Plasmin-mediated Macrophage Reversal of Low Density Lipoprotein Aggregation*

Evidence suggests that aggregated low density lipoprotein (AgLDL) accumulates in atherosclerotic lesions. Previously, we showed that AgLDL induces and enters surface-connected compartments (SCC) in human monocyte-derived macrophages by a process we have named patocytosis. Most AgLDL taken up by these macrophages in the absence of serum is stored in SCC and remains undegraded. We now show that macrophages released AgLDL (prepared by vortexing or treatment with phospholipase C or sphingomyelinase) from their SCC when exposed to 10% human lipoprotein-deficient serum (LPDS). Macrophages also took up AgLDL in the presence of LPDS, but subsequently released it. In both cases, the released AgLDL was disaggregated. Although the AgLDL that macrophages took up could not pass through a 0.45-μm filter, >60% of AgLDL could pass this filter after release from the macrophages. Disaggregation of AgLDL was verified by gel-filtration chromatography and electron microscopy that also showed particles larger than LDL, reflecting fusion of LDL that aggregates. The factor in serum that mediated AgLDL release and disaggregation was plasmin generated from plasminogen by macrophage urokinase plasminogen activator. AgLDL release was decreased >90% by inhibitors of plasmin (ε-amino caproic acid and anti-plasminogen mAb), and also by inhibitors of urokinase plasminogen activator (plasminogen activator inhibitor-1 and anti-urokinase plasminogen activator mAb). Moreover, plasminogen could substitute for LPDS and produce similar macrophage release and disaggregation of AgLDL. Because only plasmin bound to the macrophage surface is protected from serum plasmin inhibitors, interaction of AgLDL with macrophages was necessary for reversal of its aggregation by LPDS. The released disaggregated LDL particles were competent to stimulate LDL receptor-mediated endocytosis in cultured fibroblasts. Macrophage-mediated disaggregation of aggregated and fused LDL is a mechanism for transforming LDL into lipoprotein structures size-consistent with lipid particles found in atherosclerotic lesions.

Focal vessel wall retention of plasma-derived low density lipoprotein (LDL) 1 contributes to cholesterol buildup in athero-sclerotic lesions. This leads to atherosclerosis, a disease that represents the response of the vessel wall to retained LDL (1). Aggregation of LDL may contribute to its retention in developing atherosclerotic lesions (1,2). LDL recovered from lesions shows an increased tendency to aggregate (3,4), and aggregates of spherical particles presumed to be lipoproteins have been demonstrated in early developing lesions (5)(6)(7).
In contrast to what is usually the case for monomeric native LDL, AgLDL is readily taken up by macrophages and causes macrophage cholesterol accumulation (3, 8 -14). Previously, we showed that human monocyte-derived macrophages accumulate AgLDL by a unique endocytic pathway (15). In this actindependent process AgLDL induces a labyrinth of surface-connected compartments (SCC) and accumulates within them. We have named this uptake pathway patocytosis, after the Latin word patere, meaning to lie open. The protein component of LDL, apoB, and other hydrophobic materials can stimulate patocytosis (16,17). After accumulating within macrophage SCC, some AgLDL is transported to lysosomes where it undergoes degradation followed by acyl-CoA:cholesterol acyltransferase-dependent re-esterification of LDL cholesterol (15). 2 However, most AgLDL remains undegraded within SCC.
Not all cholesterol in atherosclerotic lesions accumulates within macrophages. Much of this cholesterol accumulates as cholesteryl ester-rich lipid particles in the extracellular spaces of atherosclerotic lesions, and these particles are particularly enriched in the lipid-rich core of lesions. Evidence suggests that the cholesteryl ester-rich lipid particles are derived from LDL rather than from cellular lipid droplets (reviewed in Ref. 18). The particles resemble LDL in having linoleate as the major cholesteryl ester fatty acyl group, while cellular lipid droplets have oleate as their major cholesteryl ester fatty acyl group. One important difference between extracellular cholesteryl ester-rich particles and LDL or cellular lipid droplets is that while LDL are 22 nm in diameter and cellular lipid droplets are Ͼ400 nm, the cholesteryl ester-rich lipid particles range 40 to 200 nm in diameter.
