Serotonin in the rabbit ileum: Localization, uptake, and effect on motility



Repeated experiments to localise serotonin in the myenteric plexus of rabbit ileum failed. After preincubation in serotonin (10−5 M), an extensive varicose fibre system was detected by immunocytochemical methods. Stained fibres left the myenteric plexus and ran to the muscle layers. Labelled cell bodies could not be found, even after pretreatment with colchicine or pargyline. Application of reserpine (10−5 M) and fluoxetine (10−5 M) prevented serotonin uptake. Antisera against tryptophan hydroxylase revealed a rich fibre system, including those processes that entered the tertiary plexus. These fibres were able to accumulate serotonin, but again the cell bodies could not be detected. Serotonin caused concentration-dependent contraction in the longitudinal muscle layer of the rabbit ileum. Pretreatment with tetrodotoxin strongly reduced the effect of serotonin. Preapplication of atropine caused a slight decrease of response evoked by serotonin. Combined administration of tetrodotoxin and atropine significantly reduced the responses to serotonin, but did not abolish them. At the same time, agonists of 5-HT2 and 5-HT4 receptors caused concentration-dependent contractions. Our studies show that: 1) Without pretreatment, serotonin cannot be detected in the myenteric plexus of rabbit ileum. 2) An extensive uptake system works in this plexus. If released from myenteric nerve fibres, serotonin may evoke contractions in indirect and direct ways. 3) There may be an extrinsic serotoninergic innervation from the mesenteric ganglia. 4) Serotonin exerts its effect through 5-HT2 and 5-HT4 receptors on smooth muscle cells and nerve elements. Anat Rec Part A 271A:368–376, 2003. © 2003 Wiley-Liss, Inc.

The motility of the gastrointestinal tract of the rabbit deserves special attention because of its unique features. Its basal tone is low and there is a neurally evoked component of the rhythmic contraction of the small intestine, which results in pendular movements (Kehl, 1984; Deloof and Rousseau, 1985). The anatomy of the rabbit myenteric plexus is also unusual. There are only about 2,500 neurons/cm2 in the rabbit small intestine (Gábriel et al., 1998), which is significantly different from the most commonly studied laboratory animals (rats, guinea-pigs, and mice) (Gabella, 1987). Indeed, the innervation pattern and chemical neuroanatomy of the rabbit intestine are different from those of the above-mentioned animals (Gábriel et al., 1998; Wilhelm et al., 1998).

Serotonin (5-hydroxytryptamine, 5-HT) has been identified as a transmitter in the gut by Gershon et al. (1965). They found that 5-hydroxytryptophan was taken up by the intestine and converted to 5-HT, and that a major site of retention of 5-HT was the myenteric plexus (Gershon et al., 1965). Since then a considerable number of studies have strengthened the supposition that 5-HT acts as a neurotransmitter in the enteric nervous system (ENS) (Wood and Mayer, 1978; Wood, 1979; Gershon et al., 1980). Early works using aldehyde-induced fluorescence failed to convincingly demonstrate the presence of 5-HT in enteric neurons (Robinson and Gershon, 1971; Furness and Costa, 1978); subsequently, however, by the use of antibodies raised against 5-HT, the presence of this substance was demonstrated in the small intestine and colon of the mouse, rat, guinea-pig, and pig (Costa et al., 1982; Timmermans et al., 1990, 1991).

Parallel to morphological studies, numerous pharmacological experiments provided evidence that 5-HT has extremely diverse roles in the gastrointestinal tract. This monoamine influences the transport of nutrients, and inhibits the Na+-dependent system of transport of D-galactose and L-leucin in rabbit ileum in vitro (Salvador et al., 1996, 1997). Furthermore, 5-HT alters the intestinal electrolyte transport; increases fluid, Cl, Na+, and K+ secretion into the intestinal lumen (Donowitz et al., 1980; Kellum et al., 1994; Borman and Burleigh, 1997; McLean and Coupar, 1998); inhibits Cl/HCOmath image exchange in villus cells; and stimulates Na+/H+ exchange in crypts (Sundaram et al., 1991). Exogenous 5-HT induces excitatory mechanical responses in different parts of the gastrointestinal tract in mammals when administered in vivo (Eglen et al., 1990; Tamura et al., 1996; Briejer et al., 1997), and additionally enhances mechanical activity and the tone of muscle fibres (Ng et al., 1991; Salvador et al., 2000). These and other experiments raised the possibility that 5-HT may be an endogenous activator and modulator of enteric neuronal circuits (Gershon et al., 1990). According to the above-mentioned studies, modulatory effects of 5-HT are mediated by 5-HT2, 5-HT3, and 5-HT4 receptors.

Most of the morphological articles regarding 5-HT distribution focused on the digestive tract of the guinea-pig and other rodents. Little is known about the serotonergic elements in the rabbit. The present work demonstrates the absence of 5-HT immunoreactivity from the myenteric plexus, and the presence of 5-HT accumulating nerve fibres by using the vesicular 5-HT transporter inhibitor, reserpine, and a blocker of specific plasmalemmal 5-HT transporter, fluoxetine. Our study also investigates the effect of 5-HT on the motility of this gut segment.



We used adult white rabbits of both sexes. The animals were housed in individual cages, and fed and watered ad libitum. The experimental animals were anaesthetised with an i.p. injection of ketamine solution (10 mg/kg body weight) and pithed. In some cases colchicine (10 mg/kg; Sigma, St. Louis, MO) or pargyline (30 mg/kg; Sigma) was injected twice i.p., 40 and 16 hr before the animals were killed. Pargyline was also used in in vitro experiments as described below.

Tissue Preparation

The last 10 cm of the small intestine just before the ileocaecal junction was discarded and the rest of the ileum was used for experiments. Gut segments were first washed thoroughly and then filled with Krebs solution. Pieces of gut were placed into 4% paraformaldehyde in phosphate buffer and fixed at room temperature for 8 hr. For immunohistochemistry, the tissues were washed in phosphate-buffered saline (PBS) for 10 × 10 min. Myenteric plexus/longitudinal muscle layer whole-mount preparations were then made and treated with 100% dimethyl sulfoxide or 1% Triton X-100 for 6 × 10 min to enhance the penetration of antibodies.

Cryostat sections.

Fixation was the same as that for the whole-mounts. After fixation the gut segments were stored in 30% sucrose until the preparations fully submerged. Tissues were placed into special tissue freezing medium (OCT; Triangle Biomedical Sciences, Durham, NC) and frozen. Gut pieces were cut in a cryostat at −20°C. Longitudinal sections (10 μm thick) were picked up on gelatine-coated slides, air-dried overnight at room temperature, and stored at −80°C until they were used.


Processing of untreated tissues.

Whole-mounts and sections were preincubated for 1 hr in antibody diluent solution (1% bovine serum albumin, 0.4% Triton X-100, 0.1% NaN3 in PBS). Subsequently they were soaked with primary antisera overnight at 4°C. The following primary antisera were used: anti-serotonin raised in rabbit (1:1000; Incstar, Stillwater, MN), anti-tryptophan hydroxylase (TrpOH) raised in sheep (1:500; Chemicon, Temecula, CA), and anti-tyrosine hydroxylase (TH) raised in mouse (1:1000; Chemicon, Temecula, CA). The tissues were then rinsed 6 × 10 min in PBS and placed into biotinylated anti-species secondary antibodies (goat anti-rabbit, donkey anti-mouse, or donkey anti-sheep IgG at 1:100 dilution) for 5–6 hr. Then we used extravidin-peroxidase complex (1:100; Sigma) overnight for single diaminobenzidine (DAB) staining. For fluorescence double-labelling, Texas Red dye-conjugated affinity purified donkey anti-sheep (1:100; Jackson, West Grove, PA) and FITC-conjugated affinity purified F(ab)2 anti-rabbit (Jackson) were used as secondary antisera, for TrpOH and for 5-HT and TH, respectively. To study 5-HT uptake (see below) by noradrenergic fibres, the primary antiserum against TH was visualised by using DAB followed by detection of 5-HT with FITC, as described above. Finally, the whole-mounts were coverslipped in Vectashield (Vector, Burlingame, CA) and photographs were taken with an Olympus BH2 or a Nikon Microphot FXA light microscope equipped with an epifluorescence setup.

5-HT uptake experiments.

Pieces of freshly removed gut segments were placed into Krebs solution containing 5-HT in different concentrations (0.1–10 μM). The gut segments were incubated for 25 min while the medium was constantly oxygenated. In parallel with the preceding experiment, other pieces of ileum were incubated in Krebs solution containing 10 μM 5-HT and 10 μM reserpine (Sigma) or fluoxetine (Sigma) for 25 min. Some gut pieces were preincubated in pargyline (5 × 10−5 M) dissolved in Krebs solution for 60 min. After pretreatment, the gut pieces were filled with Krebs solution and fixed in 4% paraformaldehyde for 8 hr at room temperature. Processing and photography were performed as described above.

Immunocytochemical controls.

Both negative and positive controls were performed. For positive control we processed brain stem sections that contained the raphe nuclei to ensure that our antisera (anti-5-HT and anti-TrpOH) worked properly. As negative controls we omitted the incubation step with primary antisera or replaced it with non-immune sera. The cross-reactivity of primary antibodies with the noncorresponding secondary antibodies, as well as leakage of nonspecific lights through the microscope filter sets, were checked and ruled out.

Pharmacological Experiments

Pieces of ileum were cut into segments 1.5–2.0 cm long. Myenteric plexus-longitudinal muscle preparations were made according to the method of Paton and Vizi (1969) and maintained in Krebs-Henseleit solution (in mM: NaCl 119, NaHCO3 25, KH2PO4 1.2, CaCl2 2.5, MgSO4 1.5, KCl 2.5, and glucose 11). The solution was oxygenated by bubbling with a mixture of 5% CO2/95% O2, and was kept at 37°C. The preparations were connected to isotonic transducers, under a load of 1 g. Mechanical activity was recorded on compensographic recorders by means of measuring bridges. For the drugs used, a concentration series was tested from 10−8 to 10−3 M, and the optimal working dilutions were used in further studies. Contractions were compared to and expressed as a percentage of the maximal longitudinal spasm due to KCl (80 mM) administered at the end of the experiment.


The following drugs were used: 5-HT maleate, atropine sulfate, tetrodotoxin (TTX), buspirone, methysergide, 5-carboxyaminotryptamine, α-methyl-5-hydroxytryotamine (a-m-5-HT), tropisetron, and 5-metoxytryptamine (MOT) (all from Sigma). These drugs were all dissolved in distilled water. The contact times used for pretreatments were as follows: atropine, 20 min; tetrodotoxin, 15 min; and methysergide, 20 min. The drugs were put into the bath fluid with microsyringes. Concentrations are given as the final dilutions in the bath.

Data Processing

Data are presented as means ± SEM. Statistical comparisons were made by using Wilcoxon's signed-rank test (two samples) and the Quade-test for several related samples. A difference of P < 0.05 was considered statistically significant. Any group of results contains at most two data from the same animal. Each group of pretreated preparations was compared with a control group of preparations taken from the same animals.



After the specimens were incubated with anti-5-HT antibodies, immunoreactive nerve elements could not be found (Fig. 1a), even after colchicine or pargyline pretreatment. To exclude false negativity, we performed immunolabelling on cross-sections of the gut, under the same conditions as in the control study. A few stained entero-endocrine cells and many immunopositive mast cells were found in the epithelium (Fig. 1b) and the connective tissue of the mucosa, respectively (Fig. 1c).

Figure 1.

Photographs of whole-mount preparations and cross-sections of rabbit ileum. a: 5-HT-immunoreactivity could not be found in the whole-mounts of the myenteric plexus. b: 5-HT immunopositive entero-endocrine cell in the epithelium (arrow) in cross-section. c: 5-HT immunopositive mast cells in the connective tissue (arrows) in control cross-section. d: 5-HT immunopositive fibre network after pre-incubation in 5-HT solution. Numerous fibres take up 5-HT (arrowheads). Stained fibres contain many varicosities but no cell bodies, and surround nonreactive nerve cells (asterisks). e: Varicose 5-HT immunopositive fibres (arrowheads) run to the muscle layer. Scale bar = (a–d) 5 μm and (e) 8 μm.

In a 5-HT uptake experiment, living-gut segments were preincubated in Krebs solution containing 5-HT, to reveal potential monoamine-accumulating cells in the myenteric plexus of the rabbit. Incubation with 0.1 and 1 μM 5-HT gave no staining. However, after preincubation in 10 μM 5-HT, an extensive 5-HT immunopositive fibre system without cell bodies was detected, in both the ganglia and the tertiary plexus. These processes bore densely labelled varicosities that formed pericellular baskets around unlabelled somata (Fig. 1d). Immunopositive nerve bundles ran out of the ganglia to the muscle layer (Fig. 1e). Co-application of 5-HT and reserpine (10 μM), and 5-HT and fluoxetine (10 μM) did not result in any labelling with the anti-5-HT antibody (data not shown).

Staining with anti-TrpOH revealed a moderately immunopositive fibre system. Cell bodies were not seen (Fig. 2a), even after colchicine treatment. Stained fibres left the myenteric plexus and ran to the muscle layers (Fig. 2c). Because this arrangement somewhat resembled the system of TH-positive fibres described earlier (Gábriel et al. 1998), a comparison could be made. The anti-TH antiserum labelled a scarcer fibre system than the anti-TrpOH antibodies. Occasionally cell bodies were also seen (Fig. 2b). The TH-containing varicosities were larger than their TrpOH-labelled counterparts. The TH-immunoreactive processes almost never left the large nerve bundles, and never formed a mesh-like tertiary plexus (Fig. 2d). Thus, the TH-positive and TrpOH-positive fibre systems were clearly distinct from each other.

Figure 2.

Comparison of TrpOH- and TH-immunoreactive structures in the rabbit myenteric plexus. a: A rich meshwork of TrpOH-immunoreactive nerve fibres (arrowheads), many of which surround unlabelled cell bodies (asterisks) in the ganglia of the myenteric plexus. No cell bodies were seen in the ganglia or the interconnecting strands. b: A scarce fibre system (arrowheads) can be labelled for TH. A nerve cell positive for TH (arrow) can also be seen. c: Elements of the tertiary plexus are strongly labelled for TrpOH (arrowheads), and pericellular baskets (asterisk) are also seen in the interconnecting strands (arrowheads) of the myenteric plexus. d: Although some TH-positive fibres can be observed in interconnecting strands (arrowheads), TH-positive neurites cannot be seen in the tertiary plexus. Arrow: TH-positive cell body. Scale bar = 10 μm.

Double labelling between 5-HT-accumulating structures and TrpOH immunopositive fibres showed an almost complete overlap correspondence. Most 5-HT-accumulating nerves had TrpOH immunoreactivity (Fig. 3a and b). The fibre system in the tertiary plexus appeared to be identical with the two markers (Fig. 3c and d).

Figure 3.

Serotonin uptake and synthesis, and relation of TH-immunoreactivity to TrpOH content and 5-HT accumulation in the rabbit ileal myenteric plexus. a and b: Almost all 5-HT-immunoreactive bundles are positive for TrpOH (some identical structures are marked with arrows). c and d: All 5-HT-accumulating fibres of the tertiary plexus contain TrpOH (arrows). e and f: TH-immunopositive cells never colocalise TrpOH (arrows). g and h: Nerve fibres never contain both enzymes (TH, arrowhead; TrpOH, arrow). i and j: One of the two 5-HT accumulating pericellular baskets contains TH (arrows) while the other does not (arrowheads). k and l: Most fibres accumulate 5-HT (arrows) but do not contain TH (arrowheads). Scale bar = 10 μm.

Colocalization between TrpOH and TH showed that all structures labelled with the respective antibodies were different (Fig. 3e–h). Analysis of 5-HT-accumulating nerve elements and TH-immunoreactive processes showed that the vast majority of TH-immunopositive fibres (especially those in the tertiary plexus) could not take up 5-HT (Fig. 3k and l). Only in the case of pericellular baskets were double-labelled TH-positive structures seen (Fig. 3i and j). At the same time, numerous pericellular baskets contained 5-HT only (Fig. 3i and j).


Exogenous 5-HT (10 nM–100 μM) caused concentration-dependent contractions (Fig. 4a). 5-HT (1 μM) given four times at 30-min intervals caused contractions in a reproducible manner (n = 6; data not shown).

Figure 4.

Effect of 5-HT and its agonists. a: Effect of 5-HT (0.01–100 μM) on spontaneous mechanical activity in the rabbit ileum in vitro. b: Contractions evoked by the 5-HT2 agonist α-methyl-5-hydroxytryptamine can be fully blocked by methysergide. c: The 5-HT4 agonist 5-metoxytryptamine has a strong contractile effect on the smooth muscles. Its action is blocked by methysergide.

The 5-HT receptor type responsible for the above-described processes was further investigated by 5-HT receptor subtype agonists and antagonists (5-HT1: buspirone and 5-carboxyaminotryptamine; 5-HT2: a-m-5-HT and methysergide; 5-HT3: tropisetron; 5-HT4: MOT). Both a-m-5-HT and MOT caused concentration-dependent contractions (Fig. 4b and c). Their effect could be blocked by methysergide (Fig. 4b and c). Contractions could not be evoked with the 5-HT1-receptor agonists used (10−6 to 10−4 M; n = 5; data not shown), and the contractions induced by 5-HT could not be reduced or blocked by the 5-HT1 or 5-HT3 receptor antagonists (10−6 to 10−4 M; n = 4; data not shown). TTX application in the bath fluid significantly reduced the contractile effect of 5-HT (n = 6; Fig. 5a). Pretreatment with cholinergic muscarinic receptor blocker atropine (1 μM) resulted in a slight but significant decrease in the response evoked by 5-HT (1 μM) was observed (n = 6; Fig. 5b). The combined administration of atropine and TTX significantly reduced, but did not abolish, the response to 5-HT (n = 6; Fig. 5b).

Figure 5.

Combined drug applications: identification of direct and indirect effects of 5-HT. a: TTX reduced the 5-HT-induced contraction significantly. b: After pretreatment with atropine, the contraction evoked by 5-HT decreased significantly. The combined blockade of atropine and TTX reduced the response. c: TTX reduced the 5-HT2 agonist-induced contraction. d: TTX cannot block the action of MOT.

The presence of TTX reduced the contraction evoked by the 5-HT2 agonist (n = 8; Fig. 5c), but did not influence the response to the 5-HT4 agonist (n = 6; Fig. 5d).


In the present study we examined the distribution and potential roles of 5-HT in the rabbit ileum with immunocytochemical and pharmacological methods. We could not detect 5-HT immunopositive nerve elements in the myenteric plexus without pretreatment, although 5-HT-accumulating fibres were present. TrpOH immunocytochemistry revealed numerous immunopositive fibres in the ganglia and the internodal strands, and on the surface of the longitudinal muscle layer, but cell bodies could not be visualised. The contractile effect of 5-HT on the gut musculature has a direct and an indirect component.

Serotonin Is a Potential Neurotransmitter/Modulator in Rabbit Myenteric Plexus

There is evidence that 5-HT is one of the most prominent neuroactive substances in the gut (Dreyfus et al., 1977; Costa et al., 1982; Wardell et al., 1994; Schemann et al., 1995). The present data confirmed our previous experiments that failed to reveal 5-HT in the myenteric plexus of the rabbit small intestine (Gábriel et al., 1998). However, negative results of 5-HT-immunolabelling must be interpreted with caution, because they may be caused by the fast biochemical turnover of this monoamine. We also tried to apply pargyline to block the oxydation of 5-HT, but without any success. Despite pargyline treatment, 5-HT in the fibres remained below the level of immunocytochemical detectability. At the same time, as we have proven in this study, an uptake system works even though it is restricted to a rich fibre network. Many varicose axons innervating the longitudinal muscle layer may be able to release the accumulated 5-HT, which means that this substance may still have a functional role in gut motility. Futhermore, the uptake system could play a role in the inactivation of 5-HT released from non-neuronal elements, since in the absence of an adequate inactivating mechanism, receptors for 5-HT would be likely to desensitise. In some instances, noradrenergic fibres located at key positions in the myenteric plexus may also participate in this process. It has been shown in this work that some TH-immunopositive pericellular baskets are also able to accumulate 5-HT, while the rest of the noradrenergic fibre system is not. Prevention of 5-HT accumulation by reserpine and fluoxetine clearly proves the existence of 5-HT transporters on both the cell membrane and the intracellular vesicles of the myenteric nerve elements. The coexistence of 5-HT synthesis and accumulation has not yet been reported in the ENS (Costa et al., 1982; Erde et al., 1985; Wardell et al., 1994; Meedeniya et al., 1998). Therefore, we performed double labelling to examine whether TrpOH-positive fibres have a 5-HT-uptake system. There was an almost complete overlap observed between these fibre populations, so it appears that these features can be tied to the same set of nerve fibres.

Serotonin Has a Prominent Motoric Effect in the Rabbit Small Intestine

Previous studies have investigated the effects of 5-HT on movements of the gut in several species and different gut segments. The results show that 5-HT can induce contractions in the intestinal smooth muscle both directly and through neuronal release of acetylcholine (Brownlee and Johnson, 1963; Ádám-Vizi and Vizi, 1978; Costa and Furness, 1979; Ng et al., 1991). Experiments performed on rabbit colon provided evidence that 5-HT stimulated the distal and proximal circular muscles in a dose-dependent manner, but that the responses of the longitudinal muscle were much smaller (Ng et al., 1991). The pharmacological results of the present study clearly show a stimulating effect of 5-HT on the longitudinal muscle in the small intestine. While this study was in progress, Salvador et al. (2000) reported that 5-HT increases the amplitude of contractions and integrated mechanical activity in the duodenum, jejunum, and ileum of the rabbit. The action mechanism of 5-HT in the rabbit ileum is somewhat similar to that described in the guinea-pig small intestine (Brownlee and Johnson, 1963; Gaddum and Picarelli, 1975).

The effects of 5-HT in the intestine may be mediated via 5-HT1, 5-HT2, 5-HT3, and 5-HT4 receptor types (Gershon et al., 1991; Ng et al., 1991; Kuemmerle et al., 1993; Michel et al., 1997; Salvador et al., 1997). We found that after administration of TTX, the effect of 5-HT was significantly reduced but not fully abolished. Therefore, it is obvious that contractions evoked by 5-HT are mediated via both direct and indirect mechanisms. Our results indicate that stimulation of muscle contractility appears to be stronger through the nerve elements than on the muscles. Interestingly, TTX only slightly influenced the 5-HT-mediated stimulation of proximal and distal colon, which suggests that neural pathways play a less prominent role in those gut segments (Ng et al., 1991). The direct effects of 5-HT may have been caused by release of 5-HT from the long varicose axons we observed on the surface of the longitudinal muscle layer. We have shown in this study that these effects in rabbit are mediated through 5-HT2 (on both muscles and nerve elements) and 5-HT4 (on smooth muscle cells exclusively) receptors.

Neuronal pathways activated by 5-HT appear to consist of cholinergic and noncholinergic components, as shown by the reduction in the responses by atropine and atropine + TTX. It is therefore probable that in the rabbit ileum 5-HT exerts its effect at least partially by stimulation of cholinergic motoneurons, through 5-HT2 receptors. The transmitter(s) used by the noncholinergic mechanism remain to be identified.

The gastrointestinal tract is one of the body's most abundant sources of 5-HT. Previous studies have shown that 5-HT can be produced in and released from mast cells and entero-endocrine cells of the gut wall (Kirchgessner et al., 1992; Yuan et al., 1994; Wang et al., 1995). The release of 5-HT from entero-endocrine cells may be stimulated by muscarinic cholinergic receptors (Forsberg and Miller, 1983). These cells, in addition to 5-HT-handling enteric neurons, may contribute to the regulation of intestinal motility in the rabbit. Indeed, despite the maintenance of 5-HT in the intestine in serotonin-transporter knockout mice, these animals show (apart from disturbances in cation transport) serious abnormalities in intestinal motility (Chen et al., 2001). A similar situation would also be expected in rabbits.


This study was supported by OTKA T 34160 (to R.G.) and ETT 03-383/2000 (to L.B.). R.G. also received a János Bolyai Fellowship.