Marked changes in signal transduction upon heteromerization of dopamine D₁ and histamine H₃ receptors

Abstract

Background and purpose:

Functional interactions between the G protein-coupled dopamine D₁ and histamine H₃ receptors have been described in the brain. In the present study we investigated the existence of D₁–H₃ receptor heteromers and their biochemical characteristics.

Experimental approach:

D₁–H₃ receptor heteromerization was studied in mammalian transfected cells with Bioluminescence Resonance Energy Transfer and binding assays. Furthermore, signalling through mitogen-activated protein kinase (MAPK) and adenylyl cyclase pathways was studied in co-transfected cells and compared with cells transfected with either D₁ or H₃ receptors.

Key results:

Bioluminescence Resonance Energy Transfer and binding assays confirmed that D₁ and H₃ receptors can heteromerize. Activation of histamine H₃ receptors did not lead to signalling towards the MAPK pathway unless dopamine D₁ receptors were co-expressed. Also, dopamine D₁ receptors, usually coupled to G_s proteins and leading to increases in cAMP, did not couple to G_s but to G_i in co-transfected cells. Furthermore, signalling via each receptor was blocked not only by a selective antagonist but also by an antagonist of the partner receptor.

Conclusions and implications:

D₁–H₃ receptor heteromers constitute unique devices that can direct dopaminergic and histaminergic signalling towards the MAPK pathway in a G_s-independent and G_i-dependent manner. An antagonist of one of the receptor units in the D₁–H₃ receptor heteromer can induce conformational changes in the other receptor unit and block specific signals originating in the heteromer. This gives rise to unsuspected therapeutic potentials for G protein-coupled receptor antagonists.

Introduction

Although with some initial resistance from the scientific community, the existence of neurotransmitter receptor heteromers is becoming accepted. Neurotransmitter receptors cannot only be considered as single functional units, but as forming part of multimolecular aggregates localized in the plane of the plasma membrane, which can contain other interacting proteins, including receptors for the same or other neurotransmitters (Agnati et al., 2003; 2005; Franco et al., 2003; Bockaert et al., 2004). The functional significance of receptor heteromers is however just beginning to be understood. It is becoming clear that heteromerization of neurotransmitter receptors leads to functional entities that possess different biochemical characteristics with respect to the individual components of the heteromer. Thus, the quantitative or qualitative aspects of the signalling generated by stimulation of either receptor unit in the heteromer are different from those obtained during co-activation (Ferréet al., 2007; 2009; Franco et al., 2007; Rashid et al., 2007).

The striatum is the main input structure of the basal ganglia, which are subcortical structures involved in the processing of information related with the performance and learning of complex motor acts. GABAergic striatal efferent neurons constitute more than 95% of the striatal neuronal population (Gerfen, 2004). There are two subtypes of GABAergic striatal efferent neurons: GABAergic dynorphinergic neurons, which express the peptide dynorphin and dopamine D₁ receptors, and GABAergic enkephalinergic neurons, which express the peptide enkephalin and dopamine D₂ receptors (Gerfen, 2004). Histamine is an important neuromodulator of striatal function, and the striatum contains one of the highest densities of histamine H₃ receptors in the brain (Pollard et al., 1993; Anichtchik et al., 2001; Brown et al., 2001). Both D₁ receptors and H₃ receptors are co-expressed in striatal GABAergic dynorphinergic neurons (Ryu et al., 1994; Pillot et al., 2002), where they have been reported to establish functional interactions (Arias-Montano et al., 2001; Sanchez-Lemus and Arias-Montano, 2004). In the present study we show that heteromerization of dopamine D₁ receptors and histamine H₃ receptors, produces dramatic changes in G protein coupling and signalling in human cell lines. Furthermore, both D₁ receptor and H₃ receptor antagonists could block the heteromer-mediated signalling, a fact that highlights new possibilities for G protein-coupled receptor (GPCR) pharmacology.

Methods

Expression vectors

A plasmid encoding the cDNA of the human H₃ receptor was provided by Johnson & Johnson Pharmaceutical Research & Development, L.L.C. (San Diego, CA, USA). The H₃ receptor cDNA without its stop codon was amplified by using sense and antisense primers harbouring a unique EcoRI site. The fragment was then subcloned to be in-frame with enhanced yellow variant of green fluorescent protein (EYFP) into the EcoRI site of pEYFP-N1 (Clontech, Heidelberg, Germany) to provide the plasmid H₃ receptor–YFP, which expresses EYFP on the C-terminal ends of the receptor. The human cDNAs for cannabinoid CB₁ receptors, 5HT_2B receptors or D₁ receptors cloned in pcDNA3.1 were amplified without their stop codons using sense and antisense primers harbouring unique BamHI and EcoRI to clone D₁ receptors and CB₁ receptors in EYFP vector or to clone 5HT_2B receptors or D₁ receptors in a Renilla luciferase-expressing vector (pcDNA3.1-RLuc). A pcDEF3 plasmid encoding the human cDNA of the H₄ receptor fused to EYFP was also used as negative control. The cDNA for the human D₁ receptor was also subcloned into BamHI and ApaI restriction sites of the pcDNA3.1/Hygro (Invitrogen, Grand Island, NY, USA) for the cell line stably expressing D₁ receptors and H₃ receptors. All constructs were verified by nucleotide sequencing. Nomenclature for receptors conforms to the BJP's Guide to Receptors and Channels (Alexander et al., 2008)

Cell culture and transfection

Human embryonic kidney (HEK)-293 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS), 100 units·mL⁻¹ penicillin, 100 µg·mL⁻¹ streptomycin, 2 mmol·L⁻¹ L-glutamine and 100 µg·mL⁻¹ sodium pyruvate (all from Invitrogen), at 37°C in a humidified atmosphere of 5% CO₂ For Bioluminescence Resonance Energy Transfer (BRET) experiments cells were seeded in 35 mm diameter wells of 6-well plates, and transient transfection with the corresponding fusion protein cDNAs was performed the following day by using the calcium phosphate precipitation method (Jordan et al., 1996). Cells were harvested for 48 h after transfection and used for BRET experiments. The empty vector pcDNA3.1 was used to equilibrate the total amount of transfected DNA. For extracellular signal-regulated kinase (ERK) experiments, HEK-293 cells were grown to 80% confluence and transfected by using linear polyethylenimine, MW 25 000 (PEI, Polysciences, Eppelheim, Germany) with 5 µg of cDNA corresponding to human H₃ receptors or human D₁ receptors or both cDNAs at the same time. The empty vector pcDNA3.1 was used to equilibrate the total amount of transfected DNA. Briefly, the plasmid DNA was diluted in 50 µL of medium containing no additives (serum, antibiotics or other protein), and PEI was added (ratio µg DNA : µg PEI, 1:7.5) and incubated for 8 min at room temperature. Medium with 10% FBS was added to the DNA/PEI complex, and the mixture was applied to the cultures. After 2 h incubation, the mixture was replaced for grown medium.

SK-N-MC cells were grown in Eagle's minimal essential medium, supplemented with 10% FBS, 50 units·mL⁻¹ penicillin, 50 µg·mL⁻¹ streptomycin, non-essential amino acids, 2 mmol·L⁻¹ L-glutamine and 50 µg·mL⁻¹ sodium pyruvate at 37°C in a humidified atmosphere of 5% CO₂ to 80% confluence. Cells were transiently transfected with 5 µg of cDNA corresponding to human D₁ receptors (SK-N-MC/D₁) using Lipofectamine™ 2000 (Invitrogen), according to the manufacturer's protocol. To obtain the SK-N-MC cells stably expressing human H₃ receptors and human D₁ receptors (SK-N-MC/D₁H₃), the SK-N-MC cells stably expressing the human H₃R (SK-N-MC/H₃) (provided by Johnson & Johnson Pharmaceutical Research & Development, L.L.C.) were grown to 30–40% confluence in 60 cm² dishes in presence of 600 µg·mL⁻¹ G418 (Invitrogen) and transfected with the cDNA corresponding to human D₁ receptors using Lipofectamine™ 2000. SK-N-MC/D₁H₃ receptor cells were allowed to recover for 24 h before the addition of G418 and 300 µg·mL⁻¹ hygromycin B (Invitrogen), and the colonies that survived selection were grown and tested by binding experiments and Western blotting.

Immunostaining

For immunocytochemistry, HEK-293 cells were grown on glass coverslips and transiently transfected with 0.1 µg of D₁ receptor–RLuc and 0.1 µg H₃ receptor–YFP constructs. After 48 h the cells were fixed in 4% paraformaldehyde for 15 min and washed with phosphate-buffered saline containing 20 mmol·L⁻¹ glycine (buffer A) to quench the aldehyde groups. Then, after permeabilization with buffer A containing 0.05% Triton X-100 for 15 min, cells were treated with phosphate-buffered saline containing 1% bovine serum albumin. After 1 h at room temperature, cells expressing D₁ receptor–RLuc were labelled with the primary rat monoclonal anti-D₁ receptor antibody (1:200, Sigma, St. Louis, MO, USA) for 1 h, washed and stained with the secondary antibody Alexa Fluor®350 Goat anti-rat (1:1000, Invitrogen). The H₃ receptor–YFP construct was detected by its fluorescence properties. Samples were rinsed and observed in a Leica SP5 confocal microscope (Leica Microsystems, Mannheim, Germany).

Bioluminescence Resonance Energy Transfer (BRET)

HEK-293 cells were transfected with 250 ng·well⁻¹ of the cDNA construct coding for D₁ receptor–RLuc, acting as BRET donor, and increasing amounts (0.5–9 µg·well⁻¹) of the cDNA construct coding for the BRET acceptor H₃ receptor–YFP or the negative control H₄ receptor–YFP. After 48 h of transfection cells were washed twice with Hanks' balanced salt solution HBSS (137 mmol·L⁻¹ NaCl, 5 mmol·L⁻¹ KCl, 0.34 mmol·L⁻¹ Na₂HPO₄.12H₂O, 0.44 mmol·L⁻¹ KH₂PO₄, 1.26 mmol·L⁻¹ CaCl₂.2H₂O, 0.4 mmol·L⁻¹ MgSO₄.7H₂O, 0.5 mmol·L⁻¹ MgCl₂, 10 mmol·L⁻¹ HEPES, pH 7.4) supplemented with 0.1% glucose (w·v⁻¹), detached by gently pipetting and resuspended in the same buffer. Sample protein concentration was determined to control cell number, using a Bradford assay kit (Bio-Rad, Munich, Germany) using bovine serum albumin dilutions as standards. Cell suspension (20 µg of protein) was dispensed in duplicates into 96-well black microplates with a transparent bottom (Porvair, King's Lynn, UK), and the fluorescence was measured using a Mithras LB940 fluorescence-luminiscence detector (Berthold, Bad Wildbad, Germany) with an excitation filter of 485 nm and an emission filter of 535 nm. For BRET measurement, 20 µg of cell suspension were distributed in duplicates into 96-well white opaque microplates (Porvair), and coelenterazine H (Molecular Probes Europe, Leiden, The Netherlands) was added at a final concentration of 5 µmol·L⁻¹. After 1 min the readings were collected by using sequential integration of signals detected at 440–500 nm and 510–590 nm. The same samples were incubated for 10 min, and the luminescence was measured. Cells expressing BRET donors alone were used to determine background. The BRET ratio is defined as [(emission at 510–590)–(emission at 440–500)]–Cf where Cf corresponds to (emission at 510–590)–(emission at 440–500) for the D₁ receptor–RLuc construct expressed alone in the same experiment. Curves were fitted by using a non-linear regression equation, assuming a single phase with GraphPad Prism software (San Diego, CA, USA).

Membrane preparation and protein determination

SK-N-MC/D₁H₃ receptor or transfected HEK-293 cells were harvested by centrifugation at 1500×g for 5 min. Cell pellet was washed twice with phosphate-buffered saline and resuspended in 10 volumes of 50 mmol·L⁻¹ Tris-HCl buffer, pH 7.4. Cell suspensions were disrupted with a Polytron homogenizer (PTA 20 TS rotor, setting 3; Kinematica, Basel, Switzerland) for three 5 s periods, and membranes were obtained by centrifugation at 105 000×g (40 min, 4°C). The pellet was resuspended and centrifuged under the same conditions, stored at −80°C until use. Membranes were washed once more as described above and resuspended in 50 mmol·L⁻¹ Tris-HCl buffer for immediate use. Protein was quantified by the bicinchoninic acid method (Pierce Chemical Co., Rockford, IL, USA) using bovine serum albumin dilutions as standard.

Radioligand binding experiments

Membrane suspensions (0.3 mg of protein per millilitre) were incubated for 1 h at 25°C in 50 mmol·L⁻¹ Tris-HCl buffer, pH 7.4, containing 10 mmol·L⁻¹ MgCl₂ with the indicated radioligand in the presence or absence of competing ligands. To obtain competition curves, membranes were incubated with 2.2 nmol·L⁻¹ of the D₁ receptor antagonist [³H]SCH 23390 (NEN Perkin Elmer, Wellesley, MA, USA) or with 2.0 nmol·L⁻¹ of the H₃ receptor agonist [³H]R-α-methyl histamine ([³H]RAMH, Amersham, Buckinghamshire, UK) and increasing concentrations of the D₁ receptor agonist SKF 38393 (Tocris, Ellisville, MO, USA) or H₃ receptor agonist R-α-methyl histamine (RAMH) (triplicates of 13 different competitor concentrations from 0.1 nmol·L⁻¹ to 10 µmol·L⁻¹) in the absence or the presence of 10 nmol·L⁻¹ of the H₃ receptor agonist RAMH or 100 nmol·L⁻¹ of the D₁ receptor agonist SKF 38393 respectively. In all cases, non-specific binding was determined in the presence of an excess of unlabeled ligand [10 µmol·L⁻¹ SCH 23390 (Sigma) for [³H]SCH 23390 binding or 10 µmol·L⁻¹ RAMH for [³H]RAMH binding], and in competition experiments it was confirmed that the value was the same as calculated by extrapolation of the competition curves. Free and membrane-bound ligand were separated by rapid filtration of 500 µL aliquots in a cell harvester (Brandel, Gaithersburg, MD, USA) through Whatman GF/C filters (Brandel) soaked in 0.3% PEI, which were subsequently washed for 5 s with 5 mL of ice-cold Tris-HCl buffer. The filters were incubated with 10 mL of Ecoscint H scintillation cocktail (National Diagnostics, Atlanta, GA, USA) overnight at room temperature, and radioactivity counts were determined by using a Tri-Carb 1600 scintillation counter (PerkinElmer, Boston, MA, USA) with an efficiency of 62%.

Binding data analysis

Due to the homodimeric nature of D₁ receptors (O'Dowd et al., 2005; Kong et al., 2006) and H₃ receptors (Bakker et al., 2006), binding data from competition experiments were analysed by non-linear regression using the commercial Grafit curve-fitting software (Erithacus Software, Surrey, UK), by fitting data to the two-state dimer receptor model (Franco et al., 2005; 2006; Casadóet al., 2007) and not to the classical two-independent-site model for monomeric receptors that considers two binding sites (high and low affinity binding sites). To calculate the macroscopic equilibrium dissociation constants involved in the binding of the agonist SKF 38393 or RAMH to the D₁ receptor or H₃ receptor dimer respectively, the following equation for a competition binding experiment (Casadóet al., 2007) was considered:

(1)

where A represents the radioligand (the D₁ receptor antagonist [³H]SCH 23390 or the H₃ receptor agonist [³H]RAMH) concentration, R_T is the total amount of receptor dimers and K_DA1 and K_DA2 are the macroscopic dissociation constants describing the binding of the first and the second radioligand molecule (A) to the dimeric receptor; B represents the assayed competing compound (the D₁ receptor agonist SKF 38393 or the H₃ receptor agonist RAMH) concentration, and K_DB1 and K_DB2 are, respectively, the equilibrium dissociation constants of the first and second binding of B; K_DAB can be described as a hybrid equilibrium dissociation constant, which is the dissociation constant of B binding to a receptor dimer semi-occupied by A.

Because the radioligand A ([³H]RAMH or [³H]SCH 23390) showed non-cooperative behaviour (Franco et al., 2006); (Casadóet al., 2007), Eqn 1 was simplified to Eqn 2 due to the fact that K_DA2= 4K_DA1 (see Casadóet al., 2007):

(2)

The dimer homotropic cooperativity (D_C) index for the competing ligand B (the agonist SKF 38393) was calculated (see Casadóet al., 2007; Gracia et al., 2008) according to the following expression:

Goodness of fit was tested according to reduced χ² value given by the non-linear regression programme. The test of significance for two different model population variances was based upon the F-distribution (see Casadóet al., 1990, for details). Using this F-test, a probability greater than 95% (P < 0.05) was considered the criterion to select a more complex model (cooperativity) over the simplest one (non-cooperativity). In all cases, a probability of less than 70% (P > 0.30) resulted when one model was not significantly better than the other.

cAMP determination

The SK-N-MC or transfected HEK-293 cells were grown in 25 cm² flasks to 80% confluence and incubated in serum-free medium for 16 h before the experiment. The day of experiment the cells were pre-incubated with 50 µmol·L⁻¹ zardaverine (a phosphodiesterase inhibitor; Tocris) for 10 min at 37°C and treated for 10 min with 100 nmol·L⁻¹ RAMH or 1 µmol·L⁻¹ SKF 81297 (Tocris) in the presence or the absence of 10 µmol·L⁻¹ forskolin (Sigma). When indicated, the H₃ receptor antagonist thioperamide (Sigma) or the D₁ receptor antagonist SCH 23390 (Tocris) were added at 10 µmol·L⁻¹ final concentration and pre-incubated for 5 min before agonist addition. To stop the reaction cells were placed on ice and washed with ice-cold phosphate-buffered saline. The cells were incubated with 200 µL of HClO₄ (4%) for 30 min, 1.5 mol·L⁻¹ KOH was added to reach neutral pH, and samples were centrifuged. The supernatant was frozen at −20°C. The accumulation of cAMP was measured with cyclic AMP (³H) assay system (Amersham Biosciences, Uppsala, Sweden) as described in the manual from the manufacturer.

ERK phosphorylation assay

Cells were grown in 25 cm² flasks to 80% confluence and cultured in serum-free medium for 16 h before the addition of any agent. Cells were treated or not with 10 µmol·L⁻¹ SCH 23390 or 10 µmol·L⁻¹ thioperamide for 30 min before the addition of the agonists 1 µmol·L⁻¹ RAMH or 1 µmol·L⁻¹ SKF 81297 for 2 min. In experiments evaluating Pertussis toxin (PTX), cells were pretreated with the toxin (100 ng·mL⁻¹) for 16 h before ligand addition and in experiments evaluating cholera toxin (CTX), cells were pretreated with the toxin (1 µg·mL⁻¹) for 30 min before ligand addition. At the end of the incubation periods, cells were rinsed with ice-cold phosphate-buffered saline and lysed by the addition of 500 µL of ice-cold lysis buffer (50 mmol·L⁻¹ Tris-HCl pH 7.4, 50 mmol·L⁻¹ NaF, 150 mmol·L⁻¹ NaCl, 45 mmol·L⁻¹β-glycerophosphate, 1% Triton X-100, 20 µmol·L⁻¹ phenyl-arsine oxide, 0.4 mmol·L⁻¹ NaVO₄ and protease inhibitor cocktail). The cellular debris was removed by centrifugation at 13 000×g for 5 min at 4°C, and the protein was quantified by the bicinchoninic acid method by using bovine serum albumin dilutions as standard. To determine the level of ERK1/2 phosphorylation, equivalent amounts of protein (10 µg) were separated by electrophoresis on a denaturing 7.5% SDS-polyacrylamide gel and transferred onto PVDF membranes. The membranes were then probed with a mouse anti-phospho-ERK1/2 antibody (Sigma, 1:5000). In order to rule out that the differences observed were due to the application of unequal amounts of lysates, PVDF blots were stripped and probed with a rabbit anti-ERK1/2 antibody that recognizes both, phosphorylated and non-phosphorylated ERK1/2 (Sigma, 1:40 000). Bands were visualized by the addition of anti-mouse HRP conjugated (Dako, Glostrup, Denmark) or anti-rabbit HRP conjugated (Sigma) secondary antibodies, respectively, and SuperSignal West Pico Chemiluminescent Substrate (Pierce). Bands densities were quantified with a LAS-3000 (Fujifilm, Madrid, Spain), and the level of phosphorylated ERK1/2 isoforms was normalized for differences in loading using the total ERK protein band intensities. Quantitative analysis of detected bands was performed by Image Gauge V4.0 software.

Data analysis

Results are given as mean ± SEM. Differences between group means have been tested for significance (P < 0.05) by using Student's t-test for unpaired samples.

Results

Dopamine D₁–histamine H₃ receptor heteromerization

The BRET approach was used to demonstrate the ability of H₃ receptors to heteromerize with D₁ receptors. BRET measurements were performed in transiently co-transfected HEK-293 cells by using a constant amount of D₁ receptor–RLuc and increasing amounts of H₃ receptor–YFP. The subcellular localization of fusion proteins was investigated and the D₁ receptor–RLuc and H₃ receptor–YFP membrane expression and co-localization is shown in Figure 1A. Fusion of RLuc and YFP to D₁ receptors or to H₃ receptors did not modify receptor binding parameters (results not shown) or receptor function as determined by cAMP assays (Figure 1B). The correlation between properly folded receptors, determined by ligand binding, and fluorescence or luminescence is shown in Figure 1C. The expression level of the fusion proteins was in the range of 0.05 pmol·mg⁻¹ protein for D₁ receptor–RLuc and between 0.3 and 4.5 pmol·mg⁻¹ protein for the different amounts of the transfected cDNA corresponding to H₃ receptor–YFP. These data demonstrate that the fusion proteins are not strongly over-expressed at BRET₅₀. A positive and saturable BRET signal was found for the pair D₁ receptor–RLuc and H₃ receptor–YFP (Figure 1D). From the saturation curve, a BRET_max of 0.034 ± 0.005 units and a BRET₅₀ of 10 ± 4 were calculated. As the human histamine H₄ receptor is closely related to the human H₃ receptor [31% sequence identity at the protein level, which increases to 54% in the transmembrane region; de Esch et al. (2005)], the pair D₁ receptor–RLuc and H₄ receptor–YFP was used as a negative control. Also as negative controls the BRET pairs D₁ receptor–RLuc and cannabinoid CB₁ receptor–YFP or 5HT_2B receptor–RLuc and H₃ receptor–YFP were used. As shown in Figure 1D the negative controls gave a linear non-specific BRET signal, thus confirming the specificity of the interaction between D₁ receptor–RLuc and H₃ receptor–YFP in HEK-293 cells.

Figure 1.

Heteromerization of functional D₁ and H₃ receptors. (A) Confocal microscopy images of HEK-293 cells expressing D₁ receptor–RLuc (0.1 µg plasmid) and H₃ receptor–YFP (0.1 µg plasmid). Proteins were identified by fluorescence or by immunocytochemistry. D₁ receptor–RLuc immunoreactivity in shown in blue (a), H₃ receptor–YFP fluorescence in shown in green (b) and co-localization of D₁ receptor–RLuc and H₃ receptor–YFP is shown in light blue (c). (B) Functionality of D₁ receptor–RLuc (D₁R–RLuc) and H₃ receptor–YFP (H₃R–YFP) constructs. HEK-293 cells transfected with 5 µg of cDNA corresponding to D₁ receptors or D₁ receptor–RLuc were stimulated with the D₁ receptor agonist SKF 81297 (10 µmol·L⁻¹), and HEK-293 cells transfected with 5 µg of cDNA corresponding to H₃ receptors or H₃ receptor–YFP were treated with 10 µmol·L⁻¹ forskolin plus the H₃ receptor agonist RAMH (0.1 µmol·L⁻¹). Results (mean ± SEM; n= 2–4) are expressed as percentage over basal (upper panel) or over forskolin (FK) alone (lower panel); significantly different compared with the basal for D₁ receptors and D₁ receptor–RLuc or compared with forskolin alone for H₃ receptors or H₃ receptor–YFP, (non-paired Student's t-test: *P < 0.05, **P < 0.01 and ***P < 0.001). (C) Correlation between 2.1 nmol·L⁻¹[³H]SCH 23390 binding and luminiscence expression (upper panel) or 1.9 nmol·L⁻¹[³H]RAMH binding and fluorescence expression (lower panel) in HEK-293 cell transfected with increasing amounts of cDNA for D₁ receptor–RLuc (upper panel) or H₃ receptor–YFP (lower panel) (D) D₁–H₃ receptor heteromerization in HEK-293 cells. BRET experiments were performed with HEK-293 cells co-expressing D₁ receptor–RLuc and H₃ receptor–YFP, D₁ receptor–RLuc and CB₁ receptor–YFP, 5HT_2B receptor–RLuc and H₃ receptor–YFP or D₁ receptor–RLuc and H₄ receptor–YFP constructs. Co-transfections were performed with increasing amounts of plasmid–YFP (0.5–9 µg cDNA) whereas the plasmid–RLuc construct was maintained constant (250 ng cDNA). Both fluorescence and luminiscence of each sample were measured before every experiment to confirm similar donor expressions (about 250 000 luminescent units) while monitoring the increase acceptor expression (5000–80 000 fluorescent units). The relative amount of BRET is given as the ratio between the fluorescence of the acceptor and the luciferase activity of the donor. YFP₀ corresponds to the fluorescence value of cells expressing the donor alone. BRET data are expressed as means ± SD of 3–13 different experiments grouped as a function of the amount of BRET acceptor. [³H]RAMH, [³H]R-α-methyl histamine; BRET, Bioluminescence Resonance Energy Transfer; HEK, human embryonic kidney; RLuc, Renilla luciferase; Veh, vehicle.

Intracellular crosstalk between histamine H₃ and dopamine D₁ receptors in HEK-293 cells

To investigate potential functional consequences of D₁–H₃ receptor heteromerization, HEK-293 cells expressing the human D₁ receptor and/or the human H₃ receptor at amounts giving approximately maximum BRET (Figure 1D) were treated with dopamine or histamine receptor agonists, and signalling was assayed by ERK1/2 phosphorylation. When cells expressing H₃ receptors were treated with the selective H₃ receptor agonist RAMH, no phosphorylation of ERKs was found (Figure 2A). On the other hand, when cells expressing D₁ receptors were activated with the selective D₁ receptor agonist SKF 81297, we observed a significant level of ERK1/2 phosphorylation, which was antagonized by the selective D₁ receptor antagonist SCH 23390 (Figure 2A). When HEK-293 cells were transfected simultaneously with D₁ receptors and H₃ receptors, the D₁ receptor agonist also activated the mitogen-activated protein kinase (MAPK) pathway, and this effect was blocked by SCH 23390 (Figure 2B). Interestingly, the H₃ receptor agonist was also able to induce a significant ERK1/2 phosphorylation in co-transfected cells expressing D₁–H₃ receptor heteromers (Figure 2B). The specificity of the effect was proven by the blockade of the RAMH-induced effect by the H₃ receptor antagonist, thioperamide (Figure 2B). These results indicate that the H₃ receptor is able to couple to the MAPK-signalling pathway only in HEK-293 cells expressing D₁ receptors and H₃ receptors.

Figure 2.

Crosstalk between H₃ receptors and D₁ receptors in HEK-293 cells. HEK-293 cells transiently expressing H₃ receptors (HEK-H₃) or D₁ receptors (HEK-D₁) (A) or both (HEK-D₁H₃) (B) were treated for 2 min with the H₃ receptor agonist RAMH (1 µmol·L⁻¹) or with the D₁ receptor agonist SKF 81297 (1 µmol·L⁻¹, SKF), in the presence or in the absence of the H₃ receptor antagonist thioperamide (10 µmol·L⁻¹, Thiop) or the D₁ receptor antagonist SCH 23390 (10 µmol·L⁻¹, SCH), and ERK1/2 phosphorylation (P-ERK) was determined as indicated in Methods. A representative Western blot is shown in each panel. The immunoreactive bands from three independent experiments were quantified, and values represent the mean ± SEM of fold increase of phosphorylation over the basal levels found in untreated cells. Significant differences with respect to the treatment with vehicle, were calculated by Student's t-test for unpaired samples (*P < 0.05 and **P < 0.01). ERK, extracellular signal-regulated kinase; HEK, human embryonic kidney; RAMH, R-α-methyl histamine; Veh, vehicle.

D₁–H₃ receptor heteromers in human neuroblastoma cells

For some receptor pairs it is possible to detect the heteromer receptor fingerprint (Ferréet al., 2007; Franco et al., 2007). This fingerprint often consists of intramembrane receptor–receptor interactions, in which changes in ligand binding characteristics of one receptor are obtained when the partner receptor is activated by using membrane preparations in which no intracellular crosstalk occurs (Agnati et al., 2003; El-Asmar et al., 2005; Ferréet al., 2007; Springael et al., 2007; Vilardaga et al., 2008). We investigated the possible existence of D₁–H₃ receptor intramembrane receptor interactions in SK-N-MC cells as a neuronal cell model. SK-N-MC cells have been used as a good model to transfect H₃ receptors (Bongers et al., 2007); nevertheless, some authors have described the presence of D₁ receptors in SK-N-MC cells (Sidhu et al., 1999; Chen et al., 2003; Moussa et al., 2006; Robinson et al., 2008) and some controversy exists about the functionality of these receptors, whether they couple to different G proteins (Kimura et al., 1995) and whether they signal (Chen et al., 2004) or not (Chan et al., 2005) towards the MAPK cascade. Our SK-N-MC cell clone expresses less than 0.030 pmol·(mg protein)⁻¹ of D₁ receptors [0.009 ± 0.004 pmol·(mg protein)⁻¹ in the parental cell clone and 0.026 ± 0.005 pmol·(mg protein)⁻¹ in the SK-N-MCMC/H3 cell clone], determined as [³H]SCH 23390 maximum binding, that is, at near saturating (>90%) concentrations of the radioligand. It should be noted that the D₁ receptor agonist did not induce ERK1/2 phosphorylation neither in SK-N-MCMC/H3 (Figure 3A) or in parental cells (results not shown). Therefore it seems that different SK-N-MC cell clones may give different results.

Figure 3.

ERK1/2 phosphorylation (P-ERK) via the D₁–H₃ receptor heteromer in human neuroblastoma cells. SK-N-MC cells expressing H₃ receptors (SK-N-MC/H₃) or D₁ receptors (SK-N-MC/D₁) or both (SK-N-MC/D₁H₃) were treated with the H₃ receptor agonist, RAMH (1 µmol·L⁻¹), or with the D₁ receptor agonist, SKF 81297 (1 µmol·L⁻¹, SKF) alone or in combination, in the presence or in the absence of the H₃ receptor antagonist, thioperamide (10 µmol·L⁻¹, Thiop) or the D₁ receptor antagonist, SCH 23390 (10 µmol·L⁻¹, SCH). ERK1/2 phosphorylation was determined as indicated in Methods after 2 min of agonist treatment (A, B and C). In (D) a time–course response of ERK1/2 phosphorylation induced by 1 µmol·L⁻¹ SKF 81297 or 1 µmol·L⁻¹ RAMH in SK-N-MC/D₁H₃ cells is shown. A representative Western blot is shown in each panel. The immunoreactive bands from three to four experiments were quantified, and values represent the mean ± SEM of fold increase of phosphorylation over the basal levels found in untreated cells. Significant differences were calculated by Student's t-test for unpaired samples (*P < 0.05, **P < 0.01, ***P < 0.001). ERK, extracellular signal-regulated kinase; RAMH, R-α-methyl histamine; Veh, vehicle.

Membranes prepared from SK-N-MC human neuroblastoma cells stably expressing human versions of H₃ receptors and D₁ receptors (SK-N-MC/D₁H₃ cells) were used in binding competition experiments with [³H]SCH 23390 (2.2 nmol·L⁻¹) as radioligand and increasing concentrations of SKF 38393 as competitor in the presence and in the absence of RAMH (10 nmol·L⁻¹). Binding data were fitted to the two-state dimer receptor model (Franco et al., 2005; 2006; Casadóet al., 2007), to calculate the macroscopic equilibrium dissociation constants and the cooperativity index. The competition curve was biphasic in the absence of RAMH (significantly better than monophasic; F-test: P < 0.05), showing cooperativity in the D₁ receptor agonist binding, but monophasic in the presence of RAMH. Variations in binding parameter values are shown in Table 1. These results indicate that an intramembrane crosstalk occurs between these receptors by which H₃ receptor activation induces a shift from a cooperative to a non-cooperative binding and an overall decrease of affinity for the D₁ receptor agonist binding. In contrast, D₁ receptor stimulation did not influence the agonist binding to H₃ receptor. In fact, competition experiments of 2 nmol·L⁻¹[³H]RAMH binding versus increasing RAMH concentrations, performed as indicated in Methods, gave similar R_T and K_DB1 values for the non-cooperative RAMH binding both in the absence [0.46 ± 0.05 pmol·(mg protein)⁻¹ and 2.9 ± 0.3 nmol·L⁻¹] or presence [0.42 ± 0.04 pmol·(mg protein)⁻¹ and 3.0 ± 0.3 nmol·L⁻¹] of 100 nmol·L⁻¹ SKF 38393.

Table 1.

Parameter values from competition experiments of [³H]SCH 23390 versus SKF 38393 in the presence and in the absence of RAMH (two-state dimer model)

Agonists	Parameters
	RT[pmol·(mg protein)−1]	KDB1 (nmol·L−1)	KDB2 (µmol·L−1)	DCB
SKF 38393	0.436 ± 0.011	41 ± 3	1.3 ± 0.1	−0.85
SKF 38393 + RAMH	0.404 ± 0.007	95 ± 9*	−	0

Data are mean ± SEM values of three experiments; D_CB, dimer cooperativity index for the binding of SKF 38393; K_DB1 and K_DB2, equilibrium dissociation constants for the first and second bindings of SKF 38393; RAMH, R-α-methyl histamine; R_T, total amount of receptor dimers.

^*Significantly different compared with the K_DB1 value of SKF 38393 alone, P < 0.05.

Signal transduction via D₁–H₃ receptor heteromers in human neuroblastoma cells

As described above, in HEK-293 cells, H₃ receptors were able to mediate activation of the MAPK signalling pathway only through D₁–H₃ receptor heteromerization, demonstrated by BRET. This characteristic of the heteromer can also be used as a signalling fingerprint to identify the D₁–H₃ receptor heteromers. Thus, similar biochemical experiments were performed in SK-N-MC/D₁H₃ cells and cells transfected with only one receptor. As shown in Figure 3B, cells expressing D₁ receptors are able to induce ERK1/2 phosphorylation in response to the treatment with the D₁ receptor agonist SKF 81297, an effect that was blocked by SCH 23390. In SK-N-MC/H₃ cells, RAMH had no effect on ERK1/2 phosphorylation (Figure 3B). However, in SK-N-MC/D₁H₃ cells both RAMH and SKF 81297 were able to activate the MAPK pathway, and co-activation of the two receptors did not result in synergism (Figure 3C). As shown in Figure 3D, there was no change in the time course of ERK1/2 phosphorylation when the agonists for D₁ receptors or H₃ receptors were used individually in SK-N-MC/D₁H₃ cells; the maximum phosphorylation was reached at 2 min and disappeared after 10 min stimulation. Overall the results were qualitatively identical to those obtained by using transiently transfected HEK-293 cells, demonstrating D₁–H₃ receptor heteromerization in neuroblastoma cells. These results also indicate that H₃ receptors are able to couple to the MAPK pathway only in neuroblastoma cells expressing D₁–H₃ receptor heteromers. Similar experiments were performed by using a mutant version of H₃ receptors (R3.50A; Arg 132 substituted by Ala) that is neither able to bind full agonists nor to signal (Appendix S1; Figure S1). The D₁ receptor agonist was not able to provide any ERK1/2 phosphorylation signal when D₁ receptors were co-expressed with the H₃ R3.50A receptors (data not shown). This indicates that the D₁ receptor signals towards MAPK via H₃ receptors in cells co-expressing both receptors. Interestingly, in SK-N-MC/D₁H₃ cells, SKF 81297-induced ERK1/2 phosphorylation was reversed not only by SCH 23390, the specific D₁ receptor antagonist, but also by thioperamide, the H₃ receptor antagonist. Furthermore, RAMH-induced ERK1/2 phosphorylation in these cells was not only antagonized by thioperamide but also by SCH 23390 (Figure 4). It should be noted that both SKF 81297 and SCH 23390 are specific ligands for D₁ receptors and do not appreciably interact with H₃ receptors, as in SK-N-MC/H₃ cells they were not able to reduce the 1.9 nmol·L⁻¹[³H]RAMH binding [0.61 ± 0.02 vs. 0.57 ± 0.02 and 0.54 ± 0.04 pmol·(mg protein)⁻¹ in the presence of 10 µmol·L⁻¹ SKF 81297 or 10 µmol·L⁻¹ SCH 23390, respectively]. Analogously, thioperamide and RAMH are specific H₃ receptor ligands, as they were not able to reduce the 1.9 nmol·L⁻¹[³H]SCH 23390 binding to SK-N-MC/D₁H₃ cells [0.72 ± 0.03 vs. 0.71 ± 0.02 and 0.73 ± 0.01 pmol·(mg protein)⁻¹ in the presence of 10 µmol·L⁻¹ thioperamide or 10 µmol·L⁻¹ RAMH, respectively].

Figure 4.

Effect of receptor antagonists on ERK1/2 phosphorylation (P-ERK) via the D₁–H₃ receptor heteromer in human neuroblastoma cells. SK-N-MC cells expressing H₃ receptors and D₁ receptors (SK-N-MC/D₁H₃) were treated with the H₃ receptor agonist, RAMH (1 µmol·L⁻¹), or the D₁ receptor agonist, SKF 81297 (1 µmol·L⁻¹. SKF), in the presence or in the absence of the H₃ receptor antagonist, thioperamide (10 µmol·L⁻¹, Thiop) or the D₁ receptor antagonist, SCH 23390 (10 µmol·L⁻¹, SCH). ERK1/2 phosphorylation was determined as indicated in Methods after 2 min of agonist treatment. A representative Western blot is shown. The immunoreactive bands from four experiments were quantified, and values represent the mean ± SEM of percentage of phosphorylation of agonist-treated cells. Significant differences were calculated by Student's t-test for unpaired samples (**P < 0.01, ***P < 0.001). ERK, extracellular signal-regulated kinase; RAMH, R-α-methyl histamine.

As expected from the known coupling of H₃ receptor to heterotrimeric G_i proteins (Lovenberg et al., 1999; Drutel et al., 2001; Leurs et al., 2005), RAMH markedly inhibited the 10 µmol·L⁻¹ forskolin-stimulated production of cAMP in SK-N-MC/H₃ cells, and this effect was effectively blocked by thioperamide (Figure 5A), showing that in these neuroblastoma cells the H₃ receptors are functional. Consistent with the very low D₁ receptor expression in parental SK-N-MC cells and with the reported coupling of D₁ receptors to G_s proteins (Neve et al., 2004), SKF 81297 was not able to increase cAMP in our SK-N-MC cell clone but was able to increase the intracellular levels of cAMP in SK-N-MC/D₁ cells, an effect that was fully blocked by SCH 23390 (Figure 5B). Interestingly, in SK-N-MC/D₁H₃ cells, RAMH was still able to inhibit the cAMP accumulation induced by forskolin, and thioperamide blocked this action. In the same cell clone, which co-expresses the two receptors, SKF 81297 did not have any significant effect on cAMP production but reduced the forskolin-induced cAMP levels (Figure 5C). This suggests that D₁ receptors are signalling in the D₁–H₃ receptor heteromer by coupling to an inhibitory G protein.

Figure 5.

cAMP production by D₁–H₃ receptor heteromer in human neuroblastoma cells. SK-N-MC cells expressing (A) H₃ receptors (SK-N-MC/H₃) or (B) D₁ receptors (SK-N-MC/D₁) or (C) both (SK-N-MC/D₁H₃) were treated or not with 10 µmol·L⁻¹ forskolin (FK) and the H₃ receptor agonist, RAMH (0.1 µmol·L⁻¹), and/or the D₁ receptor agonist, SKF 81297 (1 µmol·L⁻¹, SKF). The effect of the H₃ receptor antagonist, thioperamide (10 µmol·L⁻¹, Thiop) or the D₁ receptor antagonist, SCH 23390 (10 µmol·L⁻¹, SCH) was also assayed. cAMP levels were determined as indicated in Methods. Results are expressed as fold increase over basal levels obtained in untreated cells (mean ± SEM of three to five experiments). Significant differences were calculated by Student's t-test for unpaired samples (*P < 0.05, **P < 0.01). RAMH, R-α-methyl histamine; Veh, vehicle.

Based on the data described above, it is likely that a single heterotrimeric G protein, probably of the G_i/o type, is transducing the signal generated by either dopamine or histamine receptor agonists through the H₃–D₁ receptor heteromer. To check for this possibility SK-N-MC/D₁H₃ cells were pretreated with PTX, which specifically inactivates G_i/G_o-mediated signalling pathways, or with CTX, which activates adenylyl cyclase by catalysing ADP-ribosylation of the stimulatory Gα_s protein. After pretreatment with these toxins, H₃ receptors and D₁ receptors were activated by using respectively RAMH or SKF 81297. Whereas PTX inhibited the phosphorylation of ERK1/2 induced by RAMH and SKF 81297, CTX had no significant effect on the activation induced by any of the agonists (Figure 6). These results suggest that the activation of MAPK pathway through any of the two receptors in the D₁–H₃ receptor heteromer depends on G_i coupling.

Figure 6.

Effect of PTX and CTX on SKF- or RAMH-induced ERK1/2 phosphorylation (P-ERK). SK-N-MC cells expressing H₃ receptors and D₁ receptors (SK-N-MC/D₁H₃) were treated with PTX (100 ng·mL⁻¹) for 16 h or with CTX (1 µg·mL⁻¹) for 30 min prior to the addition of the H₃ receptor agonist, RAMH (1 µmol·L⁻¹), or the D₁ receptor agonist, SKF 81297 (1 µmol·L⁻¹, SKF). ERK1/2 phosphorylation was determined as indicated in Methods. A representative Western blot is shown. The immunoreactive bands from four experiments were quantified, and values represent the mean ± SEM of fold increase of phosphorylation over basal levels found in untreated cells. Significant differences were calculated by Student's t-test for unpaired samples (*P < 0.05 and **P < 0.01). CTX, cholera toxin; ERK, extracellular signal-regulated kinase; RAMH, R-α-methyl histamine; PTX, Pertussis toxin; Veh, vehicle.

Discussion

It seems that most, if not all, members of the GPCR superfamily can exist as homodimers (Bouvier, 2001; Devi, 2001; Marshall, 2001; Rios et al., 2001; George et al., 2002; Franco et al., 2003; Terrillon and Bouvier, 2004; Prinster et al., 2005; Milligan, 2006). The first indication of the existence of GPCR heteromers was obtained with radioligand binding experiments, which showed the existence of biochemical interactions between different GPCRs in brain membrane preparations (Agnati et al., 2003). In this kind of interactions, initially known as ‘intramembrane receptor–receptor interactions’, stimulation of one receptor changes the binding characteristics of another receptor for endogenous or exogenous ligands in crude membrane preparations (Agnati et al., 2003). This implied the lack of involvement of intracellular signalling and suggested some kind of allosteric interaction between adjacent receptors. Thus, at the beginning of the 90s, it was hypothesized that this intramembrane interaction could result from an intermolecular crosstalk, implying receptor heteromerization (Zoli et al., 1993). This is now considered as a biochemical fingerprint of a receptor heteromer (Ferréet al., 2007; Franco et al., 2007). Here we show that D₁ receptors and H₃ receptors are able to form D₁–H₃ receptor heteromers by BRET, in transiently transfected human embryonic cells, and by radioligand experiments in SK-N-MC/D₁H₃ cells, in which a specific H₃ receptor agonist led to the disappearance of the cooperative D₁ receptor agonist binding and to a significant change in the affinity of the D₁ receptor for the agonist.

The crosstalk occurring via receptor heteromers has different components. One of them is the above discussed change in binding characteristics of one receptor upon activation of the partner receptor. Another is the crosstalk at the level of second messengers. For heteromers in which one of the constituent receptors is coupled to G_i/o whereas the other is coupled to G_s proteins, co-activation of the receptors would result not in a functional antagonism but in contradictory messages for the cell. Recent reports are providing clues to solve this conundrum. Significant advances in the case of heteromers for the same neurotransmitter have been achieved (Jordan and Devi, 1999; George et al., 2000; Fan et al., 2005; Ciruela et al., 2006; Rashid et al., 2007). Recent data indicate that in neurons co-expressing D₁ receptors, a G_s-coupled receptor, and D₂ receptors, a G_i-coupled receptor, D₁–D₂ receptor heteromers are formed that couple to a G_q protein (Rashid et al., 2007). This makes possible that a single neurotransmitter may increase cAMP levels, decrease cAMP levels or modify intracellular calcium levels depending on whether a given neuron (or microdomain in a neuron) expresses, respectively, the D₁ receptor, the D₂ receptor or the D₁–D₂ receptor heteromer. Two different neurotransmitters, dopamine and histamine, can interact with D₁–H₃ receptor heteromers. In neuroblastoma cells co-expressing D₁ receptors and H₃ receptors there is a change in the D₁ receptor coupling from the G_s to the G_i protein, to which H₃ receptors are already coupled. In fact, in the presence of the H₃ receptor, D₁ receptors were no longer coupled to G_s, and could not activate adenylyl cyclase, but were coupled to G_i, which transduced the signal towards the MAPK pathway. On the other hand, H₃ receptors in cells co-expressing the two receptors could signal through both adenylyl cyclase (inhibiting enzyme activity) and MAPK (increasing ERK1/2 phosphorylation). These results indicate that D₁–H₃ receptor heteromers constitute unique devices to direct dopaminergic and histaminergic signalling towards the MAPK pathway in a G_s-independent and G_i-dependent manner. In the SK-N-MC cell clone stably expressing the human H₃ receptors near to physiological receptor densities [0.1–1 pmol·(mg protein)⁻¹], the H₃ receptor agonist did not promote ERK1/2 phosphorylation unless the D₁ receptor was co-expressed. It has been described that agonist binding to H₃ receptors expressed at high densities in Chinese hamster ovary or in COS-7 cells can phosphorylate ERK1/2 (Drutel et al., 2001; Gbahou et al., 2003). In contrast to the cAMP response, the H₃ receptor did not exhibit constitutive activation of the MAPK pathway (Gbahou et al., 2003). Whether ERK1/2 phosphorylation in these cells is solely due to the action of G_βγ subunits or to crosstalk with another receptor in these cells remains to be elucidated. In vivo, the first evidence of a positive correlation between ERK phosphorylation and memory improvement was given by Giovannini et al. (2003), who demonstrated an improvement in fear memory by H₃ receptor-elicited ERK2 phosphorylation in hippocampal CA3 neurons in which the D₁ receptor is co-expressed (Pantazopoulos et al., 2004)

Our results would be in agreement with the recently suggested 1:2 stoichiometry for the G protein: receptor interaction (Herrick-Davis et al., 2005). The results obtained by co-expressing D₁ receptors and the mutant version of H₃ receptors unable to activate MAPK indicate that GPCR activation results from a dynamic intersubunit interplay as shown in dimeric metabotropic glutamate receptors (Brock et al., 2007). The possibility that better explains the overall results is that D₁ receptors are able to signal to the MAP kinase in the absence of the H₃ receptor, but that in the presence of this receptor the signalling to ERK is mediated by the H₃ receptor and not via the D₁ receptor. Then, in the presence of non-functional H₃ receptors, D₁ receptor agonists are unable to produce ERK phosphorylation. Interestingly, not only the antagonist of their respective receptors but also the antagonist of the partner receptor counteracted the effect of D₁ receptor or H₃ receptor activation. Thus, an antagonist of one of the receptor units in the D₁–H₃ receptor heteromer can induce conformational changes in the other receptor unit and block specific signals originating in the heteromer. This fact broadens the therapeutic potential for GPCR antagonists.

Glossary

Abbreviations:

[³H]RAMH

[³H]R-α-methyl histamine

BRET

Bioluminescence Resonance Energy Transfer

CTX

cholera toxin

EYFP

enhanced yellow variant of green fluorescent protein

GPCR

G protein-coupled receptor

PEI

polyethylenimine

PTX

Pertussis toxin

RAMH

R-α-methyl histamine

RLuc

Renilla luciferase

Conflict of interest

The authors state no conflict of interest.

Supporting Information

Additional Supporting Information may be found in the online version of this article:

Appendix S1

Figure S1 Binding and signalling of wild type or mutant (R3.50A) H₃R in transfected HEK-293 cells. (A, B) HEK- 293 cells co-transfected with pTATAlucNEO/CRE121-3 (pTLNC121-3) CRE-luciferase reporter gene, and either the wild type or the mutant version of human H₃R (Arg 132 substituted by Ala; see Methods) were treated with a full (R-αmethyl histamine) or an inverse (A-349821) agonist and the activity of the reporter gene was recorded (see Methods). (C, D) Binding to membranes from cells transfected with either the wild type or the mutant version of human H3R were performed by using (see Methods) either radiolabelled full (NAMH) or inverse (A-349821) agonists.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.