Human Homolog of Drosophila Expanded, Hex, Functions As a Putative Tumor Suppressor in Human Cancer Cell Lines Independently of the Hippo Pathway

Oncogene (2012) 31, 1189–1195 & 2012 Macmillan Publishers Limited All rights reserved 0950-9232/12 www.nature.com/onc SHORT COMMUNICATION Human homolog of Drosophila expanded, hEx, functions as a putative tumor suppressor in human cancer cell lines independently of the Hippo pathway

S Visser-Grieve, Y Hao and X Yang

Department of Pathology and Molecular Medicine, Queen’s University, Kingston, Ontario, Canada

The Hippo signaling network is proving to be an essential Keywords: expanded; FRMD6; Willin; Hippo; LATS1; regulator within the cell, participating in multiple cellular tumor suppressor phenotypes including cell proliferation, apoptosis, cell migration and organ size control. Much of this pathway is conserved from flies to mammals; however, how the upstream components, namely Expanded, affect downstream processes in mammalian systems has remained Introduction elusive. Only recently has human Expanded (hEx), also known as FRMD6 or Willin, been identified. However, A mutated expanded (ex) gene was originally identified its functional significance with respect to its putative in 1926 and shown to enhance the growth of imaginal tumor suppressor function and activation of the Hippo discs in Drosophila, suggesting that ex functions as a pathway has not been studied. In this study, we show for tumor suppressor (Stern and Bridges, 1926). Since then, the first time that hEx possesses several tumor suppressor ex has been definitely classified as a tumor suppressor properties. First, hEx dramatically inhibits cell prolifera- due to its abilities to restrict growth, differentiation tion in two human cancer cell lines, MDA-MB-231 and and enhance apoptosis in the wings and eyes of flies MDA-MB-436 cells, and sensitizes these cells to the (Boedigheimer and Laughon, 1993; Blaumueller and chemotherapeutic drug Taxol. Furthermore, downregula- Mlodzik, 2000; McCartney et al., 2000). Due to the tion of hEx in the immortalized MCF10A breast cell line similar overgrowth phenotypes seen in the imaginal leads to enhanced proliferation and resistance to Taxol discs of flies exhibiting mutations in the core Hippo treatment. As evidence for its tumor suppressor function, pathway components, ex was added as a key player in overexpression of hEx inhibits colony formation, soft this pathway (Hamaratoglu et al., 2006). agar colony growth in vitro and in vivo tumor growth in The evolutionarily conserved Hippo signaling path- nude mice. Although Drosophila expanded (ex) can way is an area of intense research due to its fundamental activate the Hippo pathway, surprisingly no significant roles in organ size control and tumorigenesis (Zhao alterations were discovered in the phosphorylation status et al., 2010). In Drosophila, the Ser/Thr kinases Hippo of any of the Hippo pathway components, including (hpo) and dlats along with adapter proteins sav and downstream tumor suppressor LATS1, upon overexpres- mats make up the core components. In this system, sion of hEx. In addition, knockdown of both LATS1 active Hpo phosphorylates and activates dlats, enhanced and LATS2 in hEx-overexpressing cells was unable to by interacting sav and mats. The primary target is Yki, rescue the hEx phenotype, suggesting that hEx functions a transcriptional co-activator and oncogene whose independently of the Hippo pathway in this cell line. Alter- overexpression phenocopies loss of dlats. Several down- natively, we propose a mechanism through which hEx stream targets of Yki have been described, including inhibits progression through the S phase of the cell cycle bantam, cyclin E and diap1. Activation of this pathway by upregulating p21Cip1 and downregulating Cyclin A. This involves the WW and C2-domain containing protein is the first study to functionally characterize hEx and Kibra and two FERM domain proteins ex and mer, show that hEx acts in a distinct manner compared with which all localize to the apical membrane and coopera- Drosophila expanded. tively promote dlats phosphorylation and activation. Oncogene (2012) 31, 1189–1195; doi:10.1038/onc.2011.318; Connecting these membrane-associated proteins with published online 25 July 2011 external growth cues is the atypical cadherin fat or the apical transmembrane protein crumbs, both of which activate ex (reviewed in Pan, 2010 and Halder and Johnson, 2011). Correspondence: Dr X Yang, Department of Pathology and Molecular Importantly, most of this pathway is conserved in Medicine, Queen’s University, Richardson Labs 201, 88 Stuart Street, mammalian systems. Hpo homologs MST1 and MST2 Kingston, Ontario, Canada K7L 3N6. phosphorylate and activate LATS1 and LATS2 (Chan E-mail: [email protected] Received 14 April 2011; revised and accepted 21 June 2011; published et al., 2005). Once activated LATS1/LATS2 bind and online 25 July 2011 phosphorylate the Yki homolog YAP or its paralog hEx functions as a tumor suppressor in human cancer cell lines S Visser-Grieve et al 1190 TAZ, inhibiting their transcriptional activity (Hao et al., 28 WPI

