BRET-Based Biosensors to Measure Agonist Efficacies in Histamine H1 Receptor-Mediated G Protein Activation, Signaling and Interactions with GRKs and β-Arrestins

The histamine H1 receptor (H1R) is a G protein-coupled receptor (GPCR) and plays a key role in allergic reactions upon activation by histamine which is locally released from mast cells and basophils. Consequently, H1R is a well-established therapeutic target for antihistamines that relieve allergy symptoms. H1R signals via heterotrimeric Gq proteins and is phosphorylated by GPCR kinase (GRK) subtypes 2, 5, and 6, consequently facilitating the subsequent recruitment of β-arrestin1 and/or 2. Stimulation of a GPCR with structurally different agonists can result in preferential engagement of one or more of these intracellular signaling molecules. To evaluate this so-called biased agonism for H1R, bioluminescence resonance energy transfer (BRET)-based biosensors were applied to measure H1R signaling through heterotrimeric Gq proteins, second messengers (inositol 1,4,5-triphosphate and Ca2+), and receptor-protein interactions (GRKs and β-arrestins) in response to histamine, 2-phenylhistamines, and histaprodifens in a similar cellular background. Although differences in efficacy were observed for these agonists between some functional readouts as compared to reference agonist histamine, subsequent data analysis using an operational model of agonism revealed only signaling bias of the agonist Br-phHA-HA in recruiting β-arrestin2 to H1R over Gq biosensor activation.


Introduction
The histamine H 1 receptor (H 1 R) is a membrane-associated G protein-coupled receptor (GPCR) that is ubiquitously expressed on vascular endothelial, smooth muscle, immune, and brain cells and involved in, for example, acute allergic reactions, uterine contraction in preterm labor, and awakening in response to the biogenic amine histamine [1,2]. Since the 1940s, H 1 R became a blockbuster drug target for the treatment of allergic responses by two generations of antihistamines, with the first-generation of these antagonists also being sedative and reducing motion sickness by crossing the blood-brain barrier and acting on H 1 R in the brain [3]. H 1 R signals primarily through heterotrimeric G q/11 proteins to activate the phospholipase C signaling cascade resulting in the conversion of phosphatidylinositol 4,5-bisphosphate into inositol 1,4,5-triphosphate (InsP 3 ) and 1,2-diacylglycerol. These socalled second messengers, in turn, increase intracellular Ca 2+ levels and protein kinase C activity, resulting in various downstream cellular responses [2]. GPCR-mediated G protein activation is attenuated by phosphorylation of the receptor at intracellular serine and/or threonine residues, resulting in the recruitment of β-arrestin1 and/or 2 that sterically hinder further G protein coupling and facilitate receptor internalization [4,5]. Indeed, H 1 R recruits

BRET-Based Detection of G q Activation and Signaling by H 1 R
The H 1 R signals predominantly via heterotrimeric G q/11 proteins to stimulate phospholipase C activity, resulting in the production of InsP 3 and a subsequent increase of intracellular Ca 2+ levels [2]. To monitor these signaling events in real time we used (1) a recently developed tricistronic BRET-based G q activation sensor plasmid that measures the dissociation of Gβ 3 Gγ 9 -cpVenus heterodimer from Gα q -Nanoluc (Nluc) luciferase upon receptor activation [22]; and (2) intramolecular BRET-based InsP 3 and Ca 2+ conformation biosensors consisting of the InsP 3 -binding domain of the human type-I InsP 3 receptor with a R 265 K mutation and a modified Cameleon D3 sensor containing the MLCK calmodulin binding peptide 13 and the D3 variant of calmodulin, respectively [23]. Stimulation of H 1 R with 10 µM histamine (HA) results in rapid activation of heterotrimeric G q