Publications
The cerebellum: cortical processing and theory
CURRENT OPINION IN NEUROBIOLOGY
6: 759-764, 1996
Copyright © Current Biology Ltd, 1996
E. De Schutter, R. Maex
Born-Bunge Foundation, University of Antwerp - UIA, B2610 Antwerp, Belgium
Abstract
Several advances over the past year have made a complete reevaluation of the function of the cerebellum and of the role of cerebellar synaptic plasticity necessary. These include the discovery of parallel fiber-induced long term depression, the lack of effect on motor coordination of the absence of cerebellar long term depression in knockout mice and the strong activation of the cerebellar nuclei during sensory tasks.
Introduction
These are exciting times for cerebellar research. The field has been dominated for several decades by two dogmas. The first one is that the cerebellum is exclusively involved in motor control. The second dogma, the theory of cerebellar motor learning first proposed by Marr [1] and Albus [2] and subsequently elaborated by Ito [3] and others, is more controversial. Standard textbooks may give the impression that this theory has been confirmed by classical conditioning experiments [4] and hence became generally accepted, but large parts of the cerebellar research community have never believed the cerebellum to be the site of memory storage for motor learning [5-6, **7].
The Marr-Albus-Ito theory proposes that only a subset of the more than 150,000 parallel fiber (PF) synapses [8] contacting any particular Purkinje cell (PC) controls its output and ultimately the motor system. This would be accomplished by weakening the strength of the PF synapses activated during an erroneous motor command. The error signal leading to such weakening was proposed to be the climbing fiber (CF) input causing a complex spike in the PC. In other words, conjunctive activation of CF and PF synapses should lead to a reduced strength of the PF synapse, a process called long-term depression (LTD) [9].
Both dogmas are now being challenged by recent experimental findings.
LTD and the timing of the CF input
The Marr-Albus-Ito theory critically depends on the conjunctive nature of CF-induced LTD (CF-LTD). The CF input activates dendritic and somatic calcium channels leading to calcium influx [10, 11] which induces LTD of the activated PF synapses [9]. Many groups have demonstrated that an increased dendritic calcium concentration is sufficient to induce LTD; the CF input itself is not necessary (reviewed in [**7]). But if CF input is used, LTD is induced in cerebellar slices solely if the CF input precedes or coincides with the PF input [12]. Only the CF->PF sequence of activation leads to an increased dendritic calcium concentration during the PF input, but it is opposite to what would be expected if the CF signals an error in motor performance linked to the PF-activation. Neither does it conform to the order in classical conditioning experiments, where the conditioned stimulus (PF input) precedes the unconditioned stimulus (CF input) [4, 12].
It has been known for quite some time that a CF->PF sequence is needed for LTD-induction in slice [**7], which has led some theoreticians to propose unsubstantiated alternative induction schemes [13]. Many researchers, however, seem to find the results from classical conditioning experiments [4] more convincing than the slice experiments [**7, 12]. This is exemplified in a model of the biochemical pathways responsible for the induction of LTD in PCs [14], which depends on release of calcium from inositol triphosphate (IP3) sensitive stores [15] to learn particular time intervals between the PF and the CF input [14, 16]. This supposedly realistic model shows LTD induction for the PF-CF sequence [14] which is ineffective in slice [12]. Moreover it is unclear how robust the model would be to noisy input.
Figure 1
Schematic representation of the pathways thought to be involved in the induction of three forms of
LTD of PF synapses in PCs. Each pathways is coded by shadow (CF-LTD), italic (PF-LTD) or bold
(NO-LTD) print. Note that these pathways may overlap to a large degree, e.g. CF-LTD requires
activation of the AMPA and mGluR1 receptors of the PF synapse [9]. There are contradictory reports
on whether the same holds true for NO-LTD [**21, *24], but it is likely that PKG activity alone
(without PKC) is insufficient [25]. Activation of the AMPA receptor of the PF synapse is needed to
induce CF-LTD (and PF-LTD), but the exact pathways involved are unknown (Na+ influx
through the receptor might be important [9]). It has not been demonstrated that mGluR1 activation
is necessary for induction of PF-LTD, but this seems likely. Finally, it is quite possible that
all these pathways are activated together to variable degrees during experiments which use
conjunctive CF-PF stimuli to induce LTD. Abbreviations (see also text): DAG = 1,2 diacylglycerol;
PKC = protein kinase C; PKG = protein kinase G.
