Activation of vascular endothelial cells (ECs) plays an important pathogenic role in the development of atherosclerosis. Monocyte chemoattractant protein-1 (MCP-1) is a potent chemoattractant of monocytes. Besides induction of monocyte recruitment, it has been suggested that MCP-1 can also affect the cellular responses of ECs. We investigated whether MCP-1 activated the three major mitogen activated protein (MAP)-kinases extracellular signal-regulated kinase (ERK), c-Jun amino terminal kinase (JNK) and p38 MAPK. Stimulation of ECs with MCP-1 induced a time- and concentration-dependent activation of all three MAP-kinases, concentrations as low as 0.1 ng/ml were sufficient for this mechanism. MCP-1 also induced secretion of matrix metalloproteinase (MMP)-2 which along with ERK activation was inhibited by PD098059. The results demonstrate that MCP-1 can lead to direct activation of MAP kinases together with induction of MMP2 in ECs. Our data thus propose a new mechanism for the proatherogenic effect of MCP-1.
Time for primary review 29 days.
Monocyte chemoattractant protein-1 (MCP-1) belongs to the CC chemokine family which mediates leukocyte trafficking at sites of inflammation in a variety of chronic inflammatory diseases . MCP-1 also plays an important role in lesion formation during atherogenesis . Besides its potent chemoattracting property, MCP-1 can increase cytotoxic lymphocyte and natural killer cell activity in vitro , affect the phenotype of vascular smooth muscle cells  and activate cytokine secretion in monocytes .
The biological effects of chemokines are mediated by heterotrimeric G-protein coupled seven transmembrane spanning receptors such as the chemokine receptor 2 (CCR2) [6,7]. Even though MCP-1 can bind to different chemokine receptors  in vivo studies with CCR2 −/− mice indicate that the proatherogenic effects of MCP-1 are primarily mediated through binding to CCR2 . The observation of a decrease in atherosclerotic lesion size in CCR2 −/− deficient mice supports the in vivo relevance of MCP-1 in the initiation of atherosclerosis .
So far, it has been assumed that chemokine receptors are mainly expressed on leukocytic cells. Recent studies demonstrated expression of CC receptors in vascular smooth muscle cells  and endothelial cells . CC receptor expression was associated with the induction of endothelial migration in an experimental model of wound repair by MCP-1 . Despite considerable evidence that MCP-1 can affect endothelial cell function, signal transduction pathways initiated by MCP-1 remain elusive.
A wide variety of different inflammatory stimuli exert their cellular effects through phosphorylation of a family of mitogen activated protein (MAP)-kinases . The three major subfamilies of MAP-kinases are the extracellular signal-regulated kinases (ERKs), the c-Jun amino terminal kinases (JNKs), and p38 MAP kinases. Once they are activated by specific mediators such as cytokines, MAP-kinases can affect numerous downstream mediators including transcription factors like AP-1  leading to the activation of specific cellular genes being involved in a number of proatherogenic mechanism. The aim of the present study was therefore to determine the influence of MCP-1 on MAP-kinase activation in human endothelial cells (ECs).
MCP-1 was purchased from R&D Systems (Minneapolis, MN, USA). Pertussis toxin (PTX), cholera toxin (CTX) and PD098059 were obtained from Calbiochem (La Jolla, CA, USA). Anti-von Willebrand factor was from Dako (Hamburg, Germany). Anti-phospho-ERK 1/2, anti-ERK 1/2, c-Jun-gluthadione-S-transferase (GST) fusion protein, anti-phospho-c-Jun, anti-ATF-2, and the p38 MAPK assay kit were obtained from New England Biolabs. (Schwalbach, Germany). Anti-phospho JNK and anti-phospho p38 MAPK were from Cell Signaling Technology (Bevery, MA, USA), anti-α Tubulin was obtained from Sigma-Aldrich (St. Louis, MO, USA). Anti-matrix metalloproteinase (MMP)-1, -2, -3 and -9 were purchased from Calbiochem (La Jolla, CA, USA). Horseradish-peroxidase-conjugated goat anti rabbit immunoglobulin was from Dianova (Hamburg, Germany). Enhanced chemiluminescence detection reagents and Hyperfilm were from Amersham Pharmacia Biotech (Little Chalfont, Buckinghamshire, UK). Cell culture media were purchased from Cell Systems (St. Katharinen, Germany), cell culture supplements were from Gibco BRL (Karlruhe, Germany); the mycoplasma- and acheloplasma-detection kit was from Boehringer Mannheim (Mannheim, Germany).
