Truth Seeker
New Member
Abstract
The effects of cannabinoid receptor agonists and antagonists on smooth muscle resting membrane potentials and on membrane potentials following electrical neuronal stimulation in a myenteric neuron/smooth muscle preparation of wild-type and cannabinoid receptor type 1 (CB1)-deficient mice were investigated in vitro. Double staining for CB1 and nitric oxide synthase (neuronal) was performed to identify the myenteric CB1-expressing neurons. Focal electrical stimulation of the myenteric plexus induced a fast (f) excitatory junction potential (EJP) followed by a fast and a slow (s) inhibitory junction potential (IJP). Treatment of wild-type mice with the endogenous CB1 receptor agonist anandamide reduced EJP while not affecting fIJP and sIJP. EJP was significantly higher in CB1-deficient mice than in wild-type littermate controls, and anandamide induced no effects in CB1-deficient mice. N-arachidonoyl ethanolamide (anandamide), R-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3,-de]- 1,4-benzoxazin-6-yl]-1-naphtalenylmethanone, a synthetic CB1 receptor agonist, nearly abolished EJP and significantly reduced the fIJP in wild-type mice. N-piperidino-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-pyrazole-caroxamide (SR141716A), a CB1-specific receptor antagonist, was able to reverse the agonist effects induced in wild-type mice. SR141716A, when given alone, significantly increased EJP in wild-type mice without affecting IJP in wild-type and EJP in CB1-deficient mice. Interestingly, SR141716A reduced fIJP in CB1-deficient mice. In the mouse colon, nitrergic myenteric neurons do not express CB1, implying that CB1 is expressed in cholinergic neurons, which is in line with the functional data. Finally, excitatory and inhibitory neurotransmission in the mouse colon is modulated by activation of CB1 receptors. The significant increase in EJP in CB1-deficient mice strongly suggests a physiological involvement of CB1 in excitatory cholinergic neurotransmission.
TO THE PRESENT, THE ENDOCANNABINOID system comprises two G protein-coupled receptors [cannabinoid receptor types 1 (CB1) and 2 (CB2)], three known endogenous agonists (anandamide, 2-arachidonoyl glycerol, nolandin ether), one endogenous CB1 antagonist/CB2 agonist (virodhamine), and enzymes degrading endocannabinoids [fatty acid amide hydrolase for anandamide and monoglycerol lipase for 2-arachidonoyl glycerol (10, 12, 21, 30)]. CB1 is present in the peripheral and central nervous system and is distinctly involved in numerous physiological effects, whereas CB2 is expressed mainly in immune cells. A general feature of CB1 activation is the reduction of the release of a variety of neurotransmitters (e.g., acetylcholine from enteric nerves) (16, 32). Whereas there is currently no evidence of a role of CB2 in the gastrointestinal tract, CB1 appears to be involved in several physiological processes in this tissue (9, 14, 28). The cannabinoid system is involved, e.g., in intestinal motility, gastrointestinal transit, gastric secretion, and colonic propulsion. In agreement with these observations, CB1 was shown to be localized in the gastrointestinal tract of many species including humans (1, 5, 6, 18, 19, 26, 29). In the gastrointestinal tract of pigs, which is functionally similar to that of humans, CB1 is present in neurons of the myenteric and submucosal plexus of the ileum and the colon as shown by immunohistochemistry. In these neurons, CB1 is coexpressed with cholinergic markers (18). In fact, cannabinoids, including anandamide and the synthetic agonist R-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3,-de]-1,4-benzoxazin-6-yl]-1-naphtalenylmethanone (WIN 55,212—2), reduce electrically induced contractions by activation of CB1 that is located presynaptically on enteric nerves (8, 13). Furthermore, in the rat gastric fundus, cannabinoids depress excitatory neurotransmission but also modulate inhibitory neurotransmission (34).
Apart from the effects of CB1 on small intestinal motility (4), little is know about cannabinoid action on colonic motility. Two studies (23, 29) described that CB1 on myenteric neurons is involved in the reduction of colonic propulsion in mice, supporting the notion that CB1 is involved in the regulation of colonic motility.
Given the scarce knowledge on the action of cannabinoids on neurotransmission in the gastrointestinal tract, the effects of cannabinoids on neurotransmission in the colon were evaluated in the present study by intracellular recordings using CB1-deficient mice (25) and the specific antagonist SR141716A (31).
MATERIALS AND METHODS
Tissue preparation for electrophysiological experiments. Eight- to ten-week-old female CB1-deficient mice or wild-type littermates in C57BL/6N background (25) were anesthetized by sodium phenobarbital and killed by cervical dislocation in accordance with the recommendations of the animal ethics committee at the Technical University of Munich. Colon was exposed through an abdominal midline incision. The complete large bowel was removed and placed into oxygenated Krebs solution I of the following composition (in mM): 120.35 NaCl, 5.9 KCl, 2.5 MgCl2, 1.2 NaH2PO4, 15.5, NaHCO3, 2.5 CaCl2, and 11.5 glucose (pH 7.4). Colon was opened along the mesenteric border, washed off from remaining fecal material, and pinned out in a Sylgard-lined (Dow Corning; Midland, MI) dissecting dish containing oxygenated Krebs solution I. Proximal colon was separated, and the mucosa and submucosa were removed, resulting in sheets of tissue consisting of circular and longitudinal muscle layers together with the attached myenteric plexus.
Intracellular electrical recording. Sheets or strips of muscle (proximal colon) were pinned using ∼150—200 micropins (15—25 μm in thickness), obtained from wolfram wire, to the Sylgard-based electrophysiological chamber, with the exposed circular muscle layer uppermost. The chamber was perfused (5 ml/min; Kwik Pump, World Precision Instruments; Sarasota, FL) with prewarmed (37°C) oxygenated (95% O2-5% CO2) Krebs solution I. Tissues were allowed to equilibrate for 90—120 min before the experiments were started. Nifedipine (1 μM) was present throughout all experiments. In experiments investigating sIJPs and fIJPs, atropine (1 μM) and guanethidine (1 μM) were present continuously. Capillary glass microelectrodes (borosilicate glass capillaries, 1.0-mm outer diameter × 0.58-mm inner diameter, Clark Electromedical Instruments; Edenbridge, UK) were made using a microelectrode puller (model P-97, 3-mm-wide filament, Sutter Instruments; Novato, CA) filled with KCl (3 M) and had resistances in the range of 80—120 MΩ. Neurons were stimulated (15 V, 0.3 ms, single pulse) via platinum electrodes arranged perpendicularly to the circular muscle layer, which were connected to a Grass S11 stimulator via a stimulus isolation unit (Grass SIU59, Grass Instruments; Quincy, MA). The responses were recorded against a "ground" Ag-AgCl electrode placed in the bath medium. Evoked electrical events were amplified (DUO 733 microelectrode amplifier, World Precision Instruments; Sarasota, FL) and digitalized with an analog-to-digital converter (SCB 68 interface, National Instruments; Austin, TX). Permanent recordings of membrane potentials were made on a personal computer running the LABVIEW 5.0 program (National Instruments; Austin, TX).
