Phospho-selective mechanisms of arrestin

Full text

Turn on search term navigation

ARTICLE

Received 13 May 2015 | Accepted 28 Jul 2015 | Published 8 Sep 2015

Fan Yang1,2,3, Xiao Yu2,3, Chuan Liu2, Chang-Xiu Qu2, Zheng Gong2, Hong-Da Liu2, Fa-Hui Li1, Hong-Mei Wang3, Dong-Fang He2,3, Fan Yi4, Chen Song5, Chang-Lin Tian6, Kun-Hong Xiao7,8, Jiang-Yun Wang1 & Jin-Peng Sun2,3

Specic arrestin conformations are coupled to distinct downstream effectors, which underlie the functions of many G-protein-coupled receptors (GPCRs). Here, using unnatural amino acid incorporation and uorine-19 nuclear magnetic resonance (19F-NMR) spectroscopy, we demonstrate that distinct receptor phospho-barcodes are translated to specic b-arrestin-1 conformations and direct selective signalling. With its phosphate-binding concave surface,

b-arrestin-1 reads the message in the receptor phospho-C-tails and distinct phospho-interaction patterns are revealed by 19F-NMR. Whereas all functional phosphopeptides interact with a common phosphate binding site and induce the movements of nger and middle loops, different phospho-interaction patterns induce distinct structural states of

b-arrestin-1 that are coupled to distinct arrestin functions. Only clathrin recognizes and stabilizes GRK2-specic b-arrestin-1 conformations. The identied receptor-phospho-selective mechanism for arrestin conformation and the spacing of the multiple phosphate-binding sites in the arrestin enable arrestin to recognize plethora phosphorylation states of numerous GPCRs, contributing to the functional diversity of receptors.

1 Laboratory of Quantum Biophysics and Laboratory of RNA Biology, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Chaoyang District, 100101, China. 2 Key Laboratory Experimental Teratology of the Ministry of Education and Department of Biochemistry and Molecular Biology, Shandong University School of Medicine, 44 Wenhua Xi Road, Jinan, Shandong 250012, China. 3 Department of Physiology, Shandong University School of Medicine, Jinan, Shandong 250012, China. 4 Department of Pharmacology, Shandong University School of Medicine, Jinan, Shandong 250012, China. 5 Department of Biochemistry, University of Oxford, Oxford OX13QU, UK. 6 Hefei National Laboratory for Physical Science at Microscale and School of Life Science, University of Science and Technology of China, Hefei, Anhui 230027, China. 7 Department of Pharmacology and Chemical Biology, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania 15261, USA. 8 Department of Medicine, School of Medicine, Duke University, Durham, North Carolina 27705, USA. Correspondence and requests for materials should be addressed to J.-Y.W. (email: mailto:[email protected]

Web End [email protected] ) or to J.-P.S. (email: mailto:[email protected]

Web End [email protected] ).

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 1

DOI: 10.1038/ncomms9202 OPEN

Phospho-selective mechanisms of arrestin conformations and functions revealed by unnatural amino acid incorporation and 19F-NMR

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202

G-protein-coupled receptors (GPCRs) convert extracellular stimuli to intracellular signalling cascades primarily through G proteins or arrestin-mediated pathways14.

G proteins transduce signals by regulating the levels of second messengers, whereas arrestins recruit distinct downstream proteins to either desensitize receptors or initiate their own signalling pathways510. Recently, signicant conformational changes in arrestin have been observed following specic phosphopeptide binding or the formation of a receptor/arrestin complex. For example, the crystal structure of the V2-vasopressin receptor carboxy-terminalphosphopeptide (V2Rphosphopeptide (V2Rpp))/b-arrestin-1 complex revealed that the binding of

V2Rpp induced the rotation of the amino domain of b-arrestin-1 with respect to its C-terminal domain11. In another study, data obtained by electron microscopy (EM) and hydrogendeuterium exchange mass spectrometry studies revealed increased dynamics in both the N- and C-terminal domains of b-arrestin-1 after b2-adrenergic receptor (b2AR)/b-arrestin-1 complex formation2. These results suggest that the structural plasticity of b-arrestins underlies their important cellular functions.

Arrestins are multi-functional proteins7,12. Previous studies have indicated that two distinct features of these proteinsligand-induced receptor conformation and receptor phosphorylation barcodescontribute to the specic arrestin conformations that dictate selected arrestin functions5,13. Questions regarding these elements are core issues in the study of signal transduction by GPCRs, in particular given the plethora of phosphorylation states and receptor conformations of numerous receptors5,1417. However, the precise mechanism by which arrestin conformation is determined based on either a ligand-induced receptor-specic conformation or a selective phospho-barcode remains uncertain. Moreover, various receptors have no dened phosphorylation sequence information that correlates with their distinct arrestin-mediated functions, despite the existence of a myriad of evidence supporting the essential roles of phosphorylation and of negatively charged residues in the cytoplasmic regions of receptors in arrestin-mediated receptor endocytosis and other functions14,1618. These ndings raise the question of whether specic phospho-barcodes exist to direct barcode-selective arrestin functions. If such barcodes exist, then the method by which they are decoded by arrestins and translated into particular arrestin conformations remains unknown. The structural heterogeneity and exibility of active arrestins have hampered the characterization of arrestin conformations by crystallography or EM, and all active arrestin conformations determined to date have been obtained by stabilizing arrestin complexes with conformationally selective antibodies2,11. Therefore, it is desirable to develop alternative approaches to detect conformational changes in arrestins and decipher the phospho-selective mechanisms underlying distinct arrestin functions.

Recently, site-directed uorine-19 nuclear magnetic resonance (19F-NMR) spectroscopy has been used as a powerful approach for characterizing the dynamic conformational changes of large signalling protein complexes or membrane proteins1. In addition, we have developed an efcient method for incorporating the unnatural amino acid 3,5-diuorotyrosine (F2Y) into proteins by expanding the genetic code of Escherichia coli19,20. In the present study, by using this method to incorporate F2Y at specic locations in b-arrestin-1, we demonstrated that b-arrestin-1 reads receptor phospho-C-tail messages with its phosphate-binding concave surface, which harbours at least ten potential phosphate-binding sites (numbered 17 according to the binding mode of V2Rpp to b-arrestin-1 in the complex structure of V2Rpp/ b-arrestin-1, with A1A3 indicating additional phosphate-binding sites). Although all functional phosphopeptides interact

with phosphate-binding site 1 of b-arrestin-1, different GRKphosphopeptides (GRKpps) interact with b-arrestin-1 through distinct phospho-interaction patterns that are coupled to selective cellular functions. Whereas GRK2phosphopeptides (GRK2pps) interact with b-arrestin-1 within a phospho-binding sequence of the 1-4-6-7 pattern and promote clathrin recruitment, the GRK6phosphopeptide (GRK6pp) interacts with b-arrestin-1 through a different 15 sequence and provides the SRC signalling order. Moreover, 19F-NMR spectra provide direct biophysical evidence for the presence of at least three unique active b-arrestin-1 structural states corresponding to different phospho-binding patterns and only one of these structural states is recognized by clathrin. In cells, the mutation of key residues in specic phosphate-binding sites selectively abolished different b-arrestin-1 functions downstream of several receptors, suggesting that the identied phospho-pattern-selective mechanism of b-arrestin-1 conformation and function might be common to at least a subset of GPCR family members. Our results provide mechanistic insight into how the structural plasticity of b-arrestin-1 enables its recognition of specic phospho-patterns and translates these phospho-messages into distinct structural rearrangements and functional outcomes.

ResultsThe development of 19F-NMR probes of b-arrestin-1. The seven phosphate-binding sites revealed from the crystal structure of the V2Rpp/b-arrestin-1 complex allowed the identication of at least 127 potential phosphate-interacting patterns. It is useful to develop phosphate-binding sensors to reveal how b-arrestins ne-tune the phospho-pattern from a large number of receptors11. Accordingly, we selected eight positions that covered all phosphate-binding sites in the crystal structure of the V2Rpp/ b-arrestin-1 complex at which to incorporate F2Y (Fig. 1ac and

Supplementary Figs 14). The 19F-NMR spectra revealed two types of conformational states after V2Rpp binding: a slow exchange state (Y21) and a fast exchange state (Y63 and all other F2Y-incorporated positions) (Fig. 2a,b and Supplementary Tables 14). Paramagnetic titration experiments suggested that V2Rpp binding induced a structural state with greater solvent accessibility at the Y21 position (Supplementary Fig. 5). The phosphopeptide concentration dependence of the chemical shift of F2Y-Y63 and the increased area of the upeld shift peak of Y21-F2Y are consistent with concentration-dependent clathrin recruitment in the glutathione S-transferase (GST) binding assay, indicating the existence of a good structurefunction relationship (Supplementary Fig. 6 and Supplementary Table 3). In particular, all F2Y positions that directly interacted with phosphate displayed an upeld shift after V2Rpp binding (Fig. 2a and Supplementary Tables 1 and 2). By contrast, the binding of V2Rpp to V8-F2Y, which localizes to the b-strand I of arrestin and forms hydrophobic interactions with L365 of the V2Rpp, displayed a downeld shift in the 19F-NMR spectra (Fig. 2a). In addition, the binding of V2R5p to b-arrestin-1 caused an upeld shift of the 19F-NMR spectra at K107-F2Y but not at Y63-F2Y (Fig. 2c and Supplementary Table 2). Therefore, these F2Y probes may serve as phospho-pattern detectors to monitor the increasing negative charges after the binding of phosphorylated receptor or nucleic acid molecules21,22.

To study phosphopeptide-induced arrestin conformational changes, we next chose residues for F2Y incorporation that do not directly interact with V2Rpp (Fig. 1b,c and Supplementary Figs 7 and 8). Active crystal structures of the V2Rpp/b-arrestin-1 complex and a rod-arrestin splicing variant, p44, were compared with their inactive structural partners, and the results revealed a common pattern of differences11,23 (Fig. 2d). In addition to

2 NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202 ARTICLE

kDa M Y63F2Y

(63)F2YGREDLDVLGLTFR

0 200 400 600 800 1,000 1,200 1,400 1,600 m/z

63 48

F2Y

mRNA

F2Y +

y y y

b b

b y

y b b b b b

AUC

GCU UAG

CGA

AUC

F2YtRNA

F2Y RS

F2Y

Ribosome

T136 K138 F75

N375

F75

Y249

L338 1Y63

Y63

1Y63

T136 K138

K294

T13

T 6 K138

K294

K11 K294

R1665

R165

3 44

3 4

Y249

F277

L338

Y249

F277

L338

R165

F277

Y21 R7

Y209 Y209

Y21

K107

K11

Y21

K107

K11

K107

Y209

V2Rpp

N375

N37

N375

Figure 1 | The development of 19F-NMR probes via the incorporation of F2Y at specic positions of b-arrestin-1. (a) Schematic owchart of the incorporation of F2Y into b-arrestin-1. For example, the b-arrestin-1Y63-F2Y protein was acquired by co-transfection of b-arrestin-1 mutant plasmid and the pEVOL-F2YRS plasmid encoding specic Methanocaldococcus jannaschii tyrosyl amber suppressor transfer RNA/tyrosyl-tRNA synthase mutants with

