True Lies: Using Proteomics to Assess the Accuracy of Transcriptome-Based Venomics in Centipedes Uncovers False Positives and Reveals Startling Intraspecific Variation in Scolopendra subspinipes

Abstract

Centipede venoms have emerged as a rich source of novel bioactive compounds. However, most centipede species are commonly considered too small for venom extraction and transcriptomics is likely to be an attractive way of probing the molecular diversity of these venoms. Examining the venom composition of Scolopendra subspinipes, we test the accuracy of this approach. We compared the proteomically determined venom profile with four common toxin transcriptomic toxin annotation approaches: BLAST search against toxins in UniProt, lineage-specific toxins, or species-specific toxins and comparative expression analyses of venom and non-venom producing tissues. This demonstrated that even toxin annotation based on lineage-specific homology searches is prone to substantial errors compared to a proteomic approach. However, combined comparative transcriptomics and phylogenetic analysis of putative toxin families substantially improves annotation accuracy. Furthermore, comparison of the venom composition of S. subspinipes with the closely related S. subspinipes mutilans revealed a surprising lack of overlap. This first insight into the intraspecific venom variability of centipedes contrasts the sequence conservation expected from previous findings that centipede toxins evolve under strong negative selection. Our results highlight the importance of proteomic data in studies of even comparably well-characterized venoms and warrants caution when sourcing venom from centipedes of unknown origin.

Keywords

• centipede
• Scolopendra subspinipes
• venom
• intraspecific variation
• toxin annotation
• transcriptomics
• proteomics