The loss of intrinsic neurons is an early event in inflammation of the rat intestine that precedes the growth of intestinal smooth muscle cells (ISMC). To study this relationship, we cocultured ISMC and myenteric plexus neurons from the rat small intestine and examined the effect of scorpion venom, a selective neurotoxin, on ISMC growth. By 5 days after neuronal ablation, ISMC number increased to 141 ± 13% (n = 6) and the uptake of [3H]thymidine in response to mitogenic stimulation was nearly doubled. Atropine caused a dose-dependent increase in [3H]thymidine uptake in cocultures, suggesting the involvement of neural stimulation of cholinergic receptors in regulation of ISMC growth. In contrast, coculture of ISMC with sympathetic neurons increased [3H]thymidine uptake by 45–80%, which was sensitive to propranolol (30 μM) and was lost when the neurons were separated from ISMC by a permeable filter. Western blotting showed that coculture with myenteric neurons increased α-smooth muscle-specific actin nearly threefold to a level close to ISMC in vivo. Therefore, factors derived from enteric neurons maintain the phenotype of ISMC through suppression of the growth response, whereas catecholamines released by neurons extrinsic to the intestine may stimulate their growth. Thus inflammation-induced damage to intestinal innervation may initiate or modulate ISMC hyperplasia.
- sympathetic neurons
- tissue culture
- Western blotting
inflammation affects all of the cellular systems within the intestine, altering their functions as well as their ability to deliver the integrated responses necessary for normal motility. However, these changes usually become reversed as the markers of inflammation subside. For example, animal models of intestinal inflammation have shown the development of accelerated intestinal transit and hyperreactivity of the smooth muscle, which returned to baseline values when inflammation was no longer present (reviewed in Ref. 9).
Although this suggests that the effect of inflammation on the control mechanisms of intestinal motility is transient and can be met with adequate repair or replacement of cellular components, we have recently found that some aspects of intestinal structure undergo significant, essentially irreversible changes that require functional adaptation for preservation of normal intestinal motility. For example, a significant loss of myenteric and submucosal neurons was found to occur during the first 48 h following induction of colitis in the rat, with no further change observed in the subsequent 8- to 10-day period of overt inflammation (26). In earlier work in this model, as well as in experimental inflammation of the rat jejunum, we had observed that marked hyperplasia of the intestinal smooth muscle cells (ISMC) occurred through days 4–6 following induction of inflammation (4, 18). Although mitosis of the smooth muscle was normally virtually undetectable, up to 6% of the cell population were mitotic during inflammation, and this gave rise to an essentially permanent increase in the smooth muscle cell number.
Of further significance to motility, examination of the effect of growth of smooth muscle on its differentiated nature showed a significant increase in cell size as well as alterations in amount and proportion of mRNA and protein for α-smooth muscle-specific actin (α-SM actin) (3). Although a cycle of dedifferentiation, growth, and resumption of contractile phenotype is normal for repair of smooth muscle (21), an altered outcome in the case of inflammation-induced hyperplasia suggests that changes occur at the cellular level of smooth muscle, which may then affect the contractile properties of the tissue.
These findings suggest that the impact of inflammation on the enteric nervous system (ENS) is an underestimated, early challenge to intestinal function that is followed closely by large increases in the number of smooth muscle cells, which require innervation for appropriate contractile responses. These changes to both the neuronal and smooth muscle cell populations could contribute to the altered motility seen during inflammation, and these irreversible or slowly reversing changes may contribute to the frequent reports of persistence of symptoms following acute inflammatory episodes (e.g., see Ref. 16).
Elsewhere in the nervous system, the close neuron-target cell relationship is essential for mutual survival and function. This suggested to us that disruption of normal innervation of intestinal smooth muscle could be implicated in the subsequent hyperplasia. Supporting evidence comes from the observations that the neural damage precedes the onset of smooth muscle mitosis, as well as restoration of the normal density of innervation of smooth muscle thereafter (26).
Specifically, we hypothesized that innervation of ISMC may directly or indirectly influence their growth. In the absence of evidence in the literature, we developed a tissue culture model to pursue this, involving the coculture of rat ISMC with neurons from the myenteric plexus. Since extrinsic innervation of the intestine might contrast with the effects of enteric innervation, we also examined the consequences of coculture of ISMC with neurons from the sympathetic nervous system. The strong and contrasting outcomes of these studies represent evidence for a new and significant aspect of the nerve-target cell relationship within the intestine.
