Diabetes affects many aspects of gastrointestinal motility, in part due to changes in interstitial cells of Cajal (ICC). The effect of diabetes on the colon, however, is not well characterized, and the aim of the present study was to investigate possible relationships between altered colonic motility as a consequence of streptozotocin-induced diabetes and injury to ICC. Physiological, immunohistochemical, and ultrastructural techniques were employed. The motor pattern of the rat colon was dominated by rhythmic high-amplitude, low-frequency contractions that were primarily myogenic in origin. These rhythmic contractions were induced by stretch associated with increased tension; the amplitude of the superimposed rhythmic contractions increased with increasing applied tension. In diabetic rats, the stretch-induced rhythmic contractile activity remained robust and of similar frequency but was significantly higher in amplitude compared with that in control rats. At 700 mg of applied tension, the force of contraction in circular colonic muscle strips of the diabetic rats was 370% of control values. This robust presence of low-frequency contractions is consistent with the unaffected pacemaker, the ICC associated with Auerbach's plexus, and the increased amplitude correlates with loss of and injury to ICC of the submuscular plexus and intramuscular ICC. Loss of inhibitory nitrergic nerves does not appear to be a factor based on unaltered nNOS immunoreactivity.
- interstitial cells of Cajal
- immunohistochemical staining
- rhythmic contractile activity
a wide range of chronic gastrointestinal symptoms are known to frequently occur in diabetic patients, including gastroparesis, constipation, and diarrhea (33, 48). Although animal models have been used extensively to investigate the pathophysiology of diabetic motor abnormalities, few studies have addressed the colon. Preliminary studies in our laboratory using an established animal model of diabetes (the streptozotocin-induced diabetic rat, which at 8-wk duration of diabetes shows signs of gastrointestinal disturbance) found that the spontaneous contractile activity recorded from the rat proximal circular colonic muscle in vitro was increased in tissue taken from diabetic animals compared with controls. This spontaneous activity was primarily of myogenic origin (9). The present study was undertaken to further examine the origin of this spontaneous activity and to identify what may underlie its alteration in the diabetic tissue.
The motor pattern of the rat colon circular muscle is dominated by rhythmic contractions that are myogenic in origin (9, 47) with well-defined associated electrical activities possibly driven by interstitial cells of Cajal (ICC) (1, 32). The dominant motor pattern is one of low frequency (∼2.5 contractions/min in the proximal colon and declining toward the distal colon) and high amplitude, associated with slow wave activity that originates in ICC associated with Auerbach's plexus (ICC-AP) based on dissection studies in vitro (32). This is supported by studies in vitro (1) and in vivo (41) on Ws mutant rats, which show significant reductions in ICC-AP and do not exhibit high-amplitude, low-frequency rhythmic contractions seen in wild-type animals. It is this motor pattern that was dominant and studied in the present investigation. In addition to this motor pattern, low-amplitude, high-frequency contractions are observed in the rat colon that appear associated with ICC of the submuscular plexus (ICC-SMP) (32). The notion that two pacemaker systems (two types of ICC) independently orchestrate two different motor patterns has also been observed in the canine colon (37), although the dominant pacemaker resides in ICC-SMP (22, 38). In the cat colon, the dominant pacemaker also resides in ICC-SMP (6). The circular muscle of the human colon exhibits highly variable motor activity (17, 19) with evidence that ICC-SMP are associated with the rhythmic contractions (34). There is no ideal animal model for the circular musculature of the human colon. The dog and cat colon are not ideal, since the robustness of their dominant motor activity is not mimicked in human colonic preparation. In that sense, the pig colon is more like the human in that the frequency of the motor activity is easily modified by stretch and excitatory and inhibitory neural activities (16). The role of ICC in the pig colon has not yet been elucidated. The rat model is possibly a good model for the human colon, but evaluation of this awaits further studies into the role of ICC-AP and ICC-SMP in control of the human colon musculature (17, 19).
One feature of streptozotocin-induced diabetes in the rat is an enlarged small intestine, which appears primarily due to an increase in mucosal mass and secondarily to an increase in muscle mass (30). Despite a 39% (30) or 76% (21) increase in small intestinal muscle mass, no changes in contractile activity in response to carbachol were observed (30), although responses to electrical stimulation of cholinergic nerves were reduced (28), possibly associated with an increase in acetylcholinesterase (21). Interestingly, in the colon, no change in the weight of the muscle layers was observed (21).
