Masanobu Kano
1 Introduction
Activity-dependent long-term modification of transmission efficacy at synapses is thought to be a cellular basis of learning and memory (Kandel & Schwartz, 1982). Since the discovery of long-term potentiation (LTP) in the hippocampus (Bliss & Lomo, 1973), synapses that undergo plastic change have been described in various part of the brain, e.g., the visual cortex (Komatsu, Toyama, Maeda & Sakaguchi, 1981), motor cortex (Iriki, Pavlides, Keller & Asanuma, 1989), red nucleus (Tsukahara, Hultborn, Murakami & Fujito, 1975), superior colliculus (Okada & Miyamoto, 1989), and cerebellum (Ito, Sakurai & Tongroach, 1982). In the cerebellar cortex, long-term depression (LTD) of excitatory parallel fiber (PF) synapse was first described by Ito et al. (1982), and this is thought to be a cellular basis of motor learning in the cerebellum (for review, see Ito 1989; Linden 1994).
It is somewhat surprising that modifiable synapses discovered and studied to date are mostly glutamatergic excitatory synapses. Despite the importance of inhibition in brain functions, plasticity at inhibitory synapses has not been demonstrated so far except some related to epilepsy (Stelzer, Slater & ten Bruggencate, 1987). Recently, it was revealed that GABAA receptor-mediated inhibitory synaptic transmission in Purkinje cells (PCs) of the cerebellum undergoes long-lasting potentiation (Llano, Leresche & Marty, 1991b; Kano, Rexhausen, Dreessen & Konnerth, 1992; Vincent, Armstrong & Marty, 1992). Llano et al. (1991b) and Vincent et al. (1992) reported that inhibitory synaptic transmission and responsiveness of PCs to exogenous GABA are potentiated after depolarization-induced Ca2+ entry to PCs. A robust and long-lasting potentiation of inhibitory synaptic transmission is induced following stimulation of excitatory climbing fibers (CFs), (Kano et al. 1992), that was termed "rebound potentiation (RP)". This newly- found synaptic plasticity may play a role in the motor learning in concert with LTD at excitatory PF synapses. In this article, I will describe the properties and discuss possible mechanism of RP based on the data accumulated so far.
2. Properties of RP
RP is found by using whole-cell patch-clamp recording from visually identified PCs in cerebellar slices (Edwards, Konnerth, Sakmann & Takahashi, 1989; Konnerth, 1990; Kano & Konnerth. 1992b). This newly developed technique allows high-resolution recordings of membrane currents in neurons from slice preparations. Under whole-cell recording with the pipette solution containing CsCl (80mM), Cs D-gluconate (80mM), MgCl2 (2mM), EGTA (1mM), Na-ATP (4mM), Na-GTP (0.4mM) and HEPES (10mM) (pH 7.3), PCs have spontaneously occurring inward synaptic currents that are blocked in the presence of bicuculline (10mM) (Konnerth, Llano & Armstrong, 1990; Llano, Marty, Armstrong & Konnerth, 1991c), which identifies them as GABAA-receptor mediated inhibitory postsynaptic currents (IPSCs). These IPSCs are recorded at a constant frequency. However, it is often observed that the IPSC amplitudes tend to decay slowly even with ATP and GTP in the patch-pipette (see prestimulus period in Figs. 1A and 2). Under whole cell recording conditions, stimulation of CFs induced large all-or-none excitatory postsynaptic currents (EPSCs) in PCs that are accompanied by characteristic regenerative responses (Llano et al. 1991c). CF stimulation (5 shocks at 0.5Hz) readily induced a marked potentiation of IPSCs (Fig. 1A) in all cells tested (see Table 1). This phenomenon was termed 'rebound potentiation' because it is the enhancement of inhibition triggered by strong excitation of PCs, and it counteracts the tendency of 'run down' that is observed in some cells (see Figs. 1A and 2). The amplitude of RP was 139+10.9% (mean+S.E.M., n=6, measured 10 min after the conditioning CF stimulation) of the control, and the time to half-recovery was 16.5+3.6 min (n=5) (Table 1). The RP reaches its maximum in 3-15min after CF stimulation and a plateau-like phase follows (Fig. 1A).
3. Transient elevation of [Ca2+]i is required for RP to occur
Stimulation of CFs is known to induce large increases in [Ca2+]i in PCs (Ross & Werman 1987; Knpfel, Vranesic, Staub & Gahwiler, 1991; Konnerth, Dreessen & Augustin, 1992; Miyakawa, Lev-Ram, Lasser-Ross & Ross, 1992) that are spatially similar to depolarization-induced [Ca2+]i transients (Ross & Werman 1987; Tank, Sugimori, Connor & Llinas, 1988; Ross, Lasser-Ross, & Werman, 1990; Sugimori & Llinas, 1990; Lev- Ram, Miyakawa, Lasser-Ross & Ross, 1992). LTD of excitatory PFs are shown to be triggered by this CF-induced elevation of [Ca2+]i (Sakurai, 1990; Hirano, 1990b, Konnerth et al. 1992). The possibility that postsynaptic Ca2+ is required for RP is tested in the following experiments. When PCs were filled with a solution containing Ca2+ chelator BAPTA (30mM), CF stimulation failed to induce RP (Fig. 1B). Furthermore, activation of voltage-gated Ca2+ channels by a single depolarizing pulse (from a holding potential of -70mV to 0mV, duration of 500ms) induced a long-lasting potentiation of IPSCs (Fig. 1C) to an extent and with a time course similar to that of CF-induced RP. Stimulation of PC axons in the granule cell layer induced Na+ action currents. These currents did not accompany characteristic long-lasting Cl- tail currents that follow the Ca2+ entry through voltage-gated Ca2+ channels (Ca2+-activated Cl- currents, see Llano et al. 1991b). This indicates that Na+ action currents did not cause significant elevation of [Ca2+]i. In fact, antidromic activation of Na+ spikes with the same stimulation pattern (5 shocks, 0.5 Hz) as that used for CF stimulation did not induce any potentiation of IPSCs (Fig. 1D). Taken together, these findings strongly suggest that a CF-induced rise in [Ca2+]i through voltage-gated Ca2+ channels triggered RP.
