Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

5-Hydroxytryptamine (serotonin; 5-HT) is a neurotransmitter and paracrine messenger substance in the gastrointestinal tract. In the gastrointestinal tract, 5-HT can act on at least four different receptors to alter neuronal activity. 5-HT acting at 5-HT_1P receptors is a mediator of some slow excitatory synaptic potentials in the myenteric and submucosal plexuses (Gershon, 1995). The 5-HT_1P receptor is a G protein-linked receptor that couples to one or more intracellular signaling pathways, resulting in inhibition of resting potassium channels and membrane depolarization (Pan et al., 1997). 5-HT also acts at 5-HT_1A receptors, which couple via an unidentified signaling mechanism to activation of a potassium channel and neuronal inhibition (Galligan et al., 1988; Pan and Galligan, 1994; Galligan, 1996). 5-HT₄ receptors are localized on cholinergic nerves, and activation of 5-HT₄ receptors causes facilitation of acetylcholine (ACh) release (Pan and Galligan, 1994; Tonini and De Ponti, 1995). 5-HT_1A and 5-HT₄ receptors are G protein-linked receptors.

5-HT₃ receptors are ligand-gated cation channels (Derkach et al., 1989; Fletcher and Barnes, 1998). In the gastrointestinal tract, 5-HT₃ receptors are localized to enteric sensory nerve endings in the mucosal layer (Foxx-Orenstein et al., 1996; Bertrand et al., 1998). 5-HT released from enterochromaffin cells in response to stimulation of the mucosa acts on the nerve terminal 5-HT₃ receptors to initiate motor reflexes (Foxx-Orenstein et al., 1996). 5-HT₃ receptors are also localized to the nerve cell body of many enteric neurons, where they mediate rapidly developing and desensitizing depolarizations (Galligan, 1995, 1996). 5-HT₃-mediated responses have a reversal potential near 0 mV and are due to activation of a nonspecific cation conductance (Yakel and Jackson, 1988; Derkach et al., 1989). Although most enteric nerves express 5-HT₃ receptors, their contribution to synaptic excitation is unclear.

Although the macroscopic response caused by 5-HT₃-receptor activation in the myenteric plexus is similar to that mediated at nicotinic ACh receptors (Zhou and Galligan, 1996a), the pharmacological or single-channel properties of 5-HT₃ receptors in myenteric neurons have not been characterized. The purpose of our study was to determine whether 5-HT₃ receptors contribute to synaptic excitation in acutely isolated myenteric plexus preparations and to examine in detail the properties of whole-cell and single-channel currents caused by 5-HT₃-receptor activation in myenteric neurons maintained in primary culture.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Conventional Intracellular Electrophysiological Recording. Guinea pigs (300-400 g) were obtained from the Michigan Department of Public Health (Lansing, MI). The animals were sacrificed by a blow to the head and bleeding through the major neck vessels after halothane inhalation anesthesia. This procedure was approved by the All University Animal Use and Care Committee at Michigan State University. A segment of ileum taken 10 cm proximal to the ileal-cecal junction was removed and placed in oxygenated (95% O₂/5% CO₂) Krebs' solution in the following composition: 117 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 25 mM NaHCO₃, 1.2 mM NaH₂PO₄, and 11 mM glucose. The solution contained nifedipine (1 µM) and scopolamine (1 µM) to block contractions of the muscle layer during the intracellular recordings. A segment of ileum (5 cm) was cut open along the mesenteric attachment and pinned flat in a silastic elastomer-lined Petri dish. The mucosa and submucosa were carefully peeled away, and a 5-mm² section of longitudinal muscle myenteric plexus was cut out with fine scissors and forceps. The longitudinal muscle myenteric plexus preparation was transferred to a silastic elastomer-lined recording chamber and pinned flat. The chamber was then mounted on the stage of an inverted microscope (Nikon Diaphot), and the chamber was superfused continuously with warm (36°C) Krebs' solution at a flow rate of 3.5 ml/min. Neurons were impaled with 2 M KCl-containing electrodes (tip resistance, 80-100 M), and membrane potential was recorded with an Axoclamp 2A amplifier (Axon Instruments, Foster City, CA). Synaptic potentials were evoked by use of a focal electrode to stimulate the interganglionic connectives entering the ganglion containing the impaled neuron. The focal stimulating electrode was a glass pipette (tip diameter, 60 µm) filled with Krebs' solution. Single stimuli (0.5 ms duration) were used to evoke fast excitatory postsynaptic potentials (fEPSPs). Stimuli were provided by a pulse generator (Master 8, A.M.P.I.) and a constant-current stimulus isolation unit. A maximal amplitude stimulus (4-9 mA) was used to elicit fEPSPs. The initial membrane potential was adjusted to approximately 90 mV to prevent the fEPSP (or drug responses, see below) from reaching action-potential threshold. This permitted a more accurate measurement of the amplitude of the fEPSP and the effect of drugs on fEPSP amplitude. Data were acquired with a Labmaster 125 analog-digital converter, a personal computer, and Axotape 2.0 software (Axon Instruments). Membrane-potential changes were sampled at 2 kHz and were filtered at 1 kHZ (4-pole Bessel filter; Warner Instruments, New Haven, CT). A digital average of six fEPSPs was used to measure fEPSP amplitude under control conditions and after drug treatments.

