Distribution of serotonin receptors and interacting proteins in the human sigmoid colon


Helen Irving, Medicinal Chemistry and Drug Action, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville campus, Victoria 3052, Australia.
Tel: +61 3 9903 9565; fax: +61 3 9903 9638; e-mail: helen.irving@pharm.monash.edu.au


Abstract  This study aimed to examine the distribution of 5-HT receptors in the human colon. 5-HT induces desensitization of the circular muscle and as this is facilitated by G-protein coupled receptor kinases (GRKs) and other proteins, we also examined their distribution. Human sigmoid colon samples were dissected into three separate layers (mucosa, taeniae coli and intertaenial strips) and RNA was amplified by RT-PCR. The 5-HT2B receptor and all 5-HT7 receptor splice variants were expressed in all tissues. 5-HT4 a,b,c and n splice variants were also expressed in all tissues and 5-HT4d, 5-HT4g and 5-HT4i were only detected in some samples. The 5-HT2A receptor was seen predominantly in the intertaenial strips of the colon. Only one transcript of the serotonin transporter (SERT) was detected in the muscle layers. Variation was seen in GRK expression with GRK2 and 3 predominantly expressed in the mucosa, while GRK5 and 6 were found more commonly in the taeniae coli. PDZ (named after postsynaptic density protein, Drosophila disc large tumour suppressor and tight junction protein ZO-1) domain containing proteins, which may be involved in 5-HT receptor trafficking, were also detected throughout the sigmoid colon. The 5-HT3A subunit was expressed in all tissues, whereas the 5-HT3E subunit was mainly found in the mucosa layer while the 5-HT3B subunit was more common in the muscle layers. Receptor interacting chaperone (RIC-3), which is involved in transporting 5-HT3 receptor subunits, is expressed less in mucosa compared to muscle layers. In conclusion, these results show that there is variation in distribution of 5-HT receptors and interacting proteins within the sigmoid colon that may contribute to colonic function.


It is becoming increasingly apparent that several disorders of the gastrointestinal tract are associated with abnormal serotonin metabolism and/or signalling such as carcinoid syndrome, coeliac disease, bacterial induced enteritis, irritable bowel syndrome (IBS) and inflammatory bowel disease.1 The serotonin receptors of clinical relevance include 5-HT3 and 5-HT4 receptors, but more recently interest is focussing on the roles of 5-HT2B and 5HT7 receptors1–3 as it is now evident that there is a complex interplay between the 5-HT receptors in bowel function. In addressing these issues, it is important to understand that both the cellular expression as well as the function of 5-HT receptors in the human gut is significantly different from those in small laboratory animals.4

The 5-HT3 receptor is expressed on sensory nerve endings where it would be expected to contribute to sensations of bloating and pain associated with IBS. This is supported by the effectiveness of 5-HT3 antagonists in clinical trials showing decreased motility, nausea and sensitivity.2 5-HT3A and 5-HT3B receptor subunits are localized to the submucous plexus ganglion cells5 and are expressed in myenteric nerve plexus in the intertaenial layer of the human colon.6 5-HT4 receptors are present in several discrete tissue locations in the human colon. These include the mucosa where they contribute with 5-HT2A receptors to 5-HT-induced Cl secretion.7 5-HT4 receptors are also localized on the circular smooth muscle cells where agonist action induces relaxation6,8,9 and are expressed by cholinergic neurones in the human colon.10,11 The clinical efficacy of the 5-HT4 partial agonist tegaserod may also include its antagonistic action at 5-HT2B receptors.12 5-HT2B receptors are found predominantly in the longitudinal and circular smooth muscle layers and in the myenteric nerve plexus throughout the colon where they cause neuronally mediated contractile responses of longitudinal muscle.13 5-HT also acts directly on circular muscle 5-HT7 receptors to induce relaxation.14 This effect was confirmed in our recent study where we showed that 5-HT7 receptors augment the 5-HT4 receptor induced relaxation and this was associated with low, but consistent levels of 5-HT7 receptor mRNA6.

