c21  (Merck & Co)

Journal: bioRxiv

Article Title: Cholinergic–dopaminergic interplay underlies prediction error broadcasting

doi: 10.64898/2026.02.19.706866

Figure Lengend Snippet: a Schematic of the experiment. ChAT-Cre mice received bilateral injections of AAV1 viral vectors carrying the hM4D(Gi) inhibitory DREADD (n = 7) or control virus (n = 7) in the HDB, and of AAV9 vectors carrying ACh and DA sensors in the BLA and VS, respectively. This allowed us to monitor neuromodulator release using fiber photometry during chemogenetic inhibition of the BFCNs. b Left, DREADD-expressing mice receiving the C21 DREADD ligand maintained good performance on previously learned fixed go tone-reward associations (compared to control-virus injected mice: 1st new cue session, p = 0.0699; 2nd new cue session, p = 0.1923 3rd new cue session with vehicle injection, p = 0.366; compared control-virus injected mice, two-sided Mann-Whitney U-test). Right, at the same time, they failed to acquire the first novel association (compared to control-virus injected mice: 1st new cue session, p = 0.00175; 2nd new cue session, p = 0.0047; two-sided Mann-Whitney U-test). On a control day with vehicle injection (following two C21 sessions), learning deficits still persist compared to control-virus injected mice (p = 0.0047, two-sided Mann-Whitney U-test), but slightly decrease compared to the first C21 session (p = 0.0312, two-sided Wilcoxon signed-rank test). c Psychometric learning curves (see Fig.1b) comparing sessions of DREADD-expressing mice with C21 vs. vehicle injection revealed pronounced learning deficits at intermediate task difficulty levels. Lines, grand average; error shades, SEM. d Expected effects of cholinergic inhibition. Top left, dopaminergic cue responses precede synergistic BFCN responses (peak latencies; BFCN vs. T1-DAN, p = 1.8 x 10 -29 ; BFCN vs. T2-DAN, p = 1.4 x 10 -13 ; T1-DAN vs. T2-DAN, p = 1.3 x 10 -13 ; two-sided Mann-Whitney U-tests). Top right, cholinergic inhibition is expected to impair reward prediction coding and DA release after cue tones. Bottom left, fast cholinergic punishment responses are followed by slower decrease in type 1 dopaminergic activity (comparison of punishment aligned PETH peak and trough latencies, p = 9.0 x 10 -27 two-sided Mann-Whitney U-test). Bottom right, removing the cholinergic disynaptic inhibition of T1-DANs is expected to result in higher DA levels after punishments. Box-whisker plots show median, interquartile range and non-outlier range. e Average ACh release in the BLA aligned to the fixed go cue (left) and the novel go cue (right) in control (green) vs. DREADD-expressing (yellow) animals (first session with novel association; error shades, SEM). Inset, cue-evoked ACh release was reduced in DREADD-expressing mice (fixed go, p = 0.026; novel go, p = 0.0043; two-sided Mann-Whitney U-test). f New go tone-evoked ACh release in the BLA of DREADD-expressing mice, Z-scored and averaged across all novel go cues, on days with C21 (yellow) vs. vehicle (green) injection (error shades, SEM). Inset, ACh release was suppressed on days with C21 administration compared to control days with vehicle injection (p = 0.0312, two-sided Wilcoxon signed-rank test). g Same as in panels e for DA release in the VS. Cue-evoked DA release was reduced in DREADD-expressing mice (fixed go, p = 0.0087; novel go, p = 0.0022; two-sided Mann-Whitney U-test). h Same as in panels f for DA release in the VS. DA release was suppressed on days with C21 administration compared to control days with vehicle injection (p = 0.0312, two-sided Wilcoxon signed-rank test). i Average DA release in the VS aligned to punishments, comparing DREADD (orange) vs. control mice (red). Inset, DA release after punishments was enhanced by cholinergic inhibition (p = 0.0411, two-sided Mann-Whitney U-test; error shades, SEM). j Average DA release in the VS aligned to punishments, comparing C21 (orange) vs. vehicle (red) sessions in DREADD-expressing mice. Inset, DA release after punishments was enhanced following cholinergic inhibition (p = 0.0312, two-sided Wilcoxon signed-rank test; error shades, SEM). *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Article Snippet: For the chemogenetic experiments, we used C21 (Merck) as the ligand at a dose of 1 mg/kg (stock concentration: 0.2 mg/ml dissolved in physiological saline) following the dosing guidelines described in .

Techniques: Control, Virus, Inhibition, Expressing, Injection, MANN-WHITNEY, Activity Assay, Comparison, Whisker Assay