Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Abnormal smooth muscle function is often associated with IBD and may lead to diarrhea, the major clinical symptom (Snape et al., 1980; Rao, 1997). In vitro studies of smooth muscle of IBD patients suggest a defect in smooth muscle and/or neural function (Snape et al., 1980; Koch et al., 1988; Vermillion et al., 1993). Enteric nerves in Crohn's patients show evidence of both degeneration and subsequent proliferation (Dvorak et al., 1980), suggesting a possible remodeling or reorganization of neural control of smooth muscle during inflammation. This remodeling may underlie the observation that motility varies with the stage of disease activity (Koch et al., 1988; Hosseini et al., 1999). We showed previously that there are progressive alterations in smooth muscle contractility in vitro in the transition of TNBS-induced colitis from acute to chronic due, in part, to alterations in the cholinergic and in the excitatory and inhibitory nonadrenergic/noncholinergic neural control of smooth muscle function (Hosseini et al., 1999). More recently we demonstrated a TNBS-induced inflammation also results in a transient loss of nitrergic regulation (Bossone et al., 2001). There is less information on the response of adrenergic nerves to gut inflammation.

Stimulation of sympathetic nerves is generally inhibitory to gut function primarily by modulating prejunctional release of acetylcholine and less frequently by acting at postsynaptic receptors on smooth muscle (Tack and Wood, 1992). However, the reports of abundant sympathetic fibers terminating in varicosities near myenteric ganglia adjacent to the longitudinal and circular smooth muscle layers (Gershon, 1967; Furness et al., 1990) indicate a potential contribution of postjunctional mechanisms as well. Early pharmacological studies demonstrated the presence of two populations of adrenergic receptors,  and  (McIntype and Thompson, 1992; De Ponti et al., 1996). Additional studies further divided the adrenoreceptors into the subtypes ₁ and ₂, located primarily on nerves, and ₁ and ₂ on smooth muscle (Bulbring and Tomita, 1987). The discovery of a third  receptor subtype, the ₃ receptor, in the mid-1980s prompted a reexamination of the contribution of the  adrenoreceptors to gut motility. The ₃ receptor is localized and expressed on adipose tissue (Strosberg, 1997) and is also present on gut smooth muscle in a number of species, including rat and human (Kelly and Houston, 1996; Thollander et al., 1996; Bardou et al., 1998; De Ponti et al., 1999; MacDonald and Watt, 1999). This receptor is referred to as the atypical  adrenoreceptor because of its resistance to propranolol (Strosberg, 1997). Recent studies have proposed ₃ receptor agonists as a therapeutic target in IBD patients.

In the present study, we examined the ability of the physiological ligand norepinephrine as well as adrenoreceptor agonists and antagonists to modify spontaneous contractions, to prevent electrical field stimulation (EFS)-induced off contractions, and to relax muscle precontracted with carbachol. Using a rat model of TNBS-induced colitis, we next determined the inflammation-induced alterations in smooth muscle function. The aims of this study were to determine colitis-induced alterations in 1) sympathetic control of circular smooth muscle as inflammation progresses from acute to chronic, and 2) adrenergic receptor subtypes involved.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animal Model. Male Sprague-Dawley rats weighing 150 to 250 g were anesthetized with xylazine (20 mg/kg) and ketamine (100 mg/kg). Colitis was induced by intrarectal administration of TNBS (100 mg/kg in 50% ethanol). Control animals received saline. After either 4 h (acute) or 7 days (chronic) the rats were reanesthetized for surgical removal of the entire distal colon. Tissue was taken for in vitro measurements of isometric contraction.

Histology. Sections of distal colon were pinned mucosal side up, fixed in 10% neutral-buffered formalin. Tissues were embedded in paraffin and sections (6 µm) were stained with hematoxylin and eosin or with Giemsa for light microscopic evaluation of tissue architecture and differentiation of leukocytes, respectively. An ocular micrometer was used to measure the total thickness of the colon, the muscularis externa, and the circular smooth muscle layer. The ratio of circular muscle to the total muscularis externa was used to correct the muscle mass in the calculation of force per cross-sectional area.

