Review

Crosstalk between TGF-b and hedgehog signaling in cancer

Delphine Javelaud, Marie-Jeanne Pierrat, Alain Mauviel !

Institut Curie, Centre de Recherche, 91400 Orsay, FranceINSERM U1021, 91400 Orsay, FranceCNRS UMR3347, 91400 Orsay, FranceUniversité Paris XI, 91400 Orsay, France

a r t i c l e i n f o

Article history:Received 23 April 2012Revised 3 May 2012Accepted 4 May 2012Available online 15 May 2012

Edited by Joan Massagué and Wilhelm Just

Keywords:Hedgehog signalingKruppel-like transcription factor GLI2CancerTGF-bFibrosisSignaling crosstalkEMT (epithelial to mesenchymal transition)

a b s t r a c t

Hedgehog (HH) and TGF-b signals control various aspects of embryonic development and cancer pro-gression. While their canonical signal transduction cascades have been well characterized, there isincreasing evidence that these pathways are able to exert overlapping activities that challenge effi-cient therapeutic targeting. We herein review the current knowledge on HH signaling and summa-rize the recent findings on the crosstalks between the HH and TGF-b pathways in cancer.! 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.

1. Canonical Hedgehog signaling: components andmechanistics

The Hedgehog (HH) pathway plays a critical role duringembryogenesis, in particular in directing limb digit and skeletalpolarity, and in the patterned formation of cells within the ventralportion of the central nervous system [1,2]. In adulthood, the SHHpathway is involved in the maintenance of tissue homeostasis andrepair after severe injury [3]. This pathway regulates numerousimportant cellular responses such as cell proliferation, survival, dif-ferentiation, self-renewal ability, migration and epithelial to mes-enchymal transition (EMT) [4]. Disruption of the HH pathway inthe mouse embryo result in congenital anomalies affecting thecentral nervous system, axial skeleton limbs and other organs,which are found in human pathology such as holoprosencephaly[5,6]. Conversely, aberrant activations of the HH pathway havebeen described in a wide variety of human cancers, including basalcell carcinoma and medulloblastoma [7–9].

1.1. Hedgehog ligands and receptors

In mammals, there are three secreted HH ligands: Sonic HH,Indian HH and Desert HH. They are cholesterol- and palmitoyl-modified proteins expressed by a wide range of cell types, and bindthe 12-transmembrane receptor PATCHED-1 (PTCH-1). Thefunctional specificity of HH proteins is governed in part by theirexpression patterns and by regulatory mechanisms in a given celltype [10–12]. For instance, several HH binding and/or sequestrat-ing proteins such as PTCH-2, CDO, BOC or HIP and GAS1 have beencharacterized. These proteins are considered co-receptors thatmodulate ligand presentation to PTCH-1, and positively ornegatively influence cellular responses to HH ligands [13].

The current model for HH signaling is as follows: in the absenceof HH ligand, PTCH-1 localizes to the primary cilium and inhibitsthe activity of the 7-transmembrane G-protein coupled receptor-like SMOOTHENED (SMO). The primary cilium is a microtubule-based antenna-like structure that emanates from the surface of awide variety of cells in mammals [14]. Signals from the primarycilium are ultimately involved in regulating the cell cycle, cytoskel-etal organization, intra-flagellar transport and various signalingpathways such as HH and WNT [15]. In the absence of HH ligand,only repressor forms of GLI proteins are translocated into thenucleus (Fig. 1). Binding of HH ligands to PTCH-1 results in its

0014-5793/$36.00 ! 2012 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.febslet.2012.05.011

! Corresponding author. Address: Institut Curie – Centre de Recherche, Team‘‘TGF-b and Oncogenesis’’, INSERM U1021/CNRS UMR3347, University Center,Building 110, 91400 Orsay, France.

E-mail address: alain.mauviel@curie.fr (A. Mauviel).

FEBS Letters 586 (2012) 2016–2025

journal homepage: www.FEBSLetters .org

exclusion from the primary cilium, release and activation of SMO,and signal transduction leading to activation and nuclear translo-cation of GLI transcription factors, and subsequent target genetransactivation.

1.2. Hedgehog effector molecules: GLI transcription factors

GLI proteins belong to the family of Kruppel-like factors, tran-scription factors with highly conserved C2H2-Zn finger DNA-bind-ing domains. There are three mammalian GLI proteins, GLI1, GLI2,and GLI3, each encoded by distinct genes. In contrast, their Dro-sophila homolog, Cubitus interruptus (Ci) is unique [4,16]. GLI pro-teins exhibit distinct regulations, biochemical properties andtarget genes. One study based on the overexpression of eitherGLI1 or a dominant-active mutant form of GLI2 in keratinocytesdetermined both overlapping and distinct transcriptional pro-grams for these two proteins [17]. For example, while PTCH-1and TNC are induced by both GLI1 and GLI2, VGLL4 is exclusivelyregulated by GLI1 while LITAF is a GLI2-specific target. Other stud-ies have identified OPN (osteopontin) [18] and MUC5A [19] as GLI1targets, while functional GLI2 binding sites have been character-ized on BCL2 [20] PTHrP [21] Follistatin [22] and BMP2 [23]promoters.

The intrinsic molecular nature of GLI proteins largely contrib-utes to their functional specificity. Processing of GLI2 and GLI3N-terminus represents a critical mechanism allowing regulationof target genes downstream of HH signaling. In the absence ofHH ligands, GLI3 is sequentially phosphorylated by PKA, GSK3and CK1 on multiple sites in its C-terminal region. This allowsthe recruitment of the F-box protein b-TrCP, an E3 ubiquitin ligase

that targets GLI3 to a limited proteolysis by the proteasome to gen-erate a repressor form [24] that does not utilize HDACs for tran-scriptional repression [25]. HH signaling inhibits GLI2 and GLI3processing, thereby elevating full-length GLI2/3 levels [26,27]. AsGLI3 exhibits only a weak transcriptional activity, it is mainly con-sidered an inhibitor of HH activity while GLI2 serves as the maintransducer of HH gene responses. GLI2 has a composite structure,comprising an activator domain in its C-terminus and a repressiondomain in its N-terminus, flanking a central five zinc finger DNA-binding domain [26]. Cleavage of the latter (GLI2DN) unmasksthe strong transactivating potential of GLI2. GLI2 is thought to bethe primary transcriptional activator downstream of HH signaling[28]. GLI1 lacks a repressor domain and is also a potent transcrip-tional activator. It is not processed proteolytically and is directlyregulated by HH signaling [29], contributing to the amplificationof HH responses. It is a direct transcriptional target of GLI2 [30]and is able to rescue some of GLI2 functions [31].

1.3. Suppressor of Fused, master inhibitor of HH signaling

When SMO is inactivated by PTCH-1, GLI2 and GLI3 interactwith the inhibitory protein SUFU (Suppressor of Fused), and re-main sequestered in the cytoplasm so that only the repressor formscan be translocated to the nucleus [32]. SUFU is able to regulate GLIlevels by antagonizing the ubiquitin ligase SPOP, which targets GLIproteins for complete degradation by the proteasome [33,34].SUFU may also repress GLI-dependent transcription by recruitingthe histone deacetylase complex SAP18–mSin3 at the nuclear level[35]. Negative regulations of GLI protein such as proteasomal deg-radation or sequestration by SUFU are relieved in presence of HH

Fig. 1. Canonical Hedgehog signaling. In the absence of ligand, the HH pathway is maintained inactive by several mechanisms. In the cilium, PTCH-1 blocks signaltransduction by SMO. Phosphorylation of GLI proteins by PKA, GSK3 or CK1 results in their proteasomal processing into transcription repressor forms through the recruitmentof b-TRCP and SPOP. DYRK2 and DYRK1B phosphorylate GLI proteins and promote repressor form processing. GLI proteins are sequestered in the cytoplasm by theirinteraction with 14-3-3 and SUFU. RAB23 and ULK3 promote GLI inhibition by SUFU. In the presence of HH, PTCH-1 is excluded from the cilium and SMO, through GaI,induces nuclear translocation of GLI proteins. HH-transduced signal is modulated positively or negatively by the interaction of PTCH-1 with co-receptors such as CDO, BOC,HIP1 and GAS1. In the nucleus, GLI proteins act as transcription factors. Phosphorylation by DYRK1A and ULK3 and acetylation by HDACs of GLI proteins increase theirtranscriptional activity. At the chromatin level, GLI activity may be inhibited by the recruitment c-SKI/SNON, SNF5 or SAP18/mSIN3-SUFU. (For more details, refer to text andreviews in [1,4,8,12,14–16])

D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025 2017

ligand. Free GLI proteins are translocated to the nucleus wherethey can act as transcription factors. Targeted disruption of SUFUin mice results in a phenotype with Gorlin syndrome features[36], a skin cancer condition associated with constitutive HH acti-vation due to LOF mutation of PTCH-1 [37]. Thus, SUFU appears tobe a tumor suppressor and a major negative regulator of HH signal-ing as it controls the production of GLI activators and repressorsessential for graded HH signaling, such is observed in the develop-ing embryo. Remarkably, the tumor-suppressor function of SUFUmay also be mediated by its ability to abrogate the WNT/b-cateninsignaling pathway via inhibition of b-catenin nuclear translocation.Simultaneous deregulation of the WNT and HH pathways, occur-ring in response to SUFU loss of function, was shown to be criticalfor tumor formation in a multistep model of carcinogenesis [38].Inactivating mutations of the SUFU gene, and loss of SUFU expres-sion, have been identified in prostate cancer and medulloblastoma[9,39]. Noteworthy, despite ciliary localization of most HH path-way components, SUFU regulates GLI protein levels and activityin a cilium-independent manner [33].

1.4. Post-transcriptional control of GLI protein function

Maintenance of GLI proteins in an inactive state and regulationof their transcriptional activity involves multiple, highly con-trolled, mechanisms. A number of proteins implicated in positiveand negative regulation of HH signaling outcome have been iden-tified, such as RAB23, 14-3-3, STK36, ULK3, SPOP and the DYRKfamily of proteins (see Fig. 1).

Inhibition of HH signaling by Rab23 was first described duringembryogenesis where they have opposing roles in neural pattern-ing [40]. Rab23 was recently shown to regulate GLI transcriptionalactivity via direct interaction with SUFU [41].

