Truth Seeker
New Member
Abstract
Cannabinoid-1 (CB1) receptors on myenteric neurons are involved in the regulation of intestinal motility. Our aim was to investigate CB1 receptor involvement in ascending neurotransmission in mouse colon and to characterize the involved structures by functional and morphological means. Presence of the CB1 receptor was investigated by RT-PCR, and immunohistochemistry was used for colabeling studies. Myenteric reflex responses were initiated by electrical stimulation (ES) at different distances, and junction potentials (JP) were recorded from circular smooth muscle cells by intracellular recording in an unpartitioned and a partitioned recording chamber. In vivo colonic propulsion was tested in wild-type and CB1−/− mice. Immunostaining with the cytoskeletal marker peripherin showed CB1 immunoreactivity both on Dogiel type I and type II neurons. Further neurochemical characterization revealed CB1 on choline acetyltransferase-, calretinin-, and 5-HT-immunopositive myenteric neurons, but nitrergic neurons appeared immunonegative for CB1 immunostaining. Solitary spindle-shaped CB1-immunoreactive cells in between smooth muscle cells lacked specific markers for interstitial cells of Cajal or glial cells. ES elicited neuronally mediated excitatory JP (EJP) and inhibitory JP. Gradual increases in distance resulted in a wave-like EJP with EJP amplitudes being maximal at the location of stimulating electrode 6 and a maximal EJP projection distance of ∼18 mm. The CB1 receptor agonist WIN 55,212-2 reduced the amplitude of EJP and was responsible for shortening the oral spreading of the excitatory impulse. In a partitioned chamber, WIN 55,212-2 reduced EJP at the separated oral sites, proving that CB1 activation inhibits interneuron-mediated neurotransmission. These effects were absent in the presence of the CB1 antagonist SR141716A, which, when given alone, had no effect. WIN 55,212-2 inhibited colonic propulsion in wild-type mice but not in SR141716A-pretreated wild-type or CB1−/− mice. Activation of the CB1 receptor modulates excitatory cholinergic neurotransmission in mouse colon by reducing amplitude and spatial spreading of the ascending electrophysiological impulses. This effect on electrophysiological spreading involves CB1-mediated effects on motor neurons and ascending interneurons and is likely to underlie the here reported in vivo reduction in colonic propulsion.
ENDOCANNABINOIDS AND THEIR respective receptors are present throughout the gastrointestinal (GI) tract (7, 9). Whereas the cannabinoid-2 (CB2) receptor is mainly located on immune cells and its involvement in the regulation of GI motility remains speculative, there is strong evidence that cannabinoid-1 (CB1) receptors exert such roles (17, 26, 29).
In vitro experiments in various GI regions, including stomach and small and large intestine, in different species have proven that activation of the CB1 receptor by endogenous or exogenous agonists modulates motility mainly by reducing contractile activity (23, 29, 31, 41). These effects involve reduction in neurotransmitter release following activation of CB1 receptors located on a presynaptic site of the motorneuron (1, 25, 37). To date it is unknown whether this presynaptic location involves peripheral motor neurons, interneurons, or both, and it is unknown to what extent CB1 receptors are also involved in the regulation of neuronal circuits of intestinal motility rather than reduction of neurotransmitter release at the neuromuscular junction only. Beside the effects on cholinergic excitatory neurotransmission, additional CB1-mediated effects on excitatory tachykininergic and inhibitory nonadrenergic, noncholinergic neurotransmission were previously reported in the GI tract (16, 23, 41).
In more complex preparations, the cannabinoid system potently reduced small intestinal peristaltic activity and colonic bead propulsion, where activation of the CB1 receptor caused a decrease and application of respective receptor antagonists and increase in peristalsis and propulsion (14, 15, 18, 20, 21, 27, 29, 31, 48). These findings suggest pharmacological activity when targeting the CB1 receptor but also address the physiological importance of the endogenous cannabinoid system in regulation of GI motility. Antagonists of the CB1 receptor, when given alone, exert effects opposing agonist effects, suggesting permanent physiological activity of CB1 receptors.
Presently, most in vitro and in vivo evidence of CB1 involvement within the regulation of GI motility comes from studies in ileum though for GI pathophysiologies (e.g., diarrhea, irritable bowel syndrome) the colonic motility seems to be the more attractive pharmacological target. Presently, the possible CB1 receptor involvement in the myenteric circuits controlling colonic propulsion is poorly understood, and the individual cell types involved have not yet been identified.
Therefore, the aim of the present study was to localize CB1 receptors within the myenteric plexus of the colon and identify possible colocalization with other cell types involved in controlling colonic motility [myenteric neurons, interstitial cells of Cajal (ICC), and glial cells]. Furthermore, our aim was to characterize the influence of CB1 agonists and antagonists on electrophysiological phenomena underlying the neuronal circuits of the ascending myenteric pathways in mouse colon and characterize CB1 involvement beyond pure reduction of presynaptic transmitter release at a neuromuscular site. Furthermore, using a partitioned electrophysiological chamber, this study additionally intends to prove for the first time a functional influence of CB1 receptors on myenteric interneurons. Finally, we aimed to prove the concepts elaborated in the in vitro experiments in an established in vivo procedure of measuring colonic bead propulsion in wild-type and CB1−/− mice.
MATERIALS AND METHODS
Analysis of rat/mouse mRNA expression.
Total RNA was extracted using the RNeasy Mini kit (Qiagen, Valencia, CA) according to the manufacturer's protocol and converted to cDNA using the SuperScript first-strand synthesis system for RT-PCR (Invitrogen, Carlsbad, CA). PCR reactions containing 2 μl of cDNA template were amplified in a final volume of 25 μl by denaturation at 94°C for 2 min, followed by 35 cycles (CB1 receptor), respectively, 30 cycles (GAPDH) of amplification (94°C for 30 s, 58°C for 50 s, and 72°C for 1 min), and an extension at 72°C for 3 min. The following forward (f) and reverse (r) primers were used: CB1 (f), 5′-GTACCATCACCACAGACCTCCTC-3′; CB1 (r), 5′-GGATTCAGAATCATGAAGCACTCCA-3′; GAPDH (f), 5′-CCCTTCATTGACCTCAACTA-3′; GAPDH (r), 5′-AGATCCACGACGGACACATT-3′.
Tissue preparation for electrophysiological experiments.
Male mice (C57BL/6) were euthanized by cervical dislocation as approved by the government of the state of Bavaria (approval: 209.1/211—2531-2—41/01). The complete large bowel was removed through an abdominal midline incision and placed in oxygenated Krebs solution 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. The colon was then opened along the mesenteric border, washed of remaining fecal material, and pinned out in a Sylgard-lined (Dow Corning, Midland, MI) dissecting dish containing oxygenated Krebs solution. The mucosa and submucosa were removed, resulting in sheets of tissue consisting of circular and longitudinal muscle layers together with the attached myenteric plexus. Four-centimeter-long segments of proximal colon were pinned using ∼150—200 micropins (15 or 25 μm thick) obtained from tungsten wire to the Sylgard-based electrophysiological chamber, with the circular muscle layer on top. In additional experiments, the electrophysiological chamber was partitioned into two chambers after tissue was pinned as described above. The chambers were separated by a baffle and sealed. At the end of the experiment, leak tightness of the partitioning was checked by adding methylene blue on one side of the baffle. The chambers were constantly perfused (5 ml/min; Kwik Pump; World Precision Instruments, Sarasota, FL) with oxygenated (95% O2-5% CO2) Krebs solution (37°C). Tissue strips were then allowed to equilibrate for 90 min before the onset of experiments.
