Simultaneous imaging of Ca²⁺ signals in interstitial cells of Cajal and longitudinal smooth muscle cells during rhythmic activity in mouse ileum

Abstract

Electrical rhythmicity in smooth muscle cells is essential for the movement of the gastrointestinal tract. Interstitial cells of Cajal (ICC) lie adjacent to smooth muscle layers and are implicated as the pacemaker cells. However, the pace making mechanism remains unclear. To study the intercellular interaction during electrical rhythm generation, we visualized changes in intracellular Ca²⁺ concentration ([Ca²⁺]_i) in smooth muscle cells and myenteric ICC within segments of mouse ileum loaded with a fluorescent Ca²⁺ indicator, fluo-3. We observed rhythmic [Ca²⁺]_i changes in longitudinal smooth muscle cells travelling rapidly through the smooth muscle cell layer. Between the rhythmic Ca²⁺ transients, we found brief Ca²⁺ transients localized to small areas within smooth muscle cells. The amplitude but not the periodicity of rhythmic [Ca²⁺]_i transients in both cell types was partially inhibited by nicardipine, an L-type Ca²⁺ channel antagonist, suggesting that the rhythmic [Ca²⁺]_i transients reflect membrane potential depolarizations corresponding to both slow waves and triggered Ca²⁺ spikes. Longitudinal smooth muscle cells and myenteric ICC showed synchronous spontaneous [Ca²⁺]_i transients in eight out of 21 ileac preparations analysed. In the remaining preparations, the synchrony between ICC and smooth muscle cells was absent, although the rhythmicity of the smooth muscle cells was not disturbed. These results suggest that myenteric ICC may play multiple roles including pace making for physiological bowel movement.

Rhythmic spontaneous depolarizations in intestinal smooth muscle cells, often referred to as slow waves, are essential for the movement of the gastrointestinal tract. The slow waves activate ionic currents, most importantly the L-type Ca²⁺ channels, causing Ca²⁺ action potentials that are superimposed upon the slow waves in smooth muscle cells. The resultant increase in intracellular Ca²⁺ concentration ([Ca²⁺]_i) causes rhythmic smooth muscle contraction and gut movement. Spontaneous contractions are important for the segmentation or mixing of gut contents and for the basal tonus of the gut. A congenital impairment of gut rhythm results in megacolon, and rhythm disorders are implicated in common diseases such as irritable bowel syndrome (Huizinga et al. 1997; Horwitz & Fisher, 2001).

Among the various types of cells that exist in the gastrointestinal tract, interstitial cells of Cajal (ICC) have been implicated in pacemaking for rhythmic contraction of gastrointestinal smooth muscles (Thuneberg, 1982; Ward et al. 1994, 1997, 2000; Huizinga et al. 1995; Koh et al. 1998; Thomsen et al. 1998; Dickens et al. 1999; Lee et al. 1999). ICC are found within the intermuscular space between the longitudinal and circular smooth muscle layers (myenteric ICC), within the circular muscle layer and in the submucosal space (Horowitz et al. 1999). In mice that lack networks of ICC due to mutations in the dominant white spotting locus/c-kit (W/W^V mice) (Chabot et al. 1988), intestinal pacemaker activity is absent, causing megacolon (Maeda et al. 1992; Ward et al. 1994; Huizinga et al. 1995). In the small intestines of these mutant mice, networks of myenteric ICC, which are proposed to be the pacemaker cells of rhythmic contraction, are absent. However, the dysfunction of c-Kit, the product of the c-kit gene, may have multiple effects in addition to the loss of ICC and the functional relationship between ICC and smooth muscle cells remains elusive.

In this work, we simultaneously visualized intracellular Ca²⁺ signals of both myenteric ICC and longitudinal smooth muscle cells within the intact intestinal wall to study the roles of these cells in rhythm generation. We were able to observe rhythmic Ca²⁺ transients that corresponded to slow waves and triggered Ca²⁺ spikes in both cell types. Although we found synchronous rhythmic activities of ICC and smooth muscle cells in a subset of preparations, no synchronization was observed in the rest of the preparations. Our results suggest that the functions of myenteric ICC are multifarious, including rhythmic activities in mouse ileum.

