0306-4522/92$5.00+ 0.00 Pergamon Press Ltd 0 1992IBRO
NeuroscienceVol. 48, No. 4, pp. 813819, 1992 Printed in Great Britain
MEMBRANE CURRENTS INDUCED BY L-HOMOCYSTEIC ACID IN MOUSE CULTURED HIPPOCAMPAL NEURONS L. VYKLICK+* and V. VLACHOVA Institute of Physiology, Czechoslovak Academy of Sciences, Videdska 1083, 142 20, Prague 4, Czechoslovakia Abstract-The concentration-response relationship of membrane currents induced by L-homocysteic acid was studied on mouse embryonic hippocampal neurons in culture (a = 56). In the majority of neurons two phases in the dose-response relationship could be distinguished. The first was characterized by responses to 3-100 PM L-homocysteic acid which desensitized with a time-constant > 1 s in a concentration-dependent manner and were antagonized by 30 p M n-L-2-amino-5phosphonovaleric acid indicating activation of the N-methyl-D-aspartate receptors. At higher concentrations of L-homocysteic acid this component was strongly depressed. The second phase was characterized by sustained responses that were concentration-dependent (1 mM L-homocysteic acid maximum concentration tested) and were not blocked by p-L-2-amino-5-phosphonovaleric acid indicating activation of non-N-methyl-D-aspartate receptors. Eight neurons did not exhibit these two-phase characteristics in the concentration-response relationship at the beginning of the recording. The magnitude of responses to L-homocysteic acid was positively related to concentration and the responses were partially blocked by D-L-2-amino-5-phosphonovaleric acid. In these neurons, however, repeated applications of L-homocysteic acid at concentrations 30pM up to 300 PM resulted in a long-lasting, three- to four-fold increase of the membrane current. This increase was completely blocked by D-L-2-amino-5-phosphonovaleric acid (SO-100 p M) suggesting that it was produced by activation of receptors. It is suggested that the long-lasting increase of the membrane current after repeated L-homocysteic acid application reflects long-term changes in the mechanisms involved in controlling N-methyl-o-aspartate
receptors from an intracellular site in which Ca*+ plays an important but not an exclusive role.
L-homocysteic acid (L-HCA) is known for its strong excitatory action on central neurons of vertebrates from the pioneering studies of Curtis and Watkins in the early sixties. sg6Interest in the mechanisms of its action has increased after it was demonstrated that L-HCA can be released from brain slices by high concentrations of K+ in a Caz+-dependent manner which raised the possibility that it might be an endogenous transmitter. *s9However, the localization of the HCA-like immunoreactivity predominantly in glial elements4 and the controversy about the actual levels of L-HCA in the central nervous system36 raised a problem in accepting this idea. It had been shown earlier that the responses to L-HCA in cultured central neurons exhibit a negative slope conductance in the presence of Mg2+ at negative membrane potential$” and that the effects of L-HCA are blocked by Q-phosphonatesr5 which suggested that L-HCA acts at N-methyl-D-aspartate
(NMDA) receptors. In a detailed study on structure-activity relationships for a number of excitatory amino acid transmitter candidates in which agonist action on NMDA and non-NMDA receptors was examined in isolation without use of antagonists, it was shown that L-HCA selectively activates the NMDA receptors only over a narrow concentration range and that it is also a potent quisqualate receptor agonist. 24These findings are in good agreement with the results of the ligand binding experiments which demonstrated that the binding potency of L-HCA for NMDA and kainate receptors is only slightly lower than that of L-glutamate.23,25 The suggestion that L-HCA might be an endogenous transmitter has recently been corroborated by the finding that the selective uptake blocker of L-HCA, fl-P-chlorophenylglutamate (Chlorpheg),’ potentiates L-HCA- but not L-glutamate-induced responses38 and increases synaptic responses induced in CA1 hippocampal neurons by Schaffer collateral stimulation.” These findings also raised the possibility that L-HCA might be involved in self-potentiation of responses to NMDA receptor activation observed on neurons in hippocampal slices.2*28,29 The present study was designed to test the effects of L-HCA on cultured hippocampal neurons to explore whether it may induce long-lasting changes in NMDA receptors.
*To whom correspondence should be addressed. AP-5, D-L-2-amino-5-phosphonovaleric acid; BAPTA, 1,2-bis(o -aminophenoxylethane)-N, N, N’, N’tetra-acetic acid; ECS. extracellular solution: EGTA, ethyleneglycol-bis-(ta-aminoethyl-ether)-N,N,N’,N’~ tetra-acetic acid; HEPES, N-2-hydroxyethylpiperazineN’-2-ethanesulphonic acid; L-HCA, L-homocysteic acid; MEM, minimum essential media; NMDA, N-methyl-naspartate.
Abbreviations:
813
814
L. ~¢VKL]CK~'and V. VLACHOV.~ EXPERIMENTAL PROCEDURES
A
Cell culture
L-HCA
The experiments were performed on primary cultures of hippocampal neurons from embryonic mouse (BALB/c, Institute of Physiology, CAS, Prague). Nerve cell cultures, were prepared by dissociating the hippocampi of 16-17-dayold embryos and plating the suspension onto confluent hippocampal glial cell feeder layer cultures. The mothers were killed by cervical dislocation prior to removal of the fetuses. The growth medium contained minimum essential medium (MEM), 5% horse serum, and a nutrient supplement containing transferrin, insulin, selenium, corticosterone, triiodothyronine, progesterone and putrescine. ~t
Recording and perfusion techniques Experiments were performed at room temperature (22-23°C) seven to 11 days after cultures had been plated. Solitary neurons of body diameter 15-20/~m were usually selected for examination. Whole-cell membrane current recordings were carried out with a List EPC-7 amplifier. The series resistance of recording pipettes was usually around I0 MfL Responses were recorded on a pen recorder and stored on video tape after digital sampling at 20 kHz for later evaluation. The system for fast superfusion of the neurons was in principle similar to that introduced by Johnson and Asher t4 and further elaborated in other laboratories. 24'33'35It consisted of a planar array of eight parallel glass tubes 0.4 mm in diameter through which solutions were driven by a peristaltic pump. The orifice of one of the tubes was placed about I00/zm above the soma of a selected neuron which allowed exchange of the solutions with a time-constant of 50-80 ms. The step motor for changing the position of the tubes and the time of opening the solenoid valves which directed the flow of the solution either onto neurons or back to reservoirs was controlled by a microcomputer programmed to perform a sequence of applications. The culture dishes, 3 cm in diameter, were independently perfused by extracellular solution (ECS; 1 ml/min).
Experimental solutions The ECS contained (in mM) NaC1 160, KCI 2.5, CaC12 1, MgC12 2, HEPES 10, Glucose 10; pH was adjusted to 7.3 with NaOH. In Mg-free solutions, MgC12 was omitted. The pipette solution contained (in mM): CsC1 140, KC1 2.5, CaCI 2 1, MgC12 2, EGTA 5, HEPES 10, pH was adjusted to 7.3 with CsOH. Cells were washed by a rapid stream of ECS before and after each drug application for 5 s. L-HCA was applied in Mg-free extracellular solution with 3 p M glycine. Picrotoxin (10 -4 M) was added to all solutions to block GABA activated channels. Tetrodotoxin (2.10 -6 M) was added when synaptic activity or spikes became a problem. L-HCA, NMDA, D-L-2-amino-5-phosphonovaleric acid (AP-5) and glycine were provided by TocrisNeuramin.
RESULTS
Concentration-response relationship In order to prevent excessive disuse, the application o f L-HCA was limited to 5 s. The c o n c e n t r a t i o n response relationship of the whole-cell responses to L-HCA in the Mg-free solution containing 3 # M glycine exhibited two phases (Fig. 1A). T h e first phase was characterized by responses with the threshold slightly below 3/~M. The c o n c e n t r a t i o n o f L-HCA at which the largest responses were observed varied between 30 a n d 300/~M. In this range o f
~ ,~'~ ~ ~o ~ B ~
~
~
~
~
~
~
-50mY
15s
Fig. I. Series of whole-cell responses to ascending concentrations of L-HCA in a cultured hippocampal neuron. L-HCA in concentrations 3, 10, 30, 100, 300/LM and 1 mM in Mg-free extracellular solution with 3 # M glycine was rapidly superfused around the neuron for 5-s periods which were separated by 10-s intervals during which the neuron was washed with an extracellular solution containing 2 mM MgC1z. The second application of 30/tM L-HCA at the end of the series served as a control, showing that no serious changes of the responses had occurred during recording. (A) The first series of recording from the neuron. (B) The seventh series recorded about half an hour later. Membrane potential, - 5 0 mV. The bars above the records indicate time of L-HCA application.
