Cancer Research CTRC-AACR San Antonio Breast Cancer Symposium Translational Medicine Conference in Israel
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Cancer Research 67, 75-84, January 1, 2007. doi: 10.1158/0008-5472.CAN-06-2559
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Labbé, E.
Right arrow Articles by Attisano, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Labbé, E.
Right arrow Articles by Attisano, L.

Molecular Biology, Pathobiology, and Genetics

Transcriptional Cooperation between the Transforming Growth Factor-ß and Wnt Pathways in Mammary and Intestinal Tumorigenesis

Etienne Labbé1, Lisa Lock2, Ainhoa Letamendia2, Agnieszka E. Gorska4, Robert Gryfe3, Steven Gallinger3, Harold L. Moses4 and Liliana Attisano1,2

Departments of 1 Medical Biophysics and 2 Biochemistry, University of Toronto; 3 Department of Surgery, Samuel Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, Ontario, Canada; and 4 Department of Cancer Biology, Vanderbilt-Ingram Cancer Center, Nashville, Tennessee

Requests for reprints: Liliana Attisano, Department of Biochemistry, Terrence Donnelly Centre for Cellular and Biomolecular Research, Room 1008, 160 College Street, University of Toronto, Toronto, ON, Canada M5S 3E1. Phone: 416-946-3129; Fax: 416-978-8548; E-mail: liliana.attisano{at}utoronto.ca.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Transforming growth factor-ß (TGF-ß) and Wnt ligands function in numerous developmental processes, and alterations of both signaling pathways are associated with common pathologic conditions, including cancer. To obtain insight into the extent of interdependence of the two signaling cascades in regulating biological responses, we used an oligonucleotide microarray approach to identify Wnt and TGF-ß target genes using normal murine mammary gland epithelial cells as a model. Combination treatment of TGF-ß and Wnt revealed a novel transcriptional program that could not have been predicted from single ligand treatments and included a cohort of genes that were cooperatively induced by both pathways. These included both novel and known components or modulators of TGF-ß and Wnt pathways, suggesting that mutual feedback is a feature of the coordinated activities of the ligands. The majority of the cooperative targets display increased expression in tumors derived from either Min (many intestinal neoplasia) or mouse mammary tumor virus (MMTV)–Wnt1 mice, two models of Wnt-induced tumors, with nine of these genes (Ankrd1, Ccnd1, Ctgf, Gpc1, Hs6st2, IL11, Inhba, Mmp14, and Robo1) showing increases in both. Reduction of TGF-ß signaling by expression of a dominant-negative TGF-ß type II receptor in bigenic MMTV-Wnt1/DNIIR mice increased mammary tumor latency and was correlated with a decrease in expression of Gpc1, Inhba, and Robo1, three of the TGF-ß/Wnt cooperative targets. Our results indicate that the TGF-ß and Wnt/ß-catenin pathways are firmly intertwined and generate a unique gene expression pattern that can contribute to tumor progression. [Cancer Res 2007;67(1):75–84]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Growth factors belonging to the transforming growth factor-ß (TGF-ß) and Wnt family of ligands are involved in the numerous aspects of embryonic development and are important regulators of organ homeostasis in postnatal physiology (1, 2). TGF-ß factors initiate signal transduction via cell surface serine/threonine kinase receptors that directly phosphorylate receptor-regulated Smad proteins. Phosphorylated receptor-regulated Smads complex with Smad4 and then translocate to the nucleus and interact with DNA-bound transcription factors and other transcriptional regulators to control gene expression (3). Signaling downstream of Wnt ligands in the canonical Wnt/ß-catenin pathway involves stabilization of cytoplasmic ß-catenin, a protein that is normally targeted for ubiquitin-mediated degradation via the action of a destruction complex composed of proteins including adenomatous polyposis coli (APC), Axin, and glycogen synthase kinase 3ß. Wnt binding to the Frizzled and LRP5/6 receptors inhibits ß-catenin degradation, leading to ß-catenin nuclear accumulation and activation of gene transcription in association with the DNA-binding lymphoid enhancer factor (LEF)/T-cell factor (TCF) proteins and other cofactors (2).

