Fast skeletal muscle transcriptome of the Gilthead sea bream (Sparus aurata) determined by next generation sequencing
© Garcia de la serrana et al.; licensee BioMed Central Ltd. 2012
Received: 7 October 2011
Accepted: 30 March 2012
Published: 11 May 2012
The gilthead sea bream (Sparus aurata L.) occurs around the Mediterranean and along Eastern Atlantic coasts from Great Britain to Senegal. It is tolerant of a wide range of temperatures and salinities and is often found in brackish coastal lagoons and estuarine areas, particularly early in its life cycle. Gilthead sea bream are extensively cultivated in the Mediterranean with an annual production of 125,000 metric tonnes. Here we present a de novo assembly of the fast skeletal muscle transcriptome of gilthead sea bream using 454 reads and identify gene paralogues, splice variants and microsatellite repeats. An annotated transcriptome of the skeletal muscle will facilitate understanding of the genetic and molecular basis of traits linked to production in this economically important species.
Around 2.7 million reads of mRNA sequence data were generated from the fast myotomal of adult fish (~2 kg) and juvenile fish (~0.09 kg) that had been either fed to satiation, fasted for 3-5d or transferred to low (11°C) or high (33°C) temperatures for 3-5d. Newbler v2.5 assembly resulted in 43,461 isotigs >100 bp. The number of sequences annotated by searching protein and gene ontology databases was 10,465. The average coverage of the annotated isotigs was x40 containing 5655 unique gene IDs and 785 full-length cDNAs coding for proteins containing 58–1536 amino acids. The v2.5 assembly was found to be of good quality based on validation using 200 full-length cDNAs from GenBank. Annotated isotigs from the reference transcriptome were attributable to 344 KEGG pathway maps. We identified 26 gene paralogues (20 of them teleost-specific) and 43 splice variants, of which 12 had functional domains missing that were likely to affect their biological function. Many key transcription factors, signaling molecules and structural proteins necessary for myogenesis and muscle growth have been identified. Physiological status affected the number of reads that mapped to isotigs, reflecting changes in gene expression between treatments.
We have produced a comprehensive fast skeletal muscle transcriptome for the gilthead sea bream, which will provide a resource for SNP discovery in genes with a large effect on production traits of commercial interest and for expression studies of growth and adaptation.
KeywordsTeleost Gene paralogues Splice variants Newbler Roche 454 Myogenesis
The gilthead sea bream (Sparus aurata L.) is widely farmed around the Mediterranean with main centres of production in Greece, Turkey, Spain and Italy. This species which is primarily marketed as fresh fish or fillets is also cultivated in the Red Sea, the Persian Gulf, and the Arabian Sea with global production reaching circa 125,000 metric tonnes in 2008 . Gilthead sea bream is a protandrous hermaphrodite that can reach about 70 cm in length and 5 kg body mass. Males become sexually mature after 0.5 kg and by the second year most individuals have become female (>1.5 kg). The axial musculature or fillet is made up of serially arranged myotomes comprising ~65% of body mass containing slow, intermediate and fast muscle fibre types . Fast muscle fibres comprise the bulk of the myotome. The main expansion of fast muscle fibre with growth occurs by a process called mosaic hyperplasia in which myogenic progenitor cells (MPCs) fuse to form new myotubes on the surface of existing muscle fibres giving rise to a mosaic of fibre diameters as the fish matures [3–6]. MPCs also contribute additional nuclei to the muscle fibre as it expands in length and diameter . In all life history stages, myogenesis involves steps of myoblast proliferation, migration, fusion, terminal differentiation and sarcomere assembly and many of the transcription factors and signaling molecules required for the regulation of these processes have been characterised . In the majority of teleost, mosaic hyperplasia in fast muscle continues until the fish reaches around 40% of its maximum body length [3–6]. Myogenesis is a highly plastic process in which internal and external signals arising from changing environmental conditions; swimming activity and nutritional inputs are integrated to modify growth patterns . Embryonic temperature regime results in persistent changes in growth patterns in later life affecting the final number and size distribution of muscle fibres in adult fish [5, 6, 9] with potential impacts on flesh quality parameters such as texture .
The application of genomic technologies promises to revolutionise our understanding of the genetic and molecular basis of muscle growth and plasticity in farmed fish species; thereby increasing the efficiency and sustainability of aquaculture production. For example, the discovery of genetic polymorphisms associated with commercially important production traits such as growth rate and flesh quality would form the foundation for marker-assisted selection to produce superior strains for farming. Genomic studies could also enable bioactive nutritional components to be identified in commercial feeds and be used to accelerate the development of more sustainable diets with lower environmental impact. The genome of Atlantic cod (Gadus morhua) has recently been described  and several other farmed fish are in the process of being sequenced to draft level including rainbow trout (Oncorhynchus mykiss) , Atlantic salmon (Salmo salar)  and tilapia (Oreochromis niloticus) . There are also significant genetic resources available for the European sea bass (Dicentrarchus labrax) another important species in Mediterranean aquaculture. For example, Kuhl et al [15, 16] developed a complete BAC-end library from the sea bass and gilthead sea bream genomes using the three-spined stickleback (Gasterosteus aculeatus) genome as a reference for description and annotation. In contrast, there are only 1414 GenBank sequences and 74877 ESTs for the gilthead sea bream (revised on July 2011). These sequencing efforts have allowed the development of microarray platforms for gene expression studies [17–19] and sets of microsatellites for selection programs . However, the comparative lack of genetic information is a significant handicap for the development of a serious program for genetic improvement of stocks by marker assistance-selection and for a better understanding of the molecular basis of nutrition, growth, flesh quality, reproduction and disease resistance.
Next Generation Sequencing (NGS) technologies have the potential to rapidly and cost effectively expand sequence databases for non-model organisms [20–22]. In the present study we have used Roche 454 GS FLX titanium sequencing to produce a comprehensive transcriptome of fast skeletal muscle using RNA extracted from adult and juvenile gilthead sea bream subject to different nutritional states and temperatures. The resulting transcriptome with 40-times average coverage was annotated and screened for gene paralogues, alternatively spliced transcripts and microsatellite repeat sequences.
