Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 journal homepage: www.elsevier.com/locate/csbj Experimental and computational analysis of biased agonism on full-length and a C-terminally truncated adenosine A2A receptor Gemma Navarro a,b,1, Angel Gonzalez c,1, Stefano Campanacci d,1, Rafael Rivas-Santisteban a,d, Irene Reyes-Resina b,d,2, Nil Casajuana-Martin c, Arnau Cordomí c, Leonardo Pardo c,1, Rafael Franco b,e,1,⇑ a Dept. Biochemistry and Physiology, Faculty of Pharmacy and Food Science. Universitat de Barcelona. Barcelona, Spain b Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas. Instituto de Salud Carlos III, Madrid, Spain c Laboratori de Medicina Computacional, Unitat de Bioestadística, Facultat de Medicina, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain d Faculty of Chemistry, Universitat de Barcelona, Barcelona, Spain d Dept. Biochemistry and Molecular Biomedicine. School of Biology. Universitat de Barcelona. Barcelona. Spain e School of Chemistry. Universitat de Barcelona. Barcelona. Spain a r t i c l e i n f o a b s t r a c t Article history: Biased agonism, the ability of agonists to differentially activate downstream signaling pathways by sta- Received 20 May 2020 bilizing specific receptor conformations, is a key issue for G protein-coupled receptor (GPCR) signaling. Received in revised form 15 September The C-terminal domain might influence this functional selectivity of GPCRs as it engages G proteins, 2020 GPCR kinases, b-arrestins, and several other proteins. Thus, the aim of this paper is to compare the Accepted 16 September 2020 Available online 24 September 2020 agonist-dependent selectivity for intracellular pathways in a heterologous system expressing the full- length (A2AR) and a C-tail truncated (AD 40 2A R lacking the last 40 amino acids) adenosine A2A receptor, a GPCR that is already targeted in Parkinson’s disease using a first-in-class drug. Experimental data such Keywords: G protein coupled receptors as ligand binding, cAMP production, b-arrestin recruitment, ERK1/2 phosphorylation and dynamic mass Adenosine A2A receptor redistribution assays, which correspond to different aspects of signal transduction, were measured upon Functional selectivity the action of structurally diverse compounds (the agonists adenosine, NECA, CGS-21680, PSB-0777 and G protein binding LUF-5834 and the SCH-58261 antagonist) in cells expressing A2AR and AD 40 2A R. The results show that tak- b-Arrestin recruitment ing cAMP levels and the endogenous adenosine agonist as references, the main difference in bias was Molecular dynamic simulations obtained with PSB-0777 and LUF-5834. The C-terminus is dispensable for both G-protein and b- arrestin recruitment and also for MAPK activation. Unrestrained molecular dynamics simulations, at the ls timescale, were used to understand the structural arrangements of the binding cavity, triggered by these chemically different agonists, facilitating G protein binding with different efficacy. Ó 2020 The Authors. Published by Elsevier B.V. on behalf of Research Network of Computational and Structural Biotechnology. This is an open access article under the CC BY-NC-ND license (http://creative- commons.org/licenses/by-nc-nd/4.0/). 1. Introduction the first to start diverging [1] from the MECA receptor cluster [2], which is formed by the melanocortin, endothelial differentiation Adenosine triphosphate (ATP) is the main energy-transfer sphingolipid, and cannabinoid receptors. There are four identified molecule and adenosine is one of its main metabolites. Adenosine mammalian adenosine receptors (humans included) whose cog- receptors appeared early in evolution, as sensors of ATP decay, nate heterotrimeric G proteins are Gi (A1 and A3) or Gs/Golf (A2A because excess of adenosine correlates with ATP depletion. They and A2B) [3]. are considered as the most ancient within the class A (or The crystal structure of the A2A receptor (A2AR) was one of the rhodopsin-like) G-protein-coupled receptor (GPCR) family. More- first reported for the G-protein-coupled receptor (GPCR) family over, phylogenetic studies suggest that adenosine receptors were [4]. Today, there are nearly fifty structures of A1R and A2AR bound to agonists, antagonists, and/or to G proteins [5]; all of them dis- playing the common features of class A GPCRs [6]. Unfortunately, ⇑ Corresponding author at: School of Chemistry, University of Barcelona, Diagonal the N- and C-terminal domains, which are highly variable in 643, 08028 Barcelona, Spain. sequence, length, and structure [7], have been removed for crystal- E-mail addresses: rfranco123@gmail.comrfr, anco@ub.edu (R. Franco). 1 Equal contribution. lization purposes. The long C-terminal domain of A2AR (about 122 2 Current address: RG Neuroplasticity, Leibniz Institute for Neurobiology, Magde- amino acids) is known to engage G proteins, GPCR kinases, arrest- burg 39118, Germany. https://doi.org/10.1016/j.csbj.2020.09.028 2001-0370/Ó 2020 The Authors. Published by Elsevier B.V. on behalf of Research Network of Computational and Structural Biotechnology. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 ins and several other proteins [8,9]. The C-terminus is also respon- mental session, for instance of determination of cAMP levels, all sible for the constitutive activity of the receptor [10], influences agonists (plus/minus antagonist when indicated) were tested in the quaternary structure of heteromers [11,12], and is involved the same batch of transfected cells. in differential intracellular signaling [13]. A2AR offers numerous possibilities for therapeutic applications 2.3. Homogeneous HTRF binding assays [14–16]. A2AR antagonists show promise, among others, in Alzhei- mer’s and Parkinson’s diseases, attention-deficit hyperactivity dis- HEK-293T cells growing in 6-well plates were transiently trans- order, depression, and anxiety. Istradefylline, one of the most fected with 1 mg of plasmid encoding for pHALO-tagged human studied antagonists, is safe and efficacious in Parkinson and was A2AR (pHALO-A2AR, Cisbio Bioassays, Codolet, France) or 1 mg of approved in Japan as NouriastTM and by the US Food and Drug plasmid encoding for pHALO-tagged human AD 40 2A R and incubated Administration as NourianzTM [17]. A2AR agonists could be used in at 37 °C in a 5% CO2 humid atmosphere (24 h). Niemann Pick type C disease, autism-spectrum disorders, and schizophrenia. Regadenoson, a selective A2AR agonists that is a 2.4. Covalent labeling of cells expressing pHALO-tagged receptors coronary vasolidator, was approved in the United States as LexiscanTM. 