INTRODUCTION

Previously, we reported that human monocyte-macrophages display an unusual mode of endocytosis in which they sequester cholesterol crystals into an extensive labyrinth of cytoplasmic compartments (1). These compartments remain connected to the surface of the macrophage and to the extracellular space. Thus, this endocytotic process is different from phagocytosis, where particles are taken into the macrophage within vacuoles that form from pinched-off regions of the plasma membrane and do not maintain any connection to the extracellular space (2).

Besides cholesterol crystals, we observed that acetylated low density lipoprotein (AcLDL)¹ could induce and enter SCC of human monocyte-macrophages (1). However, we subsequently discovered that this occurred only for some lots of AcLDL and not with other lots. AcLDL that entered these compartments was aggregated in linear strands. We suspected that some aggregation of AcLDL occurred during its storage and that this aggregated AcLDL could have been responsible for the formation of the macrophage SCC.

Other investigators have shown that aggregation of lipoproteins enhances their uptake by macrophages. Oxidation and thiolation of LDL or treatment of LDL with sialidase, phospholipase C, or sphingomyelinase plus lipoprotein lipase (in the presence of cultured smooth muscle cells) and other treatments cause LDL to aggregate and to show enhanced degradation by macrophages (3-10). Khoo et al. (11) showed that aggregation of normal LDL by vortexing converted LDL to a form that was readily taken up by mouse peritoneal macrophages through an LDL receptor-dependent mechanism. The endocytotic pathway for uptake of the vortexed low density lipoprotein (VxLDL) was not described, but it was suggested that this LDL entered the macrophages by phagocytosis.

In the present study, we examined whether aggregated LDL has the capacity to induce and accumulate within monocyte-macrophage SCC. Not only did aggregated LDL induce and accumulate within SCC, this occurred by a mechanism that did not depend on the classical LDL receptor.

MATERIALS AND METHODS

Human LDL was prepared as described previously (12) and was obtained from PerImmune (Rockville, MD). AcLDL was prepared as described by Basu et al. (13). Methylation of LDL was performed according to Weisgraber et al. (14). ¹²⁵I-Labeled LDL and AcLDL with specific activities that ranged 70-200 µCi/mg of protein were obtained from Biomedical Technologies (Stoughton, MA). Lipoproteins that were not to be aggregated by vortexing were centrifuged at 14,000 × g for 10 min to remove any spontaneously formed aggregates from these lipoprotein preparations. To prepare aggregated lipoproteins, 40 µl of LDL or methylated LDL (5 mg/ml for unlabeled and 1-3 mg/ml for labeled lipoproteins) was placed into a 0.5-ml silicone-coated polypropylene tube and vortexed (VWR vortex mixer) at the maximal speed for 1 min. LDL aggregated with phospholipase C treatment was prepared as described previously (7, 8).

The cell association and degradation of lipoproteins by human monocyte-derived macrophages was performed according to the methods of Goldstein et al. (15) using ¹²⁵I-VxLDL and ¹²⁵I-AcLDL. Human monocyte-derived macrophages were cultured as described previously except that 2 × 10⁶ monocytes/well were initially seeded into 12-well (22-mm diameter) culture plates (Plastek C, MatTek Corp., Ashland, MA) (12). Two-week-old monocyte-macrophage cultures were rinsed three times with RPMI 1640 medium and incubated for the indicated times at 37 °C with ¹²⁵I-labeled lipoproteins added to RPMI 1640 medium containing 0.2% fatty acid-free bovine serum albumin.

Lipoprotein degradation was quantified by measurement of trichloroacetic acid-soluble organic iodide radioactivity in supernatants of media samples that were centrifuged at 15,000 × g for 10 min. Values obtained in the absence of cells were <5% those values obtained with cells. These control values were subtracted. Cell-associated ¹²⁵I-lipoproteins were determined by rinsing macrophages 5 times with Dulbecco's phosphate-buffered saline (DPBS) containing Ca²⁺, Mg²⁺, and 0.35% bovine serum albumin (3 quick rinses and two 10-min incubations on ice). After a final rinse with DPBS plus Ca²⁺ and Mg²⁺, macrophages were dissolved overnight in 0.1 N NaOH. Aliquots of NaOH-solubilized cell samples were assayed for ¹²⁵I radioactivity. Values of ¹²⁵I radioactivity determined for wells incubated with ¹²⁵I-lipoproteins but without macrophages were subtracted. These values were <1% those of the cell-associated ¹²⁵I-lipoproteins. Macrophage protein content was determined by the method of Lowry et al. (16) using bovine serum albumin as a standard. Protein contents of cultures generally ranged between 0.2 and 0.3 mg/well.

