C-Reactive Protein Promotes Monocyte Chemoattractant Protein-1—Mediated Chemotaxis Through Upregulating CC Chemokine Receptor 2 Expression in Human Monocytes
Background— Inflammation plays a crucial role in atherosclerosis. An elevated serum C-reactive protein (CRP) level is a strong marker for future atherosclerotic cardiovascular diseases. In addition, recent data suggest that CRP may directly promote atherogenesis. In this study, we investigated whether CRP can directly activate human circulating monocytes.
Methods and Results— Incubation of THP-1 monocytes with CRP (10 μg/mL) increased CC chemokine receptor 2 (CCR2) expression at both the protein and transcript levels, which in turn enhanced chemotaxis mediated by monocyte chemoattractant protein-1 (MCP-1) up to 2-fold. The CRP-induced upregulation of CCR2 expression involved binding of CRP to the FcγR, most notably FcγRI, and phospholipase D1 activation. Serum high-sensitivity CRP levels in 52 normocholesterolemic human subjects were positively correlated with CCR2 surface expression on circulating monocytes (r=0.62, P<0.001) and MCP-1–mediated monocyte chemotaxis (r=0.53, P<0.001).
Conclusions— Elevated blood CRP levels may promote accumulation of monocytes in the atherogenic arterial wall by increasing chemotactic activities of monocytes in response to MCP-1.
Received May 29, 2003; de novo received November 12, 2003; revision received January 27, 2004; accepted February 13, 2004.
C-reactive protein (CRP) is now recognized as a strong and independent marker for vascular wall inflammation and a predictor of atherosclerotic cardiovascular events.1 Recent emerging data suggest that CRP may play a direct role in the initiation and the progression of atherosclerotic lesion formation. CRP in atheroma induces release of interleukin (IL)-1β, IL-6, and tumor necrosis factor (TNF)-α by macrophages. CRP calcium-dependently binds to LDL,2 and the CRP-Ca2+-LDL complex activates the classic complement pathway and induces foam cell formation. CRP stimulates endothelial cells to express adhesion molecules and a potent chemoattractant, monocyte chemoattractant protein-1 (MCP-1). In addition, CRP per se is a chemoattractant for monocytes (for review, see Yeh and Willerson3 and Pepys and Hirschfield4).
Previous studies show that CRP concentrations as low as 5 mg/L activate cells found in atherosclerotic lesions in vitro. In steady-state conditions, a significant subset of the general population was shown to have serum CRP levels ≥5 mg/L,1 whereas those with acute inflammatory conditions were frequently found to have serum CRP levels ≥10 mg/L.4
The experiments in the present study were designed to investigate the possibility that serum levels of CRP affect circulating monocyte function. Chemotaxis is one of the most important activities of circulating monocytes. The most dominant chemotaxis receptor in monocytes is the CC chemokine receptor 2 (CCR2), which mediates chemotactic movement of monocytes in response to MCP-1. CCR2 plays a pivotal role in atherosclerotic plaque formation because genetic disruption of CCR2 in atherosclerosis-prone mice markedly decreased atherosclerosis.5
In the present study, we demonstrate that CRP regulates CCR2-mediated monocyte function. CRP induces upregulation of CCR2 expression in human monocytes, which eventually promotes monocyte chemotactic movements in response to MCP-1.
THP-1 monocytes were maintained as described previously6 using heat-inactivated FBS (Gibco) to avoid spontaneous complement-induced activation. THP-1 monocytes were incubated for 24 hours with human recombinant CRP (Calbiochem) or whole human IgG, or the Fc or Fab fragments of human IgG (Jackson Immunoresearch). The CRP stock was IgG-free and conformationally intact, as determined by electrophoresis on nondenaturing gel. To confirm CRP-specific effects, CRP in the CRP stock was preabsorbed using plates coated with anti-human CRP antibodies (Euroimmun) and used as negative control. Incubation for 24 hours with F(ab)2 fragments of monoclonal anti-human CD64 and CD32 mouse IgGs (3 μg/mL) (Research Diagnostics, Inc) was performed to inhibit CRP binding to CD64 and CD32, respectively. To block phospholipase D1 (PLD1) activity, 3 μmol/L specific antisense oligonucleotide7 was incubated with THP-1 monocytes for 36 hours (12 hours before and then for the 24-hour duration of CRP incubation). In this setting, PLD1 transcripts were nearly completely downregulated, as determined by reverse transcription–polymerase chain reaction (RT-PCR) assays. The concentration of endotoxin in cell cultures was <1 ng/mL, as determined by a timed-gel endotoxin kit assay (Sigma).
Human monocytes were isolated from freshly drawn whole blood by negative selection using the mini-MACS microbead cell isolation system (Miltenyi Biotec). Informed consent was obtained from all blood donors, as approved by the Human Subjects Committee of the Asan Medical Center (Seoul, South Korea). The purity of isolated monocytes was >90%, as estimated by flow cytometry using phycoerythrin (PE)-labeled anti-human CD14 mouse IgG (PharMingen).
CCR2 and β-integrin cell surface protein levels were estimated by flow cytometry as previously described6 using PE-conjugated mouse IgG detecting human CCR2 or β-integrins (CD11a, CD11b, and CD49d) (PharMingen). In control experiments, PE-conjugated nonspecific mouse IgG was used to measure nonspecific binding. The relative surface expression of CCR2 and β-integrins was estimated by subtracting the mean fluorescence intensity (MFI) of control cells from that of cells labeled with antibodies detecting CCR2 or β-integrins. Data were analyzed by the FACScan instrument using Cellquest software (Becton Dickinson).
For the quantitative assessment of CCR2 transcript levels, competitive RT-PCR analysis using truncated CCR2 cDNA as a competitor was performed as previously described.6 After PCR reaction using the specific primers (sense, 5′-ATGCTGTCCACATCTCGTTCTCG-3′; antisense, 5′-TTATAAACCAGCGAGACTTCCTGC-3′), densities of the CCR2 PCR product (1083 bp) and the truncated competitor cDNA PCR product (898 bp) on 2% agarose gels were measured using UN-SCAN-IT software and plotted, and the point at which both curves intersected indicated equal concentrations of both. As an internal standard, human GAPDH was amplified using the sense primer 5′-TCGGAGTCAACGGATTTGGTCGTA-3′ and the antisense primer 5′-ATGGACTGTGGTCAGAGTCCTTC-3′ to ensure equal analysis conditions for competitive RT-PCR. Human MCP-1 was amplified using the specific primers (sense, 5′-CAGCCAGATGCAATCAATGC-3′; antisense, 5′-AAGTCTTCGGAGTTTGGG-3′) under the identical PCR conditions.
For estimation of specific CRP binding to THP-1 monocytes, 105 THP-1 monocytes were incubated with CRP in PBS containing 0.1% BSA at 4°C for 30 minutes. Then, CRP bound to the cell surface was detected by subsequent incubations with murine ascites fluid containing monoclonal anti-human CRP IgG (clone CRP-8, Sigma) (1:1000 final vol/vol) and 0.1 μg FITC-conjugated F(ab)2 fragments of goat IgG against mouse IgG (Jackson Immunoresearch). To estimate nonspecific CRP binding, excess CRP (50 μg/tube) was added to the labeled cells. Specific binding of CRP to THP-1 monocytes was calculated by subtracting nonspecific binding from total binding of CRP, and the Kd of CRP binding to THP-1 monocytes was calculated using Prism software.
The chemotactic activity of THP-1 and freshly isolated human monocytes in response to 10 nmol/L MCP-1 or 1 nmol/L FMLP was measured using a 48-well microchemotaxis Boyden chamber (Neuroprobe) as described elsewhere.8 Transmigrated monocytes bound to the polycarbonate membrane were stained with 0.1% crystal violet and counted in 4 random high-power fields (×400).
For monocyte adhesion assays, monolayers of human umbilical vein endothelial cells (HUVECs) (Cell Applications Inc) between the 3rd and 5th passages were prepared and pretreated with TNF-α (25 ng/mL) for 12 hours to induce adhesion molecules such as vascular cell adhesion molecule-1 and intercellular adhesion molecule-1. THP-1 monocytes (n=105) in phenol red–free RPMI 1640 medium containing 0.1% BSA were prestimulated with 10 nmol/L MCP-1 for 30 minutes to activate membrane-bound β-integrins and were then added to pretreated HUVEC monolayers and incubated for 30 minutes at 37°C. The nonadherent THP-1 monocytes were removed by gentle washing with PBS. The number of bound THP-1 monocytes was counted in 4 random high-power fields (×100).
Serum TNF-α, IL-6, and MCP-1 levels were measured using Quantikine immunoassay kits (R&D Systems). High-sensitivity CRP (hsCRP) assays were performed with an ELISA kit (Euroimmun).
All experiments were performed at least 3 times using triplicate determinations, and data are represented as mean±SD. ANOVA and post hoc analysis were performed to assess significance of changes compared with controls. A Mann-Whitney U test was performed for the comparison of 2 groups with each other.
To examine the effect of CRP on monocyte CCR2 expression, THP-1 monocytes were incubated for 24 hours with human recombinant CRP (1 to 10 μg/mL), and CCR2 protein and transcript levels were measured. Flow cytometry analysis showed that CRP dose-dependently increased cell surface CCR2 expression, maximally by 70% over basal at concentration of 5 μg/mL CRP. Coaddition of polymyxin B (25 μg/mL) during the 24-hour incubation did not affect CRP-stimulated CCR2 expression. The selective removal of CRP in the CRP stock completely abolished the CCR2 upregulation (Figure 1A). Coincubation with IFN-γ (1000 U/mL), a known inhibitor of CCR2 expression,9 abolished CRP-stimulated CCR2 upregulation (histogram in Figure 1A), indicating that the effects were caused by CRP rather than contaminating endotoxins or cytokines. Competitive RT-PCR assays showed that CRP increased CCR2 transcript levels. CCR2 cDNA levels in CRP-treated cells (10 μg/mL per 24 hours) were 3 times higher than those in untreated cells (Figure 1B). Conversely, RT-PCR analysis showed that CCR2 expression was not induced by CRP (up to 10 μg/mL for 24 hours) in human aortic endothelial cells or human vascular smooth muscle cells, whereas CRP-induced MCP-1 expression was observed in human vascular smooth muscle cells (data not shown).
Binding studies showed that CRP specifically bound to the THP-1 monocyte cell surface with a Kd of 0.34±0.18 μg/mL (1.5×10−8 mol/L). Coincubation with Fc fragments of human IgG inhibited the CRP binding by 50% (Figure 2). However, this incubation with Fc fragments did not affect CRP-induced upregulation of CCR2 (data not shown). In the absence of CRP, flow cytometry analysis showed that either Fc fragments or whole intact IgG alone increased monocyte CCR2 expression, whereas Fab fragments did not have this effect (Figure 3A). Pretreatment of THP-1 monocytes with F(ab)2 fragments of mouse IgG, which specifically blocks CRP-binding to FcγRI (CD64), inhibited CRP-mediated CCR2 upregulation, whereas blocking CRP binding to FcγRII (CD32) had no effect (Figure 3B).
RT-PCR analysis showed that THP-1 monocytes express PLD1, not PLD2 (data not shown). Flow cytometry analysis showed that pretreatment for 36 hours with PLD1-specific antisense oligonucleotides (3 μmol/L), which nearly completely downregulated PLD1 expression, inhibited CRP-stimulated CCR2 expression (Figure 3B).
We examined the effect of CRP on THP-1 monocyte chemotaxis. CRP dose-dependently increased chemotaxis in response to MCP-1 but not in response to FMLP (Figure 4). CRP 10 μg/mL, which induced a 70% increase in CCR2 surface expression in 24 hours, almost doubled MCP-1–mediated chemotaxis (Figure 4A). We have previously shown that free cholesterol upregulated CCR2 expression in THP-1 monocytes.8 In the present study, coincubating monocytes with 10 μg/mL CRP and 25 μg/mL free cholesterol resulted in a 6-fold increase in MCP-1–mediated chemotaxis. Incubating THP-1 monocytes with the CRP-depleted CRP stock did not increase MCP-1–mediated chemotaxis (Figure 4B).
Conversely, CRP did not affect expression levels of β-integrins (CD11a, CD11b, and CD49d). Accordingly, MCP-1–induced adhesion of THP-1 monocytes on TNF-α–treated HUVEC monolayers was not enhanced by CRP (data not shown).
To determine whether there is a relationship between hsCRP serum levels and CCR2 surface expression in circulating monocytes, we recruited 52 patients (male-to-female ratio, 29:23) with coronary atherosclerosis (Table). All study subjects had been clinically stable for the previous 3 months and were neither hypercholesterolemic (LDL cholesterol ≤130 mg/dL) nor diabetic. Analysis of these patients indicated that the presence of hypertension (n=18), smoking (n=21), or history of cerebrovascular incidents (n=4) was not associated with elevated monocyte CCR2 surface expression, as determined by flow cytometry. In addition, age, body mass index, serum fasting glucose levels, and lipid profiles were not correlated with monocyte CCR2 expression. Furthermore, TNF-α, IL-6, and MCP-1 levels showed no correlation. Only serum hsCRP levels showed a strong positive correlation with monocyte CCR2 expression (r=0.62, P<0.001) (Figure 5). In a multivariate analysis, the serum hsCRP level was the only variable that significantly correlated with monocyte CCR2 expression (F=13.760, significance=0.001).
MCP-1–mediated chemotaxis of freshly isolated monocytes was positively correlated with monocyte CCR2 surface expression (r=0.35, P=0.01). Serum hsCRP levels showed a strong positive correlation with MCP-1–mediated chemotaxis (r=0.53, P<0.001) (Figure 5).
The present study provides the novel finding that CRP increases expression of human monocyte CCR2, the most dominant monocyte chemotaxis receptor. CCR2 expression is reported to be regulated by endotoxins such as lipopolysaccharide as well as by inflammatory cytokines and oxidatively modified LDL.6,9,10 It is unlikely that the data we obtained using human recombinant CRP were caused by contaminating endotoxins and cytokines, because the effect of CRP on CCR2 was observed in the presence of polymyxin B. In addition, IFN-γ, an inhibitor of CCR2 expression,9 abolished CRP-mediated CCR2 upregulation. We also confirmed that CRP-induced upregulation of CCR2 was consistently observed under lipoprotein-depleted culture conditions (data not shown), indicating that the CCR2 upregulation was not because of LDL, which had been reported as a positive regulator of monocyte CCR2 expression.8 Moreover, the removal of CRP in the human recombinant CRP stock abolished the CCR2 upregulation. Thus, our study results strongly suggest that CRP directly increases CCR2 gene expression in monocytes. Conversely, CRP-stimulated CCR2 expression was not observed in human aortic endothelial cells and vascular smooth muscle cells, suggesting that CCR2 is differentially regulated according to the cell type.
CRP has been reported to bind to the family of Fcγ receptors (FcγR) in monocytes, such as FcγRI (CD64),11 FcγRII (CD32),12 and FcγRIII (CD16).13 In this study, human IgG Fc fragments competed with CRP binding to monocytes and actually increased CCR2 expression. FcγRIII is expressed in <20% of CD14+ monocytes14 and is not expressed in THP-1 monocytes.15 We found that blocking binding of CRP to FcγRI abolished the positive regulatory effect of CRP on monocyte CCR2 expression. Taken together, activation of FcγRI, which binds both Fc fragments of IgG and CRP with high affinity, may be directly involved in CCR2 upregulation.
In leukocytes, aggregation of FcγR by immune complexes leads to phagocytosis, oxidative burst, cytokine release, and degranulation by macrophages through a signaling pathway involving PLD.16 CRP-mediated phagocytosis of phosphorylcholine-conjugated erythrocytes in U937 myeloblasts resulted from signaling through FcγRI and PLD activation.17 Activation of PLD1, the dominant isoform of PLD in THP-1 monocytes,18 was required for functional coupling of FcγRI to NADPH oxidase activation and intracellular vesicular trafficking of immune complexes from endosomes to lysosomes for degradation in U937 myeloblasts.7 Our study using antisense knockdown experiments suggested that activation of PLD1 was involved in CRP-stimulated monocyte CCR2 expression, providing additional evidence that PLD1 in monocytes/macrophages modulates key cellular responses involved in inflammatory responses.
Under steady-state conditions, human and THP-1 monocytes express ≈5000 CCR2 molecules on their plasma membranes.8 Because of such a relatively low basal expression of CCR2, small changes in receptor number can profoundly affect CCR2-mediated activities. In accordance with our previous study describing the positive regulatory effects of LDL and free cholesterol on monocyte CCR2 expression,8 here, we report that a 70% increase in CCR2 expression induced by CRP resulted in a doubling of MCP-1–mediated chemotaxis. Furthermore, coincubation of THP-1 monocytes with CRP and free cholesterol synergistically enhanced MCP-1–mediated chemotaxis. A high serum hsCRP level acts synergistically with serum LDL cholesterol level to enhance the risk of coronary artery disease up to 6 times that of normal controls.1 Thus, excessive recruitment of monocytes into the proatherogenic vascular wall, resulting from enhanced MCP-1–mediated chemotaxis, may be a mechanism by which high serum CRP and LDL cholesterol levels act synergistically to promote atherosclerosis.
We found that CRP (10 μg/mL for 24 hours) did not change expression levels of monocyte β-integrins. However, a previous study demonstrated that CRP doses equivalent to those used in this study rapidly and transiently enhanced cell surface expression of the β2-integrin CD11b.19 This may be primarily because of translocation of CD11b from intracellular stores to the plasma membrane, rather than increased protein synthesis.
It is notable in our study that the estimated ED50 value of CRP for CCR2 upregulation was as low as 2.0 mg/L. About 40% of 27 939 white women enrolled in the Women’s Health Study had serum hsCRP levels >2.0 mg/L.1 The 25th, 75th and 95th percentiles of serum hsCRP levels among 1399 Koreans were 0.34, 1.23, and 4.01 mg/L, respectively.20 In the present study, serum hsCRP levels in normocholesterolemic subjects were highly associated with monocyte CCR2 surface expression and MCP-1–mediated monocyte chemotaxis. Therefore, serum CRP levels could be a major determinant of monocyte CCR2 expression and MCP-1–mediated chemotactic activity of human circulating monocytes, particularly in normocholesterolemic subjects. Although our results showed a clear relationship between serum CRP levels and monocyte CCR2 expression and the CCR2-related chemotactic activities, the size of the study was relatively small because of the intrinsic limitations of flow cytometry analysis. We anticipate that future larger-scale studies will provide further in vivo evidence of this association.
In conclusion, our data suggest that CRP promotes monocyte chemotactic activity in response to MCP-1 via upregulation of the monocyte chemotaxis receptor CCR2. Importantly, the CRP concentration needed to promote this activity is similar to that frequently observed in the blood of both healthy and diseased individuals. It is possible that the ultimate effect of this CRP activity is to cause excessive recruitment of monocytes to the atherogenic vascular wall. Thus, CRP may be not simply a marker of atherosclerotic disease but rather an active participant in the atherogenic process.
This study was supported by grants 2003-288 and 2002-282 from the Asan Institute for Life Sciences and 02-PJ1-PG10-20707-0003 from the Korean Ministry of Health and Welfare. We thank Dr Joseph L. Witztum, University of California San Diego, for reviewing the manuscript.
This study demonstrates that CRP activates human monocytes. Incubating THP-1 monocytes with CRP upregulated expression of CCR2, the most dominant monocyte chemotaxis receptor, and enhanced the monocyte chemoattractant protein-1 (MCP-1)–mediated chemotaxis. CRP-stimulated CCR2 expression involved functional activation of FcγRI and phospholipase D1. Serum high-sensitivity CRP levels in 52 normocholesterolemic subjects were positively correlated with CCR2 expression in circulating monocytes and MCP-1–mediated monocyte chemotaxis. Our data suggest that CRP may directly cause accumulation of monocytes/macrophages in the atherosclerotic wall by increasing monocyte CCR2 expression, thereby enhancing MCP-1–mediated monocyte chemotaxis.
Chang MK, Binder CJ, Torzewski M, et al. C-reactive protein binds to both oxidized LDL and apoptotic cells through recognition of a common ligand: phosphorylcholine of oxidized phospholipids. Proc Natl Acad Sci U S A. 2002; 99: 13043–13048.
Yeh ET, Willerson JT. Coming of age of C-reactive protein: using inflammation markers in cardiology. Circulation. 2003; 107: 370–372.
Melendez A, Bruetschy L, Floto RA, et al. Functional coupling of FcγRI to nicotinamide adenine dinucleotide phosphate (reduced form) oxidative burst and immune complex trafficking requires the activation of phospholipase D1. Blood. 2001; 98: 3421–3428.
Han KH, Tangirala RK, Green SR, et al. Chemokine receptor CCR2 expression and monocyte chemoattractant protein-1 mediated chemotaxis in human monocytes: a regulatory role for plasma low-density lipoprotein. Arterioscler Thromb Vasc Biol. 1998; 18: 1983–1991.
Tangirala RK, Murao K, Quehenberger O. Regulation of expression of the human monocyte chemotactic protein-1 receptor (hCCR2) by cytokines. J Biol Chem. 1997; 272: 8050–8056.
Sica A, Saccani A, Borsatti A, et al. Bacterial lipopolysaccharide rapidly inhibits expression of C-C chemokine receptors in human monocytes. J Exp Med. 1997; 185: 969–974.
Crowell RE, Du Clos TW, Montoya G, et al. C-reactive protein receptors on the human monocyte cell line U-937: evidence for additional binding to FcγRI. J Immunol. 1991; 147: 3445–3451.
Bharadwaj D, Stein MP, Volzer M, et al. The major receptor for C-reactive protein on leukocytes is FcγRII. J Exp Med. 1999; 190: 585–590.
Khreiss T, Jozsef L, Hossain S, et al. Loss of pentameric symmetry of C-reactive protein is associated with delayed apoptosis of human neutrophils. J Biol Chem. 2002; 277: 40775–40781.
Randolph GJ, Sanchez-Schmitz G, Liebman RM, et al. The CD16(+)(FcγRIII(+) subset of human monocytes preferentially becomes migratory dendritic cells in a model tissue setting. J Exp Med. 2002; 196: 517–527.
Fleit HB, Kobasiuk CD. The human monocyte-like cell line THP-1 expresses FcγRI and FcγRII. J Leukoc Biol. 1991; 49: 556–565.