Many modifications to LDL have been shown to cause LDL to aggregate (3,8,9,19). These include oxidation of LDL and treatment of LDL with certain lipases that are present within lesions (sphingomyelinase, phospholipases A 2 and C). Simple vortexing of LDL is a convenient way to produce aggregated LDL (AgLDL) (10). Most of these modifications not only cause LDL to aggregate but also to undergo fusion (20 -23). Fusion of the LDL particles causes them to attain sizes similar to the cholesteryl ester-rich lipid particles that accumulate in the extracellular spaces of atherosclerotic lesions. Electron microscopy shows that these LDL-like cholesteryl ester-rich lipid particles can occur as individual particles in lesions (24). Also, * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 (18). Currently, no model links the occurrence of fused LDL in aggregates with the presence of lesion extracellular cholesteryl ester-rich lipid particles. Here, we report that upon exposure to and activation of plasminogen, macrophages disaggregate and release AgLDL that these cells have accumulated in SCC. Disaggregation of aggregated and fused LDL is one possible mechanism by which LDL could form cholesteryl ester-rich lipid particles larger than LDL, and resembling those lipid particles that accumulate within the extracellular spaces of atherosclerotic lesions.
Preparation of Aggregated Lipoproteins-LDL was aggregated by vortexing as described in Ref. 10 or was aggregated with phospholipase C and sphingomyelinase essentially as described before (9,20,21). Filtration of vortexed LDL through a 0.45-m filter eliminated macrophage uptake and degradation of 125 I-AgLDL. 2 This showed that only the aggregated fraction of LDL could induce and enter macrophage SCC. Therefore, only the aggregated fraction of LDL, isolated by centrifugation at 14,000 ϫ g for 15 min, was incubated with macrophages. Except where stated otherwise, experiments were carried out with AgLDL produced by vortexing.
Cell Culture and Assays-Preparation of human fibroblast and 2-week-old human monocyte-derived macrophage cultures and assays of lipoprotein metabolism and cholesterol content of cells were carried out as described previously (15,25). All data are presented as mean Ϯ S.E. determined from 3 culture wells for each data point. No error bar is shown when the error was smaller than the symbol height.
Preparation of Macrophage Disaggregated AgLDL and Incubation with Cells-Macrophages were incubated 24 h with 50 g/ml 125 I-AgLDL in the presence of RPMI 1640 medium with 10% LPDS. Then, media were collected, pooled, and dialyzed against a 200-fold greater volume of RPMI 1640 to remove most trichloroacetic acid-soluble 125 Ityrosine produced during the initial incubation of 125 I-AgLDL with macrophages. Next, the pooled media were filtered (0.45-m pore size) to produce a fraction that contained 8.5-16 g/ml disaggregated trichloroacetic acid-insoluble 125 I-AgLDL. Metabolism of the disaggregated trichloroacetic acid-insoluble 125 I-AgLDL was assessed by incubating this medium with fibroblasts and fresh macrophages. Control incubations were carried out by incubating fibroblasts and fresh macrophages with similar amounts of 125 I-LDL added to medium conditioned by macrophages 24 h without 125 I-AgLDL.
Electron Microscopy-Macrophage membranes in physical continuity with the extracellular space were labeled with ruthenium red according to the method of Luft (26), and macrophage cultures were prepared for electron microscopy as described previously (27). Lipoprotein particles in culture medium were visualized by negative staining (25).

Release of AgLDL from Macrophage SCC Was Mediated by
Plasmin Derived from Serum-After macrophages accumulated 125 I-AgLDL within SCC (for either 5 or 24 h) and were then exposed to LPDS for 24 h, macrophages released much of their accumulated 125 I-AgLDL back into the culture medium. Most 125 I-AgLDL was released as trichloroacetic acid-insoluble material indicating the presence of relatively intact LDL protein (Table I). This occurred for LDL aggregated by vortexing (Fig. 1A) and also for LDL aggregated by exposure to phospholipase C or sphingomyelinase (Fig. 2). Electron microscopy confirmed loss of AgLDL from macrophage SCC upon exposure of macrophages to LPDS (Fig. 3).
