Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Receptor, Ligand and Transducer Contributions to Dopamine D2 Receptor Functional Selectivity

  • Sean M. Peterson,

    Affiliation Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, United States of America

    ⨯
  • Thomas F. Pack,

    Affiliation Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, United States of America

    ⨯
  • Marc G. Caron

    marc.caron@dm.duke.edu

    Affiliations Department of Cell Biology, Duke University Medical Center, Durham, North Carolina, United States of America, Department of Neurobiology, Duke University Medical Center, Durham, North Carolina, United States of America, Department of Medicine, Duke University Medical Center, Durham, North Carolina, United States of America

    ⨯

Abstract

Functional selectivity (or biased agonism) is a property exhibited by some G protein-coupled receptor (GPCR) ligands, which results in the modulation of a subset of a receptor’s signaling capabilities and more precise control over complex biological processes. The dopamine D2 receptor (D2R) exhibits pleiotropic responses to the biogenic amine dopamine (DA) to mediate complex central nervous system functions through activation of G proteins and β-arrestins. D2R is a prominent therapeutic target for psychological and neurological disorders in which DA biology is dysregulated and targeting D2R with functionally selective drugs could provide a means by which pharmacotherapies could be developed. However, factors that determine GPCR functional selectivity in vivo may be multiple with receptors, ligands and transducers contributing to the process. We have recently described a mutagenesis approach to engineer biased D2R mutants in which G protein-dependent ([Gprot]D2R) and β-arrestin-dependent signaling ([βarr]D2R) were successfully separated (Peterson, et al. PNAS, 2015). Here, permutations of these mutants were used to identify critical determinants of the D2R signaling complex that impart signaling bias in response to the natural or synthetic ligands. Critical residues identified in generating [Gprot]D2R and [βarr]D2R conferred control of partial agonism at G protein and/or β-arrestin activity. Another set of mutations that result in G protein bias was identified that demonstrated that full agonists can impart unique activation patterns, and provided further credence to the concept of ligand texture. Finally, the contributions and interplay between different transducers indicated that G proteins are not aberrantly activated, and that receptor kinase and β-arrestin activities are inextricably linked. These data provide a thorough elucidation of the feasibility and malleability of D2R functional selectivity and point to means by which novel in vivo therapies could be modeled.

Citation: Peterson SM, Pack TF, Caron MG (2015) Receptor, Ligand and Transducer Contributions to Dopamine D2 Receptor Functional Selectivity. PLoS ONE 10(10): e0141637. https://doi.org/10.1371/journal.pone.0141637

Editor: Arun Shukla, Indian Institute of Technology Kanpur, INDIA

Received: June 9, 2015; Accepted: October 12, 2015; Published: October 30, 2015

Copyright: © 2015 Peterson et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: The authors wish to thank Drs. Bernard Masri, Nikhil Urs, Angela Dawn Wilkins, Olivier Lichtarge, and Lawrence Barak for helpful discussions during the course of this work. This work was supported in part by grants from the National Institutes of Health grant 5R37MH073853 to MGC. TFP is supported by an award from the Ruth K. Broad Biomedical Research Foundation. The continued support of the Pall Family Foundation is greatly appreciated. The expert technical assistance of Yushi Bai and Caroline Ray is also greatly appreciated.

Competing interests: MGC has received compensation from Lundbeck as a member of their Psychopharmacology Advisory Board and is a consultant for Omeros Corp. MGC also owns equity in Acadia Pharmaceuticals. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials. The other authors declare no potential conflict of interest.

Introduction

G protein-coupled receptors (GPCRs) are dynamic conduits of extracellular messages into complex intracellular instructions. These instructions are carried out through activation of both G protein dependent and independent signaling pathways [1]. GPCRs exhibit functional selectivity [2] in responses to natural or synthetic ligands by signaling through a subset of their normal multiple pathways. This capacity for functional selectivity has been theorized to arise from receptor conformational heterogeneity. Additionally, ligands that exhibit functional selectivity are termed biased agonists [3]. The concept of functional selectivity at GPCRs could provide a means not only to understand how GPCRs mediate their actions but also for developing more selective and efficacious therapeutic agents. Based on this concept an increasing number of functionally selective or biased ligands have been developed for a several GPCRs [4–12]. Most GPCRs, with the exception of a few [13,14], mediate the action of a single natural ligand. How in vivo pleiotropic signaling is determined following engagement of the receptor by its cognate ligand is likely to be controlled by the complement of cellular accessory proteins, such as G proteins, GPCR kinases (GRKs) and β-arrestins.

One of the major transducers of G protein independent signaling are β-arrestins, multifunctional adaptor proteins and part of the desensitization machinery that scaffold GPCRs for internalization and recycling of competent receptors to the plasma membrane [15]. β-arrestins also scaffold signaling complexes that have been shown to alter metabolic pathways [16], transcription [17], and neuronal activity leading to behavior [18]. For those GPCRs for which the consequences of functionally selective signaling have been examined, the G protein and β-arrestin signaling pathways typically subserve different cellular functions [19–21] but with some notable exceptions, as observed with the AT1A receptor in its regulation of aldosterone production [22]. The molecular details of selectivity and major signal transduction elements of β-arrestin or G protein signaling are new avenues by which GPCR pharmacology can be exploited for the development of novel pharmaceutical therapies [23].

The dopamine D2 receptor (D2R) is a prominently expressed GPCR for the biogenic amine dopamine (DA). DA is critical in many central nervous system functions and D2R is the target of many pharmaceutical interventions in which DA homeostasis is disrupted. D2R responds to DA with activation of the inhibitory family of GαO/i subunits which leads to an inhibition of cAMP production and liberation of Gβγ which leads to MAP kinase activation, as well as increased cell membrane potassium conductance through GIRK channels, among other effects [24]. Additionally, genetic and biochemical approaches have implicated β-arrestin 2 as a significant contributor to D2R signal transduction [18].

