Similar nicotinic excitability responses across the developing hippocampal formation are regulated by small-conductance calcium-activated potassium channels

Abstract

The hippocampal formation forms a cognitive circuit that is critical for learning and memory. Cholinergic input to nicotinic acetylcholine receptors plays an important role in the normal development of principal neurons within the hippocampal formation. However, the ability of nicotinic receptors to stimulate principal neurons across all regions of the developing hippocampal formation has not been determined. We show in this study that heteromeric nicotinic receptors mediate direct inward current and depolarization responses in principal neurons across the hippocampal formation of the young postnatal mouse. These responses were found in principal neurons of the CA1, CA3, dentate gyrus, subiculum, and entorhinal cortex layer VI, and they varied in magnitude across regions with the greatest responses occurring in the subiculum and entorhinal cortex. Despite this regional variation in the magnitude of passive responses, heteromeric nicotinic receptor stimulation increased the excitability of active principal neurons by a similar amount in all regions. Pharmacological experiments found this similar excitability response to be regulated by small-conductance calcium-activated potassium (SK) channels, which exhibited regional differences in their influence on neuron activity that offset the observed regional differences in passive nicotinic responses. These findings demonstrate that SK channels play a role to coordinate the magnitude of heteromeric nicotinic excitability responses across the hippocampal formation at a time when nicotinic signaling drives the development of this cognitive brain region. This coordinated input may contribute to the normal development, synchrony, and maturation of the hippocampal formation learning and memory network.

NEW & NOTEWORTHY This study demonstrates that small-conductance calcium-activated potassium channels regulate similar-magnitude excitability responses to heteromeric nicotinic acetylcholine receptor stimulation in active principal neurons across multiple regions of the developing mouse hippocampal formation. Given the importance of nicotinic neurotransmission for the development of principal neurons within the hippocampal formation, this coordinated excitability response is positioned to influence the normal development, synchrony, and maturation of the hippocampal formation learning and memory network.

INTRODUCTION

The hippocampal formation (HF) forms a cognitive circuit that is critical for learning and memory (Eichenbaum 1992; Kesner et al. 1993; Scoville and Milner 1957; Squire and Zola 1996). The classical view of information flow through the constituent regions of the HF involves a series of excitatory glutamatergic synapses beginning in the entorhinal cortex layer II/III and connecting through principal neurons of the dentate gyrus (DG), the cornu ammonis area 3 (CA3), the cornu ammonis area 1 (CA1), the subiculum (SUB), and deep layers of the entorhinal cortex (EC) (Andersen 2007). Abnormal changes to the structure and function of the HF have been linked with several neurodevelopmental disorders that involve learning and memory, including autism spectrum disorders (ASD) (Nicolson et al. 2006; Schumann et al. 2004), attention-deficit hyperactivity disorder (ADHD) (Li et al. 2014; Plessen et al. 2006), epilepsy (Seidenberg et al. 2005; Squier et al. 2003; Sun and Goodkin 2016), and Down syndrome (Belichenko et al. 2004; Pinter et al. 2001; Witton et al. 2015). Cholinergic input from the medial septum-diagonal band of Broca (MSDB) to the HF plays an important role in modulating normal development of the HF (Chang and Berg 1999; Dutar et al. 1995; Frotscher and Léránth 1985; Liu et al. 2006; Lozada et al. 2012). This is mediated in part by acetylcholine (ACh) activation of nonselective ligand-gated cation channel nicotinic acetylcholine receptors (nAChRs), which modulate neurotransmission by γ-aminobutyric acid (GABA) (Alkondon et al. 1997; Jones and Yakel 1997; Maggi et al. 2001; McQuiston and Madison 1999; Sudweeks and Yakel 2000) and glutamate (Cheng and Yakel 2015; Ge and Dani 2005) to shape the formation of the HF cognitive network.

The two major isoforms of nAChR in the HF belong to the homomeric α7 and heteromeric α4β2* families (Alkondon and Albuquerque 2004; Jones and Yakel 1997; Séguéla et al. 1993; Sudweeks and Yakel 2000; Wada et al. 1989; Zarei et al. 1999). Previous work in rodents age postnatal day 2 (P2) and older demonstrates that functional nAChRs of both isoforms are present on hippocampal interneurons (Alkondon and Albuquerque 2004; Bell et al. 2011, 2015; Frazier et al. 1998; Ji and Dani 2000; Jones and Yakel 1997; Khiroug et al. 2003; Maggi et al. 2001; McQuiston and Madison 1999; Sudweeks and Yakel 2000). Although homomeric nAChRs have been demonstrated at P10 or older to mediate inward currents in glutamatergic principal neurons of the CA1, CA3, DG, and EC layer VI (ECVI) regions (Alkondon et al. 1997, 2007; Cheng and Yakel 2015; Grybko et al. 2011; John et al. 2015; Kalappa et al. 2010), the function of heteromeric nAChRs in principal neurons across the HF is not well understood. These receptors mediate currents in principal neurons of ECVI (Tu et al. 2009) as they do in layer VI principal neurons of other cortical regions (Tian et al. 2014), but studies in principal neurons of other HF regions such as CA1 report conflicting results (He et al. 2013; Hefft et al. 1999; Tu et al. 2009). We have recently demonstrated that heteromeric nAChR-mediated inward current and excitability responses in principal neurons of the mouse CA1 follow a distinct developmental pattern in which responses are greatest in magnitude during the first 2 wk of postnatal life and decline to low adult magnitudes shortly thereafter (Chung et al. 2016). This finding is consistent with molecular studies in which expression for heteromeric nAChR subunits also peaks in the hippocampus during the first 2 wk of postnatal life and declines significantly shortly thereafter (Didier et al. 1995; Machaalani et al. 2010; Shacka and Robinson 1998; Winzer-Serhan and Leslie 2005). This early postnatal expression of heteromeric nAChR subunits varies in magnitude across regions of the HF (Didier et al. 1995; Machaalani et al. 2010; Winzer-Serhan and Leslie 2005); however, to the best of our knowledge, the function of heteromeric nAChRs during early postnatal life has not been characterized for principal neurons across all regions of the HF.

The objective of the current study was to determine whether functional heteromeric nAChRs are present in principal neurons located within each region of the young postnatal mouse HF. Whole cell electrophysiological recordings were made of visually identified principal neurons located within the CA1, CA3, DG, SUB, and ECVI regions in acute brain slices of mice aged P5–P10. Heteromeric nAChR inward current and depolarization responses were identified in all regions of the HF, and the magnitude of these responses varied across the HF with greater responses in the SUB and ECVI than in the hippocampus proper (CA1, CA3, and DG). Interestingly, although this same heteromeric nAChR stimulation increased the action potential firing frequency in active neurons, the magnitude of this response was similar across all HF regions. Mechanistic experiments focused on small-conductance calcium-activated potassium (SK) channels, which generate a medium afterhyperpolarization potential (mAHP) to modulate action potential firing frequency (Gustafsson and Wigström 1983; Hotson and Prince 1980; Sah 1996). These experiments employed the SK channel blocker apamin (Hugues et al. 1982) to show that SK channels differentially influence the mAHP in principal neurons and mediate the similar heteromeric nAChR excitability responses observed in active principal neurons across regions of the HF. This finding demonstrates that SK channels play a role to coordinate the magnitude of nicotinic excitability responses across the HF at a time when nicotinic signaling drives HF neuron development, suggesting that this coordination contributes to the normal development, synchrony, and maturation of the HF learning and memory network.

MATERIALS AND METHODS

Experimental animals.

