A mechanism for savings in the cerebellum.

The phenomenon of savings (the ability to relearn faster than the first time) is a familiar property of many learning systems. The utility of savings makes its underlying mechanisms of special interest. We used a combination of computer simulations and reversible lesions to investigate mechanisms of savings that operate in the cerebellum during eyelid conditioning, a well characterized form of motor learning. The results suggest that a site of plasticity outside the cerebellar cortex (possibly in the cerebellar nucleus) can be protected from the full consequences of extinction and that the residual plasticity that remains can later contribute to the savings seen during relearning.

Computer simulation of cerebellar information processing.

Although many functions have been ascribed to the cerebellum, the uniformity of its synaptic organization suggests that a single, characteristic computation may be common to all. Computer simulations are useful in examining this cerebellar computation, as they inherently address function at the level of information processing. Progress is facilitated by factors that make the cerebellum particularly amenable to such analysis. We review progress from two contrasting approaches. Top-down simulations begin with hypotheses about computational mechanisms and then ask how such mechanisms might operate within the cerebellum. Bottom-up simulations attempt to build a representation of the cerebellum that reflects known cellular and synaptic components as accurately as possible. We describe recent advances from these two approaches that are leading to an understanding of what information the cerebellum processes and how its neurons and synapses accomplish this task.

Timing mechanisms in the cerebellum: testing predictions of a large-scale computer simulation.

We used large-scale computer simulations of eyelid conditioning to investigate how the cerebellum generates and makes use of temporal information. In the simulations the adaptive timing displayed by conditioned responses is mediated by two factors: (1) different sets of granule cells are active at different times during the conditioned stimulus (CS), and (2) responding is not only amplified at reinforced times but also suppressed at unreinforced times during the CS. These factors predict an unusual pattern of responding after partial removal of the cerebellar cortex that was confirmed using small, electrolytic lesions of cerebellar cortex. These results are consistent with timing mechanisms in the cerebellum that are similar to Pavlov's "inhibition of delay" hypothesis.

Cerebellar function: coordination, learning or timing?

Theories of cerebellar function have largely involved three ideas: movement coordination, motor learning or timing. New evidence indicates these distinctions are not particularly meaningful, as the cerebellum influences movement execution by feedforward use of sensory information via temporally specific learning.

Mechanisms of cerebellar learning suggested by eyelid conditioning.

Classical eyelid conditioning has been used to great advantage in demonstrating that the cerebellum helps to improve movements through experience, and in identifying the underlying mechanisms. Results from recent studies support the hypotheses that learning occurs in both the cerebellar nucleus and cortex, and that these sites make different contributions. Specifically, results indicate that the cerebellar cortex is responsible for temporally specific learning. A combination of experimental and computational studies has been important for arriving at these conclusions, which seem to be applicable to the broad range of movements to which the cerebellum contributes.

Cerebellar cortex lesions prevent acquisition of conditioned eyelid responses.

We have used aspiration and electrolytic lesions to investigate the contributions of cerebellar cortex to the acquisition and expression of conditioned eyelid responses. We show that lesions of the anterior lobe of rabbit cerebellar cortex disrupt the timing of previously learned conditioned eyelid responses. These short-latency responses were used as an indication that the cerebellar cortex was sufficiently damaged and that the underlying pathways necessary for the expression of responses were sufficiently intact to support responses. Rabbits were subsequently trained for 15 daily sessions using a new conditioned stimulus. Whereas rabbits in which lesions had no significant effect on response timing showed rapid acquisition of appropriately timed eyelid responses to the new conditioned stimulus, animals with lesions that disrupt timing showed no significant increases in either amplitude or probability of responses. Histological analysis suggests that damage to the anterior lobe of the cerebellar cortex is necessary and sufficient to abolish timing and prevent acquisition. These data indicate that the cerebellar cortex is necessary for the acquisition of conditioned eyelid responses and are consistent with the hypotheses that (1) eyelid conditioning results in plasticity in both the anterior lobe of the cerebellar cortex and in the anterior interpositus nucleus and (2) induction of plasticity in the interpositus requires intact input from the cerebellar cortex.

Simulations of cerebellar motor learning: computational analysis of plasticity at the mossy fiber to deep nucleus synapse.