Number of reads obtained per experimental condition and their respective Newbler assembly results
Isotigs over 100 bp
Isotig mean lenght (bp)
Identification of full-length coding sequences (CDS) and splice variants
Transcripts with functional domains deleted that were experimentally confirmed by PCR
CDS fraction (%)
Number of exons predicted
IPR domain lost
Aspartate beta hydroxylase
Authophagic related protein 27 (ATG27)
Coagulation factor x
9 and 10
1 and 2
ATPase like ParA
Bridging integrator 1
1 to 10
autoregulation binding site
Proteinase inhibitor, cathepsin propeptide
Polyadenylate-binding protein- interacting protein 2
Transitional endoplasmic reticulum atpase (cdc48)
AAA + Atpase domain
Aspartate descarboxylase fold
Peptidase M22, Glycopeptidase
Zinc finger x-chromosomal protein
Zinc finger C2H2
Dead (asp-glu-ala-asp) box polypeptide 1
RNA recognition motif domain
Identification of microsatellite sequences
The transcriptome was screened for potential microsatellite repeats excluding adenine repetitions, which most likely correspond to polyA tails. Around 750 potential microsatellites were detected in the total isotigs (data available on request from DG). To provide information linked to known sequences, only microsatellites localized in annotated isotigs were further studied. A total of 177 non-redundant microsatellites were identified in annotated isotigs. Dinucleotide repeated motifs were the most abundant, representing 75% of the total, followed by mononucleotide (13%), trinucleotide (11%), tetranucleotide (2%) and pentanucleotide (1%) repeats (Additional file 11). All 177 microsatellite reported in this study were found in predicted UTR regions with 40% of them linked to full-coding sequence genes.
Identification of gene paralogues
List of paralogues identified in the gilthead sea bream skeletal muscle transcriptome
Paralogue Gene name
Fraction of CDS (%)
Paralogues identity (%)
Acethylcoline receptor subunit alpha 1
Adp/atp translocase (Solute carrier family 25, SLC25)
Catalyzes the exchange of ADP and ATP across the mitochondrial inner membrane
SLC25 member 5 and 6
Calpain small subunit 1
Calcium-regulated thiol-protease involved in cytoskeletal remodeling
Calpain subunit 1a/b
Carnitine O- acetyltransferase
Carnitine acetylase is specific for short chain fatty acids
Carnitine O-acetyltransferase a1/a2
Dehydrogenase reductase member 7c
Dysferlin interacting protein 1
Sarcolemma repair mechanism of both skeletal muscle and cardiomyocytes
Epithelial membranse protein 3
Probably involved in cell proliferation and cell-cell interactions
Glioblastoma amplified sequence
Widely expressed. Most abundant in heart and skeletal muscle
High mobility group box 1
DNA binding proteins that associates with chromatin
Component of the class I major histocompatibility complex (MHC)
Myomesin 185 kDa
Major component of the vertebrate myofibrillar M band
Serine threonine-protein phosphatase
Essential for cell division, and participates in muscle contractility and protein synthesis
Solute carrier family 38 member 5
Sodium-dependent, pyrimidine- and purine-selective. Involved in the homeostasis of endogenous nucleosides
Tyrosine 3 monooxigenase
Set and mynd domain- containing protein 1
Acts as a transcriptional repressor. Essential for cardiomyocyte differentiation
Dual specificity phosphatase and pro Isomerase domain containing 1
Catalyse reaction: Protein tyrosine phosphate + H2O = protein tyrosine + phosphate.
Metalloproteinase inhibitor 2 precursor
Complexes with metalloproteinases and irreversibly inactivates them
Retinoid x gamma
Rreceptor for retinoic acid
Contributes to the stabilization of the junctional membrane complexes
60s ribosomal protein l5
Required for rRNA maturation and formation of the 60 S ribosomal subunits
Reduces trans-2,3-stearoyl-CoA to stearoyl-CoA of long and very long chain fatty acids
Methylmalonate- semialdehyde Dehydrogenase
Plays a role in valine and pyrimidine metabolism. Binds fatty acyl-CoA
Eukaryotic translation initiation factor 4e type 3
Its translation stimulation activity is repressed by binding to the complex CYFIP1-FMR1
Fk506-binding protein 1a
May play a role in modulation of ryanodine receptor isoform-1 (RYR-1)
Splicing arginine serine-rich 11
May function in pre-mRNA splicing.
Kelch repeat and btb domain containing 10
Substrate-specific adapter of an E3 ubiquitin-protein ligase complex
Transcription related sequences
Transcription factor families present in the gilthead sea bream transcriptome
Transcription Factor Family
Example of family member
Number of isotigs
Percentage of total transcription factors
interleukin enhancer-binding factor 3
yy1 transcription represor factor
e1a binding protein p300
signal transducer and activator of transcription 3
transcription factor jun-d
hypoxia-inducible factor 3 alpha
General transcription factor
transcription factor 20
zinc finger and btb domain containing 33
six homeobox 1
bromodomain adjacent to zinc finger 2b
Nuclear hormone receptor
Peroxisome proliferator-activated receptor alpha
glucocorticoid receptor dna-binding factor 1
High mobility group box
transcription factor sox-6 isoform 2
interferon regulatory factor 2
forkhead box o3
tea domain family member 3
smad family member 2
e2f transcription factor 6
Partial assemblies and expression analysis
Expression analysis of libraries showing isotigs where reads from each experimental condition significantly contributed to the assembly
Orthologue accession number
Slow myosin heavy chain 2
Similar to ankyrin 2
Calcium binding and coiled coil domain
myotubularin-related protein 5
GTPase, IMAP family member 7
Myosin, heavy polypeptide 6
Adenylate kinase 1-2
Aurora kinase A-interactinng protein
Slow Troponin T2
Putative nuclease HARBI1
GTPase, IMAP family member 7
Ribosomal protein L28
Myosin light chain 2
Xin actin-binding repeat containing protein 1
Myosin-6-like isoform 1
Heat shock protein 30
Activator of 90kda heat shock protein ATPase homolog 1
eEF1A2 binding protein
Heat shock protein 4a
Heat shock protein 90
Zinc binding protein 33A
Interferon stimulated gene 15
Receptor transporting protein 3
Nicotinic acetylcholine receptor alpha 1b
Similar to ankyrin 2
G-rich sequence factor 1
rRNA promoter binding protein
The number of genes that can be obtained from Next Generation Sequencing is higher for normalised than non-normalised libraries of the kind used in the present study; however, unbiased libraries have the advantage of yielding a higher number of full-length cDNA sequences . The number of annotated isotigs in the present study was 10,465 (24% of the total) corresponding to 5,655 unique genes. The total number of annotated sequences was less than reported in the coral (Millepora arcicornis) transcriptome (17,000) , rainbow trout (Oncorhynchus mykiss) (376,238) , but similar to that obtained for eel (Anguilla anguilla) (5,530) . However, since our transcriptome was for a single tissue type (fast skeletal muscle) a lower number of unique genes would be expected than for transcriptomes based on sequencing dsDNA libraries from multiple tissues. In addition, previous studies [30, 31, 33] have considered singletons to be a valid source for gene discovery whereas the 96,000 singletons (4805 annotated) obtained in the present study were not included in further analysis.
