© 2008 Schattauer GmbH, Stuttgart

1155

High glucose conditions induce upregulation of fractalkine andmonocyte chemotactic protein-1 in human smooth muscle cellsElena Dragomir1; Ileana Manduteanu1; Manuela Calin1; Ana Maria Gan1; Daniela Stan1; Rory Ryan Koenen2;Christian Weber2; Maya Simionescu1

1Institute of Cellular Biology and Pathology “Nicolae Simionescu”, Bucharest, Romania; 2Institute for Cardiovascular Molecular Research(IMCAR), University Hospital, Aachen, RWTH Aachen University, Aachen, Germany

SummaryThe major complication of diabetes mellitus is acceleratedatherosclerosis that entails an inflammatory process, in whichfractalkine and monocyte chemotactic protein-1 (MCP-1) play akey role.We investigated the effect of diabetes-associated highglucose (HG) on these chemokines and signalling mechanismsinvolved in human aortic smooth muscle cells (SMC). Exposureof SMC to HG resulted in an increase of fractalkine and MCP-1expression and the activated mitogen-activated protein kinase(MAPK) signalling pathway, a process associated with elevatedoxidative stress.Transfection with decoy oligodeoxynucleotidesidentified the involvement of transcription factors activator pro-tein 1 (AP-1) and nuclear factor kappa B (NF-κB) in the ob-served up-regulation of chemokines. The MAPK inhibitorsblocked the phosphorylation of IkBα and c-jun, indicating therole of MAPK in NF-κB and AP-1 activation in SMC under HG

KeywordsFractalkine, MCP-1, high glucose, smooth muscle cells, MAPKsignalling

conditions.The up-regulation of MCP-1 and fractalkine was as-sociated with increased adhesive interactions between HG-ex-posed SMC and monocytes. Treatment of HG-exposed SMCwith peroxisome proliferator-activated receptors α (PPARα)activators (fenofibrate and clofibrate) resulted in a reduction ofmRNA and protein expression of MCP-1 and fractalkine. In con-clusion, HG upregulates the expression of fractalkine andMCP-1 in SMC leading to increased monocyte-SMC adhesive in-teractions by a mechanism involving activation of MAPK, acti-vator protein-1 (AP-1) and NF-κB. The increased expression ofthese two pro-inflammatory chemokines and the ensuing in-creased adhesion between SMC and monocytes may trigger theinflammatory process associated with further vascular compli-cations of diabetes.

Thromb Haemost 2008; 100: 1155–1165

Cardiovascular Biology and Cell Signalling

Correspondence to:Maya Simionescu, PhDInstitute of Cellular Biology and Pathology “Nicolae Simionescu”8, BP Hasdeu Street, PO Box 35–14, 79691-Bucharest, RomaniaTel.: +40 21 319 4518, Fax: +40 21 319 4519E-mail: maya.simionescu@icbp.ro; elena.dragomir@icbp.ro

Financial support:This work was supported by the Romanian Academy and Ministry of Education and

Research grants – CEEX (1423 and 130/2006).

Received: February 22, 2008Accepted after major revision: September 24, 2008

Prepublished online: November 13, 2008doi:10.1160/TH08-02-0104

IntroductionThe increased expression of key inflammatory chemokines,monocyte chemotactic protein-1 (MCP-1) and fractalkine, is fre-quently associated with cell activation in cardiovascular diseases(1, 2).

In several pathological conditions, such as diabetes, the vas-cular smooth muscle cells (SMC) undergo phenotypic changesswitching from the quiescent “contractile” to the active “syn-thetic” state. Moreover, diabetes accelerates the accumulation ofSMC in atherosclerotic lesion (3), a process in which chemo-kines such as MCP-1 and fractalkine play a central role. The ef-fect of high glucose (HG) on the modulation of these chemo-kines in SMC and their functional role is unknown.

In-vitro and in-vivo studies have shown that chemokines,small chemotactic cytokines (7–14 kDa) are key molecules thatattract and activate specific subsets of immune cells (which ex-press G protein-coupled receptors for these chemokines) duringnormal physiological immunosurveillance as well as during in-flammation (3, 4).

Monocyte chemotactic protein (MCP) -1 (also known asCCL2) is one of the most thoroughly characterized chemokines.Numerous studies have revealed that MCP-1 exerts its significantrole in atherogenesis through promotion of monocyte/macro-phage accumulation in the atherosclerotic plaque, leading tochronic inflammation, SMC proliferation, and ultimately toplaque instability (2, 5). Recently, an additional effect of MCP-1in atherosclerotic lesion progression was reported, namely its ca-

Dragomir et al. Chemokine upregulation by high glucose in SMC

1156

pacity to act as a chemoattractant for SMC, inducing their mi-gration to the vessel intima (6). Interestingly, MCP-1 is express-ed besides leukocytes (7) by several non-immune cells, includingvascular SMC, in which tumour necrosis factor (TNF)-α in-creases MCP-1 production in a concentration-, and time-depend-ent manner by a mechanism involving activator protein-1 (AP-1)and nuclear factor kappa B (NF-κB)-dependent pathways (8).

Fractalkine, also known as CX3CL1, is a unique chemokinebelonging to CX3C chemokine family that exist both in mem-brane-tethered and shed form (9). Trans-membrane CX3CL1 isan adhesion molecule that mediates integrin-independent cellcapture by binding to fractalkine receptor (CX3CR1) on targetcells, whereas the soluble form promotes classical chemotacticresponse of monocytes, T cells and NK cells (10). In endothelialcells, vascular SMC, mesangial cells and fibroblasts, fractalkineexpression is increased by pro-inflammatory cytokines such asinterferon gamma (IFN-γ), TNF-α and interleukin-1 beta(IL-1β) (10). A role for fractalkine–CX3CR1-mediated inflam-mation has been demonstrated in several inflammatory disordersincluding atherosclerosis (1). Thus, in mice lacking CX3CR1,the progression of atheroma formation is significantly reduced(11). SMC from human atherosclerotic plaques expressCX3CR1 and undergo chemotaxis towards fractalkine, indicat-ing the role of this chemokine in SMC migration (12).

In diabetes, the increased production of reactive oxygenspecies (ROS) and their role in the induction of pro-inflamma-tory molecules is well established (13–15). Various pathways oftissue injury caused by high glucose (HG)-produced ROS acti-vate mitogen-activated protein kinase (MAPK) pathways (16)that further activate transcription factors (NF-κB and AP-1),which in turn regulate the expression of several significant genesin the pathogenesis of diabetic complications (17). Although it iswell known that hyperglycaemia induces a major increase inROS, its effect on the modulation of MCP-1 and fractalkine inSMC has not been reported.

Given the complexity of molecular mechanisms occurring indiabetes and its associated cardiovascular complications, the aimof this study was to determine the acute and direct effect of HGon MCP-1 and fractalkine modulation in SMC and the signallingmechanism involved. We report here that in human SMC, HG-produced ROS phosphorylate ERK1/2 and p38MAPK that inturn activates NF-κB and AP-1, leading to up-regulation ofMCP-1 and fractalkine and the ensuing increased monocyte ad-hesive interaction with SMC.

