Pfl jers Archiv

Pfliigers Arch. 376, 151-157 0978)

EuropeanJournal of Physiology

9 by Springer-Verlag1978

Relaxation After a Voltage Step of Inhibitory Synaptic Current Elicited by Nerve Stimulation (Crayfish Neuromuscular Junction)* J. Dudel PhysiologischesInstitut der TechnischenUniversit/itMiinchen, Biedersteiner Strasse 29, D-8000 Mfinchen40, Federal Republic of Germany

Abstract. The clamped membrane potential of small crayfish muscle fibers was shifted in rapid steps between potentials of about - 60 and - 120 mV, and the clamp currents measured after de- and hyperpolarizing steps were averaged. In addition, the inhibitory nerve fiber was stimulated either synchronously or asynchronously with the averaging. Synchronous stimulation yielded the usual IPSCs, and asynchronous stimulation a steady state inhibitory current which relaxed to a new level after a voltage step. Fast relaxations were observed in all fibers. Their time constants "c= 15 to 20 ms at - 6 0 m V (10~ decreased on hyperpolarization and agreed with those of the decay of the IPSC at the respective potential. The relaxations could be described quantitatively by a model in which the synaptic current depends on membrane potential due to (1) the potential dependence of the life time "c of a synaptic channel, (2) to a constant channel permeability, and (3) to the potential difference from the equilibrium potential Ecl. In many fibers slow relaxations of the inhibitory synaptic current were observed also, with time constants ~ ' = 6 0 to 150ms at - 6 0 m V (10~ which decreased at more negative potentials. These slow relaxations correspond to a reported slow current noise component induced by GABA. Interpretations of this slow synaptic current component are discussed. Key words: Inhibitory postsynaptic currents Relaxation measurement - Synaptic channel kinetics.

Introduction When membrane current noise is elicited in crayfish muscle fibers by the application of the inhibitory * This investigationwas supported by the DeutscheForschungsgemeinschaft

synaptic transmitter, 7-aminobutyric acid (GABA), two noise components are observed in most fibers (Dudel et al., 1977). The fast component has an average time constant zf = 7.3 ms (at - 60 mV and 23 ~C) which is equivalent to the time constant of decay of the inhibitory postsynaptic current (IPSC) triggered by stimulation of the inhibitory nerve fiber. As in case of the noise component recorded at the vertebrate endplate (cf. Neher and Stevens, 1977), this time constant reflects the open time of the synaptic membrane channel. The slow noise component, which is not observed in all fibers, is characterized by an average time constant z s = 3 3 m s (at - 6 0 m V and 23~ This slow component is not visible in the falling phase of the IPSC. Dudel et al. (1977) suggested that this component was an extrasynaptic one. Current components due to activation of extrasynaptic receptors are well known in vertebrate muscle (Dreyer et al., 1976; Neher and Sakmann, 1976) and in glutamate receptors of locust muscle fibers (Cull-Candy, 1976). However, in crayfish muscle fibers extrasynaptic responses could not be elicited by the iontophoretic application of GABA (Dudel, unpublished). It seemed possible, therefore, that the slow noise component was of synaptic origin, but not visible in single IPSCs; i.e. it may only be activated if the transmitter is present for a longer period of time than during the few millisecond pulse of high transmitter concentration in a single IPSC. A transmitter release by the nerve terminal approaching a steady state can be achieved during long trains of high frequency stimulation. Studies of membrane noise are not practicable if the transmitter is applied intermittently. Therefore, in this study, the potential dependent time constants of the synaptic current components were determined by relaxation methods. Relaxation of membrane current after a voltage step has been shown to yield results equivalent to those obtained from noise analysis (Adams, 1975; Zingsheim and Neher, 1974;

0031-6768/78/0376/0151/$01.40

152

Pfltigers Arch. 376 (1978)

Neher and Sakmann, 1975; Sheridan and Lester, 1977; Colquhoun and Hawkes, 1977). A preliminary publication of part of these results has appeared (Dudel, 1978).

_

_

_

E 0

/

I

ii E01 ....