METHODS

Tissue preparation and Ca²⁺ imaging

Control (+/+ or ddy) and W/W^V mice (2–5 weeks old) of either sex were anaesthetized with diethyl ether and exsanguinated, as approved by the local ethics committee. Small intestines (∼5 cm from the caecum) were excised and the serosa were removed by sharp dissection. Segments (2–4 cm) of ileum were incubated in physiological salt solution (PSS) containing 20 μm fluo-3AM for 2–3 h at room temperature (22–25 °C). After the Ca²⁺ indicator loading, a flat glass capillary was inserted into the lumen of the intestine (∼10 mm in length), to form a flat surface for microscopic visualization.

The preparation was positioned securely within a trough with a cover-slip bottom so that the solution around the preparation could be perfused. The trough was then mounted on the stage of an inverted fluorescence microscope (IX70, Olympus) equipped with a cooled charge coupled device (CCD) camera (Photometrics) and with a polychromatic illumination system (T.I.L.L. Photonics). The ileac muscle was viewed under a water-immersion objective (LUMPlanFL 60, NA = 0.90, Olympus) and was illuminated at 480 nm, and the fluorescence emission (515–550 nm) was recorded as a relative measure of [Ca²⁺]_i. The size of the observation field was 209 μm × 164 μm. Two-dimensional images were obtained every 150–1000 ms with an exposure time of 60–100 ms. Because of the sufficiently small focal depth of the objective, individual fluo-3-loaded myenteric ICC and longitudinal smooth muscle cells could be visualized within the tissue. In some experiments, we used a low-magnification objective lens (UPlanFL 20, NA = 0.50, Olympus) to observe wider fields of view (612 μm × 479 μm). We preferred wide-field fluorescence microscopy to confocal laser-scanning microscopy in the present study because of the lower photobleaching and better signal-to-noise ratio. To suppress the movement of smooth muscle cells, 10 μm cytochalasin D (Saito et al. 1996), a capping agent of actin filaments, and 5 μm wortmannin, a myosin light chain kinase (MLCK) inhibitor (Nakanishi et al. 1992) were added to the PSS. The absence of an acute effect of wortmannin on the slow wave has been shown previously (Ward et al. 1999).

Image analysis was carried out on a Macintosh computer using the IPLab programme (Signal Analytics Corporation). The slow decline in fluorescence intensity due to photobleaching of fluo-3 was corrected for by linear extrapolation of the initial decay of fluorescence intensity (excluding the period of rhythmic activities). The fluorescence intensity of individual cells was normalized with respect to the averaged value (F₀) of fluorescence intensity between spontaneous Ca²⁺ transients.

Experiments were carried out at 29–31 °C. Some experiments were carried out at 37 °C and essentially the same results were obtained. However, at 37 °C it was difficult to suppress the movement of the preparations and, therefore, the analysis of the imaging experiments carried out at the lower temperature are presented below.

Immunohistochemistry

Fixed and permeabilized muscle bundles were incubated with anti-murine c-Kit rat monoclonal antibody and fluorescein isothiocyanate (FITC)-labelled anti-rat IgG. Fluorescence images of FITC (excitation wavelength, 488 nm) were viewed under a confocal microscope (Olympus Fluoview) (Yamazawa et al. 1996). FITC staining was not observed in strips treated with the secondary antibody alone. All procedures were carried out at room temperature (22–25 °C).

Measurement of contraction

Longitudinal muscle strips (1.5–2 mm in length, 400–500 μm in width) were carefully dissected in PSS. The muscle strips were tied with silk filament and then attached to a pair of stainless steel hooks, one of which was connected to a strain gauge transducer (BG-10, Kulite) and the other to a micromanipulator to adjust muscle length. The transducer and the micromanipulator were fixed to a common plastic bar that was mounted on a three-dimensional micromanipulator (M-2, Narishige) for the positioning of the preparation in the solution contained in wells (volume 0.5 ml) heated to 29–31 °C.

Materials

PSS contained (mm): 150 NaCl, 4 KCl, 2 CaCl₂, 1 MgCl₂, 5.6 glucose, and 5 Hepes, adjusted to pH 7.4 with NaOH. Fluo-3 AM and 4-(4-diethylaminostyryl)-N-methylpyridinium iodide (4-Di-2-ASP) were purchased from Molecular Probes. Anti-murine c-Kit (ACK-2) rat monoclonal antibody was obtained from Gibco BRL Life Technologies. Cytochalasin D and wortmannin were purchased from Sigma. All other chemicals were of the highest reagent grade available. W/W^V mice were purchased from Japan SLC Inc.

