Distinct Signaling Pathways for MCP-1-dependent Integrin Activation and Chemotaxis*

Transmigration of monocytes to the subendothelial space is the initial step of atherosclerotic plaque formation and inflammation. Integrin activation and chemotaxis are two important functions involved in monocyte transmigration. To delineate the signaling cascades leading to integrin activation and chemotaxis by monocyte chemoattractant protein-1 (MCP-1), we have inves-tigated the roles of MAPK and Rho GTPases in THP-1 cells, a monocytic cell line. MCP-1 stimulated b 1 integrin-dependent, but not b 2 integrin-dependent cell adhesion in a time-dependent manner. MCP-1-mediated cell adhesion was inhibited by a MEK inhibitor but not by a p38-MAPK inhibitor. In contrast, MCP-1-mediated chemotaxis was inhibited by the p38-MAPK inhibitor but not by the MEK inhibitor. The inhibitor of Rho GTPase, C3 exoenzyme, and a Rho kinase inhibitor abrogated MCP-1-dependent chemotaxis but not integrin-depend-ent cell adhesion. Further, C3 exoenzyme and the Rho kinase inhibitor blocked MCP-1-dependent p38-MAPK activation. These data indicate that ERK is responsible for integrin activation, that p38-MAPK and Rho are responsible for chemotaxis mediated by MCP-1, and that Rho and the Rho kinase are upstream of p38-MAPK in MCP-1-mediated signaling.

Several lines of evidence indicate that monocyte chemoattractant protein-1 (MCP-1) 1 is involved in the pathogenesis of atherosclerosis by promoting directed migration of inflammatory cells, such as monocytes and T lymphocytes (1,2). During the progression of atherosclerosis, there is an accumulation of low-density lipoprotein within macrophages present in the in-timal layer. Deposition of lipids within these cells leads to the formation and eventual enlargement of atherosclerotic lesions. Boring et al. (3) noted an overall decrease in atherosclerotic lesion size in mice deficient for the MCP-1 receptor, CCR2, when they are crossed with ApoE knockout mice. Gu et al. (4) also found decreased atherosclerotic lesions in MCP-1-deficient mice when they are crossed with the low-density lipoprotein receptor knockout mice. These studies have demonstrated that MCP-1 and CCR2 play a crucial role in the initiation of atherosclerosis by recruiting monocytes to the vessel wall.
According to the multistep theory, monocytes roll on the endothelial cells, interact with E-selectin, adhere to the endothelial cells by firm adhesion to ICAM-1 and VCAM-1, and then migrate into the subendothelium (5). Rolling of monocytes on endothelial cells is dependent on the binding of E-selectin and sialyl Lewis X, and adhesion to the endothelium is dependent on the interaction of integrins on monocytes and adhesion molecules on the endothelial cells, such as VCAM-1 and ICAM-1. Integrins consist of several subtypes, and each subtype is specific for its ligand. For example, ␣4␤1 integrin, very late antigen-4, binds to VCAM-1, and ␤2 integrins bind to ICAM-1. Fibronectin, one of the extracellular matrix proteins, is also known to bind to ␤1 integrins, mainly to ␣5␤1 integrin. In these serial events of the multistep theory, MCP-1 can play a key role in monocyte recruitment by both integrin activation and by promoting migration to the vessel wall. However, the signal transduction pathways leading to integrin activation and chemotaxis have not been fully elucidated.
We recently demonstrated that the ␤␥ subunit of heterotrimeric G protein, Gi, plays a key role in MCP-1-induced chemotaxis (6). In that study, we reported that activation of ERK was not involved in chemotaxis by MCP-1. MAPK family members, ERK, JNK, and p38-MAPK, have been implicated in events necessary for proliferation, differentiation, apoptosis, and certain kinds of stress responses (7). These MAPKs are activated by specific cascades responsible for certain stimuli and eventually induce a variety of cell responses. Recently, several groups have reported on the involvement of MAPK and Rho in chemotaxis (8,9). Most of the studies on signal transduction of chemotaxis and cell adhesion have been conducted in adherent cells. However, in adherent cells it would be difficult to separate chemotaxis and integrin activation, because these two functions are closely connected. Therefore, the aim of our study was to examine whether integrin activation and chemotaxis can be separated by studying the signaling cascades leading to integrin activation and chemotaxis mediated by MCP-1 in human monocytic THP-1 cells and to elucidate the role of MAPK and Rho in these biological functions. * This study was supported by grants-in-aid from the Ministry of Education, Science, Sports, and Culture of Japan; International Scientific Research Program grants from the Japanese Ministry of Education, Science, Sports, and Culture; Center of Excellence grants from the Japanese Ministry of Education, Science, Sports, and Culture; and a research grant for health sciences from the Japanese Ministry of Health and Welfare. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Cell Lines-The monocytic cell line THP-1 was a generous gift from Dr. K. Nishida (Daiichi Pharmaceuticals Co. Ltd., Tokyo, Japan) and was cultured in RPMI supplemented with L-glutamine and penicillin/ streptomycin plus 10% fetal calf serum in an atmosphere of 95% air and 5% CO 2 at 37°C.
