Neuron, Vol. 5, 247-253,September,1990,Copyright © 1990by Cell Press
Mechanisms Generating the Time Course of Dual Component Excitatory Synaptic Currents Recorded in Hippocampal Slices Shaul Hestrin,* Pankaj Sah*t~ and Roger A. Nicoll *t * Department of Physiology tDepartment of Pharmacology University of California, San Francisco San Francisco, California 94143
Summary We studied with the whole-cell recording techniques, the mechanisms underlying the time course of the slow N-methyI-D-aspartate (NMDA), and fast non-NMDA receptor-mediated excitatory postsynaptic currents (EPSCs) in hippocampal slices. The rising phase of the NMDA receptor-mediated component of the EPSC as well as the decaying phase of the NMDA and non-NMDA component were highly temperature-sensitive, suggesting that neither of these processes is determined by free diffusion of transmitter. Moreover, glutamate uptake blockers enhanced the responses to exogenously applied glutamate, but had no effect on the decay of either the NMDA or non-NMDA components of the EPSCs. On the other hand, open channel blockers known to modify NMDA channel kinetics reduced the EPSC decay time. Thus, the present results support a model in which the rise time and decay of the NMDA component are determined primarily by slow channel kinetics and the decay of the non-NMDA component is due either to channel kinetics or to desensitization. Introduction Chemical synaptic transmission either exhibits rapid onset, generated by direct activation of ion channels, or is slow, generated by receptors that are indirectly coupled to ion channels (Hille, 1984; Kandel and Schwartz, 1985; Nicoll, 1988). At virtually all synapses examined that utilize receptors which directly gate ion channels, the rise in transmitter concentration in the synaptic cleft is very brief and the duration of the synaptic current is determined by the closure of the receptor channel. This includes synapses that utilize the muscle (Magleby and Stevens, 1972a) and neuronal (MacDermott et al., 1980; Rang, 1981) nicotinic acetylcholine receptor, the glutamate receptor on crustacean muscle (Crawford and McBurney, 1977), glycine receptors (Faber and Korn, 1982) and y-aminobutyric acid (GABAA) receptors (Barker and McBurney, 1979; Collingridge et al., 1984) in CNS neurons. The glutamatergic synapses in the CNS differ in two striking ways from the synapses discussed above. First, the synaptically related glutamate activates two synaptic currents, an NMDA and a non-NMDA component, both of which result from receptors that are Presentaddress: Department of Physiologyand Pharmacology, University of Queensland, Saint Lucia Q. L. D. 4067, Australia.
directly coupled to ion channels. Second, the NMDA component, unlike all other synaptic currents resulting from receptors that are directly coupled to ion channels, has a remarkably slow rise time and long duration (for review see Collingridge and Lester, 1989). At CA1 hippocampal pyramidal neurons the nonNMDA component of the excitatory postsynaptic current (EPSC) lasts for only a few milliseconds, whereas the NMDA component requires several milliseconds to reach its peak and can last for more than a 100 ms. Given that both receptor subtypes can be colocalized at the postsynaptic membrane (Bekkers and Stevens, 1989), it is unclear what underlies the slow time course of the NMDA component. The open time of single NMDA-activated channels has been reported to be 5-15 ms (Cull-Candy and Usowicz, 1987; Jahr and Stevens, 1987; Ascher et al., 1988), although examples of long-lasting bursts of openings have been observed (Jahr and Stevens, 1987; Howe et al., 1988). We have examined the various possibilities using whole-cell recording from CA1 hippocampal pyramidal cells in slices. Our findings favor a model in which the synaptic rise in glutamate concentration is brief and the rising and decaying phases of the NMDA synaptic current are determined by the slow opening and slow closing of the NMDA receptor channels. Results Comparison of the Time Course of the NMDA and non-NMDA EPSCs Stimulation of the afferent fibers led to the generation of an EPSC that contained both a slow rising and decaying NMDA component and a fast rising and decaying non-NMDA components. The two components could be isolated using known pharmacological properties and voltage dependence (Hestrin et al., 1990). The NMDA component could be recorded in isolation by applying 6-cyano-7-nitroq uinoxaline-2 (CNQX) and holding the cell at -40 mV or at -80 mV in Mg2+-free Ringer solution. The non-NMDA component could be isolated by holding the cell at potentials more negative than -80 mV or by applying the NMDA receptor antagonist DL-2-amino-5-phosphonovalerate (APV). Figure 1A shows an EPSC recorded at -40 mV before and after blocking the non-NMDA component with CNQX. While the decay of the NMDA component at room temperature can be fitted with a single exponential function (Hestrin et al., 1990), the decay of the NMDA component of the EPSC recorded at temperatures above 30°C typically could not accurately be fitted with a single exponential function. A semi-logarithmic plot of the decay of the NMDA component clearly indicates that initially there is a rapid decay followed by a slow phase (Figure 1B). The EPSC could be fitted with a sum of two exponential functions. In six cells, recorded at 30°C-32°C, the fast
Neuron 248
A
_----
C _ ~
glutamate release and thus depend on diffusion for their activation.
