Hypoxia activates b1-integrin via ERK 1/2 and p38 MAP kinasein human vascular smooth muscle cellsq

Florian Blaschke, Philipp Stawowy, Stephan Goetze, Oliver Hintz, Michael Gr€aafe,Ulrich Kintscher, Eckart Fleck, and Kristof Graf *

Deutsches Herzzentrum Berlin, Department of Cardiology, Augustenburger Platz 1, 13353 Berlin, Germany

Received 30 July 2002

Abstract

Hypoxia plays an important role in vascular remodeling and directly affects vascular smooth muscle cell (VSMC) functions.

VSMC adhesion participates in changes of vascular structure; however, little is known about VSMC adhesion under hypoxic

conditions. It was the aim of the present study to investigate the effects of hypoxia on adhesion mechanisms in human VSMCs.

Compared to normoxic cells, hypoxia (1% O2, 24 h) significantly increased adhesion of VSMCs to collagen I by 30.2% and fibro-

nectin by 58.0%. This effect was completely inhibited in the presence of the pharmacological ERK 1/2 mitogen-activated protein

kinase (MAPK) pathway inhibitor PD98059 (30 lM) or the p38 MAPK inhibitor SB203580 (1 lM). Basal adhesion of normoxic

cells was not affected by pretreatment with PD98059 and SB203580. Hypoxia induced a time-dependent activation of ERK 1/2 and

p38 MAPK activation in human VSMCs, which were completely abolished by PD98059 or SB203580, respectively. Since adhesion

of VSMCs to fibronectin and collagen I involves b1-integrin receptors, we used a blocking antibody against b1-integrin (P5D2) to

examine potential effects of hypoxia on b1-integrins. P5D2 significantly reduced VSMC adhesion to fibronectin and collagen I in

normoxia and hypoxia in a comparable manner; however, b1-integrin protein or mRNA levels were not affected by hypoxia. As

evidenced by flow cytometry, hypoxia induced a activation of b1-integrins by exposing an conformationally sensitive epitope on the

b1-subunit. These results demonstrate that hypoxia enhances adhesion of VSMC on extracellular matrix proteins by activating

b1-integrin. � 2002 Elsevier Science (USA). All rights reserved.

Keywords: Hypoxia; Adhesion; Vascular smooth muscle cell; MAPK; Integrin

Hypoxia contributes to alterations in vascular struc-ture that define the process of vascular remodeling inhumans. Vascular remodeling events in smooth musclecells (VSMCs) are characterized by changes in VSMCphenotype, hypertrophy, proliferation, increased secre-tion of extracellular matrix proteins such as collagen Iand fibronectin, and an enhanced motility and chemo-tactic response of VSMCs. The capacity of VSMCs toadhere to the proteins of the extracellular matrix is animportant prerequisite for many of these cellular func-tions. Interactions between smooth muscle cells and theextracellular matrix proteins are mediated by the inte-

grin family of cell-surface receptors, which are com-posed of a and b subunit heterodimers [1]. Members ofthe b1-integrin family mainly mediate binding ofVSMCs to the extracellular matrix proteins collagen I(a1b1- and a2b1-integrin) and fibronectin (a5b1-integrin)[2,3]. Integrins transduce biochemical signals across thecell membrane (outside-in signaling) via activation ofintracellular signaling pathways [4–6] that include themitogen-activated protein kinase (MAPK) familymembers ERK 1/2 and p38 MAPK. Activation of theseMAPK results in activation and phosphorylation ofseveral cytosolic signaling molecules such as myosinlight chain kinases (MLCK), as well as nuclear tran-scription factors, including Egr-1, Ets-1, Elk-1, and c-fos. Thus, integrins link transmembrane signaling withcytosolic signaling events and gene induction in thenucleus [7] to regulate cell proliferation, migration, andapoptosis [8,9]. In addition, growth factors such as

Biochemical and Biophysical Research Communications 296 (2002) 890–896

www.academicpress.com

BBRC

qAbbreviations: VSMC, vascular smooth muscle cell; FCS, fetal calf

serum; MAPK, mitogen-activated protein kinase; ERK 1/2, extracel-

lular signal-regulated kinases 1 and 2; MEK, MAPK–ERK kinase.* Corresponding author. Fax: +49-30-4593-2415.

E-mail address: graf@dhzb.de (K. Graf).

0006-291X/02/$ - see front matter � 2002 Elsevier Science (USA). All rights reserved.

