The aim of this study was to characterize the pacemaker activity and inhibitory neurotransmission in the colon of Ws/Ws mutant rats, which harbor a mutation in the c-kit gene that affects development of interstitial cells of Cajal (ICC). In Ws/Ws rats, the density of KIT-positive cells was markedly reduced. Wild-type, but not Ws/Ws, rats showed low- and high-frequency cyclic depolarization that were associated with highly regular myogenic motor patterns at the same frequencies. In Ws/Ws rats, irregular patterns of action potentials triggered irregular muscle contractions occurring within a bandwidth of 10–20 cycles/min. Spontaneous activity of nitrergic nerves caused sustained inhibition of muscle activity in both wild-type (+/+) and Ws/Ws rats. Electrical field stimulation of enteric nerves, after blockade of cholinergic and adrenergic activity, elicited inhibition of mechanical activity and biphasic inhibitory junction potentials both in wild-type and Ws/Ws rats. Apamin-sensitive, likely purinergic, inhibitory innervation was not affected by loss of ICC. Variable presence of nitrergic innervation likely reflects the presence of direct nitrergic innervation to smooth muscle cells as well as indirect innervation via ICC. In summary, loss of ICC markedly affects pacemaker and motor activities of the rat colon. Inhibitory innervation is largely maintained but nitrergic innervation is reduced possibly related to the loss of ICC-mediated relaxation.

  • c-kit
  • slow waves
  • interstitial cells of Cajal
  • inhibitory junction potential
  • stereology
  • fast Fourier transform
  • colonic motility

interstitial cells of Cajal (ICC) are recognized as gut pacemaker cells and are hypothesized to be mediators of enteric innervation (8, 20, 59). In the colon of several species including rodents and humans, ICC are distributed at the level of Auerbach's plexus (ICC-AP) at the submuscular plexus (ICC-SMP), and within the musculature (ICC-IM). ICC-AP and ICC-SMP are multipolar cells with several branches, forming a network, whereas intramuscular ICC are spindle-shaped cells that run parallel to the muscle fibers (7, 16, 36, 40, 41, 49, 5153).

A role for ICC as pacemaker cells of the gut has been firmly established in the mouse small intestine at the tissue level (21, 32, 56), at the cellular level (24, 25, 48), and in vivo (11). In the small intestine, the main pacemaker area is the ICC network associated with Auerbach's plexus (AP). The characteristics of colonic pacemaker activity and its cellular origin are not fully understood, in part because of significant differences in pacemaker activity between species (19). In the canine and porcine colon, the major slow-wave activity is associated with the ICC-SMP network (24, 45), whereas the ICC-AP network has been associated with the generation of “myenteric potential oscillations” with a higher frequency (44). In the mouse and rat, ICC-SMP have been associated with the generation of high-frequency (HF) cyclic depolarizations (10–20 cpm) (36, 61), whereas in the rat colon low-frequency (LF) cyclic depolarizations have been associated with ICC-AP (36). The LF component shows a frequency gradient and presents the highest value (∼2.5 cpm) in the proximal colon; the HF component is quite constant along the colon (1). These cyclic depolarizations evoke action potentials in smooth muscle cells and combine to give rather complex colonic motor patterns. In humans, cyclic nifedipine-sensitive depolarizations have been described as well (38). The cellular basis of the different rhythmic activities of the colon has not yet been sufficiently characterized.

Cajal (39) identified ICC as being an integral part of neurotransmission (39), and ultrastructural data appear to support this (9). Interstitial cells of Cajal are intercalated between nerves and smooth muscle cells and might mediate nerve responses (9). Physiological evidence that ICC-IM participate in neurotransmission is based on studies of the stomach of W/Wv mice, which lack ICC-IM (5). ICC are thought to be essential for inhibitory neurotransmission, in particular nitrergic neurotransmission. In the gastric fundus of Sl/Sld mutant mice, ICC-IM were absent and the inhibitory neurotransmission and associated relaxations were shown to be impaired (4). A similar conclusion was also reached for the lower esophageal and pyloric sphincters (57). This concept was challenged by Goyal and coworkers (43), who showed that, in vivo, swallow-induced nitrergic inhibition of the lower esophageal sphincter was normally present in W/Wv mice. Recently two independent studies (10, 47) demonstrated that nitrergic neuromuscular relaxation is functional in the internal anal sphincter despite the absence of KIT-positive ICC. In the colon of rodents, ultrastructural studies show that ICC-IM are in close contact with nerves and smooth muscle cells. These contacts give morphological support for the hypothesis that ICC participate in neurotransmission in the colon (55). Electrophysiologically, in the rat colon, the inhibitory junction potential (IJP) consists of a fast component (apamin-sensitive) and a slow component [Nω-nitro-l-arginine (l-NNA)-sensitive]. ATP might be responsible for the fast component, and nitric oxide mediates the sustained component (37). The potential role of ICC in the generation of the colonic IJP has not been investigated. Therefore, the aim of the present work was to study the distribution of the KIT receptor and to characterize the pacemaker activity and inhibitory neurotransmission in the colon of Ws/Ws and wild-type rats (+/+). Preliminary data from this work were presented at the Digestive Disease Week meeting of the American Gastroenterology Association, New Orleans, Louisiana, 2004 (2).


