low density lipoprotein receptor-related protein and a330 bind

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 hy The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 267, No. 36, Isaue of December 25, pp. 26172-26180,1992 Printed in U. S. A.

Low Density Lipoprotein Receptor-related Protein and a 3 3 0 Bind Similar Ligands, Including Plasminogen Activator-Inhibitor Complexes and Lactoferrin, an Inhibitor of Chylomicron Remnant Clearance*

(Received for publication, July 13, 1992)

Thomas E. Willnow#$, Joseph L. Goldstein$, Kim Orthll, Michael S . Brown#, and Joachim Hem#** From the Departments of $Molecular Genetics and 11 Biochemistry, University of Texas Southwestern Medical Center, Dallas, Texas 75235

The low density lipoprotein receptor-related protein/ cYz-macroglobulin receptor (LRP) and gp330, two members of the low density lipoprotein receptor gene family, share a multitude of cysteine-rich repeats. LRP has been shown to act as an endocytosis-mediating receptor for several ligands, including protease-anti- protease complexes and plasma lipoproteins. The for- mer include az-macroglobulin-protease complexes and plasminogen activator inhibitor-activator complexes. The latter include chylomicron remnant-like particles designated &very low density lipoproteins (8-VLDL) complexed with apoprotein E or lipoprotein lipase. The binding specificity of gp330 is unknown. In the current studies we show that gp330 from rat kidney membranes binds several of these ligands on nitrocellulose blots. We also show that both LRP and gp330 bind an additional ligand, bovine lactoferrin, which is known to inhibit the hepatic clearance of chylomicron remnants. Lactoferrin blocked the LRP-dependent stimulation of cholesteryl ester synthesis in cultured human fibroblasts elicited by apoprotein E-j3-VLDL or lipoprotein lipase-&VLDL complexes. Cross-competition experiments in fibroblasts showed that the multiple ligands recognize at least three distinct, but partially overlapping sites on the LRP molecule. Binding of all ligands to LRP and gp330 was inhibited by the 39-kDa protein, which co-purifies with the two receptors, suggesting that the 39-kDa protein is a universal regulator of ligand binding to both receptors. The correlation of the inhibitory effects of lactoferrin in vivo and in vitro support the notion that LRP functions as a chylomicron remnant receptor in liver. LRP and gp330 share a multiplicity of binding sites, and both may function as endocytosis-mediating receptors for a large number of ligands in different organs.

A family of three structurally related cell surface receptors mediates the endocytosis of lipoproteins and other plasma proteins in mammalian cells (reviewed in Ref. 1). Each of

Health Grant HL 20948 and by grants from the Lucille P. Markey * This research was supported in part by National Institutes of

Charitable Trust and the Perot Family Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduer- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

§ Recipient of Postdoctoral Fellowship Wi 1158/1-1 from Deutsche Forschungsgemeinschaft.

** Recipient of a Lucille P. Markey Scholar Award and a Syntex Scholar Award.

these receptors exhibits the following five features: 1) cysteine-rich repeats of the ligand binding or complement type; 2) cysteine-rich repeats of the epidermal growth factor type; 3) YWTD (Tyr-Trp-Thr-Asp) repeats; 4) a single membrane- spanning region; and 5) at least one copy of an NPXY internalization signal that directs the receptors to coated pits. The proteins differ in the number and arrangements of the repeated sequences. All of the proteins bind Ca2+, which is required for ligand binding.

The low density lipoprotein (LDL)’ receptor is the proto- type for this family (2, 3). It contains one cluster of seven complement-type repeats and one cluster of three growth factor repeats, the latter separated by a cysteine-poor spacer that contains five copies of the YWTD sequence. The LDL receptor binds plasma lipoproteins that contain apoB-100 or apoE, and it is responsible for the removal of most interme- diate density lipoproteins and LDL from plasma. Both of these lipoproteins accumulate in plasma of patients with familial hypercholesterolemia (FH), who have mutations in the LDL receptor gene (4).

The second member of the LDL receptor gene family is designated LDL receptor-related protein/a2-macroglobulin receptor, which we refer to here as LRP (1). This protein, whose cDNA was cloned by homology with the complement repeat region of the LDL receptor sequence (5), is much larger than the LDL receptor (4525 uersus 839 amino acids). It contains 31 complement-like repeats and 22 growth factor repeats that are separated by eight spacer regions, each containing multiple YWTD repeats (1, 5). LRP does not bind LDL, but it does bind @-migrating very low density lipoproteins (P-VLDL) that have been enriched in vitro with apoE (6, 7). @-VLDL are a mixture of cholesterol-rich remnant lipoproteins derived from intestinal chylomicrons and hepatic VLDL (8). Although LRP is present on a variety of cell types and tissues, its function in the body is expressed predomi- nantly in the liver (9). Hepatic LRP has been proposed to act as a receptor for chylomicron remnants that become enriched with apoE during passage through hepatic sinusoids (1,5,9).

Strickland et al. (10) and Kristensen et al. (11) found that the receptor for a2-macroglobulin-protease complexes was identical in its structure to that of LRP. a2-Macroglobulin, a

The abbreviations used are: LDL, low density lipoprotein; a2- macroglobulin*, oc2-macroglobulin activated with methylamine; p- VLDL, @-migrating very low density lipoproteins; FH, familial hypercholesterolemia; GST, glutathione S-transferase; LPL, lipoprotein lipase; LRP, LDL receptor-related protein; PAI-1, plasminogen activator inhibitor-1; tPA, tissue-type plasminogen activator; MOPS, 4- morpholinepropanesulfonic acid; PAGE, polyacrylamide gel electrophoresis.

26172

LRP and gp330 Bind Similar Ligands 26173

plasma protease inhibitor, circulates in an inactive form. Upon binding a protease, the a2-macroglobulin is altered SO

that it binds to receptors on hepatocytes and is rapidly cleared from the circulation (12).

The binding of apoE-enriched 8-VLDL to LRP can be assayed in cultured fibroblasts from FH patients that lack LDL receptors. After binding to LRP, apoE/@-VLDL is deliv- ered to the lysosomes and the liberated cholesterol activates acyl-CoA:cholesterol acyltransferase (6). The latter is moni- tored by measuring the rate of incorporation of ['4C]oleate into cholesteryl esters by the cells. These human fibroblasts also take up activated az-macroglobulin. Preincubation of the cells with an antibody against LRP coordinately decreases the uptake of both ligands (13).

LRP from liver and placenta co-purifies with a 39-kDa protein (also known as a2-macroglobulin receptor-associated protein or az-MRAP) (14, 15). The 39-kDa protein binds to LRP in vitro, and it blocks the binding of apoE/p-VLDL and az-macroglobulin (13-16).

LRP also binds another ligand, lipoprotein lipase (LPL) (17), an enzyme that is normally bound to the surface of endothelial cells in adipose tissue and muscle (18). Under some conditions LPL can be released into the circulation attached to lipoproteins (19). Beisiegel et al. (17) showed that attachment of lipoprotein lipase to either chylomicrons or 8- VLDL enhances binding to LRP.