=t

Methods

Fibers of the opener muscle of the first walking leg of small crayfish (Orconectes limosus) were prepared as described by Dudel and Kuffler (1961) and mounted in a chamber with a superfusion system (Dudel, 1977a). The modified van Harreveld bathing solution contained in mmol/l: Na + 205, CI- 242, K + 5.4, Ca 2+ 13.5, Mg 2+ 2.6, trismaleate buffer 10; pH 7.5. It was cooled to 10~ Membrane potential was measured and current was applied through intracellular capillary microelectrodes (10 Mr2), the voltage electrode was filled with KC1, the current electrode with sodium citrate. The current was measured in the lead to the current electrode. The membrane potential was voltage clamped throughout the experiment, the potential alternating between two levels with a repetition period of 600 ms. The responses to a voltage step were summed in blocks of 64 and the average was displayed using a Nicolet 1074 instrument computer.The block averages were stored on digital tape for later processing, adding and subtracting such blocks. Since small differences in clamp current had to be evaluated, the clamp had to be very reliable. Only small muscle fibers (typical dimensions 800 gm length, 50 gm diameter) were used, which had an input resistance of at least 1 MQ.The quality of the space clamp was controlled in such preparations by injecting white noise through a third capillary at unfavorable positions. Even if the input membrane resistance would decrease to 0.5 Mr2, the clamp control was perfect for frequencies up to 100 Hz (Finger, 1978). Seventeen preparations were studied.

-

-

-L=O-

-

-

_(--0-

tE2,O 1

Fig. 1. Schematic representation of the voltage changes (uppermost diagrams) from E~ to E2, with equilibrium potential E o, and below the shapes of relaxation currents. In the middle row the time constant of relaxation at E~, vl, is larger than that at E2, %; in the bottom row zl is smaller than %. In the left column of diagrams, E 0 is more positive than E 1 and E2, and in the right hand column, E o lies between E~ and Ez. Further explanation in the text

Interpretation of Synaptic Relaxation Currents The binding of the transmitter substance to a synaptic receptor will cause a synaptic ion channel to change from a closed state to an open one_ If C denotes the number of closed channels, and O the number of open ones, and if simple kinetics are assumed, the equilibrium (1) will be established (cf. Adams, 1974, 1977; Neher and Sakmann, 1975): k~ C ~O.

(1)

described by the constant field theory (Goldman, 1943 ; Hodgkin and Katz, 1949), is constant. The value of the channel permeability 7ri = 9.2 . 1 0 - ~ c m S / s can be drived from the average value of 60mV given by Dudel et al. (1977). The potential dependent YE, Ec~ then is obtained by the relation E,-[CI],, _ _

The forward rate of the reaction, k~, increases with transmitter concentration. It is assumed to be independent of membrane potential. The lifetime of an open channel will be 1 =- k~-+ k b

(2)

and if the total number of channels is N,

o = N.k;.~.

(3)

The synaptic current flowing at a specific membrane potential E, is determined by the instantaneous number of open channels, 0,,, the unit conductance 7 of a single channel, and by the equilibrium potential E 0 of the synaptic current:

iE., O, = Om" YE, Eo (E, -- Eo).

(4)

For vertebrate endplates y usually is assumed to be a constant. In the inhibitory synaptic channel of crayfish muscle the conductance has been shown to depend on E, and on Eo, the latter shifting considerably with the holding clamp potential (Dudel, 1977b). Instead, the permeability of the inhibitory synaptic channel nl, as

YE, Ec~ = ~rl E _ E c l

F 2 eEc'F/RT--eE"F/RT RT e E"/RT- 1

(5)

The lifetime of the open channel, v, is reported to depend on the membrane potential in ali synapses studied so far, essentially due to the potential dependence of k b (Anderson and Stevens, 1973; Anderson et aI., 1976; Sheridan and Lester, 1977; Dudel, 1974, 1977b; DudeI et aI., 1977). If, in the presence of a certain transmitter concentration (k~ constant), the membrane potential is shifted from E~ to Ez, according to Eq. (3) the respective values of O 1 and O2 are proportional to z 1 and ~2, and

