Differential Affinities of Visual Arrestin, bArrestin1, and bArrestin2for G Protein-coupled Receptors Delineate Two Major Classes ofReceptors*

Received for publication, December 22, 1999, and in revised form, March 1, 2000Published, JBC Papers in Press, March 29, 2000, DOI 10.1074/jbc.M910348199

Robert H. Oakley, Stephane A. Laporte‡, Jason A. Holt, Marc G. Caron§, and Larry S. Barak

From the Howard Hughes Medical Institute Laboratories, Departments of Cell Biology and Medicine, Duke UniversityMedical Center, Durham, North Carolina 27710

Visual arrestin, barrestin1, and barrestin2 comprise afamily of intracellular proteins that desensitize G pro-tein-coupled receptors (GPCRs). In addition, barrestin1and barrestin2 target desensitized receptors to clathrin-coated pits for endocytosis. Whether arrestins differ intheir ability to interact with GPCRs in cells is notknown. In this study, we visualize the interaction ofarrestin family members with GPCRs in real time and inlive cells using green fluorescent protein-tagged ar-restins. In the absence of agonist, visual arrestin andbarrestin1 were found in both the cytoplasm and nu-cleus of HEK-293 cells, whereas barrestin2 was foundonly in the cytoplasm. Analysis of agonist-mediated ar-restin translocation to multiple GPCRs identified twomajor classes of receptors. Class A receptors (b2 adre-nergic receptor, mu opioid receptor, endothelin type Areceptor, dopamine D1A receptor, and a1b adrenergicreceptor) bound barrestin2 with higher affinity thanbarrestin1 and did not interact with visual arrestin. Incontrast, class B receptors (angiotensin II type 1A recep-tor, neurotensin receptor 1, vasopressin V2 receptor,thyrotropin-releasing hormone receptor, and substanceP receptor) bound both barrestin isoforms with similarhigh affinities and also interacted with visual arrestin.Switching the carboxyl-terminal tails of class A andclass B receptors completely reversed the affinity ofeach receptor for the visual and non-visual arrestins. Inaddition, exchanging the barrestin1 and barrestin2 car-boxyl termini reversed their extent of binding to class Areceptors as well as their subcellular distribution. Theseresults reveal for the first time marked differences inthe ability of arrestin family members to bind GPCRs atthe plasma membrane. Moreover, they show that visualarrestin can interact in cells with GPCRs other thanrhodopsin. These findings suggest that GPCR signalingmay be differentially regulated depending on the cellu-lar complement of arrestin isoforms and the ability ofarrestins to interact with other cellular proteins.

G protein-coupled receptors (GPCRs)1 comprise a large genefamily of more than 1000 members that mediate distinct phys-iological functions as diverse as phototransduction, olfaction,vascular tone, cardiac output, digestion, and pain. GPCR sig-naling is regulated by a small family of intracellular arrestinproteins that includes visual arrestin (S-antigen), barrestin1(arrestin2), and barrestin2 (arrestin3) (1, 2). Visual arrestin is60 and 65% identical in amino acid composition to barrestin1and barrestin2, respectively, and predominantly localized inrod photoreceptor cells of the retina but can also be found inother tissues (3, 4). The barrestins are 78% identical in aminoacid composition and widely expressed in tissues, but theirexpression level varies in a cell type-specific fashion (5, 6). Thevariations in arrestin homology, localization, and expressionlevel suggest that arrestin family members may differ in theirabilities to regulate GPCR signaling.

Arrestins bind to GPCRs that are phosphorylated by G pro-tein-coupled receptor kinases (GRKs) (7–9). The binding of asingle arrestin to a phosphorylated receptor competitivelyblocks agonist-mediated signal transduction by uncoupling thereceptor from heterotrimeric G proteins (5, 7, 8, 10, 11). Thisprocess is termed desensitization. In contrast to visual ar-restins that primarily desensitize light-activated rhodopsin(10), the nonvisual arrestins barrestin1 and barrestin2 alsoparticipate in processes controlling re-establishment of recep-tor responsiveness (12–15). bArrestins target desensitized re-ceptors for endocytosis and resensitization by functioning asdocking proteins that link receptors to components of the en-docytic machinery such as AP-2 and clathrin (16, 17). They alsoregulate the rate at which endosomal receptors are dephospho-rylated and recycled back to the plasma membrane throughinteractions with specific clusters of GRK-phosphorylated res-idues in the GPCR carboxyl terminus (18).

Arrestins bind numerous GPCRs at the plasma membrane(19), and this finding suggests that receptors contain commonarrestin recognition motifs. In vitro studies using purified pro-teins have shown that specificity exists between arrestin familymembers and GPCRs. For example, visual arrestin binds rho-dopsin in preference to the b2-adrenergic receptor (b2AR) andm2 muscarinic acetylcholine receptor (m2mAChR), whereasbarrestin1 and barrestin2 bind the b2AR and m2mAChR in* This work was supported in part by National Institutes of Health

Grants NS 19576 (to M. G. C.) and HL 61365 (to L. S. B.). The costs ofpublication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

‡ Recipient of a fellowship award from the Heart and Stroke Foun-dation of Canada.

§ Investigator of the Howard Hughes Medical Institute and a recipi-ent of an unrestricted Neuroscience award from Bristol-Myers SquibbCo. To whom correspondence should be addressed: Investigator of theHoward Hughes Medical Institute. Howard Hughes Medical Institute,Box 3287, Duke University Medical Center, Durham, NC 27710. Tel.:919-684-5433; Fax: 919-681-8641; E-mail: caron002@mc.duke.edu.

1 The abbreviations used are: GPCR, G protein-coupled receptor;GRK, G protein-coupled receptor kinase; b2AR, b2-adrenergic receptor;m2mAChR, m2 muscarinic acetylcholine receptor; AT1AR, angiotensinII type 1A receptor; MOR, mu opioid receptor; ETAR, endothelin type Areceptor; D1AR, dopamine D1A receptor; a1bAR, a1b-adrenergic recep-tor; NTR-1, neurotensin 1 receptor; V2R, vasopressin V2 receptor;TRHR, thyrotropin-releasing hormone receptor; SPR, substance P re-ceptor; GFP, green fluorescent protein; YFP, yellow fluorescent protein;MEM, minimal essential medium.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 275, No. 22, Issue of June 2, pp. 17201–17210, 2000© 2000 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org 17201

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

preference to rhodopsin (5, 8, 20, 21). Furthermore, barrestin1binds the b2AR with a 2.5-fold greater affinity than barrestin2,and barrestin2 binds the m2mAChR with a 1.5-fold greateraffinity than barrestin1 (21). However, whether specificity ex-ists between arrestin family members and GPCRs in cells hasnot been explored.

In the following study we investigate the dynamic interac-tions between arrestin family members and GPCRs in live cellsby assessing the redistribution of fluorescent arrestins from thecytoplasm to agonist-activated receptors at the plasma mem-brane. Two classes of GPCRs, designated A and B, are identi-fied that differ in their affinities for the arrestin isoforms. ClassA receptors, such as the b2AR, do not interact with visualarrestin and bind barrestin1 with less affinity than barrestin2.Class B receptors, such as the angiotensin II type 1A receptor(AT1AR), interact with visual arrestin and bind both barres-tin1 and barrestin2 with similar high affinities. The moleculardeterminants underlying this classification appear to reside inspecific serine residues located in the receptor carboxyl-termi-nal tail. These findings reveal a potential role for visual arres-tin in the regulation of GPCRs outside the visual system.Moreover, they suggest that the particular cellular complementof arrestin isoforms and their distinct interactions with intra-cellular proteins will play a critical role regulating the patternof GPCR desensitization, sequestration, and resensitization.

EXPERIMENTAL PROCEDURES

Materials—Isoproterenol, epinephrine, and dopamine were pur-chased from Research Biochemicals Inc. Arginine vasopressin, angio-tensin II, and substance P were obtained from Sigma. Endothelin andneurotensin were from Peninsula Laboratories. 125I-Cyanopindolol waspurchased from NEN Life Science Products. Thyrotropin-releasing hor-mone and the thyrotropin-releasing hormone receptor (TRHR) cDNAwere kindly provided by Dr. Patricia M. Hinkle (University of Roches-ter, Rochester, NY).