All animals were housed in a secure vivarium with an ambient temperature of 21–24°C and a 12-h reverse light cycle with lights on at 8:00 PM. Pregnant female CD1-strain mice were purchased from Charles River Canada (Saint-Constant, QC, Canada) and housed with ad libitum access to water and food as described previously (Chung et al. 2016). The day of birth for each litter was considered to be postnatal day 0 (P0), and male offspring were analyzed at P5–P10, which falls within the human third-trimester equivalent for brain development. All efforts were made to minimize animal suffering and to limit the number of mice used in this study. Experimental animals were cared for according to the principles and guidelines of the Canadian Council on Animal Care, and the experimental protocol was approved by the University of Guelph Animal Care Committee.

Electrophysiology.

Mice were anesthetized using isoflurane and killed by decapitation. Brains were immediately removed from the skull while submerged in 4°C sucrose artificial cerebrospinal fluid (ACSF; 254 mM sucrose, 10 mM d-glucose, 26 mM NaHCO₃, 2 mM CaCl₂, 2 mM MgSO₄, 3 mM KCl, and 1.25 mM NaH₂PO₄, pH 7.4) that had been oxygenated using carbogen (95% O₂ and 5% CO₂). Coronal slices containing the rostral/dorsal hippocampus or entorhinal cortex were cut at 400-μm thickness from approximately Bregma −1.46 to −2.18 mm or Bregma −2.92 to −3.52 mm, respectively, using a Leica VT1200 vibrating microtome (Leica Microsystems, Richmond Hill, ON, Canada) (Paxinos and Franklin 2001). Slices were transferred to oxygenated regular ACSF (128 mM NaCl, 10 mM d-glucose, 26 mM NaHCO₃, 2 mM CaCl₂, 2 mM MgSO₄, 3 mM KCl, and 1.25 mM NaH₂PO₄, pH 7.4) and maintained at 30°C in a recovery chamber for at least 2 h before use.

Slices were transferred to a modified recording chamber (Warner Instruments, Hamden, CT) that was mounted on the stage of an Axioskop FS2 Microscope (Carl Zeiss Canada, Toronto, ON, Canada) and superfused with oxygenated ACSF at a rate of 3–4 ml/min. Whole cell recording of CA1, CA3, SUB, DG, and ECVI principal neurons was performed at room temperature (21–24°C) using borosilicate glass pipette electrodes (resistance of 2–5 MΩ) filled with an internal solution containing 120 mM K-gluconate, 5 mM KCl, 2 mM MgCl₂, 4 mM K₂-ATP, 400 µM Na₂-GTP, 10 mM Na₂-phosphocreatine, 33 μM Alexa Fluor 488 hydrazide (Life Technologies, Burlington, ON, Canada), and 10 mM HEPES buffer (adjusted to pH 7.3 with KOH). Individual principal neurons were visualized using infrared differential interference contrast microscopy and selected for recording based on the location of the soma within the hippocampal principal cell layer or cerebral cortical layer. Specifically, CA1 and SUB regions were differentiated by the tightly packed principal cell layer of the CA1 compared with the diffuse principal cell layer of the SUB. CA3 neurons were located within principal cell layer before the hilus, and DG neurons were located within the suprapyramidal blade. ECVI principal neurons were located in the deep layer VI immediately lateral to the hippocampus. The morphology/type of each recorded neuron was verified during recordings with the use of epifluorescent visualization of the Alexa Fluor 488 that had diffused throughout the dendrite arbor from the recording pipette. All recordings were made using a Multiclamp 700B amplifier. Signals were acquired at 20 kHz and low-pass filtered at 2 kHz using a Digidata 1440A data acquisition system (Molecular Devices, Sunnyvale, CA). The liquid junction potential was corrected at the time of recording, and all data were analyzed post hoc using Clampfit 10.3 software (Molecular Devices).

All experiments were performed in the continuous presence of 200 nM atropine to block muscarinic acetylcholine receptors and 10 nM methyllycaconitine (MLA) to block α7-subunit-containing nAChRs. Basic electrophysiological properties were assessed in current-clamp mode by measuring responses to the injection of positive and negative current steps for 500 ms each. Nicotinic receptor-mediated responses were probed by the addition of ACh to the ACSF superfusion bath after a baseline recording period and were followed by a 5-min washout period. Inward current responses were assessed in voltage-clamp mode with neurons held at −75 mV by subtracting the mean holding current at the peak of the ACh response from the mean holding current at baseline. Depolarizing responses from rest were measured in current-clamp mode by subtracting the mean membrane potential at the peak of the ACh response from the resting membrane potential at baseline. Acceleration of action potential firing was measured in current-clamp mode by first injecting a positive current to elicit an ~1-Hz baseline firing frequency. The rate of firing was measured over a 30-s period during the peak ACh response and reported as a percent increase in frequency from the ~1-Hz baseline frequency for each neuron, using the equation (firing frequency at the peak ACh response – firing frequency at baseline)/(firing frequency at baseline) × 100. Additional pharmacological experiments were performed using the α4β2* nAChR competitive antagonist dihydro-β-erythrodine (DHβE), which was applied to the bath at 3 µM for 10 min before ACh application. For a subset of neurons, input-output curves were generated by applying 500-ms depolarizing current steps from rest in 50-pA increments. The magnitude of postburst mAHP responses was assessed following trains of action potentials that were elicited by injecting brief depolarizing current pulses (2 nA, 2 ms) at 50 Hz. Peak mAHP magnitude and timing were measured at the lowest membrane potential observed during the 50- to 1,000-ms mAHP period after of the last action potential was elicited. Peak mAHP magnitude was calculated for each neuron as the difference between its resting membrane potential and this lowest membrane potential measured during the mAHP period. For a subset of experiments, the SK channel blocker apamin was applied to slices at 100 nM for 10 min before nicotinic or mAHP responses were measured. ACh chloride and atropine were purchased from Sigma Aldrich (Oakville, ON, Canada). MLA citrate, DHβE, and apamin were purchased from Tocris Bioscience/Bio-Techne (Minneapolis, MN). All drugs were stored in stock solutions at −20°C.

Statistical analysis.

All data are reported as means ± SE for neurons within each experimental group. Statistical analyses were performed using GraphPad Prism 7 (GraphPad Software, La Jolla, CA), and a level of P < 0.05 was required to indicate statistical significance. Basic electrophysiological properties and ACh-induced nicotinic current, depolarization, and excitability responses were assessed across regions using one-way ANOVA followed by Tukey’s post hoc test. The effect of DHβE on ACh-induced nicotinic excitability responses was assessed within each region using two-tailed paired t-tests. Comparison of input-output curves was assessed using two-way ANOVA followed by Tukey’s post hoc test at each magnitude of current injection. Effects of apamin on mAHP magnitude were determined using the two-tailed paired t-test within each region, and assessment of mAHP timing across HF regions was performed using one-way ANOVA followed by Tukey’s post hoc test. The effect of apamin on ACh-induced nicotinic excitability responses was assessed using two-way ANOVA followed by the Bonferroni post hoc test within each region.

RESULTS

Basic electrophysiological properties of principal neurons across the hippocampal formation.

Basic electrophysiological properties for principal neurons in the HF regions are presented in Table 1. To the best of our knowledge, a comparison of these properties across regions of the young postnatal HF has not been reported previously. Resting membrane potential was affected by region [one-way ANOVA, F(4,125) = 7.0, P < 0.0001] and was lower in the DG than in each of the other regions (Tukey’s post hoc test, P ≤ 0.04 for each pairwise comparison). Input resistance was similarly affected by region [F(4,113) = 36.7, P < 0.0001], with resistance in DG being higher than in each of the other regions (P < 0.0001 for each pairwise comparison). Spike amplitude was also affected by region [F(4,125) = 3.3, P = 0.01], and this value was higher in ECVI than in DG (P = 0.04).

Table 1.

