Differential expression of ionic conductances in interstitial cells of Cajal in the murine gastric antrum

Abstract

Two distinct populations of interstitial cells of Cajal (ICC) exist within the tunica muscularis of the gastric antrum, and these cells serve different physiological functions. One population of ICC generates and actively propagates electrical slow waves, and the other population of ICC is innervated by excitatory and inhibitory motor neurons and mediates enteric motor neurotransmission. In spite of the key role of ICC in gastric excitability, little is known about the ionic conductances that underlie the functional diversity of these cells. In the present study we isolated ICC from the murine gastric antrum and investigated the Ca²⁺-dependent ionic conductances expressed by these cells using the patch clamp technique. Conductances in ICC were compared with those expressed in smooth muscle cells. The cells studied were identified by RT-PCR using cell-specific primers that included Myh11 (smooth muscle cells), Kit (ICC) and Uchl1 (enteric neurons) following electrophysiolgical recordings. Distinct ionic conductances were observed in Kit-positive cells. One group of ICC expressed a basal non-selective cation conductance (NSCC) that was inhibited by an increase in [Ca²⁺]_i in a calmodulin (CaM)-dependent manner. A second population of ICC generated spontaneous transient inward currents (STICs) and expressed a basal noisy NSCC that was facilitated by an increase in [Ca²⁺]_i in a CaM-dependent manner. The [Ca²⁺]_i-facilitated NSCC in ICC was blocked by the Cl⁻ channel antagonists 4,4′-diisothiocyanatostilbene-2,2′-disulphonic acid (DIDS), anthracene-9-carboxylate (9-AC) and niflumic acid. These data suggest that distinct NSCC are expressed in subpopulations of ICC and these conductances may underlie the functional differences of these cells within the gastric antrum.

In the gastric corpus and antrum, phasic mechanical activity is initiated by spontaneous, rhythmic membrane depolarizations termed slow waves (Tomita, 1981). Slow waves and associated contractile activity of gastric muscles are generated in the absence of neuronal or hormonal stimulation, but the amplitudes and frequency of these events are regulated by a variety of excitatory and inhibitory neural inputs from enteric motor nerves and by hormones and paracrine substances (Vizi et al. 1973; Debas et al. 1975; Szurszewski, 1978; Allen et al. 1984; Dickens et al. 2000; Hirst et al. 2002b,c).

Slow waves and enteric neural responses in the antrum depend upon two distinct populations of interstitial cells of Cajal (ICC). One subgroup of ICC lies in the region between the circular and longitudinal muscle layers, at the level of the myenteric plexus, and are termed ICC-MY. These cells are the dominant pacemakers within the wall of the stomach and generate electrical slow waves (Dickens et al. 1999; Ordog et al. 1999). The second population of ICC, diffusely distributed within muscle bundles of the circular muscle layer, are referred to as intramuscular ICC (ICC-im). ICC-IM are closely associated with nerve varicosities of both excitatory and inhibitory motor neurons, and these cells mediate postjunctional responses to cholinergic and nitrergic neural inputs (Burns et al. 1996; Suzuki & Hirst, 1999; Ward et al. 2000a; Ward & Sanders, 2001; Hirst et al. 2002c; Beckett et al. 2002, 2003; Suzuki et al. 2003). ICC-IM also generate a continuous discharge of electrical activity termed unitary potentials, which contribute to the excitability of the circular muscle layer (Edwards et al. 1999). Increases in [Ca²⁺]_i via cholinergic stimulation or current injection cause unitary potentials in ICC-IM to summate, leading to the formation of regenerative potentials in isolated muscle bundles lacking ICC-MY (Edwards et al. 1999; Suzuki & Hirst, 1999; Kennedy, 2000; van Helden et al. 2000; Dickens et al. 2001; Hirst et al. 2002a,b,c). Activation of regenerative potentials by release of acetylcholine (ACh) from motor neurons can phase advance slow waves in ICC-MY within intact muscle strips, and this is how excitatory nerves regulate the frequency of slow waves in the stomach (Hirst et al. 2002c; Beckett et al. 2003). Inhibitory nerve stimulation, via the release of nitric oxide, reduces unitary potentials, and this contributes to the stabilization of membrane potential (Suzuki et al. 2003; Teramoto & Hirst, 2003). ICC-IM also respond to stretch of the gastric antral wall with depolarization and a chronotrophic effect on slow wave activity (Won et al. 2005).

The electrical properties that make ICC-MY and ICC-IM unique from smooth muscle cells in the antrum have been surmised primarily from studies of muscle strips and small muscle bundles using intracellular recordings and pharmacological techniques. However, there are two major problems with these approaches. First, intracellular recordings from intact GI muscles are confounded by the fact that multiple cell types contribute to the overall excitability of these tissues (Sanders et al. 2006). Second, some pharmacological tools, such as Cl⁻ channel blocking drugs, which have been used to dissect the ionic conductances contributing to gastric excitability lack specificity (Dick et al. 1999; Welsh et al. 2000). It has been proposed that Ca²⁺-activated Cl⁻ conductances are responsible for the primary pacemaker currents in ICC-MY (Kito et al. 2002; Kito & Suzuki, 2003), and unitary or regenerative potentials in ICC-IM (Edwards et al. 1999; Hirst et al. 2002b,c; Kito et al. 2002), since these electrical events were inhibited by 4,4′-diisothiocyanatostilbene-2,2′-disulphonic acid (DIDS), anthracene-9-carboxylate (9-AC) and 1,2-bis-(o-aminophenoxy)-ethane-N,N,N′,N′-tetraacetic acid, tetraacetoxymethyl ester (BAPTA-AM). However, a Ca²⁺-activated Cl⁻ conductance has never been identified in isolated ICC (Lee & Sanders, 1993; Kim et al. 2002; Goto et al. 2004).

A non-selective cation conductance (NSCC) has been recorded in ICC-MY cultured from the murine small intestine (Koh et al. 2002). This conductance is activated by a reduction in [Ca²⁺]_i, and it is thought to be responsible for the generation of pacemaker activity in ICC-MY (Koh et al. 2002). These studies, however, are subject to the criticism that the expression of specific ionic conductances can change in primary cell cultures (Thio et al. 1993; Mitcheson et al. 1996; Ihara et al. 2002).