Asaoka and co-workers have shown that GLI proteins associatewith the 14-3-3 protein in a PKA dependent manner, and thisinteraction reduces their transcriptional activity [42].

STK36, the mammalian homolog of Drosophila Fu (Fused),antagonizes the repressive action of SUFU on HH signaling [43].However, contrary to Fu, STK36 has a limited role since its expres-sion is dispensable for proper mouse embryonic development [44].The serine/threonine kinase ULK3, which shares sequence similar-ity with STK36 and Fu, is able to phosphorylate GLI proteins,promoting their transcriptional activity [45]. However in theabsence of HH, ULK3 interacts with SUFU and promotes GLI2processing into a repressor form [45].

DYRK proteins are dual-specificity tyrosine-regulated kinasesthat have positive or negative effect on GLI transcriptional activitydepending on the family member. For instance, DYRK2 wasdescribed as a scaffold protein for an E3 ubiquitin ligase complexand phosphorylates GLI2 promoting its proteasomal degradation[46,47]. DYRK1B inhibits GLI2 transcriptional activity and pro-motes GLI3 processing into a repressor form [48]. On the otherhand, DYRK1A phosphorylates GLI1 that induces GLI1 nuclearlocalization and increases its transcriptional activity by phosphor-ylation [49].

SPOP, a substrate-binding adaptor for the cullin3-based ubiqui-tin E3 ligase, exerts competitive binding with SUFU to the N- andC-terminal regions of GLI3 or the C-terminal region of GLI2. SPOPfavors the cleavage of full-length GLI proteins into their repressorforms, thereby allowing SUFU-independent GLI3 repressorfunction [34].

The activity of the HH pathway may also be controlled at thechromatin level, by transcriptional co-activators and co-repressors.The activator form of GLI3 utilizes CBP as a co-activator of genetranscription, as in Drosophila, in which Ci uses dCBP [50,51]. c-SKI and its homolog SNO-N, both known transcriptional co-repres-sors of the TGF-b/SMAD pathway (see below), have been identified

in a yeast two-hybrid screen as interacting with both GLI2 andGLI3 [52]. When bound to GLI3, SKI recruits histone deacetylases(HDACs), thus contributing to GLI3-mediated transcriptionalrepression of target genes. The role of HDACs in HH signaling ishowever complex, as GLI1 and GLI2 are acetylated proteins, whosede-acetylation by HDACs promotes their transcriptional activity, aphenomenon that is further accentuated by a feed-forwardmechanism whereby HH induces HDAC1 expression. E3 ubiquitinligase-mediated degradation of HDAC1 suppresses HH signalingand HH-dependent growth, while GLI1 acetylation enhances cellproliferation and transformation [53]. Transcriptional activity ofGLI proteins may also be increased when they are SUMOylatedby the SUMO E3 ligase, PIAS1 [54].

Chromatin remodeling may also play a role in controlling HHsignaling outcome, either positively or negatively. For example,Brg1, an ATP-dependent chromatin remodeling factor, was shownto facilitate HDAC binding to HH target genes, enhancingGLI-dependent transcription. Yet, Brg1 may also repress HH signal-ing by allowing the recruitment of the repressor form of Gli3 toDNA, independently from its ATPase activity [55]. Another chroma-tin remodeling factor, SNF5, directly interacts with GLI1 on specificGLI-response elements to repress transcription of target genes [56].Inversely, loss of SNF5 was shown to enhance HH-dependent geneexpression and growth response.

Finally, expression of various actors of HH signaling can beregulated by miRNAs. For instance, in medulloblastoma, downreg-ulation of miR-125b, miR-326 targeting SMO and miR-324-5ptargeting GLI1 has been described [57]. Moreover, the miR-17–92cluster is overexpressed in human medulloblastoma. This up-regu-lation is controlled by N-Myc, a target of SHH pathway, and in-duces loss of PTCH-1 expression [58,59]. In non-small cell lungcancer, PTCH-1 expression may be downregulated by miR-212[60]. Also, miR-214, which is expressed during early segmentationstages of somites, is involved in the regulation of HH-regulatedgenes by controlling the expression of the negative regulator SUFU[61].

2. Autocrine and paracrine implication of HH in cancer

In human cancer, HH signaling may be constitutive due to ge-netic pathway activation via either loss-of-function mutations inPTCH-1 or SUFU [9,36,62], or gain-of-function mutations in SMO[63]. In addition, a number of cancers exhibit deregulated GLIexpression or up-regulated expression of HH ligands, without iden-tified mutations in upstream components of the pathway. Proto-typic cancers with mutations in HH pathway components aremedulloblastomas, rhabdomyosarcomas and basal cell carcinomas(BCC) [64]. Loss of function mutations of PTCH-1 have been identi-fied in Gorlin syndrome, also known as nevoid basal cell carcinomasyndrome, in which patients present several developmental disor-ders and a strong predisposition to BCC [64].

High GLI1 gene amplification has been reported in glioblastoma[65]. Also, several studies have reported that tumor cells in pancre-atic, lung, esophageal, gastric and prostate cancers secrete highamounts of HH ligand to induce tumor-promoting HH target genesin the adjacent stroma [66–70]. Yet tumor cells themselves weresometimes found not to be sensitive to HH [68]. Secreted HHligands may serve to stimulate the tumoral stroma to maintain amicroenvironment highly suitable for tumor growth [71]. Inparticular, HH ligands have been shown to stimulate angiogenesis[72], immune escape [73], stromal fibroblasts [74], as well asosteoblasts and osteoclasts whose activation promotes the estab-lishment and growth of osteolytic metastases from breast cancercells [75,76]. Inversely, stromal production of HH ligands promotestumor growth and dissemination of solid tumors such as BCC or

2018 D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025

medulloblastoma, acting on various cancer attributes, such as cellproliferation, resistance to apoptosis, increased genomic instabil-ity, as well as cell migration and invasiveness, and could also favorthe growth of hematologic malignancies such as B-cell lymphoma[69].

Experimental models mimicking aberrant HH signaling activa-tion have demonstrated the importance of this pathway for tumorinitiation and progression. For example, GLI1 and GLI2 overexpres-sion in keratinocytes of the basal epidermal layer in mice inducesmultiple skin tumors exhibiting features characteristic of humanBCC [77–80]. Conversely, GLI2 knockdown reduces colony forma-tion, anchorage-independent growth in vitro and tumor growthin vivo of prostatic tumor cells [81].

3. HH in cancer: a story of multiple crosstalks

HH signaling may however not be view as a lone effector duringtumorigenesis, as it cooperates with various other oncogenicevents for tumor formation and progression to metastasis. Theseinclude inactivation of tumor suppressors including p53 [82],NOTCH and PTEN [83,84], activated EGF receptor (EGFR) or onco-genic KRAS [85,86], as well as the WNT/b-catenin pathway [87],to cite a few.

Several lines of evidence point for a reciprocal suppressive rela-tionship between the p53 and HH pathways. p53 negatively regu-lates both the nuclear localization and the levels of GLI1 in neuralstem and tumor cells [82]. Conversely, activation of the HH path-way by overexpression of a constitutive active mutant of SMO,exogenous SHH ligand, or GLI1/GLI2 overexpression, inhibits theaccumulation of p53 protein through phosphorylation and activa-tion of the p53 inhibitor MDM2 [88]. Finally, cooperation of GLI2overexpression with p53 deficiency has been shown to contributeto the progression from benign to malignant cartilage tumors in amouse model [89].

NOTCH signaling is involved in an array of fundamental pro-cesses in both embryonic development and adult tissues [90].Depending on context, NOTCH may serve as a tumor suppressoror as an oncogenic pathway [91,92]. Remarkably, Notch1 deficiencyin murine epidermis results in a sustained expression of GLI2,which is involved in the development of BCC-like tumors [93].

Cooperation of the IGF/PI3K/AKT pathway with HH signaling ontumor development has been described in tumor mice models [94].Targeted expression of SHH and IGF2 to nestin-expressing neuralprogenitors in cerebella of newborn mice synergize for medullo-bastoma formation [94]. Moreover, PI3K/AKT activity is essentialfor SHH signaling in the specification of neuronal fates in chickenneural explants and for chondrogenic differentiation of 10T1/2cells [95]. Mechanistically, it was shown that PI3K/AKT activationinhibits GLI2 phosphorylation mediated by PKA in NIH3T3 cells,thus enhancing GLI2-dependent transcription. Inversely, PTEN, anegative regulator of the PI3K/AKT pathway often mutated in hu-man cancer, has been shown to negatively regulate transcriptionalactivity of GLI1 in glioblastoma cells [84]. Also, Stecca and co-workers have shown that RAS-MEK and AKT signaling regulatethe nuclear localization and transcriptional activity of GLI1 in mel-anoma and other cancer cells [96].

Oncogenic KRAS increases the stability of GLI1 protein and pro-motes its transcriptional activity in pancreatic cancer cells [86,97].Transgenic mouse models have shown that HH and KRAS cooper-ate at early stages of pancreatic carcinogenesis [98]. However, pan-creas cancer cells presenting KRAS mutation are insensitive to HH,yet they secrete high levels of HH ligand that induces an activationof stromal cells [68,99].

The HH and EGFR pathways cooperate to stimulate proliferationof neocortical stem cells [100,101] and invasive growth of

keratinocytes [102]. Furthermore, integration of HH and EGFRpathway signals synergistically promotes transformation and can-cer cell proliferation in vitro and in vivo, via mechanisms that mayor may not involve cooperation of GLI proteins with AP-1 tran-scription factors [16,103,104].

Several studies have described interactions between the HH andWNT/b-catenin pathways during dorso-ventral patterning of thenervous system. SHH through GLI3 controls WNT/b-catenin activa-tion and conversely WNT/b-catenin controls GLI2 and GLI3 expres-sion [105,106]. WNT protein expression can also be regulated byHH pathway [107]. Both pathways cooperate in the myogenicdetermination in the somite by inducing MYF5 expression: bindingof both a TCF/b-catenin complex and GLI1 to the enhancer of MYF5promoter cooperatively activates MYF5 expression [108]. WNT/b-catenin signaling may also regulate GLI2 and GLI3 activation inde-pendently on SHH [109]. Various human cancers exhibit a co-expression or a co-activation of HH and WNT/b-catenin signalingcomponents, such as in BCC or in pancreatic ductal adenocarci-noma [110,111]. Also, constitutive activation of HH signaling inmouse epidermis using a constitutively active mutant form ofSMO initiates tumor formation and induces activation of theWNT/b-catenin pathway, which is required for HH-driven tumori-genesis [110]. Activation of the WNT/b-catenin by HH pathway hasalso been reported in pancreatic adenocarcinoma cells [111].

