KATP channels modulate intrinsic firing activity of immature entorhinal cortex layer III neurons

Medial temporal lobe structures are essential for memory formation which is associated with coherent network oscillations. During ontogenesis, these highly organized patterns develop from distinct, less synchronized forms of network activity. This maturation process goes along with marked changes in intrinsic firing patterns of individual neurons. One critical factor determining neuronal excitability is activity of ATP-sensitive K+ channels (KATP channels) which coupled electrical activity to metabolic state. Here, we examined the role of KATP channels for intrinsic firing patterns and emerging network activity in the immature medial entorhinal cortex (mEC) of rats. Western blot analysis of Kir6.2 (a subunit of the KATP channel) confirmed expression of this protein in the immature entorhinal cortex. Neuronal activity was monitored by field potential (fp) and whole-cell recordings from layer III (LIII) of the mEC in horizontal brain slices obtained at postnatal day (P) 6–13. Spontaneous fp-bursts were suppressed by the KATP channel opener diazoxide and prolonged after blockade of KATP channels by glibenclamide. Immature mEC LIII principal neurons displayed two dominant intrinsic firing patterns, prolonged bursts or regular firing activity, respectively. Burst discharges were suppressed by the KATP channel openers diazoxide and NN414, and enhanced by the KATP channel blockers tolbutamide and glibenclamide. Activity of regularly firing neurons was modulated in a frequency-dependent manner: the diazoxide-mediated reduction of firing correlated negatively with basal frequency, while the tolbutamide-mediated increase of firing showed a positive correlation. These data are in line with an activity-dependent regulation of KATP channel activity. Together, KATP channels exert powerful modulation of intrinsic firing patterns and network activity in the immature mEC.

Neuronal excitability strongly depends on cellular energy metabolism. Lack of energy is particularly damaging during the prenatal and early postnatal periods, resulting in lasting neurological deficits throughout life (Nelson, 1989;Erecinska et al., 2004). ATP-sensitive K + channels (K ATP channels) provide a unique link between cellular energy state and electrical activity. K ATP channels are inwardly rectifying K + -selective ion channels that are inhibited by intracellular ATP. A decrease of submembrane ATP levels and accompanying rise in ADP triggers K ATP channel opening (Seino, 1999;Haller et al., 2001). K ATP channels exist in many excitable cells, including cardiac myocytes, skeletal muscle cells, pancreatic β-cells (Ashcroft and Ashcroft, 1990) and neurons (Karschin et al., 1997;Dunn-Meynell et al., 1998;Zawar et al., 1999). In excitable tissues, these channels act as metabolically controlled "excitation brakes" by hyperpolarizing cells in conditions of low ATP supply. K ATP channels are composed of four poreforming Kir6 subunits, and four regulatory sulfonylurea receptor (SUR) subunits (Aguilar-Bryan and Bryan, 1999;Nichols, 2006). While K ATP channels appear to play an important role in protecting neurons against ischemic or anoxic injury (Fujimura et al., 1997;Yamada et al., 2001;Sun et al., 2007) they are also activated in normal network states, e.g., during burst firing in respiratory neurons (Haller et al., 2001). In hippocampal granule cells open probability of single K ATP channels transiently increases in response to modest firing activity (Tanner et al., 2011).
The entorhinal cortex (EC) constitutes the major interface between the hippocampus and parahippocampal areas and plays a crucial role in spatial cognition and memory processing (Squire et al., 2004;van Strien et al., 2009;Buzsáki and Moser, 2013). Principal neurons of EC layer III (LIII) provide direct input to the apical dendritic tuft of hippocampal CA1 pyramids (Witter and Amaral, 2004) which is an important pathway for temporal association memory and fear learning (Suh et al., 2011;Kitamura et al., 2014;Lovett-Barron et al., 2014). In adult rats, medial EC (mEC) LIII principal neurons are regularly firing cells that do not discharge in bursts (Dickson et al., 1997;Gloveli et al., 1997;Yoshida and Alonso, 2007). In contrast, during early postnatal maturation a fraction of mEC LIII principal neurons spontaneously generates prolonged Ca 2+ -and voltagedependent intrinsic bursting activity (Sheroziya et al., 2009). These burst discharges involve the Ca 2+ -sensitive non-specific cationic current (I CAN ), persistent Na + current (I Nap ), and -for termination -Ca 2+ -activated K + current (I AHP ; Sheroziya et al., 2009).
