Characterisation of Elevenin-Vc1 from the Venom of Conus victoriae: A Structural Analogue of α-Conotoxins

Elevenins are peptides found in a range of organisms, including arthropods, annelids, nematodes, and molluscs. They consist of 17 to 19 amino acid residues with a single conserved disulfide bond. The subject of this study, elevenin-Vc1, was first identified in the venom of the cone snail Conus victoriae (Gen. Comp. Endocrinol. 2017, 244, 11–18). Although numerous elevenin sequences have been reported, their physiological function is unclear, and no structural information is available. Upon intracranial injection in mice, elevenin-Vc1 induced hyperactivity at doses of 5 or 10 nmol. The structure of elevenin-Vc1, determined using nuclear magnetic resonance spectroscopy, consists of a short helix and a bend region stabilised by the single disulfide bond. The elevenin-Vc1 structural fold is similar to that of α-conotoxins such as α-RgIA and α-ImI, which are also found in the venoms of cone snails and are antagonists at specific subtypes of nicotinic acetylcholine receptors (nAChRs). In an attempt to mimic the functional motif, Asp-Pro-Arg, of α-RgIA and α-ImI, we synthesised an analogue, designated elevenin-Vc1-DPR. However, neither elevenin-Vc1 nor the analogue was active at six different human nAChR subtypes (α1β1εδ, α3β2, α3β4, α4β2, α7, and α9α10) at 1 µM concentrations.


Introduction
Cone snail venoms are a rich source of toxins, including many hormone-like peptides [1][2][3]. These peptides typically function by targeting membrane proteins such as ion channels, transporters, and G-protein-coupled receptors (GPCRs) [4][5][6]. New classes of conopeptides continue to attract interest as potentially new pharmacological tools or therapeutic leads. This is exemplified by cone snail insulins, which are active at the human insulin receptor [7][8][9], and recently discovered somatostatin analogues in the venom of fish-hunting cone snails [10]. The focus of this study, elevenin-Vc1, was discovered in the venom of the mollusc-hunting cone snail Conus victoriae [1,3], but its structure, function, and physiological target have not yet been characterised.
The elevenins are a class of peptides identified in a wide range of organisms, including the venom of cone snails [3,11,12]. However, they are understudied with respect to both their structure and function. Elevenin was first identified in the L11 abdominal ganglion [11] and the head ganglia [13] of the sea hare, Aplysia californica [14]. Immunochemistry studies found the elevenin-precursor peptide in the salivary gland, suggesting that this gland could be another neurosecretory tissue region in addition to the brain [15,16]. A range of biological activities has been reported for elevenin peptides in different organisms. In the nematode Caenorhabditis elegans elevenin suppresses olfactory plasticity, thus regulating population density as the plasticity of olfactory behaviour is dependent on population density [17]. Elevenin from the annelid polychaete worm Platynereis dumerilii was found to be a neurotransmitter [18]. Bauknecht and Jekely identified P. dumerilii elevenin activity against two orphan receptors (elevenin receptors) [19] which belong to a group of related GPCRs that are generally activated by ligands containing one or two disulfide bonds [20]. Studies on the brown planthopper insect Nilaparvata lugens identified the expression of elevenin, mainly in the brain, abdominal integument, and salivary glands; knockdown of elevenin or its receptor NlA42 expression by RNA interference caused melanisation [16,21]. In the red flour beetle, Tribolium castaneum, knockdown of the elevenin receptor caused mortality during the larval stage [22]. Elevenin regulates cuticle melanisation in N. lugens through the tyrosine-mediated cuticle melanism pathway [16,21]. It is also involved in salivary secretion and the early stages of ovarian development in the black-legged tick, Ixodes scapularis [15]. Despite these extensive functional studies, no structural information is available on these peptides.
Elevenin-Vc1 was identified in the venom of C. victoriae (Uniprot id: A0A0F7YZQ7) [3]. Based on its amino acid sequence similarity to the other elevenins, elevenin-Vc1 was predicted to mimic the function of the prey's neuropeptide but was not characterised functionally, and there is no structural information available on this peptide or other members of the elevenin family. The structure of this peptide would be useful to understand its interaction with the receptor(s) and its physiological function. In this study, we have determined the solution structure of elevenin-Vc1 using nuclear magnetic resonance (NMR) spectroscopy and investigated its activity in mice and against different human nicotinic acetylcholine receptors (nAChRs).