01/Oz = ~1/%-

(6)

According to Eq. (4), iEo o = is proportional to O,, and also to z.. If therefore, as is shown in the schematic representation of relaxation currents in Fig. 1, the potential is shifted from E 1 to E 2, the current changes instantaneously from the steady state value iE,,ol at E~ to a new value iE2,0,, which is determined by YE2.Eo'(E2-Eo), but at which O~ channels remain open. The average lifetime of these channels at E 2 is r2, the current therefore will relax with the time constant % towards the new steady state current ie~,02.02 will be larger or smaller than O~ as determined by the ratio z2/zl (Eq. 6), and accordingly the current

J. Dudel: Relaxation of Inhibitory Synaptic Current

153

200ms

200ms

E2 = -55mY_ {

E C[ = -96 mV--

I

I

~

IPSC50/s +2nA

--

i I

lm~

PSC

50/s

"control

"

20nA]

I I

control

x, [PSC

i IPSC

(E2,02

I

l

I

I

*

2nA--

current

--

+nA6

-- + 2

L=0

q

.... rE"~

t

2

Qsynchronous IPSC

synchronous IPSC

Fig. 2. The uppermost tracings show identical voltage clamp programs. Below that in the left hand column, trains of IPSCs are shown, which terminate while the membrane potential is at E 2 or El, with time constants of decay ~2 and rt, respectively. In the right hand column, the middle graph shows superimposed the clamp current in the presence and in absence of asynchronously averaged IPSCs. The difference between these clamp currents is the relaxation current in the bottom graph. The graphs in the left hand column show differences in averages of 64 records, the ones in the right hand column differences in averages of 4 x 64 records

amplitude will increase or decrease during relaxation9 Analogous considerations apply for the step back from E 2 to E~. The different amplitudes of iEo,O, can be determined from Eq. (4). The following equations can be derived from Eqs. (4) and (6): 9

/-

_ _

9

r

IE ,0 /IEI,O2--IE2, O1 1E2,G2=T1/~2

iEl,01 iE~,o~_

(iE~,O, iE,,O:) (iE~,O, iE2,0:) zl/z 2 - 1

between E 1 and E 2 (Dudel, 1977b). A relaxation current with the shape shown in the bottom right hand graph o f F @ t is to be expected in this study.

(7)

Results

(8)

In the experiment shown in Fig. 2, the membrane potential was clamped alternately at E~ = - 125 mV for 400ms, and at E 2 = - - 5 5 m V for 200ms. Under these conditions the equilibrium potential for inhibition adjusted to E c l = - 96mV. The clamp current was averaged without and with stimulation of the inhibitory nerve fiber, the difference of such current records being the synaptic current, the IPSC. In the left hand part of Fig. 2 synchronous IPSC trains are shown. They end while the membrane potential is held either at E 2 or Et, and from their final decay the respective time constants z 1 = 12ms and z 2 = 17ms can be extracted. The same potential dependence of z has been measured for the decay of single IPSCs (Dudel, 1977b). The right hand

(9)

Equations (8) and (9) are useful, in case the absolute current amplitudes are not known, and only the change in current during a relaxation can be determined. The schematic representations of relaxation currents in Fig. 1 are drawn according to Eqs. (4) and (6) to (9). The shape of the relaxation current is determined by the relative ampIitudes ofz 1 and %, and also by the position of the equilibrium potential E 0 relative to E~ and E 2. The case E o > E 2 > E~ and ~1 > z2 represents the conditions and the observed relaxations at the vertebrate endplate (Adams, 1974; Neher and Sakmann, 1975). At the crayfish inhibitory synapse, z decreases with hyperpolarization, and E o = Ec~ will adjust to a potential

