Some Properties of Purkinje Neurons In Vivo
Purkinje neurons, which function as a common inhibitory output stage for cerebellar cortical signals, receive two major excitatory inputs which are organized in very different ways. Single CFs originating in the inferior olive, make powerful one-to- one synaptic contacts with Purkinje neurons on the proximal portion of the dendritic arbor. In contrast, PFs, which are the axons of cerebellar granule neurons, each make contact with many Purkinje neurons. Due to the vast numbers of granule neurons, and their divergent input to Purkinje neurons (each Purkinje neuron receives ~150,000 PF contacts), this synapse is among the most abundant in the vertebrate central nervous system. The transmitter of the PFs is thought to be glutamate (Sandoval and Cotman, 1978; Levi et al., 1985), and several types of excitatory amino acid receptor may be found at this synapse including the AMPA receptor and the metabotropic glutamate receptor. The transmitter of the CFs is still undetermined (Zhang et al., 1990; see Cunod et al., 1989 for review), but has been suggested to be an excitatory amino acid (aspartate, homocysteate and glutamate have all been proposed). The EPSCs evoked in Purkinje neurons by either CF or PF stimulation may be completely blocked by CNQX, suggesting that the AMPA receptor is the primary mediator of ion flux at both synapses (Konnerth et al, 1990; Perkel et al., 1990). NMDA receptors are developmentally down-regulated in Purkinje neurons such that very few remain in the adult (Krupa and Crepel, 1990; Rosenmund et al, 1992). In addition, the Purkinje neurons receive considerable GABAergic inhibitory drive from local interneurons.
The Purkinje neuron is highly enriched in the components of several second messenger systems. Cerebellar Purkinje neurons contain metabotropic glutamate receptors, particularly mGLUR1, a subtype linked to activation of the enzyme phospholipase C, in unusually high quantities (Masu et al., 1991), particularly in the distal dendritic spines where the parallel fiber synapses are received (Martin et al., 1992; Baude et al, 1993). The metabotropic receptor mGLUR1, the activation of which produces inositol-1,4,5-trisphosphate (IP3) and 1,2-diacylglycerol (DAG), impinges upon two of these. IP3 binds IP3 receptors and results in a mobilization of Ca from nonmitocondrial intracellular stores. IP3 receptors are highly concentrated in the Purkinje neuron endoplasmic reticulum, including that present in distal dendrites and dendritic spines (Satoh et al., 1990). Also, DAG, together with Ca, activates protein kinase C (PKC), including the g-PKC isoform present in Purkinje neurons (Hidaka et al, 1988).
Another second messenger system of interest is the nitric oxide/cGMP cascade (see Vincent, this volume, for review). Nitric oxide (NO), a gas which diffuses freely through cellular membranes, is formed by the action of NO synthase which is activated by Ca/calmodulin and acts upon arginine to form citrulline and NO. NO in turn, exerts its effects by activating soluble guanylate cyclase resulting in the production of cGMP, which can then activate cGMP-dependent protein kinase. NO synthase immunoreactivity is found in granule cells and their PFs and in basket cells, but is absent in Purkinje neurons and CFs (Bredt et al., 1990; Vincent and Kimura, 1992). Although Purkinje neurons contain guanylate cyclase (Zwiller et al., 1981; Ariano et al., 1982) cGMP-dependent protein kinase (Lohmann et al., 1981; De Camilli et al., 1984), and have detectable resting levels of cGMP (Sakaue et al., 1988), they do not significantly accumulate additional cGMP in response to application of a NO donor (de Vente et al., 1990; de Vente and Steinbusch, 1992; Southam et al., 1992). More recently, another gaseous messenger molecule that activates guanylate cyclase has been identified. Carbon monoxide (CO) is produced in brain by the conversion of heme into biliverdin and CO by the enzyme heme oxygenase-2, which is present in both Purkinje neurons and granule neurons of the cerebellar cortex (Verma et al., 1993).
Some Properties of Purkinje Neurons in Embryonic Culture
Dispersed cultures of embryonic mouse cerebellum were grown in serum-free medium by the method of Schilling et al. (1991). Neurons used for LTD experiments were typically grown for 12-21 days in culture. At this time, Purkinje neurons in culture attained an elaborate dendritic morphology. While the dendritic arbor sometimes appeared similar to the sea-fan shape seen in vivo (Figure 1A), other configurations, such as the bipolar morphology illustrated schematically in Figure 3A, were also found. On a finer scale, it may be seen that, similar to the case in vivo, most synaptic contacts from granule cells are received on dendritic spines (Dunn and Mugnaini, 1993). Purkinje neurons in culture receive spontaneous excitatory and inhibitory synaptic input as assessed by spontaneous post synaptic currents when recorded in voltage-clamp mode (in the absence of tetrodotoxin), and spontaneous firing of simple spikes in current-clamp mode (Figure 1B). They respond to stimulation of neighboring granule neurons and to exogenous excitatory amino acids with excitation. Likewise, they respond to stimulation of neighboring Purkinje neurons and exogenous GABAA agonists with inhibition. They express T, L and P-type voltage-gated Ca currents as assessed by whole-cell voltage clamp recording (Linden, unpublished observations). One notable difference between cultured embryonic and adult Purkinje neurons is that the former express large NMDA-induced inward currents (Linden and Connor, 1991) while the latter do not (Krupa and Crepel, 1990; Rosenmund et al., 1992). Interestingly, over many weeks in culture, embryonic Purkinje neurons show a down-regulation of NMDA current (Linden, unpublished observations). In culture, Purkinje neurons display a similar pattern of reactivity for immunochemical markers to that seen in vivo; they are positive for calbindin-D28K, PEP-19, g-PKC, and cGMP-dependent protein kinase. They are negative for a-PKC, and b-PKC, as well as nitric oxide synthase activity as indicated by the diaphorase reaction (Schilling et al., 1994).
Further information is available about the baseline properties of Purkinje neurons in a rat cerebellar culture preparation that has also been successfully used for gluta- mate/depolarization induced LTD. In a study that combined immunohistochemistry with electrophysiology of synaptic currents, it was determined that Purkinje neurons in culture receive inhibitory contacts on the soma and proximal dendrites while excitatory contacts from granule cells were made only on the dendrites (Hirano and Kasono, 1993). This distribution is re- markably similar to that seen in the intact cerebellum.
In sum, Purkinje neurons in embryonic culture show many of the same basic properties as do those in the intact cerebellum (Table 1). However, there has yet to be a thorough parametric comparison. For example, even though cultured and intact Purkinje cells both express AMPA receptors, it is not clear that they express the same subtypes in the same combination(s). Furthermore, it is not clear if there are pharmacological differences such as relative potencies of agonists and antagonists, or micro-anatomical differences such as distribution at synaptic versus extrasynaptic sites.