Plasmids—Constructs were generated by polymerase chain reactionfollowing standard protocols and are depicted schematically in Fig. 1.bArrestin1 with GFP conjugated to its carboxyl terminus (barr1-GFP)has been described previously (22). It was constructed by replacing theterminal stop codon of barrestin1 with a SalI restriction site and in-serting the modified cDNA in frame into p(S65T)GFP (CLONTECH). Inorder to match perfectly the 12-amino acid linker in barr1-GFP, wemodified our previously described barrestin2-GFP conjugate (19). Thisnew barr2-GFP was constructed by replacing the terminal stop codon ofbarrestin2 with a SalI restriction site and inserting the modified cDNAin frame into p(S65T)GFP. The barr1-barr2CT-GFP chimera containsthe first 333 amino acids of barrestin1 (Met-1 to Gly-333) fused to thelast 75 amino acids of barrestin2 (Asp-336 to Cys-410). The barr2-barr1CT-GFP chimera contains the first 335 amino acids of barrestin2(Met-1 to Gly-335) fused to the last 85 amino acids of barrestin1 (Leu-334 to Arg-418). For both chimeras, the cDNA was inserted in frameinto the SalI restriction site in p(S65T)GFP. bArrestin1 with YFPconjugated to its amino terminus (YFP-barr1) was constructed by in-serting the barrestin1 cDNA in frame into the SalI restriction site inpEYFPC1 (CLONTECH). bArrestin2 with YFP conjugated to its aminoterminus (YFP-barr2) was constructed by inserting the barrestin2cDNA in frame into the SalI restriction site in pEYFPC1. The YFP-barr1S412A and YFP-barr1S412D mutants were generated by replac-ing nucleotides TCT, encoding serine at amino acid 412 in YFP-barr1,with GCT or GAT that mutates the serine to an alanine or aspartic acid,respectively. Visual arrestin with GFP conjugated to its amino termi-nus (GFP-arrestin) was constructed by inserting the visual arrestincDNA in frame into the HindIII site in pEGFPC3 (CLONTECH). Con-struction of the b2AR-V2R chimera and V2R-b2AR chimera have beendescribed previously (18). The b2AR-V2R chimera contains the first 341amino acids of the b2AR (Met-1 to Cys-341) fused to the last 29 aminoacids of the V2R (Ala-343 to Ser-371). The V2R-b2AR chimera containsthe first 342 amino acids of the V2R (Met-1 to Cys-342) fused to the last72 amino acids of the b2AR (Leu-342 to Leu-413).

Cell Culture and Transfection—Human embryonic kidney (HEK-293) cells were provided by the American Type Culture Collection(ATCC). HEK-293 cells were grown in Eagle’s minimal essential me-dium with Earle’s salt (MEM) supplemented with 10% (v/v) heat-inac-tivated fetal bovine serum and gentamicin (100 mg/ml). HEK-293 cells

stably expressing the b2AR, b2AR-GFP, and AT1AR were generated bystandard procedures using G418 (400 mg/ml) selection. Transient trans-fections were performed using a modified calcium phosphate coprecipi-tation method as described previously (23).

Receptor Expression—b2AR expression was measured by Scatchardanalysis using 125I-cyanopindolol (23). The expression of the AT1ARwas measured by flow cytometry and normalized according to the b2ARexpression measured in parallel by both flow cytometry and saturationbinding as described previously (24). Receptor expression in HEK-293cells stably overexpressing the b2AR was approximately 5.0 pmol/mgwhole cell protein. Receptor expression in HEK-293 cells stably over-expressing the AT1AR was approximately 3.0 pmol/mg whole cell pro-tein. Receptor expression in HEK-293 cells stably overexpressing theb2AR-GFP was approximately 1.5 pmol/mg whole cell protein.

bArrestin Expression—Each barrestin molecule is coupled to a singleGFP moiety; therefore, the measured fluorescent intensity in the cyto-plasm of cells is proportional to the number of GFP fluorophores andhence proportional to the number of GFP-labeled barrestin molecules.Cells selected for analysis of barrestin translocation possessed low andequivalent levels of GFP fluorescence. The relative level of barrestin-GFP fluorescence (in intensity per pixel) was measured using the“range of interest” analysis provided with the Zeiss LSM-510 confocalmicroscope software. Settings on the microscope (laser power, pinholesize, detector gain, amplifier offset, amplifier gain, etc.) were heldconstant within and between experiments to ensure that cells express-ing similar amounts of the different arrestin isoforms were compared.Expression of GFP-labeled barrestins in the population of cells wasassessed by Western blotting using mouse and rabbit anti-GFP anti-bodies (CLONTECH).

Confocal Microscopy—Transfected HEK-293 cells were plated on35-mm glass-bottomed culture dishes. Two hours before analysis oftranslocation, the medium was replaced with serum-free MEM supple-mented with 10 mM HEPES. Confocal microscopy was performed on aZeiss laser scanning confocal microscope (LSM-510) using a heated(37 °C) microscope stage as described previously (18, 19). Images werecollected sequentially using single line excitation (488 nm). Saturatingconcentrations of agonist were applied directly over the selected cells 10or 15 s before the third image within a time series. The relative level ofbarrestin-GFP fluorescence (in intensity per pixel) was measured in afixed area of the cytoplasm over the duration of the experiment usingthe range of interest analysis provided with the microscope software.Data were analyzed using a “plateau with exponential decay” nonlinearregression function in GraphPad Prism.

Modeling of bArrestin Translocation—In rod photoreceptor cells, rho-dopsin is expressed at higher levels than visual arrestin (25). However,in the heart, barrestin appears to be expressed at higher levels than theb2AR so that the intracellular arrestin concentration is not a limitingfactor in receptor desensitization (26). In the latter case, most of theintracellular arrestins will remain in the cytosol with only a smallfraction translocating to agonist-activated receptors. Experimental con-ditions, however, can be modified in order to increase the translocatingfraction of arrestin proteins. The expression of endogenous receptorprotein can be augmented in a permanent or transient manner utilizingreceptor cDNA transfection methods. Receptor overexpression createsan excess of potential binding sites for arrestin. These surplus sitescreate an environment permitting much of the arrestin to bind at themembrane to activated receptors, resulting in an increased cytosolicdepletion of the arrestin protein.

Our experiments were performed in HEK-293 cells stably overex-pressing the b2AR (or AT1AR) at approximately 5.0 pmol/mg whole cellprotein and transiently overexpressing GFP-labeled barrestins. bArres-tin translocation was analyzed only in cells expressing a definedamount of the GFP-labeled barrestin (measured as described above). Byusing as a standard a permanent line of HEK-293 cells expressing aknown amount of b2AR-GFP (1.5 pmol/mg whole cell protein), we weretherefore able to estimate the receptor/barrestin ratio in the selectedcells as shown below in Equations 1 and 2.

pmol/mg of cell barrestin-GFP

5@pmol/mg of b2AR-GFP]

[fluorescence/pixel of b2AR-GFP]z

fluorescence/pixel of membrane-bound barrestin-GFP[fractional complement of membrane barrestin-GFP]

(Eq. 1)

and

Visualization of Arrestin Translocation to GPCRs17202

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

total cell b2ARtotal cell barrestin

5pmol/mg of wild-type b2AR

pmol/mg of cell barrestin-GFP(Eq. 2)

Our measurements indicate receptors were expressed at levels 2.5–5-fold greater than the level of barrestin-GFP. To ensure the simultane-ous activation of these receptors, we used a large excess of agonist.Therefore, our barrestin translocation experiments utilized both theoverexpression and simultaneous activation of receptors and can bemodeled as follows.

