Epidermal growth factor and platelet-derived growth factor-BB induce a stable increase in the activity of low density lipoprotein receptor-related protein in vascular smooth muscle cells by altering receptor distribution and recycling.

Low density lipoprotein receptor-related protein (LRP) is a multifunctional receptor, expressed by vascular smooth muscle cells (VSMCs) in normal arteries and in atherosclerotic lesions. In this investigation, we demonstrate a novel mechanism for the regulation of LRP activity in cultured rat aortic VSMCs. Cells that were treated with platelet-derived growth factor-BB (PDGF-BB) or epidermal growth factor (EGF) for 24 h bound increased amounts of the LRP ligand, activated α2-macroglobulin (α2M), at 4°C. The Bmax for activated α2M was increased from 56 ± 5 to 178 ± 24 and 143 ± 11 fmol/mg cell protein by PDGF-BB and EGF, respectively, while the KD was unchanged. Northern and Western blot analyses demonstrated that neither PDGF-BB nor EGF increase LRP mRNA or protein levels. Instead, LRP was redistributed to the cell surface and remained localized primarily in coated pits, as determined by surface protein biotinylation, affinity labeling, and immunoelectron microscopy studies. The increase in cell-surface LRP was partially explained by a 50% decrease in receptor endocytosis rate; however, at 37°C, PDGF-BB- and EGF-treated VSMCs still bound/internalized increased amounts of activated α2M and subsequently released increased amounts of trichloroacetic acid-soluble radioactivity. The cytokine-induced shifts in LRP subcellular distribution were stable when VSMCs were challenged with a saturating concentration of ligand and then incubated, in the absence of cytokine, for 2.5 h at 37°C. Regulation of LRP distribution and activity may be an important aspect of the VSMC response to the atherogenic cytokines, PDGF-BB and EGF.

Early studies showed that the ␣ 2 M receptor, which is identical to LRP (7), is recycled and reutilized (8,9). Activated ␣ 2 M-LRP complexes undergo endocytosis and then dissociate in intracytoplasmic vesicles, as a result of acidification (10,11). The receptor recycles to the cell surface while the ␣ 2 M is transferred to lysosomes. The sequence of LRP includes two copies of the NPXY motif, which functions as a signal for clustering and endocytosis of receptors in coated pits (1,5). A monoclonal antibody which binds to the LRP heavy chain has been shown to follow the same cellular processing pathway as was originally described for activated ␣ 2 M (12).
LRP is expressed by numerous cell types, including hepatocytes, macrophages, fibroblasts, neurons, and vascular smooth muscle cells (VSMCs) (1,2). Unlike the LDL receptor, LRP is not down-regulated by cholesterol in macrophages and fibroblasts (13). Expression of LRP in macrophages is regulated by specific cytokines, including interferon-␥ and colony-stimulating factor-1 (14 -16). These cytokines apparently alter LRP gene transcription (16,17). Less is known about LRP regulation in other cell types.
Immunohistochemistry studies have shown that LRP is expressed by VSMCs in neointimal atherosclerotic lesions (18,19). Cellular uptake of ␤-very low density lipoprotein by LRP may promote VSMC foam cell formation. Internalization of plasminogen activators and the urokinase receptor, uPAR, by LRP may alter VSMC migration, a process that is critical in atheroma formation (20 -22). Finally, clearance of activated ␣ 2 M, together with ␣ 2 M-associated growth factors, may regulate VSMC and macrophage gene expression (23,24).
VSMCs in atherosclerotic lesions synthesize less contractile protein, more extracellular matrix protein, and demonstrate morphologic changes referred to as the synthetic phenotype (25,26). VSMCs also acquire properties of the synthetic phenotype in cell culture and this transition is promoted by platelet-derived growth factor-BB (PDGF-BB) (26,27). When ex-posed to ␤-very low density lipoprotein, an LRP ligand, cultured VSMCs that have adopted the synthetic phenotype show a substantial increase in cholesteryl ester accumulation and lipid droplet formation (28). LDL uptake is not increased in these cells.
In this investigation, we demonstrate that the atherogenic cytokines, PDGF-BB and epidermal growth factor (EGF), increase LRP activity in cultured rat VSMCs; however, the mechanism does not involve regulation of LRP expression. Instead, EGF and PDGF-BB alter the subcellular distribution and recycling of LRP, such that the activity of the receptor is enhanced. Unlike previous examples of transient regulation of receptor recycling by cytokines (29 -31), the changes in LRP activity caused by EGF and PDGF-BB are stable. Sustained increases in LRP activity may be an important characteristic of the VSMC synthetic phenotype.

MATERIALS AND METHODS
Reagents-␣ 2 M was purified from human plasma according to the method of Imber and Pizzo (32). ␣ 2 M-MA, a form of activated ␣ 2 M, was prepared by dialyzing native ␣ 2 M against 200 mM methylamine, 50 mM Tris-HCl, pH 8.2, for 8 h at 22°C and then exhaustively against 20 mM sodium phosphate, 150 mM NaCl, pH 7.4 (phosphate-buffered saline), at 4°C. The 142-amino acid C-terminal receptor binding domain of rat ␣ 1 M, the rat homologue of human ␣ 2 M, was prepared as a recombinant polypeptide in Escherichia coli as described by Weaver et al. (23). This polypeptide (rRBD) completely inhibits specific binding of activated ␣ 2 M to rat VSMCs; the K I is 25 nM. Receptor-associated protein (RAP) is a 39-kDa ligand for LRP which blocks the binding of activated ␣ 2 M and other ligands to LRP (33). RAP was prepared as a glutathione S-transferase fusion protein, as described previously (34). The glutathione S-transferase-RAP expression construct was kindly provided by Dr. Joachim Herz (Southwestern Medical Center, Dallas). The cDNA probe for rat LRP was kindly provided by Dr. Guojun Bu (Washington University, St. Louis, MO). Polyclonal antibody R777, which readily detects rat LRP (35), was kindly provided by Dr. Dudley Strickland (American Red Cross, Rockville, MD).
Radioiodination-␣ 2 M-MA was radioiodinated using Iodo-Beads as described by the manufacturer (Pierce). Specific activities were 0.8 -2.2 Ci/g. The rRBD was radioiodinated by the lactoperoxidase/Enzymobead method, using 2 mCi of Na 125 I, according to the manufacturer's instructions (Bio-Rad). The specific activity was 30 Ci/g. VSMC Culture-VSMCs were isolated from Sprague-Dawley and spontaneously hypertensive rat aortas by enzymatic digestion (36). The cells were cultured in a 1:1 mixture of Dulbecco's modified Eagle's medium and Ham's F-12 medium, supplemented with 0.68 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 10% fetal bovine serum (HyClone Laboratories) and passaged when subconfluent with trypsin/EDTA (Life Technologies, Inc.). All cultures were maintained in a humidified 37°C incubator in 5% CO 2 , 95% air.
VSMCs that had been passaged 5-15 times were used for experiments. To induce quiescence, confluent cultures were incubated for 4 days in defined serum-free medium, which consisted of Dulbecco's modified Eagle's medium/F-12 with 0.5 M insulin, 5 g/ml transferrin, 0.2 mM ascorbate, 38 nM selenium, 0.68 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. This medium has been shown to maintain VSMCs in a quiescent state and to promote expression of VSMC contractile proteins (37).  in a ␥-counter. Cellular protein was determined by the bicinchonic acid assay (Sigma). Specific binding was calculated by subtracting nonspecific binding, detected in the presence of non-radiolabeled ␣ 2 M-MA, from total binding. Specific binding isotherms were fit to the equation for a rectangular hyperbola by nonlinear regression. The same data were also analyzed by Scatchard transformation.
Northern Blot Analyses-Equal amounts of total VSMC RNA (20 g) were denatured with glyoxal and subjected to electrophoresis in 1.0% agarose gels. After electrotransfer to Zeta-probe membranes (Bio-Rad), the RNA was cross-linked to the membrane by ultraviolet irradiation and hybridized with cDNA probes for LRP, glyceraldehyde-3-phosphate dehydrogenase, ␣-tubulin, and ␤-actin, each of which was labeled with [␣-32 P]dCTP using the Random Primers Labeling System (Life Technologies, Inc.). The membranes were washed twice with 1 ϫ SSPE, 1.0% SDS (w/v) at 42°C, and once with 0.1 ϫ SSPE, 1.0% SDS (w/v) at 65°C. Specific hybridization of cDNA probes was quantitated by Phosphor-Imager analysis (Molecular Dynamics).
Affinity Labeling with rRBD-rRBD binds to LRP and inhibits the binding of activated ␣ 2 M (38,39). rRBD is also comparable to activated ␣ 2 M in its ability to induce signal transduction responses which have been attributed to an ␣ 2 M receptor other than LRP (40). Thus, we used 125 I-rRBD as an affinity-labeling probe for ␣ 2 M receptors (the intact 718-kDa form of activated ␣ 2 M was too large to function as a probe). PDGF-BB-treated, EGF-treated, and control VSMCs were incubated with 125 I-rRBD (50 nM) for 4 h at 4°C. After washing the cultures, BS 3 (1 mM) was added for 20 min at 4°C. The VSMCs were washed again and harvested by solubilization in 1.0% Nonidet P-40, 0.25% sodium deoxycholate (with proteinase inhibitors). Solubilized proteins were subjected to SDS-PAGE on 4 -12% gels. BS 3 -stabilized complexes of 125 I-rRBD with cellular proteins were detected by PhosphorImager analysis.
Biotinylation of Cell-surface LRP-Cell-surface proteins in PDGF-BB-treated, EGF-treated, and control VSMCs were biotinylated by two sequential incubations with 0.5 mg/ml Sulfo-NHS-biotin for 15 min at 4°C. The cells were washed twice and solubilized in 0.5% Triton X-100 (with proteinase inhibitors). Biotinylated LRP was immunoprecipitated by incubation with R777, followed by Protein A-agarose (Sigma), and then subjected to SDS-PAGE. After electrotransfer to nitrocellulose, the membranes were probed with horseradish peroxidase-streptavidin complex (Pierce). Binding was detected by enhanced chemiluminescence.
Endocytosis Rate Determinations-VSMCs were surface-biotinylated with the reducible cross-linker, NHS-SS-biotin. To initiate receptor endocytosis, cold medium (4°C) was replaced with fresh medium that was pre-warmed to 37°C. Incubations proceeded for various periods of time. To terminate endocytosis, medium was replaced with fresh medium that was pre-chilled to 0°C and the culture dishes were placed in an ice-water slurry. The cells were then washed and incubated with glutathione (15.5 mg/ml) in 75 mM NaCl, pH 8.6, supplemented with 10% fetal bovine serum, to dissociate the biotin from surface-exposed LRP. Glutathione does not penetrate the plasma membrane. Thus, internalized LRP-biotin complexes are protected from reductant and may be selectively detected (41). Free thiol groups were alkylated with 5 mg/ml iodoacetamide. The cells were then solubilized in 0.5% Triton X-100 (with proteinase inhibitors). Samples of each preparation were subjected to immunoprecipitation using antibody R777 and then to SDS-PAGE. The same samples were also subjected to SDS-PAGE without prior immunoprecipitation. LRP-biotin complex was detected, after transfer to nitrocellulose, with streptavidin-horseradish peroxidase and enhanced chemiluminescence. For quantitative results, samples that were not subjected to immunoprecipitation were used, thus allowing more accurate standardization of total cellular protein load per lane. This was possible because the LRP-biotin complex was the largest complex detected and could be easily identified as an isolated band by comparison with the blots of immunoprecipitated LRP. Experiments were performed in quadruplicate. Final autoradiographs were subjected to densitometry and analyzed using ImageQuant software (Molecular Dynamics). In each experiment, cultures were treated with glutathione, without prior warming to 37°C, and then probed for LRPbiotin complex. As expected, LRP-biotin complex was typically unde-tectable in these control samples; however, the intensity of the LRP band was used to establish the "background" for our quantitative analyses.
Uptake and Digestion of ␣ 2 M-MA-VSMCs were incubated with 5.0 nM 125 I-␣ 2 M-MA in defined serum-free medium at 37°C. Some incubations were terminated after 20 min; the cultures were washed and cell-associated radioactivity, which includes cell-surface and internalized ligand, was recovered and quantitated. Other cultures were washed at 20 min and then allowed to incubate in fresh defined serumfree medium for additional periods of time at 37°C. At the specified times, conditioned medium was separated from the cells and cell-associated radioactivity was determined. Conditioned media samples were incubated with 10% trichloroacetic acid on ice for 15 min. Trichloroacetic acid soluble and precipitable radioactivity were separated by centrifugation at 9,000 ϫ g and measured in a ␥-counter. All incubations were conducted in the presence and absence of 1.0 M nonradiolabeled ␣ 2 M-MA. Specific cell associated and trichloroacetic acid soluble radioactivity were determined by subtracting that detected in the presence of non-radiolabeled ␣ 2 M-MA. Specific radioactivity was always greater than 90% of the total.
In some experiments, 125 I-␣ 2 M-MA (5 nM) was incubated with PDGF BB-treated, EGF-treated, and control VSMCs for 4 h at 4°C. The cultures were then washed in Earle's balanced salt solution, 25 mM HEPES, pH 7.4, and warmed to 37°C. At various times, conditioned medium was recovered. Trichloroacetic acid-soluble and -precipitable radioactivity in the medium and cell-associated radioactivity were determined.
Immunoelectron Microscopy Detection of LRP-VSMCs were cultured on 13-mm Thermanox coverslips (NUNC, Inc.) and then equilibrated at 4°C. Antibody R777, diluted 1:20 in culture medium, was incubated with the intact, unpermeabilized cells for 2 h. The cells were then washed and incubated with Protein A-colloidal gold adduct, 150 Å particle size (E-Y Laboratories), for 1 h. After washing the cells again, 2.0% osmium tetroxide, dissolved in phosphate-buffered saline, was diluted 1:2 into the existing medium. Subsequent fixation overnight, at 4°C, was achieved with 2.0% glutaraldehyde in 0.1 M sodium cacodylate, pH 7.4. The coverslips were processed in 70% acetone/water, followed by anhydrous acetone, a 1:1 mixture of anhydrous acetone and epoxy resin for 1 h, and then pure resin for 2 h. The coverslips were embedded in pure resin in BEEM capsules and placed in a 60°C oven overnight for polymerization. Thin sections were made perpendicular to the plane of the cells, transferred to 600-mesh nickel EM grids, and double stained with uranyl acetate and lead citrate. Specimens were examined using a Zeiss 902 electron microscope.

RESULTS
Binding of ␣ 2 M-MA to VSMCs-Quiescent cultures of rat VSMCs bound the LRP ligand, 125 I-␣ 2 M-MA, as described previously (35), and the level of binding was equivalent for cells isolated from spontaneously hypertensive rat or Sprague-Dawley rat aortas (Fig. 1). The K D and B max values were 0.8 Ϯ 0.1 nM and 56 Ϯ 2 fmol/mg cell protein, respectively, for VSMCs from Sprague-Dawley rats and 0.7 Ϯ 0.1 nM and 54 Ϯ 2 fmol/mg cell protein, respectively, for VSMCs from spontaneously hypertensive rats. All subsequent experiments were conducted using Sprague-Dawley rat VSMCs.
Confluent cultures of VSMCs that were maintained in 10% fetal bovine serum demonstrated a slightly increased B max for ␣ 2 M-MA (81 Ϯ 3 fmol/mg cell protein) compared with quiescent cultures. By contrast, quiescent VSMCs that were treated with PDGF-BB (20 ng/ml) or EGF (20 ng/ml) in defined serum-free medium for 24 h bound substantially increased amounts of 125 I-␣ 2 M-MA. The B max values for PDGF-BB-treated VSMCs (178 Ϯ 24 fmol/mg cell protein) and EGF-treated VSMCs (143 Ϯ 11 fmol/mg cell protein) were increased 3.2-and 2.6-fold, respectively, compared with quiescent cultures that were not cytokine-treated. The K D values were unchanged. Since the B max values were standardized based on total cellular protein, the increases did not reflect the growth-promoting activities of PDGF-BB or EGF.
LRP mRNA and Antigen Levels in PDGF BB-and EGFtreated VSMCs-Total cellular RNA was harvested from VSMCs treated with PDGF-BB or EGF for 6, 12, or 24 h and from control cultures that were maintained in vehicle. In four separate experiments, slight increases in LRP mRNA were detected after treating cultures with PDGF-BB or EGF for 12-24 h; however, the mRNAs for glyceraldehyde-3-phosphate dehydrogenase, ␣-tubulin, and ␤-actin were increased as well (results not shown). When LRP mRNA levels were normalized using the other mRNAs, the effects of PDGF-BB and EGF on LRP mRNA expression were not statistically significant.
A Western blot analysis of equal amounts of cellular protein isolated from control, EGF-treated, and PDGF-BB-treated VSMCs is shown in Fig. 2. Antibody R777 detected the 515-kDa LRP heavy chain and the 85-kDa light chain. PDGF-BB and EGF had no effect on the level of LRP antigen detected (two separate experiments). Therefore, the increases in equilibrium binding of ␣ 2 M-MA, caused by PDGF-BB and EGF, were either due to an uncharacterized ␣ 2 M receptor other than LRP or to a shift in the subcellular distribution of LRP such that a greater proportion of the receptor was present on the cell surface.
Expression of Cell-surface LRP-In control VSMCs, the 125 I-rRBD labeled a single polypeptide with an apparent mass equivalent to that of the LRP heavy chain (Fig. 3, panel A). RAP (170 nM) substantially inhibited labeling of the polypeptide, confirming its identity as LRP. PDGF-BB-and EGFtreated VSMCs showed increased labeling of the LRP heavy chain, suggesting that surface expression of LRP was increased in these cells. RAP inhibited 125 I-rRBD labeling of the cytokinetreated cells; the extent of inhibition (as a percent) was equivalent in control and cytokine-treated cultures. Importantly, no novel 125 I-rRBD-polypeptide complexes were identified in the PDGF-BB-or EGF-treated cells, suggesting that increased surface expression of LRP, and not a second ␣ 2 M receptor, was responsible for the increase in equilibrium binding of ␣ 2 M-MA. To confirm this finding, LRP was immunoprecipitated after biotinylation of cell surface proteins at 4°C. As shown in Panel B (representative of four separate experiments), the amount of biotinylated LRP was increased in VSMCs treated with PDGF-BB or EGF.
Localization of Cell-surface LRP-Immunoelectron microscopy was utilized to compare the localization of LRP within the plasma membranes of PDGF-BB-treated, EGF-treated, and control VSMCs. Unpermeabilized cells were analyzed in order to focus on the population of receptors detected in the ␣ 2 M-MA binding experiments. In control VSMCs, LRP was localized almost exclusively to coated pits and membrane invaginations (Fig. 4), consistent with studies of LRP in other cell types and with the known function of LRP as a recycling receptor that undergoes clathrin-dependent endocytosis (10,11). In PDGF-BB and EGF-treated cells, the localization of LRP was unchanged. Thus, the cytokine treatments did not induce a separate pool of receptors in the plasma membrane, but instead altered the steady state distribution of LRP such that more receptors were localized to the cell surface.
Analysis of LRP Function-Cytokine-treated and control VSMCs were incubated with 5.0 nM 125 I-␣ 2 M-MA at 4°C, washed, and then warmed to 37°C. Increased amounts of 125 I-␣ 2 M-MA were detected in association with the PDGF-BB-and EGF-treated cells, compared with control cells, even after incubation at 37°C for 20 min (Fig. 5). Release of digested radioligand, in the form of trichloroacetic acid-soluble peptides, was observed beginning at 60 min and was greatly increased in the cytokine-treated cultures. Thus, the cell-surface receptors which are induced by PDGF-BB and EGF are fully functional and mediate the internalization and delivery of ligand to endosomes for further processing.
To determine whether the cytokine-induced shifts in LRP distribution are stable, EGF-treated, PDGF-BB-treated, and control VSMCs were incubated with nonradiolabeled ␣ 2 M-MA (5 nM) for 4 h at 4°C, washed, and then incubated for 2.5 h at 37°C in medium that was not cytokine supplemented. The cells were then re-chilled to 4°C and binding of 125 I-␣ 2 M-MA (5 nM) was studied. In VSMC cultures that were not cytokine-treated, challenge with non-radiolabeled ligand did not affect equilibrium binding of 125 I-␣ 2 M-MA (1 Ϯ 3% decrease in six studies). In EGF-and PDGF-BB-treated cultures, ligand challenge and cytokine withdrawal had no effect on 125 I-␣ 2 M-MA binding, when compared with identical EGF-and PDGF-BB-treated cells that were not ligand-challenged (2 Ϯ 7 and 9 Ϯ 7% increases in binding for EGF-and PDGF-BB-treated cells, re- spectively, n ϭ 6). Cytokine-pretreated, ligand-challenged cultures also showed unchanged ␣ 2 M-MA binding when compared with cells that were maintained continuously in cytokine-supplemented medium (no ligand challenge). Thus, the cytokineinduced redistribution of LRP to the cell surface is stable, even when cytokine is removed for 2.5 h at 37°C and when the cells are challenged with ligand.
LRP Activity at 37°C-VSMCs were treated with 125 I-␣ 2 M-MA for 20 min at 37°C. The medium was then replaced so that turnover of bound radioligand could be followed. As shown in Fig. 6, EGF-and PDGF-BB-treated cells demonstrated increased binding/uptake of 125 I-␣ 2 M-MA at early time points; however, the same cells also degraded and released 125 I-␣ 2 M-MA at an increased rate. As a result, cell-associated radioligand dropped rapidly; within 90 min, cell-associated radioactivity in the cytokine-treated cells was less than that detected in the control cultures. When 125 I-␣ 2 M-MA was added in the presence of 250 nM RAP, ligand degradation and release (measured as trichloroacetic acid-soluble radioactivity after incubation for 75 min) was decreased by 87 Ϯ 4% in control cultures, 95 Ϯ 2% in EGF-treated cultures, and 93 Ϯ 3% in PDGF-BB-treated cultures (results not shown), confirming that LRP is the receptor responsible for ␣ 2 M-MA catabolism in control and cytokine-treated VSMCs.
Time Dependence of the VSMC Response to EGF and PDGF-BB-Rapid conversion of 125 I-␣ 2 M-MA into trichloroacetic acidsoluble radioactivity at 37°C represented a sensitive index of the change in LRP activity in VSMCs treated with PDGF-BB and EGF. To study the time dependence of the VSMC response, cells were incubated with either of the two cytokines for various periods of time. 125 I-␣ 2 M-MA digestion and release as trichloroacetic acid soluble radioactivity was then determined. The cells were allowed to process ligand for 75 min, following the protocol outlined in Fig. 6. VSMCs that were not cytokinetreated degraded/released 39 Ϯ 6 fmol/mg cell protein (n ϭ 4). Digestion/release was not significantly changed when the VSMCs were treated with EGF or PDGF-BB for 1, 3, 6, or 9 h (results not shown). By contrast, digestion/release was increased to 69 Ϯ 4 and 62 Ϯ 2 fmol/mg cell protein after incubation for 12 h, and to 137 Ϯ 15 and 174 Ϯ 21 fmol/mg cell protein after incubation for 24 h with EGF and PDGF-BB, respectively. Thus, the effects of PDGF-BB and EGF on LRP activity represent a delayed response, requiring more than 12 h to maximize. We cannot rule out the possibility that EGF or PDGF-BB cause an early, transient increase in LRP activity, as has been observed with other recycling receptors (29,31,42,43) and with LRP in insulin-stimulated adipocytes (44). Longer incubation times were studied here to focus on stable changes in LRP activity that may be associated with the ability of the cytokines to modulate VSMC phenotype.
LRP Endocytosis Rates-Changes in the rate of endocytosis can alter surface expression of a receptor. To study LRP endocytosis, cell surface proteins in EGF-treated and control VSMCs were biotinylated with NHS-SS-biotin and warmed to 37°C for 0.5, 1, 2, 4, 6, 10, 16, or 22 min. Internalized, biotinylated LRP was quantitated based on loss of reductant sensitivity. The maximum level of internalized LRP-biotin complex in the EGF-treated cells was 2-3-fold higher than that observed in control cells, reflecting the difference in surface LRP available for biotinylation. The time required for the amount of internalized LRP-biotin complex to reach a maximum was 4 -6 min in the control cultures and about 10 min in the EGFtreated cultures (n ϭ 4). For determining endocytosis rate curves, the time point which yielded the maximum level of internalized LRP-biotin complex was assumed to represent complete transfer of cell surface LRP (LRP S ) into the reductant-insensitive, internalized compartment (LRP I ). The results of four separate experiments were then averaged to generate the graphs shown in Fig. 7. The plots were nearly, but not absolutely linear, as would be expected of a simple first-order process. This result may reflect LRP endocytosis following more complex kinetics or technical limitations involved in measuring very rapid rates of endocytosis, especially after the majority of the receptor has been internalized. The rate of LRP endocytosis was decreased in the EGF-treated cultures compared with the control cultures. For comparison of the curves, linear regression was performed, selecting different time frames for analysis. The curve for LRP endocytosis in control VSMCs yielded rate constants of 0.7, 0.9, and 1.4 min Ϫ1 , re-spectively, when time points from 0 to 4, 0 to 2, or 0 to 1 min were analyzed. By comparison, the curve for LRP endocytosis in EGF-treated cells yielded rate constants of 0.4, 0.5, and 0.6 min Ϫ1 . Irrespective of the time frame selected for analysis, the difference between the two curves was approximately 2-fold. DISCUSSION In this investigation, we demonstrated for the first time that LRP activity is regulated in VSMCs by cytokines that have been implicated in the development of atherosclerosis. LRP regulation did not occur at the level of gene expression, but instead resulted from changes in the subcellular distribution of the receptor. VSMCs that were exposed to EGF or PDGF-BB for 24 h demonstrated about a 3-fold increase in cell-surface LRP. The same cells also showed increased LRP activity, as determined by radioligand binding/uptake and conversion of radioligand into trichloroacetic acid-soluble radioactivity. By altering LRP subcellular distribution and recycling, PDGF-BB and EGF may increase the ability of VSMCs to bind and internalize diverse LRP ligands.
A number of hormones and cytokines regulate the activity of recycling receptors. In adipocytes, insulin increases surface expression of GLUT4, the transferrin receptor, and the insulinlike growth factor receptor by altering the kinetics of receptor recycling (43,(45)(46)(47). LRP surface expression in adipocytes is also increased by insulin, probably by the same mechanism (44). The effects of insulin on receptor subcellular distribution are rapid (within 2-20 min) and reversible by removing the insulin (48). EGF, PDGF-BB, and insulin-like growth factor-1 cause redistribution of transferrin receptors to the cell surface in fibroblasts and A431 epidermoid carcinoma cells (42,49). Again, these responses are rapid and transient, peaking in a few minutes and then declining to control levels by 15 min to 2 h. By contrast, the changes in LRP subcellular distribution and activity, caused by PDGF-BB and EGF in VSMCs, were stable, even when the cytokines were withdrawn for 2.5 h. Exposure to saturating concentrations of ligand did not affect the subcellular localization of LRP. Furthermore, the increases in LRP activity were delayed responses, not evident until more than 12 h after adding EGF or PDGF-BB. Thus, the changes in VSMC LRP activity, induced by EGF and PDGF-BB, almost certainly reflect a stable change in cellular phenotype whereas the previous reports of rapid receptor regulation represent transient changes, probably occurring in response to signal transduction events.
In surface protein-biotinylation experiments, we demonstrated that the rate of LRP endocytosis is decreased by about a factor of two in EGF-treated VSMCs, providing a partial explanation for the increase in cell-surface LRP in these cells. In calculating the endocytosis rate constants, we assumed that exocytosis did not contribute significantly to the level of intracellular LRP detected in the first 10 min. The LRP endocytosis rate constant for control VSMCs reported here (ϳ1.0 min Ϫ1 ) is greater than that reported previously for other receptors, including GLUT4 (0.4 min Ϫ1 ) (46), the transferrin receptor (0.3 min Ϫ1 ) (49), and the EGF receptor (0.16 min Ϫ1 ) (50). When we analyzed LRP internalization without prior immunoprecipitation, at least six smaller biotinylated proteins were clearly detected on the same gels and could be quantitated. Although we did not attempt to identify these proteins, each was internalized at a slower rate than LRP. Thus, the endocytosis rate constant determined here may accurately reflect the more rapid uptake of LRP, compared with other internalized plasma membrane proteins.
An isolated decrease in receptor endocytosis rate would be expected to decrease the amount of ligand uptake observed when cells are exposed to a saturating concentration of ligand FIG. 7. Internalization of cell-surface LRP in control and EGFtreated VSMCs. Cell-surface proteins in control VSMCs (q) and VSMCs treated with EGF (20 ng/ml) for 24 h (E) were biotinylated with the reductant-sensitive cross-linker, NHS-SS-biotin. Endocytosis was initiated by replacing cold medium (4°C) with medium that was prewarmed to 37°C. After incubation for various periods of time, the warm medium was replaced with ice-cold medium and the cells were then processed to determine internalized LRP-biotin complex levels. LRP S is the amount of LRP-biotin complex initially present on the cell surface (before warming the cells to 37°C). LRP I is the amount of internalized LRP-biotin complex detected at each given time point. Values of (LRP S Ϫ LRP I /LRP S ) ϫ 100, which represent the percentage of the LRP-biotin complex remaining on the cell surface at each time, are plotted on a logarithmic scale. at 37°C; however, in the EGF-treated VSMCs, ␣ 2 M-MA uptake was increased (Fig. 6). This result indicates that the concentration of cell-surface LRP was elevated sufficiently, so as to more than compensate for the decrease in endocytosis rate. The same result also suggests that factors other than the LRP endocytosis rate contributed to the increase in cell-surface LRP. Possibilities include an increase in the LRP exocytosis rate and/or a change in the rate of transport of LRP through intracellular endosomes. Our results showing more rapid processing of ␣ 2 M-MA into trichloroacetic acid-soluble radioactivity in EGF-and PDGF-BB-treated cells suggest that LRP trafficking through endosomal compartments is accelerated by these cytokines. If early endosomes are involved, then both the rate of receptor recycling and the delivery of ligand to late endosomes would be affected. If late endosomes and/or lysosomes are involved, then the increased rate of ligand digestion would not be necessarily coupled to accelerated receptor recycling. In either case, the more rapid release of degraded ligand from cytokine-treated VSMCs suggests that EGF and PDGF-BB induce alterations in vesicular transport that are independent of changes in the rates of LRP endocytosis or exocytosis.
In the blood vessel wall or in a developing atheroma, the extent of saturation of LRP with ligand may be low. Under these conditions, binding and internalization of LRP ligands will depend to a greater extent on LRP cell-surface expression and less on the kinetics of receptor endocytosis. In the neointima, VSMCs coexist with fibroblasts and macrophages, both of which express LRP (1,2). Increased cell-surface expression of LRP, in VSMCs exposed to PDGF-BB or EGF, would be expected to increase the proportion of LRP ligands taken up by these cells, as opposed to the macrophages and fibroblasts. For some LRP ligands, such as cholesterol-enriched apoE-containing lipoproteins, distribution between the various cell types may influence phenotypic changes such as foam cell formation. For other ligands, such as the plasminogen activators, biologic potency is correlated with survival prior to cellular catabolism, and thus may be inversely related to the total level of cellsurface LRP contributed by all the various cell types in the tissue.
Finally, we consider the results of our work in relation to previously published analyses of LRP expression in arterial specimens (18,19). The earlier studies documented that VSMCs in atherosclerotic lesions express LRP; however, it was not possible to determine whether LRP expression was altered in lesion VSMCs compared with medial VSMCs. The techniques that are standardly used to study protein expression in tissue specimens, immunohistochemistry and in situ hybridization, do not detect shifts in subcellular receptor distribution, such as that demonstrated for LRP in the present study. Thus, the effects of PDGF-BB and EGF on VSMC LRP activity in vitro reveal a novel mechanism which may be operational in the regulation of this receptor in vivo, without detection by conventional tissue-analysis methodologies.