analysis of functional selectivity through g protein-dependent and -independent signaling pathways...
TRANSCRIPT
R
A-
DP
a
ARRAA
K�c�RFB
oamd[nd4my5
T
h0
Brain Research Bulletin 107 (2014) 89–101
Contents lists available at ScienceDirect
Brain Research Bulletin
j ourna l h o mepa ge: www.elsev ier .com/ locate /bra inresbul l
esearch report
nalysis of functional selectivity through G protein-dependent andindependent signaling pathways at the adrenergic �2C receptor
alma Kurko ∗, Zoltán Kapui, József Nagy, Balázs Lendvai, Sándor Kolokharmacological and Drug Safety Research, Gedeon Richter Plc., Budapest, Hungary
r t i c l e i n f o
rticle history:eceived 16 June 2014eceived in revised form 15 July 2014ccepted 17 July 2014vailable online 29 July 2014
eywords:2C-ARAMP accumulation-Arrestin recruitmenteceptor internalizationunctional selectivityias factor
a b s t r a c t
Although G protein-coupled receptors (GPCRs) are traditionally categorized as Gs-, Gq-, or Gi/o-coupled,their signaling is regulated by multiple mechanisms. GPCRs can couple to several effector pathways,having the capacity to interact not only with more than one G protein subtype but also with alternativesignaling or effector proteins such as arrestins. Moreover, GPCR ligands can have different efficaciesfor activating these signaling pathways, a characteristic referred to as biased agonism or functionalselectivity.
In this work our aim was to detect differences in the ability of various agonists acting at the �2C typeof adrenergic receptors (�2C-ARs) to modulate cAMP accumulation, cytoplasmic Ca2+ release, �-arrestinrecruitment and receptor internalization. A detailed comparative pharmacological characterization ofG protein-dependent and -independent signaling pathways was carried out using adrenergic agonists(norepinephrine, phenylephrine, brimonidine, BHT-920, oxymetazoline, clonidine, moxonidine, guan-abenz) and antagonists (MK912, yohimbine). As initial analysis of agonist Emax and EC50 values suggestedpossible functional selectivity, ligand bias was quantified by applying the relative activity scale andwas compared to that of the endogenous agonist norepinephrine. Values significantly different from0 between pathways indicated an agonist that promoted different level of activation of diverse effec-tor pathways most likely due to the stabilization of a subtly different receptor conformation from thatinduced by norepinephrine.
Our results showed that a series of agonists acting at the �2C-AR displayed different degree of functionalselectivity (bias factors ranging from 1.6 to 36.7) through four signaling pathways. As signaling via thesepathways seems to have distinct functional and physiological outcomes, studying all these stages ofreceptor activation could have further implications for the development of more selective therapeuticswith improved efficacy and/or fewer side effects.
Abbreviations: GPCR, G protein-coupled receptor; CHO cells, chinese hamstervary cells; U2OS cells, human bone osteosarcoma epithelial cells; �2C-C1 cells,drenergic �2C receptor expressing CHO-K1 cells; cAMP, adenosine 3′ ,5′-cycliconophosphate; [35S]GTP�S, guanosine 5′-O-(3-[35S]thio)triphosphate; DMSO,
imethyl sulfoxide; B-HT 920, 2-amino-6-allyl-5,6,7,8-tetra-hydro-4H-thiazolo-5,4-d]-azepine di-HCl; clonidine, 2-[2,6-dichloroaniline]-2-imidazolineHCl; gua-abenz, 1-[2,6-dichlorobenzylideneamino] guanidine; MK912, (2S,12bS)1′ ,3′-imethylspiro(1,3,4,5′ ,6,6′ ,7,12b-octahydro-2H-benzo[b]furo[2,3-a]quinolizine)-2,′-pyrimidin-2′-one; moxonidine, 4-chloro-N-(4,5-dihydro-1H-imidazol-2-yl)-6-ethoxy-2-methylpyrimidin-5-amine; oxymetazoline, 2-(3-hydroxy-2,6-dimeth-
l-4-t-butylbenzyl)-2-imidazoline; NE, norepinephrine; brimonidine or UK14,304,-bromo-N-[4,5-dihydro-1H-imidazol-2-yl]-6-quinoxalinamine.∗ Corresponding author at: H-1475 Budapest 10, POB 27, Hungary.el.: +36 1 4326188; fax: +36 1 8898400.
E-mail address: [email protected] (D. Kurko).
ttp://dx.doi.org/10.1016/j.brainresbull.2014.07.005361-9230/© 2014 Elsevier Inc. All rights reserved.
© 2014 Elsevier Inc. All rights reserved.
1. Introduction
Seven transmembrane receptors are significant targets for ther-apeutic intervention. Among them the �2-adrenergic receptors(�2-ARs) constitute a distinct family of catecholamine G-protein-coupled receptors (GPCRs). Physiological responses to endogenouscatecholamines epinephrine (adrenaline) and norepinephrine(noradrenaline) are mediated by three �2-AR subtypes: �2A, �2Band �2C (Bylund, 1992, 2005; Hieble et al., 1995; Docherty, 1998;Calzada and Artinano, 2001).
Being involved in many diverse cellular functions and phys-iological processes, these receptors are considered as importantdrug targets in the treatment of elevated blood pressure, pain,
opioid and alcohol withdrawal symptoms, in anesthetic care andalso for treatment of glaucoma, spasticity, depression, attention-deficit/hyperactivity disorder, obesity and diabetes (Serle et al.,1991; Ruffolo et al., 1993, 1995; Ruffolo and Hieble, 1994; Lakhlani
9 arch B
eSedat(aa�steT2arus
dppt�osk2
svg(�aalgeepeweb
mpaUtoeilpeptpbotiK
0 D. Kurko et al. / Brain Rese
t al., 1997; Arnsten, 1998; Kable et al., 2000; Crassous et al., 2007;anders and Maze, 2007; Arcangeli et al., 2009; Cinnamon Bidwellt al., 2010; Cottingham et al., 2011). However, currently availablerugs acting at the �2-AR show only marginal subtype selectivity,nd consequently possess several unwanted side-effects (e.g. seda-ion and cardiovascular effects) that limit their therapeutic valueKamibayashi and Maze, 2000; Maze and Fujinaga, 2000). In vitrond in vivo pharmacological analysis of adrenergic ligands, as wells recent studies on animals overexpressing or lacking different2-AR subtypes suggest specific functional roles of the receptorubtypes and highlight the possibility for developing subtype selec-ive therapeutic agents with improved efficacy and/or reduced sideffect profile (Lakhlani et al., 1997; Kable et al., 2000; Scholz andonner, 2000; Philipp et al., 2002; Wang et al., 2004; Knaus et al.,007; Cottingham et al., 2011). Furthermore, during the last decade,ccumulating experimental data have suggested that, in addition toeceptor subtype selectivity, ligands may be selective for individ-al signaling pathways as well as coupled to a particular receptorubtype (Kenakin, 1995; Berg and Clarke, 2006).
It is well known that GPCR signaling is a complex signal trans-uction network. GPCRs can couple to several different effectorathways, having the capacity to interact with a number of Grotein subtypes and with additional signaling or effector pro-eins. The most prominent G protein-independent pathway is the-arrestin system. In addition to its role in the desensitizationf G protein-mediated signaling, �-arrestin may activate distinctignaling pathways as well (e.g. nonreceptor tyrosine kinases, MAPinases, E3 ubiquitin ligases) (DeFea et al., 2000a; Luttrell et al.,001; Luttrell and Gesty-Palmer, 2010).
To date, all the three �2-AR subtypes appear to couple to theame signaling systems. Inhibition of adenylyl cyclase (AC), acti-ation of receptor operated K+ channels and inhibition of voltageated Ca2+ channels through pertussis toxin sensitive G proteinsGi1, Gi2, Gi3) are the main signal transduction mechanisms of the2-AR subtypes (Limbird, 1988; Abdulla and Smith, 1997; Saundersnd Limbird, 1999). Cell type specific stimulatory effects of �2-ARgonists have also been described, including coupling to stimu-atory cholera toxin-sensitive Gs proteins with increased cAMPeneration (Eason et al., 1992, 1994; Pohjanoksa et al., 1997; Jaspert al., 1998) and Ca2+ mobilization (Dorn et al., 1997; Kukkonent al., 1998; Reynen et al., 2000). Additional coupling of �2-ARs tohospholipase A2 (Jones et al., 1991), phospholipase D (MacNultyt al., 1992), �-arrestins (Wu et al., 1997; DeGraff et al., 1999), asell as activation of extracellular regulated kinases ERK1/2 (Alblas
t al., 1993; Flordellis et al., 1995; Stamer et al., 1996) have alsoeen reported.
However, it was quite surprising when it was first noted, that inost cases agonists do not uniformly activate all cellular signaling
athways linked to a given receptor, a characteristic referred tos biased agonism or functional selectivity (Kenakin, 2005, 2007;rban et al., 2007; Kenakin and Miller, 2010). The possible explana-
ion of biased signaling is that different agonists can either inducer stabilize distinct active conformations of the receptor (Ghanounit al., 2001; Kenakin, 2002; Vauquelin and Van Liefde, 2005), whichn turn differentially activates the effector signaling moleculeseading to different levels of stimulation of diverse downstreamathways (Kenakin, 2004; Galandrin and Bouvier, 2006; Galandrint al., 2007). Therefore, relying on a single functional assay canotentially lead to inappropriate agonist or antagonist classifica-ion. In addition, quantifying the level of bias is also important forharmacological characterization of compounds and for designingiased drugs. Especially since it has become clear that development
f ligands with pathway-selective activities may have considerableherapeutic benefit. In some instances, �-arrestin-bias confers pos-tive effect, as seen for the �-blocker carvedilol (Noma et al., 2007;im et al., 2008), the AT1aR ligand TRV120027 (Rajagopal et al.,ulletin 107 (2014) 89–101
2006; Violin et al., 2010), or the PTH1 ligand PTH-�arr (Gesty-Palmer et al., 2009; Bohinc and Gesty-Palmer, 2011). Adverse sideeffects can also be attributed to �-arrestin-dependent signaling,as it has been shown for niacin, a ligand of GPR109A (Walterset al., 2009; Kammermann et al., 2011) or TRV130 a �-opioid biasedagonist (DeWire et al., 2013). Several perfectly (e.g. carvedilol) orweakly biased ligands (e.g. formoterol, salmeterol acting at �2ARs)had been successfully used in the clinic before their biased naturewas discovered, suggesting that a number of drugs that are used inthe clinic today may also have similar biased effects (Kim et al.,2008; Rajagopal et al., 2011). Nonetheless, it can be difficult toidentify ligand bias in practice in various experimental systems.The absolute potency and efficacy of an agonist is dependent onthe properties of the biological assay system. Although a rever-sal of rank order of potency or efficacy would be evidence forligand bias (Berg et al., 1998; Kenakin, 2007), experimental condi-tions, differences in receptor reserve and amplification in differentassays have a major influence on the potency and efficacy val-ues in those assays, and therefore preclude a direct comparison ofresults to determine true ligand bias between signaling pathways.To overcome these problems, several new methods to quantifyligand bias have been recently developed and validated with exper-imental data. These rely on pairwise comparison of equimolarconcentrations, estimation of coupling efficiencies or transduc-tion coefficients derived from the operational model of agonismdeveloped by Black and Leff (1983) and determination of intrin-sic relative activity (RA) values (Ehlert, 2005, 2008; Figueroa et al.,2009; Evans et al., 2011; Rajagopal et al., 2011; Kenakin et al.,2012; Kenakin and Christopoulos, 2013; Shonberg et al., 2013). Atpresent there is an ongoing debate regarding the optimal methodfor quantifying ligand bias, as each of these approaches has advan-tages and limitations (more detailed in Section 4). However, thereis an agreement in the literature that if Hill slopes of the ago-nist concentration–response curves do not differ significantly fromunity, RA values are identical to transduction coefficients (Kenakinand Christopoulos, 2013), but estimation of bias obtained by thisapproach is not limited by the problematic fitting routine requiredfor calculating transduction coefficients (Rajagopal, 2013).
There have been a great amount of observations of functionalselectivity published by a number of groups investigating differentreceptor systems (reviewed in Violin and Lekowitz, 2007; Urbanet al., 2007; Kenakin, 2011; Whalen et al., 2011). However, toour knowledge, the present study is the first to evaluate the dif-ferences in ability of the �2C-AR agonists to modulate G proteinand �-arrestin pathways, as well as to induce receptor internal-ization. Accordingly, we compared the agonistic activity of severalligands; the endogenous agonist norepinephrine (NE), anotherphenylethylamine (phenylephrine), a guanidine (guanabenz), anox/thiazoloazepine (BHT-920) besides of four imidazolines (bri-monidine, oxymetazoline, clonidine, moxonidine) in four differenteffector systems. The inhibitory effects of two non-subtype-selective antagonists, MK912 and yohimbine, were also tested. Asdifferences in the rank order of efficacies and potencies of G protein-dependent and -independent signaling pathways were observed,the RA method was used to determine the potential of �2C-ARligands for biased signaling of and to quantify their functional selec-tivity.
