Search

The solution structure of robustoxin, the lethal neurotoxin from the Sydney funnel-web spider Atrax robustus, has been determined from 2D 1H NMR data. Robustoxin is a polypeptide of 42 residues cross-linked by four disulphide bonds, the connectivities of which were determined from NMR data and trial structure calculations to be 1-15, 8-20, 14-31 and 16-42 (a 1-4/2-6/3-7/5-8 pattern). The structure consists of a small three-stranded, anti-parallel beta-sheet and a series of interlocking gamma-turns at the C-terminus. It also contains a cystine knot, thus placing it in the inhibitor cystine knot motif family of structures, which includes the omega-conotoxins and a number of plant and animal toxins and protease inhibitors. Robustoxin contains three distinct charged patches on its surface, and an extended loop that includes several aromatic and non-polar residues. Both of these structural features may play a role in its binding to the voltage-gated sodium channel.

High resolution nuclear magnetic resonance (NMR) spectroscopy is the only method available for determining the three-dimensional structures of peptides and proteins in solution at atomic resolution. This article deals with a range of practical considerations associated with such studies, including sample preparation, instrumental setup, one- and two-dimensional NMR methods, interpretation of spectral data, and structure calculations.

B-IV is a 55-residue, crustacean-selective, neurotoxin secreted by Cerebratulus lacteus, a large marine worm found along the northeastern coast of North America. The 3D structure of this molecule in aqueous solution has been determined by 1H NMR spectroscopy at 600 MHz. The molecule has a well-defined helical hairpin structure, with the branches of the hairpin linked by four disulphide bonds. The disulphide connectivities were established from the NMR data to be 1-8/2-7/3-6/4-5, which differed from those determined previously by chemical means, where 1-7 and 2-8 connectivities were found. Each branch of the hairpin is largely alpha-helical, with the helices in the N and C-terminal branches encompassing residues 11 to 23 and 34 to 49, respectively. The loop connecting the branches of the hairpin contains two inverse gamma-turns centred on residues 24 and 25, a type I beta-turn at residues 28 to 31 and a type II beta-turn at residues 30 to 33. Arg17, -25 and -34, which are important for activity, are all on the same face of the molecule, while Trp30, which is also important for activity, is on the opposite face. Structure comparisons show that the B-IV structure is quite similar to those of Rop (ColE1 repressor of primer) and the heat-stable enterotoxin B from Escherichia coli. These structural similarities are discussed in relation to possible mechanisms of action of B-IV.

Nef is a 27 kDa myristylated phosphoprotein expressed early in infection by HIV. The N terminus of Nef is thought to play a vital role in the functions of this protein through its interactions with membrane structures. The solution structure of a 25-residue polypeptide corresponding to the N terminus of Nef (Nef1-25) has been investigated by 1H NMR spectroscopy. In aqueous solution at pH 4.8 and 281 K, this peptide underwent conformational averaging, with Pro13 existing in cis and trans conformations in nearly equal proportions. In methanol solution, however, the peptide adopted a well-defined alpha-helical structure from residues 6 to 22, with the N- and C-terminal regions having a less ordered structure. On the basis of a comparison of chemical shifts and NOEs, it appeared that this helical structure was maintained in aqueous trifluoroethanol (50% v/v) and to a lesser extent in a solution of SDS micelles. When the N-acetyl group was replaced by either an N-myristyl or a free ammonium group, there was little effect on the three-dimensional structure of the peptide in methanol; deamidation of the C terminus also had no effect on the structure in methanol. In water, the myristylated peptide aggregated. The similarity between the sequences of Nef1-25 and melittin is reflected in the similar structures of the two molecules, although the N-terminal helix of melittin is more defined. This similarity in structure raises the possibility that Nef1-25 not only interacts with membranes but also may be capable of disrupting them and causing cell lysis. This type of interaction could contribute at least in part to the killing of bystander cells in lymphoid tissues during HIV infection.

Interleukin-6 (IL-6) is a multifunctional cytokine which is involved in a broad spectrum of activities such as immune defense, hematopoiesis, and the acute phase response, as well as in the pathogenesis of multiple myeloma. A series of murine IL-6 (mIL-6) mutants, H31A, W34A, and H31A/W34A, were constructed to investigate the roles of His31 and Trp34 in the structure, conformational stability, time-dependent aggregation, folding, and spectral properties of mIL-6. The characteristic pH-dependent quenching of fluorescence of mIL-6 at low pH was shown to be caused by an interaction between Trp34 and protonated His31 at low pH and not associated with Trp157. Denaturant-induced equilibrium unfolding experiments monitored by fluorescence and far-UV CD showed that the increased quantum yield and blue shift of the wavelength of the emission maximum observed for mIL-6 at moderate denaturant concentrations were also associated with Trp34, rather than Trp157. The tendency to form aggregation-prone unfolding intermediates, as judged by poor fits to a two-state unfolding mechanism, low m values (slopes of the unfolding curve in the transition region), and the range of denaturant concentrations over which these intermediates formed, was shown to be higher for H31A than mIL-6 but significantly lower for W34A and H31A/W34A. These differences were most pronounced at pH 7.4 and correlated with the tendencies of the proteins to aggregate at high protein concentrations in the absence of denaturant. As judged by the 1H NMR chemical shifts of the aromatic residues, the global conformations of H31A and W34A were not significantly different from that of mIL-6. Nuclear Overhauser effects (NOE) between the side chains of His31 and Trp34 were consistent with the indole side chain of Trp34 being oriented toward the face of the imidazolium side chain of His31, an arrangement consistent with our estimates of a low interaction energy (0.4-0.6 kcal/mol) between these side chains. A shift in the pKa of the His31 side chain in W34A (+0.3 unit) suggested that, in the absence of Trp34, His31 could interact with other residues. Further mutations in this region should yield forms of mIL-6, even less prone to aggregation, which would be more suitable for NMR studies. Mutation of His31 and Trp34 to alanine did not significantly alter the mitogenic activity of the mutants on mouse hybridoma 7TD1 cells, even though the corresponding region of human IL-6 has been shown to be important for biological activity.

The structure-function relationships of the N-type calcium channel blocker, omega-conotoxin GVIA (GVIA), have been elucidated by structural, binding and in vitro and in vivo functional studies of alanine-substituted analogues of the native molecule. Alanine was substituted at all non-bridging positions in the sequence. In most cases the structure of the analogues in aqueous solution was shown to be native-like by 1H NMR spectroscopy. Minor conformational changes observed in some cases were characterized by two-dimensional NMR. Replacement of Lys2 and Tyr13 with Ala caused reductions in potency of more than 2 orders of magnitude in three functional assays (sympathetic nerve stimulation of rat isolated vas deferens, right atrium and mesenteric artery) and a rat brain membrane binding assay. Replacement of several other residues with Ala (particularly Arg17, Tyr22 and Lys24) resulted in significant reductions in potency (<100-fold) in the functional assays, but not the binding assay. The potencies of the analogues were strongly correlated between the different functional assays but not between the functional assays and the binding assay. Thus, the physiologically relevant assays employed in this study have shown that the high affinity of GVIA for the N-type calcium channel is the result of interactions between the channel binding site and the toxin at more sites than the previously identified Lys2 and Tyr13.

