Cannabinoid type 1 receptor modulates intestinal propulsion by an attenuation of intestinal motor responses within the myenteric part of the peristaltic reflex


PD Dr Martin Storr, Department of Internal Medicine II, University of Munich, Marchioninistr. 15, 81377 München, Germany.
Tel: +49 89 7095 2281; fax: +49 89 7095 5281;


Abstract  Cannabinoid-1 (CB1) receptor activation affects gastrointestinal propulsion in vivo. It was our aim to further characterize the involved myenteric mechanisms in vivo and in vitro. In CB1−/− mice and wild-type littermates we performed in vivo transit experiments by charcoal feeding and in vitro electrophysiological recordings in mouse small intestinal smooth muscle. Ascending neuronal contraction (ANC) following electrical field stimulation was studied in rat ileum in a partitioned organ bath separating the aboral stimulation site from the oral recording site. The knockout animals displayed an accelerated upper gastrointestinal transit compared to control animals. The CB1 receptor antagonist AM251 stimulated the force of the ANC in a concentration dependent manner when added in the oral chamber. Anandamide significantly inhibited the ANC when added in the oral chamber. Neither AM251 nor anandamide had an influence on the contraction latency. No effects were observed when drugs were added in the aboral chamber, proving a CB1 mediated action on the neuromuscular junction. Resting membrane potentials and neuronal induced inhibitory junction potentials in CB1−/− mice were unchanged as compared to wild type. However, the electrophysiological slow waves were more sensitive to blockade of Ca2+ channels in CB1−/− mice. Our data strongly suggest a physiological involvement of the CB-1 receptor in the regulation of small intestinal motility. Therefore, CB1 receptors are a promising target for the treatment of motility disorders.


Cannabis sativa has a long recreational but also a medical history. Different preparations have been used for a variety of disorders.1 Although some of these historically mentioned medical indications have to be questioned, there seems to be promising potential for cannabinoids, the active constituents of Cannabis sativa. Several years ago, an endogenous system for cannabinoids was identified, giving a functional basis for the effects of cannabinoids. This endocannabinoid system, which has been identified in various species, contains specific receptors, specific ligands and specific enzymes for synthesis and degradation of the endocannabinoids.2 There are at least two established cannabinoid receptors, termed cannabinoid-1 (CB1) and CB2, with a discrepant distribution. Cannabinoid-1 receptor is abundantly located on neuronal tissue of the central and peripheral nervous system, including the intestinal myenteric plexus,3 but also on other tissues such as colonic epithelial cells.4 In contrast, the CB2 receptor is mainly located on immune cells,3 however recently CB2 receptors were reported on central neurons.5 Several studies reported that CB1 mediated mechanisms affect motility of the intestine6–8 and are, therefore, in line with the historical use of Cannabis in gastrointestinal disorders such as, gastroenteritis and diarrhoea.1

In the myenteric and submucous plexus of embryonic rat intestine, CB1 mRNA and receptor-like immunreactivity was reported.9 For adult rats, this was confirmed for the small intestine, presently without a differentiation to the different bowel layers.10,11 Since further immunomorphological characterization in rat is lacking, investigations in porcine intestine have shown that CB1 receptors are in fact co-localized with choline acetyl transferase (ChAT)-positive neurons.12 This is consistent with functional studies in guinea pigs, where cannabinoids display inhibitory effects on cholinergic neurotransmission in smooth muscle preparations.6 In the whole gastrointestinal tract of mice, except of the pylorus, CB1 receptors were reported by immunochemistry and Western blotting.3,11,13

Presently, five different endocannabinoids with different affinities to the CB1 receptor are available.2 In the small intestine of rodents, high amounts of anandamide and 2-arachidonoyl glycerol (2-AG) exist, consistent with the notion that endocannabinoids play a role within gastrointestinal motility.14

Antagonists of CB1 receptors often show effects which are converse to CB1 receptor agonists.7,15–18 Whether this is due to an inverse agonism or the inhibition of a physiological active endocannabinoid system can be addressed in CB1−/− mice.19,20

Although CB1 receptors were localized in small intestine and in vivo and in vitro experiments strongly suggest a CB1 receptor involvement in neurotransmission and motility phenomena, an evaluation of the relevance in CB1/ mice has remained unanswered. Thus, our study aimed at evaluating the possible impact on small intestinal transit in CB1/ mice as compared with wild-type littermates. Additionally, to address the underlying mechanisms, the ascending neuronal contraction (ANC) following electrical field stimulation, being a part of the ascending reflex pathways of the myenteric part of the peristaltic reflex (ARPPR) were analysed in an in vitro setup and, the slow wave activity and junction potentials by intracellular recordings.