Our study is the first report of a skeletal muscle transcriptome in teleost fish and it contained 5655 unique transcripts including over 300 annotated transcripts related to transcription control, 750 microsatellite markers (177 associated with annotated istoigs) and 785 full-length cDNAs. The total number of microsatellites obtained was similar than in previous studies . The transcriptome contained all known components of the sarcomere and the majority of proteins were represented by multiple isoforms even though the starting tissue for library construction comprised a pure population of fast twitch muscle fibres (Figure 2). Multiple isoforms of troponins and myosin light chains have previously been reported in single fish muscle fibres . It is likely that isoforms that are expressed at specific developmental stages  or temperatures  contribute to the overall diversity of sarcomeric proteins (Figure 2). Previously, only 22 genes with splice variants have been reported in gilthead sea bream based on SANGER sequencing . In the present study, 43 genes with potential splice variants were described, including 12 that affected known functional domains. This is a relatively low discovery rate given that 30% of genes in the three-spine stickleback genome were predicted to occur as multiple transcripts . The reads containing the splice variant regions were analysed (data not show). In the majority of cases, the number of reads containing the deletion was lower than for the unspliced sequences, indicating lower levels of expression. In contrast, for cytochrome c oxidase subunit 4b and c4b binding protein the reads containing the deletion were more abundant and for a few genes, including bridging integrator 1 and cathepsin H, the proportion of splice variants was similar. The physiological effect of the splice variants with altered functional domains was not analysed in the present work and further studies are necessary to evaluate their impact on cell physiology.
A whole genome duplication occurred in basal teleosts around 300–250 million years ago resulting in duplicate copies of many genes relative to the common ancestor with tetrapods . It was estimated that in the green spotted puffer fish Tetraodon nigriviridis around 15% of the duplicate genes have been retained . Previous transcriptomic studies in Atlantic cod (Gadus morhua) , whitefish (Coregonus clupeamorfis)  and eel (Anguilla anguilla)  have not attempted to identify paralogues. In the present study with over 10,000 transcripts annotated we expected over 400 paralogues, but only 26 could be identified. Differences between expected and the actual number of paralogues found can be explained by three main factors. The first factor is linked to sequence errors in the transcriptome. We found an error rate of 1:200 bp (99.5% accuracy) similar to previous studies  with 15.2% of the transcripts having insertions or deletions in their sequence. In our study, paralogue screening was based on translated isotigs, which are dramatically affected by insertions and deletions. This is because any insertion/deletions that are not multiples of three will change the open reading frame of the isotigs or introduce an in-frame stop codon. The second factor resulting in a low rate of paralogue discovery is the short length of some of the translated peptides. The majority of automatically translated isotigs represented less than 50% of the predicted sequence length (with a large number under 20%). Thus potential paralogues with short translated isotigs failed to pass the quality filters and were not considered further. Finally a very small effect will come from the assembly. Many assemblers are designed to tolerate imperfect sequence alignment to avoid missing true joins. This tolerance for error could result in false positive joins that mask polymorphisms, including paralogues . This effect will be small due the divergence of the paralogues retained after the whole genome duplication, but cannot be completely discarded as a possibility.
Another advantage of using unbiased libraries is that it potentially allows information on gene expression levels to be obtained. The approach used here was to carry out pairwise comparisons between treatments counting the numbers of reads that contributed to isotigs in an assembly derived from the combined treatments (Figure 3; Table 5). The results indicate marked plasticity in gene expression with respect to nutritional status and temperature. In many cases, genes highly ranked for differential abundance between treatments corresponded to the activation of particular pathways. For example, in fed fish, stress chaperones including Hsp90 and Hsp70 and proteins associated with prevention of unfolded protein aggregation, and cytoskeleton structure maintenance was significantly elevated in 33°C compared to 21°C treatments (Figure 3 and 4; Table 5). Heat shock proteins function to increase thermal tolerance following acute exposure to high temperature stress . In contrast, there was no clear pattern of gene expression in the low temperature group that can be specifically associated with treatment. This may result from low temperature inhibiting feeding and inducing a similar depression of protein synthesis and metabolism as observed for fasted fish at higher temperature, thereby masking the specific effects of acute cold stress.
Food deprivation reduces gene expression of enzymes related with glycolysis in fish liver  and muscle . We found a decreased contribution of sequences to isotigs for genes associated with carbohydrate metabolism in fasted relative to fed treatments (Figure 3; Table 5). The fed library was also enriched for Notch-2 which is thought to control myoblast activity and be related to the asymmetric self-renewal of the muscle satellite cells through its inhibitor Numb [48, 49]. It has been suggested that increased Notch expression inhibits differentiation  and stimulates myoblast proliferation . The significant increase of Notch-2 expression and other genes related with metabolism (like GAPDH) could be an indication of higher metabolic rates and myoblast activity in this group compared to treatments exposed to stressful conditions. There was evidence for the upregulation of adenylate kinase-1 (AK) in fasted compared to fed libraries. AK acts as a sensor of the energy status of tissues . An increase of some of the adenylate kinase isoforms was also reported in response to the energy imbalance during fasting in rat tissues . We also found up-regulation of three sarcomeric genes (myosin polypeptide 6, slow myosin light chain 2 and slow troponin 2) consistent with shifts in myofibrillar protein isoform composition towards a slow muscle phenotype in fasted fish. Studies in Atlantic salmon also reported an increase in myosin heavy chain and the myosin light chain 2 transcripts with fasting .
