Portal de Periódicos da CAPES

Structure-Function Relationships of ω-Conotoxin GVIA

1997; Elsevier BV; Volume: 272; Issue: 18 Linguagem: Inglês

10.1074/jbc.272.18.12014

1083-351X

Michael J. Lew, James P. Flinn, Paul K. Pallaghy, Roger Murphy, Sarah L. Whorlow, Christine E. Wright, Raymond S. Norton, James A. Angus,

Nicotinic Acetylcholine Receptors Study

The structure-function relationships of the N-type calcium channel blocker, ω-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 Lys2and 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 structure-function relationships of the N-type calcium channel blocker, ω-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 Lys2and 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 ( 95% pure) were pooled, diluted and loaded onto a 250 mm × 10-mm semi-preparative column for desalting. The peptide was eluted with a linear gradient of 0.1% trifluoroacetic acid H2O/MeCN. Fractions containing the purified peptide were pooled and lyophilized. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry was used for confirmation of product identity and was performed on a Finnigan Lasermat. The absorbing matrix used was α-cyano-4-hydroxy-cinnamic acid. Capillary zone electrophoresis was used for confirmation of product homogeneity and was conducted on an Applied Biosystems model 270A instrument, using the following conditions: applied potential, 30 kV; buffer, sodium citrate (20 mm, pH 2.5); capillary length, 72 cm; capillary diameter, 50 μm; temperature, 30 °C. All polypeptide analogues were analyzed by1H NMR spectroscopy on a Bruker AMX-600 spectrometer. Two-dimensional spectra were acquired on 2–10-mg samples of polypeptide in 90% H2O, 10% 2H2O at pH 3.4 and 298 K (400 t 1 increments, 4096 data points, and 64–96 scans/t 1 increment were employed). The TOCSY spin-lock and NOESY mixing times were 70 and 400 ms, respectively. The methodology was otherwise as described by Pallaghy et al. (28Pallaghy P.K. Scanlon M.J. Monks S.A. Norton R.S. Biochemistry. 1995; 34: 3782-3794Google Scholar). NOESY spectra of analogues were analyzed in terms of which non-trivial (medium and long range) NOEs of the native molecule were lost in the analogues and which, if any, new NOEs were observed that were not seen in the native spectrum. The intensities of NOEs that were present in both spectra were not monitored. A custom-written computer program, NMRanalogue, was written to produce macros (suitable for input to the spectral analysis programs Felix 2.3 and Felix 95) that allowed this procedure to be partially automated for the series of NOESY spectra. The lost NOEs were categorized as very weak, weak, medium, strong, or very strong according to their intensities in NOESY spectra of the native molecule. Only peaks of intensity "weak" or greater for analogues of estimated sample quantity ≥7 mg were considered significant. The native sample quantity was approximately 10 mg, and only well shaped peaks were used in the native NOE list. 2P. K. Pallaghy and R. S. Norton, manuscript in preparation. The designation of a native NOE as "lost" for any given analogue was performed conservatively, so that even a very distorted peak near the required position was considered to be maintained. Insight II (version 95.0) was used for molecular graphics. Crude rat brain membranes were prepared by the method of Cruz and Olivera (29Cruz L.J. Olivera B.M. J. Biol. Chem. 1986; 261: 6230-6233Google Scholar). Individual brains were homogenized using an Ultra-Turrax homogenizer (Janke and Kunkel) in 10 volumes of buffer (0.32 m sucrose, 5 mm HEPES-Tris, pH 7.4, 0.1 mm phenylmethylsulfonyl fluoride) and the homogenate spun at 1000 × g for 10 min at 4 °C. The pellet was resuspended in another 10 volumes of buffer and spun again. The combined supernatants were then spun at 17,000 × gfor 60 min at 4 °C. The pellet was taken up in 100 volumes of buffer as above but also containing 50 mg/liter lysozyme. This gives membrane derived from about 200 μg of tissue/20 μl of homogenate. The suspension was divided into 3-ml aliquots and stored at −70 °C. Binding assays were conducted in 96-well microtiter plates with 0.65-μm filters in the bottom (Millipore multiscreen system), using a modification of the method of Cruz and Olivera (29Cruz L.J. Olivera B.M. J. Biol. Chem. 1986; 261: 6230-6233Google Scholar) and Haack et al. (30Haack J.A. Kinser P. Yoshikami D. Olivera B.M. Neuropharmacology. 