Comparative binding and degradation of lipoprotein(a) and low density lipoprotein by human monocyte-derived macrophages.

The binding and degradation of equimolar concentrations of lipoprotein(a) (Lp(a)) and low density lipoprotein (LDL) isolated from the same individual were studied in primary cultures of human monocyte-derived macrophages (HMDM). At 4 degrees C, LDL receptor-mediated binding of both Lp(a) and LDL was of low affinity, being 0.8 and 0.23 microM, respectively. Competitive binding studies indicated that the binding of Lp(a) to HMDM was competed 63% by excess LDL. In contrast to the 4 degrees C binding data, the degradation of Lp(a) at 37 degrees C was mainly nonspecific because the amount of Lp(a) processed by the LDL receptor pathway in 5 h was 17% that of LDL. According to pulse-chase experiments, this phenomenon may be accounted for by the facts that less Lp(a) is bound to HMDM at 37 degrees C and that Lp(a) has a lower intrinsic degradation rate and was not due to increased intracellular accumulation or retroendocytosis of the lipoprotein. Degradation of both lipoproteins was primarily lysosomal and only modestly affected by up- or down-regulation of the LDL receptor. The rate of retroendocytosis in HMDM was approximately equal to the degradation rate and appeared to be independent of the type of lipoprotein used, up- or down-regulation of the LDL receptor, or the presence of the lysosomotropic agent chloroquine. Overall, the results indicate that HMDM degrade Lp(a) mainly via a nonspecific pathway with only 25% of total Lp(a) degradation occurring through the LDL receptor pathway. As both 37 degrees C degradation and 4 degrees C binding of LDL are mainly LDL receptor specific, the different metabolic behavior observed at 37 degrees C suggests that Lp(a) undergoes temperature-induced conformational changes on cooling to 4 degrees C that allows better recognition of Lp(a) by the LDL receptor at a temperature lower than the physiological temperature of 37 degrees C. How apo(a) affects these structural changes remains to be established.

The binding and degradation of equimolar concentrations of lipoprotein(a) (Lp(a)) and low density lipoprotein (LDL) isolated from the same individual were studied in primary cultures of human monocyte-derived macrophages (HMDM). At 4 "C, LDL receptormediated binding of both Lp(a) and LDL was of low affinity, being 0.8 and 0.23 PM, respectively. Competitive binding studies indicated that the binding of Lp(a) to HMDM was competed 63% by excess LDL. In contrast to the 4 O C binding data, the degradation of Lp(a) at 37 "C was mainly nonspecific because the amount of Lp(a) processed by the LDL receptor pathway in 5 h was 17% that of LDL. According to pulse-chase experiments, this phenomenon may be accounted for by the facts that less Lp(a) is bound to HMDM at 37 O C and that Lp(a) has a lower intrinsic degradation rate and was not due to increased intracellular accumulation or retroendocytosis of the lipoprotein. Degradation of both lipoproteins was primarily lysosomal and only modestly affected by up-or down-regulation of the LDL receptor. The rate of retroendocytosis in HMDM was approximately equal to the degradation rate and appeared to be independent of the type of lipoprotein used, up-or down-regulation of the LDL receptor, or the presence of the lysosomotropic agent chloroquine. Overall, the results indicate that HMDM degrade Lp(a) mainly via a nonspecific pathway with only 25% of total Lp(a) degradation occurring through the LDL receptor pathway. As both 37 "C degradation and 4 "C binding of LDL are mainly LDL receptor specific, the different metabolic behavior observed at 37 "C suggests that Lp(a) undergoes temperature-induced conformational changes on cooling to 4 O C that allows better recognition of Lp(a) by the LDL receptor at a temperature lower than the physiological temperature of 37 "C. How apo(a) affects these structural changes remains to be established.