proteins reaching a maximum steady-state ∆BRET response within 2 min that remained constant for at least 30 min ( Figure 1A). Similar rapid G q activation by H 1 R in response to histamine was previously reported using a fluorescence resonance energy transfer (FRET)-based Gα q Gβ 1 Gγ 2 sensor in HeLa cells [24,25]. In contrast, recruitment of the engineered mini-G q protein to H 1 R was considerably slower as measured using split-NLuc luciferase complementation assay in HEK293T [26], which might be the consequence of the deletion of the membrane anchors and Gβγ-binding interface in its N-terminus [27]. H1R activates heterotrimeric Gq proteins, InsP3 production and Ca 2+ release in response to histamine (HA). BRET measurements in HEK293T cells transiently co-expressing H1R (0.5-1 µg plasmid/dish) in combination with the heterotrimeric Gq activation sensor (2.5 µg plasmid/dish) (A,B), InsP3 (C,D) or Ca 2+ (E,F) sensor (4.0-4.5 µg plasmid/dish) in real time upon stimulation with 10 µM histamine (A,C,E) or after 20-min stimulation with increasing concentrations histamine (B,D,F). InsP3 and Ca 2+ cells were first stimulated with vehicle (VEH) or histamine (HA) at t = 0 min followed by a second injection of vehicle (VEH) or 10 µM mepyramine (MEP) at t = 10 min, indicated by the numbered arrows 1 and 2, respectively (C-E). Data are shown as mean ± SD from at least three independent experiments performed in triplicate.
The potency (pEC50 = 7.2 ± 0.1; mean ± SD; n = 4) of histamine to activate heterotrimeric Gq proteins in our BRET assay is approximately 4-fold lower than previously observed in a G protein-based transforming growth factor-α shedding assay ( Figure 1B) [28]. Surprisingly, a 50-fold lower histamine potency was reported in essentially a similar BRET readout between Gαq-Renilla luciferase 8 (Rluc8) and the Gβ3Gγ9-green fluorescent InsP 3 and Ca 2+ cells were first stimulated with vehicle (VEH) or histamine (HA) at t = 0 min followed by a second injection of vehicle (VEH) or 10 µM mepyramine (MEP) at t = 10 min, indicated by the numbered arrows 1 and 2, respectively (C-E). Data are shown as mean ± SD from at least three independent experiments performed in triplicate.
The potency (pEC 50 = 7.2 ± 0.1; mean ± SD; n = 4) of histamine to activate heterotrimeric G q proteins in our BRET assay is approximately 4-fold lower than previously observed in a G protein-based transforming growth factor-α shedding assay ( Figure 1B) [28]. Surprisingly, a 50-fold lower histamine potency was reported in essentially a similar BRET readout between Gα q -Renilla luciferase 8 (Rluc8) and the Gβ 3 Gγ 9 -green fluorescent protein 2 (GFP2) heterodimer (called TRUPATH) in H 1 R expressing HEK293T cells [29], which might be the consequence to different protein expression levels by using a three plasmidsbased sensor setup as compared to the single tricistronic biosensor plasmid in the current study [22]. The potency of histamine to induced mini-G q protein to the H 1 R was 10-fold lower as compared our BRET-based G q activation sensor [26], which might be related to the one-to-one protein-protein interaction in the mini-G q assay as opposite to possible activation of multiple heterotrimeric G q proteins by a histamine-bound H 1 R in our BRET-based activation assay (i.e., signal amplification leading to receptor reserve).
Stimulation of H 1 R-expressing HEK293T cells with 10 µM histamine also rapidly increased InsP 3 and Ca 2+ levels to steady-state levels within 2 min as measured by ∆BRET changes of the InsP 3 and Ca 2+ biosensors, respectively, and this could be reversed by 10 µM H 1 R-antagonist mepyramine ( Figure 1C,E). Similar agonist-induced time traces were observed for both these BRET-based InsP 3 and Ca 2+ biosensors in HEK293T cells expressing the angiotensin II receptor type 1 or muscarinic acetylcholine receptor M 3 [23].

