| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Call For Papers: Comparative Genomics
1 Heymans Institute of Pharmacology, Ghent University, Ghent
2 Department of Internal Medicine, Johnson & Johnson Pharmaceutical Research and Development, Beerse, Belgium
3 Department of Functional Genomics, Johnson & Johnson Pharmaceutical Research and Development, Beerse, Belgium
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
7 transmembrane receptor; 9 transmembrane receptor; pig; G protein-coupled receptor; 5-hydroxytryptamine
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Together with 5-HT6 and 5-HT7 receptors, 5-HT4 receptors are seven transmembrane domain (7TM) receptors that are positively coupled to Gs proteins. Almost 20 years ago, the first description of the 5-HT4 receptor was made by Dumuis and coworkers (10), who reported the 5-HT4 receptor-mediated stimulation of adenylyl cyclase by 5-HT in cultured mouse embryonic colliculi neurons. To date, 5-HT4 receptors are known to be widely distributed in the body, which suggests a broad functionality. In the brain, the highest expression of 5-HT4 receptors is found in anatomical structures linked to cognition such as limbic structures, hippocampus, and basal ganglia (26, 28). Accumulated pharmacological evidence indicates that 5-HT4 receptor activation may improve memory, learning, and anxiety processes (6, 19, 27). In the periphery, involvement of 5-HT4 receptor activation was shown in the regulation of gastrointestinal motility, intestinal secretion, visceral sensitivity and antinociception, bladder contraction, and cardiac stimulation (for review see Ref. 23). The wide distribution of 5-HT4 receptors is accompanied by alternative mRNA splicing, which causes an extra level of functional diversity (Fig. 1). Since the initial cloning of two splice variants from rat brain (13), several variants of 5-HT4 receptors have been described by cDNA cloning in human and other species. The divergence in the sequence of these splice variants usually starts after amino acid Leu358, resulting in COOH-terminal splice variants. 5-HT4(a), 5-HT4(b), 5-HT4(c), 5-HT4(d), 5-HT4(g), 5-HT4(i), and 5-HT4(n) have been cloned from humans (Fig. 1); 5-HT4(a), 5-HT4(b), 5-HT4(e) in rat; and 5-HT4(a), 5-HT4(b), 5-HT4(e), and 5-HT4(f) in mouse (6, 7). Sequencing of the human 5-HT4 receptor gene (HTR4) revealed that the splice variants were indeed splice variants from the same gene (Fig. 1). Exons corresponding to the different COOH-terminal variants were found in the genomic context. Furthermore, the existence of the so-called exons "gef" suggests the existence of the putative human variants 5-HT4(e) and 5-HT4(f), which is underscored by their mRNA expression (3). In addition, one splice variant, 5-HT4(h), has an insertion of 14 amino acids in the second extracellular loop of the receptor. This variant was first described in porcine blood vessels (40). Meanwhile, the 5-HT4(h) variant has been found in human tissue, but so far only in combination with COOH-terminal exon b (3). Although this differential splicing of the human HTR4 gene attracted a lot of attention in the field, which led to the characterization of many splice variants in vitro, the full-length nature of some transcript variants has not been determined. Additionally, specific antibodies against the different 5-HT4 receptor splice variants have not been generated yet, and no experimental evidence exists that all the different identified splice variants are translated in vivo.
|
In this study we therefore examined the occurrence of 5-HT4 receptor splice variants in the pig. Apart from 5-HT4(a) and 5-HT4(b), we found nine novel splice variants, which were named 5-HT4(b2), (j), (k), (l), (m), (o), (p), (q), (r). In addition, the internal h-variant was found in combination with the COOH-terminal exons of variant a, r, and m (5-HT4(ha), 5-HT4(hr), 5-HT4(hm)). Moreover, we found splice variants composed of duplicated exons, one of which resulted in a putatively 9TM receptor (5-HT4(9TM)). Three splice variants (5-HT4(a), 5-HT4(r), and 5-HT4(9TM)) were functionally characterized.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
General Molecular Biological Methods
Unless otherwise stated, all PCR reactions were performed in a final reaction volume of 25 µl, containing 22.5 µl AccuPrime pfx SuperMix (0.55 U of AccuPrime pfx polymerase, buffer components, 330 µM dNTPs, 1.1 mM MgSO4; Invitrogen), 200 nM primers and 1 µl of template DNA. PCR amplifications were performed on a Tetrad-2 Thermal Cycler (Bio-Rad, Hercules, CA).
DNA sequencing was carried out by cycle sequencing using BigDye Terminator (Applied Biosystems) and M13F (5'-GTA-AAA-CGA-CGG-CCA-G-3') and M13R (5'-CAG-GAA-ACA-GCT-ATG-AC-3') sequencing primers. Sequences were analyzed on an ABI PRISM 3730 sequencer (Applied Biosystems) and assembled using Sequencher version 4.6 (GeneCodes, Ann Arbor, MI).