48 h post transfection culture medium was removed, cells were These facts open the possibilities for more drug approvals, both washed with PBS and incubated with 100 nM of HALO-Lumi4Tb, a agonists and antagonists, related to adenosine receptors. Thus, the terbium derivative of O6-benzylguanine (SSNPTBC, Cisbio Bioas- aim of this work was double. First, we wanted to analyze the says, Codolet, France) - previously diluted in TagLite Buffer (TLB) agonist-dependent selectivity for intracellular pathways, known (LABMED, Cisbio Bioassays, Codolet, France) for 1 h at 37 °C under as functional selectivity or biased agonism, using structurally 5% CO2 humid atmosphere. After that, cells were washed with PBS diverse compounds in a heterologous system expressing A2AR. to remove the excess of HALO-Lumi4Tb, detached with Versene Biased agonists might offer attractive therapeutic properties rela- (Gibco Life Technologies, Paisley, Scotland, UK), centrifuged at tive to their unbiased counterpart, circumstance that remains to 1,500 rpm for 5 min and resuspended in TLB. Densities of 10,000 be validated in clinical trials [18]. In fact, adenosine receptor biased cells/well were used to carry out binding assays in white opaque agonists could be used for cardioprotection without bradycardia as 384-well plates. a serious adverse effect [19,20]. Second, we analyzed in living cells the effect of the C-terminal domain of A2AR in ligand binding and in 2.5. Non-radioactive homogeneous time-resolved FRET-based binding agonist-induced signaling. assays Saturation binding experiments were performed by incubating 2. Materials and methods cells expressing Tb-labeled HALO-A2AR with increasing concentra- tions of the A2AR antagonist SCH-442416, conjugated to a fluores- 2.1. Reagents cent probe developed by Cisbio (red A2AR ligand, purchased from Cisbio Bioassays, Codolet, France). The unspecific signal was Adenosine, NECA, CGS-21680, PSB-0777, LUF-5834 and SCH- obtained by incubating cells expressing Tb-labeled HALO-A2AR 58261 were purchased from Tocris Biosciences (Bristol, UK). HEPES with increasing concentrations of red A2AR ligand in the presence was purchased from SigmaAldrich (St. Louis, MO, US). Stock solu- of 10 mM unlabeled SCH-58261. Both, fluorescent ligands and unla- tions were prepared in DMSO. Aliquots of these stock solutions beled compounds, were diluted in TLB. 10 ml of labeled cells were kept frozen at 20 °C until use. (10,000 cells/well) were loaded onto 384-well white plates and 5 ml unlabeled SCH-58261 (10 mM final concentration) or TLB were 2.2. Cell culture and transient transfection added, followed by the addition of 5 ml of increasing concentrations (0–60 nM range) of red A2AR ligand. Plates were incubated pro- Human embryonic kidney 293T (HEK-293T) cells were grown in tected from light for 2 h at room temperature before time- Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Paisley, Scot- resolved fluorescence resonance energy transfer (TR-FRET) signal land, UK) supplemented with 2 mM L-glutamine, 1 mM sodium reading. The specific binding was calculated by subtracting the pyruvate, 100 units/ml penicillin/streptomycin, MEM non- unspecific binding from the total binding. essential amino acids solution (1/100) and 5% (v/v) heat inacti- For competition binding assays, HEK-293T cells transiently vated fetal bovine serum (FBS) [all supplements were from Invitro- expressing Tb-labeled HALO-A2AR, or AD 40 2A R, were incubated with gen, (Paisley, Scotland, UK)]. Cells were incubated in a humid 20 nM fluorophore-conjugated A2AR ligand, in the presence of atmosphere of 5% CO2 at 37 °C. 24 h after seeding in 6-well or increasing concentrations (0–10 mM range) of agonists (or antago- 96-well (for ERK phosphorylation assay) plates, cells were tran- nists when indicated). Plates contained 10 ml of labeled cells, and siently transfected with cDNA coding for wild-type A2AR or AD 40 2A R 5 ml of tested compounds or TLB were added prior to the addition (amino acids 372–412 on the C-terminal domain were deleted of 5 ml of the red A2AR ligand. Plates were then incubated for at [21]) with the PEI (PolyEthylenImine, SigmaAldrich) method as least 2 h at room temperature prior to TR-FRET signal detection. previously described [13,22]. Truncation of this part of the C-tail Detailed description of the HTRF assay is found elsewhere [24]. does not remove the proposed phosphorylation codes (PxPxxP/E/ D or PxxPxxP/E/D, where P represents phospho-serine or 2.6. Signal detection and data analysis phospho-threonine) for arrestin recruitment [23]. This phosphory- lation code starts 33 amino acids after P7.50 of the NPxxY motif in Signal was detected using a PHERAstar FS (BMG Lab technolo- human rhodopsin, whereas truncation of A2AR starts 87 amino gies, Offenburg, Germany) microplate reader equipped with a FRET acids after P7.50. Cells were incubated for 4 to 5 h with the cDNA optic module allowing donor excitation at 337 nm and signal col- of interest, PEI and NaCl in serum-free medium. After that, the lection at both 665 and 620 nm. A frequency of 10 flashes/well was serum-free medium was replaced by complete culture medium selected for the xenon flash lamp excitation. The signal was col- and cells were incubated for 48 h at 37 °C in 5% CO2 humid atmo- lected at both 665 and 620 nm using the following time-resolved sphere. The sequences in the plasmids were those coding for settings: delay, 150 ms; integration time, 500 ms. HTRF ratios were human receptors. It should be noted that within a given experi- obtained by dividing the acceptor signal (665 nm) by the donor sig- 2724 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 nal (620 nm) and multiplying this value by 10,000. The 10,000- 2.9. b-Arrestin 2 recruitment multiplying factor is used solely for the purpose of easier data handling. HEK-293T cells were transiently transfected with 0.625 mg of Data were analyzed using Prism 7 software (GraphPad Soft- cDNA coding for b-arrestin 2-Rluc and 2 mg of cDNA coding for ware, Inc., San Diego, CA), and competition data were fitted by A2A-YFP or with 1.5 mg of cDNA coding for b -arrestin 2-Rluc and non-linear regression to a one site-fit logIC50, competition curves 4 mg of cDNA coding for AD 40 2A R-YFP. BRET experiments were per- were -in all cases- monophasic. KD (dissociation constant) values formed 48 h after transfection. Cells were detached using HBSS of the fluorescent ligand were obtained from the specific binding containing 0.1% glucose, centrifuged for 5 min at 3,200 rpm and saturation curves. Note that Bmax values obtained from HTRF satu- resuspended in the same buffer. Protein concentration was quanti- ration curves do not reflect absolute values of receptor binding fied by using the Bradford assay kit (Bio-Rad, Munich, Germany). sites. Ki values were determined from competition binding assays Hereafter, YFP fluorescence at 530 nm was quantified in a FluoStar by using the calculated IC50 and KD values and the Cheng-Prusoff Optima Fluorimeter (BMG Labtechnologies, Offenburg, Germany) equation [25]. to quantify receptor-YFP expression. To measure b -arrestin recruitment, cells (20 mg of protein) were distributed in 96-well microplates (Corning 3600, white plates with white bottom) and 2.7. cAMP determination were incubated for 10 min with antagonists. Cells were then stim- ulated with agonists prior to the addition of 5 mM Coelenterazine H HEK-293T cells were grown in 6-well plates and transiently (Molecular Probes, Eugene, OR). BRET between b-arrestin 2-Rluc transfected with cDNAs for A2AR or for AD 40 2A R as described in 2.2. and receptor-YFP was determined and quantified at 5 min after Two hours before initiating the experiment, medium was replaced adding coelenterazine H. This time of response was selected from by serum-free DMEM medium. Then cells were detached, isolated time-response curves (Supplementary Fig. S1C). The readings were by centrifugation (5 min at 1,500 rpm) and resuspended in DMEM collected using a Mithras LB 940 (Berthold Technologies, Bad Wild- containing 50 mM zardaverine (phosphodiesterase inhibitor) to bad, GE), which allows the integration of the signals detected in the prevent degradation of cAMP, and 5 mM HEPES (pH 7.4). Cells were short-wavelength filter (485 nm) and the long wavelength filter placed in white 384-well plates (PerkinElmer) (6,000 cells/well) (530 nm). To quantify protein-RLuc expression, luminescence and incubated with antagonists or vehicle for 15 min before treat- readings were also collected 10 min after the addition of 5 mM coe- ment with agonist or vehicle for 15 min. Finally, cAMP-Eu and the lenterazine H. fluorescent antibody were added. Readings were performed after one hour incubation at 25 °C. Homogeneous time-resolved fluores- 2.10. Dynamic mass redistribution assays (DMR) cence energy transfer (HTRF) measures were performed using the Lance Ultra cAMP kit (PerkinElmer, Waltham, MA, US). Fluores- Cell mass redistribution induced upon receptor activation was cence at 665 nm was measured in a PHERAstar Flagship plate detected by illuminating with polychromatic light the underside reader equipped with an HTRF optical module (BMG Lab technolo- of a biosensor and measuring the changes in the wavelength of gies, Offenburg, Germany). the reflected monochromatic light, that is a sensitive function of the index of refraction. The magnitude of the wavelength shift (in picometers) is directly proportional to the amount of mass 2.8. ERK1/2 phosphorylation assays redistribution. 