Finally, it should be noted that it is sometimes difficult to assess culture versus intact differences when considerable disagreement exists between different laboratories using similar preparations. As an illustrative case relevant to LTD, let us consider the electrophysiological effects of metabotropic glutamate receptor agonists on Purkinje neurons. Using the mouse embryonic culture protocol, we found that activation of Purkinje neuron metabotropic glutamate receptors by pressure- pulse application of the metabotropic agonist t-ACPD resulted in a) intradendritic calcium mobilization that could be blocked by L- AP3 or pertussis toxin and b) an inward current associated with an increase in membrane conductance, that could be blocked by L-AP3 and was dependent upon external Na and internal Ca (Linden et al., 1994). The former result generally confirms previous findings that application of metabotropic receptor agonists produces an increase in internal Ca in Purkinje neurons. However, agonist effectiveness has varied considerably in different types of preparations. For example, in Purkinje neurons of rat cerebellar acute slices, both t-ACPD and 1S,3R-ACPD are relatively ineffective in stimulating Ca mobilization compared to quisqualate which triggers large dendritic Ca increases (Llano et al., 1991; Takagi et al., 1992). In the second study AP3 was shown to block the Ca increases. In a third study, done with normal external Ca, 1S,3R-ACPD caused increases in somatic Ca but these increases appeared to reflect influx via voltage-gated channels as they were temporally associated with Ca spiking and were abolished when the membrane voltage was "manually" clamped at rest (Glaum et al, 1992). Dendritic levels were not measured. In contrast, studies on Purkinje neurons in organotypic culture have shown a large somatic Ca2+ mobilization in response to t-ACPD, with little change in the dendrites (Vranesic et al., 1991; Staub et al., 1992). Similarly, Purkinje neuron's from embryonic mouse, studied in dispersed culture, showed somatic Ca mobilization (Yuzaki and Mikoshiba., 1992). Dendritic changes in this latter study were difficult to assess because interference from the fluorescence of close neighboring neurons and glia. In our studies Ca increases often spread to the soma in response to large ZZstimuli, but the initial responses were clearly in the dendrites.
The tissue culture preparation used by Yuzaki and Mikoshiba (1992) is the most closely related to ours but the characteristics of the Ca response differ from our results in that Ca mobilization was not blocked by either L-AP3 or pertussus toxin and the response, at least to a first approximation, appeared to be localized to the soma. While these differences might represent uncontrolled factors in tissue culture, it is quite possible that they arise from developmental regulation of different receptor subtypes. That is the prominent soma response that is pertussis toxin insensitive might reflect an early developmental stage (as also suggested by these authors) with the prominent dendritic response a later stage. The experiments of Yuzaki and Mikoshiba (1992) were done predominantly on cultures before 12 DIV, and these investigators showed a sharp falloff in the number of neurons displaying the mGluR response at times > 12 DIV. In our experiments we concentrated on cultures that had been maintained for 18 DIV and over, where the Purkinje neurons display mGluR-dependent LTD. The organotypic cultures, maintained as they are in very high serum, might also be held in an immature stage.
An Incomplete Review of Mechanisms of Cerebellar LTD as Determined Using Intact and Slice Preparations
Some of the basic and less-controversial aspects of the phenomenology and mechanisms of cerebellar LTD are summarized below. Some of the more controversial findings using intact and slice preparations are discussed in later sections in specific reference to other studies conducted using cell culture techniques. LTD produced by co-activation of CFs and PFs may be detected as a depression of the PF-Purkinje neuron synapse, but not as a depression of the CF-Purkinje neuron synapse (Ito et al., 1982). Likewise, LTD is manifest as an attenuated response to test pulses of glutamate, the presumed transmitter of the PF- Purkinje neuron synapse but not aspartate (Ito et al., 1982; Crepel and Krupa, 1988). These results suggest that the synaptic modification that underlies cerebellar LTD is specific to the PF- Purkinje neuron synapse and is, at least in part, postsynaptic. The aspect of CF activation that contributes to LTD induction seems to be a prolonged depolarization of the Purkinje neuron which results in dendritic Ca entry (Konnerth et al., 1992). As such, induction of LTD is blocked when Purkinje neurons are electrically inhibited (Ekerot and Kano, 1985; Crepel and Jaillard, 1991) or loaded with a Ca chelator (Sakurai, 1990). LTD may be induced when depolarization sufficient to produce Ca entry is substituted for CF activation (Crepel and Krupa, 1988; Crepel and Jaillard, 1991; Schreurs and Alkon, 1993). Stimulation of PFs contributes to LTD induction by activating certain non-NMDA excitatory amino acid receptors. Stimulation of CFs together with iontophoretic application of glutamate or quisqualate to Purkinje neuron dendrites produces LTD of the PF-Purkinje neuron synapse (Kano and Kato, 1987). LTD is not produced if aspartate or kainate is substituted for quisqualate or glutamate, nor is it produced by iontophoresis of quisqualate alone, or by stimulation of PFs or CFs alone (Kano and Kato, 1987; Sakurai, 1987). Thus, non-NMDA excitatory amino acid receptor activation appears to be necessary, but not sufficient for LTD induction. NMDA receptors, which are not present on the Purkinje neurons of adult animals do not seem to contribute to LTD induction (Kano and Kato, 1987).
LTD is said to result from co-activation of PFs and CFs, but what are the precise timing constraints on this co-activation? One study using intracellular recording in rabbit cerebellar slice has indicated that LTD is optimally induced when CF stimulation precedes PF stimulation by 125-250 msec (Ekerot and Kano, 1989). Another study using a similar preparation has shown that LTD may be induced by CF-PF stimulation with an interval of 50 msec, but claims that LTD induced by CF-PF pairing will not occur unless disynaptic inhibition is blocked by addition of a GABAA antagonist (Schreurs and Alkon, 1993). Finally, a preliminary report using field potential recording in rabbit cerebellar slice has indicated that the optimal interval is in the opposite direction, with PF stimulation preceding CF stimulation by 250 msec (Chen and Thompson, 1992). The latter interval would require that some persistent signal from PFs, such as a consequence of metabotropic receptor activation, linger for at least 250 msec to interact with the CF signal.
A Cell Culture Preparation for the Study of Cerebellar LTD
The first studies of cerebellar LTD using a cell culture technique were conducted by Hirano (1990a,b). These very difficult experiments used a preparation in which embryonic rat olivary and cerebellar neurons were co-cultured, and Purkinje neurons were found that received input from both olivary and granule neurons. As seen in vivo, co-activation of granule neuron and olivary inputs resulted in a long-term depression of granule neuron to Purkinje neuron synaptic drive, and the latter stimulation could be substituted by direct depolarization of the Purkinje neuron sufficient to activated voltage-gated Ca influx.
