Venom down under: dynamic evolution of Australian elapid snake toxins

Abstract

Despite the unparalleled diversity of venomous snakes in Australia, research has concentrated on a handful of medically significant species and even of these very few toxins have been fully sequenced. In this study, venom gland transcriptomes were sequenced from eleven species of small Australian elapid snakes, from eleven genera, spanning a broad phylogenetic range. The particularly large number of sequences obtained for three-finger toxin (3FTx) peptides allowed for robust reconstructions of their dynamic molecular evolutionary histories. We demonstrated that each species preferentially favoured different types of α-neurotoxic 3FTx, probably as a result of differing feeding ecologies. The three forms of α-neurotoxin [Type I (also known as (aka): short-chain), Type II (aka: long-chain) and Type III] not only adopted differential rates of evolution, but have also conserved a diversity of residues, presumably to potentiate prey-specific toxicity. Despite these differences, the different α-neurotoxin types were shown to accumulate mutations in similar regions of the protein, largely in the loops and structurally unimportant regions, highlighting the significant role of focal mutagenesis. We theorize that this phenomenon not only affects toxin potency or specificity, but also generates necessary variation for preventing/delaying prey animals from acquiring venom-resistance. This study also recovered the first full-length sequences for multimeric phospholipase A2 (PLA2) 'taipoxin/paradoxin' subunits from non-Oxyuranus species, confirming the early recruitment of this extremely potent neurotoxin complex to the venom arsenal of Australian elapid snakes. We also recovered the first natriuretic peptides from an elapid that lack the derived C-terminal tail and resemble the plesiotypic form (ancestral character state) found in viper venoms. This provides supporting evidence for a single early recruitment of natriuretic peptides into snake venoms. Novel forms of kunitz and waprin peptides were recovered, including dual domain kunitz-kunitz precursors and the first kunitz-waprin hybrid precursors from elapid snakes. The novel sequences recovered in this study reveal that the huge diversity of unstudied venomous Australian snakes are of considerable interest not only for the investigation of venom and whole organism evolution but also represent an untapped bioresource in the search for novel compounds for use in drug design and development.

Figure 1

BEAST maximum credibility ultrametric tree for in-group taxa [12]. Node values indicate 95% highest posterior distributions for calibration points. Posterior probability support values are shown for each node. Species included in this study are indicated in red.

Figure 2

Homology model depicting the locations of positively selected sites from various species, indicated by different colour codes. Multiple sequence alignment of Type II α-ntxs depicting the locations of positively selected sites is also presented. Representative sequences are from Brachyurophis roperi (1. GAHA01000012, 2. GAHA01000013, 3. GAHA01000016), Cacophis squamulosus (4. GAHB01000003, 5. GAHB01000008, 6. GAHB01000008), Drysdalia coronoides (7. FJ752483, 8. FJ752485, 9. FJ752487), Hemiaspis signata (10. GAHF01000010, 11. GAHF01000011, 12. GAHF01000014), Parasuta nigriceps (13. FJ790454, 14. FJ790448, 15. FJ790450), Vermicella annulata (16. GAHJ01000013, 17. GAHJ01000014, 18. GAHJ01000015). Numerical IDs representing species lacking unique mutations are indicated by strikethrough.

Figure 3

Molecular evolution of Type I (aka: short-chain) α-neurotoxins. Three-dimensional homology models of Type I α-neurotoxins from various species, depicting the locations of positively selected sites (Model 8, PP ≥ 0.95, Bayes-Empirical Bayes approach) is presented here. Species are: (A) Brachyurophis roperi, (B) Cacophis squamulosus, (C) Drysdalia coronoides, (D) Hemiaspis signata, (E) Parasuta nigriceps and (F) Vermicella annulata.

Figure 4

Homology model depicting the locations of positively selected sites from various species, indicated by different colour codes. Multiple sequence alignment of Type II α-ntxs depicting the locations of positively selected sites is also presented. Representative sequences are from Acanthophis wellsi (1. GAGZ01000001, 2. GAGZ01000004, 3. GAGZ01000006), Brachyurophis roperi (4. GAHA01000003, 5. GAHA01000001, 6. GAHA01000002), Drysdalia coronoides (7. FJ481928, 8. FJ752461, 9. FJ752459), Echiopsis curta (10. GAHD01000001, 11. GAHD01000004, 12. GAHD01000006), Furina ornata (13. GAHE01000001, 14. GAHE01000009, 15. GAHE01000014), Hemiaspis signata (16. GAHF01000001, 17. GAHF01000005, 18. GAHF01000006), Suta fasciata (19. GAHI01000001, 20. GAHI01000004), 21. Parasuta nigriceps FJ790457, Pseudonaja modesta (22. GAHH01000040, 23. GAHH01000045, 24. GAHH01000046, 25. GAHH01000043, 26. GAHH01000042, 27. GAHH01000035) and Vermicella annulata (28. GAHJ01000009, 29. GAHJ01000010, 30. GAHJ01000016). Numerical IDs representing species lacking unique mutations are indicated by strikethrough.