For Experiment I in Table II, passaged normal human skin fibroblasts (GM09503, National Institute of General Medical Sciences Cell Repository, NIH, Camden, NJ) were seeded into 6-well (35-mm diameter) Plastek C culture plates at a density of 10⁵ cells/well. These cultures were maintained 3 days with Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and then for another 2 days with Dulbecco's modified Eagle's medium containing 10% human lipoprotein-deficient serum (PerImmune) before incubations with ¹²⁵I-lipoproteins were carried out.

Table II. Role of LDL receptor in metabolism of 125I-VxLDL Two-week-old monocyte-macrophage or passaged skin fibroblast cultures were incubated for 5 h with 20 µg/ml 125I-lipoproteins as indicated. Lipoproteins were added to 1 ml of RPMI 1640 medium containing 0.2% fatty acid-free bovine serum albumin for monocyte-macrophage cultures and to 1 ml of DMEM medium (also containing albumin) for fibroblast cultures. After incubations, the amount of TCA-soluble organic 125I in media was measured to determine degraded 125I-lipoprotein. Cultured cells were harvested, and the amount of cell-associated 125I-lipoprotein was determined. In Experiment II, monocyte-macrophages were incubated in RPMI 1640 medium containing 10% human serum without or with 50 µg/ml AcLDL for 2 days (to cholesterol-enrich the macrophages) before incubations with 125I-VxLDL. Condition Cell-associated 125I-lipoprotein Degraded 125I-lipoprotein µg/mg cell protein Experiment I   Monocyte-macrophages     normal 125I-LDL 5 h 0.1  ± 0.0 0.1  ± 0.0     methylated 125I-LDL 5 h 0.3  ± 0.1 0.0  ± 0.0     normal 125I-VxLDL 5 h 6.4  ± 0.1 1.2  ± 0.1     methylated 125I-VxLDL 5 h 9.8  ± 0.7 0.3  ± 0.0     methylated 125I-VxLDL +     cytochalasin D 5 h 2.0  ± 0.4 0.0  ± 0.0   Fibroblasts     normal 125I-LDL 0.3  ± 0.1 0.4  ± 0.0     methylated 125I-LDL 0.1  ± 0.0 0.0  ± 0.0 Experiment II   Monocyte-macrophages     AcLDL 2 day/125I-VxLDL 5 h 9.6  ± 0.4 3.0  ± 0.2     +AcLDL 2 day/125I-VxLDL 5 h 8.6  ± 0.2 0.4  ± 0.1

After incubation with the indicated amount of ¹²⁵I-lipoprotein, monocyte-macrophages were rinsed 3 times with RPMI 1640 and incubated for 30 min in 1 ml of RPMI 1640 with 25 µl of DPBS plus Ca²⁺ and Mg²⁺ or that solution containing 50 µg of trypsin (2,740 units/mg of protein; Life Technologies Inc.) at 37 °C. Then cells were rinsed and assayed for cell-associated ¹²⁵I-lipoprotein and cell protein. Organic ¹²⁵I-radioactivities in soluble and trichloroacetic acid-insoluble fractions of media were also measured.

Monocyte-macrophages were rinsed and incubated with 1 ml of RPMI 1640 medium and ¹²⁵I-labeled lipoproteins containing 4 µg/ml cytochalasin D (Calbiochem) to demonstrate the actin dependence of cell association. Chloroquine and cytochalasin D sensitivity of degradation of cell-associated ¹²⁵I-VxLDL was determined with pulse-chase experiments. Monocyte-macrophages were incubated with ¹²⁵I-VxLDL in 1 ml of RPMI 1640 for 5 h, rinsed, and then incubated in RPMI 1640 without or with 100 µM chloroquine (Sigma) or 4 µg/ml cytochalasin D. For cytochalasin D controls, macrophages were incubated with 1 ml of RPMI 1640 containing 2 µl of Me₂SO, the solvent for cytochalasin D.