This study examined the concurrent expression and distribution of all relevant serotonin receptor transcripts in different tissue layers of the human sigmoid colon. It is well established that 5-HT induces desensitization (tachyphylaxis) at different rates and magnitudes in different tissues. Homologous desensitization generally begins with agonist-induced phosphorylation of the G protein coupled receptor (GPCR) by GPCR receptor kinases (GRKs)15 which are likely to play a large role in 5-HT receptor desensitization. Receptor trafficking is also an important process in receptor expression. PDZ domains were first recognized as sequence repeats contained in three separate proteins: postsynaptic density (PSD) protein PSD-95, discs large protein and tight junction protein ZO-1. PDZ domain containing proteins interact with the extremity of the C terminal end of GPCRs including both 5-HT2A and 5-HT4 receptors.3,16 Similarly, receptor interacting chaperone (RIC-3) has a role in trafficking 5-HT3 receptor subunits.17 Hence, we also examined the distribution of transcripts of these proteins that interact with 5-HT receptors in the human colon tissues.


Human tissue

This project was approved by the Monash University Standing Committee on Ethics in Research Involving Humans (Protocol number: 98/115) and by the Human Research Ethics Committee of St Vincent’s Hospital, Melbourne (Protocol number: 001/01). Specimens of human sigmoid colon were obtained from patients who gave consent prior to surgical resection for colonic cancer at St Vincent’s Hospital, Melbourne. The specimens were taken as far from the tumour as possible and appeared to be disease free by gross visual inspection by the pathologist. Tissues were obtained from five male patients and the age of these patients ranged from 50 to 74 years (median 55). The patients received pethidine premedication and standard anaesthetics and muscle relaxants.

The specimens were acquired from the surgical theatre immediately after resection and transported to the laboratory in 4 °C Krebs–Henseleit solution (composition in mmol L−1: NaCl 118.4, KCl 4.7, MgSO4 1.2, KH2PO4 1.2, NaHCO3 25, (d)-glucose 11.1 and CaCl2 2.5) preoxygenated with carbogen (95% O2/5% CO2). The specimens were dissected to remove the mucosa and associated mesentery plus fat. The taeniae coli (longitudinal muscle bands) were dissected and the remaining intertaenial tissue was then cut into strips in the orientation of the circular smooth muscle. All tissues were stored in RNAlater® (Ambion, Austin, TX, USA) at −80 °C for the PCR experiments described below.

As previously reported,6 separate adjacent intertaenial tissue strips were mounted in a 20-mL organ bath containing Krebs–Henseleit solution (37 °C and oxygenated with carbogen) under 0.75 g of tension and washed with fresh solution every 10 min. The preparations developed tone and spontaneous activity after 30–60 min. When the basal responses had stabilized, cumulative concentration–response curves were constructed to 5-HT alone. Only one concentration–response curve was produced for each strip to avoid tachyphylaxis. The responses were recorded using isotonic transducers (Ugo Basile) connected to PowerLab™ system (ADI Instruments, Sydney, Australia). The 5-HT induced concentration-related relaxations of separate adjacent intertaenial tissue strips used in this study were determined to have an EC50 value of 136 ± 24 nmol L−1 (n = 5).

RNA extraction and cDNA preparation

Total RNA was prepared from the mucosa and taeniae coli and intertaenial strips of the human colon samples (approximately 30 mg each) that had been stored in RNAlater™ (Ambion) at −80 °C, using the RNeasy Fibrous kit (Qiagen, Doncaster, Vic., Australia) and treated with DNA-free™ DNase treatment and removal reagents (Ambion) before being quantified spectrophotometrically (UV-160A; Shimadzu, Kyoto, Japan) or by using the RNA QuantIT kit and the Qubit fluorometer (Invitrogen, Mount Waverly, Vic., Australia). Total RNA (1 μg) was reverse transcribed with SuperScript III Reverse Transcriptase (Invitrogen) using oligo dT15 primers according to the manufacturer’s specifications in a total volume of 20 μL. However, AMV Reverse Transcriptase (Promega, Annandale, NSW, Australia) was used for the quantitative PCR experiments (qPCR).

