0306-4522/92 $5.00 + 0.00 Perpmon Press Ltd 0 1992 IBRO
Neuroscience Vol. 49, No. 3, pp. 571-576, I992 Printed in Great Britain
DESENSITIZATION CULTURED
OF GABA-INDUCED CURRENTS RAT HIPPOCAMPAL NEURONS
IN
D. J. OH and M. A. DICHTER* Department of Neurology, University of Pennsylvania School of Medicine and Graduate Hospital, Philadelphia, PA 19104, U.S.A. Ahatraet-The basic characteristics of desensitization of the GABA, receptor were investigated in cultured rat hippocampal neurons (three days to four weeks in vitro) using whole cell patch clamp techniques. GABA at 10-500 pM was perfused on to neurons for 30 or 60 s, with 60 s intervals of wash with control
bath solution between perfusions. Desensitization, evaluated by peak-to-plateau ratio and time constants of current decay (7), was dosedependent and culture age-dependent. Desensitization was observed as early as three days in culture, the earliest time tested. At all ages, higher concentrations of GABA induced both larger and faster desensitization. Desensitization was markedly voltage-dependent and decmased with depolarization; peak-to-plateau ratio went from 6.3 to 1.4 and 7 went from 4.6 to 26.8 s when holding
potentials were changed from - 80 mV to + 30 mV. Low concentrations of GABA (l-2 PM) perfused for 2-60 s, which did not induce any current, had no effect on the maximal response nor desensitization produced by a subsequent application of 100p M GABA. This finding suggests that GABA receptors were not desensitized without first being activated.
The gamma-aminobutyric acid type A (GABA,) receptor, with its integral chloride ion channel, is considered to be the major inhibitory neurotransmitter receptor in the mammalian brain. GABAmediated inhibition shapes sensory input, motor output, and maintains a “basal” level of excitability in the mammalian cortex; any diminution of GABAmediated inhibition can lead to excessive excitability, which often manifests itself as seizure activity. The GABA receptor complex contains a number of modulatory sites, most of which are capable of being affected by exogenous agents (e.g. benzodiazepines, barbiturates, etc.). Although it has been hypothesized that endogenous agents also exist which can affect these sites, their identities have remained elusive to date. GABA-mediated inhibition can also be modulated by presynaptic mechanisms. GABA, receptors exist on GABA terminals and activation of these receptors can suppress the release of GABA, thereby downregulating the synapse. Other presynaptic receptors may also function in this way. GABA,-mediated inhibition can also be autoregulated by a process of desensitization. It is known that continued application of GABA to neurons produces a decrease in the GABA-induced current, due to both a change in the transmembrane chloride gradienti4J9 *To whom correspondence should be addressed. Abbreuiafions: BAPTA, 1,2-bis(2aminophenoxy)ethaneN,iV,N,N-tetra-acetic acid; DMEM, Dulbecco’s modified Eagle’s medium; HEPES, [N-(2-hydroxyethyl)piperaxine-N-(2-ethanesulfonic acid)]; IPSC, inhibitory postsynaptic current; PPR, peak-to-plateau ratio; 7, exponential current decay. 571
and a decrease in the induced conductance, a true receptor desensitization. This latter phenomenon has been observed electrophysiologically in neocorticallo*” and hippocampal neurons,6*16J7,Z and in rat brain receptors expressed in oocytesnJ* and biochemically in synaptosome preparations.2 Previous work from our laboratory demonstrated that desensitization at the GABA, receptor in neocortical neurons in culture had interesting properties.“,” It was concentration-dependent and independent of the amount of current induced or of the charge transfer induced by a given concentration of GABA. The rate and extent of desensitization was also voltage-dependent; it occurred much more slowly at depolarized membrane potentials. In addition, desensitization occurred more rapidly at the single channel level, when patches of membrane were examined in isolation, and the voltage dependence of the desensitization was lost. The cellular and molecular mechanisms underlying desensitization at the GABA, receptor are not understood. The characteristics of desensitization of GABA, receptors are not the same in all regions of the mammalian CNS. In rat retinal ganglion cells, true receptor desensitization occurs but does not show any voltage dependence.2’ In this paper we describe the basic characteristics of desensitization and resensitization of the GABA, receptor in cultured rat hippocampal neurons, attempting to develop quantitative measures which can be utilized in subsequent studies of mechanisms. We characterize the development of desensitization, its voltage dependency and the issue of whether desensitization can occur by
572
L> j.
OH
concentrations of GABA below the threshold channel opening. EXPERIMENTAL
and
for
PROCEDURES
Cell culture
Primary rat hippocampal cultures were prepared as reported previously.’ Hippocampal cells were removed from 20-21-day-old fetuses and dissociated with 0.03% trypsin in Dulbecco’s modified Eagle’s medium (DMEM) for 40 min at 37°C. Cells were subsequently dispersed by trituration and plated onto 35 mm dishes containing polylysine-coated coverslips. Cultures were maintained with DMEM, 10% hyclone calf serum, and 10% Ham’s F-12 in a 7% CO, incubator at 37°C. Cytosine arabinoside (10m5M) was added at 7-10 days after initial culture to prevent glial proliferation. Neurons cultured for three days to four weeks were used in this study. Electrophysiological recordings
Recordings were made with a Dagan 8900 patch clamp amplifier using standard whole cell patch clamp methods’* at room temperature. Kimax electrodes were used. Electrode resistance ranged from 2 to 5 MR. Neurons were routinely held at -8OmV. To analyse the voltage dependency of desensitization, membrane potentials were changed from - 80 mV to - 30 mV and + 30 mV. GABA concentrations between 1 and 500pM were applied by pressure ejection from multibarrelled micropipettes (tip diameters 5-10 pm) positioned 25-50 pm from the cell. Between applications neurons were washed with control bath solution. GABA-induced current was recorded on a Gould strip chart recorder and a computer (Indec systems). The external bath solution consisted of 140mM NaCl, 3mM KCI, 2mM CaCl,, 1 mM MgCl,, 10 mM HEPES, 8 mM glucose, 5 mM 4-aminopyridine, 4 mM CoCI,, 5 pM tetrodotoxin, pH 7.2
