Differential Composition of 5-Hydroxytryptamine3Receptors Synthesized in the Rat CNS and Peripheral Nervous System (original) (raw)

ARTICLE, Behavioral/Systems

Journal of Neuroscience 1 August 2002, 22 (15) 6732-6741; https://doi.org/10.1523/JNEUROSCI.22-15-06732.2002

Loading

Abstract

The type 3 serotonin (5-HT3) receptor is the only ligand-gated ion channel receptor for serotonin in vertebrates. Two 5-HT3 receptor subunits have been cloned, subunit A (5-HT3A) and subunit B (5-HT3B). We used in situ hybridization histochemistry and reverse transcriptase-PCR amplification to demonstrate that 5-HT3Asubunit transcripts are expressed in central and peripheral neurons. In contrast, 5-HT3B subunit transcripts are restricted to peripheral neurons. Thus, the prevalent form of 5-HT3receptor synthesized within the CNS lacks the 5-HT3Bsubunit. Because coexpression of 5-HT3A and 5-HT3B subunits produces heteromeric 5-HT3A/3Breceptors with properties that differ from those of 5-HT3Ahomomeric receptors, we investigated possible coexpression of both subunits at the cellular level. We found that near to 90% of all 5-HT3B expressing neurons coexpress the 5-HT3Asubunit in superior cervical and nodose ganglia (NG). In addition, there is a cellular population that expresses only the 5-HT3A subunit. Therefore, peripheral neurons have the capacity to synthesize two different 5-HT3 receptors, 5-HT3A+/3B− and 5-HT3A+/3B+ receptors. We also determined that neurons of NG projecting to the nucleus tractus solitarium and those of dorsal root ganglia projecting to superficial layers of the spinal cord express 5-HT3A or 5-HT3A/3B subunits. Thus, presynaptic 5-HT3 receptors containing the 5-HT3B subunit might be present in these target brain areas. The compartmentalized structural composition of the 5-HT3 receptor may be the basis of functional diversity within this receptor. This raises the possibility that 5-HT3 receptors participating in sympathetic, parasympathetic and sensory functions may be functionally different from those involved in cognition and emotional behavior.

The 5-HT3receptor is the only ligand-gated ion channel receptor for serotonin in vertebrates (Derkach et al., 1989). This receptor modulates visceral afferent information and visceral reflexes, participates in nociception and cognition (for review, see Fozard, 1992), and has been suggested to play a role in the biology of drugs of abuse (for review, see Grant, 1995; Lovinger, 1999).

Binding studies have shown 5-HT3 receptor binding sites in the CNS of rodents and primates (Kilpatrick et al., 1987;Waeber et al., 1988, 1989, 1990; Barnes et al., 1989; Pratt et al., 1990; Gehlert et al., 1991; Jones et al., 1992; Laporte et al., 1992). In recent years, cellular analysis of the pattern of distribution of the functional 5-HT3 receptor subunit A (5-HT3A) demonstrated 5-HT3A mRNA (Tecott et al., 1993; Morales et al., 1996b; Morales and Bloom, 1997) and protein (Morales et al., 1996a,1998) in several brain areas shown previously to contain 5-HT3 receptor binding sites. Neurons of peripheral ganglia are also known to contain 5-HT3 receptor binding sites (Hoyer et al., 1989;Kilpatrick et al., 1989) and 5-HT3A subunit transcripts (Tecott et al., 1993; Rosenberg et al., 1997).

In addition to the 5-HT3A subunit, which has been cloned from tissues of several animal species including humans (Maricq et al., 1991; Hope et al., 1993; Belelli et al., 1995; Miyake et al., 1995), a new class of 5-HT3 receptor subunit (5-HT3B) has also been cloned recently (Davies et al., 1999; Dubin et al., 1999; Hanna et al., 2000). In contrast to results with the 5-HT3A subunit, expression of recombinant 5-HT3B subunit alone does not produce a functional 5-HT3 receptor. However, in heteromeric receptor complexes, the 5-HT3Bsubunit confers unique pharmacological and biophysical properties. Coexpression of 5-HT3A and 5-HT3B subunits in Xenopus oocytes and mammalian cell lines yields receptors with a large single-channel conductance, low permeability to calcium ions, and a linear current–voltage relationship (Davies et al., 1999; Dubin et al., 1999).

