BK channel β1 subunits regulate airway contraction secondary to M2 muscarinic acetylcholine receptor mediated depolarization

Non-technical summary

Parasympathetic nerve activation of M3 and M2 muscarinic acetylcholine receptors initiates and modulates calcium release from the sarcoplasmic reticulum to control airway smooth muscle contraction. Here we investigate M2 acetylcholine receptors that also contribute to contraction through depolarization and recruitment of voltage-dependent calcium channels (VDCCs). We find that the calcium- and voltage-activated potassium channel (BK channel) and its β1 accessory subunit are important proteins that oppose M2-mediated contraction of airway smooth muscle. BK channels contribute to a negative baseline membrane voltage from which M2-mediated depolarization only weakly activates VDCCs. The role of BK β1 to oppose M2 signalling is evidenced by a greater than fourfold increase in the contribution of L-type VDCCs to contraction that otherwise does not occur with M2 receptor antagonist or with β1 containing BK channels. These findings provide a better understanding of how cholinergic second messenger signalling impinges on voltage-dependent mechanisms and excitation–contraction coupling of smooth muscle.

Abstract

Abstract

The large conductance calcium- and voltage-activated potassium channel (BK channel) and its smooth muscle-specific β1 subunit regulate excitation–contraction coupling in many types of smooth muscle cells. However, the relative contribution of BK channels to control of M2- or M3-muscarinic acetylcholine receptor mediated airway smooth muscle contraction is poorly understood. Previously, we showed that knockout of the BK channel β1 subunit enhances cholinergic-evoked trachea contractions. Here, we demonstrate that the enhanced contraction of the BK β1 knockout can be ascribed to a defect in BK channel opposition of M2 receptor-mediated contractions. Indeed, the enhanced contraction of β1 knockout is eliminated by specific M2 receptor antagonism. The role of BK β1 to oppose M2 signalling is evidenced by a greater than fourfold increase in the contribution of L-type voltage-dependent calcium channels to contraction that otherwise does not occur with M2 antagonist or with β1 containing BK channels. The mechanism through which BK channels oppose M2-mediated recruitment of calcium channels is through a negative shift in resting voltage that offsets, rather than directly opposes, M2-mediated depolarization. The negative shift in resting voltage is reduced to similar extents by BK β1 knockout or by paxilline block of BK channels. Normalization of β1 knockout baseline voltage with low external potassium eliminated the enhanced M2-receptor mediated contraction. In summary, these findings indicate that an important function of BK/β1 channels is to oppose cholinergic M2 receptor-mediated depolarization and activation of calcium channels by restricting excitation–contraction coupling to more negative voltage ranges.

Introduction

Physiologically, airway smooth muscle (ASM) contraction is primarily mediated by activation of M2 and M3 muscarinic cholinergic receptors that couple to G_i/o and G_q receptors, respectively (Eglen et al. 1996; Hall, 2000; Janssen, 2002; Ehlert, 2003a; Zhou et al. 2008). M3/G_q-coupled receptor signalling leads to calcium release from sarcoplasmic reticulum IP₃ receptors to activate contraction (Berridge, 1993; Daykin et al. 1993; Luo et al. 1999). M2/G_i/o activation has a modulatory role on contraction by inhibition of adenylate cyclase and decrease of cAMP levels that enhance contraction (Sankary et al. 1988; Widdop et al. 1993; Zhou et al. 2008). These well established pathways constitute the so called ‘pharmaco-contraction coupling’ of airway smooth muscle that distinguish it from other smooth muscle cells where contraction is more dependent on depolarization and calcium influx through voltage-dependent calcium channels (so called ‘excitation–contraction coupling’) (Somlyo & Somlyo, 1968).

Despite a prominent role of pharmaco-contraction coupling, airway smooth muscle does express components for excitation–contraction coupling (Farley & Miles, 1978; Zhou et al. 2008). Cholinergic agonists activate a number of depolarizing cationic and anionic currents that can initiate excitation–contraction coupling in parallel with pharmaco-contraction coupling (Benham et al. 1985; Inoue & Isenberg, 1990; Janssen & Sims, 1992; Sims, 1992; Lee et al. 1993; Wang et al. 1997a). In addition, airway smooth muscles express L-type voltage-dependent calcium channels that can contribute to voltage-dependent calcium influx (Karaki et al. 1997; Janssen, 2002). Opposing voltage-dependent calcium influx is the voltage- and calcium-activated potassium channel (BK channel). Coincident calcium rise and depolarization activate BK channels, which in turn repolarize the membrane and deactivate voltage-dependent calcium channels (Kaczorowski et al. 1996; Gribkoff et al. 1997; Calderone, 2002; Ghatta et al. 2006; Zhou et al. 2008). In ASM, BK channel function requires the accessory β1 subunit that enhances BK channel opening. Knockout of the β1 subunit increases airway contractions to a similar extent as observed with pharmacological blockade of BK channels (Semenov et al. 2006). Although ASM is not thought to undergo large depolarizations, nor does blocking L-type voltage-dependent calcium channels appear to have a large relaxant effect (Barnes, 1985; So et al. 1986), blockage of BK channels does enhance the cholinergic-induced contractions of airway (Kume et al. 1995; Semenov et al. 2006). Thus, BK/β1 channel control of membrane voltage may preclude significant activation of voltage-dependent calcium channels during cholinergic-evoked contractions.

Several studies indicate that M2 receptors may play a prominent role in mediating airway smooth muscle excitation–contraction coupling (Widdop et al. 1993; Kume et al. 1995; Wang et al. 1997a; Zhou et al. 2008). The role of BK channels downstream of M2 receptor activation is controversial. On the one hand, various signalling molecules downstream of M2 receptor activation are thought to inhibit BK channels and presumably increase contraction force. For example, M2 stimulation of G_i/o protein inhibits adenylyl cyclase thus reducing cAMP levels and protein kinase A activation (Sankary et al. 1988; Zhou et al. 2008). Because BK channels are activated by protein kinase A phosphorylation (Zhou et al. 2001), M2 receptor stimulation would result in decreased BK channel activity in ASM. Moreover, M2 receptors can inhibit BK channels through direct G_i/o interactions (Kume et al. 1992; Zhou et al. 2008). Thus, even if Ca²⁺ influx through L-type Ca²⁺ channels has a role in maintaining cholinergic-induced contraction, the relative role of BK channels remains questionable due to inhibition by M2 signalling. On the other hand, inhibition of BK channels by specific antagonists depolarizes the ASM plasma membrane (Sausbier et al. 2007) and increases contractile force (Semenov et al. 2006), arguing that BK channels are active downstream of M2 receptor activation.

