Báo cáo khoa học: Identification of sites of phosphorylation by G-protein-coupled receptor kinase 2 in b-tubulin ppt

Identification of sites of phosphorylation by G-protein-coupledreceptor kinase 2 in b-tubulin Norihiro Yoshida1,*, Kazuko Haga1,2and Tatsuya Haga1,2 1 Department of Neurochemistry, Facul

Identification of sites of phosphorylation by G-protein-coupled

receptor kinase 2 in b-tubulin

Norihiro Yoshida1,*, Kazuko Haga1,2and Tatsuya Haga1,2

1

Department of Neurochemistry, Faculty of Medicine, University of Tokyo, Japan;2Institute for Biomolecular Science,

Faculty of Science, Gakushuin University, Tokyo, Japan

G-protein-coupled receptor kinase 2 (GRK2) is known to

specifically phosphorylate the agonist-bound forms of

G-protein-coupled receptors (GPCRs) This strict specificity

is due at least partly to activation of GRK2 by

agonist-bound GPCRs, in which basic residues in intracellular

regions adjacent to transmembrane segments are thought to

be involved Tubulin was found to be phosphorylated by

GRK2, but it remains unknown if tubulin can also serve as

both a substrate and an activator for GRK2 Purified

tubulin, phosphorylated by GRK2, was subjected to

biochemical analysis, and the phosphorylation sites in

b-tubulin were determined to be Thr409 and Ser420 In

addition, the Ser444 in bIII-tubulin was also indicated to be

phosphorylated by GRK2 The phosphorylation sites in

tubulin for GRK2 reside in the C-terminal domain of

b-tubulin, which is on the outer surface of microtubules

Pretreatment of tubulin with protein phosphatase type-2A

(PP2A) resulted in a twofold increase in the phosphorylation

of tubulin by GRK2 These results suggest that tubulin is phosphorylated in situ probably by GRK2 and that the phosphorylation may affect the interaction of microtubules with microtubule-associated proteins A GST fusion protein

of a C-terminal region of bI-tubulin (393–445 residues), containing 19 acidic residues but only one basic residue, was found to be a good substrate for GRK2, like full-length b-tubulin These results, together with the finding that GRK2 may phosphorylate synuclein and phosducin in their acidic domains, indicate that some proteins with very acidic regions but without basic activation domains could serve as substrates for GRK2

Keywords: G-protein-coupled receptor kinase; protein phosphorylation; tubulin

Many G-protein-coupled receptors (GPCRs) including

rhodopsin, muscarinic acetylcholine receptors, and

b-adre-nergic receptors are known to be phosphorylated in a

light-dependent or agonist-light-dependent manner by members of the

protein kinase family called G-protein-coupled receptor

kinases (GRKs) [1] GRKs constitute a subgroup of the

serine/threonine kinase superfamily and are characterized

by their strict substrate specificity, i.e they only recognize

the stimulated forms of GPCRs The phosphorylation sites

in rhodopsin [2] and b2-adrenergic receptors [3] for GRK1

and GRK2, respectively, are located in their C-termini, and

those in muscarinic acetylcholine receptor M2subtypes (M2

receptors) [4], M3receptors [5], and a2-adrenergic receptors

[6] for GRK2 are in the central parts of their third

intracellular loops No strict consensus sequence for GRK-mediated phosphorylation has been found among these phosphorylation sites, except that acidic amino-acid resi-dues near the phosphorylation sites may be required [7] Peptides corresponding to these phosphorylation sites are generally poor substrates for GRK1 or GRK2, but their phosphorylation is greatly stimulated by rhodopsin [8],

b2-adrenergic receptors [9] or M2receptors [10] in a light-dependent or agonist-light-dependent manner GRK2, but not GRK1, is also stimulated by G-protein bc subunits, and this phosphorylation is synergistically stimulated by agonist-bound receptors and G-protein bc subunits [11–14] These results indicate that light-exposed rhodopsin, agonist-bound b-adrenergic receptors, or M2 receptors function both as substrates and activators, and explain, at least partly, why the substrates of GRK2 are restricted to agonist-bound receptors in spite of the absence of a strict consensus sequence among various phosphorylation sites As phos-phorylation site-deleted rhodopsin [15] and M2 receptors [10] also act as activators of GRK2, the activation sites are thought to be different from the phosphorylation sites Possible activation sites in M2receptors are suggested to be several portions of intracellular loops adjacent to trans-membrane segments, because the peptides corresponding to these regions stimulated phosphorylation of synthetic peptides corresponding to the phosphorylation sites in M2 receptors [14] These regions are assumed to undergo a conformational change on agonist binding and to be involved in the interaction with G-proteins [16,17] Further-more, mastoparan, which is known to mimic agonist-bound

Correspondence to T Haga, Institute for Biomolecular Science,

Faculty of Science, Gakushuin University, Mejiro 1-5-1,

Toshima-ku, Tokyo 171-8588, Japan.

Fax: + 81 35992 1034, Tel.: + 81 35992 1033,

E-mail: tatsuya.haga@gakushuin.ac.jp

Abbreviations: GPCR, coupled receptor; GRK,

G-protein-coupled receptor kinase; PP2A, phosphatase 2A; MAP,

microtubule-associated protein; PVDF, poly(vinylidene difluoride);

GST, glutathione S-transferase.

*Present address: Otsuka Pharmaceutical Co Ltd, Research Institute

for Pharmacological and Therapeutical Development,

Tokushima, Japan.

