Introduction

Top Abstract Introduction Materials Results Discussion References

Serotonin receptors display the greatest molecular diversity of any family of G protein-coupled receptors. Fourteen receptor subtypes have been cloned and subdivided into seven distinct subfamilies (5-HT_1-7) based on structural (deduced amino acid sequence, transmembrane homology), operational (binding properties, pharmacology) and transductional (effector coupling) criteria (Hoyer et al., 1994). With the exception of the 5-HT₃ receptor that is a ligand-gated ion channel, all other serotonin receptors are members of the superfamily of G protein-coupled receptors including three subtypes that couple to the stimulation of adenylate cyclase (5-HT₄, 5-HT₆ and 5-HT₇). Identifying the functional correlates of the cloned serotonin receptors and hence their therapeutic potential is challenging, particularly for the more novel subtypes such as 5-HT₆ and 5-HT₇ receptors, because selective compounds have yet to be developed.

One functional correlate of 5-HT₇ receptor activation includes smooth muscle relaxation observed in a variety of isolated tissue preparations, in which elevations of [cAMP]_i concentrations were also detected (Branchek and Zgombick, 1997; Eglen et al., 1997 for reviews). Further evidence for the mediation of the relaxant response via the 5-HT₇ receptor is provided by the localization of mRNA transcripts encoding the 5-HT₇ receptor in many blood vessels by RT-PCR (Ullmer et al., 1995). An additional physiological response that may be mediated by the 5-HT₇ receptor is the regulation of photic entrainment of circadian rhythms (Ying and Rusack, 1997), potentially via an inhibition of a GABA_A receptor-activated current (Kawahara et al., 1994). However, neither the 5-HT₇ receptor nor its mRNA has been localized to the hypothalamic suprachiasmatic nuclei (Lovenberg et al., 1993; To et al., 1995; Gustafson et al., 1996). In other in vitro and in vivo preparations, 5-HT₇ responses are intermixed with one or more additional serotonin effects. For example, the 5-HT₇ receptor mediates the high affinity portion of the biphasic stimulation of adenylate cyclase by 5-CT in guinea pig hippocampal membranes (Tsou et al., 1994). Similarly, the 5-HT₇ receptor may also be involved in the tertiary, prolonged hypotensive response to i.v. administration of 5-HT in the rat (De Vries et al., 1997).

Alternative splicing of the mRNA encoding intron-containing serotonin receptors contributes to further structural diversity, resulting in the generation of additional receptor variants. This has been demonstrated for genes encoding two G_s-coupled serotonin receptors, the 5-HT₄ (Gerald et al., 1995; Blondel et al., 1998) and 5-HT₇ (Bard et al., 1993; Heidmann et al., 1997; Jasper et al., 1997; Stam et al., 1997) subtypes, in which splice variants differ in length of their predicted carboxy termini. An intron-containing h5-HT₇ receptor gene (5-HT_7(a)) was first isolated by Bard and colleagues (1993) from a genomic library and was found to encode a protein of 445 amino acids. Subsequently, several groups reported the cloning of a cDNA encoding a second h5-HT₇ receptor isoform (5-HT_7(b)) identical to that described by Bard et al. (1993), except that it contained a truncated carboxy terminus (Heidmann et al., 1997; Stam et al., 1997; Jasper et al., 1997). A third splice variant of the h5-HT₇ receptor (5-HT_7(d)) was identified containing an additional exon and hence a longer carboxy tail than the other two 5-HT₇ receptor isoforms (Heidmann et al., 1997). The 5-HT_(7d) splice variant represents a minor molecular species (<5%) compared to the relative abundance of the other two 5-HT₇ receptor isoforms (Heidmann et al., 1997). Although the binding properties of the human 5-HT_7(a) and 5-HT_7(b) splice variants have been well characterized (Bard et al., 1993; Jasper et al., 1997), detailed functional characterization has only been reported for the h5-HT_7(b) receptor (Jasper et al., 1997). In our study, a clonal cell line was generated expressing the h5-HT_7(a) receptor isoform to determine the potencies and intrinsic activities of reference serotonergic agents using the elevation of [cAMP]_i as a functional measure of compound activity.