Animals and tissue preparation.

Male Ws/Ws and sibling wild-type rats +/+ (Sumitomo Mitsui Banking), 7 wk old, were used in the present study. Rats were kept individually and fasted for 16–18 h with ad libitum access to water. Animals were stunned before being decapitated and bled. The entire colon was removed from 1 cm below the ileocecal junction to the pelvic brim. To perform mechanical and electrophysiological studies the colon was placed in oxygenated Krebs solution (n = 6). For immunohistochemistry studies the colon was placed in PBS with nifedipine (1 μM) for 15 min to ensure relaxation (n = 4). In both types of experiments, the colon was opened along the mesenteric border and pinned to a Sylgard base with the mucosa facing upward. The colon was divided into three different parts: proximal, mid, and distal, as previously described (1). The methods of housing and handling of the animals were approved by The Ethics Committee of the Universitat Autònoma de Barcelona.


For morphological experiments, the proximal, mid, and distal segments of the colon were cut into three pieces; one piece was frozen, and the second and third were prepared for whole-mount immunohistochemistry using c-KIT antibody. The mucosa was removed by sharp dissection in all preparations. Whole-mount preparations were used to study the morphology and area density (number of cells/mm2: see Stereological analysis) of c-KIT-positive cells. Frozen sections were used to assess the distribution of KIT-positive cells throughout the thickness of the colon (1). In short, whole mounts or frozen sections for KIT staining were fixed in Zamboni's fixative. Primary antibody was rabbit anti-KIT 1:500 (Santa Cruz, Sc 168, CA). Secondary antibody was biotin-conjugated donkey anti-rabbit F(ab)2 (Jackson, ME) 1:2,000. All the antibodies were diluted in 1% human serum albumin in PBS + 0.3% Triton X-100. Negative controls included the omission of primary or secondary antibodies or preincubation of the primary antibody with the corresponding peptides.

Stereological analysis.

An established stereological technique, the fractionator technique (15, 26, 33), was used to count the number of KIT-positive cells in the proximal, mid, and distal colon of both Ws/Ws and wild-type rats. The counting was performed on systematic random fields of vision by moving an unbiased counting frame (14) through the full thickness of the whole-mount specimen. The stereological analysis of KIT-positive cells was carried out on a computer monitor by using computer-assisted interactive stereological test systems (The CAST-grid software, Olympus Denmark). KIT-positive cells present were classified and counted into two groups: 1) the KIT-positive cells at the level of the AP; 2) the KIT-positive cells adjacent to the submuscular plexus (SMP) (1). The stereological analyses were performed by a blinded procedure; the investigator performing the quantification did not know the origin of the material.

Motility studies.

Circular muscle strips (full thickness) were obtained from proximal, mid, and distal parts of the colon and were cut ∼1 cm long and 0.3 cm wide. Preparations were mounted under 1-g tension with a 2-0 silk thread in a 10-ml muscle bath containing oxygenated Krebs solution (95% O2-5% CO2) maintained at 37 ± 1°C. The strips were tied to an isometric force transducer (Harvard Apparatus, Holliston, MA) connected to a personal computer through an amplifier. Data were digitized (25 samples/s) and displayed with Data 2001 software (Panlab, Barcelona, Spain). Preparations were allowed to equilibrate for 1 h. Muscle bath experiments were performed to study 1) the pattern of mechanical activity, 2) the inhibitory neuromuscular transmission, and 3) the presence of an inhibitory neural tone. These experiments were performed both on +/+ and Ws/Ws animals in each segment (proximal, mid, and distal) of the colon. To study the pattern of mechanical activity, strips were studied in normal Krebs solution (n = 6); under nonadrenergic, noncholinergic (NANC) conditions by adding atropine, propranolol, and phentolamine (each 1 μM) to the Krebs (n = 6); and in the presence of TTX (1 μM) (n = 4). Frequency analyses of contractile activity were carried out by use of the fast Fourier transform, the short-time Fourier transform (STFT, Gabor transform), and the coherence function using a software package written by one of the authors (12). The ensemble average power spectrum was obtained for each signal. Linear trends were removed from each data segment, and the spectral coefficients of the power spectra were smoothed by using a Gaussian function to reduce their variance (12). Detailed analyses of lower frequencies were assessed following subsampling of raw data to 0.83 Hz. Changes of frequency content over the time were calculated by the STFT algorithm for 1,024 data windows to generate narrow-banded spectrograms.

