Binding of Sea Anemone Toxin to Receptor Sites Associated with Gating System of Sodium Channel in Synaptic Nerve Endings in Vitro (Scorpion Neurotoxin/Synaptosomes) J

Proc. Nati. Acad. Sci. USA Vol. 77, No. 3, pp. 1646-1650, March 1980 Neurobiology

Binding of sea anemone toxin to receptor sites associated with gating system of sodium channel in synaptic nerve endings in vitro (scorpion neurotoxin/synaptosomes) J. P. VINCENT, M. BALERNA, J. BARHANIN, M. FOSSET, AND M. LAZDUNSKI Centre de Biochimie, Universit6 de Nice, Parc Valrose, 06034 Nice cedex, France Communicated by Heinz Fraenkel-Conrat, October 9, 1979

ABSTRACT Iodination of toxin II from the sea anemone nel. These polypeptide toxins (17-19) selectively alter the Anemonia sulcata gives a labeled monoiododerivative that re- functioning of the gating system of the channel without inter- tains 80% of the original neurotoxicity. This derivative binds fering with the binding of tetrodotoxin or saxitoxin at or near specifically to rat brain synaptosomes at 20'C and pH 7.4 with a second-order rate constant of association ka = 4.6 X 104 M-' the selectivity filter (4, 8). They have been shown to interact sect and a first-order rate constant of dissociation kd = 1.1 X with a large variety of excitable membranes including my- 10-2 sect. The binding occurs on the Na+ channel at a binding elinated and nonmyelinated axons (4, 5), neuronal cells in site distinct from that of other gating system toxins like batra- culture (8), cardiac and skeletal muscle cells in culture (7), and chotoxin, veratridine, grayanotoxin, aconitine, and pyrethroids. nerve terminals (20). The maximal binding capacity Bmax is 3.2 pmol/mg of protein We describe in this paper the preparation and purification (i.e., about two sea anemone toxin binding sites per tetrodotoxin of a labeled monoiododerivative (['251]ATX11) of toxin II from binding site) and the Kd is 240 nM for the monoiododerivative This derivative and 150 nM for the native toxin. Corresponding binding pa- the sea anemone Anemonia sulcata (ATX11). rameters for the association of a 1251-labeled derivative of toxin has been used to study properties of interaction between ATXII II from the scorpion Androctonus australis Hector are Bmax = and rat brain synaptosomes. Results are compared with those 0.3 pmol/mg of protein and Kd = 1 nM, whereas the Kd of the obtained when using a labeled iododerivative ([1251]AaH11) of unmodified scorpion toxin is 0.6 nM. Competition experiments toxini from the scorpion Androctonus australis Hector involving scorpion toxins, sea anemone toxins, and synaptosomes demonstrate that, although the sea anemone toxin is able (AaH1I). to displace the scorpion toxin bound to synaptosomes, the MATERIALS AND METHODS scorpion toxin does not displace the sea anemone toxin. The sea anemone toxin but not the scorpion toxin binds to depolarized Toxins I, II, and III from the sea anemone Anemonia sulcata synaptosomes. Differences between binding properties of the and a purified fraction from the venom of the scorpion Leiurus two polypeptide toxins are analyzed in the discussion. quinquestriatus (venom obtained from Sigma) were prepared as described (16, 21). Toxin II from the scorpion Androctonus A number of toxic molecules are now available for the analysis australis Hector was kindly given to us by S. Lissitzky's group of the structural organization and function of the voltage- (Marseille, France). dependent Na+ channel in excitable membranes. These toxins The guanidine side chain of the only arginine residue of the include (i) tetrodotoxin and saxitoxin, which bind at or near the ATX11 sequence, Arg-14, has been modified by 1,2-cyclohex- selectivity filter of the Na+ channels (1, 2); (ii) veratridine, anedione, using the method of Toi et al. (22) as described for batrachotoxin, aconitine, and grayanotoxin, which depolarize the modification of the bee venom neurotoxin apamin (23). excitable membranes by causing a persistent activation of the Amino acid analysis of the ATX11 derivative obtained after Na+ channels (1); and (iii) sea anemone and scorpion neuro- cyclohexanedione treatment indicates a selective and complete toxins, which slow down the closing (inactivation) of Na+ modification of Arg-14; none of the other amino acid residues channels in axons (3-5) and in various excitable cells in culture of thie toxin was modified. Total modification of Arg-14 results (6-10). in complete loss of toxicity. The biochemical characterization of receptors that selectively lodination of Toxins. ATXII was iodinated by the chlora- recognize each class of toxins is an important step towards the mine-T method (24). ATX11 (5 mg) was incubated with 5 mCi understanding of the molecular mechanisms of action of these (1 Ci = 3.7 X 0'10 becquerels) of Na'251 (Commissariat a toxins and the mode of functioning of the Na+ channel in ex- l'Energie Atomique or Amersham) and 0.1 ,Imol of unlabeled citable tissues. Receptors that have been best characterized NaI in 0.25 ml of 20 mM Tris-HCl (pH 8.6) at room tempera- biochemically are those of tetrodotoxin and saxitoxin (2, 11, 12). ture. Three aliquots (10 ul) of a 10 mM solution of chloramine-T Radiolabeled scorpion toxins have also been prepared and used (Merck) were added at 2-min intervals. The mixture was in- to identify Na+ channels (13-15). The binding of these poly- cubated for 30 more min after addition of the last chloramine-T peptide toxins seems to be strongly voltage dependent (16) and aliquot. Bovine serum albumin (200 Atl of a 1% solution) was for that reason will probably be difficult to use in biochemical then added and the mixture was immediately loaded on a studies of the Nab channel that necessitate working with SP-Sephadex C-25 column (2 X 21 cm) equilibrated and eluted membranes or detergent-solubilized fractions. with a 50 mM NaH2PO4 buffer (pH 5.5) containing 160 mM Sea anemone toxins are among the most interesting tools for analysis of the properties of the voltage-dependent Na+ chan- Abbreviations: ATX11, toxin II from the sea anemone Anemonia sulcata; [l25I]ATXII, radiolabeled monoiododerivative of ATXI with 1 The publication costs of this article were defrayed in part by page + 0.1 (SEM) iodine atom per mol of ATX~J; AaH11, toxin II from the charge payment. This article must therefore be hereby marked "ad- scorpion Androctonus australis Hector; [l2I]AaHIl, radiolabeled io- vertisement" in accordance with 18 U. S. C. §1734 solely to indicate doderivative of AaH1j with 0.69 + 0.1 (SEM) iodine atom per mol of this fact. AaH11; LD50, mean lethal dose. 1646 Downloaded by guest on September 23, 2021 Neurobiology: Vincent et al. Proc. Natl. Acad. Sci. USA 77 (1980) 1647 NaCI. The chromatography was followed by automatic re- RESULTS cording of the absorbance at 230 and 280 nm with a Seive Characterization of 125I-Labeled ATXI and AaHj1. ATXI1 elugraph and by measurement of the radioactivity in each is a single-chain protein of 47 amino acids crosslinked by three fraction with a Packard Tri-Carb liquid scintillation spectrometer disulfide bridges (18). Although there is no tyrosine residue in (model 2450 or 3390). AaHII was iodinated by the lactoperox- the sequence, there are two histidine residues, His-32 and idase method and purified by gel filtration as described (25). His-37, which can be iodinated. ATX11 was iodinated by the Different specific radioactivities (from 12.5 to 2000 Ci/mmol) chloramine-T method with a mixture of labeled (125I) and un- were obtained for both iodotoxins by varying the ratio of Na'25I labeled (1271) NaI. Fig. 1 shows that a chromatography of the to Na'271 used in the iodination procedures. iodination mixture on sp-Sephadex C-25 at pH 5.5 not only LD5o Measurements. LD50 measurements on mice or crabs separates the toxin from excess reagent (iodine and chlora- were carried out as described (25, 26). mine-T) but also divides the toxin into two different fractions. Synaptosome Preparation. Synaptosomes were prepared In five different series of experiments, the number of iodine from brains of male Sprague-Dawley rats (200-250 g body atoms per mol of toxin was found