Data analysis

Statistical differences between groups were evaluated with the use of either Student's paired t test or Student's unpaired t test. The synchrony and frequency of spontaneous [Ca²⁺]_i activities were analysed with auto- or cross-correlograms. The values of auto-correlograms were normalized to those at time zero. The values of cross-correlograms were normalized to the geometric average obtained from values of the corresponding auto-correlograms at time zero. All results are expressed as means ± s.e.m.

RESULTS

Spontaneous Ca²⁺ transients in ileac smooth muscle cells

We observed a fluo-3-loaded longitudinal smooth muscle layer from the serosal surface of the excised mouse ileum using a wide-field fluorescence microscope and a CCD camera (Fig. 1A). We were able to distinguish individual longitudinal smooth muscle cells, and time courses of changes in [Ca²⁺]_i were measured within individual cells (areas 1–3, Fig. 1Ba) as well as in the total imaging frame. [Ca²⁺]_i increased rhythmically with synchronization among all the smooth muscle cells within the field of view (Fig. 1D). The auto-correlogram of [Ca²⁺]_i in smooth muscle cells averaged over the entire area showed a periodicity of one event per 8.5 s in this preparation (Fig. 1E). The average peak interval of the rhythmic activities in 12 samples (6.4 ± 0.6 s) was indistinguishable from that of the spontaneous contractions of longitudinal muscle strips under equivalent conditions (6.0 ± 0.8 s, n = 12).

Figure 1. Experimental set-up and Ca²⁺ imaging in ileac smooth muscle cells.

A, an illustration of the experimental arrangement. Segments of ileum were inserted with a rectangular glass capillary and were mounted on the stage of an inverted fluorescence microscope. CM, circular muscle; LM, longitudinal muscle. B, sequential fluorescence intensity images of fluo-3-loaded longitudinal smooth muscle cells. The frame interval was 500 ms and the exposure time of each frame was 100 ms. C, pixel-to-pixel differential images between two successive frames in B, indicating the areas where Ca²⁺ concentration increased. Arrows indicate local Ca²⁺ transients in smooth muscle cells. D, fluorescence intensity change plotted against time in the total area (thick line) and in small areas (1–3) of Ba. Images Ba-h correspond to the fourth Ca²⁺ transient underlined by the red horizontal bar. * indicate local Ca²⁺ transients. E, auto-correlogram of Ca²⁺ transients in the total area. The average interval of rhythmic activities was 8.5 s. F, histogram of the number of local Ca²⁺ transients between synchronized rhythmic Ca²⁺ transients. Local Ca²⁺ events in 11 cells within the imaging field were counted with respect to the time prior to the peaking of following global Ca²⁺ transients (summation of 5 intervals). The observed number of events probably underestimates the actual number of events due to the duty ratio (100 ms/500 ms = 0.2) of image acquisition. Representative results of 12 experiments.

Time-lapse Ca²⁺ images taken at 500 ms intervals during the fourth rhythmic activity (underlined by the red horizontal bar) in Fig. 1D are shown in Fig. 1Ba–h. A significant increase in [Ca²⁺]_i was observed in Bg. As shown in Fig. 1C, the pixel-to-pixel differential images of the sequential images (e.g. b′ = b − a) highlight cells in which [Ca²⁺]_i increased between successive frames. A synchronous increase in [Ca²⁺]_i in all the smooth muscle cells (Fig. 1Cg′) followed a segmental increase in [Ca²⁺]_i (Fig. 1Cf′).

In between the synchronized rhythmic Ca²⁺ transients, we found local Ca²⁺ transients in smooth muscle cells (red arrows in Fig. 1C). These local [Ca²⁺]_i transients tended to appear a few seconds before the rhythmic global Ca²⁺ responses (Fig. 1D, asterisks, and 1F). Depletion of Ca²⁺ stores by Ca²⁺-ATPase inhibitors (3–5 μm cyclopiazonic acid or 1 μm thapsigargin, n = 3) blocked both local and rhythmic Ca²⁺ responses. Furthermore, 1–10 mm caffeine blocked both local and rhythmic Ca²⁺ transients (n = 3), although 30 μm ryanodine had no effect (n = 4). These results are consistent with the notion that a compartment of intracellular Ca²⁺ stores was involved in local Ca²⁺ transients which, in turn, may have been involved in the generation of rhythmic Ca²⁺ transients (Suzuki & Hirst, 1999; van Helden et al. 2000).

