A Novel Inhibitor of α9α10 Nicotinic Acetylcholine Receptors from Conus vexillum Delineates a New Conotoxin Superfamily

Conotoxins (CTxs) selectively target a range of ion channels and receptors, making them widely used tools for probing nervous system function. Conotoxins have been previously grouped into superfamilies according to signal sequence and into families based on their cysteine framework and biological target. Here we describe the cloning and characterization of a new conotoxin, from Conus vexillum, named αB-conotoxin VxXXIVA. The peptide does not belong to any previously described conotoxin superfamily and its arrangement of Cys residues is unique among conopeptides. Moreover, in contrast to previously characterized conopeptide toxins, which are expressed initially as prepropeptide precursors with a signal sequence, a ‘‘pro’’ region, and the toxin-encoding region, the precursor sequence of αB-VxXXIVA lacks a ‘‘pro’’ region. The predicted 40-residue mature peptide, which contains four Cys, was synthesized in each of the three possible disulfide arrangements. Investigation of the mechanism of action of αB-VxXXIVA revealed that the peptide is a nicotinic acetylcholine receptor (nAChR) antagonist with greatest potency against the α9α10 subtype. 1H nuclear magnetic resonance (NMR) spectra indicated that all three αB-VxXXIVA isomers were poorly structured in aqueous solution. This was consistent with circular dichroism (CD) results which showed that the peptides were unstructured in buffer, but adopted partially helical conformations in aqueous trifluoroethanol (TFE) solution. The α9α10 nAChR is an important target for the development of analgesics and cancer chemotherapeutics, and αB-VxXXIVA represents a novel ligand with which to probe the structure and function of this protein.


Introduction
Nicotinic acetylcholine receptors (nAChRs) are pentameric ligand-gated ion channels used throughout the animal kingdom for neurotransmission. These receptors are assembled from a,b,c,d and/or e subunits to form multiple receptor subtypes with distinct pharmacological properties [1]. Elucidation of the precise structure and function of various nAChRs is challenging owing to the scarcity of ligands selective for specific receptor subtypes. In an effort to address this, we have systematically examined components of the venoms of carnivorous cone snails for selective nAChR-targeted ligands.
Molluscs of the genus Conus are comprised of .700 species. These marine snails hunt primarily polychaete worms, molluscs or fish. Each cone species produces a cocktail of .100 different compounds that enables prey capture. Despite extensive work, the vast majority of these compounds remains uncharacterized.
Conopeptides are produced in the venom duct of Conus and used offensively to immobilize prey. Their potency and selectivity for various ion channels and receptors have made them excellent pharmacological probes and drug leads [2][3]. The term conotoxin is used to describe the subset of Conus peptides that are rich in Cys residues. Conotoxins are synthesized initially as precursor proteins that are subsequently processed into the mature toxin. Previously characterized Conus peptides have been grouped into gene superfamilies based on similarities in their precursor signal sequences [4]. Within each superfamily, the toxins are grouped according to cysteine frameworks that influence their final threedimensional structure. The toxins are also grouped according to receptor or ion channel target into pharmacological families. Within a given family of conotoxins there is, characteristically, hypervariation in non-Cys residues, which is believed to enable selective action on a given target subtype. Post-translational modification or chemical synthetic modification provides further diversity [5][6].
Toxins characterized to date can be classified into one of 17 superfamilies (see Table 1) [7][8]. The current study characterizes a new conotoxin, from the worm-hunting Conus vexillum, with a unique Cys framework. As the precursor sequence does not align with any of the previously-reported gene superfamilies, this peptide represents a first-in-class compound (see (Table S1)). Total chemical synthesis was carried out to enable pharmacological and structural characterization of this novel toxin. The peptide acts as an antagonist of nicotinic acetylcholine receptors, with greatest potency at the a9a10 nAChR, a subtype expressed in a variety of tissues ranging from immune cells to sperm [9][10].

Ethics Statement
No specific permits were required for the described field studies. No specific permissions were required for Tanmen Qionghai, Hainan Province, China, which is not privately-owned or protected in any way. The field studies did not involve endangered or protected species.