48 h before the assay, HEK-293T cells were tran- siently transfected with 2 mg of cDNA coding for A2AR or AD 40 2A R. HEK-293T cells were grown on transparent Biocat Poly-D- HEK-293T cells were seeded in 384-well sensor microplates to Lysine 96-well plates (Deltalab) and kept at the incubator for obtain 70–80% confluent monolayers constituted by approximately 24 h. Then cells were transiently transfected with cDNA coding 10,000 cells/well. Prior to the assay, cells were washed twice with for A2AR or AD 40 2A R and incubated for 48 h at 37 °C in a 5% CO2 assay buffer (HBSS with 20 mM HEPES, pH 7.15 and 1% BSA) (Sig- humid atmosphere. Supplementary Fig. S1A shows that treatment maAldrich, St. Louis, MO, US) and incubated for 2 h with assay- with CGS-21680 in non-transfected HEK-293T cells is not signifi- buffer containing 0.1% DMSO (24 °C, 30 ml/well). Hereafter, the sen- cantly different to the basal condition. Thus, the expression level sor plate was scanned, and a baseline optical signature was of A2AR in non-transfected HEK-293T cells is negligible. 2 h before recorded for 10 min before adding 10 ml of the selective antagonists initiating the experiment, the medium was substituted by serum- for 30 min, followed by the addition of 10 ml of the selective ago- free DMEM. Cells were stimulated at 25 °C for 10 min with vehi- nists; all test compounds were diluted in assay buffer. Then, cle or agonists (Supplementary Fig. S1D). ERK phosphorylation DMR responses were monitored for at least 5000 s in an EnSpireÒ was measured at ‘‘short” times to detect G-protein mediated sig- Multimode Plate Reader (PerkinElmer, Waltham, MA, USA) by a nal. After that, cells were washed twice with cold PBS before the label-free technology. Results were analyzed using EnSpire Work- addition of lysis buffer (30 ml/well; Perkin Elmer, Waltham, MA, station Software v 4.10. US) and incubated for 15 min at 25 °C on a Heidolph Titramax 100 shaker. 10 ml of each cell lysate was transferred to white 2.11. Calculation of bias factor ProxiPlate 384-well microplates (PerkinElmer; Waltham, MA, USA). ERK1/2 phosphorylation was determined using AlphaS- Bias factor (bias) was calculated with the following formulas creenÒSureFireÒ kit (Perkin Elmer, Waltham, MA, US): 5 ml/well adapted from [26] in which s represents the agonist efficacy, and of acceptor beads were added. Plates, protected from light, were KA the agonist affinity [27]. incubated for 2 h at 25 °C. Finally, 5 ll/well of donor beads were     added and plates, protected from light, were incubated for 2 h s s log bias ¼ D log  D log before analysis. Fluorescence was determined using an EnSpireÒ KA j1 KA j2 Multimode Plate Reader (PerkinElmer, Waltham, MA, USA). The value of reference (100%) was that achieved in the absence of   DDlog s KA any treatment (basal). The effect of ligands was given in percent- bias ¼ 10 j1 j2 age respect to the basal value. 2725 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 log (s/KA) is defined as the transduction coefficient or synony- lipids and solvent molecules. A time step of 2.0 fs was used for mously as logR, a parameter that can be used to compare agonist the integration of equations of motions. All bonds and angles were activity between different systems. The transduction coefficient kept frozen using the LINCS algorithms. Lennard-Jones interactions is a measure of the ability of a ligand to activate the receptor [28]. were computed using a cutoff of 10 Å, and the electrostatic inter- The pathway of reference j1 was cAMP determination for Gs- actions were treated using PME with the same real-space cutoff coupling, whereas the other pathways (ERK1/2 phosphorylation, under periodic boundary conditions. MD simulations were per- ß-arrestin 2 recruitment, or DMR) were j2. s denotes the maximum formed using GROMACS v5.0.7. The AMBER99SB force field as value in each response and KA is the antilogarithm of half maximal implemented in GROMACS, Berger parameters for POPC lipids, effective concentration (EC50 if the agonist provides an increase and the GAFF parameters for the ligands (see above) were used response, IC50 if the agonist provides a reduction of the response for the MD simulations, attending the performance of this protocol induced by another reagent). on membrane-protein systems [35]. 2.12. Computational methods 2.13. Statistical analysis Preparation of protein structure and ligand parametrization. Crys- Experimental data was managed and analysed with GraphPad tal structures of the A2AR in its active intermediate state in com- Prism software version 7 (San Diego, CA, USA) or IBM SPSS Statis- plex with the adenosine (PDB id 2YDO), NECA (2YDV) and CGS- tics version 25.0 (IBM Corp., NY, USA). Unless otherwise stated data 21680 (4UHR) agonists, and the crystal structure of A2AR in its are the mean ± S.E.M (n = 5 or higher). P-values lower than 0.05 inactive state in complex with the ZM241385 antagonist (PDB id were considered statistically significant. 4EIY) were obtained from the Protein Data Bank (rcsb.org). Fusion proteins were removed and stabilizing mutations were mutated to the native sequence using MODELLER v9.12 [29]. Parameters for 3. Results and discussion adenosine, NECA, CGS-21680, PSB-0777, LUF-5834 and SCH- 58261 were obtained from the general Amber force field (GAFF) 3.1. Ligand binding experiments and HF/6-31G*//HF/6-31G*-derived RESP atomic charges calcu- lated with Gaussian09. We studied the binding of the agonists adenosine, NECA and Molecular docking. The PSB-0777 and LUF-5834 agonists were CGS-21680, the partial agonists PSB-0777 and LUF-5834, and the docked into the active intermediate state (2YDV) and the SCH- selective antagonist SCH-58261 by homogeneous time-resolved 58261 antagonist was docked into the inactive state (4EIY) using fluorescence (HTRF) experiments, schematized in Fig. 1A, in living the Molecular Operating Environment (MOE) software (Chemical HEK-293T cells expressing either wild type (wt) A2AR or a trun- Computing Group Inc., Montreal, Quebec, Canada). One hundred cated receptor (AD 40 2A R, lacking the last 40 amino acids of the C- docking solutions per ligand were generated by the triangle terminal end) tagged with HaloTag on the N-terminus (see 2.3– matcher algorithm into the active site of the receptor structures. 