Previously, we had shown that exposure of macrophages to trypsin could release AgLDL from SCC (15). Therefore, we tested whether the factor in LPDS that released AgLDL was a protease. The serum factor was sensitive to aprotinin, a serine protease inhibitor (Figs. 1A and 2), suggesting that serum contained some serine protease activity that induced release of AgLDL from macrophage SCC. Moreover, this protease could function in the presence of the many naturally occurring protease inhibitors that serum contains.
Plasmin is a serum serine protease that shows resistance to serum protease inhibitors when it is bound to the macrophage cell surface (28). Serum plasminogen is converted to active plasmin by urokinase and tissue plasminogen activators both of which are produced by macrophages within atherosclerotic lesions (29 -34). Many findings indicated that plasmin was the serum factor that mediated release of AgLDL from macrophage SCC. First, the releasing activity of LPDS was decreased 96 Ϯ 10% when LPDS was adsorbed with lysine-agarose, an agent that binds plasminogen and plasmin. Second, ⑀-amino caproic acid (25 mM), a plasmin inhibitor, decreased by 100 Ϯ 17% LPDS-induced release of 125 I-AgLDL from macrophages. Last, an anti-plasminogen/plasmin monoclonal antibody added to LPDS decreased by 99 Ϯ 10% LPDS-induced release of 125 I-AgLDL (Fig. 1B). Plasmin is reported to activate matrix metalloproteinases released by macrophages (35). However, we found no evidence that macrophage-derived matrix metalloproteinases were required for plasmin-mediated release of macrophage-associated 125 I-AgLDL. TIMP-I (1 g/ml), TIMP-II (1 g/ml), and phosphoramidon (150 g/ml) inhibitors of matrix  Fig. 5 except for medium trichloroacetic acid-insoluble data that is not shown in Fig. 5 but was obtained in the same experiment. Medium trichloroacetic acid-insoluble data following filtration is shown in Fig. 5, and medium trichloroacetic acid-insoluble data obtained from unfiltered samples is shown here for comparison with other similar data in the table.
metalloproteinases did not inhibit LPDS (or purified plasminogen)-stimulated release of 125 I-AgLDL from macrophages.
Macrophage Conversion of Serum Plasminogen to Active Plasmin Was Necessary for Release of AgLDL from Macrophages-The plasmin activity that mediated release of 125 I-AgLDL from human monocyte-macrophages was generated from serum plasminogen by the macrophages. Macrophages that had accumulated 125 I-AgLDL during a 5-h incubation with 50 g/ml 125 I-AgLDL, released 60% of their cell-associated 125 I-AgLDL when exposed to trypsin (50 g/ml) for 30 min, but released none of their 125 I-AgLDL when exposed to 10% LPDS for this short period. The fact that LPDS releasing activity was present after 24 h (Fig. 1A) but not after 30 min of incubation is consistent with a time-dependent generation of a releasing factor in LPDS. Addition of the natural inhibitor of plasminogen activator, plasminogen activator inhibitor-1 (PAI-1)(16.5 g/ml) to LPDS diminished LPDS releasing activity by 91 Ϯ 11%. Also, a purified mouse monoclonal antibody that inhibits activity of urokinase plasminogen activator (200 g/ml), decreased LPDS releasing activity by Ͼ99%. In contrast, a purified mouse monoclonal antibody that inhibits tissue plasminogen activator did not affect LPDS releasing activity. Last, the releasing activity of LPDS could be replaced by substituting purified plasminogen for LPDS ( Fig. 1C and Table I). LPDS and purified plasminogen not only caused macrophage release of trichloroacetic acid-insoluble 125 I-AgLDL protein, but these agents also caused macrophage release of LDL cholesterol (Table II). This showed that LDL particles and not just their protein components were released from macrophages.