Tissue Culture. Myenteric neurons were cultured via techniques described previously (Zhou and Galligan, 1998). Newborn guinea pigs (<36 h old) were sacrificed by severing the major neck blood vessels and spinal cord after deep halothane anesthesia. These procedures were also approved by the All University Committee on Animal Use and Care at Michigan State University. The small intestine was placed in cold (4°C) sterile-filtered Krebs' solution of the following composition: 117 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 1.2 mM NaH₂PO₄, 25 mM NaHCO₃, and 11 mM glucose. The longitudinal muscle myenteric plexus was removed from the entire length of small intestine and cut into 5-mm-long pieces. The dissected tissues were divided into four aliquots and placed in 1 ml of Krebs' solution containing 1600 U of trypsin (Sigma Chemical Co., St. Louis, MO) for 25 to 30 min at 37°C. After trypsin incubation, the tissues were triturated 30 times and then centrifuged at 900g for 5 min with a bench-top centrifuge. The supernatant was discarded, and the pellet was resuspended in sterile Krebs' solution and incubated (25-30 min, 37°C) in Krebs' solution containing 2000 U crab hepatopancreas collagenase (Calbiochem-Novabiochem, Corp., La Jolla, CA). The suspension was triturated and then centrifuged for 5 min. The pellet was resuspended in Eagle's minimum essential medium containing 10% fetal calf serum, gentamicin (10 µg/ml), penicillin (100 U/ml), and streptomycin (50 µg/ml) (all from Sigma). Cells were plated on plastic dishes coated with poly-L-lysine and maintained in an incubator at 37°C in an atmosphere of 5% CO₂ for up to 2 weeks. After 2 days in culture, 10 µM cytosine arabinoside was added to the minimum essential medium to limit smooth muscle and fibroblast proliferation, and the medium was changed twice weekly thereafter.

Whole-Cell and Single-Channel Patch-Clamp Recording. Whole-cell and outside-out patch-clamp recordings were obtained via standard methods. Recordings were carried out at room temperature with patch pipettes with tip resistances of 3 to 5 M for whole-cell and 5 to 10 M for single-channel currents in outside-out patches; seal resistances were >5 G. The tips of pipettes used for single-channel recordings were coated with Sylgard (Dow Corning, Midland, MI). The pipette solution contained the following: 160 mM CsCl, 2 mM MgCl₂, 1 mM EGTA, 10 mM HEPES, 1 mM ATP, and 0.25 mM GTP, the pH and osmolarity were adjusted to 7.4 (with CsOH) and 315 mosmol/kg (with CsCl), respectively. All recordings were made with an Axopatch 200A amplifier. Data were acquired with Axotape 2.0 and pClamp 6.0 software. Currents were sampled at 2 kHz and were filtered at 1 kHz (4-pole Bessel filter, Warner Instruments, Hamden, CT) and stored on a computer hard drive.

Drug Application. Drugs were applied in three ways in studies with conventional intracellular electrophysiological methods to record from neurons in the intact myenteric plexus. 5-HT was applied from a fine-tipped (<10 µm) pipette positioned near the neuron. The concentration of 5-HT in the pipette was 1 mM and was ejected from the pipette using a brief nitrogen pulse (Picospritzer, General Valve Inc., Fairfield, NJ). The amplitude of single 5-HT responses under control conditions and in the presence of antagonists was measured. ACh was applied by iontophoresis onto the impaled neuron. The concentration of ACh in the iontophoretic electrode was 1 M, and ACh was ejected from the electrode via positive current (50-190 nA). A holding current of 6 nA was used to minimize leak of ACh from the iontophoretic electrode. A digital average of six ACh responses was used to measure the amplitude of the ACh response under control conditions and in the presence of antagonists. Antagonists were applied in known concentrations by adding the drugs to the superfusing Krebs' solution.

