Fast skeletal muscle transcriptome of the gilthead sea bream (Sparus aurata) determined by next generation sequencing

Abstract

Background: 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.

Results: 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.

Conclusions: 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.

Figure 1

(A) Barr chart summarizing the percentage of isotigs showing any change in their sequence compared with 80 NCBI sequences (B) Distribution of the differences between NCBI and transcriptome sequences in categories of mismatch, insertion and deletion. Classification of the discrepancies between sequences was carried out by analysis of the ClustalW alignment.

Figure 2

Myofibrillar genes represented in the transcriptome mapped onto a reconstruction of a half sarcomere based on published models for filaments and M-line[25]and z-disc structure[26]. Numbers on the right side of the gene name represents isotig length (bp), isotig mean coverage and percentage of identity with the zebrafish orthologue.

Figure 3

Dot plot pairwise comparison of reads contribution to the isotigs formation from each experimental group. Each dot represents a contig with reads from one or both treatments. X vs Y graph illustrate the relation of the number of reads between treatments in function of the region where the dots are placed.

Figure 4

Barr charts summarizing transcripts with significant differences between groups in the number of reads mapped. The groups were as follows: 21°C fed (J), fasted 21°C (F), acutely transferred to 33°C fed (H) and acutely transferred to 11°C (L). All genes represented have been selected from Table 5 and have a FDR ≤ 0.01.