to be 1i 0.1 in the peak that weight) by the method of Gray and Whittaker (27) as modified is centered around fraction 24. The peak around fraction 43 by Abita et al. (20). Protein concentration was determined by consists of unmodified ATX1I. The mean lethal dose (LD5o) of the method of Hartree (28) with bovine serum albumin as a the monoiodo ATXI1 derivative is 0.21 mg/kg on mice and 2.2 standard. ,ag/kg on crabs compared to 0.17 mg/kg and 2.0 ,ag/kg, re- Binding Experiments. Cellulose acetate filters (Sartorius SM spectively, for the native toxin. Accordingly, we have prepared 11107, 0.2 1Am pore size) were used in binding experiments and purified a monoiododerivative of ATX11 that possesses more involving synaptosomes and either ['25I]ATXII or [125I]AaHII. than 80% of the original neurotoxicity. They were kept in 1% bovine serum albumin for 2 hr and then Scorpion toxin II from Androctonus australis Hector was washed once with5 ml of the incubation buffer before use. The iodinated by the lactoperoxidase method and purified by fil- standard incubation buffer consisted of 140 mM choline chlo- tration on Sephadex as described (25). The number of iodine ride, 5.4 mM KCI, 2.8 mM CaCI2, 1.3 mM MgSO4, 0.1% bovine atoms per mol of toxin is 0.69 : 0.1. We have checked that the serum albumin, and 20 mM Tris-HCI, pH 7.4 (choline medi- specific toxicity of the scorpion neurotoxin was not modified um). As indicated in the text, identical concentrations of NaCI by iodination under the conditions used, as reported (25). This (sodium medium) or KCI (potassium medium) were occasion- derivative was used for binding experiments without further ally substituted for choline chloride. purification. (i) Kinetics of association and dissociation of [125I]ATXII Association and Dissociation Kinetics of the Interaction to synaptosomes. Synaptosomes (1 mg of protein per ml) were Between [125I]ATXII and Synaptosomes. Typical kinetics of incubated at0°C or 20°C in the choline medium. Association association between [125I]ATXII and synaptosomes at0°C and kinetics were started by addition of [125I]ATXII (5-50 nM). 20°C are shown in Fig. 2A. These experiments were carried out Aliquots (0.2 ml) were taken at different times and filtered with an [125I]ATXII concentration of 11 nM. The maximal under reduced pressure through Sartorius filters. Filters were concentration of specifically bound [125I]ATXII, which corre- rapidly washed twice with 5 ml of the incubation buffer and sponds to 100% in Fig. 2A, was 0.14 nM at 0°C and 20°C. The placed in counting vials containing 8 ml of Bray's solution which free [125I]ATXII concentration varied less than 1.5% during totally dissolves them. The radioactivity retained by the filter the course of the association kinetics. Fig. 2A Inset shows that was determined at a counting efficiency of 55-60% by liquid scintillation spectrometry. After 30 min of association, the amount of specifically bound Albumin [125I]ATXii reached a plateau value. At that time, the incubation I2SI ATX medium was made 10 ,uM with unlabeled ATXII. Under these 3X105 Exess 0.3 conditions, unlabeled ATXII is in large excess compared to iodine ['25I]ATXII and displaces the labeled toxin from its association to synaptosomes. Dissociation kinetics were followed by mea- Unmodified suring the decrease in bound [125I]ATXII with the filtration ,"2X105 -ATX11 0.2 technique described above. (ii) Equilibrium binding experiments involving synaptosomes and [125I]ATXII or [25I]AaHHI. Synaptosomes (1 mg of protein per ml) were incubated with increasing concentrations i05 Ch oramine T 0.1 of [125I]ATXII or [125I]AaHI in 1 ml of choline medium for 30 min at 20°C. Quadruplicate aliquots (0.2 ml) of the incubation medium were then filtered on Sartorius filters and the bound radioactivity was measured as described above. Nonspecific 0-- 0 binding was determined in parallel experiments in the presence 0 10 20 30 40 50 60 of an excess of unlabeled toxin (10MuM for ATXII, 100 nM for Fraction FIG. 1. Purification of [125I]ATXII. ATX11 (5 mg) was iodinated AaHII). by