3'-RACE Amplification and Subcloning of 3'-RACE Products
The SMART RACE (Clontech, Mountain View, CA) protocol was adapted to amplify the 3'-ends of porcine 5-HT4 receptor cDNAs. Rapid amplification of 3'-cDNA ends (RACE)-ready cDNA was reverse transcribed from 1 µg of total RNA isolated from the mucosal and muscular fraction of the proximal stomach and the adrenal gland of newborn piglets as well as from the left and right atrium of newborn and young piglets, using a modified lock-docking oligo(dT) primer [5'-AAG-CAG-TGG-TAT-CAA-CGC-AGA-GTA-C(T)30VN-3'] and Powerscript Reverse Transcriptase (Clontech) or Superscript III Reverse Transcriptase (Invitrogen) according to the manufacturers' instructions. The first-strand reaction products were diluted with Tricine-EDTA buffer to a final volume of 110 µl. Two forward primers were designed based on the published porcine 5-HT4 receptor cDNA sequence (GenBank: AY566638): F865 and F843 (Supplemental Table S1).1
We used 1.25 µl of cDNA in a PCR reaction (5 min at 95°C followed by 20 cycles of 15 s at 95°C, 30 s at 52°C, and 1 min at 68°C, with a final extension step of 7 min at 68°C) using F865 in conjunction with the Universal Primer A Mix (UPA; long: 5'-CTA-ATA-CGA-CTC-ACT-ATA-GGG-CAA-GCA-GTG-GTA-ACA-ACG-CAG-AGT-3'; short: 5'-CTA-ATA-CGA-CTC-ACT-ATA-GGG-C-3'). We used 1 µl of this reaction as template in a nested amplification round (5 min at 95°C followed by 25 cycles of 15 s at 95°C, 30 s at 54°C, and 1 min at 68°C, followed by 7 min incubation at 68°C) with F843 and the Nested Universal Primer A (NUP; 5'-AAG-CAG-TGG-TAA-CAA-CGC-AGA-GT-3'). The target sequences of UPA and NUP lie within the adaptor sequence that is ligated to the 3'-end of the RACE-ready cDNA. PCR products were separated on a 2% agarose gel, and product bands were excised and purified using the QIAquick Gel Extraction Kit (Qiagen Benelux, Venlo, The Netherlands). Purified PCR products were ligated into pCR-Blunt II-TOPO plasmid vector and transformed into TOP10 One Shot Electrocompetent Cells (Invitrogen). For each gel-extracted PCR product, 6–12 colonies were picked randomly, and plasmid DNA was prepared from overnight grown cultures using the QIAprep Turbo BioRobot Kit on a Biorobot 8000 Workstation (Qiagen). The plasmids were screened for the presence of the 5-HT4 receptor-specific sequence located between the nested forward primer F843 and the common 3'-splice site of exon 5 on position 1118 of the open reading frame (ORF).
Tissue Localization Studies
Total RNA was prepared from porcine peripheral tissues using the Qiagen RNA isolation kit. cDNA was prepared from 1 µg total RNA using oligo(dT) primers and Superscript II reverse transcriptase (Invitrogen). The cDNAs specific for the splice variants identified in this study were detected using a nested PCR amplification. To normalize the quantities of cDNA from different tissue sources used as input in the PCR reactions, 10 ng of cDNA was PCR amplified with primers specific for the porcine glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (5'-ACC-ACA-GTC-CAT-GCC-ATC-AC-3' and 5'-TCC-ACC-ACC-CTG-TTG-CTG-TA-3'). Next, a first reaction was performed using 
10 ng of cDNA (deduced from GAPDH results) together with the exon 5-specific primer F865 and a splice variant-specific reverse primer (Supplemental Table S1). Products of this first reaction were used as templates for a nested PCR amplification with primer F843, and variant-specific 3'-end primer (Supplemental Table S1). All PCR reactions in tissue distribution studies were performed using the following protocol: 5 min at 95°C followed by 30 cycles (15 s at 95°C, 30 s at 55°C, 1 min at 68°C) and a final elongation (7 min at 68°C). The absence of a PCR product does not exclude the presence of the variant in the tissue: since we used a nested PCR approach, false negative results may appear when the concentration of target sequence is below the threshold of detection, set by the conditions of the first amplification round. To analyze the tissue distribution of the 5-HT4(h) variant, a single competitive PCR reaction was performed using AmpliTaq Gold DNA polymerase (Applied Biosystems): 10 min at 94°C followed by 41 cycles (30 s at 94°C, 30 s at 56°C, 1 min at 72°C) and a final elongation (8 min at 72°C) with primers flanking the h-exon: an exon 4-specific forward primer (Fh) and an exon 5-specific reverse primer (Rh). This results in coamplification of variants with or without the h-exon, and gives information on their relative amount because any variable affecting PCR amplification has the same effect on both molecular species.