BRET-Based Detection of β-Arrestin1/2 Recruitment to H 1 R
GPCR signaling through heterotrimeric G proteins is generally modulated by the recruitment of β-arrestin1 and/or β-arrestin2 upon phosphorylation of the intracellular GPCR C-tail, resulting in steric hindrance of G protein coupling and subsequent internalization of the GPCR-β-arrestin complex by scaffolding clathrin and the AP2 adaptor [4,5].

BRET-Based Detection of GRKs Interaction with H 1 R
To evaluate which ubiquitously expressed GRK subtypes might potentially phosphorylate the histamine-activated H 1 R and consequently promote β-arrestin1/2 coupling, a BRET assay was used to measure the interaction of H 1 R with GRK2, GRK3, GRK5, and GRK6, as previously reported for the histamine H 4 receptor (H 4 R) and the atypical chemokine receptor 3 (ACKR3) [33,39]. Histamine (10 µM) induced a very rapid increase in BRET between H 1 R-Rluc8 and GRK2-mVenus or GRK3-mVenus that peaked within 2 min and subsequently decreased within 20 min to an elevated steady-state level as compared to vehicle-stimulated cells ( Figure 3A). Hence, GRK2/3 interaction with activated H 1 R indeed precedes the recruitment of β-arrestin1/2 in time, which was also previously reported for agonist-stimulated H 4 R, atypical chemokine receptor 4 (ACKR4) oxytocin, and µ-opioid receptor [33,34,41,42], whereas GRK2/3 and β-arrestin1/2 were recruited with comparable kinetics to ACKR3 in response to chemokine CXCL12 stimulation [39]. On the contrary, histamine (10 µM) stimulation decreased BRET between H 1 R-Rluc8 and GRK5-mVenus or GRK6-mVenus with slower kinetics as compared to the increased GRK2/3 interaction to this receptor, but comparable rates as the recruitment of β-arrestin1/2 ( Figure 3A). The histamine-induced BRET decrease suggests that GRK5 and GRK6 are initially in proximity with H 1 R and dissociate upon receptor activation, but the observed response may also reflect a change in conformational orientation between BRET donor and acceptor. In line, reduced BRET between agonist-activated receptor and GRK5 and/or GRK6 has been previously observed for H 4 R, β 2 -adrenergic receptor, bile acid receptor TGR5, neurokinin-1 receptor, and protease-activated receptor 2 [33,[43][44][45][46]. In contrast, stimulation of ACKR3 with CXCL12 increased the BRET between this receptor and GRK5 [39]. No agonist-induced changes in BRET signal were observed for GRK6 with ACKR3 nor with bile acid receptor TGR5, and neither for GRK5/6 with ACKR4 [39,42,45].

H 1 R Internalization Requires GRK2/3 and β-Arrestin1/2
H 1 R is internalized upon stimulation with histamine with comparable kinetics as the recruitment of β-arrestin1/2, as measured by the increased BRET between H 1 R-Rluc8 and the early endosome marker Venus-Rab5a ( Figure 4A). Pretreatment of HEK293T cells with UBO-QIC, cmpd101, or siRNAs that decrease β-arrestin1/2 expression by 60% (Supplementary Figure S3), revealed that internalization of H 1 R is independent of G q activation but requires GRK2/3 and β-arrestins ( Figure 4B). Co-expression of a dominantnegative dynamin K 44 A mutant or hypertonic conditions using 0.16 µM sucrose completely abolished histamine-induced H 1 R translocation towards the early endosomes, indicating that H 1 R internalization is clathrin-dependent as previously reported in transfected Chinese hamster ovary (CHO) cells [48].

