ORIGINAL ARTICLE
Acetylcholine–Atropine Interactions: Paradoxical Effects on
Atrial Fibrillation Inducibility
Yu Liu, MD,* Benjamin J. Scherlag, MD,† Youqi Fan, MD,‡ Wenfang Xia, MD,*
He Huang, MD,* and Sunny S. Po, MD†
Abstract: Atropine (ATr) is well known as a cholinergic antagonist, however, at low concentrations ATr could paradoxically
accentuate the parasympathetic actions of acetylcholine (ACh). In
22 pentobarbital anesthetized dogs, via a left and right thoracotomy,
a leak-proof barrier was attached to isolate the atrial appendages
(AAs) from the rest of the atria. In group 1 (Ach+ATr+Ach), ACh,
100 mM, was placed on the AA followed by the application of ATr,
2 mg/mL. The average atrial fibrillation (AF) duration was 17 6
7 minutes. After ATr was applied to the AA and ACh again tested,
the AF duration was markedly attenuated (2 6 2 minutes, P , 0.05).
In group 2 (ATr+Ach), ATr was initially applied to the AA followed
by the application of ACh, 100 mM. There was no significant difference in AF duration (16 6 4 minutes vs. 18 6 2 minutes, P =
NS). The inhibitory effect of ATr on induced HR reduction (electrical stimulation of the anterior right ganglionated plexi and vagal
nerves) was similar between groups 1 and 2. These observations
suggest that when ATr is initially administered it attaches to the
allosteric site of the muscarinic ACh receptor (M2) leaving the orthosteric site free to be occupied by ACh. The M3 receptor that
controls HR slowing does not show the same allosteric properties.
Key Words: acetylcholine, atropine, atrial fibrillation, allosteric acetylcholine receptors
(J Cardiovasc Pharmacol Ô 2017;69:369–373)
INTRODUCTION
In 1950, Scherf et al1 found that the most effective
method for inducing sustained atrial fibrillation (AF) was
the application of high concentrations of acetylcholine
(ACh) onto the area of the sinus node or at the AV junction.
Received for publication December 23, 2016; accepted March 7, 2017.
From the *Department of Cardiology, Renmin Hospital of Wuhan University,
Wuhan, China; †Department of Medicine, Heart Rhythm Institute, The
University of Oklahoma Health Sciences Center, Oklahoma City, OK;
and ‡Department of Cardiology, Sir Run Run Shaw Hospital, Zhejiang
University School of Medicine, Hangzhou, China.
Supported by the Heart Rhythm Institute, The University of Oklahoma Health
Sciences Center to S. S. Po, the Helen and Wilton Webster Arrhythmia
Research Fund of the University of Oklahoma Foundation to B. J. Scherlag,
and the National Natural Science Foundation of China (81570459) to
Y. Liu.
The authors report no conflicts of interest.
Reprints: Yu Liu, MD, Department of Cardiology, Renmin Hospital of
Wuhan University, No. 238 Jiefang Rd, Wuhan 430060, China (e-mail:
liuyuwuda@163.com) or Sunny S. Po, MD, Heart Rhythm Institute,
1200 Everett Drive (6E103), Oklahoma City, OK 73104 (e-mail:
sunny-po@ouhsc.edu).
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
J Cardiovasc Pharmacol ä Volume 69, Number 6, June 2017
Cooling these areas terminated AF. Using this approach, we
have previously demonstrated that in the baseline state burst
pacing of the atria induced only short (9–40 seconds) of AF.
However, the application of 100 mM of ACh topically
applied to the left or right atrial appendage (AA) of the dog
heart induced AF lasting $10 minutes.2
In the present report, we initially established that ACh,
topically applied to the AA, was associated with enhanced
heart rate slowing caused by anterior right ganglionated
plexi (ARGP) or vagal nerve stimulation (VNS) and increase
in the duration of spontaneous or pacing induced AF
duration. After atropine (ATr) was applied to the AA
followed by a second application of ACh, the effects of
ACh were significantly reversed. However, when ATr was
applied first to the AA followed by the application of ACh in
the same concentrations as before, the effects of ACh,
particularly, on AF duration were similar to that shown by
ACh alone. We discuss the possible explanation for this
paradoxical finding.