Tissue culture of intestinal smooth muscle and myenteric neurons.
To obtain cultures of ISMC and neurons from the ENS, the longitudinal and circular smooth muscle layer and the enclosed myenteric plexus were removed from the small intestine of 3- to 4-wk-old male Sprague-Dawley rats (Charles River, Montreal, PQ, Canada) and dissociated using 0.25% trypsin II (Sigma, St. Louis, MO) in HEPES-buffered Hanks' saline (pH 7.35). Cell suspensions were plated onto 48-well culture plates previously coated with Matrigel (Collaborative Research). Medium (DMEM; GIBCO) containing 10% fetal calf serum (FCS) and 2.5% rat serum (RS) was then added and replaced at 48-h intervals. These were called “cocultures” to reflect the predominance of neurons and ISMC, although other cell types such as glial cells could also be present in low numbers. For subsequent use in coculture with sympathetic neurons, primary cultures of ISMC were established and maintained in DMEM plus 10% FCS alone, in which ENS neurons were undetectable after 5–7 days.
In some cultures, ENS neurons were selectively removed by treatment with the neurotoxin scorpion venom (Sigma) at 24 h. Scorpion venom contains selective agonists of the fast Na+ channel found on neurons (25), and in initial experiments, we found that a single treatment caused a dose-dependent neurotoxicity without affecting the glial or smooth muscle cells. Subsequently, we used scorpion venom at 30 μg/ml, about two times higher than the dose required for removal of all neurons in pilot experiments, and we followed this with immunocytochemistry of cohort cultures to confirm the absence of neurons.
Coculture of ISMC with sympathetic neurons.
Sympathetic neurons were established in vitro according to previously published techniques (2). Briefly, the superior cervical ganglia (SCG) from neonatal rats were dissociated and cultured on Matrigel-coated multiwell plates in medium containing 50 ng/ml 2.5 S nerve growth factor (NGF). Accessory cells were removed by an initial treatment with the antimitotic cytosine arabinoside (2.5 μM; Sigma), which resulted in a complex network of pure neurons by 2 days and thereafter.
ISMC were obtained by dissociation of 14- to 21-day-old primary cultures. These were added to 4- to 6-day-old cultures of SCG neurons or maintained in pure control on similarly treated wells. NGF was added to all wells at 48-h intervals. In coculture, the ISMC adhered and grew underneath the neurons, which responded by forming extensive contacts through the growth and rearrangement of neurites. Ganglion-like plexuses developed on top of the smooth muscle by 2–3 days, with extensive neurite contact with the ISMC. The cultures were then synchronized and tested for growth response using [3H]thymidine uptake, with the level of NGF maintained by addition of concentrated aliquots when replacement of medium was not carried out.
Immunocytochemistry and histochemistry.
Immunocytochemistry with the pan-neuronal marker PGP 9.5 (Ultraclone, Isle of Wight, UK) was used to study the presence of neurons within cultures of ISMC. Briefly, cultures were fixed with neutral buffered formalin, washed, incubated overnight at room temperature with PGP 9.5 (1/1,000 in phosphate-buffered saline with 0.2% Triton X-100), followed by a biotinylated secondary anti-rabbit antibody (1/300; DAKO) and visualized with Cy3-labeled streptavidin (1/500; Jackson Labs), each for 1 h at room temperature. Staining was observed with an Olympus BX-60 microscope, and images were digitally captured (ImagePro Plus; Media Cybernetics).
Immunocytochemistry with a mouse monoclonal antibody to α-SM actin (1A4; DAKO) was used routinely to verify the nature of cells growing in culture, as reported earlier (3). An anti-mouse secondary antibody labeled with Alexa-488 (Molecular Probes) was used for dual labeling of neurons and ISMC in coculture, with digital recombination of images.
Glyoxylic acid-induced histofluorescence was used to visualize monoaminergic nerve fibers in whole mounts of intestinal smooth muscle (14). Preparations of the intact smooth muscle wall were immersed in glyoxylic acid (0.9%, 20 min), stretched flat and dried, and then reacted at 100°C for 4 min. After mounting in paraffin oil, the preparation was viewed under fluorescent illumination to detect the presence of catecholamines.