Tissue preparation for immunocytochemistry. Adult Swiss mice of both genders (n = 4) were killed using an overdose of pentobarbital sodium (Nembutal; intraperitoneally injected). One 10-mm segment of the proximal and distal colon was removed, rinsed in Krebs solution II (in mM: 117 NaCl, 5 KCl, 2.5 CaCl2·2H2O, 1.2 MgSO4·7H2O, 25 NaHCO3, 1.2 NaH2PO4·2H2O, and 10 glucose, pH 7.4) and fixed for 2 h at room temperature in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.0). After whole mounts containing either the circular or the longitudinal muscle layer with the myenteric plexus adhered were rinsed with PBS (NaCl 0.9%-0.1 M phosphate buffer, pH 7.0), they were prepared and further processed for immunocytochemistry. All primary and secondary antibodies were diluted in PBS containing 10% normal goat serum, 0.01% bovine serum albumine, 0.05% thimerosal, and 0.01% sodium azide (named as PBS*). To obtain enhanced sensitivity and to allow combination of two antisera raised in the same host species, a Cy3 (New England Nuclear, Perkin-Elmer Life Sciences; Boston, MA)-tyramide signal amplification (TSA) kit was applied (3). In short, endogenous peroxidase activity in the whole mount was blocked by incubation in 0.3% H2O2 in 50% methanol in PBS for 30 min. Whole mounts were then preincubated in PBS containing 1% Triton X-100 for 1 h and, subsequently, incubated overnight with a rabbit primary polyclonal antibody raised against CB1 (diluted 1:5,000; Affinity BioReagents; Golden, CO). The preparations were consecutively incubated in a goat anti-rabbit biotinylated Fab fragment (diluted 1:100; Jackson ImmunoResearch Laboratory; West Grove, PA) for 2.5 h and ExtrAvidinhorseradish peroxidase (diluted 1:1,000, Sigma; Saint Louis, MO) for 2 h. Between subsequent steps, tissues were rinsed three times for 5 min in 0.05% Tween 20 in PBS. Whole mounts were then subjected to TSA-Cy3 in "amplification solution" (diluted 1:100) for 20 min and rinsed in PBS. Finally, the remaining free binding sites were blocked with an unlabeled goat anti-rabbit Fab fragment (diluted 1:100; 3 h), and a second primary antibody [rabbit polyclonal IgG raised against neuronal nitric oxide (NO) synthase (nNOS; Eurodiagnostica; Arnhem, The Netherlands; diluted 1:500)] was applied overnight and visualized by coupling to an FITC-conjugated goat anti-rabbit secondary antiserum for 2 h (Jackson; diluted 1:100).
Data presentation and statistical analysis. The electrical stimulation of intrinsic neurons was followed by a transient excitatory junction potential (EJP) and a biphasic inhibitory junction potential (IJP) that contained a transient [fast (fIJP)] and a sustained [slow (sIJP)] component. The amplitudes of the three components were measured in millivolts compared with the resting membrane potential before application of the electrical stimulus.
All data are given as means ± SE (n indicates the number of independent observations in intracellular recordings from different animals). Statistical differences between the treatment groups were tested using a commercial statistical package (SigmaStat, Jandel Scientific; San Rafael, CA) using paired or unpaired Student's t-test where appropriate. A probability of <5% that the null hypothesis is false was considered to be significant.
Drugs. The drugs used were nifedipine, TTX, guanethidine, atropine, hexamethonium (all from Sigma-Aldrich; Taufkirchen, Germany), N-arachidonoyl ethanolamide (anandamide), WIN 55,212—2 (Biotrend; Köln, Germany), and N-piperidino-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-pyrazole-caroxamide (SR141716A; gift from Sanofi Research, France).
The drugs were freshly dissolved in ethanol for anandamide, DMSO for WIN 55,212—2 and SR141716A or saline for nifedipine, TTX, guanethidine, and atropine on the day of the experiment and further diluted in Krebs solution I to organ bath concentrations. Drugs were added to the perfusate, and experiments were controlled for the effects of the drug solvents. Neither ethanol nor DMSO nor saline in the final organ bath concentrations had a significant effect on the membrane resting potential or on evoked EJP and IJP, respectively.
RESULTS
Recording of circular smooth muscle membrane potential and electrically induced junction potentials. In the absence of drugs or stimulation, circular smooth muscle cells of the proximal colon displayed stable resting membrane potentials (RMP) in wild-type (-52.8 ± 4.4; n = 6) and CB1-deficient mice (-56.7 ± 6.9; n = 6).
Nifedipine, atropine, and guanethidine did not alter RMP of the smooth muscle cells of wild-type or CB1-deficient mice, respectively.
Electrical field stimulation (EFS) was used to stimulate enteric excitatory and inhibitory neurons to demonstrate functional neuronal input to the circular smooth muscle. EFS of enteric nerves elicited TTX-sensitive junction potentials (JPs), as described in the mouse proximal colon by Shuttleworth et al. (33). In the proximal colon, single stimuli of EFS (15 V, 0.3 ms, single pulse) produced JPs consisting of three components: a fEJP followed by a fIJP and an sIJP (Fig. 1). The initial fast excitatory component was characterized by a rapid depolarization with an amplitude of 13.1 ± 4.8 vs. 27.2 ± 5.2 (CB1+/+ vs. CB1-/-; P < 0.05) followed by an inhibitory "fast" 21.6 ± 8.1 vs. 23.7 ± 8.2 and "slow" 12.1 ± 3.3 vs. 13.5 ± 4.4 component (n = 6), which was characterized by a distinct and gradual recovery to RMP (n = 6). All components of the JPs were blocked by 10-5 M TTX (n = 5, data not shown). Atropine (1 μM) abolished the EJP (12.3 ± 2.6 vs. 0.9 ± 0.5; n = 3; P < 0.01) while leaving the fIJP (25.9 ± 4.9 vs. 24.6 ± 5.3; n = 3) and the sIJP (12.7 ± 5.4 vs. 11.8 ± 3.6; n = 3) unchanged (Fig. 1E). Hexamethonium did not significantly change the EJP (13.3 ± 1.2 vs. 10.7 ± 1.4; n = 3), the fIJP (26.7 ± 6.2 vs. 27.3 ± 5.0; n = 3), or the sIJP (12.7 ± 3.1 vs. 11.7 ± 1.2; n = 3).