Intracellular electrical recording.
Intracellular recordings of smooth muscle cells of the circular muscle layer of mouse proximal colon were performed as described previously (39, 42). Guanethidine and nifedipine (both 1 μM) were present throughout all experiments. Capillary glass microelectrodes (borosilicate glass capillaries; 1.0-mm outer diameter × 0.58 mm inner diameter; Clark Electromedical Instruments, Kent, UK) were produced using a microelectrode puller (Model P-97, 3 mm-wide filament; Sutter Instrument, Novato, CA), filled with KCl (3 M), and had resistances in the range of 80—120 MΩ. Neurons were stimulated with single pulses (15 V; duration 0.3 ms) via multiple platinum electrodes arranged perpendicularly to the circular muscle layer and connected to a Grass S11 stimulator via a stimulus isolation unit (Grass SIU59; Grass Instruments, Quincy, MA). Spatial spreading of the electrophysiological responses was evaluated by applying electrical stimulation (ES) at multiple distances distal from the recording site (12 electrodes, 1 mm each, 0.7 mm apart from each other; distance covered by the 12 electrodes was 20 mm). In the partitioned electrophysiological chamber, the stimulation electrode in chamber 1 (stimulation chamber) and the recording electrode in chamber 2 (recording chamber) were 12 mm apart. Drugs were added to the caudal stimulation chamber. Following each experiment, the density of the organ bath separators was tested by adding methylene blue to one chamber.
ES of the neurons caused changes in the resting membrane potentials (RMPs) of the smooth muscle cells, and 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) 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 using the LABVIEW 5.0 program (National Instruments).
Colonic bead propulsion test.
For colonic bead propulsion testing, male CB1−/− mice (49) or wild-type littermates in C57BL/6 background, weighing 20—22 g, were used after an overnight fasting period with free access to water. Distal colonic propulsion was measured according to previously published methods (4, 32). Thirty minutes after intraperitoneal administration of drugs (or vehicle), a single 3-mm glass bead (prewarmed to 37°C) was inserted 3 cm into the distal colon of each mouse. The time to propulsion of the glass bead was determined for each animal. A higher mean propulsion time value is an index of stronger inhibition of colonic propulsion.
Immunohistochemistry.
For immunohistochemistry, whole mounts of proximal and distal colon of C57BL/6 mice were used. Colonic segments were first ligated and filled with physiological solution. Immersion fixation was done for 2 h at room temperature in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.0) followed by several rinses in 0.01 M phosphate-buffered saline (PBS, pH 7.4). Whole mounts were prepared as described previously (44). Briefly, following fixation and clearing, the tissues were dissected into two layers. In this way, the outer musculature with adhering serosa was separated from the submucosa/mucosa. The circular muscle was then removed to yield whole mounts of longitudinal muscle with the myenteric plexus attached.
All primary and secondary antibodies were diluted in PBS containing 10% normal serum, 0.01% bovine serum albumin, 0.05% thimerosal, and 0.01% sodium azide (designated as PBS*). Tissues were preincubated in PBS containing 1% Triton X-100 for 1 h and subsequently incubated overnight with the first primary polyclonal antibody. All immunohistochemical incubations were carried out at room temperature, and subsequent incubations were performed as detailed previously (43). To allow combination of two primary antisera raised in rabbit, binding sites of the first primary antibody were blocked by Cy3-conjugated Fab fragments of goat-anti-rabbit IgG (3-h incubation) and unlabeled Fab fragments of goat-anti-rabbit IgG (3-h incubation) (Fab-blocking procedure as described in Ref. 6). The second primary antibody was detected using biotin-conjugated donkey-anti-rabbit IgG and visualized by FITC-conjugated streptavidin. The antisera and streptavidin complexes used in this study as well as their respective dilutions are listed in Table 1. Negative controls in which one of the primary antibodies was omitted and cross-reactivity control stainings were performed as described by others (38).
Drugs.
Atropine, guanethidine, Nω-nitro-L-arginine, hexamethonium, nifedipine, and tetrodotoxin (TTX) were obtained from Sigma (Irvine, UK), and WIN 55,212-2 was obtained from Biotrend (Cologne, Germany). SR141716A was a gift from Sanofi-Syntelabo-Research (Montpellier, France).
Data presentation and statistical analysis.
Results are expressed as mean ± SE and were compared by using Student's t-test followed by Dunnett's post hoc test, using a commercial statistical package (SigmaStat, Jandel Scientific; San Rafael, CA). P < 0.05 was considered as statistically significant; n indicates the number of independent observations in individual animals.
RESULTS
CB1 receptor RNA expression in wild-type and CB1−/− mice.
CB1 receptor mRNA expression was confirmed in the colon of wild-type mice by RT-PCR. CB1 mRNA was detected in both the mucosa and longitudinal muscle-myenteric plexus preparations. Brain tissue was used as a positive control. CB1−/− mice lacked CB1 mRNA in all investigated regions (Fig. 1).
Intracellular recording of EJP and IJP.
Under basal conditions, circular smooth muscle cells of mouse proximal colon displayed stable RMP (−55.6 ± 5.3 mV, n = 17). ES of colonic myenteric neurons using the above-mentioned parameters gave rise to TTX-sensitive excitatory junction potentials (EJP) followed by biphasic inhibitory junction potentials (IJP). These IJP can be further subdivided into an initial fast (fIJP) and a following slow (sIJP) component. The amplitudes of the potentials were measured in millivolts (mV) compared with the membrane potential before application of the electrical stimulus. TTX (3 μM) addition in the recording chamber resulted in a significant depolarization of the membrane potential (prior, −55.8 ± 5.5 mV; TTX, −44.3 ± 4.4 mV, n = 5) and abolished all junction potentials (data not shown). The cholinergic blocker atropine (1 μM) selectively abolished the EJP, whereas RMP, fIJP, and sIJP remained unchanged, as shown previously (42). The blocker of NO synthase NG-nitro-L-arginine (100 μM) caused abolition of the sIJP, whereas RMP, EJP, and fIJP remained unchanged as shown previously (42).
Spatial distribution of reflex responses.
Upon stimulation at increasing distances from the recording site, the EJP increased up to stimulation site 6 (SE 6; distance, 10 mm), then decreased as from further distant stimulation sites (SE 7—11) and was hardly discernible anymore at SE 12 (distance, 20 mm) (Fig. 2). In contrast, the fIJP was maximal at SE 1, rapidly decreased up to SE 7 and was not seen at further distances (Fig. 2). The sIJP was maximal at SE 1 and slightly decreased up to SE 6. No sIJP were observed after SE 7 (Fig. 2).
Addition of the blocker of ganglionic transmission hexamethonium (100 μM) significantly reduced the EJP at distances >SE 3 (5 mm), and at distances >10 mm no EJP was observed [SE 1: control 9.6 ± 3.7 mV, hexamethonium (100 μM) 6.8 ± 2.5 mV; SE 3: control 12.5 ± 4.2 mV, hexamethonium (100 μM) 6.8 ± 2.1 mV; SE 5: control 21.2 ± 4.9 mV, hexamethonium (100 μM) 6.5 ± 2.8 mV*; SE 8: control 22.3 ± 4.5 mV, hexamethonium (100 μM) EJP abolished*; n = 4]. In presence of hexamethonium (100 μM), the course of fIJP and sIJP appeared unchanged, terminating at distances representing SE 6 (data not shown).