1993; 32: 1151-1159Google Scholar), as also used by Kim et al.(24Kim J.I. Takahashi M. Ogura A. Kohno T. Kudo Y. Sato K. J. Biol. Chem. 1994; 269: 23876-23878Google Scholar). 3K. Sato, personal communication. The binding buffer contained 0.32m sucrose, 5 mm HEPES-NaOH (pH 7.4), 0.1 mm phenylmethylsulfonyl fluoride, 0.3% bovine serum albumin, and 50 mg/liter lysozyme. Wash buffer consisted of 150 mm NaCl, 5 mm HEPES-NaOH, pH 7.4, 1.5 mm CaCl2, and 0.1% bovine serum albumin. Tracer solution was prepared as follows; 50 μCi of125I-[Tyr22]GVIA (DuPont NEN) was dissolved in 1 ml of water and divided into 50-μl aliquots, which were stored at −15 °C. Tracer for each assay was prepared by diluting one aliquot (2.5 μCi) into 3.5 ml of binding buffer for each 96-well microtiter plate to be used in the assay. Incubation mixtures (total volume 100 μl/well) consisted of 50 μl of binding buffer, containing displacing ligand, 20 μl of membrane preparation, and 30 μl of diluted 125I-[Tyr22]GVIA tracer solution, final concentration 0.13 nm. Nonspecific binding was determined in the presence of 1 μm unlabeled GVIA. Assays were incubated at 4 °C for 90 min and were terminated by filtration under vacuum. Each well was washed three times with 200 μl of wash buffer and then left under vacuum for 10 min to dry the filters. Filters were punched out and counted in a γ counter. Each measurement was determined in quadruplicate within an experiment, and each experiment was replicated at least three times. Data were analyzed using the computer program GraphPad Prism. Male Harlan Sprague Dawley rats were killed by CO2 anesthesia followed by decapitation, and the vasa deferentia, heart, and a portion of intestine with attached mesentery removed and placed in a dish of cool Krebs solution (composition in mm: Na+ 144, K+5.9, Mg2+ 1.2, Ca2+ 2.5, HPO4− 1.2, Cl− 129, SO4− 1.2, HCO3− 25, glucose 11, EDTA 0.026, bubbled with 5% CO2 in oxygen) for further dissection. Vasa deferentia were mounted in 5-ml organ baths, with the top of each tissue attached to an isometric force transducer (Grass FT03) and the bottom attached to a movable support and straddled with platinum stimulating electrodes. The vasa were stretched by a passive force of about 10 millinewtons and stimulated with single electrical field pulses (100 V, 0.2-ms duration) every 20 s. The resulting twitch responses were mediated by sympathetic nerves, being sensitive to inhibition by guanethidine (10 μm) or tetrodotoxin (0.1 μm), and were recorded on a chart recorder. GVIA (0.3–10 nm) caused a gradual concentration-related reduction in the size of the twitch response to electrical nerve stimulation in the rat vas deferens. Submaximally effective concentrations of GVIA elicited an initially rapid fall in the size of the twitch, and a continued very gradual decrease thereafter for over 90 min. This apparently slow equilibration meant that concentration-response curves with full equilibration of the toxin could not be constructed without interference from spontaneous fade of the twitch responses. To offset this problem, a protocol with fixed 20-min intervals between successive concentration increments was used. Cumulative concentration-response curves for GVIA or the analogues were constructed by addition of the peptides to the solution bathing the vas deferens in 10-fold concentration increments from 1 nm. A single concentration-response curve was constructed in each tissue. Under a dissecting microscope, an artery (three branch orders proximal to the arteries that enter the intestine) was carefully dissected free of the fat and connective tissue around it, and a 2-mm-long segment was mounted on 40-μm wires in a Mulvany-Halpern style isometric myograph and warmed to 37 °C. The artery was incrementally stretched radially with about four steps, and the force measured and the arterial