Influence of anandamide, WIN 55,212—2, and SR141716A on junction potentials in wild-type mice. Neither the endogenous CB1 receptor agonist anandamide [(1 mM) -52.8 ± 4.4 vs. -50.6 ± 5.4 (wild-type) and -56.7 ± 6.9 vs. -51.7 ± 6.2 (CB1-/-)] nor the CB1-specific antagonist SR141716A [(1 mM) -51.6 ± 5.6 vs. -48.9 ± 7.4 (wild-type) and -55.6 ± 5.4 vs. -51.3 ± 7.7 (CB1-/-)] nor the synthetic CB1 receptor agonist WIN 55,212—2 [(1 mM) -52.3 ± 6.7 vs. -54.4 ± 3.7 (wild-type) and -55.1 ± 4.6 vs. -52.1 ± 8.1 (CB1-/-; Fig. 1 for WIN 55,212—2 and SR141716A)] influenced the resting membrane potential of smooth muscle cells in proximal colon.
In wild-type mice, anandamide (1 mM) significantly reduced the EJP (basal: 13.1 ± 4.8; with anandamide: 7.0 ± 2.2; n = 6; P < 0.05), whereas fIJP and sIJP were not significantly affected (Fig. 2). WIN 55,212—2 (1 mM) significantly reduced the EJP (basal: 12.7 ± 3.6; with WIN 55,212—2: 3.3 ± 3.4; n = 6; P < 0.05) and the fIJP (basal 22.7 ± 2.3; with WIN 55,212—2: 14.9 ± 2.7; n = 6; P < 0.05) but not sIJP (Fig. 3). SR141716A (1 mM) significantly increased the EJP (basal: 13.1 ± 4.8; with SR141716A: 20.7 ± 7.6; n = 6; P < 0.05), whereas fIJP and sIJP remained unchanged (Fig. 4). Addition of 1 mM SR141716A 20 min before anandamide (1 mM) application significantly reduced the anandamide effect on EJPs (17.0 ± 7.2; n = 6 for SR141716A vs. 12.7 ± 3.6; n = 6 for SR141716A and anandamide; Fig. 5). SR141716A (1 mM) administered 20 min before WIN 55,212—2 (1 mM) application significantly reduced the WIN 55,212—2 effect on EJPs and abolished the effect on fIJP (Fig. 6).
Influence of anandamide, WIN 55,212—2, and SR141716A on junction potentials in CB1-deficient mice. RMP in CB1-deficient mice was not significantly different from that in wild-type mice. CB1-deficient mice showed a significantly increased EJP, whereas fIJP and sIJP were not significantly changed compared with CB1+/+ (Fig. 1). In CB1-deficient mice, anandamide (1 mM) or WIN 55,212—2 (1 mM; Figs. 1, 2, 3) did not significantly influence EJP, fIJP, or sIJP. SR141716A (1 mM) did not significantly alter EJP nor sIJP, whereas fIJP, in the presence of SR141716A, was significantly reduced compared with wild-type mice (22.7 ± 2.6 vs. 15.8 ± 2.0 with SR141716A; n = 6; P < 0.05; Fig. 4). Addition of SR141716A (1 mM) 20 min before anandamide (1 mM) or WIN 55,212—2 (1 mM) application did not change EJP nor sIJP, whereas fIJPs, in the presence of SR141716A and in the presence of a combination of SR141716A and anandamide/WIN 55,212—2, were reduced in a similar manner as with SR141716A alone (Figs. 5 and 6).
Immunocytochemical identification of CB1 receptor-expressing myenteric neurons. To characterize a possible colocalization of CB1 and nNOS, as a marker of nitrergic inhibitory neurons, we performed double-staining experiments. CB1 immunoreactivity was demonstrated in myenteric neurons of whole mounts of mouse colon by using an antibody directed toward a fusion protein containing the first 77 amino acid residues of the rat CB1 receptor. Both large- and small-sized somata of neurons were found to display CB1 immunoreactivity. Double staining with antibodies raised against neuronal NOS revealed that the nitrergic myenteric population is clearly distinct from the neuronal population containing CB1 (Fig. 7). In addition, some non-neuronal CB1 immunoreactive cells were observed at the level of the outer muscle layer, but their nature remains to be determined.
DISCUSSION
It is well accepted that endogenous cannabinoids and CB1 are involved in the regulation of smooth muscle contractibility and intestinal motility. This has been demonstrated in different regions of the gastrointestinal tract including colon in a variety of species including humans (22, 23, 27, 29). Endogenous cannabinoids in the gastrointestinal tract act via CB1, and recently, interactions between cannabinoid and other receptor systems, such as vanilloid receptors (VR1) and adenosine A1 receptors have been suggested (2, 24, 27). The mechanisms by which CB1 activation influences gastrointestinal neurotransmission mainly relate to reduction of acetylcholine release from cholinergic nerve endings, although other mechanisms, such as reduction of contractile and relaxant nonadrenergic, noncholinergic neuronal activity have also been proposed (13, 34).
CB1 involvement in colonic neurotransmission in wild-type and CB1-deficient mice. The present study demonstrates that CB1 is involved in neurotransmission in the mouse colon. Electrically induced EJPs are significantly reduced by the CB1 agonist anandamide and, to a greater extent, by the synthetic agonist WIN 55,212—2. The effects of anandamide and WIN 55,212—2 are antagonized in the presence of the CB1 antagonist SR141716A. Interestingly, the presence of SR141716A alone increases EJP, suggesting a permanent basal activation of CB1 by endogenous cannabinoids.
The antagonistic effects of SR141716A in different tissues are still the subject of much debate. It is concluded in a large number of pharmacological investigations that activation of CB1 reduces cholinergic neurotransmission by reducing acetylcholine release (24, 28). However, other studies (28) using SR141716A obtained contrasting results, i.e., either effects or no effects induced by SR141716A were observed, which are sometimes explained by antagonism to endogenous CB1 activation and sometimes by inverse agonist activity of SR141716A. Moreover, these different results were obtained using different tissues and different species, implying that at present, no clear distinction can be made between a pure antagonistic or an inverse agonist effect. Our data suggest that the SR141716A effect also involves a CB1-independent mode of action, because this drug also reduced fIJP in the CB1-deficient mice. Apart from CB1-mediated effects of anandamide described in the literature, several recent publications (2, 24, 28, 34) report effects of this endocannabinoid, which cannot be ascribed to CB1 activation but which likely involved the activation of VR1 and/or non-CB1/non-CB2 receptors.