The homologous desensitization of GPCRs first requires receptorphosphorylation by GRKs (7–9). The phosphorylation of agonist-acti-vated membrane receptors, which is essentially irreversible at theplasma membrane, can be described by first order rate Equation 3,

dR~t!dt

5 2k1 z R~t! with solution R~t! 5 RT exp(2k1t) (Eq. 3)

where R(t) is the number of non-GRK-phosphorylated receptors in thepresence of a saturating concentration of agonist; RT is the time-inde-pendent total number of membrane receptors; k1 is a measure of therate of phosphorylation, and exp is the abbreviation for the exponentialfunction. If Rphos(t) is the amount of free phosphorylated membranereceptor not bound to arrestin, and RB(t) is the number of membranereceptors bound to arrestin, conservation of mass requires the following(see Equation 4).

RT 5 R~t! 1 Rphos~t! 1 RB~t! (Eq. 4)

Likewise if BT is the total amount of cellular barrestin, BM(t) is thetime-dependent amount of receptor-associated arrestin on the mem-brane, and B(t) is the amount of cytosolic arrestin, then we determineEquation 5.

BT 5 BM~t! 1 B~t! (Eq. 5)

Experiments indicate that a single phosphorylated receptor binds asingle arrestin (i.e. BM 5 RB) (5, 7, 8). This observation and the expo-nential decay of non-GRK phosphorylated membrane receptors de-scribed by rate Equation 3 provide normalized rate equations (Equa-tions 6 or 7) for the disappearance of cytosolic arrestins X(t) 5 B(t)/BT

with time, t.

2dX~t!

dt5

dFBM~t!BT

Gdt

5 k2 z @Rphos~t!# z FB~t!BTG2 k3 z FBM~t!

BTG (Eq. 6)

or

2dXdt

5 @k3 1 k2RT~1 2 exp(2k1t!! 2 k2BT~1 2 X!]X 2 k3 (Eq. 7)

Note that the kinetic parameters k2 and k3 indicating the removal andreappearance of free cytosolic arrestins may depend on the type ofarrestin present.

At the beginning of an experiment essentially all of the arrestin iscytosolic so that X(0) 5 1, i.e. B(0) 5 BT at time t 5 0. The nonlinearterm in cytosolic arrestin, k2BT(1 2 X(t)), can be neglected with respectto the term in k2RT in Equation 7 if an experiment is designed so thatRT, the total number of membrane receptors, is significantly greaterthan BT, the total amount of arrestins in the cytosol that can possiblybind and desensitize these receptors. Moreover, the observation that fort . 0 the rate of cytosolic arrestin disappearance must always be slowerthan the rate of GRK-mediated receptor phosphorylation (i.e. 1 2 X(t) ,1 2 exp(2k1t)) further justifies neglecting this term. The rate of disap-pearance of cytosolic arrestins then simplifies to the relationship shownin Equation 8.

2dXdt

5 @k3 1 k2RT~1 2 exp(2k1t!#X 2 k3 (Eq. 8)

With the definition that a 5 k3 1 k2RT the solution to Equation 8 isgiven in Equation 9.

X~t! 5 expS2 a z t 1k2RT

k1z @1 2 exp~ 2 k1t!#D

z H1 1 k3E0

t

dt# z expSa z t# 2k2RT

k1z @1 2 exp~ 2 k1t#!#D J (Eq. 9)

Our experiments were designed to measure the rate of barrestindisappearance in a region of cytosol in single cell that is defined by thelaser-illuminated volume imaged in the confocal microscope. Since eachbarrestin molecule is coupled to a single GFP moiety, the measuredfluorescence intensity, I(t), in the cytosol will be proportional to thenumber of imaged GFP fluorophores (with proportionality constant I0)and hence proportional to the number of barrestin-GFP molecules X(t)in the illuminated volume. Therefore, the time-dependent change in themeasured fluorescence intensity, I(t) (Equation 10), or the normalizedfluorescent intensity, I(t)/I0, can be calculated from Equation 9 asfollows:

I~t! 5 I0 z X~t! 5 I0 z expS2a z t 1k2RT

k1z @1 2 exp~ 2 k1t!#D z

H1 1 k3E0

t

dt# z expSa z t# 2k2RT

k1z @1 2 exp~ 2 k1t#!#DJ (Eq. 10)

Data were analyzed by Equation 10 using the nonlinear regressionfunction Minerr of Mathcad version 6.0 (Mathsoft, Cambridge, MA).

RESULTS

Translocation of bArrestin1 and bArrestin2 to the b2AR—GFP-labeled proteins not only provide a means for analyzingintracellular protein-protein interactions without having todisrupt the plasma membrane, but they also allow for theassessment of the kinetics of interactions that occur in seconds(27, 28). Therefore, to compare the ability of barrestin1 andbarrestin2 to interact with agonist-activated GPCRs in realtime and in live cells, we fused the green fluorescent protein(GFP) to the carboxyl terminus of barrestin1 (barr1-GFP) andbarrestin2 (barr2-GFP) (see Fig. 1). The GFP-tagged bar-restins were then transiently transfected into HEK-293 cellsstably overexpressing the b2AR. These cells provide a cellularplatform in which the affinities of the different arrestin iso-forms for the b2AR can be reliably compared since each cell willexpress the same number of receptors. The transfected bar-restins were expressed at comparable levels in the populationof cells (Fig. 2C, inset). However, to ensure further an accuratecomparison of the two barrestin isoforms, single cells express-ing equivalent amounts of barr1-GFP or barr2-GFP were se-lected for analysis of translocation by matching the relative

FIG. 1. Schematic representation of the barrestin and visualarrestin fusion proteins. bArrestin1 (white) is comprised of 418amino acids; barrestin2 (black) is comprised of 410 amino acids; visualarrestin (gray) is comprised of 404 amino acids, and the GFP/YFPmoiety (gray) is comprised of 238 amino acids. The barr1-barr2CT-GFPchimera contains the first 333 amino acids of barrestin1 fused to thelast 75 amino acids of barrestin2. The barr2-barr1CT-GFP chimeracontains the first 335 amino acids of barrestin2 fused to the last 85amino acids of barrestin1. Serine 412 is mutated to an alanine inYFP-barr1S412A and to an aspartic acid in YFP-barr1S412D. Thelinkers (solid line) in barr1-GFP, barr2-GFP, barr1-barr2CT-GFP, andbarr2-barr1CT-GFP are identical. The linkers (solid line) in YFP-barr1,YFP-barr2, YFP-barr1S412A, and YFP-barr1S412D are identical.

Visualization of Arrestin Translocation to GPCRs 17203

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

intensity of GFP fluorescence as described under “Experimen-tal Procedures.” In the absence of agonist, we observed a dif-ference in the subcellular distribution of the two barrestinisoforms. barr1-GFP was distributed in both the cytoplasm andnucleus of cells, whereas barr2-GFP was distributed in thecytoplasm of cells but excluded from the nucleus (Fig. 2, A andB, compare 0-s images). Upon agonist addition, both barr1-GFP and barr2-GFP redistributed from the cytoplasm to thereceptor at the plasma membrane (Fig. 2, A and B). However,as indicated by the increase in fluorescence at the plasmamembrane, barr2-GFP translocated faster and to a greaterextent than did barr1-GFP.

To quantitate the differences in the translocation profiles ofbarr1-GFP and barr2-GFP to the b2AR, we measured the time-dependent loss of barrestin fluorescence from the cytoplasmafter treatment with agonist. These data were then analyzedusing a plateau with exponential decay nonlinear regressionfunction (GraphPad Prism) that revealed three major differ-ences in the translocation profiles of the two barrestin isoforms(Fig. 2C). First, translocation of barr2-GFP began sooner thanbarr1-GFP. The delay in time between agonist addition (arrow)

and the initial loss of barrestin from the cytoplasm was 1.5 61.0 s for barr2-GFP and 8.2 6 1.4 s for barr1-GFP. Second,translocation of barr2-GFP occurred faster than barr1-GFP.The half-life of barrestin depletion from the cytoplasm was19.2 6 1.1 s for barr2-GFP and 35.8 6 2.0 s for barr1-GFP.Third, translocation of barr2-GFP proceeded to a greater ex-tent than barr1-GFP. The fraction of cytoplasmic barr2-GFPthat translocated to the plasma membrane was 63.2 6 0.4%,whereas only 33.8 6 0.4% of the cytoplasmic barr1-GFP trans-located to the plasma membrane.