M. A.
bCHTEK
with NaOH. The internal solution contained I05 mM C& I. I mM MgC12, IO mM HEPES, 7 mM BAPTA. 70/1M Irupeptin, 14 mM phosphocreatine. 35 U/ml creatlne phosphokinase, 5 mM ATP-Mg, pH 7.2 with CsOH. We chose this internal solution to prevent rundown.‘,0.‘0 Small differences of Cl- concentration between our internal and bath solutions did not have any effect on desensitizatiorb in our experiments. RESliLTS
Prolonged application of GABA to cultured hippocampal neurons induces a current which peaks within less than 1 s and stays elevated for only a short time. The current then decreases exponentially with a single time constant. This gradual decay, desensitization. was observed in almost all cells at various culture ages from three days to four weeks, except when very low concentrations of GABA were applied. This current decay is a true receptor desensitization, as it was seen in neurons recorded with equimolar Cl on both sides of the membrane and therefore could not have been due to a simple shift in the Cl equilibrium potential. In order to compare desensitization under different conditions and between cells, we measured the time constant of exponential current decay (5) and peak-to-plateau ratio (PPR) of induced current. Figure 1 shows a typical pattern of desensitization of GABA resDonses in one neuron at four different concentrations. Higher concentrations of GABA perfused on to the cell resulted in faster and more extensive desensitization. PPR and r were dosedependent (Fig. IB) in neurons from three days to
B n Peak ! Plateau 0
Time constant
(seconds)
25 3
4 J
50pM
i”-----,
500 JJM
JL
20-
5;
ol 5
1 .O nA
i 10
0 100 GABA concentration
P 1000 QIM)
I 15 seconds
Fig. 1. Concentration dependence of GABA desensitization. (A) Representative penwriter responses from one hippocampal neuron cultured for three weeks to different concentrations of GABA applied for 1 min. The extent and rate of desensitization increased with increasing GABA concentrations. Membrane potential was held at -80 mV. (B) Desensitization evaluated by peak-to-plateau ratio (squares) and time constants of current decay (circles). Membrane potential was held at -80 mV. GABA was perfused at 10, 50, 100, and 5OOpM on to three- to four-week-old cells. Note the dose-dependent relationship. Each point represents the means + S.D. of seven to eight cells. Analysis of variance demonstrated statistical significance between the low, intermediate and high concentrations for both measures of desensitization (P < 0.01).
513
GABA desensitization in hippocampal neurons four weeks in culture. Desensitization was often extensive enough that the plateau currents after 30-45 s were smaller at higher concentrations of
GABA. Qualitatively similar dose-dependent desensitization was noted in neurons as early as three days in culture (the earliest time point tested) and as late as four weeks in culture. At concentrations of GABA from 10 to lOOpM, desensitization was larger and faster in older cells than in younger cells (Fig. 2A, B). At 500 PM GABA, however, the desensitization was similar at all culture ages, although
peak currents induced by 500 PM GABA increased significantly (P < 0.0001) from three-day-old cells (1280 + 550pA, mean f SD., N = 5) to three- to four-week-old cells (4040 + 530, N = 8). Desensitization of GABA-induced currents was markedly voltage-dependent.lO~‘l Figure 3A illustrates desensitization induced by 100 FM GABA at three different membrane potentials. Desensitization was smaller and slower as the membrane was depolarized and almost no desensitization was observed at +30 mV. PPR and t at three different membrane potentials are shown in Fig. 3B. The effect of mem-
0
3days
q
1 week
gxI 2 weeks
q
3-4 weeks
II]
3 days
g
1 week
q q
2 weeks
d 5
0
’
1OpM
’
50pM ’ 100~1M GABA concentration
70 3
p 50 8 3 40
’
5OO)tM
’
3-4 weeks
E Q 30 LO E i= 10 0 ’
~OJJM
’
50pM GABA concentration
Fig. 2. Development of desensitization from three-day- to three- to four-week-old cells to different concentrations of GABA applied for 1 min. Desensitization (A: peak-to-plateau ratio; B: T) was more prominent in older cells than in younger cells, except at high concentrations of GABA. Membrane potential was -8OmV. When no desensitization was seen, the time constant was estimated at 60s for calculation purposes. Each value in the figure is the mean f SD. of five to 14 cells. The difference between values at three days and three to four weeks was statistically significant for 10-100 PM GABA (P < 0.001) but not at the highest concentration tested. (The SD. for three-day-old neurons and 10 PM GABA was 0, as no desensitization was noted.)
574
D.
J.
OH and
A
M. A.
DICHTER
B
0 Peak /Plateau n Time constant ( seconds ) -6
-80
mV
L -6
-30
mV
63 -4 i? 3 36 -2 0 -1
2.0
nA
1 30
O-,.,,,.,..,,,,.O Membrane potantiaJ ( mV )
seconcls
Pig. 3. Voltage dependence of GABA desensitization. (A) Responses of neurons to 100 PM GABA applied for 30 s at different membrane potentials. Extent and rates of desensitization decreased with depolarization. Note that the inward current amplitude induced at -3OmV was similar to, but slightly smaller than, the outward current at + 30 mV. (B) Plots of time constants (squares) and peak-to-plateau ratio (circles) at different membrane potentials during perfusion with 1OOpM for 30 s to three-week-old neurons. (r was counted as 30 s when no desensitization was observed.) Analysis of variances was used to test the statistical significance. Peak-to-plateau ratios at - 80 mV or + 30 mV were significantly different from those at - 30 mV (P < 0.02). Time constants at -i-30 mV were significantly different from those at - 30 and - 80 mV (P c 0.001). Values plotted are the means + SD. of six cells.
brane potential was present in the younger cells, but was somewhat less prominent (Fig. 4). This voltage dependence of de~nsitization was not due to differs ences in the peak current levels. Figure 3A illustrates
that the amplitude of the inward current induced by GABA at -30 mV was very similar to the amplitude of the outward current induced at +30mV, yet the desensitization is very different.
