Developmental nicotine exposure alters potassium currents in hypoglossal motoneurons of neonatal rat

Developmental nicotine exposure is associated with increased cell excitability, which is often accompanied by compensatory changes aimed at normalizing excitability. Here we show that whole cell potassium currents are also increased in hypoglossal motoneurons from nicotine-exposed neonatal rats under conditions of increased cell and network excitability. This is consistent with a compensatory response aimed at preventing instability under conditions in which excitatory synaptic input is high and is compatible with the concept of homeostatic plasticity.

Abstract

We previously showed that nicotine exposure in utero and after birth via breast milk [developmental nicotine exposure (DNE)] is associated with many changes in the structure and function of hypoglossal motoneurons (XIIMNs), including a reduction in the size of the dendritic arbor and an increase in cell excitability. Interestingly, the elevated excitability was associated with a reduction in the expression of glutamate receptors on the cell body. Together, these observations are consistent with a homeostatic compensation aimed at restoring cell excitability. Compensation for increased cell excitability could also occur by changing potassium conductance, which plays a critical role in regulating resting potential, spike threshold, and repetitive spiking behavior. Here we test the hypothesis that the previously observed increase in the excitability of XIIMNs from DNE animals is associated with an increase in whole cell potassium currents. Potassium currents were measured in XIIMNs in brain stem slices derived from DNE and control rat pups ranging in age from 0 to 4 days by whole cell patch-clamp electrophysiology. All currents were measured after blockade of action potential-dependent synaptic transmission with tetrodotoxin. Compared with control cells, XIIMNs from DNE animals showed significantly larger transient and sustained potassium currents, but this was observed only under conditions of increased cell and network excitability, which we evoked by raising extracellular potassium from 3 to 9 mM. These observations suggest that the larger potassium currents in nicotine-exposed neurons are an important homeostatic compensation that prevents “runaway” excitability under stressful conditions, when neurons are receiving elevated excitatory synaptic input.

NEW & NOTEWORTHY Developmental nicotine exposure is associated with increased cell excitability, which is often accompanied by compensatory changes aimed at normalizing excitability. Here we show that whole cell potassium currents are also increased in hypoglossal motoneurons from nicotine-exposed neonatal rats under conditions of increased cell and network excitability. This is consistent with a compensatory response aimed at preventing instability under conditions in which excitatory synaptic input is high and is compatible with the concept of homeostatic plasticity.

developmental nicotine exposure (DNE), here defined as exposure in utero with continued exposure through breast milk after birth, is associated with alterations in the structure and function of neonatal neurons that govern many important functions, including learning and memory (Levin et al. 1996; Mitchell et al. 2012; Parameshwaran et al. 2012), the control of breathing (Eugenín et al. 2008; Fregosi and Pilarski 2008; Hafström et al. 2005; Huang et al. 2010; Jaiswal et al. 2013, 2015; Robinson et al. 2002; St-John and Leiter 1999), and the regulation of heart rate (Boychuk and Hayward 2011; Evans et al. 2005; Hafström et al. 2002; Huang et al. 2005; Neff et al. 2004; Slotkin et al. 1997). Recent studies in neonatal hypoglossal motoneurons (XIIMNs) show that DNE is associated with stunted dendrites (Powell et al. 2016), desensitization of their postsynaptic nicotinic acetylcholine receptors (nAChRs) (Pilarski et al. 2012), increased excitability (Jaiswal et al. 2013; Pilarski and Fregosi 2009), and increased spike frequency adaptation (Powell et al. 2015). Interestingly, the enhanced excitability in nicotine-exposed cells was accompanied by reduced excitatory synaptic input (Pilarski et al. 2011) as well as a reduction in the expression of glutamate receptors on the cell body (Jaiswal et al. 2013). We interpreted this as a homeostatic compensation aimed at restoring normal cell excitability, which would be important for survival given that XIIMNs are critical for the normal development of mammals, owing to their role in suckling, swallowing, and breathing. This idea is consistent with other observations showing that breathing in vivo or breathing-related motor nerve bursts recorded from in vitro preparations are at best marginally affected by DNE under nonstressed, baseline conditions (Cholanian et al. 2017; Huang et al. 2010; Wollman et al. 2016), consistent with the recruitment of compensatory mechanisms aimed at normalizing cell and network excitability.