circumference calculated at each step, producing a diameter-force curve where the diameter is that of a circle with the same circumference as the vessel at each level of stretch. The diameter of the artery was then set to be 90% of the diameter predicted for distending pressure of 100 mm Hg using standard calculations (31Mulvany M.J. Halpern W. Circ. Res. 1977; 41: 19-26Google Scholar). The rat mesenteric artery set up under these conditions does not develop any spontaneous active contractile force. Potassium depolarizing solution was applied for about 2 min to maximally activate the artery, and washed out. The prejunctional α2-adrenoreceptors were blocked with the covalent antagonist benextramine (3 μm for 5 min) in the presence of prazosin (0.1 μm) to protect the postjunctional α1-adrenoreceptors. The antagonists were then washed out, and noradrenaline (10 μm) was applied to confirm the washout of the antagonists. This procedure greatly increases the size of the responses to sympathetic nerve stimulation. Each jaw of the myograph was fitted with a platinum electrode about 1 mm away from the artery. Sympathetic nerves in the wall of the artery were stimulated with monopolar electrical field pulses of 0.25-ms duration at 30 V, stimulation parameters that give responses that are blocked >95% by tetrodotoxin. We have previously demonstrated that the responses are abolished by guanethidine and are thus mediated by sympathetic nerves. Stimulation of the sympathetic nerves produced contractile responses that had a short (<1 s) latency and decayed rapidly after the end of the train of stimuli. All responses were measured as the peak change in force. Nerve stimulation was applied as sets of three trains of 75 pulses at 25 Hz with a 60-s interval between trains, and 30 min between successive sets of stimuli. GVIA or analogues were applied in a cumulative fashion with 30 min of contact before each test stimulation. The right atrium was dissected free of the heart and placed in a 5-ml organ bath at 37 °C on a support having two fine platinum electrodes in contact with the atrium to collect the surface electrogram and another pair of electrodes for electrical stimulation. The spontaneous rate of contraction was continuously measured using the surface electrogram to trigger a period meter. Tachycardia responses mediated by sympathetic nerves were measured in response to sets of 4 electrical field pulses at 2 Hz, in the presence of atropine (1 μm) to abolish the effect of parasympathetic nerve stimulation. Increasing concentrations of GVIA or analogue were applied immediately following the second of two control stimulations, with the drug in contact with the tissue for 30 min before each stimulation. The central ear artery and marginal ear vein of New Zealand White rabbits of either sex (weight 2.5 ± 0.1 kg) were cannulated under local anesthesia (1% lignocaine hydrochloride) for measurement of blood pressure (MAP) and for drug injections, respectively. The phasic blood pressure signal triggered a rate meter for the measurement of heart rate (HR). Phasic and mean arterial pressure and HR were recorded on a Grass polygraph. Following the minor procedures as outlined above, rabbits rested quietly for about 40 min in polycarbonate restrainers. The effects of intravenous administration of selected GVIA analogues were assessed on MAP, HR, and the baroreflex. The baroreflex was measured by eliciting alternate graded steady-state increases and decreases in MAP (± 5–35 mm Hg from base line) with phenylephrine and sodium nitroprusside, respectively (32Head G.A. McCarty R. J. Autonomic Nervous System. 1987; 21: 203-213Google Scholar). GVIA analogue potencies were assessed by comparing their effects to the effect of 10 μg/kg GVIA. The GVIA analogues were administered at an initial dose of 3 or 10 μg/kg. MAP and HR were monitored for 60 min and the dose of analogue increased (cumulative half-log10 dose increments) if its effect was less than that of 10 μg/kg intravenous GVIA (33Wright C.E. Angus J.A. J. Cardiovasc. Pharmacol. 1995; 25: 459-468Google Scholar, 34Hawkes A.L. Angus J.A. Wright C.E. Clin. Exp. Pharmacol. Physiol. 