The correlation between the [Ca2+]i elevation and RP was directly examined by combining whole-cell patch clamp recording with fluorometric [Ca2+]i measurement (Fig. 2). PCs were filled with the Ca2+ indicator dye Fura-2 (200mM) through the patch-pipette and [Ca2+]i was continuously monitored by digital fluorescence video ratio imaging technique (Llano, Dreessen, Kano & Konnerth, 1991a). Fig. 2C shows that the basal level of [Ca2+]i in this cell had similar values in the soma and in the proximal dendrites of around 100 nM. On the conditioning CF stimulation, the dendritic [Ca2+]i rose steeply to about 900 nM (Fig. 2C) and stayed elevated for about 40 s, decaying then rapidly back to within about 30 nM of its initial baseline value. The [Ca2+]i in the somatic region rose only slightly to less than 200 nM (Figs. 2C). A similar pattern of [Ca2+]i change was observed in all of 5 PCs tested. After the transient rise, the [Ca2+]i in both proximal dendrites and soma maintained a stable baseline value for more than 40 min (Fig. 2C). On the other hand, the simultaneously measured IPSCs were markedly potentiated (Fig. 2A). These results indicate that RP of IPSCs is triggered by a transient elevation of [Ca2+]i in the PC dendrites, but that RP is not maintained by a persistent elevation of [Ca2+]i following CF stimulation.
Vincent et al. (1992) reported that IPSCs of PCs evoked by stimulation of putative basket cells and their axons undergo potentiation following repetitive depolarization of PCs (8 pulses of 100ms duration to 0mV at 0.5 Hz) under the recording conditions similar to those of Kano et al. (1992) (i.e., whole-cell patch clamp recording from PCs in cerebellar slices with CsCl in the patch-pipette). The peak amplitude of potentiation was 146.7+4.1% (mean+S.E.M., n=4) of the control, and the time to half-recovery was 12.9+0.5min (n=3) (Table 1). They showed that IPSCs can be fitted with two exponentials. The time constants for both the rising (ton) and decaying (toff) phases did not change during the potentiation. It should be noted that, regarding depolarization-induced potentiation of IPSCs, the peak amplitude and time to half recovery reported by Vincent et al. (1992) are comparable to those of Kano et al. (1992) (Table 1). Moreover, the values of the depolarization-induced potentiation and those of CF-induced RP are similar (Table 1). This coincidence further supports that CF-induced rise in [Ca2+]i through voltage-gated Ca2+ channels triggered RP.
Besides long-lasting RP, IPSCs are depressed transiently (for about 1min) following depolarization-induced Ca2+ entry to PCs (Llano et al. 1991b; Vincent et al. 1992; Vincent & Marty, 1993). This short-lasting depression is mainly presynaptic, since the frequency of spontaneous IPSCs decreased (Llano et al. 1991b) and the probability of failures of evoked IPSCs was greatly enhanced during the period of the depression (Vincent et al. 1992).
4. RP involves up-regulation of postsynaptic GABAA receptor function
Several lines of evidence indicate that the inhibitory transmitter used by cerebellar cortical interneurons is GABA (Ito, 1984 for review). The cause of RP may be either the enhancement of GABA release from the presynaptic terminals or the sensitivity increase at the postsynaptic receptors to GABA. The latter possibility was tested by measuring membrane currents induced by bath-applied GABA (2 mM for 10s) (Fig. 2B). At the same time, RP of IPSCs induced by CF stimulation was monitored (Fig. 2A). Following CF stimulation these currents induced by exogenous GABA were strongly enhanced. The time course of this potentiation was similar to that of RP of the IPSCs (Figs. 2A, 2B, see Table 1 for summary). Although the popssible contribution of extrasynaptic GABAA receptors to this potentiation remains, this finding supports that RP involves an up-regulation of postsynaptic GABAA receptor function. As in the control experiments for the RP of IPSCs, PC responses to bath-applied GABA were not potentiated following antidromic stimulation (5 pulses, 0.5Hz). In contrast to the marked enhancement of postsynaptic sensitivity to GABA, no significant change in IPSC frequency could be detected during RP. The relative frequencies of IPSCs obtained 20 min after conditioning (in percentage of the control values, mean+S.E.M.) were: 95 6.7 % (n=5) for CF stimulation, 105 9.5 % (n=4) for depolarization, 99 4.5 % (n=6) for CF stimulation with intracellular BAPTA, and 100 1.7 % (n=3) for the control experiments with antidromic stimulation. These results suggest that RP is mainly attributable to increased sensitivity of postsynaptic GABAA receptors on PCs.
Llano et al. (1991b) reported that GABA-induced currents are enhanced after repetitive depolarization of PCs (8 pulses of 100ms duration to 0mV at 0.5 Hz), a protocol that is also used to induce potentiation of evoked IPSCs (Vincent et al. 1992). However, the time to half recovery reported by Llano et al. (1991b) is significantly shorter than that of Kano et al. (1992) (see Table 1). One possibility for this difference would be due to the different methods for GABA application used in these two studies. Llano et al. (1991b) applied GABA presumably to the soma of PCs by "U-tube method." On the other hand, Kano et al. adopted simple bath-application and measured GABA- induced currents originating from both the dendrites and soma. Since large patch-pipettes (resistance, around 2MW) are used for recording from PCs, some intracellular factors necessary for a long-lasting up-regulation of GABAA receptors might have been "washed out" from the soma, but not from the dendrites. Llano et al. (1991b) may have measured GABA-mediated currents originating mainly from the soma that did not show long-lasting potentiation under whole-cell recording conditions. There is no other clear explanation for the difference, but this is an issue that should be addressed in future experiments.