Given that increased firing frequency or bursts can elicit opening of K ATP channels (Haller et al., 2001;Tanner et al., 2011), we investigated the role of K ATP channels for cellular and network activity in the immature mEC.
We report that excitability of immature neurons and early patterns of network oscillations are powerfully modulated via K ATP channels. These findings indicate that neuronal ATP consumption and energy demand might have important consequences for postnatal activity-dependent maturation of neurons and networks in the mEC.

ETHICAL APPROVAL
All experimental protocols were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the ethical committee of the Institute of Higher Nervous Activity and Neurophysiology, Russian Academy of Sciences (IHNA RAS) or by the state government of Baden-Württemberg, Germany. All efforts were made to minimize animal suffering and to reduce the number of animals used.

PREPARATION OF BRAIN SLICES
Horizontal brain slices (350-600 μm thick) containing the hippocampus, entorhinal and parts of perirhinal cortices were obtained from Wistar rats at postnatal day (P) 6-13 using standard procedures. P0 was taken as the day of birth. Rats were purchased from Charles River Laboratories (Sulzfeld, Germany) or from the local veterinary service (INHA RAS, Russia). Animals were decapitated, brains were rapidly removed and placed in cold (1-4 • C) oxygenated artificial cerebrospinal fluid (ACSF) containing (in mM): 124 NaCl, 3 KCl, 1.6 CaCl 2 , 1.8 MgSO 4 , 26 NaHCO 3 , 1.25 NaH 2 PO 4 , and 10 glucose (for recordings in interface-type chambers) or 130 NaCl, 3.5 KCl, 1.2 NaH 2 PO 4 , 25 NaHCO 3 , 1.3 MgCl 2 , 1 or 2 CaCl 2 , and 25 glucose (for recordings in submerged conditions). Solutions were saturated with 95% O 2 and 5% CO 2 (pH 7.4 at 37 • C). Brain slices were cut using a Vibratome (Leica VT1000S, Germany or Campden Instruments, Loughborough, UK). For extracellular fp recordings, slices were transferred into a Haas-type interface chamber, maintained at 34 ± 1 • C and superfused with ACSF at a rate of 1.5-2 ml/min. Prior to electrophysiological recordings, slices were allowed to recover for at least two hours. For whole-cell patch-clamp recordings under submerged conditions, slices were stored at 34 ± 1 • C for 10 min in a holding bath containing ACSF, before cooling down to room temperature. After incubation for at least 1 h at room temperature, individual slices were transferred into a recording chamber, superfused with oxygenated ACSF at a rate of 3-6 ml/min and maintained at 33 ± 1 • C.

RECORDING PROCEDURES
Whole-cell patch-clamp recordings were performed under visual guidance using an Olympus microscope fitted with infrared differential interference contrast optics (Olympus BX51WI). We preferentially recorded pyramidal neurons located in the deep part of LIII in order to exclude potential recording from LII pyramidal cells. The lamina dissecans, a distinct cell-free zone (sometimes referred to as layer IV), was used as a reference to identify the border between LIII and layer V. Current-clamp recordings were performed with an Axopatch 1D patch-clamp amplifier (Axon Instruments, Foster City, CA, USA) or ELC-03XS amplifier (npi electronics, Tamm, Germany). Patch electrodes were backfilled with the following solution (in mM): 115 K-gluconate, 20 KCl, 10 disodium phosphocreatine, 10 HEPES, 4 MgATP, and 0.3 GTP or 135 K-gluconate, 20 NaCl, 10 HEPES, 3.95 Mg-gluconate, 0.05 MgATP, and 0.3 GTP (tip resistance of 5-7 M ). The electrode solutions were adjusted to pH 7.3 with 1 M KOH. Data were lowpass filtered at 1-2 kHz, digitized at 5-10 kHz (Digidata 1322A, Molecular Devices), and stored on a personal computer using the AxoScope software package (Molecular Devices). Extracellular fp recordings were performed with an EXT 10-2F amplifier (npi electronics, Tamm, Germany). Signals were amplified 100×, low-pass filtered at 2 kHz and high-pass filtered at 0.3 Hz, digitized at 20 kHz with an analog-to-digital converter (Cambridge Electronic Design (CED) MICRO 1401 mkll, Cambridge, UK) and saved on a computer using Spike2 software (CED, Cambridge, UK) for offline analysis. Fp recordings were obtained with ACSF-filled borosilicate glass electrodes (tip diameter 3-5 μm) placed in LIII of the mEC. In the interface chamber, we identified the EC and its layers with a dissecting microscope.

