Structure-Guided Screening for Functionally Selective D2 Dopamine

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES

Article 2

Structure-guided screening for functionally selective D dopamine receptor ligands from a virtual chemical library Barbara Männel, Mariama Jaiteh, Alexey Zeifman, Alena Randakova, Dorothee Möller, Harald Hübner, Peter Gmeiner, and Jens Carlsson ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.7b00493 • Publication Date (Web): 28 Aug 2017 Downloaded from http://pubs.acs.org on September 1, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Structure-guided screening for functionally selective D2 dopamine receptor ligands from a virtual chemical library

Barbara Männel1, Mariama Jaiteh2, Alexey Zeifman2, Alena Randakova1, Dorothee Möller1, Harald Hübner1, Peter Gmeiner1,*, and Jens Carlsson2,*

1

Department of Chemistry and Pharmacy, Medicinal Chemistry, Friedrich-Alexander University,

Schuhstraße 19, 91052 Erlangen, Germany. 2

Science for Life Laboratory, Department of Cell and Molecular Biology, Uppsala University,

BMC, Box 596, SE-751 24 Uppsala, Sweden.

KEYWORDS: GPCR, dopamine receptor, biased signaling, molecular docking, virtual screening, virtual chemical library, functional selectivity

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 35

ABSTRACT Functionally selective ligands stabilize conformations of G protein-coupled receptors (GPCRs) that induce a preference for signaling via a subset of the intracellular pathways activated by the endogenous agonists. The possibility to fine-tune the functional activity of a receptor provides opportunities to develop drugs that selectively signal via pathways associated with a therapeutic effect and avoid those causing side effects. Animal studies have indicated that ligands displaying functional selectivity at the D2 dopamine receptor (D2R) could be safer and more efficacious drugs against neuropsychiatric diseases. In this work, computational design of functionally selective D2R ligands was explored using structure-based virtual screening. Molecular docking of known functionally selective ligands to a D2R homology model indicated that such compounds were anchored by interactions with the orthosteric site and extended into a common secondary pocket. A tailored virtual library with close to 13,000 compounds bearing 2,3dichlorophenylpiperazine, a privileged orthosteric scaffold, connected to diverse chemical moieties via a linker was docked to the D2R model. Eighteen top-ranked compounds that occupied both the orthosteric and allosteric site were synthesized, leading to the discovery of 16 partial agonists. A majority of the ligands had comparable maximum effects in the G protein and β-arrestin recruitment assays, but a subset displayed preference for a single pathway. In particular, compound 4 stimulated β-arrestin recruitment (EC50 = 320 nM, Emax = 16%), but had no detectable G protein signaling. The use of structure-based screening and virtual libraries to discover GPCR ligands with tailored functional properties will be discussed.


2

Page 3 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The last decade has seen an explosion of atomic resolution information for G protein-coupled receptors (GPCRs), a large superfamily of pharmaceutically relevant membrane proteins.1 Highresolution crystal structures have revealed inactive receptor conformations stabilized by potent drug candidates and active states with both agonist and G protein bound.2 Encouragingly, structure-based virtual screening against GPCR crystal structures3 has successfully identified both agonists and antagonists of therapeutic targets4-6 suggesting that binding site conformations captured by crystallography can guide the design of compounds with tailored signaling properties.

It has become increasingly clear that GPCR signaling is more complex than the initially proposed two-state model with distinct active and inactive conformations. GPCRs exist in a large spectrum of states and the population of these can be fine-tuned by extracellular ligands.7,

8

Functionally selective ligands are able to specifically modulate distinct, G protein dependent and independent, signaling pathways by stabilizing diverse receptor conformations.9 Ligands exhibiting signaling bias are not only relevant as molecular probes to understand GPCR function, but could also provide opportunities to develop more efficient drugs with less side effects.10-12 For example, respiratory depression and other severe side effects of opioid analgesics have been linked to β-arrestin recruitment in animal models.13 The potential to translate this finding to drug development was confirmed by the structure-guided discovery of a G protein biased µ-opioid receptor ligand exhibiting none of the adverse effects associated with β-arrestin mediated signaling in animal models.14 Conversely, the favorable therapeutic properties of the β adrenergic receptor drug carvedilol compared to other β-blockers have been attributed to its capability to promote β-arrestin related signaling.15 In the case of the D2 dopamine receptor (D2R), signaling


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 35

bias has been demonstrated to be relevant for treatment of both neurodegenerative and neuropsychiatric diseases. Evidence from animal studies supports that β-arrestin recruitment via the D2R may have beneficial effects in treatment of Parkinson’s disease.16 Moreover, partial agonists of the D2R favoring either β-arrestin recruitment or G protein signaling have been reported to exert antipsychotic effect, underpinning the therapeutic relevance of functionally selective ligands.17-20

1,4-disubstituted aromatic piperazine (1,4-DAP) scaffolds have been identified as privileged structures for dopamine receptors and related aminergic GPCRs (Figure 1a).21 Such compounds are typically composed of three moieties that contribute to their biological activity.22-24 The 1,4DAP head group serves as primary recognition element by anchoring the compound in the orthosteric binding site through electrostatic interactions with residue Asp3.32 (superscripts refer to Ballesteros-Weinstein numbering),25 which is strongly conserved among aminergic GPCRs. This pharmacophore is then linked to a second (e.g. heterocyclic) moiety via an aliphatic spacer, which influences subtype selectivity and contributes to improvement of affinity. The linker allows the ligand to extend into secondary binding pockets, which modulates the efficacy and functionally selective properties of these compounds.26-29 Dopaminergic ligands (Figure 1a) with a pyrazolo[1,5-a]pyridine connected to the spacer have been found to be biased towards G protein signaling and to antagonize β-arrestin recruitment.29, 30 In contrast, replacing this moiety by a benzothiazole resulted in the inverted functional behavior with recruitment of β-arrestin and no activation of G protein signaling.31 However, understanding of biased signaling at the dopamine receptors is limited as only a small set of functionally selective ligands are currently known and no crystal structures of complexes with relevant compounds have been determined.