Fusion of the GFP moiety to the carboxyl terminus of bar-restin1 and barrestin2 may differentially alter their affinitiesfor the b2AR. Therefore, we fused the yellow fluorescent protein(YFP) variant of GFP to the amino terminus of barrestin1(YFP-barr1) and barrestin2 (YFP-barr2), and we compared theability of these barrestin conjugates to translocate to the b2ARstably overexpressed in HEK-293 cells. In the absence of ago-nist, YFP-barr1 was distributed in both the cytoplasm andnucleus of cells, whereas YFP-barr2 was distributed in thecytoplasm but excluded from the nucleus (Fig. 3, A and B,compare 0-s images). Upon agonist addition, both barrestin

FIG. 2. Translocation of barr1-GFP and barr2-GFP to theb2AR. HEK-293 cells stably overexpressing the b2AR were transientlytransfected with barr1-GFP (A) or barr2-GFP (B). Single cells express-ing equivalent amounts of the two barrestin isoforms were selected asdescribed under “Experimental Procedures.” A and B, the distributionof barr1-GFP and barr2-GFP fluorescence was visualized with theconfocal microscope before (0 s) and after (35 s and 180 s) treatmentwith 10 mM isoproterenol. C, quantitation of barr1-GFP and barr2-GFPtranslocation to the b2AR. barr1-GFP and barr2-GFP fluorescence weremeasured in the cytoplasm of cells before and after treatment with 10mM isoproterenol. Confocal images were collected every 24.8 s (scantime 5 3.9 s), and agonist was added (arrow) 10 s before the third scan.Data represent the mean 6 S.E. of 4–5 independent experiments (n 521–24 cells) and were analyzed using a plateau with exponential decaynonlinear regression function in GraphPad Prism. The inset is a repre-sentative Western blot showing the expression level of barr1-GFP (lane1) and barr2-GFP (lane 2) in the population of cells. Mock-transfectedcells are included in lane 3.

FIG. 3. Translocation of YFP-barr1 and YFP-barr2 to the b2AR.HEK-293 cells stably overexpressing the b2AR were transiently trans-fected with YFP-barr1 (A) or YFP-barr2 (B). Single cells expressingequivalent amounts of the two barrestin isoforms were selected asdescribed under “Experimental Procedures.” A and B, the distributionof YFP-barr1 and YFP-barr2 fluorescence was visualized with the con-focal microscope before (0 s) and after (40 s and 180 s) treatment with10 mM isoproterenol. C, quantitation of YFP-barr1 and YFP-barr2translocation to the b2AR. YFP-barr1 and YFP-barr2 fluorescence weremeasured in the cytoplasm of cells before and after treatment with 10mM isoproterenol. Confocal images were collected every 24.8 s (scantime 5 3.9 s), and agonist was added (arrow) 15 s before the third scan.Data represent the mean 6 S.E. of four independent experiments (n 519–20 cells) and were analyzed using a plateau with exponential decaynonlinear regression function in GraphPad Prism. The inset is a repre-sentative Western blot showing the expression level of YFP-barr1 (lane1) and YFP-barr2 (lane 2) in the population of cells. Mock-transfectedcells are included in lane 3.

Visualization of Arrestin Translocation to GPCRs17204

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

isoforms redistributed from the cytoplasm to the receptor at theplasma membrane with profiles very similar to their GFP coun-terparts (Fig. 3, A and B). Translocation of YFP-barr2 begansooner, occurred faster, and proceeded to a greater extent thanYFP-barr1 translocation (Fig. 3C). The delay in translocationwas 1.95 6 0.8 s for YFP-barr2 and 8.34 6 3.2 s for YFP-barr1.The half-life of translocation was 28.0 6 0.8 s for YFP-barr2and 44.3 6 4.9 s for YFP-barr1. The extent of barrestin trans-location corresponded to a 66.0 6 0.3% reduction in cytoplasmicYFP-barr2 and a 39.7 6 1.0% reduction in cytoplasmic YFP-barr1. These results demonstrate that, independent of the lo-cation of the fluorescent moiety, barrestin2 has a greater af-finity for the b2AR than does barrestin1.

Translocation of bArrestin1 Phosphorylation Mutants to theb2AR—It has been recently demonstrated that barrestin1 isconstitutively phosphorylated on a carboxyl-terminal serine(Ser-412) when it is in the cytoplasm (29). However, barrestin1is dephosphorylated when it is recruited to the agonist-acti-vated b2AR at the plasma membrane (29). Since barrestin2does not possess the Ser-412 residue, it may not be subject tothe same type of regulation. To test whether the phosphoryla-tion status of barrestin1 regulates its translocation to theb2AR, we mutated Ser-412 to an alanine (S412A) or to anaspartic acid (S412D) to mimic the dephosphorylated and con-stitutively phosphorylated forms of barrestin1, respectively(29). HEK-293 cells stably overexpressing the b2AR were tran-siently transfected with YFP-barr1, YFP-barr1S412A, or YFP-barr1S412D. In the absence of agonist, no differences wereobserved in the subcellular distribution of the wild-type andmutant barrestins as each protein was distributed in both thecytoplasm and nucleus of cells (Fig. 4, A–C, compare 0-s imag-es). Upon agonist addition, the wild-type and mutant bar-restins translocated to the b2AR with very similar profiles (Fig.4, A–C). As shown in Fig. 4D, the delay in translocation was5.7 6 2.6 s for YFP-barr1, 4.9 6 1.8 s for YFP-barr1S412A, and5.2 6 2.7 s for YFP-barr1S412D. The half-life of translocationwas 44.8 6 4.2 s for YFP-barr1, 47.4 6 2.7 s for YFP-barr1S412A, and 46.9 6 4.2 s for YFP-barr1S412D. The extent

of barrestin translocation corresponded to a 42.0 6 0.9, 40.6 60.6, and 44.6 6 1.0% reduction in cytoplasmic YFP-barr1, YFP-barr1S412A, and YFP-barr1S412D, respectively. Therefore,the translocation profile of barrestin1 to the b2AR (which isdelayed in onset, slower in rate, and reduced in magnitudecompared with barrestin2 translocation) is not influenced bythe phosphorylation status of the Ser-412 residue.

Translocation of bArrestin1 and bArrestin2 Chimeras to theb2AR—Once bound to the GRK-phosphorylated b2AR, barres-tin1 and barrestin2 function as docking proteins that link thereceptor to clathrin-coated pits by interacting with componentsof the endocytic machinery such as AP-2 and clathrin (16, 17).bArrestin2 binds clathrin with a 6-fold greater affinity thanbarrestin1, and the two barrestin isoforms may differ in theircapacity to bind AP-2 (16, 17). To test whether differences inthe ability of barrestin1 and barrestin2 to interact with theseproteins contribute to their different translocation profiles, weswitched the carboxyl-terminal domains of barrestin1 and bar-restin2 because this region contains both the clathrin- andAP-2-binding sites (see Fig. 1) (16, 30). The ability of thesechimeric barrestins to translocate to the b2AR stably overex-pressed in HEK-293 cells was then assessed. In the absence ofagonist, the subcellular distribution of the chimeric barrestinswas completely reversed from their wild-type counterparts(Fig. 5, A and B, compare 0-s images). The barr1-barr2CT-GFPchimera was distributed in the cytoplasm of cells but excludedfrom the nucleus, whereas the barr2-barr1CT-GFP chimerawas distributed in both the cytoplasm and nucleus of cells. Bothchimeras translocated to the b2AR upon agonist addition (Fig.5, A and B). Similar to our findings for barr1-GFP and barr2-GFP, translocation of barr2-barr1CT-GFP began sooner andoccurred faster than barr1-barr2CT-GFP translocation (Fig.5C). The delay in translocation was 1.8 6 1.5 s for barr2-barr1CT-GFP and 8.6 6 2.3 s for barr1-barr2CT-GFP. Thehalf-life of translocation was 30.6 6 1.8 s for barr2-barr1CT-GFP and 50.7 6 4.1 s for barr1-barr2CT-GFP. However, incontrast to our findings for barr1-GFP and barr2-GFP, the twochimeric barrestins translocated to similar extents (Fig. 5C).