Analysis of rat/mouse mRNA expression

Total RNA was extracted using the RNeasy Mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol and converted into cDNA using the SuperScript first-strand synthesis system for reverse-transcription PCR (RT-PCR) (Invitrogen, Karlsruhe, Germany). 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 (CB1R), respectively, 30 cycles [glyceraldehyde-3-phosphate dehydrogenase (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. Table 1 gives details on the forward and reverse primers used.

Table 1.   Sequence of forward (f) and reverse (r) primers used in reverse-transcription PCR
SpeciesNameSequence 5′-3′
  1. CB1, cannabinoid-1; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.


Intestinal transit

Eight-week-old male mice were purchase from Charles River, Sulzbach, Germany. Gastrointestinal transit was measured in CB1−/− and wild-type mice (in a predominant C57BL/6N background).19 After an overnight fasting period (water ad libitum) a black marker (10% charcoal suspension in 5% gum arabic, 0.1 mL per 10 g body weight) was administered orally to assess upper gastrointestinal transit as described in detail by others.21,22 After 20 min, mice were killed by asphyxiation with isoflurane and the intestine was removed immediately. The distance travelled by the marker was measured in centimetre and expressed as percentage of the total length of the small intestine from pylorus to caecum.

Ascending neuronal contractions in vitro

The ANC were studied in segments of the isolated small bowel based on a established model.23,24 Briefly, male Wistar rats (250–300 g; Charles River) were anesthetized by asphyxiation with CO2 and following intraperitoneal sodium pentobarbital injection (100 mg kg−1). The ileum was immediately removed and kept in oxygenated Krebs–Ringer bicarbonate solution (KRS: NaCl 115.5 mmol L−1, MgSO4 1.16 mmol L−1, NaH2PO4 1.16 mmol L−1, glucose 11.1 mmol L−1, NaHCO3 21.9 mmol L−1, CaCl2 2.5 mmol L−1, KCl 4.16 mmol L−1). A segment of the terminal ileum (10-cm long) was carefully dissected and the mesenteric fat was removed. The gut segment was then placed in a 40 mL organ bath filled with KRS, gassed with carbogene (95% O2–5% CO2) and maintained at 37 °C. The oral and anal ends of the ileum were tied over a polyvinyl tubing (outer diameter 1.8 mm) taking into account the natural length. Two flat platinum electrodes (0.8 × 0.8 cm), which were 1 cm apart, were placed on either side 2 cm proximal of the anal end of the ileum. The specialty of the organ bath was that it could be divided by a baffle into a aboral stimulation chamber and an oral recording chamber (Fig. 1). The recording site was 3 cm orally of the stimulation site, the baffle was located 1 cm orally of the stimulation site. The denseness of the baffle was tested by application of methylene blue at the end of each experiment. Field stimulation impulses to evoke neural responses were applied using a Grass S9 stimulator (Grass, Quincy, MA, USA). The impulse signals were recorded simultaneously with the pressure curves using a Beckman A/C coupler (Beckmann, Berlin, Germany). After an equilibration period of 30 min the gut segment was stimulated every 2 min for 20 s with 20 V, three pulses per second pulse frequency and 1 ms pulse duration. This periodic stimulation was maintained throughout the experiment. When a stable response to the electrical stimulation was established (identical contractions to at least three consecutive stimuli) the experiment was started. The first three contractions served as control contractions and changes to the control values are expressed in percentage. Drugs were then added 60 s after the last stimulation. In 15-min intervals, increasing concentrations of drugs were added to the organ bath in a cumulative manner. Every segment was used for a single experiment only. For respective controls of solvents, saline, dimethyl sulfoxide (DMSO) and Tween had no effect on basal tonus or electrically induced contractions as evaluated over for a 2-h period each.

Figure 1.

