oxLDL antibody inhibits MCP‐1 release in monocytes/macrophages by regulating Ca2+/K+ channel flow

Abstract oxLDL peptide vaccine and its antibody adoptive transferring have shown a significantly preventive or therapeutic effect in atherosclerotic animal model. The molecular mechanism behind this is obscure. Here, we report that oxLDL induces MCP‐1 release in monocytes/macrophages through their TLR‐4 (Toll‐like receptor 4) and ERK MAPK pathway and is calcium/potassium channel‐dependent. Using blocking antibodies against CD36, TLR‐4, SR‐AI and LOX‐1, only TLR‐4 antibody was found to have an inhibitory effect and ERK MAPK‐specific inhibitor (PD98059) was found to have a dramatic inhibitory effect compared to inhibitors of other MAPK group members (p38 and JNK MAPKs) on oxLDL‐induced MCP‐1 release. The release of cytokines and chemokines needs influx of extracellular calcium and imbalance of efflux of potassium. Nifedipine, a voltage‐dependent calcium channel (VDCC) inhibitor, and glyburide, an ATP‐regulated potassium channel (K+ ATP) inhibitor, inhibit oxLDL‐induced MCP‐1 release. Potassium efflux and influx counterbalance maintains the negative potential of macrophages to open calcium channels, and our results suggest that oxLDL actually induces the closing of potassium influx channel – inward rectifier channel (Kir) and ensuing the opening of calcium channel. ERK MAPK inhibitor PD98059 inhibits oxLDL‐induced Ca2+/Kir channel alterations. The interfering of oxLDL‐induced MCP‐1 release by its monoclonal antibody is through its FcγRIIB (CD32). Using blocking antibodies against FcγRI (CD64), FcγRIIB (CD32) and FcγRIII (CD16), only CD32 blocking antibody was found to reverse the inhibitory effect of oxLDL antibody on oxLDL‐induced MCP‐1 release. Interestingly, oxLDL antibody specifically inhibits oxLDL‐induced ERK MAPK activation and ensuing Ca2+/Kir channel alterations, and MCP‐1 release. Thus, we found a molecular mechanism of oxLDL antibody on inhibition of oxLDL‐induced ERK MAPK pathway and consequent MCP‐1 release.


Introduction
Atherosclerosis is an inflammatory disease induced by imbalance of lipid metabolism -hyperlipaemia [1]. Oxidation of low-density lipoprotein (oxLDL) induced not only macrophage uptake and foam cell formation, but also T cell reaction and even antibody production by B cells [2][3][4]. Cytokine expression in arteriosclerotic animal model serum has a Th1 immune profile. Atherosclerosis is now regarded as an autoimmune disease [5]. Vaccination of oxLDL peptide or adoptive transfer of antibodies against oxLDL to arteriosclerotic animal model has exhibited preventive or therapeutic effects [6]. Plaque area after immune therapies reduced to 50% [7]. Moreover, macrophage staining and MCP-1 release assay have shown that inflammation decreased after oxLDL antibody treatment. The molecular mechanisms of antibody regulation of inflammatory reaction are less obvious.
We previously reported that bone marrow cells from FccRIIB-deficient mice which were transplanted to low-density lipoprotein receptor-deficient (LDLR À/À ) mice induced atherosclerotic lesion area in the descending aorta about fivefold larger than in LDLR À/À control mice [8]. Using mac-1 and P-ERK MAPK antibody staining of splenocytes, it was found that ERK activation in bone marrow-transplanted mice was significantly higher than in control mice (unpublished results). These results give a clue that inhibition of inflammatory reaction by oxLDL antibody might not be through its neutralizing effect but through its FccRIIB and ERK MAPK signal transduction pathway.
Monocyte chemoattractant protein-1 (MCP-1/CCL2) is one of the key chemokines that regulates migration and infiltration of monocytes/macrophages into the lesion area. It is overexpressed in patients with atherosclerosis [16]. MCP-1 release involves Ca 2+ activity inside the cell [17]. The Ca 2+ channel is voltage-dependent, and the extent of Ca 2+ influx depends on the degree of cell membrane potential polarization. The more the negative potential on the cell membrane, the more the Ca 2+ influx into cytoplasm when the Ca 2+ channel is activated [18,19]. The maintenance of cell membrane potential relies on the ratio of outward to inward K + current. Thus, K + outward current increase or K + inward current decrease may result in cell membrane potential hyperpolarization [20][21][22][23].
Here, we report that oxLDL mAb inhibited monocyte MCP-1 release and mRNA expression in a dose-dependent manner in the antibody treatment experiment [24]. We used in vitro study of oxLDLinduced monocyte/macrophage MCP-1 release model to investigate the molecular mechanism of oxLDL mAb on inhibition of MCP-1 release and its cellular signal transduction pathways. We found that oxLDL mAb inhibits MCP-1 release through its FccRIIB, regulating oxLDL?TLR-4?ERK MAPK?K ir closure?Ca 2+ channel openingmediated MCP-1 release. The findings may reveal the molecular mechanism of how oxLDL mAb might be able to inhibit inflammatory reaction in atherosclerotic animal model.

Ethical statement
The study design was approved by Southern Medical University ethics board, and the performance was followed according to the Helsinki declaration. Written informed consent was obtained from all donators prior to treatment.
Preparation of CD14 + human monocytes and RAW264.7 cells Monocytes were prepared from human peripheral blood mononuclear cells (PBMCs), as described previously [25]. Briefly, venous blood from healthy volunteers was heparinized and layered over Ficoll-Isopaque (Pharmacia, Freiburg, Germany) density gradient reagent according to the manufacturer's instructions; mononuclear cells were separated by centrifugation at 400 9 g for 30 min. at room temperature. Mononuclear cells were collected and washed two times with PBS without Ca 2+ and Mg 2+ by centrifugation at 250 9 g for 20 min. at 4°C. Cells were then diluted with complete RPMI 1640 medium. Human monocytes were purified with MACS CD14 microbeads (Miltenyi Biotec GmbH, Bergisch Gladbach, Germany) according to the manufacturer's instructions.
Murine macrophage cell line RAW264.7 was from American Type Culture Collection (ATCC). It was grown in complete DMEM medium under standard tissue culture conditions.
replaced with fresh media containing compounds and inhibitors as indicated. Fluorescence was measured every 0.6 sec. for 5 min. Image analysis was performed using Zeiss LCS software, and fluorescence of every cell in each field was measured. On average, 78.6 AE 12.4 cells were separately analysed per condition in each experiment. Cells exhibiting an increase in fluorescence of at least two times that of background, followed by a decrease in fluorescence and another increase in fluorescence, were scored as positive calcium oscillations. Each inhibitor was performed in duplicate within the experiment, and data shown are representative of at least three independent experiments.
The patch clamp experiments were performed as described previously [31] using Micromanipulator MP-285 and MultiClamp 700B patch clamp amplifier (Axon Instruments, Union City, CA, USA). Signals were low-pass-filtered at 5 kHz (low-pass Bessel filter), digitized (sample rate: 10 kHz) using a Digidata 1440A converter (Axon Instruments) and stored and analysed using pClamp 10.2 software (Axon Instruments). The computer and software system was also used for generating voltage and current pulses. Voltage-dependent currents were evoked by voltage pulses of 300-msec. duration delivered every 5 sec. from a holding potential of À50 mV in 10-mV increments. The steps ranged from À170 to +70 mV. Independent experiments were repeated, and 5-10 cells in each group were measured (n = 5).

Western blotting
Cells were lysed with M-PER Protein Extraction Reagent (Pierce, Rockford, IL, USA) supplemented with protease and phosphatase inhibitor cocktail, and protein concentrations of the extracts were measured by bicinchoninic acid (BCA) assay (Pierce). Forty micrograms of the protein was used and loaded per lane, subjected to sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE), transferred onto nitrocellulose membranes and then blotted as described previously [32]. Detection antibodies used (mentioned above in the section 'Materials') recognized phosphorylation of ERK, JNK, p38 and c-jun. The bactin antibody was used as a control.

MCP-1 ELISA
The Human MCP-1 ELISA Kit was employed to assay cell culture conditioned medium and carried out according to the manufacturer's instructions (Uscn Life Science Inc.). CD14 + monocytes were pre-treated with indicated treatments and further incubated under standard tissue culture conditions for 2 days. The cytokine levels in the cell culture media were detected by a biotin-labelled antibody and HRP-conjugated streptavidin and measured at a wavelength of 450 AE 10 nm.

Statistical analysis
The results were expressed as means AE S.D. One-way ANOVA and Student's t-test from an SPSS software package were used for statistical evaluation. P-values <0.05 were considered significant: *: P < 0.05; **: P < 0.01; ***: P < 0.001.

Results
oxLDL induces monocyte/macrophage MCP-1 release through TLR-4-and MAPK-dependent pathways oxLDL-containing human serum (has been tested compare to FBS in Figure S2) and oxLDL (30 mg/ml) in the presence of 10% FBS both induced CD14 + monocyte/macrophage MCP-1 release ( Fig. 1A and Figure S5), whereas human serum treated in an oxLDL antibody-conjugated agarose column, named oxLDL(-) serum, had no effect (Fig. 1A). To test which receptor mediated the oxLDL-induced MCP-1 release, we explored blocking antibodies against CD36, TLR-4, SR-AI and LOX-1. Only the TLR-4 blocking antibody showed an inhibitory effect, and simultaneous administration of the four blocking antibodies had no synergistic effect ( Fig. 1A and Figure S5). To further evaluate which MAPK might be involved in oxLDL-containing human serum-induced monocyte/macrophage MCP-1 release, p38, ERK and JNK MAPK inhibitors were investigated. All these signal transduction pathways might be involved, but blocking the ERK MAPK pathway resulted in a more dominant inhibition (Fig. 1B).
oxLDL induces monocyte/macrophage MCP-1 release that is ATP-regulated potassium/calcium channel-dependent The release of cytokines and chemokines from monocytes/macrophages needs Ca 2+ ion involvement, and Ca 2+ and K + ions have a charge balance relation in the cytosol [17,33]. Influence on the intracellular K + concentration by its ATP-regulated channel (K + ATP ) inhibitor (glyburide; Fig. 2A) or the calcium concentration by the Ca 2+ channel inhibitor (nifedipine; Fig. 2B) resulted in significantly reduced human serum-induced MCP-1 secretion from monocytes/macrophages in a dose-dependent manner.  oxLDL induces inward rectifier K + (Kir) channel repression in RAW264.7 cells RAW264.7 cells were stimulated with voltage steps in the wholecell configuration to detect ionic membrane currents. Clamp steps from the holding potential of À50 mV to voltages between À170 and +70 mV elicited the membrane currents as shown in Figure 3A. Hyperpolarizing clamp steps evoked inward rectifier currents and a current density of À24.3 AE 7.3 pA/pF at À120 mV (n = 10). These currents exhibited a prominent time-and voltagedependent inactivation. Depolarizing clamp steps between À40 and +100 mV did not induce any outward currents ( Fig. 3C and E). This is a classical Kir current with strong inward rectification, the kinetics of which are similar to those of Kir2.1-encoded K + channel current [34,35].
To verify the existence of K + -sensitive channels, the effect of the K + channel blocker CsCl (140 mM) was tested. Figure 3B and C shows the current amplitude and current density versus voltage, respectively, of the potassium channel when K + in the bath solution   2 Human serum-induced MCP-1 release is Ca 2+ /K + channel-dependent. Primary CD14 + monocytes were activated either by oxLDL-containing human serum or by lipid-depleted serum (Lipid(-)). MCP-1 released from the CD14 + monocytes into the culture medium was tested with ELISA. (A) CD14 + monocytes were pre-treated with the potassium channel inhibitor glyburide (1 nM, 10 nM, 100 nM, 1 lM respectively) and then exposed to oxLDL-containing human serum or control lipiddepleted serum. ***P = 0.0001, one-way ANOVA. All data are shown as means AE S.D. (B) CD14 + monocytes were pre-treated with the calcium channel inhibitor nifedipine (1 lM, 10 lM and 100 lM, respectively) and then exposed to oxLDL-containing human serum or control lipiddepleted serum (n = 3). **P = 0.0035, ***P = 0.0001, one-way ANOVA. All data are shown as means AE S.D. was replaced by CsCl. The application of CsCl inhibits the inward rectifier currents, which reduces the inward currents from À24.3 AE 7.3 pA/pF to À1.1 AE 1.2 pA/pF (n = 8) at À120 mV (Fig. 3C). Figure 3D and E shows the current amplitude and current density versus voltage, respectively, of the potassium channel when cells were exposed to oxLDL (30 lg/ml). The amplitude of the inward current density was reduced from À24.1 AE 9.4 pA/pF to À14.4 AE 6.8 pA/pF (n = 9) at À100 mV (Fig. 3E).
Treatment with an inhibitor of p38 (SB203580) did not have any effect on inward currents (Fig. 5A), while inhibitors of JNK (SP600125) and ERK (PD98059) (Fig. 5B and C) both increased the inward currents (P < 0.05). The amplitude of the inward current density at À100 mV in the presence of the inhibitors of JNK and ERK increased from À14.4 AE 6.8 pA/pF to À24.8 AE 6.0 pA/pF and to À22.8 AE 8.1 pA/pF, respectively (n = 9) (Fig. 5D). These results indicated that oxLDL-induced inhibition of inward rectifier currents of potassium was both ERK-and JNK MAPK-dependent.

oxLDL-induced generation of [Ca 2+ ]i oscillations in macrophages is ERK-and JNK MAPKdependent
oxLDL has been shown to induce an increase in the intracellular calcium concentration ([Ca 2+ ] i ) in bone marrow-derived macrophages [36]. Here, we tested whether oxLDL influences the calcium influx in RAW264.7 cells. Cells were loaded with the fluorescent calcium dye Fluo-4 in PBS buffer to visualize the Fig. 3 oxLDL induces inward rectifier currents of K + (Kir) channel closure in RAW264.7 cells. Representative traces of RAW264.7 cells current amplitude were documented when cells were exposed to foetal bovine serum (FBS, as control) (A) or when the K + was replaced in the bath solution by CsCl to confirm that currents were coming from the potassium channels (B) or exposed to oxLDL (30 mg/ml) in the presence of 10% FBS (D). Current density (pA/pF) was recorded either when cells were exposed to FBS or after replacement of K + in the bath solution by CsCl (C), or exposed to oxLDL (E) from a holding potential of À50 mV in 10 mV increments evoked at voltage steps from À170 mV to + 70 mV (mean AE SD from 5 to 10 cells). fluctuation of the intracellular Ca 2+ level in RAW264.7 cells. Cells were pre-treated with or without inhibitors of ERK, JNK and p38, respectively, for 30 min. and thereafter stimulated with oxLDL. Native LDL (nLDL) was used as a control. As shown in Figure 6A     An increase in [Ca 2+ ]i can be mediated by an influx of Ca 2+ from the extracellular environment or from intracellular Ca 2+ stores. To assess the contribution of extracellular Ca 2+ , RAW264.7 cells were incubated in medium containing nifedipine to block the membrane Ca 2+ channel. The percentage of Ca 2+ oscillations in response to oxLDL was reduced by nifedipine (25.2%) to about half of the levels observed in cells incubated without the blocker (Fig. 6B). This indicates that oxLDL-induced Ca 2+ oscillations are only partly dependent on the membrane Ca 2+ channels, which is in accordance with Johnny H. Chen's results showing that not only extracellular Ca 2+ , but also intracellular stores of Ca 2+ account for the oxLDL-induced [Ca 2+ ]i oscillations [36].
oxLDL activates MAPKs, while the antibody recognizing oxLDL (BI-204) inhibits the ERK MAPK pathway Western blot experiments confirmed that oxLDL-containing human serum induced p38, ERK and JNK phosphorylation, while the presence of the oxLDL monoclonal antibody (BI-204) only influenced the ERK activation. The unspecific FITC-8 antibody was used as a control.
To further evaluate the oxLDL monoclonal antibody effect, also phosphorylation of the ERK MAPK substrate, c-jun, was tested, which confirmed the inhibition of the ERK activity (Fig. 8A). To additionally test whether oxLDL (30 lg/ml for 30 min., which has been verified to be the best stimulation concentration and pre-treated time in Figure S3A and S3B) induced ERK MAPK phosphorylation, a TLR-4 blocking antibody was included. We found that the TLR-4 blocking antibody blocked Cu 2+ -modified (oxLDL) LDL-induced activation of MAPKs (Fig. 8B), while the Fe 3+ -modified LDL (minimal modified LDL, mmLDL) only showed a small effect on the ERK MAPK activation (Fig. 8C).

Discussion
Toll-like receptor 4 has been reported responsible for not only oxLDLinduced cellular signal transduction, but also the uptake of lipids by monocytes/macrophages [12]. Our data suggest ERK MAPK plays a key role in both oxLDL-induced activation and oxLDL mAb inhibition of phagocyte activation. This signal transduction hub is a centre in the signal transduction pathways for the regulation of both ionic channel permeability and eventually MCP-1 release. That oxLDL mAb inhibits MCP-1 release reflects its anti-inflammatory nature and may explain the reason for inhibition of atherosclerosis in the animal experiments.
Cytokine release from phagocytes is known to involve an increased Ca 2+ concentration in cytoplasm [37]. The increment mainly results from a stimulated influx of extracellular Ca 2+ through VDCCs [17,38]. The degree of membrane hyperpolarization determines the extent of Ca 2+ influx, while K + influx and efflux maintain the Fig. 7 BI-204 inhibits oxLDL-induced inward rectifier K + channel closure through FccRIIB. Current density (pA/pF) in RAW264.7 cells was documented when cells were exposed to foetal bovine serum (FBS, as control) or oxLDL (30 mg/ml) in the presence of 10% FBS, and in some cases pre-treated with BI-204 (4 lg/ml) (A), the CD32 antibody (3 lg/ ml) (B), the CD16 antibody (3 lg/ml) or the CD64 antibody (3 lg/ml) (C). Histogram showing current density of cells specifically evoked at À120 mV when the cells were pre-treated with these antibodies (D) (mean AE SD from 5 to 10 cells). **P = 0.0073, one-way ANOVA. All data are shown as means AE S.D. negative potential of the membrane [39]. [K + ] e and [K + ] i channel activities need to maintain sufficient negative potential to open Ca 2+ channels and ensuing cytokine release [39]. Our results indicate that oxLDL mainly repressed inward rectifier K + channel, to keep the negative charges in macrophages and open Ca 2+ channels, and we did not detect any outward K + current in RAW264.7 cells, although glyburide, which is an outward K + channel inhibitor, also attenuated oxLDLinduced MCP-1 release. It has been reported that calcium stress may activate ERK MAPK [38], and our data further show that ERK MAPK regulates closing and opening of Ca 2+ and K ir channels and thus regulates MCP-1 release. Inhibition of oxLDL-induced MCP-1 release by other groups of MAPK inhibitors also regulates opening or closing of both ion channels, but oxLDL mAb seems to inhibit oxLDL-induced ERK MAPK activity only. This is also coincident with the effect of ERK inhibitor on oxLDL-induced MCP-1 release; PD98059 is the best dominant inhibitor compared to the other groups of MAPK inhibitors in inhibition of oxLDL-induced MCP-1 release. What downstream molecule for FccRIIB could specifically down-regulate ERK MAP activity is still unclear.
We previously reported that p38 MAPK is responsible for the signal transduction on activation of macrophage expression of oxLDL-induced CD36 expression and thus is involved in lipid uptake/metabolism. The p38-specific inhibitor SB203580 dramatically inhibited oxLDL-induced foam cell formation from J774 macrophages, while PD98059 played a minor inhibitory effect [40]. The differential role of MAPKs in the regulation of lipid metabolism and imbalance of lipid metabolism-induced inflammatory reaction gives us a more clear understanding of oxLDL-induced activation of MAPK functions. oxLDL mAb specifically inhibits ERK MAPK pathway but not p38 MAPK, which is coincident with the phenomena in the vaccine and antibody treatment experiment that immune therapies influence only inflammatory effects, but not the lipid metabolisms [7].
Schiopu et al. [7] reported that antibody treatment decreased macrophage content in the plaques and MCP-1 expression in animal experiments. Although BI-204 antibody trail (GLACIER) failed in a phase II study in Europe, understanding the immune regulation of atheropathogenesis, especially on antibody-mediated anti- Fig. 8 oxLDL activates the ERK pathway through TLR-4, and the recombinant antibody recognizing oxLDL inhibits human serum-induced activation of the ERK pathway. CD14 + monocytes exposed to either foetal bovine serum (FBS, as control), oxLDL-containing human serum, Cu 2+ -oxidized LDL (oxLDL) or Fe 3+ -oxidized LDL (minimal modified, mmLDL), and in some cases pre-treated with BI-204 or the con-  inflammatory effects in arterial lesion area, still has its importance both on theoretical understanding of the disease and on future screening and selection of similar or better antibodies than BI-204, as the immune therapy against atherosclerosis still has a great prospect when we have a more detailed understanding of either molecular mechanism of immune system regulation of lipid metabolism or even more importance of subsequently induced inflammation. Li et al. [41] reported that the oxLDL antibody (MLDL1278a) inhibited oxLDL-induced MCP-1 release in a full-length IgG1 format but not in an F(ab 0 )2 format which lacks the FccR binding antibody constant region, indicating that antibody effect was by binding to Fcc receptors and it inhibited p38 MAPK. Our antibody (BI-204) used in this study was also found to inhibit oxLDL-induced MCP-1 release by FccRIIB ( Figure S1). However, BI-204 specifically inhibited ERK MAPK pathway, so these two antibodies might have different mechanisms of inhibition of macrophage activation (Fig. 9). The detailed molecular mechanism of this difference needs to have further exploration.
Taken together, we have found multiple mechanisms behind oxLDL antibody inhibition of oxLDL-induced monocyte/macrophage activation and inflammatory chemokine release (Fig. 9). The findings probably partially explained Schiopu's [7] reports that oxLDL antibody treatment had a regression effect on high-fat diet-induced atherosclerosis in Apobec-1 À/À /LDLR À/À mice, and inhibitory effect on monocyte MCP-1 release and inflammatory cell infiltration to the plaque area. As many more atherosclerotic antigens are newly found, we speculate that therapeutic effects of antibodies may have more complex mechanisms. Fig. 9 Schematic picture of the oxLDL regulation of MCP-1 release, and the mechanism of the regulatory effect of oxLDL antibody. oxLDL-induced MCP-1 release from monocytes/macrophages is regulated by Ca 2+ and K + channels, and both of them are TLR-4-and MAPK-dependent. The oxLDL antibody inhibited oxLDL-induced Kir channel closure and inhibited the oxLDL-induced ERK activation through binding to FccRIIB. Four kinds of Fcc receptors have been identified: FccRI, FccRIIA, FccRIIB and FccRIII. FccRI/III transduces activation signals through its immunoreceptor tyrosinebased activation motif (ITAM) in cell membranes. Once activated, ITAM recruits spleen tyrosine kinase (SYK) and its downstream targets, such as phospholipase C (PLCc), GTPase (Rho, Rac) and phosphatidylinositol 3-kinase (PI3K), to activate macrophages. FccRIIB transduces inhibitory signals through immunoreceptor tyrosine-based inhibitory motif (ITIM), which recruits SH2 domain-containing inositol phosphatase (SHIP), and specifically hydrolyses the PI3K product PIP3 to PIP2. We hypothesize that the oxLDL monoclonal antibody (BI-204) inhibits oxLDL-induced macrophage activation through its FccRIIB by activation of MAPK phosphatase (MKP), which might specifically inactivate the ERK MAPK pathway.