Electrophysiological properties for principal neurons of the young postnatal hippocampal formation

	CA1	CA3	DG	SUB	ECVI	P Value (One-Way ANOVA)
Resting membrane potential, mV	−70.8 ± 0.8	−74.5 ± 1.1	−79.0 ± 1.7*	−72.0 ± 1.0	−74.3 ± 0.8	<0.0001
Input resistance, MΩ	343.4 ± 18.6	353.1 ± 37.2	826.4 ± 79.1*	186.4 ± 14.5	215.0 ± 22.2	<0.0001
Spike amplitude, mV	86.9 ± 1.7	83.1 ± 1.4	82.4 ± 2.7†	89.0 ± 2.1	89.8 ± 1.2†	0.01

Values are means ± SE for principal neurons of the cornu ammonis area 1 (CA1; n = 21 neurons in 13 mice), cornu ammonis area 3 (CA3; n = 30 neurons in 12 mice), dentate gyrus (DG; n = 27 neurons in 9 mice), subiculum (SUB; n = 23 neurons in 10 mice), and layer VI of entorhinal cortex (ECVI; n = 28 neurons in 11 mice).

^*P < 0.04 indicates a statistical difference for resting membrane potential and input resistance compared with each of the other regions (Tukey’s post hoc test).

^†P = 0.04 indicates a statistical difference for spike amplitude between the DG and ECVI only (Tukey’s post hoc test).

Nicotinic inward currents in principal neurons of the hippocampal formation.

The subunits that comprise heteromeric α4β2* nAChRs are expressed in the rodent brain as early as embryonic day 18, and several studies suggest that the content of these subunits peaks within the hippocampus during early postnatal life before declining to lower adulthood levels shortly thereafter (Shacka and Robinson 1998; Winzer-Serhan and Leslie 2005; Zhang et al. 1998). Consistent with these expression studies, we have recently demonstrated that heteromeric, putative α4β2* nAChRs mediate excitatory responses in hippocampus CA1 principal neurons, with the greatest magnitude of responses also occurring during early postnatal life (Chung et al. 2016). In the present study, we first sought to determine whether functional heteromeric nAChRs are present on additional principal neurons of the HF during early postnatal life by probing for ACh-induced inward current responses in principal neurons sampled from the CA1, CA3, DG, SUB, and ECVI regions of the HF. Acute brain slices containing these regions were prepared from mice aged P5–P10, and heteromeric nAChRs were isolated pharmacologically in all experiments by the continuous bath application of 200 nM atropine (to block muscarinic acetylcholine receptors) and 10 nM MLA (to block α7-subunit-containing nAChRs). Endogenous acetylcholinesterase in mouse cortical and hippocampal brain slices metabolizes ACh as it washes into the slice preparation such that the effective ACh concentration reaching nAChRs is ~10- to 100-fold lower than the concentration applied in the ACSF bath (Bailey et al. 2010; Chung et al. 2016). We therefore applied ACh at 1 mM to facilitate comparison of our data with that from cell culture and synaptosome studies. Application of 1 mM ACh (15 s) resulted in a positive inward current response in every principal neuron that was recorded, although there was a significant effect of HF region on the magnitude responses observed [Fig. 1A; one-way ANOVA, F(4,97) = 16.2, P < 0.0001]. Nicotinic current responses were significantly greater in SUB (30.0 ± 9.1 pA, n = 18 neurons from 8 mice) and ECVI (37.5 ± 2.9 pA, n = 24 neurons from 11 mice) than in CA1 (8.6 ± 1.1 pA, n = 18 neurons from 12 mice), CA3 (5.5 ± 0.9 pA, n = 25 neurons from 12 mice), and DG (3.2 ± 0.9 pA, n = 17 neurons from 9 mice) (Tukey’s post hoc test, P ≤ 0.005 for each pairwise comparison). Previous investigation into the role of nAChRs in learning and memory has focused primarily within the hippocampus proper, where glutamatergic principal neurons of the CA1, CA3, and DG form the trisynaptic circuit (Andersen 2007). Analysis only within these three subregions of the hippocampus proper demonstrated that the magnitude of nicotinic inward currents is significantly affected by subregion [Fig. 1B; F(2,57) = 6.7, P = 0.002], with currents in CA1 being greater than those in DG (P = 0.002). Typical inward current responses are shown for one neuron from each region of the HF in Fig. 1C.

Fig. 1.

The magnitude of nicotinic inward current responses varies across regions of the hippocampal formation (HF). A: inward current responses to 1 mM ACh (15 s) were significantly affected by region (1-way ANOVA, P < 0.0001), where currents were greater in the subiculum (SUB) and layer VI of entorhinal cortex (ECVI) compared with the cornu ammonis area 1 (CA1), cornu ammonis area 3 (CA3), and dentate gyrus (DG) (Tukey’s post hoc test, ***P ≤ 0.005 for each pairwise comparison). Values are means + SE. B: analysis of neurons within the hippocampus proper, which form the trisynaptic circuit, also demonstrated a significant effect of subregion (P = 0.002), with nicotinic inward current responses in CA1 being greater than those in DG (Tukey’s post hoc test, ***P = 0.002). Values are means + SE. C: typical voltage-clamp traces are shown for 1 neuron in each region with the 15-s ACh application indicated by the gray horizontal bar. All recordings were made in the continuous presence of 200 nM atropine and 10 nM methyllycaconitine (MLA).

Nicotinic depolarization and excitability in principal neurons of the hippocampal formation.

We next sought to assess consequences of the observed nicotinic inward currents on a passive neuronal response in principal neurons of the HF by measuring depolarization from rest. Application of 1 mM ACh (15 s) resulted in current-clamp depolarization responses in every neuron tested, and the magnitude of these responses was significantly affected by HF region [Fig. 2A; one-way ANOVA, F(4,78) = 18.2, P < 0.0001]. Tukey’s post hoc test demonstrated that the magnitude of depolarization was significantly greater in principal neurons of ECVI (17.2 ± 1.4 mV, n = 15 neurons from 9 mice) than CA1 (6.4 ± 0.9 mV, n = 17 neurons from 12 mice), CA3 (5.3 ± 1.1 mV, n = 19 neurons from 9 mice), DG (2.8 ± 0.5 mV, n = 12 neurons from 7 mice), and SUB (8.9 ± 1.5 mV, n = 20 neurons from 9 mice) (P ≤ 0.0001 for each comparison) and significantly greater in principal neurons of SUB than in DG (P = 0.01). Analysis of responses within the hippocampus proper (CA1, CA3, and DG) revealed that although there was no significant effect of subregion [Fig. 2B; F(2,45) = 3.0, P = 0.06], the nicotinic depolarization response was greater in CA1 than in DG (P = 0.0496). Typical depolarization responses are shown for one neuron from each region of the HF in Fig. 2C. It should be noted that the regional pattern for measured nicotinic depolarization responses across the HF in Fig. 2A matches the anticipated regional pattern calculated using Ohm’s law (∆V = ∆I × R, where V is voltage, I is current, and R is resistance), using the measured inward current responses from Fig. 1A and the measured input resistance for each neuron from Table 1. The passive response to heteromeric nAChR stimulation of principal neurons across the HF thus relates directly to the inward current elicited in each neuron.

Fig. 2.

The magnitude of nicotinic depolarization from rest varies across regions of the HF. A: depolarization responses to 1 mM ACh (15 s) were significantly affected by region (1-way ANOVA, P < 0.0001), where depolarization was greater in ECVI than in all other regions (Tukey’s post hoc test, ***P ≤ 0.0001 for each pairwise comparison) and greater in SUB than in DG (Tukey’s post hoc test, **P = 0.01). Values are means + SE. B: analysis of neurons within the hippocampus proper, which form the trisynaptic circuit, did not demonstrate a significant effect of subregion (P = 0.06), although the depolarization response was greater in CA1 than in DG (Tukey’s post hoc test, *P = 0.0496). Values are means + SE. C: typical current-clamp traces are shown for 1 neuron in each region with the 15-s ACh application indicated by the gray horizontal bar. All recordings were made in the continuous presence of 200 nM atropine and 10 nM MLA.

