Plasma Kallikrein Promotes Epidermal Growth Factor Receptor

Plasma Kallikrein Promotes Epidermal Growth FactorReceptor Transactivation and Signaling in Vascular SmoothMuscle through Direct Activation of Protease-activatedReceptors*

Received for publication, August 3, 2010 Published, JBC Papers in Press, September 8, 2010, DOI 10.1074/jbc.M110.171769

Rany T. Abdallah‡, Joo-Seob Keum‡, Mi-Hye Lee‡, Bing Wang‡, Monika Gooz‡, Deirdre K. Luttrell‡,Louis M. Luttrell‡§¶, and Ayad A. Jaffa‡1

From the Departments of ‡Medicine and §Biochemistry & Molecular Biology, Medical University of South Carolina,Charleston, South Carolina 29425 and the ¶Research Service of the Ralph H. Johnson Veterans Affairs Medical Center,Charleston, South Carolina 29401

The kallikrein-kinin system, along with the interlocking renin-angiotensin system, is a key regulator of vascular contractility andinjury response. The principal effectors of the kallikrein-kinin sys-tem are plasma and tissue kallikreins, proteases that cleave highmolecular weight kininogen to produce bradykinin. Most of thecellular actions of kallikrein (KK) are thought to be mediated bybradykinin,whichactsviaGprotein-coupledB1andB2bradykininreceptors on VSMCs and endothelial cells. Here, we find that pri-mary aortic vascular smoothmuscle but not endothelial cells pos-sess the ability to activate plasma prekallikrein. Surprisingly,exposingVSMCs to prekallikrein leads to activation of theERK1/2mitogen-activated protein kinase cascade via a mechanism thatrequires kallikrein activity but does not involve bradykinin recep-tors. In transfected HEK293 cells, we find that plasma kallikreindirectly activates G protein-coupled protease-activated receptors(PARs) 1 and 2, which possess consensus kallikrein cleavage sites,but not PAR4. In vascular smoothmuscles, KK stimulates ADAM(a disintegrin andmetalloprotease) 17 activity via a PAR1/2 recep-tor-dependent mechanism, leading sequentially to release of theendogenousADAM17 substrates, amphiregulin and tumor necro-sis factor-�, metalloprotease-dependent transactivation of epider-mal growth factor receptors, and metalloprotease and epidermalgrowth factor receptor-dependent ERK1/2 activation. Theseresults suggest a novelmechanismof bradykinin-independent kal-likrein action that may contribute to the regulation of vascularresponses in pathophysiologic states, such as diabetesmellitus.

The kallikrein-kinin system (KKS)2 plays diverse roles inthe regulation of vascular tone, tissue inflammation, coagu-

lation, and pain (1). KKS components are expressed in manytissues and cell lines (2). Two forms of the KKS exist inhumans, tissue KKS and plasma KKS. Tissue kallikrein (KK)is expressed primarily in the kidney, vascular system, brain,and pancreas, where it acts on low molecular weight kinino-gen to release bradykinin (BK). The plasma KKS is composedof factor XII, prekallikrein (PK), and high molecular weightkininogen (HMWK). It is responsible for the activation ofthe intrinsic blood clotting pathway (3–5). The inactive formof plasma PK is bound to HMWK in circulation. When thiscomplex assembles along with other components of the KKSon endothelial cell membranes, plasma PK is activated toplasma KK, which cleaves HMWK to release BK both sys-temically and locally within the vessel wall. The cellulareffects of KK are thought to reflect primarily the action ofBK, a peptide hormone that produces proinflammatory andvasodilatory effects by activating two cell surface G protein-coupled receptors, the B1 and B2 bradykinin receptors,expressed on vascular endothelial and vascular smooth mus-cle cells (VSMCs) (6).The KKS also interfaces with the renin-angiotensin system

(RAS) (7). Plasma and tissue KK can convert prorenin to renin,the protease responsible for processing angiotensinogen toangiotensin I. Conversely, the angiotensin-converting enzyme,the protease responsible for processing angiotensin I to angio-tensin II, also cleaves BK to produce its inactive fragment, BK(1–5). Thus, factors that enhance the generation of angiotensinII tend to dampen BK signaling. Not surprisingly, given thisreciprocal regulation, the KKS and RAS generally are thoughtto exert opposing effects in the regulation of vascular tone (8).Angiotensin II, acting via G protein-coupled AT1 and AT2receptors, causes VSMC constriction and promotes hyperten-sion, whereas under normal physiological conditions, BK stim-ulates vascular endothelial cells to produce nitric oxide (NO),leading to relaxation of the underlyingVSMCs and vasodilation(9, 10). NO from endothelial cells also inhibits VSMC and renalmesangial cell proliferation (11, 12). However, in the setting ofvascular injury and endothelial denudation, BK can act directlyon VSMC to induce their contraction and activate multiplesignaling pathways in a manner similar to vasoconstrictors likeangiotensin II (13).

* This work was supported, in whole or in part, by National Institutes of HealthGrants HL087986 and HL077192 (to A. A. J.), and DK55524 (to L. M. L.). Thiswork was also supported by the South Carolina Center for BiomedicalResearch Excellence in Cardiovascular Diseases (to D. K. L.) and Depart-ment of Veterans Affairs Research Enhancement Award Program awards.

1 To whom correspondence should be addressed: Department of Biochemis-try, Faculty of Medicine, American University of Beirut, Beirut, Lebanon.Tel.: 961-350-000 (ext. 4870); E-mail: [email protected].

2 The abbreviations used are: KKS, kallikrein-kinin system; AR, amphiregulin;BK, bradykinin; EAA1, early endosomal antigen 1; VSMC, vascular smoothmuscle cell; R-VSMC, rat aortic VSMC; H-VSMC, human aortic VSMC; HB,heparin-binding; HMWK, high molecular weight kininogen; KK, kallikrein;MMP, matrix metalloprotease; PAR, protease-activated receptor; PK, prek-allikrein; RAS, renin-angiotensin system; TACE, TNF-� converting enzyme.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 285, NO. 45, pp. 35206 –35215, November 5, 2010Printed in the U.S.A.