METHODS
All animal studies were reviewed and approved by
the Institutional Animal Care and Use Committee of the
University of Oklahoma Health Sciences Center and the
Animal Studies Subcommittee of the Department of Veterans Affairs Medical Center. Twenty-two adult mongrel
dogs, weighing 20–25 kg, were anesthetized with Napentobarbital, 30 mg/kg, administered intravenously. Additional doses, 50–60 mg, were given hourly to maintain an
adequate level of surgical plane anesthesia. After endotracheal intubation, positive pressure ventilation was instituted
with a mixture of room air and 100% oxygen to obtain
oxygen saturation at least 90%. Core temperature was maintained at 36.5 6 1.58C by a sensor controlled heating pad
underneath the dog. The right and left vagosympathetic
trunks were dissected in the neck and 2 teflon coated (except
for the tips) wires were inserted into the nerve trunks for
electrical stimulation. A quadripolar electrode catheter was
introduced via the left femoral artery and positioned in the
aortic root to record His bundle electrograms. Another multielectrode catheter was inserted into a femoral vein and
passed into the right AA to record atrial electrograms. All
tracings were amplified and digitally recorded using a computerbased Bard Lab system (CR Bard Inc, Billerica, MA). Intracardiac bipolar electrograms were filtered at 30–250 Hz. ECG
leads were filtered at 0.1–250 Hz.
www.jcvp.org |
369
Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Liu et al
J Cardiovasc Pharmacol ä Volume 69, Number 6, June 2017
Procedures
Through a right thoracotomy at the fourth intercostal
space, and exposure of the heart via a pericardiotomy,
a polyethylene tube was attached to the right AA (RAA) in
a similar manner as on the left side (Fig. 1B). To assess the
function of electrical stimulation at the ARGP to slow the
heart rate, a bipolar plaque electrode was sutured to the fat
pad containing the ARGP neurons (Fig. 1B).
A gauze pad lightly moistened with ACh in various
concentrations (1, 10, and 100 mM) was applied to the LAA
or RAA (Fig. 1) and the effects on induced AF duration were
determined as a concentration/response relationship. Great
precaution was taken to avoid ACh leaking to other cardiac
structures to cause extraneous effects. The concentration of
ACh which induced an average AF duration $10 minutes
(100 mM, Fig. 2) was chosen to determine the effects of
ACh applied to the RAA or LAA on other electrophysiological parameters such as the maximal responses of heart rate
slowing induced by right VNS and ARGP stimulation which
slowed the heart rate before inducing sinus arrest, AV block
(VNS) or atrial excitation (GP stimulation).
The studies were divided into 2 groups: group 1, initial
application of ACh (100 mM) on the AA to determine the
spontaneous or induced duration of AF. ATr (2 mg/mL) was
then applied to the same AA area. After 5 minutes, another
application of ACh was made and the duration of induced AF
again determined (n = 8). Group 2: The same procedure was
followed except that a gauze pad lightly moistened with ATr
(2 mg/mL) was applied first followed by the application of
ACh and the duration of induced AF then determined (n = 8).
The effects of electrical stimulation at the ARGP and vagosympathetic trunks were performed in both group 1 and group
2 dogs.
A left thoracotomy was performed at the fourth
intercostal space to provide access to the left side of the
heart. The pericardium was incised and reflected to expose the
left atrium. A polyethylene tube was attached to the left AA
(LAA) as a barrier between the AA and the rest of the atrium.
To ensure that the barrier was leak-proof, a thin layer of tissue
glue was placed along the length of the barrier (Fig. 1A).
