In silico identification of the sea squirt selenoproteome

BACKGROUND: Computational methods for identifying selenoproteins have been developed rapidly in recent years. However, it is still difficult to identify the open reading frame (ORF) of eukaryotic selenoprotein gene, because the TGA codon for a selenocysteine (Sec) residue in the active centre of selenoprotein is traditionally a terminal signal of protein translation. Although the identification of selenoproteins from genomes through bioinformatics methods has been conducted in bacteria, unicellular eukaryotes, insects and several vertebrates, only a few results have been reported on the ancient chordate selenoproteins. RESULTS: A gene assembly algorithm SelGenAmic has been constructed and presented in this study for identifying selenoprotein genes from eukaryotic genomes. A method based on this algorithm was developed to build an optimal TGA-containing-ORF for each TGA in a genome, followed by protein similarity analysis through conserved sequence alignments to screen out selenoprotein genes form these ORFs. This method improved the sensitivity of detecting selenoproteins from a genome due to the design that all TGAs in the genome were investigated for its possibility of decoding as a Sec residue. Using this method, eighteen selenoprotein genes were identified from the genome of Ciona intestinalis, leading to its member of selenoproteome up to 19. Among them a selenoprotein W gene was found to have two SECIS elements in the 3'-untranslated region. Additionally, the disulfide bond formation protein A (DsbA) was firstly identified as a selenoprotein in the ancient chordates of Ciona intestinalis, Ciona savignyi and Branchiostoma floridae, while selenoprotein DsbAs had only been found in bacteria and green algae before. CONCLUSION: The method based on SelGenAmic algorithm is capable of identifying eukaryotic selenoprotein genes from their genomes. Application of this method to Ciona intestinalis proves its successes in finding Sec-decoding TGA from large-scale eukaryotic genome sequences, which fills the gap in our knowledge on the ancient chordate selenoproteins.