Elevenin-like Peptides Are Widely Distributed in Protostomes
Elevenins are widely distributed in diverse organisms such as molluscs, nematodes, annelids, and arthropods ( Figure 1). They are 17-19 amino acid residues in length, with two conserved Cys residues at positions 5 and 14 and a Gly residue at position 16 (numbering is based on the elevenin-Vc1 sequence). Arg15 is the next most highly conserved residue (93%), followed by Ala11 (85%), Ala19 (83%), and Pro12 (81%). The presence of similar peptides to elevenin-Vc1 in several other cone snails and molluscs, as documented in Figure 1, suggests that this class of peptide may have a common function across these species.

Peptide Synthesis, Oxidative Folding, Purification, and Mass Spectroscopy
Elevenin-Vc1 was produced by solid-phase peptide synthesis (SPPS). Peptide oxidation was performed in 50 mM HEPES (pH 7.25), with a final yield of~10%. The observed mass of the oxidised peptide ( Figure S1) was two Da less than that of the non-oxidised peptide, confirming the monomeric state of the peptide with one disulfide bond formed. When the crude elevenin-Vc1 was subjected to oxidative folding in 100 mM NH 4 HCO 3 , peptide dimers formed through intermolecular disulfide bonds.

Sequential Assignments of NMR Apectra of Elevenin-Vc1
A pH titration on elevenin-Vc1 was performed to identify the optimal pH for NMRbased structural studies. One-dimensional proton (1D 1 H) and two-dimensional total correlation spectroscopy (2D TOCSY) and 2D nuclear Overhauser spectroscopy (2D NOESY) spectra were recorded at different pH values ranging from 2 to 6.5 ( Figure S2). pHdependent chemical shift changes were observed for a few amide protons in the 1D 1 H NMR spectra recorded at pH 2 to 4 ( Figure S2). When the sample pH was increased from 4.5 to 6.5, only Ile3 showed a small upfield shift (−0.04 ppm), and the Cys14 and Ala19 amide peaks were partially overlapped, whereas the Arg2 amide peak was completely broadened ( Figure S2). Despite these small chemical shift changes, the peaks were wellresolved at pH 4.5 and the dispersion of amide proton chemical shifts was not substantially affected by varying the pH from 4.5 to 6.5, indicating good conformational stability of the peptide in this pH range. However, slight peak broadening was observed at higher pH as a consequence of the pH-dependent peak broadening that generally occurs at a higher temperature in peptides [23]. Therefore, the NMR spectra for sequential NMR assignments and the structure calculation were acquired at pH 4.5 and 283 K ( Figure S3).
Stereo-specific assignments for Hβ protons of Asp4, Cys5, Phe8, Phe10, and Cys14 and methyl groups of Ile3 were determined using 2D DQF-COSY, 2D NOESY (mixing time 50 ms) and 2D TOCSY spectra. The peaks for the Ile13 methyl groups have degenerate chemical shifts (Table S1). In addition, the Val7 and Val9 methyl proton resonances have upfield chemical shifts (Table S1 and Figure S3 Figure 3A shows a stereoview of an ensemble of 20 lowest-energy structures refined in an implicit solvent using XPLOR-NIH. Except for the N-and C-termini, the structures were well-defined, with backbone and sidechain heavy atom RMSD values of 0.4 Å and 0.8 Å, respectively, over residues 3-16 (Table 1). Elevenin-Vc1 has a five-residue bend encompassing Cys5 to Val9 and a C-terminal helix spanning residues Pro12 to Arg15 ( Figure  3B). The turn and helix are connected by the disulfide bridge between Cys5 and Cys14. The N-terminal (Arg1-Ile3) and C-terminal (Gly16-Ala19) regions, and the Val9-Ala11 region between the turn and the helix, are less well defined. NOEs between the Ile13 amide proton and the Pro12 δ protons were observed, indicating that the proline is acting as a helix inducer rather than a helix breaker [25,26]. The inter-residue NOEs between the Ala11 Hα and Pro12 Hδ resonances are diagnostic of a trans conformation of the peptide bond between Ala11 and Pro12 ( Figure 4A,B). The trans conformation was confirmed using Pro12 13 Cβ and 13 Cγ chemical shifts ( Figure 4C); the chemical shift difference between the 13 Cβ and 13 Cγ resonances was 4.3 ppm [27], confirming the trans geometry of the Ala11-Pro12 peptide bond (reference values are < 5 ppm for the trans conformation and ~10 ppm for cis [27]). The structural ensemble has been deposited in the Protein Data Bank [28] (PDB 1d: 8F04).  Figure 3A shows a stereoview of an ensemble of 20 lowest-energy structures refined in an implicit solvent using XPLOR-NIH. Except for the N-and C-termini, the structures were well-defined, with backbone and sidechain heavy atom RMSD values of 0.4 Å and 0.8 Å, respectively, over residues 3-16 (Table 1). Elevenin-Vc1 has a five-residue bend encompassing Cys5 to Val9 and a C-terminal helix spanning residues Pro12 to Arg15 ( Figure 3B). The turn and helix are connected by the disulfide bridge between Cys5 and Cys14. The N-terminal (Arg1-Ile3) and C-terminal (Gly16-Ala19) regions, and the Val9-Ala11 region between the turn and the helix, are less well defined. NOEs between the Ile13 amide proton and the Pro12 δ protons were observed, indicating that the proline is acting as a helix inducer rather than a helix breaker [25,26]. The inter-residue NOEs between the Ala11 Hα and Pro12 Hδ resonances are diagnostic of a trans conformation of the peptide bond between Ala11 and Pro12 ( Figure 4A,B). The trans conformation was confirmed using Pro12 13 Cβ and 13 Cγ chemical shifts ( Figure 4C); the chemical shift difference between the 13 Cβ and 13 Cγ resonances was 4.3 ppm [27], confirming the trans geometry of the Ala11-Pro12 peptide bond (reference values are < 5 ppm for the trans conformation and 10 ppm for cis [27]). The structural ensemble has been deposited in the Protein Data Bank [28] (PDB 1d: 8F04).