Materials and methodsReagentsDulbecco's modified Eagle medium (DMEM), RPMI-1640,antibiotics, diphenyleneiodonium (DPI), pyrrolidine dithioc-arbamate (PDTC), protein molecular weight standard, lucigenin,Ponceau S, monoclonal antibodies anti-MCP-1, and all otherchemicals were purchased from Sigma-Aldrich Chemie GmbH(Munich, Germany); fetal calf serum (FCS) was from Gibco-Life Technologies, Medist SA (Bucharest, Romania); CalceinAM Fluorescent Dye was from BD Bioscience, NovaintermedSRL (Bucharest, Romania); monoclonal antibodies againstfractalkine, the total and phosphorylated forms of p38MAPK,

ERK1/2 (p42/44MAPK) and JNK, the compatible secondaryantibodies and blocking buffer were from Santa Cruz Biotech-nology, Inc. (Santa Cruz, California, USA); monoclonal anti-Proliferating Cell Nuclear Antigen (PCNA) antibody was fromAbcam (Cambridge, UK). SuperSignalTM chemiluminescencesubstrate solution was from Pierce Biotechnology, Novain-termed SRL (Bucharest, Romania); Pertussis toxin and HighPure RNA isolation kit was purchased from ROCHE DiagnosticsGmbH (Mannheim, Germany); fenofibrate and clofibrate werefrom Calbiochem (Nottingham, UK); SB 203580 (p38MAPKinhibitor) and PD 98059 (ERK1/2 inhibitor) were from Pro-mega, Dexter Com S.R.L. (Bucharest, Romania). Oligodeoxinu-cleotides, CyQUANT proliferation kit, superfect reagent and allmolecular biology reagents were from Invitrogen Medist SA(Bucharest, Romania).

Cell cultureHuman aortic SMC were isolated from the media of foetal tho-racic aorta and characterized as a pure cell line devoid of anycontaminants (18). The cells were cultured at 37°C in DMEM,supplemented with essential and non-essential amino acids, so-dium selenite, ascorbic acid, 10% foetal calf serum (FCS,v/v)and antibiotics (100 units/ml penicillin, 100 µg/ml streptomycin,50 µg/ml neomycin) in a humidified incubator (5% CO2).Monocyte-like cell line U937 (kindly donated by Prof. S.C. Sil-verstein, Columbia University, NY, USA) were grown in suspen-sion in the RPMI 1640 culture medium containing 5% FCS andwere split 1:5, twice a week. Monocytic MonoMac6 cells werecultured in 24-well plates in RPMI 1640 medium supplementedwith 20% FCS, L-glutamine (2 mM), penicillin (200 U/ml),streptomycin (200 µg/ml), nonessential amino acids (Life Tech-nologies Inc., Bucharest, Romania) and OPI-supplement con-taining oxalacetic acid, sodium pyruvate and insulin (19).

Experimental designTo search for the effect of high glucose on MCP-1 and fract-alkine expression in SMC, in preliminary experiments, con-fluent cells were exposed for 24 hours (acute condition) to 5, 15,20, 25 and 33 mM glucose. Since the maximal effect was ob-tained at 33 mM glucose, the mechanisms involved in chemo-kine modulation was examined in cells exposed to 33 mM glu-cose or normal glucose concentration (5 mM) as control. After24 hours incubation, the cells were collected and subjected tofurther investigations. Similar glucose concentration were pre-viously employed to study the modifications of cellular pro-cesses induced by glucose (13, 15, 20, 21).

To find out the possible signalling mechanisms involved inthe modulation of chemokines, SMC were exposed to high glu-cose (as above) in the presence of specific MAPK inhibitors,SB203580 (10 µM) and PD98059 (20 µM) or ROS scavengers,diphenyleneiodonium (DPI, 5 µM) and pyrrolidine dithiocarba-mate (PDTC, 50 µM).

To examine the role of peroxisome proliferator-activated re-ceptors α (PPAR-α) in the high glucose (HG)-induced MCP-1and fractalkine expression, SMC were treated for 24 hours withHG in presence or absence of PPAR-α activators, fenofibrate(100 µM) or clofibrate (100 µM) and then subjected to RT-PCRand Western blot.

Dragomir et al. Chemokine upregulation by high glucose in SMC

1157

Reverse transcriptase-polymerase chain reaction(RT-PCR)Total cellular RNA was extracted from SMC using High PureRNA isolation kit. First-strand cDNA synthesis was performedusing 1 µg of total RNA and MMLV reverse transcriptase, ac-cording to the manufacturer’s protocol. The PCR reaction wasperformed at 56°C as annealing temperature using GoTaqpolymerase, cDNA and the corresponding primers: for MCP-1:sense 5'-AGCATCAAAGTCTCTGCCGCCCTTCTG and anti-sense 5'-ATTACTTAAGGCATAATGTTTCACA; for fract-alkine sense 5'-AACTCGAAATGGCGGCACCTT and anti-sense 5'-ATGAATTACTACCACAGCTCCG; for GAPDH:sense 5'-ACCACAGTCCATGCCATCAC and antisense5'-TCCACCACCCTGTTGCTGTA. The quantities of cDNAused for the analysis of fractalkine gene expression were twotimes those used for MCP-1 and GAPDH. The DNA fragmentwas amplified in 25 cycles (30s, 94°C; 30s, 56°C; 45sec, 68°C)followed by a final extension step of 7 minutes at 68°C. Theproducts were subjected to electrophoresis on 1.5% agarose geland analyzed on a gel analyzer system (Image MasterVDS-Phar-macia Biotech, Munich, Germany). The MCP-1 and fractalkinemRNA levels were normalized relative to GAPDH mRNA lev-els.

Real-time PCRTotal cellular RNA was isolated from cultured SMC using thetotal RNA kit. First-strand cDNA synthesis was performed using1 µg of total RNA and MMLV reverse transcriptase, according tothe manufacturer’s protocol. Quantification of NAD(P)H ox-idase subunit p22phox, was done by quantitative real-time PCRmethod using SYBR green I and Opticon 2 DNA Engine real-time thermocycler (MJ Research). Optimized amplification con-ditions were: 0.4 µM of each primer, 4 mM MgCl2, annealing at65°C and extension at 72°C for 40 cycles. The relative quantifi-cation of p22phox gene expression was carried out using thecomparative CT method (22) and the GAPDH gene as internalcontrol.

Smooth muscle cells proliferation assayCells were grown in 96 culture plates, exposed for 24 hours to5 or 33 mM D-glucose and the cell proliferation was performedwith CyQUANTCell ProliferationAssay Kit (Invitrogen, MedistSA, Bucharest, Romania) according to the manufacturer’s in-structions. Briefly, cells were frozen, thawed, and lysed with theaddition of the lysis buffer containing the green fluorescent dye,CyQUANT GR, which binds to nucleic acids. The fluorescencelevels were read on a fluorescence microplate reader (Berthold,Mithras LB 940, Dexter, Bucharest, Romania) with filters for485 nm excitation and 538 nm emission.

