Monoclonal antibodies to the low density lipoprotein receptor as probes for study of receptor-mediated endocytosis and the genetics of familial hypercholesterolemia.

Monoclonal antibodies directed against the low density lipoprotein (LDL) receptor have been prepared by immunization of mice with a partially purified receptor from bovine adrenal cortex. Spleen cells from the mice were fused with the Sp2/0-Ag14 line of mouse myeloma cells. The most extensively studied monoclonal antibody, designated immunoglobulin-C7, reacts with the human and bovine LDL receptor, but not with receptors from the mouse, rat, Chinese hamster, rabbit, or dog. 125I-labeled monoclonal antibody binds to human fibroblasts in amounts that are equimolar to 125I-LDL. In fibroblasts from 6 of 8 patients with the receptor-negative form of homozygous familial hypercholesterolemia, which have less than 5% of normal LdL binding, the amount of monoclonal antibody binding was also less than 5% of normal. Fibroblasts from the other two receptor-negative homozygotes bound an amount of monoclonal antibody that was much greater than expected on the basis of LDL binding, suggesting that these two patients produce a structurally altered receptor that binds the antibody, but not LDL. In normal fibroblasts, the receptor-bound monoclonal antibody was taken up and degraded at 37 degrees C at rapid rate similar to that for LDL. Fibroblasts from a patient with the internalization defective form of familial hypercholesterolemia bound the monoclonal antibody, but did not internalize or degrade it. The current data demonstrate the usefulness of monoclonal antibodies as probes for the study of the cellular and genetic factors involved in receptor-mediated endocytosis.

Monoclonal antibodies directed against the low density lipoprotein (LDL) receptor have been prepared by immunization of mice with a partially purified receptor from bovine adrenal cortex. Spleen cells from the mice were fused with the Sp2/0-Ag14 line of mouse myeloma cells. The most extensively studied monoclonal antibody, designated immunoglobulin-C7, reacts with the human and bovine LDL receptor, but not with receptors from the mouse, rat, Chinese hamster, rabbit, or dog. 1a51-labeled monoclonal antibody binds to human fibroblasts in amounts that are equimolar to "'I-LDL. In fibroblasts from 6 of 8 patients with the receptor-negative form of homozygous familial hypercholesterolemia, which have less than 5% of normal LDL binding, the amount of monoclonal antibody binding was also less than 5% of normal. Fibroblasts from the other two receptor-negative homozygotes bound an amount of monoclonal antibody that was much greater than expected on the basis of LDL binding, suggesting that these two patients produce a structurally altered receptor that binds the antibody, but not LDL. In normal fibroblasts, the receptor-bound monoclonal antibody was taken up and degraded at 37 "C at a rapid rate similar to that for LDL. Fibroblasts from a patient with the internalization defective form of familial hypercholesterolemia bound the monoclonal antibody, but did not internalize or degrade it. The current data demonstrate the usefulness of monoclonal antibodies as probes for the study of the cellular and genetic factors involved in receptor-mediated endocytosis.
Many of the receptors on the surface of animal cells participate in receptor-mediated endocytosis. Ligands that bind to these receptors are rapidly taken into the cell when the patch of plasma membrane that contains the receptor folds inward and pinches off to form a vesicle. The ligand enclosed in the vesicle is delivered to intracellular sites, frequently to lysosomes, where the ligand is degraded. Receptor-mediated endocytosis has been studied through the use of ligands that are altered so as to permit detection by measurements of radioactivity, fluorescence, or appearance in the electron microscope. Ligands taken up through this route include plasma transport proteins, protein hormones, glycoproteins, toxins, * This research was supported by Research Grant HL-20948 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with I8 U.S.C. Section  viruses, and other plasma proteins. Many of these ligands enter cells because their receptors are clustered in discrete regions of the surface membrane called coated pits, which invaginate into the cell to form coated endocytic vesicles (reviewed in Refs. 1 and 2).
In one system of receptor-mediated endocytosis, namely the one for plasma low density lipoprotein, several naturally occurring mutations involving the structural gene for the receptor have been described in human cells (3). These mutations affect separately the ability of the receptor to bind LDL' and its ability to be internalized after binding. Since the LDL receptor mediates the cellular uptake and degradation of plasma LDL, patients with mutations in the receptor develop high levels of plasma LDL, owing to impaired degradation, and exhibit a clinical syndrome called familial hypercholesterolemia (4).
To gain further insight into the receptor-mediated endocytosis of LDL and the genetics of this process, we have prepared monoclonal antibodies against the LDL receptor. Using technology originally described by Kohler and Milstein (5) and refined by Scharff and co-workers (6) and Kennett et al. ( 7 ) , we have injected a partially purified receptor preparation into mice, fused spleen cells from immunized mice with a line of mouse myeloma cells, and obtained clones of hybrid cells that produce antibodies against the LDL receptor. Whereas monoclonal antibodies have been produced against a variety of cell surface proteins, they have not as yet been used as probes for the study of receptor-mediated endocytosis. The great advantage of a monoclonal antibody, as opposed to the polyclonal mixture of antibodies produced by animals, is its uniform composition. Inasmuch as a monoclonal antibody consists of a single molecular species that reacts with a single antigenic determinant on the receptor molecule, it can be used to assess quantitatively the number of receptors, as well as the movement of a receptor through various membrane compartments in the cell.