L-HCA c o n c e n t r a t i o n s the responses declined with a slow time-constant of > 1 s. Five seconds o f application was n o t usually long e n o u g h for the declining phase of the responses to reach the plateau. T h e value o f the t i m e - c o n s t a n t calculated from digitalized responses varied between 1-2.2 s. A t c o n c e n t r a t i o n s higher t h a n t h a t needed to induce the m a x i m u m response, a m a r k e d decrease o f the responses was observed; the sustained c o m p o n e n t of the responses, however, became m o r e p r o m i n e n t . The second phase o f the c o n c e n t r a t i o n - r e s p o n s e relationship was d o m i n a t e d by a sustained comp o n e n t which increased in a m p l i t u d e with the L - H C A concentration. The m a x i m u m c o n c e n t r a t i o n used in o u r experiments were 600/~M or 1 m M L-HCA. After several series o f L - H C A applications, r u n d o w n o f the m e m b r a n e current was usually observed after 10-15 min. This was characterized by a higher conc e n t r a t i o n of L - H C A needed to induce the t h r e s h o l d response a n d by a m a r k e d decrease o f the declining c o m p o n e n t o f the responses to L - H C A (Fig. 1B). T w o phases in the c o n c e n t r a t i o n - r e s p o n s e relationship were, however, preserved d u r i n g the whole time o f recording from the cell. Usually f r o m five u p to 10 series o f recordings were m a d e f r o m one n e u r o n . In view o f previous studies 24"25 the m o s t likely i n t e r p r e t a t i o n o f these results was t h a t L - H C A is a mixed agonist at g l u t a m a t e receptors with a preference for N M D A - o p e r a t e d channels. T h e initial desensitizing c o m p o n e n t o f the responses could thus be a t t r i b u t e d to activation o f N M D A receptors, while the sustained c o m p o n e n t induced at high concentrations of L-HCA would be generated by cationic currents induced by activation o f n o n - N M D A receptors o f the kainate type.
815
Membrane currents induced by L-homocysteic acid in mouse cultured hippocampal neurons
A 1
f
L-HCA
h
B
-80 mV
c
3.104 w1
D lO-’ wa!s I-
-I”!-
3.10” -
>.
I”
I
XOpA
1Sr
Fig. 2. Antagonistic effects of AP-5 on responses to L-HCA. (A) L-HCA in a Mg-free extracellular solution containing 3 PM glycine was applied either alone (10, 100 and 3OOpM) or together with 30 bM AP-5 (thick bars). (B) Three-hundred micromolar L-HCA (thick bar) was interposed between 10 uM L-HCA an&cations. (C) Three-hundred L-HCA (thick bar) was in&posed be&e&r 1OOpM L-HCA applications. The bars above the records indicate the time of L-HCA applications.
The eflects of D-L-Zamino-S-phosphonovaleric acid on responses to L-homocysteic-acid To distinguish clearly between the NMDA and non-NMDA components of the responses to L-HCA we used AP-5, a competitive antagonist of the NMDA receptors. Each concentration of L-HCA
tested was applied either alone or with AE-5. We found that the declining component of the responses to L-HCA but not the sustained component induced by higher concentrations of L-HCA was strongly depressed when L-HCA was applied together with 30 PM AP-5 (Fig. 2A). A marked depression of the responses to L-HCA without AR5 was, however, observed when the L-HCA concentration was increased to 300 or 600 PM. These findings demonstrate that the declining component of the responses to L-HCA is produced by activation of NMDA receptors and, in agreement with other studies,r@5 indicates a higher affinity of L-HCA for NMDA receptors than for non-NMDA receptors. A seemingly paradoxical concentration-response relationship was observed when 300 pM L-HCA was interposed between the responses to 100 PM concentration (Fig. 2C) but not when it was interposed between the responses induced by 10 FM L-HCA (Fig. 2B). The suppressing effects of high concentrations of L-HCA on the AP-5-sensitive component were present both on the negative and the positive membrane potentials (Fig. 3A,B). In agreement with previous studies which indicated that desensitization of the responses to NMDA is concentration-dependent’~33~39 we found that the membrane current induced by L-HCA at concentrations which produced desensitization markedly increased under experimental conditions when LHCA concentration around the neuron was allowed to decrease slowly and not rapidly by washing the neuron with ECS. This was achieved by discontinuing
Fig. 3. Voltage-independence but concentration-dependence of the desensitization of the responses to L-HCA. (A, B) Responses to 300 nM L-HCA interposed between responses to 30 ,uM L-HCA at membrane potential -80 mV and +20 mV, respectively. (C, D) Responses to 30 and 100 PM L-HCA respectively. L-HCA was applied during the period indicated -o-; wocw indicates the paradigm of L-HCA application. w, The neuron was washed by extracellular solution. At the interval -o- the position of the array of the tubes was changed so that the tube containing L-HCA was facing the neuron and the valve which permitted flow of L-HCA on the neuron was open. During -c- application of L-HCA was discontinued and the tube was removed from the neuron. At -w- the array of the tubes returned to the original position and the neuron was washed by a fast stream of ECS. The diiference between -c- and -w- lies in that in c- the stream of L-HCA was only halted so that its concentration was slowly decreasing by diffusion and bath perfusion while in -w-, L-HCA was halted and rapidly removed around the neuron by washing it with extracellular solution.
application of L-HCA during the declining phase of the response and by placing the barrel away from the neuron. At this period the concentration of L-HCA decreased by diffusion and slow bath perfusion. The magnitude of the response immediately began to increase, as can be seen in Fig. 3C and D. The membrane current returned to the base line immediately after the array of tubes for application of the solutions returned to its original position and the neuron was washed with a fast stream of the extracellular solution (-w-). Long-lasting increase of the responses after repeated L-homocysteate applications
In eight neurons out of 56 which were exposed to repeated series of L-HCA applications, two phases in the dose-response relationship were not observed at the beginning of the recording from the neuron. In the first series of L-HCA applications, the magnitude of the responses was positively related to the concentration and the responses were partially blocked by AI’-5 (Fig. 4A). In the second or the third series of L-HCA applications the membrane current induced by a certain concentration of L-HCA (30-300 PM) began to increase. The increase. developed abruptly during L-HCA application and resembled breaking down of the cell membrane. The increased membrane current, however, returned to the resting level rapidly after L-HCA application had been completed and the neuron had been washed with the extracellular
816
L.
VYKLICKL
and V.
VLACHWA
L-HCA +G3pM
3 IO-' -
3 IO-~
-
-
-
-5OmV
B +50 mV
1
lOOpA
15s
I
0 “1
& 4 +
Fig. 4. Whole-cell responses to repeated applications of L-HCA. Three series of responses to L-HCA at concentration 10, 100 and 3OOpM without and with AP-5 as indicated. Ten micromolar AP-5 was applied together with 10 PM L-HCA; 30pM AP-5 was applied together with L-HCA 100 and 3OOpM. (A) tire first series, (B) the third series, and (C) the last (seventh) series of recording from the neuron. Appreciable leakage current was observed during recordings in the last series which explains that all responses are smaller. Membrane potential -50 mV. solution (Fig. 43). The magnitude of subsequent responses to higher concentrations of L-HCA in that series of applications were similar to those in the previous series. Usually not more than 2min were allowed between the series of applications. Once the increase in the response to a certain concentration of L-HCA developed, it remained unchanged during the whole time of recording from the neuron, i.e. until it was lost or discarded (Fig. 4C). Such a long-lasting increase in the responses to L-HCA was completely blocked by AP-5 (30 PM-100 PM) suggesting that it is produced by activation of NMDA receptors. The development of a long-lasting increase of the membrane current after repeated expositions of the neuron to L-HCA was not voltage-dependent, as is demonstrated in Fig. 5. In this neuron, some increase of the response could already be observed at a membrane potential of - 50 mV in the first series of applications at the end of the second response to 300 ~1M L-HCA. A marked increase of the response, however, was observed in the second series when the membrane potential was changed to +50 mV (Fig. SB). The increase of the membrane current lasted to the very end of the recording from the neuron, about 30min later, and it could be completely blocked by AP-5 during this time (Fig. 5C).