Electrophysiological studies have indicated heterogeneous properties for native 5-HT3 receptors (Derkach et al., 1989;Peters et al., 1992; Yang et al., 1992; Hussy et al., 1994; Brown et al., 1998). Although the basis of this heterogeneity is unknown, it has been speculated that receptor post-transcriptional modification such as phosphorylation or subunit composition might account for this diversity. Thus, knowledge of expression patterns for 5-HT3A and 5-HT3B subunit genes is fundamental for understanding the possible structural composition of the 5-HT3 receptors present in different cells of the nervous system. To address this question, we first investigated whether the mRNA for the 5-HT3B subunit was present in the brain, spinal cord, and peripheral ganglia. Moreover, because coexpression of 5-HT3A and 5-HT3B subunits in recombinant preparations produces heteromeric 5-HT3A/3B receptors with properties that differ from those of 5-HT3A homomeric receptors, we sought to determine possible coexpression of both subunits at the cellular level. Finally, a combination of in situ hybridization and retrograde tracing was used to investigate whether peripheral neurons that express transcripts for the 5-HT3B subunit, in addition to those of the 5-HT3A subunit, project to specific areas of the CNS.

MATERIALS AND METHODS

Tissue preparation. Adult Sprague Dawley male rats (200–250 gm body weight) were anesthetized with chloral hydrate (35 mg/100 gm) and perfused transcardially with a solution of 4% (w/v) paraformaldehyde in 0.1 m phosphate buffer (PB), pH 7.3. Brains, spinal cord, and peripheral ganglia were postfixed in fresh fixative for 15 hr at 4°C, rinsed with PB, and sequentially transferred to 12, 14, and 18% sucrose solutions. Material was frozen on dry ice and then sectioned with a cryostat. All animal procedures used were approved by the National Institute on Drug Abuse Animal Care and Use Committee.

In situ hybridization. In situ hybridization was performed as described previously (Morales and Bloom, 1997). Free-floating (25 μm) cryosections of brain tissue and spinal cord and 8–15 μm sections of peripheral ganglia on glass slides were incubated for 10 min in PB containing 0.5% Triton X-100, rinsed two times for 5 min with PB, treated with 0.2N HCl for 10 min, rinsed two times for 5 min with PB, and then acetylated in 0.25% acetic anhydride in 0.1 m triethanolamine, pH 8.0, for 10 min. Sections were rinsed two times for 5 min with PB, postfixed with 4% paraformaldehyde for 10 min, rinsed with PB, dehydrated, and hybridized at 55°C for 16 hr in hybridization buffer (50% formamide, 10% dextran sulfate, 5× Denhardt's solution, 0.62 mNaCl; 50 mm DTT, 10 mmEDTA, 20 mm PIPES, pH 6.8, 0.2% SDS, 250 μg/ml single-stranded DNA, and 250 μg/ml tRNA) containing [35S]- and [33P]-labeled single-stranded RNA probes at 107 cpm/ml. Sense and antisense riboprobes were prepared for rat 5-HT3A subunit [nucleotides 1500–2230 of the rat 5-HT3Asubunit and rat 5-HT3B subunit (nucleotides 335–1346, accession number AF155044; nucleotides 1–1948 and 739–1948, accession number AF303447)] and mouse 5-HT3B subunit (nucleotides 83–1373, accession number AF155045). Sections were treated with 4 μg/ml RNase A at 37°C for 1 hr, washed with 1× SSC and 50% formamide at 55°C for 1 hr, and washed with 0.1× SSC at 68°C for 1 hr. Sections were rinsed with PB. Free-floating sections were mounted on coated slides, air dried, dipped in nuclear track emulsion, and exposed for several weeks at 4°C before development. Sections were counterstained with toluidine blue and analyzed in a Nikon (Tokyo, Japan) Microphot-FX microscope. Material was analyzed and photographed using bright-field or dark-field microscopy.

As control for in situ hybridization specificity, sequential sections were incubated with sense or antisense radioactive riboprobes. In another type of control, sequential sections were incubated with antisense radioactive riboprobes in the absence or presence of a 100-fold excess of nonradioactive antisense riboprobe. Silver grains were scattered when hybridization was performed with radioactive sense riboprobes or with an excess of nonradioactive antisense riboprobes. This level of signal was very low and was considered unspecific background.