Western blotThe protein expression of chemokines and the signalling mol-ecules were assessed in the total extract of SMC obtained byhomogenizing the cells in Laemmli electrophoresis samplebuffer. Samples containing 50 µg of proteins were separated on20% SDS-PAGE and electroblotted onto nitrocellulose mem-branes. The membranes were incubated overnight at 4°C withthe specific monoclonal antibody for MCP-1, fractalkine,

PCNA, phospho IκB and c-jun, phospho/total p38 MAPK,ERK1/2 followed by the secondary antibody conjugated withhorseradish peroxidase (1 h). The signals were visualized usingSuperSignal West Pico chemiluminescent substrate and quanti-fied by densitometry employing a gel analyzer system.

NADPH oxidase assayThe lucigenin-enhanced chemiluminescence assay was em-ployed to determine the NADPH oxidase activity in cell homo-genates, as previously described (23). Briefly, the reaction car-ried out in a total volume of 150 µl containing 5 mM phosphatebuffer, 1 mM EGTA, pH 7.0, 5 µM lucigenin and 100 µMNADPH, was started by addition of cell homogenate (100 µgprotein); the light emission was recorded using a microplateluminometer (Berthold, Dexter, Bucharest, Romania). TheNADPH oxidase activity was expressed as relative light units(RLU)/µg of total protein.

Monocyte adhesion to SMC under static conditionsMonocytes (U937 cells) were labelled with calcein (1 µg/ml for30 min) at 37°C.To follow the role of MCP-1 and fractalkine, themonocytes were preincubated for 30 minutes with calcein inpresence of anti-CCR2 (5 µg/ml) or anti-CX3CR1 (5 µg/ml).Control and HG-treated SMC were washed with warm culturemedium and incubated with labelled fluorescent monocytes(500,000/ml) for 30 minutes at 37°C, in RPMI 1640 supple-mented with 1% FCS. After washing, the fluorescence wasmeasured with the fluorescence plate reader (Berthold, MithrasLB 940; excitation 485 nm, emission 538 nm).

Monocyte adhesion to SMC under shear flowconditionsLaminar flow assay was performed as previously described (24).Briefly, SMC were grown to confluence in 35-mm Petri dishesthat were assembled as the lower part of a parallel wall flowchamber and mounted on the stage of an Olympus IX-50 in-verted microscope (Olympus Optical, Hamburg, Germany) with4× and 10× phase contrast objectives. Monocytes (Mono Mac 6cells, 0.5 × 106/ml) suspended in Hanks' buffered salt solutioncontaining 10 mM HEPES, pH 7.4 and 0.5% human serum albu-min, kept on a heating block at 37°C were perfused into the flowchamberat a rate of 1.5 dyn/cm2 for 5 minutes; Mg2+ (1 mM) andCa2+ (1 mM) were added to monocytes shortly before the assay.The number of firmly adherent cells after 5 minutes was quanti-fied in multiple high power fields by analysis of images recordedwith a long integration JVC 3CCD video camera and a JVC SRL 900 E video recorder (JVC, Japan), and expressed as cells persquare millimetre (mean ± SD).

Oligonucleotide decoy assayTransfection of decoy oligodeoxynucleotides (ODNs) (25) wasperformed using double-stranded DNA with sequences cor-responding to the consensus NF-κB, AP-1 binding sites orscrambled. The following sequences were used: NF-κB decoyODNs, 5’-AGTTGAGGGGACTTTCCCAGGC-3’ and3’-TCAACTCCCCTGAAAGGGTCCG-5’; AP-1 decoy ODNs,5’–CGCTTGATGACTCAGCCGGAA-3’ and 3’-GCGAAC-TACTGAGTCGGCCTT-5’; scrambled decoy ODNs,

Dragomir et al. Chemokine upregulation by high glucose in SMC

1158

Figure 1: High glucose induces a dose-dependent increase inMCP-1 and fractalkine (Fkn) expression in human vascularsmooth muscle cells (SMC). A, B) Confluent SMC were exposed for24 hours to culture media containing normal (5 mM) or increasingD-glucose concentrations (15, 20, 25 and 33 mM) and mRNA of MCP-1and fractalkine was quantified by RT-PCR and normalized to GAPDHmRNA. 33 mM L-glucose was used as osmotic control for glucose, N=4,

*p<0.05 versus normal glucose (5 mM). C, D) Protein expression ofMCP-1 and fractalkine in SMC homogenates after exposure to increas-ing D-glucose concentrations or L-glucose (33 mM). As revealed byWestern blot and densitometric analysis, increasing D-glucose concen-tration upregulate protein expression of both MCP-1 and fractalkine,whereas L-glucose has no effect.

Figure 2: High D-glucose (HG) increases adhesive interactionsbetween monocyte and SMC. A) Confluent SMC were exposed 24hours to HG (33 mM) or 33 mM L-glucose, incubated with fluorescentlylabelled monocytes and adhesion assays were performed in static con-dition. Note that high D-glucose, but not L-glucose, increase monocytesadhesion. Pre-treatment of monocytes with 5 µg/ml anti-CCR2 (MCP-1receptor) or anti-CX3CR1 (fractalkine receptor) reduce the number ofadherent monocytes to SMC N=4, *p<0.05 significantly higher than

5 mM glucose, **p<0.05, significantly lower than HG by ANOVA test.B) Arrest of human monocytes (MonoMac6) to normal and HG-treatedSMC assessed in laminar flow condition. Monocytes were perfusedthrough the chamber at a constant shear rate of 1.5 dyn/cm2 for 5 min-utes. The number of firmly attached monocytes to HG-exposed SMC isfour times higher, compared with control cells. N=3, p<0.001 versuscontrol. Pre-treatment of monocytes with pertussis toxin (PTX, 100 ng/ml) significantly decreases the monocyte arrest **p<0.005 versus HG.

Dragomir et al. Chemokine upregulation by high glucose in SMC

1159

5’-CATGTCGTCACTGCGCTCAT-3’and 3’-GTACAGCAGT-GACGCGAGTA-5’. Each pair of single-stranded ODNs was an-nealed for two hours, during which time the temperature was re-duced from 95 to 25ºC. Transient transfection was performedusing 500 nM decoy ODNs and Superfect reagent. The DNA/superfect ratio was 1:7.5 (wt/wt).The uptake efficiency was evalu-ated by fluorescence microscopy using an inverted fluorescencemicroscope (Nikon Microphot-SA) and FITC-labelled ODNs.

Statistical analysisThe data obtained from the experiments were expressed as themeans ± standard error (SE). Statistical evaluation was carriedout by one-way ANOVA test. The p value for multiple compari-sons was calculated using one-way Anova and Bonferroni testfrom the OriginPro 7.5 software. p<0.05 was considered statis-tically significant.

