11-fold average coverage of the rainbow smelt (O. mordax) genome. In addition, several complementary deoxyribonucleic acid libraries were constructed, and 36,758 sequences were obtained and combined into 12,159 transcripts. Over half of these transcripts have been identified, several of which have been associated with cold adaptation. These basic resources show high levels of similarity (86%) to salmonid genes and provide initial support for genome duplication in the salmonid ancestor. They also facilitate identification of genes important to fish and direct us toward new technologies for other studies in fish biology.

Coordinated down-regulation of the antigen processing machinery in the gills of amoebic gill disease-affected Atlantic salmon (Salmo salar L.).

Several important cultured marine fish are highly susceptible to an ectoparasitic condition known as amoebic gill disease (AGD). In AGD-affected fish, modulation of IL-1beta, p53 and p53-regulated transcripts is restricted to the (multi)focal AGD-associated gill lesions. To determine whether this lesion-restricted modulation of transcripts occurs on a transcriptome-wide scale and to identify mechanisms that underpin the susceptibility of fish to AGD, we compared the transcriptome of AGD lesions with "normal" tissue from AGD-affected and healthy individuals. Global gene expression profiling using a 16K salmonid microarray, revealed a total of 176 significantly regulated annotated features and of those, the modulation of 99 (56%) was lesion-restricted. Annotated transcripts were classified according to functional gene ontology. Within the immune response category, transcripts were almost universally down-regulated. In AGD-affected tissue, significant, coordinated down-regulation of the major histocompatibility complex class I (MHC I) pathway-related genes occurred during the later stages of infection and appeared to be mediated by down-regulation of interferon-regulatory factor (IRF)-1, independent of interferon-alpha, interferon-gamma and IRF-2 expression. Within this micro-environment, suppression of the MHC I and possibly the MHC II pathways may inhibit the development of acquired immunity and could explain the unusually high susceptibility of Atlantic salmon to AGD.

Contaminant-associated disruption of vitamin A and its receptor (retinoic acid receptor alpha) in free-ranging harbour seals (Phoca vitulina).

Polychlorinated biphenyls (PCBs) have been associated with a number of toxic effects in marine mammals such as endocrine disruption and immunotoxicity that, in turn, are widely thought to have contributed to population level impacts including reproductive failure and outbreaks of disease. In this study, the dietary hormone vitamin A and expression levels of one of its receptors, retinoic acid receptor alpha (RARalpha), were used as biomarkers of PCB-associated health effects in harbour seals. Harbour seal pups (n=24) were live-captured in coastal British Columbia, Canada, and Washington State, USA, and sampled for whole blood (to obtain peripheral blood mononuclear cells, PBMCs) and blood plasma, as well as biopsies of blubber and skin. Concentrations of circulatory vitamin A (retinol) in plasma and stored vitamin A in blubber were negatively associated with blubber PCB concentrations (R=-0.518, p=0.013 and R=-0.645, p=0.009, respectively). However, vitamin A concentrations in skin, an important target tissue, remained constant, which likely reflects a compensatory transfer from blubber to maintain physiological functions. In addition, we characterized the harbour seal RARalpha, and investigated its expression levels as a potential biomarker in seals. RARalpha expression in blubber, but not on PBMCs, was elevated in more contaminated animals (R=0.580, p=0.009). This may represent a direct contaminant-related effect, or, a compensation for the contaminant-related disruption of (circulatory and/or blubber) hormone levels. Since vitamin A is critical to developmental, reproductive and immunological health, our observations of a contaminant-related disruption of its physiology in free-ranging seals may portend population level consequences. Vitamin A concentrations and RARalpha expression levels can therefore represent relevant and sensitive biomarkers of PCB-associated toxic effects in toxicological studies of marine mammals.

Toxicogenomic responses in rainbow trout (Oncorhynchus mykiss) hepatocytes exposed to model chemicals and a synthetic mixture.

As more salmon gene expression data has become available, the cDNA microarray platform has emerged as an appealing alternative in ecotoxicological screening of single chemicals and environmental samples relevant to the aquatic environment. This study was performed to validate biomarker gene responses of in vitro cultured rainbow trout (Oncorhynchus mykiss) hepatocytes exposed to model chemicals, and to investigate effects of mixture toxicity in a synthetic mixture. Chemicals used for 24h single chemical- and mixture exposures were 10 nM 17alpha-ethinylestradiol (EE2), 0.75 nM 2,3,7,8-tetrachloro-di-benzodioxin (TCDD), 100 microM paraquat (PQ) and 0.75 microM 4-nitroquinoline-1-oxide (NQO). RNA was isolated from exposed cells, DNAse treated and quality controlled before cDNA synthesis, fluorescent labelling and hybridisation to a 16k salmonid microarray. The salmonid 16k cDNA array identified differential gene expression predictive of exposure, which could be verified by quantitative real time PCR. More precisely, the responses of biomarker genes such as cytochrome p4501A and UDP-glucuronosyl transferase to TCDD exposure, glutathione reductase and gammaglutamyl cysteine synthetase to paraquat exposure, as well as vitellogenin and vitelline envelope protein to EE2 exposure validated the use of microarray applied to RNA extracted from in vitro exposed hepatocytes. The mutagenic compound NQO did not result in any change in gene expression. Results from exposure to a synthetic mixture of the same four chemicals, using identical concentrations as for single chemical exposures, revealed combined effects that were not predicted by results for individual chemicals alone. In general, the response of exposure to this mixture led to an average loss of approximately 60% of the transcriptomic signature found for single chemical exposure. The present findings show that microarray analyses may contribute to our mechanistic understanding of single contaminant mode of action as well as mixture effects, but that its use in screening of complex environmental samples will need to be further evaluated.

Identification of the sex-determining locus of Atlantic salmon (Salmo salar) on chromosome 2.

We have integrated data from linkage mapping, physical mapping and karyotyping to gain a better understanding of the sex-determining locus, SEX, in Atlantic salmon (Salmo salar). SEX has been mapped to Atlantic salmon linkage group 1 (ASL1) and is associated with several microsatellite markers. We have used probes designed from the flanking regions of these sex-linked microsatellite markers to screen a bacterial artificial chromosome (BAC) library, representing an 11.7x coverage of the Atlantic salmon genome, which has been HindIII fingerprinted and assembled into contigs. BACs containing sex-linked microsatellites and their related contigs have been identified and representative BACs have been placed on the Atlantic salmon chromosomes by fluorescent in situ hybridization (FISH). This identified chromosome 2, a large metacentric, as the sex chromosome. By positioning several BACs on this chromosome by FISH, it was possible to orient ASL1 with respect to chromosome 2. The region containing SEX appears to lie on the long arm between marker Ssa202DU and a region of heterochromatin identified by DAPI staining. BAC end-sequencing of clones within sex-linked contigs revealed five hitherto unmapped genes along the sex chromosome. We are using an in silico approach coupled with physical probing of the BAC library to extend the BAC contigs to provide a physical map of ASL1, with a view to sequencing chromosome 2 and, in the process, identifying the sex-determining gene.

Evolution of duplicated growth hormone genes in autotetraploid salmonid fishes.

A defining character of the piscine family Salmonidae is autotetraploidy resulting from a genome-doubling event some 25-100 million years ago. Initially, duplicated genes may have undergone concerted evolution and tetrasomic inheritance. Homeologous chromosomes eventually diverged and the resulting reduction in recombination and gene conversion between paralogous genes allowed the re-establishment of disomic inheritance. Among extant salmonine fishes (e.g. salmon, trout, char) the growth hormone (GH) gene is generally represented by two functional paralogs, GH1 and GH2. Sequence analyses of salmonid GH genes from species of subfamilies Coregoninae (whitefish, ciscos) and Salmoninae were used to examine the evolutionary history of the duplicated GH genes. Two divergent GH gene paralogs were also identified in Coregoninae, but they were not assignable to the GH1 and GH2 categories. The average sequence divergence between the coregonine GH genes was more than twofold lower than the corresponding divergence between the salmonine GH1 and GH2. Phylogenetic analysis of the coregonine GH paralogs did not resolve their relationship to the salmonine paralogs. These findings suggest that disomic inheritance of two GH genes was established by different mechanisms in these two subfamilies.

Genomic sequencing of the bovine T cell receptor beta locus.

The T cell, which plays an integral role in the coordination of the immune system, has a heterodimer receptor (TCR) that can exist in one of the two forms: alpha/beta or gamma/delta. Cells displaying the gamma/delta TCR comprise less than 5% of T cell populations in humans and mice. In the bovine system, however, gamma/delta populations can reach as high as 60%. Differences in T cell populations make the bovine system an excellent candidate for genomic TCR sequencing and multi-species comparisons. In an effort to characterize the bovine TCR loci, a genomic library was screened for the beta TCR gene. A shotgun sequencing library was constructed and preliminary analysis demonstrates that the organization of the bovine TCR beta constant regions is different from both humans and mice. The bovine beta locus appears to have a third constant region. Overall, the genomic characterization of the bovine TCR genes will provide insight into the evolution of T cell receptor.

ERCC1: a comparative genomic perspective.

ERCC1 plays an essential role in the nucleotide excision repair (NER) of DNA. We compare 37 kb of sequence from the ERCC1 region on human chromosome 19q13.3 to the orthologous region on mouse chromosome 7. In addition to showing the conserved gene structure between ERCC1, ASE-1, and their murine counterparts, this genomic comparison reveals a highly conserved 497 bp segment found 5 kb upstream of ERCC1 exon 1 that contains a CpG island and previously unidentified "classical" promoter elements. Additional putative regulatory elements are also found within a conserved LINE-1 (long interspersed nuclear element) sequence 800 bp upstream of exon 1 in both human and mouse. Expressed sequence tag (EST) assemblies for human ERCC1 identified numerous splice variants involving exons 1, 2, 3, 7, 8, and 9 that could affect DNA repair efficiencies of ERCC1. A previously undescribed transcript that reads through exon 9 and utilizes the polyadenylation signal of a neighboring Alu element accounts for nearly half of the total splice variants identified in the human EST database. This transcript would theoretically translate to a larger ERCC1 protein product containing a novel C-terminal end. Overall, approximately 18% of publicly available ERCC1 cDNA sequences were determined to be splice variants, while no variants were found in the mouse. The ability to assess novel transcripts and identify candidate regulatory regions demonstrates the potential utility for a catalogue archiving comparative analyses for all genes involved in DNA repair. Our comparative genomic analysis of ERCC1 can be viewed at http://web.uvic.ca/-bioweb/laj.html.

Comparative genomics of the human and mouse T cell receptor loci.

The availability of the complete genomic sequences of the human and mouse T cell receptor loci opens up new opportunities for understanding T cell receptors (TCRs) and their genes. The full complement of TCR gene segments is finally known and should prove a valuable resource for supporting functional studies. A rational nomenclature system has been implemented and is widely available through IMGT and other public databases. Systematic comparisons of the genomic sequences within each locus, between loci, and across species enable precise analyses of the various diversification mechanisms and some regulatory signals. The genomic landscape of the TCR loci provides fundamental insights into TCR evolution as highly localized and tightly regulated gene families.

Comparative analysis of the gene-dense ACHE/TFR2 region on human chromosome 7q22 with the orthologous region on mouse chromosome 5.

Chromosome 7q22 has been the focus of many cytogenetic and molecular studies aimed at delineating regions commonly deleted in myeloid leukemias and myelodysplastic syndromes. We have compared a gene-dense, GC-rich sub-region of 7q22 with the orthologous region on mouse chromosome 5. A physical map of 640 kb of genomic DNA from mouse chromosome 5 was derived from a series of overlapping bacterial artificial chromosomes. A 296 kb segment from the physical map, spanning ACHE: to Tfr2, was compared with 267 kb of human sequence. We identified a conserved linkage of 12 genes including an open reading frame flanked by ACHE: and Asr2, a novel cation-chloride cotransporter interacting protein Cip1, Ephb4, Zan and Perq1. While some of these genes have been previously described, in each case we present new data derived from our comparative sequence analysis. Adjacent unfinished sequence data from the mouse contains an orthologous block of 10 additional genes including three novel cDNA sequences that we subsequently mapped to human 7q22. Methods for displaying comparative genomic information, including unfinished sequence data, are becoming increasingly important. We supplement our printed comparative analysis with a new, Web-based program called Laj (local alignments with java). Laj provides interactive access to archived pairwise sequence alignments via the WWW. It displays synchronized views of a dot-plot, a percent identity plot, a nucleotide-level local alignment and a variety of relevant annotations. Our mouse-human comparison can be viewed at http://web.uvic.ca/~bioweb/laj.html. Laj is available at http://bio.cse.psu.edu/, along with online documentation and additional examples of annotated genomic regions.

Cloning and characterization of three novel genes, ALS2CR1, ALS2CR2, and ALS2CR3, in the juvenile amyotrophic lateral sclerosis (ALS2) critical region at chromosome 2q33-q34: candidate genes for ALS2.

Amyotrophic lateral sclerosis is a progressive neurodegenerative disease that manifests as selective upper and lower motor neuron degeneration. The autosomal recessive form of juvenile amyotrophic lateral sclerosis (ALS2) has previously been mapped to the 1.7-cM interval flanked by D2S116 and D2S2237 on human chromosome 2q33-q34. We identified three novel full-length transcripts encoded by three distinct genes (HGMW-approved symbols ALS2CR1, ALS2CR2, and ALS2CR3) within the ALS2 critical region. The intron-exon organizations of these genes as well as those of CFLAR, CASP10, and CASP8, which were previously mapped to this region, were defined. These genes were evaluated for mutations in ALS2 patients, and no disease-associated sequence alterations in either exons or intron-exon boundaries were observed. Sequence analysis of overlapping RT-PCR products covering the whole coding sequence for each transcript revealed no aberrant mRNA sequences. These data strongly indicate that ALS2CR1, ALS2CR2, ALS2CR3, CFLAR, CASP10, and CASP8 are not causative genes for ALS2.