Expression of SHH is induced by the oncogenic transcriptionfactor NF-jB in pancreatic cancer cells [112] and rhabdomyosar-coma cells [113]. In the latter case, NF-jB binding sites on SHHpromoter have been identified. Induction of SHH is involved inNF-jB-induced proliferation and resistance to TRAIL-inducedapoptosis.

4. HH-independent GLI activation

A few years ago, GLI proteins were only described as mediatorsof HH signaling. However, a growing body of evidence indicatesthat GLI expression and activation is regulated by growth factors/signaling cascades other than HH, and independently from HH.

One of the early evidences for HH ligand-independent GLI func-tion came from studies in the developing embryo, in which GLI-dependent transcription was observed in regions devoid of HH li-gands. GLI expression and activity was, for example, attributed toFGF signaling during antero-posterior embryonic patterning[114]. Recently, in a model aimed at understanding the mecha-nisms of immunoglobulin secretion by malignant cells, it wasfound that CCL5/RANTES stimulation of Waldenströmmacroglobu-linemia bone marrow stromal cells induces a rapid expression ofGLI2 via a CCR/PI3K-AKT/NF-jB mechanism. GLI2, in turn, inducesthe expression of IL-6, leading to secretion of IgM by malignant Bcells [115].

In Ewing sarcoma, GLI1 was found to be a direct transcriptionaltarget of EWS-FLI, the oncogenic transcription factor derived fromchromosomal translocation between chromosomes 11 and 22[116]. Expression of GLI1 was shown to promote the growth of Ew-ing Sarcoma tumors [117].

SMO-independent GLI activation by TNF-a was uncovered inesophageal cancer. Specifically, it was found that an activatedmTOR/S6K1 pathway downstream of TNF-a promotes GLI1 onco-genic and transcriptional activities through phosphorylation ofSer84. This, in turn, releases GLI1 from SUFU and allows for nucleartranslocation in an active form. No changes in total GLI1 levelswere observed [118]. Other signaling pathways capable of inducingthe expression of GLI proteins are the FGF, EGF, MAPK and TGF-bpathways, whose role in development and cancer progressionmay sometimes involve GLI-dependent mechanisms (reviewed in[11,119]).

D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025 2019

5. TGF-b: a potent inducer of GLI2 (and GLI1 as a consequence)

A few years ago, we identified GLI2, the most transcriptionallyactive of GLI proteins, as an early gene target of the TGF-b/SMADcascade independent of Hedgehog signaling [120]. The capacityof TGF-b to rapidly induce GLI2 expression, even in the presenceof cycloheximide, suggested that GLI2 is a direct target of SMADs,and did not require protein neosynthesis to occur. Subsequentcloning of the human GLI2 promoter confirmed this hypothesisand revealed that GLI2 transcription in response to TGF-b is med-iated by the rapid recruitment of both SMAD3 and b-catenin to dis-tinct elements of the GLI2 promoter in response to TGF-b [121].Induction of GLI2 was observed in a variety of human cell types,both normal and transformed, as well as in cell lines derived fromvarious cancers including breast and pancreatic carcinoma, glio-blastoma and cutaneous melanoma [120]. Remarkably, GLI2expression was found to be dramatically elevated in the epidermisof K5-TGF-b mice, which exhibit a hyperproliferative phenotyperesembling psoriasis [122]. Both high GLI2 expression and epider-mal hyperproliferation were lost upon crossing these mice on aSmad3 deficient background. SMAD3-dependency of GLI2 expres-sion downstream of TGF-b signaling was verified in vitro by meansof SMAD3 siRNA knockdown [120].

Interestingly, GLI1, whose expression is often considered a spe-cific marker of HH pathway activation in a plethora of publica-tions, is induced by TGF-b in a GLI2-dependent manner, andsuch induction by TGF-b is not sensitive to cyclopamine, a SMOantagonist, demonstrating a lack of requirement for HH signaling.In a study comparing the ability of cyclopamine (a HH inhibitorthat blocks SMO-dependent GLI1 expression) to inhibit thegrowth of, and induce apoptosis of, pancreatic cancer cells invitro, Hebrok and co-workers found that despite similarly ele-vated GLI1 levels, only about 50% of cell lines tested were sensi-tive to the drug [123]. Remarkably, we found that basal GLI2and GLI1 expression in these cyclopamine-resistant cell lines, aswell as their proliferative capacity, were greatly inhibited by anALK5 small molecule inhibitor and by transfected GLI2 siRNAoligonucleotides, while unaltered by cyclopamine [120]. Theseexperiments thus demonstrated that autocrine TGF-b signaling,not HH activity, was responsible for high GLI expression andresulting high proliferative capacity of these cyclopamine-resis-tant pancreatic carcinoma cell lines.

6. The TGF-b/GLI2 couple in cancer

During the course of our investigations, we compared GLI2expression in a series of human melanoma cell lines, in which con-stitutive TGF-b/SMAD signaling was previously characterized. Wefound that basal GLI2 expression in melanoma cells largely de-pends upon autocrine TGF-b signaling and that high GLI2 expres-sion is associated with loss of E-cadherin expression andincreased invasive capacities [124,125]. Loss of E-cadherin in mel-anoma is associated with the switch from radial to vertical growthphase of the primary tumor, leading to invasion of the dermis andfurther dissemination and metastasis. In epithelial cells, loss of E-cadherin is one of the key molecular events that takes place duringthe so-called epithelial to mesenchymal transition (EMT) a com-plex phenotypic conversion that involves changes in morphology,differentiation and cell–cell adhesion, and acquisition of a motilebehavior together with other mesenchymal traits. EMT is observedat various stages of embryonic development where intense tissueremodeling takes place and also represents a critical mechanismby which pre-malignant epithelial cells acquire a highly invasivephenotype that leads to metastatic spreading [126]. An EMT signa-ture is usually associated with poor prognosis in various cancers

[127]. Remarkably, the link between EMT and GLI-dependent generegulation was previously established [128], and TGF-b is aprototypic cytokine capable inducing this phenotypic switch[129], suggesting that GLI-dependent mechanisms may take placedownstream of TGF-b in the context of EMT.

GLI1 was shown previously to induce an EMT in rat kidney epi-thelial cells via induction of the E-cadherin repressor SNAIL [128].In melanoma, we found that high GLI2 expression is associated notonly with loss of E-cadherin, but also with increased cell invasive-ness in vitro and capacity to form bone metastases in mice [124].Mechanistically, we found that GLI2 knockdown in highly invasivemelanoma cells, i.e. strongly expressing GLI2, dramatically reducedtheir capacity to invade Matrigel and to form bone metastases innude mice, thus providing direct evidence for a driving role ofGLI2 in melanoma cell invasion, and for the relevance of GLI2 tar-geting to treat melanoma skeletal metastases. Remarkably, the effi-cacy of targeting TGF-b signaling to interfere with bone metastasisin the same mouse model was previously established [130–132].We identified PTHrP, IL-11, CXCR4 and OPN as TGF-b-regulatedgenes, whose expression was greatly reduced by both SMAD7 over-expression [131] and by systemic administration of the TbRI inhib-itor SD-208 [130]. These pro-metastatic genes are also critical forTGF-b-dependent establishment of bone metastases by MDA-MB-231 breast carcinoma cells in the same mouse model [133,134].Remarkably, overexpression of a dominant-negative form of GLI2targeting in the latter cells was shown to inhibit the establishmentand growth of bone metastases [135]. The authors confirmed ourinitial observations that GLI2 expression in MDA-MB-231 cells isinduced by TGF-b [120]. In both cases, the anti-metastatic potentialof GLI2 or TGF-b targeting was linked to inhibition of critical targetgenes including the osteoclastogenic factor PTHrP. Noteworthy,GLI1 has been shown to regulate OPN expression in melanomaand breast cancer cells, promoting malignant behavior, osteo-clastogenesis, and subsequent osteolysis [18,75]. A schematic ofTGF-b/HH interactions is provided in Fig. 2.

In a mouse model of breast cancer progression from ductalcarcinoma in situ (DCIS) to invasive carcinoma, others have foundthat TGF-b1 is responsible for both increased GLI2 expression andGLI-dependent transcription in MCFDCIS cells [136]. This deriva-tive of MCF10A breast carcinoma cells forms, upon implantationin mice, DCIS comedo-like tumors that spontaneously progress toinvasive tumors [137]. In their study, the authors found thatTGF-b-dependent GLI2 expression and GLI-dependent transcrip-tion, had a positive effect on myoepithelial cell differentiationand progression to invasion. In another study targeting BMP recep-tors specifically in the ovary, Bmpr1a Bmpr1b double-mutant micedeveloped granulosa cell tumors with enhanced constitutive TGF-bsignaling associated with elevated GLI2 (and GLI1) expression[138]. GLI2 was subsequently identified as a TGF-b regulated genein normal granulosa cells. Also, it was recently found that ovariancancer stem cells in recurrent tumors overexpress GLI1 and GLI2together with effectors of the TGF-b pathway including theco-receptor endoglin, as compared to the primary tumors [139].The authors found that knockdown of reduced tumor cell viabilityand resistance to cisplatin.

In a series of bladder cancer cell lines, it was found thatHH-independent GLI2 expression and function contributes to inva-siveness. The authors suggested that non-canonical GLI2 activitymay be attributed to RAS and TGF-b, both of these pathways beingcritical for bladder cancer development [140].

7. A HH–TGF-b amplification loop in cancer?

A number of situations display crosstalks between the TGF-b andHH pathways. For example, in gastric cancer cells, HH signaling

2020 D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025

induces the expression of TGF-b family members, and the latter arelikely to be involved in SMO-dependent tumorigenesis, asShh-induced cell motility and invasiveness requires TGF-b-drivenactivation of the SMAD pathway [141]. Also, expression of a consti-tutively active form of SMO under the control of the K14 promoterleads to the development of BCC in mice, accompanied with induc-tion of TGF-b2 expression and activation of the TGF-b signalingpathway. A small molecule TbRI inhibitor prevented tumor devel-opment, associated with increased lymphocyte infiltrates, suggest-ing that TGF-b is critical for SMO-mediated cancer developmentthrough its immunosuppressive activity [142].

