- Email: [email protected]
High rates of excitatory miniature currents in crayfish claw opener muscle evoked by high concentrations of γ-aminobutyric acid (GABA) in normal and Ca2+-deficient superfusions
Neuroscience Letters, 47 (1984) 251-256
251
Elsevier Scientific Publishers Ireland Ltd.
NSL 02756
HIGH RATES OF EXCITATORY MINIATURE CURRENTS IN CRAYFISH CLAW OPENER MUSCLE EVOKED BY HIGH CONCENTRATIONS OF T-AMINOBUTYRIC ACID (GABA) IN NORMAL AND Ca2+-DEFICIENT SUPERFUSIONS
WOLFGANG FINGER
Physiologisches Institut der Technischen Universitiit Miinchen, Biedersteiner Str. 29, D-8000 Miinchen 40 (F.R.G.) (Received February 27th, 1984; Revised version received March 14th, 1984; Accepted March 27th, 1984)
Key words: crayfish claw opener muscle - GABA-evoked transmitter release - excitatory miniature currents - calcium dependence o f release - presynaptic store of transmitter - depletion - noise analysis
High concentrations (0.5 tool/l) of the neutral amino acid G A B A were used to evoke release of transmitter quanta from excitatory terminals at voltage clamped crayfish muscle fibres in normal and Ca 2 + -deficient superfusions. An experiment in which the release of transmitter quanta proceeded at high rates in both normal and Ca2+-deficient superfusion was analyzed in detail indicating a Ca 2+-independent mechanism of release. In the normal superfusion, on application of G A B A , the release rates fi increased within a few seconds up to about 6000 q u a n t a / s and thereafter declined exponentially with a time constant rq = 18.5 s, most likely due to depletion of a readily releasable store of transmitter in the excitatory nerve terminals comprising at least 110,000 quanta per muscle fibre. Assuming that about 1900 excitatory synapses exist per muscle fibre [9], it results that about 58 q u a n t a can be associated with each synapse in agreement with morphological data [15] which show that between 47-117 vesicles exist in a single glutamatergic synapse of crayfish.
In previous publications [3,4], it was shown that the neutral amino acid glycine, when applied at high concentrations (0.1-0.5 mol/l) to superfusion, activates release of inhibitory transmitter quanta at inhibitory synapses in the opener muscle of the first walking leg of crayfish. This effect of glycine could not be abolished by reducing the Ca 2 ÷ -concentration in the bathing medium. In some extreme reactions with glycine, inhibitory nerve terminals were depleted of readily releasable transmitter with a time constant of between 10-20 s, and from these experiments a store of transmitter was estimated for the inhibitory nerve terminals at a single muscle fibre comprising 160,000-200,000 quanta. Furthermore, it was observed [3,5,8] that by adding high concentrations of ~/-aminobutyric acid (GABA) (0.1-0.5 mol/l) to the superfusion, release of excitatory transmitter quanta could be enhanced drastically. In the present paper, it was investigated whether the GABA-induced release of excitatory transmitter quanta could deplete the excitatory terminals of their readily 0304-3940/84/$ 03.00 © 1984 Elsevier Scientific Publishers Ireland Ltd.
