Is the Site of Action of Grayanotoxin the Sodium Channel Gating of Squid Axon?

Transcription

1 Japanese Journal of Physiology, 35, ,1985 Is the Site of Action of Grayanotoxin the Sodium Channel Gating of Squid Axon? Issei SEYAMA, Kaoru YAMAOKA, Masuhide YAKEHIRO, Yoshiake YOSHIOKA, and Kazuyuki MORIHARA* Department of Physiology, School of Medicine, Hiroshima University, Hiroshima, 734 Japan Abstract An attempt was made to elucidate the site of action of grayanotoxin (GTX) in the nerve membrane by using various endopeptidases. The experiment was conducted on squid axons isolated from Doryteuthis bleekeli with both voltage clamp and internal perfusion methods. Intracellular application of various endopeptidases for more than 30 min eliminated the gating action from both Na current and K current systems. When GTX (100 µm) was subsequently applied to the internal medium, the membranes could depolarize to various extents. This finding strongly suggests that the site of action of GTX is not confined to the channel gating but is present in a part of the Na channel having both voltage sensor and ion filter functions. With the application of trypsin, St. fradiae trypsin, pronase, BPN', and St. fradiae protease (group B), GTXinduced depolarization was much smaller than that with the application of a-chymotrypsin, N-protease, and thermolysin (group A). The difference in the sensitivity to GTX between group A and group B became remarkable as the time for application of the enzymes was prolonged. Since all enzymes belonging to group B retain trypsin-like activity and are more effective in removing the sensitivity to GTX, it is suggested that the molecular moiety around the binding site of GTX is rich in basic amino acids or the essential part for opening the Na channel should be protected by basic amino acids. Key words: grayanotoxin, depolarization, proteases, squid axon. Since HODGKIN and HUXLEY (1952) have proposed that nerve excitation is regulated by the gating and ion selective mechanisms of individual ion channels, neurotoxins which modify the properties of a particular channel have been extensively used as a pharmacological tool for elucidating the molecular mechanism Received for publication November 24, 1984 * Present address : Toho Pharmaceutical Ind. Co., Kyoto Research Laboratories, Nagaokakyo, Kyoto, 617 Japan 401

2 402 I. SEYAMA et al. of nerve excitation. It has been shown that grayanotoxin derivatives which are extracted from the leaves of the family of Ericacae specifically increase the membrane permeability to Na+ in the Na-dependent excitable membranes (NARAHASHI and SEYAMA, 1974; SEYAMA and NARAHASHI, 1981). The mode of action of GTX has been clarified as follows: (1) Na channel remains open at a membrane potential far below the ordinary threshold membrane potential by shifting the relationship between both the activation and inactivation parameters and the membrane potential to the hyperpolarizing direction, (2) a part of the Na channel loses the inactivation mechanism at the later stage of intoxication (GTX-induced slow current), and (3) the selectivity during GTX-induced slow current is much lower than that during normal Na current. Thus, GTX exerts various pharmacological actions, such as extensive transmitter release at the frog neuromuscular junction (ZusHI et al., 1983), positive inotropic effect on the myocardium (SEYAMA, 1978), and negative chronotropic effect on S-A node cells (NAKAO and SEYAMA, 1984). These findings suggest that GTX is a potential candidate for a pharmacological tool in examining the Na channel. Since the increase in membrane permeability to Na+ induced by GTX is due to modification of the Na channel, it is assumed that peptide bonds in the Na channel protein critical for GTX binding may be cleaved by certain proteases and GTX may fail to depolarize the membrane. Thus, this experiment was designed to obtain information on the molecular structure of the Na channel by correlating the loss of depolarizing response to GTX with the substrate specificity of the enzyme. A preliminary report has been published elsewhere (SEYAMA and YOSHIOKA, 1982). METHODS The experiments were conducted on squid axons isolated from Doryteuthis bleekeli. A combined method for internal perfusion and voltage clamp was used. Artificial sea water (ASW) had the following composition (in mm) : NaCI 449, KCl 10, CaCl2 50, and Tris 30. Temperature corrected ph was adjusted to 8.0 with HCI. The standard internal solution (in mm) was NaF 50, K-glutamate 320, K-phosphate buffer 15, and sucrose 333. Its ph was adjusted to 7.3. The temperature was maintained constant at 12 C throughout the experiments. The enzymes used in this experiment were a-chymotrypsin (3 x recrystallized, Worthington Biochemicals, Freehold, New Jersey), trypsin (3 x recrystallized, Worthington Biochemicals), pronase (Kaken Pharmaceutical Co., Ltd., Tokyo), subtilisin BPN' (BPN') (Nagase Seikagaku, Osaka), a neutral protease from Bacillus subtilis var. amylosacchariticus (N-protease) (recrystalized, Seikagaku Kogyo, Co., Ltd., Tokyo), Streptomyces fradiae proteases (St. fradiae proteases) (Shionogi Research Lab., Osaka; enzyme was prepared by the method developed by MORIHARA et al. Japanese Journal of Physiology