Although several studies have identified various signals through which M2 receptors could inhibit BK channels, the functional relevance to airway muscle excitation–contraction coupling has not been ascertained. Here we have utilized a combination of pharmacology, knockout animals, and ionic substitution techniques to gauge the functional relevance of M2 receptor signalling and BK channels activity on ASM contraction. Our results indicate that BK/β1 channels specifically oppose contractions occurring as a consequence of M2 signalling. We find that M2 receptors and BK channels act in an opposing fashion on voltage to affect recruitment of voltage-dependent calcium channels.

Methods

Tissue preparations and contraction recordings

The methods were similar to those described previously (Semenov et al. 2006). The BK channel β1 subunit knockout mice are congenic by seven generations of inbreeding to the C57BL/6 line of The Jackson Laboratory (Bar Harbor, ME, USA; strain C57BL/6J) and maintained as homozygous lines. Control animals used in these studies were the background C57BL/6J mouse strain from The Jackson Laboratory. All animal procedures were reviewed and approved by the University of Texas Health Science Center at San Antonio Institutional Animal Care and Use Committee. For tracheal constriction studies, animals were deeply anaesthetized with isoflurane and then killed by cervical dislocation. Trachea were quickly removed and dissected clean of surrounding tissues in ice-cold normal physiological saline solution (PSS). The tracheal tube was cut below the pharynx and above the primary bronchus bifurcation. Two metal wires, attached to a force transducer and micrometer (Radnoti, LLC), were threaded into the lumen of the trachea. The trachea was placed into an organ bath oxygenated by an O₂–CO₂ mixture (95% O₂–5% CO₂), at 37°C. Resting tension was continuously readjusted to 1 g for 1 h and then challenged with 67 mm K⁺-containing PSS twice or more until reproducible contraction responses were achieved. Subsequent experimental challenges with drugs were normalized to the constriction response to the 67 mm K⁺-PSS solution. In 67 mm K⁺-PSS, the potassium reversal potential is depolarized and therefore potassium currents are unlikely to play a role in controlling membrane potential and contraction tone. This was consistent with our earlier findings that WT and β1-KO mice exhibit no significant difference in contraction in response to 67 mm K⁺-PSS, but show a significant difference to cholinergic stimulation when bathed in normal potassium PSS (Semenov et al. 2006). For experiments that involve two cholinergic challenges, we found that the second challenge does not show significant fatigue (P = 0.19, n = 9 for WT, P = 0.18, n = 9 for KO, Student's paired t test). On average, the second challenge shows 0.95 ± 1% and 0.94 ± 1% reduction in response from that of first challenge for WT and KO trachea, respectively. Normal PSS used was (mm): 119 NaCl, 4.7 KCl, 2.0 CaCl₂, 1.0 KH₂PO₄, 1.17 MgSO₄, 18 NaHCO₃, 0.026 EDTA, 11 glucose, and 12.5 sucrose. The pH of the solution was adjusted to 7.35 by 95% O₂–5% CO₂ mixture. The 67 mm K⁺-PSS utilized reduced sodium (56.7 mm NaCl) to maintain proper osmolarity. ‘Low K⁺ PSS’ was used to measure contractions in relatively hyperpolarized conditions. Low K⁺ PSS consisted of 1.0 mm potassium derived from the KH₂PO₄ in the PSS, with no added KCl. Normal PSS has a total 5.7 mm K⁺ derived from 4.7 mm KCl plus 1.0 mm KH₂PO₄. All other ingredients were unchanged.

Tracheal smooth muscle cell isolation and sharp electrode recordings

The trachea was isolated as described above. The dorsal muscle layer was cut away from the hyaline cartilage rings and minced into ∼1 mm pieces in Ca²⁺-free Hepes-buffered Krebs solution (140 mm NaCl, 4.7 mm KCl, 1.13 mm MgCl, 10 mm Hepes, 10 mm glucose, pH 7.3). After addition of 2.5 U ml⁻¹ papain (MP Biomedicals, Solon, OH, USA), 1 mg ml⁻¹ BSA fraction V and 1 mg ml⁻¹ dithiothreitol, tracheal smooth muscle (TSM) tissue was mixed at 37°C on a rocking platform (250 moves min⁻¹) for 20 min. Tissue was washed once with the Ca²⁺-free Krebs solution and digested with 12.5 U ml⁻¹ of type VII collagenase (Sigma-Aldrich Corp.) for 10 min on a rocking platform at 37°C. Digested pieces of tissue were washed three times in Ca²⁺-free Krebs-BSA solution by spinning (750 g for 2 min), and gently triturated up to 5 min to disburse single tracheal myocytes. TSM cells were stored on ice in Ca²⁺-free Krebs-BSA solution and used on the same day. A small (50 μl) aliquot of the solution containing isolated tracheal myocytes was placed in an open 1.0 ml perfusion chamber mounted on the stage of an inverted microscope. The TSM cells were allowed to adhere to the glass bottom of the chamber for 20 min and then were perfused (0.5–1 ml min⁻¹) with the PSS containing 5.7 or 1.0 mm K⁺ (PSSs identical to those used for contraction studies). The estimated K⁺ Nernst potentials are −72 mV and −116 mV, respectively, for 5.7 or 1.0 mm K⁺ external PSS. This is based on an intracellular K⁺ concentration of 99 mm that has been reported for bovine trachea (Kirkpatrick & Bullock, 1986). The more depolarized resting voltages that were measured (Fig. 5, −44 and −53 mV, for normal and low K⁺ PSS, respectively) are likely to reflect a contribution to resting voltage by permeability to other ions such as sodium and chloride, and also a relatively high intracellular chloride concentration that is unique to trachea and other smooth muscle (120 mm intracellular chloride resulting in an estimated E_Cl of –2 mV in our PSS (Kirkpatrick & Bullock, 1986; Chipperfield & Harper, 2000). pH of the solution was maintained at 7.35 by continuous oxygenation with 95% O₂–5% CO₂ mixture and temperature was maintained at 37°C using an automatic temperature controller (TC-324B, Warner Instruments, LLC, Hamden, CT, USA).