252
releasable transmitter stores and whether the mechanism by which GABA activates release depends on extracellular Ca 2 +. The experiments were conducted at small fibres (length ~ 0.6 mm) of the crayfish claw opener muscle (Austropotamobius torrentium). Two glass microelectrodes were inserted into a muscle fibre and the membrane potential was clamped to a constant level. The preparation was superfused with a modified van Harreveld solution which contained (in mmol/l): Na + 205, C1- 242, K + 2.7, Cs + 2.7, Ca 2+ 13.5, Mg 2+ 2.6, trismaleate buffer 10; pH 7.6. Ca z + -deficient superfusions were made by replacing isotonically CaCI2 by NaCI. No EDTA was added to the Ca z + -deficient bathing solutions; therefore, the effective extracellular Ca z+-concentration probably was __<0.1 mmol/1 [10]. Before G A B A was applied continuously to the superfusion, it was applied several times for about 1 s in order to desensitize inhibitory postsynaptic receptors and to prevent a contribution of current noise from inhibitory postsynaptic currents activated by G A B A to the current noise elicited by GABA-induced spontaneous excitatory postsynaptic currents (sEPSCs) [8]. Further experimental details are described in preceding publications [2,6,16]. Figs. I-3 show an experiment in which application of 0.5 mol/1 G A B A evoked intensive, spontaneous release of excitatory transmitter quanta from excitatory nerve terminals at a muscle fibre voltage-clamped at E = - 60 mV membrane potential (T = 20°C). Fig. 1 represents current records obtained in the absence (control) and the presence of GABA. Application of G A B A resulted in a drastic increase in intensity of the current fluctuations, most likely due to superposition of sEPSCs occurring randomly at high rates [3,4,7]. The signal in Fig. 1 starts at the maximal rate of quantal release elicited by G A B A which was reached within a few seconds after the addition of G A B A to the bathing fluid. The average power spectrum calculated from the first 20 s of the GABA-evoked noise current is shown in Fig. 2A. The spectrum was fitted by a Lorentz curve (see legend of Fig. 2) with a cut-off frequency f~=85 Hz. This correlates to an average decay time constant of the sEPSCs:
oo, L!I
O.SooL,I ABA
claw open~ E=-6OmV
i.i ,51 :,~ :ix ~.~.:~;~:~. i ;:'~... ,~;:~.~
T=
20°C
5s
Fig. 1. High rates of spontaneous excitatory postsynaptic currents (sEPSCs) in a voltage-clamped fibre (E = - 6 0 mV, T = 20°C) of the crayfish claw opener muscle evoked by bath application of 0.5 m o l / l G A B A in a normal superfusion (13.5 m m o l / l Ca 2 ÷ ). Noise currents in the absence (control) and the presence of 0.5 m o l / l GABA are presented. The current fluctuations elicited by G A B A attained a maximum within a few seconds ( ~ 5 s), and the signal shown starts at about this ma xi mum.
253
,
A
S If)
r
A2s
B
Slft
~ ~,•90 Hz
~ fc,85 l-lx •
~o~o
¢a~>
"
,
,,',
~ -
OABA,IowC a 2 " - ~ , <3~>• 3t,00quantu/s
= /,000
10 .22
'1~
f I
•
•
,j....|
. . . . . . . .
1
10
l
. . . . . . . .
100
l
IO00Hz
f I
. . . . . . .
1
.l
. . . . . .
10
*.I
100
. . . . .
..,l
1000 HZ
Fig. 2. Power spectrum of the noise current evoked by GABA in the normal and the Ca 2 ÷-deficient superfusion. A: an average power spectrum was calculated for the first 20 s of the signal shown in Fig. 1. The power spectral densities S(f) were fitted by a Lorentz curve: S(f) = S(O)/[1 + (f/f¢)2], with a cut-off frequency fc =85 Hz. An average release rate ( r ) _--4000 sEPSCs/s resulted by means of eqn. 1. B: the experiment was repeated by application of GABA dissolved in the superfusion containing the low Ca 2 ÷-concentration. GABA again evoked a signal of similar intensity as before in the normal superfusion. A Lorentz curve with fc = 90 Hz was fitted to the average power spectrum calculated for a 20-s time interval, and an average rate of release ( r ) = 3 4 0 0 sEPSCs/s was estimated by eqn. 1.
fi "l~q-- 18.5 S, r =-0,98/4
IkHzl ~, -- 110000 quQnto ( t - - ~ ) 6
0
6
12
18
24
30
36
Isl
t
Fig. 3. Time course of the release rate r. According to eqn. 2 release rates, r , for shorter time intervals (1.5 s) were calculated from the GABA-evoked current fluctuations shown in Fig. 1 and plotted as a function of time. The maximum release rate was about rio=6000 sEPSCs/s and fi decreased exponentially with time (rq = 18.5 s, r = - 0.984). The number of quanta released in total was estimated by g = rq • rio--110,000 quanta• This does not include the number of quanta released from the time GABA was applied until the release was maximal ( = 5 s).