5. Activity of inhibitory synapses are necessary for RP
RP can be induced by a few numbers of CF impulses or even a single depolarizing pulse. This is apparently quite in contrast to the stimulation parameter required to induce LTD of excitatory PF synapses. For LTD to occur, repetitive stimulation of CFs (50-960 pulses at 2 to 4 Hz) is necessary (Ito et al. 1982; Ekerot & Kano, 1985; Kano & Kato, 1987; Sakurai 1987, 1990, Hirano 1990a). Moreover, PFs have to be stimulated or postsynaptic quisqualate receptors should be activated in conjunction with CF stimulation (Ito et al. 1982; Ekerot & Kano, 1985; Kano & Kato, 1987; Sakurai 1987, 1990, Hirano 1990a) or with depolarization of PCs (Hirano 1990b; Crepel & Jaillard 1991; Linden & Connor, 1991; Linden, Dickinson, Smeyne & Connor 1991; Daniel, Hemart, Jaillard & Crepel, 1992). CF stimulation alone does not induce LTD.
Since CFs fire tonically at about 0.5-1Hz in vivo, it is puzzling that such a strong and long-lasting potentiation as RP can be produced by so few numbers of CF impulses. Measurement of [Ca2+]i in PCs revealed that, under whole-cell patch-clamp recording with cesium as the major intracellular cation, Ca2+ transients induced by CF stimulation or depolarization are large in amplitude (up to 1mM) and very long in duration (longer than 10s in many cases) (Kano et al. 1992; Konnerth et al. 1992; see Fig. 2C). By contrast, CF-induced Ca2+ transients have peak amplitudes of about 100nM and durations of 200-300ms when measured under conventional intracellular recording with potassium in the recording electrode (Knpfel et al. 1991). Therefore, a single CF stimulation with cesium inside might be equal to several tens of repetitive CF impulses under recording conditions with potassium inside. Moreover, since IPSCs occur spontaneously with rather high frequency (Konnerth et al. 1990; Llano et al. 1991b; see insets of Fig1), CF stimulation alone with cesium inside may result in the conjunction of inhibitory synaptic activity and elevation of postsynaptic Ca2+. Therefore, the difference in stimulation parameters for the induction of LTD and RP cannot directly be compared.
In recent experiments in the authir's laboratory, RP was induced using whole-cell patch-clamp recording with the pipette solution containing KCl (70mM), K D-gluconate (70mM), MgCl2 (2mM), Na-ATP (4mM), Na-GTP (0.4mM), EGTA (0.2mM) and HEPES (30mM) (pH 7.3). In comparison with the pipette solution used in the experiments shown in Figs 1, 2, cesium was replaced by potassium. Because of the difficulty in stimulating inhibitory interneurons and their axons in isolation from excitatory PFs, CNQX (10mM) and APV (50mM) was used to block excitatory transmission mediated by both NMDA and non-NMDA receptors. IPSCs were recorded as inward synaptic currents at the stimulation frequency of 0.2Hz (Fig. 3A). Following repetitive depolarization of PCs (240 pulses of 50ms duration to -10mV at 4 Hz, from the holding potential of -60mV) in conjunction with stimulation of inhibitory interneurons, amplitudes of evoked IPSCs were potentiated (Fig. 3A). Interestingly, repetitive depolarization alone or stimulation of inhibitory interneurons alone at 4Hz for 1 min failed to induce a long-lasting potentiation of IPSCs (Fig. 3B). Conjunctive repetitive depolarization with stimulation of interneurons potentiated the averaged amplitude of IPSCs to 129.3+9.7% (mean+S.E.M., n=10) of control. By contrast, either repetitive depolarization alone (94.7+14.2%, n=8) or stimulation of inhibitory interneurons alone (97.6+12.1%, n=8) was ineffective to induce potentiation (Fig.3B). These results suggest that, under recording condition with potassium inside, association of inhibitory synaptic activity with depolarization of postsynaptic membrane and accompanying Ca2+ elevation are required for triggering RP. It is still unclear whether potentiation of IPSCs can occur in physiological conditions with CF stimulation instead of depolarizing pulses. Ca2+-imaging experiments show that both CF stimulations and depolarizing pulses induce similar pattern of [Ca2+]i transients due to activation of voltage-gated Ca2+ channels (Llano et al. 1991a; Kano et al. 1992). This suggests that CF stimulation also can produce potentiation of IPSCs. It should also be noted that the parameter is comparable to that required for the induction of LTD of excitatory PF synapses. Thus, these lines of evidence suggest the possibility that RP and LTD work in vivo in a cooperative way to control PC excitability.
6. Activation of protein kinase A induces up-regulation of GABAA receptor-functions
Experiments using extracellular recordings from PCs in vivo showed that 8-bromo-cAMP and forskoline reversibly potentiated the inhibitory effect of iontophoretically applied GABA on the firing of PCs (Sessler, Mouradian, Cheng, Yeh, Liu & Waterhouse, 1989; Parfitt, Hoffer & Bickford-Wimer, 1990), suggesting a modulation of GABA receptors by cAMP- dependent protein kinase (PKA). Therefore, the effect of PKA on PCs was investigated directly by using the whole-cell patch- clamp recording from PCs in cerebellar slices (Kano & Konnerth, 1992a).
Perfusing the slices with a saline containing 8-bromo- cAMP (500 mM, for 15 minutes), a membrane-permeable analogue of cAMP, significantly potentiated responses of PCs to bath-applied GABA (2-5 mM, 10 s) (Fig. 4A, upper panel). The magnitude of the potentiation was in average about 140 % of the control value before application of 8-bromo-cAMP (Fig. 4B). The potentiation always persisted until the end of the experiment. By contrast, when the internal solution contained protein kinase inhibitor peptide (PKIP, 400 mg/ml), a specific inhibitor of PKA (Cheng, Kemp, Pearson, Smith, Misconi, Van Patten & Walsh, 1986), the same manipulation failed to induce potentiation of whole-cell GABA current (Fig. 4A, lower panel). Instead, a slight reduction of the amplitudes of the currents was observed. This reduction did not seem to be due to side effects of PKIP, since a similar 'run-down' of the whole- cell current response to GABA was observed in control experiments without any treatment (Fig. 4B, control).