Functional selectivity arises from receptor conformational heterogeneity, which is the receptor’s capacity to adopt multiple related conformations that activate signaling molecules [25,26]. GPCRs undergo two major conformational processes during activation: 1) G protein stimulation and 2) β-arrestin recruitment. Ligands stabilize the transition state for guanine-nucleotide exchange factor (GEF) activity of GPCRs at G proteins, while GPCR kinases (GRKs) are efficiently recruited to agonist bound GPCRs where they phosphorylate intracellular domains, most frequently the C-terminal tail of GPCRs. Phosphorylation alters the agonist bound receptor conformational ensemble to favor β-arrestin recruitment, and this presumably initiates G protein-independent signaling.

Understanding how GPCRs propagate pleiotropic signals to generate functionally selective responses depends on the question at hand. In developing a selective ligand, recognition of receptor conformational states by a ligand may guide the experimental approaches. However, if an altered signaling mechanism mediating a specific cellular effect is desired for therapeutic benefit, distinct determinants of the selectivity process may be invoked. An important caveat in the design of biased agonists is the relative expression levels of transducer/interacting molecules, which can determine the bias because of altered coupling probability [27]. D2R’s prominence as a pharmaceutical target for many disorders makes it a good receptor candidate for precise and robust dissection of functional selectivity. We have recently reported the development and characterization of mutant D2R that are selective for G protein activation or β-arrestin recruitment [28]. These mutants, termed [Gprot]D2R and [βarr]D2R, respectively, show an unprecedented separation of function and have retained essentially the unabridged major functions of [WT]D2R.

Here, the contributions of ligand, receptor and transducer to functional selectivity are systematically assessed using several variants of the previously described mutants. These novel mutants are characterized and assessed for their unique functional selectivity properties that provide insight into the quality and determinants of D2R functional selectivity. The data demonstrate that functional selectivity is dynamic and malleable. In addition to direct manipulation of the receptor, the role of transducer levels altered the signaling profile of D2R, which suggests that the in vivo complement and expression level of transducers play a significant role in shaping the functional selectivity of D2R ligands.

Materials and Methods

Mutagenesis PCR

The Agilent Technologies (Santa Clara, CA) QuikChange mutagenesis kit was used to carry out all mutagenesis according to manufacturer’s instructions. Primers were designed as instructed, with the minimum amount of nucleotide changes required to achieve a mutation. All work was carried out on the mouse long isoform of D2R. Multiple point mutations were created by using the same primers for single point mutations on already mutated constructs. All constructs were confirmed to have no coding errors by sequencing. In addition, a previously characterized D2R mutant, termed [D80A]D2R which was previously shown to ablate sodium coordination which causes a deficit in G protein [29] and β-arrestin recruitment [28] but not ligand binding or plasma membrane trafficking was used as a negative control.

Cell culture and transfections

HEK-293T (ATCC, Manassas, VA) cells were cultured and transfected as previously reported [30].

G protein activity

D2R’s ability to inhibit cAMP production was carried out as previously described [8] using the Promega (Madison, WI) GloSensor assay with minor modifications. D2R was expressed at a mass of 1 μg of DNA (except where indicated) and the GloSensor construct was transiently transfected along with D2R at a mass of 5 μg of DNA. The luminescence was quantified with the Mithras LB940 instrument with no wavelength filter between the cells and the photomultiplier.

Bioluminescent Resonance Energy Transfer

BRET was performed as previously described [30] with some minor modifications. GRK2-YFP or β-arrestin 1-YFP replaced β-arrestin 2-YFP, where indicated. Untagged GRK2 was overexpressed at a ratio of 2-fold higher than receptor, while β-arrestin 2-YFP was always kept at the maximum allowable expression. RLuc-tagged D2R constructs were not different from untagged receptors in ligand binding and G protein coupling as previously determined for [WT]D2R [30] and confirmed for each mutant. Receptor expression levels were determined for each experiment in order to ensure comparable levels of expression.

Alternative G protein signaling

Various D2R constructs expressed in HEK 293 cells were used to determine whether D2R could mediate Gαs activation in the GloSensor assay described above. cAMP production following stimulation of endogenous β2ARs was used as a control. Whether the various D2R constructs could couple to Gαq was measured using the aequorin assay, as previously described [31]. As a control, HEK 293 cells were transfected with the angiotensin AT1A receptor and calcium was measured in response to agonist activation.

Radioligand Binding

[3H]-raclopride (Promega, Waltman, MA) binding was carried out as previously described [32]. When sodium was removed, the salt was not replaced with any other ion in the buffer. Rluc counts were conducted on the same membrane preparations, the same day that the ligand binding was carried out, using the same RLuc counting protocol as [30].

Data Analysis

Dose response curves were fit to the nonlinear regression curve y = Bottom + (Top-Bottom)/(1+10^((LogEC50-X))) for agonist curves and y = Bottom + (Top-Bottom)/(1+10^((X-LogIC50))) for antagonists in Graphpad Prism 5. Statistical tests were performed (described in Table legends) in Graphpad Prism 5. Bias quantification was carried out as previously described in [33] and [34]. For bias plots (Fig 1D and S1C and S1D Fig) the points were calculated from normalized (to [WT]D2R) responses to the two assays and fit to a quadratic equation with the constraint that B0 = 0. Each bias quantification used the same G protein activity when compared to endogenous and GRK2 overexpression data sets at β-arrestin. All values calculated in S1–S4 Tables were normalized to [WT]D2R (or control receptors in S4 Table) for each individual assay for efficacy but not potency.

thumbnail
Download:
Fig 1. Context dependent functional selectivity.

