Introduction

Summary Introduction Materials & Methods Results Discussion References

Octopamine, the monohydroxylated analogue of norepinephrine, is a well-established neurotransmitter, neurohormone, and neuromodulator in many invertebrate species (1). In particular, various insect preparations have been used to study the role of octopamine. Because many of the octopamine-mediated responses are connected to adaptation to stressful circumstances, the octopaminergic system has been considered to be the invertebrate equivalent of the vertebrate sympathetic nervous system. Although some early reports describe effects of octopamine in vertebrates (1), it is present in only trace amounts and probably plays no significant role in neurotransmission. This raises the possibility of specifically interfering with octopamine neurotransmission in invertebrates, an issue of considerable interest for the pesticide industry.

Octopamine specifically interacts with octopamine receptors, which belong to the superfamily of G protein-coupled receptors. The pharmacology of octopamine receptors is well documented in several insect species, especially in locusts. Four octopamine receptor subtypes (oct-1, oct-2A, oct-2B, oct-3) have been characterized, each with pharmacological properties (for a review, see Ref. 2). Stimulation of oct-1 receptors induces an increase in intracellular calcium concentration, whereas oct-2A, oct-2B, and oct-3 receptors are positively coupled to AC.

Octopaminergic neurotransmission has been studied not only in insects; biochemical and immunocytochemical studies have demonstrated the presence of octopamine in the brain of several molluscan species (3-7). In addition, the physiology and pharmacology of octopamine receptors have been examined (3, 8-11).

Despite the large interest in the octopaminergic system, the molecular structure of octopamine receptors has not been elucidated. However, two closely related receptors that exhibit highest affinity for tyramine were cloned from Drosophila (12, 13) and Locusta (14). Tyramine, which is the precursor of octopamine, was also suggested to play a role as a neurotransmitter in insects (15), but it is less extensively studied than octopamine. Although tyramine exhibits a higher affinity for the Drosophila receptor than does octopamine and tyramine is a more potent inhibitor of AC, both transmitters exhibit similar potencies for raising intracellular calcium levels in cells expressing this receptor (16). Therefore, the Drosophila tyramine receptor (13) has been suggested to function as a combined tyramine/octopamine receptor and will be referred to as Tyr/Oct-Dro.

We present the cloning and functional expression of a cDNA encoding a novel G protein-coupled bioamine receptor that is expressed in the CNS of the freshwater snail Lymnaea stagnalis. This receptor is designated Lym oa₁ according to the most recent recommendations by the Committee for Receptor Nomenclature of the International Union of Pharmacology (17). The predicted amino acid sequence of Lym oa₁ exhibits highest sequence similarity with the sequence of Tyr/Oct-Dro (12, 13), Tyr-Loc (14), and mammalian -adrenergic receptors. However, Lym oa₁ clearly differs from these receptors in its high affinity for octopamine and its unique signaling properties: octopamine stimulates both PLC and AC. On the basis of these data, we concluded that we had cloned an octopamine receptor.²

	Materials and Methods

Summary Introduction Materials & Methods Results Discussion References

Animals. Adult Lymnaea stagnalis (shell height, 28-34 mm) were bred in the laboratory under standard conditions (18).

Isolation and sequence analysis of the cDNA encoding Lym oa₁. Total RNA (5 µg) was isolated from L. stagnalis CNS (19) and reverse-transcribed into cDNA using oligo(dT) primers and Superscript reverse transcriptase (GIBCO BRL, Baltimore, MD). This cDNA was used as a template in a PCR with degenerate oligonucleotides that recognize conserved amino acids in TMs 6 and 7 of G protein-coupled bioamine receptors. The primers and PCR conditions were described previously (20). Sequence analysis (Sequenase, GIBCO BRL) of the PCR products revealed that one of the products showed significant similarity to the mammalian adrenergic receptors. Based on the sequence of this fragment, a specific antisense primer (5-CAGAAGGATCCGACCTGGTAAATGGTG-3, based on bp 2563-1584; Fig. 1) was used to isolate the corresponding full-length cDNA clone. A PCR-based library screening was performed on a cDNA library of L. stagnalis CNS, cloned in -ZAP according to a method described by Gibbons et al. (21). The cDNA insert was excised in vivo as a pBluescript SK phagemid (designated pBS-Lym oa₁) and sequenced on both strands (Sequenase) using a walking primer strategy.

