Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The actions of opium on the intestinal tract have been known for centuries, for it has long been used to alleviate dysentery and diarrhea. In addition, the effects of opiate analgesic drugs are often limited by their pronounced constipating action (Mancini and Bruera, 1998). These effects stem in part from the ability of opiates to reduce intestinal transit by reducing segmenting contractions of the circular smooth muscle (Kromer, 1988). They are mediated by opioid receptors (ORs) in the central and enteric nervous systems. In myenteric neurons that modulate intestinal motor function, inhibitory ORs are associated with neuronal hyperpolarization and diminished neurotransmitter release. Depending on the species and intestinal segment, ORs may be expressed on excitatory or inhibitory neurons, and therefore mediate circular muscle relaxation or tonic segmentation (Burks, 1995).

Isolated intestinal preparations containing intact myenteric ganglia, particularly that of the longitudinal muscle-myenteric plexus (LMMP) from the guinea pig ileum, have played a key role in the discovery and characterization of ORs and their endogenous ligands (Lord et al., 1977). The guinea pig ileal LMMP, which predominately expresses µ-ORs and to a lesser extent, -ORs, has become a standard bioassay preparation for the characterization of OR ligands. Opioid agonists inhibit acetylcholine release from myenteric motor neurons and attenuate twitch contractions of the longitudinal muscle in response to transmural electrical stimulation (Burks, 1995). However, myenteric -ORs predominately modulate intestinal smooth muscle contractility in the mouse, rat, cat, dog, baboon, and human small intestine (De Luca and Coupar, 1996).

For purposes of comparison with a ₁-like OR, which we have identified in the submucosa of porcine ileum (Quito and Brown, 1991), we characterized the -OR and other OR types that are expressed in the myenteric plexus of this intestinal segment. The porcine ileal myenteric plexus constitutes an enteric neural network morphologically similar to that in the human small intestine (Timmermans et al., 1992). Enkephalin-immunoreactive neurons are abundant in myenteric ganglia of the porcine small intestine and myenteric nerve fibers appear to project to the circular smooth muscle (Porcher et al., 2000). In addition, the endogenous opioid peptides dynorphin and leumorphin, which are capable of interacting with -OR, are present at high levels in the porcine small intestine (Tachibana et al., 1982; Suda et al., 1984). Finally, mRNA transcripts for the porcine -OR have been detected in full-thickness and smooth muscle-myenteric plexus preparations of porcine small intestine, and -OR-immunoreactive myenteric neurons appear to innervate the adjacent circular muscle layer of the porcine ileum (Brown et al., 1998). We were particularly interested in examining the possible interrelationships between ORs in the myenteric plexus, in light of the electrophysiological results of Egan and North (1981) demonstrating the coexistence of - and µ-ORs in myenteric neurons of the guinea pig ileum and the recent discovery of heterodimerization in recombinantly expressed ORs (Jordan and Devi, 1999).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drugs and Reagents. [D-Ala²,N-Me-Phe⁴,Gly⁵-ol]-enkephalin (DAMGO); [D-Pen²,D-Pen⁵]-enkephalin (DPDPE); [D-Ala²,D-Leu⁵]-enkephalin (DADLE); [D-Ser²,Leu⁵]-enkephalin-Thr (DSLET); and [D-Ala²]-deltorphin II (deltorphin II) were obtained from Peninsula Laboratories, Inc. (Belmont, CA). (+)-4-[(R)-((2S,5R)-4-Allyl-2,5-dimethyl-1-piperazinyl-3-methoxybenzyl]-N,N-diethylbenzamide (SNC80) was purchased from Tocris Cookson (Ballwin, MO). trans-(±)-3,4-Dichloro-N-methyl-N-(2-[1-pyrrolidinyl]cyclohexyl)benzeneacetamide methanesulfonate (U-50,488H) and (+)-(5,7, 8)-N-methyl-N-[7-(1-pyrrolidinyl)-1-oxaspiro[4.5]dec-8-yl]-benzeneacetamide (U-69,593) were obtained from Research Biochemicals International (Natick, MA). Norbinaltorphimine (norBNI), naltriben (NTB), and 7-benzylidenenaltrexone (BNTX) were synthesized in the laboratory of P. S. Portoghese (Takemori and Portoghese, 1992). Carbamylcholine chloride (carbachol) and other drugs and chemicals were obtained from Sigma Chemical Co. (St. Louis, MO).

