K_V7 channels are potential regulators of the exercise pressor reflex

Abstract

The exercise pressor reflex (EPR) originates in skeletal muscle and is activated by exercise-induced signals to increase arterial blood pressure and cardiac output. Muscle ischemia can elicit the EPR, which can be inappropriately activated in patients with peripheral vascular disease or heart failure to increase the incidence of myocardial infarction. We seek to better understand the receptor/channels that control excitability of group III and group IV muscle afferent fibers that give rise to the EPR. Bradykinin (BK) is released within contracting muscle and can evoke the EPR. However, the mechanism is incompletely understood. K_V7 channels strongly regulate neuronal excitability and are inhibited by BK. We have identified K_V7 currents in muscle afferent neurons by their characteristic activation/deactivation kinetics, enhancement by the K_V7 activator retigabine, and block by K_V7 specific inhibitor XE991. The blocking of K_V7 current by different XE991 concentrations suggests that the K_V7 current is generated by both K_V7.2/7.3 (high affinity) and K_V7.5 (low affinity) channels. The K_V7 current was inhibited by 300 nM BK in neurons with diameters consistent with both group III and group IV afferents. The inhibition of K_V7 by BK could be a mechanism by which this metabolic mediator generates the EPR. Furthermore, our results suggest that K_V7 channel activators such as retigabine, could be used to reduce cardiac stress resulting from the exacerbated EPR in patients with cardiovascular disease.

NEW & NOTEWORTHY K_V7 channels control neuronal excitability. We show that these channels are expressed in muscle afferents and generate currents that are blocked by XE991 and bradykinin (BK). The XE991 block suggests that K_V7 current is generated by K_V7.2/3 and K_V7.5 channels. The BK inhibition of K_V7 channels may explain how BK activates the exercise pressor reflex (EPR). Retigabine can enhance K_V7 current, which could help control the inappropriately activated EPR in patients with cardiovascular disease.

INTRODUCTION

The exercise pressor reflex (EPR) is an important neural control mechanism involved in regulating cardiovascular function during exercise (1). Originating in skeletal muscle, this neural reflex is activated by exercise-induced signals to cause increases in cardiac output, arterial blood pressure, and ventilation. A dysfunctional EPR is thought to play a role in pathophysiological states such as heart failure and peripheral vascular disease (claudication) (2). Inadequate muscle perfusion produces ischemia that can inappropriately activate the EPR to increase the risk for myocardial infarction (MI) in these patients (3, 4).

The chemical and physical changes that occur within muscle during ischemia are sensed by the nerve terminals of group III (myelinated) and group IV (unmyelinated) afferent neurons (5, 6). Bradykinin (BK) is an endogenous proinflammatory, pain-inducing, vasodilating peptide that is released within contracting and ischemic muscle (7). Though BK is a known activator of the EPR (8), the mechanism through which bradykinin enhances group III/IV afferent excitability is incompletely understood.

K_V7 channels form a class of slowly activating/deactivating potassium channels that are vital to the control of neuronal excitability (9, 10), but it is currently unknown if these channels are expressed by muscle afferents. K_V7 channels have previously been shown to be inhibited by BK (11), which suggests a potential mechanism for BK-induce enhancement of the EPR.

We provide evidence for functional expression of K_V7 channels by muscle afferent neurons and further demonstrate that the K_V7 current is inhibited by BK in these neurons. Thus, inhibition of K_V7 channel activity by BK could be one mechanism by which the EPR is driven by muscle ischemia.

MATERIALS AND METHODS

Animals

All animal procedures were approved by the Institutional Animal Care and Use Committee and were consistent with the National Research Council Guide for the Care and Use of Laboratory Animals. Adult male Sprague–Dawley rats (200–400 g; Hilltop Lab Animals, Scottsdale, PA) were used in these experiments. The rats were housed in a US Department of Agriculture-approved, Association for Assessment and Accreditation of Laboratory Animal Care-certified animal care facility at a constant temperature of 24°C, under controlled 12:12-h light-dark cycles, and fed a standard rat chow diet and tap water ad libitum.

Labeling and Isolation of DRG Neurons

Adult Male Sprague–Dawley rats were anesthetized via intraperitoneal injection of ketamine (43.8 mg/kg), xylazine (8.6 mg/kg), and acepromazine (1.5 mg/kg). The calf area of both legs was shaved, and the left and the right gastrocnemius muscles were injected with 100 µL of 1.5% DiI (1,1′-dioctadecyl-3,3,3′,3′-tetra-methylindocarbocyanine perchlorate) in dimethyl sulfoxide (DMSO) (12).