Reverse transcriptase-PCR. Adult Sprague Dawley male rats (200–250 gm body weight) were anesthetized with chloral hydrate (35 mg/100 gm). Brain and ganglia [nodose ganglia (NG), trigeminal ganglia (TG), superior cervical ganglia (SCG), and dorsal root ganglia (DRG)] were immediately removed and transferred to buffer RNAlater (Ambion, Austin, TX). Specific brain areas were dissected out. Entire brains, selected brain areas, and different ganglia were placed in individual Eppendorf tubes containing RNeasy lysis buffer (RNeasy kit; Qiagen, Valencia, CA) and homogenized with a rotor-stator homogenizer. Total RNA samples were isolated using the Qiagen Rneasy mini kit. cDNA synthesis and amplification of 5-HT3A and 5-HT3B subunits were performed using the PCR access kit (Promega, Madison, WI). Reverse transcriptase (RT)-PCR amplification was performed three times using duplicates of RNA samples of central and peripheral tissues. The sequences of primers used to amplify the 5-HT3 subunits were: ATCCAGGACATCAACATTTCCCTGTGGCGAACA and GTCTCAGCGAGGCTTATCACCAGCAGAG for the 5-HT3A subunit and GTGGAAGACATAGACCTGGGCTTCCTGAG and ACCCTGCGCTTCTTGGCACCTCATCAGA for the 5-HT3B subunit.

Retrograde tracing. Adult Sprague Dawley male rats (200–250 gm body weight) were anesthetized with ketamine/xylazine (40 mg/kg ketamine and 10 mg/kg xylazine, i.p.) and placed on a stereotaxic frame. A microinjector needle was lowered into the nucleus tractus solitarium (NTS) (Obex was used as reference point; NTS was located at 1 mm lateral to midline and 1 mm depth from brain surface) or into the superficial layer of the dorsal horn of the spinal cord. The tracer fluorogold (0.2–0.3 μl of 4% fluorogold in saline solution) was unilaterally injected into the NTS or into the superficial layer of dorsal horn. The injector needle was then removed. After suturing the wound and recovery from anesthesia, the animals were placed in their home cages for 6 d. Animals were subsequently anesthetized with chloral hydrate (35 mg/100 gm) and perfused transcardially as indicated under Tissue preparation. Ganglia were removed (NG from rats injected into the NTS and DRG from rats injected into the spinal cord), and 10-μm-thick cryosections were mounted on glass slides.

Data analysis. Toluidine blue-counterstained sections (10–15 μm thickness) were used to calculate the percentage of labeled neurons. Each alternate section was hybridized with antisense riboprobes for detection of either 5-HT3B or 5-HT3A subunit transcripts. Data were collected using a Nikon Eclipse E800 microscope with a ×20 objective lens, equipped with an Optronics International (Chelmsford, MA) video camera, and connected to a C-Imaging (Compix, Inc., Cranberry Township, PA) workstation. Image processing and analysis were done using C-Imaging Systems software (Compix, Inc.). The percentage of labeled neurons was calculated using two alternate sections from five corresponding ganglia of different rats; all neurons in a given field were outlined. Cell diameter was determined from labeled cells that contained a visible nucleus.

Five sets of two alternate sections (8 μm thickness) from five NG and three sets of three SCG were used to calculate the degree of coexpression of mRNA for 5-HT3B and 5-HT3A subunits at the cellular level. Each alternate section was hybridized with antisense riboprobes for detection of either 5-HT3B or 5-HT3A subunit transcripts. Before data analysis, slides were examined under bright-field and epiluminescence microscopy to assess the quality of the tissue and that of the autoradiographic signal. Selected slides were analyzed as described above to determine labeled neurons. This information was used to identify the same cells in micrographs at a final magnification of 40×. Transparency films were overlaid on micrographs, corresponding to either 5-HT3A or 5-HT3B subunit, to outline labeled neurons. Transparency films with outlines of the 5-HT3A subunit-labeled neurons were overlaid on micrographs corresponding to 5-HT3Bsubunit-labeled sections to match outlines of 5-HT3A subunit-labeled neurons with corresponding 5-HT3B subunit-labeled neurons. A similar procedure was followed to match outlines of 5-HT3B subunit-labeled neurons with 5-HT3A subunit-labeled neurons.

To calculate the distribution of 5-HT3B or 5-HT3A subunit transcripts within the total population of neurons labeled with fluorogold, before dipping slides in nuclear track emulsion, images of slides of DRG and NG were collected with a video camera under ultraviolet light. After exposure of slides for several weeks, images of DRG and NG were collected under dark-field microscopy. Both images were overlaid using Adobe Photoshop software (Adobe Systems, San Jose, CA), and the total population of fluorogold-labeled neurons with or without transcripts was obtained.