Figure 5. Tracheal relaxation by M2 receptor antagonism and calcium channel block is voltage dependent.

A, typical cholinergic (0.5 μm carbachol) contractions and M2 antagonism relaxation (100 nm AF-DX) in the absence (left panel) and presence (right panel) of nifedipine (10 μm) in normal PSS. B, typical contraction as in A but in 1 mm K⁺ PSS. C, average contraction. D, average relaxation caused by the indicated antagonists. E, nifedipine-sensitive, M2 antagonism relaxation measured as the difference of M2 antagonism (AF-DX) relaxation in the absence and presence of 10 μm nifedipine. One-way ANOVA in combination with post hoc Tukey's HSD test was used to statistically analyse experimental data. Error bars represent standard error of the mean, **P≤ 0.01, ***P≤ 0.001, n = 6 for all experimental groups. Red (KO) and blue (WT) lines show the amplitude of relaxation by M2 antagonism before and after perfusion of tracheas with 10 μm nifedipine. Contraction traces are normalized to the contractions induced by 67 mm KCl.

Membrane potentials (V_m) of the single TSM cells were measured using the sharp electrode technique with an EPC-9 amplifier (HEKA Instruments Inc., Bellmore, NY, USA). Sharp pipettes were pulled from borosilicate capillary glass (1B150F-4, WPI) to a resistance 20–60 MΩ using a Sutter P-87 pipette puller (Sutter Instrument Co., Novato, CA, USA). Pipettes were filled with the mix of 2 m KCl and 1% agarose Type IX-A to increase viscosity and reduce leak of the pipette solution into the cell. Resting membrane potentials were recorded for a minimum of 2 min to ensure a stable measurement. Impalement was apparent as a sudden drop of voltage. Recorded files were than exported to the analysis software (Igor; WaveMetrics, Inc., Lake Oswego, OR, USA). Voltage for each cell represents the average voltage during a 30 s time window.

Data analysis

Igor 5, KaleidaGraph 4.1.1 (Synergy Software, Reading, PA, USA) and Microsoft Excel 2007 were used for statistical analyses. Significance was determined with Student's t test for paired or unpaired data as appropriate. In the cases when more than two experimental groups were compared, a one-way ANOVA was applied to determine variability among groups. A post hoc Tukey's HSD test was used to compare individual groups. For comparison of two groups with multiple factors, we used a two-way ANOVA to distinguish if the groups were different. The effects were deemed significant when a P < 0.05 was obtained. The results are expressed as the means ± standard error of the mean where applicable.

Results

M2 antagonist treatment eliminates enhanced airway contraction of BK channel β1 Knockout

Previously, we showed that gene knockout of the BK channel's accessory β1 subunit increases the constrictor response to cholinergic agonist (Fig. 1A, summarized in Fig. 1B) (Semenov et al. 2006). ASM possesses both M2 and M3 receptors and in this study we wished to ascribe the contraction pathway downstream of the M2 receptor activation. 11-[[2-[(Diethylamino)methyl]-1-piperidinyl]acetyl]-5,11-dihydro-6H-pyrido[2,3-b][1,4]benzodiazepin-6-one (AF-DX 116 or AF-DX, Tocris Bioscience) is a well-characterized competitive inhibitor of M2 muscarinic receptors with a 10-fold greater affinity for M2 than M3 receptors (K_i 64 and 786 nm, respectively (Hammer et al. 1986). Application of 100 nm AF-DX effectively abolished the difference in contraction between WT and β1 KO mice (Fig. 1A and B). This suggests that M2 receptors mediate the enhanced contraction when BK channel function is diminished.

Figure 1. M2 receptor antagonism eliminates the difference in cholinergic evoked contraction force between β1 KO and WT tracheas.

A, a typical contraction stimulated by 0.5 μm carbachol (CCh). Application of 100 nm AF-DX 116 (AFDX) relaxes KO and WT tracheas to similar contraction forces. B, average data for A. C, a typical contraction in KO mouse trachea stimulated by 0.5 μm carbachol and relaxed with 100 nm AF-DX 116. The paired experiment was repeated with pre-incubation of M3 receptor antagonist J 104129 fumarate (J 104129) for 10 min before a second application of carbachol and AF-DX. Relaxation amplitude with AF-DX did not change in the presence of J 104129 (69 ± 9 before and 67 ± 4 after treatment J 104129; P = 0.4 according to paired t test). D, average data for C with regard to contraction (left panel) and relaxation (right panel). Contraction forces were normalized to the contractions induced by 67 mm KCl. CCh is 0.5 μm carbachol; AF-DX is 100 nm AF-DX 116; J 104129 is 5 nm J 104129 fumerate. Error bars represent standard error of the mean. **P < 0.01, n = 6 for B; n = 9 for D.

