GIT1 Mediates HDAC5 Activation by Angiotensin II in Vascular Smooth Muscle Cells

GIT1 Mediates HDAC5 Activation by Angiotensin II inVascular Smooth Muscle Cells

Jinjiang Pang, Chen Yan, Kanchana Natarajan, Megan E. Cavet, Michael P. Massett,Guoyong Yin, Bradford C. Berk

Objective—The G protein–coupled receptor (GPCR)-kinase2 interacting protein1 (GIT1) is a scaffold protein involved inangiotensin II (Ang II) signaling. Histone deacetylase-5 (HDAC5) has emerged as an important substrate ofcalcium/calmodulin-dependent protein kinase II (CamK II) in GPCR signaling. Here we investigated the hypothesis thatAng II–mediated vascular smooth muscle cell (VSMC) gene transcription involves GIT1-CamK II–dependentphosphorylation of HDAC5.

Methods and Results—Ang II rapidly stimulated phosphorylation of HDAC5 at Ser498 in VSMCs. Knockdown of GIT1significantly decreased HDAC5 phosphorylation induced by Ang II. The involvement of Src, phospholipase � (PLC�),and CamK II in GIT1-mediated HDAC5 phosphorylation was demonstrated. The association of GIT1 and CamK II wasconstitutive but increased after stimulation with Ang II. Moreover, the interaction of GIT1 and CamK II through theARF GTPase-activating protein (ARF-GAP) and coiled-coil domains of GIT1 was essential for the phosphorylation ofHDAC5. Finally, knockdown of GIT1 decreased myocyte enhancer factor 2 transcriptional activity induced by Ang II.

Conclusions—This study identifies a novel function for GIT1 as a mediator of Ang II–induced VSMC gene transcriptionvia a Src-PLC�-CamK II-HDAC5 signaling pathway. (Arterioscler Thromb Vasc Biol 2008;28:892-898)

Key Words: G protein–coupled receptor-kinase interacting protein1 � GIT1 � angiotensin II� histone heacetylase 5 � Ca2�-calmodulin-dependent protein kinase II � VSMC

The importance of the renin angiotensin system (RAS) incardiovascular disease has been dramatically shown by

the beneficial effects of inhibiting the RAS with angiotensin-converting enzyme inhibitors and angiotensin II (Ang II) type1 receptor (AT1R) blockers. The AT1R mediates VSMCmigration, hypertrophy, proliferation, and vascular remodel-ing.1–3 The G protein–coupled receptor (GPCR)-kinase 2interacting protein-1 (GIT1) is a multi-domain scaffold pro-tein involved in multiple GPCR signal pathways includingendocytosis, cell adhesion, and migration.4 Our laboratorywas the first to show that c-Src phosphorylates GIT15 inVSMCs after stimulation by Ang II. Furthermore, we showedthat GIT1 binds key signaling mediators including PLC� andMEK1.5–7 Conformational changes in GIT1 induced by ty-rosine phosphorylation lead to the activation of PLC�.5

Activation of PLC� is responsible for the elevation ofintracellular calcium, which results in the autophosphoryla-tion (activation) of calcium/calmodulin-dependent proteinkinase II (CamK II).8,9

Histone acetylation/deacetylation has emerged as a funda-mental mechanism for the control of gene expression. Histoneacetyltransferases stimulate transcription through acetylationof histones, resulting in relaxation of nucleosomes, and

histone deacetylases (HDACs) antagonize this activity andrepress transcription.10 Class II HDACs (HDACs 4, 5, 7,and 9) appear to be dedicated to the control of tissue growthand development. Phosphorylation of the amino termini ofclass II HDACs by CamK II creates docking sites for the14-3-3 family of chaperone proteins, which promote shuttlingof these HDACs from the nucleus to the cytoplasm, therebyderepressing HDAC target genes.11–13 HDAC5 acts as anegative regulator of cardiac growth. Transgenic mice lack-ing either HDAC5 or HDAC9 develop extremely enlargedhearts in response to pathological signals.14,15 CamK II playsan important role in HDAC5 phosphorylation induced byGPCR ligands.16 Based on these findings, we hypothesizedthat GIT1 mediates HDAC5 phosphorylation stimulated byAng II via a pathway dependent on c-Src, PLC�, and CamKII. Furthermore we propose that derepression of HDAC5increases MEF2 transcriptional activity in VSMCs.

Materials and MethodsMaterialsAntibodies to glutathione S-transferase (GST) monoclonal, GIT1polyclonal, CamK II polyclonal, 14-3-3 ployclonal, and ERK1monoclonal were from Santa Cruz Biotechnology Inc. c-Src (for

Original received December 18, 2007; final version accepted February 13, 2008.From the Aab Cardiovascular Research Institute and the Department of Medicine, University of Rochester School of Medicine and Dentistry,

Rochester, New York.Correspondence to Bradford C. Berk, MD, PhD, Aab Cardiovascular Research Institute and Department of Medicine, University of Rochester Medical

Center, 601 Elmwood Avenue, Box 679, Rochester, NY 14642. E-mail [email protected]© 2008 American Heart Association, Inc.

Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.107.161349

892 by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from

http://atvb.ahajournals.org/









chicken) and PLC� antibody was from BD Transduction Laborato-ries. FLAG M2 monoclonal antibody was from Sigma. Xpressmonoclonal antibody was from Invitrogen. Phosphospecific-ERK1/2was from Cell Signal. Phospho-Ser498 HDAC5 and Ser498 HDAC5antibodies were from Signal Antibody Technology. Ang II was fromICN Biomedicals. PP2, U73122, and KN93 were from Calbiochem.Other chemicals were purchased from Sigma.