4

Page 5 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In this study, we used homology modeling of the D2R and structure-based virtual screening to explore if a combination of these techniques could be used to discover ligands with functional selectivity. Molecular docking to a D2R model was first used to predict binding modes of G protein and β-arrestin biased ligands with the goal to identify secondary binding pockets influencing functional selectivity. In a second step, a tailor-made virtual library of synthetically tractable compounds based on a 2,3-dichlorophenylpiperazine scaffold was screened against the D2R model and a set of 18 top-ranked compounds was selected for synthesis. The functional properties of the synthesized compounds were determined by measuring [35S]GTPγS binding to assess G protein-mediated signaling and their ability to promote the recruitment of β-arrestin 2. The structural basis of functional selectivity and ligand discovery from virtual chemical libraries will be discussed.

RESULTS AND DISCUSSION Modeling of the D2R and molecular docking of functionally selective ligands. The atomic resolution structure of the D2R was predicted with MODELLER32 using the crystal coordinates of the D3R subtype (PDB accession code: 3PBL)33. The structural variation among the 100 generated homology models was relatively small, reflecting the high sequence identity between the D2- and D3R subtypes (Figure S1). Molecular docking screening with DOCK3.634 was carried out against ten D2R models. Two sets of compounds were docked to the homology models to assess their ability to recognize D2R ligands, in particular arylpiperazines as the focus of this work was to design novel ligands based on this scaffold. A set of 70 ligands containing an arylpiperazine scaffold together with property-matched decoys35 was first evaluated. Strong enrichments of ligands over decoys were obtained for all D2R models, demonstrating that the


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 35

predicted binding sites captured the features responsible for recognition of this scaffold and a representative structure was selected for further studies (Figure S2). In order to gain more specific insights into the structural-basis of functional selectivity, 51 compounds with experimentally determined bias toward either G protein (33 compounds) or β-arrestin (18 compounds) signaling pathways were docked to the homology models.29-31,

36, 37

As these

compound sets were limited in size and considering that functional selectivity has been demonstrated to show some dependence on assay conditions,38 the main goal of these docking calculations was to discover secondary binding pockets that could influence D2R signaling rather than identifying receptor conformations responsible for activation of a specific pathway. In agreement with the results obtained for the arylpiperazine ligands, both sets of biased compounds were strongly enriched over property-matched decoys (Figure S2). The ligands were anchored in the orthosteric site by a salt bridge between the charged amine group of the compounds and the conserved Asp1143.32 in transmembrane helix 3 (TM3). A flexible linker connected the anchoring scaffold, which was typically a 1,4-DAP-like moiety, to a group that could access secondary binding pockets. The main secondary binding site was predicted to be located between TM1, TM2, and TM7 (Figure 1b).23, 26 The polar contacts between the biased ligands and amino acid residues in the secondary pocket (e.g. Glu952.65, Ser4097.36, Tyr4087.37, Thr4127.41, and Tyr4167.45) were analyzed. Glu952.56 and Ser4097.36 were found to frequently interact with functionally selective ligands, in particular β-arrestin biased ligands, suggesting that design of molecules that interact with residues in the secondary pocket could lead to discovery of ligands with diverse signaling properties.


6

Page 7 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Design and molecular docking screening of a virtual chemical library. In order to explore the possibility to identify D2R ligands with functional selectivity, a focused virtual compound library was generated based on a 2,3-dichlorophenylpiperazine head group connected to either oxobutyl or amidopropyl spacers (Figure 2), which would allow interactions with the identified secondary binding pocket. Commercially available building blocks that could be attached to the other end of the spacer groups via synthesis were extracted from the ZINC database39 and were filtered to enable rapid actual synthesis of compounds and molecular weight ≤ 450 Da (for the detailed workflow, see supporting methods).23 In the final libraries, 5753 and 7232 compounds were based on the oxybutyl and amidopropyl linker, respectively.

The libraries based on the oxybutyl and amidopropyl spacer, which contained 12985 unique structures, were docked to the selected D2R model (Figure S2) and the 300 top-ranked compounds of each set were visually inspected. Molecules were only considered for synthesis if the piperazine moiety fitted well into the orthosteric site and formed a salt bridge with Asp1143.32. In addition, formation of polar contacts with Glu952.65 was also considered in the selection process as this interaction was predicted to influence the functional selectivity. A set of 18 compounds (1–18, Tables 1 and 2) was chosen for synthesis. The predicted ligands extended into the secondary binding pocket with a set of diverse moieties that differed in terms of size, shape, and polarity (Figure 3). All predicted compounds passed a filter for pan assay interference compounds (PAINS), reducing the risk of obtaining false positives in the experimental assays.40

Synthesis of predicted D2R ligands. All the 18 compounds that were selected were successfully synthesized. In the case of the compounds based on an oxybutyl linker (1–9), synthesis of the


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 35

common building block, 4-bromo-1-[4-(2,3-dichlorophenyl)piperazin-1-yl]butane, failed and alternative protocols were employed to obtain the desired products (Scheme S1). For the amidopropyl derivatives 10–18, synthesis was carried out according to the reaction used to generate the virtual library (Scheme S2). In some cases, synthesis was initiated from commercially available precursors of the building blocks used to generate the virtual library as these could readily be transformed into the desired compounds. Detailed synthetic procedures are described in supporting methods.

Biological activity. In order to avoid downstream signal amplification, two proximal test systems were selected to evaluate the capacity of the synthesized compounds to act as D2R agonists in G protein and β-arrestin signaling pathways. Canonical G protein activation was assessed in [35S]GTPγS incorporation experiments with membranes from HEK293T cells coexpressing the D2SR subtype and a Gαo1 subunit whereas the ability to recruit β-arrestin 2 was characterized using the DiscoverX Pathhunter assay. HEK293 cells stably expressing a β-arrestin 2/β-galactosidase fragment fusion protein were transfected with D2SR-tagged with a complementary β-galactosidase fragment. In addition, GRK2 was co-expressed as the presence of this kinase was recently shown to foster partial agonist-mediated recruitment of arrestins to Gi/o-coupled receptors.18, 41