Our next line of experiments aimed to determine whether an active excitability response to heteromeric nAChR stimulation in these neurons was proportional to inward current and/or depolarization responses. We measured excitability responses of active neurons in current-clamp mode by applying 1 mM ACh (15 s) to neurons that had been induced previously to fire action potentials at a frequency of 1 Hz by positive current injection and measuring the percent increase in action potential firing frequency at the peak of the ACh response relative to the ~1-Hz baseline for each neuron. In stark contrast with the regional differences observed above for inward current and passive depolarization responses, neurons in all regions showed similar active excitability responses to ACh that were not significantly affected by HF region [Fig. 3A; one-way ANOVA, F(4,78) = 0.5, P = 0.7]. The percentages by which ACh accelerated firing over baseline (Fig. 3A, right y-axis) were as follows: CA1, 196.7 ± 33.9% (n = 12 neurons from 8 mice); CA3, 281.2 ± 48.5% (n = 21 neurons from 10 mice); DG, 230.7 ± 43.4% (n = 18 neurons from 7 mice); SUB, 257.3 ± 37.4% (n = 17 neurons from 6 mice), and ECVI, 261.0 ± 33.0% (n = 15 neurons from 10 mice). These same data expressed for each neuron relative to the mean for CA1 principal neurons (Fig. 3A, left y-axis) show that the ACh-induced acceleration of action potential firing was 1.4 ± 0.2 times as strong in CA3 vs. CA1, 1.2 ± 0.2 times as strong in DG vs. CA1, 1.3 ± 0.2 times as strong in SUB vs. CA1, and 1.3 ± 0.2 times in ECVI vs. CA1. Typical traces showing the ACh-induced increase of action potential firing are shown for one neuron from each region of the HF in Fig. 3B. We also sought to determine whether these ACh-induced excitatory responses are mediated by α4β2* nAChRs by repeating this experiment in the absence and presence of the α4β2* nAChR competitive antagonist DHβE (3 μM, 10-min preexposure). Data presented in Table 2 demonstrate that DHβE decreased the nicotinic excitability response in neurons from each region of the HF (paired t-test, P = 0.03 to 0.0006 for each region). Moreover, one-way ANOVA confirmed no regional difference in the magnitude of responses either at baseline [F(4,20) = 1.0, P = 0.4] or in the presence of DHβE [F(4,20) = 1.2, P = 0.3]. These results demonstrate that, in contrast with the varied inward current and passive depolarization responses to heteromeric nAChR stimulation across regions of the HF, the active response to this same stimulation in firing neurons is quite similar across all regions of the HF.

Fig. 3.

The ability of nicotinic stimulation to increase excitability in active neurons is similar across regions of the HF. A: for principal neurons that had been induced to fire action potentials at 1 Hz by positive current injection, the increase of firing frequency from the 1-Hz baseline following application of 1 mM ACh (15 s) was not significantly affected by region (1-way ANOVA, P = 0.7). Values are means + SE. B: typical current-clamp traces are shown for 1 neuron in each region and represent typical changes in firing frequency following application of ACh (indicated by gray horizontal bar). All recordings were made in the continuous presence of 200 nM atropine and 10 nM MLA.

Table 2.

Effect of DHβE on nicotinic stimulation of active principal neurons of the young postnatal hippocampal formation

	Increase from Baseline, %
Region	Before DHβE	After DHβE	P Value (Paired t-Test)
CA1	273.5 ± 48.4	66.6 ± 21.6	0.03
CA3	356.4 ± 93.3	151.2 ± 57.4	0.02
DG	223.1 ± 35.4	70.2. ± 25.1	0.004
SUB	364.4 ± 89.9	183 ± 73.0	0.03
ECVI	373.3 ± 44.1	157.9 ± 47.5	0.0006

Values are mean ± SE for principal neurons of the CA1 (n = 5 neurons from 4 mice), CA3 (n = 5 neurons from 3 mice), DG (n = 5 neurons from 4 mice), SUB (n = 5 neurons from 4 mice), and ECVI (n = 5 neurons from 4 mice).The ability of ACh (1 mM) to increase action potential firing frequency was measured in principal neurons before and after exposure to dihydro-β-erythrodine (DHβE; 3 μM, 10 min). All recordings were made in the continuous presence of 200 nM atropine and 10 nM methyllycaconitine (MLA).

Capacity for excitation in principal neurons of the young postnatal hippocampal formation.

Given that principal neurons experience varying magnitude of nicotinic inward current and depolarization responses across the young postnatal HF, yet show very similar nicotinic increases to action potential firing, we generated input-output curves to determine whether this resulted from similar upper excitation limits for these neurons that were attained by the nicotinic inward currents. Input-output curves were generated by injecting positive current steps from 50 to 200 pA in 50-pA increments and for 500 ms each. The resulting curves for each HF region are shown in Fig. 4A, and typical traces for one neuron from each HF region in response to 50- and 150-pA current steps are shown in Fig. 4B. Comparing all regions together, two-way ANOVA revealed significant main effects of current injected [F(4,415) = 140.4, P < 0.0001] and HF region [F(4,415) = 13.8, P < 0.0001] and a significant interaction between these two factors [F(16,415) = 2.1, P = 0.008]. The lowest 50-pA current step resulted in a greater action potential firing frequency in principal neurons in CA1 (17.5 ± 2.7 Hz, n = 13 neurons from 6 mice) than in CA3 (8.8 ± 1.1 Hz, n = 17 neurons from 7 mice), SUB (8.6 ± 1.3 Hz, n = 32 neurons from 9 mice), and ECVI (8.2 ± 1.7 Hz, n = 12 neurons from 7 mice) (Tukey’s post hoc test, P ≤ 0.009 for each pairwise comparison). At the 200-pA current step, which resulted in maximal firing frequency for each neuron type, principal neurons in both CA1 (27.4 ± 3.7 Hz, n = 13 neurons from 6 mice) and SUB (26.6 ± 1.5 Hz, n = 32 neurons from 9 mice) showed a greater action potential firing frequency than in ECVI (18.5 ± 2.9 Hz, n = 12 neurons from 7 mice) (P ≤ 0.02 for each pairwise comparison). The application of depolarizing current steps beyond 200 pA elicited decreasing mean action potential frequencies across all HF regions, with some neurons presenting overexcitation depolarization block. These data clearly show that although principal neurons in all regions of the early postnatal HF are capable of firing at frequencies well beyond the 2.7- to 3.8-Hz absolute frequency observed at the peak of the nicotinic response in Fig. 3, the maximum firing frequency is lower in ECVI compared with other HF regions.

Fig. 4.

Input-output curves for principal neurons of the young postnatal HF. A: input-output curves were generated by measuring action potential firing frequency in response to the injection of positive current steps from 50 to 200 pA in 50-pA increments and for 500 ms each. The resulting pattern of firing frequency was significantly affected by the amount of current injected (2-way ANOVA, P < 0.0001) and region (2-way ANOVA, P < 0.0001), and there was a significant interaction between these two main effects (2-way ANOVA, P = 0.008). Following 50-pA injection, firing frequency was higher in CA1 than in CA3, SUB, and ECVI (Tukey’s post hoc test, P ≤ 0.009 for each pairwise comparison). Maximal firing for each region was observed following 200-pA injection, where firing frequency was higher in both CA1 and SUB than in ECVI (P ≤ 0.02 for each pairwise comparison). Values are means ± SE. B: typical current-clamp traces demonstrating action potentials elicited following injection of 50- and 150-pA current are displayed for 1 neuron from each region of the hippocampal formation.

