Keywords: Central pattern generator, Crustacean, G protein coupled receptor, Heterologous expression, Signal transduction, Stomatogastric
 |  |
| ||||
Arthropod D2 receptors positively couple with cAMP through the Gi/o protein family Correspondence to: Deborah J. Baro, Biology Department, Georgia State University, P.O. Box 4010, Atlanta, GA 30303, USA, phone: (404) 651-3107; fax: (404) 651-2509, email: dbaro/at/gsu.edu  The publisher's final edited version of this article is available at Comp Biochem Physiol B Biochem Mol Biol. | |||||||||||||||||||
Abstract The pyloric network is an important model system for understanding neuromodulation of rhythmic motor behaviors like breathing or walking. Dopamine (DA) differentially modulates neurons within the pyloric network. However, while the electrophysiological actions of DA have been well characterized, nothing is known about the signaling events that mediate its effects. We have begun a molecular characterization of DA receptors (DARs) in this invertebrate system. Here, we describe the cloning and characterization of the lobster D2 receptor, D2αPan. We found that when expressed in HEK cells, the D2αPan receptor is activated by DA, but not other monoamines endogenous to the lobster nervous system. This receptor positively couples with cAMP through multiple Gi/o proteins via two discrete pathways: 1) a Gα mediated inhibition of adenylyl cyclase (AC), leading to a decrease in cAMP and 2) a Gβγ mediated activation of Phospholipase Cβ (PLCβ), leading to an increase in cAMP. Alternate splicing alters the potency and efficacy of the receptor, but does not affect monoamine specificity. Finally, we show that arthropod D2 receptor coupling with cAMP varies with the cellular milieu. Keywords: Central pattern generator, Crustacean, G protein coupled receptor, Heterologous expression, Signal transduction, Stomatogastric | |||||||||||||||||||
The crustacean stomatogastric ganglion is extensively used as a model to understand neuromodulatory effects on motor pattern generation (Nusbaum and Beenhakker, 2002;Harris-Warrick and Marder, 1992). A wealth of information exists on the monoaminergic modulation of ion currents and neuronal firing properties (Flamm and Harris-Warrick, 1986;Harris-Warrick et al., 1995b;Harris-Warrick et al., 1995a;Harris-Warrick et al., 1998;Kloppenburg et al., 1999;Johnson et al., 2003), but nothing is known about the transduction cascades mediating these effects. To extend the usefulness of this model system and gain insight into how component neurons integrate biochemical and electrical processes, we have begun a molecular characterization of DARs in this central pattern generator (CPG). In the traditional view, DARs are classified as type-1 or type 2: type-1 DARs couple to Gs proteins, leading to a Gα-mediated increase in [cAMP]i and protein kinase A (PKA) activity, while type-2 DARs couple to Gi/o proteins to decrease [cAMP]i and PKA activity (Missale et al., 1998;Neve et al., 2004). It is now clear that this traditional view of DAR signaling is much too simple. First, DARs have been shown to couple with multiple G proteins in various heterologous and native systems (Kimura et al., 1995a;Sidhu et al., 1998;Zheng et al., 2003;O'Sullivan et al., 2004;Zhen et al., 2004;Kimura et al., 1995b). Moreover, GPCRs, including DARs, can switch G protein coupling over time in response to constant agonist application (Daaka et al., 1997;Baillie et al., 2003;Lezcano et al., 2000). Second, both the Gα and Gβγ subunits are known to mediate individual responses (Cabrera-Vera et al., 2003). Third, activated G protein subunits can directly interact with target proteins such as ion channels without altering second messenger levels (Dascal, 2001;Ivanina et al., 2004). Fourth, GPCRs are known to interact directly with target proteins. For example, DARs can physically interact with, and activate ionotropic glutamate receptors (Zou et al., 2005;Lee and Liu, 2004;Pei et al., 2004;Liu et al., 2000). Fifth, GPCRs can activate additional cascades, like the mitogen activated protein kinase (MAPK) cascade via crosstalk (Werry et al., 2005). Finally, GPCRs can directly activate G protein independent cascades. One important mechanism involves recruitment of β-arrestin scaffolds to an activated receptor and subsequent