We have produced a detailed fast skeletal muscle transcriptome for the gilthead sea bream, a commercially important aquaculture species in the Mediterranean. The transcriptome contained 5655 unique annotated genes and 785 full-length coding sequences including key transcription factors, signaling molecules and structural proteins involved in myogenesis and growth. Some limitations in the identification of gene paralogues with 454 sequencing were found. In order to facilitate future genomic studies in this species a Blast server has been made available which contains 10, 465 annotated and 35,996 un-annotated isotigs together with ~ 2,700,000 ESTs .
The juvenile gilthead sea bream (Sparus aurata L.) used in the present study originated from a fish farm brood stock kept at the Institute de Recerca i Tecnologia Agroalimentàries (IRTA) at St Carles de la Ràpita (IRTA-SCR, Spain) and were reared from the larval to juvenile stages according to the standard production procedures of this research facility. After thirteen months, two hundred juvenile gilthead sea bream, weighing 88.1 ± 7.3 g (mean ± SD, n = 35), were selected and maintained in two 400 litre tanks (22.5 kg m−3) in a temperature-controlled seawater re-circulation system (IRTAmarTM) at a mean temperature of 21°C (20.7-21.4°C) and natural photoperiod (13 L:11D). Fish were fed a commercial diet (OptiBreamTM, Skretting; pellet size: 2.6 mm; proximate biochemical composition: 46% protein, 18% fat, 7% ash) at a ration level of 3% (m/m) d−1. An adult female of 2 kg body mass that had been held at ambient temperature (annual range: 10-26°C) and natural photoperiod for several years at IRTA-SCR facilities and fed 3% body mass d-1 was also sampled.
In order to obtain the widest possible range of expressed transcript sub-sets, fish were exposed to different water temperatures and fasting. Experiments were conducted in 400 litres cylindrical tanks connected to a re-circulation unit in order to maintain constant water temperature and dissolved oxygen over 85% saturation. Fish (n = 5) were transferred from 21°C to 11°C or 33°C over 48 h. During the treatments fish were fed as previously described, however those maintained at 11°C, stopped feeding after their transfer to low temperature. Additionally, another group of fish maintained at 21°C were fasted for 5 days.
Since transcripts concentration will change over time with treatment fish were sampled at day 3 (n = 2) and day 5 (n = 3) following attainment of the new environmental conditions in order to obtain a broader representation of expressed genes. Fish were sacrificed using an overdose of 1:5,000 (m/v) of bicarbonate-buffered tricaine methanesulphonate (MS222, Sigma, Madrid, Spain) in seawater followed by spinal cord transection. Pure samples of fast skeletal muscle were dissected from dorsal epaxial myotomes at ~ 0.5 fork length (FL) on a pre-chilled glass plate maintained at 0–4°C. Muscle samples were flash frozen in liquid nitrogen and stored at −80°C until further analysis. Fish handling and trials were conducted in September 2009 in accordance with EC Directive 86/609/EEC for animal experimentation.
RNA extraction and dsDNA synthesis
RNA was extracted using QIAzol (QIAGEN, Crawley - West Sussex, UK) following the manufacturer’s recommendations. The integrity of the RNA was confirmed by ethidium bromide gel electrophoresis. RNA concentration, 260/280 and 260/230 ratios were evaluated using a NanoDrop 1000 spectrophotometer (Thermo Fischer Scientific, Waltman, MA). All RNA samples extracted had a 260/280 ratio higher than 1.9 and 260/230 above 2.2. Samples from each experimental condition were pooled in equal concentrations and the RNA integrity, concentration and ratios evaluated again. The pooled RNA samples were used for the following steps.
The dsDNA synthesis was performed using a MINT cDNA synthesis kit (Evrogen, Moscow, Russia) using cDNA synthesis primer described by Meyer et al., 2009  with a broken poly-T to avoid 454 sequencing problems in mono-nucleotide regions (5′-AAGCAGTGGTATCAACGCAGAGTCGCAGTCGGTACTTTTTTCTTTTTTV-3′). For an accurate evaluation of the dsDNA concentration Quati-IT™ PicoGreen® (Invitrogen, Pailey, UK) was used. PicoGreen® fluorescence was detected by a MSPx3000 qPCR machine as previously described .
The transcriptome for each physiological condition was determined using Roche 454 GS FLX Titanium pyrosequencing using the service run by Genepool, University of Edinburgh, School of Biological Sciences. Each physiological condition was sequenced using a half 454-plate generating around 390,000–490,000 reads with an average length of 400 bp. Because of a technical problem an initial run of the fasted sampled yielded reads with an average length of only 300 bp and therefore this plate was repeated. Both plates yielded high quality reads and were therefore used in the subsequent global assembly.
454 assembly and annotation
Around 2,700,000 reads were used to generate the sea bream transcriptome. For the partial assemblies we used the reads generated from each experimental condition. For the fasted treatment partial assembly reads from the 454 plate that yielded average read lengths (400 bp) were used. Reads were assembled using Newbler 2.5 software (Roche, 454 Life-sciences) which performs well for de novo assembly of 454 transcriptome data . Assemblies were run in a Debain Linux system, IBM x3755 8877, with 8 CPU cores (4 x dual-core AMD Opteron), 64-bit, 2.8GHz processor with 128 Gb of RAM maintained by the University of St Andrews.