The occurrence of neuropathy in diabetes is well known, but animal models show marked variability in type and extent of injury, and few studies address the colon. Neuropathy involving parasympathetic and sympathetic nerves and nonadrenergic, noncholinergic nerves have been reported (2, 21, 45, 46), as well as alterations in the responsiveness to agonists such as carbachol (3), prostaglandin F2α (42), and electrical field stimulation (29). Since spontaneous contractile activity in the gut is known to be affected by nitrergic nerves (11a, 31a), and since changes in nitrergic nerves have been reported in a model of spontaneous diabetes (47), we investigated changes in nNOS immunopositivity as possibly related to increased motor activity.
Diabetes has been shown to be associated with loss of ICC in both animal models and in patients (14, 31, 48). Loss and injury to ICC can result in defective gastrointestinal motility (5, 36, 39). Since our preliminary data suggested a myogenic (nonneural) origin of the increased motor activity, we investigated a possible role of injury to ICC. We investigated the hypothesis that increased muscle excitability was underlying the increased force of spontaneous contractions due to loss of or injury to ICC. Earlier studies on the dog colon established that removal of ICC caused depolarization of smooth muscle cells (22).
Male Wistar rats weighing 200–350 g were used in the present study. Animals were assigned to one of two groups: diabetic or age-matched controls. Housing conditions and all experimental work were conducted in accordance with the Animals (Scientific Procedures) Act 1986 under project license no. PPL 70 4649 with a project title of Gastrointestinal Research.
Induction of diabetes.
Diabetes was induced with a single intraperitoneal injection of streptozotocin (STZ; 65 mg/kg, 1 ml/100 g body wt) freshly dissolved in 20 mM citrate buffer solution (pH 4.5). Age-matched controls were injected with an equal volume of citrate buffer. The rats were housed two per cage (one control, one STZ treated) and provided with 2% sucrose water for the first 48 h after injection to prevent hypoglycemia. Diabetes was verified by a blood glucose level of ≥200 mg/dl (42), measured using an Accu-Chek active blood glucose testing kit (Roche Diagnostics) on blood taken from the tail vein. In addition, further evidence of the successful inducement of diabetes in the STZ-injected rats included weight loss, polydipsia, and polyuria.
Eight weeks after injection, control and STZ-diabetic rats were placed in separate cages for 24 h, and either food was withheld or food consumption was measured and compared using Student's t-test for unpaired observations. Rats were subsequently killed by carbon dioxide asphyxiation, and the body weight and blood glucose levels were measured. The abdominal cavity was opened via a midline incision, and the colon was removed and immediately immersed in Krebs solution (composition in mM: 118.3 NaCl, 4.7 KCl, 1.2 MgSO4, 1.2 K2HPO4, 2.5 CaCl2, 25 NaHCO3, and 11.1 d-glucose) gassed with 95%O2-5% CO2.
Pieces of colon ∼1 cm in length were removed from the proximal end and opened out, the contents were removed with the aid of a cotton bud, and segments of tissue were cut to similar size (∼10 × 2 mm) and mounted on the circular axis in 30-ml organ baths filled with gassed Krebs solution and maintained at 37°C. The tissues were attached via a thread to force displacement transducers (Dynamometer UF1), which were connected to a chart recorder (Lectromed 5041) via a preamplifier (Lectromed 3552) to record changes in isometric tension.
Whole tissue wet and dry weights and lengths.
The colon was trimmed of any debris (fat and connective tissue) and opened along a midline incision. The content was removed using a stream of running water, after which the colon was blotted dry and laid out straight on a piece of aluminum foil. The colon was allowed to dry in the open air at room temperature for one week before the weight and length was measured.
Student's t-test was used to test for significance between dry control and diabetic tissue weights and lengths. The weight was also expressed as a factor of the length of the tissue (g/cm).