stimulation of G protein-independent cascades (Lefkowitz and Shenoy, 2005). In this regard, it was recently shown that the D2 receptor modulates locomotor activity in mice via a β arrestin 2-mediated signaling complex involving Akt and PP2A, as well as by traditional G protein cascades (Beaulieu et al., 2005). There are three arthropod DARs: two type-1 receptors and one type-2 receptor (Gotzes et al., 1994;Feng et al., 1996;Han et al., 1996;Hearn et al., 2002;Blenau et al., 1998;Beggs et al., 2005;Mustard et al., 2003). A fourth arthropod receptor that responds to DA with an increase in cAMP has been cloned, but it is primarily activated by ecdysteroids and does not appear to belong to the DAR family (Srivastava et al., 2005). We have previously cloned and characterized the two type-1 DARs from the spiny lobster (Clark and Baro, 2006). Here we describe the cloning and characterization of the lobster D2 receptor, D2αPan. | |||||||||||||||||||
Cloning and expression in a heterologous system The lobster D2αPan cDNA was cloned from nervous tissue of Panulirus interruptus using a degenerate PCR strategy with conventional library screening and RACE technology as previously described (Clark et al., 2004). The D2α.1Pan sequence has been submitted to Genbank under accession number DQ900655 (Figure 1). Full length constructs were created and inserted into a pIRESneo3 vector (B.D. Biosciences Clontech, Palo Alto, CA) using standard recombinant techniques. D2αPan and AmDop3 constructs were stably expressed in HEK293 cells using methods previously described (Clark et al., 2004). AmDop3 was kindly provided by Dr. Allison Mercer, University of Otago. All tissue culture reagents were purchased from Invitrogen except the DMEM and the penicillin streptomycin solution (American Type Culture Collection), and the neomycin (Sigma).
In some experiments, the Gβγ scavengers, dexras1 (UMR cDNA resource center, University of Missouri-Rolla) or βARK495–689 (kindly provided by Dr. Robert Lefkowitz, Howard Hughes Medical Institute), were transiently expressed. In these cases, cells were maintained in DMEM supplemented with 10% dialyzed fetal bovine serum plus 600μg/mL neomycin (HEKD2αPan or HEKAmDop3) or 50 units/mL penicillin and 50μg/mL streptomycin (parental HEK cells) at 37°C, 5% CO2, and were grown to 90–95% confluency in 26x33mm wells of an 8-well plate (Fisher Scientific). One day prior to transfection, the cells received media without antibiotic. Cells were transfected with 2 μg DNA using 10μL lipofectamine in 100uL opti-MEM according to the manufacturer’s instructions. After 6 hours at 37°C, 5% CO2, cells received 1mL of DMEM containing 20% dialyzed serum. Cells received normal media (with antibiotic) 24 hours following transfection, and were assayed 24–48 hours later. The experiments described in this manuscript were conducted over the course of 2 years, during which time the properties of the parental HEK cell line varied. During the first year the parental line was insensitive to DA, even at a concentration of 100mM. The assays shown in Figures 2, 4, 6 and 7 were conducted during this initial period. There was then a long hiatus from experimentation during which time all cell lines were frozen in liquid nitrogen. Experiments were resumed during year 2. Parental HEK and HEKD2α.1Pan cells were thawed and the assays shown in Figures 3 and 5 were performed. In addition, the parental line was also transfected to generate stable HEKAmDop3, HEKD2α.2Pan, and HEKD2α.3Pan lines, and the assays shown in Figures 8 and 9 were performed. At some point during the second year the parental line began to express low and variable levels of an endogenous human D1 receptor that in some assays produced a significant increase in cAMP in response to 10−4M DA or the D1 selective agonist, 6-chloro-PB (n= 3, p< 0.05). The pharmacology of the human D1 receptor was distinct from the D2αPan receptor. The D2αPan receptor produced an increase in cAMP in response to 10-5M quinpirole (n=3, p < 0.05), a selective D2 agonist, while the parental D1 receptor did not (n=3, p> 0.05). Furthermore, the signaling properties of the two receptors were distinct: The arthropod D2 receptor relies on the Gβγ subunit to produce an increase in cAMP while the human D1 receptor does not (Figures 5 and 8).
