Elevated Cholesterol Levels in the Plasma Membranes of Macrophages Inhibit Migration by Disrupting RhoA Regulation
Objective— Atherogenesis begins as small subendothelial accumulations of foam cells that develop through unregulated uptake of modified and aggregated low-density lipoprotein (LDL). The reason why foam cells remain in the atherosclerotic plaque rather than migrating out of the area is unclear. We tested the hypothesis that elevated membrane cholesterol levels, which may result from interactions with aggregated LDL, affect macrophage migration.
Methods and Results— Cholesterol loading by incubation with cholesterol-chelated methyl-β-cyclodextrin decreased migration of J774A.1 macrophages toward complement 5a (C5a) in transwell migration assays, even though cholesterol-loaded macrophages responded to a bath application of C5a. In a micropipette polarization assay, cholesterol-loaded cells polarized toward a C5a gradient. In a transwell migration assay, cholesterol-loaded cells extended lamellae through the filter pores but were unable to translocate their cell bodies. Cholesterol loading decreased both the cellular levels of GTP-bound active RhoA and the phosphorylation of myosin light chain. Expression of constitutively active RhoA largely prevented the inhibition of cell migration by cholesterol loading.
Conclusions— These results suggest that increases in plasma membrane cholesterol content alter RhoA activation, resulting in inhibition of cell migration. These findings provide one possible explanation for the retention of foam cells in atherosclerotic lesions.
Atherogenesis is a complex process that is thought to begin when hypercholesterolemia leads to deposition of low-density lipoprotein (LDL) into the subendothelial space.1–3 Subendothelial LDL becomes oxidatively modified, aggregated, and then attached to the surrounding matrix by arterial wall cell enzymes.4,5 Macrophages take up cholesterol from the matrix-bound LDL aggregates and become lipid-filled foam cells.6 Gradual thickening of the arterial wall caused by the accumulation of foam cells leads to the development of an atherosclerotic plaque. It has been proposed that plaque progression may result in part from long-term retention of foam cells within the plaque,7 suggesting that promoting emigration of foam cells from atherosclerotic lesions might lead to plaque regression.
It is unclear why macrophage foam cells remain within the atherosclerotic lesion. A possible explanation is that the lesion environment, and specifically interaction with lipoproteins, changes the migratory behavior of macrophages. Several studies have investigated the possibility that macrophage motility is altered by interactions with modified LDLs. Oxidatively modified LDL strongly inhibited the chemotactic responses of mouse resident peritoneal macrophages,8,9 and oxidized or aggregated LDL changed the organization of the F-actin cytoskeleton and decreased the ability of macrophages to generate locomotor forces.10 However, the molecular mechanism(s) by which lipoprotein interactions caused these changes in macrophages is not understood.
The interactions of macrophages with aggregated and matrix-bound LDL have some unique features that are not apparent when macrophages bind soluble LDL. When macrophages begin to engulf aggregated matrix-bound LDL particles, they maintain prolonged contact with the LDL aggregates.11 During this prolonged contact there can be rapid transfer of cholesterol from the LDL aggregates to the macrophage. As a result, transient increases in the cholesterol content of macrophage plasma membranes may result. These findings suggest 2 possible ways that lipoprotein interactions could lead to changes in macrophage F-actin organization and migration: (1) The binding of modified lipoproteins to their receptors could initiate signals that negatively regulate the cytoskeletal changes required for migration; or (2) Cholesterol transfer from lipoproteins could modulate macrophage plasma membrane cholesterol levels, and this change in membrane cholesterol could alter signal transduction mechanisms in the macrophage. To directly test the latter possibility, in a previous study we raised macrophage membrane cholesterol levels acutely by treating cells with cholesterol-chelated methyl-β-cyclodextrin (chol-MβCD). We found that elevation of plasma membrane cholesterol levels with chol-MβCD changed F-actin organization and decreased macrophage motility, and these changes mimicked those obtained when cells were treated with soluble lipoproteins and an ACAT inhibitor, which cause cellular cholesterol levels to rise.12 In this study, we sought to identify the step at which the migration of cholesterol-loaded cells was inhibited and to investigate the molecular mechanisms of this inhibition.