View larger version (43K): [in this window] [in a new window]   Fig. 1.   Nucleotide sequence and predicted amino acid sequence of Lym oa1. Boxes, putative TM regions labeled I-VII; white letters, serine or threonine residues represent consensus sites for protein kinase C-mediated phosphorylation; gray circles, residues putatively involved in ligand binding.

Construction of Lym oa₁ expression vector and expression of Lym oa₁ in mammalian cells. The 5 part of the open reading frame of Lym oa₁ was amplified in a PCR using pBS-Lym oa₁ as a template and the following two primers: (i) a sense primer (5-GCAGGATCCACCATGGACTACAAGGACGACGATGACAAGATGTCACGTGACATCTTCATGA-3), which starts with a BamHI restriction site (bold letters), followed by a sequence comprising the start codon (italic letters), a region encoding an eight-amino acid FLAG peptide (underlined letters), and a region based on bp 956-977 (see Fig. 1); and (ii) an antisense primer (5-CAGACTCCTGCCACCAAG-3), based on bp 1394-1411 (see Fig. 1). The resulting PCR product was cloned as a BamHI/KpnI fragment into pBluescript SK⁺, and the resulting subclone (pBS-5Lym oa₁) was sequenced on both strands. The 3 part of the open reading frame of Lym oa₁ was amplified similar to the manner described for the 5 part, using the following primers: (i) a sense primer (5-TGACCTGCTCGTGTAAGG-3), based on bp 2715-2732 (see Fig. 1); and (ii) an antisense primer (5-GCTCTAGACCTAGTTTAAGTGTATCCG-3), starting with an XbaI site (bold letters) and followed by a sequence comprising the predicted stop codon (italic letters), based on bp 2860-2879 (see Fig. 1). A 113-bp AatII/XbaI fragment of the resulting PCR product was ligated to a KpnI/AatII fragment of the original cDNA clone, and the resulting subclone (pBS-3Lym oa₁) was sequenced. The final expression construct (pcDNA3-Lym oa₁) was made by ligation of a BamHI/KpnI fragment of pBS-5Lym oa₁ and a KpnI/XbaI fragment of pBS-3Lym oa₁ into the BamHI/XbaI sites of pcDNA3 (InVitrogen, San Diego, CA). pcDNA3-Lym oa₁ was introduced into HEK 293 cells by calcium phosphate-mediated transfection. Transfected cells were grown in the presence of 0.8 mg/ml geneticin (G-418, GIBCO, Grand Island, NY), and membranes of clonal cell lines were selected for binding to [³H]rauwolscine (5 nM).

Radioligand binding studies. Stably transfected HEK 293 cells were harvested in 50 mM ice-cold Tris·HCl, pH 7.4, and centrifuged for 20 min at 26,000 × g at 4°. Cells were lysed by sonication in 5 mM Tris·HCl, pH 7.4, and recentrifuged (20 min), and the membrane pellet was resuspended in 50 mM Tris·HCl, pH 7.4. To determine the dissociation constant (K_D) of [³H]rauwolscine, 20 µg of protein was incubated with increasing concentrations of [³H]rauwolscine (0.1-45 nM) in the presence of 50 mM Tris·HCl, pH 7.4, in a total volume of 250 µl for 30 min at 25°. Yohimbine (5 µM) was added to determine the nonspecific binding. To determine the affinities of several compounds for Lym oa₁, increasing concentrations of inhibitor (10¹⁰ to 10⁴ M) were used to inhibit the binding of 4 nM [³H]rauwolscine. Receptor binding reactions were terminated by filtration over Whatman (Fairfield, NJ) GF/B filters, which were then washed three times with 50 mM ice-cold Tris·HCl, pH 7.4. The radioactivity retained on the filters was measured by liquid scintillation counting. Kaleidagraph 3.0 (Abelbeck/Synergy, Reading, PA) was used to fit the obtained data. Saturation curves were fitted according to the equation: SB = B_max × [L]/([L] + K_D), in which SB is specific binding (total binding minus nonspecific binding), B_max is the total number of binding sites, and [L] is the concentration of [³H]rauwolscine. The affinity constant (K_D) was determined from Scatchard analysis plotting specific binding versus specific binding/[L]. Displacement curves were fitted according to the equation: B = TB  SB × ([I]/([I] + IC₅₀))ⁿ, in which B is bound radioactivity, TB is total binding, [I] is the concentration of the inhibitor, and n is the Hill coefficient (K_i = IC₅₀/[1 + ([L]/K_D)])