254 r(sEPSCs) = 1/(2~rfc)= 1.9 ms. The power spectral density S(O) at small frequencies (f--*0) is proportional to the average frequency (fi) by which the sEPSCs in Fig. 1 occurred during the 20-s time interval [4,7]: (fi)=
--
211+(C.V.) 21
(1)
is the average amplitude of the sEPSCs and C.V. is the coefficient of variation of the amplitudes of sEPSCs around their mean value ft. In muscle fibres of the claw opener from small crayfish ~ = - 0 . 6 6 nA and (C.V.)2 = 0.32 were found at a membrane potential E = - 60 mV in the presence of 0.5 mol/1 G A B A (Finger, W. and Martin, C., unpublished observations). The value of ~ in the presence of the high concentration of G A B A is smaller by a factor of about 1.8 than in the absence of G A B A (Finger, W. and Martin, C., unpublished observation), most likely due to an inhibitory effect of high concentrations of neutral amino acids on the excitatory receptor-channel complex in the postsynaptic m e m b r a n e [3]. Thus, with fc = 85 Hz and S ( O ) = l . 6 x 10 -z° AZs, an average frequency (fi)-_-4000 sEPSCs/s can be calculated by eqn. 1 for the release of excitatory transmitter quanta. Subsequent to this experiment, G A B A was removed from the bathing solution and the preparation was superfused for 15 min with the Ca 2 +-deficient van Harreveld solution; after that G A B A was applied in Ca 2 + -deficient solution. This again gave rise to a signal similar to that obtained during the first application of G A B A (not illustrated), in spite of the considerable reduction of the extracellular Ca / +-concentration. The average power spectrum for the signal recorded in the low CaZ+-solution is represented in Fig. 2B, and according to eqn. 1, ( f i ) = 3 4 0 0 sEPSCs/s was found for the average release rate during the first 20 s following the m a x i m u m of release. However, it was found previously that reducing the extracellular Ca 2 +-cOncentration has a profound diminishing effect on the amplitudes of nerve evoked EPSCs [1]. Furthermore, in low [Ca 2 +] a reduction in quantum current amplitude was found [1], which could mean that in low [Ca z+ ] the average release rate was even larger than 3400 sEPSCs/s. Therefore, from the present experiment it must be concluded that the mechanism by which G A B A enhances release of excitatory transmitter quanta does not depend on the external Ca z +-concentration, at least down to concentrations as low as 0.1 tool/1. As can be seen from Fig. 1, the intensity, respectively the variance, of the current fluctuations, which also is proportional to the release rate of the transmitter quanta, decreased with time indicating that the release rate decreased continuously. To evaluate this process in more detail, the variance oa o f the current fluctuations was determined for small time intervals (1.5 s), and by means of eqn. 2 [4,7]: fi =
4~fca 2 ~.2[1 + (C.V.) 2]
(2)
the release rates fi during these time intervals were estimated and plotted as a func-
255 tion of time (Fig. 3). fi declined exponentially with a time constant rq = 18.5 s, the maximum rate of release being rio = 6000 sEPSCs/s. The number of quanta released in total ( t - ~ ) in the reaction with G A B A can be estimated with the time integral over fi(t): ~-- Io fi(t) dt = fi0rq~- 111,000 quanta. Summarizing, it may be assumed that in the experiment of Fig. 1 on application of 0.5 mol/1 GABA, a readily releasable store of transmitter comprising at least 111,000 quanta was depleted with a time constant of rq = 18.5 s. Thus, like glycine at inhibitory synapses [3,4], G A B A may serve to evoke high rates of quantal release and to deplete excitatory terminals o f their transmitter. Taking into consideration that about 1900 releasing sites exist on a single muscle fibre [9], it seems that approximately 58 quanta are stored at a single synapse ready for release. This conclusion is in agreement with the observations of Nakajima and Reese [15] who observed about 42-117 vesicles at single excitatory synapses in crayfish abdominal muscle. Therefore, it may well be that in the crayfish opener muscle one quantum of transmitter is represented by one vesicle. For the mechanism by which glycine activates release of inhibitory transmitter quanta, we have suggested a Na +-dependent uptake process for glycine [4] which increases the intracellular amino acid concentration in the presynaptic terminal. It remains to be investigated whether for the GABA-evoked release a similar mechanism may be assumed. Uptake of the amino acids into the presynaptic terminal also could be electrogenic giving rise to depolarization of the presynaptic membrane and thus to activation of release of transmitter [12,14]. However, if this were true, a Ca2+-dependence of the release process should be expected [1,11], which was not observed in the present and earlier experiments [4]. Possibly, an increase o f the Na +-concentration inside the cytoplasm of the terminals as a result of activation of the uptake process for amino acids might account for the Ca 2 + -independence of a depolarization evoked transmitter release, as was observed for the release of catecholamines from the adrenal medulla [13]. I would like to thank Prof. Dr. Dudel for reading the manuscript, and Mrs. E. K6ster and Mrs. M. Griessl for technical and secretarial help. This investigation was supported by the Deutsche Forschungsgemeinschaft, Project Fi 305/1-2.