RESULTS

Expression of 5-HT3A but not 5-HT3B subunit in neurons of the CNS

Several riboprobes complementary to the rat 5-HT3A and 5-HT3B subunits (see Materials and Methods) were used to determine the pattern of expression of both subunits within the CNS. To increase sensitivity for mRNA detection, in situ hybridization histochemistry was performed in free-floating tissue sections using [35S]- and [33P]-labeled riboprobes. Regardless of hybridization conditions, 5-HT3A subunit but not 5-HT3B subunit was detected in neurons of several brain areas (Figs. 1, 2) and spinal cord (Fig. 3). At low magnification, silver grains slightly above background levels were seen in the pyramidal layer of hippocampus and olfactory tubercle in samples hybridized with riboprobes specific for detection of 5-HT3B subunit. However, at higher magnification silver grains were not clearly associated with cell bodies.

Fig. 1.

Fig. 1.

Phosphoimages comparing regional expression of mRNA encoding 5-HT3A and 5-HT3B subunits in the rat brain. A–D, Widespread expression of 5-HT3A subunit throughout the rat brain.A′–D′, Lack of detection of 5-HT3B subunit transcripts in adjacent brain sections. Acb, Nucleus accumbens; ARH, arcuate nucleus of the hypothalamus;CA1, field CA1 of the hippocampus; CA3, field CA3 of the hippocampus; Cg, cingulate cortex;COA, cortical nucleus of the amygdala;CPu, caudate putamen; DEn, dorsal endopiriform nucleus; DG, dentate gyrus;DTT, dorsal tecnia tecta; FS, fundus of the striatum; M, motor cortex; MH, medial habenular nucleus; Pir, piriform nucleus;PrL, prelimbic cortex; S, septum;SS, somatosensory cortex; Tu, olfactory tubercle; VMH, ventromedial hypothalamic nucleus. Scale bar, 1.4 mm.

Fig. 2.

Fig. 2.

Autoradiograms comparing cellular expression of mRNA encoding 5-HT3A and 5-HT3B subunits in sagittal sections of the rat brain. Bright-field (A–F) and dark-field (A′–F′) microscopy are shown. A, A′, High expression of 5-HT3A mRNA was detected in neurons of the subiculum (S; small arrows) and basket cells (large arrows) in the subgranular layer of the dentate gyrus (large arrows). B, B′, C, C′, Prominent expression of 5-HT3A mRNA was found in cortical neurons (arrows) distributed primarily in layer II. B, B′, Visual cortex. C, C′, Entorhinal cortex. Lack of expression of 5-HT3B mRNA in the hippocampal formation (D, D′), visual cortex (E, E′), and entorhinal cortex (F, F′) is shown.CPu, Caudate putamen; DG, dentate gyrus;Ent, entorhinal cortex; Pir, piriform cortex; S, subiculum; SS, somatosensory cortex; VC, visual cortex. Scale bar: bright-field sagittal sections, 750 μm; A–F, A′–F′, 75 μm.

Fig. 3.

Fig. 3.

Autoradiograms comparing cellular expression of mRNA encoding 5-HT3A and 5-HT3B subunits in the spinal cord. Bright-field (A, B) and dark-field (A′, B′) microscopy is shown. A′, Expression of 5-HT3A subunits in neurons distributed in the dorsal (small arrows) and ventral (large arrows) horn of the spinal cord. B′, Lack of detection of 5-HT3B subunit transcripts in an adjacent section. Scale bar, 290 μm.

Expression of 5-HT3A and 5-HT3B subunit in neurons of the PNS

Because previous studies have shown that the 5-HT3 receptor is abundant in peripheral neurons, we used in situ hybridization to determine whether 5-HT3A and 5-HT3B subunits were present in neurons of peripheral ganglia. In contrast to results obtained for neurons of the CNS, both 5-HT3A and 5-HT3B subunits were found in neurons of the NG (Fig. 4), SCG, TG, DRG, and myenteric plexus in the gastrointestinal tract. All tested antisense 5-HT3B subunit riboprobes revealed strong expression of 5-HT3B subunit in peripheral neurons. As controls for specificity of the _in situ_hybridization signal, addition of either a 100-fold excess of nonradioactive antisense riboprobes or hybridization with radioactive sense riboprobes were performed. Both conditions produced signal at background levels.