The growth of ISMC in vitro was evaluated by direct counting of cell number or by [3H]thymidine incorporation (reflecting the entry into S phase of mitosis). Results are reported as the average of individual experiments, and each experiment was the average from triplicate wells.
For determination of cell number, cultures were dissociated with 0.15% trypsin in HEPES-buffered Hanks' saline, and cell number was determined with a hemocytometer. To assess [3H]thymidine incorporation, a standard protocol was used involving the synchronization of mitotic cycle before testing of the growth response of ISMC, either in pure culture or in cocultures. All cultures were washed, exposed to fresh medium containing serum as appropriate, and changed to DMEM without serum 24 h later. After further incubation for 72 h, all media was replaced with either medium alone (baseline condition) or medium containing mitogenic stimuli. These were FCS, RS, or 10 ng/ml platelet-derived growth factor (PDGF; UBI). At 20 h following this, [3H]thymidine was added for 4 h, followed by routine processing for scintillation counting.
The presence and relative abundance of α-SM actin was studied using Western blotting and videodensitometry as previously described (3). Briefly, ISMC in culture or freshly dissected smooth muscle tissue were enzymatically dissociated and ISMC number was determined by hemocytometer counting in triplicates. Cell suspensions were centrifuged to a pellet and resuspended in lysis buffer containing protease inhibitors. Equal volumes of 2× sample buffer were added to the tissue lysates, and the samples were boiled for 5 min, subsequently resolved by SDS-PAGE, transferred to polyvinylidene difluoride membranes, blocked in 5% BSA, and incubated overnight at 4°C with an antibody to α-SM actin (1A4) at 1:1,000 in Tris-buffered saline (TBS) containing 0.2% Tween 20 (TBS-T). The blots were washed in TBS, incubated for 2 h at room temperature with biotinylated goat anti-mouse IgG (DAKO) at 1:3,000 in TBS-T, washed again, and incubated for 2 h with horseradish peroxidase-conjugated streptavidin antibody (1:5,000; DAKO). The blots were washed again and immersed in a substrate solution of 50 mM Tris containing 0.06% diaminobenzidine and 0.0125% H2O2 (vol/vol).
Following development, the blots were analyzed by videodensitometry (ImagePro Plus), as reported previously (3). All blots were analyzed simultaneously, expressing each band as an arbitrary value reflecting the product of density and area [integrated optical density (IOD)]. This value was then expressed as a ratio in terms of the number of ISMC loaded initially, allowing comparison of the actin content of ISMC among different experimental conditions.
ENS neurons and ISMC in coculture.
Cultures of ENS neurons and ISMC were established from the jejunum or colon from 2- to 4-wk-old rats. Initially, ISMC formed a relatively uniform layer 1–2 cells thick that showed regular and coordinated contraction. The cell shape was typically bipolar and spindle- or ribbon-shaped and was uniform throughout the cultures. By 3–4 days in vitro, spontaneous contraction was infrequent and ISMC were arranged in multilayered regions with relatively sparse intervening areas (Fig.1 A). Local retraction of the cellular sheet began to develop, and after 7 days, retraction of the entire cell mass from the culture surface began to occur.
Immunocytochemistry for the pan-neuronal marker PGP 9.5 showed that ENS neurons represented 7.8 ± 1.2% (n = 6) of the initial cells at 24 h in vitro. By 5–6 days in culture, this technique showed a uniform distribution of neurons among the ISMC, with close examination showing the extension of neurites from ENS neurons among the adjacent cells (Fig. 1). These formed a branching field with a mean radius of 160 ± 15 μm (n = 8). These were closely apposed to neighboring cells, which were identified as ISMC by positive staining with an antibody to α-SM actin (Fig.1 D). Therefore, each ENS neuron could influence ISMC within a diameter of 200–400 μm, and, typically, ISMC throughout the culture well received contact from at least one ENS neuron. These were called cocultures to reflect the predominance of neurons and ISMC, although other cell types such as glial cells could also be present in low numbers.
ENS innervation regulates the growth response of ISMC.