Experiments were performed in tissues taken from animals either fed ad libitum up to the point of death or starved (but given free access to water) for 24 h before death. After the tissues were mounted in the organ baths, no initial tension was applied to each tissue, and they were allowed to equilibrate for at least 30 min before the experiments were commenced. Cumulative tension-response curves were performed, where tissues were stretched to the point that the tension increased by 100-mg increments, with the spontaneous activity allowed to plateau before the next increment was added. The length and spontaneous activity developed were measured after each increment of tension was added, and results were expressed as both developed tension (average values, g) vs. applied tension (g) and tissue length (cm) vs. applied tension (g). The dry weights of the colon preparations were recorded, and the cross-sectional areas of the colonic tissue preparations also were calculated to ensure that any differences in the responses observed between the control and diabetic tissues were not a result of differences in the size of the tissue preparations. These experiments were repeated in the presence of TTX (1 μM) in the bathing fluid and in tissues taken from animals fed ad libitum. Two-way ANOVA followed by the Bonferroni modified t-test was used to test for significant differences between the lengths or developed tensions of the control and diabetic tissues at each applied tension. Probability levels <0.05 (P < 0.05) were taken to indicate statistical significance. Cumulative concentration-effect curves for KCl were also performed, and any alteration in tension (from baseline levels) observed was allowed to plateau before the next addition of KCl. The response to each concentration of KCl was measured from the baseline tension before drug administration. As with the tension-response curves, two-way ANOVA followed by the Bonferroni modified t-test was used to test for significant differences between the control and diabetic tissue responses at each concentration. Probability levels <0.05 (P < 0.05) were taken to indicate statistical significance.
Immunohistochemistry and quantification.
For light microscopic studies, tissues were removed from both control and 8-wk STZ-diabetic rats and washed in Krebs-Ringer bicarbonate solution. Fresh tissues were directly embedded in Tissue-Tek and frozen with isopentane immerged in liquid nitrogen. Two-in-one group serial frozen sections of 8 μm were cut with a cryostat and fixed in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 10 min at 4°C. After being washed for 30 min in 0.05 M phosphate-buffered saline (pH 7.4, with 0.3% Triton X-100), the fixed sections were either two-in-one serial sections for c-Kit and protein gene product 9.5 (PGP 9.5) immunohistochemical staining, respectively, or individual sections for neuronal nitric oxide synthase (nNOS) immunohistochemistry. After quenching of endogenous peroxidase, sections were incubated with either polyclonal rabbit anti-c-Kit (1:200; DAKO, Glostrup, Denmark), rabbit anti-PGP 9.5 (1:2,500; Chemicon International, Temecula, CA), or rabbit anti-nNOS (1:2,500; Chemicon International). All the incubation times were 18–24 h at room temperature. Secondary immunoreactions were carried out with Vectastain ABC kits (with biotinylated anti-rabbit IgG). 3,3′-Diaminobenzidine was used as a peroxidase substrate for the ABC technique. Primary antibodies were omitted from incubation solution for negative control.
Tissues from four control and four diabetic animals were examined using a conventional microscope with an attached digital camera (Sony 3CCD, model DXC-930). Fifty sections immunostained with c-Kit and PGP 9.5 were chosen, respectively. c-Kit and PGP 9.5 immunopositivities were identified and highlighted using density slicing on color scale images with Photoshop version 7.0 (Adobe Systems, Mountain View, CA). The area of immunopositive cells on each image was measured and expressed as a percentage of total area according to established methods (20, 49). Regions with increased background staining, occasionally found at the borders of the image, were excluded from the analysis. Because of stronger staining and different shape, mast cells were easily distinguished from ICC and excluded before the analysis. Mast cells were rare in the musculature of the rat colon and were easily eliminated by erasing the immunopositivity before quantification. Values are expressed as means ± SE. Means were compared using Student's unpaired t-test. Data were considered statistically significant when P < 0.05.
Both control and STZ-diabetic rats were used for electron microscopic study. Proximal colon (∼1–2 cm from the ileocolonic junction) was fixed with 2% paraformaldehyde, 2.5% glutaraldehyde, 3% sucrose, and 1.25 mM CaCl2 in 0.05 M cacodylate buffer (pH 7.4) at 4°C for 2 h. They were then postfixed in 2% osmium tetroxide for 1 h, stained en bloc with 2% aqueous uranyl acetate for 30 min, dehydrated, infiltrated, and embedded in Epon-Araldite resin. Ultrathin sections were cut parallel to the circular muscle layer and stained with lead citrate for 3 min before being viewed with a transmission electron microscope (JEOL 1200EX Biosystem).
Compounds and solvents.
N-methylnitrocarbamyl-d-glucosamine (STZ) and TTX were obtained from Sigma (Poole, UK). KCl was obtained from Fisher Chemicals (Loughborough, UK). STZ was dissolved in citrate buffer (0.9% NaCl and 60 μM citric acid, pH 4.5). TTX was dissolved in distilled water to a stock concentration of 1 mM. KCl was dissolved in distilled water to a stock of 3 M.