(Received 23 October 2002, revised 7 January 2003,

accepted 17 January 2003)

receptors and activates G-proteins [18], has been shown to

stimulate GRK1 [15] and GRK2, particularly in the

presence of G-protein bc subunits [14] All these peptides

with GRK2-stimulating activity, including mastoparan, are

basic peptides

Recent studies have suggested that GRK may

phos-phorylate substrates other than the stimulated forms of

GPCRs Tubulin is the first nonreceptor protein found to be

phosphorylated by GRK2 and GRK5, although its

phos-phorylation sites have not been identified yet [19–21] Other

nonreceptor substrates for GRK2 have been reported,

including synucleins [22], phosducin, and phosducin-like

protein [23] The phosphorylation sites in synucleins and

phosducin are located in their C-terminal domains, which

include many acidic residues but few basic residues It

remains unknown, however, whether the C-terminal

pep-tides serve as good substrates for GRK2 by themselves or

do so only in the presence of activating domains in another

part of these proteins We have attempted to identify

phosphorylation sites for GRK2 in tubulin as a first step to

determining if tubulin serves as both a substrate and an

activator for GRK2, as was shown in the case of stimulated

forms of GPCRs

Here, we show that tubulin is phosphorylated by GRK2

in a very acidic C-terminal domain and that the C-terminal

peptide of tubulin is a good substrate for GRK2, suggesting

that the presence of a basic activation domain is not

necessary for the protein to be a substrate for GRK2 In

addition, we present evidence that tubulin is phosphorylated

in situat the sites phosphorylated by GRK2

Materials and methods

Materials

Phenyl-sepharose, heparin-sepharose,

glutathione-seph-arose 4B, sephadex G-50 fine, [c-32P]ATP, the pGEX4T-3

vector, and an ECL chemiluminescence detection system

were purchased from Amersham Pharmacia Biotech

Achromobacterprotease I and endoproteinase Asp-N were

purchased from Wako Pure Chemical Industries KOD

polymerase, Pfu turbo polymerase, the pBluescript vector,

and restriction enzymes were from Toyobo C18RP-HPLC

and DEAE-5PW columns were from Tosoh A thermo

sequence fluorescent-labeled primer cycle sequencing kit

was purchased from Perkin–Elmer TLCplates were

purchased from Merck, and human erythrocyte

phospha-tase 2A (PP2A) was from Upstate Biotechnology Inc Other

reagents used were of the highest grade commercially

available

Protein expression and purification

GRK2 was overexpressed in and purified from Sf9 insect

cells as described previously with some modifications [10]

The infected Sf9 cells were homogenized in 20 mMHepes/

KOH (pH 7.0), containing 2 mMMgCl2, 1 mM

dithiothre-itol, and 0.5 mMphenylmethanesulfonyl fluoride (solution

A; 20 mL per cell pellet from 1 L culture) The homogenate

was centrifuged, and then the pellet was homogenized in

solution A supplemented with 0.5M KCl Most of the

GRK2 activity was recovered in the supernatant obtained

by centrifugation at 42 000 g for 20 min Ammonium sulfate was added to the extract to a saturation level of 20% After centrifugation, a saturated ammonium sulfate solu-tion was added to the supernatant to give a final concen-tration of 30%, and then the suspension was centrifuged and the resulting pellet dissolved in solution A (15 mL) This solution was applied to a phenyl-Sepharose column (5 mL) equilibrated with 20 mM Hepes/KOH (pH 7.0), containing 1 mM dithiothreitol, 0.5 mM phenylmethane-sulfonyl fluoride, and 1Mammonium sulfate at a flow rate

of 1 mLÆmin)1 After the column had been washed thoroughly, proteins were eluted with a linear gradient of 1.0–0Mammonium sulfate (1.0 and 0M, 20 mL each) and collected in fractions of 2 mL each The fractions containing tubulin-phosphorylating activity were combined, dialyzed against 20 mM Hepes/KOH (pH 7.0)/50 mM NaCl, and then applied to a heparin column (1 mL) equilibrated with the dialysis buffer Proteins were eluted with a linear gradient of 50–500 mM NaCl in 20 mM Hepes/KOH (pH 7.0) (30 mL in total) and collected in fractions of 1.5 mL each Each fraction was assayed for GRK2, and subjected to SDS/PAGE on a 12% acrylamide gel The purified GRK2 was mixed with an equal volume of glycerol

as a stabilizer and stored at)80 Cuntil use Crude tubulin, which contained microtubule-associated proteins (MAPs), was prepared from porcine brains by the polymerization– depolymerization procedure, which was performed three times, as described previously [24] The tubulin was further purified by phosphocellulose chromatography [25]

Phosphorylation of tubulin Tubulin was phosphorylated with GRK2 as described previously with some modifications [19] Briefly, various concentrations of tubulin were incubated with 40 nM GRK2 in a buffer comprising 50 lM [c-32P]ATP (100 c.p.m.Æpmol)1), 20 mM Tris/HC l (pH 7.4), 50 mM KCl, 2 mMEDTA, 0.5 mMEGTA, and 5 mMMgCl2at

30C, followed by SDS/PAGE Incorporation of radio-activity into the tubulin was visualized by autoradiography and quantified with a Fuji BioImage BAS2000 analyzer

Overlaying and detection of GRK2 Purified tubulin was reduced and carboxymethylated essentially as described previously [26] Purified tubulin was lyophilized and then dissolved in 140 lL 6M guanidi-nium hydrochloride in 0.1M Tris/HCl (pH 8.5), 40 lL propan-2-ol, and 2 lL 2-mercaptoethanol by incubation at room temperature for 2 h The tubulin solution was then carboxymethylated by mixing it with 1 lL 1M iodoacetic acid in 1MNaOH, followed by incubation of the mixture at room temperature in the dark for 40 min The reaction was terminated by the addition of excess 2-mercaptoethanol, and then the mixture was passed through a column of Sephadex G-50 fine (2 mL) previously equilibrated with

50 mMammonium carbonate (pH 9.0) The carboxymethy-lated tubulin was subjected to SDS/PAGE and then transferred to a poly(vinylidene difluoride) (PVDF) mem-brane [27,28] The PVDF memmem-brane was incubated in blocking buffer [0.1% (v/v) Tween 20 and 5% (w/v) nonfat dry milk in NaCl/P] for 1 h at 4C, and subsequently

washed three times with binding buffer [0.1% (v/v)

Tween 20 and 0.5% (w/v) nonfat dry milk in NaCl/Pi]

The PVDF membrane was then incubated with GRK2 in

binding buffer overnight at 4C After the PVDF

mem-brane had been washed three times with binding buffer,

GRK2 was detected by incubating the PVDF membrane

with anti-GRK2 IgG For immunological detection,

horse-radish peroxidase-conjugated anti-IgG antibodies and an

ECL chemiluminescence system were used according to the

manufacturer’s instructions

Digestion of tubulin

Phosphorylated and then carboxymethylated tubulin

(100 lg) was treated with 1 lg Achromobacter protease I

(EC3.4.21.50) in 200 lL 100 mM ammonium carbonate

buffer (pH 9.0) at 37Cfor 60 min The digested peptides

were applied to a DEAE-5PW column equilibrated with

50 mM ammonium carbonate (pH 9.0)/50 mM NaCl at a

flow rate of 0.5 mLÆmin)1 After the column had been

washed, the peptides were eluted with a linear gradient of

50–500 mMNaCl in 50 mMammonium carbonate, pH 9.0

(30 mL in total) and collected in fractions of 1 mL each

Radioactivity was detected by Cerenkov counting The

fractions containing the phosphopeptides were combined

and digested overnight with 10 lg endoproteinase Asp-N at

30C The reaction product was applied directly to a C18

RP-HPLCcolumn, which was eluted with a linear gradient

of 0–50% acetonitrile containing 0.1% trifluoroacetic acid

in 50 min at a flow rate of 0.3 mLÆmin)1 The amino-acid

sequences of the radioactive peptides were determined with

a Hewlett–Packard G1000A Protein Sequencer

Phosphoamino-acid analysis by TLC

A portion of the radioactive peptides eluted from the C18

column was lyophilized, resuspended in 6MHCl, and then

hydrolyzed by incubation at 110Cfor 60 min The

hydrolysate was lyophilized and then subjected to TLC

with pyridine/acetic acid/water (1 : 10 : 189, v/v) The

radioactive phosphoamino acids were visualized by

auto-radiography The TLCplate was sprayed with 0.7%

ninhydrin in acetone and heated in an oven at 65Cto

visualize the standard phosphoamino acids

Cloning of b-tubulin and mutagenesis of its

phosphorylation sites

Poly(A)-rich RNA was prepared from rat and mouse brains

with Moloney murine leukemia virus reverse transcriptase

(Toyobo) and then used to construct a cDNA library The

DNA fragment encoding the full-length rat bI-tubulin

(accession No AB011679) or mouse bIII-tubulin (accession

no NM_023279) was amplified by PCR using the rat or

mouse brain cDNA library as a template The PCR

products were digested with EcoRI–NotI and then cloned

into plasmid vector pBluescript II KS(–) For construction

of mutant bI-tubulin and bIII-tubulin, Thr409 and Ser420 of

bI-tubulin and Thr409, Ser420 and Ser444 of bIII-tubulin

were replaced with Ala using the inverted amplification

method [29] Oligonucleotide primers were designed in

inverted tail-to-tail directions to amplify the cloning vector

together with the inserts PCR was performed with Pfu turbo polymerase cDNAs encoding the wild-type and mutant b-tubulins were excised as EcoRI–NotI fragments and then subcloned into EcoRI–NotI-digested expression vector pGEX4T-3, followed by transformation into Escherichia coli and expression as fusion proteins with glutathione S-transferase (GST; GST-b-tubulins) A fusion protein with GST of a peptide corresponding to positions 393–445 of rat bI-tubulin was also expressed in E coli using

an expression vector, pGEX4T-3 (GST-b-tubulinC) These GST fusion proteins were purified using glutathione-Seph-arose by the procedure recommended by the manufacturer,

as described previously [10]

Dephosphorylation of tubulin Tubulin purified from porcine brains was subjected to dephosphorylation with PP2A (0.2 U) at 30Cfor 60 min The dephosphorylation buffer contained 50 mM Mes (pH 6.8), 0.1 mM EDTA, 1 mM EGTA, 5 mM MgCl2, 0.2 mgÆmL)1 BSA, and 1 mM 2-mercaptoethanol The dephosphorylation reaction was terminated by adding

10 nM okadaic acid, followed by phosphorylation of the dephosphorylated tubulin by GRK2 at 30Cas above In the control sample, PP2A was incubated with 10 nM okadaic acid before the addition of tubulin To follow the time course of dephosphorylation, tubulin was first phos-phorylated in the presence of [c-32P]ATP by GRK2 and then subjected to dephosphorylation by PP2A, followed by SDS/PAGE and quantification of the radioactivity remain-ing in the tubulin

Results

GRK2 binds specifically to b-tubulin The a and b isotypes of tubulin were separated from each other by carboxymethylation and subsequent SDS/PAGE After electrophoresis and Western blotting, the PVDF membrane was incubated with a purified preparation of GRK2 and then GRK2 antibodies As shown in Fig 1, GRK2 was found to interact only with b-tubulin This is consistent with the study by Carman et al [21], in which GRK2 phosphorylated b-tubulin but not a-tubulin There-fore, these results indicate that GRK2 binds to and phosphorylates b-tubulin specifically