Membrane preparations Stably transfected cells were harvested with trypsin (ATCC, Manassas, VA). Pellets were homogenized in 20mM HEPES (pH 7.4) containing 2mM MgCl2, 1 mM EDTA, 2mM 1,4-dithiothreitol (DTT), 1μg/mL leupeptin, 1μg/mL aprotinin, and 2mM PMSF. The homogenate was centrifuged at 2500 rpm for 5 minutes. The supernatant was recovered and centrifuged at 15,000 rpm for 30 minutes at 4°C. Pellets were resuspended in 20mM HEPES (pH 7.4) containing 0.5% 3-[(3 cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), and 2mM EDTA. Protein concentrations in each sample were determined using a BCA Protein Assay Kit (Pierce).G protein activation assay Agonist-induced activation of specific G proteins was determined using our previously described G protein activation assay (Clark and Baro, 2006). Briefly, membrane preparations from cell lines (1.5μg/ul of protein) were incubated at 37°C for 15 minutes in 10mM HEPES (pH 7.4) containing 10mM MgCl2, 100μM EDTA and 10nM GTPγ35S (Amersham) with or without DA. Reactions were terminated with ten volumes of termination buffer [10mM MgCl2, 100μM GDP, 200mM NaCl in 100mM Tris (pH 8.0)]. Fifty μl of each terminated sample were then aliquotted in triplicate to wells precoated with one antibody against a human Gα subunit [G12α, Gi1/2α, Gqα, Gi3/oα, Gsα, or Gzα (EMD/Calbiochem)] and to uncoated wells (blanks). Plates were incubated on ice for 2 hours. Wells were then rinsed three times with phosphate-buffered saline containing 0.3% Tween-20. Individual wells were placed in scintillation vials containing ScintiSafe Econo 1 (Fisher) and the radioactivity in each well was quantified with a scintillation counter. Resulting cpm from the blank wells were averaged and used as a measure of non-specific binding. The nonspecific binding was subtracted from the average cpm obtained from the coated wells. Data are expressed as cpm/μg of protein.cAMP assays cAMP levels were measured as previously described (Clark et al., 2004). Briefly, 1 × 105 cells were plated in 35mm dishes and grown to confluence. Cells were washed with 1 mL of media and preincubated at 37°C for 10 minutes in the presence of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (2.5mM) (Sigma). Cells were incubated an additional 30 minutes at 37°C with or without forskolin (2.5μM), and varying concentrations of monoamine (DA, 5-HT, tyramine, histamine, or octopamine). In some experiments, cells were pretreated for 24 hours with pertussis toxin (PTX, Calbiochem) or 15 minutes with 1-O-Octadecyl-2-O-methyl-rac-glycero-3-phosphorylcholine (Et-18-OCH3, Calbiochem). The media was removed and 0.5mL of 0.1M HCl with 0.8% Triton X-100 (Sigma) was added to the plates. After a 30 minute incubation at room temperature, the lysate was removed from the plates and centrifuged for 2 minutes. The supernatant was collected and assayed for cAMP levels using a direct cAMP enzyme immunoassay kit (Assay Designs, Inc.) according to the manufacturer’s instructions. Protein concentrations in each sample were determined using a BCA Protein Assay Kit (Pierce).Statistical analyses and curve fitting Student t-tests were performed with Excel software. Curve fitting, Kruskall-Wallis (ANOVA on ranks) tests, and Bonferroni posttests were performed with Prism (GraphPad Software, San Diego, CA, www.graphpad.com). In all cases, statistical significance was determined as p < 0.05. | |||||||||||||||||||
DARs are conserved across species Using total lobster nervous system cDNA and a combination of conventional library screening and RACE technologies, we cloned a type 2 DAR from the spiny lobster, Panulirus interruptus. We found that this receptor, D2αPan, is alternately spliced (see arrowheads in Figure 1) to produce four distinct proteins: D2α.1Pan, D2α.2Pan, D2α.3Pan, and D2α.4Pan (Table 1). We did not conduct an exhaustive search for D2αPan isoforms, and it is likely that additional splice forms exist (Hearn et al., 2002).