154

Pflfigers Arch. 376 (1978)

time constant z 2 = 17ms. The voltage step back to - 1 2 5 mV yielded a relaxation from iE, o~ to iE~,o~ with the time constant z~ = J2 ms. The small amplitude of this relaxation is mainly due to the relatively negative value of Ecv If the potential was clamped longer to E 2 than to E~, Ec~ was more positive and the relaxation current at E 1 was larger (see Fig. 3), which allowed a better evaluation of this current component. Ecl could be determined as the reversal potential for short trains of IPSCs (Dudel, 1977b). From the different ieo,oomeasured in the relaxation current shown in Fig. 2, values for O~ or 02 can be calculated by means of Eqs. (4) and (5), since all the other parameters are known. The results are presented in Table 1. The pairs of values obtained for Oa and also

part of Fig. 2 shows the results of asynchronous stimulation, i.e. the stimulating pulses were not synchronized to the clamp pulses or to the averaging. The middle record presents the total clamp current for control conditions, and also that measured during inhibitory stimulation. The IPSCs increased the clamp current by only 10 ~ , the single IPSCs being smoothed out by asynchronous averaging. The difference of these average current traces is the IPSC relaxation current displayed as the bottom graph. The relaxation current is shaped like predicted in Fig. 1 for Ea < E o < E 2 and z 1 < z 2. Immediately after the voltage step from - 125 mV to - 55 mV, iE~o~ was smaller than the final current iE~,o~,indicating that Ox is smaller than 0 2. The current relaxed to iE~,o~ with the

200ms

Eel =-9g mV

400ms

E2=-55 mV~ ....

ECI=- 80mV - -

El =-115mV--'

ZOOms

/.OOms -[

t~

_ _J-L_~ I

Jl,_

L 9

ilil

~z

~ t=O

-InA m

tI

relaxation currents

Fig.3. Voltage clamp program and relaxation currents of asynchronously stimulated (50/s) IPSCs as in Fig. 2. The relaxations were each fitted by the sums of two exponentially decaying current components, a fast one with time constants zl = 12 ms and z2 = 17 ms, and a slow one with time constants z'~ = 50 ms and z~ = 90 ms, at the respective potentials E1 and E 2. The amplitudes of these current components are given in Table 2 and in the text. Each current trace is the difference of 3 x 64 single records

Table 1. Evaluation of the number of open channels O,, during the IPSCs of Fig. 2. The values of ie.,oo, and z, were measured at the different membrane potentials E,.~/E,,,E~, was calculated using Eq. (5) and the measured Ecl and O m was calculated using Eq. (4)

n

m

E,,

ie..o~

1 1

1 2

- 125mV -125mV

- 1.3nA - 1.8nA

5.6pS 5.6pS

0 z = 11100

2 2

1 2

-

+3.6nA +5.6nA

ll.5pS ll.5pS

O ~ = 7700 02 = 12000

55mV 55mV

?e.,Ect

O., 01 =

z. 8000

z 1 = 12ms z 2 = 17 ms

J. Dudel: Relaxation of Inhibitory Synaptic Current

155

Table2. Evaluation of the fast relaxation current components iEo,o,and of the open channels O. during the IPSCs of Fig. 3. The fast transient current component (iE.,o,--iE.,O), Ec, and ~. were measured. YE,,E~,was calculated using Eq. (5), iE.,o, using Eq. (7), and O. using Eq. (4)

n

E,

(iEo,o,

Ec~

7Eo,e~,

z~

iE,,,o.

O,

- 99 mV - 80 mV -99mV - 80 mV

5.5 pS 8.6 pS 9.7pS 13.3 pS

12ms 12 ms 17ms 17 ms

- 0.6hA - 2.0 nA + 3.3nA + 2.0 nA

7200 6900 10100 9600

-- iE.,02)