The chemical shift assignments and secondary structure of a murine-human chimera, MH35, of leukaemia inhibitory factor (LIF), a 180-residue protein of molecular mass 20 kDa, have been determined from multidimensional heteronuclear NMR spectra acquired on a uniformly 13C, 15N-labelled sample. Secondary structure elements were defined on the basis of chemical shifts, NH-C alpha H coupling constants; medium-range NOEs and the location of slowly exchanging amide protons. The protein contains four alpha-helices, the relative orientations of which were determined on the basis of long-range, interhelical NOEs. The four helices are arranged in an up-up-down-down orientation, as found in other four-helical bundle cytokines. The overall topology of MH35-LIF is similar to that of the X-ray crystallographic structure for murine LIF [Robinson et al. (1994) Cell, 77, 1101-1116]. Differences between the X-ray structure and the solution structure are evident in the N-terminal tail, where the solution structure has a trans-Pro17 compared with the cis-Pro17 found in the crystal structure and the small antiparallel beta-sheet encompassing residues in the N-terminus and CD loop in the crystal structure is less stable.

The three-dimensional structure of synthetic human neuropeptide Y in aqueous solution at pH 3.2 and 37 degrees C was determined from two-dimensional 1H NMR data recorded at 600 MHz. A restraint set consisting of 440 interproton distance restraints inferred from NOEs and 11 backbone and 4 side-chain dihedral angle restraints derived from spin-spin coupling constants was used as input for distance geometry calculations on DIANA and simulated annealing and restrained energy minimization in X-PLOR. The final set of 26 structures is well defined in the region of residues 11-36, with a mean pairwise rmsd of 0.51 A for the backbone heavy atoms (N, C alpha and C) and 1.34 A for all heavy atoms. Residues 13-36 form an amphipathic alpha-helix. The N-terminal 10 residues are poorly defined relative to the helical region, although some elements of local structure are apparent. At least one of the three prolines in the N-terminal region co-exists in both cis and trans conformations. An additional set of 24 distances was interpreted as intermolecular distances within a dimer. A combination of distance geometry and restrained simulated annealing yielded a model of the dimer having antiparallel packing of two helical units, whose hydrophobic faces form a well-defined core. Sedimentation equilibrium experiments confirm the observation that neuropeptide Y associates to form dimers and higher aggregates under the conditions of the NMR experiments. Our results therefore support the structural features reported for porcine neuropeptide Y [Cowley, D.J. et al. (1992) Eur. J. Biochem., 205, 1099-1106] rather than the 'aPP' fold described previously for human neuropeptide Y [Darbon, H. et al. (1992) Eur. J. Biochem., 209, 765-771].

A derivative of the polypeptide cardiostimulant anthopleurin-B(AP-B) labeled with the spin label 1-oxyl 2,2,6,6-tetramethyl-4-piperidinyloxycarbonyl azide has been prepared and characterized. The product was found by mass spectrometry to be labeled at a single site, which amino acid sequencing showed to be the N-terminus. It also retained positive inotropic activity when assayed on isolated guinea pig atria. The spin-labeled (SL) product was found to exist in two distinct conformations by reversed-phase HPLC and in at least two conformations by electron spin resonance spectroscopy (ESR) over the pH range 2-9. The ESR data also show evidence for multimetric states of SL-AP-B over the pH range 2-9, with maximum aggregation at pH 4.5-5, and a slow disaggregation when the pH is adjusted to 8-9. The presence of multiple conformers of SL-AP-B and its tendency to aggregate render it unsuitable for high-resolution NMR structural studies of the isolated ligand, but the retention of activity may make it useful for studies of the sodium-channel-bound form of the molecule.

Charybdotoxin is a 37-residue polypeptide toxin from scorpion venom, which acts by blocking voltage-gated and Ca(2+)-activated K+ channels. We have synthesized charybdotoxin and three mono-substituted analogues using an Fmoc-tBu protocol. The Phe-2 --> Tyr analogues was chosen to introduce a site for Tyr iodination which was distinct from the K+ channel binding surface, while the Glu-12 --> Gln and Arg-19 --> His analogues were studied to probe the roles of charged residues at these positions in the structure and activity of the toxin. The synthetic native molecule was equipped with natural toxin in inhibiting the human erythrocyte Ca(2+)-dependent K+ channel. The affinities of all three analogues for the erythrocyte K+ channel were slightly reduced, with the Arg-19 --> His analogue showing the greatest increase in IC50 (2.30-fold). Two-dimensional 1H-NMR studies of these analogues showed that the Glu-12 to Gln substitution, which appeared to destabilise the N-terminal half of the alpha-helix, possibly due to the weakening of an N-terminal helix capping interaction which is apparent from our NMR data. His-21 has a pKa more than one unit below the value for a non-interacting histidine. Possible reasons for this are that the imidazolium side chain is partly buried and is located near positively charged moieties. Thus, His-21 would be neutral at physiological pH, where charybdotoxin binds to the potassium channel.

Background: The polypeptide anthopleurin-B (AP-B) is one of a number of related toxins produced by sea anemones. AP-B delays inactivation of the voltage-gated sodium channel of excitable tissue. In the mammalian heart, this effect is manifest as an increase in the force of contraction. As a result, there is interest in exploiting the anthopleurins as lead compounds in the design of novel cardiac stimulants. Essential to this endeavour is a high-resolution solution structure of the molecule describing the positions of functionally important side chains., Results: AP-B exists in multiple conformations in solution as a result of cis-trans isomerization about the Gly40-Pro41 peptide bond. The solution structure of the major conformer of AP-B has been determined by two-dimensional 1H NMR at pH 4.5 and 25 degrees C. The core structure is a four-stranded, antiparallel beta-sheet (residues 2-4, 20-23, 34-37 and 45-48) and includes several beta-turns (6-9, 25-28, 30-33). Three loops connect the beta-strands, the longest and least well defined being the first loop, extending from residues 8-17. These features are shared by other members of this family of sea anemone toxins. The locations of a number of side chains which are important for the cardiac stimulatory activity of AP-B are well defined in the structures., Conclusions: We have described the solution structure of AP-B and compared it with that of AP-A, from which it differs by substitutions at seven amino acid positions. It shares an essentially identical fold with AP-A yet is about 10-fold more active. Comparison of the structures, particularly in the region of residues essential for activity, gives a clearer indication of the location and extent of the cardioactive pharmacophore in these polypeptides.