Frontiers in Cellular Neuroscience
www.frontiersin.org

DATA ANALYSIS
Electrophysiological data were analyzed off-line using Spike2 software (CED, Cambridge, UK) and Clampfit (Molecular Devices). Fp activity was analyzed from primary data sections lasting at least 10 min. Duration of fp-bursts was measured from the onset of negative potential deflection until the peak of positivity before the field waveform return to baseline level. The amplitude of fp-bursts was calculated as the difference between baseline and the negative peak of the fp transient. Whole-cell recordings were typically started 10-15 min after break-in, when balance between intracellular millieue and patch-solution was established. When depolarizing current injections were needed to induce bursting, recording was delayed for at least 5 min after onset of stable bursting. Bursting neurons were recorded in standard bicarbonate-based ACSF containing 1 mM Ca 2+ , and regularly firing cells in ACSF containing 1 or 2 mM Ca 2+ . Spontaneous intrinsic bursting activity was analyzed from at least 3 min of recordings. Burst duration was calculated as the time between the first and the last spike within a burst. Regular firing was induced by depolarizing current steps (11 or 35 s duration, +9 to +80 pA), with intervals between pulses of 19 or 55 s, respectively. Regular firing frequency was analyzed from the 10 s of a depolarizing step (averaged data from three current pulses) under control conditions, in the presence of drugs and after washout. For statistical evaluation of drug effects, baseline values were compared to the latest phase of the interval with drugs present (at least 30 min for fp recordings/interface-type and 15 min for whole-cell recordings/submerged-type).

STATISTICAL ANALYSIS
Averaged data are given as mean ± SEM. Statistical analysis was performed using GraphPad (InStat, San Diego, CA, USA) software. Fp parameters and part of single cell data were compared by a one-way repeated-measures ANOVA followed by appropriate post hoc tests, depending on parametric or non-parametric data distribution. Paired two-tailed Student's t-test or Wilcoxon matched pairs signed ranks test was used for statistical comparison of the remaining single cell data. A p value <0.05 was regarded as significant. For all data: * p < 0.05, * * p < 0.01, * * * p < 0.001, ns, not significant.

RESULTS
We first analyzed expression of Kir6.2 (a major subunit of the K ATP channel) within the second postnatal week. We found that Kir6.2 is extensively expressed in the EC at the protein level, suggesting the presence of K ATP channels in this structure during an early developmental stage ( Figure 1A). Specificity of antibodies was shown in HEK293T cells upon silencing of Kir6.2 expression by siRNA ( Figure 1B; for details, see Materials and Methods). Immunocytochemical experiments revealed Kir6.2 expression across all layers of the immature rat mEC ( Figure 1C). Kir6.2 immunoreactivity was predominately detected on somata and proximal dendrites of individual cells located in layers II, III, and V. In order to address the functional role of K ATP channels at such early stages, we investigated network-and single-cell activity in the mEC. Fp recordings within layer III of slices from rats at P10-P13 revealed spontaneous repetitive bursts. These glutamate receptor-mediated events were characterized by prolonged fp shifts (duration 6.9 ± 0.7 s, peak amplitude 0.056 ± 0.004 mV, n = 22 slices), superimposed by fp fluctuations at ∼15-30 Hz and multiple unit discharges, as previously reported by Sheroziya et al. (2009). Importantly, the K ATP channel opener diazoxide (400 μM) reversibly reduced the duration of fp-bursts from 7.7 ± 1.3 s to 4.6 ± 0.3 s (n = 8, p < 0.05, ANOVA Bonferroni's post hoc; Figures 1D,E). The K ATP channel blocker glibenclamide (10 μM) had opposite effects and slightly prolonged fp-bursts (control vs. glibenclamide; 6.5 ± 0.8 s vs. 7.6 ± 0.5 s, n = 14, p < 0.05, ANOVA Bonferroni's post hoc; Figure 1E). The effect of glibenclamide was not reversible after washout of the drug (tested in five slices). Indeed, fp-burst duration showed a further increase following washout of the substance. We therefore opted to antagonize the effect of the K ATP channel blocker glibenclamide with the SUR1-specific K ATP channel opener NN414 (Dabrowski et al., 2003;25 μM). This led to a slight decrease of fp-burst duration which, however, did not reach significance (glibenclamide vs. NN414; 7.6 ± 0.5 vs. 7.2 ± 0.6 s, n = 14, p > 0.05, ANOVA Bonferroni's post hoc; Figure 1E). Frequency of fpbursts (i.e., burst occurrences per minute) was slightly reduced in the presence of diazoxide (control vs. diazoxide; 3.6 ± 0.4 vs. 2.4 ± 0.3, p < 0.05, ANOVA Bonferroni's post hoc), and was not affected by glibenclamide (control vs. glibenclamide; 3.6 ± 0.4 vs. 3.7 ± 0.3, p > 0.05, ANOVA Bonferroni's post hoc). Amplitude of fp-bursts was not significantly affected by both drugs (control vs. diazoxide; 0.061 ± 0.008 vs. 0.055 ± 0.005 mV, control vs. glibenclamide; 0.054 ± 0.005 vs. 0.05 ± 0.004 mV, p > 0.05 for both drugs, ANOVA Bonferroni's post hoc). Together, these results suggest that activity of K ATP channels regulates the generation and duration of spontaneous fp-bursts in the developing mEC.