the chloramine-T method. After completion of the iodination re- (iii) Competition experiments involving synaptosomes and action, 200 ul of 1% bovine serum albumin was added and the mixture labeled or unlabeled ATXII and AaHi. Synaptosomes (1 mg was loaded on an SP-Sephadex C-25 column (2 X 21 cm) equilibrated of protein per ml) were incubated for 30 min at 20°C with a and eluted with 50 mM NaH2PO4 (pH 5.5) containing 160 mM NaCl. fixed concentration of labeled iodotoxin (10 nM for [125I]ATX11 The chromatography was followed by automatic recording of the and 0.35 nM for [125I]AaH1I) and various concentrations of absorbance at 280 nm (0) and by measurement of the radioactivity unlabeled toxin in 1 ml of choline medium. The amount of la- of 10-sl aliquots of each fraction by liquid scintillation spectrometry (x). Fractions containing ATX11 were characterized by toxicity beled iodotoxin that remained bound to synaptosomes in the measurement on crabs. The ATXIIconcentration was determined by presence of the unlabeled toxin was estimated as described in using A"F0 = 27.9 (21). Fraction volume, 4.5 ml; flow rate, 1 ml/ method ii. min. Downloaded by guest on September 23, 2021 1648 Neurobiology: Vincent et al. Proc. Natl. Acad. Sci. USA 77 (1980)

_FO~ ~ ~ o2 cn 0, E E 'Z- E0E E 0 ~0Q

o6 012 4 0:3 E 0 .0 Time,-0 Time, min :3I x o -0 er _25

- Time, min 0 5 10 15 20 0 5 10 15 20 Time, min 0 .2 .4 .6 .8 1.0 1 2 3 [ free, ["'I free, nM FIG. 2. Association and dissociation kinetics for the binding of "'I1]ATX1I PM ]AaH1I [1251]ATXII to synaptosomes. (A) Kinetics of association of [125I1- FIG. 3. Comparative binding properties of [12511ATX11 and ATX11 (11 nM) to synaptosomes (1 mg protein per ml) at 00C (0) and ['25IIAaHii to synaptosomes. Synaptosomes (1 mg of protein per ml) 200C (0) in the choline medium. (Inset) Pseudo first-order repre- were incubated in the choline medium with increasing concentrations sentation of the data. X, percentage of maximal [1251]ATXII bound of [12511ATX11 (A) or [125I]AaHii (B). After an incubation time of 30 to synaptosomes. The plateau value (100%) corresponds to a binding min at 20°C, the radioactivity bound to synaptosomes was determined of 140 fmol/mg of protein at both temperatures. (B) Dissociation ki- by the filtration technique. Each value shown is the mean of qua- netics of the [125IJATXII-synaptosome complex at 00C (3) and 20'C druplicate determinations that varied less than 5%. Nonspecific (-). After a plateau value has been reached in the association kinetics binding (0) was determined in the presence of a large excess of un- described in A, dissociation of [1251]ATXII was initiated by addition labeled toxin (10 AM for ATX11, 100 nM for AaHII). Specific binding of 10 AsM unlabeled ATX11. (Inset) Pseudo-first-order representation (0) is the difference between total binding (not shown) and nonspe- of the data. cific binding. (Insets) Scatchard plots of the data. B, bound; F, free. a semilogarithmic plot of the data is linear at both temperatures assayed, which is expected for a pseudo-first-order reaction. The which increasing concentrations of [1251]ATXII (Fig. 3A) or of rate constant of the association is then: [1251]AaHii (Fig. 3B) were added to a fixed concentration of synaptosomes, either in the presence (nonspecific binding) or k = ka(['25I]ATXII) + kd, in the absence (total binding) of a large excess of the corre- where ka and kd represent the second-order rate constant of sponding unlabeled toxin (10 1M for ATX11, 100 nM for AaH11). association and the first-order rate constant of dissociation, The specific binding is defined as the difference between total respectively, of the [125I]ATXII-receptor complex. In the course and nonspecific binding. The dissociation constant of the of eight different experiments similar to those illustrated in Fig. complex formed between [125I]ATXII and synaptosomes is 240 2A, the mean values for k at 00C and 200C were found to be nM and the maximal binding capacity is 3.2 pmol/mg of pro- 2.6 i 0.7 10-3 sec-1 and 1.3 + 0.4 10-2 sec-1, respectively. tein. Corresponding values for ['25I]AaH1j are 1 nM and 0.3 Fig. 2B demonstrates that [125I]ATXII bound to synaptosomes pmol/mg of protein, respectively. Linearity of the Scatchard can be displaced by unlabeled ATXII. Because the presence of plots (Insets) demonstrates that each iodotoxin binds to a single a large excess of unlabeled ATXII (10 ,uM) prevents the reas- class of noninteracting binding sites. sociation of [125I]ATXII to synaptosomes, the dissociation process Displacement of 1251-Labeled Toxins Bound to Synapto- should be first-order with kd as rate constant. As expected somes by Unlabeled Toxins. Results presented in Fig. 4 show the were the following. (i) Unlabeled ATX1i displaces [125I]ATXII from semilogarithmic representations of kinetics data linear its association to synaptosomes (Fig. 4A). The concentration of at both temperatures studied (Fig. 2B Inset). Values of kd unlabeled ATX11 that induces half-displacement of bound measured in eight different series of experiments at each [125I]ATXii is KO.5 = 150 nM. Under our experimental condi- temperature were 3 + 0.7 10-3 sec-I (00C) and 1.1 + 0.4 10-2 tions ([125I]ATXII = 10 nM, much smaller than Kd), KO.5 rep- sec-1 (200C). Values of k and kd at 0°C or 20°C were very resents the dissociation constant Kd of the unlabeled ATX11- similar if not identical. The most simple interpretation of this synaptosome complex. This value of 150 nM is similar to the Kd result is that ka ([125I]ATXII) is always much smaller than kd so value obtained by direct binding experiment with [r25I]ATXII that k = kd. If this assumption is correct, then k should be in- [Kd = 240 nM (see Fig. 3)]. (ii) Toxins I, II, and III from Ane- dependent of [125I]ATXII. We have indeed checked that, at 0°C monia sulcata are nearly equally toxic on crustaceans, their and 200 C, the [125I]ATXII concentration can be varied from 5 LD50 being 2, 2, and 6 ,ug/kg of crab, respectively. These toxins to 50 nM without changing the value of k. Under our experi- have very different toxic activities to mice. ATX11 is very toxic mental conditions it is only possible to measure kd, the first- with a LD50 of 170 ,ug/kg of mice; toxins I and III have a low order rate constant of dissociation of the [125I]ATXII-synapto- toxicity with LD50 values higher than 15 mg/kg. It is shown in some interaction. Fig. 4A that toxins I and III are much less potent than ATXII As will be seen later, the value of Kd, the equilibrium disso- in displacing [125I]ATXII bound to rat brain synaptosomes. None ciation constant of the [1251]ATXII-synaptosome complex, is of these toxins significantly displaced [125I]ATXII at a concen- known. It is then possible to calculate the value of ka. At 200C, tration of 1 ,M. The KO.5 for neurotoxin I is 8 M-i.e., 53 times Kd = kd/ka = 240 nM and kd = 1.1 X 10-2 sec1-, so that ka = higher than the KO.5 for ATX11. Fig. 4A also shows that the kd/Kd = 4.6 X 104 M-1 sec-1. nontoxic derivative of ATX1I in which Arg-14 was selectively Specific Binding of [125I]ATXI and [125I]AaHII to Synap- modified was unable to displace [125I]ATXII from its receptor tosomes. Fig. 3 shows the results of binding experiments in site. Downloaded by guest on September 23, 2021 Neurobiology: Vincent et al. Proc. Natl. Acad. Sci. USA 77 (1980) 1649

OA A C .0 0FIG. 4. C t b w 00I o0ia

pressed iprnaomxmbni.AA B

G)50_~~~~~WZ50~_ C.) at0)- ~ -o K+ lg[ertiie 0. unable) Aoin fo bnigt Iyats .(A)C petonbe F1.5 6fetfdplrzaino.h bidn of ll5I]TX n h asne(-100 )o nMttoooio Aih03 nM 9 8 7 6 5 : 10 9 8 7 6 5 ori00 -log[unlabeled toxin] FIG. 4. Competition between 12-51-labeled ATX11 or AaH11 and unlabeled toxins for binding to synaptosomes. (A) Competition between [1251JATXII and unlabeled sea anemone or scorpion toxins for binding to synaptosomes. [1251JATX11 (10 nM) was incubated with synaptosomes (1 mg of protein per ml) and increasing concentrations of native Anemonia sulcata toxin I (u), 11 (-), or III (A), Arg-14- by~poasim~-oSyaposme an-lbletoxin weretiincuaedl o or from Lelurus modified ATX11 (v), AaHIJ (O), scorpion