Neuroscienee Letters, 114 (1990) 197-202
197
Elsevier Scientific Publishers Ireland Ltd. NSL 06944
The glycine site modulates NMDA-mediated changes of intracellular free calcium in cultures of hippocampal neurons M.W. Oliver l, J.A. Shacklock 2, M. Kessler 1, G. Lynch I and K.G. Baimbridge 2 1Centerfor the Neurobiology of Learning and Memory, Bonney Center, University of California, Irvine, CA (U.S.A.) and 2Departmentof Physiology, University of British Columbia, Vancouver,B.C. (Canada) (Received 3 January 1990; Revised version received 15 February 1990; Accepted 19 February 1990)
Key words: 7-Chlorokynurenate; Glycine; NMDA; Calcium; Hippocampus Recent evidence indicates that the N-methyl-D-aspartate (NMDA) receptor--channel complex contains a glycine subunit whose activation may be necessary for channel operation. It has been previously shown that stimulation of the NMDA receptor leads to an increase in intracellular ionic Ca 2÷ ([Ca2+]i); therefore, we examined the role of the NMDA receptor-associated glycine site in modulating [Ca2+]~ using the fluorescent dye Fura II in hippocampal neuron cultures. A 3-s pulse of 200/~M NMDA resulted in a mean lCa2+]i increase of 363 nM above the average resting concentration of 122 nM. Perfusion of the glycine site antagonist 7-chlorokynurenate (C1-Kyn) essentially eliminated the NMDA-induced alteration in [Ca2+]i. Either 40/~M glycine or 50/~M D-serine completely reversed the effect of CI-Kyn, indicating that the drug was acting at the glycine site. The NMDA receptor antagonists 2-amino-5-phosphonovalerate (AP5) and ketamine, which bind to the glutamate recognition site and the ion channel, respectively, also blocked the NMDA-mediated [Ca2+]~ response; however, glycine or D-serine did not reverse this effect. These data show that the glycine binding site coupled to the NMDA receptor modulates the NMDAmediated increase in [Ca2+]~. Antagonists of the glycine site provide a new tool to investigate and possibly control neuroplasticity and neurotoxicity related to the NMDA receptor complex.
Among the several identified receptors for the transmitter glutamate, the NMDA receptor alone allows Ca 2+ ions to pass through its channel [1]. Thus, NMDA receptors may be specialized to convert electrical activity into cellular changes, the best studied example being the induction of long-term potentiation (LTP) [2,3]. The complex regulation of the NMDA receptor, which has become apparent in recent years, may be related to the need of nerve cells to maintain receptor activation and associated Ca 2+ influx within strict limits. Whereas Mg 2+ ions are essential for establishing the unusual voltage-dependence of NMDA receptor activation [4], the function of 2 further binding sites for phencyclidine (PCP) and for glycine is not clear. The Correspondence: M.W. Oliver, Department of Anatomy and Neurobiology, University of California, Irvine, CA 92717, U.S.A. 0304-3940/90/$ 03.50 © 1990 Elsevier Scientific Publishers Ireland Ltd.
198
PCP site, presumably located within the ion channel, has no known physiological function or endogenous ligand but is of considerable pharmacological interest. The strychnine-insensitive glycine site appears to modulate the channel opening frequency without affecting either the conductance or mean open time [5]. Other studies indicate that activation of the glycine site is essential for channel operation [6] or that glycine prevents N M D A receptor desensitization [7]. It has been difficult to assess the contributions of the glycine site to N M D A receptor-mediated synaptic responses because of the likely presence of glycine in the extracellular fluid even in nominally glycine-free media. However, selective antagonists for the glycine site have become available and with these it is now possible to probe the extent to which the glycine site controls N M D A receptor function in physiologically relevant situations. Since LTP depends on the participation of N M D A receptors and involves Ca 2+ as intracellular activator, it has been speculated that Ca 2+ influx through the N M D A receptor sets in motion the sequence of processes that leads to stable potentiation. We recently found that 7-chloro-kynurenate, the most selective and potent glycine site antagonist available [8], completely blocks induction of LTP in hippocampal slices [9]. We interpreted this as evidence that activation of the glycine site is a requirement for the passage of calcium through the ionophore of synaptically activated N M D A receptors. This proposition was further tested in the present experiments by directly measuring the effect of 7-chloro-kynurenate on changes in intracellular free calcium levels produced by short pulses of NMDA. These experiments also allowed us to assess the presence of N M D A receptors not linked to a glycine site, assuming the preservation of calcium influx through these receptors. Rat hippocampal neuron cultures were prepared from 18-day-old embryos according to Banker and Cowan [! 