 Schematic drawing of the organ bath. Segments of isolated rat ileum were studied in a chambered organ bath. The ANC was initiated by electrical stimulation in the aboral located chamber, force and timing of the myenteric ANC pathways of the myenteric part of the peristaltic reflex of ileum were recorded in the oral chamber. Chambers were partitioned by a baffle only allowing the ileum to pass. This baffle partitions the stimulation and recording sites safely, as proved by methylene blue at the end of each individual experiment.

Intracellular electrical recording

Sheets of distal colon (0.5 cm) were pinned using approximately 150–200 wolfram wire micropins (15–25 μm in thickness) to a Sylgard-based electrophysiological chamber, with the circular muscle layer on top. The chamber was perfused (5 mL min−1; Kwik Pump, World Precision Instruments, Sarasota, FL, USA) with prewarmed (37 °C) oxygenated (95% O2–5% CO2) Krebs solution of the following composition (in mmolL−1): NaCl 120.35; KCl 5.9; MgCl2, 2.5; NaH2PO4 1.2; NaHCO3 15.5; CaCl2 2.5; glucose 11.5, pH 7.4. Muscle strips were allowed to equilibrate for 120 min before the experiments started. Atropine, guanethidine and nifedipine (all 1 μmol L−1) were present throughout. Capillary glass microelectrodes (borosilicate glass capillaries, 1.0 mm outer diameter × 0.58 mm inner diameter; Clark Electromedical Instruments, Reading, UK) were made by using a microelectrode puller (model P-97, 3-mm wide filament; Sutter Instrument, Novato, CA, USA), filled with KCl (3 mol L−1) and had resistances in the range of 80–120 MΩ. Neurons were stimulated (15 V; 0.3 ms duration; single pulses; 5 or 10 Hz) via platinum electrodes arranged perpendicularly to the circular muscle layer using a Grass S11 stimulator and a stimulus isolation unit (SIU59; Grass Instruments). The intracellular 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 analogue-to-digital converter (SCB 68 interface; National Instruments, Austin, TX, USA). Permanent recordings of membrane potentials (MPs) were made on a personal computer using the labview 5.0 program (National Instruments). Inhibitory junction potentials (IJP), elicited by electrical stimulation of intrinsic inhibitory neurons, were measured in millivolts compared to the MP before application of the electrical stimulus.


Anandamide and AM251 (Biotrend, Cologne, Germany) were freshly solved in one drop of Tween 80 and 1 mL DMSO (5%) and then further diluted in saline up to the final concentration. Charcoal, gum arabicum, atropine, guanethidine and nifedipine were obtained from Sigma-Aldrich, Taufkirchen, Germany.


Results are expressed as mean (±SEM) and were compared by using Student’s t-test followed by Dunnet’s post-hoc test, using a commercial statistical package (SigmaStat; Jandel Scientific, San Rafael, CA, USA). P < 0.05 was considered as statistically significant. n indicates the number of independent observations in individual mice.


CB1 receptor RNA expression in CB1+/+ and CB1−/− mice

Cannabinoid-1 receptor mRNA expression was confirmed in the ileum of rat and wild-type mice by RT-PCR. In mucosa as well as in LM-MP (longitudinal muscle myenteric preparation) preparations of both species CB1 mRNA was detected. As expected CB/ mice lack CB1 mRNA (Fig. 2).

Figure 2.

 (A) CB1-receptor and GADPH mRNA expression as determined by RT-PCR in the mouse ileal mucosa and longitudinal muscle/myenteric plexus layer (LM–MP) from wild-type (WT) and CB1-receptor knockout (KO) mice. Brain from wild type and CB1-receptor knockout mice served as positive and negative controls, respectively. (B) CB1-receptor and GADPH mRNA expression as determined by RT-PCR in the rat ileal mucosa and longitudinal muscle/myenteric plexus layer (LM–MP).

Small intestinal transit in CB1+/+ and CB1−/− mice

Upper intestinal transit was measured by the established charcoal method. The CB1+/+ had a weight ranging from 18.7 ± 0.9 g and small intestinal length of 23.7 ± 1.4 (n = 8). The CB1−/− had a weight ranging from 17.6 ± 0.6 g and small intestinal length of 22.6 ± 1.5 (n = 8). The transit in the CB1+/+ (C57BL/6N mice) was 50.2 ± 7.0%. In the CB1−/− mice, the transit was increased to 72.0 ± 7.9% (n = 8; P < 0.05 vs wild type). This represents a 43% increase of the upper gastrointestinal transit (Fig. 3).