To avoid assembly problems caused by the reads from highly expressed genes we trimed them using the –vs against a fasta file with the available sequences for these genes in gilthead sea bream (adapters and genes sequences used from trimming are in Additional file 16). Isotigs generated by the Newbler software are contigs that are consistently connected by subsets of reads. Isotigs are longer than contigs and were used for the annotation and transcriptome analysis.
Isotigs were Blasted and annotated using Blast2GO software . Sequences were blasted using Blastx against the NCBI non-redundant protein collection (nr) database with a threshold of 10-3. Annotation was done with an E-value Hit Filter of 10-6 combined with an Annotation Cutoff of 55 and GO weighting of 5. Blast2GO also annotated sequences for functional domains using InterProScan.
NGS and Sanger sequencing comparisons
Known sea bream sequences produced by the SANGER sequencing method were downloaded from GenBank  and blasted (blastn) against the sea bream transcriptome using a BLAST server  generated by the Genepool group. The best hits isotig/GeneBank were aligned using ClustalW  to determine the nature and number of differences.
Successfully annotated isotigs were introduced in the KEGG Automatic Annotation Server (KAAS) . The SBH method, optimized for ESTs annotation, was used against human, chimpanzee, orang-utan, rhesus, mouse, rat, dog, giant panda, cow, pig, horse, opossum, platypus, chicken, clawed frog, zebrafish, fruit fly and nematode pathway databases. For a more detailed reconstruction of the pathway components the PPT-Toolkit-Cell-Biology from motifolio.com was used.
Identification of full-length cDNAs
Annotated isotigs were translated to the longest amino acids sequence possible using the ORF translator tool in Blast2GO package (no longer available). Sequences with more than 150 amino acids that started with a methionine or had a methionine in the first 50 amino acids were manually blasted using NCBI Blast server against nr/nt database . Blast results were analysed to confirm that the translated isotig covered, at least 90% of the sequence with best hits and that cover the whole CDS.
Isotigs successfully annotated were used for microsatellite repeats search using msatcommander-1.0.2-alpha . An isotig was considered to contain a microsatellite if contain any of the following repeated motifs: at least 10 repeated mononucleotides (other than A), 8 repeated di- or trinucleotides, or 6 repeated tetra-, penta- or hexanucleotid motifs. Their position outside coding sequences was confirmed in those microsatellites linked to annotated isotigs by analysing the translated sequences.
Identification of splice variants
For splice variant identification we screened the list of isogroups generated during Newbler assembly. Each isogroup represents a collection of isotigs containing reads that imply connections between the isotigs. Different isotigs from a given isogroup can be used to infer splice variants. Isogroups with non-annotated isotigs were discarded. The screening was focused on detecting splice variants affecting the coding sequence. The isotigs translated sequences from each isogroup were aligned with ClustalW to detect changes in peptide sequence.
Potential splice variants were filtered a second time by blasting them against the stickleback (Gasterosteus aculeatus) genome where possible, or otherwise the green puffer fish (Tetraodon nigroviridis) genome using the Ensembl webpage BLAT algorithm . Loci positive alignments were retrieved. Splice variants sequences and loci were aligned using the Spidey mRNA/genome analyser  to predict changes in the exon composition. Splice variants with potential changes in exon composition were submitted to InterProScan annotation to detect changes in functional domains. Genes with domain annotation that were altered by splicing were experimentally confirmed using conventional PCR.
Identification of transcription factors (TF)
For the detection of transcription factors and molecules associated with transcription such as methyl transferases, histone acetyl transferases and others we screened isotigs annotated with GO levels related to transcription: GO:0006355 (regulation of cellular transcription), GO:0003700 (modulate transcription), GO:0003677 (interacts selectively with DNA), GO:0008134 (TF binding), GO:0033276 (protein complex able to transcription regulation), GO:0043425 (basic Helix-Loop-Helix interactive elements), GO:0016563 (any activity required for initiation or upregulation of transcription) and GO:0045941 (any transcription regulator activity). IDs were checked against a Transcription Factor database to confirm a role in transcription regulation and to categorize them into families  and against the Uniprot database .
Identification of gene paralogues
Because no formal software has been developed specifically for paralogue screening in assemblies from Next Generation Sequencing we used an indirect approximation using the translated isotigs. A list of protein sequences of known genes from mouse (Mus musculus) was downloaded using BioMart tool from ESEMBL . We also downloaded a list of known paralogues from different teleost species: Takifugu rubripes Tetraodon Nigroviridis Gasterosteus aculeatus Oryzias latipes and Danio rerio. Comparisons between proteins groups were performed using Inparanoid 4.0 . Comparisons were performed using the gilthead sea bream translated transcriptome against one of the datasets at time. When at least two different isotigs were identified to represent the same transcript matched with a single mouse gene they were consider as potential paralogues. In addition, if two or more teleost known paralogues matched with two different isotigs they were also considered as potential paralogues. Other relations between transcripts can give similar output from Inparanoid and be included as paralogues: redundant transcripts, splice variants, sequence fragments and wrongly translated isotigs by insertions/deletions. Inparanoid output was explored by aligning translated sequences of paralogues against each other using ClustalW. This exploration allowed us to detect and trim these “False positives” from the list of potential paralogues.
The amino acids sequences of potential paralogues were blasted against the zebrafish (Danio rerio), stickleback (Gasterosteus aculeatus), takifugu (Takifugu rubripes), medaka (Oryzias latipes), green pufferfish (Tetraodon nigroviridis), chicken (Gallus gallus), frog (Xenopus laevis) and human (Homo sapiens) genomes using Essembl . The sequences from the best hits were downloaded. Alignment of the potential paralogues and their orthologues was performed using the GUIDANCE web tool . Only fragments with an alignment confidence score over 0.93 were used for the phylogenetic analysis. The best evolutionary model was estimated for each alignment using MEGA5 software . Maximum Likelihood phylogenetic analysis was constructed, with the best evolutionary model, using the online pipeline from PhylM .
Reads from each experimental condition were mapped against the total isotigs from the global assembly using GS Reference Mapper (Roche, 454 Life Sciences). The number of reads per contig from each condition was extracted using the R statistical package . Chi-square statistic was applied to detect significant differences in the number of reads per condition per isotig. Isotigs with less than 10 reads were excluded from the analysis. A FDR correction was applied to all p-values below 0.05. Plot graphs comparing the contribution of reads from each experimental condition to the isotig formation were constructed using R package.
