Oncogene (2011) 30, 2718–2729 & 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc ORIGINAL ARTICLE Homeodomain DLX4 counteracts key transcriptional control mechanisms of the TGF-b cytostatic program and blocks the antiproliferative effect of TGF-b

BQ Trinh1,2, N Barengo1 and H Naora1,2

1Department of Systems Biology, University of Texas MD Anderson Cancer Center, Houston, TX, USA and 2Cancer Biology Program, Graduate School of Biomedical Sciences, University of Texas Health Sciences Center, Houston, TX, USA

The antiproliferative activity of transforming growth In turn, TGF-b type I phosphorylates Smad2 factor-b (TGF-b) is essential for maintaining normal and Smad3, which translocate to the nucleus and form tissue homeostasis and is lost in many types of tumors. complexes with Smad4 and other DNA-binding factors responses that are central to the TGF-b cytostatic to activate or repress transcription (Shi and Massague´, program include activation of the cyclin-dependent kinase 2003; Feng and Derynck, 2005). In most types of normal inhibitors, p15Ink4B and p21WAF1/Cip1, and repression of cells, TGF-b inhibits cell proliferation by inducing arrest c-. These gene responses are tightly regulated by a in G1 phase. The gene responses that are central to the repertoire of transcription factors that include Smad TGF-b cytostatic program are activation of the cyclin- and Sp1. The DLX4 patterning gene dependent kinase inhibitors p15Ink4B and p21WAF1/Cip1, encodes a that is absent from most and repression of the growth-promoting c-myc and ID normal adult tissues, but is expressed in a wide variety of transcription factors (Reynisdo´ttir et al., 1995; Chen malignancies, including lung, breast, prostate and ovarian et al., 2002; Kang et al., 2003). The cytostatic program is cancers. In this study, we demonstrate that DLX4 blocks tightly integrated by feedback loops that protect against the antiproliferative effect of TGF-b. DLX4 inhibited competing mitogenic signals (Siegel and Massague´, TGF-b-mediated induction of p15Ink4B and p21WAF1/Cip1 2003). For example, TGF-b activates p15Ink4B and expression. DLX4 bound and prevented Smad4 from p21WAF1/Cip1 transcription via cooperative interactions forming complexes with Smad2 and Smad3, but not between Sp1 and Smad proteins, and also prevents with Sp1. However, DLX4 also bound and inhibited repression of these by c-myc (Feng et al., 2000, DNA-binding activity of Sp1. In addition, DLX4 induced 2002; Pardali et al., 2000; Gartel et al., 2001). expression of c-myc independently of TGF-b/Smad Resistance to TGF-b-mediated growth inhibition is signaling. The ability of DLX4 to counteract key an important feature in the pathogenesis of a wide transcriptional control mechanisms of the TGF-b cyto- variety of tumors (Siegel and Massague´, 2003; Massa- static program could explain, in part, the resistance of gue´, 2008). In several types of tumors, this resistance has tumors to the antiproliferative effect of TGF-b. been attributed to mutations or deletion of core Oncogene (2011) 30, 2718–2729; doi:10.1038/onc.2011.4; components of the TGF-b signaling pathway. Inactivat- published online 7 February 2011 ing mutations in the TGF-b type II receptor have been reported in 60–90% of colon cancers associated with Keywords: cell growth; homeobox genes; Smad; Sp1; microsatellite instability (Markowitz et al., 1995; Wata- TGF-b nabe et al., 2001). Mutations or deletions of Smad4 occur in B50% of pancreatic and non-microsatellite instability colorectal cancers (Hahn et al., 1996; Introduction Woodford-Richens et al., 2001). By contrast, TGF-b type II receptor mutations have not been detected in Transforming growth factor-b (TGF-b) mediates di- lung cancers with microsatellite instability (Takenoshita verse processes, such as proliferation, differentiation et al., 1997), and the frequency of Smad4 mutations in and apoptosis, during embryogenesis and in adult lung cancers is low (7%) (Nagatake et al., 1996). tissues (Siegel and Massague´, 2003; Massague´, 2008). Similarly, many breast and prostate cancers are resistant On binding of TGF-b, the TGF-b type II receptor to TGF-b-mediated growth inhibition, but rarely con- phosphorylates and activates the TGF-b type I receptor. tain TGF-b type II receptor or Smad mutations (Schutte et al., 1996; Takenoshita et al., 1998; Bello-DeOcampo and Tindall, 2003). TGF-b receptor mutations occur in Correspondence: Dr H Naora, Department of Systems Biology, 12–31% of ovarian cancers (Wang et al., 2000; Francis- University of Texas MD Anderson Cancer Center, 7435 Fannin Thickpenny et al., 2001), but many TGF-b-resistant Street, Unit 950, Houston, TX 77054, USA. E-mail: [email protected] ovarian cancers have been found to express functional Received 13 October 2010; revised 20 December 2010; accepted 4 receptors (Yamada et al., 1999). Resistance of many January 2011; published online 7 February 2011 tumors to TGF-b-mediated growth inhibition therefore Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2719 cannot solely stem from mutations or deletions of core not inhibited by TGF-b (Figure 1g). On the other hand, components of the TGF-b signaling pathway and likely reconstitution of Smad4 in these cells increased respon- arise from other aberrations. siveness to TGF-b (Figure 1g). This responsiveness to Homeobox genes encode transcription factors that TGF-b was abrogated by DLX4 (Figure 1g). Expression control cell differentiation (McGinnis and Krumlauf, of p21WAF1/Cip1 was induced by TGF-b in MDA-MB-468 1992). DLX4 (also reported as BP1) (Haga et al., 2000) cells reconstituted with Smad4, and this induction was is a member of the DLX family of homeobox genes that blocked by DLX4 (Figure 1h). Enforced expression of controls many aspects of embryonic development, DLX4 induced c-myc levels, irrespective of the absence including bone morphogenesis and skeletal patterning or presence of Smad4 (Figure 1h). We further investi- (Panganiban and Rubenstein, 2002). Although absent gated the effect of DLX4 on c-myc induction by from most normal adult tissues, DLX4 is widely assaying c-myc promoter activity. Activity of the c-myc expressed in leukemias, lung, breast, ovarian and promoter was repressed by TGF-b in control Mv1Lu prostate cancers (Haga et al., 2000; Man et al., 2005; cells (Figure 2a). In contrast, expression of DLX4 in Hara et al., 2007; Tomida et al., 2007; Schwartz et al., Mv1Lu cells induced c-myc promoter activity irrespec- 2009). Because DLX4 is expressed in diverse types tive of TGF-b signaling (Figure 2a). Conversely, knock- of tumors, we considered the possibility that DLX4 down of DLX4 in MCF-7 cells inhibited c-myc controls a pathogenic mechanism common to multiple promoter activity and increased sensitivity to TGF-b- organ sites. Our study demonstrates that DLX4 blocks mediated repression (Figure 2b). These results suggest gene responses that are central to the TGF-b cytostatic that DLX4 blocks TGF-b/Smad-dependent repression program by sequestering Smad4 and Sp1, and by of c-myc expression, and also induces c-myc levels inducing c-myc. This blocking activity of DLX4 could independently of TGF-b/Smad signaling. explain, in part, why tumor cells are resistant to the The first 113 bp of the p15Ink4B promoter are essential antiproliferative effect of TGF-b. for induction by TGF-b and contain two Smad-binding elements (SBEs; Feng et al., 2000). Activity of this minimal promoter region was induced by TGF-b in Results control Mv1Lu cells, and this induction was abolished by mutation of the SBEs (Figure 2c). Expression of DLX4 blocks TGF-b-induced, Smad-dependent responses DLX4 in Mv1Lu cells abolished the induction of wild- To determine whether DLX4 blocks the antiprolifera- type p15Ink4B promoter activity by TGF-b (Figure 2c). tive effect of TGF-b, we initially used the non- DLX4 also modestly inhibited activity of the SBE- tumorigenic lung epithelial cell line Mv1Lu. Mv1Lu is mutant promoter (Figure 2c). To confirm that DLX4 a well-established model for studying TGF-b-mediated blocks Smad-dependent transcription, we performed induction of p15Ink4B expression and growth arrest reporter assays using MDA-MB-468 cells. Although (Reynisdo´ttir et al., 1995; Feng et al., 2002). Whereas wild-type p15Ink4B promoter activity was unresponsive to growth of Mv1Lu cells was inhibited by TGF-b in a TGF-b in Smad4-deficient MDA-MB-468 cells, respon- dose-dependent manner, enforced expression of DLX4 siveness was conferred when Smad4 was expressed in in these cells increased resistance to TGF-b (Figure 1a). these cells. This Smad4-dependent responsiveness was Expression of DLX4 in Mv1Lu cells blocked the ability eliminated when DLX4 was coexpressed (Figure 2d). of TGF-b to induce p15Ink4B expression (Figure 1b). Cell In addition to its antiproliferative effect, TGF-b is cycle analysis revealed that DLX4 expression inhibited well known to induce epithelial-to-mesenchymal transi- the induction of G1 arrest in Mv1Lu cells by TGF-b tion (EMT; Siegel and Massague´, 2003; Massague´, (Figure 1c). Enforced expression of DLX4 in the non- 2008). Because DLX4 abrogated TGF-b-mediated tumorigenic mammary epithelial cell line NMuMG also growth inhibition, DLX4 might also block the ability inhibited TGF-b-induced G1 arrest (Supplementary of TGF-b to induce EMT. To address this question, we Figure 1). In converse experiments, we determined used the NMuMG cell line, a well-established model for whether knockdown of DLX4 in tumor cells increased studying TGF-b-induced EMT. Smad4 has been de- TGF-b-mediated growth inhibition using short hairpin monstrated to be essential for TGF-b-induced EMT in RNAs that targeted different sites of DLX4 (sh90 and several cell types including NMuMG cells (Deckers sh92). The ability of short hairpin RNAs to knockdown et al., 2006). TGF-b treatment of vector-control DLX4 in MCF-7 breast cancer cells was confirmed by NMuMG cells induced profound epithelial-to-fibroblas- western blot (Figure 1d) and also by immunofluores- tic morphologic transformation (Supplementary Figure cence staining and real-time quantitative PCR (Supple- 3A), loss of E-cadherin and upregulation of N-cadherin mentary Figures 2A and B). Knockdown of DLX4 in (Supplementary Figures 3B and C). Enforced expression MCF-7 cells was observed to increase sensitivity to of DLX4 blocked downregulation of E-cadherin and TGF-b in cell viability assays (Figure 1e), and also induction of N-cadherin (Supplementary Figures 3B, C). increased the proportion of cells in G1 phase (Figure 1f). Epithelial morphology was considerably retained in Because TGF-b can also inhibit growth by Smad- TGF-b-treated þ DLX4 NMuMG cells (Supplementary independent mechanisms (Petritsch et al., 2000), we used Figure 3A). Together, our findings that DLX4 blocks the Smad4-deficient MDA-MB-468 breast cancer cell TGF-b-mediated, Smad-dependent growth inhibition line to confirm that DLX4 blocks Smad-dependent and EMT indicate that DLX4 inhibits a core component growth inhibition. Growth of MDA-MB-468 cells was of the TGF-b/Smad signaling pathway.

