Mildly Oxidized Low Density Lipoprotein Induces Contraction of Human Endothelial Cells through Activation of Rho/Rho Kinase and Inhibition of Myosin Light Chain Phosphatase*

Mildly oxidized low density lipoprotein (mox-LDL) is critically involved in the early atherogenic responses of the endothelium and increases endothelial permeability through an unknown signal pathway. Here we show that (i) exposure of confluent human endothelial cells (HUVEC) to mox-LDL but not to native LDL induces the formation of actin stress fibers and intercellular gaps within minutes, leading to an increase in endothelial permeability; (ii) mox-LDL induces a transient decrease in myosin light chain (MLC) phosphatase that is paralleled by an increase in MLC phosphorylation; (iii) phosphorylated MLC stimulated by mox-LDL is incorporated into stress fibers; (iv) cytoskeletal rearrangements and MLC phosphorylation are inhibited by C3 transferase from Clostridium botulinum, a specific Rho inhibitor, and Y-27632, an inhibitor of Rho kinase; and (v) mox-LDL does not increase intracellular Ca2+concentration. Our data indicate that mox-LDL induces endothelial cell contraction through activation of Rho and its effector Rho kinase which inhibits MLC phosphatase and phosphorylates MLC. We suggest that inhibition of this novel cell signaling pathway of mox-LDL could be relevant for the prevention of atherosclerosis.

The response to injury hypothesis of atherosclerosis proposes that the first step in atherogenesis is an endothelial dysfunction induced by stimuli such as mildly oxidized LDL (mox-LDL) 1 (1). Vascular endothelium exposed to oxidized LDL in vitro and in vivo shows an increased permeability (2,3), a hallmark of endothelial dysfunction. The signal pathways by which mox-LDL and oxidized LDL reduce endothelial barrier function are unknown. Confluent endothelial cells challenged with vasoactive substances, such as thrombin, show a rapid reorganization of their actin cytoskeleton, i.e. stimulation of myosin light chain (MLC) phosphorylation and actin stress fiber formation, leading to cell contraction and intercellular gaps (4,5). Such cytoskeletal rearrangements reduce endothelial barrier function (6) and are typically found in vivo in endothelium overlying early atherosclerotic lesions (7).
Thrombin induces endothelial cell contraction and increased endothelial permeability through binding to a serpentine receptor which is coupled via heterotrimeric G-proteins of the Gq family to phospholipase C␤ to yield inositol 1,4,5-trisphosphate, which mobilizes Ca 2ϩ from intracellular stores and thus increases cytosolic Ca 2ϩ concentration (8). This increase in intracellular Ca 2ϩ concentration leads to activation of Ca 2ϩ / calmodulin-dependent myosin light chain kinase (MLCK), which phosphorylates Thr-18 and Ser-19 of the MLC of myosin II to enable actin-myosin interaction and cell contraction (9). We and others recently found an additional pathway by which thrombin regulates endothelial cell contraction. Thrombin, probably through coupling of the thrombin receptor to heterotrimeric G-proteins of the G 12/13 family (10), was shown to activate the Ras-related GTPase Rho and its effector p160 Rho kinase in endothelial cells (5). Rho kinase phosphorylates myosin binding subunit of MLC phosphatase which is thereby inactivated (11)(12)(13). Indeed, we observed that thrombin transiently inactivated MLC phosphatase in human endothelial cells and that phosphatase inactivation correlated with peak levels in MLC phosphorylation. Hence, the Rho/Rho kinase pathway seems to regulate endothelial contractility in concert with Ca 2ϩ /calmodulin-dependent MLCK. A similar regulatory system was reported in smooth muscle cells (14 -16).
This study was aimed to investigate whether mox-LDL would mimic the action of vasoactive substances on human endothelial cells. We found that (i) mildly oxidized LDL induced a rapid reorganization of the actin cytoskeleton, the formation of intercellular gaps, and the stimulation of MLC phosphorylation within minutes and (ii) that mox-LDL induced these changes through stimulation of the Rho/Rho kinase pathway with the subsequent inhibition of MLC phosphatase and without increasing the cytosolic Ca 2ϩ concentration.
Cell Culture-Human umbilical vein endothelial cells (HUVEC) were obtained and cultured as described previously (5). Before exposure to mox-LDL or native LDL, cells were serum deprived for 15 min. Medium was then replaced by fresh serum-free medium containing mox-LDL or native LDL as indicated.
Immunofluorescence-HUVEC were stained for I-actin as described previously (5). For staining of phosphorylated MLC, cells were fixed in 3.7% formaldehyde solution in PBS and permeabilized for 5 min with 0.1% Triton X-100. Cells were treated with normal goat serum (1/10 in PBS) for 20 min and then incubated for 1 h with anti-phospho-MLC antibody, diluted 1/10 in PBS, followed by 1 h incubation with rhodamine-labeled goat anti-rabbit IgG antibody (Dianova, Hamburg, Germany), diluted 1/200 in PBS.
Myosin Light Chain Phosphorylation-MLC phosphorylation was analyzed by SDS-PAGE and Western blotting. HUVEC were grown for 10 days in Costar 6-well culture dishes and treated with mox-LDL as indicated. Cells were lysed with boiling Laemmli buffer. Proteins were then electroblotted onto polyvinylidene difluoride membranes. Membranes were incubated overnight with anti-phospho-MLC antibody (1/ 500) in Tris-buffered saline containing 0.05% Tween 20, washed three times, incubated for 1 h with horseradish peroxidase-labeled goat antirabbit IgG antibody (Amersham Pharmacia Biotech), 1/7500 in Trisbuffered saline, washed three times, and then developed with Luminol solution (Pierce) and exposed to Kodak X-Omat films. The amount of phosphorylated MLC was determined by densitometric analysis of the protein bands that reacted with anti-phospho-MLC antibody, using a Sharp XL-325 densitometer and Amersham Pharmacia Biotech Image Master software. Phosphorylated MLC appeared as a single band of 20 kDa. This 20-kDa band was also detected by a mouse monoclonal antibody to total MLC (Sigma, Deisenhofen, Germany). The lower band is because of an unspecific reaction of the secondary anti-rabbit antibody. The range of linearity of the densitometric measurements of phosphorylated myosin was determined by a dilution series of protein samples of mox-LDL-stimulated HUVEC (4 -40 g of protein/lane).
Preparation of Mox-LDL-LDL was isolated in the continuous presence of EDTA (18). LDL was dialyzed at 4°C using a N 2 -saturated buffer (18) containing EDTA (1 mM) and then stored at 4°C in darkness under N 2 . Mox-LDL was prepared from EDTA-free LDL by Cu 2ϩ -triggered oxidation as described (18). Briefly, freshly dialyzed LDL containing 1 mmol/liter EDTA was loaded onto Econo-PacI0 DG columns (Bio-Rad) and recovered in PBS. LDL was then concentrated to a final concentration of 20 mg/ml by centricon-100 concentrators (Amicon) and mildly oxidized by incubation with CuSO 4 (final concentration 640 mol/liter) for 20 h at 37°C. The endotoxin concentrations in mox-LDL and native LDL were determined by a Limulus assay and were 5 ng/ml of medium for native LDL and 3.7 ng/ml of medium for mox-LDL. These endotoxin concentrations have no effects on MLC phosphorylation or the actin cytoskeleton of HUVEC within minutes (data not shown).
Ca 2ϩ Measurement-Calcium measurements and calculation of cytosolic Ca 2ϩ -concentrations of individual cells were carried out with a Quanticell fluorimetric system (Applied Imaging Corp., Sunderland, UK) as described (19).
Measurement of Endothelial Permeability-Horseradish peroxidase (HRP) diffusion through HUVEC monolayers was determined as described previously with some modifications (5). HUVEC were cultured for 10 days on collagen-coated cell culture inserts (3-m pore size, Becton Dickinson) with medium changes every 2 days. For exposure to mox-LDL or native LDL, medium was replaced with 500 l of culture medium containing mox-LDL or native LDL. For controls mox-LDL was omitted, but otherwise cells were treated identically. After 2 min of stimulation, 500 l of medium was filled into the lower compartment, and the medium in the upper compartment was replaced with fresh medium containing HRP (0.34 mg/ml, IV-A-type, M r 44,000, Sigma, Deisenhofen, Germany). After 1 min, 60 l of medium was collected from the lower compartment and mixed with 860 l of reaction buffer (50 mM NaH 2 PO 4 , 5 mM Guaiacol) and 100 l of freshly made H 2 O 2 solution (0.6 mM in H 2 O). The reaction was allowed to proceed for 15 min at room temperature and absorbance was measured at 470 nm. Fig. 1, exposure of confluent endothelial monolayers to mox-LDL (250 g/ml) for 2 min changed their cobblestone morphology with a polygonal cell shape and a dense peripheral actin ring to a contracted phenotype with rounded cells, prominent actin stress fibers, and intercellular gaps (Fig. 1b). To test whether mox-LDL induces actin stress fibers and cell contraction via Rho and Rho kinase, we stained control cells or cells stimulated with mox-LDL for F-actin using rhodamine-phalloidin. As indicated by Fig. 1a, control cells showed a dense peripheral actin ring and almost no stress fibers. Endothelial cells stimulated with mox-LDL (2 min, 250 g/ml) showed prominent actin stress fibers and a contracted phenotype with intercellular gaps (Fig. 1b). To investigate whether these actin rearrangements are because of activation of Rho/Rho kinase, we pretreated cells with the selective Rhoinhibitor C3 transferase from Clostridium botulinum (24 h, 5 g/ml) or the Rho kinase inhibitor Y-27632 (30 min, 10 M) and then stimulated the cells with mox-LDL. C3 transferase (Fig.  1c) or Y-27632 (Fig. 1e) did not influence the actin distribution in unstimulated HUVEC, but C3 transferase (Fig. 1d) and Y-27632 (Fig. 1f) completely blocked the mox-LDL-induced actin stress fiber formation and cell contraction, indicating that mox-LDL induced these cytoskeletal rearrangements through activation of Rho and Rho kinase. In contrast to mox-LDL, native LDL had no effect on the actin cytoskeleton in human endothelial cells (Fig. 1g).

Mox-LDL Stimulates Actin Stress Fiber Formation via Rho/ Rho Kinase-As shown in
Mox-LDL Induces an Increase in Endothelial Permeability That Is Mediated by Rho/Rho Kinase-Endothelial cell con-
Mox-LDL Inactivates MLC Phosphatase and Enhances MLC Phosphorylation in Human Endothelial Cells-Phosphorylation of the light chain enables myosin to interact with and to slide against F-actin filaments in stress fibers, which are con-

FIG. 4. Mox-LDL-induced MLC phosphorylation is mediated by Rho/Rho kinase.
HUVEC were stimulated with mox-LDL (250 g/ml) as indicated. MLC phosphorylation was then determined by Western blot using an antibody to phosphorylated MLC. The lower molecular weight band is because of an unspecific cross-reaction with the secondary antibody. In cells treated with C3 transferase (24 h, 5 g/ml) or with Y-27632 (Y) (30 min, 10 M), mox-LDL-induced (2 min, 250 g/ml) MLC phosphorylation was completely blocked. A Western blot representative of three similar experiments is shown.
FIG. 6. Mox-LDL does not increase cytosolic Ca 2؉ concentration. HUVEC were loaded with Fura 2-AM and stimulated with mox-LDL (150 g/ml). Cytosolic Ca 2ϩ concentration was then traced in 38 single cells. mox-LDL yielded no increased intracellular Ca 2ϩ -levels, whereas cells responded well to thrombin (1 unit/ml).

FIG. 2. Mox-LDL increases transendothelial diffusion of HRP
in a Rho/Rho kinase-dependent manner. HUVEC were incubated without or with 5 g/ml C3 transferase (C3) from C. botulinum for 24 h or for 30 min with 10 M Rho kinase inhibitor Y-27632 and exposed to mox-LDL or native LDL (2 min, 250 g/ml). Transendothelial diffusion of HRP was determined spectrophotometrically as described under "Experimental Procedures." Each bar represents the mean Ϯ S.E. of 5-8 measurements. sidered the contractile organelles of non-muscle cells such as endothelial cells (9). If in fact mox-LDL stimulates Rho/Rho kinase, this should lead to inhibition of MLC phosphatase followed by MLC phosphorylation (5). Indeed, MLC phosphatase activity was found to be rapidly reduced to about 60% of the basal activity between 30 s and 2 min of mox-LDL treatment, which was followed by a return to base-line values after 5 min (Fig. 3a). To determine the corresponding levels of phosphorylated MLC, we performed Western blots using a phospho-MLC specific antibody. Fig. 3b shows a time course of MLC phosphorylation in endothelial cells stimulated with mox-LDL. MLC phosphorylation rose to a peak within 2 min of stimulation and then fell to base-line levels within 15 min. In contrast to mox-LDL, native LDL did not induce MLC phosphorylation in HUVEC. To demonstrate that the mox-LDL-induced MLC phosphorylation is because of Rho/Rho kinase activation, we pretreated HUVEC with C3 transferase (24 h, 5 g/ml) or Y-27632 (30 min, 10 M). The results presented in Fig. 4 demonstrate that inhibition of Rho or Rho kinase completely blocked mox-LDL-induced MLC phosphorylation. To investigate the intracellular distribution of phosphorylated MLC in HUVEC after mox-LDL stimulation, cells were stained for phosphorylated MLC. Fig. 5a indicates that in control cells almost no phosphorylated MLC can be detected. In HUVEC treated with mox-LDL (2 min, 250 g/ml), phosphorylated MLC could be detected that was organized in stress fibers (Fig.  5b). In cells pretreated with C3 transferase (Fig. 5c) or with Y-27632 (Fig. 5d), no incorporation of phosphorylated MLC into stress fibers could be observed. Incubation of cells with native LDL did not induce myosin phosphorylation (Fig. 5e).
Mox-LDL Does Not Increase Cytosolic Ca 2ϩ Concentration-To study whether an increase in cytosolic Ca 2ϩ concentration and subsequent MLCK activation also contributes to the mox-LDL-induced MLC phosphorylation, we measured cytosolic Ca 2ϩ concentrations. Fig. 6 indicates that mox-LDL does not increase cytosolic Ca 2ϩ concentration in HUVEC indicating that it enhances MLC phosphorylation in HUVEC mainly through the Rho/Rho kinase pathway. DISCUSSION Here we describe a novel signal pathway of mox-LDL in human endothelial cells. Mox-LDL induces cell contraction, i.e. formation of actin stress fibers and intercellular gaps, leading to an increase in endothelial permeability, through activation of the Rho/Rho kinase pathway. This mechanism potentially contributes to endothelial dysfunction in early atherosclerotic lesions. Down-regulation of MLC phosphatase could also be a mechanism by which mox-LDL enhances the effects of Ca 2ϩmobilizing vasoactive substances. Conceivably, continuous stimulation of the vascular endothelium by subendothelial mox-LDL may eventually lead to an enhanced susceptibility of the endothelium to such vasoactive substances.
Our data indicate that mox-LDL utilizes physiologic signal mechanisms to activate the endothelium. Although we have not determined in the present study the mox-LDL-stimulated signal pathways upstream of Rho, we suggest that mox-LDL stimulates the Rho/Rho kinase pathway in endothelial cells through activation of heterotrimeric G-protein-coupled receptors such as the lysophosphatidic acid (LPA) receptor. We have indeed recently found that LPA is formed during mild oxidation of LDL in the absence of cells and that LPA was the main ingredient in mox-LDL responsible for the induction of platelet shape change and endothelial cell activation (21). In fibro-blasts, LPA induces Rho-dependent stress fiber formation through activation of G 13 (22) and enhances contractility by this way (23).
Activation of Rho and Rho kinase by mox-LDL may not be restricted to endothelial cells. It is likely that mox-LDL activates the same pathway in smooth muscle cells and fibroblasts. Indeed, it is well documented that Rho kinase plays a pivotal role for Ca 2ϩ sensitization of vascular smooth muscle cells. Furthermore mox-LDL was found to induce vascular smooth muscle cell contraction (24). Hence mox-LDL could contribute to elevated levels of blood pressure by activation of Rho kinase and thereby enhanced contractility of smooth muscle cells. Indeed, the Rho kinase inhibitor Y-27632 has previously been found to reduce the blood pressure in hypertensive rats (25).
In conclusion, we propose that the Rho/Rho kinase pathway is critical for the activation of endothelial cells by mox-LDL and could be a new target to prevent atherogenesis and cardiovascular disease.