Influence of the CB1 receptor agonist WIN 55,212-2.
The CB1 receptor agonist WIN 55,212-2 was added to the organ bath in increasing concentrations (100 nM-10 μM). WIN 55,212-2 presence resulted in a decrease in the cholinergic excitatory EJP responses to ES in a concentration-dependent manner (Figs. 3 and 4). For SE 1—6, this decrease represented CB1 activity on the motor neurons, whereas for SE >7 the effects pertained to influences on interneurons. Additionally, the distance until which EJP could be recorded was shortened in a concentration-dependent manner, suggesting that CB1 receptor agonism depressed excitatory cholinergic responses not only in terms of amplitude but also in terms of their spatial propagation within the ascending part of the myenteric reflexes (Figs. 3 and 4). The amplitude and propagation interval of fIJP and sIJP remained unchanged (Fig. 3).
Influence of CB1 receptor antagonist SR141716A.
The CB1 receptor antagonist SR141716A (10 nM-1 μM) left the EJP, fIJP, and sIJP unchanged (Figs. 4 and 5). In the presence of SR1417161A (1 μM), WIN 55,212-2 (10 μM) was without significant effects, demonstrating that the effects of WIN 55,212-2 were mediated via the CB1 receptor (Fig. 5).
Influence of TTX, hexamethonium, atropine, and WIN 55,212-2 in the partitioned electrophysiological chamber.
TTX (3 μM) added to the stimulation chamber abolished the EJP in the recording chamber (data not shown). Hexamethonium (100 μM) added to the stimulation chamber significantly reduced the EJP in the recording chamber [control 25.1 ± 3.6 mV, hexamethonium (100 μM) 20.6 ± 3.8* mV; n = 5, *P < 0.05]. Atropine (1 μM) added to the stimulation chamber did not alter the EJP in the recording chamber (data not shown).
WIN 55,212-2 was added to the stimulation chamber in increasing concentrations (100 nM-10 μM) and significantly reduced the EJP in the anal recording chamber [control 27.2 ± 4.1 mV, WIN 55,212-2 (1 μM) 23.1 ± 3.1* mV, WIN 55,212-2 (10 μM) 21.3 ± 3.6 mV*; n = 6, *P < 0.05].
Effect of cannabinoid receptor agonist WIN 55,212-2 and antagonist SR141716A on colonic bead propulsion in vivo.
Intraperitoneal administration of WIN 55,212-2 (0.1—3 mg/kg) produced a dose-dependent inhibition of colonic propulsion in wild-type mice (Fig. 6A). This effect was not observed in CB1−/− mice (Fig. 6A). Additionally, WIN 55,212-2 had no effect in wild-type mice pretreated with the CB1 receptor-specific antagonist SR141716A (Fig. 6B), whereas SR141716A increased in vivo transit in wild-type mice, stressing the notion that the CB1 receptor is physiologically involved in colonic motility. SR141716A had no significant effect when given to CB1−/− mice (Fig. 6B).
Immunocytochemical identification of CB1 receptor-expressing myenteric neurons.
In line with the RT-PCR data, single immunolabeling and double staining with the type III intermediate filament protein peripherin showed that a large number of the myenteric neurons in mouse colon expressed immunoreactivity for the CB1 receptor (Fig. 7, A—F). This CB1 receptor labeling was observed at the level of neuronal somata (Fig. 7A) and at the level of axonal fibers running within the myenteric network and within the circular and longitudinal muscle layer (Fig. 7, B and C). Both smaller sized multidendritic neurons, part of them clearly resembling a Dogiel type I morphology, and larger sized multiaxonal neurons resembling a Dogiel type II morphology express the CB1 receptor (Fig. 7, D—F). Double-labeling experiments showed that all CB1-expressing neurons display the cholinergic marker choline acetyltransferase (ChAT) (Fig. 7, G—I), indicating that CB1-immunopositive neurons are cholinergic in nature. Double staining with the nitrergic marker (neuronal nitric oxide synthase) revealed that the large majority of the myenteric nitrergic neurons were apposed by CB1-immunoreactive nerve fibers but did not show CB1 staining (Fig. 7, J—L). A reverse image was seen after double staining with the calcium-binding protein calretinin. The large majority of calretinin-immunoreactive myenteric neurons costain for CB1. Part of these neurons, resembling a Dogiel type II morphology, are to be considered as intrinsic primary afferent neurons. The remaining of the CB1+/calretinin+ neurons have been previously assumed to function either as excitatory motor neurons or as ascending interneurons (Figs. 8, A—C). Double labeling with 5-HT, a neurotransmitter present in descending interneurons (33, 34), demonstrated that all 5-HT-immunoreactive neurons costain for CB1 (Fig. 8, D—F).
Within the muscle layers, occasionally CB1-immunoreactive, spindle-like cells were seen lying in close association with nerve bundles (Fig. 8, G—L). C-kit immunostaining clearly revealing ICC of myenteric plexus (ICC-MP) and intramuscular ICC (ICC-IM) did not show colocalization with these CB1-immunopositive spindle-shaped cells (Fig. 8, G—I). Double staining with glial fibrillary acidic protein (GFAP) (Fig. 8, J—L) or S100-protein (data not shown) was negative in terms of CB1 colabeling.
DISCUSSION
Intestinal peristalsis is coordinated by a perfect interplay of numerous structures including the enteric nervous system, ICC (pacemaker cells), glial cells, and smooth muscle cells (36). Within the enteric nervous system, interneurons coordinate transmission and modulation of neuronal impulses before a motor neuron transmits the impulse to circular and longitudinal smooth muscle cells (2, 5). Functional experiments and immunohistochemical studies have suggested CB1 receptor presence in the mouse colon (42), and by employing RT-PCR we now add that CB1 receptors are expressed throughout the colon. The data demonstrating that, apart from colonic epithelial cells, the CB1 receptor is clearly present in enteric neurons are in line with earlier observations in human colon (46).
A neuronal location of the CB1 receptor was suggested in numerous studies using functional organ bath preparations (10, 17, 23, 29). Pharmacological studies in different species rule out a smooth muscular localization of CB1 receptors, and our morphological observations strongly support this (23, 24, 31, 41). However, organ bath preparations and more complex preparations, such as the peristaltic reflex model and in vivo measurement of intestinal propulsion, are unable to functionally differentiate neuronal sites from motor neuronal and/or interneuronal sites (8, 14, 19, 32, 42). Our immunohistochemical data add that CB1 receptors are located on specific populations of cholinergic myenteric neurons, some of them resembling a clear Dogiel type I morphology, whereas others resemble intrinsic primary afferent neurons with Dogiel type II morphology, and the calretinin and nitric oxide synthase (NOS) staining suggests the presence CB1 receptors on excitatory motor neurons and ascending interneurons but not on nitrergic neurons. The 5-HT staining reveals that descending interneurons also contain CB1 receptors but presently cannot be used to distinguish ascending or descending interneurons (28, 33, 34, 47).