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2720

Figure 1 DLX4 blocks TGF-b-mediated growth inhibition. (a) Vector-control (ÀDLX4) and þ DLX4 stable Mv1Lu lines were cultured with the indicated concentrations of TGF-b for 2 days. Changes in cell growth were determined by MTT assay, and expressed relative to growth of cells incubated without TGF-b. Shown are results of two independent experiments each performed in triplicate. (b) Western blot analysis of Mv1Lu lines following treatment without and with TGF-b (10 ng/ml) for 16 h. (c) Mv1Lu lines were treated without and with TGF-b for 18 h. Indicated are the proportions of cells in G1, S and G2/M phases determined by flow cytometric analysis of propidium iodide staining. (d) MCF-7 cells were transfected with empty vector, non-targeting short hairpin RNA (shRNA) and DLX4 shRNAs (sh90 and sh92). At 2 days after transfection, DLX4 levels were assayed by western blot. (e) Transfected MCF-7 cells were cultured with the indicated concentrations of TGF-b for 2 days. Changes in cell growth were determined by MTT assay. (f) Transfected MCF-7 cells were treated without and with TGF-b (10 ng/ml) for 18 h and proportions of cells in cell cycle phases determined thereafter. (g) Vector-control (ÀDLX4) and þ DLX4 stable MDA-MB-468 lines were transfected with Smad4. At 24 h thereafter, cells were cultured without and with TGF-b (10 ng/ml) for 2 days, and changes in cell growth were examined by MTT assay. (h) Western blot analysis of MDA-MB-468 lines following treatment without and with TGF-b for 16 h.

Figure 2 DLX4 induces c-myc promoter activity and inhibits TGF-b-mediated induction of p15Ink4B promoter activity. (a) Mv1Lu cells were cotransfected with empty vector (gray bar) or DLX4 (black bar), together with empty pBV-Luc vector or with pBV- MYC(Del4) reporter plasmid that contains 900 bp of c-myc P1 and P2 promoter sequences. Transfected cells were cultured without and with TGF-b (10 ng/ml) for 18 h, and assayed for firefly luciferase (F-Luc) activity. (b) Reporter assays for c-myc promoter activity were likewise conducted using MCF-7 cells that were cotransfected with non-targeting short hairpin RNA (shRNA; gray bar) and DLX4 (sh90) shRNA (black bar). (c) Mv1Lu cells were cotransfected with empty vector (gray bar) or DLX4 (black bar), together with reporter plasmids containing no promoter (pGL2 vector), p15Ink4B promoter sequences (À113 to þ 70; p15-WT) and p15Ink4B promoter with mutated SBEs (p15-SBE-mt). Cells were cultured without and with TGF-b for 18 h, and assayed for F-Luc activity. (d) Reporter assays using the p15-WT reporter plasmid were performed using transfected MDA-MB-468 lines. Shown are relative F-Luc activities in three independent experiments, each performed in duplicate. Values were normalized by activity of co-transfected Renilla luciferase.

DLX4 blocks transcriptional activity of TGF-b-activated tion of receptor-regulated Smads (R-Smads) (Shi and receptor-regulated Smads Massague´, 2003). DLX4 did not alter levels of Smad2 Smad-dependent transcription is controlled at multiple expression or Smad2 phosphorylation (Figure 3a). levels, including phosphorylation and nuclear localiza- DLX4 also did not interfere with nuclear translocation