The steps that make up the cell migration process can be summarized as follows: sensing the chemoattractant signal, lamellar protrusion toward the signal (ie, polarization), translocation forward (which requires contractile forces), and tail detachment and retraction. We tested the effects of cholesterol loading on each stage of migration and found that the translocation step (ie, contractility) was affected in cholesterol-loaded macrophages that had been stimulated to migrate through the filters of a transwell migration chamber. Arrest at this stage of migration appears to result from inhibition of the RhoA/Rho kinase signaling pathway, which regulates cell contraction.
Plasmids and Transfection
Plasmid DNA encoding EGFP-tagged wild-type RhoA (WT-RhoA, Addgene plasmid 12965) and EGFP-tagged constitutively active (Q63L) RhoA (CA-RhoA, Addgene plasmid 12968) were obtained from Addgene (Cambridge, Mass).13 Transient transfection of RAW 264.7 cells (RAW cells) was performed using FuGENE HD transfection reagent (Roche Diagnostics) according to the manufacturer’s instructions.
Transwell Migration Assay
J774A.1 macrophage-like cells (J774 cells) or transfected RAW cells were plated into transwell migration filter inserts seated in 24-well dishes (Corning). The cells were incubated with 5 mmol/L chol-MβCD (Sigma) for various times, and some samples were subsequently incubated with 5 mmol/L MβCD (Sigma) to reverse cholesterol loading. The samples were washed extensively to remove the cholesterol modulating agents before the filter inserts were placed either in wells containing 10 nmol/L recombinant human complement 5a (C5a; Sigma) and 0.5% heat-inactivated fetal bovine serum (FBS; Hyclone) in Dulbecco Modified Eagle Medium (DMEM; Mediatech) or in wells containing FBS and media but no C5a. After 3 hours incubation at 37°C in a CO2 incubator, the samples were processed for fluorescence microscopy imaging and analysis as described (please see supplemental materials, available online at http://atvb.ahajournals.org).
Micropipette Protrusion Assay
J774 cells were plated into Poly-d-lysine-coated glass coverslip-bottom dishes (coverslip dishes) and incubated overnight. The cells were then treated with or without 5 mmol/L chol-MβCD for 15 minutes in 20 mmol/L HEPES-buffered DMEM, washed, and mounted onto the stage of a Leica DMIRB inverted microscope equipped with a microinjection system (Narishige). A micropipette filled with 1 μmol/L C5a was positioned close to the cells by using a micromanipulator, and phase-contrast time-lapse images of the cells within the field were recorded at 30-second intervals using MetaMorph image acquisition software (Molecular Devices).
RhoA Activation Assay
J774 cells, RAW cells, or human monocyte-derived macrophages (hMDM) treated with or without chol-MβCD were stimulated with 0.5 nmol/L C5a or left unstimulated, then lysed with lysis buffer (50 mmol/L Tris, 10 mmol/L MgCl2, 150 mmol/L NaCl, 1% Triton X-100, pH 7.5) on ice and clarified by centrifugation; supernatants were used for the RhoA activation assay. The amount of GTP-bound RhoA was determined using the RhoA activation assay kit from Cytoskeleton, Inc.
J774 cells were plated onto 35-mm tissue culture dishes (BD Falcon) and incubated overnight. The cells were treated with or without 5 mmol/L chol-MβCD or 10 μmol/L Y-27632 (Sigma), washed, and lysed with Laemmli buffer. After boiling, the samples were subjected to SDS-PAGE followed by immunoblotting with primary anti-myosin light chain (MLC) or anti-phospho MLC (Thr18/Ser19) antibodies (Cell Signaling Technology) and secondary HRP-conjugated goat anti-rabbit IgG antibody (Cell Signaling Technology).
All data are expressed as mean±SD and are representative of 3 independent experiments, unless otherwise noted. For individual pair-wise comparisons, Student t test was used.
For detailed Materials and Methods, please see http://atvb. ahajournals.org.
Increasing Macrophage Membrane Cholesterol Levels Impaired Chemotaxis Toward C5a
To examine the effects of cholesterol loading in the absence of any contributions from lipoprotein receptor-mediated signaling, the cholesterol content of macrophages was acutely raised or lowered by treating cells with chol-MβCD or MβCD, respectively. Supplemental Figure I (available online at http://atvb.ahajournals.org) recapitulates and extends our earlier findings12 on the effects of these treatments on cellular cholesterol content and cell migration for J774 cells and human monocyte-derived macrophages. Briefly, cholesterol loading causes a 51% decrease in the migration of J774 cells, and this decrease is largely reversible when cholesterol levels are returned to normal.
