Parallel molecular evolution in an herbivore community

Abstract

Numerous insects have independently evolved the ability to feed on plants that produce toxic secondary compounds called cardenolides and can sequester these compounds for use in their defense. We surveyed the protein target for cardenolides, the alpha subunit of the sodium pump, Na(+),K(+)-ATPase (ATPα), in 14 species that feed on cardenolide-producing plants and 15 outgroups spanning three insect orders. Despite the large number of potential targets for modulating cardenolide sensitivity, amino acid substitutions associated with host-plant specialization are highly clustered, with many parallel substitutions. Additionally, we document four independent duplications of ATPα with convergent tissue-specific expression patterns. We find that unique substitutions are disproportionately associated with recent duplications relative to parallel substitutions. Together, these findings support the hypothesis that adaptation tends to take evolutionary paths that minimize negative pleiotropy.

Fig. 1

The phylogenetic pattern of amino acid substitution at sites implicated in cardenolide-binding for ATPα1 in taxa that feed on Apocynaceae and outgroups. Numbered columns correspond to sites for which site-directed mutagenesis or protein structure analysis has suggested a role in cardenolide-binding (fig. S1). Gray sites correspond to those with only structural evidence for a role in cardenolide-binding. Dots represent identity with the consensus sequence. Letters in red are specific substitutions characterized in functional assays (fig. S1). Underlined are substitutions that require at least two nonsynoymous substitutions relative to the ancestral sequence. Orange-shaded rows correspond to specialists known to sequester cardenolides. Gray-shaded rows represent species that either are not known to sequester cardenolides and/or are nonspecialists only occasionally found on Apocynaceae. The cladogram represents the accepted species branching order (table S6), and branch lengths are not meaningful. Green circles represent inferred duplications of ATPα1. Numbers below correspond to the number of inferred substitutions associated with use of Apocynaceae. Red boxes correspond to parallel substitutions (observed in more than one independent lineage). Blue boxes correspond to unique substitutions (observed in only one lineage).

Fig. 2

Cardenolide-binding pocket structure and molecular docking–simulation results. (A) Structure of the cardenolide-binding pocket of ATP1A1 for Sus scrofa bound to the cardenolide ouabain (in red) from (31). Green lines represent possible hydrogen bonds between the protein and ouabain. (B to D) Docking simulation results for the substitutions (B) N122H, (C) N122Y, and (D) Q111V. In each case, we show the best docking position for ouabain docked to the native protein (in red) versus the protein with one of three simulated substitutions (in blue). The best docking position is defined as that closest to the coordinates of ouabain in the published cocrystal structure (31). The atomic surfaces of the derived amino acid side chains at position 111 and 122 are detailed in gold. N122H and N122Y sterically block ouabain from entering the binding pocket [root mean square deviations (RMSDs) from cocrystal position are 5.5 and 6.9 Å, respectively; RMSDs from the best wild-type docking position are 5.6 and 6.9 Å, respectively]. Three other substitutions (C104Y, R880S, and R972Q) are predicted to have similar effects (table S2). In contrast, Q111V has more subtle effects on the position of ouabain (RMSD is 2.98 Å from the cocrystal structure and 0.14 Å from the best wild-type docking position) but likely disrupts a stabilizing H-bond (A).

Fig. 3

Native protein–docking simulations and tissue-specific gene expression for ATPα1 duplicates in (A) dogbane beetles and (B) milkweed stem weevils. (Left) Best docking results for ouabain onto pig ATP1A1 (gray ligand) and estimated structures for ATPα1A (red ligand) and ATPα1B (blue ligand) for each species (33). For both species, the best dockings of ouabain to the predicted structure of native ATPα1A are within 1 Å RMSD of the best docking to pig ATP1A1 (table S3). In contrast, the best dockings of ouabain to putatively resistant ATPα1B copies for dogbane beetles (A) and milkweed stem weevils (B) are displaced by 5.9 and 6.2 Å RMSD, respectively. (Right) Tissue-specific expression means (with two standard errors) of estimates for three unrelated individuals. Positive values indicate higher expression of the putatively resistant copy, ATPα1B. Analysis of variance (ANOVA) reveals significant variation in expression pattern among tissues (dogbane beetles: F = 74.3, P = 6 × 10⁻⁵; milkweed stem weevils: F = 21.1, P = 4 × 10⁻⁴). A Tukey-Kramer test reveals significant differences in all pairwise comparisons except head versus muscle (dogbane beetles, gut-head: P = 6 × 10⁻⁵; gut-muscle: P = 3 × 10⁻⁴; head-muscle: P = 0.08; weevils, gut-head: P = 4 × 10⁻⁴; gut-muscle: P = 3 × 10⁻³; head-muscle: P = 0.36).