Am J Physiol Gastrointest Liver Physiol

281: G255G266, 2001.

Evidence supporting presence of two pacemakers in

rat colon
` ,1 ELENA ALBERTI,1 ESTER FERNA
NDEZ,1
LIDIA PLUJA
2
NEZ1
HANNE BIRTE MIKKELSEN, LARS THUNEBERG,2 AND MARCEL JIME
1
Department of Cell Biology, Physiology, and Immunology, Universitat Auto`noma
de Barcelona, 08193 Bellaterra, Barcelona, Catalunya, Spain; and 2Institute of Medical
Anatomy, University of Copenhagen, DK-2200 Copenhagen, Denmark
Received 16 December 1999; accepted in final form 28 February 2001

Pluja`, Ldia, Elena Albert, Ester Fernandez, Hanne

Birte Mikkelsen, Lars Thuneberg, and Marcel Jimenez. Evidence supporting presence of two pacemakers in rat
colon. Am J Physiol Gastrointest Liver Physiol 281:
G255G266, 2001.Intracellular microelectrodes and organ
bath techniques were used to study spontaneous cyclic electrical and mechanical activity in the rat colon. Electron
microscopy and immunohistochemical studies showed two
major populations of interstitial cells of Cajal (ICC): one
associated with Auerbachs plexus (ICC-AP) and one with the
submuscular plexus (ICC-SMP). The ICC-SMP network
partly adhered to the submucosa when removed and was
generally strongly damaged after separation of musculature
and submucosa. Similarly, longitudinal muscle removal severely damaged AP. Two electrical and mechanical activity
patterns were recorded: pattern A, low-frequency (0.51.5
cycles/min), high-amplitude oscillations; and pattern B, highfrequency (1315 cycles/min), low-amplitude oscillations.
Pattern A was recorded in preparations with intact AP but
absent in those without intact AP. Pattern B was recorded in
preparations with intact SMP but was absent in those lacking SMP. With full-thickness strips, the superimposed patterns A and B were recorded in circular muscle. When longitudinal muscle mechanical activity was recorded, only
pattern A was present. We conclude that two pacemakers
regulate rat colonic cyclic activity: the ICC-SMP network
(responsible for cyclic slow waves and small-amplitude contractions) and the ICC-AP network (which may drive the
cyclic depolarizations responsible for high-amplitude contractions). This is the first report showing consistent slow
wave activity in the rodent colon.

smooth muscle cells from

the gastrointestinal tract usually displays rhythmic
slow waves. These slow waves correlate with cyclic
contractions in many gastrointestinal smooth muscles.
For a long time, the cyclic activity was believed to be
myogenic in origin because neural blockers were not
able to modify rhythmicity. Interstitial cells of Cajal
(ICC) were described for the first time by Cajal (3). In
1982 (28), it was suggested that ICC were probably

pacemaker cells that mediate the input to smooth muscle cells to oscillate cyclically. For the last 20 years,
many physiologists have used different methodologies
to correlate ICC and slow waves, including dissection
experiments, cytotoxic chemicals to lesion ICC, ICC
isolation, and more recently, genetic and developmental models (for review, see Refs. 4, 11, 25, and 29). The
discovery that ICC express c-kit, the protooncogene
that encodes the receptor tyrosine kinase Kit, allowed
many immunohistochemical studies (12, 17, 18, 30, 31)
to be performed with Kit antibodies, improving knowledge about ICC.
The distribution of ICC within muscle layers differs
between regions of the gastrointestinal tract. In the
small intestine, ICC are located near Auerbachs
plexus (ICC-AP) and near the deep muscular plexus
(ICC-DMP) (22, 24, 28). The main pacemaker regulating peristalsis is probably the ICC-AP network (4a, 10).
However, the small intestine DMP might also contribute to rhythmicity and be involved in segmentation
(13, 28a). In the colon, ICC are located near the AP
(ICC-AP) and submuscular plexus (ICC-SMP) (1, 2,
23). In the canine colon, the slow wave activity (56
cycles/min) depends on the presence of an intact ICCSMP network (6, 14, 26). Similarly, myenteric potential oscillations (MPO; 1620 cycles/min) appear to
depend on the presence of ICC-AP (27). Accordingly,
two pacemakers might be working simultaneously, and
the final electrical activity results from the combined
activities of both pacemakers (27). In the human colon,
the slow wave activity probably depends on the interaction between ICC-SMP and ICC-AP. Slow electrical
oscillations lasting 9 s occur at a frequency of 3 cycles/
min. These oscillations were described as slow waves
generated near the SMP (21). Near the myenteric border of the circular muscle, MPO similar to those described in the canine colon have also been reported
(21).
In small rodents, the pacemaker activity of the colon
is not well understood and might involve cyclic depolarizations and slow waves. In mouse colon, cyclic de-

Address for reprint requests and other correspondence: M. Jimenez, Dept. of Cell Biology, Physiology, and Immunology, Veterinary
Faculty, Universitat Auto`noma de Barcelona, 08193 Bellaterra, Barcelona, Catalunya, Spain (E-mail: marcel.jimenez@uab.es).