Cell Adhesion Assay-Polystyrene 96-well flat-bottomed microtiter plates (Costar 3595, Corning Incorporated, Corning, NY) were coated with 25 l of soluble VCAM-1 (2.5 g/ml), soluble ICAM-1 (2.5 g/ml), soluble E-selectin (2.5 g/ml), or fibronectin (10 g/ml) for 1 h at room temperature. After incubation, wells were blocked by incubation with 225 l of 10 mg/ml heat-denatured BSA for 30 min at room temperature (10). Control wells were filled with 10 mg/ml heat-denatured BSA. 100 l of THP-1 cells suspended at a concentration of 10 6 /ml in 0.1% BSA-RPMI were incubated for the indicated times in a CO 2 incubator at 37°C in the presence or absence of MCP-1. After incubation nonadherent cells were removed by centrifugation (top side down) at 48 ϫ g for 5 min (11). Attached cells were fixed with 5% glutaraldehyde for 30 min at room temperature. Cells were washed three times with water, and 100 l of 0.1% crystal violet in 200 mM MES (pH 6.0) was added to each well and incubated at room temperature for 20 min. Excess dye was removed by washing with water three times, and the bound dye was solubilized with 100 l of 10% acetic acid (12,13). The absorbance of each well at 595 nm was then measured using a multiscan enzymelinked immunosolvent assay reader (SPECTRA classic; TECAN). Each sample was assayed in triplicate. The absorbance was linear to the cell number up to an OD of 1.9 (data not shown). For example, 0.05 of OD represents adhesion of about 2,000 cells, and 0.5 of OD represents adhesion of about 25,000 cells.
Chemotaxis Assay-The migration of THP-1 cells was determined using a modification of the method of Campbell et al. (14). Briefly, THP-1 cells were resuspended in 0.1% BSA-RPMI. After adjusting the cell density to 1 ϫ 10 6 cells/ml, 100,000 cells in 100 l were added to the top chamber of a 24-transwell apparatus (6.5-mm diameter, 5-m pore size, Costar #3421; Corning Incorporated, Corning, NY) and incubated for 2 h at 37°C in an atmosphere containing 5% CO 2 . Cells that passed through the membrane were collected from the lower well and counted in a FACScan (Becton-Dickinson, San Jose, CA).

MCP-1 Increased Adhesion of THP-1 Cells to VCAM-1 and
Fibronectin-To determine the regulation of integrin avidity by MCP-1, we studied adhesion of THP-1 cells to purified adhesion molecules. Cell adhesion to soluble E-selectin, soluble ICAM-1, and soluble VCAM-1 was determined in the presence or absence of 10 nM MCP-1 under static conditions. MCP-1 increased adhesion of THP-1 cells to VCAM-1 by more than 2-fold but not to E-selectin or ICAM-1 (Fig. 1), indicating increased avidity of the ␤1 integrin by MCP-1. In the absence of MCP-1, more THP-1 cells adhered to VCAM-1 than control. In time course experiments, we also examined cell adhesion to fibronectin as a ligand for ␣5␤1 integrin and found that MCP-1 increased cell adhesion to both VCAM-1 and fibronectin by more than 3-fold in a time-dependent manner (Fig. 2). MCP-1-dependent adhesion to VCAM-1 was also increased in a dose-dependent manner, reaching a plateau at 1 nM MCP-1 (Fig. 3). Adhesion to fibronectin was also increased by MCP-1 stimulation in a dosedependent manner (data not shown). To show that this MCP-1-mediated adhesion is dependent on ␣4␤1 and ␣5␤1 integrins, we preincubated the cells with anti-␣4 antibody and the RGDS peptide. Preincubation of THP-1 cells with anti-␣4 antibody inhibited MCP-1-dependent and -independent cell adhesion to VCAM-1 by about 80% but not with control IgG. Preincubation with the RGDS peptide, but not with the RGES peptide, inhibited MCP-1-dependent and -independent cell adhesion to fibronectin (Fig. 4). These data indicate that cell adhesion in our assay depends on the interaction between integrins and their ligands and that MCP-1 increased the avidity of both ␣4␤1 and ␣5␤1 integrins on THP-1 cells.