Propagation of intercellular Ca²⁺ waves in ileac smooth muscle cells

We also studied [Ca²⁺]_i changes in a wider area of the longitudinal smooth muscle layer at a lower magnification. Traces in Fig. 2B show the time courses of changes in [Ca²⁺]_i within areas 1–3 in Fig. 2A. The differential Ca²⁺ images taken at 200 ms intervals during the first rhythmic activity (underlined by the red horizontal bar) in Fig. 2B are shown in Fig. 2C. The Ca²⁺ transients began in a small region near area 2 and propagated as Ca²⁺ waves through the longitudinal muscle layer (Fig. 2B, arrows, and 2C). The initiation site of the intercellular waves was not fixed but migrated and the propagation velocity of Ca²⁺ transients in the circular direction was ∼300 μm s⁻¹. These results obtained by the current imaging method are in general agreement with those of previous studies, in which similar intercellular Ca²⁺ waves were observed (Stevens et al. 1999a, b, 2000; Hashitani et al. 2001).

Figure 2. Intercellular Ca²⁺ waves in longitudinal smooth muscle layer.

A, Ca²⁺ imaging at low magnification (× 20 objective). B, fluo-3 fluorescence intensity change is plotted against time in 3 areas of the image. Red arrows indicate the direction of Ca²⁺ waves (upward-pointing arrows correspond to leftward waves in A and vice versa). C, pixel-to-pixel differential Ca²⁺ images (200 ms intervals) during the first rhythmic activity underlined by the red horizontal bar in B.

Identification of myenteric ICC within intact ileum

In close association with the longitudinal smooth muscle layer of fluo-3-loaded ileum preparations, we observed cells with long processes and a morphology distinct from that of smooth muscle cells (Fig. 3Aa, arrows). No such cells were found in the ileum of W/W^V mice (Fig. 3Ab). Since W/W^V mice lack the network of myenteric ICC, it is likely that these cells correspond to myenteric ICC. When we treated the preparation with methylene blue (MB), which preferentially accumulates in ICC (Liu et al. 1993; Sanders, 1996), the cells with long processes were stained with the dye (Fig. 3Ac). No MB-stained cells were found in the W/W^V mouse ileum (Fig. 3Ad). To confirm whether these cells were myenteric ICC, we carried out an immunohistochemical analysis using an anti-c-Kit antibody (Nishikawa et al. 1991). Confocal sections obtained at 6 μm intervals from the serosal surface of the longitudinal muscle layer towards the circular muscle showed c-Kit-positive cells in the control mice but not in the W/W^V mice (Fig. 3B), in agreement with previous observations (Ward et al. 1994; Torihashi et al. 1995,Huizinga et al. 1995; Komuro & Zhou, 1996). Since the longitudinal smooth muscle layer of mouse ileum is composed of one to two cell layers with a thickness of ∼10 μm (at 2–5 weeks of age), the c-Kit-positive cells are located in close association with the longitudinal smooth muscle cells. Thus, there was an extremely good correlation between MB-positive cells and c-Kit-positive cells in terms of morphology, density and location. In the following experiments, myenteric ICC were identified on the basis of their unique morphology within the intact tissue after fluo-3-loading. At the end of each experiment, preparations were stained with MB to confirm identification of ICC.

Figure 3. Identification of ICC within intact intestinal wall.

Aa and b, fluo-3-loaded cells in longitudinal smooth muscle layer of ileum. Ac and d, staining with methylene blue reveals ICC-like cells (arrows) in the ileum of control mouse but not in that of the W/W^V mouse. Ae and f, staining with 4-Di-2-ASP reveals Auerbach's nerve plexus as a mesh-like structure beneath the longitudinal muscle layer in the ileum from both control and W/W^V mice. B, immunohistochemical analysis using anti-c-Kit antibody. Confocal images of the control and the mutant ileum stained with FITC-labelled secondary antibody. The focal plane was shifted at 6 μm intervals from the serosal surface of the longitudinal muscle layer toward the circular muscle layer. c-Kit-positive cells were found associated with the longitudinal smooth muscle layer only in wild-type ileum.

In both control and W/W^V mice, images focused between longitudinal and circular muscle layers showed Auerbach's nerve plexus as a mesh-like structure when stained with a fluorescent probe for nerve fibres, 10 μm 4-Di-2-ASP (Iino et al. 1994; Kasai et al. 1997) (Fig. 3Ae and Af). The network of Auerbach's nerve plexus and the longitudinal smooth muscle cells could be distinguished under the wide-field fluorescence microscope by shifting the focal plane at × 60 magnification, but the circular muscle layer was too deeply positioned and their signals were not detected due to absorption and scattering of light.

