Low Density Lipoprotein Receptor-related Protein 1 (LRP1) Controls Endocytosis and c-CBL-mediated Ubiquitination of the Platelet-derived Growth Factor Receptor β (PDGFRβ)*

The low density lipoprotein receptor-related protein 1 (LRP1) has been implicated in intracellular signaling functions as well as in lipid metabolism. Recent in vivo and in vitro studies suggest that LRP1 is a physiological modulator of the platelet-derived growth factor (PDGF) signaling pathway. Here we show that in mouse fibroblasts LRP1 modulates PDGF-BB signaling by controlling endocytosis and ligand-induced down-regulation of the PDGF receptor β (PDGFRβ). In LRP1-deficient fibroblasts, basal PDGFRβ tyrosine kinase activity was derepressed, and PDGF-BB-induced endocytosis and degradation of PDGFRβ were accelerated as compared with control cells. This was accompanied by rapid uptake of receptor-bound PDGF-BB into the cells and by attenuated ERK activation in response to PDGF-BB stimulation. Pulse-chase analysis indicated that the steady-state turnover rate of PDGFRβ was also accelerated in LRP-deficient fibroblasts. The rapid degradation of PDGFRβ in the LRP1-deficient fibroblasts was prevented by MG132 and chloroquine. Furthermore, the association of PDGFRβ with c-Cbl, a ubiquitin E3-ligase, as well as the ligand-induced ubiquitination of PDGFRβ were increased in LRP1-deficient fibroblasts. We show that LRP1 can directly interact with c-Cbl, suggesting a Sprouty-like role for LRP1 in regulating the access of the PDGFRβ to the ubiquitination machinery. Thus, LRP1 modulates PDGF signaling by controlling ubiquitination and endocytosis of the PDGFRβ.

The low density lipoprotein receptor-related protein 1 (LRP1) 1 is a member of the low density lipoprotein receptor gene family. It consists of a 515-kDa heavy chain containing four clusters of ligand binding domains and a non-covalently associated 85-kDa light chain containing a trans-membrane and cytoplasmic domain. LRP1 cooperates with the low density lipoprotein receptor in the endocytosis and clearance of cholesterol-rich chylomicron remnants from the circulation (1,2). However, the large number and functional diversity of LRP1 ligands (3) and the lethality in LRP1 conventional knock-out mice during early to mid-gestation (4) suggest that LRP1 is involved in essential physiological processes other than lipid metabolism.
In contrast to the low density lipoprotein receptor, there is now substantial evidence indicating a role for LRP1 in cellular signaling pathways. For instance, LRP1 has been shown to form complexes with other cell surface signaling proteins, such as the urokinase-type plasminogen activator (uPA) and the uPA receptor (uPAR) complex (5,6). Furthermore, the cytoplasmic domain of LRP1 also interacts with cytoplasmic adaptor proteins that are involved in the regulation of MAP kinase activity, cytoskeletal reorganization, and cell adhesion. These proteins include Shc, FE65, PSD-95, and c-Jun amino-terminal kinase-interacting proteins (JIPs) (7)(8)(9). Recent studies indicate that LRP1 is required for the activation of MAP kinase and phosphatidylinositol 3-kinase and for focal adhesion disassembly induced by thrombospondin binding to calreticulin (10) as well as for MAP kinase activation by lactoferrin (11).
Another prominent cellular signaling pathway that can be regulated by LRP1 is the mitogenic platelet-derived growth factor (PDGF) pathway. We have shown that the disruption of LRP1 expression in vascular smooth muscle cells in the mouse resulted in elevation of PDGF receptor ␤ (PDGFR␤) expression, accelerated development of atherosclerotic lesions, and destruction of the elastic layer in large arteries (12). PDGF potently induces smooth muscle cell migration, which can be inhibited by apolipoprotein E (ApoE) in an LRP1-dependent manner (13,14).
The mechanism by which LRP1 controls PDGFR␤ activity is not completely understood. Recently, we have shown that LRP1 is present in clathrin-coated pits as well as in caveolae, where it forms a complex with the PDGFR␤ (15). PDGF-BB binds directly, albeit weakly, to the extracellular domain of LRP1 (9). Activation of PDGFR␤ induces tyrosine phosphorylation of the LRP1 cytoplasmic domain. This phosphorylation requires Src and phosphatidylinositol 3-kinase activity and is inhibited by some LRP1 ligands such as ␣2-macroglobulin and ApoE-enriched ␤ very low density lipoprotein (␤-VLDL) (15).
PDGF was originally identified in serum and platelets as a strong mitogen for fibroblasts, smooth muscle cells, and glial cells (16). PDGFs are dimeric proteins consisting of four isoforms. PDGF binding to cognate receptors results in receptor dimerization, transphosphorylation, and activation of intracellular signaling cascades. Two structurally related receptors, PDGFR␣ and PDGFR␤, have different affinities for PDGF isoforms and elicit similar, but not identical, cellular responses (16,17). PDGF signaling plays important roles in the pathogenesis of several proliferative and degenerative diseases such as tumorigenesis, arteriosclerosis, and fibrosis (17).
Activation of the intrinsic receptor protein-tyrosine kinase activity in response to PDGF binding is required for the rapid internalization and degradation of PDGFRs (16). Ubiquitination of PDGFR␤ in response to ligand binding (18) has been implicated in the efficient degradation of the receptor through proteasomal activity (19,20). Other recent studies have also suggested that monoubiquitination of receptor tyrosine kinases, including PDGFR, mediates internalization and intracellular sorting of the receptor (21)(22)(23)(24). In any case, ubiquitination of PDGFR results in a decreased mitogenic stimulus by modulating the duration and intensity of the signal (25)(26)(27).
We hypothesized that LRP1 regulates the kinetic properties of PDGFR␤ endocytosis or trafficking in the presence or absence of PDGF-BB stimulation. For this reason, we compared the turnover, trafficking, and ubiquitination of PDGFR␤ in wild type (MEF1) and LRP1-deficient (MEF2) mouse embryonic fibroblasts. We show that PDGF-BB stimulation results in accelerated down-regulation of PDGFR␤ in MEF2 cells. The rapid degradation of the receptor in MEF2 cells was accompanied by a more rapid cellular uptake of PDGF-BB from the cell surface and attenuated ERK activation. Interaction of PDGFR␤ with c-Cbl, an E3 ubiquitin ligase, and the ubiquitination of PDGFR␤ were also increased in LRP1-deficient cells. We also found that the cytoplasmic domain of LRP1 can directly interact with c-Cbl. These observations suggest that LRP1 controls the PDGFR␤ signaling pathway by modulating the access of the receptor to the ubiquitination machinery and, thereby, the time the receptor spends on the cell surface.

EXPERIMENTAL PROCEDURES
Materials-The anti-PDGFR␤ polyclonal antibody was from Upstate Cell Signaling Solutions (Lake Placid, NY). Recombinant human PDGF-BB, protein A-agarose beads, and chloroquine were from Sigma. Protein G-agarose beads came from Roche Applied Science. The anti-c-Cbl monoclonal antibody (clone 17) and the recombinant phosphotyrosine-specific monoclonal antibody (RC20) conjugated to horseradish peroxidase were from BD Biosciences. MG132 was from Calbiochem. The L-[ 35 S]methionine and L-[ 35 S]cysteine mixture and the 125 I-labeled human recombinant PDGF-BB were from Amersham Biosciences. The anti-ubiquitin monoclonal antibody (P4D1) and the anti-insulin receptor ␤ polyclonal antibody were from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-p44/42 MAP kinase polyclonal antibody and the anti-phospho p44/42 MAP kinase polyclonal antibody came from Cell Signaling Technology (Beverly, MA). Cysteine-and methionine-free Dulbecco's modified Eagle's medium (DMEM) and insulin (human, recombinant) were obtained from Invitrogen. The Enlightning® autoradiography enhancer came from PerkinElmer Life Sciences. Goat antirabbit IgG conjugated to AlexaFluor488 was from Molecular Probes (Eugene, OR).
Cell Lines-LRP1-deficient MEF cells (MEF2 cells) and LRP1-expressing MEF cells (MEF1 cells) were cultured in DMEM with 10% fetal bovine serum and penicillin/streptomycin. Prior to stimulation with PDGF-BB, cells were cultured for 18 h in DMEM containing 0.5% fetal bovine serum.
Immunoblotting-Cells were washed twice with chilled PBS and lysed in TNE buffer (20 mM Tris-HCl (pH7.4), 0.15 M NaCl, 1% Nonidet P-40, 1 mM EDTA, 5 mM ␤-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, and 10 g/ml aprotinin). Cell homogenates were centrifuged at 15,000 ϫ g for 10 min, and aliquots of the supernatants containing 10 g of protein were resolved by SDS-PAGE in a Laemmli system and transferred onto nitrocellulose membranes (Schleicher & Schuell). Subsequently, the membrane was treated with blocking solution (Tris-buffered saline containing 0.1% Tween 20 and 3% skim milk) for 1 h at 4°C. The blocked membrane was probed with primary antibodies and further incubated with a secondary antibody conjugated with horseradish peroxidase. Bound IgG was detected using an enhanced chemiluminescence system (Amersham Biosciences).
Immunoprecipitation-The cell lysates (containing 500 g of protein) were incubated with 20 l of protein A-agarose beads for 1 h at 4°C, after which the beads were removed by centrifugation to clear the lysate. Precleared lysates were incubated with the respective antibodies for 1 h at 4°C and subsequently incubated with 15 l of protein A-agarose beads or protein G-agarose beads for 1 h at 4°C, followed by washing with TNE buffer. Immunocomplexes were resuspended in 15 l of SDS-PAGE sample buffer.
Pulse Labeling and Immunoprecipitation of PDGFR␤-Confluent monolayers of MEF1 and MEF2 cells plated on 100-mm culture dishes were pre-incubated for 1 h with methionine-and cysteine-free DMEM followed by the addition of 250 Ci of L-[ 35 S]methionine and L-[ 35 S]cysteine mixture for 30 min at 37°C. Cells were washed two times with methionine-and cysteine-free DMEM. The chase time was initiated by replacing with culture media containing a 5-fold excess of unlabeled methionine and cysteine. Cells were harvested at the indicated times by lysis in TNE buffer (Fig. 1A). The 35 S-labeled PDGFR␤ was immunoprecipitated as described above and separated by electrophoresis on SDS-polyacrylamide 4 -15% gradient gels. The gels were fixed with 25% methanol and 7.5% acetate and incubated with Enlightning® (autoradiography enhancer) for 30 min, dried, and exposed to x-ray film at Ϫ80°C.

125
I-PDGF-BB Internalization Assay-Cells were grown in 12-well plates to 80% confluence and cooled by washing twice in ice-cold DMEM. Cells were incubated with 125 I-labeled PDGF-BB (10 ng/ml; Ͼ800 Ci/mmol) in 0.5 ml DMEM containing 1 mg/ml bovine serum albumin for 90 min at 4°C. Unbound ligand was removed by washing three times with ice-cold binding medium. Cells were shifted to 37°C to allow internalization of cell surface-bound ligand. At the indicated time points (Fig. 3), cells were rapidly cooled by placing the wells on ice followed by three washes with ice-cold binding medium. Ligand remaining at the cell surface was removed by incubating with acidic buffer (20 mM sodium acetate and 150 mM NaCl, pH 3.7) for 5 min followed by ice-cold PBS (two times). Cells were lysed in TNE buffer, and cell associated radioactivity was determined by ␥ counting.
Immunofluoresence Staining-MEF1 and MEF2 cells were plated on glass coverslips. Cells were allowed to grow for 2 days, followed by serum starving for 18 h prior to PDGF-BB stimulation. Cells were washed with ice-cold PBS, fixed in 4% paraformaldehyde in PBS for 20 min, and permeabilized with methanol for 10 min at Ϫ20°C. After blocking with 8% bovine serum albumin in PBS for 60 min, cells were incubated with anti-PDGFR␤ antibody (1:100 dilution in PBS containing 1% bovine serum albumin) for 60 min at room temperature, washed, and incubated with Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:100 dilution in PBS containing 1% bovine serum albumin) for 60 min at room temperature.
Transient Expression of Recombinant LRP ␤ Chain in CHO-K1 Cells-A cDNA corresponding to the carboxyl-terminal 601 amino acids of the human LRP1 ␤ chain (residues 3944 -4544) were subcloned into the pcDNA3.1 (ϩ) vector (Invitrogen). The authentic 19 amino acid LRP1 signal peptide sequence (MLTPPLLLLLPLLSALVAA) was added to the amino terminus. This construct was a kind gift from Maria Kounnas (Neurogenetics, Inc., La Jolla, CA). Transient expression of the LRP1 ␤ chain in CHO-K1 cells was achieved by transfection of 10 6 cells with 6 g of plasmid DNA using FuGENE 6 transfection reagent (Roche Applied Science).

Ligand-induced Down-regulation of PDGFR␤ Is Accelerated in LRP1-deficient
Fibroblasts-Ligand binding of PDGFR␤ leads to activation of an intrinsic protein-tyrosine kinase activity and rapid internalization and degradation of the receptor (16). To study the role of LRP1 in this process, LRP1-deficient fibroblasts (MEF2) and wild type control cells (MEF1) were stimulated with 10 ng/ml PDGF-BB, and the change of total cellular PDGFR␤ levels was determined at the indicated times (Fig. 1A). The constitutive expression level of PDGFR␤ was comparable in MEF1 and MEF2 cells. Time-dependent downregulation of PDGFR␤ was observed in both types of cells. However, the rate of PDGFR␤ down-regulation was accelerated in MEF2 cells. Densitometric analysis of the bands (Fig. 1B) showed that the level of PDGFR␤ was decreased to ϳ20% of non-stimulated cells within 30 min of exposure to PDGF-BB in MEF2 cells.
We also compared the effect of PDGF-BB stimulation on ERK1/2 phosphorylation in both MEF1 and MEF2 cells (Fig.  1A). In MEF1 cells, robust ERK phosphorylation was observed 5 min after the stimulation, which gradually decreased after 10 min and returned to baseline levels by 60 min. By contrast, the intensity of ERK phosphorylation in response to PDGF-BB stimulation in MEF2 cells was attenuated and correlated with the level of phosphorylated PDGFR␤. The rapid down-regulation of PDGFR␤ in MEF2 cells was reversed in MEF2 cells that had been transfected with full-length human LRP1 cDNA (Fig. 1C). However, the down-regulation of the insulin receptor ␤ and the duration and intensity of ERK1/2 phosphorylation in response to insulin stimulation were similar in MEF1 and MEF2 cells (Fig. 1D). We also could not observe any differences in the time course of EGFR degradation and ERK1/2 phosphorylation in response to epidermal growth factor stimulation (data not shown). These observations suggest that the effect of LRP1 on ligand-induced down-regulation is specific to PDGFR␤.
The Turnover Rate of PDGFR␤ Is Accelerated in the Absence of LRP1-To further explore the effect of LRP1 on the regulation of PDGFR␤, we studied the rate of turnover of PDGFR␤ using pulse-chase analysis. Cells were metabolically labeled for 30 min with 35 S-labeled methionine and cysteine, followed by a chase period in the absence ( Fig. 2A) or presence (Fig. 2B) of 10 ng/ml PDGF-BB. Subsequently, cell lysates were immunoprecipitated with an anti-PDGFR␤ antibody. The amounts of radiolabeled PDGFR␤ in the immunoprecipitates were estimated by fluorography. Initially, only the 160-kDa precursor form of PDGFR␤ was detected in MEF1 and MEF2 cells. The 180-kDa mature form of the receptor first appeared after a 1-h chase in both types of cells. In MEF1 cells, PDGFR␤ levels were sustained even after a 6-h chase. However, PDGFR␤ levels markedly declined in MEF2 over the same time period, indicating that the turnover rate of PDGFR␤ in LRP1-deficient cells was accelerated even in the absence of PDGF-BB stimulation ( Fig.  2A). PDGF stimulation accelerated the degradation of PDGFR in both types of cells, but the turnover rate in MEF2 cells was much faster than that of MEF1 cells (Fig. 2B). These observations are consistent with the results shown in Fig. 1 and suggest that LRP1 negatively regulates the turnover rate of PDGF␤ in murine embryonic fibroblasts.
Internalization Rate of PDGF-BB Is Increased in LRP1-deficient Cells-The ligand-induced down-regulation of receptor tyrosine kinases (RTKs) involves the endocytosis of the ligandreceptor complex, intracellular sorting of the receptors to the lysosome, and degradation. To further investigate the mechanism that mediates the accelerated down-regulation of PDGFR␤, we compared the rate of PDGF-BB internalization in MEF1 and MEF2 cells. Cells were allowed to bind 125 I-PDGF-BB at 4°C for 60 min. PDGF-BB endocytosis was initiated by shifting the incubation temperature to 37°C. Cells were washed in acidic buffer to remove radioactive ligand that had not been internalized. As shown in Fig. 3, internalization of 125 I-PDGF-BB was markedly faster in cells lacking LRP1 (MEF2 cells).
Ligand-induced Degradation of PDGFR␤ Is Attenuated by Both Proteasome and Lysosome Inhibitors-Ligand-induced degradation of PDGFR␤ depends on endocytosis by clathrincoated pits and delivery to lysosomes (28). On the other hand, proteasomal activity has also been implicated (19,20). To de- termine which mechanism was responsible for the accelerated PDGFR degradation in MEF2 cells, we investigated the effects of MG132, a commonly used proteasome inhibitor, and chloroquine, a weak base that inhibits lysosomal proteolysis, on the down-regulation of PDGFR␤ in LRP1-deficient fibroblasts. MEF2 cells were incubated in the presence of 10 M MG132 for 30 min or 200 M chloroquine for 2 h before PDGF-BB was added, and the cells were then further incubated for the indicated times (Fig. 4, A and B). Both MG132 (Fig. 4A) and chloroquine (Fig. 4B) inhibited the ligand-induced degradation of PDGFR␤ in MEF2 cells.
We next used immunofluorescence to follow the trafficking of PDGFR␤ in MEF1 and MEF2 cells exposed to chloroquine, MG132, or both of these drugs (Fig. 5). In the absence of PDGF-BB stimulation, a diffuse staining pattern was seen in both MEF1 (Fig. 5A) and MEF2 (Fig. 5B) cells stained with anti-PDGFR␤ IgG. This pattern did not change significantly when MG132 was added (Fig. 5, E and F). The addition of the lysosomal inhibitor chloroquine to the cells, alone (Fig. 5, I and J) or in combination with MG132 (Fig. 5, M and N), resulted in the accumulation of small perinuclear vesicles, which were more prominent in the MEF2 cells. This finding is consistent with the observed faster turnover rate of PDGFR␤ in these cells. Exposure to PDGF-BB (10 ng/ml for 10 min) resulted in a reduction of cell surface staining and the accumulation of PDGFR␤ in a perinuclear vesicular compartment (Fig. 5, C and  D), which was more pronounced in MEF2 cells (Fig. 5D). MG132 significantly inhibited the loss of apparent cell surface staining and the accumulation of PDGFR␤ in vesicles (Fig. 5, G  and H). By contrast, chloroquine dramatically increased the accumulation of PDGFR␤ in large vesicular compartments in MEF1 (Fig. 5K) and MEF2 cells (Fig. 5L). Importantly, PDGFR␤ accumulation in chloroquine-treated MEF1 cells was much less pronounced than in MEF2 cells. When MG132 and chloroquine were added together prior to exposure to PDGF-BB (Fig. 5, O and P), PDGFR␤ transport to the large perinuclear vesicular compartment seen in Fig. 5, K and L was blocked. The resulting staining pattern was very similar to that seen with MG132 alone, suggesting that PDGFR␤ internalization had been arrested at a more superficial compartment. Thus, both MG132 and chloroquine reduce degradation of PDGFR␤, but they appear to exert their effects at two different steps in the PDGFR␤ trafficking pathway.
Ligand-induced Ubiquitination of PDGFR␤-Ubiquitination of PDGFR␤ is a necessary step for receptor internalization and subsequent degradation in response to PDGF stimulation (19,20), which raises the possibility that the receptor might be hyper-ubiquitinated in MEF2 cells. Therefore, we compared the levels of ubiquitinated PDGFR␤ in MEF1 and MEF2 cells after PDGF-BB stimulation (Fig. 6A). To protect PDGFR␤ from

LRP1 Regulates the Ligand-induced Degradation of PDGFR␤
ligand-induced internalization and degradation, serum-deprived cells were stimulated with 10 ng/ml PDGF-BB at 4°C for the indicated times (Fig. 6A). PDGFR␤ was then immunoprecipitated from cell lysates. Ubiquitinated PDGFR␤ was detected by immunoblotting with an anti-ubiquitin antibody (Fig.  6A). In both MEF1 and MEF2 cells, ligand-induced ubiquitination of the receptor was observed in a time-dependent manner. Importantly, the intensity of the anti-ubiquitin immunoblot was markedly greater in MEF2 cells compared with MEF1 cells, indicating that ubiquitination of PDGFR␤ is enhanced in the absence of LRP1.
Association of c-Cbl with PDGFR␤ Is Increased in LRP1deficient Cells-c-Cbl is a cytoplasmic ubiquitin ligase that has been implicated as a negative regulator of RTK signaling (29). Activation of RTKs stimulates its E3 ligase activity, which appears to be necessary for down-regulation of PDGFR␤ (24,26). To determine whether c-Cbl was involved in the enhanced PDGFR␤ ubiquitination in LRP1 deficient cells, we investigated the ability of PDGFR␤ to interact with c-Cbl in MEF1 and MEF2 cells. Cells were stimulated with PDGF-BB for the indicated times (Fig. 6B) at 37°C, and PDGFR␤ was then immunoprecipitated from the cell lysate and processed for immunoblotting with anti c-Cbl IgG (Fig.  6B). In MEF2 cells, c-Cbl co-precipitated with PDGFR␤ even in the absence of PDGF-BB stimulation (Fig. 6B, lane 4), and a small amount of PDGFR␤ was phosphorylated even under these baseline conditions. By contrast, very little c-Cbl was detected when PDGFR␤ was immunoprecipitated from MEF1 cells under the same conditions (Fig. 6B, lane 1). When MEF2 cells were stimulated with PDGF-BB there was a dramatic loss of receptor, but the remaining receptor was phosphorylated and associated with c-Cbl (Fig. 6B, lanes 5 and 6). Comparatively little c-Cbl was associated with PDGFR␤ in MEF1 cells, regardless of PDGF stimulation (Fig. 6B, lanes  1-3). Although significantly more c-Cbl was immunoprecipitated with PDGFR␤ independently of PDGF-BB stimulation in MEF2 cells, it was less extensively phosphorylated than in MEF1 cells. When LRP1 expression was restored in MEF2 cells by retransfection with a full-length LRP1 expression construct (Fig. 6C, lanes 3 and 4), c-Cbl association with the PDGFR␤ was reduced to levels comparable with those observed in MEF1 cells (compare Fig. 6C, lanes 1 and 3 with Fig. 6B, lanes 1 and 4). We conclude that the presence of LRP1 reduces the association of c-Cbl with PDGFR␤ independently of PDGF-BB signaling (12)(13)(14). Although the association of c-Cbl with PDGFR␤ did not require PDGF-BB stimulation, activation of c-Cbl by tyrosine phosphorylation appears to be required to productively ubiquitinate PDGFR␤.
LRP1 Directly Interacts with c-Cbl-To gain further insight into the mechanism by which LRP1 controls the association of c-Cbl with the PDGFR␤, we tested whether LRP1 itself can interact with this ubiquitin ligase (Fig. 7). CHO-K1 cells were used for initial experiments because of their superior transfection efficiency. A truncated form of LRP1 encoding the carboxyl-terminal 601 amino acids, which includes the complete cytoplasmic domain (Fig. 7A, lanes 3 and 4), was overexpressed in cells that had been exposed (Fig. 7A, lane 4) or not exposed (Fig.  7A, lane 3) to PDGF-BB. Cell lysates were prepared, and endogenous c-Cbl was immunoprecipitated followed by immunoblotting with antibodies directed against c-Cbl (upper section) and LRP1 (middle section), respectively. Empty vector was used as a control (Fig. 7A, lanes 1 and 2). Endogenous c-Cbl was constitutively associated with the LRP1 subunit independent of PDGF-BB stimulation (Fig. 7A, lanes 3 and 4). This result was confirmed in MEF2 cells that had been transfected with the LRP1 ␤-chain (Fig. 7B).

DISCUSSION
Earlier studies have shown that LRP1 regulates PDGFR␤ signal transduction, but the mechanism by which this occurs is not known (9,12,15). The results of the current study indicate that in mouse fibroblasts LRP1 functions as a PDGFR␤ anchor protein that controls the traffic of the receptor from the cell surface to intracellular compartments. The absence of LRP1 caused both a faster internalization of PDGFR after ligand binding (Fig. 1) and an accelerated rate of turnover (Fig. 2). A direct consequence of the altered kinetics is a shorter duration of the PDGFR␤ signal in response to PDGF-BB. The effect of LRP1 on PDGFR␤ is selective, as we did not detect any alteration in the down-regulation of either the insulin receptor or the EGFR.
The effect of LRP1 on PDGFR␤ expression and signaling is also cell type-specific. We have previously reported that smooth muscle cell-specific elimination of LRP1 resulted in the eleva-FIG. 6. PDGF-BB-induced ubiquitination of PDGFR␤ is enhanced in LRP1-deficient cells. A, MEF1 and MEF2 cells were incubated with or without 10 ng/ml PDGF-BB at 4°C for 0 -120 min. Cells were lysed in 500 l of TNE buffer. Cell lysates (containing 1 mg of protein) were processed for immunoprecipitation (IP) with an anti-PDGFR␤ antibody, followed by immunoblotting (IB) with an anti-ubiquitin antibody (upper section) or an anti-PDGFR␤ antibody (lower section). B, the association of c-Cbl with PDGFR␤ is enhanced in LRP1deficient fibroblasts. MEF1 and MEF2 cells were stimulated with 10 ng/ml PDGF-BB at 37°C and lysed at the indicated times in TNE buffer. Cell lysates (containing 1 mg of protein) were processed for immunoprecipitation with an anti-PDGFR␤ antibody (upper section) or c-Cbl antibody (lower section), followed by immunoblotting with anti-c-Cbl, anti-phosphotyrosine (PY), or an anti-PDGFR␤ antibody. C, LRP1 expression in MEF2 cells reduces the association of c-Cbl with PDGFR␤. MEF2 cells and MEF2 cells transfected with human LRP1 were stimulated with 10 ng/ml PDGF-BB at 37°C and lysed at the indicated times in TNE buffer. Cell lysates were processed for immunoprecipitation with an anti-PDGFR␤ antibody. tion of PDGFR␤ expression in the tissue (12). On the other hand, the knockdown of LRP1 expression by small interfering RNA (siRNA) in HT1080 fibrosarcoma cells decreased PDGFR␤ expression level (32). In the same study, Wu et al. also noted reduced LRP1 levels in MEF2 cells, although this finding was not attributed to increased degradation. Nevertheless, these studies show that LRP1 can influence PDGFR kinetics with either positive or negative effects on PDGFR␤ signaling, depending on the cellular context. PDGFR␤ is predominantly localized to caveolae (33), where it is available to interact with adapter proteins such as Shc and Grb2, as well as Src family kinases that are enriched in this membrane domain (33,34). LRP1 is also present in caveolae where it is tyrosine-phosphorylated when PDGF-BB binds to the PDGFR␤ (9, 15). PDGFR␤ exits the caveolar compartment following stimulation with PDGF-BB. The rate at which this occurs is markedly slower for PDGFR␤ than for EGFR (35), an RTK that does not interact with LRP1. Once PDGFR␤ exits the caveolae, it most likely moves to clathrin-coated pits where it is internalized. LRP1 also appears to escort the urokinase-type plasminogen activator receptor from caveolae (36) to clathrin coated pits (37). Unlike PDGFR␤, however, internalization of GPI anchored urokinase-type plasminogen activator receptor depends on the ability of the cytoplasmic tail of LRP1 to interact with coated pit proteins.
Studies addressing the role of RTK activity in endocytosis indicate that RTK activity is required for the ligand-induced internalization of PDGFR␤ (30). Moreover, dephosphorylation of activated PDGFR appears to play a role in regulating receptor down-regulation (38). Therefore, the accelerated turnover of PDGFR␤ in LRP1-deficient cells could be due, at least in part, to an elevation in RTK activity. We found that in LRP1-deficient fibroblasts a small fraction of PDGFR␤ and c-Cbl was tyrosine-phosphorylated even in the absence of PDGF-BB stimulation (Fig. 6B). These observations imply that the negative regulation of basal PDGFR␤ tyrosine kinase activity by LRP1 results in an attenuation of PDGFR␤ trafficking from the cell surface to intracellular compartments. Because LRP1 is physiologically present in a complex with the PDGFR␤ (9, 15), a plausible explanation for this suppression of basal kinase activity might lie in the prevention of ligand-independent receptor dimerization and transphosphorylation. This unphysiological event frequently occurs during malignant transformation under conditions where RTKs, such as the PDGFR or the EGFR, are amplified and overexpressed.
On the other hand, we also observed a clear increase in the level of PDGF-BB-dependent PDGFR␤ ubiquitination in LRP1deficient cells compared with wild type cells (Fig. 6A). Previous studies have shown that PDGF-BB causes polyubiquitination of PDGFR␤ (19,20). Polyubiquitination can occur by either the formation of ubiquitin polymers attached to a single lysine residue or the addition of single ubiquitins to multiple lysine residues (monoubiquitination). Recently Haguland et al. (21) presented strong evidence that PDGFR␤ is monoubiquitinated at multiple sites. Ubiquitination can control protein function and is a known regulator of receptor-mediated endocytosis in both yeast and mammalian cells (38). Exactly how endocytosis is regulated by ubiquitin is not known, but it may involve interactions between ubiquitinated receptors and ubiquitin binding motifs such as UIM (ubiquitin-interacting motif) and CUE domains (39). Thus, increased binding of the ubiquitin ligase c-Cbl to PDGFR␤ in LRP-deficient cells results in an increased number of ubiquitinated sites on each receptor. The presence of more ubiquitins increases the affinity of the receptor for the ubiquitin-interacting motifs in adaptor molecules such as eps15/epsin, an essential molecule in coated pit endocytosis, and internalization is accelerated. Interestingly, interaction between c-Cbl and PDGFR␤ may occur in caveolae (40).
This increased interaction of PDGFR␤ with c-Cbl was more apparent in MEF2 cells even in the absence of PDGF-BB stimulation (Fig. 6B). Hence, c-Cbl-induced ubiquitination may play a causative role in the rapid ligand-induced degradation of PDGFR␤ in LRP1-deficient cells. One possible explanation is that LRP1 may regulate the c-Cbl induced ubiquitination of PDGFR␤, similar to the way in which human Sprouty2 (hSpry2) has been proposed to regulate EGFR. Drosophila Sprouty was originally discovered as a negative regulator of EGFR signaling (41,42). Human Sprouty 2 is tyrosine-phosphorylated in response to epidermal growth factor stimulation (43). It enhances the cell surface retention of the EGFR, resulting in sustained signal transduction by inhibiting c-Cbl-mediated ubiquitination of the EGFR (44). LRP1 appears to have a similar Sprouty-like function on  4). Cells were serum-starved for 18 h, stimulated with 10 ng/ml PDGF-BB for 5 min, and lysed in 500 l TNE buffer. Cell lysates (1 mg protein) were immunoprecipitated (IP) with anti-c-Cbl monoclonal IgG, followed by immunoblotting (IB) with anti-c-Cbl (upper section) or anti-LRP1 antibodies (middle section). Expression of the recombinant LRP1 ␤-chain was confirmed by immunoblotting (lower section). B, MEF2 cells (100-mm dishes) were transfected with 20 g of pcDNA3.1 (ϩ) vector control (lanes 1 and 2) or pcDNA3.1-LRP1 ␤-chain (lane 3-6). Cells were serum-starved for 18 h, stimulated with 10 ng/ml PDGF-BB for 5 min, and lysed in 500 l TNE buffer. Cell lysates (1 mg protein) were immunoprecipitated with anti-c-Cbl monoclonal IgG (lanes 1-4) or mouse non-immune serum (lanes 5 and 6), followed by immunoblotting with anti-c-Cbl (upper section) or anti-LRP1 antibodies (middle section). Expression of the recombinant LRP1 ␤-chain was confirmed by immunoblotting (lower section). PDGFR␤ signaling. LRP1 can interact with c-Cbl independently (Fig. 7), which may competitively or sterically interfere with the association of c-Cbl with PDGFR␤ (Fig. 6B), thereby decreasing the rate at which the receptor is ubiquitinated (Fig.  6A) and removed from the cell surface by endocytosis. In fibroblasts, this results in stabilization of PDGFR␤ at the cell surface and extension of the duration of the mitogenic signal upon PDGF-BB stimulation. On the other hand, LRP1 also appears to suppress basal kinase activity of the PDGFR␤, as illustrated by the low level of tyrosine-phosphorylated PDGFR␤ and c-Cbl in the absence of PDGF-BB stimulation in LRP1-deficient cells (Fig. 6B, and Ref. 12). This finding explains the higher growth rates of MEF2 as compared with that of MEF1 cells.
Ligand-induced degradation of PDGFR␤ was blocked by the proteasome inhibitor MG132 (Fig. 4A). It is tempting to assume that the effect of the proteasome inhibitors is linked to receptor ubiquitination, because polyubiquitination of proteins will trigger proteasome-mediated degradation. However, MG132 inhibited degradation of PDGFR␤ by preventing the delivery of the receptor to a chloroquine-sensitive compartment, most likely the lysosome (Fig. 4A). Proteasome inhibitors have been found to affect receptor trafficking in several systems. For example, proteasome inhibitors block EGFR degradation by inhibiting the delivery of the receptor to the interior of multivesicular bodies (23). Furthermore, MG132 attenuates delivery of the growth hormone receptor to the lysosome for degradation even though this receptor is not a proteasome target (45).
Thus, proteasome inhibitors impede the traffic of these molecules independently of their ability to inhibit proteasomemediated degradation, explaining the increased amount of receptor at the cell surface following exposure to PDGF (Fig. 5, G  and H). Recently, a new connection between the proteasome and membrane traffic has been established with the discovery that a proteasome ␣-subunit (XAPC7/PSMA7/HSPC) interacts with Rab7 on multivesicular bodies and that overexpression of this subunit interferes with the transport of membrane proteins from early to late endosomes (46). Exactly how proteasome inhibitors affect this interaction and thereby alter subcellular trafficking remains to be determined.
In conclusion, we have shown in this study that LRP1, a multifunctional endocytic receptor and modulator of cellular signaling, controls PDGFR␤ signaling at the plasma membrane. This has important physiological consequences for the regulation of cell proliferation and migration in diverse cell types in vitro and in vivo and, specifically, for the process of atherosclerotic transformation of the vascular wall (12).