The present study for the first time shows involvement of CB1 receptors in electrophysiological circuits within the ascending myenteric neuronal system of the mouse colon and extends present understanding where CB1 involvement in myenteric neurotransmission seems to be limited to presynaptic neurotransmitter release at the neuromuscular junction. Addition of the CB1 receptor agonist WIN 55,212-2 reduced the EJP, which represents excitatory cholinergic neurotransmission (39, 42), in a concentration-dependent manner in all sites of stimulation. The reduction in amplitude caused by WIN 55,212-2 at the shorter distances (SE 1—6) represented an action on CB1 receptors located on motor neurons. Additionally, shortening of the overall transmission distance of the EJP resulted from CB1 receptor actions on interneurons, in which CB1 receptor activation reduced cholinergic excitatory neurotransmission and spreading of the signal. The finding that all these WIN 55,212-2 effects can be reversed by the specific CB1 receptor antagonist SR141716A clearly shows that the effects are mediated through CB1 receptors.
CB1 receptors located on myenteric neurons reduce smooth muscular excitation by decreasing neurotransmitter release (1, 11, 25, 37). However, the shortened transmission distances reported here are suggestive of other neuronal mechanisms, e.g., reduced neuronal excitability as an underlying mechanism. Reduction of neuronal excitability was recently attributed to CB1 receptor activation, showing that CB1 receptor activation decreases excitatory synaptic potentials and neuronal excitability in enteric S neurons but not in AH neurons of guinea pig ileum (22). Since AH neurons are believed to correspond to sensory neurons and S neurons to motor neurons, these findings are in good agreement with our observations. However, since S neurons include both motor neurons and interneurons (3, 5, 13), these electrophysiological recordings could not distinguish between these two neuronal types.
The experiments involving hexamethonium indicated that ganglionic transmission is involved in transmission of the EJP at distances further than SE 3 (5 mm) and obligatorily involved at distances >10 mm, beyond which EJP are abolished by hexamethonium. In our experiments, the EJP at the stimulation sites (SE 1—7) was significantly reduced, and, since this reduction was higher than the remaining amplitude in the presence of hexamethonium, the CB1 receptor modulated cholinergic excitatory neurotransmission at an additional site located on motor neurons. The CB1 receptor-specific antagonist SR141716A does not change the observed electrophysiological parameter but antagonizes the effect of WIN 55,212-2, further demonstrating that the observed effects are mediated by CB1 receptors located on interneurons and motor neurons. Additionally, using a partitioned electrophysiological recording chamber, we prove that CB1 receptor activation reduces myenteric interneuron neurotransmission to an extent seen with the ganglionic blocker hexamethonium and thereby for the first time provide evidence that CB1 receptors are functionally located on interneurons of the myenteric plexus and not only on motor neurons of the neuromuscular junction.
Immunohistochemistry presently lacks specific markers to distinguish motor neurons from interneurons. In the mouse colon, calretinin appears to display a widespread neuronal distribution and is found in multidendritic uniaxonal cholinergic excitatory motor neurons, ascending interneurons, and intrinsic primary afferent neurons resembling a type II morphology (12, 33, 35, 45). In our stainings, double labeling clearly indicated that in the myenteric plexus of the mouse colon CB1 is colocalized with calretinin, both at the level of perikarya and nerve fibers, whereas the large majority of the NOS-positive neurons were not coexpressing CB1 immunoreactivity. In agreement with other regions of the GI tract, CB1-positive neurons are ChAT positive, demonstrating that they are cholinergic in nature (9). Furthermore, the immunostainings provided evidence that CB1 was localized at individual nerve fibers of both intramuscular and interganglionic fiber bundles, further stressing the notion that CB1 is involved in neuronal circuits.
IJPs, which correspond to inhibitory neurotransmission, were observed in the ascending direction up to SE 7. We did not further analyze the IJP responses since they were most likely due to retrograde neuronal activation, as strengthened by the insensitivity of the IJP to hexamethonium. In further agreement with this notion was the immunohistochemical observation that NOS and CB1 did not colabel.
Additional CB1 receptors, which might be located on cells interfering with the complex phenomena of enteric neurotransmission, such as ICC or glial cells, can presently not be ruled out because of the lack of specific pharmacological tools. However, our double stainings for CB1 and c-kit ruled out the presence of CB1 receptors on ICC-MP and ICC-IM, and our double stainings with GFAP or S100 protein made the presence of CB1 receptors in glial cells unlikely in the mouse colon.
The in vitro experiments showed that activation of CB1 receptors in mouse colon reduced the ascending part of the myenteric reflex pathways. This observation suggested that colonic propulsion should be reduced in vivo. Using an established model of colonic bead propulsion measurement, we showed that this holds true for the in vivo situation and demonstrated that WIN 55,212-2 dose dependently inhibits colonic propulsion. This strengthens the electrophysiological concepts of CB1 receptor involvement for the translation to the in vivo situation. CB1 receptor activation resulted in inhibition of the electrophysiological circuits within the peristaltic reflex, which in turn caused inhibition of colonic propulsion. To disclose the nature of the involved receptor, we showed the lack of WIN 55,212-2 in CB1−/− mice and the lack of effect in wild-type mice after treatment with the CB1 receptor antagonist SR141716A, all in agreement with the notion that the CB1 receptor is involved. It has to be kept in mind that other than testing of small intestinal transit for ileum motility, colonic bead expulsion is an in vivo technique that involves mechanisms beyond motility phenomena. Colonic expulsion is amongst others influenced by rectal perception, central effects, secretion, willingness, and motility including propulsion and sphincter opening. Therefore, though the in vivo experiments using the CB1 receptor agonist WIN 55,212-2 are in line with the effects suggested from the in vitro experiments, it has to be considered that other phenomena are influenced in the in vivo experiments as well. The fact that in vivo propulsion is influenced by multiple mechanisms may explain why SR141716A increases colonic propulsion in vivo but does not influence spatial spreading of EJP in the myenteric plexus. Since SR141716A does not influence spatial spreading of EJP, other mechanisms like rectal perception, secretion, willingness, or central regulatory mechanisms may account for the effects of SR141716A on colonic propulsion. Since at higher concentrations SR141716A also elicits inverse agonist and nonspecific effects, no higher concentrations were used to clarify effects of such concentrations on spatial spreading of EJP (30, 40). The additional observation that CB1 antagonism in wild-type mice causes an increase of propulsion time, whereas, in CB1−/− mice, propulsion time is not increased, is noteworthy, and may result from chronic adaptation processes in CB1−/− mice that are absent when acute treatment with SR141716A is performed in wild-type mice.
In summary, our experiments demonstrate that CB1 receptors are physiologically involved in excitatory enteric neurotransmission circuits underlying the intestinal peristaltic reflex. For the cholinergic ascending excitatory pathways, we provide evidence that the involved CB1 receptors are located on ascending interneurons and on motor neurons. Immunohistochemistry rules out additional involved sites like ICC or glial cells. It should, however, be noted that spindle-shaped CB1-immunopositive c-kit/GFAP/S100 immunonegative cells were encountered within the outer muscle layer, and the origin of these cells is still unclear. By measuring the colonic propulsion in vivo we have proven that activation of the CB1 receptor results in delayed colonic propulsion, as suggested by our in vitro experiments. Thus our findings further strengthen the notion that the CB1 receptor is a promising pharmacological target for the treatment of colonic motor disturbances.