1 1

2 2

- 115 mV 115 mV - 65mV -- 65 mV

+ 0.3 nA + 0.8 nA - 1.0nA - 0.6 nA

for 0 z agree well. It should be noted that each value was determined on the basis of an independent measurement. In view of the fact that Ye~ is nearly twice as large as Ye,, it is obvious that consistent values of O1 and O2 can only be obtained if 7 depends on the membrane potential [Eq. (5)]. Furthermore, the values ofza and ~2 obtained for the relaxation current (Table 1) are the same as those derived from the decay of the synchronous IPSCs. Finally, as predicted by Eq. (6), the ratio 0 1 / 0 2 = 1.47 agrees with the ratio ~1/~2 = 1.42. These multiple agreements between theoretical predictions and independently determined experimental results seem to support the theoretical interpretation given in the preceding section. In the preparation of Fig. 2 only one time constant of relaxation was visible at each membrane potential. Fig. 3 shows an example of one of m a n y preparations in which additional slow relaxation components were obvious. In the two columns of the figure, only the timing of the voltage changes is different. In the left column the potential is held at E~ = - 115 mV for 400 ms and then shifted to E 2 = - 6 5 mV for 200 ms, while in the right hand one it stays at E 2 for 400 ms and at Et for 200 ms. Accordingly, Ecl adjusts to a level near to the time average of the potential, namely to - 99 mV and to - 8 0 mV, respectively (cf. Dudel, 1977b). This difference in Ecl has great consequences for the relaxation currents which are larger and more negative in the right hand record for both the fast and slow current components. The relaxation currents in Fig. 3 seem to be a mixture of components, and the absolute amplitudes of these components are not known. However, two transient exponential components with a pair of time constants ~, and ~, which are identical for the respective potential in both graphs, can be fitted to the relaxations. F r o m the amplitudes (iE.,ol-ieo,o) of these transients, and from the respective values of ~,, the absolute current amplitudes can be calculated by means of Eq. (7) to (9). This in turn allows an estimate of the respective O, values to be made [Eq. (4)]. Table 2 presents an evaluation for the fast current components. Although the amplitudes of the equivalent current components in both relaxations differ

greatly, the pairs of values obtained for O1 and 02 agree quite well. The larger current amplitudes in the right hand graph are mainly due to the higher level of 7 at E c l = - 80inV. The values of ~, and "E2 obtained from both relaxations agree with the time constants determined in this preparation from the decay of the IPSCs at the respective potential (not illustrated). The slow current components in Fig. 3 relax with approximate time constants of ~] = 50ms a n d z~ = 90ms at the membrane potentials E, and E2, respectively. Their shape, overshooting both for de- and hyperpolarization steps, agrees with the case va < ~'2 and E 1 < E2 < E0 in Fig. 1. The slow component therefore can formally be described as a synaptic current component with an equilibrium potential more positive t h a n - 60 mV, flowing through slow synaptic membrane channels to which the model of Eq. (1) to (4) is applicable. The steady state currents of this component can be determined from the amplitude of the slow relaxations by means of Eq. (7) to (9). For the left hand graph i'el,ol = - 0.5 nA and i'e2,o2--- 0.2nA, for the right hand graph i'El,ol = - 1.4 nA and i'E2,02 = -- 0.6 nA result. O] and O~ cannot be determined without further assumptions on the potential dependence of 7' and on E;. However, if the steady state currents of the slow component i'e,,o, are added to the respective values of iE,,o, of the fast component (Table 2), the respective sums agree quite well with the appropriate total steady state currents in both relaxation currents of Fig. 3 (i~1 = --1.1 nA, i~ = 3.2nA for the left hand graph, i~, = - 3.5 nA, i~ = + 1.8 nA for the right hand graph). This interpretation of the slow current component thus is fully consistent with the experimental results. However, it should also be possible to describe the slow component by alternative models, e.g. by a mechanism that slowly diminishes the absolute amplitude of a postsynaptic current pulse. The results presented in Figs. 2 and 3 are typical of those obtained in all 17 IPSC relaxation experiments. Six of these experiments were evaluated quantitatively as shown in detail above; and all showed a similar good agreement between values derived from independent measurements. In all experiments the time constants of the fast components of the relaxations agreed with

156

those of the decay of the IPSC at the same potential. The slow component varied in amplitude from not discernible to relatively large as is shown in the example of Fig. 3. The time constants of the slow component were in the range of 40 to 150ms, increasing on depolarization. In many experiments relaxation currents elicited by superfused GABA were also measured. These GABA relaxation currents showed fast and slow current components with the same time constants as the IPSC relaxations. Further details will be reported elsewhere. Discussion

