The human cerebellum and associative learning: dissociation between the acquisition, retention and extinction of conditioned eyeblinks.

The present paper is part of a systematic exploration of the neural substrates of conditioned eyeblink responses in humans. Normal subjects and patients with lesions restricted to the cerebellum were examined for their ability to acquire new classically conditioned eyeblinks to an auditory conditioned stimulus and whether they were able to perform and extinguish a previously learned natural anticipatory eyeblink response - the kinesthetic threat eyeblink response (KTER). In classical conditioning to an auditory conditioned stimulus, cerebellar patients failed to acquire new conditioned responses. In contrast to this impairment, in the KTER task both cerebellar patients and control subjects exhibited a high incidence of anticipatory eyeblinks which were initiated before the forehead tap. These results indicate that the cerebellar circuits, which are critical for the acquisition of new conditioned responses, are not essential for the storage and expression of naturally acquired conditioned responses. In the extinction experiment, cerebellar patients failed to extinguish their KTERs. This finding suggests that in humans, the acquisition of new and the extinction of previously learned conditioned responses depends on a similar set of cerebellar circuits.

The kinesthetic threat eyeblink: a new type of anticipatory eyeblink response.

The present paper is part of a systematic exploration of naturally acquired conditioned eyeblink responses in human subjects. Normal human subjects were examined for the presence of anticipatory eyeblinks in a new paradigm. They were instructed to move their hand quickly toward their face and tap their forehead. In this situation, subjects generated anticipatory eyeblinks which were initiated before the forehead tap. Additional experiments revealed that visual stimuli and internal movement-planning cues are not required for the initiation of this response. The kinesthetic information from the passively moving arm, however, was sufficient to trigger this kinesthetic threat eyeblink response (KTER). The KTER extinguished when the forehead tap did not reinforce it. These data indicate that the KTER is a unique type of naturally acquired conditioned response system which is maintained by aversive reinforcing events.

Inactivation of interposed nuclei in the cat: classically conditioned withdrawal reflexes, voluntary limb movements and the action primitive hypothesis.

The cerebellar interposed nuclei are considered critical components of circuits controlling the classical conditioning of eyeblink responses in several mammalian species. The main purpose of the present experiments was to examine whether the interposed nuclei are also involved in the control of classically conditioned withdrawal responses in other skeletomuscular effector systems. To achieve this objective, a unique learning paradigm was developed to examine classically conditioned withdrawal responses in three effector systems (the eyelid, forelimb and hindlimb) in individual cats. Trained animals were injected with muscimol in the cerebellar interposed nuclei, and the effects on the three conditioned responses (CRs) were examined. Although the effects of muscimol were less dramatic than previously reported in the rabbit eyeblink preparation, the inactivation of the cerebellar nuclei affected the performance of CRs in all three effector systems. In additional experiments, animals were injected with muscimol at the sites affecting classically conditioned withdrawal responses to determine the effects of these injections on reaching and locomotion behaviors. These tests demonstrated that the same regions of the cerebellar interposed nuclei which control withdrawal reflexes are also involved in the control of limb flexion and precision placement of the paw during both locomotion and reaching tasks. The obtained data indicate that the interposed nuclei are involved in the control of ipsilateral action primitives and that inactivating the interposed nuclei affects several modes of action of these functional units.

Microinjections of anisomycin into the intermediate cerebellum during learning affect the acquisition of classically conditioned responses in the rabbit.

The purpose of this study was to examine the effects of protein synthesis inhibition in the intermediate cerebellum on the acquisition and expression of classically conditioned nictitating membrane responses in the rabbit. Animals were conditioned for three days in a standard delay paradigm. Before each training session, either a solution of anisomycin (a protein synthesis inhibitor) or vehicle was bilaterally injected into the interposed cerebellar nuclear. Following these three training sessions, rabbits were tested to determine whether the previous training under the influence of anisomycin or vehicle resulted in the acquisition of conditioned responses. In this test, animals that were injected previously with the protein synthesis inhibitor exhibited significantly less retention of conditioned responses than rabbits injected with vehicle. Additional experiments demonstrated that anisomycin does not block the expression of conditioned responses during conditioning or in well-trained animals. Microinjections of muscimol at the same sites of the previous drug infusions suppressed the expression of conditioned responses, indicating that the protein synthesis inhibitor was applied to the eyeblink-related parts of cerebellar circuits. The obtained data are the first to demonstrate that a manipulation of cerebellar circuits, which does not affect the performance of learned behavior, can affect the process of learning. These results suggest that the synthesis of new proteins in the intermediate cerebellum participates in the formation of plastic changes responsible for eyeblink conditioning.