Statistical Analysis
All data are expressed as mean 6 SD. Statistical analysis was performed using a Student t test for the effects of
ACh and ATr on maximal heart rates during ARGP and VNS
and on the duration of induced AF. One-way analysis of
FIGURE 1. Diagrammatic representations of the left and right
side of the heart. Panel A, A polyethylene barrier was attached
across the left AA (LAA). Tissue glue applied on the LAA side of
the barrier provided a leak-proof area for the application of
a gauze pad moistened with various concentrations of ACh
onto the LAA. Other abbreviations: CS, coronary sinus; LA, left
atrium; ILGP, inferior left ganglionated plexi; LIPV, left inferior
pulmonary vein; LOM, ligament of marshall; LPA, left pulmonary artery; LSPV, left superior pulmonary vein; LV, left ventricle;
RV, right ventricle; SLGP, superior left ganglionated plexi. Panel
B, A similar barrier was attached across the right AA (RAA) to
provide a leak-proof area for the application of a gauze pad
moistened with various concentrations of ACh onto the RAA. A
plaque electrode was attached to the ARGP for electrical stimulation to slow the heart rate. Other abbreviations: IRGP, inferior right ganglionated plexi; IVC, inferior vena cava; RA, right
atrium; RIPV, right inferior pulmonary vein; RSPV, right superior
pulmonary vein; SVC, superior vena cava.
FIGURE 2. A comparison of the effect of various concentrations of ACh applied on the AAs on the duration of induced
atrial fibrillation. The bars represent the mean AF duration
induced by the application of ACh to the AAs, 1 mM ACh,
0.9 6 0.7 minutes; 10 mM, 5.6 6 3.1 minutes; 100 mM,
17.9 6 5.5 minutes.
370
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
| www.jcvp.org
Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol ä Volume 69, Number 6, June 2017
Acetylcholine/Atropine Paradox
FIGURE 3. The effect of ACh and ATr on heart
rate. A, shows that when ACh (100 mM) was
applied to the AA, electrical stimulation of the
ARGP induced an average 44% decrease of the
maximal heart rate (starting heart rate, 158 6 22–
88 6 32/min). After ATr was applied to the AA,
a second application of ACh (100 mM) showed
a significant attenuation (P , 0.05) of electrical
stimulation induced heart rate slowing over the
same voltage range. A similar response was shown
for VNS, (B).
variance followed by Duncan post hoc test was used for
multiple comparisons. P values ,0.05 were considered
significant.
RESULTS
The graph in Figure 2 depicts the concentration of ACh
applied to the AA plotted against the induced duration of AF.
There was a consistent increase in AF duration as the ACh
concentration applied to the AA increased. At ACh concentrations of 1, 10, and 100 mM, induced AF durations progressively increased from 0.9 6 0.7 to 5.6 6 3.1 to 17.9 6
5.5 minutes). On this basis we chose 100 mM as the concentration to determine the responses to the other electrophysiological parameters.
Group 1: Response to ACh, 100 mM, Applied
First to the AAs
We assessed the ability of electrical stimulation of the
ARGP to achieve maximal heart rate slowing before inducing
atrial excitation.3 Figure 3A shows that when ACh, 100 mM,
was applied to the AA a new baseline state was achieved by
electrical stimulation of the ARGP, which induced an average
44% decrease of the maximal heart rate (starting heart rate,
158 6 22–88 6 32/min). After ATr was applied to the AA,
a second application of ACh, 100 mM, was tested. Subsequently, there was a significant attenuation (P , 0.05) of
electrical stimulation to induce heart rate slowing over the
same voltage range either for ARGP or VNS (n = 4).
To assess both qualitative and quantitative aspects of
AF inducibility, ACh, 100 mM was applied to the AA. If
spontaneous AF did not occur within 2–5 minutes, electrical
or mechanical stimulation was used to induce sustained AF
($10 minutes). Figure 4 illustrates the average duration of AF
in response to ACh applied to the AA (17 6 7 minutes). After
a gauze pad moistened with ATr (2 mg/mL) was applied to
the AA and ACh again tested, the duration of AF was
markedly attenuated (2 6 2 minutes, P , 0.05, n = 8).
Group 2: Response to ACh Applied to the AAs
After the Initial Application of ATr
ATr, 2 mg/mL, applied first to the AA followed by
ACh, 100 mM, showed similar inhibitory responses in regard
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
to maximal heart rate slowing induced by either ARGP or
VNS (Fig. 5). On the other hand, there was no significant
difference in AF duration when ATr was applied first followed by ACh, 100 mM, application to the AA (Fig. 6, n = 8).
