R

Yd

Qa

b

c

d

a

ARAA

KHLY

1

aYDsrDHPwafhadt

Sf

h1

Digestive and Liver Disease 47 (2015) 826–835

Contents lists available at ScienceDirect

Digestive and Liver Disease

journa l h om epage: www.elsev ier .com/ locate /d ld

eview Article

es-associated protein in the liver: Regulation of hepaticevelopment, repair, cell fate determination and tumorigenesis

uy Nguyena, Robert A. Andersd, Gianfranco Alpinia,b,c, Haibo Baia,b,c,∗

Research, Central Texas Veterans Health Care System, Temple, TX, United StatesDigestive Diseases Research Center, BaylorScott&White Healthcare, Temple, TX, United StatesDepartment of Internal Medicine and Medical Physiology, Texas A&M Health Science Center, Temple, TX, United StatesDepartment of Pathology, Johns Hopkins University, Baltimore, MD, United States

r t i c l e i n f o

rticle history:eceived 17 March 2015ccepted 14 May 2015vailable online 22 May 2015

eywords:ippo signalling pathway

a b s t r a c t

The liver is a vital organ that plays a major role in many bodily functions from protein production andblood clotting to cholesterol, glucose and iron metabolism and nutrition storage. Maintenance of liverhomeostasis is critical for these essential bodily functions and disruption of liver homeostasis causesvarious kinds of liver diseases, some of which have high mortality rate. Recent research advances ofthe Hippo signalling pathway have revealed its nuclear effector, Yes-associated protein, as an importantregulator of liver development, repair, cell fate determination and tumorigenesis. Therefore, a precise

iverAP

control of Yes-associated protein activity is critical for the maintenance of liver homeostasis. This reviewis going to summarize the discoveries on how the manipulation of Yes-associated protein activity affectsliver homeostasis and induces liver diseases and the regulatory mechanisms that determine the Yes-associated protein activity in the liver. Finally, we will discuss the potential of targeting Yes-associatedprotein as therapeutic strategies in liver diseases.

Published by Elsevier Ltd on behalf of Editrice Gastroenterologica Italiana S.r.l.

. Introduction

YAP (Yes-associated protein) was discovered 20 years ago inn effort to identify interacting proteins to Src family kinasees [1]. It burgeoned into a hot research objective after itsrosophila ortholog, Yki, was revealed as the effector of Hippoignalling pathway in 2005 [2]. The Hippo signalling pathway wasecently discovered as a mechanism that controls organ size inrosophila [2–4]. For a detailed review about the discovery of theippo signalling pathway in Drosophila, see reviews from Duojiaan [5,6]. This review will focus on the Hippo signalling path-ay in mammals. Components of the Hippo signalling pathway

re highly conserved throughout evolution. Accordingly, loss-of-unction mutants flies can be rescued with their correspondinguman orthologs [3,4,7]. The core Hippo signalling pathway is

kinase cascade (Fig. 1). The apical membrane-associated FERMomain protein NF2 (neurofibromin 2), directly binds and recruitshe Nuclear Dbf2-related family kinase, LATS1/2 (large tumour

∗ Corresponding author at: Olin E. Teague Medical Center, 1901 South 1sttreet, Bldg. 205, 1R58, Temple, TX 76504, United States. Tel.: +1 254 743 2502;ax: +1 254 743 0378.

E-mail address: hbai@sw.org (H. Bai).

ttp://dx.doi.org/10.1016/j.dld.2015.05.011590-8658/Published by Elsevier Ltd on behalf of Editrice Gastroenterologica Italiana S.r.

suppressor homolog 1/2), to the plasma membrane. Membranerecruitment, in turn, promotes LATS1/2 phosphorylation by theSte-20 family protein kinase, MST1/2 (Mammalian STE20-Like Pro-tein Kinase 1/2), together with the adaptor protein SAV1 (salvadorhomolog 1) [8]. In turn, LATS1/2, in a complex with small regulatorprotein MOB1 (Mps One Binder Homolog A), phosphorylates YAP.Phosphorylated YAP is sequestered in cytoplasm via interactionto 14-3-3 [9]. Cytoplasmic retention of YAP renders its interac-tion with TEA domain family transcription factors TEADs (TEAdomain family members) [10] to another transcription cofactor,Vgl4 (Vestigial like protein 4), thus stops YAP’s transcriptional out-puts [9,11,12].

Increasing Yki activity by overexpressing Yki as well as byablating Yki’s upstream negative regulators, leads to a similar over-growth phenotype in Drosophila due to increased proliferation andinhibited apoptosis. These observations raised an exciting ques-tion whether the Hippo pathway might play an analogous rolein mammals. To answer this question, mouse liver was chosenas the model organ, since liver size is known to be tightly regu-lated, constituting about 5% of the body weight [13]. Liver is a vital

organ that plays a major role in metabolism and has a number offunctions in the body, including glycogen storage, decompositionof red blood cells, plasma protein synthesis, hormone production,and detoxification. It contains two kinds of epithelia, hepatocyte

l.

Q. Nguyen et al. / Digestive and Liver Disease 47 (2015) 826–835 827

Fig. 1. The Hippo signalling pathway in mammals. The Hippo signalling pathway negatively regulates Yes-associated protein activity. When the Hippo signalling is inactive,Yes-associated protein accumulates in the nucleus and functions as transcriptional co-activator of the TEA domain family transcription factors. When the Hippo signallingis active, the apical membrane-associated neurofibromin 2, directly binds and recruits the Nuclear Dbf2-related family kinase, large tumour suppressor homolog 1/2, tothe plasma membrane. Membrane recruitment, in turn, promotes large tumour suppressor homolog 1/2 phosphorylation by mammalian STE20-Like Protein Kinase 1/2,together with the adaptor protein salvador homolog 1. In turn, large tumour suppressor homolog 1/2, in a complex with small regulator protein Mps One Binder Homolog A,phosphorylates Yes-associated protein. Phosphorylated Yes-associated protein is sequestered in the cytoplasm via interaction to 14-3-3 proteins. Cytoplasmic retention ofYes-associated protein renders its interaction with transcription partner TEA domain family transcription factors to another transcription cofactor, Vestigial like 4, thus stopsY , neurS omo

acmatlocsswqsm

2

2

lauYcsisaatgs

[21–24]. Notch signalling deficiency results in a same biliary phe-

es-associated protein’s transcriptional outputs. YAP, Yes-associated protein; NF2TE20-Like Protein Kinase 1/2; SAV1, salvador homolog 1; MOB1, Mps One Binder H

nd cholangiocyte, both differentiated from hepatoblast. Hepato-ytes make up 70–85% of the liver’s cytoplasmic mass and carry outost functions of the liver. A specific feature of hepatocyte is that

s a terminally differentiated cell, it retains ability to proliferateo replenish damaged liver. In normal conditions, hepatocytes areargely quiescent dividing approximately once per year. Transgenicverexpression of YAP in hepatocyte immediately induces quies-ent hepatocytes into cell cycle and dramatically expands the liverize to as much as 5 folds of the regular mouse liver weight [9]. Thistudy provided the first direct evidence implicating the Hippo path-ay in mammalian organ size control. And it evoked an interesting

uestion – whether YAP is normally required for liver homeosta-is, which might be answered using liver-specific Yap knockoutice.

. Physiological role of YAP in the liver

.1. Role of YAP in liver development

Complete knockout of Yap results in early embryonic letha-ity [14]. To analyze YAP’s function at later stages of developmentnd in adult tissue, conditional Yap knockout mice were generatedsing loxP technique (Yapf/f) [15]. By crossing with Alb-Cre mice,ap deletion (Alb-Cre; Yapf/f) in both cholangiocytes and hepato-ytes starts from E13.5 and achieves efficient deletion at perinataltage (E18 and P1) [16,17], which is a critical time point duringntrahepatic bile duct development [18,19]. Intrahepatic bile ductstart developing as early as around E13.5, hepatic progenitor cellsdjacent to the portal veins undergo ductal commitment, forming

structure known as the ductal plate. Around birth (E18 to P1),

ubulogenesis occurs at discrete points along the ductal plates,iving rise to primitive ductal structures. These primitive ductaltructures will incorporate into portal mesenchyme to become

ofibromin 2; LATS1/2, large tumour suppressor homolog 1/2; MST1, Mammalianlog A; TEAD, TEA domain family transcription factors; Vgl4, Vestigial like 4.

mature bile ducts during the first 2 weeks of life while theremaining unincorporated ductal plates regress [18,19]. The duc-tal plates form normally but the tubulogenesis fails in Alb-Cre;Yapf/f liver, which results in bile duct paucity 2 weeks afterbirth [15]. However, whether YAP affects ductal commitmentis not clear as Yap deletion in Alb-Cre; Yapf/f liver is not effi-cient until E18. Bile duct paucity in Alb-Cre; Yapf/f mice inducescompensatory ductular reaction, accompanied by steatosis andfibrosis in the adult liver. The hepatocytes develop normallyin these mice but with a higher turnover rate in the adultliver. This high hepatocyte turnover may be due to reducedhepatocyte survival ability and bile duct paucity combinedtogether.

