Voltage sensitivity of M2 muscarinic receptors underlies the delayed rectifier-like activation of ACh-gated K+ current by choline in feline atrial myocytes


E. G. Moreno-Galindo: C.U.I.B., Universidad de Colima, Av. 25 de Julio 965, Colonia Villa San Sebastián, C.P. 28045 Colima, Col., México. Email: eloy@ucol.mx

Key points

  • • Choline (Ch) is a precursor and metabolite of the neurotransmitter acetylcholine (ACh).
  • • Previously, in cardiomyocytes Ch was shown to activate an outward K+ current in a delayed rectifier fashion, which has been suggested to modulate cardiac electrical activity and to play a role in atrial fibrillation pathophysiology. However, the identity of this current remains elusive.
  • • Single-channel recordings, biophysical profiles and specific pharmacological inhibition indicate that the current activated by Ch is the ACh-activated K+ current (IKACh).
  • • Membrane depolarization increased the potency and efficacy of IKACh activation by Ch and thus gives the appearance of a delayed rectifier activating K+ current at depolarized potentials.
  • • Our findings support the emerging concept that IKACh modulation is both voltage- and ligand-specific and reinforce the importance of these properties in understanding cardiac physiology.

Abstract  Choline (Ch) is a precursor and metabolite of the neurotransmitter acetylcholine (ACh). In canine and guinea pig atrial myocytes, Ch was shown to activate an outward K+ current in a delayed rectifier fashion. This current has been suggested to modulate cardiac electrical activity and to play a role in atrial fibrillation pathophysiology. However, the exact nature and identity of this current has not been convincingly established. We recently described the unique ligand- and voltage-dependent properties of muscarinic activation of ACh-activated K+ current (IKACh) and showed that, in contrast to ACh, pilocarpine induces a current with delayed rectifier-like properties with membrane depolarization. Here, we tested the hypothesis that Ch activates IKACh in feline atrial myocytes in a voltage-dependent manner similar to pilocarpine. Single-channel recordings, biophysical profiles, specific pharmacological inhibition and computational data indicate that the current activated by Ch is IKACh. Moreover, we show that membrane depolarization increases the potency and efficacy of IKACh activation by Ch and thus gives the appearance of a delayed rectifier activating K+ current at depolarized potentials. Our findings support the emerging concept that IKACh modulation is both voltage- and ligand-specific and reinforce the importance of these properties in understanding cardiac physiology.





V h

holding potential


acetylcholine-activated K+ current


muscarinic type-2 receptor


pertussis toxin


Activation of the acetylcholine (ACh)-gated inwardly rectifying K+ current (IKACh) or KACh channels is an essential component of the physiological control of cardiac function by the parasympathetic nervous system (Löffelholz & Papano, 1985). In cardiomyocytes, KACh channels are coupled to muscarinic type-2 receptors (M2Rs) via Gi/o proteins (Pfaffinger et al. 1985; Breitwieser & Szabo, 1985). When ACh binds to M2R, the βγ subunits of the Gi/o proteins directly activate the KACh channels to increase their open probability, thus decreasing cardiac excitability (Hibino et al. 2010).

Recently, M2R and other G-protein-coupled receptors were noted to display an intrinsic voltage dependence (Ben-Chaim et al. 2003; Martinez-Pinna et al. 2005; Ohana et al. 2006). For example, the affinity of M2R for ACh increases at hyperpolarized membrane potentials, while decreasing at depolarized potentials (Ben-Chaim et al. 2003). The intrinsic ability of M2R to ‘sense’ membrane potential was confirmed by recording gating charge displacement that represents conformational changes in the receptor induced by changes in membrane potential (Ben-Chaim et al. 2006; Navarro-Polanco et al. 2011). In addition, the voltage sensitivity of M2R was noted to be ligand dependent. As opposed to ACh, the affinity for pilocarpine increased with membrane depolarization (Navarro-Polanco et al. 2011). Membrane depolarization probably induces a conformational change in the M2R orthosteric binding site that facilitates the binding of some ligands (e.g. pilocarpine), while hindering the binding of other ligands (e.g. ACh). Based on the opposite voltage-dependent response of M2Rs for pilocarpine, this agonist activates IKACh at depolarized potentials and thus gives the appearance of a delayed rectifying outward K+ current in feline atrial myocytes (Moreno-Galindo et al. 2011) and in rabbit sinoatrial node (Rodríguez-Martínez et al. 2011).

