Chloride channels gated by extrajunctional glutamate receptors (H-receptors) on locust leg muscle

Brain Research, 481 (1989) 215-220

215

Elsevier BRE 14259

Chloride channels gated by extrajunctional glutamate receptors (H-receptors) on locust leg muscle J. Dudel 1, C. Franke 1, H. Hatt I and P.N.R. Usherwood 2 1physiologisches Institut der Technischen Universitiit Miinchen, Miinchen (F.R. G.) and 2Department of Zoology, University of Nottingham, Nottingham ( U. K. ) (Accepted 2 August 1988)

Key words: Locust muscle; Outside-out patch; Glutamate receptor; Chloride channel; L-Glutamate; Ibotenate; Channel conductance; Channel open time

Outside-out patches of extrasynaptic membrane were isolated from leg muscles of locusts. L-Glutamate and its agonists were applied to such patches either continuously or in rapidly switched pulses. When the pipette contained a high chloride concentration, 2.5 x 10-5 M glutamate triggered single-channel currents (gated by H-receptors) with a conductance of 25 pS which were carried by chloride, in addition to cationic channels (gated by D-receptors). For the chloride channels, the distribution of channel open times had components of about 2 and 12 ms. Pulses of higher glutamate concentrations elicited many superimposed channel openings, and the approximately saturating concentration of 10-3 M glutamate opened 100-200 channels simultaneously. When the pipette contained low chloride, channel conductance was reduced, and the current voltage relation was shifted towards the now negative chloride equilibrium potential. H-Receptor-gated chloride channels were activated by glutamate, ibotenate and aspartate, but not by GABA, quisqualate, kainate, N-methyl-D-aspartateand carbachol. The currents declined in the continued presence of agonist showing a time constant of desensitization >1 s. Recovery from desensitization after removal of the agonist was tested with double pulses and was found to have a time constant of about 300 ms.

INTRODUCTION L-Glutamate is usually classed as an excitatory n e u r o t r a n s m i t t e r yet there is evidence, particularly from studies of invertebrate excitable systems, that this amino acid also gates channels which could reduce m e m b r a n e excitability in nerve and muscle tissues. Extrajunctional regions of locust leg muscle contain glutamate receptors ( G l u R ) or H-receptors which gate chloride channels 9 and which co-exist with excitatory G l u R or D-receptors 1'14. The latter have almost identical physiological properties to the G l u R found postjunctionally at excitatory synapses on locust muscle, but like the H - r e c e p t o r s their physiological role is not yet u n d e r s t o o d 13. In this p a p e r we report on the properties of the H - r e c e p t o r s studied in excised, outside-out patches of locust muscle membrane at the level of the single r e c e p t o r channel using

Correspondence:

the liquid filament technique 6 for rapid application and removal of drugs. A similar study was p e r f o r m e d on the excitatory channels 2. MATERIALS AND METHODS Metathoracic extensor tibiae muscles from adult locusts (Schistocerca gregaria) were treated with 0 . 5 2 mg.m1-1 collagenase (Sigma 1A) in o r d e r to remove connective tissue 3,7. A f t e r removal of the enzyme outside-out m e m b r a n e patches were obtained with fire-polished patch pipettes coated with sylgard. These excised patches were exposed to pulses of Lglutamate and related compounds using the liquid filament switch technique 6. In this a 'liquid filament' is ejected into the superfusion from a small tube which can be shifted by a piezo crystal to pass or to hit the outside-out patch at the tip of fixed electrode.

P.N.R. Usherwood, Department of Zoology, University of Nottingham, University Park, Nottingham NG7 2RD,

U.K. 0006-8993/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

A

20ms

i

~

events per bin

events per bin

1000

50(:

800

20£ ~ . 9 m s IOC

60[

5O

40(

2C1c

20(

5 2 i

I

0

-2

-4

-(5

I

,

-8pA

0

I

I

I

I

8

16

24

32

current amplitude 1

-100 i

II

,

40rns

open time

r--h--

,

b

-80

-60

-40

,

,

,

-20mY i

0

100 ms G,u

lO'3M

|

E _

o

•-2 ~_--, E O.