In Vitro Contractility. The distal colon was opened along the mesenteric border, pinned mucosa side up in oxygenated (95% O₂, 5% CO₂) Krebs' solution, and the mucosa carefully dissected away from the muscularis externa. Muscle strips (1.0 × 0.4 cm) were cut along the circular axis, mounted in individual 8-ml organ baths, and maintained in oxygenated Krebs' solution at 37°C. One end of the tissue was attached to an isometric tension transducer (model FT03; Grass Instruments, Quincy, MA) and the other to the bottom of the bath. The bath solution was replaced every 10 min throughout each study. Tension was recorded using a Grass model 79 polygraph (Grass Instruments) and expressed as force per cross-sectional area (Hosseini et al., 1999).

Reagents. Krebs' buffer contained 118.5 mM NaCl, 4.75 mM KCl, 2.54 mM CaCl₂, 1.19 mM MgSO₄, 25 mM NaHCO₃, 1.19 mM NaH₂PO₄, and 11.0 mM glucose. Stock solutions (1 mM) of carbachol, phentolamine, propranolol, norepinephrine, prazosin, yohimbine, CGP20712A, ICI118551, dobutamine, clenbuterol, and BRL37344 were prepared and stored at 4°C and diluted in water on the day of the experiment. Cyanopindolol was dissolved in dimethyl sulfoxide (1 mM). Atropine, tetrodotoxin, L-NNA, carbachol, phentolamine, propranolol, norepinephrine, prazosin, yohimbine, CGP20712A, ICI118551, dobutamine, clenbuterol, and BRL37344 were purchased from Sigma Chemical (St. Louis, MO), and cyanopindolol was obtained from Tocris Cookson (Ballwin, MO).

Statistics. Concentration- and frequency-response curves were compared among control, acute, and chronic inflamed groups, by using a multivariate analysis of variance (SYSTAT software; SYSTAT, Inc., Evanston, IL) with repeated measures followed by post hoc testing for comparisons of multiple means. All other data were assessed using a one-way analysis of variance followed by post hoc testing with the Bonferroni correction (GraphPad Software, San Diego, CA) where appropriate. All results are expressed as means ± S.E. where applicable. The n represents the number of animals. Carbachol and EFS responses were fitted to sigmoid curves (GraphPad Software). EC₅₀ and 50% of the maximum response to EFS (EF₅₀) values (with 95% confidence limits), respectively, were determined from these curves.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Histology

Giemsa-stained sections of distal colon taken 4 h after treatment with TNBS/ethanol (acute) showed microscopic changes similar to those reported previously in this model (Hosseini et al., 1999). In acute inflammation, damage was confined to the mucosa and was characterized by neutrophil infiltration and epithelial cell detachment with mild-to-moderate vasocongestion with occasional areas of exposed lamina propria. At 7 days post-treatment with TNBS/ethanol (chronic), the mucosa showed marked edema and pronounced lymphocytic infiltration in the lamina propria, submucosa, and mucosa. There were areas of intact epithelia but more frequent areas of denuded mucosa with prominent thickening of the muscularis externa.

Smooth Muscle

EFS: Adrenergic Antagonists. As described previously in rabbit colon (Snape et al., 1989), in response to EFS, circular smooth muscle showed a small or no contraction during the stimulus (on response), but exhibited frequency-dependent contractions at the termination of the stimulus (off response) to EFS that were maximal at 5 Hz. Therefore, in the following experiments, responses to EFS are referred to as the "off" response unless specified. Both the "on" and off responses were blocked by TTX, indicating their neural dependence. The on response was abolished by atropine, demonstrating that it is mediated by cholinergic nerves. The off response was only partially reduced by atropine (~20%), suggesting it is primarily noncholinergic.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Frequency-dependent responses to EFS (1-20 Hz, 80 V, 1 ms) of colonic circular smooth muscle in control (A), acute colitis (B), and chronic colitis (C). Tissues were incubated with vehicle, L-NNA (10 µ M, nitric-oxide synthase inhibitor), phentolamine (10 µ M,  adrenergic antagonist), propranolol (1 µ M,  adrenergic antagonist), or P+P and incubated for 20 min before EFS. *p < 0.05 for the entire curve compared with vehicle (n  5 for each group).