FIG. 4. Translocation of barrestin1 phosphorylation mutants to the b2AR. HEK-293 cells stably overexpressing the b2AR were tran-siently transfected with YFP-barr1 (A), YFP-barr1S412A (B), or YFP-barr1S412D (C). Single cells expressing equivalent amounts of the wild-typeand mutant barrestin1 isoforms were selected as described under “Experimental Procedures.” A–C, the distribution of YFP-barr1, YFP-barr1S412A, and YFP-barr1S412D fluorescence was visualized with the confocal microscope before (0 s) and after (180 s) treatment with 10 mM

isoproterenol. D, quantitation of YFP-barr1, YFP-barr1S412A, and YFP-barr1S412D translocation to the b2AR. YFP-barr1, YFP-barr1S412A, andYFP-barr1S412D fluorescence were measured in the cytoplasm of cells before and after treatment with 10 mM isoproterenol. Confocal images werecollected every 24.8 s (scan time 5 3.9 s), and agonist was added (arrow) 15 s before the third scan. Data represent the mean 6 S.E. of threeindependent experiments (n 5 11–14 cells) and were analyzed using a plateau with exponential decay nonlinear regression function in GraphPadPrism.

Visualization of Arrestin Translocation to GPCRs 17205

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

We measured a 53.5 6 0.5% reduction in cytoplasmic barr2-barr1CT-GFP which is significantly less than the 63% reduc-tion observed for barr2-GFP, and we measured a 57.0 6 1.3%reduction in cytoplasmic barr1-barr2CT-GFP which is signifi-cantly more than the 34% reduction observed for barr1-GFP.

These results demonstrate that residues in the carboxyl-terminal domains of barrestin1 and barrestin2 mediate thedistinct subcellular distribution patterns of the two barrestinisoforms in the absence of agonist. These residues also deter-mine the extent to which each isoform translocates to theagonist-activated b2AR presumably through interactions withAP-2 and/or clathrin. In addition, they indicate that the initialaffinity of barrestins for the b2AR is mediated primarily byresidues in the amino-terminal domains of the two barrestinisoforms which is consistent with findings from in vitro studies(21).

Translocation of bArrestin1 and bArrestin2 to OtherGPCRs—Once barrestin binds to agonist-activated GPCRs atthe plasma membrane, the fate of the receptor-barrestin com-

plex differs among receptors. For some GPCRs, such as theb2AR, the receptor-barrestin complex dissociates at or near theplasma membrane shortly after the formation of clathrin-coated pits, and barrestin is excluded from receptor-containingendocytic vesicles (18, 22). However, for other GPCRs, thereceptor-barrestin complex remains intact, and barrestin isinternalized with the receptor into endosomes (18, 22). Theangiotensin II type 1A receptor (AT1AR) belongs to the lattergroup of receptors; therefore, we compared the ability of bar-restin1 and barrestin2 to translocate to the AT1AR stablyoverexpressed in HEK-293 cells. In marked contrast to ourfindings for the b2AR (Figs. 2 and 3), both barrestins translo-cated to the AT1AR with nearly identical profiles (Fig. 6, A andB). As shown in Fig. 6C, the delay in translocation was 5.4 61.5 s for barr2-GFP and 10.2 6 0.9 s for barr1-GFP. Thehalf-life of translocation was 26.6 6 1.9 s for barr2-GFP and25.1 6 1.2 s for barr1-GFP. The extent of barrestin transloca-tion corresponded to a 61.3 6 0.8% reduction in cytoplasmicbarr2-GFP and a 61.7 6 0.6% reduction in cytoplasmic barr1-GFP. Moreover, with a longer agonist treatment, both barres-

FIG. 5. Translocation of barrestin chimeras to the b2AR. HEK-293 cells stably overexpressing the b2AR were transiently transfectedwith barr1-barr2CT-GFP chimera (A) or barr2-barr1CT-GFP chimera(B). Single cells expressing equivalent amounts of the two barrestinchimeras were selected as described under “Experimental Procedures.”A and B, the distribution of barr1-barr2CT-GFP and barr2-barr1CT-GFP fluorescence was visualized with the confocal microscope before (0s) and after (35 s and 180 s) treatment with 10 mM isoproterenol. C,quantitation of barr1-barr2CT-GFP and barr2-barr1CT-GFP translo-cation to the b2AR. barr1-barr2CT-GFP and barr2-barr1CT-GFP fluo-rescence were measured in the cytoplasm of cells before and aftertreatment with 10 mM isoproterenol. Confocal images were collectedevery 24.8 s (scan time 5 3.9 s), and agonist was added (arrow) 10 sbefore the third scan. Data represent the mean 6 S.E. of three inde-pendent experiments (n 5 11–13 cells) and were analyzed using aplateau with exponential decay nonlinear regression function in Graph-Pad Prism. For comparison, translocation profiles determined forbarr1-GFP and barr2-GFP in Fig. 2C are included (broken gray lines).

FIG. 6. Translocation of barr1-GFP and barr2-GFP to theAT1AR. HEK-293 cells stably overexpressing the AT1AR were tran-siently transfected with barr1-GFP (A) or barr2-GFP (B). Single cellsexpressing equivalent amounts of the two barrestin isoforms wereselected as described under “Experimental Procedures.” A and B, thedistribution of barr1-GFP and barr2-GFP fluorescence was visualizedwith the confocal microscope before (0 s) and after (30 s and 120 s)treatment with 1 mM angiotensin II. C, quantitation of barr1-GFP andbarr2-GFP translocation to the AT1AR. barr1-GFP and barr2-GFPfluorescence were measured in the cytoplasm of cells before and aftertreatment with 1 mM angiotensin II. Confocal images were collectedevery 20.2 s (scan time 5 3.9 s), and agonist was added (arrow) 10 sbefore the third scan. Data represent the mean 6 S.E. of 5–7 independ-ent experiments (n 5 27–36 cells) and were analyzed using a plateauwith exponential decay nonlinear regression function in GraphPadPrism. D, the distribution of barr1-GFP and barr2-GFP was visualizedwith the confocal microscope 20 min after treatment with 1 mM angio-tensin II.

Visualization of Arrestin Translocation to GPCRs17206

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

tin isoforms were observed to internalize with the AT1AR intoendocytic vesicles (Fig. 6D). These results show that barr1-GFP has the capacity to translocate as quickly and to the sameextent as barr2-GFP and that barrestin1 and barrestin2 havesimilar high affinities for the AT1AR.

We next evaluated the translocation of barrestin1 and bar-restin2 to a variety of other GPCRs including the mu opioidreceptor (MOR), endothelin type A receptor (ETAR), dopamineD1A receptor (D1AR), a1b-adrenergic receptor (a1bAR), neu-rotensin 1 receptor (NTR-1), vasopressin V2 receptor (V2R),thyrotropin-releasing hormone receptor (TRHR), and sub-stance P receptor (SPR). With the exception of the MOR, whichwas stably overexpressed, all other receptors were transientlyoverexpressed in HEK-293 cells. A qualitative comparison wasthen made between the levels of barr1-GFP and barr2-GFPthat translocated to the plasma membrane 2 min after agonisttreatment (Fig. 7). The amount of barr2-GFP that translocatedto the MOR, ETAR, D1AR, and a1bAR was greater than theamount of barr1-GFP. In contrast, the amount of barr2-GFPand barr1-GFP that translocated to the NTR, V2R, TRHR, andSPR was very similar as both barrestin isoforms translocatedto a large extent. The robust translocation of barr1-GFP tothese receptors at the plasma membrane cleared out the cyto-plasm and revealed the pool of barr1-GFP distributed in thenucleus. With longer agonist incubations, both barrestin iso-forms dissociated from the MOR, ETAR, D1AR, and a1bAR atthe plasma membrane and did not traffic with these receptorsinto endocytic vesicles (Fig. 8 and data not shown). In contrast,both barrestin isoforms remained associated with the NTR,V2R, TRHR, and SPR and internalized with these receptorsinto endocytic vesicles (Fig. 8 and data not shown).