The costs of publication of this article were defrayed in part by the

payment of page charges. The article must therefore be hereby
marked advertisement in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.

interstitial cell of Cajal; slow waves; smooth muscle

THE MEMBRANE POTENTIAL OF

http://www.ajpgi.org

0193-1857/01 $5.00 Copyright 2001 the American Physiological Society

G255

G256

PACEMAKER ACTIVITY IN RAT COLON

polarizations occur every 3 min. These depolarizations

allow the muscle to reach the threshold to open L-type
Ca2 channels, giving rise to smooth muscle action
potentials. However, the origin of the cyclic activity is
controversial. One hypothesis involves a neurogenic
mechanism that is partially resistant to L-type Ca2
channel blockers (16) and a second hypothesis derived
from developmental studies (32) involves ICC probably
located near the myenteric plexus. In the rat colon, we
(20) have reported the presence of cyclic depolarizations similar to those described in the colon of the
mouse. It is very difficult to record slow waves in
isolated mouse circular muscle (15, 32). Small slow

Fig. 1. Organ bath recordings showing the spontaneous cyclic mechanical activity displayed by the circular (A) and longitudinal muscles (B) in a rat colonic
preparation with the submucosa kept intact and
therefore with both the submuscular plexus (SMP)
and Auerbachs plexus (AP) intact. C: mechanical
recording showing the spontaneous contractile activity of the circular muscle in a colonic preparation
devoid of submucosa, but with AP intact. D: mechanical recording showing the spontaneous contractile
activity of the circular muscle in a colonic preparation
devoid of AP, but with the SMP intact.

waves (1518 cycles/min) 310 mV in amplitude can

only be recorded in a small percentage of the strips
(15). This study aimed to characterize the ICC from the
rat colon and to investigate their relation to the recorded spontaneous electrical and mechanical activity
of rat colonic muscle.
MATERIALS AND METHODS

Tissue Preparation
Male Sprague-Dawley rats (weighing 300350 g), fasted
overnight (18 h) but allowed ad libitum access to water, were
killed by decapitation and bled (this procedure was approved

G257

PACEMAKER ACTIVITY IN RAT COLON

Table 1. Characteristics of contractions observed in rat colonic muscle strips

Low Frequency
Muscle
Strip

Amplitude,
g

Frequency,
contractions/min

SMP AP
AP
SMP

1.81 0.19
2.13 0.25
NP

0.61 0.07
0.61 0.06
NP

SMP AP

2.06 0.46

0.78 0.17

High Frequency
Duration, s

Amplitude,
g

Frequency,
contractions/min

Duration,
s

0.12 0.03
NP
0.12 0.02

13.7 0.9
NP
12.8 0.8

4.9 0.2
NP
4.8 0.4

NP

NP

NP

Circular muscle
27.30 1.21
29.1 2.28
NP
Longitudinal muscle
40.28 2.42

Values are means SE. SMP, submuscular plexus; AP, Auerbachs plexus; NP, not present.

by the Ethics Committee of the Universitat Auto`noma de

Barcelona). The colon was then removed and placed in Krebs
solution consisting of (in mM) 10.10 glucose, 115.48 NaCl,
21.90 NaHCO3, 4.61 KCl, 1.14 NaH2PO4, 2.5 CaCl2, and 1.16
MgSO4 bubbled with a mixture of 5% CO2-95% O2 (pH 7.4).
Microelectrode and organ bath studies were carried out under nonadrenergic noncholinergic conditions (1 M of atropine, propranolol, and phentolamine). To perform these studies the following three preparations were used: full-thickness
strips; muscle strips without submucosa (circular and longitudinal muscle with AP); and muscle strips without the
longitudinal layer (circular muscle with SMP). In all cases,
muscle strips were 1 cm long and 0.3 cm wide.

procedure because it was impossible to penetrate the submucosa without damaging the microelectrode. In some experiments, we measured the electrical and mechanical activity of
the preparation simultaneously. In these experiments, one
end was pinned for intracellular recordings and the opposite
end was attached to an isometric transducer and preloaded
with 1 g.
Immunohistochemistry
Specimens from the colon were quick-frozen in isopentane
cooled in liquid nitrogen and stored at 80C until freeze