Besides the potentiation of currents induced by bath- applied GABA, 8-bromo-cAMP produced a similar potentiation of the amplitudes of spontaneous IPSCs measured in the presence of 0.5mM TTX (sIPSC, Fig. 5A, upper panel for specimen record; Fig. 5B for summary). The threshold for detecting the sIPSCs (upward arrow in the lower panel of Fig. 5A) was set to -10 pA so that it was well over the background noise level. This potentiation lasted as long as the potentiation of whole-cell GABA current. The inclusion of PKIP in the patch-pipette prevented the potentiation of sIPSCs (Fig. 5B) similar to current responses induced by bath-applied GABA. It is important to note that bath-application of 500 mM cAMP (the membrane impermeable compound) affected neither GABA- mediated whole-cell current responses (Fig. 4B) nor sIPSCs (Fig. 5B), indicating that cAMP has only an intracellular site of action in PCs. The parallelism between the potentiation of the GABA- mediated currents and that of sIPSCs suggests that the same mechanism is mediating both effects (Figs 4B, 5B). Moreover, the percentage changes of the GABA-mediated currents are strongly correlated with those of the sIPSCs (Fig. 6, p<0.01 by Spearman's rank correlation test, r=0.77), supporting that both reflect the same effect, namely the change in the sensitivity of postsynaptic GABAA receptors.
The results presented above strongly suggest that GABAA receptors of cerebellar PCs are up-regulated by PKA through phosphorylation of the receptor protein itself or closely related protein(s). Thus, PKA is a strong candidate to play a role in RP. However, at present, a link between the elevation of Ca2+ and activation of PKA is unclear. To clarify whether PKA is involved in RP, blocking effects of PKA inhibitors (like PKIP) on Ca2+- induced RP should be examined. Furthermore, it should also be tested whether PKA activation occludes the subsequent induction of RP.
GABAA receptor function is modulated in different ways in different regions of the nervous system. Interestingly, Ca2+ elevation has opposite effect on GABAA receptors in other cell types so far examined. In hippocampal (Chen, Stelzer, Kay & Wong, 1990), habenular (Mulle, Choquet, Korn & Changeux, 1992) and dorsal root ganglion neurons (Inoue, Oomura, Yakushiji & Akaike, 1986), Ca2+ elevation depresses GABA responses. Effects of PKA are also opposite in hippocampal (Harrison and Lambert, 1989), spinal cord (Porter, Twyman, Uhler & MacDonald, 1990) and sympathetic ganglion (Moss, Smart, Blackstone & Huganir, 1992b) neurons. This suggests either that PCs possess GABAA receptors with a special molecular structure that undergo modulation in an opposite way to that for other cell types, or that GABAA receptor functions are regulated by yet unknown mechanisms that are influenced by Ca2+ or PKA in a way different from other cell types.
PKC and PKG are also rich in PCs and shown to be involved in LTD of excitatory PF synapses. (Crepel & Kruppa, 1988; Crepel & Jaillard, 1990; Ito & Karachot, 1990, 1992; Shibuki & Okada, 1991; Linden & Connor, 1991). However, no data are available to date that positively suggest the involvement of these kinase systems in RP of inhibitory synapses. In oocyte expression systems, PKC has been shown to depress GABAA receptors by directly phosphorylating the receptor protein (Whiting, McKernan & Iversen, 1990; Sigel, Baur & Malherbe, 1991; Moss, Doherty & Huganir, 1992a). Thus, it is possible that PKC is somehow involved in the regulation of GABAA receptor functions in PCs.
7. Plasticity of inhibitory synapses in other synapses
In the hippocampus, plasticity of inhibitory synapses was studied in relation to the mechanism of epilepsy, so that GABAergic inhibitory transmission was depressed during kindling (Stelzer et al. 1987). However, recently appeared reports in some brain regions suggest physiological role of plasticity of inhibitory synapses (Kano, 1995, for review). Korn, Oda and Faber (1992) reported that, in the goldfish, tetanic stimulation induces LTP at glycinergic inhibitory synapses on the Mauthner cells. Morishita and Sastry (1993) found the LTD of IPSCs of cells in the deep cerebellar nuclei induced by tetanic stimulation of PC axons. Komatsu and Iwakiri (1993) showed that, in the visual cortex of young rat, LTP of inhibitory transmission is induced by tetanus of inhibitory inputs, while NMDA receptor activation leads to LTD of the same synapses. Recently, Mitoma et al. (1994) reported that bath-application of serotonin causes enhancement of inhibitory synaptic transmission in cerebellar Purkinje cells, which is presumably via presynaptic mechanisms. Thus, it appears that plasticity of inhibitory synapses exists in various parts of the brain, and that it may have important physiological functions such as learning and development.
8. Conclusion
In this article, I have reviewed the properties and possible mechanisms of RP of inhibitory synapses of PCs, a newly found neural plasticity in the cerebellar cortex. The well-known form of synaptic plasticity in the cerebellum is LTD of excitatory PF- PC transmission. Evidence has been accumulated to support that LTD is a cellular basis of motor learning in the cerebellum (for review, see Ito 1989; Linden 1994). LTD is induced by PF stimulation in conjunction with CF stimulation or depolarization of postsynaptic membrane. It appears that RP also requirs conjunction of inhibitory synaptic activity and postsynaptic depolarization. The parameter for the induction of RP (4Hz, 240 conjunctive stimuli) is comparable to that of LTD (2-4Hz, 100-960 conjunctive stimuli). Thus it is likely that, in concert with LTD of excitatory PF synapses, RP of GABAergic inhibitory synapses also play a role in motor learning.
The initial trigger for RP is CF activity that causes an elevation of [Ca2+]i due to Ca2+ inflow through voltage gated Ca2+ channels. RP is expressed mainly as an up-regulation of postsynaptic GABAA receptors. PKA might play a role in the induction of RP, since activation of PKA by a cAMP analog induced potentiation of GABAA receptor function similar to RP. On the other hand, no evidence is available so far that suggest the involvement of PKC and PKG in RP, although both of these kinases are shown to be involved in LTD of excitatory PF synapses (Crepel & Kruppa, 1988; Crepel & Jaillard, 1990; Ito & Karachot, 1990, 1992; Shibuki & Okada, 1991; Linden & Connor, 1991). Thus, it is likely that two distinct processes are present in PCs that lead to RP of inhibitory, and LTD of excitatory PF synapses.