Blood glucose levels.
Blood glucose levels were significantly elevated in diabetic rats (517.5 ± 0.7 mg/dl, n = 18) compared with controls (106.1 ± 0.1 mg/dl, n = 18). All the STZ-treated rats used also showed other symptoms of diabetes, including lack of weight gain, polydipsia, polyuria, and distension of the intestines.
In rats fed ad libitum up to the point of death, the food consumption measured in the final 24 h was found to be significantly greater in diabetic rats (65.8 ± 2.6 g) compared with controls (31.3 ± 1.7g, P < 0.05; Fig. 1). It is known that the deficiency of insulin leads to leptin deficiency in both diabetic patients and STZ-induced diabetic rats (12, 13), which in turn prompts hyperphagia.
Dry weights and lengths.
After the tissues were allowed to dry for 1 wk, the lengths and weights were measured, and it was observed that in the colons taken from diabetic rats, the lengths (12.8 ± 0.6 cm, control vs. 16.8 ± 1.1 cm, STZ; Fig. 2A) and weights (0.34 ± 0.01 g, control vs. 0.47 ± 0.02 g, STZ; Fig. 2B) were significantly increased compared with control values. When the dry weight of the colon was expressed as a factor of the dry length, there was no significant difference between the control and diabetic tissues expressed per centimeter of dry tissue length (27 ± 1 mg/cm, control vs. 28 ± 1 mg/cm, STZ; Fig. 2C). This indicates that changes in the weight of the tissue were in fact due to the increased length measured and not to either hyperplasia or edema. Increased colonic length has not previously been reported in the literature.
Effect of applied tension on spontaneous activity.
Cumulative tension-response curves were performed on strips of circular muscle in the control and diabetic proximal colon. Tension was applied by increments of 100 mg. Without any tension applied, only rhythmic phasic contractile activity of very low amplitude was present, but with increasing tension, rhythmic phasic contractile activity developed with increasing amplitudes until a plateau was reached (n = 10; Fig. 3A). The activity that developed was significantly greater in amplitude in the diabetic tissues from an applied tension of 200 mg and above (Fig. 3B). For example, at 700 mg of applied tension, control strips attained a force of contraction of 0.35 g, whereas the diabetic tissue attained a force of contraction of 1.30 g (370% of control value). The lengths of both the control and diabetic tissue preparations were similar at the beginning of the experiments and were shown to increase approximately proportionally with the tension applied to the tissue. There was no difference between the lengths of the control and diabetic tissues at any applied tension (Fig. 3C). Similarly, the frequency of contractions remained comparable between control and diabetic tissue preparations (11.0 ± 2.3 contractions/5 min, control vs. 12.5 ± 1.7 contractions/5 min, STZ).
In the presence of TTX (1 μM) in tissues taken from rats fed ad libitum, the amplitude of the tension-induced rhythmic phasic contractile activity generated in the diabetic colon was significantly greater than that in the control tissues at all increments of applied tension (P < 0.05; Fig. 3, D and E). There also was a proportional relationship between tissue length and applied tension that was not different between control and diabetic colon preparations, as seen in experiments performed in the absence of TTX (n = 8; Fig. 3F). Compared with the studies performed in the absence of TTX, the amplitude of the control activity in the presence of TTX was increased at all increments of tension (P < 0.05). However, in the diabetic tissues, TTX caused a greater increase in the amplitude of the spontaneous activity than that observed in control colon preparations. For example, at 500 mg of applied tension, the force of contraction in the control colon attained 0.51 g, compared with 0.38 g in the absence of TTX, a 25% increase. In the diabetic colon, the values were 1.66 and 1.22 g, respectively, a 35% increase. The average values of the increase at all levels of tension were 17.9 ± 4.2% (control) and 31.0 ± 1.0% (STZ) (P < 0.05). Hence, the difference in the neuronal component between control and diabetic colon was 13%, a significant difference but small compared with the increase in myogenic activity.
To avoid prior colonic distention due to content that might change sensitivity to distension, we performed tension-response curves in colon preparations taken from control and diabetic rats starved for 24 h before death (Fig. 4, A and B). These were shown to produce tension-induced rhythmic phasic contractile activity in both control and diabetic tissues that was not significantly different in amplitude from that observed in tissues taken from rats fed ad libitum in response to similar increments of tension (Fig. 3, A and B; P < 0.05). Again, as previously, the lengths of the tissues were not significantly different between control and diabetic at any point in the experiment (Fig. 4C; P < 0.05).