Macrophage-generated Plasmin Caused Some Degradation of AgLDL-Besides inducing release of macrophage accumulated 125 I-AgLDL, LPDS also increased degradation of 125 I-AgLDL. Previously, we showed that following 125 I-AgLDL entry into macrophage SCC, chloroquine-sensitive (presumably lysosomal) degradation of some 125 I-AgLDL occurred (15). However, the increase in 125 I-AgLDL degradation stimulated by exposure of macrophages to LPDS or plasminogen was not mediated by macrophage lysosomes because chloroquine (100 M) did not inhibit this degradation. The degradation was due to macrophage-generated plasmin because inhibition of plasmin with 25 FIG. 1. Plasmin-mediated release of AgLDL accumulated by macrophages. Two-week-old monocyte-macrophage cultures were incubated 5 h with 50 g/ml 125 I-AgLDL in 1 ml of RPMI 1640 medium. Then, cultures were rinsed 3 times with 1 ml of RPMI 1640 medium, and incubated 1 day in RPMI 1640 medium with the additions as indicated in the figure: 1 unit/ml human plasminogen, 0.15 unit/ml aprotinin, 10% human lipoprotein-deficient serum, 200 g/ml purified mouse anti-plasminogen monoclonal antibody, and 200 g/ml purified isotype-matched control mouse monoclonal antibody (MOPC 21 mouse myeloma protein). Following incubations, cell-associated 125 I-AgLDL was determined and media were analyzed for trichloroacetic acid (TCA)-soluble and insoluble 125 I-AgLDL.
FIG. 2. Macrophage release of AgLDL produced by phospholipase C and sphingomyelinase. Two-week-old monocyte-macrophage cultures were incubated for 5 h with 50 g/ml 125 I-AgLDL aggregated with either phospholipase C (A) or with sphingomyelinase (B) in order to accumulate AgLDL in macrophage SCC (this was confirmed by electron microscopy). Then, cultures were rinsed and incubated 1 day in RPMI 1640 medium with no addition, 10% LPDS, or 10% LPDS ϩ aprotinin (0.15 unit/ml). Shown is the amount of trichloroacetic acid (TCA)-insoluble 125 I-AgLDL released into the medium.
FIG. 3. Serum caused release of AgLDL from macrophage surface-connected compartments. Two-week-old monocyte-macrophage cultures were incubated 5 h with 200 g/ml AgLDL, rinsed 3 times with RPMI 1640 medium, and then incubated 1 day in RPMI 1640 medium with either no addition (A) or 10% LPDS (B). Then cultures were rinsed, fixed with glutaraldehyde, and exposed to ruthenium red to label cellular membranes in continuity with the extracellular space. Cultures were processed for electron microscopic analysis without counterstaining. In A, ruthenium red-stained AgLDL remained within ruthenium red-stained SCC (arrows) during postincubation without LPDS, while in B, LPDS treatment caused release of AgLDL from SCC that appear collapsed (arrows). Upper right hand corner of A shows where a surface-connected compartment opens to the extracellular space. Bar is 2 m and applies to A and B. mM ⑀-amino caproic acid (data not shown) or an anti-plasminogen monoclonal antibody (Fig. 1B) prevented the LPDS-induced increase in 125 I-AgLDL degradation. That plasmin degrades 125 I-AgLDL was shown by incubating 125 I-AgLDL (50 g/ml) with plasmin (0.1 unit/ml) for 24 h. About 6% of the 125 I-AgLDL converted to trichloroacetic acid-soluble protein showing that plasmin could degrade the protein component (i.e. apoB) of LDL as previously reported (36).
Macrophage-generated Plasmin Caused Reversal of LDL Aggregation-Only AgLDL induced and entered macrophage SCC (14). However, the 125 I-AgLDL released from SCC following exposure to LPDS or plasminogen was much less aggregated. Greater than 60% of the released 125 I-AgLDL passed through a 0.45-m (pore size) filter (Table III, part A), while less than 10% of the original 125 I-AgLDL passed through the 0.45-m filter. Direct exposure of 125 I-AgLDL (50 g/ml) to plasmin (but not plasminogen) for 1 day converted Ͼ84% of the 125 I-AgLDL (both vortexed or lipase-treated) to a filtrable form. This showed that plasmin was sufficient to disaggregate 125 I-AgLDL. The size distribution of macrophage-disaggregated AgLDL (the fraction that passed through a 0.45-m filter) was compared with LDL by gel-filtration chromatography (Fig. 4). It was not possible to assess the size of the initial AgLDL added to macrophage cultures because it was too large to enter the gel-filtration column. Macrophage disaggregation of AgLDL generated lipid particles that eluted both within the elution range (fractions 30 -60) of LDL (62% of eluted disaggregated AgLDL cholesterol) and earlier than the LDL elution range (38% of eluted disaggregated AgLDL cholesterol).