Figure 3.

 This graph shows the increased gastrointestinal transit measured with the standardized charcoal method in CB1−/− mice as compared to their wild-type littermates (n = 8 each; P < 0.05).

Ascending neuronal contractile response in rat small intestine

Within 15 min after initiating the rhythmic electrical stimulation, the intestinal segments responded with stable ascending contractions which could be characterized by contraction amplitude and the area under the contraction curve. Additionally, latency, defined as the interval between onset of electrical stimulation and occurrence of the contractile response, was recorded. The recording site was located in the oral chamber and was 3 cm orally of the stimulation site in the aboral chamber. The observed parameters were stable for at least 2 h and remained unaltered in respective solvent controls. All motility responses caused by electrical field stimulation were abolished by TTX (10−6 mol L−1).24 Blockade of muscarinic receptors with atropine (10−6 mol L−1) abolished contractile responses when added to the recording chambers.24 Hexamethonium when added in the aboral chamber reduced contraction amplitude in the oral chamber by 70.1 ± 8.8%* and increased the latency by 30.9 ± 5.1%* (n = 12; *P < 0.05). Hexamethonium when added in the oral chamber reduced contraction amplitude in the recording chamber by 20.8 ± 12.5%* and increased the latency by 12.3 ± 8.3% (n = 10; *P < 0.05). The latency under standard conditions was 11.5 ± 1.1 s. When added into the oral recording chamber, the CB1 receptor antagonist AM251 (10−9–10−5 mol L−1) significantly increased contractile force and the area under the curve in a concentration dependent manner (n = 6). For the maximal concentration of AM251 used (10−5 mol L−1) the increase was +34.5 ± 11.4% (P < 0.05) for the contractile force and +49.9 ± 9.4% (P < 0.05) for the area under the curve (n = 6) (Figs 4 and 5). Interestingly, this effect on ANC was not present when AM251 was added in the aboral stimulation chamber [AM251 (10−5 mol L−1): contractile force: −3.2 ± 4.6%; area under the curve: −3.6 ± 5.6%; n = 6] (Fig. 5B). Latency was not significantly changed by AM251 (10−9–10−5 mol L−1) neither when added in the aboral chamber nor when added to the oral chamber [AM251 (10−5 mol L−1): latency aboral chamber: +5.8 ± 8.2%; latency oral chamber: +5.3 ± 10.1%; n = 6] (Table 2).

Figure 4.

 Representative traces showing the ANC response elicited by electrical field stimulation (20 V; 20 s; 3 pulses per second; 1 ms pulse duration) in the presence of (A) AM251 (10−9–10−5 M) and (B) anandamide (AEA) (10−9–10−5 M). In these experiments, drugs were added in the oral located chamber and stimulation is performed in the aboral located chamber.

Figure 5.

 This figure shows the concentration dependent effects of the endocannabinoid anandamide (10−9–10−5 M) and the CB1 receptor antagonist AM251 (10−9–10−5 M) on amplitude and area under the curve of the ascending neuronal contraction (ANC) elicited by EFS initiated by electrical field stimulation (20 V; 20 s; three pulses per second; 1 ms pulse duration). In panel A, drugs were added in the orally located chamber, whereas stimulation was performed in the aboral located chamber. Whereas anandamide caused a concentration dependent inhibition of the ANC, AM251 significantly increased the ANC. In panel B, drugs were added in the aboral located chamber, whereas recording was performed in the oral located chamber (*P < 0.05; n = 6 each).

Table 2.   Effect of the endocannabinoid anandamide and the cannabinoid-1 (CB1) receptor antagonist AM251 on latency of the ascending neuronal contraction (ANC)
DrugDrug in aboral chamber (%)Drug in oral chamber (%)
  1. CB1, cannabinoid-1; ANC, ascending neuronal contraction.

  2. Latency is defined as the interval between initiation of the ANC by electrical stimulation and the beginning of the contractile response and is expressed in percentage of control. n = 6 each.