Concentration-effect curves to KCl.
Concentration-effect curves to KCl (30, 60, 90, and 120 mM) were carried out in both control and diabetic colon taken from rats fed ad libitum. The diabetic colon preparations showed significantly larger contractile responses to K+ depolarization compared with control preparations at all concentrations tested (n = 8, P < 0.05; Fig. 5).
Immunohistochemistry: distribution of ICC and nerves.
In the control rat colon, there are three groups of c-Kit-positive ICC: ICC-AP are associated with the myenteric (Auerbach's) plexus, ICC-IM are scattered within both muscle layers, and ICC-SMP are located in submuscular plexus lining the surface of the inner circular muscle layer (Fig. 6 a). In STZ-diabetic rats, both ICC-IM and ICC-SMP were found to be scarce compared with the control tissues (Fig. 6c). Quantification showed that there was no significant difference in ICC-AP immunoreactivity between control and STZ-diabetic colon (P = 0.43). However, the area occupied by ICC-IM and ICC-SMP was significantly lower in STZ-diabetic colon compared with control colon (P < 0.0002 for ICC-IM and P < 0.004 for ICC-SMP) (Fig. 7A). In both control and STZ-diabetic rats, there was no detectable difference in the densities of ICC-IM between circular and longitudinal muscle layers (Fig. 6, a and c).
Neuronal elements were stained using a general neuronal maker, PGP 9.5. Figure 6b shows numerous enteric neurons within the myenteric plexus and abundant nerve fibers within both muscle layers in control rat colon. In contrast, the amount of nerve structures in diabetic colon was lower, especially those nerves scattered within the musculature (Fig. 6d). Quantification of PGP 9.5 immunoreactivity showed that in the diabetic colon, the area occupied by enteric nerves was significantly lower outside the myenteric plexus (P < 0.0002), but no difference was detected within the myenteric plexus (P = 0.30; Fig. 7B). The hypothesis was investigated that nitrergic nerves were lost. nNOS immunohistochemistry showed marked presence of nNOS-positive nerves in the myenteric plexus as well as both muscle layers. No difference was found in the density of nitrergic nerves between control (Fig. 8 a) and diabetic (Fig. 8b) rats (n = 5).
Electron microscopy: ultrastructure of ICC and nerves.
In control rat colon, ICC were identified at the sites of the myenteric plexus (ICC-AP, Fig. 9 a) and submuscular plexus at the submucosal surface of the circular muscle (ICC-SMP, Fig. 9c), as well as within the musculature (ICC-IM, Fig. 9b). ICC had a cytoplasm of high electron density, rich in cell organelles, including mitochondria, endoplasmic reticulum, and caveolae (Fig. 9, a2–c2). ICC-IM and ICC-SMP also displayed basal lamina and showed synapse-like connections with enteric nerves (Fig. 9b and inset) and close connection to smooth muscle cells. In STZ-diabetic colon, ICC-AP showed either a normal or a slightly injured structure (Fig. 10, a and b). Different degrees of injury of ICC-IM (Fig. 10, c and e) and ICC-SMP (Fig. 11, a–c) were observed in the diabetic colon. Swollen mitochondria (Figs. 10e and 11a), lamina bodies (Figs. 10c and 11b), and partial depletion (Fig. 11, b and c) were present within the cell bodies or processes of both ICC-IM (Fig. 10, c and e) and ICC-SMP (Fig. 11, a–c). Many ICC-IM lost synapse-like connection with the enteric nerves but usually still kept gap-junctional connection with smooth muscle cells (Fig. 10c and inset). In the usual location of ICC-IM, some fibroblast-like cells (with fibroblast ultrastructure) around the enteric nerves made gap-junctional connection with adjacent smooth muscle cells (Fig. 10d).
Enteric nerves also displayed remarkable injury in the diabetic colon. Swollen mitochondria were frequently found within the cytoplasm of enteric neurons (Fig. 10a) and axons (Fig. 10e). Synapse-like connections between ICC and nerve varicosities with special ultrastructural features such as thickened prejunctional membrane, narrow space between pre- and postjunctional membrane (20–25 nm), and synaptic vesicles being oriented toward the junction were rare compared with control tissues. This was likely due to injury to varicosities, since varicosities with extensive vacuolization were seen adjacent to ICC that kept gap-junctional connections with structurally normal smooth muscle cells, in the diabetic colon (Fig. 10c).