Incubation of 125 I-AgLDL with macrophages in the presence of LPDS as a source of plasminogen did not prevent initial macrophage uptake of 125 I-AgLDL (Fig. 5A). However, over time the cell-associated 125 I-AgLDL decreased during incubation with LPDS. Simultaneously, the 125 I-AgLDL was progressively released into the medium in disaggregated form shown by the increasing amount of medium trichloroacetic acid-insoluble 125 I-AgLDL that could be filtered (Fig. 5C and Table I). This was accompanied by an increase in degradation of 125 I-AgLDL that was mostly due to plasmin activity because degradation was inhibited 73% by anti-plasminogen monoclonal antibody (Figs. 5B and 6). Degradation of monomeric 125 I-LDL (50 g/ml) incubated 24 h with macrophages did not increase in the presence of 10% LPDS (degradation was 0.5 Ϯ 0.0 g/mg of cell protein both without and with 10% LPDS).
Macrophages accumulated about 75% as much cell-associated 125 I-AgLDL when incubated 5 h with 125 I-AgLDL (50 g/ml) and plasminogen (1 unit/ml) as when incubated with 125 I-AgLDL alone. However, as was the case for macrophages incubated with 125 I-AgLDL in the presence of LPDS, during prolonged incubation (24 h) of macrophages with 125 I-AgLDL and plasminogen, Ͼ90% of cell-associated 125 I-AgLDL was subsequently released. Macrophages exposed to 125 I-AgLDL for 48 h in the presence of 1 unit/ml plasminogen also disaggregated 125 I-AgLDL such that Ͼ50% of the total trichloroacetic acid-insoluble 125 I-AgLDL passed through a 0.45-m filter. Electron microscopy of negatively stained samples of unfiltered culture media showed that AgLDL was not disaggregated when exposed to plasminogen without macrophages (Fig. 7A), but was converted to individual lipid particles that ranged in size from 22 to 75 nm when exposed to plasminogen in the presence of macrophages (Fig. 7B). Similar-sized lipoprotein particles were released from macrophages that first were allowed to accumulate AgLDL (100 g/ml) for 5 h, and then were exposed to plasminogen or LPDS for 24 h to cause disaggregation and release of AgLDL from macrophages.
When serum was present, interaction of 125 I-AgLDL with macrophages was required for reversal of its aggregation. Medium removed from macrophage cultures after conditioning for 24 h in the presence of 10% LPDS did not disaggregate subsequently added 125 I-AgLDL (50 g/ml) during a 24-h incubation. This finding is consistent with the fact that serum contains plasmin inhibitors, and in the presence of serum, only macrophage-bound plasmin is active (28). When serum was absent, culture medium containing plasminogen conditioned 24 h by macrophages could disaggregate 125 I-AgLDL (72% of this treated 125 I-AgLDL passed a 0.45-m filter) (Table III, part B). This showed that macrophages could generate active plasmin in the culture medium but that in the presence of serum, the activity of plasmin in the culture medium was inhibited.