Partial digestion of phosphorylated tubulin Tubulin phosphorylated by GRK2 was cleaved on the C-terminal side of Lys with Achromobacter protease I and

on the C-terminal sides of Lys and Arg with trypsin The cleaved peptides were subjected to analysis by SDS/PAGE with 18% acrylamide in Tricine The smallest fragment obtained on treatment with Achromobacter protease I or trypsin had an apparent molecular mass of 6 kDa (data not shown) We examined the amino-acid sequence of b-tubulin, looking for the region with the expected length after the Achromobacter protease I or trypsin treatment, and found that the C-terminus of the b-tubulin had the most likely sequence to be phosphorylated by GRK2 The sequence between 392 and 430 is the same in the b and

bIIItubulin isotypes of pig, and bI, bII, bIII, and bIVisotypes

of mouse (Table 1) The sequence between residue 431 and

the C-terminus differs from one isotype to another, but

there are no Ser or Thr residues in the region except for

Ser444 in bIII-tubulin [30] Neither Lys nor Arg is present

between Lys392 and the C-terminus except for Lys450 in

bIII-tubulin Thus, the fragment obtained by treatment with

Achromobacterprotease I or trypsin is expected to have

53–58 residues, which corresponds to the phosphorylated

6-kDa band shown by SDS/PAGE This C-terminal region

of b-tubulin from Ala393 to Lys450 is extremely acidic with

20 acidic residues and only two basic His396 and Lys450 residues, and it has three Ser and three Thr residues

Separation of radiolabeled peptides and analysis

of phosphoamino acids Phosphorylated tubulin was carboxymethylated and then digested with Achromobacter protease I The digested sample was loaded on to a DEAE column All the radioactivity

Table 1 Sites and potential sites for GRK2-mediated phosphorylation Phosphorylation sites for GRK2 were identified for rhodopsin [2],

b 2 -adrenergic receptors [3], and a-synuclein and b-synuclein [22] Potential phosphorylation sites for GRK2 are indicated for a 2A -receptors [39], M 2

receptors [36], M 3 receptors [5], phosducin and phosducin-like protein [23] The phosphorylation sites and potential sites are indicated in bold type

as S or T Acidic and basic amino acids are denoted by italics and underlining, respectively.

Rhodopsin (bovine, 309–348, C-terminal) MNKQFRNCMLTTICCGKNPLGDDEA S ATV S KT E T S QVAPA-OH

b 2 -Adrenoceptor

(human, 362–413, C-terminal)

E QSGYHV E Q E K E NKLLC ED LPG TED FVGHQGTVP S NI DS QGRNC S TN D SLL-OH

a 2A -Adrenoceptor (human, 279–323,

third intracellular loop)

E PAPAGPR D T D AL D L EESSSSD HA E RPPGPRRP E RGPRGKGKARA

M 2 (human, 278–321, third intracellular loop) EE K E SSN DSTS VSAVASNMR DDE ITQ DE N T V STS LGHSK DE NSK

M 3 (rat, 317-361, third intracellular loop) KSWKPSA E QM D Q D H SSSD SWNNN D AAASL E N S A SSDEED IGS E TR

b III -Tubulin (mouse, 392–450, C-terminal) KAFLHWYTG E GM DE M E FT E A E SNMN D LV S YQQYQ D ATA DEE G E MY

EDDDEE S AQGPK-OH a-Synuclein (human, 103–140, C-terminal) N EE GAPQ E GIL ED MPV D P D N E AY E MP SEE GYQ D Y E P E A-OH

b-Synuclein (human, 95–134, C-terminal) P EE VAQ E AA EE PLI E PLM E P E G ES Y ED PPQ EE YQ E Y E P E A-OH

Phosducin (rat, 206–246, C-terminal) E QFA EE FFAA D V ES FLN E YGLLP E R E IH D LGQ T N TEDED I E -OH

Phosducin-like protein

(rat, 279–301, C-terminal)

VLVL TS VRNSATCH SEDSD L E I D -OH

Fig 1 b-Tubulin binds to GRK2 Purified

tubulin and carboxymethylated tubulin

(CM-tubulin) were subjected to SDS/PAGE

and then stained with Coomassie Brilliant

Blue or transferred to PVDF membranes The

PVDF membranes were incubated with

puri-fied GRK2 (0.6 lgÆmL)1) and then with

GRK2 antibodies, as described in Materials

and methods.

bound to the column, none being detected in the

flow-through fraction As shown in Fig 2A, most of the

radioactivity was eluted as a single peak The N-terminal

sequence determined for the 6-kDa fragment was

AFLHWYTGEG-, which was identical with the sequence

of residues 393–402 in the C-terminus of porcine b-tubulins

[30]

The 6-kDa fragment was further digested with endopro-teinase Asp-N and then subjected to RP-HPLCon a C18 column Phosphopeptides were eluted at 20% acetonitrile

as two peaks with a linear gradient of 0–50% acetonitrile (Fig 2B) The phosphopeptides obtained were analyzed by two-dimensional mapping on TLCplates Each of the two peak fractions from the C18column gave a single spot on TLCmapping (data not shown) Edman sequence analysis

of each peptide (peptide 1 and peptide 2) revealed that peptide 1 had the sequence DEMEFTEAESNMN(404– 416), and peptide 2 had the sequence DLVSEYQQYQ(417– 426) Acid hydrolysis followed by TLCanalysis revealed only labeled phosphothreonine on peptide 1 and labeled phosphoserine on peptide 2 (Fig 3) These results indicate that the phosphorylation sites for GRK2 are Thr409 and Ser420, but not Ser413

Phosphorylation by GRK2 of GST fusion proteins

of full-length bI-tubulin, bIII-tubulin, and C-terminal peptides of bI-tubulin expressed inE coli