The effects of solvent, pH and temperature on the 1H-NMR spectra of recombinant murine interleukin-6 (IL-6) are described. Assignments made from two-dimensional homonuclear spectra are presented for resonances of the fifteen aromatic amino-acid side chains. A time-dependent loss of intensity was observed for all resonances in the spectrum of IL-6, probably as a result of aggregation. This aggregation is markedly temperature-dependent. The pKa values of the four histidine residues in murine IL-6 has been measured; one has a value of 5.5, approx. one pH unit less than the value exhibited by the other three. Analysis of the NOESY spectra has allowed a preliminary characterisation of the nature of interactions among the aromatic side chains within the protein fold. 1H and 15N resonances of residues Thr-4 to Val-21 are assigned from three-dimensional 1H-15N correlated spectroscopy, and evidence is presented for these residues comprising a mobile N-terminal tail with little ordered structure. An N-terminal mutant lacking the first 22 residues of the murine IL-6 sequence and known to possess full biological activity was also examined and shown to have essentially retained the tertiary fold of the native molecule.

The three-dimensional structure of human insulin-like growth factor (IGF) II in aqueous solution at pH 3.1 and 300 K has been determined from nuclear magnetic resonance data and restrained molecular dynamics calculations. Structural constraints consisting of 502 NOE-derived distance constraints, 11 dihedral angle restraints, and three disulfide bridges were used as input for distance geometry calculations in DIANA and X-PLOR, followed by simulated annealing refinement and energy minimization in X-PLOR. The resulting family of 20 structures was well defined in the regions of residues 5 to 28 and 41 to 62, with an average pairwise root-mean-square deviation of 1.24 A for the backbone heavy-atoms (N, C2, C) and 1.90 A for all heavy atoms. The poorly defined regions consist of the N and C termini, part of the B-domain, and the C-domain loop. Resonances from these regions of the protein gave stronger cross peaks in two dimensional NMR spectra, consistent with significant motional averaging. The main secondary structure elements in IGF-II are alpha-helices encompassing residues 11 to 21, 42 to 49 and 53 to 59. A small anti-parallel beta-sheet is formed by residues 59 to 61 and 25 to 27, while residues 26 to 28 appear to participate in intermolecular beta-sheet formation. The structure of IGF-II in the well-defined regions is very similar to those of the corresponding regions of insulin and IGF-I. Significant differences between IGF-II and IGF-I occur near the start of the third helix, in a region known to modulate affinity for the type 2 IGF receptor, and at the C terminus. The IGF II structure is discussed in relation to its binding sites for the insulin and IGF receptors and the IGF binding proteins.

The VH domain of anti-influenza neuraminidase antibody NC41, with and without a C-terminal hydrophilic marker peptide (FLAG), has been expressed in high yield (15-27 mg/L) in Escherichia coli. Both forms were secreted into the periplasm where they formed insoluble aggregates which were solubilized quantitatively with 2 M guanidine hydrochloride and purified to homogeneity by ion-exchange chromatography. The VH-FLAG was composed of three isoforms (pI values of approximately 4.6, 4.9, and 5.3) and the VH molecule was composed of two isoforms with pI values of 5.1 and 6.7; the difference between the VH isoforms was shown to be due to cyclization of the N-terminal glutamine residue in the pI 5.1 isoform. At 20 degrees C and concentrations of 5-10 mg/ml the VH domain dimerized in solution and then partly precipitated, resulting in the broadening of resonances in its 1H NMR spectrum. Reagents such as CHAPS, n-ocytylglucoside, and ethylene glycol, which presumably mask the exposed hydrophobic interface of the VH molecule, prevented dimerization of the VH and permitted good-quality NMR spectra on isotope-labeled protein to be obtained.

The Plasmodium falciparum antigen SPAM (secreted polymorphic antigen associated with merozoites) contains an unusual set of heptad repeat units with alanine at the a and d positions. Twelve heptads with the consensus sequence AXXAXXX occur in three blocks of four, linked by short nonrepetitive sequences. A 38-residue polypeptide comprising the first block of four heptad units and five flanking residues at either end, SPAM-H1, has been synthesized and its structure in aqueous solution determined from 1H NMR data. Sedimentation equilibrium showed the peptide to be monomeric in aqueous solution. Its structure was determined from 1H NMR-derived distance and dihedral angle constraints by using distance geometry calculations, restrained simulated annealing, and conjugate gradient energy minimization in the CHARMm force field. The polypeptide contains an alpha-helix extending from Ser10 (position e of the first heptad) to at least Lys32 (position f of the fourth heptad) and possibly as far as Val35. The helix is bent, partly as result of a kink around residues 19-20. The conformations of the nine N-terminal residues and the six C-terminal residues are not well defined by the NMR data. The rms deviation from the average of the 20 best structures over the well-defined region (residues 11-31, which have backbone angular order parameters > 0.8) was 1.56 A for backbone heavy atoms (N, C alpha, and C) and 2.12 A for all heavy atoms. 2H2O exchange experiments identified slowly exchanging amide protons near the C-terminus and the last two turns of the helix. The unusual stability of the C-terminus reflects the presence of a new C-capping motif, which may involve the side chain of an asparagine in a position external to the C-cap residue. Possible interactions of the H1 sequence with the other two heptad repeat units in the intact merozoite antigen are discussed.

The three-dimensional structure in aqueous solution of the 49-residue polypeptide anthopleurin-A (AP-A), from the sea anemone Anthopleura xanthogrammica, has been determined from 1H NMR data. A restraint set consisting of 411 interproton distance restraints inferred from NOEs and 19 backbone and 13 side chain dihedral angle restraints from spin-spin coupling constants, as well as 15 lower bound restraints based on the absence of NOEs in the spectra, was used as input for distance geometry calculations in DIANA and simulated annealing and restrained energy minimization in X-PLOR. Stereospecific assignments for 12 beta-methylene pairs were also included. The final set of 20 structures had mean pairwise rms differences over the whole molecule of 2.04 A for the backbone heavy atoms (N, C alpha, and C) and 2.59 A for all heavy atoms. For the well-defined region encompassing residues 2-7 and 17-49, the corresponding values were 0.82 and 1.27 A, respectively. AP-A adopts a compact structure consisting of four short strands of antiparallel beta-sheet (residues 2-4, 20-23, 34-37, and 45-48) connected by three loops. The first loop commences with a type I beta-turn which includes two important Asp residues; this loop is the least well-defined region of the protein, although a beta-turn involving residues 13-16 is observed in nearly half the structures. The loop linking the second and third strands is constrained by the 29-47 disulfide bond and contains two well-defined beta-turns, while the third loop contains the Gly40-Pro41 sequence, which has been identified previously as the site of cis-trans isomerism. The carboxylate group of Asp7 is close to the epsilon-ammonium group of Lys37, suggesting that they may form a salt bridge. A pH titration monitored by 2D NMR supports this by showing that Asp7 has a low pKa. It is proposed that this region of the molecule and the nearby residues Asp9 and His39 form part of the molecular surface which interacts with the mammalian cardiac sodium channel.

