On the importance of oxidative folding in the evolution of conotoxins: cysteine codon preservation through gene duplication and adaptation

Abstract

Conotoxin genes are among the most rapidly evolving genes currently known; however, despite the well-established hypervariability of the intercysteine loops, the cysteines demonstrate significant conservation, with a site-specific codon bias for each cysteine in a family of conotoxins. Herein we present a novel rationale behind the codon-level conservation of the cysteines that comprise the disulfide scaffold. We analyze cysteine codon conservation using an internal reference and phylogenetic tools; our results suggest that the established codon conservation can be explained as the result of selective pressures linked to the production efficiency and folding of conotoxins, driving the conservation of cysteine at the amino-acid level. The preservation of cysteine has resulted in maintenance of the ancestral codon in most of the daughter lineages, despite the hypervariability of adjacent residues. We propose that the selective pressures acting on the venom components of cone snails involve an interplay of biosynthetic efficiency, activity at the target receptor and the importance of that activity to effective prey immobilization. Functional redundancy in the venom can thus serve as a buffer for the energy expenditure of venom production.

Figure 1. Codon usage statistics for the cysteine residues in two classes of four-loop conotoxin genes.

(A) shows the codon bias for the -conotoxins, and (B) shows the codon bias for the -conotoxins. The lines above show the characteristic disulfide connectivity of the cysteine residues. While both - and -conotoxins maintain the same general cysteine scaffold, they are readily distinguished by the precursor sequences (both signal and propeptide), length of the intercysteine loops, as well as the tendency of the -conotoxins to be considerably more hydrophobic. The GenBank accession numbers of all sequences used are available in File S1.

Figure 2. Comparison of in-frame and out-of-frame cysteine codons for - and -conotoxins.

(A) shows the data for the -conotoxins, and (B) shows the data for the -conotoxins. White bars show in-frame cysteine codons, green bars show cysteine codons in the +1 position, and blue bars show cysteine codons in the +2 position. ‘Site in sequence’ starts counting at the first cysteine in the mature toxin region, and considers only sites that have at least 100 sequences in the alignment. These numeric designations are preserved in Figures 3, S1 and S2. Sites that are specific to one class of sequence are marked with that class (e.g. 14() contains data for only the -conotoxins). Cysteine sites are marked with **; these sites are also readily identifiable by the large white bars that indicate in-frame cysteine codons.

Figure 3. Codon bias for - and -conotoxins.

(A) shows the data for the -conotoxins, and (B) shows the data for the -conotoxins. Values presented are the weighted average of values (sensu , Eq. 2) for codons encoding amino acids that can be represented by two (and only two) codons. Because the number of residues in the intercysteine loops differ between the - and -Conotoxins, sites that are specific to one class of conotoxins are followed by the class to which they are specific (e.g. site 14() only shows data for -conotoxins). Yellow bars and **indicate cysteine sites. Analysis was only conducted on sites with at least one hundred sequences in the alignment, at least thirty of which were residues that could be encoded by two codons. A value of 1 indicates that only a single codon is used, and a value of 2 indicates that there is no codon bias among the two possible codons for each residue. The same data for the other redundancy classes is presented in File S1 (Figures S1 and S2), along with the GenBank accession number of each sequence included in the analysis.

Figure 4. Phylogenetic trees of a represenatitve subclade of -conotoxin genes used in this study, showing the codon usage patterns for each cysteine.

Each tree shows the codon usage for a single cysteine, and is labeled accordingly. Shaded lines indicate TGT, and hollow lines indicate TGC. Bootstrap values are also shown on the tree for Cysteine 1. Each branch in the tree for Cysteine 1 is labeled with the GenBank accession number of the sequence and the Conus species. Phylogenetic trees for the entire O1-Superfamily are shown in File S1.