Endpoint PCR

The primers used for all the experiments are listed in Table 1. All PCR reaction conditions in this study were standardized. A standard PCR reaction mixture contained 1 μL cDNA (or negative RT reaction), 1× PCR buffer, 0.1 μmol L−1 primers, 200 μmol L−1 dNTPs and 0.5 U Taq DNA polymerase (Qiagen) and 1× Q-solution (Qiagen) per 25 μL reaction. All tubes were denatured at 94 °C for 15 min and then 35 cycles of amplification were performed (60 s denaturation at 94 °C, 60 s annealing at 55 °C and 60 s extension at 72 °C for all primer sets unless otherwise stated) with a final extension at 72 °C for 8 min in a FTS-960 DNA Thermal sequencer (Corbett Research, Mortlake NSW, Australia) or a MyCycler thermal Cycler (BioRad Laboratories, Gladesville, NSW, Australia). The quality of cDNA produced was assessed by amplifying cDNA for the house-keeping gene, glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The h5-HT2A transcript was detected following one round of PCR. Preliminary studies showed a weakly expressed band (if any) after one round of PCR using the h5-HT3 primer sets. As a result, a 1-μL aliquot of the first round PCR products was used as the new template for a second round of amplification using the same h5-HT3 primers and conditions. Similarly, one round of PCR failed to amplify several of the h5-HT4 products using the outer common forward primer and specific reverse primer, so partial nested amplification was carried out using a 1-μL aliquot of the first round PCR products and the inner common forward primer with the specific reverse primer (Table 1). The h5-HT7 receptor transcripts were amplified by nested PCR using the outer pair of primers in the first round and amplifying 1 μL of this template with the inner pair of primers in the nested round (Table 1). The h5-HT2B transcripts were not consistently observed after one round, so a second round were performed, using the same primers and 1 μL of the first round PCR products as a template. Both the h5-HT2B and h5-HT7 receptor transcripts were denatured for 10 min at 94 °C, followed by 35 cycles of amplification (60 s denaturation at 94 °C, 50 s annealing at 65 °C and 60 s extension at 72 °C) with a final extension at 72 °C for 8 min.

Table 1.   Primer oligonucleotide sequences employed for endpoint PCR studies
GeneAccession no.PrimersPosition
  1. These primer sequences were described in *Ref. (43); †Ref. (44); ‡Ref. (45) and §Ref. (31).

β actin*M10277GCCCTGAGGCACTCTTCCA2435–2453
5-HT4 outerCommon forwardCACCAATATTGTGGATCCTT 
5-HT4 innerCommon forwardGTTGAACCCTTTTCTCTACG 
5-HT4a reverseNM_001040169TTTCTCGAGTTCCTGATGAT1284–1264
5-HT4b reverseNM_000870GTGACACTGACTCTCCCACT1284–1264
5-HT4c reverseY12506AGTTTCTTCTGTCGGTTTCA1124–1104
5-HT4d reverseNM_001040171CAATAAGAATTGGCCACAG1278–1260
5-HT4g reverseNM_199453TTTCTCGAGTTCCTGATGAT1334–1314
5-HT4i reverseNM_001040173GTGACACTGACTCTCCCACT1404–1384
5-HT4n reverseNM_001040174CAACCAAATCAATGAACTCC1275–1255
5-HT7 outer†NM_019860.2AGTACCGGAATATCAACCGGAAG1448–1471

Quantitative PCR

The qPCR experiments were conducted using a Rotorgene RT-3000 (Corbett Research) and the data analysed with Rotorgene (version 6) software. Only h5-HT3 receptor subunit expression was tested by qPCR and the same primers were used as in the endpoint PCR experiments (Table 1). The reactions contained: 1 μL cDNA, 1 mmol L−1 forward and reverse primers, 4 mmol L−1 MgCl2, and 2× QuantiTech SYBR Green PCR Master Mix (Qiagen) in a final volume of 30 μL. The cDNA was amplified by one cycle at 95 °C for 15 min followed by 35 cycles of 95 °C for 15 s (denaturing), 55 °C for 20 s (annealing) and 72 °C for 25 s (extension). Data were acquired at 80 °C during cycling. Melting point analysis was undertaken on each set of reactions to determine that only a single product was produced. No primer–dimers were detected during the 35 real-time amplification cycles by melting point analysis and this was confirmed in preliminary runs with gel electrophoresis (and this also established that only one product of the correct size was produced). To ensure that DNA contamination of samples was not affecting qPCR results, each qPCR analysis contained non-template controls consisting of RNA samples where reverse transcriptase was not added (such that no cDNA was produced). Each sample was analysed in triplicate. Baseline correction was performed by subtracting the fluorescence value of the mean for the first five cycles to all values. Efficiency of reactions was determined using linear regression of the Log (fluorescence) per cycle number data with the LinRegPCR program18 which revealed that when amplification occurred, the efficiency was 1.814 ± 0.007 across all h5-HT3 primer pairs. Expression data for each h5-HT3 receptor subunit mRNA in all samples was calculated relative to two housekeeping genes (β actin and GAPDH) adapting the method described by19 and expressed as the following ratio, where Ct is the crossing point threshold of the sample for the amplified genes:


