affecting conductance

Transcription

1 Pro. Nati. Aad. Si. USA Vol. 81, pp , May 1984 Neurobiology Halothane shortens aetylholine reeptor hannel kinetis without affeting ondutane (general anesthetis/path lamp/myoytes/ell ulture) JAMES LECHLEITER AND RAPHAEL GRUENER* Department of Physiology, College of Mediine, University of Arizona, Tuson, AZ Communiated by S. Hagiwara, January 26, 1984 ASTRACT The extraellular path-lamp tehnique was used to examine how halothane, a general anestheti, affets the properties of single niotini aetylholine reeptor hannels of embryoni Xenopus skeletal musle ells grown in ulture. Under ontrol onditions, single-hannel events showed a bimodal distribution on the basis of urrent amplitudes. This distribution was maintained during exposure to halothane and its washout. In addition, the mean urrent value of the lowand high-amplitude hannels was unaffeted by the presene of the anestheti at linially relevant onentrations. In ontrast, halothane shortened the burst durations of both hannel types in a onentration-dependent manner. This shortening of burst durations may be an expression of the more rapid relaxation of the hannel protein to the nononduting state, possibly due to the disordering effet of the anestheti on membrane lipids in whih the reeptor protein is embedded. This funtional hange, in the behavior of the synapti reeptor, provides further diret information on the mode of ation of general anesthetis. Although the exat mehanism(s) of ation of general anesthetis (GAs) is still largely unknown, several hypotheses have been suggested to aount for the ability of a diverse lass of ompounds to produe loss of sensation and onsiousness. The lipid solubility theory was originally proposed beause of the remarkable orrelation between the oil/water partitioning oeffiient and the poteny of struturally unrelated agents (1-3). More reently, several variants of this theory have been formulated (4, 5). Of these, the membrane fluidity theory explains anesthesia by the disordering of the lipid bilayer resulting in inreased membrane fluidity (6-9). This effet may hange the behavior of membrane-bound proteins that are ritial to the funtioning of the neuron. Despite the diversity of proposed mehanisms (for reviews, see refs. 10 and 11), it is now generally aepted that the synapse itself is the most likely site of ation of GAs (12). This idea is derived from the finding that GAs blok synapti transmission at muh lower onentrations than required for blokade of the ation potential (13). Inasmuh as entral synapses are less aessible than peripheral ones, muh of the evidene for the ation of GAs on synapti transmission has been derived from detailed studies of the neuromusular juntion. Thus, several GAs have been shown to redue nerve-evoked postsynapti depolarizations at various vertebrate end plates (14-16). A loser examination of this derease in end plate potential amplitude, using the voltagelamp tehnique, has revealed the redution to be due to an inrease in the deay rate of miniature end plate urrents (16, 17). Using the noise-analysis tehnique, Landau et al. (18) surmised that end plate urrent redution, at the vertebrate The publiation osts of this artile were defrayed in part by page harge payment. This artile must therefore be hereby marked "advertisement" in aordane with 18 U.S.C solely to indiate this fat. neuromusular juntion, likely reflets the shortening of the time that aetylholine reeptor (AChoR) ion hannels spend in the open onfiguration. These findings are onsistent with the idea that GAs at on synapti proesses and demonstrate unequivoally the usefulness of the neuromusular juntion as a prototypi synapse. However, these data do not yield detailed and diret information onerning the speifi mehanism of GA ation beause the measurements are made on the entire ensemble of postsynapti reeptors. The reent development of the path-lamp, single-hannel tehnique (19, 20) provides diret measurements of single reeptor hannel ativity. We have used the path-lamp, single-hannel tehnique to examine how a volatile anestheti, halothane, affets the properties of single niotini AChoR hannels embedded in the membranes of embryoni Xenopus myoytes grown in ell ulture. We found that halothane, at linially relevant onentrations, shortens the single-hannel burst durations (i.e., the time spent by the hannel in the open, onduting onfiguration) without altering hannel ondutanes. MATERIALS AND METHODS Uninnervated musle ells from Xenopus laevis embryos (stages 19-22) were grown in ulture as desribed (21). Membrane pathes were sealed with eletrodes of 1-2 tim in internal diameter with resistanes of 3-15 Mfl. Eletrodes were filled with a high sodium solution whose omposition was 120 mm NaCl/1.6 mm KCl/1 mm CaCl2/8 mm Hepes HCl, ph 7.4, ontaining ,uM aetylholine (ACho). At this onentration