Cell Culture and TransfectionVSMCs were obtained from rat aorta as described,17 and passages 6to 10 were used in the experiments. HEK293 cells and VSMCs werecultured in Dulbecco modified Eagle medium supplemented with10% fetal bovine serum, penicillin, and streptomycin at 37°C in 5%CO2. HEK293T cells were transfected by Lipofectamine/plus (In-vitrogen). For cotransfection with AT1R, a ratio of 3:1 was used.After allowing protein expression for 24 hours, cells were serum-deprived for 16 hours and stimulated with 100 nmol/L Ang II. GIT1siRNA (AAGCTGCCAAGAAGAAGCTAC) and control nonsilenc-ing siRNA(AATTCTCCGACACGTGTCACT) were designed asdescribed and synthesized by Ambion. VSMCs were transfectedusing Lipofectamine 2000 (Invitrogen) according to the manufactur-er’s protocol as described previously.18 GIT1 siRNA were preparedand transfected at 1 �mol/L for 48 hours as previously described,6then serum-deprived for 16 hours, and stimulated with 100 nmol/LAng II.

Plasmid ConstructsThe mGIT1 expressed sequence tag clone (GenBank accessionnumber AI414223) was purchased and completely sequenced. Then,full-length mGIT1 (GIT1[WT]) was cloned into the NotI and Xbalsites of Xpress-pcNDA3.1 vector (Invitrogen; Xpress-GIT1[WT]).Using polymerase chain reaction (PCR), GIT1(1 to 420aa),GIT1(420 to 770aa) were cloned into Xpress-pcDNA3.1 vector(Xpress-GIT1 [1 to 635aa], Xpress-GIT1 [1 to 420aa], Xpress-GIT1[420 to 770aa]). pEBB-Flag-CamK II was a generous gift from DrRichard A. Mauzer (Department of Cell and Developmental Biology,L215, Oregon Health & Science University, 3181 South West SamJackson Park Road, Portland, Oregon, 97239).

Immunoprecipitation and ImmunoblottingFor immunoprecipitations, cells were lysed in RIPA buffer(150 mmol/L NaCl, 1% Nonidet P-40, 0.5% deoxycholic acid, 0.1%SDS, 50 mmol/L Tris-HCl, pH 8.0). Protein concentrations in thelysates were determined as described.14,19 The protein samples wereseparated by SDS-PAGE, transferred to nitrocellulose membranes,and incubated with appropriate primary antibodies. After incubatingwith fluorescence-conjugated secondary antibodies, immunoreactiveproteins were visualized by an Odyssey infrared imaging system(LI-COR Biotechnology). Densitometric analysis of the blots wasperformed with Odyssey software (LI-COR Biotechnology). Resultswere normalized by arbitrarily setting the densitometry of controlsamples to 1.0.

ImmunofluorescenceVSMCs cells were starved with serum-free DMEM overnight andthen stimulated with Ang II in 0, 5, 10, 30, and 60 minutes. Cellswere fixed with 4% formaldehyde for 10 minutes, washed withphosphate-buffered saline 3 times, permeabilized with 0.05% Tritonfor 5 minutes, and blocked with 10% normal goat serum for 1 hour.Cells were incubated with GIT1, and CamK II antibodies diluted inphosphate-buffered saline followed by Alexa Fluor 546 antirabbitIgG for red fluorescence or by Alexa Fluor 488 goat antimouse forgreen fluorescence (Molecular Probes Inc) in phosphate-bufferedsaline at a final concentration of 1.5 to 2 �g/mL each.

Luciferase Reporter AssayTo assess myocyte enhancer factor 2 (MEF2) transcriptional activa-tion, we used 3� MEF2-dependent reporter gene [generous gift fromDr Joseph Miano (Aab Cardiovascular Research Institute, Universityof Rochester School of Medicine and Dentistry, Rochester, NY

14642)] in which 3 tandem repeats of MEF2 sites were locatedupstream of the thymidine kinase gene promoter. VSMCs cultured in24-well dishes were cotransfected with 3� MEF2 luciferase reportergene, thymidine kinase-renilla-luciferase (an internal control fromPromega), Myc-HDAC5-WT plasmid (gift from Dr Eric N. Olson,University of Texas Southwestern Medical Center, Dallas, Texas)control siRNA, or GIT1 siRNA in each experiment using electropo-ration (Bio-Rad). At 32 hours posttransfection, cells were treatedwith Ang II for another 16 hours. The luciferase activities in celllysates were determined using the Dual-Luciferase Reporter Assaykit (Promega) and Wallac 1420 multilabel counter (PerkinElmer).

Statistical AnalysisAll values are expressed as means�SD from 3 to 6 independentexperiments. The significance of the results was assessed by t test. Aprobability value �0.05 was considered statistically significant.

ResultsAng II Stimulates Phosphorylation of HDAC5 inRat VSMCsBecause HDAC5 is phosphorylated by CamK II and weshowed that GIT1 phosphorylation by Ang II is required forPLC�-mediated calcium mobilization and CamK II activa-tion, we studied HDAC5 phosphorylation in VSMCs inresponse to Ang II. Phosphorylation of HDAC5 was deter-mined by phospho-HDAC5–specific antibody, which recog-nizes phosphorylation at Ser498. In response to 100 nmol/LAng II, HDAC5 phosphorylation rapidly increased by 2.6-fold within 2 minutes, and reached a maximum at 5 minutes(3.3-fold; Figure 1A). HDAC5 phosphorylation returned tobaseline after 30 minutes (Figure 1A). HDAC5 expressiondid not change during this time course.

GIT1 Is Required for Phosphorylation of HDAC5by Ang IITo show a role for GIT1 in HDAC5 phosphorylation, we usedrat GIT1 siRNA to decrease GIT1 expression in VSMCs aspreviously described.6 Rat control siRNA and rat GIT1siRNAwere designed based on unique sequences and ability toinhibit mRNA expression. Transfection of GIT1 siRNAsignificantly decreased GIT1 protein expression at 48 hours,whereas HDAC5 expression was not altered (Figure 1B).Treatment with rat GIT1 siRNA significantly decreasedHDAC5 phosphorylation induced by Ang II (80% inhibition),whereas control siRNA had no significant effect (Figure 1B).