35206 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 285 • NUMBER 45 • NOVEMBER 5, 2010

by guest on March 17, 2018

http://ww

w.jbc.org/

Dow

nloaded from


http://ww

w.jbc.org/

Dow

nloaded from


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

http://www.jbc.org/

http://www.jbc.org/

The protease-activated receptors (PARs) are a family of fourG protein-coupled receptors that are activated upon enzymaticcleavage of their N termini by specific serine proteases (14, 15).Each receptor contains an internal “tethered” ligand buriedwithin its N terminus that is exposed upon proteolysis. PARsare widely expressed, with PAR1, PAR3, and PAR4 present onplatelets, endothelial cells, myocytes, and astrocytes, whereasPAR2 is also present on epithelium. PARs are activated orinhibited by a host of proteases from the coagulation cascade,inflammatory cells, and the digestive tract (15). PAR1, PAR3,and PAR4 are all activated by thrombin, whereas PAR2, whichis present on intestinal epithelium, is activated by trypsin andtryptase. Depending on their location, PARs regulate diverseprocesses, including hemostasis, inflammation, cell migration,and pain transmission (16–19). Although their physiologicroles have not been well studied, KK and KK-related proteaseshave been recently implicated as endogenous regulators ofPARs in the central nervous system and prostate cancer(20–22).EGF receptors ErbB1–4 comprise a family of classical recep-

tor tyrosine kinases that control cell proliferation and survivalvia the Ras and phosphatidylinositol-3 kinase/Akt pathways(23). There are at least 11 endogenous EGF receptor ligands:EGF, transforming growth factor-�, heparin-binding (HB)-EGF, amphiregulin (AR), epiregulin, epigen, and neuregulin1–4, each of which contains a 6-kDa domain that is homolo-gous to EGF. All are synthesized as transmembrane precursorsthat must be cleaved to generate the mature ligand (24). Theprocessing of EGF family ligands is carried out by metallopro-tease-disintegrins of the ADAM family (25). Unlike most zinc-binding matrix metalloproteases (MMPs), which are secretedproteins, ADAMs are membrane-anchored proteases com-posed of an extracellular Zn-MMP domain, a single transmem-brane domain and a short cytoplasmic tail.ManyADAMs func-tion as “sheddases” that catalyze intramembrane proteolysis ofgrowth factors, cytokines, receptors, and adhesion molecules.Importantly, ADAM-dependent ectodomain shedding is a reg-ulated process, enabling diverse extracellular stimuli to activateEGF receptors by triggering ADAM-dependent release of EGFfamily ligands (26, 27).Substantial evidence supports a role for EGF receptor trans-

activation in the pathogenesis of vascular injury and atherogen-esis. Plasma HB-EGF levels are elevated in coronary artery dis-ease (28). Expression of pro-HB-EGF, along with ErbB1, andADAM17 is increased in the shoulder region of atheromas (29,30), and pathway components can be up-regulated by shearstress (31) and endothelial injury (32). ADAM17, also known asTNF-� converting enzyme (TACE), is a major sheddase forHB-EGF as well as the proinflammatory cytokine TNF-� (29,33). Conversely, blockade of EGF receptor activity inhibits inti-mal hyperplasia in experimental vascular injury (34). G protein-coupled receptors, like the AT1 receptor (35) and other stimuli,including lipoprotein remnants, oxidative stress, and fluidshear (36–38) all promote EGF receptor ligand shedding andEGF receptor transactivation.Here, we describe a novel mechanism whereby plasma KK

promotes EGF receptor transactivation and activation of theERK pathway independent of BK. In primary vascular smooth

muscle cells, KK acts directly on PAR1 and -2, stimulating PAR-dependent activation of ADAM17, EGF receptor ligand shed-ding, and EGF receptor transactivation. Given the sometimesopposing and sometimes complementary effects of the RASand KKS in the vasculature, we speculate that BK-independenteffects of KK may play a role in the pathogenesis of vasculardisease and therefore present a viable therapeutic target.

MATERIALS AND METHODS

cDNAConstructs—Full-length cDNA clones of PAR1, PAR2,and PAR4 were purchased from the University of MissouriRolla cDNA Resource Center. cDNAs were amplified by PCRusing primers (IDT Technologies, Coralville, IA) designed tointroduce Hind3 and EcoR1 sites at the 5� and 3� ends, respec-tively, and remove the stop codon, and subcloned into thepEGFP-N3 plasmid vector (BD Biosciences, Franklin Lakes,NJ). All sequences were verified by dideoxynucleotidesequencing.Primary Cell Isolation and Culture—Primary rat aortic

VSMCs (R-VSMC) were isolated from 75–100 g body weightSprague-Dawley rats (Charles River Laboratories,Wilmington,MA) as described previously (39). R-VSMCs were maintainedinminimum essential medium (Lonza, Basel, Switzerland) sup-plemented with 10% fetal bovine serum and 1% antibiotic/antimycotic solution (Sigma). Cells were fed every 2 days andsubcultured upon reaching 90% confluence. Human aorticVSMC (H-VSMC), human umbilical vein endothelial cells, andhuman aortic endothelial cells were purchased from Clonetics(Lonza, Basel, Switzerland) and maintained as recommendedby the supplier. H-VSMC were cultured in smooth muscle cellbasal medium supplemented with 5% FBS, 0.1% insulin, 0.2%human fibroblast growth factor, 0.1% gentamicin, and 0.1%human EGF. Growth medium was replaced every 2 days, andthe cells were subcultured upon reaching 80% confluence. Priorto each experiment, cells were seeded into multiwell plates asappropriate and incubated for 24–48 h in serum-free growthmedium supplemented with 0.1% bovine serum albumin and1%antibiotic/antimycotic solution.All experiments onprimarycells were performed between passages four and nine.HEK293 Cell Culture and Transfection—HEK293 cells were

obtained from the American Type Culture Collection andmaintained in minimum essential medium with Earle’s saltssupplemented with 10% fetal bovine serum and 1% antibiotic/antimycotic solution. Transient transfection of HEK293 cellswith plasmids encodingGFP-taggedPAR1–4was performed in6-cm dishes using FuGENE 6 (Roche Applied Science) accord-ing to the manufacturer’s instructions with 5 �g of plasmidDNAper dish and 3�l FuGENE per�g of DNA. Prior to exper-imentation, transfected cells were incubated for 24 h in serum-free growth medium supplemented with 10 mM HEPES (pH7.4), 0.1% bovine serum albumin, and 1% antibiotic/antimy-cotic solution.Activation of Plasma Prekallikrein—Serum-deprived R-

VSMC, H-VSMC, human umbilical vein endothelial cells, orhuman aortic endothelial cells cultured in black wall clear-bot-tomed 96-well plates (Corning, Inc., Corning, NY) were incu-bated with 100 nM human PK (Enzyme Research Laboratories,Inc., South Bend, IN) in the presence of 0.6 mM of the chromo-

Kinin-independent Kallikrein Signaling

NOVEMBER 5, 2010 • VOLUME 285 • NUMBER 45 JOURNAL OF BIOLOGICAL CHEMISTRY 35207


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

genic substrate for plasma KK, S2302 (Chromogenix, Milano,Italy). Immediately upon addition of the reagents, plates wereplaced in a SpectraMax 340PCplate reader (MolecularDevices,Sunnyvale, CA) and absorption at 405 nm was measured at2-min intervals for 2 h.Metalloprotease Activity—ADAM activity was measured

using two approaches. For the fluorogenic substrate assay,R-VSMCs at 90% confluence in 96-well plates were serum-starved for 48 h, after which the medium was replaced withHank’s balanced salt solution containing 20 nM human plasmaKK (Enzyme Research Laboratories, Inc.) and 20 �M of theTACE II fluorogenic peptide substrate, MCA-KPLGL-Dpa-AR-NH2 (EMD Biosciences; Gibbstown, NJ). Fluorescence ofthe cleaved product, MCA-KPLG, was measured using aMolecular Devices SpectraMax 2 plate reader (MolecularDevices, Sunnyvale, CA) at 324 nmexcitation and 400 nmemis-sion wavelengths. The second method used the InnozymeTACE Activity KitTM (EMD Biosciences). Serum starved