Data were analysed by one-way anova using GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA). The number of observations used to derive mean values is expressed by n and arithmetic mean values are given as mean ± SE.


Tissue distribution of 5-HT3 receptor subunits

Previously, we had shown that the level of expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was comparable between human colonic tissue layers6 and so we used this gene as the reference gene in this study. A typical example of GAPDH expression for one individual indicates that similar levels of mRNA were present in the extracts from different tissues (Fig. 1). That comparable levels of mRNA were present in extracts from different individuals was further confirmed by the qPCR analysis (Fig. 2).

Figure 1.

 An example of the RT-PCR analysis of the 5-HT3 receptor subunit mRNA expressed in the human colon of one patient. Control amplifications of the GAPDH gene from RNA samples with (plus) or without (minus) reverse transcriptase (RT) to monitor for DNA contamination. The size of the expected PCR products is indicated. The labelling is as follows: M, mucosa layer; TC, taeniae coli; IT, intertaenial tissue. No bands were observed for the 5-HT3AL spice variant or for the 5-HT3C and 5-HT3D receptor subunits in any of the tissues tested.

Figure 2.

 Relative expression levels of 5-HT3 receptor subtypes in human colon mucosa (A), taeniae coli (B) and intertaenial tissue (C) detected using quantitative PCR. Data are expressed as a ratio relative to β actin as described in the Methods section. Bars indicate arithmetic mean values. The 5-HT3D receptor subtype was not detected in any tissue tested. For each subunit and region, n = 3.

The levels of the 5-HT3 receptor subunits expressed in the different layers of the colon for one individual are shown in Fig. 1. This is a typical example as 5-HT3Ashort splice variant was expressed in all tissue regions in all individuals (Table 2). The 5-HT3B subunit was most commonly co-expressed in the muscular layers whereas the 5-HT3E subunit was seen in the mucosa layers (Fig. 1, Table 2). The 5-HT3Along splice variant, 5-HT3C and 5-HT3D subunits were not observed in any of the individuals using end-point PCR.

Table 2.   Distribution of transcripts of 5-HT receptors and interacting proteins detected by endpoint PCR
Gene expressedMucosa layerCircular muscleLongitudinal muscle
  1. Numbers indicate the number of individuals from whom the RNA was detected but samples from all five patients were tested unless otherwise stated (n = 5); *In these samples n = 4 (intertaenial tissue only).

5-HT3A subunit555
5-HT3B subunit133
5-HT3E subunit501
5-HT2A receptor142
5-HT2B receptor555
5-HT4a receptor545
5-HT4b receptor544
5-HT4c receptor555
5-HT4d receptor333
5-HT4g receptor143
5-HT4i receptor322
5-HT4n receptor545
5-HT7 (variants a, b, d)555

Therefore, the mRNA expression was examined in three individuals using quantitative PCR to further investigate receptor subunit distribution (Fig. 2). Relative mRNA expression levels were normalized to β-actin expression and confirm the end-point results. In addition, it is evident that 5-HT3Ashort splice variants were expressed more strongly and often than 5-HT3Along variants in all tissues. Interestingly, the 5-HT3C receptor subunit was also observed and it appeared to be more common in the mucosal layer. All 5-HT3 receptor subunit expression was significantly lower (< 0.01; one-way anova) than GAPDH expression.

