Predatory robber flies (Diptera, Asilidae) have been suspected to be venomous due to their ability to overpower well-defended prey. However, details of their venom composition and toxin arsenal remained unknown. Here, we provide a detailed characterization of the venom system of robber flies through the application of comparative transcriptomics, proteomics and functional morphology. Our results reveal asilid venoms to be dominated by peptides and non-enzymatic proteins, and that the majority of components in the crude venom is represented by just ten toxin families, which we have named Asilidin1–10. Contrary to what might be expected for a liquid-feeding predator, the venoms of robber flies appear to be rich in novel peptides, rather than enzymes with a putative pre-digestive role. The novelty of these peptides suggests that the robber fly venom system evolved independently from hematophagous dipterans and other pancrustaceans. Indeed, six Asilidins match no other venom proteins, while three represent known examples of peptide scaffolds convergently recruited to a toxic function. Of these, members of Asilidin1 closely resemble cysteine inhibitor knot peptides (ICK), of which neurotoxic variants occur in cone snails, assassin bugs, scorpions and spiders. Synthesis of one of these putative ICKs, U-Asilidin₁-Mar1a, followed by toxicity assays against an ecologically relevant prey model revealed that one of these likely plays a role as a neurotoxin involved in the immobilization of prey. Our results are fundamental to address these insights further and to understand processes that drive venom evolution in dipterans as well as other arthropods.

Cone snails produce a fast-acting and often paralyzing venom that is usually injected into their prey or predator through a hypodermic needle-like modified radula tooth. Many diverse compounds are found in their venom including small molecules, peptides and enzymes. However, peptidic toxins called conotoxins (10–40 residues and 2–4 disulfide bonds) largely dominate these cocktails. These disulfide rich toxins are very valuable pharmacological tools for investigating the function of ions channels, G-protein coupled receptors, transporters and enzymes. Here, we report on the synthesis, structure determination and biological activities of two α-conotoxins, CIA and CIB, found in the predatory venom of the piscivorous species Conus catus. CIA is a typical 3/5 α-conotoxin that blocks the rat muscle type nAChR with an IC₅₀ of 5.7 nM. Interestingly, CIA also inhibits the neuronal rat nAChR subtype α3β2 with an IC₅₀ of 2.06 μM. CIB is a 4/7 α-conotoxin that blocks rat neuronal nAChR subtypes, including α3β2 (IC₅₀ = 128.9 nM) and α7 (IC₅₀ = 1.51 μM). High resolution NMR structures revealed typical α-conotoxin folds for both peptides. We also investigated the in vivo effects of these toxins on fish, since both peptides were identified in the predatory venom of C. catus. Consistent with their pharmacology, CIA was highly paralytic to zebrafish (ED₅₀ = 110 μg/kg), whereas CIB did not affect the mobility of the fish. In conclusion, CIA likely participates in prey capture through muscle paralysis, while the putative ecological role of CIB remains to be elucidated.

Cone snails produce a fast-acting and often paralyzing venom, largely dominated by disulfide-rich conotoxins targeting ion channels. Although disulfide-poor conopeptides are usually minor components of cone snail venoms, their ability to target key membrane receptors such as GPCRs make them highly valuable as drug lead compounds. From the venom gland transcriptome of Conus miliaris, we report here on the discovery and characterization of two conopressins, which are nonapeptide ligands of the vasopressin/oxytocin receptor family. These novel sequence variants show unusual features, including a charge inversion at the critical position 8, with an aspartate instead of a highly conserved lysine or arginine residue. Both the amidated and acid C-terminal analogues were synthesized, followed by pharmacological characterization on human and zebrafish receptors and structural investigation by NMR. Whereas conopressin-M1 showed weak and only partial agonist activity at hV1bR (amidated form only) and ZFV1a1R (both amidated and acid form), both conopressin-M2 analogues acted as full agonists at the ZFV2 receptor with low micromolar affinity. Together with the NMR structures of amidated conopressins-M1, -M2 and -G, this study provides novel structure-activity relationship information that may help in the design of more selective ligands.