The cerebellum is accepted to have a crucial role in classical conditioning. Following the classical work of Eccles and co-workers in elucidating its detailed neurophysiology (Eccles et al., 1967), several theorists developed computational models of cerebellar learning (Albus, 1971). In Albus’ original conception, the so-called ‘inactivation response’ of a Purkinje cell (PC), a pause in PC spontaneous activity associated with a climbing-fibre response (CFR), could be interpreted as the internal neural representation of the overt unconditional response (UR). He further suggested that mossy/parallel fibre (MF/PF) activity produced by the conditional stimulus (CS) could be considered its internal neural representation. The effect of learning, by changing PF-PC synaptic weights with conjunctive CF/PF inputs, and the acquisition of a conditioned response would, he hypothesized, be accompanied by a conditioned pause of Purkinje neurones. Direct recordings from animal models have since provided evidence to support the Albus hypothesis. There is also strong evidence from the effects of lesions in humans that the cerebellum is required for the acquisition of classically conditioned eye blink responses. However, to date the Albus model has not been directly tested in intact human subjects.
It had been widely thought that the cerebellum is particularly difficult to record from non-invasively. However, recent work using EEG/MEG techniques supports the view that non-invasive electrophysiology of the cerebellum is indeed viable. In our own work we have reported cerebellar evoked potentials produced by vestibular and axial stimuli from scalp electrodes from placed over the posterior fossa the properties of which are consistent with a CFR (Todd et al., 2017). In addition, we also observe an ‘inactivation response’ manifest as changes in the high frequency electrocerebellogram (ECeG: Todd et al, 2018) which has a higher frequency content than cerebral EEG. These observations suggest that it may be possible to directly test the Albus hypothesis in humans and this was the aim of the study reported here.
Electrophysiological activity was recorded in 14 healthy subjects (compliant with the Declaration of Helsinki) before, during and after a classical eye-blink conditioning procedure with a 500 ms auditory tone as CS and a maxillary nerve US (Todd et al 2023). Electrodes recorded EMG/EOG at peri-ocular sites, EEG over frontal eye-fields and the ECeG over the posterior fossa. ECeG high frequency power was computed using the continuous wavelet transform and each epoch segmented in time for statistical analysis before and after conditioning.
Of the 14 subjects half strongly conditioned while the other half were resistant. However, inhibition of cerebellar activity in the form of a significant reduction in high frequency ECeG power was observed in all subjects, prior to the CR, as shown in Figure 1. Pairwise comparisons of baseline, with conditioned and unconditioned pausing are given: *, p <.05, **, <.01, ***, <.005, ns = not significant.
We conclude that while conditioned cerebellar pausing may be necessary, it is not sufficient alone to produce overt behavioural conditioning, implying the existence of another central mechanism. The outcomes of this experiment indicate the value of the non-invasive electrophysiology of the cerebellum.