R-VSMCs in 10-cm culture disheswere incubated in the presence orabsence of 20 nMhumanplasmaKK,lysed as recommended by the man-ufacturer, and normalized for pro-tein content. Equal amounts of sam-ple were loaded into 96-well platesprecoated with an anti-humanTACE monoclonal antibody, incu-bated for 2 h at room temperature,and washed, after which the activityof immobilizedADAM17 was mea-sured using MCA-KPLGL-Dpa-AR-NH2 as above.Protein Immunoblotting—Endo-

genous phosphoproteins were de-tected by immunoblotting wholecell lysates using phosphorylationsite-specific IgG. Serum-starvedR-VSMC in six-well plates werestimulated at 37 °C as described inthe figure legends, lysed with 1�hypotonic sample buffer (Invitro-gen), sonicated, boiled, resolved by4–12% gradient SDS-PAGE (In-vitrogen), and transferred to poly-vinylidine difluoride membrane(Millipore,Billerica,MA).Phosphor-ylated ERK1/2 was detected usingrabbit polyclonal anti-phospho-ERK1/2 IgG (Cell Signaling Tech-nology, Beverly, MA) with horse-radish peroxidase-conjugated goatanti-rabbit IgG secondary antibody(Jackson ImmunoResearch Labora-tories, West Grove, PA). Phosphor-ylated JNK1/2 was detected usingrabbit polyclonal antiphospho-JNK1/2 IgG (Cell Signaling Tech-nology). Phosphorylated EGF re-

FIGURE 1. Activation of plasma prekallikrein in primary vascular smoothmuscle but not endothelial cells. Serum-deprived R-VSMCs, H-VSMCs,human umbilical vein endothelial cells (HUVEC), or human aortic endothelialcells (HAEC) in 96-well plates were incubated with 100 nM human PK in thepresence of the chromogenic substrate for plasma KK, S2302 (0.6 mM), and KKactivity was quantified as change in absorption at 405 nm measured at 2-minintervals over 2 h. Data shown are from one of three experiments, whichproduced comparable results.

FIGURE 2. Activation of ERK1/2 in primary rat aortic vascular smooth muscle cells by prekallikrein isindependent of bradykinin receptors. A, serum-deprived R-VSMCs in six-well plates were treated withhuman plasma PK at increasing concentrations for 15 min (left panel) or with 200 nM plasma PK for varying times(right panel) and ERK1/2 phosphorylation was determined by immunoblotting whole cell lysates as described.Total cell ERK1/2 was blotted as a control for equal protein loading. B, serum-deprived R-VSMCs in six-wellplates were preincubated for 15 min in the presence or absence of the BK receptor antagonist, HOE140 (10 �M;Sigma), prior to stimulation with 10 nM BK (left panel) or 200 nM human plasma PK (right panel) for 15 min.ERK1/2 phosphorylation was determined from whole cell lysates as described. Total cell ERK1/2 was blotted asa control for equal protein loading. Each bar graph depicts the mean � S.E. for three separate experiments. *,greater than untreated control, p � 0.05; †, greater than control, p � 0.01.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

ceptor was detected using rabbitpolyclonal anti-ErbB1 phospho-Tyr1068 IgG (Cell Signaling Tech-nology). Total ERK1/2, measured toconfirm equal loading of whole celllysate samples, was detected us-ing polyclonal anti-ERK1/2 IgG(Upstate Biotechnology, Waltham,MA). Immune complexes werevisualized on x-ray film by en-zyme-linked chemiluminescence.Multiple exposures were made ofeach immunoblot and band inten-sities on optimally exposed filmswere quantified using a Fluor-SMultiImager (Bio-Rad).ADAM17 Ligand Shedding—

H-VSMCs at 90% confluence in10-cm culture dishes were serum-starved for 24 h, after which themedium was replaced with 7 mlof medium containing 0.5% fetalbovine serum and monolayers wereincubated for 3 h,with orwithout 20nM human plasma KK. Conditionedmedium was collected and concen-trated using iCONTM Concentrator7ml/9K tubes (Pierce) centrifugedat 4000 rpm at 4 °C for 1 h. Aliquotsof concentrated medium weremixed with 2� sample buffer andresolved by SDS-PAGE. Amphi-regulin and TNF-� release werequantified by immunoblotting us-ing rabbit polyclonal antiamphi-regulin (Santa Cruz Biotechnology)or anti-TNF� IgG (Chemicon Inter-national, Inc., Billerica, MA) withhorseradish peroxidase-conjugatedgoat anti-rabbit IgG secondaryantibody.Confocal Microscopy—To visual-

ize GFP-PAR1–4 internalization,HEK293 cells were transfectedwith plasmids encoding GFP-tagged PAR1, PAR2, or PAR4 asdescribed. Transfected cells wereplated onto collagen coated 35-mm glass-bottomed culture dishes(MatTek Co., Ashland, MA), se-rum-starved overnight, stimulatedas described in the figure legends,washed with Hank’s balanced saltsolution, and fixed with freshly pre-pared 4% paraformaldehyde for 30min at room temperature. For co-staining, cells were permeabilizedwith 0.1% Triton X-100 (Sigma-

FIGURE 3. Plasma kallikrein directly activates PAR1 and PAR2 receptors. A, schematic depicting theN-terminal proteolytic cleavage sites and internal ligand sequences of PAR1– 4. B, serum-deprived HEK293cells expressing PAR1-GFP (P1) were treated for 15 min with vehicle (Veh), 1 nM thrombin (Thr) as a positivecontrol, or 20 nM plasma KK, after which cells were fixed, permeabilized, co-stained with early endosome(EAA1) and nuclear markers, and examined under confocal fluorescence microscopy. The distribution ofGFP-PAR1 (green), early endosomes (red), and cell nuclei (blue) are shown in individual channels alongwith a composite image. C, serum-deprived HEK293 cells expressing PAR1-GFP (P1), PAR2-GFP (P2), orPAR4-GFP (P4) were treated with 1 nM thrombin (Thr; PAR1/PAR4) or 100 �M trypsin (Trp; PAR2) for 15 minas positive controls or were treated with 20 nM plasma KK for 5 or 30 min, after which cells were fixed, andthe distribution of GFP-PAR1/2/4 was examined under confocal fluorescence microscopy. Representativefields are shown from one of three experiments that produced comparable results. D, internalization ofGFP-PAR1/2/4 in response to thrombin, trypsin, or KK was quantified by measuring the ratio of cytosolicGFP fluorescence to total cellular GFP fluorescence in confocal images of cells exposed to each condition,as described. The bar graph depicts the mean � S.E. for three separate experiments. *, greater thanuntreated control, p � 0.05. NS, nonstimulated.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Aldrich) in phosphate-buffered saline for 5 min, after whichnonspecific binding sites were blocked with 3% serum in PBSfor 1 h. Cells were labeled for 1 h at room temperature in block-ing buffer containing 3–4 �g/ml mouse monoclonal anti-earlyendosomal antigen 1 (EEA1; BD Transduction Laboratories,Franklin Lakes, NJ), washed, and then incubated with a 1:300dilution of goat anti-mouseAlexa Fluor 568 (Invitrogen), also inblocking buffer for 1 h at room temperature. Nuclear DNAwasstained using 4�,6-diamidino-2-phenylindole (Invitrogen).Confocal images were captured on a Zeiss LSM510 laser scan-ning microscope using a Zeiss 63� 1.4 numerical aperturewater immersion lens using excitation, 488 nm; emission, 505–530 nm; and, excitation, 543 nm, emission, 560–615 nm filter

sets. Images were analyzed usingMetaMorph software (MolecularDevices). At least 10 individual cellsper condition from three separateexperiments were analyzed. Foreach image, regions of interest weremanually defined encompassingeither the entire cell (total GFP fluo-rescence) or the region of the cellexcluding the plasma membrane(cytosolic GFP fluorescence). Datawere then expressed as the ratio ofcytosolic GFP fluorescence to totalfluorescence � 100% (% internal-ized receptor).