2. Material and methods
2.1. Cell lines and culturing
Chinese hamster ovary (CHO-K1) cells expressing recombinanthuman �2C-ARs (�2C-C1 cell line) were purchased from Euro-screen (cat. nr. ES-032-C, Brussels, Belgium). CHO-K1 cells stably
arch B
epCwi1L1ag
lCsp
2
t(vc1TrNMiSbScStbtscpic3trlwbtsgwloeettmap
2
t
D. Kurko et al. / Brain Rese
xpressing recombinant human �2C-ARs and the chimeric G�qi5rotein were generated as described previously (Kurko et al., 2009).ells were cultured in Ham’s F12 nutrient mixture supplementedith l-glutamine (Gibco, Life Technologies, Carlsbad, CA) contain-
ng 10% fetal bovine serum (FBS, Gibco), 2.5 �g/ml amphotericin B,00 U/ml penicillin G, 100 �g/ml streptomycin (Sigma–Aldrich, St.ouis, MO), 1× non-essential amino acid mixture (Sigma), 1× RPMI-640 vitamin solution (Sigma–Aldrich), 400 �g/ml G418 (Gibco)nd for �2C/G�qi5 expressing cells 200 �g/ml hygromycin (Invitro-en, Life Technologies, Carlsbad, CA).
The PathHunter® ADRA2C total GPCR internalization U2OS celline (cat. nr. 93-0923C3) was purchased from DiscoveRx (Fremont,A, USA). Human bone osteosarcoma epithelial cells (U2OS line)tably expressing human �2C-ARs were cultured according to therotocol provided by DiscoverX.
.2. Measurement of intracellular cAMP levels
cAMP measurements were performed with a homogeneousime-resolved fluorescence (HTRF®) kit from Cisbio InternationalCodolet, France). The assay procedures followed the protocol pro-ided with the cAMP Dynamic kit (cat. nr: 62AM6PEC). Briefly,ells were seeded in 96-well half-area plates at a density of0,000 cells/well and cultured overnight at 37 ◦C and 5% CO2.he next day culture medium was removed from the cells andeplaced with agonist or vehicle containing assay buffer (140 mMaCl, 5 mM KCl, 5 mM HEPES-Na, 5 mM HEPES, 2 mM CaCl2, 2 mMgCl2, 10 mM glucose, pH = 7.4) supplemented with 100 �M 3-
sobutyl-1-methylxanthine (IBMX, phosphodiesterase inhibitor;igma–Aldrich) and 0.1% BSA (Sigma–Aldrich). When applica-le, cells were incubated with 100 ng/ml pertussis toxin (PTX;igma–Aldrich) in culture medium for 18 h before performing theAMP assay. All agonists and antagonists were purchased fromigma–Aldrich. 20 �l of agonist solution was added from a stockwo-fold more concentrated than the final concentration and incu-ated for 20 min at room temperature (RT). When antagonists wereested, an additional pre-incubation step of 10 min at RT was neces-ary and the stocks of agonists and antagonists were 4-fold higheroncentrated than the final concentration due to the dilution. Com-ounds were dissolved in DMSO (except NE, which was dissolved
n distilled water) and further diluted in buffer solution to a finaloncentration of 0.3% (v/v) DMSO (Sigma–Aldrich). After additional0 min incubation step at RT with forskolin (20 �l, final concentra-ion 1 �M) cell stimulation was stopped by adding the detectioneagents (20 �l cAMP-d2 and 20 �l anti-cAMP cryptate) diluted inysis buffer. The time-resolved fluorescence signal was quantified
ith the multi-mode reader PHERAstar FS (BMG Labtech, Orten-erg, Germany) after 60 min of incubation at RT. The ratio betweenhe acceptor fluorescence signal (A665 nm) and donor fluorescenceignal (A620 nm) × 104, representing the FRET between the conju-ated cAMP and the anti-cAMP antibody, was calculated for eachell of the assay plate. 1 �M forskolin-stimulated cAMP accumu-
ation in the absence of agonist was defined as 100%. Emax values (%f maximal inhibition of forskolin-stimulated cAMP) achieved forach drug were normalized to the response evoked by a maximallyffective concentration of brimonidine (UK14,304; 1 �M) tested inhe same experiment. The EC50 values (the concentration of agonisthat produces 50% inhibition of forskolin-stimulated cAMP accu-
ulation) were calculated by Microcal Origin 6.0 software. Datare presented as mean ± S.E.M. of four independent experimentserformed in triplicate on different days.
.3. Measurement of cytosolic free Ca2+ concentration
Fluorometric measurements of cytoplasmic calcium concentra-ion ([Ca2+]i) in �2C and G�qi5 co-expressing CHO-K1 cells were
ulletin 107 (2014) 89–101 91
carried out without modifications as described previously (Kurkoet al., 2009). All experiments were performed at least three times,on different days. Data are expressed as mean ± S.E.M.
2.4. PathHunter® ˇ-arrestin recruitment assay
The �-arrestin translocation assay was performed using thePathHunter® assay from DiscoveRx. The assay principle is basedon the enzyme fragment complementation (EFC) between two por-tions of �-galactosidase (�-gal). The receptor is labeled with a small42 amino acid fragment of �-gal, called ProLink (PK), and is stablyco-expressed in cells with �-arrestin-2 labeled with the N-terminaldeletion mutant of �-gal, called enzyme acceptor (EA). Activationof the �2C-AR stimulates �-arrestin recruitment and forces comple-mentation of the two enzyme fragments, leading to an increase inholoenzyme activity detectable as a chemiluminescent signal (RLU)due to enzyme substrate hydrolysis.
Experiments were carried out using the PathHunter® eXpressADRA2C �-arrestin GPCR Assay kit (cat. nr: 93-0218E2) accordingto the manufacturer’s recommendations. Briefly, frozen CHO-K1cells expressing tagged human �2C-AR were thawed and seeded inwhite-walled, clear-bottomed 96-well plates at 10,000 cells/well in100 �l PathHunter® cell plating 0 reagent (CP0, DiscoverX). Whenapplicable, cells were pre-treated overnight with 100 ng/ml PTXbefore performing the assay. After 48 h incubation in a humidifiedatmosphere at 37 ◦C and 5% CO2 cells were stimulated with agonistor vehicle (10 �l) for 90 min at 37 ◦C. When evaluated in antag-onistic format, cells were preincubated with antagonist (5 �l) for30 min before agonist (5 �l) addition. Compounds were dissolvedin DMSO (except norepinephrine, which was dissolved in distilledwater) and further diluted in CP0 reagent to a final concentrationof 0.9% (v/v) DMSO. Following compound incubation PathHunter®
detection reagent solution (55 �l) was added to each well and lumi-nescence was read on the multi-mode reader Synergy4 (BioTek,Winooski, VT, USA) after an additional 60 min incubation at RT.Agonist responses were normalized to the maximum responseobtained with 30 �M UK14,304 in the same experiment. EC50 andIC50 values were calculated from concentration–response curvesby sigmoidal fitting using Microcal Origin 6.0 software. All exper-iments were performed three times, on different days. Data areexpressed as mean ± S.E.M.
2.5. PathHunter® internalization assay
Receptor internalization was monitored using the PathHunter®
ADRA2C total GPCR internalization assay from DiscoveRx (cat. nr:93-0923C3). This assay provides a direct and quantitative mea-surement of internalized �2C-AR localized in early endosomesusing �-gal enzyme fragment complementation. Activated enzymeacceptor (EA) tagged �2C-AR is internalized in PK-tagged endo-somes forcing complementation (EFC) of the two �-galactosidaseenzyme fragments (EA and PK). The resulting increase in enzymeactivity is measured using chemiluminescent PathHunter® detec-tion reagents. Experiments were performed on U2OS cells stablyexpressing the labeled human �2C-ARs. Cells were grown overnightat a density of 10,000 cells per well in 55 �l PathHunter® cell plating5 reagent (CP5, DiscoverX) in white-walled, clear-bottomed 96-well plates at 37 ◦C and 5% CO2. The next day, 55 �l of agonist orvehicle solution was added to the cells for 180 min at 37 ◦C. Whenstudying the role of Gi protein activation on receptor internaliza-tion, cells were pretreated overnight with 100 ng/ml PTX beforeaddition of agonists. Compounds dissolved in DMSO (except NE
which was dissolved in distilled water) were diluted in CP5 reagentto a final concentration of 0.9% (v/v) DMSO. For antagonist testing,cells were treated with antagonist (27.5 �l) for 30 min before ago-nist (27.5 �l) addition. After 180 min compound incubation 55 �l
9 arch Bulletin 107 (2014) 89–101
pElmnwvmwa
2
i1dbcncwllcf�bpbscSvwwe1
3
3t
3p
cdt(fcatowmoppbUic
A
B
1E-3 0,0 1 0, 1 1 10 10 0 100 0 100 00
0
20
40
60
80
100
cAM
P re
spon
seno
rmal
ized
fors
kolin
inhi
bitio
n (%
)
[Lig and] (nM)
UK14 ,30 4 Clon idine Moxon idine Oxym etazoli ne
0,01 0, 1 1 10 100 1000 10000
0
20
40
60
80
100
cAM
P re
spon
seno
rmal
ized
fors
kolin
inhi
bitio
n (%
)
[Ligand] (n M)
No repinep hrine BH T-92 0 Gua nabenz Phen ylephr ine
Fig. 1. Inhibition of forskolin-induced cAMP accumulation in CHO-K1 cells express-ing recombinant �2C-ARs. Concentration–response curves for UK14,304, clonidine,moxonidine, oxymetazoline (panel A) and NE, BHT-920, guanabenz, phenylephrine(panel B) are shown. Data are expressed as relative inhibition of cAMP response
2 D. Kurko et al. / Brain Rese
er well of PathHunter® detection reagent was added, and theFC reaction was carried out for 60 min at RT in the dark. Chemi-uminescence, indicated as relative luminescence units (RLU), was
easured on the multi-mode reader Synergy4 (BioTek) and wasormalized for each agonist to the maximum response obtainedith 180 �M UK14,304 in the same experiment. EC50 and IC50
alues were calculated from concentration–response curves by sig-oidal fitting using Microcal Origin 6.0 software. All experimentsere performed three times, on different days. Data are expressed
s mean ± S.E.M.
.6. Bias factor determination
For quantifying ligand bias we have chosen the relative activ-ty (RA) scale, proposed by Ehlert and co-workers (Ehlert et al.,999; Ehlert, 2005, 2008; Figueroa et al., 2009). RA values wereefined as the maximal response of the agonist (Emax) dividedy its EC50 value. Emax and EC50 values were estimated fromoncentration–response curves of agonists with slopes not sig-ificantly different from unity. In order to cancel the impact ofell-dependent effects on the observed agonism, log(RA) valuesere then normalized to the log(RA) values of the endogenous
igand NE obtaining log(RAn) in the form of �log(RA) values. Onceog(RAn) values were obtained for each pathway, cross-pathwayomparisons were made by calculating ��log(RA) values asollows: ��Log(RA) = log(RAn)[pathway1] − log(RAn)[pathway2].
�log(RA) values significantly different from 0 revealed ligandias for an agonist for one pathway over another, com-ared to the reference agonist NE. For the characterization ofias, bias factors were defined as 10��log(RA) values. For thetandard agonist NE the bias factor was 1. Statistical signifi-ance was determined using unpaired Student’s t test. As thetudent’s t test assumes normal distribution, calculation of RAalues as logarithms allowed valid statistical comparison. Thisas carried out on the raw log(RA) values, and determinedhether the log(RA)[pathway1] − log(RA)[pathway2] value for
ach agonist was significantly different from the log(RA)[pathway] − log(RA)[pathway 2] value for the reference agonist.
. Results
.1. Analysis of compound activity at the ˛2C-AR signalinghrough different G proteins
.1.1. ˛2C-AR mediated inhibition of cAMP accumulation via G˛iroteins
Agonist activity for Gi protein signaling was tested in CHO-K1ells stably expressing �2C-ARs by measuring concentrationependent inhibition of forskolin-stimulated cAMP accumula-ion. EC50 values deduced from concentration–response curvesFig. 1A and B, Table 1) revealed a rank order of potency asollows: NE ≥ UK14,304 > BHT-920 > oxymetazoline > guanabenz >lonidine > moxonidine > phenylephrine. While NE, UK14,304nd BHT-920 inhibited forskolin-stimulated cAMP accumula-ion almost completely (> 94%), other compounds tested werenly partially effective (Table 1). Their rank order of efficacyas: UK14,304 ≥ NE > BHT-920 > oxymetazoline ≥ clonidine >oxonidine > phenylephrine ≥ guanabenz. The inhibitory effect
f UK14,304 on the forskolin-induced cAMP response was com-letely blocked in cells pretreated with 100 ng/ml PTX for 18 hrior agonist stimulation (data not shown). MK912 and yohim-
ine concentration-dependently reversed the agonist effect ofK14,304 (Fig. 7A and B), with estimated IC50 values presentedn Table 1. Functional activities of compounds in the cAMP assayorrelated well with their binding affinities (Ki) obtained from a
(induced by 1 �M forskolin in the absence of compounds) normalized to the effectof 1 �M UK14,304. Data points represent the mean ± S.E.M. of 4 independent exper-iments performed in triplicate.
competitive binding assay using [3H]-UK14,304 on cell membranesprepared from �2C-C1 cells (Kurko et al., 2009).