Choline (Ch) is a precursor and ACh metabolite that is an important component of phospholipids, plasma lipoproteins and cell membranes (Danne & Möckel, 2010). In canine atrial myocytes, Ch activates a background K+ current comparable to IKACh, as well as a delayed rectifier-like outward K+ current (Fermini & Nattel, 1994). A similar delayed rectifier K+ current was described in guinea-pig atrial myocytes activated by Ch (Shi et al. 1999) and the muscarinic agonist pilocarpine (Wang et al. 1999). The delayed rectifier-like current was postulated to represent a novel current, named IKM3, distinct from IKACh and other known voltage-dependent cardiac K+ currents. These authors proposed that Ch was a highly selective agonist of muscarinic type-3 receptors, coupled to Gq/11 proteins (Shi et al. 1998, 1999). Gq/11 proteins activate the phospholipase C–diacylglycerol–inositol phosphate system (Caulfield & Birdsall, 1998). IKM3 was suggested to modulate cardiac electrical activity (Shi et al. 1999; Wang et al. 1999) to produce cytoprotective effects against myocardial injuries (Yang et al. 2005) and to play a role in the pathophysiology of atrial fibrillation (Yeh et al. 2007). However, despite the promising cardiac pathophysiological relevance of the IKM3, the specific K+ channel subunits that carry such current remain unknown. Based on our recent description of the ligand-specific voltage dependence of M2R activation of IKACh, we hypothesized that Ch activates IKACh in a time- and voltage-dependent manner similar to pilocarpine.


Ethical approval

All animal studies were performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85–23, revised 1996) and after securing approval by the University of Colima and University of Utah Institutional Animal Care and Use Committees. All animal studies conform to the principles of UK regulations, as described in Drummond (2009). The experiments presented in this paper involved the isolation of left atrial myocytes from adult cats. The experimental protocol, including the use of sodium pentobarbitone as the anaesthetic agent, was approved by the Institutional Animal Care and Use Committee of the University of Colima. Animals were killed by excision of the heart en bloc while under anaesthesia, which was induced with sodium pentobarbitone (35–45 mg kg−1, i.p.) 30 min after having received heparin (1000 U kg−1, i.p.). The level of anaesthesia was monitored by confirming the absence of pedal withdrawal reflexes.

Cell isolation and preparation

Isolated left atrial myocytes from adult cats of either sex (>2 kg) were prepared using the enzymatic perfusion method of Isenberg & Klöckner (1982) as modified inMoreno-Galindo et al. (2011). After enzymatic isolation, cells were maintained in Kraft-Brühe (KB) solution at 4°C for later electrophysiological experiments.


The KB solution had the following composition (in mm): 80 potassium glutamate, 40 KCl, 20 taurine, 10 KH2PO4, 5 MgSO4, 10 glucose, 10 Hepes, 0.5 creatine, 10 succinic acid and 0.2 EGTA; pH was adjusted to 7.4 with KOH. The solution was bubbled with 100% O2. Patch pipettes were filled with (in mm): 80 potassium aspartate, 10 KH2PO4, 1 MgSO4, 20 KCl, 5 Hepes, 5 K4-BAPTA, 0.2 GTP-Na and 3 ATP-Na2 (pH 7.25 with KOH). The standard bath solution contained (in mm): 136 NaCl, 4 KCl, 1 MgCl2, 10 Hepes, 0.5 CaCl2, 2 CoCl2 and 11 glucose (pH 7.35 with NaOH). For assessment of Ch concentrations higher than 3 mm, isotonic conditions were maintained with replacement of NaCl by choline chloride. For single-channel currents in the cell-attached configuration, the pipette solution used was (in mm): 140 KCl, 1.8 CaCl2 and 5 Hepes (pH 7.4 with KOH). Ch was added directly to the pipette filling solution.

IKACh recording in isolated left atrial myocytes

Macroscopic currents were recorded in the whole-cell configuration of the patch-clamp technique by using an Axopatch-200B amplifier (Molecular Devices, Sunnyvale, CA, USA). Experiments were performed at a temperature of 36 ± 0.5°C. Data acquisition and command potentials were controlled by pCLAMP 10.0 software (Molecular Devices). Patch pipettes with a resistance of 1.5–2.5 MΩ were made from borosilicate capillary glass (WPI, Sarasota, FL, USA). The capacitance and series resistance with the cell membrane were compensated to provide the fastest possible capacitive transient, without ringing (oscillations). Currents were filtered with a four-pole Bessel filter at 1 kHz and digitized at 5 kHz. An agar–KCl bridge was used to earth the bath. Sequential concentration–response curves were determined as previously described (Navarro-Polanco et al. 2011).

Recordings were carried out in the presence of 3 μm E-4031 and 50 μm chromanol 293B to block rapid delayed rectifier current (IKr) and slow delayed rectifier current (IKs), respectively. External cobalt (2 mm) and internal BAPTA (5 mm) were included to block L-type calcium and calcium-activated currents, respectively. The Ch- and ACh-induced currents traces were obtained by digitally subtracting currents recorded under control conditions from those recorded in the presence of agonist.