Animals. Intestinal tissues were obtained from Yorkshire pigs (6-10 weeks of age; 10-18 kg of body weight) of each sex that were not fasted before sacrifice. Animals were sedated with an intramuscular injection of tiletamine hydrochloride-zolazepam (Telazol, 8 mg/kg; Fort Dodge Laboratories, Fort Dodge, IA), in combination with xylazine (8 mg/kg). The animals were subsequently euthanized by barbiturate overdose in accordance with approved University of Minnesota Animal Care Committee protocols. A midline laparotomy was performed to expose the intestine and a portion of the ileum, identified by its attachment to the ileo-cecal ligament, was removed and placed in an oxygenated physiological salt solution approximating the composition of porcine extracellular fluid (118 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 0.5 mM MgCl₂, 25 mM NaHCO₃, 1.0 mM NaH₂PO₄, and 11 mM D-glucose; pH 7.4).

Measurement of Smooth Muscle Contractility. Ileal segments were cut longitudinally along the antimesenteric border and tissues were placed in ice-cold, oxygenated physiological salt solution. Ileal segments were pinned out as a flap, with the mucosa uppermost. Both the mucosa and submucosa were removed and a 3 × 10-mm muscle strip was cut parallel to the circular muscle layer around the entire circumference of the ileum. The strip therefore contained the circular muscle, from which isometric recordings were made, as well as the myenteric plexus and the longitudinal smooth muscle.

Immunohistochemistry. Ileal strips from five pigs used in the pharmacological experiments described above were cut in blocks of 1 to 2 cm² and immersed in ice-cold 2% paraformaldehyde in phosphate-buffered saline (PBS) at pH 7.4 for 2 h. The tissues were then cryoprotected in graded (10-30%) concentrations of sucrose in PBS, embedded in TissueTek O.C.T. compound (Baxter Healthcare Corp., McGaw Park, IL), and frozen. Longitudinal or transverse cryostat sections (14 µm in thickness) were thaw-mounted onto Superfrost-plus slides (Fisher Scientific, Pittsburgh, PA) and stored at 20°C until use. Tissues were rehydrated in PBS for 15 min and preincubated in PBS containing 0.4% Triton X-100 (Sigma Chemical Co.) and 3% normal donkey serum (Jackson ImmunoResearch Laboratories, West Grove, PA) for 30 min at room temperature to block nonspecific binding. Sections were incubated overnight at 4°C with one or more of primary anti-OR antibodies at 1:1000 dilutions referenced in Table 1. In some cases, an antibody against the neuronal marker protein gene product 9.5 (1:150 dilution; Chemicon International Inc., Temecula, CA) raised in rabbits was used in adjacent sections to confirm neuronal morphology. All antibodies were diluted in 0.4% Triton X-100 and 3% normal donkey serum. Sections were washed in PBS for 15 min, and then incubated with appropriate secondary antibodies (donkey anti-rabbit indocarbocyanine 3-conjugated IgG at 1:400 dilution or donkey anti-goat fluorescein isothiocyanate-conjugated IgG at 1:40 dilution) in PBS for 1 h in the dark. Sections were subsequently washed in PBS for 15 min, coverslipped with Vectashield (Vector Laboratory, Burlingame, CA), and stored at 20°C. Control experiments included the omission of primary antibodies from the staining protocol or the preabsorption of primary antibodies with their relevant blocking peptides in 100-fold molar excess.

Data Analysis. Changes in the average peak amplitudes of EFS-induced circular muscle contractions were expressed as a mean percentage change relative to predrug responses to EFS. Determinations of agonist concentration-effect relationships through nonlinear regression methods and statistical analyses of data were performed using the PRISM computer software program (version 2.0; GraphPad Software, Inc., San Diego, CA). Antagonist equilibrium constants (K_e) were calculated according to the method of Kosterlitz and Watt (1968). Agonist potencies are expressed as the negative logarithm of the 50% inhibitory concentration (pIC₅₀) and this parameter was used in all statistical comparisons of agonist potency. Comparisons between a single control and treatment mean were made by paired or unpaired two-tailed Student's t tests when appropriate. Comparisons of a control mean with multiple treatment means were made by analysis of variance followed by Dunnett's test. In all cases, the limit for statistical significance was set at P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Electrically Evoked Mechanical Responses of Smooth Muscle Strips. The delivery of EFS evoked a biphasic contractile response in smooth muscle-myenteric plexus strips oriented in the plane of the circular muscle. This included an initial contraction during EFS delivery, which will hitherto be termed an ON response, and a rebound contraction following stimulus cessation, referred to as an OFF response. In a representative group of 36 tissues, EFS produced ON and OFF contractions of 28.4 ± 3.9 and 40.2 ± 4.0 mN tension. Norepinephrine reduced EFS-evoked ON and OFF contractions by 93.6 ± 4.9 and 97.5 ± 2.0%, respectively, at a concentration of 1 µM (P < 0.05, paired t test, n = 3 tissues from three pigs).

Characterization of Opioid Receptors Modulating EFS-Evoked ON Contractions. The -OR agonists U-50,488H and U-69,593 were the most effective substances tested in decreasing ON contractions (Figs. 1 and 2, top).

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Representative chart records depicting the concentration-related effects of DAMGO (top), DSLET (middle) and U-69,593 (bottom) on electrically evoked contractions. Each contraction represents the first contractile response to electrical field stimulation in ileal strips following 3 min of agonist exposure at the concentration (in nM) indicated. Preagonist control responses of the tissues are indicated as zero. Reference axes indicate force generation (in millinewtons) versus time (in min). Data are representative of responses measured in 4 to 15 tissues from 4 to 14 pigs.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Top, concentration-effect relationships of µ-, -, and -opioid agonists for inhibition of electrically induced ON contractions. Middle, antagonism of the suppressant action of U-69,593 on EFS-evoked ON responses by norBNI and NTB. Bottom, antagonism of the suppressant action of SNC80 on EFS-evoked ON responses by BNTX, norBNI, or naltriben. Each point represents the mean ± S.E.M. of responses in 3 to 15 tissues from 3 to 14 pigs.

Characterization of Opioid Receptors Modulating EFS-Evoked OFF Contractions. Peptide-based agonists interacting with ₂-OR, particularly deltorphin II and DSLET, were more potent than the nonpeptidic -OR agonist SNC80 (P < 0.05 versus deltorphin II pIC₅₀ value, Dunnett's t test) in inhibiting OFF contractions. These three -OR agonists were also more effective than the peptidic ₁-OR agonists DPDPE and DADLE or agonists interacting primarily with -OR or µ-OR (Fig. 3, top left; Table 2). The -OR agonists exhibited a rank order of potency of deltorphin II > DSLET  DADLE > SNC80  DPDPE and produced maximal reductions in OFF contraction amplitudes with a rank order of DSLET  deltorphin II  SNC80 > DADLE  DPDPE (Table 2). Like ₁-OR agonists, agonists selective for either - or µ-ORs were considerably less effective in inhibiting OFF contractions than the ₂-OR agonists (Fig. 3, top left; Table 2).

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Top, left, concentration-effect relationships of µ-, -, and -opioid agonists for inhibition of electrically induced OFF contractions. Each point represents the mean ± S.E.M of responses in 4 to 19 tissues from 4 to 16 pigs. Top, right, antagonism of the suppressant action of DSLET on EFS-evoked OFF responses by BNTX, NTB, or norBNI. Bottom, left, antagonism of the suppressant action of SNC80 on EFS-evoked OFF responses by BNTX, NTB, or norBNI. Bottom, right, antagonism of the suppressant action of U-69,593 on EFS-evoked OFF responses by NTB or norBNI. Tissues were pretreated with each antagonist at the concentrations indicated for 10 min before the determination of agonist concentration-effect relationships. The antagonist equilibrium constants (Ke) for each agonist-antagonist combination that produced a significant change in agonist potency appear in Table 3. Each point represents the mean ± S.E.M. of responses determined in three to six tissues from three to five pigs.

