2007 The Authors

Doi: 10.1111/j.1742-7843.2006.00009.x
Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224

Inhibitory Effects of Artemisinin on Voltage-Gated Ion Channels in

Intact Nodose Ganglion Neurones of Adult Rats

Blackwell Publishing, Ltd.

Guofen Qiao1, Shuo Li2, Baofeng Yang1 and Baiyan Li1,3

1

Department of Pharmacology, Harbin Medical University, Biopharmaceutical Engineering Key Laboratory of Heilongjiang Province,
Incubator of State Key Laboratory, Harbin, China, 2DePauw University, Greencastle, IN, USA, and 3Department of Biomedical
Engineering, Indiana University Purdue University Indianapolis, Indianapolis, IN, USA
(Received June 14, 2006; Accepted August 30, 2006)
Abstract: Recent data show that artemisinin has anti-arrhythmic and local anaesthetic effects. To better understand the
mechanisms, the effects of artemisinin on action potential discharge and voltage-gated ion channels properties were studied
on nodose ganglion neurones of adult rats with known sensory afferent fibre type using whole cell patch and vagus nodose
slice preparation. The present data show that both depolarization and repolarization of action potentials were markedly
inhibited by artemisinin in a concentration- and time-dependent manner in either A-type or C-type nodose ganglion
neurones without change in conduction velocity. Both tetrodotoxin-sensitive (TTX-S) Na+ and tetrodotoxin-resistant (TTX-R)
Na+ currents were significantly reduced by micro-perfusion of artemisinin; the steady-state half-activation and halfinactivation for both TTX-S and TTX-R Na+ currents were shifted towards the right without changing slope factors.
Median inhibition concentration (IC50) are 68.1 M and 236.2 M for TTX-S and TTX-R Na+ currents, respectively. Total
outward K+ currents from C-type nodose ganglion neurones were blocked by artemisinin 30300 M concentrationdependently, IC50 being 104.7 M. This effect was mimicked by tetraethylammonium 15 mM. Peak currents of N-type Ca 2+
channels were also reduced significantly (IC50=344.6 M) in the presence of artemisinin, which was less effective than that
induced by 1 M -conotoxin (CTX) GIVA. Our data demonstrate that depolarization and repolarization of action potentials
recorded from either A- or C-type nodose ganglion neurones were inhibited by artemisinin in a concentration- and timedependent manner, and that this inhibitory effect of artemisinin is probably due to the non-selective inhibition of all major
ion channels functionally expressed in nodose ganglion neurones.

Artemisinin has been widely used as an effective antimalarial

[1] agent for several centuries. It is an endoproxide-containing
sesquiterpene purified from the Chinese herbal plant, Artemisia
annua L., or sweet wormwood [2]. The pharmacology and
toxicology of artemisinin has been widely studied. Except
for the antimalarial effect, other pharmacological effects have
been documented, including cytotoxicity against tumours
and cancers [3 6] and an anti-arrhythmic effect [7]. Recent
literature shows that artemisinin mobilizes intracellular
calcium [8] and modulates ion channel activities [9]. In
addition, local anaesthetic action has been reported [10].
All recent studies and our previous data [11] suggest that
artemisinin may be involved in pain regulation in the visceral
sensory system. Nodose ganglia and dorsal root ganglia have
been used to study the ion channel activity in the sensory
system. Nodose ganglia are known to contain primary sensory neurones that receive sensory information from both
mechanoreceptors and chemoreceptors, including nociceptors and baroreceptors of visceral organs. Two populations
of nodose ganglion neurones have been categorized by visceral
afferent fibre conduction velocity (CV) [12,13] and action
potential waveform characteristics [1214]. All major voltage-

Author for correspondence: Baiyan Li, Department of Biomedical

Engineering, IUPUI, 723 West Michigan Street, SL-174, Indianapolis, IN 46202, USA (fax +1-317-278-2032, e-mail
baili@iupui.edu).

gated ion channels, such as Na+ [15], K+ [16] and Ca2+ [17]
channels including channel subtypes, have been observed on
both A-type (myelinated) and C-type (unmyelinated) nodose
ganglion neurones with our nodose slice preparation [12,13].
Thus, nodose ganglion neurones have been considered as a
suitable model to functionally study the ion channel activity
of visceral sensory neurones. In the present study, the
pharmacological effects of artemisinin on action potential
discharge characteristics, the properties of voltage-gated
Na+, K+ and Ca2+ channels, and ion channel activation and
inactivation profiles were investigated. The data may provide
comprehensive insight into the ionic mechanisms of antiarrhythmic and local anaesthetic actions of artemisinin.
Materials and Methods
Chemicals and reagents. Artemisinin was provided by Shenyang
Pharmaceutical University (Shenyang, China) and the stock solution of artemisinin was prepared to 10 mM with recording solution
[7] and diluted into 11000 M. Tetrodotoxin (TTX) 1.0 M was
applied for the separation of TTX-R Na+ channel subtypes. Tetraethylammonium (TEA) (Sigma Chemical Co., St. Louis, MO, USA)
15 mM and -conotoxin (CTX) GIVA (Alomone Labs, Jerusalem,
Israel) 1.0 M were used as reference drugs. The reagents for preparation of nodose ganglion neurones (vagus nerve-nodose ganglion
slice preparation) were exactly the same as in the previous study
[11-13]. Na-GTP and Na-ATP (Sigma Chemical Co.) were used to
prepare the pipette solution. Earles balance salt solution (Sigma
Chemical Co.) was used for dissolving enzymes (collagenase type II