	Methods and Materials

Top Abstract Introduction Materials Results Discussion References

Generation of the stable cell line. The entire coding region of the h5HT_7(a) receptor gene was subcloned into the expression vector pcEXV-3 (Bard et al., 1993). Murine fibroblasts (LMtk) were used as the transfection host to establish a stable cell line expressing this subtype by the calcium phosphate method using G-418 as a selection marker (Zgombick et al., 1991). Clonal cells were grown as adherent monolayers under standard conditions (5% CO₂, 37°C) in serum-free media (AIM V, GIBCO BRL, Grand Island, NY). Transfected cells reached approximately 90% confluency prior to use in radioligand binding and second messenger assays.

[³H]5-HT and [³H]LSD binding assays. Membranes were prepared from clonal cells using standard techniques (Branchek et al., 1990). Radioligand binding studies were performed according to the methodology outlined by Zgombick and colleagues (1991). Expression levels of the h5-HT_7(a) receptor were determined from both [³H]5-HT (agonist) and [³H]LSD (antagonist) saturation studies using eight concentrations of radioligand (0.2-20 and 0.5-60 nM, respectively). Competition studies were performed using 5 nM [³H]5-HT and 10 concentrations of competitor. Unlabeled 5-HT (10 µM) defined nonspecific binding. Membranes were incubated for 30 min at 37°C and the assay was terminated by vacuum filtration. Protein concentrations were determined by the method of Bradford (1976) using bovine serum albumin as the standard.

Intracellular cAMP assays. The h5-HT_7(a) receptor-mediated elevations of [cAMP]_i were determined using a protocol described previously (Zgombick et al., 1993), with the exception that forskolin was omitted from the assay. Transfected cells were plated in 96-well microtiter plates (5 × 10³ cells/well) and allowed to grow for 96 hr. Cells were preincubated for 20 min in N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES)-buffered saline (150 mM NaCl, 10 mM HEPES, 5 mM theophylline, 10 µM pargyline, pH 7.4 at 37°C). Antagonists were also incubated during this preincubation period in a subset of experiments. In separate experiments, cholera toxin (20 µg/ml) was added 18 hr before the assay. Cells were incubated with 10 concentrations of agonist for an additional 10 min (5% CO₂, 37°C) to generate CRCs. The assay was terminated by the removal of medium and the addition of 160 µl of 0.1 M HCl. Intracellular cAMP was determined by radioimmunoassay (PerSeptive Diagnostics, Cambridge, MA).

Data analysis. Binding (K_d, B_max and IC₅₀ values) and response (EC₅₀ and E_max) parameters were determined by nonlinear regression analysis (GraphPAD Prism, San Diego, CA). IC₅₀ values were converted to K_i values using the Cheng-Prusoff equation (1973). Apparent dissociation constants of antagonists (K_b values) were determined from the shift of the EC₅₀ value of 5-HT in the absence and presence of one concentration of antagonist (~10 × K_i values) by the method of Arunlakshana and Schild (1959).

Drugs. [³H]LSD (specific activity 76.7 Ci/mmol) and [³H]5-HT (25.2 Ci/mmol) were obtained from New England Nuclear (Boston, MA). 5-Carboxamidotryptamine, 5-hydroxytryptamine, 5-methoxytryptamine, N,N-dipropyl-5-carboxamidotryptamine, N,N-dimethyl-5-methoxytryptamine, 2-methyl-5-hydroxytryptamine, -methyl-5-hydroxytryptamine, tryptamine, 8-hydroxy N,N-dipropyl aminotetralin, 1-naphthylpiperazine, methiothepin, dihydroergotamine, mesulergine, metergoline, clozapine, ritanserin, spiperone, cyproheptadine, ketanserin and rauwolscine were purchased from Research Biochemicals Inc. (Natick, MA). d-LSD was supplied from the National Institute of Drug Abuse (Bethesda, MD).

	Results

Top Abstract Introduction Materials Results Discussion References

Expression levels (B_max values) of the recombinant h5-HT_7(a) receptor in membranes derived from stably transfected murine fibroblasts were determined from parallel saturation experiments utilizing both [³H]5-HT (agonist) and [³H]LSD (antagonist). Both [³H]LSD and [³H]5-HT bound with high affinity and in a saturable manner to membranes derived from this clonal cell line (fig. 1, A and B). Binding parameters for the [³H]radioligands (K_d and B_max values) are summarized in table 1. The site densities of the 5-HT_7(a) receptor in clonal cell membranes determined from [³H]antagonist and [³H]agonist saturation studies were very similar.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   [3H]LSD (A) and [3H]5-HT (B) binding isotherms using membranes harvested from stable transfectants expressing the recombinant human 5-HT7(a) receptor. Transformed data are represented in the form of Scatchard plots (inset). Binding parameters (Kd and Bmax values) are summarized in table 1.