Ang II Stimulates HDAC5 Phosphorylation viaSrc-Dependent PathwayPrevious data from our laboratory suggested an essential rolefor c-Src in AT1R signal transduction.20 To investigate therole of c-Src in HDAC5 phosphorylation, the c-Src inhibitor4-amino-5-(4-chlorophenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine (PP2) was administrated for 30 minutes inVSMCs, before stimulation with 100 nmol/L Ang II for 5minutes. Phosphorylation of HDAC5 was significantly de-creased by PP2 treatment (supplemental Figure IA). Tofurther confirm the role of c-Src, the dominant negative (DN)chicken c-Src adenovirus (Ad.DN-Src) was used to infectVSMCs.20 Infection of VSMCs with Ad.DN-Src resulted inrobust expression of chicken Src (supplemental Figure IB).Infection at MOI of 100 and 300 almost completely blocked

Pang et al GIT1 Mediates HDAC5 Activation by Angiotensin II 893

by guest on June 30, 2015http://atvb.ahajournals.org/Downloaded from


Ang II–induced HDAC5 phosphorylation, whereas Ad.Lac-Zinfection had no significant effect. In contrast to the dramaticeffect of Ad.DN-Src on p-HDAC5, there was significantlyless inhibition of p-ERK1/2 (supplemental Figure IB).

PLC� and CamK II Are Required for HDAC5Phosphorylation by Ang IIc-Src phosphorylates GIT1, which is bound to PLC� in acomplex under basal conditions. Conformational changes inGIT1 induced by tyrosine phosphorylation lead to the acti-vation of PLC� which is required for elevation of intracellu-lar calcium, and autophosphorylation of CamK II. To deter-mine the involvement of PLC� and CamK II in HDAC5phosphorylation in VSMCs, the effects of the PLC� inhibitor,U73122, and the CamK II inhibitor KN93 were investigated.

Both U73122 and KN93 dose dependently inhibited thephosphorylation of HDAC5 (supplemental Figure IC and ID).These results suggest that PLC� and CamK II play importantroles in the Ang II–stimulated HDAC5 signaling pathway.

Association of CamK II, HDAC5, and 14-3-3 IsIncreased by Ang IIPhosphorylation of HDAC5 creates binding sites for the14-3-3 proteins, which escort p-HDAC5 from the nucleus tothe cytoplasm, with consequent activation of HDAC5 targetgenes. To investigate the association of CamK II withHDAC5 and 14-3-3 after stimulation of Ang II in VSMCs,coimmunoprecipitation was performed. The interaction ofCamK II with HDAC5 and 14-3-3 increased rapidly (within1 minute, Figure 2A), and peaked at 5 minutes, similar to thepeak phosphorylation of HDAC5.

The Interaction of GIT1 and CamK II IsIncreased by Ang IIBecause GIT1 is a multidomain scaffold protein,4 we hypoth-esized that GIT1 functions as a scaffold for CamK II bybinding to CamK II. The association of GIT1 and CamK IIwas assayed by immunoprecipitating CamK II from VSMClysates. Whereas the binding of GIT1 and CamK II wasconstitutive (Figure 2B, time 0, supplemental Figure II), inresponse to Ang II binding rapidly increased and peaked at 5minutes (2.6-fold increase, Figure 2B, supplemental FigureII), similar to the HDAC5 phosphorylation time course. Thetime course for GIT1-CamK II binding was similar to thetime course for CamK II binding to HDAC5 and 14-3-3(Figure 2A, supplemental Figure II), suggesting a multipro-tein complex. To investigate further the potential interactionof HDAC5 and GIT1, we coexpressed them in HEK293 cells(which have low expression of CamK II). There was no basalinteraction, nor increase in response to Ang II (supplementalFigure III).

GIT1 Recruited CamK II and PLC� to Form a“Calciosome” Complex, Which Mediated HDAC5Phosphorylation by Ang IIBecause GIT1 can associate with both CamK II and PLC�,we hypothesized that GIT1, CamK II, and PLC� could forma complex required for phosphorylation of HDAC5, whichwe will term the “ Calciosome.” The association of GIT1,CamK II, PLC�, and HDAC5 was assayed by immunopre-cipitating PLC� and GIT1 from VSMC lysates. In VSMCs,GIT1, CamK II, and PLC� were present in the same complexas shown by the findings that any one of these proteinscoprecipitated the other two proteins (Figure 2C and 2D) ifGIT1 was present. This presumed role of GIT1 as a scaffoldmolecule for CamK II and PLC� was proved using 293 cells,which express no detectable GIT1. In 293 cells transfectedwith vector control (pcDNA), there was no interaction be-tween CamK II and PLC� (supplemental Figure IV). How-ever, when cells were transfected with GIT1 cDNA copre-cipitation of PLC� and CamK II was readily apparent(supplemental Figure IV). The binding of HDAC5 to thecalciosome increased in response to Ang II stimulationthrough binding to CamK II (Figure 2C and 2D, supplemental

IB: p-HDAC5

IB: HDAC5

0

1

2

3

4

0 1 2 5 10 15 30Ang II,min

5C

AD

H/5C

AD

Hp

esaer cnI

dlo

F

*

*

*

*

IB: p-HDAC5

IB: GIT1

IB: HDAC5

*

#

0

1

2

3

4

5

- + - +ControlsiRNA

GIT1siRNA

AngII

B

A VSMC

VSMC

5C

AD

H/5C

AD

Hp

esaercnI

dlo

F

Figure 1. GIT1-mediated HDAC5 phosphorylation is inducedby Ang II. A, Ang II stimulates HDAC5 phosphorylation inVSMCs. B, Knockdown of GIT1 decreased Ang II–stimulatedHDAC5 phosphorylation. The lower panel shows relativeincrease of HDAC5 phosphorylation compared to control*P�0.05 (mean�SD; n�3).

894 Arterioscler Thromb Vasc Biol May 2008



Figure II). The peak time was at 5 minutes, similar to peaktime of HDAC5 phosphorylation. These data suggested thatthe calciosome is essential for HDAC5 phosphorylation.

