The pharmacologically increased low γ (either with low doses of PTX or with TBOA) enhanced this difference by increasing coherence in low γ but not in high γ (Figure 5B; Figure S4). We also computed the spike-triggered average of distant γ oscillations and spike-field coherence, which estimates the coherence between unit firing and the distant LFP independently of changes in oscillation power or spike rate (Fries et al., 2001; Figure 5C). Again, baseline

spike-field coherence was higher in low γ compared to high γ and drug injection increased coherence specifically in the low-γ range after PTX or TBOA treatments (Figure 5C; Figure S4). This selective effect on low γ was also observed on the phase

preference of MC spiking activity relative to the distant γ cycle (Figures S4A and S4B). In contrast selleck screening library to the weak distant γ phase preference in the baseline condition (n = 16/25 cells for PTX and n = 6/8 for TBOA with Rayleigh test, p < 0.005), the pharmacologically enhanced low γ was associated with a dramatic enhancement of the strength of distant γ phase modulation in the low-γ range (+494.1% ± 93.0% with PTX and +158.1% ± 45.8% with TBOA compared to baseline) but not in the high-γ range (Figure S4). We next measured spike synchronization between pairs of distant MCs. Under baseline conditions, this website pairs of MCs displayed a nearly flat cross-correlation histogram (Figure 5Di), indicating a lack of temporal relationship between MCs and confirming that recorded pairs of MCs do not belong to the same glomerulus (Schoppa and Westbrook, 2001).

When low-γ oscillations increased, the cross-correlograms of MC pairs displayed a peak centered on zero (lag: 0.2 ± 0.3 ms, n = 9 pairs), two side peaks (mean period, 19.6 ± 0.3 ms, n = 9), and a strong oscillatory pattern specifically in the low-γ regime (Figure 5Dii). A significant increase in the correlation index confirmed that increased low γ was associated with the emergence of synchrony in the low-γ band from distant and previously unsynchronized MC pairs (Figure 5Diii). To test whether coherent MYO10 MC activity is sufficient to drive γ oscillations, we selectively manipulated MC firing activity by targeted optogenetic stimulation in transgenic mice expressing ChR2 in the MC population (Thy1:ChR2-YFP mice, line18; Figures 6A and 6B). Targeting the dorsal surface of the OB, we first examined the reliability of light-induced firing activity in response to light-train stimuli (5 ms light pulse duration) with increasing frequency. Light pulses reliably triggered action potentials with stereotyped spike latencies ( Figure 6C). Firing activity followed light pulses from 25 to 90 Hz, with a slight decrease in fidelity at higher frequencies (−22.6% ± 9.4% between 25 and 90 Hz stimulation, p = 0.015 with a paired t test, n = 9; Figure 6C).