Molecular cloning and pharmacological characterization of guinea pig 5-HT1B and 5-HT1D receptors

Human 5-HT1B and 5-HT1D receptors have been implicated as molecular targets for the treatment of acute migraine based upon the pharmacological actions and clinical efficacy of sumatriptan, an agonist for human 5-HT1B/1D receptors. The guinea pig has served as an animal model to assess 5-HT1B/1D receptor function, most recently in evaluating 5-HT1B/1D receptor agonists as potential anti-migraine agents. Since two distinct, but closely-related receptors displaying "5-HT1D receptor pharmacology" have been cloned previously from most mammalian species, the genes encoding these receptors were isolated from a guinea pig liver genomic DNA library using oligonucleotide probes targeted to nonconserved regions of recombinant human 5-HT1B and 5-HT1D receptors. Sequence analysis indicates that guinea pig 5-HT1B and 5-HT1D receptors are comprised 390 and 378 amino acids, respectively. Comparison of the deduced amino acid sequences of guinea pig 5-HT1B and 5-HT1D receptor subtypes show that they display overall and transmembrane (TM) identities of 63% and 77%, respectively. Both clones contain a conserved threonine residue in TM7, a structural feature imparting "5-HT1D receptor pharmacology". Guinea pig 5-HT1B and 5-HT1D receptor genes were transiently expressed in Cos-7 cells and their binding properties were evaluated using [3H]5-HT. Both cloned receptor subtypes displayed "5-HT1D receptor pharmacology" with the following rank order of binding affinities: 5-CT > 5-HT > sumatriptan > 8-OH-DPAT > (-)-pindolol. Ketanserin displayed modest (five-fold) 5-HT1D receptor selectivity, while methiothepin exhibited a similar selectivity for the 5-HT1B subtype. In particular, ketanserin exhibits profound differences in 5-HT1D receptor affinity (and selectivity) across species. High correlations were observed between the binding affinities of serotonergic ligands for 5-HT1D binding sites measured in guinea pig cortical membranes and both cloned guinea pig 5-HT1B (r2 = 0.88) and 5-HT1D (r2 = 0.80) receptors, indicating that the development of subtype selective compounds (i.e. 5-HT1B versus 5-HT1D) using native tissues may be more difficult to achieve without the advantage of using recombinant receptor subtypes. Additionally, there is a good correspondence between binding profiles of recombinant guinea pig 5-HT1B and 5-HT1D receptor subtypes and to their respective cloned human homologs. However, species differences in binding affinities of a subset of compounds are evident. These data extend previous observations that subtype selective (i.e. 5-HT1D) compounds identified in one species may not discriminate between closely related receptors (i.e. 5-HT1B and 5-HT1D) in all animal model systems.

Pharmacological characterizations of recombinant human 5-HT1D alpha and 5-HT1D beta receptor subtypes coupled to adenylate cyclase inhibition in clonal cell lines: apparent differences in drug intrinsic efficacies between human 5-HT1D subtypes

