Interstitial cells of Cajal - pacemakers of the gastrointestinal tract

Abstract

Gastrointestinal (GI) organs display spontaneous, non-neurogenic electrical and mechanical rhythmicity that underlies fundamental motility patterns, such as peristalsis and segmentation. Electrical rhythmicity (aka slow waves) results from pacemaker activity generated by interstitial cells of Cajal (ICC). ICC express a unique set of ionic conductances and Ca²⁺ handling mechanisms that generate and actively propagate slow waves. GI smooth muscle cells lack these conductances. Slow waves propagate actively within ICC networks and conduct electrotonically to smooth muscle cells via gap junctions. Slow waves depolarize smooth muscle cells and activate voltage-dependent Ca²⁺ channels (predominantly Ca_V1.2), causing Ca²⁺ influx and excitation-contraction coupling. The main conductances responsible for pacemaker activity in ICC are ANO1, a Ca²⁺-activated Cl⁻ conductance, and Ca_V3.2. The pacemaker cycle, as currently understood, begins with spontaneous, localized Ca²⁺ release events in ICC that activate spontaneous transient inward currents due to activation of ANO1 channels. Depolarization activates Ca_V3.2 channels, causing the upstroke depolarization phase of slow waves. The upstroke is transient and followed by a long-duration plateau phase that can last for several seconds. The plateau phase results from Ca²⁺ induced Ca²⁺ release and a temporal cluster of localized Ca²⁺ transients in ICC that sustains activation of ANO1 channels and clamps membrane potential near the equilibrium potential for Cl⁻ ions. The plateau phase ends, and repolarization occurs, when Ca²⁺ stores are depleted, Ca²⁺ release ceases and ANO1 channels deactivate. This review summarizes key mechanisms responsible for electrical rhythmicity in gastrointestinal organs.

Graphical Abstract

A. Interval between slow waves. small amplitude spontaneous transient depolarizations (STDs) summate to generate upstroke B. Upstroke phase: activation of voltage dependent Ca²⁺ conductance and Ca²⁺ entry. C. Plateau phase: Ca²⁺ entry initiates Ca²⁺-induced Ca²⁺ release and activation of Ca²⁺-activated Cl conductance (ANO1 or a Ca²⁺-dependent nonselective cation conductance). D. Repolarization: When Ca²⁺ release from ER is exhausted, ANO1 is deactivated and cells repolarize to the inter-slow wave potential.

Introduction

Portions of the gastrointestinal tract display spontaneous, rhythmic electrical events, known as slow waves. Slow waves underlie important motility patterns, such as peristalsis in the corpus and antrum of the stomach, segmentation in the small intestine and segmental and propulsive contractions in the colon. Slow waves are periodic electrical depolarizations resulting from pacemaker activity intrinsic to interstitial cells of Cajal (ICC). ICC form electrically coupled networks of cells within the plane of the myenteric plexus in the corpus and antrum of the stomach, small intestine and colon (Fig. 1). These cells are known as ICC-MY. Another set of pacemaker cells lies along the submucosal surface of the circular muscle layer in the colon (ICC-SM). Conduction of slow waves to smooth muscle cells (SMCs) causes depolarization and increases the open probability of voltage-dependent Ca²⁺ channels (Ca_V1.2, encoded by Cacna1c) that are ubiquitously expressed by gastrointestinal SMCs.

Figure 1.

Interstitial cells (ICC-MY; Green) in the plane of the myenteric plexus (Red) in guinea pig ileum. ICC-MY form an extensive electrically-coupled network between and around myenteric ganglia (MG). Redrawn from Figure 3.3 in (Komuro, 2012).

Recognition that slow waves drive motility patterns motivated investigation of the pacemaker mechanism and the ion channels responsible for excitation-contraction (E-C) coupling. Older studies relied on sucrose gap or intracellular impalements of cells to measure slow wave activities. These techniques, essentially recording the integrated activities of many syncytial cells, were not adequate to determine either the cellular source of slow waves or their mechanisms. Enzymatic dispersion of GI muscles and utilization of the patch clamp technique on single cells revealed that SMCs do not generate slow waves or express conductances necessary to regenerate slow waves (Sanders et al., 2014). Such observations suggested that another type of cell provides pacemaker activity in GI muscles. SMCs are electrically coupled via gap junctions to two types of interstitial cells, ICC and cells known as platelet-derived growth factor receptor alpha positive (PDGFRα⁺) cells, which together form a multicellular electrical syncytium, known as the SMC: ICC: PDGFRα⁺ cell (SIP) syncytium (Sanders et al., 2012; Sanders et al., 2014). ICC are the component of the SIP syncytium that generates and propagates slow waves (Langton et al., 1989; Ward et al., 1994; Huizinga et al., 1995; Sanders et al., 2014) (Fig. 2).

Figure 2.

Comparison of slow waves recorded from ICC-MY in rabbit (A) and mouse (B) small intestine in situ. Slow wave parameters are noted: (1) upstroke depolarization; (2) duration of slow wave at 50% repolarization – aka half duration; (3) plateau amplitude; (4) inter-slow wave interval. Note that upstroke phase of slow wave in rabbit ICC-MY(A) repolarized significantly before plateau phase was initiated. Slow waves in mouse ICC-MY (B) did not display this repolarization phase. Redrawn from Figure 1 in: (Kito et al., 2015)

Finding that c-Kit is a biomarker for ICC and necessary for the development of these cells made it possible to utilize loss-of-function Kit mutants to study the roles of ICC in GI muscles (Ward et al., 1994; Huizinga et al., 1995; Torihashi et al., 1995). These experiments provided the first definitive evidence that ICC are pacemakers in GI muscles. A few groups succeeded in making microelectrode impalements of ICC within intact muscles and confirmed that slow waves originated in these cells and conducted to SMCs (Dickens et al., 1999; van Helden et al., 2000; Kito & Suzuki, 2003b; Kito et al., 2015). These studies also demonstrated the importance of Ca²⁺ handling and the likely contribution of a Cl⁻ conductance in slow waves. Generation of a reporter strain of mice that expressed copGFP driven by the endogenous promoter for Kit made it possible to recognize ICC in the mixed cell populations that occurred after enzymatic digestion of GI muscles (Ro et al., 2010), and patch clamp experiments performed on ICC further demonstrated the importance of Ca²⁺-dependent conductances in the generation of electrical rhythmicity (Goto et al., 2004; Zhu et al., 2009). More recent experiments using genetically-encoded Ca²⁺ sensors (GCaMPs) expressed selectively in ICC made it possible to characterize the Ca²⁺ handling mechanisms responsible for pacemaker activity in ICC (Drumm et al., 2017; Baker et al., 2021a; Baker et al., 2021b).

Syncytial network of cells makes it possible for ICC to pace SMCs

GI muscles depend upon gap junctions between ICC and SMCs to allow slow waves excitation-contraction (E-C) coupling in SMCs. Injection of current into ICC causes conduction of electrotonic potentials to SMCs (Cousins et al., 2003), and a similar decay in amplitude occurs when slow waves are conducted to SMCs (Dickens et al., 1999; Cousins et al., 2003; Kito et al., 2005; Kito et al., 2015) (Fig. 3). No mechanism for regeneration of slow waves in SMCs has been reported. So active propagation of slow waves in GI organs occurs through the network of electrically coupled ICC. This is an important distinction between heart and GI muscles. Because atrial and ventricular myocytes can regenerate cardiac action potentials, replacing damaged pacemaker cells in heart can be accomplished by electrically pacing the myocardium. This approach does not work in GI muscles; the continuity of ICC-MY networks (and ICC-SM in the colon) is necessary for normal generation and transmission of slow waves and coordination of contractions. Disruptions in the continuity of the ICC networks, as in diabetes (Ordog et al., 2000) or after surgical resection (Moon et al., 2022), disrupts normal propagation and negatively impacts patterns of motility.

Figure 3.

Electrical activity recorded simultaneously from an ICC-MY (A) and SMC (B) in Guinea pig gastric antrum. Events denoted by black bars in A&B are superimposed in C. The events occurred in phase, with ICC-MY slightly preceding the event in the SMC. Events were recorded in the presence of nifedipine (1 mM). Time scale in B also applies to trace in A. Redrawn from Figure 3 in (Dickens et al., 1999)

Gap junctions in mouse small intestinal SMCs and ICC are encoded by Gja1 (Cx43) and Gjc1 (Cx45) and both cells express these connexin genes (Lee et al., 2017). Immunohistochemical studies have also reported expression of connexins in various regions of the GI tract (Seki & Komuro, 1998, 2001; Wang & Daniel, 2001; Cousins et al., 2003). Few studies, other than the use of non-specific antagonists of gap junction coupling, have been performed on regulation of electrical coupling in GI muscles. The mechanisms responsible for the formation of gap junctions between cells of the SIP syncytium during development, factors regulating coupling moment-to-moment during physiological behaviors, and alterations in coupling that possibly occur in disease remain poorly understood.

Conductances involved in GI pacemaking

Ca²⁺-activated conductance(s) necessary for pacemaker activity

Electrophysiological studies of GI muscle strips tested many ideas about the ionic conductances and transporters responsible for slow waves, but space is too short to review the full history of investigation here. These studies are described in prior reviews (Connor et al., 1974; Tomita, 1981; Prosser, 1982; Szurszewski, 1987; Sanders et al., 2014). Major obstacles for early studies were: i) syncytial nature of GI muscles (i.e. SIP syncytium); ii) inadequate voltage-clamp of cells in GI muscles; iii) lack of understanding of the molecular phenotypes of pacemaker cells; and iv) lack of knowing whether drug effects were due to effects on SMCs or on other cells in the SIP syncytium. Direct intracellular recording from ICC within intact muscles improved the precision of electrical recordings (Dickens et al., 1999; Kito et al., 2002; Kito & Suzuki, 2003b; Kito et al., 2005; Kito et al., 2014; Kito et al., 2015) (Fig. 2; Table 1). Pharmacological tests in these studies suggested that voltage-dependent Ca²⁺ influx and Ca²⁺-activated Cl⁻ channels contribute to the upstroke and plateau components of slow waves. From negative inter-slow wave potentials that differ from site to site in GI muscles, slow waves depolarize ICC by at least 60 mV to a maximum level between −10 and −20 mV. No experimental evidence was obtained suggesting that outward currents are responsible for repolarization, and electrophysiological studies of ICC in situ did not reveal why the plateau phase of slow waves was sustained for a second or more. Intracellular recording from ICC, which are tiny cells that make up less than 10% of the tissue mass, is very difficult, limiting the range of experiments that could be performed with this technique. Furthermore, electrical coupling between cells of the SIP syncytium meant that it was difficult to be sure that effects of drugs or treatments occurred selectively on a specific type of cell. Patch clamp experiments performed on isolated ICC yielded direct measurements of slow waves and revealed the conductances responsible for this activity.

Table 1.

Properties of spontaneous slow waves recorded from small intestinal and gastric ICC-MY and from ICC-MY isolated from mouse small intestine by enzymatic dispersion

Species	Inter-slow wave membrane potential (mV)	Membrane potential at peak of plateau phase (mV)	Frequency (min−1)	Half duration (sec)	dV/dtmax (V sec−1)	Number of tissues/cells
Rabbit*	−66 +/− 4	−16	10.6 +/− 1.8	2.47 +/− 0.32	10.7 +/− 1.7	38
Mouse*	−67 +/− 3	−12	24.4 +/− 1.9	0.81 +/− 0.08	2.2 +/− 0.3	22
Guinea Pig+	−66 +/− 1.1	−20	2.5	NR	1.1 +/− 0.2	14
Mouse&	−72.5 mV	−1.6 +/− 3.8	26.8 +/− 6.2	0.553 +/− 174	6.7 +/− 2.5	7

Values are means +/− SD, where reported; number of cells also reflects number of animals used; dV/dt_max is the maximum rate-of-rise of the upstroke phase of slow waves; NR – not reported

^*Data from intracellular recordings from ICC within intact mouse and rabbit small intestinal muscles (Kito et al., 2015)

⁺Data from intracellular recordings from ICC within intact guinea pig gastric muscles (Dickens et al., 1999)

^&Data from perforated patch recordings from enzymatically dispersed cells (Goto et al., 2004)

Enzymatic dispersion of canine colonic muscles resulted in typical SMCs, and multi-polar cells with ultrastructural features of ICC (e.g. clusters of mitochondria, caveolae and cisternae of smooth endoplasmic reticulum, and abundance of thin filaments but few thick filaments and dense bodies) (Langton et al., 1989). At the time morphological criteria, however non-specific, represented the most reliable means of identification of ICC. ICC from canine colon displayed voltage-dependent currents under voltage clamp and rhythmic depolarizations similar in waveform to slow waves recorded from intact muscles under current clamp (Langton et al., 1989). Rhythmic electrical events with these characteristics were not recorded from SMCs. It was much easier to identify ICC when a biomarker, c-Kit, for these cells was identified (Ward et al., 1994; Huizinga et al., 1995; Torihashi et al., 1995). ICC-MY were identified with c-Kit antibodies after enzymatic dispersion of mouse small intestine, and patch clamp recordings were performed (Goto et al., 2004). Spontaneous rhythmic depolarization events were observed in some cells (Table 1). These events were preceded by small amplitude spontaneous transient depolarizations (STDs). Under voltage-clamp, depolarization elicited inward currents with an autonomous nature (Fig. 4). Once evoked the inward current was sustained for about 500 ms regardless of the duration of the depolarizing pulse. The autonomous current displayed inactivation properties, as the amplitude of the current was reduced by changing holding potential from −80 to −40 mV, and hyperpolarizing pre-pulses removed inactivation. The autonomous current reversed at about 0 mV and was depressed in amplitude and duration by omission of extracellular Ca²⁺ from the bathing solution. These studies concluded that the autonomous current was a Ca²⁺ dependent, non-selective cation conductance, based on the reversal potential of the current and a Cl⁻ equilibrium potential assumed from other studies (Huizinga et al., 2002). The autonomous nature of the conductance differed from the kinetics of known voltage-dependent conductances.

Figure 4.

Autonomous inward (AI) currents recorded under voltage clamp from enzymatically dispersed ICC-MY from mouse small intestine (see original article for description of Methods). Cell was held at −80 mV. Hyperpolarization to −100 or depolarizations to −70 and −60 mV failed to evoke AI. Depolarization steps (−50 - −10 mV) activated AI that persisted longer than the duration of the voltage steps. Dotted line 1 shows amplitude of AI at each potential. AI reversed at about 0 mV and was predominantly outward at +10 and +20 mV. Dotted line 2 shows amplitude of maximal AI upon return to holding potential. AI persisted for an average duration of about 500 mS (see Table I for additional properties). Redrawn from Figure 5 in (Goto et al., 2004)

An array study of expressed genes in ICC of the mouse small intestine, revealed that TMEM16a is highly expressed (Chen et al., 2007), but the function of the encoded protein was unknown at the time these expression studies were performed. A breakthrough in understanding the ionic conductances involved in electrical rhythmicity in ICC was based on the recognition that TMEM16a (referred to from here as ANO1) encodes Ca²⁺-activated Cl⁻ channels (Caputo et al., 2008; Schroeder et al., 2008; Yang et al., 2008). Immunohistochemical studies showed that ANO1 protein is ubiquitously and uniquely expressed in ICC at all levels of the GI tract in mouse, monkey and human (Gomez-Pinilla et al., 2009; Hwang et al., 2009; Rhee et al., 2011; Blair et al., 2012). Several splice variants are differentially expressed in regions of the GI tract (Hwang et al., 2009), but the functional significance of these differences is poorly understood. The expression of ANO1 in ICC was consistent with the pharmacological and ion replacement studies performed previously by the research teams of David Hirst, Hikaru Suzuki and Dirk van Helden, suggesting that pacemaker activity is dependent upon a Ca²⁺-dependent Cl⁻ conductance (Edwards et al., 1999; van Helden et al., 2000; Hirst et al., 2002; Kito et al., 2002; van Helden & Imtiaz, 2003).

Mice expressing copGFP in ICC have served as a powerful tool for studies of ICC because they have simplified isolation and purification of ICC by fluorescence-activated cell sorting and for genome-wide expression studies (Ro et al., 2010). Isolated ICC from mouse small intestine exhibited a rounded appearance with a cell capacitance averaging 5.5 pF (Zhu et al., 2009). Whole-cell patch clamp recording from these cells showed currents similar to those observed by Goto and colleagues (Goto et al., 2004), however the properties of the currents were consistent with a Ca²⁺-activated Cl⁻ conductance. Reversal potentials followed the Cl⁻ equilibrium potential, and the currents were blocked by removal of extracellular Ca²⁺ or replacement of Ca²⁺ with Ba²⁺. The inward currents were also blocked by Ni²⁺ suggesting, as experiments on intact muscles had, that initiation of pacemaker activity in ICC includes activation of a voltage-dependent Ca²⁺ conductance. In current clamp, ICC generated spontaneous depolarization events with characteristics similar to slow waves in intact muscles, and these events were blocked by niflumic acid. Because of the specific expression of ANO1 in ICC, disruption of pacemaker activity by genetic deactivation of ANO1 (see below) and characteristics of inward currents activated by depolarization, the autonomous currents attributed to the Ca²⁺-activated Cl⁻ conductance in ICC were referred to as slow wave currents.

Gene knockout experiments confirmed the importance of ANO1 in electrical slow waves in small intestine and stomach. A congenital knock-out, developed by Jason Rock (Rock et al., 2008), displayed normal networks of ICC-MY in Ano1^−/− mice, but slow waves failed to develop in these animals (Hwang et al., 2009). Congenital knock-out of Ano1 causes a high mortality rate in mice; 80% of homozygote null mice die within the first week after birth. Slow waves developed normally in wildtype and Ano1^+/− siblings but were absent in Ano1^−/− mice. In a few mice that survived for at least 3 weeks, slow waves with adult-like amplitude and frequency developed in small intestinal and gastric muscles of wildtype and Ano1^+/− animals, but still failed to develop in Ano1^−/− mice.

Pharmacological studies performed on adult mice were less convincing about the role of ANO1 in pacemaker activity (Hwang et al., 2009). At the time of these studies antagonists of ANO1 were relatively non-selective, but an obvious difference in sensitivity was observed between small intestine and gastric muscles. Relatively high concentrations of niflumic acid or 4,4′-diisothiocyano-2,2′-stillbene-disulfonic acid (DIDS) were required to affect the frequency or block slow waves in muscles of mouse, monkey or human jejunum, whereas much lower concentrations of these drugs were effective in reducing slow wave frequency or blocking slow waves in gastric antral muscles. Differences in sensitivity to ANO1 antagonists of small intestinal and antral muscles were also apparent when more potent and selective 2^nd generation drugs (Namkung et al., 2011; Huang et al., 2012) were tested (Hwang et al., 2016). Ca²⁺-activated Cl⁻ channel inhibitors, like CaCC_inh-A01 and benzbromarone, were effective blocking slow waves in gastric muscles at low micromolar concentrations. The same compounds were less effective on small intestinal slow waves. Reasons for these pharmacological differences in sensitivity were unclear, but one study suggested that differences in Ano1 splice variants and/or levels of intracellular Ca²⁺ might affect the potency of ANO1 antagonists (Sung et al., 2016).

The role of ANO1 in slow waves was also tested using Cre/loxP technology to induce knockdown of Ano1 in ICC of mature animals (iAno1 KO). Ano1 expression was reduced by tamoxifen treatment but not abolished. ANO1 immunoreactivity in small intestines of iAno1 KO mice occurred in a non-uniform, mosaic pattern. Electrophysiological effects of iAno1 KO in small intestine were variable. Slow waves were not resolved during intracellular recording from about one quarter of SMCs impaled, but these events persisted in the remainder of cells (Malysz et al., 2017). Some SMCs displayed slow waves of irregular amplitude, and adjacent regions of muscle from other animals appeared to be uncoupled and displayed slow waves of increased frequency and irregular amplitudes. In muscles with relatively high levels of ANO1 immunoreactivity, slow wave amplitude and frequency were normal and well synchronized with Ca²⁺ transients in ICC.

Another study utilizing iAno1 KO mice confirmed the effects of ANO1 knockdown on small intestinal muscles (Hwang et al., 2019). However, this study also compared the effectiveness of Ano1 knockdown in small intestine and gastric antrum. Although the efficiency of Ano1 knockdown was similar in both regions, slow waves were present in most small intestinal muscles and abolished in antral muscles. Contractions were also abnormal in gastric muscles of iAno1 KO mice, and defects in slow waves and contractions markedly delayed gastric emptying and disrupted patterns of gastric motility (Hwang et al., 2019).

Studies on inducible knockdown of Ano1 were consistent with different sensitivities of small bowel and stomach to ANO1 antagonists. The original studies of congenital Ano1 KO mice showed that slow waves are abolished in juvenile small intestinal and gastric muscles, but inducible knockdown experiments were conducted on adult mice. Thus, it is possible the ANO1 is important for development of slow waves in the small intestine, but another conductance may develop as mice mature. The sensitivity of slow waves in the small bowel to ANO1 antagonists therefore may decrease with age. This was demonstrated by comparing the effects of CaCC_inh-A01 on small intestinal muscles from juvenile and adult mice. CaCC_inh-A01 inhibited slow wave amplitude and frequency in juvenile small intestinal muscles in a concentration-dependent manner, with total block at 5 μM. Adult muscles were less sensitive to CaCC_inh-A01. Development of a conductance in addition to ANO1 in small intestinal ICC might explain the reduced sensitivity of ANO1 antagonists and may explain the different properties of the conductances observed in studies of juvenile (Zhu et al., 2009) and adult small intestinal ICC (Goto et al., 2004). This question was addressed by patch clamp studies of adult small intestinal ICC (Hwang et al., 2019). In cells with E_Cl adjusted to −40 mV, large amplitude inward currents were observed at negative holding potentials, and amplitudes decreased with depolarization. At −35 mV, a small outward current preceded the inward current. The inward current at potentials positive to −40 mV was blocked when [Na⁺]_i was reduced to 0 mM. These findings suggest that small intestinal ICC develop a non-selective cation conductance as animals mature, and this is likely the conductance described by Goto and colleagues (Goto et al., 2004). The molecular identity of the channels responsible for this conductance has not been determined.

Voltage-dependent conductances necessary for coordination of Ca²⁺-activated conductances

Slow waves propagate over significant distances in GI muscles to coordinate the contractions responsible for peristaltic and segmental patterns. Propagation of slow waves is an active process, requiring cell-to-cell regeneration within ICC networks. Active propagation is due to voltage-dependent inward current(s) that cause the upstroke depolarization of slow waves. In canine gastric muscles slow waves propagate at 65 mm s⁻¹ in the axis of the circular muscle (CM) and depend upon [Ca²⁺]_o (Bayguinov et al., 2007). Slow waves were blocked when [Ca²⁺]_o was reduced to 0.5 mM. Nicardipine did not affect propagation velocity or the rate-of-rise of the upstroke depolarization, but T-type Ca²⁺ antagonists slowed upstroke velocity, reduced propagation velocity and inhibited slow wave propagation at higher doses tested. Slow wave propagation is dependent upon T-type Ca²⁺ channels in stomach and small intestine, but ICC along the submucosal surface in the colon (ICC-SM) have a mixed dependence on T-type and L-type currents for slow wave propagation (Baker et al., 2021b). This is an asset for colonic pacemaker activity. Membrane potentials of colonic muscles are in a range where most T-channels are inactivated, but L-type channels, which inactivate at less negative potentials, provide the inward current necessary for slow waves. If neural or hormonal inputs hyperpolarize the SIP syncytium and reduce the open probabilities of L-type channels, inactivation of T-type channels is reduced, and they become available to contribute to the slow wave upstroke. Having these 2 conductances with different activation/inactivation properties provides a safety factor, ensuring a broader range of potentials over which pacemaker activity can successfully active slow waves.

Integration of Ca²⁺ signaling and activation of conductances that generate waveforms of slow waves

There are 2 general classes of ICC in GI muscles, cells that mediate inputs from enteric motor neurons and cells that generate pacemaker activity (Sanders et al., 2014). All types of ICC display spontaneous Ca²⁺ transients that couple to activation of ANO1 channels. Activation of ANO1 channels causes spontaneous transient inward currents (STICs) and the voltage-responses to STICs are spontaneous transient depolarizations (STDs). Recordings from ICC in situ or isolated ICC have shown that STDs precede generation of slow waves (Kito & Suzuki, 2003a; Goto et al., 2004; Kito et al., 2005; Kito et al., 2015). The important difference between ICC responding to neural inputs and cells generating pacemaker activity is the expression of the voltage-dependent conductance (as described above) by the pacemaker cells. Thus, the underlying pacemaker mechanism in ICC is stochastic Ca²⁺ transients and generation of STICs, but these events do not produce slow waves unless the STDs are able to activate a voltage-dependent conductance. As described above, the voltage-dependent conductances in pacemaker ICC are Ca²⁺ conductances, and entry of Ca²⁺ is also an important step in determining the ultimate waveforms of slow waves.

Slow waves typically exceed 1 sec in duration and can persist for many secs in muscles of some species (Szurszewski, 1987; Sanders, 2019). The durations of slow waves are inconsistent with the kinetics of voltage-dependent inward or outward current channels, and this raises further questions about the mechanism of slow waves, such as: i) to what is the long-duration plateau phase attributed? and ii) what delays the repolarization phase? Pharmacological studies suggested that the long duration of slow waves is due to a Ca²⁺-dependent mechanism, and both Ca²⁺ entry and Ca²⁺ release from stores were implicated (Goto et al., 2004; Bayguinov et al., 2007; Zhu et al., 2015; Baker et al., 2021a). Current information suggests that sustained activation of a Ca²⁺-activated Cl⁻ conductance (ANO1) is responsible for the plateau phase and long durations of slow waves (Zhu et al., 2015; Hwang et al., 2016) (Fig. 5), however a Ca²⁺-dependent non-selective current may also contribute in small intestine (as above). Sustained activation of Ca²⁺-dependent conductances requires mechanisms to maintain elevated levels of [Ca²⁺]_i. Recent optogenetic experiments with ICC-specific expression of GCaMPs have provided insights on how [Ca²⁺]_i is sustained for several seconds to establish the durations of slow waves.

Figure 5.

Effects of an ANO1 antagonist on slow waves recorded from SMCs of the mouse gastric antrum. A shows excerpts of a continuous recording of slow waves before and after addition of 1-5 mM CaCC_inh-A01. B & C show tabulations of the effects of the ANO1 antagonist on slow wave frequency and half duration. ANO1 antagonist reduced slow wave frequency and the duration of the plateau phase. Redrawn from Figure 1 in (Hwang et al., 2016)

Ca²⁺ handling mechanisms in ICC have been reviewed in detail elsewhere (Sanders et al., 2023). Briefly, all ICC, studied to date, generate stochastic Ca²⁺ transients due to release from intracellular stores. The Ca²⁺ transients are responsible for the STICs and STDs generated by ICC (Zhu et al., 2015). As discussed above, these events are the driver that activates the upstroke phase of slow waves, and Ca²⁺ entry initiates Ca²⁺-induced Ca²⁺ release, setting off a cluster of localized Ca²⁺ transients from many sites within the ICC network (Drumm et al., 2017; Baker et al., 2021a; Baker et al., 2021b) (Fig. 6). The clusters of Ca²⁺ transients, maintain increased open probability of Ca²⁺-dependent channels within the ICC network. The durations of slow waves are defined by the durations of the Ca²⁺ transient clusters (Drumm et al., 2017). As Ca²⁺ release events decrease and Ca²⁺ transient clusters terminate, Ca²⁺-dependent conductances deactivate and slow wave repolarization occurs. After termination of a Ca²⁺ transient cluster, store refilling is needed, and a refractory period ensues, during which the probability of Ca²⁺ transients is low. It should also be noted that a substantial slow wave refractory period has also been noted in electrophysiological recordings, during which time the upstroke phase resets rapidly (consistent with the relatively rapid removal of inactivation of T-type Ca²⁺ channels), but the plateau phase takes several seconds for full restoration (Publicover & Sanders, 1986; Suzuki & Hirst, 1999; Nose et al., 2000). Thus, the long duration of slow waves is a result of the Ca²⁺ handling mechanisms that support activation of the Ca²⁺-dependent conductance(s).

Figure 6.

Clustering of Ca²⁺ transients in ICC defines the duration of the slow wave. A shows 2 cells from a mouse small intestine with GCaMP3 expression in ICC. ICC-MY were imaged with confocal microscopy at 60x. Ca²⁺ transients occur in clusters from multiple sites in these cells. B shows a heat map documenting the accumulation of Ca²⁺ transients during the recording period. C shows particle analysis of Ca²⁺ transient sites as determined from B. Each particle is color coded and then lined up in occurrence plots, as in D. Occurrence plots show the temporal characteristics of each Ca²⁺ transient particle in C. From another experiment E shows simultaneous recordings of a slow wave cycle (black trace) and summation of Ca²⁺ transient particles (red trace) that occurred during the slow wave. The upstroke of the slow wave preceded the initiation of the cluster of Ca²⁺ transients by about 175 ms. Ca²⁺ entry during the upstroke depolarization (which is due to activation of voltage-dependent Ca²⁺ channels) sets off Ca²⁺ release events that cause sustained activation of ANO1. At completion of Ca²⁺ release events, slow wave repolarizes to inter-slow wave potential. F shows occurrence plot of Ca²⁺ transient particles that occurred during the slow wave in E. Redrawn from Figs. 3 & 5 in: (Drumm et al., 2017)

Restorative mechanisms must be available to maintain ionic gradients and Ca²⁺ stores necessary for pacemaker activity and active propagation. For a Cl⁻ conductance to yield inward currents, E_Cl must be maintained positive to the inter-slow wave membrane potentials of ICC. This is accomplished by the actions of the Na⁺/K⁺/2Cl⁻ cotransporter (NKCC1) that is highly expressed in pacemaker ICC (Wouters et al., 2006; Kito et al., 2015; Zhu et al., 2016). Antagonists of NKCC1 caused a negative shift in E_Cl, reduced the amplitude of STICs and STDs, reduced or blocked slow waves and inhibited the underlying slow wave currents in ICC. Genetic deactivation of Slc12a2 (gene encoding NKCC1) resulted in small amplitude slow waves and loss of sensitivity to bumetanide (Wouters et al., 2006).

A consequence of Cl⁻ recovery by NKCC1 is the co-transport of Na⁺ ions into cells. Ca²⁺ release is localized suggesting that it occurs in microdomains due to close associations between the plasma membrane and endoplasmic reticulum (ER/PM junctions). These sites are likely to be where Ca²⁺-dependent conductances are activated by Ca²⁺ transients, and within this restricted volume recovery of Cl⁻ might be accompanied by periodic increases in [Na⁺]_i. ICC express Slc8a1-3, encoding Na⁺/Ca²⁺ exchangers (NCX1-3). During the inter-slow wave interval, NCX may serve to maintain low levels of Ca²⁺ restricting the activation of Ca²⁺-dependent conductances (Zheng et al., 2020). However, NCX might flip into Ca²⁺ entry mode during slow wave depolarization to between −10 and −20 mV (see Table 1) and from the influx of Na⁺, the consequences of which are discussed below.

Another restorative process involves reloading of Ca²⁺ stores. While multiple sources of Ca²⁺ might contribute to store replenishment, the STIM/Orai mechanism of store-operated Ca²⁺ entry appears very important (Zheng et al., 2018). All paralogs of Stim and Orai are expressed in small intestinal ICC, and depletion of Ca²⁺ stores yields Ca²⁺ influx (due to a Ca²⁺ release-activated Ca²⁺ current or I_CRAC). I_CRAC was blocked in ICC by an Orai antagonist or a dominant-negative STIM1 peptide. Inhibition of Orai Ca²⁺ channels caused inhibition of slow wave currents in single ICC and rapid run-down of clustered Ca²⁺ transients in ICC in situ. SarcoEndoplasmic reticulum Ca²⁺ ATPase (SERCA) is also necessary in this phase of restoration for uptake of Ca²⁺ into the ER.

Ca²⁺ entry through T-type Ca²⁺ channels is brief due to relatively fast voltage-dependent inactivation (Perez-Reyes, 2003). Thus, Ca²⁺ entry during the upstroke phase of slow waves does not explain why Ca²⁺ transient clusters persist for a second or more. We have suggested that there are additional Ca²⁺ entry mechanisms that serve to sustain Ca²⁺ induced Ca²⁺ release. As discussed above, when Ca²⁺ stores begin to empty, STIM/Orai interaction are likely to provide a source for Ca²⁺, however this mechanism may not be functional until store unloading makes Ca²⁺-induced Ca²⁺ release unlikely. The membrane potential level reached during the plateau phase of slow waves is within the range of potentials where L-type Ca_V1.2 channels are activated, but inactivation is incomplete, generating sustained window current (Cohen & Lederer, 1987; Fleischmann et al., 1994). Ca²⁺ entry due to window current might be capable of sustaining Ca²⁺-induced Ca²⁺ release. In gastric muscles, for example, nifedipine had no effect on the upstroke depolarization, but reduced the amplitude and duration of the plateau phase (Bayguinov et al., 2007) and shortened the duration of Ca²⁺ transient clusters in gastric ICC (Baker et al., 2021a). Dihydropyridines do little to slow waves in ICC-MY in the small intestine, and nicardipine did not affect the amplitude or duration of Ca²⁺ transient clusters (Drumm et al., 2017). Modeling of pacemaker activity in small bowel suggested that Na⁺/Ca²⁺ exchange (NCX) is liable to provide Ca²⁺ influx into ICC during the plateau phase (Youm et al., 2019). Na⁺ entry via NKCC1 (see above) and depolarization to about −10 mV during slow waves appears to flip NCX from Ca²⁺ extrusion to Ca²⁺ entry (Zheng et al., 2020). NCX3 was found by proximity ligation assay to be closely associated with ANO1 channels in ICC, possibly in ER/PM junctions, and antagonists of NCX inhibited and shortened the durations of slow wave currents and decreased the durations of Ca²⁺ transient clusters in ICC. Reduction in elevated [Na⁺]_i is accomplished by the α2 isoform of the sodium pump (encoded by Atp1a2), and a low concentration of ouabain increased the amplitude and duration of slow wave currents (Zheng et al., 2020).

Looking ahead in GI pacemaker research

Loss or damage to the integrity of ICC networks has been reported in a variety of GI motility disorders (see summary in Table 5 in (Sanders et al., 2023). Morphological damage or loss of ICC are the only assays currently used to determine the fate of ICC in disease, but remodeling of cells may also occur and include breakdown in mechanisms responsible for generation or propagation of slow waves. As discussed, key proteins driving pacemaker activity are likely clustered into ER/PM junctions (Zhu et al., 2015), and modeling of pacemaker activity supports this hypothesis (Youm et al., 2019). Pharmacology and gene deactivation studies suggest that ER/PM junctions contain the ion channels and transporters illustrated in (Fig. 7), and close associations between these proteins in the excluded volume of a microdomain may be necessary for pacemaker function. Characterizing the structure and function of ER/PM junctions is an important next step in understanding GI pacemaking, because then it will be possible assess whether changes in the ER/PM phenotype, as might occur with inflammation (Won et al., 2006; Kaji et al., 2016) or other factors, could be a source of motility dysfunction. Little is currently known about the integration of functions amongst key proteins in ER/PM junctions of ICC and there is much to be learned about what other proteins might provide essential regulation of function.

Figure 7.

Minimum constituents of ER/PM junctions, as supported by experimental evidence. Expression analysis has shown multiple paralogs of many components, however when one paralog is dominant, it is designated. Specific organization of proteins in ER/PM junction has not been determined. Conductances required for pacemaker activity (Pacemaker components): Ca²⁺ is released from IP₃R1 in all ICC (RyR also participates in some cells) and activates ANO1, creating STICs and STDs. STDs summate and depolarization activate Ca_V3.2 to produce the slow wave upstroke. Ca²⁺ entry sets off Ca²⁺-induced Ca²⁺ release, activating ANO1 through the plateau phase of slow waves. Restorative components: Cl⁻ lost from efflux through ANO1 is recovered by NKCC1. This also causes Na⁺ entry into the excluded volume and flips NCX into Ca²⁺ entry mode. This extends duration of plateau potential. Na⁺ is extruded by a2 isoform of sodium pump (NKX). In some ICC, Ca_V1.2, activated by depolarization during the plateau phase (window current) also serves to preserve the duration of the plateau phase. When Ca²⁺ stores are exhausted, ANO1 is deactivated and membrane potential is restored to inter-slow wave level (slow wave repolarization). STIM binds and activates ORAI, causing Ca²⁺ entry. Ca²⁺ are replenished through uptake by SERCA.

How frequency and rhythmicity are regulated is another important topic for investigation. Pacemaker ICC live in a complex environment. They are electrically coupled to cells that produce stochastic inward (intramuscular ICC) and outward currents (PDGFRα⁺ cells). Neural and other inputs control the output of these current generators and modifications in these inputs could conceivably alter GI rhythmicity or shut it off entirely. As stated above, major GI motility patterns, such as peristalsis and segmentation, depend upon the cyclic depolarization/repolarization of slow waves. Pacemaker activity persists during a variety of gut stimuli, and we currently don’t know the impact of many conditions, such as stretch, on pacemaker function. Persistence of pacemaker activity in the face of many dynamic changes appears to be due to a broad-range safety factor achieved by the conductances involved in pacemaker activity (voltage-dependent Ca²⁺ conductances) and relative independence of Ca²⁺-dependent conductances (ANO1or non-selective conductances) from changes in membrane potential and membrane resistance. More information is needed about the resilience of the pacemaker mechanism to better understand how disease-activated factors (e.g. abnormal stretch, inflammatory cytokines, etc.) might affect pacemaker function.

There has been controversy in the literature regarding mechanisms of propagation of slow waves. While current evidence suggests that voltage-dependent Ca²⁺ conductances are responsible for propagation within continuous ICC networks (Sanders et al., 2014), voltage-dependent generation of IP₃ or sensitization of IP₃ receptors has also been proposed (van Helden et al., 2000; Hirst et al., 2006). Others have suggested that slow waves do not propagate as in other excitable cells but rather ICC behave as coupled oscillators and the time between occurrence of events at different points within muscles is due to phase lag (Sarna & Daniel, 1975; Daniel et al., 1994; Imtiaz et al., 2007; Imtiaz et al., 2010). Slow wave activity might be simulated by coupled oscillator equations, but this approach lacks adjustable parameters representing real biophysical entities, such as ion channels and transporters. Our view is that propagation occurs by a process of entrainment, where each slow wave cycle is initiated at a single cell or small cluster of cells, and the ensuing depolarization due to activation of voltage-dependent inward current is regenerated, cell-to-cell through the ICC network. A pacemaker mapping study performed many years ago showed that the site of origin of slow waves in a gastric muscle sheet was quasi-stable; slow waves were generated from one site for several cycles and then shifted to a new site for several cycles (Publicover & Sanders, 1984). As above, regulation of frequency in ICC in different regions of the GI tract is not fully understood, but we suggest this will be related to variations in Ca²⁺ dynamics: abundance and composition of ER/PM junctions, kinetics of store refilling between slow waves, dynamics and interactions of Ca²⁺ channels in the ER and availability of Ca²⁺ entry mechanisms that affect Ca²⁺ release.

Recent work on the sino-atrial node has led to the formulation of an intriguing model for cardiac pacemaking (Clancy & Santana, 2020; Grainger et al., 2021; Bychkov et al., 2020; Guarina et al., 2022). In this model, the periodic electrical signal that drives sino-atrial node pacemaking seems to emerge from a high degree of functional heterogeneity through the node. Accordingly, periodic pacemaking activity may not result from dissipation of electrical heterogeneity, as proposed by the entrainment model described above, but rather is caused by electrical heterogeneity through a stochastic resonance mechanism. At present, it is unclear whether a similar mechanism could tune ICC pacemaker activity. Because stochastic resonance is a new concept in GI pacemaker activity, a general definition of this phenomenon is called for before briefly delving into the details of how it could potentially lead to pacemaking in ICC. Stochastic resonance describes a process in which a small-amplitude voltage fluctuation is amplified by the assistance of biological noise. In general terms, a system can undergo stochastic resonance if it meets three conditions. First, the system must have a threshold for signal detection or generation. Second, it must be capable of generating a relatively weak periodic signal that fails to reach threshold, at least occasionally. Third, it must have a source of noise that adds to the periodic signal.

ICC meet these three conditions. A threshold depolarization must be accomplished to activate voltage-dependent Ca²⁺ channels. Ongoing variations in membrane potential largely fail to reach threshold, but there are multiple sources of noise from both ICC and other cells within the SIP syncytium that can influence whether threshold is reached. Within the pacemaker cells there are moment-to-moment variations in Ca²⁺ release from the ER, varying levels of Ca²⁺ in microdomains that influence openings of Ca²⁺ dependent channels in the plasma membrane, and dynamic changes in Ca²⁺ entry, extrusion and uptake (Sanders et al., 2023). Other cells within the SIP syncytium produce stochastic inward (ICC) (Zhu et al., 2015) or outward (PDGFRα⁺ cells) (Kurahashi et al., 2011) that likely have dynamic effects on the net membrane potential of the syncyium.

Thus, it would be interesting to determine the relationship between noise magnitude and ICC pacemaking periodicty. Stochastic resonance would predict that, when noise is low, the probability of ICC of reaching threshold for generation of slow waves is low. As noise levels increase, the frequency and periodicity of action potentials increases. However, as noise levels increase past an optimal point (i.e. resonance), it would dominate slow wave periodicity, making the timing of these signals more random and decrease the performance of the system (i.e. reduced periodicity). The model also predicts that activation of signaling pathways could shift this relationship to the left (i.e. less noise needed to reach threshold) or to the right (i.e. more noise needed). It is important to consider the intriguing possibility that entrainment and stochastic resonance mechanisms are interdependent, such that the capacity of the ICC to undergo entrainment is tuned by the amplitude of biological noise in different regions of the ICC network. Perhaps this could be responsible for the difference in pacemaker frequency between gastric corpus and antrum. Future work should provide an answer to this important question.

Finally, a major obstacle to learning about ICC impairment in human GI motility disorders is that it is not typically possible to study affected cells and tissues early in the disease process. By the time patient symptoms are severe enough to warrant surgery, ICC may be damaged beyond repair. Use of animal models, particularly those with cell-specific expression of optogenetic and other reporters of function, can indicate what defects develop in the pacemaker mechanism at the onset of motility disorders, and some of the damaging steps might be treatable. Development and careful phenotypic characterization of human organoids may also provide opportunities to better understand disease processes in human ICC (Workman et al., 2017). Studies seeking repair or restoration of ICC networks may also be necessary if cell damage cannot be prevented. This might be accomplished with stem cells that have been suggested to replenish the ICC phenotype (Lorincz et al., 2008). However, another possibility is that ICC are not actually killed in diseases affecting GI motility. Perhaps the cells are remodeled to a non-functional phenotype. Bowel obstruction, for example, was found to cause loss of c-Kit⁺ ICC, possibly due to inflammatory factors (Won et al., 2006), but these cells recovered when the obstruction was relieved (Chang et al., 2001). Studies of the plasticity of ICC in health and disease are needed to understand and possibly repair motility disorders associated with ICC loss-of-function.

Biographies

Author Profiles

Kenton M. Sanders, Ph.D. is Professor in the Department of Physiology and Cell Biology at the University of Nevada, Reno School of Medicine. He has contributed pioneering studies of the role of interstitial cells in the regulation of smooth muscle function in the gastrointestinal and urinary tracts.

Luis Fernando Santana, Ph.D. is Professor, Chair of Physiology and Membrane Biology and Vice Dean for Basic Sciences at the University of California, Davis School of Medicine. He uses a variety of techniques to study mechanisms regulating excitation-contraction coupling in cardiac and vascular smooth muscles.

Salah A. Baker, PhD, is an Associate Professor in the Dept of Physiology and Cell Biology at the University of Nevada, Reno. He studies the role of interstitial cells and Ca²⁺ signaling in controlling smooth muscle contractility in the gastrointestinal tract with emphasis on functions of the colon motor patterns, gastric peristalsis and small bowel segmentation.