The major findings in our study on cultured hippocampal neurons are the concentration-dependent desensitization of the NMDA component of the
L-HCA
i&s--
3- 10-8-
&*
---
3- 10-S
n
3 10.’
+G~IIM
Fig. 5. Three series of responses to 30 and 300 PM L-HCA without and with AP-5 (50 and IOOpM, respectively). (A) The first series of recordings from the neuron at - 50 mV. (B, C) The third and the sixth series of recordings from the same neuron at membrane potential +50mV.
responses to L-HCA and the long-lasting increase of the responses to L-HCA after repeated applications which was observed in some neurons. Desensitization of N-methyl-o-aspartate during the action of L-homocysteate
receptors
A large range of concentrations of L-HCA and other excitatory amino acids were used by Patneau and Mayeti4 to quantify concentration-response relationships under experimental conditions which allowed us to study the generation of the membrane currents either through the NMDA- or quisqualateoperated channels in isolation without use of antagonists. To avoid Ca*+-dependent desensitization of NMDA receptors, they used a low concentration of Ca2+ (0.2 mM) in the extracellular solution and a very potent calcium chelating agent, BAPTA, in the pipette solution. On the other hand, 1 mm Ca2+ in ECS and a less potent chelating agent, EGTA, in the pipette solution were used in our experiments which can be expected to allow appreciable increases in intracellular Ca*+ during massive activation of NMDA channels. ” These differences in the composition of experimental solutions seem to be the most likely explanation why the NMDA component of the responses to L-HCA desensitized in our experiments. Admixture of Mg*’ from the ECS containing 2 mM Md+ to the Mg-free solution in which L-HCA was applied can be excluded as a possibility for the fade of the responses in our experiments because similar results were obtained at positive membrane potentials or when ECS without M&+ was used for a rapid superfusion of the neuron. Moreover, the
Membrane currents induced by L-homocysteic acid in mouse cultured hippocampal neurons amplitude of the desensitized response increased when application of L-HCA was discontinued which gave a better chance of mixing the Mg-free and Mg-containing solutions (see later). There are two reasons which indicate that the component of the responses to L-HCA which showed a marked desensitization is really accounted for by activation of the NMDA and not the quisqualate receptors. First, the rate of desensitization is relatively slow with a time-constant characteristic for the whole-cell responses to NMDA,32,33*39 while the rate of desensitization of quisqualate receptors is very fast with a time-constant of several milliseconds.26~31 Second, the desensitizing component of responses to L-HCA was effectively antagonized by AP-5, a competitive antagonist of NMDA receptors. At low concentrations of L-HCA, however, the sustained component of the responses was also partially blocked by AP-5 indicating that NMDA channels contribute to its generation. On the other hand, the sustained component of the responses to high L-HCA concentrations was insensitive to AP-5 suggesting that it is induced by activation of non-NMDA receptors. There are at least two possibilities which should be considered for explaining why the NMDA component of the responses to L-HCA was not detected at high concentrations. The first is that L-HCA, in addition to agonist recognition sites of glutamate receptors, binds to a low-affinity site on the outside of the NMDA receptor complex from where it can be. inhibited with an action similar to that of Zn or protons.19,27*~~3S The main difference between L-HCA and L-glutamate lies in the presence of an atom of sulphur in L-HCA which might be suspected to have modulatory effects on NMDA receptors. However, despite its appeal, this idea does not seem likely because no desensitization even to high L-HCA concentrations was observed when a low concentration of Ca*+ was used in the extracellular solution.24 The second possibility is that the kinetics of desensitization with high L-HCA concentrations exceeds the rate of drug application of the system used in our experiments. This explanation is favoured by the evidence that desensitization of NMDA receptors is dependent on agonist wncentration’*39 and by our finding that the magnitude of the responses to L-HCA increased when its concentration around the neuron was slowly decreasing (Fig. 3C, D). The relatively rapid rate of desensitization of the NMDA component of the responses to high L-HCA concentration can explain the seemingly paradoxical finding that a higher agonist concentration can produce a smaller response than a lower concentration, similar to findings described for quisqualate receptors.tO Apparently such a paradox can be expected to occur at all receptors which exhibit a bell-shaped dose-response relationship due to agonist concentration-dependent desensitization. Its mechanisms have already been
817
analysed at the single channel level of the nicotinic acetylcholine receptor.3 Long-lasting increase of the membrane current during repeated L-homocysteate application It does not seem likely that artifacts in recording whole cell currents such as a sudden change of the series resistance or transient release of the gigaseal could account for the long-lasting increase of the membrane current which was observed in some neurons after repeated applications of L-HCA because recordings remained stable, without marked changes which could not be explained by irreversible rundown due to lack of MgATP in the pipette solution.*’ The increase in the membrane current induced by repeated applications of L-HCA was observed in both pyramidal and granule-like neurons. It was, however, essential that NMDA receptors were abundantly developed, i.e. that the neurons exhibited large membrane currents to NMDA which was always tested in one dish of a new batch of cultures. Development of the long-lasting increase of the membrane current was observed only under the conditions in which the concentration of L-HCA was sufficient to activate massively both the NMDA and the non-NMDA receptors and L-HCA was applied repeatedly. The finding that the increase of the membrane current was completely blocked by AP-5, however, indicates that it is produced exclusively by activation of the NMDA channels. Once the increase of the membrane current to L-HCA developed, it lasted the whole time of recording from the cell. Our experiments, however, cannot answer the question on the reversibility unequivocally, because the testing for its presence represented at the same time a stimulus for its development. Two alternatives should be considered in explaining mechanisms possibly involved in this long-lasting increase of the membrane current: relief of desensitization of NMDA receptors and activation of a subset of dormant NMDA receptors, There are at least two different mechanisms which underlie desensitization of NMDA receptors. The calcium-mediated desensitization in which Ca*+ directly or indirectly, via activating cytoplasmic enzymes regulates the NMDA receptors from the interior of the neuron17s39 and the glycine-sensitive desensitization which modulates activity of the NMDA receptors from the extracellular side.33 The mechanisms suggested to underlie the reduction of the desensitization of glutamate receptors by diazoxide,” aniracetam, wheatgerm agglutini# and concanavalin A’* do not seem to apply to NMDA receptors because they concern fast desensitization of the quisqualate receptors. One possibility for explaining the development of a sudden increase of the membrane current through NMDA channels is that intracellular calcium controls processes of desensitization only to a critical
818
L.VYKLICKC~~~
concentration which, when exceeded, results in blocking the mechanisms, possibly cytoplasmic enzymes, involved in regulation of NMDA receptor desensitization. Our finding that a critical but not a higher concentration of L-HCA induced the long-lasting increase of the responses would be compatible with this idea when assuming that the enzymes for regulating NMDA receptors escape washing out during whole-cell recording and that the rate of desensitization on the extracellular side depends on agonist concentration and effectively blocks the flow of Ca* + through NMDA channels. Although the NMDA receptor has recently been cloned’6,22 there is still not enough information which
V.VLACHOVA
would allow meaningful speculation whether thcrc might exist a subclass of dormant NMDA receptors which could be activated only after repeated I -HCA applications and thus produce a long-lasting increase in the membrane current. However, in view q>f the complexity of the non-NMDA receptors’-’ it can be envisaged that the diversity of NMDA receptors ~111 still be larger in order to comply with the plastic processes in which they are likely to be involved.
Acknowledgements-This
work was supported by GA CSAS (71114). The authors thank Dr P. Hnik for reading the manuscript.