The rats were euthanized 4–10 days postinjection by CO₂ inhalation, followed by decapitation using a laboratory guillotine (Kent Scientific Corp., Torrington, CT). The Lumbar 4 (L4) and L5 dorsal root ganglia (DRG) were isolated and dissociated in Earle’s balanced salt solution containing (in mg/mL): 0.7 collagenase, 1 trypsin, and 0.1 DNase at 37°C for 60 min. The dissociated neurons were washed in minimum essential media (MEM) containing 10% fetal bovine serum (FBS) and plated onto polylysine-coated glass coverslips (Thermo Fisher Scientific, St. Louis, MO). The isolated neurons were maintained overnight in a 5% CO₂ cell culture incubator at 37°C in MEM supplemented with 10% FBS and 1% penicillin-streptomycin and used within 12–24 h.

Electrophysiological Recordings from Muscle Afferent Neurons

The extracellular recording solution for most patch-clamp experiments contained (in mM): 5 KCl, 3.2 MnCl₂, 130 N-methyl-d-glucamine (NMG)·Cl, 10 NMG·4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), and 15 glucose, with pH = 7.4 and osmolarity = 350 mOsmol/L. For some experiments, the KCl concentration was decreased from 5 to 2 mM to increase the potassium driving force at −50 mV. The intracellular solution contained (in mM): 90 K·acetate, 20 KCl, 14 creatine·PO₄, 6 MgCl₂, 10 NMG·HEPES, 5 Tris·ATP, 10 NMG₂·ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA), 0.3 Tris₂·GTP with pH = 7.4 and osmolarity = 335 mOsmol/L. For some experiments, 180 µM phosphatidylinositol 4,5 bisphosphate diC8 (PIP2) was added to the internal solution, as K_V7 channels require PIP2 to maintain activity (13, 14) and its addition can reduce K_V7 current rundown (15). PIP2 and retigabine and XE991 were made up as stock solutions in DMSO and stored as aliquots at −30 C. For most experiments, the DMSO concentration in the external solution was ≤1% but was 2% for the experiments using 100 µM XE991. All solutions applied to a single neuron contained the same DMSO concentration so only the drug concentration changed when switching solutions between control and test. The potassium reversal potential (E_K) was calculated using the Nernst equation to be −79 mV with a room temperature of 23°C.

DiI-positive neurons were identified under a microscope using epifluorescence and ionic currents were recorded from those cells using the whole cell configuration of the patch-clamp technique with an Axopatch 200 A amplifier (Molecular Devices, Sunnyvale, CA). The currents were digitized using an ITC-18 A/D converter (Instrutech Corp, Port Washington, NY). Microelectrodes with a resistance of 3–7 MΩ (mean = 5 MΩ) were pulled from Schott 8250 glass (King Precision Glass, Claremont, CA) on a Sutter P-97 puller (Sutter Instruments, Novato, CA). The holding potential was set to either −30 or −20 mV (16). Experiments were controlled by a Macintosh computer (Apple Computer, Cupertino, CA) running S5 or F6 data acquisition software written by Dr. Stephen Ikeda (NIH, NIAAA, Bethesda, MD). All experiments were conducted at room temperature.

For measuring the time course of a drug response, the holding current was measured at either −30 or −20 mV (I_Hold), whereas the step current was measured at 5 ms into a step to −50 mV (I_1st) and at the end (I_End) of that 1-s voltage step. The difference between the current at the beginning (I_1st) versus the end (I_End) of the voltage step measures the slow closing of K_V7 channels (17). We refer to this current as deltaI (ΔI). The current-voltage relationship was recorded using 1-s voltage steps ranging from −20 to −100 mV from a holding potential of −20 mV. Step current measurements were the same as described for the time-course experiment. A holding potential of −30 mV was used for the retigabine experiments to better visualize the current enhancement, as −20 mV is close to maximal activation for K_V7 currents (Fig. 4B). Data were analyzed with IgorPro (WaveMetrics, Lake Oswego, OR) running on a Macintosh computer. Cell diameter was calculated from the cell capacitance as measured by the Axopatch circuitry, assuming a specific membrane capacitance of 1 µF/cm² and that the neuron was spherical (12).

Figure 4.