Statistics. Data are expressed as means ± S.E. Student's t test for paired data or ANOVA were used to establish significant differences between control and treatment groups. Agonist concentration-response curves obtained from individual neurons were fit with the following logistic function:
where Y_min and Y_max are the minimum and maximum responses, respectively, and n is the slope factor. Mean values obtained from a number of neurons in different treatment groups were compared with Student's t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Conventional Intracellular Recordings. Recordings of fEPSPs were obtained from 100 neurons. At an initial membrane potential of 96 ± 2 mV, the average amplitude of the fEPSP in these neurons was 23 ± 1 mV. In 37 of these neurons, hexamethonium (C₆) (100 µM) reduced the fEPSP to 4.0 ± 0.4 mV, or by 83 ± 1%. These neurons were not studied further. In 67 of these neurons, C₆ reduced the fEPSP to 15 ± 1 mV, or by 34 ± 2%. The contribution of nerve-released 5-HT to the noncholinergic fEPSP was investigated in the latter group of neurons.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   An ondansetron-sensitive, noncholinergic fEPSP. A, three superimposed fEPSPs recorded in normal Krebs' solution (Control), in the presence of C6, and in the combined presence of C6 and ondansetron. Ondansetron inhibited the C6-resistant component of the fEPSP. B, C6 completely inhibited the depolarization caused by iontophoretically applied ACh (1 M). Histograms represent means + S.E. of data obtained from five neurons. *P < .01 versus control.

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Pharmacological inhibition of fEPSPs recorded from a subset of myenteric neurons. A, inhibition of fEPSP by C6 and tetrodotoxin (TTX). Data are means + S.E. fEPSP amplitude in control conditions and in the presence of C6 and TTX. *Significantly different from control (P < .01; n = 11). B, inhibition of noncholinergic fEPSP by ondansetron. Data are means + S.E. percent inhibition of C6-resistant fEPSPs (middle column in A) caused by the indicated concentrations of ondansetron. *Significantly different from control amplitude (P < .05; n = 6-9 cells/bar).

Whole-Cell Currents Caused by 5-HT and 2-Me-5-HT. 5-HT-induced currents were studied with whole-cell patch-clamp methods to record from myenteric neurons maintained in primary culture for between 7 and 20 days. 5-HT (30 µM) caused an inward current in 95 of 120 (79%) cells tested. The 5-HT-induced current was biphasic, with an initial rapidly developing peak that desensitized in the presence of agonist and a slower developing, sustained inward current (Fig. 3A). At 70 mV, the peak amplitude of the 5-HT-induced current was 517 ± 28 pA (Fig. 3A). Complete concentration-response curves for 5-HT were obtained in seven neurons in which the EC₅₀ for 5-HT was 8 ± 2 µM (Fig. 3B).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   .. Whole-cell currents caused by 5-HT and 2-Me-5HT. A, inward currents caused by 5-HT (left) and 2-Me-5HT (right). Downward deflections are current responses caused by 10 mV hyperpolarizing commands (200-ms duration). Increased amplitude of these currents indicates an increased conductance during the 5-HT response. B, concentration-response curves for 5-HT and 2 Me-5-HT. Data are expressed as a percentage of the response caused by 100 µM 5-HT in each neuron. Data are means ± S.E. obtained from seven neurons for each agonist. When not visible, error bars are smaller than symbols. Curves are best fits with a logistic function (see Materials and Methods). C, inhibition of 5-HT-induced currents by d-TC and by 2-Me-5-HT; d-TC (n = 4) reduced the maximum response caused by 5-HT. 2-Me-5-HT (n = 5) caused a 4-fold rightward shift in the 5-HT concentration-response curve. When not visible, error bars are smaller than symbols.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Currents caused by 5-HT and 2-Me-5HT are inhibited by ondansetron. A, ondansetron (1 µM) inhibits the rapidly developing and desensitizing 5-HT (30 µM)-induced current leaving a small amplitude-sustained current. B, summary of data from experiments shown in A. Columns show mean + S.E. amplitude of the rapidly developing current caused by 5-HT. Ondansetron inhibits the peak current. *Significantly different from control (P < .05). C, ondansetron inhibits the rapidly developing component of the inward current caused by 2-Me-5HT (30 µM). D, summary of data from experiments shown in C. *Significantly different from control (P < .05).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Current-voltage relationship for 5-HT3-mediated responses. A, current recordings obtained at holding potentials between 110 and 70 mV. B, current-voltage plots from experiments similar to that shown in A. Current-voltage relationship is linear, with a reversal potential of 2.7 ± 1.6 mV (n = 6). When not visible, error bars are smaller than symbols.