0]. Approximately 2-3 weeks following the initial plating, the cultured cells were pre-loaded for 1.5 h with Fura II-AM (Molecular Probes; Eugene, OR, U.S.A.) as described by Poenie et al. [11], washed and transferred to a perfusion chamber maintained at 26 + 0.5°C. The culture medium contained 1.8 mM Ca 2+ with no added Mg 2* . Individual pyramidal looking neurons were isolated with an adjustable circular diaphragm and viewed with a 100 × oil immersion lens attached to a Jenalumar microscope equipped for epifluorescence. The ratio of Fura II fluorescence following excitation at 350 nm and 380 nm was determined every 1.8 s; the signal was digitized and stored on computer for later analysis of [Ca2+]i. The method of calibration and the formula used to calculate [Ca2+]i have been described [12]. Rmin, Rmax and beta were determined to be 0.45, 6.4 and 9.72, respectively, and a Kd of 220 nM was used for Fura II. Baseline [Ca2+]i responses to 3-s pulses of 200 /tm N M D A were collected at approximately 7-min intervals and the effect of a 7-min perfusion of 20 /~M 7chlorokynurenate (C1-Kyn; 7-chloro-4-hydroxy-quinoline-2-carboxylic acid; Tocris Neuramin, Essex, U.K.) was assessed on NMDA-induced increases in [Ca2+]i. In several experiments, C1-Kyn, D-AP5 or ketamine were simultaneously perfused with glycine (40/~M) or o-serine (50/tM) to test whether these latter compounds could reverse the effects of the antagonists.
199
The mean (___ S.E.M.) basal levels of [Ca2+]i from 16 neurons was 122 + 13 nM and N M D A application resulted in an average increase of 363 + 34 nM. Perfusion of 20 #M CI-Kyn reduced the NMDA-mediated [Ca2+]i enhancement to 29 + 8 nM (n = 9). A typical dose-response experiment demonstrating the effect of CI-Kyn on N M D A responses in a single neuron is shown in Fig. 1. Consecutive pulses of 200/zM NMDA (squares) increased [Ca2+]i from a baseline value of 71 nM to an average maximum of 644 nM (i.e., a change of 573 nM). The perfusion of 5 #M, 10 aM and 20 #M CI-Kyn (bars) caused a sequential decrease in NMDA-induced alterations in [Ca2+]i to 230 nM, 180 nM and 5nM, respectively. In order to determine whether Cl-Kyn was acting on the NMDA-associated glycine site or some other binding sites (e.g., the NMDA receptor) a series of experiments were carried out using glycine, D-serine, D-AP5 and ketamine. Figure 2 shows the effect of 40/zM glycine on the CI-Kyn-induced depression of the NMDA [Ca2+]i response. A control response to NMDA exhibited a 536 nM increase in [Ca2+]i that was almost entirely eliminated by 20/zM CI-Kyn (increase in [Ca2+]i = 24 nM). Perfusion of glycine (40 #M), in the continued presence of C1-Kyn, reversed the antagonism in that [Ca2+]i increased to 449 nM when N M D A was applied. Glycine and D-serine had comparable effects and reversed C1-Kyn antagonism by 89 + 5% (n = 9). NMDA-mediated increases in [Ca2+]i were also blocked by the receptor antagonist D-AP5 and the channel blocker ketamine (mean change in [Ca2+]i --- 13 + 4 nM; n= 5). Figure 3 shows the increase (mean = 326 nM) in [Ca2+]i in a single neuron. Addition of 40 #M D-AP5 to the perfusion medium for a period of 7 min completely blocked the N M D A response and this effect was unaltered by 50 #M gly8OO ¢-
CL-KYN 5 uM '
10 uM I
600
/
20 uM
,--I
I
L) LU LU cc IJ.
I
400
n,< .-I
,_1 .J
200
UJ
C~ nz --
ta
0
D
i
0
J J
10
i
20
i
30
ta i
40
50
MINUTES Fig. I. The effect of C1-Kyn on the NMDA-mediated increase in [Ca2+]~. In normal medium, the first two 3-s applications of 200 aM NMDA (squares) resulted in a 573-nM mean enhancement of [Ca2+]i. Peffusion of 5, 10 and 20/zM C1-Kyn antagonized subsequent pulses of NMDA, with 20/zM CI-Kyn completely blocking any response to the agonist.
