Figure 2

Optogenetic inhibition of PV+ interneurons affects ongoing brain activity and increases LFP responses to single-whisker deflections. (A) Five example traces from one animal showing background activity and light-evoked negative deflections of the LFP. The duration of light application is indicated by the horizontal gray bar. (B) Average LFP activity with and without light application (each trace averaged over n = 8 mice, 25 trials per stimulus, shading represents SEM over mice). In this and all other LFP panels of Fig. 2, LFP traces were high-pass filtered at 4 Hz to remove light artifacts (see Methods). (C) Spectrograms of 100 ms LFP segments during light application (light on for 153 ms, first 20 and last 33 ms omitted from analysis to exclude laser onset and offset artifacts). Light increases LFP power across all frequencies. (D) Average LFP responses in deep L2/3 to a single 30 Hz cosine pulse to whisker W1 with and without light application. (E) Comparison of the combined effect of whisker deflection and light application (dark gray, same trace as in (D) and the calculated summation of the LFP responses to whisker deflection and light application in isolation (light gray). (F) Difference between the two traces in E. (G) Boxplot of average peak amplitudes within the first 60 ms after stimulus onset; data from individual mice shown as gray lines. Corresponds to the traces shown in panel D. Asterisks in this and other panels indicate significant (p < 0.05) differences; n.s., not significant. In this and all other figures of the main text, boxplots were constructed such that the box ranges from the 1st to the 3rd quartile of the data distribution, the horizontal line represents the median, and the whiskers encompass either the entire data distribution or extend until 1.5 times the interquartile range, whichever is shorter. In the latter case, more extreme data points (outliers) are shown as + signs. (H) Same as in G, but for peak latency. (I) As in G, but calculated from the data shown in E. (J) As in I, but for peak latency.