To study the effects of release of inhibitory neurotransmitters on mechanical activity, circular muscle strips were placed in a muscle bath under NANC conditions and were stimulated by electrical field stimulation (EFS): at 28 and 40 V, 4 Hz, 0.3 ms pulse duration, 2–3 min stimulus duration through two platinum electrodes (n = 6). To study the presence of an inhibitory neural tone, the amplitude of contractions was measured in Krebs under NANC conditions and in the presence of TTX (1 μM) (n = 4), l-NNA (1 mM) (n = 5), and the small conductance calcium-activated K+ channel blocker apamin (1 μM) (n = 5). The amplitude of contractions was measured before and after drug addition.

Microelectrode studies.

Strips from the midcolon were dissected with fine forceps by sharp dissection under a magnifying glass to remove the mucosa. These strips with both plexuses kept intact were obtained from Ws/Ws and +/+ rats (n = 6). The tissue was pinned with the serosa facing upward in a Sylgard-coated chamber and was continuously perfused with oxygenated Krebs solution (95% O2-5% CO2) at 37 ± 1°C. Strips were allowed to equilibrate for ∼1 h. Electrical recordings were obtained after impaling circular muscle cells with glass microelectrodes (40–60 MΩ of resistance) filled with 3 M KCl. Data were registered using a standard electrometer Duo 773 (WPI), an oscilloscope 4026 (Racal-Dana) and simultaneously digitized (100 Hz) and collected using EGAA software coupled to an ISC-16 analog-to-digital card (RC Electronics, Santa Barbara, CA). To study the pattern of electrical activity, the resting membrane potential and the amplitude and frequency of slow waves and cyclic depolarizations were determined. To evaluate the effect of release of inhibitory neurotransmitters, EFS was performed: total duration 100 ms, frequency 20 Hz, pulse duration 0.3 ms, and increasing voltage strengths (5, 10, 12, 15, 17, 20, and 25 V). IJPs elicited by EFS were recorded in Krebs solution (n = 6), in NANC conditions (n = 6), and in presence of nifedipine (1 μM) (n = 6). To characterize the IJP, the amplitude and duration of the transient hyperpolarization were measured (both in Ws/Ws and +/+) and when l-NNA (1 mM) and apamin (1 μM) were consecutively added to the recording chamber. The duration of the IJP was measured at the level of the resting membrane potential.

Solutions and drugs.

The composition of Krebs solution was (in mM): 10.10 glucose, 115.48 NaCl, 21.90 NaHCO3, 4.61 KCl, 1.14 NaH2PO4, 2.50 CaCl2, and 1.16 MgSO4 (pH 7.3–7.4). The solution was bubbled with carbogen (95% O2-5% CO2). The following drugs were used: phentolamine and l-NNA (Sigma Chemical, St. Louis, MO), atropine sulfate (Merck, Darmstadt, Germany), TTX and apamin (Latoxan, Valence, France), propranolol (Tocris, Tocris Cookson, Bristol, UK). Stock solutions were prepared by dissolving drugs in distilled water, except TTX, which was diluted in 1% glacial acetic acid.


Data are expressed as means ± SE. A P value <0.05 was considered to indicate statistically significant differences. Statistics were performed with GraphPad Prism v.3.0 (San Diego, CA) software. Differences in the amplitude or duration of the IJPs before and after drug infusion were compared by ANOVA for repeated measurements (two-way ANOVA) followed by Bonferroni posttest. A paired Student's t-test or one-way ANOVA was used to compare mechanical activity in the absence and presence of drugs. To test for regional differences of the three parts (proximal, mid, and distal) in morphological studies, a one-way ANOVA was used within each group of animals. Post hoc comparisons were performed with a paired t-test given that significant difference was found with the ANOVA. To avoid “mass significance,” the null hypothesis was rejected when 3P ≤ 0.05. An unpaired t-test was used to test regional differences between +/+ and Ws/Ws animals.


