Our understanding of the physiological roles played by interstitial cells of Cajal (ICC) in relation to gastrointestinal (GI) motility is still rudimentary. Nevertheless, studies into the pathophysiology of ICC are emerging at a rapid pace. Caution should be exercised, however, in assuming correlations between changes in Kit immunoreactivity, findings of ultrastructural abnormalities in ICC, and the pathophysiology and symptoms of the patients. Recent studies have revealed reduced numbers or the absence of ICC in small intestine and colon that do not exhibit normal peristaltic activity. Furthermore, important evidence is emerging that motor abnormalities in newborns may be associated with delayed maturation of the ICC network. These preliminary clinical studies provide plausible hypotheses toward the pathophysiology of certain motor disorders and strongly encourage basic scientific studies directed toward discovering the intrinsic properties of ICC as well as obtaining a deeper understanding of the physiological roles played by these cells.
- intestinal motility
- motility disorders
our understanding of the physiological role of interstitial cells of Cajal (ICC) is still rudimentary. ICC and smooth muscle cells are derived from common mesenchymal precursor cells (13,17, 35). ICC are present as networks of cells associated with the neural plexuses within the gut musculature, where they are connected to each other through long processes via gap junctions. In addition, they are found dispersed inside most circular muscle layers, including those of the sphincters. Some of the networks of ICC play a role in the generation of electrical pacemaker activity and thereby control peristalsis that is coordinated by slow waves. This occurs predominantly in the distal stomach and proximal small intestine. Evidence for such a role of ICC has accumulated in recent years and has been the subject of recent reviews (9, 26). Although the evidence has been largely indirect, single isolated ICC have recently been shown to generate pacemaker currents, proving their identity as pacemaker cells (27).
ICC communicate with smooth muscle cells through long processes that terminate on the cell body, and ICC are themselves densely innervated. From this structural arrangement, the hypothesis has long been held that ICC play a role in the innervation of smooth muscle (23). Recent data from mutant mice lacking ICC within the circular muscle of the stomach support the notion that ICC can play a role in inhibitory innervation (1). Structural evidence for inhibitory innervation has been found, but no inhibition of muscle activity occurs when these nerves are stimulated in vitro.
Although the physiological roles of ICC are just beginning to be identified, studies into roles for ICC in the pathophysiology of gut motor disorders have already begun. Several studies have found altered structural arrangements of ICC associated with a site of dysfunction. However, caution should be taken not to overinterpret the role of ICC in the absence of studies that show a causal relationship between abnormal ICC and abnormal function. Nevertheless, these preliminary studies, discussed below, are exciting and set the stage for our eventual understanding of the role of ICC in motor dysfunction and the potential of ICC as target cells for pharmacological restoration of normal motor function (9).
PSEUDO-OBSTRUCTION IN THE SMALL INTESTINE
In the mouse small intestine a direct relationship exists between propagating electrical slow waves, propagating waves of intraluminal pressure, and pulsatile outflow of intestinal contents, illustrating how slow waves can control peristaltic activity (6). Because slow wave generation has been convincingly linked to the presence of ICC, the role of ICC in certain patterns of peristaltic motor activity seems proven (9, 26, 27). Malfunctioning ICC could therefore be associated with chronic idiopathic intestinal pseudo-obstruction. In support of this possibility, the number of Kit-positive cells in the external smooth muscle layers of the intestine was shown to be 3% of normal in two patients with pseudo-obstruction, suggesting that absence of ICC may lead to pseudo-obstruction (10).
One has to be cautious in assuming simple links between loss of Kit immunoreactivity and loss of normal motor function. The experience of my own laboratory and that of others is that the degree of immunoreactivity as well as the number of ICC in patients with pseudo-obstruction are widely variable, indicating that the role of ICC in the pathophysiology of pseudo-obstruction is likely to vary. Absence of ICC or structural abnormalities are rarely seen in isolation. Often, neural abnormalities are also observed, so a role for ICC per se is difficult to establish without further studies. Furthermore, global assessment of the degree of Kit immunoreactivity may not reveal an accurate estimate of the number of ICC or the health of ICC. Despite apparently normal distributions of ICC, ICC can be of abnormal ultrastructure, as determined by electron microscopy (24). The opposite is also important; when no staining with Kit antibodies can be found, there is no certainty that ICC are absent. Absence of Kit immunoreactivity may be due to a specific change in ICC resulting in loss of the Kit protein from the membrane. In certain animal species it may be due to the inability of available antibodies to stain the ICC. It is tempting to assume that all Kit-positive cells in the GI musculature, except mast cells, are ICC. However, some of these cells have never been examined ultrastructurally, and this is urgently needed. Not all Kit-positive cells may turn out to fit the electron microscopic criteria of ICC. This may result in changing the scope of ultrastructural criteria for ICC, a new definition of identifying features, subclassification of ICC, or the recognition that certain Kit-positive cells should not be classified as ICC. Finally, the detection of a structural change in ICC may not directly relate to a functional abnormality. ICC associated with neural plexuses are organized in a network; changes in some ICC might not affect the functioning of the network as a whole.
