Pacemaker function and neural responsiveness of subserosal interstitial cells of Cajal in the mouse colon

Abstract

Much is known about myogenic mechanisms in circular muscle (CM) in the GI tract, but less is known about longitudinal muscle (LM). Two Ca²⁺ signalling behaviors occur in LM, localized intracellular waves not causing contractions and intercellular waves leading to excitation-contraction (E-C) coupling. An Ano1 channel antagonist inhibited intercellular Ca²⁺ waves and LM contractions. Ano1 channels are expressed by interstitial cells of Cajal (ICC) but not by smooth muscle cells (SMCs). We investigated Ca²⁺ signalling in a novel population of ICC that lies along the subserosal surface of LM (ICC-SS) in mice expressing GCaMP6f in ICC. ICC-SS fired stochastic localized Ca²⁺ transients. Such events have been linked to activation of Ano1 channels in ICC. Ca²⁺ transients in ICC-SS occurred by release from stores most likely via IP₃ receptors. This activity relied on influx via store-operated Ca²⁺ entry and Orai channels. No voltage-dependent mechanism that synchronized Ca²⁺ transients in a single cell or between cells was found. Nitrergic agonists inhibited Ca²⁺ transients in ICC-SS, and stimulation of intrinsic nerves activated nitrergic responses in ICC-SS. Cessation of stimulation resulted in significant enhancement of Ca²⁺ transients when compared to the pre-stimulus activity. No evidence of innervation by excitatory, cholinergic motor neurons was found. Our data suggest that ICC-SS contribute to regulation of LM motor activity. Spontaneous Ca²⁺ transients activate Ano1 channels in ICC-SS. Resulting depolarization conducts to SMCs, depolarizing membrane potential, activating L-type Ca²⁺ channels and initiating contraction. Rhythmic electrical and mechanical behaviors of LM are an emergent property of SMCs and ICC-SS.

Introduction

Circular muscle (CM) contractions contribute to propulsive contractions in the colon, but the role of longitudinal muscle (LM) is less well understood. Previous studies have shown that smooth muscle cells (SMCs) in LM are coupled to contraction via the occurrence of action potentials generated by activation of L-type (dihydropyridine-sensitive) Ca²⁺ channels (El‐Sharkawy, 1983; Chow & Huizinga, 1987; Smith et al., 1987a; Liu & Huizinga, 1993; Spencer et al., 2002). Spontaneous transient depolarizations (STDs) and action potentials occur spontaneously in LM strips devoid of myenteric ganglia and CM layers. Propagation of action potentials appears to be confined within bundles of cells running in the longitudinal axis, and propagation rarely occurs more than 1 mm in the circumferential direction (Spencer et al., 2002).

Intracellular Ca²⁺ transients occur spontaneously in longitudinal SMCs (LSMCs) due to Ca²⁺ release from stores (Hennig et al., 2002). These events are localized, asynchronous in adjacent cells, and are not coupled to contractile events. It is likely that the localized, intracellular transients couple to activation of large-conductance Ca²⁺-activated K⁺ (BK) channels, as in many types of SMCs (Wellman & Nelson, 2003), and BK channel antagonists have been shown to increase contractions and action potential firing in LSMCs (Thornbury et al., 1992; Carl et al., 1995). Intercellular Ca²⁺ waves in LSMC bundles have also been observed. These events are fast and of much greater amplitude than localized, intracellular transients, and likely result from action potential firing (Hennig et al., 2002). Intercellular Ca²⁺ waves, which along with action potentials can occur quite rhythmically (Spencer et al., 2002), are coupled to LM contractions. Whether an external pacemaker drives LSMC intercellular Ca²⁺ waves and contractions, as is the case for SMCs in the small bowel and stomach (Sanders et al., 2014), or whether LSMCs manifest intrinsic pacemaker capability are unanswered questions.

Previous morphological examinations have reported 4 distinct types of interstitial cells of Cajal (ICC) in colonic muscles of several species (Berezin et al., 1990; Christensen et al., 1992; Rumessen et al., 1993; Ishikawa & Komuro, 1996; Burns et al., 1997; Faussone-Pellegrini & Thuneberg, 1999; Aranishi et al., 2009; Blair et al., 2012a). A group of cells generating slow waves exists at the submucosal surface (ICC-SM) of the CM (Smith et al., 1987b; Serio et al., 1991; Yoneda et al., 2004); intramuscular ICC (ICC-IM) are intermingled with motor nerve terminals and are coupled electrically to SMCs and interstitial cells distinguished by expression of platelet-derived-growth-factor-receptor-alpha (PDGFRα, aka SIP syncytium; (Sanders et al., 2012). A network of multi-polar ICC lies in the myenteric region between the CM and the LM also displays pacemaker activity (ICC-MY (Smith et al., 1987b)); and a final population lies along the serosal surface (ICC-SS) of the LM (Toma et al., 1999; Vanderwinden et al., 2000; Aranishi et al., 2009; Blair et al., 2012a; Rumessen et al., 2013). ICC-SS were found both in taenia coli and in the inter-taenia regions in colons of cynomolgous monkeys (Blair et al., 2012a). It is known from these previous studies that ICC-SS are present along the subserosal surface of the proximal colon (Vanderwinden et al., 2000; Aranishi et al., 2009; Rumessen et al., 2013). Aside from their morphology and anatomical localization, virtually nothing is currently known about the physiological function or neural regulation of ICC-SS. We hypothesized that these cells participate in pacing and neural regulation of LM.

ICC-SS are stellate, multi-processed cells that are distributed within the connective tissue beneath the mesothelium. The cells are oriented along the axis of the LM muscle fibers in the proximal portion of murine, guinea-pig and primate colons, forming an electrical network via gap junctions between the cells (Burns et al., 1997; Vanderwinden et al., 2000; Aranishi et al., 2009) (Blair et al., 2012a). In the present study, mice were generated by crossing iCre-Kit mice with GCaMP6f^fl/fl mice to generate ICC-specific expression of the Ca²⁺ sensor, GCaMP6f. Video records of Ca²⁺ transients in ICC-SS were collected via fluorescence microscopy to better understand the role of these cells in regulating LSMCs. ICC-SS express the Ca²⁺-activated Cl⁻ conductance, Ano1, selectively (Gomez-Pinilla et al., 2009; Blair et al., 2012a), and evaluating the specific properties and mechanisms of Ca²⁺ transients, shown in other ICC to regulate the activation of Ano1 (Zhu et al., 2015; Baker et al., 2016; Drumm et al., 2017, 2018b, 2019b, 2019c), provides a novel means of investigating the role ICC-SS in LM motility.

Methods

Ethical Approval

All animals used and the protocols performed throughout this study were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. All procedures were approved by the Institutional Animal Use and Care Committee at the University of Nevada, Reno. All experiments conform to the principles and regulations for reporting animal experiments as described by Grundy (2015).

Animals

Ai95 (RCL-GCaMP6f)-D (GCaMP6f^fl/fl mice) and their wild-type siblings (C57BL/6) were purchased from the Jackson Laboratory (Bar Harbor, MN, USA). c-Kit^+/Cre-ERT2 (Kit-Cre mice) were gifted from Dr. Dieter Saur of the Technical University Munich, Germany. Mice expressing an inducible Cre Recombinase driven by the smooth muscle heavy chain promoter (SmMHC‐Cre mice) were generated by Dr. Stefan Offermann (Max‐ Planck‐ Institute, Germany) and gifted to our lab by Dr. James Stull of Southwestern University (TX, USA).

Tamoxifen preparation and administration

GCaMP6f^fl/fl mice were crossed with either Kit-Cre mice or SmHC-Cre mice, as outlined above, and the resulting offspring are referred to as Kit-Cre-GCaMP6f or SmHC-Cre-GCaMP6f mice throughout the manuscript. The mice were injected with tamoxifen at 6–8 weeks of age to induce Cre Recombinase and activate expression of GCaMP6f. Tamoxifen (Sigma T5648; 80mg) was dissolved in 800 μL of ethanol (Pharmco-Aaper 200 Proof - Absolute, Anhydrous) by vortexing for 20 minutes. Then 3.2 ml of Safflower (generic) was added to create solutions of 20 mg/ml, which were then sonicated for 30 minutes prior to injection.

Mice were injected (Intraperitoneal injection; IP) with 0.1 ml of tamoxifen solution (2 mg tamoxifen) for three consecutive days. Mice were used for experiments 10 days after the initial injection. Expression of GCaMP6f was confirmed by genotyping and imaging. Mice were anaesthetized by inhalation with isoflurane (Baxter, Deerfield, IL, USA) and killed by cervical dislocation before excision of tissues.

Tissue preparation

Abdomens were opened and the entire colon was removed and placed in Krebs-Ringer bicarbonate (KRB) solution. A segment of proximal colon was opened along the mesenteric border and intra-luminal contents were washed away with KRB solution. The mucosa and submucosa were removed by sharp dissection.

Calcium imaging

Colonic muscle strips were pinned with the LM layer facing upward to the bottom of a 60 mm dish coated with Sylgard elastomer (Dow Corning, Midland, MI). The dish was perfused with warmed KRB solution (37°C) for 60 mins prior to the beginning of experiments. Following this equilibration period, Ca²⁺ imaging was performed on cells in situ with an Eclipse E600FN microscope (Nikon Inc., Melville, NY, USA) equipped with a 60× 1.0 CFI Fluor lens (Nikon instruments INC, NY, USA). GCaMP6f was excited at 488 nm (T.I.L.L. Polychrome IV, Grafelfing, Germany). The pixel size using this acquisition configuration was 0.225 μm. Image sequences were collected at 33 fps with TILLvisION software (T.I.L.L. Photonics GmbH, Grafelfing, Germany). Movement artefacts were stabilized digitally with custom made Volumetry software prior to analysis of Ca²⁺ transients. For experiments involving pharmacological treatments, control video sequences were collected for 20–30 sec, and then KRB solution containing the drug concentration to be tested was perfused into the bath for 12–15 mins before another 20–30 sec period of imaging was performed. As reported previously, imaging GCaMP for 20–30 s of consecutive recordings did not lead to a decrease in Ca²⁺ transients (Drumm et al., 2018a).

Analysis of Ca²⁺ transients

Ca²⁺ release/entry events in colonic cells were imaged and analyzed as described previously (Drumm et al., 2019a). Briefly, movies of Ca²⁺ events were converted to a stack of TIFF (tagged image file format) images and imported into custom software (Volumetry G8c, GW Hennig) for initial processing. Whole-cell ROIs were created to generate spatio-temporal maps (STMs) of Ca²⁺ transients in individual cells within a FOV. These STMs were imported as TIFF files into Image J (version1.52a, National Institutes of Health, MD, USA, http://rsbweb.nih.gov/ij) for post hoc analysis. Basal fluorescence was acquired from regions of cells that displayed the most uniform and least intense fluorescence (F₀). Then fluorescence values throughout the rest of the cell were divided by the F₀ value to calibrate the STM for the amplitudes of Ca²⁺ transients as F/F₀. Ca²⁺ event amplitude, duration and spread were then calculated from the STM. Ca²⁺ transient frequency was expressed as the number of events fired per cell per minute (min⁻¹). The amplitude of Ca²⁺ transients was expressed as ΔF/F₀, the duration of Ca²⁺ transients was expressed as full duration at half maximum amplitude (FDHM) and the spatial spread of Ca²⁺ transients was expressed as μm of cell propagated per Ca²⁺ transient. Ca²⁺ transients were included in measurements if their peak amplitude reached > 20% of maximum during control period.

In some examples Ca²⁺ transients observed in colonic cells were quantified using particle (PTCL) analysis, as described previously (Drumm et al., 2017, 2018b, 2019b). Briefly, movies were firstly imported into Volumetry G8d and motion stabilized to minimize residual motion artifacts. A differential (Δt = ± 66–70 ms) and Gaussian filter (1.5 × 1.5μm, StdDev 1.0) was applied to accurately distinguish Ca²⁺ transients from the background. A particle analysis routine was applied by using of a flood-fill algorithm, which marked the structure of all adjoining pixels that had intensities above the threshold. Ca²⁺ transient PTCLs were brighter and larger than noise particles. The threshold at which noise particles emerged and reduced the average particle size was thresholded and then valid Ca²⁺ PTCLs, which were above this threshold, were then saved as a coordinate based PTCL movie. To identify Ca²⁺ firing sites, only those particles that did not overlap with any particles in the previous frame but overlapped with particles in the next 70 ms were considered Ca²⁺ firing sites.

Contractile recordings

LM muscle strips (5–8mm in length and ~2mm in diameter) were cut parallel to the LM fibers from the proximal colon of wildtype mice, attached to a stable mount and to a Gould strain gauge and immersed in jacketed 15 ml tissue baths. The LM muscles were continuously oxygenated and maintained in KRB solution at 37^oC. LM tissues were stretched to an initial tension of 5 mN, and experiments began after 60 mins of equilibration. Isometric contractions were recorded using AcqKnowledge software (3.9.1; Biopac Systems, Goleta, CA) and then their area under the curve (AUC), frequency, amplitude and duration were quantified using pClamp 9.0. Contractions at least 10% of maximum during control periods were measured for quantification.

Nerve stimulation

Neural responses were evoked either in Ca²⁺ imaging experiments or isometric tension recording experiments by electrical field stimulation (EFS). EFS was delivered by two parallel platinum electrodes on either side of a colon muscle sheet (imaging) or a LM strip (tension). EFS was evoked a by Grass stimulator S48 (Quincy MA, USA), at 0.3ms pulse duration and 100 V. Responses to EFS were abolished by tetrodotoxin (1 μM; data not shown).

Drugs and solutions

Tissues used for imaging and electrophysiological experiments were perfused with KRB solution containing (mmol/L): NaCl, 120.35; KCl, 5.9; NaHCO₃, 15.5; NaH₂PO₄, 1.2; MgCl₂, 1.2; CaCl₂, 2.5; and glucose, 11.5. KRB solution was bubbled with a mixture of 97% O₂ – 3% CO₂ and warmed to 37 ± 0.2 °C. For experiments using 0 mM [Ca²⁺]_o, CaCl₂ was excluded from the KRB solution and 0.5 mM EGTA was added, resulting in a Ca²⁺ free KRB solution. 2-aminoethyl-diphenylborinate (2-APB), atropine, caffeine, cyclopiazonic acid (CPA), L-NNA, pinacidil and nicardipine were purchased from Sigma-Aldrich (St Louis, MO, USA). Tetrodotoxin (TTX), Ani 9, thapsigargin, ODQ, ryanodine, SKF 96365, U73343 and U73122 were purchased from Tocris Bioscience (Ellisville, Missouri, USA). GSK 7975A was purchased from Aobious. Xestospongin C was purchased from Cayman Chemical. All drugs were dissolved as recommended by the manufacturers and then diluted to the desired concentrations with KRB solution.

Statistics

Data is represented as mean ± standard deviation (S.D.). Statistical analysis was performed using paired student’s t-tests (comparing two paired groups) or a one-way ANOVA with a Tukey post hoc test as appropriate (comparing three paired groups). In all statistical analyses, P<0.05 was taken as significant. Probabilities < 0.05 are represented by a single asterisk (*), probabilities < 0.01 are represented by two asterisks (**), probabilities < 0.001 are represented by three asterisks (***) and probabilities < 0.0001 are represented by four asterisks (****). Throughout the text, “n” refers to the number of animals included in datasets, and “c” refers to the number of individual cells analyzed within a given dataset.

Results

Ca²⁺ signaling patterns in LSMCs

We firstly examined basal Ca²⁺ signalling in LSMCs in proximal colons from SmHC-Cre-GCaMP6f mice by performing in situ imaging of Ca²⁺ transients. We confirmed two distinct patterns of Ca²⁺ signalling in LSMCs; intracellular Ca²⁺ waves and intercellular Ca²⁺ waves, as described previously (Hennig et al., 2002). Intracellular Ca²⁺ waves were imaged using a 60x objective (Fig. 1A–B), and occurred asynchronously between cells and were not associated with contraction of the LM. LSMC intracellular Ca²⁺ waves had an average frequency of 180.7 ± 89.6 min⁻¹, amplitude of 0.65 ± 0.47 ΔF/F₀, a FDHM of 171.1 ± 45 ms and an average spatial spread of 21.1 ± 12.35 μm (c=34, n=5). The distribution of all values for frequency, amplitude, duration and spatial spread of LSMC intracellular Ca²⁺ waves is displayed as histograms in Fig. 1C.

Fig. 1: LSMC intracellular Ca²⁺ waves.

A Raw image of LSMCs (left panel) recorded in situ from the proximal colon of a SmHC-Cre-GCaMP6f mouse (60x objective; scale bar pertains to all panels in A). The middle panel shows summed intracellular Ca²⁺ waves occurring during a 20 sec recording. The panel on the right shows the initiation sites from which intracellular Ca²⁺ waves originated from in this FOV. B Time-lapse images showing the occurrence of intracellular Ca²⁺ waves over 0.16 sec with the firing of these events indicated by the white arrows. C Frequency histograms showing the range of values of intracellular Ca²⁺ wave frequency, amplitude, duration and spatial spread in LSMC, c=34, n=5.

The spatio-temporal maps (STMs) in Fig. 2A demonstrate that the firing of LSMC intracellular Ca²⁺ waves did not rely on extracellular Ca²⁺ influx via L-type Ca²⁺ channels, as application of the Cav_1.2 channel antagonist, nicardipine (1 μM), had no effect on the frequency (P=0.94), amplitude (P=0.62), duration (P=0.08) or spatial spread (P=0.12) of the events (Fig. 2B, paired student t-tests, c=14, n=5). Similarly, incubation with a selective and potent antagonist of the Ca²⁺-activated-Cl⁻ channel Ano1, Ani 9 (Fig. 2C, 1 μM, a potent antagonist of Ano1 channels that is >18 times more potent than T16A_inh-A01 or MONNA (Seo et al., 2016)) also had no effect on LSMC intracellular Ca²⁺ wave frequency (P=0.8), amplitude (P=0.15), duration (P=0.98) or spatial spread (P=0.39) (Fig. 2D, paired student t-tests, n=10, n=5), suggesting that the generation of these Ca²⁺ events was not reliant on the activation of Ano1 channels in colonic ICC. However, all intracellular Ca²⁺ waves observed in LSMCs were abolished when intracellular Ca²⁺ stores were disrupted with the sarcoplasmic/endoplasmic reticulum (ER) Ca²⁺-ATPase (SERCA) pump inhibitor cyclopiazonic acid (CPA, 10 μM, Fig. 2E–F, P<0.0001, paired student t-test, c=10, n=5), suggesting, as previously described, that intracellular Ca²⁺ waves in LSMCs originate from Ca²⁺ release mechanisms in the ER.

Fig. 2: LSMC intracellular Ca²⁺ waves do not rely on voltage-dependent Ca²⁺ entry or Ano1 channels.

A STMs of intracellular Ca²⁺ waves in an LSMC recorded in situ, showing the effects of the Cav_1.2 channel antagonist, nicardipine (1 μM). B Summary effects of nicardipine on LSMC intracellular Ca²⁺ waves. Control values: Frequency: 181.3 ± 100.3 min⁻¹; Amplitude: 0.7 ± 0.41 ΔF/F₀; FDHM: 181.3 ± 30.7 ms; Spread: 21.85 ± 6.7 μm, c=14, n=5. Nicardipine values: Frequency: 182.4 ± 95.7 min⁻¹; Amplitude: 0.7 ± 0.5 ΔF/F₀; FDHM: 173.3 ± 26 ms; Spread: 19.4 ± 6 μm, c=14, n=5. C STMs of intracellular Ca²⁺ waves occurring in a LSMC recorded in situ, showing the effect of the Ano1 channel antagonist Ani 9 (1 μM). D Summary effects of Ani 9 on LSMC intracellular Ca²⁺ waves. Control values: Frequency: 156.6 ± 72.4 min⁻¹; Amplitude: 0.6 ± 0.3 ΔF/F₀; FDHM: 167.4 ± 20.9 ms; Spread: 20.67 ± 7.7 μm, c=10, n=5. Ani 9 values: Frequency: 151.8 ± 70.7 min⁻¹; Amplitude: 0.7 ± 0.4 ΔF/F₀; FDHM: 167.7 ± 30.3 ms; Spread: 23.1 ± 11.3 μm, c=10, n=5. E STMs of intracellular Ca²⁺ waves occurring in a LSMC recorded in situ, showing the effect of the SERCA pump inhibitor CPA (10 μM). F Summary effect of CPA on LSMC intracellular Ca²⁺ wave frequency, c=10, n=5.

The second pattern of Ca²⁺ signalling observed in LSMCs was intercellular Ca²⁺ waves. Intercellular Ca²⁺ waves spread rapidly from cell to cell, propagating across the LM when imaged at low power (10–20x, Fig. 3A). The intercellular Ca²⁺ waves occurred in rapid bursts, and were associated with contractions of the LM. STMs of these events could be drawn by reconstructing an x, t plot across the entire field of view (FOV) of the LM (Fig. 3B). In contrast to intracellular Ca²⁺ waves, intercellular Ca²⁺ waves were sensitive to nicardipine (1 μM, Fig. 3C), which reduced their frequency from 7.1 ± 2.1 to 1.7 ± 1.3 min⁻¹ (Fig. 3D, P=0.008, paired student t-test, n=4). These data suggested that intercellular Ca²⁺ waves and contractions of LM depended on Ca²⁺ influx via L-type Ca²⁺ channels. We tested the hypothesis that activation of Ano1 channels in ICC may be responsible for increased excitability of LSMCs, making it possible to reach the threshold for activation of L-type Ca²⁺ channels. Ani 9 (1 μM) significantly reduced the frequency of intercellular Ca²⁺ waves from 6.1 ± 1.7 to 2.1 ± 2.9 min⁻¹ (Fig. 3E–F, P=0.015, paired student t-test, n=7). Another approach to examine the role of ICC in generating LSMC intercellular Ca²⁺ waves was to inhibit store-operated-Ca²⁺ entry (SOCE) in ICC. SOCE has been demonstrated to be critical for the maintenance of pacemaker activity and stochastic Ca²⁺ release and Ano1 channel activation in other classes of ICC in the small intestine (Zheng et al., 2018) and colon (Drumm et al., 2019b). SOCE in ICC is mediated by interactions between the ER membrane protein STIM and the plasma membrane bound Ca²⁺ influx channels of the Orai family (Zheng et al., 2018). We inhibited SOCE in ICC using a selective antagonist of Orai channels, GSK 7975A, and found that at 1 μM LSMC intercellular Ca²⁺ waves were significantly reduced in frequency from 5.1 ± 2.1 to 2.6 ± 1.8 min⁻¹ (Fig. 3G–H, P=0.03, one way ANOVA, Tukey post hoc test, n=5). Further reduction in frequency was observed was observed with 10 μM GSK 7975A (i.e. to 1.4 ± 1.2 min⁻¹ (Fig. 3H, P=0.017, one way ANOVA, Tukey post hoc test, n=5).

Fig. 3: LSMC intercellular Ca²⁺ waves.

A Time-lapse images showing intercellular Ca²⁺ waves in the LM of the proximal colon of a SmHC-Cre-GCaMP6f mouse (20x objective; scale bar in the far left panel pertains to all panels). The white arrow shows the direction of the propagation of the intercellular Ca²⁺ wave in the far left panel. B STM of intercellular Ca²⁺ waves occurring in an LM sheet recorded in situ. C STMs of intercellular Ca²⁺ waves occurring in the LM recorded in situ and the inhibitory effect of nicardipine (1 μM). D Summary data showing the effect of nicardipine on the frequency of LSMC intercellular Ca²⁺ waves, n=4. E STMs of intercellular Ca²⁺ waves in LM recorded in situ and the effect of Ani 9 (1 μM). F Summary data showing the effect of Ani 9 on the frequency of LSMC intercellular Ca²⁺ waves, n=7. G STMs of intercellular Ca²⁺ waves in the LM recorded in situ, showing the effect of the Orai channel antagonist GSK 7975A (1 μM). H Summary data showing the effect of GSK 7975A (1 & 10 μM) on the frequency of LSMC intercellular Ca²⁺ waves, n=5.

We further investigated a role of ICC in generating normal LM contractions by performing isometric tension recordings of LM strips from the mouse proximal colon. Mouse colonic LM strips exhibited spontaneous rhythmic contractions (Fig. 4A) and the frequency of contractions was reduced from 2.2 ± 1.4 to 1.2 ± 1.4 min⁻¹ by Ani 9 (1 μM, Fig. 4A–B, P=0.02, paired student t-test, n=5). Ani 9 also reduced the amplitude of LM contractions from 2.2 ± 0.7 to 0.8 ± 0.8 mN (Fig. 4B, P= 0.0006, paired student t-test, n=5), and duration of contractions was reduced from 10.7 ± 8.8 to 2.3 ± 2.2 s (Fig. 4B, P=0.02, paired student t-test, n=5). GSK 7975A (1 μM) led to a reduction in LM contraction frequency from 2.9 ± 1.7 to 2.2 ± 1.7 min⁻¹ (Fig. 4C–D, P=0.0001, one way ANOVA, Tukey post hoc test, n=6) and this was further reduced to 1.9 ± 1.65 min⁻¹ by 10 μM GSK 7975A (Fig. 4C–D, P=0.0032, one way ANOVA, Tukey post hoc test, n=6). While there was a trend of a reduced amplitude and increased duration of LM contractions by GSK 7975A, these changes did not reach statistical significance (Fig. 4D, Amplitude: P=0.34 (1 μM), P=0.15 (10 μM), one way ANOVA, Tukey post hoc test, n=6; Duration: P=0.13 (1 μM), P=0.053 (10 μM), one way ANOVA, Tukey post hoc test, n=6). These observations suggest that Ca²⁺ store loading and Ano1 channels are required for maintenance of L-type Ca²⁺ channel activation and excitation-contraction coupling in LM.

Fig. 4: LM contractions rely on Ano1 channels and SOCE.

A Mouse LM contractions and the effects of Ani 9 (1 μM). B Summary data for the effects of Ani 9 on LM contraction frequency, amplitude and duration, n=5. C LM contractions and the effects of GSK 7975A (1 &10 μM). D Summary data for the effects of GSK 7975A on LM contraction frequency, amplitude and duration, n=6.

ICC-SS run in parallel with LM bundles and are electrically connected to LSMCs by gap junctions (Vanderwinden et al., 2000; Aranishi et al., 2009; Rumessen et al., 2013). Like other classes of ICC in the GI tract, ICC-SS display exclusive expression of Ano1 channels (Gomez-Pinilla et al., 2009; Blair et al., 2012a). Thus, ICC-SS are in a prime location and express a key ion channel that appears to regulate the excitability of the LM layer in the proximal colon. We hypothesized that if ICC-SS regulate LM activity through activation of Ano1 channels, they must exhibit intracellular Ca²⁺ signalling behaviours. We investigated this possibility by imaging Ca²⁺ signaling in ICC-SS using Kit-Cre-GCaMP6f mice.

Basal Ca²⁺ signaling in ICC-SS

In accordance with previous morphological studies (Vanderwinden et al., 2000; Aranishi et al., 2009; Blair et al., 2012a; Rumessen et al., 2013), we found that ICC-SS in the mouse colon were stellate in shape with a large ovoid soma and multiple processes (Fig. 5A). We imaged these cells along the serosal surface of the proximal colon from Kit-Cre-GCaMP6f mice. ICC-SS were spontaneously active, firing hundreds of Ca²⁺ transients per minute. Fig. 5A illustrates a typical example of ICC-SS in the proximal colon (left panel) with Ca²⁺ transients summated over a 20 sec recording (middle panel) and areas of initiation of Ca²⁺ transients (right panel). To analyze Ca²⁺ transients in ICC-SS, we employed spatio-temporal maps (STMs), as described previously (Drumm et al., 2019a). One such STM from a cell recorded in situ is shown in Fig. 5B. The STM is colour coded to illustrate more intense fluorescence as warm colours (red, orange) and less intense fluorescence as cold colours (blue, black). ICC-SS were highly active, expressing many Ca²⁺ firing sites, and each typically generating multiple transients throughout recording periods. Four such sites are indicated on the STM in Fig. 5B, and activities of each site are plotted as traces in Fig. 5C. The STM is also plotted as a 3D graph in Fig. 5D. From this example it is clear that ICC-SS exhibit Ca²⁺ transients from multiple initiation sites with a range of firing frequencies, amplitudes, durations and spatial spread. On average, an ICC-SS fired Ca²⁺ transients from at least 6 different sites during 20 sec recording periods, although some cells displayed as many as 13 firing sites during this time frame (Fig. 5E, c=165, n=33). The average firing frequency of Ca²⁺ transients in ICC-SS was relatively high, 149.5 ± 82 min⁻¹, although in some cases this could be as low as 6 min⁻¹ and as high as 510 min⁻¹ (Fig. 5E, c=165, n=33). The mean amplitude of Ca²⁺ transients in ICC-SS was 0.65 ± 0.5 ΔF/F₀, the average FDHM was 181.8 ± 53.1 ms and the average spatial spread was 12.7 ± 8.2 μm (Fig. 5E, c=165, n=33).

Fig. 5: ICC-SS exhibit dynamic Ca²⁺ signaling.

A Left panel shows ICC-SS in the proximal colon of a Kit-Cre-GCaMP6f mouse in situ (60x objective). The middle panel shows summed Ca²⁺ signals in the ICC-SS during a 20 sec recording. Right panel shows an accumulation map of Ca²⁺ transients and indicates sites of Ca²⁺ transient initiation during a 20 sec recording period. B STM of Ca²⁺ transients in a single ICC-SS recorded in situ. Four different firing sites are indicated and their activity plotted as traces in panel C. D 3-D representation of the STM shown in panel B. E Frequency histograms showing the range of values of ICC-SS Ca²⁺ transient firing site number, frequency, amplitude, duration and spatial spread, c=165, n=33.

Ca²⁺ transients in ICC-SS were not entrained, forming FOV-wide Ca²⁺ waves across multiple cells. We examined this further, as some have implied that Ca²⁺ transients in ICC must be coupled and entrained to achieve significant influence on GI motility (Van Helden et al., 2000; Imtiaz et al., 2002; van Helden & Imtiaz, 2003). Fig. 6A shows a FOV containing 5 ICC-SS (far left panel). All 5 cells exhibited spontaneous Ca²⁺ transients, and the summed signal in these cells over a 30 sec recording is shown in Fig. 6A. Using PTCL analysis (see Methods), PTCL maps were created illustrating initiating sites from which Ca²⁺ transients originated (Fig. 6A) and individual Ca²⁺ firing sites were colour coded (Fig. 6A, far right panel). Dozens of firing sites were observed in this example, and their activities were plotted as an occurrence map, with firing at each site occupying an individual coloured ‘lane’ and plotted against time (Fig. 6B). From such analysis, no obvious coordination or entrainment of ICC-SS firing sites was evident to indicate network-wide rhythmic Ca²⁺ signalling. Close examination of the occurrence maps showed that many firing sites displayed quite regular patterns, but the activities of other firing sites in the FOV did not appear to influence or be influenced by this activity.

Fig. 6: ICC-SS Ca²⁺ signaling is not entrained or coupled.

A Far left panel shows a raw image of multiple ICC-SS in a FOV. The panel to the right shows summed Ca²⁺ transients in the ICC-SS during a 20 sec recording period. The next panel shows an accumulation map of Ca²⁺ transient initiation sites during 20 sec (initiating PTCLs). The far right panel shows colour coded Ca²⁺ firing sites from which Ca²⁺ transients were initiated. B Occurrence map of Ca²⁺ firing sites shown in the field of view (FOV) in the far right panel of A. Ca²⁺ firing sites are placed into individual colour coded lanes and plotted against time. C The far left panel is a raw image of a single ICC-SS taken from an in situ recording. The middle panel shows an accumulation map of Ca²⁺ transient initiation sites during a 20 sec recording. The far right panel shows colour coded Ca²⁺ firing sites where Ca²⁺ transients were initiated. D Occurrence map of all Ca²⁺ firing sites in a single ICC-SS shown in C. The Ca²⁺ firing sites are placed into individual colour coded lanes and plotted against time. E STMs showing the effect of the gap junction inhibitor 18 β-glycyrrhetinic acid (18 β-GA, 10 μM) on Ca²⁺ transients in ICC-SS. F Summary data showing the effect of 18 β-GA on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=13, n=4.

We then applied this analysis to the activities of single cells. For example, Fig. 6C shows a representative single ICC-SS in the FOV shown in Fig. 6A, and the Ca²⁺ firing sites within this cell were identified, as described above (Fig. 6A–B). 9 Ca²⁺ firing sites were identified in this ICC-SS (Fig. 6C), and their firing patterns were plotted as an occurrence map in Fig. 6D. All 9 firing sites were spontaneously active, but the activity of any one firing site showed no apparent influence on the activities of neighbouring sites within the same cell. Thus, while some firing sites in ICC-SS display intrinsic rhythmicity; firing sites in a single cell or within multiple cells in a FOV were not entrained to produce coupled oscillations. A further demonstration of the lack of coupling was that the gap junction inhibitor 18 β-glycyrrhetinic acid (18 β-GA, 10 μM), had no effect on Ca²⁺ transients in ICC-SS (Fig. 6E), and no significant effects on the frequency (P=0.68), amplitude (P=0.29), duration (P=0.86) or spatial spread (P=0.54) of Ca²⁺ transients were observed (Fig. 6F, paired student t-tests, c=13, n=4).

Sources of Ca²⁺ required for ICC-SS activity

Ca²⁺ release from the ER is a fundamental behavior of ICC that is linked to activation of Ano1 channels and required for the regulatory functions of ICC (Zhu et al., 2015; Baker et al., 2016; Drumm et al., 2017, 2019b). Ca²⁺ release from the ER also appeared to be essential for the activity of ICC-SS, as incubation with SERCA pump inhibitors, thapsigargin (1 μM; Fig. 7A–B, c=16, n=5) or CPA (10 μM; Fig. 7C–D, c=12, n=4) abolished Ca²⁺ transients in ICC-SS (Fig. 7B,D, P<0.0001, paired student t-tests).

Fig. 7: SERCA pump inhibitors abolish ICC-SS Ca²⁺ transients.

A STMs showing the effect of the SERCA pump inhibitor, thapsigargin (1 μM), on ICC-SS Ca²⁺ transients. B Summary effect of thapsigargin on ICC-SS Ca²⁺ transient frequency, c=16, n=5. C STMs showing the effect of the SERCA pump inhibitor, CPA, on ICC-SS Ca²⁺ transients. D Summary effect of CPA on ICC-SS Ca²⁺ transient frequency, c=12, n=4.

ICC-SS were also reliant on Ca²⁺ influx from the external environment for the maintenance of Ca²⁺ transients, as incubating colon tissues with Ca²⁺ free solution (containing 0.5 mM EGTA) led to a significant reduction in the firing frequency of Ca²⁺ transients from 121.9 ± 60.9 to 56.9 ± 45.8 min⁻¹ after just 4 mins (Fig. 8A–B, P=0.0002, one way ANOVA, Tukey post hoc test, c=21, n=5). There was also a reduction in the duration of Ca²⁺ transients from 189.1 ± 40.2 ms to 98.2 ± 80.8 ms after 10 mins (Fig. 8B, P=0.011, one way ANOVA, Tukey post hoc test, c=21, n=5). The spatial spread of Ca²⁺ transients was reduced after 6 mins from 12.2 ± 4.2 to 8.1 ± 5.6 μm (Fig. 8B, P=0.0035, one way ANOVA, Tukey post hoc test, c=21, n=5). Conversely, increasing the external Ca²⁺ concentration from 2.5 mM to 5 mM increased the activity of ICC-SS (Fig. 8C). While there was no effect on the amplitude (P=0.97), duration (P=0.27) or spatial spread (P=0.39), increasing external Ca²⁺ increased the firing frequency of Ca²⁺ transients from 187.9 ± 96.6 to 223.5 ± 91.1 min⁻¹ (Fig. 8D, P=0.0011, paired student t-tests, c=16, n=5).

Fig. 8: The frequency of ICC-SS Ca²⁺ transients is modulated by Ca²⁺ influx.

A STMs showing the effect of Ca²⁺ free solution (buffered with 0.5mM EGTA) on ICC-SS Ca²⁺ transients after 4 minutes exposure to the Ca²⁺ free conditions. B Summary data showing the effect of Ca²⁺ free solution (0.5mM EGTA) on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread after 2–10 min exposures, c=21, n=5. All statistical comparisons are against control. C STMs showing the effect of 5 mM external Ca²⁺ solution on ICC-SS Ca²⁺ transients. D Summary data showing the effect of 5 mM external Ca²⁺ solution on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=16, n=5.

The specific Ca²⁺ release channels responsible for generating Ca²⁺ transients in ICC-SS were examined. Ryanodine (100 μM) had no significant effects on the frequency (P=0.1), amplitude (P=0.73), duration (P=0.13) or spatial spread (P=0.26) of Ca²⁺ transients in ICC-SS (Fig. 9A–B, paired student t-tests, c=8, n=3). The role of IP₃Rs was examined by testing the effects of a phospholipase C inhibitor (U73122) to decrease IP₃ production (Fig. 9C) and an inactive analogue of this drug (U73343). The inactive analogue, U73343 (10 μM), had no effect on ICC-SS Ca²⁺ transients (Fig. 9D, P=0.061, one way ANOVA, Tukey post hoc test, c=9, n=3), but U73122 (10 μM) reduced the firing frequency of Ca²⁺ transients from 87 ± 60.2 to 45.7 ± 63.3 min⁻¹ (Fig. 9D, P=0.047, one way ANOVA, Tukey post hoc test, c=9, n=3). U73122 also inhibited the spatial spread of Ca²⁺ transients from 12.9 ± 6.8 to 4.8 ± 5.3 μm (Fig. 9D, P=0.016, one way ANOVA, Tukey post hoc test, c=9, n=3).

Fig. 9: ER channels contributing to Ca²⁺ release in ICC-SS.

A STMs showing the effect of the ryanodine receptor antagonist, ryanodine (100 μM), on ICC-SS Ca²⁺ transients. B Summary data showing the effect of ryanodine on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=8, n=3. C STMs showing the effect of the PLC inhibitor, U73122 (10 μM), on ICC-SS Ca²⁺ transients. D Summary data showing the effect of U73122 and its inactive analogue, U73343 (10 μM), on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=9, n=3. E STMs showing the effect of caffeine (10 mM) on ICC-SS Ca²⁺ transients. F Summary effect of caffeine on ICC-SS Ca²⁺ transient frequency, c=9, n=5.

Xestospongin C (XeC, 10 μM), a direct antagonist of IP₃ receptors (IP₃Rs) had mixed effects. Across multiple batches of XeC, the drug had dramatic inhibitory effects in some experiments, excitatory effects in others and no effect in others still (data not shown). Recently, the efficacy of XeC as a selective inhibitor of IP₃Rs has been questioned (Saleem et al., 2014) and off target effects other than IP₃R inhibition have been reported for this agent (Sergeant et al., 2001). Experiments performed on isolated isoforms of the IP₃R (IP₃R1–3) in expressed systems and on IP₃R dependent Ca²⁺ signaling in vascular endothelial cells have demonstrated that caffeine is an effective antagonist of IP₃R1 (dominant isoform in ICC (Baker et al., 2016; Drumm et al., 2017, 2019b)) at concentrations of 10 mM (Saleem et al., 2014; Heathcote et al., 2019). While caffeine is also known to be a modulator of RyRs, the lack of effect of the RyR antagonist on ICC-SS Ca²⁺ signals (Fig. 9A–B) led us to test caffeine as an antagonist of IP₃R1 with any effects on RyRs presumably being minimal. As shown in Fig. 9E–F, incubation with 10 mM caffeine abolished all Ca²⁺ transients in ICC-SS (Fig. 9F, P<0.0001, paired student t-test, c=9, n=5).

The high frequency firing of Ca²⁺ transients in ICC-SS begs the questions of how this activity is sustained. This activity clearly requires a mechanism to reinitiate IP₃-mediated Ca²⁺ release from ER stores during long periods of multiple Ca²⁺ release events. In ICC from other organs, Ca²⁺ release may be sustained by regular influx of Ca²⁺ through a voltage-dependent pathway (Park et al., 2006; Bayguinov et al., 2007; Sanders et al., 2014) or by a voltage dependent generation of IP₃ (van Helden et al., 2000; Imtiaz et al., 2002; van Helden & Imtiaz, 2003). We tested the role of these voltage-dependent mechanisms in ICC-SS by hyperpolarizing colonic muscles by activation of K_ATP channels. K_ATP channels contribute to setting the resting membrane potential in the colon (Koh et al., 1998) and are highly expressed in colonic SMC, but this conductance is not resolved in ICC (Koh et al., 1998; Huang et al., 2018). We reasoned that activation of K_ATP channels with pinacidil, which hyperpolarizes colonic muscles (Koh et al., 1998), would reduce either voltage dependent Ca²⁺ influx or voltage-dependent IP₃ production. Pinacidil (10 μM; given in the presence of 1 μM tetrodotoxin (TTX; Fig. 10A) to prevent interference from possible voltage-dependent effects on neurotransmitter release) had no effect on Ca²⁺ transient frequency (P=0.06), amplitude (P=0.7), duration (P=0.9) or spatial spread (P=0.2), (Fig. 10B, paired student t-tests, c=21, n=5). Similarly, inhibiting Ca²⁺ influx via L-type Ca²⁺ channels with nicardipine (1 μM) had no effect on ICC-SS Ca²⁺ transient frequency (P=0.17), amplitude (P=0.48), duration (P=0.57) or spread (P=0.4) (Fig. 10C–D, paired student t-tests, c=13, n=5).

Fig. 10: Ca²⁺ transients in ICC-SS are voltage-independent.

A STMs showing the effect of the K_ATP channel agonist, pinacidil (10 μM), on ICC-SS Ca²⁺ transients, in the presence of 1 μM TTX. B Summary data showing the effect of pinacidil on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=21, n=5. C STMs showing the effect of nicardipine (1 μM), on ICC-SS Ca²⁺ transients. D Summary data showing the effect of nicardipine on Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=13, n=5.

Role of SOCE in sustaining ICC-SS activity

The data above suggested that ICC-SS do not require voltage dependent Ca²⁺ influx for the generation of spontaneous activity. However, we also demonstrated that a Ca²⁺ influx pathway must exist to sustain Ca²⁺ release in these cells, as spontaneous activity was acutely dependent on extracellular Ca²⁺. SOCE has been shown to be an important mechanism for store refilling and maintaining Ca²⁺ release in several types of muscles (Trebak et al., 2013), including smooth muscles (Gibson et al., 1998; Drumm et al., 2018a).

SKF 96365 (10 μM), a broadly selective SOCE antagonist, inhibited the frequency, amplitude and spatial spread of Ca²⁺ transients in ICC-SS (Fig. 11A); the frequency of Ca²⁺ transients was reduced from 216.9 ± 100.6 min⁻¹ to 139.1 ± 53.5 min⁻¹ (Fig. 11B, P=0.0017, paired student t-test, c=14, n=5), amplitudes were reduced from 0.5 ± 0.2 ΔF/F₀ in control to 0.4 ± 0.1 ΔF/F₀ by SKF 96365 (Fig. 11B, P=0.013, paired student t-test, c=14, n=5). SKF 96365 decreased the spatial spread of Ca²⁺ transients from 11.3 ± 2.6 to 9.1 ± 3.2 μm (Fig. 11B, P=0.014, paired student t-test, c=14, n=5), but did not affect Ca²⁺ transient duration (Fig. 11B, P=0.64, paired student t-test, c=14, n=5). SOCE is mediated by interactions between the stromal interaction molecular (STIM) proteins and Orai Ca²⁺ influx channels (Trebak & Putney, 2017). We took advantage of the dual effects of 2-aminoethoxydiphenyl borate (2-APB) on Orai channels and STIM proteins to further elucidate the role of SOCE in sustaining Ca²⁺ transients in ICC-SS. Low concentrations of 2-APB (5–10 μM) increase the conductance of Orai channels, while higher concentrations (50–100 μM) inhibit Orai (Peinelt et al., 2008; Yamashita et al., 2010; Xu et al., 2016; Ali et al., 2017). 2-APB (10 μM) increased the frequency of ICC-SS Ca²⁺ transients from 177.2 ± 83.9 min⁻¹ to 227.5 ± 74.9 min⁻¹ (Fig. 11C–D, P=0.006, one way ANOVA, Tukey post hoc test, c=19, n=6), and 50 μM 2-APB reduced the firing frequency to 63.9 ± 57.6 min⁻¹ (Fig. 11D, P<0.0001, one way ANOVA, Tukey post hoc test, c=19, n=6). 2-APB (10 μM) also increased spatial spread from 10.7 ± 4.5 to 14.5 ± 5.1 μm (Fig. 11D, P=0.039, one way ANOVA, Tukey post hoc test, c=19, n=6).

Fig. 11: Ca²⁺ transients in ICC-SS are modulated by SOCE.

A STMs showing the effect of the broadly selective SOCE inhibitor SKF 96365 (10 μM), on ICC-SS Ca²⁺ transients. B Summary data showing the effect of SKF 96365 on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=14, n=5. C STMs showing the effect of the different concentrations of 2-APB (10 & 50 μM) on ICC-SS Ca²⁺ transients. D Summary data showing the effect of 2 concentrations of 2-APB on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=19, n=6.

GSK 7975A, a selective antagonist of Orai channels, had dramatic inhibitory effects on ICC-SS Ca²⁺ transients (Fig. 12A). GSK 7975A (1μM) reduced the firing frequency of Ca²⁺ transients from 164.3 ± 62.9 min⁻¹ to 92.65 ± 55.9 min⁻¹ (Fig. 12B, P=0.006, one way ANOVA, Tukey post hoc test, c=17, n=5), and frequency was further reduced to 26.3 ± 33 min⁻¹ by 10 μM GSK 7975A (Fig. 12B, P<0.0001, one way ANOVA, Tukey post hoc test, c=17, n=5). 10 μM GSK 7975A also significantly reduced the duration (P=0.046) and spatial spread (P=0.0014) of Ca²⁺ transients (Fig. 12B, one way ANOVA, Tukey post hoc test, c=17, n=5). GSK 7975A (1μM) also reduced the stimulatory effects of 10 μM 2-APB on ICC-SS Ca²⁺ transients (Fig. 12C). For example, as shown in Fig. 12D, 2-APB (10 μM) increased the frequency of ICC-SS Ca²⁺ transients from 49.7 ± 84.3 to 82.3 ± 89.1 min⁻¹ (Fig. 12D, P=0.033, one way ANOVA, Tukey post hoc test, c=10, n=3) and this was then reduced to 23.3 ± 34.5 min⁻¹ by GSK 7975A (Fig. 12D, P=0.032, one way ANOVA, Tukey post hoc test, c=10, n=3).

Fig. 12: ICC-SS Ca²⁺ transients are sustained by Ca²⁺ influx via Orai channels.

A STMs showing the effect of the Orai channel inhibitor GSK 7975A (1 & 10 μM), on ICC-SS Ca²⁺ transients. B Summary data showing the effect of GSK 7975A, on ICC-SS Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=17, n=5. C STMs showing that the stimulatory effect of 2-APB (10 μM) on ICC-SS Ca²⁺ transients was reversed by GSK 7975A (1 μM). D Summary data showing the effect of GSK 7975A on ICC-SS Ca²⁺ transient frequency, amplitude, duration and spatial spread in the presence of 2-APB (10 μM), c=10, n=3.

Neural responsiveness of ICC-SS

Some types of ICC are known to transduce inputs from enteric motor neurons in the stomach (Burns et al., 1996; Ward et al., 2000, 2006; Beckett et al., 2002; Sung et al., 2018), small intestine (Ward et al., 2006; Baker et al., 2018a, 2018b) and colon (Wang et al., 2000; Drumm et al., 2019c). Previous studies have found anatomical associations between ICC-SS and nNOS⁺ neurons in primates (Blair et al., 2012b). We tested responses of ICC-SS to enteric nerve stimulation by observing the effect of neurotransmitter agonists and electrical field stimulation (EFS) on ICC-SS Ca²⁺ transients.

In contrast to ICC-IM in the proximal colon (Drumm et al., 2019c), we found no evidence that ICC-SS undergo tonic inhibition from nitrergic neurons, as basal Ca²⁺ transient activity was unaffected by either TTX (1 μM, Fig. 13A–B; Frequency (P=0.053), Amplitude (P=0.44), Duration (P=0.48), Spatial Spread (P=0.25), paired student t-tests, c=18, n=5) or the nNOS synthase inhibitor N_ω-nitro-L-arginine (L-NNA, 100 μM, Fig. 13C–D, Frequency (P=0.074), Amplitude (P=0.77), Duration (P=0.096), Spatial Spread (P=0.55), paired student t-tests, c=15, n=5).

Fig. 13: ICC-SS do not undergo tonic neural inhibition.

A STMs showing the effect of TTX (1 μM), on ICC-SS Ca²⁺ transients. B Summary effects of TTX on ICC-SS Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=18, n=5. C STMs showing the effect of the nNOS synthase inhibitor L-NNA (100 μM) on ICC-SS Ca²⁺ transients. D Summary effects of L-NNA on ICC-SS Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=15, n=5.

ICC-SS were imaged in situ before, during and after EFS (10 Hz; 10 sec) in the presence of atropine (1 μM; Fig. 14A–B). At initiation of EFS, an immediate cessation of Ca²⁺ transient firing occurred and lasted for the duration of stimulation (Fig. 14A–B). EFS reduced the firing frequency of Ca²⁺ transients from 130.8 ± 78.6 to 1.3 ± 6.5 min⁻¹ (Fig. 14C, P<0.0001, one way ANOVA, Tukey post hoc test, c=45, n=10). The amplitude, duration and spatial spread were reduced from 0.5 ± 0.2 to 0.03 ± 0.1 ΔF/F₀ (Fig. 14C, P<0.0001, one way ANOVA, Tukey post hoc test, c=45, n=10); from 168.5 ± 30 to 12.2 ± 47.4 ms (Fig. 14C, P<0.0001, one way ANOVA, Tukey post hoc test, c=45, n=10) and from 11.1 ± 3.8 to 0.3 ± 1.6 μm (Fig. 14C, P<0.0001, one way ANOVA, Tukey post hoc test, c=45, n=10), respectively. Upon cessation of EFS, a post-stimulus enhancement (‘rebound’) in Ca²⁺ transients was observed (Fig. 14A–C), and the frequency (P<0.0001), amplitude (P<0.0001), duration (P=0.0004) and spatial spread (P<0.0001) increased, as compared to pre-EFS values (Fig. 14C, one way ANOVA, Tukey post hoc test, c=45, n=10).

Fig. 14: Inhibitory innervation of ICC-SS.

A STM of the effect of electrical field stimulation (EFS, 10 Hz, 10 sec) on ICC-SS Ca²⁺ transients in the presence of the muscarinic receptor antagonist atropine (1 μM). B 3-D plots of the effect of EFS (10 Hz, 10 sec) on ICC-SS Ca²⁺ transients in the presence of atropine. C Summary data of EFS effects in the presence of atropine on the frequency, amplitude, duration and spatial spread of Ca²⁺ transients, comparing EFS and post EFS periods to pre EFS, c=45, n=10.

These data suggest that ICC-SS are innervated by inhibitory motor neurons. We tested whether the inhibitory neurotransmitter could be nitric oxide (NO), as is the case in many regions of the GI tract (Bult et al., 1990; Sanders & Ward, 2019). Ca²⁺ transients in ICC-SS were abolished when muscles were treated with an NO donor, DEA-NONOate (10 μM; Fig. 15A–B, P=0.0001, paired student t-test, c=12, n=4). The inhibitory effects of EFS on Ca²⁺ transients were blocked by pre-incubation with L-NNA (Fig. 15C–D, P<0.0001, one way ANOVA, Tukey post hoc tests, c=18, n=6). The soluble guanylate cyclase (downstream receptor for NO) activator (sGC, 1 μM Bay 58–2667, Fig. 16A) also abolished Ca²⁺ transients in ICC-SS (Fig. 16A–B, P<0.0001, paired student t-test, c=11, n=4), and ODQ (10 μM) an inhibitor of sGC, blocked the inhibitory effects of EFS on Ca²⁺ transients (Fig. 16C–D, P<0.0001, one way ANOVA, Tukey post hoc tests, c=15, n=7). Inhibiting purinergic neurotransmission in the proximal colon by antagonizing P₂Y₁ receptors with MRS 2500 (1 μM, Fig. 17A) had no effect on EFS induced inhibitory responses on ICC-SS (Fig. 17B; Frequency (P=0.59), Amplitude (P=0.59), Duration (P=0.59), Spatial Spread (P=0.59), one way ANOVA, Tukey post hoc tests, c=9, n=3).

Fig. 15: Inhibitory neural effects on ICC-SS Ca²⁺ transients are due to NO.

A STMs showing the effect of the NO donor DEA-NONOate (10 μM) on ICC-SS Ca²⁺ transients. B Summary effect of DEA-NONOate on ICC-SS Ca²⁺ transient frequency, c=12, n=4. C STMs of the effect of EFS (10 Hz, 10 sec) on ICC-SS Ca²⁺ transients in the presence of atropine (1 μM) before and after addition of L-NNA (100 μM). D Summary effects of EFS in the presence of atropine on the frequency, amplitude, duration and spatial spread of Ca²⁺ transients, comparing Pre EFS period to the EFS period and comparing the EFS period to EFS + LNNA, c=18, n=6.

Fig. 16: Inhibitory neural effects on ICC-SS Ca²⁺ transients are due to sGC.

A STMs showing the effect of the sGC activator Bay 58–2667 (1 μM) on ICC-SS Ca²⁺ transients. B Summary effect of Bay 58–2667 on ICC-SS Ca²⁺ transient frequency, c=11, n=4. C STMs of the effect of EFS (10 Hz, 10 sec) on ICC-SS Ca²⁺ transients in the presence of atropine (1 μM) before and after addition of the sGC inhibitor ODQ (10 μM). D Summary effects of EFS in the presence of atropine on the frequency, amplitude, duration and spatial spread of Ca²⁺ transients, comparing activity before and during EFS and comparing the effects of EFS and the effects of EFS + ODQ, c=15, n=7.

Fig. 17: Purinergic neurotransmission is not involved in inhibitory neural responses in ICC-SS.

A STMs of the effect of EFS (10 Hz, 10 sec) on ICC-SS Ca²⁺ transients in the presence of atropine (1 μM) before and after addition of the P2Y1 receptor antagonist MRS 2500 (1 μM). B Summary effects of EFS in the presence of atropine on the frequency, amplitude, duration and spatial spread of Ca²⁺ transients, comparing Pre EFS period to the EFS period and comparing the EFS period to EFS + MRS 2500, c=9, n=3.

We also tested the effects of excitatory neurotransmission on ICC-SS. The cholinergic receptor agonist carbachol (CCh, 1 μM, given in the presence of TTX) failed to yield any change in Ca²⁺ transient frequency (P=0.17), amplitude (P=0.51), duration (P=0.6) or spatial spread (P=0.2), (Fig. 18A–B, paired student t-tests, c=12, n=4). Furthermore, we imaged ICC-SS in situ before, during and after EFS (10 Hz; 10 sec) in the presence of L-NNA (100 μM) and MRS 2500 (1 μM) to block inhibitory nitrergic and purinergic neurotransmission (Fig. 18C). In stark contrast to the observations above under conditions favoring inhibitory neurotransmission, we found that EFS under these conditions produced no change in the frequency (P=0.7), amplitude (P=0.94), duration (P=0.63) or spatial spread (P=0.96) of ICC-SS Ca²⁺ transients (Fig. 18D, one way ANOVA, Tukey post hoc tests, c=13, n=5).

Fig. 18: Excitatory neurotransmission does not affect ICC-SS Ca²⁺ transients.

A STMs showing the effect of the cholinergic agonist, carbachol (CCh, 1 μM), on ICC-SS Ca²⁺ transients, in the presence of 1 μM TTX. B Summary data showing the effect of CCh on ICC Ca²⁺ transient frequency, amplitude, duration and spatial spread, c=12, n=4. C STMs of the effect of EFS (10 Hz, 10 sec) on ICC-SS Ca²⁺ transients in the presence of L-NNA (100 μM) and MRS 2500 (1 μM). D Summary effects of EFS in the presence of L-NNA + MRS 2500 on the frequency, amplitude, duration and spatial spread of ICC-SS Ca²⁺ transients, c=13, n=5.

We also tested the hypothesis that ‘rebound’ excitation in ICC-SS after inhibitory EFS is responsible for the responses manifest in contractions of LM. A strong increase in Ca²⁺ transients in ICC-SS after cessation of EFS should enhance openings of Ano1, and the depolarization, conducted to LSMC should result in increased Ca²⁺ action potential firing and contraction.

‘Rebound’ responses in LSMCs were explored using two approaches: Ca²⁺ imaging of LSMCs in muscles of SmHC-Cre-GCaMP6f mice and contractile experiments. EFS in the presence of atropine resulted in inhibition of Ca²⁺ transients during stimulation and ‘rebound’ excitation manifested as an intense burst of high amplitude intercellular Ca²⁺ waves upon cessation of EFS (Fig. 19A). Ani 9 (1 μM) caused significant a reduction in the ‘rebound’ response (Fig. 19A–C), and normalized area under the curve (AUC) of the post-EFS Ca²⁺ response was reduced from 1.0 to 0.57 ± 0.26 (Fig. 19C, paired student t-test, P=0.02, n=5). In contractile experiments, EFS (in the presence of atropine) was applied to LM strips at 1, 3, 5, 10 and 20 Hz (10 s train). ‘Rebound’ contractions were observed at each frequency tested (Fig. 19D, n=5), and these responses were significantly reduced by Ani9 (1 μM) at all frequencies (Fig. 19D–E, n=5). The AUC of ‘rebound’ contraction was reduced by Ani 9 from 1 to 0.3 ± 0.2 at 1 Hz, (Fig. 19E, P<0.0001, n=5), 0.26 ± 0.19 at 3 Hz (Fig. 19E, P<0.0001, n=5), 0.4 ± 0.2 at 5 Hz (Fig. 19E, P<0.0001, n=5), 0.46 ± 0.25 (Fig. 19E, P<0.0001, n=5) at 10 Hz and 0.5 ± 0.3 at 20 Hz (Fig. 19E, P=0.003, n=5).

Fig. 19: Activation of Ano1 channels contributes to LM post EFS ‘rebound’ excitation.

A STMs showing effects of EFS (10 Hz, 10 sec) on intercellular Ca²⁺ waves in LSMC in the presence of atropine (1 μM) before and after addition of Ani 9 (1 μM) recorded with a 20x objective. Responses are recorded in situ from a SmHC-Cre-GCaMP6f mouse. B Traces of the Ca²⁺ activity from the STMs in A, showing the decrease in ‘rebound’ excitation after addition of Ani 9. C Summary data showing the effects of Ani 9 on ‘rebound’ excitation in LSMC, measured as AUC from traces such as those shown in B, n=5. D Isometric tension recordings showing ‘rebound’ contractions of proximal colon LM following EFS (3, 5, 10, 20Hz, 10 sec) in the presence of atropine, before and after addition of Ani 9 (1 μM). E Summary data showing the effect of Ani 9 on ‘rebound’ contractions of proximal colon LM post EFS in the presence of atropine, n=5.

Discussion

This study investigated the behavior and functions of a novel class of ICC in the colon and how LM is regulated by ICC-SS. LSMCs generated 2 types of Ca²⁺ waves, as reported previously (Hennig et al., 2002). Intracellular Ca²⁺ transients were stochastic in nature, consisted of localized waves and were not coupled to contractions. Intercellular Ca²⁺ waves were more intense and spread within bundles of LSMCs causing contraction. Unlike intracellular Ca²⁺ waves, intercellular Ca²⁺ waves in LSMCs and LM contractions were inhibited by an Ano1 channel antagonist and by a pharmacological antagonist of SOCE. Previous studies described morphological features of ICC-SS and showed that these cells are arranged along LM bundles (Burns et al., 1997; Toma et al., 1999; Vanderwinden et al., 2000; Aranishi et al., 2009; Blair et al., 2012a; Rumessen et al., 2013), however no information about their function(s) was provided previously. The present study hypothesized that due to the cellular expression of Ca²⁺ signalling pathways, ion channel expression and anatomical connectivity of ICC-SS and LSMCs in the proximal colon, ICC-SS could be involved in driving the firing of intercellular Ca²⁺ waves and contractions in LSMCs. ICC-SS, like ICC in other regions of the GI tract, express Ano1 (Gomez-Pinilla et al., 2009; Blair et al., 2012a). Ca²⁺ transients are likely coupled to activation of Ano1, a Ca²⁺-activated Cl⁻ channel, in ICC-SS. We characterized the properties of Ca²⁺ transients in ICC-SS using an optogenetic approach. Ca²⁺ transients in ICC-SS were very sensitive to agents that modulate SOCE and specifically by a blocker of Orai channels (GSK 7975A). Antagonists of Ano1 or block of Ca²⁺ transients in ICC-SS reduced intercellular Ca²⁺ waves and contractions of LM, suggesting that ICC-SS provide a pacemaker-like input to LM. Ca²⁺ transients in ICC-SS were also inhibited by nitrergic agonists and during EFS of intrinsic enteric motor neurons via a nitrergic mechanism. A strong post-stimulus increase in Ca²⁺ transients (rebound response) followed periods of nitrergic motor neuron stimulation that appears to be responsible, at least in part, for post-stimulus contractions of LM. No innervation by enteric excitatory neurons or responsiveness to excitatory motor neurotransmitters was detected in ICC-SS.

Ca²⁺ transients occurred in ICC-SS spontaneously in a stochastic manner and were unaffected by TTX or L-NNA, suggesting that these cells are not under the influence of tonic inhibition, as was observed previously in intramuscular ICC (ICC-IM) of the CM layer (Drumm et al., 2019c). ICC-SS form an apparent network with electrical connectivity provided by gap junctions between cells (Vanderwinden et al., 2000; Aranishi et al., 2009), but Ca²⁺ transients were not coordinated or ‘entrained’, as sometimes suggested by investigators that have viewed ICC as coupled oscillators (Huizinga et al., 2014). Even within single ICC-SS, Ca²⁺ transients occurred independently of events originating from other initiation sites in the same cell. We also found that the gap junction uncoupler, 18β-glycyrrhetinic acid, which dramatically affects Ca²⁺ signalling in coupled ICC-MY of the small intestine (Park et al., 2006), did not affect firing of ICC-SS Ca²⁺ transients, suggesting no influence of electrical coupling on release of Ca²⁺ from stores. In pacemaker ICC from the small intestine, cell-to-cell coordination of Ca²⁺ firing occurs via Ca²⁺ entry through voltage-dependent Ca²⁺ channels (Drumm et al., 2017). Lack of coordination between Ca²⁺ transients in ICC-SS suggests that the depolarizing influence of Ca²⁺ transients and activation of Ano1 channels does not activate voltage-dependent Ca²⁺ channels in ICC-SS, perhaps due to low expression of these channels. This behavior is reminiscent of the behavior of ICC-DMP (Baker et al., 2016) in the small intestine and ICC-IM in the colon (Drumm et al., 2019b).

If ICC-SS provide pacemaker input to the LM, this activity takes a different form than activities of ICC-MY in the stomach and small intestine, as ICC-MY provide classical pacemaker-like activity setting the frequency and duration of the excitable events responsible for the phasic contractions that underlie peristalsis and segmentation. Due to the apparent lack of a coordinating, voltage-dependent conductance, Ca²⁺ events in ICC-SS occur independently in a stochastic manner, but summation of depolarizing currents activated in ICC-SS may drive membrane potentials of LSMCs into a range where action potential generation occurs. The voltage-dependent conductances in LSMCs may develop this activity into a rhythmic pattern. A model has shown that driving membrane potential of SMCs into a certain range of potentials, can induce voltage-dependent conductances expressed by SMCs to develop a rhythmic output (Lang et al., 1992). Thus, while ICC-SS may not directly pace the rhythmicity of LSMCs (i.e. providing the timing for excitable events in the muscle) they may contribute to pacemaker activity by conditioning membrane potential to support rhythmic generation of muscle action potentials. Thus, electrical and mechanical rhythmicity may be an emergent property of ICC-SS and LSMC functioning within an integrated cellular network.

The responsiveness of ICC-SS to enteric neurotransmission was somewhat surprising based on prior morphological reports. Sparse innervation was observed in the subserosal layer of mice and guinea pig proximal colon, and few nNOS⁺ neurons were observed (Vanderwinden et al., 2000; Aranishi et al., 2009). However, ICC-SS were found to be associated with nNOS⁺ neurons at the subserosal surface of the taenia coli in primates (Blair et al., 2012b). Our study provided evidence of nitrergic innervation of ICC-SS, as EFS in the presence of atropine caused profound inhibition of Ca²⁺ transients, an effect that was mimicked by application of an exogenous NO donor and sGC agonist and reversed by nNOS synthase and sGC antagonists. In contrast, we observed no evidence suggesting that ICC-SS receive purinergic, cholinergic or excitatory peptidergic neural inputs. Longitudinal muscles clearly receive excitatory neural input, however, and this input might be mediated directly by direct innervation of LSMCs.

It is of historical interest to note that post-stimulus excitatory responses (called ‘rebound’ excitation) was first characterized in colonic longitudinal muscles (Bennett, 1966; Burnstock et al., 1966). Rebound excitation was elicited immediately after a period of inhibitory neurotransmission and persisted in the presence of atropine. Mechanisms to explain rebound excitation included anode break, where hyperpolarization during inhibitory neurotransmission removes inactivation of channels responsible for action potentials (in this case L-type Ca²⁺ channels). Cessation of the inhibitory period, repolarization and the increased availability of L-type Ca²⁺ channels can result in enhanced action potentials firing. This may be the case for single stimuli, which are predominantly due to purinergic hyperpolarization. However, nitrergic responses are elicited by multiple stimuli (Gallego et al., 2008), and rebound excitation in LM may result from the dramatic increase in Ca²⁺ transients in ICC-SS. The post-stimulus increase in ICC-SS Ca²⁺ transients failed to develop in the presence of L-NNA or ODQ, which inhibited nitrergic neurotransmission. Rebound excitation, in the form of bursts of LSMCs intercellular Ca²⁺ waves and large amplitude LM contractions after EFS, was reduced by ~ 50% by an Ano1 channel antagonist.

A major conclusion about the role of ICC-SS in regulating LM is derived from the use of 2 reagents, an Ano1 antagonist (Ani 9) and an Orai antagonist (GSK 7975A). In the case of Ano1 antagonists, caution is required because all of the compounds we have tested have non-specific effects on L-type Ca²⁺ channels at concentrations above 5 μM. However, we found significant reduction in intercellular Ca²⁺ waves and contractions of LM in response to Ani 9 (1 μM), a dose effective in blocking Ano1 currents, but without effects on the responses of colonic muscles to elevated external K⁺ (Drumm et al., 2019c). SMCs In the small intestine do not exhibit SOCE properties and SOCE currents could only be resolved in intestinal ICC (Zheng et al., 2018), however it remains unknown if this also holds true in the colon. The other cells that comprise the SIP synticum in the colon (SMCs and PDGFRα⁺ cells) also express Ca²⁺ dependent conductances, and activation of these conductances may also rely on SOCE and Ca²⁺ release. However, in colonic SMCs and PDGFRα⁺ cells, Ca²⁺ release is coupled to the activation of outward K⁺ currents that stabilize membrane potentials or hyperpolarize colonic muscles (BK for SMCs (Thornbury et al., 1992; Hennig et al., 2002) and SK in PDGFRα⁺ cells (Kurahashi et al., 2011)). Thus, one would predict that inhibiting SOCE in these cells would lead to an increase in muscle excitability. However, we found that GSK 797A decreased LM contractions and intercellular Ca²⁺ waves, suggesting that the dominant effect of inhibiting SOCE was to disrupt Ca²⁺ handling in ICC.

A common theme is emerging about the powerful inhibitory regulation imposed by nitrergic stimuli on Ca²⁺ release events in ICC (Baker et al., 2018a; Drumm et al., 2019c). At present, we know that these effects are mediated by generation of cGMP, but steps down-stream from generation of the 2^nd messenger have not been clarified. Previous studies have linked cGMP-dependent protein kinase G (PKG) to inhibition of IP₃ dependent Ca²⁺ release in SMCs, interstitial cells and endothelial cells (Murthy & Zhou, 2003; Sergeant et al., 2006; Borysova & Burdyga, 2015). This mechanism utilizes direct phosphorylation of IP₃Rs by PKG to inhibit IP₃ receptors and reduce their open probability (Murthy & Zhou, 2003; Borysova & Burdyga, 2015). A signaling molecule activated by PKG is inositol triphosphate receptor (IP₃R)-associated cGMP-kinase substrate (IRAG; encoded by Mrvi1). IRAG co-precipitates with IP₃Rs and is essential for the cGMP-dependent inhibition of Ca²⁺ release in cultured human colonic SMCs and for the regulation of IP₃Rs more generally (Schlossmann et al., 2000; Fritsch et al., 2004; Zhang et al., 2011). The molecular elements of this pathway, IP₃Rs, PKG and IRAG are expressed in colonic and small intestinal ICC (Chen et al., 2007; Baker et al., 2016, 2018a; Drumm et al., 2017, 2019b; Lee et al., 2017), so it is possible that a mechanism similar to that in SMCs provides inhibitory regulation of Ca²⁺ release in ICC. While Ca²⁺ release may be inhibited during nitrergic stimulation, Ca²⁺ uptake into stores or Ca²⁺ through influx channels may be unimpeded. Thus, during nitrergic stimulation Ca²⁺ stores may be overloaded, and the burst of Ca²⁺ transients at cessation of nitrergic input may result from the increased Ca²⁺ load in stores. Increased store load may increase the activity of luminal Ca²⁺ binding sites within the ER of ICC, increasing the sensitivity and open probability of IP₃Rs, as occurs for the activity of RyRs in cardiac and SMCs under store overload conditions (Cheng et al., 1993; Györke & Györke, 1998; Dargan et al., 2004; Györke et al., 2017; Ríos, 2017). Store overload may initiate enhanced Ca²⁺ release when nitrergic stimulation ceases and the inhibitory drive on IP₃Rs is relieved. Conversely, the period of nitrergic inhibition may allow for an increased refractory time for cytosolic IP₃Rs themselves on the ER membrane, which are activated and deactivated by Ca²⁺ in a bell-shaped manner (Berridge et al., 2000; Berridge, 2009). Thus, a long period of limited Ca²⁺ release would increase the recovery time of IP₃Rs, and allow them to be ‘primed’ for greater release when the inhibitory influence of NO is removed compared to their recovery states under basal conditions, when continual rhythmic Ca²⁺ release leads to a somewhat constant refractory state for at least some IP₃Rs.

Shortcomings of the present study include a lack of evaluation of key genes in ICC-SS that encode the necessary elements of the proposed pacemaker mechanism and cell-to-cell communication apparatus that would be required for ICC-SS to control activation of LM and to mediate neural responses to LM. Prior morphological studies showed that ICC-SS formed close associations with LSMCs and peg-and-socket junctions were also noted (Vanderwinden et al., 2000; Rumessen et al., 2013). Functional evidence, however: i) inhibition of intercellular Ca²⁺ transients in LM cells by Ano1 channel and Orai antagonists; ii) inhibition of contractions of LM by an Ano1 antagonist; iii) alterations in frequency and duration of LM contractions by an Orai antagonist, suggest that communication occurs between ICC-SS and LSMCs. We found it very difficult to disperse sufficient numbers of ICC-SS unequivocally from serosal tissues that would be necessary for sorting and gene expression assays. Thus, a clear picture of the molecular apparatus to accomplish Ca²⁺ entry and communication between cells has not yet been determined. While Ca²⁺ transients were recorded from ICC-SS and LSMCs, these recordings were made from different muscles. Simultaneous recordings of Ca²⁺ transients in both cell layers or recordings of Ca²⁺ transients in ICC-SS while recording action potentials in LSMCs would provide a tighter correlation between activities. Cellular studies are also desirable to show that Ca²⁺ transients are coupled to activation of spontaneous transient inward currents (STICs), however such measurements have not been accomplished to date with any type of ICC. It is not entirely clear how stochastic firing of Ca²⁺ transients in ICC-SS can result in the rhythmic firing of intercellular Ca²⁺ transients in LSCMs in intact muscles. This is likely to be a complicated outcome of the kinetic features of Ca²⁺ release channels, uptake mechanisms and ion channel refractoriness. Modeling of Ca²⁺ release events in ICC-SS and integration of this activity with the electrophysiological properties of LSMCs may provide a better understanding of this emergent behavior.

In summary, we have confirmed that the LM of the proximal colon exhibits two distinct patterns of Ca²⁺ signaling. Intracellular Ca²⁺ waves do not elicit contractions and are probably coupled to activation of BK channels in LSMCs. Intercellular Ca²⁺ waves propagate through LM bundles, initiate contractions and are inhibited by antagonists of Ano1 and Orai channels. ICC-SS are in anatomical close proximity to LM bundles, express Ano1 channels, exhibit stochastic, Ca²⁺ transients that originate from Ca²⁺ release from ER stores, and are sustained by store refilling via Orai channels (SOCE). By activating Ano1 channels, Ca²⁺ transients in ICC-SS increase the excitability of electrically coupled LSMCs. ICC-SS are innervated by nitrergic nerves and the NO-sGC pathway provides powerful inhibitory regulation of Ca²⁺ signalling in ICC-SS. A large post-stimulus increase in Ca²⁺ transients leads to LSMC depolarization and a robust series of action potentials, intercellular Ca²⁺ waves and post-stimulus ‘rebound’ contraction. A diagram depicting the major features of this model is shown in Fig. 20.

Fig. 20: Proposed regulatory influence of ICC-SS on longitudinal muscle excitability.

Colonic ICC-SS fire Ca²⁺ transients from ER stores via IP₃Rs. Maintaining Ca²⁺ transients requires ER refilling by SOCE (via Orai channels) and SERCA. Ca²⁺ release activates Ano1 and Cl⁻ efflux, leading to ICC-SS depolarization. Depolarization is conducted to LSMCs via gap junctions and increases the open probability of voltage-dependent Cav_1.2 channels, leading to action potentials, Ca²⁺ influx and rapidly propagating Ca²⁺ waves and excitation-contraction coupling (EC) in LSMC bundles. NO released from nNOS⁺ enteric inhibitory motor neurons modulates the activity of ICC-SS and LSMC by activating sGC and inhibiting Ca²⁺ release in ICC-SS. LSMC also fire intracellular Ca²⁺ waves that are not affected by Ano1 and Cav_1.2 channels and are not linked to E-C coupling. These likely represent Ca²⁺ signals that activate large K⁺ channels (BK) in SMCs that serve to maintain negative resting potentials in LSMCs.

Key Points:

• Rhythmic action potentials and intercellular Ca²⁺ waves are generated in smooth muscle cells of colonic longitudinal muscles (LSMC).

• Longitudinal muscle excitability is tuned by input from subserosal ICC (ICC-SS), a population of ICC with previously unknown function.

• ICC-SS express Ano1 channels and generate spontaneous Ca²⁺ transients in a stochastic manner.

• Release of Ca²⁺ and activation of Ano1 channels causes depolarization of ICC-SS and LSMC, leading to activation of L-type Ca²⁺ channels, action potentials, intercellular Ca²⁺ waves and contractions in LSMC.

• Nitrergic neural inputs regulate the Ca²⁺ events in ICC-SS.

• Pacemaker activity in longitudinal muscle is an emergent property due to integrated processes in ICC-SS and LSMC.

Abbreviations

CM

Circular Muscle

FOV

Field of view

GI

Gastrointestinal

ICC

Interstitial Cells of Cajal

ICC-DMP

Interstitial cells of Cajal at the level of the deep muscular plexus

ICC-MY

Interstitial cells of Cajal at the level of the myenteric plexus

ICC-IM

Intramuscular interstitial cells of Cajal

ICC-SS

Sub-serosal interstitial cells of Cajal

GCaMP

Genetically encoded Ca²⁺ indicator composed of a single GFP

IP₃R

Inositol triphosphate receptor

KRB

Krebs Ringer Bicarbonate

LM

Longitudinal Muscle

PDGFRα

Platelet derived growth factor receptor α

PTCL

Ca²⁺ transient particle

ROI

Region of interest

SERCA

Sarco/endoplasmic reticulum Ca²⁺-ATPase

SIP syncytium

Electrical syncytium formed by smooth muscle cells, ICC and PDGFRα⁺ cells in GI muscles

SMC

Smooth muscle cell

SOCE

Store-operated-Ca²⁺ Entry

TTX

Tetrodotoxin