toxin mediu intsence) orin the absence (clofsym quinquestriatus venom (0) in the choline medium. (B) Competition wermincuatee0Cithaof erdith 1otinn0nM M21TXrinsthe horid tandv)8 between [l25I]AaHII and unlabeled AaH1j and ATX1I for binding to concentratioctKeplaiandthe conplementindcoing (preecso51AI7(4, orintht-Cabsence[veratridine-intag(M, 0)f1004e of bvnMtetrodotoxiniorbinding.the ofdwithc0.35inMlarizat synaptosomes. [125I]AaH11 (0.35 nM) was incubated with synapto- at 1]avalusynapttof 4. osomes.Effptoectoseardn.M B SnaptosomesSynaptosomes(Imandrtenpe( oreinc pereml)l somes (1 mg of protein per ml) and increasing concentrations of un- ioooiswere[oncentratins['2potaHsumtoincubatedinKeithepencthewih30n )orpleinthmina120°CAeitheabsnchiein(0,or0)prsencdiumVthe sof100an labeled AaHjj (0) or ATX11 (0) in the choline medium. The radio- percentageo maxma activity bound to synaptosomes was determined after incubation for pressaledin15. M bindriingnAfectodenptolarizatind si x 30 min at 20'C. byiotasu.Snpoomsadlbldtoxinswthethedtoindicthed bindingfbtrationsofvrtridine.osnposwerefor30mna20Cetrinhesduincubated 30dimmthetsrodooxineato2pe n aybl)obinofd ofcotanig20MorsedlpnbodtImediumhhodotoxtnionshl wsynaos7.4),o2.8ismex- (iii) Neither AaHI nor L. quinquestriatus unlabeled scorpion wihteidctdconcentrationsKvnof h opementrinhoine.clrd ota toxin was able to displace [125IATXII bound to synaptosomes, thaemsum ofchoin chlorideyandKl cncentrationsalwayspremaints even at concentrations as high as 10 MiM (Fig. 4A). (iv) Unla- beled AaHII displaced [125I]AaHIj with a KO.5 of 0.8 nM (Fig. at asau f155m.()Efcynaptosomesfvrtiie an 4B), which corresponds to a Kd of 0.6 nM, similar to the Kd mhediumsopetinsmols or2 in1the cholnemediumei(cose symbols) value found in direct binding experiments involving synaptosomes and ['25j]AaHIj (Kd = 1 nM, Fig. 3B). (v) Unlabeled ATXII was also able to displace [125IAaHII associated to synaptosomes with a KO.5 of 210 nM (Fig. 4B), which corresponds by the following concentrations of other toxins: tetrodotoxin to a Kd of 160 nM, in satisfactory agreement with the KO.5 value (100 nM), saxitoxin (100 nM), batrachotoxin (10 MM), veratri- (150 nM) obtained in competition with [12'I]ATXII (Fig. 4A) dine (300,uM), grayanotoxin (100,M), aconitine (100,M), or and with the Kd value (240 nM) obtained by direct binding pyrethroids (10,M). experiments (Fig. 3A). Membrane Potential Dependence for the Binding of DISCUSSION [1251]ATXI and [125I]AaHI to Synaptosomes. Synaptosomes This paper shows that it is possible to prepare a monoiodo- isolated from rat brain are able to keep their original membrane derivative of ATX11 by modification of one of the two histidine potential provided that they are incubated in a medium of residues of the toxin structure. This derivative retains a toxic suitable osmolarity and potassium concentration (29). Because activity very similar to that of the native toxin. Binding studies the mode of action of scorpion toxins has been described as of [125I]ATXII to synaptosomes have given the following results: membrane potential-dependent (16), we have studied the effect (i) The binding is saturable and specific. Anemonia toxins I and of depolarizing agents on the binding of both toxins to synap- III, which are only slightly toxic to mammals, have a much tosomes. Fig. 5A shows that an increase of the potassium con- lower affinity than ATX11 for the Na+ channel of rat synapto- centration in the incubation medium, which induces depolar- somes. Moreover, the nontoxic derivative of ATXII obtained ization of synaptosomes (29), does not modify the binding of by the selective modification of Arg-14 is unable to displace [125I]ATXII but totally prevents the binding of ['25I]AaHIj to [125I]ATXII binding from its receptor. (ii) The dissociation synaptosomes. The potassium concentration that induced process of the ATX11-receptor complex is slow enough at 0°C; half-dissociation of the [12'I]AaHij-synaptosome complex was a half-time of 4 min for the dissociation permits crosslinking 8 mM. experiments of [125I]ATXII to its receptor similar to those that Another way to depolarize synaptosomes is to use veratridine have been described by Biesecker for the snake neurotoxin- in the presence of Na+ (29). Fig. 5B shows that veratridine nicotinic receptor interaction (30). Such studies would give very concentrations up to 300,uM are without effect on the binding useful biochemical information on the chemical structure of of [125I]ATXi either in the choline or in the sodium medium. the ATX11 receptor. (iii) There is a large number of ATX11 Conversely, the binding of [12'I]AaHII was decreased by 50% binding sites in synaptosomes. The stoichiometry of binding, or 80% when synaptosomes were incubated with veratridine 3.2 pmol/mg of protein, compared to the known stoichiometry at 6 ,M or 300 ,M, respectively, in the sodium medium. This of binding of [3H]tetrodotoxin to synaptosomes [1.7 pmol/mg effect can be fully reversed in the presence of 100 nM tetro- of protein (20)] indicates a ratio of about two ATX11 binding sites dotoxin, which is known to prevent the depolarizing action of per tetrodotoxin binding site. This fairly large stoichiometry veratridine by closing Na+ channels (1). That this effect is due of binding suggests that brain synaptosomes can be a good to depolarization and not to a competition of veratridine for the starting material for the biochemical isolation of the gating [125I]AaH11 binding sites is demonstrated by the fact that re- system machinery of the Na+ channel. (iv) Neither tetrodotoxin Downloaded by guest on September 23, 2021 1650 Neurobiology: Vincent et al. Proc. Natl. Acad. Sci. USA 77 (1980) nor saxitoxin, which act at or near the selectivity filter, displaces supported by the Centre National de la Recherche Scientifique, the ATX11 from its receptor. Moreover ATX11 binding sites are Commissariat a I'Energie Atomique, the Delegation a la Recherche distinct from those of other toxins that interact with the gating Scientifique et Technique, the Fondation pour la Recherche M6dicale, system of the Na+ channel-i.e., veratridine, batrachotoxin, and the Institut National de la Sante et de la Recherche Medicale. grayanotoxin, aconitine, or pyrethroids. The 1. Narahashi, T. (1974) Physiol. Rev. 54, 813-889. comparison of the binding properties of [L251]ATXI and 2. Ritchie, J. M. & Rogart, R. B. (1977) Rev. Physiol. Biochem. [11'I]AaH11 to synaptosomes has provided a series of interesting Pharmacol. 79, 1-50. observations. These two polypeptide toxins are known to have 3. Romey, G., Chicheportiche, R., Lazdunski, M., Rochat, M., Mi- similarities in their mode of action. Both of them slow down the randa, F. & Lissitzky, S. (1975) Biochem. Biophys. Res. Commun. inactivation process (i.e., the closing of the Na+ channel in 64, 115-121. axons) (3, 4) and AaH11 stimulates neurotransmitter efflux from 4. Romey, G., Abita, J. P., Schweitz, H., Wunderer, G. & Lazdunski, synaptosomes (31) by a mechanism identical to the ATX11- M. (1976) Proc. Natl. Acad. Sci. USA 73, 4055-4059. induced efflux (20). In spite of these physiological analogies, 5. Bergman, C., Dubois, J. M., Rojas, E. & Rathmayer, W. (1976) no sequence homology can be found in the chemical structure Biochim. Biophys. Acta 455, 173-184. 6. Catterall, W. A. (1976) J. Biol. Chem. 251, 5528-5536. of these polypeptide toxins (18, 32). 7. De Barry, J., Fosset, M. & Lazdunski, M. (1977) Biochemistry Results presented in this paper show that the structural dif- 16,3850-3855. ferences between ATX11 and AaH1j are reflected by differences 8. Jacques, Y., Fosset, M. & Lazdunski, M. (1978) J. Biol. Chem. in binding properties to synaptosomes. (i) ATX11 binds to sy- 253, 7383-7392. naptosomes with an affinity (Kd = 150 - 240 nM) that is about 9. Romey, G., Jacques, Y., Schweitz, H., Fosset, M. & Lazdunski, '/iooth that of AaH1j (Kd = 0.6 - 1 nM). (ii) The number of M. (1979) Biochim. Biophys. Acta 556,344-353. binding sites for ATX1I (3.2 pmol/mg of protein) is more than 10. Bernard, P., Couraud, F. & Lissitzky, S. (1977) Biochem. Biophys. 