ResultsHigh glucose increases the fractalkine and MCP-1 geneand protein expression in SMCTo test the effect of glucose on fractalkine and MCP-1 gene ex-pression, SMC were incubated for 24 hours with normal (5 mM)or increasing concentrations of D-glucose (up to 33 mM) or 33mM L-glucose, followed by RT-PCR assay. The experiments re-vealed that control SMC expressed low message encodingMCP-1 and fractalkine mRNA. When the cells were exposed toincreasing glucose concentration, the MCP-1 gene expressionincreased as a function of glucose concentration attaining amaximal at 33 mM D-glucose (Fig. 1A), whereas the fractalkinegene expression started to increase significantly at 25 mMD-glucose. The gene expression of both chemokines was not in-fluenced by L-glucose (maximal concentration, Fig. 1A, B) ormannitol (to view Supplementary Fig. 1A, B, please visit www.thrombosis-online.com).

The protein expression of fractalkine and MCP-1 in SMC ex-posed to increasing glucose concentrations was evaluated byWestern blot assay. Compared to normal glucose, exposure ofSMC to HG for 24 hours increased the protein expression of bothchemokines in a dose-dependent manner (Fig. 1C, D). The pro-tein expression of MCP-1 and fractalkine was not significantlymodified by 33 mM L-glucose. In addition, HG did not induceSMC proliferation as determined by Western blot for PCNA andCyQUANT assay (Supplementary Fig. 1C, D).

Interactions between monocytes and high glucose-exposed SMCThe biological relevance of HG-induced MCP-1 and fractalkineexpression in SMC was tested by assaying the adhesive interac-tion between fluorescently labelled monocytes (U937 andMonoMac6) and HG-treated SMC, in static and shear flow con-ditions. These assays revealed that in both conditions monocytesadhered at a significantly higher rate to HG-exposed SMC thanto resting cells. Thus, under static conditions, exposure of SMCto HG resulted in a 34% increase in fluorescence intensity (Fig.2A). Blockade of MCP-1 or fractalkine receptor by pre-incu-bation of monocytes with anti-CCR2 or anti CX3CR1, respect-ively, led to a significant reduction of the adhesive interactions

between monocytes and HG-exposed SMC, suggesting the in-volvement of these chemokines in the process (Fig. 2A).

Similar experiments were performed under laminar flowconditions in which the monocytes (MonoMac6) were perfusedthrough the chamber at a constant shear rate of 1.5 dyn/cm2 for 5minutes. As determined by morphometrical analysis, under flowconditions the arrest of monocytes to HG-exposed SMC was ~four fold above the values obtained in controls (Fig. 2B). Pre-

Figure 3: Effect of HG on NADPH oxidase activity and NADPHoxidase subunit p22phox. A) Employing chemiluminescence assay,NADPH oxidase activity was determined in homogenates of SMC ex-pose to 5 mM or 33 mM glucose. Note that, as compared to 5 mM glu-cose, the 33 mM glucose increase the NADPH oxidase activity by 116%(*p<0.05, N=4). Effect of HG on gene and protein expression ofNADPH subunit p22phox were determined by real time PCR (B) andWestern blot (C), respectively. The p22phox gene (*p<0.005, N=3) andprotein expression (*p<0.05, N=4) is increased in SMC treated with HGas compared with control cells.

Dragomir et al. Chemokine upregulation by high glucose in SMC

1160

treatment of monocytes with pertussis toxin (100 ng/ml) signifi-cantly reduced the number of monocytes adherent to HG-ex-posed SMC, indicating the role of G protein-coupled receptorssuch as chemokine receptors in the process (Fig. 2B).

Effect of high glucose on NADPH oxidase activityTo search for the mechanism(s) involved in the up-regulation ofMCP-1 and fractalkine through HG, we first investigated the role ofthe oxidative stress by evaluating the nicotinamide adenine dinucleo-tide phosphate (NADPH) oxidase activity in HG-treated SMC. Theexperiments revealed that, as compared to control cells, exposure ofSMC to HG highly increased the NADPH oxidase activity (Fig. 3A).Since the up-regulation of NADPH oxidase activity has been shownto correlate with increases in its essential subunit, the p22phoxmRNA (26, 27), we examined the expression of this subunit in con-trol and HG-exposed SMC. RealTime PCR revealed that the level ofmRNA of p22phox expression was increased by 80% in HG-treatedSMC (Fig. 3B). Furthermore, Western blot experiments showed thatHG increased by three-fold the protein expression of p22phox inHG-exposed SMC as compared to control cells (Fig. 3C).

Involvement of ROS-induced p38 MAPK and ERK1/2activation on fractalkine and MCP-1 modulation byhigh glucose in SMCIt has been reported that ROS induced by HG promotes the in-flammatory effects by activating MAPK signalling pathway inendothelial cells (28). Under the conditions used in the presentstudy, we found that HG significantly increased phosphoERK1/2 and phospho p38MAPK in SMC (Fig. 4E, F), whereasphospho JNK was not affected (Supplementary Fig. 1E). Toexamine the role of the MAPK signalling pathway in the up-regulation of fractalkine and MCP-1 in SMC, we exposed cells toHG in the absence or presence of ERK1/2 and p38MAPK in-hibitors, SB203580 and PD98059, respectively. The RT-PCR as-says revealed that SB203580 and PD98059 reduced the gene ex-pression of fractalkine by 47% and 60%, respectively, and ofMCP-1 by 74% and 81%, respectively (Fig. 4A, B). Fur-thermore, both inhibitors decreased the fractalkine and MCP-1protein expression (Fig. 4C, D). These results indicate the role ofp38 MAPK and ERK1/2 in the up-regulation of these chemo-kines in HG-treated SMC. To further assess the effect of ROS on

Figure 4: MAPK signalling pathway isimplicated in HG-increase fractalkineand MCP-1 expression in SMC. A, B) SMCwere exposed to normal or HG (33 mM)concentration in the absence or presence ofp38MAPK inhibitor, SB203580 (SB) or ERK1/2inhibitor, PD98059 (PD) for 24 hours. Asquantified by RT-PCR, the SB203580 andPD98059 reduced the fractalkine gene ex-pression by 47% and 60%, respectively, and theMCP-1 gene expression by 74% and 81%,respectively. The results are representative forthree independent experiments.C, D) The p38MAPK inhibitor (SB) and ERK1/2inhibitor (PD) reduced the protein expressionof fractalkine (p<0.05, N=4) and MCP-1(p< 0.01, N=4) in HG-exposed SMC.E, F) Confluent SMC were incubated with5 mM or 33 mM glucose in the presence ofreactive oxygen species inhibitors, DPI andPDTC. As shown by Western blot, both in-hibitors reduce the HG-induced p38MAPK(p<0.01 and p<0.05, respectively) and ERK1/2expression (p<0.05, p<0.05). The results arerepresentative of three independentexperiments.