(A) β-arrestin 2 recruitment comparing [WT]D2R and [IYIV]D2R as determined by bioluminescent resonance energy transfer (BRET). (B) GRK2 overexpression enhances β-arrestin 2 recruitment by BRET for [IYIV]D2R and [WT]D2R, but only slightly for [Gprot]D2R, [βarr]D2R, and [D80A]D2R when compared to [28]. All data are presented with SEM from n = 3–4 independent experiments, with statistical significance calculated in S1 Table. Quantification of bias between G protein activity (data presented in S1 Fig and[28]) and β-arrestin 2 recruitment (data presented in Fig 1A and [28]) using (C) a statistical formalism where KA, calculated from EC50 = 1 [33] or (D) bias plot mapping under normal (solid lines) and GRK2 overexpression enhanced (broken lines) conditions.

https://doi.org/10.1371/journal.pone.0141637.g001

Results

A rich landscape of receptor dictated functional selectivity

The functional selectivity of a receptor can be simply viewed as its propensity to engage one signaling mechanism over another. The ability of the GPCR kinase 2 (GRK2) to affect the interaction of various D2R mutants with β-arrestin 2 was compared to a previously characterized biased D2R that was engineered [35,36] by mutating a motif unique to D2-like receptors (IYIV) to four alanines ([IYIV]D2R). As seen in Fig 1A and S1A Fig, when expressed in HEK293 cells [IYIV]D2R displays a decrease in G protein signaling activity, along with a nearly complete absence of β-arrestin 2 recruitment, as previously observed [36]. Thus, [IYIV]D2R is a G protein-preferring mutant receptor under these conditions. However, when GRK2 is overexpressed (Fig 1B), β-arrestin 2 recruitment potency is enhanced at [WT]D2R and [IYIV]D2R, and only slightly potentiated for the previously characterized [Gprot]D2R and [βarr]D2R when compared to HEK293 cells expressing endogenous levels of GRK2 [28]. Quantifying the bias between [Gprot]D2R and [IYIV]D2R using a statistical formalism [33] (Fig 1C and S1B Fig), a bias plot [34] (Fig 1D and S1C and S1D Fig) and ΔΔlog(τ/KA) calculations (S2 Table) reveal the quality of G protein bias. These different bias quantifications allow for comparisons of efficacy (calculated from EMAX to τ) and potency (calculated from EC50 to KA) in different ways, which facilitates conclusions based on their relationship. For instance, [Gprot]D2R and [IYIV]D2R display similar degrees of bias using each method because while G protein activity is slightly perturbed at [IYIV]D2R, at [Gprot]D2R the G protein activity is completely intact, while a small amount of β-arrestin activity remains. However, when GRK2 is overexpressed [Gprot]D2R does not gain appreciable β-arrestin efficacy, whereas [IYIV]D2R gains significantly in efficacy, this is revealed by the greater shift observed using the bias statistical formalism (Fig 1C). In contrast, the bias plot (Fig 1C) reveals that [IYIV]D2R is biased toward G protein activity but the mutant never reaches 100% activity, and ΔΔlog(τ/KA) values reveal that [IVIV]D2R bias is dependent on GRK2 levels (S2 Table). Since the GRK2 overexpression assay demonstrates the receptor’s capacity for β-arrestin recruitment under the most favorable conditions, it shows that [IYIV]D2R can display partial agonism at both G protein and β-arrestin activities. In contrast, [Gprot]D2R retains its original biased signaling profile: full agonism at the G protein pathway and weak partial agonism at the β-arrestin pathway even when GRK2 is highly expressed.

The controlled perturbation of G protein and β-arrestin pathways allows for a more detailed examination of partial agonism and how it can be utilized in generating receptor bias. The conserved residue A3.53 (A135) was previously mutated into all 19 possible amino acids [28] because it conferred remarkable functional selectivity properties. In fact, when the proximity of A135 to the G protein in the receptor/G protein complex (Fig 2A) is compared to its proximity to arrestin in the receptor/arrestin complex (Fig 2B) it is clear that the G protein more closely associates with A135. When A135 is mutated to basic residues, D2R lost ~50% of its G protein activity, while still retaining complete, and slightly more potent β-arrestin activity (S2A and S2B Fig, respectively). In contrast, acidic substitutions ablated activity at both pathways. Furthermore, substitution with a bulky polar residue (tyrosine) yielded a balanced reduction in both G protein and β-arrestin 2 activity to roughly 75% (S2C and S2D Fig) and substitution with a bulky nonpolar residue (phenylalanine) yielded a balanced 50% reduction (S2C and S2D Fig). These mutants were combined with one residue substitution from [Gprot]D2R (L125N) or [βarr]D2R (M140D) to generate mutants that maintain the loss of signaling associated with either [Gprot]D2R or [βarr]D2R while also exhibiting partial agonism at the retained pathway. In other words, these mutants selectively lose signaling at on pathway nearly completely, while allowing precise control of the degree of efficacy (partial agonism) still present at the other pathway.

thumbnail
Download:
Fig 2. Receptor control of partial agonism at D2R with A135 mutations.