PII: S0006 -291X(02 )02033 -8

platelet-derived growth factor (PDGF) or angiotensin IIcan activate intracellular signaling cascades, leading tobinding of an integrin activation complex to the cyto-plasmic domain of the integrin subunit. Binding of thiscomplex causes a change in the activation status of in-tegrin receptors (inside-out signaling) [10,11].

In this study, we examined the effects and the cellularmechanisms of hypoxia on b1-integrin-mediated adhe-sion of human VSMCs to fibronectin and collagen I.

Materials and methods

Chemicals. Dermal human collagen I was obtained from Vitrogen

and fibronectin from Gibco. The pharmacological ERK 1/2 MAPK

pathway inhibitor PD98059 and the p38 MAPK inhibitor SB203580

were from Calbiochem. Hybond enhanced chemiluminescence (ECL)

nitrocellulose membrane, horseradish peroxidase-linked anti-rabbit

and anti-mouse antibody, and ECL Western blotting detection re-

agents were from Amersham Life Science. Antibodies were purchased

from the following vendors: anti-b1-integrin (P5D2) and anti-a5-inte-

grin (BIIG2) from the Developmental Studies Hybridoma Bank; anti-

b3-integrin (7E3) was provided by Centocor; anti-a1-integrin (F-19)

and anti-a2-integrin (N-19) were from Santa-Cruz; anti-b1-integrin

(9EG7) was from Pharmingen; anti-FITC-conjugated secondary anti-

bodies (anti-goat, anti-rabbit) were from Sigma and anti-mouse was

from Pharmingen and Sigma; antibodies against phosphorylated

(Thr202/Tyr204) and total ERK 1/2 were from Promega; the phospho-

specific (Thr180/Tyr182) and total p38 MAPK antibodies were from

Cell Signalingand anti-a-smooth muscle actin was from DAKO.

Cell culture. Human vascular smooth muscle cells (VSMCs) were

prepared from the umbilical vein, as described previously [12]. The

cells were cultured in DMEM containing 10% fetal calf serum, 100U/

ml penicillin, 100lg/ml streptomycin, and 200mmol/L L-glutamine.

For all experiments, VSMCs (passage 2–6) were grown to 60–70%

confluence and made quiescent by serum starvation (0.2% FCS/

DMEM) for at least 48 h. When used, inhibitors were added 30min

before exposure to hypoxia. Each individual experiment represented in

the n value was performed with an independent preparation of

VSMCs. Cells were identified as VSMCs through their characteristic

growth pattern and detection of a-smooth muscle actin.

Exposure of human VSMC to hypoxia. A humified temperature-

controlled incubator (model no. 3165; Forma Scientific, Labotect,

G€oottingen, Germany) was used as hypoxic chamber. This incubator

allows control of internal oxygen levels between 0% and 21% while the

CO2 level is continously controlled at 5%. For hypoxic stimulation,

subconfluent cells were placed in the incubator, flushed with a gas

mixture containing 5% CO2, 1% O2, and 94% N2, and incubated at

37 �C for different times. Normoxia was defined as 95% air and 5%

CO2.The viability of human VSMCs, assessed by trypan blue exclusion

and by measuring lactate dehydrogenase activity in cell supernatants,

was not affected by exposure to hypoxia (data not shown).

Flow cytometry. Integrin cell surface expression was evaluated by

indirect immunofluorescence using flow cytometry [13]. Cells were in-

cubated on ice with primary antibodies at predetermined saturating

concentrations, followed by incubation with appropriate fluorescein

isothiocyanate (FITC)-conjugated secondary antibodies (dilutions:

anti-goat 1:100, anti-mouse 1:30, anti-rabbit 1:40, and anti-rat 1:50).

Non-specific fluorescence was determined using a non-specific mouse

or rat IgG. Analysis was performed using a Becton–Dickinson FAC-

Scalibur flow cytometer.

Reverse-transcription polymerase chain reaction assay. Total RNA

was isolated from VSMCs through the use of RNAzol. Reverse tran-

scription polymerase chain reaction (RT-PCR) was performed fol-

lowing standard protocols. PCR was performed in a Perkin–Elmer PE

9700 thermal cycler as a hot-start PCR. After initial denaturation at

95 �C for 5min, PCR amplification was performed by using denatu-

rating steps for 30 s at 95 �C, annealing for 40 s at 58 �C, primer ex-

tension for 30 s at 72 �C, and a final extension for 10min at 72 �C. b1,

b3, and pyruvate dehydrogenase (PDH) were amplified for 30 cycles.