Tissue distribution of other 5-HT receptors

The same cDNA extracts were used to examine the co-expression of other 5-HT receptors in the human sigmoid colon. A typical example of 5-HT2A, 5-HT2B, 5-HT4 and 5-HT7 receptor expression is shown in Fig. 3. It should be noted that with the exception of the 5-HT2A receptor, all the other receptors (including 5-HT3 receptor subunits) were scored following partial or fully nested PCR (Table 2). 5-HT2A receptors were more commonly seen in the intertaenial strips whereas 5-HT2B receptors were seen throughout the layers. The 5-HT4 receptor has at least 11 splice variants3,20 and as there has been speculation about their relative tissue distribution, we examined the distribution of seven of these in the colon. The 5-HT4 receptor a,b,c and n splice variants were consistently detected in all tissue layers, while the other splice variants were observed less often (Table 2). A previous report indicated that the 5-HT4d receptor splice variant was expressed at very low levels21 but its expression level in the sigmoid colon appeared similar to the other splice variants (Fig. 3). Interestingly, the 5-HT4g receptor splice variant appeared to be restricted to muscle layers. Three splice variants of the 5-HT7 receptor have been described and shown to be expressed in the colon.22 We confirm that all of these 5-HT7 splice variants were consistently expressed in each of the layers of the sigmoid colon (Fig. 3).

Figure 3.

 An example of RT-PCR analysis of the 5-HT2A, 5-HT2B, 5-HT4 and 5-HT7 receptors in the human colon of one patient. The amplifications were from the same set of cDNA used in Fig 1 where GAPDH expression is shown. The size of the expected PCR products is indicated. The labelling is as follows: M, mucosa layer; TC, taeniae coli; IT, intertaenial tissue.

Tissue distribution of proteins interacting with 5-HT receptors

The GRKs have an important and interchangeable role in phosphorylating 5-HT receptors to activate β-arrestin mediated receptor internalization.15 The same cDNA extracts used to detect 5-HT receptor transcript expression were used to examine the co-expression of GRKs in the colon. All GRKs were present in the colon (Fig. 4, Table 2). Although there were variations in GRK distribution between the tissue layers with GRK4 being expressed at very low levels in the muscular layers only. GRK5 and 6 were also more common in the muscle.

Figure 4.

 An example of RT-PCR analysis of the serotonin transporter (SERT) compared with expression of Receptor interacting chaperone (RIC-3), G-protein coupled receptor kinases (GRKs), Na+/H+ exchange regulatory cofactor (NHERF), sorting nexin 27 (SNX27), dihydropyrimidinase-related protein 2 (DPYSL2), MPP3 and SEC23A in the human colon of one patient. The size of the expected PCR products is indicated. The labelling is as follows: M, mucosa layer; TC, taeniae coli; IT, intertaenial tissue.

Different RNA extracts from adjacent tissue samples were used for the remaining PCR analyses. The levels of GAPDH expression in these samples were similar to that shown in Fig. 1 indicating comparable mRNA expression (data not shown). Interestingly, we observed SERT being expressed in the taeniae coli and intertaenial strips once and it was not expressed in the mucosa (n = 5). As a positive control for this set of primers, we tested ileum samples where SERT was clearly expressed in the mucosa but not the muscular layers (n = 2; data not shown).

Proteomic studies have revealed several proteins that specifically interact with 5-HT2A and 5-HT4 receptor splice variants in the mouse.23,24 BLAST searches were used to identify human homologues for five of these proteins and their transcript expression was analysed in the different layers of the sigmoid colon. Sorting nexin 27 (SNX27), Na+/H+ exchange regulatory cofactor (NHERF), dihydropyrimidinase-related protein 2 (DPYSL2) and protein transport protein Sec23 homologue A (SEC23A) homologue transcripts were strongly expressed in each layer in most individuals while transcripts of membrane protein, palmitoylated 3 (MPP3), a member of the p55 membrane associated guanylate kinase family, were typically more strongly expressed in the muscular layers (Fig. 4, Table 2). Transport of 5-HT3 receptor subunits to the cell surface is also enhanced by the receptor interacting chaperone protein RIC-3.17 RIC-3 transcripts were typically expressed more strongly in the muscular layers than the mucosa (Fig. 4, Table 2).