Retigabine (RTG) left-shifts the K_V7 I-V relationship. A: the current measured at the end of 1-s steps is plotted versus the step voltage for control (Cntl, upward triangles), 10 µM RTG (solid squares), and washout (WO, downward triangles). The average (solid circles) of Cntl and WO is also shown. B: the G-V relationships were derived from the I-V relationships shown in A to highlight the left-shift induced by RTG. The smooth lines are Boltzmann equations fits. The −80-mV point for the RTG G-V relationship is undefined since it is at the K_V7 current reversal potential for this cell.

The time course of the exchange of compounds between the pipet and neuron during whole cell recording can be estimated by an equation derived by Pusch and Neher (18):

$$\tau \text{ = 0.042}\hspace{0.17em}{\text{R}}_{\text{S}}{\text{M}}^{\text{1/3}}{\text{C}}^{\text{3/2}},$$

where τ is the time constant of exchange (s), R_S is the series resistance of the electrode (MΩ), M is the molecular weight of the compound (Da), and C is the capacitance (pF) of the neuron. We previously used this equation to estimate the exchange τ for ATP from the pipet (19). As Ca²⁺ release from intracellular stores is a proposed mechanism for receptor-induced inhibition of K_V7 current (20, 21), we estimated the exchange τ for EGTA to determine if it would be expected to disrupt our measurement of the BK response. This equation yielded an average exchange τ = 6.6 min based our average pipet resistant (5 MΩ) and cell capacitance (39 pF) for the 10 cells exposed to BK. As all BK responses were recorded within 4 min (mean = 1.96 min), it is likely that there was insufficient time for EGTA to interfere with the BK response. Consistent with this conclusion, Liu et al. (21) used a 12-min dialysis of sensory neurons with BAPTA (a more potent Ca²⁺ chelator than EGTA) to block the BK-induced inhibition of K_V7 current.

Statistics

All group data are presented as means ± standard deviation (SD). A Student’s t test was used to compare groups, and a P < 0.05 was determined to be statistically significant. “ns” indicates not significantly different.

Drugs and Chemicals

MEM, FBS, DiI, and penicillin-streptomycin were purchased from Life Technologies (Carlsbad, CA). Collagenase was purchased from Roche Molecular Biochemicals (Indianapolis, IN), and trypsin was from Worthington Inc (Lakewood, NJ). XE991 and retigabine were purchased from Alomone Labs (Jerusalem, Israel). PIP₂ was purchased from Echelon Bioscience Incorporated (Salt Lake City, UT). All other chemicals were obtained from Sigma-Aldrich (St. Louis, MO).

RESULTS

Identification of a Slowly Deactivating Potassium Current in Muscle Afferent Neurons

Currents were recorded from identified small and medium diameter muscle afferents neurons (DiI labeled) that could give rise to the EPR. Neurons were voltage clamped at a holding potential of −20 mV and deactivation currents were elicited via 1-s hyperpolarizing steps to voltages ranging from −20 to −100 mV (Fig. 1). Holding the neuronal membrane potential at a depolarized voltage affords the isolation of K_V7 channels by inactivating other K_V channels (22). Putative K_V7 current was identified kinetically as a noninactivating voltage-dependent potassium current that was active at the holding potential (−20 mV) and deactivated slowly at hyperpolarized voltages (Fig. 1A). As expected for a voltage-dependent potassium current, the putative K_V7 current deactivated faster as the step voltage became more hyperpolarized and the deactivating current reversed direction from outward (positive current) to inward (negative current) at voltages near the potassium equilibrium potential (E_K) (Fig. 1, B and C). The deactivation time constant (τ) measured from 6 neurons showed a monotonic decrease with voltage from ∼180 ms at −50 mV to ∼50 ms at −100 mV (Fig. 2A). The current reversal potential was determined by linear regression of the peak ΔI (Fig. 1C) measured from −100 to −80 mV to be −82 ± 3 mV (means ± SD; n = 6; Fig. 2B). E_K was calculated using the Nernst equation to be −79 mV (external K⁺ = 5 mM, internal K⁺ = 110 mM). Thus, the kinetic and permeation data support the identification of this as K_V7 current.

Figure 1.

