Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2003 Nov 21;554(Pt 3):829–839. doi: 10.1113/jphysiol.2003.056523

Repetitive firing of rat cerebellar parallel fibres after a single stimulation

Philippe Isope 1, Romain Franconville 2, Boris Barbour 1, Philippe Ascher 2
PMCID: PMC1664784  PMID: 14634204

Abstract

The excitatory postsynaptic currents (EPSCs) evoked in Purkinje cells (PCs) by stimulating parallel fibres (PFs) usually show a single peak, but EPSCs with multiple peaks (polyphasic EPSCs) can be observed in slices from animals older than 15 days. The EPSCs remain polyphasic when the postsynaptic current is reduced (either by reducing the intensity of the PF stimulation or by adding AMPA receptor antagonists) and when the PC membrane potential is made positive. Thus the late peaks are not due to postsynaptic active currents generated in the imperfectly clamped PC, and must arise from repetitive action potentials in the PF. Extracellular recordings from granule cell (GC) somata showed that a single PF stimulation can elicit a doublet or a train of action potentials. Both the late action potentials recorded in the GCs and the late peaks of the polyphasic EPSCs recorded in the PCs were reduced or abolished by paired-pulse stimulation of the PF or by bath application of the GABAA agonist muscimol. The late action potentials in the GCs were also suppressed by local application of muscimol around the cell body. We propose that after a single stimulation of a PF, the antidromic invasion of the ascending axon and the granule cell can trigger a doublet or a burst of action potentials which back-propagate into the PF (except for the first, which finds the PF still in its refractory period). The repetitive activation of the PF by a single stimulation could play a role in the induction of long-term depression.

The synapse of parallel fibres (PFs) on Purkinje cells (PCs) is the most studied of cerebellar synapses. The PF–PC excitatory synaptic potential (EPSP) was first characterized in vivo by stimulating a ‘beam’ of PFs (Eccles et al. 1966b) and its pharmacological analysis indicated that it is a glutamatergic EPSP involving only α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors (at least in the rat cerebellum, and after 15 days) (Llano et al. 1991; Rosenmund et al. 1992; Momiyama et al. 1996). The time course of the AMPA conductance is difficult to evaluate from in vivo voltage recordings, partly because the EPSP is always followed by a di-synaptic GABAergic IPSP. The first detailed analysis of the time course of the excitatory conductance change was performed in voltage-clamp and under conditions in which the inhibitory input could be selectively eliminated. The whole-cell configuration of the patch-clamp technique (Hamill et al. 1981) was first used in cerebellar slices in the early 1990s (Konnerth et al. 1990; Perkel et al. 1990; Llano et al. 1991). In slices taken from young rats (9–15 days old), Llano et al. (1991) observed EPSCs with a fast rising phase (10–90% rise time of about 1.8 ms), a single peak and a monophasic exponential decay (time constant of 6–7 ms at −70 mV). The reversal potential was close to 0 mV. In slices from animals older than 15 days the decay of the EPSC was slower, and the reversal potential was often positive.

Several explanations of the time course of the parallel fibre compound EPSC recorded in Purkinje cells have been advanced. It decays more slowly (Llano et al. 1991) than predicted from channel kinetics (Barbour et al. 1994; Häusser & Roth, 1997). The decay is under some conditions limited by dendritic filtering (Llano et al. 1991; Roth & Häusser, 2001), but a long-lasting receptor conductance has also been demonstrated (Barbour et al. 1994). However, none of the mechanisms proposed for generating the compound EPSC are consistent with the observations of Perkel et al. (1990) whose records (their Fig. 2) often displayed ‘bumps’ on the decay phase.

Figure 2. The shape of a two-component PC EPSC is unaltered by changes in the holding potential.

Figure 2

Open in a new tab

A, when the membrane potential was changed from −40 mV to −60 mV and to −80 mV, the response increased but remained biphasic. The three records have been scaled to the same peak size. The inflexion on the rising phase had exactly the same latency at the three potentials. The decay could be approximately fitted with a single exponential with a time constant of 16.2 ms at −40 mV, 17.4 ms at −60 mV, 14.1 ms at −80 mV. The records are averages of 10 successive traces at each potential. Stimulation rate 0.1 Hz; parasagittal slice; 5-week-old rat. B, when the membrane potential was changed from −60 mV (lower records) to +30 mV (upper records) the EPSCs were inverted but remained biphasic. The records of the left column (a) and of the centre column (b) were obtained from the same cell but with different positions of the stimulating electrode and different intensities of stimulation (the stimulation in the centre column was close to threshold). The records of the third column (c) were obtained from a different cell. Each record is an average of 5−20 traces. Stimulation rate 0.1 Hz. a and b, para-sagittal slice; 32 day-old rat; the responses reversed near 0 mV; bicuculline 10 μm.c, parasagittal slice; 20-day-old rat; bicuculline 20 μm. The reversal potential was about +10 mV. Times of stimulation are indicated by the triangles; artefacts have been blanked.

In the course of experiments on slices from animals older than 15 days, we observed EPSCs in which the decay was slow as reported by Llano et al. (1991), but was frequently also ‘bumpy’ as in the records of Perkel et al. (1990). The ‘bump’ was most often on the falling phase, but could also indent the rising phase, or create a plateau between the rising phase and the decay phase. We have investigated these ‘polyphasic’ structures. An obvious possible explanation for the multiple peaks would be poor clamp of the dendrites, allowing the activation of voltage-dependent conductances, such as Ca2+ currents (Llinás & Sugimori, 1980; De Schutter & Bower, 1994; Eilers et al. 1995; Denk et al. 1995). However, the experiments reported below support another interpretation: that the multiple peaks of the PC EPSCs reflect multiple action potentials in the PFs. We were particularly interested by this possibility because of our recent observation that, in some conditions, the induction of long-term depression (LTD) at the PF–PC synapse requires trains of action potentials in the PF (Casado et al. 2002).