Therefore, we additionally performed the experiments in CB1-deficient mice, allowing us to monitor unambiguously which effects of the respective agonists or antagonists can be truly attributed to CB1. In the CB1-deficient mice, EJP was found to be significantly increased compared with wild-type littermate controls. This finding might be explained by the possibility that in wild-type animals, a permanent suppressant activity on cholinergic mechanisms, e.g., acetylcholine release, via permanently activated CB1 is present, whereas this mechanism is lacking in CB1-deficient mice. This notion is supported by other experiments on CB1-deficient mice, e.g., reporting an increased acetylcholine release in hippocampal slices in these mutants (17) and further reinforced by the effect of SR141716A on EJP in wild-type animals and by the effect of SR141716A in other investigations, where the CB1 antagonist significantly increased acetylcholine-mediated smooth muscle responses and where SR141716A increased acetylcholine release (7). The finding that EJPs in our tissue preparations were abolished by the presence of either atropine or TTX led us to conclude that CB1 receptors are tonically activated in the mouse proximal colon and exert a suppressant effect on cholinergic excitatory neurotransmission. Additional activation of CB1 results in a reduction and, in the case of the synthetic agonist WIN 55,212—2, nearly an abolition of the cholinergic EJP. This effect is antagonized by SR141716A in wild-type animals and is absent in the CB1-deficient animals, suggesting that the effect of endogenous and exogenous cannabinoids on cholinergic neurotransmission is exclusively mediated by CB1. This is interesting, because in other preparations such as guinea-pig ileum, anandamide is believed to reduce cholinergic transmission not only by CB1 activation but also by activation of VR1 (24). However, such mechanisms seem unlikely in the mouse proximal colon, because neither anandamide nor WIN 55,212—2 exerted any effects on EJP in CB1-deficient mice.
No further actions of anandamide on fIJP or sIJP could be demonstrated either in wild-type or in CB1-deficient animals. However, WIN 55,212—2, which is known to be more potent than anandamide, significantly reduced fIJP in wild-type mice. Because this WIN 55,212—2 effect on fIJP is reversed by SR141716A in wild-type animals and because this effect is lacking in CB1-deficient mice, this effect is likely to be mediated by CB1 receptors. fIJP in this mouse colonic preparation represents one part of the nonadrenergic, noncholinergic inhibitory neurotransmission, i.e., the apamine-sensitive component, whereas the unaffected sIJP is NO dependent (11, 35). Although most of the CB1-mediated actions reported within the gastrointestinal tract appear to have suppressant effects on cholinergic excitatory neurotransmission, other reports (13, 34) also suggest an involvement of CB1 in nonadrenergic, noncholinergic neurotransmission.
Furthermore, although SR141716A did not exert any influence on fIJP and sIJP in wild-type mice, it significantly reduced fIJP in CB1-deficient mice. Any discussion on the nature of this effect remains speculative, because this finding questions the selectivity of SR141716A. However, it is possible that in CB1-deficient mice, a structurally related receptor is overexpressed and thus explains this effect, because this SR141716A effect on IJP was not observed in wild-type mice. It is the matter of further research to evaluate the SR141716A effect on fIJP in CB1-deficient mice and to characterize the involved mechanisms. It might be speculated that a structurally related receptor, e.g., the non-CB1/non-CB2 receptor supposed in numerous publications, may be underlying.
Localization of the CB1 receptor. The present report describes effects in smooth muscle cells following electrical stimulation of myenteric neurons. Only few studies have investigated cannabinoid actions by measuring changes in membrane potentials. In the guinea pig ileum, neuronal recordings revealed that WIN 55,212—2 reduces the fast cholinergic excitatory postsynaptic potentials, providing evidence that CB receptors are not only involved in neuromuscular but also in neuroneuronal cholinergic transmission in the gut (20). In half of the population of myenteric neurons, WIN 55,212—2 had no effect (20). As such, the question of whether the presence of cannabinoid receptors is confined to one or more functional classes of neurons has remained unanswered to the present. Although these findings on neuronal stimulation are in agreement with our findings, it has to be kept in mind that in our setup, recordings were made from the smooth muscle cell and not from the neurons directly. According to our findings in the hexamethonium experiments, the most likely location of the involved CB1 receptors is varicosities of the motor nerve terminals. However, our data do not suggest that CB1 receptors within the myenteric plexus are exclusively located on varicosities of the motor nerve terminals. Additional locations within the myenteric plexus as suggested by others (20) seem likely.
Apart from established localization of CB1 on non-nitrergic myenteric neurons, both our functional and immunocytochemical data do not exclude another non-neuronal location of this receptor. In agreement with earlier immunocytochemical reports (6, 18, 28), no direct immunocytochemical evidence could be provided for the expression of CB1 on smooth muscle cells. From our present results, both locations of CB1 seem to be possible. Besides different peripheral receptor locations, the literature suggests that additionally, central CB1 receptors might contribute to the regulation of colonic motility in analogy to the finding that both central and peripheral CB1 receptors are involved in the regulation of intestinal motility (15).
Within the rat stomach, CB1 immunoreactivity was detected on neuronal elements innervating blood vessels of mucosa and submucosa as well as on smooth muscle (1). Within the enteric nervous system, CB1 immunoreactivity was found to be colocalized with immunoreactivity to choline acetyltransferase, a marker of cholinergic neurons (6, 18, 29). This colocalization is in agreement with functional experiments demonstrating a suppresant effect of CB1 activation on neurally mediated cholinergic mechanisms or on neuronal acetylcholine release (24). Our finding that CB1 and NOS immunoreactivity are not colocalized in the same myenteric neurons is in line with our electrophysiological recordings showing no change in sIJP after treatment with cannabinoids.
Regarding the suppressant effect on the excitatory postsynaptic potential, our findings are in good agreement with other studies. In the guinea pig ileum, activation of CB1 by WIN 55,212—2 or the synthetic CB1 agonist CP 55,940 suppressed excitatory postsynaptic potentials by 37—46% in concentration- and CB1-dependent manners (20). SR141716A antagonized the effect of WIN 55,212—2 but, when given alone, caused a 40—50% reduction of excitatory postsynaptic potentials similar to the reduction caused by WIN 55,212—2.
In summary, the present investigation for the first time provides evidence that in the murine proximal colon, cannabinoids are involved in both excitatory cholinergic and inhibitory nonadrenergic, noncholinergic neurotransmission and that these effects are mediated via the CB1 receptor. The specific involvement of CB1 in the suppressant effect on cholinergically mediated EJP of CB1 agonists has clearly been shown, because the effects induced by anandamide and WIN 55,212—2 were reversed by the selective CB1 antagonist SR141716A and because anandamide and WIN 55,212—2 were not able to induce effects in CB1-deficient mice. The physiological impact of the reported modulation of colonic neurotransmission via peripheral CB1 receptors has yet to be established, because modulators of enteric neurotransmission might serve as therapeutic tools in a variety of functional gastrointestinal disorders (9).