Competition experiments were performed using [³H]5-HT to determine the affinity constants (K_i values) of serotonergic compounds for the h5-HT_7(a) receptor in a stably expressing cell line since we previously reported K_i values using membranes from transiently-transfected Cos-7 cells. These binding studies permitted a direct comparison between compound affinities and their potencies (agonist and/or antagonist) determined in the cAMP second messenger assay. Binding affinities (K_i values) for agonists and antagonists are listed in tables 2 and 3, respectively. The rank order of binding affinities was consistent with 5-HT₇ receptor pharmacology: 5-CT > 5-HT  5-MeOT  methiothepin > mesulergine > 8-OH-DPAT > spiperone > sumatriptan > ketanserin > rauwolscine. Higher binding affinities (2- to 4-fold) were noted for compounds that were identified as agonists in our cAMP assay relative to their previously reported K_i values. This discrepancy is due to the lower affinity (K_d = 8 nM) of [³H]5-HT for the h5-HT_7(a) receptor in Cos-7 membranes (Bard et al., 1993) than that obtained with the stable cell line used in this study (table 1). Much higher expression levels of the h5-HT_(7a) receptor were obtained in Cos-7 cells (7 pmol/mg protein; Bard et al., 1993) and may have contributed to the lower affinity for [³H]5-HT if endogenous G_s were rate-limiting. In contrast, no significant differences in binding affinities were noted for agents identified as antagonists in the cAMP second messenger assay since these compounds are less sensitive to the affinity state of the receptor as compared to agonists.

In stably transfected LM(tk) cells, 5-HT produced a concentration-dependent elevation of [cAMP]_i, with an EC₅₀ value of 80 nM and a maximal stimulation of 5.4 ± 0.4-fold above basal values (n = 50; fig. 2; table 2). The rank order of agonist potencies in the 5-HT_7(a) receptor second messenger assay paralleled their rank order of binding affinities: 5-CT > 5-HT  5-MeOT > 8-OH-DPAT > sumatriptan (fig. 2; table 2). Agonist potencies to stimulate cAMP production was more than 25-fold lower relative to their respective binding affinities (table 2). For tryptamine derivatives, the magnitude of the potency/affinity ratios (EC₅₀/K_i) was inversely related to their binding affinity constants (i.e., the higher the binding affinity the smaller the discrepancy in potency). A high correlation index (r² = 0.81) was obtained between the K_i values and EC₅₀/K_i ratios for a series of tryptamine congeners (fig. 3).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   CRCs for the stimulation of [cAMP]i by serotonergic agonists in intact cells stably expressing the recombinant human 5-HT7(a) receptor. Data was normalized to 100% relative to the maximal response elicited by 5-HT. Each data point represents a single determination. Experiments (n = 5-10) were performed as described in "Materials and Methods." , 5-CT; , 5-HT; , 5-MeOT; , DP-5-CT; , 5-MeO-DMT; , 1-NP; , 8-OH-DPAT; , Sumatriptan.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Correlation between potency/binding affinity ratios (EC50/Ki) and binding affinities (Ki) of tryptamine derivatives obtained with the recombinant human 5-HT7(a) receptor. The correlation index (r2) is listed in the graph. Numbers correspond to the following compounds: 1, 5-CT; 2, 5-HT; 3, 5-MeOT; 4, DP-5-CT; 5, 5-MeO-DMT; 6, tryptamine; 7, -Me-5-HT.

Methiothepin and several ergot alkaloids (i.e., d-LSD) antagonized the 5-HT-mediated elevation in [cAMP]_i (fig. 4). In contrast to agonists in which large variations were obtained between K_i and EC₅₀ values, the apparent K_b values of antagonists determined for the h5-HT_7(a) receptor in the second messenger assay were in close agreement with their respective K_i values in binding assays (table 3). Compounds identified as antagonists did not display inverse agonism in our heterologous expression system.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Determination of the apparent dissociation constant (Kb value) of d-LSD for the recombinant human 5-HT7(a) receptor. CRCs for the stimulation of [cAMP]i by 5-HT were generated in the absence () and presence () of 50 nM d-LSD. Data were normalized to 100% relative to the maximal response elicited by 5-HT. Each data point represents a single determination. Experiments (n = 5) were performed as described in "Materials and Methods." Apparent Kb values were determined from the Schild equation and are summarized in table 3.

