Canonical and noncanonical Wnts use
a common mechanism to activate
completely unrelated coreceptors
Luca Grumolato,1
Guizhong Liu,1
Phyllus Mong,1
Raksha Mudbhary,1
Romi Biswas,1
Randy Arroyave,1
Sapna Vijayakumar,1
Aris N. Economides,2
and Stuart A. Aaronson1,3
1
Department of Oncological Sciences, Mount Sinai School of Medicine, New York, New York 10029, USA; 2
Regeneron
Pharmaceuticals, Inc., Tarrytown, New York 10591, USA
Wnt ligands signal through b-catenin and are critically involved in cell fate determination and stem/progenitor
self-renewal. Wnts also signal through b-catenin-independent or noncanonical pathways that regulate crucial
events during embryonic development. The mechanism of noncanonical receptor activation and how Wnts trigger
canonical as opposed to noncanonical signaling have yet to be elucidated. We demonstrate here that prototype
canonical Wnt3a and noncanonical Wnt5a ligands specifically trigger completely unrelated endogenous coreceptors—LRP5/6
and Ror1/2, respectively—through a common mechanism that involves their Wnt-dependent
coupling to the Frizzled (Fzd) coreceptor and recruitment of shared components, including dishevelled (Dvl), axin,
and glycogen synthase kinase 3 (GSK3). We identify Ror2 Ser 864 as a critical residue phosphorylated by GSK3 and
required for noncanonical receptor activation by Wnt5a, analogous to the priming phosphorylation of low-density
receptor-related protein 6 (LRP6) in response to Wnt3a. Furthermore, this mechanism is independent of Ror2
receptor Tyr kinase functions. Consistent with this model of Wnt receptor activation, we provide evidence that
canonical and noncanonical Wnts exert reciprocal pathway inhibition at the cell surface by competition for Fzd
binding. Thus, different Wnts, through their specific coupling and phosphorylation of unrelated coreceptors,
activate completely distinct signaling pathways.
[Keywords: Ror1/2; LRP5/6; Wnt3a; Wnt5a; receptor activation; noncanonical Wnt signaling]
Supplemental material is available at http://www.genesdev.org.
Received June 8, 2010; revised version accepted September 20, 2010.
The Wnt family, which includes 19 members in mammals,
is essential for embryonic development and tissue
homeostasis (Clevers 2006; Angers and Moon 2009).
Ligands such as Wnt1, Wnt3a, and Wnt8 couple the seventransmembrane
domain receptor Frizzled (Fzd) and the
single-membrane-spanning low-density receptor-related
protein 5/6 (LRP5/6) (MacDonald et al. 2009) to activate
Wnt–b-catenin signaling. In this pathway, Fzd recruits
the intracellular protein dishevelled (Dvl), which in turn
brings to the membrane the axin–GSK3 (glycogen synthase
kinase 3) complex, thereby promoting the initial
phosphorylation of LRP6 (Zeng et al. 2005, 2008). Further
phosphorylation of LRP6 by casein kinase 1 (CK1)
(Davidson et al. 2005; Zeng et al. 2005) is associated with
clustering of different proteins—including LRP6, Dvl, and
Axin—to form what has been defined as the LRP6 signalosome
(Bilic et al. 2007). Phosphorylated LRP5/6 leads to
inhibition of the so-called b-catenin destruction complex
(which includes axin, GSK3, Dvl, CK1, and the tumor
suppressor adenomatous polyposis coli), resulting in the
stabilization and translocation of b-catenin in the nucleus,
where it activates target genes through binding to
TCF/LEF transcription factors (Clevers 2006; MacDonald
et al. 2009). Among its different functions, the b-catenin
or canonical Wnt pathway is a major regulator of stem/
progenitor cell maintenance, expansion, and lineage specification
in both embryonic and adult tissues (Grigoryan
et al. 2008). Aberrations of this pathway that lead to its
constitutive activation by a variety of mechanisms—such
as gene mutations, ligand overexpression, or inhibitor
down-regulation—are involved in many types of cancer
(Clevers 2006).
Other so-called noncanonical Wnt pathways, defined
as Wnt- and/or Fzd-mediated signaling independent of
b-catenin transcriptional activity (Semenov et al. 2007),
are diverse and include the Wnt polarity, Wnt-Ca2+
, and
Wnt-atypical protein kinase C pathways. These pathways
have been reported to contribute to developmental processes
such as planar cell polarity in Drosophila, convergent
extension movements during gastrulation, or neuronal
3
Corresponding author.
E-MAIL stuart.aaronson@mssm.edu; FAX (212) 987-2240.
Article is online at http://www.genesdev.org/cgi/doi/10.1101/gad.1957710.
GENES & DEVELOPMENT 24:2517–2530 Ó 2010 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/10; www.genesdev.org 2517
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
and epithelial cell migration (Semenov et al. 2007; Simons
and Mlodzik 2008; Kikuchi et al. 2009). The recent
identification of Wnt-binding domains in different cell
membrane proteins such as Ror2 and Ryk has suggested
that noncanonical family members—including Wnt4,
Wnt5a, and Wnt11—may be able to trigger receptors other
than Fzd (Green et al. 2008; Angers and Moon 2009).
Genetic interactions between Ror2 and Wnt5a were
suggested by observations that mice with inactivation of
the gene encoding Ror2 showed striking similarities to
Wnt5aÀ/À
mice, including perinatal lethality, dwarfism,
facial abnormalities, and limb shortening (Green et al.
2008). Of note, Ror2À/À
embryos also exhibited defects in
the orientation of the sensory hair cells in the inner ear, a
hallmark of Wnt planar polarity pathway aberrations observed
in Wnt5aÀ/À
, Fzd3À/À
, and Fzd6À/À
mice (Wang
et al. 2006; Qian et al. 2007; Yamamoto et al. 2008). In
addition, Ror2 was shown to be involved in the effects of
Wnt5a in some cellular and developmental processes, such
as polarized cell migration and convergent extension movements
(Hikasa et al. 2002; Nishita et al. 2006; Schambony
and Wedlich 2007). Recent studies have also suggested that
Ror2 and Wnt5a may play a role in the progression of
different types of cancers (Nishita et al. 2010a).
Unlike the case with b-catenin signaling, the mechanisms
that underlie activation of noncanonical Wnt
pathways are not well understood. Despite the lack of
typical canonical defects in Wnt5aÀ/À
mice (Yamaguchi
et al. 1999; Grigoryan et al. 2008), prototype noncanonical
Wnt5a has been reported to signal to b-catenin in the
presence of overexpressed Fzd5 (He et al. 1997) or Fzd4
and LRP5 (Mikels and Nusse 2006). These findings argue
that the pathways initiated by different Wnts may depend
on the context of receptors expressed in a given target
cell, rather than on intrinsic properties of the ligands (van
Amerongen et al. 2008). We show that, in the same cell
expressing endogenous receptors, activation of the canonical
or noncanonical pathway is determined by Wnt ligand
specificity for coupling Fzd to different and completely
unrelated coreceptors. We further establish that prototype
canonical and noncanonical Wnt3a and Wnt5a use common
intracellular components—including Dvl, axin, and
GSK3—to activate LRP6 and Ror2 coreceptors, respectively,
and trigger their different phenotypic responses. Finally,
we provide evidence for reciprocal pathway inhibition
by competition of canonical and noncanonical Wnt
ligands for cell surface binding of Fzd. Given the large
number of Wnts, as well as an increasing array of putative
noncanonical pathways and receptors, this mechanism
may represent a general paradigm underlying the activation
of other, yet-to-be characterized, Wnt signaling pathways.
Results
Wnt3a and Wnt5a specifically trigger Ser/Thr
phosphorylation of LRP6 and Ror2, respectively,
by their coupling to a common coreceptor, Fzd
We initially compared the effects of prototype Wnt3a and
Wnt5a ligands under physiological conditions in the same
cells expressing endogenous canonical receptors LRP5
and LRP6 (Supplemental Fig. 1A,B), and Ror2, implicated
as a noncanonical Wnt receptor (Green et al. 2008). These
cells, 53S and 293T, also endogenously express Fzd4 and
Fzd5 (Supplemental Fig. 1C,D; Pan et al. 2008), as Wnt5a
has been reported to activate the canonical pathway in
the presence of overexpressed Fzd5 (He et al. 1997) or
Fzd4 and LRP5 (Mikels and Nusse 2006). Under the same
conditions, Wnt3a, but not Wnt5a, induced b-catenin
stabilization, while both ligands triggered phosphorylation
of the scaffold protein Dvl2 to similar extents, as
measured by its altered gel mobility (Supplemental Fig.
2A–C), in accordance with previous reports for other cells
(Gonzalez-Sancho et al. 2004). Wnt3a, but not Wnt5a,
also induced phosphorylation of LRP5/6 (Fig. 1A; Supplemental
Fig. 2D,E), consistent with Wnt5a’s lack of canonical
activity in either cell line. In contrast, Wnt5a, but
not Wnt3a, induced retardation in the mobility of Ror2
(Fig. 1A; Supplemental Fig. 2D,E), which reflected its
phosphorylation (Supplemental Fig. 2F; Yamamoto et al.
2007). Similar effects were observed in each of several
other cell types, including mouse embryonic fibroblasts
(MEFs), Saos-2, MDAMB-157, and HeLa cells (Supplemental
Fig. 2G; data not shown). These results indicated
that prototype canonical and noncanonical Wnts, respectively,
trigger the specific phosphorylation of completely
unrelated endogenous receptors.