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2721

Figure 3 DLX4 blocks Smad transcriptional activity. Vector-control (ÀDLX4) and þ DLX4 Mv1Lu lines were serum starved overnight and then treated without and with TGF-b (10 ng/ml) for 30 min. (a) Total and phosphorylated Smad2 were detected by western blot. (b) Intracellular localization of Smad2 was detected by immunofluorescence staining. Bar, 20 mm(c and d) Mv1Lu and HepG2 cells were cotransfected with empty vector (gray bar) or DLX4 (black bar), together with GAL4-driven firefly luciferase (F- Luc) reporter plasmid and with (c) GAL4–Smad2 or (d) GAL4–Smad3. Transfected cells were cultured without and with TGF-b for 18 h, and assayed for F-Luc activity. (e) GAL4–Smad2 and (f) GAL4–Smad3 activities were likewise assayed in MCF-7 cells that were cotransfected with non-targeting short hairpin RNA (shRNA; gray bar) or DLX4 (sh90) shRNA (black bar). (g, h) HepG2 cells were cotransfected with empty vector (gray bar) or DLX4 (black bar), together with (g) pSBE4-Luc or (h) BRE-Luc reporter plasmids, and then cultured without and with TGF-b (10 ng/ml) or BMP-4 (80 ng/ml) for 18 h. Shown are relative F-Luc activities in three independent experiments, each performed in duplicate and normalized by activity of cotransfected Renilla luciferase. of Smad2 following TGF-b stimulation (Figure 3b). promoter comprising four tandem SBEs (pSBE4-Luc). DLX4 was predominantly localized in the nucleus, and DLX4 blocked induction of this promoter by TGF-b its localization was not affected by TGF-b stimulation and by BMP-4 (Figure 3g). TGF-b- and BMP-specific (Supplementary Figures 4A and B). These findings R-Smads preferentially bind distinct DNA sequences indicate that DLX4 likely inhibits nuclear events down- (Kusanagi et al., 2000; Korchynskyi and ten Dijke, stream of the TGF-b signaling pathway. 2002). Indeed, BMP-4 was not as effective as TGF-b in We determined whether DLX4 inhibits transcrip- inducing pSBE4-Luc activity (Figure 3g). We further tional activity of TGF-b R-Smads. A construct was tested the effect of DLX4 on BMP-induced transcription generated in which the GAL4 DNA-binding domain by using a reporter plasmid that contained the BMP- (DBD) was fused to the linker region and Mad responsive promoter of the Id1 gene. DLX4 blocked homology 2 (MH2) transcriptional activation domain BMP-induced Id1 promoter activity (Supplementary of Smad2 (amino acids 173–467). This chimera was Figure 5). The blocking effect of DLX4 was confirmed coexpressed in Mv1Lu cells with a firefly luciferase by using a synthetic promoter comprising two tandem reporter controlled by five tandem GAL4-binding sites. copies of Id1 BMP response elements (BRE-Luc) Transcriptional activity of GAL4–Smad2 was induced (Figure 3h). Because TGF-b- and BMP-specific R- by TGF-b, and this activation was abolished when Smads utilize Smad4 as the common and essential DLX4 was expressed (Figure 3c). Similar results were partner for formation of functional transcriptional obtained using the hepatoma cell line HepG2 complexes (Shi and Massague´, 2003), our findings raise (Figure 3c). Expression of DLX4 also abolished TGF- the possibility that DLX4 inhibits Smad4. b-mediated activation of a chimera comprising the GAL4–DBD fused to the linker region and MH2 domain of Smad3 (amino acids 133–425) (Figure 3d). DLX4 prevents Smad4 from binding R-Smads Conversely, knockdown of DLX4 in MCF-7 cells We subsequently determined whether DLX4 interferes increased TGF-b-mediated induction of GAL4–Smad2 with the binding of Smad4 to R-Smads. Following and GAL4–Smad3 activity (Figures 3e and f). transfection with FLAG-tagged DLX4 or empty vector, Smad2 and Smad3 serve as substrates for the TGF-b HepG2 cells were treated with or without TGF-b. receptors, whereas other R-Smads (Smads 1, 5 and 8) Smad2 was immunoprecipitated, and precipitates ana- are utilized by the bone morphogenetic protein (BMP) lyzed by immunoblotting using antibody (Ab) to and anti-Mullerian receptors (Shi and Massague´, 2003; Smad4. Binding of Smad4 to Smad2 was inhibited when Feng and Derynck, 2005). We initially tested the ability DLX4 was expressed (Figure 4a). We confirmed these of DLX4 to inhibit transcription induced by other findings by immunoprecipitating Smad4 and detecting members of the TGF superfamily by using a synthetic Smad2 in precipitates (Figure 4a). DLX4 also prevented

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2722

Figure 4 DLX4 prevents Smad4 from binding R-Smads and blocks interactions of Smad and Sp1 proteins with the p15Ink4B promoter. HepG2 cells were transfected with empty vector or with FLAG-tagged DLX4. At 24 h thereafter, cells were serum starved overnight and then treated without and with TGF-b (10 ng/ml) for 30 min. (a) Smad2 was immunoprecipitated from nuclear extracts, and precipitates analyzed by immunoblotting using Ab to Smad4. Conversely, Smad4 was pulled down, and precipitates analyzed by immunoblotting using Smad2 Ab. Because HepG2 cells express low levels of Smad3, immunoprecipitation (IP) assays to detect binding of Smad3 to Smad4 were performed using extracts of cells that had been transfected with Smad3. (b) Western blot of DLX4, Smad proteins and Sp1 in nuclear extracts. (c) Sp1 was immunoprecipitated from nuclear extracts, and precipitates analyzed by immunoblotting using Ab to Smad4. Conversely, Smad4 was pulled down, and Sp1 detected in precipitates. (d) Sequence of the minimal p15Ink4B promoter indicating Sp1-binding sites and SBEs (adapted from Feng et al., 2000). Underlined are sequences contained in the oligonucleotides used for oligonucleotide pulldown assays (solid line) and gel-shift assays (dashed line). (e) Biotinylated oligonucleotide containing sequences À108 to À39 of the p15Ink4B promoter was incubated with HepG2 nuclear extracts and pulled down. DNA-bound proteins in precipitates were analyzed by immunoblotting. (f) ChIP analysis of interactions of Smad and Sp1 proteins with the p15Ink4B promoter. The input fraction corresponded to 1% of the chromatin solution of each sample before IP.

binding of Smad4 to Smad3 (Figure 4a). DLX4 did not association of R-Smads, Smad4 and Sp1 with the alter Smad expression levels (Figure 4b). These results p15Ink4B promoter was detected by ChIP assays in indicate that DLX4 likely inhibits transcriptional ÀDLX4 cells following TGF-b treatment. In contrast, activity of Smad2 and Smad3 by preventing Smad4 interactions of Smad and Sp1 proteins with the p15Ink4B from interacting with these R-Smads. promoter were abrogated in þ DLX4 cells (Figure 4f). Binding affinity and selectivity of Smad complexes for Together, these findings raise the possibility that DLX4 target gene promoters are governed by interactions with binds Smads and/or Sp1. other DNA-binding factors (Shi and Massague´, 2003; Feng and Derynck, 2005). The minimal p15Ink4B promo- ter contains two Sp1-binding sites adjacent to the SBEs, DLX4 directly binds Smad4 and its induction by TGF-b involves binding of Smad4 To initially investigate whether DLX4 associates with to Sp1 as well as to R-Smads (Feng et al., 2000). As Smad4 in cells, immunoprecipitation assays were performed reported by others (Feng et al., 2000; Pardali et al., using extracts of HepG2 cells that expressed FLAG– 2000), TGF-b stimulation induced binding of Smad4 to DLX4. DLX4 associated with Smad4 in the absence and Sp1 in ÀDLX4 cells. This binding was not inhibited by presence of TGF-b stimulation (Figure 5a). Association of DLX4 (Figure 4c). Surprisingly, we observed increased endogenous DLX4 with Smad4 was detected in immuno- association of Smad4 with Sp1 in þ DLX4 cells in the precipitation assays using extracts of MCF-7 cells absence of TGF-b stimulation (Figure 4c). DLX4 did (Figure 5b). This interaction was abrogated when DLX4 not alter expression of Sp1 (Figure 4b). To determine short hairpin RNA was expressed in these cells (Figure 5b). whether DLX4 alters formation of Smad–Sp1–DNA To determine whether DLX4 interacts with Smad4 by complexes, we performed DNA pulldown assays using a direct binding, we tested the ability of in vitro translated biotinylated oligonucleotide that contained nucleotides 35S-labeled DLX4 to bind GST–Smad4 protein. GST- À108 to À39 of the p15Ink4B promoter and included the pulldown assays demonstrated that DLX4 directly binds two Sp1 binding sites and SBEs (Figure 4d). Increased Smad4 (Figure 5c). DLX4 also bound Smad2, albeit levels of Smads were detected in DNA–protein com- more weakly (Figure 5c). We sought to identify the plexes when ÀDLX4 cells were stimulated with TGF-b Smad4-binding domain of DLX4 by testing truncated (Figure 4e). In contrast, this induction was inhibited in GST–DLX4 fusion proteins for their ability to bind þ DLX4 cells (Figure 4e). DLX4 also inhibited inter- in vitro translated 35S-labeled Smad4 (Figure 5d). Dele- action of Sp1 with the p15Ink4B promoter (Figure 4e). To tion of the C-terminal tail of DLX4 only weakly affected confirm the blocking activity of DLX4 in a more its ability to bind Smad4 (Figure 5e). In contrast, physiological context, we performed chromatin immu- deletion of the DNA-binding homeodomain of DLX4 noprecipitation (ChIP) assays. As shown in Figure 4f, markedly inhibited its Smad4-binding ability. Binding of

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2723

Figure 5 DLX4 binds to Smad4. (a) HepG2 cells were transfected with empty vector or with FLAG-tagged DLX4. At 24 h thereafter, cells were serum starved overnight and then treated without and with TGF-b (10 ng/ml) for 30 min. FLAG–DLX4 was immunoprecipitated using FLAG Ab, and precipitates analyzed by immunoblotting using Ab to Smad4. Conversely, FLAG– DLX4 was detected in precipitates following immunoprecipitation (IP) using Smad4 Ab. (b) MCF-7 cells were transfected with non- targeting short hairpin RNA (shRNA) or with DLX4 (sh90) shRNA. Endogenous DLX4 was immunoprecipitated using DLX4 Ab, and precipitates analyzed by immunoblotting using Smad4 Ab. IP using mouse immunoglobulin G was included as a negative control. (c) Expression of GST–Smad2 and GST–Smad4 proteins was confirmed by SDS–polyacrylamide gel electrophoresis (left). GST-fusion proteins were assayed for direct binding to in vitro translated 35S-labeled full-length (FL) DLX4 (right). (d) GST–DLX4 constructs comprising the transactivation domain (TA), homeodomain (HD) and C-terminal tail (C), and Smad4 constructs comprising MH1 and MH2 domains and linker (LK) region. (e) FL DLX4 and portions thereof were expressed as GST-fusion proteins (left), and assayed for binding to 35S-labeled FL Smad4 (right). (f) The 35S-labeled FL and truncated Smad4 were translated in vitro (left), and assayed for binding to FL GST–DLX4 protein (right). the DLX4 homeodomain to Smad4 was detected but (Figure 6c). We also tested the ability of GST–DLX4 not as strongly as observed with full-length DLX4 protein to directly bind in vitro translated 35S-labeled (Figure 5e). We also investigated which domain of portions of Sp1 (Figure 6d). DLX4 did not bind the Smad4 interacts with DLX4 by testing the ability of N-terminal domain of Sp1 (amino acids 1–557), but GST–DLX4 protein to bind in vitro translated portions bound to its C-terminal DBD (amino acids 557–778; of Smad4 protein (Figure 5d). Direct binding of DLX4 Figure 6e). In the presence of DLX4, decreased levels of was detected to the Smad4 Mad homology 1 (MH1) Sp1 were detected in complexes associated with the domain, but not to the MH2 domain (Figure 5f). p15Ink4B promoter in both oligonucleotide pulldown and ChIP assays (Figures 4e and f). In gel-shift assays, we DLX4 also directly binds Sp1 observed that DLX4 did not bind p15Ink4B promoter Because DLX4 directly binds Smad4, which also binds sequences, but prevented Sp1 from binding the promo- Sp1, we determined whether DLX4 interacts with Sp1. ter (Figure 6f). These observations are consistent with In immunoprecipitation assays using lysates of trans- the ability of DLX4 to partially inhibit activity of the fected HepG2 cells, DLX4 was found to associate with SBE-mutant p15Ink4B promoter that contains intact Sp1- Sp1 (Figure 6a). This association was not dependent binding sites (Figure 2c). Enforced expression of DLX4 on TGF-b stimulation. We confirmed the ability of also inhibited activity of a synthetic promoter that endogenous DLX4 to interact with Sp1 in immunopre- comprised tandem Sp1-binding sites (Figure 6g). Con- cipitation assays using extracts of MCF-7 cells versely, knockdown of DLX4 stimulated Sp1-driven (Figure 6b). GST–DLX4 protein directly bound to promoter activity (Figure 6g). These results indicate that in vitro translated 35S-labeled Sp1 (Figure 6c). Whereas DLX4 inhibits p15Ink4B transcription by sequestering the N-terminal domain of DLX4 did not bind Sp1, Sp1, in addition to preventing Smad4 from interacting binding by the homeodomain was strongly detected with R-Smads.

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2724

Figure 6 DLX4 binds to the DBD of Sp1. (a) Lysates were prepared from HepG2 cells as described in Figure 5a. Interaction of FLAG–DLX4 with Sp1 was detected by reciprocal immunoprecipitation (IP) using FLAG and Sp1 Abs. (b) MCF-7 cells were transfected with non-targeting short hairpin RNA (shRNA) or with DLX4 (sh90) shRNA. Endogenous DLX4 was immuno- precipitated using DLX4 Ab, and precipitates analyzed by immunoblotting using Sp1 Ab. IP using mouse immunoglobulin G was included as a negative control. (c) GST–DLX4 proteins (described in Figure 5d) were assayed for binding to 35S-labeled full-length (FL) Sp1. (d) Sp1 constructs comprising the transactivation domain (TA) and DBD. (e) The 35S-labeled FL and truncated Sp1 were translated in vitro (left), and assayed for binding to FL GST–DLX4 protein (right). (f) Gel-shift analysis using a 32P-labeled oligonucleotide containing nucleotides À88 to À64 of the p15Ink4B promoter (refer Figure 4d). Recombinant Sp1 protein was incubated with increasing amounts of in vitro translated FLAG–DLX4 and FLAG-tag. Gel-shifted DNA-bound Sp1 is indicated. (g) MDA-MB- 468 cells were cotransfected with empty vector (ÀDLX4) or DLX4, together with the Cignal Sp1 reporter construct driven by a synthetic promoter comprising tandem Sp1-binding sites (left). Sp1-driven promoter activity was likewise assayed in MCF-7 cells that were cotransfected with non-targeting shRNA or DLX4 (sh90) shRNA (right). Shown are average relative firefly luciferase activities in three independent experiments, each performed in duplicate. Values were normalized by activity of cotransfected Renilla luciferase.

Discussion the MH2 domain of Smad4 is unable to interact with R-Smads. The TGF-b cytostatic program is essential for main- Transcription factors encoded by homeobox genes are taining normal tissue homeostasis, and is tightly characterized by their helix-turn-helix DNA-binding regulated by a network of transcription factors that homeodomain (McGinnis and Krumlauf, 1992). Our include Smad proteins, Sp1 and c-myc (Siegel and findings that binding of DLX4 to Smad4 is mediated, in Massague´, 2003; Massague´, 2008). In this study, we part, through its homeodomain raises the question of report that DLX4, a homeodomain protein that is specificity. Indeed, Hoxc8 interacts with Smad1 through expressed in a broad range of malignancies, blocks the its homeodomain (Yang et al., 2000). DLX1 has been antiproliferative effect of TGF-b by counteracting key reported to bind Smad4, but its binding region and transcriptional control mechanisms of the TGF-b mechanism have not been defined (Chiba et al., 2003). cytostatic program. Although the homeodomain is conserved throughout Our studies indicate there are several mechanisms by the homeobox gene family, the specificity of family which DLX4 inactivates transcriptional control of the members for different Smads and different Smad TGF-b cytostatic program. One mechanism is by domains is striking. DLX3 binds Smad6 but not Smad4 sequestering Smad4 and preventing Smad4 from binding (Berghorn et al., 2006). Hoxc8 interacts with the MH1 R-Smads. Smad interactions might also be prevented by domain of Smad1 (Yang et al., 2000), whereas Hoxa13 binding of DLX4 to R-Smads as we observed binding, binds the MH2 domains of Smad1, Smad2 and Smad5 albeit weak, of DLX4 to Smad2. Because Smad4 and R- but does not bind Smad4 (Williams et al., 2005). Within Smads interact with one another via their MH2 domains the homeodomain, the residues of the third helix (Shi and Massague´, 2003), our finding that DLX4 are the most highly conserved. Less conserved residues binds the Smad4 MH1 domain is surprising. One ex- of the first and second helices might govern preferential planation could be that binding of DLX4 to the MH1 binding to a specific Smad protein or Smad domain induces a conformational change such that domain. Although other homeodomain proteins might

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2725 potentially block TGF-b-mediated growth inhibition by transcription and is silenced in breast cancers binding Smads, the specificity of this inhibition could (Raman et al., 2000). Overexpression of HOXB7 and largely depend on the context on their expression. HSIX1 in ovarian and breast tumors induces expression Although most homeobox genes are expressed in a of fibroblast growth factor-2 and cyclin A1, respectively highly tissue-specific manner, DLX4 is expressed across (Naora et al., 2001; Coletta et al., 2004). Aberrant diverse malignancies (Haga et al., 2000; Man et al., expression of homeobox genes in tumors is thought to 2005; Hara et al., 2007; Tomida et al., 2007; Schwartz reflect an inappropriate recapitulation of embryonic et al., 2009). No other homeobox gene has been reported pathways (Abate-Shen, 2002; Samuel and Naora, 2005). to be commonly expressed in tumors of lung, breast, The ability of DLX4 to block TGF-b signaling might be ovarian, prostate and hematologic origin. Interference related to the functions of DLX genes in controlling by DLX4 of TGF-b-mediated growth inhibition bone morphogenesis and skeletal patterning (Pangani- could therefore be a mechanism common to multiple ban and Rubenstein, 2002). These processes are tightly organ sites. regulated by members of the TGF superfamily. Because Binding affinity and selectivity of Smad complexes for DLX4 binds Smad4, DLX4 also likely blocks signaling target gene promoters are principally dictated by emanating from other receptors of the TGF super- interactions with other DNA-binding factors (Shi and family. Indeed, DLX4 inhibited induction of transcrip- Massague´, 2003, Feng and Derynck, 2005). Similarly, tional activity by BMP-4 (Figures 3g and h; Supple- binding affinity and selectivity of several homeodomain mentary Figure 5). Several cross-regulatory interactions proteins have been reported to be modulated by have been reported between BMPs and DLX interactions with other transcriptional regulators (Shen genes during normal cell differentiation. BMP-2 acti- et al., 1999; Boogerd et al., 2008). Our study is the first vates Dlx3 transcription (Park and Morasso, 2002), report that demonstrates that a homeodomain protein whereas Smad6, an antagonist of BMP signaling, directly interacts with Sp1 and modulates Sp1 activity. inhibits DLX3 transcriptional activity (Berghorn et al., Our findings indicate that DLX4 directly binds the DBD 2006). Interestingly, we observed that levels of DLX4 of Sp1 and impairs the DNA-binding ability of Sp1, but protein decreased in cells following TGF-b stimu- does not prevent Sp1 from associating with Smad4. lation (Figure 1h). DLX4 might be a component of a Because DLX4 binds to Smad4 and to Sp1, we speculate regulatory loop that blocks TGF-b signaling and is that DLX4 inhibits p15Ink4B transcription, in part, conversely regulated by TGF-b. This feedback mechan- by forming an inactive DLX4–Smad–Sp1 complex. ism might have an important role in controlling normal Because transcription of p21WAF1/Cip1 is also induced by embryogenesis and homeostasis. TGF-b via cooperative interactions between Sp1 and Resistance to TGF-b-mediated growth inhibition is Smad proteins (Pardali et al., 2000), DLX4 could likely an important feature in the pathogenesis of most types inhibit p21WAF1/Cip1 transcription by a similar mechanism. of tumors. This resistance has been attributed to TGF-b Our finding that DLX4 induces expression of c-myc receptor or Smad mutations in several types of tumors, independently of TGF-b signaling has two implications. particularly those of gastrointestinal and pancreatic Firstly, induction of c-myc provides a competing origin (Markowitz et al., 1995; Hahn et al., 1996; mitogenic signal against the TGF-b cytostatic program. Watanabe et al., 2001; Woodford-Richens et al., 2001). Secondly, DLX4-induced c-myc expression might lead Resistance of tumor cells to the antiproliferative effect to downregulated expression of p15Ink4B and p21WAF1/Cip1, of TGF-b can also stem from downregulation of TGF-b as transcription of these genes is repressed by c-myc receptor expression (Tokunaga et al., 1999) and p15Ink4B (Feng et al., 2002; Gartel et al., 2001). The repression of deletion (Gemma et al., 1996). The ability of DLX4 to p15Ink4B transcription by c-myc has a notable similarity disable key transcriptional control mechanisms of the to our observations with DLX4. Feng et al. likewise TGF-b cytostatic program could explain why tumors reported that c-myc does not compete with Sp1 for that lack aberrations in core components of the TGF-b interaction with Smads (Feng et al., 2002). These signaling pathway can become resistant to the anti- authors speculated that c-myc forms an inactive com- proliferative effect of TGF-b. plex with Sp1 and Smad proteins. However, in contrast A striking aspect of the TGF-b signaling pathway in to our findings with DLX4, c-myc does not inhibit tumors is its biphasic function. Many tumors are interactions between Smad4 and R-Smads nor inhibits resistant to the antiproliferative effect of TGF-b but transcriptional activity of R-Smads per se (Feng et al., retain other TGF-b-mediated mechanisms that promote 2002). Furthermore, unlike DLX4, c-myc does not affect EMT and metastasis (Siegel and Massague´, 2003; binding of Sp1 to the p15Ink4B promoter (Feng et al., Massague´, 2008). It has been thought that core 2002). DLX4 might therefore repress p15Ink4B and components of the TGF-b pathway remain functional p21WAF1/Cip1 transcription by increasing c-myc expression, in these tumors, whereas downstream aberrations (such and also by directly interfering with Smad4–R-Smad as p15Ink4B deletion) disable the growth-inhibitory arm of interactions and Sp1 DNA-binding activity. the pathway (Siegel and Massague´, 2003; Massague´, Homeobox genes have essential roles in controlling 2008). By sequestering Smad4, DLX4 inactivates the cell differentiation during embryonic development. core pathway and might also block the metastasis- Increasing evidence indicates that aberrant expression promoting function of TGF-b. Indeed, DLX4 markedly of specific sets of homeobox genes in tumors can but not completely inhibited TGF-b-induced EMT deregulate cell growth. For example, HOXA5 regulates in NMuMG cells (Supplementary Figures 3A–C).

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2726 The ability of DLX4 to inhibit TGF-b-induced EMT was purchased from SABiosciences (Frederick, MD, USA). could explain the association of DLX4 with favorable Firefly luciferase constructs containing wild-type and mutant prognosis in lung cancer patients and its metastasis- p15Ink4B promoter sequences (Feng et al., 2000) were provided suppressive activity reported by Tomida et al. (2007). by Xiao-Fan Wang (Duke University Medical Center) and However, DLX4 levels in ovarian and breast cancers have Xin-Hua Feng (Baylor College of Medicine). The Id1 been reported to correlate with disease progression (Man promoter construct was provided by Robert Benezra (Memor- ial Sloan-Kettering Cancer Center; Addgene plasmid 16048; et al., 2005; Hara et al., 2007). There are several possible Tournay and Benezra, 1996). The BRE-Luc reporter construct explanations for this paradox. TGF-b not only induces (Korchynskyi and ten Dijke, 2002) was provided by Peter ten EMT by Smad-dependent mechanisms, but also via Dijke (Netherlands Cancer Institute). Smad-independent pathways that involve MAP kinase and RhoA activation (Moustakas and Heldin, 2005). Antibodies and other reagents TGF-b-induced, non-Smad pathways that promote cell Sources of Abs were as follows: DLX4 (Abcam, Cambridge, migration might not be inhibited by DLX4. DLX4 could MA, USA; Abnova, Taipei, Taiwan); Smad2, phospho Smad2 also promote tumor progression by other mechanisms (Ser465/467), Smad3, Smad4, p15Ink4b and c-myc (Cell Signaling such as sustained induction of c-myc. In addition, we have Technology, Danvers, MA, USA); p21WAF1/Cip1 (Calbiochem, found that DLX4 promotes tumor angiogenesis by San Diego, CA, USA); Sp1 and E-cadherin (Zymed, San inducing expression of vascular endothelial growth factor Francisco, CA, USA); N-cadherin (BD Biosciences, San Jose, and fibroblast growth factor-2 (Hara et al., 2007). The CA, USA); actin and FLAG (Sigma-Aldrich, St Louis, MO, mechanism that gives rise to overexpression of DLX4 in USA); and Smad2/3 and lamin A/C (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Recombinant Sp1 protein, TGF-b and tumors is unclear. The DLX4 gene maps to the 17q21.3- BMP-4 were purchased from Promega, Sigma-Aldrich and q22 region, a chromosomal ‘hot spot’ that is amplified in R&D Systems (Minneapolis, MN, USA), respectively. B10% of breast and ovarian cancers (Hyman et al., 2002). However, DLX4 overexpression occurs in >50% Cell lines and transfection of these tumors (Man et al., 2005; Hara et al., 2007), Sources of cell lines were as follows: Mv1Lu, MDA-MB-468 indicating that gene amplification is not the sole mechan- and NMuMG (American Type Culture Collection, Manassas, ism underlying this overexpression. VA, USA); HepG2, Ampho-293 and MCF-7 (Michelle Like the Hedgehog, Wnt and Notch signaling path- Barton, Douglas Boyd and Francois-Xavier Claret, MD ways, the homeobox gene network is increasingly Anderson Cancer Center). For transient expression, cells were thought to be an important hub in the intimate transfected using FuGENE6 reagent (Roche, Indianapolis, IN, relationship between embryonic development and can- USA). For generating stable lines, FLAG-tagged DLX4 complementary DNA was subcloned into the pRetroQ vector cer. Developmental patterning is known to be governed (Clontech, Palo Alto, CA, USA), and the retroviral construct by interactions between homeobox genes and members was used to transfect Ampho-293 cells. Supernatants were of the TGF superfamily, but this is the first report that harvested 2 days thereafter and used to infect target cells. functionally links a homeobox gene that is aberrantly Stable lines were selected by puromycin (0.5 mg/ml). expressed in tumors with resistance to the cytostatic activity of TGF-b. Future studies of pathways con- Growth assays and cell cycle analysis trolled by this intriguing class of patterning regulators Cells (4000 per well) were seeded in 96-well plates, and cultured will provide important insights into the multiple steps of for 2 days in complete medium containing TGF-b at the neoplastic process. concentrations ranging from 0 to 100 ng/ml. Proliferation was measured by the 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetra- zolium bromide (MTT) assay (Roche). For cell cycle analysis, Materials and methods cells were seeded in 10 cm dishes to reach 30% confluence the following day. Cells were serum starved overnight and then Plasmids cultured in complete medium with and without addition of The DLX4 complementary DNA is described elsewhere (Haga TGF-b for 18 h. Cells were harvested, fixed in 70% ethanol, and et al., 2000; Hara et al., 2007). DLX4 fragments were distribution throughout the cell cycle was determined by flow subcloned into pET41 GST vectors (Novagen, Gibbstown, cytometric analysis of propidium iodide staining. NJ, USA) as described in the text. DLX4 and non-targeting short hairpin RNAs were purchased from OriGene Technol- Reporter assays ogy (Rockville, MD, USA). GST–Smad2 and GST–Smad4 Cells were plated in 12-well plates and cotransfected with ex- plasmids were provided by Fang Liu (Rutgers University). pression plasmid (400 ng), firefly luciferase reporter plasmid Smad complementary DNAs were purchased from OriGene (100 ng) and pRL-CMV Renilla luciferase reporter plasmid Technology and subcloned into the pFA-CMV plasmid (0.5 ng; Promega) for normalizing transfection efficiency. At 24 h containing the GAL4 DBD (Stratagene, La Jolla, CA, USA) after transfection, cells were cultured for an additional 18 h as described in the text. The GAL4-driven pRF-Luc reporter without and with TGF-b or BMP-4. Luciferase activities were construct was purchased from Stratagene. The pGL2 firefly assayed using the Dual-luciferase reporter assay kit (Promega). luciferase reporter vector was purchased from Promega (Madison, WI, USA). The pBV-Luc, pSBE4-Luc and pBV- Immunofluorescence staining MYC(Del4) reporter constructs (He et al., 1998; Zawel et al., Cells were fixed in 4% paraformaldehyde and permeabilized in 1998) were provided by Bert Vogelstein (Johns Hopkins 0.1% Triton X-100. Cells were stained for 1 h with Abs to University; Addgene plasmids 16539, 16495 and 16604). Sp1 Smad2, DLX4, E-cadherin or FLAG (1:200), and staining complementary DNA was provided by Keping Xie (MD detected by Alexa Fluor 594-conjugated secondary Ab Anderson Cancer Center). The Cignal Sp1 reporter construct (Invitrogen, Carlsbad, CA, USA).

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2727 Immunoprecipitation and immunoblotting from the T7 promoter, and incubated for 2 h with 1 mgof Whole-cell extracts were prepared by lysing cells in M-PER GST-fusion protein bound to glutathione-sepharose beads buffer (Pierce Biotechnology, Rockford, IL, USA). Nuclear (GE Healthcare, Piscataway, NJ, USA). Beads were washed extracts were prepared by lysing cells in cell lysis buffer (10 mM with binding buffer (20 mM Tris–HCl, pH 8.0, 100 mM NaCl, HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), 0.1% NP-40 and 2 mM EDTA). Associated proteins were pH 7.9, 1.5 mM MgCl2,10mM KCl, 1 mM dithiothreitol plus subjected to SDS–polyacrylamide gel electrophoresis and protease and phosphatase inhibitors), followed by centrifuga- visualized by autoradiography. tion at 800 g for 10 min. Nuclear pellets were lysed in nuclear buffer (20 mM Tris–HCl, pH 8.0, 100 mM NaCl, 1% nonidet Gel-shift assay P-40 (NP-40), 10% glycerol and 2 mM EDTA), centrifuged at 12 000 g for 10 min and supernatants collected. Nuclear Recombinant Sp1 (100 ng) was incubated with increasing extracts were precleared with protein G-agarose. In all, amounts of in vitro translated FLAG–DLX4 and FLAG-tag at 1 500 mg of nuclear extract was incubated with Ab for 4–12 h 23 C for 20 min in binding buffer (10 mM Tris–HCl, pH 7.5, at 4 1C. Immunoprecipitates were washed and subjected to 1mM MgCl2,50mM NaCl, 0.5 mM EDTA, 4% glycerol, 0.5 mM SDS–polyacrylamide gel electrophoresis and immunoblotting. dithiothreitol and 1 mg poly(dI-dC).poly(dI-dC). In all, 1 ng of 32P-labeled oligonucleotide containing sequences À88 to À64 of the p15Ink4B promoter (Figure 4d) was added to the binding Oligonucleotide pulldown assay reaction and incubated for an additional 20 min. DNA–protein Nuclear extracts were precleared with streptavidin agarose for complexes were separated by polyacrylamide gel electrophor- 30 min at 4 1C and incubated overnight with biotinylated esis and visualized by autoradiography. oligonucleotide corresponding to nucleotides À108 to À39 of the p15Ink4B promoter (Figure 4d). Beads were washed with binding buffer (10 mM HEPES, pH 7.9, 100 mM KCl, 5 mM MgCl2,1mM EDTA, 10% glycerol, 1 mM dithiothreitol Abbreviations and 0.5% NP-40). DNA-bound proteins were eluted and subjected to SDS–polyacrylamide gel electrophoresis and Ab, antibody; BMP, bone morphogenetic protein; DBD, immunoblotting. DNA-binding domain; EMT, epithelial-to-mesenchymal transition; F-Luc, firefly luciferase; IP, immunoprecipitation ChIP R-Luc, Renilla luciferase; R-Smad, receptor-regulated Smad; ChIP assays were performed using the ChIP Assay kit TGF-b, transforming growth factor-b. (Upstate Biotechnology, Temecula, CA, USA) following the manufacturer’s protocol with modifications. Cells were cross-linked using 1% formaldehyde and then neutralized Conflict of interest with glycine. Cells were lysed in SDS lysis buffer and sonicated to generate DNA fragments of 200–1000 bp in length. Sheared Dr Naora’s work has been funded by the NIH and US chromatin was precleared with protein G-agarose and incu- Department of Defense. The remaining authors declare no bated overnight with 4 mg of Sp1 and Smad Abs. Protein– conflict of interest. DNA complexes were precipitated with protein G-agarose, followed by elution of immunoprecipitated complexes and reversal of cross-links. Purified DNA was used in PCR reactions to amplify a 535 bp fragment of the p15Ink4B promoter Acknowledgements using the following primers: sense, 50-TATGGTTGACTAA 0 0 TTCAAACAG-3 ; antisense, 5 -GCAAAGAATTCCGTTTT This work was supported by a Schissler Foundation Fellow- 0 CAGCT-3 . The amplified fragment was confirmed by DNA ship (B Trinh), the Vietnam Education Foundation (B Trinh), sequencing. the US Department of Defense grant W81XWH-06-1-0259 (H Naora) and the National Institutes of Health grant R01 In vitro binding assays CA141078 (H Naora). We thank Sabine Thonard for technical GST–DLX4 and GST–Smad fusion proteins were produced in assistance, and Song Yi Ko, Gary Gallick, Michelle Barton, Escherichia Coli. The 35S-labeled DLX4, Smad4 and Sp1 were Janet Price, Peng Huang and Miles Wilkinson (MD Anderson synthesized by in vitro transcription/translation (Promega) Cancer Center) for helpful discussions.

References

Abate-Shen C. (2002). Deregulated homeobox in Chiba S, Takeshita K, Imai Y, Kumano K, Kurokawa M, Masuda S cancer: cause or consequence? Nat Rev Cancer 2: 777–785. et al. (2003). Homeoprotein DLX-1 interacts with Smad4 and Bello-DeOcampo D, Tindall DJ. (2003). TGF-b/Smad signaling in blocks a signaling pathway from activin A in hematopoietic cells. prostate cancer. Curr Drug Targets 4: 197–207. Proc Natl Acad Sci USA 100: 15577–15582. Berghorn KA, Clark-Campbell PA, Han L, McGrattan M, Weiss RS, Coletta RD, Christensen K, Reichenberger KJ, Lamb J, Micomonaco Roberson MS. (2006). Smad6 represses Dlx3 transcriptional activity D, Huang L et al. (2004). The Six1 homeoprotein stimulates through inhibition of DNA binding. J Biol Chem 281: 20357–20367. tumorigenesis by reactivation of cyclin A1. Proc Natl Acad Sci USA Boogerd KJ, Wong LY, Christoffels VM, Klarenbeek M, Ruijter JM, 101: 6478–6483. Moorman AF et al. (2008). Msx1 and Msx2 are functional DeckersM,vanDintherM,BuijsJ,QueI,Lo¨wik C, van der Pluijm G interacting partners of T-box factors in the regulation of connex- et al. (2006). The tumor suppressor Smad4 is required for transforming in43. Cardiovasc Res 78: 485–493. growth factor-b-induced epithelial-to-mesenchymal transition and Chen CR, Kang Y, Siegel PM, Massague´J. (2002). /5 and p107 bonemetastasisofbreastcancercells.Cancer Res 66: 2202–2209. as Smad cofactors linking the TGFb receptor to c-myc repression. Feng XH, Derynck R. (2005). Specificity and versatility in TGF-b Cell 110: 19–32. signaling through Smads. Annu Rev Cell Dev Biol 21: 659–693.

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2728 Feng XH, Liang YY, Liang M, Zhai W, Lin X. (2002). p21Waf1/Cip1 regulation by transforming growth factor-b. J Biol Chem Direct interaction of c-Myc with Smad2 and Smad3 to inhibit 275: 29244–29256. TGF-b-mediated induction of the CDK inhibitor p15Ink4B. Mol Cell Park GT, Morasso MI. (2002). Bone morphogenetic protein-2 (BMP- 9: 133–143. 2) transactivates Dlx3 through Smad1 and Smad4: alternative mode Feng XH, Lin X, Derynck R. (2000). Smad2, Smad3 and Smad4 for Dlx3 induction in mouse keratinocytes. Nucleic Acids Res 30: cooperate with Sp1 to induce p15Ink4B transcription in response to 515–522. TGF-b. EMBO J 19: 5178–5193. Petritsch C, Beug H, Balmain A, Oft M. (2000). TGF-b inhibits p70 S6 Francis-Thickpenny KM, Richardson DM, van Ee CC, Love DR, kinase via protein phosphatase 2A to induce G1 arrest. Genes Dev Winship IM, Baguley BC et al. (2001). Analysis of the transforming 14: 3093–3101. growth factor-b functional pathway in epithelial ovarian carcinoma. Raman V, Martensen SA, Reisman D, Evron E, Odenwald WF, Br J Cancer 85: 687–691. Jaffee E et al. (2000). Compromised HOXA5 function can Gartel AL, Ye X, Goufman E, Shianov P, Hay N, Najmabadi F et al. limit p53 expression in human breast tumors. Nature 405: 974–978. (2001). Myc represses the p21WAF1/Cip1 promoter and interacts with Reynisdo´ttir I, Polyak K, Iavarone A, Massague´J. (1995). Kip/Cip Sp1/Sp3. Proc Natl Acad Sci USA 98: 4510–4515. and Ink4 Cdk inhibitors cooperate to induce cell cycle arrest in Gemma A, Takenoshita S, Hagiwara K, Okamoto A, Spillare EA, response to TGF-b. Genes Dev 9: 1831–1845. McMemamin MG et al. (1996). Molecular analysis of the cyclin- Samuel S, Naora H. (2005). Homeobox gene expression in cancer: dependent kinase inhibitor genes p15INK4B/MTS2, p16INK4/MTS1, p18 and insights from developmental regulation and deregulation. Eur J p19 in human cancer cell lines. Int J Cancer 68: 605–611. Cancer 41: 2428–2437. Haga SB, Fu S, Karp JE, Ross DD, Williams DM, Hankins WD et al. Schutte M, Hruban RH, Hedrick L, Cho KR, Nadasdy GM, (2000). BP1, a new homeobox gene, is frequently expressed in acute Weinstein CL et al. (1996). DPC4 gene in various tumor types. leukemias. Leukemia 14: 1867–1875. Cancer Res 56: 2527–2530. Hahn SA, Schutte M, Hoque AT, Moskaluk CA, da Costa LT, Schwartz AM, Man YG, Rezaei MK, Simmens SJ, Berg PE. (2009). Rozenblum E et al. (1996). DPC4, a candidate tumor suppressor at BP1, a homeoprotein, is significantly expressed in prostate human 18q21.1. Science 271: 350–353. adenocarcinoma and is concordant with prostatic intraepithelial Hara F, Samuel S, Liu J, Rosen D, Langley R, Naora H. (2007). neoplasia. Mod Pathol 22: 1–6. A homeobox gene related to Drosophila Distal-less promotes Shen WF, Rozenfeld S, Kwong A, Ko¨m ves LG, Lawrence HJ, ovarian tumorigenicity by inducing expression of vascular endothelial Largman C. (1999). HOXA9 forms triple complexes with PBX2 and growth factor and fibroblast growth factor-2. Am J Pathol 170: MEIS1 in myeloid cells. Mol Cell Biol 19: 3051–3061. 1594–1606. Shi Y, Massague´J. (2003). Mechanisms of TGF-b signaling from cell He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Costa LT membrane to the nucleus. Cell 113: 685–700. et al. (1998). Identification of c-MYC as a target of the APC Siegel PM, Massague´J. (2003). Cytostatic and apoptotic actions of pathway. Science 281: 1509–1512. TGFb in homeostasis and cancer. Nat Rev Cancer 3: 807–820. Hyman E, Kauraniemi P, Hautaniemi S, Wolf M, Mousses S, Takenoshita S, Hagiwara K, Gemma A, Nagashima M, Ryberg D, Rozenblum E et al. (2002). Impact of DNA amplification on gene Lindstedt BA et al. (1997). Absence of mutations in the transform- expression patterns in breast cancer. Cancer Res 62: 6240–6245. ing growth factor b type II receptor in sporadic lung cancers with Kang Y, Chen CR, Massague´J. (2003). A self-enabling TGFb microsatellite instability and rare H-ras1 alleles. Carcinogenesis 18: response coupled to stress signaling: Smad engages stress 1427–1429. response factor ATF3 for Id1 repression in epithelial cells. Mol Takenoshita S, Mogi A, Tani M, Osawa H, Sunaga H, Kakegawa H Cell 11: 915–926. et al. (1998). Absence of mutations in the analysis of coding Korchynskyi O, ten Dijke P. (2002). Identification and functional sequences of the entire transforming growth factor-b type II characterization of distinct critically important bone morphogenetic receptor gene in sporadic human breast cancers. Oncol Rep 5: protein-specific response elements in the Id1 promoter. J Biol Chem 367–371. 277: 4883–4891. Tokunaga H, Lee DH, Kim IY, Wheeler TM, Lerner SP. (1999). Kusanagi K, Inoue H, Ishidou Y, Mishima H, Kawabata M, Decreased expression of transforming growth factor b receptor type Miyazono K. (2000). Characterization of a bone morpho- I is associated with poor prognosis in bladder transitional cell genetic protein-responsive Smad-binding element. Mol Cell Biol 11: carcinoma patients. Clin Cancer Res 5: 2520–2525. 555–565. Tomida S, Yanagisawa K, Koshikawa K, Yatabe Y, Mitsudomi T, ManYG,FuSW,SchwartzA,PinzoneJJ,SimmensSJ,BergPE. Osada H et al. (2007). Identification of a metastasis signature (2005). Expression of BP1, a novel homeobox gene, correlates with and the DLX4 homeobox protein as a regulator of metastasis breast cancer progression and invasion. Breast Cancer Res Treat 90: by a combined transcriptome approach. Oncogene 26: 241–247. 4600–4608. Massague´J. (2008). TGF-b in cancer. Cell 134: 215–230. Tournay O, Benezra R. (1996). Transcription of the dominant-negative Markowitz S, Wang J, Myeroff L, Parsons R, Sun L, Lutterbaugh J helix-loop-helix protein Id1 is regulated by a protein complex et al. (1995). Inactivation of the type II TGF-b receptor in colon containing the immediate-early response gene Egr-1. Mol Cell Biol cancer cells with microsatellite instability. Science 268: 1336–1338. 16: 2418–2430. McGinnis W, Krumlauf R. (1992). Homeobox genes and axial Wang D, Kanuma T, Mizunuma H, Takama F, Ibuki Y, Wake N et al. patterning. Cell 68: 283–302. (2000). Analysis of specific gene mutations in the transforming Moustakas A, Heldin CH. (2005). Non-Smad TGF-b signals. growth factor b signal transduction pathway in human ovarian J Cell Science 118: 3573–3584. cancer. Cancer Res 60: 4507–4512. Nagatake M, Takagi Y, Osada H, Uchida K, Mitsudomi T, Saji S Watanabe T, Wu TT, Catalano PJ, Ueki T, Satriano R, Haller DG et al. (1996). Somatic in vivo alterations of the DPC4 gene at 18q21 et al. (2001). Molecular predictors of survival after in human lung cancers. Cancer Res 56: 2718–2720. adjuvant chemotherapy for colon cancer. N Engl J Med 344: Naora H, Yang YQ, Montz FJ, Seidman JD, Kurman RJ, Roden RB. 1196–1206. (2001). A serologically identified tumor antigen encoded by a Williams TM, Williams ME, Heaton JH, Gelehrter TD, Innis JW. homeobox gene promotes growth of ovarian epithelial cells. Proc (2005). Group 13 HOX proteins interact with the MH2 domain of Natl Acad Sci USA 98: 4060–4065. R-Smads and modulate Smad transcriptional activation functions Panganiban G, Rubenstein JL. (2002). Developmental functions of the independent of HOX DNA-binding capability. Nucleic Acids Res Distal-less/Dlx homeobox genes. Development 129: 4371–4386. 33: 4475–4484. Pardali K, Kurisaki A, More´n A, ten Dijke P, Kardassis D, Moustakas Woodford-Richens KL, Rowan AJ, Gorman P, Halford S, Bicknell A. (2000). Role of Smad proteins and transcription factor Sp1 in DC, Wasan HS et al. (2001). SMAD4 mutations in colorectal cancer

Oncogene Homeodomain protein DLX4 blocks TGF-b signaling BQ Trinh et al 2729 probably occur before chromosomal instability but after divergence Yang X, Ji X, Shi X, Cao X. (2000). Smad1 domains interacting of the microsatellite instability pathway. Proc Natl Acad Sci USA with Hoxc-8 induce osteoblast differentiation. J Biol Chem 275: 98: 9719–9723. 1065–1072. Yamada SD, Baldwin RL, Karlan BY. (1999). Ovarian carcinoma cell Zawel L, Dai JL, Buckhaults P, Zhou S, Kinzler KW, Vogelstein B cultures are resistant to TGF-b1-mediated growth inhibition despite et al. (1998). Human Smad3 and Smad4 are sequence-specific expression of functional receptors. Gynecol Oncol 75: 72–77. transcription activators. Mol Cell 1: 611–617.

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

Oncogene