1995; 22: 711-716Google Scholar). The baroreflex curve was then reassessed 60 min after the highest dose of peptide was administered and the reflex parameters (gain, location, and plateaus) obtained by fitting the baroreceptor-heart rate reflex curve to a logistic equation (32Head G.A. McCarty R. J. Autonomic Nervous System. 1987; 21: 203-213Google Scholar). Synthesis of all linear peptides proceeded smoothly, as indicated by monitoring of the Fmoc deprotection peak at each cycle. Oxidative refolding of linear peptide analogues gave a complex mixture of products, but the structures corresponding to the native fold were readily separated from non-native structures by the TEAP 2.25/MeCN chromatography system (35Hoeger C. Galyean R. Boublik J. McClintock R. Rivier J. Biochromatography. 1987; 2: 134-143Google Scholar) (Fig.1 A). Refolding of most of the linear peptide analogues gave purified yields of cyclized peptides of approximately 5% of the crude material. On the basis of HPLC analysis, only three of the oxidized analogues (G5A, N20A, and T23A) failed to show a major product with native fold that could be purified using the TEAP 2.25/MeCN chromatography system (Fig. 1, B–D). The amount of material with native-like structure in the oxidized mixtures of G5A, N20A, and T23A was therefore estimated by 1H NMR spectroscopy. The NMR spectrum of native GVIA contains five backbone amide proton resonances in the region 9.2 and 9.5 ppm (Fig. 2 A), which were also present in correctly folded Ala-containing analogues (e.g. Fig. 2 B) and thus served as a convenient marker of the native fold. Disulfide isomers or other incorrectly folded forms of the molecule could be distinguished from the native-like fold on the basis of these resonances (9Pallaghy P.K. Duggan B.M. Pennington M.W. Norton R.S. J. Mol. Biol. 1993; 234: 405-420Google Scholar). Thus, in cases where the purification scheme described above did not yield a significant quantity of the desired isomer, one-dimensional1H NMR spectra were employed to estimate the proportion of "native-like" fold in the oxidized mixture, as illustrated for T23A in Fig. 2 C. The estimate was based on the average height of the five downfield amide resonances indicative of native-like structure, with the spectrometer calibrated using a sample of native GVIA. By this means the contents of correctly-folded product in G5A, N20A, and T23A were calculated to be, respectively, 0.1, 1, and 3% of the oxidized peptide mixture (i.e. <0.15% of the crude material). Substitution of d-Ala at position 5, instead ofl-Ala, resulted in a normal refolding pattern to give a readily-isolated component corresponding to the native fold of the peptide. Sequential assignments were derived from two-dimensional TOCSY and NOESY NMR spectra for all of the analogues except G5A, N20A, and T23A. The chemical shift deviations from native were small enough to suggest that the native fold was maintained for all of these analogues. The average of the magnitude of the chemical shift deviations from native, ‖Δδ‖, provides a convenient single-parameter characterization of the structure and is tabulated in Table I for the NH and CαH resonances. The deviations of the CαH resonances were typically half the size of the NH deviations and so were less sensitive to conformational changes. The largest deviations occurred for the K2A, O4A, G5dA, O10A, S12A, Y13A, R17A, O21A, and Y27A analogues.Table ISummary of NMR chemical shift and NOE data for the 22 Ala scan analoguesAnalogueApproximate NMR quantity1-aApproximately proportional to the yield of oxidized polypeptide.‖Δδ‖av from native (NH)‖Δδ‖av from native (CαH)39 very weak peaks lostNo. of weak and medium NOEs lostNo. of NOEs gained1-bNo gained NOEs were observed for analogues with an NMR quantity of less than 7 mg.mgppmppm%K2A7.00.0940.045012S3A7.00.0130.016510O4A2.00.0840.05441G5dA3.00.0640.04956S6A7.00.0350.023801S7A5.00.0150.01110S9A1.80.0340.01626O10A4.00.0670.05626T11A7.00.0440.016800S12A7.00.0870.039321Y13A7.00.0580.040511N14A7.00.0430.0262120R17A7.00.0730.0451521S18A6.00.0200.02415O21A3.50.0850.03944Y22A7.00.0570.0271000K24A4.50.0470.023