What are implications of plasticity at inhibitory synapses to functions of the cerebellum? This newly found memory device will give the cerebellar cortical circuitry an additional computational performance. In comparison with the only one site of plasticity at excitatory PF synapses, wider-range control of PC excitability by CFs would be possible. Some 30% reduction of PF-mediated excitation occurs during LTD. If RP causes some 30% increase of inhibition at the same time, the sum would result in stronger depression of PC excitability than that achieved by LTD alone. Moreover, LTD and RP may not summate simply, because the sites of excitatory PF synapses and inhibitory basket/stellate cell synapses are spatially different. Some 10000 PFs form synapses on dendritic spines of a single PC, while basket-cell and stellate-cell axons make relatively sparse contacts on the soma and proximal dendrites respectively (Ito, 1984). This suggests that LTD of PF synapses is used for the fine control, while RP of inhibitory synapses is used for the robust and rough control of PC excitability.
The link between RP to motor learning is currently missing. Direct demonstration by the single-unit analysis in behaving animals is not easy, since it is very difficult to distinguish between the reduction of excitation and the enhancement of inhibition. A pharmacological approach would be a possible alternative. This was taken to link LTD of excitatory PF synapses to motor learning. In cerebellar slice preparation, LTD is shown to be blocked by hemoglobin (a NO scavenger). When hemoglobin was injected to the flocculus of monkeys, these animals showed no adaptation of the vestibuloocular reflex (Nagao & Ito, 1992). Moreover, when hemoglobin was injected into the cerebellar vermis of walking decerebrate cats, these animals showed no adaptation to limb perturbation (Yanagihara, Kondo & Yoshida 1994). By analogy of these works, if some pharmacological agents are found that specifically blocks RP, and if the agents act from the extracellular space, they can be tested in behaving animals. Besides pharmacological agents, a recently developed gene-targeting technique will provide new experimental tools. Genetically created mutant mice that lack a specific molecule such as a- CaM-KII and PKC-g have already been used to link hippocampal LTP and spatial learning (Silva et al., 1992a, 1992b; Grant et al., 1992; Abeliovich et al., 1993a, 1993b; Sakimura et al., 1994). Recently, this approach was also applied for linking LTD to cerebellar functions (Aiba et al., 1994; Conquet et al., 1994) Although it is not easy to assess the cerebellar functions in mice at the behavioral level, mutant mice would be a powerful new tool for studying the link between RP and motor learning.
ACKNOWLEDGEMENTS The author thanks Dr. N. Kawai for his kind support for this work. This work was partly supported by grants from the Japanese Ministry of Education, Science and Culture (05454677, 05260221, 05267242), and a garnt from the Brain Science Foundation.
Abeliovich, A., Chen, C., Goda, Y., Stevens, C., & Tonegawa, S. (1993a). Modified hippocampal long-term potentiation in PKCg mutant mice. Cell 75, 1253-1262.
Abeliovich, A., Paylor, R., Chen, C., Kim, J.J., Wehner, J.M., & Tonegawa, S. (1993b). PKCg mutant mice exhibit mild deficits in spatial and contextual learning. Cell 75, 1263-1271.
Aiba, A., Kano, M., Chen, C., Stanton, M. E., Fox, G., Herrup, K., Zwingman, T. A., & Tonegawa, S. (1994). Deficient cerebellar long-term depression and impaired motor learning in mGluR1 mutant mice. Cell 79, 377-388.
Bliss, T.V.P. & Lomo, T. (1973) Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. Journal of Physiology (London) 232: 331-356.
Chen, Q.X., Stelzer, A., Kay, A.R. & Wong, R.S.K. (1990) GABAA receptor function is regulated by phosphorylation in acutely dissociated guinea-pig hippocampal neurons. Journal of Physiology (London) 420: 207-221.
Cheng, H.C., Kemp, B.E., Pearson, R.B., Smith, A.J., Misconi, L., Van Patten, S.M. & Walsh, D.A. (1986) A potent synthetic peptide inhibitor of the cAMP-dependent protein kinase. Journal of Biological Chemistry 261: 989-992.
Crepel, F. & Krupa, M. (1988) Activation of proteik kinase C induces a long-term depression of glutamate sensitivity of cerebellar Purkinje cells. an in vitro study. Brain Research 458: 397-401.
Crepel, F. & Jaillard, D. (1990) Protein kinases, nitric oxide and long-term depression of synapses in the cerebellum. NeuroReport 1: 133-136.
Crepel, F. & Jaillard, D. (1991) Pairing of pre and postsynaptic activities in cerebellar Purkinje cells induces long-term changes in synaptic efficacy. An in vitro study. Journal of Physiology (London) 432: 123-141.
Daniel, H., Hemart, N., Jaillard, D. & Crepel, F. (1992) Coactivation of metabotropic glutamate receptors and voltage-gated calcium channels induces long-term depression in cerebellar Purkinje cells in vitro. Experimental Brain Research 90: 327-331.
Edwards, F.A., Konnerth, A., Sakmann, B. & Takahashi T. (1989) A thin slice preparation for patch-clamp recordings from neurones of the mammalian central nervous system. Pflgers Archiv 414: 600 - 612.
Ekerot, C.F. & Kano, M. (1985) Long-term depression of parallel fibre synapses following stimulation of climbing fibres. Brain Research 342: 357-360.
Grant, S.G., O'Dell, T.J., Karl, K.A., Stein, P.L., Soriano, P., & Kandel, E.R. (1992). Impaired long-term potentiation, spatial learning, and hippocampal development in fyn mutant mice. Science 258, 1903-1910.
Harrison, N.L. & Lambert, N.A. (1989) Modification of GABAA receptor function by an analog of cyclic AMP. Neuroscience Letters 105: 137-142.