of ACho, onurrent ativation of multiple hannels was rarely seen; when observed, these were omitted from the data base, failitating data analysis. Channel events were reorded from ell-attahed and ell-free pathes. The path-lamp iruit was built aording to the design of Hamill et al. (20). Single-hannel events were stored on FM magneti tape (RACAL Store SD4, Fullerton, CA). Analogue reords were digitized at 50-tkse intervals by a omputer (Dynabyte D8/4; Milpitas, CA). A downward defletion in the urrent trae, lasting for a minimum of 0.6 mse (iruit time onstant = 0.12 mse), and whih ourred within an adjustable window (usually 2-15 pa), was onsidered to be a resolvable event. When an event was deteted, the digitized reord and its trailing baseline (usually 2.5 mse) were stored on a floppy disk. Amplitudes were alulated from the mean value of all digitized points at the peak of an event omitting the two initial and final points to ross the seleted threshold. Eah event was then displayed and aepted or eliminated at the operator's disretion. The riterion for aeptane was an abrupt transition from baseline to a onstant urrent amplitude. The event was onsid- Abbreviations: GA, general anestheti; ACho, aetylholine; A- ChoR, ACho reeptor. *To whom reprint requests should be addressed. 2929

2 2930 Neurobiology: Lehleiter and Gruener ered losed when the urrent level remained at baseline for at least 1 mse. During experiments, performed at room temperature, ultures were ontinuously superfused (1-2 ml/min) with reording medium [high sodium solution: 120 mm NaCl/1.6 mm KCl/1 mm CaCl2/8 mm Hepes HCl, ph 7.4; for exised, ell-free pathes, 5 mm EGTA was added and CaCl2 was omitted. On oasion, a high potassium solution (70 mm K2SO4/4 mm KCl/8 mm Hepes HCl) was used]. Thymolfree halothane (Haloarbon Laboratories, Hakensak, NJ) 30 - I 0 L T Pro. NatL Aad SL USA 81 (1984) was added to the perfusion medium by bubbling through a vaporizer (Forreger, Smithtown, NY) at onentrations of 1-4% [perent of vaporized halothane mixed with air and bubbled through the reording medium reservoir; this onentration range orresponds to linially relevant doses (8) and to aqueous onentrations of mm, in the reording hamber, as measured by ultraviolet spetrosopy (22)]. After 4-7 min of exposure to halothane, and following a similar ontrol period, the anestheti was washed away and hannel events from the reovery period were usually reorded. It has reently been shown (23-25) that a given hannel may open repeatedly when ACho is bound to its reeptor. These openings are alled bursts sine they are interrupted by short, sometimes inomplete, losures, whose resolution depends, in part, on the time-onstant of the path-lamp iruit. To distinguish between a burst, representing the ativity of a single hannel, and the opening of two different hannels, Dionne and Leibowitz (25) assembled openingprobability histograms. They showed that, at the neuromusular juntion of the garter snake, two suessive openings appear as independent events (and hene likely to emanate from two separate hannels) when the losed times exeed about 2 mse. As a onservative definition of the burst, we 0, 4 ċ I1~ A o L0 3 - E z Fe1, II C a) -v 5- e. E 3- I0 - L LL Amplitude (pa) FIG. 1. Amplitude histograms of single-hannel events. Amplitude histograms, from a typial path, reveal a bimodal distribution. When a separation value of 4.6 pa is hosen, by eye, two hannel populations may be defined. These are alled low-amplitude and high-amplitude hannels. In this ell-free path, the ontrol mean amplitudes (-+SD) of the low-amplitude and high-amplitude hannels were 3.7 ± 0.5 pa (235 events) and 5.3 ± 0.3 pa (103 events), respetively (membrane potential was held at -120 mv; path superfused with high sodium solution). These values remained onstant at 3.6 ± 0.6 pa (134 events) and pa (103 events) during exposure of the path to 3% halothane. During the reovery period, the respetive amplitude values were 3.7 ± 0.6 pa and 5.8 ± 0.6 pa (104 and 76 events, respetively). (Insets) Representative reordings of single-hannel events before (A), during (), and after (C) exposure to halothane. eause these traes are of short durations, they do not reflet the relative frequeny of low- and high-amplitude hannel events. Calibration bars are 6 pa and 5 mse ~~~~~~~~~~~ urst Duration (mse) FIG. 2. Relationship between event amplitudes and their durations. (A) Data (from the path in Fig. 1) are distributed into two distint lusters on the basis of event amplitudes. During exposure to halothane (3%; ), the amplitude position of the lusters remains unaltered but the distribution of burst durations is shifted to shorter values. The initial distribution of burst durations is reovered (C) after washout of halothane. Events <0.6 mse were exluded due to the limitation of the iruit's response time.