RESULTS

Plasma Prekallikrein MediatesBradykinin-independent Activationof the ERK1/2 Cascade in PrimaryVascular Smooth Muscle but NotEndothelial Cells—To determinewhether vascular smooth muscle orendothelial cells are able to acti-vate plasma PK, we assayed KKactivity in R-VSMC, H-VSMC,human umbilical vein endothelialcells, and human aortic endothe-lial cells in cultures incubated with100 nM human plasma PK. Asshown in Fig. 1, VSMC from eithersource were able to activateplasma PK, leading to cleavage ofthe fluorogenic KK substrate,S2302. In contrast, umbilical veinand aortic endothelial cells wereunable to activate PK in vitro.As shown in Fig. 2A, R-VSMC

exposed to plasma PK at physio-logic concentrations (40) exhib-ited a time- and dose-dependentincrease in ERK1/2 phosphoryla-tion that paralleled the timecourse of PK activation. The effectrequired KK proteolytic activity,

as it was blocked by the KK inhibitor, aprotinin (data notshown). Because the PK-induced ERK1/2 response occurredin the absence of an exogenous source of kininogen, wehypothesized that it resulted from proteolysis of an endoge-nous VSMC substrate. To exclude a role for BK receptors, wetested whether the BK receptor antagonist, HOE140,affected KK-induced ERK1/2 activation. As shown in Fig. 2B,HOE140 had no effect on KK-stimulated ERK1/2 phospho-rylation at a concentration sufficient to block the effect ofexogenously supplied BK.PlasmaKallikreinActivates PAR1andPAR2Leading to PAR-

dependent ADAMActivation—Plasma KK is a trypsin fold ser-ine protease that cleaves substrates following Arg or Lys resi-

FIGURE 4. Plasma kallikrein causes PAR-dependent ADAM17 activation in vascular smooth musclecells. A, serum-deprived R-VSMCs in 96-well plates were treated with varying concentrations of plasma KKfor 4 h in the presence of the ADAM10/17-specific fluorogenic peptide substrate, MCA-KPLGL-Dpa-AR-NH2 (20 �M). ADAM activity was quantified as the increased fluorescence of the cleaved substrate andexpressed as percent of the activity observed when the incubation was performed in the absence of KK.Data shown represent mean � S.D. of triplicate determinations in one of three experiments that producedcomparable results. *, greater than untreated control, p � 0.05. B, serum-deprived R-VSMCs in 10-cmplates were treated with 20 nM plasma KK for 5–15 min, after which cells were lysed, and ADAM17 activitywas measured using the Innozyme TACE Activity KitTM as described. Data shown represent mean � S.D. oftriplicate determinations in one of three experiments that produced comparable results. *, greater thanuntreated control, p � 0.05. C, 20 nM human plasma KK was mixed with the chromogenic substrate forplasma KK, S2302 (0.6 mM), in the presence or absence of the PAR1 antagonist FLLRN (51) (P1Inh; 200 �M)or PAR2 antagonist FSLLRY (52) (P2Inh; 200 �M) (Peptides International; Louisville, KY), and KK activity wasmeasured as the change in absorption measured at 405 nm. D, serum-deprived R-VSMCs in 96-well plateswere preincubated for 15 min in the presence or absence of PAR1 or PAR2 antagonist peptides (200 �M)prior to treatment with 20 nM plasma KK for 4 h in the presence of MCA-KPLGL-Dpa-AR-NH2. ADAM activitywas quantified as the increased fluorescence of the cleaved substrate and expressed as percent of theactivity observed when the incubation was performed in the absence of KK. Data shown representmean � S.D. of triplicate determinations in one of three experiments that produced comparable results.*, greater than untreated control, p � 0.05. †, less than KK treated, p � 0.05.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

dues (41). As shown in Fig. 3A, the tethered ligand buriedwithin the N terminus of PARs is bounded by Arg or Lys, andalmost all known PAR activation is carried out by trypsin foldserine proteases. Moreover, several kallikrein-related pepti-dases have been shown to activate PARs with variable specific-ity (42–44), and kallikrein-related peptidase 4 was recentlyreported to activate PAR1 and PAR2, but not PAR4, in prostatecancer cells (21). To test whether plasmaKK activates PARs, weassayed for KK-dependent internalization of GFP-taggedPAR1, PAR2, and PAR4 expressed inHEK293 cells. Activation-dependent internalization is a characteristic feature of mostGPCRs (45). As shown in Fig. 3B, in vehicle-treated cells GFP-PAR1 was found primarily on the plasma membrane, with alesser amount of GFP fluorescence, probably representing nas-cent GFP-tagged receptors, present in the cytosol. Exposure to

either thrombin or KK for 15 mincaused a striking loss of GFP-PAR1from the plasma membrane andaccumulation within cytosolicpuncta that partially colocalizedwith the early endosomal markerEAA1. Fig. 3C compares the abilityof KK to stimulate internalization ofGFP-PAR1, -PAR2, and -PAR4.GFP-PAR1 and GFP-PAR4 movedfrom the plasmamembrane into thecytosol upon exposure to a knownactivator, thrombin, consistent withactivation-dependent receptor in-ternalization. GFP-PAR2 similarlyinternalized when exposed to tryp-sin, an endogenous activator ofPAR2.When exposed to plasmaKK,GFP-PAR1 and GFP-PAR2 but notGFP-PAR4 internalized, suggestingthat plasma KK activates PARs withsubstrate specificity similar to thatreported for kallikrein-related pep-tidase 4. Fig. 3D presents theseresults quantitatively.As shown in Fig. 4A, exposing