Immunohistochemistry and stereology.

Using whole mount immunohistochemistry, KIT-positive cells were found in the colon of wild-type rats at the level of AP, the SMP, and intramuscular (IM). At the level of AP, KIT-positive cells were multipolar with long branching processes, which formed a network (Fig. 1) and sometimes extended into the circular and longitudinal muscle layers. The processes at the circular muscle side seemed to be in contact with cells in the circular muscle layer as previously shown (84). Intramuscular KIT-positive cells were spindle shaped and ran parallel to the muscle fibers. At the level of the SMP, ramified KIT-positive cells formed a network of cells (not shown). In the colon of Ws/Ws rats, a reduced number of KIT-positive cells was observed in the AP region and at the level of the SMP and within the muscle layers (Fig. 1). To quantify the number of ICC at different levels of the colon in both wild-type and Ws/Ws rats, a stereological analysis was performed using random fields and a blinded methodology (see materials and methods). In wild-type rats, ANOVA analysis did not indicate any difference in the density between proximal, mid, and distal colon at the level of AP and SMP (Fig. 2). However, in mutant rats, ANOVA analysis indicated differences between the proximal colon vs. mid and distal colon at AP (Fig. 2). In Ws/Ws rats, the loss of KIT-positive cells amounted to more than 90% (all ICC subtypes: P < 0.01) at the mid and distal parts of the colon (Fig. 2). In the proximal colon the loss of KIT-positive cells was 99% (P < 0.001) at the submuscular plexus and 52% at the AP region (P < 0.01). Frozen sections on cross section were used to confirm this distribution of KIT-positive cells throughout the thickness of the colon in both Ws/Ws and +/+ rats (Fig. 3). ICC were virtually absent at the level of SMP and within muscle layers.

Fig. 1.

Whole-mount preparations showing KIT receptor-positive cells in the proximal, mid, and distal colon from wild-type +/+ (top) and Ws/Ws (bottom) rats. Arrowheads point to interstitial cells of Cajal (ICC) within the musculature (ICC-IM; out of focus) near the Auerbach's plexus (AP) in the circular muscle layer. Scale bar = 50 μm (all panels). Notice that in F the picture was taken from an area where some residual ICC were present. Randomized quantitative data are expressed in Fig. 2.

Fig. 2.

Areal densities (cells/mm2) of KIT receptor-positive cells in the colon of wild-type +/+ (left) and Ws/Ws (right) rats. Horizontal lines show group means. Values show area density means of KIT receptor-positive cells (means: number of cells/mm2 surface area). Coefficients of variation (CV = SD/group mean) are shown in parentheses. ANOVA: differences between proximal, mid, and distal colon within each segment.

Fig. 3.

Frozen sections from the proximal colon of control +/+ (left) and Ws/Ws (right) rats. Notice the presence of KIT receptor-positive cells at the level of the AP [between circular (CM) and longitudinal muscle (LM) layers], at the submuscular (SM) border, and within the musculature in +/+ rats. Mutant rats only showed KIT receptor-positive cells at the AP region. Scale bar = 50 μm.

Patterns of spontaneous mechanical activity in the rat colon.

Circular muscle strips from colon from wild-type rats showed a regular pattern of spontaneous mechanical activity characterized by LF and HF contractions in proximal, mid, and distal colon (Fig. 4A). In circular muscle strips from wild-type colon, both LF and HF contractions were of higher amplitude in the proximal colon compared with those from the mid and distal colon (Table 1). LF contractions had a higher frequency in proximal segments whereas the frequency of HF contractions was constant along the colon. In Ws/Ws animals, we could not establish a regular pattern of spontaneous contractions. In the majority of recordings, irregular contractions characterized by low amplitude were observed in Krebs solution (Fig. 4B). In the presence of TTX (n = 4), the same patterns were maintained in both +/+ and Ws/Ws rats (Fig. 5), suggesting a nonneural origin of these contraction types, as recorded in nonstimulating conditions.

Fig. 4.

Muscle strip recordings showing the spontaneous cyclic mechanical activity displayed by the circular muscle layers from the proximal, mid, and distal colon of +/+ (A) and Ws/Ws (B) rats. Tissue was mounted with preserved AP and submuscular plexus (SMP) and incubated in Krebs solution.