All synthesized compounds were first tested at a single high concentration of 10 µM to estimate Emax values (Tables 1 and 2). The measured activities were compared to the reference agonist quinpirole and the β-arrestin biased agonist UNC0006 (Figure 1a). Quinpirole was a full agonist (Emax = 100%) with a balanced signaling profile whereas UNC0006 displayed activities of 7±1%


8

Page 9 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


in the GTPγS assay and 18±1% in the β-arrestin 2 recruitment assay, which was in good agreement with literature data.18 Compounds with an oxybutylene linker showed a spectrum of signaling profiles and reached up to roughly a third of the effect of the reference agonist in both assays. Compounds 2 and 6 were inactive in both the GTPγS and β-arrestin 2 assays at 10 µM. Compound 1, the benzamidine containing compound 3, and piperidyl analog compound 5 had substantial Emax values and these were similar in both tested assays. Of these, compound 3 showed the largest effect with Emax values of 30% and 31% in the GTPγS and β-arrestin 2 assays, respectively. Compounds 4 and 7–9 appeared to have some preference for either G protein (7–9) or β-arrestin (4) signaling. For example, compound 9 was a partial agonist (Emax = 10%) in the GTPγS assay, but did not show any activity in the β-arrestin 2 assay. In contrast, compound 4 was active in the β-arrestin 2 recruitment assay (Emax = 12%), but showed no activity in the GTPγS assay. Compounds based on the amidopropyl linker (10–18) acted as partial agonists in both GTPγS assay and β-arrestin 2 assays. The phenol-based compound 10 acted as a partial agonist with similar activities in both test systems (Emax = 18% and 16% for the recruitment of βarrestin 2 and GTPγS binding, respectively). Compounds 11–13, bearing substituted-pyrazole moieties of different shape and polarity, exhibited partial agonist activity with Emax values ranging from 8 to 31% in the β-arrestin 2 assay and similar activities were observed for G protein signaling. Similarly, compounds 15–16, which both contained imidazole moieties, and the Nalkylamino cyclobutyl analogs 17–18 acted as partial agonists with no clear preference for one signaling pathway over the other. In contrast, the triazole-containing compound 14 was a partial agonist in the GTPγS assay (Emax = 15%), but almost completely inactive for the recruitment of β-arrestin 2 (Emax = 3.6%).


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 35

The results of the single point activity measurements (Figure 4) demonstrated that the majority of the synthetized compounds did not show clear preference for G protein or β-arrestin signaling. Dose-dependent activation assays were performed for a subset of the discovered ligands and particular focus was put on compounds that showed activity in the β-arrestin 2 recruitment assay. Six compounds that either showed an efficacy greater than 20% in the β-arrestin 2 recruitment assay (3, 5, 11, 12, and 17) or a difference in efficacy between the two signaling pathways greater than 5% together with a preference for β-arrestin 2 recruitment (4) were selected. Dose response curves were also obtained for the reference ligands quinpirole and β-arrestin biased UNC0006 (Table 3). UNC0006 displayed almost no activity in the GTPγS assay (EC50 = 1600 nM, Emax = 7.0%) and showed partial agonist activity with moderate potency (EC50 = 470 nM, Emax = 17%) for the recruitment of β-arrestin 2 (Figure 5a). Compound 4, which was predicted to interact with the secondary pocket with an imidazole moiety, had a similar signaling profile. Compound 4 did not promote G protein binding (Figure 5b) and exhibited a potency and efficacy (EC50 = 320 nM, Emax = 16%) similar to UNC0006 in the β-arrestin 2 assay. Compounds 3, 5, 11, 12, and 17 displayed good to moderate potencies (EC50 = 30 to 600 nM) in the β-arrestin assay and acted as partial agonists with Emax values up to 40%. In all cases, the compounds were less efficacious in the GTPγS assay and the maximal effects ranged between 12 and 24%. The efficacies of the compounds were higher in both the β-arrestin 2 recruitment and GTPγS assays compared to UNC0006 and three ligands (3, 5, and 17) had twofold higher Emax values for the former signaling pathway. Compound 3 was a potent partial agonist (EC50 = 54 nM, Emax = 23%) for the activation of G protein coupling and displayed the highest observed ligand efficacy in the β-arrestin assay, but with moderate potency (EC50 = 280 nM, Emax = 40%). Compounds 5 and 17 exhibited similar potencies in both assays with EC50 = 30 and 220 nM for β-arrestin 2


10

Page 11 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


recruitment and EC50 = 64 and 330 nM for G protein coupling, respectively (Figure 5c). In both cases, the efficacies were more than twofold higher in the β-arrestin assay (Emax = 32 and 28% for β-arrestin 2 recruitment and Emax = 13 and 12% for the GTPγS assay).

Discussion. The main result of this work is the discovery of novel ligands by integrating library design, structure-based compound prioritization, chemical synthesis, and biological testing. The cornerstones of our approach are automated design of virtual chemical libraries and structurebased screening to select compounds for experimental evaluation. We applied this strategy to predict functionally selective D2R ligands. Synthesis and testing of 18 compounds in experiments assessing G protein or β-arrestin mediated signaling resulted in the discovery of 16 ligands. These compounds displayed a spectrum of activities for the two pathways and several showed preference for one signaling pathway. The strength of the proposed approach is the possibility to rapidly explore a large number of readily synthesizable compounds, but rational discovery of functionally selective ligands remains limited by the lack of knowledge regarding the structural basis of biased signaling.