F2Y in the culture medium. The purity of the protein was determined by electrophoresis (middle panel). The puried protein was subjected to trypsin digestion and analysed by MS/MS spectroscopy, which indicates the presence of the F2Y-G-R fragment, MW 413, b31, F2Y-G-R-E-D, MW 657 and b5 1, for example. These results conrmed that F2Y was incorporated into b-arrestin-1 at Y63. m/z, mass/charge ratio. (b) Frontal view of the F2Y incorporation sites in the inactive b-arrestin-1 crystal structure (PDB:1G4M, yellow). F2Y incorporation sites as phospho-interaction probes (R7, K11, Y21, Y63, K107,

K138, R165 and K294) that directly interact with V2Rpp are shown in red. F75 in the nger loop, T136 in the middle loop, F277 in the lariat loop, Y209 between b strands XIII and XIV, Y249 between b strands XV and XVI, L338 in the splice loop and N375 adjacent to the classic clathrin-binding box are shown in cyan. The dashed lines depict the structural portion not resolved by crystallography. (c) F2Y incorporation sites in the active b-arrestin-1 structure (PDB: 4JQI). The bound V2Rpp is shown in magenta. The seven phosphate-binding sites are indicated by green circles.

selecting residues that might exhibit broad conformational changes after activation, we attempted to study positions that are naturally occupied by Y or F to minimize the effects of mutation. All of the substitutions were tested using cellular and in vitro analyses to ensure that arrestin functions were not affected (Supplementary Figs 4 and 912). The 19F-NMR spectra of b-arrestin-1 alone at the Y209-F2Y position revealed a state of slow exchange between two peaks; these peaks were reduced to a single peak after V2Rpp binding (Fig. 2e). Furthermore, the amplitude of the 19F-NMR chemical shifts induced by V2Rpp binding at the F2Y-incorporated sites increased in the order F277, Y209, Y249, F75 and T136, and these increases were proportional to the increased root mean square deviation differences between the crystal structures of the b-arrestin-1/V2Rpp complex and b-arrestin-1 alone at the corresponding positions (Fig. 2e,f and Supplementary Tables 1 and 3). Although the chemical shift and root mean square deviation are different variables, both increased together, reecting a substantial structural rearrangement at these positions following arrestin activation. Therefore, the insertion of F2Y at these positions can be used as a 19F-NMR probe to detect dynamic arrestin conformational changes.

GRK phosphorylation determines b-arrestin-1 function. The individual receptor seven-transmembrane core and specic phosphorylation patterns are two determinants of specic arrestin functions2,5,13. The GRK6-phosphorylated b2AR imparts b-arrestin-2 conformations that specically activate ERK2, whereas GRK2-phosphorylated b2AR exerts more pronounced effects on receptor internalization (Table 1)5,16,24,25. Although these studies demonstrated that the functions of b-arrestin-2 are dened by different GRKs in the cells, whether the conformation

and signalling of b-arrestin-1 are regulated by selective GRK-encoded receptor phosphorylation pattern has not been examined. Therefore, we generated a BRET sensor (Rluc-barrestin-1-YFP) to monitor the conformational change of b-arrestin-1 in cells. Stimulation with 10 mM isoproterenol (ISO) signicantly increased the BRET signal of Rluc-b-arrestin-1-YFP, indicating a conformational change of b-arrestin-1 after b2AR activation (Fig. 3a). However, the ISO-induced BRET signal was signicantly reduced by the knockdown of either GRK2 or GRK6, indicating important roles of GRK in mediating the b-arrestin-1 conformational change after receptor activation (Fig. 3a).

We next investigated whether receptor phosphorylation by different GRKs produced different signalling outcomes downstream of b-arrestin-1. Formation of the receptor/ b-arrestin-1/SRC (or SRC family member HCK, FGR or YES) ternary complexes has been demonstrated to regulate multiple cellular functions downstream of various receptors, including, but not limited to, anti-apoptotic effects, granule exocytosis, colorectal carcinoma cell migration, bacterial adhesion to endothelial cells and the proliferation of pancreatic b-cells2631.

Therefore, we inspected the roles of different GRKs in b-arrestin-1-mediated SRC activation. Whereas GRK6 knockdown signicantly decreased SRC activation downstream of b2AR activation, GRK2 knockdown increased SRC activation (Fig. 3b). Together with previous research, these experiments demonstrate that GRK2 and GRK6 encode distinct b-arrestin-1 conformations and functions in cells5,8,2830,32.

To further evaluate whether these distinct cellular outcomes regulated by different GRKs are due to the generation of specic phosphorylation information in the receptor C-tail that is decoded by b-arrestin-1, we synthesized specic GRK2- or

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 3

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202

a b c

Y63F2Y K138F2Y K11F2Y

K294F2Y R7F2Y K107F2Y

Y21F2Y V8F2Y

R165F2Y

Ratio (Protein:V2Rpp)

1:0

1:0.03

1:0.3

1:3

1:9

Y63F2Y

K107F2Y

No peptide

V2R5p

No peptide

V2R5p

No peptide

V2Rpp

No peptide

V2Rpp

Y21F2Y Y63F2Y

130

140

130 140 130 140 130 140

130 140

130

140

130

140

130

140

130

140

130

140

130

140

130

140

19F chemical shift (p.p.m.)

d e f

F75F2Y T136F2Y Y209F2Y

Finger loop

F75 Y249

Middle loop

Lariat loop

-arrestin1

Rod

01 15 510 >10

No peptide

V2Rpp

T136

Residue heavy atom RMSD,

16 14 12 108

64 2

16 14 12 108

64 20

Y209

Y21

F277

130 140 130 140 130 140

Y249F2Y

Residue heavy atom RMSD,

F277F2Y L338F2Y N375F2Y

F75

T136

V139

No peptide

V2Rpp

Y249

F277

0.5

Y209

Y254

F79

L283

0 1

Residues position

130

140

130

140

130

140

19F chemical shift (p.p.m.)

Figure 2 | Application of 19F-NMR probes of b-arrestin-1 allowed the detection of phosphate binding and structural changes by 19F-NMR spectroscopy. (a) Upeld shifts in the 19F-NMR spectra of F2Y-phospho-sensing-probes of b-arrestin-1 are observed in response to V2Rpp binding.

The chemical shifts are referenced to triuoroacetic acid (TFA, 76.5 ppm) as an internal standard. (b) 19F-NMR spectra of b-arrestin-1Y21-F2Y and

b-arrestin-1Y63-F2Y after titration with V2Rpp. Left: two distinct peaks were observed for Y21 after titration with V2Rpp. The peak at 135.216 p.p.m.

increased (representing the V2Rpp-bound state) as the peak at 130.181 p.p.m. (representing the state of b-arrestin-1 alone) decreased, indicating a slow

conformational exchange. Right: a single peak was observed for Y63; the chemical shift varied in response to differences in the concentration of the V2Rpp, suggesting that rapid conformational changes occur at this site after V2Rpp binding. (c) Binding of V2R5p, which lacks the rst three phosphates in V2Rpp (pT347, pS350 and pS357), caused an upeld shift at the K107-F2Y position but no detectable change at the Y63-F2Y position (Dp.p.m.o0.05). (d) Plots of the distance root mean square deviations (RMSDs) for individual residues between inactive b-arrestin-1 and the V2Rpp-bound b-arrestin-1 complex (upper panel) and between inactive rod arrestin and its constitutively active p44 variant (lower panel). The vertical axis shows all heavy-atom RMSDs per arrestin residue. The colour code shown in the upper right corner of the panel indicates the Ca deviation of each residue. The three loops that exhibit major conformational changes are highlighted: nger loop, red; middle loop, light blue; and lariat loop, light yellow. (e) The 19F-NMR spectra of b-arrestin-1 incorporating site-directed F2Y revealed dynamic conformational changes after activation by V2Rpp. (f). Heavy-atom RMSD versus the relative amplitude of the 19F-NMR chemical shift for specic F2Y incorporation sites.

GRK6-phosphorylated b2AR C-tails and examined their effects on the biochemical properties of b-arrestin-1 (Table 1). The GRK2-phosphorylated b2AR C-tail were separated into two segments, GRK2App and GRK2Bpp, owing to difculties encountered with synthesis. Protein kinase A (PKA) also phosphorylates b2AR and we used the PKA-phosphorylated b2AR peptide (PKApp) as a control, because PKA does not affect arrestin recruitment32. In vitro, the V2Rpp and GRK2pp induce formation of the b-arrestin-1/clathrin complex, whereas the

GRK6pp and V2Rpp promote the formation of the b-arrestin-1/SRC complex (Fig. 3c,d). By contrast, the PKAphosphopeptide did not promote the formation of either the arrestin/clathrin complex or the arrestin/SRC complex (Fig. 3c,d). Taken together, these ndings reveal that the GRK6-mediated receptor phosphorylation pattern specically activates SRC, whereas the

GRK2-mediated receptor phosphorylation pattern selectively recruits clathrin, thus facilitating receptor endocytosis (Table 1).

Phospho-interaction patterns are deciphered by19F-NMR. V2Rpp, GRK2App, GRK2Bpp and GRK6pp have 11, 8, 9 and 4 phosphates or negatively charged residues, respectively. Specic phospho-binding patterns may correspond to distinct arrestin functions. We next used 19F-NMR spectroscopy of the F2Y phospho-sensors to detect the negative charge-sensing patterns of b-arrestin-1 after stimulation with specic phosphopeptides.

All of the GRKphosphopeptides, but not the PKA phosphopeptides, induced an upeld chemical shift at Y63 (phosphate-binding site 1) (Fig. 4a). Specically, GRK2App and GRK2Bpp caused upeld chemical shifts in 19F-NMR

4 NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202 ARTICLE

Table 1 | Summary of the signalling and biochemical properties of receptor phosphorylation barcodes encoded by various

kinases.

Phosphobarcode

Sequence Biochemical properties

Cellular functions

Clathrin binding

SRC binding

Desensitization

Arrestin conformational change

Internalization SRC activatin

Full- -V2R

V2Rpp:

343ARGpTPPpSLGPQDEpSCpTpTApSpSpSLAKDTSS371

ND ND

GRK2- -b2AR

GRK2Bpp:

357NGNpTGEQpSGYHVEQEKENKLLCEDLPGTE385 GRK2App:

386DFVGHQGTVPpSDNIDpSQGRNCpSTNDpSLL413

GRK6- -b2AR

GRK6pp:

343RRSIKAYGNGYpSpSNGNTGEQSGYHVEQ370 PKA- -b2AR

PKApp:

251DGRTGHGLRRpSpSKFCLKEHKALKTLGII280

, no signicant change; , signicant change but the change is less than twofold; and , signicant change with more than twofold or with more than threefold, respectively; ND, not

determined.

