Aggregated LDL in Contact With Macrophages Induces Local Increases in Free Cholesterol Levels That Regulate Local Actin Polymerization
Objective— Interaction of macrophages with aggregated matrix-anchored lipoprotein deposits is an important initial step in atherogenesis. Aggregated lipoproteins require different cellular uptake processes than those used for endocytosis of monomeric lipoproteins. In this study, we tested the hypothesis that engagement of aggregated LDL (agLDL) by macrophages could lead to local increases in free cholesterol levels and that these increases in free cholesterol regulate signals that control cellular actin.
Methods and Results— AgLDL resides for prolonged periods in surface-connected compartments. Although agLDL is still extracellular, we demonstrate that an increase in free cholesterol occurs at sites of contact between agLDL and cells because of hydrolysis of agLDL-derived cholesteryl ester. This increase in free cholesterol causes enhanced actin polymerization around the agLDL. Inhibition of cholesteryl ester hydrolysis results in decreased actin polymerization.
Conclusions— We describe a novel process that occurs during agLDL–macrophage interactions in which local release of free cholesterol causes local actin polymerization, promoting a pathological positive feedback loop for increased catabolism of agLDL and eventual foam cell formation.
A key step in the progression of atherosclerotic lesions is formation of lipid-loaded macrophage foam cells.1 Soluble lipoproteins accumulate at sites of atherosclerotic plaque formation where they undergo modification, aggregation, and anchoring to the extracellular matrix.2–4 Monocytes migrate from the blood, differentiate into macrophages that degrade LDL, and become filled with re-esterified cholesterol droplets.5,6 These cholesteryl ester (CE)-filled foam cells acquire new biological properties, particularly loss of motility7; secretion of cytokines,8 growth factors,6 and proteases9; and induction of apoptosis.10,11 These processes contribute to early lesion growth and late complications leading to plaque rupture.5,12
It has been proposed that aggregated lipoproteins (agLDL) tightly linked to the extracellular matrix play an important role in the development of atherosclerotic lesions.4 The interaction of macrophages with retained and aggregated lipoproteins differs significantly from the uptake of monomeric lipoproteins. For example, aggregates and associated extracellular matrix components are too large to be taken up by endocytosis or even by phagocytosis without breaking the aggregates or matrix into smaller pieces. The breakdown of the lipoproteins requires the actin cytoskeleton and activation of the Rho-family GTPases Rac1 or Cdc42.13
Furthermore, for 1 to 2 hours after initial contact of a macrophage with a retained and aggregated LDL particle the agLDL remains topologically outside the macrophage even though it may be in deep plasma membrane invaginations.14,15 Interestingly, there is significant hydrolysis of CE while the agLDL remain extracellular, and this hydrolysis requires lysosomal acid lipase (LAL).14 To explain this, it has been suggested that selective CE uptake might deliver the CE to lysosomes, but such a process has not been observed. An alternative mechanism is that hydrolysis of CE in agLDL occurs extracellularly. Various types of macrophages and macrophage-related osteoclasts are able to form extracellular lytic compartments.16–18 Consistent with this, we have observed that macrophages can create extracellular acidic compartments on contact with agLDL, and that they can secrete lysosomal contents into these contact zones (Haka AS, Grosheva I, Chiang E, Buxbaum AR, Baird BA, Pierini LM, Maxfield FR, unpublished data, 2009).
Herein, we test the hypothesis that macrophage interactions with agLDL lead to extracellular CE hydrolysis and a consequent increase in free cholesterol (FC) which induces local actin polymerization.
Materials and Methods
A complete description of Materials and Methods is supplied in the supplemental material (available online at http://atvb.ahajournals.org), including supplemental Figures I and II.
Interaction of agLDL With Macrophages and Fluorescent Labeling of Cells
LDL was aggregated by vortexing, and aggregates were centrifuged and resuspended in serum-free DMEM/HEPES. Approximately 50 μg/mL of agLDL was added to cells at 37°C.
Fluorescent Labeling of Cells
F-actin was labeled with Alexa488-phalloidin, FC was labeled with filipin, and the plasma membrane was labeled with Alexa488-cholera toxin subunit B (CtB).