For single-channel current measurements, the resistances of the electrodes were 8–12 MΩ. Signals were filtered at 2 kHz and sampled at 10 kHz. Once the cell-attached configuration was obtained, the standard bath solution was changed to KB solution to maintain the membrane potential near 0 mV. In most single-channel experiments, channel activity was elicited by depolarizing steps, but measured after subsequent repolarizations. Our standard protocol consisted of 30 pulses that defined a 1 min time window (for the repolarization step; 2 s). The channel activity in the patch was expressed as the product of channel number (N) and channel open probability (Po). During repolarization, NPo was calculated by dividing the number of samples where channel activity was above the 50% threshold by the total number of samples. All single-channel recordings were performed at room temperature (21–23°C).


Ch and ACh were dissolved directly in the external solution at the desired concentrations and prepared fresh daily. E-4031, chromanol 293B and tertiapin Q were obtained from Tocris Bioscience (Ellisville, MO, USA) and all other reagents were from Sigma-Aldrich (St Louis, MO, USA). Drugs were prepared as stock solutions and kept at −20°C until use: E-4031 (3 mm in water), chromanol 293B (50 mm in DMSO), tertiapin Q (300 mm in water). Pertussis toxin (PTX, 500 μg ml−1 in water) was kept at 4°C. Aliquots of the stock solutions were diluted to final desired concentrations in the standard bath solution. In experiments using PTX, half of the left atrial cells were stored in KB solution at 36°C, and the other half in KB solution plus PTX at a final concentration of 5 μg ml−1 for 2–6 h. Before recording, PTX-treated myocytes were washed for 20–30 min in the recording chamber. Additionally, PTX (5 μg ml−1) was added to the internal solution.

Modelling of IKACh

A Markov model of IKACh was developed to reconstruct experimental data in the presence of 1 mm Ch. The model is based on a more complete model of receptor systems (Weiss et al. 1996) and a 4-state model recently used to define ACh and pilocarpine modulation of IKACh (Moreno-Galindo et al. 2011). A diagram of the 4-state model is shown in Fig. 10A. Rate coefficients were either dependent on transmembrane voltage Vm or concentration of the ligand [L]. Voltage-dependent forward rates α and backward rates β were defined as:

display math

with the rates α0 and β0 at 0 mV, the charges zα and zβ, temperature T, Faraday constant F, and gas constant R. Ligand concentration [L]-dependent forward rates γ and backward rates δ were defined as:

display math

with the parameter γL and rate constant δL. IKACh was defined based on Zhang et al. (2002):

display math

with the conductance GACh, the extracellular K+ concentration [K+]o, the reversal voltage EK, and the states B1 and B2. The value 10 is millimolar, and the value 140 is in millivolts. Parameters of the Ch model (Supplemental Table SI, available online only) were determined by computational methods using an iterative stochastic approach for numerical fitting of feature vectors extracted from experimental and model data as in (Abbruzzese et al. 2010).

Data analysis

Data are reported as mean ± SEM (n= number of cells). pCLAMP 10.0 software was used to perform non-linear least-squares kinetic analyses of time-dependent currents according to a single exponential equation and based on the simplex algorithm. Where appropriate, Student's t test (paired or unpaired) or ANOVA, followed by Bonferroni's or Dunnett's test, was used for evaluating statistical difference. A two-tailed probability value of less than 0.05 (P < 0.05) was considered statistically significant.


Time and voltage dependence of Ch-activated currents in feline atrial myocytes

In an initial series of experiments, we assessed the effects of Ch on isolated feline atrial myocytes in the whole-cell configuration by means of a standard voltage-clamp protocol: from a holding potential (Vh) of −40 mV, 2 s hyper- and depolarizing test pulses were applied to potentials ranging from −120 to +60 mV with 20 mV increments, followed by return to the Vh of −40 mV, at an interpulse interval of 10 s (Fig. 1). With depolarizing pulses, Ch (1–10 mm) induced outward currents in a time- and voltage-dependent fashion. When returning to the Vh, outward tail currents slowly decayed to reach the current level of the Vh (Fig. 1AC). In addition, 3 and 10 mm Ch increased the holding current at −40 mV and induced inward currents with hyperpolarizing pulses that decayed to baseline upon return to Vh (Fig. 1B and C). As illustrated in Fig. 1D, the currents induced by Ch are concentration dependent and show an outward rectification behaviour. Ch-activated currents in feline cardiomyocytes are carried by K+ channels and require muscarinic receptors, since the reversal potential of the currents is dependent on the external K+ concentration and atropine (100 nm) prevented such activation (data not shown), similar to previous reports in dog and guinea pig atrium (Fermini & Nattel, 1994; Shi et al. 1999).

Figure 1.

Currents activated by Ch in feline atrial myocytes 
Traces showing the currents activated by 1 mm (A), 3 mm (B) and 10 mm Ch (C) applying the voltage-pulse protocol shown in the inset. For this and subsequent figures the dashed line indicates the zero current level. D, current–voltage (I–V) relationships for the currents activated by three Ch concentrations. The amplitude of the currents was measured at the end of the test pulse and expressed in current density (pA/pF). n= 5 for each concentration.