Simultaneous Ca²⁺ imaging of ICC and smooth muscle cells during synchronized rhythmic activity

Because longitudinal smooth muscle cells and myenteric ICC were present at nearly the same depth in the ileum preparations, it was possible to obtain their fluorescence images simultaneously within the same imaging frames (Fig. 4A, areas 1–3: smooth muscle cells; areas 4–6: myenteric ICC). Figure 4B shows the changes in [Ca²⁺]_i with time within individual smooth muscle cells and in myenteric ICC. In this preparation, smooth muscle cells and myenteric ICC produced synchronized and rhythmic increases in [Ca²⁺]_i. The differential Ca²⁺ images taken at 500 ms intervals during the first, second, third and sixth rhythmic activities (marked by rectangular boxes a-d) in Fig. 4B are shown in Fig. 4C. A concomitant initiation of the increase in [Ca²⁺]_i in both smooth muscle cells and ICC can be seen (Fig. 4C). A prolonged increase in [Ca²⁺]_i in myenteric ICC can be observed as the remaining fluorescence images of these cells in the right-hand panels of Fig. 4C. These results clearly showed that we were able to distinguish Ca²⁺ signals of myenteric ICC from those of neighbouring smooth muscle cells.

Figure 4. Simultaneous imaging of Ca²⁺ transients in smooth muscle cells and ICC.

A, images of fluo-3-stained cells within the longitudinal smooth muscle layer. B, fluo-3 fluorescence intensity change is plotted against time in 6 areas of the image. Smooth muscle cells and myenteric ICC show synchronized rhythmic Ca²⁺ transients. * indicates local Ca²⁺ transients. ♦ indicates low-amplitude Ca²⁺ transients. C, differential images (as in Fig. 1C) during the period indicated by rectangles in B. Images were acquired every 500 ms.

Figure 5 shows the results obtained from another ileum preparation in which longitudinal smooth muscle cells and myenteric ICC produced synchronized and rhythmic increases in [Ca²⁺]_i. The images were acquired every 250 ms in this experiment and the changes in [Ca²⁺]_i within representative cells (four smooth muscle cells (SMCs) and three myenteric ICC) are shown (Fig. 5A). The thick black traces show mean fluorescence intensity changes with time in smooth muscle cells (12 cells) and myenteric ICC (three cells). The red traces represent the first derivatives of the mean fluorescence intensities. Figure 5B shows the auto-correlogram of the time courses of [Ca²⁺]_i in both smooth muscle cells and ICC. The mean peak interval of the rhythmic activities of myenteric ICC (5.6 ± 0.1 s, 10 intervals) was indistinguishable from that of the smooth muscle cells (5.6 ± 0.08 s, 10 intervals). The temporal relationship of the [Ca²⁺]_i transients between longitudinal smooth muscle cells and myenteric ICC was examined by a cross-correlogram (Fig. 5C) that showed a prominent peak near time zero, suggesting a close temporal association of the activities of the two types of cells. The peak of the cross-correlogram is slightly shifted toward the positive time. This may reflect the longer duration of the ICC signals than the smooth muscle cell signals. We therefore analysed the temporal relationship by means of another method that involved overlaying the 11 transients by aligning the peak of the first derivative of the mean Ca²⁺ transients of smooth muscle cells (Fig. 5D left panel). The right panel of Fig. 5D shows the average of thus-aligned Ca²⁺ transients and reveals general agreement between the rising phase in smooth muscle cells and myenteric ICC.

Figure 5. Analysis of spontaneous Ca²⁺ transients in smooth muscle cells and ICC.

A, fluo-3 fluorescence intensity change is plotted against time in 7 areas of the image. Longitudinal smooth muscle cells (areas 1–4) and myenteric ICC (areas 5–7) showed synchronized rhythmic Ca²⁺ transients. The thick black traces represent mean fluo-3 fluorescence intensities in longitudinal smooth muscle cells and myenteric ICC. The red traces represent the first derivatives of mean fluo-3 intensities (d ΔF/dt). * indicates local Ca²⁺ transients. ♦ indicates low-amplitude Ca²⁺ transients. B, auto-correlograms of smooth muscle cells and myenteric ICC revealed a marked periodicity of one event per 5.6 s in the absence and presence of 500 nm nicardipine. C, cross-correlogram of smooth muscle cells and myenteric ICC. D, results of the overlay method to compare Ca²⁺ transients in smooth muscle cells with those in ICC. E and F, Ca²⁺ transients in the presence of 500 nm nicardipine and the results of analysis by the overlay method. Note that the frequency of the rhythmic Ca²⁺ transients was not altered, although the amplitude was significantly reduced. Figures show the representative results of 8 experiments.