To ensure that the relaxant effect of AF-DX is specific to M2 but not M3 receptor inhibition, we utilized the specific M3 antagonist (αR)-α-cyclopentyl-α-hydroxy-N-[1-(4-methyl-3-pentenyl)-4-piperidinyl]benzeneacetamide fumarate (J 104129). J 104129 displays ∼120-fold higher selectivity to M3 receptors than to M2 receptors (K_i values are 4.2 and 490 nm for M3 and M2, respectively) (Mitsuya et al. 1999a,b, 2000). Although M2 receptors are historically regarded as modulating M3 receptor-initiated contractions (Fryer & Jacoby, 1998; Ehlert, 2003b), 0.5 μm carbachol application nevertheless contracted trachea during M3 receptor antagonism with J 104129 (5 nm, Fig. 1C and D). These findings are consistent with M3 receptor knockout studies that indicate M2 receptors can contribute to airway contractions in the absence of M3 receptor activation (Stengel et al. 2002; Struckmann et al. 2003). Importantly, the magnitude of relaxation evoked by 100 nm of AF-DX was not significantly altered by pretreatment with J 104129 (69.7 ± 9 before and 67.5 ± 4 with J 104129) (Fig. 1C and D). These results indicate that AF-DX antagonist effects can be ascribed to specific M2 receptor inhibition.

We investigated relaxation induced by M2 antagonist further using a dose–relaxation protocol for contractions produced by low and high cholinergic agonist concentrations (Fig. 2A and B, 0.5 μm and 3 μm carbachol, respectively). As expected for a competitive antagonist, the threshold for AF-DX effects strongly depended on cholinergic agonist concentration. Upon increase of carbachol concentration from 0.5 to 3 μm, the threshold concentration of M2 antagonist that produced relaxation shifted from 0.05 to 0.25 μm (compare Fig. 2A and B). At both concentrations of carbachol, the threshold concentrations of AF-DX produced a much greater relaxation of KO than WT tracheas. M2 antagonism effectively eliminated the difference in contractile force between WT and KO tracheas at low concentrations where AF-DX is presumably more specific for M2 than M3 receptors (Fig. 2A and B, summarized in Fig. 2C and D). Higher concentrations of AF-DX showed no further relaxation in KO relative to WT. In summary, these data suggest that enhanced airway contraction caused by β1 knockout is downstream of M2 signalling.

Figure 2. M2 antagonist AF-DX 116 relaxation of β1 KO trachea depends on carbachol concentrations.

A and B, typical concentration–relaxation curves obtained by cumulative increases in AF-DX 116 concentration during contractions stimulated by 0.5 μm carbachol (A) or 3 μm carbachol (B). KO and WT tracheas were pretreated with carbachol for 10 min until contractions reached a plateau. Cumulative increases in concentrations of AF-DX were obtained at 5 min intervals. C and D, average concentration–relaxation relationships for contractions of tracheas stimulated by 0.5 μm (C) and 3 μm (D) of carbachol. On the X-axis the log10 molar concentration of AF-DX is plotted; on the Y-axis force normalized to the contraction induced by 67 mm KCl is plotted. Error bars represent standard error of the mean. **P < 0.01, n = 6 for all experimental groups.

M2 receptors show significant desensitization (Eglen et al. 1996). Similarly, relaxation evoked by M2 antagonist also exhibited a time-dependent reduction that was more apparent at higher concentrations of AF-DX. For example, relaxation induced by 1–2 μm AF-DX (from tracheas pre-contracted in 3 μm carbachol) demonstrate a time-dependent decline in relaxation (Fig. 2B). This time-dependent loss of relaxation following M2 antagonism also occurs after pre-incubating tracheas with AF-DX for long periods preceding administration of carbachol (data not shown). Therefore, to maintain AF-DX specificity to M2 receptors and minimize the apparent decline of relaxation, subsequent studies used 100 nm of AF-DX applied after a plateau of contraction with 0.5 μm carbachol was reached.

An additional observation was oscillatory contractions in KOs evoked by M2 or M3 antagonists during low but not high concentrations of carbachol. Oscillations were prominent in 0.5 μm carbachol with either 100 nm of M2 antagonist AF-DX (Fig. 2A) or 5 nm of M3 antagonist J 104129 (Fig. 1C). Increases in carbachol concentration to 3 μm disabled oscillatory activity during contraction. Oscillations did not occur in the absence of muscarinic antagonists or in WT tracheas. Thus, blocking muscarinic signalling combined with reduced BK channel function in the β1 KO resulted in increased phasic contractions. One may conclude that BK channels are involved in stabilizing airway tonic contractions when either M2 or M3 pathways are perturbed. In these and subsequent experiments where oscillations occurred, we used time-averaged contractile tension (∼3 min) as a quantitative measure of contraction.

Relaxation of trachea by M2 antagonist does not occur via relief of BK channel inhibition

Given that M2 signalling has been reported to inhibit BK channels (Kume et al. 1992; Zhou et al. 2008), one may hypothesize that tracheal relaxation in β1 KO by M2 antagonism may be mediated by relief of BK channel inhibition. This is a concern for the β1 KO that enhances contraction through a reduction of BK channel open probability, rather than complete elimination of BK current (Semenov et al. 2006). A relaxation mechanism via relief of inhibition predicts that blocking BK channel should reduce relaxation induced by M2 antagonism. We examined this possibility by comparing M2 antagonist-induced tracheal relaxation in the absence and presence of the BK channel blocker paxilline (Sigma Aldrich, St. Louis, MO, USA) (1.0 μm). In contrast to the above prediction, the results of paired experiments (before and after application of paxilline) demonstrate that M2 antagonist-induced relaxation of KO trachea does not require BK channel activity (Fig. 3A, summarized in Fig. 3B). Although this does not exclude the possibility that M2 receptor activation inhibit BK channels, it suggests that a large component of relaxation due to M2 receptor antagonism occurs independent of BK channels. As well, the similar amplitude of relaxation in β1 KO tracheas and WT, paxilline-treated tracheas (Fig. 3B) suggests that compensatory effects in BK channels do not affect the enhanced M2 antagonist-induced relaxation in the KO.

Figure 3. Tracheal relaxation by M2 receptor antagonist is not due to relief of BK channel inhibition.