Spontaneous Mechanical Activity Recordings

Muscle strips were attached with silk threads to a stable
mount in the bottom of a 10-ml organ bath filled with carbogenated Krebs solution at 37 1C. The upper end was tied
to an isometric force transducer (Harvard UF-1) connected to
an amplifier and then to a computer. Data were digitized (25
Hz) and simultaneously displayed and collected using
Datawin2 software (Panlab-Barcelona) coupled to an ISC-16
A/D card installed in a personal computer with a Pentium
processor. A tension of 1 g was applied, and the tissue was
allowed to equilibrate for 1 h. After this period, strips displayed spontaneous phasic activity. To estimate the responses to drugs, the amplitude, duration, and frequency of
the contractions were measured before and after drug addition. We studied the activity of both the circular and the
longitudinal muscle layers by holding the muscle strips in
the direction of the circular and longitudinal fibers, respectively.
Resting Membrane Potential Recordings
Muscle strips were placed in a Sylgard-coated chamber
and continuously perfused with carbogenated Krebs solution
at 37 1C. In all cases, preparations were allowed to
equilibrate for 1 h before experiments started. Circular
muscle cells were impaled with glass microelectrodes (resistance 4060 M) filled with 3 M KCl. Membrane potential
was measured using a standard electrometer Duo773 (WPI).
Data were displayed on a digital storage oscilloscope 4026
(Racal-Dana) and simultaneously digitized (100 Hz) and collected using EGAA software coupled to an ISC-16 A/D card
(RC Electronics) installed in a 486 personal computer.
In the preparations without submucosa (circular muscle
with AP) the tissue was pinned with the circular muscle
facing up. In the preparations with submucosa (circular muscle without the longitudinal layer and full-thickness strips),
cells were impaled from the longitudinal side. We used this

Fig. 2. Mechanical recordings showing the effect of TTX (1 M; A)

and nifedipine (1 M; B) on the spontaneous cyclic activity displayed
by the circular muscle with preserved AP and SMP.

G258

PACEMAKER ACTIVITY IN RAT COLON

sectioning. The immunohistochemical techniques were as

described previously (19). To demonstrate the c-Kit receptor
polyclonal rabbit antibodies, c-kit (C-19, Santa Cruz Biotechnology) was used. Before incubation with the antibodies, all
sections were covered with a solution of either swine or goat
serum (1:5 or 1:10) for 20 min. Experimental sections were
incubated overnight with the specific antibody (1:500). Control sections were incubated with rabbit IgG (Dako X 903).
Immunoreactivity was demonstrated with the streptavidinbiotin (avidin-biotin complex) method or by the indirect fluorescence technique. To detect primary antibodies, the second and third layers were biotinylated swine anti-rabbit
(Dako E 353) and streptavidin-avidin-biotin complex-horseradish peroxidase (Dako K377) (incubation time, 1 h). For
secondary antibodies with the indirect fluorescence technique, we used goat anti-rabbit IgG (heavy and light chain)
conjugated with rhodamine (Jackson ImmunoResearch Laboratories). For light and fluorescence microscopy, we used a
Zeiss Axioplan 2. Both intact tissue and specimens in which
the submucosa or the longitudinal muscle had been fully or
partially removed were examined as described above.
Electron Microscopy
The colonic tissue was prepared by two methods. In
method A, colonic segments were cut and immersed in a
fixative solution containing 2.5% glutaraldehyde in 0.1 M
phosphate buffer, pH 7.1, at 4C for 4 h. After immersion,
pieces (1 2 mm) were cut, rinsed with 0.1 M phosphate
buffer containing 6% saccharose, and postfixed in 1% OsO4 in

Fig. 3. Microelectrode recording from

a full-thickness strip. A: overview of
the electrical activity showing the
presence of two rhythmic cyclic activities. B: detail of the low-frequency oscillation triggering muscular action potentials. C: detail of the slow wave
activity.

0.1 M phosphate buffer, pH 7.1 for 1.5 h. In method B, the

entire colon was isolated and immediately immersed in 2%
OsO4 in 0.1 M phosphate buffer, pH 7.1, for 1 min with
constant agitation, followed by addition of 45 vol aldehyde
fixative (2% glutaraldehyde, 2% formaldehyde, 0.1% picric
acid, and 0.1 M phosphate buffer, pH 7.3). After 23 h at
room temperature, the colon was rinsed with and transferred
to fresh aldehyde fixative of the same composition. Small
pieces of tissue (1 2 mm) were cut and postfixed in 2% OsO4
for 1 h. Tissue prepared by either method was dehydrated in
a graded series of alcohol, block stained in uranyl acetate (2%
in ethanol 70%) for 1 h at 4C, and embedded in Epon 812 R
(Merck). Thin sections were stained with toluidine blue for
light-microscopy investigation. Suitable areas were selected
for transmission electron microscopy, and ultrathin (5070
nm) sections were cut, mounted on copper grids, and contrasted with uranyl acetate and lead citrate. The grids were
examined in Philips 300 or 400 electron microscopes (Eindhoven, The Netherlands).
Tissue preparations from which the submucosa had been
fully or partially separated (n 3), as well as preparations
without the longitudinal muscle layer (n 4), were processed
by method B. Sections were examined by light and electron
microscopy. Particular care was exerted to obtain sections
that included the cleavage region.
Drugs
The following drugs were used: nifedipine and phentolamine (Sigma Chemical, St. Louis, MO), and TTX, atropine

PACEMAKER ACTIVITY IN RAT COLON

sulfate, and propranolol (Research Biochemicals International, Natick, MA). Stock solutions were prepared by dissolving drugs in distilled water, except for nifedipine, which
was dissolved in ethanol.
Data Analysis and Statistics
Data are expressed as means SE. A paired Students
t-test was used to compare mechanical activity in the absence
and presence of drugs. P 0.05 was considered to be statistically significant.
RESULTS

Spontaneous Mechanical Activity

The circular muscle of the rat colonic strips with the
submucosal layer kept intact showed spontaneous cyclic mechanical activity composed of two types of phasic
contractions (Fig. 1A): low-frequency and high-amplitude contractions with superimposed high-frequency
and small-amplitude contractions (n 9; Table 1). In
contrast, the longitudinal muscle in all the full-thickness preparations just showed the low-frequency and
high-amplitude contractions (n 4; Fig. 1B and Table
1). The circular muscle strips devoid of submucosa
showed the low-frequency and high-amplitude cyclic
contractions exclusively (n 30; Fig. 1C and Table 1).
The circular muscle with submucosa and without the
longitudinal muscle layer displayed the high-frequency
and low-amplitude contractions (n 5; Fig. 1D and
Table 1).
In the presence of 1 M TTX (n 9), the lowfrequency contractions increased in amplitude (1.81
0.19 vs. 2.72 0.47 g; P 0.05) and frequency (0.61
0.07 vs. 0.79 0.07 contractions/min; P 0.005)
whereas the duration was not affected. In contrast, the
high-frequency contractions were not modified in the
presence of this neural blocker (n 9) (Fig. 2A). In the
presence of 1 M nifedipine, both types of contraction
were abolished (n 20; Fig. 2B). Similar results for
TTX and nifedipine were obtained when both types of
activities were studied separately using the circular
muscle without AP or SMP, respectively (n 4 each).

G259

52.1 0.6 mV (n 6). The circular muscle cells

showed cyclic depolarizations (4.9 0.3 mV) of the
membrane potential that triggered a number of action
potentials at their peaks (17.4 3.2 action potentials/
depolarization). Each episode of depolarization with
associated action potentials correlated with a low-frequency and high-amplitude contraction and appeared
at a frequency of 0.78 0.11 episodes/min (Fig. 4A).
These cyclic depolarizations and the corresponding action potentials were abolished in the presence of 1 M
nifedipine (n 5; Fig. 4B).
Strips without longitudinal muscle. When the longitudinal muscle was removed, the slow-frequency cyclic
depolarization disappeared (n 5). Only slow waves at
a frequency of 14.2 1.93 cycles/min were recorded.
Figure 5 shows an example from three different recordings. Some slow waves carried one to three muscular
action potentials (Fig. 5, A and B). The resting membrane potential measured at the bottom of the slow
waves was 52 2 mV. The mean amplitude of the
slow waves was 5.3 0.3 mV, and the duration was
3.8 0.5 s. This electrical rhythm was similar to the

Spontaneous Electrical Activity

Full-thickness strips. In this preparation, the electrical activity consisted of two types of cyclic depolarizations (n 6; Fig. 3 A). Cyclic depolarizations (9.3 1.6
mV; 19 4s) at a frequency of 1.4 0.2 cycles/min
triggered several muscular action potentials (20 3)
(Fig. 3B). In between the cyclic depolarizations, slow
waves at a frequency of 15 1 cycles/min were recorded (Fig. 3C). These slow waves lasted 3.3 0.4 s
and showed an amplitude of 4.3 0.4 mV. The minimum resting membrane potential measured at the
bottom of the slow waves was 53 4 mV. This
electrical activity, with two putative pacemakers, was
related to the mechanical activity described before
when the circular muscle with AP and SMP was studied (Fig. 1A).
Strips without SMP. The resting membrane potential of circular muscle cells from the rat colon was

Fig. 4. A: intracellular recording showing simultaneously the spontaneous electrical and mechanical activity of the circular muscle of a
rat colonic preparation with the submucosa stripped away. B: microelectrode recording showing the effect of nifedipine (1 M) on the
spontaneous electrical activity displayed by the circular muscle from
a colonic muscle strip devoid of submucosa. RMP, resting membrane
potential.

G260

PACEMAKER ACTIVITY IN RAT COLON

Fig. 5. Slow wave activity in circular smooth

muscle with intact submucosal border from 3
different preparations (with different time
scales). Notice the presence of a few inhibitory
junction potentials in B.

mechanical rhythm of the circular muscle when the

longitudinal muscle was removed (Fig. 1D). When 1
M nifedipine (n 4) was added to the bath, a residual
cyclic activity was still recorded but action potentials
at the top of the slow waves were abolished.
Immunohistochemistry
In the rat colon, Kit immunoreactivity was present
in cells that enveloped the ganglia and fascicles of AP
and in cells at the border between the circular muscle
layer and the submucosa (n 3; Fig. 6A). Weaker
staining was observed in subserosal and ramified cells
running parallel with circular muscle cells (Fig. 6B).
Removal of submucosa resulted in abolition of Kit
immunoreactivity at the submucosal border of the
musculature (n 3; Fig. 6C).
Electron microscopy
ICC were identified throughout the colon at the sites
of the AP between the main muscle layers and the SMP

at the submucosal surface of the circular muscle (Figs.

7, 8, and 9). These ICC were referred to as ICC-AP and
ICC-SMP, respectively. Both populations of ICC were
arranged in cellular networks with branching processes in close contact with smooth muscle cells and
nerves. ICC have a cytoplasm rich in cell organelles,
including smooth and granular endoplasmic reticulum,
Golgi apparatus, and a large number of mitochondria
mainly in the branching processes. ICC displayed a
basal lamina, most prominent at the submuscular site,
and many caveolae in both sites. Bundles of thin filaments were prominent in all parts of the cytoplasm and
attachment plaques were seen, whereas thick filaments were not identified. In contrast, fibroblast-like
cells did not show close contact with muscle cells,
nerves, or ICC and were located close to collagen bundles. Although the fibroblasts also showed a differentiated cytoplasm, they generally had larger, dilated
cisternae of granular endoplasmic reticulum and
formed long, thin (in 3 dimensions: sheet-like) exten-

PACEMAKER ACTIVITY IN RAT COLON

G261

Fig. 6. Kit immunoreactivity in rat colonic musculature. A and B: sections of

the intact colonic wall. A positive reaction is present in interstitial cells of
Cajal (ICC) from 3 locations: between
the muscle layers (ICC-AP), along
nerve fascicles supplying the circular
muscle (ICC-CM), and at the submucosal border of the circular muscle (ICCSMP). C: section of the colonic wall
after partial separation of submucosa
and musculature, as indicated by the
dashed line (border region intact at
right). Note the absence of stained
ICC-SMP where submucosa was separated from the muscle (compare with
Fig. 10). Magnification: A, phase contrast, 140; B (enlargement of boxed
area in A), bright field, 250; and C,
bright field, 140.

sions. They were devoid of basal lamina and caveolae,

and filaments were less prominent (Fig. 8). In addition
to ICC-AP and ICC-SMP, some intramuscular ICC
were also identified (by similar criteria) among circular
fibers.
In those preparations where the submucosa had
been stripped off, no intact ICC could be identified at
the submucosal surface of the circular muscle. However, several unidentified dead cells, possibly ICC, with
a broken plasma membrane remained at the submucosal border of the circular muscle. Some ICC and dead
cells remained attached to the submucosa (Fig. 10).
Tissue strips, from which the longitudinal muscle
had been removed and in which the electrical activity
from the circular muscle had been recorded, were processed and sectioned. Examination by light and electron microscopy of the central region of the strips (the
region used for recording) confirmed the successful
removal of longitudinal muscle. A few ganglia still
adhered to the circular muscle. The majority of adhering interstitial cells were grossly damaged, and very
few ICC-AP could be recognized.

DISCUSSION

In the present study, we show that two pacemakers

are involved in the electrical activity of the rat colon.
When the full-thickness strip was studied, a low-frequency electrical rhythm that provokes muscular action potential and high-amplitude contractions is superimposed with slow waves corresponding to the lowamplitude mechanical activity. The interaction between
these two pacemakers determines the electrical and
mechanical activity of the smooth muscle. The electrical activity is correlated to the ICC distribution.
The pacemaker activity of the colon has been well
studied (6, 14, 26, 27) in the dog. In the canine colon,
the ICC network located near the SMP is responsible
for cyclic slow waves (56 slow waves/min) whereas
high-frequency cyclic depolarizations may originate
from the ICC network located at the site of AP (6, 14,
26, 27) . In small rodents, the pacemaker activity is not
so well characterized. In the mouse, cyclic depolarizations occur at a frequency similar to the one we found
in the rat colon (15, 16, 32). However, the origin of this

G262

PACEMAKER ACTIVITY IN RAT COLON

Fig. 7. Electron micrographs of region between the muscle layers.

LM, longitudinal muscle; CM, circular muscle. A: 2 interconnected
ICC-AP. B: detail of boxed area in A. The cytoplasm of the ICC-AP is
rather dense, but mitochondria and caveolae are seen. Note the
connection of ICC with a smooth muscle cell. Magnification: A,
6,900; B, 22,000.

cyclic activity is controversial. When the whole colon of

the mouse was studied (16), cyclic depolarizations
seemed to be neurogenic because TTX and hexamethonium abolished the cyclic activity. In contrast to these
results, the cyclic activity of the mouse colon was well
correlated to the development of ICC (32), suggesting
that interstitial cells might be the pacemaker of this
activity. In small strips from the circular muscle of the
rat colon, cyclic depolarizations insensitive to TTX
have been described (Ref. 20 and present study). These
results suggest that the pacemaker activity is not neurogenic when small strips are studied. Cyclic depolarizations can be recorded in strips without intact ICCSMP. This result shows that the ICC-SMP is not the
pacemaker of this cyclic activity but does not exclude
the ICC-AP network as the origin. In agreement with
this hypothesis, Ward and co-workers (32) suggested
that the pacemaker responsible in mouse colon for the
cyclic activity is the ICC network found near the my-

enteric plexus. Another possibility is that the cyclic

activity originated in the smooth muscle itself. However, previous data from our laboratory, using the
patch-clamp technique, show that isolated myocytes do
not oscillate cyclically (unpublished data). It is interesting to notice that the cyclic activity found in small
rodents is very similar to nonneurogenic cyclic depolarizations (24 cycles/min) found in the circular muscle of the human colon (21). In the latter study, these
cyclic depolarizations were described as slow waves
although it is not clear if the cyclic activity was sensitive to L-type Ca2 channel blockers. In the rat colon,
cyclic depolarizations are abolished in the presence of
nifedipine and consequently they are not classical slow
waves. In contrast to what we found in the rat, the
low-frequency cyclic depolarizations might originate
near the SMP in the human colon (21).
Another interesting finding was that the cyclic contractions we found in the circular muscle were also
observed in the longitudinal muscle. As we discovered
in the circular muscle, this cyclic mechanical activity
was increased by TTX and abolished by nifedipine.
This shows that this cyclic activity is not neurogenic.
An interesting hypothesis is that the cyclic activity
found in both the circular and the longitudinal muscle
might have the same origin. Accordingly, the ICC network between the two layers might be responsible for
the cyclic activity. In this case, coupling between the
ICC and both muscular layers should be expected. In
agreement with this hypothesis, MPO originating in
the ICC-AP network of canine colon spread into the
longitudinal and circular muscle (26). It is interesting
to notice that electromyographic recordings display
cyclic spike bursts that have the same frequency as the
cyclic mechanical activity found in either the circular
or longitudinal layer (9). According to our results, these
spike bursts originate near the myenteric plexus and
are probably related to action potentials from both
circular and longitudinal muscles that might contract
simultaneously.
In the dog colon, high-amplitude low-frequency (56
cycles/min) slow waves can be recorded in smooth muscle cells near the submucosal border. We were able to
record slow waves and cyclic contractions when we
used strips with an intact submucosal border region. In
contrast, when we used preparations devoid of intact
ICC-SMP this cyclic electrical and mechanical activity
was not recorded. These results suggest that the ICCSMP network is probably the pacemaker that generates slow waves and the corresponding mechanical
activity. It is important to notice that it is extremely
difficult to record slow waves in small rodents (15, 32).
This might well be a purely technical problem. Our
results demonstrate that it is important to preserve the
submucous plexus area to record slow waves. Obviously, it is very hard to impale circular muscle cells by
penetration of the submucosa because the microelectrode breaks. To avoid this problem, we impaled circular muscle cells in strips that had the longitudinal
muscle removed by fine dissection. Using this procedure, we exposed the circular muscle with the submu-

PACEMAKER ACTIVITY IN RAT COLON

G263

Fig. 8. Medium-power electron micrograph. ICC-AP (at right) could be identified and distinguished from fibroblasts (at left) by the presence in
ICC-AP of caveolae, a partial basal
lamina, a higher number of mitochondria, and fewer and nondilated endoplasmic reticulum cisternae. The cytoplasmic density of the ICC may vary
between preparations, depending on
fixation methods. Magnification, 23,000.

cosal border intact, and recordings displayed slow

waves and the corresponding mechanical activity.
Electrical slow waves were partially resistant to nifedipine, although the mechanical activity and action potentials triggered at the top of the slow waves were
nifedipine sensitive.
To confirm our hypothesis suggesting the presence of
two pacemakers, we tried to record both pacemakers
simultaneously. According to our previous arguments,
we attempted to impale circular muscle cells through
the longitudinal muscle, which is extremely thin in the
rat. With the electrode somewhat deep in the preparation we were able to record both pacemakers at the
same time. It is likely that under these conditions we
actually recorded from the circular muscle, because we
never found similar mechanical activity when the longitudinal muscle was studied. This, together with the
correspondence between the electrical and mechanical
activities, supports our hypothesis that slow waves
originate in the ICC at the submucosal border. The

slow waves spread into the circular muscle but do not

reach the longitudinal layer. Moreover, the slow-frequency cyclic oscillations producing high-amplitude
contractions probably originate in the AP and spread
into the circular and longitudinal layers.
Our results using microelectrodes and muscle bath
experiments are consistent with the distribution of
ICC. We were able to identify two networks of ICC
associated with AP and the SMP. This distribution is
very similar to those described in other species, including humans (8, 23), dogs (1, 2), and mice (7). The
ICC-SMP network was not present in the preparations
devoid of submucosa.
In conclusion, this study indicates that two pacemakers exist in the circular muscle of the rat colon.
Cyclic depolarizations occur at a low frequency, inducing cyclic contractions. Both cyclic depolarizations and
contractions are abolished in the presence of L-type
Ca2 channel blockers and consequently they are not
classical slow waves (the pacemaker input to drive

G264

PACEMAKER ACTIVITY IN RAT COLON

Fig. 9. Medium-power electron micrograph. ICC-SMP were identified by position and cytoplasmic detail: numerous mitochondria, many caveolae
(* indicates prominent areas), and a
distinct and complete basal lamina (arrows). Magnification, 16,000.