Chemical modification studies have been carried out on the sea anemone polypeptide anthopleurin-A in order to clarify the role of Arg-14 in its cardiac stimulatory activity. Reaction with 1,2-cyclohexanedione at 37 degrees C produced a range of protein products, including some with amino group modifications. These side-reactions were eliminated by prior citraconylation of the amino groups, which, following reaction with cyclohexanedione, could be reversed under conditions which preserved the cyclohexanedione adduct. Citraconylation of the three amino groups, one from the N-terminus and two from Lys-37 and Lys-48, destroyed the cardiac stimulatory activity of the molecule, but this was fully recoverable upon reversal of this reaction. It appears that one or more of the amino groups is essential for activity. Anthopleurin-A contains only one arginine residue, and this was confirmed as the site of modification by cyclohexanedione by showing that the product was refractory to proteolysis by trypsin, which normally cleaves the molecule at this residue. The positive inotropic activity of the cyclohexanedione adduct on isolated guinea-pig atria was identical to that of unmodified anthopleurin-A, indicating that the side-chain of Arg-14 is not required for cardiotonic activity.

A common structural motif consisting of a cystine knot and a small triple-stranded beta-sheet has been defined from comparison of the 3-dimensional structures of the polypeptides omega-conotoxin GVIA (Conus geographus), kalata BI (Oldenlandia affinis DC), and CMTI-I (Curcurbita maxima). These 3 polypeptides have diverse biological activities and negligible amino acid sequence identity, but each contains 3 disulfide bonds that give rise to a cystine knot. This knot consists of a ring formed by the first 2 bonds (1-4 and 2-5) and the intervening polypeptide backbone, through which the third disulfide (3-6) passes. The other component of this motif is a triple-stranded, anti-parallel beta-sheet containing a minimum of 10 residues, XXC2, XC5X, XXC6X (where the numbers on the half-cysteine residues refer to their positions in the disulfide pattern). The presence in these polypeptides of both the cysteine knot and antiparallel beta-sheet suggests that both structural features are required for the stability of the motif. This structural motif is also present in other protease inhibitors and a spider toxin. It appears to be one of the smallest stable globular domains found in proteins and is commonly used in toxins and inhibitors that act by blocking the function of larger protein receptors such as ion channels or proteases.

The three-dimensional structure in aqueous solution of the 27-residue polypeptide neurotoxin Anemonia sulcata toxin III (ATX III) has been determined from 1H NMR data. As ATX III self-associates in the millimolar concentration range, causing a marked concentration dependence for the chemical shifts of several residues [Norton, R. S., Cross, K., Braach-Maksvytis, V., & Wachter, E. (1993) Biochem. J. 293, 545-551], it was necessary to record NOESY spectra over a range of concentrations in order to eliminate any intermolecular interactions from the NOE restraint set. The pairings of the six half-cystine residues were also unknown and had to be determined (as 3-17, 4-11, and 6-22) from preliminary structure calculations performed using both upper bound distance restraints from NOESY data and a substantial number of lower bound restraints inferred from the absence of NOESY cross-peaks. Final structures were determined, using the program X-PLOR, from interproton distance restraints inferred from NOEs, backbone and side chain dihedral angle restraints from spin-spin coupling measurements, and a smaller number of lower bound restraints. Stereospecific assignments for 11 beta-methylene pairs were also included. The final set of 28 structures had an average pairwise RMS difference of 1.32 A over the backbone heavy atoms (N, C alpha, and C) and 2.18 A over all heavy atoms. For the well-defined region encompassing residues 3-22, the corresponding values were 0.62 and 1.28 A, respectively. ATX III adopts a compact structure containing four reverse turns (a distorted type I beta-turn at residues 6-9, a type I beta-turn at residues 8-11, and inverse gamma-turns at residues 12-14 and 15-17) and two other chain reversals, but no regular alpha-helix or beta-sheet. Several of the residues most affected by aggregation are located on the surface of the molecule, forming a hydrophobic patch which may constitute part of the sodium channel binding surface. Possible relationships between the structure of ATX III and those of other sea anemone toxins that interact with the same site on the voltage-gated sodium channel are considered.

Human leukaemia inhibitory factor (LIF) is a glycoprotein with a diverse range of activities on many cell types. A molecular model of LIF has been constructed based mainly on the structure of the related cytokine granulocyte colony-stimulating factor, and refined using simulated annealing and molecular dynamics in water. The model was stable during molecular dynamics refinement and is consistent with known stereochemical data on proteins. It has been assessed by comparison with 1H NMR data on the ionization behaviour of the six histidine residues in LIF, the imidazolium pKa values of which range from 3.6 to 7.4. These pKa values were assigned to individual histidine residues from NMR studies on a series of His-->Ala mutants. The environments of the histidine residues in the model account very well for their observed ionization behaviour. Furthermore, the model is consistent with mutagenesis studies which have defined a group of amino acid residues involved in receptor binding.

The structure-function relationships of the neurotoxic polypeptide Sh I, from the sea anemone Stichodactyla helianthus, have been studied using limited proteolysis with trypsin and endoproteinase Lys-C. Major products from each of the proteolytic digests were characterised using N-terminal peptide sequencing and amino-acid analysis or mass spectrometry. Of the six possible tryptic cleavage sites in Sh I, the bonds adjacent to Arg-13 and Lys-47 were found to be the most susceptible, complete cleavage occurring within minutes. Cleavages adjacent to Lys-32 and Lys-46 proceeded more slowly and cleavage adjacent to Arg-45 was the slowest. The sixth potential site, adjacent to Lys-4, was not cleaved at all. All derivatives were inactive as crustacean neurotoxins. Cleavage with endoproteinase Lys-C generated two major products. Derivatives cleaved adjacent to Lys-32 and either Lys-46 or Lys-47 were isolated. Both were inactive, indicating that either cleavage adjacent to Lys-32 or the removal of the C-terminal lysine residue(s) was sufficient to abolish activity. Lys-4 again was refractory to cleavage. The sequence of cleavage events correlated well with the static accessibility of the lysyl and arginyl side chains and to a lesser extent with the accessibility of the carbonyl oxygen of susceptible peptide bonds, as measured from the solution structure of Sh I determined by 1H-NMR. In the case of Lys-4, the lack of cleavage by trypsin and endoproteinase Lys-C may reflect a lack of flexibility in this region. The effects of the various cleavages on biological activity emphasise that the surface of the protein near the reverse turn encompassing Asp-6, Asp-7 and Glu-8 is essential for activity.

Anthopleurin-A (AP-A) is a member of a family of sea anemone-derived polypeptides that interact with sodium channels in a voltage-dependent manner, producing a positive inotropic effect on the mammalian heart. There has been considerable interest in this molecule as a lead compound for the development of novel therapeutic agents. Earlier attempts to define the 3-dimensional structure of AP-A were complicated by the fact that it was found to exist in 2 conformations in solution. Using 1H- and 13C-NMR spectroscopy, we have now shown that this conformational heterogeneity arises from cis-trans isomerization about the Gly 40-Pro 41 peptide bond and that in the major form of the protein this peptide bond adopts a cis conformation. Furthermore, the increased sensitivity afforded by higher-field NMR has allowed identification of additional minor conformations of AP-A, the origin of which is presently unknown. We believe there will be many more examples of the detection by high-field NMR of previously unobserved minor conformations of proteins in solution.