A
B
a
30
7
n
1 week
0
3 weeks
-60 -30 0 Membrane potential (mV)
30
‘3
O-,,.
Membrane potential (mV)
Fig. 4. Development of voltage-dependent desensitization. (A) Peak-to-plateau ratios of desensitization induced by 100 ,UM GABA, measmed at difTerent membrane potentials for neurons one and three weeks in culture. Note that neurons at both ages showed clear voltage dependence to the desensitization but desensitization was more extensive in the older neurons. Peak-to-plateau ratios at three weeks were significantly different from those at one week at - 80 mV (P < 0.001) and - 30 mV (P < 0.05, Student’s r-test). In one-week-old cells, the peak-to-piateau ratio at - 80 mV was signiticantly d&rent from those at - 30 and + 30 mV (P u 0.05, an&y& of vati was used). Bach point represents the mean f SD. of six to 15 cells. (B) Desensitization time constants at three different membrane potentials for neurons one and three weeks in culture. (t was estimated at 30 s when there was no detectable desenaitimtion at + 30 mV.) The nenrons desensitixed mom quickly at three weeks than at one week at - 80 mV (ZJ < 6.02, Student’s r-test).The difFerenca of time constants between - 80 and + 30 mV at three weeks (22.2 + 4.2 s, mean f S.D. of six cells) was signi8cantly larger (P < 0.02, Student’s i-test) than those at one week (12.1 + 1.8 s, the mean _+SD. of six to 15 ceils). Bach point represents the mean + SD. of six to 15 cells.
GABA desensitization in hippocampal neurons
GABA 100pM
GABA 1 pM
A
GABA 100 FM
h
UnAL
10 seconds GABA 100yM
GABA 1 PM
GABA 100 pM
Fig. 5. Low concentrations of GABA do not affect the responses of the GABA, receptor. Penwriter traces illustrating the effects of 1 PM GABA perfused for 10 and 30 s before application of 100 PM GABA (1 FM GABA did not induce any detectable current by itself). Neither the maximal response nor the desensitization produced by subsequent 1OOpM GABA was affected. (The very small decreases in maximal responses between the first and second application of 100 PM GABA were due to the slow but progressive rundown of GABA responses commonly seen in cultured neurons.)
Recovery of sensitivity of neurons to GABA occurred slowly in 2-3 min after submaximal desensi-
tization. The resensitization process did not seem to have the same marked voltage dependence as the desensitization process. The effects of low concentrations of GABA were studied to examine whether GABA, receptors could be desensitized without evidence for first opening of channels. GABA at 1 and 2 p M was perfused on to neurons for 2-60 s and then the perfused GABA concentration was rapidly switched to 100 PM. The low concentrations of GABA did not induce any detectable current and had no effect on subsequent maximal response or desensitization induced by 100 PM GABA (Fig. 5). Perfusion with 5 PM GABA induced small currents but also had no effect on the maximal response to 1OOpM GABA or on the desensitization characteristics. DISCUSSION
We evaluated desensitization of the GABA, receptor with two indicators, PPR and r, which represents the extent and rate of desensitization, respectively. Both of these measures observed in cultured hippocampal neurons were dose-dependent and culture age-dependent. In our study, desensitization was seen at very early ages of culture, three days, as early as we first examined the responses to GABA. This corresponds to a period when the hippocampal neurons were still immature and unlikely to have developed synaptic connections. This finding suggests that, in situ, desensitization of GABA,, receptors is present in the perinatal period, at the latest, in the rat. To our knowledge, the ontogenetic development of desensitization has not been reported previously. Although
the amount of desensitization for a given concentration of GABA increased with culture age, desensitization to high concentrations of GABA resulted in the same rate and extent of desensitization in the very young neurons as in the older neurons. These data suggest that the mechanism of desensitization increases its sensitivity with age, while the maximal desensitization remains constant. In this study, desensitization of the GABAA responses was clearly voltage-dependent. This is similar to what has been seen in cultured neocortical neurons.‘“,L’ In contrast, in retinal ganglion cells from 7-l l-day-old rats, desensitization of GARA-induced current was reported not to be altered by membrane potentials.*’ Thus, the properties of the GABAA receptors appear to be different in different brain regions, and the differences in desensitization characteristics could have significant consequences for the normal regulation of GABA-induced inhibition. It has recently been shown, for example, that both inhibitory postsynaptic currents (IPSCs) among cortical neurons (Frosch and Dichter, unpublished observations) and mini-IPSCs in hippocampal neurons (Wilcox and Dichter, unpublished observation) have decay phases which are voltage-dependent in much the same manner as the direct GABA responses, but at a much faster time scale. Whether these synaptic effects are due to desensitization at subsynaptic receptors or to some other factor remains to be determined. Such voltage dependency might be important in the regulation of interaction between excitatory and inhibitory neurotransmitters in CNS. Neurons depolarized by excitatory neurotransmitters would exhibit less desensitization with GABA and therefore GABA could exert a more profound inhibitory effect on neurons which are excessively excited.Lo*‘l The
576
D. .I. OH
and
mechanism of membrane potential dependency remains to be investigated. Membrane potential dependency was less in younger cells than in older cells, which suggests that younger cells do not maintain their sensitivity to GABA under depolarized conditions as well as older cells. This developmental profile might be one reason that young animals and humans are more susceptible to the development of seizures than their more mature counterparts. The kinetics of desensitization have been reported for other ligand gated channeIs.4,5,‘5 Recently, Dude1 et af.‘.’ demonstrated that very low or slowly rising concentrations of glutamate could induce desensitization of glutamate receptor/channels without first opening the channels in crayfish and locust muscle. This indicates that desensitization of glutamate receptors can proceed from a closed, but openable, channel state, to a closed and desensitized channel without
M. A. Dtcrrrra first opening. In our experiments, howcvcr. ION concentrations of GABA did not influence responses to subsequent application of larger concentrations of GABA, suggesting that desensitization of the GABA, receptors observed in our study starts not from a closed state but from an open state of the channels. Thus, the kinetics of desensitization and, likely, the underlying molecular mechanisms of desensitization, appear to differ significantly between the amino acid receptors which are responsible for most short latency excitatory and inhibitory neurotransmission in the mammalian CNS. The functional consequences of these differences remain to be determined. Acknow&dgements-We
would like to thank MS Kay Cherian for help in preparing and maintaining the cultures used for this work. This work was supported by NS 24927 and a grant from the Pew Foundation Program in Neuroscience.
