The Low Density Lipoprotein Receptor-1, LRP1, Interacts with the Human Frizzled-1 (HFz1) and Down-regulates the Canonical Wnt Signaling Pathway*

Members of the low density lipoprotein receptor family (LDLR), LRP5/6, were shown to interact with the Frizzled (Fz) receptors and to function as Wnt coreceptors. Here we show that mLRP4T100, a minireceptor of LRP1, another member of the LDLR family, interacts with the human Fz-1 (HFz1), previously shown to serve as a receptor transmitting the canonical Wnt-3a-induced signaling cascade. However, in contrast to LRP5/6, mLRP4T100, as well as the full-length LRP1, did not cooperate with HFz1 in transmitting the Wnt-3a signaling but rather repressed it. mLRP4T100 inhibitory effect was displayed also by endocytosis-defective mLRP4T100 mutants, suggesting that LRP1 repressive effect is not attributable to LRP1-mediated enhanced HFz1 internalization and subsequent degradation. Enforced expression of mLRP4T100 decreased the capacity of HFz1 cys-teine-rich domain (CRD) to interact with LRP6, in contrast to HFz1-CRD/Wnt-3a interaction that was not disrupted by overexpressing mLRP4T100. These data suggest that LRP1, by sequestering HFz1, disrupts the receptor/coreceptor complex formation, leading to the repression of the canonical Wnt signaling. Thus, this study implies that the ability to interact

The low density lipoprotein receptor (LDLR) 1 related protein-1 (LRP1), a member of the LDLR endocytic receptor family, is a large multifunctional cysteine-rich receptor that binds and endocytoses a variety of structurally and functionally distinct ligands (1,2). The LDLR family comprises two major subfamilies: one comprised of "small" receptor members of ϳ120 kDa (the LDL receptor, apoE receptor-2, and the very low density lipoprotein receptor) and a second one encompassing "large" receptor members of ϳ600 kDa (LRP1, LRP1B, and LRP2 (Megalin)). In addition, two less related members, LRP5 and LRP6, were recently identified as Wnt coreceptors (3)(4)(5)(6).
LRP1 is synthesized as a single chain precursor that undergoes post-translational proteolytic processing within the trans-Golgi compartment by a furin-like protease (7). This results in the formation of mature LRP1 as a noncovalently associated heterodimer, consisting of an extracellular 515-kDa subunit and a transmembrane 85-kDa subunit with a transmembrane domain and a cytoplasmic tail. The ectodomain of LRP1 comprises four clusters of complement-type repeats (CR), I, II, III, and IV that consist of 2, 8, 10, and 12 cysteine-rich ligand binding repeats, respectively. These clusters are separated one from another by three clusters of epidermal growth factor (EGF) precursor repeats and (F/Y)WXD spacer repeats. Whereas clusters of several CR constitute the ligand binding domain, the EGF-like repeats are required for the dissociation of the receptor within endosomes (8). In addition to these extracellular modules, each member of the LDLR family also holds a relatively short cytoplasmic tail with potential endocytosis signals (7, 9 -11). Traditionally, all members of the LDLR family have been regarded as cell surface endocytosis receptors that function in delivering their ligands to lysosomes for degradation (12)(13)(14). However, recent studies have attributed new roles to these receptors in mediating various signal transduction pathways (2), among which is the canonical Wnt signaling (3)(4)(5)(6).
The secreted cell-signaling glycoproteins, Wnts, play a central role in development and homeostasis (15)(16)(17). In the canonical Wnt signaling pathway, binding of Wnt to its receptor, Frizzled (Fz), results in the activation of Dishevelled, which in turn inactivates the glycogen synthase kinase 3␤ within the ␤-catenin degradation complex, leading to ␤-catenin stabilization (18). The stabilized ␤-catenin accumulates in the nucleus, where it interacts with members of the lymphoid enhancer factor-1/T-cell factor (TCF) transcription factor family. This activates expression of Wnt target genes, such as c-myc, cyclin D1, and others, which triggers cell proliferation, oncogenic transformation, and inhibition of apoptosis (www.stanford.edu/ ϳrnusse/wntwindow.html) (19,20).
