Platelet-derived Growth Factor (PDGF)-induced Tyrosine Phosphorylation of the Low Density Lipoprotein Receptor-related Protein (LRP) EVIDENCE FOR INTEGRATED CO-RECEPTOR FUNCTION BETWEEN LRP AND THE PDGF*

The low density lipoprotein receptor-related protein (LRP) functions in the catabolism of numerous ligands including proteinases, proteinase inhibitor complexes, and lipoproteins. In the current study we provide evidence indicating an expanded role for LRP in modulat-ing cellular signaling events. Our results show that platelet-derived growth factor (PDGF) BB induces a transient tyrosine phosphorylation of the LRP cytoplasmic domain in a process dependent on PDGF receptor activation and c-Src family kinase activity. Other growth factors, including basic fibroblast growth factor, epidermal growth factor, insulin-like growth factor-1, were unable to mediate tyrosine phosphorylation of LRP. The basis for this selectivity may result from the ability of LRP to bind PDGFBB, because surface plasmon resonance experiments demonstrated that only PDGF, and not basic fibroblast growth factor, epidermal growth factor, or insulin-like growth factor-1, bound to purified LRP immobilized on a sensor chip. The use of LRP mini-receptor mutants as well as in vitro phosphorylation studies demonstrated that the tyrosine located within family kinases. The phosphotyrosine within the NPVY provides a docking site for the Shc PTB domain, and association of this molecule with LRP may modulate signaling properties of the PDGF receptor.

The low density lipoprotein receptor-related protein (LRP) 1 is a large endocytic receptor containing a 515-kDa heavy chain to which ligands bind and a non-covalently associated 85-kDa light chain containing a transmembrane and cytoplasmic domain (for review see Ref. 1). LRP is one of 12 or more receptors that make up the LDL receptor superfamily and is essential for embryonic development in mice (2). A remarkable feature of LRP is its ability to bind and mediate the internalization of a diverse array of ligands, including proteinases (3,4), proteinase-inhibitor complexes (5,6), and lipoproteins (7). After binding to the LRP, the ligands are transported into endosomes where they uncouple in the reduced pH environment and are sorted to lysosomes for degradation. LRP recycles back to the cell surface where it is once again available to bind ligands.
Recent studies indicate that in addition to their cargo transport function, certain LDL receptor family members also participate in signaling pathways. For example, the very low density lipoprotein receptor and apoE receptor 2 both participate in a signal transduction pathways mediated by reelin (8 -10). Reelin is secreted by Cajal-Retzius cell in the outermost layer of the cerebral cortex and controls the final position of neurons that migrate from the ventricular zone. Binding of reelin to either the very low density lipoprotein receptor or apoE receptor 2 induces tyrosine phosphorylation of disabled-1 (Dab1) (9,10), an adaptor protein that interacts with the cytoplasmic domains of LDL receptor family members (11,12) and functions in tyrosine kinase signaling pathways.
In the case of LRP, accumulating evidence suggests a prominent but undefined role for this receptor in regulating cell physiology by facilitating signal transduction pathways. For example, LRP has been implicated as a component of the receptor complex for midkine (13), a heparin binding growth factor with migration-promoting and survival-promoting activities. Another LRP ligand, tissue type plasminogen activator, promotes late phase long term potentiation (14), and this activity appears to require its association with LRP (15). Finally, the binding of activated ␣ 2 M (␣ 2 M*) to LRP mediates calcium influx in neurons in a process that also involves N-methyl-D-aspartate receptors (16). The exact role LRP plays in all of these processes is not known. Recently, Barnes et al. (17) demonstrated that LRP is tyrosine-phosphorylated in v-Src-transformed cells and provided evidence suggesting that phosphorylated LRP binds to Shc, an adaptor protein that is important in the activation of the Ras (18) and c-Myc signaling pathways (19).
In the present investigation we demonstrate that platelet derived growth factor (PDGF) BB directly binds to LRP and promotes the transient tyrosine phosphorylation of the LRP cytoplasmic domain via activation of the PDGF receptor and c-Src. This phosphorylation occurs on a tyrosine residue located within the second NPXY motif found in the LRP cytoplasmic domain and generates a docking site for adaptor proteins such as Shc. In the accompanying paper Boucher et al. (20) demonstrate that, like the PDGF receptor, LRP also localizes in caveolae and the LRP ligand apoE-enriched ␤ very low density lipoprotein blocks PDGF-mediated tyrosine phosphorylation of LRP. Taken together, these findings suggest an integrative co-receptor function between the PDGF receptor and LRP, indicating that LRP and certain of its ligands may modulate signal transduction pathways mediated by the PDGF receptor.

EXPERIMENTAL PROCEDURES
Proteins, Antibodies, and Expression Constructs-A rabbit polyclonal IgG prepared against purified human LRP (R2629) was affinity-purified over LRP-Sepharose as described (21). Monoclonal antibody 5A6, which recognizes the LRP light chain (or ␤ subunit), was prepared against human LRP and has been described (22). Cells producing the anti-Myc IgG 9E10 were obtained from the American Type Culture Collection (Manassas, VA), and the IgG was purified by chromatography on protein G-Sepharose. Anti-PDGF receptor ␤ rabbit polyclonal IgG was purchased from Santa Cruz Biotechnology. The phosphotyrosine-specific monoclonal antibody 4G10 (23) conjugated to horse-radish peroxidase and anti-Src rabbit polyclonal IgG were purchased from Upstate Biotechnology. Anti-Shc rabbit polyclonal IgG was purchased from Transduction Laboratories, whereas phospho-specific and total extracellular signal-regulated kinase polyclonal antibodies were obtained from New England Biolabs.
Basic FGF, PDGFBB, PDGFAA, IGF-1, and EGF were purchased from R&D Systems. In all experiments, PDGFBB was utilized. LRP was isolated from human placenta as described by Ashcom et al. (24) and labeled with [ 125 I]iodine to a specific activity ranging from 2 to 10 Ci/g protein using iodogen (Pierce). Human receptor-associated protein (RAP) was expressed in bacteria as fusion proteins with glutathione S-transferase and was cleaved and purified as described previously (25). The cytoplasmic domain of LRP was expressed as a fusion protein with glutathione S-transferase in Escherichia coli using pGEX-2T expression vector (Promega). Construction of this expression vector was accomplished by preparing a cDNA fragment encoding amino acid residues 4426 -4525 of human LRP (numbering is based on the mature protein, as defined in Herz et al. (26)) by polymerase chain reaction using 21-base synthetic oligonucleotide primers and an LRP cDNA (27) as a template. The fusion protein was expressed and purified as described (25). Substitutions of asparagine and tyrosine to alanines in the two NPXY motifs of the cytoplasmic domain of the GST cytoplasmic domain were performed using Transformer site-directed mutagenesis kit (CLONTECH) and confirmed by sequencing. The cDNA of human LRP (27) was also used as a template to generate expression vectors for the LRP ␤ essentially as described (28). Briefly, the fragment of cDNA that encodes amino acids 3844 -4525 of LRP (GenBank TM access number X13916) was generated by PCR amplification and subcloned into pSecTag expression vector (Invitrogen) modified to produce a protein with two copies of Myc epitope at its amino terminus. The mini-receptor contains a portion of the LRP extracellular domain (including membrane proximal YWTD ␤-propeller and EGF-like repeats), transmembrane domain, and cytoplasmic tail. Substitutions of asparagine and tyrosine to alanines in the two NPXY motifs of the cytoplasmic domain of the mini-receptor were performed using the Transformer site-directed mutagenesis kit (CLONTECH) and confirmed by sequencing. All PCR products were sequenced before using to confirm that no errors were introduced by the PCR. Expression constructs encoding wild-type c-Src and kinase-inactive c-Src (K279R) in pUSEamp(Ϫ) were pur-chased from Upstate Biotechnology. A plasmid containing HA-tagged Shc was a generous gift from Dr. K. S. Ravichandran (University of Virginia, Charlotte, VA).
Solid-phase Binding Assay-Microtiter wells were coated with PDG-FBB (2 g/ml in Tris-buffered saline (TBS), pH 8.0, 50 l/well) overnight and then blocked with 300 l of 3% bovine serum albumin in TBS. 100 l of 125 I-labeled LRP (200 nM) was then added to the wells in the absence or presence of RAP (20 M) and incubated overnight at 4°C. After incubation, the microtiter wells were washed and counted.
Surface Plasmon Resonance-Binding of PDGFBB, bFGF, EGF, IGF-1 to purified LRP was measured using a BIA 3000 optical biosensor (Biacore AB, Uppsala, Sweden). For these studies, the BIAcore sensor chip (type CM5; Biacore AB) was activated with a 1:1 mixture of 0.2 M N-ethyl-NЈ- (3-dimethylaminopropyl)carbodiimide and 0.05 M N-hydroxysuccinimide in water as described by the manufacturer. Purified human LRP was immobilized at the level of 3000 response units in a working solution of 10 g/ml in 10 mM sodium acetate, pH 4.0, through the BIAcore flow cell at a rate of 5 l/min. The remaining binding sites were blocked by 1 M ethanolamine, pH 8.5, whereas unbound protein was washed out with 0.5% SDS. An additional flow cell, similarly activated and blocked without immobilization of protein, served as a negative control. A flow cell with immobilized ovalbumin at the level of 500 response units was used as a control for nonspecific protein binding. All binding reactions were performed in 10 mM HEPES, 0.15 M NaCl, 0.005% Tween 20, pH 7.4 (HBS-P buffer) (Biacore AB), containing 0.005% Tween 20. Binding of PDGFBB and selected growth factors to LRP was measured at 25°C at a flow rate of 30 l/min for 4 min, followed by 4 min of dissociation. The bulk shift due to changes in refractive index measured on blank surfaces was subtracted from the binding signal at each condition to correct for nonspecific signals. Chip surfaces were regenerated with subsequent 1-min pulses of 10 mM sodium acetate, pH 4.0, containing 1 M NaCl and 10 mM NaOH containing 1 M NaCl followed by 2 min of washing with running buffer to remove this high salt solution. All injections were performed using Application Wizard in the automated method. Binding of PDGFBB was measured using 2-fold dilutions in HBS-P buffer over a range of concentrations (20 -0.6 nM). Other growth factors as bFGF, EGF, and IGF-1 were injected at concentrations of 50 nM. All collected data were analyzed with BIA evaluation 3.0 software (Biacore) using global analysis to fit 1:1 Langmuir binding with mass transfer limitation and heterogeneous ligand models.