A, comparison of M2 antagonist (AF-DX) relaxation without (left panel) and with BK channel block with paxilline (right panel). B, summarized data of paxilline effect on M2 antagonism relaxation. One-way ANOVA was used in combination with post hoc Tukey's HSD test to statistically analyse experimental data. M2 antagonist relaxation does not significantly change after paxilline application in KO TSM (69.4 ± 1 before and 79.7 ± 3 after paxilline application, P = 0.93). In WT TSM M2 antagonism relaxation significantly increases (28.9 ± 4 before and 81.5 ± 21 after paxilline application, P = 0.01). There was no significant difference (P = 0.98) between KO and WT relaxation in tracheas pre-treated with paxilline. Error bars represent standard error of the mean, **P < 0.01, n = 5 KO, 6 WT.

Relaxation of β1 KO trachea by M2 antagonist occurs by reducing the contribution of L-type voltage-dependent calcium channels

The observation that M2 receptor antagonism normalizes contractions between β1-KO and WT tracheas is reminiscent of previous studies showing that L-type Ca²⁺ channel block eliminated enhanced contraction of β1-KO tracheas (Fig. 4A, Semenov et al. 2006). The contribution of L-type Ca²⁺ channels can be evaluated using nifedipine, a relatively specific blocker for L-type but not T-type Ca²⁺ channels (Furukawa et al. 2009), which are also expressed in airway smooth muscle (Janssen 1997). Like M2 receptor antagonism, relaxation following Ca²⁺ channel block with nifedipine occurs at low and high carbachol concentrations (Fig. 4B). Interestingly, the amplitude of relaxation following block of L-type Ca²⁺ channels is of equal magnitude at both low and high carbachol concentrations (Fig. 4C). As well, the increase in contraction in β1 KO relative to WT is of similar magnitude at both low and high carbachol concentrations (Fig. 4C). These results suggest that L-type calcium channels are maximally recruited and contribute a more substantial component to contraction at low concentrations of cholinergic agonist. Similarly, the increase of the nifedipine-sensitive contraction due to β1 KO is also maximal at low cholinergic agonist.

Figure 4. L-type calcium channel antagonist nifedipine eliminates the difference in cholinergic evoked contraction between β1 KO and WT tracheas.

A, contractions were stimulated with 3 μm carbachol (top panel) and 0.5 μm carbachol (bottom panel) before (left panels) and after (right panels) nifedipine administration. Application of 10 μm nifedipine relaxed KO and WT tracheas to similar contraction forces. B, average data for A. Relative difference in contraction force between WT and KO is maintained in low and high carbachol. C, nifedipine relaxation in high (3 μm) and (0.5 μm) low carbachol. Contractions and relaxations were normalized to the contractions induced by 67 mm KCl. One-way ANOVA in combination with post hoc Tukey's HSD test was used to statistically analyse experimental data. Error bars represent standard error of the mean. **P < 0.01, n = 6 for both KOs and WTs at 0.5 μm carbachol; n = 11 for KOs and n = 9 for WTs at 3 μm carbachol.

We tested the possibility that M2 antagonism reduces β1-KO contraction by reducing Ca²⁺ influx via L-type calcium channels. We compared relaxations following M2 antagonism in the absence and presence of the L-type Ca²⁺ channel blocker nifedipine. Figure 5 shows that contraction of both WT and KO tracheas is markedly reduced by nifedipine (Fig. 5A, compare left and right panels; mean data are summarized in Fig. 5C). We measured the relative relaxation produced by nifedipine as an estimate of L-type voltage-dependent calcium channel contribution to contraction using paired experiments (Nf, Fig. 5D). Consistent with previous studies, the larger relaxation in KO as compared with WT tracheas suggested that KO tracheas have a greater contribution of L-type voltage-dependent calcium channels to contraction (Fig. 5D). A similar effect was observed with M2 antagonism (AF-DX, Fig. 5D) suggesting that M2 signalling also confers a larger contraction in KO than WT tracheas presumably due to a larger recruitment of nifedipine-sensitive L-type Ca²⁺ channels. In contrast, pretreatment with nifedipine reduced M2 antagonism relaxations in the KO tracheas to similar values as observed in WT tracheas (Fig. 5A, and summarized in Fig. 5D, Nf+AF-DX). Indeed, subtraction of the paired experiments (M2 antagonist-relaxation without versus with nifedipine pretreatment) provides an estimate of the so-called nifedipine-sensitive component of relaxation induced by M2 antagonism. The nifedipine-sensitive component is presumably the component of M2 antagonist relaxation that depends on activity of voltage-dependent calcium channels and thereby can be abolished by nifedipine pretreatment. KO tracheas have an approximately 4.5 times greater nifedipine-sensitive component of M2 antagonist relaxation (Fig. 5E, 43.6 ± 6 in KO as compared with 9.3 ± 4 in WT P < 0.001). In contrast, the nifedipine-insensitive component of M2 antagonist relaxation is more similar between KO and WT (23 ± 1 versus 27 ± 1, respectively). Thus, the normalization of KO contraction by M2 receptor inhibition is largely due to reduced contribution of voltage-dependent calcium channels.

Relaxation of trachea by M2 antagonism is dependent on membrane potential of airway smooth muscles

The above observation that M2 receptor antagonism reduces the contribution voltage-dependent calcium channels suggests that M2 receptor activation leads to depolarization of tracheal smooth muscle. Thus, inhibition of M2 receptors may reduce membrane potential differences that otherwise cause an enhanced contraction in the KO trachea. In order to evaluate the effect of membrane potential on contractions we compared contractions in normal PSS (5.7 mm K⁺) to contractions in low K⁺ PSS (1.0 mm K⁺). The altered potassium concentration shifts the potassium equilibrium potential to a hyperpolarized membrane voltage. The effect of the external solutions on membrane voltage was confirmed by sharp electrode recording of isolated tracheal smooth muscle cells (TSMCs). The membrane potential of KO tracheal smooth muscle cells in normal K⁺ PSS (5.7 mm) was significantly depolarized (Fig. 6, 11 mV difference in voltage, P < 0.01; WT is –44 ± 2 mV, KO is –33 ± 2 mV) relative to that of the WT TSMCs. This depolarization is consistent with a greater contribution of voltage-dependent calcium channels in KO contraction (Fig. 5). Further, the application of BK channel blocker paxilline resulted in depolarization of WT to KO levels (Fig. 6). This indicates that BK channels affect resting membrane potential of TSMC. There was no significant effect of paxilline on KO TSMCs suggesting that the β1 KO mice eliminate BK channel function with regard to control of TSMC resting membrane voltage. In contrast, 1.0 mm K⁺ PSS significantly hyperpolarized and eliminated the difference in membrane voltage between KO and WT TSMCs at these conditions (Fig. 6, WT is −53 ± 5 mV, KO is −52 ± 6 mV).