REFERENCES 1. Clark G. D., Clifford D. B. and Zorumski C. F. (1990) The effect of agonist concentration,
membrane voltage and calcium on N-methyl-D-aspartate receptor desensitization. Neuroscience 39, 787-797. 2. Collingridge G. L., Herron C. E. and Lester R. A. J. (1988) Frequency-dependent N-methyl-o-aspartate receptormediated synaptic transmission in rat hippocampus. .I. Physiol. 399, 301-312. 3. Colquhoun D. and Ogden D. C. (1988) Activation of ion channels in the frog end-plate by high concentrations of acetylcholine. J. Physiol. 395, 131-159. 4. Cubnod M., Do K. Q., Streit P. (1990) Homocysteic acid as an endogenous excitatory amino acid. Trenlls pharmac. Sci. 11, 477-478. 5. Curtis D. R., Phillis J. W. and Watkins J. C. (1961) Actions of amino acids on the isolated hemisected spinal cord of the toad. Br. J. Phurmac. 16, 262-283. 6. Curtis D. R. and Watkins J. C. (1963) Acidic amino acids with strong excitatory actions on mammalian neurons. J. Physiol. 166, I-14. 7. Davies J., Francis A. A., Oakes D. J., Shardown M. J. and Watkins J. C. (1985) Selective potentiating effect of
B-p-chlorophenylglutamate Neurophannacology
on responses induced by certain sulphur-containing excitatory amino acids and quisqualate.
24, 177-180.
8. Do K. Q., Herrling P. L., Streit P. and Cuenod M. (1988) Release of neuroactive substances: homocysteic acid as an endogenous agonist of the NMDA receptor. J. neural Transm. 72, 185-190. 9. Do K. Q., Mattenberger M., Streit P. and Cuenod M. (1986) In vitro release of endogenous excitatory sulphur containing amino acids from various rat brain regions. J. Neurochem. 46, 779-786.
10. Geoffroy M., Lambolez B., Audinat E., Hamon B., Crepe1 F., Rossier J. and Kado R. T. (1991) Reduction of desensitization of the AMPA receptor by CNQX and DNQX. Molec. Pharmac. 39, 587-591. 11. Guthrie P. B., Brenneman D. E. and Neale E. A. (1987) Morphological and biochemical differences expressed in separate dissociated cultures of dorsal and ventral halves of the mouse spinal cord. Brain Res. 420, 313-323. 12. Hollmann M., Hartley M. and Heinemann S. (1991) Ca*+ permeability of KA-AMPA-gated glutamate receptor channels depends on subunit composition. Science 252, 851-853. 13. Ito S., Provini L. and Cherubini E. (1991) L-Homocysteic acid mediates synaptic excitation at NMDA receptors in the hippocampus. Neurosci. Let?. 124, 157-161. 14. Johnson J. W. and Ascher P. (1987) Glycine potentiates the NMDA response of mouse central neurones. Nature 325, 529-531. 15. Kniipfel T., Zeise M. L., Cuenod M. and Zieglgiinsberger W. (1987) L-Homocysteic acid but not L-glutamate is an endogenous N-methyl-D-aspartic acid receptor preferring agonist in rat neocortical neurons in vitro. Neurosci. Left. 81, 188-192. 16. Kumar K. N., Tilakaratne N., Johnson P. S., Allen A. E., Michaelis E. K. (1991) Cloning of cDNA for the glutamate-binding subunit of an NMDA receptor complex. Na:ure 354, 70-73. 17. Mayer M. L., MacDermott A. B., Westbrook G. L., Smith S. J. and Barker J. L. (1987) Agonist- and voltage-gated calcium entry in cultured mouse spinal cord neurons under voltage clamp measured using arsenazo III. J. Neurosci. 7, 3230-3244. 18. Mayer M. L. and Vyklicki L. Jr (1989) Concanavalin A selectively reduces desensitization of mammalian neuronal quisqualate receptors. Proc. n&n. Acad. 5’ci. U.S.A. 86, 141 I-1415. 19. Mayer M. L., VyklickL L. Jr and Westbrook G. L. (1989) Modulation of excitatory amino acid receptors by group IIB metal cations in cultured mouse neurones. J. Physiol. 415, 329-350. 20. Mayer M. L. and Westbrook G. L. (1984) Mixed-agonist action of excitatory amino acids on mouse spinal cord neuiones under voltage clamp. J. Phy&oZ. ?54, 29-53. 21. Modv I.. Salter M. W. and MacDonald J. F. (1988) Reauirement of NMDA receDtor/channels for intracellular high-energy phosphates and the extent of intranduron~l cakium buffering in cultured-mouse hippocampal neurons. Neurosci. Lett. 93, 73-78.
22. Moriyoshi K., Masu M., Ishii T., Shigemoto R., Mizuno N., Nakanishi S. (1991) Molecular cloning and characterization of the rat NMDA receptor. Nature 354, 31-37. 23. Olverman H. J., Jones A. W., Mewett K. N. and Watkins J. C. (1988) Structure/activity relations of NMDA receptor ligands as studied by their inhibition of 3H-AP5 binding in rat brain membranes. Neuroscience 26, 17-31. 24. Patneau D. K. and Mayer M. L. (1990) Structure-activity relationships for amino acid transmitter candidates acting at N-methyl-D-aspartate and quisqualate receptors. J. Neurosci. 10, 2385-2399.
Membrane currents induced by L-homocysteic acid in mouse cultured hippocampal neurons
819
25. Schwarz S., Zhou G. Z., Katki A. G. and Rodbard D. (1990) L-homocysteate stimulates rH]MK-801 binding to the phencyclidine recognition site and is thus an agonist for the N-methyl-D-aspartate-operated cation channel. Neuroscience 37, 193-200. 26. Tang C. M., Dichter M. and Morad M. (1989) Quisqualate activates a rapidly inactivating high conductance ionic channel in hippocampal neurons. Science 243, 1474-1477. 27. Tang C. M., Dichter M. and Morad M. (1990) Modulation of N-methyl-D-aspartate channels by extracellular H+. Proc. natn. Acad. Sci. U.S.A. 87, 6445-6449. 28. Thomson A. M. (1986) A magnesium-sensitive post-synaptic potential in rat cerebral cortex resembles neuronal response, to N-methylaspartate. J. Physiol. 370, 531-549. 29. Thomson A. M., Radpour S. (1991) Excitatory connections between CA1 pyramidal cells revealed by spike triggered averaging in slices of rat hippocampus are partially NMDA receptor mediated. Eur. J. Neurosci. 3, 587-601. 30. Traynelis S. F. and Cull-Candy S. G. (1990) Proton inhibition of N-methyl-n-aspartate receptors in cerebellar neurons. Nature 345, 347-350. 31. Trussell L. O., Thio L. L., Zorumski C. F. and Fischbach G. D. (1988) Rapid desensitization of glutamate receptors in vertebrate central neurons. Proc. nafn. Acad. Sci. U.S.A. 85, 2834-2838. 32. Vlachova V., Vyklicky L., Vyklicky L. Jr and VyskoEil F. (1987) The action of excitatory amino acids on chick spinal cord neurones in culture. J. Physiol. 386, 425-438. 33. Vyklicky L. Jr, Beneveniste M. and Mayer M. L. (1990) Modulation of N-methyl-D-aspartic acid receptor desensitization by glycine in cultured mouse hippocampal neurones. J. Physiol. 428, 313-331. 34. Vyklickjl L. Jr, Patneau D. K. and Mayer M. L. (1991) Modulation of excitatory synaptic transmission by drugs which
reduce desensitization at AMPA/kainate receptors. Neuron 7, 971-984. 35. Vyklicky L. Jr, Vlachovl V. and Kr%ek J. (1990) The effect of external pH changes on responses to excitatory amino acids in mouse hippocampal neurones. J. Physiol. 430, 497-517. 36. Waller S. J., Kilpatrick I. C., Chan M. W. J., Evans R. H. (1991) The influence of assay conditions on measurement of excitatory dibasic sulphinic and sulphonic alpha-amino acids in nervous tissue. J. Neurosci. Meth. 36, 167-176. 37. Yamada K. A. and Rothman S. M. (1991) D&oxide reversibly blocks glutamate desensitization and prolongs excitatory postsynaptic currents. Sot. Neurosci. Abstr. 17, 896. 38. Zeise M. L.. Knoofel T. and Ziealelnsberaer W. (1988) (+)-B-Parachloronhenvlalutamate selectively enhances the depolarizing’ response to L-homocysteic acid in neocortidai neurons of the rat: evidence for a specific uptake system. Brain Res. 443, 373-376. 39. Zorumski C. F., Yang J. and Fischbach G. D. (1989) Calcium-dependent, types of glutamate receptors. Cell. molec. Neurobiol. 9, 95-104. (Accepted 7 January 1992)
slow desensitization distinguishes different
NeuroscienceVol. 48, No. 4, pp. 813819, 1992 Printed in Great Britain
MEMBRANE CURRENTS INDUCED BY L-HOMOCYSTEIC ACID IN MOUSE CULTURED HIPPOCAMPAL NEURONS L. VYKLICK+* and V. VLACHOVA Institute of Physiology, Czechoslovak Academy of Sciences, Videdska 1083, 142 20, Prague 4, Czechoslovakia Abstract-The concentration-response relationship of membrane currents induced by L-homocysteic acid was studied on mouse embryonic hippocampal neurons in culture (a = 56). In the majority of neurons two phases in the dose-response relationship could be distinguished. The first was characterized by responses to 3-100 PM L-homocysteic acid which desensitized with a time-constant > 1 s in a concentration-dependent manner and were antagonized by 30 p M n-L-2-amino-5phosphonovaleric acid indicating activation of the N-methyl-D-aspartate receptors. At higher concentrations of L-homocysteic acid this component was strongly depressed. The second phase was characterized by sustained responses that were concentration-dependent (1 mM L-homocysteic acid maximum concentration tested) and were not blocked by p-L-2-amino-5-phosphonovaleric acid indicating activation of non-N-methyl-D-aspartate receptors. Eight neurons did not exhibit these two-phase characteristics in the concentration-response relationship at the beginning of the recording. The magnitude of responses to L-homocysteic acid was positively related to concentration and the responses were partially blocked by D-L-2-amino-5-phosphonovaleric acid. In these neurons, however, repeated applications of L-homocysteic acid at concentrations 30pM up to 300 PM resulted in a long-lasting, three- to four-fold increase of the membrane current. This increase was completely blocked by D-L-2-amino-5-phosphonovaleric acid (SO-100 p M) suggesting that it was produced by activation of receptors. It is suggested that the long-lasting increase of the membrane current after repeated L-homocysteic acid application reflects long-term changes in the mechanisms involved in controlling N-methyl-o-aspartate
receptors from an intracellular site in which Ca*+ plays an important but not an exclusive role.