Using cultured mouse Purkinje neurons, a preparation has been developed in which iontophoretic glutamate pulses and Purkinje neuron depolarization may be substituted for PF and CF stimulation, respectively (Linden et al, 1991). The depression so induced may be seen as a reduction of the glutamate-induced current as measured with a perforated-patch electrode attached to the Purkinje neuron soma. Recordings were typically made in tetrodotoxin/ picrotoxin saline to prevent spontaneous synaptic transmission and to prevent the Purkinje neurons from firing Na spikes. An iontophoresis electrode was aimed at a large caliber dendrite, at a point within one diameter-length of the soma and glutamate pulses (30-80 msec duration) were applied at a frequency of 0.05 Hz. Following 15 min of stable recording, LTD was induced by pairing 6 successive glutamate test pulses with six, 4 sec clamp depolarizations to -10 mV. The depolarization onset preceded the glutamate pulse by 500 msec. The glutamate- induced current was almost always completely decayed by the end of the depolarization step. This treatment induced a reliable depression of the glutamate-induced inward current to 50-60% of its baseline value, was usually evident in the first test pulses after conjunction, and persisted as long as the recording could be maintained. This preparation has several advantages that further the analysis of neuronal information storage. First, as LTD is induced and monitored without synaptic stimulation, alterations in Purkinje neuron responsiveness may be unambiguously attributed to post-synaptic processes. Second, the use of the perforated-patch recording technique (Horn and Marty, 1988) allows for effective voltage-clamping of the Purkinje neuron while avoiding perfusion of the intracellular second messenger and calcium-regulating systems as would occur with conventional whole-cell recording. Third, this preparation allows for simultaneous voltage-clamp recording and optical imaging using fluorescent dyes. These latter two advantages are, of course, not restricted to a cell culture preparation.
Mechanisms of LTD Induction as Determined Using Cultured Purkinje Neurons
The experiments that have been undertaken with this technique have allowed for the elucidation of some principles of LTD induction, as well as to confirm and extend some of the previous observations detailed above. LTD induced by glutamate/depolarization conjunction was found to produce a depression specific to the AMPA subtype of excitatory amino acid receptor (Linden et al., 1991; Linden and Connor, 1991). Further support for the notion that CF activation (or Purkinje neuron depolarization) exerts its effects via Ca influx was provided by the observation that LTD induced by glutamate/depolarization conjunction could be blocked in Ca-free external saline. By using selective agonists of excitatory amino acid receptors, it was shown that both AMPA and metabotropic receptors must be acti- vated to induce LTD. Likewise, application of selective antagonists of either the AMPA or the metabotropic receptors were sufficient to completely block LTD induction (Linden et al., 1991, 1993). Recently, these findings have been confirmed and extended by the observation that application of an inactivating antibody directed against mGLUR1a blocks induction of LTD by glutamate/depolarization conjunction in cultured Purkinje neurons (Shigemoto et al., 1994). Therefore, three processes appear to be necessary and sufficient for LTD induction in voltage-clamped Purkinje neurons in culture: depolarization sufficient to cause Ca influx via voltage-gated channels, AMPA receptor activation, and metabotropic receptor activation.
As one consequence of metabotropic receptor activation is protein kinase C (PKC) activation via diacylglycerol liberation, PKC inhibitors and activators were applied to determine their effects on LTD induction in culture (Linden and Connor, 1991). Inhibitors which act at both the catalytic site (RO-31-8220) and the regulatory sites (calphostin C or peptide pseudosubstrates) of PKC blocked LTD induced by glutamate/depolarization conjunction when applied during the conjunctive stimulus, but had no effect when applied 10 min after the conjunctive stimulus. These compounds did not exert their effects by attenuating voltage- gated Ca influx or the amplitude of AMPA-induced inward current. Application of phorbol-12,13-diacetate (PDA), a PKC activator, induced a depression of AMPA, but not NMDA test pulses. This selectivity was also seen when LTD was induced by quisqualate/depolarization conjunction. In addition, depression induced by PDA and quisqualate/depolarization conjunction were demonstrated to be non-additive, suggesting that they share common mechanisms. These observations confirm and extend a previous report that LTD in slice may be induced by phorbol esters (Crepel and Krupa, 1988). Interestingly, it has been shown in trigeminal neurons that activation of PKC potentiates NMDA receptor-mediated current by relieving the Mg-dependent blockade (Chen and Huang, 1992). It is not clear why PKC activation does not produce a similar effect in cerebellar Purkinje cells, but it is possible that it may result from a differential distribution of NMDA receptor subtypes.
These experiments indicate that activation of PKC is necessary for LTD induction. It is likely that PKC activation is not required for continued LTD expression. It is not clear if PKC activation is sufficient for LTD induction, or if other processes, possibly mediated by AMPA receptor activation, are required. In the simplest scenario it might be imagined that PKC phosphorylates AMPA receptors or associated proteins at the PF synapse resulting in receptor desensitization. At present, there is no evidence either to support or to eliminate consideration of this idea.
More recently, the role of AMPA receptor activation in LTD induction has been addressed (Linden et al, 1993). In Purkinje neurons voltage-clamped to -80mV in tetrodotoxin (TTX) saline, LTD of AMPA currents may be produced by Purkinje neuron depolarization together with pulses of glutamate or quisqualate but not trans-ACPD or (quisqualate + CNQX), suggesting that AMPA receptor activation is necessary for LTD induction. The AMPA receptor in these cultured Purkinje neurons does not appear to exert its effects by directly gating Ca influx as its associated channel is only weakly permeable to Ca as determined by reversal potential measurements (PCa/PNa = 0.17 for AMPA as compared to PCa/PNa = 5.5 for NMDA in these same cells). Replacement of external Na during quisqualate/depolarization conjunction with either the impermeant ions NMG or TEA or with the permeant ions Li or Cs, caused a blockade of LTD induction suggesting that Na influx through the AMPA associated channel is necessary for this process. Similarly, pairing quisqualate pulses with depolarizing steps near ENa also failed to induce LTD. To determine whether activation of voltage-gated Na channels could substitute for AMPA receptor activation, responses to AMPA test pulses were measured in current clamp mode following ACPD/depolarization conjunction in TTX-free saline. LTD was induced 3/16 times in normal (Na and Ca containing) medium and 7/16 times in medium containing the Na channel opener veratridine. This compares with an LTD induction frequency of 15/16 produced by quisqualate/depolarization conjunction. This finding is somewhat consistent with a recent report using the slice preparation which demonstrated LTD induction in 5/8 cells following direct depolarization (in TTX-free saline in current- clamp mode) together with bath application of ACPD (Daniel et al., 1992). Therefore, I suggest that while Na influx via voltage-gated channels may suffice to induce LTD infrequently, activation of AMPA receptors is more effective.
Na influx through either channel might exert its effects by activating Nai/Cao exchange, a process not stimulated by Lii. (Baker and Dipolo, 1984). Suggestively, antiserum to a bovine cardiac Na/Ca exchanger shows strong immunoreactivity in our cultured Purkinje neurons. Alternatively, Na might function to stimulate phospholipase C activity, as has been demonstrated in synaptosomal (Gusovsky et al., 1986; Guiramand et al., 1991) and slice (Benuck et al., 1989) preparations in vitro .