REFERENCES
5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
Buchhalter J. and Dichter M. (1991) Electrophysiological comparison of pyramidal and stellate nonpyramidal neurons in dissociated cell culture of rat hipoocamuus. Brain Res. Bull. 26, 333-338. Cash D. and Subbarao K. (1987) I%sensit&ation of GABA receptor from rat brain: two distinguishable receptors on the same membrane. Biochemistry 26, 75567562. Chen J., Stelzer A., Kay A. and Wong R. (1990) GABA, receptor function is regulated by phosphorylation in acutely dissociated guinea pig hippocampal neurons. J. Physiol. 420, 207-221. Colquhoun D., Mathie A., Mulrine N. and Ogden D. (1989) Studies on single acetylcholine receptor channels in muscle end plate and sympathetic neurons. In Neuromusculur Junction (eds Sellin L., Libelius R. and Thesleff S.), pp. 217-234. Elsevier, Amsterdam. Daoud M. and Usherwood P. (1978)Desensitization and potentiation during glutamate application to locust skeletal muscle. Comp. Biochem. Physiol. 5!X, 105.-110. Dichter M. and Frey J. (1989) GABA desensitization in hippocampal neurons in culture. Sot. Neurosci. Abstr. 15,997. Dude1 J., Franke Ch. and Hatt H. (1990) Rapid activation, desensitization, and resensitization of synaptic channels of crayfish muscle after glutamate pulses. Eiophys. J. 57, 533-545. Dude1 J., Franke Ch., Hatt H., Ramsey R. and Usherwood P. (1990) Glutamatergic channels in locust muscle show a wide time range of desensitization and resensitization characteristics. Neurosci. Left. 114, 207-212. Forscher P. and Oxford G. (1985) Modulation of calcium channels by norepinephrine in internally dialyzed avian sensory neurons. J. gen. Physiol. 85, 743-763. Frosch M. (1987) GABA activated currents and channels in cultured rat cortical neurons. Ph.D. thesis, Harvard University, Cambridge, MA. Frosch M., Lipton S. and Dichter M. (1992) Desensitization of GABA activated currents and channels in cultured cortical neurons. J. Neurosci. (in press). Hamill O., Marty A., Neher E., Sakmann B. and Sigworth F. (1981) Improved patch clamp techniques for high resolution current recording from cells and cell free membrane patches. Pjitiger’s Arch. ges. Physiol. 391, 85-100. Houamed K., Bilbe G., Smart T., Constanti A., Brown D., Bernard E. and Richards B. (1984) Expression of functional GABA, glycine and glutamate receptors in Xenopus oocytes injected with rat brain mRNA. Nature 310, 318-321. Huguernard J. and Alger B. (1986) Whole cell voltage clamp study of the fading of GABA activated currents in acutely dissociated hippocampal neurons. J. Neurophysiol. 56, 1-18. Katz B. and Thesleff S. (1957) A study of the ‘desensitization’ produced by acetylcholine at the motor end-plate. J. Physiol. 138, 63-80. Numann R. and Wong R. (1984) Voltage clamp study on GABA response desensitization in single pyramidal cells dissociated from the hippocampus of adult guinea pigs. Neurosci. Left. 47, 289-294. Ozawa S. and Yuzaki M. (1984) Patch clamp studies of chloride channels activated by gamma-aminobutyric acid in cultured hippocampal neurones of the rat. Neurosci. Res. 1, 2755293. Parker I., Gundersen C. and Miledi R. (1986) Action of pentobarbital on rat brain receptors expressed in Xenopus oocytes. J. Neurosci. 6, 2290-2297. Segal M. and Barker J. (1984) Rat hippocampal neurons in culture: properties of GABA activated chloride ion conductances. J. Neurophysiol. 51, 500-515. Stelzer A., Kay A. and Wong R. (1988) GABA,-receptor function in hippocampal cells is maintained by phosphorylation factors. Science 241, 339-341. Tauck D., Frosch M. and Lipton S. (1988) Characterization of GABA and glycine-induced currents of solitary rodent retinal ganglion cells in culture. Neuroscience 27, 193-203. Thalmann R. and Hershkowitz N. (1985) Some factors that influence the decrement in the response to GABA during its continuous iontophoretic application to hippocampal neurons. Brain Res. 342, 219-233. (Accepted 29 January 1992)
Neuroscience Vol. 49, No. 3, pp. 571-576, I992 Printed in Great Britain
DESENSITIZATION CULTURED
OF GABA-INDUCED CURRENTS RAT HIPPOCAMPAL NEURONS
IN
D. J. OH and M. A. DICHTER* Department of Neurology, University of Pennsylvania School of Medicine and Graduate Hospital, Philadelphia, PA 19104, U.S.A. Ahatraet-The basic characteristics of desensitization of the GABA, receptor were investigated in cultured rat hippocampal neurons (three days to four weeks in vitro) using whole cell patch clamp techniques. GABA at 10-500 pM was perfused on to neurons for 30 or 60 s, with 60 s intervals of wash with control
bath solution between perfusions. Desensitization, evaluated by peak-to-plateau ratio and time constants of current decay (7), was dosedependent and culture age-dependent. Desensitization was observed as early as three days in culture, the earliest time tested. At all ages, higher concentrations of GABA induced both larger and faster desensitization. Desensitization was markedly voltage-dependent and decmased with depolarization; peak-to-plateau ratio went from 6.3 to 1.4 and 7 went from 4.6 to 26.8 s when holding
potentials were changed from - 80 mV to + 30 mV. Low concentrations of GABA (l-2 PM) perfused for 2-60 s, which did not induce any current, had no effect on the maximal response nor desensitization produced by a subsequent application of 100p M GABA. This finding suggests that GABA receptors were not desensitized without first being activated.