A putative K_V7 current in muscle afferent neurons. A: the currents were generated using a current-voltage (I-V) protocol consisting of 1-s steps to voltages ranging from −20 mV to −100 mV (see protocol below the current traces). The holding potential was −20 mV and the interval between steps was 10 s. Note the characteristically slow deactivation of the K_V7 currents during the 1-s voltage steps, and the current reversal between −80 and −90 mV. *Where the current measurements shown in B were taken. B: the I-V relationship is from the currents shown in A. Current was measured at the beginning (I_1st; open boxes) and end (I_End, closed boxes) of the voltage steps and plotted vs. the step voltage. C: the delta current (ΔI) is the difference between the current at the beginning (I_1st) vs. end (I_End) of the 1-s voltage step and shows the isolated K_V7 current that deactivated during voltage steps < the holding potential.

Figure 2.

Kinetic and reversal potential data for the putative K_V7 current. A: a single exponential equation was fit to the deactivating current during the 1-s voltage steps ranging from −30 to −100 mV. The mean deactivation τ (± SD, n = 6 neurons) is plotted against the step voltage. B: the K_V7 current reversal potential was calculated from a linear regression of ΔI data between −100 and −80 mV (Fig. 1C) and the mean ± SD is shown (n = 6 neurons). The dashed line indicates E_K calculated from the Nernst equation. The reversal potential is not significantly different from E_K. E_K, potassium reversal potential; ΔI, difference between the current at the beginning vs. end of the 1-s voltage step; τ, deactivation time constant.

Retigabine

To support our electrophysiological identification of K_V7 current, we tested the effect of the K_V7 channel enhancing drug, retigabine (RTG) (23, 24). We applied 10 µM RTG (23, 24) and found a significant enhancement of K_V7 current at the holding potential of −30 mV (Fig. 3). The time course of the enhancement was rapid with the maximum effect reached within 10 s of initiating the application (Fig 3B). The step current at −60 mV was also increased by RTG, although the ΔI current was reduced (Fig. 3A). Both of these effects are consistent with the RTG-induced left shift in the K_V7 current-voltage relationship (I-V) (24, 25), which was verified by comparing the I-V relationships with and without 10 µM RTG (Fig. 4A). The conductance-voltage (G-V) relationship showed a monotonic increase in G that was maximal near −20 mV in control and −50 mV in RTG (Fig. 4B). Fitting the G-V relationship with the Boltzmann equation (Fig. 4B) yielded half-maximal voltage (V½) = −46.5 ± 2.3 mV and −74.8 ± 3.6 mV for control and RTG, respectively. The 10 µM RTG-induced ∼30 mV left-shift in the V½ was statistically significant (P < 0.05, n = 4). Of the five neurons exposed to RTG, two had diameters between 20–30 µm, two between 30–40 µm, and one had a diameter of >40 µm, which supports the enhancement of K_V7 channel activity in both group III and group IV afferents (12).

Figure 3.

The effect of retigabine on K_V7 current in muscle afferents. A: superimposed traces show K_V7 current in control (Cntl), during application of 10 µM retigabine (RTG), and upon washout (WO). The ΔI at −60 mV decreased from 164 pA in control to 32 pA in RTG. The dashed line indicates zero current. B: the plot shows the time course of RTG enhancement of the holding current at −30 mV. The step interval is 10 s. C: the bar graph shows the mean enhancement of K_V7 current from 5 muscle afferent neurons ± SD. *Significant enhancement (P < 0.05) by 10 µM RTG. ΔI, difference between the current at the beginning vs. end of the 1-s voltage step.

XE991 Inhibits the Potassium Current

As a second pharmacological test of the K_V7 current, we applied the K_V7 channel blocker XE991 (26–29). Application of 10 µM XE991 significantly inhibited ΔI by 57.4 ± 17.7% (n = 18, P < 0.05, Fig. 5). The holding current was also inhibited in the presence of XE991. This effect on holding current is expected for K_V7 current, as these channels do not inactivate and, therefore, generate a steady current at depolarized voltages (30). The XE991 time course reached maximal block within 20 s of application initiation (Fig. 5, A and B) and recovery, while partial, also appeared to be complete within 20–30 s following washout. The partial inhibition of ΔI was interesting, as the remaining ΔI in 10 µM XE991 also slowly deactivated like the control current (Fig. 5C), which suggested that K_V7 current was incompletely blocked. Indeed, single exponential equation fitting of ΔI in control and 10 µM XE991 (n = 10) produced deactivation τs = 148 ± 78 ms and 150 ± 126 ms for control versus XE991, respectively (ns).

Figure 5.