Histamine H 1 R Agonists Display Distinct Efficacies in BRET-Based Responses in HEK293T Cells
The aforementioned BRET-based biosensors were then used to measure H 1 R activation by a selection of agonists (Supplementary Figure S4). Similar to reference agonist histamine, all tested agonists displayed higher potencies to activate G q and in particular the increase in InsP 3 /Ca 2+ levels, as compared to modulating H 1 R protein-protein interactions with GRK2, GRK3, GRK5, GRK6, β-arrestin1, and β-arrestin2 (Figures 5 and 6A; Supplementary Table S1). Indeed, signal amplification has been previously reported for H 1 R-mediated G protein signaling but was not observed for one-to-one protein-protein interactions such as β-arrestin2 coupling [6]. The tested H 1 R agonists display comparable potencies (0.99-to 3.2-fold difference in EC 50 values) between InsP 3 and Ca 2+ responses, whereas 5-to 37fold lower potencies were observed for the activation of heterotrimeric G q protein biosensor. However, Br-phHA had only a 1.7-and 5.4-fold lower potency to activate G q as compared InsP 3 and Ca 2+ signaling, respectively ( Figure 6A; Supplementary Table S1). Interestingly, the difference in potency between GRK2/3 recruitment and G q activation was smaller for Br-phHA, Br-phHA-HA, CF3-phHA, HP, and HPHA (0.6-to 2.5-fold) as compared to histamine (18-and 21-fold, respectively) and CF3-phHA-HA (85-and 36-fold, respectively) ( Figure 6A; Supplementary Table S1).
The intrinsic activities of Br-phHA and CF3-phHA to activate G q activation were 0.7 ± 0.0 and 0.8 ± 0.0 (mean ± SD), respectively ( Figure 6B; Supplementary Table S2), which were slightly higher than the α values that were previously observed in the steady-state GTPase activity assay (0.62 and 0.61, respectively) or CRE-driven reporter gene activity (0.56 and 0.62, respectively) [8,51]. As with the reference agonist histamine, both Br-phHA-HA and CF3-phHA-HA acted as full agonists in all used H 1 R assays, whereas Br-phHA, CF3-phHA, HP, and HPHA displayed clear partial agonism in GRK2/3 and β-arrestin1/2 recruitment (Figures 5 and 6B; Supplementary Table S2). In line, partial agonism was previously reported for Br-phHA, CF3-phHA, and HP in a luciferase-complementation-based assay to measure β-arrestin1/2 recruitment to H 1 R [8]. Interestingly, Br-phHA, CF3-phHA, HP, and HPHA might display a higher intrinsic activity to modulate the interaction of H 1 R with GRK5/6 as compared to GRK2/3 and β-arrestin1/2 (Figures 5 and 6B; Supplementary Table S2), although the low potencies of Br-phHA, CF3-phHA, and HP in GRK5/6 BRET assay did not allow accurate determination of their maximum responses. The presented data in Figure 6A,B suggest signaling bias by some of the H 1 R agonists. For example, relative to reference agonist histamine, full agonist CF3-phHA-HA displays slightly higher potencies (2.4-to 4.0-fold) towards G q , InsP 3 , Ca 2+ , and β-arrestin1/2, as compared to the GRK2/3/5/6 responses (1.0-to 2.0-fold higher), suggesting signaling bias towards the G q -dependent responses. Oppositely, full agonist Br-phHA-HA shows slightly higher (1.7-to 3.2-fold higher than histamine) normalized potencies for GRK2/3 and β-arrestin1/2 in comparison to G q -mediated responses (1.8-to 5-fold lower than histamine), suggesting signaling bias towards GRK2/3 and β-arrestin1/2. Interestingly, HPHA acted as a nearly full agonist in G q -mediated responses with 2-to 5-fold lower normalized potencies as compared to histamine, but a partial agonist in the interaction with GRK2/3 and β-arrestins with 2.6-to 14-fold higher normalized potencies than