Unesterified and esterified cholesterol contents of macrophages were determined enzymatically according to the fluorometric method of Gamble et al. (17). For these assays, macrophages were harvested by scraping into distilled water and processed as described previously (12).

Extracellular and intracellular membranes were differentiated with ruthenium red according to the method of Luft (18). Monocyte-macrophage cultures were ruthenium red-stained and prepared for electron microscopy as described previously (1).

All data are presented as means ± S.E. The means were determined from 3 culture wells for each data point. No standard error is shown when it was smaller than the height of the graphic symbol.

RESULTS

Ruthenium red is an electron dense membrane stain. Because ruthenium red does not penetrate membranes of aldehyde-fixed cells, it stains only membranes in continuity with the extracellular space. Electron microscopic examination of macrophages incubated with VxLDL showed accumulation of lipoprotein particles within ruthenium red-outlined SCC not within phagocytic vacuoles (Fig. 1). As we reported previously (1), SCC were tortuous interconnecting membranous compartments. VxLDL also accumulated in some ruthenium red negative membranous compartments that had the same appearance as SCC. These ruthenium red negative membranous compartments most likely are SCC that were not open to the extracellular space at the time of fixation of the macrophages. The lipoprotein particles in all SCC were often larger than native LDL (which is about 22 nm) and often resembled beads on a string. The increased size of LDL particles can be explained by fusion of LDL particles induced by vortexing (19). LDL that aggregated by treatment with phospholipase C also induced and accumulated within SCC (Fig. 1c). The aggregates produced by phospholipase C were more rounded rather than linear as was VxLDL. No macrophage SCC were induced by LDL that had not been vortexed or by AcLDL centrifuged to remove any spontaneously formed aggregates.

[View Larger Version of this Image (148K GIF file)]

Because VxLDL entered SCC and centrifuged AcLDL did not enter these compartments, it was possible to compare the metabolism of AcLDL and VxLDL as they were processed by two different endocytosis pathways. The time courses of the metabolism of ¹²⁵I-VxLDL and ¹²⁵I-AcLDL were compared. Cell association of ¹²⁵I-VxLDL and ¹²⁵I-AcLDL both began to reach a plateau by 2 h. However, the cell association of ¹²⁵I-VxLDL was greater than 5-times that of ¹²⁵I-AcLDL by this time (Fig. 2a). Cell association of nonvortexed LDL was small compared with that of ¹²⁵I-AcLDL and ¹²⁵I-VxLDL.

[View Larger Version of this Image (18K GIF file)]

Though cell association of ¹²⁵I-VxLDL was high compared with ¹²⁵I-AcLDL, degradation of ¹²⁵I-VxLDL was less (about one-half at 4 h) than the degradation of AcLDL (Fig. 2b). Degradation of ¹²⁵I-VxLDL compared with ¹²⁵I-AcLDL could even be lower than this. In another experiment not shown, although ¹²⁵I-VxLDL cell association was 4 times that of ¹²⁵I-AcLDL by 3 h, degradation of the ¹²⁵I-VxLDL was only about one-fifth that of ¹²⁵I-AcLDL and was not significantly different from the degradation of nonvortexed ¹²⁵I-LDL.

Part of the difference in the amounts of degradation of ¹²⁵I-VxLDL and ¹²⁵I-AcLDL could be explained by differences in their initial rates of degradation. ¹²⁵I-AcLDL showed a linear increase in its degradation during incubation with the macrophages. In contrast, there was a delay in degradation of ¹²⁵I-VxLDL. Its maximal rate of degradation was not reached until 1 h after its addition to macrophage cultures. However, even when comparing their maximal rates of degradation (rather than their absolute amounts of degradation), ¹²⁵I-VxLDL was degraded at <55% the rate of degradation of ¹²⁵I-AcLDL.

