High Glucose-treated Macrophages Augment E-Selectin Expression in Endothelial Cells*

E-selectin expression by endothelial cells (ECs) is crucial for leukocyte recruitment during the inflammatory response. Macrophage accumulation and serum E-selectin elevation are features of type 2 diabetes mellitus. However, the interactions between macrophages and ECs in regulating vascular endothelial function are not clearly understood. We investigated the mechanisms underlying the modulation of EC E-selectin expression by high glucose (HG)-treated macrophages. Macrophage-conditioned media (MCM) were prepared from HG-treated macrophages. EC stimulation with HG-MCM induced increases the expression and secretion of E-selectin. By using specific inhibitors and small interfering RNAs, we demonstrate that the activation of the JNK and p38 MAPK pathways are critical for HG-MCM-induced E-selectin expression. Transcription factor ELISA and chromatin immunoprecipitation assays further showed that HG-MCM increases the NF-κB- and AP-1 DNA-binding activities in ECs. The inhibition of NF-κB and AP-1 activation by specific siRNAs blocks the HG-MCM-induced E-selectin promoter activity and expression. Protein arrays and blocking assays using neutralizing antibodies demonstrated that macrophage inflammatory protein 1α and 1β in HG-MCM are major mediators for the induction of EC E-selectin expression. These data support the hypothesis that E-selectin up-regulation stimulated by macrophages may play an active role in atherogenesis in the HG condition and suggest a new mechanism by which arterial disease is accelerated in diabetes.

Among the cardiovascular risk factors documented in diabetes, hyperglycemia appears to be an independent risk factor for diabetic vascular complications (3,4). The early recruitment of monocytes to the arterial intima and their subsequent differentiation into macrophages, which is a critical sequence of events in the development and progression of atherosclerotic lesions, involve the interaction of specific adhesion molecules on vascular endothelial cells (ECs) with their counterreceptors on monocytes (5,6). A previous study has demonstrated that high glucose (HG) treatments of human monocytes lead to the increased expression of inflammatory cytokine and chemokine genes (7). However, the effects of HG versus normal glucose (NG) conditions on the release of inflammatory mediators after the transformation of monocytes into macrophages have not yet been clearly evaluated.
There is increasing evidence that the production and secretion of proinflammatory factors in vascular cells play an important role in atherogenesis (8 -10). E-selectin is a major EC adhesion molecule that regulates the binding and extravasation of leukocytes from the bloodstream to sites of inflammation. When ECs are activated in response to cytokines, the expression of cell adhesion molecules on their surface is increased markedly (11). The appearance of soluble cell adhesion molecules (i.e. intercellular adhesion molecule 1 (ICAM-1), vascular adhesion molecule 1 (VCAM-1), and E-selectin) in the circulation is thought to be the consequence of their release from the surface of activated ECs because of increased expression (12). Several reports have demonstrated the importance of soluble (s) E-selectin in the microvascular and macrovascular complications that can arise in patients with type 2 diabetes (13,14). High serum concentrations of sE-selectin have also been reported in type 2 diabetic patients (15). The expanding body of evidence in this regard has continued to focus on the role of macrophages in the development of atherosclerotic lesions. However, the contribution of macrophages under HG conditions to the process of atherosclerosis remains unclear.
Because high serum concentration of sE-selectin is related to hyperglycemic conditions, it was hypothesized that macrophages differentiated from monocytes in HG condition may alter their gene expression and that the soluble mediators released from HG-treated macrophages may up-regulate EC E-selectin expression. To gain insights into the mechanisms by which factors released by macrophages after HG treatment may up-regulate EC E-selectin expression, macrophage-conditioned medium (MCM) from patients or from HG and NG treatments were subjected to cytokine protein array analysis to determine the proinflammatory factors produced by macrophages after differentiation from monocytes under these conditions. We found that the chemokines macrophage inflammatory protein (MIP)-1␣ and MIP-1␤ produced by HG-treated macrophages exert paracrine effects on ECs to enhance the E-selectin expression and secretion. The E-selectin up-regulation induced by MIP-1␣ and MIP-1␤ released from HG-treated macrophages is mediated through the intracellular signaling cascades JNK and p38 MAPK, and the transcription factors NF-B and activated protein 1 (AP-1). Therefore, our current findings provide a molecular basis for the mechanisms by which HG-treated macrophages enhance E-selectin expression and secretion in ECs.

EXPERIMENTAL PROCEDURES
Materials-All culture materials were purchased from Invitrogen. PD98059 (ERK inhibitor), SP600125 (JNK inhibitor), and SB203580 (p38 inhibitor) were purchased from Calbiochem. Mouse mAB against JNK1 and phospho-JNK were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit polyclonal antibodies against p38 and mouse monoclonal phospho-p38 antibody were purchased from Cell Signaling Technology (Beverly, MA). sE-selectin ELISA kits and mAB against E-selectin were obtained from R&D Systems (Minneapolis, MN). The ERK siRNA, JNK siRNA, p38 siRNA, NF-B p65 siRNA, c-Jun siRNA, and control siRNA (scrambled negative control containing random DNA sequences) were purchased from Invitrogen. All other chemicals of reagent grade were obtained from Sigma.
Human Monocyte Isolation-Human monocytes from the buffy coat were isolated as described previously (9). Peripheral blood mononuclear cells were isolated by Histopaque 1077 density-gradient centrifugation. Monocytes were purified from peripheral blood mononuclear cells by negative selection using the magnetic-activated cell sorting monocyte isolation kit (Miltenyi Biotech, Auburn, CA).
Preparation of Human MCM-Monocytes were cultured in fresh RPMI 1640 medium containing 5.5 (NG) or 25 (HG) mmol/liter glucose or 5.5 mmol/liter glucose plus 19.5 mmol/ liter mannitol (M) and supplemented with 10% FBS. After 4 days in culture (the media were collected and defined as NG-4D or HG-4D medium), the macrophages were incubated for a further 48 h in fresh serum-free RPMI medium. The conditioned media were then collected and defined as NG-MCM, M-MCM, or HG-MCM (Fig. 1). The cell viability was quantified by using an 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide assay (16). As shown in supplemental Fig. S1, the viability of macrophages was over 75.6% when cultured in HG-MCM. In addition, the cell viability was similar when macrophages were cultured in NG-, M-, and HG-MCM (supplemental Fig. S1). The final concentrations of glucose in MCM are summarized in supplemental Table S1.
EC Culture-ECs were isolated from human umbilical cords as described previously (10). Briefly, human umbilical cords were washed, filled with 0.1% collagenase, and incubated at 37°C for 10 min. ECs were resuspended in M199 supplemented with penicillin/streptomycin and 20% FBS. Cultures were maintained at 37°C in a humidified atmosphere containing 5% CO 2 . Cells were kept static, and the culture medium was changed every 3 days. HUVECs of passage 2 were used for the experiments. To investigate the effect of MCM on ECs, the M199 media was removed, and fresh RPMI medium (control), NG-MCM, and HG-MCM containing 2% FBS were added according to different treatments of cells.
ELISA for Cell Surface E-Selectin Expression-E-selectin expression on the EC surface was measured by cell surface ELISA as described previously (18). Briefly, HUVECs cultured in 96-well plates were fixed by 4% paraformaldehyde. Cell surface E-selectin expression was assessed by using the mouse anti-human E-selectin mAb followed by a horseradish-peroxidase-conjugated secondary antibody. The absorbance of each well was measured at 490 nm after the reactions were stopped.
ELISA for sE-Selectin-The level of sE-selectin in EC culture medium was determined by using sandwich ELISA (sensitivity 18 pg/ml; R&D Systems) according to the manufacturer's protocols, as described previously (9).
For siRNA transfection, HUVECs were transfected with the designated siRNA by using a RNAiMAX transfection kit (Invitrogen).
NF-B p65 and AP-1 Transcription Factor Assays (TF ELISA)-Nuclear extracts of cells were prepared as described previously (10). Equal amounts of nuclear extracts were used for quantitative measurements of NF-B p65 and AP-1 activation using commercially available ELISA kits (Panomics, Redwood City, CA) that measure NF-B p65 and AP-1 DNA binding activities.
ChIP-After cross-linking with 1% formaldehyde, the cells were centrifuged and then resuspended in a lysis buffer for sonication three times at 15 s each. Supernatants were recovered by centrifugation. Aliquots of the precleared sheared chromatin were then immunoprecipitated using 2 g of antibodies against IgG, NF-B p65, or c-Jun. The resulting DNA was used for PCR analysis, and the amplified DNA fragments were visualized on an agarose gel. PCR was performed with primers that amplify the part of the human E-selectin promoters that contain the NF-B binding sites (5Ј-GGGAA AGTTT TTGGA TGCCA TT-3Ј and 5Ј-TGTCC ACATC CAGTA AAGAG GAAAT-3Ј) and AP-1 binding sites (5Ј-ATCTA CCTTG TGAGT CATTC-3Ј and 5Ј-TAGTT GTGGT AGTAA TTAGA AT-3Ј).
Monocytic Cell Adherence Experiment-The human monocytic cell line THP-1 was obtained from the American Type Culture Collection and maintained in RPMI 1640 culture medium supplemented with 10% FBS. Before the adhesion experiments, THP-1 cells were suspended in RPMI 1640 containing 0.1% FBS and labeled with 1,10-dioctadecyl-3,3,30,30tetramethylindocarbocyanine (Molecular Probes, Eugene, OR) for 20 min. The labeled THP-1 cells (1 ϫ 10 5 cells/ml) were added to NG-MCM-or HG-MCM-treated ECs and incubated for 30 min. In parallel experiments, ECs were treated with isotype-matched IgG or neutralizing antibody against E-selectin (20 g/ml), or HG-MCM were treated with neutralizing antibodies against MIP-1␣ or MIP-1␤ during HG-MCM stimulation. Non-adherent cells were removed by washing with RPMI. The adherent THP-1 cells on the EC surface were identified and counted in 10 randomly selected microscopic fields (1.37 ϫ 1.07 mm) under a Nikon Ti-E inverted epifluorescence microscope with 10ϫ objective, and the adhesion was expressed as fold compared with controls.
Statistical Analysis-The results are expressed as mean Ϯ S.E. Statistical analysis was determined by using an independent Student t test for two groups of data and analysis of variance followed by Scheffe's test for multiple comparisons. p values less than 0.05 were considered significant.

HG-MCM Enhanced E-Selectin Expression in ECs-
The effects of macrophages from a high-glucose concentration on the expression and secretion of E-selectin in ECs were studied using HG-MCM (versus NG-and M-MCM). The time courses determined for the E-selectin mRNA levels ( Fig. 2A) revealed an increase after 1 h of HG-MCM stimulation and a peak expression at 4 h, followed by a gradual reduction thereafter. At the 4-h time point, the culturing of the ECs in HG-MCM induced increases in the E-selectin mRNA levels by 5.3-fold compared with NG-MCM-treated ECs ( Fig. 2A). The exposure of ECs to HG-MCM also caused significant increases in the surface E-selectin expression at 6 h after treatment (Fig. 2B) and the sE-selectin secretion at 12 h (C). Fig. 2D shows the dosedependent induction of E-selectin mRNA expression by HG-MCM. In addition, HG-MCM also induced slight increases in the EC ICAM-1 and VCAM-1 transcript levels (by 1.5-and 1.43-fold, respectively) compared with NG-MCM-treated ECs (Fig. 3A). NG-4D medium, HG-4D medium, and NG-MCM had marginal effects only on EC E-selectin expression, whereas HG-MCM augmented this expression significantly in ECs (Fig.  3B). The effect of glucose and insulin on EC E-selectin expression is shown in Fig. 3C. The increases in E-selectin mRNA expression were 2.12-fold in HUVECs stimulated with 25 mM glucose for 4 h. HUVECs stimulated with 50 microunits insulin had no effect on E-selectin expression.

HG-MCM-induced E-Selectin Expression in ECs Is Mediated by the JNK and p38
Pathways-Members of the MAPK superfamily (i.e. ERK, JNK, and p38) are known to regulate gene expression and cellular functions (19). To determine whether HG-MCM-induced E-selectin expression is mediated through the MAPK-dependent pathways, ECs were incubated with specific inhibitors for ERK (PD98059, 30 M), JNK (SP600125, 20 M), or p38 (SB203580, 10 M) for 1 h before and during stimulation with HG-MCM. The HG-MCM-induced mRNA expression (Fig. 4A), EC surface protein expression (B), and sE-selectin secretion (C) were found to be significantly inhibited by SP600125 and SB203580 but not by PD98059. Treatment of ECs with SP600125 and SB203580 resulted in the additive inhibition of HG-MCM-induced E-selectin expression and secretion (Fig. 4, A-C). To further confirm the involvement of JNK and p38, but not ERK, in the modulation of E-selectin expression by HG-MCM stimulation, we examined the effects of expressing specific siRNAs that target these signaling pathways upon HG-MCM-induced E-selectin expression in ECs. HG-MCM-induced E-selectin mRNA (Fig. 4A) and protein expression (B) and sE-selectin secretion (C) were inhibited by transfection with JNK-and p38-specific siRNAs but not by transfection with ERK-specific or control siRNAs (100 mol/ml for each). The effectiveness of the silencing was vali-dated because ERK siRNA, JNK siRNA, and p38 siRNA (compared with control siRNA) caused an 85% reduction in ERK, JNK, and p38 protein expressions, respectively (supplemental Fig. S2, 100 mol/ml for each).
The phosphorylation of JNK and p38 in ECs increased rapidly after HG-MCM stimulation, reaching maximal levels at 10 min (Fig. 4D). After such transient increases, the levels of phosphorylation decreased to nearly basal levels by 1 h. In addition, HG-MCM significantly caused JNK and p38 phosphorylation after 10 min of treatment, whereas NG-MCM had a minor effect on JNK and p38 phosphorylation in HUVECs (Fig. 4E).

NF-B and AP-1 Binding Sites Are Essential for the HG-MCM Induction of Human E-Selectin Promoter Activity-
To identify the cis-acting elements in the E-selectin gene promoter that mediate HG-MCM-induced E-selectin transcription, luciferase assays were conducted with the p540-Luc plasmid and several deletion promoter constructs (Fig. 5A). In HUVECs, the Ϫ540/ϩ1 region of E-selectin was found to direct maximum luciferase activity. A sequence deletion from Ϫ310 to Ϫ130 caused this reporter activity to decrease to 40 -45% of control levels, whereas a further deletion to the Ϫ40 position almost completely abolished the activity (Fig. 5A).
To investigate whether NF-B and AP-1 binds the E-selectin promoter region in HUVECs, we performed quantitative analysis of the NF-B p65 and AP-1 binding activities in vitro using TF ELISA kits from Panomics. The treatment of ECs with HG-MCM caused both NF-B p65-and AP-1 DNA binding activities to increase after 1 h and remain elevated for at least 2 h (Fig.  5B). These results were confirmed by ChIP analysis. Immunoprecipitated chromosomal DNA with a p65 antibody was sub- jected to PCR using primers designed to amplify the E-selectin promoter region harboring the NF-B binding site. NF-B was indeed found to bind the E-selectin promoter region containing the NF-B sites (Fig. 5C, left panel). Similarly, the region containing the AP-1 sites was specifically immunoprecipitated with c-Jun antibodies (Fig. 5C, right panel).  We further tested whether NF-B and AP-1 activation are involved in the signal transduction pathway leading to the HG-MCM induction of E-selectin gene expression. HUVECs were transfected with siRNAs for p65 or c-Jun, followed by stimulation with HG-MCM for 4 h. The HG-MCM-induced E-selectin mRNA and E-selectin p540-Luc promoter activity levels ( Fig.  5D) were significantly down-regulated by the inhibition of p65 or c-Jun by siRNA, indicating that NF-B and AP-1 are involved in the regulation of E-selectin gene expression. The p65-and c-Jun-specific siRNAs (compared with control siRNA) caused an 80% reduction in p65 and c-Jun protein expressions, respectively (supplemental Fig. S3).
The JNK and p38 Signaling Pathways Are Involved in HG-MCM-induced E-Selectin Promoter Activity-To evaluate whether the inhibition of E-selectin expression by the JNK and p38 signaling pathways occurs at the transcriptional level, we studied the effects of inhibitors or siRNAs against JNK and p38 upon HG-MCM-induced E-selectin p540-Luc promoter activity. Culturing of the HUVECs in HG-MCM increased the luciferase activity by 4.5-fold compared with unstimulated cells after normalization with a transfection control (Fig. 6A). Pretreatment of the cells with SP600125 and SB203580 or transfection with JNK siRNA and p38 siRNA resulted in a marked inhibition of HG-MCM-induced E-selectin promoter activity (Fig. 6A). The simultaneous treatment of ECs with SP600125 and SB203580 resulted in the additive inhibition of HG-MCMinduced E-selectin promoter activity (Fig. 6A). In HUVECs, transfection with siRNA or promoter constructs at the concen-trations used in these experiments was found not to cause cytotoxicity, on the basis of cell numbers and morphology (data not shown).
To explore whether JNK and p38 activate the promoter responsible for E-selectin transcription via NF-B and AP-1, HUVECs were pretreated with MAPK inhibitors followed by HG-MCM stimulation. NF-B p65 and AP-1 activation was then assessed by TF ELISA. Pretreatment of the cells with SP600125 and SB203580 significantly inhibited the HG-MCMinduced p65 DNA and AP-1 DNA binding activity (Fig. 6B). Pretreatment with these inhibitors simultaneously caused an additive inhibition of HG-MCM-induced p65 DNA and AP-1 DNA binding activity (Fig. 6B). In addition, ChIP analysis revealed that the combined pretreatment of ECs with SP600125 and SB203580 inhibited the HG-MCM induction of the p65 (Fig. 6C) and c-Jun promoter (D) binding activities.

HG-MCM-induced E-Selectin Expression Is Mediated by MIP-1␣ and MIP-1␤-The increases in E-selectin expression
in ECs by HG-MCM suggested that macrophages under high glucose condition release soluble mediators to exert paracrine effects on ECs and thereby induce the E-selectin expression. We examined the expression levels of cytokines in NG-MCM and HG-MCM using a human cytokine array system. Using this array, we identified that the MIP-1␣ and MIP-1␤ proteins are released at significantly higher levels in HG-MCM cultured cells compared with those exposed to NG-MCM (Fig. 7A). To confirm these results, the protein levels of MIP-1␣ and MIP-1␤ in cells grown in NG-, M-, and HG-MCM were analyzed by

. Luciferase reporter analysis of the E-selectin promoter and the roles of NF-B and AP-1 in HG-MCM-induction of E-selectin promoter activity and transcription in HUVECs.
A, HUVECs were cotransfected with 5Ј-deletion constructs and stimulated with HG-MCM for 4 h. E-selectin promoter activity was measured by luciferase assay normalized to ␤-galactosidase activity and is shown relative to that of HUVECs transfected with p540-Luc (set at 100%). B, NF-B p65 and AP-1 activation levels were determined by TF ELISA. All bar graphs represent fold increases over the levels in control ECs (CL), calculated as the mean Ϯ S.E. from three independent experiments. *, p Ͻ 0.05 versus CL p65. #, p Ͻ 0.05 versus CL AP-1. C, ChIP assays were performed for NF-B using p65 antibodies and for AP-1 using c-Jun antibodies. ELISA. Culturing of monocyte-derived macrophages for 4 days in NG-4D, M-4D, or HG-4D conditions caused only slight increases in MIP-1␣ and MIP-1␤ in the culture medium, but significant increases of MIP-1␣ and MIP-1␤ were observed in HG-MCM (Fig. 7B).
The incubation of ECs with neutralizing antibodies against MIP-1␣ or MIP-1␤ significantly inhibited HG-MCM-induced E-selectin mRNA expression and promoter activity (Fig. 7C) in addition to HG-MCM-induced NF-B p65 and AP-1 activation (D). The simultaneous treatment of ECs with neutralizing antibodies against MIP-1␣ and MIP-1␤ caused further inhibition of E-selectin promoter activity and mRNA expression (Fig. 7C) and NF-B p65 and AP-1 activation (D). In addition, ECs directly stimulated with MIP-1␣ or MIP-1␤ (5 ng/ml each) had no effect on the expression of E-selectin (Fig. 7E). However, the costimulation of ECs with NG-MCM and MIP-1␣ or MIP-1␤ had effects upon E-selectin expression that were similar to HG-MCM (Fig. 7E).

DISCUSSION
Diabetic patients have an increased susceptibility to atherosclerosis (20,21). E-selectin, which plays significant roles in atherosclerosis and type 2 DM, is a major adhesion molecule expressed by vascular ECs. In this study, we cultured primary human macrophages under HG and NG conditions and elucidated the molecular mechanisms underlying the roles of HGtreated macrophages in modulating E-selectin expression in ECs. The major findings of our present analyses are 1) that conditioned media from HG-treated macrophages induced slight increases in the EC ICAM-1 or VCAM-1 transcript levels compared with NG-MCM-treated Ecs, 2) that HG-MCM significantly increases the expression and secretion of EC E-selectin compared with NG-MCM-treated Ecs, 3) that HG-4D medium and NG-MCM only caused moderate increases in E-selectin expression, 4) that this increase in E-selectin expression is attributable to the paracrine effect of the chemokines MIP-1␣ and MIP-1␤ released by HG-treated macrophages, and 5) that the HG-MCM-increased expression of E-selectin is mediated by the JNK/p38 and NF-B/AP-1 pathways.
Endothelial dysfunction or damage can be evaluated by the measurement of biomarkers released from ECs, such as sE-selectin, sICAM-1, sVCAM-1, and von Willebrand factor (15,22). Several studies have now demonstrated that the plasma level of sE-selectin is also increased in patients with hypertension, type 2 DM, atherosclerosis, and dyslipidemia, all of which are well established risk factors for cardiovascular diseases (23)(24)(25)(26). In diabetic patients in several earlier studies, sE-selectin levels have also been found to be elevated and to correlate with the onset of glycemia (13,23,27). In particular, sE-selectin is a strong independent predictor of the incidence of DM after adjustment for obesity and other clinical and lifestyle parameters in both men and women (28). The results of this study demonstrate that conditioned media from HG-treated macrophages not only promotes the secretion of sE-selectin but also induces their gene transcription and expression in human ECs. Our analysis of the E-selectin promoter activity with different plasmid constructs further revealed that the NF-B and AP-1binding regions function as the cis-element for HG-MCM responsiveness via JNK and p38 phosphorylation. Previous studies have shown that NF-B and AP-1 can be activated through the JNK and p38 pathways in ECs (29,30). The regu-lation of gene expression through the use of combinations of different transcription factors, including NF-B and AP-1, has been reported, and JNK and p38 have been shown to be involved in the activation of p65 and c-Jun (31,32). In our present experiments, we performed luciferase assays and found that NF-B and AP-1 cooperate to activate the human E-selectin promoter. We employed TF ELISA and ChIP assays to demonstrate that the regulation of E-selectin gene expression in ECs is mediated by increased NF-B p65 and AP-1-DNA binding activities following JNK and p38 phosphorylation.
Because there is considerable evidence that inflammation plays a major role in atherogenesis, it was important to evaluate the proatherogenic activity of monocytes/macrophages in diabetes. Macrophage accumulation is a feature of type 2 DM and is associated with the development of atherosclerosis (33). Monocytes in a proinflammatory state have also been detected in severe hyperglycemic type 2 DM patients, indicating that the proinflammatory status of circulating monocytes can contribute to the development of the atherosclerotic plaques when they are recruited to the vascular lesion. It is also becoming evident that chronic inflammation is associated with type 2 DM conditions and is involved in all stages of the atherosclerotic process (34,35). The macrophage effector chemokines MIP-1␣ and MIP-1␤, produced and secreted by activated macrophages to attract other proinflammatory cells, are crucial in recruiting macrophage themselves to sites of inflammation. By using protein arrays to perform a systematic analysis and ELISA assays of the expression levels of soluble mediators produced by macrophages after HG stimulation, we have identified MIP-1␣ and MIP-1␤ as the proteins whose expression in HG-MCM was significantly higher than that in NG-or M-MCM. By using neutralizing antibodies against MIP-1␣ and MIP-1␤, we have demonstrated that these two chemokines in HG-MCM are the major factors contributing to the HG-MCM-increased E-selectin expression in ECs. A previous study has demonstrated that MIP-1␣ cannot induce E-selectin expression in ECs (27). Our current data also indicate that MIP-1␣ and MIP-1␤ have no direct effects upon E-selectin expression in ECs. However, the costimulation of ECs with NG-MCM and MIP-1␣ or 1␤ had effects similar to those of HG-MCM upon E-selectin expression. Therefore, our results suggest that MIP-1␣ and 1␤ are major factors in HG-MCM-induced E-selectin expression and that other minor factors released from macrophages may also have synergistic roles with MIP-1␣ and MIP-1␤.
The modulatory effects of a high-glucose environment on monocyte/macrophage function have been documented previously. Glucose has been shown to increase the production of cytokines by human monocytes (36). HG treatments of THP-1 monocytes have also been found to induce changes in the expression levels of several cytokines, chemokines, and related molecules (7). In this study, we observed for the first time that MIP-1␣ and MIP-1␤ are significantly increased only in macrophages after differentiation in an HG environment. It has been shown previously that glycated human serum albumin enhances MIP-1␤ mRNA expression in macrophage-like differentiated U937 cells (34). Our present data also demonstrate that the protein levels of MIP-1␣ and MIP-1␤ are similar when monocytes are cultured in an HG, NG, or mannitol environment for 4 days. HG-4D medium and NG-MCM only caused moderate increases in E-selectin expression. These results suggest that the gene expression of macrophages may be altered after transformation from monocytes under hyperglycemic conditions. Hence, we speculate that activated monocytes under high-glucose conditions may bind to ECs and, following transmigration, may differentiate into macrophages in the subendothelial space. The accumulation of macrophages in the arterial wall thereby increases the MIP-1␣ and MIP-1␤ levels, and contributes to proinflammatory responses (Fig. 9). We have also demonstrated the functional consequence of the MIP-1␣ and MIP-1␤ modulation of the HG-MCM-induced EC gene expression by measuring the adhesiveness of the monocytic THP-1 cells to HG-MCM-treated ECs. HG-MCM pre-  In summary, our results characterize the mechanisms by which HG-treated macrophages induce E-selectin expression in ECs. HG-treated macrophages induce the activation of JNK/ p38 and NF-B/AP-1 signaling pathways and ultimately enhance E-selectin expression in ECs. These findings provide insights into the mechanisms underlying the interplay between hyperglycemic macrophages with ECs in modulating EC signaling and gene expression, which may well be involved in the development of vascular complications in patients with diabetes.