The gamma-aminobutyric acid type A (GABA,) receptor, with its integral chloride ion channel, is considered to be the major inhibitory neurotransmitter receptor in the mammalian brain. GABAmediated inhibition shapes sensory input, motor output, and maintains a “basal” level of excitability in the mammalian cortex; any diminution of GABAmediated inhibition can lead to excessive excitability, which often manifests itself as seizure activity. The GABA receptor complex contains a number of modulatory sites, most of which are capable of being affected by exogenous agents (e.g. benzodiazepines, barbiturates, etc.). Although it has been hypothesized that endogenous agents also exist which can affect these sites, their identities have remained elusive to date. GABA-mediated inhibition can also be modulated by presynaptic mechanisms. GABA, receptors exist on GABA terminals and activation of these receptors can suppress the release of GABA, thereby downregulating the synapse. Other presynaptic receptors may also function in this way. GABA,-mediated inhibition can also be autoregulated by a process of desensitization. It is known that continued application of GABA to neurons produces a decrease in the GABA-induced current, due to both a change in the transmembrane chloride gradienti4J9 *To whom correspondence should be addressed. Abbreuiafions: BAPTA, 1,2-bis(2aminophenoxy)ethaneN,iV,N,N-tetra-acetic acid; DMEM, Dulbecco’s modified Eagle’s medium; HEPES, [N-(2-hydroxyethyl)piperaxine-N-(2-ethanesulfonic acid)]; IPSC, inhibitory postsynaptic current; PPR, peak-to-plateau ratio; 7, exponential current decay. 571
and a decrease in the induced conductance, a true receptor desensitization. This latter phenomenon has been observed electrophysiologically in neocorticallo*” and hippocampal neurons,6*16J7,Z and in rat brain receptors expressed in oocytesnJ* and biochemically in synaptosome preparations.2 Previous work from our laboratory demonstrated that desensitization at the GABA, receptor in neocortical neurons in culture had interesting properties.“,” It was concentration-dependent and independent of the amount of current induced or of the charge transfer induced by a given concentration of GABA. The rate and extent of desensitization was also voltage-dependent; it occurred much more slowly at depolarized membrane potentials. In addition, desensitization occurred more rapidly at the single channel level, when patches of membrane were examined in isolation, and the voltage dependence of the desensitization was lost. The cellular and molecular mechanisms underlying desensitization at the GABA, receptor are not understood. The characteristics of desensitization of GABA, receptors are not the same in all regions of the mammalian CNS. In rat retinal ganglion cells, true receptor desensitization occurs but does not show any voltage dependence.2’ In this paper we describe the basic characteristics of desensitization and resensitization of the GABA, receptor in cultured rat hippocampal neurons, attempting to develop quantitative measures which can be utilized in subsequent studies of mechanisms. We characterize the development of desensitization, its voltage dependency and the issue of whether desensitization can occur by
572
L> j.
OH
concentrations of GABA below the threshold channel opening. EXPERIMENTAL
and
for
PROCEDURES
Cell culture
Primary rat hippocampal cultures were prepared as reported previously.’ Hippocampal cells were removed from 20-21-day-old fetuses and dissociated with 0.03% trypsin in Dulbecco’s modified Eagle’s medium (DMEM) for 40 min at 37°C. Cells were subsequently dispersed by trituration and plated onto 35 mm dishes containing polylysine-coated coverslips. Cultures were maintained with DMEM, 10% hyclone calf serum, and 10% Ham’s F-12 in a 7% CO, incubator at 37°C. Cytosine arabinoside (10m5M) was added at 7-10 days after initial culture to prevent glial proliferation. Neurons cultured for three days to four weeks were used in this study. Electrophysiological recordings
Recordings were made with a Dagan 8900 patch clamp amplifier using standard whole cell patch clamp methods’* at room temperature. Kimax electrodes were used. Electrode resistance ranged from 2 to 5 MR. Neurons were routinely held at -8OmV. To analyse the voltage dependency of desensitization, membrane potentials were changed from - 80 mV to - 30 mV and + 30 mV. GABA concentrations between 1 and 500pM were applied by pressure ejection from multibarrelled micropipettes (tip diameters 5-10 pm) positioned 25-50 pm from the cell. Between applications neurons were washed with control bath solution. GABA-induced current was recorded on a Gould strip chart recorder and a computer (Indec systems). The external bath solution consisted of 140mM NaCl, 3mM KCI, 2mM CaCl,, 1 mM MgCl,, 10 mM HEPES, 8 mM glucose, 5 mM 4-aminopyridine, 4 mM CoCI,, 5 pM tetrodotoxin, pH 7.2