Ionic conductances expressed in specific populations of ICC have not been well studied because of several confounding factors. ICC contribute only a small fraction of the total number of cells within the gastric tunica muscularis, and identification of these cells after enzymatic dispersion is difficult. Specific markers for ICC include antibodies against the receptor tyrosine kinase, c-kit; however, this receptor appears to be cleaved and labelling lost during enzymatic dispersion of cells. A number of investigators have attempted to label ICC using c-kit antibody conjugated to a fluorescent probe prior to cell dispersion; however, labelling of cells becomes unspecific during the cell dispersion process and macrophages present in the cellular dispersion can phagocytose the conjugated c-kit antibody (S. M. Ward, unpublished observation). In the present work, we performed patch clamp studies on cells that we identified as possible ICC in fresh enzymatic dispersions, and then verified the identity of the cells using single cell RT-PCR. Kit-positive cells (ICC) expressed NSCC with distinctive properties that separated these cells into two populations. These conductances were regulated by calmodulin, but exhibited opposite responses to changes in [Ca²⁺]_i. Our data suggest that the two ICC populations (ICC-MY and ICC-IM) in gastric antrum may exhibit different behaviours due to the expression of conductances with different regulatory properties.

Methods

Animals

Balb/C mice between 7 and 12 days old were used for the described studies. Animals were obtained from the Jackson Laboratory (Bar Harbor, MN, USA). Mice were anaesthetized with isoflurane (Baxter, Deerfield, IL, USA) prior to cervical dislocation and decapitation. The animals were maintained and the experiments performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The Institutional Animal Use and Care Committee at the University of Nevada approved all procedures used.

Preparation of isolated ICC

To prepare freshly dispersed ICC, antral tissues were cut open along the longitudinal axis of the lesser curvature, pinned in a Sylgard-lined dish and washed with Ca²⁺-free solution containing (mm): 125 NaCl, 5.4 KCl, 15.5 NaOH, 0.34 Na₂HPO₄, 0.44 KH₂PO₄, 10 glucose, 2.9 sucrose, and 11 Hepes, adjusted to pH 7.4 with Tris buffer. Mucosa and submucosal layers were removed by sharp dissection and the remaining tunica muscularis was incubated in 1 ml of Ca²⁺-free solution supplemented with 1.3 mg collagenase (Worthington, Lakewood, NJ, USA), 2 mg fatty acid-free bovine serum albumin, and 2 mg trypsin inhibitor (Sigma-Aldrich, St Louis, MO, USA) for 8–12 min at 37°C. Following incubation in the enzyme solution, tissues were washed repeatedly (3–5 times) with Ca²⁺-free solution and gently agitated to create a cell suspension. Dispersed cells were stored at 4°C in the Ca²⁺-free Hepes solution.

Voltage-clamp patch experiments

The amphotericin B-perforated whole-cell patch-clamp technique was used to record membrane currents from dissociated murine antral cells. Amphotericin B (Sigma) was dissolved in dimethyl sulphoxide (Sigma) as a stock solution (0.08 mg μl⁻¹) and added to the pipette solution (0.4 mg ml⁻¹). Voltage ramps (400 ms duration from −100 to +100 mV or from −80 mV to +50 mV) or step depolarizations (where indicated) were applied and currents recorded. Currents were amplified with an Axopatch 200B (Axon Instruments, Union City, CA, USA). Data were digitized with a 16-bit analog to digital converter (Digidata 1322A, Axon Instruments) and stored directly on-line using pCLAMP software (version 9.2, Axon Instruments). Data were sampled at 4 kHz and filtered at 2 kHz using an eight-pole Bessel filter. Continuous holding currents (basal currents) were monitored by miniDigi (Axon Instruments) using Axoscope 9.2 software (Axon Instruments). These data were sampled at 1 kHz with no filtering. All data were analysed using pCLAMP (v. 9.2), Graphpad Prism (version 3.0, Graphpad Software Inc., San Diego, CA, USA), and Origin (v. 5.0, OriginLab Corp., Northampton, MA, USA) software.

Solutions for patch clamp experiments

In order to measure currents, mixed isolated cells (ICC, enteric neurons, smooth muscle cells) were bathed in a Ca²⁺-containing physiological salt solution (CaPSS) containing (mm): 5 KCl, 135 NaCl, 2 CaCl₂, 10 glucose, 1.2 MgCl₂, and 10 Hepes adjusted to pH 7·4 with Tris. In some experiments cells were perfused in 0 mm Ca²⁺ (BaPSS; Ca²⁺ was replaced with Ba²⁺ (2 mm)). In Na⁺ replacement experiments, Na⁺ was replaced with equimolar TEA. The pipette solution contained (mm): 135 CsCl, 5 NaCl, 2.5 MgCl₂ and 10 Hepes, adjusted to pH 7·2 with Tris. For low Cl⁻ solution (10 mm), CsCl was replaced with equimolar caesium aspartate (E_Cl was −70 mV). The calculated junction potential was −17 mV. The junction potential was digitally corrected during experiments. The effects of gadolinium (Gd³⁺), cacmidazolium (CMZ), DIDS 9-AC and niflumic acid (Sigma, St Louis, MO, USA) were tested by bath perfusion. Whole cell inward currents were performed between 29 and 30°C with the use of a CL-100 bath heater (Warner Instruments; Hamden, CT, USA) within 6 h of dispersing cells.

Calcium imaging analysis

A stock solution of Fluo-4 AM (FluoroPure AM; Molecular Probes, Eugene, OR, USA) was dissolved in DMSO (50 μg Fluo-4 AM in 10 μl DMSO). One microlitre of Fluo-4 stock solution (5 μg) was added to dispersed gastric antral cells in 1 ml of Ca²⁺ free solution. Cells were incubated in Fluo-4 for 15 min (4°C) after which they were perfused with CaPSS solution at 24 ± 0.5°C for 10 min to allow for de-esterification of the dye. Cells were imaged under an inverted fluorescence microscope (Nikon, TE2000-S; Technical Instruments, Burlingame CA, USA), using excitation and emission suitable for Fluo-4 (excitation 460–490 nm and emission > 515 nm), delivered via a xenon arc from a Lambda DG-5 (Sutter Instruments, Novato, CA, USA). Neutral density filters were used to adjust excitation intensity. Images of Ca²⁺-induced fluorescence changes were recorded at 30 ± 0.5°C using a Hamamatsu ORCA digital camera (Bridgewater, NJ, USA) and SIMPLE PCI (version 5.3.1; Compix Inc. Imaging Systems, PA, USA). For most experiments standard ramp protocols were applied and voltage-dependent changes in [Ca²⁺]_i recorded.

Analysis of electrophysiological data

Data are expressed as means ± standard errors of the mean. Student's t test was used where appropriate to evaluate differences in the data. P-values < 0.05 were taken as statistically significant. For patch clamp experiments, peak currents were analysed before and during application of drugs.

