Mechanisms of Inhibition by Apolipoprotein C of Apolipoprotein E-dependent Cellular Metabolism of Human Triglyceride-rich Lipoproteins through the Low Density Lipoprotein Receptor Pathway*

The mechanism of inhibition by apolipoprotein C of the uptake and degradation of triglyceride-rich lipoproteins from human plasma via the low density lipo- protein (LDL) receptor pathway was investigated in cultured human skin fibroblasts. Very low density li- poprotein (VLDL) density subfractions and intermediate density lipoprotein (IDL) with or without added exogenous recombinant apolipoprotein E-3 were used. Total and individual ((3-1, C-11, C-111-1, and C-111-2) apoC molecules effectively inhibited apoE-3-mediated cell metabolism of the lipoproteins through the LDL receptor, with apoC-I being most effective. When the incubation was carried out with different amounts of exogenous apoE-3 and exogenous apoC, it was shown that the ratio of apoE-3 to apoC determined the uptake and degradation of VLDL. Excess apoE-3 overcame, at least in part, the inhibition by apoC. ApoC, in contrast, did not affect LDL metabolism. Neither apoA-I nor apoA-11, two apoproteins that do not readily associate with VLDL, had any effect on VLDL cell metabolism. The inhibition of VLDL and IDL metabolism cannot be fully explained by interference of association of exog- enous apoE-3 with or displacement of endogenous apoE from the lipoproteins. IDL is a lipoprotein that apoC-I free ApoC-I apoC apoC-11, apoC-111-1, apoC-111-2 ApoC-I1

Mechanisms of Inhibition by Apolipoprotein C of Apolipoprotein Edependent Cellular Metabolism of Human Triglyceride-rich Lipoproteins through the Low Density Lipoprotein Receptor Pathway* (Received for publication, September 10, 1990) Ephraim Sehayek and Shlomo Eisenberg From the Lipid Research Laboratory, Department of Medicine B, Hadassah University Hospital,Israel The mechanism of inhibition by apolipoprotein C of the uptake and degradation of triglyceride-rich lipoproteins from human plasma via the low density lipoprotein (LDL) receptor pathway was investigated in cultured human skin fibroblasts. Very low density lipoprotein (VLDL) density subfractions and intermediate density lipoprotein (IDL) with or without added exogenous recombinant apolipoprotein E-3 were used. Total and individual ((3-1, C-11, C-111-1, and C-111-2) apoC molecules effectively inhibited apoE-3-mediated cell metabolism of the lipoproteins through the LDL receptor, with apoC-I being most effective. When the incubation was carried out with different amounts of exogenous apoE-3 and exogenous apoC, it was shown that the ratio of apoE-3 to apoC determined the uptake and degradation of VLDL. Excess apoE-3 overcame, at least in part, the inhibition by apoC. ApoC, in contrast, did not affect LDL metabolism. Neither apoA-I nor apoA-11, two apoproteins that do not readily associate with VLDL, had any effect on VLDL cell metabolism. The inhibition of VLDL and IDL metabolism cannot be fully explained by interference of association of exogenous apoE-3 with or displacement of endogenous apoE from the lipoproteins. IDL is a lipoprotein that contains both apoB-100 and apoE. By using monoclonal antibodies 4G3 and 1D7, which specifically block cell interaction by apoB-100 and apoE, respectively, it was possible to assess the effects of apoC on either apoprotein. ApoC dramatically depressed the interaction of IDL with the fibroblast receptor through apoE, but had only a moderate effect on apoB-100. The study thus demonstrates that apoC inhibits predominantly the apoE-3-dependent interaction of triglyceride-rich lipoproteins with the LDL receptor in cultured fibroblasts and that the mechanism of inhibition reflects association of apoC with the lipoproteins and specific concentration-dependent effects on apoE-3 at the lipoprotein surface.
Regulation of the uptake and degradation of very low density lipoproteins (VLDL)' and their remnants by cellular * This work was supported in part by a grant from the National Council for Research and Development (Israel) and by G. S. F.
(Munchen, Federal Republic of Germany) Dismed Grant GR858. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
' The abbreviations used are: VLDL, very low density lipoprotein(s); LDL, low density lipoprotein(s); IDL, intermediate density lipoprotein(s); HDL, high density lipoprotein(s1; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; LPDS, lipoprotein-deficient serum; PBS, phosphate-buffered saline; LRP, LDL receptor-related protein. receptors determines the amount of LDL that is generated by the apoB-100 VLDL + LDL cascade (1-3). Although these lipoproteins contain apoB-100, it appears that their interaction with cell receptors depends on the presence of functional apoE molecules (E-3 and E-4) at the surface of the lipoprotein particle (4-8). Indeed, enrichment of VLDL and IDL particles from either normolipidemic human subjects (9,lO) or patients with hypertriglyceridemia (11) with apoE-3 has recently been shown to cause a manyfold enhancement of their cellular metabolism. This mechanism is presumably responsible for the findings that injection of apoE in Watanabe and cholesterol-fed rabbits causes a dramatic reduction in the VLDL and IDL levels of the animals (12, 13).
Some lipoproteins that are naturally rich in apoE, e.g. p-VLDL and HDL, from cholesterol-fed animals, exhibit exceptionally high affinity toward cell receptors, including the B/ E (LDL) receptor (2,8,(14)(15)(16)). Yet, neither human plasma VLDL ( 9 , l l ) nor IDL (10,17), even when maximally enriched with apoE-3, are taken up and degraded by cultured human skin fibroblasts or HepG2 cells at rates higher than those of LDL, a lipoprotein that contains only apoB-100. One possible explanation for this unexpected observation is that apoE-3 molecules, when associated with triglyceride-rich lipoprotein particles, are not fully expressed in receptor-binding processes. Alternatively, it can be speculated that the apoE-mediated uptake of triglyceride-rich particles by cell receptors is partially inhibited by other apoproteins that are associated with these lipoproteins. Specifically, apoC molecules that were shown to interfere with the uptake of chylomicrons and nascent hepatic VLDL during perfusion of rat liver (18,19) may also inhibit the interaction of human VLDL with the B/ E (LDL) receptor. This possibility has been explored in this study.
Monoclonal Antibodies-Monoclonal antibodies that specifically block the ligand-binding domain of apoE (antibody 1D7) or apoB-100 (antibody 4G3) were a generous gift of Drs. Milne and Marcel (Clinical Research Institute of Montreal, Montreal, Canada). The characteristics of the antibodies are described elsewhere (25).
Reassociation of Apolipoproteins with Lipoproteins-To determine the association of apoC, apoE-3, and a mixture of apoC and apoE-3 with VLDL or IDL, the procedure previously described by us (9) was used. Lipoproteins (300 pg of protein) were incubated with different amounts of apolipoproteins for 60 min at 37 "c in 0.9% NaCl, 20 mM Tris (pH 7.4), and 0.01% EDTA buffer. After the incubation, the mixture (1 ml) was applied to a 0.9 X 25-cm column packed with 8% agarose (Bio-Gel . Lipoproteins eluted at the void volume and 1-5 ml of the column, while free apoproteins in tubes 10-14. The lipoproteins were recovered; concentrated; dialyzed against water, 0.01% EDTA (pH 7.4) buffer; and analyzed for apoprotein content by SDS-PAGE. In some experiments, "'I-VLDL and unlabeled apoE-3 and/or apoC were used, whereas in other experiments, unlabeled VLDL was incubated with '251-apoE-3 without or with unlabeled apoC. In these last experiments, the tubes collected after chromatography were assessed for radioactivity content prior to electrophoresis.
Binding, Cell Association, and Degradation of "'I-Lipoprotein.-All studies were performed in cultured human skin fibroblasts from the fifth to the fifteenth subculture as previously described (26). The cells (3.5 X IO4 cells) were grown in 35-mm dishes (Falcon Labware) in 2 ml of Dulbecco/Vogt medium supplemented with 10% fetal calf serum. After 5 days of culture, the cells were washed with 1-2 ml of phosphate-buffered saline (PBS), and the medium was replaced by a medium containing 5 mg/ml lipoprotein-deficient serum (LPDS). Experiments were performed on day 7 of culture. The cells were washed with PBS and incubated with '*'I-labeled lipoprotein (10 pg of protein/ml) in 1 ml of LPDS medium at 37 "C for 6 h. Unlabeled apoE-3 and/or apoC or individual apoC proteins were added to the medium containing the "'1-lipoproteins, and the medium was transferred to the culture dishes. In some experiments, apoE-3 or apoC was added first to the "'I-lipoprotein-containing medium and allowed to incubate for 30 min at 37 "C in a thermostated water bath. The medium was supplemented with the second apoprotein immediately prior to the incubation. At the end of the 6-h incubation period, the medium was removed and examined for non-iodide "'I-protein degradation products (27). The cells were chilled on ice and washed six times with ice-cold PBS, 0.2% bovine albumin solution. Bound lZ5Ilipoproteins were released after incubation of the cells with medium containing 10 mg/ml sodium heparin for 1 h at 4 "C following the procedure of Brown and Goldstein (28) and Goldstein et al. (29). Radioactive lipids were extracted with chloroform/methanol (l:l, v/ v), and the cell-associated '2sII-lipoprotein protein was defined as the radioactivity remaining in the delipidated cells. Dishes without cells and cultures incubated with a 40-fold excess of unlabeled lipoproteins were processed in parallel to determine blank (no cells) and nonspecific (40-fold excess unlabeled lipoproteins) metabolic activities. Results are expressed as nanograms of '2sII-lipoprotein bound, associated, or degraded per milligram of cell protein after substraction of blank and nonspecific values.
In some experiments, monoclonal antibody 1D7 or 4G3 was added to the medium containing the '251-lipoproteins prior to the addition of apoC and/or apoE-3, and the medium was incubated at 37 "C for 60 min. Thereafter, apoC and/or apoE-3 was added, and the samples were treated as described above.
Incorporation of [14C]Acetate into Sterol.-Sterol synthesis from ["Clacetate was determined in up-regulated fibroblasts after a 6-h incubation with unlabeled lipoproteins in 1 ml of LPDS medium as previously described (9). The lipoprotein-containing medium was removed and the cells were washed with PBS, 0.2% albumin and reincubated for 2 h with medium containing 10 pCi of [2-14C]acetate (55 mCi/mmol). The cells were extracted with hexane/isopropyl alcohol (3:2, v/v), and the amount of [14C]acetate incorporated into sterols was determined (26). Cultures incubated without lipoproteins served to determine the capacity of the cells to synthesize 14C-labeled sterol, and the values obtained from these cultures were taken as 100%.
Analytical Methods-Lipoprotein and cell protein were determined by the procedure of Lowry et al. (30); phospholipids, triglycerides, and cholesterol were determined by standard methods (31)(32)(33). SDSand urea-PAGE of apoproteins were performed as described (9,23,34). Apoprotein content in lipoproteins was determined using a Quick Scan apparatus (Helena Laboratories, Beaumont, TX). Relative absorbance was linear for apoB and apoE at ranges of 2-8 and 0.5-4.0 pg of protein, respectively. NalZ5I was purchased from Du Pont-New England Nuclear, and [2-14C]acetate was from The Radiochemical Centre (Amersham, Great Britain). Culture flasks and dishes were purchased from Falcon Labware, and medium was from GIBCO. All chemicals and reagents were of analytical grade.

RESULTS
Studies with Exogenous upoE-3"The chemical composition and apoprotein profile of VLDL-I, -11, and -111 and of IDL and LDL were similar to those recently published (9, 10). VLDL and IDL contained apoB-100, apoE, and apoC (data not shown). LDL contained only apoB-100. In agreement with our previous reports, we observed a manyfold increase in the binding, association, and degradation of VLDL and IDL labeled proteins upon the addition of optimal concentrations of recombinant apoE-3 to the culture medium (Table I).
The nature of the fibroblast receptor responsible for the uptake and degradation of VLDL and apoE-3-enriched VLDL was investigated under conditions that discern between the LDL receptor and the LDL receptor-related protein (LRP). The results for the proteolytic degradation data are shown in Table 11. In up-regulated fibroblasts, apoE-3 (but not apoE-2) enhanced the degradation of VLDL by 8-9-fold. Downregulation of the LDL receptor was achieved by growing the cells in the presence of LDL (80 pg of protein/ml). The degradation of VLDL and apoE-3-enriched VLDL in downregulated fibroblasts was depressed by 80%. Yet, the enhancing effect of apoE-3 (but not apoE-2) on VLDL metabolism was evident in the down-regulated cells. In both up-and down-regulated cells, an excess of unlabeled LDL effectively depressed the degradation of apoE-3-enriched VLDL. Finally, in LDL receptor-negative fibroblasts, neither VLDL nor apoE-3-enriched VLDL was degraded to any appreciable extent. Similarly, neither rabbit @-VLDL nor apoE-3-enriched VLDL was degraded in these cells. All these observations are consistent with the uptake and degradation of VLDL and apoE-3-enriched VLDL through the LDL receptor pathway, and not the LRP.
The effects of individual C apoproteins (C-I, C-11, C-111-1, and C-111-2) and total apoC on the binding and proteolytic degradation of apoE-enriched VLDL-I and VLDL-I11 and of LDL are shown in Fig. 1. In this experiment, 10 pg of VLDL or LDL protein were included in the culture medium without or with 4 pg of apoE-3 and 4 pg of total or individual C proteins. The C proteins caused a 50-90% reduction in the apoE-3-stimulated VLDL metabolism, but had no effect on LDL metabolism. Whereas all C proteins were effective in reducing the binding and degradation of apoE-3-enriched VLDL by the cells, apoC-I was considerably more effective.

TABLE I Effects of exogenous recombinant apolipoprotein E-3 on the cellular metabolism of VLDL density subfractions and
IDL Data are means f S.E. of four to nine different experiments. Binding, cell association, and proteolytic degradation of '2sII-lipoproteins were determined as described under "Experimental Procedures." The lipoproteins (10 pg of protein/ml) were incubated with up-regulated human skin fibroblasts for 6 h at 37 "C in the absence or presence of exogenous recombinant apolipoprotein E-3 (4 pg/ml).  Similar results were found for cell-associated labeled apoproteins (data not shown). In an additional experiment, the effects of increasing concentrations of total and individual C apoproteins on the apoE-3 (4 pg/ml)-stimulated VLDL metabolism were determined. Fig. 2 presents the proteolytic degradation data. ApoC and individual C proteins were already effective at a concentration of l pg/ml. At concentrations of 4 pg/ml, VLDL proteolytic degradation was reduced to <50% of the original values. With a further increase in apoC concentration in the medium to 8 pg/ml, the degradation of VLDL apoproteins was barely above that found when no apoE-3 was added. Again, apoC-I was more effective than the other C apoproteins.
The effects of mixtures of apoE-3 and total apoC at different protein concentration ratios on VLDL-I and VLDL-I11 are shown in Tables I11 and IV. Two observations can be made. First, at any concentration of apoE-3, the addition of increasing amounts of apoC causes higher degrees of inhibition of binding, association (data not shown), and degradation of VLDL. Second, at all different apoC concentrations, including the highest of 8 pg/ml, increasing amounts of apoE-3 cause stimulation of VLDL cell metabolism. This was especially prominent for the proteolytic degradation of VLDL-111; when at a concentration of 8 pg of apoC and 8 pg of apoE-3, the value was 7-8 times higher than that observed with VLDL-I11 alone. Yet, this value was less than one-tenth that In each experiment, labeled human plasma LDL was studied in parallel to VLDL. apoC was without any effect on the binding, cell association, and proteolytic degradation of LDL. Frequently, a slight increase in LDL metabolism was observed.
To determine whether other apoproteins exert a similar effect on apoE-3-stimulated VLDL cell metabolism, the experiments as detailed in Tables I1 and I11 were repeated with apoA-I and apoA-11. Table V presents the results with VLDL-I and VLDL-III(l0 pg of protein/ml) incubated with 4 pg/ml exogenous recombinant apoE-3 and increasing concentrations (2-8 pg/ml) of apoA-I. apoA-I had no effect on the binding, cell association, and proteolytic degradation of VLDL. In addition, no effect of apoA-I on the cell metabolism of VLDL-I or VLDL-I11 was found in incubations without apoE-3 or at apoE-3 concentrations of either 2 or 8 pg/ml (data not shown). Similarly, apoA-I1 was without any effect on VLDL-I or VLDL-I11 cell metabolism either in the absence or presence of different concentrations of exogenous apoE-3 (data not shown). Neither apoA-I (Table V) nor apoA-I1 (data not shown) had any effect on LDL metabolism, although a tendency toward higher metabolic activity was noted.
In the previous experiments, apoC and apoE-3 were added together to incubation media that contained the lZ5I-labeled Each 1 ml of the incubation medium contained 10 pg of VLDL-I1 protein, 4 pg of recombinant apoE-3, and 0, 1, 2, 4, or 8 pg of total or individual apoC proteins. Binding, cell association, and proteolytic degradation of 9 -V L D L were determined after 6 h of incubation with up-regulated human skin fibroblasts at 37 "C as described under "Experimental Procedures." lipoproteins immediately prior to the beginning of the incubation. Since the order of introduction of the apoproteins to VLDL may have affected the results, the following experiment was carried out. ApoC or apoE-3 at a concentration of 2,4, or 8 pg of protein/ml was added to incubation mixtures containing 10 pg/ml '251-VLDL-II; and the mixtures were preincu-bated for 30 min at 37 "C in a thermostated water bath. After preincubation, apoE-3 or apoC was added at a concentration of 2,4, or 8 pg/ml to mixtures preincubated with the opposing apoprotein (apoC or apoE-3); and the media were applied to duplicate cultured fibroblast dishes. Binding, cell association, and degradation were determined after 6 h of incubation. Fig.  3 presents the proteolytic degradation data. As is evident from Fig. 3, apoC inhibited the stimulated degradation of "' 1-VLDL-I1 preincubated with apoE-3 at all three concentrations, whereas increasing concentrations of apoE-3 stimulated the activity of '''I-VLDL-I1 samples preincubated with different amounts of apoC. In fact, the shape of the curves demonstrates trends similar to those observed with mixtures of apoC and apoE-3. Although the results suggest that the metabolic activity of VLDL was lower when apoC was introduced first to the incubation media, this observation was not consistently seen at all apoC/apoE-3 ratios. The data for binding and cell-associated apoproteins (data not shown) were similar to those described in the legend to Fig. 3 for degradation.
To ascertain that apoC did not merely decrease VLDL-'*'Iapoprotein binding, internalization, and degradation by the cells, we studied the effects of apoC on the ability of apoE-3enriched VLDL-I, -11, and -111 to down-regulate cellular sterol synthesis (Table VI). In agreement with our previous observations (9), VLDL were incapable of down-regulating cellular sterol synthesis in the absence of exogenous apoE-3, but considerable inhibition of the incorporation of [14C]acetate into sterols was observed when apoE-3 was added to the medium. The addition of apoC to the apoE-3-containing incubation medium caused a pronounced decrease in the ability of VLDL to induce down-regulation of cellular sterol synthesis.
The association of apoE-3 alone, apoC alone, and mixtures of apoE-3 plus apoC with VLDL was investigated by chromatography on agarose columns. In the first experiment, lZ5I-VLDL-I11 was incubated with unlabeled apoC (37 "C, 60 min), and VLDL was separated from unassociated apoC. At the three VLDL/apoC protein concentrations investigated (300 pg/60 pg, 300 pg/120 pg, and 300 pg/240 pg), the amount of radioactivity in VLDL was decreased only slightly by 3.9,9.0, and 8.8%, respectively, indicating a maximum exchange of labeled endogenous apoC by unlabeled exogenous apoC of 10-20%. Of interest, the small amount of radioactivity associated with apoE in VLDL (3.0% of total radioactivity) remained unchanged after incubation with apoC and was 2.9, 3.2, and 2.8% of total radioactivity, respectively. In the second experiment, VLDL-I and VLDL-I1 (300 pg of protein) were incubated for 1 h at 37 "C with 120 pg of apoE-3 alone, 120 pg of apoC alone, or a mixture of 120 pg of apoE-3 plus 120 pg of apoC. After incubation, VLDL was separated from unassociated protein, and aliquots were taken for SDS-PAGE. As shown in Fig. 4, apoE-3 and apoC became associated with VLDL when incubated separately (lunes 2 and 3 ) or together (lune 4 ) . Gel scanning yielded the following percent contribution of apoB, apoE, and apoC to total VLDL proteins in the four samples: 53.3, 2.1, and 44.9% for VLDL incubated with 0.9% NaCl (lane 1 ); 42.5, 18.7, and 32.3% for VLDL incubated with apoE-3 (lune 2 ) ; 41.9,2.5, and 53.7% for VLDL incubated with apoC (lune 3 ) ; and 36.9, 10.7, and 45.5% for VLDL incubated with apoE and apoC (lune 4 ) , respectively. Similar results were obtained with other VLDL density subfractions. In another experiment, association of apoA-I with VLDL-I and VLDL-I1 was determined (data not shown). Association of apoA-I with VLDL was evident, but apoA-I accounted for only 7.1% of total VLDL proteins, about onethird that of apoE-3 or apoC.

TABLE 111
Effect of exogenous apoC and apoE-3 on the binding and degradation of VLDL-I and LDL by cultured fibroblasts Data are means f S.E. of three experiments. Cells were incubated with "'I-VLDL-I or '"1-LDL (10 pg of protein/ml), and the culture medium was supplemented with either exogenous recombinant apoE-3 or total human apoC fraction, or both, at the protein concentrations indicated. Binding and degradation were determined after 6 h of incubation at 37 "C as described under "Experimental Procedures." Results for cell-associated proteins (data not shown) showed trends similar to those for binding and degradation.  Data are means f S.E. of three experiments or of two experiments (Footnote a). Cells were incubated with VLDL-111 (10 pglml), and the culture medium was supplemented with either exogenous recombinant apoE-3 or total human apoC fraction, or both, at the protein concentrations indicated. Binding and degradation were determined after 6 h of incubation at 37 "C as described under "Experimental Procedures." Results for cellassociated protein (data not shown) showed trends similar to those for binding and degradation. The different VLDL preparations obtained after gel filtration were tested for biological activity by their ability to suppress cholesterol synthesis in up-regulated fibroblasts (Fig.  5). Incorporation of [14C]acetate into sterols was effectively inhibited by apoE-3-enriched VLDL (column D), and this capacity to inhibit cholesterol synthesis was prevented by apoC added either after (column E ) or prior (column F ) to the gel filtration procedure.
From the gel scanning data, it can be calculated that the addition of apoC to an incubation system containing VLDL and apoE-3 caused a moderate decrease in the amount of

FIG. 3. Effects of preincubation of VLDL with either apoE-3 or apoC on degradation of lipoprotein in presence of opposing apoprotein, apoC or apoE-3.
Aliquots of "'I-VLDL-II(l0 pg/ ml of LPDS medium) were preincubated in test tubes with 2, 4, or 8 pg/ml apoE-3 or apoC for 30 min at 37 "C. At the end of the preincubation, the opposing apoprotein (apoC or apoE-3) was added to samples of the preincubation media at protein concentrations of 2, 4, or 8 pg/ml. The media containing '"I-VLDL-I1 and the apoproteins were transferred to culture dishes with up-regulated human skin fibroblasts. Binding, cell association, and proteolytic degradation of '251-VLDL-II were determined after 6 h of incubation at 37 "C as described under "Experimental Procedures." Shown are the degradation data. A, 1251-VLDL-II preincubated with recombinant apoE-3 (2 pg/ml ( 0 ) , 4 pg/ml (A), and 8 pg/ml (0)) followed by the addition of apoC; B, '2'I-VLDL-II preincubated with apoC (2 pg/ml (O), 4 pg/ ml (A), and 8 pg/ml (W)) followed by the addition of recombinant apoE-3. apoE-3 that became associated with VLDL (a decrease in the apoE/apoB ratio from 0.44 to 0.29 (Fig. 4, lanes 2 and 4). To elucidate whether apoC can displace apoE that is already associated with VLDL, the following experiment was performed. First, lZ5I-apoE-3 was allowed to associate with  (30 pg of protein/ml) or LDL (15 pg/ml) with or without the addition of exogenous apoE-3 (12 pg of protein/ml) and exogenous apoE-3 plus apoC (12 pg of protein/ml) for 6 h at 37 "C. At the end of the incubation, the lipoprotein-containing medium was replaced by LPDS medium containing [2-'T]acetate, and sterol synthesis was determined as described under "Experimental Procedures." Cultures incubated in LPDS medium without lipoproteins served to determine the capacity of up-regulated cells to synthesize "C-labeled sterols. The values obtained from these cultures were taken as 100%. . SDS-PAGE of VLDL-I1 incubated at 37 "C for 60 min with or without exogenous apoproteins followed by gel filtration on agarose columns. Lane I , VLDL incubated without exogenous apoproteins; lane 2, VLDL incubated with exogenous apoE-3 (300 pg of protein/l20 pg of protein); lane 3, VLDL incubated with exogenous apoC (300 pg of protein/l20 pg of protein); lane 4, VLDL incubated with exogenous apoE-3 and exogenous apoC (300 pg of protein/l20 pg of protein/l20 pg of protein). The gels were loaded with an equal amount of cholesterol (15 pg).
Identical results were found with apoC-111-1. With apoC-I, less radioactivity separated with VLDL (63.0 uersu.9 76.4-80.7%), but ""I-apoE-3 distributed along the remaining column volume and not toward the salt volume. Thus, although apoC appears to interfere, to some extent, with the association of apoE-3 with VLDL, it did not displace appreciable amounts of associated apoE-3 molecules from the particles.
Studies with Endogenous apoE-The possibility that apoC may have different effects on endogenous apoE-3 than those described above for exogenous apoE-3 was evaluated with two apoE-containing lipoproteins that exhibit apoE-dependent cell metabolism: total VLDL from a subject with borderline high plasma triglyceride levels (-200 mg/dl) and IDL from a normotriglyceridemic subject with profile E-4/3. For each lipoprotein, we determined the effects of adding increasing concentrations (2-12 pg/ml) of apoC on the binding, cell association, and degradation of the lipoprotein (10 pg of protein) in the absence or presence of exogenous apoE-3 (4 pg/ml). The results of these experiments are shown in Fig. 7.
The data demonstrate an apoC-dependent inhibition of cell FIG. 5. Effect of VLDL incubated with recombinant apoE-3 and with apoE-3 and apoC on incorporation of ["'Clacetate into sterols in up-regulated human skin fibroblasts. The cells were exposed to the different lipoprotein preparations (25 pg of cholesterol/ml) for 6 h at 37 "C. The lipoprotein-containing medium was removed, and incorporation of ["'Clacetate into sterols was determined as described under "Experimental Procedures." Data are the percent of sterol synthesis determined in cells exposed to medium without lipoproteins (bar A ) . The following conditions are shown: bar B, LDL; bar C, VLDL incubated in 0.9% NaCl and isolated on an agarose column (see Fig. 4, lane I ) ; bar D, VLDL incubated with apoE-3 (10:4 protein ratio) prior to gel filtration (see Fig. 4, lane 2 ) ; barE, same VLDL as in bar D supplemented with apoC (10:4 protein ratio); bar F, VLDL incubated with both apoE-3 and apoC (1044 protein ratio) prior to gel filtration (see Fig. 4, lane 4 ) . Data are means f S.D. of triplicate dishes.
Yet, for most experimental data points, especially IDL, the effects of apoC were marginally less pronounced for particles not enriched with apoE-3. T o ascertain that the effects of apoC on endogenous apoE are not due to displacement of apoE from the particles, IDL was incubated at 37 "C for 60 min without or with apoC (300 pg of protein/240 pg of protein) and then isolated by gel filtration on an agarose column. No apparent displacement of endogenous apoE was found (  apoE-3, 61, 267, and 1940;IDL, 30, 99, and 933;and IDL + apoE-3,41,247, and 1925, respectively. The chemical composition (percent) of VLDL was: protein, 10.2%; triglyceride, 61.0%; esterified cholesterol, 7.4%; free cholesterol, 3.9%; and phospholipids, 17.6% of total VLDL mass. The corresponding values for IDL were: 16.3, 26.8, 22.8, 8.9, and 25.2%, respectively.  42.5, 11.2, and 41.1% in lane 2, respectively. Gels were loaded with 15 pg of protein for each sample. 8). The apoE/apoB absorbance ratio determined by gel scanning was 0.28 in IDL preincubated without apoC and 0.27 in IDL preincubated with apoC. The corresponding values for apoC/apoB absorbance ratios, in contrast, increased from 0.54 to 0.97, indicating considerable association of added apoC to IDL without a measurable decrease in apoE content.
IDL contains two proteins that may serve as ligands for the LDL receptor: apoB-100 and apoE. To determine whether apoC inhibits interaction of IDL with the receptor via apoB-100 or apoE, or both, the effects of apoC on IDL cell metabolism were determined after preincubation (37 "C, 60 min) of the lipoprotein with a monoclonal antibody that specifically  (0). ""I-IDL (10 pg of protein/ml) was preincubated with either monoclonal antibody 1D7 or 4G3 (10 pg of protein/ml) for 60 min at 37 "C. ApoC at the indicated protein concentrations was added to aliquots of the two incubation mixtures, and the media were transferred to culture dishes containing up-regulated human skin fibroblasts. Binding, cell association, and proteolytic degradation of the "'I-IDL proteins were determined after 6 h of incubation a t 37 "C as described under "Experimental Procedures." Shown are the degradation data as percent of the activity in samples containing the monoclonal antibodies, but not apoC. For comparison, also included are the effects of apoC on IDL degradation in the absence of antibodies (None). Antibody 4G3 reduced the binding of ""I-LDL from 69 to 0 ng of protein/mg of cell protein, the cell association from 174 to 0, and the degradation from 749 to 11 ng of protein/mg of cell protein/ 6 h. Antibody 1D7 was without effect on ""I-LDL metabolism. blocks receptor interactions by apoE (antibody 1D7) or an antibody that is specific for apoB-100 (antibody 4G3). The results of this experiment are shown in Fig. 9. For comparison, Fig. 9 also includes the results obtained with IDL alone, in the absence of antibodies. In the presence of antibody 4G3, apoC was as effective or more effective as an inhibitor of IDL catabolism by the cells as compared to IDL alone. In contrast, in the presence of antibody 1D7, the inhibition of IDL degradation was only one-fourth to one-third that observed with IDL alone or with IDL and antibody 4G3. The difference was especially pronounced at low, more physiological concentrations of apoC. It thus appears that C apoproteins exert their inhibitory effects on IDL metabolism predominantly through apoE-dependent interactions, but not apoB-100-dependent processes.

DISCUSSION
Lipoprotein particles along the apoB-100 cascade may follow one of two metabolic routes: interaction with endothelium-bound lipases and delivery of triglyceride fatty acids to tissues or interaction with cellular receptors followed by internalization and lysosomal degradation of the lipoproteins (3, 35). The large apoB-100 lipoproteins (VLDL) contain, at their surfaces, protein molecules that are responsible and essential for either metabolic route. These are apoC-11, which are necessary for activation of lipoprotein lipase-mediated triglyceride hydrolysis (36), and apoB-100 and apoE-3 (or apoE-4), which serve as ligands for interactions of lipoproteins with the LDL receptor and other receptors (2,8,14-16). Thus, theoretically, the interaction of VLDL with lipoprotein lipase or cell receptors should occur at random. Yet, many studies demonstrated that intact VLDL, especially large-size and light VLDL-I and VLDL-11, possess a very limited capacity to interact with cell receptors (4-9). Apparently, the ligandbinding domains of apoB-100 and apoE-3 (or apoE-4) in VLDL are not available for interaction with cellular receptors either because of inadequate exposure of the proteins to the receptor or because of inhibition of receptor interaction by other apoproteins, e.g. apoC. This study was carried out to critically examine the latter mechanism.
Several studies have demonstrated that apoE-dependent uptake of rat chylomicrons and VLDL by perfused rat liver is inhibited by individual and total human and rat apoC molecules (18,19,37). In two studies, the apoE/apoC ratio was reported to be without effect on the liver uptake of chylomicrons and their remnants (19,37). This study was carried out with human lipoproteins in cultured human skin fibroblasts, where the predominant receptor for the uptake of apoB-100and apoE-containing lipoproteins is the LDL receptor (9,lO). When using VLDL density subfractions from normolipidemic human subjects that lack the capacity to interact with the receptors unless enriched with exogenous apoE-3, we were able to precisely determine the stoichiometry of the effects of combinations of apoC and apoE on the binding, internalization, and degradation of the lipoproteins. Our data unequivocally demonstrate that the ratio of apoE to apoC determines the degree of cell metabolism of VLDL. At any apoE concentration, increasing the apoC concentration in the incubation mixture caused progressive inhibition of the cell metabolism of VLDL. Conversely, increasing apoE-3 concentrations enhanced VLDL metabolism at all apoC concentrations. It is interesting to note that, in the cell culture system, the order of introducing apoC and apoE to VLDL had no or only a minor effect on the resulting metabolic activity. Thus, VLDL that had already acquired apoE was inhibited by apoC, whereas VLDL that acquired apoC was stimulated by apoE.
In experiments on perfused rat liver (18,19,37), it was impossible to determine whether the inhibitory effect of apoC is on the LDL receptor or on additional receptor(s) specific for apoE, e.g. LRP (38)(39)(40). Previous studies from this laboratory (9, 10) and this study (Table 11) show that the metabolism of apoE-3-enriched VLDL and IDL in cultured human fibroblasts occurs through the LDL receptor (9, 10). Therefore, this study proves that apoC inhibits the metabolism of apoE-enriched human VLDL by the LDL receptor pathway. apoC was recently shown to inhibit the interactions of apoEenriched rabbit P-VLDL with the LRP (41). This effect was especially pronounced with apoC-I (42). Our results, however, differ from those of Kowal et al. (41) and Weisgraber et al. (42) in several respects. First, we show a very pronounced effect of apoC on VLDL metabolism through the LDL receptor pathway, whereas in the assays used by Kowal et al. and Weisgraber et al., there are minimal effects, if any, of apoC on the binding of P-VLDL to the LDL receptor. Second, in our studies, although apoC-I was most effective, all other C apoproteins including apoC-111-1 and apoC-111-2 were also effective. Third, the concentrations of total and individual C apoproteins used by us as well as the concentration of apoE-3 are considerably lower than those used in the other studies. Last, we observed a partial recovery of the cell metabolism of VLDL upon addition of increasing concentrations of apoE-3 to an apoC-inhibited system. These differences perhaps reflect one or more of the following differences between the studies: ( a ) methodology (uptake and degradation uersus cholesterol esterification); ( b ) lipoproteins (human VLDL as compared to rabbit 0-VLDL; ( c ) cells (normal skin fibroblasts uersus LDL receptor-negative cells); and ( d ) receptors (LDL receptor uersus LRP).
Another aspect of this study was the question of whether apoC similarly affects apoE-3 that is incorporated into lipoproteins by an exogenous route as compared to endogenous apoE-3 associated with particles isolated from the plasma. The results reported here demonstrate a similar, although not identical effect of exogenous apoC on both endogenous and exogenous E apoproteins. This observation indicates that the inhibition of cell metabolism of VLDL and IDL by apoC is not an artifact of added exogenous apoE-3, but is of physiological significance. Yet, it should be pointed out that, in this as well as in previous studies (18,19,37,41,42), apoC is added by an exogenous route. This route of adding apoC to lipoproteins may result in different behavior of the apoC molecules at the lipoprotein surface as compared to endogenous apoC. It is of significance to note in this context that the amounts of apoC necessary to produce a measurable inhibition of cellular metabolism of VLDL and IDL (1-4 pg/ 10 pg of lipoprotein) are one-fourth or equal to the amounts of apoC present. In contrast, the addition of similar amounts of exogenous apoE-3 without loss of endogenous apoC causes a dramatic stimulation of the same metabolic activities (9, 10).
The mechanisms by which C apoproteins exert their inhibitory effects on apoE-dependent cell metabolism of lipoproteins were partially elucidated in this study. This effect was specific for C apoproteins that associate with VLDL and IDL and was not observed with apoA-I and apoA-11, proteins that poorly associate with VLDL (16). Although apoC-I was most effective, all apoC molecules (C-I, C-11, C-111-1, and C-111-2) were inhibitory. apoC molecules do not associate and had no effect on LDL metabolism in the same cells. Thus, direct interaction of apoC with the receptor can be ruled out as a mechanism of interference with the uptake of lipoproteins. Moreover, the observations with IDL and monoclonal antibodies that specifically block apoE (antibody 1D7) and apoB-100 (antibody 4G3) indicate that apoC has only a mild effect on the binding of lipoproteins to the LDL receptor via apoB-100 even when both are present on the same particle. In contrast, apoC had a very substantial effect on interactions via apoE. A potential mechanism for this specific apoC effect is displacement of apoE-3 from the particles as proposed in the studies on the LRP (42). This mechanism, however, is not compatible with our results. Although apoC interfered to some extent (-30-35%) with the association of apoE with VLDL, the amount of apoE remaining in the particles should have enhanced their metabolism to a very considerable extent, and that was not observed. More important, no or only minimal displacement of endogenous (IDL) or of associated exogenous (VLDL) apoE-3 by apoC was found. We therefore suggest that the apoC effects are due predominantly to interactions with apoE at the lipoprotein surface. Such interactions are apparently specific and do not readily occur between apoC and apoB-100. Whether apoC somehow affects the conformation of apoE-3 molecules integrated at the outer coat of lipoproteins or whether there is an actual protein/protein interaction between apoC and apoE is not known. It is also unknown whether deletion of apoC from VLDL without triglyceride hydrolysis would result in activation of the uptake of the lipoprotein by cell receptors. Yet, we cannot rule out the possibility that apoC displaces a minute amount of apoE-3 that is especially active in receptor-dependent metabolic processes. Noteworthy, no evidence supporting such a hypothesis exists in the literature or can be inferred from our results. Regardless of mechanism, however, our observations