) hEx 2008; Lei et al., 2008). Numerous transcriptional targets 4 24 hEx-shEx for LATS1 and LATS2 (Visser and Yang, 2010a) as well 20 as YAP (Ota and Sasaki, 2008; Zhao et al., 2008) and 16 TAZ (Lai et al., 2011) have been identiﬁed with several WPI hEx hEx-shEx 12 hEx (α-FLG) proven to mediate key Hippo pathway functions. 8

Cell number (x10 4 Although this core kinase cascade is well conserved, β-actin 0 upstream regulation of the mammalian Hippo pathway 0 152 3 4 is less understood. Homologs exist for each membrane- Days associated protein: Merlin/NF2 for mer, FRMD6/ 15 Willin/Human Expanded (hEx) for ex, KIBRA for WPI

) hEx Kibra as well as FAT4 for fat and Crumbs1–3 for 4 12 hEx-shEx Crumbs. Although Merlin/NF2 is a well-characterized tumor suppressor and has been shown to enhance 9 LATS1/2 phosphorylation (Lau et al., 2008; Yu et al., WPI hEx hEx-shEx 6 hEx (α-FLG) 2010) and human KIBRA has also been found to bind 3 LATS1 and LATS2 enhancing their phosphorylation Cell number (x10 β -actin 0 and kinase activity (Yu et al., 2010; Xiao et al., 2011), 0 1 2345 thereby functioning in parallel to their Drosophila Days homologs, little is known about hEx in human systems. hEx is a FERM domain protein similar to Merlin pLKO.1 1 60 shEx-1 ) and the ERM (Ezrin, Radixin, Moesin) family of cyto- 4 50 shEx-2 0.8 skeletal crosslinkers and is localized through- 40 out the cytoplasm or along the plasma membrane 0.6 30 (Gunn-Moore et al., 2005). In this study, we outline 0.4 20 the key cellular functions of hEx and suggest that 10 Relative mRNA 0.2 Cell number (x10 hEx possesses tumor suppressor properties. By analyz- 0 ing individual components of the Hippo pathway, 0 0 12345 pLKO.1 shEx-1 shEx-2 we provide the first evidence that hEx does not activate Days or function through this pathway, suggesting that the Figure 1 hEx is a regulator of cell proliferation. (a) Western blot upstream regulation of the Hippo pathway in MDA- analysis of hEx expression levels in MDA-MB-231 cells. Cells were MB-231 cells is not conserved. This study is the first infected with lentivirus expressing vector alone (WPI), hEx-FLG to characterize hEx and provides novel insights into (hEx) or hEx-FLG followed by the subsequent infection with its functions. lentivirus expressing shRNA against hEx (hEx-shEx). Cell lysate extracted from established cell lines was subjected to western blot using anti-FLG M2 monoclonal antibody (Sigma, Oakville, ON, Canada) and b-actin (Sigma) as internal loading control. (b) Cell proliferation analysis. Cells were plated in triplicate with equal cell Results and discussion number into 24-well plates and cell numbers were counted on days 1, 2, 3, 4 and 5 days postplating. Experiments were performed at least three times. Data are mean±s.d. (c) Western blot analysis of To examine the cellular functions of hEx in breast hEx expression in MDA-MB-436 cells. Cell lines establishment and cancer cell lines, we established hEx-expressing MDA- western blot analysis were as described above. (d) Cell proliferation MB-231 cell lines using lentivirus expressing vector analysis was as described for MDA-MB-231 cells. (e) Quantitative alone (WPI) or FLAG-tagged hEx (Figure 1a). MDA- RT–PCR analysis of hEx mRNA levels in MCF10A cells. Cells were infected with lentivirus expressing vector alone (pLKO.1) or MB-231 is an aggressive breast cancer cell line but when expressing one of the two shRNAs targeting hEx (shEx-1 and hEx is overexpressed their growth is severely inhibited shEx-2) (targeting sequences: shEx-1, 50-GATGAAGTTCCAGA (Figure 1b). Importantly, this phenotype is specific to GTTTGTT-30; shEx-2, 50-CCACAGACTATATGTCGGAAA-30). expression of hEx and not due to any non-specific virus mRNA was extracted and qRT–PCR was performed as previously described (Visser and Yang, 2010a) using rRNA as internal effects since knockdown of hEx in the hEx-over- control. (f) Cell proliferation analysis was performed as described expressing cells (hEx-shEx) using lentivirus expressing above. shRNA targeting hEx completely reverses the growth inhibitory phenotype. To show the effect of hEx on cell proliferation is not cell line specific, we repeated these Traditionally, tumor suppressors are defined by their experiments using another aggressive breast cancer cell loss of function in cancer. To recapitulate this in our cell line MDA-MB-436 (Figure 1c). As with the MDA-MB- system we downregulated endogenous hEx in the 231 cells, expression of hEx dramatically inhibits MDA- MCF10A immortalized breast cell line using lentivirus MB-436 cell proliferation whereas the hEx-shEx cell line expressing vector alone (pLKO.1) or one of the two proliferates at a similar rate compared with the WPI shRNAs targeting hEx for degradation (shEx-1 and control cell line (Figure 1d). In addition, hEx inhibits shEx-2). Due to the understudied nature of hEx, a cell proliferation in a dose-dependent manner (Supple- reliable antibody for hEx does not exist. Therefore, to mentary Figure S1), showing that rates of cell prolifera- assess the relative levels of hEx in MCF10A cells we tion correlate with hEx expression. used qRT–PCR to measure mRNA levels and show that