3 PROTEASES AND GRAYANOTOXIN 403 (1967)), Streptomyces fradiae trypsin (St, fradiae trypsin) (Shionogi Research Lab., Osaka; the enzyme was prepared by the method developed by MORIHARA and Tsuzuiu (1968)), and thermolysin (recrystallized, Daiwa Kasei, Co., Ltd., Tokyo), prepared immediately before use. The stock solution of GTX I was prepared by dissolving GTX I into ethanol and diluting with distilled water. Final concentration of stock solution was 6 x 10 M in 60% ethanol solution. RESULTS Time course of decay of both peak and steady-state currents after intracellular application of proteases It has been shown that the gating mechanism for the Na channel is on the intracellular surface (TASAKI and TAKENAKA, 1964; ARMSTRONG et al., 1973). Furthermore, NARAHASHI and SEYAMA (1974) have suggested that GTX acts from the internal surface and shown that GTX acts more strongly and quickly with internal application than with external application and the recovery after washing is also faster with internal application. Thus, application of proteases for eliminating or modifying the Na channel gating should be limited to the internal medium. Proteases (1 mg/ml) tested in this experiment were classified into two groups, Time (min) Fig, 1. Time course of decay of both INa and Ig during the intracellular application of St. fradiae proteases (group B). The membrane was held intermittently at the membrane potential of -70 mv and was then stepped up to -30 mv. The duration of depolarizing pulse was 6 msec. As the digestion proceeds, both currents decrease, indicating that enzyme digests both current systems. Vol. 35, No. 3, 1985

4 404 I. SEYAMA et al. Fig. 2. Time course of decay of both INa and T during the intracellular application of a- chymotrypsin. The experimental procedure was the same as that described in Fig. 1. Fig. 3. Effect of intracellular application of a-chymotrypsin on,-v relationship of squid axon. INa ( ) and Ig (.) during intracellular digestion by a-chymotrypsin were sampled in 30 min. Holding potential was -70 my. The duration of stimulating pulses was 6 msec. Upper inset figure shows the family of voltage clamp records in the control and lower inset figure those after enzymatic treatment. Japanese Journal of Physiology

5 PROTEASES AND GRAYANOTOXIN ms c Fig. 4. Effect of intracellular application of St. fradiae proteases on I V relationship of squid axon. The experimental procedure was the same as that described in Fig. 3. depending on whether depolarization induced by GTX after intracellular digestion with proteases remained constant (group A) or not (group B), as described later. When the axons were subjected to all the examined proteases, a small depolarization in the range between 1 and 15 mv occurred depending on the exposure time. During the action of proteases, the membrane was intermittently held at the membrane potential of -70 mv and was stepped up to -30 mv. The membrane currents associated with step depolarization were recorded in order to detect the degree of enzymatic modification of the Na channel gating. Both transient (INa) and steady state (Ix) currents decreased and eventually disappeared after subjecting the membrane to proteases for more than 30 min (Figs. 1 and 2). As shown in Figs. 3 and 4, I -V curve (filled symbols) at the phase when the axon lost its excitability was remarkably different from control I -V curve (open symbols). I -V relation became ohmic, indicating that the gating processes for both current systems were either eliminated or failed to respond to change in membrane potential. Depolarization induced by GTX on the axon which had been treated by proteases After confirming the elimination of the gating function of the Na channel, GTX (100,uM) was introduced intracellularly. In the case of treatment with a- chymotrypsin, N-protease, and thermolysin (group A), GTX was able to depolarize Vol. 35, No. 3, 1985