'-3

•-4pA

I

W '5xlO-5M

Fig. 1. Recordings and evaluations of channel currents triggered by L-glutamate at H-receptors of one outside-out patch of locust muscle membrane• High chloride solution in pipette (intracellular). A: samples of single-channel currents recorded during continuous application of L-glutamate (2.5 × 10-5 M) with a pipette potential of-75 inV. Filter at 1.5 kHz. B: histogram of single H-receptor channel amplitudes obtained from recording in A. The peak at ca. - 2 pA represents the current associated with a single open channel, that at ca. - 4 pA the current generated by two simultaneous channel openings. C: current-voltage characteristic for single chloride channels as measured in A. D: frequency distribution of open times of single channels in A. Only openings which could be resolved unequivocally (i.e. with no superpositions) are included in the distribution• The histogram is fitted approximately with 2 exponentials. E: currents elicited by 100 ms pulses of L-glutamate 10-3, 10-4 and 5 × 10-5 M respectively. Clamp potential -40 mV, same current calibration in all 3 traces. Filter at 1 kHz.

217 Switching times can be arranged to be in the order of 0.1 ms. In a few experiments agonists were applied continuously. The preparations and patches were superfused with locust saline 8 (180 mM NaCI, 10 mM KCi, 2 mM CaC1 e and 10 mM trismaleate buffer; the pH was adjusted to 6.8 with 1 M NaOH). The patch pipettes contained either a low-chloride solution 4 (150 mM potassium propionate, 5 mM sodium propionate, 2 mM MgC12, 1 mM CaCI 2, 10 mM E G T A (Ca 2÷ buffered at 10-s M), 10 mM trismaleate buffer, pH adjusted to 7.2 with K O H so that the total K ÷ concentration was 190 mM) or a high-chloride solution 5 (150 mM KC1, 5 mM NaC1, 2 mM MgCI2, 1 mM CaC12, 10 mM E G T A and 10 mM trismaleate buffer, pH adjusted to 7.2 with K O H to give a final K ÷ concentration of 190 mM and a C1- concentration of 161 mM). The experiments were done at room temperature (20-23 °C). RESULTS When L-glutamate was pulsed onto an outside-out patch, two types of transmembrane current were often seen (see Fig. 3A). However, at the chloride equilibrium potential (Ecl), with either high or low chloride medium in the patch pipette only one of the currents was present, i.e. that due to activation of the cation-selective channels of D-receptors 2. The other current was most obvious at pipette potentials distant from Ecl, and with low concentrations (ca. 10-5 M) of L-glutamate it could be resolved into its single-channel components (Fig. 1A). This low glutamate concentration rarely elicited openings of excitatory (Dreceptor) channels (not shown) which could be easily discerned due to their higher current amplitude of about -15 pA 2. Fig. 1B shows an amplitude histogram of the H-receptor channel currents, with peaks at ca. - 2 a n d - 4 pA. The amplitude o f - l . 9 pA represents the single-channel openings with a conductance of 25 pS, while the -3.8 pA peak was generated by superposition of two-channel openings. In these experiments the chloride concentration in the pipette was high, with chloride concentrations being almost symmetrical across the patch. When the single-channel currents, illustrated in Fig. 1B, were studied at different clamp potentials (Fig. 1C), the extrapolated current-voltage relation showed zero current at about 0 mV, i.e. at the calculated chloride equilibri-

20pA ~LA \

i patch

Epat~

60pA

x

E Patch - 9 0 mV

'

20 mV

- 20

Fig. 2. A: patch-clamp currents elicited by pulses of 1 mM glutamate of 50 ms duration (liquid-filament switch6). Each pulse activated many channels. The patch-clamp potential (Epatch) is indicated to the right of the respective trace. Low-chloride solution in patch pipette. Filter at 5 kHz. B: dependence of peak clamp currents, ipa~ch,from measurements like in A, on the potential, Epatch. All values from same patch as that in A.

um potential. The channel currents were not reversed in this experiment, since measurable, positive single-channel currents would have required fairly large positive potentials with danger of destabilization of the patch (but see Fig. 2). With a low L-glutamate concentration (2.5 × 10-5 M) on average only 0.49 channels were open at any instant (Fig. 1A). To study the effects of higher glutamate concentrations, pulses of glutamate of 100 ms duration were flashed at the same patch. The sample recordings in Fig. 1E show that with each glutamate pulse there was a steep rise of current with a delay of about 10 ms from the beginning of the pulse. After the termination of the pulse it took more than 100 ms for channel openings to cease. These delays are not due to delayed concentration changes, since the same glutamate pulses could elicit maximum activation of D-receptors within about 1 ms, and similarly after the end of the pulse channel activity ceased within a few milliseconds e. The dependence of the chloride current amplitude on the glutamate concentration in the pulse was very steep below 10-4 M (Fig. 1E) the current amplitude rising about 10-fold from 5 × 10-5 to 10 -4 M. Above 10-4 M glutamate, the effect approached a saturation level. At 10-3 M glutamate the maximum current observed was -125 pA which corresponds to the simultaneous opening of about 140 channels (at a potential o f - 4 0 mV, see Fig. 1C). This fairly high channel density is typical for the Hreceptor. If the probability that a channel is activated after a glutamate pulse is assumed to be 1 at saturation, i.e. at 10-3 M, then the probability of activation would be 0.75 at 10-4 M, 0.064 at 5 x 10 -5 M and 0.004 at 2.5 × 10-5 M glutamate, respectively.