Concentration–response relationship for Ch activation of IKACh is voltage dependent

The time- and voltage-dependent properties of Ch-activated currents are reminiscent of those evoked by the muscarinic agonist pilocarpine (Moreno-Galindo et al. 2011) and thus we wondered if Ch might activate IKACh in a similar voltage-dependent manner. Accordingly, we carried out concentration–response curves to determine the IKACh activation by Ch at two holding potentials, −100 and +50 mV. Current elicited by increasing Ch concentrations was normalized to that obtained with a saturating concentration of ACh (10 μm). Consistent with the results of pilocarpine in Moreno-Galindo et al. (2011), we found that the IKACh activation by Ch was indeed dependent on the membrane potential (Fig. 2A and B). For instance, recordings in Fig. 2A illustrate that the response of 1 mm Ch is barely detectable at −100 mV, whilst at +50 mV it is one-third of the maximal response. In summary, the EC50 for IKACh activation by Ch at Vh−100 mV (20 ± 2 mm) was 4-fold larger than at +50 mV (5 ± 1 mm). Furthermore, the maximal asymptote of the response induced by Ch at −100 mV was 0.52 ± 0.04, whereas at +50 mV it nearly reached the response evoked by 10 μm ACh (0.98 ± 0.06). This observation suggests that depolarization increases the efficacy of Ch for activating IKACh (Fig. 2B), similar to that observed for pilocarpine (Navarro-Polanco et al. 2011).

Figure 2.

Voltage-dependent IKACh activation by Ch in concentration–response curves 
A, IKACh traces recorded from atrial myocytes elicited by increasing Ch concentrations at a Vh of −100 mV (bottom traces) and +50 mV (top traces). Ch was perfused for the duration indicated (horizontal bars at the top of the traces). B, concentration–response curves for IKACh activation by Ch at −100 mV (squares) and +50 mV (circles). Data were normalized to current evoked by a maximal ACh (10 μm) response recorded in that cell and plotted as a function of Ch concentration. The lines represent data fits to a Hill equation. The Emax, EC50 and Hill coefficient were 0.52 ± 0.04, 21 ± 3 mm and 1.1 ± 0.1, respectively (Vh−100 mV), while for Vh+50 mV values were 0.98 ± 0.03, 5 ± 1 mm and 0.8 ± 0.1, respectively. n= 8 cells, in both membrane voltages.

Ch-induced currents activate in a voltage-dependent manner, but deactivate independently of voltage

As mentioned earlier, the voltage-dependent currents activated by Ch were considered to be delayed rectifying K+ currents (Fermini & Nattel, 1994; Shi et al. 1999; Yeh et al. 2007). To clarify the kinetics of Ch-evoked currents, we analysed the kinetics of activation and deactivation induced by three Ch concentrations (1, 3 and 10 mm). Activation kinetics were assessed by a hyperpolarized pre-pulse (−120 mV) to close the channels, followed by depolarizing pulses to open the channels; in contrast, the deactivation process was evaluated by a depolarized pre-pulse (+50 mV) to activate the channels, followed by hyperpolarizing pulses to measure the rate of channel closure. Figure 3 illustrates typical current tracings showing current activation (Fig. 3A and B) and deactivation (Fig. 3D and E) by 1 and 3 mm Ch. Activation was slower at hyperpolarized potentials and accelerated with depolarization for the three Ch concentrations (Fig. 3C). Also, the activation time course was Ch concentration dependent, with faster kinetics at higher concentrations. On the other hand, deactivation kinetics were independent of membrane voltage for the three Ch concentrations (Fig. 3F). The voltage- and ligand-concentration-dependent activation with voltage-independent deactivation properties is similar to IKACh induced by ACh and pilocarpine (Moreno-Galindo et al. 2011), but not typical for a classic delayed rectifying K+ current.

Figure 3.

Activation and deactivation kinetics of the Ch-evoked currents 
Representative recordings of 1 mm (A) and 3 mm Ch-elicited currents (B) by means of 3 s depolarizing steps to −20, 0, +20, +40 and +60 mV from a Vh of −120 mV (inset at the bottom of the traces). The activation time courses were fitted to a single exponential standard equation to obtain the time constants (τact). C, relationships of voltage and τact produced by the test depolarizing pulses at three Ch concentrations (n= 5 for each concentration). Tail currents induced by 1 mm (D) and 3 mm Ch (E) applying hyperpolarizing pulses to membrane potentials between −120 and +20 mV in steps of 20 mV from a Vh of +50 mV (inset at the bottom of the traces). The time courses of tail current decays (deactivation) were fitted to a single exponential standard to obtain the time constants (τdeact). F, τdeact are plotted as a function of the hyperpolarizing membrane potentials at three Ch concentrations. n= 6 at each concentration.