Likewise, tolbutamide increased firing frequency of regularly firing neurons when these were strongly activated (Figures 4C,D). Adding tolbutamide enhanced high frequency firing from 7.1 ± 0.8 to 8.6 ± 0.9 Hz (121%, n = 7, p = 0.004, t-test). However, tolbutamide was not effective at low frequency firing (control vs. tolbutamide; 1.8 ± 0.2 vs. 2.0 ± 0.3 Hz, n = 12, p = 0.095, t-test). In line with these activity-dependent effects, tolbutamidemediated elevation of firing frequency correlated positively with control frequency (Pearson's r = 0.61, p = 0.001; Figure 4D, bottom panel). RMP of regularly firing neurons was not affected by tolbutamide (n = 12, p = 0.397, t-test). In addition, in 4 out of 16 recording neurons, tolbutamide switched neuronal activity from regular firing into bursting mode. It has been reported that the widely used K ATP channel blocker tolbutamide does also affect several Ca 2+ -and voltage-dependent K + currents in adult hippocampal neurons, including I M , I AHP, and D-type K + currents (Crépel et al., 1993;Erdemli and Krnjević, 1996). We therefore tested effects of a second K ATP channel blocker, glibenclamide, on intrinsic bursts of immature LIII neurons (Figures 5A,B). Similar to tolbutamide, glibenclamide (1 μM) strongly enhanced duration of bursts (control vs. glibenclamide; 6.1 ± 0.9 vs. 19.0 ± 4.7 s, n = 8, p = 0.017, t-test), whereas no significant effects were found for frequency within bursts (5.5 ± 0.7 vs. 5.8 ± 0.5 Hz, p = 0.341, t-test) and for occurrence of bursts (2.2 ± 0.3 vs. 1.4 ± 0.2, p = 0.078, t-test). The effect of glibenclamide was, however, not reversible after washout of the drug. As a control, we therefore measured stability of burst parameters during prolonged recordings without drugs (50 μM ATP in the patch electrode). We did not find any significant changes during time-matched control recordings (n = 6, burst duration: p = 0.430, frequency within bursts: p = 0.101, number of bursts per minute: p = 0.119, t-test for all values; Figures 5C,D), indicating that the irreversible effects of glibenclamide were indeed specifically induced by the drug. These data show that K ATP channel blockers efficiently enhance intrinsic firing activity. In summary, K ATP channels exert powerful modulation of intrinsic bursting and regular firing activity as well as spontaneous early network oscillations in the immature mEC.
Dashed lines indicate 95% confidence interval. Recordings were performed in the presence of kynurenic acid and picrotoxin and with pipette solution containing 4 mM ATP.

DISCUSSION
We show that K ATP channels are strongly involved in the modulation of spontaneous network oscillations and intrinsic firing activity in the immature rat mEC in vitro. Based on extracellular and intracellular recordings in the presence of K ATP -affecting drugs we report that these channels efficiently modulate electrical activity in layer III of the developing mEC. K ATP channel openers reduced the duration of spontaneous fpbursts and suppressed intrinsic neuronal firing activity, including both prolonged bursting and regular firing patterns. In contrast, Frontiers in Cellular Neuroscience www.frontiersin.org Data are given as mean ± SEM. Paired t -test or Wilcoxon test, *p < 0.05, ns, not significant.