We cloned the bI-tubulin and bIII-tubulin genes GST fusion proteins of bI-tubulin (GST-bI-tubulin), its C-terminal peptide (393–445) (GST-bI-tubulinC), and bIII-tubulin (GST-bIII-tubulin) were expressed in E coli and then subjected to phosphorylation by GRK2 with different substrate concentrations (Fig 4) Each substrate was found to be phosphorylated by GRK2 at similar rates The Kmvalues for GST-bI-tubulin, GST-bIII-tubulin and GST-bI-tubulinCwere estimated to be 2.6, 6 and 12 lM, respectively The K values for GST-b-tubulin and

Fig 2 Elution profile of 32 P-labeled peptides on DEAE and C 18

RP-HPLC columns (A) Phosphorylated tubulin was

carboxymethyl-ated, digested with Achromobacter protease I at 37 Cfor 30 min, and

then applied to a DEAE column (0.75 · 7.5 cm) The elution of

peptides and radioactivity was monitored by measuring A 280 (upper

line) and Cerenkov radiation (lower line), respectively Edman

degra-dation of the fraction, which included the major phosphopeptide,

revealed the N-terminal sequence to be AFLHWYTGEG(393–402).

(B) The pooled fractions from the DEAE column were digested

overnight with endoproteinase Asp-N at 30 C and then applied to a

C 18 RP-HPLCcolumn (0.46 · 25 cm) Elution was monitored by UV

absorption at 214 nm (upper line), and radioactivity was measured by

Cerenkov counting (lower line) Two phosphopeptides were eluted

from the C 18 RP-HPLCcolumn, which were subjected to Edman

degradation and sequence determination The sequence of peptide 1

was determined to be DEMEFTEAESNMN(404–416) and that of

peptide 2 to be DLVSEYQQYQ(417–426).

Fig 3 [ 32 P]Phosphoamino-acid analysis on TLC plates [32 P]Phospho-peptides eluted from a DEAE or C 18 RP-HPLCcolumn (peptide 1 and peptide 2) were partially hydrolyzed with HCl and then analyzed by TLC Autoradiograms of the TLC plates are shown, together with standard phosphoamino acids.

GST-bIII-tubulin are comparable to those reported for the

phosphorylation of tubulin purified from porcine brains

(0.4–3 lM) [19–21]

Phosphorylation by GRK2 of GST fusion proteins

of b-tubulin mutants

To confirm that Thr409 and Ser420 are the only

phos-phorylation sites for b-tubulin, we constructed mutants of

bI-tubulin with Ala409 and/or Ala420 in place of Thr409

and Ser420 In addition, we constructed mutants of bIII

-tubulin with Ala409, Ala420, and Ala444 or Ser444 to

examine whether Ser444 in bIII-tubulin is phosphorylated by

GRK2 These mutant forms were expressed in E coli as

GST fusion proteins and then analyzed with respect to their

phosphorylation by GRK2 As demonstrated in Fig 5A,

compared with the wild-type bI-tubulin (GST-bI-tubulin),

the mutant bI-tubulin (T409A and S420A) was less than

50% phosphorylated by GRK2 and the double mutant

bI-tubulin (T409A/S420A) was hardly phosphorylated at

all These results confirm that Thr409 and Ser420 are the

only residues in bI-tubulin phosphorylated by GRK2 On

the other hand, compared with wild-type bIII-tubulin

(GST-bIII-tubulin), the double mutant bIII-tubulin (T409A/

S420A) was 30% phosphorylated, and the triple mutant

bIII-tubulin (T409A/S420A/S444A) was hardly

phosphory-lated at all (Fig 5B) This result indicates that Ser444 of

bIII-tubulin is also a site of phosphorylation

Phosphorylation of phosphatase-treated tubulin

Tubulin purified from porcine brains was phosphorylated

with GRK2 and then dephosphorylated with PP2A About

80% of the phosphate was removed from the tubulin on treatment with 0.2 U PP2A for 40 min at 30C , as was the case for bIII-tubulin [31,32] (Fig 6) We treated the purified tubulin with PP2A and then phosphorylated it with GRK2

As shown in Fig 7, the amount of phosphorylation doubled

on pretreatment with PP2A This result indicates that tubulin had been phosphorylated when purified and that the endogenous phosphorylation is susceptible to PP2A and that the site can be phosphorylated by GRK2

Discussion

In this report, we have shown that the phosphorylation sites

in tubulin for GRK2 reside in the C-terminal domain of b-tubulin, and that two (Thr409 and Ser420) of five Ser or Thr residues in this domain are phosphorylated As the four isotypes of b-tubulin, bI, bII, bIIIand bIV, have the same sequence around the phosphorylation sites, it is most likely that all these isotypes serve as substrates for GRK2 The extent of phosphorylation by GRK2 was found to be increased when tubulin was pretreated with PP2A after its purification from porcine brain This result indicates that tubulin is phosphorylated in situ at sites from which phosphate may be removed by PP2A and to which phosphate may be added by GRK2 One of the most likely candidate sites is Ser444 in bIII-tubulin, although it is also possible that Thr409, Ser420, and other residues are the relevant sites Evidence for this is that Ser444 has been identified as the phosphorylation site in brain-specific bIII -tubulin phosphorylated in cultured cells [33] and in the brain [34] Furthermore, Khan and coworkers have reported that phosphate on the Ser444 residue of bIII-tubulin is resistant

to a wide variety of phosphatases, except human erythrocyte

Fig 4 Phosphorylation by GRK2 of GST fusion proteins of b I -tubulin and b III -tubulin (GST-b-tubulin) and the C-terminal peptide of b I -tubulin

(GST-b I -tubulinC) The indicated concentrations of GST fusion proteins were subjected to phosphorylation with GRK2 in the presence of 50 l M [c-32P]ATP and 40 n M GRK2 for 10 min, followed by SDS/PAGE, and radioactivity counting of the tubulin band Molar concentrations of fusion proteins were calculated from the molecular mass: GST, 27.5 kDa; GST-b I -tubulin and GST-b III -tubulin, 82.5 kDa; GST-b I -tubulinC, 33.5 kDa.