l Dudel, J., The effect of reduced calcium on quantal unit current and release at the crayfish neuromuscular junction, Pflfigers Arch., Europ. J. Physiol., 391 (1981) 35-40. 2 Dudel, J., Finger, W. and Stettmeier, H., Inhibitory synaptic channels activated by y-aminobutyric acid (GABA) in crayfish muscle, Pflfigers Arch., Europ. J. Physiol., 387 (1980) 143-151. 3 Finger, W., Enhanced release of inhibitory and excitatory transmitter quanta in the crayfish neuromuscular junction by glycine and GABA, Neurosci. Lett., 34 (1982) 33-38. 4 Finger, W., Effects of glycine on the crayfish neuromuscular junction II. Release of inhibitory transmitter activated by glycine, Pflfigers Arch., Europ. J. Physiol., 397 (1983) 128-134. 5 Finger, W., Glutamate-operated postsynaptic currents in crayfish claw opener muscle, Neurosci. Lett., 36 (1983) 163-168. 6 Finger, W. and Stettmeier, H., Efficacy of the two-microelectrode voltage clamp technique in crayfish muscle, Pfliigers Arch., Europ. J. Physiol., 387 (1980) 133-141.
256 7 Finger, W. and Stettmeier, H., Analysis of miniature spontaneous inhibitory postsynaptic currents (slPSCs) from current noise in crayfish opener muscle, Pfliigers Arch., Europ. J. Physiol., 392 (1981) 157-162. 8 Finger, W. and Stettmeier, H., Postsynaptic actions of ethanol and methanol in crayfish neuromuscular junctions, Pfltigers Arch., Europ. J. Physiol., 400 (1984) 113-120. 9 Florey, E. and Cahill, M.A., The innervation pattern of crustacean skeletal muscle. Morphometry and ultrastructure of terminals and synapses, Cell Tiss. Res., 224 (1982) 527-541. 10 Hubbard, J.l., Jones, S.F. and Landau, E.M., On the mechanism by which calcium and magnesium affect the spontaneous release of transmitter from mammalian motor nerve terminals, J. Physiol. (Lond.), 194 (1968) 355-380. 11 Katz, B. and Miledi, R., Release of acetylcholine from a nerve terminal by electric pulses of variable strength and duration, Nature (Lond.), 207 (1965) 1097-1098. 12 Kehoe, J., Electrogenic effects of neutral amino acids on neurons of Aplysia californica, Cold Spring Harb. Symp. quant. Biol., 40 (1976) 145-155. 13 Lastowecka, A. and Trifar6, J.M., The effect of sodium and calcium ions on the release of catecholamines from the adrenal medulla: sodium deprivation induces release by exocytosis in the absence of extracellular calcium, J. Physiol. (Lond.), 236 (1974) 681-705. 14 Martin, D.L., Carrier-mediated transport and removal of GABA from synaptic regions. In Roberts, E., Chase, T.N. and Tower, D.B. (Eds.), GABA in Nervous System Function, Raven Press, New York, 1976, pp. 347-386. 15 Nakajima, Y. and Reese, T.S., Inhibitory and excitatory synapses in crayfish stretch receptor organs studied with direct rapid-freezing and freeze-substitution, J. comp. Neurol., 213 (1983) 66-73. 16 Stettmeier, H., Finger, W. and Dudel, J., Glutamate activated postsynaptic channels in crayfish muscle investigated by noise analysis, Pfltigers Arch., Europ. J. Physiol., 397 (1983) 13-19.