Considerable work on homeostatic synaptic plasticity in a range of neuron types shows that it can occur by balancing inhibitory and excitatory synaptic inputs and/or by changing intrinsic neuron properties (Hengen et al. 2013; Turrigiano 2008, 2012). Potassium channels play a prominent role in determining intrinsic cell excitability, by adjusting resting membrane potential, firing threshold, repetitive spiking behavior, etc. A recent study in Drosophila motoneurons and interneurons (Ping and Tsunoda 2012) showed that inactivity-induced upregulation of nAChRs, leading to an increase in excitatory synaptic input, was associated with a compensatory increase in the expression of potassium channels, with subsequent stabilization of cell excitability. Although cell excitability was not completely restored, excessive excitability was nonetheless avoided by a homeostatic adjustment of intrinsic neuron properties (Ping and Tsunoda 2012).

Considering the totality of the above observations, we designed experiments to test the hypothesis that the previously observed increase in the excitability of XIIMNs from DNE animals is associated with an increase in the total whole cell potassium current (including transient, rapidly inactivating and sustained, noninactivating components). DNE did increase both transient and sustained potassium currents in XIIMNs, though only under conditions of elevated extracellular potassium. These observations suggest that the increased potassium currents in nicotine-exposed neurons are an important homeostatic compensation that prevents “runaway” excitability under stressful conditions, when neurons are receiving elevated excitatory synaptic input.

MATERIALS AND METHODS

Animals.

We used 91 Sprague-Dawley rat pups of either sex, ranging in age from postnatal day 0 (P0) through postnatal day 4 (P4). This neonatal period corresponds to a gestational age of 29-34 wk in humans (Ballanyi et al. 1999). All neonates were born via spontaneous vaginal delivery from adult female rats purchased from Charles River Laboratories (Wilmington, MA). The nicotine-exposed neonates were taken from 30 separate litters; the saline-exposed and unexposed neonates (see below) were taken from 34 separate litters. We used one to three neonates from each litter. The pups were housed together with their mothers and siblings in the animal care facility at the University of Arizona under a 12:12-h light-dark cycle (lights on 0700) in a quiet room maintained at 22°C and 20-30% relative humidity and with water and food available ad libitum. All procedures and protocols described in this report were approved by the University of Arizona Institutional Care and Use Committee and conformed to National Institutes of Health guidelines.

Developmental nicotine exposure.

DNE was achieved by subcutaneous implantation of Alzet 1007D mini-osmotic pumps (Alzet) into pregnant dams on gestational day 5, as previously described (Huang et al. 2004; Luo et al. 2004, 2007). The rats were anesthetized with a subcutaneous injection of a mixture of xylazine (8.0 mg/kg), ketamine (25 mg/kg), and acepromazine (1.0 mg/kg); buprenorphine (0.5 mg/kg) was given for control of postoperative pain. Pregnant dams and their fetuses were exposed to an average dose of 6 mg·kg⁻¹·day⁻¹ of nicotine bitartrate, which produces plasma levels of cotinine ranging from 60 to 92 ng/ml (Powell et al. 2015). This is comparable to the average cotinine level (88 ng/ml) found in the umbilical cord blood of newborns whose mothers smoked an average of 95 cigarettes/week (Berlin et al. 2010). Control animals were derived from pregnant dams into which we implanted Alzet pumps filled with sterile saline (sham control dams) and from true control dams that did not receive the pump. As in numerous earlier studies, we did not see any differences between neonatal XIIMNs from pups born to sham or true control dams for any of the variables that we studied. As a result, data from the two control groups were combined.

Medullary slice preparation.

Pups of either sex were removed from their home cages at random and weighed. Animals were anesthetized with hypothermia and decerebrated at the coronal suture. The vertebral column and ribcage were exposed and moved to a dissection dish filled with cold (4–8°C) oxygenated (95% O₂-5% CO₂) artificial cerebrospinal fluid (aCSF), composed of (in mM) 120 NaCl, 26 NaHCO₃, 30 glucose, 1 MgSO₄, 3 KCl, 1.25 NaH₂PO₄, and 1.2 CaCl₂ with pH adjusted to 7.4 and osmolarity to 300–325 mosM. The spinal cord and brain stem were extracted, and all tissue rostral to the pontomedullary junction was removed. The preparation was glued to an agar block, rostral surface up, for serial microsectioning in a vibratome (VT1000; Leica). Transverse medullary slices were taken in chilled oxygenated aCSF until the most rostral hypoglossal nerve rootlets were near the tissue surface. A single 700-μm slice was taken to capture the majority of the hypoglossal motor nucleus. The slice was then transferred to a container with fresh aCSF that was continuously aerated with 95%O₂-5%CO₂ and allowed to equilibrate at room temperature for 1–2 h before recording.