Source, Graphs and Figures: Cannabinoid-1 (CB1) receptors regulate colonic propulsion by acting at motor neurons within the ascending motor pathways in mouse colon
Cannabinoid-1 (CB1) receptors on myenteric neurons are involved in the regulation of intestinal motility. Our aim was to investigate CB1 receptor involvement in ascending neurotransmission in mouse colon and to characterize the involved structures by functional and morphological means. Presence of the CB1 receptor was investigated by RT-PCR, and immunohistochemistry was used for colabeling studies. Myenteric reflex responses were initiated by electrical stimulation (ES) at different distances, and junction potentials (JP) were recorded from circular smooth muscle cells by intracellular recording in an unpartitioned and a partitioned recording chamber. In vivo colonic propulsion was tested in wild-type and CB1−/− mice. Immunostaining with the cytoskeletal marker peripherin showed CB1 immunoreactivity both on Dogiel type I and type II neurons. Further neurochemical characterization revealed CB1 on choline acetyltransferase-, calretinin-, and 5-HT-immunopositive myenteric neurons, but nitrergic neurons appeared immunonegative for CB1 immunostaining. Solitary spindle-shaped CB1-immunoreactive cells in between smooth muscle cells lacked specific markers for interstitial cells of Cajal or glial cells. ES elicited neuronally mediated excitatory JP (EJP) and inhibitory JP. Gradual increases in distance resulted in a wave-like EJP with EJP amplitudes being maximal at the location of stimulating electrode 6 and a maximal EJP projection distance of ∼18 mm. The CB1 receptor agonist WIN 55,212-2 reduced the amplitude of EJP and was responsible for shortening the oral spreading of the excitatory impulse. In a partitioned chamber, WIN 55,212-2 reduced EJP at the separated oral sites, proving that CB1 activation inhibits interneuron-mediated neurotransmission. These effects were absent in the presence of the CB1 antagonist SR141716A, which, when given alone, had no effect. WIN 55,212-2 inhibited colonic propulsion in wild-type mice but not in SR141716A-pretreated wild-type or CB1−/− mice. Activation of the CB1 receptor modulates excitatory cholinergic neurotransmission in mouse colon by reducing amplitude and spatial spreading of the ascending electrophysiological impulses. This effect on electrophysiological spreading involves CB1-mediated effects on motor neurons and ascending interneurons and is likely to underlie the here reported in vivo reduction in colonic propulsion.
ENDOCANNABINOIDS AND THEIR respective receptors are present throughout the gastrointestinal (GI) tract (7, 9). Whereas the cannabinoid-2 (CB2) receptor is mainly located on immune cells and its involvement in the regulation of GI motility remains speculative, there is strong evidence that cannabinoid-1 (CB1) receptors exert such roles (17, 26, 29).
In vitro experiments in various GI regions, including stomach and small and large intestine, in different species have proven that activation of the CB1 receptor by endogenous or exogenous agonists modulates motility mainly by reducing contractile activity (23, 29, 31, 41). These effects involve reduction in neurotransmitter release following activation of CB1 receptors located on a presynaptic site of the motorneuron (1, 25, 37). To date it is unknown whether this presynaptic location involves peripheral motor neurons, interneurons, or both, and it is unknown to what extent CB1 receptors are also involved in the regulation of neuronal circuits of intestinal motility rather than reduction of neurotransmitter release at the neuromuscular junction only. Beside the effects on cholinergic excitatory neurotransmission, additional CB1-mediated effects on excitatory tachykininergic and inhibitory nonadrenergic, noncholinergic neurotransmission were previously reported in the GI tract (16, 23, 41).
In more complex preparations, the cannabinoid system potently reduced small intestinal peristaltic activity and colonic bead propulsion, where activation of the CB1 receptor caused a decrease and application of respective receptor antagonists and increase in peristalsis and propulsion (14, 15, 18, 20, 21, 27, 29, 31, 48). These findings suggest pharmacological activity when targeting the CB1 receptor but also address the physiological importance of the endogenous cannabinoid system in regulation of GI motility. Antagonists of the CB1 receptor, when given alone, exert effects opposing agonist effects, suggesting permanent physiological activity of CB1 receptors.
Presently, most in vitro and in vivo evidence of CB1 involvement within the regulation of GI motility comes from studies in ileum though for GI pathophysiologies (e.g., diarrhea, irritable bowel syndrome) the colonic motility seems to be the more attractive pharmacological target. Presently, the possible CB1 receptor involvement in the myenteric circuits controlling colonic propulsion is poorly understood, and the individual cell types involved have not yet been identified.
Therefore, the aim of the present study was to localize CB1 receptors within the myenteric plexus of the colon and identify possible colocalization with other cell types involved in controlling colonic motility [myenteric neurons, interstitial cells of Cajal (ICC), and glial cells]. Furthermore, our aim was to characterize the influence of CB1 agonists and antagonists on electrophysiological phenomena underlying the neuronal circuits of the ascending myenteric pathways in mouse colon and characterize CB1 involvement beyond pure reduction of presynaptic transmitter release at a neuromuscular site. Furthermore, using a partitioned electrophysiological chamber, this study additionally intends to prove for the first time a functional influence of CB1 receptors on myenteric interneurons. Finally, we aimed to prove the concepts elaborated in the in vitro experiments in an established in vivo procedure of measuring colonic bead propulsion in wild-type and CB1−/− mice.
MATERIALS AND METHODS
Analysis of rat/mouse mRNA expression.
Total RNA was extracted using the RNeasy Mini kit (Qiagen, Valencia, CA) according to the manufacturer's protocol and converted to cDNA using the SuperScript first-strand synthesis system for RT-PCR (Invitrogen, Carlsbad, CA). PCR reactions containing 2 μl of cDNA template were amplified in a final volume of 25 μl by denaturation at 94°C for 2 min, followed by 35 cycles (CB1 receptor), respectively, 30 cycles (GAPDH) of amplification (94°C for 30 s, 58°C for 50 s, and 72°C for 1 min), and an extension at 72°C for 3 min. The following forward (f) and reverse (r) primers were used: CB1 (f), 5′-GTACCATCACCACAGACCTCCTC-3′; CB1 (r), 5′-GGATTCAGAATCATGAAGCACTCCA-3′; GAPDH (f), 5′-CCCTTCATTGACCTCAACTA-3′; GAPDH (r), 5′-AGATCCACGACGGACACATT-3′.
Tissue preparation for electrophysiological experiments.
Male mice (C57BL/6) were euthanized by cervical dislocation as approved by the government of the state of Bavaria (approval: 209.1/211—2531-2—41/01). The complete large bowel was removed through an abdominal midline incision and placed in oxygenated Krebs solution 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. The colon was then opened along the mesenteric border, washed of remaining fecal material, and pinned out in a Sylgard-lined (Dow Corning, Midland, MI) dissecting dish containing oxygenated Krebs solution. The mucosa and submucosa were removed, resulting in sheets of tissue consisting of circular and longitudinal muscle layers together with the attached myenteric plexus. Four-centimeter-long segments of proximal colon were pinned using ∼150—200 micropins (15 or 25 μm thick) obtained from tungsten wire to the Sylgard-based electrophysiological chamber, with the circular muscle layer on top. In additional experiments, the electrophysiological chamber was partitioned into two chambers after tissue was pinned as described above. The chambers were separated by a baffle and sealed. At the end of the experiment, leak tightness of the partitioning was checked by adding methylene blue on one side of the baffle. The chambers were constantly perfused (5 ml/min; Kwik Pump; World Precision Instruments, Sarasota, FL) with oxygenated (95% O2-5% CO2) Krebs solution (37°C). Tissue strips were then allowed to equilibrate for 90 min before the onset of experiments.
