Investigation of cerebellar involvement in cognitive sequencing

Although there is increasing recognition that the cerebellum is involved in cognition as well as motor function, the manner in which the cerebellum contributes to cognition is uncertain. The cerebellar sequencing hypothesis posits that the cerebellum acquires sequence information, makes sequence predictions, and detects sequence violations via a forward model. Sequencing requirements are prominent in both verbal working memory (VWM) and language acquisition. Although cerebellar activation has been observed in many VWM and language investigations, the brainstem/cerebellar neural correlates of sequencing in cognition, and the influence of cerebellar sequence predictions on neocortical targets, are poorly understood. In Aim 1, we will investigate the cerebellar sequencing hypothesis in VWM. Our preliminary data indicate inferior olive (IO) involvement in sequence acquisition, whereas sequence violations are characterized by large increases in cerebro-cerebellar functional connectivity (FC). We hypothesize that cerebellar ataxia patients, who exhibit gray and white matter degeneration, should exhibit neural and behavioral abnormalities in sequencing- dependent VWM performance. Using fMRI, we hypothesize group differences in IO activation during sequence acquisition and FC between cerebellum and known critical forebrain regions (CFRs) for VWM (left inferior frontal gyrus [LIFG], supplementary motor area, and left temporal/parietal cortex) during sequence violation. Using structural neuroimaging, we predict group differences in cerebellar gray matter and peduncle integrity, as well as in structural connectivity between cerebellum and CFRs. We will correlate these structural and functional measures with task performance to infer brain-behavior relationships. In Aim 2, we will test the generalization of cerebellar sequencing in a statistical learning/language acquisition paradigm. Ataxia and control subjects will receive fMRI during a learning phase in which they listen to a pseudo language to learn transition probabilities of sequences of phonemes, and a test phase in which they will detect phoneme sequence violations. We hypothesize that ataxia patients will be impaired in detecting incorrect sequences or novel transitions of syllables. As in Aim 1, we hypothesize group differences in IO activation during sequence acquisition and changes in FC between cerebellum and known CFRs during sequence violation, but for statistical learning those CFRs are left superior temporal gyrus, striatum, and LIFG. Finally, the cerebellum is hypothesized to provide its sequence prediction computation to neocortical targets, and in Aim 3, using cerebellar transcranial magnetic stimulation (TMS) with concurrent functional MRI, we will test the hypothesis that TMS disruption during a sequencing task will produce greater changes in neocortical activation relative to an analogous control task that does not have the predictive component. These investigations will improve our understanding of cerebellar involvement in cognitive function as well as cognitive abnormalities in patients with cerebellar damage, or with diseases linked to the cerebellum.