histamine. towards the Gq-dependent responses. Oppositely, full agonist Br-phHA-HA shows slightly higher (1.7-to 3.2-fold higher than histamine) normalized potencies for GRK2/3 and β-arrestin1/2 in comparison to Gq-mediated responses (1.8-to 5-fold lower than histamine), suggesting signaling bias towards GRK2/3 and β-arrestin1/2. Interestingly, HPHA acted as a nearly full agonist in Gq-mediated responses with 2-to 5-fold lower normalized potencies as compared to histamine, but a partial agonist in the interaction with GRK2/3 and β-arrestins with 2.6-to 14-fold higher normalized potencies than histamine. To quantitatively compare the efficacy and potential signaling bias of these H1R agonists, all concentration-response data were analyzed using the operational model to retrieve a transduction ratio (τ/KA) for each ligand in each response ( Figure 6C; Supplementary Table S3) [53,54]. The log (τ/KA) values for all agonists were first normalized to the log(τ/KA) values of reference agonist histamine within each response to yield Δlog(τ/KA) values as measure of relative effectiveness ( Figure 6D; Supplementary Table S4). Both Br-phHA-HA and CF3-phHA-HA displayed slightly higher relative effectiveness in most induce heterotrimeric Gq biosensor activation versus all other tested responses, only a significant bias towards β-arrestin2 recruitment was observed. The relative effectiveness of CF3-phHA-HA to induced GRK2 recruitment to H1R was slightly lower as compared to all other tested responses, however, this difference found not to be statistically different. Hence, all H1R agonist-induced responses were compared to Gq activation by calculating the ΔΔlog (τ/KA) values to quantify bias ( Figure 6E; Supplementary Table S5) [53][54][55].  Tables S1 and S2). Intrinsic activities were calculated versus reference full agonist histamine. Transduction ratio (Log(τ/K A ) (C), relative effectiveness (∆Log(τ/K A ) (D) and Log bias factor (∆∆Log(τ/KA) (E) as calculated after analyzing the concentration response curves in Figure 5 using the operational model of agonism to retrieve the transduction ratio (Log(τ/K A ; Supplementary To quantitatively compare the efficacy and potential signaling bias of these H 1 R agonists, all concentration-response data were analyzed using the operational model to retrieve a transduction ratio (τ/K A ) for each ligand in each response ( Figure 6C; Supplementary Table S3) [53,54]. The log (τ/K A ) values for all agonists were first normalized to the log(τ/K A ) values of reference agonist histamine within each response to yield ∆log(τ/KA) values as measure of relative effectiveness ( Figure 6D; Supplementary Table S4).
Both Br-phHA-HA and CF3-phHA-HA displayed slightly higher relative effectiveness in most responses as compared to reference agonist histamine. However, histamine was more effective in G q and GRK2 than Br-phHA-HA and CF3-phHA-HA, respectively. In contrast, Br-phHA, CF3-phHA, and HP displayed slightly lower relative effectiveness than histamine in most responses. Although Br-phHA-HA displayed lower relative effectiveness to induce heterotrimeric G q biosensor activation versus all other tested responses, only a significant bias towards β-arrestin2 recruitment was observed. The relative effectiveness of CF3-phHA-HA to induced GRK2 recruitment to H 1 R was slightly lower as compared to all other tested responses, however, this difference found not to be statistically different. Hence, all H 1 R agonist-induced responses were compared to G q activation by calculating the ∆∆log (τ/KA) values to quantify bias ( Figure 6E; Supplementary Table S5) [53][54][55].