Intracellular electrical recording.
Intracellular recordings of smooth muscle cells of the circular muscle layer of mouse proximal colon were performed as described previously (39, 42). Guanethidine and nifedipine (both 1 μM) were present throughout all experiments. Capillary glass microelectrodes (borosilicate glass capillaries; 1.0-mm outer diameter × 0.58 mm inner diameter; Clark Electromedical Instruments, Kent, UK) were produced using a microelectrode puller (Model P-97, 3 mm-wide filament; Sutter Instrument, Novato, CA), filled with KCl (3 M), and had resistances in the range of 80—120 MΩ. Neurons were stimulated with single pulses (15 V; duration 0.3 ms) via multiple platinum electrodes arranged perpendicularly to the circular muscle layer and connected to a Grass S11 stimulator via a stimulus isolation unit (Grass SIU59; Grass Instruments, Quincy, MA). Spatial spreading of the electrophysiological responses was evaluated by applying electrical stimulation (ES) at multiple distances distal from the recording site (12 electrodes, 1 mm each, 0.7 mm apart from each other; distance covered by the 12 electrodes was 20 mm). In the partitioned electrophysiological chamber, the stimulation electrode in chamber 1 (stimulation chamber) and the recording electrode in chamber 2 (recording chamber) were 12 mm apart. Drugs were added to the caudal stimulation chamber. Following each experiment, the density of the organ bath separators was tested by adding methylene blue to one chamber.
ES of the neurons caused changes in the resting membrane potentials (RMPs) of the smooth muscle cells, and 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) 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 using the LABVIEW 5.0 program (National Instruments).
Colonic bead propulsion test.
For colonic bead propulsion testing, male CB1−/− mice (49) or wild-type littermates in C57BL/6 background, weighing 20—22 g, were used after an overnight fasting period with free access to water. Distal colonic propulsion was measured according to previously published methods (4, 32). Thirty minutes after intraperitoneal administration of drugs (or vehicle), a single 3-mm glass bead (prewarmed to 37°C) was inserted 3 cm into the distal colon of each mouse. The time to propulsion of the glass bead was determined for each animal. A higher mean propulsion time value is an index of stronger inhibition of colonic propulsion.
Immunohistochemistry.
For immunohistochemistry, whole mounts of proximal and distal colon of C57BL/6 mice were used. Colonic segments were first ligated and filled with physiological solution. Immersion fixation was done for 2 h at room temperature in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.0) followed by several rinses in 0.01 M phosphate-buffered saline (PBS, pH 7.4). Whole mounts were prepared as described previously (44). Briefly, following fixation and clearing, the tissues were dissected into two layers. In this way, the outer musculature with adhering serosa was separated from the submucosa/mucosa. The circular muscle was then removed to yield whole mounts of longitudinal muscle with the myenteric plexus attached.
All primary and secondary antibodies were diluted in PBS containing 10% normal serum, 0.01% bovine serum albumin, 0.05% thimerosal, and 0.01% sodium azide (designated as PBS*). Tissues were preincubated in PBS containing 1% Triton X-100 for 1 h and subsequently incubated overnight with the first primary polyclonal antibody. All immunohistochemical incubations were carried out at room temperature, and subsequent incubations were performed as detailed previously (43). To allow combination of two primary antisera raised in rabbit, binding sites of the first primary antibody were blocked by Cy3-conjugated Fab fragments of goat-anti-rabbit IgG (3-h incubation) and unlabeled Fab fragments of goat-anti-rabbit IgG (3-h incubation) (Fab-blocking procedure as described in Ref. 6). The second primary antibody was detected using biotin-conjugated donkey-anti-rabbit IgG and visualized by FITC-conjugated streptavidin. The antisera and streptavidin complexes used in this study as well as their respective dilutions are listed in Table 1. Negative controls in which one of the primary antibodies was omitted and cross-reactivity control stainings were performed as described by others (38).
Drugs.
Atropine, guanethidine, Nω-nitro-L-arginine, hexamethonium, nifedipine, and tetrodotoxin (TTX) were obtained from Sigma (Irvine, UK), and WIN 55,212-2 was obtained from Biotrend (Cologne, Germany). SR141716A was a gift from Sanofi-Syntelabo-Research (Montpellier, France).
Data presentation and statistical analysis.
Results are expressed as mean ± SE and were compared by using Student's t-test followed by Dunnett's post hoc test, using a commercial statistical package (SigmaStat, Jandel Scientific; San Rafael, CA). P < 0.05 was considered as statistically significant; n indicates the number of independent observations in individual animals.
RESULTS
CB1 receptor RNA expression in wild-type and CB1−/− mice.
CB1 receptor mRNA expression was confirmed in the colon of wild-type mice by RT-PCR. CB1 mRNA was detected in both the mucosa and longitudinal muscle-myenteric plexus preparations. Brain tissue was used as a positive control. CB1−/− mice lacked CB1 mRNA in all investigated regions (Fig. 1).
Intracellular recording of EJP and IJP.
Under basal conditions, circular smooth muscle cells of mouse proximal colon displayed stable RMP (−55.6 ± 5.3 mV, n = 17). ES of colonic myenteric neurons using the above-mentioned parameters gave rise to TTX-sensitive excitatory junction potentials (EJP) followed by biphasic inhibitory junction potentials (IJP). These IJP can be further subdivided into an initial fast (fIJP) and a following slow (sIJP) component. The amplitudes of the potentials were measured in millivolts (mV) compared with the membrane potential before application of the electrical stimulus. TTX (3 μM) addition in the recording chamber resulted in a significant depolarization of the membrane potential (prior, −55.8 ± 5.5 mV; TTX, −44.3 ± 4.4 mV, n = 5) and abolished all junction potentials (data not shown). The cholinergic blocker atropine (1 μM) selectively abolished the EJP, whereas RMP, fIJP, and sIJP remained unchanged, as shown previously (42). The blocker of NO synthase NG-nitro-L-arginine (100 μM) caused abolition of the sIJP, whereas RMP, EJP, and fIJP remained unchanged as shown previously (42).
Spatial distribution of reflex responses.
Upon stimulation at increasing distances from the recording site, the EJP increased up to stimulation site 6 (SE 6; distance, 10 mm), then decreased as from further distant stimulation sites (SE 7—11) and was hardly discernible anymore at SE 12 (distance, 20 mm) (Fig. 2). In contrast, the fIJP was maximal at SE 1, rapidly decreased up to SE 7 and was not seen at further distances (Fig. 2). The sIJP was maximal at SE 1 and slightly decreased up to SE 6. No sIJP were observed after SE 7 (Fig. 2).
Addition of the blocker of ganglionic transmission hexamethonium (100 μM) significantly reduced the EJP at distances >SE 3 (5 mm), and at distances >10 mm no EJP was observed [SE 1: control 9.6 ± 3.7 mV, hexamethonium (100 μM) 6.8 ± 2.5 mV; SE 3: control 12.5 ± 4.2 mV, hexamethonium (100 μM) 6.8 ± 2.1 mV; SE 5: control 21.2 ± 4.9 mV, hexamethonium (100 μM) 6.5 ± 2.8 mV*; SE 8: control 22.3 ± 4.5 mV, hexamethonium (100 μM) EJP abolished*; n = 4]. In presence of hexamethonium (100 μM), the course of fIJP and sIJP appeared unchanged, terminating at distances representing SE 6 (data not shown).