(A) Relative proximity of G proteins (green spheres) and A135 (red sphere) in D3 (blue ribbon, PDB ID: 3PBL [37]) as determined by alignment of D3R to β2AR in receptor/G protein complex (PDB ID: 3SN6, [38]). (B) Arrestin (yellow spheres) does not reside close to A135 when D3R is aligned to rhodopsin in receptor/arrestin complex (PDB ID: 4ZWJ [39]). (C) G protein activity as determined by inhibition of isoproterenol-induced cAMP accumulation is titrated by substitution of A135 with a bulky polar (tyrosine) or nonpolar (phenylalanine) residue and combined with L125N or M140D to impart controlled loss of G protein function. (C) β-arrestin 2 recruitment as determined by BRET is similarly controlled. All data are presented with SEM from n = 3–5 independent experiments, with statistical significance calculated in S1 Table.

https://doi.org/10.1371/journal.pone.0141637.g002

Agonist texture reveals novel modes of functional selectivity

A unique G protein-biased mutant (termed [Gprot4PM]D2R) displayed remarkable retention of G protein activity and loss of β-arrestin 2 recruitment when the reference agonist quinpirole was used to probe activity (Fig 3A and 3B, respectively). This mutation was generated by the same iterative Evolutionary Trace process that produced [Gprot]D2R and [βarr]D2R [28]. However, when the endogenous ligand DA was used, the G protein activity was unchanged, but the β-arrestin activity was ~50% of [WT]D2R (Fig 3A and 3B, S3 Table). At [Gprot4PM]D2R both DA and quinpirole were able to recruit β-arrestin 2 to almost the same extent of [WT]D2R when GRK2 was overexpressed (Fig 3C, S3 Table). However, the recruitment of GRK2 by [Gprot4PM]D2R showed the same agonist selectivity between DA and quinpirole (Fig 3D) as observed with β-arrestin 2. These data demonstrate the concept of agonist texture [40], which predicts that full agonists induce distinct receptor activation states and conformations.

thumbnail
Download:
Fig 3. A unique G protein biased mutant demonstrates agonist texture.

(A) Dopamine (DA) and quinpirole equivalently inhibit cAMP production, which is equivalent to [WT]D2R for [Gprot4PM]D2R (T69F Y133L Y209N A372S). (B) [Gprot4PM]D2R has roughly 50% efficacy in response to DA but not quinpirole for β-arrestin 2 recruitment. (C) GRK2 overexpression rescues both DA and quinpirole β-arrestin 2 recruitment activity nearly to [WT]D2R levels (dotted line, from Fig 1B). (D) GRK2 recruitment as determined by BRET (where GRK2 is tagged with YFP) shows the same ligand discrepancy as β-arrestin 2. All data are presented with SEM from n = 3 independent experiments, with statistical significance calculated in S3 Table.

https://doi.org/10.1371/journal.pone.0141637.g003

While [Gprot]D2R and [βarr]D2R display unprecedented separation of signal in response to DA [28], various additional D2R agonists and antagonists were used to probe the extent of agonist texture between the mutants at G protein and β-arrestin 2 activation. Each agonist tested at cAMP inhibition was effectively equivalent for [WT]D2R and [Gprot]D2R while being severely disrupted for [βarr]D2R and [D80A]D2R (Fig 4A, 4C and 4E, S3 Table). In contrast, [βarr]D2R was nearly as effective at β-arrestin 2 recruitment as [WT]D2R whereas [Gprot]D2R and [D80A]D2R were deficient (Fig 4B, 4D and 4F, S3 Table). Similarly, the well characterized antagonist raclopride, typical antipsychotic haloperidol, and atypical antipsychotic aripiprazole inhibited DA-induced cAMP reduction for [Gprot]D2R and [WT]D2R (Fig 4G, 4I and 4K, S3 Table), and β-arrestin 2 recruitment for [βarr]D2R and [WT]D2R (Fig 4H, 4J and 4L, S3 Table). These diverse D2R ligands behave as expected for each assay and provide evidence that the complex activation states of [Gprot]D2R and [βarr]D2R remain intact, as opposed to [Gprot4PM]D2R, which has lost responsiveness to quinpirole at β-arrestin 2 recruitment selectively.

thumbnail
Download:
Fig 4. Agonists and antagonists with diverse pharmacophores elicit predictable responses at [Gprot]D2R and [βarr]D2R.

The D2R agonists quinpirole, apomorphine, and N-propylapomorphine (NPA) were tested for G protein activity (A,C,E) and β-arrestin 2 recruitment (B,D,F). For each agonist, [Gprot]D2R showed a response similar to [WT]D2R at G protein activation and more similar to [D80A]D2R for β-arrestin recruitment, while [βarr]D2R was not active at the G protein pathway but retained activity at the β-arrestin pathway. The antagonists raclopride (G,H) haloperidol (I,J) and partial antagonist aripiprazole (K,L) were able to block DA elicited D2R activation at the G protein pathway (G,I,K) for [Gprot]D2R and [WT]D2R to the same extent, while [D80A]D2R and [βarr]D2R had no effect to inhibit. In contrast, these antagonists block DA elicited β-arrestin 2 recruitment (H,J,L) for [βarr]D2R and [WT]D2R. All data are presented with SEM from n = 3–4 independent experiments, with statistical significance calculated in S3 Table.

https://doi.org/10.1371/journal.pone.0141637.g004

The status of receptor interacting partners in extremely biased mutant D2Rs

Allosteric modulators were assessed at [Gprot]D2R and [βarr]D2R while being compared to positive and negative controls ([WT]D2R and [D80A]D2R, respectively). The interaction of each receptor with other components of the desensitization machinery mirrored the recruitment of β-arrestin 2. GRK2 (Fig 5A) and β-arrestin 1 (Fig 5B) showed a similar slight potentiation and loss of efficacy at [βarr]D2R when compared to [WT]D2R, while [Gprot]D2R was severely deficient (S4 Table). To test whether the mutagenesis mediated loss of function at G protein and β-arrestin interactions achieved with [Gprot]D2R and [βarr]D2R could potentially have induced aberrant activation of normally inactive receptor interacting proteins, two other non-conventional D2R signaling avenues (Gαs and Gαq) were assessed and neither were activated by any of the D2R mutants (Fig 5C and 5D, respectively, S4 Table).

thumbnail
Download:
Fig 5. Interacting partners and allosteric D2R determinants of functional selectivity.