251
Elsevier Scientific Publishers Ireland Ltd.
NSL 02756
HIGH RATES OF EXCITATORY MINIATURE CURRENTS IN CRAYFISH CLAW OPENER MUSCLE EVOKED BY HIGH CONCENTRATIONS OF T-AMINOBUTYRIC ACID (GABA) IN NORMAL AND Ca2+-DEFICIENT SUPERFUSIONS
WOLFGANG FINGER
Physiologisches Institut der Technischen Universitiit Miinchen, Biedersteiner Str. 29, D-8000 Miinchen 40 (F.R.G.) (Received February 27th, 1984; Revised version received March 14th, 1984; Accepted March 27th, 1984)
Key words: crayfish claw opener muscle - GABA-evoked transmitter release - excitatory miniature currents - calcium dependence o f release - presynaptic store of transmitter - depletion - noise analysis
High concentrations (0.5 tool/l) of the neutral amino acid G A B A were used to evoke release of transmitter quanta from excitatory terminals at voltage clamped crayfish muscle fibres in normal and Ca 2 + -deficient superfusions. An experiment in which the release of transmitter quanta proceeded at high rates in both normal and Ca2+-deficient superfusion was analyzed in detail indicating a Ca 2+-independent mechanism of release. In the normal superfusion, on application of G A B A , the release rates fi increased within a few seconds up to about 6000 q u a n t a / s and thereafter declined exponentially with a time constant rq = 18.5 s, most likely due to depletion of a readily releasable store of transmitter in the excitatory nerve terminals comprising at least 110,000 quanta per muscle fibre. Assuming that about 1900 excitatory synapses exist per muscle fibre [9], it results that about 58 q u a n t a can be associated with each synapse in agreement with morphological data [15] which show that between 47-117 vesicles exist in a single glutamatergic synapse of crayfish.
In previous publications [3,4], it was shown that the neutral amino acid glycine, when applied at high concentrations (0.1-0.5 mol/l) to superfusion, activates release of inhibitory transmitter quanta at inhibitory synapses in the opener muscle of the first walking leg of crayfish. This effect of glycine could not be abolished by reducing the Ca 2 ÷ -concentration in the bathing medium. In some extreme reactions with glycine, inhibitory nerve terminals were depleted of readily releasable transmitter with a time constant of between 10-20 s, and from these experiments a store of transmitter was estimated for the inhibitory nerve terminals at a single muscle fibre comprising 160,000-200,000 quanta. Furthermore, it was observed [3,5,8] that by adding high concentrations of ~/-aminobutyric acid (GABA) (0.1-0.5 mol/l) to the superfusion, release of excitatory transmitter quanta could be enhanced drastically. In the present paper, it was investigated whether the GABA-induced release of excitatory transmitter quanta could deplete the excitatory terminals of their readily 0304-3940/84/$ 03.00 © 1984 Elsevier Scientific Publishers Ireland Ltd.