Electrophysiology.

Equilibrated slices were transferred to the recording chamber and perfused at a rate of 1.5–2 ml/min with aCSF that was maintained at 27°C (TC-324B temperature controller, Warner Instrument), and aerated with 95%O₂-5%CO₂. XIIMNs were visualized with an Olympus BX-50WI fixed-stage microscope (×40 water-immersion objective, 0.75 NA) with infrared and differential interference contrast optics and a video camera (C25400-07, Hamamatsu). Recordings were made with glass pipettes (tip resistance 4–8 MΩ) pulled from thick-walled borosilicate glass capillary tubes (OD: 1.5 mm, ID: 0.75 mm) and filled with a solution containing (in mM) 135 K-gluconate, 4 KCl, 10 HEPES, 5 ATP (Mg²⁺ salt), 0.375 GTP, and 12.5 phosphocreatine, with pH adjusted to 7.2 and osmolarity of 275–300 mosM. Filled pipettes were attached to a preamplifier mounted in a micromanipulator (Siskiyou, MC1000E). The preamplifier was connected to a Multiclamp 700B amplifier, and the signals were digitized with a Digidata 1440 A/D converter (Molecular Devices, Sunnyvale, CA).

The following procedures pertain to all recordings described below (see Protocols). First, XIIMNs were identified on the basis of their size and location within the XII motor nucleus. Pipettes were apposed to the soma, and after a gigaohm seal was achieved and the membrane under the seal was ruptured the cell was maintained in voltage-clamp mode (holding potential −70 mV). In all experiments, the slow and fast components of the capacitive current transient were cancelled to the extent possible, and series resistance was compensated with a correction coefficient of 60%. After these corrections were made, a square-wave voltage step (from −70 to −75 mV) was introduced to calculate the input resistance of the cell. We then switched to current-clamp mode to record the resting membrane potential (V_m), and then switched back to voltage-clamp mode. The holding potential was set to −70 mV, and the perfusate was switched to aCSF containing tetrodotoxin (TTX). After 5 min of TTX application, we ran a voltage-step protocol, which consisted of twelve 10-mV square-wave steps from a holding potential of −70 mV. Each step was 400 ms in duration, with the first step −100 mV and the last step +10 mV.

Drugs.

Drugs were purchased from Sigma (St. Louis, MO), except for TTX, which was purchased from R&D Chemicals (Minneapolis, MN). The drugs were mixed on the day of the experiment in either standard (3 mM K⁺) or modified (9 mM K⁺) aCSF (see below) with the following concentrations: TTX, 1 μM; cadmium chloride hemi(pentahydrate) (Cd²⁺, 200 μM); and tetraethylammonium chloride (TEA, 10 mM). The drug solutions were aerated with 95%O₂-5%CO₂, maintained at 27°C, and perfused into the recording chamber at a rate of 1.5–2 ml/min.

Protocols.

We report data obtained from 179 XIIMNs, with 93 cells from control neonates and 86 from DNE neonates. The main protocol was designed to measure transient and sustained outward currents (see below and see Fig. 3 for an example) in either physiological (3 mM) or elevated (9 mM) extracellular potassium. A total of 52 control (from 18 dams) and 57 DNE (from 21 dams) XIIMNs were studied at physiological extracellular potassium (3 mM), and a total of 30 control (from 15 dams) and 21 DNE (from 19 dams) XIIMNs were studied in elevated extracellular potassium (9 mM). Some of the XIIMNs studied in elevated potassium (DNE, n = 14 of 21; control, n = 21 of 30) were used in an earlier study designed to examine postsynaptic responses to AMPA or glutamate before the experiments described below. However, these cells were used only when AMPA or glutamate effects could be completely reversed with washout. Moreover, comparison of the potassium current in cells that were or were not first exposed to AMPA/glutamate did not reveal any systematic differences. We used elevated potassium in some of the experiments in an effort to evoke rhythmic excitatory synaptic input from the respiratory central pattern generator to the XIIMNs, although this also reduces the potassium driving force. All XIIMNs studied in elevated potassium were confirmed to be rhythmically active before TTX was applied.