Several 5-HT receptors contribute to the normal function of the human colon and as their distribution relative to each other is poorly documented, we investigated 5-HT receptor distribution in different tissue layers of the sigmoid colon at a transcript level. This work has previously been reported in abstract form.25 The 5-HT-induced concentration-related relaxations of separate adjacent intertaenial strips used in this study were determined to have an EC50 value of 136 nmol L−1 which is consistent with literature values. The relaxant effect of 5-HT on the circular muscle of the human colon is achieved by actions at 5-HT4 and 5-HT7 receptors.6,8,9,14 We observed differences in 5-HT receptor transcript distribution that provides encouragement to the notion that specific receptor subunits or splice variants could be potential targets for ‘super drugs’ that are specific for particular splice variants as mooted by Pindon et al.26

The 5-HT3 receptor is a ligand gated ion channel that mainly carries sodium and potassium ions to mediate a rapidly activated and desensitizing inward rectifying current. Functional 5-HT3 receptors are assembled as a pentamer of five symmetrically arranged subunits surrounding a central pore.27 Five different subunits have been identified in humans; the 5-HT3A receptor subunit can form functional homo-oligomers and is an essential component of functional hetero-oligomer channels.28 Transcripts of the genes encoding the 5-HT3 receptor subunits have been detected in the human colon6,29,30 and also at relatively low levels in the stomach compared with the duodenum.31 Functional and immunohistochemical experiments have identified 5-HT3 receptors in the submucous plexus of the human colon.5 However, 5-HT3 receptors have not been shown to contribute to functional relaxation or contraction in vitro8 unlike rodent intestines.32,33 In spite of this, we demonstrated here that the 5-HT3 subunit transcripts were present in the intertaenial and taeniae coli strips presumably in the myenteric plexus. Interestingly, there was a spatial distribution of the 5-HT3 subunits with the 5-HT3A subunit distributed throughout all tissues, the 5-HT3E subunit being found predominantly in the mucosa tissue and the 5-HT3B subunit in the muscular tissue. Unlike the duodenum and stomach where there is a correlation between the 5-HT3C and 5-HT3E subunit expression,31 we observed that the 5-HT3C subunit was expressed at lower levels in the sigmoid colon. Our results suggest that 5-HT3A/3B hetero-oligomers are more likely to form in the myenteric plexus whereas 5-HT3A/3E hetero-oligomers are likely in the submucous plexus of the sigmoid colon.

5-HT4 receptors are widespread and present in several discrete tissue locations in the human colon. 5-HT4 receptors are localized on the circular smooth muscle cells where agonist action induces relaxation and inhibition of spontaneous contractions.6,8,9 Paradoxically, 5-HT4 receptors expressed by cholinergic neurones in the human colon simultaneously enhance acetylcholine and NO release.10,11 It has been suggested that 5-HT itself and 5-HT4 agonists facilitate colonic propulsion via a coordinated combination of circular muscle relaxation and longitudinal muscle contraction.10,34 Another less well-established location of the 5-HT4 receptor in the human colon is on sensory nerve endings, where its function may be to increase sensory perceptions arising from the abdomen leading to altered motility patterns. Our recent clinical studies lend some support to this hypothesis and showed that IBS patients have significantly lower perception and defecation thresholds to rectal thermal and pressure stimuli compared to age- and gender-matched control subjects.35–37 5-HT4 receptors also are present in the mucosa where they contribute to 5-HT-induced Cl secretion.7 The most compelling evidence establishing the 5-HT4 receptor as a therapeutic target for treating IBS comes from clinical trials showing that 5-HT4 agonists reduce abdominal pain and give a degree of relief from other symptoms.2 The human small intestine and colon express several 5-HT4 receptor splice variants.3 This has led to speculations that different 5-HT4 receptor splice variants are expressed in various locations and so contribute to the regulation of bowel functions in different ways. Hence we investigated the distribution of transcripts of several splice variants of the 5-HT4 receptor in the sigmoid colon. All transcripts tested were present in all tissues with the exception of the 5-HT4g receptor splice variant (generally relatively low expression and only detected once in the mucosa layer). It is relevant to note that the intestinal specific splice variant, 5-HT4d receptor splice variant occurred in all tissues of the sigmoid colon. This information is important in the search for potential 5-HT4 receptor splice variant ligands as it offers the opportunity to develop splice variant-selective drugs termed ‘super drugs’26 where for instance an agonist preferentially targets the “d” over the “g” splice variant.