218

GUOFEN QIAO ET AL.

and Trypsin-3X, Worthington Biochemical Co., Lakewood, NJ, USA).

Other chemicals used in this experiment for recording solution were
from Sigma and Fisher (Woodlands, TX, USA).
Experimental animals. Adult (150 250 g) Sprague-Dawley rats of
either sex were selected for nodose slice preparation. All experimental
animals were from the Experimental Animal Center of Harbin
Medical University, Harbin, China, and housed in an animal facility
at least 3 days before use. All experimental protocols used in this
experiment have previously been approved by the School of Medical
Science, Harbin Medical University.
Vagus nerve-nodose ganglion preparation. Briefly, adult rats were
placed in an airtight induction chamber for inhalation of the anaesthetic metofane (methoxyflurane, Schering-Plough Animal Health,
Union, NJ, USA). Upon lack of reflex response to tail pinch, the
animals were immediately sectioned at the mid-auxiliary region for
preserving at least 2 cm of vagus. Nodose ganglia with attached
vagus nerves were carefully excised. The ganglia were sliced without
damaging the vagus, and then treated with 1 mg/ml collagenase and
0.5 mg/ml trypsin at 37C for 40 and 20 min., respectively. The
nodose slices were placed in a 48C recording solution at least
1 hr before patch recordings. All surgical procedures were conducted
under stereomicroscope (magnification 40).
Recording solutions. Extracellular solution (in mM): (i) for action
potential recordings: 137.0 NaCl, 5.4 KCl, 1.0 MgCl2, 10.0 glucose
and 10.0 HEPES; (ii) for K+ channel recordings: 140.0 K-Aspartate
(K-ASP), 3.0 MgCl2 and 10.0 HEPES; (iii) for Na+ channel recordings:
50.0 NaCl, 10.0 MgCl2, 10.0 glucose and 25.0 HEPES; and (4) for
Ca2+ recordings: 140.0 TEA-Cl, 5.0 4-aminopyridine, 15.0 HEPES,
20.0 glucose and 2.0 CaCl2. Pipette solution (in mM): (i) for action
potential recordings: 6.0 NaCl, 50.0 KCl, 50.0 K2SO4, 5.0 MgCl2
and 10 HEPES; (ii) for K + channel recordings: 137 NMDG,
5.4 KCl, 1.0 MgCl2, 10.0 glucose and 10.0 HEPES; (iii) for Na+
channel recordings: 7.0 NaF, 195.2 NMDG, 165.7 TEA-Cl, 2.0 MgCl2
and 10.0 HEPES; and (iv) for Ca2+ recordings: 124.0 CsCl, 5.0
MgCl2, 10.0 glucose and 10.0 HEPES. The pH for extracellular and
intracellular solutions was adjusted to 7.30 and 7.20, respectively.
-Mannitol (Sigma Chemical Co.) was used for adjustment of
osmolarity of extracellular and intracellular solutions to 310 and 290,
respectively. For providing a physiological intracellular environment,

the final [Ca2+]i for all intracellular solutions were adjusted to

10 nM by adding 4.0 mM Bapta-K, (Sigma Chemical Co.),
0.25 mM CaCl2, and 2.0 mM each of Na-ATP and Na-GTP to the
pipette solution just before the experiments.
Electrophysiological techniques. Briefly, current and voltage clamp
experiments were performed with the whole cell mode of patch
clamp. Patch pipettes were pulled (P-87, Sutter Instruments,
Novato, CA, USA) and polished (MF-830, Narishigi), electrode
resistance being about 1.5 2.5 M. All patch recordings were
conducted with Axopatch 200B amplifier (Axon). All experiments
were conducted and visualized under Axioskop routine microscope
(Zeiss). The details regarding individual protocols, holding potentials, and test intervals are presented in the Results section. The
entire experimental protocol, data acquisition, storage, displaying
and preliminary waveform analysis were accomplished using the
pClamp (version 9.0.2, Axon) software package operating on a PC
platform. All experiments were conducted at 2123C.
Data analysis and statistics. Data acquisition and analysis were performed with pClamp and Microsoft Excel. Conductancevoltage
(GV) relationships according to g = I/(VVr), where I is the peak
current at voltage (V), and Vr is the reversal potential extrapolated
from the linear portion of the currentvoltage (IV) relationships.
A Boltzamann function of the form 1/{1 + exp [(VV1/2)/k]}, where
V1/2 is the voltage of half activation or inactivation and k is a slope
factor, was used to fit GV relationships and inactivation curves.
The pooled data were presented as mean S.D., and t-tests were
used to evaluate the statistical significance between control and
tested groups.