We question the widely accepted assumption that a molecular mechanism for long-term expression of synaptic plasticity is sufficient to explain the persistence of memories. Instead, we show that learning and memory require that these cellular mechanisms be correctly integrated within the architecture of the neural circuit. To illustrate this general conclusion, our studies are based on the well characterized synaptic organization of the cerebellum and its relationship to a simple form of motor learning. Using computer simulations of cerebellar-mediated eyelid conditioning, we examine the ability of three forms of plasticity at mossy fiber synapses in the cerebellar nucleus to contribute to learning and memory storage. Results suggest that when the simulation is exposed to reasonable patterns of "background" cerebellar activity, only one of these three rules allows for the retention of memories. When plasticity at the mossy fiber synapse is controlled by nucleus or climbing fiber activity, the circuit is unable to retain memories because of interactions within the network that produce spontaneous drift of synaptic strength. In contrast, a plasticity rule controlled by the activity of the Purkinje cell allows for a memory trace that is resistant to ongoing activity in the circuit. These results suggest specific constraints for theories of cerebellar motor learning and have general implications regarding the mechanisms that may contribute to the persistence of memories.

Inhibitory control of LTP and LTD: stability of synapse strength.

Although much is known about the induction of synaptic plasticity, the persistence of memories suggests the importance of understanding factors that maintain synaptic strength and prevent unwanted synaptic changes. Here we present evidence that recurrent inhibitory connections in the CA1 region of hippocampus may contribute to this task by modulating the relative ability to induce long-term potentiation and depression (LTP and LTD). Bath application of the gamma-aminobutyric acid (GABA) type A agonist muscimol to hippocampal slices increased the range of frequencies that produce LTD, whereas in the presence of the GABA type A antagonist picrotoxin LTD was induced only at very low stimulation frequencies (0.25-0.5 Hz). Because one source of GABAergic input to CA1 pyramidal cells is via recurrent inhibition, we tested the prediction that elevated postsynaptic spike activity would increase feedback GABA inhibition and favor the induction of LTD. By using an induction stimulation of 8 Hz, which alone produced no net change in synaptic strength, we found that stimulation presented during antidromic activation of pyramidal cell spikes induced LTD. This effect was blocked by picrotoxin. The influence of recurrent inhibition on LTP and LTD displays properties that may decrease the potential for self-reinforcing, runaway changes in synapse strength. A mechanism of this sort may help maintain patterns of synaptic strengths despite the ongoing opportunities for plasticity produced by synapse activation.

Pharmacological analysis of cerebellar contributions to the timing and expression of conditioned eyelid responses.

Contradictory results have been reported regarding the effects of cerebellar cortex lesions on the expression of conditioned eyelid responses--either no effect, partial to complete abolition of responses, or disruption of response timing. This uncertainty is increased by debates regarding the region(s) of cerebellar cortex that are involved, by the likelihood that cortex lesions can inadvertently include damage to the interpositus nucleus or other pathways necessary for response expression, and by potential confounds from the degeneration of climbing fibers produced by cerebellar cortex lesions. We have addressed these issues by reversibly blocking cerebellar cortex output via infusion of the GABA antagonist picrotoxin into the interpositus nucleus. After picrotoxin infusion, conditioned responses are spared but their timing is disrupted and their amplitude diminished. In the same animals, conditioned responses were abolished by infusion of the GABA agonist muscimol and were unaffected by infusion of saline vehicle. These results are consistent with the hypothesis that (i) plasticity in the interpositus nucleus contributes to the expression of conditioned responses, as suggested by the responses seen with the cortex disconnected, and (ii) plasticity in the cerebellar cortex also contributes to conditioned response expression, as suggested by disruption of response timing.

A mathematical model of the cerebellar-olivary system I: self-regulating equilibrium of climbing fiber activity.

We use a mathematical model to investigate how climbing fiber-dependent plasticity at granule cell to Purkinje cell (gr-->Pkj) synapses in the cerebellar cortex is influenced by the synaptic organization of the cerebellar-olivary system. Based on empirical studies, gr-->Pkj synapses are assumed to decrease in strength when active during a climbing fiber input (LTD) and increase in strength when active without a climbing fiber input (LTP). Results suggest that the inhibition of climbing fibers by cerebellar output combines with LTD/P to self-regulate spontaneous climbing fiber activity to an equilibrium level at which LTP and LTD balance and the expected net change in gr-->Pkj synaptic weights is zero. The synaptic weight vector is asymptotically confined to an equilibrium hyperplane defining the set of all possible combinations of synaptic weights consistent with climbing fiber equilibrium. Results also suggest restrictions on LTP/D at gr-->Pkj synapses required to produce synaptic weights that do not drift spontaneously.

A mathematical model of the cerebellar-olivary system II: motor adaptation through systematic disruption of climbing fiber equilibrium.

The implications for motor learning of the model developed in the previous article are analyzed using idealized Pavlovian eyelid conditioning trials, a simple example of cerebellar motor learning. Results suggest that changes in gr-->Pkj synapses produced by a training trial disrupt equilibrium and lead to subsequent changes in the opposite direction that restore equilibrium. We show that these opposing phases would make the net plasticity at each gr-->Pkj synapse proportional to the change in its activity during the training trial, as influenced by a factor that precludes plasticity when changes in activity are inconsistent. This yields an expression for the component of granule cell activity that supports learning, the across-trials consistency vector, the square of which determines the expected rate of learning. These results suggest that the equilibrium maintained by the cerebellar-olivary system must be disrupted in a specific and systematic manner to promote cerebellar-mediated motor learning.