252
releasable transmitter stores and whether the mechanism by which GABA activates release depends on extracellular Ca 2 +. The experiments were conducted at small fibres (length ~ 0.6 mm) of the crayfish claw opener muscle (Austropotamobius torrentium). Two glass microelectrodes were inserted into a muscle fibre and the membrane potential was clamped to a constant level. The preparation was superfused with a modified van Harreveld solution which contained (in mmol/l): Na + 205, C1- 242, K + 2.7, Cs + 2.7, Ca 2+ 13.5, Mg 2+ 2.6, trismaleate buffer 10; pH 7.6. Ca z + -deficient superfusions were made by replacing isotonically CaCI2 by NaCI. No EDTA was added to the Ca z + -deficient bathing solutions; therefore, the effective extracellular Ca z+-concentration probably was __<0.1 mmol/1 [10]. Before G A B A was applied continuously to the superfusion, it was applied several times for about 1 s in order to desensitize inhibitory postsynaptic receptors and to prevent a contribution of current noise from inhibitory postsynaptic currents activated by G A B A to the current noise elicited by GABA-induced spontaneous excitatory postsynaptic currents (sEPSCs) [8]. Further experimental details are described in preceding publications [2,6,16]. Figs. I-3 show an experiment in which application of 0.5 mol/1 G A B A evoked intensive, spontaneous release of excitatory transmitter quanta from excitatory nerve terminals at a muscle fibre voltage-clamped at E = - 60 mV membrane potential (T = 20°C). Fig. 1 represents current records obtained in the absence (control) and the presence of GABA. Application of G A B A resulted in a drastic increase in intensity of the current fluctuations, most likely due to superposition of sEPSCs occurring randomly at high rates [3,4,7]. The signal in Fig. 1 starts at the maximal rate of quantal release elicited by G A B A which was reached within a few seconds after the addition of G A B A to the bathing fluid. The average power spectrum calculated from the first 20 s of the GABA-evoked noise current is shown in Fig. 2A. The spectrum was fitted by a Lorentz curve (see legend of Fig. 2) with a cut-off frequency f~=85 Hz. This correlates to an average decay time constant of the sEPSCs:
oo, L!I
O.SooL,I ABA
claw open~ E=-6OmV
i.i ,51 :,~ :ix ~.~.:~;~:~. i ;:'~... ,~;:~.~
T=
20°C
5s
Fig. 1. High rates of spontaneous excitatory postsynaptic currents (sEPSCs) in a voltage-clamped fibre (E = - 6 0 mV, T = 20°C) of the crayfish claw opener muscle evoked by bath application of 0.5 m o l / l G A B A in a normal superfusion (13.5 m m o l / l Ca 2 ÷ ). Noise currents in the absence (control) and the presence of 0.5 m o l / l GABA are presented. The current fluctuations elicited by G A B A attained a maximum within a few seconds ( ~ 5 s), and the signal shown starts at about this ma xi mum.
253
,
A
S If)
r
A2s
B
Slft
~ ~,•90 Hz
~ fc,85 l-lx •
~o~o
¢a~>
"
,
,,',
~ -
OABA,IowC a 2 " - ~ , <3~>• 3t,00quantu/s
= /,000
10 .22
'1~
f I
•
•
,j....|
. . . . . . . .
1
10
l
. . . . . . . .
100
l
IO00Hz
f I
. . . . . . .
1
.l
. . . . . .
10
*.I
100
. . . . .
..,l
1000 HZ
Fig. 2. Power spectrum of the noise current evoked by GABA in the normal and the Ca 2 ÷-deficient superfusion. A: an average power spectrum was calculated for the first 20 s of the signal shown in Fig. 1. The power spectral densities S(f) were fitted by a Lorentz curve: S(f) = S(O)/[1 + (f/f¢)2], with a cut-off frequency fc =85 Hz. An average release rate ( r ) _--4000 sEPSCs/s resulted by means of eqn. 1. B: the experiment was repeated by application of GABA dissolved in the superfusion containing the low Ca 2 ÷-concentration. GABA again evoked a signal of similar intensity as before in the normal superfusion. A Lorentz curve with fc = 90 Hz was fitted to the average power spectrum calculated for a 20-s time interval, and an average rate of release ( r ) = 3 4 0 0 sEPSCs/s was estimated by eqn. 1.
fi "l~q-- 18.5 S, r =-0,98/4
IkHzl ~, -- 110000 quQnto ( t - - ~ ) 6
0
6
12
18
24
30
36
Isl
t
Fig. 3. Time course of the release rate r. According to eqn. 2 release rates, r , for shorter time intervals (1.5 s) were calculated from the GABA-evoked current fluctuations shown in Fig. 1 and plotted as a function of time. The maximum release rate was about rio=6000 sEPSCs/s and fi decreased exponentially with time (rq = 18.5 s, r = - 0.984). The number of quanta released in total was estimated by g = rq • rio--110,000 quanta• This does not include the number of quanta released from the time GABA was applied until the release was maximal ( = 5 s).