3.1.2. ˛2C-AR mediated cytoplasmic Ca2+ response via chimericG˛qi5 proteins
Agonists for Gi protein coupled 7TMRs are often identified inhigh-throughput screens based on receptor coupling to alternativeG proteins that mobilize Ca2+ (e.g. G�qi5, G�16). In a prior studywe have shown that signaling of �2C-ARs can be diverted to Ca2+
mobilization through chimeric G�qi5 proteins without alteringreceptor pharmacology (Kurko et al., 2009). In CHO-K1 cells stablyco-expressing �2C-ARs and G�qi5 proteins BHT-920, guanabenz,moxonidine and phenylephrine concentration dependently stim-ulated Ca2+ release (Fig. 2). Concentration-dependence of NE,UK14,304, clonidine and oxymetazoline in elevating cytoplasmicCa2+ and the inhibitory effect of antagonists were determinedpreviously (Kurko et al., 2009). Similar to the cAMP assay, full andpartial agonists could have been discriminated with the Ca2+ assayas well. Emax values relative to UK14,304 ranged from 97% (for NE)down to 76% (for phenylephrine) (Table 1), yielding a rank orderof efficacy as follows: UK14,304 ≥ NE > BHT-920 > moxonidine
≥ guanabenz ≥ oxymetazoline > clonidine > phenylephrine. Therank order of potency was: UK14,304 ≥ NE > oxymetazoline > BHT-920 > clonidine > guanabenz > moxonidine > phenylephrine.
D. Kurko et al. / Brain Research Bulletin 107 (2014) 89–101 93
Table 1Comparison of the pharmacological profiles of ligands acting at the human �2C-AR determined with multiple assay formats.
Compounds Ca2+ release cAMP �-Arrestin recruitment Internalization
pEC50/pIC50 Emax/Imax (%) pEC50/pIC50 Emax/Imax (%) pEC50/pIC50 Emax/Imax (%) pEC50/pIC50 Emax/Imax (%)
Norepinephrine 8.61 ± 0.06a 97 ± 0.4a 8.60 ± 0.03 97 ± 2 6.60 ± 0.03 103 ± 3 6.02 ± 0.04 108 ± 0.5Phenylephrine 6.44 ± 0.04 76 ± 4 6.45 ± 0.03 81 ± 1 5.72 ± 0.06 52 ± 5 4.64 ± 0.01 45 ± 0.3UK14,304 8.77 ± 0.03a 100a 8.51 ± 0.02 99 ± 1 6.88 ± 0.02 100 ± 0.2 5.88 ± 0.05 100 ± 0.9Clonidine 8.16 ± 0.04a 77 ± 1a 7.74 ± 0.03 88 ± 2 6.74 ± 0.03 36 ± 3 4.64 ± 0.03 35 ± 4Moxonidine 6.94 ± 0.04 84 ± 1 6.60 ± 0.03 82 ± 3 5.35 ± 0.01 65 ± 2 4.73 ± 0.02 43 ± 3Oxymetazoline 8.56 ± 0.08a 82 ± 3a 8.22 ± 0.08 88 ± 4 6.92 ± 0.02 25 ± 2 n.d. 23 ± 0.1BHT-920 8.24 ± 0.02 94 ± 4 8.40 ± 0.01 94 ± 4 6.68 ± 0.08 64 ± 2 5.56 ± 0.04 65 ± 3Guanabenz 7.67 ± 0.04 83 ± 5 7.86 ± 0.05 80 ± 6 6.57 ± 0.07 60 ± 1 5.14 ± 0.04 64 ± 7MK912 9.12 ± 0.03a 101 ± 1a 9.34 ± 0.02 102 ± 3 9.55 ± 0.02 98 ± 2 8.20 ± 0.02 100 ± 0.3Yohimbine 7.76 ± 0.07a 101 ± 1a 8.16 ± 0.03 97 ± 2 7.90 ± 0.03 101 ± 1 6.48 ± 0.01 101 ± 2
pEC50 and Emax values (maximal responses expressed as percentage of the maximal response of UK14,304) were deduced from the agonist concentration–response curves ofvarious compounds. pIC50s were determined against an ∼EC80 concentration of UK14,304. cAMP data were obtained using �2C-AR expressing CHO-K1 cells; Ca2+ data werederived from CHO-K1 cells co-expressing �2C-AR and chimeric G�qi5 proteins; �-arrestin data were determined in CHO-K1 cells stably co-expressing PK-tagged �2C-AR andEA-tagged �-arrestin-2, while internalization data were derived from U2OS cells expressing EA-tagged �2C-AR and PK-tagged endosomes. Data in the table represent themean ± S.E.M. from 3 to 7 independent experiments. n.d., not determined.
a Data from Kurko et al. (2009).
1000010001001010,10,01
0
20
40
60
80
100
Ca2+
resp
onse
norm
aliz
ed re
spon
se (%
)
[Lig and] ( nM)
BHT -92 0 Guanaben z Moxon idine Phen ylephr ine
Fig. 2. Concentration-dependence of agonists-induced Ca2+ responses in �2C-ARand G�qi5 co-expressing CHO-K1 cells. Agonist responses were expressed as apercentage of the response observed with a maximally effective concentration ofUK14,304 (1 �M). Agonists tested were: BHT-920, guanabenz, moxonidine andpor
3t
3
ctwsdnipt9dmopccl
A
B
0,1 1 10 100 1000 10000 100000
0
20
40
60
80
100
arre
stin
recr
uitm
ent
norm
aliz
ed re
spon
se (%
)
[Ligand ] (nM)
UK14 ,30 4 Clon idine Moxon idine Oxymetazoli ne
0,1 1 10 100 1000 10000 10 0000
0
20
40
60
80
100
arre
stin
recr
uitm
ent
norm
aliz
ed re
spon
se (%
)
[Ligand] (nM)
Norep inephr ine BHT -92 0 Guana ben z Phen ylephrin e
Fig. 3. Concentration-dependence of agonist-induced �-arrestin recruitment in�2C-AR and �-arrestin expressing CHO-K1 cells. Agonist responses for UK14,304,clonidine, moxonidine, oxymetazoline (panel A) and NE, BHT-920, guanabenz,phenylephrine (panel B) were expressed as percentage of the response observed
henylephrine. Concentration-dependent effects of UK14,304, NE, clonidine andxymetazoline have been determined previously (Kurko et al., 2009). Data pointsepresent the mean ± S.E.M. of 4 independent experiments performed in triplicate.
.2. Analysis of compound activity at the ˛2C-AR signalinghrough G protein-independent pathways
.2.1. ˛2C-AR mediated ˇ-arrestin recruitment�-Arrestin recruitment was studied in PathHunter eXpress
ells engineered for human �-arrestin-2 translocation in responseo �2C-AR activation. Proximity of the �2C-ARs and �-arrestin-2as measured by �-gal fragment complementation. Agonist
timulation of the cells for 90 min resulted in a concentration-ependent increase of �-gal activity (Fig. 3A and B). EC50 andormalized Emax values for all the agonists tested are shown
n Table 1. The rank order of potency was quite different com-ared to that of agonist potencies observed in the cAMP orhe Ca2+ assay: oxymetazoline > UK14,304 > clonidine > BHT-20 > NE > guanabenz > phenylephrine > moxonidine. Similarly,ifference in rank order of efficacy was observed: NE ≥ UK14,304 >oxonidine ≥ BHT-920 > guanabenz > phenylephrine > clonidine >
xymetazoline. The �-arrestin assay was able to detect full and
artial agonists, but lower efficacy values were obtained whenompared to the former cAMP or Ca2+ assays. Moreover, poten-ies of agonists for recruiting �-arrestin were up to 1.6 log unitower than those measured for inhibiting cAMP accumulation orwith UK14,304 at a maximally effective concentration (30 �M). Data points repre-sent the mean ± S.E.M. of 3 independent experiments performed in duplicate.
elevating cytoplasmic Ca2+. UK14,304-induced �-arrestin asso-
ciation was concentration-dependently inhibited by pretreatingthe cells with MK912 or yohimbine (Fig. 7A and B, Table 1). Theconcentration–response relationship of UK14,304 for �-arrestin
94 D. Kurko et al. / Brain Research Bulletin 107 (2014) 89–101
10 100 100 0 10000 10 0000
0
20
40
60
80
100A
B
inte
rnal
izat
ion
norm
aliz
ed re
spon
se (%
)
[Ligand ] (nM)
UK14 ,30 4 clon idine Moxon idine
1 10 100 1000 1000 0 100000
0
20
40
60
80
100
inte
rnal
izat
ion
norm
aliz
ed re
spon
se (%
)
[Lig and] (nM)
Norepinephr ine BHT -920 Guan aben z Phen ylephr ine
Fig. 4. Concentration-dependence of agonist-induced internalization in �2C-ARexpressing U2OS cells. Agonist responses for UK14,304, clonidine, moxonidine(panel A) and NE, BHT-920, guanabenz, phenylephrine (panel B) were expressedas percentage of the response observed at the maximal effective concentrationoe
t1
3
ripisnwgzgzz1wal�Ust
-20 0 20 40 60 80 100 120 140 16 0 180 200
100
200
300
400
500
Rel
ativ
e in
crea
se in
RLU
sig
nal
Time /min
UK14,304 NA Moxonidine Ox ymetazo lin e Clonidine BHT-920 Phenylephrine Guanabenz
Fig. 5. Time course of the agonist-induced internalization of �2C-AR. U2OS cells
f UK14,304 (180 �M). Data points represent the mean ± S.E.M. of 3 independentxperiments performed in duplicate.
ranslocation was not affected by an overnight pretreatment with00 ng/ml PTX (data not shown).
.2.2. ˛2C-AR mediated receptor internalizationTo examine the effects of agonist stimulation on cell surface
eceptor internalization we used the PathHunter® total GPCRnternalization assay based on �-gal enzyme fragment com-lementation technology. All agonists tested induced �2C-AR
nternalization as evidenced by the increase in chemiluminescentignal due to �-gal activity. After 180 min pretreatment partial ago-ists induced less internalization than full agonists (Fig. 4A and B)ith a rank order of efficacy as follows: NE > UK14,304 > BHT-920 >
uanabenz > phenylephrine > moxonidine > clonidine > oxymeta-oline. The rank order of potency was: NE > UK14,304 > BHT-920 >uanabenz > moxonidine > clonidine > phenylephrine > oxymeta-oline. It was not possible to determine an EC50 value for oxymeta-oline because of its very low activity (23% activity at 180 �M and2% at 90 �M). Agonist Emax values for receptor internalizationere almost identical to those for �-arrestin recruitment to
ll agonists tested. However, agonists proved to be 1 log unitess potent in inducing receptor internalization than in evoking
-arrestin translocation (Table 1). In the presence of antagonistsK14,304 failed to induce internalization, demonstrating thepecificity of the assay (Fig. 7A and B). Estimated IC50 values forhe inhibitory activity of MK912 and yohimbine are presented
were treated with agonists (180 �M) for 15–180 min and the relative increase inthe RLU signal was plotted against time. Data points represent the mean ± S.E.M. of3 independent experiments performed in triplicate.
in Table 1. Overnight treatment of cells with PTX did not pre-vent internalization, neither efficacies nor potencies of NE andUK14,304 were significantly affected in the presence of 100 ng/mlPTX (data not shown).
Kinetics of the �2C-AR internalization was also examined.�2C-AR expressing U2OS cells were preincubated with agonists(180 �M) for various time periods and the extent of receptorinternalization was assessed (Fig. 5). The extent of internalizationcaused by NE after 180 min (the maximal time period studied)was similar to that evoked by UK14,304, although the time coursefor NE was significantly slower, reaching about 3, 14 and 33%of maximal internalization for the time period studied after 15,30 and 60 min, respectively. At these time points UK14,304 pro-duced about 54, 59 and 69% internalization (P < 0.001). Althoughpartial agonists induced less receptor internalization, the time-course for internalization induced by other imidazoline compounds(i.e. oxymetazoline, moxonidine and clonidine) were very similarto that of UK14,304, showing a significantly more rapid internal-ization pattern compared to NE. Representative data of �2C-ARinternalization induced by 15 min agonist treatment are shown inFig. 6. ANOVA followed by Tukey’s multiple comparisons post hocanalysis revealed significant differences in time-course of inter-nalization for UK14,304, oxymetazoline and phenylephrine up to60 min, and for clonidine and moxonidine up to 30 min comparedto that of NE.