Human umbilical venous endothelial cells (HUVECs) were isolated from umbilical cord as described . Cells were cultured in endothelial basal medium (EBM) supplemented with 10% FCS, 3 μg/ml bovine brain extract, 1 μg/ml hydrocortison, 10 μg/ml human epidermal growth factor and 50 μg/ml amphotericin B/gentamicin. For all experiments HUVEC culture passages 2–5 were used at 80–90% confluence. Culture purity was confirmed by the typical cobblestone morphology and positive immunochemical staining (≥ 95%) for von Willebrand factor. Prior to use representative endothelial cell cultures were determined to be mycoplasma- and acheloplasma-free.
For the experiments HUVECs were cultivated in the absence of FCS for 12 h and were exposed to different concentrations of MCP-1 for 0–120 min. In some experiments, cells were pretreated with inhibitors of intracellular signal transduction molecules for different time periods prior to MCP-1 addition. Immediately before addition of MCP-1 the medium was replaced with fresh EBM containing 0.1% bovine serum albumin and antibiotics.
HUVECs were washed with ice-cold PBS buffer and cells were lysed as described . Following quantification of protein concentration by a bicinchoninic acid protein assay kit from Pierce, 10 μg of cell extracts were subjected to 10% SDS-PAGE and subsequently immunoblotted for phospho-ERK, phospho-JNK and phospho-p38 MAPK determination.
Secreted matrix MMPs were detected in EC supernatant derived from 6-cm tissue culture dishes. Equal amounts of collected supernatant were subjected to 10% SDS-PAGE and immunoblotted for MMP-1, -2, -3 and -9. Horseradish-peroxidase-conjugated secondary antibody and chemiluminescence were used for detection.
Measurement of JNK activity
In some experiments the activity of JNK was measured by solid-phase kinase assay as described . In brief, cell-lysates were incubated with GST-c-Jun (1–79) fusion protein bound to Glutathione–Sepharose beads at 4 °C overnight. Complexes were recovered by centrifugation, washed and resuspended with kinase lysis buffer containing 100 μM ATP. Reactions were terminated by the addition of Laemmli sample buffer and boiling at 100 °C for 5 min. Samples were separated by SDS–PAGE and transferred to nitrocellulose membranes. To detect phosphorylated c-Jun, the membranes were incubated with an anti-phospho-c-Jun antibody following ECL detection.
Measurement of p38 MAPK activity
The activity of p38 MAPK was assayed as described . Cell lysates were incubated overnight with anti-p38 MAPK. The immunocomplexes were precipitated with protein G-Sepharose and recovered by centrifugation, washed and incubated with kinase buffer. P38 MAPK activity in immunoprecipitates was measured using the p38 MAPK assay kit according to the manufacturer’s protocol. Phosphorylated ATF-2 was detected using a polyclonal phospho-ATF-2 specific antibody.
Activation patterns of MAP-kinases were similar in kinase assays and immunoblots using antibodies directed against phosphorylated MAP-kinases.
RNA isolation and RT-PCR
Total RNA from HUVECs and vascular smooth muscle cells (VSMCs) was extracted. Following DNAse treatment to eliminate possible residual DNA, 1 μg RNA was converted into cDNA by standard techniques as described . Amplification was performed at 94 °C (3 min) for denaturation, followed by a 32 cycles set of 1 min at 94 °C, 1 min at 64 °C and 2 min at 72 °C with a final extension at 72 °C for 3 min. Primers were synthesised according to published sequences  resulting in a 255 bp fragment (primer pair: sense 5′CCAACTCCTGCCTCCGCTCTA3′, antisense 5′CCGCCAAAATAACCGATGTGATAC3′) PCR products were analysed by gel electrophoresis and sequenced by MWG (Ebersberg, Germany).
Determination of MCP-1 release
Cells were grown in 6-well plates to confluency and kept in serum-free medium for 12 h prior to stimulation. Following MCP-1 exposure (10 ng/ml) for 15 min fresh medium was added. Quantification of MCP-1 secretion in cell culture supernatants was performed by ELISA (Quantikine Immunoassay, R&D Systems, Minneapolis, MN, USA) according to the manufactor’s instructions.
Data are presented as mean±standard deviation as indicated. Multiple comparisons were evaluated with ANOVA, followed by Fisher’s protected least significant difference model. Values of P<0.05 were considered statistically significant.
Expression of CCR2 mRNA in ECs
To determine the expression of MCP-1 receptor CCR2 in human ECs, RT PCR was performed to amplify RNA isolated from ECs and arterial (a) and venous (b) smooth muscle cells (SMCs) as positive controls using specific primers for CCR2. PCR products were analysed by agarose gel electrophoresis. As shown in Fig. 1 a band migrating at an expected molecular weight of 255 bp was observed in both cell types. To assure correct amplification PCR products were finally sequenced (data not shown).