Some studies using the cerebellar slice technique have suggested that release of the gaseous second messenger, nitric oxide (NO), in the cerebellar molecular layer and the consequent activation of soluble guanylate cyclase in the Purkinje neuron, is necessary for LTD produced by PF/CF conjunctive stimulation (Shibuki and Okada, 1991) or PF/depolarization conjunctive stimulation (Crepel and Jaillard, 1990; Daniel et al., 1993). These studies demonstrated that an LTD-like phenomenon could be induced when climbing fiber stimulation or direct depolarization of the Purkinje neuron was replaced by the application of NO (via donor molecules such as sodium nitroprusside) or membrane- permeable analogs of cGMP. Likewise, induction of LTD by more conventional means could be blocked by inhibitors of NO synthase (such as NG-nitro-L-arginine) or agents that bind NO in the extracellular fluid (such as hemoglobin). Recently, it has been demonstrated that NO synthase inhibitors and NO scavengers fail to block LTD induction in slices treated with a gliotoxin, leading to the proposal that Bergmann glia exert a tonic negative regulation of LTD induction which is relaxed by NO (Shibuki, 1993).
In contrast, it has been reported that LTD of glutamate currents produced without synaptic stimulation in cultured Purkinje neurons is unaffected by reagents that stimulate (sodium nitroprusside) or inhibit (hemoglobin, NG-nitro-L-arginine) NO signaling (Linden and Connor, 1992). What might underlie the difference between the lack of effect of NO-modulating reagents that we report and the actions found in previous studies? One possibility is that the difference lies in the type of stimulus used to monitor LTD. Both investigations from other laboratories used PF stimulation to monitor LTD, while our study used glutamate pulses. Therefore if the target of NO action were a site other than the Purkinje neuron dendrite (such as the PF terminal) then the effect of NO-modulating reagents would be detected with PF stimulation but not with glutamate pulses. However, two observations suggest that this explanation might not be correct. First, preliminary data indicate that LTD induced in culture by stimulating a granule cell input to a Purkinje neuron together with direct depolarization is not blocked by hemoglobin or NG-nitro-L- arginine (Linden, unpublished observations). Second, a recent report using the cerebellar slice preparation showed no effect of an NO donor on PF-evoked EPSPs recorded in Purkinje neurons (Glaum et al., 1992). It should be noted that similar confusion about the role of NO may be found in recent studies of hippocampal LTP, all of which have used the slice preparation (Lum-Ragan and Gribkoff, 1993; Williams et al., 1993). This suggests that the divergent results using NO-modulating reagents in cerebellar LTD is probably not directly attributable to differences in slice vs. culture preparations.
Typically, one of the most frustrating aspects of experimentation with primary cultured neurons is cell-to-cell variation. As such, the embryonic mouse cerebellar culture protocol has been refined to attempt to minimize variation in the electrophysiological properties of Purkinje neurons. While this strategy has been largely successful in mouse, considerable variation remains in certain electrophysiological properties of Purkinje neurons recorded using a similar preparation of rat tissue (Linden et al., 1992). It should be noted that this does not represent any fundamental distinction between mouse and rat tissue, but rather that our preparation for rat is not optimal; other investigators have produced a better preparation of rat tissue (Shigemoto et al., 1994). While this degree of variation precludes many types of experiments on cerebellar LTD, it does allow for a certain type of correlational study that exploits the variation inherent in some cell culture preparations. A form of variation present in our rat Purkinje neurons was the amplitude of voltage- gated Ca current. In addition, much to our annoyance, LTD could only be induced in about half of the rat Purkinje neurons tested. As previous work had indicated that activation of voltage-gated Ca channels was necessary for induction of LTD, we sought to determine if the amplitude of the voltage-gated Ca conductance and the probability of successful LTD induction were positively correlated.
Voltage-gated Ca current was recorded by applying depolarizing voltage steps from a holding potential of -90 mV. Recordings were made using the whole-cell patch-clamp technique with N-methylglucamine and tetraethylammonium- containing internal saline at t = -10 min. When Ca currents were present, they typically consisted of both a transient, low-threshold component and a larger, sustained high-threshold component. Measures of peak Ca current largely reflected the amplitude of the latter component. Following establishment of a baseline response to glutamate test pulses, LTD induction by glutamate/depolarization conjunctive stimulation was attempted at t = 0 min. Measures of % of baseline current amplitude were made at t = 20 min. Analysis of these experiments, illustrated in Figure 2, indicates that LTD induction is more likely to occur in cells with larger voltage-gated Ca currents. However, the amplitude of LTD once induced is fairly constant, and does not vary with Ca current amplitude. Ca current amplitude was not correlated with input resistance, degree of dendrtic elaboration or somatic diameter. Of course, a correlation such as this does not prove a causal link between Ca conductance and LTD induction. Ca current amplitude could merely co-vary with a third parameter such as expression of an enzyme, which is necessary for LTD induction. However, together with interventive evidence, correlational results exploiting cell culture variability can be an additional tool to evaluate the substrates of LTD.
Cerebellar LTD is attractive as a model system for the study of information storage, not only because of its duration, but also because it demonstrates input specificity; LTD is confined to those PFs that are active at the time of CF stimulation (Ekerot and Kano, 1985; Chen and Thompson, 1992). This property confers enormous computational power upon the Purkinje neuron because it allows the strength of small groups of the ~150,000 PF inputs to be independently attenuated. Recently, it has been demonstrated that input specificity is preserved when LTD is induced by quisqualate/depolarization conjunction in cultured Purkinje neurons (Linden, 1994a). When multiple, discrete quisqualate application sites are used, LTD is confined to those sites that are stimulated together with depolarization (Figure 3). As these experiments are conducted in TTX/picrotoxin saline, it is unlikely that evoked release from presynaptic terminals is involved in determining specificity. While this experiment suggests that presynaptic processes are not required for input-specific induction of LTD, it is not definitive. The Purkinje neurons grown in this culture system receive synaptic contacts and while most synaptic transmission is abolished by the addition of tetrodotoxin/picrotoxin saline, some release persists under these conditions as indicated by the presence of spontaneous EPSCs which are blocked by kynurenate (data not shown). It is possible that the level of spontaneous release could be increased either directly, as a result of a presynaptic action of quisqualate, or indi- rectly, by activation of the Purkinje neuron resulting in K-efflux or some other retrograde signal. Therefore two manipulations were performed to minimize the contribution of presynaptic processes. First, at the outset of the experiment, adenosine (100 mM) was included in the external saline to suppress spontaneous transmitter release (see Dunwiddie, 1990, for review). In the cerebellum, large numbers of adenosine receptors are present on PF terminals (Goodman et al., 1983) and application of adenosine has been shown to potently block evoked PF-mediated, but not CF-mediated, synaptic drive (Kocsis et al, 1984). Addition of adenosine (100 mM) to the external saline reduced the frequency of spontaneous synaptic currents to <10% of baseline values measured using either perforated-patch or conventional whole- cell recording, the latter using a Cs-containing recording electrode saline. Second, Purkinje neurons were physically isolated by scraping away adjacent cellular material in the culture dish, returning the dish to the incubator for 18-24 hours, and then conducting two-site LTD induction as in Figure 1. Following experiments with scraping, living cultures were stained with rhodamine 123, a vital stain for mitochondria which may be used to image presynaptic terminals in neuronal culture (Yoshikami and Okun,1984), to confirm the absence of viable terminals on the isolated Purkinje cell. Neither application of adenosine nor physical isolation, nor the two treatments in combination inter- fered with the induction of input-specific LTD. These findings strongly suggest that postsynaptic processes are sufficient to confer input-specificity upon cerebellar LTD in culture.
This finding suggests that during the induction of LTD, there exists some spatially restricted signal in the postsynaptic cell. The question remains as to which signal or signals are required to in- duce LTD in an input-specific manner. As Ca influx through voltage-gated Ca channels causes an increase in internal Ca concentration that is distributed widely across the dendritic arbor (Konnerth et al., 1992), it is not a good candidate for such a process. The consequences of local glutamate release (or quisqualate application in the present experiments), namely activation of AMPA and metabotropic receptors and the consequent stimulation of PKC seem most likely to fulfill this requirement.
To evaluate this possibility, the PKC-activating phorbol ester phorbol-12,13-dibutyrate (PDBu) was applied both locally using the two-site stimulation protocol. Local application of PDBu (0.2 mM, at 0-5 min) to site 1 by micro-perfusion resulted in a specific depression of site 1 responses that was not altered by subsequent conjunctive stimulation of site 2 . While phorbol esters such as PDBu are potent exogenous activators of PKC, it is not clear that they mimic endogenous routes of PKC activation. Whereas phorbol esters activate PKC by reducing the Ca concentration required for activation to intracellular resting levels, the natural activators of PKC (diacylglycerols, unsaturated fatty acids) are likely to act synergistically with intracellular Ca transients, and with each other, to cause PKC activation (Nishizuka, 1992). In addition, phorbol esters have effects other than PKC activation. For example, they are activators of the enzyme phospholipase A2 (Mallorga et al., 1980), which may or may not be activated as a consequence of metabotropic receptor activation in Purkinje neurons.
To address this issue, 1-oleoyl-2-acetylglycerol (OAG), a somewhat more physiological activator of PKC was employed. Unlike PDBu, local application of OAG (10 mM) to site 1 together with AMPA test pulses did not result in a depression of Purkinje neuron responsivity monitored at either site. However, when OAG was applied to site 1 together with step depolarizations (10 steps to -10 mV, 4 sec long, 20 sec interstep interval) and AMPA test pulses, a depression specific to site 1 was induced. AMPA test pulses were used instead of quisqualate in this case because quisqualate/depolarization conjunction alone is sufficient to induce LTD (Linden et al., 1991).
Taken together, these experiments suggest that input- specificity of LTD is conferred, at least in part, by spatially constrained activation of PKC in the Purkinje neuron dendrite. It is likely that this spatially constrained activation results from the conjunction of a broad Ca signal contributed by direct depolarization of the Purkinje neuron (or by CF activation in the slice or intact cerebellum) and a constrained signal or set of signals contributed by AMPA and metabotropic receptor activation (or PF activation). The role of PFs seems limited to simple release of glutamate upon stimulation. Alteration in glutamate release as a consequence of LTD induction does not appear to be necessary, nor does release of any signal in addition to glutamate.
Conclusion: Some Thoughts on Method
There are obvious limitations to the study of a complex, multisynaptic phenomenon such as cerebellar LTD using a cell culture system. Even if the biochemistry of particular cell types, the pharmacology of particular synapses and the gross morphology of the Purkinje cell itself remain similar to that in vivo, there are still significant problems of interpretation. Certainly, the cell culture system does not maintain the wiring diagram of the cerebellar cortex, and as a consequence, aspects of information processing and/or storage that are dependent on this level of structure cannot be examined using such a reduced preparation. There are several other limitations that should be explicitly considered. 1) The monitoring of Purkinje neuron responsivity through the application of exogenous agonist is potentially problematic in that one will monitor the function of certain number of extrasynaptic receptors which may not be relevant to alterations in synaptic strength. It should be noted, however, that the detection of LTD by test pulses of exogenous excitatory AMPA receptor agonists may be seen following LTD induction using either a glutamate/depolarization conjunction protocol (Linden et al., 1991; Shigemoto et al., 1994) or a PF activation/depolarization protocol (Crepel and Krupa, 1988) suggesting that these test pulses activate a high proportion of synaptic (as opposed to extrasynaptic) receptors. 2) The cell culture system may be less useful in evaluating the effects on LTD of extrinsic modulatory systems that project to the cerebellum, such as the noradrenergic (Bloom et al., 1971; Siggins et al., 1971) and serotonergic (Strahlendorf et al., 1979) systems. For example the b-adrenergic receptor, which is present on Purkinje neurons in vivo is not strongly expressed in cultured embryonic Purkinje neurons as assessed electrophysiologically (Linden, unpublished observations). 3) The cell culture system is probably not ap- propriate for parametric analysis of LTD induction for several reasons. First, the timing constraints present with synaptic stimulation cannot be analyzed with the application of exogenous transmitters because the evoked currents are many times slower in the latter case. Second, as has been discussed previously, these parameters are likely to be dependent upon disynaptic inhibitory function along specific delay lines not recruited in culture. Third, the present protocol seems to induce an unusually consistent amplitude of LTD (about 50% of baseline) suggesting that it is maximally induced with a single treatment, making it difficult as assess additivity of multiple depressive events. 4) There have been differing conclusions reached with culture versus slice/intact protocols in some cases. However, there is no case where all of the studies conducted in the slice system are in agreement and all of those conducted in the culture system are in disagreement. Hence, there is not at present an example where disagreement between labs can be definitively attributed to choice of preparation.
This is not to say that the cell culture preparation is without use. It is extremely well suited to analysis at the level of the single synapse or single postsynaptic neuron. It the case of cerebellar LTD induction, where presynaptic alterations have not been found using any preparation, it has been particularly useful in dissecting out certain aspects of this phenomenon. The demonstration that input specificity of cerebellar LTD may be conferred by local activation of PKC in the postsynaptic compartment is perhaps the best example of this. In addition, in certain cases the inherent variability of cell culture preparations may be exploited for correlational studies (Figure 2). In my view, analysis of cerebellar LTD in cell culture should be regarded as one useful tool in determining the biochemical and biophysical underpinnings of this plastic process.
Ms. Dorit Gurfel provided skillful technical assistance. Figure 1A was provided by Dr. Karl Schilling, and Figure 1B by Dr. Michael Dickinson. This work was supported in part by Public Health Service grant MH51106, a Klingenstein Fellowship, an Alfred P. Sloan Research Fellowship, and a McKnight Scholarship.
Note
This manuscript represents the state of the field in August, 1994, the time of its final revision.
Ariano, M.A., Lewicki, J.A., Brandwein, H.J., and Murad, F. (1982) Immunohistochemical localization of guanylate cyclase within neurons of rat brain. Proceedings of the National Academy of Sciences of the USA 79:1316-1320.
Baude A., Nusser, Z., Roberts, J.D.B., Mulvihill, E., McIlhinney, R.A.J., Somogyi, P. (1993) The metabotropic glutamate receptor (mGlur1a) is concentrated at perisynaptic membrane of neuronal subpopulations as detecxted by immunogold reaction. Neuron 11:771-787.
Benuck, M., Reith, M.E.A., and Lajtha, A. (1989) Phosphoinositide hydrolysis induced by depolarization and sodium channel activation in mouse cerebrocortical slices. Neuropharmacology 28:847-854.
Bloom, F.E., Hoffer, B.J., and Siggins, G.R. (1971) Studies on norepinephrine-containing afferents to Purkinje cells of rat cerebellum. I. Localization of the fibers and their synapses. Brain Research 25:501-521.
Bredt, D.S., Hwang, P.M., and Snyder, S.H. (1990) Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature 347:768-770.
Chen C., and Thompson, R.F. (1992) Associative long-term depression revealed by field potential recording in rat cerebellar slice. Society for Neuroscience 1992, 18:1215.
Chen, L., and Huang L.-Y. M. (1992) Protein kinase C reduces Mg2+ block of NMDA-receptor channels as a mechanism of modulation. Nature 356:521-523.
Crepel, F., Daniel, H., Hemart, N., and Jaillard, D. (1991) Effects of ACPD and AP3 on parallel fibre-mediated EPSPs of Purkinje cells in cerebellar slices in vitro. Experimental Brain Research 86:402- 406.
Crepel, F., Daniel, H., Hemart, N., and Jaillard, D. (1993) Mechanisms of synaptic plasticity in the cerebellum. In: Long- term potentiation: a debate of current issues, vol. 2., ed. M. Baudry and J. Davis, MIT Press.
Crepel, F., and Jaillard, D. (1990) Protein kinases, nitric oxide and long-term depression of synapses in the cerebellum. NeuroReport 1:133-136.
Crepel, F., and Jaillard, D. (1991) Pairing of pre- and postsynaptic activities in cerebellar Purkinje cells induces long-term changes in synaptic efficacy in vitro. Journal of Physiology (London) 432:123-141.
Crepel, F., and Krupa, M. (1988) Activation of protein kinase C induces a long-term depression of glutamate sensitivity of cerebellar Purkinje cells. An in vitro study. Brain Research 458:397-401.
Cunod, M., Do, K.-Q., Vollenweider, F., Zollinger, M., Klein, A., and Streit, P. (1989) The puzzle of the transmitters in the climbing fibers. Experimental Brain Research Series 17:161-176.
Daniel, H., Hemart, N., Jaillard, D., and Crepel, F. (1992) Coactivation of metabotropic glutamate receptors and of voltage- gated calcium channels induces long-term depression in cerebellar Purkinje cells in vitro. Experimental Brain Research 90:327-331.
Daniel, H., Hemart, N., Jaillard, D., and Crepel, F. (1993) Long-term depression requires nitric oxide and guanosine 3'-5' cyclic monophosphate production in cerebellar Purkine cells. European Journal of Neuroscience 5:1079-1082.
De Camilli, P., MIller, P.E., Levitt, P., Walter, U., and Greengard, P. (1984) Anatomy of cerebellar Purkinje cells in the rat determined by a specific immunohistochemical marker. Neuroscience 11:761-817.
de Vente, J., Bol, J.G.J.M., Berkelmans, H.S., Schipper, J., and Steinbusch H.M.W. (1990) Immunocytochemistry of cGMP in the cerebellum of the immature, adult, and aged rat: The involvement of nitric oxide. A micropharmacological study. European Journal of Neuroscience 2:845-862.
de Vente, J., Steinbusch, H.H.W. (1992) On the stimulation of soluble and particulate guanylate cyclase in the rat brain and the involvement of nitric oxide as studied by cGMP immuno- cytochemistry. Acta Histochemica 92:13-38.
Dunn, M.E., and Mugnaini, E. (1993) Influence of granule cells on the survival and differentiation of Purkinje cells in dissociated cerebellar cultures. Society for Neuroscience Abstracts19:1723.
Dunwiddie, T.V. (1990) In Adenosine and Adenosine Receptors, M. Williams, ed., (Totowa, NJ:Humana Press), pp. 143-172.
Ekerot, C.-F., and Kano, M. (1985) Long-term depression of parallel fibre synapses following stimulation of climbing fibres. Brain Research 342:357-360.
Ekerot, C.-F., and Kano, M. (1989). Stimulation parameters influencing climbing fibre induced long-term depression of parallel fibre synapses. Neuroscience Research 6:264-268.
Garthwaite, J., and Garthwaite, G. (1987) Cellular origins of cyclic GMP responses to excitatory amino acid receptor agonists in rat cerebellum in vitro. Journal of Neurochemistry 48:29-39.
Glaum, S.R., Slater, N.T., Rossi, D.J., and Miller, R.J. (1992) The role of metabotropic glutamate receptors at the parallel fiber-Purkinje cell synapse. Journal of Neurophysiology 68:1453-1462.
Goodman, R.R., Kuhar, M.J., Hester, L., and Snyder, S.H. (1983) Adenosine receptors: Autoradiographic evidence for their location on axon terminals of excitatory neurons. Science 220:967-969.
Guiramand, J., Vignes, M., Mayat, E., Lebrun, F., Sassetti, I., and Recasens, M. (1991) A specific transduction mechanism for the glutamate action on phosphoinositide metabolism via the quisqualate metabotropic receptor in rat brain synaptoneurosomes: I. External Na+ requirement. Journal of Neurochemistry 57:1488-1500.
Gusovsky, F., Hollingsworth, E.B., and Daly, J.W. (1986) Regulation of phosphatidylinositol turnover in brain synaptoneurosomes: stimulatory effects of agents that enhance influx of sodium ions. Proceedings of the National Academy of Science of the USA 83:3003-3007.
Hiadaka, H., Tanaka, T., Onoda, K., Hagiwara, M., Watanabe, M., Ohta, H., Ito, Y., Tsurudome, M., and Yoshida, T. (1988) Cell type specific expression of protein kinase C isozymes in the rabbit cerebellum. Journal of Biological Chemistry 263:4523-4526.
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.