Histamine H 1 R Agonists Induced Ca 2+ Mobilization in HeLa Cells Endogenously Expressing H 1 R
Finally, HeLa cells were stimulated with the H 1 R agonists to measure their efficacy in the rapid and transient Ca 2+ mobilization via endogenous H 1 R using the Ca 2+sensitive dye Fluo4NW ( Figure 7A) [56]. Endogenous expression of H 1 R in HeLa cells (B max = 150-200 fmol/mg protein [24,56]) is approximately 20-34-fold lower as compared to HEK293T cells transfected with 1 µg H 1 R or H 1 R-Rluc8 plasmid (B max = 3000-5000 fmol/mg protein; data not shown). Not surprisingly, all ligands displayed less efficacy for Ca 2+ mobilization in HeLa cells as compared to G q -activation responses in transfected HEK293T as revealed by their considerably lower pEC 50 values (>30-fold) for most ligands resulting in the (almost) absence of responsiveness to 100 µM Br-phHA and CF3-phHA, while responses by HP and HPHA did not allow accurate determination of their intrinsic activities ( Figure 7B,C; Supplementary Tables S1 and S2). However, these lower efficacies in this fast and short term (within 20-30 s) Ca 2+ response in HeLa cells as opposed to higher efficacies in the more prolonged responses (20-60 min) of the BRET-based assays in HEK293T cells might also be (in part) the consequence of binding kinetics of these agonists to the H 1 R, as observed for dopamine D2 receptor agonists [57]. Both CF3-phHA-HA and Br-phHA-HA displayed higher potency and intrinsic activities as compared to reference full agonist histamine ( Figure 7B,C; Supplementary Table S2). In contrast to all other H 1 R agonists, Br-phHA-HA displayed only a 2-fold lower potency to induce Ca 2+ mobilization in HeLa cells as compared to G q -activation in HEK293T cells (Supplementary Table S1). Similar to the BRET-based assays in HEK293T cells, CF3-phHA-HA and Br-phHA-HA displayed higher transduction ratios (log(τ/K A ) values) and relative effectiveness (∆log(τ/KA) values) in Ca 2+ signaling in HeLa cells in comparison to histamine, whereas transduction ratios and relative effectiveness values of HP and HPHA were lower ( Figure 7D,E and Supplementary Table S4).
In conclusion, BRET-based assays were used to measure G q activation, InsP 3 and Ca 2+ signaling, internalization, and interaction with GRKs2/3/5/6 and β-arrestins1/2 by the H 1 R in response agonist stimulation. Small differences in agonist efficacy were observed between some of these H 1 R responses. However, analysis of these functional responses using the operational model only revealed significant bias of a single agonist, BR-phHA-HA, towards β-arrestin2 recruitment over G q biosensor activation. Nevertheless, these BRET-based biosensors are valuable tools in the pharmacological characterization of H 1 R ligands that can be used in both real-time and end-point format at a reasonably high throughput.
as compared to Gq-activation in HEK293T cells (Supplementary Table S1). Similar to the BRET-based assays in HEK293T cells, CF3-phHA-HA and Br-phHA-HA displayed higher transduction ratios (log(τ/KA) values) and relative effectiveness (Δlog(τ/KA) values) in Ca 2+ signaling in HeLa cells in comparison to histamine, whereas transduction ratios and relative effectiveness values of HP and HPHA were lower ( Figure 7D,E and Supplementary Table S4). In conclusion, BRET-based assays were used to measure Gq activation, InsP3 and Ca 2+ signaling, internalization, and interaction with GRKs2/3/5/6 and β-arrestins1/2 by the H1R in response agonist stimulation. Small differences in agonist efficacy were observed

Materials
Fetal bovine serum (FBS) was bought from Bodinco (Alkmaar, The Netherlands) and penicillin/streptomycin was obtained from GE Healthcare (Uppsala, Sweden). Dulbecco's modified Berlin, Germany) and also resynthesized at AstraZeneca (Macclesfield, UK), as previously described [58][59][60]. All other chemicals were of analytical grade and purchased from standard commercial suppliers. Cell culture plastics were purchased from Greiner Bio-One GmbH (Frickenhausen, Germany).