C urves were fitted to the Michaelis–Menten equation, and K m values were estimated to be 2.5 l M (GST-b I -tubulin), 6 l M (GST-b III -tubulin) and

12 l M (GST-b I -tubulinC) These experiments were repeated three times with essentially the same results.

PP2A [31], which is known to bind to polymerized tubulin

[32] Moreover, we demonstrated phosphorylation of

Ser444 by GRK2 using recombinant tubulin mutants,

although the phosphorylation was not detected for tubulin

purified from porcine brain These results suggest that

GRK2 is the kinase that phosphorylates Ser444, although

we cannot exclude the involvement of other kinases such as

casein kinase II [35] This assumption is supported by the

observation that GRK2 is localized with microtubules in

intact cells and the localization is facilitated by

agonist-bound GPCRs [20]

The phosphorylation sites for GRK2 have been

deter-mined for rhodopsin [2], b2-adrenergic receptors [3], and

synucleins [22] Serine and threonine clusters have also been

shown to be phosphorylation sites for GRK2 in M2

receptors [36], M3 receptors [5], a2A-adrenergic receptors

[37], and phosducin [23], although the phosphorylated

amino-acid residues have not been determined definitely

These phosphorylation sites and phosphorylation site

candidates are shown in Table 1 Each of these

phosphory-lation sites resides in an acidic domain with a fairly long

span It may be a prerequisite for phosphorylation by

GRK2 that the phosphorylation sites are in an acidic

domain However, it is not the only condition for

phosphorylation that the Ser and Thr residues are in the

acidic domain, as GRK2 phosphorylated Thr409, Ser420

and Ser444 but not Thr399, Ser413, and Thr429 in the

C-terminal domain of b -tubulin Further research is

necessary into what discriminates phosphorylated from nonphosphorylated residues

Initially, we hypothesized that tubulin may serve as both

a substrate and an activator for GRK2 and that it contains

a basic GRK2-activating domain besides a substrate domain This working hypothesis is not supported by the present findings that the C-terminal peptide of bI-tubulin (bI-tubulinC), which is very acidic and does not contain a basic domain, is as good a substrate as full-length tubulin Even if tubulin contains a basic GRK2-activating domain, the effect of the putative domain should not be important because the Kmvalues for bI-tubulin and bI-tubulinConly differ by a factor of 5 Therefore, it is likely that tubulin is a substrate for GRK2 for a different reason from that in the case of agonist-bound GPCRs

Synucleins and phosducin have been reported to be substrates for GRK2, but it is not known if they have basic domains which serve as activators for GRK2 However, we have noticed a common characteristic of tubulin, synucleins, and phosducin, i.e all three proteins have very acidic C-terminal domains that include phosphorylation sites The C-terminal domain of bIII-tubulin contains 20 acidic residues in a span of 58 residues (35%) with only two basic residue (His and Lys) The C-terminal domains of synuc-leins (a and b) also contain phosphorylation sites in very acidic domains, with 37–40% of acidic residues and no basic residues (Table 1) The phosphorylation sites in GPCRs are also in an acidic domain, but the acidic nature is much less

Fig 5 Phosphorylation of GST-b I -tubulin and

b III -tubulin mutants (A) Residues Thr409 and Ser420, and both Thr409 and Ser420 residues

in b I -tubulin were replaced with alanine resi-dues, yielding mutants T409A, S420A, and T409A/S420A, respectively (B) Residues Thr409 and Ser420 and/or Ser444 residues were replaced with alanine residues, yielding mutants T409A/S420A and T409A/S420A/ S444A GST fusion proteins of these mutants were expressed in E coli and then purified as described in Materials and methods These b-tubulin mutants were subjected to phos-phorylation with GRK2 The values are the means of three independent experiments for each b I -tubulin and b III -tubulin mutants with similar results and are expressed as percent-ages of the control value for wild-type GST-b I

or b III -tubulin Error bars represent means

± SD.

evident The presence of very acidic domains, particularly in

the C-termini of nonreceptor substrates, may constitute a

criterion for phosphorylation by GRK2

The C-terminal domain containing Thr409 and Ser420

has been shown to form an a-helix (H12, 408–423 residues)

and to be located on the outermost surface of microtubules

[38] The C-terminal residues including Ser444 of b-tubulin,

which are lacking in the structure model, are also thought to

be located on the outermost surface of microtubules This is

consistent with the findings that microtubules as well as

tubulin dimer can be phosphorylated by GRK2 and that

phosphorylated tubulin can polymerize into microtubules

[19] The C-termini of a and b tubulin are thought to

be involved in the binding of MAPs and motor proteins

[39,40] MAPs and motor proteins are known to have major

roles in microtubule assembly, organelle transport,

and mitosis It is possible that phosphorylation of the

C-terminus of b-tubulin by GRK2 affects the microtubule

dynamics or cellular mechanisms by affecting the binding of

MAPs or motor proteins In addition, a series of recent

studies have demonstrated that Ga or Gbc subunits interact

directly with tubulin [41–43] and that muscarinic receptor

activation induces transient translocation of tubulin to the

plasma membrane [44] Furthermore, microtubules have

been suggested to mediate the internalization of b-adrener-gic receptors [45] The GRK-mediated phosphorylation of tubulin may affect physiological processes including GPCRs, and the interaction of GRK2 with tubulin may have an effect on the function of GRK2

Acknowledgements

We thank Professor K Matsushima and Mr Y Terashima for their help in determining the peptide sequences This work was supported in part by grants from the Japan Society for the Promotion of Science (Research for Future Program), and from the Japan Science and Technology Corporation (CREST).