Fz proteins are seven transmembrane domain-containing receptors (22), whose cysteine-rich extracellular domains (CRD) bind to Wnts. Studies in Drosophila, Xenopus, and mammalian cells showed that Arrow in Drosophila and the mammalian paralogs LRP5/6 interact with Fz and serve as Wnt coreceptors for the transduction of the canonical Wnt signaling pathway (3)(4)(5)(6). LRP5/6 are LDLR family members whose different structural properties placed them in a distinct subfamily (2,23). Although they contain similar elements in their extracellular portion, these elements are arranged in a reverse order, such that the YWTD and the EGF repeats are N-terminal to the ligand-binding repeats preceding immediately the FIG. 1. LRP1 represses the canonical Wnt pathway. A, 293T cells were transfected with Wnt 3a-HA (0.1 g), wtHFz1 (1 g), increasing amounts of mLRP4T100-HA (as indicated below the ramp) or empty vector, TCF/Luc reporter (1 g), and ␤-galactosidase (0.1 g). Cells were harvested 48 h after transfection, the luciferase and ␤-galactosidase levels were measured, and the results were normalized for transfection efficiency relative to ␤-galactosidase levels. Values are the fold increase relative to a negative control co-transfected with empty vector, arbitrarily determined as 1. Lysates were subjected to Western analysis, and blots were incubated with anti-HA antibody (bottom panel). B, 293T cells were transfected with Wnt 3a-HA (0.1 g), HFz1-varFLAG (1 g), mLRP4T100-HA or empty vector (0.5 g), TCF/Luc reporter (1 g), and ␤-galacto-plasma membrane (2). Despite the presence of similar core elements in their ectodomains, LRP1 could not substitute for LRP5/6 in transmitting the canonical Wnt signal (3).
Previously, we have molecularly cloned the human Fz-1 (HFz1) and showed that it transduced the canonical Wnt signaling following exposure to Wnt-3a (24). Here we present data showing that LRP1, similarly to LRP6, interacts with HFz1; however, in contrast to LRP5/6, LRP1 does not participate in the transmission of the canonical Wnt signaling but rather represses it, presumably via disrupting HFz1⅐LRP6 complex formation.

EXPERIMENTAL PROCEDURES
Plasmids-The full-length LRP1 (25), subcloned in pcDNA3.1, was obtained from Dr. Joachim Herz (University of Texas, Dallas, TX). A minireceptor of LRP1, mLRP4T100, with an N-terminal HA epitope, as well as the endocytosis-defective HA-tagged mutants, mLRP4Ttailless and mLRP4T100(Y63A, L86A, L87A) (26), were obtained from Drs. Guojun Bu and Yongh Li (Washington University School of Medicine, St Louis, MO). mLRP4T100 comprises the fourth cluster of ligand binding repeats (residues 3274 -4525) of the full-length LRP1 (27) and the entire C terminus, including the transmembrane domain and the cytoplasmic tail (26,28,29). The endocytosis-defective mLRP4T100 (Y63A, L86A, L87A) mutant bears three point mutations at the cytoplasmic tail, whereas mLRP4Ttailless mutant lacks the cytoplasmic tail (26). LRP6 (30) was obtained from Dr. John F. Hess (Merck Research Laboratories, West Point, PA) and was FLAG-tagged by Drs. Liu Guizhong and Stuart A. Aaronson, of the Derald H. Ruttenberg Cancer Center, Mount Sinai School of Medicine, New York, NY as described (31). The HFz1 in pCEV29 was previously described (24). HFz1-varFLAG was constructed by inserting the FLAG epitope between the restriction sites DraIII and StyI located inside the HFz1 variable region. To enable detection by Western analysis, HFz1 was FLAG-tagged at its C terminus (Full HFz1-CЈ-FLAG) by inserting HFz1 open reading frame into pcDNA3, upstream of and in-frame to a FLAG epitope. HFz1 (⌬327)-FLAG (here designated FLAG-tagged HFz1-CRD) was previously described (24). HA-tagged HFz1-CRD was similarly constructed by inserting HFz1-CRD downstream of and in-frame to a HA tag in pcDNA3. Cyt-HFz1-FLAG was constructed by PCR amplification of residues 2161-2344, followed by insertion into pFLAG-CMV1 (Sigma) downstream of and in-frame to a FLAG tag sequence bordered by an Nterminal signal peptide. Wild type and mutated TCF/Luc reporter constructs (24), containing three copies of the TCF-binding domain upstream of a firefly luciferase reporter gene, were obtained from Dr. B. Vogelstein (Johns Hopkins Oncology Center, Baltimore, MD). HAtagged Wnt-3a in pLNCX vector previously reported (32) was obtained from Dr. J. Kitajewski (Columbia University, New York, NY).
