Lipoprotein lipase binds to low density lipoprotein receptors and induces receptor-mediated catabolism of very low density lipoproteins in vitro.

Lipoprotein lipase (LPL), the major enzyme responsible for the hydrolysis of plasma triglycerides, promotes binding and catabolism of triglyceride-rich lipoproteins by various cultured cells. Recent studies demonstrate that LPL binds to three members of the low density lipoprotein (LDL) receptor family, including the LDL receptor-related protein (LRP), GP330/LRP-2, and very low density lipoprotein (VLDL) receptors and induces receptor-mediated lipoprotein catabolism. We show here that LDL receptors also bind LPL and mediate LPL-dependent catabolism of large VLDL with Sf 100-400. Up-regulation of LDL receptors by lovastatin treatment of normal human foreskin fibroblasts (FSF cells) resulted in an increase in LPL-induced VLDL binding and catabolism to a level that was 10-15-fold greater than in LDL receptor-negative fibroblasts, despite similar LRP activity in both cell lines. This indicates that the contribution of LRP to LPL-dependent degradation of VLDL is small when LDL receptors are maximally up-regulated. Furthermore studies in LRP-deficient murine embryonic fibroblasts showed that the level of LPL-dependent degradation of VLDL was similar to that in normal murine embryonic fibroblasts. LPL also promoted the internalization of protein-free triglyceride emulsions; lovastatin-treatment resulted in 2-fold higher uptake in FSF cells, indicating that LPL itself could bind to LDL receptors. However, the lower induction of emulsion catabolism as compared with native VLDL suggests that LPL-induced catabolism via LDL receptors is only partially dependent on receptor binding by LPL and instead is primarily due to activation of apolipoproteins such as apoE. A fusion protein between glutathione S-transferase and the catalytically inactive carboxyl-terminal domain of LPL (GST-LPLC) also induced binding and catabolism of VLDL. However GST-LPLC was not as active as native LPL, indicating that lipolysis is required for a maximal LPL effect. Mutations of critical tryptophan residues in GST-LPLC that abolished binding to VLDL converted the protein to an inhibitor of lipoprotein binding to LDL receptors. In solid-phase assays using immobilized receptors, LDL receptors bound to LPL in a dose-dependent manner. Both LPL and GST-LPLC promoted binding of VLDL to LDL receptor-coated wells. These results indicate that LPL binds to LDL receptors and suggest that the carboxyl-terminal domain of LPL contributes to this interaction.


Lipoprotein lipase (LPL), the major enzyme responsible for the hydrolysis of plasma triglycerides, promotes binding and catabolism of triglyceride-rich lipoproteins by various cultured cells. Recent studies demonstrate that LPL binds to three members of the low density lipoprotein (LDL) receptor family, including the LDL receptor-related protein (LRP), GP330/LRP-2, and very low density lipoprotein (VLDL) receptors and induces receptor-mediated lipoprotein catabolism. We show here that LDL receptors also bind LPL and mediate LPL-dependent catabolism of large VLDL with S f 100 -400. Up-regulation of LDL receptors by lovastatin treatment of normal human foreskin fibroblasts (FSF cells) resulted in an increase in LPL-induced VLDL binding and catabolism to a level that was 10 -15-fold greater than in LDL receptor-negative fibroblasts, despite similar LRP activity in both cell lines. This indicates that the contribution of LRP to LPL-dependent degradation
of VLDL is small when LDL receptors are maximally up-regulated. Furthermore studies in LRP-deficient murine embryonic fibroblasts showed that the level of LPLdependent degradation of VLDL was similar to that in normal murine embryonic fibroblasts. LPL also promoted the internalization of protein-free triglyceride emulsions; lovastatin-treatment resulted in 2-fold higher uptake in FSF cells, indicating that LPL itself could bind to LDL receptors. However, the lower induction of emulsion catabolism as compared with native VLDL suggests that LPL-induced catabolism via LDL receptors is only partially dependent on receptor binding by LPL and instead is primarily due to activation of apolipoproteins such as apoE. A fusion protein between glutathione S-transferase and the catalytically inactive carboxyl-terminal domain of LPL (GST-LPLC) also induced binding and catabolism of VLDL. However GST-LPLC was not as active as native LPL, indicating that lipolysis is required for a maximal LPL effect. Mutations of critical tryptophan residues in GST-LPLC that abolished binding to VLDL converted the protein to an inhibitor of lipoprotein binding to LDL receptors. In solidphase assays using immobilized receptors, LDL receptors bound to LPL in a dose-dependent manner. Both LPL and GST-LPLC promoted binding of VLDL to LDL receptor-coated wells. These results indicate that LPL binds to LDL receptors and suggest that the carboxyl-terminal domain of LPL contributes to this interaction.
There are conflicting data regarding the ability of LPL to promote lipoprotein catabolism via LDL receptors (19 -23). LDL receptors are the major mediators of LDL clearance from plasma, but they have also been implicated as mediators of VLDL clearance via apoE (24). Aviram et al. (19) concluded that LPL enhances LDL uptake and degradation in both fibro-blasts and human monocyte-derived macrophages via LDL receptors. They suggest that LPL's lipolytic activity is necessary for its ability to enhance LDL degradation, but they did not study the effect of LPL on the catabolism of VLDL, which is a better substrate for LPL than LDL (25). Several investigators determined that LPL-mediated LDL degradation is independent of the LDL receptor and is not regulated by factors affecting LDL receptor expression (20 -22). These studies also focused on the effect of LPL on LDL rather than VLDL catabolism. Mulder et al. (23) concluded that internalization of LPL⅐VLDL complexes in HepG2 cells was mediated by LDL receptors, since it was negligible in LDL receptor-lacking fibroblasts. However, they did not study lipoprotein degradation or the ability of LPL itself to bind to LDL receptors. The relative contributions of LRP versus LDL receptors to the effects of LPL have not been clearly shown in cells expressing both receptors.
The present studies were aimed at resolving these discrepancies. Normal human foreskin fibroblasts (FSF cells) were studied in which both LRP and LDL receptor-dependent pathways are active. Data were also obtained in LDL receptorlacking (FH) human fibroblasts, HUVECs, and normal (MEF) and LRP-deficient (PEA13) murine embryonic fibroblasts. None of these cell lines is known to express GP330/LRP-2, but HUVECs express VLDL receptors. Our results establish a significant role for LDL receptors in LPL-mediated binding and degradation of VLDL. In addition, both solid-phase assays using immobilized LDL receptors and internalization studies using protein-free lipid emulsions demonstrate that LPL binds directly to LDL receptors and thereby may contribute to the catabolism of native VLDL via the LDL receptor pathway.

EXPERIMENTAL PROCEDURES
Materials-VLDL with S f 100 -400 were isolated as described previously by ultracentrifugation of plasma from fasted normolipidemic human subjects with the most common apoE phenotype (E3/3) (26). These are the largest, most triglyceride-rich particles present in the fasted state (27). Bovine milk LPL was isolated using a published method (28). The carboxyl-terminal fragment of human LPL (amino acid residues 313-448) was produced in Escherichia coli as a fusion protein with glutathione S-transferase (GST) as described previously (18) and designated as GST-LPLC. Using site-directed mutagenesis, tryptophan residues at positions 393 and 394 were changed to alanine to generate GST-LPLC ww (17,18) and tryptophans 390, 393, and 394 were replaced with alanine to generate GST-LPLC www . Recombinant human receptorassociated protein (RAP) was produced as described previously (29). A monoclonal antibody against the catalytic domain of LPL (mAb7) was prepared from a hybridoma kindly provided by Dr. Louis Smith (Methodist Hospital, Houston, TX) (30). Monoclonal antibodies IgG-C7 and IgG-4A4 directed against the first cysteine-rich repeat and the cytoplasmic terminal 14 amino acids, respectively, of the LDL receptor were obtained as described previously (31). Polyclonal rabbit antihuman LRP antibody Rb777 was produced as described previously (7). Human ␣ 2 M was isolated and activated by methylamine treatment as described previously (32). Monoclonal antibodies and ␣ 2 M were 125 I-labeled using IODO-GEN (Pierce) as described previously (18). Iodination of S f 100 -400 particles to specific activities of 300 -500 cpm/ng was done by the iodine-monochloride method (33). Protein-free particles with S f 100 -400 were isolated from a 10% intralipid emulsion (Travenol) by ultracentrifugal flotation (26), and their triglyceride content was estimated by an enzymatic calorimetric assay (Sigma) (34). They were labeled with [ 3 H]cholesteryl oleyl ether, a nondegradable marker of cellular uptake (21). A glass tube containing 0.5 ml of minimum Eagle's medium, 4 mg/ml BSA and 35 Ci of [ 3 H]cholesteryl oleyl ether (Amersham Corp.) was sonicated for 10 min at room temperature. Intralipid particles with S f 100 -400 containing 3-4 mg of triglycerides were added, and the mixture was incubated at 37°C for 20 min and then returned to room temperature. This treatment resulted in the incorporation of [ 3 H]cholesteryl oleyl ether in the emulsion. The tritiated lipid emulsions were stored at 4°C overnight before use. The sizes of S f 100 -400 particles from plasma and those from Intralipid were similar as indicated by scanning electron microscopy on a Hitachi S-4000 instrument (data not shown) using previously described techniques (35).
Cell Binding Assays-Normal human foreskin fibroblasts (FSF cells) were cultured as described (36,37). Mutant skin fibroblasts that do not express LDL receptors (FH cells) (38), were obtained from the National Institute of General Medical Sciences Human Genetic Mutant Cell Repository (GM00486A), Camden, NJ. MEF and mutant MEF cells that lack LRP (PEA13) (39) were provided by Dr. Joachim Herz (Dallas, TX). HUVECs were isolated from human umbilical cords as described previously (40). LDL receptors were usually up-regulated by incubation prior to the assay for 48 h with media containing 2 mg/ml lipoproteindeficient serum (LPDS) and 24 h in the presence of 1 g/ml of lovastatin (41,42). In some studies with murine fibroblasts, lovastatin was omitted and the duration of LPDS treatment was reduced to 24 h. In a few experiments, as indicated in the figure legends, LDL receptors were down-regulated by the addition of 20 g/ml LDL (41). Surface binding to metabolically inactive cells was studied after incubating cells with 125 I-ligands for 3 h at 4°C as described previously (36,37). Steady-state ligand internalization and degradation were measured after incubating cells with radiolabeled ligands at 37°C for 5 h (36,37). Degradation was defined as the trichloroacetic acid-soluble radioactivity in the incubation medium. Internalization was defined as radioactivity that remained cell-associated after incubating cells at 4°C in buffer containing 10 mg/ml tripolyphosphate. Total protein in each well as determined by the Lowry assay (43) varied by less than 15% within each experiment.
Total DNA in each well as determined spectrofluorometrically using the DNA-binding dye Hoechst 33258 (44) varied by Ͻ10% within each experiment. Wells treated with lovastatin or LPDS contained 30 -60% of the protein and DNA amounts present in untreated wells. Thus, when the two treatments were compared, results were corrected for protein or DNA per well. When LDL receptors were not up-regulated, results are presented in units of ng/well. Solid-phase Binding Assays-LDL receptors were partially purified by fractionation of total cell extracts over DE52-cellulose (Whatman) as described previously (45). Microtiter wells (96-well plates, Immulon 2, Dynatech) were coated for 30 min at 37°C with 100 l of buffer containing 3 g of anti-LDL receptor antibody, IgG-4A4. The wells were then blocked with 1% BSA at 37°C for 1 h. LDL receptors from DE52 eluants were specifically immobilized by incubating IgG-coated wells for 16 h at 4°C with 0 -100 g of DE52 eluant protein in 100 l of buffer containing 50 mM Tris-HCl, pH 8.0, 2 mM CaCl 2 , 0.5% BSA (46). To examine the direct binding of LPL to LDL receptors, the wells were washed and incubated first for 3 h at 4°C with buffer containing increasing concentrations of LPL then for 2 h at 4°C with 5 g/ml of 125 I-mAb7. After washing, bound ligands were desorbed in 0.1 N NaOH and quantitated for radioactivity. The extent of LPL binding to LDL receptors was reflected by the amount of 125 I-mAb7 bound. In some experiments, immobilized LDL receptors were incubated with 125 I-S f 100 -400 in the absence or presence of LPL or GST-LPLC.
LPL binding to LDL receptors was also detected by an enzyme-linked immunoabsorbant assay as described previously (46). Wells were coated with 0 -10 g of LPL for 5 h at 4°C, blocked with 1% BSA for 2 h at 4°C, washed, and incubated for 16 h at 4°C with 100 l of buffer with or without 1 g/l DE52 eluant (partially purified LDL receptor preparation). Then, wells were washed and incubated for 2 h at 4°C with 100 l of buffer containing 30 g/ml of anti-LDL receptor antibody (IgG-C7) followed by another 2-h incubation at 4°C with 100 l of a 1:10,000 dilution of alkaline phosphatase-conjugated rabbit anti-mouse IgG (Sigma). Wells were then washed and 100 l of substrate (1 mg/ml pnitrophenyl phosphate, Sigma) was added. After 30 min at 37°C, relative LDL receptor binding was estimated by measuring the absorbance at 405 nm. As a control, nonimmune mouse IgG was used instead of IgG-C7. Standard curves prepared simultaneously by coating wells with known amounts of IgG-C7 were linear (correlation coefficients Ͼ0.98). All data from cellular and solid-phase assays were measured in duplicate.

LPL-stimulated VLDL Binding and Degradation Is Influenced by LDL Receptor Number-Normal foreskin fibroblasts
(FSF cells) were used to investigate the role of LDL receptors in LPL-stimulated degradation of VLDL particles. Lovastatin treatment of FSF cells up-regulated their LDL receptors and increased 125 I-LDL binding and degradation by 8 -10-fold (data not shown and Ref. 36). Lovastatin had no effect on LRP expression as indicated by catabolism of 125 I-labeled activated ␣ 2 M (␣ 2 M*) (data not shown and Ref. 47). In five control experiments, we found that expression of catabolic data in terms of cellular DNA instead of protein gave results with identical interpretation. This was true in the presence or absence of lovastatin/LPDS treatment. Half-maximal LPL-induced 125 I-VLDL binding to cell surfaces at 4°C shifted to a ϳ5-fold lower LPL concentration when LDL receptors were up-regulated by lovastatin (Fig. 1A). This surprising result suggests that binding affinity was dependent on the number of LDL receptors, as would be expected if LPL simultaneously interacted with multiple LDL receptors. Another interesting observation in Fig. 1A is that GST-LPLC, a fusion protein containing the carboxylterminal segment (residues 313-448) of LPL, was also able to enhance VLDL binding in a dose-dependent and lovastatinregulated manner, albeit with a lower maximal binding. This indicates that the carboxyl-terminal domain of LPL, which lacks catalytic activity, contributes to LPL's interaction with LDL receptors. LPL also significantly promoted degradation of 125 I-VLDL by LDL receptors in a dose-dependent manner (Fig.  1B). Up-regulation of LDL receptors enhanced LPL-dependent degradation up to 4-fold. Either with or without lovastatin treatment, significantly higher concentrations of GST-LPLC than LPL were required to enhance 125 I-VLDL degradation, and the magnitude of stimulation was much lower than that obtained with LPL.
To further examine the contribution of LDL receptors to LPL-dependent VLDL degradation, studies were performed in cells lacking LRP. Results in Fig. 2 show degradation of 125 I-LDL and 125 I-␣ 2 M* by wild-type murine embryonic fibroblasts (MEF cells) and by mutant MEF cells lacking LRP (PEA13 cells). Unlike MEF cells, PEA13 cells are completely devoid of LRP activity and are unable to catabolize 125 I-␣ 2 M* ( Fig. 2A). In contrast, 125 I-LDL degradation is comparable in both cell lines and is stimulated 5-7-fold by lovastatin treatment (Fig.  2B). This indicates that the lack of LRP in PEA13 cells does not affect the level of LDL catabolism or its up-regulation by lovastatin. As expected, catabolism of 125 I-␣ 2 M* was not affected by lovastatin treatment.
We compared the effect of LPL and lovastatin treatment in normal (FSF) and LDL receptor-lacking (FH) human fibroblasts and wild-type (MEF) and LRP-deficient (PEA13) murine embryonic fibroblasts (Fig. 3). LPL promoted 125 I-VLDL degradation in a dose-dependent manner in untreated FSF cells. LPL stimulated VLDL degradation up to 15-fold in lovastatintreated normal fibroblasts. In three separate experiments, up-regulation of LDL receptors with lovastatin resulted in a 4 -5fold increase in LPL-induced 125 I-VLDL degradation by FSF cells (Fig. 3A). In control experiments (data not shown), catabolism of 125 I-LDL in the absence of LPL was increased by 8 -10-fold by lovastatin treatment. These data suggest that LDL receptors mediate the vast majority of LPL-dependent 125 I-VLDL degradation in lovastatin-treated cells. Previously, we showed that LPL induced uptake of 125 I-VLDL via LRP (7,9), and this finding was confirmed here (Fig.  3B) with a 10 -20-fold induction of 125 I-VLDL catabolism by gradually increasing the concentration of LPL. However, compared with normal fibroblasts, the magnitude of 125 I-VLDL degradation was ϳ10-fold lower and not significantly altered by lovastatin treatment. Even in the absence of lovastatin treatment, normal fibroblasts displayed 2.5-4-fold higher catabolism in the presence of LPL than FH cells (Fig. 3, A and B). This difference probably reflects degradation by the LDL-receptor pathway, since degradation of 125 I-␣ 2 M* by normal and FH cells was nearly identical, indicating that both cell lines express similar levels of LRP (data not shown). LPL-dependent 125 I-VLDL degradation by LRP-deficient PEA13 cells was upregulated by lovastatin treatment (Fig. 3C). In the presence or absence of lovastatin treatment, LPL-induced 125 I-VLDL degradation by PEA13 cells was comparable with that seen in MEF cells. Taken together, these results suggests that a majority of LPL-dependent VLDL catabolism occurs via the LDL receptor pathway.
We next studied the catabolism of increasing concentrations of 125 I-VLDL in the presence of a fixed LPL concentration (Fig.  4). At all concentrations of 125 I-VLDL, degradation was enhanced severalfold by the presence of 1 g/ml LPL in FSF, HUVECs, MEF, and PEA13 cells. LPL-dependent degradation of 125 I-VLDL was strongly up-regulated by lovastatin treatment in all four cell lines, suggesting a role for LDL receptors. Unlike normal fibroblasts, HUVECs express very low levels of LRP and are unable to catabolize significant amounts of 125 I- deficient PEA13 cells (Fig. 5). When cells were incubated with 20 g/ml LDL to down-regulate LDL receptors prior to the assay, MEF cells displayed ϳ30% greater LPL-induced VLDL degradation (Fig. 5). This agrees with results shown in Fig. 3C. But in Fig. 4, C and D, LPL-induced VLDL degradation was slightly greater in PEA13 cells than in MEF cells. Thus the difference between MEF and PEA13 was too small to be seen in all experiments, despite the fact that in most experiments MEF cells displayed slightly greater LPL-dependent VLDL catabolism than PEA13 cells.
The receptor-mediated component of VLDL catabolism was measured by the ability of RAP to compete for the effects of LPL. RAP has previously been shown to inhibit catabolism via all members of the LDL receptor family (46, 48 -50). As shown in Fig. 5, RAP inhibited Ͼ50% of the catabolism in both MEF and PEA13 cells. The ability of RAP to inhibit degradation in LRP-lacking PEA13 cells provides evidence that the LDL receptor pathway contributes to LPL-dependent VLDL degradation, because no other members of the LDL receptor family are expressed by these cells.
In control experiments (data not shown) LPL-stimulated 125 I-VLDL degradation was partially inhibited by anti-LDL receptor monoclonal antibody IgG-C7, anti-LPL antibody mAb7, and RAP, but not by nonimmune murine IgG or anti-LRP antibody Rb777. However, Rb777 completely inhibited degradation of 125 I-␣ 2 M*.

LPL-stimulated Degradation of Protein-free Intralipid Emulsion Is Mediated by LDL Receptors-
To determine if the ability of LPL to induce 125 I-VLDL catabolism via LDL receptors could be due to a direct interaction between LPL and LDL receptors, we studied catabolism of protein-free triglyceride and phospholipid emulsions. In these experiments, neither apoE nor apo B100 were present. Likewise, the activator of LPL, apoC-II, was not present. The emulsion was subjected to ultracentrifugation to isolate particles with S f 100 -400, which were labeled with [ 3 H]cholesteryl oleyl ether, a nondegradable marker of cellular binding or uptake. Results in Fig. 6 show that in FSF cells, basal uptake of the emulsion was low, but LPL promoted the internalization of these particles in a dose-dependent manner, even in the absence of any apolipoproteins. This suggests a direct interaction between LPL and cell surface receptors. FSF cells treated with lovastatin were able to take up almost twice as many emulsion particles than untreated cells, suggesting an involvement of the LDL receptor pathway. The average induction by lovastatin was 1.7-fold Ϯ 0.3 in three separate experiments. The fold induction was the same even when the data were corrected for DNA instead of protein. Likewise MEF and PEA13 cells also internalized protein-free lipid emulsion in the presence of LPL (data not shown). Internalization was ϳ2-fold higher in MEF and PEA13 cells treated with LPDS to up-regulate LDL receptors (data not shown).

VLDL Binding and Degradation via LDL Receptors Is Inhibited by Mutant GST-LPLCs-
We showed previously that in the absence of LPL, VLDL particles bind to LDL receptors but not to LRP (24). We have also shown that GST-LPLC contains both lipoprotein and receptor binding domains (17). Two tryptophans at positions 393 and 394 are required for binding lipoproteins (17). A site-directed double mutant, GST-LPLC ww , with both tryptophans changed to alanines, and a triple mutant, GST-LPLC www , with tryptophans at positions 390, 393, and 394, replaced with alanine were tested for their effects on VLDL binding and catabolism. GST-LPLC ww (17) and GST-LPLC www (data not shown) do not bind VLDL. After lovastatin treatment, the vast majority of 125 I-VLDL binding and degradation in FSF cells is LDL receptor-mediated (24). GST did not have any effect on either binding at 4°C or degradation at 37°C. On the other hand, GST-LPLC ww and GST-LPLC www inhibited both 125 I-VLDL binding (Fig. 7A) and degradation (Fig. 7B) in a dose-dependent manner. Similarly, both mutants completely inhibited 125 I-LDL degradation as well (data not shown).

LPL Binds to LDL Receptors in Solid-phase
Assays-Next, we tested the ability of LPL to bind directly to LDL receptors in solid-phase assays. Microtiter plates coated with increasing amounts of LPL were incubated with partially purified LDL receptors. The amount of receptor bound was then quantitated using IgG-C7, a specific anti-LDL receptor antibody. Fig. 8A shows that the amount of LDL receptor bound is directly proportional to the amount of immobilized LPL. When IgG-C7 was exposed to LPL-coated wells in the absence of added LDL receptors, there was relatively little binding. Also, when nonimmune mouse IgG was used as a control instead of IgG-C7, there was no detectable binding (data not shown). In another approach, LDL receptors from partially purified cell extracts were specifically immobilized on microtiter wells using anti-LDL receptor antibody IgG-C7 (46). The wells were subsequently incubated with increasing amounts of LPL, and the amount of LPL bound was quantitated using 125 I-labeled IgG mAb7, a monoclonal antibody against LPL (Fig. 8B). In the presence of LDL receptors, the amount of LPL bound increased with the dose of LPL added; binding of LPL in the absence of LDL receptors was significantly lower.
We investigated the ability of LPL or GST-LPLC to promote binding of 125 I-VLDL to LDL receptors in solid-phase assays (Fig. 9). These assays were done at 4°C, a temperature which inhibits lipolytic activity. In the absence of either protein, 125 I-VLDL bound to LDL receptor-coated wells (Fig. 9A), as shown previously (24). Both LPL and GST-LPLC stimulated the binding of 125 I-VLDL to immobilized LDL receptors (Fig. 9A). Nonspecific binding to BSA-coated wells was Ͻ5% and has been subtracted from these data. In Fig. 9B, wells coated with anti-LDL receptor antibody IgG-4A4 were used to immobilize increasing amounts of LDL receptors from DE52 eluants (LDL receptor preparation) of normal fibroblasts. At a fixed concentration of 125 I-VLDL, binding was dependent on the amount of LDL receptors added and was enhanced by 2-3-fold in the presence of LPL (Fig. 9B).

DISCUSSION
It is well established that LPL promotes the catabolism of triglyceride-rich lipoproteins by a variety of cultured cells. Pre- vious studies show that LPL promotes lipoprotein degradation via LRP (6 -10). The VLDL receptor and GP330/LRP-2 are implicated in the process as well (10 -13). Here we demonstrate that LDL receptors also contribute to the LPL-dependent enhancement of lipoprotein catabolism.
The present data establish that LPL stimulates VLDL catabolism via LDL receptors and binds directly to LDL receptors. LPL induces catabolism of normal VLDL in FSF, FH, HUVECs, MEF, and PEA13 cells. Since PEA13 cells do not express either LRP, GP330, or VLDL receptors (39,51), LPLdependent VLDL catabolism by these cells is probably mediated by LDL receptors. Treatment with lovastatin to up-regulate LDL receptors greatly enhances LPL-stimulated VLDL degradation in all four cell lines possessing an intact LDL receptor pathway. Up-regulation of LDL receptors results in a 2-fold increase in LPL-dependent binding and uptake of protein-free intralipid emulsion by normal fibroblasts, MEF, and PEA13 cells. Since intralipid emulsion do not contain any apolipoprotein, this result provides evidence for a direct interaction between LPL and LDL receptors. We confirmed the direct binding of LPL to LDL receptors in solid-phase assays. LDL receptors bind to LPL-coated microtiter wells in a dose-dependent manner. Both LPL and a fusion protein containing its carboxyl-terminal domain (GST-LPLC) promote binding of 125 I-VLDL to LDL receptor-coated wells. A ligand specific for LRP, 125 I-␣ 2 M*, does not bind to these wells (46), eliminating the possibility that the LPL effect may be due to LRP contamination.
Studies with various competitors provide further evidence that LPL stimulates VLDL degradation via LDL receptors. Mutations of tryptophan to alanine in GST-LPLC at positions 393, 394, and 390 (GST-LPLC www ) or at positions 393 and 394 (GST-LPLC ww ) abolish its binding to VLDL without affecting binding to LDL receptors or LRP (data not shown and Refs. 16 and 17). Both mutants competitively inhibit degradation of 125 I-VLDL by normal fibroblasts in which the LDL receptor pathway is up-regulated. RAP also competes for LPL-induced 125 I-VLDL degradation in both MEF and PEA13 cells, because it is a ligand for LDL receptors (46). Similarly, antibodies against the LDL receptor, but not LRP, partially compete for LPL-mediated lipoprotein degradation (data not shown).
Lipolysis is not required for LPL-dependent promotion of VLDL binding and catabolism. This is suggested by solid-phase and cell experiments at 4°C when the enzymatic activity is negligible. Moreover, LPL promotes the internalization of protein-free intralipid emulsions that are devoid of apoC-II, an essential co-factor for LPL's lipolytic activity. Rumsey et al. (21) also showed that LPL enhances the internalization of proteinfree triglyceride emulsion particles by macrophages. GST-LPLC, a fusion protein containing the carboxyl-terminal noncatalytic domain of LPL, also stimulates 125 I-VLDL binding at 4°C and degradation at 37°C as was shown previously in studies on LRP (17,18). However, GST-LPLC was significantly less potent than LPL. This may be explained by the inability of GST-LPLC to undergo dimerization (17) or optimal protein folding. Nykjaer et al. (52) showed that dimeric LPL is much more potent than monomeric LPL at stimulating 125 I-VLDL binding to immobilized LRP. The effect of LPL appears to depend on the ratio of VLDL to LPL. Thus, increasing concentrations of LPL to ϳ30 nM or GST-LPLC to ϳ700 nM induces degradation of 3.5 nM VLDL in a dose-dependent manner. At higher concentrations, when LPL and GST-LPLC are present in even larger molar excess over VLDL, they behave as competitive inhibitors of LDL receptor-mediated catabolism (data not shown) as observed previously in studies of LRP (18).
Despite the fact that lipolysis is not required for LPL-dependent VLDL catabolism, it clearly enhances catabolism. Lipolysis may expose or activate the receptor-recognition domains of apoE or apoB100 and contribute to LPL's effect (53)(54)(55)(56)(57). This is consistent with the finding that the lipid component of lipoprotein particles may modulate the reactivity of the apoprotein component (58). Also, Takahashi et al. (12) showed that the structurally unrelated Pseudomonas lipase can enhance VLDL catabolism via the VLDL receptor. Thus, GST-LPLC's lack of catalytic activity may be partly responsible for its relatively lower potency than LPL to the stimulate VLDL catabolism at 37°C. In our study, internalization of proteinfree emulsion increased only 2-fold following up-regulation of LDL receptors in FSF cells, whereas under similar conditions VLDL catabolism was increased up to 5-fold. Thus, the magnitude of LPL-induced catabolism of the emulsion is less than that obtained with native VLDL. This suggests that the effect of LPL on VLDL catabolism is only partially dependent on direct interaction of LPL with the receptor.
Our studies indicate that the affinity of LPL for LDL receptors is lower than its affinity for LRP, GP330/LRP-2, or VLDL receptors. The K D values reported for these receptors are, respectively, 34, 6, and 1 nM (11,12,17). However, due to the low affinity of LPL for LDL receptors as compared with its high nonspecific binding, we could not accurately determine its affinity in solid-phase or cell surface assays. Also, we were unable to demonstrate LPL binding to LDL receptors by ligand-blotting due to high nonspecific binding of LPL to the membrane support at the concentrations of LPL required.
Our studies indicate a major role for LDL receptors as compared with LRP in LPL-dependent VLDL catabolism. By comparing FSF and MEF cells, which express both LDL receptors and LRP, to their mutant counterparts that lack either LDL receptors or LRP, we were able to demonstrate that LDL receptors contribute to the majority of LPL-dependent VLDL catabolism. Only when LDL receptors were not up-regulated could a significant contribution of LRP to LPL-dependent VLDL catabolism be demonstrated. However, this may reflect the situation in vivo where the LDL receptor pathway is usually down-regulated (59).
Although we studied VLDL, it is likely that these observations apply to chylomicron catabolism as well, because the remnants of large VLDL and chylomicrons appear to share the same catabolic pathway (60). The relative contributions of LDL receptors and LRP to remnant catabolism in vivo are not clearly defined, but evidence presented here and elsewhere implicates both receptors (61)(62)(63)(64)(65)(66)(67). Accumulation of triglyceride-rich precursors of LDL has been reported in humans with severe LDL receptor deficiency (61). Treatment of hypertriglyceridemic subjects with lovastatin to increase their LDL receptor number results in a reduction of their VLDL levels (62). Furthermore chylomicron remnant catabolism is impaired in Watanabe heritable hyperlipidemic rabbits, which lack functional LDL receptors and in cholesterol-fed rabbits with low LDL receptor activity (63,64). Willnow et al. (65) implicated LRP in the catabolism of remnant particles when they found that over-expression of RAP, an inhibitor of LRP, inhibits chylomicron remnant clearance by transgenic mice lacking LDL receptors. However, this effect was not seen in mice having normal LDL receptor activity, suggesting a role for LDL receptors in chylomicron remnant clearance (66). Consistent with this idea, Willnow et al. recently found that chylomicron remnants accumulate in RAP knock-out mice that have deficient LRP activity (67). However, the remnants accumulate only if the mice also lack LDL receptors.
The observation that LPL can bind directly to LDL receptors as well as to other members of the LDL receptor family is consistent with other studies that provide evidence for an apolipoprotein-like role for LPL in remnant catabolism (68 -71). LPL is present in nanomolar concentrations in human plasma and is associated with lipoproteins (68). Nordenstrom et al. (69) found that plasma LPL activity directly correlates with plasma triglyceride levels during infusion of triglyceride emulsion in humans. Peterson et al. (70) demonstrated that plasma LPL activity dramatically increases in rats after intravenous administration of intralipid emulsions. Elevated plasma LPL caused by heparin injection or intravenous infusion of Intralipid results in increased LPL activity in the liver. Moreover, hepatectomy of hypertriglyceridemic rats causes a progressive increase in plasma LPL activity (70). Vilaro et al. (71) confirmed these findings by showing that injection of intralipid in starved rats enhances LPL activity in the liver. These data suggest that a fraction of LPL can bind to triglyceride-rich lipoproteins under normal circumstances and undergo transport to the liver where LPL may function as a recognition signal for the hepatic uptake of lipoproteins.
In addition to its potential role in hepatic clearance of triglyceride-rich particles, another interesting potential role for LPL is its participation in atherogenesis. Corey and Zilversmit (72,73) showed a coordinate increase in the LPL activity and cholesterol content of rabbit aortas with cholesterol feeding and proposed that atherogenesis may result from abnormal postprandial lipolytic processing of triglyceride-rich particles. LPL is synthesized by macrophages and macrophage-derived foam cells present in atherosclerotic plaques and to a lesser degree by smooth muscle cells as well (74). Rutledge and Goldberg (75) recently reviewed evidence that in these tissues, LPL may play a role in the pathogenesis of atherosclerosis by promoting the retention of VLDL in the arterial wall. Furthermore Rapp et al. (76) and Chung et al. (77) have found intact VLDL and surface remnants of triglyceride-rich particles in human atherosclerotic plaques. Studies in HUVECs reported here indicate that LPL induces catabolism of VLDL via LDL receptors in endothelial cells. Other cells present in atheromas, including macrophages and smooth muscle cells also participate in LPL-induced uptake of lipoproteins (21, 78 -80). The potential role of vascular LPL in atherogenesis is an important area for future research.