In this terminology, “steady-state” current corresponds to what has traditionally

been called “persistent” current, as defined by slow ramps, and we use the terms interchangeably. We were somewhat surprised that a model of a single uniform population of sodium channels could give a good prediction of the experimentally observed currents, because CA1 pyramidal neurons (whose experimental data were used to tune the model) probably express current from multiple types of sodium channels. Subthreshold current in CA1 neurons is JAK cancer partly from Nav1.6 channels (Royeck et al., 2008), which are prominently expressed in many neuronal types with large persistent currents (e.g., Raman et al., 1997; Maurice et al., 2001; Enomoto et al., 2007; Osorio et al., 2010; Gittis et al., 2010; Kodama et al., 2012) but persistent sodium current in CA1 pyramidal neurons from Nav1.6 null mice is reduced by only SCH772984 concentration ∼40% (Royeck et al., 2008), suggesting substantial contributions from other channel types also. The persistent sodium current in the Nav1.6 null animals has almost identical voltage dependence with that in wild-type animals (Royeck et al., 2008), suggesting that the voltage dependence of non-Nav1.6 channels must be very similar to that from Nav1.6. This

makes it plausible that a single model can account reasonably well for current from mixed sources. The model does not account for resurgent sodium current, a component of sodium current expressed in Purkinje neurons (Raman and Bean, 1997) and some CA1 pyramidal neurons (Castelli et al., 2007; Royeck et al., 2008). Resurgent current requires depolarizations depolarized to −30mV to be activated significantly (Raman and Bean, 2001; Aman and Raman, 2010) and should be minimally engaged by the protocols we used for exploring subthreshold current or by EPSP waveforms (Figures 7D–7I), where all voltages were below −40mV. The model also does not account for a process of slow inactivation, which affects both transient and persistent sodium current

(Fleidervish and Gutnick, 1996; Mickus et al., 1999; Aman and Raman, 2007) and produces roughly parallel changes in the two components (Taddese and Bean, 2002; Do and Bean, 2003). Modeling slow inactivation accurately second (e.g., Menon et al., 2009; Milescu et al., 2010) was not feasible as we did not use protocols designed to characterize it under our conditions. In many neurons slow inactivation was minimal with the protocols we used (e.g., Figure 4A), so it is unlikely to be important for the essential relationship between persistent and transient current studied here. TTX-sensitive sodium current has been shown to amplify EPSPs in many neuronal cell types, including cortical pyramidal neurons (Deisz et al., 1991; Stuart and Sakmann, 1995; González-Burgos and Barrionuevo, 2001), hippocampal CA1 pyramidal neurons (Lipowsky et al.