Valine-phenylalanine Interactions
Atoms oriented perpendicular to the plane of aromatic rings in peptide/protein structures experience a shielding effect; therefore, the resonances from these atoms appear at a higher field compared to random coil chemical shifts [29][30][31]. In the 1 H NMR spectrum, the Val7 methyl peak was shifted by −0.63 ppm, and the Val9 methyl peak was shifted by

Valine-Phenylalanine Interactions
Atoms oriented perpendicular to the plane of aromatic rings in peptide/protein structures experience a shielding effect; therefore, the resonances from these atoms appear at a higher field compared to random coil chemical shifts [29][30][31]. In the 1 H NMR spectrum, the Val7 methyl peak was shifted by −0.63 ppm, and the Val9 methyl peak was shifted by −0.48 ppm compared to the random coil chemical shifts ( Figure 5A and Table S1). In the elevenin-Vc1 structure, one of the two methyl groups from Val7 and Val9 residues is oriented perpendicular to the aromatic rings of Phe8 and Phe10 residues, respectively ( Figure 5B). The measured distances between the centroid of the Phe8/Phe10 aromatic ring and the Val7/Val9 methyl groups were 4.1 Å and 4.2 Å, respectively ( Figure 5B).

Valine-phenylalanine Interactions
Atoms oriented perpendicular to the plane of aromatic rings in peptide/protein structures experience a shielding effect; therefore, the resonances from these atoms appear at a higher field compared to random coil chemical shifts [29][30][31]. In the 1 H NMR spectrum, the Val7 methyl peak was shifted by −0.63 ppm, and the Val9 methyl peak was shifted by −0.48 ppm compared to the random coil chemical shifts ( Figure 5A and Table S1). In the elevenin-Vc1 structure, one of the two methyl groups from Val7 and Val9 residues is oriented perpendicular to the aromatic rings of Phe8 and Phe10 residues, respectively (Figure 5B). The measured distances between the centroid of the Phe8/Phe10 aromatic ring and the Val7/Val9 methyl groups were 4.1 Å and 4.2 Å, respectively ( Figure 5B).   Table S1.