PDH was used as internal standard for RNA loading. Products were

separated by gel electrophoresis (2% agarose gel) and visualized by

ethidium bromide staining. Primer sequences for b1 were 50-ACAC

GTCTCTCTCTGTCG-30, and 50-CAGTTGTTACGGCACTCT-30,

for b3 50-TTCTCCATCCAAGTGCGGCA-30 and 50-GCCAATCCG

CAGGTTACTGG-30, and for PDH 50- GGTATGGATGAGGAGCT

GGA-30 and 50-CTTCCACAGCCCTCGACTAA-30. Primers for am-

plifying VEGF (516-bp amplicon for VEGF189 and 444-bp amplicon

for VEGF165) were purchased from R&D. Genbank Accession Codes

were M84237 for human b1-integrin, S49379 for human b3-integrin,

and X62568 for VEGF.

Adhesion. The efficiency of cell attachment was determined by

measuring the number of cells that adhered to a substrate, as described

previously [13]. Ninty-six-well plates were coated overnight at 4 �Cwith fibronectin (20lg/mL) or collagen I (20 lg/mL). When used as

inhibitors, cells were pretreated with blocking antibodies against b1-

and b3-integrin for 30min at room temperature. A total of 30 000 cells

were placed in each well and allowed to adhere at 37 �C for 60min.

After adhesion, cells were stained with 0.5% toluidine blue, followed by

solubilization in 1% SDS, and quantification in a microtiter plate

reader at 590 nm.

Western blot analysis. Cells were made quiescent by serum starva-

tion for 48 h and then exposed to hypoxia for the indicated times. For

inhibitor studies, cells were pretreated for 30min with PD98059

(30lM) or SB203580 (1 lM). Cells were lysed with radioimmunopre-

cipitation (RIPA) buffer supplemented with proteinase inhibitors for

20min on ice. Equal amounts of protein (30lg) were separated by

SDS–PAGE (7.5% standard gel) and Western blot analysis was per-

formed following standard protocols. Proteins were blotted onto ni-

trocellulose membranes and then incubated with antibodies that

recognize either (a) total ERK 1/2, (b) ERK 1/2, when phosphorylated

on threonine 202 and tyrosine 204 or (a) total p38 MAPK, (b) p38

MAPK, when phosphorylated on threonine 180 and tyrosine 182.

Antibodies were used at a concentration of 1:500 (phospho p38

MAPK), 1:2000 (ERK 1/2), and 1:5000 (phospho-ERK 1/2). Immu-

noreactive bands were visualized with horseradish peroxidase-conju-

gated secondary antibodies (1:1000 dilution). The peroxidase reaction

was developed with an ECL detection system (Amersham Corp). All

Western blot experiments were repeated at least three times with dif-

ferent cell preparations. Band intensity was analyzed by densitometry

using the NIH Image Program 1.62 for a Macintosh PC and expressed

in arbitrary units (arb.U.).

Statistical analysis. Analysis of variance (ANOVA) and paired or

unpairedt tests were performed for statistical analysis, as appropriate.

A P value less than 0.05 was considered to be statistically significant.

Data are expressed as means� SEM.

Results

Hypoxia-induced increase of adhesion of human VSMCsto fibronectin and collagen I is ERK 1/2- and p38 MAPK-dependent

To investigate the influence of hypoxia on cell–matrixinteractions, adhesion assays were performed. HumanVSMCs were made quiescent by serum starvation (0.2%FCS) for 48 h and then exposed to hypoxia (1% O2).Short-term hypoxia (up to 12 h) did not affect adhesion tothe extracellular matrix proteins fibronectin or collagen I

F. Blaschke et al. / Biochemical and Biophysical Research Communications 296 (2002) 890–896 891

(Figs. 1A and 2A). However, exposure to hypoxia for24 h induced a significant increase in adhesion to fibro-nectin (58.0%, P < 0:01 versus unstimulated cells) andcollagen I (30.2%, P < 0:01 versus unstimulated cells)(Figs. 1A and 2B). Pretreatment with the MEK inhibitor