Dragomir et al. Chemokine upregulation by high glucose in SMC

1161

p38MAPK and ERK1/2 activation in HG-stimulated SMC, wetested the effect of ROS scavengers (DPI and PDTC) on MAPKactivation. Western blot experiments illustrated that both ROSscavengers (Fig. 4E, F) inhibited the phosphorylation of ERK1/2and p38MAPK induced by HG.

Role of NF-κB and AP-1 in the modulation offractalkine and MCP-1By employing transfection of doubled-stranded oligodeoxynu-cleotides containing cis-acting elements for the indicated tran-scription factors, the involvement of NF-κB and AP-1 pathwaysin the regulation of fractalkine and MCP-1 expression in HG-treated SMC could be determined. Three hours after transfectionwith DNA and Superfect, the cells were exposed to HG. First, weverified that double-stranded ODN tagged with FITC at their5’end were efficiently introduced into SMC nuclei using thedendrimer-based transfection reagent. The results showed that24 hours after transfection, FITC-labelled ODN were detectablein the nuclei of ~75% of the cells (Fig. 5A). No fluorescence wasfound in non-transfected cells. Transfected cells were homogen-ized and the fractalkine and MCP-1 expression was determinedby Western blot. The results showed that blocking of NF-κB andAP-1 binding sites with specific decoy ODN, downregulated thefractalkine and MCP-1 protein expression (Fig. 5B, C), indicat-ing the direct participation of these transcription factors in theregulation of chemokines in SMC.

Contribution of MAPK to NF-κB and AP-1 activation inHG-exposed SMCSince it has been reported that the phosphorylated forms ofAP-1subunit (c-jun) and of the NF-κB inhibitory protein (IκB) areequivalent with the activation state of AP-1 and NF-κB, respect-

ively, we evaluated the phospho-c-jun and phospho-IκB ex-pression. SMC exposed to HG in the presence or absence ofMAPK inhibitors were subjected to Western blot. The resultsshowed that SB203580 and PD98059 reduced the phosphory-lated form of c-jun by 70% and 80%, respectively, and IκB by70% and 74%, respectively (Fig. 6A, B). These data indicate theup-stream involvement of p38MAPK and ERK1/2 in NF-κB andAP-1 activation.

PPARα activators reduce MCP-1 and fractalkineexpression induced by high glucose in SMCSince we have demonstrated in previous reports that PPARα acti-vators reduce the expression of MCP-1 in HG-treated endothelialcells (29), we searched for the effect of these agonists on MCP-1and fractalkine up-regulation by HG in SMC. The experimentsshowed that the increased MCP-1 and fractalkine gene expressioninduced by HG was considerably reduced by exposure of the cellsto fenofibrate or clofibrate (Fig. 7A, B); in addition, Western blotassays revealed that both agonists also decreased the protein ex-pression of these chemokines (Fig. 7C, D). Exposure of SMC to100 µM fenofibrate or clofibrate for 24 hours did not affect cell vi-ability as detected by MTT assay (Supplementary Fig. 1F).

To assess the mechanism by which fenofibrate and clofibratereduced the MCP-1 and fractalkine expression we examined theeffect of these agonists on ERK1/2 and p38MAPK activation byWestern blot, using phospho-specific antibodies. We found thatthe HG-increased phosphorylation of p38MAPK and ERK1/2induced by HG in SMC was significantly inhibited by fenofi-brate or clofibrate (Fig. 7E, F). These data indicated that PPARαactivators effectively counteracted the HG up-regulation of fract-alkine and MCP-1 expression in SMC by decreasing the acti-vation of the MAPK pathway.

Figure 5: Blocking of NF-kB and AP-1binding sites by transfection with decoyoligodeoxynucleotides (ODN) inhibitsHG-induced fractalkine and MCP-1expression in SMC. A) FITC-labelled ODNtransfected SMC seen in fluorescence andphase contrast microscopy (a,b) and uponmerging the two images (c). Note that 24hours after transfection, the fluorescent ODNis localized in the cell nuclei. B, C) Non-transfected (C) or cells transfected with decoyODN directed to the NF-κB, AP-1 orscrambled (SC) binding sites were exposed for24 hours to HG. Protein expression offractalkine (B) and MCP-1 (C) was evaluated byWestern blot. As shown, inhibition of NF-κBand AP-1 binding sites led to a significantlyreduced protein expression of fractalkine(p<0.05, p<0.001) and MCP-1 (p<0.05,p<0.005).

Dragomir et al. Chemokine upregulation by high glucose in SMC

1162

DiscussionThe production of chemokines by inflammatory cells has beenwell-documented; there are also reports indicating that SMCstimulated with pro-inflammatory cytokines are a potent sourceof chemokines (30, 31). However, little is known on the effect ofhigh glucose on chemokine expression in SMC.

Our results revealed that HG upregulates the gene and pro-tein expression of MCP-1 and fractalkine in SMC, resulting inthe increased adhesive interaction between monocytes and HG-treated SMC. In addition, we explored the signalling mech-anisms involved in HG-induced up-regulation of these chemo-kine and demonstrated that ERK1/2 and p38MAPK protein ki-nases and the NF-κB and AP-1 transcription factors, are instru-

mental in this process. These data suggest that HG may act, atleast in part, by regulating the production and consequently thefunction of MCP-1 and fractalkine in SMC. The proliferation as-says that showed no differences between control and HG ex-posed SMC indicated that the up-regulation of both chemokinesexpression was not due to an immediate increased SMC prolifer-ation.

Whereas the mechanisms of leukocyte adhesion to the vascu-lar endothelium have been intensively investigated, less is knownabout the pro-adhesive role of SMC. The potential of vascularSMC to interact with monocytes is suggested by the reports thatshow that SMC from atherosclerotic lesion (but not of normalvessels) express adhesion molecules, chemokines and chemo-kine receptors (12, 32). Recent reports indicate a 2.5-fold higherarrest of monocytes to neointimal SMC by a process triggered byGRO-α and fractalkine (33). We have found that HG increasesthe adhesive interaction between SMC and monocytes, in bothstatic and laminar flow conditions. Blockade of MCP-1 or fract-alkine receptor on monocytes (by their pre-incubation with anti-bodies to chemokine receptors) decreased the adhesive inter-actions, suggesting their involvement in this process. Based onthese findings, we propose that the increased interactions be-tween monocytes and SMC exposed to HG contribute to themonocyte-derived macrophages retention within the artery wall.Moreover, since fractalkine and MCP-1 are chemoattractants forSMC (6, 12), the HG-induced increased expression of these che-mokines may provide a mechanism of SMC accumulation in theatherosclerotic plaques.