In the current paper, we show that a monoclonal antibody binds to the LDL receptor in an amount stoichiometric with that of LDL. The receptor-bound monoclonal antibody is internalized and degraded in lysosomes in a fashion similar to that of LDL. Moreover, mutant cells from patients with FH, which fail to bind or internalize LDL, also fail to bind or internalize the monoclonal antibody. glutamine, 100 units/ml of penicillin, 100 pg/ml of streptomycin, and 15% FCS). All other materials were obtained from sources as previously reported (9, 10).

Monoclonal Antibodies
Immunization-The LDL receptors used for immunization were solubilized from bovine adrenal cortex with octylglucoside, partially purified by DEAE-cellulose and agarose gel column chromatography, and precipitated with acetone in the presence of phosphatidylcholine as described (IO). The final preparation bound -50 pg of '*'I-LDL/ mg of protein under standard assay conditions (10). The liposomes containing the receptor were suspended in buffer containing 20 mM Tris-chloride, 50 m~ NaC1, and 1 mM CaCIP (pH 8) to a final protein concentration-of 50 pg/ml and were mixed with an equal volume of Freund's adjuvant. Each mouse was immunized at intervals of three weeks with a series of three injections of LDL receptors (10 pg of protein for each injection). The fwst two injections were given intraperitoneally and the third was given intravenously into the tail vein. Freund's complete adjuvant was used for the primary immunization, incomplete adjuvant was used for the second injection, and no adjuvant was used for the final immunization. Three days after the third injection, each mouse was killed, and the spleen was removed under sterile conditions. Cell Fusion-Spleen cells from an immunized mouse were fused with myeloma cells according to standard methods (6,7). In brief, the splenic lymphoid cells were collected by perfusing the spleen with medium B (Eagle's minimal essential medium containing 20 mM N-2-4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid at pH 7.4). Contaminating erythrocytes were removed by suspending the cells in icecold 0.17 M NttC1, followed by centrifugation (7). For each fusion, viable splenic lymphoid cells (2 X lo8) and viable myeloma cells (4 X 10') were mixed in 0.5 ml of 30% polyethylene glycol 4000 in medium A adjusted to pH 8 and centrifuged for 6 min (7). (The total exposure time to polyethylene glycol was 8 min.) The cell pellet was diluted with 10 ml of medium B and again subjected to centrifugation.
The pellet was resuspended in HAT medium containing 20% heatinactivated FCS. Aliquots of the diluted cell suspension (200 pl containing lo5 cells) were pipetted into 96-well Falcon plates and incubated at 37 "C in a humidified 5% COP incubator. Wells that contained visible cell clusters were given fresh HAT medium 10 to 15 days after fusion, and two days later the medium was removed and assayed for anti-LDL receptor activity as described below.
Solid-Phase Indirect Binding Assay-In the initial screening step, antibodies to the LDL receptor were detected by a solid-phase indirect binding assay. Aliquots (25 pl) of partially purified bovine adrenal LDL receptors (2.5 pg of protein of the DEAE-cellulose fraction (10) containing sufficient receptor to bind 25 ng of IZ5I-LDL) were plated into 96-well polyvinyl microtiter plates and dried overnight at 37 "C. Just prior to assay, the protein was fmed to the bottom of each well by addition of 100 pl of 50% methanol, followed by two washes with 100 pl of Buffer C (PBS containing 20 mg/ml of bovine serum albumin). Aliquots of medium from each growing culture (50-100 pl) were added to each precoated well and incubated for 2 h at room temperature. Each well was then washed 5 times with 100 pl of Buffer C, after which a saturating amount of '251-labeled goat anti-mouse IgG (25 pg, 1.5 to 3 X lo5 cpm) in 25 pl of PBS was added. After incubation for 2 h at room temperature, each well was washed twice with Buffer C, followed by 3 washes with 100 p1 of PBS. Blank binding values were determined by substituting fresh HAT medium for the culture supernates. All culture supernates that gave values for lZ5I-IgG binding that were at least 2-fold above the blank values (-1000 cpm/well) were subjected to a second screening step.
In the second screening step, the assay was identical to the first assay, except that 40 pg of protein of octylglucoside-solubilized extracts (10) from either normal or receptor-negative FH homozygote fibroblasts (SV-40 transformed) were fixed to each well and used as the solid-phase antigen in place of bovine adrenal LDL receptors. The LDL receptor activity of the soluble extracts from the normal and FH homozygote cells was 1.5 and <0.06 pg of '"I-LDL binding activityhg of protein, respectively. Aliquots of medium from each hybridoma culture that was positive in the fmt screening step were assayed in pairs of wells precoated with normal and FH homozygote fibroblast extracts. Nonspecific binding was -2500 cpm/well. Each culture supernate that gave values for lP51-IgG binding that were at least 2-fold higher in wells coated with normal fibroblast extracts as compared with wells coated with FH homozygote extracts were considered as positive hybridomas and were processed further as described below.