Fig. 10. Low-power electron micrograph. Before fixation, the submucosa

had been carefully separated from the
musculature up to the site marked by
the full line (the cleavage marked by
the dashed line). Note the disrupted
border layer of the musculature at left,
compared with the much less affected
cells at right, including intact ICCSMP (compare with Fig. 6C). Magnification: 1,800.

PACEMAKER ACTIVITY IN RAT COLON

slow waves in the intestine or colon is nifedipine insensitive). The ICC-SMP network is clearly not essential
for this cyclic activity, because it is present in strips
devoid of intact ICC-SMP. When the submuscular region is preserved and the longitudinal muscle removed,
slow waves and muscular contractions (1314/min),
can be recorded. These slow waves are partially nifedipine insensitive. Consistent with our electrophysiological and mechanical studies, we identified two ICC
networks, associated with the SMP and AP, using
electron microscopy and immunohistochemistry. We
suggest that these ICC constitute two networks of
pacemaker cells responsible for the observed cyclic
activities.
This coordination between two pacemakers is not all
that different from what has been described in the dog
(1, 2, 27) or cat (5): the ICC-SMP network is responsible for the 67/min slow waves and mechanical contractions as it has been described in these species.
Although the ICC-AP is responsible for high-frequency
contractions in the dog and cat, trains of MPO at 1- to
2-min intervals also exist in the cat (5). These trains of
MPO induce cyclic phasic contractions similar to those
we described in the rat. In this sense, MPO and cyclic
depolarizations described in this study might have the
same origin, spread to both circular and longitudinal
layers, and cause similar cyclic contractions. However,
two major questions about the origin of both cyclic
activities remain to be answered: 1) which input, probably coming from ICC-SMP, allows the slow wave generation and 2) which input, probably coming from ICCAP, allows the smooth muscle to reach the threshold to
open L-type Ca2 channels? Answer to these questions
should help us understand how the rat colon works
and, as rat colon has been extensively used as a model
of colitis, to pinpoint the origin of motility changes
found in such conditions.
This work was supported by Direccion General de Ensenanza
Superior e Investigacion Cientfica Grant PM-980171, Generalitat
de Catalunya: Grup de Recerca per lEstudi de la Motilitat Grant
SGR-9900272, and by a grant from the Universitat Auto`noma de
Barcelona (L. Pluja`).
REFERENCES
1. Berezin I, Huizinga JD, and Daniel EE. Interstitial cells of
Cajal in the canine colon: a special communication network at
the inner border of the circular muscle. J Comp Neurol 273:
4251, 1988.
2. Berezin I, Huizinga JD, and Daniel EE. Structural characterization of interstitial cells of Cajal in myenteric plexus and
muscle layers of canine colon. Can J Physiol Pharmacol 68:
14191431, 1990.
3. Cajal SR. Sur les ganglions et plexus nerveux de lintestin. C R
Soc Biol (Paris) 45: 217223, 1893.
4. Daniel EE and Berezin I. Interstitial cells of Cajal: are they
major players in control of gastrointestinal motility? J Gastrointest Motil 4: 124, 1992.
4a.Der-Silaphet T, Malysz J, Hagel S, Arsenault LA, and
Huizinga JD. Interstitial cells of Cajal direct normal propulsive
contractile activity in the mouse small intestine. Gastroenterology 114: 724736, 1998.
5. Du C and Conklin JL. Origin of slow waves in the isolated
proximal colon of the cat. J Auton Nerv Syst 28: 167178, 1989.
6. Durdle NG, Kingma YJ, Bowes KL, and Chambers MM.
Origin of slow waves in the canine colon. Gastroenterology 84:
375382, 1983.