Cloning of Full-Length cDNAs and Generation of Fusion Proteins With the EGFP
Full-length cDNAs of 5-HT4(a) and 5-HT4(r) were generated by RT-PCR, using cDNA generated with Superscript III and random hexamers (Invitrogen) (5-HT4(a): F1022/R968, 5-HT4(r): F1022/R1014; Supplemental Table S1). The resulting PCR products were gel fractionated, excised, purified, and used as a template in a second reaction with a nested reverse primer (R906 and R1024 respectively; Supplemental Table S1). Products were again size-separated on an agarose gel, excised, purified, cloned into pENTR-D-TOPO (Invitrogen) and sequence verified. The full-length cDNA of 5-HT4(9TM) was assembled by ligation. The HaeII restriction fragment of PCR product F843/R1039 (Supplemental Table S1) was ligated (T4 DNA ligase; Roche Diagnostics, Mannheim, Germany) with the HaeII digested PCR product F1022/R1028 (Supplemental Table S1). The resulting ligation product was cloned into pENTR-D-TOPO and transformed into TOP10 One Shot Electrocompetent Cells. Several colonies were picked; Miniprep DNA was prepared using the QIAprep Spin Miniprep Kit (Qiagen), and the presence of an insert was verified by restriction enzyme digestion; the sequence was verified by sequencing. The PCR products of the porcine 5-HT4(a), 5-HT4(r), and 5-HT4(9TM) were then shuttled into the mammalian expression vector pDEST40 by LR clonase (Invitrogen). The plasmids containing the full length sequences were used to generate a full-length sequence with a reverse primer in which the TGA stop codon was mutated to remove the translation termination signal and to provide an in-frame sequence with enhanced green fluorescent protein (EGFP). PCR products were purified using the QIAquick PCR Purification Kit (Qiagen), cloned into pENTR-D-TOPO and sequence verified. These products were subcloned into an in house generated pEGFP-N1/pDEST expression vector to generate NH2-terminal fusions to the EGFP.
Cell Culture and Transfection
HEK-293 tsA cells were grown in DMEM (Invitrogen) containing 10% fetal calf serum (Hyclone, Logan, UT), 105 IU/l penicillin (Serva, Heidelberg, Germany), 0.1 g/l streptomycin (Serva), 0.1 g/l pyruvate (Sigma, St. Louis, MO), and 0.292 g/l L-glutamine (Sigma), under 5% CO2 at 37°C. Cells were transfected overnight in suspension using Lipofectamine 2000 reagent (Invitrogen) and cultured posttransfection in DMEM containing 10% dialyzed fetal calf serum (Invitrogen).
Membrane Preparations and Radioligand Binding
Cells were cultured on 145 mm Petri dishes and incubated in medium containing 5 mM sodium butyrate (Acros Organics, Geel, Belgium) for 20 h before experiments to boost receptor expression before membrane preparation for radioligand saturation binding (33). Cells were washed twice with ice-cold PBS. The cells were then scraped from the plates with a cell scraper, suspended in 50 mM Tris·HCl buffer (pH 7.4), and harvested by centrifugation for 10 min at 16,000 g. The pellet was resuspended in 5 mM Tris·HCl (pH 7.4) and homogenized with an ultra Turrax homogenizer. The resulting membranes were collected by centrifugation at 25,000g for 20 min and stored in 50 mM Tris HCl buffer (pH 7.4) or 50 mM HEPES-NaOH (pH 7.4), at a protein concentration of 1 mg/ml. The Bradford protein assay (Bio-Rad) was used for protein determination with bovine serum albumin as a standard.
Assay mixtures (0.5 ml) contained 50 µl of the tritiated ligand [either the 5-HT4 receptor antagonist [3H]GR-113808 or the agonist [3H]5-HT (Amersham Biosciences, Little Chalfont, UK)] and 0.4 ml of the membrane preparation (at 0.003, 0.006, or 0.013 mg/ml protein for [3H]GR-113808 binding of 5-HT4(a), 5-HT4(r), or 5-HT4(9TM), respectively, and 0.025, 0.075, or 0.13 mg/ml protein for [3H]5-HT binding of 5-HT4(a), 5-HT4(r), or 5-HT4(9TM), respectively). Furthermore, 50 µl of solvent was added for total binding or 50 µl of 10 µM SB-207266 for [3H]GR-113808 binding or 50 µl of 10 µM GR-113808 for [3H]5-HT binding, to determine nonspecific binding. The [3H]GR-113808 assay buffer was 50 mM HEPES-NaOH (pH 7.4). The [3H]5-HT assay buffer was 50 mM Tris·HCl (pH 7.4) containing 10 mM MgCl2, 1 µM pargyline, and 1 µM paroxetine. The mixture was incubated for 1 h at 25°C. The incubation was terminated by rapid filtration over Whatman GF/B filters, presoaked in 0.1% polyethylenimine, followed by three washing steps with 50 mM HEPES-NaOH (pH 7.4) for [3H]GR-113808 binding or 50 mM Tris·HCl (pH 7.4) for [3H]5-HT binding. Ligand concentration isotherms were obtained with twelve concentrations of [3H]GR-113808 ranging from 0.02 to 3 nM or 12 concentrations of [3H]5-HT ranging from 0.2 to 14 nM.