  3. All values in table are not significant.

Anandamide (mol L−1)
10−9+10.6 ± 7.0−9.4 ± 8.2
10−8+7.6 ± 7.5−5–7 ± 11.5
10−7+15.3 ± 8.0−6.9 ± 12.4
10−6+16.7 ± 7.8−0.8 ± 7.9
10−5+13.9 ± 10.3−5.3 ± 8.0
AM251 (mol L−1)
10−9−0.7 ± 9.8+8.0 ± 5.7
10−8+7.6 ± 5.8+8.4 ± 6.8
10−7+13.5 ± 4.5+14.4 ± 6.5
10−6+3.7 ± 5.0+16.0 ± 9.6
10−5+5.8 ± 8.2+5.3 ± 10.1

When added into the oral chamber the endocannabinoid anandamide (10−9–10−5 mol L−1) significantly reduced contractile force and the area under the curve in a concentration dependent manner (n = 6). For the maximal concentration of anandamide used (10−5 mol L−1), the reduction was −28.3 ± 4.5% (P < 0.05) for the contractile force and −32.7 ± 7.8% (P < 0.05) for the area under the curve (n = 6) (Figs 4 and 5A). Interestingly, this effect on ANC was not present when anandamide was added in the aboral chamber [anandamide (10−5 mol L−1): contractile force: −9.5 ± 4.5%; area under the curve: −5.0 ± 8.7%; n = 6] (Fig. 5B). Latency was not significantly altered by anandamide (10−9–10−5 mol L−1) neither when added in the aboral chamber nor when added in the oral chamber [anandamide (10−5 mol L−1): latency aboral chamber: +13.9 ± 10.3%; latency oral chamber: −5.3 ±8.0%; n = 6] (Table 2).

Intracellular recordings in CB1+/+ and CB1−/− mice

The resting MP was not significantly different, when CB1+/+ and CB1−/− were compared (RMP CB1+/+: 59.4 ± 5.3 mV; RMP CB1−/−: 57.7 ± 6.8 mV; n = 4). Circular smooth muscle cells of both genotypes displayed the typical tetrodotoxin insensitive slow wave activity (CB1+/+ frequency: 39.5 ± 1.4 min−1; amplitude: 22.1 ± 4.9 mV; CB1−/−: 42.5 ± 3.3 min−1; amplitude: 25.4 ± 3.2 mV; n = 4). Generally, nifedipine is added in the recording chamber to abolish spiking activity.25 This spiking activity causes contractile responses of the tissue and thus results in dislocation of the intracellular electrode. Interestingly, when adding nifedipine (10−6 mol L−1) in the organ bath spiking activity was abolished in the CB1+/+ mice as expected without affecting slow waves (CB1+/+ after nifedipine 10−6 mol L−1; RMP: 53.3 ± 7.2 mV; frequency: 40.1 ± 1.7 min−1; amplitude: 25.7 ± 5.1 mV; n = 4), whereas in CB1−/− mice, slow wave activity was completely abolished (CB1−/− after nifedipine 10−6 mol L−1; RMP: 54.6 ± 4.4 mV; frequency: abolished; amplitude: abolished; n = 4) (Fig. 6). This abolition of slow waves was not observed in CB1+/+ mice when the CB1 receptor antagonist SR141716A was present (n = 4), suggesting developmental changes in the CB1-deficient mice, causing these changes in sensitivity to calcium channel blockade.

Figure 6.

 Intracellular recordings in circular smooth muscle cells. Left: CB1+/+, right CB1−/−. (A) shows intestinal slow waves without and with nifedipine (10−6 M) presence. Note the abolition of slow waves in CB1−/− mice. (B) shows a higher time resolution, demonstrating the spiking activity on the top of the slow waves, which is submaximally reduced by the chosen nifedipine concentration to reduce muscle contractions and thus, dislocation of the electrodes. (C) shows ES (electrical stimulation) induced IJP in presence of nifedipine, which are qualitative present in both genotypes.

To investigate neuromuscular interaction, we induced neuronal mediated junction potentials. When eliciting IJP in both genotypes we found IJP to be present in both genotypes. This displays that inhibitory pathways of neuromuscular transmission are still effective. However, quantitative comparison of IJP in both genotypes is not applicable due to slow wave activity in the wild-type animals and thus different starting situations (Fig. 6).