Τhe D2αPan receptor is orthologous to the Drosophila receptor DD2R (Hearn et al., 2002) and the Apis receptor AmDop3 (Beggs et al., 2005). A BLAST against the Homo sapien Reference Proteins database showed that the D2αPan receptor was most homologous to the long form of the human D2 receptor (NP_000786) with an E value of 3e-58. Figure 1 illustrates that all D2 receptors are well conserved across species. When compared to its fly, honeybee and human homologs, the D2αPan receptor shows 45%, 39% and 37% amino acid identity over the entire protein, respectively. As expected, the 7 transmembrane regions are among the most conserved portions of the protein. In addition, the cytoplasmic domains known to interact with G proteins show a fairly high degree of identity, including intracellular loops 1 and 2, the amino and carboxy portions of intracellular loop 3, and the cytoplasmic C-terminal domain (Limbird, 2004;Cabrera-Vera et al., 2003). Most amino acid substitutions in these regions are conservative (Figure 1). D2α.1Pan couples with multiple members of the Gi/o family in HEK cells As stated above, the G protein interaction domains are well conserved between arthropodal and mammalian D2 receptors. Similarly, the G protein domains that interact with receptors are conserved across species (reviewed in Cabrera-Vera et al., 2003). The last five residues of the Gα C-terminus is an important mediator of receptor-G protein interactions. This domain shows 100% amino acid identity between human and arthropod Gs, Gi, Go and Gq homologs (reviewed in Clark and Baro, 2006). While receptor-G protein interactions are not mediated solely by this structural feature, the extreme conservation suggests that the mechanisms for receptor-G protein interactions will be similar in mammals and arthropods. This predicts that arthropodal GPCRs will activate the same G protein(s) as homologous mammalian receptors when expressed in mammalian cell lines. Mammalian type-2 DARs stimulate both PTX sensitive (Gαo, Gαi1, Gαi2, Gαi3) and insensitive (Gαz) members of the Gαi/o family (Obadiah et al., 1999;Banihashemi and Albert, 2002;Ghahremani et al., 1999). We stably expressed the D2α.1Pan construct in a HEK cell line (HEKD2α.1Pan) and performed our previously described G protein activation assay (Clark and Baro, 2006) to determine receptor-G protein coupling. Figure 2 illustrates that D2α.1Pan couples with both PTX-sensitive and PTX-insensitive members of the Gαi/o family, but not with Gαs, Gαq, or Gα12. The receptor appeared to couple most strongly with Gαi3/o, producing a significant 2.4-fold increase in activity in response to a 15-minute application of 10μM DA (p < 0.007). There was also significant coupling with Gαz (1.8-fold increase in activity, p < 0.02) and Gαi1/2 (1.4-fold increase in activity, p < 0.02).DA activates D2α.1Pan to produce an increase in cAMP D2αPan orthologs have been shown to respond to multiple monoamines. In addition to DA, tyramine stimulates the DD2R and AmDop3 receptors, and DD2R responds to serotonin (5-HT) (Hearn et al., 2002;Beggs et al., 2005). Gαi/o proteins are known to decrease AC activity and reduce [cAMP]i. We therefore further characterized the D2α.1Pan receptor by measuring [cAMP]i in cells after a 30 minute exposure to one of five monoamines that are endogenous to lobster nervous tissue. Figure 3 illustrates that at a concentration of 1mM, 5-HT, octopamine, tyramine and histamine had no significant effect on [cAMP]i in HEKD2α.1Pan relative to parental cells. On the other hand, DA produced a significant 2.3-fold increase in [cAMP]i in HEKD2α.1Pan relative to parental cells (p < 3 x 10−4). Collectively, these data suggest that DA is the only endogenous monoamine that activates the D2α.1Pan receptor when expressed in HEK cells.The DA-induced increase in [cAMP]i is mediated by Gi/o proteins Interestingly, DA produced an increase in cAMP levels (Figure 3), despite the fact that the D2α.1Pan receptor couples with the Gi/o family (Figure 2). This is contrary to previous studies on the fly and bee orthologs of D2αPan, which show that when these receptors are expressed in HEK cell lines, exposure to DA produces a decrease in cAMP (Beggs et al., 2005;Hearn et al., 2002). It is not clear whether the difference lies in the cell lines or the arthropod D2 receptors. In order to elucidate the mechanism responsible for this difference, we further characterized the lobster D2 signaling cascade(s) in HEK cells.DARs can signal through mechanisms independent of the traditional, G protein mediated pathways (Beaulieu et al., 2005;Zou et al., 2005;Lee and Liu, 2004;Pei et al., 2004;Liu et al., 2000;Lefkowitz and Shenoy, 2005). To determine if the DA-induced increase in cAMP is due to Gi/o proteins (Figure 2) and/or G protein independent cascades, we examined the effect of DA in the presence of PTX, which specifically blocks the activation and dissociation of all members of the Gi/o family, except Gz. We hypothesized that if PTX can partially block the DA-induced increase in cAMP, it would suggest that Gi/o proteins help to mediate the response. Figure 4 (solid line) shows that HEKD2α.1Pan cells produced a dose-dependent increase in cAMP levels in response to DA, with an EC50 of 9.2 x 10−7M, while the parental HEK cell line did not respond to DA. The dashed line illustrates that application of PTX significantly attenuated the DA response, and reduced the maximal fold change in cAMP from 4.4 to 3 (p < 0.0008). On the other hand, PTX had no effect on cAMP levels in the parental HEK cell line. These data, along with Figure 2, suggest that the DA-evoked increase in cAMP depends, in part, on PTX sensitive trimeric Gi/o proteins. The DA response was not completely eliminated despite the fact that a saturating concentration of PTX was applied (500 ng/mL; see Figure 4 inset for PTX dose response curve). This is at least partially due to coupling between the D2α.1Pan receptor and the PTX insensitive Gz protein (Figure 2); however, we cannot rule out the possibility that the D2αPan receptor activates additional G protein independent cascades to increase [cAMP]i. The Gβγ subunits of Gi/o proteins contribute to DA-induced alterations in [cAMP]i AC can be regulated by both Gα and Gβγ subunits (Federman et al., 1992). While Gαi/o subunits decrease or have no effect on AC activity, Gβγ can increase or decrease AC activity depending on the AC and Gβγ isozymes involved (Cabrera-Vera et al., 2003). We tested the hypothesis that Gβγ subunits mediate the DA-induced increase in [cAMP]i in the HEKD2α.1Pan cell line by blocking the Gβγ pathway with known Gβγ scavengers. Dexras1 has been shown to specifically block agonist-stimulated GPCR activation of Gβγ signaling (Nguyen and Watts, 2005). Similarly, the carboxyl-terminal domain of βARK1 and βARK2 (βARK495–689) suppresses Gβγ-mediated responses by scavenging free βγ subunits (Koch et al., 1994). We transiently expressed βARK495–689 or dexras1 in HEK and HEKD2α.1Pan cell lines and measured [cAMP]i in the presence of 10−4M DA. Figure 5 shows that expression of either βARK495–689 or dexras1 significantly inhibited the DA-induced increase in cAMP in HEKD2α.1Pan cells by 64 ± 9% (p < 0.04) and 51 ± 3% (p < 0.008) respectively, but had no significant effect on the parental cell line (p > 0.7). These data suggest that Gβγ subunits contribute to the DA-induced changes in [cAMP] in HEKD2αPan cells.Blocking the Gβγ cascade reveals a DA-induced, Gαi/o mediated decrease in cAMP Gβγ can have many immediate effectors, including PLCβ, ACs, ion channels, kinases and components of the synaptic vesicle release machinery (Blackmer et al., 2005;Cabrera-Vera et al., 2003;Stehno-Bittel et al., 1995;Sullivan, 2005;Gerachshenko et al., 2005). It has been previously demonstrated that D2 receptors can regulate ACII activity via Gβγ mediated activation of PLCβ (Tsu and Wong, 1996). To determine whether Gβγ subunits act via PLCβ in HEKD2αPan cells, we applied the PLCβ inhibitor, ET-18-OCH3, and measured cAMP levels in the presence of increasing concentrations of DA. Figure 6 (dashed line) demonstrates that inhibiting PLCβ also inhibited the DA induced increase in cAMP, and revealed a dose-dependent decrease in cAMP.The dose-dependent decrease in cAMP was largely PTX sensitive. Figure 7 illustrates that in the presence of forskolin (an AC activator) and ET-18-OCH3, DA evokes a clear dose dependent decrease in cAMP in HEKD2α.1Pan cells with an EC50 of 1.4 x 10−7M. The total inhibition by saturating levels of PTX was 77% of the maximal response. The response that remained in the presence of PTX was most likely mediated by Gz (Fig. 2). Collectively, these data suggest that DA initiates parallel signaling cascades in HEKD2α.1Pan cells with opposing effects on cAMP levels: the Gαi/o subunits cause a decrease in cAMP while Gβγ subunits activate PLCβ to cause an increase in cAMP. Figure 7 also suggests that the D2α.1Pan receptor, like