Cloning and Expansion of Hybridomas-Monoclonal hybridoma cell lines were cloned by limiting dilution (7). The C7 clone, which was the source of the anti-receptor antibody used for the majority of the experiments in this paper, was cloned and subcloned. The HI0 clone was cloned once. The cloned cells were grown in suspension in medium A. The cells were then expanded for collection of large volumes of antibody-containing ascites fluid by injection into pristane-primed BALB/C mice (7). A mouse injected with lo7 cells of the C7 clone yielded, on average, -5 ml of ascitic fluid containing 1.5 mg/ ml of IgG (see below).
Purification and Zodination of Monoclonal ZgG-Cells were removed from the ascites fluid by centrifugation (18,000 rpm, 15 min, room temperature). An IgG fraction was isolated from the supernate on columns of Protein A-Sepharose CL-4B that were equilibrated with 0.1 M sodium phosphate (pH 7) (11). The IgG was eluted with a solution containing 1 M acetic acid and 0.1 M glycine (pH 3), after which the pH was raised immediately by addition of 1 M potassium phosphate (dibasic). The IgG fraction was dialyzed against Buffer D (10 m~ sodium phosphate and 50 mM NaCl at pH 7.4) and stored at -70 "C at a concentration of 2-6 mg/ml. The class and subclass of the antibody produced by the C7 clone were determined by Ouchterlony analysis using antisera specific for p chains and for IgG subclasses. The monoclonal antibodies were radiolabeled with lZ5I by a solid-phase iodinating reagent as described by Fraker and Speck (12). A typical iodination reaction contained 200 pg of Iodo-Gen absorbed to a glass scintillation vial to which was added 1 ml of Buffer D containing 3 mg of IgG-C7 and 1 mCi of carrier-free sodium [Iz5I] iodide. After incubation for 20 min at 4 "C, lZ5I-labeled IgG-C7 was separated from free ' ' ' 1 by Sephadex G-25 column chromatography, followed by dialysis against 12 liters of Buffer D. The specific radioactivity averaged 700 cpm/ng (105 cpm/fmol).

Lipoproteins
Human LDL (density, 1.019 to 1.063 g / d ) and lipoprotein-deficient serum (LPDS, density > 1.215 g / d ) were prepared from plasma of individual healthy subjects by ultracentrifugation (IO). Human LDL was radiolabeled with ' ' ' 1 by the iodine monochloride method (10). Rabbit P-VLDL was prepared as previously described (13 For lZ5I-LDL and lZ5I-IgG-C7 binding experiments performed at 4 "C, cell monolayers were placed for 30 min in a 4 "C cold room, after which the medium was removed and replaced with 2 ml of icecold medium B containing 10% human LPDS (14).

Assays
The total cellular binding of lZ5I-LDL to fibroblast monolayers at 4 "C was determined using the standard wash procedure as previously described (14). The total cellular binding of ?-IgG-C7 to fibroblast monolayers at 4 "C was determined exactly as described for lZ5I-LDL binding (14) except that the cells were washed 4 times rapidly, then incubated for 3 min in wash buffer, followed by a rapid final wash. The amounts of surface-bound and internalized lz5I-LDL in fibroblast monolayers at 37 "C were determined by the dextran sulfate release assay (14). The total cellular content (surface-bound + internalized) of lZ5I-LDL and lZ5I-IgG-C7 at 37 "C was determined as described for I-LDL (14) except that the wash procedure for Iz5I-IgG-C7 was modified as described above. The proteolytic degradation of lZ5I-LDL and "'I-IgG-C7 by fibroblast monolayers at 37 "C was determined by measuring the amount of Iz5I-labeled trichloroacetic acid-soluble (noniodide) material formed by the cells and extracted into the culture medium (15). The protein content of lipoproteins, antibodies, and cell extracts was determined by the method of Lowry et al. (16) with bovine serum albumin as a standard. Unless otherwise stated in the legends, each data point represents the average of duplicate incubations.
Localization of Monoclonal Anti-receptor Binding Sites in Fibroblasts by Indirect Immunofluorescence-On Day 7 of cell growth, monolayers of fibroblasts grown on glass coverslips were exposed to 50 pg/ml(340 nM) of anti-receptor IgG-C7 for 1 h at either 4 or 37 "C. The cells were then washed (17) and fixed for 15 min at room temperature in 3% paraformaldehyde in Buffer F (10 m~ sodium phosphate, 0.15 M NaC1, and 2 m~ MgC12 at pH 7.4), after which they were rinsed once with 2 ml of 50 m~ NKCl and twice with Buffer F (Fig. 7, A to D). Coverslips that had been incubated with IgG-C7 at 37 "C ( Fig. 7, C and D) were then treated with 3 ml of Buffer F containing 0.05% Triton X-100 for 10 min at -10 "C. Both sets of coverslips ( A to D) were then placed in a fresh Petri dish (cell side up) covered with 60 pl of goat anti-mouse IgG coupled to fluorescein isothiocyanate (Meloy) at a concentration of 0.5 mg/ml for 1 h at 37 "C. Coverslips that had been incubated with IgG-C7 at 4 "C ( A and B ) were incubated for an additional 1 h at 37 "C with 0.5 mg/ml of rabbit anti-goat IgG coupled to fluorescein isothiocyanate (Cappel Laboratories). Both sets of coverslips (A to D) were washed at room temperature 4 times (15 min/wash) with 2 ml of Buffer F and mounted on glass slides with 90% glycerol in 0.1 M Tris-HC1 at pH 9.4. The coverslips were viewed with a Zeiss photomicroscope 111 equipped with a 100-watt mercury epifluorescence light source and the appropriate fdter package. Photographs were taken on Kodak Tri-X fb at an ASA of 400 using the automatic photometer of the Zeiss instrument and were developed in Microdol X (Kodak) (17).

RESULTS
Preparation of Monoclonal Antibody, IgGX7"Mice were immunized with a partially purified preparation of LDL receptors from bovine adrenal cortex. The spleen cells were fused with the Sp2/0-Ag14 line of mouse myeloma cells. Approximately 8% (126/1680) of the wells produced viable clusters of hybridomas. About 12% of these wells (15/126) produced an antibody that bound to the partially purified LDL receptors from bovine adrenal cortex. Of these 15 positive hybridomas, 5 were positive when further screened for the presence of an antibody that would bind to solubilized extracts of plasma membranes from normal human fibroblasts, but not fibroblasts from a patient with receptor-negative homozygous FH. Antibody-producing clones from these 5 hybridomas were isolated as described under "Experimental Procedures." These clones were then expanded by injection into the peritoneal cavity of mice. Immunoglobulin G was isolated from the ascitic fluid on columns of Sepharose-bound Protein A from Staphylococcus aureus. The IgG was iodinated with lZ5I and used for further studies. The monoclonal antibody used for the current studies is designated IgG-C7. The antibody contains heavy chains of the IgG2b subclass. Fig. 1A shows a saturation curve for the binding of lZ5I-IgG-C7 monolayers of normal human fibroblasts at 4 "C. The curve had two components: a high affinity component that approached saturation at an IgG concentration of -3 n~ (0.45 pg/ml), and a nonspecific component that rose linearly at concentrations of lZ5I-IgG-C7 up to 16 n~. In cells from a patient with receptor-negative homozygous FH, no high affinity binding of the antibody was observed. There was only a linear binding that paralleled the nonspecific component in the normal cells. Previous studies have shown that the number of LDL receptors in normal fibroblasts is markedly reduced when the cells are grown in the presence of a mixture of 25hydroxycholesterol plus cholesterol (4). Fig. 1B shows that the high affinity binding of 1251-IgG-C7 was also abolished by this sterol treatment. Fig. 2 compares the saturation curves for binding of lZ5I-LDL ( Fig. 2A) and IZ5I-IgG-C7 (Fig. 2B) to normal fibroblasts at 4 "C. The binding experiments were performed in the absence or presence of an excess of unlabeled ligand. The difference between binding in the absence and presence of unlabeled ligand is the high affinity binding and it is indicated b y the dashed lines in Fig. 2. Half-maximal binding for lZ5I-LDL to the high affinity receptor occurred at -2 n~ ( Fig. 2A). Half-maximal binding for lZ5I-IgG-C7 occurred at -1 nM (Fig.  2B). At saturation, the cells bound equimolar amounts of lZ5I-LDL and lZ5I-IgG-C7 (-175 fmol/mg of cell protein). Thus, in normal fibroblasts, there is one high affinity monoclonal antibody binding site for each high affinity LDL binding site.
We next performed a series of experiments in which parallel dishes of fibroblasts were incubated at 4 "C with saturating  levels of either lZ5I-LDL or '251-IgG-C7 and the amount of high affinity binding of each ligand at saturation was measured (Fig. 3). One of the control cell strains was studied on 8 different occasions (closed circle with error bars). The maximal binding of '251-IgG-C7 in this strain averaged 221 fmol/ mg of cell protein. The maximal binding of lZ5I-LDL was similar, 258 fmol/mg. In cells from three other control subjects, there was also a close correlation between LDL and antibody binding (0). In cells from 2 patients with heterozygous FH, there was a proportionate reduction in the binding of both ligands (0). Cells from 8 patients with the receptornegative form of homozygous FH showed less than 5% of the normal '251-LDL binding activity. Six of these eight mutant strains also bound less than 5% of the normal amount of lz5I-IgG-C7 (A). However, cells from two of the eight receptornegative patients appeared to be different (indicated by asterisk). In these two subjects, the amount of lZ5I-IgG-C7 binding was much higher than would be expected on the basis of the small amount of 1251-LDL binding. We also studied fibroblasts from seven patients with the receptor-defective form of homozygous FH, whose cells bound 10-25% of the normal amount of *251-LDL (A). In general, these cells also bound 10-25% of the normal amount of '251-IgG-C7. However, there was one receptor-defective patient whose cells appeared to bind significantly more antibody than expected on the basis of the level of lZ5I-LDL binding (indicated by asterisk). It is possible that the individuals whose fibroblasts show inappropriately high antibody binding activity relative to LDL binding activity (indicated by asterisks in Fig. 3) represent patients whose mutations destroy the LDL binding site of the receptor, but leave a residual amount of antigenic activity (see "Discussion"). Fig. 4 shows a time course experiment designed to determine the fate of the receptor-bound 1251-IgG-C7 in intact normal fibroblasts at 37 "C. The cellular content of 1251-IgG-C7 reached a plateau by I h and remained constant over the ensuing 5 h. After a brief lag, degradation products appeared in the medium in the form of '251-labeled acid-soluble noniodide radioactivity. This time course is similar to the time course previously observed for the degradation of receptorbound '251-LDL in fibroblasts (14, 15). By 6 h, -4 times as much lZ5I-IgG-C7 had been degraded as was present in the cells in the steady state. Degradation, but not uptake, of the '251-IgG-C7 was completely blocked by 75 PM chloroquine, an agent that disrupts lysosomal function (18), indicating that degradation occurred within lysosomes (data not shown).  Iz5I-IgG-C7 (25 cpm/fmol) in the absence or presence of IO00 nM of unlabeled IgG-C7 (Experiment B). After incubation for 5 h at 37 "C, the cellular content of Iz5I-IgG-C7 and the amount of '"I-IgG-C7 degraded by the cells and released into the medium were determined. Specific (or high affinity) values were calculated as described in the legend to Fig. 2 /fmol) (A, B ) or '251-lgG-C7 (9 cpm/fmol) (C, D). After incubation for 5 h at 37 "C, the total cellular content of '251-protein (upperpanels) and the amount of "' 1protein degradation products excreted into the medium (lower panels) were determined. The dashed lines show the receptor-dependent values which were determined by subtracting the values in sterol-treated cells from those in untreated cells.

Receptor-mediated Endocytosis of Monoclonal IgG-C7 in Fibroblasts-
As expected from their failure to bind lZ51-IgG-C7 at 4 "C, fibroblasts from a receptor-negative FH homozygote failed to take up significant amounts of lZ51-IgG-C7 at 37 "C or to degrade the antibody with high affinity (Table I, A). Fibroblasts from a patient with the internalization-defective form of FH are known to bind 60 to 80% of the normal amount of lZ5I-LDL at the cell surface, but to internalize and degrade less than 10% of the normal amount (3,19). When these cells were incubated with lZ5I-IgG-C7 at 37 "C, the total cellular content of radioactivity (surface-bound plus internalized) was -20% of normal (Table I, experiment B). The amount of high affinity degradation of the 1251-IgG-C7 was only 6% of normal. These findings are consistent with the notion that the internalization-defective cells bind the IZ5I-IgG-C7 at the receptor site, but fail to internalize or degrade it, just as they do for lZ51-LDL.

experiment
To further probe the relation between metabolism of the lZ5I-IgG-C7 and IZ5I-LDL, we incubated normal fibroblasts with increasing concentrations of either ligand for 5 h at 37 "C, after which the amount of lZ5I-ligand associated with the cells and the amount degraded were measured (Fig. 5). As previously reported (14,15), the saturation curves for lZ51-LDL uptake (Fig. 5A) and IZ5I-LDL degradation (Fig. 5B) were similar. On average, at each lZ5I-LDL concentration, -3-fold more IZ5I-LDL had been degraded than was associated with the cell. When the cells were treated with 25-hydroxycholesterol plus cholesterol, the high affinity uptake and degradation of IZ51-LDL were abolished, and only a nonspecific linear uptake was obtained. The difference between the uptake or degradation of IZ5I-LDL in the absence and presence of sterols is equivalent to the specific receptor-mediated uptake or degradation and is shown by the dashed lines.
The results with lZ51-IgG-C7 were similar to those obtained with lZ5I-LDL (Fig. 5, C and D ) . When the data for the steroltreated cells were subtracted from the data from the untreated cells, the absolute value for high affinity uptake of IZ5I-IgG-C7 at saturation (dashed line) was similar to that for lz5I-LDL (-1.8 pmol/mg of cell protein) (cf Figs. 5A and 5C). Similarly, the amount of degradation at saturation was the same for Iz5I-  normal (A, B) and internalization-defective FH homozygote (C, D) fibroblasts at 4 and 37 "C. On Day 7 of cell growth, each monolayer of normal and mutant fibroblasts received 1.5 ml of ice-cold medium B containing 10% LPDS and 14 nM "'II-IgG-C7 (67 cpm/fmol). After incubation for 2 h at 4 "C, the monolayers were washed extensively as described under "Experimental Procedures." A triplicate set of dishes was then harvested for measurement of cell-bound "'I-IgG-C7 at zero time (a). The remaining dishes received 2 ml of medium B containing 10% LPDS and were incubated at either 4 (0) or 37 "C (0, A, 0). At the indicated time, each monolayer was washed with 1 ml of ice-cold medium B, and the amount of '*'II-IgG-C7 remaining bound to the cells (0,O) was determined as described under "Experimental Procedures." The amount of intact (A) and degraded (0) Iz5I-IgG-C7 released from the cells at 37 "C was measured by precipitating the combined medium and wash in 10% trichloroacetic acid for 1 h at 4 "C and separating the acid-precipitable and acid-soluble fractions by centrifugation.

Monoclonal Antibodies
to the LDL Receptor LDL and '*'I-IgG-C7 (-6.5 pmol/mg) (cf Figs. 5B and 50). Differences in Binding Affinity of Monoclonal IgG-C7 a t 4 a n d 37 "C-Previous studies have shown that the affinity of the fibroblast LDL receptor for LDL is -10-fold lower at 37 "C, as compared with 4 "C (14). A comparison of the 37 "C uptake data for I2'I-LDL in Fig. 5 with the 4 "C binding data of Fig. 2 shows this 10-fold decrease in affinity at 37 "C, as compared with 4 "C. An even more striking difference was observed in the apparent affinity of the "'I-IgG-C7 for the receptor at the two temperatures. Half-maximal binding at 4iC was achieved at a "'I-IgG-C7 concentration of -1 nM (Fig. l ) , whereas a concentration of 75 nM was required for half-maximal uptake at 37 "C ( Fig. 7 C and D), suggesting that the affinity of the receptor for the antibody was -75-fold lower at 37 "C, as opposed to 4 "C.
One reason for the lower affinity of the antibody for the receptor at 37 "C was the rapid dissociation of the receptorantibody complex at the higher temperature. Fig. 6 shows an experiment in which fibroblasts were incubated at 4 "C with "'II-IgG-C7, washed, and then incubated further at either 4 or 37 "C. When the normal cells were incubated at 4 "C, the dissociation of the "'II-IgG-C7 from the receptor was extremely slow (Fig. 7A). No detectable dissociation occurred over the first 60 min. At 37 "C, there was a biphasic loss of I2'I-IgG-C7 from the cells. About one-third of the cell-bound radioactivity left the normal cell within the first 10 min and appeared in the medium in a form that was precipitable by trichloroacetic acid (Fig. 6B). Over the ensuing 2 h, there was a slight further loss of acid-precipitable radioactivity. A small amount of acid-soluble radioactivity also appeared in the medium.
To study the dissociation of cell-bound "'I-IgG-C7 in the absence of internalization, we performed the same experiment using fibroblasts from a patient with the internalization-defective form of FH. In these mutant cells, the amount of binding of '251-IgG-C7 at 4 "C was -50% of normal (compare Fig. 6C with Fig. 6A). At 4 "C, the rate of dissociation of the lZsII-IgG-C7 from the receptor was slow, as in the normal cells. However, at 37 "C, most of the radioactivity left the cell within 15 min. This radioactivity appeared in the culture medium entirely as acid-precipitable radioactivity (Fig. 6D). No degradation occurred. Thus, in the internalization-defective FH cells, as well as in the normal cells, the rate of dissociation of the "'I-IgG-C7 from the receptor was much faster at 37 "C than at 4 "C.
Immunofluorescence Staining Pattern of Monoclonal'IgG-C7 in Fibroblasts-Monolayers of fibroblasts were incubated with IgG-C7 for 1 h at 4 "C and then incubated with fluorescein-coupled goat anti-mouse IgG followed by fluoresceincoupled rabbit anti-goat IgG. The surface fluorescence appeared as discrete dots organized in linear arrays (Fig. 7A). This appearance is the same as that seen when the receptor is labeled directly with fluorescent reconstituted LDL (20). Only faint nonspecific binding was seen in cells from a receptor-negative FH homozygote (Fig. 7B). Another group of normal fibroblasts were incubated with the IgG-C7 for 1 h at 37 "C, then fwed and permeabilized with Triton X-100 prior to incubation with the fluorescein-coupled goat anti-mouse IgG (Fig. 7 0 . T h e IgG-C7 was localized within the cell in perinuclear vacuoles that correspond to the general distribution of lysosomes. Receptor-negative FH homozygote cells contained no detectable intracellular fluorescence after incubation with the IgG-C7 (Fig. 70).
Competition between Monoclonal Antibody a n d Lipoproteins for Binding to the LDL Receptor-Unlabeled IgG-C7 competed with "'I-LDL for binding to the LDL receptor of intact fibroblasts at 4 "C (Fig. 8). Binding of "'I-LDL was reduced by 70% at an IgG-C7 concentration of 2 nM, which is in the same range as the concentration required for saturation of "'II-IgG-C7 binding at 4 "C (Fig. 2). Complete inhibition of "'I-LDL binding was not achieved, even when the IgG-C7 concentration was increased to 670 n~. A t this concentration,

Effect ofprio: incubation with unlabeled lipoproteins and IgG preparations in preventing the binding of '251-labeled monoclonal
anti-receptor IgG-C7 to human fibroblasts On Day 7 of cell growth, each monolayer of normal fibroblasts received 1.5 ml of ice-cold medium B containing 10% LPDS and the indicated addition at the indicated concentration. After a prior incubation for 1 h at 4 "C, 14 nM of 1251-IgG-C7 (62 cpm/fmol) was added to each dish. The cells were incubated further for 2 h at 4 "C, after which the total amount of 1251-IgG-C7 bound to the cells was determined. a small amount of lz5I-LDL binding persisted. This residual binding was specific in that it could be competitively reduced by an excess of unlabeled LDL ( X on the ordinate of Fig. 8).
In other experiments, we found that lz5I-LDL binding was not reduced by a control mouse monoclonal antibody not directed against the LDL receptor.
Although the IgG-C7 competitively reduced the binding of I-LDL to the cells at 4 "C, the converse was not true. As shown in Table 11, LDL did not reduce the binding of lZ5I-IgG-C7, even when the cells were incubated with the LDL at supersaturating levels for 1 h prior to addition of the lZ5Ilabeled monoclonal antibody. We also tested two ligands that have extremely high affinities for the LDL receptor, rabbit P-VLDL ( K d = -0.4 m), and canine apo E-HDL, (Kd = -0.1 nM). Neither of these ligands reduced the binding of '251-IgG-C7, even when added at concentrations 12 to 100-fold above their respective Kds. To ensure that the lZ5I-IgG-C7 binding observed in this experiment was specific, we showed that the binding was reduced by an excess of unlabeled IgG-C7. Binding of lZ5I-IgG-C7 also was reduced by a monoclonal antibody derived from another clone, IgG-H10. Binding was not reduced by human IgG. In other experiments, we have found that the binding of lZ5I-IgG-C7 is not reduced by IgG isolated from the serum of nonimmunized mice or by a control mouse monoclonal antibody not directed against the LDL receptor (data not shown).

DISCUSSION
This paper describes the preparation of a monoclonal antibody against the LDL receptor and its use in the study of the genetics of receptor-mediated endocytosis of LDL. The first conclusion of these studies is that the monoclonal antibody, designated IgG-C7, binds to human fibroblasts in amounts that are equimolar to the binding of LDL. This suggests that each LDL receptor has one antibody binding site/LDL binding site. When the receptors are suppressed, as when the cells are treated with 25-hydroxycholesterol plus cholesterol, the binding of LDL and IgG-C7 are reduced proportionately. Moreover, when the receptor number is reduced by mutation, as occurs in FH heterozygotes and homozygotes, the binding of IgG-C7 is reduced proportionately (with certain notable exceptions, as discussed below).
After binding to the receptor, monoclonal IgG-C7 was rapidly internalized at 37 "C by the cells and degraded in lysosomes in a manner similar to the receptor-mediated uptake of LDL. An unexpected finding was the observation that the monoclonal antibody apparently can be detached from the receptor within the cell, allowing the receptor to undergo its usual recycling process in which it returns to the surface and binds another molecule of antibody. This conclusion follows from the kinetic data showing that the cells continue to bind, internalize, and degrade the IgG-C7 at a steady rate without any depletion of receptors for at least 6 h at 37 "C (Fig. 4). Continuous uptake of LDL in this fashion has been shown to be due to recycling of the LDL receptor to the surface after its dissociation from LDL within the cell (17,21).
The ability of the LDL receptor to recycle in the presence of the monoclonal antibody may be dependent on its rapid dissociation from the receptor at 37 "C (Fig. 6 ) . If this dissociation occurs within the cell, it would liberate the receptor so that it could return to the surface and bind another molecule of antibody. Previously we have studied a polyclonal rabbit antibody to the LDL receptor (9). This antibody does not readily dissociate from the receptor at 37 'C.' Since dissociation of the rabbit antibody does not occur, the receptor cannot return to the surface. When fibroblasts are incubated with Unpublished observations. this polyclonal antibody, the receptors on the surface immediately decline and remain low for several hours. Under these conditions, the receptors can be shown by indirect immunofluorescence to be localized within the cell in an intact form As a result of the rapid dissociation of the monoclonal antibody from the LDL receptor at 37 "C, its apparent affinity for the receptor is -75-fold lower at 37 "C than at 4 "C. The affinity of the receptor for LDL is also -10-fold lower at 37 "C, as opposed to 4 "C, and this may also be due, at least in part, to a more rapid dissociation at the higher temperature (19). In the case of LDL, the dissociation rate is further increased when the pH is lowered below 6 (22). The ability of the ligand (either LDL or IgG-C7) to dissociate from €he receptor is crucial if recycling of the receptor is to occur.
Cells from a patient with the internalization-defective form of FH bound the monoclonal antibody but failed to degrade significant amounts of it. This failure of degradation is consistent with a failure to internalize the ligand. Also consistent with this interpretation was the finding that the amount of '251-IgG-C7 bound to the surface of the internalization-defective cells at 4 "c was 50% of normal ( Fig. 6C), whereas at 37 OC, the total cellular content of *251-IgG-C7 (surface-bound plus internalized) was only 20% of normal (Table I). The greater degree of abnormality at 37 "C is presumably due to a lack of internalization. The rapid internalization of the monoclonal antibody in normal cells and the apparent lack of internalization in the internalization-defective FH cells strongly suggests that the IgG-C7 enters the cell through the normal LDL uptake mechanism, and not by inducing an abnormal internalization of the receptor. Studies with ferritinlabeled IgG-C7 are currently underway to c o d i that this rapid internalization takes place through coated pits.
Although the LDL receptor has one IgG-C7 binding site for each LDL binding site, the two sites do not appear to be identical, as indicated by the following observations. Under conditions in which the receptor was fully occupied by prior incubation with LDL or with the higher affinity ligands p-VLDL and apo E-HDL,, the Iz5I-IgG-C7 was still able to bind in normal amounts. Moreover, when lZ5I-LDL was pre-bound to the receptor at 4 "C, the addition of a saturating amount of IgG-C7 did not cause the LDL to dissociate at 4 "C (data not shown). On the other hand, when the cells were first allowed to bind the IgG-C7 at 4 "C, the subsequent binding of lZ5I-LDL was reduced by -80% (Fig. 8). Thus, although the IgG-C7 does not bind directly to the LDL binding site, the binding of the antibody may cause a structural change that occludes the LDL binding site. This could involve cross-linking of adjacent receptor molecules, a conformational change in the receptor, or simple steric hindrance by the antibody. It is of interest that these changes reduce the binding of Iz5I-LDL at 4 "C by a maximum of -80%, but not by 100% (Fig. 8). At 37 "C, the monoclonal antibody completely blocks Iz5I-LDL binding, uptake, and degradation (Fig. 9). Half-maximal inhibition is achieved at an IgG-C7 concentration of -15 nM, a concentration at which only -20% of the receptors are occupied with antibody (Fig. 5). The complex stoichiometry of the interaction between the antibody binding site and the LDL binding site may be related to the possible organization of the LDL receptor in clusters of 4 binding sites (23, 24). Such an organization has been postulated on the basis of kinetic studies showing that 1 molecule of apoE-HDL, can bind to 4 LDL binding sites (23,24). Further studies with monoclonal antibodies should allow testing of this hypothesis.
One of the major uses of the monoclonal antibody will be as a probe for study of possible mutations in the structural gene for the LDL receptor in subjects with FH. In most of the (21): mutant cell strains so far studied, the observed reduction in LDL binding sites is associated with a parallel reduction in binding sites for lZ5I-IgG-C7. This reduction in antibody binding suggests one of two possibilities: 1) the absolute amount of receptor protein on the cell surface is drastically reduced, or 2) the receptor protein is present in near normal amounts, but is altered so as to lose its antigenic site, as well as its LDL binding site. We should be able to distinguish between these two possibilities by obtaining a variety of monoclonal antibodies that recognize different antigenic determinants (epitopes) on the receptor protein. Absence of multiple antigenic determinants would suggest an absence of the entire protein.
Fibroblasts from three FH homozygotes appear to possess at least one gene that produces a receptor that binds much more antibody than it does LDL. Two of these subjects are classified as receptor-negative with no significant LDL binding activity and one is receptor-defective with 25% of normal LDL binding activity. The findings in these 3 cell strains have been reproduced several times. Further studies are now in progress to determine whether these cells possess a structurally altered receptor with loss of the LDL binding site and preservation of the antibody binding site.
The monoclonal antibody used in the current studies was produced by immunizing a mouse with an LDL receptor partially purified from cows. We selected a clone of cells that produced an antibody that reacted with human fibroblasts. It is of interest that this antibody does not bind to the LDL receptor of cultured mouse L cells, mouse adrenal Y-1 cells, or Chinese hamster ovary cells (data not shown), even though these cells contain a receptor that binds human LDL. In addition, the IgG-C7 blocks binding of '251-LDL to bovine adrenal membranes in vitro, but not to membranes from the adrenals of the mouse, rat, dog, or rabbit (data not shown). Thus, the antibody recognizes a determinant that is present on the LDL receptors of humans and cows, but not rats, mice, hamsters, dogs, and rabbits. The nature of this determinant is unknown, but its existence allows a clear differentiation between the human and bovine LDL receptors on the one hand and the receptors from the rodent and canine species on the other. The monoclonal antibody, therefore, will be useful for somatic cell genetic studies designed to determine whether a cultured rodent cell has acquired the human gene for the LDL receptor.