IAJ conceived the study, DGDLS was responsible for RNA extraction, dsDNA synthesis, sequence assembly and bioinformatics, AE and KA were responsible for fish husbandry and AE assisted with sample preparation, IAJ and DGDLS wrote the manuscript. All authors read and approved the manuscript.
We would like to thank staff at IRTA for providing Gilthead sea bream for this study, particularly Enric Gisbert and Maria Darias. DNA sequencing was carried out in the GenePool genomics Facility in the University of Edinburgh. We thank GenePool Staff for assistance, especially from Dr Stephen Bridge, with sequencing and bioinformatics advice. Also thank to Dr Paris Vestos for his support in the design of the R scripts. The research was funded by the European Community’s Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 222719 – LIFECYCLE. This work also received funding from the MASTS pooling initiative (The Marine Alliance for Science and Technology for Scotland) and their support is gratefully acknowledged. MASTS is funded by the Scottish Funding Council (grant reference HR09011) and contributing institutions.
- López-Albors O, Gil F, Ramírez-Zarzosa G, Vázquez JM, Latorre R, García-Alcázar A, Arencibia A, Moreno F: Muscle development in Gilthead sea bream (Sparus aurata L.) and sea bass (Dicentrarchus labrax, L.): further histochemical and ultrastructural aspects. Anat Histol Embryol. 1998, 27: 223-229. 10.1111/j.1439-0264.1998.tb00185.x.View ArticlePubMed
- Rowlerson A, Mascarello F, Radaelli G, Veggetti A: Differentiation and growth of muscle in the fish Sparus aurata (L): II. Hyperplastic and hypertrophic growth of lateral muscle from hatching to adult. J Muscle Res Cell Motil. 1995, 16: 223-236. 10.1007/BF00121131.View ArticlePubMed
- Weatherley AH, Gill HS, Lobo AF: Recruitment and maximal diameter of axial muscle fibres in teleosts and their relationship to somatic growth and ultimate size. J Fish Biol. 1988, 33: 851-859. 10.1111/j.1095-8649.1988.tb05532.x.View Article
- Johnston IA, Manthri S, Alderson R, Smart A, Campbell P, Nickell D, Robertson B, Paxton CG, Burt ML: Freshwater environment affects growth rate and muscle fibre recruitment in seawater stages of Atlantic salmon (Salmo salar). J Exp Biol. 2003, 206: 1337-1351. 10.1242/jeb.00262.View ArticlePubMed
- Johnston IA, Lee HT, Macqueen DJ, Paranthaman K, Kawashima C, Anwar A, Kinghorn JR, Dalmay T: Embryonic temperature affects muscle fibre recruitment in adult zebrafish: genome-wide changes in gene and microRNA expression associated with the transition from hyperplastic to hypertrophic growth phenotypes. J Exp Biol. 2009, 212: 1781-1793. 10.1242/jeb.029918.View ArticlePubMed
- Johnston IA, Bower NI, Macqueen DJ: Growth and the regulation of myotomal muscle mass in teleost fish. J Exp Biol. 2011, 214: 1617-1628. 10.1242/jeb.038620.View ArticlePubMed
- Johnston IA: Environment and plasticity of myogenesis in teleost fish. J Exp Biol. 2006, 209: 2249-2264. 10.1242/jeb.02153.View ArticlePubMed
- Steinbacher P, Marschallinger J, Obermayer A, Neuhofer A, Sänger AM, Stoiber W: Temperature-dependent modification of muscle precursor cell behaviour is an underlying reason for lasting effects on muscle cellularity and body growth of teleost fish. J Exp Biol. 2011, 214: 1791-1801. 10.1242/jeb.050096.PubMed CentralView ArticlePubMed
- Johnston IA, Alderson R, Sandham C, Dingwall A, Mitchell D, Selkirk C, Nickell D, Baker R, Robertson B, Whyte D, Springate J: Muscle fibre density in relation to the colour and texture of smoked Atlantic salmon (Salmo salar L.). Aquaculture. 2000, 189: 335-349. 10.1016/S0044-8486(00)00373-2.View Article
- Star B: The genome sequence of Atlantic cod reveals a unique immune system. Nature. 2011, 477: 207-210. 10.1038/nature10342.PubMed CentralView ArticlePubMed
- INRA Biotechnology Laboratories:http://locus.jouy.inra.fr/,
- Genomic Research in All Salmon:http://web.uvic.ca/grasp/,
- Broad Institute:http://www.broadinstitute.org/,
- Kuhl H, Beck A, Wozniak G, Canario AVM, Volckaert FAM, Reinhardt R: The European sea bass Dicentrarchus labrax genome puzzle: comparative BAC-mapping and low coverage shotgun sequencing. BMC Genomics. 2010, 11: 68-10.1186/1471-2164-11-68.PubMed CentralView ArticlePubMed
- Kuhl H, Sarropoulou E, Tine M, Kotoulas G, Magoulas A, Reinhardt R: A comparative BAC map for the gilthead sea bream (Sparus aurata L.). J Biomed Biotechnol. 2011, 2011: 329025-PubMed CentralView ArticlePubMed
- Sarropoulou E, Kotoulas G, Power DM, Geisler R: Gene expression profiling of Gilthead sea bream during early development and detection of stress-related genes by the application of cDNA microarray Technology. Physiol Genomics. 2005, 23: 182-191. 10.1152/physiolgenomics.00139.2005.View ArticlePubMed