Fig. 4.

Fig. 4.

Expression of 5-HT3A and 5-HT3B subunits in peripheral ganglia. Bright-field (A–D) and dark-field (A′–D′) microscopy are shown. A, A′, B, B′, SCG. C, C′, D, D′, NG. Expression of 5-HT3A subunit was detected in neurons of the SCG (A′) and NG (C′). Neuronal expression of 5-HT3B subunit was found in the SCG (B′) and NG (D′). Scale bar, 175 μm.

RT-PCR amplification of 5-HT3A and 5-HT3Bsubunit transcripts

RT-PCR assays were performed to amply transcripts for 5-HT3A and 5-HT3B subunits using RNA isolated from the entire brain, anterior olfactory nucleus, frontal cortex, striatum, hippocampus, and peripheral ganglia (TG, SCG, NG, and DRG). Although transcripts for the 5-HT3Asubunit were amplified from RNA samples of the entire brain, specific brain areas and peripheral ganglia amplification of transcripts for the 5-HT3B subunit was obtained only from peripheral ganglia RNA (Fig. 5). Thus, RT-PCR results confirm data derived from in situ hybridization analysis showing expression of mRNA for 5-HT3Asubunit in central and peripheral neurons, whereas the 5-HT3B subunit mRNA was detected only in peripheral neurons.

Fig. 5.

Fig. 5.

Amplification of 5-HT3A and 5-HT3B subunit transcripts from RNA of the CNS and PNS. RT-PCR amplification of 5-HT3A and 5-HT3Bsubunit transcripts using RNA isolated from the entire brain, anterior olfactory nucleus (AON), frontal cortex (FC), striatum, hippocampus (Hipp), TG, SCG, NG, and DRG is shown. PCR products of the predicted size (425 bp) for 5-HT3A transcripts (3A) were obtained from RNA of central and peripheral origin. However, PCR products of predicted size (354 bp) for 5-HT3B transcripts (3B) were amplified only from peripheral RNA.

Coexpression of 5-HT3A and 5-HT3B subunits in peripheral neurons

Cellular analysis of neurons expressing 5-HT3A subunit (Fig.6) demonstrated that approximately one-half of the total population of neurons expressed the 5-HT3A subunit in TG (45.73 ± 6.56%, mean ± SEM; 1557 neurons) and SCG (45.73 ± 3.72%; 1957 neurons). In contrast, a larger population of 5-HT3A-expressing neurons was found in the NG (Fig. 6), where 88.64 ± 1.99% (1435 neurons) of the total population of neurons expressed the 5-HT3Asubunit (Fig. 6). A smaller population of peripheral neurons expressed the 5-HT3B subunit; approximately one-third of the total population of neurons expressed the 5-HT3B subunit in the SCG (29.95 ± 3.69%; 1489 neurons) and TG (35.99 ± 3.89%; 2551 neurons). As with the 5-HT3A subunit, the NG had the highest proportion of 5-HT3B-expressing neurons (51.11 ± 2.41%; 1804 neurons). Thus, within the three ganglia, the population of neurons expressing the 5-HT3Asubunit was larger than the one expressing 5-HT3B; the difference between these two populations was 15.78% for TG, 9.74% for SCG, and 37.53% for NG.

Fig. 6.

Fig. 6.

Frequency histogram showing percentage of neurons expressing either 5-HT3A or 5-HT3B subunits in the total neuronal population of TG, SCG, and NG. The data are presented as mean ± SEM; n = number of cells.

Additional cellular analysis of cell diameter and expression of 5-HT3A and 5-HT3B subunits showed that, in the three ganglia, both subunits were expressed in neurons of the same diameter (Table 1), suggesting that the two subunits are coexpressed in some neurons. Thus, serial sections of NG (Fig. 7) and SCG were used to determine the degree of coexpression of 5-HT3A and 5-HT3B subunits within single neurons. Analysis of the proportion of neurons coexpressing 5-HT3A/5-HT3Bsubunits in the total population of 5-HT3B-expressing neurons showed that the vast majority (93.67–96%) of 5-HT3B-labeled neurons also expressed the 5-HT3A subunit in the SCG (93.67 ± 1.45%; 143 neurons) and NG (96 ± 0.89%; 330 neurons). Comparison of these results with those detailed above (Fig.6) indicated that of the total population of neurons containing 5-HT3A subunit, a small proportion (9.74% for SCG and 37.53% for NG) lacks the 5-HT3Bsubunit.