10 times higher than the number of binding sites for AaH11 (0.3 Res. Commun. 77,782-788. 11. Agnew, W. S., Levinson, S. R., Brabson, J. S. & Raftery, M. A. pmol/mg of protein). (iii) Competition experiments show that, (1978) Proc. Natl. Acad. Sci. USA 75, 2606-2610. although ATX11 is able to fully displace AaH1j from its associ- 12. Chicheportiche, R., Balerna, M., Lombet, A., Romey, G. & ation to synaptosomes, the reverse is not true. No displacement Lazdunski, M. (1979) J. Biol. Chem. 254, 1552-1557. of ATX11 by AaH1j or by a scorpion toxin isolated from Leiurus 13. Ray, R., Morrow, C. S. & Catterall, W. A. (1978) J. Biol. Chem. quinquestriatus venom has been observed even at very high 253, 7307-7313. concentrations (10 ,uM) of the scorpion toxins. (iv) The binding 14. Catterall, W. A. & Beress, L. (1978) J. Biol. Chem. 253, 7393- of ATX11 is strongly dependent upon the membrane potential 7396. of synaptosomes, whereas the binding of ATX11 occurs even on 15. Jover, E., Martin-Moutot, N., Couraud, F. & Rochat, H. (1978) completely nerve Biochem. Biophys. Res. Commun. 85, 377-382. depolarized terminal particles. 16. Catterall, W. A. (1977) J. Biol. Chem. 252, 8660-8668. The most intriguing result concerns differences in stoichi- 17. Wunderer, G. & Eulitz, M. (1978) Eur. J. Biochem. 89, 11-17. ometries between scorpion toxin and sea anemone toxin re- 18. Wunderer, G., Fritz, H., Wachter, E. & Machleidt, W. (1976) ceptors. Only two explanations can be given to this observation. Eur. J. Biochem. 68, 193-198. The first one is that there are 10 times more ATX11 receptors 19. Martinez, G., Kopeyan, C., Schweitz, H. & Lazdunski, M. (1977) in a Na+ channel than there are AaH11 receptors. The second FEBS Lett. 84, 247-252. one is that there are different classes of Na+ channels in sy- 20. Abita, J. P., Chicheportiche, R., Schweitz, H. & Lazdunski, M. naptosomes, only one of them being recognized by AaH11. The (1977) Biochemistry 16, 1838-1844. second interpretation seems much more realistic than the first. 21. Beress, L., Beress, R. & Wunderer, G. (1975) FEBS Lett. 50, AaH11 would recognize only the less abundant class of sites (0.3 311-314. 22. Toi, K., Bynum, E., Norris, E. & Itano, H. A. (1967) J. Biol. Chem. pmol/mg of protein) in which the gating system of the Na+ 242, 1036-1043. channel is in a polarized conformation. ATX11 would obviously 23. Vincent, J. P., Schweitz, H. & Lazdunski, M. (1975) Biochemistry also recognize this class of sites because it can displace [125I1] 14, 2521-2525. AaH11 binding. But it would also bind to another larger class of 24. Hunter, W. M. & Greenwood, F. C. (1962) Nature (London) 194, sites (3.2 - 0.3 = 2.9 pmol/mg of protein) that are unable to 495-496. bind AaH11 either because they are not in a polarized confor- 25. Rochat, H., Tessier, M., Miranda, F. & Lissitzky, S. (1977) Anal. mation or because they are devoid of a component that is nec- Biochem. 82, 532-548. essary for AaHIJ binding. The existence of two classes of sites, 26. Beress, L. & Beress, R. (1971) Kieler Meeresforsch. 27, 117- if they do exist, could be due to a heterogeneity of polarization 127. In sea anemone toxins 27. Gray, E. G. & Whittaker, V. P. (1962) J. Anat. 96,79-88. of the nerve terminal particles. any case, 28. Hartree, E. F. (1972) Anal. Biochem. 48, 422-427. seem to be very good tools for titration of the gating system of 29. Blaustein, M. P. & Goldring, J. M. (1975) J. Physiol. (London) the Na+ channel in any state of polarization, whereas scorpion 247, 589-615. toxins can be used for the titration of polarized channels. 30. Biesecker, G. (1973) Biochemistry 12, 4403-4409. 31. Romey, G., Abita, J. P., Chicheportiche, R., Rochat, H. & Laz- We thank Dr. Schweitz for kindly providing the sea anemone toxin dunski, M. (1976) Biochim. Biophys. Acta 448,607-619. used in this work, M. C. Lenoir and N. Alenda for expert assistance, 32. Rochat, H., Rochat, C., Sampieri, F., Miranda, F. & Lissitzky, and R. Corda for skillful fishing of the sea anemone. This work was S. (1972) Eur. J. Biochem. 28,381-388. Downloaded by guest on September 23, 2021