These results identify two classes of GPCRs. “Class A” re-ceptors include the b2AR, MOR, ETAR, D1AR, and a1bAR.These receptors, which recognize three different G proteins (Gs,Gi, and Gq), bind barrestin2 with higher affinity than barres-tin1. Complexes between class A receptors and barrestin1 orbarrestin2 are relatively unstable, however, and dissociate ator near the plasma membrane and are excluded from receptor-containing vesicles. “Class B” receptors include the AT1AR,NTR-1, V2R, TRHR, and SPR. These receptors, which recog-nize Gs and Gq, bind both barrestin1 and barrestin2 with high

affinity. Complexes between class B receptors and barrestin1or barrestin2 are stable and internalize together into endocyticvesicles.

Translocation of Visual Arrestin to GPCRs—Although visualarrestin preferentially binds rhodopsin, in vitro studies haveshown that it can also bind the b2AR but with an affinity5–15-fold lower than barrestin1 and barrestin2 (5, 8, 21). Totest whether this interaction occurs in cells, we constructed a

FIG. 7. Translocation of barr1-GFPand barr2-GFP to various GPCRs.HEK-293 cells stably expressing the MORor transiently expressing the ETAR,D1AR, a1bAR, NTR-1, V2R, TRHR, orSPR were transiently transfected withbarr1-GFP or barr2-GFP. Shown are rep-resentative confocal microscopic imagesof barr1-GFP and barr2-GFP fluores-cence approximately 2 min after treat-ment with the appropriate agonist.

FIG. 8. Distribution of barr1-GFP and barr2-GFP 20 min afteragonist treatment. HEK-293 cells were transiently transfected withbarr1-GFP or barr2-GFP and either the D1AR, a1bAR, V2R, or SPR.Shown are representative confocal microscopic images of barr1-GFPand barr2-GFP fluorescence approximately 20 min after treatmentwith the appropriate agonist.

Visualization of Arrestin Translocation to GPCRs 17207

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

GFP-arrestin fusion protein (see Fig. 1), and we evaluated itsability to translocate to the b2AR stably overexpressed in HEK-293 cells. In the absence of agonist, GFP-arrestin (like barres-tin1) was distributed in both the cytoplasm and nucleus of cells(Fig. 9A, upper panel). Upon agonist addition, however, GFP-arrestin did not translocate to the b2AR (Fig. 9A, upper panel).Even after 60 min of agonist treatment, no translocation wasobserved (data not shown). Similarly, GFP-arrestin did nottranslocate to MOR, ETAR, D1AR, and a1bAR (data notshown).

We next evaluated the ability of visual arrestin to translo-cate to the AT1AR stably overexpressed in HEK-293 cells sincethis receptor is less selective and binds both barrestin isoformswith high affinity. As shown in the lower panel of Fig. 9A,GFP-arrestin translocated to the AT1AR within 2 min of ago-nist stimulation. Moreover, GFP-arrestin translocated to theother class B receptors (Fig. 9B). The robust translocation ofGFP-arrestin to the V2R, TRHR, and SPR at the plasma mem-brane cleared out the cytoplasm and revealed the pool of GFP-arrestin distributed in the nucleus. Interestingly, in contrast toour observations for the barrestin isoforms, translocation ofvisual arrestin resulted in a more diffuse and less punctatepattern of fluorescence at the plasma membrane. This suggeststhat although visual arrestin binds class B receptors at theplasma membrane, it may not target these receptors to clath-rin-coated pits which is consistent with its reported inability tobind clathrin (17). These results not only demonstrate thatvisual arrestin can interact in cells with GPCRs other thanrhodopsin, but they also further establish the existence of twoclasses of GPCRs that differ in their affinity for arrestin familymembers.

Translocation of Visual and Nonvisual Arrestins to ChimericGPCRs—Binding of arrestin isoforms to many GPCRs appearsto be mediated by GRK-phosphorylated serine and threonineresidues located in the receptor carboxyl-terminal tail. To testwhether the carboxyl-terminal tail of class A and class B re-ceptors contributes to their differential affinities for the visualand nonvisual arrestins, we employed chimeric receptors inwhich the tails of the b2AR (class A) and V2R (class B) wereswapped (18). The b2AR with the V2R tail (b2AR-V2R chimera)and the V2R with the b2AR tail (V2R-b2AR chimera) weretransiently expressed in HEK-293 cells. We then comparedqualitatively the levels of barr1-GFP, barr2-GFP, and GFP-arrestin at the plasma membrane 2 min after agonist treat-ment. The amount of barr2-GFP that translocated to the V2R-b2AR chimera was much greater than the amount of barr1-GFP, and no translocation was observed for GFP-arrestin (Fig.10, upper panel). Therefore, the V2R-b2AR chimera displays a“b2AR-like” affinity for the visual and nonvisual arrestins. Inmarked contrast, both barrestin isoforms translocated to asimilarly large extent to the b2AR-V2R chimera (Fig. 10, lower

panel). Moreover, GFP-arrestin translocated to the b2AR-V2Rchimera at the plasma membrane. Therefore, the b2AR-V2Rchimera displays a “V2R-like” affinity for the visual and non-visual arrestins. These results demonstrate that residues inthe receptor carboxyl-terminal tail play a central role in deter-mining the affinity of GPCRs for arrestin family members.

DISCUSSION

In this study we identify two major classes of GPCRs thatdemonstrate marked differences in their interactions with vis-ual arrestin, barrestin1, and barrestin2. Class A receptors,which bind both biogenic amines and peptide ligands, demon-strate little to no affinity for visual arrestin and have a higheraffinity for barrestin2 than barrestin1. Class B receptors,which bind peptide ligands, bind visual arrestin and have asimilar high affinity for both barrestin1 and barrestin2. Resi-dues in the receptor carboxyl-terminal tail appear entirelyresponsible for determining class A or B characteristics ratherthan the nature of the ligand. This classification paradigmprovides a foundation for predicting the manner in which thecellular complement of arrestin isoforms will regulate the de-sensitization, sequestration, and resensitization of a particularGPCR.

For many GPCRs, homologous desensitization involves bothGRK phosphorylation and arrestin binding (1, 2). Our findingsfit a model in which the kinetics of GRK phosphorylation arerate-limiting and facilitate the interaction of arrestin with re-ceptors (Fig. 11A). In support of this interpretation, GRK phos-phorylation of rhodopsin and the b2AR is necessary for highaffinity binding of arrestins both in vitro and in cells (7–9, 19,

FIG. 9. Translocation of GFP-arres-tin to various GPCRs. A, HEK-293 cellsstably overexpressing the b2AR (upperpanel) or the AT1AR (lower panel) weretransiently transfected with GFP-arres-tin. The distribution of GFP-arrestin flu-orescence was visualized with the confo-cal microscope before (0 s) and after (120s) treatment with the appropriate ago-nist. B, HEK-293 cells were transientlytransfected with GFP-arrestin and eitherthe NTR-1, V2R, TRHR, or SPR. Shownare representative confocal microscopicimages of GFP-arrestin fluorescence ap-proximately 2 min after treatment withthe appropriate agonist.

FIG. 10. Translocation of barr1-GFP, barr2-GFP, and GFP-ar-restin to chimeric GPCRs. HEK-293 cells were transiently trans-fected with barr1-GFP, barr2-GFP, or GFP-arrestin and either theV2R-b2AR chimera (upper panel) or the b2AR-V2R chimera (lower pan-el). Shown are representative confocal microscopic images of barr1-GFP, barr2-GFP, and GFP-arrestin fluorescence approximately 2 minafter treatment with the appropriate agonist.