Initially, direct determination of cell number was used to evaluate ISMC growth, and this was found to increase by two- to threefold over a 6-day period. To study the effect of loss of innervation on growth of ISMC, some cultures were treated with scorpion venom (30 μg/ml), a selective toxic agonist that targets the Na+ channels present on neurons. Immunocytochemistry showed that ENS neurons were undetectable after 18 h, without direct damage to ISMC. ISMC number was determined in cultures exposed to scorpion venom 6 days earlier, and this was compared with ISMC number in untreated cohort cultures. Figure 2 shows the results of a typical experiment, in which the treatment with scorpion venom and the resulting loss of enteric neurons led to an increased ISMC number compared with controls. On average, treatment of cocultures of ISMC from the jejunum caused a 41 ± 13% increase in ISMC number (n = 6 experiments; P < 0.05).
We hypothesized that the presence of innervation by ENS neurons in vitro might correlate with a decreased response to serum factors present in the culture medium. To test this, we assessed the uptake of [3H]thymidine by ISMC following addition of a mitogenic stimulus in control cultures, finding that addition of FCS or RS caused a reproducible and dose-dependent uptake of [3H]thymidine. In experiments evaluating the effect of treatment with scorpion venom, ISMC in cultures were typically more responsive to stimuli such as 2.5% RS, as well as to the control manipulation of addition of medium alone (Fig.3 A). Therefore, the responses to stimuli were calculated relative to appropriate internal controls (i.e., serum-free medium alone). Figure 3 B shows that the treatment with scorpion venom at day 1 strongly affected the subsequent (day 6) uptake of [3H]thymidine by ISMC, nearly doubling the response to 5% FCS, whereas the response to 2.5% RS increased by 50% (Fig. 3 B). Therefore, coculture of ENS neurons with ISMC suppressed the ISMC growth response, thus explaining the increase in ISMC number that was measured following loss of the ENS neurons.
ENS neurons might regulate the growth response of ISMC by the release of neurotransmitters, which could act either acutely or chronically to modify the growth patterns. To test the consequences of neurotransmitter release on ISMC growth response, we first examined the acute effect of scorpion venom, added simultaneously with the mitogenic factor on day 5 of culture, and compared this with untreated cohort cultures. No effect was seen (Fig. 3 B), indicating that under these circumstances at least, the acute release of neurotransmitters from ENS neurons did not influence the response of ISMC to growth factors.
Innervation in vitro might modify the growth response by the chronic release of neurotransmitters. To partially test this, hexamethonium or atropine was added twice daily to cause the chronic blockade of nicotinic or muscarinic cholinergic receptors, respectively. Hexamethonium caused no significant change in subsequent [3H]thymidine uptake in response to 2.5% RS at doses of 1, 10, or 100 μg/ml (n = 3, P > 0.05; Fig. 4). However, atropine at doses >10−6 M (10−5 and 10−4 M) caused significant increase in [3H]thymidine uptake (Fig.4). Although this suggested that stimulation of muscarinic receptors on the ISMC was involved, daily application of the stable cholinergic agonist carbachol to ISMC in pure culture did not cause significant change in ISMC growth response (10−6–10−4 M; n = 3 per dose). Therefore, the cholinergic input to the ISMC from ENS neurons is implicated in the suppression of the growth response of ISMC but could not be mimicked by exogenous application.
Sympathetic innervation increases ISMC growth response.
The sympathetic nervous system has a trophic influence on vascular smooth muscle, and the neural mechanism is thought to involve local or circulating levels of catecholamines (5). Sympathetic innervation of the intestine is inhibitory, causing relaxation of smooth muscle largely by α-adrenergic receptor-mediated input to ENS neurons (15). To justify examination of a possible influence of sympathetic innervation on rat ISMC growth, we looked for evidence for direct innervation of ISMC using histochemistry for catecholamines in whole mounts of the smooth muscle layers of rat jejunum. As expected, this showed dense innervation of the myenteric plexus neurons but also revealed the spread of nerves within the smooth muscle (Fig. 5 A). These progressively decreased in caliber as they branched, and single axons with highly fluorescent varicosities were seen to terminate among the ISMC (Fig. 5 A). Therefore, we examined the possibility that sympathetic innervation might directly influence ISMC growth response.