Figure 6. External potassium concentrations alter resting membrane potentials in tracheal smooth muscle.

Membrane potentials of acutely isolated tracheal smooth muscle cells (TSMCs) from β1 KO and WT mice in 1 mm external potassium, 5.7 mm (normal) potassium and 5.7 mm (normal) potassium with paxilline block (PX) of BK channels. Results were statistically analysed by one-way ANOVA to determine variability among the groups with the following post hoc Tukey's HSD test to compare individual groups. In normal PSS (5.7 mm K⁺) β1 KO membrane was depolarized to V_m=−32.8 ± 2 mV, n = 33, in comparison to WT membrane potential: V_m=−44.0 ± 2 mV, n = 40 (P = 0.0082). In 1 mm K⁺ PSS, membrane potentials were hyperpolarized in both WT (V_m= 52.5 ± 5 mV, n = 14) and KO (V_m= 52.2 ± 5 mV, n = 6) and no statistical difference were found between KO and WT (P = 1). Application of 1 μm paxilline significantly depolarized WT TSMCs to V_m=−33.5 ± 3 mV (n = 29, P = 0.022) while in KO cells V_m=−30.5 ± 2 mV (n = 22, P = 0.98), and no statistically significant difference was observed in comparison to normal PSS. Also, there was no statistically significant difference of membrane potential in the presence of paxilline between KO and WT cells (P = 0.97). Error bars represent standard error of the mean.

If M2 antagonism reduces KO constriction by reversing M2-mediated depolarization, then hyperpolarization with 1.0 mm K⁺ PSS may be expected to occlude these effects. In the absence of cholinergic activation, 1.0 mm K⁺ PSS evoked no measurable changes in resting tension (data not shown). However, 1.0 mm K⁺ PSS substantially reduced cholinergic-evoked contractile force (compare Fig. 5B, 1 mm K⁺ PSS versusFig. 5A normal PSS, summarized in Fig. 5C). In addition, the increased contraction in KO (as seen in Fig. 5A), presumably due to a more depolarized trachea, is eliminated in the low K⁺ PSS (Fig. 5B). Rather, we see a small but significant increase of contraction in WT relative to KO. Finally, the voltage-dependent Ca²⁺ channel component of contraction (nifedipine sensitive-component) is reduced substantially and becomes similar in WT and KO (Fig. 5D). These results indicate that 1.0 mm K⁺ PSS effectively hyperpolarizes membrane potential during cholinergic-evoked contractions.

Utilizing 1.0 mm K⁺ PSS, we investigated the relaxant effect of M2 antagonism under hyperpolarizing conditions. Consistent with a decreased recruitment of voltage-dependent calcium channels in 1.0 mm K⁺ PSS, relaxation in the KO by M2 antagonism is reduced to WT levels (Fig. 5B and D). Also, the increased nifedipine-sensitive component of M2 antagonist relaxation seen in KO at normal PSS (9.3 ± 4 in WT and 43.6 ± 6 in KO, P < 0.001) is substantially reduced to near WT levels in 1.0 mm K⁺ PSS (Fig. 5E, 9.3 ± 1 in WT and 13.6 ± 3 in KO. P = 0.1). In summary, these results are consistent with membrane repolarization as a mechanism to relax tracheal muscle following antagonism of M2 receptor signalling.

We directly examined M2 antagonism effects on membrane potentials during cholinergic activation. Figure 7 shows a typical voltage measurement using sharp electrode recordings of isolated tracheal smooth muscle cells. Recordings indicate that TSMC membrane potentials widely oscillate around the means (Fig. 7A, dashed lines). As shown previously, KO TSMCs are more depolarized at resting conditions (Fig. 7A, time A, also Fig. 6). Carbachol (0.5 μm) evoked a transient hyperpolarization that was followed by a depolarization (Fig. 7A, time B). Although KOs were more depolarized than WT TSMCs, the relative carbachol induced depolarization in both KO and WT TSMCs was very similar (Fig. 7A, approximately 6 mV, summary data in Fig. 7C). The effect of M2 antagonist was to transiently repolarize membrane potential to pre-carbachol levels (Fig. 7A time C, and Fig. 7C). Thus, although effects of cholinergic activation and M2 antagonism are similar in WT and KO, the KO trachea voltages are significantly different from WT (Fig. 7B, P = 0.007 in comparison of WT and KO groups, two-way ANOVA) in that they operate from a more depolarized baseline voltage.

Figure 7. M2 receptor antagonism transiently opposes cholinergic evoked depolarizations.

A, β1 KO and WT membrane potentials were measured using sharp electrode recording of acutely isolated tracheal smooth muscle cells following superfusion with 0.5 μm carbachol, and carbachol with 100 nm AF-DX 116. Lines A, B and C indicate regions where voltages were summarized in panels B and C. B, average membrane potentials measured in normal PSS (ctrl, A), following addition of carbachol (carbachol, B), and carbachol with AF-DX 116 (carbachol+AF-DX, C). KO and WT (KO A, B, C versus WT A, B, C) groups show statistically significant difference using a two-way ANOVA (P = 0.007). Individual treatments within groups are not statistically different using one-way ANOVA. C, relative change in membrane potential due to carbachol administration (A-B), carbachol vs. carbachol + AF-DX (B-C), or pre-carbachol vs. carbachol + AF-DX (A-C). Relative changes were statistically analysed using one-way ANOVA with the following post hoc Tukey's HSD test. A-B (KOs: 5.5 ± 1, WTs: 5.7 ± 1.5) and B-C (KOs: −6.1 ± 1.3, WTs: –4.4 ± 0.6) have opposite polarities but values themselves are not statistically different from each other for both KO and WT. This implies that depolarization due to carbachol administration (A-B) is effectively abolished by application of AF-DX (B-C). On the other hand, both A-B and B-C significantly differ from A-C (KO: –0.6 ± 1, WT: 1.2 ± 1.2), which has a value close to zero. This implies that AF-DX abrogates membrane depolarization evoked by carbachol. Error bars represent standard error of the mean, **P≤ 0.01, *P = 0.02; n = 7 (KOs), n = 8 (WTs).