(A) GRK2 and (B) β-arrestin 1 recruitment as assessed by BRET show a similar profile as β-arrestin 2: [βarr]D2R recruits normally, while [Gprot]D2R is severely deficient. (C) Each D2R construct was expressed in HEK 293T cells and assessed for its ability to stimulate cAMP in response to DA. Stimulation of endogenous receptor by isoproterenol was used as a control response. (D) Gαq mediated Ca2+ flux, as measured by the aequorin luminescence assay, is not stimulated by [WT]D2R, [Gprot]D2R, [βarr]D2R or [D80A]D2R, compared to AngII induced Ca2+ flux induced by transient expression of AT1AR. (E) BMAX was determined by binding, while luciferase-tagged receptors provided a BMAX-independent measure of receptor number. In this assay, the responsiveness to sodium is retained for all mutants (except [D80A]D2R). All data are presented with SEM from n = 3–4 independent experiments, with statistical significance calculated in S4 Table.

https://doi.org/10.1371/journal.pone.0141637.g005

The effect of the GPCR allosteric modulator sodium has been previously shown to enhance dopamine binding to D2R [29] and has been proposed to function as an efficacy switch for receptor bias [41]. As previously demonstrated [29], when sodium is removed from D2R binding buffer this results in a 50% reduction in BMAX, similar to the phenomenon observed in the A2AR [42]. To assess sodium dynamics in D2R, radioligand binding was used to determine BMAX with and without sodium on the Renilla luciferase tagged D2R constructs [30]. In this way, luminescence counts yield total receptors available for binding (relative to [WT]D2R) and BMAX yields total binding sites, modulated by sodium. This assay yielded a 50% reduction in [WT]D2R and [βarr]D2R apparent BMAX (Fig 5E) when binding was performed without sodium and [Gprot]D2R also showed a sodium dependent reduction, although [Gprot]D2R expresses lower than [WT]D2R as previously described [28]. [D80A]D2R BMAX did not change regardless of the presence of sodium, which validates the experiment as [D80A]D2R is mutated at the presumed site of sodium interaction and has been previously shown to not bind sodium [29].

Discussion

The functional selectivity of D2R is dynamic and malleable. Ligands that target D2R have significant therapeutic impact [24] and the quality of agonism or antagonism of these ligands can be operationally defined in monitoring systems [8,30]. The agonist or antagonist quality of a ligand/receptor pair is dependent upon the assay used for detection [43]. Here, the capacity, quality and character of D2R functional selectivity has been examined for different mutated D2Rs.

GPCR bias is operationally defined by the degree of engagement of a given signaling pathway versus another. It is clear that [IYIV]D2R and [Gprot]D2R both share G protein bias, however the quality of bias differs, depending on their operational definition. For example, in some cases it may be beneficial to retain all of the G protein activity, while in others it may be crucial to abolish all β-arrestin recruitment at the sacrifice of some G protein activity. In fact, [IYIV]D2R was mutated to contain fewer alanine substitutions which resulted in full G protein activity and partial β-arrestin activity [36]. Nevertheless, both [IYIV]D2R and [Gprot]D2R were used to dissect D2R-mediated ERK phosphorylation [28,35] and both mutants yielded a similar conclusion: that D2R-mediated ERK phosphorylation is largely G protein mediated.

Operational consistency allows for meaningful conclusions about D2R signaling pathways to be drawn. The partial agonist activity of aripiprazole has raised the possibility that complete blockade of D2R is not necessary to impart antipsychotic efficacy [44]. This partial agonism allows for effective targeting of G protein or β-arrestin pathways [8]. However, the causal relationship of partial agonism and biased partial agonism (as opposed to partial antagonism) has not been explored. The mutants described here that are derivations of A135-mutated D2R are tools that would allow for operationally defined agonism.

While receptor manipulation is desirable to demonstrate causal relationships, biased ligands provide valuable insight and are a more reasonable avenue toward therapy development. However, precise details and principles governing ligand action remain elusive. Here, the phenomenon of agonist texture [40] is demonstrated with [Gprot4PM]D2R. As previously observed, full agonists can stabilize different receptor activation states that have functional consequences [40]. Therefore, functional selectivity could occur from the loss of function at one pathway or it could be thought of as a gain of new receptor activity. This phenomenon represents a valuable conceptual framework for a fundamental property of receptor activation that is relevant to functional selectivity.

Intracellular signal transduction proteins are key elements in dictating bias. Their interactions with receptors dictate agonist efficacy [45] and targeting their activation with biased agonists is an avenue by which already validated receptor targets can be leveraged for improved therapies. Here, the related desensitization allosteric modulators (β-arrestin 1, β-arrestin 2, and GRK2, Fig 5A and 5B) were shown to fall into a common activation family using [βarr]D2R. These findings could have an impact on future studies of more detailed elements of D2R’s β-arrestin signaling arm, such as barcoded phosphorylation patterns [46], GRK subfamily contributions [47], pleiotropic β-arrestin conformation states [48], and other posttranslational modifications [49]. Additionally, related allosteric modulators (different G proteins, Fig 5C and 5D) remained inactive at each mutant receptor, which indicates that none of the mutants have a gross abnormal gain of function.

Interactions with small molecule allosteric modulators are also exciting avenues by which functional selectivity can be modulated. Sodium, an intracellular GPCR allosteric modulator, binds both [Gprot]D2R and [βarr]D2R, indicating that both G protein and β-arrestin activation require dynamic sodium regulation. However, allosteric biased ligands may confer functional selectivity by exploiting the recently solved extracellular vestibule [50] to generate noncompetitive negative or positive allosteric modulators [51] or bitopic ligands [52].