3.3. Analysis of functional selectivity at the ˛2C-AR
Our data indicated that there were interesting differences inagonist activity between the four signaling endpoints measured.Among data presented in Table 1, there were both reversals inrank order of efficacies and reversals in rank order of potenciesbetween agonists. The most striking differences were observedwith oxymetazoline, UK14,304 and clonidine having higher poten-cies for �-arrestin recruitment than NE but lower potencies forcAMP accumulation, Ca2+ release or receptor internalization com-pared to the endogenous ligand. Although these reversals inpotency may indicate functional selectivity, there is evidence thatsuch reversals are not a necessary condition (Evans et al., 2010).
Therefore we determined the intrinsic relative activities (RA) of�2 adrenergic agonists to derive quantitative measures of func-tional selectivity through the four different signaling pathways.The logarithms of the RA values (log(RA)) of agonists for eliciting
D. Kurko et al. / Brain Research Bulletin 107 (2014) 89–101 95
Table 2Calculation of logarithm of relative activity (log(RA)) values for ligands acting at the �2C-AR.
Compounds Ca2+ release cAMP �-Arrestin recruitment Internalization
log RA log RAn log RA log RAn log RA log RAn log RA log RAn
Norepinephrine 1.71 ± 0.07 0 ± 0.10 1.59 ± 0.03 0 ± 0.04 −0.38 ± 0.03 0 ± 0.04 −0.96 ± 0.03 0 ± 0.04Phenylephrine −0.68 ± 0.05 −2.40 ± 0.09 −0.65 ± 0.03 −2.24 ± 0.04 −1.57 ± 0.01 −1.19 ± 0.03 −2.71 ± 0.01 −1.75 ± 0.03UK14,304 1.77 ± 0.03 0.06 ± 0.08 1.51 ± 0.02 −0.08 ± 0.04 −0.12 ± 0.02 0.26 ± 0.04 −1.12 ± 0.05 −0.16 ± 0.06Clonidine 0.07 ± 0.04 −0.67 ± 0.08 0.66 ± 0.01 −0.93 ± 0.03 −0.68 ± 0.05 −0.3 ± 0.06 −2.82 ± 0.06 −1.86 ± 0.07Moxonidine −0.13 ± 0.05 −1.85 ± 0.09 −0.48 ± 0.01 −2.08 ± 0.03 −1.87 ± 0.02 −1.48 ± 0.04 −2.64 ± 0.02 −1.69 ± 0.03Oxymetazoline 1.47 ± 0.08 −0.25 ± 0.11 1.18 ± 0.07 −0.41 ± 0.08 −0.69 ± 0.003 −0.31 ± 0.03 n.a. n.a.BHT-920 1.22 ± 0.03 −0.5 ± 0.08 1.38 ± 0.02 −0.21 ± 0.04 −0.51 ± 0.09 −0.13 ± 0.1 −1.63 ± 0.04 −0.68 ± 0.05Guanabenz 0.61 ± 0.06 −1.1 ± 0.09 0.78 ± 0.05 −0.81 ± 0.06 −0.67 ± 0.07 −0.28 ± 0.08 −2.04 ± 0.04 −1.08 ± 0.05
RA values were calculated from data presented in Table 1 and were defined as the maximal response of the agonist (Emax) divided by its potency expressed as an EC50 value.log(RA) was the logarithm of RA values. In order to cancel observational bias resulting from different sensitivities of the assays, log(RA) values were normalized to the log(RA)values of the endogenous ligand NE obtaining log(RAn) values. n.a., not applicable.
NA
oxymetazo
lineUK14
,304moxo
nidinecloni
dine
phenyleph
rineBHT
-920guan
abenz
0
20
40
60
80
100
120
****
***
***
inte
rnal
izat
ion
norm
aliz
ed re
spon
se (%
)
15 min agon ist trea tment
Fig. 6. Internalization of the �2C-AR in U2OS cells after 15 min exposure to vari-ous agonists. Enzyme activity (RLU) values at different time points were expressedas a percentage of the maximal internalization seen after 180 min. ANOVA fol-lowed by Tukey’s post hoc analysis comparing normalized RLU values of differentagonists at various times to the values of NE revealed significant differences in time-course of internalization for UK14,304, oxymetazoline, moxonidine, clonidine andphenylephrine after 15 min treatment. The same significantly different internal-ization pattern was observed for UK14,304, oxymetazoline and phenylephrine upto 60 min, and for clonidine and moxonidine up to 30 min when compared to NE.Ct
rasctiotsbosia(rntspan
0,01 0,1 1 10 100 10 00
0
20
40
60
80
100A
B
inhi
bitio
n (%
)
[MK-912] (nM)
cA MP Ca2+
ββ-arres tin internali zati on
0,1 1 10 100 100 0 10000
0
20
40
60
80
100
inhi
bitio
n (%
)
[Yoh imbine] (n M)
cA MP Ca2+
ββ-arres tin internali zati on
Fig. 7. Concentration-dependent inhibition of various �2C-AR mediated responsesby MK912 (panel A) and yohimbine (panel B) determined with multiple assays.�2C-AR expressing cells pretreated with increasing concentrations of MK912 oryohimbine were incubated with a submaximal concentration of UK14,304 (∼EC80)and inhibitory potencies were assessed in cAMP, Ca2+, �-arrestin and internalizationassays. Preliminary experiments confirmed that both MK912 and yohimbine were
olumns represent the mean ± S.E.M. of 3 independent experiments performed inriplicate. *P < 0.05; **P < 0.01; ***P < 0.001.
esponses through G�i (Fig. 1), G�qi5 (Fig. 2), �-arrestin (Fig. 3)nd receptor internalization(Fig. 4) were calculated and analyzedtatistically as described under Section 2. We found that agonistoncentration–response curves typically exhibited Hill slopes closeo unity, suggesting that the log(RA) calculation was valid in thesenstances (with the sole exception of receptor internalization dataf oxymetazoline, which was so ineffective in inducing internaliza-ion, that the model was not applicable). As different intracellularignaling pathways have various sensitivities to agonists, ligandias must be detected and quantified by comparing the activitiesf agonists within one assay to a selected standard agonist. Table 2hows log(RA) ratios for each pathway and log(RA) values normal-zed to that of NE (log(RAn)). Bias factors calculated accordinglyre summarized in Table 3. Values significantly different from 1.0marked in bold in Table 3) indicated an agonist that promotedeceptor conformation distinct from that induced by NE. Althoughone of the tested agonists showed perfectly biased propertiesoward multiple signaling pathways, several compounds displayedome level of bias between the different readouts. When com-
aring bias factors derived from the RA scale for cAMP or Ca2+nd �-arrestin pathways, UK14,304, moxonidine, clonidine, gua-abenz and phenylephrine were identified as having bias toward
devoid of agonist activity in all assays. Data points represent the mean ± S.E.M. of3–4 independent experiments.
�-arrestin recruitment compared to NE. Phenylephrine displayeda high bias factor of 11.3 between �-arrestin translocation andcAMP accumulation and a bias factor of 16.2 between �-arrestintranslocation and Ca2+ release, suggesting that it was 11.3 and16.2 times more active than NE at recruiting �-arrestin. UK14,304,moxonidine, clonidine and guanabenz did not match the same
level of apparent bias of phenylephrine (bias factors from 1.6 to6.58), however, RA analysis ultimately showed that these valueswere significantly different for the marked pathways relative to
96 D. Kurko et al. / Brain Research B
Tab
le
3C
alcu
lati
on
of
bias
fact
ors
usi
ng
the
rela
tive
acti
vity
(RA
)
met
hod
for
liga
nd
s
acti
ng
at
the
�2C
-AR
.
Com
pou
nd
s
Log
RA
n[p
ath
way
1
−
pat
hw
ay
2]
[arr
esti
n-c
AM
P]
Bia
s
fact
or[a
rres
tin
-Ca2+
]
Bia
s
fact
or
[arr
esti
n-i
nte
rnal
]
Bia
s
fact
or
[Ca2+
-cA
MP]
Bia
s
fact
or
[Ca2+
-in
tern
al]
Bia
s
fact
or
[cA
MP-
inte
rnal
] B
ias
fact
or
Nor
epin
eph
rin
e
0
±
0.06
1
0
±
0.11
1
0
±
0.06
1
0
±
0.11
1
0
±
0.11
1
0
±
0.06
1Ph
enyl
eph
rin
e
1.05
±
0.05
***
11.3
1.21
±
0.1**
*16
.2
0.56
±
0.05
**3.
66
−0.1
6
±
0.10
0.7
−0.6
5
±
0.10
*0.
23
−0.4
9 ±
0.05
**0.
32U
K14
,304
0.35
±
0.06
*2.
220.
21
±
0.09
*1.
610.
43
±
0.07
*2.
67
0.14
±
0.09
1.33
0.22
±
0.09
1.66
0.08
± 0.
07
1.2
Clo
nid
ine
0.63
±
0.07
***
4.28
0.37
±
0.1*
2.33
1.56
±
0.09
***
36.7
0.26
±
0.09
1.83
1.2
±
0.11
**15
.7
0.93
± 0.
07**
8.57
Mox
onid
ine
0.59
±
0.05
***
3.92
0.36
±
0.09
*2.
30.
2
±
0.05
*1.
59
0.23
±
0.09
1.71
−0.1
6
±
0.09
0.69
−0.3
9
±
0.05
0.41
Oxy
met
azol
ine
0.10
±
0.08
1.26
−0.0
6±
0.11
0.86
n.a
.
n.a
.
0.16
±
0.13
1.46
n.a
.
n.a
.
n.a
.
n.a
.B
HT-
920
0.09
±
0.1
1.22
0.37
±
0.12
2.32
0.55
±
0.11
**3.
53
−0.2
8
±
0.09
0.53
0.18
±
0.09
1.52
0.46
±
0.06
2.89
Gu
anab
enz
0.53
±
0.1**
*3.
350.
82
±
0.12
***
6.58
0.80
±
0.09
**6.
28
−0.2
9±
0.11
0.51
−0.0
2
±
0.10
0.95
0.27
±
0.08
1.87
For
calc
ula
tion
of
pos
sibl
e
bias
in
sign
alin
g
tow
ard
s
mu
ltip
le
sign
alin
g
pat
hw
ays,
cros
s-p
ath
way
com
par
ison
s
wer
e
mad
e
by
calc
ula
tin
g
log(
RA
n)[
pat
hw
ay1]
−
log(
RA
n)[
pat
hw
ay2]
valu
es. B
ias
fact
ors
wer
e
defi
ned
as
10�
�lo
g(R
A)
valu
es. S
.E.M
. for
log(
RA
n)[
pat
hw
ay1]
−
log(
RA
n)[
pat
hw
ay2]
was
calc
ula
ted
as√
(S.E
.M. :
log(
RA
n)[
pat
hw
ay1]
)2+
(S.E
.M. :
log(
RA
n)[
pat
hw
ay2]
)2. S
tud
ent’
s
t
test
s
wer
e
carr
ied
out
on
the
raw
log(
RA
)
valu
es, a
nd
det
erm
ined
wh
eth
er
the
log(
RA
)[p
ath
way
1]
−
log(
RA
)[p
ath
way
2]
valu
e
for
each
agon
ist w
as
sign
ifica
ntl
y
dif
fere
nt f
rom
the
log(
RA
)[p
ath
way
1]
−
log(
RA
)[p
ath
way
2]
valu
e
for
the
refe
ren
ce
agon
ist N
E.
Bia
s
fact
ors
in
bold
den
otes
stat
isti
call
ysi
gnifi
can
t
dif
fere
nce
in
bias
com
par
ed
to
NE.
n.a
.,
not
app
lica
ble.
*P
<
0.05
.**
P
<
0.01
.**
*P
<
0.00
1.
ulletin 107 (2014) 89–101
the endogenous agonist. Bias factors for cAMP and Ca2+ pathwaysversus receptor internalization indicated statistically significantbias for phenylephrine toward internalization by factors of 3.1 and4.5, respectively. In contrast, clonidine was more active in inhibitingcAMP accumulation and elevating cytoplasmic Ca2+ than in pro-ducing internalization compared with NE. This translated in biasfactors of 8.6 and 15.7, consistent with cAMP and Ca2+ bias, respec-tively. All agonists were identified as having bias toward �-arrestinrecruitment versus receptor internalization (except oxymetazolinewhere no bias factor could be calculated). In contrast, there were nostatistically significant ligand bias observed between the cAMP andCa2+ pathways, suggesting that the compounds tested were aboutequieffective in the stabilization of signaling states of G�i and G�qi5proteins.
4. Discussion
Over the past few years, a growing number of articles havebeen published describing the identification of biased agonists at awide variety of GPCRs (Kenakin, 2011; Whalen et al., 2011; Wisleret al., 2014). However, there is no available data about the abilityof agonists acting at the �2C-AR to selectively stimulate G protein-dependent and -independent signaling pathways. We show in thisstudy that a series of agonists acting at the �2C-AR display func-tional selectivity through four signaling pathways.