Cholesterol-Loaded J774 Cells Retained the Ability to Respond to C5a and Polarize in the Direction of the Source of Chemoattractant
Control and cholesterol-loaded cells were stimulated with a bath application of C5a, and the resulting reorganization of F-actin was evaluated by confocal microscopy. Figure 1A shows that raising membrane cholesterol led to an increase in membrane ruffling and cell spreading even in the absence of exogenous stimulation, consistent with our previous findings.12 In response to C5a, control cells extend F-actin-rich lamellae. Because the concentration of C5a in this assay was equivalent in all directions, the cells often extended several lamellae, and the cells that polarized did so in random directions. When C5a was added to cholesterol-loaded cells, the cells spread more than untreated C5a-stimulated cells, and they put out many more F-actin-containing lamellae. Under these conditions, almost no polarized cells were found. Therefore, the impaired chemotaxis of cholesterol-loaded cells was not caused by an inability to respond to C5a.
To test the ability of the cells to sense a gradient of chemoattractant and extend a protrusion toward the chemoattractant source (ie, polarize), a micropipette polarization assay was used. A micropipette was filled with C5a and then placed in the proximity of cells within the field of view of the microscope. Time-lapse images of the field of cells were recorded. As shown in Figure 1B, control cells extended protrusions toward the micropipette tip emitting C5a. When the position of the micropipette was changed, the cells rapidly responded by changing the direction of their lamellae (see supplemental Movie I). Similarly, although cholesterol-loaded J774 cells generally had larger lamellipodia than untreated cells, the cholesterol-loaded cells still showed polarization toward the micropipette (Figure 1B) and responded to changes in the position of the chemoattractant source (see supplemental Movie II).
J774 Cells With Increased Membrane Cholesterol Levels Showed an Impaired Ability to Undergo Translocation
After extending a lamella toward the source of chemoattractant, cells complete migration by translocating their bulk forward and subsequently retracting their uropods or tails.14 The effect of cholesterol modulation on translocation and tail retraction was evaluated in the transwell migration model described for supplemental Figure IB. The positions and morphologies of C5a-stimulated J774 cells were evaluated by confocal imaging of fluorescently-labeled cell nuclei and F-actin cytoskeletons. As indicated by the F-actin staining, the number of protrusions extended by the cells was not altered by cholesterol modulation (Figure 1D, i, ii, v, vi). Quantification of these images (Figure 1E) revealed that 21% (s.d. 5%, n=3) of untreated cells had protrusions in the absence of C5a stimulation, and for cholesterol loaded cells the number of spontaneous protrusions was about the same (18%, s.d. 5%, n=3). In the presence of C5a, the fraction of both untreated and cholesterol-loaded cells with protrusions was increased to nearly identical levels (untreated: 82%, s.d. 9%; cholesterol-loaded: 81%, s.d. 5%). The lamellae extended by untreated and cholesterol-loaded cells in response to C5a were phenotypically distinct: cholesterol-loaded cells could extend very large pseudopodia through the pores without translocating their cell bodies (note absence of nuclei; arrows in Figure 1D, panel vi), but this was rarely seen for untreated cells. These results suggest that the defect in migration of cholesterol-loaded cells was attributable to a defect in the translocation step of migration.
Rho/Rho Kinase Signaling Pathway Was Inhibited by Cholesterol Loading
It has been shown that Rho family small GTPases are involved in migration, mainly through regulation of F-actin organization and actin-myosin contraction.15,16 Rac, Cdc42, and RhoA are well-characterized members of the Rho GTPase family, and it was shown that for migrating fibroblasts RhoA is necessary for cell translocation and tail retraction (see supplemental Figure II).17 We investigated the activation state of RhoA in J774 cells before and after cholesterol loading using a RhoA pull-down assay.18 As shown in Figure 2A, the amount of active (ie, GTP-bound) RhoA in resting J774 cells was decreased in cells with elevated membrane cholesterol. After 3 minutes of activation with a bath application of C5a, untreated cells showed approximately a 25% decrease in active RhoA, which returned close to the level of resting cells by 30 minutes after C5a addition. On the other hand, in cholesterol-loaded cells, the amount of active RhoA remained well below that of untreated cells even after C5a treatment, which caused a slight increase in active RhoA levels. Lowering the cellular cholesterol to control levels with MβCD reversed the inactivation of RhoA seen in cholesterol- loaded cells (supplemental Figure IIIA). The reduction in RhoA activity by cholesterol loading was also found in hMDM and RAW cells (supplemental Figure IIIB).