Uptake into SCC was an active cell-mediated process that was dependent on temperature and actin microfilament function. Cell association of ¹²⁵I-VxLDL was 10 times greater at 37 °C compared with 4 °C and 5 times greater at 37 °C compared with 20 °C (Fig. 3). Cytochalasin D, an inhibitor of actin microfilament function (20), inhibited cell association of ¹²⁵I-VxLDL by 72 ± 3.5% (n = 3 experiments) during 5 h of incubation with 50 µg/ml ¹²⁵I-VxLDL (see Fig. 4 for results of one experiment). Nocodazole, an inhibitor of microtubule function, did not inhibit cell association (data not shown). Electron microscopy showed that cytochalasin D blocked formation of SCC induced by VxLDL similar to what we have observed for compartments induced by cholesterol crystals (1). The temperature dependence and cytochalasin D sensitivity of ¹²⁵I-VxLDL cell association showed that most cell association occurred through an active cell-mediated process and was not due to nonspecific sticking of lipoprotein aggregates to the culture dish.

[View Larger Version of this Image (15K GIF file)]

[View Larger Version of this Image (38K GIF file)]

After cell association of 20 µg/ml ¹²⁵I-VxLDL with macrophages for 5 h, then up to 90% could be released from the macrophages with subsequent trypsin treatment (Fig. 5). Incubation of macrophages for 5 h with 50 µg/ml ¹²⁵I-VxLDL showed an average (of three experiments) cell association of 19.9 ± 1.3 µg ¹²⁵I-VxLDL/mg of cell protein, 67 ± 2% of which could be released by trypsin. There was a progressive decrease with time in the percentage of cell-associated ¹²⁵I-VxLDL that could be released by trypsin (Fig. 4). 86, 70, and 64% cell-associated ¹²⁵I-VxLDL could be released after 1-, 3-, and 5 h-incubations with 50 µg/ml ¹²⁵I-VxLDL, respectively. Between 5 and 24 h, there was no further decrease in the percent of cell-associated ¹²⁵I-VxLDL that could be released by trypsin. Cell association of ¹²⁵I-LDL that had been aggregated with phospholipase C was less than that of ¹²⁵I-VxLDL. However, similar to ¹²⁵I-VxLDL, most cell association of phospholipase C-treated ¹²⁵I-LDL was inhibited by cytochalasin D and could be released by trypsin (Fig. 6).

[View Larger Version of this Image (22K GIF file)]

[View Larger Version of this Image (30K GIF file)]

About one-third of cell-associated ¹²⁵I-VxLDL could not be released with trypsin. However, this trypsin-resistant cell-associated ¹²⁵I-VxLDL did accumulate within macrophages by an actin-dependent process. This was shown by incubating macrophages for 5 h with 50 µg/ml ¹²⁵I-VxLDL in the presence of cytochalasin D and then exposing the macrophages to trypsin. In this case, release of cell-associated ¹²⁵I-VxLDL was almost complete, being 4 ± 0% that of the amount found with macrophages not treated with cytochalasin D or trypsin. The occurrence of some trypsin-resistant cell-associated ¹²⁵I-VxLDL could be explained by the electron microscopic findings of some VxLDL having accumulated within ruthenium red negative SCC that lacked connections to the cell surface (see electron microscopy results above), incomplete release by trypsin of VxLDL from open SCC, and some uptake of VxLDL into non-SCC endocytic compartments. However, at a minimum, one-third VxLDL was taken up into SCC (i.e. one-third of cell-associated ¹²⁵I-VxLDL was both cytochalasin D-inhibitable and trypsin-releasable).

Degradation of cell-associated ¹²⁵I-VxLDL was incomplete. When macrophages were incubated with 50 µg/ml ¹²⁵I-VxLDL for 5 h and then incubated in RPMI 1640 medium without VxLDL for 1 and 2 days, only 38.2 ± 7.1% (n = 5 experiments) and 39.0 ± 9.6% (n = 3 experiments), respectively, cell-associated ¹²⁵I-VxLDL was subsequently degraded.