In some experiments it was unclear whether these Kit-positive cells which produced small inward currents in response to membrane depolarization were either ICC exhibiting small current responses or other cell types which produced false Kit-positives via cDNA contamination during the single-cell RT-PCR amplification procedure. Cells that expressed [Ca²⁺]_i-facilitated currents greater than 100 pA were predominantly Kit-positive, and therefore cells that exhibited [Ca²⁺]_i-facilitated currents greater than −180 pA (average amplitude of kit-positive cells showing calcium facilitated current) were treated as ICC and data obtained from these cells was included in the analysis.

Total RNA isolation and RT-PCR

After patch clamp experiments were performed, each antral cell that was studied was collected using applied suction to the pipette which resulted in aspiration of the cell into the patch pipette. The pipette contents were subsequently ejected into a sterile 500 μl tube. Cells were rapidly frozen down in liquid nitrogen and stored at −20°C until single cell RT-PCR was performed. Total RNA was isolated from each cell using the Cells-to-cDNA II (Ambion Inc., Austin, TX, USA), according to the manufacturer's instructions. Total RNA was also extracted from whole murine antral tissue (mucosa and submucosa removed) using Trizol reagent (Gibco BRL, Carlsbad, CA, USA). First-strand cDNA was synthesized and polymerase chain reaction (PCR) was performed with specific primers (Table 1, NWG Biotech, High Point, NC, USA) using AmpliTaq Gold reagents (PE Applied Biosystems, Foster City, CA, USA). For single cell gene amplification, secondary PCR was performed using 1 μl of first PCR product. All PCR products were analysed on 2% agarose gels and visualized by ethidium bromide fluorescence. The identity of PCR amplification products obtained were confirmed by DNA sequencing. Each primer pair was designed to span at least one intron, which controlled for genomic DNA contamination in the source RNA.

Table 1.

PCR primer pairs

Official symbol for mRNA	Official name for encoded protein	GenBank accession no.	Primer sequence (5′ to 3′)	Binding position
GAPDH	Mus musculus similar to glyceraldehyde-3-phosphate dehydrogenase (GAPDH)	XM_001004821	S-GTCTTCACCACCATGGAGA A-AAGCAGTTGGTGGTGCAG	S–823-841 A–975-992
Myh11	Myosin, heavy polypeptide 11, Smooth muscle	NM_013607	S-GAGAAAGGAAACACCAAGGTCAAGC, A-AACAAATGAAGCCTCGTTTCCTCTC	S–5721-5745, A–5930-5954
Kit*	c-kit†	NM_021099	S-CGCCTGCCGAAATGTATGACG A-GGTTCTCTGGGTTGGGGTTGC	S–2706-2726 A–2847-2867
Uchl1	Ubiquitin carboxy-terminal hydrolase L1 or protein gene product 9.5 (PGP 9.5)	NM_011670	S-CGATGGAGATTAACCCCGAGATG A-TTTTCATGCTGGCCGTGAG	S–22-44 A–171-190

^*Kit refers to the mRNA which was previously termed c-kit. This terminology has been changed in accordance with the NCBI database.

^†c-kit refers to the protein product which was previously termed Kit. This terminology has been changed in accordance with the NCBI database. S, Sense; A, Antisense. MHC primers also detected a 39bp smaller splice variant of the MHC gene (BC026142).

Results

Basal currents recorded from antral smooth muscle cells

Whole cell patch clamp recordings were performed on freshly dispersed antral myocytes using the perforated patch technique. Antral smooth muscle cells were initially identified under an inverted microscope as phase-bright elongated cells with a spindle-shaped morphology. Basal currents were recorded at a holding potential of −70 mV using intracellular solutions which contained either 145 mm or 10 mm Cl⁻ (E_Cl= 0 or −70 mV, respectively). Sustained periods at this holding potential (20–30 min) resulted in no detectible spontaneous transient inward current (STIC) activation. Ramping membrane potential from −100 mV to +100 mV repetitively (5 × 400 ms ramps applied every 1 s; Fig. 1A) activated Ca²⁺ currents in response to each depolarization (Fig. 1B and B′). Simultaneous imaging of cells loaded with Fluo-4 detected an increase in [Ca²⁺]_i associated with the inward current and summation of Ca²⁺ transients in response to multiple ramps (Fig. 1C). Increasing [Ca²⁺]_i did not initiate an inward current or affect the basal currents at the −70 mV holding potential (Fig. 1B). After recordings of currents and Ca²⁺ transients, cells were collected into patch pipettes and single-cell RT-PCR performed using primers specifically designed to detect specific phenotypes: Myh11 for smooth muscle cells, Kit for ICC, and Uchl1 for enteric neurons. Amplification of the housekeeping gene, GAPDH, was also performed to test the quality of the cDNA. Transcripts for these cell specific markers were also amplified from cDNA prepared from whole antral tissues (mucosa and submucosa removed) as a positive control. Non-template control (NTC) experiments were also included as a negative control to test for primer contamination. Amplified products of GAPDH, Myh11, Kit and Uchl1 were detected in gastric antral tissues but not in non-template controls (NTC). Patched cells with an elongated spindle shaped morphology were Myh11- and GAPDH-positive, but did not demonstrate Kit and Uchl1 expression, indicating that these cells were smooth muscle cells (Fig. 1D).

Figure 1. Basal currents recorded from antral smooth muscle cells.

A, a representative basal current recorded from an gastric antral myocyte held at −70 mV. During the recording, ramp protocols (5 × 400 ms) were applied twice. The first 5 trains of ramps are shown on faster time scale in B and B ′ along with simultaneous measurements of intracellular Ca²⁺ (C). B and C are shown on the same time scale. Ca²⁺ currents were recorded in response to ramp depolarizations (B′), and Ca²⁺ imaging confirmed an increase in [Ca²⁺]_i. Increases in [Ca²⁺]_i did not initiate or inhibit basal currents in antral myocytes. After electrophysiological recordings, single cell RT-PCR detected the expression of Myh11 and GAPDH but not Kit or Uchl1, indicating the cell had a smooth muscle phenotype. Signals below 100 bp are primer dimers due to the number of amplification steps (D).