EFS: Adrenergic Agonists. The  adrenoreceptor appears to be dominant in the adrenergic regulation of EFS-induced contractions. To determine the inflammation-induced alterations of specific  adrenoreceptors, EFS (5 Hz) off contractions were assessed in the presence of specific agonists (Table 1). In control rats, norepinephrine (Fig. 2A) and BRL37344 (Fig. 2B) abolished EFS responses (Table 1). In contrast, the ₁ and ₂ receptor agonists dobutamine (Fig. 2C) and clenbuterol (Fig. 2D), respectively, had little effect on EFS contractions.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effects of 1 µM norepinephrine (A) or the specific adrenergic agonists 0.1 µM BRL37344 (B, 3 agonist), 1 µM dobutamine (C, 1 agonist); or 1 µM clenbuterol (D, 2 agonist) on the response to EFS (5 Hz, 80 V, 1 ms) in control smooth muscle strips. Left, response in vehicle; right, the response in the presence of the various agonists. Tissues were incubated with vehicle or agonists for 10 min before EFS. Each panel is representative of at least four independent experiments.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Colitis-induced changes in 3 adrenergic control of smooth muscle. Each panel is representative of at least four independent experiments. A, responses to EFS. Tissues were treated with 0.1 µM BRL37344 and subjected to EFS (5 Hz, 80 V, 1 ms). B, spontaneous contractions. Tissues were equilibrated in oxygenated Krebs' solution for 0.5 to 1 h and then 0.1 µM BRL37344 was added (arrow). The response to norepinephrine (NE; arrow) is shown in the inset.

Spontaneous Contractions. To determine the effect of colitis on  adrenoreceptor regulation of spontaneous colonic contractions, strips were incubated in the presence and absence of  adrenergic agonists or antagonist. In control strips, spontaneous contractions were suppressed completely by 1 µM norepinephrine (Fig. 3B, inset), by the ₂ adrenergic agonist clenbuterol (data not shown), and by 0.1 µM ₃ adrenergic agonist BRL37344 (Fig. 3B). The effect of BRL37344 was observed in the presence of TTX, indicating that ₃ action is primarily on smooth muscle as reported previously (Anthony et al., 1998; Luckensmeyer and Keast, 1998). Only the ₃ antagonist cyanopindolol increased the amplitude of spontaneous contractions more than 2-fold (238 ± 47% of vehicle). There was little or no effect of individual selective antagonists of the ₁, ₂ adrenoreceptors on spontaneous contractions (data not shown).