A model of Pavlovian eyelid conditioning based on the synaptic organization of the cerebellum.

We present a model based on the synaptic and cellular organization of the cerebellum to derive a diverse range of phenomena observed in Pavlovian eyelid conditioning. These phenomena are addressed in terms of critical pathways and network properties, as well as the sites and rules for synaptic plasticity. The theory is based on four primary hypotheses: (1) Two cerebellar sites of plasticity are involved in conditioning: (a) bidirectional long-term depression/potentiation at granule cell synapses onto Purkinje cells (gr-->Pkj) in the cerebellar cortex and (b) bidirectional plasticity in the interpositus nucleus that is controlled by inhibitory inputs from Purkinje cells; (2) climbing fiber activity is regulated to an equilibrium level at which the net strength of gr-->Pkj synapses remains constant unless an unexpected unconditioned stimulus (US) is presented or an expected US is omitted; (3) a time-varying representation of the conditioned stimulus (CS) in the cerebellar cortex permits the temporal discrimination required for conditioned response timing; and (4) the ability of a particular segment of the CS to be represented consistently across trials varies as a function of time since CS onset. This variation in across-trials consistency is thought to contribute to the ISI function. The model suggests several empirically testable predictions, some of which have been tested recently.

The cerebellum: a neuronal learning machine?

Comparison of two seemingly quite different behaviors yields a surprisingly consistent picture of the role of the cerebellum in motor learning. Behavioral and physiological data about classical conditioning of the eyelid response and motor learning in the vestibulo-ocular reflex suggests that (i) plasticity is distributed between the cerebellar cortex and the deep cerebellar nuclei; (ii) the cerebellar cortex plays a special role in learning the timing of movement; and (iii) the cerebellar cortex guides learning in the deep nuclei, which may allow learning to be transferred from the cortex to the deep nuclei. Because many of the similarities in the data from the two systems typify general features of cerebellar organization, the cerebellar mechanisms of learning in these two systems may represent principles that apply to many motor systems.

LTP induced by activation of voltage-dependent Ca2+ channels requires protein kinase activity.

We have examined the requirement for protein kinase activity in long-term potentiation (LTP) induced by activation of voltage-dependent Ca2+ channels (VDCCs) in hippocampal slices. We previously demonstrated that LTP induced by application of the K+ channel blocker tetraethylammonium (TEA-LTP) consisted of two distinct components, an NMDA receptor-dependent component and a VDCC-dependent component. The results herein demonstrate that both the NMDA and VDCC-dependent components of TEA-LTP are blocked by K-252a, a broad spectrum protein kinase inhibitor. Furthermore, VDCC-dependent TEA-LTP is attenuated by KN-62, a specific inhibitor of Ca2+/calmodulin dependent protein kinase II (CaM-KII). These results demonstrate that LTP induced by VDCC activation requires protein kinase activity and suggest that different routes of postsynaptic Ca2+ influx activate protein kinases to trigger the induction of LTP but that these enzyme systems may be contained in different cell compartments.

Extinction of conditioned eyelid responses requires the anterior lobe of cerebellar cortex.

We test the hypothesis that the cerebellar cortex is required for the extinction of conditioned eyelid responses in rabbits trained using standard Pavlovian delay procedures. Following 10 daily training sessions during which rabbits achieved asymptotic performance, lesions of the ipsilateral hemisphere of the cerebellar cortex were made by aspiration. The target of these lesions was the anterior lobe, as suggested by previous observations that this region is necessary for the learning-dependent timing of conditioned eyelid responses (Perrett et al., 1993). We report that anterior lobe damage, as indicated by disrupted response timing and confirmed by tissue analysis, produces severe deficits in conditioned response extinction. Postlesion responses show no significant decline over ten training sessions, whereas response timing and extinction are unaffected by lesions that do not include the anterior lobe. These conditioned responses that do not extinguish display stimulus specificity, excluding the possibility that they are unlearned responses unmasked by cerebellar cortex lesions. These observations suggest that Pavlovian eyelid conditioning is mediated by synaptic plasticity in at least two sites and the anterior lobe of the cerebellar cortex influences one of these sites during extinction. Based on these and previous data, we propose the hypothesis that eyelid conditioning can involve plasticity in both the cerebellar cortex and interpositus nucleus and that plasticity in the nucleus is controlled by input from Purkinje cell activity in the cortex. This hypothesis is consistent with observations that the cerebellar cortex may not always be required for the expression of conditioned responses, but it is necessary for response timing and for extinction.