Rat sympathetic neurons were obtained from the SCG of neonatal rats and cultured with 50 ng/ml of NGF for 6 days. The initial 48-h exposure to the antimitotic cytosine arabinoside (10−6 M) removed all accessory cells, and the pure cultures of SCG neurons formed a well-dispersed, interconnected network. Rat ISMC were obtained by enzymatic dispersion of 14- to 21-day-old primary cultures, maintained without RS (thus lacking ENS neurons), and added to cultures of SCG neurons. Figure 5 B shows the development of numerous close interactions among the ISMC and the SCG neurons at 4 days in vitro.
In these cocultures (SCG + ISMC), it was possible that the presence of SCG neurons alone or the NGF required for their culture could directly affect the uptake of [3H]thymidine, and this was examined in control experiments. Pure cultures of SCG neurons took up insignificant amounts of [3H]thymidine that did not increase with the addition of 10% FCS. Also, the addition of NGF required for SCG culture (50 ng/ml) to pure cultures of ISMC did not affect their subsequent thymidine uptake in response to stimulation (similar to untreated controls; P > 0.05). Therefore, we proceeded to examine the response to growth factors in cocultures of SCG + ISMC, comparing this with cohort cultures of pure ISMC.
ISMC cocultured with SCG neurons showed a significant and consistent increase in [3H]thymidine uptake in response to growth factors, up to 1.8 times greater than the cohort ISMC cultures (Fig.5 C). In separate experiments, the responses to FCS at 5% and 10% increased by nearly 50%, whereas a greater increase was seen with PDGF (10 ng/ml). Therefore, sympathetic neurons cocultured with ISMC can enhance their response to growth factors.
Catecholamines exert a trophic influence on vascular smooth muscle in vivo and in vitro (17). In the present study, the SCG neurons are adrenergic (30) and therefore their coculture with ISMC might lead to increased responsiveness to mitogens by a similar mechanism. To test this, the effect of the β-adrenoceptor antagonist propranolol (10 μM) on the mitogenic response of ISMC was tested, both in pure culture or with ISMC cocultured with sympathetic neurons. Propranolol was added either 60 min before addition of a mitogenic stimulus to the ISMC or chronically (from the time of initiation of coculture). Both acute and chronic treatment with propranolol reduced the uptake of [3H]thymidine by pure ISMC in response to FCS, and there was a significantly greater effect seen with prolonged treatment (Fig. 5 D). However, the degree of effect was similar between ISMC cocultured with neurons and those in pure culture, and the hyperresponsiveness of innervated ISMC remained present in each case (Fig. 5 D).
To investigate the requirement for close contact, SCG neurons were grown on permeable filter inserts before being placed into wells containing ISMC. Experiments were carried out to determine whether this form of coculture also affected ISMC response to growth factors, but results showed that [3H]thymidine uptake was equal among ISMC in pure culture and cohort ISMC maintained with SCG neurons grown on inserts (n = 3; P > 0.8). Therefore, close contact was a requirement for growth response.
ENS innervation maintains ISMC phenotype.
Prolonged mitosis of vascular smooth muscle cells in vitro and in vivo is associated with alterations in intracellular protein content and an increased emphasis on protein synthesis (8). Therefore, inhibition of ISMC growth by ENS neurons in vitro might preserve the differentiated contractile phenotype. To test this, we examined the effects of coculture with ENS neurons on the α-SM actin content of ISMC.
At 6 days after neuronal ablation using scorpion venom as before, ISMC were freed from treated or cohort control cultures by enzymatic dissociation. Following homogenization, Western blotting with the 1A4 antibody was used to detect α-SM actin. Since it was possible that significant alterations in ISMC size might influence the interpretation of these blots, total protein per ISMC was determined in parallel and used as an indicator of cell size. These values were similar between cultures of control ISMC (8.6 ± 1.6 ng/cell; n = 7) and scorpion venom-treated ISMC (7.6 ± 1.5 ng/cell;n = 7). Therefore, we used video densitometry to analyze the levels of α-SM actin per ISMC (expressed as IOD/cell), with data derived from experiments using a range of 10,000–60,000 ISMC per lane (Fig. 6).