Fig. 3.

Transient and sustained outward potassium currents in XIIMNs. A: representative recording demonstrating transient (enclosed by dotted rectangle) and sustained (arrow) outward current components recorded in the presence of TTX. B: zoomed-in view of the transient outward currents, which quickly follow the capacitive current. Arrow indicates where the peak transient current was measured. Voltage steps were initiated from a holding potential of −70 mV and began by stepping to a command potential of −100 mV, followed by 11 depolarizing steps, delivered in 10-mV increments, up to +10 mV. The leak current was subtracted from all traces.

In the second series of studies, the voltage-clamp protocol described above was repeated before and after the addition of Cd²⁺ to block the Ca²⁺-dependent K⁺ current. We first repeated the voltage step protocol in TTX alone and then repeated it 5 min after bath application of a cocktail containing TTX and Cd²⁺. This protocol was done in 11 cells at normal extracellular potassium (5 control, 6 DNE) and 11 cells in elevated potassium (6 control, 5 DNE). In the cells studied in 3 mM extracellular potassium, we also ran a protocol that consisted of a prepulse to −30 mV from a holding potential of −70 mV, with the prepulse followed by twelve 400-ms-duration square-wave steps as described above. The prepulse protocol was done to estimate the contribution of voltage-gated, rapidly inactivating A-type potassium currents to the total outward current.

Finally, in the third series of studies we measured transient and sustained outward currents in TTX alone and again 5 min after bath application of a cocktail containing TTX and TEA. These experiments were done to confirm that the outward currents were carried by potassium channels and were done in four cells from control neonates and two cells from DNE neonates; all TEA studies were done in normal (3 mM) extracellular potassium concentration. At the end of each of the experiments described above, measurement of input resistance was repeated to confirm that the gigaohm seal was still intact.

Data analysis and statistics.

Voltage-clamp data were obtained with Clampex software and analyzed with Clampfit software (Molecular Devices). First, the leak current was subtracted from all traces and from these leak-subtracted traces we measured the transient and sustained outward currents, as shown in Fig. 3. Note that the transient outward current was measured immediately after the capacitive artifact, while the sustained current, also termed the “steady-state” current, was measured at the end of the 400-ms voltage step. We note that the leak subtraction protocol was imperfect, as indicated by small outward currents at hyperpolarized command potentials (see Fig. 4). We used the Clampfit program to subtract leak current, but small currents of inconsistent polarity remained. We are not sure why this is the case, but there are at least three possibilities: 1) errors in the subtraction process; 2) the presence of calcium currents; or 3) the cells are coupled by gap junctions, resulting in variable leaks. It is well known that XIIMNs have gap junctions (Bou-Flores and Berger 2001; Cifra et al. 2009) and that that there will be leak subtraction errors when neurons are electrically coupled (Rabbah et al. 2005). In addition, the leak subtraction process relies in part on measures of input resistance, and we used the value of input resistance obtained right after break-in for the leak subtraction calculation. To determine whether input resistance changed by the time the actual voltage-clamp protocol was run, we randomly selected data from five neurons and manually calculated input resistance in these cells using the voltage and current values obtained during the voltage-clamp protocol. The input resistance values measured at the beginning and end of the protocol were remarkably consistent, indicating that changes in input resistance over the course of the experiment are unlikely to explain errors in leak subtraction. The presence of gap junctions, calcium currents, and random errors appear to be the most plausible explanations. All leak-subtracted currents were plotted as a function of the command potential (see Fig. 4) and analyzed statistically with two-way ANOVA (Prism, GraphPad Software, La Jolla, CA) with the main factors treatment (control, DNE) and command potential. Data were considered statistically significant if P ≤ 0.05. Data in all figures are represented as means ± SE unless stated otherwise.

Fig. 4.