2.5). First, a saturation curve was performed using a Top-ranking solutions were inspected and conformations in which fluorescence-conjugated antagonist (Fig. 1B). The obtained KD val- the central moiety of PSB-0777 and the aminopyridine group of ues were 1.7 nM and 1.2 nM for full-length A2AR and AD 40 2A R, respec- LUF-5834 were located in the same region as the adenine moiety tively (Table 1). Subsequently, competition assays were performed of adenosine and NECA were selected (Supplementary Fig. S2). using 20 nM of the fluorescence-conjugated antagonist and The binding pose of SCH-58261 was selected in such a manner that increasing amounts of the ligands (Fig. 1C-1D). The Ki value of each the orientation of its pyrazolo-triazolo-pyrimidine central moiety compound was calculated from IC50 and KD values using the Cheng- was similar to the bicyclic triazolotrazine core of the highly potent Prusoff expression [25] (Table 1). The obtained Ki values are in the selective antagonist ZM241385 found in the 4EIY structure (Sup- range of those obtained in radioligand binding assays for isolated plementary Fig. S2). membranes from tissues or transfected cells. According to this Molecular dynamics (MD) simulations. The complexes of adeno- assay, the Ki values of agonists are in the 13–219 nM range, LUF- sine, NECA, CGS-21680, PSB-0777 and LUF-5834 with the interme- 5834 being the compound that binds A2AR the strongest and diate active structure of A2AR and the complex of SCH-58261 with PSB-0777 the weakest. Clearly, removal of the C-terminal domain the inactive structure of A2AR (see above) were embedded in a pre- results in weaker agonist binding to the orthosteric site (a ratio equilibrated lipid bilayer box containing 1-palmitoyl-2-oleoyl-sn- of Ki values in the 1.4–4.4 range, which correspond to binding free glycero-3-phosphatidylcholine (POPC), water molecules (TIP3P) energy differences between 0.2 and 0.9 Kcal/mol). On the other and monoatomic Na + and Cl ions (0.2 M). Assignment of ioniza- hand, removal of the C-terminal domain has a significant impact tion states and hydrogens at physiological pH for the selected on the binding of the antagonist SCH-58261 (0.7 nM vs 113.9 nM structures was conducted with the Protonate3D method [30] as for full-length A2AR and AD 40 2A R, respectively; binding free energy implemented in MOE. D2.50 was deprotonated (negatively difference of 3.0 Kcal/mol). charged) for antagonist-bound receptor simulations and proto- nated (neutral) for agonist-bound receptor simulations [31], 3.2. Molecular models of agonist binding whereas H264 was protonated (positively charged) in all simula- tions [32]. A sodium ion was placed, near the negatively charged The binding of these compounds to A2AR has been extensively D2.50, in the inactive, antagonist-bound, structure of A2AR studied by site-directed mutagenesis, computer simulations, and [33,34]. Molecular systems were subject to a 1000 cycles of energy supported by the recent crystallographic data of adenosine, NECA minimization, followed by 20 ns of gradual relaxation of positional and CGS in complex with A2AR (reviewed in [36]). PSB-0777 and restraints (corresponding to 100, 50, 25 and 10 kJ.mol1nm2) at LUF-5834 were docked (see 2.12) using, as template, ligands solved protein backbone coordinates before the production phase in order in crystal structures with homologous chemical scaffolds (see Sup- to hydrate the receptor cavities and allow lipids to pack around the plementary Fig. S2). All molecules but LUF-5834 are structurally protein. After equilibration, unrestrained MD simulation (3 replicas related through a common adenosine core. Fig. 2 summarizes the of 1 ls giving a total of 3 ls of sampling time) were produced for binding modes of all compounds, emphasizing common/different each ligand-receptor complex at a constant temperature of 300 K areas of the receptor occupied by each ligand. In order to under- using separate v-rescale thermostats for the receptor, ligands, stand the molecular mechanism of receptor activation (see 3.5), 2726 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 Fig. 1. Competition curves in HTRF-based assays. Panel A: Scheme of the homogeneous binding technology performed in living HEK-293T cells. Panel B. Saturation isotherm of binding of fluorophore-conjugated red A2AR ligand to HEK-293T cells transiently transfected with HALO-A2AR in the absence (red) or presence of 10 mM SCH-58261 (black); specific binding is depicted in green. Panels C-D: Non-radiolabeled HTRF-based competition curves of specific binding of 20 nM fluorophore-conjugated red A2AR ligand in the presence of increasing concentrations of different agonists and of the selective antagonist SCH-58261, in cells expressing A2AR (C) or AD 40 2A R (D). Data represent the mean ± SEM of a representative experiment (n = 4). HTRF ratio = (665 nm acceptor signal/620 nm donor signal)  10,000. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Table 1 IC50 and Ki values obtained from competition binding assays in HEK-293T cells As shown in Fig. 2, three regions (blue, green and yellow rectan- expressing wild type A2AR or AD 40 2A R. Values reported are the mean of three gles) are invariably occupied by functional groups of all the studied experiments conducted in triplicates fitted to a monophasic competition model. compounds, illustrated by the binding of adenosine in the central A2AR AD40 2A R panel. The heterocyclic site (blue rectangle) corresponds to the IC50 Ki IC50 Ki KD CT / DDGa central region and accommodates the adenine moiety of adeno- i (nM) (nM) (nM) (nM) Kwt i (kcal/mol) sine, NECA, CGS-21680, and PSB-0777, and the aminopyridine adenosine 1658 183 4804 531 2.9 0.6 group of LUF-5834 through aromatic/hydrophobic interactions NECA 1118 124 2190 242 2.0 0.4 with F168, M5.38, M7.35, I7.39 (numbering in Ballesteros and CGS-21680 746 82.5 1690 187 2.3 0.5 Weinstein scheme [37], except for residues outside helices in PSB-0777 1984 219 8680 960 4.4 0.9 which sequence numbers are used) and polar interactions of the LUF-5834 115 13 166.0 18 1.4 0.2 exocyclic nitrogen with N6.55 (Supplementary Fig. S3). The second SCH-58261 6.3 0.7 1030 114 162.3 3.0 region (green rectangle) accommodates the 50 -hydroxymethyl A2AR AD40 2A R (adenosine and PSB-0777), N-ethylcarboxamido (NECA and CGS- Bmax KD (nM) Bmax KD (nM) 21680) and hydroxyphenyl group of LUF-5834 (Fig. 2). The 50 - 6959 1.677 9214 1.22 hydroxymethyl group attached to the ribose ring of adenosine a Experimental binding free energy differences between wild type (A2AR) and and PSB-0777 and the hydroxyphenyl group of LUF-5834 hydrogen truncated (AD 2A R) receptors, calculated as DDG = -RT ln (Ki 40 DCT wt /Ki ) where R is bond H6.52 during all the simulation time (Supplementary Fig. S3). 1.987 cal mol1 K1 and T is the temperature of 298°K (-RT = 0.592 kcal mol1) NECA and CGS-21680 replace the 50 -hydroxymethyl group by the longer N-ethylcarboxamido group that can directly interact with both H6.52 and T3.36 (Supplementary Fig. S3). Several X-ray struc- tures of A2AR contain crystallized water molecules in the environ- we studied by molecular dynamics (MD) simulations (see 2.12) the ment of T3.36 and N5.42 [4] that are important for the activity of interactions of crucial ligand moieties to key receptor amino acids GPCRs [38,39]. Supplementary Fig. S6 shows that during the MD in all these regions. Detailed analysis of the simulations shows that simulations adenosine, PSB-0777 and LUF-5834 maintain this these binding modes were stable during the unbiased 1 ls MD water bridge, whereas NECA and CGS-21680 cannot because the simulations (three replicas) as shown by the relatively low move- N-ethylcarboxamido group interacts with T3.36. Finally, the third ment in root mean-square deviation (rmsd) plots of the receptor region (yellow rectangle) is occupied by the ribose moiety of ade- and ligand heavy atoms, as well as the conservation of the sec- nosine derivatives and by the carbonitrile substituent of LUF-5834. ondary structure elements of A2AR (Supplementary Figs. S3-S5). Most important interactions in this site imply the formation of 2727 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 Fig. 2. Ligand-receptor complexes. Cross-section through the A2AR, highlighting the agonists adenosine (white sticks), NECA (cyan), CGS-21680 (pink), PSB-0777 (green) and LUF-5834 (blue) occupying the binding site. Color rectangles highlight regions of the orthosteric binding cavity that are occupied by functional groups of the studied agonists (see 3.2). The PIF motif (in yellow) located below the orthosteric binding cavity, and the side chain of the highly conserved R3.50 (in green) of the DRY motif near the G protein binding site are highlighted. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) hydrogen bonds from the hydroxyl groups at positions C2 and C3 much later in the signaling pathway such as protein trafficking, of the ribose moiety of adenosine derivatives with the S7.42 and rearrangement of cytoskeleton and adhesion or morphological H7.43 residues. changes. The amount of transfected cDNA for A2AR and AD 40 2A R Three additional regions are occupied by the ligands (orange, was adjusted to obtain similar receptor expression levels (Supple- cyan and black in Fig. 2). The ethyl group of the N- mentary Fig. S1B). b-arrestin recruitment and ERK1/2 phosphoryla- ethylcarboxamido moiety of NECA and CGS-21680 extends deep tion were analyzed by time response curves (Supplementary into the binding pocket (orange), forming hydrophobic interactions Figs. S1C-S1D). In our cAMP assay conditions, adenosine, NECA, with C5.46 (Supplementary Fig. S3). These interactions between and CGS-21680 behaved as full agonists, whereas PSB-0777 and aliphatic chains and sulfur-containing amino acids have been LUF-5834 were partial agonists on full-length A2AR. The effect shown to be of high energy [40]. The long substituents of CGS- was specific for all agonists as shown by the blockade of cAMP pro- 21680 and PSB-0777 expand toward the extracellular loops duction using the selective receptor antagonist SCH-58261 (Sup- through a cavity within TMs 1, 2 and 7 (black rectangle), which plementary Fig. S8). b-arrestin recruitment assays showed that has been shown to influence G protein and b-arrestin signaling in PSB-0777 and NECA are more efficient in recruitment than adeno- other GPCRs [41,42]. The negatively charged SO 3 group of PSB- sine and CGS-21680. Remarkably, LUF-5834, which is a partial ago- 0777 interacts with His264 in ECL3, whereas the longer chain of nist in the cAMP assay, is as efficient as CGS-21680 and adenosine CGS-21680 permits the negatively charged COO to form an ionic in b-arrestin recruitment. pERK1/2 dose–response curves follow interaction with K153 in ECL2 (Supplementary Fig. S7). Finally, the similar patterns as cAMP curves, with the exceptions of PSB-0777 imidazole substituent of LUF-5834 interacts with residues Y1.35, that behaves as a full agonist, and LUF-5834 that is a very weak A2.61, I2.64, S2.65, and I7.39 in the sideward region of the binding partial agonist. DMR responses were all very similar with PSB- cavity (cyan rectangle). This binding orientation of LUF-5834 has 0777 providing a slightly larger signal at higher concentrations. been previously proposed using mutagenesis studies [43]. This In agreement with previous reports [10], removal the last 40 region is generally occupied by inverse agonists and, thus, we amino acids of the C-terminus in the AD 40 2A R construct caused a believe contributes to the low activity of LUF-5834 regarding the right-shift of the cAMP dose–response curves (a ratio of EC50 values rest of the agonists. in the 2.2–9.8 range, similar to the range of Ki ratios) with almost no change in Emax for all agonists, with the exception of LUF- 3.3. Agonist-induced signaling responses 5834 whose effect became almost negligible in the truncated form of the receptor (Fig. 3 and Table 2). C-terminal truncation provides Next, we measured four functional read-outs that correspond to a small decrease in b-arrestin recruitment with no changes in Emax different steps of the signaling pathways in cells expressing full- rank order relative to full-length A2AR. pERK1/2 dose–response length A2AR or AD 40 2A R: cAMP production, b-arrestin recruitment, curves of AD 40 2A R were very similar to full-length A2AR except that ERK1/2 phosphorylation, and dynamic mass redistribution (DMR) pEC50 increased for the truncated receptor, which goes against assays (Fig. 3 and Table 2). Immediately following receptor activa- the general trend, and the potency of NECA and PSB-0777 became tion, cAMP levels increase as the result of adenylyl cyclase activa- close to that of adenosine and CGS-21680. Finally, DMR read-outs tion by Gs. Next, receptor phosphorylation by G protein kinases of AD 40 2A R were almost negligible for NECA and adenosine while we triggers b-arrestin recruitment. Later, ERK signaling is regulated could observe a higher signal for PSB-0777 compared to that of the by G protein or/and b-arrestin. DMR accounts for events that occur other compounds in A2AR and AD 40 2A R. 2728 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 nal, i.e. cAMP response elicited by adenosine (see 2.8). Bias factors for CGS-21680, PSB-0777, LUF-5834 and NECA agonists in cAMP, b- arrestin recruitment, ERK1/2 phosphorylation and DMR signaling responses are summarized in the radar plot of Fig. 4. Adenosine and CGS-21680 are balanced agonists with similar bias factors for the four pathways, whereas NECA has small factors for DMR and b-arrestin recruitment and LUF-5834 has small factors for all responses other than cAMP. By contrast, PSB-0777 has a higher bias for the pERK1/2 response. Truncation of the C-terminus makes LUF-5834 and PSB-0777 more balanced. In fact, LUF-5834 acquires the largest bias factors for b-arrestin recruitment, ERK1/2 phospho- rylation and DMR. By contrast, NECA keeps the small factors for b- arrestin recruitment and DMR. Overall the results are consistent with the C-terminus being dispensable for both G-protein and arrestin recruitment and also for MAPK activation (ERK1/2 phos- phorylation and DMR), with only minor differences in signaling compared to full-length A2AR. 3.5. Molecular mechanisms of agonist-induced receptor activation The first step to understand the mechanisms of agonist-induced receptor activation, was to compare by MD simulations the trajec- tories of amino acids at positions 3.40, 5.50, and 6.44 (Fig. 5), which have been named as the ‘‘P-I-F” motif [44], the ‘‘transmission switch” [45], the ‘‘triad core” [46] or the connector region [6], in the presence of the agonists adenosine, NECA, CGS-21680, PSB- 0777 and LUF-5834 and the antagonist SCH-58261. These residues, located below the ligand binding cavity and above the G protein or arrestin binding cavity (see Fig. 2), adopt different positions upon binding of agonists or antagonists [47,48], and we have been using them to predict the effect of the ligand on the conformational state of the receptor [49–51]. Clearly, the agonist-bound, active-like complexes are characterized relative to the antagonist-bound, inactive-like, complex by the proposed inward movement of TM5 at the highly conserved P5.50, rotation of TM3 due to a steric clash with the bulky I3.40, and an outward movement of F6.44 in TM6 (Fig. 5). Unfortunately, these movements are similar for full and partial agonists and cannot be used to explain the different agonist Fig. 3. Signaling in cells expressing either A2AR or AD40 2A R. Dose response curves on pharmacological profiles. 