DISCUSSION
In his 1936 Nobel lecture in which Loewi described his
discovery that the substance released by vagal stimulation
was ACh he also stated that, “We were able to determine that
its effect is inhibited by atropine.”4 Subsequent studies revealed that ATr is a competitive antagonist of the muscarinic
ACh receptors (mAChRs). Early investigations5–7 suggested
that ATr, particularly in low concentrations, could paradoxically enhance cholinergic activation; whereas, based on
numerous experiments it was well documented that ATr
was an effective inhibitor of ACh activation. The paradoxical
action of ATr, as cholinergic inhibitor or activator, has
recently been explained by the emergence of the property
of certain molecules to manifest allosteric regulation. This
is defined as regulation of an enzyme or other protein by
FIGURE 4. The effect of ACh and ATr on AF duration. ACh
(100 mM) applied to the AA induced AF lasting 17 6 7 minutes which was markedly attenuated when ATr was applied
to the AA followed by a second application of ACh (2 6
2 minutes, P , 0.01).
www.jcvp.org |
371
Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Liu et al
J Cardiovasc Pharmacol ä Volume 69, Number 6, June 2017
FIGURE 5. In group 2 dogs, ATr was applied first
to the AA then followed by ACh, ARGP stimulation, panel (A) and VN stimulation showed the
same diminished heart rate response.
binding at the allosteric site (other than the protein’s active or
orthosteric site).
In a recent study of the allosteric properties of ATr,
May et al8 demonstrated that “prolonged exposure to allosteric modulators (viz., ATr) can cause up-regulation of cell
surface M2, mAChRs expression and cellular responsiveness.
These effects are independent of the well-documented actions
of (allosteric) modulator administration on othosteric (sites).”
They go on to state, “Allosteric modulators possess a number
of theoretical advantages over orthosteric drugs. For instance,
they can either inhibit or potentiate ligand binding affinity
and/or function.” depending on which site they occupy.8,9
In the present study, we found that in contrast to the
previously mentioned studies,5–7 relatively high concentrations of ACh, 100 mM, applied to the AA after the initial
application of ATr to the same area did not significantly
change the duration of induced AF (Fig. 6); whereas, when
ATr was applied after ACh had significantly increased AF
duration determined at baseline, there was a marked inhibition of AF duration in response to the next application of ACh
(Fig. 4). We hypothesize that ATr when administered initially
acted as an allosteric modulator and was bound to the allosteric site of M2 mAChRs leaving the orthosteric site
FIGURE 6. In group 2 dogs, AF duration was not significantly
changed compared with the AF duration in response when
ACh alone was applied to the AAs. See text for further
discussion.
available for the ACh to attach to this site and allowing
essentially no change in the duration of induced AF as seen
when ACh was applied to the AA before ATr. On the other
hand, when ACh was the initial application, ATr acted as
a competitive antagonist, which resulted in inhibition of the
ACh induced AF duration (Fig. 4).
Another unexpected finding was the effect on slowing
of the heart rate in response to the application of ACh onto the
AA. In many of our previous experiments we used ACh
application in various concentration to induce prolonged
durations of AF.2,10 We did not find any consistent change
in heart rate even when ACh at concentrations of 100 mM
was applied to the AA. However, in those studies we never
compared the response to slowing of the heart rate with
ARGP stimulation and VNS before and after ACh or ATr
was applied to the AA. In the present study, we found 2
different responses in regard to AF duration, ie, inhibition
versus activation, in group I compared with group 2, respectively. Yet, in regard to heart rate slowing by ARGP stimulation or VNS, ATr showed only inhibition in both Groups.
How can these findings be reconciled?
The longstanding axiom that M2 receptors were the
only functional mAChRs in the heart11 has been challenged
by the finding that there are multiple subtypes found in the
dog heart, specifically M2, M3, and M4.12 We hypothesize that
the M2 ACh receptors play an important role in provoking AF
inducibility and interacting with ATr based on its allosteric
properties. In contrast, activation of M3 ACh receptors, which
are responsible for heart rate slowing13 may not participate in
ATr allostery. Wang et recently reported functional and
molecular evidence for the presence of M3 and M4 receptors
in canine and M3 receptors in guinea pig atrial myocytes.
