Clusters of cerebellar Purkinje cells control their afferent climbing fiber discharge

Edited by Shigetada Nakanishi, Osaka Bioscience Institute, Suita, Japan, and approved August 29, 2013 (received for review February 5, 2013)
September 17, 2013
110 (40) 16223-16228

Significance

The inferior olive, one of the major source of inputs to the cerebellum, sends climbing fibers to Purkinje cells, the key processing units of cerebellar-dependent motor control. Using an optogenetic strategy, we demonstrate that Purkinje cells disinhibit their climbing-fiber afferents via a poly-synaptic circuit. These findings identify a functional closed-loop organization in the olivo-cerebellar circuits that is potentially important for cerebellar motor learning.

Abstract

Climbing fibers, the projections from the inferior olive to the cerebellar cortex, carry sensorimotor error and clock signals that trigger motor learning by controlling cerebellar Purkinje cell synaptic plasticity and discharge. Purkinje cells target the deep cerebellar nuclei, which are the output of the cerebellum and include an inhibitory GABAergic projection to the inferior olive. This pathway identifies a potential closed loop in the olivo-cortico-nuclear network. Therefore, sets of Purkinje cells may phasically control their own climbing fiber afferents. Here, using in vitro and in vivo recordings, we describe a genetically modified mouse model that allows the specific optogenetic control of Purkinje cell discharge. Tetrode recordings in the cerebellar nuclei demonstrate that focal stimulations of Purkinje cells strongly inhibit spatially restricted sets of cerebellar nuclear neurons. Strikingly, such stimulations trigger delayed climbing-fiber input signals in the stimulated Purkinje cells. Therefore, our results demonstrate that Purkinje cells phasically control the discharge of their own olivary afferents and thus might participate in the regulation of cerebellar motor learning.
The cerebellar cortex is involved in a wealth of functions, from the control of posture to higher cognitive processes (13). Purkinje cells (PCs) are key processing units of the cerebellar cortex (4): each PC receives more than 175,000 parallel fiber synaptic inputs carrying information about the ongoing sensory-motor context. It also receives a single inferior olive afferent, the climbing fiber, which triggers a complex spike (CS), modulates PC firing (5), controls synaptic input plasticity, and has been proposed to carry error and clock signals to the cerebellum (2, 48). PCs are grouped in multiple parasagittal microzones, each receiving projections from separate areas of the inferior olive and projecting to subregions of the cerebellar nuclei (CN) (912). In the CN, PCs make inhibitory contacts on excitatory neurons that project to various premotor areas and propagate cerebellar computations to the motor system. Anatomical evidence indicates that PC terminals also contact CN inhibitory neurons that target inferior olive cells (13, 14). This nucleo-olivary pathway is topographically organized in multiple parallel projections to the inferior olive subnuclei (15), suggesting the existence of closed olivary-cortico-nuclear loops. Therefore, the discharge of a population of PCs in a microzone might not only shape the output of the cerebellum but also control its afferent climbing-fiber signal. Previous studies have shown that stimulation of the nucleo-olivary pathway significantly reduces olivary cell firing (1618) and that pharmacological and genetic manipulations of PCs or olivary cell activity induce reciprocal modulations of the firing rate of PCs and climbing fibers (19, 20). These results indicate that PCs may tonically modulate the nucleo-olivary pathway. However, whether the cerebellar cortex can phasically recruit this pathway and whether this circuit functions as a closed loop is currently unknown. We thus set out to study the impact of phasic stimulations of PCs on cortico-nucleo-olivary loops. To control selectively PC firing rates, we engineered a mouse line expressing Channelrhodopsin-2 (ChR2) specifically in PCs. By combining optogenetic stimulation and in vivo electrophysiological recordings, we show that stimulating a set of PCs in a region of the cerebellar cortex triggers a restricted inhibition in the cerebellar nuclei and a transient disinhibition of the inferior olive cells that project to this set of PCs.

Results

L7-ChR2-eYFP Mice Engineering.

A genetically modified mouse (L7-ChR2) that expresses ChR2(H134R)-Yellow Fluorescent Protein (YFP) under the control of the regulatory elements of the pcp2 gene was created using the bacterial artificial chromosome modification strategy (SI Methods, Fig. 1A, and Fig. S1A) (21). Specific expression of the ChR2-YFP fusion protein in all PCs of the cerebellar cortex was detected by YFP fluorescence on cerebellar sections and confirmed by anti-calbindin immunostaining (Fig. 1B and Fig. S1B). GABA immunostaining and whole-cell patch-clamp recordings in acute cerebellar slices showed no expression in molecular layer interneurons or Golgi cells (Fig. 1D). Behavioral tests were performed to assess motor performance of the mutant mice. No difference was found between L7-ChR2 mice and wild-type littermates (Fig. 1C), indicating that this transgene expression does not perturb motor functions.
Fig. 1.
Generation and characterization of L7-ChR2(H134R)-eYFP mice. (A) Diagram of the modified BAC containing the ChR2(H134R)-eYFP cDNA under the control of the L7/pcp2 gene regulatory elements. (B) Transgene expression monitored by YFP fluorescence in cerebellar sections (Upper Left); (Lower Left) calbindin immunolabeling; (Upper Right) GABA immunolabeling; (Lower Right) overlay. (C) Falling latencies in the rotarod test during three sessions of three trials at 0–45 rpm (tested by multifactorial ANOVA, P = 0.392; WT, n = 10; ChR2, n = 9). (D) Molecular layer interneurons, Golgi cells, and Purkinje cells recorded in acute cerebellar slices visualized using Alexa 594 in the pipette and wide-field illumination at 473 nm. (Scale bar: 20 µm.) No current was observed in molecular layer interneurons (n = 4) and Golgi cells (n = 6).