PD98059 (30 lmol/L) completely inhibited the hypoxia-mediated increase in adhesion to fibronectin (P < 0:05,Fig. 1B) or collagen I (P < 0:05, Fig. 2B). Reoxygenation(24 h) following hypoxia restored adhesion to fibronectinor collagen I to normoxic levels (both P < 0:05 versushypoxia). The effect of hypoxia was also completelyabolished in the presence of the p38 MAPK inhibitorSB203580 (1 lmol/L). Basal adhesion of unstimulatedcells was not altered by PD98059 and SB203580 (Figs. 1Band 2B), indicating that ERK 1/2- and p38 MAPK-dependent signaling events in response to hypoxia arerequired for VSMC adhesion. Preincubation with the

Fig. 1. Hypoxia-induced increase of adhesion of human VSMCs

(30 000/well) to fibronectin-coated (20 lg/ml) plates is ERK 1/2- and

p38 MAPK-dependent. Cells were made quiescent by serum starvation

(0.2 FCS) for 48 h, exposed to hypoxia (1% O2) for the indicated time

period, and allowed to adhere for 1 h at 37 �C in 96-well plates. (A)

Cells were exposed to hypoxia for 3, 6, 12, and 24 h. (B) Cells were

preincubated with blocking antibodies against b1- (10lg/ml) or b3-

integrin antibody (10lg/ml) for 30min, prior to exposure to hypoxia

(24 h). (C) Cells were incubated with PD98059 [PD] (30lM) or

SB203580 [SB] (1lM) before exposure to hypoxia for 24 h. Experi-

ments were performed in quadruplicate, with three different sets of

cells. Data are expressed as means� SEM. *P < 0:01 vs normoxia (N),

**P < 0:05 vs normoxia alone, and #P < 0:05 vs hypoxia (Hy) alone.

Fig. 2. Hypoxia-induced increase of adhesion of human VSMCs

(30 000/well) to collagen I-coated (20lg/ml) plates is ERK 1/2-and p38

MAPK-dependent. Cells underwent the same treatment protocol, as

described in figure legend 1. Groups (2A–C) were formed according to

figure legend 1. Experiments were performed in quadruplicate, with

three different sets of cells. Data are expressed as means� SEM.

*P < 0:01 vs normoxia (N), **P < 0:05 vs normoxia alone, and

#P < 0:05 vs hypoxia (Hy) alone.

892 F. Blaschke et al. / Biochemical and Biophysical Research Communications 296 (2002) 890–896

b1-integrin blocking antibody P5D2 (10 lg/mL) inhibitedadhesion to fibronectin of normoxic (63.3% inhibition,P < 0:05 versus normoxia alone) and hypoxic cells(69.2% inhibition, P < 0:05 versus hypoxia alone, Fig.1C). Adhesion to collagen I of normoxic cells (35.1%inhibition, P < 0:05 versus normoxia alone) and cellsexposed to hypoxia (28.5% inhibition, P < 0:05 versusnormoxia alone) was also inhibited by pretreatment withP5D2 (Fig. 2C). The inhibitory effect of the b1-integrinantibody was not different between normoxic and hyp-oxic cells. In contrast, preincubation with the b3-integrinblocking antibody 7E3 (10 lg/mL) did not significantlyalter adhesion of cells exposed to normoxia or hypoxia onfibronectin and collagen I (Figs. 1C and 2C). Non-specific IgG did not also alter adhesion to both matrixproteins (data not shown).

Hypoxia induces ERK 1/2 and p38 MAPK activation inhuman VSMCs

To elucidate further the signaling events involved inhypoxia-induced effects on cell adhesion, we investigatedthe activation of ERK 1/2 and p38 MAPK by hypoxiain human VSMCs. Cells were made quiescent by serumstarvation (0.2% FCS) for 48 h and subjected to hypoxia(1% O2). Activated, phosphorylated ERK 1/2 was in-vestigated by immunoblotting with a phospho-specificERK 1/2 MAPK antibody. Hypoxia induced a rapidactivation of ERK 1/2, reaching a maximum after30min with a 3.0-fold induction (P < 0:05 versus un-stimulated cells) (Figs. 3A and C). Phosphorylation ofERK 1/2 remained elevated for 2 h of hypoxia (2.5-foldover unstimulated cells, P < 0:05). The total amount of

Fig. 3. Hypoxia stimulates phosphorylation of ERK 1/2 and p38 MAPK. Serum-starved VSMCs (0.2% FCS, 48 h) were exposed to hypoxia (Hy, 1%

O2) for the times indicated. (A) Representative immunoblots are shown with antibodies that recognize total ERK 1/2, phosphorylated ERK 1/2, total

p38 MAPK, and phosphorylated p38 MAPK. (B) Hypoxia-induced ERK 1/2 phosphorylation is blocked by PD98059 [PD] (30 lM). Activation of

p38 MAPK is inhibited by SB203580 [SB] (1lM) C, Densitometric analysis of the phosphorylated kinases ERK 1/2 and p38 MAPK after exposure to

hypoxia. Experiments were repeated 3 times; data are expressed as means� SEM. *P < 0:05 vs normoxia (N).