Inhibition of Rho Kinase Largely Recapitulates the Phenotype of Cholesterol-Loaded Cells
GTP-bound RhoA subsequently activates the serine/threonine kinase Rho kinase (supplemental Figure II),19–21 so inhibition of Rho kinase might mimic the phenotype elicited by RhoA inactivation through cholesterol loading. J774 cells were pretreated with Rho kinase inhibitor Y-27632, and F-actin organization of the cells before and after C5a stimulation was examined (Figure 2B). Y-27632 treatment on its own induced cells to form single or multiple F-actin-rich lamellipodia (Figure 2B, ii). C5a stimulation of the Y-27632–treated cells caused the cells to spread, and this spreading was greater than that of untreated C5a-stimulated cells (Figure 2B, iii, iv). Quantification of cell area revealed that C5a-stimulated Y-27632–treated cells spread to a similar extent as C5a-stimulated cholesterol-loaded cells (Figure 2C).
Transwell migration of Y-27632–treated J774 cells was decreased as compared with untreated cells (supplemental Figure IVA). Y-27632–treated cells had a number of large lamellipodia extending through the filter pores, but they showed reduced translocation across the pore because many of the nuclei remained above the filter (supplemental Figure IVB). However, the fraction of cells with protrusions was unchanged by Y-27632 treatment (supplemental Figure IVC). Similar phenotypes were observed in J774 cells in which RhoA was inhibited by C3 transferase treatment (supplemental Figure V).
Phosphorylation of MLC Was Decreased by Cholesterol Loading
Rho kinase phosphorylates MLC phosphatase and inhibits its activity, which results in an increase in MLC phosphorylation, and consequently an increase in actomyosin contractility.22 Because cholesterol-loaded cells exhibited a phenotype similar to cells in which Rho kinase was inhibited (ie, Y-27632–treated cells), we probed the phosphorylation status of MLC by Western blot. The phosphorylation of MLC was slightly but significantly decreased by cholesterol loading as shown in Figure 3. As a control, the phosphorylation of MLC in Y-27632–treated cells was measured and found to be decreased a similar amount as in cholesterol-loaded cells.
Expression of Constitutively Active RhoA Protects Macrophage Migration From Inhibition by Cholesterol Loading
RAW cells were used to test the effects of the expression of constitutively active RhoA (CA-RhoA) on migration of cholesterol-loaded cells. Like J774 cells, RAW cells showed inactivation of RhoA after cholesterol loading (supplemental Figure IIIB). As shown in Figure 4, cholesterol loading greatly inhibited the migration of untransfected RAW cells (light gray bars, 41% migration compared with untreated cells, s.d. 17%), but cholesterol loading was less effective at inhibiting the migration of cells expressing CA-RhoA (dark gray bar on right, 84% migration compared with untreated cells, s.d. 20%). Expression of WT-RhoA did not affect the inhibition of migration by cholesterol loading (Figure 4, dark gray bar on left), and neither transfection of WT-RhoA nor CA-RhoA affected the number of migrated cells with normal cholesterol levels (data not shown).
Previously, we showed that raising the cholesterol content of macrophage membranes either by chol-MβCD or by modified LDL (and ACAT inhibition) resulted in Rac-mediated membrane ruffling and increased macropinocytic activity in the absence of cell stimulation by chemoattractants.12 We also showed that chemoattractant-stimulated migration was decreased in cells with elevated cholesterol levels.12 In the present study, we have investigated the mechanism by which cholesterol loading decreases the C5a-stimulated migration of J774 macrophages. We show here that cholesterol loading inactivates RhoA in macrophages, thereby inhibiting migration of the cells attributable to impaired translocation/ contraction.
