Differential Agonist-Mediated Internalization of the Human 5-Hydroxytryptamine 7 Receptor Isoforms

Serotonin (5-hydroxytryptamine; 5-HT) is a small molecule involved in a number of physiological and pathological processes. Fourteen subtypes of receptors and the serotonin transporter mediate these processes.

The human 5-HT₇ receptor was cloned in 1993 (Bard et al., 1993). 5-HT₇ receptors are expressed in the central nervous system (thalamus, hypothalamus, limbic, and cortical regions), but also in peripheral tissues (vascular and gastrointestinal smooth muscle and certain leukocytes) (Eglen et al., 1997). A variety of approaches has provided evidence that 5-HT₇ receptors may be involved in regulation of affect, body temperature, circadian rhythms and rapid eye movement sleep, primary sensory neuronal sensitivity, and relaxation of smooth muscle in a variety of tissues (Terron, 1996, 2002; Prins et al., 1999; Vanhoenacker et al., 2000; De Ponti and Tonini, 2001; Inoue et al., 2003; Centurion et al., 2004; Janssen et al., 2004; Thomas and Hagan, 2004; Varnas et al., 2004). Three splice variants of the human 5-HT₇ receptor have been described (Heidmann et al., 1997; Jasper et al., 1997; Stam et al., 1997) and designated 5-HT_7(a), 5-HT_7(b), and 5-HT_7(d), which differ only in the length and amino acid composition of their carboxy-terminal tail. Although levels of expression differed, all three isoforms were detected in most tissues examined (Heidmann et al., 1998).

In previous studies, the three isoforms showed pharmacology and functional coupling to adenylyl cyclase to be essentially identical. However, the human 5-HT_7(d) isoform did display a diminished 5-HT-mediated stimulation of adenylyl cyclase activity (Krobert et al., 2001). Levy's group also extensively characterized the ability of the three isoforms to constitutively activate adenylyl cyclase and the inverse agonist properties of a series of antagonists. Again, 5-HT_7(a), 5-HT_7(b), and 5-HT_7(d) displayed similar profiles for both constitutive activity and inverse agonist responses (Krobert and Levy, 2002).

Previous studies have all failed to show any differences in pharmacology or function among the three isoforms of the human 5-HT₇ receptors. One investigation, in the laboratory of Levy and colleagues (Krobert et al., 2001), showed that the 5-HT_7(d) isoform had reduced efficacy in 5-HT-mediated stimulation of adenylyl cyclase. We chose to examine the trafficking of the human 5-HT₇ receptors in response to treatment with agonist to see whether there were differences among the three isoforms. Receptor trafficking has been investigated and shown to occur among other members of the serotonin receptor family, particularly 5-HT_1(a), 5-HT_2(a), and 5-HT_2(c) receptors (Xia et al., 2003; Schlag et al., 2004; Zimmer et al., 2004). In this study, we demonstrate that the 5-HT_7(d) isoform exhibits receptor trafficking that is distinct from that of 5-HT_7(a) or 5-HT_7(b). Human 5-HT_7(d) receptors display agonist-independent internalization and indeed internalize even in the presence of antagonist. We employed a novel double-labeling procedure to label two populations (surface and internal) of receptor carrying the same epitope tag. These studies reveal near total removal of 5-HT_7(d) receptors initially on the surface of the cell to the interior and replacement with a population of receptors previously found inside the cell. Under identical conditions, the majority of 5-HT_7(a) and 5-HT_7(b) remain on the surface. The consequence of this removal of 5-HT_7(d) receptors from the surface is a reduced efficacy of cAMP-responsive reporter gene activity. Agonist-independent internalization provides a possible mechanism for the previous observation that the human 5-HT_7(d) isoform has a reduced ability to stimulate the second messenger pathway. These data further show that the 5-HT_7(d) carboxy tail may mediate an interaction with a trafficking pathway for internalization other than that used by 5-HT_7(a) and 5-HT_7(b) receptors.