The sea anemone polypeptide anthopleurin-A (AP-A) at nanomolar concentrations enhances myocardial contractility without affecting automaticity. It has a therapeutic index higher than that of the digitalis glycosides, and may serve as a molecular model for designing a new class of inotropic drugs acting on the myocardial Na channel at site 3. AP-A is a 49 residue peptide crosslinked by three disulfide bonds; its tertiary structure has been determined by NMR. Here we report the solid-phase synthesis of this polypeptide. Synthetic AP-A displayed CD and NMR spectra identical with those of the natural toxin; it possessed 94 +/- 15% of the inotropic activity of natural AP-A. Therefore, it is feasible to prepare various type 1 sea anemone toxin analogs by solid-phase chemical synthesis in order to identify side chains important for peptide folding and interaction with sodium channels.

Two synthetic peptides corresponding to the C-terminal 19 residues of human and murine interleukin-6, respectively, have been synthesized and their structures in solution investigated using high-resolution 1H-NMR spectroscopy. Both peptides show a marked dependence of chemical-shift dispersion on pH, with a greater degree of structure apparent above pH 4.5, where their glutamate carboxyl groups are ionised. In purely aqueous solution, neither peptide adopts a well-defined structure, although the murine peptide has characteristics of a nascent helix. Titration of the murine peptide with trifluoroethanol produced a significant increase in structure, which was then investigated using two-dimensional NMR. In 50% (by vol.) trifluoroethanol the murine peptide consists of a well-defined central helix of 12 residues with unstructured N-terminal and C-terminal regions. These observations lend experimental support to the current model of the interleukin-6 structure, which proposes a four-helical bundle with the last helix encompassing the C-terminal 20-30 residues. Furthermore, the fact that synthetic peptides corresponding to part of the putative receptor-binding surface of interleukin-6 are able to adopt a similar conformation in solution to that proposed for the intact protein suggests that such peptide analogues should be useful starting points in the design of peptide agonists and antagonists of interleukin-6.

Leukaemia inhibitory factor (LIF) is a polyfunctional cytokine active on many cell types. We present here 1H NMR studies on the solution properties and stability of MH35, a chimera of murine and human LIF which can be expressed at high levels in Escherichia coli, thus enabling efficient labelling of the protein with the stable isotopes 13C and 15N. MH35 has 85% sequence identity with human LIF and similar activity in biological assays. The 1H chemical shifts of the 12 conserved aromatic residues and the pKa values of the five conserved histidine residues in MH35 and human LIF are very similar. Temperature dependence studies indicate that both proteins are stable, with significant conformational changes occurring only above 70 degrees C. These results show that these proteins have a similar overall structure and stability and that MH35 is therefore a suitable analogue of human LIF for structural studies in solution.

A high resolution structure in solution has been determined for the polypeptide neurotoxin I, from the sea anemone Stichodactyla helianthus, using NMR data, distance geometry calculations, and refinement by back-calculation of two-dimensional nuclear Overhauser enhancement (NOE) spectra. A set of 913 distance constraints derived from NOEs was used, together with a large set of lower distance bounds based on the absence of NOEs in the spectrum. Eight published structures for neurotoxin I were refined independently to give structures which agree better with the experimental data, as reflected in reduced R factors calculated over well resolved cross-peaks of the two-dimensional NOE spectra and a lower total volume of peaks in back-calculated spectra that are absent from experimental spectra. The refined structures are also more precisely defined, with mean pairwise root mean square differences over backbone heavy atoms of 0.62 A for well defined residues and 1.14 A for all residues, compared with previous values of 1.09 and 2.41 A, respectively. The consensus constraint set from the 8 refined structures was also used to generate 12 new structures, with corresponding root mean square differences of 0.76 and 1.26 A. In all 20 structures the loop linking the first and second strands of the beta-sheet is considerably better defined than before. A type I beta-turn encompassing the functionally important residues Asp6, Asp7, and Glu8 is evident in the refined structures.

The 27 amino acid residue polypeptide omega-conotoxin GVIA, from venom of the cone shell Conus geographus, blocks neuronal voltage-activated calcium channels at picomolar concentrations. The three-dimensional structure in aqueous solution of synthetic omega-conotoxin has been determined from two-dimensional 1H n.m.r. data recorded at 600 MHz. Structural constraints consisting of interproton distances inferred from NOEs and dihedral angles from spin-spin coupling constants were used as input for distance geometry calculations with the program DSPACE. The structures were then refined using back-calculation of NOESY spectra. The family of structures obtained in this way is well defined by the n.m.r. data, the best 12 structures having pairwise root-mean-square differences of 0.68 (+/- 0.15) A over the backbone heavy atoms (N, C alpha and C) and 1.15 (+/- 0.17) A over all heavy-atoms. The molecule adopts a compact structure consisting of a small, triple-stranded, anti-parallel beta-sheet and several reverse turns. All three tyrosine residues are located on the molecular surface, which is noteworthy for its abundance of side-chain hydroxyl groups. There is no negatively charged group in conotoxin, but the five positively charged groups are distributed in three small patches on the surface, one of which, made up of the ammonium moieties of the N terminus and Lys2, may contribute to the receptor-binding surface of the molecule. An isomer of conotoxin with the same amino acid sequence, but different disulfide pairings, has also been investigated. Its structure is less well ordered than that of native conotoxin and it shows significant heterogeneity, probably as a result of cis-trans isomerism preceding hydroxyproline residues.

Two murine interleukin-6 (mIL-6) variants were constructed using the polymerase chain reaction (PCR), one lacking the last five residues (183-187) at the C-terminus (pMC5) and another with the last five residues of mIL-6 substituted by the corresponding residues of human IL-6 (pMC5H). The growth stimulatory activity of pMC5 on the mouse hybridoma cell line 7TD1 was < 0.05% of mIL-6, whereas pMC5H and mIL-6 were equipotent. The loss of biological activity of pMC5 correlated with its negligible receptor binding affinity on 7TD1 cells, while the binding of pMC5H was comparable to that of mIL-6. Both pMC5 and pMC5H, like mIL-6, failed to interact with recombinant soluble human IL-6 receptor when assayed by surface plasmon resonance-based biosensor analysis. These studies suggest that the C-terminal seven amino acids of human IL-6, alone, do not define species specificity for receptor binding. A variety of biophysical techniques, as well as the binding of a conformational-specific monoclonal antibody, indicated that the global fold of the mIL-6 variants was similar to that of mIL-6, although small changes in the NMR spectra, particularly for pMC5, were observed. Some of these changes involved residues widely separated in the primary structure. For instance, interactions involving Tyr-22 were influenced by the C-terminal amino acids suggesting that the N- and C-termini of mIL-6 are in close proximity. Equilibrium unfolding experiments indicated that pMC5 was 0.8 kcal/mol less stable than mIL-6, whereas pMC5H was 1.4 kcal/mol more stable. These studies emphasize the structural importance of the C-terminal amino acids of IL-6 and suggest that truncation or mutation of this region could lead to small but significant alterations in other regions of the molecule.