The specific K_V7 channel blocker XE991 partially inhibits the K_V7 current. A: the time course is shown for the effect of 10 µM XE991 on current measured at the holding potential (−20 mV, I_Hold) and at the beginning (I_1st) and end (I_End) of 1-s voltage steps to −50 mV. *Currents displayed in C. B: ΔI (see Fig. 1 legend) is shown to illustrate the effect of XE991 on the isolated K_V7 current. C: averaged currents from three control (Cntl) sweeps and three sweeps in 10 µM XE991 are shown (indicated by the asterisks in A). The dashed line indicates zero current. D: the mean inhibition of ΔI at −50 mV is shown along with the individual data points (n = 18). *Significant inhibition (P < 0.05).

Of the subunits that can comprise K_V7 channels, K_V7.2, 7.3, and 7.5 have been shown to be expressed in sensory neurons with K_V7.2 and 7.3 thought to form heterodimers (26, 29, 31–33). The published XE991 IC₅₀ for K_V7.2/7.3 heterodimers is 0.6 µM (34), whereas that for homomeric K_V7.2 and 7.3 channels is 1 µM (10). Interestingly, the reported XE991 IC₅₀ for K_V7.5 homomers is 65 µM (35). This led us to hypothesize that the slowly deactivating current (ΔI) remaining in 10 µM XE991 resulted from K_V7.5 channel activity, which should be blocked by 100 µM XE991. We initiated a second set of experiments in which we exposed muscle afferent neurons to 10 µM and then 100 µM XE991 (Fig. 6). For these experiments, 10 µM XE991 blocked ΔI by 47.1% ± 27.2%, whereas the block increased to 74.5% ± 18.3% in 100 µM XE991 (P < 0.05, n = 18, Fig. 6B). There were two neurons (31 and 41 µm diameters) recorded with little or no ΔI inhibition by 10 µM XE991, but showed strong inhibition (69% and 39%, respectively) by the higher concentration (Fig. 6, B and C). One neuron showed complete inhibition of ΔI by 10 µM XE991 (Fig. 6B).

Figure 6.

Additional block of K_V7 current by 100 µM XE991. A: averaged currents from 3 sweeps each in control (Cntl), 10 µM, and 100 µM XE991 are shown. The holding current at −20 mV and the slowly deactivating K_V7 current at –50 mV were partially blocked by 10 µM XE991 and more completely blocked by 100 µM XE991. The dashed line indicates zero current. B: the inhibition by both 10 µM and 100 µM XE991 is plotted for 18 muscle afferent neurons. The lines link the data points from individual neurons and the gray bars indicate the average. *Indicates that the inhibition by 10 µM XE991 is significant. **Indicates that the inhibition by 100 µM XE991 is significantly larger than that by 10 µM XE991 (P < 0.05). C: the percent inhibition of ΔI at −50 mV by 10 µM XE991 (n = 36) versus the diameter of the tested muscle afferent neurons is shown. Note that two neurons responded with little or no inhibition (31 and 41 µm diameters), but the K_V7 current in both neurons was blocked by 100 µM XE991 (see B). D: the percent increase of K_V7 current block by 100 µM XE991 over that produced by 10 µM XE991 is plotted versus muscle afferent diameter (n = 18).

Figure 6C shows the distribution of ΔI block by both 10 µM XE991 versus neuronal diameter. The majority of neurons had diameters between 20 and 40 µm, which we have argued is a size range that includes both group III (30–40 µm) and group IV (<30 µm) neurons (12). The percentage increase in block of K_V7 current block by 100 µM XE991 was plotted versus cell diameter to determine if neurons within a certain size range showed differential responses (Fig. 6D). However, this was not observed. Although the increased block by 100 µM XE991 was variable, the range was similar between group IV (<30 µm diameter) and group III (30–40 µm diameter) neurons.

Bradykinin Inhibits K_V7 Channels in Rat DRG

BK is a known activator of the EPR, but the mechanism through which BK enhances group III/IV afferent excitability is unknown. BK has previously been shown to inhibit K_V7 current in rat sympathetic neurons (11). Therefore, we were interested to see if BK would inhibit K_V7 current in muscle afferent neurons. Using the same voltage protocol used to examine the effect of XE991, 300 nM BK significantly inhibited the K_V7 current (Fig. 7). The time course shows that inhibition was complete 40–50 s after initiation of BK application (Fig. 7B), but the recovery was only partial due to rundown of the K_V7 current. The ΔI at −50 mV was inhibited by 64.7% ± 26.8% (P < 0.05, n = 10, Fig. 7C) and the holding current (steady-state current) at −20 mV was also inhibited (Fig. 7, A and B).

Figure 7.