Since G protein and �-arrestin pathways were found to be dis-tinct signaling routes (Luttrell et al., 1999; DeFea et al., 2000b;Luttrell and Gesty-Palmer, 2010), there is an opportunity to mod-ulate them independently by ligands having pathway-selectiveactivities. Recent studies suggest that the complex relationshipbetween G proteins and �-arrestins can influence both the efficacyand the side effects of a GPCR targeted drug, and combination ofthese activities determines in vivo effectiveness of the compound(Taylor et al., 2012). Therefore, besides classical second-messengerreadouts, �-arrestin-based assays are now widely applied in theanalysis of GPCR pharmacology (van Der Lee et al., 2008; Hansonet al., 2009; Rajagopal et al., 2011). However, compounds selectedon the basis of their potencies and/or efficacies alone can have otheractivities (e.g. receptor desensitization and internalization) thatmay limit their clinical effectiveness (e.g. tolerance, tachyphylaxis).
Based on this knowledge, our primary objective was to performa detailed pharmacological analysis of �2C-ARs by measuring theeffects of structurally different ligands on G�i and G�qi5 proteinsignaling, �-arrestin recruitment and receptor internalization. Wefound little difference in agonist activity for triggering responsesthrough the �2C-AR coupled to G�i (cAMP accumulation) or G�qi5(Ca2+ release), in agreement with previously reported results (Jeonet al., 1995; Parsley et al., 1999; Pauwels and Colpaert, 2000;Umland et al., 2001). Agonist potencies obtained by these assayswere also very similar to the Ki values determined in a competi-tive radioligand binding assay using [3H-UK14,304] (Kurko et al.,2009). This can be attributed to the fact that the binding assaydetected a G protein-coupled, high-affinity complex. In contrast tothe high affinity binding, EC50 values of �2C-AR agonists from the�-arrestin assay were largely in line with those reported for low-affinity binding sites or with those determined in the presence of Gprotein complexation to avoid the formation of a receptor ternarycomplex (Pohjanoksa et al., 1997; Jasper et al., 1998; Umlandet al., 2001; Audinot et al., 2002) or obtained from GTP�S studies(Jasper et al., 1998; Audinot et al., 2002). So far there have been noEC50 values published for �-arrestin recruitment of �2C-ARs that
could be compared to our results. We found that the �-arrestinassay was up to 1.6 log unit less sensitive than the cAMP or Ca2+assay, this shift being a general phenomenon for all agonists tested(i.e. observational bias). Similar findings were observed with the
arch B
�etbincrttotapsaTia
lbAitWpc11ltoiienstDoFioamp(arr(p(d(e(
Pcwtmic
D. Kurko et al. / Brain Rese
-opioid (McPherson et al., 2010), EDG1 and EDG2 (van Der Leet al., 2008; Yin et al., 2009), S1P1 (Yin et al., 2009), and S1P3 recep-ors (Riddy et al., 2012). This shift in potency and efficacy appears toe partly due to the greater receptor reserve of �2C-ARs expressed
n CHO cells used for the cAMP or Ca2+ mobilization assays (dataot shown). The �-arrestin assay operates at or near to a 1:1 stoi-hiometric ratio between the activation of the receptor and theecruitment of �-arrestin (Eglen et al., 2007; Bassoni et al., 2012),herefore arrestin binding may require a higher degree of recep-or activation than G protein-dependent signaling. For instance, inur �2C-C1 cell line 2.5% receptor occupancy by UK14,304 provedo be sufficient to elicit a half-maximal response for full receptorctivation measured in the cAMP assay (data not shown). It is alsoossible that the �-galactosidase detection system has lower sen-itivity. An alternative explanation for the detected shift in agonistctivity using different readouts is proposed by Riddy et al. (2012).hey assume that �-arrestin recruitment may alter S1P3 receptorsnto a conformation distinct from that measured in the Ca2+ releasessay (Riddy et al., 2012).
Because of the possible role of receptor internalization in theimitation of clinical effectiveness of several agonists, studies ofiased agonism often use receptor internalization as end-point.lthough G protein coupling of �2C-ARs has been extensively
nvestigated, there is less information regarding their desensitiza-ion, internalization or down-regulation due to agonist stimulation.
hile �2A-ARs and �2B-ARs have been shown to localize in thelasma membrane, a large amount of �2C-ARs appear to be in intra-ellular compartments in a variety of cell lines (von Zastrow et al.,993; Wozniak and Limbird, 1996; Daunt et al., 1997; DeGraff et al.,999; Olli-Lähdesmäki et al., 1999). The amount of intracellularly
ocalized receptors may depend on the cell type and the tempera-ure (Daunt et al., 1997; Jeyaraj et al., 2001). Therefore, detectionf agonist-induced internalization of the �2C-AR by conventionalmmunocytochemical techniques is quite complicated, often lead-ng to contradictory results (Daunt et al., 1997; Olli-Lähdesmäkit al., 1999). However, using multiple epitope tags – a tech-ique applied for studying a range of phenomena from prematureynapse formation (e.g. Zhang et al., 2013) to receptor internaliza-ion (Daunt et al., 1997) – and labeling of cell surface receptors,aunt et al. (1997) observed agonist-promoted internalizationf the plasma membrane-localized �2C-ARs in HEK-293 cells.urthermore, DeGraff et al. (1999) demonstrated that internal-zation of �2C-ARs was significantly promoted upon coexpressionf �-arrestin-2, suggesting that this subtype internalizes in anrrestin-dependent and probably a phosphorylation-independentanner. Besides the human �2C-AR displays no agonist-promoted
hosphorylation by G protein-coupled receptor kinases (GRKs)Eason and Liggett, 1992; Kurose and Lefkowitz, 1994; Jewell-Motznd Liggett, 1996), its heterodimerization with the �2A-AR subtypeesults in a decrease in the level of agonist-promoted �-arrestinecruitment and GRK2-mediated �2A-AR phosphorylation as wellSmall et al., 2006). However, it has also been shown that phos-horylation is not an absolute requirement for arrestin bindingRichardson et al., 2003; Jala et al., 2005). Even phosphorylation-eficient mutants still maintain the ability to interact with arrestinsMilasta et al., 2005; Stalheim et al., 2005) due to the pres-nce of negatively charged residues acting as phospho-mimeticsMukherjee et al., 2002; Galliera et al., 2004; Tobin, 2008).
Consistent with the results of DeGraff et al. (1999) we found thatTX completely abolished the cAMP response in �2C-AR expressingells, but neither the efficacy nor the potency of NE and UK14,304ere affected in the receptor internalization assay, indicating
hat this response occurred primarily through a G�i independentechanism. Although internalization of the �2C-AR has been stud-
ed by a number of groups, no EC50 values were published wean compare our results with. While efficacy values of receptor
ulletin 107 (2014) 89–101 97
internalization were virtually the same as for �-arrestin translo-cation, potency data differed up to 1 log unit in these assays.Many factors such as differences between experimental condi-tions cell type-specific regulatory proteins of CHO or U2OS cellsmay account for the observed differences. Since different cell typesexpress different levels of endogenous �-arrestin-1 and �-arrestin-2 (Santini et al., 2000; Oakley et al., 2012), and �-arrestin bindinghas been shown to be the most important rate-limiting step of�2C-AR internalization (DeGraff et al., 1999), the efficiency and theextent of receptor internalization may also depend on the cell lineemployed. In our studies all agonists tested induced �2C-AR inter-nalization in the U2OS cell line, but the extent and time-course ofendocytosis was ligand specific. Partial agonists induced less inter-nalization of �2C-ARs than full agonists, and imidazoline derivativesevoked internalization of �2C-ARs with significantly faster kinet-ics than NE. Although both of the full agonists UK14,304 and NEinduced the same degree of internalization after a 180 min treat-ment, NE had a significantly weaker internalization effect up to60 min than UK14,304. A similar pattern to that of UK14,304 wasobserved for the partial agonist imidazoline compounds (i.e. cloni-dine, moxonidine, and oxymetazoline) and also for phenylephrine.It is noteworthy that agonists displaying a rapid time-course of �2C-AR internalization were also more active for triggering responsesthrough the �-arrestin signaling pathway compared to NE. Similardifferential regulation in �-arrestin signaling leading to recep-tor desensitization and internalization have been implicated forin vivo responses of sensitivity, duration and tolerance of cloni-dine vs. guanfacine in �2A-AR expressing knock-in mice (Lu et al.,2009). Concerning antagonists, their potencies appeared to be moreconsistent across the different assay formats, being in agreementwith published data (Schaak et al., 1997; Uhlén et al., 1994, 1997;Lalchandani et al., 2002; Mustafa et al., 2005).
During the last decade, accumulating experimental data havedemonstrated that where multiple pathways are activated by thesame receptor, different agonists may cause different activation ofspecific signaling pathways (Kenakin, 2004; Galandrin and Bouvier,2006; Shonberg et al., 2014). Binding of these ligands to a receptormay either induce or stabilize different active conformations thatmay lead to specific signaling outcomes; a phenomenon referredas functional selectivity or biased agonism (Kenakin, 2005; Urbanet al., 2007). Biased agonism has been primarily reported as aphenomenon of synthetic ligands but in some cases even natu-ral ligands can display marked differences in their efficacies forG protein- or �-arrestin-mediated signaling or receptor inter-nalization (Gurwitz et al., 1994; Reiner et al., 2010; Rajagopalet al., 2013). Rajagopal and co-workers’ findings demonstrate thatbiased agonism of endogenous ligands acting at chemokine recep-tors is a common and likely evolutionarily conserved mechanism(Rajagopal et al., 2013).
However, most of the biased agonists described to date havebeen identified by comparing rank orders of potencies or effica-cies. This simple procedure can be useful for the identificationof extremely biased compounds but is also inherently prone toobservational bias. Recently several more sophisticated methodsfor quantifying ligand bias have been developed, some of themshowing the potential for not only the more precise characteri-zation of previously described biased ligands, but also identifyingnovel (and possibly weakly) biased ligands as well (Rajagopalet al., 2010, 2011; Shonberg et al., 2014). However, each of thedescribed methods has its own advantages and limitations. In theoperational model, bias factors can be calculated by either fittingconcentration–response data using dissociation constants from
independent binding experiments to obtain a coupling efficiency(Rajagopal et al., 2011) or fitting the data with more complicatedroutines to define a transduction ratio for each agonist and pathway(Evans et al., 2011; Kenakin et al., 2012). However, if the affinity of
9 arch B
atmCinsRtrseCeltasC
pfnR�anctceraieipatisictbcta1tSirhgmmr(eec�fiab
8 D. Kurko et al. / Brain Rese
n agonist depends on the signaling proteins coupled to the recep-or, the use of an independent single estimate of agonist affinity
ay introduce an error into the calculation of bias (Kenakin andhristopoulos, 2013a). On the other hand, if the data of full agonists
s fitted to the operational model to obtain a transduction ratio,on-physiological constraints have to be used such as arbitrarilyetting dissociation constant values to zero (Rajagopal et al., 2011;ajagopal, 2013). Another parameter that has been used to quan-ify biased agonism is the intrinsic relative activity (RA) which is aelative measure of the affinity constant of an agonist for the activetate of the receptor relative to that of a standard agonist (Ehlertt al., 1999; Ehlert, 2005, 2008; Figueroa et al., 2009; Kenakin andhristopoulos, 2013). RA values are calculated from the maximalffect produced by a ligand (Emax) and the concentration of theigand that produces half-maximal response (EC50), and are agreedo yield similar results with the above-mentioned approaches forgonists that produce concentration–response relationships withlopes that are not significantly different from unity (Kenakin andhristopoulos, 2013; Shonberg et al., 2014).
Since differences and even reversals in the rank order ofotencies and efficacies between signaling pathways suggestedunctional selectivity for some of the compounds tested, and ago-ists typically exhibited Hill slopes close to one, we applied theA method for quantifying the level of biased signaling at the2C-AR. This approach allowed us to perform a comprehensivenalysis of ligand bias and to identify several weakly biased ago-ists towards �-arrestin recruitment or receptor internalization. Inontrast, we found no evidence for functional selectivity betweenhe cAMP and Ca2+ pathways. The RA analysis indicated thatonformations associated with the agonists tested were coupledqually well to pathways mediating cAMP accumulation and Ca2+
elease, supporting our previous results which showed that �2C-ARctivation via G�qi5 was an appropriate substitute for estimat-ng agonist activity at the �2C-AR signaling through G�i (Kurkot al., 2009). Regarding �-arrestin signaling or receptor internal-zation, none of the agonists tested exhibited perfectly biasedroperties. Nevertheless, UK14,304, moxonidine, clonidine, guan-benz and phenylephrine displayed a substantially higher capacityhan NE to recruit �-arrestin than to activate Ca2+ release or tonhibit cAMP accumulation. Furthermore, all the agonists showedignificant bias toward �-arrestin translocation versus receptornternalization. Interestingly, high bias factors were observed forlonidine towards the Ca2+ and cAMP pathways compared to recep-or internalization. In contrast, phenylephrine showed significantias toward receptor internalization compared to Ca2+ release orAMP accumulation. Our data showed that agonists with only sub-le differences in structure can still display functional selectivity,s seen for NE and phenylephrine. Phenylephrine was 11.3 and6.2 times more active than NE to induce �-arrestin recruitmenthan to inhibit cAMP accumulation or to stimulate Ca2+ release.imilarly, it was 3 and 4.5 times more active at inducing receptornternalization compared to cAMP accumulation and Ca2+ release,espectively. It is reasonable to suppose that the lack of the para-ydroxyl group on the catechol ring and the presence of the methylroup in the structure of phenylephrine are able to affect confor-ational changes in the cytoplasmic domains of the receptor and toodulate the coupling to effector proteins. Similar findings were
eported for other GPCRs, e.g. HT2C, D2, �1A, mGlu1, 5 receptorsMiller et al., 2000; Gay et al., 2004; Evans et al., 2011; Emeryt al., 2012). Evans et al. (2011) found that phenylephrine but notpinephrine showed substantial bias toward extracellular acidifi-ation rate compared with two G-protein-mediated pathways in
1A adrenoreceptor expressing cells. Emery et al. (2012) identi-ed that glutaric acid and succinic acid, analogs of glutamate andspartate, respectively, with a deleted �-amino group were fullyiased toward �-arrestin mediated signaling of mGlu1 receptors.ulletin 107 (2014) 89–101
Similarly, relatively small structural modification of both 5-HT2Cand D2 ligands elicited more than 100-fold functional selectiv-ity without affecting binding affinity (Miller et al., 2000; Gayet al., 2004). These observations highlight how subtle differences inchemical structure of various ligands can ultimately affect down-stream signaling responses through conformational changes.