M. A.
bCHTEK
with NaOH. The internal solution contained I05 mM C& I. I mM MgC12, IO mM HEPES, 7 mM BAPTA. 70/1M Irupeptin, 14 mM phosphocreatine. 35 U/ml creatlne phosphokinase, 5 mM ATP-Mg, pH 7.2 with CsOH. We chose this internal solution to prevent rundown.‘,0.‘0 Small differences of Cl- concentration between our internal and bath solutions did not have any effect on desensitizatiorb in our experiments. RESliLTS
Prolonged application of GABA to cultured hippocampal neurons induces a current which peaks within less than 1 s and stays elevated for only a short time. The current then decreases exponentially with a single time constant. This gradual decay, desensitization. was observed in almost all cells at various culture ages from three days to four weeks, except when very low concentrations of GABA were applied. This current decay is a true receptor desensitization, as it was seen in neurons recorded with equimolar Cl on both sides of the membrane and therefore could not have been due to a simple shift in the Cl equilibrium potential. In order to compare desensitization under different conditions and between cells, we measured the time constant of exponential current decay (5) and peak-to-plateau ratio (PPR) of induced current. Figure 1 shows a typical pattern of desensitization of GABA resDonses in one neuron at four different concentrations. Higher concentrations of GABA perfused on to the cell resulted in faster and more extensive desensitization. PPR and r were dosedependent (Fig. IB) in neurons from three days to
B n Peak ! Plateau 0
Time constant
(seconds)
25 3
4 J
50pM
i”-----,
500 JJM
JL
20-
5;
ol 5
1 .O nA
i 10
0 100 GABA concentration
P 1000 QIM)
I 15 seconds
Fig. 1. Concentration dependence of GABA desensitization. (A) Representative penwriter responses from one hippocampal neuron cultured for three weeks to different concentrations of GABA applied for 1 min. The extent and rate of desensitization increased with increasing GABA concentrations. Membrane potential was held at -80 mV. (B) Desensitization evaluated by peak-to-plateau ratio (squares) and time constants of current decay (circles). Membrane potential was held at -80 mV. GABA was perfused at 10, 50, 100, and 5OOpM on to three- to four-week-old cells. Note the dose-dependent relationship. Each point represents the means + S.D. of seven to eight cells. Analysis of variance demonstrated statistical significance between the low, intermediate and high concentrations for both measures of desensitization (P < 0.01).
513
GABA desensitization in hippocampal neurons four weeks in culture. Desensitization was often extensive enough that the plateau currents after 30-45 s were smaller at higher concentrations of
GABA. Qualitatively similar dose-dependent desensitization was noted in neurons as early as three days in culture (the earliest time point tested) and as late as four weeks in culture. At concentrations of GABA from 10 to lOOpM, desensitization was larger and faster in older cells than in younger cells (Fig. 2A, B). At 500 PM GABA, however, the desensitization was similar at all culture ages, although
peak currents induced by 500 PM GABA increased significantly (P < 0.0001) from three-day-old cells (1280 + 550pA, mean f SD., N = 5) to three- to four-week-old cells (4040 + 530, N = 8). Desensitization of GABA-induced currents was markedly voltage-dependent.lO~‘l Figure 3A illustrates desensitization induced by 100 FM GABA at three different membrane potentials. Desensitization was smaller and slower as the membrane was depolarized and almost no desensitization was observed at +30 mV. PPR and t at three different membrane potentials are shown in Fig. 3B. The effect of mem-
0
3days
q
1 week
gxI 2 weeks
q
3-4 weeks
II]
3 days
g
1 week
q q
2 weeks
d 5
0
’
1OpM
’
50pM ’ 100~1M GABA concentration
70 3
p 50 8 3 40
’
5OO)tM
’
3-4 weeks
E Q 30 LO E i= 10 0 ’
~OJJM
’
50pM GABA concentration
Fig. 2. Development of desensitization from three-day- to three- to four-week-old cells to different concentrations of GABA applied for 1 min. Desensitization (A: peak-to-plateau ratio; B: T) was more prominent in older cells than in younger cells, except at high concentrations of GABA. Membrane potential was -8OmV. When no desensitization was seen, the time constant was estimated at 60s for calculation purposes. Each value in the figure is the mean f SD. of five to 14 cells. The difference between values at three days and three to four weeks was statistically significant for 10-100 PM GABA (P < 0.001) but not at the highest concentration tested. (The SD. for three-day-old neurons and 10 PM GABA was 0, as no desensitization was noted.)
574
D.
J.
OH and
A
M. A.
DICHTER
B
0 Peak /Plateau n Time constant ( seconds ) -6
-80
mV
L -6
-30
mV
63 -4 i? 3 36 -2 0 -1
2.0
nA
1 30
O-,.,,,.,..,,,,.O Membrane potantiaJ ( mV )
seconcls
Pig. 3. Voltage dependence of GABA desensitization. (A) Responses of neurons to 100 PM GABA applied for 30 s at different membrane potentials. Extent and rates of desensitization decreased with depolarization. Note that the inward current amplitude induced at -3OmV was similar to, but slightly smaller than, the outward current at + 30 mV. (B) Plots of time constants (squares) and peak-to-plateau ratio (circles) at different membrane potentials during perfusion with 1OOpM for 30 s to three-week-old neurons. (r was counted as 30 s when no desensitization was observed.) Analysis of variances was used to test the statistical significance. Peak-to-plateau ratios at - 80 mV or + 30 mV were significantly different from those at - 30 mV (P < 0.02). Time constants at -i-30 mV were significantly different from those at - 30 and - 80 mV (P c 0.001). Values plotted are the means + SD. of six cells.
brane potential was present in the younger cells, but was somewhat less prominent (Fig. 4). This voltage dependence of de~nsitization was not due to differs ences in the peak current levels. Figure 3A illustrates
that the amplitude of the inward current induced by GABA at -30 mV was very similar to the amplitude of the outward current induced at +30mV, yet the desensitization is very different.
A
B
a
30
7
n
1 week
0
3 weeks
-60 -30 0 Membrane potential (mV)
30
‘3
O-,,.
Membrane potential (mV)
Fig. 4. Development of voltage-dependent desensitization. (A) Peak-to-plateau ratios of desensitization induced by 100 ,UM GABA, measmed at difTerent membrane potentials for neurons one and three weeks in culture. Note that neurons at both ages showed clear voltage dependence to the desensitization but desensitization was more extensive in the older neurons. Peak-to-plateau ratios at three weeks were significantly different from those at one week at - 80 mV (P < 0.001) and - 30 mV (P < 0.05, Student’s r-test). In one-week-old cells, the peak-to-piateau ratio at - 80 mV was signiticantly d&rent from those at - 30 and + 30 mV (P u 0.05, an&y& of vati was used). Bach point represents the mean f SD. of six to 15 cells. (B) Desensitization time constants at three different membrane potentials for neurons one and three weeks in culture. (t was estimated at 30 s when there was no detectable desenaitimtion at + 30 mV.) The nenrons desensitixed mom quickly at three weeks than at one week at - 80 mV (ZJ < 6.02, Student’s r-test).The difFerenca of time constants between - 80 and + 30 mV at three weeks (22.2 + 4.2 s, mean f S.D. of six cells) was signi8cantly larger (P < 0.02, Student’s i-test) than those at one week (12.1 + 1.8 s, the mean _+SD. of six to 15 ceils). Bach point represents the mean + SD. of six to 15 cells.
GABA desensitization in hippocampal neurons
GABA 100pM
GABA 1 pM
A
GABA 100 FM
h
UnAL
10 seconds GABA 100yM
GABA 1 PM
GABA 100 pM
Fig. 5. Low concentrations of GABA do not affect the responses of the GABA, receptor. Penwriter traces illustrating the effects of 1 PM GABA perfused for 10 and 30 s before application of 100 PM GABA (1 FM GABA did not induce any detectable current by itself). Neither the maximal response nor the desensitization produced by subsequent 1OOpM GABA was affected. (The very small decreases in maximal responses between the first and second application of 100 PM GABA were due to the slow but progressive rundown of GABA responses commonly seen in cultured neurons.)
Recovery of sensitivity of neurons to GABA occurred slowly in 2-3 min after submaximal desensi-
tization. The resensitization process did not seem to have the same marked voltage dependence as the desensitization process. The effects of low concentrations of GABA were studied to examine whether GABA, receptors could be desensitized without evidence for first opening of channels. GABA at 1 and 2 p M was perfused on to neurons for 2-60 s and then the perfused GABA concentration was rapidly switched to 100 PM. The low concentrations of GABA did not induce any detectable current and had no effect on subsequent maximal response or desensitization induced by 100 PM GABA (Fig. 5). Perfusion with 5 PM GABA induced small currents but also had no effect on the maximal response to 1OOpM GABA or on the desensitization characteristics. DISCUSSION
We evaluated desensitization of the GABA, receptor with two indicators, PPR and r, which represents the extent and rate of desensitization, respectively. Both of these measures observed in cultured hippocampal neurons were dose-dependent and culture age-dependent. In our study, desensitization was seen at very early ages of culture, three days, as early as we first examined the responses to GABA. This corresponds to a period when the hippocampal neurons were still immature and unlikely to have developed synaptic connections. This finding suggests that, in situ, desensitization of GABA,, receptors is present in the perinatal period, at the latest, in the rat. To our knowledge, the ontogenetic development of desensitization has not been reported previously. Although
the amount of desensitization for a given concentration of GABA increased with culture age, desensitization to high concentrations of GABA resulted in the same rate and extent of desensitization in the very young neurons as in the older neurons. These data suggest that the mechanism of desensitization increases its sensitivity with age, while the maximal desensitization remains constant. In this study, desensitization of the GABAA responses was clearly voltage-dependent. This is similar to what has been seen in cultured neocortical neurons.‘“,L’ In contrast, in retinal ganglion cells from 7-l l-day-old rats, desensitization of GARA-induced current was reported not to be altered by membrane potentials.*’ Thus, the properties of the GABAA receptors appear to be different in different brain regions, and the differences in desensitization characteristics could have significant consequences for the normal regulation of GABA-induced inhibition. It has recently been shown, for example, that both inhibitory postsynaptic currents (IPSCs) among cortical neurons (Frosch and Dichter, unpublished observations) and mini-IPSCs in hippocampal neurons (Wilcox and Dichter, unpublished observation) have decay phases which are voltage-dependent in much the same manner as the direct GABA responses, but at a much faster time scale. Whether these synaptic effects are due to desensitization at subsynaptic receptors or to some other factor remains to be determined. Such voltage dependency might be important in the regulation of interaction between excitatory and inhibitory neurotransmitters in CNS. Neurons depolarized by excitatory neurotransmitters would exhibit less desensitization with GABA and therefore GABA could exert a more profound inhibitory effect on neurons which are excessively excited.Lo*‘l The
576
D. .I. OH
and
mechanism of membrane potential dependency remains to be investigated. Membrane potential dependency was less in younger cells than in older cells, which suggests that younger cells do not maintain their sensitivity to GABA under depolarized conditions as well as older cells. This developmental profile might be one reason that young animals and humans are more susceptible to the development of seizures than their more mature counterparts. The kinetics of desensitization have been reported for other ligand gated channeIs.4,5,‘5 Recently, Dude1 et af.‘.’ demonstrated that very low or slowly rising concentrations of glutamate could induce desensitization of glutamate receptor/channels without first opening the channels in crayfish and locust muscle. This indicates that desensitization of glutamate receptors can proceed from a closed, but openable, channel state, to a closed and desensitized channel without
M. A. Dtcrrrra first opening. In our experiments, howcvcr. ION concentrations of GABA did not influence responses to subsequent application of larger concentrations of GABA, suggesting that desensitization of the GABA, receptors observed in our study starts not from a closed state but from an open state of the channels. Thus, the kinetics of desensitization and, likely, the underlying molecular mechanisms of desensitization, appear to differ significantly between the amino acid receptors which are responsible for most short latency excitatory and inhibitory neurotransmission in the mammalian CNS. The functional consequences of these differences remain to be determined. Acknow&dgements-We
would like to thank MS Kay Cherian for help in preparing and maintaining the cultures used for this work. This work was supported by NS 24927 and a grant from the Pew Foundation Program in Neuroscience.