Three forms of cerebellar LTD
One of the questions with CF-LTD was how a CF input on the smooth dendrite can raise the calcium concentration close to the PF synapses located on the spiny dendrite. This was solved when calcium imaging experiments demonstrated that the complex spike causes activation of voltage-gated calcium channels everywhere in the dendritic tree [10], including the spiny dendrite. Computer modeling showed that localized PF inputs should also activate calcium channels [17] because of the high density of these channels in the spiny dendrite. This prediction was recently confirmed by two groups [**18, *19]. Eilers et al. [**18] used confocal laser scanning microscopy to show that small focal PF inputs lead to transient localized increases in the dendritic calcium concentration, proportional to the size of the PF stimulus. Another study using two-photon excitation imaging demonstrated that even a single PF input activates calcium channels located on the dendritic spine [*19]. Additionally, a subset of PF synapses was found where calcium influx through the AMPA receptor channel might be a second source of increased calcium [*19]. More quantitative studies suggest, however, that the calcium fraction of the total synaptic current through AMPA receptor channels is less than 1% in PCs [20].
Whatever the contribution of calcium influx through synaptic channels, focal PF activation can lead to local calcium influx. As was pointed out by us [**7], this PF-induced calcium inflow fulfills the biophysical requirements for the induction of LTD of the activated synapses (PF-LTD). The existence of PF-LTD has recently been confirmed experimentally [**21]. PF stimulation at strengths sufficient to cause calcium inflow similar to that demonstrated by Eilers et al. [**18] induced a LTD of the activated PF synapses which largely occluded further depression by CF-LTD [**21]. Interestingly, under the same experimental conditions a second form of LTD was induced. This nitric oxide-dependent LTD (NO-LTD) affected PF synapses presumably not activated during the stimulus presentation. The role of NO in cerebellar LTD has been controversial for some time [9, 22] as NO synthase is notably absent in PCs and in the CFs which were assumed to release it [23]. Neuronal NO synthase is, however, present in PFs [23] and one study claims that NO application can replace the PF input in LTD induction [*24]. There remain important issues to be resolved, like whether the NO application needs to be combined with an increased postsynaptic calcium concentration [*24] or not [**21, 25], if activated PFs always release NO and how NO-LTD interacts with CF-LTD [25].
As summarized in Fig. 1, we may now distinguish three forms of LTD at the PF synapse. CF-LTD, the `Marr-Albus-Ito form', is conjunctive and specific to the activated PF synapses. PF-LTD is non-conjunctive but also specific; while NO-LTD is maybe conjunctive but also diffuse, affecting distant non-activated PF synapses. Moreover, PF-LTD is more easibly induced in cerebellar slices than CF-LTD (Hartell, personal communication). In our opinion, the existence of PF-LTD and NO-LTD is an even more serious objection to the Marr-Albus-Ito theory of cerebellar learning than the CF->PF sequence needed for CF-LTD. It is indeed difficult to imagine how a non-conjunctive change in synaptic plasticity can co-exist with learning through conjunctive CF-LTD, as the PF-LTD would occlude learned patterns [**21].
