The results with dissociated CA1 neurons predict minimal engagement of either persistent or transient components of sodium current FG-4592 purchase at voltages negative to −80mV and increasing engagement at more depolarized voltages in the range from −70mV to −60mV. To test the dependence

of uncaging-evoked EPSPs (uEPSPs) on membrane potential in this range, the resting potential of the neuron was adjusted to different voltages in each experiment using direct current from the amplifier. Figure 6A shows the mean ± SEM of uEPSPs from spines recorded in control solutions from holding potentials of −83mV (light gray), −73mV (gray), and −63mV (black). The peak voltage change of the uEPSP evoked by stimulation of a single spine was ∼1mV when the membrane potential was −83mV, and the peak uEPSP increased Ibrutinib progressively when the holding potential was depolarized to −73mV or −63mV, with a ∼20% enhancement when elicited from −63mV (Figure 6B). The enhancement at −63mV compared

to −83mV was statistically significant (p = 0.024, paired t test, n = 18). Consistent with originating by engagement of voltage-dependent sodium current, this effect was absent when the same experiment was performed in the presence of TTX (Figures 6C and 6D; p = 0.91, n = 21, paired t test comparing −63mV and −83mV). As expected from this comparison, the size of the uEPSP was significantly smaller in TTX than control when elicited from −63mV (p = 0.04, unpaired t test) but not when elicited from −83mV (p = 0.63, unpaired t test). The effect of TTX to reduce EPSPs evoked in spines of CA1 neurons is similar to previous results seen with stimulation of spines in neocortical pyramidal neurons (Araya et al., 2007). Do the components of subthreshold transient and steady-state sodium current come from the same channels that carry suprathreshold transient current? To explore

whether this is likely in principle, we tested the prediction of kinetic models for sodium channel gating. Figure 7A shows a Markov model for sodium channel gating based on previous models formulated to match experimental measurements of suprathreshold transient sodium current (Kuo and Bean, 1994) or both persistent and transient current (Taddese and Bean, 2002; Milescu et al., 2010) in other types of central neurons. The rate constants Mannose-binding protein-associated serine protease were adjusted so that the predicted suprathreshold transient current (Figure 7B) matched the voltage dependence and kinetics of current recorded in CA1 neurons under our experimental conditions. The model predicted a midpoint of activation of transient current of −36mV and a midpoint of inactivation of −65mV (Figure 7C), corresponding to typical experimental values. We found that the model predicts both subthreshold steady-state and subthreshold transient current, with kinetics and voltage dependence similar to the experimentally measured currents. The model predicts steady-state conductance with a midpoint of −63mV and a slope factor of 3.