0.5 mM forskolin-induced cAMP levels, on ß-arrestin recruitment, on ERK1/2 phosphorylation, and on dynamic mass redistribution (DMR). Data (n = 12, each in Thus, in order to understand the structural arrangements of the triplicates) for cAMP are given in percentage (100% represents the forskolin effect). binding cavity, triggered by these chemically different agonists, Data (n = 10, each in triplicates) for BRET assays, used to determine b-arrestin facilitating G protein binding or b-arrestin recruitment with differ- recruitment, are given in milliBRET Units (mBU). Data (n = 6, each in triplicates) for ent efficacy, we studied the trajectories of a selected group of 34 ERK1/2 phosphorylation are expressed as % with respect to basal levels. DMR amino acids either located above (in the ligand binding cavity) tracings are representing the picometer (pm)-shifts of reflected light wavelengths over time upon ligand treatment. and below (in the G protein or arrestin binding cavity) the ‘‘trans- mission switch” amino acids in the presence of agonists and the antagonist (see Supplementary Table S1). In the bivariate correla- tion analysis, dependent variables are Emax values measured in 3.4. Addressing biased agonism cAMP production and b-arrestin recruitment, whereas indepen- dent variables are the movement of the chosen amino acid in the Bias factors were calculated taking as reference the effect of agonist-bound, active-like complexes relative to the antagonist- adenosine, the endogenous compound, and the Gs-mediated sig- bound, inactive-like complex. This is measured as the distance Table 2 pEC50 and Emax values obtained in HEK-293T cells expressing wild type A2AR or truncated AD40 2A R for cAMP production, b-arrestin II recruitment, ERK1/2 phosphorylation and dynamic mass redistribution (DMR) response (Fig. 3). cAMP assays b -arrestin assays pERK1/2 assays DMR assays A2AR AD40 2A R A2AR AD40 2A R A2AR AD40 2A R A2AR AD 40 2A R pEC50 Emax pEC50 Emax pEC50 Emax pEC50 Emax pEC50 Emax pEC50 Emax pEC50 Emax pEC50 Emax Adenosine 8.0 156.6 8.1 163.5 7.8 24.1 7.7 23.7 7.5 373.0 7.7 326.9 6.4 1231 6.4 517.8 NECA 9.0 156.6 9.0 166.0 7.9 34.9 7.8 31.3 8.3 392.1 8.2 302.6 6.4 1416 6.5 407.9 CGS-21680 8.0 155.1 8.2 163.4 7.9 25.9 7.8 26.8 7.4 373.5 7.7 315.2 6.4 1374 6.4 941.5 PSB-0777 8.1 127.7 8.6 133.5 7.9 37.6 7.7 33.8 8.5 374.7 8.3 300.7 6.2 1887 6.2 1332 LUF-5834 9.7 131.1 8.2 113.6 8.0 27.8 7.8 28.0 8.2 183.9 8.2 175.8 6.3 1306 6.3 988.6 2729 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 Fig. 4. Radar plot representation of bias factors. Plots show the bias factors of the different compounds in the different functional outcomes in cells expressing A2AR or AD40 2A R. Adenosine and the Gs-cAMP signaling pathway were used as reference for calculations. Fig. 5. Receptor side-chain movements in response to agonists. Plot of the centroids (calculated from 100 snapshots) of the Cb atoms of a selected group of 34 amino acids (Table S1) located above (in the ligand binding cavity) and below (in the G protein or b-arrestin binding cavity) the ‘‘transmission switch” amino acids obtained during 1 ls of MD simulations of A2AR in the presence of the agonists adenosine, NECA, CGS-21680, PSB-0777 and LUF-5834 and a selective antagonist SCH-58261. The distances between the centroids of the agonist-bound conformations and the centroid of the antagonist-bound conformation were statistically correlated with Emax values measured in cAMP production and b-arrestin recruitment (Table S1). (A) Movement of the salt bridge between E169ECL2 and H264ECL3 that is proposed to govern the residence time of ligands. (B) Movement of E131.39, and the nearby H2787.43, that correlates with b-arrestin recruitment. (C) The proposed mechanism of receptor activation at the ‘‘transmission switch” amino acids (inward movement of TM 5, an anticlockwise rotation of TM 3, and an outward movement of TM 6, see arrows) is observed but similar for full and partial agonists. (D) Movement of T883.36 and W2466.48 that correlate with cAMP production (Supplementary Fig. S9). between the centroid (calculated from 100 snapshots obtained and Y5.58 (p = 0.039) below the ‘‘transmission switch” (Supple- during 1 ls of unbiased MD simulations) of the Cb positions of mentary Table S1). Y5.58 is a key amino acid in the process of the chosen amino acids of A2AR bound to agonists and the centroid receptor activation as it stabilizes the extended conformation of of A2AR bound to the antagonist. No deviations from normality R3.50 in the active state [52]. Positions 3.36 and 6.48 have been (Shapiro-Wilk test) are observed in all variables, thus, correlation described as conformational toggle or trigger switches involved analyses were performed using a Pearson test for continuous, nor- in the initial agonist-induced receptor activation in other GPCRs mally distributed variables. such as cannabinoid CB1R [53,54], serotonin 5HT4R [55], melano- Although ligand efficacy is a function of multiple factors, we corin MC4R [56], mGlu2R [57], or A2AR [58]. With the aim of under- found a clear statistically significant correlation between Emax in standing at the molecular level the different activation trends cAMP assays and the movement of the side chains of T3.36 among agonists, we explored the amount of time, in the MD sim- (p = 0.033) and W6.48 (p = 0.015) above the ‘‘transmission switch” ulations, the side chains of T3.36 and W6.48 spend in the gauche 2730 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 + (g+, v1 = -60°), gauche- (g-, v1 = 60°) or trans (t, v1 = 180°) con- domí: Investigation, Validation, Supervision. Leonardo Pardo: formations (see Supplementary Fig. S9). Clearly, the antagonist- Conceptualization, Validation, Project administration, Resources, bound structure favors the g+/g+ conformations of T3.36 and Writing - original draft, Funding acquisition. Rafael Franco: Con- W6.48, whereas full agonists adenosine, NECA, and CGS-21680 ceptualization, Validation, Project administration, Resources, Writ- favor the g-/g+ conformations. Notably, partial agonists PSB-0777 ing - original draft, Funding acquisition. and LUF-5834 cannot achieve these g-/g+ conformations as fre- quently as full agonists. In this respect, it should be noted that Thr residues in this g- conformation are capable of hydrogen bond- Appendix A. Supplementary data ing the backbone carbonyl in the previous turn of the helix, which is known to trigger a local opening of the helix [59] that might be Supplementary data to this article can be found online at necessary for receptor activation. Importantly, the T3.36A muta- https://doi.org/10.1016/j.csbj.2020.09.028. tion impedes signaling [43,60] and the W6.48A mutation has sig- nificant deleterious effects on receptor function [43]. In the case of Emax in b-arrestin recruitment, we found a statis- References tically significant correlation only with the movement of the side chain of E1.39 (p = 0.014) above the ‘‘transmission switch” (Supple- [1] Mickael ME, Rajput A, Steyn J, Wiemerslage L, Burglin T. An optimised phylogenetic method sheds more light on the main branching events of mentary Table S1). Because E1.39 and H7.43 might form an ionic rhodopsin-like superfamily. Comp Biochem Physiol Part D Genomics interaction we calculated the distance between these side chains Proteomics 2016;20:85–94. in the presence of the different agonists. However, we have not [2] Fredriksson R, Lagerstrom MC, Lundin LG, Schioth HB. The g-protein-coupled receptors in the human genome form five main families. Phylogenetic analysis, observed a statistically significant correlation between this dis- paralogon groups, and fingerprints. Mol Pharmacol 2003;63:1256–72. tance and Emax in b-arrestin recruitment. [3] Fredholm BB, IJzerman AP, Jacobson KA, Linden J, Muller CE. International Union of Basic and Clinical Pharmacology. LXXXI. Nomenclature and classification of adenosine receptors–an update. Pharmacol Rev 2011;63:1–34. 