Control of Purkinje Cells by Light.

To characterize PC activation by ChR2, in vitro recordings were performed on acute cerebellar slices using whole-cell patch-clamp. Wide-field illumination with blue light (Fig. 2A and SI Methods) systematically activated an inward current in all PCs tested (n = 65; Fig. 2 B–D). By restricting the field of illumination (SI Methods), we could show that ChR2 channels are expressed throughout PC dendrites (Fig. 2B) and estimate their density to be 150–300 channels per µm2 illuminated [Fig. S2, assuming a unitary conductance of 100 fS (22)]. We then set out to characterize the current induced in PCs by illumination as a function of irradiance and pulse duration. For a pulse duration of 1 ms, increasing irradiance enhanced the amplitude of the photocurrent up to several hundred picoamps (Fig. 2C and SI Methods) with a decay time constant of 17.3 ± 8.5 ms (n = 11). For longer pulses (100 ms), the current rapidly desensitized to a stationary current at 58 ± 3% of the peak for stimulation frequency below 0.05 Hz (n = 5) (Fig. 2 C, iv), in agreement with previous work (23). The pulse frequency used in most of the experiments reported in this work was above 0.05 Hz, leading to a smaller peak current but a similar stationary current (Fig. 2C). Current-clamp recordings showed that light-evoked current carrying a charge above 14.7 ± 10 pC (n = 4) elicited an action potential in PCs (Vm = −60 mV) and that action potentials were repeated during burst illuminations of up to 10 Hz for irradiance above 8 mW/mm2 (Fig. 2D). These results demonstrate that, despite its very low conductance, the density of ChR2 channels is high enough to produce large currents and reliably elicit action potentials in PCs of L7-ChR2 mice.
Fig. 2.
Electrophysiological characterization of L7-ChR2(H134R)-eYFP mice in vitro and in vivo. (A) Schematics of acute cerebellar slices illustrating whole-cell recordings combined with laser-point scan photostimulation (Upper) or wide-field illumination (Lower). (B, Upper) Point-scan photostimulation (1-ms pulses) of Purkinje-cell dendrites and average current elicited at identified color-code points. (Lower) normalized current (± SEM) plotted against the distance from the soma (n = 7, Mann-Whitney test, P < 0.05). (C, i) Average whole-cell currents recorded in a PC with 1-ms wide-field LED illumination at various irradiances. (ii) Average charge against irradiance (n = 8 cells; black, mean ± SEM). (iii) Example of 10 successive traces superimposed from one cell. (iv) Desensitization of peak current at 0.5 Hz (100-ms pulses). Panels iii and iv are from different cells. (D, i) Action potential recorded in current-clamp mode elicited by a 1-ms light pulse (top) and corresponding photocurrent recorded in voltage clamp (bottom). (ii) Action potential initiation for burst of illumination (1-ms pulses at 10 Hz). (E) Schematic of in vivo juxtacellular recordings of PCs combined with optical fiber illumination. (F) Example traces recorded in juxtacellular mode during 100-ms pulses. (G) Relationship between PC firing rate and irradiance intensity (n = 14 cells). Triangle, cells deeper than 400 µm. (H) Delay of occurrence (D) of the first spike during illumination, compared with the preceding spike (BS) at various irradiances (n = 20 cells, 16 mice). Box plots show median values and 25/75% percentiles. Control median delay = 10.9 ms (quartiles: 8.4, 17.1); median delay = 5.5 ms (quartiles: 3.9, 8.8) at 2.5 mW/mm2 and 3 ms (quartiles 2.2, 4.2) at 19 mW/mm2, Kruskal–Wallis One-Way ANOVA, P < 0.001. Pairwise multiple comparison, Dunn's method, *P < 0.05. Median delays during illumination were not significantly different. (I) Time course of the effect of illumination on the firing rate. Post stimulus time histograms (PSTHs) were normalized to the baseline frequency before averaging. Only cells with strong increases in firing rate (greater than twofold increase) were used in this panel. Kruskal–Wallis One-Way ANOVA, P < 0.001. Pairwise multiple comparisons, Dunn's method. *P < 0.05.
Next, the activation of PCs was characterized in vivo using juxtacellular recordings in anesthetized mice. Blue-light illumination was delivered through optical fibers coupled to a LED and positioned in the vicinity of the brain surface (SI Methods and Fig. 2E). In contrast to in vitro conditions, the irradiance received by the PCs in vivo depends strongly on the scattering of light in the tissue and varies as a function of the depth and orientation of the dendrite of the PC. PCs were identified by the occurrence of complex spikes (CSs), produced by the climbing-fiber input, and by the frequency of spontaneous simple spikes (mean CS rate = 0.44 ± 0.63 Hz; mean simple spike rate = 39.7 ± 3.9 Hz, n = 19, n = 16 mice). Increasing irradiance intensity led to an increase in simple spike firing rate in all cells (n = 14) whose somata lie in the first hundreds of micrometers from the surface of the tissue (Fig. 2 F and G). This increase in simple spike-firing rate could continue up to 500 ms (the longest duration tested) and reach up to 250 Hz. Further increasing the intensity induced a depolarization block consisting of a burst of spikes at light onset followed by a complete suppression of PC firing during the remaining time of illumination (Fig. 2F). Because of the diffusion of light (SI Methods; and see Fig. S5), depolarization block was not observed for cells distant from the tip of the optic fiber by more than 300 µm even at tip irradiance above 30 mW/mm2. Also, no depolarization block was elicited with short pulses of light (≤2 ms). To quantify the onset of the effect of illumination, the latency of the first spike after the onset of illumination was measured. This latency was significantly shortened to a median delay of 5.5 ms (n = 19) (Fig. 2H) for irradiance as low as 2.5 mW/mm2, and further decreased to a minimum median of 3.0 ms for an irradiance intensity of 19 mW/mm2 (Fig. 2H). At the offset of the illumination, the time required to resume basal firing rate ranged between 10 and 30 ms (Fig. 2I). These findings demonstrate that ChR2-expressing PCs can be excited by light with high temporal precision in vivo.