Discussion

A major finding of this study is that BK potassium channel and β1 subunit control of excitation–contraction coupling is contingent on M2 receptor activation. Knockout of the β1 subunit has little consequence when M2 receptors are inhibited. The prominent role of M2 receptors in excitation–contraction coupling has also been observed in ileum smooth muscle studies where contraction in M2 receptor knockout mice is only weakly sensitive to block of L-type calcium channels. However, animals that are WT for M2 receptors and lacking M3 are highly sensitive to L-type calcium channel block (Unno et al. 2005). Nevertheless, our results indicate that M2 receptor activation only partly accounts for recruitment of voltage-dependent calcium channels. The L-type calcium channel component of contraction is approximately two-fold larger than the M2 antagonist-sensitive component of contraction in both WT and KO (Fig. 5D). One explanation may be that the AF-DX concentrations used (100 nm used, K_i 64 nm) to avoid non-specific inhibition of M3 receptors (K_i 786 nm) may not completely inhibit M2 receptors (Hammer et al. 1986). This may also explain why we did not completely eliminate contractions in tracheas treated with both M3 antagonist and M2 antagonists (Fig. 1D). Alternatively, it may be that M2 receptor depolarization and BK channel-mediated repolarization occur over only part of the voltage ranges that recruit voltage-dependent calcium channels.

Comparison of β1 knockout and WT membrane potentials indicates that BK channels do not particularly oppose the depolarization evoked by cholinergic agonist. Rather, BK current creates a lower resting membrane potential from which cholinergic evoked depolarization occurs. Interestingly, the voltage window reported for calcium influx of airway smooth muscle (Fleischmann et al. 1994) correlates well with voltage differences between β1 KO and WT mice, and resulting effects on contraction. Voltage-dependent calcium influx has a threshold at approximately −40 mV with a steep voltage dependence that peaks at −30 mV (Fleischmann et al. 1994). We measured a resting membrane potential of −33 ± 2 mV in the β1 knockout that depolarizes following cholinergic activation by ∼6 mV to the peak voltage for calcium influx in the cells (−27 mV). In contrast, in the WT cells resting membrane potentials are more negative (−44 mV). Although cholinergic agonist depolarizes WT cells by similar voltages to KOs, the voltage nevertheless only rises to the foot of the voltage window where calcium influx is moderate. This steep voltage dependence for calcium influx combined with the depolarized resting membrane potential probably explains the much larger contribution of voltage-dependent calcium channels to contraction of KO than WT tracheas (Fig. 5). As well, this may explain why ∼6 mV hyperpolarization by M2 receptor antagonist (Fig. 7) is sufficient to largely reduce differences in contraction between KO and WT mice.

Several studies have characterized M2 receptor signalling pathways that inhibit BK channels either directly through G_i protein, or indirectly via reduction of cAMP levels (Kume et al. 1992; Wang & Kotlikoff, 1996; Wang et al. 1997b; Zhou et al. 2008). One may therefore hypothesize that contraction induced by M2 receptor activation can be partly mediated by inhibition of BK channels, depolarization, and activation of voltage-dependent calcium channels. Relaxation with M2 antagonist would thus occur by relief of BK channel inhibition. Although we indeed see that M2 antagonism relaxes trachea through a voltage-dependent mechanism resulting in reduced contribution of voltage-dependent calcium channels, we however observed that M2 antagonism decreases contraction in the presence of the BK channel blocker paxilline (Fig. 3). As well, the hyperpolarization due to M2 antagonism was of equal magnitude in both WT and β1 KO trachea (Fig. 7). This suggests that M2 antagonism does not relax airway via relief of BK channel inhibition, but rather through hyperpolarization that occurs independent of BK channels. The mechanisms may be through deactivation of the G_i-protein-dependent non-specific cationic current I_cat that is present in many types of smooth muscles including TSM (Benham et al. 1985; Inoue et al. 1987; Inoue & Isenberg, 1990; Pacaud & Bolton, 1991; Janssen & Sims, 1992; Sims, 1992; Lee et al. 1993; Fleischmann et al. 1997). I_cat shows sustained cholinergic activation and is sensitive to M2 inhibition (Fleischmann et al. 1997; Wang et al. 1997a). It has been shown that I_cat effectively depolarizes plasma membrane and therefore can trigger Ca²⁺ influx via voltage-dependent Ca²⁺ channels (Benham et al. 1985; Inoue et al. 1987; Inoue & Isenberg, 1990). It also has been shown that inhibition of M2 receptors in equine TSMCs by pertussis toxin not only prevents recruitment of I_cat but also completely abolishes the sustained Ca²⁺ influx (Wang et al. 1997a).

Utilization of the β1 subunit knockout in these and other studies provided an opportunity to investigate BK channel function in smooth muscle where β1 is enriched and otherwise promotes channel activation. However, the possibility exists that compensatory mechanisms may belie the severity of the β1 knockout phenotype. Such precedence exists with knockout of the pore-forming α subunit of BK channels. It has been reported that knockout of the BK channel in airway contributes to enhanced cGMP signalling to reduce rather than enhance cholinergic airway constriction (Sausbier et al. 2007). A partial compensatory mechanism is also observed in bladder smooth muscle, but this is eliminated with a smooth muscle and inducible knockout of the channel gene (Sprossmann et al. 2009). This is in contrast to β1 knockout mice where we have not observed dramatic evidence of compensation. For example, β1 knockout enhances contraction to a similar extent as BK channel block of wild-type trachea (Semenov et al. 2006). In this study, we also observed that the depolarization in β1 knockout smooth muscle was indistinguishable from depolarization of wild-type with BK channel blocker. Thus, there is a discrepancy between the observation of compensatory effects in α knockout but not β1 knockout mice. One may speculate that the β1 subunit may have a less essential role than BK α subunits in development of airway smooth muscle and therefore result in smaller compensatory effects in adult airway.