Activation of MAP-kinases in ECs
It has been shown that MAP-kinase activation can mediate cellular responses of G-protein coupled seven transmembrane spanning receptors. Therefore we investigated whether MCP-1 affects MAP-kinase activity in ECs. MCP-1 stimulation induced a rapid and transient 3.3-fold increase of ERK activation, which peaked at 10 min and returned to baseline levels within 1 h (Fig. 2A). An involvement of JNK and p38 MAPK in MCP-1-mediated endothelial activation was assessed by kinase assays using c-Jun and ATF-2 as substrates. As shown in Fig. 2B MCP-1 also induced JNK activation whose kinetic profile was comparable with the time course observed for ERKs. In contrast to ERK and JNK, p38 MAPK activity was biphasic with a rapid initial increase after 5 min which, after return to baseline, was followed by a second activation 1 h after MCP-1 exposure (Fig. 2C).
To assess the sensitivity of MAP-kinase activation different MCP-1 concentrations were used. Whereas concentrations as low as 0.1 ng/ml were capable to stimulate a significant increase in JNK (Fig. 3B) and p38 MAPK (Fig. 3C) activity, ERK activation needed 10-fold higher concentrations (Fig. 3A).
Mechanisms of MCP-1-induced activation of ERK1/2
For ERK activation by MCP-1 the pathways involved were studied. PD098059, an inhibitor of MEK, completely blocked MCP-1-induced ERK activity, indicating that ERK phosphorylation was indeed specifically mediated by its immediate ‘upstream’ kinase (Fig. 4A). Since previous studies suggested a central role for small G-proteins in MCP-1-mediated signal transduction in peripheral blood cells , we finally determined whether G-proteins were necessary to mediate MCP-1-induced ERK activation in ECs. HUVECs were preincubated with either pertussis toxin (PTX), an inhibitor of Gi-proteins or cholera toxin (CTX), an activator of Gs protein before exposure to MCP-1. As demonstrated in Fig. 4B, MCP-1-dependent signal transduction pathways required Gi protein activity to stimulate ERK activity, whereas activation of Gs-proteins did not influence MCP-1-mediated activation of ERK, but already activated basal ERK activity.
Regulation of MCP-1 secretion by MCP-1 in ECs
In order to determine the capacity of MCP-1 to act as a trigger as well as a secretory product in this experimental setting we investigated the effect of this chemokine on MCP-1 secretion in ECs. As shown in Table 1 MCP-1 stimulation showed a trend to increase the secretion of MCP-1 in ECs, however, the differences versus control were not significant.
|MCP-1 secretion (pg/ml) (mean±S.D.)|
|Controls||MCP-1 stimulated ECs|
|MCP-1 secretion (pg/ml) (mean±S.D.)|
|Controls||MCP-1 stimulated ECs|
MCP-1 induces secretion of MMP-2
To examine functional consequences of MCP-1 stimulation in ECs we studied secretion of matrix metalloproteinases (MMPs). HUVECs were incubated with MCP-1 (10 ng/ml) for indicated time periods, supernatant was collected and finally analysed for the content of MMP-1, -2, -3 and -9.
Stimulation of ECs with MCP-1 (10 ng/ml) resulted in a time-dependent, about 7-fold increase of MMP-2 secretion peaking at 24 h (Fig. 5A). MMP 1, 3 and 9 were also detected in the supernatant of EC, yet MCP-1 failed to further enhance secretion of those MMPs (data not shown).
Mechanism of MCP-1-induced MMP-2 secretion
Considering that induced MMP expression was shown to dependent on sustained ERK phosphorylation in the tumor cell line SCC-12F , we investigated the role of ERK phosphorylation in endothelial MMP-2 expression induced by MCP-1. ECs were preincubated with the MEK inhibitor PD098059 (30 μmol/l, 30 min) before addition of MCP-1 (10 ng/ml, 24 h). Pretreatment with PD098059 significantly reduced the MMP-2 release suggesting that ERK phosphorylation was necessary for MMP-2 expression (Fig. 5B).
It was shown here that MCP-1 activates MAP-kinases in ECs. In addition, MCP-1 potently induced MMP-2, proteins which have been implicated in lesion progression and plaque destabilization.
In accordance with previously published data we also found that ECs express the CCR2 gene providing the prerequisite for ligand receptor interaction. The identification of CCR2 in ECs and vascular smooth muscle indicates that MCP-1 not only participates in monocyte recruitment and trafficking but may also have functional implications for other vascular cells during atherosclerosis [4,11].