Although there are no published data for quantifying ligand biasat the �2C-AR, similar observations were published in previousreports with recombinant cell lines expressing different subtypesof �-adrenergic receptors (Kukkonen et al., 2001; Evans et al.,2011). Kukkonen et al. (2001) found differences in the abilityof chemically different �2-adrenoceptor agonists to activate twosignaling pathways in �2A-AR expressing HEL 92.1.7 cells. Theendogenous ligands, NE and epinephrine, were several time lesspotent in decreasing cAMP levels than in elevating Ca2+ concentra-tion, whereas synthetic imidazoline and ox/thiazoloazepine ligandswere more potent in decreasing cAMP levels. Evans et al. (2011)tested the capacity of agonists to stimulate Ca2+ release, cAMPaccumulation, and changes in extracellular acidification rate at thehuman �1A-adrenoceptor. Quantifying functional selectivity by theapplication of the operational model they showed significant biaseffects for several agonists through the three signaling pathways.
Over the last decade it has become clear that developmentof ligands with pathway-selective activities may hold the cluefor the development of therapeutically more advantageous drugs(Kim et al., 2008; Violin et al., 2010; Bohinc and Gesty-Palmer,2011; Kammermann et al., 2011; DeWire et al., 2013). However,the major bottleneck in determining the therapeutic potential ofG-protein and �-arrestin-biased ligands may be the lack of knowl-edge which signaling pathway is the therapeutically relevant one.Testing in animal models could help to delineate the physiologicconsequences of different signaling mechanism. For instance, ithas been reported that �-arrestin KO mice are resistant to �2-AR-stimulated sedation (Wang et al., 2004). On the contrary, �-arrestinmediated �2A-AR down-regulation without activation of G proteinsaccounts for the therapeutically desirable antidepressant effect ofdesipramine (Cottingham et al., 2011).
Guided by drug discovery efforts made in the field of biased ago-nism, in this work we have demonstrated that biased ligands actingat �2C-ARs could be identified by exploring receptor activationusing different signaling readouts and by applying an appropri-ate analytical method to quantify the degree of biased signaling.These new observed differences in signal transduction broaden ourunderstanding on �2C-AR signaling at all these stages of receptoractivation. This may help link specific signaling properties to giventherapeutic activities or unwanted effects, having the potential topredict better activity and fewer side effects when designing drugs.
Conflict of interest statement
The authors declare that there are no conflicts of interest.
Acknowledgments
We thank for the excellent technical assistance of Mrs. E. Széll,Mrs. P. Unghy and Mrs. A. Garai. We are grateful to I. Tarnawafor invaluable help in discussion and for critically reading themanuscript.
References
Abdulla, F.A., Smith, P.A., 1997. Ectopic alpha2-adrenoceptors couple to N-type Ca2+
channels in axotomized rat sensory neurons. J. Neurosci. 17, 1633–1641.Alblas, J., van Corven, E.J., Hordijk, P.L., Milligan, G., Moolenaar, W.H., 1993. Gi-
mediated activation of the p21ras-mitogen-activated protein kinase pathwayby alpha 2-adrenergic receptors expressed in fibroblasts. J. Biol. Chem. 268 (30),22235–22238.
arch B
A
AA
B
B
B
B
B
B
B
C
C
C
C
D
D
D
D
D
D
D
E
E
E
E
E
E
E
E
E
D. Kurko et al. / Brain Rese
rcangeli, A., D’Alo, C., Gaspari, R., 2009. Dexmedetomidine use in general anaes-thesia. Curr. Drug Targets 10, 687–695.
rnsten, A.F., 1998. The biology of being frazzled. Science 280, 1711–1712.udinot, V., Fabry, N., Nicolas, J.P., Beauverger, P., Newman-Tancredi, A., Millan,
M.J., Try, A., Bornancin, F., Canet, E., Boutin, J.A., 2002. Ligand modulation of[35S]GTP�S binding at human alpha(2A), alpha(2B) and alpha(2C) adrenocep-tors. Cell. Signal. 14 (10), 829–837.
assoni, D.L., Raab, W.J., Achacoso, P.L., Loh, C.Y., Wehrman, T.S., 2012. Measure-ments of �-arrestin recruitment to activated seven transmembrane receptorsusing enzyme complementation. Methods Mol. Biol. 897, 181–203.
erg, K.A., Clarke, W.P., 2006. Development of functionally selective agonists as noveltherapeutic agents. Drug Discov Today Ther. Strateg. 3 (4), 421–428.
erg, K.A., Maayani, S., Goldfarb, J., Scaramellini, C., Leff, P., Clarke, W.P., 1998. Effec-tor pathway-dependent relative efficacy at serotonin type 2A and 2C receptors:evidence for agonist-directed trafficking of receptor stimulus. Mol. Pharmacol.54 (1), 94–104.
lack, J.W., Leff, P., 1983. Operational models of pharmacological agonism. Proc. R.Soc. Lond. B: Biol. Sci. 220, 141–162.
ohinc, B.N., Gesty-Palmer, D., 2011. �-Arrestin-biased agonism at the parathy-roid hormone receptor uncouples bone formation from bone resorption. Endocr.Metab. Immune Disord. Drug Targets 11 (2), 112–119.
ylund, D.B., 1992. Subtypes of �1- and �2-adrenergic receptors. FASEB J. 6 (3),832–839.
ylund, D.B., 2005. Alpha-2 adrenoceptor subtypes: are more better? Br. J. Pharma-col. 144 (2), 159–160.
alzada, B.C., Artinano, A.A., 2001. Alpha-adrenoceptor subtypes. Pharm. Res. 44 (3),195–208.
innamon Bidwell, L., Dew, R.E., Kollins, S.H., 2010. Alpha-2 adrenergic receptors andattention-deficit/hyperactivity disorder. Curr. Psychiatry Rep. 12 (5), 366–373.
ottingham, C., Chen, Y., Jiao, K., Wang, Q., 2011. The antidepressant desipramine isan arrestin-biased ligand at the �2A-adrenergic receptor driving receptor down-regulation in vitro and in vivo. J. Biol. Chem. 286 (41), 36063–36075.
rassous, P.A., Denis, C., Paris, H., Sénard, J.M., 2007. Interest of alpha2-adrenergicagonists and antagonists in clinical practice: background, facts and perspectives.Curr. Top. Med. Chem. 7 (2), 187–194.
aunt, D.A., Hurt, C., Hein, L., Kallio, J., Feng, F., Kobilka, B.K., 1997. Subtype-specificintracellular trafficking of alpha2-adrenergic receptors. Mol. Pharmacol. 51 (5),711–720.
eFea, K.A., Vaughn, Z.D., O’Bryan, E.M., Nishijima, D., Déry, O., Bunnett, N.W., 2000a.The proliferative and antiapoptotic effects of substance P are facilitated by for-mation of a �-arrestin-dependent scaffolding complex. Proc. Natl. Acad. Sci.U.S.A. 97 (20), 11086–11091.
eFea, K.A., Zalevsky, J., Thoma, M.S., Déry, O., Mullins, R.D., Bunnett, N.W., 2000b. �-Arrestin-dependent endocytosis of proteinase-activated receptor 2 is requiredfor intracellular targeting of activated ERK1/2. J. Cell Biol. 148, 1267–1281.
eGraff, J.L., Gagnon, A.W., Benovic, J.L., Orsini, M.J., 1999. Role of arrestins in endo-cytosis and signaling of alpha2-adrenergic receptor subtypes. J. Biol. Chem. 274(16), 11253–11259.
eWire, S.M., et al., 2013. A G protein-biased ligand at the �-opioid receptor ispotently analgesic with reduced gastrointestinal and respiratory dysfunctioncompared with morphine. J. Pharmacol. Exp. Ther. 344 (3), 708–717.
ocherty, J.R., 1998. Subtypes of functional �1- and �2-adrenoceptors. Eur. J.Pharmacol. 361, 1–15.
orn II, G.W., Oswald, K.J., McCluskey, T.S., Kuhel, D.G., Liggett, S.B., 1997. Alpha2A-adrenergic receptor stimulated calcium release is transduced by Gi-associatedG��-mediated activation of phospholipase C. Biochemistry 36 (21), 6415–6423.
ason, M.G., Liggett, S.B., 1992. Subtype-selective desensitization of alpha2-adrenergic receptors: different mechanisms control short and long-termagonist-promoted desensitization of alpha2C10, alpha2C4 and alpha2C2. J. Biol.Chem. 267, 25473–25479.
ason, M.G., Kurose, H., Holt, B.D., Raymond, J.R., Liggett, S.B., 1992. Simultaneouscoupling of �2-adrenergic receptors to two G-proteins with opposing effects.Subtype-selective coupling of �2C10, �2C4, and �2C2 adrenergic receptors to Giand Gs. J. Biol. Chem. 267, 15795–15801.
ason, M.G., Jacinto, M.T., Liggett, S.B., 1994. Contribution of ligand structure to acti-vation of �2-adrenergic receptor subtype coupling to Gs. Mol. Pharmacol. 45 (4),696–702.
glen, R.M., Bosse, R., Reisine, T., 2007. Emerging concepts of guanine nucleotide-binding protein-coupled receptor (GPCR) function and implications for highthroughput screening. Assay Drug Dev. Technol. 5 (3), 425–451.
hlert, F.J., 2005. Analysis of allosterism in functional assays. J. Pharmacol. Exp. Ther.315, 740–754.
hlert, F.J., 2008. On the analysis of ligand-directed signaling at G protein-coupledreceptors. Naunyn Schmiedebergs Arch. Pharmacol. 377 (4–6), 549–577.
hlert, F.J., Griffin, M.T., Sawyer, G.W., Bailon, R., 1999. A simple method for estima-tion of agonist activity at receptor subtypes: comparison of native and clonedM3 muscarinic receptors in guinea pig ileum and transfected cells. J. Pharmacol.Exp. Ther. 289, 981–992.
mery, A.C., DiRaddo, J.O., Miller, E., Hathaway, H.A., Pshenichkin, S., Takoudjou, G.R.,Grajkowska, E., Yasuda, R.P., Wolfe, B.B., Wroblewski, J.T., 2012. Ligand bias atmetabotropic glutamate 1a receptors: molecular determinants that distinguish
�-arrestin-mediated from G protein-mediated signaling. Mol. Pharmacol. 82 (2),291–301.vans, B.A., Broxton, N., Merlin, J., Sato, M., Hutchinson, D.S., Christopoulos, A.,Summers, R.J., 2011. Quantification of functional selectivity at the human �(1A)-adrenoceptor. Mol. Pharmacol. 79 (2), 298–307.
ulletin 107 (2014) 89–101 99
Evans, B.A., Sato, M., Sarwar, M., Hutchinson, D.S., Summers, R.J., 2010. Liganddirected signalling at �1-adrenoceptors. Br. J. Pharmacol. 159, 1022–1038.
Figueroa, K.W., Griffin, M.T., Ehlert, F.J., 2009. Selectivity of agonists for the activestate of M1 to M4 muscarinic receptor subtypes. J. Pharmacol. Exp. Ther. 328,331–342.
Flordellis, C.S., Berguerand, M., Gouache, P., Barbu, V., Gavras, H., Handy, D.E., Berezia,G., Masliah, J., 1995. Alpha2 adrenergic receptor subtypes expressed in Chinesehamster ovary cells activate differentially mitogen-activated protein kinase bya p21ras independent pathway. J. Biol. Chem. 270 (8), 3491–3494.
Galandrin, S., Bouvier, M., 2006. Distinct signaling profiles of beta1 and beta2adrenergic receptor ligands toward adenylyl cyclase and mitogen-activated pro-tein kinase reveals the pluridimensionality of efficacy. Mol. Pharmacol. 70 (5),1575–1584.