REFERENCES
5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
Buchhalter J. and Dichter M. (1991) Electrophysiological comparison of pyramidal and stellate nonpyramidal neurons in dissociated cell culture of rat hipoocamuus. Brain Res. Bull. 26, 333-338. Cash D. and Subbarao K. (1987) I%sensit&ation of GABA receptor from rat brain: two distinguishable receptors on the same membrane. Biochemistry 26, 75567562. Chen J., Stelzer A., Kay A. and Wong R. (1990) GABA, receptor function is regulated by phosphorylation in acutely dissociated guinea pig hippocampal neurons. J. Physiol. 420, 207-221. Colquhoun D., Mathie A., Mulrine N. and Ogden D. (1989) Studies on single acetylcholine receptor channels in muscle end plate and sympathetic neurons. In Neuromusculur Junction (eds Sellin L., Libelius R. and Thesleff S.), pp. 217-234. Elsevier, Amsterdam. Daoud M. and Usherwood P. (1978)Desensitization and potentiation during glutamate application to locust skeletal muscle. Comp. Biochem. Physiol. 5!X, 105.-110. Dichter M. and Frey J. (1989) GABA desensitization in hippocampal neurons in culture. Sot. Neurosci. Abstr. 15,997. Dude1 J., Franke Ch. and Hatt H. (1990) Rapid activation, desensitization, and resensitization of synaptic channels of crayfish muscle after glutamate pulses. Eiophys. J. 57, 533-545. Dude1 J., Franke Ch., Hatt H., Ramsey R. and Usherwood P. (1990) Glutamatergic channels in locust muscle show a wide time range of desensitization and resensitization characteristics. Neurosci. Left. 114, 207-212. Forscher P. and Oxford G. (1985) Modulation of calcium channels by norepinephrine in internally dialyzed avian sensory neurons. J. gen. Physiol. 85, 743-763. Frosch M. (1987) GABA activated currents and channels in cultured rat cortical neurons. Ph.D. thesis, Harvard University, Cambridge, MA. Frosch M., Lipton S. and Dichter M. (1992) Desensitization of GABA activated currents and channels in cultured cortical neurons. J. Neurosci. (in press). Hamill O., Marty A., Neher E., Sakmann B. and Sigworth F. (1981) Improved patch clamp techniques for high resolution current recording from cells and cell free membrane patches. Pjitiger’s Arch. ges. Physiol. 391, 85-100. Houamed K., Bilbe G., Smart T., Constanti A., Brown D., Bernard E. and Richards B. (1984) Expression of functional GABA, glycine and glutamate receptors in Xenopus oocytes injected with rat brain mRNA. Nature 310, 318-321. Huguernard J. and Alger B. (1986) Whole cell voltage clamp study of the fading of GABA activated currents in acutely dissociated hippocampal neurons. J. Neurophysiol. 56, 1-18. Katz B. and Thesleff S. (1957) A study of the ‘desensitization’ produced by acetylcholine at the motor end-plate. J. Physiol. 138, 63-80. Numann R. and Wong R. (1984) Voltage clamp study on GABA response desensitization in single pyramidal cells dissociated from the hippocampus of adult guinea pigs. Neurosci. Left. 47, 289-294. Ozawa S. and Yuzaki M. (1984) Patch clamp studies of chloride channels activated by gamma-aminobutyric acid in cultured hippocampal neurones of the rat. Neurosci. Res. 1, 2755293. Parker I., Gundersen C. and Miledi R. (1986) Action of pentobarbital on rat brain receptors expressed in Xenopus oocytes. J. Neurosci. 6, 2290-2297. Segal M. and Barker J. (1984) Rat hippocampal neurons in culture: properties of GABA activated chloride ion conductances. J. Neurophysiol. 51, 500-515. Stelzer A., Kay A. and Wong R. (1988) GABA,-receptor function in hippocampal cells is maintained by phosphorylation factors. Science 241, 339-341. Tauck D., Frosch M. and Lipton S. (1988) Characterization of GABA and glycine-induced currents of solitary rodent retinal ganglion cells in culture. Neuroscience 27, 193-203. Thalmann R. and Hershkowitz N. (1985) Some factors that influence the decrement in the response to GABA during its continuous iontophoretic application to hippocampal neurons. Brain Res. 342, 219-233. (Accepted 29 January 1992)