The solution properties, secondary structure and global fold of the 27-residue polypeptide neurotoxin III (ATX III), from the sea anemone Anemonia sulcata, have been investigated using high-resolution 1H-n.m.r. spectroscopy. Studies of the concentration dependence of the n.m.r. spectrum indicate that the molecule self-associates in the millimolar concentration range useable for n.m.r. analysis, the association being less pronounced at acidic pH values. The dependence on pH of association implies that electrostatic interactions play a role in this process, while the significant concentration-dependent shifts of the aromatic resonances of Tyr-7 and Trp-13 indicate that hydrophobic interactions also contribute. Individual pKa values have been determined for most ionizable groups in the molecule. Sequence-specific resonance assignments were obtained for all protons using a range of two-dimensional homonuclear-correlated and nuclear-Overhauser-effect (nOe) spectra. The secondary structure of the polypeptide was identified from sequential (i, i+1) and medium-range (i, i+2/3/4) nOe connectivities, NH to C alpha H coupling constants, C alpha H chemical shifts, and the location of slowly exchanging backbone-amide protons. ATX III contains no regular alpha-helix or beta-sheet, consisting instead of a network of reverse turns. nOe connectivities between half-cystine residues are consistent with the disulphide pairings 3-17, 4-11 and 6-22. ATX III has a well-defined structure and appears to lack the disordered loop which, in the longer sea anemone toxins (46-49 residues), may be part of the receptor-binding surface.

The complete sequence-specific assignment of resonances in the 1H-NMR spectrum of the polypeptide neurotoxin III (Hm III) from the sea anemone Heteractis macrodactylus is described. Comparison of the chemical shifts and pattern of NOEs for Hm III with those for the related toxin Hp III from Heteractis paumotensis, which differs only in the substitution of Asn for Tyr at position 11, shows that the overall secondary and tertiary structures are conserved. The largest differences in chemical shift caused by the substitution at position 11 are observed for the NH resonances of Arg-13, Thr-14, Ala-15, Leu-17, and Cys-26. The C alpha H resonances influenced most are those of ASP-6, Gly-9, Leu-17, and Glu-42, while the most affected C beta H resonances are from Leu-17, Glu-28, and Lys-32. The absence of long-range NOEs to the aromatic ring of Tyr-11 as well as the lack of significant chemical shift effects on residues outside the loop comprising residues 7-16 confirm that this part of the loop makes no long-lived contacts with the rest of the molecule. The deviations from random coil shifts of Hm III are compared with those of the related anemone toxins Hp II, Hp III, and toxin I from Stichodactyla helianthus (Sh I). The similarity in deviations in chemical shift as a function of sequence position for these four toxins emphasizes the overall structural homology among these polypeptides.

Ciguatoxin-2, a major ciguatoxin present in the flesh and viscera of ciguateric fishes, has been shown by 1H nuclear magnetic resonance studies (2-dimensional homonuclear Hartman Hahn, nuclear Overhauser effect and decoupling difference experiments) to be a diastereomer of ciguatoxin-3, differing only in stereochemistry at carbon 52 (a quaternary carbon). This difference accounts for the significant changes in the chemical shift of resonances for protons in this region of ciguatoxin-2. Differences between ciguatoxin-1, -2 and -3 involve modifications at only one end of the ciguatoxins (ring M) and modest differences in potency, indicating that this ring contributes to, but is not critical for, high affinity binding of the ciguatoxins to voltage-dependent sodium channels. It is proposed that ciguatoxin-2 originates from a different precursor to the precursor (presumably gambiertoxin-4b) for ciguatoxin-1 and -3, and that both precursors are produced by a common biosynthetic pathway in Gambierdiscus toxicus.

Sequence-specific resonance assignments are reported for the 500-MHz 1H-NMR spectrum of the 55-residue neurotoxin B-IV, isolated from the heteronemertine worm Cerebratulus lacteus. A range of two-dimensional homonuclear correlated and NOE spectra was used in making these assignments, which include NH, C alpha H and C beta H resonances, as well as most resonances from longer-chain spin systems, with the exception of the ten Lys residues, where spectral overlap prevented complete, unambiguous assignments. The secondary structure of B-IV was identified from the pattern of sequential (i, i + 1) and medium range (i, i + 2/3/4) NOE connectivities and the location of slowly exchanging backbone amide protons. Two helices are present, incorporating residues 13-26 and 33-49, and the C-terminal five residues form a helix-like structure. A type-I reverse turn, involving residues 28-31 is present in a small loop linking the two major helices, and the N-terminus appears to be unordered at 27 degrees C, although it may adopt a more ordered conformation at lower temperatures. These elements of secondary structure, together with the four disulfide bonds in the protein, provide sufficient information to define the global fold of the molecule in solution. The pH and temperature dependence of the toxin have been investigated by 1H-NMR and the pKa values of several ionisable sidechains determined.

A loop corresponding to residues 8-17 in the polypeptide cardiac stimulant anthopleurin-A is known to be important for the cardiostimulant activity of this molecule. To investigate the activity and possible conformations of this loop in isolation, two synthetic peptides have been studied. The first corresponds to residues 6-20 of anthopleurin-A with Cys6 replaced by Thr, and the second to residues 6-21 of anthopleurin-A, with Thr21 replaced by Cys. The introduction of an additional cysteine in the latter peptide enabled an intramolecular disulfide to be formed between the N- and C-terminal residues. Both linear peptides and the disulfide-containing analogue lack the cardiostimulant and Na(+-)-channel binding activity in the parent molecule, anthopleurin-A, indicating that although the loop is important for the function of anthopleurin-A, other regions of the molecule must also be involved in activity. Assignments of the 1H-NMR spectra of both peptides are presented, and their pH and temperature dependences investigated. The results show that the amide protons of Gly5 and Asn11 (corresponding to Gly10 and Asn16 in anthopleurin-A) sample hydrogen-bonded conformations in solution. Based on these NMR data, two regions of non-random structure, encompassing residues 2-5 and 8-11, respectively, are proposed, and the possible involvement of such structures in the activity of anthopleurin-A is discussed.

The chemical, physical and biological properties of the cytolysin tenebrosin-C from Actinia tenebrosa have been compared with those of equinatoxin II from Actinia equina. The two proteins are indistinguishable by reverse-phase and cation-exchange HPLC and capillary zone electrophoresis, and give similar peptide fragments upon cyanogen bromide cleavage (as judged by the chromatographic behaviour, ultraviolet absorption spectra, amino acid composition and N-terminal amino acid sequences of the peptides). Their cardiac stimulatory activities are identical, and their haemolytic activities are similar, with equinatoxin II having slightly greater activity. These data indicate that the two molecules are either identical in all 179 amino acid positions, or differ by no more than one or two residues. These findings are discussed in the context of the taxonomic relationship between the two species of sea anemone.

BDS II, a 43-residue polypeptide from the sea anemone Anemonia sulcata, is reported to have both antihypertensive and antiviral activity. This polypeptide possesses a number of sequence and structural similarities to a class of cardiotonic proteins which bind to receptor site 3 of the voltage-gated sodium channel. In contrast to these cardiostimulant proteins, which produce positive inotropic effects at concentrations of 2-15 nM, BDS II produced a weak negative inotropic effect upon isolated guinea-pig atria, with doses of 90 and 180 nM depressing contractile strength by 15 and 28%, respectively. BDS II also competed with a 125-iodine labelled derivative of AP-A (a representative of the cardiostimulant proteins) bound to sodium channels in rat brain synaptosomes. The IC50 for BDS II versus AP-A was 5.2 microM. BDS II may therefore be considered an antagonist for receptor site 3 of the voltage-gated sodium channel. Structural differences between BDS II and the agonist AP-A which may give rise to their different effects on the sodium channel are considered.