3 Neurobiology: Lehleiter and Gruener report our data as burst durations when a series of hannel openings is interrupted by losed times shorter than 1 mse. A burst that is uninterrupted by losures is a speial ase of suh events and is ommonly referred to as the hannel open time. The Student t test was used to determine statistial signifiane. RESULTS When single-hannel events were assembled into histograms, on the basis of urrent amplitudes, a bimodal distribution was obtained (Fig. 1; see also refs. 26 and 27). The resultant two populations of events are designated low-amplitude and high-amplitude hannels. The low-amplitude hannel has a relatively long open time and a small amplitude, whereas the high-amplitude hannel has a shorter open time and a larger amplitude (see Fig. 1 Insets). The bimodel amplitude distribution and the mean amplitude values are maintained during exposure to halothane (Fig. 1) and the reovery period following the washout of halothane (Fig. 1C). In another set of experiments, the mean hannel amplitude from four separate ell-free pathes (membrane potentials were held at -120 mv and pathes were superfused with high potassium solution; see Materials and Methods for omposition) averaged 5.5 ± 0.1 pa for the low-amplitude and 8.2 ± 0.1 pa for high-amplitude hannels (mean ± SEM; n = 4 pathes) during the ontrol period. When exposed to 4% halothane, the mean amplitudes remained unaltered at 5.4 ± 0.2 pa and 8.2 ± 0.2 pa, respetively. Pro. NatL Aad Si. USA 81 (1984) i A o00 80 j 50 - The mean event amplitude, at a given membrane potential, and from a given path, remained onstant throughout the reording period (Fig. 1) despite some variability, in the absolute amplitudes, from path to path. The total number of events, however, dereased as a funtion of time and exposure to halothane (Fig. 1). When the urrent amplitude of eah event (data as in Fig. 1) is plotted against the orresponding single-hannel burst duration, the low-amplitude and high-amplitude hannels lustered about their respetive ontrol amplitude values (Fig. 2A). In the presene of halothane (Fig. 2) the distribution of burst durations was shifted to shorter values, and during the reovery period, the initial distribution of burst durations was almost ompletely reestablished (Fig. 2C). To quantitate the redution in single-hannel burst durations, integrated histograms were assembled and fit with a single exponential (see legend of Fig. 3). The mean burst duration was estimated from the inverse of the time onstant of the fit (nonlinear least squares method). A omparison of these histograms for the low-amplitude (Fig. 3 A and ) and high-amplitude hannels (Fig. 3 C and D), before and after exposure to 3% halothane, shows a marked derease in burst durations from a mean value of 3.7 mse to 2.7 mse and from 1.3 mse to 1.0 mse, respetively, for the two hannel populations. Table 1 summarizes the onentration-dependent derease in the mean burst duration for pooled data in whih both low- and high-amplitude hannel events were ombined. Even though these data (and those in Fig. 4 below) were obtained from ell-attahed and ell-free pathes, the rela a LI ;z 120 o00 D urst Duration (mse) FIG. 3. Integrated burst duration histograms. urst durations (same path as in Fig. 1) are grouped on the basis of their amplitudes. The lowamplitude and high-amplitude hannels are plotted separately during ontrol onditions (A and C) and during exposure to halothane ( and D), respetively. The bin width of the histograms is 100,use, and eah bin ontains the umulative values of all events with burst durations greater than the value of a partiular bin. The mean hannel burst duration was estimated from the inverse of the time onstant of the fitted exponential N(t) = NTexp(-at), using a nonlinear least squares method, where N(t) is the number of events at time t, NTiS the total number of events, and a is the time onstant (28, 29). An independent evaluation of burst durations, using-the maximum likelihood method (30), produed values that were numerially very lose and statistially idential with those obtained from the method used here. Thus, the possible error introdued by this method appears to be very small.