They found these receptors to be functionally coupled to 2
novel and distinct K+channels. Activation of M3 receptors in
guinea pig atrial preparations promotes membrane repolarization and slows sinus rate.13 The potential responses to ATr in
regard to these different subtypes remains to be determined,
particularly as it relates to the chronotropic, inotropic, or other
properties of heart function. In any event, the paradoxical
finding of the ACh/ATr interactions in regard to AF duration
372
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
| www.jcvp.org
Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
J Cardiovasc Pharmacol ä Volume 69, Number 6, June 2017
Acetylcholine/Atropine Paradox
provides another approach for further studies to glean new
insights into the nature of allosteric and orthosteric mechanisms of action of various modulators and their interaction
with cholinergic agonists.
2. Scherlag BJ, Hou YL, Lin J, et al. An acute model for atrial fibrillation
arising from a peripheral atrial site: evidence for primary and secondary
triggers. J Cardiovasc Electrophysiol. 2008;19:519–527.
3. Lazzara R, Scherlag BJ, Robinson MJ, et al. Selective parasympathetic
control of the canine sinoatrial and atrioventricular nodes. Circ Res.
1973;32:393–401.
4. Loewi O. Über humorale Übertragbarkeit der Herznervenwirkung.
Pflügers Arch Ges Physiol. 1921;189:239–242.
5. Dyablova PE. The sensitizing effect of atropine to acetylcholine. Bull
Exp Biol Med. 1964;58:956–957.
6. Ashford A, Penn GB, Ross JW. Cholinergic activity of atropine. Nature.
1962;193:1082–1083.
7. Hazard R, Savini E, Renier-Cornec A. Increase in the acetylcholinesensitivity of the isolated intestine by minimal doses of atropine. Arch
Int Pharmacodyn Ther. 1959;120:369–373.
8. May LT, Lin Y, Sexton PM, et al. Regulation of M2 muscarinic acetylcholine receptor expression and signaling by prolonged exposure to allosteric modulators. J Pharmacol Exp Ther. 2005;312:382–390.
9. May LT, Christpoulos A. Allosteric modulators of G-protein-coupled
receptors. Curr Opin Pharmacol. 2003;3:551–556.
10. Lu Z, Scherlag BJ, Lin J, et al. Autonomic mechanism for complex
fractionated atrial electrograms: evidence by fast fourier transform analysis. J Cardiovasc Electrophysiol. 2008;19:835–842.
11. Sharma VK, Colecraft HM, Rubin LE, et al. Does the mammalian heart
contain only the M2 muscarinic reptor subtype. Life Sci. 1997;60:
1023–1029.
12. Shi H, Wang H, Wang Z. Identification and characterization of multiple
subtypes of muscarinic acetylcholine receptors and their physiological
functions in canine hearts. Mol Pharmacol. 1999;55:497–507.
13. Wang H, Han H, Zhang L, et al. Expression of multiple subtypes of
muscarinic receptors and cellular distribution in the human heart. Mol
Pharmacol. 2001;59:1029–1036.
14. Mao J, Scherlag BJ, Liu Y, et al. The atrial neural network as a substrate
for atrial fibrillation. J Interv Card Electrophysiol. 2012;35:3–9.
Limitations
It should be noted that we did not test the effect of ATr
alone applied to the AA on the inducibility of AF duration.
However, in a previous report14 ATr application to the AA
caused a significant reduction in the mean duration of electrical stimulation induced AF. A possible explanation for this
response may be that receptor antagonists possess certain
properties of their agonists.
CONCLUSION
Previous studies have shown that when ATr is initially
administered it attaches to the allosteric site of the M2 AChRs
leaving the orthosteric site free to be occupied by ACh. Given
after ACh, ATr becomes an effective and specific competitive
inhibitor of ACh at the orthosteric M2 AChRs. However, this
allosteric property may not apply to other AChRs which
control heart rate.
REFERENCES
1. Scherf D, Morgenbesser LJ, Nightingale EJ, et al. Further studies on the
mechanism of auricular fibrillation. Proc Soc Exp Biol Med. 1950;73:
650–654.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
www.jcvp.org |
373
Copyright Ó 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.