The important role of YAP in bile duct development was furthersupported by the phenotype of liver specific Nf2 knockout mice(Alb-Cre; Nf2f/f) [15]. NF2 is one of the upstream tumour suppressorsof the Hippo signalling pathway and negatively regulates Yki inDrosophila [20]. Consistently, liver specific Nf2 deficiency results ina reverse phenotype to Yap deficiency on bile duct development.As a portion of cholangiocytes on the ductal plates form primitiveducts in WT livers, and no cholangiocyte on the ductal plates formsprimitive ducts in Alb-Cre; Yapf/f livers, in Alb-Cre; Nf2f/f livers all thecholangiocytes on the ductal plates become bile ducts [15]. Thisbile ducts over-tubulogenesis phenotype of Alb-Cre; Nf2f/f liver isdue to increased YAP activity and can be suppressed by partial Yapdeletion [15].

The underlying mechanism of YAP-regulated bile duct tubulo-genesis remains to be identified. Several studies have suggestedthat the Notch signalling is one of the downstream effectors of YAP

notype as Yap deficiency [16,25]. However, direct genetic studiesare necessary to put Notch signalling downstream of YAP in bileduct development.

8 d Live

2

pbalCdNritpartadaba[h

2

tiatrehdhrltcaattraYtzeer

FgYdt

3

rM

28 Q. Nguyen et al. / Digestive an

.2. Role of YAP in repairing damaged liver

Although YAP is critical for developing liver, it is largely dis-ensable in quiescent adult liver. Yap deletion in adult mouse livery injecting Poly dI-dC to 8 weeks old Mx1-Cre; Yapf/f mice did notffect liver histology [26]. Similarly, Nf2 deletion in adult mouseiver with adenovirus delivered Cre recombinase (Ade-Cre) or Mx1-re only leads to very mild periporal hyperplasia compared toramatic and widespread biliary hyperplasia in Alb-Cre mediatedf2 deletion [27]. However, YAP is required to ensure sufficient

egenerative proliferation in response to liver injury. Cholestaticnjury induced by ligation of the mouse common bile duct (BDL)riggers repair responses including cholangiocyte and hepatocyteroliferation, both of which peak at day 5 post-BDL and calm downfter 14 days post-BDL [28]. YAP protein levels first increase theneturn to baseline correspondingly with cholangiocyte and hepa-ocyte proliferation levels in isolated cholangiocytes, hepatocytesnd whole liver [26]. Yap deletion with Mx1-Cre; Yapf/f significantlyecreases cholangiocyte proliferation, delays hepatocyte prolifer-tion and enhances parenchymal damage after BDL [26]. It has alsoeen shown that YAP protein levels increase during the regener-tion response in both mouse and rat partial hepatectomy model29–31]. However, the role of YAP in the repair process after partialepatectomy remains to be evaluated.

.3. Role of YAP in determining liver cell fate

Hepatocytes and cholangiocytes, both differentiated from hepa-oblasts and/or putative liver progenitors, are the only two epithelian the liver. The relationship among hepatocytes, cholangiocytesnd putative liver progenitors, as well as the factors that determineheir cell fate specification are long standing questions in the liveresearch field [32,33]. Recently, Yimlamai et al. provided definitivevidence showing different YAP levels/activity could determineepatic cell fates and mature hepatocytes retain the capacity ofedifferentiation [24]. In adult mouse liver, YAP is expressed atigh levels in bile ducts, with many cholangiocytes displayingobust nuclear YAP expression. Conversely, YAP is expressed atow levels in hepatocytes, where the distribution is throughouthe cell [15]. Further addition of active YAP protein in cholangio-ytes leads to hyperplasia, but not to a ductal/progenitor-cell-likeppearance. The interesting observation came from overexpressingctive form of YAP in mature hepatocytes, which dedifferentiateshe mature hepatocytes into ductal/progenitor cells. More impor-antly, this hepatocyte-dedifferentiated ductal/progenitor cell canedifferentiate into hepatocyte after YAP activity levels drop [24]. Ingreement, Fitamant et al. also found that increasing endogenousAP activity due to Mst1/2 deficiency results in conversion of hepa-ocytes in periportal area to atypical ductal cells and override theonation programme in pericentral hepatocytes [34]. Their studiesvoked another interesting question: whether reducing YAP lev-ls in cholangiocytes will transdifferentiate them into hepatocytes,emains to be answered.

YAP also plays important role in other gastrointestinal tissues.or example, YAP levels are critical for intestinal regeneration, pro-enitor cells maintenance and differentiation [35,36]. IncreasingAP levels in pancreas affects pancreatic architecture [37,38]. Aetailed review about the Hippo signalling in other gastrointestinalissues can be found in reviews of Yu et al. [39]

. YAP activity and liver diseases

Consistent with the critical role of YAP in liver physiology, dys-egulation of YAP activity induces liver diseases in mouse models.ore importantly, elevated YAP levels were constantly found in

r Disease 47 (2015) 826–835

human liver diseases, especially liver cancer. In this section, we willsummarize the studies revealed the links between YAP activity andliver diseases.

3.1. YAP and liver cancer

Role of YAP in liver cancer demonstrated by animal models. Thefirst evidence of YAP in driving liver cancer formation came from amouse model of liver cancer initiated from progenitor cells har-bouring p53 loss and c-myc overexpression [40]. Genome wideanalysis of tumours in this mouse model revealed a recurrentamplification at mouse chromosome 9qA1 that contains Yap andcIAP1 (cellular inhibitor of apoptosis protein-1). Interestingly, thesyntenic region of human chromosome 11q22 is amplified in a vari-ety of human cancers [41–44]. Functional analysis indicated thatboth cIAP1 and Yap contribute to tumorigenesis and are required tosustain rapid growth of this amplicon-containing tumour.

The direct evidences of YAP in driving liver cancer camefrom genetic modified mouse models with manipulations ofHippo signalling pathway components (phenotypes summarizedin Tables 1 and 2). Transgenic overexpression of YAP in hepatocytes(ApoE-rtTA; TRE Yap) leads to liver enlargement and hepatocellularcarcinoma (HCC) formation [9]. Similarly, increasing YAP activity byliver-specific knockout of the Hippo signalling pathway upstreamtumour suppressors also leads to HCC and/or cholangiocarcinoma(CC) formation. Two studies of Alb-Cre; Nf2f/f mice revealed liver-specific deletion of Nf2 induces HCC and biliary tumour althoughwith some differences in mouse survival, time course of tumourprogression and the malignancy of biliary tumour, which is prob-ably due to different mouse genetic background [15,27]. Germlinemutants of mammalian Hippo kinase orthologs Mst1 (Mst1−/−) orMst2 (Mst2−/−) are viable and fertile, without liver abnormality[45], which may be due to functional redundancy of MST1 andMST2 as they share 76% identity in amino acid sequence [46].However, liver specific deletion of Mst2 on Mst1 null background,Alb-Cre; Mst1−/−;Mst2f/f and Ade-Cre;Mst1−/−;Mst2f/f, develop HCCat 3 months of age or 10 weeks after Ade-Cre administration, respec-tively [47]. It was also reported that CAGGCre-ER; Mst1−/−;Mst2f/−

mice develop both HCC and CC by 6 months of age after postnatalTamoxifen administration [48]. Similarly but less aggressively, liverspecific knockout of Sav1 also resulted in mixed types of tumoursaround one year of age [49,50]. More importantly, heterozygousdeletion of Yap, which results in no phenotypical manifestation byitself, significantly suppressed tumour formation in Alb-Cre; Nf2f/f

or Alb-Cre;Mst1−/−;Mst2f/f mice [15,34] (Table 2). These findingsprovided strong evidences that YAP is the driver of tumour for-mation in Nf2 or Mst1/2-deficient livers.

Further evidences of YAP’s involvement in liver cancer camefrom the hydrodynamic injection model and the carcinogeninduced mouse and rat HCC models. YAP and �-catenin showsconcurrent nuclear localization in human hepatoblastoma (HB)specimens and interacts with each other in HB cell line. Combinedhydrodynamic delivery of constitutively active Yap (YapS127A) andconstitutively active �-catenin (�N90/�-catenin) to mouse liverleads to HCC formation as early as 3 weeks after injection [51].Increased YAP protein levels were found in HCCs developed in micegiven diethylnitrosamine (DENA) followed by repeated treatmentswith 1,4-Bis[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP) [52].Nuclear YAP accumulation was observed in preneoplastic hepa-tocytes generated in rats 4 weeks after DENA treatment, whichsuggested YAP activation is an early event in liver tumorigenesis[53]. Taken together, all these studies provided solid evidences that

Yap is a proto-oncogene in the context of liver.

Role of YAP in liver cancer demonstrated by human studies. Inhumans, elevated YAP activity in different types of primary livercancers was observed by many studies (summarized in Table 3).

Q. Nguyen et al. / Digestive and Liver Disease 47 (2015) 826–835 829

Table 1Liver phenotypes of genetically modified mice with manipulations of single Hippo pathway components.

Gene Genotype Phenotype Reference

Amot1 Alb-Cre;Amotf/f Normal [111]Alb-Cre;Amotf/f + DDC13 Reduced levels of ductular reactions [111]

Fat2 Fat4−/− Death at birth [98]Alb-Cre;Fat4f/f Normal [99]

RASSF3 Rassf1a−/− Increased tumour susceptibility [100,101]Rassf5a−/− Normal [102]

Nf24 Nf2−/− Embryonically lethal between E6.5 and E7.0 [103]Nf2+/− Develop a variety of tumours, 9% HCC16 [104]AAV-Cre; Nf2f/f Hepatocyte dedifferentiation [24]Alb-Cre;Nf2f/f Over-developed bile ducts, hepatomegaly, bile duct harmatoma or CC17, HCC [15,27]Ade-Cre;Nf2f/f Mild periportal hyperplasia [27]Ade-Cre;Nf2f/f + PH14 HCC and CC [27]Mx1-Cre;Nf2f/f Mild OC18 proliferation [27]Mx1-Cre;Nf2f/f +PH HCC and CC [27]

Mst1/25 Mst1−/− or Mst2−/− Normal [45,47,48,50]Mst1−/−;Mst2+/− Hepatomegaly, HCC by 7 months [47,48]Mst1+/−;Mst2−/− Low incidence HCC by 15 months [47,48]Mst1−/−;Mst2−/− Embryonically lethal at E8.5 [47,48,50,105]Alb-Cre; Mst1/2f/f Hepatomegaly, HCC at 5–6 months [48,50]Alb-Cre; Mst1−/−;Mst2f/f Hepatomegaly, HCC by 3 months [47]Alb-Cre; Mst1−/−;Mst2f/− Hepatomegaly, HCC at 1 month [48]Ade-Cre; Mst1−/−;Mst2f/f or Ade-Cre; Mst1−/−;Mst2f/− Hepatomegaly, lethal HCC 10 weeks after Ade-Cre injection [47]CAGGCre-ER; Mst1−/−;Mst2f/− Hepatomegaly, HCC and CC at 6 months [48]

Sav16 Sav1−/− Embryonically lethal around E17.5 to E18.5 [106]Sav1+/− 74% mice developed liver tumours [49]Alb-Cre; Sav1f/f Hepatomegaly, HCC and CC by 13–14 months [49,50]MMTV-Cre; Sav1f/f Liver tumours by 14 months [50]Caggs-creER(T2); Sav1f/f Liver tumours by 14 months [50]

Lats1/27 Lats1−/− Neonatal death, infertility and growth retardation [107]Lats2−/− Embryonically lethal on or before E12.5 [108]

Mob18 Mob1a−/− or Mob1btr/tr Normal [109]Mob1a−/−;Mob1btr/tr Embryonically lethal [109]Mob1a−/−;Mob1btr/+ 50% liver tumours [109]

Yap9 ApoE-rtTA; TRE Yap Hepatomegaly, HCC [9]LAP1-tTA;TetO-YapS127A Hepatomegaly [97]Ck19-CreERT;TetO-Yap S127A Ductal hyperplasia [24]AAV-Cre; TetO-YapS127A Hepatomegaly, Hepatocyte dedifferentiation [24]Yap−/− Embryonically lethal at E8.5 [14]Alb-Cre;Yapf/f Bile duct paucity [15]Mx1-Cre;Yapf/f Normal [26]Mx1-Cre;Yapf/f + BDL15 More severe liver injury [26]

Taz10 Taz−/− Minor skeletal defects, renal cysts [110]TEAD11 Alb-rtTA;TRE TEAD2-DN Normal [11]Vgl412 ApoE-rtTA; TRE Vgl4 Normal [12]

1Amot, angiomotin; 2Fat, FAT atypical cadherin; 3RASSF, Ras Association (RalGDS/AF-6) Domain Family; 4Nf2, neurofibromin 2; 5Mst1/2, Mammalian STE20-Like ProteinK 1/2;1 rbonyh

Ch[vi

TL

1

M

inase 1/2; 6Sav1, salvador homolog 1; 7Lats1/2; large tumour suppressor homolog1TEAD, TEA domain family members; 12Vgl4, Vestigial like 4; 13DDC, 3,5-diethoxycaepatocellular carcinoma; 17CC, cholangiocarcinoma; 18OC, oval cell.

onsistently, most research groups observed more than 50% of

uman liver cancer tumours show elevated nuclear YAP expression9,51,54–60]. Multiple mechanisms were attributed to the ele-ated nuclear YAP expression, including gene amplification [40,56],ncreased Yap mRNA levels [51,55,56] and the dysregulation of the

able 2iver phenotypes of genetically modified mice with manipulations of combined Hippo pa

Gene Genotype

Nf21&Amot2 Alb-Cre;Nf2f/f;Amotf/f

Nf2&Fat43 Alb-Cre;Nf2f/f; Fat4f/f

Nf2&Yap4 Alb-Cre;Nf2f/f; Yapf/f

Alb-Cre;Nf2f/f; Yapf/+

Nf2&TEAD5 Alb-Cre;Nf2f/f;Alb-rtTA;TRE TEAD2-DN

Nf2&Vgl46 Alb-Cre;Nf2f/f;ROSA26-loxP-STOP-loxP-rtTA;TREVgl4

Mst1/27&Yap Alb-Cre; Mst1−/−;Mst2f/f;Yapf/+

Alb-Cre; Mst1−/−;Mst2f/f;Yapf/f

Yap&TEAD ApoE-rtTA;TRE Yap; TRE TEAD2-DNYap&Vgl4 ApoE-rtTA;TRE Yap; TRE Vgl4

Yap&Rbpj8 AAV-Cre;TetO YapS127A; Rpbjf/f

Nf2, neurofibromin 2; 2Amot, angiomotin; 3Fat4, FAT atypical cadherin 4; 4Yap, Yes-assocammalian STE20-Like Protein Kinase 1/2; 8Rbpj, recombination signal binding protein f

8Mob1, Mps One Binder Homolog A; 9Yap, Yes-associated protein; 10Taz, tafazzin;l-1,4-dihydrocollidine; 14PH, partial hepatectomy; 15BDL, bile duct ligation; 16HCC,

Hippo signalling with reduced expression of MST1 [47], LATS [57]

and RASSF1A [59]. Several genes were proposed as YAP down-stream targets in driving liver cancer development, including Axl(AXL receptor tyrosine kinase) [61], Survivin [56] and Amphiregulin[59]. Combining human and mouse data together, YAP is a very

thway components.

Phenotype Reference

Suppressed Nf2 phenotype [111]Nf2 phenotype is not suppressed [99]Suppressed Nf2 phenotype [15]Suppressed Nf2 phenotype [15]Suppressed Nf2 phenotype [11]Completely suppressed Nf2 phenotype [12]Suppressed Mst1/2 phenotype [34]Suppressed Mst1/2 phenotypeSuppressed YAP overexpression phenotype [11]Suppressed YAP overexpression phenotype [12]Suppressed YAP overexpression phenotype [24]

iated protein; 5TEAD, TEA domain family members; 6Vgl4, Vestigial like 4; 7Mst1/2,or immunoglobulin kappa J region.

830 Q. Nguyen et al. / Digestive and Liver Disease 47 (2015) 826–835

Table 3Summary of Yes-associated protein expression survey studies in human liver cancer samples.

Cancer type Patients origin No. controlsamples

Nuclear YAP8

expressionNo. tumoursamples

Nuclear YAPexpression

Paired con-trol/tumour?

Reference

HCC1 USA6 N/A7 N/A 20 65% N9 [9]HCC USA 42 5% 115 54% N [54]HCC Hong Kong 177 9.0% 177 62.1% Y10 [55]HCC USA 38 25% 70 85% N [57]HCC USA 83 28% 87 71% Y [56]HCC China 20 5% 40 42.5% N [58]HCC China 10 Very low 70 51.4% N [60]HCC Korea 15 20% 22 73% N [59]HCC Korea N/A N/A 100 31% N [59]HCC Europe N/A N/A 103 65% N [51]CC2 USA 8 N/A 16 98% N [57]ICC3 USA 5 None 10 90% N [56]ICC Europe N/A N/A 62 98% N [51]cHCC-CC4 Korea N/A N/A 58 67.2% N [59]HB5 USA 5 N/A 22 73% Y [57]HB Europe/USA N/A N/A 94 85% N [51]

1 angiob prot

pt

3

stecacYi

4

dftYpttl

4

terHtf[itlrsTlh

HCC, hepatocellular carcinoma; 2CC, cholangiocarcinoma; 3ICC, intrahepatic chollastoma; 6USA, United States of America; 7N/A, not available; 8YAP, Yes-associated

romising therapeutic target for liver cancer. Further studies needo focus on how to efficiently and specifically inhibit YAP activity.

.2. YAP and other liver diseases

Besides liver cancer, dysregulation of YAP activity was con-tantly observed in biliary diseases, which is in consistence withhe role of YAP in bile duct homeostasis. Elevated nuclear YAPxpression was found in ductular reactions of primary sclerosingholangitis (PSC) and primary biliary cirrhosis (PBC) [26] as wells in bile duct epithelium of neonates with biliary atresia [62]. Inholestatic livers of young patients, hepatocytes also display robustAP staining [63]. Whether YAP plays a role in other liver diseases

s an interesting question and remains to be investigated.

. Regulatory mechanisms of YAP activity in the liver

In the previous sections, we summarized accumulative evi-ences showing that precise regulation of YAP activity is criticalor liver homeostasis and tumorigenesis. Therefore, it is impor-ant to completely understand the mechanisms that regulateAP activity. Multiple mechanisms including mRNA transcription,ost-transcriptional modification, nucleocytoplasmic transporta-ion and protein stability participate in regulating YAP activity. Inhis section, we will review up-to-date discoveries on these regu-atory mechanisms in the liver (Fig. 2).

.1. Yap mRNA transcription

Several transcription factors including GABP (GA-binding pro-ein), homeobox protein NANOG and CREB (cAMP responselement-binding protein) were reported to occupy Yap promoteregion and upregulate Yap mRNA transcription in mouse liver orCC cells. Wu et al. used Yap promoter region as bait to pull down

he binding proteins from mouse liver lysate and identified Etsamily transcription factor complex GABP�/GABP� as candidate31]. Yap promoter contains 16 Ets family transcription factor bind-ng site (EBS) and multiple EBS sequences were bound by GABPo regulate Yap transcription. Interestingly, GABP� is phosphory-ated and inhibited by LATS1, which indicates that Hippo signallingegulates YAP activity through both phosphorylation and tran-

cription [31]. Functional screening for oncogenes regulated byLR4 (Toll-like receptor 4) and NANOG using a lentiviral cDNAibrary from tumour-initiating stem-like cells (TICs) in mouse anduman HCC tumours revealed YAP as one of the candidates [64].

carcinoma; 4HCC-CC, combined hepatocellular-cholangiocarcinoma; 5HB, hepato-ein; 9N, no; 10Y, yes.

Chip-qPCR assay indicated that NANOG binds to two NANOG con-sensus sites in Yap promoter (−1294 and −323 nt). Two studiesusing HCC cell lines suggested CREB regulates Yap mRNA transcrip-tion but mapped CREB binding sites to different regions of Yappromoter [58,65]. Zhang et al. mapped the CREB binding regionon Yap promoter to be −232/+115, where is also occupied by HBx(Hepatitis B virus X protein) to activate Yap transcription throughCREB [58]. Wang et al. mapped the CREB binding region on Yappromoter to be −608/−439 [65]. Amphiregulin and HB-EGF (epi-dermal growth factor) treatment is able to increase Yap mRNAlevels through EGFR (epidermal growth factor receptor) and MEK1(Dual specificity mitogen-activated protein kinase kinase 1), but itis not clear how EGFR signalling increases Yap transcription [66].

Yap transcription is regulated epigenetically by Menin. As a scaf-fold protein, Menin partners with mixed-lineage leukemia (MLL)histone methyltransferase to regulate the transcription of targetgenes by altering histone tail modifications [67]. In an effort toidentify the promoters occupied by Menin, Xu et al. performedChIP-on-chip screens with Menin antibody in HepG2 cells [68].Yap is one of the top picks with promoter occupied by Menincoinciding with the transcriptional activation of H3 modification.Interestingly, Menin epigenetically upregulates Yap transcriptionspecifically in HCC cell lines as although Menin binds to the pro-moter of Yap in other types of cell lines including MCR-7, Wilm’sand A549, there are no obvious effects on H3K4me3 levels at Yap1foci by Menin overexpression.

4.2. Post transcriptional modification to Yap mRNA

Liu et al. used bioinformatics technology to screen miRNAs thatmay target the 3′ UTR region of Yap gene for degradation andidentified miR-375 as a candidate [69]. Luciferase reporter assayconfirmed that miR-375 indeed targets the 3′ UTR region of Yapgene. Ectopic expression of miR-375 decreased endogenous YAPprotein levels as well as YAP transcriptional target – Ctgf (connec-tive tissue growth factor) mRNA levels in PLC and 97L HCC cell lines.Consistently, elevated YAP activity correlate with down-regulationof miR-375 in human HCC tissues, chemical-induced mouse HCCmodel and DENA-induced rat preneoplastic hepatocytes [52,53,69].

4.3. YAP protein phosphorylation/nucleocytoplasmic shuttling

YAP phosphorylation mediated by the Hippo signalling rep-resents the major regulation strategy of YAP activity. As atranscription coactivator, nuclear accumulation is necessary for

Q. Nguyen et al. / Digestive and Liver Disease 47 (2015) 826–835 831

Fig. 2. Mediators of Yes-associated protein activity in the liver. In the liver, Yes-associated protein activity is regulated through multiple mechanisms including transcription,post transcriptional modification, phosphorylation and protein degradation. Transcriptional factors GA-binding protein, homeobox protein NANOG, cAMP response element-binding protein/Hepatitis B virus X protein and scaffold protein Menin bind to Yes-associated protein promoter region to promote Yes-associated protein mRNA transcription.miR375 interacts with 3′ UTR of Yes-associated protein mRNA to promote mRNA degradation and inhibit translation. Association between Yes-associated protein and itstranscription factor partner TEA domain family members can be inhibited by competitively binding of Vestigial like protein 4 or through phosphorylation by the Hippopathway. Yes-associated protein Phosphorylation by the Hippo signalling pathway primes its further phosphorylation by casein kinase 1�/�, which leads to Yes-associatedprotein degradation through proteasome. Yes-associated protein degradation can be inhibited by TEA domain family members interaction. Bile acids inhibit Yes-associatedprotein phosphorylation through IQ Motif Containing GTPase Activating Protein 1. Active Hippo signalling pathway can also leads to GA-binding protein � phosphorylation,w otein;H 4; CK

YdaSsturtaR1

ppumaYolI�itYa

u

hich in turn reduces Yes-associated protein transcription. YAP, Yes-associated prepatitis B virus X protein; TEADs, TEA domain family members; Vgl4, Vestigial like

AP’s function. Nuclear accumulation or cytoplasmic retentionepends on YAP’s phosphorylation status. Dephosphorylated YAPccumulates in nucleus while phosphorylation at Ser127 (human,er112 in mouse YAP) by LATS1/2 kinase due to activated Hippoignalling sequesters YAP in the cytoplasm via 14-3-3 interac-ion [9]. In mouse liver, ablation of the Hippo signalling pathwaypstream tumour suppressors Nf2, Mst1/2 or Sav1 all resulted ineduced levels of phosphorylated YAP, which in turn drives liverumorigenesis [15,47,49]. Using HCC cell lines, overexpression ofnother Hippo signalling pathway upstream tumour suppressorASSF1A (Ras Association (RalGDS/AF-6) Domain Family Member) increased YAP phosphorylation [59].

Other than the Hippo pathway, bile acid regulates YAP phos-horylation through IQGAP1 (IQ motif containing GTPase activatingrotein 1) in the liver [63]. Phosphor-YAP levels were downreg-lated in the presence of elevated bile acid levels in Fxr−/−Shp−/−

ice or WT mice fed with 1% cholic acid. IQGAP1 is required for bilecid-induced YAP activation as 1% cholic acid diet failed to induceAP activation in Iqgap1−/− mice. Furthermore, adenoviral Iqgap1verexpression in WT liver was sufficient to increase YAP proteinevels. As a key scaffolding protein that interacts with E-cadherin,QGAP1 overexpression reduces cellular adhesion by dissociating-catenin from the cadherin–catenin complex [70]. These results

ndicated that elevated bile acid levels reduce hepatic cell adhesionhrough disrupting E-cadherin junctions, which in turn activates

AP. E-cadherin junction disruption has been previously reporteds an evoking event of YAP activation in breast cancer cell lines [71].

In Drosophila or non-hepatic mammalian cells, multiplepstream inputs including extracellular matrix area and stiffness

GABP, GA-binding protein; CREB, cAMP response element-binding protein; HBx,1�/�, casein kinase 1�/�; IQGAP1, IQ Motif Containing GTPase Activating Protein 1.

[72,73], cell density [54], cell detachment [74], F-actin cytoskele-ton [75], spectrin cytoskeleton [76,77], and cellular energy level[78,79] were shown to regulate YAP nuclear accumulation depend-ent or independent of the canonical Hippo signalling. Detailedreviews of these upstream YAP regulators can be found elsewhere[80,81]. Serum-derived small molecules sphingosine-1-phosphate(S1P) and lysophosphatidic acid (LPA) are particularly interestingbecause they are the first extracellular diffusible factors regulatingYAP activity [82,83]. S1P and LPA act through GPCRs (G-protein-coupled receptors) G12/13 to inhibit the Hippo pathway kinaseLATS1/2, which in turn activates YAP [83]. Whether these upstreaminputs regulate YAP activity in the context of liver requires furtherinvestigation.

4.4. YAP protein stability

Another strategy that cell employs to regulate YAP protein levelsis protein stability [11,84]. Overexpression of YAP’s transcriptionalpartner, TEAD, significantly increased YAP protein levels but notYap mRNA levels in mouse liver and as well as in HEK293 cells,suggesting TEAD binds and stabilizes YAP, which is further sup-ported by the observation that YAP S94A (human, Ser79 in mouseYAP), a mutant form defective in TEAD binding, had a shorterhalf-life than wild-type YAP [11]. Binding to TEAD may keep YAPin active form thus inhibit YAP protein degradation initiated by

LATS1/2 phosphorylation. LATS1/2 can phosphorylate YAP at all thefive consensus HxRxxS motifs. Among these phosphorylatable ser-ine/threonine residues, phosphorylation at S381 (human, Ser363in mouse YAP) primes this region for further phosphorylation by

8 d Live

CYplY

fYismsirc

5

bipm[foYTmoldtfCpYot

tcbitbreYuicb9ibetw

YYcoi

32 Q. Nguyen et al. / Digestive an

K1�/� (casein kinase 1�/�) at Ser384 (human, Ser366 in mouseAP) and S387 (human, Ser369 in mouse YAP) to generate a phos-horylated phosphodegron that will recruit SCF�-TRCP E3 ubiquitin

igase. Ubiquitination by SCF�-TRCP E3 ubiquitin ligase then leadsAP to proteasome degradation [84].

Although multiple mechanisms have already been revealed,urther efforts are still needed for a complete understanding ofAP activity regulation. For example, YAP protein levels elevate

n mouse livers following BDL [26]. However, the phosphorylationtatus of YAP and Yap mRNA levels remain stable excluding involve-ent of Hippo signalling and transcriptional regulation, which

uggested other mechanisms of YAP activity regulation. Moreover,t will be also important to define the contributions of variousegulating mechanisms to YAP activity in different physiologicalontexts.

. YAP transcriptional partners and targets in the liver

As a transcription coactivator, YAP does not bind to DNA directlyut, rather, partners with DNA-binding transcription factors. Stud-

es in mammalian cells have revealed several transcription factorartners for YAP including the p53 family p73 [85], the Runt familyember Runx2 [86], and the TEAD/TEF family transcription factors

10]. Among these reported YAP-interacting partners, TEAD/TEFamily transcription factors are identified as the major mediatorsf YAP function in mammalian cells [87,88]. To investigate whetherAP functions through TEAD factors in intact mammalian tissues,EAD proteins were inactivated specifically and conditionally in theouse liver [11]. The mouse genome contains four highly homol-

gous TEAD family members, all of which are expressed in theiver. Because different TEAD members appear to function redun-antly [89], dominant-negative scheme was employed to blockhe activity of all four TEAD factors with engineering a truncatedorm of TEAD2 that lacks its DNA-binding domain (TEAD2-DN) [11].o-overexpression of YAP and TEAD2-DN in mouse hepatocytesotently suppressed hepatomegaly and tumorigenesis caused byAP overexpression or loss of NF2. Thus, TEAD is the major partnerf YAP in the liver to induce the transcription of YAP’s downstreamargets.

mRNA microarray analysis has revealed a large number of geneshat are induced by YAP in mouse livers [9,24,50]. Among theseandidate YAP target genes, Axl is a direct target of YAP and haseen shown to contribute to YAP-dependent oncogenic functions

n HCC cell lines [61]; Ctgf [54] and Amphiregulin [90] were showno be direct targets of YAP using non-liver cell lines and found toe induced by YAP in mouse liver [91] and HCC cell lines [59],espectively; in human liver cancer survey studies, expression lev-ls of CTGF, Survivin and Glypican 3 were found to correlate withAP activity levels [56,57]. Two independent studies have put YAPpstream of the Notch signalling pathway [23,24]. YAP activation

n mouse livers increases mRNA levels of several Notch pathwayomponents including Notch1/2, Jag1 (Jagged 1), Hes1 (hes familyHLH transcription factor 1) and Sox9 (sex determining region Y-box) [23,24], among which Notch2 is a direct target of YAP [24]. Moremportantly, Notch inhibition through Rbpj (recombination signalinding protein for immunoglobulin kappa J region) deletion damp-ns the hepatocyte-to-ductal dedifferentiation phenotype in Yapransgenic liver, which put the Notch signalling downstream of YAPith genetic studies in intact mammalian tissue.

Other than inducing target gene expression transcriptionally,AP also exerts its influence through protein-protein interactions.

AP interaction with SMAD7/SMAD3 in the cytoplasm results inytoplasmic retention of SMAD3, then in turn leads to inactivationf the canonical TGF-� signalling, which contributes to tumour-nitiating stem-like cell formation in HCC caused by HCV and/or

r Disease 47 (2015) 826–835

alcohol-related chronic liver injury [64]. In HCC cell lines, YAP pro-motes CREB protein stability through interaction with MAPK14/p38(mitogen-activated protein kinase 14) and BTRC (beta-transducinrepeat containing E3 ubiquitin protein ligase) [65]. c-Myc transcrip-tion is indirectly regulated by YAP through interaction with c-Abl[92].

6. YAP as therapeutic target in liver cancer

The current available data including YAP upregulation in humanliver cancer, the tumorigenic phenotypes of mouse liver-specificNf2/Mst/Sav knockouts and Yap transgenic overexpression, sug-gests that increased YAP activity is a driving event for liver cancerformation. The fast advance of our knowledge to YAP’s regulatorsand regulations made it possible to inhibit YAP activity with mul-tiple schemes.

6.1. Reducing YAP protein levels

Reducing YAP protein levels is the most direct strategy todecrease YAP activity. Using Yap siRNA encapsulated into LNPs(siYap-LNPs), which improves the cellular uptake, stability andpharmacokinetics of the siRNA, Fitamant et al. demonstrated thattail vein injection of siYap-LNPs dramatically reduced liver tumourburden of Mst1/2-deficient mice [34]. Importantly, the treatmentregimen was well tolerated only inducing a slight inflammation inperiportal areas that was fully resolved within a short time follow-ing discontinuation of treatment.

6.2. Disrupting YAP–TEAD interaction

Using genetic modified mouse models, Liu-Chittenden andcolleagues demonstrated that the tumorigenic potential of YAPin mouse liver requires association to TEAD [11], which makesYAP–TEAD interaction an ideal drug target. By screening the JohnsHopkins Drug Library, they identified members of the porphyrinfamily as small molecule inhibitors of YAP–TEAD interactions.One of these inhibitors, verteporfin (VP) is used clinically as aphotosensitizer in photodynamic therapy for neovascular macu-lar degeneration. Administration of VP to YAP transgenic mice andNf2 knockout mice significant suppressed hepatomegaly caused byhepatocyte and biliary cell proliferation, respectively, in these twomouse models, without overt adverse effects in other organs [11].Additionally, VP treatment reduced cell proliferation more pro-nouncedly in HuCCT1 cholangiocarcinoma cell line than it did tonormal cholangiocyte cell line H69, suggesting the potential of VPin treating cholangiocarcinoma [62].

Although YAP mainly functions through TEAD, it is not the onlytranscription cofactor of TEAD. Vgl4 competes with YAP on bind-ing to TEAD via TDU domains [12,93,94]. Animal studies indicatedthat Vgl4 overexpression significantly suppressed transgenic YAP-induced liver overgrowth and tumorigenesis [12], which suggesteda potential therapeutic application of Vgl4 in cancer therapy. Jiaoand colleagues resolved the crystal structure of mouse Vgl4 tan-dem TDU region in complex with the YBD domain of mouse TEAD4,and based on this complex structure, they designed a “Super-TDU”peptide to inhibit YAP–TEAD interaction [94]. Administration ofSuper-TDU peptide to Helicobacter pylori and N-methyl-N′-nitro-N-nitrosoguanidine (MNNG) induced-gastric cancer mouse modelssignificantly suppressed tumour numbers and tumour cell prolif-eration.

6.3. Targeting YAP upstream regulators

The Hippo signalling pathway. The Hippo signalling path-way is the major upstream negative regulator of YAP activity.

d Live

EaspiQYa

ttfmti[t

tetraan

6

gsYTNsi

7

stmehfidid

oaimoYsPppIsus

Q. Nguyen et al. / Digestive an

levated Hippo signalling activities will efficiently inhibit YAPctivity. Integrin-linked kinase (ILK) suppresses the Hippoignalling pathway via phospho-inhibition of MYPT1-PP1 (myosinhosphatase target subunit 1-protein phosphatase 1), leading to

nactivation of NF2 [95]. Pharmaceutical inhibition of ILK withLT0267 in tumour cell lines and mouse xenograft models suppressAP-dependent tumour growth. However, this approach requiresn intact Hippo pathway signal flow in tumour cells.

F-actin cytoskeleton. F-actin cytoskeleton is fundamental to sus-ain YAP function. A mild disruption of the actin cytoskeleton leadso dramatic inhibition of YAP nuclear localization [72,74,96]. There-ore, factors those contribute to the actin cytoskeleton integrity

ay have therapeutic potential in inhibiting YAP activity. Indeed,he ROCK inhibitor Y27632 and the nonmuscle myosin heavy chainnhibitor blebbistatin have been shown to inhibit YAP activity72,96]. The potential of these inhibitors in treating YAP-dependentumours needs to be further evaluated in animal models.

AMP-activated kinase (AMPK). Cellular energy stress nega-ively regulates YAP activity through AMPK, which is a criticalnergy level sensor and mediator [78,79]. Pharmacological activa-ion of AMPK by AMPK activators, aminoimidazole carboxamideibonucleotide (AICAR) or metformin, significant suppressednchorage-independent growth of Lats-null cells with high YAPctivity. Furthermore, metformin inhibited tumour growth of Lats-ull cells xenografted mice [79].

.4. Targeting YAP downstream events

Downstream targets that are essential in YAP-driven tumori-enesis also have therapeutic potentials. For example, the Notchignalling has been indicated to be important downstream ofAP for the development of YAP-driven liver hyperplasia [24].herefore, YAP-driven tumours might benefit from treatment withotch inhibitors. Indeed, administration of �-secretase inhibitor

uppressed transgenic YAP induced dysplastic phenotype in thentestine [97].

. Concluding remarks

Benefited from serving as the model organ for the Hippoignalling research in the past several years, our understanding ofhe physiological function and molecular mechanism of YAP in the

ouse liver has been greatly advanced. These studies have firmlystablished YAP as a critical regulator of liver development andomeostasis. The importance of YAP in the liver is further solidi-ed by the realization that elevated YAP activity drives liver cancerevelopment, which makes small molecules inhibiting YAP activ-

ty have very promising therapeutic potentials for this devastatingisease.

Despite recent progress, our knowledge about this importantncogene remains incomplete. First of all, comparing to the rapiddditions of YAP’s upstream regulators, the outputs of YAP remainncompletely defined, especially in intact mammalian tissues. A

ajor challenge in the future is to elucidate the molecular naturef these outputs, and the molecular mechanisms that contribute toAP’s physiological function. Another challenge is to fully under-tand the molecular mechanisms that regulate YAP protein level.reviously most attention was paid to YAP activation by dephos-horylation. However, clinical survey indicated elevated total YAProtein levels are commonly observed in human tumour samples.

n combination with the realization that mutations in the Hippoignalling pathway are rarely observed in human cancer, dysreg-lation of YAP protein levels may be a more commonly employedcheme by cancer cells.

r Disease 47 (2015) 826–835 833

Conflict of interestNone declared.

Funding

Portion of this work was supported by the Dr. Nicholas C. High-tower Centennial Chair of Gastroenterology from Scott & White,a VA Research Career Scientist Award, a VA Merit award andNIH R01DK076898 grant to Dr. G. Alpini, and a grant award fromBaylorScott&White operational funds to H. Bai. This work is alsosupported by R01DK080736 to R. Anders.

This material is the result of work supported by resources at theCentral Texas Veterans Health Care System. The views expressed inthis article are those of the authors and do not necessarily representthe views of the Department of Veterans Affairs.

Acknowledgements

We apologize that we are unable to cite some important papersdue to the scope of this review.

References

[1] Sudol M. Yes-associated protein (YAP65) is a proline-rich phosphoproteinthat binds to the SH3 domain of the Yes proto-oncogene product. Oncogene1994;9:2145–52.

[2] Huang J, Wu S, Barrera J, et al. The Hippo signaling pathway coordinately reg-ulates cell proliferation and apoptosis by inactivating Yorkie, the DrosophilaHomolog of YAP. Cell 2005;122:421–34.

[3] Lai ZC, Wei X, Shimizu T, et al. Control of cell proliferation and apoptosis bymob as tumor suppressor, mats. Cell 2005;120:675–85.

[4] Wu S, Huang J, Dong J, et al. Hippo encodes a Ste-20 family protein kinasethat restricts cell proliferation and promotes apoptosis in conjunction withsalvador and warts. Cell 2003;114:445–56.

[5] Pan D. Hippo signaling in organ size control. Genes & Development2007;21:886–97.

[6] Pan D. The hippo signaling pathway in development and cancer. Develop-mental Cell 2010;19:491–505.

[7] Tao W, Zhang S, Turenchalk GS, et al. Human homologue of the Drosophilamelanogaster lats tumour suppressor modulates CDC2 activity. Nature Genet-ics 1999;21:177–81.

[8] Yin F, Yu J, Zheng Y, et al. Spatial organization of Hippo signaling at theplasma membrane mediated by the tumor suppressor Merlin/NF2. Cell2013;154:1342–55.

[9] Dong J, Feldmann G, Huang J, et al. Elucidation of a universal size-controlmechanism in Drosophila and mammals. Cell 2007;130:1120–33.

[10] Vassilev A, Kaneko KJ, Shu H, et al. TEAD/TEF transcription factors utilizethe activation domain of YAP65, a Src/Yes-associated protein localized in thecytoplasm. Genes & Development 2001;15:1229–41.

[11] Liu-Chittenden Y, Huang B, Shim JS, et al. Genetic and pharmacological dis-ruption of the TEAD-YAP complex suppresses the oncogenic activity of YAP.Genes & Development 2012;26:1300–5.

[12] Koontz LM, Liu-Chittenden Y, Yin F, et al. The Hippo effector Yorkie controlsnormal tissue growth by antagonizing scalloped-mediated default repression.Developmental Cell 2013;25:388–401.

[13] Michalopoulos GK, DeFrances MC. Liver regeneration. Science1997;276:60–6.

[14] Morin-Kensicki EM, Boone BN, Howell M, et al. Defects in yolk sac vasculo-genesis, chorioallantoic fusion, and embryonic axis elongation in mice withtargeted disruption of Yap65. Molecular and Cellular Biology 2006;26:77–87.

[15] Zhang N, Bai H, David KK, et al. The Merlin/NF2 tumor suppressor functionsthrough the YAP oncoprotein to regulate tissue homeostasis in mammals.Developmental Cell 2010;19:27–38.

[16] Geisler F, Nagl F, Mazur PK, et al. Liver-specific inactivation of Notch2, but notNotch1, compromises intrahepatic bile duct development in mice. Hepatol-ogy 2008;48:607–16.

[17] Xu X, Kobayashi S, Qiao W, et al. Induction of intrahepatic cholangiocellularcarcinoma by liver-specific disruption of Smad4 and Pten in mice. Journal ofClinical Investigation 2006;116:1843–52.

[18] Lemaigre FP. Development of the biliary tract. Mechanisms of Development2003;120:81–7.

[19] Zong Y, Stanger BZ. Molecular mechanisms of liver and bile duct development.Wiley Interdisciplinary Reviews Developmental Biology 2012;1:643–55.

[20] Hamaratoglu F, Willecke M, Kango-Singh M, et al. The tumour-suppressor

genes NF2/Merlin and expanded act through Hippo signalling to regulate cellproliferation and apoptosis. Nature Cell Biology 2006;8:27–36.

[21] Yu J, Poulton J, Huang YC, et al. The hippo pathway promotes Notch signalingin regulation of cell differentiation, proliferation, and oocyte polarity. PLoSONE 2008;3:e1761.

8 d Live

34 Q. Nguyen et al. / Digestive an

[22] Chen HJ, Wang CM, Wang TW, et al. The Hippo pathway controls polar cellfate through Notch signaling during Drosophila oogenesis. DevelopmentalBiology 2011;357:370–9.

[23] Tschaharganeh DF, Chen X, Latzko P, et al. Yes-associated protein up-regulatesJagged-1 and activates the Notch pathway in human hepatocellular carci-noma. Gastroenterology 2013;144:1530–42, e12.

[24] Yimlamai D, Christodoulou C, Galli GG, et al. Hippo pathway activity influ-ences liver cell fate. Cell 2014;157:1324–38.

[25] McCright B, Lozier J, Gridley T. A mouse model of Alagille syndrome:Notch2 as a genetic modifier of Jag1 haploinsufficiency. Development2002;129:1075–82.

[26] Bai H, Zhang N, Xu Y, et al. Yes-associated protein regulates the hepaticresponse after bile duct ligation. Hepatology 2012;56:1097–107.

[27] Benhamouche S, Curto M, Saotome I, et al. Nf2/Merlin controls progen-itor homeostasis and tumorigenesis in the liver. Genes & Development2010;24:1718–30.

[28] Georgiev P, Jochum W, Heinrich S, et al. Characterization of time-relatedchanges after experimental bile duct ligation. British Journal of Surgery2008;95:646–56.

[29] Apte U, Gkretsi V, Bowen WC, et al. Enhanced liver regeneration followingchanges induced by hepatocyte-specific genetic ablation of integrin-linkedkinase. Hepatology 2009;50:844–51.

[30] Wang C, Zhang L, He Q, et al. Differences in Yes-associated protein and mRNAlevels in regenerating liver and hepatocellular carcinoma. Molecular MedicineReports 2012;5:410–4.

[31] Wu H, Xiao Y, Zhang S, et al. The Ets transcription factor GABP is a compo-nent of the hippo pathway essential for growth and antioxidant defense. CellReports 2013;3:1663–77.

[32] Michalopoulos GK, Barua L, Bowen WC. Transdifferentiation of rat hepa-tocytes into biliary cells after bile duct ligation and toxic biliary injury.Hepatology 2005;41:535–44.

[33] Greenbaum LE. The ductal plate: a source of progenitors and hepatocytes inthe adult liver. Gastroenterology 2011;141:1152–5.

[34] Fitamant J, Kottakis F, Benhamouche S, et al. YAP inhibition restores hep-atocyte differentiation in advanced HCC, leading to tumor regression. CellReports 2015;10:1692–707.

[35] Cai J, Zhang N, Zheng Y, et al. The Hippo signaling pathway restricts the onco-genic potential of an intestinal regeneration program. Genes & Development2010;24:2383–8.

[36] Barry ER, Morikawa T, Butler BL, et al. Restriction of intestinal stem cell expan-sion and the regenerative response by YAP. Nature 2013;493:106–10.

[37] George NM, Day CE, Boerner BP, et al. Hippo signaling regulates pancreasdevelopment through inactivation of Yap. Molecular and Cellular Biology2012;32:5116–28.

[38] Gao T, Zhou D, Yang C, et al. Hippo signaling regulates differentiation andmaintenance in the exocrine pancreas. Gastroenterology 2013;144:1543–53,e1.

[39] Yu FX, Meng Z, Plouffe SW, et al. Hippo pathway regulation of gastrointestinaltissues. Annual Review of Physiology 2015;77:201–27.

[40] Zender L, Spector MS, Xue W, et al. Identification and validation of onco-genes in liver cancer using an integrative oncogenomic approach. Cell2006;125:1253–67.

[41] Imoto I, Yang ZQ, Pimkhaokham A, et al. Identification of cIAP1 as a candi-date target gene within an amplicon at 11q22 in esophageal squamous cellcarcinomas. Cancer Research 2001;61:6629–34.

[42] Dai Z, Zhu WG, Morrison CD, et al. A comprehensive search for DNA ampli-fication in lung cancer identifies inhibitors of apoptosis cIAP1 and cIAP2 ascandidate oncogenes. Human Molecular Genetics 2003;12:791–801.

[43] Bashyam MD, Bair R, Kim YH, et al. Array-based comparative genomichybridization identifies localized DNA amplifications and homozygous dele-tions in pancreatic cancer. Neoplasia 2005;7:556–62.

[44] Snijders AM, Schmidt BL, Fridlyand J, et al. Rare amplicons implicate fre-quent deregulation of cell fate specification pathways in oral squamous cellcarcinoma. Oncogene 2005;24:4232–42.

[45] Zhou D, Medoff BD, Chen L, et al. The Nore1B/Mst1 complex restrains antigenreceptor-induced proliferation of naive T cells. Proceedings of the NationalAcademy of Sciences of the United States of America 2008;105:20321–6.

[46] Dan I, Watanabe NM, Kusumi A. The Ste20 group kinases as regulators of MAPkinase cascades. Trends in Cell Biology 2001;11:220–30.

[47] Zhou D, Conrad C, Xia F, et al. Mst1 and Mst2 maintain hepatocyte quiescenceand suppress hepatocellular carcinoma development through inactivation ofthe Yap1 oncogene. Cancer Cell 2009;16:425–38.

[48] Song H, Mak KK, Topol L, et al. Mammalian Mst1 and Mst2 kinases playessential roles in organ size control and tumor suppression. Proceedingsof the National Academy of Sciences of the United States of America2010;107:1431–6.

[49] Lee KP, Lee JH, Kim TS, et al. The Hippo–Salvador pathway restrainshepatic oval cell proliferation, liver size, and liver tumorigenesis. Proceedingsof the National Academy of Sciences of the United States of America2010;107:8248–53.

[50] Lu L, Li Y, Kim SM, et al. Hippo signaling is a potent in vivo growth and tumor

suppressor pathway in the mammalian liver. Proceedings of the NationalAcademy of Sciences of the United States of America 2010;107:1437–42.

[51] Tao J, Calvisi DF, Ranganathan S, et al. Activation of beta-catenin and Yap1in human hepatoblastoma and induction of hepatocarcinogenesis in mice.Gastroenterology 2014;147:690–701.

r Disease 47 (2015) 826–835

[52] Kowalik MA, Saliba C, Pibiri M, et al. Yes-associated protein regulation of adap-tive liver enlargement and hepatocellular carcinoma development in mice.Hepatology 2011;53:2086–96.

[53] Perra A, Kowalik MA, Ghiso E, et al. YAP activation is an early event and apotential therapeutic target in liver cancer development. Journal of Hepatol-ogy 2014;61:1088–96.

[54] Zhao B, Wei X, Li W, et al. Inactivation of YAP oncoprotein by the Hippo path-way is involved in cell contact inhibition and tissue growth control. Genes &Development 2007;21:2747–61.

[55] Xu MZ, Yao TJ, Lee NP, et al. Yes-associated protein is an independent pro-gnostic marker in hepatocellular carcinoma. Cancer 2009;115:4576–85.

[56] Bai H, Gayyed MF, Lam-Himlin DM, et al. Expression of Yes-associated pro-tein modulates survivin expression in primary liver malignancies. HumanPathology 2012;43:1376–85.

[57] Li H, Wolfe A, Septer S, et al. Deregulation of Hippo kinase signalling in humanhepatic malignancies. Liver International: Official Journal of the InternationalAssociation for the Study of the Liver 2012;32:38–47.

[58] Zhang T, Zhang J, You X, et al. Hepatitis B virus X protein modulates onco-gene Yes-associated protein by CREB to promote growth of hepatoma cells.Hepatology 2012;56:2051–9.

[59] Ahn EY, Kim JS, Kim GJ, et al. RASSF1A-mediated regulation of AREG viathe Hippo pathway in hepatocellular carcinoma. Molecular Cancer Research:MCR 2013;11:748–58.

[60] Bai N, Zhang C, Liang N, et al. Yes-associated protein (YAP) increaseschemosensitivity of hepatocellular carcinoma cells by modulation of p53.Cancer Biology & Therapy 2013;14:511–20.

[61] Xu MZ, Chan SW, Liu AM, et al. AXL receptor kinase is a mediator ofYAP-dependent oncogenic functions in hepatocellular carcinoma. Oncogene2011;30:1229–40.

[62] Gurda GT, Zhu Q, Bai H, et al. The use of Yes-associated protein expression inthe diagnosis of persistent neonatal cholestatic liver disease. Human Pathol-ogy 2014;45:1057–64.

[63] Anakk S, Bhosale M, Schmidt VA, et al. Bile acids activate YAP to promote livercarcinogenesis. Cell Reports 2013;5:1060–9.

[64] Chen CL, Tsukamoto H, Liu JC, et al. Reciprocal regulation by TLR4 andTGF-beta in tumor-initiating stem-like cells. Journal of Clinical Investigation2013;123:2832–49.

[65] Wang J, Ma L, Weng W, et al. Mutual interaction between YAP and CREBpromotes tumorigenesis in liver cancer. Hepatology 2013;58:1011–20.

[66] Urtasun R, Latasa MU, Demartis MI, et al. Connective tissue growth factorautocriny in human hepatocellular carcinoma: oncogenic role and regula-tion by epidermal growth factor receptor/yes-associated protein-mediatedactivation. Hepatology 2011;54:2149–58.

[67] Hughes CM, Rozenblatt-Rosen O, Milne TA, et al. Menin associates with atrithorax family histone methyltransferase complex and with the hoxc8 locus.Molecular Cell 2004;13:587–97.

[68] Xu B, Li SH, Zheng R, et al. Menin promotes hepatocellular carcinogenesis andepigenetically up-regulates Yap1 transcription. Proceedings of the NationalAcademy of Sciences of the United States of America 2013;110:17480–5.

[69] Liu AM, Poon RT, Luk JM. MicroRNA-375 targets Hippo-signaling effector YAPin liver cancer and inhibits tumor properties. Biochemical and BiophysicalResearch Communications 2010;394:623–7.

[70] Kuroda S, Fukata M, Nakagawa M, et al. Role of IQGAP1, a target of the smallGTPases Cdc42 and Rac1, in regulation of E-cadherin-mediated cell-cell adhe-sion. Science 1998;281:832–5.

[71] Kim NG, Koh E, Chen X, et al. E-cadherin mediates contact inhibition ofproliferation through Hippo signaling-pathway components. Proceedingsof the National Academy of Sciences of the United States of America2011;108:11930–5.

[72] Dupont S, Morsut L, Aragona M, et al. Role of YAP/TAZ in mechanotransduc-tion. Nature 2011;474:179–83.

[73] Calvo F, Ege N, Grande-Garcia A, et al. Mechanotransduction and YAP-dependent matrix remodelling is required for the generation and mainte-nance of cancer-associated fibroblasts. Nature Cell Biology 2013;15:637–46.

[74] Zhao B, Li L, Wang L, et al. Cell detachment activates the Hippo pathwayvia cytoskeleton reorganization to induce anoikis. Genes & Development2012;26:54–68.

[75] Aragona M, Panciera T, Manfrin A, et al. A mechanical checkpoint controlsmulticellular growth through YAP/TAZ regulation by actin-processing factors.Cell 2013;154:1047–59.

[76] Fletcher GC, Elbediwy A, Khanal I, et al. The spectrin cytoskeleton regulatesthe Hippo signalling pathway. The EMBO Journal 2015;34:940–54.

[77] Deng H, Wang W, Yu J, et al. Spectrin regulates Hippo signaling by modulatingcortical actomyosin activity. Elife 2015;4:e06567.

[78] DeRan M, Yang J, Shen CH, et al. Energy stress regulates hippo-YAP signalinginvolving AMPK-mediated regulation of angiomotin-like 1 protein. CellReports 2014;9:495–503.

[79] Mo JS, Meng Z, Kim YC, et al. Cellular energy stress induces AMPK-mediated regulation of YAP and the Hippo pathway. Nature Cell Biology2015;17:500–10.

[80] Yu FX, Guan KL. The Hippo pathway: regulators and regulations. Genes &

Development 2013;27:355–71.

[81] DuPont CM, Coppola JJ, Kaercher RM, et al. Impaired trace fear condition-ing and diminished ERK1/2 phosphorylation in the dorsal hippocampusof adult rats administered alcohol as neonates. Behavioral Neuroscience2014;128:187–98.

d Live

a targeted inactivation of Wwtr1. Proceedings of the National Academy of

Q. Nguyen et al. / Digestive an

[82] Miller E, Yang J, DeRan M, et al. Identification of serum-derived sphingosine-1-phosphate as a small molecule regulator of YAP. Chemistry & Biology2012;19:955–62.

[83] Yu FX, Zhao B, Panupinthu N, et al. Regulation of the Hippo-YAP pathway byG-protein-coupled receptor signaling. Cell 2012;150:780–91.

[84] Zhao B, Li L, Tumaneng K, et al. A coordinated phosphorylation by Lats andCK1 regulates YAP stability through SCF(beta-TRCP). Genes & Development2010;24:72–85.

[85] Strano S, Munarriz E, Rossi M, et al. Physical interaction with Yes-associatedprotein enhances p73 transcriptional activity. Journal of Biological Chemistry2001;276:15164–73.

[86] Yagi R, Chen LF, Shigesada K, et al. A WW domain-containing yes-associatedprotein (YAP) is a novel transcriptional co-activator. The EMBO Journal1999;18:2551–62.

[87] Ota M, Sasaki H. Mammalian Tead proteins regulate cell proliferation and con-tact inhibition as transcriptional mediators of Hippo signaling. Development2008;135:4059–69.

[88] Zhao B, Ye X, Yu J, et al. TEAD mediates YAP-dependent gene induction andgrowth control. Genes & Development 2008;22:1962–71.

[89] Sawada A, Kiyonari H, Ukita K, et al. Redundant roles of Tead1 and Tead2 innotochord development and the regulation of cell proliferation and survival.Molecular and Cellular Biology 2008;28:3177–89.

[90] Zhang J, Ji JY, Yu M, et al. YAP-dependent induction of amphiregulin identifiesa non-cell-autonomous component of the Hippo pathway. Nature Cell Biology2009;11:1444–50.

[91] Shimomura T, Miyamura N, Hata S, et al. The PDZ-binding motif of Yes-associated protein is required for its co-activation of TEAD-mediated CTGFtranscription and oncogenic cell transforming activity. Biochemical and Bio-physical Research Communications 2014;443:917–23.

[92] Xiao W, Wang J, Ou C, et al. Mutual interaction between YAP and c-Myc is crit-ical for carcinogenesis in liver cancer. Biochemical and Biophysical ResearchCommunications 2013;439:167–72.

[93] Guo T, Lu Y, Li P, et al. A novel partner of Scalloped regulates Hipposignaling via antagonizing Scalloped-Yorkie activity. Cell Research 2013;23:1201–14.

[94] Jiao S, Wang H, Shi Z, et al. A peptide mimicking VGLL4 function actsas a YAP antagonist therapy against gastric cancer. Cancer Cell 2014;25:166–80.

[95] Serrano I, McDonald PC, Lock F, et al. Inactivation of the Hippo tumoursuppressor pathway by integrin-linked kinase. Nature Communications2013;4:2976.

[96] Wada K, Itoga K, Okano T, et al. Hippo pathway regulation by cell morphologyand stress fibers. Development 2011;138:3907–14.

r Disease 47 (2015) 826–835 835

[97] Camargo FD, Gokhale S, Johnnidis JB, et al. YAP1 increases organsize and expands undifferentiated progenitor cells. Current Biology: CB2007;17:2054–60.

[98] Saburi S, Hester I, Fischer E, et al. Loss of Fat4 disrupts PCP signaling andoriented cell division and leads to cystic kidney disease. Nature Genetics2008;40:1010–5.

[99] Bossuyt W, Chen CL, Chen Q, et al. An evolutionary shift in the regulation ofthe Hippo pathway between mice and flies. Oncogene 2014;33:1218–28.

[100] Tommasi S, Dammann R, Zhang Z, et al. Tumor susceptibility of Rassf1a knock-out mice. Cancer Research 2005;65:92–8.

[101] van der Weyden L, Tachibana KK, Gonzalez MA, et al. The RASSF1A isoformof RASSF1 promotes microtubule stability and suppresses tumorigenesis.Molecular and Cellular Biology 2005;25:8356–67.

[102] Park J, Kang SI, Lee SY, et al. Tumor suppressor ras association domain family 5(RASSF5/NORE1) mediates death receptor ligand-induced apoptosis. Journalof Biological Chemistry 2010;285:35029–38.

[103] McClatchey AI, Saotome I, Ramesh V, et al. The Nf2 tumor suppressor geneproduct is essential for extraembryonic development immediately prior togastrulation. Genes & Development 1997;11:1253–65.

[104] McClatchey AI, Saotome I, Mercer K, et al. Mice heterozygous for a mutation atthe Nf2 tumor suppressor locus develop a range of highly metastatic tumors.Genes & Development 1998;12:1121–33.

[105] Oh S, Lee D, Kim T, et al. Crucial role for Mst1 and Mst2 kinases inearly embryonic development of the mouse. Molecular and Cellular Biology2009;29:6309–20.

[106] Lee JH, Kim TS, Yang TH, et al. A crucial role of WW45 in developing epithelialtissues in the mouse. The EMBO Journal 2008;27:1231–42.

[107] St John MA, Tao W, Fei X, et al. Mice deficient of Lats1 develop soft-tissue sarcomas, ovarian tumours and pituitary dysfunction. Nature Genetics1999;21:182–6.

[108] McPherson JP, Tamblyn L, Elia A, et al. Lats2/Kpm is required for embryonicdevelopment, proliferation control and genomic integrity. The EMBO Journal2004;23:3677–88.

[109] Nishio M, Hamada K, Kawahara K, et al. Cancer susceptibility and embryoniclethality in Mob1a/1b double-mutant mice. Journal of Clinical Investigation2012;122:4505–18.

[110] Hossain Z, Ali SM, Ko HL, et al. Glomerulocystic kidney disease in mice with

Sciences of the United States of America 2007;104:1631–6.[111] Yi C, Shen Z, Stemmer-Rachamimov A, et al. The p130 isoform of angiomotin

is required for Yap-mediated hepatic epithelial cell proliferation and tumori-genesis. Science Signaling 2013;6:ra77.