If cerebellar LTD is not a mechanism for motor learning, what is then its function? We have recently proposed that LTD of the PF synapse, complemented by the potentiation of inhibitory synapses on PCs [26], might be important in normalizing the total excitatory input onto a single PC [**7]. The typical circuitry of the cerebellum does not provide for a feedback mechanism to regulate the balance between excitation and inhibition onto PCs. It seems that such a regulation would be essential in a cell which receives more than 150,000 excitatory inputs [8] onto a highly excitable dendrite [**18, *19]. PF-LTD [**21] can provide such an autoregulatory mechanism by slowly reducing the strength of PF synapses when excessive excitation of a PC dendrite leads to calcium influx [**7].
Figure 2
A comparison between (A) the predictions made by the Marr-Albus-Ito theory of cerebellar
motor learning and (B) the performance of knockout mice on tests of motor coordination and
conditioning of the eyeblink reflex. The abbreviations in (B) refer to the mutant gene (see
text) and the causal effects the specific knockout mice establish or disprove. Important
differences between (A) and (B) are the strong dependence of motor function on
single CF inputs, the weak link between LTD and classical conditioning and the absent connection
between cerebellar LTD and motor coordination (GFAP mutant). It is unclear if the persistence of
multiple CF inputs changes the properties of cerebellar LTD.
Knockout of classical conditioning
Thompson and others have proposed that classical conditioning of the eyeblink reflex is experimental proof that the cerebellum is the memory site of motor learning [4, 16, 27] This work, which is largely based on lesion studies [27], is very controversial [5, 6, 28]. Available space does not allow us to reiterate these criticisms, but recent results from gene targeting experiments shed new light on this issue. The interpretation of motor performance and behavior in knockout mice is, however, complicated by the use of embryonic stem cells derived from a neurologically impaired strain of mice [*29]. Therefore we will mainly compare different knockout experiments instead of considering the function of specific genes.
Over the last two years several mutants with impaired cerebellar LTD have been described. These include knockouts of genes encoding for metabotropic glutamate receptor type 1 (mGluR1) [30, 31], for the glutamate receptor d subunit (GluRd2) [32] and for glial fibrillary acidic protein (GFAP) [33]. Surprisingly, knockouts of protein kinase C gamma (PKC gamma), an enzyme necessary for the induction of LTD [9], did not show reduced LTD [**34]. More important is that three of these mutant strains (mGluR1, GluRd2 and PKC gamma) showed persistent multiple CF innervation of PCs [35]. In normal rodents PCs are innervated by three to four CFs at postnatal day five. The supernumerary CFs are eliminated during the second and third postnatal weeks [36], resulting in a one-to-one relation between CFs and PCs at the age when rodents show adult motor behavior. These three mutants showed severe deficits in motor coordination [30-32, **34], while the GFAP mutant with normal CF innervation did not [33]. The mutants with impaired cerebellar LTD showed reduced learning of the eyeblink reflex [30, 33], but these deficits are only partial and it remains to be proven [**7] that they are caused solely by the absence [30] or decrease [33] of LTD. In this context the PKC gamma mutant may be more illuminating. It showed normal cerebellar LTD and normal, even facilitated, eyeblink conditioning, but severely impaired motor coordination [**34]. These results, which are summarized in Fig. 2, suggest that the one-to-one relation between CFs and PCs is much more important than LTD for the normal motor function of the cerebellum. The GFAP and PKC gamma mutants suggest that LTD deficits and impaired eyeblink conditioning have very little impact on the motor behavior of the animal.
The mysterious function of the CF input
The results just described demonstrate the importance of the single CF input to PCs. But if this input is not the teacher in motor learning as suggested by Marr [1] and Albus [2], what is then its function [37]? Llinás and co-workers have proposed that the CF is a clock which activates the proper PC assemblies at the right time [38, *39, 40]. Using multiple-microelectrode recordings of PCs they showed that the spatial and temporal organization of CF inputs is rhythmic and time-locked to motor activity [*39]. Synchronous CF inputs show complex spatial patterns which change during sequences of movements. The authors conclude that these spatiotemporal activation patterns allow the control of individual muscles. It remains to be demonstrated, however, that CF activations in this area of the cerebellum are not just a reflection of sensory input instead of motor command signals [*41]. It looks as if some of their results may be species specific, as rhythmic activation of CF inputs was not found in monkeys [42]. Similarly, this group has reported that the conduction time in rat CFs is independent of their length [38] supporting the clock-like synchronous arrival of CF signals everywhere in the cortex [40], but this result was not confirmed in cats [43].