TGF-ß and Wnt ligands regulate a number of common processes during embryonic development, such as patterning of imaginal discs in Drosophila and tissue specification and organogenesis in vertebrate embryos (2). These pathways can coordinate developmental events by regulating the expression of target genes such as cerberus, chordin, crescent, goosecoid, noggin, siamois, and Xtwn in the Xenopus Spemann organizer, and Id2 and Msx1 in mammalian cells (4, 5). Further, mutations affecting the integrity of both the Wnt and TGF-ß pathways in human malignancies have been extensively reported. For instance, mutations in the human TGF-ß type II receptor, Smad2, and Smad4 occur in colon tumors; Smad4 mutations are found in a high percentage of pancreatic cancers (6). Activation of the Wnt/ß-catenin pathway through a variety of mechanisms is also widespread in human tumors. For instance, in sporadic colorectal cancers, biallelic loss of APC occurs in up to 85% of cases, with most remaining cases harboring ß-catenin mutations (2). Although mutations in APC, Axin, and ß-catenin are rare in human breast cancers, accumulation of nuclear ß-catenin, a hallmark of Wnt pathway activation, is seen in ~60% of clinical breast cancer samples (7). The most common mechanisms of Wnt pathway activation in breast cancers involve loss of expression of secreted frizzled-related protein 1 and Wnt inhibitory factor 1 through promoter methylation and overexpression of Wnt ligands (8–10). In addition, a number of signaling pathways with known alterations in breast cancers can affect ß-catenin function. These include the epidermal growth factor, insulin-like growth factor, integrin-linked kinase, and Akt pathways, as well as the tumor suppressors p53 and PTEN (10).

Despite an established role in suppressing tumor formation, the TGF-ß pathway can also act as a promoter of tumor invasion at later stages of tumorigenesis (11). Whereas most tumors of epithelial origin lose the ability to be growth inhibited by TGF-ß through various molecular or mutational mechanisms, retention of a partially functional pathway favors a more aggressive tumor behavior and consequently a less favorable prognosis (11). TGF-ß has been shown to synergize with oncogenic pathways in various systems. For example, transgenic mice overexpressing the HER2/Neu oncogene and TGF-ß1 under the control of the MMTV promoter have increased invasiveness and a higher frequency of lung metastasis compared with single transgenics (12). Similarly, evidence suggests that Wnts can also cooperate with other signaling pathways during tumorigenesis. For instance, in MMTV-Wnt1 mice, targeted deletion of the estrogen receptor {alpha} delays tumor formation, and bigenic mice carrying fibroblast growth factor-3 driven by MMTV and MMTV-Wnt1 develop tumors faster than mice with either transgene alone (13, 14). Interestingly, MMTV-Wnt1 mice crossed with MMTV-TGF-ß1 mice show no alterations in tumor formation,5 suggesting that TGF-ß cannot inhibit the growth of epithelial cells stimulated by Wnt1. We and others have previously described a mechanism for cooperative transcriptional regulation by mediators of the TGF-ß and Wnt pathways (15). Synergistic gene induction involves direct physical interaction between Smads and LEF/TCF proteins coincident with DNA binding by the complex to target promoters such as Xtwin, Msx2, and Gastrin (16–18). Despite strong evidence that TGF-ß and Wnt factors cooperatively regulate developmental and homeostatic processes, little is known about the specific transcriptional targets that are controlled by a combination of these ligands and whether a direct interplay of these two pathways regulates the various stages of tumorigenesis.

In this study, we have defined a novel transcriptional program induced by TGF-ß and Wnt that could not have been predicted by the analysis of the single ligand treatments alone. Further, we show that many of these cooperative targets display increased expression in tumors, suggesting that these two pathways may cooperate to promote tumor development.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Production of Wnt3a conditioned medium. Mouse L cells stably expressing mouse Wnt3a were generated using pPGK-neo-Wnt3a (19). Control or Wnt3a conditioned medium was collected from confluent cells cultured for 3 days in DMEM with 0.2% fetal bovine serum (FBS), centrifuged to remove cell debris, and stored at 4°C for up to 3 months. Human colorectal cell lines LS1034, HT29, and SW480 were grown in {alpha}-MEM with 10% FBS.