Novel mutants and rigorous examinations of strongly biased mutants provided a conceptual framework for the feasibility of such exercises. Furthermore, these studies provide robust and versatile tools for further investigations of partial agonism or agonist texture in complex, physiologically relevant systems. The actions of dopamine, and D2R, are dysregulated in many neurological and psychiatric disorders, yet the complete and precise molecular actions of D2R remain elusive. The work presented here highlights new tools available for the molecular dissection of D2R and provides valuable insight into methods that can be used to increase the utility of targeting D2R pharmacologically for improved therapeutics.

Supporting Information

S1 Fig. Comparison of receptor bias.

(A) cAMP partial agonism at [IYIV]D2R recapitulates previously published values [36]. Data are presented with SEM from n = 3 independent experiments. (B) Comparison of each biased mutant quantified using a statistical formalism [33] with endogenous GRK levels (solid lines) compared to GRK2 overexpression (broken lines). (C) and (D) bias plots to compare each receptor with and without GRK2 overexpressed, respectively. The data presented in B,C, and D is the full data set of mutants, while Fig 1C and 1D show only the G protein-biased mutants of these data.

https://doi.org/10.1371/journal.pone.0141637.s001

(TIF)

S2 Fig. Molecular determinants of signal efficacy at A135.

(A) G protein activity, as assessed by cAMP inhibition and (B) β-arrestin 2 recruitment, as assessed by BRET are compared to [WT]D2R efficacy for the G protein pathway (dotted line, A) and potency for the β-arrestin 2 recruitment (dotted line, B) respectively. Basic residue substitutions (blue) strongly bias D2R toward β-arrestin with an increase in potency, while acidic residues ablate signaling at both pathways. All data are presented with SEM from n = 3 independent experiments.

https://doi.org/10.1371/journal.pone.0141637.s002

(TIF)

S1 Table. Receptor controlled perturbation of functional selectivity.

Values derived from Figs 1 and 2 to demonstrate the receptor’s contributions to functional selectivity. *p<0.05 when compared to [WT]D2R for efficacy and potency as determined by Bonferroni post-hoc test after p<0.05 for one-way ANOVA.

https://doi.org/10.1371/journal.pone.0141637.s003

(DOCX)

S2 Table. Quantifying bias at G protein preferring mutant D2R.

BINF and ΔΔlog(τ/KA) were calculated according to references in the table. Some control data (DA at cAMP inhibition and β-arrestin 2 recruitment for [WT]D2R) calculated from [28].

https://doi.org/10.1371/journal.pone.0141637.s004

(DOCX)

S3 Table. Ligand contributions to functional selectivity.

Calculated from Figs 3 and 4. *p<0.05 when compared to [WT]D2R for efficacy and potency at each ligand as determined by Bonferroni post-hoc test after p<0.05 by one-way ANOVA.

https://doi.org/10.1371/journal.pone.0141637.s005

(DOCX)

S4 Table. Transducer contributions to functional selectivity.

Calculated from Fig 5. *p<0.05 when compared to [WT]D2R or control receptors (β2AR for Gαs or AT1AR for Gαq) for efficacy and potency as determined by Bonferroni post-hoc test after p<0.05 for one-way ANOVA.

https://doi.org/10.1371/journal.pone.0141637.s006

(DOCX)

Acknowledgments

The authors wish to thank Drs. Bernard Masri, Nikhil Urs, Angela Dawn Wilkins, Olivier Lichtarge, and Lawrence Barak for helpful discussions during the course of this work. This work was supported in part by grants from the National Institutes of Health grant 5R37MH073853 to MGC. TFP is supported by an award from the Ruth K. Broad Biomedical Research Foundation. The continued support of the Pall Family Foundation is greatly appreciated. The expert technical assistance of Yushi Bai and Caroline Ray is also greatly appreciated.

Author Contributions

Conceived and designed the experiments: SMP TFP MGC. Performed the experiments: SMP TFP. Analyzed the data: SMP MGC. Wrote the paper: SMP TFP MGC.