Fig. 5.

Mechanical recordings showing the spontaneous cyclic mechanical activity displayed by circular muscle strips from wild-type and Ws/Ws rats in the presence of the neural Na+ channel blocker TTX (1 μM). Colonic preparations were taken from the mid part of the colon with preserved AP and SMP. Notice the presence of myogenic high-frequency and low-frequency contractions in wild-type animals and the presence of irregular contractions in Ws/Ws rats (left). Right: expansion of segment A.

View this table:
Table 1.

Motility patterns observed in colonic muscle strips from +/+ rats, with both plexuses intact in Krebs solution

To investigate the characteristics of mechanical activity from +/+ and Ws/Ws rats, we performed analyses in the frequency domain using Fourier methods. Application of the fast Fourier transform algorithm to raw and subsampled signals revealed the existence of discrete harmonics in both +/+ and Ws/Ws rats (Fig. 6). Wild-type (+/+) rats always exhibited high-energy LF components (1.33–0.68 cpm) together with lower energy HF components in the bandwidth of 10–17 cpm. When compared along the colon, LF components but not HF components significantly decreased in frequency distally (Table 1). Analyses using the STFT revealed the independence of low and high frequencies, i.e., whereas HF components are maintained over time, LF are episodic and are superimposed on HF components (Fig. 6A). On the other hand, signals from Ws/Ws rats featured prominent discrete high-energy HF components (proximal: 13.92 ± 0.39 cpm; mid: 15.62 ± 0.92 cpm and distal: 16.02 ± 0.38 cpm, n = 6) superimposed with other more irregular components (Fig. 6B). Although the HF component was maintained over time, different components with different frequencies were transiently observed (shown for distal colon in Fig. 6B). LF components in Ws/Ws rats either were absent or displayed considerably lower energy than HF components in all colonic segments studied. Finally, although HF components in Ws/Ws rats tended to have an increased frequency (and energy) in all segments compared with +/+, coherence analyses of power spectral densities revealed that they fell in the same bandwidth as +/+ rats (not shown). The HF contractions did not display the regularity observed in the wild-type tissues in most muscle strips (Fig. 5).

Fig. 6.

Frequency analysis on spontaneous contractile activity from distal colon in wild-type +/+ and Ws/Ws rats. In each panel an example is shown of spontaneous mechanical activity (A), a power spectral density by fast Fourier transform (FFT; B) and a time-frequency spectrogram obtained by application of the short-time Fourier transform (STFT) to the isometric signal (C). Note the different scales of the y-axes (B) and gray scales (C).

Patterns of electrical activity in the midcolon.

Intracellular electrical activity of circular smooth muscle cells from wild-type colons displayed a resting membrane potential of −53.5 ± 1.5 mV. The electrical recordings were dominated by cyclic depolarizations at 1.2 ± 0.1 cpm (n = 6) with marked superimposed action potentials with an amplitude of 24.0 ± 1.3 mV (Fig. 6). In addition, slow waves were recorded at 19.6 ± 1.4 cpm (n = 6) with an amplitude of 3.5 ± 0.3 mV. The resting membrane potential of the circular muscle cells from Ws/Ws rats was −49.3 ± 3.2 mV (n = 6) (not significantly different from wild type). In five of the six mutant animals, the regular pattern observed in +/+ rats was absent, and the cyclic depolarizations with superimposed action potentials were not observed; action potentials occurred either continuously or in irregular bursts. In one preparation, however, regular cyclic depolarizations occurred. Action potentials occurred at amplitudes of 25.7 ± 2.6 mV (n = 6) (Fig. 7), not significantly different from wild-type tissue. The HF slow-wave activity was not evident in Ws/Ws tissues.

Fig. 7.

Intracellular microelectrode recordings showing the spontaneous electrical activity displayed by circular muscle strips from the midcolon with both plexuses kept intact in +/+ (left) and Ws/Ws (right). In the colonic musculature of wild-type rats, the pattern of low-frequency, slow depolarizations with superimposed action potentials is dominant. Each graph is from a different animal (n = 6).

Effect of stimulation of inhibitory neurons on the spontaneous activity.