4 2932 Neurobiology: Lehleiter and Gruener Table 1. Effets of halothane on mean burst duration % halothane urst duration 1 95 ± 8* (3, 2, 1402) 2 67 ± 7t (1, 4, 1049) 3 52 ± 9t(1, 4, 647) 4 52 ± 14t (0, 4, 659) Values are % of ontrol and represent the mean ± SEM. Values in parentheses show numbers of ell-free pathes, ell-attahed pathes, and events, respetively. *Not signifiantly different from ontrol. tsignifiantly different from ontrol (P < 0.005). tive (% ontrol) redution, by halothane, of the burst durations was independent of path type or its membrane potential. When the effets of halothane are plotted separately for the low-amplitude and high-amplitude hannels (Fig. 4), a differential effet is observed in whih the redution in the burst durations of low-amplitude hannels (at 2% halothane) is signifiantly larger than that of the high-amplitude hannels. Conversely, at the highest halothane onentration (4%), low-amplitude hannels from two of three pathes showed a prominent slow omponent in the burst duration histogram. To obtain a best least squares fit to the low-amplitude hannel data, at this onentration of halothane, a double exponential fit was utilized and the burst duration was estimated from only the initial, fast time onstant (dashed irle in Fig. 4). The solid irle datum point, at 4% halothane, was derived from the single exponential fit and it ontains the subpopulation of hannel burst durations that were atually prolonged by the high onentration of halothane. DISCUSSION The independene of anestheti poteny on size, shape, and the hemial nature of GAs suggests that anestheti ation 0L Q 4-02p o L- :3 a 4., L- :3 m % Halothane FIG. 4. Conentration-dependent effet of halothane on hannel burst durations. Data from low-amplitude (o) and high-amplitude (v) hannels are plotted separately as a funtion of halothane onentration. The redutions in burst durations (both low-amplitude and high-amplitude hannels) are statistially different from their respetive ontrol values at 2% (P < 0.005), 3% (P < 0.05), and 4% (P < 0.005) halothane. At 2% halothane, the redution in the burst durations of low-amplitude hannels is signifiantly larger than that of the high-amplitude hannels (P < 0.05). urst durations are expressed as % of ontrol (bars represent the SEM). Numbers, in parentheses, next to data symbols show the number of pathes and the number of events used to alulate the respetive values for eah datum point. The asterisk (*) represents pooled low-amplitude hannel data from three pathes fit to a single exponential. In two of the three pathes, a prominent slow omponent in the burst duration histograms was observed. These data ould be better fit by a double exponential. When only the fast omponent is used, the data show a signifiant redution in burst duration (broken symbol) as ompared to ontrol values. Pro. NatL Aad SL USA 81 (1984) may depend on a nonspeifi interation with exitable membranes. GAs, regardless of their moleular struture, ause disordering of lipid membranes (refs. 31 and 32; but see ref. 33), and this may lead to "fluidization" of the membrane in whih reeptor proteins are embedded (8, 9). Suh a perturbation may be predited to permit a faster relaxation of the altered onformation of the reeptor protein indued by the agonist-reeptor interation. This would result in faster hannel losures and therefore in a redution of mean hannel open time (8). Consistent with this idea, Maleque et al. (34) have reported that ketamine aelerated the deay rate of miniature end plate urrents, at the frog neuromusular juntion, without affeting their ondutanes. Similarly, Haydon and Urban (35) have shown that halothane aelerates the urrent kinetis of the voltage-sensitive sodium hannel in the squid giant axon, resulting in a redution in inward urrent flow. In ontrast, Fernandez