Materials
Specimens of Conus vexillum were collected from the South China Sea off Tanmen Qionghai, Hainan Province, China. Venom ducts were frozen and stored at 280uC. Creator SMART cDNA Library Construction Kit was from CLONTECH Laboratories, Inc. Acetylcholine chloride, atropine, and bovine serum albumin (BSA) were from Sigma. The reverse-phase HPLC analytical Vydac C18 column (5 mm, 4.6 mm6250 mm) and preparative C18 Vydac column (10 mm, 22 mm6250 mm) were from Shenyue. Reagents for peptide synthesis were from GL Biochem. Acetonitrile was from Fisher. Trifluoroacetic acid (TFA) was from Tedia. All other chemicals used were of analytical grade. Clones of rat a2-a7 and b2-b4, as well as mouse muscle a1b1de cDNAs were kindly provided by S. Heinemann (Salk Institute, San Diego, CA). Clones for a9 and a10 were generously provided by A.B. Elgoyen (Instituto de Investigaciones en Ingeniería Genética y Biología Molecular, Buenos Aires, Argentina). Clones of b2 and b3 subunits in the high expressing pGEMHE vector were kindly provided by C.W. Luetje (University of Miami, Miami, FL). Total RNA was extracted from individual ducts and purified as described previously [11]. Venom duct cDNA library construction followed the kit manufacturer's suggested protocol. Briefly, the first-strand cDNA was synthesized with the SMART IV Oligonucleotide and transcriptase. Full-length, double-stranded (ds) cDNA (SMART cDNA) was generated by long-distance PCR. SMART cDNA was ligated into the Sfi I predigested pDNR-LIB vector. The signal and mature peptide sequences of the conotoxin precursors were predicted using online ProP 1.0 Server [12].

Peptide Synthesis
The linear peptide was assembled by solid-phase methodology on an ABI 433A peptide synthesizer using FastMoc (N-(9fluorenyl) methoxycarbonyl) chemistry and standard side-chain protection, except for cysteine residues. Cys residues of the three possible isomers were protected in pairs with either S-trityl on Cys3 and Cys19 (designated aB-VxXXIVA [1,2]), Cys3 and Cys20 (designated aB-VxXXIVA [1,3]), Cys19 and Cys20 (designated aB-VxXXIVA [1,4]) or S-acetamidomethyl on Cys20 and Cys32, Cys19 and Cys32, Cys3 and Cys32, respectively. The peptides were removed from a solid support by treatment with reagent K (TFA / water / ethanedithiol / phenol / thioanisole; 90: 5 : 2.5: 7.5: 5,v / v / v / v / v). The released peptide was precipitated and washed three times with cold ether. A two-step oxidation protocol was used to fold the peptides selectively, as described previously [13]. Briefly, the disulfide bridge between Cys3 and Cys19, Cys3 and Cys20, or Cys19 and Cys20, respectively, was closed by dripping the peptide into an equal volume of 20 mM potassium ferricyanide, 0.1 M Tris, pH 7.5. The solution was allowed to react for 45 min, and the monocyclic peptide was purified by reverse-phase HPLC. Simultaneous removal of the S-acetamidomethyl groups and closure of the disulfide bridge between Cys20 and Cys32, Cys19 and Cys32, or Cys3 and Cys32, respectively, was carried out by iodine oxidation as follows: the monocyclic peptide in HPLC eluent was dripped into an equal volume of iodine (10 mM) in H 2 O:TFA:acetonitrile (74:2:24 by volume) and allowed to react for 10 min. The reaction was terminated by the addition of ascorbic acid, diluted 10-fold with 0.1% TFA, and the bicyclic peptide was purified by HPLC on a reversed-phase C18 Vydac column using a linear gradient of 20-60% B60 in 40 min. Solvent B was 60% ACN, 0.092% TFA, and H 2 O; Solvent A 0.1% TFA in H 2 O. Peptide concentration was measured using absorbance at 280 nm, and calculated using the Beer-Lambert equation and a calculated molar extinction coefficient of 3040 cm 21 M 21 .

cRNA Preparation and Injection
Capped cRNA for the various subunits were made using the mMessage mMachine in vitro transcription kit (Ambion) following linearization of the plasmid. The cRNA was purified using the Qiagen RNeasy kit. The concentration of cRNA was determined by absorbance at 260 nm. Oocytes were injected within one day of harvesting and recordings were made 1-4 days post-injection.