Conditioned and unconditioned forelimb reflex systems in the cat: involvement of the intermediate cerebellum.

Temporary inactivation of the cerebellar interposed nuclei was used to assess the role of the intermediate cerebellum in the performance of forelimb cutaneo-muscular reflexes in the cat. The following types of reflexive responses were evaluated: the classically conditioned and unconditioned forelimb withdrawal responses and the forelimb tactile placing, hopping and magnet responses. The experiments tested the hypothesis that the intermediate cerebellum is involved in the performance of all the above forelimb reflexes. The forelimb withdrawal reflex was classically conditioned in a newly developed paradigm in which animals were first operantly conditioned to stand on four elevated platforms. Trained animals were microinjected with a gamma-aminobutyric acid (GABA) agonist, muscimol, in the interposed nuclei, and the effects of inactivation of the intermediate cerebellar output on the forelimb reflexes were examined. The main findings of these experiments are that unilateral muscimol inactivation of the interposed nuclei in the cat abolished the expression of the classically conditioned limb flexion reflex, suppressed the performance of the unconditioned withdrawal reflex and, in parallel, down-regulated the tactile placing, hopping and magnet postural responses in the ipsilateral forelimb. These observations are inconsistent with concepts indicating exclusive involvement of the intermediate cerebellum in the classically conditioned reflexes elicited by aversive stimuli. On the contrary, they support the hypothesis of a more global involvement of this structure in learned and unlearned defensive flexion reflexes and in automatic postural response systems.

Effects of muscimol inactivation of the cerebellar interposed-dentate nuclear complex on the performance of the nictitating membrane response in the rabbit.

Intracranial microinjections of the GABAA agonist muscimol were used to assess the involvement of the dentato-interposed cerebellar nuclear complex in the performance of the conditioned (CR) and unconditioned (UR) nictitating membrane responses in the rabbit. Specifically, the experiments test the hypothesis that the cerebellar nuclei are involved in the performance of both the CRs and URs. The experiments employed temporary nuclear lesions to disrupt the CRs in order to examine parallel effects on URs. Animals were conditioned in a standard delay conditioning paradigm. Injection sites at which the muscimol application disrupted execution of the CRs were identified in each rabbit. Once these sites were found, the effects of muscimol and saline injections were evaluated while alternating paired trials with unpaired trials in which only the unconditioned stimuli were applied. There are two main findings in the present study. First, the activation of the GABAA receptors in the dentato-interposed cerebellar nuclear region reduced the amplitude and increased the latency of the UR. This change in the UR closely paralleled the disruption of the CR. This observation is consistent with the notion that the cerebellum is involved in the regulation of defensive flexion reflexes. Second, cerebellar nuclear inactivation did not eliminate the tone-induced enhancement of the UR. This finding suggests the presence of cerebellum-independent circuits subserving the intermodal interaction between the conditioned and unconditioned stimuli.

Distribution of c-fos expression in brainstem neurons associated with conditioning and pseudo-conditioning of the rabbit nictitating membrane reflex.

Experiments were performed to test the hypothesis that there is a characteristic distribution of neuronal c-fos expression associated with the classical conditioning of the rabbit nictitating membrane reflex (NMR). Rabbits were divided into two groups: a conditioning group that received paired tone and airpuff stimuli in a traditional delay NMR conditioning paradigm and a pseudo-conditioning group in which the same number of tone and airpuff stimuli were applied but without being paired. Labeling was present in similar brainstem nuclei in both groups of animals. The labeled sites included trigeminal and auditory nuclei in the classical pathway for the nictitating membrane reflex as well as other nuclei such as the raphe nuclei and those in the ventrolateral medulla (VLM). However, there were quantitative differences in the labeling between the two groups. There were significantly more labeled nuclear profiles in the trigeminal nucleus of the pseudoconditioned rabbits, but more labeled nuclear profiles in the raphe nuclei in the conditioned animals. Interestingly, the ratio of the labeling in the raphe versus the VLM strongly differed between the two groups.