F. Blaschke et al. / Biochemical and Biophysical Research Communications 296 (2002) 890–896 893

ERK 1/2 was not affected by hypoxia, as shown in Fig.3A. Hypoxia-mediated phosphorylation of ERK 1/2MAPK was completely abolished by the MEK inhibitorPD98059 (30 lM) (Fig. 3B). The p38 MAPK-specificinhibitor SB203580 (1 lM) had no effect on hypoxia-induced ERK 1/2 phosphorylation (Fig. 3B). Thephosphorylation and activation of p38 MAPK in re-sponse to hypoxia was examined with a phospho-spe-cific p38 MAPK antibody. Hypoxia induced a rapid andtransient activation of p38 MAPK with a return tobaseline levels at 90min. The maximum of p38 MAPKphosphorylation was observed after 30min (2.5-foldinduction, P < 0:05 versus unstimulated cells, Figs. 3Aand C). Treatment with SB203580 (1 lM) inhibited hy-poxia-induced phosphorylation of p38 MAPK (Fig.3B). Reoxygenation (24 h) following 30min of hypoxiadid not induced phosphorylation of ERK 1/2 and p38MAPK compared to unstimulated cells (data notshown).

Hypoxia does not increase integrin expression in humanVSMCs

We demonstrated that binding of VSMCs to fibro-nectin and collagen I is antagonized by b1-integrinblocking antibodies. However, increased adhesion ofVSMCs to fibronectin and collagen I in response tohypoxia could result not only from an upregulation ofb1-integrin, but also from its heterodimeric partners a1,a2, or a5. We therefore investigated the expression ofthese integrins on VSMCs in response to hypoxia. Hu-man VSMCs were made quiescent by serum starvation(0.2% FCS) for 48 h and exposed to hypoxia (1% O2) for24 h (Fig. 4A) and 48 h (data not shown). Using flowcytometry, exposure to hypoxia revealed no significantchanges of a1-, a2-, a5-, b1-, and b3-integrin expression.Furthermore, RT-PCR assays revealed no significanteffect on the expression of b1- and b3-integrin mRNAsafter exposure to hypoxia for 6, 12, and 24 h, as shownin Fig. 4B. Vascular endothelial growth factor (VEGF)mRNA expression, known to be upregulated by hypoxiain human VSMCs [14], was used as positive control.Hypoxia increased VEGF mRNA expression within 6 h,reaching a maximum after 12–24 h with a 2.1-fold in-duction (P < 0:05 versus unstimulated cells) (Fig. 4B).

Hypoxia induces activation of b1-integrin

Since hypoxia does not change the quantitative ex-pression of integrins involved in cell adhesion, surfaceexpression of an activation-dependent epitope on b1-subunit (9EG7) was quantitated by flow cytometry (Fig.5). The epitope recognized by the monoclonal antibody9EG7 is conformationally sensitive and exposed in theactivated b1-integrin chain [15]. Human VSMCs wereserum starved for 48 h and exposed to hypoxia (1% O2)

Fig. 4. Hypoxia does not increase integrin protein or mRNA expres-

sion on human VSMCs. Cells were made quiescent by serum starva-

tion (0.2% FCS) for 48 h. (A) Cells were exposed to normoxia or

hypoxia (1% O2) for 24 h. Cells were then incubated with a non-specific

mouse IgG (gray area) or with specific anti-a1-, anti-a2-, anti-a5-, anti-

b1-, anti-b3-, or anti-b5-integrin antibodies at predetermined saturating

concentrations. The x-axis corresponds to fluorescence intensity on a

logarithmic scale and the y-axis to the cell number. Experiments were

repeated three times. (B) Cells were exposed to normoxia (N) or hy-

poxia (Hy, 1% O2) for 6, 12, or 24 h. Representative RT-PCR blots

from three separate experiments show transcription levels for b1- and

b3-integrins. VEGF was used as positive control for hypoxia induced

increase of mRNA expression. PDH, a ubiquitously, generally non-

regulated housekeeping gene was used as internal standard for RNA

loading.

