Fig. 2: Ach and DA signals are dynamically correlated during reward-based decision-making.
From: Dopamine and glutamate regulate striatal acetylcholine in decision-making
a, DA release detected by dLight1.1 and rDAh recorded from bilaterally injected mice (left). The average z-scored sensor signal ± s.e.m. is shown (errors are often smaller than line thickness) (n = 3 mice). AAV, adeno-associated virus. b, Overlay of simultaneously recorded DA and Ach dynamics, and schematic of the injection strategy. Data are shown as in a (n = 6 mice). c, Confocal images of sensor expression in neurons of the VLS for a representative mouse recorded in b. DAPI is a nuclear marker. Scale bar, 10 µm. d, Covariance of DA and Ach signals from a 2ABT session in which DA lags Ach. Ach signals are compared to DA signals from another session (green) or randomly shifted signals from the same session (orange). The average covariance ± s.e.m. is depicted (n = 6 mice). e, Covariance of trial-segregated DA and Ach signals (left) and their noise (right) in which DA lags Ach by the indicated time. Data are shown as in d. The insets highlight the time offset of the minimum covariance signal. f, Full time-dependent covariance analysis of DA and Ach signals. The average signals ± s.e.m. are shown within the top and left subplots. An enlarged view of the outlined region in white is shown to the right with the time (s) indicated (n = 6 mice). g, Summary of the off-diagonal negative covariance calculated from the matrices in e. h, Photometry kernels produced by a GLM that incorporates behavioural, history and photometry variables. The mean kernels ± s.d. that predict Ach signals from rDAh signals (rDA to gAch) and DA signals from Ach3.0 signals (gAch to rDA) are shown (n = 6 mice).