Voltage-clamp Recording
Oocytes were voltage-clamped and exposed to ACh and peptide as described previously [14]. Briefly, the oocyte chamber consisting of a cylindrical well (,30 ml in volume) was gravity perfused at a rate of ,2 ml/min with ND96 buffer (96.0 mM NaCl, 2.0 mM KCl, 1.8 mM CaCl 2 , 1.0 mM MgCl 2 , 5 mM HEPES, pH 7.1-7.5) containing 0.1 mg/ml BSA. The Ba ++ -ND96 had 1.8 mM BaCl 2 in place of CaCl 2 . The membrane potential of the oocytes was clamped at 270 mV. The oocyte was subjected once a minute to a 1 s pulse of 100 mM ACh. In the case of the a9a10 and muscle a1b1de subtypes, there is a 1 s pulse of 10 mM Ach, and for the a7 subtype a 200 mM ACh pulse. For toxin concentrations $10 mM, once a stable baseline was achieved, either ND-96 alone or ND-96 containing conotoxin was applied manually for 5 min prior to the addition of the agonist. All recordings were done at room temperature (,22uC).

Data Analysis
The average of five control responses just preceding a test response was used to normalize the test response to obtain ''% response.'' Each data point of a dose-response curve represents the average 6 S.E. of at least three oocytes. The dose-response data were fit to the equation, % response = 100/[1+ ([toxin]/IC50)ˆn H ], where n H is the Hill coefficient, by non-linear regression analysis using GraphPad Prism (GraphPad Software).

NMR Spectroscopy
1 H NMR spectra were recorded on aB-VxXXIVA isomers at a concentration of , 350 mM in 20 mM phosphate/10% 2 H 2 O buffer at pH , 5.8. The 1D spectra were acquired on a Bruker Avance 600 MHz NMR spectrometer equipped with cryogenic probe fitted with a z axis gradient. The NMR spectra were collected at 5uC using the excitation sculpting pulse sequence [15]. Spectra were acquired over 4K data points with 64 scans and a 1 H spectral width of 14 ppm. All spectra were processed in TOPSPIN (version 3.0) and referenced to the water resonance.

Circular Dichroism Analysis
aB-VxXXIVA isomers were dissolved in 20 mM phosphate buffer (pH 5.9) and CD spectra were recorded on a Jasco-815 spectropolarimeter at a concentration of 43 mM at 20uC. Spectra were collected at 0.05 nm intervals over the wavelength range 260-195 nm in a 10 mm pathlength cuvette. Three scans were collected and averaged for each peptide sample with scanning rate of 100 nm/min 21 . The spectra were then smoothed using a thirdorder polynomial function. In order to investigate the effect of trifluoroethanol (TFE) on the conformation, CD spectra for aB-VxXXIVA [1,2] were also acquired following the addition of 10, 20, 50 and 87% TFE. The % a-helix and b-sheet content were calculated from the CD data using the CDPro program [16].

Discovery and Sequence Analysis of a cDNA of the Precursor of aB-VxXXIVA
In general, conotoxins are translated initially as prepropeptide precursors [7], with proteolytic cleavage yielding the final product(s). Peptides in the same superfamily are characterized by highly conserved prepropeptide precursor sequences. This conservation has allowed direct identification of new peptides belonging to a particular superfamily by cDNA sequencing of family or superfamily genes [17]. However, a large fraction of conotoxins present in the Conus genus has yet to be sequenced and several additional families of toxins remain to be identified. Most of the cone snails investigated to date are fish-or mollusc-hunters.
In an effort to discover novel conotoxin families, we examined the worm-hunting C. vexillum. Specimens were collected from the South China Sea and dissected venom ducts were used to construct a cDNA library. Approximately 50 clones from the cDNA library were chosen randomly for sequencing and inspected for previously unreported sequences. In the present study with C. vexillum, several members of the previously characterized aand v-superfamilies were identified. In addition, however, an unusual precursor sequence was noted (Fig. 1, Table 1, GenBank accession number JX297421). A sequence similarity search detected no homology with precursors of the known superfamilies of conotoxins [7] (Table S1). The sequence was analyzed with DNAstar software and online ProP 1.0 Server [12], which indicated a 28residue signal sequence followed by a previously unreported 40residue mature toxin (Table 2 and see also Table S2 for sequence alignment). For other conotoxins, the encoding cDNA has a characteristic three-region organization, including a signal sequence, a ''pro'' region, and the toxin-encoding region [7,18] The generation of the mature toxin requires proteolytic cleavage of the N-terminal prepro-region of the precursor. In contrast to previously characterized conopeptide toxins, the precursor of aB-VxXXIVA has no ''pro'' region. The putative proteolytic processing site between prepropeptide and mature region for conotoxins is usually a basic amino acid (K or R). In contrast, the predicted cleavage site for the aB-VxXXIVA precursor is -LG- (Fig. 1). The predicted mature peptide exhibited a new cysteine framework, not previously reported for conotoxins, C-CC-C ( Table 2 and Table S2). The predicted mature toxin sequence was VRCLEKSGAQPNKLFRPPCCQKGPSFARHSRC-VYYTQSRE.