L-homocysteic acid (L-HCA) is known for its strong excitatory action on central neurons of vertebrates from the pioneering studies of Curtis and Watkins in the early sixties. sg6Interest in the mechanisms of its action has increased after it was demonstrated that L-HCA can be released from brain slices by high concentrations of K+ in a Caz+-dependent manner which raised the possibility that it might be an endogenous transmitter. *s9However, the localization of the HCA-like immunoreactivity predominantly in glial elements4 and the controversy about the actual levels of L-HCA in the central nervous system36 raised a problem in accepting this idea. It had been shown earlier that the responses to L-HCA in cultured central neurons exhibit a negative slope conductance in the presence of Mg2+ at negative membrane potential$” and that the effects of L-HCA are blocked by Q-phosphonatesr5 which suggested that L-HCA acts at N-methyl-D-aspartate
(NMDA) receptors. In a detailed study on structure-activity relationships for a number of excitatory amino acid transmitter candidates in which agonist action on NMDA and non-NMDA receptors was examined in isolation without use of antagonists, it was shown that L-HCA selectively activates the NMDA receptors only over a narrow concentration range and that it is also a potent quisqualate receptor agonist. 24These findings are in good agreement with the results of the ligand binding experiments which demonstrated that the binding potency of L-HCA for NMDA and kainate receptors is only slightly lower than that of L-glutamate.23,25 The suggestion that L-HCA might be an endogenous transmitter has recently been corroborated by the finding that the selective uptake blocker of L-HCA, fl-P-chlorophenylglutamate (Chlorpheg),’ potentiates L-HCA- but not L-glutamate-induced responses38 and increases synaptic responses induced in CA1 hippocampal neurons by Schaffer collateral stimulation.” These findings also raised the possibility that L-HCA might be involved in self-potentiation of responses to NMDA receptor activation observed on neurons in hippocampal slices.2*28,29 The present study was designed to test the effects of L-HCA on cultured hippocampal neurons to explore whether it may induce long-lasting changes in NMDA receptors.
*To whom correspondence should be addressed. AP-5, D-L-2-amino-5-phosphonovaleric acid; BAPTA, 1,2-bis(o -aminophenoxylethane)-N, N, N’, N’tetra-acetic acid; ECS. extracellular solution: EGTA, ethyleneglycol-bis-(ta-aminoethyl-ether)-N,N,N’,N’~ tetra-acetic acid; HEPES, N-2-hydroxyethylpiperazineN’-2-ethanesulphonic acid; L-HCA, L-homocysteic acid; MEM, minimum essential media; NMDA, N-methyl-naspartate.
Abbreviations:
813
814
L. ~¢VKL]CK~'and V. VLACHOV.~ EXPERIMENTAL PROCEDURES
A
Cell culture
L-HCA
The experiments were performed on primary cultures of hippocampal neurons from embryonic mouse (BALB/c, Institute of Physiology, CAS, Prague). Nerve cell cultures, were prepared by dissociating the hippocampi of 16-17-dayold embryos and plating the suspension onto confluent hippocampal glial cell feeder layer cultures. The mothers were killed by cervical dislocation prior to removal of the fetuses. The growth medium contained minimum essential medium (MEM), 5% horse serum, and a nutrient supplement containing transferrin, insulin, selenium, corticosterone, triiodothyronine, progesterone and putrescine. ~t
Recording and perfusion techniques Experiments were performed at room temperature (22-23°C) seven to 11 days after cultures had been plated. Solitary neurons of body diameter 15-20/~m were usually selected for examination. Whole-cell membrane current recordings were carried out with a List EPC-7 amplifier. The series resistance of recording pipettes was usually around I0 MfL Responses were recorded on a pen recorder and stored on video tape after digital sampling at 20 kHz for later evaluation. The system for fast superfusion of the neurons was in principle similar to that introduced by Johnson and Asher t4 and further elaborated in other laboratories. 24'33'35It consisted of a planar array of eight parallel glass tubes 0.4 mm in diameter through which solutions were driven by a peristaltic pump. The orifice of one of the tubes was placed about I00/zm above the soma of a selected neuron which allowed exchange of the solutions with a time-constant of 50-80 ms. The step motor for changing the position of the tubes and the time of opening the solenoid valves which directed the flow of the solution either onto neurons or back to reservoirs was controlled by a microcomputer programmed to perform a sequence of applications. The culture dishes, 3 cm in diameter, were independently perfused by extracellular solution (ECS; 1 ml/min).
Experimental solutions The ECS contained (in mM) NaC1 160, KCI 2.5, CaC12 1, MgC12 2, HEPES 10, Glucose 10; pH was adjusted to 7.3 with NaOH. In Mg-free solutions, MgC12 was omitted. The pipette solution contained (in mM): CsC1 140, KC1 2.5, CaCI 2 1, MgC12 2, EGTA 5, HEPES 10, pH was adjusted to 7.3 with CsOH. Cells were washed by a rapid stream of ECS before and after each drug application for 5 s. L-HCA was applied in Mg-free extracellular solution with 3 p M glycine. Picrotoxin (10 -4 M) was added to all solutions to block GABA activated channels. Tetrodotoxin (2.10 -6 M) was added when synaptic activity or spikes became a problem. L-HCA, NMDA, D-L-2-amino-5-phosphonovaleric acid (AP-5) and glycine were provided by TocrisNeuramin.
RESULTS
Concentration-response relationship In order to prevent excessive disuse, the application o f L-HCA was limited to 5 s. The c o n c e n t r a t i o n response relationship of the whole-cell responses to L-HCA in the Mg-free solution containing 3 # M glycine exhibited two phases (Fig. 1A). T h e first phase was characterized by responses with the threshold slightly below 3/~M. The c o n c e n t r a t i o n o f L-HCA at which the largest responses were observed varied between 30 a n d 300/~M. In this range o f
~ ,~'~ ~ ~o ~ B ~
~
~
~
~
~
~
-50mY
15s
Fig. I. Series of whole-cell responses to ascending concentrations of L-HCA in a cultured hippocampal neuron. L-HCA in concentrations 3, 10, 30, 100, 300/LM and 1 mM in Mg-free extracellular solution with 3 # M glycine was rapidly superfused around the neuron for 5-s periods which were separated by 10-s intervals during which the neuron was washed with an extracellular solution containing 2 mM MgC1z. The second application of 30/tM L-HCA at the end of the series served as a control, showing that no serious changes of the responses had occurred during recording. (A) The first series of recording from the neuron. (B) The seventh series recorded about half an hour later. Membrane potential, - 5 0 mV. The bars above the records indicate time of L-HCA application.