References

  1. 1. Lefkowitz RJ, Shenoy SK (2005) Transduction of receptor signals by beta-arrestins. Science 308: 512–517. pmid:15845844
  2. 2. Urban JD, Clarke WP, von Zastrow M, Nichols DE, Kobilka B, et al. (2007) Functional selectivity and classical concepts of quantitative pharmacology. J Pharmacol Exp Ther 320: 1–13. pmid:16803859
  3. 3. Kenakin T (2013) New concepts in pharmacological efficacy at 7TM receptors: IUPHAR review 2. Br J Pharmacol 168: 554–575. pmid:22994528
  4. 4. Conroy JL, Free RB, Sibley DR (2015) Identification of G Protein-Biased Agonists That Fail To Recruit beta-Arrestin or Promote Internalization of the D1 Dopamine Receptor. ACS Chem Neurosci 6: 681–692. pmid:25660762
  5. 5. Free RB, Chun LS, Moritz AE, Miller BN, Doyle TB, et al. (2014) Discovery and characterization of a G protein-biased agonist that inhibits beta-arrestin recruitment to the D2 dopamine receptor. Mol Pharmacol 86: 96–105. pmid:24755247
  6. 6. Felker GM, Butler J, Collins SP, Cotter G, Davison BA, et al. (2015) Heart failure therapeutics on the basis of a biased ligand of the angiotensin-2 type 1 receptor. Rationale and design of the BLAST-AHF study (Biased Ligand of the Angiotensin Receptor Study in Acute Heart Failure). JACC Heart Fail 3: 193–201. pmid:25650371
  7. 7. Soergel DG, Subach RA, Burnham N, Lark MW, James IE, et al. (2014) Biased agonism of the mu-opioid receptor by TRV130 increases analgesia and reduces on-target adverse effects versus morphine: A randomized, double-blind, placebo-controlled, crossover study in healthy volunteers. Pain 155: 1829–1835. pmid:24954166
  8. 8. Allen JA, Yost JM, Setola V, Chen X, Sassano MF, et al. (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. pmid:22025698
  9. 9. Tschammer N, Bollinger S, Kenakin T, Gmeiner P (2011) Histidine 6.55 is a major determinant of ligand-biased signaling in dopamine D2L receptor. Mol Pharmacol 79: 575–585. pmid:21163968
  10. 10. Shonberg J, Herenbrink CK, Lopez L, Christopoulos A, Scammells PJ, et al. (2013) A structure-activity analysis of biased agonism at the dopamine D2 receptor. J Med Chem 56: 9199–9221. pmid:24138311
  11. 11. Moller D, Kling RC, Skultety M, Leuner K, Hubner H, et al. (2014) Functionally selective dopamine D(2), D(3) receptor partial agonists. J Med Chem 57: 4861–4875. pmid:24831693
  12. 12. Weichert D, Banerjee A, Hiller C, Kling RC, Hubner H, et al. (2015) Molecular determinants of biased agonism at the dopamine D(2) receptor. J Med Chem 58: 2703–2717. pmid:25734236
  13. 13. Thompson GL, Canals M, Poole DP (2014) Biological redundancy of endogenous GPCR ligands in the gut and the potential for endogenous functional selectivity. Front Pharmacol 5: 262. pmid:25506328
  14. 14. Scholten DJ, Canals M, Maussang D, Roumen L, Smit MJ, et al. (2012) Pharmacological modulation of chemokine receptor function. Br J Pharmacol 165: 1617–1643. pmid:21699506
  15. 15. Laporte SA, Oakley RH, Zhang J, Holt JA, Ferguson SS, et al. (1999) The beta2-adrenergic receptor/betaarrestin complex recruits the clathrin adaptor AP-2 during endocytosis. Proc Natl Acad Sci U S A 96: 3712–3717. pmid:10097102
  16. 16. Luttrell LM, Ferguson SS, Daaka Y, Miller WE, Maudsley S, et al. (1999) Beta-arrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science 283: 655–661. pmid:9924018
  17. 17. Maudsley S, Martin B, Gesty-Palmer D, Cheung H, Johnson C, et al. (2015) Delineation of a Conserved Arrestin-Biased Signaling Repertoire In Vivo. Mol Pharmacol. pmid:25637603
  18. 18. Beaulieu JM, Sotnikova TD, Marion S, Lefkowitz RJ, Gainetdinov RR, et al. (2005) An Akt/beta-arrestin 2/PP2A signaling complex mediates dopaminergic neurotransmission and behavior. Cell 122: 261–273. pmid:16051150
  19. 19. Walters RW, Shukla AK, Kovacs JJ, Violin JD, DeWire SM, et al. (2009) beta-Arrestin1 mediates nicotinic acid-induced flushing, but not its antilipolytic effect, in mice. J Clin Invest 119: 1312–1321. pmid:19349687
  20. 20. Gesty-Palmer D, Yuan L, Martin B, Wood WH 3rd, Lee MH, et al. (2013) beta-arrestin-selective G protein-coupled receptor agonists engender unique biological efficacy in vivo. Mol Endocrinol 27: 296–314. pmid:23315939
  21. 21. Bohn LM, Lefkowitz RJ, Gainetdinov RR, Peppel K, Caron MG, et al. (1999) Enhanced morphine analgesia in mice lacking beta-arrestin 2. Science 286: 2495–2498. pmid:10617462
  22. 22. Lymperopoulos A, Rengo G, Zincarelli C, Kim J, Soltys S, et al. (2009) An adrenal beta-arrestin 1-mediated signaling pathway underlies angiotensin II-induced aldosterone production in vitro and in vivo. Proc Natl Acad Sci U S A 106: 5825–5830. pmid:19289825
  23. 23. Whalen EJ, Rajagopal S, Lefkowitz RJ (2011) Therapeutic potential of beta-arrestin- and G protein-biased agonists. Trends Mol Med 17: 126–139. pmid:21183406
  24. 24. Beaulieu JM, Gainetdinov RR (2011) The physiology, signaling, and pharmacology of dopamine receptors. Pharmacol Rev 63: 182–217. pmid:21303898
  25. 25. Nygaard R, Zou Y, Dror RO, Mildorf TJ, Arlow DH, et al. (2013) The dynamic process of beta(2)-adrenergic receptor activation. Cell 152: 532–542. pmid:23374348
  26. 26. Manglik A, Kim TH, Masureel M, Altenbach C, Yang Z, et al. (2015) Structural Insights into the Dynamic Process of beta2-Adrenergic Receptor Signaling. Cell 161: 1101–1111. pmid:25981665