Chemical Synthesis and Oxidative Folding of VxXXXIVA
With four Cys residues there are three possible disulfide bond arrangements:Cys3-Cys19, Cys20-Cys32 (aB-VxXXIVA [1,2]); Cys3-Cys20, Cys19-Cys32 (aB-VxXXIVA [1,3]), and Cys3-Cys32, Cys19-Cys20 aB-VxXXIVA [1,4] (Fig. 2). Fmoc chemistry was used to synthesize the linear aB-VxXXIVA peptides. The cysteine side chains were protected in pairs with orthogonal protecting groups that could be removed selectively under different conditions, allowing the formation of one disulfide bridge at a time. The first and second, first and third, or second and third cysteine residues were protected with acid-labile groups, which were simultaneously removed during cleavage from the resin. Ferricyanide was used to close the first disulfide bridge. Reversephase HPLC was used to purify the monocyclic peptide; subsequently, the acid-stable acetometomethyl groups were removed from the remaining two cysteines by iodine oxidation, which also closed the second disulfide bridge. The three fully folded peptide isomers were individually purified by HPLC. Electrospray mass spectrometry was utilized to confirm the identity of the products. The monoisotopic masses in Da were: calculated, 4622.27; observed 4622.3 (aB-VxXXIVA [1,2]), 4622.2 (aB-VxXXIVA [1,3]), and 4622.4 (aB-VxXXIVA [1,4]).
Synthetic peptides with these disulfide bond arrangements were used in all subsequent studies.
a9a10 nAChRs are known to be highly permeable to calcium. Entry of Ca ++ through the nAChR elicits a response by Ca ++activated chloride currents. The magnitude of this response in Xenopus oocytes is large and can comprise .90% of the observed current. In contrast, the closely-related divalent cation Ba ++ does not elicit a response. We therefore assessed whetheraB-VxXXIVA blocked the response to ACh when Ba ++ was substituted for Ca ++ in the buffer. Consistent with previous observations, the ACh response of a9a10 nAChRS in Ba ++ ND96 was substantially smaller than that observed in Ca ++ ND96 (data not shown). Using Ba ++ ND96, the a9a10 nAChR was most potently blocked by aB-VxXXIVA [1,2] with an IC 50 of 1.49 mM; under these conditions, aB-VxXXIVA [1,3] had an IC 50 of 3.15 mM and aB-VxXXIVA [1,4] did not potently block the a9a10 nAChR subtype (Fig. 5). Thus, the potency of the aB-VxXXIVA isomers in the presence of Ba ++ was similar to that seen in Ca ++ , consistent with the toxin effect being due to blockade of the nAChR rather than blockade of the Ca ++ -activated Cl 2 channel.

NMR Studies
The 1D 1 H NMR spectra of aB-VxXXIVA isomers in phosphate buffer at pH 5.8 show that the majority of the amide protons fall within the 8.0-9.0 ppm range (Fig. 6); the lack of chemical shift dispersion here and elsewhere in the spectrum indicates that these isomers lack any significant tertiary structure. The same was true at pH 7.0 (Fig. S1, S2, S3). NMR spectra were also acquired in the presence of 3-10 mM CaCl 2 to ascertain whether calcium had any effect on their conformation, but no change in chemical shift dispersion was observed (Fig. S4).