Brain mapping of new cerebellar functions
If it is unclear how the cerebellum works, do we at least understand its function? Recent human brain mapping results suggest we do not, as the cerebellum is activated during many cognitive and language tasks without overt motor components [44], though some argue that some subconscious contribution of the motor system must be involved [45]. A recent functional MRI study [**46] increases the challenge to any theory claiming an exclusive motor function for the cerebellum [45]. This study, which was specifically designed to test Bower's hypothesis that the cerebellum controls sensory acquisition instead of motor function [*41], confirms and extends older PET-studies [47]. Dentate nucleus activity was increased whenever the subject had to make either a passive or active tactile discrimination [**46]. Conversely, the lowest activity was present during a pure motor task. Bower interprets this as evidence that the cerebellum tunes sensory systems, sometimes through activation of the motor system, to improve perception [*41]. Similarly, increased activation of the cerebellar vermis during explicit learning of a visuomotor task could be interpreted as an optimization of visual sensory acquisition [48-49]. Interestingly, suppression of (badly tuned ?) sensory input may reduce tremors in cerebellar patients [50].
Conclusion
In our opinion, the implications of recent findings on the properties of cerebellar LTD [**21] and the lack of correlation between deficient eyeblink conditioning and motor dysfunction [33] are too fundamental to retain the Marr-Albus-Ito theory. Attempts to redefine the theory of cerebellar motor learning [13, 51] fail to account for these findings and lack the clarity and detail of the original theories. New theories are needed, which may be found more easily by a bottom-up approach where first the physiological properties of cerebellar neurons are elucidated. Our detailed computer modeling of the PC [17, 52], for example, has led us to propose a new theory on the role of cerebellar synaptic plasticity [**7].
Such an approach will prevent the use of inappropriate metaphores from computer technology [44, 51] when a new unifying theoretical framework is formulated to explain the cerebellar contribution to motor [6, 53], sensory [**46], cognitive [44] and even autonomic [54] functions. But this will require more refined experimental studies to provide a better understanding of how, for example, motor control is executed by the brain [55].
Acknowledgements
Our work is supported by the University of Antwerp, the National Fund for Scientific Research (Belgium) and by NIMH RO1-MH52903.
References
1. Marr DA: A theory of cerebellar cortex. J Physiol (London) 1969, 202: 437-470.
2. Albus JS: A theory of cerebellar function. Math Biosci 1971, 10: 25-61.
3. Ito M: The cerebellum and neural control. New York: Raven Press; 1984.
4. Thompson RF, Krupa DJ: Organization of memory traces in the mammalian brain. Annu Rev Neurosci 1994, 17: 519-549.
5. Welsh JP, Harvey JA: The role of the cerebellum in voluntary and reflexive movements. In The cerebellum revisited. Edited by RR Llinás, C Sotelo. New York: Springer-Verlag; 1992: 301-334.
6. Bloedel JR, Bracha V: On the cerebellum, cutaneomuscular reflexes, movement control and the elusive engrams of memory. Behav Brain Res 1995, 68: 1-44.
**7. De Schutter E: Cerebellar long-term depression might normalize excitation of Purkinje cells: a hypothesis. Trends Neurosci 1995, 18: 291-295.
This paper reviews the experimental literature contradicting the Marr-Albus-Ito theory and presents a new theory on the role of cerebellar LTD. It is proposed that under physiological conditions LTD is autoinduced by PF inputs. LTD and other forms of Purkinje cell synaptic plasticity are part of a local negative feedback loop that prevents overstimulation of PCs by PF inputs, but they are not involved in any behavioral learning. The theory explains why it is difficult to induce LTD when normal inhibition is present and why inhibitory inputs are potentiated by the same calcium influx which induces LTD.