Colocalization of GIT1 and CamK IITo further confirm the interaction of GIT1 and CamK II,immunofluorescence experiments were performed. After se-rum starvation for 24 hours, VSMCs were incubated withAng II for up to 60 minutes. GIT1 and CamK II location wasassayed by immunohistochemistry. In the absence of Ang II,GIT1 distributed across the entire cell (Figure 3A). Thelocalization of CamK II was similar to GIT1 (Figure 3B).This is consistent with a previous report that CamK II�c

locates in cytoplasm while CamK II�b is nuclear.8,21 Dualimmunofluorescent detection revealed that GIT1 colocalizedwith CamK II basally (Figure 3C). After administration ofAng II for 5 to 10 minutes, GIT1 translocated to theperinuclear and nuclear area (Figure 3D, G). CamK II alsotranslocated to the perinuclear and nuclear areas (Figure 3Eand 3H). GIT1 and CamK II were mostly colocalized duringthis period (Figure 3F and 3I), consistent with the imunopre-cipitation results (Figure 2). From 30 to 60 minutes, bothGIT1 and CamK II returned to the cytoplasmic compartment(Figure 3J through 3O). The similar translocation of GIT1and CamK II suggests a significant functional interaction.

Domains of GIT1 That Mediate Interaction WithCamK IIGIT1 is composed of an ARF GAP domain, an ankyrin repeatregion, 2 carboxyl paxillin-binding subdomains, a SpaII

homology domain (SHD), synaptic localization domain (SLD),and 3 putative coiled-coil (CC) domains (Figure 4A). To definethe domains responsible for the GIT1-CamK II interaction, wetransfected HEK 293 cells with Flag-CamK II and the GIT1deletion mutants described in methods. Immunoprecipitation ofCamK II with anti-Flag antibody coprecipitated GIT1(1–770),GIT1(1–635), but not GIT1(1–420), GIT1(250–770), orGIT1(420–770; Figure 4B and 4C). These results suggest thatboth the ARF-GAP and CC2 domains are required for GIT1-CamK II interaction because the only GIT1(1–635) has both theARF-GAP domain and the CC2 domain, whereas the othermutants lack one of the these domains. GIT1 functions as ascaffold for CamK II by binding to CamK II, which is essentialfor phosphorylation of HDAC5.

To determine the functional significance of GIT1 bindingwith CamK II, HDAC5 phosphorylation induced by Ang II wasstudied in 293 cells transfected with Flag-CamK II, Xpress-GIT1 or both (note the AT1R was also cotransfected; Figure 5).Overexpression of GIT1 or CamK II alone had no significanteffect on HDAC5 phosphorylation (Figure 5A and 5B). How-ever, when both were overexpressed, HDAC5 phosphorylationsignificantly increased (2.8-fold, Figure 5A and 5B). Based onthe finding that the ARF GAP domain and CC2 domain areessential for GIT1 and CamK II interaction, we determined theability of GIT1 mutants to increase phosphorylation of HDAC5.

AngII IgG 0 1 2 5 10 15 30 (min)

IB HDAC5

IB 14-3-3

IB CamKII

IP CamKII

B

VSMC

C

IB:PLCγγγγ

IB: HDAC5

IB: CamKIIIP PLCγγγγ

VSMC

D

IB: GIT1

IB: HDAC5

IB: CamKII

IP GIT1

VSMC

AngII IgG 0 1 2 5 15 30 (min)

AngII IgG 0 1 2 5 15 30 (min)

IB GIT1

IB CamKIIIP CamKII

VSMC

AngII IgG 0 2 5 10 15 30 (min)

A

Figure 2. Association of GIT1, CamK II, and PLC� wasincreased by Ang II. The association of endogenous CamK II,HDAC5, 14-3-3, PLC�, and GIT1 in VSMCs was assayed byimmunoprecipitation (IP) with CamK II antibody and probing forHDAC5 and 14-3-3 (A), GIT1 (B). The binding of CamK II, PLC�,and HDAC5 or the interaction of GIT1, CamK II, and HDAC5were assayed by immunoprecipitation respectively with PLC�antibody (C) or GIT1 antibody (D).

GIT1 CamKII MergeAngII

0 min

5 min

10 min

30 min

60 min

A B C

D E F

G H I

J K L

M N O

25µm

Figure 3. Colocalization of GIT1 and CamK II by immunofluores-cence. VSMCs were incubated with Ang II for the indicatedtimes. The cells were washed, fixed, and stained using GIT1and CamK II antibodies. GIT1 was visualized as red fluores-cence, CamK II visualized as green fluorescence, and colocal-ization of GIT1 and CamK II as yellow fluorescence. Bar repre-sents 25 �m.




As shown in Figure 5C and 5D, GIT1 mutants lacking theARF-GAP domain (eg, GIT1[420–770]) or CC2 domain (eg,GIT1[1–420]) had significantly less effect on phosphorylationof HDAC5 compared with WT GIT1 (P�0.05, Figure 5C and5D), but still substantial effect on phosphorylation of HDAC5compared to pcDNA group (P�0.05, Figure 5C and 5D).However in Figure 4 we showed that these mutants could notbind to CamK II. How can the mutants that do not bind to CamKII still induce HDAC5 phosphorylation? Our hypothesis is thatthese 2 mutants may bind to CamK II through 1 binding site.However, with loss of 1 required binding domain, this binding ismuch weaker than the binding of GIT1 (WT) or GIT1(1–635)(including required 2 binding domain) to CamK II. This weakinteraction is difficult to detect by immunoprecipatation. Allthese data suggest that GIT1 functions as a scaffold for CamK II,thereby increasing phosphorylation of HDAC5.

GIT1 Is Required for Ang II–Induced MEF2Transcriptional Activity in VSMCsIn the nucleus, HDAC5 associates with the myocyte enhancerfactor-2 (MEF2), which represses MEF2 transcriptional ac-tivity.22,23 To determine the role of GIT1 in HDAC5-mediated regulation of MEF2 transcriptional activation inVSMCs, we transfected VSMCs with a 3� MEF2-luciferasereporter plasmid and myc-HDAC5 WT. We then determinedthe effect of decreasing GIT1 expression with GIT1 siRNAon MEF2 activity. Ang II significantly increased MEF2transcriptional activity in VSMCs (Figure 6), which wasdecreased by knockdown of GIT1 (Figure 6).