R-VSMCs to plasma KK produced adose-dependent increase in matrixmetalloprotease activity assayedusing the ADAM10/ADAM17-spe-cific fluorogenic substrate, KPLGL-Dpa-AR-NH2. As shown in Fig. 4B,repeating the experiment usingthe ADAM17/TACE-specific In-nozyme TACE activity assay con-firmed that the increase was due, atleast in part, to KK-dependent acti-vation of ADAM17. The KK effectwas detectable within 5 min ofexposure and persisted for at least15 min.Because ADAM17 is regulated by

multiple stimuli, including many Gprotein-coupled receptors (26, 27), we next tested whetherthere was a link between KK-dependent PAR activation and itsactivation of ADAM17 in R-VSMC. For these experiments,R-VSMCs were preincubated in the presence of PAR1 or PAR2antagonist peptides prior to exposing the cells to KK. As thesepeptides are related to the internal PAR1/2 ligand sequence, wefirst had to exclude the possibility that they would also behaveas pseudosubstrate inhibitors of plasma KK. As shown in Fig.4C, neither peptide had any effect upon KK activity whenassayed in a cell-free system. Similarly, neither peptide inter-feredwith the ability of recombinantADAM17/TACE to cleaveits fluorogenic substrate, KPLGL-Dpa-AR-NH2, in a cell-freeassay (data not shown).However, as shown in Fig. 4D, the PAR1and PAR2 antagonist peptides blocked KK-induced ADAMactivation in R-VSMCs, suggesting that KK cleavage of PARs

TNFα

TN

Fα

FIGURE 5. Plasma kallikrein stimulates ADAM-dependent amphiregulin and TNF-� release from vascularsmooth muscle cells. A, serum-deprived H-VSMC in 10-cm dishes were incubated in the presence or absenceof 50 nM human plasma KK for 1–3 h, after which the medium was collected, concentrated, and immunoblottedfor AR or TNF�, as described. B, serum-deprived H-VSMC were incubated with or without 50 nM human plasmaKK in the presence or absence of GM6001 (10 �M; Calbiochem), and AR shedding was determined. Data areexpressed as the percent increase in AR or TNF-� abundance compared with incubations performed in theabsence of KK. Each bar graph depicts the mean � S.E. for three separate experiments. *, greater than untreatedcontrol, p � 0.05; †, less than KK treated, p � 0.05. NS, nonstimulated.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

occurs upstream of ADAM activation. The two inhibitors incombination were somewhat more effective that either alone;however, the difference was not statistically significant.PlasmaKallikrein Stimulates ADAM-dependentAmphiregu-

linRelease andEGFReceptorTransactivation in PrimaryAorticVascular SmoothMuscle—ADAMsare known to play a key rolein the regulated shedding of at least six of the knownEGF familyligands, transforming growth factor-�, EGF, HB-EGF, betacel-lulin, epiregulin, and AR (46). Of these, HB-EGF and AR havebeen implicated in the control of VSMC proliferation anddevelopment of vascular disease (28–30, 47–49). To determinewhether the KK-induced increase in ADAM activity increasedVSMC shedding of EGF family growth factors, we looked forthe presence of AR and HB-EGF in the culture medium afterKK exposure. Because species-specific antibodies recognizingrat AR and HB-EGF are not available, we employed primaryH-VSMC for these experiments. As shown in Fig. 5A, KK expo-sure increased in AR shedding compared with vehicle-treatedcontrol cells. As shown in Fig. 5B, KK-stimulated AR releasewas ADAM-dependent, because it was sensitive to the broadspectrum MMP inhibitor GM6001. Although we were unableto detect HB-EGF shedding in response to plasma KK (data notshown), we did observe increased shedding of anotherADAM17/TACE substrate, TNF-�, consistent with the resultsof our direct assay of ADAM17/TACE activity. As shown in Fig.6A, GM6001 had no effect on KK activity in a cell-free assay,confirming that the inhibition of KK-induced AR shedding didnot result from direct inhibition of plasmaKK. As shown in Fig.6B, we observed a rapid 3-fold increase in EGF receptor auto-phosphorylation (ErbB1 pTyr1068) in R-VSMC upon KK treat-ment that was sensitive to GM6001, consistent with ADAM-dependent EGF receptor ligand shedding in KK-treated cells.

Kallikrein-stimulated ERK1/2 andJNK Activity in Primary Aortic Vas-cular Smooth Muscle Is PartiallyADAM- and EGF Receptor-de-pendent—EGF receptor transacti-vation contributes to ERK1/2 acti-vation by diverse stimuli, includingmany G protein-coupled receptors(26, 27). To test whether transacti-vation was involved in KK-inducedactivation of the ERK1/2 and JNKcascades in primary VSMC, wedetermined whether the activationof these pathways by PK or KK inR-VSMC was sensitive to inhibitionby GM6001 or the EGF receptorkinase inhibitor, AG1478. As shownin Fig. 7A, 15 min of exposure toeither PK or KK produced ERK1/2activation that was significantlyinhibited in the presence ofGM6001. Fig. 7B illustrates similareffects on PK- and KK-inducedJNK1/2 phosphorylation, implicat-ing MMP-dependent ectodomainshedding in both signals. As shown

in Fig. 7C, KK-induced ERK1/2 activationwas substantially, butincompletely, inhibited by AG1478 at a concentration suffi-cient to completely block ERK1/2 activation by exogenouslysupplied EGF. These results indicate that MMP-dependentEGF receptor transactivation plays a significant role in theVSMC response to KK exposure. The incomplete effect ofGM6001 and AG1478 indicates that other pathways, possiblyalso mediated via KK-activated PAR1/2, contribute to ERK1/2activation, as has been shown for other G protein-coupledreceptors (50).

DISCUSSION

The RAS and KKS constitute an interlocking signaling net-work involved in the regulation of vascular function (7–10).Coordinate regulation is achieved through shared pathwaycomponents. Plasma and tissue KK, which generate BK fromHMWKand lowmolecular weight kininogen, respectively, alsoconvert prorenin to renin, enhancing RAS activity, whereasangiotensin-converting enzyme, the source of angiotenisn II,also cleaves BK to an inactive fragment. The principal effectorsof the two systems, angiotensin II and BK, exert opposingeffects on vascular tone, with angiotensin II directly promotingVSMC contraction and BK indirectly causing vasodilationthrough BK receptor-mediated NO generation by vascularendothelial cells. Despite the generally salutary effects of BK, itis clear that dysregulation of theKKS is associatedwith progres-sion of the vascular and renal complications of diabetesmellitus(40). In the setting of endothelial denudation, BK can actdirectly on B1 and B2 bradykinin receptors expressed byVSMCto promote vasoconstriction in amanner similar to angiotensinII (13). The complex roles of the KKS in health and disease and

FIGURE 6. Plasma kallikrein stimulates ADAM-dependent EGF receptor transactivation in vascularsmooth muscle cells. A, 20 nM human plasma KK was combined with the chromogenic substrate for plasmaKK, S2302 (0.6 �M), in the presence or absence of GM6001 (10 �M), and KK activity was measured as the changein absorption measured at 405 nm. B, serum-deprived R-VSMCs in six-well plates were preincubated in thepresence or absence of GM6001 for 15 min prior to treatment with 20 nM plasma KK for 1–5 min. EGF receptorTyr1068 phosphorylation was determined by immunoblotting whole cell lysates as described. Data shownrepresents the mean � S.E. for five separate experiments. *, greater than untreated control, p � 0.05; †, lessthan KK treated, p � 0.05. NS, nonstimulated (vehicle).