Materials and Methods

Construction of Plasmids

Human 5-HT_7(a) [h5-HT_7(a)] receptors with a C-terminal epitope tag derived from influenza hemagglutinin (HA) were constructed by PCR from a construct containing the previously cloned h5-HT_7(a) cDNA (Heidmann et al., 1997) in the pcD-SRα expression vector (Takebe et al., 1988). The sequence MGYPYDVPDYA with a consensus Kozak sequence and a 5′ EcoRI site was added at the amino terminus of the receptor coding sequence. This primer has the sequence, 5′GAATTCGAATTCATCATGGGGTACCCATATGACGTCCCAGACTACGCCATGGACGTTAACAGCAGCGGC3′. The primer for the 3′ end of the PCR product was designed to hybridize to a sequence surrounding the pcD-SRα multiple cloning site with the sequence 5′ATGTCTGGATCCTCGACGAGCTCG3′. The product was cut with EcoRI and BamHI and ligated back into pcD-SRα vector with the resulting expression vector being designated SRα-HA 5-HT_7A. Epitope-tagged expression vectors for the 5-HT_7(b) and 5-HT_7(d) were generated by domain swapping coding sequence corresponding to the C terminus of the 5-HT_7(b) and 5-HT_7(d) isoforms with SRα-HA 5-HT_7A to generate SRα-HA 5-HT_7B and SRα-HA 5-HT_7D. Constructs were validated by restriction analysis and sequencing.

Determination of Relative Abundance of mRNA Isoforms with Reverse Transcriptase-PCR

Total RNA was extracted from homogenized tissues or cells using a guanidine isothiocyanate/phenol extraction protocol (Chomczynski and Sacchi, 1987), digested for 30 min with ribonuclease-free DNaseI (Roche Diagnostics, Indianapolis, IN), and heat denatured for 15 min at 65°C. Poly(A)⁺ RNA was prepared using Dynabead Oligo (dT)25 selection (Dynal Biotech, Brown Deer, WI). Poly(A)⁺ RNA (1 mg) was reverse-transcribed to cDNA as described previously (Heidmann et al., 1997). cDNA was synthesized using SuperScript II reverse transcriptase (Invitrogen, Carlsbad, CA). Radiolabeled PCR was performed using the primers 184F 3′AGTACCGGAATATCAACCGGAAG5′ and 184R 3′GTTTATTTCATCTCCATTGTTCTGC5′. This produces three products amplified from a single template corresponding to the three human isoforms of the 5-HT₇ receptor. Products were separated on 12% polyacrylamide gels. Relative abundance of the isoforms was determined by phosphorimaging and densitometry using the GS363 Molecular Imaging System (Bio-Rad, Hercules, CA).

Generation of Stable Cell Lines

The expression vectors for selection of HA-5-HT_7(a), HA-5-HT_7(b), and HA-5-HT_7(d) stable cell lines were constructed by subcloning of the epitope-tagged receptors (SRα-HA-5-HT_{7A, -7B, -7D}) into the expression vector pIRESpuro (BD Biosciences Clontech, Palo Alto, CA). 293T cells, an SV40 T antigen-expressing variant of the 293 adenovirus-transformed human embryonic kidney cell line HEK 293 from American Type Culture Collection (ATCC CRL-1573), were transfected using the calcium phosphate method. Positive clones were selected and maintained in Dulbecco's modification of Eagle's medium (DMEM) + 10% fetal bovine serum + penicillin (100 IU/ml)/streptomycin (100 μg/ml) + 10 μg/ml puromycin. Serum products contain large amounts of serotonin, making it necessary to remove serum for experiments. Therefore, 24 h prior to experiments cell cultures were washed extensively and fed with COMPLETE serum-free medium (Mediatech, Herndon, VA).

Radioligand Binding Assay

Saturation binding assays were performed as described previously (Heidmann et al., 1998). Briefly, stable cell lines were grown in COMPLETE serum-free medium for 24 h, harvested, and crude membranes were prepared by Polytron homogenization. Aliquots of membrane suspension were incubated with increasing concentrations of [³H]5-HT (26–30 Ci/mmol) (PerkinElmer Life and Analytical Sciences, Boston, MA) in a total volume of 0.5 ml, filtered on GF/C filters, washed, and counted by liquid scintillation. Nonspecific binding was determined in the presence of 2 μM methiothepin. Bio-Rad protein assay was used to determine protein content. Saturation binding data were analyzed using nonlinear regression analysis in GraphPad Prism.

