ACEA, ghrelin (human), rimonabant hydrochloride, and YIL 781 hydrochloride were purchased from Tocris Bioscience (Bristol, United Kingdom). Concentrated (10 mM) stock solutions prepared in DMSO, milli-Q^® H₂O (Merck/Millipore, Darmstadt, Germany), or ethanol were stored at −20°C. In each experimental session, aliquots of concentrated solutions of compounds were thawed and conveniently diluted in the appropriate experimental solution.

C57BL/6J female mice were used for the experiments. All animals were subjected to a 12 h/12 h light/ dark cycle in a temperature- and humidity-controlled room and were allowed free access to water and standard laboratory chow. C57BL/6J mice were randomly assigned to a high-fat diet (HFD; 60% kcal from fat; catalog no. D12492, Research Diets, New Brunswick, NJ, USA) or standard diet (STD; 10% kcal from fat; catalog no. D12450B, Research Diets) for 60 days. Primary striatal neurons were obtained from fetuses of mothers on STD or HFD diets. Pregnant animals were killed by cervical dislocation during the light phase. All animal procedures were performed in agreement with European guidelines (2010/63/EU) and approved by the University of Barcelona Ethical Committee, which reports to the regional Government (Protocol #9659; Generalitat de Catalunya, May 24, 2019).

Human embryonic kidney HEK-293T (lot 612968) cells were acquired from the American Type Culture Collection (ATCC). They were amplified and frozen in liquid nitrogen in several aliquots. Cells from each aliquot were used until passage 19. HEK-293T cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Paisley, Scotland, United Kingdom) supplemented with 2 mM L-glutamine, 100 U/ml penicillin/streptomycin, MEM Non-Essential Amino Acid Solution (1/100), and 5% (v/v) heat-inactivated Fetal Bovine Serum (FBS; all supplements were from Invitrogen, Paisley, Scotland, United Kingdom). Cells were maintained in a humid atmosphere of 5% CO₂ at 37°C.

Cells were transiently transfected with the corresponding cDNAs using the PEI (PolyEthylenImine, Sigma-Aldrich, St. Louis, MO, USA) method as previously described (Carriba et al., 2008; Hradsky et al., 2011; Navarro et al., 2012). Four hours after transfection, growth medium was replaced by a complete medium. Experiments were carried out 48 h later.

To prepare primary striatal neurons, brains from fetuses of pregnant mice were removed (gestational age: 17 days). Neurons were isolated as described in Hradsky et al. (2013) and plated at a confluence of 40,000 cells/0.32 cm². Briefly, the samples were dissected and, after careful removal of the meninges, digested for 20 min at 37°C with 0.25% trypsin. Trypsinization was stopped by adding an equal volume of culture medium (Dulbecco’s modified Eagle medium-F-12 nutrient mixture, Invitrogen). Cells were brought to a single cell suspension by repeated pipetting followed by passage through a 100 μm-pore mesh. Pelleted (7 min, 200× g) cells were resuspended in supplemented DMEM and seeded at a density of 3.5 × 10⁵ cells/ml. The next day, the medium was replaced by neurobasal medium supplemented with 2 mM L-glutamine, 100 U/ml penicillin/streptomycin, and 2% (v/v) B27 medium (Gibco). Neuronal cultures were used for assays after 15 days of culture. Using NeuN as a marker, the percentage of neurons in the cultures was >90%.

The human cDNAs for the CB₁R, GHS-R1a, and D₁R cloned in pcDNA3.1 were amplified without their stop codons using sense and antisense primers. The primers harbored either unique BamHI and KpnI sites for CB₁R and HindIII and BamHI sites for GHS-R1a and D₁R. The fragments were subcloned to be in frame with an enhanced yellow fluorescent protein (pEYFP-N1; Clontech, Heidelberg, Germany) and the Renilla luciferase protein (Rluc; pRluc-N1; PerkinElmer, Wellesley, MA) on the C-terminal end of the receptor to produce CB₁R-YFP, D₁R-Rluc, and GHS-R1a-Rluc.

HEK-293T cells transfected with cDNAs for CB₁R-YFP and GHS-R1a-Rluc were fixed in 4% paraformaldehyde for 15 min and then washed twice with PBS containing 20 mM glycine before permeabilization with the same buffer containing 0.2% Triton X-100 (5 min incubation). The samples were treated for 1 h with blocking solution (PBS containing 1% bovine serum albumin) and labeled with a mouse anti-Rluc (1/100; MAB4400, Millipore, Burlington, MA, USA) as primary antibody and subsequently treated with Cy3-conjugated anti-mouse IgG (1/200; 715-166-150; Jackson ImmunoResearch) as the secondary antibody (1 h each). The samples were washed several times and mounted with 30% Mowiol (Calbiochem, San Diego, USA). Nuclei were stained with Hoechst (1/100). Samples were observed under a Zeiss 880 confocal microscope (Carl Zeiss, Oberkochen, Germany).

HEK-293T cells growing in 6-well plates were transiently cotransfected with a constant amount of cDNA encoding for GHS-R1a fused to Renilla luciferase (GHS-R1a-Rluc) and with increasing amounts of cDNA corresponding to CB₁ receptor fused to the yellow fluorescent protein (CB₁R-YFP). For negative control, cells were cotransfected with a constant amount of cDNA encoding for D₁R-Rluc and with increasing amounts of cDNA for CB₁R-YFP. Forty-eight hours post-transfection cells were washed twice in quick succession with HBSS (137 mM NaCl; 5 mM KCl; 0.34 mM Na₂HPO₄; 0.44 mM KH₂PO₄; 1.26 mM CaCl₂; 0.4 mM MgSO₄; 0.5 mM MgCl₂; and 10 mM HEPES, pH 7.4) supplemented with 0.1% glucose (w/v), detached by gently pipetting and resuspended in the same buffer. To have an estimation of the number of cells per plate, protein concentration was determined using a Bradford assay kit (Bio-Rad, Munich, Germany) with bovine serum albumin dilutions for standardization. To quantify YFP-fluorescence expression, cells were distributed (20 μg protein) in 96-well microplates (black plates with a transparent bottom; Porvair, Leatherhead, UK). Fluorescence was read using a Mithras LB 940 (Berthold, Bad Wildbad, Germany) equipped with a high-energy xenon flash lamp, using a 10-nm bandwidth excitation and emission filters at 485 and 530 nm, respectively. YFP-fluorescence expression was determined as the fluorescence of the sample minus the fluorescence of cells expressing only protein-Rluc. For the BRET measurements, the equivalent of 20 μg of cell suspension was distributed in 96-well microplates (white plates; Porvair), and 5 μM coelenterazine H was added (PJK GMBH, Kleinblittersdorf, Germany). Then, 1 min after coelenterazine H addition, the readings were collected using a Mithras LB 940 (Berthold, Bad Wildbad, Germany), which allowed the integration of the signals detected in the short-wavelength filter at 485 nm (440–500 nm) and the long-wavelength filter at 530 nm (510–590 nm). To quantify receptor-Rluc expression, luminescence readings were collected 10 min after 5 μM coelenterazine H addition. The net BRET is defined as [(long-wavelength emission)/(short-wavelength emission)]-Cf where Cf corresponds to [(long-wavelength emission)/(short-wavelength emission)] for the Rluc construct expressed alone in the same experiment. The BRET curves were fitted assuming a single phase by a non-linear regression equation using the GraphPad Prism software (San Diego, CA, USA). BRET values are given as milli-BRET units (mBU: 1000 × net BRET).

HEK-293T cells transfected with the cDNAs for CB₁R (1 μg) and/or GHS-R1a (1.5 μg) and neuronal primary cultures were plated in 6-well plates. Two hours before initiating the experiment, the cell-culture medium was replaced by the non-supplemented DMEM medium. Then, cells were detached, resuspended in the non-supplemented DMEM medium containing 50 μM zardaverine, and plated in 384-well microplates (2,500 cells/well). Cells were pretreated (15 min) with the corresponding antagonists (1 μM rimonabant for CB₁R and 1 μM YIL 781 for GHS-R1a) or vehicle and stimulated with agonists (1 nM, 10 nM, 100 nM, and 1 μM ACEA for CB₁R or 200 nM ghrelin for GHS-R1a; 15 min) before the addition of 0.5 μM FK or vehicle. Finally, the reaction was stopped by the addition of the Eu-cAMP tracer and the ULight-conjugated anti-cAMP monoclonal antibody prepared in the “cAMP detection buffer” (PerkinElmer). All steps were performed at 25°. Homogeneous time-resolved fluorescence energy transfer (HTRF) measures were performed after 60 min incubation at RT using the Lance Ultra cAMP kit (PerkinElmer, Waltham, MA, USA). Fluorescence at 665 nm was analyzed on a PHERAstar Flagship microplate reader equipped with an HTRF optical module (BMG Lab Technologies, Offenburg, Germany).