894 F. Blaschke et al. / Biochemical and Biophysical Research Communications 296 (2002) 890–896

for 24 h. Hypoxia increased binding of 9EG7 to humanVSMCs compared to unstimulated cells. These resultsdemonstrate that hypoxia induces an activation ofb1-integrin (Fig. 5B), without changing its quantitativesurface expression (Fig. 4A).

Discussion

The present study demonstrates that exposure tohypoxia for 24 h (1% O2) induces a significant increase inb1-integrin mediated adhesion of human VSMCs to fi-bronectin and collagen I. This effect is preceded by aphosphorylation of ERK 1/2 and p38 MAPK. The

hypoxia-mediated increase of adhesion requires signal-ing through ERK 1/2 and p38 MAPK, since both theERK1/2 MAPK pathway inhibitor PD98059 andSB203580, an inhibitor of p38 MAPK, completely pre-vented the hypoxic effects on cell adhesion to fibronectinand collagen I. Although adhesion of VSMCs to thesematrix proteins is mainly mediated via b1-integrin re-ceptors, the quantitative cell surface expression ofb1-subunit was not altered by hypoxia. As evidenced byflow cytometry, hypoxia induced the activation ofb1-integrins by exposing an activation-dependent epi-tope on the b1-subunit on human VSMCs.

Recently, it has has been reported that exposure tohypoxia increased expression of avb3-integrin in humanumbilical vein endothelial cells [16]. In contrast, in thepresent study, we did not detect increased mRNA or cellsurface protein levels for b1- or b3-integrins after hypoxiafor 6–48 h. These different observations in endothelialcells and human VSMCs suggest cell type-specific dif-ferences in response to hypoxia. Our group and othershave demonstrated that adhesion of VSMCs to fibro-nectin and collagen I is b1-integrin-dependent [2,13]. Thisis confirmed by our present observation that adhesion ofnormoxic and hypoxic human VSMCs to fibronectin andcollagen I is inhibited by the specific b1-integrin blockingantibody P5D2. We have previously shown that b1-in-tegrin mediated adhesion is increased by angiotensin IIand PDGF without altering integrin expression [13]. Themechanism of changing affinity and avidity of integrinreceptors by intracellular signals, termed ‘‘inside-out’’signaling, has been characterized by several authors [17].Initiated by non-integrin receptors, an integrin activationcomplex binds to the cytoplasmic domain of the integrinsubunits. This leads to increased integrin affinity byconformational changes in the integrin subunits or in-creased avidity by integrin clustering in the cell mem-brane. Affinity modulation of avb3 and b2-integrins hasbeen described previously [18,19]. Our data demonstratethat hypoxia induces an activation of b1-integrin withoutchanging its quantitative surface expression. Binding ofthe specific antibody 9EG7 [15], which detects a con-formationally sensitive epitope exposed in activated b1-integrin heterodimers, was increased by hypoxia.

In previous reports, MAPK pathway has been re-ported to mediate integrin activation [20,21]. This is inaccordance with our results, demonstrating an involve-ment of ERK 1/2 and p38 MAPK in hypoxia modulatedcell–matrix interactions.

Previous studies have demonstrated different timepatterns of ERK 1/2 and p38 MAPK activation in re-sponse to hypoxia, depending on the cell type studied.Hypoxia has been reported to activate ERK 1/2 and/orp38 MAPK in PC12 cells, cultured rat cardiac myocytesand fibroblasts [22–24]. In PC12 cells, hypoxia-inducedphosphorylation of both kinases [23], whereas in adultbovine adventitial fibroblasts hypoxia activated only

Fig. 5. Hypoxia induces activation of b1-integrin. Cells were made

quiescent by serum starvation (0.2% FCS) for 48 h and then exposed to

normoxia or hypoxia (1% O2) for 24 h. Cell surface expression of an

activation-dependent epitope (9EG7) on b1-subunit was evaluated by

flow cytometry analysis. (A) Normoxic cells were incubated with a

non-specific rat IgG (dotted line) or with specific anti-b1-integrin an-

tibody (clone 9EG7). (B) Cells, exposed to normoxia (N, thin line) or

hypoxia (Hy, thick line), were incubated with b1-integrin antibody

(clone 9EG7).