Influence of the CB1 receptor agonist WIN 55,212-2.
The CB1 receptor agonist WIN 55,212-2 was added to the organ bath in increasing concentrations (100 nM-10 μM). WIN 55,212-2 presence resulted in a decrease in the cholinergic excitatory EJP responses to ES in a concentration-dependent manner (Figs. 3 and 4). For SE 1—6, this decrease represented CB1 activity on the motor neurons, whereas for SE >7 the effects pertained to influences on interneurons. Additionally, the distance until which EJP could be recorded was shortened in a concentration-dependent manner, suggesting that CB1 receptor agonism depressed excitatory cholinergic responses not only in terms of amplitude but also in terms of their spatial propagation within the ascending part of the myenteric reflexes (Figs. 3 and 4). The amplitude and propagation interval of fIJP and sIJP remained unchanged (Fig. 3).
Influence of CB1 receptor antagonist SR141716A.
The CB1 receptor antagonist SR141716A (10 nM-1 μM) left the EJP, fIJP, and sIJP unchanged (Figs. 4 and 5). In the presence of SR1417161A (1 μM), WIN 55,212-2 (10 μM) was without significant effects, demonstrating that the effects of WIN 55,212-2 were mediated via the CB1 receptor (Fig. 5).
Influence of TTX, hexamethonium, atropine, and WIN 55,212-2 in the partitioned electrophysiological chamber.
TTX (3 μM) added to the stimulation chamber abolished the EJP in the recording chamber (data not shown). Hexamethonium (100 μM) added to the stimulation chamber significantly reduced the EJP in the recording chamber [control 25.1 ± 3.6 mV, hexamethonium (100 μM) 20.6 ± 3.8* mV; n = 5, *P < 0.05]. Atropine (1 μM) added to the stimulation chamber did not alter the EJP in the recording chamber (data not shown).
WIN 55,212-2 was added to the stimulation chamber in increasing concentrations (100 nM-10 μM) and significantly reduced the EJP in the anal recording chamber [control 27.2 ± 4.1 mV, WIN 55,212-2 (1 μM) 23.1 ± 3.1* mV, WIN 55,212-2 (10 μM) 21.3 ± 3.6 mV*; n = 6, *P < 0.05].
Effect of cannabinoid receptor agonist WIN 55,212-2 and antagonist SR141716A on colonic bead propulsion in vivo.
Intraperitoneal administration of WIN 55,212-2 (0.1—3 mg/kg) produced a dose-dependent inhibition of colonic propulsion in wild-type mice (Fig. 6A). This effect was not observed in CB1−/− mice (Fig. 6A). Additionally, WIN 55,212-2 had no effect in wild-type mice pretreated with the CB1 receptor-specific antagonist SR141716A (Fig. 6B), whereas SR141716A increased in vivo transit in wild-type mice, stressing the notion that the CB1 receptor is physiologically involved in colonic motility. SR141716A had no significant effect when given to CB1−/− mice (Fig. 6B).
Immunocytochemical identification of CB1 receptor-expressing myenteric neurons.
In line with the RT-PCR data, single immunolabeling and double staining with the type III intermediate filament protein peripherin showed that a large number of the myenteric neurons in mouse colon expressed immunoreactivity for the CB1 receptor (Fig. 7, A—F). This CB1 receptor labeling was observed at the level of neuronal somata (Fig. 7A) and at the level of axonal fibers running within the myenteric network and within the circular and longitudinal muscle layer (Fig. 7, B and C). Both smaller sized multidendritic neurons, part of them clearly resembling a Dogiel type I morphology, and larger sized multiaxonal neurons resembling a Dogiel type II morphology express the CB1 receptor (Fig. 7, D—F). Double-labeling experiments showed that all CB1-expressing neurons display the cholinergic marker choline acetyltransferase (ChAT) (Fig. 7, G—I), indicating that CB1-immunopositive neurons are cholinergic in nature. Double staining with the nitrergic marker (neuronal nitric oxide synthase) revealed that the large majority of the myenteric nitrergic neurons were apposed by CB1-immunoreactive nerve fibers but did not show CB1 staining (Fig. 7, J—L). A reverse image was seen after double staining with the calcium-binding protein calretinin. The large majority of calretinin-immunoreactive myenteric neurons costain for CB1. Part of these neurons, resembling a Dogiel type II morphology, are to be considered as intrinsic primary afferent neurons. The remaining of the CB1+/calretinin+ neurons have been previously assumed to function either as excitatory motor neurons or as ascending interneurons (Figs. 8, A—C). Double labeling with 5-HT, a neurotransmitter present in descending interneurons (33, 34), demonstrated that all 5-HT-immunoreactive neurons costain for CB1 (Fig. 8, D—F).
Within the muscle layers, occasionally CB1-immunoreactive, spindle-like cells were seen lying in close association with nerve bundles (Fig. 8, G—L). C-kit immunostaining clearly revealing ICC of myenteric plexus (ICC-MP) and intramuscular ICC (ICC-IM) did not show colocalization with these CB1-immunopositive spindle-shaped cells (Fig. 8, G—I). Double staining with glial fibrillary acidic protein (GFAP) (Fig. 8, J—L) or S100-protein (data not shown) was negative in terms of CB1 colabeling.
DISCUSSION
Intestinal peristalsis is coordinated by a perfect interplay of numerous structures including the enteric nervous system, ICC (pacemaker cells), glial cells, and smooth muscle cells (36). Within the enteric nervous system, interneurons coordinate transmission and modulation of neuronal impulses before a motor neuron transmits the impulse to circular and longitudinal smooth muscle cells (2, 5). Functional experiments and immunohistochemical studies have suggested CB1 receptor presence in the mouse colon (42), and by employing RT-PCR we now add that CB1 receptors are expressed throughout the colon. The data demonstrating that, apart from colonic epithelial cells, the CB1 receptor is clearly present in enteric neurons are in line with earlier observations in human colon (46).
A neuronal location of the CB1 receptor was suggested in numerous studies using functional organ bath preparations (10, 17, 23, 29). Pharmacological studies in different species rule out a smooth muscular localization of CB1 receptors, and our morphological observations strongly support this (23, 24, 31, 41). However, organ bath preparations and more complex preparations, such as the peristaltic reflex model and in vivo measurement of intestinal propulsion, are unable to functionally differentiate neuronal sites from motor neuronal and/or interneuronal sites (8, 14, 19, 32, 42). Our immunohistochemical data add that CB1 receptors are located on specific populations of cholinergic myenteric neurons, some of them resembling a clear Dogiel type I morphology, whereas others resemble intrinsic primary afferent neurons with Dogiel type II morphology, and the calretinin and nitric oxide synthase (NOS) staining suggests the presence CB1 receptors on excitatory motor neurons and ascending interneurons but not on nitrergic neurons. The 5-HT staining reveals that descending interneurons also contain CB1 receptors but presently cannot be used to distinguish ascending or descending interneurons (28, 33, 34, 47).