The fact that GPCR ligands do not uniformly activate signaling pathways has sparked interest in rational design of functional selectivity.42, 43 In the case of the D2R, functionally selective ligands are typically larger than balanced reference agonists (e.g. dopamine and ropinirole) and have been predicted to achieve their unique signaling properties by interacting with the orthosteric site and secondary (allosteric) pockets at the same time.28,

29, 44-46

Given the high molecular


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 35

complexity of such “bitopic” scaffolds, the chemical space for this compound class is likely not covered well in commercial chemical space. Functionally selective ligands may hence be more rare than balanced agonists in general chemical libraries and less likely to emerge from highthroughput screens.47 An alternative design route is iterative medicinal chemistry optimization based on D2R ligands that possess a subset of the desired signaling properties.28,

44-46

For

example, the antipsychotic drug aripiprazole has been used as a starting point for generation of D2R ligands with different signaling profiles.18, 19, 29, 31, 37, 48 In the present work, bitopic D2R ligands targeting the orthosteric as well as secondary binding pockets were designed by screening of virtual chemical libraries, which allowed consideration of a larger number of compounds for synthesis than possible using a traditional medicinal chemistry approach. A phenylpiperazine scaffold was selected as the core element of a focused library as this ligand has been demonstrated to bind to the orthosteric site with high affinity.26 Since such bitopic phenylpiperazine analogs are not readily commercially available, we created a virtual library with synthesizable compounds that were able to access allosteric pockets by connecting the orthosteric scaffold via two different linkers to diverse moieties from commercial sources. An alternative strategy, which was recently used by Vass et al. to design selective dopamine receptor ligands,49 is to carry out separate docking screens of a chemical library against the orthosteric and secondary binding pocket, followed by identification of chemical linkers to connect two compounds. This technique explores orders of magnitude larger numbers of possible ligands due to the combinatorial explosion resulting from combining any two commercial fragments. However, it is challenging from a practical viewpoint as an optimal linker that allows the two fragments to maintain their predicted binding modes has to be identified for each compound pair. In this work, the screening library was restricted to a single orthosteric ligand and two linkers to


12

Page 13 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


virtually connect the scaffold to building blocks, which gave us the advantage of screening the final product directly. These criteria resulted in a focused library with close to 13,000 compounds, which could easily be expanded by considering additional reactions, linkers, and orthosteric ligand scaffolds. As synthesizing the entire focused library would be unfeasible, molecular docking screening against a homology model was used to guide compound selection. Encouragingly, the 18 predicted ligands were successfully synthesized and 16 of these displayed activity for either one or both the β-arrestin and G protein signaling pathways (Emax ≥ 10%). The hit rate from the docking screen was excellent, but this result was to a large extent due to the use of the potent phenylpiperazine scaffold. What was more encouraging was that targeting the secondary binding pocket resulted in ligands with different signaling profiles, both in terms of the level of activation and bias. The discovered ligands were partial agonists with up to 40% of the activity of the reference agonist for both pathways and overall there was a strong linear correlation between the Emax values from β-arrestin and G protein signaling with a slope close to unity. This result demonstrated that increasing activation of one pathway typically also led to more signaling via the second pathway and that a majority of the synthesized compounds hence exhibited a balanced signaling profile. Compound 4 was a notable exception as it had similar signaling profile and potency as the reference compound UNC0006, which was one of the first discovered β-arrestin biased D2R compounds.18 This partial agonist had an EC50 value of 320 nM (Emax = 16%) in the β-arrestin recruitment assay, but no detectable G protein signaling. Similarly, compound 5 and 17 displayed some signaling preference for the β-arrestin pathway with improved potency compared to compound 4 and UNC0006. These compounds hence represent starting-points for development of chemical probes for understanding functional selectivity and could contribute to development of novel dopaminergic antipsychotics with less side effects.17, 18,


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

48

Page 14 of 35

It should be acknowledged that our studies on D2R activation were performed in cells

overexpressing D2R and either the G protein subunit Gαo1 or β-arrestin-2 and GRK2. How these results can be translated into in vivo outcomes and whether the ligands are able to stimulate G protein activation or β-arrestin recruitment in native D2R-expressing neurons will require further experiments.

Advances in two areas could further improve predictions of functional selectivity using structurebased modeling. First of all, experimental quantification of signaling bias is still challenging because biological responses appear to be dependent on assay conditions, cellular environment, and kinetics.38 Reliable structure-functional selectivity relationships and multiple scaffolds will be crucial for identification of relevant receptor-ligand interactions at the molecular level. In this work, ligands displaying bias towards either G protein or β-arrestin mediated signaling were docked to a homology model to identify allosteric pockets responsible for functional selectivity. In agreement with previous studies,27-29 our results indicated that functionally selective compounds interacted with a secondary pocket formed by TM1, TM2, and TM7. This predicted binding mode is in agreement with the crystal structure of the β1 adrenergic receptor in complex with the biased agonist carvedilol, which extends from the orthosteric site into the same allosteric pocket.50 Our models pointed to that specific interactions in this region may determine functional selectivity and several β-arrestin biased ligands were found to form hydrogen bonds with Glu952.65. The novel ligands discovered from our prospective screen were all predicted to interact with Glu952.65, but only a few showed preference for a specific signaling pathway, indicating that either interactions with the residue are not sufficient to stabilize the relevant receptor conformation or that the compounds were simply not able to form this hydrogen bond.


14

Page 15 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The challenges involved in predicting functionally selective ligands is further highlighted by the high chemical similarity between ligands that exhibit biased and balanced signaling profiles,18, 19, 28, 29, 31, 37, 44, 45, 48

which suggests that small perturbations to the binding site can stabilize distinct

intracellular conformations. Considering the small binding site differences in crystal structures of the active and inactive β2 adrenergic receptor,51 the conformational changes connected activation of a specific pathway may be subtle and hence difficult to capture with the static receptor structures used in molecular docking applications.

The second factor limiting design of functionally selective ligands is the lack of high-resolution crystal structures of the D2R. Returning to the example of the β2 adrenergic receptor, Weiss et al. used molecular docking screens against a crystal structure determined in an active conformation to identify ligands from a commercial chemical library and, remarkably, the discovered compounds recapitulated the biased signaling profile of the co-crystallized agonist.6 Virtual screening against homology models has been successful in identifying ligands of dopamine receptors and related GPCRs,52 but such structures may not necessarily capture the interactions determining signaling or selectivity properties53,

54

and, in our particular case, we were also

limited to modeling active receptor conformations based on a structure determined in an inactive state. D2R crystal structures, ideally in complex with both functionally selective and balanced ligands, may enable more successful screening for lead candidates with biased signaling properties. As the number of GPCR crystal structures is increasing rapidly, the combination of virtual screening libraries and structure-based docking can readily be applied to guide discovery of ligands to many therapeutic targets.