Oncogene hEx functions as a tumor suppressor in human cancer cell lines S Visser-Grieve et al 1191 hEx expression is reduced (Figure 1e; Supplementary 28.53±3.46% cell death after treatment with 200 nM Figure S2). Significantly, when hEx is downregulated, Taxol, cells expressing hEx were significantly more the rate of proliferation in MCF10A cells dramatically sensitive with 42.03±2.97% cell death. Importantly, increases (Figure 1f), providing final evidence that hEx knockdown of hEx in hEx-expressing cells is also resis- is directly involved in regulating cell proliferation. tant to Taxol with a maximum of 29.29±3.21% cell In addition to the importance of cell proliferation in death. Similar results were also obtained when cells were tumor progression, apoptosis also has a vital role, both treated with lower concentrations of Taxol, showing that in eliminating tumor cells during the early onset of sensitivity to Taxol correlates with hEx expression tumorigenesis, as well as during the development of drug (Figure 2a). resistance. A widely used chemotherapeutic in the A role of hEx in mediating drug response is further treatment of breast and ovarian cancer is the micro- supported by analysis of the MCF10A cells with reduced tubule inhibitor Taxol (McGrogan et al., 2008). Using hEx expression. Compared with MDA-MB-231 cells, the the MDA-MB-231 stable cell lines expressing hEx, cells MCF10A cells are relatively sensitive to Taxol treatment, were treated with increasing concentrations of Taxol. a property consistent with their non-tumorigenic nature. As shown in Figure 2a, although the control cells However, when hEx expression is reduced, the cells are relatively resistant to Taxol with a maximum become resistant to Taxol treatment (Figure 2b). Unlike control cells (pLKO.1) that respond dramatically to low concentrations of Taxol with significantly increasing cell death with increasing Taxol concentrations, both shEx-1 45 WPI and shEx-2 have very little cell death at low drug hEx 40 concentrations. The most dramatic difference is seen at hEx-shEx 10 nM Taxol with 39.26±4.60% cell death in pLKO-1 35 expressing MCF10A cells compared with an almost 30 twofold reduction in cell death in both shEx-1 and shEx- 2 cell lines with 23.04±1.51 and 23.33±1.06% cell 25 death, respectively. This is the first evidence to delineate 20 the cellular functions of hEx, showing that hEx not only regulates cell proliferation but also is an important 15