Reporter Gene Assays

293T cells were plated in 24-well dishes at 1 × 10⁵ cells/well in DMEM + 10% FBS and grown at 37°C/5% CO₂ for 18 h. Medium was reduced to 250 μl and cells were transfected with expression vectors for the three 5-HT₇ isoforms (SRα5-HT_{7a, 7b, 7d}) by calcium phosphate precipitation. A total of 250 ng of DNA was added to each well, which in addition to the 5-HT₇ receptor expression vectors, contained 2.5 ng of an α-168 luciferase reporter construct and 40 ng of RSV-β-galactosidase as a control for transfection efficiency. After overnight transfection, culture medium was changed to COMPLETE serum-free medium for 24 h prior to drug treatment. Treatments were performed in triplicate wells and luciferase and β-galactosidase activities in cell extracts were measured on a luminometer (Berthold AutoLumat 953). Dose response curves were generated and statistical analyses performed using GraphPad Prism.

Immunocytochemistry

These experiments were performed using two different techniques. The first was a standard technique where cells were grown on coverslips and subjected to various treatments. The cells were then fixed, permeabilized, and labeled with primary antibodies directed at epitope tags or organelles, followed by detection with appropriate fluorochrome-conjugated secondary antibodies. The second approach was to label receptors on the surface of the cell with primary and secondary antibodies prior to any treatments. Cells were removed from incubator and placed on ice to inhibit internalization of receptors. Primary and secondary antibodies were applied, cells were returned to the incubator, and drug treatments were initiated. Upon completion of treatment, cells were fixed, permeabilized, and receptors that were not on the surface initially were labeled as above.

Total Receptor Stain. Stable cell lines expressing the three HA-5HT₇ isoforms were grown on 12-mm glass coverslips. The slides were coated with 100 μg/ml poly-d-lysine for 60 min prior to plating cells. After 24 h in DMEM + 10% FBS, cells were washed three times with DMEM and grown for 24 h in COMPLETE serum-free medium (Mediatech). Cells were drug treated at 37°C for 120 min, then washed two times with PBS/Ca²⁺/Mg²⁺ (PBS + 0.1 g/l CaCl₂ + 1.0 g/l MgCl₂) on ice. Cells were fixed in 4% paraformaldehyde for 20 min at room temperature. Coverslips were washed 4 × 5 min with PBS/Ca²⁺/Mg²⁺. Cells were permeabilized in PBS/Ca²⁺/Mg²⁺ + 0.1% Triton X-100 for 15 min. Coverslips were blocked with PBS/Ca²⁺/Mg²⁺ + 10% normal goat serum for 60 min. Coverslips were incubated in primary antibody, 3F10 anti-HA rat monoclonal (Roche Diagnostics), 1:1000 in PBS/Ca²⁺/Mg²⁺ + 10% normal goat serum for 60 min. Slides were then washed 3 × 5 min with PBS/Ca²⁺/Mg²⁺ and reblocked for 15 min. Secondary antibody, goat anti-rat Alexa Fluor 488 (Molecular Probes, Eugene OR), 1:2000 in PBS/Ca²⁺/Mg²⁺ + 10% normal goat serum was added, and slides were incubated for 20 min. Coverslips were again washed 3 × 5 min with PBS/Ca²⁺/Mg²⁺, counterstained with 4′,6-diamidino-2-phenylindole, and mounted with ProLong Antifade (Molecular Probes).

Double Stain of HA-Tagged 5-HT₇ Surface Receptors versus HA-Tagged 5-HT₇ Cytoplasmic Receptors (Fig. 6). This modification of a live cell surface labeling protocol allowed us to examine two populations of human 5-HT₇ receptors carrying the same epitope tag. Therefore, the images and data in Fig. 6 differ from Fig. 3, in that the receptors labeled green are those that were on the cell surface prior to any treatments.