When exponentially growing Saccharomyces cerevisiae was transferred from a normal high water activity growth medium (aw 0.997) to a medium containing 8% NaCl low water activity growth medium (aw 0.955), glycerol accumulation during the first eight hours of the adaptation was both retarded and greatly diminished in magnitude. Investigation of the underlying reasons for the slow onset of glycerol accumulation revealed that not only was overall glycerol production reduced by salt transfer, but also the rates of ethanol production and glucose consumption were reduced. Measurement of glycolytic intermediates revealed an accumulation of glucose-6-phosphate, fructose-6-phosphate, fructose 1,6 bisphosphate and phosphoenolpyruvate in S. cerevisiae 3 to 4 h after transfer to salt, suggesting that one or more glycolytic enzymes were inhibited. Potassium ions accumulated in S. cerevisiae after salt transfer and reached a maximum about 6 h after transfer, whereas the sodium ion content increased progressively during the adaptation period. The trehalose content also increased in adapting cells. It is suggested that inhibition of glycerol production during the initial period of adaptation could be due to either the inhibition of glycerol-3-phosphate dehydrogenase by increased cation content or the inhibition of glycolysis, glycerol being produced glycolytically in S. cerevisiae. The increased accumulation of glycerol towards the end of the 8-h period suggests that the osmoregulatory response of S. cerevisiae involves complex sets of adjustments in which inhibition of glycerol-3-phosphate dehydrogenase must be relieved before glycerol functions as a major osmoregulator.

Sea anemones produce a series of toxic polypeptides and proteins with molecular weights in the range 3000-5000 that act by binding to specific receptor sites on the voltage-gated sodium channel of excitable tissue. This article reviews our current knowledge of the molecular basis for activity of these molecules, with particular emphasis on recent results on their receptor binding properties, the role of individual residues in activity and receptor binding, and their three-dimensional structures as determined by nuclear magnetic resonance spectroscopy. A region of these molecules that constitutes at least part of the receptor binding domain is proposed.

Viscera (48.3 kg) from moray eels (Lycodontis javanicus) collected in a ciguatera endemic area were extracted and the ciguatoxins characterized. Three major ciguatoxins, CTX-1, CTX-2 and CTX-3, were isolated and purified to homogeneity on reverse phase high performance liquid chromatography. Several minor toxins were also detected. CTX-1 (490 micrograms) was comparable by both 1H nuclear magnetic resonance (1H NMR) and mass spectroscopy (MH+ m/z = 1111) to ciguatoxin isolated previously from moray eels. CTX-2 (280 micrograms) and CTX-3 (100 micrograms) were less polar ciguatoxins not previously characterized. CTX-2 and CTX-3 differed from CTX-1 by 16 mass units, suggesting that they were less oxygenated analogues. 1H NMR revealed that the hydroxyl at C54 in CTX-1 was absent in CTX-2 and CTX-3. An additional change in the chemistry of CTX-2 compared to CTX-1 and CTX-3 was also suggested on the basis of 1H NMR, indicating that CTX-2 may arise from a different precursor to CTX-1. CTX-3 is likely to be an intermediate in the oxidation of a gambiertoxin (sodium channel toxins from Gambierdiscus toxicus) to CTX-1. The i.p. LD50 values for CTX-1, CTX-2 and CTX-3 were 0.25, 2.3 and 0.9 micrograms/kg, respectively. The signs induced in mice by the ciguatoxins were similar, except that CTX-2 and CTX-3 induced hind-limb paralysis that was absent with CTX-1. Each ciguatoxin was potent orally. CTX-1, CTX-2 and CTX-3 competitively inhibited the binding of [3H]brevetoxin-3 to voltage-dependent sodium channels with relative potencies qualitatively (but not quantitatively) comparable to mouse lethality. This study reveals that the relatively small chemical differences between CTX-1, CTX-2 and CTX-3 give rise to significant structure-activity and pharmacokinetic differences.

The sea anemone Stichodactyla helianthus neurotoxin I, a 48-residue polypeptide, was synthesized by automated solid phase methodology. The fully reduced polypeptide was subsequently refolded in the presence of a glutathione oxidoreduction buffer to the biologically active species containing three disulfide bonds. The overall yield after rigorous purification was 12.5%. The circular dichroism (CD), and proton nuclear magnetic resonance (1H NMR) spectra of the HPLC-purified synthetic toxin were indistinguishable from those obtained concurrently with the natural toxin. A subtilisin digest of the synthetic neurotoxin generated peptide fragments identical to that of a sample of the natural toxin subjected to the same treatment. The toxicity of the synthetic polypeptide was identical to that of the natural toxin (crab LD50, 3.1 micrograms/kg). The equilibrium dissociation constant (28 nM) for interaction of the synthetic toxin with crab axolemma vesicles was nearly identical to that of the natural toxin (25 nM).

The three-dimensional structure of the sea anemone polypeptide Stichodactyla helianthus neurotoxin I in aqueous solution has been determined using distance geometry and restrained molecular dynamics simulations based on NMR data acquired at 500 MHz. A set of 470 nuclear Overhauser enhancement values was measured, of which 216 were used as distance restraints in the structure determination along with 15 dihedral angles derived from coupling constants. After restrained molecular dynamics refinement, the eight structures that best fit the input data form a closely related family. They describe a structure that consists of a core of twisted, four-stranded, antiparallel beta-sheet encompassing residues 1-3, 19-24, 29-34, and 40-47, joined by three loops, two of which are well defined by the NMR data. The third loop, encompassing residues 7-16, is poorly defined by the data and is assumed to undergo conformational averaging in solution. Pairwise root mean square displacement values for the backbone heavy atoms of the eight best structures are 1.3 +/- 0.2A when the poorly defined loop is excluded and 3.6 +/- 1.0A for all backbone atoms. Refinement using restrained molecular dynamics improved the quality of the structures generated by distance geometry calculations with respect to the number of nuclear Overhauser enhancements violated, the size of the total distance violations and the total potential energies of the structures. The family of structures for S. heliathus neurotoxin I is compared with structures of related sea anemone proteins that also bind to the voltage-gated sodium channel.