Adult airway smooth muscle is generally regarded as contracting tonically, and therefore it is quite intriguing that another function of BK channels is to prevent oscillatory contractions of airway muscle. We observed oscillatory contractions in β1 KO mice and upon application of BK channel blocker. Oscillatory contractions have also been described in the BK channel α subunit knockout mice and upon inhibition of BK channels by specific antagonists (Yagi et al. 2002, 2003; Sausbier et al. 2007). Oscillatory contractions only occur with sub-micromolar carbachol doses and depend on voltage-sensitive Ca²⁺ influx since L-type calcium channel blockers prevent oscillations (Sausbier et al. 2007). This is consistent with our observation that oscillatory contractions occur at low carbachol concentrations (Fig. 2A) where L-type Ca²⁺ channels have a greater relative contribution to cholinergic evoked contractions (Fig. 3). The mechanism by which BK channels stabilize contraction of otherwise tonic ASM requires further study.

The broad aim of this study was to understand how airway contraction that is initiated by metabotropic cholinergic signalling affects recruitment of voltage-dependent Ca²⁺ channels. The role of voltage-dependent calcium channels in cholinergic evoked contractions is somewhat controversial given that L-type calcium channels are high voltage-activated and airway is not regarded as depolarizing to a sufficient extent to activate these channels (Janssen, 2002; Roux et al. 2006). As well, although L-type calcium channels maintain contractions, high potassium depolarization is not sufficient to initiate contractions of bronchioles (Perez & Sanderson, 2005). Rather, voltage-dependent Ca²⁺ influx may serve a secondary role to contraction by refilling Ca²⁺ stores rather than directly affecting microfilament Ca²⁺. Nevertheless, our experiments measured a relatively large component (∼53%) of contraction that is sensitive to L-type calcium channel blocker and also a similar component of contraction that was sensitive to hyperpolarization by low external potassium (Fig. 5). Interestingly, this large contribution by voltage-dependent mechanisms at moderate cholinergic agonist (0.5 μm, roughly 50% of maximal dosage) becomes a smaller component of total contraction at higher cholinergic agonist (Fig. 4, and Semenov et al. 2006). Thus, the concentration of cholinergic agonist used may impact how investigators view the contribution of L-type calcium channels to airway contraction.

Recently, investigators have identified a voltage-dependent contractile mechanism in ASM that occurs independent of L-type calcium channels to increase cytosolic calcium (Liu et al. 2009). The mechanism is a direct voltage activation of unliganded M3 muscarinic acetylcholine receptors that increase calcium release through IP₃ receptors, and secondarily to ryanodine receptors to contract muscle. Voltage-dependent activation of M3 muscarinic acetylcholine receptors would predict that contractions would relax to a greater level with hyperpolarization (due to combined deactivation of L-type calcium channels and voltage-sensitive M3 receptors) compared to block with L-type Ca²⁺ channel blocker nifedipine alone. However our experiments did not reveal this phenomenon. Contractions following nifedipine block were similar to contractions observed in hyperpolarizing conditions (Fig. 5, normalized contraction amplitudes of 74 versus 83, respectively). This suggests that voltage activation by M3 receptor may not be physiologically relevant when the receptors are already liganded by cholinergic agonist.

In summary, the focus of these experiments was to understand the role of M2 muscarinic acetylcholine receptors and BK channels during excitation–contraction coupling in ASM. A summary diagram that displays M2 signalling and BK channels is presented in Fig. 8. A major finding is that BK channels functionally oppose the consequence of M2 receptor signalling that leads to depolarization and the recruitment of voltage-dependent calcium channels (Fig. 8A). Knockout of the β1 subunit results in greater depolarization evoked by the added effects of M2 signalling and reduced BK potassium current resulting in enhanced excitation–contraction coupling (Fig. 8B). Inhibition of M2 receptors by AF-DX 116 eliminates M2 mediated depolarization, and coincidently reduces the role of BK channels and β1 subunits in the control of ASM contraction (Fig. 8C).

Figure 8. Diagrammatic representation of signalling in airway smooth muscle cells.

A, pharmaco-contraction and excitation–contraction coupling in WT cells. Acetylcholine binding to muscarinic receptors evokes calcium from SR via M3 receptors and also inhibits cAMP-dependent relaxation mechanisms through M2 receptors (pharmaco-contraction coupling). M2 receptors also activate depolarization through G_i sensitive non-selective cation currents (I_cat). BK potassium channels and β1 subunits maintain voltages below threshold for activation of voltage-dependent calcium channels despite depolarization via M2 receptors/I_cat. B, knockout of the BK channel β1 subunit depolarizes membrane voltages. This, in combination with M2/I_cat activation, recruits voltage-dependent calcium channels and enhances airway constriction. C, in the absence of M2 receptor/I_cat signalling (AF-DX 116 antagonist), membrane voltage is below threshold for voltage-dependent calcium channel activation. β1 knockout, or BK channel block do not affect cholinergic-evoked contractions, which are mediated principally via M3-mediated events under these conditions.

Glossary

Abbreviations

ASM

airway smooth muscle

TSM

tracheal smooth muscle

TSMC

tracheal smooth muscle cell

VDCC

voltage-dependent calcium channel

Author contributions

I.S., R.B. and J.T.H. contributed to the conception of the experiments. All authors contributed to the design of the study. The contraction experiments were performed by I.S., and voltage measurements were performed by I.S. and B.W. Statistical analysis was performed by I.S. and J.T.H. The manuscript was prepared and edited by all authors, who approved the submitted version of the manuscript. The experiments were conducted at the Department of Physiology, University of Texas Health Science Center at San Antonio.