Visualization of Arrestin Translocation to GPCRs17208

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

21). Moreover, GFP analogues of GRK2 translocate to the SPRat the plasma membrane prior to the binding of barrestin (28).Some barrestin binding may also occur, however, with agonist-activated receptors in the absence of GRK phosphorylation. Forexample, in vitro studies have shown that barrestin binds withlow affinity to the unphosphorylated b2AR and m2mAChR (21).In addition, in experiments with the SPR under conditions thatshould significantly reduce GRK-mediated phosphorylation inresponse to agonist (60 min agonist treatment at 4 °C), bothtransfected and endogenous barrestins were found to associatewith the receptor at the plasma membrane (31). bArrestinbinding to GPCRs is thought to involve the engagement of tworegions of the barrestin molecule with two corresponding re-gions of the receptor (2, 21). A large region within the amino-terminal half of barrestin (termed the activation-recognitiondomain) appears to bind the third loop of agonist-activatedGPCRs (32). A positively charged smaller domain (termed thephosphorylation-recognition domain) appears to interact withGRK-phosphorylated residues in the GPCR carboxyl terminus(33). Thus, although barrestins may bind intracellular loops of

agonist-activated receptors that are not phosphorylated, theaffinity of this interaction is presumably weak. GRK phospho-rylation and the engagement of both barrestin recognition do-mains with the receptor appear to be necessary for a highaffinity interaction to take place.

From the data presented in Fig. 2, we observed that trans-location of barrestin1 to the b2AR was delayed in onset andslower in rate compared with barrestin2 translocation. Re-analysis of this data using the model described under “Exper-imental Procedures” indicates that these qualitative differ-ences correspond to an approximately 10-fold higher affinity ofbarrestin2 for the b2AR relative to barrestin1 (Fig. 11B). Incontrast to these findings in cells, in vitro studies have reportedthat barrestin1 has a 2.5-fold greater affinity for the b2AR thanbarrestin2 (21). The discrepancy may arise from differences inthe manner in which barrestin1 and barrestin2 interact intra-cellularly with the GRK-phosphorylated b2AR carboxyl-termi-nal tail. The b2AR tail contains 13 putative phosphate acceptorsites, and barrestin1 binding in vivo may require that either agreater number of sites or different residues be phosphoryl-ated. Moreover, proteins other than GRKs apparently influencethe interaction between barrestins and GPCRs. The barrestincarboxyl-terminal domain contains AP-2 and clathrin-bindingsites (16, 30). A stronger interaction between barrestin2 andthese endocytic proteins may explain why swapping the bar-restin1 and barrestin2 carboxyl-terminal domains increasesthe fraction of the barrestin1 chimera and decreases the frac-tion of the barrestin2 chimera that translocate to the b2AR atthe plasma membrane.

The difference in the ability barrestin1 and barrestin2 tointeract in cells with the b2AR is typical of class A receptors.However, receptors belonging to class B, such as the AT1AR,appear to bind barrestin1 and barrestin2 equally well (Figs. 6Cand 11C). Two lines of evidence suggest that their affinity forbarrestins is greater than that possessed by class A receptors.First, AT1ARs and barrestins form stable endocytic complexes,whereas the b2ARzbarrestin complexes dissociate at or near theplasma membrane. The ability of barrestins to internalize withthe AT1AR into endocytic vesicles may indicate that this re-ceptor internalizes in a barrestin-dependent pathway. Consist-ent with this possibility, overexpression of barrestin1 potenti-ates AT1AR sequestration in HEK-293 cells (24). Second,visual arrestin translocates to class B but not class A receptors.The ability of class B receptors, such as the AT1AR and V2R, tobind arrestin family members with such high affinity appearsto be mediated entirely by clusters of phosphorylated residuesin the receptor carboxyl-terminal tail (18, 22). Similar clustersare noticeably absent from the tails of class A receptors. Thus,it is not unexpected that switching tails between class A andclass B receptors completely reverses their affinity for both thevisual and nonvisual arrestins.

Differences in barrestin affinities for class A receptors sug-gest that the tissue complement of barrestin1 and barrestin2may provide an additional degree of receptor regulation. bAr-restin1 is more abundant than barrestin2 in many bovinetissues including the heart, lung, and spleen and is the pre-dominant isoform in rat respiratory epithelial cells (6, 34).Conversely, barrestin2 is more abundant than barrestin1 inthe rat central nervous system, olfactory epithelial cells, andspermatozoa (5, 34). bArrestin2 was also found to be 3–5-foldmore abundant than barrestin1 in a variety of transformed celllines (35). Since barrestin1 and barrestin2 bind class A recep-tors with different affinities, signaling of class A receptors willbe more affected by relative changes in the ratio of barrestin1and barrestin2 than class B receptors. An important challengefor future research will be to determine the relationship be-

FIG. 11. Modeling of arrestin translocation and desensitiza-tion of agonist-activated GPCRs. A, schematic depicting the homol-ogous desensitization of GPCRs in the presence of saturating concen-trations of agonist (a). The kinetics of arrestin binding tophosphorylated membrane receptors, resulting in their homologousdesensitization, is mathematically described under “Experimental Pro-cedures.” B and C, time course of barrestin1 and barrestin2 disappear-ance from the cytoplasm after agonist activation and phosphorylation ofthe b2AR (B) and AT1AR (C). Data from Fig. 2 (b2AR) and Fig. 6(AT1AR) were re-analyzed using the relationship in Equation 10 (“Ex-perimental Procedures”). The kinetic parameter k1 (corresponding toGRK phosphorylation) was set at 0.02 for the b2AR and 0.025 for theAT1AR (23). The rate constants k2RT and k3 (corresponding to barres-tin1 and barrestin2 association and dissociation, respectively) weredetermined by nonlinear regression analysis. For the b2AR, k2RT and k3values were 0.014 and 0.034 for barrestin1 and 0.195 and 0.114 forbarrestin2. For the AT1AR, k2RT and k3 values were 0.058 and 0.038 forbarrestin1 and 0.083 and 0.056 for barrestin2.

Visualization of Arrestin Translocation to GPCRs 17209

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from

tween GPCR behavior and the absolute and relative levels ofbarrestin1 and barrestin2 in normal cells and in diseasedtissues.

The ability of visual arrestin to interact with class B recep-tors suggests that it may play an unappreciated role regulatingthe signaling of GPCRs other than rhodopsin. Visual arrestin isexpressed primarily in rod photoreceptor cells of the retina, butit is also found in other brain regions such as the pineal glandand in peripheral tissues including the heart, kidney, and lung(3, 4). In contrast to barrestin1 and barrestin2, visual arrestinis unlikely to promote receptor sequestration as it does not bindclathrin and interacts very poorly with AP-22 (17). Moreover,we observe that visual arrestin does not appear to target classB receptors to clathrin-coated pits. Therefore, visual arrestinby desensitizing class B receptors without promoting internal-ization may provide a more permanent scaffold for membranesignaling events and an endocytic independent mechanism toresensitize receptors more rapidly than either barrestin1 orbarrestin2 (36–39).

An unanticipated finding of these studies is that the agonist-independent subcellular distributions of the arrestin isoformsdiffer. Visual arrestin and barrestin1 are distributed in boththe cell cytoplasm and nucleus, whereas barrestin2 is found inthe cytoplasm but is excluded from the nucleus. The amino acidresidues mediating arrestin compartmentalization appear toreside in the arrestin carboxyl terminus. Switching the barres-tin1 and barrestin2 carboxyl-terminal domains completely re-verses their subcellular distributions. Notably, reports haveshown that functional prostaglandin E2 receptors reside on thenuclear envelope and other GPCRs, including muscarinic andangiotensin receptors, may be there as well (40–43). Thus, thenuclear pools of visual arrestin and barrestin1 may specificallyregulate nuclear GPCR signaling.

In summary, we have shown in real time and in live cellsthat specificity exists between arrestin family members andGPCRs. Therefore, these findings provide a basis for predictingthe arrestin-mediated pattern of GPCR desensitization, se-questration, and resensitization. Moreover, they explain whythe regulation of certain GPCRs will be more susceptible tovariations in the cellular complement of arrestin isoforms.