Examining MCP-1-dependent signal transduction in leukocytes, it has been demonstrated that MCP-1 can affect tyrosine kinases, induce changes of intracellular calcium  or cAMP levels  and activate MAP-kinases [11,20]. However, due to the availability of specific effector molecules chemokine-mediated cellular effects can vary depending on the cell type in which the receptor is expressed in. These differential activation patterns might be due to the availability of specific G-protein subunits and other downstream mediators. Besides its activating properties on leukocytes, it was still unknown if MCP-1 is able to modulate the activation of specific MAP-kinase cascades in other cells lines. Since ECs play a pivotal role in the development and progression of atherosclerosis, we therefore focused on MCP-1-induced activation profiles of MAP-kinases in EC. Our findings show that MCP-1 stimulation induces a potent activation of ERK, JNK and p38 MAPK in vascular ECs. The time course observed for ERK, JNK and p38 MAPK was similar to what has been described in leukocytes . In contrast to ERK and JNK the kinetic profile of p38 MAPK showed a biphasical pattern of activation which has also been reported for this MAP-kinase after stimulation with cytokines or vasoactive agents [21,22]. Our results indicate that MCP-1-CCR2 interactions may be able to trigger different regulatory signal transduction mechanism. Remarkably, already low concentrations of MCP-1 were effective in this signal transduction pathway indicating the pathophysiological significance of our present findings. Extended to an in vivo situation it can therefore be hypothesized that small amounts of MCP-1 secreted by monocytes after first contact with the vascular wall may lead to a change in the activation state of the luminal EC layer. Since recent studies of Parissis et al. found similar CC chemokine concentrations in patient suffering an acute myocardial infarction, in vitro activation of EC MAP-kinases by low concentrations of MCP-1 may therefore reflect the in vivo relevance of our results .
CC receptors couple via heterotrimeric G-proteins to exert a variety of cellular responses . It has been reported that MCP-1 signaling involves Go/Gi proteins for the regulation of migration , cellular calcium flux  or the activation of phospholipase C  in monocytes. Therefore we were interested to investigate the possible involvement of pertussis toxin sensitive G-proteins to evaluate the coupling mechanism in ECs. Similar to recent findings in human monocytes , pertussis toxin blocked the MCP-1-induced EKR activation in vascular ECs, suggesting a Go/Gi dependent coupling mechanism. Moreover the observation that PD098059, an inhibitor of MEK, completely inhibited MCP-1-induced ERK activity implied that ERK was indeed downstream of MEK in these human ECs.
Previous studies mainly focused on the induction of MCP-1 in ECs by inflammatory agents [1,26,27]. Data presented here now suggest that MCP-1 itself can function as an activating agent for ECs. Monocyte adherence to endothelium and infiltration of the vessel wall are probably the first steps leading to the development of the fatty streaks, which depend on endothelial activation and expression of certain proteins such as adhesion molecules. It is conceivable that MCP-1 secreted by monocytes leads to this activation of ECs through recruitment of MAP-kinase pathways as evidenced in the present paper.
Since activation of vascular ECs is known to be associated with the induction of specific genes we were interested in the functional consequences of MCP-1 stimulation for EC. ERK dependent signaling was shown to participate in MMP expression in different cell types like keratinocytes, UM-SCC-1 cells  or human ECs [17,29–31]. Even though it has been demonstrated that ECs can secrete different MMPs [32–34], little is known about the regulation of endothelial MMP release. Recently, MMP-1 gene expression and synthesis was proposed to be stimulated by MCP-1 in human fibroblasts . The present results demonstrate that MCP-1 not only activates ERK but also induces secretion of MMP-2 in human ECs, thereby further enhancing its atherogenic properties. As evidenced by inhibition of both ERK phosphorylation and MMP-2 secretion through PD098059, ERK activation seemed to be directly involved in secretion of MMP-2, possibly through activation of transcription factor AP-1 as described in other cell systems . MMPs have been proposed to play a dominant role in atherogenesis promoting destabilization and rupture of plaque in all stages of the lesion. Hence, MCP-1 may promote destabilization and rupture of plaques by induction of matrix-degrading enzymes, ultimately leading to thrombosis and complete obstruction of the vessel . Activation of AP-1 by MCP-1 may not only affect MMPs but also other aspects of cellular inflammation such as cytokine secretion. To test if MCP-1 may also affect its own secretion, we tested the presence of an autocrine loop, however, could not detect increased release of MCP-1 from ECs after stimulation with MCP-1.
In conclusion, our results demonstrate that apart from its chemotactic properties for monocytes, MCP-1 promotes activation of EC through phosphorylation of MAP-kinases. As MCP-1 is already present in very early lesions, it may be a key regulator for early EC dysfunction in the course of atherogenesis. This could be one of the underlying mechanisms by which local MCP-1 release through monocytes can foster atherogenesis.
The authors thank their colleagues from the Department of gynaecology, University of Heidelberg, for supplying human umbilical cords. We also thank Jianwei Fei for her excellent technical assistance.