254 r(sEPSCs) = 1/(2~rfc)= 1.9 ms. The power spectral density S(O) at small frequencies (f--*0) is proportional to the average frequency (fi) by which the sEPSCs in Fig. 1 occurred during the 20-s time interval [4,7]: (fi)=
--
211+(C.V.) 21
(1)
is the average amplitude of the sEPSCs and C.V. is the coefficient of variation of the amplitudes of sEPSCs around their mean value ft. In muscle fibres of the claw opener from small crayfish ~ = - 0 . 6 6 nA and (C.V.)2 = 0.32 were found at a membrane potential E = - 60 mV in the presence of 0.5 mol/1 G A B A (Finger, W. and Martin, C., unpublished observations). The value of ~ in the presence of the high concentration of G A B A is smaller by a factor of about 1.8 than in the absence of G A B A (Finger, W. and Martin, C., unpublished observation), most likely due to an inhibitory effect of high concentrations of neutral amino acids on the excitatory receptor-channel complex in the postsynaptic m e m b r a n e [3]. Thus, with fc = 85 Hz and S ( O ) = l . 6 x 10 -z° AZs, an average frequency (fi)-_-4000 sEPSCs/s can be calculated by eqn. 1 for the release of excitatory transmitter quanta. Subsequent to this experiment, G A B A was removed from the bathing solution and the preparation was superfused for 15 min with the Ca 2 +-deficient van Harreveld solution; after that G A B A was applied in Ca 2 + -deficient solution. This again gave rise to a signal similar to that obtained during the first application of G A B A (not illustrated), in spite of the considerable reduction of the extracellular Ca / +-concentration. The average power spectrum for the signal recorded in the low CaZ+-solution is represented in Fig. 2B, and according to eqn. 1, ( f i ) = 3 4 0 0 sEPSCs/s was found for the average release rate during the first 20 s following the m a x i m u m of release. However, it was found previously that reducing the extracellular Ca 2 +-cOncentration has a profound diminishing effect on the amplitudes of nerve evoked EPSCs [1]. Furthermore, in low [Ca 2 +] a reduction in quantum current amplitude was found [1], which could mean that in low [Ca z+ ] the average release rate was even larger than 3400 sEPSCs/s. Therefore, from the present experiment it must be concluded that the mechanism by which G A B A enhances release of excitatory transmitter quanta does not depend on the external Ca z +-concentration, at least down to concentrations as low as 0.1 tool/1. As can be seen from Fig. 1, the intensity, respectively the variance, of the current fluctuations, which also is proportional to the release rate of the transmitter quanta, decreased with time indicating that the release rate decreased continuously. To evaluate this process in more detail, the variance oa o f the current fluctuations was determined for small time intervals (1.5 s), and by means of eqn. 2 [4,7]: fi =
4~fca 2 ~.2[1 + (C.V.) 2]
(2)
the release rates fi during these time intervals were estimated and plotted as a func-
255 tion of time (Fig. 3). fi declined exponentially with a time constant rq = 18.5 s, the maximum rate of release being rio = 6000 sEPSCs/s. The number of quanta released in total ( t - ~ ) in the reaction with G A B A can be estimated with the time integral over fi(t): ~-- Io fi(t) dt = fi0rq~- 111,000 quanta. Summarizing, it may be assumed that in the experiment of Fig. 1 on application of 0.5 mol/1 GABA, a readily releasable store of transmitter comprising at least 111,000 quanta was depleted with a time constant of rq = 18.5 s. Thus, like glycine at inhibitory synapses [3,4], G A B A may serve to evoke high rates of quantal release and to deplete excitatory terminals o f their transmitter. Taking into consideration that about 1900 releasing sites exist on a single muscle fibre [9], it seems that approximately 58 quanta are stored at a single synapse ready for release. This conclusion is in agreement with the observations of Nakajima and Reese [15] who observed about 42-117 vesicles at single excitatory synapses in crayfish abdominal muscle. Therefore, it may well be that in the crayfish opener muscle one quantum of transmitter is represented by one vesicle. For the mechanism by which glycine activates release of inhibitory transmitter quanta, we have suggested a Na +-dependent uptake process for glycine [4] which increases the intracellular amino acid concentration in the presynaptic terminal. It remains to be investigated whether for the GABA-evoked release a similar mechanism may be assumed. Uptake of the amino acids into the presynaptic terminal also could be electrogenic giving rise to depolarization of the presynaptic membrane and thus to activation of release of transmitter [12,14]. However, if this were true, a Ca2+-dependence of the release process should be expected [1,11], which was not observed in the present and earlier experiments [4]. Possibly, an increase o f the Na +-concentration inside the cytoplasm of the terminals as a result of activation of the uptake process for amino acids might account for the Ca 2 + -independence of a depolarization evoked transmitter release, as was observed for the release of catecholamines from the adrenal medulla [13]. I would like to thank Prof. Dr. Dudel for reading the manuscript, and Mrs. E. K6ster and Mrs. M. Griessl for technical and secretarial help. This investigation was supported by the Deutsche Forschungsgemeinschaft, Project Fi 305/1-2.