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 35

METHODS Homology modeling, molecular docking, and virtual library generation. The D2R homology models were generated using MODELLER32 based on a crystal structure of the D3R (PDB code 3PBL33). All docking screens against the D2R homology models were performed with the program DOCK3.6.34 The virtual chemical library was generated with the JChem software from ChemAxon (JChem version 5.11.4, http://www.chemaxon.com). For detailed computational methods, see supporting methods. Chemistry. Detailed synthetic procedures for compounds 1–18 are described in supporting methods. [35S]GTPγS binding and β -arrestin 2 recruitment assays. The [35S]GTPγS binding assay was performed with membrane preparations from HEK-293T cells which transiently coexpressed the dopamine receptor D2S and the Gαo protein. The measurement of receptor stimulated β-arrestin 2 recruitment was performed using the PathHunter assay (DiscoverX, Birmingham, U.K.) according to the manufacturer’s protocol. For detailed experimental methods, see supporting methods.

AUTHOR INFORMATION Corresponding author’s email: Jens Carlsson ([email protected]) Peter Gmeiner ([email protected])


16

Page 17 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ACKNOWLEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft (DFG) grant GRK1910 to P.G. and by the Swedish Research Council (2013-5708), the Science for Life Laboratory, and Åke Wibergs Stiftelse (M15-0287, M16-0233) to J.C. Computational resources were provided by the Swedish National Infrastructure for Computing. We thank OpenEye Scientific Software for the use of OEChem and OMEGA at no cost. J.C. and P.G. participate in the European COST Action CM1207 (GLISTEN).

ASSOCIATED CONTENT Supporting information. Detailed experimental and computational methods, Supporting Tables, Figures, Schemes, and 1H and 13C NMR spectra for the synthesized compounds. This material is available free of charge via the Internet at http://pubs.acs.org.

ABBREVIATIONS 1,4-DAP, 1,4-disubstituted-aromatic piperazine; DOPE, discrete optimized protein energy; DxR, Dx dopamine receptor; Gαx, G protein alpha subunit subtype x; GRK2, G protein-coupled receptor kinase subtype 2; GTPγS, guanosine 5’-O-(thiotriphosphate); PAINS, pan assay interference compounds; SEM, standard error of the mean; TMX, transmembrane helix X; UNC0006, 7-{4-[4-(2,3-dichlorophenyl)-1,4-diazepan-1-yl]butoxy}-3,4-dihydroquinolin-2(1H)one.


17


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 35

TABLES AND FIGURES Table 1. Structures and functional properties at the D2SR of ligands with oxybutyl spacer. The general structure of the compounds is shown in Figure 2. [35S]GTPγS bindinga

β-arrestin 2 recruitmentb

docking rank

Emax [%]c

Emax [%]c

1

169

10 ± 1

7.8 ± 0.5

2

64

3±2

-0.36 ± 0.7

3

65

30 ± 1

31 ± 2

4

28

1 ± 0.4

12 ± 2

5

130

24 ± 2

24 ± 2

6

62

4±1

1.2 ± 1

7

3

11 ± 2

4.3 ± 2

8

7

14 ± 2

4.9 ± 2

9

31

10 ± 3

-0.1 ± 0.7

compd

R

a

Data represent mean values ± SEM of 6 to 7 individual experiments each performed in triplicate. bData represent mean values ± SEM of 3 to 5 individual experiments each performed in triplicate. c Emax displayed as mean values ± SEM are relative (%) to the maximal effect of quinpirole.


18

Page 19 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2. Structures and functional properties at the D2SR of ligands with amidopropyl spacer. The general structure of the compounds is shown in Figure 2. [35S]GTPγS bindinga


docking rank

Emax [%]c

Emax [%]c

10

82

16 ± 2

18 ± 1

11

181

29 ± 3

31 ± 2

12

28

33 ± 3

26 ± 2

13

140

14 ± 3

7.9 ± 2

14

50

15 ± 1

3.6 ± 0.6

15

146

13 ± 1

9.8 ± 0.4

16

176

23 ± 5

20 ± 1

17

252

19 ± 1

24 ± 2

18

79

16 ± 1

20 ± 2

compd

R

a

Data represent mean values ± SEM of 6 to 7 individual experiments each performed in triplicate. bData represent mean values ± SEM of 3 to 5 individual experiments each performed in triplicate. cEmax displayed as mean values ± SEM are relative (%) to the maximal effect of quinpirole.


19


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 35

Table 3. Intrinsic activities and potencies of compounds 3, 4, 5, 11, 12, and 17 determined by [35S]GTPγS binding and recruitment of β-arrestin 2 after stimulation of D2SRs. [35S]GTPγS bindinga


compd

EC50 [nM]

Emax [%]c

EC50 [nM]

Emax [%]c

quinpirole

98 ± 20

100 ± 0.4

18 ± 0.7

100 ± 3

UNC0006

1600 ± 750

7.0 ± 2

470 ± 250

17 ± 2

3

54 ± 28

23 ± 2

280 ± 25

40 ± 2

4.5 ± 2

320 ± 130

16 ± 1

d

4

n.d.

5

64 ± 33

13 ± 1

30 ± 4.8

32 ± 1

11

460 ± 190

20 ± 3

600 ± 150

34 ± 2

12

380 ± 140

24 ± 3

51 ± 6.8

27 ± 3

17

330 ± 100

12 ± 2

220 ± 49

28 ± 2

a

Data represent mean values ± SEM of four to twelve individual experiments each performed in triplicate bData represent mean values ± SEM of three to five individual experiments each performed in triplicate. cEmax are relative (%) to the maximal effect of quinpirole. dn.d. not determined.


20

Page 21 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. (a) General chemical structure of 1,4-DAP ligands and structures of the G protein biased ligand A and β-arrestin biased ligands B and C. (b) Predicted binding mode of a β-arrestin biased ligand in the D2R homology model.


21


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 35

Figure 2. Design of the virtual chemical library. To allow a convenient synthesis, building blocks bearing a hydroxyl or carboxylic acid group were selected from the ZINC database and were connected in silico to the 2,3-dichlorophenylpiperazine head group via nucleophilic substitution in case of the oxybutyl spacer or an amide coupling for the amidopropyl spacer. A set of 18 compounds (1–18) was selected from docking screens against a D2R homology model.