The inwardly rectifying nature of Ch-evoked currents

Inward rectification is a hallmark of IKACh induced by ACh (as shown in Fig. 5CF). However, Ch-induced currents apparently did not show inward rectification, but rather gave the impression of a delayed rectifier current (Fig. 1). Thus, the next experiments were performed using voltage protocols designed to demonstrate the inwardly rectifying nature of Ch-induced currents. First, we derived a ‘fully-activated’ current–voltage (IV) relationship using a depolarizing pre-pulse to fully activate the current, followed by voltage steps to various test potentials to measure the peak recovered current. The peak current recovered during the test pulse provides a measure of available current at the test potential (similar to protocols used to study inward rectification in hERG channels; Spector et al. 1996). This ‘fully activated’I–V relationship clearly reveals the nature of inward rectification (Fig. 4A and B). Next, we used an alternative approach to evaluate inward rectification of Ch- and ACh-activated currents. From a Vh of +40 mV, a voltage ramp was applied during brief applications of Ch and ACh (Fig. 4C). As illustrated in Fig. 4D, Ch (5 mm) activated an inwardly rectifying K+ current that was similar in magnitude and rectification to current activated by ACh (300 nm).

Figure 4.

Inward rectifiying nature of IKACh activated by Ch and ACh 
A, currents evoked by 1 mm Ch using the voltage protocol shown in the inset. B, fully activated I–V curves of the 1 mm Ch-induced currents measured as a relationship between the instantaneous current (arrow) and the membrane voltage of the test pulse (n= 5). C, IKACh activation by ACh (300 nm) and Ch (5 mm) at a Vh of +40 mV. Rapid vertical deflections represent changes in membrane current induced by 50 ms voltage ramps from −140 to +50 mV. D, I–V curves from Ch- and ACh-induced currents obtained by substracting currents recorded in the absence of agonist from currents in the presence of agonist in response to linear voltage ramps from −140 to +50 mV.

Finally, we evaluated the effects of the duration of a hyperpolarizing voltage step on inward rectification and the development of apparent delayed rectifier current. From a Vh of −40 mV, 1 s depolarizing pulses to +50 mV were applied to activate channels, followed by a short 200 ms pulse to −120 mV. This pulse duration was brief enough to allow channel unblock from Mg2+/polyamines, but too short for channels to deactivate. Subsequently, depolarizing test pulses to potentials between −110 and +50 mV were applied to measure peak currents (Fig. 5A). Using this protocol in the same myocyte, Ch- and ACh-evoked currents displayed very similar characteristics: the magnitude of recovered current and inward rectification (Fig. 5A, C and E and inset). In both cases, current following the hyperpolarized interpulse was much larger than preceding the interpulse, as a result of unblock of Ch- and ACh-activated channels.

Figure 5.

Influence of deactivating pre-pulse duration on Ch-elicited currents determines the rectification behaviour of IKACh
Representative families of 3 mm Ch-activated currents (A and B) and 1 μm ACh-activated currents (C and D) obtained from the same myocyte using a three-pulse protocol (top of panels): from a Vh of −40 mV, the membrane potential was stepped 500 ms to +50 mV and then to −120 mV during 200 ms (A and C) or 4 s (B and D), before stepping to test potentials between −110 and +50 mV. Note that the decaying phase of Ch- and ACh-induced currents at +60 mV overlapped (inset). E, current density–voltage relationships of the current measured at the peak (arrow) from A and C. F, current density–voltage relationships of the current measured at the end of test pulses (arrow) from B and D. (n= 5.)

When the duration of the hyperpolarizing interpulse (−120 mV) was increased from 200 ms to 4 s, the ACh-induced current remained activated and exhibited inward rectification with subsequent depolarizations (Fig. 5D and F). By contrast, Ch-induced current deactivated during the 4 s hyperpolarizing pre-pulse and then slowly activated with subsequent depolarizations giving the appearance of a delayed rectifier current (Fig. 5B and F). Again, the current magnitude induced by these concentrations of ACh and Ch was similar, when measured in the same myocyte. Taken together, these results imply that the apparent delayed rectifier property is the result of current deactivation in the interpulse, followed by slow activation during the test pulses.

Ch and ACh induce slowly activating currents with opposite voltage dependence

To further characterize how the Ch-induced inward rectifying current (IKACh) transforms progressively into an apparent delayed rectifying current, we used the following voltage protocol. In the presence of 3 mm Ch, a 1 s depolarizing pulse to +50 mV was applied from −40 mV to activate the channels, then an interpulse of varying duration (250, 1250, 2250 and 4250 ms) to −120 mV was given to gradually deactivate the channels, followed by a 1 s depolarizing step to +50 mV, and then a return to a Vh of −40 mV (Fig. 6A). As the duration of the 2nd (deactivating) pulse to −120 mV was increased, the initial peak transient current elicited by the subsequent depolarizing step decreased (see arrows in Fig. 6A), and at the same time a delayed rectifying-like current became progressively more evident.