The mean IOD per cell for ISMC grown in pure culture (i.e., without ENS neurons) was 0.012 ± 0.004 (n = 3), which was more than threefold lower than the value for freshly isolated ISMC [determined from aliquots of cells taken before culturing; 0.042 ± 0.001 (n = 3)]. However, the mean IOD/cell for cultured ISMC was substantially higher following coculture with ENS neurons [0.031 ± 0.004 (n = 3)]. This indicates that ISMC cocultured with ENS neurons contain more α-SM actin than ISMC in pure culture and are closer to, but still significantly less than, that of freshly isolated ISMC (P > 0.05).
Therefore, examination of the effect of coculture of ISMC with ENS neurons has shown consistent changes in their growth response, total cell number, and intracellular contents. We conclude that innervation in vitro leads to the development and/or maintenance of a more differentiated cell that is relatively less responsive to growth factors.
Recent work in animal models of intestinal inflammation showed that damage to the ENS was an early event, preceding the onset of hyperplasia of smooth muscle in both the circular and longitudinal muscle layers (26). We suspected a causal relationship, proposing that ENS-derived factors normally regulate the proliferative state of smooth muscle, and investigated this through development of a tissue culture model. In this model, myenteric plexus neurons survived dissociation from the tissue well and rapidly extended dense arrays of neurites among the neighboring smooth muscle cells. The smooth muscle cells were uniform in appearance, and all showed at least some expression of α-SM actin following immunocytochemistry with the 1A4 antibody. The initial ratio of 1:8 of neurons among the smooth muscle cells allowed good opportunity for interactions and the possible development of growth control.
To study the effect of neuronal coculture on smooth muscle growth, we used scorpion venom to cause the selective ablation of neurons and compared the growth of ISMC in these cultures with that in untreated parallel but otherwise identical cultures. After 5 days, the culture wells treated with scorpion venom contained significantly more ISMC than the untreated controls. The possibility that neuronal coculture led to suppression of the response of ISMC to growth factors was investigated through growth assays involving [3H]thymidine incorporation, showing that the presence of enteric neurons was associated with a significant suppression of the ISMC response to serum mitogens.
Our investigation of the possible mechanisms involved the testing of the role of cholinergic neurotransmitter release in regulation of ISMC growth. Hexamethonium had no effect over the dose range used, whereas atropine increased the growth response of ISMC in cocultures. This is interpreted as evidence that suppression of nicotinic cholinergic receptors, involved in interneuronal interactions, did not affect neurally mediated growth control and provided support for a role for direct neuron-smooth muscle interactions acting via muscarinic receptors on the ISMC. Although daily application of carbachol to pure cultures of ISMC had no effect, this may not adequately mimic the local release of acetylcholine into the cellular microenvironment.
Clearly, the complex array of neuropeptides that are present in the ENS are expected to be present in vitro as well, requiring a more extensive examination of the effect of both excitatory and inhibitory factors. For example, ATP has recently been described both as a neurotransmitter in guinea pig enteric neurons (1) and as a growth factor for vascular smooth muscle (12). In addition, vasoactive intestinal polypeptide inhibited the growth response of rabbit colonic smooth muscle (29), whereas neuropeptide Y stimulated growth in human vascular smooth muscle cells (13).
Examination of the consequences of coculture of ISMC with pure cultures of rat sympathetic neurons showed significant enhancement of the ISMC growth response, the opposite effect from that seen with enteric neurons. The close proximity of these neurons among the ISMC was required, since their separation by a permeable filter removed the stimulatory effect. Under these culture conditions, sympathetic neurons remain adrenergic (30), which suggests that neurally released catecholamines stimulate ISMC growth. In support of this, both the acute and chronic addition of propranolol to ISMC in cocultures caused significant inhibition of the growth response, showing the sensitivity of ISMC to catecholamines. However, a similar response was seen among ISMC in pure culture when exposed to propranolol, which prevents the clearer identification of a neural mechanism for this effect. Although the presence and significance of direct adrenergic innervation of the rodent ISMC has been debated (15,27, 31), recent functional evidence has shown that sympathetic nerves innervate adrenoceptors of various types in the intestinal muscle layers (20). Although adrenergic innervation of ISMC has been difficult to demonstrate in standard sections (27), this was readily seen in whole mounts of rat jejunum and thus may reflect technical differences.
Elsewhere, a trophic influence of the sympathetic nervous system on vascular smooth muscle is well established, and sympathetic hyperinnervation is thought to be a major determinant of intimal hyperplasia and increased blood pressure in the spontaneously hypertensive rat model (17). Catecholamines released by these neurons are held responsible for this effect and may act directly to cause increased smooth muscle growth. For example, supersensitivity to catecholamines without change in receptor number is indicated in hyperplasia of rabbit aortic smooth muscle in vivo (22). An indirect effect is also possible, since noradrenaline increased the number of receptors for PDGF in rat aortic smooth muscle in vitro (6). Studies in vitro have shown that sympathetic innervation acts via adrenergic neurotransmitters to cause hypertrophy of the cardiac myocyte (19) as well as regulation of ion channel physiology (32).
Our evidence suggests that neurally mediated release of neurotransmitters was responsible for growth control in vitro, rather than a nonspecific effect of contact with neural membrane. This was shown by the loss of neural growth control following the application of receptor antagonists and, most strikingly, by the opposite effects of coculture of ISMC with ENS vs. sympathetic neurons.
Neurally mediated inhibition or stimulation of smooth muscle growth in the intestine may also be a specific phenomenon that is linked to the neuronal phenotype. The outcome of a net increase in smooth muscle cell number during inflammation may potentially reflect an imbalance among the intrinsic and extrinsic neural influences on the ISMC. For example, the loss of enteric neurons requires sprouting from survivors into a new target field, which may take substantially longer than the regeneration of damaged sympathetic axons whose extrinsic cell bodies are intact.
There is extensive remodeling of innervation of the intestine, due in part to transient or permanent damage affecting both intrinsic and extrinsic innervation. For example, inflammation has been shown to impair the release of acetylcholine (10) and noradrenaline (28) from the muscle wall of the rat jejunum. In jejunitis, we have shown earlier that inflammation caused a permanent upregulation of choline acetyltransferase (11), evidence for increased synthesis of acetylcholine and potentially reflecting the response of intrinsic cholinergic enteric neurons to the increased target population of smooth muscle. As evidence of axonal proliferation, we have shown that the density of innervation of smooth muscle is maintained following inflammation-induced hyperplasia in the rat colon (26).
Smooth muscle cells are highly differentiated, which is evident from the large proportion of contractile filaments within the cell, and must undergo a process of dedifferentiation, referred to as “phenotypic modulation” for mitosis and an increase in cell number to occur (8, 23). In general, smooth muscle proliferation is associated with a decrease of contractile proteins and an increase in protein synthesis. Although these changes are normally reversed, smooth muscle in disease states such as atherosclerosis continue to show increased growth responsiveness, altered lipid metabolism, increased matrix production, and loss of contractile proteins (24). The dedifferentiation of smooth muscle into proliferative, synthetic myoblasts and their subsequent resumption of a myogenic program is necessary for normal homeostasis as well as pathogenesis.
Cultured gastrointestinal smooth muscle cells display a coordinated program of gene expression reflecting this process (7). Factors that regulate the extent of cell division may influence the outcome of cell division in vivo, leading to significant alterations in phenotype that are possibly only slowly reversed and may contribute to altered motility. For example, we found that smooth muscle cells in the inflamed intestine were substantially larger, with altered α-SM actin content (3). The inverse relationship between differentiation and cell division led us to examine the effect of innervation in vitro on the α-SM actin content of ISMC in vitro, using similar techniques as before (3). Although examination of total protein content as a measure of cell size showed no significant changes with coculture, semiquantitative Western blotting and video densitometry provided clear evidence for greatly increased actin content in ISMC cocultured with ENS neurons. This is interpreted as evidence that the innervation of ISMC has a critical role in maintenance of the contractile phenotype through suppression of proliferation.
Overall, the available evidence suggests strongly that neurons are effective regulators of intestinal smooth muscle growth and the maintenance of their differentiated state. This emphasizes the importance of increasing our knowledge of the effects of both acute and chronic inflammation of the intestine on its innervation.
This work was supported by the Canadian Association of Gastroenterology and the Medical Research Council of Canada.
Address for reprint requests and other correspondence: M. G. Blennerhassett, Gastrointestinal Diseases Research Unit, Queens Univ., Hotel Dieu Hospital, 166 Brock St., Kingston, ON K7L 5G2, Canada.
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. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society