Microscopy and Image Quantification
A Zeiss LSM510 laser scanning confocal microscope and a Leica DMIRB epi-fluorescence microscope were used for fluorescence imaging. Image analysis is described in the supplemental materials.
Transfection of siRNA
RNAi oligonucleotides (Dharmacon) were transfected using Lipofectamine for mouse peritoneal macrophages and HiPerFect for RAW 264.7 cells.
Recombinant Pichia-derived Human LAL (phLAL) was added to RAW 264.7 cells as described.21
Hydrolysis of Cholesteryl-[4-14C]-Oleate-Containing LDL
LDL reconstituted with cholesteryl-[4-14C]-oleate was vortex-aggregated and incubated with cells in the presence of an acyl-CoA:cholesterol acyltransferase (ACAT) inhibitor followed by treatment with plasmin to release extracellular agLDL.22 Lipids were extracted, and radiolabeled FC and CE were measured by thin layer chromatography (TLC).
Extracellular agLDL in Surface-Connected Compartments Leads to the Formation of F-Actin–Rich Structures
We examined the formation of F-actin in cells near sites of contact with agLDL. J774 cells were incubated with suspensions of Alexa546-agLDL followed by fixation and labeling of F-actin by Alexa488-phalloidin (Figure 1A and 1B). After 30 minutes, an enrichment of F-actin was detected near the sites of contact with agLDL (arrowheads, Figure 1A). An enlarged view (Figure 1A, inset) shows that F-actin (green) is closely associated with the agLDL (red), indicating that actin polymerization is enhanced in the immediate vicinity of the contact with agLDL. Approximately half of the aggregates touching cells had F-actin enriched structures in their immediate vicinity within 30 minutes, and the fraction of agLDL surrounded by F-actin increased further at later times (not shown). Macrophage-like RAW cells, mouse peritoneal macrophages, and mouse bone marrow–derived macrophages all exhibit similar responses to agLDL (supplemental Figure III). To investigate a more physiological model, J774 cells were plated on top of agLDL bound to a smooth-muscle cell matrix. Similar F-actin structures near contact sites were observed (supplemental Figure III). The interaction with agLDL did not induce apoptosis (assessed by annexin V binding) or cell permeabilization (assessed by propidium iodide staining).
Under the conditions used in these experiments, the sites of contact between cells and agLDL were surface-connected membrane invaginations (also called surface-connected compartments or SCCs), similar to those described previously.14,15 Cells were incubated with Alexa546-agLDL (red, Figure 1C) at 37°C and then labeled on ice with Alexa488-CtB (green, Figure 1C) to label the plasma membrane. Because the cells are not fixed or permeabilized, the macromolecular cholera toxin can only label glycolipids on the surface of the cells. Sites of contact with agLDL were labeled by CtB (Figure 1C, arrowhead), indicating that they are connected to the cell surface. An axial slice through the confocal stack at the position of the green line shows that the part of the aggregate resembling a separate vesicle in the projection image is actually a deep SCC (Figure 1C′, arrowhead). After a 2-hour incubation of J774 cells with agLDL, some SCCs containing agLDL can still be observed, but Alexa546-agLDL can also be seen in organelles that are not connected to the surface (supplemental Figure IV). Individual SCCs could persist for more than 30 minutes as confirmed by live-cell observation (supplemental Figure V).
FC Is Enriched at the Site of Contact Between agLDL and Macrophages
In a previous study it was shown that there was significant CE hydrolysis while the agLDL particles were still outside the macrophage.14 This hydrolysis was not seen on incubation with macrophages lacking LAL. It was hypothesized that the hydrolysis was a consequence of selective uptake of CE into cells and hydrolysis in lysosomes. To determine whether CE hydrolysis might actually take place in SCCs, we used filipin, a fluorescent sterol-binding polyene that can be used for quantitative measurement of FC,23 to detect FC.
Cells were incubated with Alexa546-agLDL, fixed, and labeled with filipin and Alexa488-phalloidin (Figure 2). Cells incubated with agLDL on ice for 30 minutes (Figure 2A and 2D) showed weak filipin staining in the region contacting agLDL (arrowheads, Figure 2A and 2D), and there was no increased F-actin near the agLDL (Figure 2D). When cells were incubated with agLDL at 37°C, there was intense filipin staining near the agLDL (arrows, Figure 2B and 2C), and these regions were enriched in F-actin (Figure 2E and 2F). The bright filipin labeling indicates there is an increase in FC at sites of contact between agLDL and the cell. The agLDL itself contains low amounts of FC as shown by the weaker intensity of filipin labeling of regions of agLDL that are not surrounded by F-actin (arrowheads, Figure 2A through 2F).
