LRP1 [LDL (low-density lipoprotein) receptor-related protein 1]-null CHO cells (Chinese-hamster ovary cells) (13-5-1 cells) exhibited accelerated cell growth and severe tumour progression after they were xenografted into nude mice. Reconstitution of LRP1 expression in these cells, either with the full-length protein or with a minireceptor, reduced growth rate as well as suppressed tumour development. We tested the role of the tyrosine residue in the FXNPXY63 motif within the LRP1 cytoplasmic domain in signal transduction and cell growth inhibition by site-specific mutagenesis. The LRP1 minireceptors harbouring Tyr63 to alanine or Tyr63 to phenylalanine substitution had diametrically opposite effects on cell growth, cell morphology and tumour development in mice. The Y63F-expressing cells showed suppressed cell growth and tumour development, which were associated with decreased β-catenin and cadherin concentrations in the cells. On the other hand, the Y63A-expressing cells lacked inhibition on cell growth and tumour development, which were associated with hyperactivation of ERKs (extracellular-signal-regulated kinases), FAK (focal adhesion kinase) and cyclin D1 in the cells. The mutant Y63A minireceptor also exhibited reduced capacity in binding to the Dab2 (disabled 2) adaptor protein. In addition, the Y63A mutant showed increased caveolar localization, and cells expressing Y63A had altered caveolae architecture. However, tyrosine to alanine substitution at the other NPXY29 motif had no effect on cell growth or tumorigenesis. These results suggest that the FXNPXY63 motif of LRP1 not only governs cellular localization of the receptor but also exerts multiple functional effects on signalling pathways involved in cell growth regulation.

INTRODUCTION

LRP1 [LDLR (low-density lipoprotein receptor)-related protein 1] is a large (4525 amino acids) type I membrane protein composed of an extracellular α-chain (515 kDa) and a transmembrane β-chain (85 kDa) in the form of a non-covalently linked heterodimer [1]. The extracellular α-chain is composed of three modular structural repeats, namely class A ligand-binding (complement-type) repeats, epidermal growth factor precursor-type repeats and the β-propeller repeats [2,3]. The class A ligand-binding repeats in the α-chain are present in four clusters (designated I, II, III and IV), of which cluster II and cluster IV possess the ability to bind various structurally and functionally unrelated ligands [4]. The cytoplasmic tail of the β-chain of LRP1 contains NPXY and FXNPXY motifs, which resemble those required for receptor-mediated endocytosis in other LDLR family members [5,6]. In LRP1, the FXNPXY motif binds to endocytic adaptor proteins, including AP2 (adaptor protein 2), ARH (autosomal recessive hypercholesterolaemia) and Dab2 (disabled 2) [5,79], and mediates endocytosis through clathrin-coated pits.

In addition to its endocytic function, LRP1 has been increasingly recognized to play a role in cell growth regulation through signal transduction (for a review, see [10]). Thus many protein factors previously identified as ligands for LRP1-mediated endocytosis, such as urokinase plasminogen activator, plasminogen activator inhibitor 1, MMPs (matrix metalloproteinases), ApoE (apolipoprotein E), lactoferrin and TGF-β (transforming growth factor-β), also play roles in regulating cell migration, contraction and adhesion. Often, binding of these ligands to their respective cognate receptors requires LRP1 as a co-receptor in the same cell. In smooth-muscle cells, LRP1 may act as a co-receptor with PDGF (platelet-derived growth factor) receptor in mediating the PDGF-triggered signalling pathway [11]. Mice harbouring tissue-specific knockout of LRP1 in smooth-muscle cells developed severe atherosclerosis in the LDLR−/− background and exhibited profound up-regulation of PDGF receptor in the LRP1-deficient cells [11]. Similar co-receptor functions have been suggested for LRP1 in the Wnt signalling mediated through the Frizzled/LRP5/6 system [12] and in the TGF-β signalling cascade [13].

The endocytosis and signalling events of an endocytic receptor can be functionally interconnected [5]. The FXNPXY motif of LRP1 not only binds to endocytic APs but also serves as a binding site for signal transduction adaptor proteins, such as Dab1, SH3 domain (Src homology 3 domain) homology protein and insulin receptor substrate [6,8,14,15]. Thus it is of interest to determine if the FXNPXY motif within the LRP1 cytoplasmic domain plays a dual role in receptor-mediated endocytosis and in signal transduction. In the present study, we tested the role of tyrosine residues of the two NPXY motifs within LRP1 in signal transduction and cell growth using site-specific mutagenesis. Our results showed that replacement of tyrosine with alanine or phenylalanine at the FXNPXY63 motif of LRP1 had opposite effects on cell growth inhibition and suggested the involvement of this motif in multiple signal transduction pathways.

EXPERIMENTAL

Construction and expression of LRP1 minireceptors

The expression plasmid encoding the full-length human LRP1 (pcDNA3.1-hLRP1) (a gift from Dr Dudley K. Strickland) was used as a template to generate the minireceptor LRP1-IV that contained the cluster IV ligand-binding repeats. Two cDNA fragments were PCR-amplified with the Pfu Turbo polymerase; they encoded (from 5′ to 3′) LRP1 signal peptide, N-terminal five amino acids (AIDAP) of the mature LRP1 (GenBank® accession no. X13916), HA (haemagglutinin) epitope (YPYDVPDYA), a three-amino-acid [GAP (Gly-Ala-Pro)] linker and the cluster IV ligand-binding repeats. Primer pair used to amplify the XhoI site, the LRP1 signal peptide sequence and the HA tag–AscI restriction site were 5′-TAATACGACTCACTATAGGG (forward) and 5′-GGGCGCGCCAGCGTAATCCGGAACATCGTATGGGTAAGGGGCGTCGATAGCCGCCGCGAC (reverse). The primer pair used to amplify the cluster IV ligand-binding repeats and the AscI and NotI restriction sites were 5′-GGCGCGCCCTCCAACTGCACGGCTAGCCAG (forward) and 5′-AAAAGGAAAAGCGGCCGCCTATGCCAAGGGGTCCCCTATC (reverse). The two amplified cDNA fragments were inserted into the pGEM-T vector, sequenced and subcloned into pcDNA3.1 at the multiple cloning sites XhoI and NotI. The resulting construct was designated LRP1-IVwt (or IVwt). The Tyr29 to alanine, Tyr63 to alanine or Tyr63 to phenylalanine substitution at the two NPXY motifs was introduced using the QuikChange® mutagenesis approach (Stratagene, La Jolla, CA, U.S.A.) to create Y29A, Y63A and Y63F variants (numbering starts at the first amino acid immediately adjacent to the transmembrane domain) [16,17]. The sequences of the mutagenesis primers for Y29A were: 5′-GATTGGAAACCCCACCGCGAAGATGTACGAAGGC (forward) and 5′-GCCTTCGTACATCTTCGCGGTGGGGTTTCCAATC (reverse), primers for Y63A were: 5′-CTTCACCAACCCCGTGGCCGCCACACTCTACATG (forward) and 5′-CATGTAGAGTGTGGCGGCCACGGGGTTGGTGAAG (reverse), and primers for Y63F were 5′-CTTCACCAACCCCGTGTTTGCCACACTCTACATG (forward) and 5′-CATGTAGAGTGTGGCAAACACGGGGTTGGTGAAG (reverse). The LRP1-IV minireceptors were stably expressed in LRP-null CHO (Chinese-hamster ovary) cells (13-5-1 line) [18] by transfection with FuGENE™ 6 reagent (Roche, Laval, QC, Canada). Control cells were transfected with pcDNA3.1 vector. Multiple stable clones resistant to G418 (700 μM) were selected, and those showing comparable levels of LRP1-IV expression were maintained in the presence of 500 μM G418 for further analysis.

Cell growth and the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide] assay

Cell growth was determined at 24 h intervals after initial seeding of 104 cells per 60 mm dish in triplicate [19]. The number of viable cells was determined by the MTT assay that measures the reduction of a tetrazolium compound MTT into an insoluble formazan product by mitochondria. The amount of colour produced is directly proportional to the number of viable cells [20]. The MTT assay was performed using the CellTiter 96 Assay kit (Promega, Madison, WI, U.S.A.) and the Vi-CELL cell viability analyser (Beckman Coulter, Fullerton, CA, U.S.A.) according to the manufacturer's instructions.

Soft-agar assay

Anchorage-independent growth was determined by quantifying colony formation in soft agar [19]. Cells were suspended and plated at 2×103 cells/well of a 96-well plate on soft agar. After colony formation, the soft agar was solubilized and cells were lysed and mixed with CyQuant working solution (Cell Biolabs, San Diego, CA, U.S.A.). The number of cells was detected and quantified by CyQuant GR dye at 480 nm/520 nm with a fluorescence plate reader according to the manufacturer's instructions.

Cell migration assay

Cells were cultured in serum-free media for 16 h prior to seeding at 5×105 cells/ml (300 μl) into a Boyden chamber with 8.0 μm pore membranes (Chemicon, Billerica, MA, U.S.A.) that were coated with fibronectin or collagen I [21]. Control membranes were coated with BSA. The membranes were washed with serum-free F12 medium containing 2 mM CaCl2 and 2 mM MgCl2. Cell migration was allowed to proceed for 4 h with fibronectin-coated membranes or for 5 h with collagen I-coated membranes at 37 °C. Both the upper and lower chambers contained serum-free medium. At the end of the incubation period, non-migrating cells were removed from the upper surface of the membranes with cotton swabs. The membranes were fixed and stained with 0.1% (w/v) Crystal Violet. The dye was eluted with 10% (v/v) acetic acid, and the attenuance (D) at 595 nm was determined.

Tumour xenografts

Female Balb/c nude mice (3–4 weeks of age) were purchased from Charles River and housed under specific pathogen-free facilities approved by the Canadian Association for Accreditation of Laboratory Animal Care. Cells expressing LRP1-IV minireceptors were grown to subconfluence, briefly trypsinized with 0.25% (w/v) trypsin in EDTA, washed with serum-free medium and resuspended in PBS. A total of 2.5×106 cells in 150 μl of PBS were injected subcutaneously into the lower-dorsal region of the mice. Tumour development was measured in two dimensions (length and width) with a digital calliper, and tumour volume was calculated using the formula: (π/6)×width2×length [19]. The mice with tumour appearance were killed at day 16 post-injection, and tumours were excised and fixed in 10% (v/v) formalin and embedded in paraffin. Mice that showed no tumour appearance after 32 days were considered negative.

Immunohistology of tumour

Five micrometre slices of paraffin-embedded tumour samples were deparaffinized with xylene, dehydrated in alcohol and rehydrated in PBS. Endogenous peroxidase was blocked with 3% (w/v) H2O2 in PBS. The specimens were blocked with 5% (w/v) BSA in PBS and incubated overnight at 4 °C with an anti-LRP1 β-chain antibody. After 1 h incubation at room temperature with peroxidase-conjugated secondary anti-rabbit antibody, positive reaction was detected by exposure to stable 3,3′-diaminobenzidine tetrahydrochloride (Sigma, Oakville, ON, Canada). Slips were counterstained with Shandon Instant Hematoxylin. Sections stained for immunoperoxidase or H/E (haematoxylin and eosin) were examined using the Nikon Eclipse 80i microscope equipped with a QIMAGING MicroPublisher 5.0 RTV camera. Digital images were captured using Image Pro Plus version 5 software.

Immunocytochemistry

For detection of LRP1 cell surface presentation, cells were grown on fibronectin-coated coverslips for 24 h, washed three times with ice-cold PBS, and fixed on ice with 4% (w/v) paraformaldehyde for 30 min. The fixed cells were washed three times with 100 mM glycine, blocked with 1% BSA for 1 h, incubated with a monoclonal anti-HA antibody and subsequently with goat anti-mouse IgG conjugated with Alexa Fluor® 488 (Invitrogen, Burlington, ON, Canada). For detection of intracellular F-actin, the fixed cells were treated with Triton X-100 [0.1% (v/v) in PBS] for 5 min at room temperature (20 °C). The permeabilized cells were incubated in blocking solution [10% (w/v) FBS (fetal bovine serum) in PBS] for 20 min, and stained with Alexa Fluor® 660 phalloidin (2.5%, w/v) (Invitrogen) in blocking solution for 1 h at room temperature. For caveolin-1 staining, the fixed and permeabilized cells were blocked with FBS (10% in PBS) for 1 h, and incubated with anti-caveolin-1 polyclonal antibody and subsequently with goat anti-mouse IgG conjugated with Alexa Fluor® 599 (Invitrogen). The coverslips were mounted on to a glass slip using SlowFade Light AntiFade (Invitrogen), and the images were captured using OLYMPUS FluoView 1000 and the MellesGriot laser scanning confocal imaging system.

Immunoblot analysis

Cells were collected into ice-cold PBS containing 2 mM NaF/2 mM Na3VO4 and lysed with a lysis buffer containing 50 mM Tris (pH 8.0), 150 mM NaCl, 1% (v/v) Triton X-100, 2 mM MgCl2, 2 mM EGTA, 2 mM EDTA, 2 mM glycerophosphate, 2 mM NaF, 2 mM Na3VO4 and a cocktail of protease inhibitors. The cell lysate was separated by PAGE in the presence of SDS (SDS/PAGE) and the resolved proteins were transferred on to 0.2 μm pore nitrocellulose membranes (Bio-Rad, Hercules, CA, U.S.A.). The membranes were incubated with appropriate primary antibodies for 16 h at 4 °C and secondary antibody for 2 h at room temperature. Primary antibodies used were: anti-pY 4G10 (Upstate, Billerica, MA, U.S.A.), polyclonal anti-LRP1 β-chain [22], monoclonal anti-HA (BabCO, Richmond, CA, U.S.A.), anti-FAK (focal adhesion kinase), anti-phospho-FAK (Tyr397), anti-ERK (extracellular-signal-regulated kinase) [p44/42 MAPK (mitogen-activated protein kinase)], anti-phospho-ERK (Thr202/Tyr204) anti-cadherin, anti-cyclin D1 (Santa Cruz Biotechnology, Santa Cruz, CA, U.S.A.), anti-caveolin-1 (C13630) and anti-p96 (Dab2) (BD Transduction Laboratories, Mississauga, ON, Canada), goat anti-human/rat β-catenin (R&D Systems, Minneapolis, MN, U.S.A.) and monoclonal anti-HSP60 (Sigma). The immunocomplexes were visualized by the BM chemiluminescence method (Roche).

Co-immunoprecipitation

Cell lysates were precleared with preimmune IgG and Protein A beads. The lysate was adjusted to a final concentration of 20% (w/v) sucrose and incubated with primary antibody for 1 h and Protein A beads (eBioscience, San Diego, CA, U.S.A.) for another 1 h at 4 °C. The beads were washed five times with the lysis buffer. The co-immunoprecipitated proteins were eluted with reducing SDS sample buffer [25 mM Tris/HCl, pH 8.0, 150 mM NaCl, 10% (w/v) SDS, 50 mM DTT (dithiothreitol) and 40 mM 2-mercaptoethanol], separated by SDS/PAGE (8% gel) and detected using TrueBlot (eBioscience). Anti-Dab2 polyclonal antibody used in the co-immunoprecipitation experiments was a gift from Dr Xiang-Xi Xu (Department of Medical Oncology, Fox Chase Cancer Center, Philadelphia, PA, U.S.A.) or obtained from Santa Cruz Biotechnology (sc-13982).

Kinase expression profiling

Kinase expression was profiled by Western-blot analysis, a service provided by Kinexus (Vancouver, BC, Canada).

Other assays

RAP (receptor-associated protein)-specific binding to cell surface LRP1 was performed using 125I-RAP (5.0 nM) at 4 °C in the absence or presence of 50-fold excess of unlabelled RAP as described previously [23,24]. Isolation of membrane rafts was performed using the detergent (Triton X-100) solubilization method as described previously [22]. Cholesterol oxidase (0.5 unit/ml) treatment was conducted at 37 °C for 60 min as described previously [25].

RESULTS

Restoration of LRP1 expression in LRP1-null CHO cells reduced cell growth and tumorigenesis

The LRP1-null CHO cells, designated 13-5-1, have been used as a model system for LRP1 structure–function relationship studies [13,23,26,27]. We noted that the LRP1-null cells exhibited accelerated growth rate as compared with parental CHO-K1 cells (results not shown). It has been shown previously by several clinical studies that LRP1 expression is either aberrant or entirely lost in several cancers [2832]. Since these studies suggest a potential role of LRP1 in regulating cell growth, we thus first tested the effect of restoration of LRP1 expression on cell growth using two LRP1 constructs, namely the full-length human LRP1 (hLRP) and a minireceptor termed LRP1-IV (IVwt) that contains cluster IV of the ligand-binding repeats (see the top two lines in Figure 1A). The LRP1-IV minireceptor has also been used for LRP1 structure–function studies and is shown to possess normal endocytic functions for many LRP1 ligands [17,22,33]. Stable transfection of 13-5-1 with either hLRP or IVwt resulted in normal expression and post-translational cleavage into the α- and β-chains, and the level of transgene expression was comparable with that in 3T3-L1 and HEK-293 cells (human embryonic kidney cells) (bottom panels in Figure 1A). Noticeably, F-actin staining showed that morphological changes occurred from fibroblastoid spindle shape in vector-transfected control cells to squamous shape in hLRP- or IVwt-expressing cells (Figure 1B). Moreover, expression of hLRP or IVwt resulted in markedly reduced stress fibre formation. Similar morphological changes in LRP1-null cells after LRP transfection have been observed previously [13]. Stable expression of hLRP and IVwt resulted in decreased cell growth by 40 and 30% respectively as compared with cells transfected with vector alone (Figure 1C). Likewise, anchorage-independent colony formation in soft-agar was decreased by 50 and 40% respectively (Figure 1D). Expression of hLRP or IVwt also markedly inhibited fibronectin-induced cell migration as compared with vector-transfected control cells (Figure 1E). [These cells showed limited migration on type I collagen-coated substratum (Figure 1E).] These results provide the first indication that restoration of LRP1 expression in the LRP1-null cells, either with the full-length protein or a minireceptor, resulted in cell growth inhibition.

Restoration of LRP1 expression reduced cell growth

Figure 1
Restoration of LRP1 expression reduced cell growth

(A) Schematic representation of human LRP1 (hLRP) and the wild-type LRP1-IV minireceptor (IVwt) (top panel). Clusters of ligand-binding repeats are presented as horizontal rectangles and marked I, II, III and IV respectively. The empty vertical bar denotes the transmembrane domain. The arrow denotes the furin cleavage site that gives rise to α- and β-chains. The expressed hLRP and IVwt in 13-5-1 (LRP-null) cells was detected by immunoblot analysis using appropriate antibodies as indicated (middle panels). Samples derived from equal cell protein of 3T3-L1 or HEK-293 cells were used to compare LRP1 expression (bottom panel). (B) Cells were stained with Alexa Fluor® 660 phalloidin to visualize F-actin. (C) Cell growth rate. **P<0.01 (hLRP and IVwt compared with vector control, n=3). (D) Anchorage-independent cell growth. Cells were grown on soft agar for 14 days and quantified using the CyQuant dye by fluorescence plate reader. RFU, relative fluorescence units. Results are derived from two independent experiments performed in triplicate. ***P<0.001 (hLRP and IVwt compared with vector control). (E) Cell migration assay using a Boyden chamber with either collagen I or fibronectin as substratum. ***P<0.001 (hLRP and IVwt compared with vector control).

Figure 1
Restoration of LRP1 expression reduced cell growth

(A) Schematic representation of human LRP1 (hLRP) and the wild-type LRP1-IV minireceptor (IVwt) (top panel). Clusters of ligand-binding repeats are presented as horizontal rectangles and marked I, II, III and IV respectively. The empty vertical bar denotes the transmembrane domain. The arrow denotes the furin cleavage site that gives rise to α- and β-chains. The expressed hLRP and IVwt in 13-5-1 (LRP-null) cells was detected by immunoblot analysis using appropriate antibodies as indicated (middle panels). Samples derived from equal cell protein of 3T3-L1 or HEK-293 cells were used to compare LRP1 expression (bottom panel). (B) Cells were stained with Alexa Fluor® 660 phalloidin to visualize F-actin. (C) Cell growth rate. **P<0.01 (hLRP and IVwt compared with vector control, n=3). (D) Anchorage-independent cell growth. Cells were grown on soft agar for 14 days and quantified using the CyQuant dye by fluorescence plate reader. RFU, relative fluorescence units. Results are derived from two independent experiments performed in triplicate. ***P<0.001 (hLRP and IVwt compared with vector control). (E) Cell migration assay using a Boyden chamber with either collagen I or fibronectin as substratum. ***P<0.001 (hLRP and IVwt compared with vector control).

It has been observed previously that expression of full-length LRP1 in LRP1-null cells attenuated tumorigenicity in nude mice (J.S. Huang, personal communication). In the present studies, we confirmed that subcutaneous xenografting of LRP1-null cells indeed resulted in severe tumour development in nude mice (top panel in Figure 2A). All mice started to develop tumours at day 7 after injection, and the tumour volume progressed at the rate of ∼40mm3/day to day 16 (Figure 2B). Expression of IVwt, as was observed for the full-length LRP1, resulted in significantly reduced tumour development (bottom panel in Figure 2A). Only half of the mice injected with IVwt cells developed tumours, and the tumours grew at a low rate of ∼10mm3/day (Figure 2B). At day 16, the tumour volume in mice injected with vector-transfected control cells was 751.6±161.6mm3, whereas, in mice injected with IVwt cells, it was 233.0±116.6mm3 (P<0.05) (compare top and bottom panels of Figure 2A). Mice injected with untransfected parental LRP-null (13-5-1) cells exhibited the same tumour growth rate as those injected with vector-transfected control cells (results not shown). Immunostaining with anti-LRP1 β-chain antibody showed clear signals of the transgene product in tumour specimens obtained from mice injected with IVwt cells (Figure 2C). Histological analysis with H/E staining did not reveal significant differences in the phenotypes (e.g. area of necrosis) of tumour cells between samples obtained from the two groups of mice (results not shown). These in vivo results, consistent with the reported growth inhibition by LRP1 expression [13,34], indicate that the LRP1-IV minireceptor possesses functional elements essential for the inhibition of cell growth and tumour development.

In vivo tumorigenesis assay

Figure 2
In vivo tumorigenesis assay

(A) Subcutaneous tumour developed in nude mice at day 16 post-injection. Individual tumour volume was measured at the indicated days after injection of vector-transfected control cells or IVwt cells (2.5×106 cells per injection). Average volume of tumours (means±S.E.M.) is shown at the indicated days post injection. Between two groups of mice, *P<0.05 at days 11, 14 and 16. (B) The rates of tumour growth between the two groups of mice. *P<0.05 (IVwt compared with vector control). (C) Immunohistochemical analysis of IVwt expression with tumour biopsy samples 16 days after injection.

Figure 2
In vivo tumorigenesis assay

(A) Subcutaneous tumour developed in nude mice at day 16 post-injection. Individual tumour volume was measured at the indicated days after injection of vector-transfected control cells or IVwt cells (2.5×106 cells per injection). Average volume of tumours (means±S.E.M.) is shown at the indicated days post injection. Between two groups of mice, *P<0.05 at days 11, 14 and 16. (B) The rates of tumour growth between the two groups of mice. *P<0.05 (IVwt compared with vector control). (C) Immunohistochemical analysis of IVwt expression with tumour biopsy samples 16 days after injection.

Tyr63 to alanine substitution abolished the ability of LRP1-IV to inhibit cell growth

The cell growth inhibition by IVwt expression suggests that the C-terminal portion of LRP1 contains the information essential for this function. We postulated that the cell growth inhibition function might be linked to one of the two NPxY motifs residing within the cytoplasmic domain of LRP1. To test this hypothesis, we performed site-specific mutagenesis at tyrosine residues (i.e. the putative phosphorylation site) to introduce Tyr29 to alanine, Tyr63 to alanine and Tyr63 to phenylalanine substitutions (top lines in Figure 3A). These constructs were stably transfected into 13-5-1 cells, and multiple clones that expressed comparable levels of LRP1-IV variants were selected (representative clones are shown in the bottom panels of Figure 3A). Immunocytochemistry analysis with non-permeabilized cells showed that the LRP1-IV variants were present on the cell surface, with Y63A-expressing cells having noticeably higher cell surface presentation than other cells (Figure 3B). Increased cell surface presentation of a similar Y63A mutant form of LRP1-IV has been observed previously [17]. Ligand binding experiments with the 125I-RAP confirmed that the cell-surface-expressed LRP1-IV variants were functional (Figure 3C). Pulse (30 min)–chase (up to 2 h) experiments with [35S]methionine showed that all of the LRP1-IV variants were processed normally (i.e. from the precursor endoplasmic reticulum form into the α- and β-chains) (Figure 3D). There was no noticeable difference in their turnover rate among these variants during chase. Results from these kinetic studies ruled out the possibility that replacing the amino acid tyrosine with alanine or phenylalanine residues causes protein misfolding.

Loss of cell growth inhibition in cells expressing a Y63A mutant

Figure 3
Loss of cell growth inhibition in cells expressing a Y63A mutant

(A) Schematic representation of tyrosine to alanine or tyrosine to phenylalanine mutants (top panel). The expressed LRP1-IV variants were detected by immunoblot analysis (bottom panels) as in Figure 1(A). (B) Immunocytochemistry of cell surface LRP1-IV variants using monoclonal anti-HA antibody under non-permeabilizing conditions. (C) Specific binding of 125I-RAP at 4 °C to cells expressing different LRP1-IV variants. (D) Pulse–chase analysis of post-translational processing of LRP1-IV variants. (E) Cell growth rates. **P<0.01 (IVwt compared with vector control, n=3); ***P<0.001 (Y63F compared with vector control, n=3). (F) Anchorage-independent cell growth. ***P<0.001 (IVwt, Y29A and Y63F compared with vector control, n=6).

Figure 3
Loss of cell growth inhibition in cells expressing a Y63A mutant

(A) Schematic representation of tyrosine to alanine or tyrosine to phenylalanine mutants (top panel). The expressed LRP1-IV variants were detected by immunoblot analysis (bottom panels) as in Figure 1(A). (B) Immunocytochemistry of cell surface LRP1-IV variants using monoclonal anti-HA antibody under non-permeabilizing conditions. (C) Specific binding of 125I-RAP at 4 °C to cells expressing different LRP1-IV variants. (D) Pulse–chase analysis of post-translational processing of LRP1-IV variants. (E) Cell growth rates. **P<0.01 (IVwt compared with vector control, n=3); ***P<0.001 (Y63F compared with vector control, n=3). (F) Anchorage-independent cell growth. ***P<0.001 (IVwt, Y29A and Y63F compared with vector control, n=6).

However, unlike that of IVwt, expression of Y63A in the LRP1-null cells failed to exert an inhibitory effect on cell growth or anchorage-independent colony formation. Thus Y63A cells exhibited a growth rate that was similar to the vector-transfected control cells but ∼40 and 100% higher than IVwt and Y63F cells, respectively (Figure 3E). Likewise, the formation of Y63A cell colonies on soft-agar was similar to vector control, but was ∼35% higher than IVwt, Y29A or Y63F cells (Figure 3F). Enhanced cell growth and anchorage-independent colony formation were observed in multiple clones of Y63A cells, ruling out the possibility that abnormal cell growth was attributable to clonal variation. These results suggest that Tyr63 to alanine mutation, but not Tyr63 to phenylalanine or Tyr29 to alanine mutation, severely compromises LRP1's ability to inhibit cell growth.

The loss of growth inhibition by Tyr63 to alanine mutation was further exemplified by in vivo analysis of the progression of tumorigenesis in xenografted nude mice. Stably transfected cell lines expressing comparable levels of the respective LRP1 minireceptors were injected into the mice at two different doses. At low dose (0.5×106 cells), the incidence of tumour development was 66% (4 out of 6 mice) for vector transfected control (LRP-null), 14% (1 out of 7) for IVwt, 14% (1 out of 7) for Y29A, 17% (1 out of 6) for Y63F and 33% (2 out of 5) for Y63A. At high dose (2.5×106 cells), all of the mice (8 out of 8) injected with LRP-null cells developed tumours within 5–7 days post-injection. The incidence of tumour development was reduced to 56% (5 out of 9) in IVwt-injected mice (P<0.001 compared with LRP-null). The incidence of tumour development in Y63A-injected mice was 89% (8 out of 9), almost as high as in mice injected with LRP-null (0.05<P<0.1 compared with LRP-null). Mice injected with Y29A or Y63F cells also exhibited low frequency in tumour development (78% for Y29A and 75% for Y63F) as compared with LRP-null (**P<0.01). In addition to differences in the frequency of tumour development, there were also marked differences in the rate of tumour growth (Figure 4B). Mice injected with Y63A cells exhibited fast tumour growth rate (∼30mm3/day) that was similar to that observed in mice injected with LRP1-null (i.e. vector-injected control) (top panel in Figure 4A). In contrast, mice injected with Y63F or Y29A cells displayed lowered tumour growth rates (∼10mm3/day) that resembled the growth rate observed in mice injected with IVwt cells (middle and bottom panels in Figure 4A). Immunostaining of Y63A cells present in the tumour specimens showed more condensed Y63A signals in the cells as compared with that in Y29A, Y63F, (Figure 3C) and IVwt cells (Figure 2C). Together, mutational analysis of the FXNPXY motifs with cell culture and nude mice models suggests a functional domain that is critical for receptor function in cell growth inhibition. Moreover, retention of normal cell growth inhibition in the Y63F mutant suggests that the presence of the aromatic tyrosine side chain, not a potential phosphorylation site, is essential for FXNPXY function.

In vivo tumorigenesis assay

Figure 4
In vivo tumorigenesis assay

(A) Subcutaneous tumour developed in nude mice. The experiment was performed as in Figure 2, except that cells expressing different LRP1-IV variants were used. Average volume of tumours (means±S.E.M.) is shown at the indicated days post injection. *P<0.05 (Y29A compared with Y63A at day 16); **P<0.01 (Y63F compared with Y63A at day 16). (B) Rates of tumour growth. Values for vector-transfected control cells and IVwt cells shown in Figure 2(B) are included here for comparison purposes. The rate of tumour growth for Y63A cells is not statistically different from vector-transfected control cells. *P<0.05 (IVwt and Y29A compared with vector control); **P<0.01 (Y63F compared with vector control). (C) Immunohistochemical analysis of expression of LRP1-IV variants within tumour biopsy samples at day 16 post-injection.

Figure 4
In vivo tumorigenesis assay

(A) Subcutaneous tumour developed in nude mice. The experiment was performed as in Figure 2, except that cells expressing different LRP1-IV variants were used. Average volume of tumours (means±S.E.M.) is shown at the indicated days post injection. *P<0.05 (Y29A compared with Y63A at day 16); **P<0.01 (Y63F compared with Y63A at day 16). (B) Rates of tumour growth. Values for vector-transfected control cells and IVwt cells shown in Figure 2(B) are included here for comparison purposes. The rate of tumour growth for Y63A cells is not statistically different from vector-transfected control cells. *P<0.05 (IVwt and Y29A compared with vector control); **P<0.01 (Y63F compared with vector control). (C) Immunohistochemical analysis of expression of LRP1-IV variants within tumour biopsy samples at day 16 post-injection.

Hyperactivation of the MEK (MAPK/ERK kinase)/ERK pathway by Y63A expression

To gain insights into mechanisms responsible for the link between the FXNPXY motif and receptor function in cell growth, we profiled the expression and activation pattern of protein kinases. We contrasted the IVwt and Y63A cells that showed inhibition and lack-of-inhibition respectively of cell growth. Of 75 kinases examined, 11 showed increases and two showed decreases (by ≥40%) in Y63A cells as compared with IVwt cells. The kinases involved in the MAPK pathway [ERK1, MEK2 and JNKs (c-Jun N-terminal kinases)] were mostly up-regulated in Y63A cells [Figure 5A(a)], whereas, the kinases involved in the Wnt signalling pathway [e.g. protein kinase CK2 and GSK3 (glycogen synthase kinase 3)] were down-regulated [Figure 5A(b)]. Down-regulation of protein kinase CK2 and GSK3 probably resulted in increased stability of β-catenin in Y63A cells (see below). Kinases involved in the PKB (protein kinase B) pathway and other receptor tyrosine kinases {e.g. FAK, Lyn [a Src family tyrosine kinase], Csk (C-terminal Src kinase) and Src} were variably altered by Y63A expression [Figure 5A(c)]. Kinases involved in the PKC (protein kinase C)/PKD (protein kinase D) pathway, particularly PKCα and CaMK1D (Ca2+/calmodulin-dependent protein kinase 1D), were up-regulated [Figure 5A(d)]. These kinase profiling results provide the first indication that expression of IVwt, not the Y63A mutant, may suppress activation of several major signalling pathways in the cells.

Hyperactivation of the ERK pathway in Y63A-expressing cells

Figure 5
Hyperactivation of the ERK pathway in Y63A-expressing cells

(A) Comparative analysis of kinase expression between Y63A cells and cells expressing IVwt. Kinases are grouped into one of the four pathways: (a) ERK, (b) Wnt, (c) FAK and other kinases and (d) PKC/PKD. Results are presented relative to cells expressing IVwt. (B) Immunoblot analysis of ERK and phospho-ERK (p44/42). Shown on top of the blots are relative intensities of pERK bands with respect to IVwt that were semi-quantified by scanning densitometry. (C) Immunoblot analysis of phospho-ERK and phospho-cyclin D1 (pCD1). Cells were serum-starved for 18 h and treated with (+) or without (–) 10% FBS for 30 min. Shown on top of the blots are relative intensities of pCD1 with respect to IVwt cells that were treated with serum. (D) Colony formation of Y63A cells treated with PD98059 (+PD) or LY290042 (+LY). Number of PD- or LY-treated cells are presented as a percentage of the number of cells incubated with the control growth medium (GM). Immunoblot analysis demonstrating the effect of PD98059 treatment on ERK phosphorylation is shown in the inset.

Figure 5
Hyperactivation of the ERK pathway in Y63A-expressing cells

(A) Comparative analysis of kinase expression between Y63A cells and cells expressing IVwt. Kinases are grouped into one of the four pathways: (a) ERK, (b) Wnt, (c) FAK and other kinases and (d) PKC/PKD. Results are presented relative to cells expressing IVwt. (B) Immunoblot analysis of ERK and phospho-ERK (p44/42). Shown on top of the blots are relative intensities of pERK bands with respect to IVwt that were semi-quantified by scanning densitometry. (C) Immunoblot analysis of phospho-ERK and phospho-cyclin D1 (pCD1). Cells were serum-starved for 18 h and treated with (+) or without (–) 10% FBS for 30 min. Shown on top of the blots are relative intensities of pCD1 with respect to IVwt cells that were treated with serum. (D) Colony formation of Y63A cells treated with PD98059 (+PD) or LY290042 (+LY). Number of PD- or LY-treated cells are presented as a percentage of the number of cells incubated with the control growth medium (GM). Immunoblot analysis demonstrating the effect of PD98059 treatment on ERK phosphorylation is shown in the inset.

Since aberrant activation of the ERK pathway is involved in accelerated cells growth and transformation, we examined further the effect of FXNPXY mutations on ERK activation. Constitutive ERK activation was observed in Y63A and vector-transfected control cells as compared with that in IVwt and Y63F cells when serum was present in the media (Figure 5B), indicative of hyperactivation of ERK in the absence of functional LRP1. In another experiment where the cells were serum-deprived for 18 h and then cultured in the media±FBS for 30 min, robust ERK phosphorylation in response to serum treatment occurred in IVwt and Y63A cells but not in Y63F cells (top panel in Figure 5C). Under the same conditions, phosphorylation of cyclin D1, which is subjected to regulation of ERK and Wnt/β-catenin signalling pathways [35], was slightly activated in Y63A cells under serum-deprived conditions (as compared with IVwt cells) and was further activated by serum supplementation. However, in Y63F cells, cyclin D1 phosphorylation was increased under serum-free conditions (as compared with IVwt cells) yet its response to serum was entirely absent (middle panel in Figure 5C). The lack of an effect of serum treatment on cyclin D1 phosphorylation in Y63F cells corroborates further that cell growth inhibition by LRP1 expression is related to blunted ERK activation.

To identify factors responsible for enhanced anchorage-independent colony formation and constitutive activation of ERK in Y63A cells, we determined the effect of treatment with the MEK inhibitor PD98059. PD98059 treatment completely inhibited colony formation (Figure 5D) and diminished ERK phosphorylation in Y63A cells (inset in Figure 5D). Colony formation of Y63A-expressing cells was only partially inhibited by the PI3K (phosphoinositide 3-kinase) inhibitor LY290042 (Figure 5D). Treating the cells with the p38 MAPK inhibitor SB202190 had no effect on ERK phosphorylation (results not shown). Thus upstream regulation of hyperactivation of ERK in Y63A cells was probably mediated by MEK. These results suggest that the loss of cell growth inhibition in Y63A cells was associated with constitutive activation of the MEK→ERK cascade.

Opposing effects of Y63A and Y63F on β-catenin levels

It has been shown that deregulation of both the Wnt/β-catenin and ERK pathways can lead to cell transformation [36,37]. We thus examined the level of β-catenin and its phosphorylation status in cells expressing LRP1-IV variants. The level of β-catenin was slightly increased in Y63A cells compared with that in IVwt cells. On the other hand, β-catenin levels were markedly decreased in multiple stable Y63F clones and this was accompanied by decreased cadherin and FAK (as well as phosphorylated FAK) (data derived from a representative Y63F stable clone are shown in Figure 6A). The decreased β-catenin in Y63F cells was not merely attributable to enhanced proteasome-mediated degradation, because treating cells with the proteasome inhibitor MG-132 (the proteasome inhibitor carbobenzoxy-L-leucyl-L-leucyl-leucinal) did not equalize the β-catenin levels between Y63F cells and other cells (top panel in Figure 6B). The level of phosphorylated Ser37/Thr41 β-catenin, the substrate of proteasomal degradation, was similar among all cells in the presence of MG-132 (bottom panel in Figure 6B). Immunocytochemistry analysis showed prominent staining of β-catenin at the cell–cell boundaries in all cells except in Y63F cells where plasma membrane localization of β-catenin was markedly decreased (Figure 6C). The diminished association of β-catenin with plasma membrane might be the result of lowered cadherin levels in these cells. A striking contrast in morphology and stress fibre formation was observed between Y63A and Y63F cells (Figure 6D), the former resembled vector-transfected control cells and the latter resembled IVwt cells (see Figure 1B). These results suggest that expression of Y63A and Y63F variants exerted opposing effects on plasma-membrane-associated proteins that are important for cytoskeleton arrangement, focal adhesion and extracellular matrix formation.

Opposing effects of Y63A and Y63F on β-catenin and other plasma-membrane-associated proteins

Figure 6
Opposing effects of Y63A and Y63F on β-catenin and other plasma-membrane-associated proteins

(A) Immunoblot analysis of β-catenin, cadherin, phospho-FAK [anti-phospho-FAK (Tyr397)] and FAK. (B) Immunoblot analysis of β-catenin and phospho-β-catenin in cells treated with (+) or without (–) MG-132. (C) Immunocytochemical analysis of β-catenin (green). Nucleus was stained with DAPI (4′,6-diamidino-2-phenylindole) (blue). (D) Cells were stained with Alexa Fluor® 660 phalloidin to visualize F-actin. (E) Co-immunoprecipitation of Dab2 and LRP1 β-chain. An anti-Dab2 polyclonal antibody was used to pull down complexes and the precipitates were probed for LRP1 β-chain (top panel) and Dab2 p96 (bottom panel) respectively.

Figure 6
Opposing effects of Y63A and Y63F on β-catenin and other plasma-membrane-associated proteins

(A) Immunoblot analysis of β-catenin, cadherin, phospho-FAK [anti-phospho-FAK (Tyr397)] and FAK. (B) Immunoblot analysis of β-catenin and phospho-β-catenin in cells treated with (+) or without (–) MG-132. (C) Immunocytochemical analysis of β-catenin (green). Nucleus was stained with DAPI (4′,6-diamidino-2-phenylindole) (blue). (D) Cells were stained with Alexa Fluor® 660 phalloidin to visualize F-actin. (E) Co-immunoprecipitation of Dab2 and LRP1 β-chain. An anti-Dab2 polyclonal antibody was used to pull down complexes and the precipitates were probed for LRP1 β-chain (top panel) and Dab2 p96 (bottom panel) respectively.

To search for potential link(s) between mutations within the LRP1 FXNPXY motif and the profound changes in cell morphology and in ERK activation, we determined interaction of LRP1 β-chain and various adaptor proteins by co-immunoprecipitation. Among various adaptor proteins examined (including AP2, ARH and Dab2), only Dab2 showed a drastic difference in binding between Y63A and Y63F. Immunoprecipitation of the cell lysates with anti-Dab2, and subsequent immunoblot analysis for LRP1 β-chain and Dab2 showed markedly reduced interaction between Y63A mutant and Dab2 (Figure 6E, lane 8), whereas interaction of Y29A (lane 6) and Y63F (lane 10) with Dab2 was normal as compared with IVwt (lane 4). Notably, the level of Dab2 was decreased in Y63F cells (lane 10).

The Y63A-expressing cells had altered caveolae architecture

Finally, we determined the effect of mutations at the YXNPXY motif on the receptor protein localization to lipid rafts or caveolae. Lipid rafts are membrane microdomains rich in cholesterol and signalling molecules [38]. Caveolin-1 is a membrane-bound protein that specifies the lipid rafts termed caveolae. We have shown previously that LRP1 localization to lipid rafts is sensitive to insulin and PDGF treatment in differentiated 3T3-L1 adipocytes [22]. Examination of caveolin-1 distribution showed a dramatic increase in caveolin-1 association with lipid rafts in cells that exhibited loss of growth inhibition (i.e. vector control and Y63A cells) (the four left-hand side panels in Figure 7A). Increased caveolin-1 association with lipid rafts was observed in Y63A cells even after treatment with cholesterol oxidase (four right-hand side panels in Figure 7A), suggesting that some of the lipid rafts were not present on the cell surface. There were no significant changes in the total cellular caveolin-1 concentrations between different cell lines as determined by immunoblot analysis of total cell lysates (results not shown). Immunocytochemical analysis showed prominent caveolin-1 signals present on the surface of Y63A cells (Figure 7B). A small proportion of Y63A (<1% of total) was reproducibly observed in lipid rafts (Figure 7C). These results suggest that the architecture of lipid rafts or caveolae, a platform of signal transduction, is altered by expression of the Y63A mutant.

Altered caveolar architecture in Y63A-expressing cells

Figure 7
Altered caveolar architecture in Y63A-expressing cells

(A) Cell homogenates were subjected to lipid raft fractionation using the detergent method as described in the Experimental section. After sucrose-density-gradient ultracentrifugation, the fractionated samples were probed for caveolin-1 (left-hand side panels). Samples derived from cells treated with cholesterol oxidase (CO) prior to fractionation are shown in the right-hand side panels. (B) Immunocytochemical analysis of caveolin-1 using a polyclonal anti-caveolin-1 antibody. (C) Immunoblots of LRP1 β-chain. IB, immunoblot.

Figure 7
Altered caveolar architecture in Y63A-expressing cells

(A) Cell homogenates were subjected to lipid raft fractionation using the detergent method as described in the Experimental section. After sucrose-density-gradient ultracentrifugation, the fractionated samples were probed for caveolin-1 (left-hand side panels). Samples derived from cells treated with cholesterol oxidase (CO) prior to fractionation are shown in the right-hand side panels. (B) Immunocytochemical analysis of caveolin-1 using a polyclonal anti-caveolin-1 antibody. (C) Immunoblots of LRP1 β-chain. IB, immunoblot.

DISCUSSION

Lack of functional LRP1 is associated with hyperactivation of ERK

The current site-specific mutagenesis analysis attempts to unravel functional roles of the tyrosine residues within the two NPXY motifs of human LRP1 in cell growth regulation and signal transduction. We have identified the distal FXNPXY motif as a functional domain for cell growth inhibition and ERK activation. Cells deficient in LRP1, either because of the lack of LRP1 expression (in LRP1-null cells) or the lack of functional FXNPXY motif (in Y63A cells), displayed aberrantly constitutive activation of MEK→ERK→cyclin D1 (Figures 5B and 5C). Although mechanisms by which deficiency in normal LRP1 expression can lead to hyperactivation of the ERK pathway remain to be defined, we found that restoring LRP1 expression could effectively suppress cell growth and blunt ERK activation, suggesting strongly that LRP1 indeed plays a role in antiproliferation. It was reported previously that LRP1-deficient MEFs (murine embryonic fibroblasts) had elevated phospho-ERK when cells were serum-starved for 18 h, and restoration of LRP1 expression in these cells suppressed ERK hyperactivation [39]. The level of ERK phosphorylation, however, was equivalent between LRP1-deficient and normal MEFs when serum was present in the media [39]. In the present study, we found that ERK hyperactivation occurred in vector-transfected LRP1-null cells and Y63A cells even when the media were supplemented with serum, and the expression of IVwt or Y63F suppressed ERK phosphorylation (Figure 5B). Differences in ERK phosphorylation between Y63A and Y63F cells were also observed when the cells were first deprived and then re-supplemented with serum (Figure 5C). Thus our results, in agreement with previous observations with the LRP1-deficient MEF cells, suggest that LRP1 plays a role in signal transduction through regulating the ERK pathways. Recently, several studies have shown that ERK activation through LRP1 occurs in other cell type as well. For example, the LRP1 ligand lactoferrin-induced mitogenesis of rat osteoblastic cells requires functional LRP1 and is associated with increased ERK phosphorylation [40]. Also, TGF-β-mediated differentiation of rat kidney fibroblasts can be potentiated by connective tissue growth factor, another LRP1 ligand, that can activate ERK in a process associated with transient phosphorylation of tyrosine within the LRP1 β-chain [41]. Our mutational analysis of the Tyr63 to alanine or Tyr63 to phenylalanine substitution showing differential responses in ERK activation provides new evidence implicating the involvement of the FXNPXY motif of LRP1 in ERK activation.

Signalling function of LRP1 is specifically linked to the FXNPXY motif

The present studies also provide new evidence linking the FXNPXY motif of LRP1 to cell growth inhibition (Figures 3E and 3F) and tumour progression (Figure 4). The opposing effects of Y63A and Y63F mutations on cell growth regulation are accompanied by differential alterations in the levels of β-catenin, cadherin and FAK (Figure 6A), in addition to altered ERK activation. Two pieces of experimental evidence suggest that involvement of the FXNPXY motif is not attributable to tyrosine phosphorylation. First, in the cultured CHO cells, we did not detect phosphorylation of the Tyr63 residue under any culture conditions tested (H. Zhang, unpublished work). Secondly, the fact the Y63F mutant, but not the Y63A mutant, retained the normal LRP1 function in suppressing cell growth and tumour development suggests that the presence of an aromatic side chain is probably more important than phosphorylation in conferring LRP1 action. The lack of an effect of Tyr29 to alanine substitution on the LRP1 cell growth inhibition indicates that there is no functional redundancy between the NPXY29 and FXNPXY63 motifs. Moreover, the loss of cell growth inhibition by Tyr63 to alanine substitution observed in the present study was unlikely to be attributable to abnormally high or low expression of the mutant, because the effect was observed with multiple clones of which the level of Y63A expression was comparable with other variants tested. Thus our work suggests a specific cell growth inhibition function associated with the FXNPXY63 motif. Recently, an in vivo transgenic mice work in which the NPVY63 motif of the mouse LRP1 was replaced (using knock-in technology) with amino acids AAVA63 showed that Tyr63 to alanine substitution caused no observable phenotypic changes in the resulting animals [42]. However, mutations at the NPXY29 motif (NPTY-to-AATA) resulted in destruction of liver tissue between E16.5 (embryonic day 16.5) and E18.5 [42]. The reason for the lack of a manifestation of the NPXY63 function in cell growth inhibition in vivo remains to be explained. It is possible that selective interaction between NPXY29 and adaptor proteins for cytokine signalling is organ-specific and also developmental stage-specific [42], which may not be revealed in ovarian-originated CHO cells in vitro. Therefore the selective interaction of adaptor proteins with NPXY motifs in a tissue- and time-specific manner may play important roles in diverse functions of LRP1.

Pleiotropic effects of FXNPXY mutations on signalling pathways

The two NPXY motifs in the cytoplasmic domain of LRP1 β-chain have been shown to interact with a variety of adaptor and scaffold proteins, and some of the adaptor proteins are potentially involved in mitogenic signal transduction in addition to mediating endocytosis [14,43]. However, the biological significance of LRP1 binding to these adaptor proteins remains largely unclear. We have discovered in the present study that mutation within the FXNPXY motif has a profound effect on the level of Dab2 (e.g. in the case of Y63F) and its interaction with LRP1 β-chain (e.g. in the case of Y63A mutation) (Figure 6E). Dab2 is involved in regulation of endocytosis of LDLR family members such as LDLR, ApoER2 (ApoE receptor 2) and megalin [44], through interacting with the FXNPXY motif irrespective of the phosphorylation status of the tyrosine residue [6,15]. Dab2 is a known tumour suppressor; its expression in human ovarian carcinoma cells suppresses cell growth rate and tumour formation in nude mice [45,46], whereas homozygous knockout of Dab2 results in early embryonic lethality [45]. Factors that are known to regulate ERK activation include Grb2 (growth-factor-receptor-bound protein 2) and SOS (Son of Sevenless), and the ERK activation by Grb2 and SOS can be attenuated by Dab2 through its interacting with and sequestering Grb2 [44,47]. In the present mutagenesis studies, hyperactivation of ERK in cells deficient in LRP1 (either because LRP1-null or Y63A mutation) was closely associated with reduced interaction between LRP1 and Dab2 (Figures 5B and 6E). It remains to be determined if the decreased LRP1–Dab2 interaction in LRP1-null and Y63A cells is responsible for hyperactivation of ERK.

LRP1 has been recognized as a member of the TβR (TGF-β receptor) family and it can mediate TGF-β-induced growth inhibition independent of other TβRs [13,34]. During the early stage of tumorigenesis, TGF-β acts as a tumour suppressor principally through its ability to cause growth arrest and apoptosis, whereas at late stage of tumorigenesis, TGF-β promotes tumour progression through its ability to induce morphological changes termed EMT (endothelial mesenchymal transdifferentiation) [48]. The process of EMT is characterized by loss of cell–cell interaction and acquisition of fibroblastic morphology. The present experiments with cells expressing the full-length LRP1 or the minireceptor IVwt (Figure 1B) show that cytoskeleton reorganization (visualized by F-actin) occurred in these cells that exhibited cell growth inhibition. Thus lack of LRP1 expression results in cell transdifferentiation under non-stimulated conditions (i.e. in the absence of exogenous TGF-β). Our results are in agreement with previously reported morphological changes (from fibroblastoid spindle shape to wild-type squamous shape) observed in 13-5-1 cells transfected with the LRP1 minigene [13]. In mouse mammary gland epithelial cells, Dab2 is shown to play a role in TGF-β-induced EMT, probably through activation of integrin [49].

Dab2 also plays a role in the Wnt/β-catenin signalling [50]. Notably, concomitant changes in β-catenin protein levels and ERK phosphorylation were observed in the Y63A and Y63F cells in the absence of extracellular stimuli (Figures 5B and 6A). The Wnt/β-catenin pathway is known to play a role in ERK activation during growth stimulation [36,37]. Defective Wnt signalling is associated with cancer, osteoporosis and other diseases [51]. Members of the LDLR family, LRP5/6, act as co-receptors of Frizzled in the canonical Wnt/β-catenin signalling [52]. On the other hand, LRP1 has been shown to interact with Frizzled and negatively regulate the signalling process by sequestering Frizzled [12]. The Y63A-expressing cells that lost cell growth inhibition displayed up-regulated intracellular β-catenin, and the increased β-catenin was not merely a result of decreased proteasomal degradation (Figure 6B). These results provide additional evidence that expression of LRP1 indeed negatively regulates the Wnt/β-catenin signalling pathway. Thus ERK hyperactivation observed in the Y63A-expressing cells under non-stimulated conditions is probably achieved through interplay of multiple signalling molecules including Dab2 and β-catenin as suggested by other studies [36,53].

We thank Dr Xiang-Xi Xu for the anti-Dab2 antibody, Dr Dudley K. Strickland (Center for Vascular and Inflammatory Diseases, University of Maryland School of Medicine, Baltimore, MD, U.S.A.) for the human LRP1 expression plasmid, and Lin Chen and Shumei Zhong for excellent technical assistance. This work was supported by the Canadian Institutes for Health Research grant number GR-11471.

Abbreviations

     
  • AP2

    adaptor protein 2

  •  
  • ApoE

    apolipoprotein E

  •  
  • ARH

    autosomal recessive hypercholesterolaemia

  •  
  • CHO cell

    Chinese-hamster ovary cell

  •  
  • Dab2

    disabled 2

  •  
  • EMT

    endothelial mesenchymal transdifferentiation

  •  
  • ERK

    extracellular-signal-regulated kinase

  •  
  • FAK

    focal adhesion kinase

  •  
  • FBS

    fetal bovine serum

  •  
  • Grb2

    growth-factor-receptor-bound protein 2

  •  
  • GSK3

    glycogen synthase kinase 3

  •  
  • HA

    haemagglutinin

  •  
  • H/E

    haematoxylin and eosin

  •  
  • HEK-293 cells

    human embryonic kidney cells

  •  
  • LDLR

    low-density lipoprotein receptor

  •  
  • LRP1

    LDLR-related protein 1

  •  
  • MAPK

    mitogen-activated protein kinase

  •  
  • MEF

    murine embryonic fibroblast

  •  
  • MEK

    MAPK/ERK kinase

  •  
  • MTT

    3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide

  •  
  • PDGF

    platelet-derived growth factor

  •  
  • PKC

    protein kinase C

  •  
  • PKD

    protein kinase D

  •  
  • RAP

    receptor-associated protein

  •  
  • SOS

    Son of Sevenless

  •  
  • TGF

    transforming growth factor

  •  
  • TβR

    TGF-β receptor

References

References
1
Herz
J.
Hamann
U.
Rogne
S.
Myklebost
O.
Gausepohl
H.
Stanley
K. K.
Surface location and high affinity for calcium of a 500-kD liver membrane protein closely related to the LDL-receptor suggest a physiological role as lipoprotein receptor
EMBO J.
1988
, vol. 
7
 (pg. 
4119
-
4127
)
2
Sun
F.
Avramoglu
R. K.
Vassiliou
G.
Brown
R. J.
Ko
K. W.
McPherson
R.
Yao
Z.
Do clustered beta-propeller domains within the N-terminus of LRP1 play a functional role?
Biochim. Biophys. Acta
2005
, vol. 
1721
 (pg. 
139
-
151
)
3
Springer
T. A.
An extracellular beta-propeller module predicted in lipoprotein and scavenger receptors, tyrosine kinases, epidermal growth factor precursor, and extracellular matrix components
J. Mol. Biol.
1998
, vol. 
283
 (pg. 
837
-
862
)
4
Herz
J.
Strickland
D. K.
LRP: a multifunctional scavenger and signalling receptor
J. Clin. Invest.
2001
, vol. 
108
 (pg. 
779
-
784
)
5
Bonifacino
J. S.
Traub
L. M.
Signals for sorting of transmembrane proteins to endosomes and lysosomes
Annu. Rev. Biochem.
2003
, vol. 
72
 (pg. 
395
-
447
)
6
Uhlik
M. T.
Temple
B.
Bencharit
S.
Kimple
A. J.
Siderovski
D. P.
Johnson
G. L.
Structural and evolutionary division of phosphotyrosine binding (PTB) domains
J. Mol. Biol.
2005
, vol. 
345
 (pg. 
1
-
20
)
7
Nagai
M.
Meerloo
T.
Takeda
T.
Farquhar
M. G.
The adaptor protein ARH escorts megalin to and through endosomes
Mol. Biol. Cell
2003
, vol. 
14
 (pg. 
4984
-
4996
)
8
Stolt
P. C.
Bock
H. H.
Modulation of lipoprotein receptor functions by intracellular adaptor proteins
Cell. Signalling
2006
, vol. 
18
 (pg. 
1560
-
1571
)
9
Harasaki
K.
Lubben
N. B.
Harbour
M.
Taylor
M. J.
Robinson
M. S.
Sorting of major cargo glycoproteins into clathrin-coated vesicles
Traffic
2005
, vol. 
6
 (pg. 
1014
-
1026
)
10
May
P.
Bock
H. H.
Herz
J.
Integration of endocytosis and signal transduction by lipoprotein receptors
2003
 
Science STKE 2003, E12
11
Boucher
P.
Gotthardt
M.
Li
W. P.
Anderson
R. G.
Herz
J.
LRP: role in vascular wall integrity and protection from atherosclerosis
Science
2003
, vol. 
300
 (pg. 
329
-
332
)
12
Zilberberg
A.
Yaniv
A.
Gazit
A.
The low density lipoprotein receptor-1, LRP1, interacts with the human frizzled-1 (HFz1) and down-regulates the canonical Wnt signalling pathway
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
17535
-
17542
)
13
Tseng
W. F.
Huang
S. S.
Huang
J. S.
LRP-1/TbetaR-V mediates TGF-beta1-induced growth inhibition in CHO cells
FEBS Lett.
2004
, vol. 
562
 (pg. 
71
-
78
)
14
Gotthardt
M.
Trommsdorff
M.
Nevitt
M. F.
Shelton
J.
Richardson
J. A.
Stockinger
W.
Nimpf
J.
Herz
J.
Interactions of the low density lipoprotein receptor gene family with cytosolic adaptor and scaffold proteins suggest diverse biological functions in cellular communication and signal transduction
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
25616
-
25624
)
15
Howell
B. W.
Lanier
L. M.
Frank
R.
Gertler
F. B.
Cooper
J. A.
The disabled 1 phosphotyrosine-binding domain binds to the internalization signals of transmembrane glycoproteins and to phospholipids
Mol. Cell. Biol.
1999
, vol. 
19
 (pg. 
5179
-
5188
)
16
Loukinova
E.
Ranganathan
S.
Kuznetsov
S.
Gorlatova
N.
Migliorini
M. M.
Loukinov
D.
Ulery
P. G.
Mikhailenko
I.
Lawrence
D. A.
Strickland
D. K.
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
J. Biol. Chem.
2002
, vol. 
277
 (pg. 
15499
-
15506
)
17
Li
Y.
Marzolo
M. P.
van Kerkhof
P.
Strous
G. J.
Bu
G.
The YXXL motif, but not the two NPXY motifs, serves as the dominant endocytosis signal for low density lipoprotein receptor-related protein
J. Biol. Chem.
2000
, vol. 
275
 (pg. 
17187
-
17194
)
18
FitzGerald
D. J.
Fryling
C. M.
Zdanovsky
A.
Saelinger
C. B.
Kounnas
M.
Winkles
J. A.
Strickland
D.
Leppla
S.
Pseudomonas exotoxin-mediated selection yields cells with altered expression of low-density lipoprotein receptor-related protein
J. Cell Biol.
1995
, vol. 
129
 (pg. 
1533
-
1541
)
19
Anand
N.
Murthy
S.
Amann
G.
Wernick
M.
Porter
L. A.
Cukier
I. H.
Collins
C.
Gray
J. W.
Diebold
J.
Demetrick
D. J.
Lee
J. M.
Protein elongation factor EEF1A2 is a putative oncogene in ovarian cancer
Nat. Genet.
2002
, vol. 
31
 (pg. 
301
-
305
)
20
Mosmann
T.
Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays
J. Immunol. Methods
1983
, vol. 
65
 (pg. 
55
-
63
)
21
Huttenlocher
A.
Ginsberg
M. H.
Horwitz
A. F.
Modulation of cell migration by integrin-mediated cytoskeletal linkages and ligand-binding affinity
J. Cell Biol.
1996
, vol. 
134
 (pg. 
1551
-
1562
)
22
Zhang
H.
Links
P. H.
Ngsee
J. K.
Tran
K.
Cui
Z.
Ko
K. W.
Yao
Z.
Localization of low density lipoprotein receptor-related protein 1 to caveolae in 3T3-L1 adipocytes in response to insulin treatment
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
2221
-
2230
)
23
Ko
K. W.
McLeod
R. S.
Avramoglu
R. K.
Nimpf
J.
FitzGerald
D. J.
Vukmirica
J.
Yao
Z.
Mutation at the processing site of chicken low density lipoprotein receptor-related protein impairs efficient endoplasmic reticulum exit, but proteolytic cleavage is not essential for its endocytic functions
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
27779
-
27785
)
24
Ko
K. W.
Avramoglu
R. K.
McLeod
R. S.
Vukmirica
J.
Yao
Z.
The insulin-stimulated cell surface presentation of low density lipoprotein receptor-related protein in 3T3-L1 adipocytes is sensitive to phosphatidylinositide 3-kinase inhibition
Biochemistry
2001
, vol. 
40
 (pg. 
752
-
759
)
25
Smart
E. J.
Ying
Y. S.
Conrad
P. A.
Anderson
R. G.
Caveolin moves from caveolae to the Golgi apparatus in response to cholesterol oxidation
J. Cell Biol.
1994
, vol. 
127
 (pg. 
1185
-
1197
)
26
Avramoglu
R. K.
Nimpf
J.
McLeod
R. S.
Ko
K. W.
Wang
Y.
FitzGerald
D.
Yao
Z.
Functional expression of the chicken low density lipoprotein receptor-related protein in a mutant Chinese hamster ovary cell line restores toxicity of Pseudomonas exotoxin A and degradation of alpha2-macroglobulin
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
6057
-
6065
)
27
Li
Y.
Lu
W.
Marzolo
M. P.
Bu
G.
Differential functions of members of the low density lipoprotein receptor family suggested by their distinct endocytosis rates
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
18000
-
18006
)
28
Desrosiers
R. R.
Rivard
M. E.
Grundy
P. E.
Annabi
B.
Decrease in LDL receptor-related protein expression and function correlates with advanced stages of Wilms tumors
Pediatr. Blood Cancer
2006
, vol. 
46
 (pg. 
40
-
49
)
29
Koong
A. C.
Denko
N. C.
Hudson
K. M.
Schindler
C.
Swiersz
L.
Koch
C.
Evans
S.
Ibrahim
H.
Le
Q. T.
Terris
D. J.
Giaccia
A. J.
Candidate genes for the hypoxic tumor phenotype
Cancer Res.
2000
, vol. 
60
 (pg. 
883
-
887
)
30
Li
Y.
Wood
N.
Grimsley
P.
Yellowlees
D.
Donnelly
P. K.
In vitro invasiveness of human breast cancer cells is promoted by low density lipoprotein receptor-related protein
Invasion Metastasis
1998
, vol. 
18
 (pg. 
240
-
251
)
31
McGarvey
T.
Hussain
M. M.
Stearns
M. E.
In situ hybridization studies of alpha 2-macroglobulin receptor and receptor-associated protein in human prostate carcinoma
Prostate
1996
, vol. 
28
 (pg. 
311
-
317
)
32
Yamamoto
M.
Ikeda
K.
Ohshima
K.
Tsugu
H.
Kimura
H.
Tomonaga
M.
Expression and cellular localization of low-density lipoprotein receptor-related protein/alpha 2-macroglobulin receptor in human glioblastoma in vivo
Brain Tumor Pathol.
1998
, vol. 
15
 (pg. 
23
-
30
)
33
Obermoeller-McCormick
L. M.
Li
Y.
Osaka
H.
FitzGerald
D. J.
Schwartz
A. L.
Bu
G.
Dissection of receptor folding and ligand-binding property with functional minireceptors of LDL receptor-related protein
J. Cell Sci.
2001
, vol. 
114
 (pg. 
899
-
908
)
34
Huang
S. S.
Ling
T. Y.
Tseng
W. F.
Huang
Y. H.
Tang
F. M.
Leal
S. M.
Huang
J. S.
Cellular growth inhibition by IGFBP-3 and TGF-beta1 requires LRP-1
FASEB J.
2003
, vol. 
17
 (pg. 
2068
-
2081
)
35
Tetsu
O.
McCormick
F.
Beta-catenin regulates expression of cyclin D1 in colon carcinoma cells
Nature
1999
, vol. 
398
 (pg. 
422
-
426
)
36
Yun
M. S.
Kim
S. E.
Jeon
S. H.
Lee
J. S.
Choi
K. Y.
Both ERK and Wnt/beta-catenin pathways are involved in Wnt3a-induced proliferation
J. Cell Sci.
2005
, vol. 
118
 (pg. 
313
-
322
)
37
Jeon
S. H.
Yoon
J. Y.
Park
Y. N.
Jeong
W. J.
Kim
S.
Jho
E. H.
Surh
Y. J.
Choi
K. Y.
Axin inhibits extracellular signal-regulated kinase pathway by Ras degradation via beta-catenin
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
14482
-
14492
)
38
Anderson
R. G.
The caveolae membrane system
Annu. Rev. Biochem.
1998
, vol. 
67
 (pg. 
199
-
225
)
39
Ma
Z.
Thomas
K. S.
Webb
D. J.
Moravec
R.
Salicioni
A. M.
Mars
W. M.
Gonias
S. L.
Regulation of Rac1 activation by the low density lipoprotein receptor-related protein
J. Cell Biol.
2002
, vol. 
159
 (pg. 
1061
-
1070
)
40
Grey
A.
Banovic
T.
Zhu
Q.
Watson
M.
Callon
K.
Palmano
K.
Ross
J.
Naot
D.
Reid
I. R.
Cornish
J.
The low-density lipoprotein receptor-related protein 1 is a mitogenic receptor for lactoferrin in osteoblastic cells
Mol. Endocrinol.
2004
, vol. 
18
 (pg. 
2268
-
2278
)
41
Yang
M.
Huang
H.
Li
J.
Li
D.
Wang
H.
Tyrosine phosphorylation of the LDL receptor-related protein (LRP) and activation of the ERK pathway are required for connective tissue growth factor to potentiate myofibroblast differentiation
FASEB J.
2004
, vol. 
18
 (pg. 
1920
-
1921
)
42
Roebroek
A. J.
Reekmans
S.
Lauwers
A.
Feyaerts
N.
Smeijers
L.
Hartmann
D.
Mutant Lrp1 knock-in mice generated by recombinase-mediated cassette exchange reveal differential importance of the NPXY motifs in the intracellular domain of LRP1 for normal fetal development
Mol. Cell. Biol.
2006
, vol. 
26
 (pg. 
605
-
616
)
43
Barnes
H.
Ackermann
E. J.
van der Geer
P.
v-Src induces Shc binding to tyrosine 63 in the cytoplasmic domain of the LDL receptor-related protein 1
Oncogene
2003
, vol. 
22
 (pg. 
3589
-
3597
)
44
Zhou
J.
Hsieh
J. T.
The inhibitory role of DOC-2/DAB2 in growth factor receptor-mediated signal cascade. DOC-2/DAB2-mediated inhibition of ERK phosphorylation via binding to Grb2
J. Biol. Chem.
2001
, vol. 
276
 (pg. 
27793
-
27798
)
45
Yang
D. H.
Fazili
Z.
Smith
E. R.
Cai
K. Q.
Klein-Szanto
A.
Cohen
C.
Horowitz
I. R.
Xu
X. X.
Disabled-2 heterozygous mice are predisposed to endometrial and ovarian tumorigenesis and exhibit sex-biased embryonic lethality in a p53-null background
Am. J. Pathol.
2006
, vol. 
169
 (pg. 
258
-
267
)
46
Mok
S. C.
Wong
K. K.
Chan
R. K.
Lau
C. C.
Tsao
S. W.
Knapp
R. C.
Berkowitz
R. S.
Molecular cloning of differentially expressed genes in human epithelial ovarian cancer
Gynecol. Oncol.
1994
, vol. 
52
 (pg. 
247
-
252
)
47
Xu
X. X.
Yi
T.
Tang
B.
Lambeth
J. D.
Disabled-2 (Dab2) is an SH3 domain-binding partner of Grb2
Oncogene
1998
, vol. 
16
 (pg. 
1561
-
1569
)
48
Derynck
R.
Akhurst
R. J.
Balmain
A.
TGF-beta signalling in tumor suppression and cancer progression
Nat. Genet.
2001
, vol. 
29
 (pg. 
117
-
129
)
49
Prunier
C.
Howe
P. H.
Disabled-2 (Dab2) is required for transforming growth factor beta-induced epithelial to mesenchymal transition (EMT)
J. Biol. Chem.
2005
, vol. 
280
 (pg. 
17540
-
17548
)
50
Hocevar
B. A.
Mou
F.
Rennolds
J. L.
Morris
S. M.
Cooper
J. A.
Howe
P. H.
Regulation of the Wnt signalling pathway by disabled-2 (Dab2)
EMBO J.
2003
, vol. 
22
 (pg. 
3084
-
3094
)
51
Moon
R. T.
Kohn
A. D.
De Ferrari
G. V.
Kaykas
A.
WNT and beta-catenin signalling: diseases and therapies
Nat. Rev. Genet.
2004
, vol. 
5
 (pg. 
691
-
701
)
52
He
X.
Semenov
M.
Tamai
K.
Zeng
X.
LDL receptor-related proteins 5 and 6 in Wnt/beta-catenin signalling: arrows point the way
Development
2004
, vol. 
131
 (pg. 
1663
-
1677
)
53
Kim
S. E.
Choi
K. Y.
EGF receptor is involved in WNT3a-mediated proliferation and motility of NIH3T3 cells via ERK pathway activation
Cell. Signalling
2007
, vol. 
19
 (pg. 
1554
-
1564
)