Evidence indicating that high plasma levels of Lp(a)' place * This work was supported by United States Public Health Service Program Project Grant HL-185'77 and was carried out in part in the National Institutes of Health Shared Research Facility supported by National Heart, Lung, and Blood Institute Grant 35361. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adoertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
)I To whom correspondence should be addressed Dept. of Medicine, Box 81, University of Chicago 5841 S. Maryland Ave., Chicago, IL 60637. The abbreviations used are: Lp(a), lipoprotein(a); LDL, low density lipoprotein; DMEM, Dulbecco's modified Eagle's medium; HSA, human serum albumin; PBS(CMF), phosphate-buffered saline minus calcium and magnesium; LPDS, lipoprotein-deficient serum; EACA, 6-amino hexanoic acid; HEPES, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; SDS, sodium dodecyl sulfate; HMDM, human monocyte-derived macrophages. individuals at risk for cardiovascular disease has led to increased interest in the role played by the LDL receptor in the catabolism of Lp(a) by cultured cells. To date, most studies on the binding and degradation of Lp(a) have been conducted in fibroblasts without coming to a consensus as to the role played by the LDL receptor pathway (1)(2)(3)(4)(5)(6)(7)(8)(9). In particular, quantitative comparisons of Lp(a) and LDL have been highly variable. This may have been due to a lack of appreciation in some of the earlier studies of Lp(a) heterogeneity, Lp(a) stability, and the difficulty of obtaining Lp(a) preparations that are free of LDL. For example, some Lp(a) species with large apo(a) polymorphs aggregate in the cold, thus potentially compromising the results of the 4 "C binding studies (10, 11).
Also, Lp(a) that is isolated in the absence of protease inhibitors may undergo degradation of the apo(a) moiety thereby creating an LDL-like particle (12).
Lacking knowledge of the molecular weight of Lp(a) or Lp(a) protein, comparative studies of Lp(a) and LDL have been made on an equal protein or cholesterol basis or on equal apoB concentration (determined immunochemically) (1)(2)(3)(4)(5)(6)(7)(8)(9). When protein is used as the basis of comparison, the molar Lp(a) concentration is actually 2-%fold lower than that of LDL, whereas when comparisons are made on equal apoB concentrations, Lp(a) levels may actually be up to 2-fold higher than those of LDL because of the reduced affinity of anti-apoB antibodies for Lp(a) relative to LDL (13). Although the variablity in the molarity of Lp(a) and LDL may be less when comparisons are made on an equal cholesterol basis, this is also not ideal because the cholesterol content of Lp(a) and LDL is not fixed (10,14).
In previous studies, we determined the molecular weight of Lp(a) protein from an individual having Lp(a) with a fast apo(a) polymorph (M, = 281,000) by compositional and ultracentrifugal analysis (10,14). In the present study we have utilized this well-characterized Lp(a) along with LDL isolated from the same individual in order to determine the role of the LDL receptor pathway in the cellular binding and degradation of these two lipoproteins. In addition, experiments were performed and analyzed on a molar basis. We chose the human monocyte-derived macrophage (HMDM) as a model cell type because it is an important component of atheromatous lesions (15,16) and is known to possess receptors of high specificity for LDL (17-19) which may, however, not be identical to the LDL receptor present on fibroblasts (20). Our studies show that HMDM degrade Lp(a) at a rate slower than LDL. A preliminary account of these results has been reported (21).
Preparation of Lpfa) and LDL-Autologous LDL and Lp(a) were isolated from the plasma of two subjects with a fast apo(a) polymorph (M, = 281,000 (14)). Prior to donation, all subjects gave informed consent. The molecular mass of the protein moiety of this Lp(a) was determined previously from the molecular mass and chemical composition of Lp(a) (10,14) and was found to be 927,000 daltons. This molecular weight is also consistent with the presence of 2 mol of apo(a) and 1 mol of apoB/Lp(a) molecule. Because 28.1% of the mass of apo(a) is carbohydrate, the protein portion of 2 mol of apo(a) is equivalent to 404,000 daltons and when added to the mass of apoB results in a mass of 918,000 daltons for the protein moiety of Lp(a).
Lp(a) was purified from the d < 1.21 g/ml total lipoproteins by lyslne-Sepharose chromatography. Total lipoproteins in 0.1 M phosphate buffer containing 0.01% Na,-EDTA, 0.01% NaN3, and 1 mM benzamidine, pH 7.4, were passed over a column (20 x 5 cm') containing lysine-Sepharose at 50 ml/h. Following application of the lipoproteins, the column was washed with 50 ml of 0.5 M NaCl, 0.1 M NaHCO,, 1 mM benzamidine, pH 8.3, to remove nonspecifically bound lipoproteins. Lp(a) was eluted with 20 mM 6-amino hexanoic acid (EACA) dissolved in 0.1 M phosphate, 1 mM benzamidine, and 0.01% Na2-EDTA and NaN3, pH 7.4. This heterogenous Lp(a) fraction was made 7.5 wt % with CsCl and centrifuged 20 h at 49,000 rpm a t 20 "C in the 50.2 Ti rotor in quick seal tubes. The CsCl forms a selfgenerating density gradient which allows the separation of Lp(a) species with different apo(a) polymorphs and also serves to remove any contaminating plasminogen. LDL was isolated from the unbound lipoproteins obtained after lysine-Sepharose chromatography by density gradient centrifugation in 3.75 wt % CsCl (Ti-50.2 rotor, 49,000 rpm, 20 h, 20 "C). Lp(a) and LDL preparations were checked for purity by SDS-gradient gel electrophoresis on 2-16% acrylamide gels (Pharmacia, Uppsala, Sweden). The Lp(a) and LDL (15 pglsample) were incubated with 1% SDS for 5 min at 95 "C. The gels were stained with Coomassie Blue R-250. Both lipoproteins were stored filter sterilized (0.45 pm) in Sarsted vials, filled to allow no airspace, in 0.15 M NaCI, pH 7.4, containing 0.01% Na,-EDTA and NaN,.