We quantified the average filipin intensity in pixels that contained agLDL. The filipin intensity for the cells incubated on ice was similar to the filipin intensity in aggregates that did not touch cells. After 30 minutes at 37°C, the filipin labeling in pixels that contained agLDL in contact with cells was significantly higher than for cells incubated on ice, and the filipin intensity remained elevated after a 1-hour incubation (Figure 2G). These data indicate that at least a portion of the CE hydrolysis occurs while agLDL is outside the cell.14 After a 2-hour incubation, increased intracellular filipin labeling is seen, and an increase in staining of lipid droplets by Nile Red is observed (not shown). Using biochemical assays, it was reported that about 10% of CE associated with agLDL enmeshed in extracellular matrix is hydrolyzed by J774 cells within 40 minutes, a time when most of the agLDL is still extracellular.14
Under the conditions of our experiments, the majority of cell-associated cholesterol transferred from agLDL is FC. We incubated cells with cholesteryl-[4-14C]-oleate reconstituted agLDL. After 90 minutes, extracellular labeled agLDL was removed by plasmin treatment, which has been reported to release agLDL residing in the SCC.22 More than 60% of the cell-associated radiolabeled cholesterol was FC (supplementary Table I). As discussed below, we have also demonstrated hydrolysis of CE in the agLDL that remains extracellular.
Formation of F-Actin Structures Around agLDL and Increase of Filipin Intensity Requires Rac or Cdc42 Activities but Not Rho Activity
Inhibition of all Rho family GTPases by Clostridium difficile toxin B has been shown to inhibit the degradation of matrix-retained and agLDL by >90%, whereas the selective inhibition of Rho by C3 transferase had no effect.13 We evaluated the role of Rho GTPases in forming the F-actin containing structures at sites of contact with agLDL and in the localized increase in FC. Cells pretreated with toxins were incubated with a suspension of Alexa546-agLDL for 30 minutes in the presence of toxins, fixed, and labeled with filipin and Alexa488-phalloidin (Figure 2H through 2M). In control cells, filipin intensity increased, and F-actin rich compartments formed at sites of contact between agLDL and cells (Figure 2H and 2K arrows). Treatment with toxin B inhibited the increase in filipin labeling (arrowheads, Figure 2I), and the ability of cells to form an F-actin rich compartment around agLDL was blocked (arrowheads, Figure 2L). Quantitative analysis showed that the average filipin intensity in pixels containing aggregates in contact with cells was significantly lower in cells treated with toxin B than in untreated cells (Figure 2N). Treatment with C3 transferase inhibited activation of Rho as confirmed by an effector pull-down assay (not shown), but there was no effect on the increase in filipin intensity or actin polymerization around agLDL (Figure 2J, 2M, and 2N). These data indicate that actin polymerization around agLDL and the increase in FC require the activity of Rac or Cdc42 but not Rho.
Hydrolysis of CE Is Required for Actin Polymerization Around agLDL
We examined whether a local increase of FC at the sites of contact between cells and agLDL is responsible for the increased actin polymerization observed around agLDL. We incubated J774 cells with agLDL containing a nonhydrolysable analog of CE – cholesteryl-oleyl ether. LDL reconstituted with either cholesteryl-oleyl ester or cholesteryl-oleyl ether was aggregated and incubated with cells for 30 minutes. The cells were fixed and labeled with filipin (Figure 3A through 3D). Sites of contact between agLDL and cells were identified in phase-contrast microscopy images. A much greater filipin signal was observed in areas of contact with ester-containing aggregates (arrowheads, Figure 3A and 3B) as compared with ether-containing aggregates (arrowheads, Figure 3C and 3D).