Stable cell lines were grown on coverslips and serum-starved as above. Cells were removed from the incubator and placed on ice for 15 min, then washed once with cold COMPLETE. Cells were incubated on ice with primary antibody, 3F10 anti-HA rat monoclonal, (Roche Diagnostics) 1:1000 in COMPLETE for 60 min. Cells were washed three times with COMPLETE. Secondary antibody, goat anti-rat Alexa Fluor 488, 1:1000 in COMPLETE was applied for 60 min on ice. Cells were then removed to an incubator where prewarmed medium with or without appropriate drugs was added. After 120 min, cells were immediately fixed in 4% paraformaldehyde for 20 min at room temperature. Cells were permeabilized in PBS/Ca²⁺/Mg²⁺ + 0.1% Triton X-100 for 15 min. The second primary antibody, HA.11 anti-HA mouse monoclonal (CRP, Cumberland, VA), 1:1000 in COMPLETE was applied for 60 min at room temperature. Cells were washed 3 × 5 min with PBS/Ca²⁺/Mg²⁺ and reblocked for 15 min in COMPLETE. Secondary antibody, goat anti-mouse Alexa Fluor 568, 1:2000 in COMPLETE was applied for 30 min at room temperature. Cells were again washed 3 × 5 min with PBS/Ca²⁺/Mg²⁺, counterstained with 4′,6-diamidino-2-phenylindole, and mounted with ProLong Antifade.

Immunofluorescence Microscopy

Observation and imaging of fixed or live cells were done on a Nikon Eclipse TE300 epifluorescent microscope. Image acquisition, processing, and analysis were done with IPLab (Scanalytics, Fairfax, VA) and Adobe Photoshop. Images (512 × 512 pixels) were acquired using a Photometrics SenSys cooled CCD camera and IPLab acquisition software. Images were deconvolved using MicroTome deconvolution software (Scanalytics). Fluorescence intensity measurements were made by hand-segmenting images into plasma membrane (300-nm wide) and perinuclear regions. Values are expressed as a percentage of total fluorescence intensity for each region. Each cell was segmented separately, and 15 to 30 fields (2–10 cells/field) were measured for each data set. Data sets were analyzed using the Mann-Whitney test, data expressed as a mean ± standard error of the mean.

Surface Biotinylation Assay

For surface biotinylation experiments, cells were grown to 80% confluence on 100-mm dishes and serum-starved overnight by washing three times and feeding with COMPLETE serum-free medium. Plates were chilled to 4°C and washed twice with PBS/Ca²⁺/Mg²⁺. Plates were then incubated with 300 μg/ml sulfo-NHS-SS-biotin (Sigma-Aldrich, St. Louis, MO) in PBS/Ca²⁺/Mg²⁺ for 30 min. Unreacted biotin was quenched by washing three times with Tris-buffered saline (10-mm Tris pH 7.4, 154 mM NaCl). Plates were returned to the incubator in COMPLETE serum-free medium containing agonist 5-carboxamidotryptamine (5-CT), antagonist (R)-3-(2-(2-(4-methylpiperidin-1-yl)ethyl)pyrrolidine-1-sulfonyl) phenol (SB 269970), or no drugs (untreated) for 120 min. Plates were then removed and chilled to 4°C. Remaining surface biotin was cleaved by incubating in glutathione reducing buffer (50 mM glutathione, 75 mM NaCl, 75 mM NaOH, 10% FBS) for 30 min followed by quenching of unreacted glutathione with iodoacetamide buffer (0.9% iodoacetamide and 1% bovine serum albumin in PBS/Ca²⁺/Mg²⁺). The cells were then immediately lysed in cold lysis buffer [Tris-buffered saline, 0.1% Triton X-100, 3 μg/ml aprotonin, 2 μg/ml pepstatin, 1 μg/ml leupeptin, and 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride] in a volume of 0.5 ml.