DNE significantly increased both transient and sustained outward currents in high extracellular potassium. A and C: transient (A) and sustained (C) potassium currents recorded in normal (3 mM) extracellular potassium were the same in control and DNE XIIMNs (control, n = 48 cells; DNE, n = 55 cells). B and D: in 9 mM extracellular potassium, both transient (A) and sustained (C) potassium currents were larger in cells from DNE animals than cells from control animals (control, n = 30 cells; DNE, n = 21 cells). For transient current P = 0.001; for sustained current P = 0.0103 (2-way ANOVA). In all panels, data are shown as means ± SE.

We also conducted fluctuation analysis of the whole cell currents to obtain a rough estimate of the mean single-channel current and the total number of channels, as described in Alvarez et al. (2002). Using Clampex software, we conducted fluctuation analysis on the peak current in 21 control cells and 15 DNE cells that were studied in high potassium and in which acceptable measurements of variance could be made. We report data for peak current, making the assumption that all channels were activated under these conditions (Noceti et al. 1996), although even under these conditions channel number will be underestimated if there is significant inactivation. The average estimated values of single-channel current and whole cell channel number in cells from control and DNE animals were compared statistically with the unpaired t-test.

RESULTS

Age and weight of animals used and V_m and input resistance in control and DNE cells studied in normal or elevated extracellular potassium.

As shown in Fig. 1, top left, the DNE animals used in experiments done in normal extracellular potassium were slightly but significantly older than the control animals. Control and DNE animals studied in high extracellular potassium were the same age (Fig. 1, top right). There were no differences in body weight between the control and DNE animals used in either the low- or high-potassium studies (Fig. 1, second row). The resting membrane potential recorded in normal extracellular potassium was slightly but significantly more hyperpolarized in cells from DNE animals than in control animals (Fig. 1, third row, left, P = 0.0098), and although 9 mM extracellular potassium made the resting potential more depolarized, there were no differences between control and DNE cells (Fig. 1, third row, right). There were no significant differences in input resistance in cells from control and DNE animals, whether studied in a normal or elevated extracellular potassium concentration (Fig. 1, bottom).

Fig. 1.

Age, weight, resting membrane potential (V_m), and input resistance (R_in) in XIIMNs from control and DNE animals, recorded in either 3 or 9 mM extracellular potassium. Left: preparations from which we recorded cells in 3 mM extracellular potassium (n = 48 control cells and 52 DNE cells); none of these cells showed rhythmic activity. Right: cells studied in 9 mM extracellular potassium; all of these cells had rhythmic, respiration-related activity. In all panels, the horizontal line represents the mean. The cells studied in 3 mM extracellular potassium in the DNE group were derived from somewhat older animals than in the control group (age; control vs. DNE, P = 0.003, unpaired t-test), and they had a slightly more hyperpolarized V_m (control vs. DNE, P = 0.0098, unpaired t-test). There were no differences in weight or R_in. Values for all 4 variables were the same in control (n = 30 cells) and DNE (n = 21 cells) animals used in the 9 mM extracellular potassium experiments (right).

Potassium channels in neonatal rat XIIMNs.

To determine whether the potassium currents recorded in XIIMNs from the young neonatal rats studied here share attributes of potassium currents previously recorded in XIIMNs (Jiang et al. 1994; Lape and Nistri 2000; Viana et al. 1993), we ran three protocols. First, since in many neurons both transient and sustained outward currents are partially carried by Ca²⁺-activated potassium channels, we blocked Ca²⁺-dependent currents by bath application of CdCl₂ in a subset of cells (n = 5 control and 6 DNE in 3 mM extracellular potassium; n = 6 control and 5 DNE in 9 mM extracellular potassium). The outward currents obtained after application of CdCl₂ were subtracted from the total current measured in TTX alone (Fig. 2, A and B). Bath application of CdCl₂ reduced both transient and sustained outward currents, and this was true whether the extracellular potassium concentration was 3 mM or 9 mM. For example, in 3 mM external potassium CdCl₂ reduced the transient and sustained outward currents in XIIMNs from control animals by 37% and 30%, respectively, which is in the range observed in XIIMNs previously (Lape and Nistri 2000). Nevertheless, although Ca²⁺-activated potassium channels make a significant contribution to the total outward current, this contribution was not significantly different in control and DNE cells, either in 3 mM or 9 mM extracellular potassium.

Fig. 2.

