A cerebellar granule cell–climbing fiber computation to learn to track long time intervals

In classical cerebellar learning, Purkinje cells (PkCs) associate climbing fiber (CF) error signals with predictive granule cells (GrCs) active just prior (~150ms). Cerebellum also contributes to behaviors characterized by longer timescales. To investigate how GrC-CF-PkC circuits might learn seconds-long predictions, we imaged simultaneous GrC-CF activity over days of forelimb operant conditioning for delayed water reward. As mice learned reward timing, numerous GrCs developed anticipatory activity ramping at different rates until reward delivery, followed by widespread time-locked CF spiking. Relearning longer delays further lengthened GrC activations. We computed CF-dependent GrC→PkC plasticity rules, demonstrating that reward-evoked CF spikes sufficed to grade many GrC synapses by anticipatory timing. We predicted and confirmed that PkCs could thereby continuously ramp across seconds-long intervals from movement to reward. Learning thus leads to new GrC temporal bases linking predict..., , , # A cerebellar granule cell–climbing fiber computation to learn to track long time intervals

[https://doi.org/10.5061/dryad.bk3j9kdm6](https://doi.org/10.5061/dryad.bk3j9kdm6)

Data files (stored in matlab (.mat) format) are used to produce main and supplementary figures. The description of all variables is as follows:

For each mouse on each session, there is a data structure with many fields, some of which are present only contextually for some session types.

Call the current structure session’s data structure \"curd\"

\"curd\" contains many fields, among them:

• nIC_GrC, nIC_CF - number of GrCs or CFs

• pixh, pixw - image height and width in pixels

• ICmat_CF, ICmat_GrC – Npixh X Npixw X Ngrc or Ncf

• dtb, dtimCb, dtDLC - sampling time step in sec for NIdaq behavioral data, imaging data, and behavioral video data

• ntb, ntimCb - total # of samples; NIdaq or microscope

• nc - total number of detected forelimb movements

• \"midAlgn\" and \"rewAlgn.\" Each of these fields is in turn a...