L-HCA c o n c e n t r a t i o n s the responses declined with a slow time-constant of > 1 s. Five seconds o f application was n o t usually long e n o u g h for the declining phase of the responses to reach the plateau. T h e value o f the t i m e - c o n s t a n t calculated from digitalized responses varied between 1-2.2 s. A t c o n c e n t r a t i o n s higher t h a n t h a t needed to induce the m a x i m u m response, a m a r k e d decrease o f the responses was observed; the sustained c o m p o n e n t of the responses, however, became m o r e p r o m i n e n t . The second phase o f the c o n c e n t r a t i o n - r e s p o n s e relationship was d o m i n a t e d by a sustained comp o n e n t which increased in a m p l i t u d e with the L - H C A concentration. The m a x i m u m c o n c e n t r a t i o n used in o u r experiments were 600/~M or 1 m M L-HCA. After several series o f L - H C A applications, r u n d o w n o f the m e m b r a n e current was usually observed after 10-15 min. This was characterized by a higher conc e n t r a t i o n of L - H C A needed to induce the t h r e s h o l d response a n d by a m a r k e d decrease o f the declining c o m p o n e n t o f the responses to L - H C A (Fig. 1B). T w o phases in the c o n c e n t r a t i o n - r e s p o n s e relationship were, however, preserved d u r i n g the whole time o f recording from the cell. Usually f r o m five u p to 10 series o f recordings were m a d e f r o m one n e u r o n . In view o f previous studies 24"25 the m o s t likely i n t e r p r e t a t i o n o f these results was t h a t L - H C A is a mixed agonist at g l u t a m a t e receptors with a preference for N M D A - o p e r a t e d channels. T h e initial desensitizing c o m p o n e n t o f the responses could thus be a t t r i b u t e d to activation o f N M D A receptors, while the sustained c o m p o n e n t induced at high concentrations of L-HCA would be generated by cationic currents induced by activation o f n o n - N M D A receptors o f the kainate type.
815
Membrane currents induced by L-homocysteic acid in mouse cultured hippocampal neurons
A 1
f
L-HCA
h
B
-80 mV
c
3.104 w1
D lO-’ wa!s I-
-I”!-
3.10” -
>.
I”
I
XOpA
1Sr
Fig. 2. Antagonistic effects of AP-5 on responses to L-HCA. (A) L-HCA in a Mg-free extracellular solution containing 3 PM glycine was applied either alone (10, 100 and 3OOpM) or together with 30 bM AP-5 (thick bars). (B) Three-hundred micromolar L-HCA (thick bar) was interposed between 10 uM L-HCA an&cations. (C) Three-hundred L-HCA (thick bar) was in&posed be&e&r 1OOpM L-HCA applications. The bars above the records indicate the time of L-HCA applications.
The eflects of D-L-Zamino-S-phosphonovaleric acid on responses to L-homocysteic-acid To distinguish clearly between the NMDA and non-NMDA components of the responses to L-HCA we used AP-5, a competitive antagonist of the NMDA receptors. Each concentration of L-HCA
tested was applied either alone or with AE-5. We found that the declining component of the responses to L-HCA but not the sustained component induced by higher concentrations of L-HCA was strongly depressed when L-HCA was applied together with 30 PM AP-5 (Fig. 2A). A marked depression of the responses to L-HCA without AR5 was, however, observed when the L-HCA concentration was increased to 300 or 600 PM. These findings demonstrate that the declining component of the responses to L-HCA is produced by activation of NMDA receptors and, in agreement with other studies,r@5 indicates a higher affinity of L-HCA for NMDA receptors than for non-NMDA receptors. A seemingly paradoxical concentration-response relationship was observed when 300 pM L-HCA was interposed between the responses to 100 PM concentration (Fig. 2C) but not when it was interposed between the responses induced by 10 FM L-HCA (Fig. 2B). The suppressing effects of high concentrations of L-HCA on the AP-5-sensitive component were present both on the negative and the positive membrane potentials (Fig. 3A,B). In agreement with previous studies which indicated that desensitization of the responses to NMDA is concentration-dependent’~33~39 we found that the membrane current induced by L-HCA at concentrations which produced desensitization markedly increased under experimental conditions when LHCA concentration around the neuron was allowed to decrease slowly and not rapidly by washing the neuron with ECS. This was achieved by discontinuing
Fig. 3. Voltage-independence but concentration-dependence of the desensitization of the responses to L-HCA. (A, B) Responses to 300 nM L-HCA interposed between responses to 30 ,uM L-HCA at membrane potential -80 mV and +20 mV, respectively. (C, D) Responses to 30 and 100 PM L-HCA respectively. L-HCA was applied during the period indicated -o-; wocw indicates the paradigm of L-HCA application. w, The neuron was washed by extracellular solution. At the interval -o- the position of the array of the tubes was changed so that the tube containing L-HCA was facing the neuron and the valve which permitted flow of L-HCA on the neuron was open. During -c- application of L-HCA was discontinued and the tube was removed from the neuron. At -w- the array of the tubes returned to the original position and the neuron was washed by a fast stream of ECS. The diiference between -c- and -w- lies in that in c- the stream of L-HCA was only halted so that its concentration was slowly decreasing by diffusion and bath perfusion while in -w-, L-HCA was halted and rapidly removed around the neuron by washing it with extracellular solution.
application of L-HCA during the declining phase of the response and by placing the barrel away from the neuron. At this period the concentration of L-HCA decreased by diffusion and slow bath perfusion. The magnitude of the response immediately began to increase, as can be seen in Fig. 3C and D. The membrane current returned to the base line immediately after the array of tubes for application of the solutions returned to its original position and the neuron was washed with a fast stream of the extracellular solution (-w-). Long-lasting increase of the responses after repeated L-homocysteate applications
In eight neurons out of 56 which were exposed to repeated series of L-HCA applications, two phases in the dose-response relationship were not observed at the beginning of the recording from the neuron. In the first series of L-HCA applications, the magnitude of the responses was positively related to the concentration and the responses were partially blocked by AI’-5 (Fig. 4A). In the second or the third series of L-HCA applications the membrane current induced by a certain concentration of L-HCA (30-300 PM) began to increase. The increase. developed abruptly during L-HCA application and resembled breaking down of the cell membrane. The increased membrane current, however, returned to the resting level rapidly after L-HCA application had been completed and the neuron had been washed with the extracellular
816
L.
VYKLICKL
and V.
VLACHWA
L-HCA +G3pM
3 IO-' -
3 IO-~
-
-
-
-5OmV
B +50 mV
1
lOOpA
15s
I
0 “1
& 4 +
Fig. 4. Whole-cell responses to repeated applications of L-HCA. Three series of responses to L-HCA at concentration 10, 100 and 3OOpM without and with AP-5 as indicated. Ten micromolar AP-5 was applied together with 10 PM L-HCA; 30pM AP-5 was applied together with L-HCA 100 and 3OOpM. (A) tire first series, (B) the third series, and (C) the last (seventh) series of recording from the neuron. Appreciable leakage current was observed during recordings in the last series which explains that all responses are smaller. Membrane potential -50 mV. solution (Fig. 43). The magnitude of subsequent responses to higher concentrations of L-HCA in that series of applications were similar to those in the previous series. Usually not more than 2min were allowed between the series of applications. Once the increase in the response to a certain concentration of L-HCA developed, it remained unchanged during the whole time of recording from the neuron, i.e. until it was lost or discarded (Fig. 4C). Such a long-lasting increase in the responses to L-HCA was completely blocked by AP-5 (30 PM-100 PM) suggesting that it is produced by activation of NMDA receptors. The development of a long-lasting increase of the membrane current after repeated expositions of the neuron to L-HCA was not voltage-dependent, as is demonstrated in Fig. 5. In this neuron, some increase of the response could already be observed at a membrane potential of - 50 mV in the first series of applications at the end of the second response to 300 ~1M L-HCA. A marked increase of the response, however, was observed in the second series when the membrane potential was changed to +50 mV (Fig. SB). The increase of the membrane current lasted to the very end of the recording from the neuron, about 30min later, and it could be completely blocked by AP-5 during this time (Fig. 5C).