From previous publications, we know that CB1 receptor activation reduces cholinergic excitatory neurotransmission most likely by prejunctional modulation of acetylcholine release.26 Using different CB1 receptor agonists, postjunctional effects of cannabinoids on unstimulated and/or stimulated smooth muscle cells were ruled out.15,27,28 Interestingly, when antagonists are used cholinergic neurotransmission in small intestinal smooth muscle strips of rodents is sometimes increased, suggesting an inverse agonist activity.7 Still, it can also be speculated that the antagonists block a permanent inhibiting tone mediated by permanent endocannabinoid activity of small intestine. If this second notion holds true, one would expect that in CB1−/− mice gastrointestinal transit should be increased. At the very beginning, we performed RT-PCR and report evidence of CB1 receptor by reporting CB1 mRNA in rat as well as in mouse ileum. CB1−/− mice lacked CB1 mRNA as expected.

By performing the established method of measurement of upper gastrointestinal transit, our present study demonstrates that the assumptions of a permanent inhibiting tone is mediated by endocannbinoids seems correct, since in CB1−/− mice the upper gastrointestinal transit is increased by 43%. This increased transit in the CB1−/− mice points to the physiological relevance of the CB1 receptor within small intestinal motility. Furthermore, the increased transit argues for a permanent suppressive tone in small intestinal motility mediated via the CB1 receptor. Our findings are consistent with the recently published concepts on transit experiments performed in wild-type mice by others. Calignano et al.29 found that the endocannabinoid anandamide (5 mg kg−1) inhibits upper gastrointestinal transit by 50% as compared to vehicle. This effect was sensitive to the CB1 receptor antagonist SR141716A and, therefore, is CB1 receptor-mediated. Given alone SR141716A increased in upper gastrointestinal transit ranging from 17.5% to 48%8,29,30 and the 43% increase in our experiments is well in this range. The increased transit reported here for the CB1−/− mice is furthermore in agreement to other studies where a tendency to an increased transit was reported.31,32 The lack of significance in the other studies might be related to gender differences or methodological differences.

Cannabinoid-1 receptor-mediated effects are also observed in pathological states, such as intestinal inflammation. In a croton oil inflammation model, the increased upper gastrointestinal transit is reduced by CP55,940, a CB1/CB2 receptor agonist. Interestingly, this reduction is more pronounced in the inflamed intestines.33 Therefore, CB1 receptor seems to be also important for the maintenance of gastrointestinal motility under pathophysiological conditions. However, little is known on how CB1 receptors are involved in more complex circuits underlying intestinal propulsion besides reduction of cholinergic excitatory neurotransmission via CB1 receptors, which are located presynaptically.

To elucidate possible underlying mechanisms, we performed in vitro organ bath experiments where the ANC is elicited and development of force and the latency until occurrence of this ANC is evaluated. This complex model of measurement of peristaltic activity is established in rats and guinea pigs but not for mice. In our modification of this method, the organ bath is partitioned in two chambers by a baffle allowing only the ileum to pass (Fig. 1). This allows to observing separately pharmacological effects in the aboral stimulation chamber where electrical stimulation is applied and thus effects in the electrically stimulated neurons and interneurons. In the oral chamber, where the recordings are performed, effects on the final motor neurons and the smooth muscle are observed.

The ANC was significantly increased by the CB1 receptor antagonist AM251 in a concentration dependent manner. This stands in good agreement to the increased small intestinal transit in the charcoal experiments and to reports from others, suggesting the CB1 receptor effects being important for the regulation of small intestinal motility. In other setups such CB1 antagonist effects are sometimes denoted as inverse agonist activity. Although from our organ bath experiments such an inverse agonist effect seems possible, the findings in the CB1−/− mice may argue for a permanent CB1 receptor activation. If lacking an increased transit occurs. Thus, the effect of the antagonist in the organ bath experiments may be judged as a blockade of the permanently activated endogenous cannabinoid system.

In another setup observing the peristaltic reflex induced by intraluminal perfusion, CB1 receptor activation reduced contractility comparable to our setup.27 SR141716A, a comparable CB1 receptor antagonist, in low concentrations alone, had no effect on the ascending peristaltic reflex, higher concentrations were not tested.27 leaving the presently discovered effect of antagonists unobserved. The stand-alone effect of the CB1 receptor antagonist, which is opposite to the effect of anandamide, stresses the physiological relevance of this receptor in the regulation of intestinal motility. Findings from other experimental designs directed to peristalsis are in agreement to our observation that CB1 antagonists increase peristalsis. In isolated guinea-pig ileum segments, the CB1 antagonists SR141716A and AM281 slightly but significantly increased maximal ejection pressure during the empty phase of peristalsis34 and in isolated mouse colon SR 141716A enhanced both tonic and phasic motor activities in a model of peristalsis evoked by electrical stimulation.35

The endocannabinoid anandamide had opposite effects as compared to the antagonist. The ANC was significantly inhibited by anandamide in a concentration dependent manner. This inhibition is present when anandamide is added in the oral chamber but not when anandamide is added in the aboral chamber. This provides evidence that anandamide activates receptors at a neuromuscular site. Since no effect is observed when anandamide is added to the aboral chamber where only neurons but not the neuromuscular site is in contact with anandamide, an additional site of anandamide action at a neuro-neuronal site is unlikely. The finding that propagation latency is unchanged further supports the notion that neuro-neuronal sites are not involved in the action of anandamide. From organ bath experiments on intestine, which were published for rats and guinea pigs, it is well accepted that CB1 receptors reduce neuromuscular interaction. The receptors can functionally be localized on a prejunctional nerve ending, where CB1 receptor activation reduces acetylcholine release towards the smooth muscle.36 A smooth muscular localization of CB1 receptors was never reported. Our results reflect this. In addition, we report that within the ascending myenteric pathways an effect of anandamide on neuro-neuronal interaction is unlikely. However, such an effect cannot be completely ruled out by our experiments since due to the lipid nature of anandamide, some parts of the tissue might not be penetrable to anandamide. In a study using potent synthetic CB1 receptor agonists neuronal excitatory postsynaptic potentials (EPSP) were reduced in a subpopulation of myenteric S-neurones in a SR141716A sensitive manner, however, in another subpopulation, the EPSP were reduced by the CB1 antagonist SR141716A.37

Slow wave activity is an electrophysiological phenomenon in the regulation of gastrointestinal motility. Slow waves are generated by the interstitial cells of Cajal (ICC) as has been demonstrated by experiments in cKit−/− mice38 slow waves are necessary for coordinated contraction and peristalsis.39 Interestingly, CB1−/− mice are sensitive to calcium channel blockade as demonstrated by the presence of nifedipine, a phenotype comparable to cKit−/− mice. This increased sensitivity to the Ca2+-channel blocker nifedipine selectively concerns the slow wave activity, whereas neuromuscular junction potentials seem unaltered and can be elicited in mice of both genotypes. This argues against a general sensitivity to Ca2+-channel blockade, but rather in favour of a specific sensitivity of the slow waves in the CB1−/− mice. The finding that the CB1 receptor antagonist did not alter slow waves in the CB1+/+ mice strongly argues for developmental changes caused by the lack of CB1 receptors. Although IJP are not comparable in the different genotypes due to the different starting conditions, they are well present in both genotypes, arguing against the possibility that differences in inhibitory neurotransmission are involved in the CB1 mediated effects on small intestinal motility. Comparable, in mouse colon, where the electrophysiological mechanisms underlying inhibitory neurotransmission are easier accessible, IJP are not reduced by CB1 receptor activation with anandamide.40

In summary, our present study supports the notion that CB1 receptors are physiologically involved in small intestinal motility. The transit experiments in the CB1−/− mice demonstrate that the lack of CB1 receptors causes a significant augmentation of gastrointestinal transit. This stands in good agreement with a permanent inhibitory tone on gastrointestinal motility, mediated by CB1 receptors. The CB1 receptors hereby involved modulate small intestinal motility by modulation of the ANC, most likely by CB1 receptors located on motor neurons. The presently reported findings suggest the endocannabinoid system as an important component in the regulation of gastrointestinal motility.


This study was supported by grants from the Deutsche Forschungsgemeinschaft, Bonn, Germany (STO 645/2-1 to M.S.) and the Society of Gastroenterology in Bavaria, Germany (to M.S.).