Although the detailed molecular events that control high glu-cose-induced modification of the vascular cells are unknown,numerous data demonstrate the central role of oxidative stress inthis process. Accumulation of superoxide anion (O2

-) is one ofthe reported candidates for the pathogenesis of vascular compli-cations in diabetic environment. It was demonstrated that highglucose level stimulates reactive oxygen species production viaprotein kinase C (PKC)-dependent activation of NADPH ox-idase in vascular cells and in renal mesangial cells (34, 35). In-side the cell, high glucose concentration increases the synthesisof diacylglycerol, which is a critical activating cofactor for theclassic isoforms of protein kinase C, -β, -δ and -α (36). WhenPKC is activated by intracellular high glucose concentration, inaddition to activation of NADPH oxidase, it has a variety of ef-fects on gene expression, producing cellular dysfunction in dia-betic conditions (36). In our study, we found that in SMC, highglucose increases the NADPH oxidase activity and the gene andprotein expression of its subunit, p22phox. We assumed that theintracellular signalling pathways leading to HG-dependent fract-alkine and MCP-1 increased expression in SMC might includeROS-activated ERK1/2, p38MAPK, and JNK.

It was reported that HG activates MAPK signalling pathwaysin SMC (28). Since no data exist on the involvement of MAPKpathway (or other pathways) in MCP-1 and fractalkine regu-lation in HG-exposed SMC, we searched for a possible role ofERK1/2, p38MAPK and JNK in the up-regulation of these che-mokines. The result showed that HG activates ERK1/2 andp38MAPK but not JNK. The pharmacological inhibitors ofERK1/2 and p38MAPK significantly reduced the expression ofchemokines induced by HG. In addition, we tested the effect of

Figure 6:Effect of high glucose on NF-κB andAP-1 activation de-termined by the phosphorylated forms of c-jun and IκB.Confluent SMC exposed to 5 mM or 33 mM glucose concentration werelysed and the phospho c-jun and phospho IκB protein expression (pc-junand pIκB) were determined byWestern blot and analysed by densitometry.A) As compared to normal, HG increases the expression of c-jun phos-phorylated form.Activation of c-jun is abolished by p38MAPK and ERK1/2inhibitors (SB203580 and PD98059, respectively). B) HG increases IκBphosphorylation as compared to control, a feature that is abolished byPD98059 and SB203580. (N=3,*p<0.01, significantly higher than 5 mM glu-cose, **p<0.01 significantly lower than 33 mM glucose).

Dragomir et al. Chemokine upregulation by high glucose in SMC

1163

ROS scavengers on ERK1/2 and p38MAPK expression andfound that incubation of SMC with HG in the presence of PDTCand DPI reduced the ERK1/2 and p38MAPK phosphorylation,underlining a role of ROS in MAPK signalling.

Recently, Chen and colleagues have shown that in rat SMCactivated by TNF-α, the CCL2/MCP-1 expression is dependenton the activation of ERK1/2 pathway but not of p38MAPK (8).By contrast, the results obtained on human endothelial and mes-angial cells show that p38MAPK, but not ERK1/2, mediate theinduction of CCL2/MCP-1 elicited by TNF-α and IL1-β, re-spectively (37). In addition, SP-610025 (JNK inhibitor) andSB203580, but not PD98095, significantly reduced TNF-α-in-duced fractalkine protein synthesis, demonstrating the role ofJNK and p38MAPK (38). These data suggest that MCP-1 andfractalkine modulation vary according to the cell type (endothe-lial cell, mesangial cells or SMC) and the agonists (TNF-α,IL1-β) that act on these cells.

It is known that many cellular responses induced by HGrequire intracellular signals transduced to two major transcrip-

tion factors, AP-1 and NF-κB (39, 40). The activation of AP-1 byTNF-α is regulated by three of the five known MAPK, theERK1/2, p38MAPK, and JNK (41, 42). A p38MAPK inhibitorhas been shown to reduce lysophosphatidic acid dependent phos-phorylation of IκB and NF-κB translocation to the nucleus (43),suggesting that NF-κB activation can be dependent on MAPKsignalling. We hypothesized that in SMC, MAPK signalling cas-cade activated by HG (via ROS production) may converge toAP-1 and/or NF-κB activation and induce the increased fractal-kine and MCP-1 gene and protein expression. The experimentsrevealed that treatment of HG-treated SMC with SB203580 orPD98095 inhibited phosphorylation of both IκBα and c-jun, in-dicating the role of p38MAPK and ERK1/2 in the activation ofNF-κB and AP-1, respectively. The role of these transcriptionfactors in fractalkine and MCP-1 modulation was established byexperiments in which SMC were transfected with decoy oligo-deoxynucleotides (for NF-κB, AP-1 and scrambled) that showedthat blocking of NF-κB and AP-1 binding sites downregulatedboth chemokines expression.

Figure 7:The inhibitory effectof fenofibrate (F) and clofibrate (C) onfractalkine and MCP-1 mRNA gene andprotein expression in SMC exposedto HG concentration.A, B) Confluent cells were exposed for 24hours to normal glucose (5 mM) or HGconcentration (33 mM) in the presence orabsence of fenofibrate (100 µM) or clofibrate(100 µM). The fractalkine and MCP-1 mRNAlevels were evaluated by RT-PCR andnormalized to GAPDH mRNA. As shown,fenofibrate and clofibrate downregulate thegene expression of fractalkine (63% and 85%,respectively) and of MCP-1 (61% and 68%,respectively). C, D) Protein expression ofMCP-1 and fractalkine in cellular homogenateof SMC treated with HG in the presence orabsence of PPARα activators. Western blot anddensitometric analysis reveal the inhibitoryeffect of F and C on fractalkine and MCP-1protein expression in HG-treated SMC (N=3,*p<0.01, **p<0.05). E, F) Western blot assaydisplaying the effect of F and C on ERK1/2 andp38MAPK activation in SMC exposed to HG.Treatment of SMC with 33 mM glucose inpresence of F or C reduced significantly thep38MAPK (p<0.005) and ERK1/2 activation(p<0.001).

4. Suzuki LA, Poot M, Gerrity RG, et al. Diabetes ac-celerates smooth muscle accumulation in lesions ofatherosclerosis: lack of direct growth-promoting ef-fects of high glucose levels. Diabetes 2001; 50:851–860.5. Libby P. Changing concepts of atherogenesis. J In-tern Med 2000; 247: 349–358.6. Lo IC, Shih JM, Jiang MJ. Reactive oxygen speciesand ERK 1/2 mediate monocyte chemotactic protein-

1-stimulated smooth muscle cell migration. J BiomedSci 2005; 12: 377–388.7. Cipriani B, Borsellino G, Poccia F, et al. Activationof C-C beta-chemokines in human peripheral bloodgammadelta T cells by isopentenyl pyrophosphate andregulation by cytokines. Blood 2000; 95: 39–47.8. Chen YM, Chiang WC, Lin SL, et al. Dual regu-lation of tumor necrosis factor-alpha-inducedCCL2/monocyte chemoattractant protein-1 expression

References1. Combadiere C, Potteaux S, Gao JL, et al. Decreasedatherosclerotic lesion formation in CX3CR1/apolipo-protein E double knockout mice. Circulation 2003;107: 1009–1016.2. Schober A, Zernecke A. Chemokines in vascularremodeling. Thromb Haemost 2007; 97: 730–737.3. Braunersreuther V, Mach F, Steffens S. The specificrole of chemokines in atherosclerosis. Thromb Hae-most 2007; 97: 714–721.