Circular Dichroism Analysis
CD spectra were acquired on all threeaB-VxXXIVA isomers in phosphate buffer. All peptide isomers exhibited minima at around 200 nm (Fig. 7A), indicative of a random coil conformation with no a-helical and b-sheet content, and consistent with our NMR results. As TFE is known to stabilize the a-helical structure in proteins and peptides [19], CD spectra of one of the isomers (aB-VxXXIVA [1,2]) were recorded in increasing concentrations of TFE. Upon addition of 50-85% TFE, aB-VxXXIVA [1,2]    Xenopus oocytes expressing a9a10 nAChR were voltage clamped at -70 mV and subjected to a 1 s pulse of ACh every min as described in Materials and Methods. A representative response in a single oocyte is shown. After control responses to ACh, the oocyte was exposed to 10 mM toxin for 5 min (arrow). After the 5 min toxin application, a response to ACh was measured (a). After 1 min of toxin washout, another response to ACh was measured (b). Note that the response to ACh recovered to control level after 1 min of toxin washout. (B) Concentration response of a9a10 nAChRs exposed to the three different isomers of aB-VxXXIVA (see Fig 2). Values shown in the graph are mean 6 SEM from 3-5 separate oocytes. The IC 50  showed slightly increased ordered structure, as evident by the shift in the minimum towards 208 nm, with some ellipticity also developing at 222 nm (Fig. 7B). The CD data were fitted using three algorithms (CDSSTR, CONTINLL, and SELCON) in CDPro [16]. The outputs obtained from all three algorithms gave very similar values and indicated that the aB-VxXXIVA [1,2] isomer in the presence of 87% TFE had , 42% a-helix, , 8% bstrand and , 50% unordered structure (including turns), whereas, in the absence of TFE it had , 7% a-helix, , 31% b-strand and , 62% unordered structure.