While TGF-b likely contributes to some of HH biological effects,it is also highly likely that the opposite is true, as TGF-b-drivenEMT in non-small cell lung cancer cells was shown to be inhibitedby pharmacologic inhibitors of the HH pathway. The authorsshowed that TGF-b induced SHH expression in these cells, whichin turn activated GLI1 and GLI1-dependent EMT [143]. Inversely,we recently showed that TGF-b inhibits PKA activity while con-comitantly inducing GLI2 and GLI1 expression [144]. PKA blockademay contribute to increase the pool of full-length activator GLIproteins in the cell, thus allowing a possible HH response. Taken

together, these reports indicate that TGF-b and HH signaling mayform a vicious cycle promoting and amplifying the metastatic pro-cess whereby GLI2, and its downstream target GLI1, play a majorrole in allowing promoting tumor cell invasion and resistance toapoptosis.

Noteworthy, neuropilins may represent a critical node linkingHH and TGF-b signaling in cancer and fibrosis. Neuropilin-1(NRP1) has recently been shown to function as a co-receptor forTGF-b on the membrane of cancer cells, and to enhance thecanonical SMAD2/3 signaling in response to both latent and ac-tive TGF-b [145]. This activity of NRP1 is not limited to cancercells as it also enhances TGF-b-dependent regulatory T cell activ-ity [146] and promotes liver cirrhosis and associated vascularchanges by enhancing PDGF and TGF-b signaling in hepatic stel-late cells [147]. At the mean time, NRPs have been shown to pos-itively regulate HH signaling, mediating HH transduction betweenactivated SMO and SUFU, resulting in increased maximal HHtarget gene activation [148]. Furthermore, as NRP1 transcriptionitself is augmented by HH signaling [149], it suggests the exis-tence of a feed-forward mechanism of HH signal amplificationvia NRP.

Fig. 2. TGF-b and HH signaling crosstalks. TGF-b ligands induce cellular responses through membrane-bound, heteromeric serine/threonine kinase receptor complexes,which phosphorylate and activate SMAD proteins, whose subsequent accumulation in the nucleus allows target gene transactivation. SMAD3/SMAD4 complexes, incooperation with b-catenin, induces GLI2 expression [121]. The TGF-b pathway may contribute to GLI-specific responses by at least two mechanisms: TGF-b-induced GLI2may participate in the signal transduction activated by HH ligand and reinforce HH-inducible transcriptional response. TGF-b may inhibit PKA activity, thus limiting GLIprotein processing by the proteasome. TGF-b-induced GLI2 may also act independently on HH pathway and directly activate its transcriptional targets. GLI2-induced genesparticipate in EMT, tumor and bone metastasis development. The TGF-b and HH pathways may thus cooperate to promote cancer disease progression.

D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025 2021

8. Is tissue fibrosis a HH disease?

Remarkably, a few recent reports now implicate HH signalingin the development of tissue fibrosis affecting various organs,including skin, lung and liver. For example, it has been shown thatthe small molecule inhibitor of HH signaling LDE223 and SMOsiRNAs inhibit bleomycin-induced dermal fibrosis and exertsanti-fibrotic activities in tight-skin-1 mice [150]. Not only didthese approaches prevent progression of fibrosis, they alsoinduced regression of pre-established disease. In non-alcoholichepatic fibrosis, it has been shown that HH pathway activationparallels the severity of tissue injury [151], as estimated by thenumber of SHH and GLI2 positive cells that was found to increasewith disease stage, as well as expression of HH target genes. HHantagonists were shown to promote the regression of both liverfibrosis and hepatocellular carcinoma in mice [152]. Similarly, arole for HH signaling in the pathogenesis of renal fibrosis wasrecently described [153,154].

Fibrosis has long been considered a prototypic pathological con-dition of TGF-b-driven disease, mainly through its capacity to acton extracellular matrix deposition and remodeling [155–158].GLI2 is directly and transcriptionally induced by the TGF-b path-way in a wide variety of cell types and in various physio-patholog-ical situations [120,136–138]. Since GLI2 is the prime substrate forHH signaling [4], it becomes highly likely that TGF-b will be able toinduce HH responsiveness to cells that did not express GLI2 priorto TGF-b exposure. Interestingly, simultaneous expression ofTGF-b1 and SHH was detected by immunohistochemistry in epi-thelial cells from both human lung fibrosis (cryptogenic fibrosingalveolitis and bronchiectasis), and allergen-induced lung fibrosisin mice [159]. It cannot be excluded that TGF-b may contributeto the expression of GLI2 in injured tissue where it may serve asa substrate for strengthen HH signals in a self-amplifying patho-genic loop.

9. Concluding remarks

Targeting the HH pathway for cancer treatment by means ofSMO antagonists has shown remarkable efficacy in the pre-clinicaland clinical settings, against tumors (basal cell carcinoma (BCC) ofthe skin, pancreatic carcinoma) with identified mutations in theupstream components of the pathway [160]. Yet, a number of tu-mors exhibiting high expression of GLI1 are oblivious to HH signal-ing inhibition, suggesting the existence of alternate pathwaysleading to expression of downstream HHmediators. We and othershave contributed to identify TGF-b as a potent transcriptional reg-ulator of GLI2, resulting in GLI1 activation independently from theHH signaling cascade. In addition to facilitating direct GLI2-depen-dent oncogenic events, it is plausible that TGF-b (and other signalscapable of inducing GLI2 expression) may prime or potentiate HHresponsiveness by elevating the available pool of GLI2, a criticalcomponent of the HH response.

The vast majority of publications describing the roles of TGF-bor HH signaling in cancer (and fibrosis) examined these cytokinesindependently from each other. It will now be important to simul-taneously evaluate the respective contribution of both pathways indisease to characterize the proper target(s) for therapeutic inter-vention at a given stage of disease progression.

Acknowledgements

Supported by the Donation Henriette et Emile Goutière, InstitutNational du Cancer (INCa, PLBIO-2008), INSERM, CNRS, LigueNationale Contre le Cancer (Equipe Labellisée LIGUE), and Associa-tion pour la Recherche contre le Cancer (to D.J.). M.J. Pierrat

received a PhD scholarship from Canceropole Ile-de-France, com-plemented by Fondation pour la Recherche Médicale.

References

[1] Varjosalo, M. and Taipale, J. (2008) Hedgehog: functions and mechanisms.Genes Dev. 22, 2454–2472.

[2] McMahon, A.P., Ingham, P.W. and Tabin, C.J. (2003) Developmental roles andclinical significance of hedgehog signaling. Curr. Top. Dev. Biol. 53, 1–114.

[3] Lees, C., Howie, S., Sartor, R.B. and Satsangi, J. (2005) The hedgehog signallingpathway in the gastrointestinal tract: implications for development,homeostasis, and disease. Gastroenterology 129, 1696–1710.

[4] Hui, C.C. and Angers, S. (2011) Gli proteins in development and disease. Annu.Rev. Cell Dev. Biol. 27, 513–537.

[5] Monuki, E.S. (2007) The morphogen signaling network in forebraindevelopment and holoprosencephaly. J. Neuropathol. Exp. Neurol. 66, 566–575.

[6] Hayhurst, M. and McConnell, S.K. (2003) Mouse models ofholoprosencephaly. Curr. Opin. Neurol. 16, 135–141.

[7] Yang, L., Xie, G., Fan, Q. and Xie, J. (2010) Activation of the hedgehog-signalingpathway in human cancer and the clinical implications. Oncogene 29, 469–481.

[8] Marini, K.D., Payne, B.J., Watkins, D.N. and Martelotto, L.G. (2011)Mechanisms of Hedgehog signalling in cancer. Growth Factors 29, 221–234.

[9] Taylor, M.D., Liu, L., Raffel, C., Hui, C.C., Mainprize, T.G., Zhang, X., Agatep, R.,Chiappa, S., Gao, L., Lowrance, A., et al. (2002) Mutations in SUFU predisposeto medulloblastoma. Nat. Genet. 31, 306–310.

[10] Pathi, S., Pagan-Westphal, S., Baker, D.P., Garber, E.A., Rayhorn, P., Bumcrot,D., Tabin, C.J., Blake Pepinsky, R. and Williams, K.P. (2001) Comparativebiological responses to human Sonic, Indian, and Desert hedgehog. Mech.Dev. 106, 107–117.

[11] Jenkins, D. (2009) Hedgehog signalling: emerging evidence for non-canonicalpathways. Cell. Signal. 21, 1023–1034.

[12] Varjosalo, M. and Taipale, J. (2007) Hedgehog signaling. J. Cell Sci. 120, 3–6.[13] Allen, B.L., Tenzen, T. and McMahon, A.P. (2007) The Hedgehog-binding

proteins Gas1 and Cdo cooperate to positively regulate Shh signaling duringmouse development. Genes Dev. 21, 1244–1257.

[14] Drummond, I.A. (2012) Cilia functions in development. Curr. Opin. Cell Biol.24, 24–30.

[15] Berbari, N.F., O’Connor, A.K., Haycraft, C.J. and Yoder, B.K. (2009) The primarycilium as a complex signaling center. Curr. Biol. 19, R526–R535.

[16] Kasper, M., Regl, G., Frischauf, A.M. and Aberger, F. (2006) GLI transcriptionfactors: mediators of oncogenic Hedgehog signalling. Eur. J. Cancer 42, 437–445.

[17] Eichberger, T., Sander, V., Schnidar, H., Regl, G., Kasper, M., Schmid, C.,Plamberger, S., Kaser, A., Aberger, F. and Frischauf, A.M. (2006) Overlappingand distinct transcriptional regulator properties of the GLI1 and GLI2oncogenes. Genomics 87, 616–632.

[18] Das, S., Harris, L.G., Metge, B.J., Liu, S., Riker, A.I., Samant, R.S. and Shevde, L.A.(2009) The hedgehog pathway transcription factor GLI1 promotes malignantbehavior of cancer cells by up-regulating osteopontin. J. Biol. Chem. 284,22888–22897.

[19] Inaguma, S., Kasai, K. and Ikeda, H. (2011) GLI1 facilitates the migration andinvasion of pancreatic cancer cells through MUC5AC-mediated attenuation ofE-cadherin. Oncogene 30, 714–723.

[20] Regl, G., Kasper, M., Schnidar, H., Eichberger, T., Neill, G.W., Philpott, M.P.,Esterbauer, H., Hauser-Kronberger, C., Frischauf, A.M. and Aberger, F. (2004)Activation of the BCL2 promoter in response to Hedgehog/GLI signaltransduction is predominantly mediated by GLI2. Cancer Res. 64, 7724–7731.

[21] Sterling, J.A., Oyajobi, B.O., Grubbs, B., Padalecki, S.S., Munoz, S.A., Gupta, A.,Story, B., Zhao, M. and Mundy, G.R. (2006) The hedgehog signaling moleculeGli2 induces parathyroid hormone-related peptide expression and osteolysisin metastatic human breast cancer cells. Cancer Res. 66, 7548–7553.

[22] Eichberger, T., Kaser, A., Pixner, C., Schmid, C., Klingler, S., Winklmayr, M.,Hauser-Kronberger, C., Aberger, F. and Frischauf, A.M. (2008) GLI2-specifictranscriptional activation of the bone morphogenetic protein/activinantagonist follistatin in human epidermal cells. J. Biol. Chem. 283, 12426–12437.

[23] Zhao, M., Qiao, M., Harris, S.E., Chen, D., Oyajobi, B.O. and Mundy, G.R. (2006)The zinc finger transcription factor Gli2 mediates bone morphogeneticprotein 2 expression in osteoblasts in response to hedgehog signaling. Mol.Cell. Biol. 26, 6197–6208.

[24] Bhatia, N., Thiyagarajan, S., Elcheva, I., Saleem, M., Dlugosz, A., Mukhtar, H.and Spiegelman, V.S. (2006) GLi2 is targeted for ubiquitination anddegradation by beta-TrCP ubiquitin ligase. J. Biol. Chem. 281, 19320–19326.

[25] Tsanev, R., Tiigimagi, P., Michelson, P., Metsis, M., Osterlund, T. andKogerman, P. (2009) Identification of the gene transcription repressordomain of Gli3. FEBS Lett. 583, 224–228.

[26] Sasaki, H., Nishizaki, Y., Hui, C., Nakafuku, M. and Kondoh, H. (1999)Regulation of Gli2 and Gli3 activities by an amino-terminal repressiondomain: implication of Gli2 and Gli3 as primary mediators of Shh signaling.Development 126, 3915–3924.

[27] Pan, Y., Bai, C.B., Joyner, A.L. and Wang, B. (2006) Sonic hedgehog signalingregulates Gli2 transcriptional activity by suppressing its processing anddegradation. Mol. Cell. Biol. 26, 3365–3377.

2022 D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025

[28] Mill, P., Mo, R., Fu, H., Grachtchouk, M., Kim, P.C., Dlugosz, A.A. and Hui, C.C.(2003) Sonic hedgehog-dependent activation of Gli2 is essential forembryonic hair follicle development. Genes Dev. 17, 282–294.

[29] Dahmane, N., Lee, J., Robins, P., Heller, P. and Ruiz i Altaba, A. (1997)Activation of the transcription factor Gli1 and the Sonic hedgehog signallingpathway in skin tumours. Nature 389, 876–881.

[30] Ikram, M.S., Neill, G.W., Regl, G., Eichberger, T., Frischauf, A.M., Aberger, F.,Quinn, A. and Philpott, M. (2004) GLI2 is expressed in normal humanepidermis and BCC and induces GLI1 expression by binding to its promoter. J.Invest. Dermatol. 122, 1503–1509.

[31] Bai, C.B. and Joyner, A.L. (2001) Gli1 can rescue the in vivo function of Gli2.Development 128, 5161–5172.

[32] Cheng, S.Y. and Yue, S. (2008) Role and regulation of human tumorsuppressor SUFU in Hedgehog signaling. Adv. Cancer Res. 101, 29–43.

[33] Chen, M.H., Wilson, C.W., Li, Y.J., Law, K.K., Lu, C.S., Gacayan, R., Zhang, X., Hui,C.C. and Chuang, P.T. (2009) Cilium-independent regulation of Gli proteinfunction by Sufu in Hedgehog signaling is evolutionarily conserved. GenesDev. 23, 1910–1928.

[34] Wang, C., Pan, Y. and Wang, B. (2010) Suppressor of fused and Spop regulatethe stability, processing and function of Gli2 and Gli3 full-length activatorsbut not their repressors. Development 137, 2001–2009.

[35] Cheng, S.Y. and Bishop, J.M. (2002) Suppressor of Fused represses Gli-mediated transcription by recruiting the SAP18-mSin3 corepressor complex.Proc. Natl. Acad. Sci. USA 99, 5442–5447.

[36] Svard, J., Heby-Henricson, K., Persson-Lek, M., Rozell, B., Lauth, M., Bergstrom,A., Ericson, J., Toftgard, R. and Teglund, S. (2006) Genetic elimination ofsuppressor of fused reveals an essential repressor function in the mammalianHedgehog signaling pathway. Dev. Cell 10, 187–197.

[37] Hahn, H., Wicking, C., Zaphiropoulous, P.G., Gailani, M.R., Shanley, S.,Chidambaram, A., Vorechovsky, I., Holmberg, E., Unden, A.B., Gillies, S.,et al. (1996) Mutations of the human homolog of Drosophila patched in thenevoid basal cell carcinoma syndrome. Cell 85, 841–851.

[38] Taylor, M.D., Zhang, X., Liu, L., Hui, C.C., Mainprize, T.G., Scherer, S.W.,Wainwright, B., Hogg, D. and Rutka, J.T. (2004) Failure of a medulloblastoma-derived mutant of SUFU to suppress WNT signaling. Oncogene 23, 4577–4583.

[39] Sheng, T., Li, C., Zhang, X., Chi, S., He, N., Chen, K., McCormick, F., Gatalica, Z.and Xie, J. (2004) Activation of the hedgehog pathway in advanced prostatecancer. Mol. Cancer 3, 29.

[40] Eggenschwiler, J.T., Espinoza, E. and Anderson, K.V. (2001) Rab23 is anessential negative regulator of the mouse Sonic hedgehog signallingpathway. Nature 412, 194–198.

[41] Chi, S., Xie, G., Liu, H., Chen, K., Zhang, X., Li, C. and Xie, J. (2012) Rab23negatively regulates Gli1 transcriptional factor in a Su(Fu)-dependentmanner. Cell. Signal. 24, 1222–1228.

[42] Asaoka, Y., Kanai, F., Ichimura, T., Tateishi, K., Tanaka, Y., Ohta, M., Seto, M.,Tada, M., Ijichi, H., Ikenoue, T., et al. (2010) Identification of a suppressivemechanism for Hedgehog signaling through a novel interaction of Gli with14-3-3. J. Biol. Chem. 285, 4185–4194.

[43] Murone, M., Luoh, S.M., Stone, D., Li, W., Gurney, A., Armanini, M., Grey, C.,Rosenthal, A. and de Sauvage, F.J. (2000) Gli regulation by the opposingactivities of fused and suppressor of fused. Nat. Cell Biol. 2, 310–312.

[44] Merchant, M., Evangelista, M., Luoh, S.M., Frantz, G.D., Chalasani, S., Carano,R.A., van Hoy, M., Ramirez, J., Ogasawara, A.K., McFarland, L.M., et al. (2005)Loss of the serine/threonine kinase fused results in postnatal growth defectsand lethality due to progressive hydrocephalus. Mol. Cell. Biol. 25, 7054–7068.

[45] Maloverjan, A., Piirsoo, M., Kasak, L., Peil, L., Osterlund, T. and Kogerman, P.(2010) Dual function of UNC-51-like kinase 3 (Ulk3) in the Sonic hedgehogsignaling pathway. J. Biol. Chem. 285, 30079–30090.

[46] Varjosalo, M., Bjorklund, M., Cheng, F., Syvanen, H., Kivioja, T., Kilpinen, S.,Sun, Z., Kallioniemi, O., Stunnenberg, H.G., He, W.W., et al. (2008) Applicationof active and kinase-deficient kinome collection for identification of kinasesregulating hedgehog signaling. Cell 133, 537–548.

[47] Maddika, S. and Chen, J. (2009) Protein kinase DYRK2 is a scaffold thatfacilitates assembly of an E3 ligase. Nat. Cell Biol. 11, 409–419.

[48] Lauth, M., Bergstrom, A., Shimokawa, T., Tostar, U., Jin, Q., Fendrich, V.,Guerra, C., Barbacid, M. and Toftgard, R. (2010) DYRK1B-dependentautocrine-to-paracrine shift of Hedgehog signaling by mutant RAS. Nat.Struct. Mol. Biol. 17, 718–725.

[49] Mao, J., Maye, P., Kogerman, P., Tejedor, F.J., Toftgard, R., Xie, W., Wu, G. andWu, D. (2002) Regulation of Gli1 transcriptional activity in the nucleus byDyrk1. J. Biol. Chem. 277, 35156–35161.

[50] Dai, P., Akimaru, H., Tanaka, Y., Maekawa, T., Nakafuku, M. and Ishii, S. (1999)Sonic Hedgehog-induced activation of the Gli1 promoter is mediated by GLI3.J. Biol. Chem. 274, 8143–8152.

[51] Akimaru, H., Chen, Y., Dai, P., Hou, D.X., Nonaka, M., Smolik, S.M., Armstrong,S., Goodman, R.H. and Ishii, S. (1997) Drosophila CBP is a co-activator ofcubitus interruptus in hedgehog signalling. Nature 386, 735–738.

[52] Dai, P., Shinagawa, T., Nomura, T., Harada, J., Kaul, S.C., Wadhwa, R., Khan,M.M., Akimaru, H., Sasaki, H., Colmenares, C., et al. (2002) Ski is involved intranscriptional regulation by the repressor and full-length forms of Gli3.Genes Dev. 16, 2843–2848.

[53] Canettieri, G., Di Marcotullio, L., Greco, A., Coni, S., Antonucci, L., Infante, P.,Pietrosanti, L., De Smaele, E., Ferretti, E., Miele, E., et al. (2010) Histonedeacetylase and Cullin3-REN(KCTD11) ubiquitin ligase interplay

regulates Hedgehog signalling through Gli acetylation. Nat. Cell Biol. 12,132–142.

[54] Cox, B., Briscoe, J. and Ulloa, F. (2010) SUMOylation by Pias1 regulates theactivity of the Hedgehog dependent Gli transcription factors. PLoS ONE 5,e11996.

[55] Zhan, X., Shi, X., Zhang, Z., Chen, Y. and Wu, J.I. (2011) Dual role of Brgchromatin remodeling factor in Sonic hedgehog signaling during neuraldevelopment. Proc. Natl. Acad. Sci. USA 108, 12758–12763.

[56] Jagani, Z., Mora-Blanco, E.L., Sansam, C.G., McKenna, E.S., Wilson, B., Chen, D.,Klekota, J., Tamayo, P., Nguyen, P.T., Tolstorukov, M., et al. (2010) Loss of thetumor suppressor Snf5 leads to aberrant activation of the Hedgehog–Glipathway. Nat. Med. 16, 1429–1433.

[57] Ferretti, E., De Smaele, E., Miele, E., Laneve, P., Po, A., Pelloni, M., Paganelli, A.,Di Marcotullio, L., Caffarelli, E., Screpanti, I., et al. (2008) Concerted microRNAcontrol of Hedgehog signalling in cerebellar neuronal progenitor and tumourcells. EMBO J. 27, 2616–2627.

[58] Northcott, P.A., Fernandez, L.A., Hagan, J.P., Ellison, D.W., Grajkowska, W.,Gillespie, Y., Grundy, R., Van Meter, T., Rutka, J.T., Croce, C.M., et al. (2009) ThemiR-17/92 polycistron is up-regulated in sonic hedgehog-drivenmedulloblastomas and induced by N-myc in sonic hedgehog-treatedcerebellar neural precursors. Cancer Res. 69, 3249–3255.

[59] Uziel, T., Karginov, F.V., Xie, S., Parker, J.S., Wang, Y.D., Gajjar, A., He, L., Ellison,D., Gilbertson, R.J., Hannon, G., et al. (2009) The miR-17!92 clustercollaborates with the Sonic Hedgehog pathway in medulloblastoma. Proc.Natl. Acad. Sci. USA 106, 2812–2817.

[60] Li, Y., Zhang, D., Chen, C., Ruan, Z. and Huang, Y. (2012) MicroRNA-212displays tumor-promoting properties in non-small cell lung cancer cells andtargets the hedgehog pathway receptor PTCH1. Mol. Biol. Cell 23,1423–1434.

[61] Flynt, A.S., Li, N., Thatcher, E.J., Solnica-Krezel, L. and Patton, J.G. (2007)Zebrafish miR-214 modulates Hedgehog signaling to specify muscle cell fate.Nat. Genet. 39, 259–263.

[62] Xie, J., Johnson, R.L., Zhang, X., Bare, J.W., Waldman, F.M., Cogen, P.H., Menon,A.G., Warren, R.S., Chen, L.C., Scott, M.P., et al. (1997) Mutations of thePATCHED gene in several types of sporadic extracutaneous tumors. CancerRes. 57, 2369–2372.

[63] Xie, J., Murone, M., Luoh, S.M., Ryan, A., Gu, Q., Zhang, C., Bonifas, J.M., Lam,C.W., Hynes, M., Goddard, A., et al. (1998) Activating smoothened mutationsin sporadic basal-cell carcinoma. Nature 391, 90–92.

[64] Epstein, E.H. (2008) Basal cell carcinomas: attack of the hedgehog. Nat. Rev.Cancer 8, 743–754.

[65] Kinzler, K.W., Bigner, S.H., Bigner, D.D., Trent, J.M., Law, M.L., O’Brien, S.J.,Wong, A.J. and Vogelstein, B. (1987) Identification of an amplified, highlyexpressed gene in a human glioma. Science 236, 70–73.

[66] Domenech, M., Bjerregaard, R., Bushman, W. and Beebe, D.J. (2012) Hedgehogsignaling in myofibroblasts directly promotes prostate tumor cell growth.Integr. Biol. (Camb.) 4, 142–152.

[67] Yauch, R.L., Gould, S.E., Scales, S.J., Tang, T., Tian, H., Ahn, C.P., Marshall, D., Fu,L., Januario, T., Kallop, D., et al. (2008) A paracrine requirement for hedgehogsignalling in cancer. Nature 455, 406–410.

[68] Tian, H., Callahan, C.A., DuPree, K.J., Darbonne, W.C., Ahn, C.P., Scales, S.J. andde Sauvage, F.J. (2009) Hedgehog signaling is restricted to the stromalcompartment during pancreatic carcinogenesis. Proc. Natl. Acad. Sci. USA106, 4254–4259.

[69] Lindemann, R.K. (2008) Stroma-initiated hedgehog signaling takes centerstage in B-cell lymphoma. Cancer Res. 68, 961–964.

[70] Ma, X., Chen, K., Huang, S., Zhang, X., Adegboyega, P.A., Evers, B.M., Zhang, H.and Xie, J. (2005) Frequent activation of the hedgehog pathway in advancedgastric adenocarcinomas. Carcinogenesis 26, 1698–1705.

[71] Harris, L.G., Samant, R.S. and Shevde, L.A. (2011) Hedgehog signaling:networking to nurture a promalignant tumor microenvironment. Mol.Cancer Res. 9, 1165–1174.

[72] Geng, L., Cuneo, K.C., Cooper, M.K., Wang, H., Sekhar, K., Fu, A. and Hallahan,D.E. (2007) Hedgehog signaling in the murine melanoma microenvironment.Angiogenesis 10, 259–267.

[73] Rowbotham, N.J., Hager-Theodorides, A.L., Furmanski, A.L. and Crompton, T.(2007) A novel role for Hedgehog in T-cell receptor signaling: implications fordevelopment and immunity. Cell Cycle 6, 2138–2142.

[74] Shaw, A., Gipp, J. and Bushman, W. (2009) The Sonic Hedgehog pathwaystimulates prostate tumor growth by paracrine signaling and recapitulatesembryonic gene expression in tumor myofibroblasts. Oncogene 28, 4480–4490.

[75] Das, S., Samant, R.S. and Shevde, L.A. (2011) Hedgehog signaling induced bybreast cancer cells promotes osteoclastogenesis and osteolysis. J. Biol. Chem.286, 9612–9622.

[76] Das, S., Tucker, J.A., Khullar, S., Samant, R.S. and Shevde, L.A. (2012) Hedgehogsignaling in tumor cells facilitates osteoblast-enhanced osteolyticmetastases. PLoS ONE 7, e34374, http://dx.doi.org/10.1371/journal.pone.0034374.

[77] Nilsson, M., Unden, A.B., Krause, D., Malmqwist, U., Raza, K., Zaphiropoulos,P.G. and Toftgard, R. (2000) Induction of basal cell carcinomas andtrichoepitheliomas in mice overexpressing GLI-1. Proc. Natl. Acad. Sci. USA97, 3438–3443.

[78] Grachtchouk, M., Mo, R., Yu, S., Zhang, X., Sasaki, H., Hui, C.C. and Dlugosz,A.A. (2000) Basal cell carcinomas in mice overexpressing Gli2 in skin. Nat.Genet. 24, 216–217.

D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025 2023

[79] Hutchin, M.E., Kariapper, M.S., Grachtchouk, M., Wang, A., Wei, L., Cummings,D., Liu, J., Michael, L.E., Glick, A. and Dlugosz, A.A. (2005) Sustained Hedgehogsignaling is required for basal cell carcinoma proliferation and survival:conditional skin tumorigenesis recapitulates the hair growth cycle. GenesDev. 19, 214–223.

[80] Sheng, H., Goich, S., Wang, A., Grachtchouk, M., Lowe, L., Mo, R., Lin, K., deSauvage, F.J., Sasaki, H., Hui, C.C., et al. (2002) Dissecting the oncogenicpotential of Gli2: deletion of an NH(2)-terminal fragment alters skin tumorphenotype. Cancer Res. 62, 5308–5316.

[81] Thiyagarajan, S., Bhatia, N., Reagan-Shaw, S., Cozma, D., Thomas-Tikhonenko,A., Ahmad, N. and Spiegelman, V.S. (2007) Role of GLI2 transcription factor ingrowth and tumorigenicity of prostate cells. Cancer Res. 67, 10642–10646.

[82] Stecca, B. and Ruiz i Altaba, A. (2009) A GLI1-p53 inhibitory loop controlsneural stem cell and tumour cell numbers. EMBO J. 28, 663–676.

[83] Castellino, R.C., Barwick, B.G., Schniederjan, M., Buss, M.C., Becher, O.,Hambardzumyan, D., Macdonald, T.J., Brat, D.J. and Durden, D.L. (2010)Heterozygosity for Pten promotes tumorigenesis in a mouse model ofmedulloblastoma. PLoS ONE 5, e10849.

[84] Xu, Q., Yuan, X., Liu, G., Black, K.L. and Yu, J.S. (2008) Hedgehog signalingregulates brain tumor-initiating cell proliferation and portends shortersurvival for patients with PTEN-coexpressing glioblastomas. Stem Cells 26,3018–3026.

[85] Lauth, M. and Toftgard, R. (2011) Hedgehog signaling and pancreatic tumordevelopment. Adv. Cancer Res. 110, 1–17.

[86] Ji, Z., Mei, F.C., Xie, J. and Cheng, X. (2007) Oncogenic KRAS activateshedgehog signaling pathway in pancreatic cancer cells. J. Biol. Chem. 282,14048–14055.

[87] Takebe, N., Harris, P.J., Warren, R.Q. and Ivy, S.P. (2011) Targeting cancer stemcells by inhibiting Wnt, Notch, and Hedgehog pathways. Nat. Rev. Clin. Oncol.8, 97–106.

[88] Abe, Y., Oda-Sato, E., Tobiume, K., Kawauchi, K., Taya, Y., Okamoto, K., Oren,M. and Tanaka, N. (2008) Hedgehog signaling overrides p53-mediated tumorsuppression by activating Mdm2. Proc. Natl. Acad. Sci. USA 105, 4838–4843.

[89] Ho, L., Stojanovski, A., Whetstone, H., Wei, Q.X., Mau, E., Wunder, J.S. andAlman, B. (2009) Gli2 and p53 cooperate to regulate IGFBP-3- mediatedchondrocyte apoptosis in the progression from benign to malignant cartilagetumors. Cancer Cell 16, 126–136.

[90] Artavanis-Tsakonas, S. and Muskavitch, M.A. (2010) Notch: the past, thepresent, and the future. Curr. Top. Dev. Biol. 92, 1–29.

[91] Koch, U. and Radtke, F. (2007) Notch and cancer: a double-edged sword. Cell.Mol. Life Sci. 64, 2746–2762.

[92] Lobry, C., Oh, P. and Aifantis, I. (2011) Oncogenic and tumor suppressorfunctions of Notch in cancer: it’s NOTCH what you think. J. Exp. Med. 208,1931–1935.

[93] Nicolas, M., Wolfer, A., Raj, K., Kummer, J.A., Mill, P., van Noort, M., Hui, C.C.,Clevers, H., Dotto, G.P. and Radtke, F. (2003) Notch1 functions as a tumorsuppressor in mouse skin. Nat. Genet. 33, 416–421.

[94] Rao, G., Pedone, C.A., Del Valle, L., Reiss, K., Holland, E.C. and Fults, D.W.(2004) Sonic hedgehog and insulin-like growth factor signaling synergize toinduce medulloblastoma formation from nestin-expressing neuralprogenitors in mice. Oncogene 23, 6156–6162.

[95] Riobo, N.A., Lu, K., Ai, X., Haines, G.M. and Emerson Jr., C.P. (2006)Phosphoinositide 3-kinase and Akt are essential for Sonic Hedgehogsignaling. Proc. Natl. Acad. Sci. USA 103, 4505–4510.

[96] Stecca, B., Mas, C., Clement, V., Zbinden, M., Correa, R., Piguet, V., Beermann, F.and Ruiz, I.A.A. (2007) Melanomas require HEDGEHOG–GLI signalingregulated by interactions between GLI1 and the RAS-MEK/AKT pathways.Proc. Natl. Acad. Sci. USA 104, 5895–5900.

[97] Nolan-Stevaux, O., Lau, J., Truitt, M.L., Chu, G.C., Hebrok, M., Fernandez-Zapico, M.E. and Hanahan, D. (2009) GLI1 is regulated through smoothened-independent mechanisms in neoplastic pancreatic ducts and mediates PDACcell survival and transformation. Genes Dev. 23, 24–36.

[98] Pasca di Magliano, M., Sekine, S., Ermilov, A., Ferris, J., Dlugosz, A.A. andHebrok, M. (2006) Hedgehog/Ras interactions regulate early stages ofpancreatic cancer. Genes Dev. 20, 3161–3173.

[99] Lauth, M., Bergstrom, A., Shimokawa, T. and Toftgard, R. (2007) Inhibition ofGLI-mediated transcription and tumor cell growth by small-moleculeantagonists. Proc. Natl. Acad. Sci. USA 104, 8455–8460.

[100] Palma, V., Lim, D.A., Dahmane, N., Sanchez, P., Brionne, T.C., Herzberg, C.D.,Gitton, Y., Carleton, A., Alvarez-Buylla, A. and Ruiz i Altaba, A. (2005) Sonichedgehog controls stem cell behavior in the postnatal and adult brain.Development 132, 335–344.

[101] Palma, V. and Ruiz i Altaba, A. (2004) Hedgehog-GLI signaling regulates thebehavior of cells with stem cell properties in the developing neocortex.Development 131, 337–345.

[102] Bigelow, R.L., Jen, E.Y., Delehedde, M., Chari, N.S. and McDonnell, T.J. (2005)Sonic hedgehog induces epidermal growth factor dependent matrixinfiltration in HaCaT keratinocytes. J. Invest. Dermatol. 124, 457–465.

[103] Schnidar, H., Eberl, M., Klingler, S., Mangelberger, D., Kasper, M., Hauser-Kronberger, C., Regl, G., Kroismayr, R., Moriggl, R., Sibilia, M., et al. (2009)Epidermal growth factor receptor signaling synergizes with Hedgehog/GLI inoncogenic transformation via activation of the MEK/ERK/JUN pathway.Cancer Res. 69, 1284–1292.

[104] Eberl, M., Klingler, S., Mangelberger, D., Loipetzberger, A., Damhofer, H., Zoidl,K., Schnidar, H., Hache, H., Bauer, H.C., Solca, F., et al. (2012) Hedgehog-EGFRcooperation response genes determine the oncogenic phenotype of basal cell

carcinoma and tumour-initiating pancreatic cancer cells. EMBO Mol. Med. 4,218–233.

[105] Alvarez-Medina, R., Cayuso, J., Okubo, T., Takada, S. and Marti, E. (2008) Wntcanonical pathway restricts graded Shh/Gli patterning activity through theregulation of Gli3 expression. Development 135, 237–247.

[106] Ulloa, F., Itasaki, N. and Briscoe, J. (2007) Inhibitory Gli3 activity negativelyregulates Wnt/beta-catenin signaling. Curr. Biol. 17, 545–550.

[107] Mullor, J.L., Dahmane, N., Sun, T. and Ruiz i Altaba, A. (2001) Wnt signals aretargets and mediators of Gli function. Curr. Biol. 11, 769–773.

[108] Borello, U., Berarducci, B., Murphy, P., Bajard, L., Buffa, V., Piccolo, S.,Buckingham, M. and Cossu, G. (2006) The Wnt/beta-catenin pathwayregulates Gli-mediated Myf5 expression during somitogenesis.Development 133, 3723–3732.

[109] Borycki, A., Brown, A.M. and Emerson Jr., C.P. (2000) Shh and Wnt signalingpathways converge to control Gli gene activation in avian somites.Development 127, 2075–2087.

[110] Yang, S.H., Andl, T., Grachtchouk, V., Wang, A., Liu, J., Syu, L.J., Ferris, J., Wang,T.S., Glick, A.B., Millar, S.E., et al. (2008) Pathological responses to oncogenicHedgehog signaling in skin are dependent on canonical Wnt/beta3-cateninsignaling. Nat. Genet. 40, 1130–1135.

[111] Pasca di Magliano, M., Biankin, A.V., Heiser, P.W., Cano, D.A., Gutierrez, P.J.,Deramaudt, T., Segara, D., Dawson, A.C., Kench, J.G., Henshall, S.M., et al.(2007) Common activation of canonical Wnt signaling in pancreaticadenocarcinoma. PLoS ONE 2, e1155.

[112] Nakashima, H., Nakamura, M., Yamaguchi, H., Yamanaka, N., Akiyoshi, T.,Koga, K., Yamaguchi, K., Tsuneyoshi, M., Tanaka, M. and Katano, M. (2006)Nuclear factor-kappaB contributes to hedgehog signaling pathway activationthrough sonic hedgehog induction in pancreatic cancer. Cancer Res. 66,7041–7049.

[113] Kasperczyk, H., Baumann, B., Debatin, K.M. and Fulda, S. (2009)Characterization of sonic hedgehog as a novel NF-kappaB target gene thatpromotes NF-kappaB-mediated apoptosis resistance and tumor growthin vivo. FASEB J. 23, 21–33.

[114] Brewster, R., Mullor, J.L. and Ruiz i Altaba, A. (2000) Gli2 functions in FGFsignaling during antero-posterior patterning. Development 127, 4395–4405.

[115] Elsawa, S.F., Almada, L.L., Ziesmer, S.C., Novak, A.J., Witzig, T.E., Ansell, S.M.and Fernandez-Zapico, M.E. (2011) GLI2 transcription factor mediatescytokine cross-talk in the tumor microenvironment. J. Biol. Chem. 286,21524–21534.

[116] Beauchamp, E., Bulut, G., Abaan, O., Chen, K., Merchant, A., Matsui, W., Endo,Y., Rubin, J.S., Toretsky, J. and Uren, A. (2009) GLI1 is a direct transcriptionaltarget of EWS-FLI1 oncoprotein. J. Biol. Chem. 284, 9074–9082.

[117] Zwerner, J.P., Joo, J., Warner, K.L., Christensen, L., Hu-Lieskovan, S., Triche, T.J.and May, W.A. (2008) The EWS/FLI1 oncogenic transcription factorderegulates GLI1. Oncogene 27, 3282–3291.

[118] Wang, Y., Ding, Q., Yen, C.J., Xia, W., Izzo, J.G., Lang, J.Y., Li, C.W., Hsu, J.L.,Miller, S.A., Wang, X., et al. (2012) The crosstalk of mTOR/S6K1 and hedgehogpathways. Cancer Cell 21, 374–387.

[119] Lauth, M. and Toftgard, R. (2007) Non-canonical activation of GLItranscription factors: implications for targeted anti-cancer therapy. CellCycle 6, 2458–2463.

[120] Dennler, S., Andre, J., Alexaki, I., Li, A., Magnaldo, T., ten Dijke, P., Wang, X.J.,Verrecchia, F. and Mauviel, A. (2007) Induction of sonic hedgehog mediatorsby transforming growth factor-beta: Smad3-dependent activation of Gli2and Gli1 expression in vitro and in vivo. Cancer Res. 67, 6981–6986.

[121] Dennler, S., Andre, J., Verrecchia, F. and Mauviel, A. (2009) Cloning of thehuman GLI2 Promoter: transcriptional activation by transforming growthfactor-beta via SMAD3/beta-catenin cooperation. J. Biol. Chem. 284, 31523–31531.

[122] Li, A.G., Wang, D., Feng, X.H. and Wang, X.J. (2004) Latent TGFbeta1overexpression in keratinocytes results in a severe psoriasis-like skindisorder. EMBO J. 23, 1770–1781.

[123] Thayer, S.P., di Magliano, M.P., Heiser, P.W., Nielsen, C.M., Roberts, D.J.,Lauwers, G.Y., Qi, Y.P., Gysin, S., Fernandez-del Castillo, C., Yajnik, V., et al.(2003) Hedgehog is an early and late mediator of pancreatic cancertumorigenesis. Nature 425, 851–856.

[124] Alexaki, V.I., Javelaud, D., Van Kempen, L.C., Mohammad, K.S., Dennler, S.,Luciani, F., Hoek, K.S., Juarez, P., Goydos, J.S., Fournier, P.J., et al. (2010) GLI2-mediated melanoma invasion and metastasis. J. Natl. Cancer Inst. 102, 1148–1159.

[125] Javelaud, D., Alexaki, V.I., Pierrat, M.J., Hoek, K.S., Dennler, S., Van Kempen, L.,Bertolotto, C., Ballotti, R., Saule, S., Delmas, V., et al. (2011) GLI2 and M-MITFtranscription factors control exclusive gene expression programs andinversely regulate invasion in human melanoma cells. Pigment CellMelanoma Res. 24, 932–943.

[126] Guarino, M., Rubino, B. and Ballabio, G. (2007) The role of epithelial–mesenchymal transition in cancer pathology. Pathology 39, 305–318.

[127] Yang, J. and Weinberg, R.A. (2008) Epithelial–mesenchymal transition: atthe crossroads of development and tumor metastasis. Dev. Cell 14, 818–829.

[128] Li, X., Deng, W., Nail, C.D., Bailey, S.K., Kraus, M.H., Ruppert, J.M. and Lobo-Ruppert, S.M. (2006) Snail induction is an early response to Gli1 thatdetermines the efficiency of epithelial transformation. Oncogene 25, 609–621.

[129] Zavadil, J. and Bottinger, E.P. (2005) TGF-beta and epithelial-to-mesenchymaltransitions. Oncogene 24, 5764–5774.

2024 D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025

[130] Mohammad, K.S., Javelaud, D., Fournier, P.G., Niewolna, M., McKenna, C.R.,Peng, X.H., Duong, V., Dunn, L.K., Mauviel, A. and Guise, T.A. (2011) TGF-{beta}-RI kinase inhibitor SD-208 reduces the development and progressionof melanoma bone metastases. Cancer Res. 71, 175–184.

[131] Javelaud, D., Mohammad, K.S., McKenna, C.R., Fournier, P., Luciani, F.,Niewolna, M., Andre, J., Delmas, V., Larue, L., Guise, T.A., et al. (2007) Stableoverexpression of Smad7 in human melanoma cells impairs bone metastasis.Cancer Res. 67, 2317–2324.

[132] Javelaud, D., Delmas, V., Moller, M., Sextius, P., Andre, J., Menashi, S., Larue, L.and Mauviel, A. (2005) Stable overexpression of Smad7 in human melanomacells inhibits their tumorigenicity in vitro and in vivo. Oncogene 24, 7624–7629.

[133] Kang, Y., He, W., Tulley, S., Gupta, G.P., Serganova, I., Chen, C.R., Manova-Todorova, K., Blasberg, R., Gerald, W.L. and Massague, J. (2005) Breast cancerbone metastasis mediated by the Smad tumor suppressor pathway. Proc.Natl. Acad. Sci. USA 102, 13909–13914.

[134] Kang, Y., Siegel, P.M., Shu, W., Drobnjak, M., Kakonen, S.M., Cordon-Cardo, C.,Guise, T.A. and Massague, J. (2003) A multigenic program mediating breastcancer metastasis to bone. Cancer Cell 3, 537–549.

[135] Johnson, R.W., Nguyen, M.P., Padalecki, S.S., Grubbs, B.G., Merkel, A.R.,Oyajobi, B.O., Matrisian, L.M., Mundy, G.R. and Sterling, J.A. (2011) TGF-{beta}promotion of Gli2 induced PTHrP expression is independent of canonicalHedgehog signaling. Cancer Res. 71, 822–831.

[136] Hu, M., Yao, J., Carroll, D.K., Weremowicz, S., Chen, H., Carrasco, D.,Richardson, A., Violette, S., Nikolskaya, T., Nikolsky, Y., et al. (2008)Regulation of in situ to invasive breast carcinoma transition. Cancer Cell13, 394–406.

[137] Miller, F.R., Santner, S.J., Tait, L. and Dawson, P.J. (2000) MCF10DCIS.comxenograft model of human comedo ductal carcinoma in situ. J. Natl. CancerInst. 92, 1185–1186.

[138] Edson, M.A., Nalam, R.L., Clementi, C., Franco, H.L., Demayo, F.J., Lyons,K.M., Pangas, S.A. and Matzuk, M.M. (2010) Granulosa cell-expressedBMPR1A and BMPR1B have unique functions in regulating fertility butact redundantly to suppress ovarian tumor development. Mol.Endocrinol. 24, 1251–1266.

[139] Steg, A.D., Bevis, K.S., Katre, A.A., Ziebarth, A., Dobbin, Z.C., Alvarez, R.D.,Zhang, K., Conner, M. and Landen, C.N. (2012) Stem cell pathways contributeto clinical chemoresistance in ovarian cancer. Clin. Cancer Res. 18, 869–881.

[140] Mechlin, C.W., Tanner, M.J., Chen, M., Buttyan, R., Levin, R.M. and Mian, B.M.(2010) Gli2 expression and human bladder transitional carcinoma cellinvasiveness. J. Urol. 184, 344–351.

[141] Yoo, Y.A., Kang, M.H., Kim, J.S. and Oh, S.C. (2008) Sonic hedgehog signalingpromotes motility and invasiveness of gastric cancer cells through TGF-beta-mediated activation of the ALK5–Smad 3 pathway. Carcinogenesis 29, 480–490.

[142] Fan, Q., He, M., Sheng, T., Zhang, X., Sinha, M., Luxon, B., Zhao, X. and Xie, J.(2010) Requirement of TGFbeta signaling for SMO-mediated carcinogenesis.J. Biol. Chem. 285, 36570–36576.

[143] Maitah, M.Y., Ali, S., Ahmad, A., Gadgeel, S. and Sarkar, F.H. (2011) Up-regulation of sonic hedgehog contributes to TGF-beta1-induced epithelial tomesenchymal transition in NSCLC cells. PLoS ONE 6, e16068.

[144] Pierrat, M.J., Marsaud, V., Mauviel, A. and Javelaud, D. (2012) Expression ofMicrophtalmia-Associated Transcription Factor (MITF), which is critical formelanoma progression, is inhibited by both transcription factor GLI2 andtransforming growth factor-beta. J. Biol. Chem., http://dx.doi.org/10.1074/jbc.M112.358341.

[145] Glinka, Y., Stoilova, S., Mohammed, N. and Prud’homme, G.J. (2011)Neuropilin-1 exerts co-receptor function for TGF-beta-1 on the membraneof cancer cells and enhances responses to both latent and active TGF-beta.Carcinogenesis 32, 613–621.

[146] Glinka, Y. and Prud’homme, G.J. (2008) Neuropilin-1 is a receptor fortransforming growth factor beta-1, activates its latent form, and promotesregulatory T cell activity. J. Leukoc. Biol. 84, 302–310.

[147] Cao, Y., Wang, L., Nandy, D., Zhang, Y., Basu, A., Radisky, D. andMukhopadhyay, D. (2008) Neuropilin-1 upholds dedifferentiation andpropagation phenotypes of renal cell carcinoma cells by activating AKT andsonic hedgehog axes. Cancer Res. 68, 8667–8672.

[148] Hillman, R.T., Feng, B.Y., Ni, J., Woo, W.M., Milenkovic, L., Hayden Gephart,M.G., Teruel, M.N., Oro, A.E., Chen, J.K. and Scott, M.P. (2011) Neuropilins arepositive regulators of Hedgehog signal transduction. Genes Dev. 25, 2333–2346.

[149] Hochman, E., Castiel, A., Jacob-Hirsch, J., Amariglio, N. and Izraeli, S. (2006)Molecular pathways regulating pro-migratory effects of Hedgehog signaling.J. Biol. Chem. 281, 33860–33870.

[150] Horn, A., Kireva, T., Palumbo-Zerr, K., Dees, C., Tomcik, M., Cordazzo, C., Zerr,P., Akhmetshina, A., Ruat, M., Distler, O., et al. (2012) Inhibition of hedgehogsignalling prevents experimental fibrosis and induces regression ofestablished fibrosis. Ann. Rheum. Dis. 71, 785–789.

[151] Guy, C.D., Suzuki, A., Zdanowicz, M., Abdelmalek, M.F., Burchette, J., Unalp, A.and Diehl, A.M. (2011) Hedgehog pathway activation parallels histologicseverity of injury and fibrosis in human nonalcoholic fatty liver disease.Hepatology, http://dx.doi.org/10.1002/hep. 25559.

[152] Philips, G.M., Chan, I.S., Swiderska, M., Schroder, V.T., Guy, C., Karaca, G.F.,Moylan, C., Venkatraman, T., Feuerlein, S., Syn, W.K., et al. (2011) Hedgehogsignaling antagonist promotes regression of both liver fibrosis andhepatocellular carcinoma in a murine model of primary liver cancer. PLoSONE 6, e23943.

[153] Fabian, S.L., Penchev, R.R., St-Jacques, B., Rao, A.N., Sipila, P., West, K.A.,McMahon, A.P. and Humphreys, B.D. (2012) Hedgehog-Gli PathwayActivation during Kidney Fibrosis. Am. J. Pathol. 180, 1441–1453.

[154] Ding, H., Zhou, D., Hao, S., Zhou, L., He, W., Nie, J., Hou, F.F. and Liu, Y. (2012)Sonic hedgehog signaling mediates epithelial–mesenchymal communicationand promotes renal fibrosis. J. Am. Soc. Nephrol., http://dx.doi.org/10.1681/ASN.2011060614.

[155] Mauviel, A. (2005) Transforming growth factor-beta: a key mediator offibrosis. Methods Mol. Med. 117, 69–80.

[156] Schiller, M., Javelaud, D. and Mauviel, A. (2004) TGF-beta-induced SMADsignaling and gene regulation: consequences for extracellular matrixremodeling and wound healing. J. Dermatol. Sci. 35, 83–92.

[157] Verrecchia, F. and Mauviel, A. (2002) Transforming growth factor-betasignaling through the Smad pathway: role in extracellular matrix geneexpression and regulation. J. Invest. Dermatol. 118, 211–215.

[158] Verrecchia, F., Mauviel, A. and Farge, D. (2006) Transforming growth factor-beta signaling through the Smad proteins: role in systemic sclerosis.Autoimmun. Rev. 5, 563–569.

[159] Stewart, G.A., Hoyne, G.F., Ahmad, S.A., Jarman, E., Wallace, W.A., Harrison,D.J., Haslett, C., Lamb, J.R. and Howie, S.E. (2003) Expression of thedevelopmental Sonic hedgehog (Shh) signalling pathway is up-regulated inchronic lung fibrosis and the Shh receptor patched 1 is present in circulatingT lymphocytes. J. Pathol. 199, 488–495.

[160] Rubin, L.L. and de Sauvage, F.J. (2006) Targeting the Hedgehog pathway incancer. Nat. Rev. Drug Discov. 5, 1026–1033.

D. Javelaud et al. / FEBS Letters 586 (2012) 2016–2025 2025