Source, Graphs and Figures: Cannabinoid receptor type 1 modulates excitatory and inhibitory neurotransmission in mouse colon
The effects of cannabinoid receptor agonists and antagonists on smooth muscle resting membrane potentials and on membrane potentials following electrical neuronal stimulation in a myenteric neuron/smooth muscle preparation of wild-type and cannabinoid receptor type 1 (CB1)-deficient mice were investigated in vitro. Double staining for CB1 and nitric oxide synthase (neuronal) was performed to identify the myenteric CB1-expressing neurons. Focal electrical stimulation of the myenteric plexus induced a fast (f) excitatory junction potential (EJP) followed by a fast and a slow (s) inhibitory junction potential (IJP). Treatment of wild-type mice with the endogenous CB1 receptor agonist anandamide reduced EJP while not affecting fIJP and sIJP. EJP was significantly higher in CB1-deficient mice than in wild-type littermate controls, and anandamide induced no effects in CB1-deficient mice. N-arachidonoyl ethanolamide (anandamide), R-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3,-de]- 1,4-benzoxazin-6-yl]-1-naphtalenylmethanone, a synthetic CB1 receptor agonist, nearly abolished EJP and significantly reduced the fIJP in wild-type mice. N-piperidino-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-pyrazole-caroxamide (SR141716A), a CB1-specific receptor antagonist, was able to reverse the agonist effects induced in wild-type mice. SR141716A, when given alone, significantly increased EJP in wild-type mice without affecting IJP in wild-type and EJP in CB1-deficient mice. Interestingly, SR141716A reduced fIJP in CB1-deficient mice. In the mouse colon, nitrergic myenteric neurons do not express CB1, implying that CB1 is expressed in cholinergic neurons, which is in line with the functional data. Finally, excitatory and inhibitory neurotransmission in the mouse colon is modulated by activation of CB1 receptors. The significant increase in EJP in CB1-deficient mice strongly suggests a physiological involvement of CB1 in excitatory cholinergic neurotransmission.
TO THE PRESENT, THE ENDOCANNABINOID system comprises two G protein-coupled receptors [cannabinoid receptor types 1 (CB1) and 2 (CB2)], three known endogenous agonists (anandamide, 2-arachidonoyl glycerol, nolandin ether), one endogenous CB1 antagonist/CB2 agonist (virodhamine), and enzymes degrading endocannabinoids [fatty acid amide hydrolase for anandamide and monoglycerol lipase for 2-arachidonoyl glycerol (10, 12, 21, 30)]. CB1 is present in the peripheral and central nervous system and is distinctly involved in numerous physiological effects, whereas CB2 is expressed mainly in immune cells. A general feature of CB1 activation is the reduction of the release of a variety of neurotransmitters (e.g., acetylcholine from enteric nerves) (16, 32). Whereas there is currently no evidence of a role of CB2 in the gastrointestinal tract, CB1 appears to be involved in several physiological processes in this tissue (9, 14, 28). The cannabinoid system is involved, e.g., in intestinal motility, gastrointestinal transit, gastric secretion, and colonic propulsion. In agreement with these observations, CB1 was shown to be localized in the gastrointestinal tract of many species including humans (1, 5, 6, 18, 19, 26, 29). In the gastrointestinal tract of pigs, which is functionally similar to that of humans, CB1 is present in neurons of the myenteric and submucosal plexus of the ileum and the colon as shown by immunohistochemistry. In these neurons, CB1 is coexpressed with cholinergic markers (18). In fact, cannabinoids, including anandamide and the synthetic agonist R-[2,3-dihydro-5-methyl-3-(4-morpholinylmethyl)pyrrolo[1,2,3,-de]-1,4-benzoxazin-6-yl]-1-naphtalenylmethanone (WIN 55,212—2), reduce electrically induced contractions by activation of CB1 that is located presynaptically on enteric nerves (8, 13). Furthermore, in the rat gastric fundus, cannabinoids depress excitatory neurotransmission but also modulate inhibitory neurotransmission (34).
Apart from the effects of CB1 on small intestinal motility (4), little is know about cannabinoid action on colonic motility. Two studies (23, 29) described that CB1 on myenteric neurons is involved in the reduction of colonic propulsion in mice, supporting the notion that CB1 is involved in the regulation of colonic motility.
Given the scarce knowledge on the action of cannabinoids on neurotransmission in the gastrointestinal tract, the effects of cannabinoids on neurotransmission in the colon were evaluated in the present study by intracellular recordings using CB1-deficient mice (25) and the specific antagonist SR141716A (31).
MATERIALS AND METHODS
Tissue preparation for electrophysiological experiments. Eight- to ten-week-old female CB1-deficient mice or wild-type littermates in C57BL/6N background (25) were anesthetized by sodium phenobarbital and killed by cervical dislocation in accordance with the recommendations of the animal ethics committee at the Technical University of Munich. Colon was exposed through an abdominal midline incision. The complete large bowel was removed and placed into oxygenated Krebs solution I of the following composition (in mM): 120.35 NaCl, 5.9 KCl, 2.5 MgCl2, 1.2 NaH2PO4, 15.5, NaHCO3, 2.5 CaCl2, and 11.5 glucose (pH 7.4). Colon was opened along the mesenteric border, washed off from remaining fecal material, and pinned out in a Sylgard-lined (Dow Corning; Midland, MI) dissecting dish containing oxygenated Krebs solution I. Proximal colon was separated, and the mucosa and submucosa were removed, resulting in sheets of tissue consisting of circular and longitudinal muscle layers together with the attached myenteric plexus.
Intracellular electrical recording. Sheets or strips of muscle (proximal colon) were pinned using ∼150—200 micropins (15—25 μm in thickness), obtained from wolfram wire, to the Sylgard-based electrophysiological chamber, with the exposed circular muscle layer uppermost. The chamber was perfused (5 ml/min; Kwik Pump, World Precision Instruments; Sarasota, FL) with prewarmed (37°C) oxygenated (95% O2-5% CO2) Krebs solution I. Tissues were allowed to equilibrate for 90—120 min before the experiments were started. Nifedipine (1 μM) was present throughout all experiments. In experiments investigating sIJPs and fIJPs, atropine (1 μM) and guanethidine (1 μM) were present continuously. Capillary glass microelectrodes (borosilicate glass capillaries, 1.0-mm outer diameter × 0.58-mm inner diameter, Clark Electromedical Instruments; Edenbridge, UK) were made using a microelectrode puller (model P-97, 3-mm-wide filament, Sutter Instruments; Novato, CA) filled with KCl (3 M) and had resistances in the range of 80—120 MΩ. Neurons were stimulated (15 V, 0.3 ms, single pulse) via platinum electrodes arranged perpendicularly to the circular muscle layer, which were connected to a Grass S11 stimulator via a stimulus isolation unit (Grass SIU59, Grass Instruments; Quincy, MA). The responses were recorded against a "ground" Ag-AgCl electrode placed in the bath medium. Evoked electrical events were amplified (DUO 733 microelectrode amplifier, World Precision Instruments; Sarasota, FL) and digitalized with an analog-to-digital converter (SCB 68 interface, National Instruments; Austin, TX). Permanent recordings of membrane potentials were made on a personal computer running the LABVIEW 5.0 program (National Instruments; Austin, TX).
