Inactivation of YAP oncoprotein
by the Hippo pathway is involved
in cell contact inhibition and tissue
Bin Zhao,1,2,10Xiaomu Wei,3Weiquan Li,1Ryan S. Udan,4,5Qian Yang,1,2Joungmok Kim,1,10
Joe Xie,1Tsuneo Ikenoue,1Jindan Yu,6Li Li,2,10Pan Zheng,6,7Keqiang Ye,8Arul Chinnaiyan,6
Georg Halder,4,5Zhi-Chun Lai,3and Kun-Liang Guan1,2,9,10,11
1Life Sciences Institute, University of Michigan, Ann Arbor, Michigan 48109, USA;2Department of Biological Chemistry,
University of Michigan, Ann Arbor, Michigan 48109, USA;3Department of Biology and Intercollege Graduate Program in
Genetics, The Pennsylvania State University, University Park, Pennsylvania 16802, USA;4Program in Developmental
Biology, Baylor College of Medicine, Houston, Texas, 77030, USA;5Department of Biochemistry and Molecular Biology,
M.D. Anderson Cancer Center, Houston, Texas 77030, USA;6Department of Pathology, University of Michigan, Ann Arbor,
Michigan 48109, USA;7Department of Surgery, University of Michigan, Ann Arbor, Michigan 48109, USA;8Department of
Pathology, Emory University School of Medicine, Atlanta, Georgia 30322, USA;9Institute of Gerontology, University of
Michigan, Ann Arbor, Michigan 48109, USA
The Hippo pathway plays a key role in organ size control by regulating cell proliferation and apoptosis in
Drosophila. Although recent genetic studies have shown that the Hippo pathway is regulated by the NF2 and
Fat tumor suppressors, the physiological regulations of this pathway are unknown. Here we show that in
mammalian cells, the transcription coactivator YAP (Yes-associated protein), is inhibited by cell density via
the Hippo pathway. Phosphorylation by the Lats tumor suppressor kinase leads to cytoplasmic translocation
and inactivation of the YAP oncoprotein. Furthermore, attenuation of this phosphorylation of YAP or Yorkie
(Yki), the Drosophila homolog of YAP, potentiates their growth-promoting function in vivo. Moreover, YAP
overexpression regulates gene expression in a manner opposite to cell density, and is able to overcome cell
contact inhibition. Inhibition of YAP function restores contact inhibition in a human cancer cell line bearing
deletion of Salvador (Sav), a Hippo pathway component. Interestingly, we observed that YAP protein is
elevated and nuclear localized in some human liver and prostate cancers. Our observations demonstrate that
YAP plays a key role in the Hippo pathway to control cell proliferation in response to cell contact.
[Keywords: YAP; Lats; Mst; contact inhibition; NF2; Hippo]
Supplemental material is available at http://www.genesdev.org.
Received August 9, 2007; revised version accepted September 14, 2007.
Body and organ size of metazoans is determined by cell
number and cell size. The opposing action of cell prolif-
eration and apoptosis controls cell number in particular
tissue and organs (Conlon and Raff 1999). Recent genetic
studies in Drosophila have shown that the Hippo signal-
ing pathway plays a key role in restricting organ size by
controlling both cell proliferation and apoptosis (Edgar
2006; Harvey and Tapon 2007; Pan 2007). Hippo (Hpo) is
a Ste20 family protein kinase that complexes with a
regulatory scaffold protein Salvador (Sav) (Kango-Singh
et al. 2002; Tapon et al. 2002; Harvey et al. 2003; Panta-
lacci et al. 2003; Udan et al. 2003; Wu et al. 2003). The
Hpo/Sav complex phosphorylates and activates Warts
(Wts), a NDR (nuclear Dbf2-related) family protein ki-
nase. Wts has an activating subunit Mats (Mob as tumor
suppressor) (Lai et al. 2005; Wei et al. 2007). The Wts/
Mats complex inhibits Yorkie (Yki), a transcription co-
activator (Huang et al. 2005), possibly via direct phos-
phorylation, although the precise mechanism has yet to
be determined. In Drosophila, key downstream targets of
Yki include cyclin E, diap-1, and the bantam microRNA
(Huang et al. 2005; Nolo et al. 2006; Thompson and Co-
Although elusive for several years, the signals up-
stream of Hpo are now emerging. The NF2 tumor sup-
pressor, also known as Merlin (Mer), and Expanded (Ex),
two ezrin/radixin/moesin (ERM) family actin-binding
proteins (McClatchey and Giovannini 2005; Okada et al.
2007), have been shown to positively regulate the Hippo
10Present address: Department of Pharmacology and Moores Cancer Cen-
ter, University of California at San Diego, La Jolla, CA 92093, USA.
E-MAIL firstname.lastname@example.org; FAX (734) 647-9702.
Article is online at http://www.genesdev.org/cgi/doi/10.1101/gad.1602907.
GENES & DEVELOPMENT 21:2747–2761 © 2007 by Cold Spring Harbor Laboratory Press ISSN 0890-9369/07; www.genesdev.org 2747
pathway in Drosophila (Hamaratoglu et al. 2006). Inter-
estingly, genetic data indicate that Fat, a protocadherin
tumor suppressor, also functions upstream of Hpo (Ben-
nett and Harvey 2006; Cho et al. 2006; Hariharan 2006;
Silva et al. 2006; Willecke et al. 2006; Tyler and Baker
2007; Yin and Pan 2007). The fact that Fat may interact
with another protocadherin, Dachsous, at the cell sur-
face (Matakatsu and Blair 2004; Halbleib and Nelson
2006) suggests an exciting possibility that the Hippo
pathway may be involved in cell growth regulation in
response to cell–cell contact.
Components of the Hippo pathway are highly con-
served in mammals, including YAP (Yes-associated pro-
tein), Lats1/2, Mob, Mst1/2, Sav, Merlin, Ex1/2, and Fat4
(Yki, Wts, Mats, Hpo, Sav, Mer, Expanded, and Fat ho-
mologs, respectively). Human YAP, Lats1, Mst2, and
Mob1 can functionally rescue the respective Drosophila
mutants, suggesting the functional conservation of these
proteins in mammals (Edgar 2006). Interestingly, YAP
has been shown recently to be a candidate oncogene in
the human chromosome 11q22 amplicon (Overholtzer et
al. 2006; Zender et al. 2006). In addition, mutations of
Lats1/2, Sav, and Mob have been implicated in tumori-
genesis (St John et al. 1999; Tapon et al. 2002; Lai et al.
2005; Takahashi et al. 2005; Harvey and Tapon 2007). In
spite of their conservation and intimate relationship
with cancer, the Hippo pathway has not been systemati-
cally studied in mammalian cells.
A fundamental property of a normal cell is to cease
proliferation upon reaching confluence, a phenomenon
referred to as cell contact inhibition (Eagle and Levine
1967). In contrast, cancer cells are able to escape cell
contact inhibition, which enhances their ability to in-
vade host tissues and metastasize (Hanahan and Wein-
berg 2000). This is also one of the most commonly used
criteria for cellular transformation in vitro (Abercrombie
1979). Although activation of oncogenes and inactiva-
tion of tumor suppressor genes can prevent contact in-
hibition, the precise molecular mechanism is not clear.
In this report we show that YAP is regulated by the
Hippo pathway and may play an important role in me-
diating cell contact inhibition. YAP is phosphorylated
and inhibited by the Lats tumor suppressor, and this
phosphorylation results in its association with 14–3–3
and cytoplasmic localization. This regulatory mecha-
nism is utilized in YAP regulation by cell density and is
likely conserved in Drosophila. Furthermore, overex-
pression of YAP antagonizes density-dependent gene
regulation and contact inhibition, whereas expression of
dominant-negative YAP restores contact inhibition in a
human cancer cell line bearing a deletion of Sav. More-
over, we showed that YAP expression levels and nuclear
localization are strongly elevated in some human cancers.
YAP localization and phosphorylation are regulated
by cell density
YAP is a transcription coactivator and a candidate onco-
gene, but neither its function in cancers nor its physi-
ological regulation has been established. Interestingly,
we found that YAP localization was regulated by cell
density (Fig. 1A). At low density, YAP was predomi-
nantly localized in the nuclei of NIH-3T3 cells. In con-
trast, YAP translocated to the cytoplasm at high density.
Similar observations were made in the MCF10A human
breast epithelial cell line (Fig. 1A). This translocation
was unlikely due to differential medium conditions, be-
cause in cell colonies YAP was preferentially localized to
nuclei in cells at the edge but displayed cytoplasmic lo-
calization in cells toward the center (Fig. 1B). Given the
fact that YAP is a transcription coactivator acting in the
cell nucleus (Yagi et al. 1999), our results indicate that
YAP may be inhibited by high cell density.
Besides translocation, YAP from high-density cultures
displayed a slower electrophoretic migration (Fig. 1C).
This density-dependent mobility shift was due to phos-
phorylation because phosphatase treatment converted
YAP to the fast migrating form, suggesting that YAP
phosphorylation is regulated by cell density. Together,
the above observations indicate a possible relationship
between YAP phosphorylation and cytoplasmic localiza-
tion upon high cell density.
The Hippo pathway regulates YAP phosphorylation,
activity, and localization
In Drosophila, it has been reported that Yki, the YAP
homolog, is inhibited by the Hippo pathway, possibly via
phosphorylation (Huang et al. 2005). Therefore, we
tested effects of the Hippo pathway on YAP phosphory-
lation. All cDNAs used in the cell culture studies are of
human or mouse origin. YAP2, one of the two alterna-
tively spliced forms of human yap, was coexpressed with
the Hippo pathway kinases Mst2 or Lats2. We found that
expression of Mst2 or Lats2 caused a modest mobility
shift of YAP2 that was further enhanced by Sav and Mob,
the respective regulatory subunits of Mst2 and Lats2
(Fig. 2A). Moreover, coexpression of both Mst2 and Lats2
resulted in a dramatic mobility shift of YAP2. These re-
sults indicate that ectopic expression of Mst2 and Lats2
induces YAP2 phosphorylation.
In order to test the possibility of direct phosphoryla-
tion of YAP2 by Lats2, we performed an in vitro kinase
assay using purified GST-YAP2 and immunoprecipitated
Lats2. As shown in Figure 2B, Lats2, but not the kinase-
inactive Lats2-KR, phosphorylated YAP2. In contrast,
Mst2 poorly phosphorylated GST-YAP2, even though it
had much stronger autophosphorylation than that of
Lats2 (Supplementary Fig. S1A). These data demonstrate
that Lats2 directly phosphorylates YAP2, while Mst2
stimulates YAP2 phosphorylation indirectly in vivo, per-
haps by activating Lats2.
YAP has been shown to interact with and activate the
TEAD family transcription factors, which have four
highly conserved members (Vassilev et al. 2001). To as-
sess the effect of phosphorylation on YAP activity, we
utilized a reporter system consisting of a 5× UAS-lucif-
erase reporter and a Gal4 DNA-binding domain fused to
TEAD4 (Gal4-TEAD4). In the absence of YAP, Gal4-
Zhao et al.
2748GENES & DEVELOPMENT
TEAD4 had low basal activity. However, when YAP2
was cotransfected, the reporter was strongly activated
(Fig. 2C). Coexpression of Lats2 or Mst2, but not the
kinase-inactive mutants, resulted in a dose-dependent
inhibition of the reporter (Supplementary Fig. S1B,C).
Reminiscent of the effect seen on phosphorylation,
YAP2 activity was further inhibited by coexpressing
Mst2/Sav or Lats2/Mob, and even more dramatically in-
hibited by a combination of all four proteins (Fig. 2C).
This inhibition of YAP2 activity was also observed in
COS7 and HeLa cells (data not shown). We also tested
the effect of Merlin and Expanded on YAP2 activity. Co-
expression of either caused a modest but reproducible
inhibition of YAP2 activity (Fig. 2D). Furthermore, Mer-
lin and Expanded enhanced the inhibition of YAP2 by
Mst2 and Lats2. Consistently, Merlin also caused a mo-
bility shift of YAP2 (Supplementary Fig. S1D).
Next, we addressed whether Mst2 and Lats2 affected
YAP localization. In HeLa cells, endogenous YAP was
localized in the nucleus at low cell density (Fig. 2E).
However, expression of Lats2, but not the kinase-inac-
tive mutant, caused a dramatic redistribution of YAP to
the cytoplasm (Fig. 2E). Similarly, expression of Mst2 but
not the kinase-inactive mutant increased cytoplasmic
YAP, although less dramatically. Expression of Merlin
also resulted in YAP cytoplasmic translocation, support-
ing the role of Merlin in the Hippo pathway in mamma-
lian cells. These results suggest that activation of the
Hippo pathway may cause cytoplasmic translocation of
YAP through phosphorylation by Lats.
Schwannoma is the major tumor type associated with
Merlin mutation. We examined YAP localization in the
RT4-D6-P2T rat Schwannoma cell line, which is inca-
pable of inducing Merlin expression at high cell density
as normal Schwann cells do (Morrison et al. 2001). We
observed that the majority of Schwannoma cells showed
nuclear YAP localization even under high density (Fig.
2F). Interestingly, expression of Merlin wild-type, but
not a cancer-derived L64P mutant, restored YAP cyto-
plasmic translocation. Together, these results further
support the involvement of Merlin and the Hippo path-
way in the regulation of YAP translocation in response
to cell density.
Lats inhibits YAP by phosphorylating HXRXXS motifs
Lats belongs to the NDR family of protein kinases (Her-
govich et al. 2006). Previous biochemical studies have
shown that the yeast Dbf2 kinase recognizes an RXXS
motif in its substrates (Mah et al. 2005). Interestingly, in
search of such a consensus, we noticed that YAP2 con-
tains five HXRXXS motifs (Fig. 3A), of which three are
conserved in Drosophila. It is worth noting that the pep-
tides utilized in elucidating the Dbf2 recognition motif
also had a histidine at position −5 (Mah et al. 2005).
We mutated YAP2 by replacing individual serine resi-
dues in the HXRXXS motifs with alanine. Among the
single mutants tested, S127A, which is conserved in
Drosophila Yki, was most resistant to Mst2/Lats2-in-
duced mobility shift (Fig. 3B). Mutation of all five serine
and MCF10A cells were cultured sparsely or to confluence. YAP was stained with anti-YAP antibody. (B) MCF10A cells at the edge
of a large colony have high nuclear YAP. YAP was stained with anti-YAP antibody. (C) High cell density induces YAP phosphorylation.
NIH-3T3 cell lysates from cells at different densities were probed with anti-YAP antibody. ? phosphatase treatment is indicated.
YAP localization and phosphorylation are regulated by cell density. (A) YAP localization is affected by cell density. NIH-3T3
YAP in cell contact inhibition
GENES & DEVELOPMENT 2749
residues (YAP2-5SA) produced a YAP2 downshift more
dramatic than that of any single mutant. Furthermore, ?
phosphatase treatment abolished the Mst2/Lats2-in-
duced mobility shift of YAP2, therefore verifying the role
of phosphorylation in this mobility shift (Fig. 3B). These
results indicate that Ser127 is the primary phosphoryla-
tion site in YAP2, while serines in other HXRXXS motifs
may also be phosphorylated.
To further confirm the phosphorylation of YAP2
HXRXXS motifs by Lats2, an in vitro kinase assay was
YAP2 mobility. Flag-YAP2 was cotransfected with indicated plasmids into HEK293 cells. Western blot was performed as indicated. (B)
In vitro phosphorylation of YAP2 by Lats2. HA-Lats2 was immunoprecipitated from transfected HEK293 cells. In vitro kinase assay
was performed using purified GST-YAP2 as a substrate in the presence of [32P]ATP. GST-Sin1 was used as a negative control. (KR)
Kinase-inactive mutant. (C) YAP2 activity is inhibited by Mst2 and Lats2. Indicated plasmids were cotransfected with a 5× UAS-
luciferase reporter and a CMV-?-gal construct into 293T cells. Luciferase activity was measured and normalized to ?-galactosidase
activity. (D) YAP2 activity is inhibited by Merlin and Expanded. Experiments are similar to those in C. The Ex used is human FRMD6.
(E) Activation of the Hippo pathway causes YAP cytoplasmic localization. HeLa cells were transfected with indicated plasmids.
Endogenous YAP2 was stained to visualize the localization. (F) Cell density-induced YAP translocation is Merlin dependent. RT4-
D6-P2T Schwannoma cell lines with empty vector, inducible wild-type Merlin, or a Merlin-L64P mutant were cultured to confluence.
Merlin expression was induced by doxycycline for 2 d. (Left panel) Expression of Merlin was determined by Western blot. Endogenous
YAP was stained and YAP localization was quantified.
The Hippo pathway regulates YAP phosphorylation, activity, and localization. (A) Coexpression of Mst2 and Lats2 decreases
Zhao et al.
2750GENES & DEVELOPMENT
performed. Mutation of S127 reduced and mutation of all
five serine residues abolished YAP2 phosphorylation by
Lats2 as determined by32P incorporation (Fig. 3C). Phos-
phorylation of S127 was also verified by immunoblotting
with a phospho-YAP (S127)-specific antibody (Fig. 3C).
The specificity of this antibody was confirmed by
phosphatase treatment (Supplementary Fig. S2A). These
data demonstrate that Lats2 directly phosphorylates YAP2
on S127 and other serine residues in the HXRXXS motifs.
The functional significance of YAP2 phosphorylation
was evaluated by the TEAD4 reporter assay. As shown
before, wild-type YAP2 was potently inhibited by coex-
pression of Mst2/Lats2; however, the S127A mutant
showed resistance to this inhibition (Fig. 3D). Further-
more, the YAP2-5SA mutant was not only resistant to
inhibition by Mst2 and Lats2, but also displayed an el-
evated basal activity (Fig. 3D). Together, our data dem-
onstrate that YAP2 activity is inhibited by phosphoryla-
target sequence was aligned with the five HXRXXS motifs of human YAP2. (B) Ser127 is the major phosphorylation site in YAP2.
Wild-type or mutant Flag-YAP2 was cotransfected with HA-Mst2 and HA-Lats2 as indicated. YAP2 mobility shift was determined by
anti-Flag Western blot. (C) Lats2 directly phosphorylates YAP2 on HXRXXS motifs. In vitro phosphorylation of YAP2 mutants with
immunoprecipitated HA-Lats2 was performed. Phosphorylation of GST-YAP2 was detected by either
phospho-YAP (S127) Western blot. (Bottom panel) GST-YAP2 input was shown by Coomassie Blue staining. (D) YAP2 phosphoryla-
tion-defective mutants S127A and 5SA are resistant to inhibition by Mst2 and Lats2. The reporter assay is similar to those in Figure
2C. The fold activity inhibition of each mutant by Mst2/Lats2 is indicated at the top of this panel. (E) Coexpression of Mst2 and Lats2
increases YAP2 S127 phosphorylation. Flag-YAP2 was cotransfected with HA-Lats2 and Flag-Mst2 into HEK293 cells as indicated.
Flag-YAP2 was immunoprecipitated and phosphorylation of S127 was detected by pYAP (S127) antibody. (F) Knockdown of Lats
decreases endogenous YAP S127 phosphorylation. HeLa cells were transfected twice with small interfering RNA for Lats1 and Lats2
as indicated. Phosphorylation and protein levels of endogenous YAP were determined by Western blot. Knockdown of Lats was verified
by the anti-Lats antibody, which recognizes both Lats1 and Lats2. (G) YAP S127 phosphorylation increases with cell density. NIH-3T3
and MEF cells were harvested at different densities, and YAP phosphorylation was assayed. (H) Lats2 kinase activity increases with
cell density. NIH-3T3 cells were harvested at different densities. Endogenous Lats2 was immunoprecipitated and used in an in vitro
kinase assay. Phosphorylation of GST-YAP2 was detected by anti-phospho-YAP (S127) Western blot. Rheb IP was included as a
Lats inhibits YAP by phosphorylating HXRXXS motifs. (A) YAP2 contains five HXRXXS motifs. The yeast Dbf2 optimal
32P incorporation or anti-
YAP in cell contact inhibition
GENES & DEVELOPMENT 2751
tion of the HXRXXS motifs, especially S127. Since YAP2-
5SA was also partially inhibited by Mst/Lats, additional
levels of regulation may exist.
To confirm the phosphorylation of YAP2 S127 in vivo
by Lats2, we did coexpression in cultured cells. Co-
transfection of Lats2 alone or together with Mst2 in-
creased YAP2 S127 phosphorylation (Fig. 3E). Interest-
ingly, expression of the inactive Lats2-KR mutant de-
creased YAP2 basal phosphorylation, perhaps through a
dominant-negative effect. This supports a role of endog-
enous Lats in YAP2 phosphorylation. To further confirm
this, Lats1 and Lats2 were down-regulated by RNA in-
terference. Knockdown of Lats2 caused a significant re-
duction in S127 phosphorylation of transfected Flag-
YAP2, while knockdown of both Lats1 and Lats2 abol-
ished its phosphorylation (Supplementary Fig. S2B).
Similarly, knockdown of both Lats1 and Lats2 decreased
endogenous YAP phosphorylation (Fig. 3F), thus estab-
lishing an important role of Lats in YAP phosphorylation
After the determination of a Lats target phosphoryla-
tion site on YAP, we re-examined the cell density-in-
duced phosphorylation of YAP. Along with the reduced
electrophoretic migration of YAP as shown in Figure 1C,
we also observed that YAP S127 phosphorylation was
increased by cell density in both NIH-3T3 and mouse
embryonic fibroblast (MEF) cells (Fig. 3G). To determine
whether cell density regulates Lats kinase activity, we
immunoprecipitated endogenous Lats2 from NIH-3T3
cells and measured its kinase activity toward YAP in
vitro. Interestingly, Lats2 from high-density culture dis-
played an elevated activity (Fig. 3H). This result directly
suggests the activation of Lats and possibly the Hippo
pathway under high cell density, which nicely explains
the increased phosphorylation of YAP.
It has been previously reported that YAP2 S127 is
phosphorylated by Akt/PKB (Basu et al. 2003). However,
the reported YAP inhibition by Akt-dependent phos-
phorylation is inconsistent with recent genetic data that
demonstrate YAP as an oncogene. We tested the func-
tion of Akt in YAP phosphorylation. Surprisingly, nei-
ther LY294002 nor wortmannin (two PI3K inhibitors) de-
creased YAP2 S127 phosphorylation, although they po-
tently blocked the phosphorylation of Akt and GSK3, a
physiological Akt substrate (Supplementary Fig. S3A). In
addition, neither EGF nor insulin stimulated YAP2 phos-
phorylation, while both strongly stimulated the phos-
phorylation of Akt and GSK3 (Supplementary Fig.
S3A,B). Phosphorylation of Akt T308 by PDK1 is essen-
tial for Akt activity (Williams et al. 2000). However, YAP
phosphorylation was not affected by PDK1 knockout
(Supplementary Fig. S3C). In addition, coexpression of
wild-type or constitutively active myristoylated Akt did
not increase YAP2 phosphorylation (Supplementary Fig.
S3D). We also observed that Akt did inhibit YAP2 activ-
ity, but in a kinase activity-independent manner, sug-
gesting that Akt overexpression could inhibit YAP2 in-
directly (Supplementary Fig. S3E). Together, our results
demonstrate that Akt is unlikely to be responsible for
YAP2 S127 phosphorylation.
Phosphorylation promotes YAP cytoplasmic
localization and inhibits its transcription factor
To directly determine the effect of phosphorylation on
YAP localization, we performed immunofluorescence
staining of transfected wild-type or phosphorylation-de-
ficient YAP2. Flag-YAP2 showed prominent nuclear lo-
calization in transfected cells, while coexpression of
Lats2 induced nearly complete cytoplasmic transloca-
tion (Fig. 4A). Interestingly, Lats2 had only minor effects
on YAP2-S127A and 5SA localization.
To determine whether Lats-dependent phosphoryla-
tion is indeed responsible for YAP translocation under
high cell density, we examined MCF10A cells stably ex-
pressing Myc-YAP2 or Myc-YAP2-5SA. Similar to en-
dogenous YAP, Myc-YAP2 showed density-dependent
subcellular localization (Fig. 4B). In contrast, Myc-YAP2-
5SA displayed both nuclear and cytoplasmic staining un-
der high density. Together, our studies suggest that phos-
phorylation of HXRXXS motifs by Lats is at least in part
responsible for the nuclear-to-cytoplasm translocation of
YAP in response to cell contact signals.
YAP is a transcription coactivator; therefore, we hy-
pothesized that the Lats-induced cytoplasmic transloca-
tion of YAP inhibits its function by attenuating its in-
teraction with nuclear-localized transcription factors. In-
deed, we observed that coexpression of Mst2 and Lats2
decreased the association between TEAD4 and YAP2
(Supplementary Fig. S4), but had no effect on the inter-
action between TEAD4 and YAP2-S127A or 5SA (Fig.
4C). To exclude the possibility that YAP2 phosphoryla-
tion directly affects YAP2/TEAD4 interaction affinity,
we tested whether dephosphorylation affects YAP2/
TEAD4 association in vitro. Immunoprecipitated YAP2
was treated with ? phosphatase and incubated with Myc-
TEAD4 containing cell lysate. As shown in Figure 4D,
dephosphorylation of YAP2 had little effect on its inter-
action with TEAD4 in vitro. Therefore, we conclude that
YAP2 phosphorylation by Lats2 leads to decreased inter-
action with TEAD secondary to cytoplasmic retention.
S127 phosphorylation regulates YAP and 14–3–3
One commonly seen mechanism of cytoplasmic reten-
tion of nuclear proteins is 14–3–3 binding (Muslin and
Xing 2000). Interestingly, YAP S127 phosphorylation has
been reported to create a 14–3–3-binding site (Basu et al.
2003). We observed that YAP2 interacts with 14–3–3,
and this interaction is completely abolished by phospha-
tase treatment (Fig. 5A). Furthermore, expression of
Mst2 and Lats2, but not Akt, increased the interaction
between YAP2 and 14–3–3 in a S127-dependent manner,
as neither YAP2-S127A nor 5SA showed any binding to
14–3–3 (Fig. 5B). Our results suggest a model in which
Lats2 promotes YAP2 cytoplasmic localization by in-
creasing S127 phosphorylation and 14–3–3 binding.
Study of the yeast Dbf2 kinase has shown that R at the
−3 position of target S/T is critical for kinase recognition
Zhao et al.
2752GENES & DEVELOPMENT
(Fig. 3A). However, the function of the H at the −5 posi-
tion is unknown. We tested the importance of this his-
tidine by examining the phosphorylation of YAP2-
H122Y and H122L mutants. Mutation of H122 to either
Y or L significantly decreased S127 phosphorylation in
vitro and in vivo (Fig. 5C,D), indicating the importance
of the histidine at the −5 position. These two mutations
also attenuated interaction with 14–3–3 (Fig. 5D), which
is likely due to decreased S127 phosphorylation. A pro-
line at the +2 position of a phosphorylated serine is criti-
YAP2 cytoplasmic localization induced by Lats2. Flag-YAP2 wild type or mutants were transfected alone or together with HA-Lats2
into HeLa cells. Cells were stained with Flag and HA antibodies. (B) Phosphorylation is required for cell density-induced YAP2
cytoplasmic translocation. MCF10A cells stably expressing Myc-YAP2 or Myc-YAP2-5SA were cultured at low or high density.
Myc-YAP2 was stained with anti-Myc antibody. (C) Lats and Mst decrease YAP2/TEAD4 interaction in vivo in a S127-dependent
manner. Indicated plasmids were transfected into HEK293 cells. Flag-YAP2 was immunoprecipitated, and coprecipitated Myc-TEAD4
was detected by Western blot. (D) YAP2 dephosphorylation does not affect its interaction with TEAD4 in vitro. Flag-YAP2 (cotrans-
fected with Mst2 and lats2) immunoprecipitated from HEK293 cells were treated with ? phosphatase as indicated and then used in an
in vitro TEAD pull-down assay. Myc-TEAD4 was prepared from transfected HEK293 cells. The final products were analyzed by
Phosphorylation promotes YAP cytoplasmic localization and inhibits transcription factor binding. (A) Ser127 is required for
YAP in cell contact inhibition
GENES & DEVELOPMENT2753
cal for 14–3–3 binding. To investigate the importance of
this proline, we assayed 14–3–3 interaction with a YAP2-
P129D mutant. As expected, mutation of P129 com-
pletely eliminated 14–3–3 binding and also decreased
recognition by the pYAP antibody (Fig. 5E). However, in
vitro phosphorylation assays showed that the P129D
mutation did not affect phosphorylation by Lats2 (Fig.
5C). Based on the above data, we conclude that P129 is
important for 14–3–3 binding, but is not directly in-
volved in YAP phosphorylation by Lats. In contrast,
H122 plays a critical role in YAP phosphorylation by
S127 phosphorylation regulates YAP and Yki
In a genetic screen for suppressors of phenotypes caused
by Hippo overexpression, we recovered three alleles of
yki. Remarkably, our yki alleles all affect the highly con-
served region surrounding S168, which corresponds to
S127 in YAP2. As shown in Figure 6A, two of the alleles
harbor mutations of H163 to Y and L, while the other
allele harbors a mutation of P170 to S. All three alleles
strongly suppress the small wing phenotype caused by
Hippo overexpression and exhibit weak semidominant
phenotypes (Fig. 6B, panels a–d). The most noticeable
dominant phenotypes were in the wings, which were
slightly larger and often had defects in the posterior
cross-vein. Due to the large wing phenotypes, we named
these yki alleles after the large ears of the Dumbo car-
toon character. The suppression of Hippo-hyperactivated
phenotypes suggests that these mutant Yki proteins are
constitutively active and evade suppression by Hippo
signaling, which is consistent with our biochemical
studies of YAP2. If this were the case, we would expect
that cells homozygous mutant for these yki alleles phe-
nocopy hippo loss-of-function mutations. Indeed, we
found that ykiDbomutant cells had a growth advantage
over wild-type cells (Fig. 6B, panels e–h). This effect was
apparent in adult eyes showing that ykiDbomutant cells,
marked by the lack of pigmentation, outcompeted red
wild-type cells (Fig. 6B, panels e,f) as well as in develop-
ing eye tissues at larval stages, where eyFLP-induced
ykiDbomutant clones occupied nearly the entire disc tis-
sues, in contrast to wild-type control clones, which oc-
cupied less than half of the discs (Fig. 6B, panels g,h). In
addition, ykiDbomutant eye tissue exhibited ectopic cell
proliferation posterior to the morphogenetic furrow (Fig.
6B, panels i,i?), a region where wild-type cells exit the
cell cycle and start to differentiate, and produced an ex-
cess number of interommatidial cells (Fig. 6B, panels j,j?).
Characteristic for mutations in Hippo signaling compo-
nents, cyclin E and ex, transcriptional targets of Hippo
signaling, were up-regulated in ykiDboclones (Fig. 6B,
panels k,l). Interestingly, this up-regulation was also ob-
served in heterozygous cells, perhaps due to the semi-
dominant nature of the ykiDboalleles.
The combination of these phenotypes is very distinc-
tive for loss of Hippo signaling, although they are not as
severe as those observed for null mutants of hippo or
wts. The ykiDboalleles thus mimic hypomorphic hippo
alleles. We conclude that the YkiDbomutations produce
YAP and 14–3–3 interaction. (A) Dephos-
phorylation abolishes the interaction be-
tween YAP2 and 14–3–3 in vitro. Flag-
YAP2 immunoprecipitated from trans-
fected HEK293 cells was treated with ?
phosphatase as indicated and then used to
HEK293 cell lysate. The products were
analyzed by Western blot. (B) Lats2 but not
Akt enhances YAP2 and 14–3–3 interac-
tion. Flag-YAP2 plasmids were cotrans-
fected with Myc-14–3–3 and other indi-
cated plasmids into HEK293 cells. Myc-
coimmunoprecipitated Flag-YAP2 was de-
tected. (C) Mutation of H122 but not P129
decreases YAP2 S127 phosphorylation by
Lats2. In vitro phosphorylation of YAP2
Lats2 was performed. Phosphorylation of
GST-YAP2 was detected by32P incorpora-
tion. (Bottom panel) GST-YAP2 input was
shown by Coomassie Blue staining. 4SA
(S127) denotes that four of the five Lats
phosphorylation sites were mutated to ala-
nine except Ser127. (D) Mutation of His122 in YAP2 impairs Ser127 phosphorylation and 14–3–3 binding. Indicated plasmids were
transfected into HEK293 cells. Flag-YAP2 was immunoprecipitated, and the immunoprecipitates were probed as indicated. (E) Pro129
of YAP2 is required for 14–3–3 binding. Experiments were similar to those in D.
S127 phosphorylation regulates
Zhao et al.
2754GENES & DEVELOPMENT
Drosophila melanogaster Yki wild-type and Dbo mutant proteins around the S127 (YAP2) residue. Mutated residues are shown in
green. (B) Dominant active yorkie mutations around the phosphorylation site S168 mimic hippo loss-of-function phenotypes. (Panel
a) Wild-type wing. (Panel b) Hpo overexpression driven by nubbin-Gal4. (Panel c) nubbin-Gal4 UAS-Hpo, ykiDbo/+. (Panel d) ykiDbo/+.
(Panel e) A fly with an eye mosaic for a mutation in the white gene. Clones were induced using the eye-specific FLP driver (eyFLP),
and a cell-lethal mutation on the homologous (w+) chromosome was used to eliminate twin spot clones, which increased the area of
the w−cell clones. (Panel f) A fly with a mosaic eye induced by the same method as in e. However, this fly carries a ykiDbomutation
on the w−chromosome. (Panels g,h) Eye imaginal discs from third instar larvae containing wt and ykiDbomutant clones that were
marked by the absence of GFP (gray). (Panels i–l) ykiDbomutant clones marked by the absence of GFP. (Panel i) Eye imaginal disc
containing ykiDbomutant clones and labeled for BrdU incorporation (red in panel i, and grayscale in panel i?). Asterisks indicate the
morphogenetic furrow, arrows indicate the second mitotic wave, and arrowheads point to ectopic cell proliferation in ykiDbomutant
clones posterior to the second mitotic wave. (Panle j) Mid-pupal retina stained with Discs large (Dlg) antibodies to visualize cell
outlines (red in panel j, and grayscale in panel j?). ykiDbomutant clones showed extra interommatidial cells (arrowhead). (Panel k)
ykiDbomutant clones showed up-regulated expression of Cyclin E (arrowheads) (red in panel k, and grayscale in panel k?), most
conspicuously behind the second mitotic wave (arrows). (Panel l) ykiDbomutant clones showed increased Ex (red in panel l, and
grayscale panel l?) levels in the eye imaginal disc. (C) The phosphorylation-defective YAP2-S127A is more active in promoting tissue
growth in Drosophila. (Panels a–d) Third instar larval eye discs were analyzed for the transcriptional activities of diap1-lacZ reporter
genes. Anterior is to the left. Red arrows indicate the morphogenetic furrow. (Panels e–h) Mid-pupal eye discs were stained with Discs
large (Dlg) antibody to outline cells. SEM (scanning electron microscopy) images of fly adult eyes are presented in panels i–l. Genotypes
of the fly tissues are GMR-Gal4/+; diap1-lacZ/+ (panel a), GMR-Gal4/UAS-Flag-YAP2; diap1-lacZ/+ (panel b), GMR-Gal4/UAS-Flag-
YAP2S127A; diap1-lacZ/+ (panel c), GMR-Gal4/UAS-yki-V5; diap1-lacZ/+ (panel d), wild-type (Canton S) (panels e,i), GMR-Gal4/UAS-
Flag-YAP2 (panels f,j), GMR-Gal4/UAS-Flag-YAP2S127A,(panels g,k), and GMR-Gal4/UAS-yki-V5 (panels h,l).
S127 phosphorylation regulates YAP and Yki biological function in vivo. (A) Alignment of the Homo sapiens YAP2 and the
YAP in cell contact inhibition
GENES & DEVELOPMENT2755
dominant active proteins that are not as efficiently sup-
pressed by Hippo signaling. This is likely due to reduced
phosphorylation by Wts (Yki-Dbo1 and Dbo2) and re-
duced 14–3–3 binding (Yki-Dbo3), as observed for the
respective YAP mutants.
In another line of evidence, we compared the activity
of YAP/Yki and the phosphorylation-deficient S127A
mutant in transgenic flies that overexpressed these pro-
teins in developing eyes. As expected, overexpression of
YAP2 or Yki increased the transcription of diap1-lacZ
(Fig. 6C, panels a–d) and CycE-lacZ (Supplementary Fig.
S5, panels a–d) reporter genes, transcriptional readouts
for Yki activity (Huang et al. 2005). Overexpression of
YAP2 had a moderate effect on eye size and slightly in-
creased the size of larval eye discs and adult eyes (Fig.
6C, panels f,j). The phosphorylation-defective YAP2-
S127A was more potent and caused a significant increase
in the size of eye discs and in the number of interomma-
tidial cells (Fig. 6C, panels g,k). The adult eyes of such
animals were overgrown, but folded and had a severe
morphological defect (Fig. 6C, panels a–c,i–k). All of
these phenotypes are reminiscent of warts and mats mu-
tants, and the YAP2-S127A was in fact as potent as the
fly Yki protein in promoting tissue growth (Fig. 6C).
These data, together with ones from the Dbo mutants,
suggest a critical role of YAP/Yki phosphorylation by
Lats/Wts in the negative regulation of YAP/Yki in vivo.
YAP regulates density-dependent gene expression
and alteration of YAP activity affects cell contact
As a transcription coactivator, YAP functions by regu-
lating gene expression. Gene expression microarray ex-
periments were performed to compare genes that are
regulated by YAP and by cell density. We found that the
set of genes induced by YAP2 significantly overlaps with
the set of genes that are repressed by high cell density
(Fig. 7A). Similarly, the set of genes repressed by YAP2
(possibly by indirect means) significantly overlaps with
the set of genes induced by high density. However, the
set of genes induced (or repressed) by YAP2 does not
significantly overlap with the set of genes induced (or
repressed) by high density. The opposite regulation of
gene expression by YAP and high cell density was con-
firmed by quantitative RT–PCR of selected genes (Fig.
7B). These observations indicate that YAP and cell den-
sity regulate many genes in opposite manners.
Our data indicate that YAP may play a role in cell
contact inhibition. To further investigate YAP regula-
tion by cell contact, scratch wounds were generated in
confluent cell cultures to relieve contact inhibition. As
shown in Figure 7C, both YAP staining intensity and
nuclear localization were significantly elevated in cells
at the border of the wound, while cells further away
showed cytoplasmic localization of YAP. Interestingly,
the nuclear YAP-positive cells were also positive for
Ki67, a marker of cell proliferation, indicating that these
cells have re-entered the cell division cycle. The above
data further demonstrate that YAP localization is regu-
lated by cell density and that nuclear YAP may promote
cell cycle entry.
To test the effect of increased YAP activity on contact
inhibition, we stably expressed YAP2 in NIH-3T3 cells.
YAP2-overexpressing NIH-3T3 cells kept proliferating
even after reaching confluency and resulted in a higher
saturation density than vector control cells (Fig. 7D).
Confluent cells were also analyzed for cell cycle distri-
bution by 5-bromo-2?-deoxyuridine (BrdU) incorporation
assay. Many more YAP2-expressing cells (23%) were still
in S phase compared with vector control cells (6%) (Fig.
7E). The above observations suggest that YAP overex-
pression may overcome contact inhibition, and therefore
further supports an important role of YAP inactivation
by the Hippo pathway in cell contact inhibition.
We tested whether interference of YAP activity was
able to restore contact inhibition in cancer cells that are
otherwise deficient of it. ACHN is a cancer cell line from
a metastatic human renal adenocarcinoma. It has been
reported to bear a deletion of Sav (Tapon et al. 2002),
which suggests a deficient Hippo pathway. Consistently,
the cells in this line clearly grow on top of each other and
pile up, even under low cell density (Fig. 7F), suggesting
loss of contact inhibition. We generated ACHN cells sta-
bly express YAP2-5SA-?C, which is the YAP2-5SA
nucleus-localizing form with a deletion of the C-termi-
nal transcription activation domain. This mutant YAP2
is insensitive to the Hippo pathway-induced cytoplasmic
translocation and cannot activate gene expression, and
therefore may act as a dominant-negative form. Al-
though the expression of this mutant YAP2 was as low
as endogenous YAP (Fig. 7F), its effect was dramatic. The
YAP2-5SA-?C-expressing ACHN cells grow as a single
layer and do not pile up even after confluent (Fig. 7F).
This result indicates, at least in Hippo pathway-deficient
cancer cells, that the loss of cell contact inhibition can
be restored by blocking endogenous YAP function, and
therefore further supports the function of YAP in contact
Elevated YAP protein and nuclear localization
Although YAP has been implicated as a candidate onco-
gene, it has not been reported whether YAP is indeed
activated in human cancers. We evaluated YAP expres-
sion in human cancers by immunohistochemical stain-
ing of tissue microarrays. Among the 115 cases of hepa-
tocellular carcinoma (HCC) samples examined, 63
samples (54%) showed strong YAP staining, while 95%
of normal liver tissue samples (40 out of 42 cases)
showed very weak staining, indicating a significant dif-
ference in YAP protein levels between normal and can-
cerous tissues (P < 0.001, Fisher exact test) (Fig. 8A,B).
Furthermore, the majority of HCC cells displayed stron-
ger nuclear YAP staining. These observations show that
dysregulation of YAP protein level and localization in-
deed occurs in human HCC. Similar observations were
made in prostate cancer tissues (P = 0.004) (Fig. 8A,B).
We speculate that YAP activation in cancer tissues is
Zhao et al.
2756GENES & DEVELOPMENT
likely due to mutation or dysregulation of the Hippo
pathway including YAP itself, and that uncontrolled
YAP activation may contribute to cancer development.
Recent studies have implicated YAP as an oncogene
(Overholtzer et al. 2006; Zender et al. 2006). However,
neither the precise biological function nor the physi-
ological regulation of YAP is clear. Our study demon-
strates that YAP functions downstream from and is in-
hibited by the Hippo pathway in cell contact inhibition
It has been reported previously that YAP2 S127 is
phosphorylated by Akt in response to growth factor
stimulation (Basu et al. 2003). However, we found that
phosphorylation of YAP2 S127 is not affected in PDK1
knockout cells, in which the Akt activity is abolished.
Furthermore, EGF, insulin, PI3K inhibitors, and Akt
overexpression did not affect YAP phosphorylation in
our experiments. In contrast, we presented data to show
that S127 in YAP2 is directly phosphorylated by Lats.
Given the fact that YAP is an oncogene and its activity is
inhibited by S127 phosphorylation, the previous model
that Akt phosphorylates YAP S127 is inconsistent with
the known function of Akt. We conclude that Lats plays
a direct role in YAP phosphorylation and inhibition.
However, it is still possible that YAP might be phos-
phorylated by Akt under some physiological or patho-
dent gene expression and alteration of YAP
activity affects cell contact inhibition. (A)
High cell density and YAP affect gene ex-
pression in opposite manners. YAP-regu-
lated genes were revealed by microarray
analyses of control and YAP-overexpress-
genes were also identified by microarray
analysis of sparse and confluent cells.
Genes that show more than twofold differ-
ences were used in the comparison. P val-
ues were calculated by Fisher exact test. (B)
YAP and cell density-regulated genes. To-
tal RNA isolated from NIH-3T3 cells sta-
bly expressing YAP2 or vector control (top
chart) and from low- or high-density cul-
tures (bottom chart) were analyzed by
quantitative RT–PCR and normalized to
HPRT (hypoxanthine phosphoribosyltrans-
ferase 1). (C) Correlation of cell prolifera-
tion and nuclear YAP localization. Conflu-
ent MCF10A culture was scratched. Six
hours later, cells were fixed and stained for
YAP and Ki67. (D) YAP promotes cell
growth and elevates saturation density.
Growth curves of NIH-3T3 cells stably ex-
pressing YAP2 or vector were determined.
Confluent density is indicated. (E) YAP
promotes proliferation of confluent cells.
Vector and YAP-overexpressing NIH-3T3
cells were cultured to confluence. Cells at
a similar density were pulse-labeled with
BrdU followed by staining with anti-BrdU
and 7-AAD (a fluorescent dye for total
DNA) for flow cytometric analysis. (F)
Dominant-negative YAP restores contact
inhibition in ACHN cancer cells. ACHN
cells stably expressing vector or Myc-
YAP2-5SA-?C were cultured to low den-
sity or confluence. Cell morphologies are
shown in the left panels. The loss of con-
tact inhibition in ACHN cells is evidence
that cells pile on top of each other. Myc-
YAP2-5SA-?C expression level is shown
by Western blot in the right panels.
YAP regulates density-depen-
YAP in cell contact inhibition
GENES & DEVELOPMENT2757
YAP in cell contact inhibition
An important observation described in this study is the
dramatic translocation of YAP between the nucleus and
the cytoplasm in response to cell density status. We pro-
pose a model that upon cell–cell contact, certain cell
surface receptors (Fat is a possible candidate) are acti-
vated via interaction with other surface proteins (such as
Dachsous) (Matakatsu and Blair 2004). The activated re-
ceptor then stimulates Merlin and Expanded, which in
turn stimulate Mst/Sav protein kinase activity. Active
Mst/Sav phosphorylates and activates the Lats/Mob
complex, which directly
HXRXXS motifs. Phosphorylated YAP then associates
with 14–3–3 and is sequestered in the cytoplasm (Fig.
8C). Remarkably, genetic screens in Drosophila identi-
fied three Yki gain-of-function alleles, and mutation of
the corresponding residues in YAP decreases phosphory-
lation or 14–3–3 binding. These results demonstrate the
functional importance and the inhibitory nature of the
phosphorylation of YAP-S127 (Yki-S168) by the Hippo
pathway. However, additional mechanisms of YAP regu-
lation may also exist, because YAP-5SA can be partially
inhibited by Mst and Lats in the reporter assays, and the
subcellular localization of this mutant still shows partial
response to cell density.
Several lines of evidence support the function of
YAP in contact inhibition. First, cell density regulates
Lats kinase activity and YAP nuclear/cytoplasmic shut-
tling. Second, scratching of confluent cultured cells in-
duces YAP nuclear localization in cells at the wound
edge. Those cells with nuclear YAP also enter the cell
cycle. Third, YAP-overexpressing cells fail to exit the
cell cycle when confluent and grow to a much higher
density. In fact, there is a strong correlation between
nuclear YAP protein levels and staining with the pro-
liferation marker Ki67 (data not shown). Fourth, ex-
pression of dominant-negative YAP restores contact in-
hibition in a human cancer cell line bearing deletion of
Sav. Fifth, YAP regulates many genes in a manner oppo-
site to high cell density. Furthermore, YAP is inhibited
by Merlin, which has been implicated in mediating
cell contact inhibition (Lallemand et al. 2003; Okada
et al. 2005). Together, our study indicates that YAP
plays a critical role in cell contact inhibition and that
the Hippo pathway may relay cell contact signals to in-
nuclear localization in human cancers.
Tissue microarrays of liver and prostate
cancer were stained with anti-YAP anti-
body (brown). Cell nuclei were counter-
Nuclear YAP protein is significantly el-
evated in human cancers. Samples were
scored based on median nuclear staining
intensity, ranging from 0 to 6 (0 for nega-
tive, and 6 for very strong staining). Strong
staining was considered a score of 2 or
higher for liver and 3 or higher for prostate.
P values (Fisher exact test) indicate the dif-
ferences in the proportions of strong YAP
staining between cancer and normal speci-
mens. (C) A model for YAP regulation by
cell contact via the Hippo pathway.
(A) Elevated YAP protein and
Zhao et al.
2758 GENES & DEVELOPMENT
activate YAP, thereby inducing contact inhibition
YAP in tumorigenesis
Our data show that YAP expression is frequently el-
evated in human cancers. More than 50% of HCC exam-
ined have increased nuclear YAP protein levels. Prostate
cancers also have significant elevation of YAP protein
levels and nuclear localization, although at a lower fre-
quency. These data indicate that YAP may play an im-
portant role in human tumorigenesis. However, in spite
of the high frequency of YAP overexpression we ob-
served, a relatively low incidence (5%–15%) of amplifi-
cation of the human chromosome 11q22 amplicon has
been reported in human tumors (Baldwin et al. 2005;
Snijders et al. 2005; Zender et al. 2006). Thus, we specu-
late that elevation of YAP protein levels in cancer is not
entirely due to gene amplification, but may instead re-
sult from dysregulation of the Hippo pathway. For ex-
ample, mutation of NF2 should result in inhibition of
the Hippo pathway and subsequent activation of YAP by
abrogation of inhibitory phosphorylation. We propose
that Merlin functions as a tumor suppressor, at least in
part by inactivating the YAP oncoprotein. It has also
been reported that both Sav and Mob are mutated in
tumor cell lines (Tapon et al. 2002; Lai et al. 2005).
Therefore, mutation or dysregulation of Mst/Sav and
Lats/Mob may contribute to uncontrolled YAP activa-
tion in human cancers.
Constitutive activation of YAP may cause evasion of
contact inhibition, therefore providing a growth advan-
tage for YAP-overexpressing cancer cells. YAP may also
be activated in other cancer types. Future studies to elu-
cidate the Hippo pathway and YAP regulation will not
only provide new insights into cell growth regulation,
especially cell contact inhibition, but will also be valu-
able in understanding tumorigenesis. Pharmacologic in-
tervention in the Hippo pathway—for example, inhibi-
tion of YAP—may be an effective strategy to treat can-
cers exhibiting YAP activation and/or overexpression.
Many important questions in the Hippo pathway re-
main to be addressed. For example, little is known about
how Merlin is activated by cell contact. One possibility
is CD44, a cell surface receptor implicated in cell contact
inhibition, acts upstream of Merlin (Morrison et al.
2001). It is also possible that Fat4 may function upstream
of Merlin or Ex1 to initiate the cell contact signaling
pathway. Another key question is the mechanism of Mst
activation by Merlin. Also elusive is the critical tran-
scription factor(s) mediating the physiological function
of YAP. We speculate that TEAD may have a role in YAP
function. Interestingly, Scalloped, the Drosophila TEAD
homolog, plays some roles in regulating cell proliferation
and apoptosis (Delanoue et al. 2004), suggesting an in-
triguing possibility of Scalloped as a Yki target transcrip-
tion factor. The bantam microRNA plays a critical role
in Drosophila to mediate the Hippo pathway signaling,
but there is no obvious bantam homolog in human ge-
nome (Nolo et al. 2006; Thompson and Cohen 2006). It
will be interesting to see whether functionally similar
microRNA exists in humans to mediate the physiologi-
cal function of YAP.
In summary, our study demonstrates that inactivation
of the YAP oncoprotein may play a critical role in cell
contact inhibition. This is at least partially accom-
plished by the Hippo pathway-dependent phosphoryla-
tion that promotes YAP binding to 14–3–3 and cytoplas-
mic localization. Dysregulation of YAP evades contact
inhibition and may contribute to tumorigenesis.
Materials and methods
Cell culture, transfection, and retroviral infection
HEK293 cells, 293T cells, HeLa cells, NIH-3T3 cells, MEF cells,
ACHN cells, and the RT4-D6-P2T Schwannoma cells were cul-
tured in DMEM (Invitrogen) containing 10% FBS (Invitrogen)
and 50 µg/mL penicillin/streptomycin (P/S). MCF10A cells
were cultured in DMEM/F12 (Invitrogen) supplemented with
5% horse serum (Invitrogen), 20 ng/mL EGF, 0.5 µg/mL hydro-
cortisone, 10 µg/mL insulin, 100 ng/mL cholera toxin, and 50
µg/mL P/S. Transfection with lipofectamine was performed ac-
cording to the manufacturer’s instructions.
To generate wild-type or mutant YAP2-expressing cells, ret-
rovirus infection was performed by transfecting 293 Phoenix
retrovirus packaging cells with empty vector or pQCXIH-YAP2
constructs. Forty-eight hours after transfection, retroviral su-
pernatant was supplemented with 5 µg/mL polybrene, filtered
through a 0.45-µm filter, and used to infect MCF10A, NIH-3T3,
or ACHN cells. Thirty-six hours after infection, cells were se-
lected with 200 µg/mL hygromycin (Roche) in culture medium.
The RT4-D6-P2T Schwannoma cells with inducible Merlin ex-
pression have been described before (Morrison et al. 2001; Rong
et al. 2004).
For immunofluorescence staining, cells were cultured on cov-
erslips to appropriate density. Cells were fixed with 4% para-
formaldehyde for 15 min and then permeabilized with 0.1%
Triton X-100. After blocking in 3% BSA for 30 min, slides were
incubated with first antibody diluted in 1% BSA for 1.5 h. After
washing with PBS, slides were incubated with Alexa Fluor 488-
or 594-conjugated secondary antibodies (1:1000 dilution) for 1.5
h. The slides were then washed and mounted.
Immunoprecipitation and kinase assay
For the Lats2 and Mst2 kinase assays, HEK293 cells were trans-
fected with HA-Lats2 or Flag-Mst2. Forty-eight hours post-
transfection, cells were lysed with lysis buffer (50 mM HEPES at
pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 10 mM pyro-
phosphoate, 10 mM glycerophosphate, 50 mM NaF, 1.5 mM
Na3VO4, protease inhibitor cocktail [Roche], 1 mM DTT, 1 mM
PMSF) and immunoprecipitated with anti-HA or anti-Flag anti-
bodies. The immunoprecipitates were washed three times with
lysis buffer, followed by once with wash buffer (40 mM HEPES,
200 mM NaCl) and once with kinase assay buffer (30 mM
HEPES, 50 mM potassium acetate, 5 mM MgCl2). The immu-
noprecipitated Lats2 or Mst2 was subjected to a kinase assay in
the presence of 500 µM cold ATP, 10 µCi [?-32P]ATP, and 1 µg
of GST-YAP2 expressed and purified from Escherichia coli as
substrate. The reaction mixtures were incubated for 30 min at
YAP in cell contact inhibition
GENES & DEVELOPMENT2759
30°C, terminated with SDS sample buffer, and subjected to
SDS-PAGE and autoradiography. The same procedure was used
for endogenous Lats2 kinase assay, except that endogenous
Lats2 immunoprecipitated from NIH-3T3 cells was used.
For the luciferase reporter assay, 293T cells were seeded in 12-
well plates. 5× UAS-luciferase reporter, CMV-?-gal, and indi-
cated plasmids were cotransfected as described previously.
Thirty-six hours after transfection, cells were lysed and lucifer-
ase activity was assayed using the enhanced luciferase assay kit
obtained from BD Biosciences following the manufacturer’s in-
structions. All luciferase activities were normalized to ?-galac-
BrdU labeling and flow cytometric analysis
For cell cycle progression analysis, cells were cultured to desired
confluence. Cells were then labeled with BrdU and analyzed by
flow cytometry using the FITC BrdU Flow Kit obtained from BD
Biosciences following the manufacturer’s instructions. Briefly,
cells were pulse-labeled with 10 µM BrdU in culture medium
for 30 min. After trypsinization and PBS wash, cells were fixed
and permeabilized. Incorporated BrdU was exposed by DNase
treatment and then stained by FITC-conjugated anti-BrdU anti-
body. Total DNA was stained by 7-AAD (7-amino-actinomycin
D). Data were collected on a BD FACSCalibur and analyzed
with CellQuest Pro software.
RNA isolation and real-time PCR
Total RNA was isolated from cultured cells using Trizol reagent
(Invitrogen). cDNA was synthesized by reverse transcription us-
ing random hexamers and subjected to real-time PCR with
gene-specific primers in the presence of Cybergreen (Applied
Biosystems). Relative abundance of mRNA was calculated by
normalization to hypoxanthine phosphoribosyltransferase 1
Gene expression microarray analysis
For analysis of gene expression in YAP overexpression cells, the
cells were cultured to 90% confluency before harvest. For com-
paring gene expression in low- and high-density cultures, cells
were seeded at different densities and harvested at 30% or com-
plete confluency. Total RNA was extracted with Trizol (Invit-
rogen) followed by further purification using the RNeasy kit
(Qiagen). Biotinylated cRNAs were then prepared according to
the Affymetrix standard labeling protocol. The biotinylated
cRNAs were then fragmented and hybridized to the Affymetrix
GeneChip Mouse Genome 430 2.0 Array or the Human Ge-
nome U133 Plus 2.0 Array, respectively. Chips were washed and
stained with Streptavidin R-phycoerythrin (Invitrogen). After
scanning the chips, the data were analyzed using GCOS soft-
ware. Scaling was performed with a target intensity of 500 to
facilitate the comparison of multiple arrays. A cut-off value of
0.05 was applied to the detection P value to assign a present (P),
marginal (M), or absent (A) call to each probe set. A signal value
was calculated using the One-Step Tukey’s Biweight Estimate
to represent the relative abundance of a transcript. Up- or down-
regulation of a gene is determined by two criteria: first, at least
one P call in the two samples being compared; second, at least
a twofold change (or indicated) of the signal value. The micro-
array analysis was done at the Molecular Biology Core Labora-
tory (University of Michigan, School of Dentistry).
We thank Drs. Marius Sudol for the pCMV-Flag-YAP2 con-
struct, Duojia Pan for a full-length yki cDNA, Tian Xu for Lats2,
Brian Seed and Joseph Avruch for Mst2, Jiandie Lin for the 5×
UAS-luciferase reporter and the Gal4-TEAD4 and Gal4-Foxo3
constructs, and Dario R. Alessi for the PDK1+/+and PDK1−/−
embryonic stem cell lysates. We thank the Developmental
Studies Hybridoma Bank at the University of Iowa for Dlg an-
tibody. We also thank Drs. Stephen J. Weiss, Patrick Hu, Ken
Inoki, Chung-Han Lee, and Theresa A. Reno for critical reading
of the manuscript. This work was supported by grants from NIH
(to K.L.G), and the National Science Foundation (IOS-0641914)
Abercrombie, M. 1979. Contact inhibition and malignancy. Na-
ture 281: 259–262.
Baldwin, C., Garnis, C., Zhang, L., Rosin, M.P., and Lam, W.L.
2005. Multiple microalterations detected at high frequency
in oral cancer. Cancer Res. 65: 7561–7567.
Basu, S., Totty, N.F., Irwin, M.S., Sudol, M., and Downward, J.
2003. Akt phosphorylates the Yes-associated protein, YAP,
to induce interaction with 14–3–3 and attenuation of p73-
mediated apoptosis. Mol. Cell 11: 11–23.
Bennett, F.C. and Harvey, K.F. 2006. Fat cadherin modulates
organ size in Drosophila via the Salvador/Warts/Hippo sig-
naling pathway. Curr. Biol. 16: 2101–2110.
Cho, E., Feng, Y., Rauskolb, C., Maitra, S., Fehon, R., and Irvine,
K.D. 2006. Delineation of a Fat tumor suppressor pathway.
Nat. Genet. 38: 1142–1150.
Conlon, I. and Raff, M. 1999. Size control in animal develop-
ment. Cell 96: 235–244.
Delanoue, R., Legent, K., Godefroy, N., Flagiello, D., Dutriaux,
A., Vaudin, P., Becker, J.L., and Silber, J. 2004. The Dro-
sophila wing differentiation factor vestigial-scalloped is re-
quired for cell proliferation and cell survival at the dorso–
ventral boundary of the wing imaginal disc. Cell Death Dif-
fer. 11: 110–122.
Eagle, H. and Levine, E.M. 1967. Growth regulatory effects of
cellular interaction. Nature 213: 1102–1106.
Edgar, B.A. 2006. From cell structure to transcription: Hippo
forges a new path. Cell 124: 267–273.
Halbleib, J.M. and Nelson, W.J. 2006. Cadherins in develop-
ment: Cell adhesion, sorting, and tissue morphogenesis.
Genes & Dev. 20: 3199–3214.
Hamaratoglu, F., Willecke, M., Kango-Singh, M., Nolo, R.,
Hyun, E., Tao, C., Jafar-Nejad, H., and Halder, G. 2006. The
tumour-suppressor genes NF2/Merlin and Expanded act
through Hippo signalling to regulate cell proliferation and
apoptosis. Nat. Cell Biol. 8: 27–36.
Hanahan, D. and Weinberg, R.A. 2000. The hallmarks of cancer.
Cell 100: 57–70.
Hariharan, I.K. 2006. Growth regulation: A beginning for the
hippo pathway. Curr. Biol. 16: R1037–R1039.doi: 10.106/
Harvey, K. and Tapon, N. 2007. The Salvador–Warts–Hippo
pathway—An emerging tumour-suppressor network. Nat.
Rev. Cancer 7: 182–191.
Harvey, K.F., Pfleger, C.M., and Hariharan, I.K. 2003. The Dro-
sophila Mst ortholog, hippo, restricts growth and cell prolif-
eration and promotes apoptosis. Cell 114: 457–467.
Hergovich, A., Stegert, M.R., Schmitz, D., and Hemmings, B.A.
2006. NDR kinases regulate essential cell processes from
yeast to humans. Nat. Rev. Mol. Cell Biol. 7: 253–264.
Zhao et al.
2760 GENES & DEVELOPMENT
Huang, J., Wu, S., Barrera, J., Matthews, K., and Pan, D. 2005. Download full-text
The Hippo signaling pathway coordinately regulates cell pro-
liferation and apoptosis by inactivating Yorkie, the Dro-
sophila Homolog of YAP. Cell 122: 421–434.
Kango-Singh, M., Nolo, R., Tao, C., Verstreken, P., Hiesinger,
P.R., Bellen, H.J., and Halder, G. 2002. Shar-pei mediates cell
proliferation arrest during imaginal disc growth in Dro-
sophila. Development 129: 5719–5730.
Lai, Z.C., Wei, X., Shimizu, T., Ramos, E., Rohrbaugh, M., Ni-
kolaidis, N., Ho, L.L., and Li, Y. 2005. Control of cell prolif-
eration and apoptosis by mob as tumor suppressor, mats.
Cell 120: 675–685.
Lallemand, D., Curto, M., Saotome, I., Giovannini, M., and Mc-
Clatchey, A.I. 2003. NF2 deficiency promotes tumorigenesis
and metastasis by destabilizing adherens junctions. Genes &
Dev. 17: 1090–1100.
Mah, A.S., Elia, A.E., Devgan, G., Ptacek, J., Schutkowski, M.,
Snyder, M., Yaffe, M.B., and Deshaies, R.J. 2005. Substrate
specificity analysis of protein kinase complex Dbf2–Mob1
by peptide library and proteome array screening. BMC Bio-
chem. 6: 22. doi: 10.1186/1471-2091-6-22.
Matakatsu, H. and Blair, S.S. 2004. Interactions between Fat and
Dachsous and the regulation of planar cell polarity in the
Drosophila wing. Development 131: 3785–3794.
McClatchey, A.I. and Giovannini, M. 2005. Membrane organi-
zation and tumorigenesis—The NF2 tumor suppressor, Mer-
lin. Genes & Dev. 19: 2265–2277.
Morrison, H., Sherman, L.S., Legg, J., Banine, F., Isacke, C., Hai-
pek, C.A., Gutmann, D.H., Ponta, H., and Herrlich, P. 2001.
The NF2 tumor suppressor gene product, merlin, mediates
contact inhibition of growth through interactions with
CD44. Genes & Dev. 15: 968–980.
Muslin, A.J. and Xing, H. 2000. 14–3–3 proteins: Regulation of
subcellular localization by molecular interference. Cell. Sig-
nal. 12: 703–709.
Nolo, R., Morrison, C.M., Tao, C., Zhang, X., and Halder, G.
2006. The bantam microRNA is a target of the hippo tumor-
suppressor pathway. Curr. Biol. 16: 1895–1904.
Okada, T., Lopez-Lago, M., and Giancotti, F.G. 2005. Merlin/
NF-2 mediates contact inhibition of growth by suppressing
recruitment of Rac to the plasma membrane. J. Cell Biol.
Okada, T., You, L., and Giancotti, F.G. 2007. Shedding light on
Merlin’s wizardry. Trends Cell Biol. 17: 222–229.
Overholtzer, M., Zhang, J., Smolen, G.A., Muir, B., Li, W., Sgroi,
D.C., Deng, C.X., Brugge, J.S., and Haber, D.A. 2006. Trans-
forming properties of YAP, a candidate oncogene on the
chromosome 11q22 amplicon. Proc. Natl. Acad. Sci. 103:
Pan, D. 2007. Hippo signaling in organ size control. Genes &
Dev. 21: 886–897.
Pantalacci, S., Tapon, N., and Leopold, P. 2003. The Salvador
partner Hippo promotes apoptosis and cell-cycle exit in Dro-
sophila. Nat. Cell Biol. 5: 921–927.
Rong, R., Surace, E.I., Haipek, C.A., Gutmann, D.H., and Ye, K.
2004. Serine 518 phosphorylation modulates merlin intra-
molecular association and binding to critical effectors im-
portant for NF2 growth suppression. Oncogene 23: 8447–
Silva, E., Tsatskis, Y., Gardano, L., Tapon, N., and McNeill, H.
2006. The tumor-suppressor gene fat controls tissue growth
upstream of expanded in the hippo signaling pathway. Curr.
Biol. 16: 2081–2089.
Snijders, A.M., Schmidt, B.L., Fridlyand, J., Dekker, N., Pinkel,
D., Jordan, R.C., and Albertson, D.G. 2005. Rare amplicons
implicate frequent deregulation of cell fate specification
pathways in oral squamous cell carcinoma. Oncogene 24:
St John, M.A., Tao, W., Fei, X., Fukumoto, R., Carcangiu, M.L.,
Brownstein, D.G., Parlow, A.F., McGrath, J., and Xu, T.
1999. Mice deficient of Lats1 develop soft-tissue sarcomas,
ovarian tumours and pituitary dysfunction. Nat. Genet. 21:
Takahashi, Y., Miyoshi, Y., Takahata, C., Irahara, N., Taguchi,
T., Tamaki, Y., and Noguchi, S. 2005. Down-regulation of
LATS1 and LATS2 mRNA expression by promoter hyper-
methylation and its association with biologically aggressive
phenotype in human breast cancers. Clin. Cancer Res. 11:
Tapon, N., Harvey, K.F., Bell, D.W., Wahrer, D.C., Schiripo,
T.A., Haber, D.A., and Hariharan, I.K. 2002. salvador Pro-
motes both cell cycle exit and apoptosis in Drosophila and is
mutated in human cancer cell lines. Cell 110: 467–478.
Thompson, B.J. and Cohen, S.M. 2006. The Hippo pathway
regulates the bantam microRNA to control cell proliferation
and apoptosis in Drosophila. Cell 126: 767–774.
Tyler, D.M. and Baker, N.E. 2007. Expanded and fat regulate
growth and differentiation in the Drosophila eye through
multiple signaling pathways. Dev. Biol. 305: 187–201.
Udan, R.S., Kango-Singh, M., Nolo, R., Tao, C., and Halder, G.
2003. Hippo promotes proliferation arrest and apoptosis in
the Salvador/Warts pathway. Nat. Cell Biol. 5: 914–920.
Vassilev, A., Kaneko, K.J., Shu, H., Zhao, Y., and DePamphilis,
M.L. 2001. TEAD/TEF transcription factors utilize the acti-
vation domain of YAP65, a Src/Yes-associated protein local-
ized in the cytoplasm. Genes & Dev. 15: 1229–1241.
Wei, X., Shimizu, T., and Lai, Z.C. 2007. Mob as tumor sup-
pressor is activated by Hippo kinase for growth inhibition in
Drosophila. EMBO J. 26: 1772–1781.
Willecke, M., Hamaratoglu, F., Kango-Singh, M., Udan, R.,
Chen, C.L., Tao, C., Zhang, X., and Halder, G. 2006. The fat
cadherin acts through the hippo tumor-suppressor pathway
to regulate tissue size. Curr. Biol. 16: 2090–2100.
Williams, M.R., Arthur, J.S., Balendran, A., van der Kaay, J.,
Poli, V., Cohen, P., and Alessi, D.R. 2000. The role of 3-phos-
phoinositide-dependent protein kinase 1 in activating AGC
kinases defined in embryonic stem cells. Curr. Biol. 10: 439–
Wu, S., Huang, J., Dong, J., and Pan, D. 2003. hippo encodes a
Ste-20 family protein kinase that restricts cell proliferation
and promotes apoptosis in conjunction with salvador and
warts. Cell 114: 445–456.
Yagi, R., Chen, L.F., Shigesada, K., Murakami, Y., and Ito, Y.
1999. A WW domain-containing yes-associated protein
(YAP) is a novel transcriptional co-activator. EMBO J. 18:
Yin, F. and Pan, D. 2007. Fat flies expanded the hippo pathway:
A matter of size control. Sci. STKE 2007: pe12. doi: 10.1126/
Zender, L., Spector, M.S., Xue, W., Flemming, P., Cordon-
Cardo, C., Silke, J., Fan, S.T., Luk, J.M., Wigler, M., Hannon,
G.J., et al. 2006. Identification and validation of oncogenes in
liver cancer using an integrative oncogenomic approach.
Cell 125: 1253–1267.
YAP in cell contact inhibition
GENES & DEVELOPMENT 2761