Figure 6.

Effect of deactivating pre-pulse duration on the Ch- and ACh-activated current 
Current recordings induced by 3 mm Ch (A) and 100 nm ACh (B) in the same myocyte applying opposite voltage commands: two activating pulses to +50 mV (Ch) or −120 mV (ACh) for 1 s, separated by a deactivating pulse to −120 mV (Ch) or +50 mV (ACh) with variable duration (see inset at the bottom of the current traces). Note that, in the presence of Ch or ACh, IKACh activation is more evident as long as the deactivating interpulse duration increases.

After washout of Ch, we used an opposite voltage protocol to study the effect of the 2nd voltage pulse duration on the degree of deactivation using 100 nm ACh (Fig. 6B). In this case, a 1 s activating voltage step to −120 mV was followed by a 2nd pulse of varying duration to +50 mV to allow channels to deactivate, after which the voltage was returned to −120 mV to assess the degree of current re-activation. As the duration of the 2nd voltage pulse (to +50 mV) was lengthened, the re-activation process of the current at −120 mV (IKACh relaxation) was increasingly manifested, but in the opposite direction to the Ch-induced current. Similar results were obtained in five other myocytes interchanging the order of agonist application. These experiments illustrate that both Ch and ACh evoked currents that activate slowly, but with opposite voltage dependence. Stated another way, both ACh and Ch exhibit the property of ‘IKACh relaxation’, but the opposite voltage dependence of Ch-induced IKACh relaxation gives the impression of a delayed rectifier current.

Ch-elicited currents are PTX- and tertiapin-sensitive, consistent with activation of IKACh

Pertussis toxin (PTX) and tertiapin Q are selective pharmacological tools used to study M2R and IKACh, respectively. PTX specifically blocks the interaction between M2R and Gi/o proteins (Pfaffinger et al. 1985; Breitwieser & Szabo, 1985), while tertiapin Q is a potent and selective blocker of IKACh in cardiac myocytes (Kitamura et al. 2000; Drici et al. 2000). To determine if Ch activates IKACh via M2R, we tested the sensitivity of Ch-activated currents to tertiapin Q and PTX. Preincubation with PTX prevented the activation of currents by Ch (Fig. 7A, B and E). Moreover, tertiapin Q (300 nm) completely abolished the currents previously induced by 1 mm Ch (Fig. 7C, D and E). When tertiapin Q was perfused before the application of Ch, it prevented the activation of Ch-activated currents (data not shown). These data support the notion that Ch activates IKACh.

Figure 7.

PTX and tertiapin Q inhibit the Ch-induced currents 
Macroscopic currents evoked by 1 mm Ch in myocytes pre-incubated without PTX (A) or with PTX (B). Ch-activated currents before (C) and during perfusion with 300 nm tertiapin Q (D). Currents were obtained applying the voltage-pulse protocol shown in the inset. Arrows point to the zero current level. E, summary of the Ch-induced currents inhibited by PTX or tertiapin Q. Current amplitude was measured at the end of the +60 mV test potential and expressed in current density (pA/pF). n= 5–6 for each group. ***P < 0.001, compared with its own control.

Ch-induced K+ single-channel current

To further characterize the current activated by Ch we obtained single-channel recordings in the cell-attached configuration, using equimolar external and internal K+ concentrations (see Methods). In these conditions, the reversal potential was close to 0 mV, which permitted us to study single-channel inward currents. When Ch was absent from the pipette filling solution, very low activity was recorded (data not shown). Conversely, when 3 mm Ch was added to the pipette solution, single-channel activity characteristic of the fast open–close kinetics of KACh channels was recorded, and showed strong inward rectification (Fig. 8A). The measured conductance was 39.1 pS for Vh negative to −40 mV (Fig. 8B) and the mean open time was 1.1 ms at −80 mV (Fig. 8C). The NPo (the product of the number of channels in a patch and the open probability) increased as a function of the transmembrane voltage between −100 and −40 mV (Fig. 8D). Finally, no Ch-induced channel activity was observed when atropine (100 nm) was also included in the pipette solution (data not shown). These data further support the notion that Ch-activated channels in feline cardiomyocytes represent KACh channels.

Figure 8.

Properties of single channels activated by Ch are similar to IKACh
A, recordings of single-channel activity obtained in a cell-attached patch at various membrane potentials induced by 3 mm Ch in the pipette filling solution. B, I–V relationship for single Ch-activated channels. The unitary slope conductance at negative voltages was 39.1 pS. C, histogram of the open time distribution at a membrane potential of −80 mV. The lifetimes of openings were distributed according to a single exponential function with time constant (τ) of 1.1 ms. D, NPo (the product of the number of channels in a patch and the open probability) is plotted as a function of the membrane potential. NPo was obtained from data segments of 30 s duration (n= 6 patches). *P < 0.05 and ***P < 0.001, compared with activity recorded at −100 mV.