Hirano, T. (1990a) Depression and potentiation of the synaptic transmission between a granule cell and a Purkinje cell in rat cerebellar culture. Neuroscience Letters 119: 141-144.
Hirano, T. (1990b) Effects of postsynaptic depolarization in the induction of synaptic depression between a granule cell and a Purkinje cell in rat cerebellar culture. Neuroscience Letters 119: 145-147.
Inoue, M., Oomura, Y., Yakushiji, T. & Akaike, N. (1986) Intracellular calcium ions decrease the affinity of the GABA receptor. Nature 324: 156-158.
Iriki, A., Pavlides, C., Keller, A. & Asanuma, H. (1989) Long- term potentiation in the motor cortex. Science 245: 1385-1387.
Ito, M. (1984) The Cerebellum and Neural Control. Raven Press, New York.
Ito, M. (1989) Long-term depression. Annual Review of Neuroscience 12: 85 - 102.
Ito, M. & Karachot, L. (1990) Messengers mediating long-term desensitization in cerebellar Purkinje cells. NeuroReport 1: 129-132.
Ito, M. & Karachot, L. (1992) Protein kinases and phosphatase inhibitors mediating long-term desensitization of glutamate receptors in cerebellar Purkinje cells. Neuroscience Research 14: 27-38.
Ito, M., Sakurai, M. & Tongroach, P. (1982) Climbing fibre induced depression of both mossy fibre responsiveness and glutamate sensitivity of cerebellar Purkinje cells. Journal of Physiology (London) 324: 113-134.
Kandel, E. & Schwartz, J. (1982) Molecular biology of learning: modulation of transmitter release.Science 218: 433-443.
Kano, M. (1995) Plasticity of inhibitory synapses in the brain: a possible memory mechanism that has been overlooked. Neuroscience Research 21: 177-182.
Kano, M. & Kato, M. (1987) Quisqualate receptors are specifically inolved in cerebellar synaptic plasticity. Nature 325: 276-279.
Kano, M. & Konnerth, A. (1992a) Potentiation of GABA- mediated currents by cAMP-dependent protein kinase. NeuroReport 3: 563-566.
Kano, M. & Konnerth, A. (1992b) Cerebellar slices for patch clamp recording. In: Practical Electrophysiological Methods, ed. H. Kettenmann & R. Grantyn. Wiley-Liss.
Kano, M., Rexhausen, U., Dreessen, J. & Konnerth, A. (1992) Synaptic excitation produces a long-lasting rebound potentiation of inhibitory synaptic signals in cerebellar Purkinje cells. Nature 356: 601-604.
Knpfel, T., Vranesic, I., Staub, C. & Gahwiler, B.H. (1991) Climbing fibre responses in olivo-cerebellar slice cultures. II. dynamics of cytosolic calcium in Purkinje cells. European Journal of Neuroscience 3: 343-348.
Korn, H., Oda, Y. & Faber, D.S. (1992) Long-term potentiation of inhibitory circuits and synapses in the central nervous system. Proceedings of National Academy of Sciences U.S.A. 89: 440-443.
Komatsu, Y. & Iwakiri, M. (1993) Long-term modification of inhibitory synaptic transmission in developing visual cortex. NeuroReport 4: 907-910.
Komatsu, Y., Toyama, K., Maeda, J. & Sakaguchi, H. (1981) Long-term potentiation investigated in a slice preparation of striate cortex of young kittens. Neuroscience letters 26: 269- 274.
Konnerth, A. (1990) Patch-clamping in slices of mammalian CNS. Trends in Neuroscience 13: 321-323.
Konnerth, A., Dreessen, J. & Augustin, G.J. (1992) Brief dendritic calcium signals initiate long-lasting synaptic depression in cerebellar Purkinje cells. Proceedings of National Academy of Sciences U.S.A. 89: 7051-7055.
Konnerth, A., Llano, I. & Armstrong, C. (1990) Synaptic currents in cerebellar Purkinje cells. Proceedings of National Academy of Sciences U.S.A. 87: 2662 - 2665.
Lev-Ram, V., Miyakawa, H., Lasser-Ross, N. & Ross, W.N. (1992) Calcium transients in cerebellar Purkinje neurons evoked by intracellular stimulation. Journal of Neurophysiology 68: 1167-1177.
Linden, D.J. (1994). Long-term synaptic depression in the mammalian brain. Neuron 12, 457-472.
Linden, D.J. & Connor, J.A. (1991) Participation of postsynaptic PKC in cerebellar long-term depression in culture. Science 254: 1656-1659.
Linden, D.J., Dickinson, H., Smeyne, M. & Connor, J.A. (1991) A long-term depression of AMPA currents in cultured cerebellar Purkinje neurons. Neuron 7: 81-89.
Llano, I., Dreessen, J., Kano, M. & Konnerth, A. (1991a) Intradendritic release of calcium induced by glutamate in cerebellar Purkinje cells. Neuron 7: 577-583.
Llano, I., Leresche, N. & Marty, A. (1991b) Calcium entry increases the sensitivity of cerebellar Purkinje cells to applied GABA and decreases inhibitory postsynaptic currents. Neuron 6: 565 - 574.
Llano, I., Marty, A., Armstrong, C. & Konnerth, A. (1991c) Synaptic- and agonist-induced excitatory currents of Purkinje cells in rat cerebellar slices. Journal of Physiology (London) 424: 183-213.
Mitoma, H., Kobayashi, T., Song S-Y. & Konishi, S. (1994) Enhancement by serotonin of GABA-mediated inhibitory synaptic currents in cerebellar Purkinje cells. Neuroscience Letters 173: 127-130.
Miyakawa, H., Lev-Ram, V., Lasser-Ross, N. & Ross, W.N. (1992) Calcium transients evoked by climbing fiber and parallel fiber synaptic inputs in guinea pig cerebellar Purkinje neurons. Journal of Neurophysiology 68: 1178-1189.
Morishita, W. & Sastry, B.R. (1993) Long-term depression of IPSCs in rat deep cerebellar nuclei. NeuroReport 4: 719-722.