We also performed a theoretical experiment by approximating the total amount of inward current received by neurons at the peak of the nicotinic response in Fig. 3 by adding 1) the amount of current required to elicit baseline firing at ~1 Hz in the experiment from Fig. 3 plus 2) the mean nicotinic inward current received by each neuron type in Fig. 1. This resulted in the following theoretical total inward currents: CA1, 25.2 ± 2.0 pA (n = 11 neurons from 3 mice); CA3, 19.2 ± 1.1 pA (n = 14 neurons from 2 mice); DG, 12.9 ± 1.7 pA (n = 12 neurons from 2 mice); SUB, 49.2 ± 9.9 (n = 11 neurons from 3 mice), and ECVI, 66.7 ± 4.1 pA (n = 11 neurons from 4 mice). With the use of this theoretical total current and extrapolation from the input-output curves in Fig. 4, this experiment suggests that ACh-induced nicotinic currents applied in a 500-ms step to neurons at rest would result in varying firing frequencies between 15% and 60% of the maximum firing frequency for each principal neuron type across the HF. This contrasts with results from Fig. 3 demonstrating that actual ACh-induced nicotinic currents received by active neurons lead to very similar firing frequencies, between 10% and 20% of the maximum firing frequency for each principal neuron type across the HF. We next focused on mechanisms that regulate excitability in active neurons.

SK channels are widely expressed throughout the brain and regulate neuron excitability following the firing one or more action potentials (Sah 1996). They are activated by a rise in intracellular calcium during the action potential and mediate potassium efflux contributing to the generation of an afterhyperpolarization of medium duration (mAHP) (Nicoll 1988; Schwindt and Crill 1998). The mAHP activates within milliseconds and decays with a time constant in the range of hundreds of milliseconds, until the onset of the slow AHP, which is active from one to several seconds (Faber 2009; Lancaster and Adams 1986; Lancaster and Zucker 1994; Matthews et al. 2009). Because SK channels are expressed in HF neurons that receive glutamatergic neurotransmission (Babiec et al. 2017; Lappin et al. 2005; Mateos-Aparicio et al. 2014; Ngo-Anh et al. 2005) and we have observed regional differences across the HF for action potential firing frequency in response to low and high current injection, we next determined whether SK channels differentially regulate the mAHP in principal neurons of the young postnatal HF. This was accomplished by measuring mAHP magnitude within single neurons before and after application of the SK channel blocker apamin at 100 nM for 10 min. Action potentials were elicited from rest by applying brief depolarizing current pulses (trains of 8 or 16 pulses of 2 nA for 2 ms each at 50 Hz) and measuring peak mAHP magnitude and timing at the lowest membrane potential observed during the 50- to 1,000-ms mAHP period after the last action potential was elicited. Peak mAHP magnitude was calculated for each neuron as the difference between its resting membrane potential and this lowest membrane potential measured during the mAHP period. Data for peak mAHP magnitude and representative current-clamp traces are shown in Fig. 5 for each neuron type before (baseline) and after apamin treatment following trains of 8 (Fig. 5A) and 16 pulses (Fig. 5B). In CA1 principal neurons, peak mAHP magnitude for 8 current pulses was −0.6 ± 0.6 mV at baseline and 0.2 ± 0.8 mV after apamin treatment (two-tailed paired t-test, n = 6 neurons from 3 mice, P = 0.4), and for 16 current pulses was −1.1 ± 0.5 mV at baseline and −0.9 ± 0.6 mV after apamin treatment (n = 6 neurons from 3 mice, P = 0.8). Apamin also did not affect peak mAHP magnitude in SUB principal neurons, which for 8 current pulses was −0.004 ± 0.4 mV at baseline and 1.3 ± 0.8 mV after apamin treatment (n = 6 neurons from 3 mice, P = 0.2), and for 16 current pulses was −0.7 ± 0.4 mV at baseline and 0.4 ± 0.2 mV after apamin treatment (n = 6 neurons from 3 mice, P = 0.07). In ECVI principal neurons, however, apamin treatment did significantly reduce peak mAHP amplitude. Peak mAHP was reduced for 8 current pulses from −0.6 ± 0.2 mV at baseline to 0.3 ± 0.4 mV after apamin treatment (n = 6 neurons from 4 mice, P = 0.007) and was reduced for 16 current pulses from −1.3 ± 0.4 mV at baseline to 0.2 ± 0.2 mV after apamin treatment (n = 6 neurons from 4 mice, P = 0.01).

Fig. 5.

The medium afterhyperpolarization (mAHP) is differentially affected by the SK channel blocker apamin in principal neurons of the young postnatal HF. Trains of 8 (A) or 16 (B) action potentials were elicited from rest by applying brief depolarizing current pulses (2 nA, 2 ms) at 50 Hz and measuring the peak mAHP magnitude during the 50- to 1,000-ms mAHP period after the last action potential was elicited, relative to resting membrane potential. Data for this mAHP magnitude are shown in histogram plots for each neuron type at baseline (open bars) and following application of 100 nM apamin for 10 min (red bars). Current-clamp traces are shown below each histogram plot for a typical neuron at baseline (black trace) and after apamin administration (red trace), with the timing for the peak mAHP magnitude at baseline indicated by the arrow. Apamin treatment increased mAHP magnitude in DG neurons following 8 current pulses (2-tailed paired t-test, *P = 0.05) and decreased mAHP magnitude in ECVI neurons following both 8 (2-tailed paired t-test, **P = 0.007) and 16 current pulses (2-tailed paired t-test, *P = 0.01). There was a significant effect of HF region on peak mAHP time at baseline following 8 (1-way ANOVA, P = 0.0003) and 16 (1-way ANOVA, P = 0.002) current pulses. All values are means + SE.

We found that in young postnatal CA3 and DG neurons, membrane potential following multiple action potentials did not return to rest within the 50- to 1,000-ms time frame for the mAHP. As such, measured peak mAHP magnitudes at baseline were positive relative to the resting membrane potential. Apamin did not affect peak mAHP magnitude in CA3 principal neurons, which for 8 current pulses was 2.6 ± 1.2 mV at baseline and 1.9 ± 0.6 mV after apamin treatment (n = 5 neurons from 2 mice, P = 0.6), and for 16 current pulses was 1.3 ± 1.1 mV at baseline and 2.0 ± 0.7 mV after apamin treatment (n = 5 neurons from 2 mice, P = 0.6). Interestingly, apamin treatment increased peak mAHP magnitude in DG principal neurons. Peak mAHP was increased for 8 current pulses from 3.4 ± 0.7 mV at baseline to 0.4 ± 1.24 mV after apamin treatment (n = 6 neurons from 2 mice, P = 0.05) and was increased for 16 current pulses from 1.9 ± 0.7 mV at baseline to 0.3 ± 1.0 mV after apamin treatment (n = 6 neurons from 2 mice, P = 0.07).

The timing for the peak mAHP magnitude at baseline varied across regions of the HF, as indicated by arrows in Fig. 5, with peaks generally occurring earlier for CA1 and SUB than in the CA3, DG, and ECVI. One-way ANOVA for peak times after 8 current pulses reveals a significant effect of region [F(4,24) = 8.1, P = 0.0003]. mAHP peak time occurred earlier in CA1 and ECVI than in CA3 (Tukey’s post hoc test, P = 0.02 for each comparison) and earlier in CA1, SUB, and ECVI than in DG (P < 0.01 for each comparison). The peak mAHP time was also affected by region following 16 current pulses [F(4,24) = 5.7, P = 0.002], with the peak occurring earlier in CA1, SUB, and ECVI than in DG (P < 0.05 for each comparison). Results from this experiment using apamin suggest that SK channels increase mAHP magnitude in ECVI neurons and decrease mAHP magnitude in DG neurons to differentially regulate firing frequencies for principal neurons of the young postnatal HF.