6 406 I. SEYAMA et al. Fig. 5. Relationship between GTX-induced depolarization and duration of application of group A enzyme solutions. Each mark represents one example after treatment of the enzyme shown by a particular symbol for a given time. Straight line was adopted to all data by the least square method. Fig. 6. Relationship between GTX-induced depolarization and duration of application of group B enzyme solutions. Each mark represents one example after the treatment of enzyme shown by symbols for a given time. Although trypsin has a strict substratespecificity, the response of the axonal membrane to GTX after trypsin treatment ( ) is different from that after treatment with enzymes belonging to group B. Each straight line was fitted to the data shown in different symbols by the least square method. Japanese Journal of Physiology

7 PROTEASES AND GRAYANOTOXIN 407 the membrane substantially (Fig. 5). In most cases, the membrane potential reversed polarity and attained a value between +2 and +5 mv. However, in normal axons, the membrane potential at which depolarization by the internal application of GTX was finally attained was always about 10 mv more positive than that by GTX after enzyme treatment. Since generation of a small depolarization after treatment with enzymes, as suggested in the previous section, implies that the membrane becomes leaky, it is understandable that depolarization attained after enzyme treatment plus GTX application is always smaller than that by application of GTX alone. On the other hand, the treatment of axons with BPN', trypsin, pronase, St. fradiae trypsin, and St. fradiae proteases (group B) weakened and eventually eliminated the action of GTX (Fig. 6). It is interesting to test whether the depolarization induced by GTX on the enzyme-treated axons is qualitatively similar to that induced by GTX on the normal axon. During GTX-induced depolarization after treatment with BPN' for 35 min, application of 1, tm tetrodotoxin could repolarize the membrane to the level of the original resting potential, suggesting that depolarization observed after enzyme treatment is due to increase in the membrane permeability to Nat DISCUSSION The main result obtained by this experiment is that GTX is able to depolarize the membrane after proteolytic enzymes destroyed or inactivated the proteins which compose the Na channel gating. Therefore, it does not agree with the supposition that GTX exerts depolarizing action through the modification of the Na channel gating (SEYAMA and NARAHASHI, 1981). One of the possible mechanisms of GTX binding with the Na channel is that there are two sites of action; that is, the Na channel gating and the rest of the Na channel functioning as an ion filter. Since depolarization, induced by GTX after intracellular digestion with group A proteases, such as a-chymotrypsin, thermolysin, and N-protease, became almost a slarge as that in the normal condition, the site of action of GTX is most likely the protein of the Na channel having an ion filter rather than the channel gating. The finding by MATSUTANI et al. (1981) that GTX molecule possesses a hydrophobic part and may be able to gain access to the Na channel through the lipid layer offers another support for this notion. In order to depolarize the membrane having no channel gating, GTX may bind with the voltage sensor from which proteases had disconnected the Na channel gating, and GTX-attached voltage sensor can restore the response to the change in membrane potential. In this experiment, treatment of axonal membrane with group B proteases such as BPN', pronase, and Ste fradiae proteases tends to decrease the GTX-induced depolarization. However, treatment of axonal membrane with group A enzymes does not affect it. It has been shown that trypsin and St. fradiae trypsin (MORIHARA, 1974) exhibit a strict substrate-specificity toward basic amino acid Vol. 35, No. 3, 1985