Discussion
Conotoxins are a highly specialized set of disulfide-bonded peptides that are structurally and functionally diverse. Despite this diversity, toxins identified to date may be grouped into approximately 17 gene superfamilies based on conservation of the signal sequence. Within these gene superfamilies, the mature peptides adopt one of 23 patterns of arrangement of cysteine residues. Pharmacological targets within a gene superfamily may differ. For example, in the A superfamily, there are both paralytic and excitotoxic peptides [20].
It is very likely that the previously described superfamilies and Cys frameworks represent only a small fraction of the total chemical space of conotoxins. C. vexillum inhabits waters up to 70 m deep in Hainan province of the South China Sea and feeds on eunicid worms. Here, we describe the discovery and characterization from this species of aB-VxXXIVA, a peptide that differs in substantial aspects from previously-reported conotoxins.
The clone for aB-VxXXIVA was obtained from random sequencing of a cDNA library prepared from venom ducts. The signal sequence of aB-VxXXIVA does not align well with the signal sequence of other known conotoxins. Conservation of the signal sequence has previously been exploited as a means of cloning novel conotoxins from different species of cone snails [21]. The unique signal sequence of aB-VxXXIVA explains why this novel conotoxin has not been detected previously with screening primers designed to recognize known gene superfamilies. The discovery of aB-VxXXIVA expands the known complexity of this group of ion channel-and receptor-targeted ligands. Interestingly, the precursor for aB-VxXXIVA is unique among conotoxins in that it lacks a pro region. The pro region of disulfide-bonded Figure 4. aB-Conotoxin VxXXIVA differentially blocks a9a10, a7, a3b4 and a4b2 nAChRs. nAChR subtypes were expressed as described in Materials and Methods. ''C'' indicates control responses to ACh. Oocytes were then exposed to 10 mM peptide for 5 min, followed by application of ACh. The peptide blocked a9a10 but not a7, a3b4 or a4b2 nAChRs. doi:10.1371/journal.pone.0054648.g004    peptides has been shown to facilitate oxidative folding [22]. Consequently, the pro region of conotoxins was originally proposed as a means by which these peptides could fold into the same three-dimensional scaffold with identical disulfide connectivity [23]. However, evidence from studies with the two-disulfide a-conotoxin GI and three disulfide v-conotoxin MVIIA [24] indicates that the propeptide sequence does not necessarily contribute directly to folding thermodynamics but rather plays a facilitative role when folding is catalyzed by a disulfide isomerase [25]. The pro domain has also been implicated in the secretory pathway of hydrophobic O-superfamily conotoxins [26]. Apparently, such a mechanism is not necessary for the more hydrophilic aB-VxXXIVA.
The mature aB-VxXXIVA toxin is 40 amino acid residues in length and has a previously unreported arrangement of four Cys residues, C-CC-C. We synthesized the three possible disulfide isomers (Fig. 2) and assessed their activity at nAChRs. There are no reported examples of conotoxins that contain a vicinal disulfide bridge, and in the present case, the isomer that was synthesized with linkage between the adjacent second and third Cys residues was inactive. Both of the other two possible disulfide connectivities, aB-VxXXIVA [1,2] with a disulfide connectivity I-II, III-IV, and aB-VxXXIVA [1,3] with I-III, II-IV, blocked aa9a10 nAChRs, with the I-II, III-IV connectivity being 2-fold more active than the I-III, II-IV form.
There is precedent for conotoxins that selectively block the a9a10 over other nAChR subtypes. a-Conotoxin Vc1.1 from C. victoriae and a-conotoxin RgIA from C. regius block the a9a10 nAChR with IC 50 values of 5 and 19 nM, respectively [27]. Vc1.1 also blocks a6/a3b2b3 and a3b4 nAChRs with IC 50 values of 140 and 4200 nM, respectively. Both a-CTx Vc1.1 and a-CTx RgIA were subsequently found to activate GABA B receptors [27,28,29]. In addition, other conotoxins that block nAChRs have also been reported to block voltage-gated ion channels including sodium and potassium channels [30,31]. The IC 50 values for the aB-VxXXIVA isomers against a9a10 nAChRS are in the micromolar range. It is therefore possible that these peptides, in addition to blocking nAChRs, will subsequently be found to act on other ligand-or voltage-gated ion channels.
Although cone snails hunt fish, molluscs and worms, worms are the most common prey. The nAChR subunits from these polychaete marine worms have not been cloned; however, it is of note that aB-VxXXIVA preferentially targets the a9a10 subtype of nAChR. The a9 subunit is a member of the nAChR family although it is more distantly related; indeed it appears to be the closest subunit to the ancestor that gave rise to the nAChR family [32]. Thus, it is tempting to speculate that, among Conus, the worm-hunting species may be particularly likely to produce toxins that target a9 receptors.
The a9 subunit is also of increasing interest in biomedicine. Conotoxins that target the a9 nAChR have been shown to be analgesic [10,27] and to accelerate the recovery of function after nerve injury, possibly through immune-mediated mechanisms [33,34]. In addition, small molecule antagonists of a9a10 nAChRs are analgesic in models of neuropathic pain [35,36].
The a9a10 receptor is present in keratinocytes and is implicated in the pathophysiology of wound healing [37]. Recently it has been shown that the a9 subunit is overexpressed in breast cancer tissue. a9 antagonists reduce tumour growth [38,39]. Moreover, variants of the a9 subunit affect transformation and proliferation of bronchial cells [1,40]. Thus, novel antagonists of the a9a10 nAChR are not only of value to structure/function analysis of this receptor subtype but may also help inform development of novel therapeutics.
The aB-VxXXIVA toxins are atypical among disulfide-bridged conotoxins in showing largely disordered structures in aqueous solution over a range of temperature and pH values. While unusual, this is consistent with structure predictions that show no significant ordered secondary structure for this amino acid sequence (Fig. S5); presumably this is also why the addition of a helix-stabilizing co-solvent like TFE did not induce significant helical structure in aB-VxXXIVA (Fig. 7). There are, however, precedents for disulfide-bridged conotoxins with poorly ordered structures and potent biological activity. Synthetic a2AuIB, for example, formed both a globular (native) isomer and a ribbon isomer upon oxidative refolding, and the ribbon isomer, although having a less well-defined structure, had approximately 10 times greater potency than the native peptide on nACh-evoked currents in rat parasympathetic neurons [41]. More recently, three different disulfide-bridge isomers of the m-conotoxin PIIIA, which contains three disulfides, were found to block the skeletal muscle voltagegated sodium channel Na V 1.4 with similar, yet distinct potencies [42] even though one of them was disordered and gave a poorly dispersed 1 H NMR spectrum akin to those observed for all three aB-VxXXIVA disulfide isomers.
The concept of intrinsically disordered proteins is well established now [43], although it is quite unusual to find a conotoxin containing two disulfide bridges that displays these properties, as in the case of aB-VxXXIVA. It is believed that most intrinsically disordered proteins adopt a more ordered structure upon binding to their physiological targets [44], although evidence is emerging that this is not always the case. It remains to be seen if aB-VxXXIVA becomes more ordered upon binding to a9a10 nAChR. This might be assessed by studying the interaction of ACh-binding proteins engineered to resemble the a9a10 nAChR [45] and/or by creating conformationally constrained analogues of aB-VxXXIVA.  Figure S4 1 H NMR spectra of aB-VxXXIVA [1,2] in the presence and absence of CaCl 2, in 90% H 2 O/10% 2 H 2 O at pH 5.5, acquired on a Varian 600 MHz NMR spectrometer at 22uC. (TIFF) Figure S5 Secondary structure prediction of aB-VxXXIVA isomer, using the PSIPRED protein structure prediction server (http://bioinf.cs.ucl.ac.uk/psipred/). (TIFF)