Recombinant human 5-HT1D alpha and 5-HT1D beta receptor subtypes were stably expressed in NIH-3T3 fibroblasts (1D alpha cell line) and Y-1 adrenocortical tumor cells (1D beta cell line), respectively, for pharmacological evaluations of serotonergic compounds to inhibit forskolin-stimulated cAMP accumulation (FSCA). [3H]LSD saturation studies indicated that 5-HT1D receptor expression levels were slightly higher in the 1D beta cell line (Bmax = 1334 +/- 134 fmol/mg protein) than in the 1D alpha cell line (Bmax = 900 +/- 218 fmol/mg protein). 5-HT inhibited FSCA with similar potencies (EC50 approximately 2 nM) in both assay systems. The rank order of agonist potencies in both clonal cell lines matched their pharmacological profiles previously determined in binding studies: dihydroergotamine > or = 5-carboxamidotryptamine (5-CT) > LSD > or = 5-HT > sumatriptan > 1-naphthylpiperazine (1-NP) > yohimbine > 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH DPAT) > 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane (DOI), with Ki/EC50 ratios greater than unity. Methiothepin acted as a silent antagonist at both human 5-HT1D alpha and 5-HT1D beta receptors with apparent dissociation constants (Kb values) of 12 +/- 1 nM and 3 +/- 1 nM, respectively. Whereas GR 127,935, metergoline, DOI, and quipazine acted as full agonists in the 1D alpha cell line, these compounds behaved as partial agonists in the 1D beta cell line. To determine whether high levels of receptor reserve might mask partial agonist activity in the two second messenger assay systems, studies were performed using the irreversible receptor alkylating agent N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline (EEDQ). The relationships between receptor occupancy and inhibition of FSCA were determined for 5-HT, sumatriptan, and 1-NP in both clonal cell lines after partial receptor inactivation using Furchgott analysis. Hyperbolic relationships between receptor occupancy and second messenger response were determined for 5-HT in both transfected cell lines. Steep hyperbolic relationships were also found for sumatriptan and 1-NP in the 1D beta cell line whereas nearly linear relationships were observed for these two compounds in the 1D alpha cell line. Moreover, KA/EC50 ratios of these compounds were significantly larger in the 1D beta (10-32) as compared to the 1D alpha (0.9-2.5) cell line. These data are consistent with the hypothesis that the two heterologous expression systems contain a differential amount of receptor reserve. Despite the presence of an apparently larger-receptor reserve in the 1D beta cell line, GR 127,935, metergoline, DOI, and quipazine behaved as partial agonists. Although the potencies (EC50 values) of compounds matched their respective affinity constants (Ki values) for the closely-related 5-HT1D subtypes, differences in intrinsic activities were observed for a few compounds between the two 5-HT1D receptor expression systems. Since receptor reserve is dependent on the properties of both the assay system and drug, the observed variations in intrinsic activity, although influenced by the variable amounts of receptor reserve in the two transfected cell lines, reflect primarily system-independent differences in the intrinsic efficacy of the tested compounds at the two human 5-HT1D receptors. Higher intrinsic efficacies of compounds at the human 5-HT1D alpha receptor relative to the human 5-HT1D beta subtype may be responsible for the higher intrinsic activities observed in the 1D alpha cell line, even though receptor reserve is apparently lower in this system.

Differences in ligand binding profiles between cloned rabbit and human 5-HT1D alpha and 5-HT1D beta receptors: ketanserin and methiothepin distinguish rabbit 5-HT1D receptor subtypes

The study of serotonin receptor function has been complicated by the extreme molecular diversity of serotonin receptor subtypes, the lack of selective agonists and antagonists for many of the subtypes, and divergence in the pharmacological properties of a single receptor subtype across different animal species. An example of this pharmacological diversity between species homologues is provided by the 5-HT1D receptor subfamily. To further advance the ability to characterize and pharmacologically compare functional responses mediated by native 5-HT1D receptors, we have cloned the 5-HT1D alpha and 5-HT1D beta receptor subtypes from the rabbit and evaluated their pharmacological profiles using radioligand binding assays. The deduced amino acid sequences of the rabbit 5-HT1D alpha and 5-HT1D beta receptor genes displayed 60% overall identity [75% transmembrane (TM) identity] to each other and > 90% overall identity (95% TM identity) to their corresponding human homologues. Two compounds were identified in binding assays which discriminated between the closely-related 5-HT1D receptors. Ketanserin exhibited high affinity (pKi = 7.66) and selectivity (> 20-fold) for the 5-HT1D alpha receptor while methiothepin displayed high affinity (pKi = 7.86) and selectivity (16-fold) for the 5-HT1D beta receptor subtype. The rabbit and human recombinant 5-HT1D receptors showed significant intraspecies (rabbit 5-HT1D alpha vs. 5-HT1D beta) and interspecies (i.e. rabbit vs. human 5-HT1D alpha) similarities in their ligand binding profiles. These data suggest that 5-HT1D-mediated responses in rabbit preparations may provide information relevant to the pharmacology of the 5-HT1D receptor subtypes in humans.