Unsynchronized rhythmic [Ca²⁺]_i activities in smooth muscle cells and ICC

In some preparations (seven out of 21 preparations), although both smooth muscle cells and myenteric ICC showed rhythmic [Ca²⁺]_i activities, no temporal synchronization between the two types of cells was observed and an example of this is shown in Fig. 6. While myenteric ICC (areas 1–5, Fig. 6A) showed regular activities, the frequency of the rhythmicity was about half that of smooth muscle cells (Fig. 6B). This was further supported by the auto-correlograms of the mean Ca²⁺ signals in the two types of cells (Fig. 6C). Furthermore, not all Ca²⁺ transients of ICC showed close temporal association with those of smooth muscle cells (Fig. 6B). Indeed, the cross-correlogram of the Ca²⁺ transients in the two types of cells does not produce a prominent peak near time zero (Fig. 6C), showing that the Ca²⁺ transients of smooth muscle cells and myenteric ICC are not synchronized in this preparation.

Figure 6. Low -frequency rhythmic Ca²⁺ transients in myenteric ICC.

A representative result of ileum preparations in which ICC generated rhythmic activities that were not synchronized with those of longitudinal smooth muscle cells. A, fluo-3-stained cells within the longitudinal muscle layer. Five ICC are marked by circles. B, Thick lines, mean Ca²⁺ transients in smooth muscle cells (n = 28) and in ICC (n = 5). Thin lines, Ca²⁺ transients in individual ICC. Although myenteric ICC showed rhythmic Ca²⁺ transients, they were not in phase with those of longitudinal smooth muscle cells. C, auto- and cross-correlograms of smooth muscle cells and myenteric ICC. Images were taken every 250 ms. Figures show representative results of 7 experiments.

Source of Ca²⁺ in rhythmic Ca²⁺ responses

We then analysed the effect of L-type Ca²⁺ channel antagonists, which are known to inhibit the triggered Ca²⁺ spikes but have no effect on the slow waves (Huizinga et al. 1997; Horowitz et al. 1999). When 500 nm nicardipine was applied to block the L-type Ca²⁺ channels, the amplitude of the rhythmic Ca²⁺ transients was decreased in both longitudinal smooth muscle cells and myenteric ICC (Fig. 5E). However, the peak interval of the rhythmic Ca²⁺ transients of both cell types was not altered by nicardipine, as shown by the auto-correlograms (Fig. 5B). When each Ca²⁺ signal was overlaid as in Fig. 5D, we again observed synchronization of the rising phase of Ca²⁺ signals between smooth muscle cells and ICC (Fig. 5F). The specific effect of the Ca²⁺ channel antagonist on the amplitude but not on the periodicity of Ca²⁺ transients in both smooth muscle cells and myenteric ICC was confirmed in several other preparations (with either synchronized or unsynchronized Ca²⁺ transients) (Fig. 7). These results suggest that the nicardipine-insensitive component of the rhythmic Ca²⁺ transients corresponds to the slow waves. When we visualized [Ca²⁺]_i in the ileum of W/W^V mice, regular and rhythmic activity of the longitudinal smooth muscle was observed in some limited fields of observation. However, these rhythmic Ca²⁺ transients in mutant mice were completely blocked by 500 nm nicardipine (n = 4).

Figure 7. Effect of a Ca²⁺ channel antagonist on rhythmic Ca²⁺ transients.

Summary of the effect of 500 nm nicardipine on the periodicity (interval) and amplitude of rhythmic Ca²⁺ transients in longitudinal smooth muscle cells (open columns) and in myenteric ICC (black columns). Means ± s.e.m., n = 6.

The spontaneous Ca²⁺ transients of smooth muscle cells and myenteric ICC were not changed by atropine (1 μm, n = 2) or tetrodotoxin (10 μm, n = 2), supporting the view that these Ca²⁺ transients are myogenic in origin. The spontaneous Ca²⁺ transients in ileac preparations were abolished when the preparations were treated with MB and were exposed to light (50 μm, n = 3). This is consistent with the results of the previous work that MB suppresses slow waves (Liu et al. 1993).