Membrane depolarization increases [Ca²⁺]_i and inhibits basal currents in a subpopulation of Kit-positive cells

Enzymatic dispersions of gastric antral muscles also contained a subpopulation of cells with a morphology that was distinct from smooth muscle cells. These cells were often stellate in shape with processes that extended from the central body and had a characteristic phase-dark appearance (Langton et al. 1989; Sanders, 1996). Basal currents were also recorded from a holding potential of −70 mV using an intracellular solution containing either high (145 mm) or low (10 mm) Cl⁻ concentrations (Fig. 2A). Cells exhibited large basal currents (−391 ± 40 pA; n = 15) and occasional bursts of STIC activity when recorded under conditions with symmetrical Cl⁻ concentrations. Similar to that observed in antral myocytes, ramp depolarizations (5×, 400 ms ramps from −100 to 100 mV) caused an increase in [Ca²⁺]_i (Fig. 2B). However unlike antral myocytes, membrane depolarizations decreased basal currents by an average of 64 ± 7 pA (Fig. 2A, n = 15; P < 0.0001). Basal currents often remained suppressed if [Ca²⁺]_i remained slightly elevated following membrane depolarization. The amplitudes of the [Ca²⁺]_i-inhibited current (69 ± 12 pA) and basal current (−359 ± 45 pA) were not significantly different when E_Cl was set at 0 mV (n = 3). These data indicate that the basal current expressed in this population of cells was not carried by Cl⁻ ions.

Figure 2. Increases in [Ca²⁺]_i inhibits a basal current in a subpopulation of Kit-positive cells.

Voltage clamp experiments performed on a population of Kit-positive cells detected large basal currents and low input resistances at a holding potential of −70 mV (A). During basal current recordings, 5 trains of ramp depolarizations (400 ms duration) were applied. Unlike antral myocytes, depolarization-induced increases in [Ca²⁺]_i inhibited or decreased basal currents (B). The basal current remained suppressed as [Ca²⁺]_i did not return to the prestimulus level. The amplitudes of [Ca²⁺]_i-inhibited current and basal current were not significantly different when E_Cl was changed from −70 mV to 0 mV. After the recordings, cells were confirmed as ICC using single-cell RT-PCR (C). Amplification signals below 100 bp are the consequence of primer dimerization. The [Ca²⁺]_i-inhibited current recorded in CaPSS solution (Da) was abolished when the solution was perfused with BaPSS (Db). The calmodulin inhibitor calmidizolium (CMZ; 2 μm) also blocked the calcium-inhibited current in this population of ICC (E).

After recordings were performed, cells were collected for molecular studies using primers that were designed to identify the cell phenotype. Single cell RT-PCR revealed that these cells were Kit- and GAPDH-positive, but negative for Myh11 and Uchl1 (Fig. 2C). Kit-positive cells within the tunica muscularis include ICC and mast cells. However the latter are few in number in the tunica muscularis of healthy mice and have a small rounded appearance (Nissinen & Panula, 1995) that distinguishes them from the phase-dark spindle-shaped cells that recordings were obtained from in the present study.

In order to determine if the inhibition of the basal currents expressed in this population of Kit-positive cells was due to membrane depolarization or depolarization-induced increases in [Ca²⁺]_i via Ca²⁺ influx, basal currents were monitored during ramp depolarizations in control CaPSS and with BaPSS. Under symmetrical Cl⁻ solutions, basal currents with an average amplitude of −441 ± 41 pA were recorded in CaPSS solution and ramp depolarizations inhibited these currents by 68 ± 9 pA (Fig. 2Da). When these cells were perfused in BaPSS, ramp depolarizations did not cause significant inhibition of basal currents (6 ± 3 pA, n = 4, Fig. 2Db; P < 0.0007 when compared to inhibition of the current in CaPSS), suggesting that membrane depolarization did not directly inhibit basal currents but depolarization-induced Ca²⁺ influx caused an increase in [Ca²⁺]_i which inhibited basal conductances expressed in this population of Kit-positive cells. Calmidazolium (2 μm) reduced the inhibition of basal currents by ramp depolarizations indicating that Ca²⁺–CaM regulates the basal currents expressed in this population of Kit-positive cells (Fig. 2Ea and b). These data suggest that membrane depolarization causes an influx in Ca²⁺ which inhibits basal currents, in a Ca²⁺–CaM-dependent manner.

Reduced extracellular Na⁺ decreases basal currents and Ca²⁺-inhibited currents in Kit-positive cells

Since basal currents expressed in this population of Kit-positive cells were not carried by Cl⁻ ions, the cation permeability of the basal conductance was investigated by reducing [Na⁺]_o. Under symmetrical Cl⁻ conditions, basal currents (−381 ± 48 pA, n = 6) were recorded in high Na⁺ containing control CaPSS (Fig. 3A), and basal currents were inhibited (63 ± 9 pA, n = 6) in response to increases in [Ca²⁺]_i produced by ramp depolarizations (Fig 3Aa). During basal current recordings, extracellular Na⁺ was reduced from 135 mm to 50 mm by replacing Na⁺ (85 mm) with equimolar TEA. Under these conditions, basal currents were reduced from −381 ± 48 pA to −69 ± 11 pA (n = 4, P < 0.008, Fig. 3A) and Ca²⁺-inhibited currents in response to ramp depolarizations (Figs 3Ab) were also reduced from 63 ± 9 pA to 12 ± 2 pA (P < 0.01). The reversal potential (the average current response from 5 trains of ramp depolarizations) of the basal currents in control solution was 2 ± 2 mV (Fig. 3Aa and B). When the extracellular Na⁺ was reduced to 50 mm, the reversal potential of the basal current shifted to more negative potentials (−9 ± 3 mV; Fig. 3Ab and B; P < 0.006) indicating that a sodium permeable current was responsible for the basal current as well as the [Ca²⁺]_i-inhibited current in this group of Kit-positive cells. The average amplitude of depolarization-induced inhibition of basal currents in Kit-positive cells was significantly different from the amplitude of current responses in Myh11 or Uchil1 positive cells (P < 0.0001). Cells that express [Ca²⁺]_i-inhibited currents were predominantly Kit-positive, and cells that displayed [Ca²⁺]_i-inhibited currents of 52 pA (average current amplitude) or greater were treated as Kit-positive ICC.

Figure 3. Low extracellular Na⁺ decreases the amplitude of basal currents and [Ca²⁺]_i-inhibited currents in a subpopulation of Kit-positive cells.

Inhibition of basal currents due to an increase in [Ca²⁺]_i was observed using an intracellular solution which contained symmetrical Cl⁻ (A). During the recording, trains of ramp depolarizations (400 ms duration) were applied 5 times (Aa). Aa, the inhibition of basal current by an increase in [Ca²⁺]_i. During recordings of basal current, extracellular Na⁺ was reduced to 50 mm. Under this condition, the basal and [Ca²⁺]_i-inhibited currents were reduced (Ab). B, the voltage-dependent responses of the currents recorded in A. The reversal potential of the basal currents recorded in normal sodium (Aa) shifted to more negative potentials in low extracellular Na⁺ solution (Ab).

Basal currents recorded from a second subpopulation of Kit-positive cells

As stated in the introduction, there are two populations of ICC in the murine gastric antrum and enzymatic dispersions yielded two populations of Kit-positive cells. The second population of Kit-positive cells was spindle shaped with a phase-dark appearance with a phase-dark appearance. This sub-population of ICC exhibited distinct electrical properties that distinguished them from the first group of Kit expressing cells described in the previous sections. These Kit-positive cells displayed noisy basal currents. Spontaneous transient inward currents (STICs) also generated by these cells (approximately 60% of cells) were occasionally superimposed on the noisy basal currents (Fig. 4Aa, E_Cl= 0 mV; and Fig. 4Ba, E_Cl=−70 mV). Figure 4Ab and Bb shows an amplitude histogram generated from basal current recordings (1–3 min) from the same cells shown in Fig. 4Aa and Ba, respectively. Basal currents and STICs were qualitatively very similar when they were recorded when E_Cl was set at 0 or −70 mV. For the qualitative analysis of basal currents (including STICs), the width of the amplitude histogram at 80% from the peak was measured. The averaged values of the amplitude histogram analysis obtained from basal currents recorded under symmetrical Cl⁻ (12 ± 2, n = 24) or low [Cl⁻]_i (12 ± 2, n = 19) conditions were not significantly different. This analysis suggests that STICs were not due to the activation of a Cl⁻ conductance. Single-cell RT-PCR performed after patch clamp experiments revealed that these cells were Kit- and GAPDH-positive, but negative for Myh11 and Uchl1 (Fig. 4C).

Figure 4. Basal currents recorded from a subpopulation of Kit-positive cells.

Aa and Ba, representative currents recorded from Kit-positive ICC held at −70 mV when E_Cl was set at 0 mV and −70 mV, respectively. Unlike the first population of Kit-positive ICC, these cells expressed noisy basal currents and more prominent STICs. Ab and Bb, amplitude histograms generated from basal current recordings shown in Aa and Ba, respectively. For the quantitative measurements of basal currents and STICs, the width of the amplitude histogram at 80% from the peak was measured. The averaged value of the amplitude histogram analysis was not significantly different between cells at the two different E_Cl conditions. After electrophysiological recordings, single-cell RT-PCR detected the expression of Kit but not the expression of Myh11 or Uchil1 (C). Signals below 100 bp are a result of amplification primer dimerization.

Properties of the STICs in Kit-positive cells

The properties of STICs were further examined using symmetrical 140 mm Cl⁻ solutions at a holding potential of −70 mV (E_Cl= 0 mV). In order to determine their reversal potential, currents were monitored at various potentials, from −80 to +40 mV in 20 mV increments. Representative STICs recording at each potential, are shown in Fig. 5A. Amplitude histograms (Fig. 5B) were also generated from basal current recordings at each potential from the same cell shown in Fig. 5A. The average current value at 80% from the peak of the amplitude histograms plotted against each potential is shown in Fig. 5C (n = 4). This analysis indicated that the degree of membrane current fluctuation was reduced when cells were depolarized toward 0 mV from −80 mV and increased again when cells were further depolarized towards +40 mV. The data also showed that the reversal potentials of the basal current and STICs was approximately 0 mV. The amplitude histogram analysis also indicated that STICs were reduced from 9 ± 1 to 4 ± 1 pA (n = 4, P < 0.005) by application of Gd³⁺ (50 μm; Fig. 5D), suggesting these spontaneous events resulted from the periodic activation of a NSCC.

Figure 5. Properties of STICs recorded from a subpopulation of Kit-positive cells.

A, the voltage dependence of STICs recorded between −80 mV to +40 mV in 20 mV increments. B, amplitude histogram generated from basal current recordings at −80, −60, −20 and 0 mV from the same cell. The averaged value at 80% from the peak of the amplitude histograms was plotted against each potential (C). Fluctuations in basal currents decreased towards 0 mV from larger amplitudes at more hyperpolarized or depolarized potentials. Bath application of Gd³⁺ (50 μm) significantly reduced STIC activity (D).

Increases in [Ca²⁺]_i facilitates an inward current in Kit-positive cells

The effects of intracellular Ca²⁺ on basal currents generated in the second population of Kit-positive ICC were investigated using trains of five ramp depolarizations delivered every second (−80 mV to +50 mV for 400 ms) from a holding potential of −70 mV. Basal current was monitored under conditions where Cl⁻ equilibrium potentials were set at −70 mV (Fig. 6Aa) or 0 mV (Fig. 6B). During ramp depolarizations two distinct components of the Ca²⁺ current were observed: a low threshold Ca²⁺-activated current (which peaked between −35 and ∼−40 mV) and high threshold-activated Ca²⁺ current (which peaked at approximately 0 mV, Fig. 6Aa′), as previously reported in ICC (Lee & Sanders, 1993; Kim et al. 2002). Simultaneous analysis of [Ca²⁺]_i using Fluo 4-AM demonstrated that the ramp depolarizations also caused an increase in [Ca²⁺]_i (Fig. 6Ab). Unlike the first population of Kit-positive ICC, ramp depolarizations activated a noisy inward current which remained activated for several seconds and occurred whether E_Cl was 0 mV (182 ± 50 pA, n = 15) or −70 mV (234 ± 36 pA, n = 17; P > 0.05). Representative currents in response to the first and fifth ramp potentials are shown in Fig. 6C. Current responses increased in amplitude from the 1st to the 5th ramp as successive depolarizations were applied. Subtraction of the first from the fifth current trace demonstrated the facilitation of the inward current in this type of Kit-positive ICC (Fig. 6D). Depolarization-induced currents reversed at −8 ± 3 mV when E_Cl was 0 mV (n = 5) and −13 ± 2 mV when E_Cl was −70 mV (n = 14). There was no statistical differences in the amplitude (P > 0.4) or the reversal potentials (P > 0.1) of the depolarization-induced currents at either E_Cl indicating the currents expressed in the second population of Kit-positive cells were not carried by Cl⁻ ions. Figure 6E shows the amplitude of the [Ca²⁺]_i-facilitated current plotted against the number of Kit-positive cells that revealed an activated basal inward current in response to membrane depolarization. The currents activated in this subpopulation of Kit-positive cells ranged between −20 and −544 pA with an average amplitude of −188 ± 33 pA (n = 23). The amplitude of this current was significantly different from the amplitude of current responses in Myh11- or Uchil1-positive cells (P < 0.0001) (Fig. 6E).

Figure 6. Increasing [Ca²⁺]_i facilitates an inward current in Kit-positive cells.

Aa, currents activated by 5 trains of ramp protocols (−80 mV to +50 mV; 400 ms duration) applied to the second subpopulation of Kit-positive ICC at a holding potential of −70 mV (E_Cl=−70 mV). Current responses to the ramp depolarizations are shown on faster time scales in Aa′. Unlike smooth muscle cells, these Kit-positive cells exhibited low and high voltage-activated Ca²⁺ conductances. Simultaneous Ca²⁺ imaging using Fluor-4 detected an increase in [Ca²⁺]_i during the ramp depolarization (Ab). The depolarization induced an increase in [Ca²⁺]_i-facilitated inward currents (Aa). [Ca²⁺]_i-facilitated inward currents were also observed using an intracellular solution which contained symmetrical 140 mm Cl⁻ (B). C, current responses to the 1st and the 5th ramp depolarizations. D, the Ca²⁺-facilitated current (obtained by subtracting the 1st from the 5th current obtained in response to the ramp depolarizations obtained in C). E, the number of Kit-positive cells which exhibited facilitation of basal currents in response to membrane depolarization, plotted against the amplitude of the [Ca²⁺]_i-facilitated current. Unlike antral myocytes or enteric neurons, Kit-positive cells (23 cells) tended to facilitate larger inward currents.

[Ca²⁺]_i-facilitated basal currents are regulated by calmodulin in a subpopulation of Kit-positive ICC

In order to investigate whether the activation of basal currents expressed in the second population of Kit-positive cells was directly due to a change in membrane potential or indirectly due to a depolarization-induced increase in [Ca²⁺]_i via Ca²⁺ influx, basal currents were monitored during application of ramp depolarizations in control CaPSS and in BaPSS. Again, under symmetrical Cl⁻ solution conditions, noisy basal currents with an average amplitude of −68 ± 13 pA were recorded in CaPSS solution. Since low and high voltage-activated Ca²⁺ conductances were recorded during ramp depolarizations in this population of Kit-positive ICC (Fig. 6Aa′), a double depolarization protocol was applied during basal current recordings at a holding potential of −70 mV described as follows. Cells were stepped to −40 mV from −70 mV for 400 ms before returning to −70 mV and subsequently to 0 mV for the same duration after a 1 s interval (Fig. 7B′). Currents recorded in response to this voltage step protocol are shown in Fig. 7B. In CaPSS, a small Ca²⁺ current mixed with the basal noisy current was recorded when cells were depolarized to −40 mV. When cells were stepped back to −70 mV, activation of an inward tail current was observed. Depolarization to 0 mV activated larger Ca²⁺ currents which were followed by larger amplitude and more sustained tail currents after repolarization back to −70 mV (Fig. 7Aa and B). The amplitude of the tail current following repolarization from 0 mV back to −70 mV averaged 313 ± 50 pA (n = 5). When CaPSS was substituted for BaPSS the noisy basal current was reduced to −35 ± 20 pA (Fig. 7Ab, P < 0.04). In BaPSS, the double step depolarization-induced tail currents were significantly reduced at both potentials but small tail currents averaging 68 ± 13 pA still persisted in BaPSS (Fig. 7Ab and B, P < 0.01). These observations suggest that Ca²⁺ influx via both low and high threshold voltage-activated Ca²⁺ channels may facilitate inward currents.

Figure 7. [Ca²⁺]_i-facilitated currents are regulated by calmodulin in the second subpopulation of Kit-positive cells.

Under low [Cl⁻]_i conditions, basal currents were recorded in CaPSS solution at a holding potential of −70 mV (A). A double depolarization protocol (B′) was applied during basal current recordings and current responses to this protocol are shown in B. Aa and B, tail currents (recorded at −70 mV) activated after depolarization to −40 and 0 mV. When CaPSS was substituted for BaPSS the noisy basal current was reduced (Ab and B). In BaPSS the double step depolarization-induced tail currents were significantly reduced (Ab and B). CMZ at 2 μm reduced the [Ca²⁺]_i-facilitated currents generated in response to ramp depolarizations (Cb compared to Ca).

The dependence of this [Ca²⁺]_i-facilitated current on calmodulin (CaM) was investigated using the CaM inhibitor, calmidazolium (CMZ). The amplitude of [Ca²⁺]_i-facilitated currents was significantly reduced from 261 ± 51 pA to 53 ± 5 pA when CMZ (2 μm) was added to the CaPSS solution (Fig. 7Ca and b, n = 4, P < 0.03). These data suggest that the depolarization-induced increase in [Ca²⁺]_i activates a Ca²⁺–CaM-dependent inward conductance in this population of ICC.

Low extracellular Na⁺ decreases the amplitude of STICs and basal currents

Since basal currents and STICs expressed in this second population of Kit-positive cells were not carried by Cl⁻ ions, cation permeability of these conductances was investigated by reducing [Na⁺]_o. At a holding potential of −70 mV (E_Cl was set to −70 mV), basal currents and depolarization-induced, [Ca²⁺]_i-facilitated currents were both recorded (Fig. 8A). During current recordings, [Na⁺]_o was reduced to 70 mm (Figs 8Ab) by replacing 65 mm Na⁺ with equimolar TEA. The amplitudes of the basal current and the [Ca²⁺]_i-facilitated current were reduced from −116 ± 42 pA and 301 ± 70 (Fig. 8Aa) to −49 ± 24 (n = 7, P < 0.018) and 121 ± 38 (n = 4, P < 0.02), respectively, in 70 mm[Na⁺]_o solutions (Figs 8Ab). Amplitude histogram analysis of the basal noisy currents (width of amplitude histogram obtained at 80% from peaks) indicated that membrane current fluctuations including STIC activities were reduced in 70 mm[Na⁺]_o (from 18 ± 5 to 11 ± 3, n = 5, P < 0.05 data not shown). The averaged reversal potentials of the net whole-cell currents (i.e. net inward current) recorded on the fifth ramp depolarization was shifted to the left by 8 ± 3 mV in 70 mm[Na⁺]_o solution (P < 0.007). These data indicate that a sodium permeable current is responsible for the basal current, STICs and [Ca²⁺]_i-facilitated currents expressed in this population of Kit-positive cells.

Figure 8. Low extracellular Na⁺ decreases the amplitude of STICs and basal currents generated in Kit-positive ICC.

A, basal currents and STICs recorded at a holding potential of −70 mV using intracellular solutions which contained 10 mm Cl⁻ (E_Cl=−70 mV). Ramp depolarizations (at points a and b) induced an increase in [Ca²⁺]_i-facilitated inward currents in this cell. Aa, [Ca²⁺]_i-facilitated currents recorded under control conditions on a faster time scale. During the basal current recording, extracellular Na⁺ was reduced to 70 mm (A). STICs, basal current and the amplitude of [Ca²⁺]_i-facilitated currents were inhibited in reduced extracellular Na⁺ (A and Ab). Reduction in extracellular Na⁺ to 70 mm caused a shift of the reversal potential of basal currents to the left (B; currents taken during the 5th ramp depolarization).

[Ca²⁺]_i-facilitated non-selective cation currents are blocked by Cl⁻channel antagonists

Unitary potentials and their summation into regenerative potentials are inhibited when [Ca²⁺]_i is buffered with BAPTA-AM (Edwards et al. 1999; Hirst et al. 2002b) and therefore appear to need an increase in [Ca²⁺]_i for their activation. It has been suggested that the rise in [Ca²⁺]_i in antral tissues activates a calcium-dependent Cl⁻ conductance as both unitary potentials and regenerative potentials are inhibited by known Cl⁻ channel blockers, such as DIDS (Hirst et al. 2002b). Therefore we sought to determine the effects of these compounds on the [Ca²⁺]_i-facilitated current expressed in the second population of Kit-positive ICC. The double depolarization protocol described above was utilized (see Fig. 9B′). This protocol was applied every 30 s in the absence and presence of DIDS (100 μm). Under control conditions depolarization-induced, [Ca²⁺]_i-facilitated tail currents were revealed during the repolarization from − 40 and 0 mV to the − 70 mV holding potential (Fig. 9Aa and B). Under control conditions, depolarizations to 0 mV activated tail currents that averaged 243 ± 45 pA from a basal current of −67 ± 39 pA. In the presence of DIDS, tail current amplitude decreased to 51 ± 6 pA (P < 0.02) and basal currents were also reduced to −14 ± 9 pA (P < 0.05; Fig. 9Ab and B; n = 4). In two experiments 9-AC and in two additional experiments niflumic acid also reduced [Ca²⁺]_i-facilitated currents (data not shown). These data demonstrate that Cl⁻ channel blockers are not selective in their inhibition of currents expressed in Kit-positive ICC.

Figure 9. [Ca²⁺]_i-facilitated NSCC are blocked by the Cl⁻ channel antagonist DIDS.

Under conditions when E_Cl was set to −70 mV, basal currents and [Ca²⁺]_i-facilitated currents were recorded at a holding potential of −70 mV (A) and in response to a double depolarization protocol (B′). The current responses to this double depolarization protocol are shown in B. Aa and B, tail currents (recorded at −70 mV) activated after membrane depolarization to −40 and 0 mV. When DIDS was applied to the bath solution, the noisy basal current and amplitude of [Ca²⁺]_i-facilitated tail currents was reduced (Ab and B).

Discussion

In the present study we utilized the patch clamp technique and single cell reverse transcriptase polymerase chain reaction to identify Kit-positive ICC in enzymatic dispersions of the murine gastric antrum. We found two different populations of Kit-positive cells based on their electrical characteristics. The two classes of ICC had fundamentally different resting basal ionic conductances which displayed opposite responses to changes in intracellular calcium.

The first population of Kit-positive ICC expressed large basal currents. During current recordings occasional bursts of STIC activity were recorded in these Kit-positive ICC with an irregular low frequency. The basal current expressed in this population of Kit-positive ICC was inhibited when [Ca²⁺]_i was increased in a calmodulin-dependent manner in response to membrane depolarization. The Ca²⁺-sensitive basal conductances expressed in the subgroup of ICC were identified as a NSCC based on their reversal potentials and their permeability to Na⁺. The possibility of the current being carried by Cl⁻ ions was ruled out, as modification of E_Cl did not affect either the basal conductances or [Ca²⁺]_i-inhibited currents expressed in this population of ICC. The actual Cl⁻ concentration within ICC will, however, depend upon the rate of dialysis of Cl⁻ ions into the cell versus accumulation of intracellular Cl⁻ via plasmalemmal pumps.

A NSCC similar to that observed in murine gastric antral ICC observed in the present study has been identified in ICC cultured from the murine small intestine (Koh et al. 2002). In acute cultures (2–3 days), these cells grew into networks resembling ICC-MY and generated STICs that produced voltage oscillations which were similar to the slow wave activity recorded from intact tissues (Koh et al. 1998; Thomsen et al. 1998). Cell dialysis of ICC with 10 mm BAPTA activated a NSCC that was reduced in low extracellular Na⁺ solutions and in high [Ca²⁺]_i. On cell patch recordings from these ICC revealed periodic clusters of channel openings that generated distinct inward currents whose average frequency approximated that of spontaneous pacemaker currents recorded in these ICC. The channel clusters which were activated with BAPTA dialysis had a unitary conductance of 13 pS and reversed around 0 mV. Typical oscillations in membrane potential or current were not regularly observed in single freshly dispersed ICC from the gastric antrum during the present study. This was likely to be a consequence of recordings being performed on single isolated cells or that the pacemaker current was already activated and contributing to the basal current observed in these cells. Activation of pacemaker currents in cultured ICC from the small intestine with BAPTA was shown to cause a summation of spontaneous currents into a sustained large inward current (Koh et al. 2002). Thus the basal current observed in this population of antral Kit-positive ICC may be similar to the sustained summation of pacemaker currents observed in ICC from the small intestine following BAPTA treatment.

One of the current hypotheses that describes pacemaker mechanisms in gastrointestinal muscles depicts pacemaker currents that are activated by [Ca²⁺]_i handling between ‘pacemaker units’ that consist of the endoplasmic reticulum (ER), mitochondria and a finite cytoplasmic volume between the ER, mitochondria and the plasma membrane (Ward et al. 2000b; Suzuki et al. 2000; Malysz et al. 2001; Sanders et al. 2006). Activation of IP₃ type I receptors initiates calcium release from IP₃-sensitive stores (Suzuki et al. 2000; Ward et al. 2000b), leading to its uptake into the mitochondria through a mitochondrial calcium uniporter. Mitochondrial uptake reduces the calcium concentration in the anatomically defined space between the mitochondria and the plasma membrane (< 10 nm), which leads to the activation of a [Ca²⁺]_i-inhibited pacemaker conductance. In the present study, ramp depolarizations were utilized to increase [Ca²⁺]_i and this led to an inhibition of basal currents. Therefore this basal current in antral ICC may represent the [Ca²⁺]_i-inhibited current that has been described in the small intestine (Koh et al. 2002).

In contrast, the second subpopulation of Kit-positive ICC exhibited smaller but very noisy basal currents that often had STICs superimposed upon them. Unlike the first population of ICC, increases in [Ca²⁺]_i activated an inward current that was modulated in a calmodulin-dependent manner. This [Ca²⁺]_i-facilitated conductance and STICs were both NSCCs since they were permeable to Na⁺, reversed close to 0 mV and were inhibited by gadolinium. Cl⁻ was also ruled out as a charge carrier as there was no change in the current reversal potentials when E_Cl was modified over a 70 mV range.

In muscle bundles isolated from the gastric antrum, ICC-IM generate an ongoing discharge of membrane noise termed unitary potentials (Edwards et al. 1999; Suzuki et al. 2003). In the gastric antrum, these unitary potentials can summate into regenerative potentials when [Ca²⁺]_i is increased by depolarizing current injection or via stimulation of cholinergic nerves (Suzuki & Hirst, 1999; Hirst et al. 2002c; Beckett et al. 2004). We have speculated that the noisy basal currents and STICs expressed in this second population of Kit-positive cells are responsible for the unitary potentials recorded in gastric antral tissues and that the [Ca²⁺]_i-facilitated current, possibly through the summation of STICs, might be the conductance responsible for the generation of regenerative potentials. Thus, it is likely that this second population of ICC represents ICC-IM

There is considerable controversy in gastrointestinal tissues over what current is responsible for the generation of slow waves and/or regenerative potentials. It has been suggested that a rise in [Ca²⁺]_i in ICC activates a Ca²⁺-activated Cl⁻ conductance that is responsible for slow wave generation (Edwards et al. 1999; Hirst et al. 2002c). This hypothesis was proposed based on intracellular recordings on multicellular preparations and is supported by the observations that DIDS, 9-AC, niflumic acid, and BAPTA-AM all inhibited unitary potentials or pacemaker activity in gastric antral tissues. Cl⁻ conductances have been reported in cultured ICC from the murine small intestine, including a high conductance, inwardly rectifying and a volume-activated Cl⁻ conductance, but their physiological roles are uncertain (Tokutomi et al. 1995; Huizinga et al. 2002). Ca²⁺-activated Cl⁻ channels, which were suggested as the conductance responsible for pacemaker potentials in ICC-MY and regenerative potentials in ICC-IM, have never been identified in patch clamp studies on ICC (Lee & Sanders, 1993; Kim et al. 2002; Goto et al. 2004). In the present study of freshly dispersed ICC, a Ca²⁺-activated Cl⁻ conductance was not detected, but prominent NSCCs were observed in all Kit positive cells. Cl⁻ channel blocking drugs have non-specific effects on a variety of ionic conductances (Dick et al. 1999; Welsh et al. 2000), and these blockers also inhibited the NSCC in cultured ICC (Koh et al. 1998, 2002). In the present study we found that DIDS, 9-AC and niflumic acid (data not shown) inhibited the [Ca²⁺]_i-facilitated NSCC in gastric antral ICC. These findings might explain why these drugs block unitary and regenerative potentials in intact muscle bundles, and the data obtained in the present study suggest that a Ca²⁺-activated Cl⁻ conductance is not expressed in antral ICC and is not involved in the spontaneous activity of ICC.

NSCC in both populations of ICC appear to be regulated by a Ca²⁺–CaM-dependent process providing an insight into the cellular mechanisms underlying the generation of pacemaker activity and unitary/regenerative potentials. Since their discovery, a number of investigators have suggested that transient receptor potential (TRP) channels are the molecular candidates of NSCC. TRPC 1–7 (Tang et al. 2001; Clapham, 2003; Zhu, 2005), TRPV1 (Numazaki et al. 2003; Rosenbaum et al. 2004), V4 (Strotmann et al. 2003) and V6 (Niemeyer et al. 2001; Lambers et al. 2004), along with TRPM2 (Tong et al. 2006) and M4 (Nilius et al. 2005) have been shown to bind with CaM and are positively (facilitated) and negatively (inhibited) regulated in a Ca²⁺–CaM-dependent manner. The Ca²⁺–CaM regulation of the NSCC observed in the present study suggests that TRP channels may be involved in the Ca²⁺–CaM-inhibited or -activated conductances expressed in gastric ICC.

It has recently been suggested that TRPM7 encodes for the NSCC responsible for pacemaker activity in acutely cultured ICC from the small intestine (Kim et al. 2005). TRPM7 is involved in Mg²⁺ homeostasis and is inhibited by [Mg²⁺]_i (Nadler et al. 2001). At the C-terminus, TRPM7 contains an α-kinase domain (chaK1), which negatively regulates channel activity when [Mg²⁺]_i increases (Schmitz et al. 2003). At physiological concentrations of Ca²⁺, only Mg²⁺ and not other divalents can directly modulate TRPM7/ChaK1 kinase (Ryazanova et al. 2004). At high concentrations of [Ca²⁺]_i (> 100 μm), CaM (5 μm) can inhibit ChaK1 phosphotransferase activity. Binding of CaM to its substrates or competitive inhibition of the α-kinase domain by CaM was suggested to inhibit the phosphotransferase activity (Ryazanova et al. 2004). Although the CaM binding sequence on TRPM7 has not been identified, it is known that Ca²⁺–CaM inhibits ChaK1 and therefore it is possible that Ca²⁺–CaM could enhance TRPM7 activity. Since pacemaking was activated by a reduction in [Ca²⁺]_i in ICC cultured from the small intestine (Koh et al. 2002), it is unlikely that TRPM7 activity is responsible for this pacemaker activity.

In summary, the present study has described two distinct types of NSCC in different populations of Kit-positive ICC from the murine gastric antrum. Both conductances were sensitive to [Ca²⁺]_i. A [Ca²⁺]_i-inhibited conductance expressed in one population of ICC was similar to the conductance previously described in cultured ICC from the small intestine. A conductance of this type appears to be responsible for spontaneous pacemaker activity in the small bowel and may serve the same function in the stomach. A Ca²⁺-facilitated conductance was also observed in a second population of ICC that typically generated a continuous discharge of transient inward currents. This conductance may be responsible for the generation of unitary potentials that can summate into regenerative potentials which contribute to the overall excitability of the gastric antrum. At the present time, due to lack of specific cellular markers, it is not possible to firmly assign the Ca²⁺-dependent NSCC we observed to specific ICC, either ICC-MY or ICC-IM.