The motor pattern of the rat colon is dominated by high-amplitude, low-frequency contractions that are myogenic in origin (9, 47). These rhythmic contractions are linked to rhythmic slow depolarizations and to the presence of ICC-AP (1, 32, 32). Since distension is one of the major stimuli for induction of motor activity, we set out to investigate possible differences in distension-induced motor activity after induction of diabetes. We opted for a highly controlled and reproducible experimental protocol by applying increments of stretch induced by sequential addition of 100 mg of tension to in vitro muscle strips. Circular muscle strips were examined, since it is primarily the circular musculature that orchestrates peristaltic activity in the colon. The present study shows that without applied tension, no contractile activity is present and that a rhythmic myogenic motor pattern is induced by the first 100 mg of applied tension induced by stretching the tissue. The amplitude of these contractions increases with increased applied tension. In diabetic rats, the stretch (tension)-induced rhythmic contractile activity remained robust and of similar frequency but was significantly higher in amplitude compared with that in control rats.
The following three points warrant discussion. First, in the diabetic colon, the rhythmicity of contraction and the frequency of contraction remained unaffected, suggesting that the pacemaker system of the colon remained intact. Second, the increased force of contraction in the diabetic rats was primarily nonneural in origin, suggesting a cause within the musculature and possibly ICC. Third, a relatively minor increase in force of contraction appears to be due to alteration in neural innervation of the musculature, also possibly linked to ICC.
The rhythmic low-frequency contractions are associated with the presence of ICC-AP (1, 32). The present study observed little change in ICC-AP. Minor injury to individual ICC-AP is unlikely to have a major effect on the regulation of contractile activity, since ICC-AP are arranged in a network and a network can overcome some individual cells to be injured. If damage would have been more extensive but in relatively discreet areas, this could have affected proper propagation of electrical activities; however, such a type of injury was not observed.
The marked increase in force of rhythmic contractions in diabetic rats was not due to hyperplasia, since the muscle weight per length did not change; this is consistent with a previous study that also noted no change in weight per length in the colon despite hyperplasia in the small intestine (21). Increased muscle weight in the small intestine was also noted in a subsequent study (30). Furthermore, in the small intestine, proliferation of epithelial cells and hypertrophy of the mucosa have been observed in conjunction with alterations of the smooth muscle layer in rat duodenum (35). The development of rhythmic contractions in both control and diabetic colon was primarily due to myogenic mechanisms and/or ICC. Ultrastructural and immunohistochemical investigations did not show obvious changes in the smooth muscle. However, a marked change in ICC-SMP was evident. ICC-SMP execute a hyperpolarizing force onto smooth muscle cells, and injury or loss of ICC-SMP will cause depolarization and increased excitability, which can explain the increased force of contraction observed in the diabetic colon. First, when ICC were removed from the circular muscle of the dog colon, the adjacent musculature depolarized (22). The reason for this phenomenon may be twofold. First, the intrinsic membrane potential of ICC is likely hyperpolarized compared with smooth muscle cells, which influences neighboring smooth muscle cells (22). In addition, ICC are a source of carbon monoxide that hyperpolarizes smooth muscle cells (8, 11). The hypothesis that loss of ICC increases muscle excitability is also consistent with data from the mouse small intestine, where it was shown that the musculature was strongly depolarized in W/WV mutant mice (18, 24). It also is consistent with a study in nonobese diabetic (NOD)/LtJ mice, which found a relationship between loss of ICC and depolarization of antral smooth muscle (31). In summary, the loss of and injury to ICC-SMP observed in the present study is consistent with increased excitability of the musculature, which can explain the marked, TTX-insensitive increase in force of contraction induced by stretch. It also is consistent with the increased response of the diabetic colonic muscle preparations to KCl.
There also was a TTX-sensitive component in the increased force of the rhythmic contractions, although this was minor compared with the myogenic component. Spontaneous and/or tension-induced neural activity in the preparations under study was primarily inhibitory (or at least the net effect was inhibitory), since TTX increased force of contractions. Therefore, the hypothesis appears warranted that spontaneous and tension-induced neural activity in the diabetic colon harbors a larger inhibitory component compared with control tissue. Although the inhibitory component may well be nitrergic in origin, there is no evidence that the nitrergic component is larger in diabetic rats, since no changes in nNOS immunoreactivity were observed (present study) and Western blots did not show changes in nNOS expression (9). However, differences between control and diabetic rats were small, and the methods used to assess nNOS activity may not be sensitive enough to detect small (functional) changes. It is possible that a difference in the neural inhibitory component involves ATP or VIP. The loss of ICC may not be a major player in inhibitory neural motor activity, since purinergic innervation is not mediated by ICC (40), and increasing evidence suggests that nitrergic innervation also occurs in the absence of ICC (1, 4, 7), although this is still controversial (40, 44).