K ATP channel blockers slightly prolonged fp-bursts and strongly enhanced the duration of intrinsic bursts as well as firing frequency of regularly firing neurons. In this latter case, effects at the cellular level seemed to be stronger than effects at the network (fp) level. Indeed, intrinsic bursts of single neurons contribute to, but not fully reflect the spontaneous fp-bursts, which arise from complex interactions between excitatory and inhibitory transmission as well as intrinsic neuronal properties. Termination of fp-bursts in immature EC is likely partly associated with activation of fast spiking GABAergic interneurons, as it has been reported for adult animals (Tahvildari et al., 2012). Indeed, a selective blockade of GABA A -receptors always elicited paroxysmal field discharges in the immature EC (Sheroziya et al., 2009). It is thus possible, that K ATP channel blockers increase activity of fast spiking interneurons that restricts prolongation of fp-bursts. Importantly, regular firing activity was modulated in a frequency-dependent manner, with strongest effects of the K ATP channel blocker tolbutamide on highly active neurons, and strongest effects of the K ATP channel opener diazoxide on slowly firing cells. These correlations might reflect transient changes in intracellular (submembrane) ATP concentration during electrical activity (Howarth et al., 2012). High-frequency firing activity might significantly reduce local ATP levels and, therefore, increase . Averaged data are given as mean ± SEM. ANOVA followed by Dunn's post hoc test (for B; 4 mM ATP) or Paired t -test, *p < 0.05, **p < 0.01, ns, not significant. Bottom: correlation between basal firing frequency and tolbutamide-mediated increase of firing ( frequency = frequency under 300 μM tolbutamide -control frequency). Solid line indicates linear correlation (**two-tailed p < 0.01). Dashed lines indicate 95% confidence interval. All recordings were obtained in the presence of kynurenic acid and picrotoxin.
opening of K ATP channels. In this situation, blockers of K ATP are highly efficient while drugs increasing channel opening lose effect. At low firing frequencies, higher intracellular ATP levels would exert opposite effects. Prolonged intrinsic bursting activity is a characteristic feature of developing, but not mature, rat mEC LIII neurons (Sheroziya et al., 2009). The ionic mechanisms underlying the generation of burst firing include activation of various currents: Ca 2+ -sensitive non-specific cationic current (I CAN ), persistent Na + current (I Nap ) and Ca 2+ -activated K + current (I AHP , BK-current; Sheroziya and Egorov, 2008;Sheroziya et al., 2009). Our present data show that ATP-sensitive K + current is also involved in the generation of prolonged intrinsic bursts. Number and frequency of spikes during prolonged bursts appear to be more than enough to induce opening of K ATP channels, that, together with I AHP , finally mediate burst termination. Effects of the K ATP channel blocker tolbutamide were prominent at low intracellular concentrations of ATP (50 μM). However, the drug was effective even at high ATP concentrations in the patch electrode (4 mM), as frequently used in whole-cell recordings. This value is much higher than the half-maximal inhibitory concentration of ATP for K ATP channels (∼25 μM; Nichols et al., 1991). Thus, our data support the proposal that open probability of K ATP channels reflects activity-dependent fluctuations of ATP/ADP concentrations within local submembrane domains which are not entirely controlled by the solution in the patch pipette (Haller et al., 2001;Mollajew et al., 2013). It is likely that such local changes in ADP/ATP ratio represent submembrane ATP consumption during increased Na + -K + -ATPase activity (Haller et al., 2001).
In the adult rat hippocampus, the K ATP channel blocker tolbutamide also affects several Ca 2+ -and voltage-dependent K + currents, including M-type K + current and I AHP (Crépel et al., 1993;Erdemli and Krnjević, 1996). We can not exclude that such mechanisms do also play some role in the immature mEC. It has been reported, however, that K ATP channels themselves participate in the slow afterhyperpolarization (sAHP) in hippocampal dentate granule cells of juvenile mice, and thereby affect neuronal excitability following elevated firing (Tanner et al., 2011). The role of M-currents in immature neuronal excitability seems to be not Frontiers in Cellular Neuroscience www.frontiersin.org significant. For example, the Kv7/M channel blocker linopirdine has only a minor effect on neonatal activity, in contrast to juvenile CA3 pyramidal neurons (Safiulina et al., 2008). In addition, the hyperpolarization-activated current (Ih) plays an important role in modulation of excitability of adult LIII neurons (Shah et al., 2004;Huang et al., 2009). Altered expression of Ih underlying hyperpolarization-activated cyclic nucleotide-gated (HCN) channel subunits has been reported from models of temporal lobe epilepsy where it affects seizure threshold (Shah et al., 2004). Expression of HCN1 subunits in EC LIII neurons is, however, dominant in distal dendrites, and is quite moderate in somata of these cells, particular at early stages (Vasilyev and Barish, 2002). Nevertheless, we do not exclude a contribution of Ih to prolonged burst firing of immature LIII neurons.
