Back to U19 Main Page
DOPE Bernardo Sabatini Towards a unified framework for dopamine signaling in the striatum
Title: Dopamine recordings during a self-timed behavior
Abstract:
GCaMP6f fiber photometry recordings from genetically-defined dopaminergic cell bodies in the substantia nigra pars compacta (SNc), the ventral tegmental area (VTA), and/or dopaminergic axon terminals in the dorsolateral striatum (DLS) were collected from water-deprived, head-fixed mice as they executed a self-timed movement task (n=12 mice). In a separate cohort of animals, dopamine release in the DLS was monitored during the self-timed movement task by fluorescence of one of the two novel dopamine indicators, dLight1.1 (n=5 mice) or DA2m (n=4 mice). We additionally co-expressed tdTomato as a control fluorophore to detect optical artifacts. Ongoing body movements were monitored by neck EMG, high speed video, and back-mounted accelerometer. Mice were given a 5 uL juice reward if the first lick following a start timing cue occurred within a reward window (3.333-7s after the cue). If the mice first-licked before or after the reward window, the mouse was not rewarded for the trial and had to wait the full trial duration before entering a 10 s intertrial interval. Each animal completed up to 26 behavioral sessions with 400-1500 trials each. Mice learned to target their licking toward the reward window, and we sought to relate the natural variability in the timing of these licks to the dopaminergic signal unfolding during the timing interval. Dopaminergic signals were analyzed by aligning to the cue and first-lick events. For analyses of averaged data, trials were pooled based on first-lick time, and we also repeated these analyses on single trial data, Two features were apparent in the data: a baseline offset in dopaminergic signal predictive of single-trial movement timing as well as a ramping signal reminiscent of a threshold process, in which the rate of ramping toward an apparent threshold level was likewise predictive of single-trial movement time. We quantified the relationship between the dopaminergic signal, ongoing body movements/artifacts and movement timing with a generalized linear encoding model. We quantified the predictive power of dopaminergic signals on movement timing with two complementary decoding models: 1) a single-trial threshold model, and 2) a generalized linear decoding model whose predictors included the dopaminergic signal as well as other task variables and movement signals.