SK channels differentially regulate nicotinic excitability in active principal neurons of the young postnatal hippocampal formation.

Given that SK channels differentially regulate the mAHP across regions of the young postnatal HF, primarily to increase mAHP magnitude in ECVI, we next sought to determine whether SK channels influence the magnitude of ACh-induced nicotinic excitation responses in active neurons within this cognitive circuit. The experiment from Fig. 3 was repeated by measuring the percent increase in firing frequency following application of 1 mM ACh (15 s) to neurons that had been induced by positive current injection to fire action potentials at a baseline of ~1 Hz. We compared results from Fig. 3 with those from this new experiment, which was performed following blockade of SK channels by exposing slices to 100 nM apamin for 10 min. To directly compare results from this new experiment and results from Fig. 3, the percentage by which ACh increased firing frequency was normalized relative to the mean for CA1 neurons within each experiment. Results are shown in Fig. 6A, with data from Fig. 3A represented by open bars and new data following apamin treatment represented by red bars. There was a significant main effect of region [two-way ANOVA, F(4, 104) = 8.0, P < 0.0001] but no main effect of apamin treatment [F(1, 104) = 0.3, P = 0.6] on relative ACh-induced acceleration of firing, although there was a significant interaction between these two variables [F(4, 104) = 7.0, P < 0.0001]. Bonferroni’s post hoc test revealed a significant effect of apamin treatment to increase the ACh-induced acceleration of firing in ECVI neurons compared with untreated ECVI neurons (P < 0.0001). However, apamin treatment did not significantly affect the magnitude of the ACh response in neurons of CA1 (P = 1.0), CA3 (P = 1.0), DG (P = 0.1), or SUB (P = 1.0). One-way ANOVA within each treatment group showed that although there was no effect of HF region on ACh-induced acceleration of firing in untreated neurons (as per data for Fig. 3 described above), a significant effect of HF region emerged following apamin treatment [F(4,26) = 6.6, P = 0.0009]. Tukey’s post hoc test demonstrated that after apamin treatment, relative ACh-induced acceleration of firing was significantly greater in principal neurons of ECVI than CA1 (P = 0.02), CA3 (P = 0.007), DG (P = 0.0004), and SUB (P = 0.02). One-way ANOVA within the hippocampus proper after apamin treatment showed an effect of subregion [F(2,16) = 3.9, P = 0.04], with relative ACh-induced acceleration of firing being greater in CA1 than DG (P = 0.04). It is most interesting that this pattern of regional differences for the magnitude of ACh-induced acceleration of firing after apamin treatment in this experiment (red bars in Fig. 6A) is strikingly similar to the pattern of regional differences for ACh-induced depolarization from rest as shown in Fig. 2. These findings suggest that SK channel regulation of neuron excitability normalizes nicotinic responses in active neurons by increasing the relative response in DG and limiting the relative response in ECVI, with an end result to facilitate similar nicotinic excitability responses in active principal neurons across the young postnatal HF.

Fig. 6.

The ability of nicotinic stimulation to increase excitability in active neurons is differentially regulated by SK channels across regions of the HF. A: principal neurons were induced to fire action potentials at 1 Hz by positive current injection, and the increase in firing frequency in response to 1 mM ACh (15 s) was measured. The magnitude of this nicotinic response was normalized relative to CA1 for neurons that were not treated (open bars) and for neurons that had been treated with 100 nM apamin for 10 min (red bars). ACh-induced acceleration of firing frequency was affected by HF region (2-way ANOVA, P < 0.0001) but not by apamin treatment (2-way ANOVA, P = 0.6), and there was a significant interaction between these two variables (2-way ANOVA, P < 0.0001). Apamin treatment increased the relative response to ACh in ECVI only (Bonferroni’s post hoc test, ^†P < 0.0001). Analysis of each treatment group separately showed no effect of HF region in untreated neurons (1-way ANOVA, P = 0.7) and a significant effect of HF region after apamin treatment (1-way ANOVA, P = 0.0009), where the nicotinic response was greater in ECVI than in CA1, CA3, DG, and SUB (Tukey’s post hoc test, P ≤ 0.02 for each comparison). Analysis of apamin-treated neurons within the hippocampus proper also demonstrated a significant effect of subregion (P = 0.04) with greater nicotinic responses in CA1 than in DG (Tukey’s post hoc test, P = 0.04). Values are as means + SE. B: typical current-clamp traces are shown for one neuron from each region after apamin treatment, which represent typical changes in firing frequency following application of ACh (indicated by gray bar). All recordings were made in the continuous presence of 200 nM atropine and 10 nM MLA.

DISCUSSION

This study provides evidence that glutamatergic principal neurons in the CA1, CA3, DG, SUB, and ECVI regions of the HF are excited by heteromeric nAChRs during early postnatal development. SK channels regulate this excitability response, resulting in similar changes to action potential firing frequency elicited by nicotinic input across regions of the HF. These findings were obtained through a series of physiological and pharmacological experiments performed with the use of whole cell electrophysiological recording of visually identified principal neurons of each HF region within acute hippocampal brain slices collected from mice aged P5–P10. We first found that heteromeric nAChR stimulation elicits direct inward current and depolarization responses in principal neurons of all HF regions investigated and that the magnitude of these passive responses is greater in the SUB and ECVI than in the hippocampus proper regions CA1, CA3, and DG. Interestingly, this same heteromeric nAChR stimulation applied to active principal neurons increased action potential firing frequency in a manner that was similar across all HF regions. Mechanistic experiments demonstrated that SK channels differentially influence mAHP magnitude in principal neurons of the HF, which appears to offset the varied nicotinic inputs to produce the similar excitability responses observed across regions of the HF. The comprehensive characterization of heteromeric nAChR function presented in this study demonstrates that SK channels regulate the magnitude of nicotinic excitability responses in active principal neurons during a developmental period when nicotinic signaling drives HF neuron development. This regulation of excitability by SK channels may therefore contribute to normal development, synchrony, and maturation of the HF learning and memory network.

Heteromeric nicotinic receptor signaling within the developing hippocampal formation.

Afferent cholinergic neurotransmission from the MSDB plays an important role in the normal development and mature function of the HF (Berger-Sweeney et al. 2001; Chang and Berg 1999; Dutar et al. 1995; Frotscher and Léránth 1985; Mitsushima et al. 2013; Nott and Levin 2006; Placzek et al. 2009). Functional nAChRs have been identified on GABAergic interneurons (Alkondon and Albuquerque 2004; Bell et al. 2011, 2015; Frazier et al. 1998; Ji and Dani 2000; Jones and Yakel 1997; Khiroug et al. 2003; McQuiston and Madison 1999; Sudweeks and Yakel 2000) and on glutamatergic principal neurons (Alkondon et al. 1997, 2007; Cheng and Yakel 2015; Grybko et al. 2011; He et al. 2013; Kalappa et al. 2010; Tu et al. 2009) of the HF. We provide evidence in this current study that functional heteromeric, putative α4β2* nAChRs are present on principal neurons within the CA1, CA3, DG, SUB, and ECVI regions of the HF during a period equivalent to the human third trimester for brain development (Dobbing and Sands 1979). This is consistent with previous in situ reports that α4-, β2-, and α5-subunits are expressed within hippocampus principal neurons (Heath et al. 2010; Marks et al. 1992; Salas et al. 2003; Sudweeks and Yakel 2000; Wada et al. 1989; Winzer-Serhan and Leslie 2005). Nicotinic responses to bath-applied ACh in our slice preparation were likely not mediated by homomeric α7 nAChRs, because this isoform desensitizes rapidly (Quick and Lester 2002) and all experiments were conducted in the continuous presence of the α7-subunit-containing nAChR antagonist MLA (Macallan et al. 1988; Murray et al. 2012). We have recently reported, using the same experimental design as this current study, that heteromeric nAChRs mediate postsynaptic nicotinic signaling in young postnatal CA1 principal neurons and that these responses are inhibited by the α4β2* nAChR-selective antagonist DHβE (Chung et al. 2016). This is consistent with findings from this current study, which show statistically significant inhibition of nicotinic responses by DHβE in principal neurons of all HF regions examined. These previous findings, in combination with the findings of the current study, suggest that heteromeric nAChRs of the α4β2* isoforms mediate excitation of principal neurons within CA1, CA3, DG, SUB, and ECVI during early postnatal development.