8 408 I. SEYAMA et al. residues at carbonyl-side of the splitting point. BPN' has been reported to retain a substrate-specificity toward basic amino acids (MORIHARA,1974). Furthermore, pronase and Ste fradiae proteases contain trypsin-like enzymatic activity (MORIHARA et al., 1967). However, enzymes belonging to group A do not have such a substrate-specificity. Thus, it is reasonable to assume that splitting of the peptide bonds containing arginine residue and lysine residue is closely related with the decrease in sensitivity to GTX. Another noteworthy point is that arginil or lysil residue commits not only to binding of GTX with Na channel but also to inactivation process of Na channel (EATON et al., 1978; OXFORD et al., 1978). In taking into account that the inactivation mechanism of Na channel is eliminated at the late stage of intoxication by GTX, GTX may combine with the part of the Na channel mainly consisting of arginine or lysine residue and attachment of GTX may in turn disturb the functional connection between GTX binding site of the Na channel and the gating site of inactivation process. Recently, extensive efforts have been undertaken for reconstituting purified Na channel from various sources, such as Panulirus argus lobster walking legs (VILLEGAS and VILLEGAS, 1981), rat sarcolemma (WEIGELE and BARCHI, 1982), rat brain (TALVENHEIMO et al., 1982), and electric organ of Electrophorus electricus (ROSENBERG et a!., 1984). It has been reported that the reconstituted Na channel in the lipid bilayer lacks the gating function, indicating that in the process of purification of protein, the labile part of the Na channel is most likely taken off and/or inactivated. In order to activate the reconstituted Na channel, biological toxins must be used which open the Na channel. These biological toxins include grayanotoxin, batrachotoxin (KHoDoRov and REVENKO,1979), veratridine (ULBRICHT, 1969), and aconitine (SCHMIDT and SCHMITT, 1974). They exert common physiological actions on the Na channel; that is, the Na channel remains open at the membrane resting potential by shifting I -V curve to the hyperpolarizing direction. For example, confirmation of the Na channel has been made by the observation that veratridine increases 22Na trapped into the lipid vesicles which contain protein extracted from rat brain synaptosomes (TAMKUN et al., 1984). Since chemicals like veratridine can increase the permeability of the Na channel which lacks channel gating, the responsiveness of extracted protein to these chemicals implies that the concerned protein may be a part of the Na channel, that is, the Na channel having no channel gating. NOTE IN ADDENDUM After submitting this paper, we read with interest a paper published in Nature, 312: (1984) by Noda et al. They have predicted that the Na channel consists of four repeats, each of which contains six segments. In each segment, there is a unique segment (S4 according to their nomenclature), which is rich in arginine or lysine residue, facing only toward the intracellular space. It is in- Japanese Journal of Physiology

9 PROTEASES AND GRAYANOTOXIN 409 triguing to speculate that this segment corresponds to the place digested by trypsinlike endopeptidases, leading to the loss of GTX action. The authors wish to express their gratitude to Professor A. Watanabe, National Institute for Physiological Science, Okazaki, Japan, Professor A. Inaba, Marine Biological Station, Hiroshima University, Mukaijima, Onomichi, Japan, and Professor T. Iga, Department of Biology, Faculty of Science, Shimane University, Matsue, Japan for the supply of squid and constant encouragement. REFERENCES ARMSTRONG, C. M., BEZANILLA, F., and ROJAS, E. (1973) Destruction of sodium conductance inactivation in squid axons perfused with pronase. J. Gen. Physiol., 62: EATON, D. C., BRODWICK, M. S., OXFORD, G. S., and RUDY, B. (1978) Arginine-specific reagents remove sodium channel inactivation. Nature, 271: HODGKIN, A. L. and HUXLEY, A. F. (1952) A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. (Lond.),117: KHODOROV, B. I, and REVENKO, S. V. (1979) Further analysis of the mechanisms of action of batrachotoxin on the membrane of myelinated nerve. Neuroscience, 4: MATSUTANI, T., SEYAMA, I., NARAHASHI, T., and IWASA, J. (1981) Structure-activity relationship for grayanotoxin derivatives in frog skeletal muscle. J. Pharmacol. Exp. Ther., 217: MORIHARA, K. (1974) Comparative specificity of microbial proteinases. Adv. Enzymol., 41: MORIHARA, K., OKA, T., and TsuzuKI, H. (1967) Multiple proteolytic enzymes of Streptomyces fradiae: Production, isolation and preliminary characterization. Biochim. Biophys. Acta, 139: MORIHARA, K. and TSUZUKI, H. (1968) A trypsin-like protease from Streptomyces fradiae. Arch. Biochim. Biophys., 126: NAKAO, M. and SEYAMA, I. (1984) Effect of a-dihydro-grayanotoxin-ii on the electrical activity of the rabbit sino-atrial node. J. Physiol. (Loud.), 357: NARAHASHI, T. and SEYAMA, I. (1974) Mechanism of nerve membrane depolarization caused by grayanotoxin I. J. Physiol. (Loud.), 242: OXFORD, G. S., Wu, C. H., and NARAHASHI, T. (1978) Removal of sodium channel inactivation in squid giant axon by N-bromoacetamide. J. Gen. Physiol., 71: ROSENBERG, R. L., TOMIKO, S. A., and AGNEW, W. S. (1984) Reconstitution of neurotoxinmodulated ion transport by the voltage-regulated sodium channel isolated from the electroplax of Electrophorus electricus. Proc. Natl. Acad. Sci. U.S.A., 81: SCHMIDT, H. and SCHMITT, 0. (1974) Effect of aconitine on the sodium permeability of the node of ranvier. Pflugers. Arch., 349: SEYAMA, I. (1978) Effect of grayanotoxin I on SA node and right atrial myocardia of the rabbit. Am. J. Physiol., 235: C136-C142. SEYAMA, I, and NARAHASHI, T. (1981) Modulation of sodium channel of squid nerve membranes by grayanotoxin I. J. Pharmacol. Exp. Ther., 219: SEYAMA, I. and YOSHIOKA, Y. (1982) Modulation of the sensitivity of squid axon membrane to grayanotoxin (GTX) after the treatment with various endopeptidases. J. Physiol. Soc. Jpn., 44: 347. TALVENHEIMO, J. A., TAMKUN, M. M., and CATTERALL, W. A. (1982) Reconstitution of neurotoxin-stimulated sodium channel purified from rat brain. J. Biol. Chem., 257: TAMKUN, M. M., TALVENHEIMO, J. A., and CATTERALL, W. A. (1984) The sodium channel from Vol. 35, No. 3, 1985

10 410 I. SEYAMA et al. rat brain. Reconstitution of neurotoxin-activated ion flux and scorpion toxin binding from purified components. J. Biol. Chem., 259: TASAKI, I. and TAKENAKA, T. (1964) Effect of various potassium salts and proteases upon excitability of intracellularly perfused squid giant axon. Proc. Natl. Acad. Sci. U.S.A., 52: ULBRICHT, W. (1969) The effect of veratridine on excitable membranes of nerve and muscle. Ergeb. Physiol., 61: VILLEGAS, R. and VILLEGAS, G. M. (1981) Nerve sodium channel incorporation in vesicles. Annu. Rev. Biophys. Bioeng.,10: WEIGELE, J.B. and BARCHI, R.L. (1982) Functional reconstitution of the purified sodium channel protein from rat sarcolemma. Proc. Natl. Acad. Sci. U.S.A., 79: ZUSHI, S., MIYAGAWA, J., YAMAMOTO, M., KATAOKA, K., and SEYAMA, I. (1983) Effect of grayanotoxin on the frog neuromuscular junction. J. Pharmacol. Exp. Ther., 226: Japanese Journal of Physiology