References

1 Pitcher, J.A., Freedman, N.J & Lefkowitz, R.J (1998) G protein-coupled receptor kinases Annu Rev Biochem 67, 653–692.

2 Ohguro, H., Rudnicka-Nawrot, M., Buczylko, J., Zhao, X., Taylor, J.A., Walsh, K.A & Palczewski, K (1996) Structural and enzymatic aspects of rhodopsin phosphorylation J Biol Chem.

271, 5215–5224.

Fig 7 Effects of PP2A treatment on phosphorylation of tubulin by GRK2 Purified tubulin was incubated with or without PP2A for

60 min at 30 Cfollowed by phosphorylation by GRK2 at 30 Cfor

30 min In samples designated PP2A + okadaic acid, PP2A was preincubated with okadaic acid before the phosphorylation reaction The indicated values are means of triplicate determinations in one of three independent experiments with similar results Error bars repre-sent means ± SD.

Fig 6 Dephosphorylation by PP2A of phosphorylated tubulin Tubulin

purified from porcine brains was phosphorylated with [c-32P]ATP by

GRK2 at 30 Cfor 30 min as described above and then incubated with

or without PP2A (0.2 U) for the indicated time at 30 C, followed by

SDS/PAGE and counting of radioactivity on tubulin band The

indicated values are means of triplicate determinations in one of three

independent experiments with similar results and are expressed as

percentages of the amount of tubulin phosphorylated before addition

of PP2A Error bars represent means ± SD.

3 Fredericks, Z.L., Pitcher, J.A & Lefkowitz, R.J (1996)

Identifi-cation of the G protein-coupled receptor kinase phosphorylation

sites in the human b 2 -adrenergic receptor J Biol Chem 271,

13796–13803.

4 Nakata, H., Kameyama, K., Haga, K & Haga, T (1994)

Loca-tion of agonist-dependent-phosphorylaLoca-tion sites in the third

intracellular loop of muscarinic acetylcholine receptors (m2

subtype) Eur J Biochem 220, 29–36.

5 Wu, G., Bogatkevich, G.S., Mukhin, Y.V., Benovic, J.L.,

Hilde-brandt, J.D & Lanier, S.M (2000) Identification of G bc binding

sites in the third intracellular loop of the M(3)-muscarinic receptor

and their role in receptor regulation J Biol Chem 275, 9026–

9034.

6 Liggett, S.B., Ostrowski, J., Chesnut, L.C., Kurose, H., Raymond,

J.R., Caron, M.G & Lefkowitz, R.J (1992) Sites in the third

intracellular loop of the a 2A -adrenergic receptor confer short term

agonist-promoted desensitization J Biol Chem 267, 4740–4746.

7 Onorato, J.J., Palczewski, K., Regan, J.W., Caron, M.G.,

Lef-kowitz, R.J & Benovic, J.L (1991) Role of acidic amino acids in

peptide substrates of the b-adrenergic receptor kinase and

rho-dopsin kinase Biochemistry 30, 5118–5125.

8 Fowles, C., Sharma, R & Akhtar, M (1988) Mechanistic studies

on the phosphorylation of photoexcited rhodopsin FEBS Lett.

238, 56–60.

9 Chen, C.Y., Dion, S.B., Kim, C.M & Benovic, J.L (1993)

Beta-adrenergic receptor kinase Agonist-dependent receptor binding

promotes kinase activation J Biol Chem 268, 7825–7831.

10 Kameyama, K., Haga, K., Haga, T., Moro, O & Sadee, W (1994)

Activation of a GTP-binding protein and a

GTP-binding-protein-coupled receptor kinase (b-adrenergic-receptor kinase-1) by a

muscarinic receptor m2 mutant lacking phosphorylation sites.

Eur J Biochem 226, 267–276.

11 Haga, K & Haga, T (1992) Activation by G protein beta gamma

subunits of agonist- or light-dependent phosphorylation of

mus-carinic acetylcholine receptors and rhodopsin J Biol Chem 267,

2222–2227.

12 Pitcher, J.A., Inglese, J., Higgins, J.B., Arriza, J.L., Casey, P.J.,

Kim, C., Benovic, J.L., Kwatra, M.M., Caron, M.G & Lefkowitz,

R.J (1992) Role of bc subunits of G proteins in targeting the

beta-adrenergic receptor kinase to membrane-bound receptors Science

257, 1264–1267.

13 Kameyama, K., Haga, K., Haga, T., Kontani, K., Katada, T &

Fukada, Y (1993) Activation by G protein beta gamma subunits

of beta-adrenergic and muscarinic receptor kinase J Biol Chem.

268, 7753–7758.

14 Haga, K., Kameyama, K & Haga, T (1994) Synergistic activation

of a G protein-coupled receptor kinase by G protein beta gamma

subunits and mastoparan or related peptides J Biol Chem 269,

12594–12599.

15 Palczewski, K., Buczylko, J., Kaplan, M.W., Polans, A.S &

Crabb, J.W (1991) Mechanism of rhodopsin kinase activation.

J Biol Chem 266, 12949–12955.

16 Moro, O., Lameh, J., Hogger, P & Sadee, W (1993) Hydrophobic

amino acid in the i2 loop plays a key role in receptor-G protein

coupling J Biol Chem 268, 22273–22276.

17 Wess, J., Liu, J., Blin, N., Yun, J., Lerche, C & Kostenis, E (1997)

Structural basis of receptor/G protein coupling selectivity studied

with muscarinic receptors as model systems Life Sci 60, 1007–1014.

18 Higashijima, T., Uzu, S., Nakajima, T & Ross, E.M (1988)

Mastoparan, a peptide toxin from wasp venom, mimics receptors

by activating GTP-binding regulatory proteins (G proteins).