Elevenin-Vc1 Made Mice Hyperactive upon Intracranial Injection
Elevenin-Vc1 was identified in the venom of C. victoriae [3], but its function in the venom is not known. Many conotoxins display activity in mice when injected intracranially [32]. Here we tested the effect of elevenin-Vc1 in mice by intracranial injection (Table 2). Intracranial injection of elevenin-Vc1 in mice at doses of 5 or 10 nmol caused increased activity over the 1 h period of observation compared to negative control animals, suggesting the presence of a mammalian target receptor for this peptide. The peptide did not produce any observable changes to normal or depolarisation-induced intracellular Ca 2+ levels in cultured mouse dorsal root ganglion cells, indicating that the target receptor is not expressed in these sensory neurons (data not shown). Table 2. Summary of results from intracranial mouse injections. Mice were 14-17 days old and weighed 6.7-8.9 g. Injections were carried out in duplicate unless stated otherwise.
The orientations of the Ala11, Pro12, and Ile13 residues in the elevenin-Vc1 structure are similar to that of Asp5, Pro6, and Arg7 residues in α-RgIA and α-ImI ( Figure 7C). This similarity is intriguing given that α-RgIA and α-ImI contain an additional disulfide bond.
Most conotoxins have multiple disulfide bonds, although there are some that lack disulfide bonds [2,[43][44][45][46]. Those peptides with multiple disulfide bonds are relatively stable compared to those lacking them [44]. Elevenins are slightly longer than α-conotoxins, but they share some structural similarities with that class of peptide toxins. These similarities in structure may be an example of convergent evolution, but their different targets-nAChRs for α-conotoxins and most likely GPCRs for elevenins-imply that they have evolved independently. There are some examples of α-conotoxins with single disulfide bonds. A mutant of the cyclic analogue of α-conotoxin Vc1.1 lacking one disulfide bond was shown to be similar to wild-type cyclic α-conotoxin cVc1.1 (with two disulfide bonds) in activity against human α9α10 nAChR and the Cav2.2 and Cav2.3 calcium channels [47]. The removal of the first disulfide bond (C2-C8) also eliminated the formation of multiple isomers due to disulfide bond shuffling during peptide folding. Czon1107 from C. zonatus, which contains a single disulfide bridge, is an allosteric, non-competitive inhibitor of hα3β4 and α7 nAChRs [48,49].
An analogue of elevenin-Vc1, L11 from A. californica, is known to act on GPCRs [20], and since elevenin-Vc1 has 73% amino acid sequence identity to L11, it will be interesting to test elevenin-Vc1 on these receptors. Information on the structure of peptide toxins alone and in complex with their receptors is valuable in guiding the development of target-specific analogues that could be useful as pharmacological tools [50][51][52][53][54].

Conclusions
We have described the first structure of an elevenin peptide (elevenin-Vc1), determined in solution using NMR spectroscopy. The structure has a short helix near the C-terminus stabilised by a single disulfide bond, and the structural fold is similar to the α-conotoxins such as α-RgIA, α-ImI, and α-Vc1.1. The behaviour induced by the elevenin-Vc1 in mice suggests that the peptide may act on receptors in the brain. Both elevenin-Vc1 and the analogue elevenin-Vc1-DPR lack inhibitory activity against human nAChR subtypes, but further studies focusing on the effect of elevenin-Vc1 on a range of GPCRs and other membrane protein receptors in different cell lines may provide insights into its primary target(s). Crude peptide at 0.3 mg/mL concentration was subjected to air oxidation in 50 mM HEPES buffer (pH 7.3) containing 50 mM NaCl for 18 h. Disulfide bond formation was confirmed by LC-MS. The reaction was centrifuged to separate non-dissolved material, or precipitate, and the supernatant was filtered and purified by reversed-phase highperformance liquid chromatography (RP-HPLC). A 5-60% linear gradient of solvent B (99.9% ACN, 0.1% TFA) against solvent A (0.1% TFA in water) over 60 min through a C18 column (Vydac, 10 × 300 mm) was used to purify the peptides. Liquid chromatographymass spectroscopy (LC-MS) was used to assess the purity and mass of peptides ( Figure S1).