While the mechanism involved in Wnt3a-induced
phosphorylation of LRP6 is well characterized (MacDonald
et al. 2009), the mechanism by which Wnt5a mediates
Ror2 phosphorylation is not known. Ror2 is a receptor
Tyr kinase (RTK), and there are conflicting reports that
Wnt5a induces Ror2 phosphorylation on Tyr (Liu et al.
2008; Mikels et al. 2009) or Ser/Thr (Yamamoto et al.
2007) residues. In the presence of their ligands, RTKs
generally are clustered and trans-autophosphorylated,
which triggers the activation of downstream effectors
(Schlessinger 2000). We compared the effects of treating
cells expressing N-terminal Flag-tagged Ror2 with Wnt5a
or anti-Flag antibody. Whereas antibody-induced Ror2
clustering triggered its Tyr autophosphorylation (Fig. 1B),
Wnt5a treatment failed to do so and instead promoted
Ror2 Ser/Thr phosphorylation (Fig. 1B). Of note, Wnt5ainduced
Ser/Thr phosphorylation of Ror2 exhibited relatively
slow kinetics (Supplemental Fig. 3A), similar to the
time course of LRP6 activation by Wnt3a (Supplemental
Fig. 3B; Khan et al. 2007).
It is well established that Wnt-induced coupling of
LRP5/6 and Fzd coreceptors is required for canonical
signaling (Tamai et al. 2000; Liu et al. 2005; Bilic et al.
2007). The fact that both Wnt3a and Wnt5a induced
Dvl phosphorylation (Fig. 1A; Supplemental Fig. 2A–D;
Gonzalez-Sancho et al. 2004) and increased the levels of
phosphatidylinositol 4,5-bisphosphates (PIP2) (Fig. 1C),
which are hallmarks of Wnt3a-induced activation of Fzd
(Pan et al. 2008; MacDonald et al. 2009), suggested that
Wnt5a also possessed the inherent ability to bind and
trigger Fzd. Indeed, when fused to the LRP5/6-binding
domain of Dkk2, Wnt5a activated b-catenin signaling
(Liu et al. 2005) and induced phosphorylation of LRP6
Grumolato et al.
2518 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
(Fig. 1D), indicating that the inability of Wnt5a to stimulate
the b-catenin pathway was due to its lack of functional
interaction with endogenous LRP5/6.
Since Wnt5a promoted Ror2 Ser/Thr phosphorylation,
we speculated that Wnt5a might function by a mechanism
analogous to that of canonical ligands and couple
Fzd to Ror2. To explore this possibility, we generated
Ror2–LRP6, a fusion receptor containing the Ror2 extracellular
domain and the LRP6 transmembrane and intracellular
domains. If Wnt5a triggered the coupling of
the Ror2 extracellular domain to endogenous Fzd, we
reasoned that we might observe phosphorylation of the
LRP6 intracellular domain of this chimeric receptor and
stimulation of the b-catenin pathway. Indeed, in the presence
of Ror2–LRP6, Wnt5a strongly stimulated b-catenin/
TCF reporter activity (Fig. 1E). Dkk1, an antagonist of
canonical signaling through interactions at the cell surface
with LRP5/6 (Clevers 2006), abrogated the ability of
Wnt3a, but not Wnt5a, to trigger signaling in Ror2–LRP6expressing
cells (Fig. 1E). Furthermore, Wnt5a specifically
induced Ser phosphorylation of Ror2–LRP6, while, under
the same conditions, Wnt3a triggered phosphorylation of
endogenous LRP6 (Fig. 1F). Thus, by replacing the Ror2
intracellular domain with that of LRP6, we converted
Ror2 into a canonical Wnt receptor specifically activated
by Wnt5a.
We next investigated the role of endogenous Fzd in
Wnt5a-induced triggering of Ror2. The protein Shisa,
which specifically inhibits Fzd translocation to the membrane
by retaining it in the endoplasmic reticulum
(Yamamoto et al. 2005), has been used to demonstrate
the involvement of Fzd in Wnt3a-mediated phosphorylation
of LRP6 (Zeng et al. 2008). Figure 2A confirms Shisa’s
inhibition of b-catenin/TCF reporter activity triggered by
Wnt3a, and shows further that Shisa antagonized the
ability of Wnt5a to activate Ror2–LRP6. Moreover, Shisa
abrogated both the phosphorylation of Ror2 induced by
Wnt5a (Fig. 2B,C) and Wnt3a-mediated activation of
LRP6 (Fig. 2B). As a control for Fzd inhibition, Shisa
expression blocked the upshift of Dvl2 induced by both
Figure 1. Prototype canonical and noncanonical
Wnt ligands activate distinct endogenous
receptors. (A) Opposite effects of
Wnt3a and Wnt5a on LRP5/6 and Ror2
phosphorylation. 53S cells were treated
with control, Wnt3a, or Wnt5a CM for 3 h,
and immunoblot analysis was performed
using the indicated antibodies for endogenous
proteins. The nonphosphorylated
forms of LRP6, LRP5, Ror2, and Dvl2
(arrow) and the upshifted bands corresponding
to the phosphorylated forms
(asterisk) are indicated. (B) Wnt5a induces
Ror2 phosphorylation on Ser/Thr residues.
Ror2À/À
MEFs stably expressing GFP or
N-terminal Flag-tagged Ror2 were treated
in the presence of the Tyr phosphatase
inhibitor sodium vanadate (40 mM) with
either control or Wnt5a CM, or with anti-HA
(control) or anti-Flag (50 mg/mL) monoclonal
antibodies for 1 h or 4 h. The cell lysates
were subjected to immunoprecipitation using
anti-Flag antibody, and immunoblot analysis
was performed with anti-phospho-Tyr
antibody, followed by two rounds of stripping
and reprobing with anti-phospho-Ser/
Thr and anti-Ror2 antibodies. (C) Effects
of Wnt3a and Wnt5a on PIP2 levels. The
results normalized to the control represent
the mean values 6 SEM of five independent
experiments. (*) P < 0.05 compared
with control (Student’s t-test). (D) The DKK2C–W5a secreted fusion protein can trigger LRP6 phosphorylation. Rat1 cells were
treated for 3 h with control, Wnt5a, DKK2C, or DKK2C–W5a CM, and immunoblot analysis was performed for endogenous LRP6 and
Dvl2. The amounts of Wnt5a and DKK2C–W5a proteins in same volumes of CM are shown. (E) Wnt5a stimulates the b-catenin/TCF
reporter in the presence of the Ror2–LRP6 chimeric receptor. Rat1 cells were cotransfected with SuperTop luciferase reporter, renilla
luciferase plasmid, and Ror2–LRP6 or control vector. Two days after transfection, the cells were treated overnight for maximal
stimulation with Wnt CM in the presence or absence of Dkk1 CM, and luciferase assay was performed. The renilla-normalized
values are expressed as fold increase compared with the vector-transfected control cells and correspond to the mean values (6SD, in
triplicate) of one representative of three independent experiments. (F) Wnt5a, but not Wnt3a, induces the phosphorylation of Ror2–
LRP6. Two days after transfection with Ror2–LRP6, Rat1 cells were treated for 3 h with control, Wnt3a, or Wnt5a CM, and the cell
lysates were subjected to immunoblot analysis using anti-phospho-LRP6 and anti-LRP6 antibodies. The bands close to the 225-kDa
and 76-kDa markers correspond to endogenous LRP6 and Ror2–LRP6, respectively.
Mechanism of Ror1/2 receptor activation
GENES & DEVELOPMENT 2519
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
Wnt3a and Wnt5a (Fig. 2B). Despite the fact that there are
10 Fzd genes encoded within the mammalian genome
(MacDonald et al. 2009), siRNAs targeting Fzd2, Fzd4,
and Fzd5 were shown recently to partially inhibit Wnt3ainduced
phosphorylation of LRP6 in 293T cells (Pan et al.
2008). Figure 2D and Supplemental Figure 4A show that
these same siRNAs also inhibited the phosphorylation of
Ror2 induced by Wnt5a in these cells. Together, these
results indicated that Fzd is required for the activation of
both canonical and noncanonical Wnt coreceptors.
Previous studies have shown that Wnt3a directly
coupled Fzd and LRP6 in an in vitro binding assay using
the secreted extracellular domains of these receptors
(Tamai et al. 2000). Figure 2, E and F shows that Wnt3a,
but not Wnt5a, was able to bridge LRP6 and Fzd. Conversely,
Wnt5a, but not Wnt3a, formed a tripartite complex
with Ror2 and Fzd (Fig. 2G). All of these findings
demonstrated that Wnt3a and Wnt5a physically couple
Fzd, respectively, to two different and completely unrelated
coreceptors.
Wnt-induced phosphorylation of LRP6 and Ror2
is mediated by a common intracellular machinery
Our findings that Wnt5a mediated coupling of Fzd to
Ror2 and promoted activation of Ror2–LRP6 suggested
that this ligand might induce Ser/Thr phosphorylation of
Ror2 through a mechanism analogous to Wnt3a triggering
of LRP6, which involves intracellular components
Dvl, axin, and GSK3. Figure 3A shows that Wnt5a treatment
promoted colocalization of aggregates staining for
Ror2 and Dvl2 at the cell membrane. These Ror2 membrane
aggregates were not detected in response to Ror2
shRNA (Supplemental Fig. 5). Consistent with a role of
Dvl in the activation of both canonical and noncanonical
receptors, down-regulation of the three Dvl isoforms by
lentiviral shRNAs or expression of a dominant-negative
Dvl construct (DN-Dvl), which binds to Fzd but is unable
to induce phosphorylation of LRP6 (Zeng et al. 2008),
inhibited the triggering of Ror2, LRP6, or Ror2–LRP6
induced by their respective ligands (Fig. 3B; Supplemental
Fig. 4B,C).
In the canonical Wnt pathway, axin functions as a
component of the b-catenin destruction complex as well
as a scaffold protein involved in the recruitment of GSK3
to the Fzd–LRP6 complex, required for LRP6 activation
by phosphorylation (Zeng et al. 2008; MacDonald et al.
2009). Figure 3C and Supplemental Figure 6A show that
shRNA-mediated down-regulation of axin inhibited the
phosphorylation of Ror2 or LRP6 specifically induced by
Wnt5a and Wnt3a, respectively. Moreover, Wnt5a, but
not Wnt3a, induced the formation of a complex containing
endogenous Ror2 and axin (Fig. 3D), suggesting that
Wnt5a triggered recruitment of axin-GSK3 to the Fzd–
Ror2 complex to promote Ror2 phosphorylation, similarly
to what has been described for LRP6 (Zeng et al.
2008). In fact, a recent study reported that GSK3 was
involved in Ror2 phosphorylation (Yamamoto et al. 2007).
We confirmed these data and showed that different
Figure 2. Wnt3a and Wnt5a couple Fzd
to LRP6 or Ror2 coreceptors, respectively,
to induce their phosphorylation. (A) Rat1
cells stably expressing GFP or myc-tagged
Shisa were cotransfected with SuperTop
luciferase reporter, renilla luciferase plasmid,
and Ror2–LRP6 or control vector. Two
days after transfection, the cells were treated
overnight with control, Wnt3a, or Wnt5a
CM, and a luciferase assay was performed.
The renilla-normalized values are expressed
as fold increase compared with the vectortransfected
control cells and correspond
to the mean values (6SD, n = 4) of one
representative of two independent experiments.
(B) Saos-2 cells stably expressing
GFP or myc-tagged Shisa were treated for
3 h with control, Wnt3a, or Wnt5a CM,
and immunoblot analysis was performed
with the indicated antibodies. The nonphosphorylated
(arrow) and phosphorylated
(asterisk) Ror2 and Dvl2 are indicated. (C)
Ror2À/À
MEFs stably expressing Flag-tagged
Ror2 were transduced with GFP or Shisa
lentiviruses and treated with control or
Wnt5a CM overnight, followed by immunoprecipitation
using anti-Flag antibody.
Immunoblot analysis was performed using the indicated antibodies. (D) Two days after transfection with control or Fzd2,4,5
siRNAs, 293T cells were treated for 3 h with Wnt3a or Wnt5a CM. Immunoblot analysis was performed using anti-Ror2 antibody.
(E–G) Wnt3a and Wnt5a couple Fzd to LRP6 or Ror2, respectively. CM containing receptor extracellular domains (LRP6N-Fc, Fzd8CRD-Flag,
and Ror2-extr-HA) or Wnt ligands were mixed as indicated, and immunoprecipitation was performed with anti-Fc (E),
anti-Flag (F), or anti-HA (G) antibodies, followed by immunoblot analysis using the indicated antibodies.
Grumolato et al.
2520 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
GSK3-specific inhibitors, SB216763 and BIO, each dramatically
decreased the ability of Wnt3a and Wnt5a to
specifically trigger phosphorylation of LRP6 or Ror2, respectively,
under the same conditions in Saos-2 cells
(Supplemental Fig. 6B,C). All of these findings strongly
supported a model in which the binding of Fzd to either
canonical or noncanonical coreceptors mediated by Wnt
ligands specific for each coreceptor leads to their phosphorylation
by the same intracellular machinery.
Wnt5a-induced phosphorylation of Ror2 on Ser 864
is required for noncanonical coreceptor activity
It has been shown that GSK3 phosphorylation serves as
a priming event, followed by CK1 phosphorylation, in the
activation of LRP6 (MacDonald et al. 2009). We used
a bioinformatics approach to analyze the Ror2 intracellular
domain and found various GSK3 and CK1 putative
phosphorylation sites conserved in vertebrates (Supplemental
Fig. 7A). We individually substituted alanine for
the two Ser residues that showed the highest scores as
GSK3 sites and exogenously expressed wild-type Ror2 or
these mutants in rat1 cells, which lack detectable endogenous
Ror2 expression (data not shown). Figure 3E shows
that the Ror2-S864A mutant failed to exhibit the Wnt5ainduced
upshift indicative of phosphorylation, while the
S772A mutant showed the characteristic upshift. The
S864A mutation also inhibited Ror2 in vitro phosphorylation
by GSK3 (Supplemental Fig. 7B). Furthermore,
Figure 3. The same intracellular machinery
is used for Wnt-induced phosphorylation
of LRP6 and Ror2. (A) Colocalization of
endogenous Ror2 (green) and Dvl2 (red)
upon Wnt5a treatment (2 h) in Saos-2 cells.
(B) Dvl proteins are required for LRP6 and
Ror2 phosphorylation. Saos-2 cells were
transduced with shRNA lentiviral vectors
targeting the three Dvl genes or a control
vector and treated for 2 h with Wnt CM,
followed by immunoblot analysis. Nonphosphorylated
(arrow) and phosphorylated
(asterisk) Ror2 and Dvl2,3 are indicated. (C)
Axin is involved in ligand-induced triggering
of both Ror2 and LRP6. Two days after
transduction with control or shAxin lentiviruses,
Saos-2 cells were treated for 3 h
with control, Wnt3a, or Wnt5a CM followed
by immunoblot analysis. (D) Wnt5a
induces formation of a complex including
endogenous Ror2 and axin. MDA-MB157
cells were treated for 90 min with CM,
followed by Ror2 immunoprecipitation and
immunoblot analysis with anti-axin and
anti-Ror2 antibodies. (E) Effects of putative
GSK3 phosphorylation site mutations on
Ror2 upshift. Two days following transfection
with the indicated Ror2 constructs,
Rat1 cells were treated with control or
Wnt5a CM for 2 h, followed by immunoblot
analysis. (F) A slot blot generated using serial
dilutions of Ror2 nonphosphopeptide and
phosphopeptide was stained with Ponceau
solution (loading control), followed by immunoblot
analysis with anti-non-PhosphoRor2
antibody. (G) Ror2À/À
MEFs stably
expressing Flag-tagged wild-type (wt) or
S864A Ror2 constructs were treated with
control or Wnt5a CM, followed by immunoprecipitation
with anti-Flag antibody and
immunoblot analysis. (H) Ror2À/À
MEFs
stably expressing Flag-tagged Ror2 were
treated with control or Wnt5a CM followed
by immunoprecipitation with anti-Flag antibody.
The immunoprecipitates were treated
in the presence or absence of l protein phosphatase, and immunoblot analysis was performed. (I) Ror2 Tyr kinase activity is not
required for Wnt5a-induced Ror2 phosphorylation. Ror2À/À
MEFs stably expressing Flag-tagged Ror2 constructs were treated with
control or Wnt5a CM, and immunoblot analysis was performed using anti-Ror2 and anti-b-catenin (loading control) antibodies.
Mechanism of Ror1/2 receptor activation
GENES & DEVELOPMENT 2521
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
a peptide antibody that specifically recognized only the
nonphosphorylated Ror2 Ser 864 (Fig. 3F) detected wildtype
Ror2 in the absence of Wnt5a treatment, but failed to
recognize Wnt5a-triggered wild-type Ror2 or the S864A
mutant receptor expressed at similar levels (Fig. 3G). Of
note, the ability of this non-Phospho-Ror2 antibody to
detect Wnt5a-triggered Ror2 was restored by treating the
cell lysate with l protein phosphatase (Fig. 3H). Together,
these results demonstrate that Wnt5a promotes the
phosphorylation of Ror2 on Ser 864. Of note, a kinasedead
Ror2 construct containing a mutation in the ATPbinding
domain was phosphorylated similarly to wildtype
Ror2 in response to Wnt5a, while, under the same
conditions, the upshift of Ror2-S864A was strongly impaired
(Fig. 3I), confirming that Ror2 Tyr kinase activity
is not required for the triggering of this receptor by
Wnt5a.
Wnt5a has been reported to stimulate polarized cell
migration by a mechanism involving Ror2 and Dvl
(Nishita et al. 2006; Schlessinger et al. 2007; Nomachi
et al. 2008). Figure 4A shows that Wnt5a stimulated the
migration of rat1 cells containing wild-type Ror2 but not
the same cells expressing Ror2-S864A at comparable
levels. The kinase-dead Ror2 behaved similarly to the
wild-type receptor (Supplemental Fig. 8A; Nishita et al.
2006), indicating that Ror2 Tyr kinase activity is dispensable
for this function. Conversely, Ror2 mutants with
S864D or S864E substitutions, which can mimic phosphorylated
residues, increased rat1 cell motility even in
the absence of Wnt5a treatment (Fig. 4B). Consistent with
these findings, exogenous expression of wild-type Ror2 in
MEFs derived from mice with inactivation of the Ror2
gene (DeChiara et al. 2000) partially rescued their impaired
cell motility, while Ror2-S864A had no effect
(Supplemental Fig. S8B,C). These results provide biological
evidence that Wnt-induced phosphorylation of Ror2
is required for activation of noncanonical signaling functions
involved in cell motility.
Identification of Ror1 as another noncanonical
receptor specifically phosphorylated
in the presence of Wnt5a
Wnt5a has been reported to bind Ror1 in vitro (Fukuda
et al. 2008). Our in vivo analysis revealed that Wnt5a, but
not Wnt3a, provoked an upshift of endogenous Ror1 and
stimulated its phosphorylation on Ser/Thr residues in
HCC1500 and HCC3153 breast cancer cells (Fig. 5A,B).
Moreover, cotreatment with BIO or SB216763 inhibited
both the upshift of Ror1 specifically induced by Wnt5a
and LRP6 phosphorylation induced by Wnt3a (Fig. 5C;
Supplemental Fig. 9), strongly suggesting that the mechanism
involved in Wnt5a activation of Ror2 applied to
Ror1 as well. To confirm this conclusion, we generated a
Ror1–LRP6 chimeric receptor containing the Ror1 extracellular
domain with LRP6 transmembrane and cytoplasmic
domains. Figure 5D shows that Wnt5a stimulated
b-catenin/TCF reporter activity in cells expressing Ror1–
LRP6, but not vector control cells, indicating that this
prototype noncanonical Wnt ligand functionally coupled
the Ror1 moiety of the chimeric receptor to endogenous
Fzd and connected the chimeric receptor to canonical
signaling. Furthermore, Wnt5a specifically induced phosphorylation
of Ror1–LRP6, while Wnt3a instead promoted
Figure 4. Ror2 phosphorylation is required
for Wnt5a-induced cell migration.
(A) Confluent rat1 cells stably expressing
GFP, Ror2, or Ror2-S864A were wounded
and treated with control or Wnt5a CM
(0.5% of fetal bovine serum [FBS]), and
the relative migration of the wound edge
was measured after 15 h. Mean values 6
SEM (n = 18) from one representative of
five independent experiments are shown.
(B) Rat1 cells stably expressing GFP, Ror2,
Ror2-S864E, or Ror2-S864D were wounded
and incubated in 0.5% FBS medium, and
the relative migration was measured after
14 h. Mean values 6 SEM (n = 22) from one
representative of three independent experiments
are shown. (***) P < 0.001 compared
with all other conditions (A) or with
GFP-expressing cells (B) (Student’s t-test).
Immunoblot analysis was performed to
assess the expression levels of Ror2 con-
structs.
Grumolato et al.
2522 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
the phosphorylation of endogenous LRP6 (Fig. 5E). Together,
these results establish that Wnt5a activates Ror1
by the same mechanism responsible for Wnt3a activation
of canonical signaling through LRP5/6 and Wnt5a triggered
noncanonical signaling through Ror2.
Cross-inhibition at the cell membrane between
canonical and noncanonical Wnts
Recent studies using exogenous Ror2 overexpression indicated
that Wnt5a inhibits the canonical Wnt pathway
downstream from b-catenin by a mechanism involving
signaling through Ror2 (Mikels and Nusse 2006; Mikels
et al. 2009). We found that neither Ror1/2 shRNAmediated
knockdown (Fig. 6A) nor Ror2 knockout (Supplemental
Fig. 10A) affected the ability of Wnt5a to
inhibit Wnt3a-induced b-catenin/TCF reporter activity,
indicating that endogenous Ror1/2 are not strictly required
for inhibition of the b-catenin pathway by Wnt5a.
Based on our evidence that canonical and noncanonical
Wnt receptor activation involves shared use of Fzd, we
reasoned instead that an excess of either type of Wnt
might interfere with the other pathway as a result of
ligand competition for binding to Fzd. Consistent with
this hypothesis, Wnt5a antagonized Wnt3a interaction
with Fzd in an in vitro binding assay (Supplemental Fig.
10B). Moreover, Wnt5a antagonized both LRP6 phosphorylation
and b-catenin stabilization induced by Wnt3a
(Fig. 6B) and mimicked the effects of Dkk1 in decreasing
b-catenin/TCF reporter activity in tumor cells with Wnt
autocrine pathway activation, but not in those containing
a b-catenin-activating mutation (Fig. 6C). Reciprocally,
Figure 6, D and E, and Supplemental Figure 10C show
that cotreatment with Wnt3a inhibited the ability of
Wnt5a to trigger phosphorylation of both endogenous
and exogenous Ror2, as assessed by analysis of Ror2
upshift or anti-phospho-Ser/Thr antibody. All of these
findings demonstrate that canonical and noncanonical
Wnts cross-compete for activation of the reciprocal
pathway at the cell surface.
Human mesenchymal stem cells (hMSCs) can give rise
in vitro and in vivo to different lineages, including osteoblasts,
chondrocytes, myocytes, and adipocytes (Pittenger
et al. 1999). We showed that canonical Wnts maintain
hMSCs in a proliferative and undifferentiated state, while
Wnt5a promotes their osteogenic differentiation (Liu
et al. 2009). In fact, decreased Ror2 expression inhibited
differentiation of these cells (Supplemental Fig. 11A; Liu
et al. 2007) in accordance with a role of this receptor in
osteogenic differentiation (Green et al. 2008). Cocultivation
of hMSCs with vector control or Ror2–LRP6-expressing
reporter cells revealed evidence of constitutive noncanonical
Wnt ligand activity in hMSCs (Fig. 7A). To investigate
whether inhibition of osteogenic differentiation
induced by high levels of Wnt3a might be caused in part
by canonical Wnt ligand interference with endogenous
Ror2 signaling at the receptor level, we inhibited the
canonical pathway downstream from LRP5/6 by expression
of either lentiviral b-catenin shRNA or dominantnegative
TCF4 (DN-TCF4). Despite the resulting abrogation
of Wnt3a-induced b-catenin/TCF reporter activity
(Fig. 7B; Supplemental Fig. 11C), we observed no impairment
in the ability of Wnt3a to decrease hMSC differentiation
(Fig. 7C; Supplemental Fig. 11B). These results
imply that Wnt3a can inhibit osteogenic differentiation
by directly antagonizing Wnt5a/Ror2 signaling at the
receptor level, further supporting the concept of crosscompetition
at the cell surface between canonical and
noncanonical Wnts.
Discussion
Our present studies establish a novel mechanism by
which distinct classes of Wnt ligands activate very
Figure 5. Ror1 is specifically phosphorylated by Wnt5a. (A)
HCC1500 cells were treated for 3 h with control, Wnt3a, or
Wnt5a CM, followed by immunoblot analysis. The nonphosphorylated
(arrow) and phosphorylated (asterisk) Ror1 is indicated.
(B) HCC3153 cells were treated for 3 h with control
or Wnt5a CM, followed by immunoprecipitation with anti-Ror1
antibody and immunoblot analysis. (C) HCC1500 cells were
pretreated for 1 h with or without the GSK3-specific inhibitor
BIO (20 mM), followed by 2 h treatment with control, Wnt3a, or
Wnt5a CM in the presence or absence of BIO (20 mM). Immunoblot
analysis was performed with the indicated antibodies. (D)
Rat1 cells cotransfected with SuperTop luciferase reporter,
renilla luciferase plasmid, and Ror1–LRP6 or control vector
were treated overnight with control or Wnt5a CM, and a luciferase
assay was performed. The renilla-normalized values are
expressed as fold increase compared with the vector-transfected
control cells and correspond to the mean values (6SD, n = 4) of
one representative of three independent experiments. (E) Rat1
cells transfected with Ror1–LRP6 were treated for 3 h with
control, Wnt3a, or Wnt5a CM, and immunoblot analysis was
performed. The bands close to the 225-kDa and 76-kDa markers
correspond to endogenous LRP6 and Ror1–LRP6, respectively.
Mechanism of Ror1/2 receptor activation
GENES & DEVELOPMENT 2523
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
different signaling pathways determined by their specific
abilities to couple a common receptor, Fzd, with completely
unrelated endogenous coreceptors. Moreover, we
established that the mechanism of activation of these
unrelated receptors involves their Ser/Thr phosphorylation
by shared intracellular components (Fig. 8). As one
demonstration of this model, we were able to convert
Wnt5a from a noncanonical to a canonical ligand by
switching the extracellular domain of LRP6 with Ror1
or Ror2 external domains. This allowed Wnt5a, but not
Wnt3a, to specifically induce Ser/Thr phosphorylation of
the LRP6 intracellular domain of the Ror/LRP6 chimeric
receptor and consequently activate the canonical pathway.
Previous studies involving overexpression of cotransfected
Wnt ligands and Ror2 led to evidence that Ror2
could bind various Wnts, including canonical Wnt1,
Wnt3a, and Wnt8 (Hikasa et al. 2002; Billiard et al.
2005). In contrast, in vitro analysis indicated that this
Figure 6. Cross-competition at the cell surface between canonical and noncanonical Wnt ligands. (A) 293T cells stably expressing
control or Ror1/2 shRNAs and transfected with SuperTop luciferase reporter and renilla luciferase plasmid were treated overnight with
Wnt3a in the presence or the absence of Wnt5a CM in a 1-to-2 ratio, and a luciferase assay was performed. The renilla-normalized
values are expressed as fold decrease compared with the Wnt3a-treated cells and correspond to the mean values (6SD, n = 4) of one
representative of two independent experiments. Ror1/2 knockdown was assessed by immunoblot analysis. (B) Wnt5a inhibits Wnt3ainduced
phosphorylation of LRP6 and b-catenin stabilization. Wnt5a 293T cells were treated for 3 h with Wnt3a and/or Wnt5a CM in
a 1-to-2 ratio, and the cell lysates were subjected to immunoblot analysis or GST-E-cadherin pull-down for analysis of uncomplexed
b-catenin. (C) PA1 ovarian teratocarcinoma cells (autocrine Wnt signaling) and AGS gastric adenocarcinoma cells (activating b-catenin
mutation) were transduced with lentiviral mTOP- or mFOP-luciferase reporters together with a PGK-driven renilla luciferase virus and
treated for 22 h with Wnt5a, Dkk1, or their corresponding control CM. The reporter activity was calculated by dividing the mTOP/
renilla ratio by the mFOP/renilla ratio, and the mean values (6SEM, in triplicate) of one representative of two independent experiments
are shown. (**) P < 0.01; (***) P < 0.001 compared with control (Student’s t-test). (D,E) Wnt3a inhibits Wnt5a-induced phosphorylation
of Ror2. (D) Ror2À/À
MEFs stably expressing Flag-tagged Ror2 were treated with Wnt5a in the presence or absence of Wnt3a CM in a
1-to-2 ratio, followed by immunoprecipitation with anti-Flag antibody and immunoblot analysis. (E) Saos2 cells were treated with the
indicated amounts of purified Wnt5a and Wnt3a for 2 h, followed by immunoblot analysis for endogenous Ror2 and LRP6.
Nonphosphorylated (arrow) and phosphorylated (asterisk) Ror2 are indicated.
Grumolato et al.
2524 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
receptor specifically interacted only with noncanonical
Wnt5a (Oishi et al. 2003; Mikels and Nusse 2006). We
demonstrated here that Wnt5a, but not Wnt3a, physically
bridged Ror2 to Fzd extracellular domains. Conversely,
under the same conditions, Wnt3a, but not Wnt5a, coupled
Fzd to LRP6. While LRP5/6 coreceptors bind canonical
Wnts through their large extracellular b-propeller
domains (Clevers 2006; MacDonald et al. 2009), both Fzd
and Ror1/2 bind Wnt by means of their cysteine-rich
domains (CRDs) (Logan and Nusse 2004; Green et al.
2008). This 120- to 125-amino-acid domain is present in
a variety of proteins, including Fzd and Smoothened
seven-transmembrane domain receptors, Ror1/2 and the
muscle-specific kinase (MuSK) RTKs, and the secreted
Fzd-related proteins (FRPs) (Green et al. 2008). Previous
biochemical and crystallographic studies have suggested
that the CRDs of Fzd and FRP could interact to form
dimers (Bafico et al. 1999; Dann et al. 2001). In this regard,
it has been reported that cotransfected Ror2 and Fzd
coimmunoprecipitated in the absence (Oishi et al. 2003)
or presence of Wnt5a (Nishita et al. 2010b). Under stringent
in vitro binding conditions, we showed that Wnt5a is
required for the formation of a complex containing the
Fzd8 CRD and the Ror2 extracellular domain. These findings
are consistent with predictions from the crystal structure
of Fzd and FRP CRDs that a ligand is likely required to
increase the otherwise weak dimerization affinity of this
type of domain (Dann et al. 2001). While our in vitro
binding data do not exclude formation of Ror1/2 and Fzd
homodimers in the presence of Wnt5a, our findings that
Ror2 Ser/Thr phosphorylation requires endogenous Fzd
and is not triggered by forced antibody-induced dimerization
strongly argue that Ror1/2 activation depends on
Wnt5a-induced clustering of Ror1/2 and Fzd. Further studies
will be needed to determine the stoichiometry of Ror2–
Wnt5a–Fzd and LRP6–Wnt3a–Fzd tripartite complexes,
and the structural determinants responsible for the specific
coreceptor-binding properties of different Wnts.
Dvl, a core component of both canonical and noncanonical
Wnt signaling, possesses N-terminal DIX, central
PDZ, and C-terminal DEP domains (Wallingford and
Habas 2005). Whereas DIX and PDZ are required for
canonical Wnt signaling, the PDZ and DEP domains have
Figure 7. b-Catenin is dispensable for Wnt3a-mediated inhibition
of hMSC osteogenic differentiation. (A) hMSCs display
endogenous noncanonical activity. Rat1 cells were cotransfected
with vector (V) or Ror2–LRP6 (R–L) together with SuperTop
or SuperFop and renilla luciferase plasmids. One day after
transfection, reporter Rat1 cells were transferred to plates
containing undifferentiated or differentiated (4 d in OS medium)
hMSCs, parental L cells (negative control), or Wnt5a-expressing
L cells (positive control) for an additional day before the
luciferase assay. Paracrine Ror2–LRP6 mean activities (6SD)
from one representative of two experiments are expressed as
a ratio between Top and Fop reporters, both normalized with the
renilla luciferase values. (B) hMSCs containing the mTOP- or
mFOP-luciferase reporters were transduced with a lentiviral
shRNA vector targeting b-catenin or control vector and were
treated overnight with or without purified Wnt3a (200 ng/mL)
in OS medium. The reporter activity was calculated by dividing
the mTOP/renilla ratio by the mFOP/renilla ratio, and the mean
values (6SEM, n = 4) of one representative of two independent
experiments are represented. b-Catenin down-regulation was
assessed by immunoblot analysis. (C) hMSCs transduced with
control or b-catenin shRNA lentiviruses were grown in normal
medium or OS differentiation medium with or without purified
Wnt3a (200 ng/mL). hMSC osteogenic differentiation was
assessed for alkaline phosphatase activity by staining (top panel,
10 d of differentiation) or using a luminescent substrate (bottom
panel, 6 d of differentiation). The results represent the mean
values (6SEM, n = 5) of one representative of three independent
experiments. (*) P < 0.05 compared with hMSCs differentiated
in the absence of Wnt3a (Student’s t-test). Figure 8. Model for receptor activation by canonical and noncanonical
Wnt ligands. See the text for details.
Mechanism of Ror1/2 receptor activation
GENES & DEVELOPMENT 2525
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
been shown to participate in noncanonical pathways
(Wallingford and Habas 2005). These results have suggested
that Dvl acts as a molecular switch to activate
canonical or noncanonical signaling (Wallingford and
Habas 2005). Our present findings that Dvl knockdown
inhibited the phosphorylation of Ror2 as well as that of
LRP6 (Bilic et al. 2007; Zeng et al. 2008) imply that this
scaffold protein plays an important role in the activation
of both classes of Wnt coreceptors. Moreover, a Dvl construct
containing only the PDZ and DEP domains served
as a dominant negative for Wnt-triggered phosphorylation
of both Ror2 and LRP6 (Supplemental Fig. 4B–D).
While the DIX domain is known to be involved in Wnt3ainduced
phosphorylation of LRP6 (Zeng et al. 2008;
Metcalfe et al. 2010), these results argue that this domain
also plays an important role in the triggering of noncanonical
coreceptors. Of note, Dvl has been shown to
activate b-catenin signaling in the absence of LRP5/6,
while expression of the Dvl DEP domain alone stimulated
the noncanonical pathway (Wallingford and Habas
2005; Metcalfe et al. 2010). In light of our present
findings, it is tempting to speculate that Dvl plays a dual
role in both canonical and noncanonical signaling: one as
a common component of the shared machinery involved
in coreceptor activation, and another downstream and
specific for each pathway.
Axin is a crucial component of the b-catenin destruction
complex (Clevers 2006; MacDonald et al. 2009). This
protein also acts as a scaffold protein required for GSK3
phosphorylation of LRP5/6 (Zeng et al. 2008). Our studies
provide the first demonstration that axin is required for
activation of noncanonical Ror2 as well, and that Wnt5a
induces the formation of a complex containing endogenous
Ror2 and axin. These findings suggest that, following
Wnt3a and Wnt5s triggering of Fzd, Dvl and axin
allow the recruitment of GSK3 into proximity with
LRP5/6 or Ror1/2, respectively, to induce their phosphorylation
(Fig. 8). This mechanism is consistent with the
‘‘initiation and amplification’’ model proposed for LRP5/6
activation (Baig-Lewis et al. 2007; Zeng et al. 2008;
MacDonald et al. 2009) and would generalize it to noncanonical
signaling. Following specific coreceptor activation,
the two pathways likely diverge such that, in
canonical Wnt signaling, the phosphorylated intracellular
domain of LRP5/6 binds to axin and inhibits GSK3
phosphorylation of b-catenin, resulting in its stabilization
(MacDonald et al. 2009). In the Wnt5a noncanonical
pathway, the phosphorylated cytoplasmic tails of Ror1
and Ror2 may act as docking sites for other, yet-to-be
identified, cytoplasmic effectors.
The Ser/Thr kinase GSK3 is involved in a variety of
cellular processes and signaling pathways (Kockeritz
et al. 2006). While GSK3 has been reported to participate
in Ror2 phosphorylation (Yamamoto et al. 2007), this
study provided no evidence for mechanistic or biological
implications. Using a phospho-specific antibody generated
by us, we identified Ror2 Ser 864 as a GSK3 site
phosphorylated in response to Wnt5a. Ser 864 most likely
represents the noncanonical equivalent of the GSK3priming
phosphorylation sites in LRP6 required for further
phosphorylation and activation of this canonical
coreceptor (Zeng et al. 2005). Indeed, we demonstrated
that Ser 864 inactivating and phosphomimetic substitutions,
respectively, abolished and enhanced the effects of
Ror2 on cell motility. Together, these findings provide the
first demonstration that Wnt5a-induced Ser phosphorylation
is functionally required for activation of Ror biological
function.
The role of the Ror1/2 Tyr kinase domain in noncanonical
Wnt signaling is controversial (Green et al.
2008). High, nonphysiological levels of Wnt5a were required
to induce Ror2 phosphorylation on Tyr, but not Ser/
Thr, residues (Liu et al. 2008), while others (Yamamoto
et al. 2007) reported that Wnt5a concentrations sufficient
to trigger Ror2 Ser/Thr phosphorylation did not promote
Tyr phosphorylation. We showed that Wnt5a induced
Ser/Thr, but not Tyr, phosphorylation of Ror2 under
conditions in which this receptor exhibited biological
function. Conversely, forced Ror2 dimerization by an
antibody resulted in Tyr phosphorylation in the absence
of Ser/Thr phosphorylation. Furthermore, we showed
that wild-type and kinase-dead Ror2 acted similarly upon
Wnt5a treatment in both biochemical and functional experiments.
These findings as well as other evidence presented
here and elsewhere (Nishita et al. 2006; Yamamoto
et al. 2007) argue strongly that Wnt5a does not trigger
endogenous Ror by the typical mechanism of RTK dimerization
to activate noncanonical signaling, but instead
couples completely unrelated Ror and Fzd coreceptors.
The fact that the Ror1/2 Tyr kinase domain has been
conserved throughout evolution implies its likely biological
role. One possibility is that Ror1/2 has functions
independent of any triggered by Wnts. Alternatively, it is
possible that, while not directly required for ligand-induced
receptor activation, the Ror1/2 Tyr kinase domain
may be involved in the phosphorylation of downstream
effectors in certain aspects of its signaling. In this regard,
it has been reported that a kinase-dead Ror2 construct
was unable to rescue the decreased expression of paraxial
protocadherin induced by Ror2 morpholinos in Xenopus
embryos (Schambony and Wedlich 2007).
While it is well established that Wnt5a can inhibit the
canonical Wnt pathway, the mechanism involved is
controversial, with studies reporting effects either downstream
from or upstream of b-catenin (Ishitani et al. 2003;
Topol et al. 2003; Westfall et al. 2003; Mikels and Nusse
2006; Mikels et al. 2009; Sato et al. 2010). We found that
Wnt5a and Wnt3a cross-competed for Fzd binding in
vitro, and that Wnt5a antagonized Wnt3a signaling in
the absence of a requirement for endogenous Ror1/2,
consistent with cell surface Wnt ligand competition for
Fzd binding (Sato et al. 2010). Moreover, our present
findings that Wnt3a similarly inhibits the phosphorylation
of Ror2 induced by Wnt5a provides strong evidence
that competition at the level of Fzd binding is involved in
reciprocal regulation of canonical and noncanonical pathways.
Finally, we showed that Wnt3a inhibited hMSC
osteogenic differentiation independently of a requirement
for b-catenin activation, implying that canonical Wnt
ligands can antagonize the Ror2 pathway at the cell
Grumolato et al.
2526 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
surface, and that this effect has biological relevance in
cell fate specification. Wnt canonical signaling has important
functions in normal bone development (Grigoryan
et al. 2008; Liu et al. 2009), and our findings suggest that
canonical ligands may act in part by inhibiting the Ror2
pathway, whose role in bone development is strongly
supported by analysis of knockout mice as well as identification
of Ror2 mutations in human skeletal syndromes
(Green et al. 2008).
Other signaling pathways, such as TGF-b/BMP (Feng
and Derynck 2005), share common as well as distinct
coreceptors, but, to our knowledge, none show the unique
strategy identified here involving use of the same intracellular
components to activate completely unrelated
coreceptors. Given the large number of Wnts and the
existence of various receptors containing Wnt-binding
domains (Green et al. 2008; Angers and Moon 2009), as
well as an increasing array of putative noncanonical
pathways (Semenov et al. 2007), the common mechanism
of coreceptor activation delineated here may represent
a general paradigm underlying the activation of other,
yet-to-be characterized, Wnt pathways. In this regard, we
found that, unlike Wnt5a, other noncanonical Wnts such
as Wnt4 and Wnt11 were unable to trigger Ror1–LRP6
or Ror2–LRP6 fusion receptors (L Grumolato and SA
Aaronson, unpubl.), suggesting that these ligands may
signal through other noncanonical coreceptor(s). Of note,
it has been reported recently that Wnt11 functions
through Dvl and the Musk receptor during the establishment
of zebrafish neuromuscular junctions (Jing et al.
2009). In our model (Fig. 8), the combinatorial nature
of Wnt ligand/coreceptor interactions, together with a
shared activation mechanism involving a limited number
of transduction molecules, allows specificity of signaling
through the diversity of ligands, while preserving the
principle of economy in receptor activation.
Materials and methods
Cell culture, transfection, and lentivirus production
Cell lines were maintained in Dulbecco’s modified Eagle’s
medium (Lonza) supplemented with 10% fetal bovine serum
(Sigma). Control, Wnt3a, and Wnt5a conditioned media (CM)
were collected from L cells (American Type Culture Collection
[ATCC]) grown according to ATCC instructions. Purified Wnt3a
and Wnt5a were purchased from R&D Systems. SB216763 was
purchased form Cayman Chemical. BIO was purchased from
Sigma. Control and Dkk1 CM were generated by transient
transfection of 293T cells with pcDNA3 or Dkk1 (pcDNA3)
and collection of the media 2 d and 3 d after transfection. Ror2À/À
MEFs were derived from mice containing a truncation of Ror2
induced by the insertion of a lacZ cassette after the transmembrane
domain that behave as null mutants (DeChiara
et al. 2000). We showed previously that PA1 ovarian cancer cells
exhibit autocrine activation of Wnt signaling (Bafico et al. 2004),
while AGS gastric cancer cells contain a b-catenin-activating
mutation (Caca et al. 1999). Primary hMSC cultures (Lonza) were
expanded in basal medium according to the manufacturer’s
instructions or were differentiated using osteogenic medium
(Lonza) diluted 1:3 in normal medium. Alkaline phosphatase
activity was detected histochemically with the Leukocyte
Alkaline Phosphatase kit (Sigma-Aldrich) or quantified with a
luminometer using a chemiluminescent substrate (BioFX Laboratories)
and normalized to total proteins.
Transient transfection was performed using Fugene 6 (Roche)
according to the manufacturer’s instructions or with polyethylenimine
(Polysciences). For lentivirus production, 293T cells
were cotransfected with the lentiviral vector and pCMV D8.91
and pMD VSV-G plasmids (kindly provided by I. Weissman). Two
days after transfection, CM was collected, supplemented with
8 mg/mL polybrene, and added for overnight incubation to freshly
plated cells to be transduced. Two days after infection, the cells
were selected in 2 mg/mL puromycin or 10 mg/mL blasticidin.
siRNAs (Dharmacon) were used to knock down the three
predominant Fzd coreceptors in 293T cells—i.e., Fzd2, Fzd4, and
Fzd5—as reported previously (Pan et al. 2008). 293T cells were
transfected using Lipofectamine 2000 (Invitrogen) according to
the manufacturer’s instructions, and 2 d later the cells were
treated with CM or purified Wnt3a and Wnt5a (R&D Systems).
As a control, an siRNA targeting GFP was used.
Constructs
The DKK2C–Wnt5a chimera was generated by fusion of the LRPbinding
domain of mouse Dkk2 (amino acids 141–259) and
Wnt5a (Liu et al. 2005). The Ror2–LRP6 and Ror1–LRP6 constructs
(pcDNA3) were generated by fusion of the human Ror2
or Ror1 extracellular domains and human LRP6 transmembrane
and intracellular domains. Human Dkk1 was cloned in
pcDNA3-Flag. The b-catenin/TCF reporter Super8XTOPFlash,
containing eight TCF-responsive elements driving the firefly
luciferase gene, was kindly provided by R. Moon. The LRP6
extracellular domain fused to the IgG heavy chain (extLRP6-Fc)
was cloned in pcDNA3 vector, and the CRD of Fzd8 was cloned
in pcDNA3-Flag (Fzd8-CRD-Flag). All lentiviral shRNA constructs
were generated in VIRHD-EP (kindly provided by L.
Gusella) containing either puromycin or blasticidin resistance
genes, mostly using target sequences from the literature (sequences
available on request). Other constructs generated using
this lentiviral vector include the extracellular domain of human
Ror2 (amino acids 1–394) with an HA tag at the C terminus
(extRor2HA); a dominant-negative Dvl (including the PDZ and
DEP domains of human Dvl1) fused to GFP at the N terminus;
human wild-type Ror2; Ror2-S864A, Ror2-K507A, Ror2-S864D,
and Ror2-S864E with an N-terminal Flag tag; dominant-negative
TCF4 fused to GFP; and Shisa, cloned by reverse transcription
and PCR from Xenopus embryos and containing a myc tag at the
C terminus. GSK-3 putative phosphorylation sites in the Ror2
intracellular domain were identified using the Scansite motif
Scanner, and Ror2 mutant constructs were generated using the
QuikChange II Site-Directed Mutagenesis kit (Stratagene) with
N-terminal Flag-tagged Ror2 constructs (pCS2 vector or VIRHDEP
lentiviral vector) as templates. The lentiviral b-catenin/TCF
luciferase reporter containing 14 wild-type (MegaTop-LentiLuc)
or mutant (MegaFop-LentiLuc) TCF consensus sequences was
generated by PCR from a seven-repeat-containing construct
described previously (Akiri et al. 2009). All constructs were
sequence-verified.
Antibodies
The anti-LRP5 mouse monoclonal antibody was generated following
the procedure previously used for the anti-LRP6 antibody
(Khan et al. 2007). Nonphospho-Ror2 antibody was generated by
Proteintech using the peptide VPKPSSHHS*GSGSTS (asterisk
indicates phosphorylated residue) conjugated to keyhole limpet
hemocyanin. The bleeds were purified on a nonphosphopeptide
Mechanism of Ror1/2 receptor activation
GENES & DEVELOPMENT 2527
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
column using the CarboxyLink Immobilization kit (Thermo
Fisher Scientific), and the specificity of the antibody was assessed
by slot blot using serial dilutions of the nonphosphopeptide and
phosphopeptide. Rabbit anti-LRP6, rabbit anti-phospho-LRP6
(S1490), rabbit anti-Wnt5a, rabbit anti-phospho-Thr (Cell Signaling
Technology), rabbit anti-Dvl2, mouse anti-Dvl3 (Santa Cruz
Biotechnology), mouse anti-Flag, mouse anti-tubulin (Sigma),
rabbit anti-phospho-Ser (Invitrogen), goat anti-Ror2, goat antiRor1,
rat anti-Wnt3a, goat anti-axin (R&D Systems), mouse antiphospho-Tyr
clone 4G10 (Millipore), and mouse anti-b-catenin
(BD Biosciences) were from commercial sources. Mouse antimyc
clone 9E10 was obtained from the Mount Sinai Hybridoma
Core Facility.
Free b-catenin, PIP2 quantification, and luciferase assays
Uncomplexed b-catenin was measured by the GST-E-cadherinbinding
assay as described previously (Liu et al. 2003). The
amounts of PIP2 were measured by ELISA as described previously
(Pan et al. 2008). For luciferase assay, cells transfected or transduced
with the b-catenin/TCF firefly luciferase reporters and the
renilla luciferase control reporters were lysed, and luciferase
activity was measured using the Dual-Luciferase Reporter Assay
System (Promega) according to the manufacturer’s instructions.
In vitro binding assay
CM was collected from 293 cells stably expressing extLRP6-Fc;
L cells stably expressing extRor2-HA, Wnt3a, or Wnt5a; and 293T
cells transiently transfected with Fzd8-CRD-Flag. ExtLRP6-Fc,
Fzd8-CRD-Flag, and extRor2-HA CM were purified overnight
with anti-Fc (Sigma), anti-Flag (Sigma), or anti-HA (Covance)
beads, respectively. The beads were then washed and incubated
with equal volumes of Wnt CM. After washing, the beads were
incubated overnight with CM containing the other coreceptor
and washed. Loading buffer was then added to the beads, and
SDS-PAGE was performed, followed by immunoblot analysis.
Wound healing assay
Confluent rat1 cells and MEFs plated on fibronectin-coated
(Sigma) wells were maintained for 1 d (rat1) or overnight in
0.5% FBS medium. The cells were then scratched with a sterile
1-mL tip, washed twice in PBS, and incubated with control or
Wnt5a medium containing 0.5% FBS. Photographs of the cells
were taken at time 0 and 15 h (rat1) or 8 h (MEFs) after scratch
using an inverted microscope. The relative distance was calculated
using ImageJ (http://rsb.info.nih.gov/ij/index.html).
Acknowledgments
We thank L. Gusella, I. Weissman, R. Moon, A. Yamamoto, A.
Yaniv, A. Gazit, T. Villanueva de la Torre, Z. Khan, B. Zhao,
S.K.A. Divakar, G. Akiri, and S. Asciutti for providing reagents.
We also thank M. Mlodzik, S. Sokol, H. Hikasa, J. Chen, L. Chen,
M. Swan, J. Lee, R. Diotti, C. Munoz-Fontela, and S.K. Mungamuri
for helpful comments and discussion. L.G. is recipient of a DOD
Prostate Cancer Post-doctoral Fellowship (PC050673) and a New
York State Department of Health Breast Cancer Research Postdoctoral
Fellowship Award. G.L. received support from the Leo
and Julia Forchheimer Foundation and Breast Cancer Alliance,
Inc. S.V. received a Post-doctoral Fellowship from the American
Urological Association Foundation. This study was supported by
grants from the Breast Cancer Research Foundation, the Leo and
Julia Forchheimer Foundation, and the National Cancer Institute
(CA071672 and T32CA078207).
References
Akiri G, Cherian MM, Vijayakumar S, Liu G, Bafico A, Aaronson
SA. 2009. Wnt pathway aberrations including autocrine Wnt
activation occur at high frequency in human non-small-cell
lung carcinoma. Oncogene 28: 2163–2172.
Angers S, Moon RT. 2009. Proximal events in Wnt signal
transduction. Nat Rev Mol Cell Biol 10: 468–477.
Bafico A, Gazit A, Pramila T, Finch PW, Yaniv A, Aaronson SA.
1999. Interaction of frizzled related protein (FRP) with Wnt
ligands and the frizzled receptor suggests alternative mechanisms
for FRP inhibition of Wnt signaling. J Biol Chem 274:
16180–16187.
Bafico A, Liu G, Goldin L, Harris V, Aaronson SA. 2004. An
autocrine mechanism for constitutive Wnt pathway activation
in human cancer cells. Cancer Cell 6: 497–506.
Baig-Lewis S, Peterson-Nedry W, Wehrli M. 2007. Wingless/Wnt
signal transduction requires distinct initiation and amplification
steps that both depend on Arrow/LRP. Dev Biol 306:
94–111.
Bilic J, Huang YL, Davidson G, Zimmermann T, Cruciat CM,
Bienz M, Niehrs C. 2007. Wnt induces LRP6 signalosomes
and promotes dishevelled-dependent LRP6 phosphorylation.
Science 316: 1619–1622.
Billiard J, Way DS, Seestaller-Wehr LM, Moran RA, Mangine A,
Bodine PV. 2005. The orphan receptor tyrosine kinase Ror2
modulates canonical Wnt signaling in osteoblastic cells. Mol
Endocrinol 19: 90–101.
Caca K, Kolligs FT, Ji X, Hayes M, Qian J, Yahanda A, Rimm DL,
Costa J, Fearon ER. 1999. b- and g-catenin mutations, but not
E-cadherin inactivation, underlie T-cell factor/lymphoid enhancer
factor transcriptional deregulation in gastric and
pancreatic cancer. Cell Growth Differ 10: 369–376.
Clevers H. 2006. Wnt/b-catenin signaling in development and
disease. Cell 127: 469–480.
Dann CE, Hsieh JC, Rattner A, Sharma D, Nathans J, Leahy DJ.
2001. Insights into Wnt binding and signalling from the
structures of two Frizzled cysteine-rich domains. Nature
412: 86–90.
Davidson G, Wu W, Shen J, Bilic J, Fenger U, Stannek P, Glinka A,
Niehrs C. 2005. Casein kinase 1 g couples Wnt receptor activation
to cytoplasmic signal transduction. Nature 438: 867–872.
DeChiara TM, Kimble RB, Poueymirou WT, Rojas J, Masiakowski
P, Valenzuela DM, Yancopoulos GD. 2000. Ror2, encoding a
receptor-like tyrosine kinase, is required for cartilage and
growth plate development. Nat Genet 24: 271–274.
Feng XH, Derynck R. 2005. Specificity and versatility in TGF-b
signaling through Smads. Annu Rev Cell Dev Biol 21: 659–693.
Fukuda T, Chen L, Endo T, Tang L, Lu D, Castro JE, Widhopf
GF II, Rassenti LZ, Cantwell MJ, Prussak CE, et al. 2008.
Antisera induced by infusions of autologous Ad-CD154leukemia
B cells identify ROR1 as an oncofetal antigen and
receptor for Wnt5a. Proc Natl Acad Sci 105: 3047–3052.
Gonzalez-Sancho JM, Brennan KR, Castelo-Soccio LA, Brown
AM. 2004. Wnt proteins induce dishevelled phosphorylation
via an LRP5/6- independent mechanism, irrespective of their
ability to stabilize b-catenin. Mol Cell Biol 24: 4757–4768.
Green JL, Kuntz SG, Sternberg PW. 2008. Ror receptor tyrosine
kinases: Orphans no more. Trends Cell Biol 18: 536–544.
Grigoryan T, Wend P, Klaus A, Birchmeier W. 2008. Deciphering
the function of canonical Wnt signals in development and
disease: Conditional loss- and gain-of-function mutations of
b-catenin in mice. Genes Dev 22: 2308–2341.
He X, Saint-Jeannet JP, Wang Y, Nathans J, Dawid I, Varmus H.
1997. A member of the Frizzled protein family mediating
axis induction by Wnt-5A. Science 275: 1652–1654.
Grumolato et al.
2528 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
Hikasa H, Shibata M, Hiratani I, Taira M. 2002. The Xenopus
receptor tyrosine kinase Xror2 modulates morphogenetic
movements of the axial mesoderm and neuroectoderm via
Wnt signaling. Development 129: 5227–5239.
Ishitani T, Kishida S, Hyodo-Miura J, Ueno N, Yasuda J,
Waterman M, Shibuya H, Moon RT, Ninomiya-Tsuji J,
Matsumoto K. 2003. The TAK1–NLK mitogen-activated
protein kinase cascade functions in the Wnt-5a/Ca2+
pathway
to antagonize Wnt/b-catenin signaling. Mol Cell Biol 23:
131–139.
Jing L, Lefebvre JL, Gordon LR, Granato M. 2009. Wnt signals
organize synaptic prepattern and axon guidance through the
zebrafish unplugged/MuSK receptor. Neuron 61: 721–733.
Khan Z, Vijayakumar S, de la Torre TV, Rotolo S, Bafico A. 2007.
Analysis of endogenous LRP6 function reveals a novel feedback
mechanism by which Wnt negatively regulates its
receptor. Mol Cell Biol 27: 7291–7301.
Kikuchi A, Yamamoto H, Sato A. 2009. Selective activation
mechanisms of Wnt signaling pathways. Trends Cell Biol 19:
119–129.
Kockeritz L, Doble B, Patel S, Woodgett JR. 2006. Glycogen
synthase kinase-3—An overview of an over-achieving protein
kinase. Curr Drug Targets 7: 1377–1388.
Liu G, Bafico A, Harris VK, Aaronson SA. 2003. A novel
mechanism for Wnt activation of canonical signaling
through the LRP6 receptor. Mol Cell Biol 23: 5825–5835.
Liu G, Bafico A, Aaronson SA. 2005. The mechanism of
endogenous receptor activation functionally distinguishes
prototype canonical and noncanonical Wnts. Mol Cell Biol
25: 3475–3482.
Liu Y, Bhat RA, Seestaller-Wehr LM, Fukayama S, Mangine A,
Moran RA, Komm BS, Bodine PV, Billiard J. 2007. The
orphan receptor tyrosine kinase Ror2 promotes osteoblast
differentiation and enhances ex vivo bone formation. Mol
Endocrinol 21: 376–387.
Liu Y, Rubin B, Bodine PV, Billiard J. 2008. Wnt5a induces
homodimerization and activation of Ror2 receptor tyrosine
kinase. J Cell Biochem 105: 497–502.
Liu G, Vijayakumar S, Grumolato L, Arroyave R, Qiao H, Akiri
G, Aaronson SA. 2009. Canonical Wnts function as potent
regulators of osteogenesis by human mesenchymal stem
cells. J Cell Biol 185: 67–75.
Logan CY, Nusse R. 2004. The Wnt signaling pathway in
development and disease. Annu Rev Cell Dev Biol 20: 781–
810.
MacDonald BT, Tamai K, He X. 2009. Wnt/b-catenin signaling:
Components, mechanisms, and diseases. Dev Cell 17: 9–26.
Metcalfe C, Mendoza-Topaz C, Mieszczanek J, Bienz M. 2010.
Stability elements in the LRP6 cytoplasmic tail confer
efficient signalling upon DIX-dependent polymerization.
J Cell Sci 123: 1588–1599.
Mikels AJ, Nusse R. 2006. Purified Wnt5a protein activates or
inhibits b-catenin–TCF signaling depending on receptor context.
PLoS Biol 4: e115. doi: 10.1371/journal.pbio.0040115.
Mikels A, Minami Y, Nusse R. 2009. Ror2 receptor requires
tyrosine kinase activity to mediate Wnt5A signaling. J Biol
Chem 284: 30167–30176.
Nishita M, Yoo SK, Nomachi A, Kani S, Sougawa N, Ohta Y,
Takada S, Kikuchi A, Minami Y. 2006. Filopodia formation
mediated by receptor tyrosine kinase Ror2 is required for
Wnt5a-induced cell migration. J Cell Biol 175: 555–562.
Nishita M, Enomoto M, Yamagata K, Minami Y. 2010a. Cell/
tissue-tropic functions of Wnt5a signaling in normal and
cancer cells. Trends Cell Biol 20: 346–354.
Nishita M, Itsukushima S, Nomachi A, Endo M, Wang Z, Inaba
D, Qiao S, Takada S, Kikuchi A, Minami Y. 2010b. Ror2/
Frizzled complex mediates Wnt5a-induced AP-1 activation
by regulating Dishevelled polymerization. Mol Cell Biol 30:
3610–3619.
Nomachi A, Nishita M, Inaba D, Enomoto M, Hamasaki M,
Minami Y. 2008. Receptor tyrosine kinase Ror2 mediates
Wnt5a-induced polarized cell migration by activating c-Jun
N-terminal kinase via actin-binding protein filamin A. J Biol
Chem 283: 27973–27981.
Oishi I, Suzuki H, Onishi N, Takada R, Kani S, Ohkawara B,
Koshida I, Suzuki K, Yamada G, Schwabe GC, et al. 2003.
The receptor tyrosine kinase Ror2 is involved in noncanonical
Wnt5a/JNK signalling pathway. Genes Cells 8:
645–654.
Pan W, Choi SC, Wang H, Qin Y, Volpicelli-Daley L, Swan L,
Lucast L, Khoo C, Zhang X, Li L, et al. 2008. Wnt3a-mediated
formation of phosphatidylinositol 4,5-bisphosphate regulates
LRP6 phosphorylation. Science 321: 1350–1353.
Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R,
Mosca JD, Moorman MA, Simonetti DW, Craig S, Marshak
DR. 1999. Multilineage potential of adult human mesenchymal
stem cells. Science 284: 143–147.
Qian D, Jones C, Rzadzinska A, Mark S, Zhang X, Steel KP, Dai
X, Chen P. 2007. Wnt5a functions in planar cell polarity
regulation in mice. Dev Biol 306: 121–133.
Sato A, Yamamoto H, Sakane H, Koyama H, Kikuchi A. 2010.
Wnt5a regulates distinct signalling pathways by binding to
Frizzled2. EMBO J 29: 41–54.
Schambony A, Wedlich D. 2007. Wnt-5A/Ror2 regulate expression
of XPAPC through an alternative noncanonical signaling
pathway. Dev Cell 12: 779–792.
Schlessinger J. 2000. Cell signaling by receptor tyrosine kinases.
Cell 103: 211–225.
Schlessinger K, McManus EJ, Hall A. 2007. Cdc42 and noncanonical
Wnt signal transduction pathways cooperate to
promote cell polarity. J Cell Biol 178: 355–361.
Semenov MV, Habas R, Macdonald BT, He X. 2007. SnapShot:
Noncanonical Wnt signaling pathways. Cell 131: 1378.e1–
1378.e2. doi: 10.1016/j.cell.2007.12.011.
Simons M, Mlodzik M. 2008. Planar cell polarity signaling:
From fly development to human disease. Annu Rev Genet
42: 517–540.
Tamai K, Semenov M, Kato Y, Spokony R, Liu C, Katsuyama Y,
Hess F, Saint-Jeannet JP, He X. 2000. LDL-receptor-related
proteins in Wnt signal transduction. Nature 407: 530–535.
Topol L, Jiang X, Choi H, Garrett-Beal L, Carolan PJ, Yang Y.
2003. Wnt-5a inhibits the canonical Wnt pathway by promoting
GSK-3-independent b-catenin degradation. J Cell
Biol 162: 899–908.
van Amerongen R, Mikels A, Nusse R. 2008. Alternative Wnt
signaling is initiated by distinct receptors. Sci Signal 1: re9.
doi: 10.1126/scisignal.135re9.
Wallingford JB, Habas R. 2005. The developmental biology of
Dishevelled: An enigmatic protein governing cell fate and
cell polarity. Development 132: 4421–4436.
Wang Y, Guo N, Nathans J. 2006. The role of Frizzled3 and
Frizzled6 in neural tube closure and in the planar polarity of
inner-ear sensory hair cells. J Neurosci 26: 2147–2156.
Westfall TA, Brimeyer R, Twedt J, Gladon J, Olberding A,
Furutani-Seiki M, Slusarski DC. 2003. Wnt-5/pipetail functions
in vertebrate axis formation as a negative regulator of Wnt/
b-catenin activity. J Cell Biol 162: 889–898.
Yamaguchi TP, Bradley A, McMahon AP, Jones S. 1999. A Wnt5a
pathway underlies outgrowth of multiple structures in the
vertebrate embryo. Development 126: 1211–1223.
Yamamoto A, Nagano T, Takehara S, Hibi M, Aizawa S. 2005.
Shisa promotes head formation through the inhibition of
Mechanism of Ror1/2 receptor activation
GENES & DEVELOPMENT 2529
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from
receptor protein maturation for the caudalizing factors, Wnt
and FGF. Cell 120: 223–235.
Yamamoto H, Yoo SK, Nishita M, Kikuchi A, Minami Y. 2007.
Wnt5a modulates glycogen synthase kinase 3 to induce
phosphorylation of receptor tyrosine kinase Ror2. Genes
Cells 12: 1215–1223.
Yamamoto S, Nishimura O, Misaki K, Nishita M, Minami Y,
Yonemura S, Tarui H, Sasaki H. 2008. Cthrc1 selectively
activates the planar cell polarity pathway of Wnt signaling
by stabilizing the Wnt–receptor complex. Dev Cell 15: 23–
36.
Zeng X, Tamai K, Doble B, Li S, Huang H, Habas R, Okamura H,
Woodgett J, He X. 2005. A dual-kinase mechanism for Wnt
co-receptor phosphorylation and activation. Nature 438:
873–877.
Zeng X, Huang H, Tamai K, Zhang X, Harada Y, Yokota C,
Almeida K, Wang J, Doble B, Woodgett J, et al. 2008.
Initiation of Wnt signaling: Control of Wnt coreceptor Lrp6
phosphorylation/activation via frizzled, dishevelled and axin
functions. Development 135: 367–375.
Grumolato et al.
2530 GENES & DEVELOPMENT
Cold Spring Harbor Laboratory Presson November 11, 2013 - Published bygenesdev.cshlp.orgDownloaded from