Ligand concentration-binding isotherms were calculated by nonlinear regression analysis. Data were fit to one-and two-binding site models. The fit with the lowest squares of sum of residuals (F-test) was chosen as the best fit of the data. The null hypothesis was rejected at a P < 0.05 level. All data are reported as means ± SE.
Measurements of cAMP Formation
After an overnight incubation in medium containing dialyzed fetal calf serum, cells that were transiently transfected with 5-HT4(a), 5-HT4(r), and 5-HT4(9TM) were used for cAMP measurements performed by homogeneous time-resolved fluorescence (Cisbio International, Bedford, MA) (12). We added 25 µl of cell suspension (10,000 cells per well) and 25 µl of either vehicle or compound (1 pM–10 µM), all in stimulation buffer composed of HBSS (Invitrogen) plus 1 mM isobutylmethylxanthine (Sigma), 5 mM HEPES (Sigma), 1 mM MgCl2 (Merck, Darmstadt, Germany), and 1 g/l BSA (Sigma) to the wells of a black Costar 96-well plate and incubated that for 30 min at room temperature. In experiments where the effect of 5-HT was assessed in the presence of 0.3 µM GR-113808, we preincubated cells with the antagonist (or its solvent) for 20 min, before adding 5-HT. To assess the effect of direct activation of adenylyl cyclase, 3 µM forskolin (Sigma) was added to the transfected cells and incubated for 30 min at room temperature. Effects were quantified as the 665 nm/620 nm ratio and reported on a standard curve to deduce respective cAMP concentrations. cAMP concentrations [one (5-HT4(a), (r) experiment or two experiments (5-HT4(9TM)) with n = 6 independent wells assayed for each concentration point] were used to fit the data to a three-parameter logistic model using GraphPad Prism version 4.02. To compare the curve location parameters of 5-HT and prucalopride, the fit of a three-parameter logistic model with common curve location parameters was compared with the fit of a three-parameter logistic model with separate curve location parameters. The goodness of fit of both models was compared with an F-test (assessed by the sum of squares), adjusting for difference in the number of degrees of freedom. If the F-test resulted in a P value <0.05, the individual curve location parameters (the maximal effect 
, pEC50) were compared by a t-test, and the corresponding P values are shown.
Confocal Microscopy
HEK-293 tsA cells (8.106 cells) were transfected with 3 µg of the fusion constructs of 5-HT4(a), 5-HT4(r), 5-HT4(9TM) with EGFP, diluted, and grown on polylysine-treated Lab-Tek chambered cover glasses (Nunc, Naperville, IL) at a density between 40,000 and 80,000 cells/cm2. At 48 h posttransfection, the cells were labeled with Alexa Fluor 594 wheat germ agglutinin (Molecular Probes, Eugene, OR) to label the plasma membrane, fixed, and covered with a drop of Vectashield antifade solution containing DAPI (Vector Laboratories, Burlingame, CA) and a cover glass. Cells were analyzed using the Zeiss LSM 510Meta confocal imaging system.
Data Mining
The following bioinformatics tools were used: genome browsing (http://www.ensembl.org/index.html; http://trace.ensembl.org; http://www.sanger.ac.uk/cgi-bin/Projects/S_scrofa/BESsearch.cgi) and transmembrane topology prediction (http://www.cbs.dtu.dk/services/TMHMM-2.0/; http://bioweb.pasteur.fr/seqanal/interfaces/toppred.html; http://bioinfo.si.hirosaki-u.ac.jp/ConPred2/); alignments between the porcine splice variants found in our study and the human HTR4 genomic region were performed using several alternative alignment algorithms: Spidey, an mRNA-to-genomic alignment program (http://www.ncbi.nlm.nih.gov/IEB/Research/Ostell/Spidey/), ClustalW (http://www.ebi.ac.uk/Tools/clustalw/index.html), and EMBOSS-Align (Smith-Waterman algorithm; http://www.ebi.ac.uk/emboss/align/). To allow for interspecies comparisons, Spidey was run with lowered mismatch, gap opening, and gap extension penalties. When using ClustalW and EMBOSS-Align, the gap opening penalty and gap extension penalty were varied during the optimization of the alignments.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
|
900 bp long UTR. At the protein level, the terminal six amino acids of 5-HT4(b) were substituted by two valine residues in 5-HT4(b2) (Fig. 3B). Nevertheless, a PCR that was set up with specific primers designed to amplify the extreme 3'-end of the coding sequence of the published porcine 5-HT4(b) variant only, produced a single band of expected size, suggesting the presence of 5-HT4(b) (data not shown). Besides 5-HT4(a) and 5-HT4(b), none of the COOH-terminal variants currently known in other species were identified in the RACE experiments (nor in RT-PCR experiments with a reversed primer based on the human sequences; see supplemental data). However, several sequences were found that comprised the common sequence up to exon 5 but did not correspond to any known COOH-terminal splice variant of the 5-HT4 receptor. In accordance with the current nomenclature of 5-HT4 receptor splice variants (http://www.iuphar-db.org/GPCR/ReceptorDisplayForward?receptorID=2329), we named these new porcine variants 5-HT4(j), 5-HT4(k), 5-HT4(l), 5-HT4(m), 5-HT4(o), 5-HT4(p), 5-HT4(q), and 5-HT4(r) (Figs. 2A, 3B). All the new COOH-terminal variants but 5-HT4(m) could be aligned to the human HTR4 genomic region (Supplemental Fig. S1). Sequence identities were, however, rather low, in the range of 60–70%, indicating the low conservation of these sequences between human and porcine DNA. For comparison, the sequence identity of exon 2 (90%), 3 (90%), 4 (85%), h (84%), 5 (91%), a (83%), and b (99%) are much higher (Supplemental Fig. S2). The presence of the new variants in human tissue is therefore unlikely.