To investigate the role of CE hydrolysis in the increase of F-actin near agLDL, we examined F-actin in cells incubated with either ester- or ether-containing agLDL (Figure 3E and 3F). F-actin-rich structures that formed around agLDL reconstituted with CE were prominent (Figure 3E), and their morphology was similar to that observed in cells interacting with agLDL made from native LDL. Contact with the cholesteryl ether-containing agLDL caused some actin polymerization, but the area of the F-actin structures and the brightness of the phalloidin labeling were much less than for agLDL containing CE (Figure 3F and 3G). Quantitative analysis showed that the amount of F-actin around agLDL containing cholesteryl ether is significantly lower than around CE-reconstituted agLDL. These data show that formation of FC attributable to hydrolysis of CE at sites of contact between cells and agLDL is important for inducing local actin polymerization.
Depletion of FC by treatment with methyl-β-cyclodextrin (MβCD), a cholesterol chelator, also decreased local actin polymerization during interaction of macrophages with agLDL (supplemental Figure VI).
Knockdown of LAL Reduces Actin Polymerization Around agLDL
To test whether LAL-mediated CE hydrolysis is required for stimulation of actin polymerization around agLDL, we performed siRNA-mediated knockdown of LAL expression in RAW cells. The amount of LAL protein in the cells was reduced by more than 50% after treatment with LAL-specific siRNAs, whereas treatment with nontargeting siRNA did not cause a significant change in LAL expression (Figure 4A). Samples were incubated with Alexa546-agLDL for 30 minutes, fixed, and labeled for F-actin (Figure 4B and 4C). Treatment with nontargeting siRNA did not affect formation of F-actin structures around agLDL, but treatment with LAL-specific siRNA significantly reduced the formation of F-actin structures, as confirmed by quantitative analysis (Figure 4D). To confirm the specificity of the siRNA-mediated effect, we added back recombinant phLAL (human LAL produced in Pichia pastoris)21 to LAL siRNA-treated cells. This mannosylated phLAL should enter macrophage cells via the mannose receptor and be delivered to lysosomes. Restoration of LAL to cells reversed the effect of siRNAs on the F-actin surrounding agLDL (Figure 4D), demonstrating that a decrease in LAL specifically caused the decreased F-actin accumulation in siRNA knockdown cells.
Similar results were obtained on siRNA-mediated knock-down of LAL in mouse peritoneal macrophages (supplemental Figure VII).
Treatment With Bafilomycin A1 Inhibits FC Production and Formation of F-Actin at Sites of Contact Between Cells and agLDL
Proper functioning of LAL requires an acidic pH. AgLDL-containing SCCs could be acidified by H+-pumping vacuolar ATPase (V-ATPase), which resides in the plasma membrane and in internal organelles.24 To test whether acidification of the agLDL-containing compartment is needed for extracellular CE hydrolysis and actin polymerization around agLDL, we treated cells with bafilomycin A1, an inhibitor of the V-ATPase. J774 cells were incubated with Alexa546-agLDL in the absence or presence of bafilomycin A1 (1 μmol/L) for 30 minutes, fixed, and labeled with filipin and Alexa488-phalloidin. Little increase of filipin labeling was observed near aggregates touching bafilomycin A1-treated cells in comparison with nontreated cells (arrowheads, Figure 5A and 5B), and the size of F-actin structures around agLDL was smaller in bafilomycin A1-treated cells (Figure 5C and 5D). Quantification confirmed these results (Figure 5E and 5F). Consistent with this, bafilomycin A1 reduced the amount of radiolabeled FC delivered to cells attributable to labeled CE hydrolysis by 90% in a biochemical assay (supplemental Table I). (It should be noted that bafilomycin A1 treatment increases the lysosomal pH in J774 cells under these conditions; data not shown.) We have also measured the generation of FC from radiolabeled CE in extracellular aggregates by vigorously rinsing to remove agLDL from SCCs (Haka AS, Grosheva I, Chiang E, Buxbaum AR, Baird BA, Pierini LM, Maxfield FR, unpublished data, 2009). In that study, 3% to 4% of radiolabeled CE in agLDL was converted into radiolabeled FC extracellularly, and bafilomycin A1 inhibited the production of radiolabeled FC. Taken together, these data show the requirement for acidification by V-ATPase for hydrolysis of CE and induction of actin polymerization around agLDL.