Figure 5

Molecular evolution of Type II (aka: long-chain) α-neurotoxins. Three-dimensional homology models of Type II α-neurotoxins from various species, depicting the locations of positively selected sites (Model 8, PP ≥ 0.95, Bayes-Empirical Bayes approach). Species are: (A) Acanthophis wellsi, (B) Brachyurophis roperi, (C) Drysdalia coronoides, (D) Echiopsis curta, (E) Furina ornata, (F) Hemiaspis signata, (G) Parasuta nigriceps and (H) Pseudonaja modesta.

Figure 6

Structural and functional evolution of Type III α-neurotoxins. Multiple sequence alignment of Type III α-ntxs depicting the locations of positively selected sites (Model 8, PP ≥ 0.95, Bayes-Empirical Bayes approach) in various species of Australian elapids is presented here. Homology model depicting the locations of positively selected sites from various species, indicated by different colour codes, is also presented. Representative sequences are from Brachyurophis roperi (1. GAHA01000009, 2. GAHA01000010, 3. GAHA01000011), Cacophis squamulosus (4. GAHB01000009, 5. GAHB01000010, 6. GAHB01000011), Furina ornata (7. GAHE01000022, 8. GAHE01000023, 9. GAHE01000020, Pseudonaja modesta (10. GAHH01000009, 11. GAHH01000015, 12. GAHH01000022) and Vermicella annulata (13. GAHJ01000001, 14. GAHJ01000003, 15. GAHJ01000004). Numerical IDs representing species lacking unique mutations are indicated by strikethrough.

Figure 7

Molecular evolution of Type III α-neurotoxins. Three-dimensional homology models of Type III α-neurotoxins from various species, depicting the locations of positively selected sites (Model 8, PP ≥ 0.95, Bayes-Empirical Bayes approach). Species are: (A) Furina ornata, (B) Pseudonaja modesta and (C) Vermicella annulata.

Figure 8

Sequence alignment of natriuretic peptides. (1). P68515 Bothrops insularis, (2). K4J3K2 Azemmiops feae, (3). K4IT20 Azemmiops feae, (4). A8YPR6 Echis ocellatus, (5). Q09GK2 Philodryas olfersii, (6). GAHI01000013 Suta fasciata, (7). P83228 Oxyuranus scutellatus, (8). GAHI01000016 Suta fasciata, (9). A8YPR9 Cerastes cerastes Post-translationally cleaved peptides in shaded in gray.

Figure 9

Phylogenetic reconstruction of the molecular evolutionary history of natriuretic peptides. Non-toxin outgroup sequences (P23582 and P55207) not shown. Representative sequences obtained in this study are shown in red. Node labels indicate posterior probabilities.

Figure 10

Sequence alignment of ‘taipoxin/paradoxin’-like presynaptic complex subunits: α-subunit (1). Q45Z43 Oxyuranus microlepidotus, (2). Q45Z48 Oxyuranus scutellatus, (3). GAGZ01000028 Acanthophis wellsi, (4). A6MFM9 Rhinoplocephalus nigrescens, (5). GAHI01000025 Suta fasciata, (6). B5G6G1 Tropidechis carinatus, β-subunit (7). Q45Z46 Oxyuranus microlepidotus, (8). Q45Z53 Oxyuranus scutellatus, (9). GAGZ01000024 Acanthophis wellsi, (10). GAHI01000027 Suta fasciata and γ-subunit (11). Q4VRI6 Oxyuranus scutellatus, (12). GAGZ01000027 Acanthophis wellsi, (13). Q9PUG7 Austrelaps superbus, (14). GAHI01000030 Suta fasciata.

Figure 11

Phylogenetic reconstruction of the molecular evolutionary history of snake venom Type I phospholipase A₂ toxins. Non-toxin outgroup sequences (Q8JFB2 and Q8JFG2) not shown. Representatives of sequences obtained in this study are shown in red. Node labels indicate posterior probabilities.

Figure 12

Sequence alignment of precursors encoding: dual-domain kunitz (1). B2BS84 Austrelaps labialis, (2). GAHG01000009 Hoplocephalus bungaroides; mono-domain kunitz (3). GAGZ01000019 Acanthophis wellsi, (4). GAGZ01000017 Acanthophis wellsi, (5). GAHB01000016 Cacophis squamulosus, (6). GAHD01000011 Echiopsis curta, (7). GAHG01000008 Hoplocephalus bungaroides, (8). GAHH01000051 Pseudonaja modesta, (9). GAHI01000010 Suta fasciata; dual-domain waprin (10). A7X4K1 Philodryas olfersii; mono-domain waprin (11). GAHC01000021 Denisonia devisi, (12). A7X4J4 Rhabodophis tigrinus, (13). A7X4K7 Philodryas olfersii, (14). A7X4I7 Thrasops jacksonii, (15). B5G6H4 Notechis scutatus, (16). B5G6G8 Oxyuranus scutellatus; kunitz-waprin fusion (17). D3U2B9 Sistrurus catenatus edwardsii, (18). D3U0D3 Sistrurus catenatus tergeminus, (19). GAHB01000034 Cacophis squamulosus, (20). GAHI01000009 Suta fasciata.