Table 1.

Cell diameter of neurons expressing mRNA for the 5-HT3A and 5-HT3B subunits in peripheral ganglia1-a

Fig. 7.

Fig. 7.

Coexpression of 5-HT3A and 5-HT3B subunits in the NG. Pairs of micrographs of adjacent serial sections show expression of 5-HT3A (A, A′) and 5-HT3B (B, B′) subunits. Bright-field (A, B) and dark-field (A′, B′) microscopy is shown. Arrows indicate examples of neurons coexpressing 5-HT3A and 5-HT3B subunits. Scale bar, 50 μm.

Peripheral neurons expressing 5-HT3A and 5-HT3B subunits innervate specific areas of the CNS

We have found previously that 5-HT3A and 5-HT3B subunits are expressed in the DRG in a pattern similar to that of the NG (Morales et al., 2001). Because both ganglia project to the PNS and CNS, a combination of in situ hybridization and retrograde labeling was used to determine whether DRG and NG neurons expressing 5-HT3A or 5-HT3B subunits project to defined target areas in the CNS. Within the DRG (Table 2), nearly one-half of all fluorogold-labeled neurons expressed the 5-HT3A subunit (41.91 ± 0.06%), and approximately one-third expressed the 5-HT3Bsubunit (28.65 ± 0.03%). For the NG (Table 2), one-third of all fluorogold-labeled neurons expressed the 5-HT3Asubunit (33.88 ± 0.06%), and one-quarter expressed the 5-HT3B subunit (26.03 ± 0.05%). Expression of 5-HT3A and 5-HT3Bsubunits was often seen within the same retrogradely labeled neuron (Fig. 8), indicating coexpression of both subunits in neurons of the NG projecting to the NTS and neurons of the DRG innervating superficial layers of the spinal cord.

Table 2.

Percentage of neurons expressing mRNA for 5-HT3A or 5-HT3B subunits in the total population of fluorogold-labeled neurons in DRG and NG

Fig. 8.

Fig. 8.

Detection of 5-HT3A and 5-HT3B transcripts in fluorogold retrogradely-labeled neurons in the NG. A, B, Fluorogold-labeled neurons.A′, B′, Dark-field microscopy indicating detection of transcripts encoding 5-HT3A (A′) or 5-HT3B (B′) subunits. _Arrows_indicate examples of retrogradely labeled neurons coexpressing 5-HT3A and 5-HT3B transcripts. Scale bar, 60 μm.

DISCUSSION

Differential expression of 5-HT3A and 5-HT3B subunits in the CNS and PNS

In the present study, we used in situ hybridization histochemistry and RT-PCR amplification to demonstrate that 5-HT3A subunit transcripts are expressed in central and peripheral neurons in the rat. In contrast, 5-HT3B subunit transcripts are restricted to peripheral neurons. The lack of detectable levels of mRNA encoding the 5-HT3B subunit in neurons of the CNS suggests that 5-HT3 receptors synthesized in the CNS might be 5-HT3A homomeric receptors or heteromeric receptors containing 5-HT3A subunits in combination with subunits different from the 5-HT3B subunit. These putative subunits may participate in the formation of 5-HT3 receptors with high conductance, such as those reported for hippocampal primary neurons in culture (Jones and Surprenant, 1994). The absence of 5-HT3B subunit mRNA in rat central neurons might reflect a species-specific difference in the pattern of expression of this subunit, because 5-HT3B subunit mRNA in human brain tissue has been detected by Northern blot analysis (Davies et al., 1999) and RT-PCR amplification (Dubin et al., 1999).