Representative recordings demonstrating methods used to isolate Ca²⁺ dependent, A-type, and TEA sensitive potassium currents. A and B: Ca²⁺ dependence. First, after TTX was bath applied, 12 voltage steps were initiated from a holding potential of −70 mV, starting with a command step to − 100 mV and increasing by 10 mV with each step. Under these conditions, both transient and sustained outward currents are observed (A). After bath application of a cocktail containing TTX and CdCl₂ the voltage step protocol was repeated. The addition of CdCl₂ resulted in diminished transient and sustained outward currents (B). C: identification of A-type currents. To inactivate A-type potassium currents, a brief prepulse from −70 mV to −30 mV was delivered before initiation of the voltage step protocol. Blocking A-type currents further reduced the sustained current and virtually eliminated the transient current, as shown in the zoomed-in version. Leak current was subtracted from all traces. The recordings shown here were done in a XIIMN from a DNE animal in 9 mM extracellular potassium. D and E: identification of TEA-sensitive currents. First, after TTX was bath applied, 12 voltage steps were initiated from a holding potential of −70 mV (D), as depicted in inset below E. Under these conditions, both transient and sustained outward currents are observed. After bath application of a cocktail containing TTX and TEA the voltage step protocol was repeated (E). It is clear that the addition of TEA was associated with a marked reduction of both transient and sustained outward currents. The recordings shown here were from a XIIMN from a DNE animal and studied in 9 mM extracellular potassium. The leak current was first subtracted from all traces. See text for further explanation.

To determine whether the transient outward current component was inactivated rapidly at depolarizing potentials, as is common for A-type potassium currents, in a small subset of cells (5 control, 6 DNE) studied in 3 mM extracellular potassium we conducted a protocol that began with a prepulse to −30 mV, followed by the same step protocol described above (Fig. 2C). The prepulse to −30 mV essentially abolished the transient potassium current (Fig. 2C, magnified traces), and further reduced the sustained outward currents, suggesting that XIIMNs also express A-type currents. However, the change in current evoked by the prepulse was the same in cells from control and DNE animals.

We then asked whether a component of the potassium current could be blocked by TEA, given that previous studies have shown that XIIMNs express TEA-sensitive channels (Lape and Nistri 2000; Viana et al. 1993). Accordingly, we bath applied TEA and repeated the step protocol in four control and two DNE XIIMNs, all studied in 3 mM extracellular potassium. As shown in Fig. 2, D and E, TEA powerfully suppressed both transient and sustained outward currents.

Measurement of transient and sustained outward currents in neonatal hypoglossal motoneurons.

An example of the outward currents that were the main focus of this study is shown in Fig. 3. In the presence of TTX, a large, rapidly activating and inactivating transient outward current was observed soon after the capacitive artifact. These currents are shown within the dotted rectangle in Fig. 3A and on expanded current and time scales in Fig. 3B. The downward arrow in Fig. 3B illustrates where we measured the amplitude of this transient current. We also consistently observed a sustained outward current, which was measured near the end of each voltage step, as indicated by the downward arrow in Fig. 3A. Transient and sustained outward currents measured in 3 mM extracellular potassium were virtually identical in XIIMNs from control (n = 48) and DNE (n = 52) animals (Fig. 4, A and C, respectively). In contrast, measurement of outward currents in 9 mM extracellular potassium revealed significant differences between control (n = 30) and DNE (n = 21) cells. As shown in Fig. 4, B and D, both transient (F = 10.95, P = 0.001; Fig. 4B) and sustained (F = 6.6, P = 0.0103; Fig. 4D) outward currents were higher in DNE animals than in control cells. There were no significant interactions between the command potential and either transient or sustained outward currents under either low- or high-potassium conditions, suggesting that the treatment effect was independent of the voltage command step (Fig. 4, B and D).

Given the different potassium driving force in experiments done in 3 and 9 mM extracellular potassium, we also calculated whole cell conductance under all conditions. The results of the conductance analysis mirrored the data in Fig. 4, inasmuch as there were no differences between control and DNE cells in 3 mM extracellular potassium but a significant difference at 9 mM extracellular potassium. For example, peak sustained conductance at 10 mV, the highest command potential that we used, averaged 2.3 ± 0.24 and 3.5 ± 1.15 nS in control and DNE cells, respectively, and two-way ANOVA revealed a significant treatment effect (F = 5.3, P = 0.022). As with the current analysis (Fig. 4), the interaction between treatment and command potential was not significant.