NMR Sample Preparation
The sample for NMR experiments was prepared by dissolving lyophilised peptide at a concentration of 0.8 mM in 95% H 2 O/5% 2 H 2 O and adjusting the pH to 4.5 by adding 0.1 M NaOH. The sample for the deuterium exchange experiment was prepared by dissolving the peptide at 0.5 mM concentration in 100% 2 H 2 O and adjusting the pH to 4.5 (uncorrected for deuterium isotope effect) by adding 0.1 M 2 HCl or 0.1 M NaO 2 H.

NMR Data Collection, Processing, and Analysis
One-dimensional 1 H NMR spectra were recorded at 283 K, with different pH conditions ranging from 2 to 7. The two-dimensional (2D) total correlation spectroscopy (TOCSY, 80 ms and 100 ms spin lock times) and nuclear Overhauser spectroscopy (NOESY, 50 ms and 300 ms mixing times) NMR spectra were recorded at different temperatures for residue-specific and sequential NMR assignments, respectively. Double-quantum filtered correlation spectroscopy (DQF-COSY), TOCSY (30 ms spin lock time), and NOESY (50 ms mixing time) NMR spectra were recorded for stereospecific assignments of Hβ protons.
[ 13 C-1 H]-HSQC and [ 15 N-1 H]-HSQC spectra were recorded for obtaining 13 C and 15 N chemical shifts, respectively. Spectra were processed using TopSpin (version 3.5), and chemical shifts and NOE assignments were made in CcpNmr software.

Structure Determination
The structure of elevenin-Vc1 was calculated using distance restraints derived from intensities of NOE cross-peaks in a 2D NOESY spectrum with a mixing time of 300 ms. In total, 246 NOE-derived distances were converted into unambiguous structural restraints. Dihedral angle restraints were estimated from 3 J HN-Hα coupling constants measured from 1D 1 H or 2D DQF-COSY spectra, using the following ranges: −120 ± 40 • when 3 J HN-Hα ≥ 8Hz and −65 ± 25 • when 3 J HN-Hα ≤ 6Hz. Three distance restraints were used for disulfide connectivity as follows: 2, 3, and 3Å for S(i)-S(j), S(i)-Cβ(j), and S(j)-Cβ(i), respectively. No hydrogen bond restraints were included in the structure determination. Initial structures were calculated by CYANA (version 3.97) using only NOE distance constraints [55]. The structures generated using CYANA were refined by simulated annealing, first in vacuo, then in an implicit solvent using the EEFx force field in XPLOR-NIH (version 2.45) [56]. Ramachandran statistics were generated using MolProbity Ramachandran analysis in the protein structure validation suite version 1.5 [57]. The structure figures were generated using the PyMOL Molecular graphics system, version 2.2.0 Schrödinger, LLC (http://www.pymol.org, accessed on 4 December 2022).

Mice behavioural Experiments
Swiss Webster mice (14-17 days old; 6.7-8.9 g) were injected intracranially (IC) with different doses of synthetic peptide dissolved in 12 µL of 0.9% NaCl, and behaviour was observed for 1 h to determine differences between treated and control animals. The screening was performed in duplicate at an initial dose of 10 nmol and repeated at lower concentrations until there were no observable differences from negative control animals (injected with 12 µL of 0.9% NaCl). All experiments involving the use of animals were approved by the Institutional Animal Care and Use Committee of the University of Utah (IACUC #14-08018). While the peptide used in these and other functional assays was highly pure ( Figure S1), concentrations were not corrected for peptide content (likely to be around 70% of the dry weight following purification by RP-HPLC).