J Biol Chem 263, 6491–6494.

19 Haga, K., Ogawa, H., Haga, T & Murofushi, H (1998)

GTP-binding-protein-coupled receptor kinase 2 (GRK2) binds and

phosphorylates tubulin Eur J Biochem 255, 363–368.

20 Pitcher, J.A., Hall, R.A., Daaka, Y., Zhang, J., Ferguson, S.S., Hester, S., Miller, S., Caron, M.G., Lefkowitz, R.J & Barak, L.S (1998) The G protein-coupled receptor kinase 2 is a microtubule-associated protein kinase that phosphorylates tubulin J Biol Chem 273, 12316–12324.

21 Carman, C.V., Som, T., Kim, C.M & Benovic, J.L (1998) Binding and phosphorylation of tubulin by G protein-coupled receptor kinases J Biol Chem 273, 20308–20316.

22 Pronin, A.N., Morris, A.J., Surguchov, A & Benovic, J.L (2000) Synucleins are a novel class of substrates for G protein-coupled receptor kinases J Biol Chem 275, 26515–26522.

23 Ruiz-Gomez, A., Humrich, J., Murga, C., Quitterer, U., Lohse, M.J & Mayor, F Jr (2000) Phosphorylation of phosducin and phosducin-like protein by G protein-coupled receptor kinase 2.

J Biol Chem 275, 29724–29730.

24 Shelanski, M.L., Gaskin, F & Cantor, C.R (1973) Microtubule assembly in the absence of added nucleotides Proc Natl Acad Sci USA 70, 765–768.

25 Weingarten, M.D., Lockwood, A.H., Hwo, S.Y & Kirschner, M.W (1975) A protein factor essential for microtubule assembly Proc Natl Acad Sci USA 72, 1858–1862.

26 Cresfield, A.M., Moore, S & Stein, W.H (1963) The preparation and enzymatic hydrolysis of reduced and S-carboxymethylated proteins J Biol Chem 238, 622–627.

27 Laemmli, U.K (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4 Nature (London) 227, 680–685.

28 Towbin, H., Staehelin, T & Gordon, J (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications Proc Natl Acad Sci USA 76, 4350–4354.

29 Imai, Y., Matsushima, Y., Sugimura, T & Terada, M (1991) A simple and rapid method for generating a deletion by PCR Nucleic Acid Res 19, 2785–2785.

30 Sullivan, K.F (1988) Structure and utilization of tubulin isotypes Annu Rev Cell Biol 4, 687–716.

31 Khan, I.A & Luduena, R.F (1996) Phosphorylation of beta III-tubulin Biochemistry 35, 3704–3711.

32 Hiraga, A & Tamura, S (2000) Protein phosphatase 2A is asso-ciated in an inactive state with microtubules through 2A1-specific interaction with tubulin Biochem J 346, 433–439.

33 Luduena, R.F., Zimmermann, H.P & Little, M (1988) Identifi-cation of the phosphorylated beta-tubulin isotype in differentiated neuroblastoma cells FEBS Lett 230, 142–146.

34 Diaz-Nido, J., Serrano, L., Lopez-Otin, C., Vandekerckhove, J & Avila, J (1990) Phosphorylation of a neuronal-specific beta-tubulin isotype J Biol Chem 265, 13949–13954.

35 Serrano, L., Diaz-Nido, J., Wandosell, F & Avila, J (1987) Tubulin phosphorylation by casein kinase II is similar to that found in vivo J Cell Biol 105, 1731–1739.

36 Pals-Rylaarsdam, R., Gurevich, V.V., Lee, K.B., Ptasienski, J.A., Benovic, J.L & Hosey, M.M (1997) Internalization of the m2 muscarinic acetylcholine receptor Arrestin-independent and -dependent pathways J Biol Chem 272, 23682–23689.

37 Eason, M.G., Moreira, S.P & Liggett, S.B (1995) Four consecutive serines in the third intracellular loop are the sites for b-adrenergic receptor kinase-mediated phosphorylation and desensitization of the a 2A -adrenergic receptor J Biol Chem 270, 4681–4688.

38 Nogales, E., Wolf, S.G & Downing, K.H (1998) Structure of the alpha beta tubulin dimer by electron crystallography Nature (London) 391, 199–203.

39 Serrano, L., Avila, J & Maccioni, R.B (1984) Controlled pro-teolysis of tubulin by subtilisin: localization of the site for MAP2 interaction Biochemistry 23, 4675–4681.

40 Wang, Z & Sheetz, M.P (2000) The C-terminus of tubulin

increases cytoplasmic dynein and kinesin processivity Biophys.

J 78, 1955–1964.

41 Wang, N., Yan, K & Rasenick, M.M (1990) Tubulin binds

specifically to the signal-transducing proteins, Gs alpha and Gi

alpha 1 J Biol Chem 265, 1239–1242.

42 Popova, J.S., Garrison, J.C., Rhee, S.G & Rasenick, M.M (1997)

Tubulin, Gq, and phosphatidylinositol 4,5-bisphosphate interact

to regulate phospholipase Cbeta1 signaling J Biol Chem 272,

6760–6765.

43 Roychowdhury, S & Rasenick, M.M (1997) G protein beta1-gamma2 subunits promote microtubule assembly J Biol Chem.

272, 31576–31581.

44 Popova, J.S & Rasenick, M.M (2000) Muscarinic receptor acti-vation promotes the membrane association of tubulin for the regulation of Gq-mediated phospholipase Cbeta (1) signaling.

J Neurosci 20, 2774–2782.

45 Limas, C.J & Limas, C (1985) Carbachol induces desensitization

of cardiac beta-adrenergic receptors through muscarinic M1 receptors Biochem Biophys Res Commun 128, 699–704.