Another class of recently recognized ligands for LRP are complexes of plasminogen activators and their inhibitors (20- 22). By ligand blot analysis, LRP binds complexes of tissue- type plasminogen activator (tPA) and plasminogen activator inhibitor (PAI-1) (20). In uiuo, complexes of tPA-PAI-1 are cleared rapidly in the liver (23). In cultured liver cells, tPA is taken up by several mechanisms, one of which requires complex formation of the protease with PAI-1 (20, 21). LRP also binds complexes of urokinase-type plasminogen activator and

The third member of the LDL receptor gene family, called gp330, is less well characterized. This protein was originally identified as a target molecule in kidney for autoimmune antibodies in rats with Heymann's nephritis (24). gp330 is normally located in certain epithelial cells, including those of the kidney (renal glomerulus and proximal tubule), yolk sac, and epididymis, but not in liver (24, 25). Antibodies against a 3 3 0 cause it to slough from the surface and to deposit in the basement membrane, causing nephritis. Fragments of a partial cDNA for gp330 have been cloned (26). All of the fragments cloned to date show amino acid sequences homol- ogous with LRP, including the complement-like repeats, growth factor repeats, YWTD repeats, and NPXY internalization motif. Thus far, it is not known whether any of the ligands that bind to LRP or the LDL receptor also bind to gp330.

In the current study we use a nitrocellulose blotting assay to show that gp330 binds several of the known ligands for LRP. We also show that both proteins bind an additional ligand, bovine lactoferrin. This protein was previously reported to inhibit the hepatic uptake of chylomicron remnants when infused into the circulation of intact animals (27, 28). The multiple ligands that bind to LRP appear to recognize at least three distinct sites on the LRP molecule. The 39-kDa protein is capable of displacing ligands from all three sites.

PAI-1 (20, 22).

EXPERIMENTAL PROCEDURES

Materials-@-VLDL (d < 1.006 g/ml) was prepared from the plasma of rabbits fed for 4 days with a 2% (w/w) cholesterol, 10% (v/ w) coconut oil diet as described (7). The ratio of total cholesterol to protein in @-VLDL preparations ranged from 10 to 17. Complexes of

t-PA. PAI-1 were prepared as previously described (20). Recombinant human apoE (E3 isoform), obtained from Escherichia coli (lot No. 16-074-R07) (29), was kindly provided by Tikva Vogel (Biotechnology General, Rehovot, Israel). IgG-2E1, a mouse monoclonal antibody directed against purified rat LRP, was prepared as described (13). Three different mouse monoclonal antibodies directed against rat gp330 were kindly provided by Marilyn G. Farquhar (University of California San Diego, La Jolla, CA). A rabbit polyclonal antibody directed against rat LRP was prepared as described (6). Human 012-

macroglobulin activated with methylamine (az-macroglobulin') (14) was kindly provided by Dudley Strickland (American Red Cross, Rockville, MD). Bovine a,-macroglobulin was purified from plasma on a Zn2+-chelate affinity column and activated with methylamine (30). Recombinant glutathione S-transferase (GST) and rat 39-kDa fusion protein (designated GST-39 kDa protein) were produced in DH5a bacteria transformed, respectively, with a GST expression plasmid and a GST-39 kDa expression plasmid as previously described (13). We obtained '%aC12 (16 mCi/mg) from Amersham; streptavidin from Pierce; bovine lactoferrin (catalog No. L-4765), imipramine (catalog. No. I-7379), and bovine milk lipoprotein lipase (affinity purified; 1320 units/mg protein; catalog No. L-2254) from Sigma. The lipoprotein lipase (LPL) was prepared as follows: -4 mg of enzyme suspended in 3.8 M ammonium sulfate, 20 mM Tris-HC1 at pH 8 was centrifuged (10,000 rpm, 10 min, 4 "C), and the resulting pellet was resuspended in 1 ml of 20 mM Tris-HC1,O.l M NaCl at pH 8, dialyzed at 4 "C against 12 liters of buffer for 16 h, stored at 4 "C at a final protein concentration of 1.8 mg/ml, and used within 10 days. All Iz5I-labeled ligands were radiolabeled by the IODO-GEN method (31).

LRP and gp330-A DEAE-cellulose fraction of solubilized rat liver membranes containing both LRP and LDL receptors was prepared as previously described (6). Extracts from rat kidney membranes enriched in gp330 were prepared at 4 "C by homogenizing whole kidneys in a Dounce homogenizer in 3 volumes (v/w) of buffer containing 20 mM Tris-HC1, 2 mM MgC12, 0.25 M sucrose, 1 mM phenylmethylsulfonyl fluoride at pH 7.5. The preparation was centrifuged at 4 "C successively at 1000 X g for 10 min, 10,000 X g for 10 min, and 100,000 X g for 45 min. The 100,000 X g membrane pellet was then resuspended in 50 mM Tris-HC1, 2 mM CaCI2, and 80 mM NaCl at pH 8; passed through multiple needles of decreasing diameter; and stored at -20 "C at a protein concentration of 32 mg/ml.

Ligand Blotting-Partially purified membrane preparations from rat liver and kidney were subjected to SDS-PAGE under nonreducing conditions. The samples (nonboiled) were subjected to electrophoresis either on a 5% Mini-PROTEAN I1 Electrophoresis Cell (Bio-Rad) a t 150 V for 2 h at room temperature or on a 3-8% gradient gel as described (6, 7). After electrophoresis, the proteins were transferred to nitrocellulose paper at 4 "C (7). Individual strips of nitrocellulose (containing -30 pg of protein/strip) were processed for ligand blotting by incubation in 1 ml of ligand buffer (50 mM Tris-HC1, 2 mM CaC12, 80 mM NaCl, 5% (w/v) bovine serum albumin, 0.1% (v/v) Triton X- 100 at pH 8) for 1 h at room temperature. The solution was then replaced by 1 ml of ligand buffer containing the indicated '251-labeled ligand. After incubation for 1 h at room temperature, the strips were washed (7), and subjected to autoradiography.

Cells-Diploid human skin fibroblasts were grown in monolayer at 37 "C in a 5% COP atmosphere and set up for experiments as described (6,7). FH 1003 fibroblasts were from a French-Canadian subject who is homozygous for a >lO-kilobase deletion that removes the promoter and first exon of the LDL receptor gene (32). FH 1003 fibroblasts were derived from a new skin biopsy obtained from the same subject whose fibroblasts were previously referred to as FH 808 (6).

Cholesterol Esterification Assay-On day 0, 3-4 X 10' cells were seeded into 60-mm Petri dishes and grown in 10% (v/v) fetal calf serum as described (33). On day 7, each monolayer received a final volume of 2 ml of medium A (Dulbecco's modified Eagle's medium (-glutamine) supplemented with 1 mM P-mercaptoethanol and 5% fetal calf serum) containing a mixture of either apoE/@-VLDL or lipoprotein lipase/@-VLDL as indicated in the legends. The apoE/@- VLDL or lipoprotein lipase/@-VLDL mixture was preincubated for 1 h at 37 "C in 0.5-0.6 ml of medium A before addition to the culture medium (6, 7). After the indicated time, cells were pulse-labeled with 0.2 mM ["Cloleate bound to albumin at a specific activity of 7,500 to 10,500 dpm/nmol and harvested for measurement of cholesteryl ["C] oleate and ["Cltriglycerides as described (33).

Proteolytic Degradation of 1251-Labeled Proteins by Fibroblasts Monolayers-Cell-dependent proteolysis of Iz5I-labeled ligands was measured as previously described (33) and is expressed as micrograms

26174 LRP and gp330 Bind Similar Ligands

of 'Z51-labeled trichloroacetic acid-soluble (non-iodide) material released into the culture medium per mg of total cell protein.

RESULTS

Fig. 1 shows the results of nitrocellulose blotting assays comparing the binding of various ligands to liver LRP and kidney gp330. The LRP preparation consisted of a crude fraction from a single DEAE column separation of rat liver membrane proteins. The gp330 preparation consisted of a detergent extract of kidney membranes without further puri- fication. Both protein preparations exhibited many Coo- massie-stained bands on SDS-polyacrylamide gels. We inten- tionally used these crude preparations because the contami- nating proteins serve as controls for specificity of ligand binding to LRP and gp330. A monoclonal antibody to LRP visualized the high molecular weight protein in liver, but not in kidney (lanes 4 and 5). In contrast, a mixture of three monoclonal anti-gp330 antibodies visualized gp330 in the kidney extract, but not in the liver (lanes 6 and 7 ) . Both proteins bound 45Ca2' (lanes 8 and 9) . Both proteins also bound bovine Iz5I-labeled lactoferrin (lanes 10 and I l ) , lZ5I- LPL (lanes 12 and 13), and lZ5I-GST-39 kDa (lanes 14 and 15). This fusion protein, which contains the entire coding

Coomassie Anti-LRP Antigp330 "Ca

I L K ' n - ?q i d f P '

% Q;; 200 kDa - -

1 2 3 4 5 6 7 8 9

200 kDa - r

92 kDa - 10 11 12 13 14 15 16 I 7

FIG. 1. Binding of multiple ligands to LRP and a 3 3 0 in rat liver and kidney membrane extracts. Partially purified membranes from rat liver ( L ) and rat kidney ( K ) (each containing -0.5 mg of protein) were subjected to preparative nonreducing 5% SDS- PAGE, and either stained with Coomassie Brilliant Blue (lanes 2 and 3) or transferred to nitrocellulose as described under "Experimental Procedures" (lanes 4-1 7). In lane 1, protein molecular weight markers (RAINBOW, Amersham) of the indicated molecular weight are shown. The arrow denotes the position of migration of LRP and gp330, which correspond approximately to the position of migration of apoB-100 of LDL (-515 kDa). Replicate nitrocellulose strips in lanes 4-7 were subjected to immunoblot analysis with either 5 pg/ml monoclonal anti-LRP IgG-2E1 (lanes 4 and 5) or 5 pg/ml each of a mixture of three monoclonal anti-gp 330 antibodies (IgG-8, -C287.20B, and -D155F21) (lanes 6 and 7), followed by incubation with '251-labeled rabbit anti-mouse IgG (1 X IO6 cpm/ml). The strips were exposed at -70 "C to Kodak XAR-5 film for 3 h (lanes 4 and 5) or 16 h (lanes 6 and 7). Nitrocellulose strips in lanes 8 and 9 were incubated with 2 pCi/ml "CaC12 in 1 ml of buffer containing 10 mM MOPS, 50 mM KCI, 5 mM MgCI2, 0.02% (v/v) Tween 20 at pH 7. After washing two times for 3 min in water, the strips were exposed at -70 "C to film for 72 (lane 8 ) or 5 (lane 9) h. Strips in lanes 10- 17 were subjected to ligand blotting as described under "Experimental Procedures" with 1 X lo6 cpm/ml of the indicated '251-labeled ligand. The exposure times of the films were 2 h for lanes 10-14 and 16 and 30 min for lanes 15 and 17.

region of the 39-kDa protein was prepared in E. coli. In a control experiment neither LRP nor gp330 bound '251-GST (lanes 16 and 17).

As previously reported (6, 7), LRP from liver binds rabbit P-VLDL, but only after the lipoprotein has been enriched in vitro with apoE (Fig. 2, lune 2) . This binding was not inhibited by GST, but it was abolished by the GST-39 kDa fusion protein. Binding of P-VLDL to the LDL receptor in the same preparations did not require additional apoE, nor was it affected by the GST-39 kDa protein. Identical results were obtained with gp330 in the kidney membranes, i.e. P-VLDL bound to gp330, but only after enrichment with apoE (lane 6) , and binding was abolished by the GST-39 kDa protein (lune 8). These assays employ biotinylated 8-VLDL, which is visualized by the binding of 1251-streptavidin. The kidney membranes contain a streptavidin binding protein (denoted by the asterisk in lanes 5-8), which is not dependent on the presence of P-VLDL.

Another previously described ligand for LRP consists of a complex between tPA and its inhibitor, PAI-1. Fig. 3 shows that LRP in the liver membrane preparation bound '251-tPA, but only when it was present in a complex with PAI-1 (lane 2 ) . Binding was abolished by the GST-39 kDa protein (lane 4 ) , but not by GST (lane 3). Identical results were obtained with gp330 in the kidney preparation (lanes 5-8).

The above results indicate that gp330 and LRP both bind the same spectrum of ligands as determined by ligand blotting on nitrocellulose. To confirm that the binding of these ligands to LRP is functionally significant, we performed a series of experiments with cultured human fibroblasts that express LRP, but not gp330. The experiments were designed to determine whether the various ligands for LRP are taken up and metabolized and whether these processes could be inhibited by the GST-39 kDa protein. To prevent the possible con- founding of these results by LDL receptor binding, we used a strain of human fibroblasts obtained from a patient who is homozygous for a deletion that prevents any expression of LDL receptors (5).

To measure the receptor-mediated uptake of P-VLDL, we

LIVER ' KIDNEY

1 - gp330

LDL + - Receptor

*

FIG. 2. Ligand binding of apoE-enriched 8-VLDL to rat liver and kidney membrane extracts. Partially purified membranes from rat liver (3-8% gradient gel) or rat kidney (5% gel) were subjected to SDS-PAGE, and the proteins were transferred to nitrocellulose (-30 pg of protein/strip) and subjected to ligand blot analysis. Each replicate strip was preincubated a t room temperature for 30 min with either blotting buffer (lanes I , 2, 5, and 6 ) or blotting buffer containing 20 pg/ml GST (lanes 3 and 7 ) or 20 pg/ml GST-39 kDa fusion protein (lanes 4 and 8). Each strip was then incubated for 1 h at room temperature with 10 pg of protein/ml of biotinylated 0-VLDL that had been preincubated in the absence (lanes I and 5) or presence (lanes 2-4 and 6-8) of apoE at a final concentration of 20 pg/ml. Bound ligand was detected with '2511-labeled streptavidin (1 X lo6 cpm/ml), after which the strips were exposed to XAR film a t -70 "C for 2 (lanes 5-8) or 16 (lanes 1-4) h. Gels were calibrated with molecular weight markers as described in the legend to Fig. 1. The asterisk (*) denotes a 1251-streptavidin binding protein detected in kidney membrane extracts that migrates a t a position of -100 kDa.


FIG. 3. Ligand binding of tPA.PA1-1 complex to LRP and a 3 3 0 in rat liver and kidney membrane extracts. Partially purified membranes of rat liver or rat kidney were subjected to 5% SDS-PAGE, and the proteins were transferred to nitrocellulose (-30 pg of protein/strip) and subjected to ligand blot analysis as described under “Experimental Procedures.” Individual strips were incubated with 50 ng/ml 1251-labeled tPA (1 X lo5 cpm/ml) (lanes 1 and 5) or a complex of 50 ng/ml 1Z51-tPA with 0.5 pg/ml PAL1 (lanes 2-4 and 6- 8) that was prepared as described (20). Prior to addition of the Iz5I- labeled tPA or 1251-labeled tPA. PAI-1 complex, the strips were preincubated for 30 min at room temperature with 20 pg/ml GST (lanes 3 and 7) or 20 pg/ml GST-39 kDa fusion protein (lanes 4 and 8). The strips were exposed to XAR film for 48 h at -70 “C. Gels were calibrated with molecular weight markers as described in the legend to Fig. 1.

5 1 I

+Lp Lipase

0 10 20 30 40 50 p-VLDL (pg protein/ml)

FIG. 4. Cholesteryl [‘4C]oleate formation in FH fibroblasts lacking LDL receptors: stimulation by apoE/&VLDL (0) or lipoprotein lipase/fl-VLDL (0). On day 7, each monolayer received 2 ml of medium A containing the indicated concentration of j3-VLDL that had been preincubated with a fixed ratio of either apoE (protein ratio, P-VLDLlapoE = 1) (0) or lipoprotein (Lp) lipase (protein ratio; j3-VLDLllipoprotein lipase = 3.3) (0) as indicated. After incubation for 5 h at 37 “C, the cells were pulse-labeled for 2 h with 0.2 mM [“C] oleate, and their content of cholesteryl [‘4C]oleate was measured. Each value is the average of duplicate incubations. The amount of cholesteryl [“Cloleate formed in the absence of additions (0.04 nmol. h” mg protein”) was subtracted from each value.

used the cholesterol esterification assay (6, 7). This assay takes advantage of the observation that fibroblasts synthesize very few cholesteryl esters in the absence of an external source of cholesterol (33). Delivery of 8-VLDL to lysosomes releases cholesterol, and this increases the incorporation of [“Cloleate into cholesteryl [14C]oleate.

Fig. 4 shows that the addition of rabbit 8-VLDL to the LDL receptor-negative fibroblasts did not appreciably stimulate the incorporation of [14C]oleate into cholesteryl [“C] oleate. As reported previously (6, 7), when the P-VLDL had been preincubated with apoE, the lipoprotein caused a marked stimulation (Fig. 4, closed circles). A similar degree of stimulation was observed when the 8-VLDL was incubated with lipoprotein lipase instead of apoE (Fig. 4, open circles). At high concentrations the stimulatory activity of the lipoprotein lipase-enriched P-VLDL declined.

The stimulation of cholesteryl ester synthesis by apoE/P-

VLDL was abolished by the GST-39 kDa fusion protein (Fig. 5A). Similarly, the stimulation by LPL/P-VLDL was also reduced by GST-39 kDa protein (Fig. 58) . The reduction, however, was not as complete as the reduction achieved when apoE -8-VLDL complexes were used.

Preincubation of the fibroblasts with a polyclonal antibody against LRP led to a decrease in the ability of both apoE/P- VLDL and LPL/P-VLDL to stimulate cholesteryl ester synthesis (Fig. 6, A and B ) . We have previously shown that preincubation with this antibody depletes the cell surface of LRP (7). Nonimmune IgG had no inhibitory effect.

The ability of internalized lipoproteins to stimulate cholesteryl esterification is abolished by hydrophobic amines such as imipramine. These agents are thought to act by blocking the emergence of free cholesterol from the lysosome (34). Fig. 7A shows that imipramine abolished the stimulation of cholesteryl ester synthesis that was achieved by apoE/P-VLDL and LPL/P-VLDL. Imipramine did not affect the cholesteryl esterification when it was stimulated by 25-hydroxycholesterol plus cholesterol (Fig. 78) , nor did it inhibit [‘4C]oleate incorporation into [14C]triglycerides (data not shown). These data add further confirmation to the idea that LPL/P-VLDL required uptake and degradation in lysosomes in order to stimulate cholesteryl ester synthesis. We also observed that the LPL/B-VLDL-mediated stimulation of cholesteryl [“C] oleate synthesis, but not of [14C]triglyceride synthesis, was abolished when the cells were incubated with chloroquine, an inhibitor of lysosomal function (Table I).

Considered together, the studies in Figs. 4-7 indicate that the binding of LPL/@-VLDL to LRP that was observed previously in intact cells (17) is functionally significant in that it delivers P-VLDL to lysosomes where the cholesteryl esters of the lipoprotein are hydrolyzed and made available for cellular cholesteryl ester synthesis.

Bovine lactoferrin is an extremely potent inhibitor of the

A. APO E/&VLDL

100

GST-39 kDa

I GST-39 kDa -0

A ”

0 2 4 6 6 1 0 0 2 4 6 6 1 0 I ,

GST or GST-39 kDa Fusion Protein (pg/ml)

FIG. 5. Inhibition by 39-kDa protein of apoE/@-VLDL and lipoprotein lipase/&VLDL-stimulated cholesteryl [“Cloleate formation in FH fibroblasts lacking LDL receptors. On day 7, each monolayer received 1.5 ml of medium A containing the indicated concentration of GST (0) or GST-39 kDa fusion protein (0). After incubation for 1 h at 37 “C, each dish received 0.5 ml of medium A containing j3-VLDL at a final concentration of 25 pg of protein/ml that had been preincubated with either apoE at a final concentration of 25 pg/ml (panel A ) or lipoprotein (Lp) lipase at 7.5 pg/ml (panel B ) . After incubation for 5 h at 37 “C, the cells were pulse-labeled for 2 h at 37 “C with a 0.2 mM [14C]oleate, and their content of cholesteryl [“Cloleate was measured. A blank value of 0.07 nmol. h”.mg protein” corresponding to the amount of cholesteryl [14C]oleate formed in parallel dishes receiving j3-VLDL without apoE or lipoprotein lipase was subtracted from each value. Each value is a single incubation except for the 100% of control value, which represents the mean of triplicate incubations. The 100% of control value was 3.1 nmol . h” . mg protein”.


1 A. ApoE/pVLDL 1 8. LP Lipase/b-VLDL 1

0.6 Nonirnrnune

0.4

2.4

1.8

1.2

- I

IgG (re/rnl) FIG. 6. Inhibition by anti-LRP polyclonal antibody of the

apoE/@-VLDL and lipoprotein lipase/@-VLDL-stimulated cholesteryl [''Cloleate formation in FH fibroblasts lacking LDL receptors. On day 6, each monolayer received 2 ml of growth medium containing the indicated concentration of either nonimmune rabbit IgG (0) or polyclonal anti-LRP rabbit IgG (0). After incubation for 16 h, each monolayer received 1.5 ml of fresh medium A containing the same concentration of IgG followed by 0.5 ml of medium containing 0-VLDL at a final concentration of 15 pg of protein/ml that had been preincubated with either apoE at a final concentration of 15 pg/ ml (panel A ) or lipoprotein ( L p ) lipase at 5 pg/ml (panel B ) . After incubation for 5 h at 37 "C, the cells were pulse-labeled for 2 h with 0.2 mM ['4C]oleate, and the content of cholesterol [14C]oleate was measured. A blank value of 0.07 nmol. h".mgprotein" corresponding to the amount of cholesteryl ["Cloleate formed in dishes receiving no additions was subtracted from each value. Each value is the average of duplicate incubations.

- I

Imipramine (pM)

FIG. 7. Inhibition by imipramine of apoE/@-VLDL and lipoprotein lipase/@-VLDL-stimulated cholesteryl ['4C]oleate formation in FH fibroblasts lacking LDL receptors. On day 7, each monolayer received 1.5 ml of medium A containing the indicated concentration of imipramine. Panel A, after 15 min at 37 "C each dish received 0.5 ml of medium A containing P-VLDL at a final concentration of 30 pg of protein/ml that had been preincubated with either apoE at a final concentration of 40 pg/ml (A) or lipoprotein lipase (LPL) at 8 pg/ml (A). Panel E , each dish received 0.5 ml of medium A containing either 8-VLDL at a final concentration of 30 pg/ml that had been preincubated with apoE at 40 pg/ml (0) or a mixture of sterols containing 25-hydroxycholestero1 at a final concentration of 5 pg/ml and cholesterol at 10 pg/ml (0). After incubation for 5 h at 37 "C, the cells were pulse-labeled for 2 h at 37 "C with 0.2 mM ["Cloleate, and their content of cholesteryl ["Cloleate was measured. Blank values of 0.11 and 0.17 nmol. h",mg protein" for panels A and E , respectively, corresponding to the amount of cholesteryl [14C]oleate formed in parallel dishes receiving no additions were subtracted from each value. Each value represents the average of duplicate (no imipramine) or single incubations (+ imipramine). The 100% of control values were 7.3 (A), 4.6 (A), 2.1 (O), and 2.9 (0) nmol. h". mg protein".

ability of apoE/p-VLDL to stimulate cholesteryl esterification (Fig. 8). Complete inhibition was achieved at a lactoferrin concentration of 10 pg/ml. The lactoferrin also inhibited the ability of LPLIP-VLDL to stimulate cholesteryl esterification, but the effect was much less pronounced than was the effect with apoE/p-VLDL. A 50% inhibition was observed at 30 pg/ ml bovine lactoferrin, and the inhibition reached only 75% at 80 Fg/ml. The ability of lactoferrin to inhibit cholesteryl oleate formation was specific for apoE/P-VLDL. Lactoferrin

did not inhibit cholesteryl esterification when it was stimulated by a mixture of cholesterol and 25-hydroxycholesterol, which enters cells independently of any receptors (Fig. 8B) .

Table I1 shows that the human fibroblasts degraded 1251-~2- macroglobulin'. We have previously shown that this degradation requires binding to LRP, uptake by endocytosis, and degradation in lysosomes (13). The process was quantitatively inhibited by the GST-39 kDa protein (Table 11). Inclusion of lactoferrin at concentrations up to 100 pg/ml had no effect. '261-Labeled lactoferrin was degraded by the fibroblasts, and the degradation could be inhibited by about 50% with the GST-39 kDa protein. a2-Macroglobulin* at concentrations of up to 300 pg/ml had no inhibitory effect on the degradation of lZ5I-labeled lactoferrin.

Fig. 9 shows that the degradation of 1251-a2-macroglobulin* was completely inhibited by the GST-39 kDa protein in fibroblasts. On the other hand, lactoferrin had no inhibitory effect.

Another ligand for LRP in vitro is '251-labeled tPA.PA1-1 complex. Fig. 10 shows that Iz5I-labeled tPA.PA1-1 was degraded by the human fibroblasts. Degradation was completely abolished by low concentrations of the GST-39 kDa protein, indicating that it was mediated by LRP. Degradation was not inhibited, however, by any of the other ligands for LRP, including lipoprotein lipase/P-VLDL, lactoferrin, apoE/P- VLDL, or a,-macroglobulin' (Fig. 10, A and B) . To make certain that the lactoferrin preparation was active, in the same experiment we showed that lactoferrin inhibited the apoE/P-VLDL-mediated stimulation of cholesteryl ["C] oleate synthesis (Fig. 10B, inset).

To further explore the receptor specificity of '251-labeled lactoferrin metabolism in fibroblasts, we incubated the cells with increasing amounts of lZ5I-labeled lactoferrin in the absence and presence of an excess of GST-39 kDa protein, and measured the degradation (Fig. 1lA). lz5I-Labeled lactoferrin was degraded by a relatively high affinity process that was inhibited by about 50% by GST-39 kDa protein. Simi- larly, when different concentrations of GST-39 kDa protein were tested, we observed that the GST-39 kDa protein reduced the degradation of Iz5I-labeled lactoferrin by 60% at a concentration of 3 pg/ml (Fig. 11B). Increasing the concentration of the GST-39 kDa protein up to 60 pg/ml did not produce any further inhibition. These results suggest that LRP accounts for about one-half of the uptake of lactoferrin in fibroblasts. The LRP-independent process is not inhibited by GST-39 kDa, even a t high concentrations.

DISCUSSION

The current data illustrate the unusual complexity in the ligand binding specificities of the members of the LDL receptor gene family. The multiple ligands that bind to each receptor appear to attach to distinct, but partially overlapping binding sites. This pattern is likely attributable to the binding of different ligands to different combinations of cysteine-rich repeats. A precedent for this type of combinatorial binding was established earlier in studies of the LDL receptor (35).

The LDL receptor binds two structurally unrelated ligands: apoB-100, which contains 4536 amino acids and occurs in a single copy in each lipoprotein particle, and apoE, which contains 299 amino acids and occurs in multiple copies in each lipoprotein particle. Both ligands bind to the region of the LDL receptor that contains the seven complement-type repeats. Mutagenesis studies revealed that binding of apoB- 100-containing lipoproteins was markedly reduced upon deletion or alteration of any one of the complement-type ligand binding repeats except the first repeat (35). This suggested


TABLE I Inhibition by chloroquine of apoE/P- VLDL and lipoprotein lipase/& VLDL stimulated

choksteryl ['4CJokate formation in FH fibroblasts lacking LDL receptors On day 7, each monolayer received 1.5 ml of medium A in the absence or presence of 75 pM chloroquine as indicated. After 15 min at 37 "C,

each dish received 0.5 ml of medium A containing the indicated addition. After incubation for 5 h at 37 "C, the cells were pulse-labeled for 2 h at 37 "C with 0.2 mM ['4C]oleate, and their content of cholesteryl ['4C]oleate and ['4C]triglycerides was measured. Each value represents the average of duplicate incubations. No blank value was subtracted from the data.

Addition Cholesteryl ["Cloleate formed ["C]Triglycerides formed to medium -Chloroquine +Chloroquine -Chloroquine +Chloroquine

nmol. h" . mg protein" nmol. h" . mg protein"

p-VLDL, 30 pg/ml 0.19 0.03 20.6 14.7 P-VLDL + apoE, 40 pg/ml 7.4 0.04 20.7 20.2 LPL, 8 pg/ml 0.21 0.03 20.6 19.6 @-VLDL + LPL 4.7 0.03 23.0 22.8

1 Y Br A t t

z Bovlrw, Lsdofenln (pg/ml)

FIG. 8. Inhibition by bovine lactoferrin of apoE/@-VLDL and lipoprotein lipase/fl-VLDL-stimulated cholesteryl ['"C] oleate formation in FH fibroblasts lacking LDL receptors. Panel A , each monolayer received 1.5 ml of medium A containing the indicated concentration of bovine lactoferrin. After incubation for 15 min at 37 "C, each dish received 0.5 ml of medium containing 8- VLDL at a final concentration of 40 pg of protein/ml that had been preincubated with either apoE at a final concentration of 40 pg/ml (0) or lipoprotein (Lp) lipase at 9 pg/ml (0). Panel B, incubations were similar to those in panel A except that each dish received either a mixture of sterols containing 25-hydroxycholesterol of a final concentration of 5 pg/ml and cholesterol at 10 pg/ml or 0-VLDL at a final concentration of 40 pg of protein/ml that had been preincubated with apoE at 50 pg/ml. After incubation for 5 h at 37 "C, the cells were pulse-labeled for 2 h at 37 "C with 0.2 mM ["Cloleate, and their content of cholesteryi ["Cloleate was measured. Blank values of 0.05 and 0.09 nmol- h".mg protein" for panets A and 8, respectively, corresponding to the amount of cholesteryl [14C]oleate formed in parallel dishes receiving no additions were subtracted from each value. The 100% of control values in panel A were 4.4 (0) and 3.6 (0) nmol. h" . mg protein" and, in panel B, 1.1 (0) and 2.8 (A) nmol. h" . mg protein". Each value is the average of duplicate or triplicate incubations. Lp Lipase, lipoprotein lipase.

that the large apoB-100 particle binds simultaneously to multiple repeats. In contrast, the binding of apoE was not abolished by the destruction or functional elimination of any single repeat. The greatest reduction in binding (60%) was observed when the fifth repeat was eliminated (35). This finding suggested that apoE can bind to any one of several different repeats, although it may prefer the fifth repeat.

LRP contains 31 ligand binding repeats and 22 growth factor repeats (1, 5 ) , and therefore it has the potential for even more complex patterns of binding. On ligand blots, LRP is now known to bind at least seven macromolecules. ApoE/ P-VLDL, LPL, tPA/PAI-1, lactoferrin, and the 39-kDa protein were studied in the current paper. In addition, LRP in intact fibroblasts binds ap-macroglobulin' (13) and Pseudom- onas exotoxin A (36). All of the ligands that bind to LRP on ligand blots have been shown in the current study or in previous studies (6, 13, 20) to be taken up and degraded in lysosomes in an LRP-dependent fashion in cultured fibroblasts. The uptake of each of these molecules (except for

Pseudomonas exotoxin A which has not yet been studied) is reduced by prior incubation of cells with a polyclonal anti- LRP antibody that depletes the cell surface of LRP. All of the agents that block ligand binding to LRP on ligand blots also inhibit the delivery of these ligands to lysosomes in fibroblasts. We therefore believe that all of the ligands that bind to LRP on ligand blots are functionally cleared from the extracellular fluid by LRP-mediated endocytosis.

Only one potential ligand for gp330 has been previously identified. Kanalas and Makker (37) reported the presence of a plasminogen binding site on rat gp330 in ligand blot and enzyme-linked immunoabsorbent assays. Using the nitrocellulose binding assay, in the current study we show that gp330 binds many of the same ligands that bind to LRP, including 45Ca2+, 1251-labeled tPA. PAI-1 complexes, apoE. P-VLDL complexes, '251-LPL, and '251-labeled lactoferrin. Binding of all ligands was inhibited by the GST-39 kDa fusion protein, but not GST, indicating that the ligands bound to corresponding sites on gp330 and LRP. The current ligand blots were performed with crude kidney membrane preparations and it is theoretically possible that some of the ligands bind to other proteins that co-migrate with gp330 following SDS electrophoresis. We believe this to be unlikely because all of the ligands bound to an extremely large (>515 kDa) and abundant protein (visible on stained SDS gels of crude membranes) that showed an identical pattern of migration irrespective of the ligand used to visualize it. gp330 is the only known protein of this size and abundance in the kidney (see lane 3 in Fig. 1). Moreover, none of the ligands bound to any other protein on the blots, indicating the specificity of the binding to gp330. For unknown reasons, neither LRP nor gp330 binds 1251-cy2- macroglobulin* reproducibly on ligand blots, even though LRP clearly binds '251-cy2-macroglobulin' in intact fibroblasts and in other assay systems (13, 38). We therefore cannot determine from these studies whether gp330 can bind to a2-macroglobulin*.

Available evidence suggests that LRP must contain multiple independent binding sites. One of these sites binds apoE/ P-VLDL, and another binds a2-macroglobulin'. There is no clear evidence for high affinity cross-competition between these ligands (39, 40).2 Moreover, lactoferrin, which potently inhibits the cellular uptake of apoE/P-VLDL, does not inhibit the uptake of cy2-macroglobulin', supporting the notion that these two sites are independent. I t is highly likely that LRP contains additional binding sites that are at least partially distinct. For example, the binding of LPLIP-VLDL to LRP is relatively resistant to competition by lactoferrin, whereas apoE/P-VLDL binding is extremely sensitive, suggesting that

* J. L. Goldstein, J. Herz, and M. S. Brown, unpublished observa- tions.


TABLE 11 Degradation of bovine '251-aZ-macroglobulin* (a2"*) and bovine 1251-lactoferrin by FH fibroblnsts lacking LDL receptors

On day 7, each monolayer received 2 ml of Dulbecco's modified Eagle's medium (without glutamine) containing 2 mg/ml bovine serum albumin and either bovine 1251-a~-macroglobulin* (155 cpm/ng) or bovine 1Z51-lactoferrin (182 cpm/ng) in the presence of the indicated concentration of unlabeled protein. After incubation for 5 h at 37 "C, the total amount of 1251-labeled degradation products excreted into the medium was measured.

'2sI-Labeled ligand Unlabeled protein

IZ5I-a2-M*, 10 pg/ml None 1251-~2-M*, 10 pg/ml GST-39 kDa, 30 pg/ml Iz5I-a2-M*, 10 pg/ml Lactoferrin, 50 pg/ml Iz5I-a2-M*, 10 pg/ml Lactoferrin, 100 pg/ml Iz5I-Lactoferrin, 15 pg/ml None Iz5I-Lactoferrin, 15 pg/ml GST-39 kDa, 30 pg/ml Iz6I-Lactoferrin, 15 pg/ml a,"*, 100 pg/ml 1251-Lactoferrin. 15 wdml a2-M*. 300 wdml

Degradation of 12sI-a2M* or '261-lactoferrin g. 5 h" . mg protein"

1.0 0.003 1.2 1.4 2.5 1.2 2.6 2.6

~~~~

GST-39 kDa 0.5

Unlabeled Protein (pg/ml)

FIG. 9. Degradation of human 'Z61-~z-macroglobulin' by FH fibroblasts lacking LDL receptors: inhibition by 39-kDa fusion protein but not by bovine lactoferrin. On day 7, each monolayer received 2 ml of Dulbecco's modified Eagle's medium (without glutamine) containing 2 mg/ml bovine serum albumin and the indicated concentration of either bovine lactoferrin (0) or GST- 39 kDa fusion protein (0). After 15 min at 37 "C, each dish received 15 pg/ml human '251-a~-macroglobulin' (594 cpm/ng). After incubation for 5 h at 37 "C, the total amount of 1251-a2-macroglobulin* degradation products excreted into the medium was measured. Each value represents the average of duplicate or triplicate incubations.

the binding sites for apoE/@-VLDL and LPL/P-VLDL are different. Unfortunately, technical difficulties preclude a di- rect study of competition between apoE/P-VLDL and LPL/ 0-VLDL. It is likely that tPA-PAI-1 complexes bind to still another site, as evidenced by the lack of competition with az- macroglobulin', lactoferrin, apoE/P-VLDL, and LPL/p- VLDL (Fig. 10).

Remarkably, the binding sites for all ligands are occluded by the binding of the 39-kDa protein. This likely means that the 39-kDa protein can bind to all sites, which would require a high multiplicity of 39 kDa binding to LRP. Williams et al. (16) have published data suggesting that a single LRP binds only two molecules of the 39-kDa protein. If this finding can be confirmed, it would suggest that the 39-kDa protein causes a conformational change that excludes binding to all sites.

Several of the ligands that bind to LRP also bind to heparin. These include apoE (41), LPL (18), and tPA - PAI-1 complexes (42). Moreover, the 39-kDa protein was originally purified as a heparin-binding protein (43). Thus each of these proteins must possess a region of concentrated positive charges as does bovine lactoferrin (27, 28). Each of the complement-type ligand binding repeats of LRP has a conserved negatively- charged sequence, SDE (Ser-Asp-Glu), which is also conserved in all of the ligand binding repeats of the LDL receptor

(35). In the latter receptor, the SDE sequences are crucial for binding of apoE and apoB-100 (both of which are heparin- binding proteins). It seems likely that the SDE sequences in LRP are also the binding sites for the positively-charged regions on some of its ligands.

The current data with lactoferrin add additional support to the hypothesis that hepatic LRP participates in the clearance of chylomicron remnants from plasma. Infusion of bovine lactoferrin into rats retarded the clearance of chylomicron remnants into the liver by 50 (27) to 75% (28). In the current studies bovine lactoferrin was shown to be a potent inhibitor of the ability of LRP to recognize apoE/P-VLDL, a surrogate for chylomicron remnants. Moreover, lactoferrin also inhibited, albeit with less potency, the binding of LPL/P-VLDL to LRP. We have previously shown that other inhibitors of hepatic uptake of chylomicrons in uiuo (44), namely the C apoproteins, also block binding of apoE/P-VLDL to LRP (7, 45). Thus, the two agents that are known to inhibit chylomicron remnant uptake in animals (lactoferrin and the C apoproteins) also inhibit binding to LRP.

Beisiegel et al. (17) showed previously that LPL/P-VLDL complexes bind to LRP as determined by chemical cross- linking. We have extended these findings in the current studies by showing that LPL. p-VLDL complexes deliver cholesterol to cells and thereby stimulate cholesteryl ester synthesis. This delivery is abolished by the GST-39 kDa protein, suggesting that it is mediated by LRP. It is also reduced, but not abolished, when the cells are incubated with a polyclonal antibody to LRP. In previous studies we have shown that this antibody also reduces the binding of az-macroglobulin', but also only partially (13). We believe that these partial effects indicate that the antibody only partially depletes the cell surface of LRP (7). We cannot exclude the alternative possi- bility, namely, that cells have an additional receptor that resembles LRP and is inhibited by the 39-kDa protein, but is not recognized by the polyclonal antibody. This explanation seems unlikely because we have found no evidence for a second apoE/P-VLDL binding protein in ligand blots from membranes of these fibroblasts. We also have found no evidence that these fibroblasts express gp330, as indicated by ligand blotting with lZ5I-labeled GST-39 kDa protein and 45Ca2+ or by immunoblotting with a polyclonal antibody raised against purified rat gp330.

The ability of LPL. /3-VLDL complexes to stimulate cholesterol esterification was dependent on lysosomal function, as indicated by chloroquine inhibition (Table I). The stimulation was also abolished by imipramine (Fig. 7), which blocks the export of cholesterol from lysosomes (34). These data


Lp Lipase / P-VLDL

GST- 39 kDa Protein

, I I I, - 0 20 40 60 80 100 " 0 50 100 150 200 Unlabeled Protein (pg/ml) Unlabeled Protein (pg/ml)

FIG. 10. Degradation of '251-labeled tPA-PAI-1 complex by FH fibroblasts lacking LDL receptors: inhibition by 39-kDa fusion protein but not by lactoferrin and other ligands for LRP. On day 7, each monolayer received 2 ml of Dulbecco's modified Eagle's medium (without glutamine) containing 2 mg/ml bovine serum albumin, 1 mM P-mercaptoethanol, and human '251-labeled tPA. PAI- 1 complex (50 ng/ml tPA; 2214 cpm/ng tPA in experiment A and 2952 cpm/ng in experiment B) in the presence of the indicated concentration of one of the following unlabeled proteins: A, GST-39 kDa fusion protein; 0, bovine lactoferrin; A, 8-VLDL that had preincubated for 1 h with lipoprotein lipase a t a protein ratio of 1 pg of P-VLDL/0.33 pg of lipoprotein (Lp) lipase; 0, P-VLDL that had been preincubated for 1 h with apoE at a protein ratio of 1 pg of P-VLDLIl pg of apoE; or ., bovine a2-macroglobulin*. After incubation for 6 h a t 37 "C, the total amount of lZ5I-tPA degradation products excreted into the medium was measured. Each value represents the average of duplicate incubations. The inset in panel B shows the effect of bovine lactoferrin on apoE/P-VLDL-stimulated cholesteryl ['4C]oleate formation in the same cells that were used to determine the effect of lactoferrin on the degradation of lZ5I-tPA. PAI-1 complex shown in panel E. The cells were incubated for 5 h at 37 "C with P-VLDL at a final concentration of 30 pg/ml that had been preincubated with apoE at a final concentration of 30 pg/ ml, followed by a pulse labeling for 2 h with 0.2 mM ['4C]oleate. A blank value of 0.11 nmol. h-' .mg protein-' corresponding to the amount of cholesteryl ['4C]oleate formed in dishes receiving 0-VLDL without apoE was subtracted from each value. Each value represents a single incubation.

I i3 1'

l251-Lamfenin (pg/rnl) Un~abIed Protein (pg/ml)

FIG. 11. Degradation of bovine '261-labeled lactoferrin by FH fibroblasts lacking LDL receptors: inhibition by 39-kDa fusion protein. Panel A , on day 7 each monolayer received 2 ml of Dulbecco's modified Eagle's medium (without glutamine) containing 2 mg/ml bovine serum albumin and either 60 pg/ml GST (0) or 60 pg/ml GST-39 kDa fusion protein (0) as indicated. After 15 min at 37 "C, each dish received the indicated concentration of bovine lZ5I- labeled lactoferrin (222 cpm/ng). Panel B, on day 7 each monolayer received the same medium as above containing 2 mg/ml bovine serum albumin in the presence of the indicated concentration of either GST (A) or GST-39 kDa fusion protein (A). After 15 min a t 37 "C, each dish received 10 pg/ml lZ5I-labeled bovine lactoferrin (211 cpm/ng). For panels A and B, the monolayers were incubated for 5 h at 37 "C, after which the total amount of '251-labeled lactoferrin degradation products excreted into the medium was measured. In panel A each value represents the average of duplicate incubations. The dotted line i n panel A denotes the amount of 39-kDa protein-dependent degradation. In panel B, each value represents the average of triplicate (no unlabeled protein) or single (+unlabeled protein) incubations.

suggest that stimulation of cholesteryl ester synthesis results from the LRP-mediated uptake of LPLIP-VLDL and its delivery to lysosomes.

An important unanswered question relates to the functional distinction, if any, between LRP and gp330. Thus far, our studies of ligand binding on nitrocellulose blots have failed to show any difference in the ligand binding specificities of the two receptors. gp330 differs in its tissue distribution from LRP. It is found on glomerular and tubular epithelial cells in the kidney and in lung epithelial cells and in yolk sac epithe- lium, but notably not in the liver (24-26). In contrast, LRP is expressed on hepatic epithelial cells and on connective tissue cells such as fibroblasts (5 , 6). It is also present on the invading trophoblasts of the mouse blastocyst (46). It is possible that some as yet unknown functional difference exists

between these two proteins that explains their selective expression in different cell types.

Acknowledgments-We thank Wen-Ling Niu, Debra Noble-Mor- gan, and Richard Gibson for excellent technical assistance. Edith Womack and Tracye Martin provided invaluable help in growing cultured cells.

REFERENCES 1. Brown, M. S., Herz, J., Kowal, R. C., and Goldstein, J. L. (1991) Curr.

2. Yamamoto, T., Davis, C. G., Brown, M. S., Schneider, W. J., Casey, M. L., Opin. Lipidology 2,65-72

3. Goldstein, J. L., Brown, M. S., Anderson, R. G. W., Russell, D. W., and Goldstein, J. L., and Russell, D. W. (1984) Cell 39, 27-38

4. Goldstein, J. L., and Brown, M. S. (1989) in The Metabolic Basis oflnherited Schneider, W. J. (1985) Annu. Reu. Cell Bid. 1, 1-39

Disease (Scriver, C. R., Beaudet, A. L., Sly, W. S., and Valle, D., eds) pp. 1215-1250, McGraw-Hill Publishing Co., New York

5. Herz, J., Hamann, U., Rogne, S., Myklehost, O., Gausepohl, H., and Stanley, K. K. (1988) EMBO J. 7,4119-4127

6. Kowal, R. C., Herz, J., Goldstein, J. L., Esser, V., and Brown, M. S. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 5810-5814

7. Kowal, R. C., Herz, J., Weisgraber, K. H., Mahley, R. W., Brown, M. S., and Goldstein, J. L. (1990) J. Biol. Chem. 266, 10771-10779

8. Mahley, R. W., and Innerarity, T. L. (1983) Biochim. Biophys. Acta 737,

9.

10.

11.

12. 13.

14.

15. 16.

17.

18. 19.

20.

21.

22.

23.

24. 25.

Herz, J., Kowal, R. C., Ho, Y. K., Brown, M. S., and Goldstein, J. L. (1990)

Strickland, D. K., Ashcom, J. D., Williams, S., Burgess, W. H., Migliorini,

Kristensen, T., Moestrup, S. K., Gliemann, J., Bendtsen, L., Sand, O., and

Sottrup-Jensen, L. (1989) J. Biol. Chem. 264,11539-11542 Herz, J., Goldstein, J. L., Strickland, D. K., Ho, Y. K., and Brown, M. S.

(1991) J. Biol. Chem. 266, 21232-21238 Strickland, D. K., Ashcom, J. D., Williams, S., Battey, F., Behre, E.,

McTigue, K., Battey, J. F., and Argraves, W. S. (1991) J. Biol. Chem.

Moestrup, S. K., and Gliemann, J. (1991) J. Biol. Chem. 266,14011-14017 266, 13364-13369

Williams, S. E., Ashcom, J. D., Argraves, W. S., and Strickland, D. K.

Beisiegel, U., Weber, W., and Bengtsson-Olivecrona, G . (1991) Proc. Natl.

Bensadoun, A. (1991) Annu. Reu. Nutr. 11, 217-237 Felts, J. M., Itakura, H., and Crane, R. T. (1975) Biochem. Biophys. Res.

Orth, K., Madison, E. L., Gething, M-J., Sambrook, J. F., and Herz, J. Commun. 66,1467-1475

Bu, G., Williams, S., Strickland, D. K., and Schwartz, A. L. (1992) Proc. (1992) Proc. Natl. Acad. Sci. U. S. A. 89,7422-7426

Nykjaer, A,, Petersen, C. M., Moller, B., Jensen, P. A,, Moestrup, S. K., Natl. Acad. Sci. U. S. A. 89, 7427-7431

A., and Gliemann, J. (1992) J. Biol. Chem. 267,14543-14546 Holtet, T. L., Etzerodt, M., Thogersen, H. C., Munch, M., Andreasen, P.

Wing, L. R., Hawksworth, G. M., Bennett, B., and Booth, N. A. (1991) J. Lab. Clin. Med. 117, 109-114

197-222

J. Biol. Chem. 266, 21355-21362

M., and Argraves, W. S. (1990) J. Biol. Chem. 266,17401-17404

Sottrup-Jensen, L. (1990) FEES Lett. 276, 151-155

(1992) J. Biol. Chem. 267,9035-9040

Acad. Sci. U. S. A. 88,8342-8346

Kerjaschki, D., and Farquhar, M. G. (1983) J. Exp. Med. 167,667-686 Chatelet, F., Brianti, E., Ronco, P., Roland, J., and Verroust, P. (1986) Am.

J. Pathol. 122, 512-519

26180 LRP and gp330 Bind Similar Ligands 26. Ra chowdhury, R., Niles, J. L., McCluskey, R. T., and Smith, J. A. (1989) 35. Russell, D. W., Brown, M. S., and Goldstein, J. L. (1989) J. Biol. Chem.

Jcience 244,1163-1166 27. Huettinger, M., Retzek, H., Eder, M., and Goldenberg, H. (1988) Clin. 36. Kounnas, M. Z., Morris, R. E., Thompson, M. R., FitzGerald, D. J.,

264,21682-21688

Binchem. 21,87-92 28. van Dijk, M. C. M., Ziere, G. J., and van Berkel, T. J. C. (1992) Eur. J .

Strickland, D. K., and Saelinger, C. B. (1992) J. Biol. Chem. 267,12420-

Biochem. 206,775-784 12423

29. Vogel, T., Weisgraber, K. H., Zeevi, M. I., Ben-Artzi, H., Levanon, A. Z., 38. Pastan, I. H., and Willingham, M. C. (1981) Science 214,504-509 37. Kanalas, J. J., and Makker, S. P. (1991) J. Biol. Chem. 266,10825-10829

Rall, S. C., Jr., Innerarity, T. L., Hui, D. Y., Taylor, J. M., Kanner, D., 39. Hussain, M. M., Maxfield, F. R., ML-Oliva, J., Tabas, I., Ji, Z-S., Inner- Yavin, Z., Amit, B., Aviv, H., Gorecki, M., and Mahley, R. W. (1985) arity, T. L., and Mahley, R. W. (1991) J. Biol. Chem. 2 6 6 , 13936-13940 Pmc. Natl. Acad. Sci. U. S. A. 82,8696-8700 40. van Dijk, M. C. M., Ziere, G. J., Boers, W., Linthorst, C., Bijsterbosch, M.

Lonblad, P. B., Kristensen, T., Mortensen, S. B., Petersen, T. E., and 41. Mahley, R. W. (1988) Science 240,622-630 Magnusson, S. (1983) Ann. N. Y. Acad. Sci. 421,41-60

00,849-857

J. L. (1987) N. Engl. J. Med. 317,734-737

98,241-260

30. Sottrup-Jensen, L., Stepanik, T. M., Wierzbicki, D. M., Jones, C. M., K., and van Berkel, T. J. C. (1991) Biochem. J. 279,863-870

31. Fraker, P. J., and Speck, J. C., Jr. (1978) Biochem. Biophys. Res. Commun. H. (1991) Biochemistry 30,1021-1028

32. Hobbs, H. H., Brown, M. S., Russell, D. W., Davignon, J., and Goldstein, Biochem. (Tokyo) 108,297-302

33. Goldstein, J. L., Basu, S. K., and Brown, M. S. (1983) Methods Enzymol. 45. Weis aber, K. H., Mahley, R. W., Kowal, R. C., Herz, J., Goldstein, J. L.,

34. Liscum, L., and Dahl, N. K. (1992) J. Lipid Res. 33,1239-1254 46. Herz, J., Clouthier, D. E., and Hammer, R. E. (1992) Cell, in press anyBrown, M. S. (1990) J. Biol. Chem. 266,22453-22459

42. Ehrlich, H. J., Keijer, J., Preissner, K. T., Beggink, R. K., and Pannekoek,

43. Furukawa, T., Ozawa, M., Huang, R-P., and Muramatsu, T. (1990) J.

44. Windler, E., and Havel, R. J. (1985) J. Lipid Res. 26,556-565

low density lipoprotein receptor-related protein and a330 bind

Documents