Inhibition of Cerebellar Nuclei Neurons by Purkinje Cells.

To assess whether the photostimulation of PCs was able to modulate the activity of their target neurons in the cerebellar nuclei (CN), in vivo extracellular recordings were performed with tetrodes in the CN while illuminating the ipsilateral cerebellar cortex (SI Methods and Fig. 3 A and B). To increase the probability of stimulating the group of PCs that target the recording site in the CN, we used illumination parameters that maximize the number of PCs excited (optical fiber of 600 µm diameter, 100-ms pulses, irradiance at 60 mW/mm2) (SI Methods). Using these conditions, only 10% of activated PCs are likely to experience intensities that trigger a depolarization block (SI Methods; and see Fig. S5). By illuminating the lateral part of Crus I, sites that induced inhibition in recorded units from the interposed nucleus were found (26/48 tested sites; 21 animals) (Fig. 3 CG). The suppression of firing was often accompanied by a small deflection in the field potential, presumably due to synaptic activation (red arrow, Fig. 3C). Usually, several cells were recorded simultaneously by the tetrode at each recording site (3.7 ± 1.7 cells per site) (Fig. 3B). Among the 97 cells recorded at responding sites (i.e., where at least one cell was inhibited), 70 cells were inhibited by Crus I illumination whereas 27 did not respond (average of 76 ± 26% inhibited cells at each site). These cells exhibited a mean firing rate of 18.0 ± 13.1 Hz and a coefficient of variation of 0.58 ± 0.29 for the interspike interval.
Fig. 3.
Effective silencing of cerebellar nuclear cells by photostimulation of Purkinje cells in vivo. (A) Propagation of blue light in the cerebellar cortex (color coded part of the image) for a 600-µm optical fiber placed on the surface of the cerebellum (grayscale part of the image). In the color-coded part, the color of each pixel corresponds to its value expressed as a percentage of the intensity of the brightest pixel. The 10% isoline is delineated in white (i.e., corresponding to ∼2 mW/mm2 if the brightest pixel immediately under the fiber corresponds to 20 mW/mm2). (B) Illustration of tetrode recordings in cerebellar nuclei. (Left) Schematics of the experiment. (Center) Example traces (vertical ticks of different colors signal different spikes; numbers identify the four channels of the tetrode). (Right) Average unfiltered waveforms on the four channels (same color code as for Center), superimposed on 10 successive events (gray traces). (C) Example of a recording obtained on one channel in the interposed nucleus. The average extracellular potential (LFP, local field potential, red) recorded in CN is superimposed on 30 successive sweeps (gray lines) aligned on the onset of photo-stimulation (2 ms, 60 mW/mm2). Note the suppression of extracellular (negative) spikes after the stimulation. The average LFP was calculated from 232 sweeps. The red arrow signals the deflection of LFP. (D) Example of a raster plot (Upper) and corresponding PSTH (Lower; bin, 1 ms), showing a complete suppression of firing following the stimulation (blue bar); the example is the same as in C. (E, Left) Example of PSTHs for one cell tested with different light intensities (pulse duration, 25 ms). (Right) Average duration of full inhibition plotted against the irradiance at the fiber tip (at 25 ms) or pulse duration (at 60 mW/mm2) (n = 74 cells) (SI Methods). The shaded area indicates the SEM. (F) Example cell inhibited in response to the illumination of Crus I (Upper, PSTH), but not Crus II (Lower, PSTH). Bin, 5 ms. Blue bar, light pulse (100 ms, 60 mW/mm2). (G) A 3D view of the distribution of CN recording sites where inhibited cells were found (red) or not (gray) during Crus I illumination (100 ms, 19 mW/mm2) on the ipsilateral cerebellar Crus I region (AP, −6.1; ML, 3.3). The 3D rectangle arrow identifies the region where most responsive sites are clustered. See Fig. S3 planar projections.
To quantify the effectiveness of the inhibition, the parameters of illumination were then varied (Fig. 3E and Fig. S3). Decreasing the intensity of irradiance to 19 mW/mm2 did not change the proportion of responding cells at responding sites in the CN (72.8 ± 31% responding cells from 57 cells recorded at 25 responding sites in four mice), showing that this intensity is enough to recruit the minimum number of PCs necessary for CN inhibition. The duration of full inhibition (complete suppression of firing) increased with pulse duration and irradiance intensity (Fig. 3E). For 25-ms pulses, a complete inhibition of some CN units could be observed at irradiance higher than 3 mW/mm2. The mean latency to full inhibition was slow but variable (mean = 18.9 ± 11.7 ms at 19 mW/mm2 and 19.7 ± 12.15 ms at 60 mW/mm2) (Fig. S3 DF), which is consistent with the requirement of a large number of activated PCs for complete inhibition of CN firing. Finally, trains of 20-Hz and 30-Hz light pulses induced a rhythmic modulation of CN firing (Fig. S4). Taken together, these data show that illumination of PCs allows a dynamic control of CN firing (24) and that the duration of inhibition increases with the intensity and duration of light stimulations. Because PC inhibition has been proposed to produce a rebound excitation in CN neurons (25), we examined whether the instantaneous frequency was significantly enhanced relative to baseline values when the cells resumed their firing. The inhibition produced by PC inputs did not trigger a detectable rebound excitation even at high irradiance (60 mW/mm2) in most CN cells (about 90%) but was followed by either a progressive or a direct return of CN discharge to its baseline value after the end of the stimulation (Fig. S3). To test the spatial specificity of the responses, the light source at the surface of the cerebellum was displaced from Crus I to Crus II, once a responding site was found; in most cells, the light-induced inhibition was lost (n = 12/14 cells recorded, six recording sites, four mice) (Fig. 3F), consistent with the zonal anatomical organization of the cortico-nuclear pathway. We then mapped the region of the CN inhibited by activation of ∼1000–3000 PCs in Crus I (pulse duration: 100 ms, irradiance 19 mW/mm2; see Fig. 3A, Fig. S5, and SI Methods). The cellular responses were explored by systematic penetrations in ranges of stereotaxic coordinates [antero-posterior (AP) −5.9 to 6.4, medio-lateral (ML) 1.5; 2.5, depth 2.1, 3.3], and the most responsive sites were located in a small area (AP −6.2, −6.4, ML 1.7, 1.9, depth 2.7; 3.2) (n = 106 cells, 53 recording sites, four mice) (Fig. 3G and Fig. S3).

Activation of Purkinje Cells Controls Inferior Olivary Neurons Discharge.

We then set out to assess whether the stimulation of a population of PCs can stop the firing of CN neurons and then influence the discharge of inferior olivary neurons. Complex spikes (CS) in PCs were monitored in juxtacellular recordings in vivo as a readout of olivary cells discharge. The recorded PC and its neighbors were then excited using small optical fibers (diameter 50–200 µm), and the irradiance was adjusted to obtain a strong activation without depolarization block (mean = 20.6 ± 11.7 mW/mm2). Local illumination by pulses of light lasting 35–500 ms elicited a CS in 25 out of 42 PCs (n = 25 mice) (Fig. S6) with a mean latency of 138 ± 39 ms after the onset of the stimulation (Fig. 4 A and B). No response was observed for pulse durations of 10 and 20 ms. For pulse durations of 100 ms or less, most stimulation-induced CSs were elicited tens of milliseconds after the end of illumination, thus ruling out the possibility that CSs are a direct (artifactual) consequence of light stimulation. Indeed, no effect of illumination was detected in wild-type control mice either in PCs or in other cell types (n = 4 PCs, 7 non-PCs, presumably inhibitory interneurons from four wild-type mice). Interestingly, once a rebound CS was observed, the response rate was independent of the size of the fiber used (F(1,23) = 2.85, P = 0.10) and the irradiance at the tip of the optic fiber (F(1,52) = 0.73, P = 0.40). Therefore, data obtained using different irradiance intensities and fiber diameters were pooled together. No difference was observed in the onset of the CS responses for pulse durations between 35 and 500 ms whereas the offset of the response increased with pulse duration (Fig. 4C), suggesting that the beginning and the end of the response are locked respectively to the beginning—with a long delay—and to the end of the stimulation. Therefore, CS were evoked by pulses of irradiance around 20 mW/mm2 using a 100-µm optical fiber that should activate about 200–500 PCs (Fig. 4A and Fig. S5). Taken together, these results indicate that, provided a critical number of PCs were illuminated for a minimal period (35 ms), the disinhibition of olivary neurons resulted in CSs evoked with a minimal delay of 80–100 ms. Moreover, experiments in which the optical fiber was moved along the transverse axis above the surface showed that the evoked CS response (i.e., the inferior olive disinhibition) decreased and disappeared together with the simple spike response (i.e., the direct activation of the recorded cell) (Fig. 4E). These results indicate that the set of PCs that control the climbing fiber afferent to a PC is localized close to that PC (Fig. 4F), consistent with the topographical organization of the cortico-nucleo-olivary circuit.
Fig. 4.
Photostimulation of Purkinje cells controls inferior olivary cell discharge. (A, i) Blue–light propagation observed in a transverse section of the cerebellar cortex using a 100-µm optical fiber. The same code as in Fig. 3A. (ii) Individual trace from in vivo juxtacellular PC recordings and corresponding raster plot for both simple spikes (gray ticks) and CSs (black ticks, the signature of olivary cell discharge in PCs). Light illumination elicited CSs in the recorded PC. The arrow indicates the evoked complex spikes. (B) Example raster plot and average peri-stimulation firing rate (obtained from the density of spikes convolved with a Gaussian kernel with a 10-ms variance) of CSs recorded with 10-, 50-, and 100-ms illumination pulses (blue rectangles). Red lines identify the time periods with significant increase in the CS firing rate relative to baseline. (C) Mean latency from stimulation onset to response onset (Top, no significant difference, F1,56 = 0.074, P = 0.79) and response end (Middle, significant difference across conditions, F1,56 = 465, P < 10−4), and mean response rate (spike/stim, Bottom, significant difference across conditions, F1,56 = 56, P < 10−4) for 35-, 50-, 100-, and 500-ms pulses (n = 4, 13, 19, and 11 cells, respectively; 50–200 trials by conditions). *, **, ***P < 0.05, 0.01, and 0.001, respectively. (D) Influence of the recent history of CS firing on the timing of the disinhibitory response. For each cell, the PSTH (Top) was recalculated to realign the trials on the time of the first CS after stimulus onset (Middle, green CSs are aligned, trials without CS are not represented). In the example (Middle), the trials are sorted by decreasing ISIs before the response CS showing that the first response spikes occurred preferentially after specific ISIs (red arrows); in some trials, the illumination triggered two CSs yielding a peak at positive times. The realigned PSTH were averaged for eight cells (Bottom) and compared with the expected values after randomly shifting the preresponse CS (±300 ms). The disinhibitory response occurred preferentially after specific intervals (*) after the last prestimulation CS spikes. Dashed and dotted lines represent shuffled PSTH ± 1 SD and 2 SD respectively. (E) Increase of CS frequency against the modulation of simple spike frequency after blue-light illumination at different optical-fiber position. Points acquired from the same cell are connected by a gray line. A linear fit (black line, r2 = 0.7) with confidence interval at 95% (dashed black line) is superimposed on the data showing the correlation between direct PC activation and CS response (and the simultaneous disappearance of both responses when the fiber is moved away). (F) Diagram of the synaptic connections in the olivo-cortico-nucleo-olivary loop and illustration of discharge in Purkinje cells, CN neurons, and olivary cells based on our results. Blue light activates PCs (simple spike firing rate increased during stimulation, gray ticks) that stop CN neuron discharge and consequently disinhibit inferior olive neurons. A CS is then observed in PCs. Ticks illustrate neuronal discharge.
There is a substantial jitter in the time of occurrence of the CS after the onset of illumination (several tens of milliseconds) (Fig. 4 B and D), suggesting a temporal fluctuation of the responsiveness of the inferior olivary neurons to disinhibition. There was indeed a correlation between the probability of occurrence of the CS following the stimulation of PCs and the CS baseline firing rate (Fig. S6B), indicating that PC illumination is more efficient when olivary cells are more depolarized. Because inferior olivary neurons are known to express subthreshold oscillations spontaneously (26), we examined whether such oscillations would condition the timing of the CS evoked by PCs stimulations. Poststimulus time histograms were constructed by aligning individual trials on the time of the first CS after the onset of stimulation (Fig. 4D, green ticks; 100 ms pulse, n = 8), and the evoked CS were found to occur preferentially ∼225 ms and ∼450 ms after a preceding spontaneous CS (Fig. 4D, Bottom). These results show that the disinhibition of olivary neurons triggers CS spikes in phase with an ∼4-Hz subthreshold oscillation.

Discussion

In this study, a transgenic mouse line expressing ChR2 specifically in PCs is described and used to investigate the functional organization of the olivo-cerebellar circuit. The level of ChR2(H134) expression yielded reliable and sustained currents (23, 27) in PCs and increased the cells’ discharge. Focal illumination in one specific part of the cerebellar cortex (crus I) transiently inhibited 75% of the cells recorded simultaneously from the same tetrode in a restricted area of the cerebellar nuclei. Lowering the intensity of stimulation did not change the percentage of responsive cells at responsive sites, suggesting that the nonresponsive cells receive weaker PC inputs, as it has been suggested for nuclear GABAergic interneurons (28). Although light stimulations took a few milliseconds to evoke spikes in PCs, the complete suppression of firing in CN cells was variable among cells and was reached on average in ∼20 ms, indicating the need for a high convergence of presynaptic PCs or a temporal summation of inhibitory PC inputs to fully inhibit CN neurons. Surprisingly, although sustained inhibition was induced in CN neurons following activation of PCs, rebound excitation (29, 30) was observed in only ∼10% of the cells, in agreement with the study of Alviña et al. (31). Anesthesia might alter physiological properties of CN neurons. However, recent studies (25, 32) have demonstrated that rebound excitation in CN neurons requires the extensive activation of the presynaptic PCs. Thus, a more likely interpretation is that our protocol of illumination leads to poorly synchronized PC discharges and/or fails to recruit enough presynaptic PCs.
The stimulation of PCs during tens of milliseconds elicited olivary discharge with an onset latency close to 100 ms and an offset latency that increased with stimulation duration. Such long-latency phenomenon has been previously suggested in behaving monkeys (33) by analyzing the simple spike discharge preceding complex spikes. In our experiments, the CS response started while the CN neurons were still inhibited, which rules out the recruitment of an (indirect) excitatory output of the CN to the inferior olive (15, 34). However, the evoked CS is consistent with properties of the nucleo-olivary inhibitory pathway (35), which is dominated by asynchronous release (17) leading to long latencies for inhibition (18); this mode of transmission should delay the transmission of changes in CN firing rate and filter out very transient suppressions of CN firing. Moreover, the variability of the latency of the CS response across trials seems to be partly due to intrinsic subthreshold fluctuations (26) (in our case, 4-Hz oscillation) of the membrane potential of inferior olivary neurons. Indeed, the maximal probability of occurrence of CS response was found in phase with olivary oscillations. Overall, the characteristics of the CS response to increased PC discharge are consistent with a disinhibition of the inferior olive via the nucleo-olivary pathway. Our results are also consistent with a topographical organization of the cortico-nucleo-olivary pathway. CN neurons inhibition was lost when the illumination was moved to a neighboring lobule. Moreover, the CS response in the recorded PCs disappeared as soon as direct excitation (evidenced by simple spike modulation) was lost when the optic fiber was moved away in the transverse direction. The combination of these results and those from previous studies (9, 16, 24, 26, 27, 3032, 36) strongly argue for the existence of closed cortico-nucleo-olivary loops where spatially restricted sets of PCs control their afferent climbing fibers. The spatial extension of the set of PCs excited by our smallest illumination (around 300 µm diameter) is closer to the scale of cerebellar zones than of microzones (11). Indeed, recent imaging studies in vivo describing synchronized CSs during sensory stimulation identified microbands narrower than 100 µm (37, 38). Testing whether the control of the CS is segregated across such microzones may require the use of specific strategies (39) to restrict the expression of the opsin to single microzones and therefore circumvent the difficulty of illuminating very narrow bands of cells. The closed cortico-nucleo-olivary loops will favor the triggering of a CS volley in a subset of PCs shortly after their firing rate is increased. Moreover, because the nucleo-olivary inhibitory pathway targets gap junctions between olivary cells and promotes their decoupling (7, 40), the inhibition of this pathway should promote CS synchrony (19, 41). This feedback excitatory loop could also modulate plasticity in the cerebellar cortex, for example, at the parallel fiber to PC synapse because the delay between the onset of the increased PCs firing and the rebound CS (around 100 ms) matches the optimal interval between parallel-fiber and climbing-fiber discharges required for the induction of long-term depression (42, 43). Indeed, the nucleo-olivary pathway has been proposed to play important roles in conditioning (44, 45), and an appealing possibility is that the cortico-nuclear pathway controls the contribution of the nucleo-olivary pathway to motor learning. A wide range of theoretical and experimental evidence has led to the proposal that learning mechanisms in the cerebellar cortex underlie the formation and the storage of internal models of the sensory-motor system (46). These models convert desired movements into motor commands or predict sensory outcomes of planned movements (47); the adjustment of the models is performed by supervised learning via the climbing fibers. The cortico-nucleo-olivary pathway described in our study provides a way to propagate the predictions computed in the cerebellar cortex to the inferior olive, where it can be compared, with an appropriate delay, with the actual outcome of the ongoing task carried by other inputs to the inferior olive. Therefore, the closed cortico-nucleo-olivary loops may play an essential role in adjusting the internal models in the cerebellum.

Methods

All experimental procedures conform to Centre National de la Recherche Scientifique and National Institutes of Health guidelines on animal experimentation. A BAC transgenic mouse expressing channelrhodopsin-2(H134R) under the control of the regulatory elements of the L7-pcp2 gene was generated (21) (SI Methods and Fig. S1). In vivo extracellular recordings of PCs were performed in anesthetized animals (SI Methods). Means were given ± SD unless otherwise stated.

Acknowledgments

We thank Pr Karl Deisseroth for kindly providing the pAAV-double floxed-hChR2(H134R)EYFP-WPRE-pA vector. We thank the Mouse Clinical Institute, Laurence Huck, and Cassidy Fiford for technical assistance. We thank Samuel Garcia for support on software and Boris Barbour and Nancy Grant for comments on this manuscript. This work was supported by the Centre National pour la Recherche Scientifique, Ecole Normale Supérieure, Institut National de la Santé et de la Recherche Médicale, Université de Strasbourg, Ministère de la Recherche, Agence Nationale pour la Recherche Grants ANR-09-MNPS-038 CeCoMod, ANR-2010-JCJC-1403-1 MicroCer, and ANR-GUI-AAP-04 Sensocode; the INTERREG IV Rhin Supérieur Program; and the European Funds for Regional Development, TIGER project. Publication costs are supported by the Neurex network (www.neurex.org).

Supporting Information

Supporting Information (PDF)
Supporting Information

References

1
M Ito, Cerebellar circuitry as a neuronal machine. Prog Neurobiol 78, 272–303 (2006).
2
WT Thach, HP Goodkin, JG Keating, The cerebellum and the adaptive coordination of movement. Annu Rev Neurosci 15, 403–442 (1992).
3
P Dean, J Porrill, C-F Ekerot, H Jörntell, The cerebellar microcircuit as an adaptive filter: Experimental and computational evidence. Nat Rev Neurosci 11, 30–43 (2010).
4
M Ito The Cerebellum and Neural Control (Raven, New York, 1984).
5
CI De Zeeuw, et al., Spatiotemporal firing patterns in the cerebellum. Nat Rev Neurosci 12, 327–344 (2011).
6
RR Llinás, Cerebellar motor learning versus cerebellar motor timing: The climbing fibre story. J Physiol 589, 3423–3432 (2011).
7
Y Yarom, D Cohen, The olivocerebellar system as a generator of temporal patterns. Ann N Y Acad Sci 978, 122–134 (2002).
8
NH Barmack, Inferior olive and oculomotor system. Prog Brain Res 151, 269–291 (2006).
9
O Oscarsson, Functional units of the cerebellum—sagittal zones and microzones. Trends Neurosci 2, 143–145 (1979).
10
J Voogd, M Glickstein, The anatomy of the cerebellum. Trends Neurosci 21, 370–375 (1998).
11
R Apps, R Hawkes, Cerebellar cortical organization: A one-map hypothesis. Nat Rev Neurosci 10, 670–681 (2009).
12
M Glickstein, F Sultan, J Voogd, Functional localization in the cerebellum. Cortex 47, 59–80 (2011).
13
BJ Nelson, JC Adams, NH Barmack, E Mugnaini, Comparative study of glutamate decarboxylase immunoreactive boutons in the mammalian inferior olive. J Comp Neurol 286, 514–539 (1989).
14
TM Teune, J van der Burg, CI de Zeeuw, J Voogd, TJ Ruigrok, Single Purkinje cell can innervate multiple classes of projection neurons in the cerebellar nuclei of the rat: A light microscopic and ultrastructural triple-tracer study in the rat. J Comp Neurol 392, 164–178 (1998).
15
TM Teune, J van der Burg, J van der Moer, J Voogd, TJ Ruigrok, Topography of cerebellar nuclear projections to the brain stem in the rat. Prog Brain Res 124, 141–172 (2000).
16
F Bengtsson, G Hesslow, Cerebellar control of the inferior olive. Cerebellum 5, 7–14 (2006).
17
AR Best, WG Regehr, Inhibitory regulation of electrically coupled neurons in the inferior olive is mediated by asynchronous release of GABA. Neuron 62, 555–565 (2009).
18
P Bazzigaluppi, T Ruigrok, P Saisan, CI De Zeeuw, M de Jeu, Properties of the nucleo-olivary pathway: An in vivo whole-cell patch clamp study. PLoS ONE 7, e46360 (2012).
19
SP Marshall, EJ Lang, Local changes in the excitability of the cerebellar cortex produce spatially restricted changes in complex spike synchrony. J Neurosci 29, 14352–14362 (2009).
20
X Chen, et al., Disruption of the olivo-cerebellar circuit by Purkinje neuron-specific ablation of BK channels. Proc Natl Acad Sci USA 107, 12323–12328 (2010).
21
S Gong, et al., A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature 425, 917–925 (2003).
22
JY Lin, A user’s guide to channelrhodopsin variants: Features, limitations and future developments. Exp Physiol 96, 19–25 (2011).
23
A Berndt, et al., High-efficiency channelrhodopsins for fast neuronal stimulation at low light levels. Proc Natl Acad Sci USA 108, 7595–7600 (2011).
24
AL Person, IM Raman, Purkinje neuron synchrony elicits time-locked spiking in the cerebellar nuclei. Nature 481, 502–505 (2012).
25
F Bengtsson, CF Ekerot, H Jörntell, In vivo analysis of inhibitory synaptic inputs and rebounds in deep cerebellar nuclear neurons. PLoS ONE 6, e18822 (2011).
26
S Khosrovani, RS Van Der Giessen, CI De Zeeuw, MT De Jeu, In vivo mouse inferior olive neurons exhibit heterogeneous subthreshold oscillations and spiking patterns. Proc Natl Acad Sci USA 104, 15911–15916 (2007).
27
S Zhao, et al., Cell type–specific channelrhodopsin-2 transgenic mice for optogenetic dissection of neural circuitry function. Nat Methods 8, 745–752 (2011).
28
M Uusisaari, T Knöpfel, Functional classification of neurons in the mouse lateral cerebellar nuclei. Cerebellum 10, 637–646 (2011).
29
R Tadayonnejad, et al., Rebound discharge in deep cerebellar nuclear neurons in vitro. Cerebellum 9, 352–374 (2010).
30
CD Aizenman, DJ Linden, Regulation of the rebound depolarization and spontaneous firing patterns of deep nuclear neurons in slices of rat cerebellum. J Neurophysiol 82, 1697–1709 (1999).
31
K Alviña, JT Walter, A Kohn, G Ellis-Davies, K Khodakhah, Questioning the role of rebound firing in the cerebellum. Nat Neurosci 11, 1256–1258 (2008).
32
FE Hoebeek, L Witter, TJ Ruigrok, CI De Zeeuw, Differential olivo-cerebellar cortical control of rebound activity in the cerebellar nuclei. Proc Natl Acad Sci USA 107, 8410–8415 (2010).
33
RC Miall, JG Keating, M Malkmus, WT Thach, Simple spike activity predicts occurrence of complex spikes in cerebellar Purkinje cells. Nat Neurosci 1, 13–15 (1998).
34
TJ Ruigrok, CI de Zeeuw, J van der Burg, J Voogd, Intracellular labeling of neurons in the medial accessory olive of the cat. I. Physiology and light microscopy. J Comp Neurol 300, 462–477 (1990).
35
EJ Lang, I Sugihara, R Llinás, GABAergic modulation of complex spike activity by the cerebellar nucleoolivary pathway in rat. J Neurophysiol 76, 255–275 (1996).
36
I Sugihara, et al., Projection of reconstructed single Purkinje cell axons in relation to the cortical and nuclear aldolase C compartments of the rat cerebellum. J Comp Neurol 512, 282–304 (2009).
37
I Ozden, MR Sullivan, HM Lee, SS Wang, Reliable coding emerges from coactivation of climbing fibers in microbands of cerebellar Purkinje neurons. J Neurosci 29, 10463–10473 (2009).
38
KK Ghosh, et al., Miniaturized integration of a fluorescence microscope. Nat Methods 8, 871–878 (2011).
39
T Tsubota, Y Ohashi, K Tamura, Optogenetics in the cerebellum: Purkinje cell-specific approaches for understanding local cerebellar functions. Behav Brain Res, 2013).
40
CI de Zeeuw, JC Holstege, TJ Ruigrok, J Voogd, Ultrastructural study of the GABAergic, cerebellar, and mesodiencephalic innervation of the cat medial accessory olive: Anterograde tracing combined with immunocytochemistry. J Comp Neurol 284, 12–35 (1989).
41
JP Welsh, R Llinás, Some organizing principles for the control of movement based on olivocerebellar physiology. Prog Brain Res 114, 449–461 (1997).
42
P Safo, WG Regehr, Timing dependence of the induction of cerebellar LTD. Neuropharmacology 54, 213–218 (2008).
43
SS Wang, W Denk, M Häusser, Coincidence detection in single dendritic spines mediated by calcium release. Nat Neurosci 3, 1266–1273 (2000).
44
A Rasmussen, DA Jirenhed, G Hesslow, Simple and complex spike firing patterns in Purkinje cells during classical conditioning. Cerebellum 7, 563–566 (2008).
45
JJ Kim, DJ Krupa, RF Thompson, Inhibitory cerebello-olivary projections and blocking effect in classical conditioning. Science 279, 570–573 (1998).
46
M Ito, Control of mental activities by internal models in the cerebellum. Nat Rev Neurosci 9, 304–313 (2008).
47
M Kawato, Internal models for motor control and trajectory planning. Curr Opin Neurobiol 9, 718–727 (1999).