- Ferraresso S, Vitulo N, Mininni AN, Romualdi C, Cardazzo B, Negrisolo E, Reinhart R, Canario AVM, Patarnello T, Bargelloni L: Development and validation of a gene expression oligo microarray for the gilthead sea bream (Sparus aurata). BMC Genomics. 2008, 9: 580-10.1186/1471-2164-9-580.PubMed CentralView ArticlePubMed
- Calduch-Giner JA, Davey G, Saera-Vila A, Houeix B, Talbot A, Prunet P, Cairns MT, Pérez-Sánchez J: Use of microarray technology to assess the time course of liver stress response after confinement exposure in gilthead sea bream (Sparus aurata L.). BMC Genomics. 2010, 11: 193-10.1186/1471-2164-11-193.PubMed CentralView ArticlePubMed
- Fraser BA, Weadick CJ, Janowitz I, Rodd FH, Hughes KA: Sequencing and characterisation of the guppy (Poecilia reticula) transcriptome. BMC Genomics. 2011, 12: 202-10.1186/1471-2164-12-202.PubMed CentralView ArticlePubMed
- Santure AW, Gratten J, Mossman JA, Sheldon BC, Slate J: Characterisation of the transcriptome of a wild great tit Parus major population by next generation sequencing. BMC Genomics. 2011, 12: 283-10.1186/1471-2164-12-283.PubMed CentralView ArticlePubMed
- Vogel H, Altincicek B, Glöckner G, Vilcinskas A: A comprehensive transcriptome and immune-gene repertoire of the lepidopteran model host Galleria mellonella. BMC Genomics. 2011, 12: 308-10.1186/1471-2164-12-308.PubMed CentralView ArticlePubMed
- Sequence Read Archive:http://www.ncbi.nlm.nih.gov/sra,
- Milne I, Bayer M, Cardle L, Shaw P, Stephen G, Wright F, Marshall D: Tablet- next generation sequence assembly visualization. Bioinformatics. 2010, 26: 401-402. 10.1093/bioinformatics/btp666.PubMed CentralView ArticlePubMed
- Clark KA, McElhinny AS, Beckerle MC, Gregorio CC: Striated muscle cytoarchitecture: an intricate web of form and function. Annu Rev Cell Dev Biol. 2002, 18: 637-706. 10.1146/annurev.cellbio.18.012502.105840.View ArticlePubMed
- Hoshijima M: Mechanical stress–strain sensors embedded in cardiac cytoskeleton: Z disk, titin and associated structure. Am J Physiol Heart Circ Physiol. 2006, 290: 1313-1325.View Article
- Ensembl BioMart:http://www.ensembl.org/biomart/,
- Wall PK, Leebens-Mack J, Chanderbali AS, Barakat A, Wolcott A, Liang H, Landherr L, Tomsho LP, Hu Y, Carlson JE, Ma H, Schuster SC, Soltis DE, Soltis PS, Altman N, dePamphilis CW: Comparison of next generation sequencing Technologies for transcriptome characterization. BMC Genomics. 2009, 10: 347-10.1186/1471-2164-10-347.PubMed CentralView ArticlePubMed
- Meyer E, Aglyamova GV, Wang S, Buchanan-Carter J, Abrego D, Colbourne JK, Willis BL, Matz MV: Sequencing and de novo analysis of a coral larval transcriptome using 454 GSFlx. BMC Genomics. 2009, 10: 219-10.1186/1471-2164-10-219.PubMed CentralView ArticlePubMed
- Salem M, Rexroad CE, Wang J, Thorgaard GH, Yao J: Characterization of the rainbow trout transcriptome using Sanger and 454-pyrosequencing approaches. BMC Genomics. 2010, 11: 564-10.1186/1471-2164-11-564.PubMed CentralView ArticlePubMed
- Coppe A, Pujolar JM, Maes GE, Larsen PF, Hansen MM, Bernatchez L, Zane L, Bortolluzzi S: Sequencing, de novo annotation and analysis of the first Anguilla anguilla transcriptome: EelBase opens new perspectives for the study of the critical endangered European eel. BMC Genomics. 2010, 11: 635-10.1186/1471-2164-11-635.PubMed CentralView ArticlePubMed
- Vera JC, Wheat CW, Fescemyer HW, Frilander MJ, Crawford DL, Hanski I, Marden JH: Rapid transcriptome characterization for a nonmodel organism using 454 pyrosequencing. Mol Ecol. 2008, 17: 1636-1647. 10.1111/j.1365-294X.2008.03666.x.View ArticlePubMed
- Vogiatzi E, Lagnel J, Pakaki V, Louro B, Canario AV, Reinhardt R, Kotoulas G, Magoulas A, Tsigenopoulos CS: In silico mining and characterization of simple sequence repeats from gilthead sea bream (Sparus aurata) expressed sequence tags (EST-SSRs); PCR amplification, polymorphism evaluation and multiplexing and cross-species assays. Mar Genomics. 2011, 4: 83-91. 10.1016/j.margen.2011.01.003.View ArticlePubMed
- Crockford T, Wommack KE, Johnston IA, McAndrew BJ, Mutungi G, Johnson TP: Inter- and intra-specific variation in myosin light chain and troponin I composition in fast muscle fibres from two species of fish (genus Oreochromis) which have different temperature-dependent contractile properties. J Muscle Res Cell Motil. 1991, 12: 439-446. 10.1007/BF01738328.View ArticlePubMed
- Brooks S, Johnston IA: Influence of development and rearing temperature on the distribution, ultrastructure and myosin sub-unit composition of myotomal muscle fibre types in the plaice, Pleuronectes platessa. Mar Biol. 1993, 117: 501-513.