Activity of Ch-induced single channels is increased by membrane depolarization

The strong inward rectification of Ch-induced channels complicates analysis of outward single-channel currents at membrane voltages positive to the reversal potential (0 mV, in this case). To clarify how Ch augments the activity of single-channel currents at positive potentials, we used two separate strategies in cell-attached patches to assay current activity at a hyperpolarized potential preceded by a depolarizing potential. First, channels were activated by 1.5 s pre-pulses to various potentials in the presence of 3 mm Ch, followed by a voltage step to −100 mV to measure the extent of channel activation by the pre-pulse (Fig. 9A). At −100 mV, the NPo was 0.041 ± 0.023 (n= 16). The NPo measured at the test pulse of −100 mV was plotted as a function of the pre-pulse voltage (Fig. 9C). The NPo value was higher as the activating pre-pulse became more positive, indicating an increase in channel activity with membrane depolarization. Next we measured single-channel activity as a function of time, by applying an activating pre-pulse to +50 mV of variable duration and measuring channel activity at −100 mV (Fig. 9B and D). The NPo value was enhanced as the activating pre-pulse duration became longer. These voltage protocols reveal that time and voltage increase the open probability of Ch-induced channels, but only when channels are assayed after removal of polyamine/Mg2+ blockade (i.e. at hyperpolarized potentials preceded by depolarization). Stated another way, the time- and voltage-dependent increased channel activity is masked by polyamine/Mg2+ blockade.

Figure 9.

The Ch-activated single-channel activity is increased by depolarization 
Typical examples of single-channel current recorded from a cell-attached patch at −100 mV after depolarizing steps to various membrane potentials with fixed duration of 1.5 s (A), and after a depolarization to +50 mV of increasing duration (0.05, 0.5, 1.0, 1.5 and 2.0 s) (B). NPo is plotted as a function of the membrane potential of depolarizing pre-pulses (C), and as a function of the depolarizing pre-pulse duration (D). For C and D, n= 6 patches. *P < 0.05 and ***P < 0.001, compared with activity recorded at −100 mV.

A Markov model of Ch-activated IKACh

To further emphasize how Ch gives rise to the apparent delayed rectifier current, a Markov model was developed to reconstruct the effects of Ch on IKACh, using a 4-state model previously designed to study ACh and pilocarpine (Moreno-Galindo et al. 2011). The model incorporates a published paradigm of IKACh that recapitulates inward rectification (Zhang et al. 2002). The model comprises four states coupled with rate coefficients dependent on Vm and ligand concentration (Fig. 10A). Featured parameters of the model are listed in Supplemental Table SI. Simulated IKACh elicited by step depolarizations in the presence of 1 mm Ch is presented in Fig. 10B using the pulse protocol from Fig. 1. The Ch model recapitulated the slowly activating currents elicited by membrane depolarization in the presence of Ch that resemble a ‘delayed rectifier’ K+ current. Likewise, the simulations recapitulate the experimental features of activation and deactivation of Ch-induced current (Fig. 10C and D). The computational features of Ch-induced IKACh are similar to those reported for pilocarpine (Moreno-Galindo et al. 2011), supporting the notion that both Ch and pilocarpine activate IKACh with a voltage dependence that is opposite to that of ACh. Note that the model does not alter the fundamental inward rectification parameters of IKACh, but rather incorporates ligand-specific voltage-dependent parameters. We previously showed that by changing the ligand-specific voltage-dependent parameter, the inwardly rectifying current activated by ACh could be converted into an apparent delayed rectifier current using pilocarpine (Moreno-Galindo et al. 2011), similar to our Ch-induced currents.

Figure 10.

Simulated IKACh elicited by Ch 
A, diagram of 4-state Markov model to reconstruct IKACh. Rate coefficients were dependent on membrane voltage or ligand concentration. Simulated IKACh elicited by 1 mm of Ch using the voltage-clamp protocols of Figs 1, and 3A and D (panels B, C and D, respectively). U1 represents the receptor state in the absence of ligand and at a hyperpolarized membrane potential; U2, the absence of ligand at a depolarized potential; B1, the presence of ligand at a hyperpolarized membrane potential; and B2, the presence of ligand at a depolarized membrane potential.


Here, we clarify the nature of Ch-evoked currents in feline atrial myocytes within the context of recent discoveries regarding the nature of voltage- and ligand-dependent modulation of IKACh. Based on experimental and computational findings, we conclude that the apparent delayed rectifier current elicited by Ch represents slow IKACh activation (the so-called ‘relaxation’ gating) with a voltage dependence opposite to ACh, but similar to pilocarpine. These observations further underscore the idea that voltage-dependent conformational changes in the muscarinic receptor modulate the affinity of the receptor for ligands in a ligand-specific manner. The consequence of this notion is that membrane depolarization may activate or deactivate IKACh at sub-saturating concentrations of ligands, giving the appearance of a delayed rectifier or inward rectifier current, respectively.