218 In recordings with low glutamate concentrations it was possible to identify single openings and closings of the H-receptor channel (Fig. I A ) which gave insight into the open-channel kinetics of this receptor 3-5. The distribution of open times for 2.5 x 10 -5 M L-glutamate is illustrated in Fig. 1D. It contains a short component r t of 2 ms and a larger long component r2 of 12 ms. However, the recording of Fig. 1A was filtered at 1.5 kHz which will have attenuated brief openings and closings. It follows, therefore, that the distribution illustrated in Fig. 1D must be only an approximation. Closed times were not evaluated because the patch contained a large number (at least 140) of channels which presumably contributed to the recording. The recordings above were obtained with a high chloride concentration in the pipette. In order further to substantiate the claim that the recorded currents are carried by CI-, patch clamp currents were measured also in low intracellular C1- solution, in which propionate was the main anion. In this solution the single-channel conductance goes down, which made it impractical to measure the voltage dependence of single-channel currents. Instead, 50 ms pulses of 1 mM glutamate were applied to the patch, which elicited the simultaneous opening of many channels generating a peak current of +52 p A at +20 mV in Fig. 2A. At this potential the current trace was

1 A

t ~

1 3

M

noisy, possibly due to activation of voltage-dependent channels. The chloride current reversed polarity between - 4 0 and - 9 0 mV, with the interpolated current-voltage relation in Fig. 2B suggesting a reversal potential o f - 6 0 inV. According to the chloride-concentration gradient across the patch, a reversal potential of about - 9 0 mV is expected for a pure CIcurrent. However, locust muscle is not completely impermeant to large anions 1°. Since the H-receptor channel is slightly permeable to propionate then this would account for the deviation of the reversal potential from the chloride equilibrium potential. The only alternative would be a flow of K +, but glutamate-activated currents like those shown in Figs. 1 - 3 A are not affected by replacing K + with Cs +, a cation which blocks K + currents (see also ref. 10). Lea and Usherwood have previously shown that the H-receptor channel of locust muscle is sparingly permeable to methylsulphate and even less so to sulphate. In this respect it differs from the chloride channel gated by the locust muscle G A B A receptor where methylsulphate is impermeant I°. Unlike its excitatory GIuR counterpart on locust muscle (see Discussion), the H-receptor channel is gated by D L - i b o t e n a t e (10 -4 M) t° but not by L-quisqualate (10 -2 M) (Fig. 3) (see also refs. 13, 14). Chloride channels were never seen when N-methyiI>aspartate ( N M D A ) (10 -2 M), L-kainate (10 -2 M),

t

D

.

1 Ibo

10-4M

t

G

50 pA

C

Quis 10"3M

';VI1

F

:............

oi~ 100

1 ' ,' ~ 200 300 ms

I

'

I ' 1 ' ', ' 100 200 3 0 0 ms

[ 0

! 0.5

I 1

I 1.5 s

Fig. 3. Pharmacology of the H-receptor studied in an outside-out patch. Low chloride in pipette. Compounds were applied as brief pulses using the liquid filament technique 6. Pulse duration in A - F 100 ms (arrows) pulses repeated at 4 s intervals. A: 10-3 M L-glutamate gated a large outward (chloride) current (containing many single channel currents) seen as an upward deflection, superimposed on which is an initial burst of inward (cationic) channel currents. B: 10-2 M aspartate also gated an outward current, with rare openings of the cationic channel (arrow). C: 10-4 M L-quisqualate evoked a burst of inward (cationic) channel current. D: 10-4 M DL-ibotenate gated an outward current only. No currents were gated by 10-2 M N-methyl-D-aspartate (F) and L-kainate (E), respectively. In all of these recordings the patch potential was -29 inV. G: outward current during a 1 s pulse of 10-2 M aspartate. Inactivation of the response (desensitization) with an approximate time constant >1 s. Patch depolarized to 0 mV. All recordings filtered at 5 kHz.