200
~"
700
GLYCINE
t-
CL-KYN
I
600 ' L.) ,-J
500
< o uJ
w
400
CE LL
a:
197
Elsevier Scientific Publishers Ireland Ltd. NSL 06944
The glycine site modulates NMDA-mediated changes of intracellular free calcium in cultures of hippocampal neurons M.W. Oliver l, J.A. Shacklock 2, M. Kessler 1, G. Lynch I and K.G. Baimbridge 2 1Centerfor the Neurobiology of Learning and Memory, Bonney Center, University of California, Irvine, CA (U.S.A.) and 2Departmentof Physiology, University of British Columbia, Vancouver,B.C. (Canada) (Received 3 January 1990; Revised version received 15 February 1990; Accepted 19 February 1990)
Key words: 7-Chlorokynurenate; Glycine; NMDA; Calcium; Hippocampus Recent evidence indicates that the N-methyl-D-aspartate (NMDA) receptor--channel complex contains a glycine subunit whose activation may be necessary for channel operation. It has been previously shown that stimulation of the NMDA receptor leads to an increase in intracellular ionic Ca 2÷ ([Ca2+]i); therefore, we examined the role of the NMDA receptor-associated glycine site in modulating [Ca2+]~ using the fluorescent dye Fura II in hippocampal neuron cultures. A 3-s pulse of 200/~M NMDA resulted in a mean lCa2+]i increase of 363 nM above the average resting concentration of 122 nM. Perfusion of the glycine site antagonist 7-chlorokynurenate (C1-Kyn) essentially eliminated the NMDA-induced alteration in [Ca2+]i. Either 40/~M glycine or 50/~M D-serine completely reversed the effect of CI-Kyn, indicating that the drug was acting at the glycine site. The NMDA receptor antagonists 2-amino-5-phosphonovalerate (AP5) and ketamine, which bind to the glutamate recognition site and the ion channel, respectively, also blocked the NMDA-mediated [Ca2+]~ response; however, glycine or D-serine did not reverse this effect. These data show that the glycine binding site coupled to the NMDA receptor modulates the NMDAmediated increase in [Ca2+]~. Antagonists of the glycine site provide a new tool to investigate and possibly control neuroplasticity and neurotoxicity related to the NMDA receptor complex.
Among the several identified receptors for the transmitter glutamate, the NMDA receptor alone allows Ca 2+ ions to pass through its channel [1]. Thus, NMDA receptors may be specialized to convert electrical activity into cellular changes, the best studied example being the induction of long-term potentiation (LTP) [2,3]. The complex regulation of the NMDA receptor, which has become apparent in recent years, may be related to the need of nerve cells to maintain receptor activation and associated Ca 2+ influx within strict limits. Whereas Mg 2+ ions are essential for establishing the unusual voltage-dependence of NMDA receptor activation [4], the function of 2 further binding sites for phencyclidine (PCP) and for glycine is not clear. The Correspondence: M.W. Oliver, Department of Anatomy and Neurobiology, University of California, Irvine, CA 92717, U.S.A. 0304-3940/90/$ 03.50 © 1990 Elsevier Scientific Publishers Ireland Ltd.
198
PCP site, presumably located within the ion channel, has no known physiological function or endogenous ligand but is of considerable pharmacological interest. The strychnine-insensitive glycine site appears to modulate the channel opening frequency without affecting either the conductance or mean open time [5]. Other studies indicate that activation of the glycine site is essential for channel operation [6] or that glycine prevents N M D A receptor desensitization [7]. It has been difficult to assess the contributions of the glycine site to N M D A receptor-mediated synaptic responses because of the likely presence of glycine in the extracellular fluid even in nominally glycine-free media. However, selective antagonists for the glycine site have become available and with these it is now possible to probe the extent to which the glycine site controls N M D A receptor function in physiologically relevant situations. Since LTP depends on the participation of N M D A receptors and involves Ca 2+ as intracellular activator, it has been speculated that Ca 2+ influx through the N M D A receptor sets in motion the sequence of processes that leads to stable potentiation. We recently found that 7-chloro-kynurenate, the most selective and potent glycine site antagonist available [8], completely blocks induction of LTP in hippocampal slices [9]. We interpreted this as evidence that activation of the glycine site is a requirement for the passage of calcium through the ionophore of synaptically activated N M D A receptors. This proposition was further tested in the present experiments by directly measuring the effect of 7-chloro-kynurenate on changes in intracellular free calcium levels produced by short pulses of NMDA. These experiments also allowed us to assess the presence of N M D A receptors not linked to a glycine site, assuming the preservation of calcium influx through these receptors. Rat hippocampal neuron cultures were prepared from 18-day-old embryos according to Banker and Cowan [! 