Iodination of Lpfa) and LDL-The radioiodination of Lp(a) and LDL was performed using the iodine monochloride method of Mc-Farlane (22) as modified by Bilheimer et al. (23). The specific radioactivity for both lipoproteins averaged 600 cpm/ng lipoprotein protein. The integrity of the radiolabeled Lp(a) and LDL were checked by SDS-gradient gel electrophoresis (2-16%) as described above. The "'I-labeled Lp(a) and LDL (approximately 100,000 cpm/sample) were transferred to nitrocellulose, and autoradiography was performed with Kodak XAR-5 film. The percent of the radioactive label incorporated into apo(a) and apoBloo was determined by two different methods. Apo(a) was separated from the parent radioiodinated lipoprotein after reduction of Lp(a) with 50 mM dithiothreitol and rate zonal density gradient centrifugation (14). The floating lipoprotein particle containing apoBlOn but not apo(a), and apo(a) from the centrifuge tube bottom were dialyzed against 0.15 M NaCl and counted for radioactivity. After correcting for the radioactivity found in the lipid moiety (4.5%), the percentage of label in apo(a) was 55 and 45% in apoBllll,. We also determined this distribution for two different preparations of radioiodinated Lp(a) reduced with 5% mercaptoethanol by SDS-gradient polyacrylamide gel electrophoresis. After visualizing the two apoprotein bands with Coomassie Blue, they were excised from the gel and counted for radioactivity. The mean percent distribution of radioactivity in apo(a) was 58.2 f 1.6 and 41.8 f 1.6 in apoBlm.
These numbers corresponded well with the distribution of tyrosines in the two apoproteins as calculated from their respective amino acid sequences which were 62.3% in apo(a) and 37.7% in ap~Bloo and appears to indicate that neither apoprotein is preferentially labeled.
The mock iodination of Lp(a) was carried out using the same conditions employed with the radioiodinations except that for each milligram of lipoprotein protein, 0.6 nmol of cold NaI was used in place of 0.6 nmol of Na'2"I.
Isolation and Culture of HMDM-Isolation of human monocytes from donors who gave informed consent was carried out according to Fogelman et al. (24) with minor changes. Briefly, freshly drawn human blood (300 ml) was collected into sterile 50-ml plastic centrifuge tubes at a final heparin concentration of 6 units/ml. An additional sample (50 ml) was collected into a citrate/phosphate/dextrose solution at a blood to citrate/phosphate/dextrose solution ratio of 6:l (v/v). A final 150 ml of blood was collected without anticoagulant for the preparation of autologous serum and LPDS.
The heparinized blood was mixed with 100 ml of Plasmagel (Cellular Products, Inc.) in a 500-ml Teflon separatory funnel. The red cells were allowed to settle 60-75 min at room temperature at which time they were slowly run out the bottom of the funnel. After removal of the red cells, the preparation of leukocytes, citrate/phosphate/ dextrose containing plasma, the hypertonic treatment of the cells, density gradient centrifugation, and preparation of monocytes for culture proceeded as outlined by Fogelman et al. (24). After the cells were counted, they were resuspended in DMEM (20% autologous serum) to a final concentration of 1.0 X lo6 cells/ml; 1.0 ml of cell suspension was plated into each well of Linbro 12-well tissue culture plates (area/well = 4.5 cm'). The cells were placed in a 37 "C humidified-air incubator (95% air, 5% CO,) and allowed to attach for 30 min, after which time they were washed twice with serum-free DMEM to remove any contaminating lymphocytes and platelets and finally overlaid with 1.0 ml of DMEM (20% autologous serum). The medium was replaced on the second and fifth day of culture for 8-day-old cells and the first, fourth, and seventh day for 10-day-old cells.
Twenty-four h prior to the initiation of experiments, the cells were washed twice with serum-free DMEM and placed into DMEM containing either 20% autologous LPDS to upregulate the LDL receptor or into 20% autologous serum to down-regulate the LDL receptor. For the competitive binding and degradation experiments in which 10-day-old cells were used, the preincubation with 20% autologous LPDS was eliminated. We came to this decision after discovering that our schedule of medium changes mimicked up-regulated cells so that LDL-specific binding of either LDL or Lp(a) was unchanged, and LDL-specific degradation of LDL and Lp(a) was only 12.1 and 10.5% lower, respectively, than in the up-regulated cells. The preparation of autologous serum and autologous lipoprotein-deficient serum was carried out as previously described (19).
HMDM were classified and their viability determined as described by Fogelman et al. (25). The cells were >99% monocyte-macrophages by the time the experiments were performed. There was no change in cell viability after either binding or degradation experiments. 4 "C Binding-Lipoprotein binding assays at 4 "C with 8-10-dayold HMDM were carried out essentially according to Innerarity et al. (26). Briefly, macrophage monolayers were washed once with DMEM, then twice with DMEM (0.2% human serum albumin (HSA)) and cooled to 4 "C. The medium was then removed and DMEM (0.2% HSA) containing Iz5I-labeled lipoprotein with and without unlabeled competitor was added. Dishes were incubated at 4 "C, with rocking for 4 h. The cells were then washed four times with 1.0 ml/well of ice-cold PBS (0.2% bovine serum albumin) and once with 1.0 ml/well of ice-cold PBS(CMF); the first wash was allowed to sit 10 min before proceeding with the remaining four washes in rapid succession as described by Haberland et al. (27). The cell monolayer was then dissolved with two additions of 0.5 ml of 0.1 N NaOH and counted for radioactivity, and an aliquot was removed to determine cellular protein by the method of Lowry et al. (28)  Briefly, cell monolayers were washed once with DMEM then twice with DMEM (0.2% HSA). DMEM (0.2% HSA) containing various concentrations of either T -L p ( a ) or '*'I-LDL with and without unlabeled competitor was added to the cells, and the cells were returned to the 37 "C incubator for 5 h. Proteolytic degradation of 1251-labeled lipoproteins was measured by assaying the amount of '?'Ilabeled trichloroacetic acid-soluble (noniodide) material formed by the cells and excreted into the medium. Cell monolayers were then washed, lysed, and protein determined as described for the binding assays.
Retroendocytosis-Retroendocytosis experiments were performed essentially as described by Greenspan et al. (31). Briefly, 8-day-old HMDM were incubated for 24 h in DMEM containing either 20% autologous LPDS or 20% autologous serum. The cells were washed once with DMEM then twice with DMEM (0.2% HSA). Fresh DMEM (0.2% HSA) containing either 100 nM '*'I-Lp(a) or "'I-LDL was added to the cells which were then returned to the 37 "C incubator for 4 h. This period is referred to as the "pulse period." The culture medium from the pulse period was removed and the amount of ' 9labeled trichloroacetic acid-soluble (noniodide) material formed was determined as described by Goldstein et al. (30). The cells were then chilled on ice, washed four times with 1.0 ml/well of ice-cold PBS(CMF), and incubated a t 4 "C with heparin (10 mg/ml) with shaking to release the cell surface bound "'1-lipoproteins. After 1 h, the heparin containing medium was removed and counted to determine the amount of "'I-Lp(a) or "'1-LDL released. The cells were washed again several times with ice-cold PBS(CMF), once with DMEM containing 20% LPDS, and warmed to 37 "C before being chased for 2 h with DMEM (20% autologous LPDS) at 37 "C. The chase period culture medium was removed after 2 h, and lipoprotein degradation was determined as described above. The resulting precipitate was washed once with 10% trichloroacetic acid and retroendocytosis was determined as picomoles of trichloroacetic acid-precipitable lZ'II-material released into the chase medium/mg cell protein.
The cells were again washed several times with ice-cold PBS(CMF) and lysed with 0.1 N NaOH. This material was counted to determine the amount of internalized "'I-Lp(a) or "'I-LDL remaining in the cells after the chase period, and an aliquot was taken for the determination of protein. When employed, chloroquine (100 p~) was dissolved in the culture medium of the pulse and chase periods.

RESULTS
4 "C Binding Studies-The 4 "C binding of Lp(a) to 8-dayold up-regulated human monocyte-derived macrophages was compared to that of LDL. All comparisons were made on a molar basis using a molecular weight of 514,000 for the protein moiety of LDL (32), and a molecular weight of 918,000 for the protein moiety of Lp(a) (10, 14). The integrity and purity of the hot and cold ligands as determined by SDS-gradient gel electrophoresis are shown in Fig. 1. Representative binding curves indicate substantial binding of Lp(a) to LDL receptors (see Fig. 2). Total binding for Lp(a) at 100 nM lipoprotein was 0.324 f 0.126 pmol/mg cell protein (n = 5) and 0.388 & 0.106 pmol/mg cell protein ( n = 5) for LDL. Specific binding was derived by subtracting the nonspecific binding curve (binding obtained in the presence of a 50-fold molar excess unlabeled LDL) from the total binding curve. Specific binding of Lp(a) at 100 nM lipoprotein as measured in three experiments was 79% that of LDL (0.253 f 0.055 pmol/mg cell protein for Lp(a) and 0.320 f 0.103 pmol/mg cell protein for LDL). In contrast to normal human skin fibroblasts which usually exhibit saturable binding with 10-20 nM LDL (5-10 pg/ml) (33), HMDM required more than 10-fold higher LDL concentrations for saturation. Since Lp(a) did not appear to saturate LDL-specific sites at 200 nM, three experiments were carried out using higher concentrations of the ligand up to 400 nM (see Fig. 3). Note that in panels A and B, lipoprotein concentration is plotted on a logarithmic scale in order to  total binding. Panels A and C refer to ""I-Lp(a) and panels R and D to "'I-LDL. In panels C and D, the solid line represents a computergenerated curve derived from analysis of the data by non-linear least square regression; the solid circles are the actual data points. 100 nM lipoprotein is equivalent to 91.8 pg/ml Lp(a) protein and 51.4 pg/ml LDL protein.
generate sigmoidal binding curves that plateau when all the binding sites are occupied or saturated (34). As can be seen, neither specific binding curve has reached saturation although the one for LDL appears to pass through an inflection point. Analysis of the two curves by non-linear least square regression (panels C and D) indicated that the binding of both Lp(a) and LDL to HMDM was of low affinity having a K,I of 0.8 PM for Lp(a) and 0.23 PM for LDL. Maximal lipoprotein binding for Lp(a) was 2.23 f 0.44 pmol/mg cell protein and for LDL 1.05 f 0.07 pmol/mg cell protein.
4 "Competitve Binding Assays-The LDL-specific nature of Lp(a) binding to HMDM was confirmed by competitive binding assays performed at 4 "C. As shown in Fig. 4, a 50-fold molar excess of unlabeled LDL was as effective as Lp(a) in inhibiting the binding of 25 nM '251-Lp(a) to HMDM. In fact, LDL inhibited the binding of '251-Lp(a) by 61.8 f 11.9% (n = 5) compared to 63.3 f 7.3% (n = 4) for Lp(a). Although not as effective as LDL, Lp(a) was nevertheless an efficient competitor for the binding of 25 nM '251-LDL. Unlabeled LDL inhibited binding of '251-LDL by 80.3 f 5.7% (n = 4) while Lp(a) competed with the binding of '251-LDL by 53.8 f 7.2% (n = 4). This difference between Lp(a) and LDL for competition with '251-LDL was significant at p 5 0.001.
37 "C Degradation Studies-In contrast to its cell surface binding at 4 "C, Lp(a) was degraded much less efficiently than LDL both in terms of total degradation and LDL-specific degradation. Representative degradation curves from one experiment are shown in Fig. 5 indicating that the rate of degradation of Lp(a) was much lower than that of LDL. Results averaged from five experiments indicate that at 100 nM lipoprotein, the total degradation of Lp(a) was only 31.8% that of LDL (1.51 f 0.47 pmol/mg cell protein for Lp(a) compared to 4.74 k 1.78 pmol/mg cell protein for LDL). LDLspecific degradation was obtained by subtracting the nonspecific degradation (degradation obtained in the presence of a 50-fold molar excess unlabeled LDL) from the total degradation of either lipoprotein. The LDL-specific degradation of Lp(a) was only 17.3% that of LDL. For Lp(a), the specific degradation was 0.738 f 0.062 pmol/mg cell protein (n = 3) and 4.26 f 0.18 pmol/mg cell protein (n = 3) for LDL. Both the total and specific degradation curves for l2'1-Lp(a) in HMDM were linear and non-saturable, whereas those of 1251-LDL appeared to saturate albeit at much higher concentrations than have been observed in normal fibroblasts (33). Formation of Cholesteryl r4C]Esters-The poor degradation of Lp(a) is reflected in its diminished capacity to induce the formation of cholesteryl esters relative to LDL. As shown in Fig. 6, at 100 nM lipoprotein, Lp(a) stimulated only 17% as much cholesterol esterification as did LDL. The doseresponse curve for Lp(a) appears to be linear up to 200 nM while the curve for LDL approaches saturation.
37 "C Competitive Degradation Studies-The results obtained from the 37 'C competitive degradation assays are at variance with those obtained from the 4 "C competitive bind-  ing assays. As expected, unlabeled LDL was an excellent competitor for the degradation of '251-LDL (25 nM) in that a 50-fold molar excess inhibited degradation by 75.6 f 3.6% (n = 6) (Fig. 7). In contrast, LDL was able to inhibit the degradation of 25 nM '"I-Lp ( In order to rule out this possibility, we mock iodinated Lp(a) with cold NaI in order to compare its ability to compete for lZ5I-Lp(a) with native Lp(a) at a 20-fold molar excess. The results indicated that native Lp(a) inhibited degradation 26% and mock iodinated Lp(a) 30%. This difference was not statistically different suggesting that the poor competition was not due to lipoprotein modification but more likely due to the inability to saturate all binding sites. Retroendocytosis Experiments-Several pulse-chase experiments were performed in an effort to resolve the discrepancy between the results obtained from the 4 "C binding and the 37 "C degradation assays and to determine specifically whether ( a ) the binding of Lp(a) differs at 37 "C from that at 4 "C, or ( b ) Lp(a) binds to macrophages but follows a degradation pathway different from that of LDL. The results showed, that while the total and specific binding of LDL and Lp(a) at 4 "C were approximately equal, at 37 "C the binding of Lp(a) was only one-third that of LDL (0.045 f 0.01 pmol/ mg cell protein for Lp(a) and 0.135 f 0.03 pmol/mg cell protein for LDL) (Table I). Second, the amount of Lp(a) internalized after the pulse was only 39% that of LDL indicating that Lp(a) was not accumulating inside the cell. This also demonstrated that the internalization rate of Lp(a) did not differ significantly from that of LDL because the binding ratio of Lp(a) to LDL was approximately equal to the internalization ratio. Third, the total degradation of Lp(a) during both the pulse and chase was only 25 and 28% that of LDL, respectively. Finally, the total amount of Lp(a) excreted into the medium was only 40% that of LDL indicating that there was no preferential retroendocytosis of Lp(a). The amount excreted was significant and for each lipoprotein represents 22% of the quantity that was internalized during the 4-h pulse. This value was approximately equal to the amount of each lipoprotein that was degraded during the 2-h chase, being 15% for Lp(a) and 22% for LDL. This difference in degradation was highly significant as shown by the paired Student's t test ( p < 0.001) indicating that LDL was degraded more efficiently than Lp(a).
The lysosomotropic agent chloroquine was used to deter-  chase Percent change refers to the difference between the processing in the presence of chloroquine and that when chloroquine is not present (Table I). mine whether Lp(a) was degraded in lysosomes. In the presence of 100 PM chloroquine, approximately 100% more LDL and Lp(a) were bound (Table 11). The percent change refers to the differences between the results presented in Table I1 and those in Table I. There was also a 200% increase in the amount of Lp(a) and LDL internalized during the pulse period and a 270% increase in the amount of Lp(a) and LDL that remained internalized after the chase period. The total degradation of LDL was drastically decreased by 82% during the pulse when chloroquine was present, whereas that of Lp(a) was decreased by 68%. The intrinsic degradation rate of Lp(a), as measured during the 2-h chase, decreased by 19% in the presence of chloroquine and that of LDL dropped by 29%. When expressed as the amount degraded relative to the amount of internalized lipoprotein, the degradation of both Lp(a) and LDL were drastically reduced from 15 to 4.2% and from 22 to 4.9%, respectively. In addition, the amount of Lp(a) and LDL excreted by retroendocytosis was increased 132 and 260%, respectively, when chloroquine was present. However, the percentage of either lipoprotein that was subject to retroendocytosis remained basically unchanged.
Regulation of Degradative Pathway-To determine whether the degradation of LDL and Lp(a) is influenced by LDL receptor activity in HMDM, half of the cells in three experiments were preincubated for 24 h with 20% autologous serum and the other half with 20% autologous LPDS. Values for the total and specific degradation of both lipoproteins at 100 nM are given in Table 111. For cells growing in the presence of an external cholesterol source, the specific degradation rate was decreased %fold for LDL and 2.2-fold for Lp(a) indicating that the regulation of LDL receptor activity on HMDM was relatively modest. The total degradation rate of LDL was decreased 2.4-fold and that of Lp(a) 1.7-fold.
We also performed a pulse-chase experiment to determine whether the amount of Lp(a) bound, internalized, and retroendocytosed was dependent upon up-or down-regulation of LDL receptor activity. In the down-regulated cells, the 37 "C heparin releasable binding of Lp(a) was reduced 18% in comparison to the up-regulated cells while LDL was reduced by 53% (compare Table IV to Table I). The amount of internalized Lp(a) and LDL was not significantly changed in the down-regulated cells. The degradation rate of LDL was only 40% that seen in the up-regulated cells during both the pulse and the chase while the Lp(a) degradation rate was reduced by 50% during the pulse and by only 13% during the chase. Lastly, the rate of retroendocytosis for both lipoproteins remained relatively unchanged being 23% for Lp(a) and 24% for LDL compared to 22% for both lipoproteins in the up-regulated cells.   (Table I).

DISCUSSION
The results presented here indicate that the LDL receptor pathway of HMDM is only minimally involved in the degradation of Lp(a) when compared on an equimolar basis to LDL. This is exemplified by the fact that unlabeled LDL at a 50-fold molar excess was able to inhibit the degradation of "'I-Lp(a) by only 26.2%. This finding is particularly surprising considering that at 4 "C, LDL was a good competitor for '"I-Lp(a) by reducing its binding by 61.8%. In contrast to the dichotomous behavior of Lp(a), excess LDL competed effectively for "'I-LDL in both 4 "C binding and 37 "C degradation experiments. Several mechanisms could account for these differences: 1 ) the binding of Lp(a) to HMDM at 37 "C is less than that observed at 4 "C; 2) Lp(a) binding is normal, however, the rate of internalization is less than that of LDL; 3 ) Lp(a) is bound and internalized normally but degraded slower than LDL; and 4 ) a greater percentage of Lp(a) than LDL could escape degradation by entering a retroendocytotic pathway.
The pulse-chase experiments designed to investigate these possibilities supported the first mechanism, namely, that at 37 "C, binding was less than at 4 "C. These experiments also ruled out the second hypothesis that was based on the possibility that the internalization rate of Lp(a) was less than that of LDL. In addition, the results partially supported the third mechanism that, in addition to lower binding, Lp(a) also had a lower intrinsic degradation rate when compared to LDL as determined during the 2-h chase period. Thus, both the decreased binding and slower degradation contributed to the overall lower degradation rate of Lp(a) and prevented overt accumulation of Lp(a) inside the cells. It is possible that the slow degradation rate of Lp(a) is related to the extensive glycosylation of apo(a) based on the fact that glycoproteins with high sugar content are known to be resistant to proteolysis (35). Finally, the fourth mechanism was ruled out by the observation that there was no preferential exocytosis of Lp(a) which could divert Lp(a) from the degradative pathway and account for the lower total degradation of this lipoprotein.
The conclusion that reduced degradation of Lp(a) relative to LDL can be accounted for by lower binding and degradation does not explain why the binding of Lp(a) to the LDL receptor of HMDM was so much better at 4 than at 37 "C. One possibility is that the conformation of these two lipoproteins is differentially affected by temperature. Some Lp(a) species with large apo(a) polymorphs are sensitive to changes in temperature, e.g. they can undergo a reversible cold-induced aggregation (ll), a process different from irreversible aggregation caused by foaming, oxidation, proteolysis, or bacterial degradation. It is conceivable that Lp(a) polymorphs that do not aggregate in the cold undergo conformational changes that affect the receptor-binding domain of apoB resulting in better interaction with the LDL receptor when the temperature is lowered to 4 "C.
Our studies showed that the affinity of LDL for the LDL receptor on HMDM is at least two orders ofmagnitude lower than in fibroblasts indicating that the LDL receptor in these two cells is probably different. In fact, in an experiment designed to compare the efficacy of heparin in releasing LDL uersus Lp(a) from the macrophage cell surface at 37 "C, we found that 20 mg/ml heparin was no more effective than PBS in releasing either lipoprotein from their binding sites which is indicative of low affinity binding. This conclusion is in keeping with the findings of Knight and Soutar (36), and Van Lenten and Fogelman (20) who demonstrated that HMDM have receptors with low affinity for LDL. Van Lenten and Fogelman proposed the existence of two classes of LDL recep- tor sites on HMDM, one class with high affinity that may correspond to the classical LDL receptor and a second, also specific for LDL, but of lower affinity. The latter was hypothesized to channel ligands into a different pathway and to operate a t high ligand concentrations such as those used in the present study.
With this in mind and because the degradation of Lp(a) has a large nonspecific component, it was of interest to determine the involvement of lysosomes in the processing of Lp(a). When we employed the lysosomotropic agent chloroquine which impairs vesicular transport of lipoproteins to lysosomes and raises organelle pH (37), we noted a substantial decrease in the total degradation of both Lp(a) and LDL, although that of LDL was comparatively greater, possibly because of a much smaller nonspecific component. It should be noted that the observed decrease in degradation of LDL in the presence of chloroquine was similar to that described by Shechter et al. (38). Since specific degradation of Lp(a) was one-quarter of the total degradation, one would expect only a 25% reduction in degradation if just the specific pathway was lysosomal in nature. However, because the observed decrease was approximately 70%, it was apparent that a sizeable portion of Lp(a) taken up nonspecifically was also degraded in lysosomes. In agreement with the observations of Shechter et al. (38), we found that there was a substantial intracellular accumulation of LDL. This increase was probably caused by the block in degradation, which also raised the cellular content of Lp(a) and accelerated the rate of retroendocytosis of both lipoproteins.
In contrast to the retroendocytosis rate, the percentage of internalized Lp(a) or LDL that was excreted into the medium appeared to be the same and was constant as indiciated by the linear relationship obtained from the plot of the quantity of retroendocytosed uersus internalized lipoprotein (P = 0.97) (see Fig. 8). From the slope, we calculated that the fraction of either internalized lipoprotein that was excreted by HMDM was 24.6% and appeared to be independent of the functional state of the LDL receptor, the presence of chloroquine, or the type of lipoprotein. It appears, therefore, that the early endocytic pathway of the two lipoproteins is similar since retroendocytosis did not differentiate between Lp(a) and LDL.
Finally, it should be noted that the retroendocytic pathway in HMDM is highly active since it is, at least for LDL, 3-10fold higher than that reported in fibroblasts or smooth muscle cells during similar pulse-chase periods (31,39,40). No comparisons for Lp(a) can be made since retroendocytosis of Lp(a) has not been studied in these cell lines.
Unlike the classic LDL receptor of fibroblasts, the LDL receptor activity of HMDM is comparatively less responsive to regulation by cellular cholesterol content. In agreement with Fogelman et al. (17) and Knight and Soutar (36), we also observed only a 2-3-fold increase in LDL degradation upon preincubation of HMDM with LPDS. Similarly, down regulation of the LDL receptor of HMDM resulted in decreased degradation of Lp(a). However, based on the low percentage of Lp(a) degraded by the LDL receptor pathway (26%)) the observed decrease in degradation appeared to be greater than expected. This decrease may be explained in part by the fact that down-regulation also caused a 30% decrease in nonspecific degradation of Lp(a) for reasons that are not clear at this time.
Low degradation of Lp(a) and an inability to induce formation of cholesteryl esters in comparison to LDL do not appear to support a significant involvement of the scavenger receptor of HMDM in the processing of Lp(a). This conclusion is in keeping with the preliminary results by Haberland et al. (41) that indicate that Lp(a) (same Lp(a) as was used in the present study) modified with malondialdehyde, but not native Lp(a), is taken up by the scavenger receptor of HMDM. These findings, however, are a t variance with the preliminary observations of Powell et al. (42), Gavish et al. (43), and Zioncheck et al. (44) who reported that Lp(a) binds to the scavenger receptor. The reason for this discrepancy is not apparent but may be due to differences in experimental conditions.
In summary, our results indicate that HMDM degrade Lp(a) mainly through a nonspecific mechanism(s) with only one-quarter being degraded by the LDL receptor pathway. In quantitative terms, only one-third as much Lp(a) as LDL is degraded when the two lipoproteins are compared on a molar basis: moreover, when LDL-specific degradation is compared, Lp(a) degradation is 6-fold lower than that of LDL. The exact reason for the low LDL-specific degradation of Lp(a) is not known but may be related to temperature-induced conformational changes in Lp(a) (which do not occur in LDL) that permit a better recognition by the LDL receptor of HMDM at 4 than at 31 "C.
We would like to stress that these results were obtained with only one species of Lp(a) that was chosen because of its stability and known physical parameters. As more Lp(a) species with different apo(a) polymorphs are characterized, we plan to extend these studies in order to determine the effect of apo(a) polymorphism on the uptake and degradation of Lp(a) by macrophages and other cell lines, i.e. human skin fibroblasts and primary human hepatocytes.