Moss, S.J., Doherty, C.A. & Huganir, R.L. (1992a) Identification of the cAMP-dependent protein kinase and protein kinase C phosphorylation sites within the major intracellular domains of the b1, g2S, and g2L subunits of the g- aminobutyric acid type A receptor. Journal of Biological Chemistry 267: 14470-14476.
Moss, S.J., Smart, T.G., Blackstone, C.D. & Huganir, R.L. (1992b) Functional modulation of GABAA receptors by cAMP- dependent protein phosphorylation. Science 257: 661-665.
Mulle, C., Choquet, D., Korn, H. & Changeux, J.P. (1992) Calcium influx through nicotinic receptor in rat central neurons: its relevance to cellular regulation. Neuron 8: 135- 143.
Nagao, S. & Ito, M. (1991) Subdural application of hemoglobin to the cerebellum blocks vestibuloocular reflex adaptation. NeuroReport 2: 193-196.
Okada, Y. & Miyamoto, T. (1989) Formation of long-term potentiation in superior colliculus slices from the guinea pig. Neuroscience Letters 96: 108-113.
Parfitt, K.D., Hoffer, B.J., Bickford-Wimer, P.C. (1990) Potentiation of gamma-aminobutyric acid-mediated inhibition by isoproterenol in the cerebellar cortex: receptor specificity.Neuropharmacology 29: 909-916.
Porter, N.M., Twyman, R.E., Uhler, M.D. & MacDonald, L. (1990) Cyclic AMP-dependent protein kinase decreases GABAA receptor current in mouse spinal neurons. Neuron 5: 789-796.
Ross, W.N. & Werman, R. (1987) Mapping calcium transients in the dendrites of Purkinje cells from the guinea pig cerebellum in vitro. Journal of Physiology (London) 389: 319-336.
Ross, W.N., Lasser-Ross, N. & Werman, R. (1990) Spatial and temporal analysis of calcium dependent electrical activity in guinea pig Purkinje cell dendrites. Proceedings of Royal Society London Series B Biological Science 240: 173-185.
Sakimura, K., Kutsuwada, T., Ito, I., Manabe, T., Takayama, C., Kushiya, E., Yagi, T., Aizawa, S., Inoue, Y., Sugiyama, H. & Mishina, M. (1994) Reduced hippocampal LTP and spatial learning in mice lacking NMDA receptor e1 subunit. Nature 373: 151-155.
Sakurai, M. (1987) Synaptic modification of parallel fibre- Purkinje cell transmission in in vitro guinea pig cerebellar slices. Journal of Physiology (London) 394: 463-480.
Sakurai, M. (1990) Calcium is an intracellular mediator of the climbing fibre induction of cerebellar long-term depression. Proceedings of National Academy of Sciences U.S.A. 87: 3383 - 3385.
Sessler, F.M., Mouradian, R.D., Cheng, J.T., Yeh, H.H., Liu, W. & Waterhouse, B.D. (1989) Noradrenergic potentiation of cerebellar Purkinje cell responses to GABA: evidence for mediation through the b-adrenoceptor-coupled cyclic AMP system. Brain Research 499: 27-38.
Shibuki, K. & Okada, D. (1991) Endogenous nitric oxide release required for long-term synaptic depression in the cerebellum. Nature 349: 326-328.
Sigel, E., Baur, R. & Malherbe, P. (1991) Activation of protein kinase C results in down-modulation of different recombinant GABAA-channels. FEBS Letters 291: 150-152.
Silva, A.J., Stevens, C.F., Tonegawa, S. & Wang, Y. (1992) Deficient hippocampal long-term potentiation in a-calcium- calmodulin kinase II mutant mice. Science 257: 201-206.
Silva, A.J., Stevens, C.F., Tonegawa, S., & Wang, Y. (1992a). Deficient hippocampal long-term potentiation in a-calcium- calmodulin kinase II mutant mice. Science 257, 201-206.
Silva, A.J., Paylor, R., Wehner, J.M., & Tonegawa, S. (1992b). Impaired spatial learning in a-calcium-calmodulin kinase II mutant mice. Science 257, 206-211.
Stelzer, A., Slater, N.T. & ten Bruggencate, G. (1987) Activation of NMDA receptors blocks GABAergic inhibition in an in vitro model of epilepsy. Nature 326: 698-701.
Sugimori, M. & Llinas, R.R. (1990) Real-time imaging of calcium influx in mammalian cerebellar Purkinje cells in vitro.Proceedings of National Academy of Sciences U.S.A. 87: 5084-5088.
Tank, D.W., Sugimori, M., Connor, J.A. & Llinas, R.R. (1988) Spatially resolved calcium dynamics of mammalian Purkinje cells in cerebellar slice. Science 242: 773-777.
Tsukahara, N., Hultborn, H., Murakami, F. & Fujito, Y. (1975) Electrophysiological study of formation of new synapses and collateral sprouting in red nucleus neurons after partial denervation. Journal of Neurophysiology 38: 1359-1372.
Vincent P., Armstrong, C.M. & Marty, A. (1992) Inhibitory synaptic currents in rat cerebellar Purkinje cells: modulation by postsynaptic depolarization. Journal of Physiology (London) 456: 453-471.
Vincent P. & Marty, A. (1993) Neighboring cerebellar Purkinje cells communicate via retrograde inhibition of common presynaptic interneurons. Neuron 11: 885-893.
Whiting, P., McKernan, R.M. & Iversen L.L. (1990) Another mechanism for creating diversity in g-aminobutyrate type A receptors: RNA splicing directs expression of two forms of g2 subunit, one of which contains a protein kinase C phosphorylation site.Proceedings of National Academy of Sciences U.S.A. 87: 9966-9970.
Yanagihara, D., Kondo, I. & Yoshida, T. (1994) Nitric oxide- mediated cerebellar synaptic plasticity plays an important role in adaptive interlimb coordination during perturbed locomotion. Japanese Journal of Physiology 44, suppl. 1: S222.