0]. Approximately 2-3 weeks following the initial plating, the cultured cells were pre-loaded for 1.5 h with Fura II-AM (Molecular Probes; Eugene, OR, U.S.A.) as described by Poenie et al. [11], washed and transferred to a perfusion chamber maintained at 26 + 0.5°C. The culture medium contained 1.8 mM Ca 2+ with no added Mg 2* . Individual pyramidal looking neurons were isolated with an adjustable circular diaphragm and viewed with a 100 × oil immersion lens attached to a Jenalumar microscope equipped for epifluorescence. The ratio of Fura II fluorescence following excitation at 350 nm and 380 nm was determined every 1.8 s; the signal was digitized and stored on computer for later analysis of [Ca2+]i. The method of calibration and the formula used to calculate [Ca2+]i have been described [12]. Rmin, Rmax and beta were determined to be 0.45, 6.4 and 9.72, respectively, and a Kd of 220 nM was used for Fura II. Baseline [Ca2+]i responses to 3-s pulses of 200 /tm N M D A were collected at approximately 7-min intervals and the effect of a 7-min perfusion of 20 /~M 7chlorokynurenate (C1-Kyn; 7-chloro-4-hydroxy-quinoline-2-carboxylic acid; Tocris Neuramin, Essex, U.K.) was assessed on NMDA-induced increases in [Ca2+]i. In several experiments, C1-Kyn, D-AP5 or ketamine were simultaneously perfused with glycine (40/~M) or o-serine (50/tM) to test whether these latter compounds could reverse the effects of the antagonists.
199
The mean (___ S.E.M.) basal levels of [Ca2+]i from 16 neurons was 122 + 13 nM and N M D A application resulted in an average increase of 363 + 34 nM. Perfusion of 20 #M CI-Kyn reduced the NMDA-mediated [Ca2+]i enhancement to 29 + 8 nM (n = 9). A typical dose-response experiment demonstrating the effect of CI-Kyn on N M D A responses in a single neuron is shown in Fig. 1. Consecutive pulses of 200/zM NMDA (squares) increased [Ca2+]i from a baseline value of 71 nM to an average maximum of 644 nM (i.e., a change of 573 nM). The perfusion of 5 #M, 10 aM and 20 #M CI-Kyn (bars) caused a sequential decrease in NMDA-induced alterations in [Ca2+]i to 230 nM, 180 nM and 5nM, respectively. In order to determine whether Cl-Kyn was acting on the NMDA-associated glycine site or some other binding sites (e.g., the NMDA receptor) a series of experiments were carried out using glycine, D-serine, D-AP5 and ketamine. Figure 2 shows the effect of 40/zM glycine on the CI-Kyn-induced depression of the NMDA [Ca2+]i response. A control response to NMDA exhibited a 536 nM increase in [Ca2+]i that was almost entirely eliminated by 20/zM CI-Kyn (increase in [Ca2+]i = 24 nM). Perfusion of glycine (40 #M), in the continued presence of C1-Kyn, reversed the antagonism in that [Ca2+]i increased to 449 nM when N M D A was applied. Glycine and D-serine had comparable effects and reversed C1-Kyn antagonism by 89 + 5% (n = 9). NMDA-mediated increases in [Ca2+]i were also blocked by the receptor antagonist D-AP5 and the channel blocker ketamine (mean change in [Ca2+]i --- 13 + 4 nM; n= 5). Figure 3 shows the increase (mean = 326 nM) in [Ca2+]i in a single neuron. Addition of 40 #M D-AP5 to the perfusion medium for a period of 7 min completely blocked the N M D A response and this effect was unaltered by 50 #M gly8OO ¢-
CL-KYN 5 uM '
10 uM I
600
/
20 uM
,--I
I
L) LU LU cc IJ.
I
400
n,< .-I
,_1 .J
200
UJ
C~ nz --
ta
0
D
i
0
J J
10
i
20
i
30
ta i
40
50
MINUTES Fig. I. The effect of C1-Kyn on the NMDA-mediated increase in [Ca2+]~. In normal medium, the first two 3-s applications of 200 aM NMDA (squares) resulted in a 573-nM mean enhancement of [Ca2+]i. Peffusion of 5, 10 and 20/zM C1-Kyn antagonized subsequent pulses of NMDA, with 20/zM CI-Kyn completely blocking any response to the agonist.
200
~"
700
GLYCINE
t-
CL-KYN
I
600 ' L.) ,-J
500
< o uJ
w
400
CE LL
a:
