Limits...
Target-Driven Evolution of Scorpion Toxins.

Zhang S, Gao B, Zhu S - Sci Rep (2015)

Bottom Line: By using maximum-likelihood models of codon substitution, we analyzed molecular adaptation in scorpion sodium channel toxins from a specific species and found ten positively selected sites, six of which are located at the core-domain of scorpion α-toxins, a region known to interact with two adjacent loops in the voltage-sensor domain (DIV) of sodium channels, as validated by our newly constructed computational model of toxin-channel complex.This work presents an example of atypical co-evolution between animal toxins and their molecular targets, in which toxins suffered from more prominent selective pressure from the channels of their competitors.Our discovery helps explain the evolutionary rationality of gene duplication of toxins in a specific venomous species.

View Article: PubMed Central - PubMed

Affiliation: Group of Peptide Biology and Evolution, State Key Laboratory of Integrated Management of Pest Insects &Rodents, Institute of Zoology, Chinese Academy of Sciences, 1 Beichen West Road, Chaoyang District, 100101 Beijing, China.

ABSTRACT
It is long known that peptide neurotoxins derived from a diversity of venomous animals evolve by positive selection following gene duplication, yet a force that drives their adaptive evolution remains a mystery. By using maximum-likelihood models of codon substitution, we analyzed molecular adaptation in scorpion sodium channel toxins from a specific species and found ten positively selected sites, six of which are located at the core-domain of scorpion α-toxins, a region known to interact with two adjacent loops in the voltage-sensor domain (DIV) of sodium channels, as validated by our newly constructed computational model of toxin-channel complex. Despite the lack of positive selection signals in these two loops, they accumulated extensive sequence variations by relaxed purifying selection in prey and predators of scorpions. The evolutionary variability in the toxin-bound regions of sodium channels indicates that accelerated substitutions in the multigene family of scorpion toxins is a consequence of dealing with the target diversity. This work presents an example of atypical co-evolution between animal toxins and their molecular targets, in which toxins suffered from more prominent selective pressure from the channels of their competitors. Our discovery helps explain the evolutionary rationality of gene duplication of toxins in a specific venomous species.

No MeSH data available.


Mapping of PSSs of the M. martensii α-toxins on the secondary structure of BmKM1.Residue variability of PSSs is shown in blue for the core-domain and orange for the NC-domain. Relative frequencies of each PSS are shown in bracket. Only the PSSs convergently predicted by M2a and M8 are shown here. The core-domain refers to two loops (boxed in grey): J-loop, the region between Cys2 and Cys3, structurally preceding the α-helix; B-loop, the region between Cys5 and Cys6, structurally linking two β-strands. The NC-domain includes the N-turn and C-tail.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4595728&req=5

f1: Mapping of PSSs of the M. martensii α-toxins on the secondary structure of BmKM1.Residue variability of PSSs is shown in blue for the core-domain and orange for the NC-domain. Relative frequencies of each PSS are shown in bracket. Only the PSSs convergently predicted by M2a and M8 are shown here. The core-domain refers to two loops (boxed in grey): J-loop, the region between Cys2 and Cys3, structurally preceding the α-helix; B-loop, the region between Cys5 and Cys6, structurally linking two β-strands. The NC-domain includes the N-turn and C-tail.

Mentions: The completion of the whole genome sequencing of the first scorpion species (M. martensii)17 allows us to gather nearly all scorpion α-toxin genes of this species to analyze their evolutionary selection pattern and potential driving force. A total of 29 non-redundant toxin sequences were collected and aligned, which cover all three different pharmacological subgroups, e.g. α-like BmKM1 and BmKM10, insect-specific BmKαIT1, and classic BmKαTX11 (Fig. S1). To test positive selection of the M. martensii α-toxin multigene family, we employed maximum likelihood (ML) models of codon substitution to identify potential PSSs (sites with a nonsynonymous to synonymous substitution rate ratio, dN/dS = ω, > 1 significantly)18. As shown in Table 1, the ML estimates (MLEs) under M0 predicts that all amino acid sites have a ω of 0.85, approximately equals 1, indicative of neutral selection. However, M0 fits the data worse than other models, suggesting that ω values (i.e. selective pressure) may be heterogeneous among sites. The MLEs under M2a suggest that 27% of sites are under positive selection with ω = 2.47 (Table 1). The likelihood ratio test (LRT) statistic between M1a and M2a is 16.6 (2Δl), much greater than critical values from a χ2 distribution with degree of freedom = 2 (p < 0.001), indicating the presence of PSSs. M2a did identify 10 PSSs with P ≥ 0.9 by methods of the Bayes Empirical Bayes (BEB) (Table 1) and the Naïve Empirical Bayes (NEB) (Table S1). The test using M7 and M8 models results in a highly similar conclusion. Although M8 predicts two more PSSs, 10 of them are identical to those predicted by M2a (Table 1 and Table S1). It is remarkable that these PSSs are scattered in the primary structure of the toxin, yet they cluster in the functional regions previously characterized19202122, in which six (15E, 17A, 18R, 37Q, 38W and 39V, numbered according to BmKM1) are seated on the core-domain comprising two loops: B- and J-loop; and others on the NC-domain (Fig. 1). When compared with the PSSs identified previously131415, it was found that the majority of them were overall similar in positions, suggesting that Nav channel toxins from different scorpion species might suffer from similar selective pressure.