- Johnston IA, Temple GK: Thermal plasticity of skeletal muscle phenotype in ectothermic vertebrates and its significance for locomotory behaviour. J Exp Biol. 2002, 205: 2305-2322.PubMed
- Louro B: Gilthead sea bream (Sparus auratus) and European sea bass (Dicentrarchus labrax) expressed sequenced tags: characterization, tissue-specific expression and gene markers. Mar Genomics. 2010, 3: 179-191. 10.1016/j.margen.2010.09.005.View ArticlePubMed
- Lu J, Peatman E, Wang W, Yang Q, Abernathy J, Wang S, Kucuktas H, Liu Z: Alternative splicing in teleost fish genomes: same-spicies and cross-species analysis and comparisons. Mol Genet Genomics. 2010, 283: 531-539. 10.1007/s00438-010-0538-3.View ArticlePubMed
- Jaillon O: Genome duplication in teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype. Nature. 2004, 431: 946-957. 10.1038/nature03025.View ArticlePubMed
- Johansen SD, Coucheron DH, Andreassen M, Karlsen BO, Furmanek T, Jørgensen TE, Emblem A, Breines R, Nordeide JT, Moum T, Nederbragt AJ, Stenseth NC, Jakobsen KS: Large-scale sequence analyses of Atlantic cod. N Biotechnol. 2009, 25: 263-271. 10.1016/j.nbt.2009.03.014.View ArticlePubMed
- Jeukens J, Renaut S, St-Cyr J, Nolte AW, Bernatchez L: The transcriptomics of sympatric dwarf and normal lake whitefish (Coregonus clupeaformis spp. Salmonidae) divergence as revealed by next-generation sequencing. Mol Ecol. 2010, 19: 5389-5403. 10.1111/j.1365-294X.2010.04934.x.View ArticlePubMed
- Huse SM, Huber JA, Morrison HG, Sogin ML, Welch DM: Accuracy and quality of massively parallel DNA pyrosequencing. Genome Biol. 2007, 8: 143-10.1186/gb-2007-8-7-r143.View Article
- Miller JR, Koren S, Sutton G: Assembly algorithms for next-generation sequencing data. Genomics. 2010, 95: 315-327. 10.1016/j.ygeno.2010.03.001.PubMed CentralView ArticlePubMed
- Basu N, Todgham AE, Ackerman PA, Bibeau MR, Nakano K, Schulte PM, Iwama GK: Heat shock protein genes and their functional significance in fish. Gene. 2002, 95: 173-183.View Article
- Salem M, Silverstein J, Rexroad CE, Yao J: Effect of starvation on global gene expression and proteolysis in rainbow trout (Oncorhynchus mykiss). BMC Genomics. 2007, 8: 328-10.1186/1471-2164-8-328.PubMed CentralView ArticlePubMed
- Bower NI, Taylor RG, Johnston IA: Phasing on muscle gene expression with fasting-induced recovery growth in Atlantic salmon. Front Zool. 2009, 6: 19-10.1186/1742-9994-6-19.View Article
- Kuang S, Kuroda K, Le Grand F, Rudnicki MA: Asymmetric self-renewal and commitment of satellite stem cells in muscle. Cell. 2007, 129: 999-1010. 10.1016/j.cell.2007.03.044.PubMed CentralView ArticlePubMed
- Buas MF, Kadesch T: Regulation of skeletal myogenesis by Notch. Exp Cell Res. 2010, 18: 3028-3033.View Article
- Conboy IM, Rando TA: The regulation of Notch signaling controls satellite cell activation and cell fate determination in postnatal myogenesis. Dev Cell. 2002, 3: 397-409. 10.1016/S1534-5807(02)00254-X.View ArticlePubMed
- Dzeja P, Terzic A: Adenylate kinase and AMP signalling networks: metabolic monitoring, signal communication and body energy sensing. Int J Mol Sci. 2009, 10: 1729-1772. 10.3390/ijms10041729.PubMed CentralView ArticlePubMed
- de Lange P, Ragni M, Silvestri E, Moreno M, Schiavo L, Lombardi A, Farina P, Feola A, Goglia F, Lanni A: Combined cDNA array/RT-PCR analysis of gene expression profile in rat gastrocnemius muscle: relation to its adaptive function in energy metabolism during fasting. FASEB J. 2004, 18: 350-352.PubMed
- Blotta I, Prestinaci F, Mirante S, Cantafora A: Quantitative assay of total dsDNA with PicoGreen reagent and real-time fluorescence detection. Ann Ist Super Sanita. 2005, 41: 119-123.PubMed
- Kumar S, Blaxter ML: Comparing de novo assemblers for 454 transcriptome data. BMC Genomics. 2010, 11: 571-10.1186/1471-2164-11-571.PubMed CentralView ArticlePubMed
- Götz S, García-Gómez JM, Terol J, Williams TD, Nagaraj SH, Nueda MJ, Robles M, Talón M, Dopazo J, Conesa A: High-throughput functional annotation and data mining with the Blast2GO suite. Nucleic Acids Res. 2008, 36: 3420-3435. 10.1093/nar/gkn176.PubMed CentralView ArticlePubMed
- EBI ClustalW:http://www.ebi.ac.uk/Tools/msa/clustalw2/,
- KEGG Automatic Annotation Server:http://www.genome.jp/tools/kaas,
- NCBI BLAST server:http://blast.ncbi.nlm.nih.gov/Blast.cgi,
- Faircloth BC: msatcommander: detection of microsatellites repeat arrays and automated, locus-specific primer design. Mol Ecol Res. 2008, 8: 92-94. 10.1111/j.1471-8286.2007.01884.x.View Article
- Ensembl BLAST server:http://www.ensembl.org/Multi/blastview,
- Spidey mRNA/genome analyser:http://www.ncbi.nlm.nih.gov/spidey/,
- Transcription factor genes & associated conserve noncoding elements database.http://tfcones.fugu-sg.org/index.htm,
- O’Brien KP, Remm M, Sonnhammer ELL: Inparanoid: A Comprehensive Database of Eukaryotic Orthologs. Nucl Acids Res. 2005, 33: 476-480. Inparanoid website http://inparanoid.sbc.su.se/cgi-bin/index.cgiView Article
- Penn O, Privman E, Ashkenazy H, Landan G, Graur D, Pupko T: GUIDANDE: a web server for assessing alignment confidence scores. Nucl Acids Res. 2010, 38: 23-28. Guidence websitehttp://guidance.tau.ac.il/index.htmlView Article
- Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S: MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance and maximum parsimony methods. Mol Biol Evol. 2011, 28: 2731-2739. 10.1093/molbev/msr121.PubMed CentralView ArticlePubMed
- Guindon S, Gascuel O: A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol. 2003, 52: 696-704. 10.1080/10635150390235520. PhyML South France Bioinformatic platform http://www.atgc-montpellier.frView ArticlePubMed
- R statistical package.www.R-project.org,
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.