Pretreatment of clonal cells with cholera toxin (20 µg/ml for 18 hr) abolished the 5-HT-mediated elevation of [cAMP]_i (fig. 5). No changes in inositol phosphate production or intracellular calcium concentrations were observed after exposure of stable transfectants to 1 µM 5-HT (data not shown). These data indicate that the h5-HT_7(a) receptor directly interacts with G_s protein(s) and does not couple to G_q/G₁₁ or G_i pathways. Similar observations were reported previously in which the h5-HT_7(b) receptor isoform coupled only to adenylate cyclase stimulation in transfected HEK 293 cells (Jasper et al., 1997).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Effect of cholera toxin pretreatment on the 5-HT-induced elevation of [cAMP]i in intact cells stably expressing the recombinant human 5-HT7(a) receptor. CRCs for the stimulation of [cAMP]i by 5-HT was generated in control () and cholera toxin-pretreated () cells. Data were normalized to 100% relative to the maximal response elicited by 5-HT. Each data point represents a single determination. Experiments (n = 3) were performed as described in "Materials and Methods."

	Discussion

Top Abstract Introduction Materials Results Discussion References

RT-PCR studies have assisted in the identification of four mammalian 5-HT₇ receptor isoforms that are structurally divergent in their predicted intracellular carboxy termini. The 5-HT_7(a) receptor was the first isoform cloned from human with a predicted length of 445 amino acids (Bard et al., 1993). Alternative splicing of a second intron in the coding region located near the carboxy terminus of the human gene generates a 13 amino acid truncated receptor isoform (5-HT_7(b)) due to a five nucleotide base insertion which introduces an in-frame stop codon (Jasper et al., 1997; Heidmann et al., 1997; Stam et al., 1997). Two additional splice variants, designated 5-HT_7(c) (rat) and 5-HT_7(d) (human), are produced by a retained exon cassette, which is distinct between the two species. These other isoforms encode gene products with unique carboxy-terminal amino acid sequences different from both the 5-HT_7(a) and 5-HT_7(b) receptors as well as from each other (Heidmann et al., 1997). The 5-HT_7(a) and 5-HT_7(b) receptor isoforms represent the predominant splice variants in both rat and human (Heidmann et al., 1997).

The binding properties of the recombinant human 5-HT_7(a) (Bard et al., 1993) and 5-HT_(7b) (Jasper et al., 1997) receptor isoforms have been reported previously. The discriminating properties of h5-HT₇ receptor pharmacology include a rank order of binding affinities: 5-CT > 5-HT  5-MeOT  methiothepin > 8-OH-DPAT > spiperone > sumatriptan > ketanserin > rauwolscine. In addition, a series of antipsychotic and tricyclic antidepressant compounds display high affinity for the 5-HT₇ receptor (Roth et al., 1994). Because detailed functional information (intrinsic activities and potencies) of reference serotonergic compounds has been determined only for the h5-HT_(7b) splice variant (Jasper et al., 1997), the h5-HT_(7a) receptor isoform was stably expressed in murine fibroblasts (LMtk) and second messenger responses were investigated to gain further insight into the pharmacological properties of the different splice variants of the h5-HT₇ receptor.

Low site densities of the h5-HT_7(a) receptor in LM(tk) membranes were obtained from saturation studies using both [³H]5-HT and [³H]LSD (163 and 189 fmol/mg protein, respectively) (table 1). The similar (and low) receptor expression levels obtained with both [³H]agonist and [³H]antagonist radioligands are suggestive that endogenous G_s protein(s) is not rate limiting in this cell line and that h5-HT_7(a) receptors exist in the high affinity (G protein-coupled) state. However, the relationship between the high affinity state of the receptor (detected from [³H]5-HT competition studies) and the affinity state of the h5-HT_7(a) receptor that mediates the stimulation of adenylate cyclase(s) is not known at present.