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

its potential as a therapeutic target underscore the importanceof understanding its mechanisms of action in vascular tissue.Under normal physiologic conditions, VSMCs are not

directly exposed to plasma PK due to the presence of an intactendothelial barrier. As we demonstrate, endothelial cells lackthe ability to activate PK in the absence of factor XIIa generatedthrough activation of the intrinsic clotting system (1–3). Incontrast, we find that VSMCs, which could be exposed to PK inthe setting of vascular injury and endothelial denudation, donot require exogenous clotting cascade factors to generate KK.Activation of plasma PK by VSMC does involve proteolysis, asincubation of PK with VSMC leads to the generation of cleavedPK fragments, which is similar to that produced when PK isactivated by factor XIIa in a cell-free system (data not shown).However, the identity of the VSMC protease(s) responsible forPK activation remains to be determined. Unlike endothelialcells, activation of PK by VSMC also does not require HMWK,

whichmediates the binding of PK toendothelial cell membranes, indi-cating that the mechanism of PKbinding, as well as its cleavage/acti-vation by VSMC is distinct fromthat employed by endothelial cells.Our data suggest that PK activa-

tion by VSMC initiates a cascade ofsignaling events leading to EGFreceptor transactivation and activa-tion of the ERK1/2 and JNK cas-cades that, ironically, involves nei-ther BK or B1/B2 receptors. Asdepicted schematically in Fig. 8, PKbinding to an as yet unknown PKactivator on the surface of VSMCsgenerates KK. Among the sub-strates cleaved by KK are the N ter-mini of PAR1/2. One consequenceof PAR activation is increasedADAM activity, specifically includ-ing ADAM17/TACE, which acts oncell surface EGF receptor ligandprecursors, including AR, to trans-activate EGF receptors, and onunprocessed TNF� to enhance itsrelease. Although it is likely that invivo activation of PK on VSMC alsoleads to BK production and B1/B2receptor activation, our resultsdemonstrate a direct role for plasmaKK in the activation of PAR1/2 andstimulation of pathways involved inthe control of cell proliferation, apo-ptosis, and inflammation.Like BK receptors, PARs are

expressed on both endothelial cellsand VSMCs. Endothelial cells pri-marily express PAR1, althoughPAR2, PAR3, and PAR4 also arepresent (16, 18, 17). In normal arter-

ies, thrombin can trigger either endothelium-dependent relax-ation (53, 54) or endothelium-dependent contraction (55). Thedominant effect varies between vascular beds, with human andporcine coronary arteries undergoing vasodilation (54, 56),whereas in porcine renal interlobular arteries, thrombininduces a biphasic effect resulting from initial NO-dependentrelaxation followed by calcium-dependent contraction (55).Although direct thrombin-induced contraction of canine cor-onary arteries has been reported, VSMCsnormally express onlylow levels of PAR1 (57). VSMCexpression of PAR1 and PAR2 isup-regulated; however, under pathophysiologic conditions, forexample following balloon injury (58, 59) or in human athero-sclerotic lesions (60, 61), and is associated with an exaggeratedcontractile response to thrombin in vitro (62). Thrombin stim-ulates VSMCproliferation, hypertrophy, andmigration in vitro(63, 64) through Ca2�- and PKC-dependent effects on theexpression of egr-1, c-fos, c-jun, JunB, FosB, and fra-1 (65, 66).

FIGURE 7. Activation of ERK1/2 by plasma kallikrein in vascular smooth muscle cells is partially ADAM-and EGF receptor-dependent. A, serum-deprived R-VSMCs in 6-well plates were preincubated in the pres-ence or absence of GM6001 (10 �M) for 15 min prior to treatment with 200 nM PK or 20 nM plasma KK for 15 min.ERK1/2 phosphorylation was determined by immunoblotting whole cell lysates as described. Total cell ERK1/2was blotted as a control for equal protein loading. B, serum-deprived R-VSMCs were preincubated in thepresence or absence of GM6001 for 15 min prior to treatment with 200 nM PK or 20 nM plasma KK for 15 min. JNKphosphorylation was determined by immunoblotting whole cell lysates, and total cell JNK was blotted as acontrol for equal protein loading. C, serum-deprived R-VSMCs were preincubated in the presence or absence ofAG1478 (100 nM) for 15 min prior to treatment with 10 nM EGF or 20 nM plasma KK for 15 min. ERK1/2 phos-phorylation was determined by immunoblotting whole cell lysates, and total cell ERK1/2 was blotted as acontrol for equal protein loading. Each bar graph represents the mean � S.E. for three separate experiments. *,greater than untreated control, p � 0.05; †, less than stimulated in the absence of inhibitor, p � 0.05. NS,nonstimulated (vehicle).




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

As vascular injury would both expose VSMC to plasma PK andup-regulate PAR1/2 expression, our findings suggest thatPAR1/2 activation by VSMC-activated plasma KKmay exacer-bate the vascular injury response.One consequence we find of KK-mediated PAR1/2 activa-

tion in VSMCs is activation of ADAM family MMPs, notablyADAM17/TACE. PAR-dependent ADAM activation is associ-atedwith increasedAR andTNF-� secretion,MMP-dependentEGF receptor activation, and EGF receptor-dependent activa-tion of the ERK1/2 and JNK pathways. ADAM17 is the majorsheddase for HB-EGF and AR in VSMC. As with the PARs,expression of ADAM17, along with other components of theEGF receptor signaling network, is up-regulated in the settingof endothelial injury, shear stress, and atherosclerosis (29–33).Trans-activation of EGF receptors, triggered by diverse stimuli,including G protein-coupled receptor activation, oxidativestress, and fluid shear, has been shown to play an important rolein the vascular injury response, promoting VSMC hypertrophyand hyperplasia, as well as fibroblast proliferation (34–38).Whereas in normal arteries, an intact endothelium shields

the underlying VSMCs from exposure to plasma PK, endothe-lial cell BK receptors and PARs, on balance, produce vasodila-tory responses that oppose RAS-mediated vasoconstriction.However, loss of endothelial integrity may set up a “perfectstorm” wherein these normally protective factors exacerbatevascular injury and contribute to disease progression. Withboth PAR1/2 and EGF receptor pathway components up-regu-lated, VSMC are poised to respond to circulating factors in amanner that promotes vascular injury. Our data suggest thatplasma PK, activated on the surface of exposed VSMC, maybe one such factor, not only generating BK but causing directPAR1/2 activation, EGF receptor transactivation, and therelease of proinflammatory cytokines. As such, KK, whichalso increases RAS activity by activating prorenin, may be anattractive therapeutic target for the treatment of vasculardisease in high risk settings.

REFERENCES1. Sainz, I. M., Pixley, R. A., and Colman, R.W. (2007)Thromb. Haemost. 98,

77–832. Hermann, A., Arnhold, M., Kresse, H., Neth, P., and Fink, E. (1999) Im-

munopharmacology 45, 135–1393. Schmaier, A. H. (2008) Int. Immunopharmacol. 8, 161–1654. Motta, G., Rojkjaer, R., Hasan, A. A., Cines, D. B., and Schmaier, A. H.