22

Page 23 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Predicted binding modes for compounds 3, 4, 5, 11, 12, and 17 (a–f) in the D2R homology model. The D2R is represented with grey cartoons and key residues are shown in sticks. All ligands are represented as sticks with yellow carbon atoms and hydrogen bonds are depicted as black dashed lines.


23


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 35

Figure 4. Comparison between the Emax of compounds 1–18 and UNC0006 in β-arrestin 2 recruitment and in GTPγS binding assays. Compounds 1–9 (oxybutylene linker) and UNC0006 are shown with filled green circles. The propyl carboxamide containing compounds 10–18 are represented with blue circles. Compound numbers are shown for the ligands with full doseresponse curves. The gray dashed line represents equal Emax values for both pathways.


24

Page 25 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Dose-dependent stimulation of β-arrestin 2 recruitment and [35S]GTPγS binding for (a) UNC0006, (b) compounds 4, and (c) 5. Data represent mean ± SEM from three to five independent experiments, each performed in triplicate.


25


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 35

REFERENCES (1) Katritch, V., Cherezov, V., and Stevens, R. C. (2013) Structure-Function of the G Protein– Coupled Receptor Superfamily, Annu. Rev. Pharmacol. Toxicol. 53, 531–556. (2) Rasmussen, S. G. F., DeVree, B. T., Zou, Y., Kruse, A. C., Chung, K. Y., Kobilka, T. S., Thian, F. S., Chae, P. S., Pardon, E., Calinski, D., Mathiesen, J. M., Shah, S. T. A., Lyons, J. A., Caffrey, M., Gellman, S. H., Steyaert, J., Skiniotis, G., Weis, W. I., Sunahara, R. K., and Kobilka, B. K. (2011) Crystal structure of the β2 adrenergic receptor-Gs protein complex, Nature 477, 549–555. (3) Rodríguez, D., Ranganathan, A., and Carlsson, J. (2015) Discovery of GPCR Ligands by Molecular Docking Screening: Novel Opportunities Provided by Crystal Structures, Curr. Top. Med. Chem. 15, 2484–2503. (4) Kolb, P., Rosenbaum, D. M., Irwin, J. J., Fung, J. J., Kobilka, B. K., and Shoichet, B. K. (2009) Structure-based discovery of β2-adrenergic receptor ligands, Proc. Natl. Acad. Sci. U.S.A. 106, 6843–6848. (5) Kooistra, A. J., Leurs, R., de Esch, I. J. P., and de Graaf, C. (2015) Structure-Based Prediction of G-Protein-Coupled Receptor Ligand Function: A β-Adrenoceptor Case Study, J. Chem. Inf. Model. 55, 1045–1061. (6) Weiss, D. R., Ahn, S., Sassano, M. F., Kleist, A., Zhu, X., Strachan, R., Roth, B. L., Lefkowitz, R. J., and Shoichet, B. K. (2013) Conformation Guides Molecular Efficacy in Docking Screens of Activated β-2 Adrenergic G Protein Coupled Receptor, ACS Chem. Biol. 8, 1018–1026.


26

Page 27 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(7) Liu, J. J., Horst, R., Katritch, V., Stevens, R. C., and Wüthrich, K. (2012) Biased Signaling Pathways in β2-Adrenergic Receptor Characterized by 19F-NMR, Science 335, 1106–1110. (8) Manglik, A., Kim, Tae H., Masureel, M., Altenbach, C., Yang, Z., Hilger, D., Lerch, Michael T., Kobilka, Tong S., Thian, Foon S., Hubbell, Wayne L., Prosser, R. S., and Kobilka, Brian K. (2015) Structural Insights into the Dynamic Process of β2-Adrenergic Receptor Signaling, Cell 161, 1101–1111. (9) Rajagopal, K., Lefkowitz, R. J., and Rockman, H. A. (2005) When 7 transmembrane receptors are not G protein-coupled receptors, J. Clin. Invest. 115, 2971–2974. (10) Whalen, E. J., Rajagopal, S., and Lefkowitz, R. J. (2011) Therapeutic potential of β-arrestinand G protein-biased agonists, Trends Mol. Med. 17, 126–139. (11) Kenakin, T., and Christopoulos, A. (2013) Signalling bias in new drug discovery: detection, quantification and therapeutic impact, Nat. Rev. Drug Discov. 12, 205–216. (12) Shonberg, J., Lopez, L., Scammells, P. J., Christopoulos, A., Capuano, B., and Lane, J. R. (2014) Biased agonism at G protein-coupled receptors: The promise and the challenges - a medicinal chemistry perspective, Med. Res. Rev 34, 1286–1330. (13) Raehal, K. M., Walker, J. K. L., and Bohn, L. M. (2005) Morphine Side Effects in βArrestin 2 Knockout Mice, J. Pharmacol. Exp. Ther. 314, 1195–1201. (14) Manglik, A., Lin, H., Aryal, D. K., McCorvy, J. D., Dengler, D., Corder, G., Levit, A., Kling, R. C., Bernat, V., Hübner, H., Huang, X.-P., Sassano, M. F., Giguère, P. M., Löber, S., Da, D., Scherrer, G., Kobilka, B. K., Gmeiner, P., Roth, B. L., and Shoichet, B. K. (2016) Structure-based discovery of opioid analgesics with reduced side effects, Nature 537, 185–190.