A BLAST search of the known human variants against porcine genomic shotgun sequences resulted in the identification of the porcine homologs of the "gef" exons (SS_WGS-375j07.q1k). Similar as in human, the sequence is preceded by a splice acceptor site (AG), and 5-HT4(g) might thus be present in porcine tissue. Based on this shotgun sequence, 5-HT4(e) and 5-HT4(f) cannot be generated in the pig since the required alternative acceptor sites are not present in the porcine sequence (Supplemental Fig. S3).
Identification of Splice Variants Composed of Duplicated Exons of the Porcine 5-HT4 Receptor, Including 5-HT4(9tm) and 5-HT4(15tm)
In addition to the splice variants described above, the RACE experiments with cDNA from left atrium revealed splice variants composed of (partial) duplicates of one or more exons of the porcine HTR4 gene. For example, variants in which the common splice donor site of exon 5 was followed by exons 4–5 or exons 4-h-5 (Fig. 2B: variant 1). Exon 5 was only partially duplicated and followed by a short sequence that contained an in-frame stop codon, a polyadenylation signal, and a poly-A tail. The duplicated exons were in frame with the common sequence. Consequently, translation results in duplication of part of the transmembrane region of the receptor: the sequence between the second and the third intracellular loop, including two transmembrane helices, is repeated after the common intracellular COOH terminus (Supplemental Fig. S2 and Fig. 4A). This homofusion variant is predicted to contain nine transmembrane helices by all prediction servers tested (Fig. 4, B and C) and we therefore named this variant 5-HT4(9TM). Additional evidence for the existence of this variant was found by the amplification of products with an exon succession 5-4-5 (variant 1, 5-HT4(9TM)) and 5-5 (variant 2) using primers located within exon 5 (F1045-R1046) (Fig. 2B).
|
Finally, the existence of splice variants composed of duplicated exons fused to the complete common 5-HT4 region was given a solid base by the amplification of a variant with a rank order of exons (1)-2-3-4-5-1-4, which suggests a duplicated region located downstream of the HTR4 gene. This variant was found in a heminested PCR reaction on the muscular fraction of the ileum using F1022 (exon 1 from ATG onwards) and R1015 in a first amplification round, and forward primers located in exon 2 (F932) and exon 3 (F934 or F858) in a second round (Fig. 2B, variant 4).
Genomic Context of the Newly Identified HTR4 Exons
To confirm that the newly identified variants originated from alternative splicing of genomic exons, we elucidated the genomic context of the new cDNA nucleotide sequences. Several of the 5-HT4(a) UTR exons could be traced down to their genomic context on chromosome 2q29 of the porcine physical map via the Ensembl Trace Server (e.g., end sequences of BACs CH242-19O19.T7 and PigE-146O22.T7; detailed results not shown). The common exons 1, 4, and 5 were also mapped in this genomic region (BACs CH242-91N19.SP6, CH242–369H11.T7, and CH242-400M6.T7 respectively; Supplemental Table S2).