Effects of DSLET and U-69,593 on Carbachol-Evoked Smooth Muscle Contractions. To confirm the neuronal site of opioid action, the abilities of - and -OR agonists to alter smooth muscle contractions elicited by the cholinergic agonist carbachol were examined. At the relatively high concentration of 0.1 µM, neither DSLET nor U-69,593 altered the potency or effectiveness of carbachol to produce contractions in ileal muscle strips (Fig. 4).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Contractile effects of carbachol in ileal muscle strips alone and after pretreatment with 0.1 µM U-69,593 (left) or DSLET (right). Each point represents the mean percentage ± S.E.M. of the maximum response to carbachol alone in 3 (U-69,593) and 6 (DSLET) tissues from three to four pigs. Maximal responses to carbachol alone averaged 143.4 ± 20.0 mN in nine tissues from seven pigs.

Expression and Colocalization of Opioid Receptor-Like Immunoreactivities in Myenteric Plexus. Using a primary antibody directed against an identical peptide sequence in the predicted second extracellular loop of murine and porcine -OR (Arvidsson et al., 1995; Brown et al., 1998), intense -OR-like immunoreactivity was observed in myenteric neurons contained in smooth muscle-myenteric plexus strips from the porcine ileum (Fig. 5, A, D, and G). Ganglia containing several large immunoreactive neurons as well as smaller neurons could be observed; in addition, numerous -OR immunoreactive fibers were seen within the myenteric plexus and smooth muscle layers. All neural elements coexpressed immunoreactivity toward protein gene product 9.5 (data not shown). In contrast, smooth muscle cells did not appear to express -OR immunoreactivity.

View larger version (49K): [in this window] [in a new window]   Fig. 5.   Digitized confocal micrographs of opioid receptor (OR)-like immunoreactivities in neurons and nerve fibers in the myenteric plexus and smooth muscle coats of porcine ileum. A and D, localization of -OR-like immunoreactivity (red) in myenteric plexus (MP) and nerve fibers coursing through circular muscle (CM; long arrow). Arrowhead in D shows a representative myenteric neuron and short arrow points to a nerve fiber in myenteric plexus. Photomicrograph in G represents area encompassed by box in D under higher magnification to exhibit myenteric neurons. B and C, identical section as seen in A coincubated with anti-µ-OR antibody (green, B). Note lack of µ-OR-like immunoreactivity and no colocalization (yellow, C) of - and µ-OR immunoreactivities. E and F, identical section as seen in D coincubated with anti--OR antibody (green, B). Note colocalization (yellow, F) of - and -OR immunoreactivities. H and I, photomicrographs represent areas encompassed by boxes in E and F under higher magnification to exhibit myenteric neurons. Micrographs are representative of observations in longitudinal sections of ilea taken from three to five pigs. Scale bar, A-C, 150 µm; D-F, 115 µm; and G-I, 43 µm.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Low-frequency, intermittent electrical field stimulation of porcine ileal smooth muscle-myenteric plexus strips produced ON and OFF contractions as has been reported previously in several other intestinal preparations in vitro, including that of the human colon (Chamouard et al., 1993, 1994). The amplitudes of both contractions are reduced by >90% in the presence of the neuronal Na⁺ channel blocker tetrodotoxin, but OFF responses are more sensitive to inhibition by the muscarinic cholinergic antagonist atropine and the neurokinin-1 receptor blocker CP-96,345-1 than are ON contractions (Brown et al., 1998). Thus, both contractile responses to EFS appear to be neurogenic and mediated by muscarinic cholinergic receptors. However, OFF responses are also mediated by neurokinin-1 receptors, whereas ON responses are associated with noncholinergic, nonadrenergic neurotransmitters that have not yet been identified. In the present study, -OR agonists most effectively inhibited ON contractions and -OR agonists attenuated OFF responses. It is likely that the OR agonists interacted with ORs present on myenteric neurons or nerve fibers rather than those previously described on rabbit circular smooth muscle cells (Kuemmerle and Makhlouf, 1992), because neither the -OR agonist U-69,593 nor the -OR agonist DSLET affected carbachol-induced myogenic contractions. Furthermore, smooth muscle cells did not display immunoreactivity toward µ-, -, or -ORs. In rat intestine, immunoreactivities to - or µ-ORs similarly were not associated with smooth muscle cells (Fickel et al., 1997). In contrast to opioids, norepinephrine inhibited both ON and OFF contractions by a similar magnitude. Adrenergic receptors in ileal strips mediating the effects of norepinephrine, a major inhibitory neurotransmitter substance in the enteric nervous system, may differ from ORs in their cellular distribution, coupling to effectors, and physiological role in gut motility.

The -OR agonists U-50,488H and U-69,593 inhibited EFS-evoked ON contractions to a greater extent than - or µ-OR agonists, but were relatively ineffective in inhibiting OFF responses. U-69,593 actions were competitively inhibited by the selective -OR antagonist norBNI with a K_e of 4.2 nM, but were unaffected by the -OR antagonist NTB. The K_e value of norBNI calculated from the present results is similar to its K_i value (3.5 nM) in displacing [³H]naloxone benzoylhydrazone from U-50,488H-preferring ₁-OR binding sites in guinea pig cerebellum (Clark et al., 1989). These results strongly suggest that -ORs modulate the ON response to EFS in porcine ileal strips. Although it was less effective than either -OR agonist, the -OR agonist SNC80 was nearly twice as effective as the peptidic -OR agonists tested in inhibiting ON contractions. Neither norBNI nor the putative ₁-OR antagonist BNTX significantly affected SNC80 potency. This result indicates that -OR, and specifically those of the putative ₂-subtype (Zaki et al., 1996), may additionally modulate ON responses to EFS. OFF contractions occurring after EFS cessation were most effectively inhibited by DSLET, deltorphin II, and SNC80, which act as agonists at putative ₂-ORs. The putative ₁-OR agonists DPDPE and DADLE were partially effective, as were - and µ-OR agonists. In comparison to their effects on ON contractions, both -OR agonists were >10-fold less potent and >2-fold less effective in inhibiting OFF contractions. At equimolar concentrations, the ₂-OR antagonist NTB, but not the ₁-OR antagonist BNTX, reduced the potencies of DSLET and SNC80 in inhibiting OFF contractions. Differences in agonist activities on ON and OFF contractions might arise from agonist binding to different sites on -OR (Quock et al., 1999), and in terms of NTB antagonism, may be attributable to the relative receptor reserve for each agonist in this preparation. The mechanisms by which peptide and nonpeptide -OR agonists function in this preparation clearly require further examination.

The inhibitory actions of DSLET and SNC80 on OFF contractions were also reduced by norBNI. The -OR antagonist produced dextral shifts in the concentration-effect curves of these -OR agonists that were equal to or greater than those produced by an equimolar concentration of NTB. norBNI decreased the respective potencies of U-69,593 and DSLET in inhibiting ON and OFF responses by an equivalent magnitude. It has been reported previously that norBNI is >2 orders of magnitude more potent at -OR than at -OR in standard functional bioassays and >150-fold more selective in binding assays (Takemori and Portoghese, 1992). It is possible that the concentration of norBNI used (100 nM) was outside of its selectivity window for -ORs and the drug was blocking -ORs.

The interaction between -OR agonists and a -OR antagonist was further examined in immunohistochemical experiments with anti-OR primary antisera. An antibody raised against an identical epitope in the second extracellular loops of murine and porcine -OR (Table 1) detected specific -OR-like immunoreactivity in myenteric neurons and in nerve fibers within the myenteric plexus and circular smooth muscle. This distribution pattern is similar to that previously reported with an antibody raised against the N terminus of the cloned murine -OR (Brown et al., 1998). Immunoreactivity to -OR was also localized in myenteric neurons but was present in fewer nerve fibers; this finding is in general agreement with the localization of -OR immunoreactivity in the rat proximal colon (Fickel et al., 1997). A subset of myenteric neurons coexpressed both -OR and -OR immunoreactivities. To our knowledge, this is the first example of such OR colocalization in the peripheral nervous system, and provides additional evidence supporting an association between these two receptors. Heterodimeric /-ORs and homodimeric -ORs have been recombinantly expressed, albeit at high receptor densities, in cultured cell lines (Cvejic and Devi, 1997; Jordan and Devi, 1999). The interaction between the selective -OR agonists and a -OR antagonist could be explained by invoking the existence of allosterically coupled /-OR dimers in the porcine ileum. The "address" residue for norBNI binding to the -OR is Glu²⁹⁷, which resides at the top of transmembrane (TM) domain VI (Horth et al., 1995). Substitution of a glutamate residue in an equivalent position of the µ- or -OR enhances the affinity of these mutant receptors for norBNI (Metzger et al., 2001). If an interlocking receptor dimer model with a TM V/VI interface is used (Gouldson et al., 1998), then one of the "hybrid" seven TM bundles in the /-OR heterodimer will contain the Glu²⁹⁷ residue and should therefore recognize norBNI (Fig. 6A). The second hybrid bundle, containing residues on outer loop 3 of -OR, may be responsible for the recognition of the -OR agonists SNC80 and DSLET because a substantial decrease in -OR-selective ligand affinity is associated with point mutations in this loop (Valiquette et al., 1996). Contact dimers between - and -ORs also are possible (Fig. 6B; Gouldson et al., 1998). In either receptor dimer model, antagonism could be mediated by norBNI binding to the seven TM bundle that contains Glu²⁹⁷; this event would allosterically alter the conformation of the associated seven TM bundle (7TM-B or ) that interacts with -OR agonists (Fig. 6). These models are conceptually identical to that proposed for allosterically coupled agonist and antagonist sites in µ-OR dimers (Portoghese and Takemori, 1983) and provide a mechanism for the antagonism of -OR agonists by the prototypical -antagonist norBNI. Clearly, additional experiments will be required to establish whether this phenomenon is attributable to a heterodimeric receptor, different neural circuits mediating the contractile responses to electrical stimulation, differential distribution of ORs on myenteric neural elements, or another mechanism.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Schematic illustration of an extracellular view of a proposed /-OR heterodimer containing an interface at TM helices V and VI. Possible motifs include the interlocking dimer (A) and the contact dimer (B) models. In A, the filled circles represent TM helices of the -OR, whereas open circles indicate TM helices of the -OR. The seven TM-A hybrid bundle (A) or the -OR monomer component (B) contains Glu (E297), which is the key residue determinant for norbinaltorphimine binding, and the third outer loop (OL3) of TM-B or the -OR monomer component confers selectivity for -OR agonists (SNC80 and DSLET). Antagonism of -OR agonism mediated by OL3 would occur when norBNI binds to E297 of the seven TM-A bundle or -OR monomer component.

The porcine ileal myenteric plexus, unlike the guinea pig or rat LMMP, does not appear to express µ-ORs. Morphine, the antidiarrheal agent loperamide, and the highly selective µ-OR agonist DAMGO were the three drugs with the weakest inhibitory actions on contractile responses to EFS. Therefore, µ-OR protein may not be expressed or is expressed in low abundance in this intestinal subregion. Indeed, immunoreactivity to µ-OR, unlike that for -OR and -OR, could not be detected in myenteric neurons and fibers. It is conceivable that the primary antibody used for these studies, raised against an N-terminal epitope in human µ-OR, may not have detected the homologous peptide sequence in porcine µ-OR, which displays 70% sequence identity with human µ-OR (Table 1). Immunoreactivity to µ-OR was also absent in the neural elements of the porcine ileal submucosa, although this antibody recognized specific receptor-like immunoreactivity in sections of the porcine hypothalamus and guinea pig ileum (Poonyachoti et al., 2001). Thus, the present tissue preparation appears to distinguish, in the absence of µ-ORs, opioid activities mediated by - or -ORs independently and possibly in functional association.