DiscussionThe major finding of this study is that GIT1 is a novelmediator of Ang II–mediated VSMC gene transcription.Specifically we show that GIT1 participates in an Ang II

signaling pathway that involves phosphorylation of HDAC5and activation of MEF2, downstream of a pathway thatrequires Src, PLC�, and CamK II (supplemental Figure V).Our results suggest that GIT1, via its multi-domain scaffold-ing function, coordinates Ang II signaling events that controlcalcium-dependent signaling (PLC� and CamK II).

The focus of the present study is on CamK II which is afamily of cytosolic serine/threonine protein kinases that existas multimers consisting of �, �, �, or � subunits, eachencoded by a different gene.24,25 Whereas CamK II� and �are mainly expressed in neuronal tissues, CaMKII�b,CaMKII�c, and CaMKII� are abundant in the heart8,21,26 andVSMCs.27 CaMKII can phosphorylate type II HDACs.11

These HDACs (HDAC 4, 5, 7, and 9) normally represstranscriptional activity (eg, activation driven by MEF2) andfavor condensed DNA. When HDAC is phosphorylated inresponse to neurohumoral stimuli, it is exported from thenucleus (in association with the chaperone protein 14-3-3),MEF2 is derepressed, and a hypertrophic program of geneexpression is activated.28 Indeed, genetic knockout ofHDAC5 results in marked cardiac hypertrophy.14,15 Ang II isan important cardiovascular hormone that induces cardio-myocyte and VSMC hypertrophy.3,29,30

Here we investigated the role of HDAC5 phosphorylationin MEF2 activation induced by Ang II. Ang II rapidly inducesphosphorylation of HDAC5 in rat VSMCs, with a peak at 5minutes. In VSMCs, Ang II binding to the AT1R activatesc-Src. c-Src phosphorylates GIT1, which is bound to PLC� ina complex under basal conditions. Conformational changes inGIT1 induced by Src-dependent tyrosine phosphorylationlead to the activation of PLC�. Activation of PLC� isresponsible for elevation of intracellular calcium, whichresults in the autophosphorylation of CamK II. Activation ofCamK II causes HDAC5 phosphorylation and changes intarget gene expression. Data to support this pathway includethe finding that knockdown of GIT1 by siRNA significantlyinhibited phosphorylation of HDAC5. Furthermore, DN-Srcadenovirus infection, Src inhibitor PP2, PLC� inhibition, andCamK II inhibition decreased phosphorylation of HDAC5induced by Ang II. These data show that GIT1 plays animportant role in HDAC5-dependent signaling by regulatingthe activation of PLC�.

We previously showed that GIT1 acted as a scaffoldprotein for the MEK1-ERK1/2 pathway. Specifically GIT1exhibits 4 scaffold characteristics4,31 including (1) assemblyof specific modules; (2) excluding (or insulating) othermolecules; (3) promoting sequential activation of enzymes byphysical interaction; and (4) providing feedback regulation ofcell surface receptors. Here we show that GIT1 acts as ascaffold protein for CamK II. The interaction of GIT1 andCamK II was constitutive and increased after Ang II stimu-lation. The interaction of GIT1 and CamK II was furtherconfirmed by immunofluorescence colocalization. Both GIT1and CamK II (CamK II�b or CamK II�c) translocated fromcytoplasm to nucleus in a similar manner after stimulationwith Ang II. In contrast, there was no direct interactionbetween GIT1 and HDAC5, nor between PLC� and CamK II.Rather GIT1 acts as a scaffold protein to assemble a “calciumsignaling complex” (eg, “Calciosome”), which includes

C

AARF GAP ANK 250 375

SHD

253 274 424 474 649 669 770

CC1 CC2 CC3

1

B

PBS

IP: Flag-CamKII IB: FlagCamKII

IP: Flag CaMKIIIB: Xpress-GIT1

-GIT1CamKII - - + + + + + +

HEK293

97 kd70 kd

1-42

025

0-77

0

420-

770

1-63

5

1-77

0

1-77

0

-

596

SLD

420-

770

Total Cell lysate

IB:Xpress-GIT1

GIT1

97 kd70 kd

46 kd39 kd

57 kd

- + - + 1-42

025

0-77

01-

635

Figure 4. Domains of GIT1 required for interaction with CamK II.A, GIT1 domain structure B, Xpress-GIT1 constructs werecotransfected into HEK 293 cells with Flag-CamK II. CamK IIwas immunoprecipitated (IP). The expression of GIT1 wasdetected by anti-Xpress antibodies. The arrows show interac-tions of GIT1 constructs. C, Total cell lysate probed with anti-Xpress showed levels of expression of GIT1.




PLC� as well as CamK II. This complex facilitates phosphor-ylation of HDAC5 by CamK II. Although it has been reportedthat PLC� translocates to the nucleus in VSMCs,32 it isunclear how this complex could translocate from cytoplasmto nucleus.

GIT1 is composed of an ARF GAP domain, an ankyrinrepeat region, 2 carboxyl paxillin-binding subdomains, aSpaII homology domain (SHD), and 3 putative coiled-coil(CC) domains. The ARF-GAP domain has been shown to befunctionally important for endocytosis.33 The CC2 region isimportant for interactions with PAK,34 the CC3 region(including residues 646 to 770) is essential for interactionswith paxillin,35 and the SHD region overlaps with bindingsites defined for FAK and PIX.35 The domains of GIT1required for CamK II interaction are located in theN-terminal ARF GAP domain and CC2 domains. Althoughbinding of CamK II to GIT1 is constitutive, it appears thatspecific conformational changes occur in GIT1 or CamK II inresponse to tyrosine phosphorylation of GIT1 that enableGIT1 (and CamK II) translocation to the nucleus. It islikely that dephosphorylation contributes to export fromthe nucleus.