Transient neonatal pseudo-obstruction has been associated with a delayed maturation of ICC (12). In mice at birth, ICC are not fully matured with respect to ultrastructural features, and a complete network of cells has not yet been formed. In the 2 days after birth, cells mature fully and increase in number to form a complete network. In fact, mitoses in ICC were observed in the first 2 days (L. Thuneberg, unpublished data). When an antibody to the Kit receptor is given to mice at day 4, the ICC network does not develop normally within the expanding musculature and intestinal motor abnormalities develop (20). The precise time of maturation of the ICC networks in humans is not known. Although in humans Kit immunoreactivity can normally already be found after 9 wk of gestation (12), in one child born at 27 wk of gestation no ICC were observed in the musculature 3 wk after birth, coincident with features of small bowel obstruction (12). After a temporary ileostomy, at 42 wk, the motor activity became normal and ICC had developed in the musculature. These data indicate that delayed maturation of ICC, through premature birth or otherwise, may be the cause of motility disorders in infants.
MOTOR ABNORMALITIES IN THE COLON
Although generation of slow waves in the dog colon is clearly related to ICC associated with the submuscular plexus, the role of ICC as well as the role of slow waves in human colon motor control is not clear. Colonic motor patterns are not dominated by the fast rhythmic slow wave-driven peristalsis observed during stomach emptying or transit through the proximal small intestine. Furthermore, there is no continuously present electrical activity in the human colon in vivo. Distension and/or neural activity induce electrical activity. The role of ICC in the human colon is likely to be extensively integrated with neural activity and may therefore be more difficult to unravel. It is important to note that in the human colon a band of special smooth muscle cells exists at the inner surface of the circular muscle layer traversing septa and connecting the circular muscle lamellae (7, 25) (Fig. 1). This structure encompasses the submuscular plexus nerves and also winds into the submucosa, making contact with submucosal neural elements. ICC are structurally an integral part of this muscle layer, although their precise physiological role has not yet been defined (28). The ICC network likely plays an important role in the generation of electrical activity (22) as well as in the coordination of electrical and motor activity along the long axis of the colon (2, 18). Consistent with this view were observations in a 1-mo-old infant that presented with colonic obstruction (Fig. 2) (12). In the colon, no peristaltic activity was observed and no Kit immunoreactivity was present within the circular muscle and in the submuscular plexus area. Immunohistochemistry revealed a normal nerve distribution and normal ICC (i.e., presence of Kit immunoreactivity) associated with the myenteric plexus area. In the ileum, normal peristaltic activity was seen and the ICC distribution was normal. By 6 mo, normal peristaltic activity had developed in the colon and the ICC were fully developed.
Constipation due to colonic hypomotility is often associated with anorectal malformations in children (11). In 7 of 12 patients, no ICC were observed associated with the submuscular plexus, whereas the presence of ICC in other locations was affected to varying degrees. Innervation as judged by neuronal immunohistochemical markers was normal. Kenny et al. (11) proposed that the colonic hypomotility was due to the absence of ICC. However, hypomotility would likely also involve a muscle or nerve lesion, and no difference in symptoms was indicated between patients with or without normal ICC.
Hirschsprung’s disease is associated with obstruction of colonic transit, and the affected segment is characterized by the absence of intramural ganglia in the enteric neural plexuses. However, the direct cause of the abnormal motility is not known. It has been recognized that aganglionic segments have a lack of inhibitory junction potentials (15), no functional inhibitory innervation (21), and absence of nitric oxide (NO)-containing nerves (29), but cholinergic input is present (8,15). The electrical activity of the aganglionic segments was also abnormal in that slow wave-like activity was absent (15). This is consistent with the notion that a very low density of ICC was observed in the submuscular plexus area and within the circular muscle layer (33, 34). A combination of a lack of normal myogenic control and a lack of inhibitory innervation, which makes normal peristalsis impossible, may therefore contribute to motor abnormalities in Hirschsprung’s disease.