Bradykinin (BK) inhibits K_V7 current in muscle afferents. A: superimposed traces of K_V7 current recorded in the presence (black) and absence (gray) of 300 nM BK. Each trace is an average of three sweeps. The current at −50 mV is inhibited as well as the current at the −20-mV holding potential. The dashed line indicates zero current. B: the time course of inhibition by 300 nM BK is shown for I_Hold (closed squares) and ΔI (open circles). *Points from which the currents in A were averaged. C: the means ± SD inhibition of ΔI by 300 nM BK (n = 10, P < 0.05). D: the percent inhibition of K_V7 current by 300 nM BK is plotted against the neuronal diameter. The different symbols show data from muscle afferent neurons recorded with (gray) or without (black) internal PIP2.

The use of internal PIP2 in some of our recordings could have impaired BK-induced inhibition, as receptor-induced inhibition can result from hydrolysis and depletion of PIP2 via G_q/11 activation of phospholipase C (20). However, we found no significant difference in the BK-induced inhibition between muscle afferent neurons recorded with (50.7% ± 32.8%, n = 3) or without (70.7% ± 24.1%, n = 7) internal PIP2 (Fig. 7D), which is consistent with the conclusion that the BK-induced inhibition depends more on inositol 1,4,5-trisphosphate (IP3)-induced Ca²⁺ release from internal stores than on PIP2 depletion (20, 21). In addition, all BK responses were recorded within 4 min of establishing the whole cell recording (mean = 1.95 min, n = 10). Based on our calculation, this interval is likely too short for large molecules such as EGTA to dialyze into cytoplasm to interfere with the BK response (see materials and methods).

To better understand the type of muscle afferent neurons (i.e., groups I–IV) responding to BK, the percent inhibition of ΔI was plotted against cell diameter (Fig. 7D). The recorded neurons evenly distributed between group IV (n = 3, <30 µm), group III (n = 4, 30–40 µm), and large (n = 3, >40 µm) diameter muscle afferents. Within each group, the size of the inhibition ranged from 25%–40% to 80%–100% (Fig. 7D). The variable nature of the BK-induced inhibition could reflect neurons with low bradykinin receptor expression or poor coupling between receptors and K_V7 channels due to faster than estimated dialysis of EGTA from the pipet.

DISCUSSION

Because K_V7 channels function to control neuronal excitability, we were interested to see if these channels are expressed in muscle afferent neurons, some of which give rise to the EPR. We identified a noninactivating voltage-dependent potassium current in muscle afferent neurons that showed slow deactivation kinetics expected for K_V7 channels (11, 26). Consistent with the literature, this potassium current was inhibited by the K_V7 channel blocker XE991 (17, 29, 36) and enhanced by RTG (24, 25). Similar to that observed in sympathetic and sensory neurons (11, 21), we found that BK inhibited the K_V7 current in small- to medium sized muscle afferent neurons (presumably group III and IV). Since K_V7 channels are expressed in the muscle afferents and sensitive to BK, the inhibition of K_V7 channels may be one mechanism by which BK can enhance afferent excitability to drive the EPR.

K_V7 Channels

K_V7 current (also called M-current) was first studied in bullfrog sympathetic neurons and was described as a noninactivating potassium current that slowly activates and deactivates in a time- and voltage-dependent manner (9, 37). K_V7 channels are active near normal membrane resting potential and remain active at subthreshold (depolarized) membrane potentials (9, 22). The unique voltage dependence of this channel allows the current to be isolated by voltage clamping the cells at depolarized potentials (e.g., −20 mV). In contrast to the rapid activation/deactivation kinetics of most other voltage-gated ion channels (<10 ms), K_V7 channels activate/deactivate slowly (9, 26). The properties of this channel allow the current to oppose depolarization and reduce excitability of neurons (9, 10, 26).

K_V7 channels are composed of four subunits that can come from five genes (KCNQ1-5), which make the K_V7.1–7.5 proteins (10, 31, 38). Of these, K_V7.2, 7.3, and 7.5 have been shown to be expressed in sensory neurons with K_V7.2 and 7.3 proteins reported to be the most abundant in small- to medium-sized sensory neurons (26, 29, 31–33). It has been surmised that the primary composition of the K_V7 channels in small nociceptor-like sensory neurons is most likely either homomeric K_V7.2 channels and/or heteromeric K_V7.2/7.3 channels (26). There is also evidence for the expression of K_V7.5 channels in small, unmyelinated sensory neurons (i.e., C-type and group IV) (39). We did not determine the subunit composition of K_V7 channels in muscle afferent neurons but noted differences in the block by XE991 that supports expression of both K_V7.2/7.3 and K_V7.5 channels in muscle afferent neurons (see XE991 Block of KV7 Current).