The end sequences of BACs CH242-338C6.T7 and CH242-478H10.T7 on the one hand and the genomic shotgun sequence SS_WGS-258f21.q1k on the other hand provide additional evidence for the presence of two different b-variants that are derived from separate exons: an alternative splice-donor site in the b-exon may lead to the alternative 5-HT4(b2) variant, while the unspliced exon b generates the known 5-HT4(b) sequence (Fig. 5A, Supplemental Fig. S1C).
|
1 kb downstream of exon d (Supplemental Fig. S1B). Porcine splice variant 5-HT4(m) was found with two alternative UTRs originating from internal splicing of exon m, both containing the canonical polyadenylation signal (Fig. 5C). None of these sequences showed homology to the human genome. 5-HT4(p) was generated from an alternative splice acceptor site within one of the 5-HT4(a) UTR variants, which causes a UTR exon to become a coding exon (Fig. 5D). The end sequence of BAC clone CH242-19O19.T7 contains part of the p sequence. The COOH-terminal sequence of 5-HT4(q) (Fig. 5E) was also found within the publicly available genomic sequences. Unexpectedly, however, an insertion of 26 nucleotides was present in the cDNA of 5-HT4(q) compared with this genomic sequence. The porcine homologs of the human "gef" exons are found 120 bp upstream of the q-exon in the porcine genome (Supplemental Fig. S3). Finally, part of the 3'-UTR of 5-HT4(r), containing a polyadenylation consensus sequence, was also identified in porcine genomic sequences (Fig. 5F). A homologous sequence to this porcine exon r was found between exon 5 and exon i on the human HTR4 gene (Supplemental Fig. S1B).
Finally, in support of our assumption that the identified 5-HT4 variants did not represent genomic sequences, we determined the intron sequence located at the 3'-end of exon 5 from porcine BAC clone CH242-369H11 and confirmed that it did not correspond to any of the sequences found in this study, nor to an exon duplication (data not shown).
Tissue Distribution of COOH-Terminal Splice Variants of the Porcine 5-HT4 Receptor
We then examined the expression profile in different tissues of the variant COOH-terminal sequences by RT-PCR. Except for the 5-HT4(j) variant, all other variants identified in the RACE experiments were confirmed in at least one tissue (Fig. 6A). The applied method does unfortunately not allow for a quantitative analysis of the results, and false negative results cannot be excluded when the concentration of target sequence was below the threshold of detection. Splice variants 5-HT4(a) and 5-HT4(b) are the predominant variants in terms of expression level and distribution across various tissues, followed by 5-HT4(m), 5-HT4(r), and 5-HT4(9TM). Left and right atrium were the only tissues that expressed all variants in at least one of the examined animals.
|
Cloning and Functional Characterization of the Porcine 5-HT4(a), 5-HT4(r), and 5-HT4(9TM) Splice Variants
Cloning and confocal imaging.
Full-length cDNAs from 5-HT4(a), 5-HT4(r), and 5-HT4(9TM) were generated using RT-PCR. To compare the pharmacological profiles of the 5-HT4(a), 5-HT4(r), and 5-HT4(9TM) variants, the corresponding pDEST40 constructs were transiently transfected into HEK-293 tsA cells. In addition, to visualize cellular distribution, these variants were fused to EGFP and transfected into HEK-293 tsA cells (Fig. 7A). EGFP-fused 5-HT4(a) and 5-HT4(r) were clearly expressed at the cell membrane. In addition, also intracellular structures (presumably ER and Golgi) were stained, which was most pronounced in 5-HT4(r). 5-HT4(9TM) was mainly located intracellularly, suggestive for a different topology and less effective membrane trafficking or instability of this variant compared with the other two variants; weak membrane staining was, however, also observed.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Although 5-HT4 receptors are expressed in embryonic murine hearts (16) and in failing rat hearts (36), the porcine heart is currently the only available tool for studying cardiac 5-HT4 receptors under physiological conditions in a nonprimate animal model. However, the findings of this study suggest that additional validation at the molecular level might be needed to give a solid basis for the use of swine as a model for human 5-HT4 receptor function. In human, several 5-HT4 receptor variants are produced by alternative splicing of a complex gene of >185 kb that includes at least 14 exons (6). While the in vivo relevance and functional role of the multiplicity of 5-HT4 receptor splice variants remains to be elucidated, differences between 5-HT4 receptor splice variants are becoming increasingly recognized in heterologous expression systems (5, 6). These differences occur at the level of receptor palmitoylation (34, 35), downstream coupling (33, 37), receptor desensitization (29, 32), and constitutive receptor activity (8). Additionally, human 5-HT4 receptors interact with specific intracellular proteins in a splice variant-specific manner (15), which further supports the view that such isoforms are in charge of fine regulation of signal transduction and are not the result from transcriptional leakage (5). This study is the first to address the splicing behavior of the porcine 5-HT4 receptor.