27


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 35

(15) Wisler, J. W., DeWire, S. M., Whalen, E. J., Violin, J. D., Drake, M. T., Ahn, S., Shenoy, S. K., and Lefkowitz, R. J. (2007) A unique mechanism of β-blocker action: Carvedilol stimulates β-arrestin signaling, Proc. Natl. Acad. Sci. U.S.A. 104, 16657–16662. (16) Urs, N. M., Bido, S., Peterson, S. M., Daigle, T. L., Bass, C. E., Gainetdinov, R. R., Bezard, E., and Caron, M. G. (2015) Targeting β-arrestin2 in the treatment of L-DOPA–induced dyskinesia in Parkinson’s disease, Proc. Natl. Acad. Sci. U.S.A. 112, E2517–E2526. (17) Park, S. M., Chen, M., Schmerberg, C. M., Dulman, R. S., Rodriguiz, R. M., Caron, M. G., Jin, J., and Wetsel, W. C. (2016) Effects of β-Arrestin-Biased Dopamine D2 Receptor Ligands on Schizophrenia-Like Behavior in Hypoglutamatergic Mice, Neuropsychopharmacology 41, 704–715. (18) Allen, J. A., Yost, J. M., Setola, V., Chen, X., Sassano, M. F., Chen, M., Peterson, S., Yadav, P. N., Huang, X. P., Feng, B., Jensen, N. H., Che, X., Bai, X., Frye, S. V., Wetsel, W. C., Caron, M. G., Javitch, J. A., Roth, B. L., and Jin, J. (2011) Discovery of beta-arrestin-biased dopamine D2 ligands for probing signal transduction pathways essential for antipsychotic efficacy, Proc. Natl. Acad. Sci. U.S.A. 108, 18488–18493. (19) Möller, D., Banerjee, A., Uzuneser, T. C., Skultety, M., Huth, T., Plouffe, B., Hübner, H., Alzheimer, C., Friedland, K., Müller, C. P., Bouvier, M., and Gmeiner, P. (2017) Discovery of G Protein-Biased Dopaminergics with a Pyrazolo[1,5-a]pyridine Substructure, J. Med. Chem. 60, 2908–2929. (20) Masri, B., Salahpour, A., Didriksen, M., Ghisi, V., Beaulieu, J. M., Gainetdinov, R. R., and Caron, M. G. (2008) Antagonism of dopamine D2 receptor/beta-arrestin 2 interaction is a


28

Page 29 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


common property of clinically effective antipsychotics, Proc. Natl. Acad. Sci. U.S.A. 105, 13656–13661. (21) Welsch, M. E., Snyder, S. A., and Stockwell, B. R. (2010) Privileged scaffolds for library design and drug discovery, Curr. Opin. Chem. Biol. 14, 347–361. (22) Löber, S., Hübner, H., Tschammer, N., and Gmeiner, P. (2011) Recent advances in the search for D3- and D4-selective drugs: probes, models and candidates, Trends Pharmacol. Sci. 32, 148–157. (23) Ehrlich, K., Götz, A., Bollinger, S., Tschammer, N., Bettinetti, L., Härterich, S., Hübner, H., Lanig, H., and Gmeiner, P. (2009) Dopamine D2, D3, and D4 selective phenylpiperazines as molecular probes to explore the origins of subtype specific receptor binding, J. Med. Chem. 52, 4923–4935. (24) Loaiza, P. R., Löber, S., Hübner, H., and Gmeiner, P. (2006) Click Chemistry on Solid Phase: Parallel Synthesis of N-Benzyltriazole Carboxamides as Super-Potent G-Protein Coupled Receptor Ligands, J. Comb. Chem. 8, 252–261. (25) Ballesteros, J. A., and Weinstein, H. (1995) [19] Integrated methods for the construction of three-dimensional models and computational probing of structure-function relations in G proteincoupled receptors, In Methods in Neurosciences (Stuart, C. S., Ed.), pp 366–428, Academic Press. (26) Newman, A. H., Beuming, T., Banala, A. K., Donthamsetti, P., Pongetti, K., LaBounty, A., Levy, B., Cao, J., Michino, M., Luedtke, R. R., Javitch, J. A., and Shi, L. (2012) Molecular


29


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 35

Determinants of Selectivity and Efficacy at the Dopamine D3 Receptor, J. Med. Chem. 55, 6689–6699. (27) Lane, J. R., Chubukov, P., Liu, W., Canals, M., Cherezov, V., Abagyan, R., Stevens, R. C., and Katritch, V. (2013) Structure-based ligand discovery targeting orthosteric and allosteric pockets of dopamine receptors, Mol. Pharmacol. 84, 794–807. (28) Weichert, D., Banerjee, A., Hiller, C., Kling, R. C., Hübner, H., and Gmeiner, P. (2015) Molecular Determinants of Biased Agonism at the Dopamine D2 Receptor, J. Med. Chem. 58, 2703–2717. (29) Möller, D., Kling, R. C., Skultety, M., Leuner, K., Hübner, H., and Gmeiner, P. (2014) Functionally selective Dopamine D2, D3 receptor partial agonists, J. Med. Chem. 57, 4861– 4875. (30) Möller, D., Salama, I., Kling, R. C., Hübner, H., and Gmeiner, P. (2015) 1,4-Disubstituted aromatic piperazines with high 5-HT2A/D2 selectivity: Quantitative structure-selectivity investigations, docking, synthesis and biological evaluation, Bioorg. Med. Chem. 23, 6195–6209. (31) Chen, X., Sassano, M. F., Zheng, L., Setola, V., Chen, M., Bai, X., Frye, S. V., Wetsel, W. C., Roth, B. L., and Jin, J. (2012) Structure-functional selectivity relationship studies of betaarrestin-biased dopamine D(2) receptor agonists, J. Med. Chem. 55, 7141–7153. (32) Šali, A., and Blundell, T. L. (1993) Comparative Protein Modelling by Satisfaction of Spatial Restraints, J. Mol. Biol. 234, 779–815. (33) Chien, E. Y. T., Liu, W., Zhao, Q., Katritch, V., Won Han, G., Hanson, M. A., Shi, L., Newman, A. H., Javitch, J. A., Cherezov, V., and Stevens, R. C. (2010) Structure of the Human