Our results are also in line with previous reports showing that K ATP channel can be activated in response to ATP consumption during normal (i.e., physiological) levels of neuronal activity. The open state probability of K ATP channel augments in response to ATP consumption during moderate neuronal firing (Haller et al., 2001;Tanner et al., 2011) or via prolonged activation of glutamate receptors (Mollajew et al., 2013). K ATP channel are strongly involved in the modulation of various neurophysiological functions. In dopaminergic neurons of the medial substantia nigra, activity of K ATP channels enables burst firing in vitro and in vivo, thereby controlling novelty-induced exploratory behavior (Schiemann et al., 2012). In hypothalamic neurons expressing proopiomelanocortin (POMC) the age-dependent up-regulation of K ATP channels causes hyperpolarization and neuronal silencing, contributing to obesity of aged animals (Yang et al., 2012). K ATP channel are also involved in pathologically altered network activity such as epileptic seizures (Hernández-Sánchez et al., 2001;Yamada et al., 2001).
In the immature EC, spontaneous fps and large-scale propagating oscillatory calcium transients mediated by ionotropic glutamate (but not GABA) receptors have been reported (Jones and Heinemann, 1989;Garaschuk et al., 2000;Sheroziya et al., 2009;Namiki et al., 2013). At mature stages, mEC neurons can generate slow-wave network oscillations (Dickson et al., 2003), which are initiated by selective activation of GluR5 kainate receptors in a recurrent network, and terminated by activation of K ATP channels in active neurons (Cunningham et al., 2006). This example illustrates the tight coupling between neuronal activity and energy homeostasis. In the immature mEC, however, network oscillations are generated by different mechanisms, including the joined activation of both NMDA and AMPA/kainate receptors (Sheroziya et al., 2009). Moreover, in contrast to the observations in adult rats, in our hands the K ATP channel opener diazoxide suppressed, but never completely blocked fp-bursts. This finding indicates a differential impact of ATP-sensitive K + currents on immature network activity versus mature sleep-related slow-wave oscillations in the mEC. Importantly, it has been shown that LIII principal neurons are critically involved in the processes of initiation, propagation, termination, and reflection of synchronized traveling calcium waves in the immature EC (Namiki et al., 2013). This mechanism may be related to the ability of these neurons to generate prolonged intrinsic bursting activity at early postnatal stages.
The functional significance of prolonged intrinsic bursts in immature EC networks are still unknown. They may well contribute to the functional maturation of the EC and anatomically connected areas. Recurrent excitatory connections and electrical coupling between EC LIII pyramidal neurons have been demonstrated in adult rats using paired intracellular recordings (Dhillon and Jones, 2000). Intralaminar excitatory recurrent connections within LIII of juvenile rats have been recently shown using scanning photostimulation (Beed et al., 2010). Thus, activity of immature LIII neurons feeds back onto neurons within the same layer, and therefore may contribute to the activitydependent maturation of this local network in the EC. In addition, EC LIII neurons provide direct input to the apical dendrites of CA1 pyramids. Therefore, bursts of LIII neurons may induce dendritic spikes in CA1 pyramidal neurons that could then propagate to the soma and trigger action potentials (Jarsky et al., 2005). Strong firing of LIII neurons can also contribute to the distal dendritic enrichment of HCN1 channels in CA1 pyramids (Shin and Chetkovich, 2007). In adult animals, excitatory input from EC LIII to CA1 is essential for the formation of temporal association memory and fear learning (Suh et al., 2011;Kitamura et al., 2014;Lovett-Barron et al., 2014).
To conclude, our results show that activation of K ATP channels plays an important, limiting role for network activity and intrinsic neuronal firing in the immature mEC. This finding provides new insights into the mechanisms of activity-dependent maturation of neurons and networks in the EC. In addition, the key role of K ATP channels constitutes an important link between neuronal activity and neurometabolic state in this important area of the medial temporal lobe.