Fig. 3: DA activation to aversive air puff is explained by bidirectional force changes.

a We exposed mice to a mild air puff to the face in a separate session to test the effect of stimulus valence on phasic DA activity. b Mice respond to air puffs with a backward change in force followed by a forward change in force (rebound). c Average change in force trace across mice (n = 24 sessions). d Backward force was generated before the rebound during the air puff trials (paired two-tailed t-test, p < 0.0001, n = 24 sessions from 4 mice). e Left, Rasters for a representative Backward DA neuron and a representative Forward DA neuron showing activation at different latencies. Right, average firing rate traces for the neurons whose rasters are shown to the left. Note latency differences. f Latency for Backward DA neuron firing was lower than Forward DA (unpaired two-tailed t-test, p = 0.018, n = 41 Forward DA, n = 35 Backward DA). g Forward DA and Backward DA peak firing rates were not different (unpaired two-tailed t-test, p = 0.33, n = 41 Forward DA, n = 35 Backward DA). h Forward and Backward DA neuron firing activity was tuned for each population’s preferred force direction during air puff trials (Left, linear regression; Non-preferred, Forward DA, R² = 0.32, Backward DA, R² = 0.47; Right, linear regression; Preferred, Forward DA, R² = 0.94, Backward DA, R² = 0.92). Error bars represent SEM. Source Data provided.