Two-thirds of the degradation of cell-associated ¹²⁵I-VxLDL could be accounted for by the degradation of the pool of ¹²⁵I-VxLDL that was trypsin-releasable (i.e. was within SCC or associated with macrophage cell surfaces). This was shown by incubating monocyte-macrophages with 50 µg/ml ¹²⁵I-VxLDL for 5 h to produce a large pool of cell-associated ¹²⁵I-VxLDL. Then, the macrophages were treated without or with trypsin for 30 min as described in Fig. 4 to remove ¹²⁵I-VxLDL from SCC. The amount of ¹²⁵I-VxLDL degraded during a subsequent 24-h incubation in RPMI 1640 medium was determined for those macrophages with ¹²⁵I-VxLDL accumulated in SCC and for those macrophages that had their ¹²⁵I-VxLDL removed by trypsin treatment. Trypsin treatment removed 66 ± 1% cell-associated ¹²⁵I-VxLDL, and this decreased subsequent ¹²⁵I-VxLDL degradation during the chase period by a similar amount at 64 ± 1%.

Degradation of ¹²⁵I-VxLDL occurred in lysosomes. When macrophages were first incubated 5 h with ¹²⁵I-VxLDL and then chased 1 day in RPMI 1640 medium with the lysosome inhibitor chloroquine, degradation of cell-associated ¹²⁵I-VxLDL was inhibited >90% (Table I). Chloroquine inhibition of ¹²⁵I-VxLDL degradation resulted in a 2-fold increase in loss of nondegraded (i.e. trichloroacetic acid-insoluble) ¹²⁵I-VxLDL from the macrophages during the chase period. However, most (80%) of the ¹²⁵I-VxLDL that chloroquine blocked from being degraded remained associated with the macrophages. This was shown by the observation that cell-associated ¹²⁵I-VxLDL was increased for macrophages incubated with chloroquine compared with cell-associated ¹²⁵I-VxLDL for macrophages incubated without chloroquine (Table I).

Table I. Effect of chloroquine on degradation of 125I-VxLDL Two-week-old monocyte-macrophage cultures were incubated with 50 µg/ml 125I-VxLDL for 5 h, and then the amount of cell-associated 125I-VxLDL was determined. Other duplicate sets of cultures were rinsed and incubated without 125I-VxLDL for 24 h with or without 100 µM chloroquine. Then media were collected from these cultures for determination of TCA-soluble and insoluble organic 125I, and macrophages were harvested for determination of cell-associated 125I-VxLDL. Condition Cell-associated 125I-VxLDL TCA-soluble organic 125I in medium TCA-insoluble organic 125I in medium Cell-associated 125I-VxLDL plus organic 125I in medium µg/mg cell protein 125I-VxLDL 5 h 16.6  ± 2.1 125I-VxLDL 5 h/RPMI 1640    without chloroquine 1 day 8.9  ± 0.2 6.2  ± 0.2 1.2  ± 0.1 16.3  ± 0.1   with chloroquine 1 day 13.3  ± 1.0 0.6  ± 0.0 2.5  ± 0.2 16.4  ± 1.2

Above, it was shown that cytochalasin D blocked uptake into SCC. We also determined whether cytochalasin D blocked subsequent degradation of ¹²⁵I-VxLDL that was already accumulated in SCC. Macrophages were first incubated 5 h with 50 µg/ml ¹²⁵I-VxLDL and then chased 1 day in RPMI 1640 medium without or with cytochalasin D. During the chase period 5.5 ± 0.2 µg/mg of cell protein of ¹²⁵I-VxLDL was degraded without cytochalasin D, but only 0.9 ± 0.1 µg/mg of cell protein of ¹²⁵I-VxLDL was degraded with cytochalasin D. Thus, transport of ¹²⁵I-VxLDL from SCC to lysosomes was also an actin-dependent process.