Dragomir et al. Chemokine upregulation by high glucose in SMC

1164

Since PPARα is a ligand-activated transcription factor thatregulates the expression of key genes involved in the inflamma-tory response in the vascular wall (44), we questioned whetherPPARα activators (fenofibrate and clofibrate) modulate the ex-pression of fractalkine and MCP-1 in SMC. We report here that inSMC exposed to HG, PPARα activators significantly reduced thefractalkine and MCP-1 protein and mRNA expression. It is knownthat the down-regulation of chemokine expression by fibrates maytake place by receptor-dependent and independent mechanisms.Fibrates activate PPARα and directly inhibit the NF-κB and AP-1subunits, p65 and c-jun, respectively, leading to inhibition ofNF-κB and AP-1 target gene expression (45). These data, corrob-orated our results showing that NF-κB and AP-1 are involved inMCP-1 and fractalkine expression, indicate that the effect of fi-brates on these chemokines is, at least in part, a receptor-depend-

ent process. The data expand on our previous results obtained onendothelial cells demonstrating that PPARα activators inhibitNF-κB and AP-1 activation and downregulate MCP-1 expression(29).

However, at the high concentrations of fibrates employed inthis study (100 µM) we can not exclude a receptor-independenteffect on SMC. The inhibition of MAPK pathway by fenofibrateor clofibrate (Fig. 7E, F) and the possible reduction of the c-fossubunit of AP-1 transcription factor can occur independently ofPPARα activation.

Although further studies are warranted, the anti-inflamma-tory effect of fenofibrate (demonstrated in this study) togetherwith previous data of the Fenofibrate Intervention and EventLowering in Diabetes (FIELD) trial (46) indicating promisingeffects with fenofibrate in prevention of the progression of dia-betes-related microvascular complications support the role of fe-nofibrate in preventing the progression of diabetes-relatedmicrovascular complications in type 2 diabetes.

A limitation of our study is the high glucose concentrationused. In the experiments using a gradient of glucose concen-trations (5–33 mM), the maximal effect was detected at 33 mMglucose, and for this reason, we employed this concentration inour study. To a lesser extent, 20 and 25 mM glucose also induceda significant increase in fractalkine and MCP-1 expression inSMC (Fig. 1).

The in-vitro data on HG increased expression of fractalkineand MCP-1 corroborate well with our preliminary results ob-tained in a diabetic mouse model, showing that in diabetic mouseaorta, the expression of these two chemokines is significantly in-creased (unpublished data), allowing the assumption that the up-regulation of these molecules also takes place in vivo.

Taken together, the experiments reported here show that inhuman aortic SMC, HG concentration induced-oxidative stresstriggers the MAPK signal transduction network, that further ac-tivate NF-κB and AP-1 transcription factors, which generate theup-regulation of fractalkine and MCP-1. The increased ex-pression of these pro-inflammatory chemokines, which are che-moattractants for other SMC and promote the monocytes – SMCadhesive interactions, may account in part, for the initiation ofthe inflammatory process and the diabetes-induced vascularcomplications. The uncovering of the sequence of molecularevents involved in high glucose induced-expression of chemo-kines in SMC may lead to new therapeutic targets in diabetes.

AcknowledgementsWe acknowledge the skilful assistance of Gabriela Mesca and Ioana Mano-lescu. The authors want to thank Adrian Manea for help with transfectionexperiments and the constructive discussion.

What is known about this topic?− Vascular smooth muscle cells (SMC) derived from a

mouse model of type 2 diabetes, display enhanced in-flammatory gene expression.

− High glucose-induces TNF-alpha shedding in aorticsmooth muscle cells and promotes vascular inflam-mation.

− High glucose level stimulates reactive oxygen speciesproduction via PKC-dependent activation of NADPH ox-idase in vascular cells.

− Fibrates exert a direct anti-atherogenic effect in vascula-ture independently of their lipid-lowering activity.

What does this paper add?− High glucose induces overexpression of MCP-1 and

fractalkine – two of the most important chemokines invascular diseases.

− High glucose facilitates monocytes- smooth muscle cellsadhesive interaction, by a process dependent of MCP-1 –CCR2 and fractalkine –CX3CR1 interactions.

− Direct inhibition of NF-κB and AP-1 binding sites de-creases MCP-1 and fractalkine expression in SMC ex-posed to high glucose.

− Activation of NF-κB and AP-1 transcription factors byglucose is dependent on oxidative stress induced –p38MAPK and ERK1/2 phosphorylation.

− PPARα activators (fenofibrate and clofibrate) inhibitMCP-1 and fractalkine expression induced by high glu-cose.

in vascular smooth muscle cells by nuclear factor-kap-paB and activator protein-1: modulation by type IIIphosphodiesterase inhibition. J Pharmacol Exp Ther2004; 309: 978–986.9. Bazan JF, Bacon KB, Hardiman G, et al. A newclass of membrane-bound chemokine with a CX3Cmotif. Nature 1997; 385: 640–644.10. 10.Umehara H, Bloom E, Okazaki T, et al. Fractal-kine and vascular injury. Trends Immunol 2001; 22:602–607.11. Lesnik P, Haskell CA, Charo IF. Decreased athero-sclerosis in CX3CR1-/- mice reveals a role for fractal-kine in atherogenesis. J Clin Invest 2003; 111:333–340.12. Lucas AD, Bursill C, Guzik TJ, et al. Smoothmuscle cells in human atherosclerotic plaques expressthe fractalkine receptor CX3CR1 and undergo chemo-taxis to the CX3C chemokine fractalkine (CX3CL1).Circulation 2003; 108: 2498–2504.13. Dragomir E, Manduteanu I,Voinea M, et al.Aspirinrectifies calcium homeostasis, decreases reactiveoxygen species, and increases NO production in highglucose-exposed human endothelial cells. J DiabetesComplications 2004; 18: 289–299.14. HinokioY, Suzuki S, Hirai M, et al. Oxidative DNAdamage in diabetes mellitus: its association with dia-betic complications. Diabetologia 1999; 42: 995–998.15. Manduteanu I, Voinea M, Antohe F, et al. Effect ofenoxaparin on high glucose-induced activation of en-dothelial cells. Eur J Pharmacol 2003; 477: 269–276.16. Srinivasan S, Bolick DT, Hatley ME, et al. Glucoseregulates interleukin-8 production in aortic endothelialcells through activation of the p38 mitogen-activatedprotein kinase pathway in diabetes. J Biol Chem 2004;279: 31930–31936.17. Mohamed AK, Bierhaus A, Schiekofer S, et al. Therole of oxidative stress and NF-kappaB activation inlate diabetic complications. Biofactors 1999; 10:157–167.18. Tirziu D, Jinga VV, Serban G, et al. The effects oflow density lipoproteins modified by incubation withchondroitin 6-sulfate on human aortic smooth musclecells. Atherosclerosis 1999; 147: 155–166.19. Weber C,Aepfelbacher M, Haag H, et al.Tumor ne-crosis factor induces enhanced responses to platelet-ac-tivating factor and differentiation in human monocyticMono Mac 6 cells. Eur J Immunol 1993; 23: 852–859.20. Maedler K, Storling J, Sturis J, et al. Glucose- andinterleukin-1beta-induced beta-cell apoptosis requiresCa2+ influx and extracellular signal-regulated kinase(ERK) 1/2 activation and is prevented by a sulfonylureareceptor 1/inwardly rectifying K+ channel 6.2 (SUR/Kir6.2) selective potassium channel opener in humanislets. Diabetes 2004; 53: 1706–1713.21. Takaishi H, Taniguchi T, Takahashi A, et al. Highglucose accelerates MCP-1 production via p38 MAPK