The major findings in our study on cultured hippocampal neurons are the concentration-dependent desensitization of the NMDA component of the
L-HCA
i&s--
3- 10-8-
&*
---
3- 10-S
n
3 10.’
+G~IIM
Fig. 5. Three series of responses to 30 and 300 PM L-HCA without and with AP-5 (50 and IOOpM, respectively). (A) The first series of recordings from the neuron at - 50 mV. (B, C) The third and the sixth series of recordings from the same neuron at membrane potential +50mV.
responses to L-HCA and the long-lasting increase of the responses to L-HCA after repeated applications which was observed in some neurons. Desensitization of N-methyl-o-aspartate during the action of L-homocysteate
receptors
A large range of concentrations of L-HCA and other excitatory amino acids were used by Patneau and Mayeti4 to quantify concentration-response relationships under experimental conditions which allowed us to study the generation of the membrane currents either through the NMDA- or quisqualateoperated channels in isolation without use of antagonists. To avoid Ca*+-dependent desensitization of NMDA receptors, they used a low concentration of Ca2+ (0.2 mM) in the extracellular solution and a very potent calcium chelating agent, BAPTA, in the pipette solution. On the other hand, 1 mm Ca2+ in ECS and a less potent chelating agent, EGTA, in the pipette solution were used in our experiments which can be expected to allow appreciable increases in intracellular Ca*+ during massive activation of NMDA channels. ” These differences in the composition of experimental solutions seem to be the most likely explanation why the NMDA component of the responses to L-HCA desensitized in our experiments. Admixture of Mg*’ from the ECS containing 2 mM Md+ to the Mg-free solution in which L-HCA was applied can be excluded as a possibility for the fade of the responses in our experiments because similar results were obtained at positive membrane potentials or when ECS without M&+ was used for a rapid superfusion of the neuron. Moreover, the
Membrane currents induced by L-homocysteic acid in mouse cultured hippocampal neurons amplitude of the desensitized response increased when application of L-HCA was discontinued which gave a better chance of mixing the Mg-free and Mg-containing solutions (see later). There are two reasons which indicate that the component of the responses to L-HCA which showed a marked desensitization is really accounted for by activation of the NMDA and not the quisqualate receptors. First, the rate of desensitization is relatively slow with a time-constant characteristic for the whole-cell responses to NMDA,32,33*39 while the rate of desensitization of quisqualate receptors is very fast with a time-constant of several milliseconds.26~31 Second, the desensitizing component of responses to L-HCA was effectively antagonized by AP-5, a competitive antagonist of NMDA receptors. At low concentrations of L-HCA, however, the sustained component of the responses was also partially blocked by AP-5 indicating that NMDA channels contribute to its generation. On the other hand, the sustained component of the responses to high L-HCA concentrations was insensitive to AP-5 suggesting that it is induced by activation of non-NMDA receptors. There are at least two possibilities which should be considered for explaining why the NMDA component of the responses to L-HCA was not detected at high concentrations. The first is that L-HCA, in addition to agonist recognition sites of glutamate receptors, binds to a low-affinity site on the outside of the NMDA receptor complex from where it can be. inhibited with an action similar to that of Zn or protons.19,27*~~3S The main difference between L-HCA and L-glutamate lies in the presence of an atom of sulphur in L-HCA which might be suspected to have modulatory effects on NMDA receptors. However, despite its appeal, this idea does not seem likely because no desensitization even to high L-HCA concentrations was observed when a low concentration of Ca*+ was used in the extracellular solution.24 The second possibility is that the kinetics of desensitization with high L-HCA concentrations exceeds the rate of drug application of the system used in our experiments. This explanation is favoured by the evidence that desensitization of NMDA receptors is dependent on agonist wncentration’*39 and by our finding that the magnitude of the responses to L-HCA increased when its concentration around the neuron was slowly decreasing (Fig. 3C, D). The relatively rapid rate of desensitization of the NMDA component of the responses to high L-HCA concentration can explain the seemingly paradoxical finding that a higher agonist concentration can produce a smaller response than a lower concentration, similar to findings described for quisqualate receptors.tO Apparently such a paradox can be expected to occur at all receptors which exhibit a bell-shaped dose-response relationship due to agonist concentration-dependent desensitization. Its mechanisms have already been
817
analysed at the single channel level of the nicotinic acetylcholine receptor.3 Long-lasting increase of the membrane current during repeated L-homocysteate application It does not seem likely that artifacts in recording whole cell currents such as a sudden change of the series resistance or transient release of the gigaseal could account for the long-lasting increase of the membrane current which was observed in some neurons after repeated applications of L-HCA because recordings remained stable, without marked changes which could not be explained by irreversible rundown due to lack of MgATP in the pipette solution.*’ The increase in the membrane current induced by repeated applications of L-HCA was observed in both pyramidal and granule-like neurons. It was, however, essential that NMDA receptors were abundantly developed, i.e. that the neurons exhibited large membrane currents to NMDA which was always tested in one dish of a new batch of cultures. Development of the long-lasting increase of the membrane current was observed only under the conditions in which the concentration of L-HCA was sufficient to activate massively both the NMDA and the non-NMDA receptors and L-HCA was applied repeatedly. The finding that the increase of the membrane current was completely blocked by AP-5, however, indicates that it is produced exclusively by activation of the NMDA channels. Once the increase of the membrane current to L-HCA developed, it lasted the whole time of recording from the cell. Our experiments, however, cannot answer the question on the reversibility unequivocally, because the testing for its presence represented at the same time a stimulus for its development. Two alternatives should be considered in explaining mechanisms possibly involved in this long-lasting increase of the membrane current: relief of desensitization of NMDA receptors and activation of a subset of dormant NMDA receptors, There are at least two different mechanisms which underlie desensitization of NMDA receptors. The calcium-mediated desensitization in which Ca*+ directly or indirectly, via activating cytoplasmic enzymes regulates the NMDA receptors from the interior of the neuron17s39 and the glycine-sensitive desensitization which modulates activity of the NMDA receptors from the extracellular side.33 The mechanisms suggested to underlie the reduction of the desensitization of glutamate receptors by diazoxide,” aniracetam, wheatgerm agglutini# and concanavalin A’* do not seem to apply to NMDA receptors because they concern fast desensitization of the quisqualate receptors. One possibility for explaining the development of a sudden increase of the membrane current through NMDA channels is that intracellular calcium controls processes of desensitization only to a critical
818
L.VYKLICKC~~~
concentration which, when exceeded, results in blocking the mechanisms, possibly cytoplasmic enzymes, involved in regulation of NMDA receptor desensitization. Our finding that a critical but not a higher concentration of L-HCA induced the long-lasting increase of the responses would be compatible with this idea when assuming that the enzymes for regulating NMDA receptors escape washing out during whole-cell recording and that the rate of desensitization on the extracellular side depends on agonist concentration and effectively blocks the flow of Ca* + through NMDA channels. Although the NMDA receptor has recently been cloned’6,22 there is still not enough information which
V.VLACHOVA
would allow meaningful speculation whether thcrc might exist a subclass of dormant NMDA receptors which could be activated only after repeated I -HCA applications and thus produce a long-lasting increase in the membrane current. However, in view q>f the complexity of the non-NMDA receptors’-’ it can be envisaged that the diversity of NMDA receptors ~111 still be larger in order to comply with the plastic processes in which they are likely to be involved.
Acknowledgements-This
work was supported by GA CSAS (71114). The authors thank Dr P. Hnik for reading the manuscript.