It also is possible that the net increase of inhibitory neural activity was due to loss of excitatory nerves. This would be consistent with the observation of the present study that PGP-9.5 immunopositivity was reduced. Since neurotransmission may be mediated in part by ICC, the TTX-sensitive component may be due to injury to ICC. If an excitatory neural pathway were to require synaptic contacts between motor neurons and ICC-IM as has been proposed in the stomach (43), then injury to ICC could cause loss of a component of excitatory innervation. Alterations in neural activity has been a focus of many studies based on the assumption that diabetes is associated with neuropathy. The studies on animal models, however, are conflicting and do not show a consistent pattern of certain types and extent of neural injury.
No abnormal rhythmicity and no abnormal ICC-AP were observed in the colon of diabetic rats in the present study. In contrast, dysrhythmia of the electrical activity of the gastric antrum occurred (23) and was correlated with the loss of ICC-AP and delayed gastric emptying in the spontaneously diabetic NOD/LtJ mouse model (31). Full antral wall thickness biopsies in human diabetic patients displaying gastroparesis also have shown a significant reduction of ICC (10). This is supported by evidence of a significant loss of ICC in both the myenteric plexus and circular muscle layers throughout the thickness of the diabetic human jejunum (14). Another study, in which the majority of the diabetic patients showed gastrointestinal symptoms such as constipation, also indicated a severe loss of ICC (only 40% of control levels) in the myenteric plexus and circular muscle regions, as detected by Kit immunohistochemistry. This study suggested that gastroenteropathies may be the result of the ICC deficiency observed (27). Comparing this literature to the present study, it is clear that diabetes may affect different organs of the gastrointestinal tract in markedly different ways. In addition, the present study shows that the development of diabetes may not affect all subtypes of ICC in a similar manner. In the colon, ICC-SMP as well as ICC-IM were affected most severely, and ICC-AP were not affected. Although in most pathologies, injury to ICC appears to be associated with reduction in motor activity, the present study shows that if ICC-SMP are affected but ICC-AP are not, the result is an increased excitability of the musculature and increased force of distension-induced rhythmic contractile activity.
The diabetic rats were eating significantly more food than the control rats within 1 wk after STZ injection. The ketone levels of the rats in this study were not measured, but previous work suggests that rats injected with STZ show clear signs of raised ketone levels after 12 wk of STZ-induced diabetes (15). However, other reports in the literature show that a lack of insulin leads to leptin deficiency in both humans and STZ-diabetic rats (12, 13) and that in both cases hyperphagia was observed. In these studies the marked increase in the presence of ketone bodies was not considered to be the cause of hyperphagia but merely as an indicator of the severe insulin deficiency responsible for reduced leptin signaling. Since STZ destroys the β-cells of the pancreas, it is safe to assume that our rats were also insulin deficient, which may explain the hyperphagia exhibited.
Megaintestine (an enlargement of the intestines) has been observed in both the colon and the ileum in a number of species and has been proposed to be secondary to diabetic autonomic neuropathy (25). It also has been suggested that this observation of megaintestine is an adaptation to the hyperphagia displayed by animals. However, there is conflicting evidence as to whether there is a mucosal or muscular proliferation (21, 30). Proliferation of both layers, in many tissues, has been reported. In rat duodenum, proliferation of the epithelial cells and hypertrophy of the mucosa have been observed in conjunction with alterations of the smooth muscle layer (35). These alterations of the muscle layer are considered to be an enlargement of the spaces between the muscle cells, and reports of enlarged basal lamina, excess collagen fibrils, increased levels of extracellular matrix proteins (particularly laminin-1 and fibronectin), and entrapped membranous materials increasing the space between smooth muscle cells in other tissues are common in diabetes (26, 35). Interestingly, in none of the above studies on megacolon or megaintestine were remarks made about changes in tissue length.
This work was funded by a grant from Diabetes UK and by operating grants from the Canadian Institutes of Health Research.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society