Cell Cultures, Transfections, and Luciferase Reporter Assays-Human 293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum. For luciferase assays, cells growing in 6-well dishes were co-transfected at 60 -70% confluence by CaPO4 precipitation transfection (33). For TCF-dependent luciferase reporter assay, cells were transfected with HFz1, Wnt 3a-HA, TCF/Luc reporter, mLRP4T100-HA, LRP1, or empty vector (1 g each or as otherwise indicated in text or figure legends) and 0.1 g of CMV-␤-galactosidase plasmid (Invitrogen) used to evaluate the efficiency of transfection. The total amount of DNA was adjusted for equal amounts with empty vector. 48 h following transfection, the luciferase levels were measured employing the luciferase reporter gene assay kit (Roche Applied Science) as described previously (24). Luciferase activities were normalized for transfection efficiency by ␤-galactosidase activity measured using the ␤-galactosidase enzyme-linked immunosorbent assay kit (Roche Applied Science). Data are presented as mean values and S.D. for at least three independent experiments done in duplicate, compared with the level of luciferase activity obtained in the presence of empty vector, presented as 1.
Immunoprecipitation and Western Blot Analysis-48 h following transfection, 293T cells in 10-cm dishes were washed with phosphatebuffered saline and solubilized in lysis buffer (100 mM NaCl, 50 mM Tris, pH 7.5, 1% Triton X-100, 2 mM EDTA) containing complete EDTAfree protease inhibitor mixture (Roche Applied Science). 100 g of cellular protein were resolved in SDS-PAGE, and proteins were transferred to nitrocellulose membranes. After blocking with 5% low fat milk, filters were incubated with the specific primary antibody for 2 h. After washing in phosphate-buffered saline/Tween 20, membranes were subjected to enhanced chemiluminescence detection analysis (Amersham Biosciences) using horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology). For coimmunoprecipitation analysis, cells were solubilized in lysis buffer (see above) and extracts were clarified by centrifugation at 12,000 ϫ g for 30 min at 4°C. 3-6 mg of total cell lysates were incubated with the specific antibody for 18 h at 4°C and then incubated, with rotation, with protein A beads for 2 h at 4°C. Beads were then collected by centrifugation, washed three times in lysis buffer, suspended in Laemmli buffer, and analyzed by SDS-PAGE as described above. Anti-FLAG monoclonal antibody and anti-FLAG M2 affinity gel were purchased from Sigma. HA polyclonal antibody as well as anti-rabbit, anti-mouse, and anti-rat IgG horseradish peroxidase-conjugated secondary antibodies were obtained from Santa Cruz Biotechnology. HA monoclonal antibody was obtained from Roche Applied Science.
Immunofluorescence Staining-48 h following transfection, cells were fixed and permeabilized with methanol/acetone. Cells were treated with primary antibodies for 1 h, washed, and exposed to Rhodamine Red anti-mouse antibody (Jackson ImmunoResearch) for 1 h. Slides were analyzed by confocal microscopy.

RESULTS
mLRP4T100 Represses the Canonical Wnt Signaling Pathway-Because the LRP1 extracellular domain contains similar motifs to those of LRP5/6 (2), we sought to investigate whether it affects the canonical Wnt pathway. Because the full-length LRP1 is an extremely large receptor and its analysis is hampered because of its inefficient expression (28), we took advantage of a membrane-anchored LRP1 minireceptor, designated mLRP4T100, demonstrated to mimic the function and trafficking of LRP1 (27,34). 293T cells were co-transfected with HFz1, Wnt-3a-HA, TCF/Luc reporter, and increasing amounts of mLRP4T100-HA. Data showed that mLRP4T100 repressed the canonical Wnt signaling in a dose-dependent manner (Fig. 1A). The diminished level of Wnt signaling was not because of differences in levels of Wnt-3a-HA expression, which was equally expressed in the absence or presence of increasing amounts of mLRP4T100-HA (Fig. 1A, bottom panel). To assure that mLRP4T100 overexpression does not interfere with HFz1 expression, we used a FLAG-tagged HFz1, HFz1-varFLAG, whose expression could be evaluated by immunofluorescence analysis. 293T cells were cotransfected with HFz1-varFLAG, Wnt-3a-HA, and either mLRP4T100-HA or empty vector. Data showed that mLRP4T100-HA repressed Wnt3a/HFz1 induced signaling (Fig. 1B, top panel). Western analysis (Fig. 1B, middle panel) or immunofluorescence analysis (Fig. 1B, bottom  panel) showed that this repression was not because of inefficient expression of either Wnt-3a-HA or HFz1-varFLAG, respectively, in the presence of mLRP4T100. Wnt signaling was repressed also by the full-length LRP1 in a dose-dependent manner (Fig. 1, C and D), whereas the levels of both Wnt-3a-HA and HFz1-varFLAG were not affected by overexpressing LRP1. We next explored whether LRP1 would repress Wnt signaling transmitted via LRP6. Consistent with previous results showing that LRP6 synergizes with Fz8 and Wnt1 (35) and with Wnt-3a (31) in transducing the canonical Wnt signaling, our data showed that LRP6 synergizes also with HFz1 and Wnt-3a to transduce the Wnt pathway ( Fig. 2A). mLRP4T100 repressed Wnt signaling in a dose-dependent manner in the presence of ectopically expressed LRP6 (Fig. 2, B and C), and this repression was not because of inefficient expression of LRP6 in the presence of mLRP4T100 (Fig. 2C).
mLRP4T100 Acts Upstream of ␤-Catenin Degradation Complex-To determine the stage of the canonical Wnt pathway that is affected by LRP1, we investigated whether blocking of Results are presented as detailed in the legend to Fig. 1. B, 293T cells were transfected with Wnt 3a-HA (0.1 g), wtHFz1 or empty vector (1 g), LRP6-FLAG or empty vector (0.1 g), increasing amounts of mLRP4T100-HA or empty vector, TCF/Luc reporter (1 g), and ␤-galactosidase (0.1 g). Results are presented as detailed in Fig. 1. C, 293T cells were transfected with Wnt 3a-HA (0.1 g), wtHFz1 (1 g), mLRP4T100-HA or empty vector (0.5 g), LRP6-FLAG (0.5 g), TCF/Luc reporter (1 g), and ␤-galactosidase (0.1 g). Luciferase activity (top panel) was measured and presented as detailed in Fig. 1. Lysates were subjected to Western analysis, and blots were incubated with anti-FLAG antibody (bottom panel). glycogen synthase kinase-3␤ activity would overcome LRP1 repressive effect. 293T cells, transiently transfected with mLRP4T100 or empty vector and with TCF/Luc reporter, were treated with LiCl (30 mM for 24 h), known to activate Wnt signaling by blocking glycogen synthase kinase-3␤ activity (36). Data showed that mLRP4T100 did not affect LiCl-stimulated Wnt signaling (Fig. 3), suggesting that mLRP4T100 affects the canonical Wnt pathway upstream of ␤-catenin degradation complex.
mLRP4T100 Does Not Repress Wnt Signaling by Modulating HFz1 Endocytic Degradation-LRP1 belongs to a class of receptors that undergo constitutive endocytosis in the presence or absence of ligand, due to multiple potential endocytosis signals residing within the cytoplasmic tail (26,37). We thus sought to investigate whether mLRP4T100 repressive effect on Wnt signal is attributable to mLRP4T100-mediated modulation of HFz1 endocytic degradation. We took advantage of endocytosis-defective mLRP4T100 mutants (26) and explored their ability to repress Wnt signaling. 293T cells were cotransfected with HFz1, Wnt-3a-HA, and either mLRP4T100 (Y63A, L86A, L87A) or mLRP4Ttailless. Data showed that the two endocytosis-defective mLRP4T100 mutants repressed Wnt signaling in a dose-dependent manner (Fig. 4, A and B), thus excluding the possibility that mLRP4T100 repressive effect on Wnt3a/HFz1 signaling results from enhanced internalization and degradation of the Wnt⅐HFz complex.
mLRP4T100 Interacts with HFz1-CRD-Previously it was shown that LRP5/6 bind to the CRD domain of Fz receptor via EGF-repeats 1 and 2 of LRP5/6 (35). Because LRP5/6 contain core elements similar to those of LRP1, although in a reverse order, we sought to investigate whether LRP1 is capable of binding to HFz1. 293T cells were transiently co-transfected with HFz1-CRD-FLAG and mLRP4T100-HA or empty vector, and coimmunoprecipitation assays were performed. Data showed that the anti-HA antibody immunoprecipitated HFz1-CRD-FLAG when co-expressed with mLRP4T100-HA (Fig. 5A,  lane a) but not with empty vector (Fig. 5A, lane b). Similarly, mLRP4T100-HA was immunoprecipitated by anti-FLAG antibody when co-expressed with HFz1-CRD-FLAG (Fig. 5A, lane e) but not with empty vector (Fig. 5A, lane f). mLRP4T100-HA efficiently bound also to the full-length HFz1 FLAG-tagged at its CЈ terminus (full HFz1-CЈ-FLAG) (Fig. 5B, lane a) but did not bind to the FLAG-tagged membrane-anchored HFz1cytoplasmic fragment (Cyt-HFz1-FLAG) (Fig. 5, lane b).
Results thus suggested that HFz1-CRD specifically binds to mLRP4T100. We next investigated whether mLRP4T100-HA and HFz1-CRD-FLAG interact when both are expressed in adjacent cells. 293T cells transiently transfected with mLRP4T100-HA were co-cultivated with 293T cells expressing HFz1-CRD-FLAG, and lysates were immunoprecipitated with anti-HA antibody. Data showed that the anti-HA immunocomplexes (Fig. 5C, lane a) contained mLRP4T100-HA as well as HFz1-CRD-FLAG, further confirming that the extracellular domains of HFz1 and LRP1 do interact.
mLRP4T100 Competes with LRP6 for Binding to HFz1-CRD-Because mLRP4T100 binds to HFz1-CRD, we next explored whether overexpressing mLRP4Ttailless would affect LRP6/HFz1 interaction. 293T cells were co-transfected with HFz1-CRD-HA, LRP6-FLAG, and either mLRP4Ttailless-HA or empty vector, and immunoprecipitation assays were performed. Data showed that despite similar levels of expression of LRP6-FLAG and HFz1-CRD-HA, in the presence or absence of mLRP4Ttailless (Fig. 6A, lanes c or d, respectively), the anti-FLAG immunoprecipitates contained only 20% of HFz1-CRD-HA in the presence of mLRP4Ttailless, as compared with empty vector alone (Fig. 6A, lane a compared with lane b). These data suggested that mLRP4Ttailless, presumably by binding to HFz1-CRD, interferes with LRP6/HFz1 interaction.

LRP1 Interacts with HFz1 and Represses Wnt Signaling
Previously, we showed that HFz1-CRD interacts with Wnt-3a (24). To determine whether overexpressing mLRP4Ttailless would affect HFz1-CRD/Wnt-3a interaction, 293T cells were co-transfected with Wnt-3a-HA, HFz1-CRD-FLAG, and either mLRP4Ttailless-HA or empty vector, and immunoprecipitation assays were performed. Data showed that the anti-FLAG immunoprecipitates contained similar levels of Wnt-3a-HA and HFz1-CRD-FLAG in the presence or absence of mLRP4Ttailless-HA (Fig. 6B, lanes b or c, respectively), suggesting that mLRP4T100 does not disrupt HFz1-CRD⅐Wnt-3a complex formation. DISCUSSION In the present study we show that mLRP4T100, a minireceptor of LRP1, interacted with the CRD of HFz1 when coexpressed. Furthermore, we show that mLRP4T100 and HFz1-CRD interacted also when both proteins were expressed in adjacent cells, further demonstrating the specificity of this interaction and ruling out the possibility that it resulted from retention of misfolded proteins within the endoplasmic reticulum (38). Moreover, the similar levels and cell surface localization of the full-length HFz1-FLAG, in the presence or absence of LRP1 or mLRP4T100-HA as detected by immunofluores-cence, excluded the possibility that the repression of the Wnt signaling by LRP1 resulted from endoplasmic reticulum trapping of HFz1⅐mLRP4T100 complexes and improper targeting of HFz1 to the plasma membrane.
Previously, it was reported that LRP1 could not substitute LRP5/6 in the transmission of the Wnt canonical pathway (3). Consistent with this observation we show that mLRP4T100 does not cooperate with HFz1 to transmit Wnt-3a signaling. Moreover, our data suggest that LRP1 negatively regulates the canonical Wnt pathway.
LRP1 is abundantly expressed in a variety of tissues, predominantly in mature tissues of the liver and central nervous system (12, 40 -41), primarily functioning in mediating the binding and transport of ligands from the cell surface to the endosomal degradation pathway (42). However, our data showing that the endocytosis-defective mLRP4T100 mutants efficiently repressed Wnt signaling suggest that LRP1 repressive effect does not result from LRP1-mediated enhanced internalization and subsequent degradation of HFz1⅐Wnt-3a complexes.
Although most LDLR family members are classically known for their involvement in the endocytic pathway, some are involved in signaling processes (2). For example, recent studies  c, d, g, h) or after (IP, lanes a, b, e, f) immunoprecipitation. B, 293T cells were co-transfected with mLRP4T100-HA (2 g) and Full HFz1-CЈ-FLAG (4 g) or Cyt-HFz1-FLAG (2 g). Lysates were immunoprecipitated with anti-HA. Blots were treated (IB) with anti-HA or anti-FLAG before (lanes c, d) or after (lanes a, b) immunoprecipitation. C, 293T cells transfected with mLRP4T100-HA or empty vector were co-cultivated with 293T cells transfected with HFz1-CRD-FLAG. Lysates were immunoprecipitated with anti-HA (lanes a, b) antibody. Blots were incubated (IB) with either anti-HA or anti-FLAG before (lanes c, d, e) or after (lanes a, b) immunoprecipitation.
showed that some LDLR members, particularly LRP1, participate in signal transmission by the plasminogen activator, leading to enhanced activity of protein kinase A (43). In addition, LRP1 binding to ␣2-macroglobulin activates Ca 2ϩ influx (44). LRP1 was also shown to modulate platelet-derived growth factor signaling (45)(46)(47). The cytoplasmic tail of LRP1 interacts with numerous adaptor molecules, including a variety of cytoplasmic adaptor or scaffold proteins, such as Fe65, PSD-95, c-Jun N-terminal kinase (JNK)-interacting proteins JIP-1/2 and Disabled-1 (10) that play a role in regulating mitogenactivated protein kinase activity, cytoskeletal reorganization, cell adhesion, and stimulation of Ca 2ϩ entry (48). In addition, mLRP1 cytoplasmic tail was shown to interact with a GTPbinding protein and induce cyclic-AMP-dependent protein kinase activity (49) and to recruit Shc followed by Ras activation (50). LRP1 also inhibits nuclear activation of JNK-target sub-strates, Elk-1 and c-Jun, by sequestering the activated JNK into the plasma membrane (51). Recent studies showed that LRP1 activated phosphoinositide 3-kinase and extracellularrelated kinase via G proteins (52). It was thus reasonable to assume that some of those LRP1-mediated signaling events might directly or indirectly lead to the repression of the canonical Wnt pathway. However, because LRP1 cytoplasmic domain plays a crucial role in LRP1-mediated signal transduction (2), our data showing that the tailless mLRP4T100 mutant preserved its repressive effect on Wnt signaling exclude the possibility that this repression is attributed to LRP1-affected signaling pathways that cross-regulate the canonical Wnt pathway.
We show that mLRP4T100 interacts with HFz1-CRD and that this interaction interferes with HFz1⅐LRP6 complex formation. Furthermore, we show that HFz1-CRD/Wnt-3a inter-FIG. 6. mLRP4Ttailless interferes with receptor-coreceptor complex formation. A, mLRP4Ttailless interferes with HFz1-CRD/LRP6 interaction. 293T cells were co-transfected with LRP6-FLAG (1 g), HFz1-CRD-HA (1 g), and mLRP4Ttailless-HA or empty vector (3 g). Cell lysates were immunoprecipitated with anti-FLAG antibody. Immunoblots were treated (IB) with anti-FLAG or anti-HA antibodies before (lanes c, d, e) or after (lanes a, b) immunoprecipitation. A representative Western blot is shown. Inset, mean value (three independent experiments) of densitometric measurements of HFz1-CRD-HA immunoprecipitated by anti-FLAG in the presence of mLRP4Ttailless (a) or empty vector (b). Each bar denotes the mean Ϯ S.D. B, mLRP4T100 does not interfere with HFz1-CRD/Wnt-3a interaction. 293⌻ cells were co-transfected with mLRP4Ttailless-HA or empty vector (3 g), HFz1-CRD-FLAG or empty vector (1 g), and Wnt-3a-HA or empty vector (1 g). Lysates were subjected to immunoprecipitation and Western blot analysis as described in panel A. action was not disrupted by mLRP4T100. Our studies thus imply that mLRP4T100 inhibits Wnt signaling by binding to the ectodomain of HFz1, thus interfering with the formation of active HFz1⅐LRP6 complex.
It is tempting to speculate that the role we ascribe to LRP1 in negatively regulating the canonical Wnt signaling pathway may extend to some other members of the LDLR family. The canonical Wnt signaling plays a crucial role in various developmental processes (15)(16)(17). Interestingly, recent observations indicated that the role of LRP5 in bone development is mediated via the Wnt canonical signaling (53). It is thus intriguing to assume that LRP1 and presumably other LDLR family members might be implicated in negatively regulating various Wnt/ Fz/LRP5/6-induced signaling events during development.
The versatility of the effects of some LDLR members on Wnt signaling are reminiscent of several other factors, such as Dickkopf family members (54) and the Frizzled-related proteins (21), which both antagonize and synergize with Wnt ligands and thus serve to fine-tune Wnt signaling pathway, whose tightly controlled regulation is essential for proper coordination of various developmental processes.