Results
Effects of artemisinin on action potential discharge properties
of identified A- or C-type nodose ganglion neurones.
A-type identified by afferent fibre CV. In this part of the
experiments, six nodose ganglion neurones with faster CVs
at a range of 1015 m/sec. (average: 11.2 3.45 m/sec.) were
classified into A-type nodose ganglion neurones (fig. 1A).
Except for fast CVs over 2.5 m/sec., A-type nodose ganglion

Fig. 1. Inhibitory effects of artemisinin on depolarization and repolarization characteristics of action potentials recorded from A-type
(myelinated, A, n = 6) and C-type (unmyelinated, B, n = 14) nodose ganglion neurones identified by sensory afferent fibre conduction
velocity (CV) in intact nodose slice preparation of adult rats.
2007 The Authors
Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224

219

ARTEMISININ/NGNS/WHOLE-CELL CURRENTS
Table 1.

Effects of artemisinin on action potential discharge properties of A-type (n = 6) and C-type (n = 14) nodose ganglion neurones identified by
afferent fibre conduction velocity (CV).
Groups

CV
(m/sec.)

APD50
(m/sec.)

PeakAP
(mV)

A-type (myelinated) nodose ganglion neurones, CV = 11.2 3.45 m/sec.

Control
11.2 3.45
0.76 0.16
52.9 4.59
Artemisinin 10 M
11.2 3.43
0.78 0.14
51.6 4.02
Artemisinin 30 M
11.1 3.47
1.29 0.45**
45.1 2.03*
Artemisinin 100 M
11.2 3.51
1.56 0.76**
35.4 2.64**
Artemisinin 100 M/1 min.
11.2 3.49
2.14 0.94**
31.9 3.03**

Artemisinin 100 M/3 min.

11.1 3.52
2.46 0.98**
28.2 4.12**
C-type (unmyelinated) nodose ganglion neurones, CV = 0.64 0.14 m/sec.
Control
0.64 0.14
2.76 1.97
56.9 4.41
Artemisinin 30 M
0.64 0.13
4.39 0.98**
53.5 3.03
Artemisinin 100 M
0.64 0.14
8.44 2.13**
44.7 4.16**
Artemisinin 300 M/1 min.
0.64 0.15
12.6 3.64**
40.9 4.08**
Artemisinin 300 M/3 min.
0.65 0.14
11.7 2.98**
36.5 4.12**

Artemisinin 300 M/10 min.

0.65 0.14
12.2 2.40**
34.4 3.31**

PeakAHP
(mV)

UVAPD50
(V/sec.)

DVAPD50
(V/sec.)

69.4 1.38
67.2 1.10*
65.7 0.82*
63.8 0.85**
62.2 0.71**
60.7 0.83**

278 76
264 63
205 38*
121 24**
98 21**
67 16**

168 26
141 16*
88 12**
49 11**
36 10**
29 8**

68.2 2.30
65.1 0.91*
61.9 1.20**
60.5 1.04**
58.2 1.34**
57.6 2.04**

94.4 22.4
67.9 17.1*
39.9 19.8**
21.8 11.0**
16.6 6.45**
14.0 3.92**

45.3 16.7
29.3 9.07**
19.9 5.96**
15.6 6.67**
11.2 4.44**
10.1 2.56**

Data were represented as mean S.D. *P < 0.05 and **P < 0.01 compared with control group. Action potentials were evoked from three of
six A-type recorded nodose ganglion neurones by vagal stimulation; Action potentials were elicited from 5 of 14 recorded C-type nodose
ganglion neurones by vagal stimulation. APD50, action potential duration at the point of 50% height of action potential; Peak AP, peak of
action potential; PeakAHP, peak of after-hyperpolarization; UVAPD50, action potential upstroke velocity at the point of APD50; DVAPD50, action
potential downstroke velocity at the point of APD50.

neurones also possessed some other unique features, which

were distinct with C-type nodose ganglion neurones, like a
brief action potential duration, much faster depolarization
and repolarization, and smooth trajectory repolarization
phase over the course of action potential waveform. 10
100 M, CVs and resting membrane potentials were clearly
unchanged, but action potential profiles were dramatically
altered in a concentration-dependent manner (table 1). With
100 M of artemisinin (fig. 1A), the peak of action potentials
were significantly reduced by 34% (P < 0.01), APD50 (action
potential duration at the point of 50% height of action
potential) was prolonged by about 100% (P < 0.01), the rate
of depolarization or repolarization was remarkably slowed
down by at least 50% (P < 0.01) or 60% (P < 0.01) respectively,
and the peak of hyperpolarization was also about 60% less
negative (P < 0.01). In addition, time-dependent inhibition
was also observed when cells were continuously exposed to
artemisinin. After a 3-min. exposure to artemisinin, action
potentials from three (50%) out of six recorded A-type nodose
ganglion neurones were no longer able to be elicited by
vagal stimulation.
C-type identified by afferent fibre CV. Nodose ganglion
neurones with slower CVs less than 2.5 m/sec. were generally
classified into C-type. Other features that also help to distinguish C-type from A-type nodose ganglion neurones are
slower rates of depolarization and repolarization of action
potentials, and the Hump over the course of repolarization.
In this experiment, 14 C-type action potentials were obtained
and identified by stimulating vagus (fig. 1B). The average of
CV was 0.64 0.14 m/sec. (range: 0.5 0.7 m/sec.), APD50 was
about three times wider than that of A-type with significant
Hump during the repolarization. After microperfusion of

artemisinin 30300 M to neurones, similar concentrationand time-dependent inhibition of artemisinin were established
but higher concentration of artemisinin was required. As
compared with A-type nodose ganglion neurones, APD50 of
C-type nodose ganglion neurones were prolonged more than
300%, the rates of depolarization and repolarization were both
reduced by at least 70%, the peaks of after-hyperpolarization
were about 130% less negative, and the reduction of peaks
of action potentials were similar to those of A-type nodose
ganglion neurones when artemisinin 300 M was applied.
Like A-type nodose ganglion neurones, artemisinin had no
effects on CVs and resting membrane potential of the C-type
neurones (table 1). When cells were exposed to artemisinin
for more than 10 min., action potentials were no longer able
to be elicited by vagal stimulation by 9 out of 14 recorded
C-type nodose ganglion neurones.
In accordance with the effects of artemisinin on action
potential discharge characteristics, it is easy for us to predict
that artemisinin may influence the Na+ channel currents
i.e. TTX-S and TTX-R Na+, K+ and perhaps Ca2+ channel
currents because both depolarization and repolarization
of action potentials from either A- or C-type nodose ganglion
neurones were inhibited. Therefore, the following voltage
clamp experiments were designed to test our hypothesis.
Artemisinin-induced modulation of TTX-S Na+ channel
currents. To study the effect of artemisinin on TTX-S Na+
channel currents, A-type nodose ganglion neurones were
targeted because TTX-S Na+ channel subtype is the only
Na+ channels functionally expressed by A-type nodose ganglion
neurones. Six typical A-type neurones were identified by
sensory afferent fibre CVs (average CV = 12.8 1.86 m/sec.;
fig. 2A, inset) recorded from cell-attached configuration just

2007 The Authors

Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224

220

GUOFEN QIAO ET AL.

Table 2.

Effects of artemisinin 30 1000 M on TTX-S (n = 6) and TTX-R (n = 11) Na+ channel properties in nodose ganglion neurones. A- and Ctype nodose ganglion neurones were identified in nodose slice preparation by afferent fibre CVs collected from cell-attached configuration
just before the ruptured patch. Average conduction velocity (CV) was 12.8 1.86 m/sec. for A-type and 0.58 0.31 m/sec. for C-type nodose
ganglion neurones.
Peak of
IV (mV)

Peak of
IV (nA)

A-type nodose ganglion neurones (myelinated fibre)

Control
26.5 3.98
7.52 0.65
Artemisinin 10 M
22.8 2.46*
6.48 0.47*
Artemisinin 30 M
19.8 2.04** 5.06 0.31**
Artemisinin 100 M
17.1 1.96** 3.11 0.29**
TTX 500 nM
17.7 1.87** 1.25 0.14**
C-type nodose ganglion neurones (myelinated fibre)
Ctrl (TTX 1 M)
15.4 2.96
3.57 0.49
Artemisinin 30 M
15.1 2.21
3.29 0.45
Artemisinin 100 M
14.7 1.84
2.37 0.33*
Artemisinin 300 M
12.2 0.94*
1.52 0.36**
Artemisinin 1000 M 10.6 0.82** 0.72 0.24**

Decrease of
IV (%)

V1/2 for
activation (mV)

Slope
factor (k)

V1/2 for
inactivation (mV)

Slope
factor (k)

13.8 4.20*
32.7 6.74**
58.6 5.83**
83.4 3.66**

34.4 1.96
32.2 1.99
29.1 2.06*
20.8 1.84**
17.6 1.26**

5.89 1.3
5.87 1.47
5.73 1.12
5.44 1.43
5.06 1.28

60.2 3.66
58.2 4.61
54.4 4.19*
47.8 3.49**
39.7 3.55**

6.71 1.21
6.82 1.12
7.21 1.03
7.43 1.40
7.76 1.38

7.88 1.22*
33.7 4.96**
57.5 5.77**
79.0 5.67**

15.1 1.44
13.0 3.12
10.4 2.64**
7.88 2.49**
5.69 1.86**

3.94 1.44
3.92 1.61
3.79 1.89
3.72 1.77
3.61 1.68

33.6 4.95
31.5 5.44
28.4 4.41*
24.6 4.09**
20.9 4.13**

3.62 0.96
3.67 1.16
3.79 1.21
3.97 1.57
4.13 1.64

Data were represented as mean S.D. *P < 0.05 and **P < 0.01 compared with control group.

before the ruptured patch. TTX-S Na+ channel currents from

those A-type nodose ganglion neurones were also recorded
by the step pulse voltage commends (fig. 2A). From IV
relationship (fig. 2C), TTX-S Na+ channel currents activated

Fig. 2. Effects of artemisinin on TTX-S Na+ currents. (A) Sample

traces of TTX-S Na+ currents recorded from identified A-type
nodose ganglion neurone (see inset). (B) Typical traces of peak
TTX-S Na+ currents before and after artemisinin 10 100 M and
TTX 500 nM. (C) Current-voltage (IV ) relationships (n = 6) of
TTX-S Na+ currents. Inset: Afferent fibre conduction velocity (CV )
obtained from cell-attached before the rupture patch.

at more hyperpolarized membrane voltage near 60 mV, the

peak currents appeared at 26.5 3.98 mV with average peak
of 7.52 0.65 nA. The reversal potential of TTX-S Na+
channel currents was near +45 mV. After locally applying
artemisinin 10100 M, the peaks of IV curve were
reduced significantly (IC50 = 68.1 M) and markedly shifted
towards the right about 10 mV (fig. 2B, table 2). To investigate
artemisinin-induced activation and inactivation characteristics,
both normalized steady-state activation and inactivation
of TTX-S Na+ channel currents were fitted with a Boltzmann equation. Half-activation and half-inactivation (V1/2)
of TTX-S Na+ channel currents were 34.4 1.96 mV and
60.2 3.66 mV, respectively, and were shifted by artemisinin
to the less negative voltage in a concentration-dependent
manner without changing slope factors (fig. 3, table 2). The
artemisinin-induced inhibitory effects on both activation and
inactivation of TTX-S Na+ channel currents were less
potent than those induced by TTX 500 nM.
Artemisinin-induced modulation of TTX-R Na+ channel
currents. In these experiments, typical C-type nodose ganglion neurones (n = 11) were identified by CVs (average
CV = 0.58 0.31 m/sec.; fig. 4A, inset) and control TTX-R
Na+ channel currents were obtained in the presence of
1.0 M of TTX (fig. 4A). From the IV relationship (fig. 4C),
the average peak of IV curve was 3.75 0.43 nA and
appeared at 15.4 2.96 mV. Artemisinin significantly and
concentration-dependently changed the characteristics of
TTX-R Na+ channel currents. The peak of current was
shifted to the right (P < 0.01) with the highest concentration
of artemisinin; the average peak current was decreased by
7.88%, 33.7%, 57.5% and 79.0% (IC50 = 236.2 M), respectively,
after microperfusion of artemisinin 301000 M (fig. 4B,
table 2). Normalized GV relationships (V1/2) for activation
(control = 15.1 1.44 mV) and steady-state inactivation
(control = 33.6 4.95 mV) of TTX-R Na+ channel currents

2007 The Authors

Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224

ARTEMISININ/NGNS/WHOLE-CELL CURRENTS

221

Fig. 3. Effects of artemisinin on steady-state activation (A) and inactivation (B) of TTX-S Na + currents. Mean data values were calculated
using a normalized chord conductance. Boltzmann curves are best fit to the mean TTX-S (n = 6) and normalized conductance-voltage
(GV) curves were obtained before and after artemisinin applications. Vertical bars show mean S.D.

were both shifted towards the right up to 5.69 1.86 mV

and 20.9 4.13 mV, respectively (fig. 5, table 2), whereas
the significant change of slope factors was not observed.
Effects of artemisinin on outward K+ currents. Because both
A- and C-type nodose ganglion neurones functionally