0m"
/
V C0nt
°I
B
150 pA
100 pA
m 100 ~
]
D
m----~10
E
Mechanisms Generating the Time Course of Dual Component Excitatory Synaptic Currents Recorded in Hippocampal Slices Shaul Hestrin,* Pankaj Sah*t~ and Roger A. Nicoll *t * Department of Physiology tDepartment of Pharmacology University of California, San Francisco San Francisco, California 94143
Summary We studied with the whole-cell recording techniques, the mechanisms underlying the time course of the slow N-methyI-D-aspartate (NMDA), and fast non-NMDA receptor-mediated excitatory postsynaptic currents (EPSCs) in hippocampal slices. The rising phase of the NMDA receptor-mediated component of the EPSC as well as the decaying phase of the NMDA and non-NMDA component were highly temperature-sensitive, suggesting that neither of these processes is determined by free diffusion of transmitter. Moreover, glutamate uptake blockers enhanced the responses to exogenously applied glutamate, but had no effect on the decay of either the NMDA or non-NMDA components of the EPSCs. On the other hand, open channel blockers known to modify NMDA channel kinetics reduced the EPSC decay time. Thus, the present results support a model in which the rise time and decay of the NMDA component are determined primarily by slow channel kinetics and the decay of the non-NMDA component is due either to channel kinetics or to desensitization. Introduction Chemical synaptic transmission either exhibits rapid onset, generated by direct activation of ion channels, or is slow, generated by receptors that are indirectly coupled to ion channels (Hille, 1984; Kandel and Schwartz, 1985; Nicoll, 1988). At virtually all synapses examined that utilize receptors which directly gate ion channels, the rise in transmitter concentration in the synaptic cleft is very brief and the duration of the synaptic current is determined by the closure of the receptor channel. This includes synapses that utilize the muscle (Magleby and Stevens, 1972a) and neuronal (MacDermott et al., 1980; Rang, 1981) nicotinic acetylcholine receptor, the glutamate receptor on crustacean muscle (Crawford and McBurney, 1977), glycine receptors (Faber and Korn, 1982) and y-aminobutyric acid (GABAA) receptors (Barker and McBurney, 1979; Collingridge et al., 1984) in CNS neurons. The glutamatergic synapses in the CNS differ in two striking ways from the synapses discussed above. First, the synaptically related glutamate activates two synaptic currents, an NMDA and a non-NMDA component, both of which result from receptors that are Presentaddress: Department of Physiologyand Pharmacology, University of Queensland, Saint Lucia Q. L. D. 4067, Australia.
directly coupled to ion channels. Second, the NMDA component, unlike all other synaptic currents resulting from receptors that are directly coupled to ion channels, has a remarkably slow rise time and long duration (for review see Collingridge and Lester, 1989). At CA1 hippocampal pyramidal neurons the nonNMDA component of the excitatory postsynaptic current (EPSC) lasts for only a few milliseconds, whereas the NMDA component requires several milliseconds to reach its peak and can last for more than a 100 ms. Given that both receptor subtypes can be colocalized at the postsynaptic membrane (Bekkers and Stevens, 1989), it is unclear what underlies the slow time course of the NMDA component. The open time of single NMDA-activated channels has been reported to be 5-15 ms (Cull-Candy and Usowicz, 1987; Jahr and Stevens, 1987; Ascher et al., 1988), although examples of long-lasting bursts of openings have been observed (Jahr and Stevens, 1987; Howe et al., 1988). We have examined the various possibilities using whole-cell recording from CA1 hippocampal pyramidal cells in slices. Our findings favor a model in which the synaptic rise in glutamate concentration is brief and the rising and decaying phases of the NMDA synaptic current are determined by the slow opening and slow closing of the NMDA receptor channels. Results Comparison of the Time Course of the NMDA and non-NMDA EPSCs Stimulation of the afferent fibers led to the generation of an EPSC that contained both a slow rising and decaying NMDA component and a fast rising and decaying non-NMDA components. The two components could be isolated using known pharmacological properties and voltage dependence (Hestrin et al., 1990). The NMDA component could be recorded in isolation by applying 6-cyano-7-nitroq uinoxaline-2 (CNQX) and holding the cell at -40 mV or at -80 mV in Mg2+-free Ringer solution. The non-NMDA component could be isolated by holding the cell at potentials more negative than -80 mV or by applying the NMDA receptor antagonist DL-2-amino-5-phosphonovalerate (APV). Figure 1A shows an EPSC recorded at -40 mV before and after blocking the non-NMDA component with CNQX. While the decay of the NMDA component at room temperature can be fitted with a single exponential function (Hestrin et al., 1990), the decay of the NMDA component of the EPSC recorded at temperatures above 30°C typically could not accurately be fitted with a single exponential function. A semi-logarithmic plot of the decay of the NMDA component clearly indicates that initially there is a rapid decay followed by a slow phase (Figure 1B). The EPSC could be fitted with a sum of two exponential functions. In six cells, recorded at 30°C-32°C, the fast
Neuron 248
A
_----
C _ ~
glutamate release and thus depend on diffusion for their activation.
0m"
/
V C0nt
°I
B
150 pA
100 pA
m 100 ~
]
D
m----~10
E