Cell-association of ¹²⁵I-VxLDL was not mediated through the classical LDL receptor, but some of its degradation may have been. Reductive methylation of LDL is known to prevent its binding to the macrophage LDL receptor (21). However, methylated ¹²⁵I-VxLDL became cell-associated with monocyte-macrophages through a cytochalasin D-inhibitable process that could not have been mediated by the LDL receptor (Experiment I in Table II). Also, a one-day incubation of monocyte-macrophages with RPMI 1640 medium containing 200 µg/ml methylated VxLDL caused the total cholesterol content of the macrophages to increase from 83 ± 2 to 260 ± 9 nmol/mg of cell protein. Eighty-eight percent of the increase in cholesterol content could be inhibited by cytochalasin D. Moreover, electron microscopy showed that methylated VxLDL induced and entered SCC similar to that shown for VxLDL in Fig. 1.

On the other hand, although cell association of methylated ¹²⁵I-VxLDL was even greater than that of ¹²⁵I-VxLDL, degradation of methylated ¹²⁵I-VxLDL was only 25% compared with normal ¹²⁵I-VxLDL. This suggested that up to 75% degradation of ¹²⁵I-VxLDL may have been mediated through the LDL-receptor. Dissociation of ¹²⁵I-VxLDL degradation mediated by the LDL receptor and ¹²⁵I-VxLDL cell association was confirmed another way. The LDL receptor in human monocyte-macrophages is down-regulated by cholesterol enrichment (22). Incubation of monocyte-macrophages with 50 µg/ml AcLDL for 2 days increased macrophage cholesterol content about 2-fold (from 94 ± 2 to 205 ± 7 nmol/mg of cell protein). Cholesterol enrichment of the macrophages decreased ¹²⁵I-VxLDL degradation by 85% but only decreased ¹²⁵I-VxLDL cell association by 10% (Experiment II in Table II).

DISCUSSION

We have shown that LDL aggregated by vortexing or treatment with phospholipase C (and vortexed AcLDL)² induced and accumulated within SCC of human monocyte-derived macrophages. Uptake of aggregated LDL into SCC is a novel endocytosis pathway for this lipoprotein. Native nonaggregated LDL did not induce nor enter these compartments. Uptake into SCC was temperature- and actin-dependent. VxLDL that accumulated in SCC was degraded slowly compared with AcLDL. Though cell association of VxLDL was an active cell-mediated process, nevertheless cell-associated VxLDL could be released from macrophages by trypsin. This confirmed the electron microscopic finding that VxLDL had accumulated within SCC rather than within phagocytic vacuoles.

Accumulation of aggregated lipoproteins in SCC may have been overlooked in previous studies. Often, cells are treated with protease to remove cell surface-associated lipoproteins that are considered to represent lipoproteins not actively taken up by the cells but rather specifically and nonspecifically bound to cell surfaces (8, 23). However, our study shows that at least for human monocyte-macrophages, protease-releasable lipoproteins can actively accumulate within SCC.

Previously, uptake of aggregated lipoproteins has been attributed to the process of phagocytosis. This was because uptake of aggregated lipoproteins was inhibited by cytochalasins and phagocytosis is an actin-dependent process that is inhibitable by cytochalasins (20). Actin dependence does not necessarily mean that phagocytosis is the uptake pathway. This is because, besides phagocytosis, other endocytotic pathways of macrophages such as sequestration into SCC and macropinocytosis are also actin-dependent (1, 24). Electron microscopy was not used in many previous studies to confirm that aggregated lipoproteins entered macrophages by direct phagocytosis. Not only did cytochalasin D inhibit uptake of VxLDL into SCC, cytochalasin D also inhibited subsequent degradation of SCC-bound VxLDL. Thus, it remains possible that VxLDL first entered SCC but then was transported from SCC into the cell cytoplasm by phagocytosis. However, we were not able to visualize by electron microscopy transport of VxLDL out of SCC, possibly because this process was slow and limited.

Uptake into SCC is a process different from phagocytosis. Phagocytosis involves uptake of material into vacuoles that form from pinched-off regions of the plasma membrane (2). SCC form in part through invaginations of the plasma membrane that do not pinch off from the plasma membrane (1). The formed SCC can be delineated with the electron-dense stain ruthenium red that distinguishes intra- from extracellular membranes by staining only the latter. Hoff et al. (25, 26) showed rapid degradation of aggregates of 4-hydroxynonenal-modified LDL taken up by mouse peritoneal macrophages through phagocytosis (verified by electron microscopy). On the other hand, these investigators reported that VxLDL showed delayed processing (i.e. degradation) by macrophages (Ref. 27; also shown in Ref. 23). The delayed processing of VxLDL was not due to a diminished capacity of lysosomal enzymes to degrade VxLDL (also shown in Ref. 28). They speculated that poor processing of VxLDL could have been due to the nature of the intracellular pathway taken by the VxLDL. Uptake of VxLDL into SCC is an intracellular pathway showing slow degradation that could explain these earlier findings.

Human monocyte-macrophages are capable of directly phagocytosing 0.5-3-µm latex beads.² So what property of VxLDL targets this aggregate to SCC rather than to phagocytic vacuoles? VxLDL is multivalent with respect to its apolipoprotein B component, and valency of a ligand may target receptors to different endocytosis pathways. Tabas and co-workers (29, 30) described a unique endocytotic pathway for  very low density lipoproteins (VLDL) in mouse peritoneal macrophages in which multi-valency of ligand was a factor determining the route of endocytosis. In this pathway, large -VLDL enter mouse peritoneal macrophages through surface-connected tubules. Surface-connected tubules are plasma membrane invaginations that do not terminate in a labyrinth of interconnected compartments as do SCC. In contrast to the uptake of VxLDL into SCC in the present study, uptake of -VLDL into surface-connected tubules is not inhibited by cytochalasin D. Also, uptake into these tubules occurs through the LDL receptor that binds the multivalent apolipoprotein E component of -VLDL (30). Thus, uptake of -VLDL by mouse peritoneal macrophages through the surface-connected tubule pathway and uptake of VxLDL by human monocyte-macrophages through the SCC pathway appear to be different but possibly related endocytic pathways.

It is possible that the specific macrophage uptake pathway taken by particles depends on the nature of the receptor that binds the particles. The receptor that mediated cell association of VxLDL was not the classical LDL receptor. Methylation of LDL, which prevents binding of LDL to the LDL receptor (21), did not block macrophage cell association of methylated ¹²⁵I-VxLDL, cholesterol accumulation by macrophages incubated with methylated VxLDL, or accumulation of methylated VxLDL within SCC. Similar to what Khoo et al. (11) found with mouse peritoneal macrophages, we observed that the LDL receptor did mediate much of the degradation of ¹²⁵I-VxLDL. Macrophage degradation of ¹²⁵I-VxLDL was partially inhibited by methylation of LDL and by macrophage cholesterol enrichment, a treatment previously shown to down-regulate the LDL receptor in macrophages (22). The results suggest that whereas uptake into SCC occurs independently of the LDL receptor, subsequent movement from SCC to lysosomes may depend in part on LDL receptor function.

It is conceivable that vortexing of LDL reveals a binding domain in LDL that is normally masked. Such a domain could then mediate LDL binding to some other receptor capable of uptake into SCC. In this regard, Gianturco et al. (31, 32) have shown the existence of a macrophage receptor that recognizes a binding domain within apolipoprotein B-48 but which is not expressed by native LDL. Alternatively, vortexing of LDL could reveal a cryptic lipid ligand. In this regard, uptake of oxidized LDL appears to be mediated in part by a lipid moiety of the oxidized LDL (33). Trypsin treatment of the VxLDL in our study did prevent its uptake by monocyte-macrophages.² However, trypsin treatment also reversed the aggregation of VxLDL. Thus, it was not clear whether loss of a protein ligand or loss of multi-valency of aggregated VxLDL was responsible for the lack of macrophage uptake of the trypsin-treated VxLDL.

Uptake into SCC is a pathway that leads to cellular accumulation of at least two different types of aggregated LDL (e.g. vortexed and phospholipase C-treated LDL). In preliminary experiments, we have observed macrophage uptake into SCC for some lots of oxidized LDL, another form of LDL that can aggregate (27). Considering the role of SCC in the uptake of aggregated lipoproteins will be important in future studies. Furthermore, uptake of aggregated LDL by the non-LDL receptor pathway shown in this study could be important in cellular lipoprotein uptake in atherosclerotic lesions where activity of the classical LDL receptor is down-regulated (34).