Effects of Macrophage Cholesterol Enrichment on Uptake, Release, and Disaggregation of AgLDL-Cholesterol-enriched macrophages retained their capacity to accumulate 125 I-AgLDL, and subsequently disaggregate and release this 125 I- Then, macrophage cultures were rinsed and incubated 24 h in RPMI 1640 medium with 1 unit/ml plasminogen to cause macrophage release and disaggregation of AgLDL. Following this incubation, media were collected, pooled, concentrated, and gel-filtered through 2% agarose (B) as described previously (25). Native LDL was eluted through the same gel for comparison (A). Fractions were assayed for cholesterol. Fraction 18 is the void volume peak. Cellular Metabolism of Macrophage-disaggregated AgLDL-Macrophage metabolism of disaggregated AgLDL particles released from macrophages was assessed. Disaggregated 125 I-AgLDL was produced by incubating macrophages with 125 I-AgLDL in the presence of 10% LPDS, dialyzing the collected media to remove trichloroacetic acid-soluble 125 I-tyrosine, then filtering media to produce a fraction containing disaggregated trichloroacetic acid-insoluble 125 I-AgLDL. Aprotinin (0.5 unit/ ml) was added to the disaggregated 125 I-AgLDL (16 g/ml) samples that were then incubated 5 h with fresh monocytemacrophages. The aprotinin inhibited any further plasmin-dependent proteolysis of the disaggregated 125 I-AgLDL so that lysosomal degradation of the lipoproteins could be assessed. No  7. Structure of AgLDL after interaction with macrophages. 100 g/ml AgLDL produced by phospholipase C treatment was incubated 24 h in RPMI 1640 medium containing 1 unit/ml plasminogen without (A) or with macrophages (B). Samples of culture medium were negatively stained and examined by electron microscopy. The large aggregates of LDL produced by phospholipase C treatment (A) were disaggregated during incubation with macrophages in the presence of plasminogen (B). Phospholipase C causes fusion of LDL (22) accounting for the presence of lipid particles larger than 22 nm in diameter, the usual size of LDL. Bar is 100 nm and applies to A and B. macrophage degradation of the disaggregated 125 I-AgLDL occurred as shown by the finding that there was no production of trichloroacetic acid-soluble 125 I-tyrosine. Plasmin treatment of LDL is reported not to affect its recognition and uptake through the LDL receptor (37). Therefore, we tested whether macrophages could degrade native 125 I-LDL. Macrophages also did not degrade fresh 125 I-LDL (16 g/ml) added to macrophage-conditioned medium without aprotinin and incubated similarly for 5 h with fresh macrophages. The lack of degradation of 125 I-LDL and macrophage-disaggregated 125 I-AgLDL is consistent with down-regulation of the LDL receptor reported to occur in differentiated human monocyte-macrophages (38). Indeed, confirming what we reported previously (15), uptake of AgLDL into human monocyte-macrophages was not mediated by the LDL receptor. An anti-LDL receptor antibody (200 g/ ml) that blocks LDL interaction with the LDL receptor did not decrease uptake of 125 I-AgLDL (50 g/ml) incubated 5 h with macrophages.
Human fibroblasts degrade 125 I-LDL taken up by the LDL receptor that is well expressed in these cells. Therefore, we compared degradation of 125 I-LDL (8.5 g/ml) and macrophage-disaggregated 125 I-AgLDL (8.5 g/ml) incubated 24 h with human fibroblasts. Human fibroblasts degraded similar amounts of 125 I-LDL and macrophage-disaggregated 125 I-AgLDL, 2.1 Ϯ 0.1 and 2.0 Ϯ 0.1 g/ml, respectively. Degradation was inhibited by a 23-fold excess of unlabeled LDL showing that the degradation was specific and mediated by the LDL receptor.

DISCUSSION
Uptake of AgLDL into macrophages by patocytosis led to its storage in SCC until macrophages were exposed to serum. Then, disaggregation and release of accumulated AgLDL occurred following macrophage-mediated conversion of serumderived plasminogen to active plasmin. On the other hand, if macrophages encountered AgLDL in the presence of serum, the macrophages initially accumulated AgLDL but over time released disaggregated LDL due to macrophage activation of serum plasminogen. Although it has been reported that macrophages possess both the urokinase and tissue types of plasminogen activators (29), only the urokinase-type of plasminogen activator-mediated plasminogen activation was involved in release and disaggregation of AgLDL. A monoclonal antibody that inhibits urokinase plasminogen activator effectively blocked release of AgLDL from macrophages, while a monoclonal antibody that inhibits tissue plasminogen activator did not block AgLDL release.