Measurements of IP formation. HEK 293 cells stably expressing Lym oa₁ were grown onto 24-well plates and incubated with 1 µCi of [³H]inositol (Amersham, Arlington Heights, IL)/ml of inositol-free DMEM (GIBCO) supplemented with 10% dialyzed fetal calf serum for 24 hr. Cells were washed once with DMEM and incubated with 10 mM LiCl for 10 min at 37°. Agonists (10³ M to 10⁹ M) were added and incubated for 60 min at 37°. The medium was removed, and the cells were lysed by sonication in ice-cold chloroform/methanol (1:2). After chloroform extraction, the aqueous phase was incubated with Dowex AG 1X8 anion exchange resin. Resin was washed thoroughly with H₂O, and IPs were eluted with 1 ml of 0.1 M formic acid/1 M ammonium formate. Radiolabeled IPs were measured by liquid scintillation counting.

Measurements of cAMP formation. The cellular concentration of cAMP in HEK 293 cells stably expressing Lym oa₁ was determined as described by Leurs et al. (23). Briefly, cells were grown onto 24-well plates and incubated for 20 min at 37° with 300 µM 3-isobutyl-1-methylxanthine and agonists (10³ M to 10⁹ M) in DMEM. After stimulation of the cells, the medium was aspired, and the cells were lysed by sonication in ice-cold 0.1 N HCl. After neutralization, the extract was incubated with protein kinase A and 30,000 dpm of [2,8-³H]cAMP for 2.5 hr on ice. Reactions were terminated by filtration over a Whatman GF/B filter, and filters were washed three times with 50 mM ice-cold Tris·HCl, pH 7.4. The bound radioactivity was compared with a calibration curve in which 0.25-32 pmol of cAMP was incubated with protein kinase A and [2,8-³H]cAMP.

	Results

Summary Introduction Materials & Methods Results Discussion References

Cloning of Lym oa₁ cDNA. PCR with degenerated oligonucleotides recognizing conserved regions in TMs 6 and 7 of bioamine receptors resulted in the isolation of several partial receptor cDNA fragments that are expressed in the CNS of the pond snail L. stagnalis. One of these fragments showed considerable similarity to the vertebrate -adrenergic receptors. A full-length cDNA clone (3416 bp) corresponding to this fragment was isolated from a L. stagnalis CNS cDNA library (Fig. 1). It consists of a leader sequence of 922 bp, an open reading frame of 1914 bp, and a trailer sequence of 580 bp. The predicted protein (638 amino acids) shows the typical seven hydrophobic regions putatively forming the seven TM domains characteristic for G protein-coupled receptors. An eighth hydrophobic region at the amino-terminal part of the protein, as is present in many invertebrate G protein-coupled neurotransmitter receptors, is not present in Lym oa₁.

View larger version (49K): [in this window] [in a new window]   Fig. 2.   Alignment of the deduced amino acid sequence of Lym oa1 with the tyramine receptors cloned from Drosophila and Locusta (Tyr/Oct-Dro and Tyr-Loc) and with the human platelet -adrenergic receptor (h 2). White letters, identical residues; overlined, putative TM regions predicted for Lym oa1 (labeled I-VII).

Stable expression of Lym oa₁ in HEK 293 cells: pharmacological profile. HEK 293 cells were transfected with pcDNA 3-Lym oa₁ and selected for growth on medium containing 0.8 mg/ml G-418. Several resistant colonies were isolated, and membrane preparations of these clonal cell lines were tested for their ability to bind [³H]rauwolscine. One stable cell line (B_max = 3.6 ± 0.2 pmol/mg of protein; five experiments) was selected for further analysis of the pharmacological and signaling properties of Lym oa₁. Fig. 3 shows the saturation binding curve and Scatchard plot for [³H]rauwolscine (K_D = 2.85 nM; five experiments). HEK 293 cells not expressing Lym oa₁ did not show any binding to [³H]rauwolscine (data not shown). To test which naturally occurring agonist could be the ligand for Lym oa₁, several bioamine neurotransmitters were tested for their ability to displace [³H]rauwolscine binding (Table 1). Hill coefficients for all agonist inhibition curves were not significantly different than 1. The rank order of potencies of the tested transmitters was (±)-p-octopamine > serotonin = p-tyramine > ()-epinephrine > ()-norepinephrine > dopamine > histamine. Because of the structural relationship between Lym oa₁ and the adrenergic receptors, we tested several other agonists and antagonists that are known to interact with adrenergic receptors (Table 1). The rank order of potency of those agonists was (±)-p-synephrine  clonidine > xylometazoline  phenylephrine  oxymetazoline > B-HT920 >methoxamine, and the rank order of the antagonists was yohimbine > chlorpromazine  spiperone > phentolamine > mianserine  rauwolscine > prazosin > propanolol  alprenolol > pindolol. This profile suggests that Lym oa₁ is more closely related to the ₂- than to the ₁- and -adrenergic receptors. To determine whether Lym oa₁ exhibits differential homology to one of the ₂-adrenergic receptor subtypes, we performed a similar pharmacological characterization for the human _2A-, _2B-, and _2C-adrenergic receptors (Table 1). No significant correlation could be found between the agonist binding properties of Lym oa₁ and the ₂-adrenergic receptors. For the antagonists, however, Lym oa₁ was more closely correlated to the _2B-adrenergic (r_S = 0.950, p < 0.001) than to the _2A-adrenergic (r_S = 0.820, p = 0.007) or _2C-adrenergic (r_S = 0.683, p < 0.042) receptor.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Saturation binding curve for [3H]rauwolscine binding to the Lym oa1 receptor expressed in HEK 293 cells. Inset, Scatchard plot of the same data; SB, specific binding (total binding minus nonspecific binding as determined with 5.106 M yohimbine binding); F, free ligand.

View this table: [in this window] [in a new window]   TABLE 1 pKi values of various compounds for Lym oa1 and human 2A-, 2B-, and 2C-adrenergic receptors as determined by displacement of [3H]rauwolscine binding Ki values are mean ± standard deviation, as deduced from at least three independent experiments, each of which was performed in triplicate. In the case of Lym oa1, the concentration of [3H]rauwolscine was 4 nM. In the case of the human 2-adrenergic receptors, the concentration was 1 nM.

Stable expression of Lym oa₁ in HEK 293 cells: signal transduction. Activation of octopamine receptors classically results in elevation of the concentration of cAMP in the cell (1). Therefore, we tested the ability of octopamine to stimulate AC in HEK 293 cells stably expressing Lym oa₁. Fig. 4 shows the dose-dependent increase in intracellular cAMP levels induced by octopamine and tyramine. Octopamine increases the intracellular cAMP concentration maximally to 9-fold over basal levels, with an EC₅₀ value of 5.1 ± 0.8 × 10⁶ M. Tyramine could induce only a slight increase in intracellular cAMP levels (2-fold over basal levels) at rather high concentrations (10³ M). Stimulation of untransfected HEK 293 cells with octopamine or tyramine did not show any elevation in cAMP concentration (data not shown). Related neurotransmitters (epinephrine, norepinephrine, dopamine, serotonin) did not stimulate cAMP formation to a significantly higher degree in HEK 293 cells expressing Lym oa₁ than in untransfected cells (data not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Concentration-effect relationship of agonist-induced increase in intracellular concentration of cAMP in HEK 293 cells stably transfected with Lym oa1. Error bars, standard deviation of the mean value obtained from three experiments, each performed in triplicate. Octopamine () and tyramine () maximally stimulated cAMP levels 9- and 2-fold over basal, respectively. Basal levels were 3 ± 1 pmol/well.

7

5

5

5

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Concentration-effect relationship of agonist-induced increase in intracellular concentration of IPs in HEK 293 cells stably transfected with Lym oa1. Error bars, standard deviation of the mean value obtained from three experiments, each performed in triplicate. Stimulation was performed with octopamine (), epinephrine (), tyramine (), and norepinephrine (). Maximal stimulation was 46-, 33-, 35-, and 31-fold over basal, respectively. Maximal stimulation is plotted as 100% for easier comparison of the concentration dependencies of the different agonist-induced effects on intracellular IP concentration. Basal IP levels were 2.3%, 3.3%, 3.3%, and 3.2%, respectively, and correspond to 292 ± 95 dpm of [3H]IPs/well.

	Discussion

Summary Introduction Materials & Methods Results Discussion References

We described the cloning and functional expression of a cDNA that encodes a novel type of G protein-coupled receptor isolated from the CNS of L. stagnalis. The predicted protein sequence exhibits highest identity to the tyramine/octopamine receptor from Drosophila, the tyramine receptor from Locusta, and the mammalian -adrenergic receptors. The extent of this identity (40-50% in TM regions), however, indicates that the L. stagnalis receptor is not a species variant of these receptors. Indeed, on permanent expression of the receptor in HEK 293 cells, it shows higher affinity for octopamine than for tyramine or (nor)epinephrine. Furthermore, octopamine is the most potent agonist in stimulation of IP and cAMP formation. These results strongly suggest that we identified an octopamine receptor, and we propose to call it Lym oa₁.

The residues that are involved in the interaction between catecholamines and their receptors have been studied extensively. An important source of binding energy between the ligand and the receptor is provided by the ionic interaction with the amine of the ligand and an aspartic acid residue in TM 3. However, both agonists and antagonists interact with this aspartic acid residue. The actual receptor activation is thought to involve an interaction between the catechol moiety of the agonist and residues in TMs 6 and 7. More specifically, the aromatic ring interacts with a phenylalanine residue in TM 6, and the catechol hydroxyl groups form hydrogen bonds with side chains of two serine residues in TM 5. In the ₂-adrenergic receptor, these two serine residues (Ser204 and Ser207) were shown to interact with the meta- and para-hydroxyl group of the catechol moiety of norepinephrine, respectively (26). In the _2A-adrenergic receptor, Ser204 was shown to be involved in hydrogen bonding with the para-hydroxyl group (like its cognate residue in the ₂-adrenergic receptor; 27). In contrast, Ser200 did not seem to participate in receptor agonist interaction. In Lym oa₁, the presence of Asp108 in TM 3 and Phe535 in TM 6 is conserved. Interestingly, only one of the two conserved serine residues in TM 5 is present (Ser246). This serine residue corresponds to the serine residues in the - and -adrenergic receptors (Ser207 and Ser204, respectively) that are thought to interact with the para-hydroxyl group of the catechol moiety of norepinephrine. Because octopamine is a mono-, para-hydroxylated catecholamine, the conservation of Ser246 can be considered as indicative for a similar binding of octopamine and norepinephrine to their receptors.

Octopamine is a known neurotransmitter in L. stagnalis. Several identified neurons in the CNS have been shown to contain octopamine (7), and recently, [³H]octopamine binding sites in the CNS were pharmacologically characterized.³ This pharmacological profile, however, differs strongly from the profile of Lym oa₁. In particular, the affinities of phentolamine and yohimbine are much higher for Lym oa₁ than for the [³H]octopamine binding sites in total L. stagnalis CNS. This suggests that in addition to Lym oa₁, other octopamine receptor or receptors will be present in L. stagnalis.

The pharmacological properties of the oct-1, -2A/-2B, and -3 receptor subtypes from insect tissues have been studied extensively (for a review, see Ref. 2). Interestingly, the ligand binding characteristics of Lym oa₁ show no homology with the binding profiles of any of the insect octopamine receptors, suggesting that Lym oa₁ is not a species variant of the previously characterized insect octopamine receptor subtypes. Therefore, we suggest that Lym oa₁ is the first member of a novel subclass of octopamine receptors.