Recently it was reported that Ang II–dependent phosphor-ylation of HDAC5 in VSMCs is through a PKC and PKD

pathway that is resistant to the calcium chelator BAPTA/AMas well as CaMK inhibitors KN93 and KN62.36 Our datasuggest that GIT1 and CamK II play an essential role inphosphorylation of HDAC5, possibly independent of the

IB: p-HDAC5

HEK293+AT1 R pcDNA GIT1 CamKII GIT1+CamKII

AngII - + - + - + - +

IB: GIT1

IB: Flag-CamKII

IB: Actin

0

0.5

1

1.5

2

2.5

3

3.5 *

n

i t c

A

/ 5 C

A

D

H

p

e s a e r c

n

i d

l

o

F

- + - + - + - + AngII pcDNA GIT1 CamKII GIT1+CamKII

IB: Myc-HDAC5

IB: Xpress-GIT1

IB: Flag-CamKII

IB: Actin

IB: p-HDAC5

AngII - + - + - + - + - +

pcDNA WT 1-420 1-635 420-770

Flag-CamKII

0

0.5

1

1.5

2

2.5

3

3.5

pcDNA WT 1-420 1-635 420-770

5 C

A

D

H

/ 5

C

A

D

H

p

e s a e r c n

i

d

l o

F

* #

*

#

GIT1

CamKII - + + + +

GIT1

HEK293+AT1R A

B

C

D

Figure 5. Binding of GIT1 and CamK II is essential for phosphorylation of HDAC5. A–B, HEK293 cells were transfected with pcDNA, Xpress-GIT1(wt), Flag-CamKII, or Xpress-GIT1(wt)�Flag-CamKII and stimulated with Ang II for 5 minutes. C–D, HEK293 cells were cotransfectedwith Xpress-GIT1WT or GIT1 mutants with Flag-CamK II and stimulated with Ang II for 5 minutes. Lysates was immunoblotted for p-HDAC5,Actin, anti-Xpress, and anti-Flag antibody B and D, Relative increase of HDAC5 phosphorylation compared to pcDNA group or WT group.*P�0.05 compared with pcDNA group; #P�0.05 compared with WT group. (means�SD; n�3).

0

5

10

15

20

25 esareficu

L evitaleR

2feM f

o ytivitcA

*

*#

- + - +ControlsiRNA

GIT1siRNA

Ang II

Figure 6. GIT1 altered Ang II–stimulated MEF2 transcriptionalactivity. Luciferase activity was determined in VSMCs transfectedwith MEF2 reporter gene and control or GIT1 siRNA.The graphs represent averaged data (means�SD, n�6). (*P�0.05vs control siRNA without Ang II; #P�0.05 vs control siRNA andAng II group.)




PKC-PKD signal pathway. The discrepancy regarding therole of calcium and CamK II could be attributable to differentcell conditions and experimental procedures.

In summary, we have demonstrated that HDAC5 is phos-phorylated in rat VSMCs by Ang II via a pathway thatinvolves PLC� and CamK II. We also found that GIT1 andHDAC5 are involved in Ang II–stimulated MEF2 transcrip-tional activity (supplemental Figure V). These findings sug-gest novel roles for GIT1 and HDAC5 in Ang II signaling inVSMCs.

Sources of FundingThis work was supported by grants from the NIH (HL63462 toB.C.B., HL77789 to C.Y.), and AHA EIA 0740021N to C.Y.

DisclosuresNone.

References1. Berk BC. Vascular smooth muscle growth: autocrine growth mech-

anisms. Physiol Rev. 2001;81:999–1030.2. Berk BC. Angiotensin II signal transduction in vascular smooth muscle:

pathways activated by specific tyrosine kinases. J Am Soc Nephrol.1999;10 Suppl 11:S62–S68.

3. Berk BC, Haendeler J, Sottile J. Angiotensin II, atherosclerosis, and aorticaneurysms. J Clin Invest. 2000;105:1525–1526.

4. Hoefen RJ, Berk BC. The multifunctional GIT family of proteins. J CellSci. 2006;119:1469–1475.

5. Haendeler J, Yin G, Hojo Y, Saito Y, Melaragno M, Yan C, Sharma VK,Heller M, Aebersold R, Berk BC. GIT1 mediates Src-dependent acti-vation of phospholipase Cgamma by angiotensin II and epidermal growthfactor. J Biol Chem. 2003;278:49936–49944.

6. Yin G, Haendeler J, Yan C, Berk BC. GIT1 functions as a scaffold forMEK1-extracellular signal-regulated kinase 1 and 2 activation byangiotensin II and epidermal growth factor. Mol Cell Biol. 2004;24:875– 885.

7. Yin G, Zheng Q, Yan C, Berk BC. GIT1 is a scaffold for ERK1/2activation in focal adhesions. J Biol Chem. 2005;280:27705–27712.

8. Ginnan R, Singer HA. CaM kinase II-dependent activation of tyrosinekinases and ERK1/2 in vascular smooth muscle. Am J Physiol CellPhysiol. 2002;282:C754–C761.

9. Van Riper DA, Schworer CM, Singer HA. Ca2�-induced redistributionof Ca2�/calmodulin-dependent protein kinase II associated with an endo-plasmic reticulum stress response in vascular smooth muscle. Mol CellBiochem. 2000;213:83–92.

10. Kuo MH, Allis CD. Roles of histone acetyltransferases and deacetylasesin gene regulation. Bioessays. 1998;20:615–626.

11. McKinsey TA, Zhang CL, Lu J, Olson EN. Signal-dependent nuclearexport of a histone deacetylase regulates muscle differentiation. Nature.2000;408:106–111.

12. McKinsey TA, Zhang CL, Olson EN. Activation of the myocyte enhancerfactor-2 transcription factor by calcium/calmodulin-dependent proteinkinase-stimulated binding of 14-3-3 to histone deacetylase 5. Proc NatlAcad Sci U S A. 2000;97:14400–14405.

13. Vega RB, Harrison BC, Meadows E, Roberts CR, Papst PJ, Olson EN,McKinsey TA. Protein kinases C and D mediate agonist-dependentcardiac hypertrophy through nuclear export of histone deacetylase 5. MolCell Biol. 2004;24:8374–8385.