Percent cell death mediator of drug response. 10 The acquisition of the sustained proliferative and apoptotic resistance properties exhibited by cells with 5 downregulated hEx levels is reminiscent of key hall- 0 marks of tumor cells (Hanahan and Weinberg, 2011). 0 10 2550 100 200 This prompted us to assess the ability of hEx to function nM Taxol as a tumor suppressor in vitro. Initial analysis with a 80 clonogenic assay measuring the ability of single cells to pLKO.1 form clones, a property attributed to tumor cells, shEx-1 70 demonstrated that whereas the control MDA-MB-231 shEx-2 breast cancer cells expressing vector alone (WPI) were 60 able to efﬁciently form colonies after 10 days with a ± 50 total number of 325 34 colonies, expression of hEx signiﬁcantly inhibited the formation of colonies with 40 a total number of only 192±31 colonies (Figure 3a). As before, loss of hEx in the hEx-overexpressing line 30 rescues the ability of cells to form colonies with 309±49

Percent cell death colonies. 20 Transformation assays provided further evidence that 10 hEx functions as a tumor suppressor in cancer cell lines. Expression of hEx severely inhibited the ability of 0 MDA-MB-231 cells to form colonies in soft agar with 0 1 2.5 5 10 25 only 24±16 colonies formed compared with 86±11 nM Taxol colonies formed in the WPI control cells and 77±19 Figure 2 hEx expression affects sensitivity to the chemotherapeu- colonies in the hEx-shEx cell line (Figure 3b). On the tic drug Taxol. (a) WPI, hEx and hEx-shEx MDA-MB-231 cell other hand, loss of hEx in the non-tumorigenic lines described in Figure 1 were plated in triplicate in 24-well plates at a density of 2 Â 104 cells per well. Cells were treated with 0, 10, MCF10A cell line causes transformation of these cells. 25, 50, 100 and 200 nM of Taxol for 48 h. Cell death was quantified Where essentially no colonies form in vector control by trypan blue analysis. Experiments were performed at least three pLKO.1 cells (12±3), there is a significant increase in times. Data are mean±s.d. (b) Loss of hEx renders MCF10A cells anchorage-independent growth in both shEx-1 (78±9) resistant to Taxol-induced cell death. MCF10A cells with decreased ± hEx expression as established in Figure 1 were plated in triplicate at and shEx-2 (52 7) cell lines (Figure 3c). Finally, in vivo, 3 Â 104 cells per well of 24-well plates and treated with 0, 1, 2.5, 5, hEx can dramatically inhibit tumor growth in mouse 10 or 25 nM Taxol for 48 h. Cell death was quantified as above. xenograph models (Figures 3d and e).

Oncogene hEx functions as a tumor suppressor in human cancer cell lines S Visser-Grieve et al 1192 WPI hEx hEx-shEx 360 300 240 MDA-MB-231 * 180 120 60 Number of colonies 0 WPI hEx hEx-shEx

WPI hEx hEx-shEx 100 80

MDA-MB-231 60

40 *

20 Number of colonies 0 WPI hEx hEx-shEx

pLKO.1 shEx-1 shEx-2 90 * 75 60 * MCF10A 45 30 15 Number of colonies 0 pLKO.1 shEx-1 shEx-2

600 WPI )

3 hEx 500 400 WPI 300 200

100 hEx Tumor volume (mm 0 0 5 10 15 20 25 Days Figure 3 hEx is a putative tumor suppressor (a) Colony formation assay. MDA-MB-231 stable cell lines expressing WPI, hEx or hEx- shEx as described in Figure 1 were plated at 1 Â 103 cells per 100 mm plate. Ten days after plating, colonies were stained with 0.005% crystal violet in 20% methanol. Pictures were obtained and colonies counted using the Bio-Rad Gel Doc System (Bio-Rad, Mississauga, Canada). Quantization is shown on the right. Data are mean±s.d. of three samples per cell line. Statistical analysis using a Student’s t-test was performed and *indicates significant difference (P-value p0.05). (b, c) Soft agar assay. About 2 Â 104 MDA-MB- 231 cells overexpressing hEx (b) or MCF10A cells with reduced hEx expression (c) were mixed with 0.4% agarose in growth media and overlaid on 0.8% agarose in a 6-well plate. Four weeks after plating, colonies were stained and quantified as above. Quantization is shown on the right. Data are mean±s.d. of three samples per cell line. Statistical analysis using a Student’s t-test was performed and *indicates a significant difference (P-value p0.05). (d) hEx inhibits tumor growth in vivo. About 4 Â 106 cells were injected into each flank (left: MDA-MB-231-WPI; right: MDA-MB-231-hEx) of 12 nude mice. Tumor size was measured every 8 days using a digital caliper and calculated as follows: 0.4 Â A Â B2 (A, length; B, wide; in mm). Data are mean±s.d. of 12 mice. (e) Representative image of tumor growth in mice.

These results are consistent with Drosophila models 1997). In addition, further studies also suggested where ex has also been implicated in cell growth that where overexpression of ex induces cell death and apoptosis as functions of its tumor suppressor (Blaumueller and Mlodzik, 2000), loss of ex along with activity. In fact, some of the first studies on ex showed its FERM domain partner mer suppressed apoptosis in that it regulated cell proliferation in the imaginal discs the developing eye (Hamaratoglu et al., 2006). Together, in the developing flies and when overexpressed led to the enhanced proliferation and the loss of apoptosis decreased cell number in the wings (Boedigheimer et al., in ex mutant flies lead to tumor overgrowth.

Oncogene hEx functions as a tumor suppressor in human cancer cell lines S Visser-Grieve et al 1193 Although the Hippo pathway is more complex in autophosphorylate on conserved threonine residues mammalian systems, the proliferation, apoptosis and (Radu and Chernoff, 2009) and once active, phosphory- tumor suppressive functions of hEx are similar to late LATS1 on T1079 leading to LATS1 autophos- several human Hippo pathway members. Not only are phorylation on S909 (Chan et al., 2005). Once active, LATS1 and LATS2 considered tumor suppressors LATS1 phosphorylates YAP on S127 (Hao et al., 2008) (Visser and Yang, 2010b), but also in cell lines over- and TAZ on S89 (Lei et al., 2008). Using specific anti- expression of either LATS1 or LATS2 inhibits proli- bodies for these phosphorylation sites, it is evident that feration and induces cell death (Yang et al., 2001; Xia hEx does not significantly enhance the phosphorylation et al., 2002; Li et al., 2003), and we have shown that loss of the canonical Hippo pathway. To further confirm of both LATS1 and LATS2 enhances cell proliferation that hEx does not function through this pathway, we and renders cells resistant to Taxol (Visser and Yang, knocked down the central kinases LATS1 and LATS2 in 2010a). In addition, phosphorylation and expression of hEx-expressing MDA-MB-231 cells (Figure 4b) and both downstream targets YAP (Li et al., 2011) and TAZ show that loss of LATS1 and LATS2 was not able to (Chan et al., 2008; Lai et al., 2011) have been associated alter the inhibitory effect of hEx on cell proliferation with breast tumor progression and Taxol sensitivity. (Figure 4c). Therefore, unlike in Drosophila, where loss Because expanded is an upstream component of of dlats rescues the ex phenotype (Feng and Irvine, the evolutionarily conserved Drosophila Hippo pathway, 2007), hEx tumor suppressor function is independent of we wanted to determine if hEx functions in an evolu- the Hippo pathway. tionarily conserved manner by activating the Hippo Importantly, although much of the Hippo pathway is pathway. In Drosophila, ex expression leads to enhanced conserved from flies to humans, complexity increases phosphorylation and activation of dlats and subsequent moving up the evolutionary tree. Therefore, differences phosphorylation and inhibition of yorkie (Hamaratoglu can arise in proteins structure, interactions and function. et al., 2006; Yu et al., 2010). Surprisingly, when we Importantly, evidence already exists to support this assessed the relative phosphorylation levels of the hypothesis. Although hEx is the closest human homolog mammalian Hippo pathway components, including to ex, their structure is considerably different. hEx and MST1/2, LATS1, YAP and TAZ, in hEx-overexpressing Drosophila ex share the conserved FERM domain, MDA-MB-231 cells we found no significant differences but hEx lacks the C-terminal region of Drosophila ex (Figure 4a). Upon stimulation, MST1 and MST2 which contains several PPXY motifs responsible for

G1: 42.5% WPI hEx 119 S: 26.8% hEx (αFLG) G2: 30.7% WPI hEx hEx-shLATS WPI hEx (αFLG) LATS1 WPI hEx hEx-shEx p15 LATS1 LATS1 pS909 0 1023 PMT4 Lin LATS2 p18 YAP 116 G1: 41.3% S: 35.4% β-actin G2: 23.3% p21 pYAP hEx p27 TAZ 27 WPI 24 hEx ) 0 1023 pTAZ 4 21 hEx-shLATS Cyclin A 111 18 G1: 44.1% MST2 15 S: 26.9% Cyclin B G2: 29.0% 12 hEx-shEx MST1 9 Cyclin E 6 pMST1/2 Cell number (x10 3 β-actin 0 1023 β 0 PMT4 Lin -actin 012345 Days Figure 4 hEx functions independently of the Hippo pathway (a). Total cell lysate was extracted from MDA-MB-231 cells infected with lentivirus expressing WPI or hEx-FLG. Western blot using speciﬁc phospho-antibodies shows hEx does not activate the Hippo pathway. Antibodies used are as follows: LATS1 rabbit polyclonal antibody, pLATS1-S909 (Y03), pYAP-127, TAZ, MST1, MST2 and pMST1/2 were purchased from Cell Signaling Technologies (Pickering, ON, Canada); YAP and pTAZ-S89 were purchased from Santa Cruz Biotechnologies (Santa Cruz, CA, USA). (b) Downregulation of LATS1 and LATS2 in hEx-expressing MDA-MB-231 cells. Western blot shows expression of hEx (anti-FLG), LATS1 and LATS2 (Bethyl Laboratories, Burlington, ON, Canada) after infection of lentivirus expressing vector alone (WPI), hEx-FLG (hEx) or hEx followed by the subsequent infection with lentivirus expressing shRNA targeting LATS1 and lentivirus expressing shRNA targeting LATS2 (hEx-shLATS). Targeting sequences for LATS1 and LATS2 are as follows: LATS1, 50-GTCTGCTTCATACATTCCTAA-30; LATS2, 50-CAGGACCAAACAGTGAC ACTT-30 (c) Cell proliferation assay as described in Figure 1. (d) Expression of hEx causes S-phase arrest. About 1 Â 106 MDA-MB- 231 stable cell lines expressing WPI, hEx or hEx-shEx as described in Figure 1 cells were harvested, treated with RNase A (200 mg/ml), stained with propidium iodide (50 mg/ml) and analyzed with EPICS ALTRA HSS Flow cytometer. The percentage of each cell-cycle phase (G1, S, G2/M) was calculated using Cylchred software. (e) Protein lysate was extracted from MDA-MB-231 stable lines expressing WPI, hEx or shEx and analyzed by western blot. Antibodies used are as follows: anti-FLAG M2 (Sigma), anti-p15, anti-18, anti-p21 and p27, anti-Cyclin A2, anti-Cyclin B1 and anti-Cyclin E2 (Cell Signaling Technologies).

Oncogene hEx functions as a tumor suppressor in human cancer cell lines S Visser-Grieve et al 1194 protein–protein interactions (Gunn-Moore et al., 2005; a panel of cell-cycle proteins. Expression of hEx most Badouel et al., 2009). Interestingly, it has been suggested significantly increases p21Cip1 and decreases Cyclin A that the tumor suppressor function of Drosophila ex is expression (Figure 4e). The p21Cip1 cyclin kinase inhi- mediated primarily via its C-terminus (Pellock et al., bitor has pleiotrophic effects, affecting all stages of cell- 2007), which is further evidence that hEx and Drosophila cycle progression (Jung et al., 2010), and the ability of ex, although both tumor suppressors, mediate their p21Cip1 to inhibit S-phase progression has been linked to effects by distinct mechanisms. the downregulation of Cyclin A expression (Ogryzko In addition, disparities between Drosophila and et al., 1997). human systems have been shown in both the down- In summary, this work characterizes the cellular stream transcriptional targets and the upstream regula- functions of hEx protein in cancer cell lines and tory proteins of the Hippo pathway. For example, shows that it is a potent inhibitor of cell proliferation, cyclin E, diap1 and bantam are key genes downstream of transformation and modulator of drug sensitivity. This Yki in Drosophila (Zhao et al., 2010), but have not suggests that hEx has tumor suppressor properties and been shown to be transcriptional targets of its orthologs future work should confirm this nature of hEx through (Ota and Sasaki, 2008; Zhao et al., 2008; Lai et al., mutational studies and/or clinical cancer sample analy- 2011). Additionally, whereas dRASSF antagonizes the sis. Importantly, hEx functions in a Hippo-independent Hippo pathway in Drosophila (Polesello et al., 2006), its manner, shedding new light on distinct Hippo signaling mammalian homolog RASSF1A activates MST1/2 and pathways in flies and mammals. the kinase cascade (Guo et al., 2007). Also upstream, Drosophila Kibra interacts with multiple Hippo pathway Conflict of interest components including mer, ex, hpo and dlats. However, its human homolog binds only Merlin but not MST2 or The authors declare no conflict of interest. hEx (Genevet et al., 2010; Yu et al., 2010). Together, this suggests that although the core Hippo kinase cassette is conserved, both the mediators and regulators of this Acknowledgements pathway are more diverse in mammalian systems. Understanding the mechanisms mediating hEx func- This work was supported by a grant from the Canadian tion is essential. To this end, we propose a model Institute of Health Research, a New Investigator Award from whereby hEx suppresses S-phase progression by upregu- the Canadian Institute of Health Research and an Early Cip1 Researcher Award from the Ontario Ministry of Research lating p21 and downregulating Cyclin A. Cell-cycle and Innovation, Canada, to Xiaolong Yang; and an Ontario analysis of hEx-overexpressing MDA-MB-231 cells Graduate Scholarship to Stacy Visser-Grieve. We would like shows an accumulation of cells in the S phase with a to thank Matt Gordon at the Queen’s Cancer Research concomitant decrease in the G2 phase (Figure 4d). To Institute for performing the FACS analysis and Dr Frank assess the mechanism, we surveyed the expression of Gunn-Moore for providing the hEx-GFP plasmid.

References

Badouel C, Gardano L, Amin N, Garg A, Rosenfeld R, Le Bihan T Gunn-Moore FJ, Welsh GI, Herron LR, Brannigan F, Venkateswarlu K, et al. (2009). The FERM-domain protein expanded regulates hippo Gillespie S et al. (2005). A novel 4.1 ezrin radixin moesin (FERM)- pathway activity via direct interactions with the transcriptional containing protein, ‘willin’. FEBS Lett 579: 5089–5094. activator yorkie. Dev Cell 16: 411–420. Guo C, Tommasi S, Liu L, Yee JK, Dammann R, Pfeifer GP. Blaumueller CM, Mlodzik M. (2000). The drosophila tumor suppres- (2007). RASSF1A is part of a complex similar to the drosophila sor expanded regulates growth, apoptosis, and patterning during Hippo/Salvador/Lats tumor-suppressor network. Curr Biol 17: development. Mech Dev 92: 251–262. 700–705. Boedigheimer M, Laughon A. (1993). Expanded: a gene involved in Halder G, Johnson RL. (2011). Hippo signaling: growth control and the control of cell proliferation in imaginal discs. Development beyond. Development 138: 9–22. 118: 1291–1301. Hamaratoglu F, Willecke M, Kango-Singh M, Nolo R, Hyun E, Boedigheimer MJ, Nguyen KP, Bryant PJ. (1997). Expanded functions Tao C et al. (2006). The tumour-suppressor genes NF2/Merlin and in the apical cell domain to regulate the growth rate of imaginal expanded act through hippo signalling to regulate cell proliferation discs. Dev Genet 20: 103–110. and apoptosis. Nat Cell Biol 8: 27–36. Chan EH, Nousiainen M, Chalamalasetty RB, Schafer A, Nigg EA, Hanahan D, Weinberg RA. (2011). Hallmarks of cancer: the next Sillje HH. (2005). The Ste20-like kinase Mst2 activates the human generation. Cell 144: 646–674. large tumor suppressor kinase Lats1. Oncogene 24: 2076–2086. Hao Y, Chun A, Cheung K, Rashidi B, Yang X. (2008). Tumor Chan SW, Lim CJ, Guo K, Ng CP, Lee I, Hunziker W et al. (2008). A suppressor LATS1 is a negative regulator of oncogene YAP. J Biol role for TAZ in migration, invasion, and tumorigenesis of breast Chem 283: 5496–5509. cancer cells. Cancer Res 68: 2592–2598. Jung YS, Qian Y, Chen X. (2010). Examination of the expanding Feng Y, Irvine KD. (2007). Fat and expanded act in parallel to pathways for the regulation of p21 expression and activity. regulate growth through warts. Proc Natl Acad Sci USA 104: Cell Signal 22: 1003–1012. 20362–20367. Lai D, Ho KC, Hao Y, Yang X. (2011). Taxol resistance in breast Genevet A, Wehr MC, Brain R, Thompson BJ, Tapon N. (2010). cancer cells is mediated by the hippo pathway component TAZ Kibra is a regulator of the Salvador/Warts/Hippo signaling and its downstream transcriptional targets Cyr61 and CTGF. network. Dev Cell 18: 300–308. Cancer Res 71: 2728–2738.

Oncogene hEx functions as a tumor suppressor in human cancer cell lines S Visser-Grieve et al 1195 Lau YK, Murray LB, Houshmandi SS, Xu Y, Gutmann DH, Yu Q. Polesello C, Huelsmann S, Brown NH, Tapon N. (2006). The (2008). Merlin is a potent inhibitor of glioma growth. Cancer Res drosophila RASSF homolog antagonizes the hippo pathway. 68: 5733–5742. Curr Biol 16: 2459–2465. Lei QY, Zhang H, Zhao B, Zha ZY, Bai F, Pei XH et al. (2008). Radu M, Chernoff J. (2009). The DeMSTification of mammalian TAZ promotes cell proliferation and epithelial-mesenchymal Ste20 kinases. Curr Biol 19: R421–R425. transition and is inhibited by the hippo pathway. Mol Cell Biol Stern C, Bridges CB. (1926). The mutants of the extreme left end of the 28: 2426–2436. second chromosome of DROSOPHILA MELANOGASTER. Li W, Wang L, Katoh H, Liu R, Zheng P, Liu Y. (2011). Identification Genetics 11: 503–530. of a tumor suppressor relay between the FOXP3 and the hippo Visser S, Yang X. (2010a). Identification of LATS transcriptional pathways in breast and prostate cancers. Cancer Res 71: 2162–2171. targets in HeLa cells using whole human genome oligonucleotide Li Y, Pei J, Xia H, Ke H, Wang H, Tao W. (2003). Lats2, a putative microarray. Gene 449: 22–29. tumor suppressor, inhibits G1/S transition. Oncogene 22: 4398–4405. Visser S, Yang X. (2010b). LATS tumor suppressor: a new governor of McCartney BM, Kulikauskas RM, LaJeunesse DR, Fehon RG. cellular homeostasis. Cell Cycle 9: 3892–3903. (2000). The neurofibromatosis-2 homologue, merlin, and the tumor Xia H, Qi H, Li Y, Pei J, Barton J, Blackstad M et al. (2002). suppressor expanded function together in drosophila to regulate LATS1 tumor suppressor regulates G2/M transition and apoptosis. cell proliferation and differentiation. Development 127: 1315–1324. Oncogene 21: 1233–1241. McGrogan BT, Gilmartin B, Carney DN, McCann A. (2008). Xiao L, Chen Y, Ji M, Dong J. (2011). KIBRA regulates hippo Taxanes, microtubules and chemoresistant breast cancer. Biochim signaling activity via interactions with large tumor suppressor Biophys Acta 1785: 96–132. kinases. J Biol Chem 286: 7788–7796. Ogryzko VV, Wong P, Howard BH. (1997). WAF1 retards S-phase Yang X, Li DM, Chen W, Xu T. (2001). Human homologue of progression primarily by inhibition of cyclin-dependent kinases. drosophila lats, LATS1, negatively regulate growth by inducing Mol Cell Biol 17: 4877–4882. G(2)/M arrest or apoptosis. Oncogene 20: 6516–6523. Ota M, Sasaki H. (2008). Mammalian Tead proteins regulate cell Yu J, Zheng Y, Dong J, Klusza S, Deng WM, Pan D. (2010). Kibra proliferation and contact inhibition as transcriptional mediators of functions as a tumor suppressor protein that regulates hippo signaling hippo signaling. Development 135: 4059–4069. in conjunction with merlin and expanded. Dev Cell 18: 288–299. Pan D. (2010). The hippo signaling pathway in development and Zhao B, Li L, Lei Q, Guan KL. (2010). The hippo-YAP pathway in cancer. Dev Cell 19: 491–505. organ size control and tumorigenesis: an updated version. Genes Pellock BJ, Buff E, White K, Hariharan IK. (2007). The drosophila Dev 24: 862–874. tumor suppressors expanded and merlin differentially regulate Zhao B, Ye X, Yu J, Li L, Li W, Li S et al. (2008). TEAD mediates cell cycle exit, apoptosis, and wingless signaling. Dev Biol 304: YAP-dependent gene induction and growth control. Genes Dev 22: 102–115. 1962–1971.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene