Cannabinoid regulation of angiotensin II-induced calcium signaling in striatal neurons

Functionality assays in HEK-293T cells expressing cannabinoid CB₁ and/or angiotensin AT₁ receptors

The HEK-293T-cell-based heterologous system was used to express the human versions of the Ang II type 1 (AT₁R) and the cannabinoid type 1 (CB₁R) receptors. First, the G_i-mediated signaling was assayed in cells expressing AT₁Rs and/or CB₁Rs using selective agonists and/or antagonists of the two receptors and treated with forskolin (FK), which activates adenylate cyclase to increase the intracellular levels of the cyclic nucleotide. In cells expressing only one receptor, the selective antagonist of the CB₁R, rimonabant, did not affect signaling of the AT₁R (Supplementary Fig. S1A), and candesartan, the antagonist or the AT₁R, did not affect signaling of the CB₁R (Supplementary Fig. S1B). The cAMP levels determined in FK-treated cells expressing the AT₁R showed G_i coupling due to the decrease produced by the 15-min incubation with Ang II. The effect was prevented by the selective antagonist, candesartan (Fig. 1A). Analogously, the cAMP levels induced by FK in CB₁R-expressing cells were reduced by the selective agonist, arachidonoyl 2′-chloroethylamide (ACEA). Pretreatment with rimonabant, the selective CB₁R antagonist, prevented the effect of ACEA (Fig. 1B). In cells coexpressing the two receptors, the two agonists, Ang II and ACEA, were able to decrease cAMP, suggesting G_i coupling of AT₁R and CB₁R. Simultaneous treatment with the two agonists did not lead to either synergism or additive effect. However, there was a finding, the prevention by rimonabant of the action of Ang II (Fig. 1C), that cannot be explained if the receptors are not interacting (physically or functionally). This cross-antagonism, i.e., the fact that the selective antagonist of one receptor is preventing the action of the agonist of the other receptor is suggestive of direct interaction, i.e., of the formation of AT₁CB₁Hets.

The measurement of the degree of extracellular signal-regulated kinase (ERK) 1/2 phosphorylation in the presence of agonists showed that the two receptors, when expressed individually in HEK-293T cells, are linked to the mitogen-activated protein kinase (MAPK) signaling pathway. The effect was specific as it was blocked by selective antagonists (Fig. 1D, E). In cells coexpressing the two receptors, there was a bidirectional cross-antagonism, i.e., candesartan, the AT₁R antagonist, prevented the action of ACEA, the CB₁R agonist, and rimonabant, the CB₁R antagonist, prevented the action of Ang II on the AT₁R. (Fig. 1F). Again, these findings are consistent with a molecular interaction of the two receptors.

Colocalization and interaction between cannabinoid CB₁ and angiotensin AT₁ receptors

Colocalization in HEK-293T cells expressing the two receptors was assayed by immunocytochemical techniques using receptors fused to Renilla luciferase (Rluc) and/or yellow fluorescent protein (YFP). First, the individual expression of each receptor was checked in cells transfected with cDNA coding for AT₁R fused to YFP or with cDNA coding for CB₁R fused to YFP; the fluorescence due to YFP excitation was observed in the two conditions (Fig. 2A, B). Coexpression was achieved by transfecting with cDNAs coding for AT₁R-Rluc and CB₁R-YFP. AT₁R-Rluc was detected by a mouse monoclonal anti-Rluc antibody and a secondary sulfo-cyanine3 (Cy3)-conjugated anti-mouse IgG antibody; CB₁R-YFP was detected by the fluorescence emitted by YFP upon excitation (see scheme in Fig. 2D). Marked overlap of the signal in the two detection channels suggested a high degree of receptor colocalization (Yellow, Fig. 2C). To check whether colocalization could result from a direct interaction between the two proteins expressed, bioluminescence resonance energy transfer 2 (BRET²) assays were performed. The experiment was performed in HEK-293T cells expressing a constant amount of CB₁R-Rluc and increasing levels of AT₁R-GFP² (see scheme in Fig. 2G). The saturation curve shown in Fig. 2E demonstrates that the two receptors may interact in a heterologous expression system. Negative control using CB₁R-Ruc and angiotensin-converting enzyme 2 (ACE2) fused to GFP² led to a linear relationship between the BRET² signal and the donor/acceptor ratio, indicating no interaction (Fig. 2F).

Alteration of AT₁R-mediated G_q signaling in cells coexpressing AT₁ and CB₁ receptors

Upon discovering that AT₁ and cannabinoid CB₁ receptors interact in HEK-293T cells, we undertook assays to determine the levels of cytosolic Ca²⁺ and to test whether the receptor–receptor interaction regulates the coupling of the AT₁R to the G_q family of transducers. As noted in the “Methods” section, the experimental setup took advantage of a calmodulin-based calcium ion sensor that emits fluorescence when bound to Ca²⁺ and it is excited at the appropriate wavelength.

In cells expressing the AT₁R, but not the cannabinoid receptor, the agonist, Ang II, caused a marked increase in fluorescence, which did not decay after 150 s of measurement (Fig. 3A). The maximal effect Ang II-mediated on calcium release was completely abolished by pretreatment with candesartan, while pretreatment with rimonabant did not alter Ang II-mediated calcium mobilization (Fig. 3D). As expected, ACEA did not produce any signal in cells expressing the CB₁ receptor (Fig. 3B); the fluorescence signal was barely detectable (Fig. 3E). In cells that coexpressed the two receptors, the effect of Ang II was significant, but with different kinetics compared to the signal observed in cells that expressed only AT₁R. In fact, in cotransfected cells, Ang II led to a significant increase in fluorescence that did decay, i.e., a peak of fluorescence was observed (at around 90 s, Fig. 3C). Furthermore, ACEA, which by itself did not lead to any effect, significantly reduced the effect of Ang II also modifying the kinetics; when cells were treated with both ACEA and Ang II the peak of calcium-calmodulin sensor fluorescence was found at <50 s. Finally, while candesartan, the selective AT₁R antagonist, completely prevented the response, rimonabant, the CB₁R antagonist, reduced the Ang II response with a fluorescence peak observed around 40 s (Fig. 3C). In cells that coexpressed the two receptors, the maximal effect of Ang II-mediated on calcium release was significantly reduced in both cells treated with ACEA and in cells pretreated with rimonabant (Fig. 3F). Taken together, the results indicate that CB₁R occupancy by agonists or antagonists reduces the effect of Ang II and modifies the kinetics of cytosolic Ca²⁺ release/reuptake.

Expression of AT₁CB₁Hets and functionality assays in primary striatal neurons

The in situ PLA is instrumental in detecting receptor–receptor complexes in primary cells or tissue sections. The procedure was performed as described in “Methods” section using primary striatal neurons incubated with anti-AT₁R and anti-CB₁R antibodies and, subsequently, with secondary antibodies conjugated to complementary oligonucleotide probes. AT₁-CB₁ receptor complexes (Fig. 4B) are identified by red fluorescence dots (nuclei are labeled in blue by Hoechst 33342) in structures also labeled using an Alexa-488 conjugated anti-NeuN antibody. The negative control was obtained without the anti-mouse PLA probe secondary antibody (Fig. 4A). The graph indicating the number of points per neuron (≈5.7) and the percentage of cells expressing receptor–receptor complexes (>82%) is found in Fig. 4C. Expression of AT₁CB₁Hets was also assessed in structures also labeled using an Alexa-488 conjugated anti-MAP2 antibody. The negative control was obtained without the anti-mouse PLA probe secondary antibody (Fig. 4D). Data in Fig. 4E suggest that the majority of the label is located more in the neuronal soma that in extension. On average 5.7 dots/cell were counted near NeuN⁺ stain (Fig. 4C) while 9.3 dots/cell were counted near MAP2⁺ structures (Fig. 4F); it may be deduced that 39% complexes are in dendrites, and 61% in neuronal somas.

Assays of determination of intracellular cAMP levels were performed in primary neurons of the rodent corpus striatum using agonists and antagonists of CB₁ and AT₁ receptors. The levels of the cyclic nucleotide induced by the treatment with FK were decreased by both, Ang II and ACEA, despite simultaneous treatment with the two agonists did not lead to a greater effect. The results obtained when neurons were pretreated with antagonists were like those found in the heterologous expression system, i.e., the antagonist of the AT₁R prevented the effect of Ang II but not the effect of ACEA, whereas the antagonist of the CB₁R prevented the effect of ACEA, but also the effect of Ang II (Fig. 4G). In ERK1/2 phosphorylation assays the results were also like those described for HEK-293T cells coexpressing the two receptors. Ang II and ACEA increased the phosphorylation levels and both the antagonists, candesartan and rimonabant, prevented the effect of each agonist. The only difference between results in primary neurons and cotransfected HEK-293T cells is that the simultaneous treatment in striatal neurons with the two receptor agonists did not result in any measurable signal (Fig. 4H). In summary, the results show that these primary neurons of the striatum express AT₁CB₁Hets and that the occupancy of the CB₁R by agonists and/or antagonists affects the signaling mediated by the AT₁R.

Cytosolic calcium levels in primary striatal neurons treated with receptor agonists

Primary striatal cultures loaded with Fluo-4 NW were treated with receptor ligands to understand whether CB₁R activation alters the AT₁R-mediated response that leads to increases in cytoplasmic Ca²⁺ concentration. Real-time measurements were performed taking images every 600 ms in a confocal microscope and the results are displayed in Fig. 5. Neurons within a given region of interest (ROI) responded to 100 nM Ang II with a rapid spike after which fluorescence slowly decreased; after 8 min of recording, the fluorescence did not return to baseline (Fig. 5A, B). The signal was completely abolished by the selective AT₁R antagonist, candesartan; in the presence of this antagonist, Ang II failed to produce a significant level of fluorescence (∆F/F₀ < 0.1, data not shown). When Ang II was administered with ACEA, the CB₁R agonist, the signal in all neurons within the ROI was much lower than that obtained in the absence of the cannabinoid (Fig. 5C, D). ACEA reduced the ∆F/F₀ peak by approximately half (Fig. 5A, C) and the fluorescence reached the baseline value in <90 s (Fig. 5C). These data confirm that cannabinoid modulation of AT₁R-mediated calcium signaling occurs not only in transfected HEK-293T cells but also in striatal neurons.

Expression of AT₁CB₁Hets in the striatum of a parkinsonian rat model

The in situ PLA was used to detect receptor complexes in sections from the striatum of the 6-OHDA hemilesioned rats. Samples were collected from four animal groups: (i) control (non-lesioned), (ii) lesioned, (iii) lesioned levodopa-treated non-dyskinetic, and (iv) lesioned levodopa-treated dyskinetic. According to the tyrosine hydroxylase (TH) labeling displayed in Fig. 6A–D, the lesion was virtually complete (>95%) in the right striatum of all animal groups. The loss of dopaminergic terminals in the lesioned hemisphere is not recovered with levodopa treatment. Dyskinesia does not occur in all animals treated with levodopa and when it does occur, serotonergic pathways are likely to be involved^35,36. The amount of AT₁R and CB₁R complexes was markedly high in striatal NeuN⁺ neurons from non-lesioned animals and less abundant in the other animal groups (Fig. 6E–H). The label was negligible in the negative control made by omitting one of the two primary antibodies (Fig. 6I). Samples from lesioned animals showed significantly fewer AT₁CB₁Hets red dots per NeuN⁺ cell (>95% in non-lesioned versus ≈45% in lesioned). Striatal neurons expressing receptor complexes were ≈23% in levodopa-treated non-dyskinetic, and ≈70% in levodopa-treated non-dyskinetic. The number of complexes per neuron decreased in lesioned animals; interestingly, the number of complexes was further reduced in levodopa-treated non-dyskinetic animals whereas it was slightly recovered in dyskinetic animals (Fig. 6J). In summary, expression of AT₁CB₁Hets in striatal neurons varies depending on both the lesion and the consequences of levodopa treatment of lesioned rats.

Reagents

Forskolin (FK), Arachidonoyl 2′-chloroethylamide (ACEA), angiotensin II (Ang II), Candesartan, Rimonabant (SR141716), PolyEthylenImine (PEI), 6-hydroxy dopamine (6-OHDA), and Hoechst 33342 were purchased from Merck (St Louis, MO, USA). Concentrated (10 mM) stock solutions of agonists/antagonists prepared in ethanol (ACEA), DMSO (Candesartan and Rimonabant) or water (Ang II) were stored at −20 °C; they were thawed and diluted before use.

Cell culture and transfection

HEK-293T cells, batch 70022180, were acquired from the American Type Culture Collection (ATCC). Cells were amplified and frozen in liquid nitrogen in several aliquots. Cells from each aliquot were used until passage 18. HEK-293T cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Paisley, Scotland, UK) supplemented with 2 mM l-glutamine, 100 μg/mL sodium pyruvate, 100 U/mL penicillin/streptomycin, MEM non-essential amino acids solution (1/100) and 5% (v/v) heat-inactivated fetal bovine serum (all supplements were from Invitrogen, Paisley, Scotland, UK) and maintained at 37 °C in a humid atmosphere of 5% CO₂.

Primary striatal neurons were obtained from 19-day mouse embryos as described in refs. ^78,79. Briefly, striata were dissected and digested in 0.25% trypsin for 15 min at 37 °C. Trypsinization was stopped by repeated washes with Hank’s Buffered Saline Solution (HBSS, Gibco). Cells were brought to a single-cell suspension by repeated pipetting followed by passage through a 100 µm-pore mesh. Cells were then resuspended in supplemented DMEM and seeded at a density of 3.5 × 10⁵ cells/mL in 6-well plates or 96-well plates for functional assays and in 12-well plates for immunocytochemistry or PLA assays. The day after, the medium was replaced by neurobasal medium supplemented with 2 mM l-glutamine, 100 U/mL penicillin/streptomycin, and 2% (v/v) B27 (Gibco) and cells were cultured for 12 days. Cultures were maintained at 37 °C in a 5% CO₂ humid atmosphere. Immunodetection of the NeuN marker showed that preparations contained >95% neurons.

Cell transfection

HEK-293T cells were transiently transfected with the corresponding cDNA(s) by the PEI method. Briefly, cDNA diluted in 150 mM NaCl was mixed (10 min) with PEI (5.5 mM in nitrogen residues) and prepared in 150 mM NaCl for 10 min. The cDNA–PEI complexes were placed in contact with HEK-293T cells and were incubated for 4 h in a serum-free medium. Then, the medium was replaced by a fresh supplemented culture medium and cells were maintained at 37 °C in a 5% CO₂ humid atmosphere. Forty-eight hours after transfection, cells were washed, detached, and resuspended in assay buffer/medium for further analysis.

Expression vectors

pcDNA3.1-based plasmids encoding for CB₁R, CB₁R-Rluc, CB₁R-YFP, AT₁R, AT₁R-Rluc, AT₁R-YFP, AT₁R-GFP² and ACE2-GFP² fusion proteins were used. Plasmids encoding fusion proteins were generated by subcloning the coding region of each receptor to be in-frame with restriction sites of pRluc-N1 (Clontech, Heidelberg, Germany), pEYFP-N1 (PerkinElmer, Waltham, MA, USA) and pEGFP²-N1 (Clontech) vectors to provide plasmids that express the receptors with Rluc, YFP or GFP² proteins fused on the C-terminal end.

6-OHDA lesion and behavior assessment

Protocols adhered to EU directives (2010/63/EU and 86/609/CEE) and were approved by the ad hoc ethical committee. Male Wistar rats were used and the experimental procedures were similar to those elsewhere described^66,80,81. Lesioning protocol was performed in 8-week-old rats.

Animals were anesthetized with 1% ketamine (75 mg/kg) and 2% xylazine (10 mg/kg). Twelve micrograms of 6-OHDA in 4 μL of saline containing 0.2% ascorbic acid was injected in the right medial forebrain bundle. Rotational behavior was used to confirm the lesion. Non-lesioned animals were subjected to a similar procedure but using saline instead of the neurotoxin.

A bank of 8 automated rotometer bowls (Rota-count 8, Columbus Instruments, Columbus, OH, USA) was used to monitor full (360°) body turns in either direction (90 min after i.p. injection of 2.5 mg/kg dextroamphetamine dissolved in saline). Six full body turns/min ipsilateral to the lesion corresponds to >90% depletion of dopamine fibers in the striatum⁸²; however dopaminergic lesions were confirmed with the cylinder test, i.e., by analyzing spontaneous use of the forelimb⁸³. A glass cylinder (20 cm in diameter) was used and the number of right or left forepaw contacts was blindly assessed. Scoring left (impaired) touches as a percentage of total touches, control animals would receive a 50% unbiased score, whereas lesioned animals that were selected for further experimentation showed <20% scores.

Levodopa-induced dyskinesia

Sixteen lesioned animals were chronically treated for 3 weeks with levodopa methyl ester (6 mg/kg) and benserazide (10 mg/kg), the administration route was subcutaneous. The treatment reliably induces dyskinetic movements in some rats. Abnormal involuntary movements that appeared in some of the treated animals were assessed by the elsewhere described dyskinesia scale⁸⁴. The severity of each limb, orolingual, and axial involuntary movements was assessed using scores from 0 to 4 (4 = continuous, 3 = continuous but interrupted by strong sensory stimuli; 2 = frequent, present >50% of the time, and 1 = occasional, present <50% of the time. Dyskinetic rats were considered when animals displayed a score ≥2 per monitoring period on at least two types of abnormal involuntary movements). Non-dyskinetic rats exhibited no levodopa-induced abnormal involuntary movements or mild/occasional ones. Animals with low scores, either dyskinetic or non-dyskinetic were excluded. In summary, four groups of animals were analyzed: (i) non-lesioned, (ii) lesioned, treated with vehicle; (iii) lesioned and dyskinetic and (iv) lesioned and non-dyskinetic despite levodopa treatment. TH immunostaining was performed in sections taken postmortem; lesioned animals showed >95% nigral dopaminergic denervation. PLA were performed in different fields of striatal sections. Images were captured in delimitation areas; the striatum was delimited using a bright field.

cAMP determination

Determination of intracellular 3′,5′-monophosphate (cAMP) levels was performed in primary neurons or HEK-293T cells transfected with the cDNA for AT₁R (1.5 μg), the cDNA for the CB₁R (1.5 μg) or both. The intracellular concentration of this first messenger was determined using the Lance Ultra cAMP kit (PerkinElmer). The method consists of a time-resolved fluorescence resonance energy transfer (TR-FRET) immunoassay in which endogenous cAMP competes with europium (Eu) chelate-labeled cAMP tracer for binding sites on a cAMP-specific antibody labeled with the ULightTM dye. Light pulses at 320 nm excite the Eu of the tracer. The energy emitted by the excited Eu is transferred by FRET to ULight molecules on the antibodies, which in turn emit light at 665 nm. In the absence of cAMP, maximal TR-FRET signal is achieved; when an agonist leads to an increase in cytosolic cAMP levels the competition between the unlabeled and the Eu-labeled cAMP species leads to a decrease in the TR-FRET signal, the emission fluorescence remains unmodified when equilibrium is achieved. cAMP concentrations per cell or mg protein were determined using a standard curve using pure unlabeled cAMP. Residual energy from the Eu chelate will produce light at 615 nm. The dynamic range covers from 10⁻¹⁰ to 10⁻⁸ M cAMP concentrations.

Neurons or transfected cells were grown in 6-well plates. Forty-eight hours post-transfection, the medium was replaced by serum-free medium (DMEM). Two hours later cells were detached, isolated by centrifugation (5 min at 1500 rpm) and resuspended in ^cAMPmedium, which consisted of DMEM containing HEPES (5 mM, pH 7.4), zardaverine (50 µM), a phosphodiesterase inhibitor to prevent degradation of cAMP, and 0.1% bovine serum albumin (BSA). Determination was performed in 384-well plates (PerkinElmer) using 4000 cells/well.

Cells were incubated for 15 min with 2 µL of ^cAMPmedium (to determine the basal levels of cAMP) or with 2 µL of ligands prepared in ^cAMPmedium. When indicated antagonists (1 µM candesartan or 1 µM rimonabant) were added to cells for 15 min before addition of agonists. Fifteen minutes after addition of agonists, cells were treated for 15 min with 500 nM FK. cAMP-Europium (cAMP-Eu) (5 µL) and fluorophore-containing ULight™ antibody (5 µL) were then added. Incubation was prolonged for 1 h at 25 °C and the PHERAstar Flagship reader equipped with an HTRF optical module (BMG Lab Technologies, Offenburg, Germany) was used for measuring the 665/620 nm ratio.

ERK1/2 phosphorylation assay

The link to the MAPK signaling pathway was assessed by a homogeneous method, which avoids immunoblotting and directly measures levels of phosphorylated proteins in a cell-based format (AlphaScreen^® SureFire^® kit; PerkinElmer). The kit contains an antibody that is specific for a phospho-epitope and another that is specific for another region (distal to the phospho-epitope) of extracellular signal-regulated kinase 1 and 2 (ERK1/2). One of the antibodies is biotinylated and binds to streptavidin-conjugated donor beads, and the second binds to protein A Sepharose beads that contain an acceptor. Only immuno-complexes that contain both antibodies can bind both beads and, therefore, donor-to-acceptor energy transfer can occur. Emission fluorescence is measured using an EnSpire Multimode Plate Reader (PerkinElmer). The kit uses the AlphaScreen^® technology that is based on the emission of singlet oxygen by the donor beads (which contain phthalo cyanine, excited by the red light at 680 nm), which can diffuse to reach the acceptor beads where the energy of singlet oxygen is used by a cascade (thioxene–anthracene–rubrene), leading to the emission of light at a shorter wavelength (520–620 nm range) than the excitation light. Transparent Biocat Poly-D Lysine 96-well plates (Deltalab) were used for cell culture. The assay was performed either in neurons or in transfected HEK-293T cells.

Complete medium was replaced by serum-free medium, and cells were treated or not for 10 min with the antagonists, candesartan or rimonabant followed by 7-min treatment with the selective agonists Ang II and/or ACEA. Cells were then washed twice with cold PBS before the addition of lysis buffer (15 min at 25 °C, 30 µL/well; PerkinElmer) and then incubated under agitation in a Polymax 2040 (Heidolph Instruments, Schwabach, Germany). Ten microliters of each cell lysate were transferred to 384-well microplates (white ProxiPlate; PerkinElmer). Five microliters per well of beads containing the acceptor were added and allowed to incubate in the dark for 2 h at 25 °C. Finally, 5 μL/well of beads containing the donor were added and plates were protected from the light. After 2 h incubation fluorescence was determined. The effect of ligands was given in percentage respect to the reference value (basal). The value achieved in the absence of any treatment (30 µL DMEM) was taken as a reference (basal = 100).

Immunocytochemistry

HEK-293T cells were seeded on glass coverslips in 12-well plates. Twenty-four hours after, cells were transfected with AT₁R-YFP cDNA (1 μg), CB₁R-YFP cDNA (1 μg), and/or AT₁R-Rluc (1 µg). Forty-eight hours after, cells were fixed in 4% paraformaldehyde for 15 min and washed twice with PBS containing 20 mM glycine before permeabilization with PBS-glycine containing 0.2% Triton X-100 (5 min incubation). Cells were blocked during 1 h with PBS containing 1% BSA. HEK-293T cells were labeled with a mouse anti-Rluc antibody (1/100; Millipore, Darmstadt, Germany) and subsequently treated with Cy3-conjugated anti-mouse (1/200; Jack- son ImmunoResearch (red)) antibody (1 h each). The AT₁R-YFP and CB₁R expression was detected by the YFP’s own fluorescence. Nuclei were stained with Hoechst 33342 (1/100 from stock 1 mg/mL; Sigma-Aldrich). Samples were washed several times and mounted with Immu-Mount™ (ref. 9990402). Images were obtained in a Zeiss LSM 880 confocal microscope (ZEISS, Germany) with the 63× oil objectives.

Bioluminescence resonance energy transfer (BRET²) assay

HEK-293T cells were transiently cotransfected with a constant amount of cDNA encoding for CB₁R-Rluc (0.2 μg) and with increasing amounts of cDNA corresponding to either AT₁R-GFP² (0.5–4.5 μg) or with ACE2-GFP² (0.25–2 μg), as a negative control for BRET² assay. To assess cell amount, protein concentration was determined using a Bradford assay kit (Bio-Rad, Munich, Germany) using BSA dilutions as standards. To quantify protein-GFP² expression, fluorescence was read in a Mithras LB 940 equipped with a high-energy xenon flash lamp, using a bandwidth excitation filter at 395 nm. For BRET measurements, the equivalent of 20 μg protein cell suspension was distributed in 96-well microplates (white plates, Porvair, Leatherhead, UK) and DeepBlueC (5 µM) was added (PJK GMBH, Kleinblittersdorf, Germany). Thirty seconds after DeepBlueC 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 395 nm and the long-wavelength filter at 515 nm. To quantify receptor-Rluc expression, luminescence readings were collected 10 min after the addition of DeepBlueC (5 µM) (Molecular Probes, Eugene, OR) using a Mithras LB 940.

Determination of cytoplasmic calcium ion (Ca²⁺) levels

HEK-293T transfected with the cDNAs for AT₁R and/or CB₁R were also transfected with the cDNA for the GCaMP6 calcium sensor (1 μg)⁸⁵. Forty-eight hours post-transfection, HEK-293T cells were detached using Mg²⁺-free Locke’s buffer (154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO₃, 2.3 mM CaCl₂, 5.6 mM glucose, 5 mM HEPES, 10 μM glycine, pH 7.4), centrifuged for 5 min at 3200 rpm and resuspended in the same buffer. Protein concentration was quantified by using the Bradford assay kit (Bio-Rad, Munich, Germany). To measure Ca²⁺ mobilization, cells (40 µg of protein) were distributed in 96-well microplates (black plates with a transparent bottom; Porvair, Leatherhead, UK) and were preincubated for 10 min with antagonists, when indicated, before adding AT₁R agonist (Ang II, 100 nM) and/or CB₁R agonist (ACEA, 100 nM) right before readings. Fluorescence emission intensity due to GCaMP6 was recorded at 515 nm upon excitation at 488 nm on the EnSpire® Multimode Plate Reader for 150 s every 5 s at 100 flashes per well.

For real-time detection of cytoplasmic calcium ion (Ca²⁺) levels in striatal neurons, cells were seeded in Lab-Tek® Chambered #1.0 Borosilicate cover glass devices (ref. 155411, ThermoFisher), coated with poly-d-lysine (ref. 16021412, Gibco), and then incubated for 15 days. After this period, neurons were washed three times with HBSS. Then, neurons were incubated with the Fluo-4 NW Ca²⁺ indicator (ref. 10266762, Fisher Scientific) for 1 h at 37 °C. Eight-well chambers were then observed under a Zeiss 880 confocal microscope (Carl Zeiss, Oberkochen, Germany) with the 40× oil immersion objective. The excitation wavelength was 488 nm, and the emission wavelength was 516 nm. To establish baseline fluorescence (F₀) images were taken for 15 s before adding ligands. Next, the AT₁R agonist (Ang II, 100 nM) and/or the CB₁R agonist (ACEA, 100 nM) were added, and images were captured at the maximum allowed speed (600 ms) for a total period of 8 min. Background fluorescence (F_B) was estimated by measuring fluorescence in cell-free spaces. The auto-focus option was activated in each observation to avoid blurring after applying the treatment. ΔF/F₀ ratio was used for quantification. This normalization helps to report variations in the initial fluorescence intensity across different conditions, allowing for a more accurate comparison of the relative changes in fluorescence.

Immunohistochemistry

TH immunohistochemistry analysis was used to verify the extent of the lesion. The sections were incubated for 1 h in 10% normal swine serum with 0.25% Triton X-100 in 20 mM potassium phosphate-buffered saline, containing 1% bovine serum albumin (KPBS-BSA), and afterward incubated overnight (at 4 °C) with anti-TH as the dopaminergic marker (mouse monoclonal anti-TH, Sigma-Aldrich Cat# T2928, RRID: AB_477569; 1:10,000). Sections were then incubated with the corresponding biotinylated secondary antibody (horse anti-mouse, Vector Laboratories, Inc., Newark, CA, USA; Cat# BA-2001, RRID: AB_2336180; 1:200) for 60 min, and subsequently with an avidin–biotin–peroxidase complex (ABC, 1:100, Vector) for 90 min. Finally, sections were revealed with 0.04% hydrogen peroxide and 0.05% 3-3′diaminobenzidine (DAB, D5637, Sigma-Aldrich). Control sections in which the primary antibody was omitted were immune-negative for TH.

Proximity ligation assay (PLA)

The interaction between AT₁ and CB₁ receptors in striatal neurons and brain sections was detected using the Duolink in situ PLA detection Kit (OLink; Bioscience, Uppsala, Sweden ref. DUO92008) following the instructions of the supplier. Primary neurons grown on glass coverslips were fixed in paraformaldehyde (4%) for 15 min, washed with PBS containing glycine (20 mM) to quench the aldehyde groups and permeabilized with the same buffer containing Triton X-100 (0.05%, 20 min for primary neurons and 30 min for brain sections) and successively washed with PBS. Then, samples were incubated (1 h) at 37 °C with a blocking solution (ref. DUO82007, Sigma-Aldrich) in a pre-heated humidity chamber. After overnight incubation with the antibody diluent medium having a mixture of equal amounts of rabbit anti-AT₁R (ab124734, Abcam) (1/100) and mouse anti-CB₁R (sc-518,035, Santa Cruz) antibodies (1/100), ligation and amplification were conducted as indicated by the supplier. Neurons were identified by staining with the Alexa Fluor^®488 conjugated anti-NeuN antibody (ab190195, 1/200) or Alexa Fluor^®488 conjugated anti-MAP2 antibody (MAB3418X, 1/200). Samples were mounted using the mounting medium with Hoechst 33342 (1/100) 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 405, 488, and 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 versus total cells was determined, and Student’s t-test was used to compare the values (red dots/cell). The percentage of cells expressing at least one red dot is indicated, along with their S.E.M., at the top of each bar of the graphs. Images of neurons were taken under bright-field microscopy to observe the proper morphology of the cells.

Cellprofiller® pipeline design

First, the ColorToGray module was used to convert the color images to grayscale for individual channels, Hoechst (blue), PLA dots (red) and NeuN or MAP2 stain (green). This was followed by the application of a GaussianFilter to smooth the images and reduce noise. Threshold module was employed to segment the NeuN or MAP2 (green) channel and PLA dots (red) channel images, distinguishing the signal from the background. The IdentifyPrimaryObjects module was used to identify primary objects in the NeuN or MAP2 (green) channel, representing the neuronal structures. Subsequently, the IdentifySecondaryObjects module was applied to identify secondary objects in the PLA (red) channel, using the primary NeuN or MAP2 objects as a reference.

The ShrinkToObjectCenters module was then applied to refine the positions by shrinking the objects to their centers, followed by the ExpandOrShrinkObjects module to adjust the size of the objects if needed to improve accuracy. The OverlayOutlines module was used to overlay the outlines of identified objects on the original images. Finally, the SaveImages module was used to save the processed images, and the ExportToSpreadsheet module was utilized to export the data for further analysis.

Data handling and statistical analysis

GraphPad Prism 10 software (San Diego, CA, USA) was used for data analysis. One-way ANOVA followed by post-hoc Bonferroni’s test, post-hoc Tukey’s test or Dunnett’s test were used when multiple comparison analysis. BRET parameters were calculated using an ad hoc online tool⁸⁶. In PLA assays the number of red dots per cell was determined using with a pipeline specifically designed for the Cellprofiller® software. Two-tailed Student’s t-test was used for PLA statistical analysis. Statistical analysis was undertaken only when each group size was at least n = 5, n being the number of independent variables (technical replicates were not treated as independent variables). Differences were considered significant when P ≤ 0.05.