REFERENCES

1. Ferguson, S. S., Barak, L. S., Zhang, J., and Caron, M. G. (1996) Can.J. Physiol. Pharmacol. 74, 1095–1110

2. Krupnick, J. G., and Benovic, J. L. (1998) Annu. Rev. Pharmacol. Toxicol. 38,289–319

3. Breitman, M. L., Tsuda, M., Usukura, J., Kikuchi, T., Zucconi, A., Khoo, W.,and Shinohara, T. (1991) J. Biol. Chem. 266, 15505–15510

4. Smith, W. C., Milam, A. H., Dugger, D., Arendt, A., Hargrave, P. A., andPalczewski, K. (1994) J. Biol. Chem. 269, 15407–15410

5. Attramadal, H., Arriza, J. L., Aoki, C., Dawson, T. M., Codina, J., Kwatra,M. M., Snyder, S. H., Caron, M. G., and Lefkowitz, R. J. (1992) J. Biol.Chem. 267, 17882–17890

6. Sterne-Marr, R., Gurevich, V. V., Goldsmith, P., Bodine, R. C., Sanders, C.,Donoso, L. A., and Benovic, J. L. (1993) J. Biol. Chem. 268, 15640–15648

7. Lohse, M. J., Benovic, J. L., Codina, J., Caron, M. G., and Lefkowitz, R. J.

(1990) Science 248, 1547–15508. Lohse, M. J., Andexinger, S., Pitcher, J., Trukawinski, S., Codina, J., Faure,

J. P., Caron, M. G., and Lefkowitz, R. J. (1992) J. Biol. Chem. 267,8558–8564

9. Gurevich, V. V., and Benovic, J. L. (1992) J. Biol. Chem. 267, 21919–2192310. Wilden, U., Hall, S. W., and Kuhn, H. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,

1174–117811. Pippig, S., Andexinger, S., Daniel, K., Puzicha, M., Caron, M. G., Lefkowitz,

R. J., and Lohse, M. J. (1993) J. Biol. Chem. 268, 3201–320812. Ferguson, S. S., Downey, W. E., III, Colapietro, A. M., Barak, L. S., Menard, L.,

and Caron, M. G. (1996) Science 271, 363–36613. Zhang, J., Barak, L. S., Winkler, K. E., Caron, M. G., and Ferguson, S. S.

(1997) J. Biol. Chem. 272, 27005–2701414. Yu, S. S., Lefkowitz, R. J., and Hausdorff, W. P. (1993) J. Biol. Chem. 268,

337–34115. Pippig, S., Andexinger, S., and Lohse, M. J. (1995) Mol. Pharmacol. 47,

666–67616. Laporte, S. A., Oakley, R. H., Zhang, J., Holt, J. A., Ferguson, S. S., Caron,

M. G., and Barak, L. S. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 3712–371717. Goodman, O. B., Jr., Krupnick, J. G., Santini, F., Gurevich, V. V., Penn, R. B.,

Gagnon, A. W., Keen, J. H., and Benovic, J. L. (1996) Nature 383, 447–45018. Oakley, R. H., Laporte, S. A., Holt, J. A., Barak, L. S., and Caron, M. G. (1999)

J. Biol. Chem. 274, 32248–3225719. Barak, L. S., Ferguson, S. S., Zhang, J., and Caron, M. G. (1997) J. Biol. Chem.

272, 27497–2750020. Gurevich, V. V., Richardson, R. M., Kim, C. M., Hosey, M. M., and Benovic,

J. L. (1993) J. Biol. Chem. 268, 16879–1688221. Gurevich, V. V., Dion, S. B., Onorato, J. J., Ptasienski, J., Kim, C. M., Sterne-

Marr, R., Hosey, M. M., and Benovic, J. L. (1995) J. Biol. Chem. 270,720–731

22. Zhang, J., Barak, L. S., Anborgh, P. H., Laporte, S. A., Caron, M. G., andFerguson, S. S. (1999) J. Biol. Chem. 274, 10999–11006

23. Ferguson, S. S., Menard, L., Barak, L. S., Koch, W. J., Colapietro, A. M., andCaron, M. G. (1995) J. Biol. Chem. 270, 24782–24789

24. Zhang, J., Ferguson, S. S. G., Barak, L. S., Menard, L., and Caron, M. G. (1996)J. Biol. Chem. 271, 18302–18305

25. Dolph, P. J., Ranganathan, R., Colley, N. J., Hardy, R. W., Socolich, M., andZuker, C. S. (1993) Science 260, 1910–1916

26. Ungerer, M., Parruti, G., Bohm, M., Puzicha, M., DeBlasi, A., Erdmann, E.,and Lohse, M. J. (1994) Circ. Res. 74, 206–213

27. Oancea, E., and Meyer, T. (1998) Cell 95, 307–31828. Barak, L. S., Warabi, K., Feng, X., Caron, M. G., and Kwatra, M. M. (1999)

J. Biol. Chem. 274, 7565–756929. Lin, F. T., Krueger, K. M., Kendall, H. E., Daaka, Y., Fredericks, Z. L., Pitcher,

J. A., and Lefkowitz, R. J. (1997) J. Biol. Chem. 272, 31051–3105730. Krupnick, J. G., Goodman, O. B., Jr., Keen, J. H., and Benovic, J. L. (1997)

J. Biol. Chem. 272, 15011–1501631. McConalogue, K., Dery, O., Lovett, M., Wong, H., Walsh, J. H., Grady, E. F.,

and Bunnett, N. W. (1999) J. Biol. Chem. 274, 16257–1626832. Wu, G., Krupnick, J. G., Benovic, J. L., and Lanier, S. M. (1997) J. Biol. Chem.

272, 17836–1784233. Kieselbach, T., Irrgang, K. D., and Ruppel, H. (1994) Eur. J. Biochem. 226,

87–9734. Dawson, T. M., Arriza, J. L., Jaworsky, D. E., Borisy, F. F., Attramadal, H.,

Lefkowitz, R. J., and Ronnett, G. V. (1993) Science 259, 825–82935. Menard, L., Ferguson, S. S., Zhang, J., Lin, F. T., Lefkowitz, R. J., Caron,

M. G., and Barak, L. S. (1997) Mol. Pharmacol. 51, 800–80836. Daaka, Y., Luttrell, L. M., Ahn, S., Della Rocca, G. J., Ferguson, S. S., Caron,

M. G., and Lefkowitz, R. J. (1998) J. Biol. Chem. 273, 685–68837. Luttrell, 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., and Lefkowitz, R. J. (1999) Science 283, 655–661

38. Palczewski, K., McDowell, J. H., Jakes, S., Ingebritsen, T. S., and Hargrave,P. A. (1989) J. Biol. Chem. 264, 15770–15773

39. Byk, T., Bar-Yaacov, M., Doza, Y. N., Minke, B., and Selinger, Z. (1993) Proc.Natl. Acad. Sci. U. S. A. 90, 1907–1911

40. Bhattacharya, M., Peri, K. G., Almazan, G., Ribeiro-da-Silva, A., Shichi, H.,Durocher, Y., Abramovitz, M., Hou, X., Varma, D. R., and Chemtob, S.(1998) Proc. Natl. Acad. Sci. U. S. A. 95, 15792–15797

41. Bhattacharya, M., Peri, K., Ribeiro-da-Silva, A., Almazan, G., Shichi, H., Hou,X., Varma, D. R., and Chemtob, S. (1999) J. Biol. Chem. 274, 15719–15724

42. Lind, G. J., and Cavanagh, H. D. (1993) Invest. Ophthalmol. & Visual Sci. 34,2943–2952

43. Booz, G. W., Conrad, K. M., Hess, A. L., Singer, H. A., and Baker, K. M. (1992)Endocrinology 130, 3641–36492 S. A. Laporte, personal communication.

Visualization of Arrestin Translocation to GPCRs17210

by Tom

maso C

osta on August 22, 2008

ww

w.jbc.org

Dow

nloaded from