Hirano, T. (1991) Differential pre- and postsynaptic mechanisms for synaptic potentiation and depression between a granule cell and a Purkinje cell in rat cerebellar culture. Synapse 7:321-323.
Hirano, T., and Kasono, K. (1993) Spatial distribution of excitatory and inhibitory synapses on a Purkinje cell in rat cerebellar culture. Journal of Neurophysiology 70:1316-1325.
Ito, M. (1991) The cellular basis of cerebellar plasticity. Current Opinion in Neurobiology 1:616-620.
Ito, M., and Karachot, L. (1990) Messengers mediating long-term desensitization in cerebellar Purkinje cells. NeuroReport 1:129- 132.
Ito, M., and 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., and Tongroach, P. (1982) Climbing fibre induced depression of both mossy fiber responsiveness and glutamate sensitivity of cerebellar Purkinje cells. Journal of Physiology (London) 324:113-134.
Kano, M., and Kato, M. (1987) Quisqualate receptors are specifically involved in cerebellar synaptic plasticity. Nature 325:276-279.
Kano, M., and Kato, M. (1988) Mode of induction of long-term depression at parallel fibre-Purkinje cell synapses in rabbit cerebellar cortex. Neuroscience Research 5:544-556.
Kocsis, J.D., Eng, D.L., and Bhisitkul, R.B. (1984). Adenosine selectively blocks parallel fiber-mediated synaptic potentials in rat cerebellar coretx. Proceedings of the National Academy of Science of the USA 81:6531- 6534
Konnerth, A., Dreessen, J., and Augustine, G.J. (1992) Brief dendritic calcium signals initiate long-lasting synaptic depression in cerebellar Purkinje cells. Proceedings of the National Academy of Science of the USA 89:7051-7055.
Konnerth, A., Llano, I., and Armstrong, C.M. (1990) Synaptic currents in cerebellar Purkinje cells. Proceedings of the National Academy of Science of the USA 87:2662-2665.
Krupa, M., and Crepel, F. (1990) Transient sensitivity of rat cerebellar Purkinje cells to N-methyl-D-aspartate during development. A voltage-clamp study in in vitro slices. European Journal of Neuroscience 2:312-316.
Levi, G., Gordon, R.D., Gallo, V., Wilkin, G.P., and Balazs, R. (1982) Putative amino acid transmitters in the cerebellum. I. Depolarization induced release. Brain Research 239:425-445.
Linden, D.J. (1994a) Input-specific induction of cerebellar long- term depression does not require presynaptic alteration. Learning and Memory 1:121-128.
Linden, D.J. (1994b) Long-term synaptic depression in the mammalian brain. Neuron 12:457-472.
Linden D.J., and Connor J.A. (1991) Participation of postsynaptic PKC in cerebellar long-term depression in culture. Science 254:1656-1659.
Linden D.J., and Connor J.A. (1992) Long-term depression of glutamate currents in cultured cerebellar Purkinje neurons does not require nitric oxide signalling. European Journal of Neuroscience 4:10-15.
Linden, D.J., and Connor, J.A. (1993) Cellular mechanisms of long- term depression in the cerebellum. Current Opinion in Neurobiology 3:401-406.
Linden, D.J. and Connor, J.A. (1995) Long-term synaptic depression. Annual Review of Neuroscience: in press.
Linden, D.J., Smeyne, M., and Connor, J.A. (1993) Induction of cerebellar long-term depression in culture requires postsynaptic action of sodium ions. Neuron 10:1093-1100.
Linden, D.J., Smeyne, M., and Connor, J.A. (1994) Trans-ACPD, a metabotropic receptor agonist, produces calcium mobilization and an inward current in cultured cerebellar Purkinje neurons. Journal of Neurophysiology 71:1992-1998.
Linden, D.J., Dickinson, M.H., Smeyne M., Connor J.A. (1991) A long-term depression of AMPA currents in cultured cerebellar Purkinje neurons. Neuron 7:81-89.
Linden, D.J., Smeyne, M., Sun, S.C., and Connor, J.A. (1992) An electrophysiological correlate of protein kinase C isozyme distribution in cultured cerebellar neurons. Journal of Neuroscience 12:3601-3608.
Llano, I., Dreessen, J., Kano, M., and Konnerth, A. (1991) Intradendritic release of calcium induced by glutamate in cerebellar Purkinje cells. Neuron 7:577-583.
Lohmann, S.M., Walter, U., Miller, P.E., Greengard, P., De Camilli, P. (1981) Immunohistochemical localization of cyclic GMP-dependent protein kinase in mammalian brain. Proceedings of the National Academy of Science of the USA 78:653-657.
Lum-Ragan, J.T., and Gribkoff, V.K. (1993) The sensitivity of hippocampal ong-term potentiation to nitric oxide synthse inhibitors is dependent upon the pattern of conditioning stimu- lation. Neuroscience 57:973-983.
Martin, L.J., Blackstone, C.D., Huganir, R.L., and Price, D.L. (1992) Cellular localization of a metabotropic glutamate receptor in rat brain. Neuron 9:259-270.
Mallorga, P., Tallman, J.F., Henneberry, R.C., Hirata, F., Strittmatter, W.T., and Axelrod, J. (1980). Mepacrine blocks beta adrenergic agonist-induced desensitization in astrocytoma cells. Proc. Natl. Acad. Sci. USA 77, 1341-1345.
Masu, M., Tanabe, Y., Tsuchida, K., Shigemoto, R., and Nakanishi, S. (1991) Sequence and expression of a metabotropic glutamate receptor. Nature 349:760-765.
Nishizuka, Y. (1992). Intracellular signaling by hydrolysis of phospholipids and activation of protein kinase C. Science 258, 607-614.
Perkel, D.J., Hestrin, S., Sah, P., and Nicoll, R.A. (1990) Excitatory synaptic currents in Purkinje cells. Proceedings of the Royal Society (London) 241:116-121.
Rosenmund C., Legendre P., and Westbrook G.L. (1992) Expression of NMDA channels on cerebellar Purkinje cells acutely dissociated from newborn rats. Journal of Neurophysiology 68:1901-1905.
Sakaue M., Kuno T., and Tanaka C. (1988) Novel type of monoclonal antibodies against cyclic GMP and application to immunocytochemistry of the rat brain. Japanese Journal of Pharmacology 48:47-56.
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 fiber in induction of cerebellar long-term depression. Proceedings of the National Academy of Science of the USA 87:3383-3385.
Sandoval, M.E., and Cotman, C.W. (1978) Evaluation of glutamate as a neurotransmitter of cerebellar parallel fibers. Neuroscience 3:199-206.
Satoh, T., Ross, C.A., Villa, A., Supattapone, S., Pozzan, T., Snyder, S.H., and Meldolesi, J. (1990) The inositol 1,4,5-trisphosphate receptor in cerebellar Purkinje cells: quantitative immunogold labeling reveals concentration in an ER subcompartment. Journal of Cell Biology 111:615-624.