(1998) Blood 91, 516–5285. Motta, G., Shariat-Madar, Z., Mahdi, F., Sampaio, C. A., and Schmaier,

A. H. (2001) Thromb. Haemost. 86, 840–8476. Mombouli, J. V., and Vanhoutte, P. M. (1995) Annu. Rev. Pharmacol.

Toxicol. 35, 679–7057. Schmaier, A. H. (2003) Am. J. Physiol. Regul. Integr. Comp. Physiol. 285,

R1–138. Schmaier, A. H. (2002) J. Clin. Invest. 109, 1007–10099. Kichuk, M. R., Seyedi, N., Zhang, X., Marboe, C. C., Michler, R. E., Ad-

donizio, L. J., Kaley, G., Nasjletti, A., and Hintze, T. H. (1996) Circulation94, 44–51

10. Siragy, H. M., Jaffa, A. A., and Margolius, H. S. (1997) Hypertension 29,757–762

11. Garg, U. C., and Hassid, A. (1989) J. Clin. Invest. 83, 1774–177712. Garg, U. C., and Hassid, A. (1989) Am. J. Physiol. 257, F60–6613. Briner, V. A., Tsai, P., and Schrier, R. W. (1993) Am. J. Physiol. 264,

F322–32714. Macfarlane, S. R., Seatter, M. J., Kanke, T., Hunter, G. D., and Plevin, R.

(2001) Pharmacol. Rev. 53, 245–28215. Ossovskaya, V. S., and Bunnett, N. W. (2004) Physiol. Rev. 84, 579–62116. Coughlin, S. R. (2005) J. Thromb. Haemost. 3, 1800–181417. Hirano, K. (2007) Arterioscler. Thromb. Vasc. Biol. 27, 27–3618. Vassallo, R. R., Jr., Kieber-Emmons, T., Cichowski, K., and Brass, L. F.

(1992) J. Biol. Chem. 267, 6081–608519. Ramachandran, R., and Hollenberg, M. D. (2008) Br. J. Pharmacol. 153,

S263–28220. Vandell, A. G., Larson, N., Laxmikanthan, G., Panos, M., Blaber, S. I.,

Blaber, M., and Scarisbrick, I. A. (2008) J. Neurochem. 107, 855–87021. Ramsay, A. J., Dong, Y., Hunt,M. L., Linn,M., Samaratunga, H., Clements,

J. A., and Hooper, J. D. (2008) J. Biol. Chem. 283, 12293–1230422. Hollenberg, M. D., Oikonomopoulou, K., Hansen, K. K., Saifeddine, M.,

Ramachandran, R., and Diamandis, E. P. (2008) Biol. Chem. 389, 643–65123. Yarden, Y., and Sliwkowski, M. X. (2001) Nat. Rev. Mol. Cell Biol. 2,

127–13724. Harris, R. C., Chung, E., and Coffey, R. J. (2003) Exp. Cell Res. 284, 2–1325. Schlondorff, J., and Blobel, C. P. (1999) J. Cell Sci. 112, 3603–361726. Carpenter, G. (2000) Sci. STKE. 2000, PE127. Zwick, E., Hackel, P. O., Prenzel, N., and Ullrich, A. (1999) Trends Phar-

macol. Sci. 20, 408–41228. Dreux, A. C., Lamb, D. J., Modjtahedi, H., and Ferns, G. A. (2006) Ather-

osclerosis 186, 38–5329. Canault, M., Peiretti, F., Kopp, F., Bonardo, B., Bonzi, M. F., Coudeyre,

J. C., Alessi,M. C., Juhan-Vague, I., andNalbone, G. (2006)Atherosclerosis187, 82–91

30. Nakata, A., Miyagawa, J., Yamashita, S., Nishida, M., Tamura, R.,Yamamori, K., Nakamura, T., Nozaki, S., Kameda-Takemura, K., Kawata,S., Taniguchi, N., Higashiyama, S., and Matsuzawa, Y. (1996) Circulation94, 2778–2786

31. Morita, T., Yoshizumi, M., Kurihara, H., Maemura, K., Nagai, R., andYazaki, Y. (1993) Biochem. Biophys. Res. Commun. 197, 256–262

32. Igura, T., Kawata, S.,Miyagawa, J., Inui, Y., Tamura, S., Fukuda, K., Isozaki,K., Yamamori, K., Taniguchi, N., Higashiyama, S., and Matsuzawa, Y.(1996) Arterioscler. Thromb. Vasc. Biol. 16, 1524–1531

33. Sunnarborg, S.W.,Hinkle, C. L., Stevenson,M., Russell,W. E., Raska, C. S.,Peschon, J. J., Castner, B. J., Gerhart, M. J., Paxton, R. J., Black, R. A., andLee, D. C. (2002) J. Biol. Chem. 277, 12838–12845

34. Chan, A. K., Kalmes, A., Hawkins, S., Daum, G., and Clowes, A.W. (2003)J. Vasc. Surg. 37, 644–649

35. Ohtsu, H., Dempsey, P. J., Frank, G. D., Brailoiu, E., Higuchi, S., Suzuki, H.,Nakashima, H., Eguchi, K., and Eguchi, S. (2006) Arterioscler. Thromb.

FIGURE 8. Proposed mechanism of bradykinin-independent kallikreineffects in vascular smooth muscle cells. Circulating plasma PK is activatedto KK upon binding to the surface of exposed VSMC. The N termini of PAR1and PAR2 undergo KK-dependent cleavage, exposing their internal tetheredligands and promoting PAR-dependent activation of ADAM17. ADAM17 acti-vation releases AR, causing EGF receptor transactivation and EGF receptor-dependent activation of downstream signaling pathways including the ERKand JNK cascades.




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Vasc. Biol. 26, e133–736. Kawakami, A., and Yoshida, M. (2005) J. Atheroscler. Thromb. 12, 73–7637. Touyz, R. M. (2006) Arterioscler. Thromb. Vasc. Biol. 26, 685–68838. Cunningham, K. S., and Gotlieb, A. I. (2005) Lab. Invest. 85, 9–2339. Tan, Y., Hutchison, F. N., and Jaffa, A. A. (2004)Am. J. Physiol. Heart. Circ.