How does plasmin cause release of AgLDL from macrophage SCC? Previously, we showed that trypsin could release AgLDL that had accumulated in SCC (15). Both trypsin and plasmin can partially degrade apoB of LDL (36) and disaggregate AgLDL in the absence of macrophages. Here, macrophagegenerated plasmin caused partial degradation of LDL protein (i.e. apoB). While some LDL contained in AgLDL may bind SCC directly, most LDL in AgLDL is presumably retained in SCC because of adherence of one LDL to another LDL within AgLDL. It is likely that plasmin degradation of apoB disrupts the non-covalent bonds that hold LDL in aggregates and causes LDL to disaggregate. Because SCC are open to the extracellular space, LDL particles disaggregated by plasmin can diffuse from SCC into the extracellular space. It is unlikely that macrophage-generated plasmin induced an active expulsion of LDL from SCC as neither cytochalasin D nor nocodazole (inhibitors of microfilaments and microtubules, respectively) blocked LDL release. 2 Interaction of AgLDL with macrophages was necessary for reversal of its aggregation in serum. This finding is consistent with previous reports that only plasmin bound to the macrophage surface is active, because when bound, plasmin is protected from the action of serum inhibitors (28). Because only macrophage-bound plasmin is active in the presence of serum, it makes sense that plasmin should be closely associated with SCC where AgLDL accumulates. Unfortunately, we have not found an anti-plasmin antibody suitable for carrying out electron immunocytochemistry to investigate this issue.
While macrophage-generated plasmin increased degradation of AgLDL in the presence of serum, macrophage-generated plasmin did not cause degradation of native (monomeric) LDL. Monomeric LDL does not enter macrophage SCC (15), and monomeric LDL is not substantially metabolized by well differentiated human monocyte-macrophages owing to limited expression of the LDL receptor (38). On the other hand, AgLDL uptake into SCC of human monocyte-macrophages is not mediated by the LDL receptor (15). Thus, the finding that macrophage-generated plasmin could not degrade monomeric LDL might be because monomeric LDL does not come in contact with macrophage-bound plasmin.
After macrophage accumulation of AgLDL, exposure of these macrophages to LPDS did not result in plasmin-mediated release of all accumulated cholesterol (see Table III). This could have occurred for several reasons. First, some AgLDL undergoes lysosomal degradation during and after incubation of macrophages with AgLDL (15). Cholesterol derived from Then the macrophages were rinsed 3 times in RPMI 1640 medium, and incubated 1 day in RPMI 1640 with either plasminogen (1 unit/ml) or 10% LPDS. Following incubations, the trichloroacetic acid-insoluble 125 I-AgLDL released into the medium was determined before and after filtration through a 0.45-m (pore-size) polysulfone (low protein-binding) filter. This assessed the decrease in size of the original 125 I-AgLDL, Ͻ10% of which could pass through the 0.45-m filter before or after a 1-day incubation in either medium without macrophages. B, macrophages were incubated in RPMI 1640 medium with either no addition, plasminogen (1 unit/ml), or 10% LPDS for 1 day. The macrophage-conditioned medium was removed, filtered, and 50 g of 125 I-AgLDL was added to each sample of medium. After a 1-day incubation at 37°C, the % of trichloroacetic acid-insoluble 125 I-AgLDL in the medium that passed through a 0.45-m filter was determined. AgLDL degraded in lysosomes could remain within macrophages. Second, exposure of macrophages to serum did not release all cell-associated AgLDL from macrophages, and thus some AgLDL may remain in SCC. Last, Tabas and colleagues (39) have recently shown that when macrophages are incubated with AgLDL, LDL-cholesteryl ester hydrolysis exceeds LDL protein degradation, a phenomenon associated with selective cellular uptake of cholesteryl esters from monomeric LDL in other studies (40). Possibly, some cholesterol transfers from AgLDL in SCC into the macrophage cytoplasm by this selective uptake process. In any case, it remains to be determined under what conditions AgLDL can transform human monocyte-derived macrophages into foam cells where most accumulated cholesteryl ester is stored in cytoplasmic lipid droplets. Under the conditions examined so far, most AgLDL (and its cholesterol) that enters human monocyte-macrophages by patocytosis remains in macrophage SCC, or is disaggregated and released when these macrophages are exposed to plasminogen in serum.