30

Page 31 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Dopamine D3 Receptor in Complex with a D2/D3 Selective Antagonist, Science 330, 1091– 1095. (34) Mysinger, M. M., and Shoichet, B. K. (2010) Rapid Context-Dependent Ligand Desolvation in Molecular Docking, J. Chem. Inf. Model. 50, 1561–1573. (35) Mysinger, M. M., Carchia, M., Irwin, J. J., and Shoichet, B. K. (2012) Directory of Useful Decoys, Enhanced (DUD-E): Better Ligands and Decoys for Better Benchmarking, J. Med. Chem. 55, 6582–6594. (36) Jin, J., Roth, B., and Frye, S. (2012) Functionally Selective Ligands of Dopamine D2 Receptors, WO2012/003418 A2 (37) Männel, B., Dengler, D., Shonberg, J., Hübner, H., Möller, D., and Gmeiner, P. (2017) Hydroxy-Substituted Heteroarylpiperazines: Novel Scaffolds for β-Arrestin-Biased D2R Agonists, J. Med. Chem. 60, 4693–4713. (38) Klein Herenbrink, C., Sykes, D. A., Donthamsetti, P., Canals, M., Coudrat, T., Shonberg, J., Scammells, P. J., Capuano, B., Sexton, P. M., Charlton, S. J., Javitch, J. A., Christopoulos, A., and Lane, J. R. (2016) The role of kinetic context in apparent biased agonism at GPCRs, Nat. Commun. 7, 10842. (39) Irwin, J. J., Sterling, T., Mysinger, M. M., Bolstad, E. S., and Coleman, R. G. (2012) ZINC: A Free Tool to Discover Chemistry for Biology, J. Chem. Inf. Model. 52, 1757–1768. (40) Baell, J. B., and Holloway, G. A. (2010) New Substructure Filters for Removal of Pan Assay Interference Compounds (PAINS) from Screening Libraries and for Their Exclusion in Bioassays, J. Med. Chem. 53, 2719–2740.


31


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 35

(41) Winpenny, D., Clark, M., and Cawkill, D. (2016) Biased ligand quantification in drug discovery: from theory to high throughput screening to identify new biased µ opioid receptor agonists, Br. J. Pharmacol. 173, 1393–1403. (42) Urban, J. D., Clarke, W. P., von Zastrow, M., Nichols, D. E., Kobilka, B., Weinstein, H., Javitch, J. A., Roth, B. L., Christopoulos, A., Sexton, P. M., Miller, K. J., Spedding, M., and Mailman, R. B. (2007) Functional Selectivity and Classical Concepts of Quantitative Pharmacology, J. Pharmacol. Exp. Ther. 320, 1–13. (43) Neve, K. A. (2009) Functional Selectivity of G Protein-Coupled Receptor Ligands - New Opportunities for Drug Discovery, In The Receptors (Neve, K. A., Ed.) 1st ed., pp 25–40, Humana Press, Totowa, NJ. (44) Szabo, M., Klein Herenbrink, C., Christopoulos, A., Lane, J. R., and Capuano, B. (2014) Structure–activity relationships of privileged structures lead to the discovery of novel biased ligands at the dopamine D2 receptor, J. Med. Chem. 57, 4924–4939. (45) Shonberg, J., Herenbrink, C. K., Lopez, L., Christopoulos, A., Scammells, P. J., Capuano, B., and Lane, J. R. (2013) A structure-activity analysis of biased agonism at the dopamine D2 receptor, J. Med. Chem. 56, 9199–9221. (46) Hiller, C., Kling, R. C., Heinemann, F. W., Meyer, K., Hübner, H., and Gmeiner, P. (2013) Functionally selective dopamine D2/D3 receptor agonists comprising an enyne moiety, J. Med. Chem. 56, 5130–5141. (47) Free, R. B., Chun, L. S., Moritz, A. E., Miller, B. N., Doyle, T. B., Conroy, J. L., Padron, A., Meade, J. A., Xiao, J., Hu, X., Dulcey, A. E., Han, Y., Duan, L., Titus, S., Bryant-Genevier,


32

Page 33 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


M., Barnaeva, E., Ferrer, M., Javitch, J. A., Beuming, T., Shi, L., Southall, N. T., Marugan, J. J., and Sibley, D. R. (2014) Discovery and Characterization of a G Protein–Biased Agonist That Inhibits β-Arrestin Recruitment to the D2 Dopamine Receptor, Mol. Pharmacol. 86, 96–105. (48) Chen, X., McCorvy, J. D., Fischer, M. G., Butler, K. V., Shen, Y., Roth, B. L., and Jin, J. (2016) Discovery of G Protein-Biased D2 Dopamine Receptor Partial Agonists, J. Med. Chem. 59, 10601–10618. (49) Vass, M., Ágai-Csongor, É., Horti, F., and Keserű, G. M. (2014) Multiple Fragment Docking and Linking in Primary and Secondary Pockets of Dopamine Receptors, ACS Med. Chem. Lett. 5, 1010–1014. (50) Warne, T., Edwards, Patricia C., Leslie, Andrew G. W., and Tate, Christopher G. (2012) Crystal Structures of a Stabilized β1-Adrenoceptor Bound to the Biased Agonists Bucindolol and Carvedilol, Structure 20, 841–849. (51) Rasmussen, S. G. F., Choi, H.-J., Fung, J. J., Pardon, E., Casarosa, P., Chae, P. S., DeVree, B. T., Rosenbaum, D. M., Thian, F. S., Kobilka, T. S., Schnapp, A., Konetzki, I., Sunahara, R. K., Gellman, S. H., Pautsch, A., Steyaert, J., Weis, W. I., and Kobilka, B. K. (2011) Structure of a nanobody-stabilized active state of the β2 adrenoceptor, Nature 469, 175–180. (52) Carlsson, J., Coleman, R. G., Setola, V., Irwin, J. J., Fan, H., Schlessinger, A., Sali, A., Roth, B. L., and Shoichet, B. K. (2011) Ligand discovery from a dopamine D3 receptor homology model and crystal structure, Nat. Chem. Biol. 7, 769-778. (53) Rodríguez, D., Brea, J., Loza, María I., and Carlsson, J. (2014) Structure-Based Discovery of Selective Serotonin 5-HT1B Receptor Ligands, Structure 22, 1140–1151.


33


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 35

(54) Rodríguez, D., Gao, Z.-G., Moss, S. M., Jacobson, K. A., and Carlsson, J. (2015) Molecular Docking Screening Using Agonist-Bound GPCR Structures: Probing the A2A Adenosine Receptor, J. Chem. Inf. Model. 55, 550–563.

TABLE OF CONTENTS GRAPHIC


34

Page 35 of 35

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC


Structure-Guided Screening for Functionally Selective D2 Dopamine

Recommend Documents