Micro and macroevolution of sea anemone venom phenotype

Abstract

Venom is a complex trait with substantial inter- and intraspecific variability resulting from strong selective pressures acting on the expression of many toxic proteins. However, understanding the processes underlying toxin expression dynamics that determine the venom phenotype remains unresolved. By interspecific comparisons we reveal that toxin expression in sea anemones evolves rapidly and that in each species different toxin family dictates the venom phenotype by massive gene duplication events. In-depth analysis of the sea anemone, Nematostella vectensis, revealed striking variation of the dominant toxin (Nv1) diploid copy number across populations (1-24 copies) resulting from independent expansion/contraction events, which generate distinct haplotypes. Nv1 copy number correlates with expression at both the transcript and protein levels with one population having a near-complete loss of Nv1 production. Finally, we establish the dominant toxin hypothesis which incorporates observations in other venomous lineages that animals have convergently evolved a similar strategy in shaping their venom.

Acontia, a Specialised Defensive Structure, Has Low Venom Complexity in Calliactis polypus

“…This indicates that the acontia of C. polypus delivers a high dose of a single toxin rather than a concoction of toxins. The over-representation of a single NaTx is in strong agreement with an ancestral state reconstruction analysis of toxin gene expression that found NaTx to be the dominant toxin expressed in the Metridioidea species analysed, which included C. polypus [ 38 ]. The low complexity of the C. polypus acontia venom confirms the validity of the ‘dominant toxin hypothesis’, where a single toxin or toxin family dominates the venom phenotype in sea anemones [ 38 ].…”

“…The presence of this cysteine-rich sequence in the venom of two separate species strongly suggests that it is a functional protein, most likely with toxin activity. Similar to many sea anemone peptide toxins [ 11 , 38 ], multiple codons from the Unknown 12C protein were found to be evolving in a manner consistent with the action of negative selection, which was most prominent for the cysteine encoding codons. The cysteine framework has no similarity to any known toxin in sea anemones [ 45 , 46 ], although it has an identical motif at two locations in the scaffold that match the last three C-X frames of ShK toxins (C-X3-C-X2-C) [ 47 ] and sea anemone 8 toxins [ 48 ].…”

Venom trade-off shapes interspecific interactions, physiology, and reproduction

“…Nematostella interactions with prey. To analyze the gut contents of Nematostella we used whole-body genomic DNA extractions using a protocol described in our previous study (4). Total genomic DNA was extracted using the AllPrep DNA/RNA Kit (Qiagen, USA) for 10 or more individuals across five locations spanning the Atlantic coast of North America (Cresent Beach, Nova Scotia; Saco, Maine; Wallis Sand, New Hampshire; Sippewissett, Massachusetts; Ft. Fisher, NC) during the months of March, June, and September in 2016.…”

“…A major component of venom are toxic proteins, henceforth referred to as toxins, which are directly encoded by the venomous animal's genome. These toxin-encoding genes have been shown to evolve under strong selective pressure at both the sequence and gene expression levels (3)(4)(5), due to their important ecological implications as well as their likely high metabolic cost of production (6,7). Furthermore, toxins are directly involved in predator-prey interactions and have been shown to have specialized functional roles.…”

“…1C and table S1; analysis of variance (ANOVA) P = 0.0001], which is the source of the common N. vectensis laboratory strain (11,17). At the protein level, Nv1 was undetectable in FL compared to being among the top five most abundant protein from NC Nematostella whole-animal proteome (4). We find that the loss of Nv1 in FL is not an anecdotal phenomenon, with additional sampling of Nematostella 12 km from the original site, revealing that all individuals (n = 10) have 0 to 2 Nv1 diploid copies (table S1).…”