In agreement with previous studies, 5-HT evoked a concentration-dependent elevation of [cAMP]_i in intact cells expressing the h5-HT_7(a) receptor (fig. 2). The rank order of agonist potencies to stimulate cAMP formation paralleled their rank order of binding affinities: 5-CT > 5-HT  5-MeOT > 8-OH-DPAT > sumatriptan. However, agonist potencies ranged from 25-950-fold lower than their respective binding affinities (table 2). Jasper and colleagues (1997) also noted significantly lower agonist potencies relative to their binding affinities for the h5-HT_7(b) receptor stably expressed in HEK 293 cells, although smaller differences were observed (8- to 75-fold) relative to that obtained in our study. Agonist potencies for the h5-HT_7(a) receptor (table 2) were 3- to 17-fold lower than for the same subset of compounds for the h5-HT_7(b) receptor (Jasper et al., 1997). The 3-fold higher binding affinities of agonists for the h5-HT_7(b) receptor (Jasper et al., 1997) relative to the h5-HT_7(a) splice variant (table 2) may underlie a portion of the observed agonist potency differences between the receptor isoforms. Indeed, Jasper and colleagues (1997) determined the binding properties of the two splice variants of the h5-HT₇ receptor in parallel using [³H]5-CT and observed small variations in receptor pharmacology, with a conservation of the rank order of binding affinities. These data would suggest that differences in experimental methodologies employed in the two investigations (i.e., radioligands, buffer, transfection host) contribute to the lower binding affinities reported in this study relative to those of Jasper et al. (1997). The much higher expression of the h5-HT_7(b) receptor in HEK 293 cells (>7 pmol/mg protein; Jasper et al., 1997) relative to the lower density of the h5-HT_7(a) receptor in LM(tk) cells (~200 fmol/mg protein; fig. 1) may contribute to the observed variations in agonist potencies between the two studies. Differential expression of native G_s and adenylate cyclase isoforms in the two transfection hosts could also lead to differences in the efficiency of receptor coupling to adenylate cyclase stimulation (and hence agonist potencies) by the two splice variants. The low potency of agonists to elicit functional responses has also been noted in native 5-HT₇ receptor systems from various species, suggestive that low efficiency in receptor-effector coupling may be an inherent property of this subtype (Tsou et al., 1994; Leung et al., 1996; Schoeffter et al., 1996; Terrón, 1996; Eglen et al., 1997 for review).

The high EC₅₀/K_i ratios reported here and by others (see above) for the 5-HT₇ receptor, is not necessarily unexpected. In the context of current receptor theory, one would expect an EC₅₀ value for an agonist in a functional assay to be less than or equal to its affinity for the uncoupled state of the receptor (K_low) calculated from antagonist competition binding studies performed in the presence of guanine nucleotides. Indeed, this affinity state (K_low) corresponds to that derived from receptor inactivation studies (K_A) using the method of Furchgott (Furchgott and Bursztyn, 1967). In contrast, the K_i values measured in our binding studies are presumably the affinity of agonists for the G protein-coupled state of the receptor (K_(high)) because [³H]5-HT was used as the radioligand in the absence of guanine nucleotides. The nature of the relationship between K_(high) and the EC₅₀ for an agonist is unclear currently. As the designation implies, the binding affinity (K_i) of a compound is a "constant" dependent primarily upon its molecular structure, whereas agonist potency (EC₅₀) varies as a function of its intrinsic efficacy and the degree of receptor reserve present in the assay system. Therefore, the EC₅₀/K_i ratios might be expected to vary similarly such that one would expect highly efficacious compounds to show a higher ratio as compared to partial agonists. The calculated EC₅₀/K_i ratios obtained in our study may be an indirect measure of the intrinsic efficacy of the agonist tested and would be expected to vary analogous to the K_low/K_high ratios. It has been demonstrated for several G protein-coupled receptors that the magnitude of agonist K_low/K_high ratio (guanine nucleotide shift) calculated from binding studies is an estimate of the intrinsic efficacy of that compound, with the most efficacious agonists displaying the largest ratios (Kenakin, 1993).

