The change in Vm power (relative to the spontaneous level) was expressed in decibels. The change in coherence within the frequency band is the same as in Figure 4H. The reduction of low-frequency synchrony was correlated with a decrease in low-frequency power (Figure 5A; r = 0.36, p = 0.017). However, the change in Vm power only accounts for 13% of the variance in change in coherence. In addition, in 25% of the cells (11/44), a decrease in low-frequency

coherence was associated with an increase in low-frequency power. Therefore, visual stimulation seems to disrupt the intrinsic low-frequency, large-amplitude fluctuations in the network (e.g., up and down state transitions), and may also introduce additional low-frequency activity, thereby interfering with the low-frequency structure of the circuit see more Hydroxychloroquine supplier dynamics. Similar phenomena may occur throughout the cerebral cortex (cf. Churchland

et al., 2010). At high frequencies (Figure 5B), the presence of visual stimulation always increased the Vm power. This increase, in turn, correlated with the increase of synchrony (r = 0.54, p = 0.0001). We noticed, however, that for a majority of pairs (38/44, 86%), the change in power was smaller than 8 dB, and for these pairs the change could not predict the change in coherence (p = 0.095; Figure 5C). To find other factors that might contribute to the change in coherence, we then separated these pairs into two groups based on whether the mean coherence change was larger or smaller than 0.05 (Figures 5C and 5D, green and yellow). Although the average power change at high frequencies was similar for these two groups of pairs (green, 5.31 dB; yellow, 5.35 dB; p = 0.92, permutation test), the shapes of the spectrum of relative power change were different in that pairs with larger coherence increase had sharper peaks, centered

near 33 Hz (Figure 5E). We calculated an index that captured how peaked the spectrum curve was—the power change at 100 Hz divided by the that at the peak, each measured with respect to the Megestrol Acetate power change between 0 and 2 Hz (Figure 5E, H1/H2). Pairs with larger coherence increases had smaller indices, meaning that their relative power spectra were on average more peaked (0.58 versus 0.75, p = 0.005, permutation test). In previous studies using paired intracellular recordings, strong Vm synchrony caused the spike-triggered Vm average (Vm STA) between neurons to straddle the spike time (cf. Gentet et al., 2010, Lampl et al., 1999 and Poulet and Petersen, 2008). This “average synchronous excitation potential,” or ASEP, was initially identified in combined intracellular-extracellular recordings from monkey motor cortex by Matsumura et al. (1996) and is distinct from the Vm STA caused by monosynaptic connections (cf. Bruno and Sakmann, 2006).