In the presence of atropine, phentolamine, and propranolol, nerve stimulation caused inhibition of the spontaneous mechanical activity in both wild-type (n = 6) and Ws/Ws (n = 6) rats (Figs. 8 and 9). This inhibition was completely prevented by prior addition of TTX (1 μM) (Fig. 8) or by prior addition of a combination of l-NNA (1 mM) and apamin (1 μM) (Fig. 9, n = 5). When measuring intracellular electrical activity, nerve stimulation caused an IJP in both +/+ and Ws/Ws animals (Figs. 10 and 11, n = 6). When the stimulus strength was gradually increased, the amplitude and duration of the IJP progressively increased. Figure 10C shows the relationship between the stimulus strength (voltage) and the amplitude (mV) and duration (s) of the IJPs in Krebs solution, NANC, and with nifedipine (1 μM). In Krebs and under NANC conditions, an off response with spiking activity was recorded after the IJP, in both wild-type and mutant animals. When nifedipine was added to the chamber, the spiking activity was abolished. In +/+ animals, the IJP clearly showed two components: an initial fast hyperpolarization and a more sustained hyperpolarization (Figs. 10A and 11). There was no significant difference in the amplitude of IJPs between +/+ and WsWs rats. The IJP duration, however, measured at the base of the IJP, was smaller in the colon of WsWs animals (Fig. 10C). In Ws/Ws animals, in most impalements, an apamin-sensitive fast component was followed by a sustained hyperpolarization sensitive to l-NNA (Fig. 11); however, in ∼33% of recordings only a fast component sensitive to apamin (Fig. 11) was observed. l-NNA (1 mM) caused a reduction (P < 0.05) in the duration of the IJP in both +/+ and WsWs rats showing a functional nitrergic neurotransmission in both groups (Fig. 11).

Fig. 8.

Mechanical recordings showing the effect of activation of enteric neurons through electrical field stimulation (EFS, 28–40 V, 4 Hz, 0.3 ms, 2–3 min) on the spontaneous cyclic mechanical activity displayed by the circular muscle in midcolon preparations. Recordings were obtained from nonadrenergic, noncholinergic (NANC) conditions (left) and in the presence of TTX (1 μM) (right), in both +/+ rats (top) and Ws/Ws rats (bottom). Sodium nitroprusside (NaNP 10 μM) was added at the end to check muscle relaxation.

Fig. 9.

Mechanical recordings showing the effect of EFS (28–40 V, 4 Hz, 0.3 ms, 2–3 min) on the spontaneous cyclic mechanical activity displayed by midcolon circular muscle strips. Recordings obtained from NANC conditions (left) and in the presence of a combination of Nω-nitro-l-arginine (l-NNA; 1 mM) and apamin (1 μM) (right), in both +/+ (top) and Ws/Ws (bottom) rats.

Fig. 10.

Intracellular recordings showing inhibitory junction potentials (IJPs) elicited by EFS with increasing strengths (5, 10, 12, 15, 17, 20, and 25 V) in +/+ (A) and Ws/Ws (B) rats, under NANC conditions and in the presence of nifedipine (1 μM). C: graphs representing the amplitude (left) and duration (right) of the IJPs elicited by EFS in circular muscle strips in Krebs solution (top), under NANC conditions (middle) and in the presence of nifedipine (1 μM) (bottom), in both +/+ (•) and Ws/Ws rats (○). Notice that mutant animals present a decrease in the duration (not the amplitude) of the IJP (ANOVA P < 0.05) in Krebs, in NANC conditions, and in presence of nifedipine (1 μM).

Fig. 11.

Effect of l-NNA and apamin on the IJPs of wild-type and Ws/Ws colon. A: intracellular recordings showing IJPs under NANC conditions (wild type) elicited by EFS in +/+ and Ws/Ws rats. Increased voltage strength of stimulation (5, 10, 12, 15, 17, 20, and 25 V) in +/+ rats under NANC conditions and in the presence of l-NNA (1 mM). Effect of apamin (1 μM) on the IJPs elicited by a repetitive stimulus of 17 V. B: IJPs elicited by EFS in Ws/Ws animals. Notice that the IJP at left has a prominent fast but no sustained nitrergic component. In contrast, the IJP at right has a fast followed by a sustained, l-NNA-sensitive component, representing 4 of the 6 animals studied. In both cases, the fast component is sensitive to apamin. C: graphs representing the effect of l-NNA (1 mM) and a combination of l-NNA (1 mM) and apamin (1 μM) on the amplitude and duration of IJPs from +/+ (left) and Ws/Ws (right) rats. All recordings were done in presence of nifedipine (1 μM).

