Alexandra Bova, Matthew A. Weber, Mackenzie M. Spicer, Vaibhav Khandelwal, Rachael A. Volkman, Madison McMurrin, Julia Kryca, Kartik Sivakumar, Braedon Q Kirkpatrick, Qiang Zhang, Nandakumar S. Narayanan
Across species, prefrontal and striatal neurons exhibit time-dependent ramping activity, defined as a consistent monotonic change in the firing rate across temporal intervals. However, it is unclear if ramping activity is related to the cognitive process of estimating time or to other behavioral factors such as anticipating reward or regulating movements. Here, we harnessed two novel approaches to determine how these factors contribute to prefrontal and striatal ramping activity in male and female mice performing an interval timing task. First, to determine how movement contributes to ramping activity, we tracked movement velocity using DeepLabCut as well as task-specific movements while recording prefrontal or striatal ensembles during interval timing. We found that time was more accurately decoded by ramping neurons than movement-modulated neurons, with the exception of prefrontal velocity-modulated neurons. Second, to disambiguate temporal signals from anticipatory reward signals, we compared activity patterns in neurons that were recorded during interval timing to the same neurons recorded during a pavlovian conditioning task. We found more ramping activity and more accurate temporal decoding by neuronal ensembles during interval timing compared with pavlovian conditioning. Together, these data quantify contributions of time estimation, movement, and reward anticipation in prefrontal and striatal ensembles, and they suggest that ramping is a cognitive signal that estimates time. Our results provide insight into how prefrontal and striatal ensembles multiplex information to effect temporal control of action.