et al. (36) reported that hloroform redues the magnitude of sodium gating urrents without affeting their kinetis. Reently, Nioll and Madison (37) reported that several GAs produed neuronal membrane hyperpolarization, albeit at quite high onentrations. In ontrast, Lynh et al. (38) showed that volatile anesthetis do not affet the resting potential of ardia ells. Examination of inhibitory, "GAA-ergi" postsynapti reeptors revealed that when exposed to pentobarbital, mean hannel open times were atually prolonged (28). These diverse findings may reflet the ation of the different lasses of anesthetis or the use of different preparations. Our data show that halothane shortens the burst durations of single ion hannels, assoiated with the niotini AChoR, in a onentration-dependent manner. These findings are onsistent with previous reports on maroend plate urrents (8, 16) and noise-analysis data (18). It is of interest to note here that Gage and Hamill (8) reported that at high halothane onentrations, miniature end plate urrents had a rapid, early deay followed by a slower-than-normal deay. These data are onsistent with our observations on the prolongation of low-amplitude hannel burst durations under similar experimental onditions. In ontrast to its effets on singlehannel burst duration, halothane does not appear to alter the single-hannel ondutanes at any of the tested onentrations. Thus, for a given path, held at a onstant membrane potential, event amplitudes are not altered by the presene of halothane. The shortening of the burst durations of postsynapti reeptor hannels, whih we report here, would result in a derease in the amplitude and duration of postsynapti potentials, and if this effet were suffiiently large, ion urrents may fall below the threshold required for ativation of the postsynapti element and result in blokade of synapti transmission. During exposure to halothane, a derease in the frequeny of hannel events was observed (ompare Fig. 1 A-C). This "drop-out" phenomenon was not observed for the high-amplitude hannel events under prolonged (up to 15 min), ontrol onditions. ut under similar onditions, the low-amplitude hannel event frequeny dereased (within the first 5 min) and then leveled off. Sine the urrent amplitudes of both low-amplitude and high-amplitude hannels remain onstant throughout the reording period, and sine the mean low-amplitude and high-amplitude hannel burst durations return to ontrol values after washout of halothane, the drop-out phenomenon is unlikely to be due simply to membrane deterioration. One possible explanation for the dropout phenomenon is that some reeptors, loated in the membrane path, might migrate to the glass-membrane interfae and remain trapped there. Ion hannel events from suh reeptors would therefore be undetetable. The more rapid disappearane of low-amplitude hannel events, in our experiments, and the halothane-indued disappearane of highamplitude hannel events may reflet an inrease in the lat-

5 Neurobiology: Lehleiter and Gruener eral mobility of the respetive reeptors. More likely, however, is the possibility that halothane, in addition to shortening hannel burst durations, also hanges the affinity of ACho for its reeptor (39). This effet would result in inreased failure to ativate hannel events either beause of a redued probability of transmitter-reeptor interation or due to the involvement of reeptor desensitization. Sine we have used low agonist onentrations ( ,uM), and sine we rarely saw simultaneous multiple hannel openings, it is unlikely that the drop-out we observed was due to reeptor desensitization. Although our findings do not identify the site of ation of halothane as a GA, they provide useful information on the possible mode of ation of the drug in attenuating or bloking synapti transmission. The