G265

7. Faussone-Pellegrini MS. Comparative study of interstitial

cells of Cajal. Acta Anat (Basel) 130: 109126, 1987.
8. Faussone-Pellegrini MS, Pantalone D, and Cortesini C.
Smooth muscle cells, interstitial cells of Cajal and myenteric
plexus interrelationships in the human colon. Acta Anat (Basel)
139: 3144, 1990.
9. Ferre JP and Ruckebusch Y. Myoelectrical activity and propulsion in the large intestine of fed and fasted rats. J Physiol
(Lond) 362: 93106, 1985.
10. Hara H, Kubota M, and Szurszewski JH. Electrophysiology
of smooth muscle of the small intestine of some mammals.
J Physiol (Lond) 372: 501520, 1986.
11. Horowitz B, Ward SM, and Sanders KM. Cellular and molecular basis for electrical rhythmicity in gastrointestinal muscles. Annu Rev Physiol 61: 1943, 1999.
12. Huizinga JD, Thuneberg L, Kluppel M, Malysz J,
Mikkelsen HB, and Bernstein A. W/kit gene required for
intestinal pacemaker activity. Nature 373: 347349, 1995.
13. Jimenez M, Cayabyab FS, Vergara P, and Daniel EE.
Heterogeneity in electrical activity of the canine ileal circular
muscle: interaction of two pacemakers. Neurogastroenterol Motil
8: 339349, 1996.
14. Liu LWC, Thuneberg L, and Huizinga JD. Selective lesioning of interstitial cells of Cajal by methylene blue and light leads
to loss of slow waves. Am J Physiol Gastrointest Liver Physiol
266: G485G496, 1994.
15. Lyster DJK, Bywater RAR, and Taylor GS. Regulation of the
frequency of myoelectric complexes in the isolated mouse colon.
J Gastrointest Motil 5: 143151, 1993.
16. Lyster DJK, Bywater RAR, and Taylor GS. Neurogenic
control of myoelectric complexes in the mouse isolated colon.
Gastroenterology 108: 13711378, 1995.
17. Maeda H, Yamagata A, Nishikawa S, Yoshinaga K, Kobayashi S, Nishi K, and Nishikawa SI. Requirement of c-kit for
development of intestinal pacemaker system. Development 116:
369375, 1992.
18. Mikkelsen HB, Malysz J, Huizinga JD, and Thuneberg L.
Action potential generation, Kit receptor immunohistochemistry
and morphology of steel-Dickie (Sl/Sld) mutant small intestine.
Neurogastroenterol Motil 10: 1126, 1998.
19. Mikkelsen HB, Mirsky R, Jessen KR, and Thuneberg L.
Macrophage-like cells in muscularis externa of mouse small
intestine: immunohistochemical localization of Fa/80, M1/70 and
Ia-antigen. Cell Tissue Res 252: 301306, 1988.
20. Pluja` L, Fernandez E, and Jimenez M. Neural modulation of
the cyclic electrical and mechanical activity in the rat colonic
circular muscle: putative role of ATP and NO. Br J Pharmacol
126: 883892, 1999.
21. Rae MG, Fleming N, McGregor DB, Sanders KM, and Keef
KD. Control of motility patterns in the human colonic circular
muscle layer by pacemaker activity. J Physiol (Lond) 510: 309
320, 1998.
22. Rumessen JJ, Mikkelsen HB, and Thuneberg L. Ultrastructure of interstitial cells of Cajal associated with deep muscular
plexus of human small intestine. Gastroenterology 102: 5668,
1992.
23. Rumessen JJ, Peters S, and Thuneberg L. Light- and electron microscopic studies of interstitial cells of Cajal and muscle
cells at the submucosal border of human colon. Lab Invest 68:
481495, 1993.
24. Rumessen JJ and Thuneberg L. Interstitial cells of Cajal in
human small intestine. Gastroenterology 100: 14171431, 1991.
rdog T, Koh SD, Torihashi S, and Ward SM.
25. Sanders KM, O
Development and plasticity of interstitial cells of Cajal. Neurogastroenterol Motil 11: 311338, 1999.
26. Smith TK, Reed JB, and Sanders KM. Origin and propagation of electrical slow waves in circular muscle of canine proximal colon. Am J Physiol Cell Physiol 252: C215C224, 1987.
27. Smith TK, Reed JB, and Sanders KM. Interaction of two
electrical pacemakers in muscularis of canine proximal colon.
Am J Physiol Cell Physiol 252: C290C299, 1987.
28. Thuneberg L. Interstitial cells of Cajal: intestinal pacemaker
cells. Adv Anat Embryol Cell Biol 71: 1130, 1982.

G266

PACEMAKER ACTIVITY IN RAT COLON

28a.Thuneberg L and Peters S. Toward a concept of stretchcoupling in smooth muscle. I. Anatomy of intestinal segmentation and sleeve contractions. Anat Rec 262: 110124, 2001.
29. Thuneberg L, Rumessen JJ, Mikkelsen HB, Peters S, and
Jessen H. Structural aspects of interstitial cells of Cajal as
intestinal pacemaker cells. In: Pacemaker Activity and Intercellular Communication, edited by Huizinga JD. Boca Raton, FL:
CRC, 1995, p. 193222.
30. Torihashi S, Ward SM, Nishikawa SI, Nishi K, Kobayashi
S, and Sanders KM. c-kit-Dependent development of intersti-

tial cells and electrical activity in the murine gastrointestinal

tract. Cell Tissue Res 280: 97111, 1995.
31. Ward SM, Burns AJ, Torihashi S, and Sanders KM. Mutations of the proto-oncogene c-kit blocks development of interstitial cells and electrical rhythmicity in murine intestine.
J Physiol (Lond) 480: 9197, 1994.
32. Ward SM, Harney SC, Bayguinov JR, McLaren GJ, and
Sanders KM. Development of electrical rhythmicity in the murine gastrointestinal tract is specially encoded in the tunica
muscularis. J Physiol (Lond) 505: 241258, 1997.