RhoA regulates contraction of migrating cells by activating Rho kinase, which in turn promotes MLC phosphorylation by blocking the activity of MLC phosphatase. The phosphorylation of MLC creates tension on actin filaments, which results in contractile force generation. Inhibition of myosin light chain kinase caused lamellar extension in all directions in neutrophils.23 This is similar to what is seen in macrophages with either cholesterol loading or Rho kinase inhibition, both of which caused a decrease in the levels of phosphorylated myosin light chain. We have demonstrated that cholesterol-loaded cells are phenotypically similar to Rho kinase inhibitor-treated cells; both cell populations exhibit increased cell spreading and a block in cell body translocation during transwell migration. In addition, we have shown that cholesterol loading leads to RhoA inactivation and dephosphorylation of MLC. Furthermore, the expression of constitutively active RhoA largely protected cells from the effects of cholesterol loading on migration. Based on these findings, we propose that impaired chemotaxis in cholesterol-loaded macrophages is, at least in part, a result of deficient RhoA/Rho kinase signaling. Because tail retraction is also an important step for cell migration and is regulated by the RhoA/Rho kinase signaling pathway,24 impaired tail retraction may also contribute to the inhibition of chemotaxis by cholesterol loading. Because the morphological changes in cholesterol-loaded cells were not identical to those in Y-27632–treated cells, it is likely that cholesterol loading may affect other regulators of cell migration, such as Rac.12 Another possible reason for reduced migration after cholesterol loading is an increase or decrease in cell adhesiveness. We found that cell adhesiveness, as measured by vigorous rinsing of cells, was neither increased nor decreased by cholesterol loading (supplemental Figure VI).
Lamellar extension is one of the first morphological changes induced by chemotactic stimulation, and it is thought to be positively regulated by the Rho-GTPases Rac and Cdc42, but negatively regulated by RhoA.25,26 Inactivation of the RhoA/Rho kinase signaling pathway caused THP-1 monocytes and human primary monocytes to extend protrusions, whereas unnecessary protrusions were prevented by active RhoA in unstimulated cells.26 C5a stimulation leads to a modest inhibition of RhoA activity (Figure 2A) yet a strong chemotaxis response (supplemental Figure IB). The current model for RhoA activation in a migrating cell calls for a local inhibition of RhoA activity within protruding lamellae but continued RhoA activation in the cell body and uropod. So, the apparently modest decrease in RhoA activation after C5a treatment may represent a large local (ie, leading edge) decrease in a subset of migrating cells. The morphological features of cholesterol-loaded cells can be explained by changes in the activation states of the Rho GTPases. In our previous report we showed that cholesterol loading increased the levels of GTP-bound active Rac and spontaneous membrane extensions were promoted.12 Here we found that a bath application of C5a led to enhanced spreading in cholesterol-loaded cells. Observed changes in morphology and Rho GTPase activation are summarized in Figure 4B. In the context of Rho GTPase regulation, membrane protrusion leading to increased spreading may be increased in these cells by both activation of Rac and inactivation of RhoA. When a gradient of chemoattractant is present, cholesterol-loaded cells could extend membrane protrusions toward the source in both 2-D and 3-D settings. The formation of especially large lamellipodia in response to chemotactic stimulation supports the idea that Rho may be misregulated in cholesterol-loaded cells. There is a negative interaction between Rac and Rho activation in neutrophils,27 and this might be related to our results in which these 2 GTPases are regulated in opposite directions by cholesterol.
In summary, our findings demonstrate that increases in plasma membrane cholesterol content interfere with RhoA activation and consequently inhibit macrophage migration attributable to reduced contractile force generation. These findings provide a possible explanation for the retention of foam cells in atherosclerotic lesions. Besides its role in the regulation of cell motility, RhoA can also regulate the cell cycle and apoptosis. It will be interesting to investigate the effects of cholesterol-induced Rho inactivation on these aspects of foam cell biology.
We thank Dr William Muller for helpful discussions.
Sources of Funding
This work was supported by National Institutes of Health grant P01-HL072942.
Original received November 21, 2006; final version accepted April 11, 2007.
Williams KJ, Tabas I. The response-to-retention hypothesis of early atherogenesis. Arterioscler Thromb Vasc Biol. 1995; 15: 551–561.
Schissel SL, Tweedie-Hardman J, Rapp JH, Graham G, Williams KJ, Tabas I. Rabbit aorta and human atherosclerotic lesions hydrolyze the sphingomyelin of retained low-density lipoprotein. Proposed role for arterial-wall sphingomyelinase in subendothelial retention and aggregation of atherogenic lipoproteins. J Clin Invest. 1996; 98: 1455–1464.
Oorni K, Pentikainen MO, Ala-Korpela M, Kovanen PT. Aggregation, fusion, and vesicle formation of modified low density lipoprotein particles: molecular mechanisms and effects on matrix interactions. J Lipid Res. 2000; 41: 1703–1714.
Osterud B, Bjorklid E. Role of monocytes in atherogenesis. Physiol Rev. 2003; 83: 1069–1112.
Llodra J, Angeli V, Liu J, Trogan E, Fisher EA, Randolph GJ. Emigration of monocyte-derived cells from atherosclerotic lesions characterizes regressive, but not progressive, plaques. Proc Natl Acad Sci U S A. 2004; 101: 11779–11784.
Quinn MT, Parthasarathy S, Steinberg D. Endothelial cell-derived chemotactic activity for mouse peritoneal macrophages and the effects of modified forms of low density lipoprotein. Proc Natl Acad Sci U S A. 1985; 82: 5949–5953.
Quinn MT, Parthasarathy S, Fong LG, Steinberg D. Oxidatively modified low density lipoproteins: a potential role in recruitment and retention of monocyte/macrophages during atherogenesis. Proc Natl Acad Sci U S A. 1987; 84: 2995–2998.
Zerbinatti CV, Gore RW. Uptake of modified low-density lipoproteins alters actin distribution and locomotor forces in macrophages. Am J Physiol Cell Physiol. 2003; 284: C555–C561.
Buton X, Mamdouh Z, Ghosh R, Du H, Kuriakose G, Beatini N, Grabowski GA, Maxfield FR, Tabas I. Unique cellular events occurring during the initial interaction of macrophages with matrix-retained or methylated aggregated low density lipoprotein (LDL). Prolonged cell-surface contact during which ldl-cholesteryl ester hydrolysis exceeds ldl protein degradation. J Biol Chem. 1999; 274: 32112–32121.
Qin C, Nagao T, Grosheva I, Maxfield FR, Pierini LM. Elevated plasma membrane cholesterol content alters macrophage signaling and function. Arterioscler Thromb Vasc Biol. 2006; 26: 372–378.
Subauste MC, Von Herrath M, Benard V, Chamberlain CE, Chuang TH, Chu K, Bokoch GM, Hahn KM. Rho family proteins modulate rapid apoptosis induced by cytotoxic T lymphocytes and Fas. J Biol Chem. 2000; 275: 9725–9733.
Ridley AJ, Schwartz MA, Burridge K, Firtel RA, Ginsberg MH, Borisy G, Parsons JT, Horwitz AR. Cell migration: integrating signals from front to back. Science. 2003; 302: 1704–1709.
Ren XD, Kiosses WB, Schwartz MA. Regulation of the small GTP-binding protein Rho by cell adhesion and the cytoskeleton. Embo J. 1999; 18: 578–585.
Leung T, Manser E, Tan L, Lim L. A novel serine/threonine kinase binding the Ras-related RhoA GTPase which translocates the kinase to peripheral membranes. J Biol Chem. 1995; 270: 29051–29054.
Kimura K, Ito M, Amano M, Chihara K, Fukata Y, Nakafuku M, Yamamori B, Feng J, Nakano T, Okawa K, Iwamatsu A, Kaibuchi K. Regulation of myosin phosphatase by Rho and Rho-associated kinase (Rho-kinase). Science. 1996; 273: 245–248.
Eddy RJ, Pierini LM, Maxfield FR. Microtubule asymmetry during neutrophil polarization and migration. Mol Biol Cell. 2002; 13: 4470–4483.
Worthylake RA, Lemoine S, Watson JM, Burridge K. RhoA is required for monocyte tail retraction during transendothelial migration. J Cell Biol. 2001; 154: 147–160.
Worthylake RA, Burridge K. RhoA and ROCK promote migration by limiting membrane protrusions. J Biol Chem. 2003; 278: 13578–13584.
Sander EE, ten Klooster JP, van Delft S, van der Kammen RA, Collard JG. Rac downregulates Rho activity: reciprocal balance between both GTPases determines cellular morphology and migratory behavior. J Cell Biol. 1999; 147: 1009–1022.