Ketanserin and ritanserin discriminate between recombinant human 5-HT1D alpha and 5-HT1D beta receptor subtypes

Compounds able to discriminate functionally between the closely related cloned human 5-HT1D alpha and 5-HT1D beta receptor subtypes have not been reported previously. In [3H]5-HT competition assays, the classical 5-HT2A receptor antagonists, ritanserin and ketanserin, displayed moderate affinity (pKi = 7.30 and 7.17, respectively) and marked selectivity (22- and 71-fold, respectively) for the recombinant human 5-HT1D alpha subtype relative to the 5-HT1D beta receptor. In contrast, the nonselective 5-HT1/2 receptor antagonist, methiothepin, exhibited similar binding affinities (pKi = 7.64-8.01) for both recombinant 5-HT1D subtypes. The antagonistic properties of these compounds were evaluated for their ability to block 5-HT-induced inhibition of forskolin-stimulated cAMP accumulation in intact cells stably expressing either 5-HT1D alpha or 5-HT1D beta receptors. All three compounds behaved as antagonists devoid of intrinsic activity in the functional assays. The apparent pKb values determined in functional assays closely matched their pKi values obtained in binding assays. Since ketanserin exhibits significant selectivity for the human 5-HT1D alpha receptor, this antagonist can be used as a pharmacological tool to discriminate between 5-HT1D alpha and 5-HT1D beta receptor-mediated responses in human tissues.

Dual coupling of cloned human 5-hydroxytryptamine1D alpha and 5-hydroxytryptamine1D beta receptors stably expressed in murine fibroblasts: inhibition of adenylate cyclase and elevation of intracellular calcium concentrations via pertussis toxin-sensitive G protein(s)

The second messenger coupling of cloned human 5-hydroxytryptamine (5-HT)1D alpha and 5-HT1D beta receptors stably expressed in murine fibroblasts (LM (tk-)) was investigated. Clonal cell lines expressing similar receptor densities (Bmax = 750-950 fmol/mg) were used in this study. 5-HT (EC50 = 1.5-2.0 nM) and sumatriptan (EC50 = 6-14 nM), a selective 5-HT1D agonist, produced dose-dependent inhibition of forskolin-stimulated cAMP accumulation in intact cells transfected with the 5-HT1D alpha or 5-HT1D beta receptor gene. The maximal inhibitory responses elicited by these agonists were slightly greater with the 5-HT1D alpha receptor (approximately 90%) than the 5-HT1D beta receptor (approximately 80%). 5-HT (EC50 = 1.7-2.4 nM) and sumatriptan (EC50 = 8-18 nM) also evoked dose-dependent elevations in intracellular calcium concentrations ([Ca2+]i), with EC50 values that were indistinguishable from those for inhibition of forskolin-stimulated cAMP accumulation. Cells expressing 5-HT1D beta receptors displayed significantly larger 5-HT-induced increases in [Ca2+]i than did cells expressing 5-HT1D alpha receptors (206 nM versus 114 nm increase; p < 0.01). Dose-dependent elevations in inositol phosphates (IP) were also observed after application of 5-HT (EC50 = 29-54 nM) or sumatriptan (EC50 = 73-481 nM); the maximal increases in IP accumulation were modest (51-69%) for both 5-HT1D subtypes. In contrast to the cAMP and calcium responses, the concentration-response curves for IP accumulation were shifted to the right at least 10-fold. Methiothepin, a nonselective 5-HT1 antagonist, competitively antagonized the cAMP response, yielding an apparent dissociation constant (Kb) of 3-4 nM for the 5-HT1D receptors. Methiothepin (10 microM) significantly reduced the elevations in [Ca2+]i (> 90%) and IP (> 75%) evoked by saturating concentrations (1 microM) of agonists. All three functional responses were significantly attenuated (> 90%) by pretreatment with 100 ng/ml pertussis toxin. The sumatriptan-induced elevation of [Ca2+]i via activation of the 5-HT1D subtypes may provide a molecular mechanism of action by which sumatriptan could directly constrict cerebral blood vessels and alleviate migraine symptoms.