its mammalian and Drosophila homologs, may constitutively activate Gi/o proteins. The forskolin activated cAMP levels are significantly higher in HEKD2α.1Pan relative to the parental HEK cell line (Figure 7; 150pmol/mg vs. 60pmol/mg; p < 10−4). This compensatory mechanism is known as heterologous sensitivity or supersensitivity (Watts, 2002;Vortherms et al., 2004;Watts and Neve, 2005). Several studies have demonstrated that chronic Gi/o activity ultimately leads to a paradoxical increase in AC activity (supersensitivity) through a number of different molecular mechanisms. The intracellular milieu determines whether arthropod D2 receptors positively or negatively couple with cAMP As previously stated, when expressed in HEK293 cells, the lobster versus fruit fly and honeybee orthologs of the D2 receptor produce opposite changes in cAMP levels in response to DA: the lobster D2 receptor positively couples with cAMP (Figure 3) while the fly and bee orthologs of the D2 receptor negatively couple with cAMP (Beggs et al., 2005;Hearn et al., 2002). We predicted that if the cellular background determines whether D2 receptors positively or negatively couple with cAMP, then expressing the honeybee ortholog of the D2 receptor, AmDop3, in our parental HEK293 cell line should produce an increase, rather than the previously described decrease in cAMP. To test the hypothesis we obtained the AmDop3 clone from the Mercer lab, transformed our HEK293 cells to generate a stable cell line, HEKAmDop3, and measured changes in cAMP in response to DA. Figure 8 shows that the HEKAmDop3 cells responded to 10−4M DA with a significant increase in cAMP (p < 0.0001). The DA-induced increase in HEKAmDop3 cells is roughly 7-fold greater than that observed in the parental HEK line (p < 0.009, HEKAmDop3 vs. HEK). The increase was attenuated by transiently expressing the Gβγ scavenger βARK495–689. Together these data suggest that the cellular milieu greatly influences D2 mediated changes in [cAMP] and that there are no obvious functional differences between the signaling properties of the honeybee and lobster D2 orthologs.Alternate splicing changes the potency and efficacy of D2αPan isoforms Figure 1 and Table 1 indicate that the D2αPan receptor can be alternately spliced to create multiple isoforms with differences in their carboxy termini and/or intracellular loop 3. The carboxy terminus and intracellular loop 3 are involved in G protein coupling (Wong, 2003). Changes in amino acid sequence in these regions can alter the strength or specificity of G protein signaling (Franke et al., 1990;Cotecchia et al., 1990). The C-terminus of GPCRs also determines the rate of receptor recycling and receptor coupling to β-arrestin mediated cascades (Oakley et al., 1999). To determine whether alternate splicing produces functional differences in G protein signaling, we established stable HEK cell lines expressing D2α.2Pan or D2α.3Pan and obtained DA dose-response curves for the resulting cell lines: HEKD2α.2Pan and HEKD2α.3Pan. Figure 9A shows that in all cases the receptor produces a dose-dependent increase in [cAMP]i that is significantly higher than in the parental HEK cell line (2.5-fold and 2-fold greater, respectively, at 10−4M DA; p < 0.05). Altering the carboxy terminal domain (D2α.1Pan vs. D2α.2Pan) reduced the EC50 by more than an order of magnitude (from 9.2 x 10−7M to 2.4 x 10−6M, respectively). In addition, removing the alternately spliced intracellular loop 3 exon (D2α.2Pan vs. D2α.3Pan) once again changed the EC50 by more than an order of magnitude (from 2.4 x 10−6 to 4 x 10−7, respectively) and significantly altered receptor efficacy. Thus, these data suggest that alternate splicing may change the potency of the D2αPan receptor. We also tested the effect of 5HT, octopamine, tyramine, and histamine on [cAMP]i in HEKD2α.2Pan and HEKD2α.3Pan cell lines. Figure 9B shows that, like the D2α.1Pan receptor (Figure 3), D2α.2Pan and D2α.3Pan respond only to DA when expressed in HEK cells. Thus, alternate splicing does not affect the monoamine specificity of these receptors. | |||||||||||||||||||
CPGs are highly modulated neural circuits that rely on GPCRs to produce a rhythmic output (Ramirez et al., 2004;Marder and Bucher, 2001). The effects of DA on a model CPG, the pyloric network, have been extremely well characterized (Harris-Warrick et al., 1998;Gruhn et al., 2005;Johnson et al., 2003;Kloppenburg et al., 2000;Kloppenburg et al., 1999;Peck et al., 2001); however, the molecular mechanisms by which DA exerts its effects are completely unknown. To begin to investigate the molecular underpinnings of the dopaminergic response in pyloric neurons, we cloned and characterized the only known arthropod type-2 DAR from Panulirus interruptus: D2αPan. Heterologous expression in HEK cells indicates that this receptor is specifically activated by DA, as opposed to other monoamines known to be endogenous to the lobster nervous system. Alternate splicing in intracellular loop 3 and at the carboxy terminus alters the potency and efficacy of the receptor. Surprisingly, we found that when expressed in HEK cells the D2αPan receptor positively couples with cAMP. The increase in cAMP is mediated, in part, by multiple Gi/o proteins. D2αPan stimulation of Gi/o activity results in the activation of two discrete pathways: 1) a Gα mediated inhibition of AC, leading to a decrease in cAMP and 2) a Gβγ mediated activation of PLCβ, leading to an increase in cAMP. We also found that contradictory to previous reports (Beggs et al., 2005), the honeybee D2 receptor can positively couple with cAMP via the Gβγ subunits of Gi/o proteins, suggesting that the intracellular environment can alter receptor coupling to cAMP. We conclude that arthropod and mammalian D2 receptor signaling is very similar, and that D2 mediated signaling is determined by both the functional properties of the receptor and the intracellular milieu. The D2αPan receptor simultaneously activates multiple cascades It is not clear whether the D2αPan receptor response is mediated entirely by G proteins in HEK cells. Figure 7 suggests that a PTX insensitive protein, probably Gαz, mediates roughly 23% of the DA induced decrease in cAMP while the PTX sensitive Gαi/o subunits are responsible for 77% of the response. However, saturating levels of PTX only reduced the DA induced increase in cAMP from 4.4- to 3-fold, rather than the predicted 1.8-fold (Figure 4). Furthermore, the EC50 for the increase in cAMP (9.2 x 10−7; Figure 4) is 6.6-fold lower than the EC50 for the decrease in cAMP (1.4 x 10−7; Figure 7). There are at least two possible explanations for these findings, and they are not mutually exclusive. First, D2αPan receptors may simultaneously activate multiple cascades, including G protein independent cascades (Beaulieu et al., 2005;Lefkowitz and Shenoy, 2005). Second, Gαz may donate the majority of Gβγ subunits that interact with PLCβ to increase cAMP. Specific Gα donors for Gβγ subunits have previously been observed in certain cell types. For example, GIRK channels are activated by Gβγ subunits that are exclusively donated by Gαi2 and Gαi3 in native tissues, though any Gα subunit can donate the Gβγ subunits in studies utilizing heterologous expression systems (Dascal, 2001). Specificity in native tissues appears to be conferred by binding of the α-subunit to the GIRK effector (Ivanina et al., 2004) and the fact that upon activation, Gi and Gz proteins undergo a conformational change, but do not dissociate into physically independent Gα and Gβγ subunits (Frank et al., 2005). Although Gα donor specificity has never to our knowledge been observed for Gβγ regulation of PLCβ, we cannot dismiss this concept a priori.Unexpectedly, the D2αPan receptor-initiated cascades regulate cAMP in opposing directions in the same cells. These cascades may be highly localized to create microdomains of cAMP gradients (Zaccolo and Pozzan, 2002;Rich et al., 2001). On the other hand, the cascades may function with different kinetics and interact to generate feedback loops. In addition, there are examples of G protein mediated cascades dominating the early portion of a response to constant agonist application, while β-arrestin cascades predominate in later portion (Ahn et al., 2004). Thus, distinct D2αPan mediated cascades may operate in different timeframes to generate multiphasic responses. Receptor signaling varies with the intracellular milieu Interestingly, a D2 receptor can produce opposite responses even when expressed in the "same" cell type. When AmDop3, the honeybee ortholog of the arthropod D2 receptor, is expressed in HEK293 cells in the Mercer lab, it produces a decrease in cAMP; however, when it is expressed in HEK293 cells in the Baro lab, it produces an increase in cAMP. Such a finding is not unique to the arthropod D2 receptor. For example, isoproterenol induced β2-adrenergic receptor signaling in HEK293 cells varies across labs (Daaka et al., 1997;Friedman et al., 2002;Lefkowitz et al., 2002). Tissue culture cell lines can often rearrange their genetic material and/or alter their genetic programs, most likely because culture conditions provide little selective pressure for maintaining a constant genome/transcriptome/proteome. Thus, receptor signaling in a given cell type may vary with the lab because cell lines diverge within and across labs over time. Indeed, in our hands the parental HEK cell line could alter its response to DA, despite the fact that it was cultured under constant conditions. Differences in the expression, localization and/or interactions of downstream effectors of the D2 receptor could account for the differences in the AmDop3 response in each HEK cell line. All of these findings reinforce the idea that GPCR signaling is context dependent. Based on these studies, we cannot predict how the D2αPan receptor will affect cAMP levels in pyloric neurons; though the data suggest that the D2αPan receptor will most likely couple with Gi and Go proteins to alter cAMP levels in pyloric neurons.Monoaminergic GPCR signaling is conserved across species Relatively little is known about invertebrate monoaminergic GPCRs compared to their vertebrate homologs. Data mining studies suggest that there are roughly 19 monoamine receptors in arthropods (Clark et al., 2004;Roeder, 2003). By the year 2004, 10 of these receptors had been cloned and characterized (Tierney, 2001;Blenau and Baumann, 2001;Clark et al., 2004). Several recent efforts have reduced the number of uncharacterized monoaminergic receptors to roughly 3 out of 19 (Balfanz et al., 2005;Srivastava et al., 2005;Maqueira et al., 2005;Cazzamali et al., 2005;Evans and Maqueira, 2005).Both receptors and G proteins show strong amino acid sequence conservation in functional domains across species. Here we have demonstrated that the arthropod D2 receptor can regulate second messenger levels by coupling to both AC, via Gα subunits, and PLCβ, via Gβγ subunits of Gi/o proteins. Similar findings were previously published for mammalian D2 receptors expressed in HEK293 cells (Tsu and Wong, 1996) and in native neurons (Hernandez-Lopez et al., 2000). In addition, we have previously shown that comparable to mammalian type-1 DARs, the D1αPan receptor couples with Gs and the D1βPan receptor couples with both Gs and Gz when expressed in HEK cells (Clark and Baro, 2006). Likewise, we have shown that the 5-HT2βPan receptor couples with Gq (Clark et al., 2004) and the 5-HT1αPan receptor couples with Gi/o (Spitzer and Baro, submitted), as is the case for their respective mammalian homologs. Collectively, these data strongly suggest that signaling mechanisms for homologous receptors are well conserved across species. | |||||||||||||||||||
Receptor expression studies in heterologous systems are important as they help to define key structure/function relationships for homologous receptors across species. Such studies are also useful and necessary in that they reveal organizing principles for signal transduction and more specifically, the repertoire of cascades available to a given receptor. However, heterologous expression studies are limited by the fact that receptor signaling is context dependent. In order to understand the function of a receptor in a specific cell type, the receptor must ultimately be studied in that cell type. We have found that when the D2αPan receptor is heterologously expressed, it couples with Gi/o proteins and can modulate cAMP levels through both Gα and Gβγ subunits, like all of its homologs. The data also suggest that D2αPan receptor signaling may involve additional Gi/o-independent mechanisms. These results set the stage for future studies aimed at understanding the role of D2 receptors in native neurons involved in rhythmic motor pattern generation. | |||||||||||||||||||
We thank Nadja Spitzer and Elizabeth Prince for useful comments on the manuscript. We are grateful to Dr. Allison Mercer, University of Otago, for providing the AmDop3 construct and to Dr. Robert Lefkowitz, Howard Hughes Medical Institute, for providing the βARK495–689. This work was supported, in part, by NIH NS38770 to DJB. MCC is a fellow of the Molecular Basis of Disease Program at Georgia State University. | |||||||||||||||||||
Abbreviations | |||||||||||||||||||
Footnotes Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. | |||||||||||||||||||
References
| |||||||||||||||||||