Ch was originally reported to activate two K+ currents in canine atrial myocytes: a background current similar to IKACh and a delayed rectifier-like outward current (Fermini & Nattel, 1994). The delayed rectifier component was later reported to require M3Rs and was postulated as a novel cardiac current, IKM3 (Shi et al. 1999). While IKM3 was implicated in several important pathophysiological cardiac conditions (see e.g. Shi et al. 1999; Wang et al. 1999, 2007; Yang et al. 2005; Yeh et al. 2007), the precise molecular basis of K+ channel subunits that conduct IKM3 remains elusive.

We recently described unique ligand- and voltage-dependent properties of M2R activation of IKACh and proposed a novel mechanism to explain the enigmatic IKACh relaxation process (Moreno-Galindo et al. 2011). IKACh relaxation refers to a slow decrease or increase in current magnitude with depolarization or hyperpolarization, respectively (Noma & Trautwein, 1978; Ishii et al. 2001). We recently demonstrated that IKACh relaxation represents a voltage-dependent change in agonist affinity as a consequence of a voltage-dependent conformational change in the muscarinic receptor. Therefore, in the current study we sought to determine if the Ch-evoked delayed rectifier-like current was a consequence of IKACh activation.

Ch-activated currents behaved similarly to IKACh in several regards. First, Ch activated channels with single-channel conductance, mean open time and inward rectification properties comparable to IKACh (Navarro-Polanco & Sánchez-Chapula, 1997). Second, the current induced by Ch was sensitive to PTX and to the selective blocker of IKACh, tertiapin Q (Kitamura et al. 2000; Drici et al. 2000). Third, when fully activated, Ch-evoked currents displayed inward rectification similar to ACh. Fourth, membrane depolarization increased the potency and efficacy of Ch to activate IKACh (while hyperpolarization decreased them) in a manner similar to that reported for pilocarpine. Finally, a Markov model of IKACh that incorporates ligand-specific voltage-dependent parameters recapitulates the features of Ch-evoked currents. Thus, we conclude that in feline atrial myocytes, Ch activates a single current, namely IKACh, in a voltage-dependent manner. The opposite voltage dependence of IKACh activation by Ch (relative to ACh) gives the appearance of a delayed rectifier activating K+ current at depolarized potentials. The fundamental mechanism underlying this phenomenon is that the affinity and efficacy of muscarinic agonists are differentially modulated by voltage.

The observation that G-protein-coupled receptors, including M2Rs, possess intrinsic voltage sensitivity (Ben-Chaim et al. 2003, 2006; Navarro-Polanco et al. 2011) has important consequences for understanding the physiology and pharmacology of excitable cells (Olcese, 2011). Adding to the complexity, the M2R voltage sensitive properties are not the same for all muscarinic agonists, but vary depending upon the specific ligand. The present study further underscores the concept of ligand-specific voltage modulation of M2Rs in that Ch behaves in an opposite way to ACh, but similar to pilocarpine. We previously hypothesized that voltage-induced conformational changes are transmitted to the agonist binding (orthosteric) site of the M2R, to facilitate or impede binding of a particular agonist, based on the distinct chemical structure of the ligand (Navarro-Polanco et al. 2011). The ligand-specific voltage-dependent modulation of IKACh is especially relevant to cardiac physiology and pathophysiology given the relatively long duration of the cardiac action potential. There is sufficient time between the diastolic (hyperpolarized) membrane potential and the plateau (depolarized) phase to allow for unique modulation of heart rate and/or action potential duration by endogenous or pharmacological ligands. The unique ligand- and voltage-dependent properties could be exploited for novel therapeutic agents and indications. Our findings support the emerging concept that IKACh modulation is both voltage- and ligand-specific and reinforce the importance of these properties in understanding parasympathetic and pharmacological modulation of cardiac physiology.


Additional information

Competing interests


Author contributions

All research studies were done at the Centro Universitario de Investigaciones Biomédicas. R.A.N.-P., J.A.S.-C. and E.G.M.-G. participated in the conception and design of the studies. R.A.N.-P., I.A.A-F., P.D.S.-F., D.E.B.-H., J.C.R.-E., F.B.S. and E.G.M.-G. contributed to the experiment performance. R.A.N.-P., I.A.A.-F., M.T.-F. and E.G.M.-G. performed data analysis and were involved in the interpretation of data. M.T.-F. and E.G.M.-G. had equal contributions in writing the paper with critical input from all the authors, who also approved the final version.


This work was supported by Consejo Nacional de Ciencia y Tecnología, México (054577 to R.A.N.-P.; 2008-01-105941 to J.A.S.-C.; 2011-01-167109 to E.G.M.-G.).


The authors wish to thank Miguel Angel Flores-Virgen for technical assistance.