in vascular endothelial cells. Biochem Biophys ResCommun 2003; 305: 122–128.22. Pfaffl MW. A new mathematical model for relativequantification in real-time RT-PCR. Nucleic Acids Res2001; 29: e45.23. Ungvari Z, Csiszar A, Edwards JG, et al. Increasedsuperoxide production in coronary arteries in hyper-homocysteinemia: role of tumor necrosis factor-alpha,NAD(P)H oxidase, and inducible nitric oxide synthase.Arterioscler Thromb Vasc Biol 2003; 23: 418–424.24. von Hundelshausen P, Koenen RR, Sack M, et al.Heterophilic interactions of platelet factor 4 andRANTES promote monocyte arrest on endothelium.Blood 2005; 105: 924–930.25. Fichtner-Feigl S, Fuss IJ, Preiss JC, et al. Treatmentof murine Th1– and Th2-mediated inflammatory boweldisease with NF-kappa B decoy oligonucleotides. JClin Invest 2005; 115: 3057–3071.26. Sorescu D, Weiss D, Lassegue B, et al. Superoxideproduction and expression of nox family proteins inhuman atherosclerosis. Circulation 2002; 105:1429–1435.27. Zalba G, San Jose G, Moreno MU, et al. NADPHoxidase-mediated oxidative stress: genetic studies ofthe p22(phox) gene in hypertension. Antioxid RedoxSignal 2005; 7: 1327–1336.28. Srivastava AK. High glucose-induced activation ofprotein kinase signaling pathways in vascular smoothmuscle cells: a potential role in the pathogenesis of vas-cular dysfunction in diabetes (review). Int J Mol Med2002; 9: 85–89.29. Dragomir E, Tircol M, Manduteanu I, et al. Aspirinand PPAR-alpha activators inhibit monocyte chemoat-tractant protein-1 expression induced by high glucoseconcentration in human endothelial cells. Vascul Phar-macol 2006; 44: 440–449.30. De Keulenaer GW, Ushio-Fukai M, Yin Q, et al.Convergence of redox-sensitive and mitogen-activatedprotein kinase signaling pathways in tumor necrosisfactor-alpha-mediated monocyte chemoattractant pro-tein-1 induction in vascular smooth muscle cells. Arte-rioscler Thromb Vasc Biol 2000; 20: 385–391.31. Jordan NJ, Watson ML, Williams RJ, et al. Chemo-kine production by human vascular smooth musclecells: modulation by IL-13. Br J Pharmacol 1997; 122:749–757.32. Braun M, Pietsch P, Schror K, et al. Cellular ad-hesion molecules on vascular smooth muscle cells.Cardiovasc Res 1999; 41: 395–401.33. Zeiffer U, Schober A, Lietz M, et al. Neointimalsmooth muscle cells display a proinflammatory pheno-type resulting in increased leukocyte recruitment me-diated by P-selectin and chemokines. Circ Res 2004;94: 776–784.34. Inoguchi T, Sonta T, Tsubouchi H, et al. Protein ki-nase C-dependent increase in reactive oxygen species

(ROS) production in vascular tissues of diabetes: roleof vascular NAD(P)H oxidase. J Am Soc Nephrol2003; 14: S227–232.35. Nakagami H, Kaneda Y, Ogihara T, et al. Endothe-lial dysfunction in hyperglycemia as a trigger of athero-sclerosis. Curr Diabetes Rev 2005; 1: 59–63.36. Brownlee M. The pathobiology of diabetic compli-cations: a unifying mechanism. Diabetes 2005; 54:1615–1625.37. Goebeler M, Kilian K, Gillitzer R, et al. TheMKK6/p38 stress kinase cascade is critical for tumornecrosis factor-alpha-induced expression of monocyte-chemoattractant protein-1 in endothelial cells. Blood1999; 93: 857–865.38. Sukkar MB, Issa R, Xie S, et al. Fractalkine/CX3CL1 production by human airway smooth musclecells: induction by IFN-gamma and TNF-alpha andregulation by TGF-beta and corticosteroids. Am J Phy-siol Lung Cell Mol Physiol 2004; 287: L1230–123439. Chen S, Mukherjee S, Chakraborty C, et al. Highglucose-induced, endothelin-dependent fibronectinsynthesis is mediated via NF-kappa B and AP-1. Am JPhysiol Cell Physiol 2003; 284: C263–722.40. Ramana KV, Friedrich B, Srivastava S, et al. Acti-vation of nuclear factor-kappaB by hyperglycemia invascular smooth muscle cells is regulated by aldose re-ductase. Diabetes 2004; 53: 2910–2920.41. Karin M. The regulation of AP-1 activity by mi-togen-activated protein kinases. J Biol Chem 1995;270: 16483–16486.42. Kyriakis JM. Activation of the AP-1 transcriptionfactor by inflammatory cytokines of the TNF family.Gene Expr 1999; 7: 217–231.43. Saatian B, ZhaoY, He D, et al. Transcriptional regu-lation of lysophosphatidic acid-induced interleukin-8expression and secretion by p38 MAPK and JNK inhuman bronchial epithelial cells. Biochem J 2006; 393:657–668.44. Zambon A, Gervois P, Pauletto P, et al. Modulationof hepatic inflammatory risk markers of cardiovasculardiseases by PPAR-alpha activators: clinical and experi-mental evidence. Arterioscler Thromb Vasc Biol 2006;26: 977–986.45. Delerive P, De Bosscher K, Besnard S, et al. Perox-isome proliferator-activated receptor alpha negativelyregulates the vascular inflammatory gene response bynegative cross-talk with transcription factors NF-kap-paB and AP-1. The Journal of biological chemistry1999; 274: 32048–32054.46. Scott R, Best J, Forder P, et al. Fenofibrate Interven-tion and Event Lowering in Diabetes (FIELD) study:baseline characteristics and short-term effects of fe-nofibrate [ISRCTN64783481]. Cardiovasc Diabetol2005; 4: 13.

Dragomir et al. Chemokine upregulation by high glucose in SMC

1165