Control siRNA GRK2 siRNA GRK6 siRNA

0 1 5 15 0 1

1 5 5 15

15 0

0.50

Control siRNA GRK2 siRNA GRK6 siRNA

7 6 5 4 3 2 1 0

0 5 10 15

Time (min)

Min IB:p-SRC(416)

IB:T-SRC IB:GRK2

IB:GRK6

IB:GAPDH

ISO

0.45

p-SRC (416)

fold above basal

0.40

* *

0.35

0.30

ISO (10 M)

H89 (10 M)

siRNA

V2R GRK2A GRK2B GRK6 PKA

Control

- + + + + +

Control

GRK2

GRK6

c d

Peptide

SRC-3D

-Arrestin-1

IB: -Arrestin-1

IB: GST

Folds compared with control

4 *** ***

Peptide

Clathrin

IB: -Arrestin-1

IB: GST 0

V2R GRK2A GRK2B GRK6 PKA

+ + + + + +

Folds compared with control

***

*** ***

+ + + + + +

-Arrestin-1

0 GRK2A

V2R GRK2B

GRK6

PKA

GRK2A

V2R GRK2B

GRK6

PKA

Figure 3 | Different phospho-barcodes encode different arrestin functions and conformations. (a) Activation of b2-AR-induced conformational changes in b-arrestin-1 as measured using intramolecular BRET. Plasmids encoding a b-arrestin-1 BRET biosensor (Luc-b-arrestin-1-YFP) and Flag-b2-AR were transiently co-transfected in HEK293 cells. After starvation, the cells were stimulated with ISO (10 mM) for 10 min. The effects of GRK phosphorylation of b2-AR on ISO-induced b-arrestin-1 conformational change were probed using siRNA-mediated GRK knockdown. (b) A representative western blotting is shown for the effects of knockdown of specic GRKs in b2-AR-mediated proto-oncogene tyrosine-protein kinase (SRC) activation. Left panel: HEK293 cells that were transiently transfected with Flag-b2-AR were stimulated with ISO for 1, 5 and 15 min. To examine the roles of GRK2 and GRK6 in b2-AR-mediated

SRC activation, siRNAs that were selective for specic GRKs were transfected into the cells. A western blot representative of at least three independent experiments is shown on the left. Right panel: western blot signals of phospho-SRC 416 (a) were quantied using densitometry and are expressed as fold change versus basal level. Whereas GRK6 is required for SRC activation downstream of b2-AR activation, GRK2 is a negative regulator of b2-AR-induced

SRC activation. (a,b) At least three independent experiments were performed. #Po0.05, ISO-stimulated cells are compared with unstimulated cells. *Po0.05; **Po0.01; ***Po0.005; GRK siRNA-treated cells are compared with control siRNA-treated cells. (c,d) Effects of phospho-peptide binding on b-arrestin-1/clathrin complex formation (c) or b-arrestin-1/SRC-SH2-SH3-kinase-three domain (SRC3D) complex formation (d). b-arrestin-1 (300 nM) was incubated with an equal concentration of various phosphopeptides and either GST-clathrin 1494 (c) or GST-SRC3D (d). The complexes were pulled down using GST beads and the amount of b-arrestin-1 bound to clathrin or SRC3D was monitored using a specic b-arrestin-1 antibody (left panel in each gure). At least three independent experiments for each assay were performed. Western blot signals of b-arrestin-1 bound to clathrin (c) or SRC3D (d) from the left panels were quantied using densitometry and are shown as columns (right panel). **Po0.01; ***Po0.005; phosphopeptide stimulated b-arrestin-1/clathrin complex formation (c) or b-arrestin-1/SRC complex formation (d) are compared with the basal state.

spectroscopy at positions K11, K294, K107 and Y21, which surround phosphate-binding sites 4, 6 and 7, whereas GRK6pp binding specically induced an upeld chemical shift at R7, corresponding to phosphate-binding site 5 (Fig. 4a,b and Supplementary Table 2). A small amplitude of an upeld chemical shift was also observed at R165 after GRK6pp binding.

However, this chemical shift was not due to the binding of a phosphate or a negative charge residue at phosphate binding site 2 or 3, as the binding of GRK6pp caused no signicant chemical shift at either K11-F2Y, K138-F2Y or R62-F2Y (Fig. 4a and Supplementary Fig. 13 and Supplementary Table 2). Rather, the chemical shift at R165-F2Y after GRK6pp binding may reect a

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 5

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202

Y63F2Y Site 1

130 140

K138F2Y Site 2

R165F2Y

Site 2/3

K11F2Y

Site 3/4

K294F2Y Site 4

R7F2Y

Site 5

Y21F2Y Site 6

No peptide

GRK2App

GRK2Bpp

GRK6pp

V2Rpp

PKApp

130 140 130 140 130 140 130 140 130 140

130

140

K107F2Y Site 6/7

130 140

19F chemical shift (p.p.m.)

No peptide

GRK2App

GRK2Bpp

GRK6pp

Site 1 2 3 4 5 7

V2Rpp

GRK2App

GRK2Bpp

GRK6pp

PKApp

V2Rpp

Clathrin

PKApp

SRC

19F chemical shift (p.p.m.)

Y63F2Y K294F2Y K107F2Y

Y63F2Y

+0.89

R7F2Y

No peptide

GRK6pp

SEP355

Y63

+0.09

+0.01

+0.02

19F chemical shift (p.p.m.)

E369

Site 1

130 140

GRK6pp-pS355S

GRK6pp-E369A

Site 5

e f

No peptide

GRK2Bpp-E373A

GRK2Bpp-E379/D380A

TPO360

Site1

GRK2Bpp

+0.04

+0.61

+0.64 +0.09

130 140

+0.28

+0.01

Y63

K294

GRK2Bpp-pT360T

E373

E379

Site4

K107

D380

130 140 130 140 19F chemical shift (p.p.m.)

Site 6/Site 7

Figure 4 | The phospho-interaction pattern of b-arrestin-1 was deciphered via the 19F-NMR spectra. (a) Effects of different phosphopeptide binding on the 19F-NMR spectra on b-arrestin-1-F2Y phospho-sensing probes. (b) Specic phospho-binding patterns revealed by 19F-NMR spectra that correlate to their biochemical and cellular properties. The red balls indicate that a phosphate or a negatively charged residue interacts with the specic phospho-binding site localized in the b-arrestin-1 N terminus. Grey balls indicate that no signicant chemical shift increase (Dp.p.m.o0.05) was detected by the phospho-sensor around the specic site. The binding of the seven phosphate sites by V2Rpp promoted both clathrin and SRC binding. Two GRK2pps displayed similar phospho-binding patterns (1-4-6-7) and promoted clathrin binding but not SRC interaction. The phospho-binding pattern for GRK6pp is sites 1 and 5, which may correlate with its SRC function. (c) Effects of binding of GRK6pp mutants on the 19F-NMR spectra on b-arrestin-1-R7-F2Y and Y63-F2Y phospho-sensing probes. (e) Effects of binding of GRK2pp mutants on the 19F-NMR spectra on b-arrestin-1 phospho-sensing probes. (d,f) Structural representation of a model of specic interactions between the negative charged GRK6pp/GRK2pp residue and the residue localized in phospho-binding sites of b-arrestin-1. The model was generated by application of the PI-LZerD algorithm using V2Rpp/b-arrestin-1 complex structure (PDB: 4JQI) as a template and adjusted by results from 19F-NMR spectra.

6 NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202 ARTICLE

conformational change at R165 or an interaction with R165 by a negatively charged residue with an unknown manner. A simplied phospho-interaction model of different phosphopeptides and their relations to arrestin functions are summarized in Fig. 4b.

To further characterize how these phosphopeptides interact with specic phosphate-binding sites of b-arrestin-1, we selectively mutated the negatively charged residues of these phosphopeptides to alanine and tested their effects on clathrin or SRC recruitment in vitro. Two mutations of GRK6pp and three mutations of GRK2pp affected phosphopeptide-induced b-arrestin-1/SRC or b-arrestin-1/clathrin interaction, respectively (Supplementary Figs 14 and 15). Importantly, the mutation of pS355 in GRK6pp selectively eliminated the upeld chemical shift at the Y63-F2Y position, whereas the mutation of E369A specically abolished the upeld chemical shift at the R7-F2Y position after phosphopeptide binding (Fig. 4c). Therefore, the pS355 of GRK6pp interacts with phosphate-binding site 1 and E369 interacts with phosphate-binding site 5 in the phosphate-binding concave surface of b-arrestin-1 (Fig. 4d and

Supplementary Table 7). Similarly, 19F-NMR spectroscopy with selective GRK2pp mutations indicated that pT360, E373 and E379/D380 specically interact with phosphate-binding sites 1, 4 and 6/7, respectively (Fig. 4e,f).

Phospho-barcodes induce structural changes in b-arrestin-1. The specic phosphopeptide/arrestin interaction pattern should trigger substantial arrestin conformational rearrangements. We next inspected the 19F-NMR spectrum of F2Y incorporation positions outside of the phosphopeptide interacting region. The binding of phosphate site 1 by GRKphosphopeptides may provide the driving force for the repositioning of the nger loop and middle loop, which is a prelude to receptor core binding2,33. All of the GRK-phosphorylated peptides consistently induced downeld chemical shifts at residue 75 in the nger loop and at residue 136 in the middle loop (Fig. 5a and Supplementary Table 5). Therefore, except for PKApp, the common conformational changes that occur in the nger loop (residue75) and middle loop (residue 136) regions that are induced in GRKphosphopeptide binding (phosphateY63 interaction) might enable the interaction with the receptor cores for fully phosphorylated receptor/arrestin associations. Another important common structural feature of GRKphosphopeptide binding is the downeld chemical shift at N375, which indicates the displacement of the C-tail of b-arrestin-1, despite the interaction of GRK2pp and GRK6pp with arrestin through different phosphate-binding sites in their C-terminal regions (sites 6 and 7 for GRK2pp and site 5 for GRK6pp; Figs 1b,c, 4b and 5a).

In particular, the interaction of GRK2pps (GRK2App and GRK2Bpp) with phosphate-binding site 4 enables the rotation of K294, releasing D290 and D297 from the polar core (Fig. 4a,b,e,f). The rearrangement of the polar core may lead to prominent twisting of the lariat loop, which facilitates the binding of the receptor transmembrane domain or clathrin. In addition, negatively charged residues or phosphates of both GRK2App and GRK2Bpp bind to phosphate sites 6 and 7, which disrupts the hydrophobic interactions between F388 and Y21/K107, resulting in the displacement of the C-tail and the conformational change at the amphipathic a-helix and the rst two b-strands of b-arrestin-1. By contrast, although the binding of GRK6pp may also partially release the C terminus of b-arrestin-1, as it signicantly increases the negative charge around R7 and causes a downeld shift at the Y21 position, it does not change the chemical shift at K294 or K107 (Fig. 4ad and Supplementary Table 2). Therefore, selective GRKphosphopeptide interaction patterns might result in distinct arrestin structural states.

Accordingly, GRK2pp specically induced a structural change at Y249 in the loop between b-strands XV and XVI, and

L338-F2Y in the splice loop (Fig. 5b and Supplementary Table 5). By contrast, the GRK6pp, but not the GRK2pp or PKAphosphopeptides, induced a specic structural change at F277-F2Y in the end of the lariat loop (Fig. 5c). Moreover, only V2Rpp induced a downeld chemical shift at the Y209-F2Y position in the 19F-NMR spectra, whereas all other tested phosphopeptides caused no signicant effect (Fig. 5d and Supplementary Table 6). This conformational change at Y209 may be correlated with the binding of V2Rpp to phosphate-binding sites 2 or 3 of b-arrestin-1, which are the only two phosphate-binding sites that are not covered by GRK or PKAphosphopeptides (Fig. 4b). Taken together, these ndings show that these phospho-barcode-induced structural rearrangements potentially enable the imparting of distinct functions of b-arrestin-1 through clathrin, SRC, ERK or other downstream effector proteins (Fig. 5e).

Clathrin-induced structural alterations in b-arrestin-1. The GRK2-encoded arrestin conformation recruits clathrin and promotes receptor endocytosis. Clathrin-mediated phospho-receptor/arrestin complex trafcking is proposed to involve two distinct regions, the classic clathrin binding (CCB) box

376LIELD380 and the eight amino acid splice loop34. All F2Y mutations in the CCB box impaired the formation of the phosphopeptide/arrestin/clathrin complex (Supplementary Fig. 11). Thus, we selected L338-F2Y in the splice loop and N375-F2Y adjacent to the CCB box, to monitor clathrin-induced arrestin conformational changes. The binding of clathrin to the activated b-arrestin-1 in the presence of V2Rpp or GRK2pp led to a substantial chemical shift of the 19F-NMR signals at Y249-F2Y, L338-F2Y and N375-F2Y (Fig. 6ac). Although GRK6pp also induced a structural change at N375-F2Y, clathrin binding caused no additional chemical shift (Fig. 6a). Moreover, the binding of clathrin to the phosphopeptide-occupied b-arrestin-1 did not affect the chemical shift at F277, a GRK6-specic structural site (Fig. 6d). These ndings suggest that clathrin recognizes GRK2-encoded specic arrestin conformations of the splice loop, the CCB box and the loop between b-strands XV and XVI;

however, clathrin does not recognize GRK6pp-encoded arrestin conformations. In particular, although V2Rpp and GRK2pp induce different amplitudes in the chemical shifts at L338-F2Y, Y249-F2Y and N375-F2Y, which suggests structural plasticity at these positions, the chemical shifts of the signals assigned to these positions in the ternary complexes of V2Rpp/arrestin/clathrin or GRK2pp/arrestin/clathrin exhibit no detectable differences (Fig. 6ac). These results indicate that clathrin recognizes different phosphopeptide/arrestin intermediates and stabilizes the ternary complex to a common structural state at these positions.

We next assessed the effect of the paramagnetic relaxation agent chromium acetylacetone (Cr) on the 19F-NMR spectra of F2Y-N375 of b-arrestin-1 in the presence of clathrin and different

GRKpp. The 19F-NMR spectra of F2Y-N375 revealed that the N375 position in the b-arrestin-1 alone or in the GRK2pp/ b-arrestin-1/clathrin ternary complex is less susceptible to line broadening by Cr than the GRK2pp/b-arrestin-1 complex or the GRK6pp/b-arrestin-1 complex plus clathrin, indicating that F2Y-N375 is less readily solvent accessible in b-arrestin-1 alone or the GRK2pp/b-arrestin-1/clathrin complex (Fig. 6e and Supplementary Fig. 16). Therefore, a model was built in which the C-terminal region of b-arrestin-1 was dislodged after GRKpp binding, which was protected by clathrin in the GRK2pp/

b-arrestin-1 complex but not in the GRK6pp/b-arrestin-1

complex (Fig. 6f).

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 7

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202

a b

F75F2Y T136F2Y Y249F2Y L338F2Y F277F2Y

N375F2Y

No peptide

GRK2App

GRK2Bpp

GRK6pp

No peptide

GRK2App

GRK2Bpp

GRK6pp

No peptide

GRK2App

GRK2Bpp

GRK6pp

V2Rpp

PKApp

130

140

130

140

130

140

130 140 130 140 130 140

19F chemical shift (p.p.m.)

Agonist

Y209F2Y

No peptide

GRK2App

GRK2Bpp

GRK6pp

V2Rpp

F75

PKApp

130 140

19F chemical shift (p.p.m.)

-Arrestin-1

Clathrin Src

Figure 5 | 19F-NMR spectra revealed the different structural states of b-arrestin-1 in response to different phospho-binding patterns. (a) Common structural changes that occur after b-arrestin-1 activation induced by the binding of functional phosphopeptides. These results suggest the existence of common conformational states representing b-arrestin-1 activation after functional phosphopeptide binding. (b) Specic structural rearrangements after

GRK2pp binding. (c) GRK6-encoded specic structural rearrangements at F277 position. (d) V2Rpp specically stimulated conformational change at Y209 position. (e) Proposed models illustrating the receptor phospho-barcode-encoded distinct arrestin conformations that dictate specic downstream signalling. The phosphorylated C-tail of the receptor (in orange) extends into the cytoplasm. Interaction with the receptor phospho-C-tail induces a structural rearrangement of b-arrestin-1 that includes residues in the nger loop (such as F75) and the middle loop (such as T136); these residues (in purple) might participate in the receptor transmemberane core interaction. Different GRKs encode different conformations in the C-lobe of b-arrestin-1.

GRK2 encodes specic conformations of Y249 and L338 (blue triangles), which are recognized by clathrin (in blue). GRK6 encodes distinct b-arrestin-1 conformations, such as the conformation of F277 (red triangle), which might be recognized by signalling proteins such as SRC (in red). Specic

V2Rpp-induced conformational change is localized at Y209 (in orange).

Figure 6 | 19F-NMR spectra revealed the structural alteration of b-arrestin-1 in response to the binding of Clathrin. Chemical shifts after phosphopeptide/b-arrestin-1/clathrin ternary complex formation at specic F2Y incorporation sites: at N375-F2Y close to the classic clathrin-binding box (a), at L388 in the splice loop (b) and at Y249-F2Y (c). (d) No signicant changes (Dp.p.m.o0.05) were observed for the chemical shifts at F277-F2Y position in the 19F-NMR spectra after clathrin binding. (e) Paramagnetic titration experiments. As the concentration of the Cr increased, the full width at half maximum of the 1D-19F-NMR spectra at the F2Y-N375 position of the b-arrestin-1/GRK2Bpp complex signicantly increased (Dp.p.m. 25 HZ),

followed by the b-arrestin-1/GRK6 complex in the presence of clathrin (Dp.p.m. 16 HZ), and then the b-arrestin-1/GRK2Bpp/clathrin ternary complex

(Dp.p.m. 5 HZ) and the b-arrestin-1 alone (Dp.p.m. 4 HZ), indicating a greater protective effect of Cr at the N375 position of the b-arrestin-1/GRK2Bpp

complex in the absence of clathrin. (f) A cartoon illustration of the phospho-pattern-encoded structural rearrangements of b-arrestin-1 and its subsequent stabilization by clathrin binding. The red ball indicates the phosphate. Although the binding of GRK2App, GRK2Bpp and V2Rpp all caused a partial dislodging of the C terminus of b-arrestin-1, the resulting structural states induced by different phosphopeptides could be different. Clathrin binding stabilized all these phosphopeptide-occupied arrestins to a single state.

8 NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202 ARTICLE

Phospho-decision mechanisms of b-arrestin-1 functions. To further characterize the functional outcomes of the specic phospho-interaction pattern and the accompanying structural rearrangements of b-arrestin-1, we generated systematic b-arrestin-1 mutants that covered all seven V2R phosphate-

binding sites and a partial lariat loop deletion construct. Mutations of all residues in phosphate-binding sites 2 or 3 had no signicant effects on clathrin or SRC binding. By contrast, destroying phosphate-binding site 1 by Y63/R65/K77-3A, which is essential for the nger loop rearrangement, eliminated both

N375F2Y

L338F2Y

+Clathrin

135.69 No shift

No shift

135.66

135.99

135.97

No shift

135.42

135.45

135.36

No peptide

V2Rpp

GRK2App

GRK2Bpp

GRK6pp

135.34

135.71

134.74

135.21

135.00

135.39

+ 0.68 shift

+ 0.21 shift

+ 0.45 shift

+Clathrin

134.93

135.85

+0.37 shift

+0.77 shift

0.14 shift

135.70

134.99

No peptide

V2Rpp

GRK2App

GRK2Bpp

GRK6pp

135.97

130 140

130 140 130 140 130 140

19F chemical shift (p.p.m.)

c d

134.49

134.59

134.58

134.57

134.80

134.17

134.02

134.07

134.79

Y249F2Y +Clathrin

+0.42 shift

0.31 shift

F277F2Y

No shift

+Clathrin

No peptide

V2Rpp

GRK2App

GRK2Bpp

GRK6pp

0.31 shift

+0.56 shift

+0.50 shift

135. 31

135.34

135.32

135.05

135.33

135.32

135.04

135. 31 No peptide

GRK2App

GRK2Bpp

GRK6pp

134.48

130 140 130 140 130 140

19F chemical shift (p.p.m.)

e Acetylacetone chromium

N375F2Y

N375F2Y+GRK2Bpp N375F2Y+GRK2Bpp+Clathrin N375F2Y+GRK6pp+Clathrin

62.25 HZ

68.92 HZ

74.80 HZ

69.87 HZ

56.45 HZ

49.53 HZ

69.42 HZ

68.52 HZ

65.51 HZ

64.01 HZ 52.85 HZ

58.35 HZ

5 mM

61.26 HZ

68.52 HZ

3 mM

58.45 HZ

1 mM

58.52 HZ

0 mM

135 136 135 136 135 136 135 136

19F chemical shift (p.p.m.)

f GRK2App

GRK2Bpp V2Rpp

-Arrestin-1 Multiple conformations Single conformation

+Clathrin

CCB box Clathrin

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 9

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202

Site 1

IB: -Arrestin-1 IB: GST

Peptide

IB: -Arrestin-1 IB: GST

Pull down : SRC

Site 3

Site 4

Site 5

Pull down : Clathrin Pull down : SRC

Peptide

IB: -Arrestin-1 IB: GST

Pull down : Clathrin

Peptide

IB: -Arrestin-1 IB: GST

K11A

K11/K160

R25A

K294A

R7A

Site 6-7

Pull down : Clathrin Pull down : SRC

K11/K160

R1653A

Pull down : Clathrin Pull down : SRC

Peptide

IB: -Arrestin-1 IB: GST

K10A

Y21/K107

K107A

Y63A

Y63/R65

K773A

Y21/K107

2AK10E

Peptide

IB: -Arrestin-1 IB: GST

Site 2

Pull down : Clathrin

Pull down : SRC

Peptide

IB: -Arrestin-1 IB: GST

R62A

R62/K160

Lariat loop deletion

Pull down : Clathrin Pull down : SRC

Peptide

R62/K160

R1653A

Peptide

Figure 7 | Specic phospho-sensors located in the N terminus of b-arrestin-1 determine its selective interaction with clathrin or SRC. Systematic Ala scanning mutations covering all seven phosphate-binding sites of b-arrestin-1 were examined for their effects on phosphopeptide-induced clathrin or SRC recruitment. The effects of phosphopeptide binding on b-arrestin-1/clathrin complex formation (left panel) and b-arrestin-1/SRC3D complex formation (right panel) were examined with a GST pull-down assay. Representative western blottings from at least three independent experiments for each assay are shown. The K10/Y21/K107-3A mutant was not expressed and therefore changed according to the combination of K21/K107-2A-K10E. Destroying phosphate-binding site 1 by Y63/R65/K77-3A abolished both phosphopeptide-induced clathrin and SRC recruitment (shown in magnate). Mutation in site 4 by R25A eliminated, and the mutation at site 6/7 by Y21/K107-2A/K10E, decreased phosphopeptide-induced clathrin binding but had no effect on SRC interaction (in green). Conversely, the mutation at site 5 by R7A and the partial deletion of the lariat loop decreased the SRC interaction but had no effect on phosphopeptide-induced clathrin binding (in red). All other mutations had no signicant effects on phosphopeptide-induced SRC or clathrin recruitment.

clathrin and SRC binding (Fig. 7 and Supplementary Fig. 17). More interestingly, the GRK2pp-stimulated clathrin binding was signicantly reduced by mutations of phosphate-binding sites 6 and 7 (Y21A/K107A/K10E) and was entirely eliminated by a single mutation of R25A in phosphate-binding site 4. However, these mutations have no signicant effects on the GRK6pp-induced SRC/b-arrestin-1 interaction. Furthermore, the interruption of the GRK6pp-specic interacting site by R7A signicantly reduced the phosphopeptide-induced SRC binding but had no noticeable effect on clathrin binding. Accordingly, partial deletion of the lariat loop close to the F277 position, which only exhibits structural rearrangement after GRK6pp binding, resulted in a decreased phosphopeptide-induced SRC interaction (Fig. 7 and Supplementary Fig. 17). These results indicated that phosphate-binding sites 4, 6 and 7 govern phosphopeptide-induced clathrin binding, whereas phosphate-binding site 5 denes the SRC interaction. The binding of phosphopeptides to phosphate-binding site 1 primes both SRC and clathrin recruitment.