ICC AND SPHINCTER FUNCTION
The role of ICC in the control of motor activity of the sphincters has not been resolved. ICC occur prominently within the circular muscle layer, where they make gap junction contact with each other and with smooth muscle cells without forming an obvious network (3). The pylorus can generate its own rhythmic activity, independent of the small intestine and stomach, and ICC may play a role in its generation. Slow wave activity from both the stomach and the intestine can propagate into the pyloric smooth muscle; hence the interaction between the ICC networks of these organs may be crucial for normal coordinated motor activity. The absence of ICC in the pylorus may not allow for normal propagation of electrical events and could thus result in abnormal peristaltic activity. In addition, ICC and inhibitory innervation may be linked in the circular muscle of the pylorus, similar to the stomach (1).
In children with infantile hypertrophic pyloric stenosis (IHPS), ICC were not present in sections of the circular muscle (30), although there were cells with some ultrastructural features of ICC, suggesting underdeveloped ICC (16). IHPS is also associated with a loss of both peptidergic nerves (3) and NO-containing nerves (32). Motor abnormalities in IHPS are likely due to a combination of a lack of inhibitory innervation and abnormal myogenic control caused by the absence of ICC. Normally a peristaltic contraction associated with gastric emptying continues through the pylorus, resulting in pulsatile emptying of the stomach. The typical postprandial projectile vomiting associated with IHPS might be caused by a defect in the coordination between the movement of the pyloric sphincter and the stomach (30). Interestingly, when two cases of resolved IHPS were examined for Kit immunohistochemistry, the distribution of ICC had become normal (31).
ICC AND INFLAMMATION
Rumessen (24) examined ICC and surrounding structures in four patients with advanced ulcerative colitis. He focused on the special inner layer of circular muscle in which the submuscular plexus and a network of ICC are embedded. The smooth muscle cells appeared normal, whereas some ICC in all of the patients showed specific abnormalities. Typical changes found in ICC were an increase in the occurrence of secondary lysosomes in close association with multiple large confluent lipid droplets and large clusters of glycogen granules. In addition to changes in ICC, nerve terminals of the submuscular plexus were swollen, with an absence or paucity of transmitter vesicles. The changes in ICC were often located in specific areas of the cell and hence were not a sign of impending cell death. They may reflect abnormal metabolic activity. The ultrastructural change was not seen in all ICC and may not have affected the electrical activity of the musculature (14). It may be an advantage that ICC are organized in a network; damage to some cells may be corrected by the rest of the network. Since nerve terminals associated with ICC were affected in ulcerative colitis, motor abnormalities could be due to alterations in ICC-nerve interactions.
Animal models of inflammation support the notion that damage to ICC may underlie motor abnormalities. During a Trichinella spiralis infection, ICC were the first cells to undergo structural changes; these were accompanied by changes in electrical activity and abnormalities in intestinal transit (4, 5). ATrichinella spiralis infection represents inflammation with extensive infiltration of immune cells, including lymphocytes, and the interaction between immune cells, muscle, and ICC is under intense investigation. Acute inflammation due to acetic acid exposure in the dog colon was associated with infiltration of neutrophils and macrophages and was associated with damage to smooth muscle cells (19) (Fig. 3). Throughout the inflamed colon sections occurred where the network of ICC and associated nerve structures were disrupted, as observed by electron microscopy (Fig.3). However, sections of normal or damaged ICC were also present; the cellular processes of ICC were affected the most, which suggests loss of communication with smooth muscle cells. The electrical activity of muscle strips was also variable; in some strips no slow wave activity was recorded, whereas in other strips slow wave activity was abnormal. These changes in the neuromuscular apparatus affect in vivo motility, although the contribution of each injured cell type to the abnormal motor activity is difficult to decipher. In these studies, structural changes are not uniformly distributed over an affected segment. Hence structural and functional studies should ideally be done in all parts of the region investigated. Injury to ICC due to inflammation is repaired over time; our preliminary data suggest that both cell division and the generation of new processes are important in this process.
Studies into the pathophysiology of ICC have begun and will likely increase markedly in the coming years. Interesting hypotheses have been put forward as to the functional consequences of the absence or delayed maturation of ICC. In many cases, however, correct interpretation of the findings as they relate to motor dysfunction may have to wait for more information from basic scientific studies directed toward discovering intrinsic properties of ICC and obtaining a deeper understanding of the physiological roles played by these cells under normal conditions.
We gratefully acknowledge support from the Medical Research Council of Canada in the form of an operating grant and a Medical Research Council Scientist Award.
Address for reprint requests: J. D. Huizinga, McMaster Univ., HSC-3N5C, 1200 Main St. West, Hamilton, ON L8N 3Z5, Canada.
↵* Fourth in a series of invited articles on Neural Injury, Repair, and Adaptation in the GI Tract.
- Copyright © 1998 the American Physiological Society