F. Blaschke et al. / Biochemical and Biophysical Research Communications 296 (2002) 890–896 895

p38 MAPK [25]. In the present study, both ERK 1/2 andp38 MAPK were rapidly activated in response to hy-poxia, indicating that both protein kinases play animportant role in the observed hypoxia-mediated effectson adhesion of human VSMC. At present, the upstreammechanisms leading to hypoxia-induced activation ofERK1/2 and p38 MAP kinase are not fully understood.Smooth muscle cells have been shown to produce reac-tive oxygen species (ROS) in response to distinct cellularstresses. ERK 1/2 and p38 MAPK are activated by ROS[26,27], suggesting a candidate role of ROS as upstreammediator. Furthermore, it has recently been demon-strated that hypoxia-inducible factor-1 (HIF-1a) in-duction by thrombin involves ROS activation viap22phox-containing NADPH oxidase. These data suggestthat ROS activation might contribute to hypoxia-in-duced signaling [28]. However, this was not the focus ofthe present study and requires further investigation.

In summary, our study demonstrates that hypoxiaaugments b1-integrin-mediated adhesion of humanVSMCs via activation of the ERK 1/2 and p38 MAPKpathways.

Acknowledgments

This work was supported by a grant (to K.G.) from the Deutsche

Forschungsgemeinschaft (GR 1368/2-2). The authors gratefully ac-

knowledge the technical assistance of Chantel Spencer–H€aansch and

Heike Kallisch.

References

[1] R.O. Hynes, Integrins: versatility, modulation, and signaling in

cell adhesion, Cell 69 (1992) 11–25.

[2] R.I. Clyman, K.A. McDonald, R.H. Kramer, Integrin receptors

on aortic smooth muscle cells mediate adhesion to fibronectin,

laminin, and collagen, Circ. Res. 67 (1990) 175–186.

[3] R.I. Clyman, D.C. Turner, R.H. Kramer, An a 1/b 1-like integrin

receptor on rat aortic smooth muscle cells mediates adhesion to

laminin and collagen types I and IV, Arteriosclerosis 10 (1990)

402–409.

[4] E.A. Clark, J.S. Brugge, Integrins and signal transduction

pathways: the road taken, Science 268 (1995) 233–239.

[5] R.L. Juliano, S. Haskill, Signal transduction from the extracellu-

lar matrix, J. Cell. Biol. 120 (1993) 577–585.

[6] K. Yamamoto, M. Yamamoto, Cell adhesion receptors for native

and denatured type I collagens and fibronectin in rabbit arterial

smooth muscle cells in culture, Exp. Cell Res. 214 (1994) 258–263.

[7] M.J. Robinson, M.H. Cobb, Mitogen-activated protein kinase

pathways, Curr. Opin. Cell Biol. 9 (1997) 180–186.

[8] R.J. Davis, The mitogen-activated protein kinase signal transduc-

tion pathway, J. Biol. Chem. 268 (1993) 14553–14556.

[9] T. Force, J.V. Bonventre, Growth factors and mitogen-activated

protein kinases, Hypertension 31 (1998) 152–161.

[10] D. Neri, S. Montigiani, P.M. Kirkham, Biophysical methods for

the determination of antibody-antigen affinities, Trends Biotech-

nol. 14 (1996) 465–470.

[11] S.J. Shattil, H. Kashiwagi, N. Pampori, Integrin signaling: the

platelet paradigm, Blood 91 (1998) 2645–2657.

[12] J. Chamley-Campbell, G.R. Campbell, R. Ross, The smooth

muscle cell in culture, Physiol. Rev. 59 (1979) 1–61.

[13] K. Kappert, G. Schmidt, G. Doerr, B. Wollert-Wulf, E. Fleck, K.

Graf, Angiotensin II and PDGF-BB stimulate b1-integrin-medi-

ated adhesion and spreading in human VSMCs, Hypertension 35

(2000) 255–261.

[14] P. Koehne, C. Willam, E. Strauss, R. Schindler, K.U. Eckardt, C.

Buhrer, Lack of hypoxic stimulation of VEGF secretion from

neutrophils and platelets, Am. J. Physiol. Heart Circ. Physiol. 279

(2000) H817–824.