Proteins were measured using the BioRad protein assay. Equivalent amounts of protein were loaded onto NeutrAvidin agarose (Pierce Biotechnology, Rockford, IL) and incubated overnight at 4°C. The beads were washed three times with lysis buffer and biotinylated proteins were eluted with SDS sample buffer, resolved with SDS-polyacrylamide gel electrophoresis, and detected using anti-HA (HA.11, Covance) antibody.

Results

Relative Expression of the Three Human 5-HT₇ Isoforms in Various Cell Lines and Tissues. Radiolabeled amplification of cDNAs reverse-transcribed from poly(A)⁺ RNA isolated from a range of tissue types known to express the 5-HT₇ receptor (Bard et al., 1993). The primer set used (Heidmann et al., 1997) produced three different bands corresponding to the differing C-termini of the 5-HT_7(a), 5-HT_7(b), and 5-HT_7(d) serotonin receptors. This permitted an examination of the relative abundance of the three human 5-HT₇ isoforms within each tissue or cell (Fig. 1). Each tissue examined expressed all three isoforms. Even when we examined transformed cell lines (various blood cell types) and a primary culture of aortic smooth muscle cells, all three isoforms were found. In all cases, the 5-HT_7(d) isoform was the least abundant. It was observed that although the 5-HT_7(b) isoform was the most prevalent, the relative amounts of 5-HT_7(a) and 5-HT_7(d) differed by tissue type, with the 5-HT_7(d) isoform being most abundant in smooth muscle and least common in brain tissues.

HA-Tagged 5-HT₇ Receptors Are Functionally Equivalent to the Native 5-HT₇ Receptors. A cAMP-responsive luciferase reporter gene assay was used to evaluate the functionality of the 5-HT_7(a) serotonin receptor with an N-terminal influenza hemagglutinin epitope tag, HA-5-HT_7(a). 293T human embryonic kidney cells were transiently transfected with vectors that expressed either native 5-HT_7(a) or HA-tagged 5-HT_7(a) receptors. Treatment of cells with the selective agonist 5-CT (Fig. 2A) showed a dose-dependent activation of the reporter gene to similar magnitude and with comparable agonist potency for the native (EC₅₀ = 0.31 nM) and the HA-tagged (EC₅₀ = 0.46 nM) 5-HT₇ receptors. Likewise, cells treated with 5-CT and the 5-HT₇-specific antagonist SB 269970 displayed dose-dependent inhibition with similar inhibition potency. Although the HA-tagged 5-HT_7(a) receptor was slightly less sensitive to SB 269970 inhibition, EC₅₀ = 7 nM versus EC₅₀ = 0.93 nM for the native receptor, both displayed complete inhibition of 5-CT-stimulated activity.

5-HT₇ Receptors Internalize in Response to Agonist. Stable cell lines overexpressing HA-5-HT_7(a), HA-5-HT_7(b), and HA-5-HT_7(d) receptors (304–482 fmol/mg) were examined using epifluorescence microscopy. After 24 h in serum-free medium, cells were subjected to one of three treatments for 2 h: untreated (change to fresh medium), 1 μM 5-CT (agonist), or 1 μM SB 269970 (antagonist). Cells on coverslips were fixed, stained with anti-HA antibody, and examined using epifluorescent microscopy. Images were acquired with a cooled CCD camera and deconvolved with MicroTome deconvolution software. In the untreated condition (Fig. 3A) both HA-5-HT_7(a) and HA-5-HT_7(b) have the majority of their receptors in the plasma membrane and fewer in the cytoplasm. In contrast, HA-5-HT_7(d) receptors have a greater proportion in an internal compartment. When treated with agonist (5-CT), all three isoforms present a picture with many receptors localized to a perinuclear compartment. Antagonist (SB 269970) treatment results in a distribution of HA-5-HT_7(a) and HA-5-HT_7(b) signal that closely resembles the untreated state. Again, HA-5-HT_7(d) differs, now exhibiting more receptor staining on the plasma membrane than in the untreated condition. Figure 3B is a quantitative representation of 15 fields of cells (avg. five cells/field) for each condition. A biochemical assay of 5-HT₇ receptor internalization using surface biotinylation (Fig. 5) showed that although HA-5-HT_7(a) and HA-5-HT_7(b) show some constitutive internalization, the HA-5-HT_7(d) isoform has a much larger proportion of receptors in the cytoplasm in the untreated condition. We observe a translocation of receptors into the cytoplasm with agonist (5-CT) exposure for HA-5-HT_7(a) and HA-5-HT_7(b). The HA-5-HT_7(d) isoform shows an apparent retention of receptors in the cytoplasm in the presence of antagonist (SB 269970).

HA-5-HT_7(d) Receptors Have Diminished Activity to Activate a cAMP-Responsive Reporter Gene. The observation that HA-5-HT_7(d) had a greater proportion of receptors localized in the cytoplasm than did either HA-5-HT_7(a) or HA-5-HT_7(b) led to further investigation into the functional activity of these stable cell lines. We found that the three isoforms had similar agonist potency: HA-5-HT_7(a) (304 ± 46 fmol/mg), HA-5-HT_7(b) (310 ± 10 fmol/mg), and HA-5-HT_7(d) (482 ± 39 fmol/mg). However, the maximum response of the HA-5-HT_7(d) isoform was only about half that of the HA-5-HT_7(a) and HA-5-HT_7(b) isoforms (Fig. 4). This result was consistent over repeated experiments with different stable clones of each isoform. HA-5-HT_7(d) receptors have reduced activity, possibly due to fewer receptors at the cell surface at any given time.

Double Labeling of External versus Internal Populations of HA-5-HT₇ Receptors. To determine whether HA-5-HT_7(d) receptors were never reaching the plasma membrane or if they were being internalized in an agonist-independent fashion, we devised a method to double-label plasma membrane versus cytoplasmic receptors. We used a rat anti-HA monoclonal antibody, 3F10 (Roche Diagnostics), to label surface receptors on live cells, then rapidly fixed and permeabilized the cells and labeled internal receptors with a mouse anti-HA monoclonal antibody, HA.11 (Covance). This method provided good resolution of the two populations of HA-tagged 5-HT₇ receptors and revealed that at the time of labeling before any treatment, all three isoforms have a population of receptors on the surface and a population in the perinuclear region (Fig. 6, top row).

HA-5-HT_7(d) Receptors Are Constitutively Internalized in the Absence of Agonist. The above double-labeling protocol enabled us to more closely examine the data presented in Fig. 3. This experiment (Fig. 6) is different in that we could examine receptors that were on the cell surface prior to treatment (those with the green label) independently from those that were initially inside the cell (those with the red label). After labeling the surface receptors and simply returning the cells to 37°C for 2 h (untreated), two different patterns of receptor distribution were evident. HA-5-HT_7(a) and HA-5-HT_7(b) receptors that began on the cell surface (green) remained there, but clustered into some sort of membrane microdomain. The surface population (green) of HA-5-HT_7(d) receptors was nearly all inside the cell and had been replaced with others (red) that were inside at the beginning of the experiment. Still, the majority of these (red) HA-5-HT_7(d) receptors either remain in or has returned to the cytoplasm. HA-5-HT_7(a) and HA-5-HT_7(b) expressing cells assume this state when they are exposed to agonist (5-CT) and surface receptors internalize in response. Agonist-treated HA-5-HT_7(d) cells (5-CT) look very much like the untreated condition. When HA-5-HT_7(d) cells are exposed to antagonist (SB 269970), a significant fraction of surface-labeled receptors (green) remain at the plasma membrane. Conversely, HA-5-HT_7(a) and HA-5-HT_7(b) expressing cells appear much as they do in the untreated condition with receptor clustering (green), but not much internalization.

Discussion

We have found that the three splice variants of the human 5-HT₇ receptor display different trafficking patterns in response to agonist treatment. The 5-HT_7(d) isoform is continuously internalized in an agonist-independent fashion. In fact, 5-HT_7(d) receptors internalize in the presence of a potent 5-HT₇ receptor antagonist. 5-HT_7(d) receptors are properly displayed on the cell surface and respond to agonist treatment with activation of a cAMP-responsive reporter gene. The constitutive internalization of the 5-HT_7(d) isoform is distinct from the signaling pathway as untreated cells show no cAMP-mediated response and a 5-HT₇ receptor antagonist inhibits agonist-mediated reporter gene activity in a dose-dependent manner. Although no such trafficking differences have been reported among the serotonin receptors, similar trafficking patterns have been reported for the α₁ adrenergic receptors (Hirasawa et al., 1997; Chalothorn et al., 2002; Hague et al., 2003; Morris et al., 2004). However, these α₁ adrenergic receptors are transcribed from three separate genes.

We established stable cell lines expressing 5-HT_7(a), 5-HT_7(b), and 5-HT_7(d) receptors. A functional assay that measured the response of a cAMP-responsive luciferase gene revealed a reduced efficacy of the 5-HT_7(d) isoform. The response was only 50% that of the 5-HT_7(a) or 5-HT_7(b) isoforms, even though the 5-HT_7(d) receptors were expressed at equivalent or slightly greater levels. Imaging studies revealed a greater proportion of 5-HT_7(d) receptors in the perinuclear region relative to the plasma membrane in cells that have not been exposed to agonist than what is seen in cell lines expressing 5-HT_7(a) or 5-HT_7(b) receptors. Two possible explanations for this distribution pattern were endocytosis of surface receptors without agonist exposure, or failure of receptors to exit the Golgi and reach the surface. Double-labeling studies of populations of receptors residing at either the surface or inside the cell prior to treatment eliminated the second possibility, as receptors that had been inside the cell moved to the surface. This was true for all three isoforms (Fig. 6A, red label). The difference in trafficking between 5-HT_7(d) receptors and 5-HT_7(a) or 5-HT_7(b) was that the 5-HT_7(d) receptors cycled to the perinuclear region without agonist exposure (Fig. 6A, green label). Surface biotinylation experiments showed some constitutive internalization of all three isoforms, but the proportion of 5-HT_7(d) receptors inside the cell compared with total receptor is much greater. In addition, this internalization of 5-HT_7(d) receptors takes place even in the presence of antagonist. It has been demonstrated by Morello and coworkers that cell permeant antagonists can act as “chemical chaperones” to bring receptors trapped intracellularly to the surface (Morello et al., 2000). Both our imaging and surface biotinylation studies show that while there may be some of this occurring with SB 269970 exposure, certainly many of the 5-HT_7(d) receptors remain inside the cell. Thus, there may be fewer 5-HT_7(d) receptors at the surface of the cell that are available for signaling at any given time, resulting in diminished efficacy in signal transduction via the 5-HT_7(d) receptor.

Amplification of the different carboxy-termini of the three isoforms revealed the presence of 5-HT_7(a), 5-HT_7(b), and 5-HT_7(d) receptors in each tissue or cell type evaluated. The relative abundance of the three isoforms was variable in the three tissue types investigated, with the greatest relative abundance of the 5-HT_7(d) isoform found in smooth muscle tissues. The significance of these distribution differences is not clear; for example, they could indicate specialized functions in tissue such as smooth muscle where expression is greater. Our previous work looking for cellular differences in splicing in rat brain was not able to show any differences (Heidmann et al., 1998). Further examination of individual cell types rather than tissues may reveal single isoform expression.

In this investigation, we have observed that the human 5-HT_7(d) isoform of the 5-HT₇ serotonin receptor has reduced activity when compared with either the 5-HT_7(a) or the 5-HT_7(b) isoform. Our results suggest that this is due to a reduced population of receptors on the surface of the plasma membrane at any given time because of a constitutive agonist-independent endocytosis of the 5-HT_7(d) receptor.

The functional significance of the reduced signaling efficacy of the 5-HT_7(d) receptor on the regulation of serotonin signaling through the 5-HT₇ receptor is unclear. Elucidation of the roles of the human 5-HT₇ receptor variants and their interaction with other cellular proteins remains an objective of further investigation.