Optimized deep-targeted proteotranscriptomic profiling reveals unexplored Conus toxin diversity and novel cysteine frameworks

Abstract

Cone snails are predatory marine gastropods characterized by a sophisticated venom apparatus responsible for the biosynthesis and delivery of complex mixtures of cysteine-rich toxin peptides. These conotoxins fold into small highly structured frameworks, allowing them to potently and selectively interact with heterologous ion channels and receptors. Approximately 2,000 toxins from an estimated number of >70,000 bioactive peptides have been identified in the genus Conus to date. Here, we describe a high-resolution interrogation of the transcriptomes (available at www.ddbj.nig.ac.jp) and proteomes of the diverse compartments of the Conus episcopatus venom apparatus. Using biochemical and bioinformatic tools, we found the highest number of conopeptides yet discovered in a single Conus specimen, with 3,305 novel precursor toxin sequences classified into 9 known superfamilies (A, I1, I2, M, O1, O2, S, T, Z), and identified 16 new superfamilies showing unique signal peptide signatures. We were also able to depict the largest population of venom peptides containing the pharmacologically active C-C-CC-C-C inhibitor cystine knot and CC-C-C motifs (168 and 44 toxins, respectively), as well as 208 new conotoxins displaying odd numbers of cysteine residues derived from known conotoxin motifs. Importantly, six novel cysteine-rich frameworks were revealed which may have novel pharmacology. Finally, analyses of codon usage bias and RNA-editing processes of the conotoxin transcripts demonstrate a specific conservation of the cysteine skeleton at the nucleic acid level and provide new insights about the origin of sequence hypervariablity in mature toxin regions.

Deep venomics of the Pseudonaja genus reveals inter- and intra-specific variation

“…Only 25% (P. textilis), 27% (P. aspidorhyncha) and 24% (P. nuchalis) of the reads could not be annotated ( Figure 2). The N50 and N90 values for the raw reads and processed toxins of the three transcriptomes (Table 1) as well as average Phred+33 quality scores of >30 [30], indicated that the processed transcripts could be utilised for small peptide and protein identification from each venom proteome. For cysteine-rich secreted proteins (CRVPs), reads were manually assembled to the complete precursor with both reads containing methionine as the first residue.…”

Targeted Sequencing of Venom Genes from Cone Snail Genomes Improves Understanding of Conotoxin Molecular Evolution

“…For the conotoxin sequences, it is known that traditional assemblers perform poorly in reconstructing all potential conotoxin gene copies (Lavergne et al 2015;Phuong et al 2016). To ameliorate this issue, we reassembled conotoxin genes using the assembler PRICE v1.2 (Ruby et al 2013), which employs iterative mapping and extension using paired read information to build out contigs from initial seed sequences.…”

An efficient transcriptome analysis pipeline to accelerate venom peptide discovery and characterisation

“…combined with the much greater sequencing depth provided by the platform (Schirmer et al, 2015) has already started to be used to sequence venom gland transcriptomes producing much larger datasets (Barghi et al, 2015;Lavergne et al, 2015).…”

“…While our reanalyses have been performed on 454 pyrosequencing data, which has been the most commonly used platform for the analysis of cone snail venom gland transcriptomes thus far (Prashanth et al, 2014), more recent studies (Barghi et al, 2015;Lavergne et al, 2015) have begun to use Illumina sequencing technology (Schirmer et al, 2015), generating shorter reads but more data and uncovering further sequence diversity (Lavergne et al, 2015). Since ConoSorter is able to handle these larger datasets after assembly (Lavergne et al, 2015), it follows that our pipeline is well placed to probe larger datasets.…”

“…Since ConoSorter is able to handle these larger datasets after assembly (Lavergne et al, 2015), it follows that our pipeline is well placed to probe larger datasets. The method can be broadened to the identification of non-secreted sequences by skipping the signal peptide requirement and proceeding directly to clustering sequences using CD-HIT, and can be adapted to transcriptomes of different venomous animals, once the Regular Expression and pHMM models are trained for their venom sequences (Lavergne et al, 2015). The H2 superfamily delineated by Lavergne et al appears to be the product of a sequencing error.…”