PT  - JOURNAL ARTICLE
AU  - Kaushik Lakshminarasimhan
AU  - Marjorie Xie
AU  - Jeremy Cohen
AU  - Britton Sauerbrei
AU  - Adam Hantman
AU  - Ashok Litwin-Kumar
AU  - Sean Escola
TI  - Specific connectivity optimizes learning in thalamocortical loops
AID  - 10.1101/2022.09.27.509618
DP  - 2022 Jan 01
TA  - bioRxiv
PG  - 2022.09.27.509618
4099  - http://biorxiv.org/content/early/2022/09/28/2022.09.27.509618.short
4100  - http://biorxiv.org/content/early/2022/09/28/2022.09.27.509618.full
AB  - Cortico-thalamo-cortical loops have a central role in cognition and motor control, but precisely how thalamus contributes to these processes is unclear. Recent studies showing evidence of plasticity in thalamocortical synapses indicate a role for thalamus in shaping cortical dynamics – and thus behavior – through learning. Since corticothalamic projections compress cortical activity into a lower-dimensional thalamic activity space, we hypothesized that the computational role of thalamus would depend critically on the structure of corticothalamic connectivity. To test this, we identified the optimal corticothalamic structure that promotes biologically plausible learning in thalamocortical synapses. We found that corticothalamic structures specialized to carry an efference copy of the cortical output benefit motor control, while corticothalamic connections that communicate the directions of highest variance in cortical activity are optimal for working memory tasks. We analyzed neural recordings from mice performing grasping and delayed discrimination, and found corticothalamic interactions consistent with these predictions. These results suggest that thalamus orchestrates cortical dynamics in a functionally precise manner through structured connectivity.Competing Interest StatementThe authors have declared no competing interest.