K_V7 Current in Sensory Neurons

K_V7 current has been recorded in dissociated rat neurons of the superior cervical ganglion (SCG) (11) and dorsal root ganglia (DRG) (17, 21, 26). In this study, we used the whole cell patch clamp technique to identify a noninactivating potassium current in isolated muscle afferent neurons that we determined resulted from K_V7 channel activity due to the current properties and pharmacology. The current was noninactivating, which resulted in a steady-state outward current at −20 mV that slowly deactivated upon stepping to more negative voltages. The slowly deactivating current reversed from outward to inward around −80 mV, and the calculated reversal potential was not significantly different from the Nernst potential for potassium.

The K_V7 current deactivation τ was previously evaluated using double exponential fitting (26). Using small, capsaicin-sensitive DRG neurons, the authors found the deactivation τ’s at −50 mV to be 74 ± 10 and 583 ± 134 ms. We evaluated the deactivation τ over a range of voltages using a single exponential equation, which has also been used to evaluate K_V7 current (40). We found single exponential fitting to provide less variable τ’s than a double exponential fit (not shown) over this voltage range. At −50 mV, the single exponential deactivation τ was 179 ± 17 ms (n = 4) for ΔI, but this value represents the weighted average of the two deactivation components. To compare with the previous report, we fit deactivation at −50 mV with a double exponential equation and found the τ’s to be 65 ± 17 and 418 ± 201 ms in the same four neurons, which are close to the values reported previously (26). The relative amplitude of the slow component was 52% ± 13% for the tested neurons, so the two components were almost equally balanced. Based on the current properties, we conclude that muscle afferent neurons express functional K_V7 channels.

XE991 Block of K_V7 Current

K_V7 channels can be “pharmacologically identified” by their sensitivity to XE991 (26, 29). We initially used 10 µM XE991, which has been previously used to study K_V7 current (26–29). However, the response was almost always a partial inhibition with a large variance among muscle afferent neurons. The deactivation τ at −50 mV was virtually identical between control ΔI and that in 10 µM XE991, which suggested that K_V7 current was only partially blocked. The published XE991 IC₅₀ for K_V7.2/7.3 heterodimers is 0.6 µM (34), while that for both the K_V7.2 and 7.3 homomers is 1 µM (10). The IC₅₀ for XE991 block of K_V7 current in DRG neurons was previously shown to be ∼0.26 µM (26), which is consistent with channels formed from K_V7.2 and 7.3 homomers or heterodimers. Interestingly in the Passmore paper (their Fig. 1), there was no increase in K_V7 current block between 1 and 10 µM XE991, but these concentrations failed to completely block the slowly deactivating current at −50 mV (26), as we have also observed. Based on the similarity of our results with those of previous investigations, we conclude that 10 µM XE991 blocked K_V7 channels comprised of K_V7.2 and 7.3 subunits alone or in combination. These channels with “high affinity” for XE991 comprised on average ∼50% of the total K_V7 current, but this varied from 0 to 100% among different muscle afferent neurons.

As mentioned in the K_V7 sections of the DISCUSSION, K_V7.5 subunits are expressed in sensory neurons (26, 39) and channels formed from this subunit could be the source of the K_V7 current that remained in 10 µM XE991. The IC₅₀ for XE991 block of K_V7.5 homomer channels is 65 µM and these subunits are not expected to combine with either K_V7.2 and 7.3 subunits to form K_V7 channels (35). If the IC₅₀ is similar for native K_V7.5 channels (low affinity), we would expect little or no block of the current with 10 µM XE991, but substantial block when the concentration was increased to 100 µM. In our experiments, the average block increased by 40 ± 29% in 100 µM XE991 (n = 18) over that measured in 10 µM, but again the increase varied widely across muscle afferent neurons with a range of 1 to 100%. It is notable that with an IC₅₀ = 65 µM, the block of K_V7.5 channels by 100 µM XE991 will be incomplete. Therefore, it is expected that K_V7 current would be incompletely blocked even at this high concentration, which is what we observed. Thus, we conclude that the XE991 “low affinity” K_V7 current is generated by K_V7.5 channels (on average ∼50% of the total K_V current). Based on the range of cell sizes, we conclude that both group III and IV afferent neurons express K_V7.2/7.3 and 7.5 channel variants.