14. Zhang CL, McKinsey TA, Chang S, Antos CL, Hill JA, Olson EN. ClassII histone deacetylases act as signal-responsive repressors of cardiachypertrophy. Cell. 2002;110:479–488.

15. Chang S, McKinsey TA, Zhang CL, Richardson JA, Hill JA, Olson EN.Histone deacetylases 5 and 9 govern responsiveness of the heart to asubset of stress signals and play redundant roles in heart development.Mol Cell Biol. 2004;24:8467–8476.

16. Wu X, Zhang T, Bossuyt J, Li X, McKinsey TA, Dedman JR, Olson EN,Chen J, Brown JH, Bers DM. Local InsP3-dependent perinuclear Ca2�signaling in cardiac myocyte excitation-transcription coupling. J ClinInvest. 2006;116:675–682.

17. Melaragno MG, Wuthrich DA, Poppa V, Gill D, Lindner V, Berk BC,Corson MA. Increased expression of Axl tyrosine kinase after vascularinjury and regulation by G protein-coupled receptor agonists in rats. CircRes. 1998;83:697–704.

18. Caplice NM, Bunch TJ, Stalboerger PG, Wang S, Simper D, MillerDV, Russell SJ, Litzow MR, Edwards WD. Smooth muscle cells inhuman coronary atherosclerosis can originate from cells administeredat marrow transplantation. Proc Natl Acad Sci U S A. 2003;100:4754 – 4759.

19. Backs J, Song K, Bezprozvannaya S, Chang S, Olson EN. CaM kinase IIselectively signals to histone deacetylase 4 during cardiomyocyte hyper-trophy. J Clin Invest. 2006;116:1853–1864.

20. Ishida M, Ishida T, Thomas S, Berk BC. Activation of extracellularsignal-regulated kinases (ERK1/2) by angiotensin II is dependent onc-Src in vascular smooth muscle cells. Circ Res. 1998;82:7–12.

21. Hoch B, Meyer R, Hetzer R, Krause EG, Karczewski P. Identification andexpression of delta-isoforms of the multifunctional Ca2�/calmodulin-dependent protein kinase in failing and nonfailing human myocardium.Circ Res. 1999;84:713–721.

22. McKinsey TA, Kuwahara K, Bezprozvannaya S, Olson EN. Class IIhistone deacetylases confer signal responsiveness to the ankyrin-repeat proteins ANKRA2 and RFXANK. Mol Biol Cell. 2006;17:438 – 447.

23. Berger I, Bieniossek C, Schaffitzel C, Hassler M, Santelli E, RichmondTJ. Direct interaction of Ca2�/calmodulin inhibits histone deacetylase 5repressor core binding to myocyte enhancer factor 2. J Biol Chem.2003;278:17625–17635.

24. Hudmon A, Schulman H. Structure-function of the multifunctionalCa2�/calmodulin-dependent protein kinase II. Biochem J. 2002;364:593–611.

25. Tombes RM, Faison MO, Turbeville JM. Organization and evolution ofmultifunctional Ca(2�)/CaM-dependent protein kinase genes. Gene.2003;322:17–31.

26. Colomer JM, Mao L, Rockman HA, Means AR. Pressure overload selec-tively up-regulates Ca2�/calmodulin-dependent protein kinase II in vivo.Mol Endocrinol. 2003;17:183–192.

27. Schworer CM, Rothblum LI, Thekkumkara TJ, Singer HA. Identificationof novel isoforms of the delta subunit of Ca2�/calmodulin-dependentprotein kinase II. Differential expression in rat brain and aorta. J BiolChem. 1993;268:14443–14449.

28. Olson EN, Schneider MD. Sizing up the heart: development redux indisease. Genes Dev. 2003;17:1937–1956.

29. Watanabe T, Barker TA, Berk BC. Angiotensin II and the Endothelium:Diverse Signals and Effects. Hypertension. 2005;45:163–169.

30. Dikalova A, Clempus R, Lassegue B, Cheng G, McCoy J, Dikalov S,San Martin A, Lyle A, Weber DS, Weiss D, Taylor WR, Schmidt HH,Owens GK, Lambeth JD, Griendling KK. Nox1 overexpression poten-tiates angiotensin II–induced hypertension and vascular smoothmuscle hypertrophy in transgenic mice. Circulation. 2005;112:2668 –2676.

31. Claing A, Perry SJ, Achiriloaie M, Walker JK, Albanesi JP, LefkowitzRJ, Premont RT. Multiple endocytic pathways of G protein-coupledreceptors delineated by GIT1 sensitivity. Proc Natl Acad Sci U S A.2000;97:1119–1124.

32. Egan CG, Wainwright CL, Wadsworth RM, Nixon GF. PDGF-inducedsignaling in proliferating and differentiated vascular smooth muscle:effects of altered intracellular Ca2� regulation. Cardiovasc Res. 2005;67:308–316.

33. Claing A, Chen W, Miller WE, Vitale N, Moss J, Premont RT, LefkowitzRJ. beta-Arrestin-mediated ADP-ribosylation factor 6 activation and beta2-adrenergic receptor endocytosis. J Biol Chem. 2001;276:42509–42513.

34. Turner CE, Brown MC, Perrotta JA, Riedy MC, Nikolopoulos SN,McDonald AR, Bagrodia S, Thomas S, Leventhal PS. Paxillin LD4 motifbinds PAK and PIX through a novel 95-kD ankyrin repeat, ARF-GAPprotein: A role in cytoskeletal remodeling. J Cell Biol. 1999;145:851–863.

35. Zhao ZS, Manser E, Loo TH, Lim L. Coupling of PAK-interactingexchange factor PIX to GIT1 promotes focal complex disassembly. MolCell Biol. 2000;20:6354–6363.

36. Xu X, Ha CH, Wong C, Wang W, Hausser A, Pfizenmaier K, Olson EN,McKinsey TA, Jin ZG. Angiotensin II stimulates protein kinaseD-dependent histone deacetylase 5 phosphorylation and nuclear exportleading to vascular smooth muscle cell hypertrophy. Arterioscler ThrombVasc Biol. 2007;27:2355–2362.




1

Supplemental Data

Supplemental Figure1. c-Src, PLCγ and CamkII are required forHDAC5 phosphorylation

VSMC were incubated with PP2(A), U73122(C) or KN93(D) for 30 min, then stimulated with

AngII for 5 min. In (B) Ad.DN-Src or Lac-Z were infected, then the VSMCs were stimulated

with AngII for 5 min. HDAC5, ERK1/2 were detected.

Supplemental Figure2. Quantitation of the binding of GIT1, CamKII, PLCγ and HDAC5

in VSMC in figure2

Ratio of GIT1 to CamKII(A); HDAC5 or 14-3-3 to CamKII(B); CamKII or HDAC5 to PLCγ(C);

HDAC5 or CamKII to GIT1(D) of different time point compared to time 0 group.

Supplemental Figure3. GIT1 didn’t associate with HDAC5 in HEK293 cells.

HEK293 cells were transfected with Xpress-GIT1 and Myc-HDAC5 and stimulated with AngII

for 5min. The association of GIT1 and HDAC5 was assayed by immunoprecipitation with

Xpress antibody. Western blot of total cell lysis showed the equal expression of HDAC5 and

GIT1.

Supplemental Figure4. PLCγ didn’t directly interact with CamKII.

HEK293 cells were transfected with Flag-CamKII and Xpress vector or Flag-CamKII and

Xpress-GIT1 (WT) .The association of PLCγ and CamKII was assayed by immunoprecipitation

(IP) with PLCγ or CamKII antibody. Lower panel showed the expression of GIT1, PLCγ and

CamKII.

Supplemental Figure5. Model for GIT1-mediated HDAC5 signaling.

AngII activates c-Src, PLCγ and CamKII by AT1R. GIT1 assembles CamKII and PLCγ to form

a “ Calciosome”, which is required for phosphorylation of HDAC5. Phosphorylated HDAC5 is

2

exported via 14-3-3, thereby increasing MEF2 transcriptional activity, derepressing HDAC5

target genes induced by AngII.

Supplemental Fig. 1

Ad:DN-Src(MOI) con con 30 100 300 30 100 300

IB: p-HDAC5

IB: c-Src(chicken)

IB: p-ERK1/2

IB: ERK1/2

Src inhibitor: PP2 (nM) - - 3 10 30 3 10 30

IB: p-HDAC5

IB:ERK1/2

PLCγ inhibitor:U73122(µM) - - 0.3 1 3 10

IB:p-HDAC5

IB: ERK2

CamKII inhibitor:KN-93 (µM) 0 0 3 3 10 10 1

IB: p-HDAC5

IB:ERK1

A

B

C

D

AngII - + - - - + + +

AngII - + - - - + + +

AngII - + + + + +

AngII - + - + - +

00.51.01.52.02.53.0

0 2 5 10 15 30AngII(min)

GIT

1/C

amK

II F

old

Incr

ease GIT1

00.51.01.52.02.53.0

0 1 2 5 15 30AngII(min)

Cam

KII/

PLCγ

Fold

Inc

reas

e

CamKII

01020304050607080

0 1 2 5 15 30AngII(min)

HD

AC

5/PL

C γ

Fol

d In

crea

se HDAC5

0102030405060

0 1 2 5 10 15 30AngII(min)

HD

AC

5/C

amK

IIFo

ld In

crea

se

00.51.01.52.02.53.0

0 1 2 5 10 15 30AngII(min)

14-3

-3/C

amK

II(F

old

Incr

ease

)

14-3-3HDAC5

02468

1012

0 1 2 5 15 30AngII(min)

HD

AC

5/G

IT1

Fol

d In

crea

se HDAC5

01234567

0 1 2 5 15 30AngII(min)

Cam

KII/

GIT

1 Fo

ld In

crea

se

CamKII

A

B

D

F

Supplemental Fig. 2

C

E

G

IB: Myc-HDAC5

IP: Xpress-GIT1 Total Lysis

IB: Xpress-GIT1

HEK293+AT1R AngII - + - +

Supplemental Fig. 3

Supplemental Fig. 4

IB: Flag-CaMKII

IB: PLCγ

IP: PLCγ Total Lysis

IB: Xpress-GIT1

HEK293 cell

IP: CamKII

Xpress-GIT1 - + - + - + - +Flag-CamKII + + + + + + + +

Total Lysis

AngII

CamKII

MEF2HDAC5

PHDAC5

Active MEF2Target genetranscript

cytosol

nucleus

Src

-14-3-3

14-3-3GIT1CamKII HDAC5 P

PLCγ

PLCγ

GIT1

GIT1

CamKII

PLCγ

Supplemental Fig. 5

Yin and Bradford C. BerkJinjiang Pang, Chen Yan, Kanchana Natarajan, Megan E. Cavet, Michael P. Massett, Guoyong

GIT1 Mediates HDAC5 Activation by Angiotensin II in Vascular Smooth Muscle Cells

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 2008 American Heart Association, Inc. All rights reserved.

Greenville Avenue, Dallas, TX 75231is published by the American Heart Association, 7272Arteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/ATVBAHA.107.1613492008;

2008;28:892-898; originally published online February 21,Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/28/5/892World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org/content/suppl/2008/04/18/ATVBAHA.107.161349.DC1.htmlData Supplement (unedited) at:

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document. Question and AnswerPermissions and Rightspage under Services. Further information about this process is available in the

which permission is being requested is located, click Request Permissions in the middle column of the WebCopyright Clearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of theArteriosclerosis, Thrombosis, and Vascular Biologyin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:


http://atvb.ahajournals.org/content/28/5/892

http://atvb.ahajournals.org/content/suppl/2008/04/18/ATVBAHA.107.161349.DC1.html

http://www.ahajournals.org/site/rights/

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


GIT1 Mediates HDAC5 Activation by Angiotensin II in Vascular Smooth Muscle Cells

Documents