Previous electrophysiological studies in rodents aged P10 and older demonstrate nicotinic responses in ECVI principal neurons (Tu et al. 2009) and either no direct nicotinic responses or nicotinic responses that occur in a low proportion of CA1, CA3, DG, and SUB principal neurons (He et al. 2013; Hefft et al. 1999; McQuiston and Madison 1999; Tu et al. 2009). Our findings in younger mice aged P5–P10 demonstrate the greatest heteromeric nAChR function in SUB and ECVI principal neurons but show that heteromeric nAChRs also mediate direct inward currents and depolarization from rest in principal neurons of the CA1, CA3, and DG at this age. Previous work suggests that endogenous cholinergic signaling plays a role in determining the pattern of glutamatergic synapse and network formation during development. For example, nicotinic stimulation of immature CA3 principal neurons or local astrocytes in young rats can convert silent synapses to a functional status, resulting in enhanced glutamatergic synaptic transmission through the CA3-CA1 Schaffer collateral pathway (Maggi et al. 2001; Wang et al. 2013). Nicotinic signaling also influences the morphological growth and maturation of neurons within the central nervous system (Campbell et al. 2010; Lipton et al. 1988; Liu et al. 2006; Pugh and Berg 1994), and β2-subunit-containing nAChRs specifically are important for the normal development of dendritic spines in principal neurons of the cerebral cortex (Ballesteros-Yáñez et al. 2010) and hippocampus (Lozada et al. 2012). This body of work suggests that nicotinic signaling not only regulates glutamatergic transmission throughout the mature HF but also may play an important role in the formation of integrated and stabilized network flow through this cognitive brain region.

Coordinated activity in the developing hippocampal formation.

Spontaneous coordinated activity of immature neuronal networks appears in the rodent hippocampus during the first postnatal week (Ben-Ari et al. 1989; Crépel et al. 2007; Garaschuk et al. 1998). These oscillations are mediated by excitatory GABAergic neurotransmission to form giant depolarizing potentials (GDPs) in the HF (Nardou et al. 2009; Sipilä et al. 2005), which are a primordial form of synchrony between neurons that precedes more organized coordinated activity such as theta and gamma rhythms (Bragin et al. 1995; Buzsáki and Draguhn 2004; Fellous and Sejnowski 2000; Kasyanov et al. 2004; Mohajerani and Cherubini 2006). Nicotinic signaling appears to play an important modulatory role in hippocampus GDP function during this developmental period. Both α7- and β2-subunit-containing nAChRs exert regulatory actions on network-driven GDPs, where stimulation of these nAChRs increases GDP frequency (Le Magueresse et al. 2006; Maggi et al. 2001), and conversely, blockade of these nAChRs decreases GDP frequency (Maggi et al. 2001). Because it is well established that α7 and α4β2* nAChRs are present on interneurons throughout regions of the hippocampus, it is proposed that the activation of these receptors on interneurons synapsing directly onto principal cells leads to an increase GABA release and increased GDP formation. Our present findings that functional heteromeric nAChRs are present on all principal neuron types of the young postnatal HF and that their activation leads to similar-magnitude excitability responses in active neurons across all regions of the HF suggest that these postsynaptic nAChRs may also be involved in the formation of coordinated and organized rhythms within the developing HF. Slightly later during postnatal development, nicotinic signaling at α7- and β2-subunit-containing nAChRs also facilitates the normal switch of GABAergic neurotransmission from excitation to inhibition by mediating increased expression of the KCC2 chloride transporter to establish the mature low intracellular chloride concentration (Liu et al. 2006). Because this switch occurs in rodent HF during the first 2 wk of postnatal life, when we observed the greatest nAChR function (Garaschuk et al. 1998; Tyzio et al. 2007), and requires calcium transients (Ganguly et al. 2001), nAChRs located directly on principal neurons during this developmental period may also aid to shape the development of mature HF neuronal networks through this mechanism.

The generation of somatic mAHPs by SK channels contributes to the excitability and firing frequency of neurons by regulating the speed and pattern of instantaneous firing frequency (Hallworth et al. 2003; Stocker et al. 1999), tonic firing frequency (Sah 1996; Wolfart et al. 2001; Zhang and McBain 1995), rhythmic burst activity, and oscillatory activity (Cingolani et al. 2002; Cueni et al. 2008; Wolfart and Roeper 2002). Our results show that during early postnatal development, SK channels mediate mAHPs of varying magnitude and timing across regions of the HF, which likely underlies their ability to differentially modulate and coordinate excitability responses to nAChR stimulation of varying magnitudes in active principal neurons across regions of the HF. The end result is quite striking, because the nAChR-mediated inward current and depolarization responses are much lower in the CA1, CA3, and DG regions of the hippocampus proper than in ECVI, yet this same stimulation leads to near-identical acceleration of action potential firing in active neurons across all HF regions. This coordinated nicotinic response may be important to the development of synchronized activity during development and the establishment of mature HF networks.

We found that principal neurons in young postnatal CA1 and SUB presented hyperpolarizing mAHPs following trains of action potentials that were not significantly affected by apamin treatment. Previous reports of studies using older rodents demonstrate that the two SK channel subtypes K_Ca2.1 and K_Ca2.2 are coexpressed at high levels in principal neurons of CA1, CA3, DG, SUB, and ECVI (Stocker and Pedarzani 2000) and that K_Ca2.2 channels are necessary and sufficient to generate a mAHP in CA1 principal neurons (Bond et al. 2004). The discrepancy between our results and those reported in the previous literature may be a factor of the young postnatal age of mice examined in this current study. In contrast with extrinsic afferents, intrinsic associational connections of the rodent hippocampus develop relatively late. Entorhinal axons innervate the hippocampus proper at embryonic day 15 (Skutella et al. 1999; Supèr and Soriano 1994), whereas principal neurons are generated in the second half of gestation and granule cells are generated postnatally (Altman and Bayer 1990; Jabès et al. 2011). To the best of our knowledge, the in situ expression of SK channels has not been reported previously for rodent HF principal neurons during P5–P10. More recent literature has provided evidence that is consistent with our finding that SK channels likely do not contribute to the mAHP in CA1 principal neurons. This work employed pharmacological channel blockers to show that the mAHP and neuronal excitability are regulated by voltage-gated Kv7/KCNQ/M potassium channels and hyperpolarization-activated cyclic nucleotide-gated (HCN) potassium channels, with little or no contribution from SK channels (Gu et al. 2005, 2008; Yue and Yaari 2004).

