Figure 3

Learning to discriminate artificial sensations using DCS resulted in changes in striatal local field potentials (LFPs). (a) Behavioral box used for training rats to discriminate two DCS patterns (Virtual narrow: 100 pulses vs. Virtual wide: 1 pulse). Each trial was divided into pre-stimulus, stimulus and post-stimulus periods. A typical trial comprised of the following sequence - DO: Door Open, BB: Beam Break, CP: Center Poke, RD: Reward delivery. Stimulation began at BB, when the rat crossed a photobeam inside the inner chamber. Spike and LFP activity from M1, S1 and STR was recorded during the training period and comparisons were made between early (1^st session) and late (last session) periods of training. (b) Rats exhibited significant improvement in discrimination performance between early and late sessions. (c) Sensory latency (measured as time between BB and CP, left graph) significantly increased, while response latency (measured as time between CP and RD) significantly decreased between early and late training sessions. (d) Representative time-frequency spectrogram of same striatal LFP channel recorded during entire early and late session of one rat (red marker above graphs indicates stimulation onset time for each trial). Session-averaged spectral power for the same channel showing no difference between early and late sessions is shown at right. (e) Trial-averaged pre-stimulus striatal LFP spectral power for one rat showing clear increase in the delta (1.5–4.5 Hz) and theta (5–9.5 Hz) frequency range from early to late learning. Time (0) corresponds to stimulation onset. (f,g) Mean pre-stimulus striatal spectral power in the delta and theta range (pink arrow in f) significantly increased from early to late sessions in all rats (n = 5). (*p < 0.05, **p < 0.01, ***p < 0.001).