Physiol. 286, H926–93240. Jaffa, A. A., Durazo-Arvizu, R., Zheng, D., Lackland, D. T., Srikanth, S.,

Garvey,W. T., Schmaier, A. H., and theDCCT/EDIC StudyGroup. (2003)Diabetes 52, 1215–1221

41. Gosalia, D. N., Salisbury, C. M., Ellman, J. A., and Diamond, S. L. (2005)Mol. Cell. Proteomics 4, 626–636

42. Oikonomopoulou, K., Hansen, K. K., Saifeddine, M., Tea, I., Blaber, M.,Blaber, S. I., Scarisbrick, I., Andrade-Gordon, P., Cottrell, G. S., Bunnett,N. W., Diamandis, E. P., and Hollenberg, M. D. (2006) J. Biol. Chem. 281,32095–32112

43. Angelo, P. F., Lima,A. R., Alves, F.M., Blaber, S. I., Scarisbrick, I. A., Blaber,M., Juliano, L., and Juliano, M. A. (2006) J. Biol. Chem. 281, 3116–3126

44. Stefansson, K., Brattsand, M., Roosterman, D., Kempkes, C., Bocheva, G.,Steinhoff, M., and Egelrud, T. (2008) J. Invest. Dermatol. 128, 18–25

45. Gaborik, Z., and Hunyady, L. (2004) Trends Endocrinol. Metab. 15,286–293

46. Sahin, U., Weskamp, G., Kelly, K., Zhou, H.M., Higashiyama, S., Peschon,J., Hartmann, D., Saftig, P., and Blobel, C. P. (2004) J. Cell Biol. 164,769–779

47. Kalmes, A., Daum, G., and Clowes, A.W. (2001)Ann. N.Y. Acad. Sci. 947,42–54

48. Shin, H. S., Lee, H. J., Nishida, M., Lee, M. S., Tamura, R., Yamashita, S.,Matsuzawa, Y., Lee, I. K., and Koh, G. Y. (2003) Circ. Res. 93, 302–310

49. Kato, M., Inazu, T., Kawai, Y., Masamura, K., Yoshida, M., Tanaka, N.,Miyamoto, K., and Miyamori, I. (2003) Biochem. Biophys. Res. Commun.301, 1109–1115

50. Luttrell, L. M. (2003) J. Mol. Endocrinol. 30, 117–12651. Martorell, L.,Martínez-Gonzalez, J., Rodríguez, C., Gentile,M., Calvayrac,

O., and Badimon, L. (2008) Thromb. Haemost. 99, 305–31552. Al-Ani, B., Saifeddine, M., Wijesuriya, S. J., and Hollenberg, M. D. (2002)

J. Pharmacol. Exp. Ther. 300, 702–70853. Tesfamariam, B. (1994) Am. J. Physiol. 267, H1962–196754. Mizuno, O., Hirano, K., Nishimura, J., Kubo, C., and Kanaide, H. (1998)

Eur. J. Pharmacol. 351, 67–7755. Derkach, D. N., Ihara, E., Hirano, K., Nishimura, J., Takahashi, S., and

Kanaide, H. (2000) Br. J. Pharmacol. 131, 1635–164256. Mizuno, O., Kobayashi, S., Hirano, K., Nishimura, J., Kubo, C., and Ka-

naide, H. (2000) Br. J. Pharmacol. 130, 1140–114657. Ku, D. D., and Zaleski, J. K. (1993) J. Cardiovasc. Pharmacol. 22, 609–61658. Wilcox, J. N., Rodriguez, J., Subramanian, R., Ollerenshaw, J., Zhong, C.,

Hayzer, D. J., Horaist, C., Hanson, S. R., Lumsden, A., and Salam, T. A.(1994) Circ. Res. 75, 1029–1038

59. Damiano, B. P., D’Andrea, M. R., de Garavilla, L., Cheung, W. M., andAndrade-Gordon, P. (1999) Thromb. Haemost. 81, 808–814

60. Nelken, N. A., Soifer, S. J., O’Keefe, J., Vu, T. K., Charo, I. F., and Coughlin,S. R. (1992) J. Clin. Invest. 90, 1614–1621

61. Napoli, C., de Nigris, F., Wallace, J. L., Hollenberg, M. D., Tajana, G., DeRosa, G., Sica, V., and Cirino, G. (2004) J. Clin. Pathol. 57, 513–516

62. Ku, D. D., and Dai, J. (1997) J. Cardiovasc. Pharmacol. 30, 649–65763. McNamara, C. A., Sarembock, I. J., Gimple, L. W., Fenton, J. W., 2nd,

Coughlin, S. R., and Owens, G. K. (1993) J. Clin. Invest. 91, 94–9864. Kanthou, C., Kanse, S. M., Newman, P., Kakkar, V. V., and Benzakour, O.

(1995) Blood Coagul. Fibrinolysis 6, 753–76065. Rothman, A., Wolner, B., Button, D., and Taylor, P. (1994) J. Biol. Chem.

269, 6399–640466. Kanthou, C., Benzakour, O., Patel, G., Deadman, J., Kakkar, V. V., and

Lupu, F. (1995) Thromb. Haemost. 74, 1340–1347




http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Luttrell, Louis M. Luttrell and Ayad A. JaffaRany T. Abdallah, Joo-Seob Keum, Mi-Hye Lee, Bing Wang, Monika Gooz, Deirdre K.

Protease-activated Receptorsand Signaling in Vascular Smooth Muscle through Direct Activation of

Plasma Kallikrein Promotes Epidermal Growth Factor Receptor Transactivation

doi: 10.1074/jbc.M110.171769 originally published online September 8, 20102010, 285:35206-35215.J. Biol. Chem.

10.1074/jbc.M110.171769Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/285/45/35206.full.html#ref-list-1

This article cites 66 references, 24 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M110.171769

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;285/45/35206&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/285/45/35206

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=285/45/35206&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/285/45/35206

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/285/45/35206.full.html#ref-list-1

http://www.jbc.org/

VOLUME 285 (2010) PAGES 35206 –35215DOI 10.1074/jbc.A110.171769

Plasma kallikrein promotes epidermal growth factorreceptor transactivation and signaling in vascularsmooth muscle through direct activation ofprotease-activated receptors.Rany T. Abdallah, Joo-Seob Keum, Hesham M. El-Shewy, Mi-Hye Lee,Bing Wang, Monika Gooz, Deirdre K. Luttrell, Louis M. Luttrell,and Ayad A. Jaffa

Dr. Hesham M. El-Shewy was inadvertently left off of the authorlist. The correct author list is shown above. Dr. El-Shewy’s affiliationis the Department of Medicine, Medical University of South Caro-lina, Charleston, South Carolina 29425.

VOLUME 286 (2011) PAGES 11909 –11918DOI 10.1074/jbc.A110.193359

An infant-associated bacterial commensal utilizes breastmilk sialyloligosaccharides.David A. Sela, Yanhong Li, Larry Lerno, Shuai Wu, Angela M. Marcobal,J. Bruce German, Xi Chen, Carlito B. Lebrilla, and David A. Mills

The grant information footnote should read as follows. Thisworkwassupported, in whole or in part, by National Institutes of Health NICHDAwards R01HD059127, R01HD065122, and R01HD061923 andNIGMS Grant R01GM076360. This work was also supported by grantsfrom theUniversity of CaliforniaDiscoveryGrant Program,DairyMan-agement Inc., and the California Dairy Research Foundation and byUnited States Department of Agriculture National Research Initiative-Cooperative State Research, Education, and Extension Service Award2008-35200-18776.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 286, NO. 26, p. 23620, July 1, 2011© 2011 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

23620 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286 • NUMBER 26 • JULY 1, 2011

ADDITIONS AND CORRECTIONS This paper is available online at www.jbc.org

We suggest that subscribers photocopy these corrections and insert the photocopies in the original publication at the location of the origi-nal article. Authors are urged to introduce these corrections into any reprints they distribute. Secondary (abstract) services are urged tocarry notice of these corrections as prominently as they carried the original abstracts.

Documents

Plasma Kallikrein Promotes Epidermal Growth Factor Receptor