Inhibition of ERK but Not p38-MAPK Abrogated MCP-1induced Adhesion-To determine whether MAPK activation is involved in MCP-1-mediated integrin activation, we next pretreated the cells with MEK-or p38-MAPK-specific inhibitors and examined the effect of ERK and p38-MAPK on MCP-1-de-

FIG. 5. MEK inhibitor (PD98059) blocks MCP-1-dependent adhesion to VCAM-1 and fibronectin in a dose-dependent manner.
THP-1 cells were pre-incubated with PD98059 or SB203580 at the indicated concentrations for 1 h in an atmosphere of 95% air and 5% CO 2 at 37°C. Both inhibitors were dissolved in Me 2 SO, and its final concentration was 0.2% including control. After the incubation, cells were subjected to adhesion assays on VCAM-1 (A) or fibronectin (B) in the presence (open columns) or absence (closed columns) of 10 nM MCP-1 for 20 min. The adhesion assays were done as described under "Experimental Procedures." Data represent the mean Ϯ S.D. of triplicate measurements. Results are representative of five separate experiments. To make comparison easier, the level of adhesion obtained without MCP-1 without inhibitors was taken as 100%. pendent integrin activation. Pretreatment of THP-1 cells with the MEK inhibitor, PD98059, inhibited MCP-1-induced adhesion to VCAM-1 and fibronectin in a dose-dependent manner (Fig. 5). In contrast, pretreatment with p38-MAPK inhibitor, SB203580, did not affect MCP-1-induced cell adhesion, indicating the involvement of ERK but not p38-MAPK in MCP-1-dependent integrin activation.
Inhibition of p38-MAPK but Not ERK Abrogated MCP-1induced Chemotaxis-To examine the role of MAPK in MCP-1-mediated chemotaxis, we pretreated the cells with MAPK inhibitors before the chemotaxis assay. We found that in the presence of MCP-1, THP-1 cells showed a typical bell-shaped pattern of chemotactic responses, and the maximal response was achieved at 1 nM MCP-1 in the lower chamber (data not shown). In contrast to the results in the cell adhesion assays, pretreatment of THP-1 cells with SB203580 abrogated MCP-1-induced chemotaxis in a dose-dependent manner (Fig. 6). In contrast, pretreatment with PD98059 did not affect the chemotaxis, indicating the involvement of p38-MAPK but not of ERK in MCP-1-mediated chemotaxis. We have also tested the effect of anti-␣4 antibody and the RGDS peptide in the chemotaxis assay, but both of them had no effect on chemotaxis induced by MCP-1 (data not shown), indicating that integrin activation does not play a role in chemotaxis induced by MCP-1 in THP-1 cells.
Next, to examine the role of Rho GTPase and the Rho kinase, we pretreated the cells with C3 exoenzyme and a Rho kinase inhibitor, Y-27632, and performed cell adhesion and chemotaxis assays. Pretreatment of the cells with C3 exoenzyme and Y-27632 abrogated MCP-1-mediated chemotaxis but not cell adhesion to fibronectin (Fig. 7) or VCAM-1 (not shown), indicating that Rho is involved in chemotaxis but not in integrin activation. To examine whether Rho and a Rho kinase are upstream of p38-MAPK, the effect of C3 exoenzyme and Y-27632 on MCP-1-induced p38-MAPK activation was determined. We found that MCP-1 could phosphorylate ERK and p38-MAPK in THP-1 cells. However, pretreatment of the cells with C3 exoenzyme and Y-27632 abrogated MCP-1-induced phosphorylation of p38-MAPK but not of ERK (Fig. 8).

SB203580 or PD98059 Did Not Affect Calcium Flux by MCP-
1-To rule out the possibility of nonspecific inhibition of MCP-1-induced signaling by these inhibitors, we checked MCP-1-dependent calcium flux after pretreatment with SB203580 and PD98059. As shown in Fig. 9, MCP-1-induced calcium flux was not affected by this pretreatment. DISCUSSION In this study we have elucidated the role of MAPK and Rho GTPase in MCP-1-mediated cell adhesion and chemotaxis. We show that MCP-1 induced activation of the integrins ␣4␤1 and ␣5␤1 in THP-1 cells and that the integrin activation is dependent on ERK activation. In contrast, MCP-1-dependent chemotaxis was dependent on activation of Rho and p38-MAPK. Thus, as depicted in Fig. 10 two important biological functions mediated by MCP-1 utilize two distinct MAPK-dependent signaling pathways.
In this study we used sensitive cell adhesion assays to demonstrate important functions of MCP-1. Although Weber et al. (16) demonstrated that binding of monocytes to VCAM-1 was reduced at 15 min under stimulation with MCP-1 in a similar assay, we found greater than a 2-fold increase in cell adhesion to this molecule at 10 to 20 min. In preliminary experiments, we have found a basal increase in cell adhesion after labeling the cells with fluorescent dye and washing the cells. We speculate that this is because of some stress on the cells. Further, a basal increase in cell adhesion after spinning down and washing the cells might be because of increased MAPK activation during these procedures. In support of this hypothesis, MacKenna et al. (17) reported that in cardiac fibroblasts ERK and JNK are activated by mechanical stretch. Therefore, it would be important to avoid stress on the cells as much as possible in this experiment. Work is now in progress to determine the effect of mechanical stress on MAPK activation and cell adhesion.