Complex [Ca²⁺]_i transients in myenteric ICC

In some preparations (six out of 21 preparations), myenteric ICC showed low-frequency irregular activities (Fig. 8). Although synchronous rhythmic activities were observed among all the smooth muscle cells within the field of observation, different parts of the network of myenteric ICC (areas 1–7) showed low-frequency irregular activities that were not well synchronized (Fig. 8B). An increase in [Ca²⁺]_i was first observed in area 6 (Fig. 8C, arrowhead) in synchrony with that in longitudinal smooth muscle cells. Then, propagation of Ca²⁺ increase from area 1 to area 5 was observed (Fig. 8C, arrows). Although area 6 of the ICC network showed a synchronous response with the longitudinal smooth muscle cells during one of the rhythmic Ca²⁺ responses (Fig. 8C, arrowhead), there was no general temporal association between the activities of longitudinal smooth muscle cells and those of myenteric ICC, either in the cell bodies of myenteric ICC (areas 1–5) or at the processes (areas 6 and 7) (Fig. 8B).

Figure 8. Intercellular Ca²⁺ waves in myenteric ICC.

A, fluorescence image of fluo-3-loaded cells within the longitudinal muscle layer. B, changes in [Ca²⁺]_i were plotted against time in mean smooth muscle cells (19 cells) and 7 small areas within a network of myenteric ICC. C, differential fluorescence images taken during the period indicated by the red horizontal bar in B, indicating the cells in which [Ca²⁺]_i rose. Note that the [Ca²⁺]_i transients in most of the myenteric ICC lagged behind those of smooth muscle cells. An arrowhead and arrows indicate the site of Ca²⁺ increases in myenteric ICC. Images were taken every 250 ms. Figure shows representative results of 6 experiments.

Comparison of [Ca²⁺]_i transients between smooth muscle cells and ICC

As described above, myenteric ICC showed three types of Ca²⁺ transients while longitudinal smooth muscle cells exhibited regular rhythmic Ca²⁺ transients. The results obtained from 21 ileum preparations are compiled in Fig. 9A to compare the mean peak intervals of the Ca²⁺ transients in smooth muscle cells with those in myenteric ICC. The average of the mean peak interval of myenteric ICC (10.1 ± 1.2 s) was significantly greater than that of smooth muscle cells (6.2 ± 0.5 s, P < 0.05). The half-width of the rhythmic Ca²⁺ transients was significantly greater in myenteric ICC (2.8 ± 0.4 s) than in longitudinal smooth muscle cells (1.5 ± 0.3 s, P < 0.05). In eight out of 21 ileum preparations, there was a 1:1 correlation between the activities of smooth muscle cells and those of myenteric ICC in terms of the mean peak intervals (Fig. 9A, red open circles a-h). The temporal relationship between the spontaneous Ca²⁺ transients of smooth muscle cells and those of myenteric ICC in these preparations was analysed by the overlay method as in Fig. 5 (Fig. 9Ba–h). The rising phases of the Ca²⁺ transients in both smooth muscle cells and ICC showed good agreement, although the smooth muscle cells Ca²⁺ transients peaked earlier than myenteric ICC Ca²⁺ transients.

Figure 9. Summary of data on simultaneous imaging of Ca²⁺ transients.

A, comparison of mean peak intervals of Ca²⁺ transients between smooth muscle cells and myenteric ICC in 21 ileum preparations. Each symbol indicates the result of an ileum preparation from a different animal. B, overlay analysis of 8 samples in which smooth muscle cells and ICC showed Ca²⁺ transients with the same periodicity (red symbols, a-h, in A). Red traces, longitudinal smooth muscle cells; blue traces, myenteric ICC. Ca²⁺ transients were normalized by their peak values before the overlay analysis. The ileum preparations analysed in Figs 4 and 5 correspond to h and e, respectively.

In some preparations (n = 7), smooth muscle cells and myenteric ICC generated rhythmic but unsynchronized [Ca²⁺]_i activities (Fig. 9A, black open circles). In six preparations, ICC showed irregular increases in [Ca²⁺]_i, while smooth muscle cells showed rhythmic Ca²⁺ transients (Fig. 9A, filled circles). Importantly in these experiments, the frequency of rhythmic activity in myenteric ICC was never greater than that observed in longitudinal smooth muscle cells.