Whereas principal neurons in CA3 and DG did not present true hyperpolarization responses during the mAHP period following trains of action potentials, apamin treatment surprisingly decreased the membrane potential of DG neurons during the period, suggesting that SK channels influence the hyperexcitable nature of DG neurons during early postnatal development. Synchronized activity in immature networks may depend on the hyperexcitabilty of developing circuits in the hippocampus proper (Ben-Ari et al. 1989, 2007; Rivera et al. 1999), because depolarizing GABA supports episodes of bursting activity underlying waves of depolarization that precede more organized forms of HF activity (Blankenship and Feller 2010; Bragin et al. 1995; Buzsáki and Draguhn 2004; Fellous and Sejnowski 2000; Kasyanov et al. 2004; Mohajerani and Cherubini 2006). The effect of apamin to decrease excitability in DG neurons seems counterintuitive to the nature of SK channels, although this is not without precedent. In approximately half of neurons tested within the nucleus tractus solitarii, apamin decreased neuron excitability in a bicuculline-dependent manner, suggesting an indirect effect of apamin to increase the excitability of local interneurons and the resulting tonic GABAergic input to recorded neurons (Butcher et al. 1999). A similar mechanism may underlie apamin effects on young postnatal DG neurons in our study, given the extensive interneuron network already present at this age (Holter et al. 2007; Szabo et al. 2017). Alternatively, blockade of SK channels may unmask the function of additional potassium channels that are present in recorded DG neurons. Similar unmasking occurs in CA1 pyramidal neurons: although Kv7/KCNQ/M channels are major contributors to the mAHP in these neurons, as described above, blockade of Kv7/KCNQ/M channels unmasks a previously shunted influence of SK channels to enhance the mAHP (Chen et al. 2014). Given that the channel contributions toward excitability are reversed in mature DG neurons compared with CA1 neurons such that SK channels are the major contributor to the mAHP and Kv7/KCNQ/M channels reduce excitation through decreased input resistance and increased firing threshold (Mateos-Aparicio et al. 2014), it is possible that SK channel blockade in young DG neurons removes a source of shunting and allows Kv7/KCNQ/M channels to powerfully lower membrane potential following action potential firing. Additional voltage-gated potassium channels that may be unmasked in young DG neurons include the Kv1/KCNA channel, for which Kv1.1 subunit expression is greatest during the first 2 postnatal weeks in mice (Prüss et al. 2010) and function can strongly regulate action potential firing at a relatively low threshold (Kirchheim et al. 2013; Ovsepian et al. 2016), and Kv3/KCNC voltage-gated potassium channels, for which Kv3.4 subunit expression and function are normally present at low levels during the first postnatal week in rat (Riazanski et al. 2001).

Principal neurons in ECVI exhibited relatively slower onset mAHPs that were significantly inhibited by apamin, suggesting that SK channels act to decrease the excitability of these neurons. The regional differences in peak mAHP kinetics across the HF may result from SK channels being found in different subcellular compartments and coupled to a variety of calcium sources across different neuron types (Bennett et al. 2000; Hallworth et al. 2003; Stocker et al. 1999; Umemiya and Berger 1994; Williams et al. 1997). Regardless, the regional differences in SK channel regulation of principal neuron excitability are consistent with the disconnect between the strength of nicotinic passive depolarization responses from rest and nicotinic active excitability responses in firing neurons, for example, because SK channels appear to augment excitability of DG neurons and limit excitability of ECVI neurons, resulting in similar excitability responses to nicotinic stimulation in active neurons.

This study measured the overall influence of SK channel activation on an ACh-induced nicotinic excitability response when SK channels were activated by all sources of calcium. It should be noted that although this was not measured in the current study, calcium-permeable ligand-gated ion channels such as the nAChR may form a direct source of calcium influx to activate SK channels. For the α4β2* nAChRs that most likely mediated ACh responses in recorded neurons, receptors with the (α4)₃(β2)₂ stoichiometry display greater calcium permeability than receptors with the (α4)₂(β2)₃ stoichiometry (Fucile 2004; Tapia et al. 2007). Receptors with the (α4)₂(β2)₂(α5) stoichiometry, which are likely present in CA1 and ECVI principal neurons (Chung et al. 2016; Wada et al. 1989), exhibit greater calcium permeability than those containing α4- and β2-subunits only (Tapia et al. 2007). In addition, nicotinic receptor activation may indirectly increase SK channel function through elevations in calcium from intracellular calcium stores (Dajas-Bailador et al. 2002; Sharma and Vijayaraghavan 2001) and may facilitate the activation of voltage-dependent calcium channels in immature neurons (Dajas-Bailador et al. 2002; Rathouz and Berg 1994). One example occurs in cochlear outer hair cells, where SK channels are tightly coupled with heteromeric nAChRs such that receptor activation by ACh induces an influx of calcium that activates SK channels, followed by the consequent efflux of potassium. The resulting fast inhibitory signal modulates the frequency of action potential trains elicited by cholinergic signaling (Oliver et al. 2000). The potential contribution of nAChR stimulation toward SK channel activation, and its role in coordinated activity in principal neurons across the developing HF, are important areas for future research.

Implications for aberrant development of the hippocampal formation.

The presence of functional heteromeric nAChRs on principal neurons during early postnatal HF development has implications for the understanding of neurodevelopmental disorders affecting HF structure and HF-dependent higher order cognitive functions such as learning, memory, and attention. For example, children diagnosed with ASD and ADHD have presented with abnormal postnatal development of the HF leading to increased hippocampus size (Plessen et al. 2006; Schumann et al. 2004). Dysfunction specifically to hippocampal nAChRs has also been linked with specific neurodevelopmental disorders, including epilepsy and schizophrenia (Labate et al. 2013; Roshan-Milani et al. 2003; Tizabi 2007; Tregellas et al. 2010). It will be important to determine whether heteromeric nAChRs present on young postnatal HF principal neurons contribute to regulation of HF network activity and synchrony during this time, because epilepsy involves dysregulation of normal oscillations and/or neuron loss in the CA1, CA3, and entorhinal cortex (Bragin et al. 1999; Coras et al. 2014; Lega et al. 2015; Schmeiser et al. 2017), and schizophrenia is linked with alterations to the morphology of dentate granule principal neurons (Lauer et al. 2003; Senitz and Beckmann 2003), the synaptic organization of CA3 principal neurons (Kobayashi 2009; Kolomeets et al. 2007), and markers of synaptic transmission for glutamatergic and GABAergic systems (Reynolds et al. 1990; Talbot et al. 2004). Heteromeric nAChRs are currently targeted for the treatment of neurodevelopmental disorders such as ASD, ADHD, and schizophrenia, each of which involves abnormal HF-dependent cognitive functions (Arneric et al. 2007; Dineley et al. 2015; Taly et al. 2009).

Conclusion.

Results from this study demonstrate that heteromeric nAChRs mediate nicotinic signaling in CA1, CA3, DG, SUB, and ECVI principal neurons of the young postnatal HF. Differential regulation of excitability by SK channels in these neurons normalizes the differential magnitude of their passive nAChR responses, leading to active nAChR responses in firing neurons that are similar in magnitude across all HF regions. These findings provide novel insight into normal HF development and place heteromeric nAChRs in a position to directly influence the development of coordinated, synchronous neurotransmission within the developing HF cognitive network.

GRANTS

This work was supported by Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant 436190 (to C.D.C. Bailey) and Canada Foundation for Innovation Project No. 30381 (to C.D.C. Bailey). B.Y.T. Chung was the recipient of a Postgraduate Scholarship-Doctoral Program from NSERC, an Ontario Graduate Scholarship from the Government of Ontario, and an Ontario Veterinary College Scholarship from the University of Guelph.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

B.Y.T.C. and C.D.C.B. conceived and designed research; B.Y.T.C. performed experiments; B.Y.T.C. and C.D.C.B. analyzed data; B.Y.T.C. and C.D.C.B. interpreted results of experiments; B.Y.T.C. and C.D.C.B. prepared figures; B.Y.T.C. drafted manuscript; B.Y.T.C. and C.D.C.B. edited and revised manuscript; B.Y.T.C. and C.D.C.B. approved final version of manuscript.