We applied CsF-DIDS in repatches of seven cells after having collected a sufficient number of ripple-associated

cPSCs under control conditions close to the potential of Cl− reversal. In line with our hypothesis, ripple-associated fast synaptic inputs indeed persisted in the repatch recording with disrupted GABAAR-mediated learn more synaptic transmission (Figure 6C). We again analyzed downward and upward slopes of putative EPSCs and compared their values before and following perfusion of the cells with CsF-DIDS. Moreover, ripple-locked downward cPSC slopes were unchanged following intracellular block of inhibition (control: 24.3 ± 0.8 pA/ms, n = 224 cPSCs; CsF-DIDS: 26.6 ± 0.7 pA/ms, n = 462 cPSCs; 7 repatched cells; p = 0.1; K-S test), whereas upward slopes were slightly enhanced (control: 12.9 ± 0.3 pA/ms; CsF-DIDS: 13.9 ± 0.2 pA/ms; Figure 6D; p < 0.0001; K-S test). Additionally, we examined the intervals between

successive downward slopes. Distributions peaked at 4–5 ms, consistent with ripple frequency, both in control conditions and after CsF-DIDS administration (Figure 6E; see Figure S6B for single-cell data). Taken together, these results derived from experimentally blocking the somatic postsynaptic action of GABAergic inputs corroborate our hypothesis that ripples are accompanied by a strong oscillation-coherent phasic excitatory component. We next asked whether Buparlisib in vivo Diclofenamide ripple-coherent cPSCs represent the spiking output of CA3 pyramidal neurons (Both et al., 2008)

or whether they are generated locally within the CA1 network. We used “minislices” where area CA1 was isolated from the adjacent CA3 and subiculum (Figures 7A and 7C). In this experimental system, we observed SWRs at a rate of 0.46 ± 0.09 Hz (median: 0.46 Hz; range: 0.13 Hz to 0.93 Hz; 8 CA1 minislices; Figure 7B). Ripple frequency in these events was 213.1 ± 6.6 Hz on average (median: 215 Hz; range: 175 Hz to 235 Hz; Figure 7B, right). To test whether ripple-coherent cPSCs survived in the isolated area CA1, we again recorded from principal neurons voltage-clamped close to the reversal potential of Cl− (−66 mV). SWRs in CA1 minislices were indeed accompanied by phasic inward currents at ripple frequency that were also phase coherent with LFP ripples (Figures 7D–7E; n = 725 cPSCs; 5 cells). Moreover, in minislices, cPSC downward slope phases with respect to LFP ripples (−101° ± 8°, Figure 7F) were comparable with those derived from intact slices (−114° ± 10°, Figure 4E). In summary, this set of experiments demonstrates the possibility of a local origin of ripple-coherent excitatory PSCs within area CA1. The observation that excitatory PSCs are phasic and ripple-locked raised the question of whether they could account for the timing of action potentials in target CA1 principal neurons.