4. Conclusion [4] Jaakola VP, Griffith MT, Hanson MA, Cherezov V, Chien EY, Lane JR, et al. The 2.6 angstrom crystal structure of a human A2A adenosine receptor bound to an antagonist. Science 2008;322:1211–7. The agonist-dependent selectivity for intracellular pathways of [5] Pandy-Szekeres G, Munk C, Tsonkov TM, Mordalski S, Harpsoe K, Hauser AS, A2AR has been studied using, in synergy, molecular modeling tools et al. GPCRdb in 2018: adding GPCR structure models and ligands. Nucleic and pharmacological assays. This combination of expertise has per- Acids Res 2018;46:D440–6. [6] Weis WI, Kobilka BK. The molecular basis of G protein-coupled receptor mitted to understand the structural arrangements of the binding activation. Annu Rev Biochem 2018;87:897–919. cavity, triggered by chemically different agonists, facilitating G [7] Gonzalez A, Cordomí A, Caltabiano G, Campillo M, Pardo L. Impact of helix protein binding with different efficacy. The mechanism of irregularities on sequence alignment and homology modelling of G protein- agonist-induced b-arrestin recruitment seems more difficult to coupled receptors. ChemBioChem 2012;13:1393–9. [8] Ciruela F, Canela L, Burgueno J, Soriguera A, Cabello N, Canela EI, et al. rationalize. First, different ligands stimulation might induce dis- Heptaspanning membrane receptors and cytoskeletal/scaffolding proteins: tinct patterns of receptor phosphorylation, which direct specific focus on adenosine, dopamine, and metabotropic glutamate receptor function. b-arrestin conformations and functional outcomes [61–63]. Sec- J Mol Neurosci 2005;26:277–92. [9] Keuerleber S, Gsandtner I, Freissmuth M. From cradle to twilight: the carboxyl ond, there is increasing evidence that biased signaling could also terminus directs the fate of the A(2A)-adenosine receptor. Biochim Biophys be a consequence of binding kinetics [64–66]. In particular, a rela- Acta 2011;1808:1350–7. tion between residence time of ligands with biased signaling in [10] Klinger M, Kuhn M, Just H, Stefan E, Palmer T, Freissmuth M, et al. Removal of the carboxy terminus of the A2A-adenosine receptor blunts constitutive both A1R and A2AR [67,68] that is governed by a salt bridge activity: differential effect on cAMP accumulation and MAP kinase between E169ECL2 and H264ECL3 in A2AR (Fig. 5A) [69,70] has been stimulation. Naunyn Schmiedebergs Arch Pharmacol 2002;366:287–98. shown. Finally, the last 40 amino acids of the C-terminal end of [11] Navarro G, Ferre S, Cordomi A, Moreno E, Mallol J, Casado V, et al. Interactions between intracellular domains as key determinants of the quaternary A2AR receptor are dispensable for both G-protein and arrestin structure and function of receptor heteromers. J Biol Chem recruitment and also for MAPK activation (ERK1/2 phosphorylation 2010;285:27346–59. and DMR). [12] Kofalvi A, Moreno E, Cordomi A, Cai NS, Fernandez-Duenas V, Ferreira SG, et al. Control of glutamate release by complexes of adenosine and cannabinoid receptors. BMC Biol 2020;18:9. 5. Notes [13] Navarro G, Cordomi A, Brugarolas M, Moreno E, Aguinaga D, Perez-Benito L, et al. Cross-communication between Gi and Gs in a G-protein-coupled receptor heterotetramer guided by a receptor C-terminal domain. BMC Biol Authors declare no conflict of interest. 2018;16:24. [14] de Lera Ruiz M, Lim YH, Zheng J. Adenosine A2A receptor as a drug discovery target. J Med Chem 2014;57:3623–50. Funding [15] Gutierrez-de-Teran H, Sallander J, Sotelo E. Structure-based rational design of adenosine receptor ligands. Curr Top Med Chem 2017;17:40–58. This work was partially supported by grants from the Spanish [16] Domenici MR, Ferrante A, Martire A, Chiodi V, Pepponi R, Tebano MT, et al. Adenosine A2A receptor as potential therapeutic target in neuropsychiatric Ministry of Economy and Competitiveness (BFU2015-64405-R, disorders. Pharmacol Res 2019;147:104338. SAF2017-84117-R, RTI2018-094204-B-I00 and PID2019- [17] Dungo R, Deeks ED. Istradefylline: first global approval. Drugs 109240RB-I00; they may include FEDER funds), the Alzheimer’s 2013;73:875–82. [18] Wisler JW, Rockman HA, Lefkowitz RJ. Biased G protein-coupled receptor Association (AARFD-17-503612) and by a grant from Fundació la signaling: changing the paradigm of drug discovery. Circulation Marató de TV3 (201413-30). 2018;137:2315–7. [19] Valant C, May LT, Aurelio L, Chuo CH, White PJ, Baltos JA, et al. Separation of on-target efficacy from adverse effects through rational design of a bitopic CRediT authorship contribution statement adenosine receptor agonist. PNAS 2014;111:4614–9. [20] Smith JS, Lefkowitz RJ, Rajagopal S. Biased signalling: from simple switches to allosteric microprocessors. Nat Rev Drug Discov 2018;17:243–60. Gemma Navarro: Conceptualization, Validation, Supervision, [21] Burgueno J, Blake DJ, Benson MA, Tinsley CL, Esapa CT, Canela EI, et al. The Project administration, Resources, Writing - original draft, Funding adenosine A2A receptor interacts with the actin-binding protein alpha-actinin. acquisition. Angel Gonzalez: Validation, Supervision, Project J Biol Chem 2003;278:37545–52. administration, Resources. Stefano Campanacci: Investigation. [22] Medrano M, Aguinaga D, Reyes-Resina I, Canela EI, Mallol J, Navarro G, et al. Orexin A/hypocretin modulates leptin receptor-mediated signaling by Rafael Rivas-Santisteban: Investigation. Irene Reyes-Resina: allosteric modulations mediated by the ghrelin GHS-R1A receptor in Investigation. Nil Casajuana-Martin: Investigation. Arnau Cor- hypothalamic neurons. Mol Neurobiol 2018;55:4718–30. 2731 G. Navarro, A. Gonzalez, S. Campanacci et al. Computational and Structural Biotechnology Journal 18 (2020) 2723–2732 [23] Zhou XE, He Y, de Waal PW, Gao X, Kang Y, Van Eps N, et al. Identification of during family A G protein-coupled receptor activation. Mol Pharmacol phosphorylation codes for arrestin recruitment by G protein-coupled 2011;79:262–9. receptors. Cell 2017;170(457–469):e13. [48] Rasmussen SG, Choi HJ, Fung JJ, Pardon E, Casarosa P, Chae PS, et al. Structure [24] Martinez-Pinilla E, Rabal O, Reyes-Resina I, Zamarbide M, Navarro G, Sanchez- of a nanobody-stabilized active state of the beta(2) adrenoceptor. Nature Arias JA, et al. Two affinity sites of the cannabinoid subtype 2 receptor 2011;469:175–80. identified by a novel homogeneous binding assay. J Pharmacol Exp Ther [49] Troupiotis-Tsailaki A, Zachmann J, Gonzalez-Gil I, Gonzalez A, Ortega- 2016;358:580–7. Gutierrez S, Lopez-Rodriguez ML, et al. Ligand chain length drives activation [25] Cheng YC, Prusoff WH. Relationship between the inhibition constant (Ki) and of lipid G protein-coupled receptors. Sci Rep 2017;7:2020. the concentration of inhibitor which causes 50 per cent inhibition (IC50) of an [50] Izquierdo C, Gomez-Tamayo JC, Nebel JC, Pardo L, Gonzalez A. Identifying enzymatic reaction. Biochem Pharmacol 1973;22:3099–108. human diamine sensors for death related putrescine and cadaverine [26] Rajagopal S, Ahn S, Rominger DH, Gowen-MacDonald W, Lam CM, Dewire SM, molecules. PLoS Comput Biol 2018;14:e1005945. et al. Quantifying ligand bias at seven-transmembrane receptors. Mol [51] Llinas Del Torrent C, Casajuana-Martin N, Pardo L, Tresadern G, Perez-Benito L. Pharmacol 2011;80:367–77. Mechanisms underlying allosteric molecular switches of metabotropic [27] Black JW, Leff P. Operational models of pharmacological agonism. Proc R Soc glutamate receptor 5. J Chem Inf Model 2019;59:2456–66. Lond B 1983;220:141–62. [52] Park JH, Scheerer P, Hofmann KP, Choe HW, Ernst OP. Crystal structure of the [28] Kenakin T, Christopoulos A. Measurements of ligand bias and functional ligand-free G-protein-coupled receptor opsin. Nature 2008;454:183–7. affinity. Nat Rev Drug Discov 2013;12:483. [53] McAllister SD, Hurst DP, Barnett-Norris J, Lynch D, Reggio PH, Abood ME. [29] Webb B, Sali A. Comparative Protein Structure Modeling Using MODELLER. Structural mimicry in class A G protein-coupled receptor rotamer toggle Curr Protoc Bioinform 2014;47. 