Assay against Heterologous Human nAChRs Expressed in Xenopus laevis Oocytes
All procedures were approved by the University of Wollongong Animal Ethics Committee (project number AE2003). Female Xenopus laevis were sourced from Nasco (Fort Atkinson, WI, USA), and a maximum of four frogs were kept in a 15 L aquarium at 20-26 • C with a 12 h light/dark cycle. Oocytes were obtained from five-year-old frogs anaesthetised with 1.7 mg/mL ethyl 3-aminobenzoate methanesulfonate (pH 7.4 with NaHCO 3 ). Stage V-VI oocytes (Dumont's classification; 1200-1300 µm diameter) were defolliculated with 1.5 mg/mL collagenase Type II (Worthington Biochemical Corp., Lakewood, NJ, USA) at room temperature for 1-2 h in OR-2 solution containing (in mM): 82.5 NaCl, 2 KCl, 1 MgCl 2 and 5 HEPES at pH 7.4.
The human muscle nAChR clones (α1, β1, δ, and ε) were purchased from Integrated DNA Technologies (Coralville, IA, USA), the human (h) α3, α9, α10, β2, and β4 clones were purchased from OriGene (Rockville, MD, USA), and all were subsequently inserted into the pT7TS vector. The human α4 and α7 clones were obtained from Prof. Jon Lindstrom (University of Pennsylvania, Philadelphia, PA, USA). Plasmid constructs of the human nAChR clones were linearised for in vitro mRNA synthesis using mMessage mMachine transcription kit (AMBION, Forster City, CA, USA).
Electrophysiological recordings were carried out 2-5 days after cRNA microinjection. Two-electrode voltage clamp recordings of X. laevis oocytes expressing human nAChRs were performed at room temperature (21-24 • C) using a GeneClamp 500B amplifier and pClamp9 software interface (Molecular Devices, Sunnyvale, CA, USA) at a holding potential −80 mV. Voltage-recording and current-injecting electrodes were pulled from GC150T-7.5 borosilicate glass (Harvard Apparatus, Holliston, MA) and filled with 3 M KCl, giving resistances of 0.3-1 MΩ. Due to the Ca 2+ permeability of α9α10 nAChRs, 100 µM BAPTA-AM incubation was carried out before recording to prevent the activation of X. laevis oocyte endogenous Ca 2+ -activated chloride channels. Oocytes expressing hα9α10 nAChRs were perfused with ND115 solution containing (in mM): 115 NaCl, 2.5 KCl, 1.8 CaCl 2 , and 10 HEPES at pH 7.4, whereas oocytes expressing all other nAChR subtypes were perfused with ND96 solution using a continuous Legato 270 push/pull syringe pump perfusion system (KD Scientific, Holliston, MA, USA) at a rate of 2 mL/min in an OPC-1 perfusion chamber of < 20 µL volume (Automate Scientific, Berkeley, CA, USA).
Initially, oocytes were washed briefly with ND115/ND96 solution, followed by three applications of acetylcholine (ACh) at a half-maximal excitatory ACh concentration (EC 50 ) for the nAChR subtypes (3 µM for hα4β2, 5 µM for hα1β1δε, 6 µM for hα3β2 and hα9α10, 100 µM for hα7 and 300 µM for hα3β4) [58]. Washout with bath solution was done for 3 min between ACh applications. Oocytes were incubated with peptides for 5 min with the perfusion system turned off, followed by co-application of ACh and compound with flowing bath solution. All peptide solutions were prepared in ND115/ND96 + 0.1% bovine serum albumin (BSA). Incubation with 0.1% BSA was performed to ensure that the BSA and the pressure of the perfusion system had no effect on nAChRs. Peak current amplitudes before (ACh alone) and after (ACh + peptide) compound incubation were measured using Clampfit version 10.7.0.3 software (Molecular Devices, Sunnyvale, CA, USA), where the ratio of ACh + peptide-evoked current amplitude to ACh alone-evoked current amplitude was used to assess the activity of the compounds at the nAChRs. All electrophysiological data were pooled (n = 6-8) and represent means ± standard deviation (SD). Data analysis was performed using GraphPad Prism 5 (GraphPad Software, La Jolla, CA, USA).