data we report here doument diretly that, in the presene of halothane, single AChoR hannels lose more rapidly. In addition to the redution in the burst duration, halothane may derease or inrease the number of brief losures (flikers) during a burst. In the former ase, our findings of redued burst durations demonstrate a net redution in the open state onfiguration. In the latter ase, the inreased probability of linking additional open times into a given burst duration would tend to produe an inrease rather than a derease in the value of the burst. Thus, we would be underestimating the effet of halothane on reduing the burst durations. Clearly, further work is needed to better understand the mode of ation of GAs. We are grateful to Steve Moffett for writing the omputer programs and to Drs. T. lank, P. rehm, R. rinton, and D. Kreulen for their onstrutive suggestions. Thymol-free halothane was a gift from Mr. Lou Ferstanding (Haloarbon Laboratories, Hakensak, NJ). This study was supported by a asi Researh Siene grant to R.G. from the Arizona Health Sienes Center and by National Institutes of Health Training Grant HL07249 to the Department of Physiology. 1. Meyer, H. H. (1901) Naunyn-Shmiedeberg's Arh. Exp. Pathol. Pharmakol. 46, Overton, E. (1901) Studien uber die Narkose (Fisher, Jena, G.D.R.). 3. Meyer, K. H. (1937) Trans. Faraday So. 33, Mullins, L. J. (1954) Chem. Rev. 54, Miller, K. W., Paton, W. D. M., Smith, R. A. & Smith, E.. (1973) Mol. Pharmaol. 9, Metalfe, J. C., Seeman, P. & urgen, A. S. V. (1968) Mol. Pharmaol. 4, Trudell, J. R., Hubbell, W. J. & Cohen, E. N. (1973) iohim. iophys. Ata 291, Gage, P. W. & Hamill, 0. (1975) Neurosi. Lett. 1, Lenaz, G., Curatola, G., Mazzanti, L., ertoli, E. & Pastuszko, A. (1979) J. Neurohem. 32, Roth, S. H. (1980) Can. Anaesth. So. J. 27, Pro. Natl. Aad Si. USA 81 (1984) Franks, N. P. & Lieb, W. R. (1982) Nature (London) 300, Judge, S. E. (1983) r. J. Anaesth. 55, Larrabee, M. G. & Posternak, J. M. (1952) J. Neurophysiol. 15, Gissen, A., Karis, J. H. & Nastuk, W. L. (1966) J. Am. Med. Asso. 197, Waud,. E. & Waud, D. R. (1975) Anesthesiology 42, Gage, P. W. & Hamill, 0. P. (1976) r. J. Pharmaol. 57, Torda, T. A. & Gage, P. W. (1977) r. J. Anaesth. 49, Landau, E. M., Rihter, J. & Cohen, S. (1979) Mol. Pharmaol. 16, Neher, E., Sakmann,. & Steinbah, J. H. (1978) Pflugers Arh. 375, Hamill, 0. P., Marty, A., Neher, E., Sakmann,. & Sigworth, F. J. (1981) Pflugers Arh. 391, Gruener, R. & Kidokoro, Y. (1982) Dev. iol. 91, lank, T. J. J. & Thompson, M. (1980) Anesth. Analg. 59, Sakmann,., Patlak, J. & Neher, E. (1980) Nature (London) 286, Colquhoun, D. & Sakmann,. (1981) Nature (London) 294, Dionne, V. E. & Leibowitz, M. (1982) iophys. J. 39, Clark, R.. & Adams, P. R. (1981) So. Neurosi. Abstr. 7, 838a. 27. rehm, P., Moody-Corbett, F. & Kidokoro, Y. (1983) So. Neurosi. Abstr. 9, 1180a. 28. Jakson, M.., Lear, H., Mathers, D. A. & arker, J. L. (1982) J. Neurosi. 2, rehm, P., Moody-Corbett, F. & Kullberg, R. (1984) J. Physiol. (London), in press. 30. Colquhoun, D. & Sigworth, F. J. (1983) in Single-Channel Reording, eds. Sakmann,. & Neher, E. (Plenum, New York), pp Pang, K., Chang, T. & Miller, K. W. (1979) Mol. Pharmaol. 15, Mountastle, D.., iltonen, R. L. & Halsey, M. J. (1978) Pro. Natl. Aad. Si. USA 75, Lieb, W. R., Kovalysik, M. & Mendelsohn, R. (1982) iohim. iophys. Ata 688, Maleque, M. A., Warnik, J. E. & Albuquerque, E. X. (1981) J. Pharmaol. Exp. Ther. 219, Haydon, D. A. & Urban,. W. (1983) J. Physiol. (London) 341, Fernandez, J. M., ezanilla, F. & Taylor, R. E. (1982) Nature (London) 297, Nioll, R. A. & Madison, D. V. (1982) Siene 217, Lynh, C., III, Vogel, S., Pratila, M. G. & Sperelakis, N. (1982) J. Pharmaol. Exp. Ther. 222, Young, A. P., rown, F. F., Halsey, M. J. & Sigman, D. S. (1978) Pro. Natl. Aad. Si. USA 75,