The addition of TTX (1 μM) in muscle bath preparations increased the amplitude of spontaneous contractions in both wild-type and mutant rats (Fig. 12), indicating the presence of an inhibitory tone in both groups of animals. The addition of l-NNA (1 mM) increased the amplitude of spontaneous contractions (Fig. 12), indicating marked presence of spontaneous release of nitric oxide. When apamin was added in the presence of l-NNA, a significant increase in contractile activity occurred (not shown), indicating that colonic muscle preparations from +/+ and Ws/Ws rats have intrinsic spontaneous activity of nitrergic and purinergic nerves.

Fig. 12.

A: muscle strip recordings under NANC conditions from midcolon circular muscle layers showing the effects of TTX (1 μM; top) and l-NNA (1 mM; bottom) on the spontaneous mechanical activity in wild-type +/+ and Ws/Ws rats. B: graphs representing the differences in the amplitude of spontaneous contractions before and after the addition of TTX and l-NNA to the muscle bath.


The pacemaker activity and the inhibitory neurotransmission were characterized in the colon of Ws/Ws rats and compared with wild-type animals. The colon of Ws/Ws rats showed a marked change in electrical activity of the musculature, associated abnormal motility patterns, a strong reduction in KIT-positive cells, and reduced inhibitory neurotransmission.

A stereological analysis was conducted to measure the density of KIT-positive cells. Stereology has been widely used to quantify the loss of neurons in the brain in several conditions, such as in aging (30). This methodology has been previously used to quantify the distribution of KIT-positive cells in the colon of Sprague-Dawley rats (1). In the colon of +/+ rats, KIT-positive cells were found at the level of AP (ICC-AP), the submucosal border (ICC-SMP), and within both muscle layers (ICC-IM). ICC-AP and ICC-SMP were multipolar cells with several branches, forming a network, whereas intramuscular ICC were spindle-shaped cells and ran parallel to the muscle fibers. We have previously reported a similar distribution of KIT-positive cells in Sprague-Dawley rats (1). In contrast, in Ws/Ws rats a marked reduction in KIT-positive cells was found but it was not the same for the different subtypes of ICC. At the level of AP, a reduction between 52% (proximal colon) and 94% (distal colon) was found. An absence of KIT immunoreactivity was observed at the level of the SMP (ICC-SMP) and within the muscle layers (ICC-IM). Absence of KIT-positive cells has been reported at the level of AP in the small intestine of mutant rodents such as Ws/Ws rats (22), W/Wv mice (32), and Sl/Sld mice (34). In the fundus of W/Wv and Sl/Sld mice, ICC-IM are missing (4, 5).

In vitro studies on the colon of the rat show regular motor patterns and well-defined associated electrical activities as recently described in Sprague-Dawley rats (1, 36). Identical activities were observed in the present study in wild-type rats. The rat colon displays slow depolarizations with superimposed action potentials associated with LF contractions of high amplitude. This activity is most closely associated with the presence of ICC-AP (1, 36). In addition, regular slow-wave activity at a faster frequency is associated with cyclic lower amplitude contraction at the same frequency. This HF component is most closely associated with ICC-SMP (1, 36). These two frequency components behave independently as judged from the analyses using STFT. In Ws/Ws rats the electrical and mechanical patterns were markedly impaired. The dominant patterns were that of irregular depolarizations and irregular contractions. In one Ws/Ws rat, slow-frequency depolarizations were observed, likely related to the persistent presence of residual ICC-AP. The Ws/Ws rats also lacked the regular slow-wave activity in the HF domain. This is likely associated with the complete lack of ICC-SMP. The Ws/Ws rats did show marked contractile activity in a frequency range that was wider but overlapped with that of the HF component. The HF contractile activity is likely associated with action potential generation of smooth muscle cells, which was shown to be unaffected in the present study. Smooth muscle cells devoid of ICC do not necessarily generate action potential and contractile activity in a random fashion. This is consistent with observations in the dog colon, where, after removal of pacemaker ICC, contractile activity could be induced in the same frequency range as ICC-dominated contractions (28). The present study is in excellent agreement with a recent paper that showed colonic motility in Ws/Ws rats in vivo (46). In this study, wild-type rats showed high-amplitude, regular contractions that reached a maximal frequency of 1.4 contractions/min in the presence of blockade of nitric oxide synthase, similar to the maximal frequency of LF contractions observed in the present study in vitro. This motor pattern was absent in Ws/Ws rats. The present study gives structural and functional support to the hypothesis that loss of ICC is associated with impairment of colonic motor function in vivo (46).