Tissue preparation for immunocytochemistry. Adult Swiss mice of both genders (n = 4) were killed using an overdose of pentobarbital sodium (Nembutal; intraperitoneally injected). One 10-mm segment of the proximal and distal colon was removed, rinsed in Krebs solution II (in mM: 117 NaCl, 5 KCl, 2.5 CaCl2·2H2O, 1.2 MgSO4·7H2O, 25 NaHCO3, 1.2 NaH2PO4·2H2O, and 10 glucose, pH 7.4) and fixed for 2 h at room temperature in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.0). After whole mounts containing either the circular or the longitudinal muscle layer with the myenteric plexus adhered were rinsed with PBS (NaCl 0.9%-0.1 M phosphate buffer, pH 7.0), they were prepared and further processed for immunocytochemistry. All primary and secondary antibodies were diluted in PBS containing 10% normal goat serum, 0.01% bovine serum albumine, 0.05% thimerosal, and 0.01% sodium azide (named as PBS*). To obtain enhanced sensitivity and to allow combination of two antisera raised in the same host species, a Cy3 (New England Nuclear, Perkin-Elmer Life Sciences; Boston, MA)-tyramide signal amplification (TSA) kit was applied (3). In short, endogenous peroxidase activity in the whole mount was blocked by incubation in 0.3% H2O2 in 50% methanol in PBS for 30 min. Whole mounts were then preincubated in PBS containing 1% Triton X-100 for 1 h and, subsequently, incubated overnight with a rabbit primary polyclonal antibody raised against CB1 (diluted 1:5,000; Affinity BioReagents; Golden, CO). The preparations were consecutively incubated in a goat anti-rabbit biotinylated Fab fragment (diluted 1:100; Jackson ImmunoResearch Laboratory; West Grove, PA) for 2.5 h and ExtrAvidinhorseradish peroxidase (diluted 1:1,000, Sigma; Saint Louis, MO) for 2 h. Between subsequent steps, tissues were rinsed three times for 5 min in 0.05% Tween 20 in PBS. Whole mounts were then subjected to TSA-Cy3 in "amplification solution" (diluted 1:100) for 20 min and rinsed in PBS. Finally, the remaining free binding sites were blocked with an unlabeled goat anti-rabbit Fab fragment (diluted 1:100; 3 h), and a second primary antibody [rabbit polyclonal IgG raised against neuronal nitric oxide (NO) synthase (nNOS; Eurodiagnostica; Arnhem, The Netherlands; diluted 1:500)] was applied overnight and visualized by coupling to an FITC-conjugated goat anti-rabbit secondary antiserum for 2 h (Jackson; diluted 1:100).
Data presentation and statistical analysis. The electrical stimulation of intrinsic neurons was followed by a transient excitatory junction potential (EJP) and a biphasic inhibitory junction potential (IJP) that contained a transient [fast (fIJP)] and a sustained [slow (sIJP)] component. The amplitudes of the three components were measured in millivolts compared with the resting membrane potential before application of the electrical stimulus.
All data are given as means ± SE (n indicates the number of independent observations in intracellular recordings from different animals). Statistical differences between the treatment groups were tested using a commercial statistical package (SigmaStat, Jandel Scientific; San Rafael, CA) using paired or unpaired Student's t-test where appropriate. A probability of <5% that the null hypothesis is false was considered to be significant.
Drugs. The drugs used were nifedipine, TTX, guanethidine, atropine, hexamethonium (all from Sigma-Aldrich; Taufkirchen, Germany), N-arachidonoyl ethanolamide (anandamide), WIN 55,212—2 (Biotrend; Köln, Germany), and N-piperidino-5-(4-chlorophenyl)-1-(2,4-dichlorophenyl)-4-methyl-3-pyrazole-caroxamide (SR141716A; gift from Sanofi Research, France).
The drugs were freshly dissolved in ethanol for anandamide, DMSO for WIN 55,212—2 and SR141716A or saline for nifedipine, TTX, guanethidine, and atropine on the day of the experiment and further diluted in Krebs solution I to organ bath concentrations. Drugs were added to the perfusate, and experiments were controlled for the effects of the drug solvents. Neither ethanol nor DMSO nor saline in the final organ bath concentrations had a significant effect on the membrane resting potential or on evoked EJP and IJP, respectively.
RESULTS
Recording of circular smooth muscle membrane potential and electrically induced junction potentials. In the absence of drugs or stimulation, circular smooth muscle cells of the proximal colon displayed stable resting membrane potentials (RMP) in wild-type (-52.8 ± 4.4; n = 6) and CB1-deficient mice (-56.7 ± 6.9; n = 6).
Nifedipine, atropine, and guanethidine did not alter RMP of the smooth muscle cells of wild-type or CB1-deficient mice, respectively.
Electrical field stimulation (EFS) was used to stimulate enteric excitatory and inhibitory neurons to demonstrate functional neuronal input to the circular smooth muscle. EFS of enteric nerves elicited TTX-sensitive junction potentials (JPs), as described in the mouse proximal colon by Shuttleworth et al. (33). In the proximal colon, single stimuli of EFS (15 V, 0.3 ms, single pulse) produced JPs consisting of three components: a fEJP followed by a fIJP and an sIJP (Fig. 1). The initial fast excitatory component was characterized by a rapid depolarization with an amplitude of 13.1 ± 4.8 vs. 27.2 ± 5.2 (CB1+/+ vs. CB1-/-; P < 0.05) followed by an inhibitory "fast" 21.6 ± 8.1 vs. 23.7 ± 8.2 and "slow" 12.1 ± 3.3 vs. 13.5 ± 4.4 component (n = 6), which was characterized by a distinct and gradual recovery to RMP (n = 6). All components of the JPs were blocked by 10-5 M TTX (n = 5, data not shown). Atropine (1 μM) abolished the EJP (12.3 ± 2.6 vs. 0.9 ± 0.5; n = 3; P < 0.01) while leaving the fIJP (25.9 ± 4.9 vs. 24.6 ± 5.3; n = 3) and the sIJP (12.7 ± 5.4 vs. 11.8 ± 3.6; n = 3) unchanged (Fig. 1E). Hexamethonium did not significantly change the EJP (13.3 ± 1.2 vs. 10.7 ± 1.4; n = 3), the fIJP (26.7 ± 6.2 vs. 27.3 ± 5.0; n = 3), or the sIJP (12.7 ± 3.1 vs. 11.7 ± 1.2; n = 3).