Comparison of the cellular population containing 5-HT3A and 5-HT3B mRNA subunits demonstrated that the population of peripheral neurons expressing the 5-HT3A subunit is larger than the one expressing 5-HT3B subunit. Moreover, analysis of the proportion of neurons coexpressing both subunits showed that >90% of 5-HT3B-labeled neurons coexpressed the 5-HT3A subunit. It is not clear whether the small number of 5-HT3B-expressing cells lacking 5-HT3A signal represent neurons with levels of 5-HT3A transcripts below the detectable threshold of in situ hybridization or express a 5-HT3 subunit distinct from the 5-HT3A. However, coexpression of 5-HT3A/3B subunits in most of the 5-HT3B-expressing cells indicates that these peripheral neurons have the potential to synthesize heteromeric 5-HT3A/3B receptors. In addition, there is a neuronal population containing 5-HT3A transcripts that lacks expression of the 5-HT3B subunit. These results provide anatomical evidence suggesting that at least two subpopulations of cells are present in the PNS: one having the potential for synthesizing homomeric 5-HT3Areceptors and the other for producing heteromeric 5-HT3A/3B receptors. Consistent with this suggestion, electrophysiological evidence indicates that 5-HT3 receptors of different conductance are present in the SCG (Yang et al., 1992; Hussy et al., 1994). In this regard, coexpression of 5-HT3A and 5-HT3B subunits in Xenopus oocytes and mammalian cell lines yields receptors with a large channel conductance and low permeability to calcium ions (Davies et al., 1999; Dubin et al., 1999), whereas the 5-HT3A monomeric receptors display inwardly rectifying currents and low single-channel conductance (Davies et al., 1999; Dubin et al., 1999). These electrophysiological observations are in agreement with our anatomical results and support the notion that homomeric 5-HT3A and heteromeric 5-HT3A/3B receptors constitute functional distinct receptors in the peripheral ganglia.

Peripheral neurons expressing 5-HT3A and 5-HT3B subunits innervate specific areas in the CNS

We have found previously that 5-HT3A and 5-HT3B subunits are expressed in the DRG in a proportion similar to that of the NG (Morales et al., 2001). Because both ganglia project to the CNS in addition to the PNS, we sought to determine whether DRG neurons innervating superficial layers of the dorsal horn and NG neurons projecting to the NTS express 5-HT3A or 5-HT3B subunits. We found that approximately one-quarter of all retrogradely labeled neurons of the DRG and NG express the 5-HT3Bsubunit and that more than one-third express the 5-HT3A subunit. Because one-half (DRG) to one-third (NG) of all 5-HT3A subunit-containing neurons expressed the 5-HT3B subunit and >90% of all 5-HT3B subunit-labeled neurons have the 5-HT3A subunit, these results suggest that DRG and NG central nerve endings might contain homomeric (5-HT3A) and heteromeric (5-HT3A/3B) receptors (Fig.9). Thus, despite the lack of detection of 5-HT3B subunit mRNA in the CNS (Fig. 9), 5-HT3B subunit protein might be present in areas innervated by NG (i.e., the NTS) (Fig. 9) and DRG (i.e., superficial layers of the dorsal horn) (Fig. 9). Although these results support the suggestion that terminals from peripheral neurons with different structures will innervate the CNS, the compartmentalization of these different receptors at the molecular and structural level remains to be determined.

Fig. 9.

Fig. 9.

The compartmentalized structural composition of the 5-HT3 receptors may be the basis of pharmacological and electrophysiological diversity within this receptor. The 5-HT3A but not 5-HT3B subunit mRNA (5-HT3A+/3B−) is expressed in central neurons [for example, neurons of the dorsal horn (DH) and ventral horn (VH) of the spinal cord and neurons of the hypoglossal nucleus (12)]. However, two major subpopulations of neurons are present in the periphery: one that coexpresses 5-HT3A and 5-HT3B subunits [(5-HT3A+/3B+); see neurons 1, 3, and 4 and 1′, 3′, and 4′] and another expressing only 5-HT3A subunit [(5-HT3A+/3B−); see neurons 2 and 2′]. Both types of neurons are present in DRG (A; 1–4) and NG (B; 1′–4′) and might have central and peripheral targets. These results suggest that central nerve endings of peripheral neurons might contain homomeric (5-HT3A) and heteromeric (5-HT3A/3B) receptors. Thus, despite the lack of detection of 5-HT3B subunit mRNA in the CNS, 5-HT3B subunit protein might be present in central areas innervated by DRG (i.e., superficial layers of the spinal cord) and NG (i.e., the NTS). Central and peripheral compartmentalization of receptors containing or lacking 5-HT3B subunit might occur in neurons endowed with both subunits (see neurons 1, 1′, 4, and 4′). We cannot discount the possibility that additional as yet unidentified subunits (5-HT3x) might be present in central and peripheral neurons. AP, Area postrema;gr, gracile fasciculus; 10, dorsal motor nucleus of the vagus.