Target-Driven Evolution of Scorpion Toxins.

Zhang S, Gao B, Zhu S - Sci Rep (2015)

Mapping of PSSs of the M. martensii α-toxins on the secondary structure of BmKM1.Residue variability of PSSs is shown in blue for the core-domain and orange for the NC-domain. Relative frequencies of each PSS are shown in bracket. Only the PSSs convergently predicted by M2a and M8 are shown here. The core-domain refers to two loops (boxed in grey): J-loop, the region between Cys2 and Cys3, structurally preceding the α-helix; B-loop, the region between Cys5 and Cys6, structurally linking two β-strands. The NC-domain includes the N-turn and C-tail.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4595728&req=5

f1: Mapping of PSSs of the M. martensii α-toxins on the secondary structure of BmKM1.Residue variability of PSSs is shown in blue for the core-domain and orange for the NC-domain. Relative frequencies of each PSS are shown in bracket. Only the PSSs convergently predicted by M2a and M8 are shown here. The core-domain refers to two loops (boxed in grey): J-loop, the region between Cys2 and Cys3, structurally preceding the α-helix; B-loop, the region between Cys5 and Cys6, structurally linking two β-strands. The NC-domain includes the N-turn and C-tail.
Mentions: The completion of the whole genome sequencing of the first scorpion species (M. martensii)17 allows us to gather nearly all scorpion α-toxin genes of this species to analyze their evolutionary selection pattern and potential driving force. A total of 29 non-redundant toxin sequences were collected and aligned, which cover all three different pharmacological subgroups, e.g. α-like BmKM1 and BmKM10, insect-specific BmKαIT1, and classic BmKαTX11 (Fig. S1). To test positive selection of the M. martensii α-toxin multigene family, we employed maximum likelihood (ML) models of codon substitution to identify potential PSSs (sites with a nonsynonymous to synonymous substitution rate ratio, dN/dS = ω, > 1 significantly)18. As shown in Table 1, the ML estimates (MLEs) under M0 predicts that all amino acid sites have a ω of 0.85, approximately equals 1, indicative of neutral selection. However, M0 fits the data worse than other models, suggesting that ω values (i.e. selective pressure) may be heterogeneous among sites. The MLEs under M2a suggest that 27% of sites are under positive selection with ω = 2.47 (Table 1). The likelihood ratio test (LRT) statistic between M1a and M2a is 16.6 (2Δl), much greater than critical values from a χ2 distribution with degree of freedom = 2 (p < 0.001), indicating the presence of PSSs. M2a did identify 10 PSSs with P ≥ 0.9 by methods of the Bayes Empirical Bayes (BEB) (Table 1) and the Naïve Empirical Bayes (NEB) (Table S1). The test using M7 and M8 models results in a highly similar conclusion. Although M8 predicts two more PSSs, 10 of them are identical to those predicted by M2a (Table 1 and Table S1). It is remarkable that these PSSs are scattered in the primary structure of the toxin, yet they cluster in the functional regions previously characterized19202122, in which six (15E, 17A, 18R, 37Q, 38W and 39V, numbered according to BmKM1) are seated on the core-domain comprising two loops: B- and J-loop; and others on the NC-domain (Fig. 1). When compared with the PSSs identified previously131415, it was found that the majority of them were overall similar in positions, suggesting that Nav channel toxins from different scorpion species might suffer from similar selective pressure.

Bottom Line: By using maximum-likelihood models of codon substitution, we analyzed molecular adaptation in scorpion sodium channel toxins from a specific species and found ten positively selected sites, six of which are located at the core-domain of scorpion α-toxins, a region known to interact with two adjacent loops in the voltage-sensor domain (DIV) of sodium channels, as validated by our newly constructed computational model of toxin-channel complex.This work presents an example of atypical co-evolution between animal toxins and their molecular targets, in which toxins suffered from more prominent selective pressure from the channels of their competitors.Our discovery helps explain the evolutionary rationality of gene duplication of toxins in a specific venomous species.

View Article: PubMed Central - PubMed

Affiliation: Group of Peptide Biology and Evolution, State Key Laboratory of Integrated Management of Pest Insects &Rodents, Institute of Zoology, Chinese Academy of Sciences, 1 Beichen West Road, Chaoyang District, 100101 Beijing, China.

ABSTRACT
It is long known that peptide neurotoxins derived from a diversity of venomous animals evolve by positive selection following gene duplication, yet a force that drives their adaptive evolution remains a mystery. By using maximum-likelihood models of codon substitution, we analyzed molecular adaptation in scorpion sodium channel toxins from a specific species and found ten positively selected sites, six of which are located at the core-domain of scorpion α-toxins, a region known to interact with two adjacent loops in the voltage-sensor domain (DIV) of sodium channels, as validated by our newly constructed computational model of toxin-channel complex. Despite the lack of positive selection signals in these two loops, they accumulated extensive sequence variations by relaxed purifying selection in prey and predators of scorpions. The evolutionary variability in the toxin-bound regions of sodium channels indicates that accelerated substitutions in the multigene family of scorpion toxins is a consequence of dealing with the target diversity. This work presents an example of atypical co-evolution between animal toxins and their molecular targets, in which toxins suffered from more prominent selective pressure from the channels of their competitors. Our discovery helps explain the evolutionary rationality of gene duplication of toxins in a specific venomous species.

No MeSH data available.