8. Harvey RJ, Napper RMA: Quantitative studies of the mammalian cerebellum. Prog Neurobiol 1991, 36: 437-463.
9. Linden DJ, Connor JA: Long-term synaptic depression. Annu Rev Neurosci 1995, 18: 319-357.
10. Miyakawa H, Lev-Ram V, Lasser-Ross N, Ross WN: Calcium transients evoked by climbing fiber synaptic inputs in guinea pig cerebellar Purkinje neurons. J Neurophysiol 1992, 68: 1178-1189.
11. Eilers J, Callewaert G, Armstrong CM, Konnerth A: Calcium signaling in a narrow somatic submembrane shell during synaptic activity in cerebellar neurons. Proc Natl Acad Sci USA 1995, 92: 10272-10276.
12. Karachot L, Kado RT, Ito M: Stimulus parameters for induction of long-term depression in in vitro rat Purkinje cells. Neurosci Res 1994, 21: 161-168.
13. Houk JC, Buckingham JT, Barto AG: Models of the cerebellum and motor learning. Behav Brain Sci 1996, 19: 368-383.
14. Fiala JC, Grossberg S, Bullock D: Metabotropic glutamate receptor activation in cerebellar Purkinje cells as substrate for adaptive timing of the classically conditioned eye-blink response. J Neurosci 1996, 16: 3760-3774.
15. Kasono K, Hirano T: Involvement of inositol trisphosphate in cerebellar long-term depression. NeuroReport 1995, 6: 569-572.
16. Perrett SP, Mauk MD: Extinction of conditioned eyelid responses requires the anterior lobe of cerebellar cortex. J Neurosci 1995, 15: 2074-2080.
17. 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 USA 1994, 91: 4736-4740.
**18. Eilers J, Augustine GJ, Konnerth A: Subthreshold synaptic Ca2+ signaling in fine dendrites and spines of cerebellar Purkinje neurons. Nature 1995, 373: 155-158.
Subthreshold activation of PF synapses leads to voltage-gated calcium influx in the spiny dendrite. The authors claim that the calcium influx stays restricted to a dendritic compartment, consisting of spines and adjacent dendritic shaft, but demonstrate this only for somatic membrane potentials of -60 mV or lower.
*19. Denk W, Sugimori M, Llinás RR: 2 types of calcium response limited to single spines in cerebellar Purkinje cells. Proc Natl Acad Sci USA 1995, 92: 8279-8282.
This study extends the preceding one by demonstrating that a presumably single PF input can activate calcium channels on a single spine. Dendritic changes in calcium concentration requires co-activation of multiple PF inputs. The second calcium response, influx through the AMPA receptor channel, is more controversial.
20. Tempia F, Kano M, Schneggenburger R, Schirra C, Garaschuk O, Plant T, Konnerth A: Fractional calcium current through neuronal AMPA-receptor channels with a low calcium permeability. J Neurosci 1996, 16: 456-466.
**21. 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.
Strong PF input can induce a form of long-term depression of the activated synapses which is independent of NO and which largely occludes LTD induced by co-activation of CF and PF inputs. At the same time a NO-dependent LTD of distant PF synapses is induced.
22. Li J, Smith SS, McElligott JG: Cerebellar nitric oxide is necessary for vestibulo-ocular reflex adaptation, a sensorimotor model of learning. J Neurophysiol 1995, 74: 489-494.
23. Vincent SR: Nitric oxide and synaptic plasticity: no news from the cerebellum. Behav Brain Sci 1996, 19: 362-367.
*24. Lev-Ram V, Makings LR, Keitz PF, Kao JPY, Tsien RW: Long-term depression in cerebellar Purkinje neurons results from coincidence of nitric oxide and depolarization-induced Ca2+ transients. Neuron 1995, 15: 407-415.