We found that in the monocytic cell line ␤1 integrins but not ␤2 integrins are activated by MCP-1. However, Weber et al. (18) have reported that MCP-1 induces a prolonged increase in the binding of monocytes to ICAM-1 in a static adhesion assay. This difference may be because of a difference in the way to remove nonadherent cells. We removed nonadherent cells by centrifugation, but they did it by plate washer. So we speculate that binding of ␤2 integrin and ICAM-1 is not strong enough to overcome the centrifugation force. Further, Chan et al. (19) have reported that activation of ␤1 integrin by chemokines might be much stronger than that of ␤2 integrin and that ␤1 integrin/VCAM-1 interaction activates ␤2 integrin-mediated cell adhesion in human T cells. Therefore, in in vivo situations, activation of ␤1 integrins might be stronger and more important in the early phase of cell migration.
In the cell adhesion assay, we could not abrogate MCP-1-dependent adhesion to fibronectin by the RGDS peptide, even though we have used sufficient concentrations of the peptide. We speculate that the inhibitory effect of this peptide on the interaction between fibronectin and integrins is not so strong (20). Because RGDS-independent adhesion of fibronectin has been reported (21), it is also possible that RGDS-independent cell adhesion is induced by MCP-1.
In this study we showed that in THP-1 cells ERK is responsible for integrin activation by MCP-1 but not p38-MAPK or Rho. Laudanna et al. (22), however, reported that Rho is also involved in integrin activation by interleukin-8 in neutrophils and lymphocytes. The reasons for these differences are not clear, but signaling through integrin activation in response to chemokines might be cell type-specific. It is also possible that different types of integrins in leukocytes might be activated in response to each chemokine.
Recently, several reports have shown that p38-MAPK is involved in chemotaxis induced by serum, lysophosphatidylcholine, and chemokines in leukocytes and smooth muscle cells (8,9). Our study has also shown that p38-MAPK is involved in chemotaxis induced by MCP-1 in THP-1 cells. In contrast, Yen et al. (23) showed that ERK is responsible for MCP-1-mediated chemotaxis. Knall et al. (24), on the other hand, have shown that ERK or p38-MAPK is not involved in interleukin-8-mediated chemotaxis. The reason for these differences is not clear, but in the system of Yen et al. (23) the activation of integrin might have been required for monocyte chemotaxis. Rho family GTPases have also been shown to be involved in cell migration (25). Thus our data are consistent with the others that p38-MAPK and Rho are involved in chemotaxis. However, the relationship between Rho and MAPK is quite complicated. For example, Zhang et al. (26) have shown that p38-MAPK is downstream of Rho in interleukin-1-mediated signaling. In contrast, Hippenstiel et al. (27) have reported that LPS-induced activation of p38-MAPK is not affected by Clostridium difficile toxin B-10463, a specific inhibitor of Rho. In terms of ERK and Rho, some reports have claimed that Rho is upstream of ERK (28 -30), whereas others have demonstrated that Rho and ERK are activated independently (31). However, few studies have been conducted to examine whether ERK and p38-MAPK are differentially affected by Rho. In this paper, we clearly show that MCP-1 phosphorylates ERK and p38-MAPK in THP-1 cells and that Rho is upstream of p38-MAPK but not of ERK in MCP-1-mediated signal transduction. We have also shown that the Rho kinase (32) is between Rho and p38-MAPK and is a key molecule for chemotaxis. However, downstream targets of p38-MAPK leading to chemotaxis still remain to be determined. Although we found that MCP-1 also activated JNK (data not shown), the role of JNK activation in MCP-1-mediated signaling was not determined in this study.
In summary, we have provided clear evidence that two distinct signaling cascades are present to mediate MCP-1-induced activation of ␤1 integrins and chemotaxis in THP-1 cells. These two distinct signaling cascades would be important for transmigration of monocytes through endothelial cells. The most intriguing aspect of this study is that we could separate two important biological functions of leukocytes, integrin activation and chemotaxis, by different assays and found that two distinct signaling cascades mediate these two functions. In adherent cells, however, segregation of integrin activation and chemotaxis would be very difficult to assess. As depicted in Fig. 10, identification of signaling molecules located at the bifurcation to ERK and p38-MAPK would be important to delineate the signaling cascades through CCR2. Further, it would be intriguing to determine the signaling cascades in a condition closer to in vivo situations.