Galandrin, S., Oligny-Longpré, G., Bouvier, M., 2007. The evasive nature of drugefficacy: implications for drug discovery. Trends Pharmacol. Sci. 28 (8), 423–430.
Galliera, E., Jala, V.R., Trent, J.O., Bonecchi, R., Signorelli, P., Lefkowitz, R.J., Man-tovani, A., Locati, M., Haribabu, B., 2004. Beta-arrestin-dependent constitutiveinternalization of the human chemokine decoy receptor D6. J. Biol. Chem. 279,25590–25597.
Gay, E.A., Urban, J.D., Nichols, D.E., Oxford, G.S., Mailman, R.B., 2004. Functional selec-tivity of D2 receptor ligands in a Chinese hamster ovary hD2L cell line: evidencefor induction of ligand-specific receptor states. Mol. Pharmacol. 66 (1), 97–105.
Gesty-Palmer, D., Flannery, P., Yuan, L., et al., 2009. A �-arrestin-biased agonist of theparathyroid hormone receptor (PTH1R) promotes bone formation independentof G protein activation. Sci. Transl. Med. 1 (1), 1ra1.
Ghanouni, P., Gryczynski, Z., Steenhuis, J.J., Lee, T.W., Farrens, D.L., Lakowicz, J.R.,Kobilka, B.K., 2001. Functionally different agonists induce distinct conforma-tions in the G protein coupling domain of the �2 adrenergic receptor. J. Biol.Chem. 276, 24433–24436.
Gurwitz, D., Haring, R., Heldman, E., Fraser, C.M., Manor, D., Fisher, A., 1994. Discreteactivation of transduction pathways associated with acetylcholine m1 receptorby several muscarinic ligands. Eur. J. Pharmacol. 267, 21–31.
Hanson, B.J., Wetter, J., Bercher, M.R., Kopp, L., Fuerstenau-Sharp, M., Vedvik, K.L.,Zielinski, T., Doucette, C., Whitney, P.J., Revankar, C., 2009. A homogeneousfluorescent live-cell assay for measuring 7-transmembrane receptor activityand agonist functional selectivity through beta-arrestin recruitment. J. Biomol.Screen. 14 (7), 798–810.
Hieble, J.P., Bondinell, W.E., Ruffolo Jr., R.R., 1995. Alpha- and beta-adrenoceptors:from the gene to the clinic. 1. Molecular biology and adrenoceptor subclassifi-cation. J. Med. Chem. 38 (18), 3415–3444.
Jala, V.R., Shao, W.H., Haribabu, B., 2005. Phosphorylation-independent beta-arrestintranslocation and internalization of leukotriene B4 receptors. J. Biol. Chem. 280,4880–4887.
Jasper, J.R., Lesnick, J.D., Chang, L.K., Yamanishi, S.S., Chang, T.K., Hsu, S.A., Daunt,D.A., Bonhaus, D.W., Eglen, R.M., 1998. Ligand efficacy and potency at recombi-nant �2-adrenergic receptors: agonist-mediated [35S]GTP�S binding. Biochem.Pharmacol. 55 (7), 1035–1043.
Jeon, Y.T., Luoa, C., Forrayb, C., Vaysseb, P.J.J., Branchekb, T.A., Gluchowskia, C., 1995.Pharmacological evaluation of UK-14,304 analogs at cloned human � adrenergicreceptors. Bioorg. Med. Chem. Lett. 5 (19), 2255–2258.
Jewell-Motz, E.A., Liggett, S.B., 1996. G protein-coupled receptor kinase specificityfor phosphorylation and desensitization of alpha2-adrenergic receptor sub-types. J. Biol. Chem. 271 (30), 18082–18087.
Jeyaraj, S.C., Chotani, M.A., Mitra, S., Gregg, H.E., Flavahan, N.A., Morrison, K.J., 2001.Cooling evokes redistribution of alpha2C-adrenoceptors from Golgi to plasmamembrane in transfected human embryonic kidney 293 cells. Mol. Pharmacol.60 (6), 1195–1200.
Jones, S.B., Halenda, S.P., Bylund, D.B., 1991. Alpha 2-adrenergic receptor stimu-lation of phospholipase A2 and of adenylate cyclase in transfected Chinesehamster ovary cells is mediated by different mechanisms. Mol. Pharmacol. 39(2), 239–245.
Kable, J.W., Murrin, L.C., Bylund, D.B., 2000. In vivo gene modification elucidatessubtype-specific functions of alpha(2)-adrenergic receptors. J. Pharmacol. Exp.Ther. 293 (1), 1–7.
Kamibayashi, T., Maze, M., 2000. Clinical uses of alpha2-adrenergic agonists. Anes-thesiology 93 (5), 1345–1349.
Kammermann, M., Denelavas, A., Imbach, A., Grether, U., Dehmlow, H., Apfel,C.M., Hertel, C., 2011. Impedance measurement: a new method to detectligand-biased receptor signaling. Biochem. Biophys. Res. Commun. 412 (3),419–424.
Kenakin, T., 1995. Pharmacological proteus? Trends Pharmacol. Sci. 16, 256–258.Kenakin, T., 2002. Drug efficacy at G protein-coupled receptors. Annu. Rev. Pharma-
col. Toxicol. 42, 349–379.Kenakin, T., 2004. Principles: receptor theory in pharmacology. Trends Pharmacol.
Sci. 25, 186–192.Kenakin, T., 2005. New concepts in drug discovery: collateral efficacy and permissive
antagonism. Nat. Rev. Drug Discov. 4 (11), 919–927.Kenakin, T., 2007. Functional selectivity through protean and biased agonism: who
steers the ship? Mol. Pharmacol. 72, 1393–1401.Kenakin, T., 2011. Functional selectivity and biased receptor signaling. J. Pharmacol.
Exp. Ther. 336 (2), 296–302.
Kenakin, T., Miller, L.J., 2010. Seven transmembrane receptors as shapeshifting pro-teins: the impact of allosteric modulation and functional selectivity on new drugdiscovery. Pharmacol. Rev. 62, 265–304.
Kenakin, T., Christopoulos, A., 2013. Signalling bias in new drug discovery: detection,quantification and therapeutic impact. Nat. Rev. Drug Discov. 12 (3), 205–216.
1 arch B
K
K
K
K
K
K
K
K
L
L
L
L
L
L
L
M
M
M
M
M
M
M
N
O
O
P
P
00 D. Kurko et al. / Brain Rese
enakin, T., Christopoulos, A., 2013a. Measurements of ligand bias and functionalaffinity. Nat. Rev. Drug Discov. 12 (6), 483.
enakin, T., Watson, C., Muniz-Medina, V., Christopoulos, A., Novick, S., 2012. A sim-ple method for quantifying functional selectivity and agonist bias. ACS Chem.Neurosci. 3 (3), 193–203.
im, I.M., Tilley, D.G., Chen, J., Salazar, N.C., Whalen, E.J., Violin, J.D., Rockman, H.A.,2008. �-Blockers alprenolol and carvedilol stimulate �-arrestin mediated EGFRtransactivation. Proc. Natl. Acad. Sci. U.S.A. 105 (38), 14555–14560.
naus, A.E., Muthig, V., Schickinger, S., Moura, E., Beetz, N., Gilsbach, R., Hein, L., 2007.�2-Adrenoceptor subtypes – unexpected functions for receptors and ligandsderived from gene-targeted mouse models. Neurochem. Int. 51 (5), 277–281.
ukkonen, J.P., Jansson, C.C., Akerman, K.E., 2001. Agonist trafficking of G(i/o)-mediated alpha(2A)-adrenoceptor responses in HEL 92.1.7 cells. Br. J. Pharmacol.132 (7), 1477–1484.
ukkonen, J.P., Renvaktar, A., Shariatmadari, R., Akerman, K.E., 1998. Ligand- andsubtype-selective coupling of human alpha-2 adrenoceptors to Ca++ elevationin Chinese hamster ovary cells. J. Pharmacol. Exp. Ther. 287 (2), 667–671.
urko, D., Bekes, Z., Gere, A., Baki, A., Boros, A., Kolok, S., Bugovics, G., Nagy, J., Szom-bathelyi, Z., Ignacz-Szendrei, G., 2009. Comparative pharmacology of adrenergic�2C receptors coupled to Ca2+ signaling through different G� proteins. Neu-rochem. Int. 55 (7), 467–475.
urose, H., Lefkowitz, R.J., 1994. Differential desensitization and phosphorylation ofthree cloned and transfected alpha2-adrenergic receptor subtypes. J. Biol. Chem.269, 10093–10099.
akhlani, P.P., MacMillan, L.B., Guo, T.Z., McCool, B.A., Lovinger, D.M., Maze, M., Lim-bird, L.E., 1997. Substitution of a mutant alpha (2a)-adrenergic receptor viahit and run gene targeting reveals the role of this subtype in sedative, anal-gesic, and anesthetic sparing responses in vivo. Proc. Natl. Acad. Sci. U.S.A. 94,9950–9955.
alchandani, S.G., Lei, L., Zheng, W., Suni, M.M., Moore, B.M., Liggett, S.B., Miller, D.D.,Feller, D.R., 2002. Yohimbine dimers exhibiting selectivity for the human alpha2C-adrenoceptor subtype. J. Pharmacol. Exp. Ther. 303 (3), 979–984.
imbird, L.E., 1988. Receptors linked to inhibition of adenylate cyclase: additionalsignaling mechanisms. FASEB J. 2 (11), 2686–2695.
u, R., et al., 2009. Epitope-tagged receptor knock-in mice reveal that differentialdesensitization of alpha2-adrenergic responses is because of ligand-selectiveinternalization. J. Biol. Chem. 284 (19), 13233–13243.
uttrell, L.M., Ferguson, S.S., Daaka, Y., Miller, W.E., Maudsley, S., Della Rocca, G.J.,Lin, F., Kawakatsu, H., Owada, K., Luttrell, D.K., Caron, M.G., Lefkowitz, R.J., 1999.�-Arrestin-dependent formation of �2 adrenergic receptor–Src protein kinasecomplexes. Science 283, 655–661.
uttrell, L.M., Gesty-Palmer, D., 2010. Beyond desensitization: physiological rele-vance of arrestin-dependent signaling. Pharmacol. Rev. 62 (2), 305–330.
uttrell, L.M., Roudabush, F.L., Choy, E.W., Miller, W.E., Field, M.E., Pierce, K.L.,Lefkowitz, R.J., 2001. Activation and targeting of extracellular signal-regulatedkinases by �-arrestin scaffolds. Proc. Natl. Acad. Sci. U.S.A. 98 (5), 2449–2454.
acNulty, E.E., McClue, S.J., Carr, I.C., Jess, T., Wakelam, M.J., Milligan, G., 1992.Alpha 2-C10 adrenergic receptors expressed in rat 1 fibroblasts can reg-ulate both adenylyl cyclase and phospholipase D-mediated hydrolysis ofphosphatidylcholine by interacting with pertussis toxin-sensitive guaninenucleotide-binding proteins. J. Biol. Chem. 267 (4), 2149–2156.
aze, M., Fujinaga, M., 2000. Alpha(2) adrenoceptors in pain modulation – whichsubtype should be targeted to produce analgesia? Anesthesiology 92, 934–936.
cPherson, et al., 2010. �-Opioid receptors: correlation of agonist efficacy forsignalling with ability to activate internalization. Mol. Pharmacol. 78 (4),756–766.
ilasta, S., Evans, N.A., Ormiston, L., Wilson, S., Lefkowitz, R.J., Milligan, G., 2005. Thesustainability of interactions between the orexin-1 receptor and beta-arrestin-2is defined by a single C-terminal cluster of hydroxy amino acids and modulatesthe kinetics of ERK MAPK regulation. Biochem. J. 387 (3), 573–584.
iller, K.J., Robichaud, A.J., Largent, B.L., 2000. 5-HT2C receptor-mediated activationof IP3 and cGMP in HEK293 cells: evidence for agonist directed trafficking. Soc.Neurosci. Abstr., Abstract 26(1–2), Number 46.6.
ukherjee, S., Gurevich, V.V., Preninger, A., Hamm, H.E., Bader, M.F., Fazleabas, A.T.,Birnbaumer, L., Hunzicker-Dunn, M., 2002. Aspartic acid 564 in the third cyto-plasmic loop of the luteinizing hormone/choriogonadotropin receptor is crucialfor phosphorylation-independent interaction with arrestin2. J. Biol. Chem. 277,17916–17927.
ustafa, S.M., Bavadekar, S.A., Ma, G., Moore, B.M., Fellerb, D.R., Millera, D.D.,2005. Synthesis and biological studies of yohimbine derivatives on human �2C-adrenergic receptors. Bioorg. Med. Chem. Lett. 15, 2758–2760.
oma, T., et al., 2007. �-Arrestin-mediated �1-adrenergic receptor transactivationof the EGRF confers cardioprotection. J. Clin. Invest. 117 (9), 2445–2458.
akley, R.H., Revollo, J., Cidlowski, J.A., 2012. Glucocorticoids regulate arrestin geneexpression and redirect the signaling profile of G protein-coupled receptors.Proc. Natl. Acad. Sci. U.S.A. 109 (43), 17591–17596.
lli-Lähdesmäki, T., Kallio, J., Scheinin, M., 1999. Receptor subtype-induced targetingand subtype-specific internalization of human alpha(2)-adrenoceptors in PC12cells. J. Neurosci. 19 (21), 9281–9288.
arsley, S., Gazi, L., Bobirnac, I., Loetscher, E., Schoeffter, P., 1999. Functional �2C-adrenoceptors in human neuroblastoma SH-SY5Y cells. Eur. J. Pharmacol. 372
(1), 109–115.auwels, P.J., Colpaert, F.C., 2000. Disparate ligand-mediated Ca2+ responses by wild-type, mutant Ser200Ala and Ser204Ala a2A-adrenoceptor: G�15 fusion proteins:evidence for multiple ligand-activation binding sites. Br. J. Pharmacol. 130 (7),1505–1512.
ulletin 107 (2014) 89–101
Philipp, M., Brede, M., Hein, L., 2002. Physiological significance of �2-adrenergicreceptor subtype diversity: one receptor is not enough. Am. J. Physiol. Regul.Integr. Comp. Physiol. 283, 287–295.