Noise fluctuation analysis did not reveal any differences in estimated single-channel current or channel number between cells from control and DNE animals. The estimated single-channel current averaged 1.8 ± 0.2 pA and 1.84 ± 0.3 pA (n = 10) in control and DNE cells, respectively (P = 0.9783). The estimated channel number averaged 1,018 ± 183 and 962 ± 219 in control and DNE cells, respectively (P = 0.8567).

DISCUSSION

This is the first study to examine the influence of DNE on whole cell potassium currents in XIIMNs of young neonatal rats. Our studies, in a large number of cells, show clearly that DNE alters both transient and sustained whole cell potassium currents in XIIMNs but only in elevated extracellular potassium (9 mM), as discussed below. Additional experiments documented the presence of A-type, Ca²⁺-dependent, and TEA-sensitive outward currents in the neonatal XIIMNs that we studied. However, at this time it is unclear what specific potassium channels are altered by DNE. Below we discuss why the differences are observed only in elevated extracellular potassium and the putative functional consequences of our observations.

Influence of extracellular potassium on outward currents in control and DNE cells.

One of our more intriguing findings is that the DNE-mediated augmentation of the outward current was only observed in XIIMNs that were studied in 9 mM extracellular potassium (Fig. 4). Elevating extracellular potassium reduces the chemical driving force on potassium and depolarizes all neurons in the slice, not just XIIMNs. Nonetheless, despite the reduced potassium driving force, high extracellular potassium is associated with increased outward current magnitude in expression systems (Wood and Korn 2000) and in neurons (Pardo et al. 1992). The reason for this paradox is uncertain, but detailed biophysical analyses of voltage-gated potassium (K_v) channels (specifically K_v2.1) indicate that this can be accounted for, in large part, by the conversion of potassium channels from the inactivated to the activated state, particularly at more positive command potentials, where potassium channel inactivation is common (Wood and Korn 2000). How the increased extracellular potassium removes inactivation has not been definitively established, but data in voltage-dependent, TEA-sensitive channels suggest that raising extracellular potassium alters the conformation of the channel’s outer vestibule in a manner that favors increased channel conductance (Immke et al. 1999; Immke and Korn 2000). Interestingly, the higher potassium current in neurons from DNE cells observed at more positive command potentials (ranging from −20 to 10 mV; Fig. 4) is consistent with the removal of channels from inactivation. Indeed, if one examines the peak transient and sustained currents in 3 and 9 mM potassium (Fig. 4), it is noteworthy that these currents tend to be smaller in 9 compared with 3 mM potassium in control cells, while the opposite is true in cells from DNE animals.

Another possibility is that high extracellular potassium depolarized neurons within the hypoglossal motor nucleus or elsewhere in the thick slice and altered the properties of motoneurons in DNE but not control preparations. Increased excitatory synaptic activity in high extracellular potassium most certainly occurs, as all XIIMNs studied under these conditions showed strong, respiration-related rhythmic activity, consistent with many previous studies (Cholanian et al. 2017; Del Negro et al. 2002; Feldman and Del Negro 2006; Pilarski et al. 2011). Although our experiments were done in the presence of TTX, the slices were bathed in 9 mM potassium soon after they were placed in the recording chamber, and typically several minutes elapsed before we were able to record from a rhythmically active motoneuron. Thus, there was more than enough time for neuromodulators to be released, and it is likely that modulators such as acetylcholine (Shao et al. 2008; Shao and Feldman 2005), serotonin (Monteau et al. 1990; Schwarzacher et al. 2002), substance P (Johnson et al. 1996), and norepinephrine (Zaczek et al. 1993) were released from interneurons. It is plausible that such “spillover” of these modulators was increased in XIIMNs from DNE animals, especially since distinct populations of potassium channels, depending on their subcellular localization, can respond differently to local changes in neurotransmitters and, in turn, modulate local activity-dependent pathways (Trimmer 2015).

Another important consideration is the physiological significance of experiments done in elevated extracellular potassium. Although the extracellular potassium concentration bathing dorsal vagal motoneurons in slices varies by only a few millimolar under baseline conditions, stimulation of the solitary tract was associated with a change in extracellular potassium from 3 to 9 mM (Brockhaus et al. 1993). Similarly, values for extracellular potassium in hippocampal neurons subjected to severe hypoxic stress can reach 50 mM or more (Müller and Somjen 2000). Thus, our findings are likely significant during hypoxic stress or other conditions associated with intense neuronal activation.