Kinetics of 5-HT₃-Mediated Whole-Cell Currents. The rate of solution exchange measured by the time to steady state of open-tip junction currents recorded after expelling the neuron from the pipette was 9 ± 1 ms (n = 5) (Fig. 6A). In these neurons, the 10 to 90% rise time of the 5-HT-induced current was 127 ± 19 ms (n = 5). The data described above indicate that 5-HT was acting at two receptors to cause an inward current: a 5-HT₃ receptor and an ondansetron-insensitive receptor. To study the rate of 5-HT₃-receptor desensitization, the response mediated at this receptor needs to be isolated. This was accomplished by recording 5-HT responses in the absence and presence of 1 µM ondansetron (Fig. 6B). Responses recorded in the presence of ondansetron were subtracted from control responses to yield the ondansetron-sensitive current (Fig. 6B). These studies revealed that the 5-HT₃-mediated inward current desensitized with double exponential time course (₁ = 1.1 ± 0.1 s; ₂ = 6.9 ± 0.9 s; n = 5).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Rise time and desensitization of 5-HT3-mediated inward current. A, upper trace shows open-tip current caused by switching the solution exposed to the open pipette tip from normal Krebs' solution to one containing an additional 5 mM NaCl and 5-HT 30 µM. The open-tip current reached steady state in 9 ms. Lower trace shows the rise time of the 5-HT-induced current with the pipette and drug solution shown in the upper trace. The 10 to 90% rise time was 127 ms. B, desensitization of the 5-HT3-mediated inward current. Upper trace (a) shows the total 5-HT current, whereas the middle trace (b) shows the 5-HT-induced current in the presence of ondansetron. The bottom trace shows the subtracted current (a-b), revealing the ondansetron-sensitive current. This protocol was used to determine the rate of desensitization of the 5-HT3-mediated inward current.

5-HT-Induced Single-Channel Currents in Outside-Out Patches. 5-HT (10-30 µM) activated single-channel currents in 18 of 24 (75%) patches tested at a patch potential of 110 mV. Single-channel currents in the absence of 5-HT were rarely observed. When recordings were obtained with ATP and GTP omitted from the pipette solution, the single-channel amplitude at 110 mV was 1.9 ± 0.1 pA (n = 4 patches; Fig. 7, A and B). This value was derived from Gaussian fits of the amplitude/frequency histograms constructed from recordings from four patches and yielded a single-channel conductance of 17 ± 1 pS. A smaller amplitude event of 0.9 ± 0.03 pA (8.2 pS) was observed in some recordings (Fig. 7A); however, this conductance occurred too infrequently to alter the overall amplitude/frequency histogram (Fig. 7A). 5-HT-induced single-channel currents were blocked completely by ondansetron (1 µM) in three patches tested (Fig. 8, A and B). Under control conditions, the open probability in the presence of 10 µM 5-HT was 0.04 ± 0.001, whereas, in the presence of ondansetron, the open probability was 0.0003 ± 0.0003 (P < .05).

View larger version (25K): [in this window] [in a new window]   Fig. 7.   5-HT (10 µM)-activated single-channel currents in outside-out patches. A, amplitude/frequency histogram for single-channel currents in a patch Vh = 110 mV. The average amplitude was 1.9 pA, with a single-channel conductance of 17 pS. Inset shows examples of the predominant single-channel event and a smaller amplitude current that was occasionally observed (arrow). B, amplitude of single-channel currents at the indicated patch potentials. C, current-voltage relationship for 5-HT-activated single-channel currents. The current-voltage relationship was linear, with a mean current reversal potential of 0.03 ± 0.7 mV (n = 10). When not visible, error bars are smaller than symbols.