219 G A B A (10 -4 M), glycine (10 -2 M) and carbachol (10 -5 M) were pulsed onto outside-out patches, but they were obtained with high concentrations (10 -2 M) of L-aspartate (Fig. 3). Apart from their different pharmacological properties the D- and H-receptors of locust muscle also respond differently to concanavalin A. Whereas desensitization of the former is inhibited by this lectin, concanavalin A has no effect on desensitization of the H-receptor 11't2. In the absence of concanavalin A, the H-receptor desensitizes more slowly than the D-receptor. For example, during a 1 s pulse of 10 -2 M aspartate the H-receptor current declined in amplitude (Fig. 3) with a time constant >1 s which is orders of magnitude greater than the inactivation time constant of the D-receptor (two kinetic types of receptors with apparent desensitization constants of 25 ms and 3 ms, respectively) 2. Of additional interest was the continuation of H-receptor channel gating, albeit at a low level, after prolonged application of L-glutamate or DL-ibotenate. Recovery from desensitization was studied in the experiment of Fig. 4 using double pulses of 1 mM glutamate, each of which lasted for 50 ms. The example in Fig. 4A shows the currents elicited by pulses applied with an interval of 200 ms, and repeated every 4 s. The rise in current due to the second pulse is much smaller than that elicited by the first pulse which indicates desensitization during the first pulse. When the pulse interval is varied, the current elicited by the second pulse is seen to be only 0.32 of that after the first pulse for a 0.1 s pulse interval, but the receptors recover from this desensitization within 1 s, with an apparent time constant of about 300 ms. DISCUSSION In the same outside-out patches of locust muscle membrane, two types of channels were activated by glutamate: excitatory channels permeable to a mixture of cations corresponding to D-receptors 13,14, and inhibitory channels permeable to chloride corresponding to H-receptors 9,13,14. These channel types differed not only in their specific permeabilities, but also in other characteristics. One such difference may be in the number of channels of each type present in an outside-out patch. While up to about 20 cationic channels could be opened by saturating concen-

pulse 1 1~ pulse 2 ~/

A

~1 ~

2x50 ms Ou 1 rrtvl 0.2 s ¢~torval

I. 0.8 ¸ 0.6.

o+4:

0~0

!~Jl,.~ 2

./

13

0.21

ol

'

0'.2

'

o14

o16

o18

i~

delay of pulse 2

Fig. 4. Recovery from desensitization. A: example of outward currents elicited by two double pulses of 1 mM glutamate, each lasting 50 ms. The delay between pulse 1 and pulse 2 was 200 ms, and the interval between pulse pairs was 4 s. Same patch as that in Fig. 2, and same recording conditions, patch potential 0 mV. B: plot of average amplitudes of currents elicited by the second pulse relative to the amplitude of current gated by the first pulse (rel./patch), in dependence on the delay of pulse 2 after pulse 1. The reactions to the second pulses were measured as rises in current over the decay of the current elicited by the first pulse. trations of agonists 2, more than 100 chloride channels were activated under analogous conditions (compare cationic and anionic channels in Fig. 3). A clear difference is that agonists of the cationic channel are quisqualate, glutamate and aspartate 2 (in order of decreasing potency), whereas those of the anionic receptor-channel are glutamate, ibotenate and aspartate. The reaction kinetics were generally much slower in the chloride channels than in the cationic ones. The same pulse of glutamate flashed at a patch could elicit maximum activation of cationic channels within less than 1 ms 2, while the first chloride channels were opened with delays in the range of 10 ms and full activation was reached only after 20-50 ms (Figs. 1, 3). Activations of cationic channels were terminated by desensitization with time constants of 25 or 3 ms, respectively2, while desensitization was slower by orders of magnitude and incomplete for the chloride channels. Even when the agonist was switched off, activity of the chloride channels subsided only slowly, with time constants in the range of 50 ms. The cationic and the anionic channels gated by glutamate thus show pronounced kinetic differences. The glutamate-activated chloride channel in locusts is similar in some respects to the glutamate-acti-

22(I vated chloride channel of crayfish muscle ~o.lt. Several hundred chloride channels can be activated in similar patches of both locust and crayfish muscle. Channel conductance is about 25 pS in both species if the chloride concentrations on both sides of the patch are

glycine, and in case of low extracellular Ca -~~ concentration also acetylcholine and carbachol ~5. The different agonists in crayfish preferentially activate characteristic multiples of the channel conductance 5,

high. Also the channel mean open times are of the

but analogous activations of multiples were not seen in patches from locust muscle.

same order of magnitude. Pulses of agonists trigger a relatively slow rise and fall of activation of chloride

ACKNOWLEDGEMENTS

channels in both species, and also desensitization is slow and incomplete in both. However, the spectrum of agonists differs: while in the locust the chloride

The authors wish to thank Miss Bettina Weich for technical and Mrs. Martina Griessl for secretarial

channels are activated by glutamate, ibotenate and

help. This work was supported by the Deutsche Forschungsgemeinschaft, SFB 220 and by a grant from

aspartate (decreasing order of potency), the respective agonists in crayfish are quisqualate, glutamate, ),-aminobutyric acid, fl-guanidino-propionic acid and

REFERENCES 1 Cull-Candy, S.G. and Usherwood, P.N.R. The site of action of ibotenic acid and the identification of two populations of glutamate receptors on insect muscle fibres, Nature (Lond.), 246 (1973) 62-64. 2 Dudel, J., Franke, Ch., Hart, H., Ramsey, R.L. and Usherwood, P.N.R., Rapid activation and desensitization by glutamate of excitatory, cation-selective channels in locust muscle, Neurosci. Lett., 88 (1988) 33-38. 3 Franke, Ch. and Dudel, J., High resolution measurements of single channel currents activated by glutamate in crayfish muscle, Neurosci. Lett., 59 (1985) 241-246. 4 Franke, Ch., Hatt, H. and Dudel, J., The excitatory glutamate-activated channel recorded in cell-attached and excised patches from the membranes of tail, leg and stomach muscles of crayfish, J. Comp. Physiol., A159 (1986) 579-589. 5 Franke, Ch., Hatt, H. and Dudel, J., The inhibitory chloride channel activated by glutamate as well as y-amino-butyric acid (GABA), J. Comp. Physiol., A159 (1986) 591-609. 6 Franke, Ch., Hatt, H. and Dudel, J., Liquid filament switch for ultra-fast exchanges of solutions at excised patches of synaptic membrane of crayfish muscle, Neurosci. Lett., 77 (1987) 199-204. 7 Huddie, P.L. and Usherwood, P.N.R., Single potassium channels of adult locust (Schistocerca gregaria) muscle recorded using the giga-ohm seal patch clamp technique, J.

the British Science and Engineering Research Council.

Physiol. (Lond.), 378 (1986) 60P. 8 Kerry, C.J., Kits, K.S., Ramsey, R.L., Sansom, M.S.P. and Usherwood, P.N.R., Single channel kinetics of a glutamate receptor, Biophys. J., 51 (1987) 137-144. 9 Lea, T.J. and Usherwood, P.N.R., The site of action of ibotenic acid and the identification of two populations of glutamate receptors on insect muscle fibres, Comp. Gen. Pharmacol., 4 (1973) 333-350. 10 Lea, T.J. and Usberwood, P.N.R., Effect of ibotenic acid on chloride permeability of insect muscle fibres, Comp. Gen. Pharmacol., 4 (1973) 351-363. 11 Mathers, D.A. and Usherwood, P.N.R., Concanavalin A blocks desensitization of glutamate receptors on insect muscle fibres, Nature (Lond.), 259 (1976) 409-411. 12 Mathers, D.A. and Usherwood, P.N.R. Effects of concanavalin A on junctional and extrajunctional L-glutamate receptors on locust skeletal muscle fibres, Comp. Biochem. Physiol., 59C (1978) 151-155. 13 Usherwood, P.N.R., Amino acids as neurotransmitters, Adv. Comp. Physiol. Biochem., 7 (1978) 227-309. 14 Usherwood, P.N.R., Glutamate synapses and receptors on insect muscle. In G. Di Chiara and G.L. Gessa (Eds.), Glutamate as a Neurotransmitter, Raven, New York, 1981, pp. 183-193. 15 Zufall, F., Franke, Ch. and Hatt, H., Acetylcholine activates a chloride channel as well as glutamate and GABA. Single channel recordings from crayfish stomach and opener muscles, J. Comp. Physiol., in press.