LTD of PF synapses was induced by combining caged NO release with PC depolarization, occluding further LTD by PF activation. This study suggests that the PF synaptic ending is the normal source of NO, which fits anatomical data better than the conventional view of release by CF.
25. Hartell NA: Inhibition of cGMP breakdown promotes the induction of cerebellar long-term depression. J Neurosci 1996, 16: 2881-2890.
26. Kano M: Long-lasting potentiation of GABAergic inhibitory synaptic transmission in cerebellar Purkinje cells: its properties and possible mechanisms. Behav Brain Sci 1996, 19: 354-361.
27. Chen L, Bao SW, Lockard JM, Kim JJ, Thompson RF: Impaired classical eyeblink conditioning in cerebellar-lesioned and Purkinje cell degeneration (pcd) mutant mice. J Neurosci 1996, 16: 2829-2838.
28. Logan CG, Grafton ST: Functional anatomy of human eyeblink conditioning determined with regional cerebral glucose metabolism and positron-emission tomography. Proc Natl Acad Sci USA 1995, 92: 7500-7504.
*29. Gerlai R: Gene-targeting studies of mammalian behavior: is it the mutation or the background genotype? Trends Neurosci 1996, 19: 177-181.
The relevance of gene-targetting techniques as they are currently employed in studying the behavioral impact of single gene mutations is questioned. The embryonic stem cells usually originate from a strain of mice with serious neurological deficits.
30. 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.
31. Conquet F, Bashir ZI, Davies CH, Daniel H, Ferraguti F, Bordi F, Franz-Bacon K, Reggiani A, Matarese V, Condé F, Collingridge GL, F. C: Motor deficit and impairment of synaptic plasticity in mice lacking mGluR1. Nature 1994, 372: 237-243.
32. Kashiwabuchi N, Ikeda K, Araki K, Hirano T, Shibuki K, Takayama C, Inoue T, Kutsuwada T, Yagi T, Kang Y, Aizawa S, Mishina M: Impairment of motor coordination, Purkinje cell synapse formation, and cerebellar long-term depression in gluRd2 mutant mice. Cell 1995, 81: 245-252.
33. Shibuki K, Gomi H, Chen L, Bao SW, Kim JSK, Wakatsuki H, Fujisaki T, Fujimoto J, Katoh A, Ikeda T, Chen C, Thompson RF, Itohara S: Deficient cerebellar long-term depression, impaired eyeblink conditioning and normal motor coordination in GFAP mutant mice. Neuron 1996, 16: 587-599.
**34. Chen C, Kano M, Abeliovich A, Chen L, Bao SW, Kim JJ, Hashimoto K, Thompson RF, Tonegawa S: Impaired motor coordination correlates with persistent multiple climbing fiber innervation in PKC gamma mutant mice. Cell 1995, 83: 1233-1242.
Persistent multiple CF innervation of PCs in mutant mice with normal cerebellar LTD and intact eyeblink conditioning leads to severe motor coordination deficits. These results are compared to other gene-targetting studies which affected cerebellar LTD.
35. Kano M, Hashimoto K, Chen C, Abeliovich A, Aiba A, Kurihara K, Watanabe M, Inoue Y, Tonegawa S: Impaired synapse elimination during cerebellar development in PKC gamma mutant mice. Cell 1995, 83: 1223-1231.
36. Crépel F: Regression of functional synapses in the immature mammalian cerebellum. Trends Neurosci 1982, 5: 266-269.
37. Simpson JI, Wylie DR, De Zeeuw CI: On climbing fiber signals and their consequences. Behav Brain Sci 1996, 19: 384-398.
38. Sugihara I, Lang EJ, Llinás R: Uniform olivocerebellar conduction time underlies Purkinje cell complex spike synchronicity in the rat cerebellum. J Physiol (London) 1993, 470: 243-271.
*39. Welsh JP, Lang EJ, Sugihara I, Llinás R: Dynamic organization of motor control within the olivocerebellar system. Nature 1995, 374: 453-457.