To determine extracellular signal-regulated kinase 1/2 (ERK1/2) phosphorylation, HEK-293T transfected cells and primary striatal neurons were plated (50,000 cells/well) in transparent Deltalab 96-well plates and kept in the incubator for 15 days. Two hours before the experiment, the medium was replaced by non-supplemented DMEM medium. Next, the cells were pre-treated at RT for 10 min with antagonists (1 μM rimonabant for CB₁R and YIL 781 for GHS-R1a) or vehicle and stimulated for an additional 7 min with selective agonists (1 nM, 10 nM, 100 nM, 1 μM ACEA for CB₁R and 200 nM ghrelin for GHS-R1a). Then, cells were washed twice with cold PBS before the addition of 30 μl/well “Ultra lysis buffer” -PerkinElmer- (15 min treatment). Afterward, 10 μl of each supernatant was placed in white ProxiPlate 384-well plates and ERK1/2 phosphorylation was determined using an AlphaScreen^®SureFire^® kit (PerkinElmer), following the instructions of the supplier, and using an EnSpire^® Multimode Plate Reader (PerkinElmer, Waltham, MA, USA). The reference value (100%) was the value achieved in the absence of any treatment (basal). Agonist effects were given in percentage with respect to the basal value.

HEK-293T cells were transfected with the cDNAs for CB₁R (1 μg) and/or GHS-R1a (1.5 μg) in the presence of 1 μg cDNA for the calmodulin-based calcium GCaMP6 sensor (Chen et al., 2013) using the PEI method. 48 h after transfection, cells were detached using Mg⁺²-free Locke’s buffer (pH 7.4; 154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO₃, 2.3 mM CaCl₂, 5.6 mM glucose and 5 mM HEPES) supplemented with 10 μM glycine. 1,500 cells per well were plated in 96-well black, clear-bottom, microtiter plates. Then, cells were incubated for 10 min with the CB₁R and GHS-R1a antagonists (1 μM rimonabant or 1 μM YIL 781), and subsequently stimulated with selective agonists (1 nM, 10 nM, 100 nM, 1 μM ACEA, or 200 nM ghrelin). Upon excitation at 488 nm, real-time 515 nm fluorescence emission due to calcium-ion complexed GCaMP6 was recorded on the EnSpire^® Multimode Plate Reader (every 5 s, 100 flashes per well).

Physical interaction between CB₁R and GHS-R1a was detected using the Duolink in situ PLA detection Kit (OLink; Bioscience, Uppsala, Sweden) following the instructions of the supplier. Primary neurons were grown on glass coverslips, fixed in 4% paraformaldehyde for 15 min, washed with PBS containing 20 mM glycine to quench the aldehyde groups, and permeabilized with the same buffer containing 0.05% Triton X-100 (20 min). Then, samples were successively washed with PBS. After 1 h incubation at 37°C with the blocking solution in a pre-heated humidity chamber, primary neurons were incubated overnight in the antibody diluent medium with a mixture of equal amounts of mouse anti-CB₁R (1/100; sc-293419, Santa Cruz Technologies, Dallas, TX, USA) and rabbit anti-GHS-R1a (1/100; ab95250, Abcam, Cambridge, United Kingdom) to detect CB₁R-GHS-R1a complexes. Neurons were processed using the PLA probes detecting primary antibodies (Duolink II PLA probe plus and Duolink II PLA probe minus) diluted in the antibody diluent solution (1:5). Ligation and amplification were done as indicated by the supplier. Samples were mounted using the mounting medium with Hoechst (1/100; Sigma-Aldrich) to stain nuclei. Samples were observed in a Zeiss 880 confocal microscope (Carl Zeiss, Oberkochen, Germany) equipped with an apochromatic 63× oil immersion objective (N.A. 1.4) and a 405 nm and a 561 nm laser lines. For each field of view, a stack of two channels (one per staining) and four Z stacks with a step size of 1 μm were acquired. The number of neurons containing one or more red spots vs. total cells (blue nucleus) was determined, and the unpaired t-test was used to compare the values (red dots/cell) obtained.

Cannabis sativa L consumption has an orexigenic effect via a mechanism in which a hormone of the endocrine system, ghrelin, participates. To identify whether or not there are functional interactions between the cannabinoidergic and the orexinergic systems, we first tested a potential interaction between the CB₁R and the functional form of the ghrelin receptor, GHS-R1a. Immunocytochemical assays in HEK-293T cells transfected with the cDNA of the GHS-R1a fused to Renilla luciferase (Rluc) and/or the cDNA for the CB₁R fused to the Yellow Fluorescent Protein (YFP) led to detect the receptors at the plasma membrane level with a marked colocalization when coexpressed (Figure 1A).

As colocalization may be found for proteins that are close (approximately 200 nm apart) but may not be directly interacting, a Bioluminescence Resonance Energy Transfer (BRET) assay was performed in HEK-293T cells cotransfected with a constant amount of the cDNA for GHS-R1a-Rluc and increasing amounts of cDNA for CB₁R-YFP. A saturation curve (BRET_max = 44 ± 4 mBU, BRET₅₀ = 280 ± 70) was obtained, demonstrating a direct interaction between the two receptors in this heterologous expression system (Figure 1B). When the same experiment was performed in cells pretreated for 30 min with the selective CB₁R agonist, arachidonyl-2’-chloroethylamide (ACEA, 100 nM), no significant differences were observed (BRET_max = 47 ± 4 mBU, BRET₅₀ = 400 ± 10). This result indicates that CB₁R-GHS-R1a interaction is not affected by the activation of the CB₁R. As a negative control, HEK-293T cells were transfected with a constant amount of dopamine D₁ receptor-Rluc cDNA and increasing amounts of CB₁R-YFP cDNA; the nonspecific linear signal indicates a lack of interaction between these two proteins (Figure 1C).

After identifying a direct interaction between CB₁R and GHS-R1a, the functional consequences of the interaction were investigated. Signaling assays were performed considering that the CB₁R couples to G_i and that, although the canonical protein that couples to GHS-R1a receptor is G_q, the ghrelin receptor may also couple to G_i. The activation of any of the receptors in the presence of forskolin (FK), which activates adenylate cyclase, led to a decrease in cytosolic cAMP levels in HEK-293T cells expressing CB₁R or GHS-R1a. In cells expressing the CB₁R, the selective agonist, arachidonyl-2’-chloroethylamide (ACEA), led to a dose-response decrease in (FK)-stimulated cAMP levels that were not affected by pretreatment with ghrelin. Moreover, ghrelin treatment induced no effect over cannabinoid CB₁R, demonstrating ligand selectivity. The effect of ACEA was specific as it was blocked by a selective CB₁R antagonist, rimonabant (Figure 2A). In cells expressing the GHS-R1a, 200 nM ghrelin induced a significant (circa 30%) decrease of FK-induced cAMP level that was not affected by ACEA but that was completely counteracted by YIL 781, a selective GHS-R1a receptor antagonist (Figure 2B). It was confirmed that ACEA did not induce any effect over the GHS-R1a. When the G_q coupling was assayed using the GCaMP6 sensor of cytoplasmic Ca²⁺, no signal was obtained in cells expressing the CB₁R (Figure 2C), while in cells expressing the GHS-R1a receptor ghrelin lead to a transient peak of cytosolic [Ca²⁺] that was not modified by preincubation with ACEA but that was prevented upon preincubation with the ghrelin receptor antagonist (Figure 2D).

In HEK-293T cells expressing CB₁ and GHS-R1a receptors, the ghrelin-induced decrease of FK-stimulated cAMP levels was significant although lower than that in cells only expressing the ghrelin receptor (37% vs. 15%; Figure 3A). Interestingly, the presence of the ghrelin receptor uncoupled CB₁R activation from G_i protein, at least at a functional level. In fact, ACEA even at the highest concentration (1 μM) was not able to significantly reduce FK-induced cAMP levels. Functionality was, however, found when cells were treated with ghrelin and there was no additive effect at concentration of 10 nM ACEA or higher; the lower concentration of ACEA (1 nM) showed a trend to increase the ghrelin-induced effect (Figure 3A). Results related to G_q coupling were noteworthy as ACEA, which did not lead to calcium ion mobilization by itself, significantly increased the effect of ghrelin. Remarkably, the potentiation was much stronger at low doses of ACEA (1 nM) than at higher concentrations (Figure 3B). In particular, the 40 s post-activation increase provided by the presence of 1 nM, 10 nM or 100 nM ACEA over the signal provided by 200 nM ghrelin, was (in percentage), 140 ± 30, 65 ± 30, and 71 ± 15, respectively. Finally, as GHS-R1a and CB₁ receptor agonists lead to activation of the mitogen-activated protein kinase (MAPK) pathway (Mousseaux et al., 2006; Daigle et al., 2008; Navarro et al., 2018b), we tested the properties of the heteromer in the link to the MAPK signaling pathway. Again, the effect of the cannabinoid receptor agonist was suppressed when the two receptors were co-expressed, while the effect of ghrelin was not modified by low doses of ACEA but was diminished when higher doses were used (Figure 3C). In a control assay, it was confirmed that ACEA does increase ERK1/2 phosphorylation, as previously reported (Navarro et al., 2018a, b). These data demonstrate that cannabinoids may regulate GHS-R1a function depending on the concentration and that GHS-R1a expression suppresses cannabinoid receptor-mediated events supposedly by the establishment of heteromeric complexes.

In the 90s different laboratories proved interactions between GPCRs to form heteromeric complexes. These complexes can be detected by energy transfer techniques in heterologous expression systems or by proximity ligation assays (PLA) in natural sources, either primary cultures or tissue sections. An often-found property of a heteromer formed by two GPCRs is that the antagonist of one of the receptors not only blocks the signaling originated at the receptor but also the signaling originated at the partner receptor. Such cross-antagonism, which is due to conformational changes transmitted from one receptor to the other, may serve as a print to detect the heteromer. Another possibility is that coactivation leads to a smaller effect than that obtained upon activating only one receptor of the complex; this phenomenon is known as negative crosstalk. Finally, in some cases, the antagonist of one receptor may restore the signaling via the partner receptor in the heteromer.

To study the effect of antagonists, HEK-293T cells expressing CB₁ and GHS-R1a receptors were pretreated with the selective antagonists before agonist stimulation. As observed in Figure 4A, the CB₁R antagonist, rimonabant, did not block G_i-mediated ghrelin-induced effect, whereas the GHS-R1a antagonist, YIL 781, which completely blocked ghrelin-induced decrease of FK-induced cAMP levels, did not restore the CB₁R-G_i coupling. Accordingly, there was no cross-antagonism in GHS-R1a/G_i-mediated signaling when the CB₁ receptor was blocked by a selective antagonist. Similar were results in Figure 4B, i.e., there was no cross-antagonism in GHS-R1a receptor/G_q-mediated signaling when the CB₁R receptor was blocked by a selective antagonist. In summary, neither cross-antagonism nor restoration of CB₁R-G_i coupling was observed when addressing direct G_i- or direct G_q-induced outputs using selective antagonists.

After demonstrating the functionality of CB₁R-GHS-R1aHet in transfected HEK-293T cells, we moved to a more physiologic environment to look for the expression of the receptor complex. We investigated the expression and function of CB₁R-GHS-R1aHet in primary striatal neurons. First, we analyzed the effect of ACEA and ghrelin on FK-induced cAMP. As observed in Figure 5, ACEA was able to induce a dose-dependent effect. Ghrelin also decreased the FK-induced cAMP levels. Coactivation with ghrelin and 1 nM ACEA resulted in an additive effect, but not at higher ACEA concentrations. Moreover, the effect of ACEA was not affected by the antagonists of the GHS-R1a receptor and, reciprocally, the effect of ghrelin was not counteracted by the antagonists of the CB₁R (Figure 5B). On the other hand, when analyzing ERK1/2 phosphorylation in striatal primary neurons, both ACEA and ghrelin led to a significant increase in phosphorylation levels. In addition, no cross-antagonism was detected in primary neurons pretreated with rimonabant followed by ghrelin stimulation (Figures 5C,D). These results suggest that the proportion of CB₁R-GHS-R1aHets in primary striatal cultures is relatively low and that CB₁ and GHS-R1a receptors may be forming complexes with other GPCRs (see “Discussion” section).

One of the aims of this study was to correlate the expression of CB₁R-GHS-R1a receptor complexes in a situation of unbalanced energy homeostasis. For this purpose, we used primary striatal neurons isolated from fetuses of mothers fed a standard (STD) or high fat (HFD) diet (see “Materials and Methods” section). The expression of CB₁R-GHS-R1aHets was assessed by in situ proximity ligation assay (PLA). An important increase in the expression of CB₁-GHS-R1a complexes was observed in striatal neurons from fetuses of HFD mothers (18 red dots/cell vs. 3 red dots/cell in neurons from fetuses of STD mothers; Figures 6A,B). Moreover, when primary neurons from fetuses of HFD mothers were treated with 1 nM ACEA, a significant decrease in complex expression (12 red dots/cell) was observed; however, a higher concentration of the compound, 100 nM, led to a marked increase in the number of complexes (36 red dots/cell). Remarkably, treatment with 200 nM ghrelin produced a robust increase in CB₁R-GHS-R1a complex expression (>40 red dots/cell; Figure 6C). In summary, the expression of the CB₁R-GHS-R1aHet was higher in the progeny of HFD mothers and may be regulated by ghrelin and, differentially, by cannabinoid concentration.