Information & Authors

Information

Published in

Go to Proceedings of the National Academy of Sciences
Go to Proceedings of the National Academy of Sciences
Proceedings of the National Academy of Sciences
Vol. 110 | No. 40
October 1, 2013
PubMed: 24046366

Classifications

Submission history

Published online: September 17, 2013
Published in issue: October 1, 2013

Keywords

  1. motor control
  2. olivo-cerebellar loop
  3. complex spikes

Acknowledgments

We thank Pr Karl Deisseroth for kindly providing the pAAV-double floxed-hChR2(H134R)EYFP-WPRE-pA vector. We thank the Mouse Clinical Institute, Laurence Huck, and Cassidy Fiford for technical assistance. We thank Samuel Garcia for support on software and Boris Barbour and Nancy Grant for comments on this manuscript. This work was supported by the Centre National pour la Recherche Scientifique, Ecole Normale Supérieure, Institut National de la Santé et de la Recherche Médicale, Université de Strasbourg, Ministère de la Recherche, Agence Nationale pour la Recherche Grants ANR-09-MNPS-038 CeCoMod, ANR-2010-JCJC-1403-1 MicroCer, and ANR-GUI-AAP-04 Sensocode; the INTERREG IV Rhin Supérieur Program; and the European Funds for Regional Development, TIGER project. Publication costs are supported by the Neurex network (www.neurex.org).

Notes

This article is a PNAS Direct Submission.

Authors

Affiliations

Joseph Chaumont
Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Unité Propre de Recherche 3212, Université de Strasbourg, 67084 Strasbourg, France;
Nicolas Guyon
Institut de Biologie de l'Ecole Normale Supérieure, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8197, Institut National de la Santé et de la Recherche Médicale, Unité 1024, 75005 Paris, France;
Antoine M. Valera
Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Unité Propre de Recherche 3212, Université de Strasbourg, 67084 Strasbourg, France;
Guillaume P. Dugué
Institut de Biologie de l'Ecole Normale Supérieure, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8197, Institut National de la Santé et de la Recherche Médicale, Unité 1024, 75005 Paris, France;
Daniela Popa
Institut de Biologie de l'Ecole Normale Supérieure, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8197, Institut National de la Santé et de la Recherche Médicale, Unité 1024, 75005 Paris, France;
Paikan Marcaggi
Institut de Biologie de l'Ecole Normale Supérieure, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8197, Institut National de la Santé et de la Recherche Médicale, Unité 1024, 75005 Paris, France;
Vanessa Gautheron
Center for Interdisciplinary Research in Biology, Collège de France, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 7241, Institut National de la Santé et de la Recherche Médicale, Unité 1050, 75005 Paris, France;
Sophie Reibel-Foisset
Chronobiotron, Unité Mixte de Service 3415, Centre National pour la Recherche Scientifique, Université de Strasbourg, 67084 Strasbourg, France;
Stéphane Dieudonné
Institut de Biologie de l'Ecole Normale Supérieure, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8197, Institut National de la Santé et de la Recherche Médicale, Unité 1024, 75005 Paris, France;
Aline Stephan
Institut de Génétique Moléculaire et Cellulaire, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 7104, Institut National de la Santé et de la Recherche Médicale, Unité 964, Université de Strasbourg, 67400 Illkirch, France; and
Michel Barrot
Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Unité Propre de Recherche 3212, Université de Strasbourg, 67084 Strasbourg, France;
Jean-Christophe Cassel
Laboratoire de Neurosciences Cognitives et Adaptatives, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 7364, Université de Strasbourg, 67084 Strasbourg, France
Jean-Luc Dupont
Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Unité Propre de Recherche 3212, Université de Strasbourg, 67084 Strasbourg, France;
Frédéric Doussau
Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Unité Propre de Recherche 3212, Université de Strasbourg, 67084 Strasbourg, France;
Bernard Poulain
Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Unité Propre de Recherche 3212, Université de Strasbourg, 67084 Strasbourg, France;
Fekrije Selimi1
Center for Interdisciplinary Research in Biology, Collège de France, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 7241, Institut National de la Santé et de la Recherche Médicale, Unité 1050, 75005 Paris, France;
Clément Léna1
Institut de Biologie de l'Ecole Normale Supérieure, Centre National de la Recherche Scientifique, Unité Mixte de Recherche 8197, Institut National de la Santé et de la Recherche Médicale, Unité 1024, 75005 Paris, France;
Philippe Isope2,1 [email protected]
Institut des Neurosciences Cellulaires et Intégratives, Centre National de la Recherche Scientifique, Unité Propre de Recherche 3212, Université de Strasbourg, 67084 Strasbourg, France;

Notes

2
To whom correspondence should be addressed. E-mail: [email protected].
Author contributions: F.S., C.L., and P.I. designed research; J.C., N.G., A.M.V., G.P.D., P.M., V.G., S.R.-F., F.D., F.S., and P.I. performed research; S.D., A.S., M.B., J.-C.C., B.P., F.S., and C.L. contributed new reagents/analytic tools; J.C., G.P.D., D.P., J.-L.D., F.S., C.L., and P.I. analyzed data; and J.C., C.L., and P.I. wrote the paper.
1
F.S., C.L., and P.I. contributed equally to this work.

Competing Interests

The authors declare no conflict of interest.

Metrics & Citations

Metrics

Note: The article usage is presented with a three- to four-day delay and will update daily once available. Due to ths delay, usage data will not appear immediately following publication. Citation information is sourced from Crossref Cited-by service.


Citation statements




Altmetrics

Citations

If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download.

Cited by

    Loading...

    View Options

    View options

    PDF format

    Download this article as a PDF file

    DOWNLOAD PDF

    Get Access

    Login options

    Check if you have access through your login credentials or your institution to get full access on this article.

    Personal login Institutional Login

    Recommend to a librarian

    Recommend PNAS to a Librarian

    Purchase options

    Purchase this article to get full access to it.

    Single Article Purchase

    Clusters of cerebellar Purkinje cells control their afferent climbing fiber discharge
    Proceedings of the National Academy of Sciences
    • Vol. 110
    • No. 40
    • pp. 15851-16283

    Media

    Figures

    Tables

    Other

    Share

    Share

    Share article link

    Share on social media