Schilling, K., Schmidt, H.H.H.W., and Baader, S.L. (1994) Nitric oxide synthase expression reveals compartments of cerebellar granule cells and suggests a role for mossy fibers in their development. Neuroscience: in press.
Schilling, K., Dickinson, M., Connor, J.A., and Morgan, J.I. (1991) Electrical activity in cerebellar cultures determines Purkinje cell dendritic growth patterns. Neuron 7:891-902.
Schreurs, B.G., and Alkon, D.L. (1993) Rabbit cerebellar slice analysis of long-term depression and its role in classical conditioning. Brain Research 631:235-240.
Shibuki, K. (1993) Nitric oxide: a multi-functional messenger substance in cerebellar synaptic plasticity. Seminars in the Neurosciences 5:217-223.
Shibuki, K., and Okada, D. (1991) Endogenous nitric oxide release required for long-term synaptic depression in the cerebellum. Nature 349:326-328.
Shigemoto, R., Nakanishi, S., Hirano, T. (1994) Antibodies inactivating mGluR1 metabotropic glutamate receptor block long- term depression in cultured Purkinje cells. Neuron 12:1245- 1255.
Siggins, G.R., Hoffer, B.J., Oliver, A.P., and Bloom, F.E. (1971) Activation of a central noradrenergic pathway to cerebellum. Nature 233:481-483.
Southam, E., Morris, R., and Garthwaite, J. (1992) Sources and targets of nitric oxide in rat cerebellum. Neuroscience Letters 137:241-244.
Staub, C., Vranesic, I., and Knopfel, T. (1992) Responses to metabotropic receptor activation in cerbellar Purkinje cells: induction of an inward current. European Journal of Neuroscience 4:832-839.
Strahlendorf, J.C., Strahlendorf, H.K., and Barnes, C.D. (1979) Modulation of cerebellar neuronal activity by raph stimulation. Brain Research 169:565-569.
Takagi, H., Takimizu, H., de Barry, J., Kudo, Y., and Yoshioka, T. (1992) The expression of presynaptic t-ACPD receptor in rat cerebellum. Biochemical and Biophysical Research Communications 189:1287-1295.
Verma, A., Hirsch, D.J., Glatt, C.E., Ronnett, G.V., and Snyder, S.H. (1993) Carbon monoxide: a putative neural messenger. Science 259:381-384.
Vincent, S.R., and Kimura, H. (1992) Histochemical mapping of nitric oxide synthase in the rat brain. Neuroscience 46:755-784.
Vranesic, I., Batchelor, A., Gahwiler, B.H., Garthwaite, J., Staub, C., and Knopfel, T. (1991) Trans-ACPD-induced Ca2+ signals in cerebellar Purkinje cells. NeuroReport 2:759-762.
Williams, J.H., Li, Y-G., Nayak, A., Errington, M.L., Murphy, K.P.S.J., and Bliss, T.V.P. (1993) The suppression of long-term potentiation in rat hippocampus by inhibitors of nitric oxide synthase is temperature and age dependent. Neuron 11:877-884.
Yoshikami, D., and Okun, L.M. (1984) Staining of living presynaptic nerve terminals with selective fluorescent dyes. Nature 310:53-56.
Yuzaki, M., and Mikoshiba, K. (1992) Pharmacological and immunocytochemical characterization of metabotropic glutamate receptors in cultured Purkinje cells. Journal of Neuroscience 12:4253-4263.
Zhang, N., Walberg, F., Laake, J.H., Meldrum, B.S., and Ottersen, O.P. (1990) Aspartate-like and glutamate-like immunoreactivities in the inferior olive and climbing fibre system: a light microscopic and semiquantitative electron microscopic study in rat and baboon (Papio anubis). Neuroscience 38:61-80.
Zwiller, J., Ghandour, M.S., Revel, M.O., and Basset, P. (1981) Immunohistochemical localization of guanylate cyclase in rat cerebellum. Neuroscience Letters 23:31-36.
[Note: Figures avaiable only in hard copy version]
Figure 1. Embryonic mouse Purkinje neurons grown in dispersed culture. A) Purkinje neurons grown in culture maintain an elaborate dendritic morphology similar to that seen in vivo, and are selectively immunoreactive for the marker calbindin-D28K. This cell was from a culture of embryonic day 16 mouse cerebellum grown in vitro for 21 days. Scale bar = 20 mm. Photograph courtesy of Dr. Karl Schilling. B) Spontaneous activity and responses to iontophoretically applied GABA and glutamate recorded in current clamp mode using a perforated-patch electrode attached to the cell soma.
Figure 2. Amplitude of voltage-gated Ca conductance predicts induction frequency of LTD in cultured embryonic rat Purkinje neurons. A) Cells were divided into three groups based on peak Ca current amplitude (measured at t = -10 min), and the % of cells in which LTD was successfully induced was tallied. Successful LTD induction was defined as a reduction of glutamate-induced current to < 80 % of baseline. B) The degree of attenuation following glutamate/depolarization conjunctive stimulation was measured (at t = 20 min) only for those cells that showed successful LTD induction as defined by the criterion above.
Figure 3. Input specificity of LTD may be seen in the absence of synaptic stimulation. A) A schematic diagram of the recording and stimulation protocol. Quisqualate iontophoretic electrodes (labeled "1" and "2" in diagram) are aimed at two widely separated sites on the Purkinje neuron dendritic arbor. Purkinje neurons with bipolar morphology are specifically chosen to facilitate non-overlapping fields of stimulation. The responses to test pulses of quisqualate delivered alternately to the two stimulation sites are recorded with a perforated patch electrode attached to the soma. This electrode also functions to deliver depolarizing voltage-clamp commands which may be paired with stimulation at either site to induce LTD. B) Pairing of quisqualate pulses delivered to site 1 with depolarization (t = 0 min) resulted in the induction of LTD as monitored with subsequent test pulses delivered to site 1 but not site 2. Site 2/depolarization con- junction (at t = 30 min) also resulted in LTD of that input with no further alteration of site 1 responsivity.
Table 1. A comparison of proteins potentially relevant to cerebellar synaptic plasticity: cultured embryonic Purkinje neurons versus adult Purkinje neurons in vivo. Determinations in this table were made either by immunocytochemistry, electrical recording or both.
Embryonic Cell Culture
Adult In Vivo Receptors
Metabotropic (mGLUR1) + +
AMPA + +
NMDA + 0
GABAA + +
b-adrenergic 0 +
IP3 + +
ryanodine + + Enzymes
a-PKC 0 0
b-PKC 0 0
g-PKC + +
Nitric oxide synthase (Neuronal) 0 0
cGMP-dependent protein kinase + + Ion Channels
ICa (T) + +
ICa (L) + +
ICa (P) + +
IK(Ca) + +
INa (TTX-sensitive) Other Proteins
Na/Ca exchanger + +
calbindin D28K + +