View larger version (23K): [in this window] [in a new window]   Fig. 8.   Single-channel currents caused by 5-HT are blocked by ondansetron. A, examples of single-channel events caused by 5-HT applied to an outside-out patch in control solution (left) in the presence of solution containing ondansetron (middle) and after ondansetron washout (right). B, amplitude/frequency histograms from recordings shown in A show that ondansetron completely blocks 5-HT-activated currents in a reversible manner.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Drugs acting at 5-HT₃ receptors can alter gastrointestinal motility (Tonini and De Ponti, 1995; Sanger, 1996). Studies in vivo showed that 5-HT₃-receptor antagonists delay gastrointestinal transit in mice and rats (Nagakura et al., 1996; Ito et al., 1997), and ondansetron, may be effective in treating gastrointestinal motility disorders in humans (Sanger, 1996; Wilde and Markham, 1996). 5-HT can also affect peristalsis in vitro (Bulbring and Crema, 1958), and initiation of peristalsis in guinea pig colon involves stimulation of sensory nerves by endogenous 5-HT acting at 5-HT₃ receptors (Foxx-Orenstein et al., 1996). Terminals of intestinal sensory nerves express 5-HT₃ receptors (Hillsley and Grundy, 1998; Bertrand et al., 1998) that are activated by 5-HT released from enterochromaffin cells in response to mucosal stimulation (Gershon, 1995). 5-HT may also contribute to ganglionic transmission responsible for peristalsis. In rabbit colon, ondansetron inhibits descending relaxation of circular muscle caused by distention oral to the recording site (Messori et al., 1995). In dog and guinea pig intestine, 5-HT₃ receptors participate in neurotransmission mediating ascending excitation (Neya et al., 1993; Yuan et al., 1994). However, there has been no direct demonstration of synaptic responses mediated by 5-HT₃ receptors in enteric nerves.

FEPSPs recorded from myenteric neurons are caused by ACh acting at nicotinic receptors and by other transmitters. The noncholinergic component of the fEPSP is mediated largely by ATP acting at P2X receptors, but some fEPSPs persist in the presence of C₆ and P2X-receptor antagonists (Galligan and Bertrand, 1994; LePard et al., 1997; LePard and Galligan, 1999). In 11% of neurons, fEPSPs are inhibited by C₆ and ondansetron or tropisetron. C₆ blocked responses to ACh, whereas neither ondansetron nor tropisetron affected ACh responses. Therefore, C₆-resistant fEPSPs are not due to incomplete block of nicotinic receptors, and ondansetron or tropisetron do not block nicotinic receptors. ACh and 5-HT must mediate fEPSPs recorded from some myenteric neurons.

It is unclear whether ACh and 5-HT are released from the same nerves. 5-HT-containing neurons contain choline acetyltransferase, suggesting that 5-HT and ACh could be cotransmitters (Steele et al., 1991; Young and Furness, 1995). In addition, there are neurons that accumulate 5-HT that also contain choline acetyltransferase (Steele et al., 1991; Meedeniya et al., 1998). The 5-HT/ choline acetyltransferase-containing neurons have long (up to 100 mm), aborally directed projections and comprise 11% of all aborally projecting neurons (Meedeniya et al., 1998). The small number of 5-HT-containing neurons would account for the infrequent observation of 5-HT₃-mediated fEPSPs detected in this study.

5-HT may be released by descending interneurons. Therefore, 5-HT₃-mediated fEPSPs would be recorded from descending interneurons or inhibitory motorneurons receiving input from descending interneurons. 5-HT-containing terminals rarely contact inhibitory motor neurons (Young and Furness, 1995), so it is likely that 5-HT₃-receptor-mediated fEPSPs were recorded from descending interneurons. However, 5-HT₃-receptor antagonists do not block the descending inhibitory component of the peristaltic reflex (Yuan et al., 1994) but do inhibit the ascending excitatory component in the same preparations. It is possible that 5-HT₃-mediated fEPSPs contribute to descending motor responses other than those involved in peristalsis. Alternatively, there may be an unidentified ligand released from ascending interneurons that is an agonist at 5-HT₃ receptors (Yuan et al., 1994).