DISCUSSION

The present work provides the first simultaneous [Ca²⁺]_i recording from myenteric ICC and smooth muscle cells side by side in situ. We observed rhythmic [Ca²⁺]_i increases in all the longitudinal smooth muscle cells and in a subset of myenteric ICC. When the L-type voltage-dependent Ca²⁺ channels were blocked by nicardipine, the amplitude of the rhythmic Ca²⁺ signals was partially decreased, although the periodicity of the remaining Ca²⁺ transients was not altered. This is in parallel with the measurement of membrane voltage in intestinal smooth muscle cells, in that the amplitude and periodicity of slow waves are insensitive to Ca²⁺ channel antagonists but the Ca²⁺ spikes that are triggered by the slow waves are inhibited by the drugs (Huizinga et al. 1997; Horowitz et al. 1999). These results suggest that the nicardipine-sensitive and insensitive components of the Ca²⁺ transients correspond to triggered Ca²⁺ spikes and slow waves, respectively. Furthermore, the rhythmic Ca²⁺ transients were inhibited by both caffeine and cyclopiazonic acid, which are well-known inhibitors of slow waves (Liu et al. 1995; Dickens et al. 1999; Ward et al. 2000; Malysz et al. 2001). The rhythmic Ca²⁺ transients in smooth muscle cells sometimes show decreased peak sizes and such low-amplitude transients may represent slow waves that failed to trigger Ca²⁺ spikes (e.g. Ca²⁺ transients marked by diamonds in Fig. 4B and Fig. 5A).

Our results clearly showed that multiple types of spontaneous Ca²⁺ transients exist in myenteric ICC of mouse ileum. A tight temporal relationship between the activities of smooth muscle cells and those of myenteric ICC was observed in eight out of 21 preparations (Fig. 9A red open circles). In those preparations, the rising phase of the increase in [Ca²⁺]_i in smooth muscle cells coincided with that of myenteric ICC (Fig. 9B). These results are consistent with the hypothesis that myenteric ICC are the pacemakers of smooth muscle cell activities. In the electron microscopic studies, gap junctions were found between longitudinal smooth muscle and myenteric ICC in canine intestine (Daniel et al. 2001). On the other hand, in the remaining ileum preparations that were examined in this study, there was no 1:1 temporal relationship between the activities of smooth muscle cells and those of myenteric ICC (Fig. 9A black open and filled symbols). The myenteric ICC observed in such preparations are unlikely to be the pacemakers of the adjacent smooth muscle cells. It is possible that pacemaker ICC are present beyond the imaging field and that electrical activity is conducted via longitudinal smooth muscle cells. Indeed, the Ca²⁺ transients in smooth muscle cells show intercellular waves (Fig. 2) and it is possible that there are certain areas that produce the rhythmicity and initiate intercellular waves that invade the remaining smooth muscle cells. The present results do not exclude the possibility that smooth muscle cells produce the rhythmicity by themselves and the resultant depolarization spreads to myenteric ICC, with some delay and with less than 100 % efficiency, via the electrical coupling between the two types of cells. Taken together, it appears that not all myenteric ICC are pacemaker cells and that these cells may have additional roles in the intestine.

The mechanism of rhythmic depolarizations in the intestine remains to be clarified. Recently, an interesting mechanism for the generation of slow waves in the stomach has been postulated, which may be referred to as the ‘IP₃-depolarization cross-coupling mechanism’ (Suzuki & Hirst, 1999; van Helden et al. 2000). In this hypothesis, local Ca²⁺ release events via the IP₃ receptors activate inward currents to generate spontaneous transient depolarizations which, in turn, stimulate IP₃ production via a mechanism that is yet to be identified (Ganitkevich & Isenberg, 1993). Because membrane depolarization can spread among neighbouring cells via the electrical coupling, these events in either ICC or smooth muscle cells are assumed to make a positive feedback loop to generate rhythmic and synchronized depolarizations. In accordance with this hypothesis, previous results suggest that the IP₃ receptor is required for slow wave generation (Liu et al. 1995; Suzuki et al. 2000; Ward et al. 2000). We observed repetitive, asynchronous and local Ca²⁺ transients between the global Ca²⁺ transients within smooth muscle cells (Fig. 1C, asterisks in Figs 1D, 4B and 5A). The frequency of the appearance of such local Ca²⁺ transients gradually increased before the initiation of rhythmic activities (Fig. 1F). When these local Ca²⁺ transients were blocked by the depletion of Ca²⁺ stores, the rhythmic activities of smooth muscle cells were abolished. It is an interesting possibility, therefore, that the local Ca²⁺ transients observed in this study are involved in the generation of spontaneous transient depolarizations. Further investigation is awaited regarding this issue