  27. 27. Zhang J, Ferguson SS, Barak LS, Bodduluri SR, Laporte SA, et al. (1998) Role for G protein-coupled receptor kinase in agonist-specific regulation of mu-opioid receptor responsiveness. Proc Natl Acad Sci U S A 95: 7157–7162. pmid:9618555
  28. 28. Peterson SM, Pack TF, Wilkins AD, Urs NM, Urban DJ, et al. (2015) Elucidation of G-protein and beta-arrestin functional selectivity at the dopamine D2 receptor. Proc Natl Acad Sci U S A 112: 7097–7102. pmid:25964346
  29. 29. Neve KA, Cox BA, Henningsen RA, Spanoyannis A, Neve RL (1991) Pivotal role for aspartate-80 in the regulation of dopamine D2 receptor affinity for drugs and inhibition of adenylyl cyclase. Mol Pharmacol 39: 733–739. pmid:1828858
  30. 30. Masri B, Salahpour A, Didriksen M, Ghisi V, Beaulieu JM, et al. (2008) Antagonism of dopamine D2 receptor/beta-arrestin 2 interaction is a common property of clinically effective antipsychotics. Proc Natl Acad Sci U S A 105: 13656–13661. pmid:18768802
  31. 31. Evron T, Peterson SM, Urs NM, Bai Y, Rochelle LK, et al. (2014) G Protein and beta-Arrestin Signaling Bias at the Ghrelin Receptor. J Biol Chem 289: 33442–33455. pmid:25261469
  32. 32. Urs NM, Snyder JC, Jacobsen JP, Peterson SM, Caron MG (2012) Deletion of GSK3beta in D2R-expressing neurons reveals distinct roles for beta-arrestin signaling in antipsychotic and lithium action. Proc Natl Acad Sci U S A 109: 20732–20737. pmid:23188793
  33. 33. Barak LS, Peterson S (2012) Modeling of bias for the analysis of receptor signaling in biochemical systems. Biochemistry 51: 1114–1125. pmid:22221218
  34. 34. Kenakin T, Christopoulos A (2013) Signalling bias in new drug discovery: detection, quantification and therapeutic impact. Nat Rev Drug Discov 12: 205–216. pmid:23411724
  35. 35. Lan H, Liu Y, Bell MI, Gurevich VV, Neve KA (2009) A dopamine D2 receptor mutant capable of G protein-mediated signaling but deficient in arrestin binding. Mol Pharmacol 75: 113–123. pmid:18809670
  36. 36. Clayton CC, Donthamsetti P, Lambert NA, Javitch JA, Neve KA (2014) Mutation of three residues in the third intracellular loop of the dopamine D2 receptor creates an internalization-defective receptor. J Biol Chem 289: 33663–33675. pmid:25336643
  37. 37. Chien EY, Liu W, Zhao Q, Katritch V, Han GW, et al. (2010) Structure of the human dopamine D3 receptor in complex with a D2/D3 selective antagonist. Science 330: 1091–1095. pmid:21097933
  38. 38. Rasmussen SG, DeVree BT, Zou Y, Kruse AC, Chung KY, et al. (2011) Crystal structure of the beta2 adrenergic receptor-Gs protein complex. Nature 477: 549–555. pmid:21772288
  39. 39. Kang Y, Zhou XE, Gao X, He Y, Liu W, et al. (2015) Crystal structure of rhodopsin bound to arrestin by femtosecond X-ray laser. Nature 523: 561–567. pmid:26200343
  40. 40. Stallaert W, Dorn JF, van der Westhuizen E, Audet M, Bouvier M (2012) Impedance responses reveal beta(2)-adrenergic receptor signaling pluridimensionality and allow classification of ligands with distinct signaling profiles. PLoS One 7: e29420. pmid:22242170
  41. 41. Katritch V, Fenalti G, Abola EE, Roth BL, Cherezov V, et al. (2014) Allosteric sodium in class A GPCR signaling. Trends Biochem Sci 39: 233–244. pmid:24767681
  42. 42. Massink A, Gutierrez-de-Teran H, Lenselink EB, Ortiz Zacarias NV, Xia L, et al. (2015) Sodium ion binding pocket mutations and adenosine A2A receptor function. Mol Pharmacol 87: 305–313. pmid:25473121
  43. 43. Kenakin TP (2009) Cellular assays as portals to seven-transmembrane receptor-based drug discovery. Nat Rev Drug Discov 8: 617–626. pmid:19609267
  44. 44. Strange PG (2008) Antipsychotic drug action: antagonism, inverse agonism or partial agonism. Trends Pharmacol Sci 29: 314–321. pmid:18471899
  45. 45. Kenakin TP (2012) Biased signalling and allosteric machines: new vistas and challenges for drug discovery. Br J Pharmacol 165: 1659–1669. pmid:22023017
  46. 46. Nobles KN, Xiao K, Ahn S, Shukla AK, Lam CM, et al. (2011) Distinct phosphorylation sites on the beta(2)-adrenergic receptor establish a barcode that encodes differential functions of beta-arrestin. Sci Signal 4: ra51. pmid:21868357
  47. 47. Li L, Homan KT, Vishnivetskiy SA, Manglik A, Tesmer JJ, et al. (2015) G Protein-coupled Receptor Kinases of the GRK4 Protein Subfamily Phosphorylate Inactive G Protein-coupled Receptors (GPCRs). J Biol Chem 290: 10775–10790. pmid:25770216
  48. 48. Shukla AK, Violin JD, Whalen EJ, Gesty-Palmer D, Shenoy SK, et al. (2008) Distinct conformational changes in beta-arrestin report biased agonism at seven-transmembrane receptors. Proc Natl Acad Sci U S A 105: 9988–9993. pmid:18621717
  49. 49. Kommaddi RP, Shenoy SK (2013) Arrestins and protein ubiquitination. Prog Mol Biol Transl Sci 118: 175–204. pmid:23764054
  50. 50. Kruse AC, Ring AM, Manglik A, Hu J, Hu K, et al. (2013) Activation and allosteric modulation of a muscarinic acetylcholine receptor. Nature 504: 101–106. pmid:24256733
  51. 51. Wootten D, Christopoulos A, Sexton PM (2013) Emerging paradigms in GPCR allostery: implications for drug discovery. Nat Rev Drug Discov 12: 630–644. pmid:23903222
  52. 52. Lane JR, Sexton PM, Christopoulos A (2013) Bridging the gap: bitopic ligands of G-protein-coupled receptors. Trends Pharmacol Sci 34: 59–66. pmid:23177916