Unexpectedly, we observed a reverse correlation between EC₅₀/K_i ratios and K_i values of agonists (table 2; fig. 3). For example, we expected this ratio to be lower for 5-MeOT, -Me-5-HT and 8-OH-DPAT as compared to 5-HT and 5-CT, because the former compounds have been shown to be either inactive or partial agonists (and therefore possess lower intrinsic efficacy than the latter compounds) in several 5-HT₇ receptor-expressing systems (Lovenberg et al., 1993; Leung et al., 1996; Schoeffter et al., 1996; Terrón, 1996). The higher intrinsic efficacies of -Me-5-HT and 8-OH-DPAT in transfected cell lines (table 2) (Jasper et al., 1997) relative to native systems implies that a degree of receptor reserve may be present in the clonal cells, even though receptor expression may be low (<200 fmol/mg protein; table 1). The significance of the relationship between EC₅₀/K_i ratios and intrinsic efficacies obtained in this study and its relevance to signal transduction is not clear at present, but our results indicate that these ratios may not be predictive of the intrinsic efficacy of compounds in mediating physiological responses at native 5-HT₇ receptors. Differences in receptor-effector coupling, species variations in pharmacology or the experimental methodologies employed may contribute to variations in the intrinsic efficacy of the above agonists in different systems. Studies using transfected cells expressing lower number of receptors may provide a more accurate estimate of intrinsic efficacy and may permit the identification of partial agonism for compounds such as -Me-5-HT, since the degree of receptor reserve could be adequately low for 1-NP (arylpiperazine; table 2) but not for a structurally divergent series as the tryptamines. Determination of K_low values of agonists in [³H]LSD (antagonist) competition binding assays using nonhydrolyzable GTP analogues and generation of K_A values by partial receptor alkylation may also assist in elucidating further the underlying molecular mechanisms involved between receptor activation and the generation of the second messenger response for this subtype.

In accordance with 5-HT₇ receptor pharmacology, methiothepin and several ergot alkaloids such as d-LSD competitively antagonized the h5-HT_7(a) receptor-mediated elevation of [cAMP]_i (fig. 4; table 3). These compounds did not display inverse agonism in our heterologous expression system. With the exception of ketanserin, antagonist potencies (K_b values) to block 5-HT-mediated cAMP production were in close agreement with affinity constants (K_i values) derived from [³H]5-HT competition binding (<3-fold; table 3), which contrast the results obtained with agonists (see above). These data differ markedly from those reported by Jasper et al. (1997) using the h5-HT_(7b) receptor, in which antagonist K_b values determined in second messenger studies were 5- to 40-fold lower than their respective affinity constants obtained from binding studies. These authors postulated that the receptor may attain different conformational states during functional interactions with G_s in second messenger and radioligand binding assays. Alternatively, the disparities may reflect differences in experimental methodology between the two investigations (radioligands, transfection hosts). To better evaluate the apparent variations in agonist and antagonist potencies observed between the two h5-HT₇ splice variants, it is critical to perform parallel studies by expressing both receptor isoforms in the same cell line at similar and physiologically relevant levels.

The physiological significance of multiple splice variants of the h5-HT₇ receptor has yet to be fully appreciated. However, recent evidence suggests that these alternatively spliced 5-HT₇ receptor isoforms may be differentially expressed in human brain and smooth muscle. RT-PCR studies have identified the 5-HT_7(a) receptor as the predominant molecular species in the human brain although the 5-HT_(7b) receptor as the primary transcript in human smooth muscle (Hamblin and Heidmann, 1997). Structural differences in the carboxy termini of the h5-HT₇ receptor splice variants indicate potential differential interactions with PDZ-domain-containing proteins, determinants of protein organization with the plasma membrane (Martin et al., 1998). Additionally, these two splice variants may display differential coupling to G_s isoforms. Splice variants of the h5-HT₇ receptor differ in their length of their C-termini, and this region of G_s-coupled receptors is a structural domain implicated in molecular interactions with G_s (Strader et al., 1994). However, it is highly unlikely that the human 5-HT_(7a) and 5-HT_(7b) receptor isoforms are differentially down-regulated by phosphorylation because neither splice variant contains known consensus sequences for potential phosphorylation in their predicted intracellular carboxy termini. In contrast, the h5-HT_(7d) possesses two phosphorylation consensus sites (Heidmann et al., 1997) and may be regulated in a phosphorylated-dependent manner.

The detailed pharmacological characterization of the cloned h5-HT_7(a) receptor and other related splice variants in heterologous expression systems will continue to facilitate the interpretation of in vivo and in vitro responses to 5-HT in native tissues and aid in the identification of additional 5-HT₇ receptor-mediated responses. The physiological roles and potential therapeutic significance of splice variants of this receptor are only now becoming appreciated. The development of selective 5-HT₇ receptor agents will accelerate the investigation of receptor function and regulation as well as assist in the delineation of the role of 5-HT₇ receptor isoforms in a human pathophysiology.