Because of the structural similarity among octopamine, tyramine, and (nor)epinephrine, we compared the ligand binding properties of Lym oa₁ with those of the cloned tyramine and mammalian adrenergic receptors. When the pharmacological profile of Lym oa₁ is compared with the profiles of Tyr/Oct-Dro and Tyr-Loc (18-20), there is a better correlation between Lym oa₁ and Tyr/Oct-Dro (r_S > 0.9, p < 0.001) (12, 13) than between Lym oa₁ and Tyr-Loc (r_S  0.8, p = 0.07) (14). However, phentolamine, chlorpromazine, and mianserine in particular have a markedly higher affinity for Lym oa₁ than for the insect receptors. When the pharmacological profile of Lym oa₁ is compared with that of the human adrenergic receptors, the affinities of both agonists and antagonists for Lym oa₁ indicate that Lym oa₁ is more closely related to the -adrenergic receptors than to the -adrenergic receptors (illustrated by the high affinity of clonidine and the low affinities of alprenolol, propanolol, and pindolol for Lym oa₁). Furthermore, the high affinities of rauwolscine and yohimbine and the moderate affinity of prazosin indicate a closer relationship between Lym oa₁ and the ₂-adrenergic receptors than between Lym oa₁ and the ₁-adrenergic receptors. Therefore, we wanted to know whether Lym oa₁ was most closely related to one of the three -adrenergic receptor subtypes. Such information could be useful in unraveling the structure-function relationships of -adrenergic receptors. Concerning the antagonist binding affinities, a better correlation can be found between Lym oa₁ and the _2B-adrenergic receptor subtype (r_S = 0.950, p < 0.001) than between Lym oa₁ and the _2A-adrenergic receptor subtype (r_S = 0.820, p = 0.007) and the _2C-adrenergic receptor subtype (r_S = 0.683, p < 0.042). However, there is no significant correlation between the agonist binding affinities of Lym oa₁ and the ₂-adrenergic receptors. The pharmacological and structural relationships between Lym oa₁ and the ₂-adrenergic receptors suggest that the corresponding genes have evolved from a common ancestor. The conservation of the functional properties of octopamine receptors and norepinephrine receptors (e.g., their key role in stress adaptation) is in agreement with such an evolutionary connection.

Most octopamine receptors (i.e., the oct-2 and oct-3 receptor subtypes) have been shown to be positively coupled to AC (1, 2). Activation of the less well-characterized oct-1 receptor subtype has been described to result in an elevation of the intracellular concentrations of both calcium and IPs (24, 25). We have shown that octopamine induces an increase in both IPs and cAMP in cells stably expressing Lym oa₁. We did observe, however, that the coupling of Lym oa₁ to AC seemed to be less efficient than the coupling to PLC. This can be concluded from the fact that the maximal production of cAMP (~20 pmol/well, ~9-fold over basal) is markedly lower than that obtained by stimulation of endogenous (G_s coupled) -adrenergic receptors in the same HEK 293 cell-line (~48 pmol/well, ~20-fold over basal; data not shown), although the latter receptors are expressed at much lower densities. In contrast, the octopamine-induced stimulation of IP formation is extremely efficient (45-fold over basal). Furthermore, octopamine is the only agonist that can induce a significant increase in cAMP levels in cells expressing Lym oa₁ compared with wild-type HEK 293 cells. Although the stimulation of AC by the activation of endogenous -adrenergic receptors might mask the effect of (nor)epinephrine on Lym oa₁, this does not hold true for tyramine. On the other hand, all tested agonists elicit a strong IP response to octopamine. Similar situations, in which receptor activation leads to a strong stimulation of phosphatidyl inositol bisphosphate hydrolysis and a weak increase in cAMP levels, have been described for the ₁-adrenergic receptors (28-30), for the m1 and m3 muscarinic acetylcholine receptors (30-36), and for the 5-hydroxytryptamine_2A serotonin receptor (37). Such dual signaling can been explained by (i) coupling of the receptor to two different G proteins; (ii) coupling of the receptor to a single G protein that activates PLC, followed by cross-activation of AC by activated protein kinase C or elevated calcium concentrations; or (iii) coupling of the receptor to a single G protein of which the  subunit stimulates AC and the subunit stimulates PLC. Further experiments are required to find out which of these three possibilities accounts for the dual coupling of Lym oa₁. Furthermore, the presence of secondary signaling pathways can depend on receptor density (28, 38-40) and cell type (29). Therefore, it is worthwhile to investigate whether cell lines expressing lower number of receptors are still capable of dual coupling and whether the coupling is specific for HEK 293 cells.