In previous studies we found that increasing macrophage plasma membrane cholesterol levels globally (by incubation of cells with cholesterol chelated to a carrier, MβCD) led to alterations in macrophage signaling and F-actin organization.23,25 Based on these observations, we speculated that contact of macrophages with agLDL in the vessel wall could lead to similar alteration in cellular F-actin organization as a consequence of FC transfer. In this study, we show directly that interactions of agLDL with macrophages lead to local increases in FC and that these localized increases in FC influence (and are influenced by) local changes in F-actin organization.
Although the uptake and degradation of agLDL was shown to be actin-dependent,13 the spatial relationship of F-actin and lipoprotein aggregates was not explored. We have shown that global changes in F-actin are not induced by interactions with agLDL, but rather, F-actin-rich structures form almost exclusively at sites of contact between agLDL and the macrophage surface. This is reminiscent of the actin polymerization associated with phagocytic cups.26 However, the process described here is distinct from phagocytosis because agLDL is not taken up immediately into a sealed, degradative compartment (ie, a phagosome). Instead, agLDL remains in SCCs that are still open to the extracellular space, as demonstrated by the accessibility of the compartments to CtB.
The intimate spatial association of F-actin with the agLDL suggests that specific signaling mechanisms stimulate actin polymerization at areas of contact. Based on our previous work,23,25 we hypothesize that the local transfer of FC from agLDL drives the elaboration of F-actin rich membrane structures. We show here that macrophage engagement of agLDL leads to increases in FC at sites of contact with agLDL and that these sites colocalize with local increases in F-actin (Figure 2). This process involves Rho-family GTPases, which are key regulators of many actin-dependent processes in cells.27 Inhibition of Rac and Cdc42 activation completely abolished actin polymerization around agLDL. It appears that actin assembly leads to an increase in the area of contact with agLDL. We cannot rule out the possibility that other processes (eg, stimulated secretion of lysosomal contents) are also affected.
The role of receptors in the interaction of agLDL with macrophages is not well defined. Low-density lipoprotein-receptor–related protein (LRP) has been shown to play a role in the uptake of agLDL.28 However, several other LDL-binding receptors were examined for their potential role in uptake of agLDL, and none of them were sufficient for this process.13 In studies in mice, it was found that knockout of the scavenger receptors SRA and CD36 did not significantly reduce macrophage foam cell formation in ApoE−/− mice on a Western diet.29,30 In addition heparan sulfate proteoglycans of the syndecan family, in particular syndecan-4, were shown to mediate uptake of lipase-modified LDL.31,32
Our prior studies showed that delivery of FC to cells via MβCD (and thus without any engagement of receptors) is sufficient for the induction of actin polymerization.23 To test the necessity of FC delivery for actin polymerization after engagement of agLDL, we inhibited CE hydrolysis by replacing CE in agLDL with nonhydrolysable cholesteryl ether. We also used 2 additional approaches suggested by our characterization of the SCCs (Haka AS, Grosheva I, Chiang E, Buxbaum AR, Baird BA, Pierini LM, Maxfield FR, unpublished data, 2009): (1) knocking-down LAL levels, and (2) application of the V-ATPase inhibitor, bafilomycin A1. Under each of these conditions, the size of the F-actin rich structures around agLDL decreased significantly, indicating that agLDL-derived FC is required for F-actin polymerization. It should be noted that none of these treatments should affect receptor-mediated interactions with agLDL, indicating that such receptor interactions by themselves are not sufficient to trigger the increase in F-actin. Additionally, actin polymerization appears to be required for local FC accumulation because toxin B treatment inhibited the local increases in FC at agLDL-membrane contact zones. Thus, extracellular CE hydrolysis and local actin polymerization induced by agLDL are mutually dependent processes.
We propose a model for the interaction of macrophages with agLDL (Figure 6). On contact with agLDL (Figure 6A), an extracellular acidified compartment is formed (Figure 6B), allowing the hydrolysis of CE from agLDL (Figure 6C). This leads to local increases in membrane FC (Figure 6C), which induce localized actin polymerization (Figure 6D) and membrane extension (Figure 6E). This, in turn, promotes more extensive contact of the cell membrane with agLDL, resulting in further CE hydrolysis, FC transfer and actin polymerization, etc; a pathological cycle becomes established. The positive feedback loop may only be broken when the cell has hydrolyzed all the CE that is within reach. This model may evoke alternate approaches for the prevention of foam cell formation and atherosclerotic lesion progression.
We thank Drs Hong Du and Gregory Grabowski (The Children’s Hospital Research Foundation, Cincinnati, Ohio) for generously providing recombinant phLAL.
Sources of Funding
This work was supported by NIH grants R37-DK27083 and R01 HL037560. A.S.H. was supported by an NRSA Fellowship from the NIH.
Received September 22, 2008; revision accepted June 15, 2009.
Williams KJ, Tabas I. Lipoprotein retention–and clues for atheroma regression. Arterioscler Thromb Vasc Biol. 2005; 25: 1536–1540.
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.
Charo IF, Taubman MB. Chemokines in the pathogenesis of vascular disease. Circ Res. 2004; 95: 858–866.
Galis ZS, Sukhova GK, Kranzhofer R, Clark S, Libby P. Macrophage foam cells from experimental atheroma constitutively produce matrix-degrading proteinases. Proc Natl Acad Sci U S A. 1995; 92: 402–406.
Tabas I. Consequences and therapeutic implications of macrophage apoptosis in atherosclerosis: the importance of lesion stage and phagocytic efficiency. Arterioscler Thromb Vasc Biol. 2005; 25: 2255–2264.
Sakr SW, Eddy RJ, Barth H, Wang F, Greenberg S, Maxfield FR, Tabas I. The uptake and degradation of matrix-bound lipoproteins by macrophages require an intact actin Cytoskeleton, Rho family GTPases, and myosin ATPase activity. J Biol Chem. 2001; 276: 37649–37658.
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). J Biol Chem. 1999; 274: 32112–32121.
Zhang WY, Gaynor PM, Kruth HS. Aggregated low density lipoprotein induces and enters surface-connected compartments of human monocyte-macrophages. Uptake occurs independently of the low density lipoprotein receptor. J Biol Chem. 1997; 272: 31700–31706.
Punturieri A, Filippov S, Allen E, Caras I, Murray R, Reddy V, Weiss SJ. Regulation of elastinolytic cysteine proteinase activity in normal and cathepsin K-deficient human macrophages. J Exp Med. 2000; 192: 789–799.
Tabas I, Lim S, Xu XX, Maxfield FR. Endocytosed beta-VLDL and LDL are delivered to different intracellular vesicles in mouse peritoneal macrophages. J Cell Biol. 1990; 111: 929–940.
Du H, Schiavi S, Levine M, Mishra J, Heur M, Grabowski GA. Enzyme therapy for lysosomal acid lipase deficiency in the mouse. Hum Mol Genet. 2001; 10: 1639–1648.
Zhang WY, Ishii I, Kruth HS. Plasmin-mediated macrophage reversal of low density lipoprotein aggregation. J Biol Chem. 2000; 275: 33176–33183.
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.
Grinstein S, Nanda A, Lukacs G, Rotstein O. V-ATPases in phagocytic cells. J Exp Biol. 1992; 172: 179–192.
Nagao T, Qin C, Grosheva I, Maxfield FR, Pierini LM. Elevated cholesterol levels in the plasma membranes of macrophages inhibit migration by disrupting RhoA regulation. Arterioscler Thromb Vasc Biol. 2007; 27: 1596–1602.
Patel M, Morrow J, Maxfield FR, Strickland DK, Greenberg S, Tabas I. The cytoplasmic domain of the low density lipoprotein (LDL) receptor-related protein, but not that of the LDL receptor, triggers phagocytosis. J Biol Chem. 2003; 278: 44799–44807.
Manning-Tobin JJ, Moore KJ, Seimon TA, Bell SA, Sharuk M, Alvarez-Leite JI, de Winther MP, Tabas I, Freeman MW. Loss of SR-A and CD36 activity reduces atherosclerotic lesion complexity without abrogating foam cell formation in hyperlipidemic mice. Arterioscler Thromb Vasc Biol. 2009; 29: 19–26.
Boyanovsky BB, Shridas P, Simons M, van der Westhuyzen DR, Webb NR. Syndecan-4 mediates macrophage uptake of group V secretory phospholipase A2-modified LDL. J Lipid Res. 2009; 50: 641–650.