Cell Culture and PDGF Treatment-WI-38 fibroblasts were cultured in 150-mm plates in DMEM containing 10% serum until they reached 60 -70% confluency. Cell layers were then washed three times with serum-free medium. After washing, the media was replaced with DMEM containing either 0.1% fetal bovine serum or 1% Nutridoma® NS, and the cells were incubated with this media for an additional 18 h. For PDGF treatment, PDGFBB was added to the cells in DMEM containing either 0.1% fetal bovine serum or 1% Nutridoma® NS.
Immunoprecipitation and Immunoblot Analysis-After stimulation, cell layers were washed 2 times with cold Dulbecco's phosphate-buffered saline containing 1 mM orthovanadate, and the lysate was prepared in lysis buffer (50 mM Tris, 150 mM NaCl, 1% Nonidet P-40, and a protease and phosphatase inhibitor mixture (Calbiochem)). After preclearing with mouse (or rabbit), IgG (10 g/ml) lysates were immunoprecipitated with either monoclonal 5A6-protein G-Sepharose or R2629-protein G-Sepharose. Immunoprecipitates were washed 3 times with lysis buffer and then boiled with nonreducing SDS-PAGE sample buffer for 10 min. Samples were separated by electrophoresis on 4 -20% or 8% SDS-PAGE precast gels (Invitrogen) and transferred to nitrocellulose membranes for immunoblot analysis. Membranes were blocked with 2% bovine serum albumin in Dulbecco's phosphate-buffered saline for 1 h and incubated with anti-phosphotyrosine monoclonal IgG 4G10horseradish peroxidase (HRP) conjugate (Calbiochem) (1:3000 dilution) in 2% bovine serum albumin in Dulbecco's phosphate-buffered saline with 0.1% Tween 20 for 1 h and washed 5 ϫ 4 in Dulbecco's phosphatebuffered saline with 0.1% Tween 20. Membranes were developed with chemiluminescent reagent (Pierce), and bands were visualized using Biomax MR film (Eastman Kodak Co.). For visualizing immunoprecipitated LRP, the membranes were first stripped with Re-blot Western blot recycling kit (Chemicon International), and then the membranes were probed with iodinated 5A6 (2 g/ml) overnight, washed, and exposed to BiomaxMR film. Extracellular signal-regulated kinase activation was measured by immunoblotting cell lysates (25 g/lane) with phospho-specific and total extracellular signal-regulated kinase polyclonal antibodies (New England Biolabs) according to manufacturer's instructions.
Inhibitor Assays-WI-38 cells were grown in 150-mm plates in DMEM containing 10% serum to 60 -70% confluency. Cell layers were then washed 3 times with serum-free medium and then incubated in 1% Nutridoma containing DMEM overnight. After serum deprivation, cells were preincubated for 15 min with inhibitors PP2 (300 nM), PP3 (300 nM), and AG1296 (900 nM) followed by incubation with 30 ng/ml PDG-FBB for 12-15 min. PP2, PP3, and AG1296 were obtained from Calbiochem. Cells were washed, lysed, and processed as mentioned above. Transient Expression of HA-tagged Shc cDNA in COS-1 Cells-COS-1 cells were cultured to 30% confluency and then transfected with 10 g of HA-Shc containing plasmid using FuGENE 6 transfection reagent (Roche Molecular Biochemicals). 24 h after transfection, the cells were cultured for 18 h in DMEM medium containing 0.1% fetal bovine serum. The cells were then treated with PDGFBB (40 ng/ml), whereas control cells received no treatment. After 15 min, cell extracts were prepared and subjected to immunoprecipitation with anti-LRP polyclonal R2629, and immunoprecipitated proteins were subjected to immunoblot analysis with anti-phosphotyrosine IgG 4G10 and anti-HA IgG.
In Vitro Phosphorylation of LRP Cytoplasmic Domain-Src kinase assay was carried out using purified active human recombinant Src kinase (Calbiochem) according to the manufacturer's instructions. Briefly 10 g of purified protein (GST-wild-type (WT), GST-(NPTY 3 APTA), GST-(NPVY 3 APVA), GST) was incubated with 10 units of purified Src and 10 Ci of [␥-32 P]ATP for 10 min at 30°C. The reactions were terminated by the addition of sample buffer containing ␤Ϫmercaptoethanol. Phospho-labeled proteins were separated on 4 -12% SDS-PAGE precast gel and blotted onto nitrocellulose membranes. The nitrocellulose membrane was stained with Ponseau S and then exposed to BiomaxMS film.
Transient Transfection of Src and LRP-␤ in COS-1 Cells-WT LRP-␤ or each of NPTY 3 APTA, NPVY 3 APVA mutant LRP expression vectors were transiently transfected into COS-1 cells with either pUSE empty vector (mock) or one of the c-Src expression plasmids (c-Src(WT), c-Src K297R (kinase inactive)). After transfection, cell extracts were prepared and subjected to immunoprecipitation with anti-LRP monoclonal 5A6. The immunoprecipitated proteins were tested for tyrosine phosphorylation by immunoblotting with anti-phosphotyrosyl antibodies. Whole cell extracts (2%) were analyzed by immunoblotting for LRP-␤ expression using anti-Myc IgG and for c-Src expression using anti-Src IgG.

RESULTS
PDGF Promotes Tyrosine Phosphorylation of the LRP ␤ Subunit-When added to fibroblasts, PDGF binds to PDGF receptors and induces rapid dimerization and tyrosine phosphorylation of these receptors (29). These events are followed by internalization and degradation of the receptor-ligand complex (30). To determine whether PDGF induces phosphorylation of the LRP cytoplasmic domain, WI-38 fibroblasts were incubated with PDGF for varying time periods. Analysis of cell extracts for the presence of phosphotyrosine using the phosphotyrosinespecific monoclonal antibody 4G10 (23) and for the PDGF receptor antigen identified the rapid tyrosine phosphorylation of a prominent band with a mobility identical to the PDGF receptor (Fig. 1A). The phosphoprotein disappeared with time, as expected for the PDGF receptor.
The same cell extracts were subjected to immunoprecipitation with anti-LRP IgG and probed for tyrosine phosphorylation. The results (Fig. 1B) demonstrated that the LRP ␤ chain becomes phosphorylated on tyrosine residues in response to PDGF treatment. Maximal LRP phosphorylation occurred 10 min after the addition of PDGF, and interestingly, the phosphorylation was transient and disappeared with time. These results indicate that binding of PDGF to cells results in a transient phosphorylation of the tyrosine residue(s) within the LRP cytoplasmic domain. Similar results were obtained in rat smooth muscle cells (data not shown).
Tyrosine Phosphorylation of LRP Is Selective for PDGF-To determine whether other growth factors are also able to mediate phosphorylation of LRP, serum-starved fibroblasts were incubated with various growth factors before subjecting cell extracts to immunoprecipitation with anti-LRP monoclonal antibodies. Immunoprecipitates were then probed with anti-phos-photyrosine. The results demonstrate that only PDGF-treated fibroblasts contain tyrosine-phosphorylated LRP (Fig. 2, upper  panel, lane 3). Other growth factors including bFGF, IGF-1, and EGF did not stimulate the tyrosine phosphorylation of LRP. When the membranes were stripped and reprobed with 125 I-labeled anti-LRP IgG we confirmed that LRP was immunoprecipitated from the cell extracts (Fig. 2, middle panel). A control experiment confirmed that all growth factors were active because they were able to induce the phosphorylation of extracellular signal-regulated kinase (Fig. 2, lower panel). These results reveal selectivity for PDGF in mediating the tyrosine phosphorylation of the LRP cytoplasmic domain.
PDGF Directly Binds to LRP-To gain insight into possible mechanisms by which PDGF promotes LRP tyrosine phosphorylation, studies were initiated to determine whether LRP can directly bind PDGF. To accomplish this, the binding of PDGF and other growth factors to purified LRP immobilized on a BIACore sensor chip was measured. The sensorgrams, shown in Fig. 3A, demonstrate that PDGF, but not bFGF, EGF, and IGF, bound LRP immobilized on a sensor chip. The binding of PDGF but not other growth factors by LRP might offer an explanation for the selectivity in LRP phosphorylation observed in Fig. 2.
The affinity of PDGF binding to LRP was estimated by injecting varying concentrations of PDGF on the LRP-immobilized chip (Fig. 3B), and the results demonstrate a concentration-dependent binding of PDGF to the sensor chip. The data did not fit a single site model but were adequately described by a model in which LRP contains two binding sites for PDGF. The binding is characterized by K D values of 12 and 17 nM, with one site displaying rapid association and dissociation rates, whereas the second site displayed slower association and dissociation rates.
The binding of all known ligands to LRP is prevented by the 39-kDa RAP (25,31). To determine whether the binding of PDGF to LRP is inhibited by RAP, an enzyme-linked immunosorbent assay was performed in which the binding of 125 I- labeled purified LRP to microtiter wells coated with PDGF was examined. The results (Fig. 4A) demonstrate that RAP partially inhibited the binding of LRP to immobilized PDGF and did not completely reduce the binding to background levels. At this time, we speculate that LRP contains two PDGF binding sites, with only one site sensitive to RAP inhibition. However, to prove this will require additional studies. We next measured the effect of RAP on PDGF-mediated phosphorylation of LRP. WI-38 fibroblasts were incubated with PDGF in the presence and absence of RAP, and the degree of tyrosine phosphorylation of the LRP cytoplasmic domain was measured. The results (Fig. 4B, lane 3) demonstrate that RAP slightly reduces but does not prevent the phosphorylation of LRP mediated by PDGF, consistent with the data presented in Fig. 4A.

PDGF-induced LRP Phosphorylation Requires the PDGF Receptor and Is Mediated by Src or Src-related Kinases-
The ability of PDGF to bind to LRP raises the possibility that phosphorylation of LRP mediated by PDGF could result either from direct association of this growth factor with LRP or, alternatively, via an integrative interaction between the PDGF receptor and LRP. To further characterize the mechanism by which PDGF initiates tyrosine phosphorylation of LRP, several inhibitors were employed. The PDGF receptor is known to activate Src family kinases, and thus, we used the cell-permeable Src family kinase inhibitor PP2 to determine whether Src family members are involved in tyrosine phosphorylation of LRP. PP2 competes with ATP for binding to Src family kinases (32), thereby inhibiting enzymatic activity. Fig. 5 shows that 300 nM PP2 inhibited LRP phosphorylation (Fig. 5, lane 3). In contrast, identical amounts of the structurally related PP3 had no effect in this assay (Fig. 5, lane 4). PP3 is an appropriate negative control since it does not alter Src family kinase activity. Together, these results suggest that Src family kinase members mediate the phosphorylation of the LRP cytoplasmic domain. To verify a role for the PDGF receptor in this process, we used the PDGF receptor-specific inhibitor, AG1296 (33). AG1296 does not effect PDGF binding to the PDGF receptor nor does it alter receptor dimerization, but it does abolish receptor autophosphorylation (34). The results demonstrate that AG1296 blocks PDGF-mediated LRP phosphorylation (Fig. 5, lane 5) and confirm the requirement for the PDGF receptor in this pathway.
The Second NPXY Motif in LRP Is a Site for Src Family Kinase-mediated Phosphorylation-We next set out to identify the site on the LRP cytoplasmic domain that is phosphorylated by Src and Src family kinase members. The LRP cytoplasmic domain has two NPXY consensus motifs within its cytoplasmic domain that are potential sites for Src family kinase-mediated phosphorylation. To determine which site is preferentially phosphorylated by Src family kinases, the LRP cytoplasmic domain and two mutants were expressed as fusion proteins with GST, and in vitro phosphorylation studies employing purified Src were carried out. These studies (Fig. 6, lane 1) demonstrated that Src readily phosphorylates the LRP cytoplasmic domain. Likewise, a molecule in which the NPTY sequence within the LRP cytoplasmic domain was converted to APTA was also readily phosphorylated by Src (Fig. 6, lane 2). In contrast, mutation of the NPVY motif to APVA abolished Srccatalyzed phosphorylation of the LRP cytoplasmic domain (Fig.  6, lane 3) strongly suggesting that tyrosine 63 within the NPVY site is a Src phosphorylation site. Identical results were obtained when other Src family kinase members (Fyn, Lyn, and Lck) were utilized to phosphorylate wild-type and mutant LRP cytoplasmic domains (data not shown), confirming first that several Src family kinases are able to phosphorylate the LRP cytoplasmic tail and, second, that the site of phosphorylation is at tyrosine 63. The serine threonine kinase, protein kinase C ␣, was also found to phosphorylate the cytoplasmic domain of LRP, but in contrast to the Src family kinases, the extent of phosphorylation was not influenced by mutations at either of the two NPXY motifs (data not shown). In vitro phosphorylation studies do show selectivity for specific kinases, because calmodulin kinase was unable to phosphorylate the LRP cytoplasmic domain (data not shown). Together, these studies indicate that the NPVY sequence is the site phosphorylated by Src family kinase members.
To confirm these results in cells, LRP mini-receptors were constructed containing the entire LRP ␤ chain and a small portion of the ␣ subunit. Plasmids containing the cDNA encoding this mini-receptor or various mutant molecules were co-  5) were added, and incubation was extended for 15 min at 37°C, whereas the control plate received no treatment (lane 1). After incubation, the cells were washed, and cell extracts were subjected to immunoprecipitation (IP) with the anti-LRP monoclonal 5A6. After electrophoresis and transfer to nitrocellulose, the membranes were analyzed with anti-phosphotyrosine (P tyrosine) monoclonal antibody 4G10-HRP conjugate. For visualizing LRP, membranes were probed with 125 I-labeled 5A6 (2 g/ml) overnight, washed, and exposed to BiomaxMR film. Extracellular signal-regulated kinase activation was measured by probing cell lysates with phospho-specific polyclonal antibodies. P-erk, phospho-extracellular signal-regulated kinase.
FIG. 3. PDGF binds to LRP immobilized on a sensor chip. A, real time binding curves for 50 nM PDGFBB, basic-FGF, EGF, and IGF to LRP immobilized on CM5 chip by amine coupling. The sensorgrams were obtained at flow rate of 30 l/min and a temperature of 25°C. Flow cells with immobilized ovalbumin and without any ligand-activated chip surfaces were used as the controls for nonspecific binding and were subtracted from the presented data. B, kinetic analysis of sensorgrams for PDGFBB (0.6, 1.2, 2.5, 5, 10, and 20 nM) to immobilized LRP. The flow rate was 30 l/min, and flow cells with immobilized ovalbumin and without any ligand-activated chip surfaces were subtracted from the presented data. Curves represent the best fit to a model in which LRP contains two binding sites for PDGF, with k a1 ϭ 1.93 ϫ 10 6 M Ϫ1 s Ϫ 1; k d1 ϭ 3.2 ϫ 10 Ϫ2 s Ϫ1 ; k a2 ϭ 7.5 ϫ 10 4 M Ϫ1 s Ϫ1 ; k d2 ϭ 8.92 ϫ 10 Ϫ4 s Ϫ1 . transfected with c-Src into COS-1 cells. These cells were chosen because they are readily transfected and because we found virtually undetectable amounts of endogenous LRP that was tyrosine-phosphorylated. After transient transfection, the extent of phosphorylation was measured by immunoprecipitation and immunoblotting experiments. The results of this experiment demonstrate extensive tyrosine phosphorylation in cells co-transfected with both the wild-type receptor and c-Src (Fig.  7, lane 2). As a control, a kinase-inactive c-Src mutant was employed, and results demonstrate this mutant is unable to mediate LPR cytoplasmic tail phosphorylation (Fig. 7, lane 3). Like the results obtained from in vitro phosphorylation studies, mutations in the NPTY sequence had little effect on LRP phosphorylation (Fig. 7, lane 5). In contrast, mutation of the NPVY sequence to APVA greatly reduced LRP tyrosine phosphorylation (Fig. 7, lane 8). These results confirm the in vitro studies, indicating that tyrosine 63 within the second NPXY motif is the target for Src-catalyzed phosphorylation.
PDGF-induced Phosphorylation of LRP Increases Its Association with Shc-Several adaptor proteins containing phosphotyrosine binding (PTB) domains recognize phosphotyrosine within an NPXY consensus motif. One of these adaptor proteins, Shc, was previously shown to associate with phosphorylated LRP cytoplasmic domain in v-Src-transformed cell lines (17), and Shc is known to be involved in PDGF receptor signaling pathways. To test whether or not tyrosine phosphorylation of LRP induced by PDGF treatment increases the association of Shc with LRP, COS-1 cells were transiently transfected with Shc and then stimulated with PDGF. In COS-1 cells we noticed that the extent of LRP tyrosine phosphorylation induced by PDGF was extremely low (Fig. 8A, lanes 1 and 2). However, in the cells transfected with Shc, we noticed an increase in the extent of LRP tyrosine phosphorylation, particularly in response to PDGF treatment (Fig. 8A, lane 3). The mechanism for increased tyrosine phosphorylation upon Shc transfection is not known at this time. The membranes were stripped and probed for Shc antigen, and these results (Fig. 8A, lane 3) revealed that Shc co-immunoprecipitates with tyrosine-phosphorylated LRP and its association is enhanced with PDGF treatment. Together, the results support the notion that PDGF-induced tyrosine phosphorylation at tyrosine 63 within the LRP cytoplasmic domain generates a binding site recognized by Shc.
To confirm that the second NPXY motif within the LRP  2-5). After a 15-min incubation with PDGF, the cells were washed, and extracts were subjected to immunoprecipitation (IP) with anti-LRP 5A6 and subjected to immunoblot analysis (WB) using anti-phosphotyrosine antibody 4G10-HRP conjugate (upper panel). For visualizing LRP, membranes were probed with 125 I-labeled 5A6 (2 g/ ml) overnight, washed, and exposed to BiomaxMR film (lower panel). P]ATP for 10 min at 30°C. After terminating the reaction, phospho-labeled proteins were separated on 4 -12% SDS-PAGE precast gel and blotted on to nitrocellulose membrane. The membrane after transfer was stained with Ponseau S and exposed to BiomaxMS film. cytoplasmic domain represents a docking site for the PTB domain of Shc, COS-1 cells were co-transfected with Shc and wild-type LRP-␤ or the two mutant receptors. After transfection, the cells were treated with PDGF, and cell extracts were subjected to immunoprecipitation with anti-LRP IgG. Probing the immunoblots for anti-HA IgG (to detect Shc) revealed that Shc co-immunoprecipitated with wild-type LRP-␤ and its NPTY 3 APTA mutant (Fig. 8B, lanes 1-4) but not with the NPVY 3 APVA mutant (Fig. 8B, lanes 5 and 6). These results confirmed that mutations within the second NPXY motif in LRP abolish Shc interaction, indicating that this motif represents the docking site between these two proteins. Curiously, co-transfection of LRP-␤ and Shc alleviated the need for PDGF stimulation to phosphorylate LRP on tyrosine residues (Fig.  8B), as constitutive phosphorylation of LRP-␤ and the NPTY 3 APTA mutant were observed. In all cases, co-immunoprecipitation of Shc correlated with the phosphorylated form of LRP. DISCUSSION PDGF is an important regulator of embryological development and also plays a critical role in wound healing and in the pathogenesis of various diseases such as tumorigenesis, atherosclerosis, fibrosis, and inflammatory disorders (35). The cellular effects of PDGF are mediated by two distinct receptors, termed PDGFR-␣ and PDGFR-␤, which recognize different isoforms of PDGF. PDGF itself exists as a dimer of two homologous chains, A and B, that are disulfide-linked. All possible isoforms (AA, AB, BB) exists and are biologically active. Recently, a new family member, termed PDGFC, has been identified that is required for appropriate kidney development (36). PDGFAA only binds to PDFGR␣, whereas PDGFAB and BB bind both PDGFR-␣ and PDGFR-␤. Upon binding to its receptor, PDGF induces receptor dimerization and autophosphorylation of the cytoplasmic domain at tyrosine residues (37,38). Phosphorylated cytoplasmic domain of the PDGF receptor provides docking sites for a vast number of adaptor molecules, including Shc and tyrosine kinases such as Src, which in turn initiate several signal transduction pathways.
In the present study we demonstrate that the addition of PDGFBB to fibroblasts results in a transient phosphorylation of the LRP cytoplasmic domain at tyrosine 63, located within its second NPXY motif. PDGFBB-mediated LRP phosphorylation occurs in fibroblasts, smooth muscle cells, and COS cells. LRP tyrosine phosphorylation requires the PDGF receptor and appears to be mediated by Src family kinases. This is supported by several lines of evidence. First, a potent antagonist of PDGFR completely blocks LRP tyrosine phosphorylation. Second, inhibitors of Src family kinases block PDGF-mediated tyrosine phosphorylation of LRP. Third, in vitro assays confirm that purified Src and other Src family kinases phosphorylate purified LRP and a GST fusion protein containing the LRP cytoplasmic domain. Mutations within each of the two NPXY motifs revealed that tyrosine 63 within the second NPXY motif is the phosphorylation site for Src family kinases.  1, 4, 7) or the c-Src expression plasmids (c-Src, lanes 2,5,8) or the kinase inactive c-Src K297R (lanes 3, 6,9). After transfection, cell extracts were subjected to immunoprecipitation (IP) with anti-LRP monoclonal 5A6 and then analyzed by immunoblotting (WB) with anti-phosphotyrosyl 4G10-HRP conjugate (upper panel). Whole cell extracts were blotted for LRP (middle panel) using anti-Myc IgG or c-Src (lower panel) expression to ensure that equal amounts of these proteins were expressed.  1 and 3). After a 15-min incubation with PDGFBB at 37°C, the cells were washed, and cell extracts were subjected to immunoprecipitation (IP) with anti-LRP R2629. Immunoprecipitates were immunoblotted (IB) with anti-phosphotyrosine 4G10-HRP conjugate (upper panel) and anti-Shc IgG (lower panel). B, COS-1 cells were transfected with 5 g of HA-Shc plasmid (lanes 1-6) and 5 g of plasmid containing LRP-␤ WT (lanes 1 and 2), LRP-␤ NPTY mutant (lanes 3 and 4), or LRP-␤ NPVY mutant (lanes 5 and 6). After 24 h, the medium was removed and replaced with DMEM containing 0.1% fetal bovine calf serum, and cells were cultured for an additional 18 h. Cells were treated plus or minus PDGFBB (50 ng/ml) for 12 min. Cell extracts were prepared and subjected to immunoprecipitation with anti-LRP monoclonal 5A6 IgG (10 g), and immunoprecipitated proteins were subjected to immunoblot analysis under nonreducing conditions with anti-phosphotyrosyl 4G10-HRP conjugate (1: 4000 dilution), anti HA-Shc polyclonal IgG (0.4 g/ml), and 125 I-labeled anti-LRP monoclonal 5A6 IgG (1 g/ml). demonstrate that phosphoinositide 3-kinase is also required for LRP tyrosine phosphorylation.
c-Src and related kinases are important in a variety of key cellular functions (39) and can be activated by a large spectrum of cell surface receptors, including many growth factor receptors (40 -42). In the current investigation, we found that tyrosine phosphorylation of LRP is selective for PDGFBB. Other growth factors such as bFGF, IGF-1, and EGF were unable to induce LRP tyrosine phosphorylation. The basis for this selectivity is not known but may result from the ability of LRP to bind PDGFBB with high affinity. The PDGFBB homodimer could co-localize LRP and the PDGFR on the cell surface ( Fig.  9), thereby bringing LRP into close proximity with c-Src activated by the PDGFR. The accompanying manuscript (20) demonstrates that LRP is located in caveolea, where phosphorylated forms of the PDGF receptor are known to reside (43). These proximity effects may account for the selective phosphorylation of LRP mediated by the PDGFR. Interestingly, recent studies indicate that the PDGFR interacts closely with another membrane receptor, the EGF receptor. These two receptors appear to form heterodimers on the cell surface (44), and this interaction appears physiologically important because disruption of PDGFR/EGFR heterodimers significantly inhibits PDGF-mediated activation of extracellular signal-regulated kinases 1 and 2 (44).
Our studies indicate that LRP tyrosine phosphorylation may also have important consequences in PDGF-initiated signaling, because PDGF-induced tyrosine phosphorylation of LRP generates a docking site for Shc. Shc is an adaptor protein that contains a carboxyl-terminal Src homology 2 (SH2) domain and an amino-terminal PTB domain that is involved in signal transduction by protein-tyrosine kinases (45). The Shc SH2 domain interacts with several phosphorylated tyrosine residues on the PDGF receptor with low to moderate affinity (46), whereas the Shc PTB domain recognizes phosphorylated tyrosine in NPXY motifs. A functional Shc PTB domain is necessary for Shc tyrosine phosphorylation in v-Src-transformed cell lines (17), suggesting that association of Shc with phosphorylated LRP may be important for subsequent phosphorylation of Shc (17). Phosphorylation of Shc on Tyr-317 allows for Grb2 binding (18), thereby activating the Ras pathway, whereas phosphorylation of Shc on Tyr-239 and Tyr-240 initiates a second signaling pathway involving the induction of c-Myc (19). Thus, PDGF-mediated phosphorylation of LRP may represent a key event in downstream signaling requiring Shc.
We noticed in our experiments that the binding of PDGF to LRP was not completely inhibited by RAP. RAP (47) is an endoplasmic reticulum-associated chaperone (48) that binds with high affinity to LRP and other members of the LDL receptor superfamily and prevents ligands from associating with these receptors (25,31). RAP is known to bind to clusters of complement-type repeats on LRP (49) where most ligands associate with this receptor. The fact that RAP cannot completely prevent the binding of PDGF to LRP suggests that PDGF may interact with a region on LRP that is distinct from the clusters of complement-like repeats; one possibility might be the ␤-propeller domains present on LRP, structures that function as protein interaction domains in a variety of molecules.
LRP contains two NPXY motifs within its cytoplasmic domain. Earlier studies on the structurally related LDL receptor revealed a necessity for its single NPXY motif for coated pitmediated internalization (50). The first NPXY motif in LRP is not required for internalization because mutation of residues within this motif have little impact on endocytosis of LRP mini-receptors (51). In contrast, mutation of tyrosine 63 within the second NPXY motif or mutation of leucine 66 impaired the ability of LRP to undergo endocytosis (51), indicating an important role for this region (Y 63 ATL 66 ) in recruiting this receptor into coated pits. Tyrosine 63 is also located in the second NPXY motif in LRP, which has also been identified as a region that interacts with Dab1. Dab1 is an adaptor protein that interacts with the cytoplasmic domains of LDL receptor family members (11,12) and functions in a tyrosine kinase signaling pathway. Like Shc, Dab1 also contains a PTB domain; however, its interaction with LRP does not require phosphorylation of the tyrosine residue within the NPXY motif. Thus, this NPXYXXL motif within the LRP cytoplasmic domain is a region that regulates endocytosis and also interacts with adaptor molecules involved in signal transduction. PDGF-mediated phosphorylation of the LRP cytoplasmic domain at this site will therefore be expected to regulate both its endocytic and signaling properties.
In summary, our studies have demonstrated a PDGFBBinduced phosphorylation of the LRP cytoplasmic domain mediated by Src or Src family kinase members. The fact that other growth factor receptors known to activate Src are not capable of initiating LRP phosphorylation suggests an integrative interaction between the PDGF receptor and LRP that may significantly influence signal transduction pathways mediated by PDGF. Further work will be required to identify these pathways.