Influence of anandamide, WIN 55,212—2, and SR141716A on junction potentials in wild-type mice. Neither the endogenous CB1 receptor agonist anandamide [(1 mM) -52.8 ± 4.4 vs. -50.6 ± 5.4 (wild-type) and -56.7 ± 6.9 vs. -51.7 ± 6.2 (CB1-/-)] nor the CB1-specific antagonist SR141716A [(1 mM) -51.6 ± 5.6 vs. -48.9 ± 7.4 (wild-type) and -55.6 ± 5.4 vs. -51.3 ± 7.7 (CB1-/-)] nor the synthetic CB1 receptor agonist WIN 55,212—2 [(1 mM) -52.3 ± 6.7 vs. -54.4 ± 3.7 (wild-type) and -55.1 ± 4.6 vs. -52.1 ± 8.1 (CB1-/-; Fig. 1 for WIN 55,212—2 and SR141716A)] influenced the resting membrane potential of smooth muscle cells in proximal colon.
In wild-type mice, anandamide (1 mM) significantly reduced the EJP (basal: 13.1 ± 4.8; with anandamide: 7.0 ± 2.2; n = 6; P < 0.05), whereas fIJP and sIJP were not significantly affected (Fig. 2). WIN 55,212—2 (1 mM) significantly reduced the EJP (basal: 12.7 ± 3.6; with WIN 55,212—2: 3.3 ± 3.4; n = 6; P < 0.05) and the fIJP (basal 22.7 ± 2.3; with WIN 55,212—2: 14.9 ± 2.7; n = 6; P < 0.05) but not sIJP (Fig. 3). SR141716A (1 mM) significantly increased the EJP (basal: 13.1 ± 4.8; with SR141716A: 20.7 ± 7.6; n = 6; P < 0.05), whereas fIJP and sIJP remained unchanged (Fig. 4). Addition of 1 mM SR141716A 20 min before anandamide (1 mM) application significantly reduced the anandamide effect on EJPs (17.0 ± 7.2; n = 6 for SR141716A vs. 12.7 ± 3.6; n = 6 for SR141716A and anandamide; Fig. 5). SR141716A (1 mM) administered 20 min before WIN 55,212—2 (1 mM) application significantly reduced the WIN 55,212—2 effect on EJPs and abolished the effect on fIJP (Fig. 6).
Influence of anandamide, WIN 55,212—2, and SR141716A on junction potentials in CB1-deficient mice. RMP in CB1-deficient mice was not significantly different from that in wild-type mice. CB1-deficient mice showed a significantly increased EJP, whereas fIJP and sIJP were not significantly changed compared with CB1+/+ (Fig. 1). In CB1-deficient mice, anandamide (1 mM) or WIN 55,212—2 (1 mM; Figs. 1, 2, 3) did not significantly influence EJP, fIJP, or sIJP. SR141716A (1 mM) did not significantly alter EJP nor sIJP, whereas fIJP, in the presence of SR141716A, was significantly reduced compared with wild-type mice (22.7 ± 2.6 vs. 15.8 ± 2.0 with SR141716A; n = 6; P < 0.05; Fig. 4). Addition of SR141716A (1 mM) 20 min before anandamide (1 mM) or WIN 55,212—2 (1 mM) application did not change EJP nor sIJP, whereas fIJPs, in the presence of SR141716A and in the presence of a combination of SR141716A and anandamide/WIN 55,212—2, were reduced in a similar manner as with SR141716A alone (Figs. 5 and 6).
Immunocytochemical identification of CB1 receptor-expressing myenteric neurons. To characterize a possible colocalization of CB1 and nNOS, as a marker of nitrergic inhibitory neurons, we performed double-staining experiments. CB1 immunoreactivity was demonstrated in myenteric neurons of whole mounts of mouse colon by using an antibody directed toward a fusion protein containing the first 77 amino acid residues of the rat CB1 receptor. Both large- and small-sized somata of neurons were found to display CB1 immunoreactivity. Double staining with antibodies raised against neuronal NOS revealed that the nitrergic myenteric population is clearly distinct from the neuronal population containing CB1 (Fig. 7). In addition, some non-neuronal CB1 immunoreactive cells were observed at the level of the outer muscle layer, but their nature remains to be determined.
DISCUSSION
It is well accepted that endogenous cannabinoids and CB1 are involved in the regulation of smooth muscle contractibility and intestinal motility. This has been demonstrated in different regions of the gastrointestinal tract including colon in a variety of species including humans (22, 23, 27, 29). Endogenous cannabinoids in the gastrointestinal tract act via CB1, and recently, interactions between cannabinoid and other receptor systems, such as vanilloid receptors (VR1) and adenosine A1 receptors have been suggested (2, 24, 27). The mechanisms by which CB1 activation influences gastrointestinal neurotransmission mainly relate to reduction of acetylcholine release from cholinergic nerve endings, although other mechanisms, such as reduction of contractile and relaxant nonadrenergic, noncholinergic neuronal activity have also been proposed (13, 34).
CB1 involvement in colonic neurotransmission in wild-type and CB1-deficient mice. The present study demonstrates that CB1 is involved in neurotransmission in the mouse colon. Electrically induced EJPs are significantly reduced by the CB1 agonist anandamide and, to a greater extent, by the synthetic agonist WIN 55,212—2. The effects of anandamide and WIN 55,212—2 are antagonized in the presence of the CB1 antagonist SR141716A. Interestingly, the presence of SR141716A alone increases EJP, suggesting a permanent basal activation of CB1 by endogenous cannabinoids.
The antagonistic effects of SR141716A in different tissues are still the subject of much debate. It is concluded in a large number of pharmacological investigations that activation of CB1 reduces cholinergic neurotransmission by reducing acetylcholine release (24, 28). However, other studies (28) using SR141716A obtained contrasting results, i.e., either effects or no effects induced by SR141716A were observed, which are sometimes explained by antagonism to endogenous CB1 activation and sometimes by inverse agonist activity of SR141716A. Moreover, these different results were obtained using different tissues and different species, implying that at present, no clear distinction can be made between a pure antagonistic or an inverse agonist effect. Our data suggest that the SR141716A effect also involves a CB1-independent mode of action, because this drug also reduced fIJP in the CB1-deficient mice. Apart from CB1-mediated effects of anandamide described in the literature, several recent publications (2, 24, 28, 34) report effects of this endocannabinoid, which cannot be ascribed to CB1 activation but which likely involved the activation of VR1 and/or non-CB1/non-CB2 receptors.
Therefore, we additionally performed the experiments in CB1-deficient mice, allowing us to monitor unambiguously which effects of the respective agonists or antagonists can be truly attributed to CB1. In the CB1-deficient mice, EJP was found to be significantly increased compared with wild-type littermate controls. This finding might be explained by the possibility that in wild-type animals, a permanent suppressant activity on cholinergic mechanisms, e.g., acetylcholine release, via permanently activated CB1 is present, whereas this mechanism is lacking in CB1-deficient mice. This notion is supported by other experiments on CB1-deficient mice, e.g., reporting an increased acetylcholine release in hippocampal slices in these mutants (17) and further reinforced by the effect of SR141716A on EJP in wild-type animals and by the effect of SR141716A in other investigations, where the CB1 antagonist significantly increased acetylcholine-mediated smooth muscle responses and where SR141716A increased acetylcholine release (7). The finding that EJPs in our tissue preparations were abolished by the presence of either atropine or TTX led us to conclude that CB1 receptors are tonically activated in the mouse proximal colon and exert a suppressant effect on cholinergic excitatory neurotransmission. Additional activation of CB1 results in a reduction and, in the case of the synthetic agonist WIN 55,212—2, nearly an abolition of the cholinergic EJP. This effect is antagonized by SR141716A in wild-type animals and is absent in the CB1-deficient animals, suggesting that the effect of endogenous and exogenous cannabinoids on cholinergic neurotransmission is exclusively mediated by CB1. This is interesting, because in other preparations such as guinea-pig ileum, anandamide is believed to reduce cholinergic transmission not only by CB1 activation but also by activation of VR1 (24). However, such mechanisms seem unlikely in the mouse proximal colon, because neither anandamide nor WIN 55,212—2 exerted any effects on EJP in CB1-deficient mice.
No further actions of anandamide on fIJP or sIJP could be demonstrated either in wild-type or in CB1-deficient animals. However, WIN 55,212—2, which is known to be more potent than anandamide, significantly reduced fIJP in wild-type mice. Because this WIN 55,212—2 effect on fIJP is reversed by SR141716A in wild-type animals and because this effect is lacking in CB1-deficient mice, this effect is likely to be mediated by CB1 receptors. fIJP in this mouse colonic preparation represents one part of the nonadrenergic, noncholinergic inhibitory neurotransmission, i.e., the apamine-sensitive component, whereas the unaffected sIJP is NO dependent (11, 35). Although most of the CB1-mediated actions reported within the gastrointestinal tract appear to have suppressant effects on cholinergic excitatory neurotransmission, other reports (13, 34) also suggest an involvement of CB1 in nonadrenergic, noncholinergic neurotransmission.
Furthermore, although SR141716A did not exert any influence on fIJP and sIJP in wild-type mice, it significantly reduced fIJP in CB1-deficient mice. Any discussion on the nature of this effect remains speculative, because this finding questions the selectivity of SR141716A. However, it is possible that in CB1-deficient mice, a structurally related receptor is overexpressed and thus explains this effect, because this SR141716A effect on IJP was not observed in wild-type mice. It is the matter of further research to evaluate the SR141716A effect on fIJP in CB1-deficient mice and to characterize the involved mechanisms. It might be speculated that a structurally related receptor, e.g., the non-CB1/non-CB2 receptor supposed in numerous publications, may be underlying.
Localization of the CB1 receptor. The present report describes effects in smooth muscle cells following electrical stimulation of myenteric neurons. Only few studies have investigated cannabinoid actions by measuring changes in membrane potentials. In the guinea pig ileum, neuronal recordings revealed that WIN 55,212—2 reduces the fast cholinergic excitatory postsynaptic potentials, providing evidence that CB receptors are not only involved in neuromuscular but also in neuroneuronal cholinergic transmission in the gut (20). In half of the population of myenteric neurons, WIN 55,212—2 had no effect (20). As such, the question of whether the presence of cannabinoid receptors is confined to one or more functional classes of neurons has remained unanswered to the present. Although these findings on neuronal stimulation are in agreement with our findings, it has to be kept in mind that in our setup, recordings were made from the smooth muscle cell and not from the neurons directly. According to our findings in the hexamethonium experiments, the most likely location of the involved CB1 receptors is varicosities of the motor nerve terminals. However, our data do not suggest that CB1 receptors within the myenteric plexus are exclusively located on varicosities of the motor nerve terminals. Additional locations within the myenteric plexus as suggested by others (20) seem likely.
Apart from established localization of CB1 on non-nitrergic myenteric neurons, both our functional and immunocytochemical data do not exclude another non-neuronal location of this receptor. In agreement with earlier immunocytochemical reports (6, 18, 28), no direct immunocytochemical evidence could be provided for the expression of CB1 on smooth muscle cells. From our present results, both locations of CB1 seem to be possible. Besides different peripheral receptor locations, the literature suggests that additionally, central CB1 receptors might contribute to the regulation of colonic motility in analogy to the finding that both central and peripheral CB1 receptors are involved in the regulation of intestinal motility (15).
Within the rat stomach, CB1 immunoreactivity was detected on neuronal elements innervating blood vessels of mucosa and submucosa as well as on smooth muscle (1). Within the enteric nervous system, CB1 immunoreactivity was found to be colocalized with immunoreactivity to choline acetyltransferase, a marker of cholinergic neurons (6, 18, 29). This colocalization is in agreement with functional experiments demonstrating a suppresant effect of CB1 activation on neurally mediated cholinergic mechanisms or on neuronal acetylcholine release (24). Our finding that CB1 and NOS immunoreactivity are not colocalized in the same myenteric neurons is in line with our electrophysiological recordings showing no change in sIJP after treatment with cannabinoids.
Regarding the suppressant effect on the excitatory postsynaptic potential, our findings are in good agreement with other studies. In the guinea pig ileum, activation of CB1 by WIN 55,212—2 or the synthetic CB1 agonist CP 55,940 suppressed excitatory postsynaptic potentials by 37—46% in concentration- and CB1-dependent manners (20). SR141716A antagonized the effect of WIN 55,212—2 but, when given alone, caused a 40—50% reduction of excitatory postsynaptic potentials similar to the reduction caused by WIN 55,212—2.
In summary, the present investigation for the first time provides evidence that in the murine proximal colon, cannabinoids are involved in both excitatory cholinergic and inhibitory nonadrenergic, noncholinergic neurotransmission and that these effects are mediated via the CB1 receptor. The specific involvement of CB1 in the suppressant effect on cholinergically mediated EJP of CB1 agonists has clearly been shown, because the effects induced by anandamide and WIN 55,212—2 were reversed by the selective CB1 antagonist SR141716A and because anandamide and WIN 55,212—2 were not able to induce effects in CB1-deficient mice. The physiological impact of the reported modulation of colonic neurotransmission via peripheral CB1 receptors has yet to be established, because modulators of enteric neurotransmission might serve as therapeutic tools in a variety of functional gastrointestinal disorders (9).
Source, Graphs and Figures: Cannabinoid receptor type 1 modulates excitatory and inhibitory neurotransmission in mouse colon