Methods

The experiments followed European Community guidelines on the care and use of animals (86/609/CEE, CE official journal L358, 18 December 1986), French legislation (decree no. 97/748, 19 october 1987, J. O. République française, 20 october 1987) and the recommendations of the CNRS.

The animals were anaesthetized either with halothane or with ketamine (75 mg kg−1) and xylazine (0.5 mg kg−1) prior to decapitation. Transverse or para-sagittal cerebellar slices (300 μm) of Wistar rats (aged 15–45 days) were prepared following the method described by Llano et al. (1991). Slices were visualized using either a 60 × or a 40 × water-immersion objective (Axioskop, Carl Zeiss) and infrared optics (illumination filter 750 ± 50 nm; Sony CCD camera).

All experiments were performed at room temperature (18–24°C). The recording chamber was continuously perfused at a rate of 1.5 ml min−1 with a solution containing (mm): NaCl 130, KCl 2.5, CaCl2 2, MgCl2 1, NaH2PO4 1.3, NaHCO3 26, glucose 10, bubbled with 95% O2 and 5% CO2 (pH 7.4). In most experiments 10 μm bicuculline methochloride (and in some cases 1 μm strychnine) were added to the bath solution to block fast inhibitory transmission. The concentration of bicuculline was increased to 20 μm in a few experiments in which the EPSCs were recorded at positive potentials and a residual (di-synaptic) GABAergic current could be detected.

PFs were stimulated by means of a glass pipette (tip diameter 5–10 μm) filled with extracellular Hepes-buffered saline. This stimulation electrode was placed at the surface of the molecular layer at a distance of 100–500 μm from the recorded PC. Stimulation intensity was usually between 3 and 15 V; duration was between 30 and 300 μs. Stimulation frequency was usually 0.1 Hz and consisted of either one or two pulses separated by 20 ms in most cases.

PCs were voltage clamped in the whole-cell configuration. Patch pipettes had resistances of 2.5–4.5 MΩ. The standard internal solution was a ‘K-based’ solution containing (mm): potassium gluconate 140, Hepes 10, EGTA 1, KCl 6, MgCl2 1, Na2ATP 4 (Na)GTP 0.4, pH adjusted to 7.3 with KOH. In some experiments involving depolarization to positive potentials potassium gluconate was replaced by caesium gluconate and the Ca2+ buffering was increased by adding 10 mm EGTA. Series resistance was maintained between 4 and 10 MΩ then compensated with settings of 95–98%. HEKA software was used for data acquisition and analysis of PC EPSCs. Whole-cell recordings were filtered at 2 kHz and digitized at 10 kHz. Analysis was performed in the IgorPro graphing environment (Wavemetrics Inc., Lake Oswego, OR, USA).

GCs were recorded in loose cell-attached mode with a home-made amplifier (Barbour & Isope, 2000) or with the HEKA amplifier. The pipettes used had a resistance of 15–20 MΩ when filled with Hepes-buffered solution (containing (mm): NaCl 130, KCl 2.5, CaCl2 2, MgCl2 1, Hepes 10; pH 7.3). The same pipettes were used for local applications of muscimol. They were filled with a solution of muscimol at 20 μm. Pressure pulses were made at 2 bars for durations of 50–500 ms.

Means are reported with s.d. 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzo(f)quinoxaline-7-sulphonamide (NBQX) and bicuculline methochloride were purchased from Tocris Cookson. All the other chemicals were from Sigma.

Results

Properties of multipeak EPSCs

Multi-peak EPSCs were observed in both transverse and parasagittal slices from rats varying in age from 19 to 45 days. In most slices it was possible to find positions of the stimulating electrode and/or intensities of stimulation which elicited simple EPSCs (with a sharp peak and a smooth monophasic decay) and other positions from which EPSCs deviating from this shape were elicited, usually in the form of ‘double-peak’ EPSCs.

The analysis presented below was based on experiments in which the double-peak structure was visible both in single traces and in average traces. Typical examples are illustrated in Figs 1, 2 and 4. The rise time (10–90%) of the first peak (measured on the averaged record) had a mean value of 2.48 ± 0.83 ms (n = 58); the late part of the decay was monotonic and smooth, with a mean time constant of 19.6 ± 7.6 ms (n = 58). However, the second peak could be smaller or larger than the first. It could indent the rising phase or the decay phase, or contribute to the formation of a plateau between the rising phase and the decay phase.

Figure 1. The shape of a two-component PF-EPSC recorded in a Purkinje cell is not changed after addition of NBQX.

Figure 1

Open in a new tab

A, typical two-component PF-EPSC recorded in a Purkinje cell under control conditions (in the presence of bicuculline). B, after addition of NBQX (0.1 μm) to the superfisate. C, after washing NBQX for 15 min. The EPSC was reduced by NBQX and recovered partially. The peak decreased from −315 pA in A to −97 pA in B and recovered to −206 pA in C, but the shape remained biphasic and the time constant of the decay was not markedly changed (A: 17.5 ms; B: 14 ms; C: 16 ms). The records are averages of 29 (A and B) and 30 (C) successive traces. Rate of stimulation 0.1 Hz; para-sagittal slice; 5-week-old rat. Times of stimulation are indicated by the triangles; artefacts have been blanked.

Figure 4. Purkinje cell responses to paired-pulse PF stimulation.

Figure 4

Open in a new tab

Superimposed responses (EPSCs) to single-pulse and paired-pulse stimulation (the filled triangle indicates the first stimulus; the open triangle indicates the timing of the second). The interval between the paired stimuli was 20 ms. The response to the second stimulus is potentiated (paired-pulse facilitation) and monophasic. Transverse slice; 5-week-old rat; holding potential −60 mV; stimulation rate 0.1 Hz. Stimulus artefacts have been blanked.

In other cases (not illustrated) the two peaks were visible in the individual EPSCs but not in the averaged record. However, inspection of the averaged record could still reveal traces of the duality of the responses, such as the presence of a plateau between the rising phase and the decay phase, or an abnormally slow rise (first peak smaller than the second in individual records), or an abnormally slow decay (second peak smaller than the first in individual records). These cases were interesting inasmuch as they suggested that in some previous experiments, the double-peak structure of the EPSC could have been masked by averaging.

In some cases the individual records showed more than two peaks but the averaged record showed only two peaks in which the second had a rounded shape, because the jitter in the latencies of the late peaks increased with the number of peaks.

The late peak is not caused by dendritic voltage-activated conductances

We tested two predictions of the hypothesis that the late peaks of the synaptic current could reflect the activation of voltage-dependent conductances in the PC dendrites. The first prediction was that a smaller EPSC should be less able than a large one to trigger a local response in PC dendrites. To reduce the EPSC amplitude we either added NBQX at a subsaturating concentration, or reduced the intensity of the PF stimulation.The second prediction was that changing the membrane potential in the dendrites should alter the behaviour of active conductances in the PC dendrite.

Effects of NBQX

At concentrations of 1 μm, NBQX nearly completely eliminated the EPSCs, whether they showed a single peak or multiple peaks. However, in the latter case, during the development of the block, the two components of the response decreased in parallel and were still visible when the response had been reduced by 95%. There was no abrupt disappearance of the late peaks as would have been expected if these peaks corresponded to voltage-activated currents occurring in unclamped dendritic regions.

When the concentration of NBQX was lowered to 0.1 μm, the response stabilized at about 10–30% of the initial amplitude and, again, the shape of the multiple-peak responses was unchanged. One of the three experiments of this type is illustrated in Fig. 1.

Threshold stimulation

When the PF stimulation was at threshold for the appearance of a postsynaptic response, the smallest EPSCs were a few tens of picoamps, i.e. in the upper range of the distribution of miniature EPSCs (Barbour, 1993; Sabatini & Regehr, 1997). These small EPSCs were frequently biphasic, as illustrated in Fig. 2B (centre column). It is unlikely that in such records the second hump could be due to a local Ca2+ action potential (for instance), since the size of the current is small. This observation supports the hypothesis that the two components correspond to two successive EPSCs.

Effects of changing the holding potential

The reponses showing multiple peaks kept their shape when the holding potential was changed. The constancy of the shape is illustrated in Fig. 2A for a case in which the holding potential was changed from −40 to −60 and −80 mV. In Fig. 2B the potential was changed from −60 to +30 mV. The records are reminiscent of those of Perkel et al. (1990) in which the ‘bumps’ of responses recorded in ‘old’ animals (4–5 weeks) were observed at both negative and positive potentials.

The interpretation of the data obtained at negative potentials could be ambiguous. Hyperpolarization simultaneously increases the synaptic current as well as the difference between holding potential and the activation voltages of any putative voltage-dependent conductances. One could argue that the ability of the EPSC to activate such conductances might not be greatly altered by hyperpolarization. However, the data obtained at positive potentials are unambiguous, since known voltage-dependent conductances which could distort the EPSC are fully inactivated in this range. We conclude that the late peaks do not reflect the activation of voltage-dependent conductances.

The multipeak structure of the EPSCs is due to repetitive firing of the parallel fibres

Since the data presented above suggested that the multipeak structure of the EPSCs is due to repetitive firing of the PFs, we decided to record action potentials from GCs following antidromic stimulation. The PFs were stimulated at 0.1 Hz either with one pulse or with two pulses at a 20 ms interval. This double-pulse protocol, used in the experiments illustrated in Figs 3 and 6, was chosen to interpret an observation which will be described below (Figs 4 and 5), the sensitivity of the late peaks of the PC EPSC to repetitive stimulation of the PFs.

Figure 3. Granule cell responses to paired-pulse stimulation of parallel fibres.

Figure 3

Open in a new tab

A, the cell responded with a doublet of action potentials to the first PF stimulation, and with a single action potential to the second PF stimulation, applied 20 ms after the first. The timing of stimulation is indicated by the triangle. The stimulation artefacts have been blanked. B, superimposed records of the responses to 43 stimulations (grey) and their average (black). In 13 cases there was a doublet of action potentials as illustrated in A. In the other 30 records the first PF stimulation only triggered one action potential. Stimulation rate 0.1 Hz; 31-day-old rat. Times of stimulation are indicated by the triangles; artefacts have been blanked.

Figure 6. Effects of muscimol on the granule cell responses to paired-pulse stimulation of parallel fibres.

Figure 6

Open in a new tab

A, control responses to paired stimuli of parallel fibres with a 20 ms interval (0.1 Hz repetition rate, in the presence of 10 μm bicuculline). The cell responded with one, two or three action potentials to the first PF stimulation, and with one or two action potentials to the second PF stimulation. B, in the presence of muscimol (1 μm; bicuculline was washed at the same time) both the first and the second PF stimulations triggered only a single action potential. C, the pattern observed in control conditions reappeared partially after washing the muscimol, but the first stimulation never elicited more than two action potentials, and the second stimulation never more than one action potential. Para-sagittal slice; 6-week-old rat. Times of stimulation are indicated by the triangles; artefacts have been blanked; individual traces are grey, and their averages black.

Figure 5. Purkinje cell responses to PF stimulation at 0.1 Hz and 1 Hz.

Figure 5

Open in a new tab

The response to stimulations at 0.1 Hz and 1 Hz have been superimposed in three examples. In cell A, the second component of the EPSC seems to be eliminated by increasing the stimulation rate to 1 Hz. In cell B, increasing the stimulation rate to 1 Hz reduces the second EPSC but the persistence of a late component is revealed by the rounded peak of the response. In C, the persistence of the second component is directly visible. Para-sagittal slices. A, 3-week-old rat; B and C, 4-week-old rats; holding potential −60 mV; bicuculline (10 μm) was present in A and B but not in C. However, complete elimination of the second component was also seen in other cells in the absence of bicuculline. Times of stimulation are indicated by the triangles; artefacts have been blanked.

Antidromic invasion of granule cells

The responses of the GC soma were recorded in the loose cell-attached mode (Barbour & Isope, 2000) in para-sagittal slices from 3- to 4-week-old animals. After verifying that direct stimulation of the cell body through the recording electrode elicited an action potential, the stimulation electrode was placed in the molecular layer and moved until an antidromic action potential was recorded in the GC. In 49 out of 141 cells tested, single stimuli (at 0.1 Hz) evoked more than one action potential in the GC for at least some of the stimuli. In 39 of the cells one could observe a doublet of the type illustrated in Fig. 3. In the 10 other cells the stimulation of the PF triggered three or more action potentials in some runs.

As can be seen in Fig. 3B for a case where the stimulation induced doublets, the latency of the first action potential was very stable, whereas the latency of the second action potential was variable. This variability was even larger for later action potentials, as shown in Fig. 6, in which the PF stimulation elicited in many cases a triplet of action potentials.

In 17 of the cells in which there was more than one action potential the interval between the first two action potentials (tgc) could be measured reliably, and had a mean value of 5.2 ± 1.5 ms.

Effects of repetitive stimulation

The late action potentials in the GC as well as the late EPSCs of the PC were very sensitive to the frequency of stimulation. Figure 3 shows the responses of a GC to pairs of pulses separated by 20 ms and repeated 30 times at 0.1 Hz. The first pulse of the pair elicited a doublet of action potentials in 13 out of 30 sweeps. No doublet was seen in response to the second pulse of the pair. Figure 6A (control) corresponds to a case in which the first pulse elicited one, two or three action potentials. The second PF stimulation never produced triplets. In most cases it induced a single action potential and in a few cases a doublet.

The elimination of the late action potentials by a double pulse was quantified by calculating the ratio of the mean number of ‘late action potentials’ elicited by the second PF stimulation over the mean number of late action potentials after the first stimulation. This ratio is 0 in the example illustrated in Fig. 3, where there were no late action potentials in any of the responses to the second pulse. The ratio was 0.17 in the records of Fig. 6A (upper traces, control) where there were a few doublets in responses to the second PF stimulation. The mean ratio for the 13 cells analysed was 0.07 ± 0.09.

The lability of the late action potentials illustrated in Figs 3 and 6 is likely to explain the lability of the late peak(s) of the PC EPSC, which is illustrated in Fig. 4. The first PF stimulation elicited a double-peak EPSC. The second PF stimulation, applied 20 ms after the first, induced an EPSC that was markedly potentiated (paired-pulse facilitation: Konnerth et al. 1990; Perkel et al. 1990; Atluri & Regehr, 1996), but in which the second peak was completely eliminated. This was observed in most experiments, although in a few cases the second peak was only partially occluded. This is in agreement with the fact that in double-pulse experiments the repetitive firing of the GC is always reduced but not always abolished.

When the interval between the two PF stimulations was increased, the occlusion of the late action potentials in GC recordings and of EPSCs in PC recordings became less marked but was still detectable at intervals of 1 s. Figure 5 illustrates three examples of the effect of a 1 Hz stimulation on dual EPSCs. In the case of Fig. 5A, when the rate of stimulation was increased from 0.1 Hz to 1 Hz, the early component was slightly reduced but the late component seemed to be completelty eliminated. In the case of Fig. 5B, the early component was unchanged while the late one seemed to disappear, but at 1 Hz the shape of the rounded peak strongly suggested that the response remained polyphasic. In Fig. 5C both components were reduced at 1 Hz, but the second component was still detectable.

The effect of increasing the stimulation frequency from 0.1 to 1 Hz was also analysed on the responses of GCs. The effect was quantified by calculating the ratio of the mean number of ‘late action potentials’ elicited at 0.1 Hz over the mean number of late action potentials elicited at 1 Hz. The ratio was 1.3 ± 0.1 (n = 14). This is consistent with the observations made on the EPSCs.

Effects of bath-applied muscimol

The emission of a burst of action potentials after a brief stimulation is not predicted by classical (Hodgkin–Huxley) models of voltage-dependent conductances and implies that the GC possesses a special set of ‘slow’ depolarizing conductances allowing the development of the late action potentials. Evidence for slow depolarizing conductances in the GCs has been provided by D'Angelo et al. (1998) using direct stimulation of the GC. Hamann et al. (2002) further showed that a burst of action potentials could be elicited in the GCs after an orthodromic stimulation, but only after block of the tonic GABAA-receptor conductance in GCs (Brickley et al. 1996; Wall & Usowicz, 1997). We thus applied muscimol to examine if an increased inhibitory conductance in the GCs reduced their probability of emitting a burst of action potentials after an antidromic stimulation.

Figure 6 illustrates an experiment in which, in the control period, the first PF stimulation induced one, two or three action potentials, and the second PF stimulation only one or two (GABAA receptors were blocked by bicuculline). In the presence of added muscimol (1 μm) (and after washing out bicuculline), both the first and the second PF stimulations triggered only single action potentials. Similar experiments were performed in six cells. The effect was quantified as the ratio of the mean number of action potentials induced by the first PF stimulation measured after 2 min in the presence of muscimol (1 μm) over the same value in control conditions. This ratio was 0.08 in the experiment of Fig. 6 and its mean value was 0.2 ± 0.28 (n = 6).

Figure 7 illustrates the parallel effect of muscimol on EPSCs recorded in a PC. The record obtained in the control solution shows a double EPSC (a). In the presence of muscimol, the EPSC became monophasic (b). The difference between the two records (a–b) shows that the second EPSC (eliminated by muscimol) started 9 ms after the first, had a more rounded peak but a similar speed of decay. The fact that the peak was rounded in this averaged record is likely to reflect the dispersion of the late action potentials of the GCs, visible in Figs 3 and 6.

Figure 7. Effects of muscimol on PF EPSCs recorded in a Purkinje cell.

Figure 7

Open in a new tab

Superimposed traces of EPSCs recorded in the presence of bicuculline (‘ctrl’), in the presence of muscimol (‘muscimol’; bicuclline washed), and the difference between them showing the muscimol-sensitive current. The application of muscimol (2 μm) triggered an outward current which developed progressively. The records were taken when the outward current had just started developing and had a amplitude of 50 pA. At this point the amplitude of the first component was unaffected, whereas the second component was completely eliminated. With time the outward current reached a value of 470 pA and by that time the size of the first component had been reduced from the initial value of 325 pA to 227 pA. The difference trace shows the complete time course of the second component, and illustrates in particular that it decays at the same rate as the first component. Five-week-old rat; transverse slice; holding potential −50 mV; PF stimulation 0.1 Hz. Each trace is the average of 16 successive records. Times of stimulation are indicated by the triangle; artefacts have been blanked.

PCs have GABAA receptors which muscimol did activate. The conductance change induced by muscimol in the PC could be expected to modify the voltage clamp control of the dendritic tree. Indeed in certain experiments (but not in Fig. 7) there was an acceleration of the decay of the monophasic EPSC associated with a reduction of the time to peak. These changes can be accounted for by the effect of muscimol on the PC conductance. They cannot explain the selective elimination of the late component, and we therefore interpret this selective elimination as reflecting an effect of muscimol on the presynaptic neurones. The analogous experiments looking at the effect of muscimol on action potentials in the GCs do not suffer from this problem.

Origin of the action potential burst. The bursts of action potentials in the GC indicate the presence of conductances allowing repetitive activity but do not localize these conductances. We considered two main possibilities (Fig. 8). In the first case, a train of action potentials originates in the PF (at the stimulation site) and propagates along the PF and then down the ascending axon to the soma. In the second case, the PF stimulation triggers a single action potential in the PF which elicits a burst of action potentials only once it reaches the ascending axon (or the soma) of the GC. The first action potential cannot reverberate in the orthodromic direction because the axon is still in its refractory period. But the following action potentials would propagate back along the axon and, when they reach the PF/PC synapse, elicit the late EPSCs.

Figure 8. Two possible origins for the repetitive activity of PFs.

Figure 8

Open in a new tab

A, the late action potentials are emitted at the site of stimulation and travel along the PF and down into the ascending axon towards the soma. B, the stimulation of the PF only elicits a single action potential with travels along the PF, goes down the ascending axon and there activates a slow depolarizing conductance which leads to the firing of one or more additional action potentials. The first (incoming) action potential cannot reverberate because the axon is in its refractory period. But the second action potential can propagate orthodromically to the synapse. For mechanism A the interval between the two action potentials at the synapse on the PC (tpc) is the same as the interval between the two action potentials recorded in the GC (tgc). For mechanism B the interval between the two action potentials at the synapse on the PC (tpc) is the sum of the conduction time from the synapse to the granule cell, the interval between the two action potentials recorded in the GC (tgc) and the conduction from the granule cell to the synapse. The conduction times from the synapse to the granule cell and from the granule cell to the synapse are assumed to be equal (ta). The diagram places the dendritic tree of the Purkinje cell between the stimulating electrode and the GC soma. This is always the case in parasagittal slices and nearly always the case in transverse slices. Thus, the travel time (ta) includes not only the travel time along the ascending axon, but also the travel time from the branch point to the synapse.

The effects of locally applied muscimol

The selective effects of muscimol on the late action potentials are likely to be due to a selective reduction of the slow depolarization responsible for the late action potentials, whether this slow depolarization occurs near the stimulation electrode or near the GC soma. To localize this slow conductance we examined the effects of a local application of muscimol on the action potentials recorded from the GCs. Muscimol was puffed either at the level of the cell body of the GC or around the stimulating electrode in the molecular layer (Fig. 9) and we compared the reduction of the frequency of occurrence of the late action potentials produced by the two applications. The difference was very clear. Although the muscimol application on the GC cell body was not always effective, and although the application of muscimol on the PF sometimes induced a reduction of the late spiking, in a given cell, the effect of muscimol applied on the GC soma was always much larger than that on the PF. This is illustrated in Fig. 9 for an experiment in which the PF was stimulated at 1 Hz. Before the application of muscimol the response was always a doublet or a triplet. Less than 1 s after the application of muscimol on the cell body of the GC, the responses were reduced to a single action potential whether the pulse of muscimol lasted 50 ms (Fig. 9A) or 200 ms (Fig. 9B). The first doublet reappeared after 5–6 s in the first case, after 6–8 s in the second. In contrast, when muscimol was applied on the PF, there was no detectable effect for a 50 ms pulse. For a 200 ms pulse the fourth response after the application was a single action potential but the next responses were doublets and triplets. Thus the inhibition produced by applying muscimol on the PF appeared later and was briefer than that produced by applying muscimol on the GC cell body. We quantified the effects of muscimol by calculating the ratio of the number of late spikes during the 10 sweeps following the muscimol pulse to the number of late pulses during the 10 sweeps preceding the pulse. For a pulse of 50 ms the ratio was 0.45 ± 0.03 (n = 7) at the GC, 0.91 ± 0.02 (n = 7) at the PF. For a pulse of 200 ms, the ratio was 0.16 ± 0.02 (n = 5) at the GC, 0.64 ± 0.02 (n = 5) at the PF.

Figure 9. Local applications of muscimol on the granule cell and on the parallel fibre.

Figure 9

Open in a new tab

The PF was stimulated at 1 Hz and in the absence of muscimol this triggered two or three action potentials recorded in the GC soma. The successive records obtained at intervals of 1 s are arranged vertically. Muscimol was applied just after the fourth stimulation by a pressure pulse lasting 50 ms (A) or 200 ms (B). The application was made either on the GC soma (left columns in A or B) or on the PF at the level of the stimulating electrode (right columns in A or B). Following application to the soma, the late action potentials were eliminated in the next 4 records (shown). Following application to the stimulation site, there was no detectable effect in the case of a short pulse (A). After a longer pulse an effect was seen but it was delayed (third record, i.e. about 3 s after the muscimol pulse) and transient (about 1 s). Para-sagittal slice; 4-week-old rat.

These experiments allow us to conclude that the effects of muscimol are due to an action on receptors situated in the granule cell layer and not under the PF stimulating electrode. The delayed and small effect observed in a few cases by applying muscimol on the PFs is likely to be due to diffusion of muscimol to the granule cell layer.

The late action potentials can be observed in the absence of GABA-A blockers. Repetitive activity of the GCs, which we observed after antidromic activation of the GCs, has also been observed after orthodromic stimulation of mossy fibres (Hamann et al. 2002). However, these authors only observed the bursts of action potentials in the presence of furosemide, which blocks a large fraction of the tonically active GABAA receptors in GCs. To test if, in our experiments, this tonic activation could prevent the emission of bursts, we repeated the experiments in the absence of bicuculline. In six slices that had never been exposed to this compound, we observed multipeak EPSCs (e.g. Fig. 5C). These EPSCs were not changed by the addition of bicuculline. Similarly, multiple action potentials in GCs were observed in the absence of bicuculline in 23 cells in five slices (e.g. Fig. 9). It thus appears that in our slices the tonic activation of the GABAA receptors in GCs was not sufficient to prevent the repetitive activation of granule cells by an antidromic stimulation.

Discussion

Dendritic filtering

Dendritic filtering undoubtedly slows the EPSCs recorded in animals older than 15 days (Llano et al. 1991; Roth & Häusser, 2001), whether the second peak is present or not. The model of Roth & Häusser (2001) predicts that at P21 a dendritic conductance change rising with an exponential time constant of 0.2 ms and decaying with an exponential time constant of 3 ms will appear in the soma as an EPSC with a 20–80% rise time of up to 3.5 ms and a decay time constant of up to 15 ms. The values that we obtained for the single peak EPSCs (in control conditions or in the presence of muscimol) are in good agreement with these predictions. In contrast, dendritic filtering cannot explain the double-peak EPSCs.

Dendritic voltage escape and active currents

If the late peaks of multiple peak records were due to the activation of voltage-dependent conductances (such as Ca2+ channels), the shape of the multiphasic EPSCs should have been altered by changing the holding potential between −90 and +40 mV and by blocking the postsynaptic receptors (with NBQX). The fact that the shape was not changed indicates that the successive peaks of the EPSCs correspond to successive volleys of action potentials in the PFs.

In a few recordings, individual traces showed postsynaptic ‘action potentiallets’ which differed from the usual late peaks by their fast rise, and their all-or-none character. Records of this type have been shown by Hartell (1996) (his Fig. 1) and interpreted as expressing a local development of active currents, as predicted by Eilers et al. (1995). The occasional presence of these action potentiallets in our experiments indicates that in some cases postsynaptic ‘unclamped’ action potentials may contribute to the complexity of the EPSCs and can make their decay longer and noisier.

The origin of the action potential burst

The effects of local applications of muscimol clearly indicate that the GABA receptors responsible for the effects of bath-applied muscimol are not on the parallel fibres. These experiments also strongly suggest the burst originates in the GC somato-dendritic compartment and/or the ascending axon. D'Angelo et al. (1998) have reported the presence in GCs of two types of TTX-sensitive Na+ currents, which in their experiments triggered a slow depolarization at around −55 mV and were responsible for subthreshold depolarizing potentials around resting potential. These Na+ currents are likely candidates for the observed effects. However, it cannot be completely excluded that a burst-generating mechanism in the PF could be blocked by an electrotonic propagation of a somatic voltage change.

In principle, the comparison of the interval between PC EPSCs (tpc, a first approximation of the interval between the two action potentials arriving at the synapse) with that between successive GC action potentials recorded at the soma (tgc) could give an additional clue regarding the site of origin of the burst. If the second action potential originates in the PF (Fig. 8A), the intervals should be identical and the difference tpc–tgc = 0. If the second action potential originates in the ascending axon or the GC soma (Fig. 8B), the mean interval between the peaks of the PC EPSCs should be slightly longer than the interval between the first and the second action potentials recorded in the GCs. In this case, the difference between tpc and tgc will depend on the relative positions of the stimulating electrode, the branch point of the ascending axon and the synaptic contact. However, in all cases it will include twice the conduction time (ta) from the branching point of the PF to the site of origin of the second GC action potential. If we assume an average conduction velocity at room temperature of about 200 μm ms−1 (Eccles et al. 1966a; Vranesic et al. 1994; D'Angelo et al. 1995) and an average ascending axon length for the stimulated GCs of about 200 μm (Harvey & Napper, 1988), the difference between tpc and tgc should be 2 ms if the second action potential originates in the soma of the GCs.

The mean value of tpc was 6.5 ± 1.3 ms (n = 34). The mean value of tgc was 5.2 ± 1.5 ms (n = 17). The two values are significantly different (Student's t test: t = 3.25, P < 0.001). On the other hand, the difference tpc–tgc = 1.3 ms is not significantly different from 2 ms. Thus, in a first approximation, these results support the hypothesis that the burst does not originate in the PF but in the granule cell or in the ascending axon.

Doublets and LTD induction

In a recent study of LTD at the PF–PC synapse, Casado et al. (2002) induced LTD by coupling PF stimulation with PC depolarization at 1 Hz for 2 min. They found that this protocol was only effective when the PF stimulation was a doublet (stimuli separated by 60 ms). This requirement for a doublet of action potentials was consistent with the proposed involvement of the presynaptic NMDA receptors of the PFs: the glutamate released by the first action potential probably activated the presynaptic NMDA receptors only after the action potential had passed, and this meant that their voltage-dependent magnesium block (Nowak et al. 1984) could only be relieved if a second action potential arrived within a given interval. However, this interpretation is difficult to reconcile with the fact that, in previous publications, coupling PF stimulation and PC depolarization had been found to induce LTD even if the PFs were stimulated at low frequency with a single pulse.

A possible way to reconcile the two sets of data is to suppose that in the experiments using single PF pulses during the induction protocol, the single stimulus actually triggered two action potentials. The duality of the input could have been missed in the experiments using current clamp (e.g. Crepel & Jaillard, 1991; Hartell, 1996). The records illustrating experiments using voltage clamp do not show multiple peaks (e.g. Aiba et al. 1994; Khodakhah & Armstrong, 1997) but many show rounded EPSCs which could correspond to a multicomponent structure smoothed by averaging. It is therefore possible that in some of the experiments in which the induction of LTD used single-pulse stimulation of the PFs, there may have been a covert repetitive activation of the PFs, and that a sufficient fraction of the doublets persisted during the 1 Hz stimulation.

Physiological conditions

Our experiments were aimed primarily at understanding the apparent contradictions in the literature regarding the requirements of LTD induction, as discussed above. Since most of the experiments in this field were performed at room temperature, we analysed the GC doublets and the PF dual EPSCs at room temperature. We have not tried to establish whether the repetitive firing of the GCs would also be observed at higher, more physiological temperatures.

The bursting behaviour of granule cells that we have analysed could be observed in the absence of bicuculline. This appears at first sight to contradict the observations of Hamann et al. 2002) who reported bursting after orthodromic stimulation, but only after blockade of the tonic activation of granule cell furosemide-sensitive GABAA receptors. The difference could be due to a difference between orthodromic and antidromic stimulation. It could also be linked to the fact that the tonic activation of GABAA receptors in granule cells increases with age: Hamann et al. (2002) used animals of 35–45 days, whereas our experiments in the absence of bicuculline were performed on animals aged 21–30 days. Finally, it could also be linked with the different temperatures of the experiments (29 ± 3°C versus 21 ± 3°C).

Although doublets are not necessarily observed in all preparations following antidromic stimulation of PFs (cf. Isope & Barbour, 2002; adult rat slices at 32°C), their possible presence should in future be taken into account when interpreting experiments involving such stimulation.

Acknowledgments

This work was supported by the CNRS, the Ecole normale supérieure, the Université René Descartes, and the European Commission (grant no. IST-2001–35271). PI was the recipient of a fellowship from the Fondation pour la Recherche Médicale. We thank Mariano Casado, Joël Chavas, Stéphane Dieudonné and Alain Marty for comments on the manuscript.

References

  1. Aiba A, Kano M, Chen C, Stanton ME, Fox GD, Herrup K, Zwingman TA, Tonegawa S. Deficient cerebellar long-term depression and impaired motor learning in mGluR1 mutant mice. Cell. 1994;79:377–388. [PubMed] [Google Scholar]
  2. Atluri PP, Regehr W. Determinants of the time course of facilitation at the granule cell to Purkinje synapse. J Neurosci. 1996;16:5661–5671. doi: 10.1523/JNEUROSCI.16-18-05661.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barbour B. Synaptic currents evoked in Purkinje cells by stimulating individual granule cells. Neuron. 1993;11:759–769. doi: 10.1016/0896-6273(93)90085-6. [DOI] [PubMed] [Google Scholar]
  4. Barbour B, Isope P. Combining loose cell-attached stimulation and recording. J Neurosci Meth. 2000;103:199–208. doi: 10.1016/s0165-0270(00)00318-6. [DOI] [PubMed] [Google Scholar]
  5. Barbour B, Keller BU, Llano I, Marty A. Prolonged presence of glutamate during excitatory synaptic transmission to cerebellar Purkinje cells. Neuron. 1994;12:1331–1343. doi: 10.1016/0896-6273(94)90448-0. [DOI] [PubMed] [Google Scholar]
  6. Brickley SG, Cull-Candy SG, Farrant M. Development of a tonic form of synaptic inhibition in rat cerebellar granule cells resuling from persistent activation of GABAA receptors. J Physiol. 1996;497:753–759. doi: 10.1113/jphysiol.1996.sp021806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Casado M, Isope P, Ascher P. Involvement of presynaptic N-methyl-D-aspartate receptors in cerebellar long-term depression. Neuron. 2002;33:123–130. doi: 10.1016/s0896-6273(01)00568-2. [DOI] [PubMed] [Google Scholar]
  8. Crepel F, Jaillard D. Pairing of pre- and postsynaptic activities in cerebellar Purkinje cells induces long-term changes in synaptic efficacy in vitro. J Physiol. 1991;432:123–141. doi: 10.1113/jphysiol.1991.sp018380. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. D'Angelo E, De Filippi G, Rossi P, Taglietti V. Synaptic excitation of individual rat cerebellar granule cells in situ: evidence for the role of NMDA receptors. J Physiol. 1995;484:397–413. doi: 10.1113/jphysiol.1995.sp020673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. D'Angelo E, De Filippi G, Rossi P, Taglietti V. Ionic mechanism of electroresponsiveness in cerebellar granule cells implicates the action of a persistent sodium current. J Neurophysiol. 1998;80:493–503. doi: 10.1152/jn.1998.80.2.493. [DOI] [PubMed] [Google Scholar]
  11. De Schutter E, Bower JM. Simulated responses of cerebellar Purkinje cells are independent of the dendritic location of granule cell synaptic inputs. Proc Natl Acad Sci U S A. 1994;91:4736–4740. doi: 10.1073/pnas.91.11.4736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Denk W, Sugimori M, Llinás R. Two types of calcium response limited to single spines in cerebellar Purkinje cells. Proc Natl Acad Sci U S A. 1995;92:8279–8282. doi: 10.1073/pnas.92.18.8279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Eccles JC, Llinás R, Sasaki K. Parallel fiber stimulation and the responses induced thereby in the Purkinje cells of the cerebellum. Exp Brain Res. 1966a;1:17–39. doi: 10.1007/BF00235207. [DOI] [PubMed] [Google Scholar]
  14. Eccles JC, Llinás R, Sasaki K. Intracellularly recorded responses of the cerebellar Purkinje cells. Exp Brain Res. 1966b;1:161–183. doi: 10.1007/BF00236869. [DOI] [PubMed] [Google Scholar]
  15. Eilers J, Augustine GJ, Konnerth A. Subthreshold synaptic Ca2+ signalling in fine dendrites and spines of cerebellar Purkinje neurons. Nature. 1995;373:155–158. doi: 10.1038/373155a0. [DOI] [PubMed] [Google Scholar]
  16. Hamann M, Rossi DJ, Attwell D. Tonic and spillover inhibition of granule cells control information flow through cerebellar cortex. Neuron. 2002;33:625–633. doi: 10.1016/s0896-6273(02)00593-7. [DOI] [PubMed] [Google Scholar]
  17. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch. 1981;391:85–100. doi: 10.1007/BF00656997. [DOI] [PubMed] [Google Scholar]
  18. Hartell NA. Strong activation of parallel fibers produces localized calcium transients and a form of LTD that spreads to distant synapses. Neuron. 1996;16:601–610. doi: 10.1016/s0896-6273(00)80079-3. [DOI] [PubMed] [Google Scholar]
  19. Harvey RJ, Napper RMA. Quantitative study of granule and Purkinje cells in the cerebellar cortex of the rat. J Comp Neurol. 1988;274:151–157. doi: 10.1002/cne.902740202. [DOI] [PubMed] [Google Scholar]
  20. Häusser M, Roth A. Dendritic and somatic glutamate receptor channels in rat cerebellar Purkinje cells. J Physiol. 1997;501:77, 95. doi: 10.1111/j.1469-7793.1997.077bo.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Isope P, Barbour B. Properties of unitary granule cell –>Purkinje cell synapses in adult rat cerebellar slices. J Neurosci. 2002;22:9668–9678. doi: 10.1523/JNEUROSCI.22-22-09668.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Khodakhah K, Armstrong CM. Induction of long-term depression and rebound potentiation by inositol trisphosphate in cerebellar Purkinje neurons. Proc Natl Acad Sci U S A. 1997;94:14009–14014. doi: 10.1073/pnas.94.25.14009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Konnerth A, Llano I, Armstrong CM. Synaptic currents in cerebellar Purkinje cells. Proc Natl Acad Sci U S A. 1990;87:2662–2665. doi: 10.1073/pnas.87.7.2662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Llano I, Marty A, Armstrong CM, Konnerth A. Synaptic- and agonist-induced excitatory currents of Purkinje cells in rat cerebellar slices. J Physiol. 1991;434:183–213. doi: 10.1113/jphysiol.1991.sp018465. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Llinás R, Sugimori M. Electrophysiological properties of in vitro Purkinje cell dendrites in mammalian cerebellar slices. J Physiol. 1980;305:197–213. doi: 10.1113/jphysiol.1980.sp013358. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Momiyama A, Feldmeyer D, Cull-Candy SG. Identification of a native low-conductance NMDA channel with reduced sensitivity to Mg2+ in rat central neurones. J Physiol. 1996;494:479–492. doi: 10.1113/jphysiol.1996.sp021507. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Nowak L, Bregestovski P, Ascher P, Herbet A, Prochiantz A. Magnesium gates glutamate-activated channels in mouse central neurones. Nature. 1984;307:462–465. doi: 10.1038/307462a0. [DOI] [PubMed] [Google Scholar]
  28. Perkel DJ, Hestrin S, Sah P, Nicoll RA. Excitatory synaptic currents in Purkinje cells. Proc R Soc Lond B Biol Sci. 1990;241:116–121. doi: 10.1098/rspb.1990.0074. [DOI] [PubMed] [Google Scholar]
  29. Rosenmund C, Legendre P, Westbrook GL. Expression of NMDA channels on cerebellar Purkinje cells acutely dissociated from newborn rats. J Neurophysiol. 1992;68:1901–1905. doi: 10.1152/jn.1992.68.5.1901. [DOI] [PubMed] [Google Scholar]
  30. Roth A, Häusser M. Compartmental models of rat cerebellar Purkinje cells based on simultaneous somatic and dendritic patch-clamp recordings. J Physiol. 2001;535:445–472. doi: 10.1111/j.1469-7793.2001.00445.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Sabatini BL, Regehr WG. Control of neurotransmitter release by presynaptic waveform at the granule cell to Purkinje cell synapse. J Neurosci. 1997;17:3425–3435. doi: 10.1523/JNEUROSCI.17-10-03425.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Vranesic I, Ijima T, Ichikawa M, Matsumoto G, Knöpfel T. Signal transmission in the parallel fiber-Purkinje cell system visualized by high-resolution imaging. Proc Natl Acad Sci U S A. 1994;91:13014–13017. doi: 10.1073/pnas.91.26.13014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wall ML, Usowicz MM. Development of action potential-dependent and independent spontaneous GABAA receptor-mediated currents in granule cells of postnatal cerebellum. Eur J Neurosci. 1997;9:533–548. doi: 10.1111/j.1460-9568.1997.tb01630.x. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES