Regulation of Gastrointestinal Smooth Muscle Function by Interstitial Cells

Abstract

Interstitial cells of mesenchymal origin form gap junctions with smooth muscle cells in visceral smooth muscles and provide important regulatory functions. In gastrointestinal (GI) muscles, there are two distinct classes of interstitial cells, c-Kit⁺ interstitial cells of Cajal and PDGFRα⁺ cells, that regulate motility patterns. Loss of these cells may contribute to symptoms in GI motility disorders.

Endothelial cells have long been recognized for their critical role in regulating vascular tone, but similar recognition of the important regulatory functions provided by various classes of interstitial cells in visceral smooth muscles has been slower to be accepted. Several classes of interstitial cells, referred to by a variety of terms, including interstitial cells of Cajal (ICC), ICC-like (ICC-LI) cells, PDGFRα⁺ cells, fibroblast-like cells, and teleocytes, have been described morphologically and provide various regulatory functions. Interstitial cells can contribute to the excitability of smooth muscle tissues because some classes of these cells form gap junctions with smooth muscle cells (SMCs). Interstitial cells express a variety of receptors for neurotransmitters, hormones, paracrine substances, and inflammatory mediators, and these cells also express second-messenger pathways and ion channels that allow them to mediate postjunctional responses to neurotransmission and other biological regulatory substances. Because of extensive morphological studies and some naturally occurring mutant animals with defects in ICC, interstitial cells in the gastrointestinal (GI) tract have been studied in greatest detail, but it should be noted that all smooth muscles display some types of interstitial cell populations. In many cases, the physiological functions and roles in pathophysiology of interstitial cells are still unknown. This short review describes various aspects of physiological regulation that have been associated with interstitial cells of GI muscles and how they enhance the motor behaviors of visceral smooth muscles.

Morphology of Interstitial Cells and Relation to Smooth Muscle Cells

Interstitial cells are distinctly different than SMCs, often displaying multiple processes and typically few thick filaments. Ultrastructural features of ICC include an abundance of mitochondria, moderately well developed Golgi, thin and intermediate filaments, and rough and smooth endoplasmic reticulum (ER) (23, 24, 66, 90). Some ICC display caveolae and a basal lamina. Mitochondria and cisternae of ER are often prevalent in the perinuclear region, and regions of close apposition between the ER and plasma membrane are common. Pacemaker functions and neural responses of ICC appear to depend on Ca²⁺ release mechanisms from internal stores (3, 39, 108, 125), and the close appositions between ER and the plasma membrane suggest that much of this signaling occurs in microdomains. There is an abundance of rough ER in PDGFRα⁺ cells, giving them a fibroblast-like appearance, and these cells also lack caveolae and a basal lamina. PDGFRα⁺ cells were referred to as fibroblast-like cells for many years, but now distinctive chemical coding (such as specific labeling with antibodies for PDGFRα) has provided a more precise means of referring to these cells.

ICC and PDGFRα⁺ cells form gap junctions with SMCs (41, 67). SMCs, ICC, and PDGFRα⁺ cells express numerous gap-junction genes and proteins (15, 35, 98). Electrical coupling causes the interstitial cells and SMCs to function as a multicellular syncytium we have called the SIP syncytium (94). This structure serves as the pacemaker in GI muscles and transducer of neural and other regulatory inputs. Electrical coupling allows conductance changes in one type of cell to affect the excitability of the other types of cells in the SIP syncytium.

Intramuscular ICC (ICC-IM) and ICC clustered within the region of the deep muscular plexus in the small intestine (ICC-DMP) are very closely associated with varicosities of motoneurons in the tunica muscularis. The close associations with neural processes were common observations in many classical ultrastructural studies of ICC (17, 21, 51, 90). Structured junctions between ICC and nerve varicosities can be found with spacing of ∼20 nm, and pre- and postjunctional synaptic proteins are present (10). Connectivity of this sort can also be found between varicosities and SMCs (28, 79); however, the only morphometric study comparing connectivity of neurons with ICC and SMCs found far more frequent junctional connections between ICC and nerve varicosities than between neurons and SMCs (16). If the regions of close spacing are sites of neurotransmitter release, then very high concentrations of transmitters might be achieved at postjunctional receptors, and metabolism or uptake of transmitters might also be accelerated. FIGURE 1 shows representative images of ICC and PDGFRα⁺ cells using immunohistochemical techniques, and their relationships with each other and the processes of enteric motoneurons.

FIGURE 1.

ICC and PDGFRα⁺ cells in the small intestine

A and B: whole mounts of c-Kit⁺ ICC-MY (arrows) and ICC-DMP (arrowheads) in the murine small intestine. Note the well defined network structure of ICC-MY, and the more distributed and parallel orientation of ICC-DMP that run parallel to the circular muscle and are closely aligned with nerve processes of the deep muscular plexus. C and D: images from the small intestine of Macaca fascicularis (cynomolgus monkey) showing relatively the same distributions of ICC-MY (arrows) and ICC-DMP (arrowheads) in the primate intestine. E and F: distributions of PDGFRα⁺ cells at the level of the myenteric plexus (E; PDGFRα⁺-MY; arrows) and PDGFRα⁺ cells in the deep muscular plexus (F; PDGFRα⁺-DMP; arrowheads) in the murine intestine. Intramuscular ICC and PDGFRα⁺ are closely associated with enteric motor nerves. G: the relationship between ICC-DMP (green; arrows) in very close contact with inhibitory motor neurons highlighted by an antibody for nNOS (red; arrowheads). ICC and PDGFRα⁺ cells are distinct populations of interstitial cells. H: ICC-MY (green; arrows) and PDGFRα⁺ cells (red; arrowheads) in the same tissue. Scale bars are shown for each panel.

Role of Interstitial Cells in Pacemaker Activity

The idea that ICC might provide pacemaker activity originated from the morphological evidence showing that ICC were electrically coupled to SMCs, and this idea was later supported by dissection experiments in the 1980s that showed that pacemaker activity was focused in regions of tissue that contained networks of ICC (6, 11, 36, 100). Removal of ICC networks rendered muscles void of pacemaker activity and inhibited active slow-wave propagation (96). Subsequent studies using mutant animals in which ICC failed to develop (44, 113) or in mice treated with neutralizing c-Kit antibodies (107) confirmed the obligatory role of ICC in pacemaking. Patch-clamp studies of GI SMCs failed to display spontaneous pacemaker activity or conductances compatible with the generation or regeneration of slow waves. Freshly isolated ICC display autorhythmicity, with ion conductances and pharmacologies consistent with the slow waves recorded in intact muscles (39, 45, 73, 124).

Selective impalement of ICC in intact muscles showed (FIGURE 2, A AND B) 1) omnipresent slow-wave activity, as also observed in smooth muscle cells (19, 59, 60); 2) faster rates of depolarization (upstroke phase) compared with slow waves recorded from SMCs in the same muscles (59, 61); 3) upstrokes of slow waves in ICC were phase advanced vs. the onset of depolarization in nearby SMCs (40). Elegant experiments in which dual impalements of ICC and SMC were accomplished clearly demonstrated electrical coupling between these cells (15, 40). Taken together, these findings supported the concept that ICC generate electrical slow waves and actively propagate these events through networks of ICC that are electrically coupled to the circular and longitudinal muscle layers. Slow waves conduct to SMCs passively, and no mechanism that is capable of regenerating slow waves has been identified in SMCs. Propagation of slow waves in ICC networks is critical for organ-level motility patterns such as gastric peristalsis and intestinal segmentation.

FIGURE 2.

Recordings of electrical activity from the different cell types of the SIP syncytium

A and B: slow waves recorded by impalements of an ICC-MY and a circular SMC in the murine small intestine (as in Refs. 60, 61), respectively. The resting potentials in these cells were −70 mV in the ICC-MY and −68 mV in the SMC. The slow-wave events in ICC are much larger in amplitude, and the upstroke depolarization is ∼1 V/s. Slow waves conduct to SMCs, and the smaller amplitude is a result of the passive nature of this communication (i.e., decremental conduction). C–E: recordings from an ICC-MY, a SMC, and a fibroblast-like cell of the rabbit small intestine (as in Refs. 58, 59), respectively. In these recordings, the resting potentials were −68 mV for the ICC-MY, −64 mV for the SMC, and −43 mV for the fibroblast-like cell. Note again that the amplitude and rise time of the slow waves recorded from ICC-MY are much greater than the events recorded in SMCs. Small oscillations at the frequency of slow waves were also observed in fibroblast-like cells, but a prominent activity in these cells was spontaneous transient hyperpolarizations (STHs; asterisks denote examples). These events appeared to be due to SK channels because they were blocked by apamin. STHs occurred during and between slow waves. Cells were identified by filling them with fluorescent dyes during impalements.

ICC are a minor component of the total cells in smooth muscle tissues, representing <10% of cells (81). Thus it was hard to recognize these cells in enzymatic dispersions of the muscle layers. This problem was solved by generation of a reporter strain of mice in which a bright green fluorescent protein was expressed in ICC (88). Isolation of ICC from these mice allowed more detailed experiments into the mechanism of slow-wave activity. Studies on intact muscles by the groups of David Hirst, Hikaru Suzuki, and Dirk van Helden had suggested that slow waves were due to a Ca²⁺-activated Cl⁻ conductance (CaCC) (39, 57, 60, 108), but other studies, performed on cultured ICC, suggested a nonselective cation conductance was the inward current likely to be responsible for slow waves (65). When it became possible to study freshly isolated ICC, long-duration “autonomous” currents were observed upon step depolarizations of these cells (31). These large-amplitude, unique currents displayed by ICC are due to a CaCC with the properties of Ano1 (aka Tmem16a) (124). Studies on intact muscles showed that genetic deactivation of Ano1 completely abolished slow-wave activity (FIGURE 3) (45). Ano1 expression was rapidly lost in cell culture, and the autorhytmicity retained by ICC was altered vs. the pacemaker activity of cells in situ.

FIGURE 3.

Loss of pacemaker activity in Ano1^−/− mice

A: genotyping of Ano1^{tm1Bdh/tm1bdh} mice. The wild-type allele was absent in animals 1–3 in this representative litter, proving that these mice were Ano1^−/−. Animals 4 and 5 were a heterozygote and a wild-type homozygote, respectively. B and C: intracellular electrical recordings from jejunal and antral circular muscles of each of the five animals. Slow waves were absent in the muscles of Ano1^−/− mice. Slow waves with a normal amplitude and frequency were recorded from animals that were heterozygous or homozygous for Ano1 (animals 4 and 5, respectively). In contrast to what has been claimed by others (102), we observed no significant differences in the development or density of ICC-MY (pacemaker class of ICC in these organs; see Table 1). D and E: ICC [ICC-MY (arrows) and ICC-DMP (arrowheads)] have a normal distribution and density in the small intestines of Ano1^−/− mice, suggesting that Ano1 is not critical for the development of ICC. Figure is redrawn from Ref. 45 with permission. Scale bar for D and E is shown in E.

Multiple electrode and imaging studies have shown that slow waves propagate actively in GI muscles and organs at rates exceeding 1 mm/s (6, 85). In the axis of the circular muscle layer, the propagation velocity can exceed 10 mm/s. This rate is too fast to be due to simple intracellular diffusion of Ca²⁺ or a second messenger causing sequential Ca²⁺ release from stores. A voltage-dependent mechanism for propagation is required. Some investigators suggested that changes in membrane potential are linked to enhanced production of IP₃ (39, 108) or perhaps voltage-dependent sensitization of IP₃ receptors that are also required for slow-wave generation and regeneration (105). Our own experiments suggested that the voltage-dependent mechanism is due to voltage-dependent Ca²⁺ entry, and the likely mechanism is activation of T-type Ca²⁺ channels. The controversy regarding the voltage-dependent mechanism is not fully resolved, but voltage-dependent IP₃ production has not yet been demonstrated in ICC, whereas these cells express Cacna1H and display T-like Ca²⁺ currents (7, 85, 123). In this concept, Ca²⁺ entry is thought to produce Ca²⁺-induced Ca²⁺ release (CICR) from ryanodine and/or IP₃ receptors. CICR appears to be the signal coupled to activation of CaCC (125). It is possible that Ca²⁺ entry could be the signal that “sensitizes” IP₃ receptors, since these channels are sensitive to cytoplasmic Ca²⁺ levels (48).

Still unresolved is the question of how the plateau phase of slow waves is maintained for long durations (one to many seconds). Openings of CaCC depend on maintenance of Ca²⁺ at the cytoplasmic surface of the channels, and Ca²⁺ release events are typically transient in nature (86, 126). The mechanism for maintenance of CaCC activation to create the plateau phase of slow waves therefore requires a sustained means of Ca²⁺ delivery. More experiments will be required to understand the full mechanism of slow-wave generation and propagation, but this is an important question. The depolarization of SMCs caused by conduction of slow waves yields activation of voltage-dependent (L-type) Ca²⁺ channels in SMCs, and this is the primary signal that initiates contraction in GI muscles (84). Ca²⁺ entry is regulated in part by the amplitude and duration of slow waves, so the mechanism regulating slow-wave duration (i.e., plateau phase) is of critical importance to the regulation of contractile force in GI muscles.

Role of Interstitial Cells in Neurotransmission

ICC and Neurotransmission

The close proximity of interstitial cells (ICC-IM and ICC-DMP; see Table 1) and apparent specialized junctions with enteric motoneurons suggested that ICC are innervated (17, 21, 51). Neurotransmitter signals transduced by these cells can be conveyed to SMCs via gap junctions. Expression of receptors and effectors by interstitial cells and SMCs suggests that postjunctional responses to motor neurotransmission in GI muscles are likely to represent integrated responses of the SIP syncytium (FIGURE 4). More detailed reviews of this topic have been provided previously (92, 97).

Table 1.

Localization and proposed functional contributions of cellular components of the SIP syncytium in GI muscles

Type of Cell and Anatomical Localization	Common Name(s)	Proposed Function(s)	Representative References
Myenteric ICC (in region between circular and longitudinal muscle layers)	ICC-MY; ICC-MP	1) Pacemaker cells that generate electrical slow waves; 2) form networks that allow active (regenerative) propagation of slow waves; 3) voltage-dependent excitability due to expression of voltage-dependent Ca2+ channels; 4) role in setting resting membrane potential and basal excitability of SIP syncytium	7, 19, 44, 83, 99, 107, 113, 123
Intramuscular ICC (within bundles of smooth muscle cells and lying in close proximity to nerve varicosities)	ICC-IM; ICC-DMP in region of the deep muscular plexus of the small intestine	1) Transduction of neurotransmitter signals from enteric motor neurons; 2) mechano-sensitivity; 3) lack of voltage-dependent excitability due to generally low expression of voltage-dependent Ca2+ currents; 4) contribution to active propagation in gastric muscles; 5) mediation of inflammatory input via protease-activated receptors	8, 10, 14, 18, 33, 76, 104, 106, 112, 114
ICC within septa between bundles of muscle	ICC-SEP	1) Active propagation of slow waves into the depth of thicker muscles (as in human GI tract); 2) interactions with enteric neurons	42, 74, 115
ICC along the submucosal surface of the circular muscle layer in colon and to small extent in stomach	ICC-SM	1) Pacemaker activity in the colon	11, 100
ICC at the serosal surface of the longitudinal muscle layer	ICC-SS	Unknown	2, 13
Fibroblast-like cells found in all regions of the GI tract; generally common localization with ICC; also in close apposition with varicosities of enteric neurons	*PDGFRα+ cells	1) Mediation of purinergic inhibition (purinergic inhibitory junction potential); 2) mediation of inflammatory input via protease-activated receptors	5, 47, 69–71, 104
Smooth muscle cells	Circular SMCs, longitudinal SMCs, sphincteric SMCs, sling SMCs in stomach	1) Generation of forces in motility; 2) maintenance of sphincter tone; 3) contributions to setting resting excitability of SIP syncytium; 4) transduction of neural and hormonal responses; 5) mediation of Ca2+ sensitization mechanisms	32, 54, 75, 84, 91, 101, 103, 117

^*PDGFRα⁺ cells also have distinct localization and cellular morphologies when found within muscle bundles or in the myenteric region. They are often intertwined with ICC in these localizations. At present, no differences in cells within muscle bundles and cells in the myenteric region have been distinguished; thus, for the purposes of this overview description, these cells are not broken down into

^*PDGFRα⁺-MY and

^*PDGFRα⁺-IM classes.

FIGURE 4.

Cartoon showing SIP syncytium (94)

c-Kit⁺ ICC (green) and PDGFRα⁺ (fibroblast-like cells; red) are intermixed with smooth muscle fibers in bundles of cells within the circular muscle layer (and in some regions in the longitudinal muscle layer as well). Intramuscular ICC and PDGFRα⁺ cells are closely associated with the processes of enteric motoneurons and express receptors, second-messenger pathways, and ion channels facilitating responses to enteric motor neurotransmitters. ICC and PDGFRα⁺ cells are electrically coupled to SMCs, so responses elicited in the interstitial cells can conduct to SMCs and regulate the excitability of the musculature. ICC-MY and PDGFRα⁺ cells are also found in the region of the myenteric plexus. ICC-MY are pacemaker cells and generate electrical slow waves. PDGFRα⁺ cells are responsive to purines. Here again, the interstitial cells are electrically coupled to circular and longitudinal SMCs, and slow waves conduct from ICC-MY to SMCs to generate a phasic pattern of contractions in the musculature.

Early studies on mutants (W/W^V and Sl/Sl^d mice) lacking most ICC-IM in the gastric fundus showed that postjunctional electrophysiological responses to nitrergic and cholinergic neurotransmission were decreased (9, 14, 112). Nitric oxide, which normally causes significant hyperpolarization in mouse fundus, had little effect on membrane potential in W/W^V mutants, suggesting that a major transduction pathway for NO occurs in ICC-IM. The fast transient excitatory junction potentials (EJPs) elicited by cholinergic neurotransmission were also depressed in W/W^V mutants.

More recent studies of rat and mouse mutants with reduced populations of ICC-IM have claimed that normal motor responses are observed in these mice and that ICC do not contribute to the postjunctional responses (1, 20, 43, 122). However, some of these studies have utilized regions of the GI tract in which there is an incomplete loss of ICC, making the results difficult to interpret (1, 122). Recent studies have also shown that the developmental defect in ICC-IM of W/W^V is not consistent among the mutant animals, but, when the lesion was severe, nitrergic responses were depressed (95).

An interesting observation was made regarding cholinergic neurotransmission in that contractile responses to cholinergic neurotransmission can actually be elevated in W/W^V muscles (12). In this study, Ca²⁺ sensitization responses to carbachol added to the solutions bathing muscles were compared with responses to cholinergic nerve stimulation. Carbachol caused phosphorylation of CPI-17 and MYPT, two components of Ca²⁺ sensitization pathways that are enhanced to inhibit myosin phosphatase (101), but cholinergic neurotransmission resulted in phosphorylation of only CPI-17. When ICC-IM were reduced in numbers in W/W^V muscles or when metabolism of ACh was reduced by cholinesterase inhibitors, then MYPT phosphorylation was observed. These data suggest that cholinergic responsiveness was altered in W/W^V muscles by recruitment of additional Ca²⁺ sensitization mechanisms, and, in these conditions, ACh released from nerve terminals reaches receptors not normally bound when ICC are present and when cholinesterases are active.

Experiments have also been performed recently utilizing the Cre/loxP technology to produce cell-specific lesions in nitric oxide receptors [soluble guanylate cyclase (sGC)] (32, 33, 75, 76). These studies have suggested that nitrergic regulation of contractile activity in GI organs results from integrated responsiveness from specific components of the SIP syncytium. For example, in the colon, arrhythmic contractions occurred in wild-type mice and in mice with specific knockdown of sGC in SMCs. Mice with global knockdown of sGC and mice with knockdown directed at ICC displayed rhythmic contractile activity. Blocking neural inputs switched the contractile pattern in wild-type and SMC knockdown mice to rhythmic contractions. These findings suggest that ICC are involved in tonic neural inhibition of the colon, a behavior observed in GI muscles of several species (116, 119). Responses to electrical field stimulation of intrinsic inhibitory neurons were, in contrast, mainly due to inhibitory responses generated by SMCs because knockdown of sGC in SMCs reduced these responses. In the esophagus, a different pattern of cell-specific regulation was noted (33). Basal tone of the lower esophageal sphincter was regulated equally by SMCs and ICC, but swallowing-induced relaxation responses appeared to be mainly mediated by ICC. These studies rely heavily on the ability of Cre recombinase to knock out expression of sGC in specific types of cells. As discussed recently, the degree of knockdown may be incomplete in studies in which inducible Cre is activated in adult animals (93).

PDGFRα⁺ Cells in Neurotransmission

There is another type of interstitial cell that was originally identified by electron microscopy that is also closely aligned with motor nerve terminals in GI muscles. These cells were called fibroblast-like cells for many years because of their morphological features. Fibroblast-like cells were shown to express small conductance Ca²⁺-activated K⁺ (SK3) channels (27, 62, 110), raising the possibility that they might have a role in responses to purine neurotransmitters in GI muscles. Recently, antibodies against PDGFRα were found to provide selective labeling of fibroblast-like cells throughout the GI tract (49, 50). A reporter strain of mice expressing a histone 2B-eGFP fusion protein driven by the endogenous Pdgfra promoter allowed identification of these cells for physiological and expression studies after enzymatic digestion of muscles.

Studies of intact murine and human GI muscles have clearly demonstrated that the purine neurotransmitter(s) binds to P2Y1 receptors to elicit the postjunctional effects of purinergic neurotransmission (i.e., the fast inhibitory junction potential that is linked to inhibition of contractions) (29, 30, 46). Purinergic effects are mediated, in part, by activation of apamin-sensitive, small-conductance Ca²⁺-activated K⁺ channels. PDGFRα⁺ cells display high expression of P2ry1 (P2Y1 receptors) and Kcnn3 (SK3 channels), and expression of P2ry1 and Kcnn1-3 genes is far lower in SMCs than in PDGFRα⁺ cells. Immunohistochemical analysis showed PDGFRα⁺ cells to be intertwined with ICC and the processes of excitatory (labeled with antibodies to vesicular acetylcholine transporter) and inhibitory (labeled with antibodies to nNOS) motoneurons. These data suggest that, when purines are released from inhibitory neurons, there is likelihood that receptors and effectors expressed by PDGFRα⁺ cells and appropriate for purinergic neurotransmission might be activated. Immunohistochemical and expression studies also demonstrated that PDGFRα⁺ cells are a population of cells distinct from ICC (49, 50, 70, 71). Similar interstitial cells with the same distributions and overlapping immunohistochemical properties were found in human colon (70) and stomach (34).

Voltage-clamp studies showed that PDGFRα⁺ cells express a Ca²⁺-dependent K⁺ conductance that was activated by purines and blocked by apamin (71). Some PDGFRα⁺ cells displayed spontaneous, periodic activation of K⁺ currents (spontaneous transient outward currents or STOCs), and these events were also blocked by apamin. Single channels with a conductance (∼10 pS) and sensitivity to intracellular Ca²⁺ (EC₅₀ = 364 nM), consistent with expression of SK3 channels, were observed in excised membrane patches. Application of P2Y1 agonists caused activation of a conductance with the same properties, and apamin and a selective P2Y1 antagonist, MRS2500, inhibited these responses.

Studies to compare the relative responsiveness of SMC and PDGFRα⁺ cells to purines were performed on cells from the murine colon (58). Previous studies have shown that SMCs express P2Y receptor-driven, apamin-sensitive responses, but the current density resolved by activation of these channels was low, and depolarized potentials were needed to resolve the outward currents initiated by purines (64). Application of ATP caused significant hyperpolarization responses in PDGFRα⁺ cells, and these responses were partially blocked by MRS2500, a selective P2Y1 antagonist. Held at potentials equivalent to the resting potentials of cells in intact muscles, ATP caused depolarization of SMCs. Selective P2Y1 agonists also caused significant hyperpolarization of PDGFRα⁺ cells and little or no response in SMCs. Taken together, these data suggest that SMCs are unlikely to mediate the signature purinergic response in GI muscles: fast transient hyperpolarization. This response more likely results from purinergic activation of P2Y1 receptors and SK3 channels in PDGFRα⁺ cells.

The question of how P2Y1 receptors link to activation of SK3 channels in PDGFRα⁺ cells was addressed in imaging studies in which these cells were loaded with Fluo-4 and identified by a reporter molecule (eGFP) expressed in cell nuclei (4, 5). Purines were found to activate Ca²⁺ transients in PDGFRα⁺ cells from gastric fundus and colon. Responses were blocked in P2ry1^−/− mice. The Ca²⁺ transients to P2Y1 selective agonists were blocked by MRS2500, but responses to ATP were not completely blocked by P2Y1 antagonists, nor were responses in muscles of P2ry1^−/− mice. These data tend to confirm other studies suggesting that another more P2Y1-selective purine, and not ATP, is the purinergic neurotransmitter in GI muscles (47, 80).

Ca²⁺ imaging studies were also used to study the sequence of events in response to purinergic neurotransmission (5). With cholinergic and nitrergic components of neural inputs blocked, stimulation of intrinsic motoneurons evoked Ca²⁺ transients in PDGFRα⁺ and SMCs. The postjunctional responses were abolished by MRS2500 and in muscles of P2ry1^−/− mice. The latency in the Ca²⁺ transients evoked in PDGFRα⁺ cells were equivalent to electrophysiological responses. The response evoked in SMCs during the stimulus was a slowly developing reduction in Ca²⁺. Gap-junction uncouplers inhibited responses of SMCs to nerve stimulation but did not affect the timing of Ca²⁺ transients evoked in PDGFRα⁺ cells. These data showed that PDGFRα⁺ cells are directly innervated by enteric inhibitory motoneurons and that these cells mediate responses to purinergic neurotransmitters by releasing Ca²⁺ from intracellular stores. Postjunctional hyperpolarization responses are conducted to SMCs via gap junctions.

Recently, cells identified morphologically as fibroblast-like cells were impaled and filled with dye to verify sites of recording in rabbit small intestine (58). In contrast to recordings from ICC and SMCs, fibroblast-like cells generated spontaneous transient hyperpolarizations (STHs). These events were often superimposed on low-amplitude oscillations at the frequency of slow waves (FIGURE 2, C-E). The STHs were reduced by MRS2500 and blocked by apamin, suggesting they were due, in part, to binding of P2Y1 receptors, possibly from tonic release of purine neurotransmitters, but some fraction of this activity may have been due to spontaneous activation of SK channels in these cells. These results are consistent with the integrative nature of the SIP syncytium and the cell-specific behavior of fibroblast-like cells. Slow waves appear to conduct to these cells with decrement, and the cells are capable of purinergic activation of SK-like currents. Spontaneous activation of SK channels also occurs via spontaneous Ca²⁺ release events that have been observed in this class of interstitial cells in the stomach and colon (4, 5). Ongoing STHs in electrically coupled cells within the SIP syncytium would tend to produce a hyperpolarizing influence (tonic inhibition) in intact GI muscles and decrease the input resistance of the SIP syncytium, reducing the amplitude of conducting slow waves and reducing the probability of reaching the threshold for generation of action potentials in SMCs.

Role of Interstitial Cells in Pathophysiology

Research into the causes of GI motility disorders has largely focused on problems that might arise from lesions in enteric neurons or SMCs. Certainly, there are motility disorders that are due to developmental defects in enteric neurons, as in Hirschsprung's disease; neuropathies that can develop with age, as in diabetes; and remodeling of sensory pathways, as may be important in several functional bowel diseases. However, there appears to be a wide range of motility disorders that may be due to causes other than those directly related to cellular pathologies of enteric neurons and SMCs. The important regulatory functions provided by interstitial cells suggest that lesions in these cells might lead to defective electrical pacing, inappropriate coordination between regions of the SIP syncytium due to lack of connectivity between ICC and loss of active propagation of slow waves in the ICC networks, reduced or unbalanced neural regulation due to reduced or altered excitatory and/or inhibitory neural inputs, altered baseline excitability of the SIP syncytium, and altered responses to stretch (Table 1). Further suggestion that interstitial cells may be important factors in GI motility disorders comes from pathological reports showing reduced numbers of ICC or abnormal distributions of ICC networks that might lead to problems of connectivity in GI muscles of human patients with motor defects. It should be stressed that this type of information does not prove cause-and-effect, but studies of animal models with reduced populations of ICC have generally supported the idea that reductions in this cell population might be causative in certain types of motility disorders. One issue of importance is the quality of morphological evaluation that has been performed to compare populations of ICC in healthy and diseased muscles. A recent paper speaks to the procedures that should be followed in this type of clinical evaluation (63). Table 2 summarizes some of the motility disorders in which pathologists have reported reduced or defective ICC populations that might be linked to symptoms.

Table 2.

Motility disorders in which ICC loss or dysfunction has been reported

Motility Disorder	Major Symptoms That Might be Related to Loss of Interstitial Cells	References
Achalasia	Reduction in swallowing-induced relaxation of lower esophageal sphincter	22, 56, 68
Diabetic gastropathy	Delayed gastric emptying; reduction in gastric accommodation and/or response to inhibitory neurotransmission	26, 38, 53, 78, 82, 111
Idiopathic gastroparesis	Weakened or dyscoordinated antral contractions leading to prolonged period of gastric emptying; loss of responsiveness to digestive neural control	121
Pyloric stenosis	Inability of the pyloric sphincter to relax to aid gastric emptying	72, 109
Intestinal pseudo-obstruction	Delayed or blocked intestinal transit	25, 52, 55, 120
Ulcerative colitis	Fecal stasis due to decreased motility in proximal colon; increased transit times in proximal colon; enhanced transit in rectosignmoid areas; reduced colonic motility following a meal	89
Slow transit constipation	Reduced force of colonic contractions; weakened propulsive contractions (mass movements); loss of coordination produced during peristaltic reflex; constipation	37, 77, 118
Internal anal sphincter achalasia	Reduction in internal anal relaxation during defecation reflex	87

Little is known about how loss or defects in PDGFRα⁺ cells might impact motility, and, as yet, this group of interstitial cells appears to have been evaluated in a very limited number of human tissue samples (34). Here again, the importance of these cells in purinergic neurotransmission and some inflammatory responses might make them candidates for pathologists to examine once accurate immunohistochemical criteria, for their evaluation has been developed for the clinical lab.