31P NMR has been used to observe phosphorus-containing compounds in both perchloric acid and KOH extracts and in whole cell suspensions of Crithidia luciliae. Cells were grown in Bone and Steinert medium, or in a modified RPMI culture medium and harvested after about 72 h in mid- to late log phase. 31P NMR spectra of the perchloric acid extracts indicated that 3-phosphoglycerate, which is normally at low concentrations in most cells, was the dominant phosphorus-containing compound in the sugar phosphate region. 3-Phosphoglycerate is the end product of glycosomal glycolysis and our finding is consistent with previous observations of the failure to detect prior glycolytic intermediates. Other metabolites observed were ATP, ADP, NAD(P)+, phosphoenolpyruvate and low molecular weight polyphosphates (PPn, n less than 20). The presence of high-molecular-weight polyphosphates was established by spectra recorded on extracts obtained through subsequent treatment of the insoluble fraction with KOH. 31P NMR experiments on whole cells indicated that the average main internal pH of cells in late-log growth phase was approx. pH 7.2 +/- 0.1, using the orthophosphate resonance as an indicator. The cells responded to the addition of glucose (final concentration approx. 35 mM) with a decrease in pH, both internal (delta pH = -0.9 (55 min)-1) and external (delta pH = -1.3 (15 min)-1). Polyphosphates and ATP could not be observed in whole cell experiments, although perchloric acid extracts of identically treated cells showed no significant depletion of these compounds.

The complete amino acid sequence of the cardiac stimulatory and haemolytic protein tenebrosin-C, from the Australian sea anemone Actinia tenebrosa, has been determined by Edman degradation of the intact molecule and fragments produced by treatment of the polypeptide chain with cyanogen bromide and enzymatic cleavage with endoproteinase Asp-N, thermolysin and trypsin. The molecule is a single-chain polypeptide consisting of 179 amino acid residues with a calculated molecular mass of 19,797 Da. Tenebrosin-C shows a high degree of amino acid sequence similarity (63%) with Stoichactis helianthus cytolysin III [Blumenthal, K. M. and Kem, W. R. (1983) J. Biol. Chem. 258, 5574-5581] and is identical to a partial sequence (90 residues) reported for equinatoxin, a cardiostimulatory and haemolytic protein isolated from the European sea anemone Actinia equina [Ferlan, I. and Jackson, K. (1983) Toxicon Suppl. 3, 141-144]. No amino acid sequence similarity was detected between tenebrosin-C and other protein sequences stored in available databases. The predicted secondary structure of tenebrosin-C suggests that it is a compact, highly structured molecule.

Selective proteolysis of the polypeptide cardiostimulant anthopleurin-A by trypsin introduces a single break in the polypeptide backbone on the C-terminal side of Arg14. The resulting derivative is devoid of any cardiostimulant activity. The structural changes which accompany this loss of activity have been examined by one- and two-dimensional 1H-NMR spectroscopy. It is shown that the overall backbone folding of anthopleurin-A is conserved on digestion, with some structural changes occurring for residues which are adjacent to the site of cleavage by trypsin. Thus, although previous NMR studies on anthopleurin-A indicate that the region surrounding Arg14 is devoid of any ordered structure, it appears that some degree of structural integrity is required to allow the essential side chains to adopt the conformation necessary to produce a cardiostimulant effect.

High-resolution nuclear magnetic resonance (NMR) spectroscopy has developed into a powerful method for determining the complete three-dimensional structures of proteins in solution. This article reviews current applications of this technique, with particular emphasis on areas of relevance to biotechnology, including protein engineering and protein design.

The sequence-specific assignment of resonances in the 500-MHz 1H-NMR spectrum of a cardioactive sea anemone polypeptide, anthopleurin-A, is described. The assignment procedure involved analysis of two-dimensional phase-sensitive multiple-quantum-filtered, double-quantum, homonuclear Hartmann-Hahn and nuclear Overhauser effect spectra. Using sequential information, specific assignments have been made for resonances arising from all 49 amino acid residues. Resonances arising from a number of residues in a minor conformer present in solution are also assigned. These results greatly extend previous resonance assignments made from spectra acquired at 300 MHz [Gooley, P. R. and Norton, R. S. (1985) Eur. J. Biochem. 153, 529-539] and provide the basis for a more accurate definition of the conformation of anthopleurin-A in aqueous solution. The secondary structure includes a four-stranded antiparallel beta-sheet encompassing residues 2-4, 21-23, 34-36 and 45-49, and possibly a beta-bulge located at Ser-19 and Gly-20. A type II beta-turn is formed by residues 30-33. These structural elements also occur within other related sea anemone polypeptides, but the conformation of the small loop region containing Pro-41 appears to be unique to anthopleurin-A.

Three new proteins with cardiac stimulatory and haemolytic activity, designated tenebrosins-A, -B and -C, have been purified from the Australian sea anemone Actinia tenebrosa. These proteins are basic (pI greater than or equal to 9.4), have mol. wt of about 20,000, and have very similar amino acid compositions and N-terminal amino acid sequences. None of the proteins contains cysteine or cystine residues. On isolated, spontaneously beating guinea pig atria they exhibit at 1-2 nM strong positive inotropic and slight to moderate chronotropic effects. In some cases a transient negative inotropic effect occurs prior to onset of the positive inotropic response. The proteins are also haemolytic, producing 50% haemolysis of guinea pig erythrocytes at concentrations similar to those showing positive inotropic effects.

Tenebrosin-C is a protein of mol. wt 19,500 that displays potent cardiac stimulatory and haemolytic activities. Its haemolytic activity is inhibited by sphingomyelin but not phosphatidylcholine, and is not affected by Ca2+. The positive inotropic effect of tenebrosin-C on isolated guinea pig right atria is inhibited by the cyclooxygenase blockers indomethacin and aspirin, the lipoxygenase blocker and leukotriene antagonist RG5901, and the phospholipase A2 inhibitor mepacrine. This activity of tenebrosin-C therefore appears to be due to stimulation of the release of arachidonic acid and subsequent formation of prostaglandins and leukotrienes. Phospholipase A2-like activity was found with some tenebrosin-C preparations, but did not correlate with their positive inotropic or haemolytic activities and was too weak to account for either of these effects. Treatment of tenebrosin-C with various proteases in order to obtain active fragments showed that the protein is remarkably resistant to proteolysis.

Natural-abundance 13C NMR spectra (at 15.04 MHz) of the polypeptide toxin II from the sea anemone Anemonia sulcata have been analysed and compared with corresponding spectra reported recently for a closely related polypeptide anthopleurin A. The spectra contain many resolved one-carbon and two-carbon resonances from carbonyl, aromatic and methyl carbons, many of which have been assigned to individual carbons in the molecule on the basis of their chemical shifts, including their pH dependence, and by comparison with the 13C NMR spectrum of anthopleurin A. Analysis of the effects of pH on the spectrum yields estimates for the pKa values of a number of functional groups in the molecule, as follows: side-chain carboxylates of the two aspartic acid residues 2 and 3.1; COOH-terminal carboxylic acid, 3.5; imidazolium moieties of the two histidine residues, 6.7 and 7.6 NH2-terminal ammonium, 8. The similarity between the pKa values of these functional groups in toxin II and those of corresponding groups in anthopleurin A, together with the close agreement between chemical shifts of conserved carbons, indicates that many local interactions are nearly identical in the two molecules, and thus supports the thesis that their overall conformations in solution are similar. However, the local interactions involving one of the aspartic acid residues are altered in toxin II. Together with other data, this leads to a proposal for the site in these two molecules which is responsible for their cardiac stimulatory activity.