Functional implications of the presence of homomeric and heteromeric 5-HT3 receptors in the nervous system

Studies with recombinant preparations indicate that 5-HT3 receptors with distinct structural composition result in receptors with different pharmacological and biophysical characteristics. Thus, knowledge of expression patterns for endogenous 5-HT3A and 5-HT3B subunit transcripts and proteins will be helpful in understanding the possible structural composition of 5-HT3 receptors present in different cells of the nervous system. In this regard, the lack of detection of 5-HT3B mRNA in the rat central neurons indicates that the 5-HT3B subunit is not a prominent component of the 5-HT3 receptor synthesized in the rat CNS. However, it does not discard the possibility that heteromeric 5-HT3A/3B receptors of peripheral origin, such as those from NG and DRG, might be present in nerve endings innervating the NTS and dorsal horn (Fig. 9). The compartmentalized structural composition of the 5-HT3 receptors may be the basis of pharmacological and electrophysiological diversity within native 5-HT3 receptors. The distinct distribution of 5-HT3A and5-HT3B subunits indicates that 5-HT3 receptors originating in central or peripheral neurons with specific pharmacological and electrophysiological properties are likely to participate in different neuronal pathways and animal behaviors.

The 5-HT3 receptor modulates visceral afferent information and visceral reflexes, participates in nociception and cognition (for review, see Fozard, 1992), and has been suggested to play a role in the biology of drugs of abuse (for review, see Grant, 1995; Lovinger, 1999). It is well established that vagal afferent fibers are the main source of 5-HT3 receptors in the NTS, because ablation of the NG in the rat and ferret results in a dramatic reduction of 5-HT3 receptor binding sites in the NTS (Pratt and Bowery, 1989; Leslie et al., 1990). The 5-HT3 receptors in the NTS and gastrointestinal tract have been suggested as the site of action of 5-HT3 receptor antagonists used in the treatment of emesis associated with cancer chemotherapy (Costall et al., 1986;Miner and Sanger, 1986; Andrews et al., 1988; Leslie et al., 1990;Naylor and Rudd, 1996). Although most of the interest on vagal 5-HT3 receptors has been focused on their role in emesis, a more widespread role in the modulation of visceral afferent information and visceral reflexes might be expected, because functional studies in the rat have shown that 5-HT3receptors located on sensory efferent fibers within the NTS regulate blood pressure (Merahi et al., 1992; Veelken et al., 1993) and heart rate (Veelken et al., 1993). Within the superficial layers of the dorsal horn, 5-HT3 receptor-binding sites are prominent (Glaum and Anderson, 1988; Gehlert et al., 1991; Laporte et al., 1992) and represent, in part, primary sensory terminals that contain 5-HT3 receptors (Hamon et al., 1989; Kidd et al., 1993). In addition to 5-HT3 receptors distributed on primary afferents, intrinsic interneurons may contribute to the pool of 5-HT3 receptors present in the dorsal horn (Alhaider et al., 1991), because these interneurons were found to express the 5-HT3A (Morales et al., 1998) but not 5-HT3B subunit (present study). Interestingly, these observations imply that 5-HT3 receptors in the dorsal horn may represent receptors not only of different origins but also of different composition. It has long been recognized that 5-HT3 receptors present on sensory nerve terminals are involved in serotonin-induced pain (Richardson et al., 1985; Giordano and Rogers, 1989; Glaum et al., 1990). In this regard, detection of the 5-HT3A subunit in DRG neurons of different sizes and coexpression of 5-HT3A and 5-HT3B subunits in medium and large neurons (Morales et al., 2001) suggest that 5-HT3receptors of different structures may convey nociceptive or proprioceptive information.

The specific contribution of homomeric (5-HT3A) versus heteromeric (5-HT3A/3B) receptors in functions mediated by the PNS and CNS is difficult to evaluate, because neither 5-HT3B knock-out mice nor specific antagonists capable of differentiating these receptors are as yet available. However, the participation of the 5-HT3B subunit in sympathetic, parasympathetic, and sensory functions is underscored by the high levels of expression of this subunit in SCG, NG, TG, and DRG. We suggest that contrary to 5-HT3 receptors originating in the PNS, 5-HT3 receptors synthesized in the CNS participating in cognition and emotional behavior will have unique structural and functional properties. Information on the distribution, electrophysiological and pharma-cological characteristics, and possible participation of the different 5-HT3receptors in peripheral and central circuits will be useful for the development of specific antiemetic and analgesic drugs. This characterization will also be important to further evaluate the controversial participation of the 5-HT3 receptor in the neuronal effects of drugs of abuse.

Footnotes