Dominant-Negative Lox-1 Blocks Homodimerization of Wild-Type Lox-1–Induced Cell Proliferation Through Extracellular Signal Regulated Kinase 1/2 Activation
C-type lectin-like oxidized low-density lipoprotein (Ox-LDL) receptor-1 (Lox-1) belongs to the same family as natural killer cell receptors Ly49A and CD94 and functionally undergoes dimerization. Although Lys262 and Lys263 in the C terminus of bovine (b)Lox-1 play an important role in the uptake of Ox-LDL, mutation of these residues has not been suggested to be a potential source of the dominant-negative property. We hypothesize that dominant-negative human (h)Lox-1 forms a heterodimer with Lox-1–wild-type (WT) and blocks Lox-1–WT–induced cell signaling. Based on the use of molecular imaging techniques with laser scanning confocal microscopy and immunoprecipitation in an hLox-1–expressing Chinese hamster ovary cell system, homodimerization of hLox-1–WT was localized in the cell membrane, and Ox-LDL activated extracellular signal regulated kinase (ERK)1/2 without the translocation of hLox-1-WT. Lys266 and Lys267 of hLox-1, corresponding with Lys262 and Lys263 of bLox-1, were mutated (hLox1-K266A/K267A), and the mutant receptor inhibited hLox-1–WT–induced thymidine incorporation and ERK1/2 activation. Although Ox-LDL binds to the dominant-negative mutant receptor and is taken up by cytoplasm, ERK1/2 activation was blocked by heterodimerization with the mutant receptor and hLox-1–WT in the cell membrane. In addition, in human coronary artery smooth muscle cells, which express hLox-1–WT, we confirmed that the activation of ERK1/2 and [3H]-thymidine incorporation was caused by the addition of Ox-LDL, and these actions were blocked by hLox1-K266A/K267A. In conclusion, the present findings constitute the first evidence that strategies aimed at blocking cell-proliferative pathways at the receptor level could be useful for impairing Lox-1–induced cell proliferation.
The endothelial receptor for oxidized (Ox) low-density lipoprotein (LDL) is Lox-1, a membrane protein that belongs structurally to the C-type lectin-like family and is expressed in vascular endothelium.1 Lox-1 is found in macrophage and vascular smooth muscle cells (VSMCs), as well as vascular endothelium cells.2 Ox-LDL stimulates the activation of extracellular signal regulated kinase (ERK)1/2 in VSMCs; the oxidation of LDL potentiates this effect, and a lipid moiety is involved. Ox-LDL activation of ERK1/2 is cell-type specific, and ERK1/2 in the pathway involving Ox-LDL contributes to the altered cellular function associated with atherogenesis.3 The overall configuration of Lox-1 is a cap formation once it transfixes conformer membrane protein. There is an amino terminus as in perikaryon, and the C terminus is extracellular. It is divided into 4 domains: a perikaryon domain, membrane penetration domain, cervical domain, and lectin-like domain, from the amino-terminal side.1
Because the crystal configurations of Ly49A and CD94, which are natural killer cell receptors in the C-type lectin-like family, have been clarified, a disulfide-linked homodimer on the cell surface crystallizes as a symmetrical oligomer, and the formation of a dimer is indispensable as an interface.4–6 A receptor function analysis of Lox-1 was reported recently, and mutagenesis and deletion studies of the C-terminal (lectin-like domain) locus suggested that it was important for functional manifestation. Ox-LDL lost its function by the deletion of 6 amino acids or substitutions of the H226, R229, and R231 residues in the C terminal of bovine (b)Lox-1.7 More recently, deletion of the 10 most C-terminal amino acid residues261–270 was enough to disrupt Ox-LDL binding activity, and substitutions of Lys262 and/or Lys263 with Ala additively attenuated the activity in bLox-1.8 In addition, when the mutation of bLox-1 failed to bind Ox-LDL, the cell-surface expression of mutant forms of Lox-1 was confirmed. These studies suggested that the function of Lox-1 may be blocked using a dominant-negative strategy. The function of Lox-1 receptor has not been clarified in terms of molecular cytology and the usefulness of the dominant-negative property. In this study, Lys266 and Lys267 of human (h)Lox-1, corresponding with Lys262 and Lys263 of bLox-1, were mutated (hLox1-K266A/K267A) as a dominant-negative candidate. Therefore, we hypothesized that dominant-negative hLox-1 forms a heterodimer with Lox-1–wild-type (WT) and blocks Lox-1–WT–induced cell ERK1/2 activation and cell proliferation.
LDL was isolated from fresh plasma obtained from healthy volunteers by sequential ultracentrifugation as a 1.019<d<1.063 g/mL fraction at 4°C in potassium bromide.9 The isolated LDL was dialyzed at 4°C in slide-A-lyzer dialysis cassettes (Pierce) against 0.15 mol/L of sodium chloride with 1 mmol/L of EDTA (pH 7.4) and, before oxidation, LDL was dialyzed in slide-A-lyzer dialysis cassettes against PBS (pH 7.4) at 4°C with 3 changes in 24 hours in the dark to remove EDTA. The ethics committee of Fukuoka University Hospital approved the study, and informed consent was obtained from each of the volunteers.
Assessment of LDL Oxidization
Minimally oxidized LDL was prepared by incubating LDL, at a concentration of 100 mg protein/mL, with 10 mmol/L CuSO4 in PBS buffer (pH 7.4) at 37°C for 90 minutes.10,11 The copper-induced oxidation of LDL was monitored by capillary electrophoresis and agarose electrophoresis.12–14 Labeling of Ox-LDL with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI, Molecular Probes) was performed as described previously.15
Plasmid Constructs and Mutagenesis
The WT and mutant hLox-1 cDNAs were amplified by PCR and subcloned into pcDNA3.1/V5-His-TOPO, pcDNA3.1/myc-His A, enhanced green fluorescent protein (pEGFP)-C1, pEGFP-N1, enhanced yellow fluorescent protein-C1, and discosoma red fluorescent protein (DsRed)-C1 vectors (Invitrogen).8 Lys266 and Lys267 in hLox-1 were substituted by alanine with mutagenic reverse primer (5′-CTGTGCTCTTAGGTTTGCCGCCGCCTGACA-3′). The mutated genes were sequenced completely to verify the mutations.
Cell Culture, Transient Transfection, and Membrane Preparation
Human coronary artery smooth muscle cells (HCSMCs) were purchased from Clonetics Corp. HCSMCs were pure, on the basis of morphology and staining for α-actin smooth muscle cell expression. The cells were 100% negative for factor VIII. Smooth muscle cell growth medium consisted of basal medium, human recombinant epidermal growth factor, insulin, human recombinant fibroblast growth factor, gentamicin, amphotericin-B, and 5% FBS, and the cells were incubated in humidified air with 5% CO2. Three to 5 passages of HCSMCs at 90% confluence were incubated in medium without FBS for 24 hours before incubation with Ox-LDL. To examine the receptor specificity of Ox-LDL action, HCSMCs were pretreated with human Lox-1 blocking antibody (Jmab92, 10 mg/mL)16 for 30 minutes, and then exposed to Ox-LDL. Chinese hamster ovary (CHO; ldlA7) cells17 were seeded at 90% confluence in 15-cm dishes before transfection with Lipofectamine 2000 (Gibco). On the second day, cells were harvested with trypsin and replated on dishes or slides. Forty-eight hours after transfection, the cells were used for further analysis. Clonal cell lines that permanently expressed Lox-1 were selected by 800 mg/mL of G418. Transfected cells were harvested, and cell membranes were prepared by the nitrogen Parr bomb disruption method in the presence of protease inhibitors.
Immunoblotting Analysis for Lox-1
The cells harvested after incubation with or without Ox-LDL for 10 minutes were used to measure the expression of ERK1/2 activity or Ox-LDL–induced Lox-1 translocation. Total cell lysates or cell membrane of CHO cells and HCSMCs from each experiment were separated by 10% SDS-PAGE and transferred to nitrocellulose membranes. After incubation in blocking solution (5% skim milk, Becton Dickinson), membranes were incubated with primary antibody overnight at 4°C. Membranes were washed and incubated with secondary antibody for 1 hour and then detected with the enhanced chemiluminescence system (Amersham). The expression and cell-surface localization of hLox-1 were detected by anti-V5 antibody (1:1000, Invitrogen, 1155797) and anti-myc antibody (1:1000, Invitrogen, 500094), followed by goat anti-mouse IgG (H+L) horseradish peroxidase conjugate (1:1000, Bio-Rad). The immunoprecipitation samples were obtained using protein A/G PLUS-Agarose (Santa Cruz Biotechnology, H0604) and affinity-purified rabbit anti-mouse IgG (Jackson Immuno Research Laboratories, Inc, 52122). The activation of ERK1/2 was detected by ERK1/2 antibody (1:1000, Invitrogen, 9102) and phospho (p)-ERK1/2 antibody (1:500, Invitrogen, 9106S), followed by goat anti-mouse IgG (H+L) horseradish peroxidase conjugate or goat anti-rabbit IgG (H+L) alkaline phosphatase conjugate (1:1000, Bio-Rad), respectively. The bands were visualized using an LAS-3000 luminous image analyzer (Fuji Film). To quantify the densitometry of the bands, the films were scanned, and the density of each band was measured using Image Gauge version 4.0.
Immunofluorescent Staining and Confocal Microscopy
To analyze the expression levels and translocation of hLox-1 and hLox1-K266A/K267A in CHO cells, DiI-labeled Ox-LDL in WT and/or mutant hLox-1–transfected CHO cells was determined by confocal laser microscopy. Microscopy was performed using a Zeiss LSM410 microscope (Carl Zeiss) and a ×63 oil immersion objective. Images were acquired using excitation and emission wavelengths of 488 and 510 nm to 525 nm for EGFP or 543 and 590 nm for DsRed, respectively. Cells were incubated with 100 μg/mL of Ox-LDL at 37°C for 10 minutes or 10 mg/mL of DiI-labeled Ox-LDL at 37°C for 10 minutes. After 3 washes with ice-cold PBS, the cells were fixed with 1% paraformaldehyde for 20 minutes and 70% ethyl alcohol for 15 minutes on ice. The cells were stained with VECTASHIELD with 4′6-diamino-2-phenyl indole (Vector Laboratories).
[3H]-Thymidine was purchased from Amersham. DNA synthesis was evaluated by [3H]-thymidine incorporation. CHO cells were transiently cotransfected with pcDNA3 (0.3 μg) and hLox-1-WT-myc (1.2 μg) or hLox-1-WT-myc (0.3 μg) and hLox-1- K266A/K267A-myc (1.2 μg) in 24-well plates (104 cells per well). pcDNA3 (1.5 μg)-transfected cells were used as a control. Two days after transfection, the cells were incubated with 100 μg/mL Ox-LDL for 18 hours and analyzed with regard to [3H]-thymidine incorporation. Cells were labeled for 12 hours with [3H]thymidine (0.5 μCi/mL) before harvest and then washed 3 times with PBS. After the addition of perchloric acid, the acid-precipitable material was dissolved in 1 N NaOH and subjected to liquid scintillation counting.
The results are expressed as the mean±SEM. Significant differences in measured values were evaluated with the unpaired Student t test. Statistical significance was set at <0.05.
EGFP-hLox1 Expressed in the Cell Membrane
We established CHO cell lines that permanently expressed EGFP-hLox-1 (EGFP fused to the C-terminal side of hLox-1, EGFP-hLox1-CHO) and hLox1-EGFP (EGFP fused to the N-terminal side of hLox-1, hLox1-EGFP-CHO). It has been reported that the configuration of Lox-1 is a cap formation once it transfixes conformer membrane protein. There is an amino terminus as in perikaryon, and the C terminus is extracellular.1 In contrast to our expectation, we found that EGFP-hLox-1 but not hLox-1-EGFP was expressed in the membrane using confocal microscopy (Figure 1a). Because EGFP in the C terminus of hLox-1 was located on the exoplasmic side in the hLox1-EGFP-CHO cell, it may be difficult to cap the Lox-1 homodimer.
ERK1/2 Activity in HCSMC Via hLox-1-WT
Native LDL and LDL oxidized by the addition of CuSO4 stimulated ERK1/2 kinase in a time- and dose-dependent manner in baboon and rat VSMCs through Lox-1.3 As in bovine and rat, human Lox-1 is found in HCSMCs, and we analyzed whether Ox-LDL activated ERK1/2 through Lox-1 in HCSMCs. We used human Lox-1 blocking antibody, which was further selected to block Ox-LDL binding and uptake in Lox-1–expressing cells.18 Ox-LDL-induced ERK1/2 activation (100 μg/mL) was moderately blocked by human Lox-1 blocking antibody in HCSMCs (Figure 1b). Next, we used EGFP-hLox1-CHO to analyze the optimal conditions for ERK activation. EGFP-hLox-1–WT cells were incubated with Ox-LDL (0 to 200 μg/mL) for 0, 5, 10, 15, or 20 minutes to determine ERK1/2 activity. We found that the optimal condition for maximum ERK1/2 activation was incubation with 100 μg/mL Ox-LDL for 10 minutes, and these conditions were used in subsequent experiments (Figure 1c and 1d).
Homodimerization of hLox-1–WT Was Localized in the Cell Membrane
We cotransfected CHO cells with EGFP-hLox1–WT and DsRed-hLox1–WT to confirm the colocalization of these hLox-1–WT in the cell membrane. EGFP-hLox1–WT and DsRed-hLox1–WT were colocalized in the cell membrane in the presence or absence of Ox-LDL for 10 minutes under serum-free conditions using confocal microscopy (Figure 2a). In addition, immunoblotting of hLox-1–WT-V5 was performed. hLox-1–WT-V5 was homodimerized in the cell membrane in the presence or absence of Ox-LDL (Figure 2b). Ox-LDL may activate ERK1/2 without the translocation of Lox-1 after being pretreated with β-mercaptoethanol under nonreducing conditions, suggesting that disulfide bonding may be important for homodimerization. In addition, hLox-1–WT-V5 showed 2 monomer bands, which may be glycosylated and unglycosylated Lox-1.
Ox-LDL Did Not Activate ERK1/2 Via hLox-1-K266A/K267A
Next, we established hLox-1-K266A/K267A as a possible contributor to the dominant-negative effect for hLox-1–WT–induced ERK activation. When we cotransfected CHO cells with EGFP-hLox1-K266A/K267A and DsRed-hLox1-K266A/K267A, these hLox-1s were colocalized in the cell membrane similar to hLox-1–WT in the presence or absence of Ox-LDL for 10 minutes under serum-free conditions using confocal microscopy (Figure 2c). In addition, hLox-1-K266A/K267A was homodimerized in the cell membrane using immunoblotting (Figure 2d). Ox-LDL did not activate ERK1/2 by immunoblotting in EGFP-hLox-1-K266A/K267A CHO cells (Figure 2e). We confirmed that the 2 residues (Lys266 and Lys267) play a role in inducing ERK activation but not homodimerization.
Dominant-Negative hLox-1 Inhibits ERK1/2 Activation and [3H]-Thymidine Incorporation Through Heterodimerization
To analyze the inhibition of Ox-LDL–induced ERK1/2 activation through hLox-1–WT by hLox-1-K266A/K267A, we established an EGFP-hLox-1-K266A/K267A CHO cell line. hLox-1–WT-V5 transfected either the EYFP-CHO cell line or the EGFP-hLox-1-K266A/K267A CHO cell line. Ox-LDL–induced ERK1/2 activation was mostly blocked in the hLox-1–WT-V5–transfected EGFP-hLox-1- K266A/K267A CHO cell line (Figure 3a). Expression levels of hLox-1–WT-V5 in the EGFP or EGFP-hLox-1-K266A/K267A cell lines were similar (data not shown). Next, we analyzed the incorporation of [3H]-thymidine in CHO cells that had been cotransfected with hLox1–WT-myc and/or hLox1-K266A/K267A-myc (Figure 3b). Ox-LDL increased [3H]-thymidine incorporation in hLox1–WT-myc–expressing CHO cells by &3-fold, whereas this stimulation was significantly blocked in CHO cells that had been cotransfected with hLox1–WT-myc and hLox1-K266A/K267A-myc.
To check the localization of hLox-1–WT and hLox-1-K266A/K267A using confocal microscopy, we cotransfected CHO cells with DsRed-hLox1–WT and EGFP-hLox1- K266A/K267A. The DsRed-hLox1–WT and EGFP-hLox1-K266A/K267A were colocalized in the cell membrane before and after the addition of Ox-LDL for 10 minutes (Figure 3c). To confirm heterodimerization using immunoprecipitation, we cotransfected CHO cells with hLox-1–WT-V5 and either hLox1–WT-myc or hLox1-K266A/K267A-myc and incubated them with or without Ox-LDL for 10 minutes. The hLox-1–WT and hLox-1-K266A/K267A dimerized in the cell membrane (Figure 3d), suggesting that hLox-1-K266A/K267A had dominant-negative action against hLox-1–WT through heterodimerization.
Dominant-Negative hLox-1 Did Not Block DiI-Ox-LDL Binding in the Cell Membrane
To analyze the lack of hLox-1 function in ERK activation, we transfected CHO cells with or without hLox-1–WT or hLox-1-K266A/K267A. The binding of DiI-LDL or DiI-Ox-LDL to Lox-1 was analyzed using confocal microscopy (Figure 4). We found that DiI-Ox-LDL accumulated in the cell membrane equally in hLox-1–WT and hLox-1-K266A/K267A. In contrast, DiI-LDL did not accumulate in the cell membrane. A dominant-negative property of hLox-1-K266A/K267A plays a role without blocking Ox-LDL bound to the Lox-1.
Dominant-Negative hLox-1 Inhibits ERK1/2 Activation and [3H]-Thymidine Incorporation in HCSMCs
To confirm that hLox-1-K266A/K267A inhibits Ox-LDL–induced ERK1/2 activation and [3H]-thymidine incorporation in native cells, we used EGFP or hLox-1-K266A/K267A–transfected HCSMCs. Ox-LDL–induced ERK1/2 activation was mostly blocked in hLox-1-K266A/K267A–transfected HCSMCs (Figure 5a). In addition, Ox-LDL increased [3H]-thymidine incorporation &2.7-fold, whereas this stimulation was significantly blocked by hLox1-K266A/K267A (Figure 5b).
Members of the C-type lectin-like receptor family (Ly49A-J, NKR-P1, CD94/NKG2, and CD69) are homodimeric or heterodimeric type 2 transmembrane glycoproteins.19,20 Lox-1 belongs structurally to the C-type lectin-like family and may undergo dimerization. The present study was performed to examine the homodimerization of hLox-1 in the cell membrane.
There is no evidence that Lox-1 dimerizes, such as that CD94 and Ly49A exist as a disulfide-linked homodimer at the cell surface, as a prerequisite for its function. In this study, the coimmunoprecipitation of differentially tagged hLox-1 implied that there was a definite association between 2 hLox-1 molecules. Ox-LDL activated ERK1/2 via hLox-1 without the translocation of Lox-1, which suggests that hLox-1–WT exists in the cell membrane to form a homodimer independent of Ox-LDL. The homodimerization of hLox-1 in the cell membrane suggests a new alternative for the treatment of atherosclerosis at a receptor level using a dominant-negative strategy. Although antisense Lox-1 completely inhibited Ox-LDL–induced ERK1/2 activation,21 this method only temporarily inhibits ERK1/2 activity and is not suitable for inducing long-term treatment.
Previous studies have shown that the overexpression of Lox-1 induced the activation of p38 mitogen-activated protein kinase (MAPK) and oxidative stress in cardiac myocytes and the inhibition of p38 MAPK by cotransfection of a dominant negative, as well as by the administration of a specific inhibitor almost completely blocked the induction of apoptosis by Lox-1 activation.22 However, these methods are not specific for Lox-1, and various other signals remained. An inhibitory strategy for a dominant negative is a well-known method for MAPK, c-Jun N-terminal kinase (JNK),C and Ras.21,23–25 In particular, dominant-negative Ras21 has been well studied, because there is no depressant drug. However, no relevant study has used the dominant-negative method to inhibit the function of the natural killer cell receptor family. The importance of receptor dimerization was highlighted by the results of a crystal configuration analysis for a receptor, and this dominant-negative method has attracted attention. For example, it was reported that tubular atrophy and fibrosis after nephritic ischemia were significantly decreased in a transgenic mouse that showed a dominant-negative EGF receptor.26 In addition, Chen et al27 reported a dominant-negative α1-adrenergic receptor, which is a G protein–coupled receptor. Lee et al28 reported that dominant-negative estrogen receptor mutants have the potential to induce apoptosis of T47D cells and the regression of tumors. In the present study, for the first time, we established a method for the targeted inhibition of hLox-1 using a dominant-negative transgenic approach.
Using a dominant-negative transgenic approach, we showed that targeted inhibition of hLox-1 blocked ERK1/2 activation and [3H]-thymidine incorporation with Ox-LDL binding. To the best of our knowledge, the present findings constitute the first evidence that strategies aimed at blocking Lox-1–induced cell proliferative pathways at the receptor level could be useful for delaying the progression of arteriosclerosis and the occurrence of coronary artery disease in humans.
We are grateful for the gifts of DiI-LDL, DiI-Ox-LDL (Dr Hiroyuki Itabe), and CHO (ldlA7) cells (Dr Monty Krieger).
- Received April 24, 2006.
- Revision received May 4, 2006.
- Accepted May 17, 2006.
Kusuhara M, Chait A, Cader A, Berk BC. Oxidized LDL stimulates mitogen-activated protein kinases in smooth muscle cells and macrophages. Arterioscler Thromb Vasc Biol. 1997; 17: 141–148.
Shi X, Niimi S, Ohtani T, Machida S. Characterization of residues and sequences of the carbohydrate recognition domain required for cell surface localization and ligand binding of human lectin-like oxidized LDL receptor. J Cell Sci. 2001; 114: 1273–1282.
Chapman MJ, Laplaud PM, Luc G, Forgez P, Bruckert E, Goulinet S, Lagrange D. Further resolution of the low density lipoprotein spectrum in normal human plasma: physicochemical characteristics of discrete subspecies separated by density gradient ultracentrifugation. J Lipid Res. 1988; 29: 442–458.
Noble RP. Electrophoretic separation of plasma lipoproteins in agarose gel. J Lipid Res. 1968; 9: 693–700.
Stocks J, Miller NE. Capillary electrophoresis to monitor the oxidative modification of low density lipoproteins. J Lipid Res. 1998; 39: 1305–1309.
Walsh BA, Mullick AE, Walzem RL, Rutledge JC. 17beta-estradiol reduces tumor necrosis factor-alpha-mediated LDL accumulation in the artery wall. J Lipid Res. 1999; 40: 387–396.
Chen H, Li D, Saldeen T, Mehta JL. Transforming growth factor-beta(1) modulates oxidatively modified LDL-induced expression of adhesion molecules: role of LOX-1. Circ Res. 2001; 89: 1155–1160.
Kingsley DM, Krieger M. Receptor-mediated endocytosis of low density lipoprotein: somatic cell mutants define multiple genes required for expression of surface-receptor activity. Proc Natl Acad Sci U S A. 1984; 81: 5454–5458.
Li D, Singh RM, Liu L, Chen H, Singh BM, Kazzaz N, Mehta JL. Oxidized-LDL through LOX-1 increases the expression of angiotensin converting enzyme in human coronary artery endothelial cells. Cardiovasc Res. 2003; 57: 238–243.
Li D, Mehta JL. Antisense to LOX-1 inhibits oxidized LDL-mediated upregulation of monocyte chemoattractant protein-1 and monocyte adhesion to human coronary artery endothelial cells. Circulation. 2000; 101: 2889–2895.
Iwai-Kanai E, Hasegawa K, Sawamura T, Fujita M, Yanazume T, Toyokuni S, Adachi S, Kihara Y, Sasayama S. Activation of lectin-like oxidized low-density lipoprotein receptor-1 induces apoptosis in cultured neonatal rat cardiac myocytes. Circulation. 2001; 104: 2948–2954.
Szeberenyi J, Cai H, Cooper GM. Effect of a dominant inhibitory Ha-ras mutation on neuronal differentiation of PC12 cells. Mol Cell Biol. 1990; 10: 5324–5332.
Cai H, Szeberenyi J, Cooper GM. Effect of a dominant inhibitory Ha-ras mutation on mitogenic signal transduction in NIH 3T3 cells. Mol Cell Biol. 1990; 10: 5314–5323.