Retigabine

An exaggerated EPR can be observed in patients with peripheral vascular disease (8). Since dysfunctional K_V7 channels result in the destabilization of the membrane potential and increased action potential firing frequency (41), activation of these channels can suppress neuronal excitability (10, 22, 42). Indeed, the K_V7 channel activator retigabine (RTG) is FDA approved for the treatment of partial seizures in adults (43), but it was recently removed from the US market by the manufacturer. We applied 10 µM RTG, which is a concentration commonly used to study K_V7 current (24–26, 29, 44), and found significant enhancement of the holding current at −30 mV as well as the current during a step to −60 mV. This enhancement resulted from a left-shift in the voltage-dependent activation of K_V7 current as previously described (24, 42). RTG has been reported to enhance current generated by all three K_V7 subunits known to be expressed in sensory neurons (24, 45, 46). Although the enhancement of K_V7.5 channel activity appears to require higher RTG concentrations than that of K_V7.2/7.3 channels, a concentration of 10 µM RTG is sufficient to enhance K_V7.5 channel activity (45).

BK and the EPR

BK is released in contracting muscles during exercise (7) and elicits a pressor response through unidentified mechanisms. We hypothesized that BK-induced inhibition of K_V7 current is a potential mechanism through which BK elicits a pressor response. Our data show that BK does inhibit K_V7 channels in muscle afferents, but the mechanism of this inhibition was not investigated. The BK-induced inhibition of K_V7 current is likely to be mediated via the bradykinin2 (B₂) receptor (47) since B₂ receptors mediate the inhibition of K_V7 current in sensory neurons (21) as well as the BK-induced activation of the EPR (4). The IC₅₀ for BK inhibition of K_V7 current was measured to be 60 nM in sensory neurons (21), and active gastrocnemius muscles at a 40% maximal workload can produce interstitial BK concentrations of 111 nM (7). Thus, the normal increase of the BK concentration in active muscles could sufficiently inhibit K_V7 channels via activation of B₂ receptors to increase action potential activity in group III and IV afferents, which may serve as part of the mechanism to elicit the EPR in healthy individuals.

Although the EPR is important to optimize cardiovascular function during exercise in healthy individuals, this reflex can also be inappropriately activated in patients with cardiovascular diseases that reduce blood flow to the muscle. Thus, diseases such as heart failure (48) and peripheral vascular disease (8) can sufficiently reduce blood flow to produce muscle ischemia and activate the EPR. The hyperactive EPR can produce a strain on the heart to increase the patient’s risk for myocardial infarction (6).

K_V7 channels are key mediators in the maintenance of the resting membrane potential and K_V7 channel dysfunction is associated with various neuronal hyperexcitability disorders (10, 29, 49). They are also regulators of sensory neuron afferent activity (44) and we showed them to be functionally active in group III and group IV neurons that likely mediate the EPR. Therefore, a K_V7 channel activator, such as RTG, could provide a treatment to reduce EPR hyperactivity in patients with cardiovascular disease by reducing group III and group IV afferent activity. However, such a drug would likely need to be more selective than RTG, which enhances the activity of all K_V7 channel isoforms (24, 45, 46). This broad activity played a role in the adverse effects that patients experienced including altered vision, urinary retention, somnolence, dizziness, difficulty in concentrating and QTc prolongation (43, 50), which eventually led the drug’s removal from the US market. However, a more selective K_V7 channel enhancing drug could inhibit the EPR activity to lower sympathetic efferent activity and reduce the risk of myocardial infarction in patients with cardiovascular disease (6).

GRANTS

This study was supported by funding from the Graduate Program Committee of the Kirksville College of Osteopathic Medicine, A.T. Still University of Health Sciences and a grant from the National Institute of Arthritis and Musculoskeletal and Skin Diseases (AR059397).

DISCLOSURES

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

AUTHOR CONTRIBUTION

A.B.W., K.Y.S., and K.S.E. conceived and designed research; A.B.W., K.Y.S., and K.S.E. performed experiments; A.B.W., K.Y.S., and K.S.E. analyzed data; A.B.W. and K.S.E. interpreted results of experiments; A.B.W. and K.S.E. prepared figures; A.B.W., K.Y.S., and K.S.E. drafted manuscript; A.B.W., K.Y.S., and K.S.E. edited and revised manuscript; A.B.W., K.Y.S., and K.S.E. approved final version of manuscript.