[15] M. Lenter, H. Uhlig, A. Hamann, P. Jeno, B. Imhof, D. Vestweber,

A monoclonal antibody against an activation epitope on mouse

integrin chain b1 blocks adhesion of lymphocytes to the endothelial

integrin a6b1, Proc. Natl. Acad. Sci. USA 90 (1993) 9051–9055.

[16] H.L. Walton, M.H. Corjay, S.N. Mohamed, S.A. Mousa, L.D.

Santomenna, T.M. Reilly, Hypoxia induces differential expression

of the integrin receptors avb3 and avb5 in cultured human

endothelial cells, J. Cell Biochem. 78 (2000) 674–680.

[17] R.S. Ross, T.K. Borg, Integrins and the myocardium, Circ. Res.

88 (2001) 1112–1119.

[18] T. Chavakis, A.E. May, K.T. Preissner, S.M. Kanse, Molecular

mechanisms of zinc-dependent leukocyte adhesion involving the

urokinase receptor and b2-integrins, Blood 93 (1999) 2976–2983.

[19] N. Pampori, T. Hato, D.G. Stupack, S. Aidoudi, D.A. Cheresh,

G.R. Nemerow, S.J. Shattil, Mechanisms and consequences of

affinity modulation of integrin avb3 detected with a novel patch-

engineered monovalent ligand, J. Biol. Chem. 274 (1999) 21609–

21616.

[20] Z. Li, X. Xi, X. Du, A mitogen-activated protein kinase-

dependent signaling pathway in the activation of platelet integrin

aIIbb3, J. Biol. Chem. 276 (2001) 42226–42232.

[21] Z. Zhang, K. Vuori, H. Wang, J.C. Reed, E. Ruoslahti, Integrin

activation by R-ras, Cell 85 (1996) 61–69.

[22] M. Das, D.M. Bouchey, M.J. Moore, D.C. Hopkins, R.A.

Nemenoff, K.R. Stenmark, Hypoxia-induced proliferative re-

sponse of vascular adventitial fibroblasts is dependent on g

protein-mediated activation of mitogen-activated protein kinases,

J. Biol. Chem. 276 (2001) 15631–15640.

[23] K.R. Laderoute, H.L. Mendonca, J.M. Calaoagan, A.M. Knapp,

A.J. Giaccia, P.J. Stork, Mitogen-activated protein kinase phos-

phatase-1 (MKP-1) expression is induced by low oxygen condi-

tions found in solid tumor microenvironments. A candidate MKP

for the inactivation of hypoxia-inducible stress-activated protein

kinase/c-Jun N-terminal protein kinase activity, J. Biol. Chem. 274

(1999) 12890–12897.

[24] Y. Seko, K. Tobe, K. Ueki, T. Kadowaki, Y. Yazaki, Hypoxia and

hypoxia/reoxygenation activate Raf-1, mitogen-activated protein

kinase kinase, mitogen-activated protein kinases, and S6 kinase in

cultured rat cardiac myocytes, Circ. Res. 78 (1996) 82–90.

[25] D.J. Welsh, P. Scott, R. Plevin, R. Wadsworth, A.J. Peacock,

Hypoxia enhances cellular proliferation and inositol 1,4, 5-

triphosphate generation in fibroblasts from bovine pulmonary

artery but not from mesenteric artery, Am. J. Respir. Crit. Care

Med. 158 (1998) 1757–1762.

[26] A.S. Baas, B.C. Berk, Differential activation of mitogen-activated

protein kinases by H2O2 and O2 in vascular smooth muscle cells,

Circ. Res. 77 (1995) 29–36.

[27] M. Ushio-Fukai, R.W. Alexander, M. Akers, K.K. Griendling,

p38 Mitogen-activated protein kinase is a critical component of

the redox-sensitive signaling pathways activated by angiotensin II.

Role in vascular smooth muscle cell hypertrophy, J. Biol. Chem.

273 (1998) 15022–15029.

[28] A. Gorlach, I. Diebold, V.B. Schini-Kerth, U. Berchner-Pfannsch-

midt, U. Roth, R.P. Brandes, T. Kietzmann, R. Busse, Thrombin

activates the hypoxia-inducible factor-1 signaling pathway in

vascular smooth muscle cells: Role of the p22phox-containing

NADPH oxidase, Circ. Res. 89 (2001) 47–54.

896 F. Blaschke et al. / Biochemical and Biophysical Research Communications 296 (2002) 890–896