Consistent with these biochemical results, the interruption of phosphate-binding site 4 by R25A selectively impaired b-arrestin-1/clathrin interaction downstream of b2AR activation, whereas destroying phosphate-binding site 5 specically eliminated ISO-induced b-arrestin-1/SRC association (Fig. 8a). A similar result for arrestin/partner interactions by specic phospho-interaction patterns was also observed in CCKAR for clathrin and in SSTR2 for clathrin and SRC (Fig. 8b,c). Therefore, this newly identied phospho-pattern decision mechanism also underlies at least a subset of GPCRs.

DiscussionUnderstanding the coding mechanisms and structural basis of arrestin-mediated GPCR signalling is currently a central issue in receptor biology that may have important implications for new drug development. In contrast to the tremendous effort

focused on understanding the functional signicance of the conformational heterogeneity of receptors1,3539, the importance of the structural plasticity of arrestin has only just begun to be analysed11,4042. Recent studies have suggested that ligandspecic receptor conformation is correlated with, and that the C-terminal phosphorylation pattern denes, specic arrestin conformations that dictate selective signalling pathways1,5,13. The marked conformational rearrangement of arrestin in the crystal structure of the b-arrestin-1/V2Rpp/Fab30 complex provides considerable insight into the structural properties of arrestin and further supports the idea that specic conformations underlie selective arrestin functions2. However, this crystal structure only presents a static picture for a single activated arrestin conformation. Certain important structural features that are highly relevant to arrestin functions, such as various loops and the nal 50 amino acids at the C terminus of arrestin, are absent from the structure of the b-arrestin-1/V2Rpp complex because of poor electron densities11. Moreover, the ways in which the specic structural states of arrestins are dened by specic phosphorylated cytoplasmic regions of a receptor or by ligand-occupied receptor conformations remain uncertain and correlations between specic arrestin conformations and distinct arrestin functions have not been established. In addition, the exible nature of the arrestin structure has hampered crystallographic analysis in the absence of stabilization by conformation-specic antibodies. In this context, we have developed a series of F2Y-based structural sensors in b-arrestin-1 through unnatural amino acid incorporation (Fig. 1ac). These sensors enabled us to use 19F-NMR spectroscopy to inspect phospho-interaction patterns and residue-specic structural information in b-arrestin in different active states (Fig. 2a,e).

Although there is no apparent phospho-sequence identity between GRK2App and GRK2Bpp, 19F-NMR spectra of F2Y incorporated into the phosphate-binding sites of b-arrestin-1 indicated that these phosphopeptides interact with b-arrestin-1

10 NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202 ARTICLE

b c

Clathrin fold over basal

pSrc(416) fold over basal

1.5

Clathrin fold over basal

2AR

CCKAR

SSTR2

IP : HA--Arrestin-1

IB: clathrin

IB: pSrc (416)

IB: HA

IB: 2AR Isoproterenol

WT R25A R7A

IP : HA--Arrestin-1 IP : HA--Arrestin-1

IB: Clathrin

IB: pSrc (416)

IB: HA

IB: FlagCCK8S

WT R25A R7A

0.5

IB: Clathrin

IB: pSrc(416)

IB: HA

IB: Flag SST

WT R25A R7A

0 WT R25A R7A

WT R25A R7A

pSrc(416) fold over basal

Cell lysate Cell lysate Cell lysate

0.5

0.4

0.3

0.2

+ + + + + + + + +

0.1

WT R25A R7A

Clathrin Src

Figure 8 | Phospho-pattern-selective mechanisms of specic arrestin functions are common for at least a subset of GPCRs. (a) The phospho-binding pattern of b-arrestin-1 determines selective arrestin functions downstream of b2AR in cells. HEK293 cells were co-transfected with Flag-b2AR and wild-type HA-b-arrestin-1 or specic phosphate-binding site mutants (R25A for site 4 and R7A for site 5), then stimulated with 10 mM isoproterenol (ISO)

for 15 min. The b-arrestin-1 was immunoprecipitated by HA-antibody-conjugated agarose and the formation of the b-arrestin-1/clathrin or b-arrestin-1/SRC was detected by specic clathrin or SRC antibodies. R25A or R7A selectively disrupted the ISO-induced b-arrestin-1/clathrin complex formation or the

ISO-induced b-arrestin-1/SRC complex formation, respectively. (b,c) The phospho-binding pattern of b-arrestin-1 determines selective arrestin interaction with clathrin or SRC downstream of somatostatin receptor 2 (SSTR2, Fig. 4c) and cholecystokinin A receptor (CCKAR, Fig. 4d). The Flag-CCKAR or

Flag-SSTR2 and HA-b-arrestin-1 plasmids were co-transfected in HEK293 cells. The cells were starved and then stimulated with 1 mM somatostatin (c) or 2 mM CCK-8s (d) for 15 min, respectively. The somatostatin promoted both clathrin and SRC recruitment for b-arrestin-1, whereas the CCK-8s only promoted b-arrestin-1/clathrin interaction. The site 4 mutation R25A selectively disrupted b-arrestin-1/clathrin interaction after somatostatin or CCK-8s application and the site 5 mutation R7A impaired the b-arrestin-1/SRC complex formation after somatostatin stimulation. (ac) Representative western blottings from at least three experiments are shown on the left and the statistics from at least three independent experiments are shown on the right.

*Po0.05, **Po0.01; the effects of mutants were compared with the wild type. (d) A proposed model for arrestin signalling encoded by the phospho-C-tail of the GPCRs. The phosphate-binding sites arranged in the N terminus of arrestin look similar to the holes in a ute. Left: the pattern 1-4-6-7 directed a specic arrestin conformation for clathrin signalling. Middle, the pattern 1-5 directed specic SRC signalling. Right: other combinations may direct other GPCR signalling.

through identical phosphate-binding sites (1-4-6-7) (Table 1 and Fig. 4b). In contrast, a different phospho-interaction pattern was observed for the interaction between GRK6pp and b-arrestin-1 (1-5) (Fig. 4b). Analyses of 19F-NMR spectra allowed us to reveal structural changes common to all functional phosphopeptide binding. These common changes include alterations to residue F75 in the nger loop, T136 in the middle loop and N375 in the C-terminal region (Fig. 5a). Recent EM and crystallographic studies have suggested that the nger loop of arrestins is inserted into transmembrane (TM) helices and intracellular loops of receptors after receptor/arrestin complex formation2,43. Therefore, structural changes at the nger loop and middle loop may be an important common mechanism in promoting the formation of tight receptor/arrestin complexes after arrestin reads specic receptor phosphorylation signals. Apart from the polar core of the central portion of the b-arrestin-1 protein, the

C-terminal tail of b-arrestin-1 (encompassing N375) runs in the opposite direction to the nger loop and the middle loop. Two potentially different modes of binding between b-arrestin-1 and clathrin in a non-stimulated state have been proposed by the complex structure of clathrin and b-arrestin-1 (ref. 34). However, the interaction mode between active arrestin and clathrin has not been dened. In the present study, relaxation experiments at position N375 with different conditions enabled us to

demonstrate that the C-terminal region of b-arrestin-1 is protected in the GRK2Bpp/b-arrestin-1/clathrin ternary complex (Fig. 6e,f). Thus, our results suggest that the receptor and clathrin interact with b-arrestin-1 in opposite directions to form a sandwich-like structure during receptor internalization.

More importantly, 19F-NMR spectra at other F2Y-incorporated sites provide direct biophysical evidence for the presence of at least three unique active b-arrestin-1 structural states corresponding to different phospho-binding patterns. The change at F277 is specic to GRK6pp (phospho pattern 1-5), the changes at the Y249 and L338 positions are specic to GRK2pp (phospho pattern 1-4-6-7) and the change at the Y209 position is unique to V2Rpp (phospho pattern-1-2-3-4-5-6-7) (Fig. 5bd). By maintaining or disrupting the polar core of b-arrestin-1, specic interactions at phosphate-binding site 4 may play a causative role in discriminating among distinct arrestin conformations induced by the binding of different GRKpps. Accordingly, the change in

19F-NMR spectra at GRK2-specic sites but not GRK6-specic sites after clathrin binding, in combination with GST pull-down and cellular studies, further conrms the coupling of distinct arrestin structural states to selective arrestin functions (Figs 6ab, 7 and 8ac). Taken together, the experiments in our study elucidate a key step in the working mechanism of GPCRs; in particular, the phospho-barcode of the receptor is harboured in

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 11

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202

the phosphate-binding concave surface of b-arrestin-1 and translated by arrestin into specic structural states that direct distinct arrestin-mediated functions (Fig. 8d).

In addition to revealing the phospho-coding mechanism of the specic functions of b-arrestin-1, 19F-NMR spectroscopy enabled us to probe for dynamic and subtle conformational changes at specic F2Y-incorporated sites. For example, the 19F-NMR spectra reveal the presence of both slow and fast exchanges between different conformations at the Y21-F2Y and Y209-F2Y positions (Fig. 2a,b,e). Specically, Y21-F2Y consists of two components after V2Rpp binding, indicating the presence of two independent equilibria between locally different conformations in the presence of the V2Rpp. Increasing the V2Rpp concentration lead to the increased size of the upeld-shifted peak of F2Y-Y21 at position 135.22 (Fig. 2b and Supplementary Table 3).

Moreover, the doseresponse curve of V2Rpp-stimulated clathrin recruitment was directly correlated with the increased peak area of F2Y-Y21 at the 135.22 position, suggesting that the

equilibrium at 135.22 represents an active arrestin conforma

tion (Supplementary Fig. 6). It is worth noting that Y21 forms a hydrophobic interaction with the F388 of the C-tail, which is important in maintaining the three elements interaction in the inactive arrestin structure (Supplementary Fig. 5c). In the previous model, binding of the phosphorylated receptor C-tail removed the proximal C-terminal region of b-arrestin. Consistent with this hypothesis, the paramagnetic relaxation experiment at Y21 suggested that V2Rpp binding induced greater solvent accessibility at the 135.22 equilibrium position (Supplementary

Fig. 5). Similar to the binding of V2Rpp, the binding of GRK2App and GRK2Bpp to b-arrestin-1 induced two independent equilibria at the Y21-F2Y position (Fig. 4a). The population of active peaks (at the 135.05 position) induced by GRK2pp binding was

smaller than that caused by V2Rpp binding, consistent with the lower efcacy of GRK2pp in promoting clathrin recruitment than that of V2Rpp (Fig. 3d). Interestingly, even in the saturated V2Rpp concentration, half of the F2Y-Y21 of b-arrestin still presented as an inactive structural state (peak at 130.18)

(Fig. 2b). This result indicated the possibility that not all b-arrestin-1 proteins are activated during the signal transduction, or that the complete activation of b-arrestin-1 requires other activating elements, such as binding of the receptor transmembrane core.

In contrast to visual arrestin that forms oligomers in physiological conditions, the full-length b-arrestin-1 is mainly a monomer in solution44. However, a truncated version of b-arrestin-1 (1382) exists as mixtures of monomer and dimer, which implies that the oligomerization state of b-arrestin-1 might be regulated under appropriate physiological conditions45. Among the 17 F2Y-incorporated positions of b-arrestin-1 in the present study, the Y249 and L338 positions are in the vicinity of the dimer interface of the truncated b-arrestin-1 (1382)

(Supplementary Fig. 18). Although our NMR experiments used full-length b-arrestin-1, which is unlikely to form dimers during data collection, the observed 19F-NMR changes at these positions may reect the fact that the binding of phosphopeptide or clathrin can change the oligomerization propensity of b-arrestin-1 under specic cellular conditions. Recently, the functional importance of homodimerization and heterodimerization has been extensively studied at the receptor level38,46,47. Although the functional relevance of arrestin oligomers remains relatively unknown, oligomeric forms of visual arrestins have been observed under physiological conditions and studies have found that b-arrestins concentrate in specic subcellular localizations to allow for dimer formation after receptor activation4850. The formation of b-arrestin oligomers might provide a platform for the binding of receptor dimers or other downstream effectors, including

Mdm2 or other proteins50, leading to the formation of higher-order functional complexes.

In conclusion, we have claried a longstanding question regarding the receptor phospho-coding mechanism that dictates selective b-arrestin-1 structural features and signalling. The phosphate-binding sites along the N terminus of b-arrestin-1 are arranged in a shape that is similar to the holes in a ute and arrestin moves differently according to the instructions of the phospho-receptor ngers. Whereas a receptor-encoded 1-4-6-7 pattern determines the specic b-arrestin-1 conformation for clathrin recruitment, a different receptor tune, 1-5, provides the SRC signalling order (Fig. 8d). Moreover, the N terminus of b-arrestin-1 harbours three potential additional phosphate-binding sites (A1A3; Supplementary Fig. 19). The combination of all phosphate-binding sites in a single arrestin allows the expression of more than 1,000 patterns (2101 1,023) that allow

it, in theory, to produce a plethora of arrestin conformations, facilitating numerous downstream protein interactions51. Most of the phosphorylation binding sites of b-arrestin-1 consist of conserved residues, indicating their common roles in the recognition of phospho-barcodes in different arrestin members and across different species (Supplementary Figs 2 and 3). Activation of the same G subtype protein-coupled GPCR can have distinct functions in cells. Although some functional differences are attributed to the duration and concentration of increased levels of second messengers52, many functional differences in receptors may result from distinct arrestin functions53,54.The revealed ability of arrestin to read and translate various phosphorylation patterns through its phosphate-binding concave surface contributes to the diverse functions of more than 800 GPCRs present in the human genome. The phospho-receptor C-tail may work together with different ligand-induced receptor conformations and orchestrate the breadth of GPCR signalling complexity.

Methods

Reagents. The monoclonal anti-phospho-SRC (pTyr-416, 2101), anti-GST (2622) and anti-YFP (2555) antibodies were from Cell Signaling. The anti-c-SRC (60315-1-Ig), anti-GRK2 (13990-1-AP) and anti-GRK6 (11439-1-AP) antibodies were from Proteintech (Chicago, USA). The monoclonal anti-Flag M2 antibody (F3165) were purchased from Sigma. The anti-b2AR(sc-569), anti-haemagglutinin (HA)(sc-7392) and anti-Clathrin (sc-12734) antibodies were from Santa Cruz. Most of the antibodies are used at a 1:1,000 dilution, except for anti-clathrin antibody at 1:500. The HA and Flag beads were purchased from Sigma. Glutathione-Sepharose 4B and Ni-NTA Agarose were from Amersham Pharmacia Biotech. V2Rpp and GRK2App were synthesized by the Tufts University core facility (Boston, USA). GRK2Bpp, GRK6pp and PKApp were synthesized by China Peptides Co., Ltd. (Shanghai, China). The monoclonal anti-b-arrestin-1 antibody was a generous gift from Dr Lefkowitz at Duke University. All of the other reagents were from Sigma.

Constructs. The pcDNA3.0-Flag-b2AR (Flag-b2AR), b-arrestin-1-YFP, pcDNA3.1-Luc-b-arrestin-2-YFP and pCDNA3.1-Clathrin constructs were generous gifts from Dr Lefkowitz at Duke University21,32. The chicken c-SRC (residues 87531, containing SH3, SH2 and catalytic domains) was a generous gift from Dr Zhong-yin Zhang at Indiana University and John Kuriyan at UC Berkeley55,56. For F2Y-incorporated protein expression, the pEVOL-F2YRS plasmid used has been described previously19. To monitor the b-arrestin-1 conformational change, the BRET sensor for b-arrestin-1 (pcDNA3.1-Luc-barrestin-1-YFP) was created by in-Fusion of the b-arrestin-1-YFP plasmid with the pcDNA3.1-Luc-b-arrestin-2-YFP construct57,58. For E. coli expression, the full-length wild-type complementary DNAs of bovine b-arrestin-1 was subcloned into the NdeI/XhoI sites of the pET22b vector with the C-terminal His tag. DNA encoding residues 1494 of human clathrin and chicken c-SRC (residues 87531, containing SH3, SH2 and catalytic domains) were sub-cloned into a PGEX-6P1 expression vector with an N-terminal GST tag for E. coli expression and GST pull-down assays. The b-arrestin-1 mutations Y21TAG, Y63TAG, F75TAG,

T136TAG, Y249TAG, F277TAG, L338TAG, L379TAG, F75Y, T136Y, F277Y, L338Y and L379Y were generated using the Quikchange mutagenesis kit (Stratagene). All of the constructs and mutations were veried by DNA sequencing.

12 NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202 ARTICLE

Cell culture, siRNA and transfection. HEK 293 cells were attained from ATCC and cultured at 37 C and 5% CO2 in DMEM containing 10% (v/v) fetal bovine serum and penicillinstreptomycin (5,000 IU l 1). GRK2 or GRK6 expression was silenced in HEK 293 cells using a specic small interfering RNA (siRNA) duplex.

The GRK2 target sequences were 50-AAGAAGUACGAGAAGCUGGAG-30 and the GRK6 target sequences were 50-AACAGUAGGUUUGUAGUGAGC-30 as previously described5,59,60. Cells were transfected with GRK siRNA, control siRNA (a non-targeting 25-nt siRNA) or other plasmids by Lipofectamine 2000 according to the manufacturers instructions.

Western blotting. To examine the roles of GRK2 and GRK6 in b2-AR-mediated SRC activation, selective siRNA towards specic GRK and Flag-b2AR plasmids were co-transfected into HEK-293 cells. After 24 h of transfection, the cells were starved for 4 h and then stimulated with ISO (10 mM) at the indicated times.

Subsequently, the cells were washed three times with cold PBS and then collected in cold lysis buffer (50 mM Tris pH 7.5, 150 mMNaCl, 1% Triton X-100, 1 mM EGTA, 1 mM Na3VO4, 50 mMNaF, 0.25% (m/v) sodium deoxycholate, 10% (v/v)

glycerol, a protease inhibitor cocktail tablet and 5 mM IAA). The cell lysates were centrifuged for 15 min after 30 min of end-to-end rotation at 4 C. Lysate protein concentrations were determined using the Bradford protein assay. Equal amounts of lysate proteins were subjected to SDSPAGE and western blotting. SRC-416 phosphorylation was detected using a specic antibody. The original pictures for uncropped gels are shown in Supplementary Figs 20 and 21.

Co-immunoprecipitation. To examine the effects of b-arrestin-1 mutants on ISO-induced b2AR/b-arrestin-1 complex formation, the co-immunoprecipitation experiment was performed61,62. In detail, HEK293 cells were co-transfected with

Flag-b2AR and b-arrestin-1-YFP mutants. Twenty-four hours after transfection, the cells were starved for 4 h and then stimulated with ISO (10 mM) for 10 min. The cell lysates were subjected to immunoprecipitation using anti-FLAG beads incubated overnight at 4 C. Immune complexes were analysed by western blotting with specic antibodies. The associated b-arrestin 1-YFP was detected using an anti-YFP antibody.

To examine the effects of b-arrestin-1 mutants on ISO, CCK-8s or somatostatin induced clathrin/b-arrestin-1 or SRC/b-arrestin-1 complex formation downstream of b2AR, CCKAR and SSTR2; HEK293 cells were co-transfected with Flag-b2AR/HA-b-arrestin-1, Flag-CCKAR/HA-b-arrestin-1 or Flag-SSTR2/ HA-b-arrestin-1, respectively. Different HA-b-arrestin-1 mutations were used to detect the contribution of the phosphate-binding sites of b-arrestin-1 during the arrestin/clathrin or arrestin/SRC complex formatin. Forty-eight hours after transfection, the cells were starved for 8 h and then stimulated with ISO (10 mM for activation of b2AR), CCK-8s (2 mM for activation of CCKAR) or somatostatin(1 mM for activation of SSTR2) for 15 min. The cell lysates were subjected to immunoprecipitation using anti-HA beads incubated overnight at 4 C. Immune complexes were analysed by western blotting with specic antibody towards clathrin or pSRC416, respectively.Representative western blots from at least three experiments are shown in Fig. 8.

Confocal microscopy. Confocal microscopy experiment was performed30. The plasmids encoding the b-arrestin-1 BRET biosensor (Luc-b-arrestin-1-YFP) or b-arrestin-1-YFP mutants were transiently co-transfected with Flag-b2AR in HEK-293 cells. Twenty-four hours after transfection, the cells were plated on bronectin-coated, 35-mm, glass-bottom plates. After 4 h of starvation, the cells were stimulated with ISO (10 mM) for 10 min. The translocation of b-arrestin-1 was monitored by VT-Innity Confocal Imaging System.

BRET assay. The plasmids Luc-b-arrestin-1-YFP, Flag-b2AR and different siRNAs were transiently co-transfected in HEK-293 cells. Twenty-four hours after transfection, HEK-293 cells were distributed in bronectin-coated 96-well microplates. Before the BRET assay, cells were washed three times with PBS and then cells were incubated with coelenterazine h (nal concentration, 5 mM) for5 min. Subsequently, the cells were stimulated with 10 mM ISO for 10 min and light emission was detected (460500 nm for Luc and 510550 nm for yellow uorescent protein (YFP)) using a multilabel reader (Mithras Luria Bertani 940; Berthold Technologies). The BRET signal was determined as the ratio of the light emitted by YFP and the light emitted by Luc.

Expression and purication of native b-arrestin-1. The plasmid pET22b-barrestin 1 was transformed into BL21 E. coli. The transformed BL21 cells were cultured and induced with 0.3 mM isopropyl-b-D-thiogalactoside (IPTG) at an

OD600 nm of 0.8. After growing overnight at 25 C, the cells were harvested, re-suspended in lysis buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl) and lysed by sonication. The bacterial lysate was next centrifuged and the supernatant was puried by Ni-NTA afnity chromatography and Superdex 200 according to the manufacturers instructions (GE Healthcare).

Expression and purication of F2Y-incorporated proteins. For the expression of b-arrestin-1 F2Y proteins, pEVOL-F2YRS was co-transformed with different pET22b-b-arrestin-1 TAG mutations into BL21 (DE3). A single colony was grown overnight at 37 C in Luria Bertani (LB) medium. Five litres of the transformed cells were then induced with 0.3 mM IPTG and 0.02% L-arabinose at an OD600 nm of 1.0 in the presence of 0.5 mM F2Y. After growing overnight at 25 C, the cells were harvested and resuspended in buffer containing 20 mM Tris-HCl pH 8.0 and 150 mM NaCl. The bacterial lysate was centrifuged and the supernatant was puried by Ni-NTA afnity chromatography and gel ltration chromatography.