The present study for the first time shows involvement of CB1 receptors in electrophysiological circuits within the ascending myenteric neuronal system of the mouse colon and extends present understanding where CB1 involvement in myenteric neurotransmission seems to be limited to presynaptic neurotransmitter release at the neuromuscular junction. Addition of the CB1 receptor agonist WIN 55,212-2 reduced the EJP, which represents excitatory cholinergic neurotransmission (39, 42), in a concentration-dependent manner in all sites of stimulation. The reduction in amplitude caused by WIN 55,212-2 at the shorter distances (SE 1—6) represented an action on CB1 receptors located on motor neurons. Additionally, shortening of the overall transmission distance of the EJP resulted from CB1 receptor actions on interneurons, in which CB1 receptor activation reduced cholinergic excitatory neurotransmission and spreading of the signal. The finding that all these WIN 55,212-2 effects can be reversed by the specific CB1 receptor antagonist SR141716A clearly shows that the effects are mediated through CB1 receptors.
CB1 receptors located on myenteric neurons reduce smooth muscular excitation by decreasing neurotransmitter release (1, 11, 25, 37). However, the shortened transmission distances reported here are suggestive of other neuronal mechanisms, e.g., reduced neuronal excitability as an underlying mechanism. Reduction of neuronal excitability was recently attributed to CB1 receptor activation, showing that CB1 receptor activation decreases excitatory synaptic potentials and neuronal excitability in enteric S neurons but not in AH neurons of guinea pig ileum (22). Since AH neurons are believed to correspond to sensory neurons and S neurons to motor neurons, these findings are in good agreement with our observations. However, since S neurons include both motor neurons and interneurons (3, 5, 13), these electrophysiological recordings could not distinguish between these two neuronal types.
The experiments involving hexamethonium indicated that ganglionic transmission is involved in transmission of the EJP at distances further than SE 3 (5 mm) and obligatorily involved at distances >10 mm, beyond which EJP are abolished by hexamethonium. In our experiments, the EJP at the stimulation sites (SE 1—7) was significantly reduced, and, since this reduction was higher than the remaining amplitude in the presence of hexamethonium, the CB1 receptor modulated cholinergic excitatory neurotransmission at an additional site located on motor neurons. The CB1 receptor-specific antagonist SR141716A does not change the observed electrophysiological parameter but antagonizes the effect of WIN 55,212-2, further demonstrating that the observed effects are mediated by CB1 receptors located on interneurons and motor neurons. Additionally, using a partitioned electrophysiological recording chamber, we prove that CB1 receptor activation reduces myenteric interneuron neurotransmission to an extent seen with the ganglionic blocker hexamethonium and thereby for the first time provide evidence that CB1 receptors are functionally located on interneurons of the myenteric plexus and not only on motor neurons of the neuromuscular junction.
Immunohistochemistry presently lacks specific markers to distinguish motor neurons from interneurons. In the mouse colon, calretinin appears to display a widespread neuronal distribution and is found in multidendritic uniaxonal cholinergic excitatory motor neurons, ascending interneurons, and intrinsic primary afferent neurons resembling a type II morphology (12, 33, 35, 45). In our stainings, double labeling clearly indicated that in the myenteric plexus of the mouse colon CB1 is colocalized with calretinin, both at the level of perikarya and nerve fibers, whereas the large majority of the NOS-positive neurons were not coexpressing CB1 immunoreactivity. In agreement with other regions of the GI tract, CB1-positive neurons are ChAT positive, demonstrating that they are cholinergic in nature (9). Furthermore, the immunostainings provided evidence that CB1 was localized at individual nerve fibers of both intramuscular and interganglionic fiber bundles, further stressing the notion that CB1 is involved in neuronal circuits.
IJPs, which correspond to inhibitory neurotransmission, were observed in the ascending direction up to SE 7. We did not further analyze the IJP responses since they were most likely due to retrograde neuronal activation, as strengthened by the insensitivity of the IJP to hexamethonium. In further agreement with this notion was the immunohistochemical observation that NOS and CB1 did not colabel.
Additional CB1 receptors, which might be located on cells interfering with the complex phenomena of enteric neurotransmission, such as ICC or glial cells, can presently not be ruled out because of the lack of specific pharmacological tools. However, our double stainings for CB1 and c-kit ruled out the presence of CB1 receptors on ICC-MP and ICC-IM, and our double stainings with GFAP or S100 protein made the presence of CB1 receptors in glial cells unlikely in the mouse colon.
The in vitro experiments showed that activation of CB1 receptors in mouse colon reduced the ascending part of the myenteric reflex pathways. This observation suggested that colonic propulsion should be reduced in vivo. Using an established model of colonic bead propulsion measurement, we showed that this holds true for the in vivo situation and demonstrated that WIN 55,212-2 dose dependently inhibits colonic propulsion. This strengthens the electrophysiological concepts of CB1 receptor involvement for the translation to the in vivo situation. CB1 receptor activation resulted in inhibition of the electrophysiological circuits within the peristaltic reflex, which in turn caused inhibition of colonic propulsion. To disclose the nature of the involved receptor, we showed the lack of WIN 55,212-2 in CB1−/− mice and the lack of effect in wild-type mice after treatment with the CB1 receptor antagonist SR141716A, all in agreement with the notion that the CB1 receptor is involved. It has to be kept in mind that other than testing of small intestinal transit for ileum motility, colonic bead expulsion is an in vivo technique that involves mechanisms beyond motility phenomena. Colonic expulsion is amongst others influenced by rectal perception, central effects, secretion, willingness, and motility including propulsion and sphincter opening. Therefore, though the in vivo experiments using the CB1 receptor agonist WIN 55,212-2 are in line with the effects suggested from the in vitro experiments, it has to be considered that other phenomena are influenced in the in vivo experiments as well. The fact that in vivo propulsion is influenced by multiple mechanisms may explain why SR141716A increases colonic propulsion in vivo but does not influence spatial spreading of EJP in the myenteric plexus. Since SR141716A does not influence spatial spreading of EJP, other mechanisms like rectal perception, secretion, willingness, or central regulatory mechanisms may account for the effects of SR141716A on colonic propulsion. Since at higher concentrations SR141716A also elicits inverse agonist and nonspecific effects, no higher concentrations were used to clarify effects of such concentrations on spatial spreading of EJP (30, 40). The additional observation that CB1 antagonism in wild-type mice causes an increase of propulsion time, whereas, in CB1−/− mice, propulsion time is not increased, is noteworthy, and may result from chronic adaptation processes in CB1−/− mice that are absent when acute treatment with SR141716A is performed in wild-type mice.
In summary, our experiments demonstrate that CB1 receptors are physiologically involved in excitatory enteric neurotransmission circuits underlying the intestinal peristaltic reflex. For the cholinergic ascending excitatory pathways, we provide evidence that the involved CB1 receptors are located on ascending interneurons and on motor neurons. Immunohistochemistry rules out additional involved sites like ICC or glial cells. It should, however, be noted that spindle-shaped CB1-immunopositive c-kit/GFAP/S100 immunonegative cells were encountered within the outer muscle layer, and the origin of these cells is still unclear. By measuring the colonic propulsion in vivo we have proven that activation of the CB1 receptor results in delayed colonic propulsion, as suggested by our in vitro experiments. Thus our findings further strengthen the notion that the CB1 receptor is a promising pharmacological target for the treatment of colonic motor disturbances.
Source, Graphs and Figures: Cannabinoid-1 (CB1) receptors regulate colonic propulsion by acting at motor neurons within the ascending motor pathways in mouse colon