l Dudel, J., The effect of reduced calcium on quantal unit current and release at the crayfish neuromuscular junction, Pflfigers Arch., Europ. J. Physiol., 391 (1981) 35-40. 2 Dudel, J., Finger, W. and Stettmeier, H., Inhibitory synaptic channels activated by y-aminobutyric acid (GABA) in crayfish muscle, Pflfigers Arch., Europ. J. Physiol., 387 (1980) 143-151. 3 Finger, W., Enhanced release of inhibitory and excitatory transmitter quanta in the crayfish neuromuscular junction by glycine and GABA, Neurosci. Lett., 34 (1982) 33-38. 4 Finger, W., Effects of glycine on the crayfish neuromuscular junction II. Release of inhibitory transmitter activated by glycine, Pflfigers Arch., Europ. J. Physiol., 397 (1983) 128-134. 5 Finger, W., Glutamate-operated postsynaptic currents in crayfish claw opener muscle, Neurosci. Lett., 36 (1983) 163-168. 6 Finger, W. and Stettmeier, H., Efficacy of the two-microelectrode voltage clamp technique in crayfish muscle, Pfliigers Arch., Europ. J. Physiol., 387 (1980) 133-141.
256 7 Finger, W. and Stettmeier, H., Analysis of miniature spontaneous inhibitory postsynaptic currents (slPSCs) from current noise in crayfish opener muscle, Pfliigers Arch., Europ. J. Physiol., 392 (1981) 157-162. 8 Finger, W. and Stettmeier, H., Postsynaptic actions of ethanol and methanol in crayfish neuromuscular junctions, Pfltigers Arch., Europ. J. Physiol., 400 (1984) 113-120. 9 Florey, E. and Cahill, M.A., The innervation pattern of crustacean skeletal muscle. Morphometry and ultrastructure of terminals and synapses, Cell Tiss. Res., 224 (1982) 527-541. 10 Hubbard, J.l., Jones, S.F. and Landau, E.M., On the mechanism by which calcium and magnesium affect the spontaneous release of transmitter from mammalian motor nerve terminals, J. Physiol. (Lond.), 194 (1968) 355-380. 11 Katz, B. and Miledi, R., Release of acetylcholine from a nerve terminal by electric pulses of variable strength and duration, Nature (Lond.), 207 (1965) 1097-1098. 12 Kehoe, J., Electrogenic effects of neutral amino acids on neurons of Aplysia californica, Cold Spring Harb. Symp. quant. Biol., 40 (1976) 145-155. 13 Lastowecka, A. and Trifar6, J.M., The effect of sodium and calcium ions on the release of catecholamines from the adrenal medulla: sodium deprivation induces release by exocytosis in the absence of extracellular calcium, J. Physiol. (Lond.), 236 (1974) 681-705. 14 Martin, D.L., Carrier-mediated transport and removal of GABA from synaptic regions. In Roberts, E., Chase, T.N. and Tower, D.B. (Eds.), GABA in Nervous System Function, Raven Press, New York, 1976, pp. 347-386. 15 Nakajima, Y. and Reese, T.S., Inhibitory and excitatory synapses in crayfish stretch receptor organs studied with direct rapid-freezing and freeze-substitution, J. comp. Neurol., 213 (1983) 66-73. 16 Stettmeier, H., Finger, W. and Dudel, J., Glutamate activated postsynaptic channels in crayfish muscle investigated by noise analysis, Pfltigers Arch., Europ. J. Physiol., 397 (1983) 13-19.