What is the fate of disaggregated AgLDL particles released from macrophages? While it is possible that disaggregated AgLDL released from macrophage SCC could re-enter macrophages through other endocytic pathways such as coated pits and undergo lysosomal degradation, this did not occur during incubation of macrophages with the concentrations of disaggregated AgLDL that we could test. Incubation of disaggregated 125 I-AgLDL with fresh macrophages in medium lacking plasmin activity did not produce any additional degradation of the disaggregated 125 I-AgLDL. However, 125 I-LDL also was not degraded significantly when incubated with macrophages under similar conditions (i.e. low 125 I-LDL concentration and short time of incubation that should detect LDL receptor-mediated uptake rather than low affinity and fluid-phase uptake of 125 I-LDL). As mentioned above, lack of significant degradation likely reflects down-regulation of the LDL receptor in differentiated monocyte-derived macrophages (38). Macrophage-released disaggregated 125 I-AgLDL and 125 I-LDL were degraded similarly by human fibroblasts showing that LDL receptor-mediated uptake of disaggregated 125 I-AgLDL could occur. Although macrophage-disaggregated AgLDL potentially can be metabolized through the LDL receptor, this receptor is down-regulated in cells of human atherosclerotic lesions (41). Thus, if disaggregated AgLDL occurs in atherosclerotic lesions, the lipid particles could accumulate in the extracellular spaces of lesions similar to their accumulation in culture medium here.
Aggregates of spherical particles, some having larger diameters (up to 118 nm) than LDL (22 nm), have been observed by freeze-etch analysis in rabbit subendothelial matrix following injection of monomeric LDL (5), and during atherosclerotic lesion development in Watanabe heritable hyperlipidemic rabbits that have genetically elevated LDL levels (7). Similar-sized non-aggregated cholesteryl ester-rich lipoprotein structures occur in atherosclerotic lesions, and do not appear to be derived from release of intracellular lipid droplets (7,18,(42)(43)(44)(45). Here, disaggregated AgLDL released by macrophages also showed particles with diameters larger than LDL. The presence of particles larger than LDL that were not aggregates of LDL can be attributed to the fact that vortexing, sphingomyelinase, and phospholipase C treatment of LDL cause LDL fusion as well as aggregation (20 -22, 46). Thus, macrophage disaggregation of aggregated and fused LDL is one mechanism that could contribute to the accumulation in lesions of those extracellular lipid particles that resemble LDL chemically but that are larger than LDL.
Plasminogen derived from the blood has been detected in atherosclerotic lesions, and modified or native LDL increase the capacity of some types of macrophages to produce plasmin (47,48). Also, we found that macrophage cholesterol accumulation did not inhibit plasminogen-dependent release and disaggregation of AgLDL. Thus, since macrophages reverse aggregation of LDL in the presence of plasminogen in vitro, one might expect disaggregated AgLDL in areas where macrophages are present. However, conversion of plasminogen to active plasmin by macrophages could be limited. PAI-1 inhibited serum plasminogen-induced release of AgLDL from macrophages. Considering that levels of PAI-1 are increased in atherosclerotic lesions compared with normal, plasmin generation may be confined to areas of lesions where plasminogen activation is not inhibited by PAI-1 (32, 49 -53). The state of aggregation of LDL surrounding macrophages has not yet been examined with freeze-etch microscopy, the only technique that can detect lipoprotein particle aggregation. Other types of immunohistochemical and routine microscopic studies of LDL in lesions cannot distinguish whether lipoprotein particles are aggregated or are packed together without being aggregated.
Efflux of lipoprotein particles from the vessel wall (including those as large as the fused LDL particles observed in the present study) is inversely proportional to their size (54). Therefore, aggregation of LDL in atherosclerotic lesions could contribute to LDL accumulation in lesions. Accordingly, plasmin-mediated disaggregation of LDL aggregates might facilitate efflux of LDL from the vessel wall. In this regard, plasmin has been shown to release a substantial fraction of LDL trapped in minced atherosclerotic plaque tissue samples (55). On the other hand, because macrophage emigration from lesions occurs (56), plasmin-mediated release of AgLDL from macrophages precludes the possibility of macrophages transporting AgLDL out of lesions. Further study is needed to learn the significance of macrophage-mediated AgLDL disaggregation for the development of atherosclerotic lesions. In any case, macrophage disaggregation of aggregated and fused LDL is a novel mechanism for generating size-consistent models of lipoprotein structures found in atherosclerotic lesions.