REFERENCES 1. Clark G. D., Clifford D. B. and Zorumski C. F. (1990) The effect of agonist concentration,
membrane voltage and calcium on N-methyl-D-aspartate receptor desensitization. Neuroscience 39, 787-797. 2. Collingridge G. L., Herron C. E. and Lester R. A. J. (1988) Frequency-dependent N-methyl-o-aspartate receptormediated synaptic transmission in rat hippocampus. .I. Physiol. 399, 301-312. 3. Colquhoun D. and Ogden D. C. (1988) Activation of ion channels in the frog end-plate by high concentrations of acetylcholine. J. Physiol. 395, 131-159. 4. Cubnod M., Do K. Q., Streit P. (1990) Homocysteic acid as an endogenous excitatory amino acid. Trenlls pharmac. Sci. 11, 477-478. 5. Curtis D. R., Phillis J. W. and Watkins J. C. (1961) Actions of amino acids on the isolated hemisected spinal cord of the toad. Br. J. Phurmac. 16, 262-283. 6. Curtis D. R. and Watkins J. C. (1963) Acidic amino acids with strong excitatory actions on mammalian neurons. J. Physiol. 166, I-14. 7. Davies J., Francis A. A., Oakes D. J., Shardown M. J. and Watkins J. C. (1985) Selective potentiating effect of
B-p-chlorophenylglutamate Neurophannacology
on responses induced by certain sulphur-containing excitatory amino acids and quisqualate.
24, 177-180.
8. Do K. Q., Herrling P. L., Streit P. and Cuenod M. (1988) Release of neuroactive substances: homocysteic acid as an endogenous agonist of the NMDA receptor. J. neural Transm. 72, 185-190. 9. Do K. Q., Mattenberger M., Streit P. and Cuenod M. (1986) In vitro release of endogenous excitatory sulphur containing amino acids from various rat brain regions. J. Neurochem. 46, 779-786.
10. Geoffroy M., Lambolez B., Audinat E., Hamon B., Crepe1 F., Rossier J. and Kado R. T. (1991) Reduction of desensitization of the AMPA receptor by CNQX and DNQX. Molec. Pharmac. 39, 587-591. 11. Guthrie P. B., Brenneman D. E. and Neale E. A. (1987) Morphological and biochemical differences expressed in separate dissociated cultures of dorsal and ventral halves of the mouse spinal cord. Brain Res. 420, 313-323. 12. Hollmann M., Hartley M. and Heinemann S. (1991) Ca*+ permeability of KA-AMPA-gated glutamate receptor channels depends on subunit composition. Science 252, 851-853. 13. Ito S., Provini L. and Cherubini E. (1991) L-Homocysteic acid mediates synaptic excitation at NMDA receptors in the hippocampus. Neurosci. Let?. 124, 157-161. 14. Johnson J. W. and Ascher P. (1987) Glycine potentiates the NMDA response of mouse central neurones. Nature 325, 529-531. 15. Kniipfel T., Zeise M. L., Cuenod M. and Zieglgiinsberger W. (1987) L-Homocysteic acid but not L-glutamate is an endogenous N-methyl-D-aspartic acid receptor preferring agonist in rat neocortical neurons in vitro. Neurosci. Left. 81, 188-192. 16. Kumar K. N., Tilakaratne N., Johnson P. S., Allen A. E., Michaelis E. K. (1991) Cloning of cDNA for the glutamate-binding subunit of an NMDA receptor complex. Na:ure 354, 70-73. 17. Mayer M. L., MacDermott A. B., Westbrook G. L., Smith S. J. and Barker J. L. (1987) Agonist- and voltage-gated calcium entry in cultured mouse spinal cord neurons under voltage clamp measured using arsenazo III. J. Neurosci. 7, 3230-3244. 18. Mayer M. L. and Vyklicki L. Jr (1989) Concanavalin A selectively reduces desensitization of mammalian neuronal quisqualate receptors. Proc. n&n. Acad. 5’ci. U.S.A. 86, 141 I-1415. 19. Mayer M. L., VyklickL L. Jr and Westbrook G. L. (1989) Modulation of excitatory amino acid receptors by group IIB metal cations in cultured mouse neurones. J. Physiol. 415, 329-350. 20. Mayer M. L. and Westbrook G. L. (1984) Mixed-agonist action of excitatory amino acids on mouse spinal cord neuiones under voltage clamp. J. Phy&oZ. ?54, 29-53. 21. Modv I.. Salter M. W. and MacDonald J. F. (1988) Reauirement of NMDA receDtor/channels for intracellular high-energy phosphates and the extent of intranduron~l cakium buffering in cultured-mouse hippocampal neurons. Neurosci. Lett. 93, 73-78.
22. Moriyoshi K., Masu M., Ishii T., Shigemoto R., Mizuno N., Nakanishi S. (1991) Molecular cloning and characterization of the rat NMDA receptor. Nature 354, 31-37. 23. Olverman H. J., Jones A. W., Mewett K. N. and Watkins J. C. (1988) Structure/activity relations of NMDA receptor ligands as studied by their inhibition of 3H-AP5 binding in rat brain membranes. Neuroscience 26, 17-31. 24. Patneau D. K. and Mayer M. L. (1990) Structure-activity relationships for amino acid transmitter candidates acting at N-methyl-D-aspartate and quisqualate receptors. J. Neurosci. 10, 2385-2399.
Membrane currents induced by L-homocysteic acid in mouse cultured hippocampal neurons
819
25. Schwarz S., Zhou G. Z., Katki A. G. and Rodbard D. (1990) L-homocysteate stimulates rH]MK-801 binding to the phencyclidine recognition site and is thus an agonist for the N-methyl-D-aspartate-operated cation channel. Neuroscience 37, 193-200. 26. Tang C. M., Dichter M. and Morad M. (1989) Quisqualate activates a rapidly inactivating high conductance ionic channel in hippocampal neurons. Science 243, 1474-1477. 27. Tang C. M., Dichter M. and Morad M. (1990) Modulation of N-methyl-D-aspartate channels by extracellular H+. Proc. natn. Acad. Sci. U.S.A. 87, 6445-6449. 28. Thomson A. M. (1986) A magnesium-sensitive post-synaptic potential in rat cerebral cortex resembles neuronal response, to N-methylaspartate. J. Physiol. 370, 531-549. 29. Thomson A. M., Radpour S. (1991) Excitatory connections between CA1 pyramidal cells revealed by spike triggered averaging in slices of rat hippocampus are partially NMDA receptor mediated. Eur. J. Neurosci. 3, 587-601. 30. Traynelis S. F. and Cull-Candy S. G. (1990) Proton inhibition of N-methyl-n-aspartate receptors in cerebellar neurons. Nature 345, 347-350. 31. Trussell L. O., Thio L. L., Zorumski C. F. and Fischbach G. D. (1988) Rapid desensitization of glutamate receptors in vertebrate central neurons. Proc. nafn. Acad. Sci. U.S.A. 85, 2834-2838. 32. Vlachova V., Vyklicky L., Vyklicky L. Jr and VyskoEil F. (1987) The action of excitatory amino acids on chick spinal cord neurones in culture. J. Physiol. 386, 425-438. 33. Vyklicky L. Jr, Beneveniste M. and Mayer M. L. (1990) Modulation of N-methyl-D-aspartic acid receptor desensitization by glycine in cultured mouse hippocampal neurones. J. Physiol. 428, 313-331. 34. Vyklickjl L. Jr, Patneau D. K. and Mayer M. L. (1991) Modulation of excitatory synaptic transmission by drugs which
reduce desensitization at AMPA/kainate receptors. Neuron 7, 971-984. 35. Vyklicky L. Jr, Vlachovl V. and Kr%ek J. (1990) The effect of external pH changes on responses to excitatory amino acids in mouse hippocampal neurones. J. Physiol. 430, 497-517. 36. Waller S. J., Kilpatrick I. C., Chan M. W. J., Evans R. H. (1991) The influence of assay conditions on measurement of excitatory dibasic sulphinic and sulphonic alpha-amino acids in nervous tissue. J. Neurosci. Meth. 36, 167-176. 37. Yamada K. A. and Rothman S. M. (1991) D&oxide reversibly blocks glutamate desensitization and prolongs excitatory postsynaptic currents. Sot. Neurosci. Abstr. 17, 896. 38. Zeise M. L.. Knoofel T. and Ziealelnsberaer W. (1988) (+)-B-Parachloronhenvlalutamate selectively enhances the depolarizing’ response to L-homocysteic acid in neocortidai neurons of the rat: evidence for a specific uptake system. Brain Res. 443, 373-376. 39. Zorumski C. F., Yang J. and Fischbach G. D. (1989) Calcium-dependent, types of glutamate receptors. Cell. molec. Neurobiol. 9, 95-104. (Accepted 7 January 1992)
slow desensitization distinguishes different