Properties of 5-HT₃-Mediated Whole-Cell Currents. 5-HT elicited a biphasic inward current in most neurons. This response was characterized by a rapidly developing and desensitizing current and a small, sustained current. Similar results were obtained with the 5-HT₃-receptor agonist 2-Me-5-HT; however, the peak response caused by 2-Me-5-HT was only 20% of that caused by 5-HT. These data indicate that 2-Me-5-HT is a partial agonist at guinea pig enteric 5-HT₃ receptors as in other cells (Butler et al., 1990; Hussy et al., 1994; Niemeyer and Lummis, 1998), and 2-Me-5-HT should block the 5-HT₃ receptor. We found that a low concentration of 2-Me-5-HT caused a rightward shift in the 5-HT concentration-response curve, indicating that 2-Me-5-HT is a 5-HT₃-receptor antagonist in enteric neurons. d-Tubocurare (d-TC) is an antagonist with low nanomolar affinity for 5-HT₃ receptors (Hussy et al., 1994; Jones and Surprenant, 1994). However, the 5-HT₃ receptor in myenteric neurons is resistant to antagonism by d-TC. The mechanism of block of 5-HT₃ receptor by d-TC is unclear, but the structural features that provide d-TC sensitivity must be absent in guinea pig myenteric plexus 5-HT₃ receptors. Ondansetron inhibited peak currents elicited by 5-HT and 2-Me-5-HT. However, in the presence of ondansetron, both agonists induced a sustained inward current. The 5-HT receptor mediating the sustained current was not identified, but the time course of this response is consistent with it being mediated by 5-HT_1P receptors (Gershon, 1995).

Current-Voltage Relationship. Currents activated by 5-HT reversed near 0 mV, consistent with this response being mediated by an increase in a nonspecific cation conductance (Yakel and Jackson, 1988; Derkach et al., 1989). The current-voltage relationship was linear and similar to that obtained from heterologously expressed guinea pig enteric 5-HT₃ receptors (Lankiewicz et al., 1998). However, 5-HT₃ receptors from human, rat, and mouse tissues exhibit marked inward rectification (Hussy et al., 1994; Fletcher and Barnes, 1998; Lankiewicz et al., 1998). This may be because of intrinsic differences in the conducting properties of the 5-HT₃-receptor subunit(s) expressed by different cells or modifications caused by unidentified 5-HT₃-receptor subunits (Fletcher and Barnes, 1998).

Kinetics of 5-HT₃-Receptor-Mediated Current. To study the properties of currents caused by 5-HT₃-receptor activation, the contribution of the sustained current to the total current needed to be removed. Because the receptor mediating the sustained current was not identified, it was not possible to block this response pharmacologically. However, the contribution of the sustained current to the total 5-HT-induced current was removed by subtracting ondansetron-insensitive currents. The 5-HT₃-mediated inward current reached a peak in <150 ms, similar to that for long- and short-form guinea pig enteric 5-HT₃ receptors expressed in HEK293 cells (Lankiewicz et al., 1998). The response mediated at native 5-HT₃ receptors in myenteric neurons desensitized with a double exponential time course, whereas desensitization of responses mediated at heterologously expressed guinea pig enteric 5-HT₃ receptors decayed with a linear time course. There are several explanations for the difference in desensitization kinetics between native and heterologously expressed 5-HT₃ receptors. Desensitization of 5-HT₃ receptors may be regulated by receptor phosphorylation (Yakel and Jackson, 1988; Boddeke et al., 1996), and intracellular regulatory mechanisms in HEK293 cells may differ from those in myenteric neurons. In addition, the subunit composition of native 5-HT₃ receptors may be different from receptors composed of single subunits expressed heterologously (van Hooft et al., 1997). Native receptors may have regulatory sites absent from expressed receptors. Finally, simultaneous activation of other 5-HT receptors expressed by myenteric neurons could activate intracellular mechanisms that regulate 5-HT₃-receptor function.

Single-Channel Currents. 5-HT activated single-channels in outside-out patches whose properties were unaffected by the absence or presence of ATP and GTP in the recording pipette solution. In addition, channel activation caused by 5-HT was blocked by ondansetron. These data indicate that 5-HT-activated single-channel currents are mediated by directly-gated 5-HT₃ receptors.

Conclusions. These studies showed that 5-HT₃ receptors contribute to fEPSPs at some synapses in guinea pig small intestinal myenteric plexus. Based on the known aboral projection of 5-HT-containing neurons in the myenteric plexus, the 5-HT₃ synapses are likely to play a role in descending motor or secretomotor responses. The properties of myenteric 5-HT₃ receptors differ in their rate of desensitization, current-voltage relationship, and gating properties from nicotinic and P2X receptors (Zhou and Galligan, 1996b). P2X and nicotinic receptors also mediate fEPSPs in the myenteric plexus. These different properties may be important during bursts of fEPSPs or during coactivation of fast and slow excitatory synaptic pathways.