The ionic basis of pacemaker activity in the colon was first explored in the canine colon, where the slow wave of the circular muscle layer is strong and omnipresent. The upstroke of the slow-wave activity in the circular muscle of the canine colon persists in the presence of L-type calcium channel blockers although it is dependent on extracellular calcium (3, 58). In the human colon studied in vitro, slow-wave activity has different components and all contractile activity and action potentials were abolished by L-type calcium channel blockers (6, 38). The HF slow-wave component is also sensitive to L-type calcium channel blockers (6). The LF component in the human colon still needs to be examined, but unpublished data from our laboratory show that nifedipine abolished all slow-wave activity. The present study shows that, in the rat colon, slow waves and cyclic depolarizations as well as action potentials are sensitive to nifedipine and depend on ICC input. It is interesting to note that despite the presence of residual ICC (mainly at the AP level) an altered motor function is present both in vitro (present study) and in vivo (46). Decreased density or volume of ICC has been described in patients with slow-transit constipation (27, 31, 50). Further studies may reveal that the rat colon is a good model for understanding the human colonic motor function, and the Ws/Ws colon may be a good model for further study of the functional consequences of loss of ICC.

The inhibitory neurotransmission was evaluated on mechanical and electrical activities of the circular muscle cells. Activation of enteric neurons through EFS caused inhibition of the spontaneous mechanical activity and elicited IJPs. We have previously shown in Sprague-Dawley rats that the inhibitory mediators released by enteric motor neurons are probably ATP and nitric oxide (37). In these animals, the IJP shows two components: a fast component followed by a sustained hyperpolarization. The fast component of the IJP is apamin sensitive and is probably mediated by ATP whereas the sustained component is apamin insensitive but l-NNA sensitive and it is probably due to nitric oxide release (37). A biphasic IJP has been described in the colon of several species including mouse (42), guinea pig (17), and humans (23). In the present paper, we report a similar result in wild-type rats, suggesting that the inhibitory neurotransmitters in these animals are also ATP and nitric oxide.

The colon of Ws/Ws rats displayed inhibitory junction potentials and mechanical relaxations induced by EFS. The IJP amplitudes of Ws/Ws and wild-type rats, associated with the fast component and dominated by the apamin-sensitive neurotransmitter, were not significantly different. Hence this part of the inhibitory neurotransmission is not mediated by ICC-IM as observed in the stomach of W/Wv mice (18). Assuming that the fast IJP is mediated by ATP (13, 37, 60), the present study shows that purinergic neurotransmission is likely independent of KIT-positive ICC-IM. The slow, nitrergic component of the IJP was present in the Ws/Ws rats despite the absence of ICC-IM, suggesting a direct communication between nitrergic inhibitory motor neurons and smooth muscle. Consistently, when NOS was blocked in the Ws colon, an increase in the mechanical activity was found, indicating the presence of ongoing neural release of nitric oxide directly affecting smooth muscle cells. This was an unexpected result and not consistent with the hypothesis that ICC-IM are essential for nitrergic innervation. In other areas of the gastrointestinal tract, in W/Wv mice, the absence of ICC-IM appears to cause a virtual absence of inhibitory neurotransmission (4, 5, 7), although nitrergic relaxation of the internal anal sphincter was noted in W/Wv mice (10, 47) and nitrergic relaxation was observed in vivo in the lower esophageal sphincter of W/Wv mice (43). In some muscle preparations from the Ws/Ws colon, a markedly reduced slow component of the IJP was observed. This may be related to a reduced number of nitrergic nerves that was observed in preliminary observations (2). It may also reflect the existence of direct and indirect (via ICC) nitrergic innervation (35). ICC are directly innervated by nitrergic neurons (54). This may primarily affect ICC function but will influence smooth muscle cells as well. Loss of the indirect component of innervation may contribute to the reduction of the slow component of the IJP.

In conclusion, the strong reduction in c-kit-positive ICC caused impairment of pacemaker activity and associated motor activity in the colon of Ws/Ws rats. Apamin-sensitive inhibitory innervation was not affected by loss of ICC. Variable presence of nitrergic innervation likely reflects the presence of direct nitrergic innervation to smooth muscle cells as well as indirect innervation via ICC.


The following operating grants supported this research: SAF2003-05830 and BFU2006-05055 from the Ministerio de Ciencia y Tecnología PI042460 from Instituto de Salud Carlos III Ministerio de Sanidad y Consumo; Leo Pharma Research Foundation, Denmark; and the Canadian Institutes of Health Research.


  • 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.


View Abstract