The average lifetimes of inhibitory synaptic channels as determined by noise analysis were zf--7.3 ms and z s = 33 ms at - 60 mV and 23 ~C. The relaxation currents in this study showed time constants in the range of z = 15 to 20 ms and z' = 60 to 150 ms at about the same potential, but at 10~ Since the Q~o of z, determined from the decay of IPSCs, is 1.9 to 2.5 (Dudel, 1977b), the time constants found here agree very well with those calculated from the GABA noise spectrum in the same preparation. In addition to in,formation also obtainable from noise analysis, the present study has yielded new data on the steady state currents and on their dependence on the membrane potential and Ecl. A large number of independent measurements could be fitted by the model expressed in Eqs. (1) to (5), which assumes essentially that the permeability of a synaptic channel is constant, and that the potential dependence of the synaptic current is due to two factors; (I) the potential dependence of the lifetime z of a synaptic channel, and (II) the potential dependence of the synaptic conductance 7 which follows from a constant channel permeability ~zi [Eq. (5)]. As stated above, the essential features of this model, with the exception of the constant ~, were developed for the nicotinic vertebrate acetylcholine synapses. The applicability of this model to inhibitory GABA synapses of crayfish may indicate that this model describes very general characteristics of the mechanism of chemical synaptic transmission. The slow noise component observed in many preparations can be reproduced in relaxations of IPSCs, and this component therefore should be a synaptic one. The slow current component may therefore represent synaptic channels which open on activation by the transmitter and whose open lifetime depends on the membrane potential as is also the case for the fast channels. The slow synaptic channels would differ from the fast ones, apart from their longer open lifetime, only in respect to the ions flowing through the channel. While the fast inhibitory channels only allow C1 ions to pass, the slow ones having an equilibrium potential

Pflfigers Arch. 376 (1978)

more positive than Ecl or EK, should have an appreciable permeability for Na +- and/or Ca2+-ions. While this interpretation is in full agreement with the experimental results, it should be substantiated by further evidence. Alternative explanations for the slow current component certainly could be proposed. Two mechanisms that tend slowly to reduce synaptic responses may be discussed. One is desensitization, a decrease of the synaptic response during the prolonged presence of transmitter. However, the inhibitory GABA receptors of the opener muscle are known to partially desensitize only in presence of very high GABA concentrations (Feltz, 1971; Dudel and Hatt, 1976). In addition, the IPSCs were recorded under steady state conditions in this study. The records used for Fig. 3, e.g., were taken in 6 blocks of 64 single voltage step cycles each, blocks with and without stimulation alternating. The consecutive blocks of records showed no trend in the amplitudes of the relaxation currents, therefore there is no indication of a progressive desensitization. However, some degree of steady state desensitization could build up during a train of IPSCs. The open channels then would also be in equilibrium with desensitized (D) ones: C,-~-O ~-D. With appropriate time constants of desensitization, slow relaxation components may result. Another mechanism for the generation of the slow synaptic current component could be based on shifts of Eci after a voltage step. Chloride current flowing after a voltage step can change the intracellular CI-concentration in the small preparations, and Ecl will shift in the same direction as the membrane potential (Dudel, 1977b). Such shifts of EC1 will reduce the synaptic current, but the observed shifts proceed much too slowly to account for the slow synaptic current component. The shortest time constant of the adjustment of EC1 seen after a voltage step is about 1 min (Dudel" 1977b, Fig. 4). Further assumptions, e.g. restricted spaces, could be made, thus explaining more rapid local transient changes in [C1-]i. However, changes in [Cl-]i after shifts in membrane potential are easily observed in all crayfish muscle fibers, while the slow synaptic current component is variable in relative amplitude and not visible in some fibers (see Fig. 2). It seems improbable that some fibers should have "restricted intracellular spaces" and others not. Acknowledgements'. I want to thank Drs. D. A. Mathers and R. Rtidel for their comments on the manuscript, Miss I. Horstmann for technical assistance and Mrs. L. Bauer and M. Griessl for secretarial help. References

Adams, P. R.: Kinetics of agonist conductance changes during hyperpolarization at frog endplates. Br. J. Pharmacol. 53, 308-310 (1975)