Radioligand Binding Experiments
HEK293T cells were collected two days after transfection with 5 µg HA-H 1 R or H 1 R-Rluc8 plasmids per 10 cm dish in ice-cold phosphate-buffered saline (137 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 and 2 mM KH 2 PO 4 ), centrifuged at 1900× g for 10 min at 4 • C, and pellets were resuspended and homogenized on ice in binding buffer (50 mM Na 2 HPO 4 and 50 mM KH 2 PO 4 , pH7.4) using a Branson sonifier 250 (Boom bv, Meppel, The Netherlands). The Protein levels in the cell homogenates was determined using the Pierce TM BCA protein assay kit. Cell homogenates (0.5-1.6 µg/well) were incubated with 0.5 to 80 nM [ 3 H]mepyramine for 4 h at 22 • C with gentle agitation in the absence or presence of 10 µM mianserin to quantify total and nonspecific binding, respectively, as previously described [6]. Incubations were terminated by rapid filtration and 3 subsequent washes with ice-cold 50 mM Tris-HCl (pH 7.4) over 0.5% branched polyethyleneimine (750 kDa)-soaked GF/C filter plates using a 96 well FilterMate-harvester (PerkinElmer; Waltham, MA, USA). GF/C filter plates were dried at 52 • C for at least 30 min before addition of 25 µL/well MicroScint-O to quantify filter-bound radioactivity using a Wallac 1450 MicroBeta Trilux counter (PerkinElmer; Waltham, MA, USA) after a delay of at least 3 h. Radioligand binding affinity (K d ) and total receptor numbers (B max ) were determined by fitting the data to the "One-site-Total and nonspecific binding" model in GraphPad Prism version 8.4.3 (GraphPad Software, San Diego, CA, USA).

Intracellular Ca 2+ Mobilization in HeLa Cells
HeLa cells endogenously expressing the H 1 R were seeded in a clear-bottom 96-well plate (2 × 10 4 cells/well). The next day, the cells were loaded with Fluo-4NW dye (1 vial/24 mL) for one hour in assay buffer (HBSS supplemented with 20 mM HEPES and 2.5 mM probenecid), subsequently washed twice to remove the excess of dye, and reconstituted in assay buffer. Fluorescence (excitation at 494 nm and emission at 516 nm) was measured every second in the NOVOstar microplate reader (BMG Labtech; Ortenberg, Germany) at 37 • C. First the background signal was measured (F b ), followed by the peak Ca 2+ mobilization response 5-10 s after agonist injection (F ago ), finally Triton X-100 (0.25% v/v) was injected 40 s later to quantify the maximum calcium levels by lysing the cells (F t ). The agonist-induced Ca 2+ response was calculated as ∆fluorescence = (F ago − F b )/(F t − F b ) [56].

Data Analysis
All results were analyzed using GraphPad Prism v8.4.3 (GraphPad Software, San Diego, CA, USA). Concentration-response curves were fitted using the "three parameters-log(agonist) vs. response" model: Intrinsic activity (α) value is calculated as: α = f itted maximum response agonist f ited maximum response histamine Concentration-response curves were globally fitted using the operational model to determine the transduction coefficient logR (R = τ/K A ) as previously described [53][54][55], with τ and K A being the index of efficacy and functional equilibrium dissociation constant of the agonist, respectively: The logR value of reference agonist histamine is subtracted from the logR values of ligands for each specific response to obtain their relative effectiveness by eliminating the impact of cell-and assay-dependent effects [53][54][55]: Standard deviation on the ∆logR is calculated using: Next, the ∆logR values of each ligand in the different functional responses were subtracted from its ∆logR value in G q activation yielding ∆∆logR values to quantify signaling bias between G q activation and the other responses (x). ∆∆logR values > 0 indicate bias towards G q activation over the other responses (x) and vice versa for ∆∆logR values < 0 [53][54][55]: Standard deviation on the ∆∆logR is calculated using: Welch ANOVA (does not assume equal variance) with a Dunnett's T3 multiple comparisons test on the ∆logR values was used to determine the statistical significance of ligand activity between pathways, where p < 0.05 was considered to be significant.