5 6 1-32. switches: the importance of the F3.36(201)/W6.48(357) interaction in [30] Labute P. Protonate3D: assignment of ionization states and hydrogen cannabinoid CB1 receptor activation. J Biol Chem 2004;279:48024–37. coordinates to macromolecular structures. Proteins 2009;75:187–205. [54] Krishna Kumar K, Shalev-Benami M, Robertson MJ, Hu H, Banister SD, [31] Rodriguez-Espigares I, Torrens-Fontanals M, Tiemann JKS, Aranda-Garcia D, Hollingsworth SA, et al. Structure of a signaling cannabinoid receptor 1-G Ramirez-Anguita JM, Stepniewski TM, et al. GPCRmd uncovers the dynamics of protein complex. Cell 2019;176(448–458):e12. the 3D-GPCRome. Nat Methods 2020;17:777–87. [55] Pellissier LP, Sallander J, Campillo M, Gaven F, Queffeulou E, Pillot M, et al. [32] Massink A, Gutierrez-de-Teran H, Lenselink EB, Ortiz Zacarias NV, Xia L, Conformational toggle switches implicated in basal constitutive and agonist- Heitman LH. Sodium ion binding pocket mutations and adenosine A2A induced activated states of 5-hydroxytryptamine-4 receptors. Mol Pharmacol receptor function. Mol Pharmacol 2015;87:305–13. 2009;75:982–90. [33] Liu W, Chun E, Thompson AA, Chubukov P, Xu F, Katritch V, et al. Structural [56] Ersoy BA, Pardo L, Zhang S, Thompson DA, Millhauser G, Govaerts C, et al. basis for allosteric regulation of GPCRs by sodium ions. Science Mechanism of N-terminal modulation of activity at the melanocortin-4 2012;337:232–6. receptor GPCR. Nat Chem Biol 2012;8:725–30. [34] Gutierrez-de-Teran H, Massink A, Rodriguez D, Liu W, Han GW, Joseph JS, et al. [57] Perez-Benito L, Doornbos MLJ, Cordomi A, Peeters L, Lavreysen H, Pardo L, et al. The role of a sodium ion binding site in the allosteric modulation of the A(2A) Molecular switches of allosteric modulation of the metabotropic glutamate 2 adenosine G protein-coupled receptor. Structure 2013;21:2175–85. receptor. Structure 2017;25(1153–1162):e4. [35] Cordomi A, Caltabiano G, Pardo L. Membrane protein simulations using [58] Rodriguez D, Pineiro A, Gutierrez-de-Teran H. Molecular dynamics simulations AMBER force field and berger lipid parameters. J Chem Theory Comput reveal insights into key structural elements of adenosine receptors. 2012;8:948–58. Biochemistry 2011;50:4194–208. [36] Jespers W, Schiedel AC, Heitman LH, Cooke RM, Kleene L, van Westen GJP, et al. [59] Deupi X, Olivella M, Sanz A, Dolker N, Campillo M, Pardo L. Influence of the g- Structural mapping of adenosine receptor mutations: ligand binding and conformation of Ser and Thr on the structure of transmembrane helices. J signaling mechanisms. Trends Pharmacol Sci 2018;39:75–89. Struct Biol 2010;169:116–23. [37] Ballesteros JA, Weinstein H. Integrated methods for the construction of three [60] Bertheleme N, Singh S, Dowell SJ, Hubbard J, Byrne B. Loss of constitutive dimensional models and computational probing of structure-function activity is correlated with increased thermostability of the human adenosine relations in G-protein coupled receptors. Methods Neurosci 1995;25:366–428. A2A receptor. Br J Pharmacol 2013;169:988–98. [38] Pardo L, Deupi X, Dolker N, Lopez-Rodriguez ML, Campillo M. The role of [61] Nobles KN, Xiao K, Ahn S, Shukla AK, Lam CM, Rajagopal S, et al. Distinct internal water molecules in the structure and function of the rhodopsin family phosphorylation sites on the beta(2)-adrenergic receptor establish a barcode of G protein-coupled receptors. ChemBioChem 2007;8:19–24. that encodes differential functions of beta-arrestin. Sci Signal 2011;4:ra51. [39] Venkatakrishnan AJ, Ma AK, Fonseca R, Latorraca NR, Kelly B, Betz RM, et al. [62] Yang F, Yu X, Liu C, Qu CX, Gong Z, Liu HD, et al. Phospho-selective Diverse GPCRs exhibit conserved water networks for stabilization and mechanisms of arrestin conformations and functions revealed by unnatural activation. Proc Natl Acad Sci USA 2019. amino acid incorporation and (19)F-NMR. Nat Commun 2015;6:8202. [40] Gomez-Tamayo JC, Cordomi A, Olivella M, Mayol E, Fourmy D, Pardo L. [63] Yang Z, Yang F, Zhang D, Liu Z, Lin A, Liu C, et al. Phosphorylation of G protein- Analysis of the interactions of sulfur-containing amino acids in membrane coupled receptors: from the barcode hypothesis to the flute model. Mol proteins. Protein Sci 2016;25:1517–24. Pharmacol 2017;92:201–10. [41] Liu JJ, Horst R, Katritch V, Stevens RC, Wuthrich K. Biased signaling pathways [64] Klein Herenbrink C, Sykes DA, Donthamsetti P, Canals M, Coudrat T, Shonberg in beta2-adrenergic receptor characterized by 19F-NMR. Science J, et al. The role of kinetic context in apparent biased agonism at GPCRs. Nat 2012;335:1106–10. Commun 2016;7:10842. [42] McCorvy JD, Butler KV, Kelly B, Rechsteiner K, Karpiak J, Betz RM, et al. [65] Grundmann M, Kostenis E. Temporal bias: time-encoded dynamic GPCR Structure-inspired design of beta-arrestin-biased ligands for aminergic GPCRs. signaling. Trends Pharmacol Sci 2017;38:1110–24. Nat Chem Biol 2018;14:126–34. [66] Jensen DD, Lieu T, Halls ML, Veldhuis NA, Imlach WL, Mai QN, et al. Neurokinin [43] Lane JR, Klein Herenbrink C, van Westen GJ, Spoorendonk JA, Hoffmann C, 1 receptor signaling in endosomes mediates sustained nociception and is a IJzerman AP. A novel nonribose agonist, LUF5834, engages residues that are viable therapeutic target for prolonged pain relief. Sci Transl Med 2017;9. distinct from those of adenosine-like ligands to activate the adenosine A(2a) [67] Guo D, Mulder-Krieger T, Ijzerman AP, Heitman LH. Functional efficacy of receptor. Mol Pharmacol 2012;81:475–87. adenosine A(2)A receptor agonists is positively correlated to their receptor [44] Wacker D, Wang C, Katritch V, Han GW, Huang XP, Vardy E, et al. Structural residence time. Br J Pharmacol 2012;166:1846–59. Features for Functional Selectivity at Serotonin Receptors. Science [68] Yun Y, Chen J, Liu R, Chen W, Liu C, Wang R, et al. Long residence time 2013;340:615–9. adenosine A1 receptor agonists produce sustained wash-resistant antilipolytic [45] Venkatakrishnan AJ, Deupi X, Lebon G, Tate CG, Schertler GF, Babu MM. effect in rat adipocytes. Biochem Pharmacol 2019;164:45–52. Molecular signatures of G-protein-coupled receptors. Nature [69] Segala E, Guo D, Cheng RK, Bortolato A, Deflorian F, Dore AS, et al. Controlling 2013;494:185–94. the dissociation of ligands from the adenosine A2A receptor through [46] Huang W, Manglik A, Venkatakrishnan AJ, Laeremans T, Feinberg EN, Sanborn modulation of salt bridge strength. J Med Chem 2016;59:6470–9. AL, et al. Structural insights into micro-opioid receptor activation. Nature [70] Guo D, Pan AC, Dror RO, Mocking T, Liu R, Heitman LH, et al. Molecular basis of 2015;524:315–21. ligand dissociation from the adenosine A2A receptor. Mol Pharmacol [47] Sansuk K, Deupi X, Torrecillas IR, Jongejan A, Nijmeijer S, Bakker RA, et al. A 2016;89:485–91. structural insight into the reorientation of transmembrane domains 3 and 5 2732