Possible functional consequences of altered potassium current in XIIMNs.

The combinatorial assembly of multiple potassium channel subunits leads to substantial structural and functional diversity (Choe 2002). Moreover, differences in the localization of a given potassium channel (e.g., at the axon initial segment, at presynaptic sites, on dendritic spines, etc.) can enhance or diminish the channel’s function (Trimmer 2015). Historically, K_v channels have been linked to overall excitability of the neuron (Tsantoulas and McMahon 2014), with mutations of K_v channels associated with disorders of neuronal hyperexcitability, including epilepsy, chronic pain and episodic ataxia-1 (Browne et al. 1994; Li and Toyoda 2015; Maljevic and Lerche 2013). Consistent with these observations, we have shown that XIIMNs from DNE animals show alterations in their excitability and firing rate behavior. These include a lower action potential firing threshold (Pilarski et al. 2011; Powell et al. 2015) and a steeper relation between action potential discharge frequency and injected current but a reduced peak discharge rate (Pilarski et al. 2011). Cells from DNE animals also showed a reduced steady-state discharge rate, defined as the average rate over the last 250 ms of a 1-s injection of depolarizing current, as well as increased spike frequency adaptation (Powell et al. 2015).

The increased potassium currents observed in cells from DNE animals may explain some of these observations. For example, the larger sustained (“steady state”) potassium current in XIIMNs from DNE animals could account for the lower steady-state discharge rate and the increase in spike frequency adaptation observed previously (Powell et al. 2015). It is also possible that the larger transient, fast-inactivating potassium current in the DNE cells underlies the previously observed increase in the relation between action potential discharge frequency and injected current (Pilarski et al. 2011), by hastening repolarization and reducing the interspike interval. At the same time, the increased sustained potassium current may account for the lower peak discharge rate observed previously in XIIMNs from DNE animals (Pilarski et al. 2011), or at least contribute to it. We hypothesize that the previously observed reduction in the complexity of the dendritic tree in XIIMNs from DNE animals (Powell et al. 2016) precipitates increased cell excitability; the increase in potassium currents observed here reflects one of several mechanisms that can counteract it, consistent with the concept of homeostatic plasticity. Nevertheless, our estimates of channel number did not reveal any differences between control and DNE neurons, which argues against increased potassium channel expression. However, given that we previously showed a smaller dendritic tree in cells from DNE animals compared with control cells (Powell et al. 2016), it is possible that the relative density of potassium channels is higher in cells from DNE animals, as the same number of channels would be dispersed over a smaller surface area. Alternative interpretations are that channel distribution is altered, or that the voltage sensitivity of channel open probability is different in cells from control and DNE animals. Similarly, single-channel current did not reveal significant differences between control and DNE cells, suggesting that the larger whole cell current in the cells from DNE animals is unlikely due to differences in channel subtype. These results, based on noise analysis, should be interpreted with caution as the analysis is based on various assumptions, including an adequate space clamp in all neurons studied.

Conclusions.

We have demonstrated a DNE-mediated enhancement of transient and sustained outward currents carried via potassium channels, but only under conditions where extracellular potassium concentration was increased threefold above normal. We have established that a component of these outward currents are Ca²⁺ sensitive, although the great majority of the outward currents were blocked by TEA, which blocks most potassium channels. Accordingly, detailed single-channel studies are necessary to elucidate the specific potassium channels targeted by DNE, as well as the mechanism behind DNE-induced augmentation of the outward current. The possibilities include, but are not limited to, increased potassium channel expression, rearrangement of the potassium channel subunits, different inactivation/activation kinetics, differences in the subcellular localization of existing K⁺ channels, and modulator-induced activation of potassium currents.

GRANTS

This work was supported by National Institute of Child Health and Human Development Grant R01 HD-071302.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

M.C. and J.W. performed experiments; M.C. and R.F.F. analyzed data; M.C., R.B.L., and R.F.F. interpreted results of experiments; M.C. and R.F.F. prepared figures; M.C. and R.F.F. drafted manuscript; M.C., R.B.L., and R.F.F. edited and revised manuscript; M.C., J.W., R.B.L., and R.F.F. approved final version of manuscript.