Pohjanoksa, K., Jansson, C.C., Luomala, K., Marjamaki, A., Savola, J.M., Scheinin, M.,1997. �2-Adrenoceptor regulation of adenylyl cyclase in CHO cells: dependenceon receptor density, receptor subtype and current activity of adenylyl cyclase.Eur. J. Pharmacol. 335 (1), 53–63.
Rajagopal, S., 2013. Quantifying biased agonism: understanding the links betweenaffinity and efficacy. Nat. Rev. Drug Discov. 12 (6), 483.
Rajagopal, S., Ahn, S., Rominger, D.H., Gowen-MacDonald, W., Lam, C.M., DeWire,S.M., Violin, J.D., Lefkowitz, R.J., 2011. Quantifying ligand bias at seven-transmembrane receptors. Mol. Pharmacol. 80 (3), 367–377.
Rajagopal, S., Bassoni, D.L., Campbell, J.J., Gerard, N.P., Gerard, C., Wehrman, T.S.,2013. Biased agonism as a mechanism for differential signaling by chemokinereceptors. J. Biol. Chem. 288 (49), 35039–35048.
Rajagopal, S., Rajagopal, K., Lefkowitz, R.J., 2010. Teaching old receptors new tricks:biasing seven-transmembrane receptors. Nat. Rev. Drug Discov. 9, 373–386.
Rajagopal, S., Whalen, E.J., Violin, J.D., Stiber, J.A., Rosenberg, P.B., Premont, R.T., Coff-man, T.M., Rockman, H.A., Lefkowitz, R.J., 2006. �-Arrestin-2 mediated inotropiceffects of the angiotensin II type 1A receptor in isolated cardiac myocytes. Proc.Natl. Acad. Sci. U.S.A. 103, 16284–16289.
Reiner, S., Ambrosio, M., Hoffmann, C., Lohse, M.J., 2010. Differential signaling of theendogenous agonists at the beta2-adrenergic receptor. J. Biol. Chem. 285 (46),36188–36198.
Reynen, P.H., Martin, G.R., Eglen, R.M., MacLennan, S.J., 2000. Characterization ofhuman recombinant alpha(2A)-adrenoceptors expressed in Chinese hamsterlung cells using intracellular Ca(2+) changes: evidence for cross-talk betweenrecombinant alpha(2A)- and native alpha(1)-adrenoceptors. Br. J. Pharmacol.129 (7), 1339–1346.
Richardson, M.D., Balius, A.M., Yamaguchi, K., Freilich, E.R., Barak, L.S., Kwatra, M.M.,2003. Human substance P receptor lacking the C-terminal domain remains com-petent to desensitize and internalize. J. Neurochem. 84 (4), 854–863.
Riddy, D.M., Stamp, C., Sykes, D.A., Charlton, S.J., Dowling, M.R., 2012. Reassessmentof the pharmacology of Sphingosine-1-phosphate S1P3 receptor ligands usingthe DiscoveRx PathHunterTM and Ca2+ release functional assays. Br. J. Pharmacol.167 (4), 868–880.
Ruffolo Jr., R.R., Bondinell, W., Hieble, J.P., 1995. Alpha- and beta-adrenoceptors:from the gene to the clinic. 2. Structure–activity relationships and therapeuticapplications. J. Med. Chem. 38 (19), 3681–3716.
Ruffolo Jr., R.R., Hieble, J.P., 1994. �-Adrenoceptors. Pharmacol. Ther. 61 (1–2), 1–64.Ruffolo Jr., R.R., Nichols, A.J., Stadel, J.M., Hieble, J.P., 1993. Pharmacologic and ther-
apeutic applications of alpha 2-adrenoceptor subtypes. Annu. Rev. Pharmacol.Toxicol. 33, 243–279.
Sanders, R.D., Maze, M., 2007. �2-Adrenoceptor agonists. Curr. Opin. Investig. Drugs8 (1), 25–33.
Santini, F., Penn, R.B., Gagnon, A.W., Benovic, J.L., Keen, J.H., 2000. Selective recruit-ment of arrestin-3 to clathrin coated pits upon stimulation of G protein-coupledreceptors. J. Cell Sci. 113, 2463–2470.
Saunders, C., Limbird, L.E., 1999. Localization and trafficking of �2-adrenergic recep-tor subtypes in cells and tissues. Pharmacol. Ther. 84 (2), 193–205.
Schaak, S., Cayla, C., Blaise, R., Quinchon, F., Paris, H., 1997. HepG2 and SK-N-MC: twohuman models to study alpha-2 adrenergic receptors of the alpha-2C subtype.J. Pharmacol. Exp. Ther. 281 (2), 983–991.
Scholz, J., Tonner, P.H., 2000. �2-Adrenoceptor agonists in anaesthesia: a newparadigm. Curr. Opin. Anaesthesiol. 13, 437–442.
Serle, J.B., Lustgarten, J.S., Podos, S.M., 1991. A clinical trial of metipranodol, anoncardioselective beta-adrenergic antagonist, in ocular hypertension. Am. J.Ophthalmol. 112, 302–307.
Shonberg, J., Herenbrink, C.K., López, L., Christopoulos, A., Scammells, P.J., Capuano,B., Lane, J.R., 2013. A structure–activity analysis of biased agonism at thedopamine D2 receptor. J. Med. Chem. 56 (22), 9199–9221.
Shonberg, J., Lopez, L., Scammells, P.J., Christopoulos, A., Capuano, B., Lane,J.R., 2014. Biased agonism at G protein-coupled receptors: the promiseand the challenges – a medicinal chemistry perspective. Med. Res. Rev.,http://dx.doi.org/10.1002/med.21318.
Small, K.M., Schwarb, M.R., Glinka, C., Theiss, C.T., Brown, K.M., Seman,C.A., Liggett, S.B., 2006. Alpha2A- and alpha2C-adrenergic receptors formhomo- and heterodimers: the heterodimeric state impairs agonist-promotedGRK phosphorylation and beta-arrestin recruitment. Biochemistry 45 (15),4760–4767.
Stalheim, L., Ding, Y., Gullapalli, A., Paing, M.M., Wolfe, B.L., Morris, D.R., Trejo, J.,2005. Multiple independent functions of arrestins in the regulation of protease-activated receptor-2 signaling and trafficking. Mol. Pharmacol. 67, 78–87.
Stamer, W.D., Huang, Y., Seftor, R.E., Svensson, S.S., Snyder, R.W., Regan, J.W., 1996.Cultured human trabecular meshwork cells express functional alpha 2A adren-ergic receptors. Invest. Ophthalmol. Vis. Sci. 34, 2803–2812.
Taylor, S., Gray, J.R., Willis, R., Deeks, N., Haynes, A., Campbell, C., Gaskin, P., Leavens,K., Demont, E., Dowell, S., Cryan, J., Morse, M., Patel, A., Garden, H., Wither-ington, J., 2012. The utility of pharmacokinetic-pharmacodynamic modeling inthe discovery and optimization of selective S1P(1) agonists. Xenobiotica 42 (7),671–686.
Tobin, A.B., 2008. G-protein-coupled receptor phosphorylation: where, when andby whom. Br. J. Pharmacol. 153 (Suppl. 1), S167–S176.
Uhlén, S., Lindblom, J., Johnson, A., Wikberg, J.E., 1997. Autoradiographic studies ofcentral alpha 2A- and alpha 2C-adrenoceptors in the rat using [3H]MK912 andsubtype-selective drugs. Brain Res. 770 (1–2), 261–266.
arch B
U
U
U
v
V
V
V
v
W
12338.
D. Kurko et al. / Brain Rese
hlén, S., Porter, A.C., Neubig, R.R., 1994. The novel alpha-2 adrenergic radioligand[3H]-MK912 is alpha-2C selective among human alpha-2A, alpha-2B and alpha-2C adrenoceptors. J. Pharmacol. Exp. Ther. 271 (3), 1558–1565.
mland, S.P., Wan, Y., Shah, H., Billah, M., Egan, R.W., Hey, J.A., 2001. Receptor reserveanalysis of the human �2C-adrenoceptors using [35S]GTP�S and cAMP functionalassays. Eur. J. Pharmacol. 411, 211–221.
rban, et al., 2007. Functional selectivity and classical concepts of quantitative phar-macology. J. Pharmacol. Exp. Ther. 320 (1), 1–13.
an Der Lee, M.M., Bras, M., van Koppen, C.J., Zaman, G.J., 2008. �-Arrestin recruit-ment assay for the identification of agonists of the sphingosine 1-phosphatereceptor EDG1. J. Biomol. Screen. 13 (10), 986–998.
auquelin, G., Van Liefde, I., 2005. G protein-coupled receptors: a count of 1001conformations. Fundam. Clin. Pharmacol. 19 (1), 45–56.
iolin, J.D., Lekowitz, R.J., 2007. �-Arrestin-biased ligands at seven transmembranereceptors. Trends Pharmacol. Sci. 28, 416–422.
iolin, J.D., DeWire, S.M., Yamashita, D., Rominger, D.H., Nguyen, L., Schiller, K.,Whalen, E.J., Gowen, M., Lark, M.W., 2010. Selectively engaging �-arrestins atthe angiotensin II type 1 receptor reduces blood pressure and increases cardiacperformance. J. Pharmacol. Exp. Ther. 335 (3), 572–579.
on Zastrow, M., Link, R., Daunt, D., Barsh, G., Kobilka, B., 1993. Subtype-specificdifferences in the intracellular sorting of G protein-coupled receptors. J. Biol.Chem. 268 (2), 763–766.
alters, R.W., Shukla, A., Kovacs, J.J., Violin, J.D., DeWire, S.M., Lam, C.M., Chen,J.R., Muelbauer, M.J., Whalen, E.J., Lefkowitz, R.J., 2009. �-Arrestin 1 mediates
ulletin 107 (2014) 89–101 101
nicotinic acid-induced flushing, but not its antilipolytic effect, in mice. J. Clin.Invest. 119, 1312–1321.
Wang, Q., Zhao, J., Brady, A.E., Feng, J., Allen, P.B., Lefkowitz, R.J., Greengard, P.,Limbird, L.E., 2004. Spinophilin blocks arrestin actions in vitro and in vivo atG protein-coupled receptors. Science 304 (5679), 1940–1944.
Whalen, E.J., Rajagopal, S., Lefkowitz, R.J., 2011. Therapeutic potential of �-arrestin-and G protein-biased agonists. Trends Mol. Med. 17 (3), 126–139.
Wisler, J.W., Xiao, K., Thomsen, A.R., Lefkowitz, R.J., 2014. Recent developments inbiased agonism. Curr. Opin. Cell Biol. 27, 18–24.
Wozniak, M., Limbird, L.E., 1996. The three alpha 2-adrenergic receptor subtypesachieve basolateral localization in Madin-Darby canine kidney II cells via differ-ent targeting mechanisms. J. Biol. Chem. 271 (9), 5017–5024.
Wu, G., Krupnick, J.G., Benovic, J.L., Lanier, S.M., 1997. Interaction of arrestins withintracellular domains of muscarinic and alpha2-adrenergic receptors. J. Biol.Chem. 272 (28), 17836–17842.
Yin, H., Chu, A., Li, W., Wang, B., Shelton, F., Otero, F., Nguyen, D.G., Caldwell,J.S., Chen, Y.A., 2009. Lipid G protein-coupled receptor ligand identifica-tion using �-arrestin PathHunterTM assay. J. Biol. Chem. 284 (18), 12328–
Zhang, Z., Zhou, W., Fan, J., Ren, Y., Yin, G., 2013. G-protein-coupled receptor kinaseinteractor-1 serine 419 accelerates premature synapse formation in corticalneurons by interacting with Ca(2+)/calmodulin-dependent protein kinase II�.Brain Res. Bull. 95, 70–77.