Trypsin digestion and MS/MS analysis. The b-arrestin-1-F2Y protein was subjected to electrophoresis and the protein band was cut into small plugs and washed twice in 200 ml distilled water for 10 min. The gel bands were dehydrated in 100% acetonitrile for 10 min and dried in a Speedvac (Labconco) for B15 min. Disulde bonds were reduced by adding 10 ml of 100 mM dithiothreitol (DTT) and subsequently alkylated by 40 mM IAA, 25 mM NH4HCO3 for 45 min at room temperature in the dark. The sample was then mixed with trypsin by a ratio of 100:1 in Tris buffer and digested at 37 C for 12 h. Digestion was stopped by adding formic acid to 1% nal concentration. Digested samples were puried and desalted, and re-dissolved in 30 ml 50% CH3CN/0.1% CF3COOH buffer before MS/MS analysis.

LC-MS/MS analysis was performed using a Thermo Finnigan LTQ linear ion trap mass spectrometer in line with a Thermo Finnigan Surveyor MS Pump Plus HPLC system. The peptides generated by trypsin digestion were loaded onto a trap column (300SB-C18, 5 0.3 mm, 5 mm particle) (Agilent Technologies, Santa

Clara, CA), which was connected through a zero dead volume union to the self-packed analytical column (C18, 100 mm i.d 100 mm, 3 mm particle)

(SunChrom, Germany). The peptides were then eluted over a gradient (045% B in 55 min, 45100% B in 10 min, where B 80% Acetonitrile, 0.1% formic acid) at a

ow rate of 500 nl min 1 and introduced online into the linear ion trap mass spectrometer (ThermoFisher Corporation, San Jose, CA) using nano electrospray ionization. MS data were analysed by Bioworks 3.2 software.

Expression and purication of clathrin and SRC 3D proteins. GST-tagged Clathrin or SRC 3D Protein was expressed in BL21 E. coli. In brief, 5 l of GST-tagged Clathrin or SRC 3D-transformed E. coli was cultured, induced by0.3 mM IPTG and pelleted by centrifugation. The cell pellets were resuspended in 50 ml of a GST buffer containing 25 mM Tris-HCl at pH 8.0, 150 mM NaCl, 0.5% Triton X-100, 5% glycerol, 2 mM EDTA and 1 mM DTT. The bacterial lysate was centrifuged and the supernatant was collected and incubated with 2 ml of glutathione-Sepharose 4B for 2 h. The beads were washed three times and the bound GST-Clathrin protein or GST-SRC protein was eluted with 10 mM GSH for GST pull-down experiments. To acquire clathrin for phospho-peptide/b-arrestin-1/clathrin complex formation, GST-Clathrin bound to GST beads was mixed with HRV-3C protease for 12 h at 4 C. The cleaved protein was concentrated to 500 ml and subjected to gel ltration using a Superdex 200 column (GE Healthcare) before being added to the phosphopeptide/b-arrestin-1 complex.

Clathrin/SRC-binding assay. Binding of clathrin or SRC to b-arrestin-1 was performed as previously described7,63,64. In detail, 300 nM wild-type b-arrestin-1 from F2Y incorporated proteins were rst mixed with specic phosphopeptides and incubated in binding buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2 mM EDTA, 1 mM DTT) at room temperature for 30 min. After the incubation, 10 mM

GST-clathrin or GST-SRC protein was added and incubated for another 1 h at room temperature. Subsequently, 10 ml of GST beads were added to the mixture.

After 2 h of end-to-end rotation at 4 C, the beads were collected and washed six times with cell lysis buffer. After removing the supernatant in the nal wash, the samples were re-suspended in 2 SDS loading buffer and boiled for 10 min. The

arrestin/clathrin or arrestin/SRC complexes were analysed by western blotting. For each experiment, a western blot representative of at least three independent experiments is shown in Figures.

NMR experiments. To detect phosphopeptide-induced arrestin conformational changes, 50 mM b-arrestin-1 F2Y proteins were mixed with or without a threefold molar ratio of phosphopeptides (V2Rpp, GRK2App, GRK2Bpp, GRK6pp or PKApp) and incubated in binding buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl, 10% D2O) with end-to-end rotation at room temperature for 30 min. To detect the arrestin conformational change after clathrin binding, an equal molar of clathrin was added to the arrestin or arrestin/phosphopeptide complex and incubated for another 30 min. The protein samples were then subjected to 19F-NMR experiments.

For the titration experiment, 50 mM b-arrestin-1-Y21-F2Y or b-arrestin-1-Y63-F2Y proteins were mixed at different ratios (0.3:1, 1:1, 3:1 and 9:1) of the V2Rpp and incubated in binding buffer (20 mM Tris-HCl pH 8.0, 150 mM NaCl, 10% D2O) at room temperature for 30 min. The proteins were then subjected to

19F NMR experiments.

All NMR data were collected using an Agilent OD2 600 spectrometer tted with a 5-mm broad band probe. The 19F 90 pulse lengths were 9.9 ms and the spectra were typically obtained using 15,000 scans and a recovery delay of 1 s. Data were

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 13

ARTICLE NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202

processed using 10-Hz Lorentzian line broadening and were referenced to the internal TFA standard ( 76.5 p.p.m.). All of the spectra were recorded at 25 C.

References

1. Liu, J. J., Horst, R., Katritch, V., Stevens, R. C. & Wuthrich, K. Biased signaling pathways in beta2-adrenergic receptor characterized by 19F-NMR. Science 335, 11061110 (2012).

2. Shukla, A. K. et al. Visualization of arrestin recruitment by a G-protein-coupled receptor. Nature 512, 218222 (2014).

3. Strachan, R. T. et al. Divergent transducer-specic molecular efcacies generate biased agonism at a G protein-coupled receptor (GPCR). J. Biol. Chem. 289, 1421114224 (2014).

4. Lefkowitz, R. J. & Shenoy, S. K. Transduction of receptor signals by beta-arrestins. Science 308, 512517 (2005).

5. Nobles, K. N. et al. Distinct phosphorylation sites on the beta(2)-adrenergic receptor establish a barcode that encodes differential functions of beta-arrestin. Sci. Signal. 4, ra51 (2011).

6. Hara, M. R. et al. A stress response pathway regulates DNA damage through beta2-adrenoreceptors and beta-arrestin-1. Nature 477, 349353 (2011).

7. Xiao, K. et al. Global phosphorylation analysis of beta-arrestin-mediated signaling downstream of a seven transmembrane receptor (7TMR). Proc. Natl Acad. Sci. USA 107, 1529915304 (2010).

8. Goodman, Jr O. B. et al. Beta-arrestin acts as a clathrin adaptor in endocytosis of the beta2-adrenergic receptor. Nature 383, 447450 (1996).

9. Perry, S. J. et al. Targeting of cyclic AMP degradation to beta 2-adrenergic receptors by beta-arrestins. Science 298, 834836 (2002).

10. Beaulieu, J. M. et al. An Akt/beta-arrestin 2/PP2A signaling complex mediates dopaminergic neurotransmission and behavior. Cell 122, 261273 (2005).

11. Shukla, A. K. et al. Structure of active beta-arrestin-1 bound to a G-protein-coupled receptor phosphopeptide. Nature 497, 137141 (2013).

12. Xiao, K. et al. Functional specialization of beta-arrestin interactions revealed by proteomic analysis. Proc. Natl Acad. Sci. USA 104, 1201112016 (2007).13. Rahmeh, R. et al. Structural insights into biased G protein-coupled receptor signaling revealed by uorescence spectroscopy. Proc. Natl Acad. Sci. USA 109, 67336738 (2012).

14. Busillo, J. M. et al. Site-specic phosphorylation of CXCR4 is dynamically regulated by multiple kinases and results in differential modulation of CXCR4 signaling. J. Biol. Chem. 285, 78057817 (2010).

15. Butcher, A. J. et al. Concomitant action of structural elements and receptor phosphorylation determines arrestin-3 interaction with the free fatty acid receptor FFA4. J. Biol. Chem. 289, 1845118465 (2014).

16. Butcher, A. J. et al. Differential G-protein-coupled receptor phosphorylation provides evidence for a signaling bar code. J. Biol. Chem. 286, 1150611518 (2011).

17. Morgan, D. J. et al. Mutation of putative GRK phosphorylation sites in the cannabinoid receptor 1 (CB1R) confers resistance to cannabinoid tolerance and hypersensitivity to cannabinoids in mice. J. Neurosci. 34, 51525163 (2014).

18. Delom, F. & Fessart, D. Role of phosphorylation in the control of clathrin-mediated internalization of GPCR. Int. J. Cell Biol. 2011, 246954 (2011).

19. Li, F. et al. A genetically encoded 19F NMR probe for tyrosine phosphorylation. Angew. Chem. Int. Ed. Engl. 52, 39583962 (2013).

20. Liu, C. C. & Schultz, P. G. Adding new chemistries to the genetic code. Annu. Rev. Biochem. 79, 413444 (2010).

21. Kotula, J. W. et al. Targeted disruption of beta-arrestin 2-mediated signaling pathways by aptamer chimeras leads to inhibition of leukemic cell growth. PLoS ONE 9, e93441 (2014).

22. Patel, D. J. & Wang, Z. Readout of epigenetic modications. Annu. Rev. Biochem. 82, 81118 (2013).

23. Kim, Y. J. et al. Crystal structure of pre-activated arrestin p44. Nature 497, 142146 (2013).

24. Ren, X. R. et al. Different G protein-coupled receptor kinases govern G protein and beta-arrestin-mediated signaling of V2 vasopressin receptor. Proc. Natl Acad. Sci. USA 102, 14481453 (2005).

25. Kim, J. et al. Functional antagonism of different G protein-coupled receptor kinases for beta-arrestin-mediated angiotensin II receptor signaling. Proc. Natl Acad. Sci. USA 102, 14421447 (2005).

26. Talbot, J., Joly, E., Prentki, M. & Buteau, J. beta-Arrestin1-mediated recruitment of c-Src underlies the proliferative action of glucagon-like peptide-1 in pancreatic beta INS832/13 cells. Mol. Cell. Endocrinol. 364, 6570 (2012).

27. Buchanan, F. G. et al. Role of beta-arrestin 1 in the metastatic progression of colorectal cancer. Proc. Natl Acad. Sci. USA 103, 14921497 (2006).

28. Coureuil, M. et al. Meningococcus Hijacks a beta2-adrenoceptor/beta-Arrestin pathway to cross brain microvasculature endothelium. Cell 143, 11491160 (2010).

29. Barlic, J. et al. Regulation of tyrosine kinase activation and granule release through beta-arrestin by CXCRI. Nat. Immunol. 1, 227233 (2000).

30. Luttrell, L. M. et al. Beta-arrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science 283, 655661 (1999).

31. DeFea, K. A. et al. The proliferative and antiapoptotic effects of substance P are facilitated by formation of a beta -arrestin-dependent scaffolding complex. Proc. Natl Acad. Sci. USA 97, 1108611091 (2000).

32. Violin, J. D., Ren, X. R. & Lefkowitz, R. J. G-protein-coupled receptor kinase specicity for beta-arrestin recruitment to the beta2-adrenergic receptor revealed by uorescence resonance energy transfer. J. Biol. Chem. 281, 2057720588 (2006).

33. Gurevich, V. V. & Gurevich, E. V. The molecular acrobatics of arrestin activation. Trends Pharmacol. Sci. 25, 105111 (2004).

34. Kang, D. S. et al. Structure of an arrestin2-clathrin complex reveals a novel clathrin binding domain that modulates receptor trafcking. J. Biol. Chem. 284, 2986029872 (2009).

35. Kahsai, A. W. et al. Multiple ligand-specic conformations of the beta2-adrenergic receptor. Nat. Chem. Biol. 7, 692700 (2011).

36. Nygaard, R. et al. The dynamic process of beta(2)-adrenergic receptor activation. Cell 152, 532542 (2013).

37. Bokoch, M. P. et al. Ligand-specic regulation of the extracellular surface of a G-protein-coupled receptor. Nature 463, 108112 (2010).

38. Xue, L. et al. Major ligand-induced rearrangement of the heptahelical domain interface in a GPCR dimer. Nat. Chem. Biol. 11, 134140 (2015).

39. Damian, M. et al. Ghrelin receptor conformational dynamics regulate the transition from a preassembled to an active receptor:Gq complex. Proc. Natl Acad. Sci. USA 112, 16011606 (2015).

40. Kim, M. et al. Conformation of receptor-bound visual arrestin. Proc. Natl Acad. Sci. USA 109, 1840718412 (2012).

41. Zhuang, T. et al. Involvement of distinct arrestin-1 elements in binding to different functional forms of rhodopsin. Proc. Natl Acad. Sci. USA 110, 942947 (2013).

42. Charest, P. G., Terrillon, S. & Bouvier, M. Monitoring agonist-promoted conformational changes of beta-arrestin in living cells by intramolecular BRET. EMBO Rep. 6, 334340 (2005).

43. Szczepek, M. et al. Crystal structure of a common GPCR-binding interface for G protein and arrestin. Nat. Commun. 5, 4801 (2014).

44. Milano, S. K., Pace, H. C., Kim, Y. M., Brenner, C. & Benovic, J. L. Scaffolding functions of arrestin-2 revealed by crystal structure and mutagenesis. Biochemistry 41, 33213328 (2002).

45. Han, M., Gurevich, V. V., Vishnivetskiy, S. A., Sigler, P. B. & Schubert, C. Crystal structure of beta-arrestin at 1.9A: possible mechanism of receptor binding and membrane Translocation. Structure 9, 869880 (2001).

46. Bellot, M. et al. Dual agonist occupancy of AT1-R-alpha2C-AR heterodimers results in atypical Gs-PKA signaling. Nat. Chem. Biol. 11, 271279 (2015).47. Bouvier, M., Heveker, N., Jockers, R., Marullo, S. & Milligan, G. BRET analysis of GPCR oligomerization: newer does not mean better. Nat. Methods 4, 34 author reply 4 (2007).

48. Hirsch, J. A., Schubert, C., Gurevich, V. V. & Sigler, P. B. The 2.8A crystal structure of visual arrestin: a model for arrestins regulation. Cell 97, 257269 (1999).

49. Schubert, C. et al. Visual arrestin activity may be regulated by self-association.J. Biol. Chem. 274, 2118621190 (1999).50. Boularan, C. et al. Beta-arrestin 2 oligomerization controls the Mdm2-dependent inhibition of p53. Proc. Natl Acad. Sci. USA 104, 1806118066 (2007).

51. Shenoy, S. K. & Lefkowitz, R. J. Beta-arrestin-mediated receptor trafcking and signal transduction. Trends Pharmacol. Sci. 32, 521533 (2011).

52. De Marinis, Y. Z. et al. GLP-1 inhibits and adrenaline stimulates glucagon release by differential modulation of N- and L-type Ca2 channel-dependent

exocytosis. Cell Metab. 11, 543553.53. Shukla, A. K., Xiao, K. & Lefkowitz, R. J. Emerging paradigms of beta-arrestin-dependent seven transmembrane receptor signaling. Trends Biochem. Sci. 36, 457469 (2011).

54. Reiter, E., Ahn, S., Shukla, A. K. & Lefkowitz, R. J. Molecular mechanism of beta-arrestin-biased agonism at seven-transmembrane receptors. Annu. Rev. Pharmacol. Toxicol. 52, 179197 (2012).

55. Levinson, N. M. et al. Structural basis for the recognition of c-Src by its inactivator Csk. Cell 134, 124134 (2008).

56. Yu, X. et al. Structure, inhibitor, and regulatory mechanism of Lyp,a lymphoid-specic tyrosine phosphatase implicated in autoimmune diseases. Proc. Natl Acad. Sci. USA 104, 1976719772 (2007).

57. Shukla, A. K. et al. Distinct conformational changes in beta-arrestin report biased agonism at seven-transmembrane receptors. Proc. Natl Acad. Sci. USA 105, 99889993 (2008).

58. Charest, P. G., Terrillon, S. & Bouvier, M. Monitoring agonist-promoted conformational changes of beta-arrestin in living cells by intramolecular BRET. EMBO Rep. 6, 334340 (2005).

59. Heitzler, D. et al. Competing G protein-coupled receptor kinases balance G protein and beta-arrestin signaling. Mol. Syst. Biol. 8, 590 (2012).

60. Chen, W. et al. Activity-dependent internalization of smoothened mediated by beta-arrestin 2 and GRK2. Science 306, 22572260 (2004).

14 NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications

NATURE COMMUNICATIONS | DOI: 10.1038/ncomms9202 ARTICLE

61. Wang, H. M. et al. A stress response pathway in mice upregulates somatostatin level and transcription in pancreatic delta cells through Gs and beta-arrestin 1. Diabetologia 57, 18991910 (2014).

62. Wang, H. M. et al. The catalytic region and PEST domain of PTPN18 distinctly regulate the HER2 phosphorylation and ubiquitination barcodes. Cell Res. 24, 10671090 (2014).

63. Xiao, K. et al. Activation-dependent conformational changes in {beta}-arrestin 2. J. Biol. Chem. 279, 5574455753 (2004).

64. Nobles, K. N. et al. The active conformation of beta-arrestin1: direct evidence for the phosphate sensor in the N-domain and conformational differences in the active states of beta-arrestins1 and -2. J. Biol. Chem. 282, 2137021381 (2007).

Acknowledgements

We thank Dr Dang-sheng Li for stimulating discussions and Dr Xin Xie for critical reading of the manuscript. We acknowledge support from the National Key Basic Research Program of China (2012CB910402 to J.-P.S., 2015CB856203 to J.-Y.W. and 2013CB967700 to X.Y.), the National Natural Science Foundation of China (21325211 toJ.-Y.W., 31271505 to J.-P.S. and 31370016 to F.-H.L.), the Shandong Natural Science Fund for Distinguished Young Scholars (JQ201320 to X.Y. and JQ201517 to J.-P.S.), the Fundamental Research Fund of Shandong University (2014JC029 to X.Y.) and the Program for Changjiang Scholars and Innovative Research Team in University (IRT13028). We thank Ms Shan-shan Zang and Dr Xue-hui Liu of the Core Facility of Protein Research, Institute of Biophysics, Chinese Academy of Sciences, for their help in the NMR data collection, analysis and valuable discussion.

Author contributions

J.-P.S., J.-Y.W. and F.Y. designed all of the experiments. F.Y. puried both F2Y incorporated b-arrestin 1 and clathrin. F.Y. also functionally characterized b-arrestin-1, optimized, collected and analysed the 19F-NMR data and performed the BRET assay.

K.-H.X., F.Y. and F. Yi. performed the HDXMS experiments and designed key experiments. F.Y. and F. Yi. performed other mass spectrometry experiments. X.Y. made the arrestin, clathrin and SRC constructs, and designed and supervised all of the cellular assays. C.L., C.-X.Q., F.-H.L. and D.F-.H. puried F2Y-incorporated b-arrestin-1, clathrin and SRC-3D. H.-M.W. performed the confocal microscopy and co-immunoprecipitation assays. Z.G. and H.-D.L. carried out the cellular studies. C.-L.T. initiated the 19F-NMR experiments. C.S., J.-P.S. and F.Y. performed molecular dynamics and molecular modelling. J.-P.S., J.-Y.W. and X.Y. supervised the overall project design and execution. F.Y., X.Y., H.-M.W. and J.-P.S. participated in data analysis and interpretation. J.-Y.W. conceived the F2Y incorporation and 19F-NMR experiments. F.Y., X.Y. and J.-P.S. wrote the manuscript. All of the authors have seen and commented on the manuscript.

Additional information

Supplementary Information accompanies this paper at http://www.nature.com/naturecommunications

Web End =http://www.nature.com/ http://www.nature.com/naturecommunications

Web End =naturecommunications

Competing nancial interests: The authors declare no competing nancial interests.

Reprints and permission information is available online at http://npg.nature.com/reprintsandpermissions/

Web End =http://npg.nature.com/ http://npg.nature.com/reprintsandpermissions/

Web End =reprintsandpermissions/

How to cite this article: Yang, F. et al. Phospho-selective mechanisms of arrestin conformations and functions revealed by unnatural amino acid incorporation and

19F-NMR. Nat. Commun. 6:8202 doi: 10.1038/ncomms9202 (2015).

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the articles Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/

NATURE COMMUNICATIONS | 6:8202 | DOI: 10.1038/ncomms9202 | http://www.nature.com/naturecommunications

Web End =www.nature.com/naturecommunications 15

Word count: 12891

Show less

Abstract

Translate

Specific arrestin conformations are coupled to distinct downstream effectors, which underlie the functions of many G-protein-coupled receptors (GPCRs). Here, using unnatural amino acid incorporation and fluorine-19 nuclear magnetic resonance (¹⁹ F-NMR) spectroscopy, we demonstrate that distinct receptor phospho-barcodes are translated to specific β-arrestin-1 conformations and direct selective signalling. With its phosphate-binding concave surface, β-arrestin-1 'reads' the message in the receptor phospho-C-tails and distinct phospho-interaction patterns are revealed by ¹⁹ F-NMR. Whereas all functional phosphopeptides interact with a common phosphate binding site and induce the movements of finger and middle loops, different phospho-interaction patterns induce distinct structural states of β-arrestin-1 that are coupled to distinct arrestin functions. Only clathrin recognizes and stabilizes GRK2-specific β-arrestin-1 conformations. The identified receptor-phospho-selective mechanism for arrestin conformation and the spacing of the multiple phosphate-binding sites in the arrestin enable arrestin to recognize plethora phosphorylation states of numerous GPCRs, contributing to the functional diversity of receptors.

Details

Title

Phospho-selective mechanisms of arrestin conformations and functions revealed by unnatural amino acid incorporation and 19F-NMR

Author

Yang, Fan; Yu, Xiao; Liu, Chuan; Qu, Chang-xiu; Gong, Zheng; Liu, Hong-da; Li, Fa-hui; Wang, Hong-mei; He, Dong-fang; Yi, Fan; Song, Chen; Tian, Chang-lin; Xiao, Kun-hong; Wang, Jiang-yun; Sun, Jin-peng

Pages

8202

Publication year

2015

Publication date

Sep 2015

Publisher

Nature Publishing Group

e-ISSN

20411723

Source type

Scholarly Journal

Language of publication

English

DOI

https://doi.org/10.1038/ncomms9202

ProQuest document ID

1710313249

Phospho-selective mechanisms of arrestin conformations and functions revealed by unnatural amino acid incorporation and 19F-NMR

Jump to:

Full text

Abstract

Details

Suggested sources