Such selective properties do not seem to be achievable at first glance with the 5-HT7 receptor where our data show that all splice variants were expressed in the different tissue layers. However, the 5-HT7d splice variant is expressed at relatively low levels in tissues such as the brain and spleen and is intestinal preferring22 and so could be a suitable intestinal target. The 5-HT4 and 5-HT7 receptors appear to augment their effects via relaxing the intertaenial muscle of the colon.6,14 Whereas high levels of both mRNA and protein for 5-HT2B receptors were found predominantly in the muscle layers and in the myenteric nerve plexus throughout the colon where they caused neuronally mediated contractile responses of longitudinal muscle.13 We have confirmed that the 5-HT2B receptor transcript is widespread in the sigmoid colon in this study; hence the clinical efficacy of tegaserod may indeed involve 5-HT2B antagonist activity in addition to 5-HT4 agonist activity.12 Mucosal sheets of the sigmoid colon contain a ketanserin-sensitive (5-HT2A) receptor, which is involved in inducing a secretory response.7 This response appears to be stimulated via 5-HT2A and 5-HT4 receptors in the ascending colon and 5-HT2A receptors predominantly in the sigmoid colon.7 Surprisingly, we only detected the 5-HT2A receptor transcript in one mucosal layer of the sigmoid colon although it was common in the intertaenial region (n = 5). This discrepancy indicates that the functional distribution of receptors regulating Cl secretion should be clarified by future investigations.

As 5-HT acts on several specific 5-HT receptor subtypes that have opposing effects, mechanisms for 5-HT removal have also evolved. 5-HT is imported into cells via the 5-HT transporter (SERT) where 5-HT is catabolized by intracellular enzymes such as monoamine oxidases and glucuronyl transferases. We were surprised to observe that SERT was rarely expressed in the human sigmoid colon despite the presence of several 5-HT receptors in this region. A recently published study confirms our observation in our colon samples, but demonstrates that SERT is functionally active in the small intestine38 and high levels of SERT expression have also been reported in the duodenum.31 This suggests that desensitization of 5-HT receptors and the presence of GRKs may have a major role in the colon. Some support for this contention comes from the observation that the intracellular levels of GRK6 are important factors in determining the onset, severity and period of dextran sodium sulphate-induced colitis in mice.39 Although there are some differences in the GRK distribution, each tissue contained at least two GRKs and either GRK2 or GRK5 (both of which can phosphorylate 5-HT4 receptor splice variants40) were present. Recently, it was demonstrated that the 5-HT2A receptor-arrestin interactions profoundly alter the response to endogenous serotonin levels in neurones41 indicating the importance of the cellular milieu in determining drug action.

Receptor trafficking is also influenced by other proteins such as PDZ domain containing proteins which interact with the extremity of the C terminal end of GPCRs such as 5-HT2A and 5-HT4 receptors.3,16 Proteomic experiments have demonstrated that the PDZ ligand of the 5-HT2A receptor interacts with a different set of PDZ proteins to that of the 5-HT2C receptor in the mouse.23 Similarly it has been shown that different proteins interact with either the mouse 5-HT4a or 5-HT4e receptor splice variants which contain different C terminals with PDZ domains.24 These specific sets of different PDZ proteins probably contribute to the different signal transduction properties and trafficking of these receptors. We observed that high levels of transcripts of the human homologues of these proteins were expressed in the different tissue layers of the sigmoid colon. One of these proteins, DPYSL2 has also recently been identified as a potential marker of colorectal cancer as it is over-expressed and secreted from cancerous tumours.42 DPYSL2 is important in neuronal tissues where it plays a role in nerve development as it modulates microtubule dynamics. Transport of 5-HT3 receptors to the cell surface is also enhanced by the receptor interacting chaperone protein RIC-317 and surprisingly transcripts of RIC-3 were expressed less in the mucosa.

In conclusion, differences in the relative distribution and actual transcript levels of 5-HT receptors were observed between the dissected tissue layers of the sigmoid colon that were particularly marked with the 5-HT3 receptor subunits and to a lesser extent with the distribution of the 5-HT4 receptor splice variants. Interestingly, in view of recent speculation about novel drug targets modulating receptor trafficking and desensitization, transcripts of SERT were rarely observed in the colon tissue, whereas the GRKs involved in receptor desensitization were common.


This project was supported by the NHMRC Australia and the Victorian College of Pharmacy Monash University Faculty small grants fund. The authors thank Dr H Liu for providing the sequences of the 5-HT7 receptor primers and Dr M Congiu for assistance in collecting tissue samples.