Carbachol Precontractions. The effects of norepinephrine and BRL37344 on smooth muscle function were evaluated further by assessing inflammation-induced alterations in their ability to relax carbachol-induced contractions. Maximum responses to carbachol were similar in controls (26,606 ± 471 mN/cm²), acute (23,853 ± 2,658 mN/cm²), and chronic (26,438 ± 3,218 mN/cm²) colitis. However, compared with controls (EC₅₀ = 33 nM) there was more than a 10-fold decrease in the sensitivity of the response in chronic (EC₅₀ = 552 nM), but not in acute inflammation (EC₅₀ = 24 nM). In control colon, norepinephrine (1 µM; Fig. 4A, inset) and BRL37344 (Fig. 4A) relaxed carbachol (1 µM)-induced contractions by 100 ± 1 and 89 ± 4%, respectively. The ability of BRL37344 to relax the carbachol-induced contractions was blocked completely by cyanopindolol. In inflamed colon, a greater concentration of BRL37344 was required to relax carbachol-induced contractions (Fig. 4B). The EC₅₀ values for relaxation were 1.3 and 1.0 µM in acute and chronic colitis, respectively, and were nearly 10-fold higher than in controls (0.16 µM). The maximum relaxation was 59 ± 5% in acute and 65 ± 6% in chronic colitis.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   BRL37344 produced a concentration-dependent relaxation of carbachol precontracted smooth muscle in control (saline), acute (4 h after TNBS), and chronic (7 days after TNBS) colitis. A, individual strips were contracted with 1 µM carbachol (added at ) and BRL37344 was added (arrows) at concentrations of 0.01 (a), 0.1 (b), 1 (c), 10 µM (d), followed by a wash (e). The response to norepinephrine (NE; arrow) is shown in the inset. Each panel is representative of at least four independent experiments. B, cumulative concentration-response curves showing percentage of relaxation by BRL37344 of carbachol precontraction. Values are expressed as the percentage (means ± S.E.) relaxation of the carbachol-induced contraction. *p < 0.05 compared with control (n  5 for each group).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Abnormal motility is induced by inflammation of the colon (Spriggs et al., 1951; Snape et al., 1980; Koch et al., 1988; Snape et al., 1988; Tomita et al., 1998) and may lead to the diarrhea characteristic of IBD. The decreased colonic motility observed generally in IBD patients may be attributed in part to the disturbed function of inhibitory nerves. We showed previously that TNBS-induced colitis produced a transient loss in nitrergic regulation of smooth muscle contractility (Bossone et al., 2001). The sympathetic nervous system, however, also constitutes an important portion of the inhibitory input to the gut. Patients with diabetes (Fedorak et al., 1985) or Crohn's disease exhibit a loss of the "sympathetic brake" (Lindgren et al., 1991), whereas altered vagal regulation, increased influence of inhibitory nerves (Tomita et al., 1998), and/or specific neurotransmitters such as vasoactive intestinal polypeptide (Koch et al., 1988) have been reported in ulcerative colitis. In the present study, we assessed changes in the contribution of adrenergic nerves to TNBS-induced alterations in circular smooth muscle contractility in rat distal colon. We found that TNBS-induced colitis in rats produced significant alterations in the  adrenergic control in addition to the changes in nitrergic regulation.

In general, stimulation of adrenergic nerves is inhibitory to nonsphincteric gut smooth muscle. Traditionally,  adrenergic control is linked to an inhibition of an excitatory cholinergic effect mediated primarily by the large number of sympathetic terminals in the myenteric plexus. This inhibition involves -input via ₁ modulation of noradrenaline released from sympathetic nerves and ₂ receptors on smooth muscle. After the identification of the atypical ₃ receptor in the gastrointestinal tract (Manara et al., 1996; MacDonald and Watt, 1999), studies showed that responses to isoprenaline in rat distal colon were mediated primarily via the ₃ receptor (Mckean and MacDonald, 1995). Although direct innervation of smooth muscle by sympathetic fibers is reportedly sparse, adrenergic receptors present on smooth muscle are sensitive to circulating catecholamines. In addition, immunohistochemical studies revealed that sympathetic fibers ramify in varicosities near myenteric ganglia and adjacent to the muscularis propria (Furness et al., 1990), suggesting a physiological role for postjunctional adrenergic modulation (Zhang et al., 1992; Spencer et al., 1999). In the present study, we investigated changes induced by inflammation of the rat distal colon in the adrenergic modulation of spontaneous contractions and responses to EFS. In this in vitro preparation, sympathetic modulation of responses to EFS is derived from the stimulation of the residual ends of extrinsic fibers with synapses on neurons within the myenteric plexus.

In control rats, responses to EFS were elevated significantly in the presence of propranolol alone, or in combination with phentolamine, but not by phentolamine alone, indicating that the principal adrenergic inhibition of distal colonic muscle is mediated by  rather than  receptors. The magnitude of the increase in responses to EFS in the presence of phentolamine and propranolol was similar to that in L-NNA, confirming the inhibitory effect of both nitrergic and adrenergic nerves. The lack of an effect of phentolamine on EFS responses in controls is consistent with the inability of ₁ or ₂ antagonists to alter off contractions in the present study. A specific ₃ antagonist is not yet commercially available, however, we used cyanopindolol, a  adrenergic antagonist that completely blocked the inhibitory effects of BRL37344 in this study and others (MacDonald and Watt, 1999). The enhanced response to EFS in the presence of cyanopindolol, but not the ₁ and ₂ antagonists, indicated that the dominant adrenergic receptor was ₃. Indeed, in control rats, low concentrations (0.1 µM) of the ₃ agonist BRL37344 abolished EFS-induced off contractions, whereas higher concentrations (1 µM) almost fully relaxed muscle strips precontracted with carbachol. These effects of the ₃ agonist were nearly identical to those of endogenous adrenergic agonist norepinephrine. These data are consistent with the finding that adrenergic inhibition of distal rat colon is mediated by postjunctional ₃ receptors on smooth muscle that are responsive to sympathetic nerve stimulation (Manara et al., 1996; Luckensmeyer and Keast, 1998).

The colon exhibits spontaneous contractions that are linked to the slow wave frequency that is intrinsic to smooth muscle. In healthy colon, the enteric nerves exert an inhibitory influence on the myogenic activity of the circular smooth muscle (Keef et al., 1993). Previous studies showed that nitrergic nerves exert a major control of spontaneous contractions in the colon in several species (Keef et al., 1993, 1997; Bossone et al., 2001). We showed that the amplitude of spontaneous contractions was enhanced in acute inflammation, an effect due, in part, to a transient loss of neural nitric oxide regulation (Bossone et al., 2001). In the colon, sympathetic nerves may also exert a tonic activity on smooth muscle function; however, there may be some species specificity. In cats, the pathway is reported to involve  adrenoreceptors, whereas in human and rats, the  receptor may be dominant (Manara et al., 1996). In the present study, the amplitude of spontaneous contractions was augmented significantly only in the presence of cyanopindolol, demonstrating a physiological role for ₃ receptors.

Colitis significantly altered the response to EFS. Off contractions were elevated significantly at all frequencies in acute, and at higher frequencies (>5 Hz) in chronic inflammation. The reduced sensitivity to EFS, as well as to carbachol, observed in strips taken from rats with chronic colitis can be attributed to the well documented hypertrophy of the muscularis propria in this group (Morris et al., 1989; Yamada et al., 1992; Hosseini et al., 1999). Inflammation produced significant alterations in the regulation of colonic smooth muscle by inhibitory nerves. Unlike controls, EFS off contractions in acute or chronic strips were not enhanced significantly by L-NNA or propranolol, indicating a reduction in nitrergic and  adrenergic regulation in colitis.

The ability of BRL37344 to inhibit contractions in responses to EFS, and of cyanopindolol to enhance EFS responses, were attenuated in colitis. Moreover, in colitis, there was a 10-fold increase in the concentration of the ₃ agonist required to relax precontracted strips. The maximal relaxations of BRL37344 on precontracted smooth muscle were also decreased in colitis. Finally, the ability of cyanopindolol to augment the amplitude of spontaneous contractions was lost in inflammation. It is of interest that although colitis attenuated the ability of BRL37344 to reduce EFS contractions, inflammation did not alter the ability of BRL37344 to abolish the myogenic spontaneous contractions. These data demonstrate that the loss of adrenergic control in inflammation involved the ₃ receptor; thus, these receptors may be a potential target for therapeutic intervention in IBD. Taken together, these data demonstrate that in both acute and chronic colitis, the enhanced response to EFS may be attributed to the persistent reduction in inhibitory nitrergic and  adrenergic regulation. This latter effect is not due to changes in ₁ or ₂ adrenoreceptors, but may be attributed to significant attenuation of ₃ receptor-mediated postjunctional inhibition, and/or reduced sensitivity or decreased number of ₃ receptors on smooth muscle.

In conclusion, the present study demonstrated that in control rats, the predominant adrenergic regulation of spontaneous contractions as well as contractions in response to nerve stimulation was mediated by  receptors, specifically, the ₃ adrenoreceptor. In the chronic stage of inflammation, there was a decrease in the sensitivity to nerve stimulation and pharmacological contractile agents attributed to the marked hypertrophy of smooth muscle that was not evident in the acute stage. Colitis significantly increased responses to EFS, an effect due to changes in both the nitrergic and ₃ adrenergic control of rat colonic circular smooth muscle. In both acute and chronic inflammation, there was a loss of adrenergic influence that was due primarily to a down-regulation in inhibitory ₃ adrenergic regulation. This loss of adrenergic regulation is similar to the loss of the "sympathetic brake" in Crohn's disease (Lindgren et al., 1991) and diabetes (Fedorak et al., 1985) and may contribute to the diarrhea in IBD.