[Note: Figures avaiable only in hard copy version]
LEGENDS
FIG. 1 Rebound potentiation (RP) of inhibitory postsynaptic currents (IPSCs) in cerebellar Purkinje cells (PCs). A, Climbing fiber (CF) stimulation (5 pulses, 0.5 Hz) induces prolonged potentiation of IPSCs. B, CF stimulation (5 pulses, 0.5 Hz) fails to induce potentiation when the intracellular solution contained 30 mM BAPTA. C, Depolarization of PCs (single pulse from holding potential -70 mV to 0 mV, 500 ms) induced potentiation of IPSCs. D, Activation of antidromic spikes (5 spikes, 0.5 Hz) does not potentiate IPSC. Each point represents the mean + S.E.M. of amplitudes of 100 to 200 consecutive IPSCs normalized to the values before conditioning. Since there is a tendency for the amplitudes of IPSCs to decay slowly during the control period, the last three values before conditioning were chosen for calculating the baseline (broken line at the 100% level). The threshold for detecting the IPSCs was set to -50 pA. Current traces are taken before (1) and 14 min after (2) conditioning. The traces in the insets were recorded at the time points indicated with thin arrows. The upward thick arrow indicates the time of conditioning (from Kano et al. 1992).
FIG. 2 The transient [Ca2+]i rises accompanying the RP of IPSCs and the potentiation of current responses to bath-applied GABA. A, RP of IPSCs following CF stimulation (5 pulses, 0.5 Hz). B, potentiation of PC current responses to bath-applied GABA (2 mM, 10 s). Insets display current traces recorded before (1) and 15 min after (2) CF stimulation. GABA was applied during the period indicated by the horizontal bars on top of each trace. C, [Ca2+]i in the soma and the proximal dendrites. The upward thick arrow indicates the time of conditioning CF stimulation (from Kano et al. 1992).
Table 1 Summary for potentiation of IPSCs and GABA- induced currents of cerebellar PCs. Comparison of the data from three different papers.
FIG. 3 Conjunctive stimulation of inhibitory interneurons with depolarization of PCs is required for RP. A, Six consecutive traces of stimulus evoked IPSCs (0.2 Hz) are superimposed at time points before (control) and 17min after repetitive stimulation of inhibitory interneurons (at 4 Hz for 1 min) in conjunction with depolarization (50ms duration to -10mV from the holding potential of -60mV). Patch-pipette contained: KCl (70mM), K D-gluconate. (70mM), MgCl2 (2mM), Na-ATP (4mM), Na-GTP (0.4mM), EGTA (0.2mM) and HEPES (30mM) (pH 7.3, adjusted with KOH). The external Ringer solution contained CNQX (10mM) and APV (50mM) to block excitatory transmission mediated by both NMDA and non-NMDA receptors. B, Average changes of IPSC amplitudes 17 min after the three types of conditioning: conjunctive stimulation of inhibitory interneurons with depolarization of PCs (conjunction), depolarization of PCs alone (depol. alone), stimulation of inhibitory interneurons alone (IPSC alone). Thirty-six consecutive trials of IPSCs were averaged in each cell to measure the amplitudes. Values represent percentage mean (+S.E.M.) change of average amplitudes of IPSCs of those taken before the conditioning. Number of recorded cells is shown in parenthesis under each column.
FIG. 4 A, upper panel: Potentiation of GABA-mediated currents of PCs by 8-bromo-cyclic AMP (8b-cAMP). Records were taken in the presence of tetrodotoxin (TTX, 0.5 mM). Incubation of the slices with 8b-cAMP (500 mM, 15 min) produces a marked potentiation of the amplitudes of whole-cell current responses induced by bath-application of GABA (5 mM, 10 s). Note that the amplitudes of spontaneously occurring IPSCs are also potentiated after 8b-cAMP application. Lower panel: Same experimental protocol, but the protein kinase inhibitor peptide (PKIP 400 mg/ml) was added to the internal solution. Note that the GABA-induced currents are depressed rather than potentiated under these conditions. Traces displayed in the right column were taken at about 20 min after the end of 8 b- cAMP application. B, Average changes of GABA-induced whole-cell currents following different experimental manipulations: 8b-cAMP, bath-application of 8b-cAMP (500 mM, 15 min) (n=5); PKIP, bath-application of 8b-cAMP (500 mM, 15 min) and the addition of 400 mg/ml PKIP to the internal solution (n=4); cAMP, bath-application of membrane impermeable cAMP (500 mM, 15 min) (n=6); control, no experimental manipulations, records taken at about 30 min after the beginning of the recording (n=5). Bars represent average values (+ S.E.M.) of current changes normalized to control values taken in the same cells at about 10 min after the beginning of the whole-cell recording (modified from Kano & Konnerth, 1992a).
FIG. 5 A, Record of spontaneous IPSCs (sIPSC, upper panel), amplitude distribution of the sIPSCs (hatched columns in the lower panel) and the background noise histogram (open columns in the lower panel). Records were taken in the presence of 0.5 mM TTX. The background noise histogram was constructed from the whole data points of the current traces where no IPSCs were present (total 1 s, sampling frequency of 2 kHz). The threshold for detecting the sIPSCs (upward arrow in the lower panel) was set to -10 pA so that it was well over the background noise level. The amplitude distribution histogram of the IPSCs was constructed from the peak amplitudes of 400 consecutive sIPSCs. For demonstration purpose, a part of the whole sIPSC distribution with the current range from 0 to -90 pA is shown. The ordinate on the left is applied to the amplitude distribution of sIPSCs and that on the right is to the background noise. Bin width, 2 pA. B, Average changes of sIPSC amplitudes following different experimental manipulations similar to Fig. 5B. Data from the same groups of cells shown in Fig. 5B, but for sIPSCs. Bars represent average values (+ S.E.M.) of current changes normalized to control values taken in the same cells at about 10 min after the beginning of the whole-cell recording (modified from Kano & Konnerth, 1992a).
FIG. 6 Scatter diagram showing the correlation between the percentage changes for the amplitudes of GABA-mediated currents (ordinate) and those of sIPSCs (abscissa) following the experimental manipulations indicated in Fig. 5B and 6B.