J. Dudeh Relaxation of Inhibitory Synaptic Current Adams, P. R. : Relaxation experiments using bath-applied suberyldicholine. J. Physiol. (Lond.) 268, 271-289 (1977) Anderson, C. R., Cull-Candy, S. G., Miledi, R.: Glutamate and quisqualate noise in voltage-clamped Iocust muscle fibres. Nature 261, 151 -- 153 (1976) Anderson, C. R., Stevens, C. F. : Voltage clamp analysis of acetylcholine produced end-plate current fluctuations at frog neuromuscular junction. J. Physiol. (Lond.) 235, 655-691 (1973) Colquhoun, D., Hawkes, A. G.: Relaxation and fluctuations of membrane currents that flow through drug-operated channels. Proc. R. Soc. Lond. [Biol.] 199, 231-262 (1977) Cull-Candy, S. G.: Two types of extrajunctional L-glutamate receptors in locust muscle fibres. J. Physiol. (Lond.) 255, 449 - 464 (1976) Dudel, J. : Nonlinear voltage dependence of excitatory synaptic current in crayfish muscle. Pflfigers Arch. 352, 227-241 (1974) Dudel, J. : Dose-response curve of glutamate applied by superfusion to crayfish muscle synapses. Pflfigers Arch. 368, 4 9 - 54 (1977a) Dudel, J.: Voltage dependence of amplitude and time course of inhibitory synaptic current in crayfish muscle. Pfliigers Arch. 371, 167-174 (1977b) Dudel, J. : Relaxation of inhibitory synaptic current after a step in membrane potential. Pflfigers Arch. 373, R63/224 (1978) Dudel, J., Finger, W., Stettmeier, H. : GABA induced membrane current noise and the time course of the inhibitory synaptic current in crayfish muscle. Neurosci. Letters 6, 203- 208 (1977) Dudel, J., Hart, H. : Four types of GABA receptors in crayfish leg muscles characterized by desensitization and specific antagonist. Pfliigers Arch. 364, 217-222 (1976) Dudel, J., Kuffier, S. W. : The quantal nature of transmission and spontaneous miniature potentials at the crayfish neuromuscular junction. J. Physiol. (Lond.) 155, 514-529 (1961)

157 Dreyer, F., Mfitler, K.-D., Peper, K., Sterz, R. : The M. omohyoideus of the mouse as a convenient mammalian muscle preparation. Pflfigers Arch. 367, 115 - 122 (1976) Feltz, A. : Competitive interaction of fl-guanidino propionic acid and " 7-aminobutyric acid on the muscle fibre of the crayfish. J. Physiol. (Lond.) 216, 391-401 (1971) Finger, W. : GABA induced conductance fluctuations in synaptic channels of crayfish muscle. Pflfigers Arch. 373, R 63/223 (1978) Goldman, D. E. : Potential, impedance and rectification in membranes. J. Gen. Physiol. 27, 37 (1943) Hodgkin, A. L., Katz, B. : The effect of sodium ions on the electrical activity of the giant axon of the squid. J. Physiol. (Lond.) 108, 37 (1949) Neher, E., Sakmann, B. : Voltage-dependence of drug-induced conductance in frog neuromuscular junction. Proc. Natl. Acad. Sci. U.S.A. 72, 2140-2144 (1975) Neher, E., Sakmann, B. : Noise analysis of drug induced voltage clamp currents in denervated frog muscle fibres. J. Physiol. (Lond.) 258, 705-729 (1976) Neher, E., Stevens, C. F.: Conductance fluctuations and ionic pores in membranes. Ann. Rev. Biophys. Bioeng. 6, 345-381 (1977) Sheridan, R. E., Lester, H. A.: A study of neuralty evoked postsynaptic currents and of voltage-jump relaxations. J. Gen. Physiol. 70, 187-219 (1977) Zingsheim, H. P., Neher, E, : The equivalence of fluctuation analysis and chemical relaxation measurements : a kinetic study of ion pore formation in thin lipid membranes. Biophys. Chem. 2, 197-207 (1974)

Received May 26, 1978