Using BAC end sequences positive for HTR4 sequences from the porcine genome project, we annotated the HTR4 gene to porcine chromosome 2q29. The common part of the porcine 5-HT4 receptor is highly conserved between human and pig, but our analysis demonstrates large differences at the 3'-end of the gene, resulting in highly different gene splice variants between these species. Indeed, we describe the expression of nine novel COOH-terminal splice variants in porcine tissue that have not been described in human tissue or any other species. Similar as in human, rat and mouse, these new variants (5-HT4(b2, j, k, l, m, o, p, q, r)) differ in their sequence from Leu358 onwards, resulting in COOH-terminal splice variants that have not been reported in the human HTR4 gene. Our data provide evidence that these variants were splice variants from a single gene since we identified intronic sequences in the public databases that surrounded the exons, excluding the possibility that the variants originated from intron retention. On the other hand, we detected only few of the splice variants that have been reported in human. Based on the porcine genomic homolog of the "gef" exons, it can be anticipated that of these three variants, only 5-HT4(g) might be present in porcine tissues. Although no porcine genomic sequences were retrieved that are homologous to the other human variants, their existence cannot be ruled out due to the unfinished status of the porcine genome project.
Splice variants in extracellular loops of G protein-coupled receptors are rare (20). One of these is the 5-HT4(h) receptor, which has an extra insertion of 14 amino acids in its second extracellular loop. The first description of a partial 5-HT4(h) sequence originates from porcine cDNA (40), and this variant has also been found in human, albeit only in combination with the b-terminal exon (5-HT4(hb)) (3). However, the authors suggested that the h-exon is probably also expressed in combination with various other COOH-terminal exons. Here, we provide experimental evidence that porcine 5-HT4(h) splice variants exist in combination with other COOH-terminal tails, 5-HT4(ha), 5-HT4(hm), and 5-HT4(hr). Interestingly, a tissue distribution analysis of the exon h-containing variants revealed that the mucosal fractions along the entire gastrointestinal tract, in contrast to the muscle layers containing the myenteric plexus, predominantly expressed variants with the h-exon. In mice, a similar distribution pattern of some 5-HT4 receptor splice variants has been described. Expression of 5-HT4(a) and 5-HT4(b) mRNA, normalized to the expression of synaptophysin to correct for the proportion of transcripts derived from neurons, was more pronounced in the submucosa than in the longitudinal muscle with adherent myenteric plexus (25). In contrast, the expression of mRNA encoding 5-HT4(e) and 5-HT4(f) was myenteric plexus specific (25). The preferential mucosal localization of exon h-containing splice variants might correlate with their presence in enterochromaffin cells (EC) and/or intrinsic primary afferent neurons. These are two well-known locations of 5-HT4 receptors, where they are involved in autofeedback of 5-HT release from EC cells and in the peristaltic reflex, respectively (14, 38).
In addition, our data demonstrate that at least part of the porcine HTR4 gene is duplicated in the genome, which can give rise to splice variants comprising duplicated exons. All of these variants contained duplicated exons from the common part of the receptor (exon 1 to 5). One of these duplication variants is predicted to be a homofusion protein containing nine putative TM domains (5-HT4(9TM)), which we also demonstrated to be functionally active. Several other findings argue in favor of a duplicated genomic region: 1) A diversity of splice variants that each contained different duplicated exons were obtained; 2) Multiple combinations of duplicated exons were found, but the duplicated exons always retained the same order; 3) The variants were found using different primer combinations, located in different exons; 4) Expression of the duplicated variants was found in multiple tissues from different animals; 5) The junction site is the common splicing site Leu358, supporting the fact that the duplication variants resulted from alternative splicing mechanisms; 6) The presence of porcine COOH-terminal splice variants that have not been found in human suggests that this genomic region is not well conserved between pig and human. This is further supported by the low sequence similarity with their human homolog sequences.
In addition, some circumstantial evidence can be found that would be in line with a duplication of the porcine HTR4 gene. The human chromosome 5q corresponds to porcine chromosomes 2 and 16, with the human 5q13–q32 region (74.0–148.2 Mb) corresponding to the porcine region 2q22–q28 (22, 39). In the human genome, the HTR4 gene is located at 5q32 (147.8–148.0 Mb), thus close to the breakpoint between the two synteny blocks in human chromosome 5. This is clearly illustrated by the location of the CSF1R and GPX3 genes. While both genes map closely to the HTR4 gene on human chromosome 5 (at 149.4 Mb and 150.4 Mb, respectively), the CSF1R maps at the porcine chromosome 2q telomere, while GPX3 maps at porcine chromosome 16 (22). In line with these results, the porcine HTR4 gene is located at the telomeric end of chromosome 2 (at 2q28 or q29). Of note, the mapping of the porcine HTR4 gene to SSC 2q28 by Shimogiri et al. (39) contrasts with the porcine physical map in the Ensembl database, where the HTR4 gene is mapped to SSC 2q29. Enrichment of segmental duplications have been shown earlier in regions of breaks of synteny between human and mouse genomes (2). In addition, the duplication frequency in subtelomeric regions is greater than the genome average (42).