Multi-microelectrode recordings in the rat cerebellar hemisphere are used to investigate spatio-temporal activation of CF inputs during a trained movement (tongue protrusion). The rhythmic, synchronous patterns are interpreted to be motor commands controlling muscle activation.
40. De Zeeuw CI, Lang EJ, Sugihara I, Ruigrok TJH, Eisenman LM, Mugnaini E, Llinás R: Morphological correlates of bilateral synchrony in the rat cerebellar cortex. J Neurosci 1996, 16: 3412-3426.
*41. Bower JM: Is the cerebellum sensory for motor's sake or motor for sensory's sake: The view from whiskers of a rat? Prog Brain Res 1996, in press.
It is argued that the cerebellum primarily controls the quality of sensory input. An extensive review presents refreshing reinterpretations of established experimental data. It is argued, for example, that the vestibulo-ocular reflex is primarily sensory, improving the quality of vision, but needs to use the motor system to achieve its goal.
42. Keating JG, Thach WT: Nonclock behavior of inferior olive neurons: interspike interval of Purkinje cell complex spike discharge in the awake behaving monkey is random. J Neurophysiol 1995, 73: 1329-1340.
43. Aggelopoulos NC, Duke C, Edgley SA: Nonuniform conduction time in the olivocerebellar pathway in the anesthetized cat. J Physiol (London) 1995, 486: 763-768.
44. Leiner HC, Leiner AL, Dow RS: The underestimated cerebellum. Hum Brain Mapp 1995, 2: 244-254.
45. Thach WT: On the specific role of the cerebellum in motor learning and cognition: clues from PET activation and lesion studies in man. Behav Brain Sci 1996, 19: 411-431.
**46. Gao JH, Parsons LM, Bower JM, Xiong JH, Li JQ, Fox PT: Cerebellum implicated in sensory acquisition and discrimination rather than motor control. Science 1996, 272: 545-547.
Functional magnetic resonance imaging shows that cerebellar dentate nucleus activation is increased whenever a passive (hands and fingers immobilized) or active (using the fingers to feel the shape of balls) sensory task requires discrimination (left-to-right comparison). A pure motor task (picking up balls) results in the smallest activation.
47. Seitz RJ, Roland PE, Bohm C, Greitz T, Stone-Elander S: Somatosensory discrimination of shape: tactile exploration and cerebral activation. Eur J Neurosci 1991, 3: 481-492.
48. Rauch SL, Savage CR, Brown HD, Curran T, Alpert NM, Kendrick A, Fischman AJ, Kosslyn SM: A PET investigation of implicit and explicit sequence learning. Hum Brain Mapp 1995, 3: 271-286.
49. Doyon J, Owen AM, Petrides M, Sziklas V, Evans AC: Functional anatomy of visuomotor skill learning in human subjects examined with positron emission tomography. Eur J Neurosci 1996, 8: 637-648.
50. Dash BMS: Role of peripheral inputs in cerebellar tremor. Movem Dis 1995, 10: 622-629.
51. Raymond JL, Lisberger SG, Mauk MD: The cerebellum: a neuronal learning machine? Science 1996, 272: 1126-1131.
52. De Schutter E, Bower JM: An active membrane model of the cerebellar Purkinje cell. I. Simulation of current clamps in slice. J Neurophysiol 1994, 71: 375-400.
53. Bastian AJ, Thach WT: Cerebellar outflow lesions: a comparison of movement deficits resulting from lesions at the levels of the cerebellum and thalamus. Ann Neurol 1995, 38: 881-892.
54. Ghelarducci B, Sebastiani L: Contribution of the cerebellar vermis to cardiovascular control. J Auton Nerv Syst 1996, 56: 149-156.
55. Gomi H, Kawato M: Equilibrium-point control hypothesis examined by measured arm stiffness during multijoint movement. Science 1996, 272: 117-120.