Fig. 4. Effects of artemisinin on TTX-R Na+ currents. (A) Sample

traces of TTX-R Na+ currents recorded from identified C-type nodose
ganglion neurone (see inset) in the presence of TTX 1.0 M. (B) Typical
traces of peak TTX-R Na+ currents from the control (in presence of
1.0 M TTX) and groups in the presence of artemisinin 30 1000 M.
(C) Current-voltage (IV) relationships (n = 11) of TTX-R Na+
currents. Inset: Afferent fibre conduction velocity (CV) (0.57 m/sec.)
obtained from cell-attached configuration before the rupture patch.

expressed similar K+ channel subtypes in adult rats [16] and

for technical reasons, C-type nodose ganglion neurones
were selected for evaluating the effect of artemisinin on total
outwards K+ currents. Before the voltage clamp experiments,
action potentials (fig. 6A) were evoked from five C-type
nodose ganglion neurones by vagal stimulation. CVs
collected from those neurones were all less than 1 m/sec. After
completely changing the recording solution, outwards wholecell K+ currents were recorded (fig. 6B). From the IV
relationships, the average peak measured at a steady-state
of K+ currents was 21.8 2.56 nA and K+ currents were
activated between 50 and 40 mV. Artemisinin 30 300 M
induced a dramatic reduction (IC50 = 104.7 M) of total K+
currents in a concentration-dependent manner (fig. 6C).
The average peaks of K+ currents measured at a steady-state
were reduced about 27.6%, 53.6% and 74.1%. The maximal
inhibition caused by artemisinin 300 M was relatively
equivalent to that (79.7%) induced by TEA 15 mM.
Inhibitory effect of artemisinin on N-type Ca2+ currents. Our
pilot experimental data with nodose slice preparation
showed that both A- (data not shown) and C-type nodose
ganglion neurones [12,13] expressed N-type Ca2+ channels.
About similar amounts (70%) of total Ca2+ channel currents
were blocked by -CTX GIVA 1.0 M, a specific N-type
Ca2+ channel blocker; therefore, the cell type is not an
important issue in this particular investigation. Total
inwards Ca2+ currents were observed (fig. 7B) on typical
C-type nodose ganglion neurones (n = 5) and CVs collected
from those five C-type neurones were all less than 1 m/sec.
(fig. 7A). The results show that the average peak of the IV
curve (fig. 7C) for N-type Ca2+ currents was 3.23 0.73 nA
and appeared at 4.46 1.28 mV. Compared with the control group, the peaks of N-type Ca2+ currents were dosedependently reduced (IC50 = 344.6 M) by microperfusion
of artemisinin 30300 M. The maximal reduction of
N-type Ca2+ currents (1.54 0.51 nA) caused by artemisinin
300 M was about 52.3% less than that induced by -CTX

2007 The Authors

Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224

222

GUOFEN QIAO ET AL.

Fig. 5. Effects of artemisinin on steady-state activation (A) and inactivation (B) of TTX-R Na + currents. Mean data values were calculated
using a normalized chord conductance. Boltzmann curves are best fit to the mean TTX-R (n = 11) and normalized conductance-voltage
(GV) curves were obtained before and after artemisinin applications. Vertical bars show mean SD.

GIVA 1.0 M (67% reduction, fig. 7D), suggesting that

N-type Ca2+ channel currents recorded from C-type nodose
ganglion neurones were only partially inhibited by artemisinin
300 M.
Discussion
It is well known that several kinds of ion channels are
involved in the course of action potential. In the nodose
neurones, the Na+ channels, including TTX-S Na+ and
TTX-R Na+ channels, N-type Ca2+ channels and K+ channels, including DTX-sensitive, 4-aminopyridine-sensitive
(transient), TEA-sensitive K+ channels, and Ca2+ activated
K+ channel, all contribute to the formation of the action

potential waveform. Any of those ion channels modulated

by the drug/chemicals will change the waveform of action
potentials. In the present study, we have shown that depolarization of action potentials from either A- or C-type
nodose ganglion neurones was inhibited by artemisinin
concentration-dependently without changing the sensory
afferent CVs. Data were consistent with our previous
findings [11] and suggest that at least voltage-gated Na+ ion
channels might be involved in this inhibitory effect of
artemisinin because, in general, the Na+ ion influx through
voltage-gated sodium channels, especially TTX-S Na+ channels,
occurring during depolarization of action potential plays a
critical role in the generation of action potential. This would
well explain the failure to induce action potentials by vagal

Fig. 6. Artemisinin modulated total outward K+ currents. (A) Identified C-type nodose ganglion neurone (n = 5) by conduction velocity
(CV). (B) Control K+ current traces were recorded from the same C-type nodose ganglion neurone shown as (A). (C) Current-voltage (IV)
curve obtained from before and after artemisinin 30 300 M or TEA 15 mM. Data were presented as mean S.D.
2007 The Authors
Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224

ARTEMISININ/NGNS/WHOLE-CELL CURRENTS

223

Fig. 7. Artemisinin modulated total Ca2+ currents. (A) The response elicited by vagal stimulation were recorded from patch electrode under
cell-attached configuration and conduction velocity (CV) indicated that this is a typical C-type nodose ganglion neurone. (B) Typical whole
cell Ca2+ currents were from the same nodose ganglion neurone. (C) A typical current-voltage (IV ) curve that was obtained from the
recording (B). (D) Effects of Artemisinin and -CTX GIVA 1.0 M on N-type Ca2+ currents. Data were presented mean S.D., n = 5.
*P < 0.05 and **P < 0.01 compared with the control group.

stimulation when the cells were treated with artemisinin

for a longer time. The present data also show that the repolarization of action potentials from A- or C-type nodose
ganglion neurones was inhibited dramatically and this effect
could be affected by artemisinin either indirectly due to
inhibition of Na+ channel resulting in less recruitment of K+
currents, or directly to the inhibition of other voltage-gated
ion channels like TTX-R Na+, K+ and Ca2+ channels, which
are the major ion channels for regulation of action potential
repolarization; in other words, the influence of artemisinin
on repolarization might be the secondary effect due to
the inhibition of artemisinin on depolarization.
To test our hypothesis, A-type (myelinated fibre) nodose
ganglion neurones was first selected and identified by visceral
sensory afferent CV (fig. 2A, inset) for studying the effect of
artemisinin on TTX-S Na+ channels because TTX-S Na+
channel is the only Na+ channel expressed by A-type nodose
ganglion neurones (fig. 2A). The data demonstrate that the
peak of TTX-S Na+ channel was dramatically reduced by
artemisinin concentration- and time-dependently (table 2)

with IC50 at 68.1 M without the effect on CVs. This effect

of artemisinin was similar to, but less potent than that of
TTX 500 nM. Our previous study [18] shows that ketamine
not only inhibited Na+ channels in nodose neurones, but
also sensory afferent CVs, as the Na+ channels are also
expressed in the membrane around the Ranvier node. At
this point, the present data still could not explain why
artemisinin blocked Na+ channels without affecting CVs.
Likewise the CVs of A-type nodose ganglion neurones were
not changed by ketamine but CVs of the C-type neurones
were [19]. The activation and inactivation gating profiles of
the TTX-S Na+ channel were also evaluated and our results
indicate that normalized activation and steady-state inactivation (V1/2, fig. 4A,B) were shifted towards the right
without changing the slope factor (k). In order to study of the
effect of artemisinin on TTX-R Na+ channel, C-type (unmyelinated fibre) nodose ganglion neurones were identified
by CVs (fig. 4A, inset) and TTX-R Na+ channels were recorded
in the presence of TTX 1.0 M (fig. 4A). The effects of artemisinin on TTX-R Na+ channels and activation/inactivation

2007 The Authors

Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224

224

GUOFEN QIAO ET AL.

profiles were very similar to those on TTX-S channel currents but less potent than those of artemisinin on A-type
nodose ganglion neurones, IC50 being 236.2 M.
According to the current clamp data from the previous
and current experiments, it is hard to exclude that artemisinin
may directly influence the repolarization of action potentials.
The documented literature show that artemisinin did increase
intracellular Ca2+ mobilization in cardiomyocytes [8], but a
clear conclusion can not be reached whether artemisinin is
involved in the modification of action potential through
Ca2+ channel. To address the question how artemisinin
affects the repolarization of action potential, voltage-gated
K+ and Ca2+ channel activities were observed through voltage clamp experiments. Our data show that not only K+
channels but also N-type Ca2+ channels were attenuated by
artemisinin in a concentration-dependent manner, IC50 for
K+ and N-type Ca2+ currents being 104.7 M and 344.6 M,
respectively. These data demonstrate that artemisinin was
directly involved in the inhibition of action potential repolarization. The IC50 data of artemisinin on TTX-S and TTXR Na+ channels and K+ channels, and the concentrationand time-dependent inhibition of artemisinin on action
potentials are well consistent with those of ketamine [18].
Together our findings suggest that artemisinin most likely
acts at ion channel levels like ketamine. The effects of
artemisinin on ion channel activities were reversible and the
results were consistent with previous findings [7,19]. The
concentration of artemisinin used in this study is relevant to
the plasma levels after a single dosage in patients [20] or
health adults [21]. The possible explanation of the controversy between our present and the previous data, which
show that artemisinin mobilised intracellular Ca2+ concentration [8], might be due to the difference between sensory
neurones and cardiomyocytes.
From the above results and discussion we may conclude
that artemisinin concentration and time dependently inhibited depolarization and repolarization of action potentials
in both myelinated A-type and unmyelinated C-type nodose
ganglion neurones, prolonged the action potential duration
with no effects on resting membrane potential (RMP) and
CVs, and that those inhibitory effects were directly due to
artemisinins non-selective blocks on voltage-gated TTX-S
and TTX-R Na+ channels, K+ channels and N-type Ca2+ channels
that were functionally expressed in nodose ganglion neurones.
The present study demonstrates the anti-arrhythmic and local
anaesthetic actions of artemisinin at the ion channel level.
References
1 Klayman DL. Qinghaosu (artemisinin): an antimalarial drug
from China. Science 1985;228:4955.