The functionally active 5-HT4(9TM) variant forms a partial 5-HT4 receptor dimer that is directly transcribed from the genome. This variant is proposed to possess a 9TM topology by a number of different prediction algorithms, including ConPred II. In this method sequences are first run through ConPred_elite, which has an accuracy of almost 100% (1). Based on these modeling approaches only, however, we cannot exclude that 5-HT4(9TM) has a 7TM topology with a large COOH-terminal tail containing the duplicated sequences. The 9TM domain structure needs experimental confirmation using for instance a series of truncated fusion constructs of 5-HT4(9TM) (4). It has been postulated that new protein topologies may originate during evolution through multistep gene rearrangements, involving duplication, fusion, and truncation, which may lead to fully functional intermediate proteins (41). Besides, evidence is accumulating that in human and in other mammalian genomes, copy number polymorphisms explain a substantial amount of genetic variation and may account for a significant proportion of normal phenotypic variation (11). It has been suggested that subtelomeres could be a place where new genes are created by duplication and subsequent modification processes and could be the source for phenotypic variation among individuals (24). This is illustrated by a study on the transcriptional activity of multiple copies of the subtelomerically located olfactory receptor gene OR-A (24). Our data might fit into these theories.
The porcine splice variants showed some degree of tissue specificity. Some of the predominant variants in terms of expression level and distribution across various tissues were selected for a functional analysis: the homofusion receptor 5-HT4(9TM), one of the new COOH-terminal variants 5-HT4(r) and 5-HT4(a), which is present throughout species examined. These experiments confirmed, at least for these variants, that the variants identified in this study were expressed at the cell membrane and that they shared some characteristics with human 5-HT4 receptors: 1) 5-HT4-EGFP fusion constructs were expressed at the cell membrane; 2) Two affinity sites for [3H]5-HT were revealed on porcine 5-HT4(a) and 5-HT4(r) receptors (GPCR fraction and noncoupled receptor fraction), similar to what has been reported for the human 5-HT4(a) and 5-HT4(b) variants (33); in contrast, only a single affinity site was found for 5-HT at 5-HT4(9TM). Possibly the low receptor density of 5-HT4(9TM) might mask the few high affinity sites; 3) [3H]GR-113808 bound to these receptors as a typical antagonist (capable of binding the GPCR fraction and the noncoupled receptor fractions) with KD values consistent with previous findings on 5-HT4 receptor-transfected cells and different brain tissues (33). Nevertheless, GR-113808 behaved as a weak agonist at these splice variants. Such a behavior has been previously reported with 5-HT4 receptor antagonists, including GR-113808, in cell systems transfected with 5-HT4 receptors (3, 33); 4) The variants all constitutively stimulated the formation of cAMP. The role of the COOH terminus in the constitutive activity of 5-HT4 receptors has been previously reported (8); 5) Prucalopride behaved as a superagonist (more efficacious than 5-HT) at the porcine 5-HT4(a) receptor, while it was equi-efficacious with 5-HT at the 5-HT4(r) variant and a partial agonist at 5-HT4(9TM). Such a reversal of efficacy order cannot be attributed to differences in receptor expression levels. At the human 5-HT4(a) receptor prucalopride behaved as a full agonist compared with 5-HT in inducing cAMP formation but as a superagonist when Ca2+ influx was measured (33). The 5-HT4 receptor agonist cisapride has also been shown to behave as a superagonist in mouse colliculi neurons (10).
Together, our results add to the complexity of the 5-HT4 receptor. As discussed above for the human variants, at present one can only speculate on the functional significance of such a diversity of splice variants. Some of the variants identified in this study are only weakly expressed at the mRNA level and might thus not lead to functionally significant protein levels. However, given the multitude of variants, a tight regulation of their differential expression can be expected, and the expression of some of these variants might only be triggered by certain external stimuli or time factors.
In conclusion, our data provide evidence for a duplicated porcine HTR4 gene. Alternative splicing generates a homofusion GPCR with a predicted 9TM topology. These results urge for caution when interpreting data on 5-HT4 receptor-related pharmacology obtained in the pig. Validation at the molecular level might be warranted before extrapolating results to human, as drug binding properties at the expressed set of splice variants can be altered. Nevertheless, in vitro and/or in vivo functional data obtained in the porcine model could still be highly similar to their human counterpart, especially because all 5-HT4 receptor splice variants characterized so far are positively coupled to cAMP, and the COOH-terminal variations appear to affect mainly the fine regulation of the signal transduction cascade.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Present addresses: J. De Maeyer, Movetis N.V., Veedijk 58, 2300 Turnhout, Belgium; J. Aerssens, Tibotec bvba, Generaal De Wittelaan L11 B3, 2800 Mechelen, Belgium.
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 The online version of this article contains supplemental material. 
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) and prucalopride (
) (left graphs) and 5-HT in the presence of 0.3 µM GR-113808 (
) (right graphs) on cAMP formation in HEK-293 tSA cells transfected with 5-HT4(a), 5-HT4(r), and 5-HT4(9TM). Note the elevated baseline in experiments with GR-113808, indicative of partial agonistic behavior of this compound. ***P < 0.001 maximal effect of prucalopride vs. 5-HT.