2 Hien TT, White NJ. Qinghaosu. Lancet 1993;341:6038.

3 Sun WC, Han JX, Yang WY, Deng DA, Yue XF. Antitumor
activities of 4 derivatives of artemisic acid and artemisinin B
in vitro. Acta Pharmacol Sin 1992;13:5413.
4 Deng DA, Xu CH, Cai JC. Derivatives of artemisinin B with
antileukemia activity. Yao Xue Xue Bao 1992;27:31720.
5 Moore JC, Lai H, Li JR, Ren RL, McDougall JA. Oral administration of dihydroartemisinin and ferrous sulfate retarded
implanted fibrosarcoma growth in the rat. Cancer Lett
1995;98:837.
6 Efferth T, Dunstan H, Sauerbrey A, Miyachi H, Chitambar CR.
The anti-malarial artesunate is also active against cancer. Int J
Oncol 2001;18:76783.
7 Yang BF, Luo DL, Bao LH, Zhang YC, Wang HZ. Artemisinin
blocks activating and slowly activating K+ current in guinea pig
ventricular myocytes. Acta Pharmacol Sin 1998;19:26972.
8 Ai J, Gao HH, Wang L, Luo DL, Yang BF. Effects of matrine,
artemisinin, and tetrandrine on [Ca2+]i in guinea pig ventricular
myocytes. Acta Pharmacol Sin 2001;22:5125.
9 Shi YL, Xu YF, Gu C, Li Y. Reversible elimination of K+ and
Na+ currents at motor nerve terminals by SM486, a derivative
of artemisinin. Acta Physiol Sin 1995;47:2530.
10 Huang FS, Hu Q, Shi YL. The inhibitory effects of artemisininderivatives on Na+ and K+ channels in comparison with those
of procaine. Acta Physiol Sin 1998;50:14552.
11 Qiao GF, Yang BF, Li WH, Li BY. Effects of artemisinin on
action potentials from C-type nodose ganglion neurons. Acta
Pharmacol Sin 2003;24:93742.
12 Li BY, Schild JH. Comparisons of somatic action potentials
from dispersed and intact rat nodose ganglia using the patch
clamp technique. Acta Pharmacol Sin 2002a;23:4819.
13 Li BY, Schild JH. Patch clamp electrophysiology in the nodose
ganglia of the adult rat. J Neurosci Methods 2002b;83:886
97.
14 Schild JH, Clark JW, Hay M, Mendelowitz D, Andresen MC,
Kunze DL. A- and C-type rat nodose sensory neurons: model
interpretations of dynamic discharge characteristics. J Neurophysiol 1994;71:233858.
15 Li BY, Alfrey KD, Schild JH. Correlation between the activation and inactivation gating profiles of the TTX-resistant Na+
current from fluorescently identified aortic baroreceptor neurons of the adult rat. FASEB J 2005;19:A606.
16 Li BY, Schild JH. Differential distribution of voltage-gated K+ ion
channels in adult rat aortic baroreceptor neurons with myelinated
and unmyelinated afferent fibre. FASEB J 2006;20:A775.
17 Schild JH, Li BY. The N-type Ca2+ current is not responsible
for the repolarization hump in the action potential of nodose
sensory neurons. FASEB J 2001;15:A1150.
18 Jin YH, Bailey TW, Doyle MW et al. Ketamine differentially
blocks sensory afferent synaptic transmission in medial nucleus
tractus solitarius (mNTS). Anesthesiology 2003;98:12132.
19 Zhou J, Du ZM, Qin XH, Sun JP, He SZ, Yang BF. Mechanism of antiarrhythmic effect of the anti-malarial artemisinin.
Acta Pharm Sin 1999;34:56972.
20 De Vries PJ, Tran KD, Nguyen XK et al. The pharmacokinetics
of a single dose of artemisinin in patients with uncomplicated
falciparum malaria. Am J Trop Med Hyg 1997;56:5037.
21 Ashton M, Goufigh G, Trinh HT et al. Artemisinin pharmacokinetics in healthy adults after 250, 500 and 1000 mg single
oral doses. Biopharm Drug Dispos 1998;19:24550.

2007 The Authors

Journal compilation 2007 Nordic Pharmacological Society. Basic & Clinical Pharmacology & Toxicology, 100, 217224