Constructs, nuclear accumulation assays, and immunoblotting. The Smad-binding elements (SBE)-TOP plasmid was constructed by inserting two SBEs upstream of the LEF/TCF binding sites in the pGL3-OT version of TOPFLASH (20). For reporter assays, normal murine mammary gland (NMuMG) cells were transfected using LipofectAMINE (Invitrogen, Carlsbad, CA) as previously described (16). LS1034 cells were transfected using Fugene 6 according to the manufacturer's instructions (Roche Diagnostics, Indianapolis, IN). For nuclear accumulation assays, NMuMG cells were incubated in control conditioned medium for 20 h, and TGF-ß or Wnt3a was added for the indicated times. The cells were fixed, permeabilized, and proteins were detected using a mouse monoclonal anti–ß-catenin antibody (1:1,000; BD Transduction Laboratories, BD Biosciences, San Jose, CA) or rabbit polyclonal anti-Smad4 antibody (1:500; Santa Cruz Biotechnology, Santa Cruz, CA) followed by a FITC-conjugated secondary antibody (Jackson ImmunoResearch, West Grove, PA) and 4',6-diamidino-2-phenylindole staining. Nuclear versus cytoplasmic fluorescence was quantitated using the ArrayScan II reader and Cellomics Nuclear Translocation software (Cellomics, Pittsburgh, PA). Phosphorylated Smad2/3 was detected by Western blotting using rabbit polyclonal anti–phospho-Smad2 antibody (1:2,000; Cell Signaling Technology, Danvers, MA).

Microarray processing and analysis. RNA was extracted from NMuMG cells with TRI reagent (Sigma-Aldrich, St. Louis, MO) then purified using the RNeasy kit (Qiagen, Valencia, CA). Indirect Cy3/Cy5 labeling of cDNAs, hybridization, and array processing was done using standard protocols6 with in-house pair-spotted 70-mer oligonucleotide arrays (SLRI_MO7K) representing 6,868 mouse genes from the UniGene database (Unigene Mm build 83; Array-Ready Oligo set, OPERON). Scanning and analysis was done with the GenePix 4000B scanner and software (Axon Instruments, Molecular Devices, Sunnyvale, CA). For analysis, fluorescence data were mean-centered and a ligand/control expression ratio was computed and log2 transformed. The data were sorted by decreasing or increasing median ratio to generate induced and repressed lists, respectively. Data were obtained for six independent experiments with reciprocal Cy3/Cy5 labeling for a total of 12 hybridizations per condition. For the final analysis, all experiments were used for individual Wnt3a and TGF-ß treatments, whereas seven hybridizations were used for the combined Wnt3a+TGF-ß treatment.

Mouse and human tumor samples. Min (many intestinal neoplasia; APC+/–) mice bred on the C57BL/6J strain were obtained from The Jackson Laboratory (Bar Harbor, ME). Representative colon and small intestinal polyps and a matched normal sample were macroscopically harvested at 5 months of age.

The MMTV-Wnt1 transgenic mice, obtained from H.E. Varmus (Program in Cancer Biology and Genetics, Memorial Sloan-Kettering Cancer Center, New York, NY), were bred to MMTV/DNIIR animals (14) to generate MMTV-Wnt1/DNIIR mice. The Wnt1 transgene was detected by Southern blot (14) or PCR, whereas the DNIIR transgene was detected by PCR as previously described (21). Virgin Wnt1 heterozygotes (24 females) and multiparous Wnt1/DNIIR heterozygotes (19 females) were used in the tumor study. Mammary tumors were removed 3 to 4 weeks after detection.

Human familial adenomatous polyposis (FAP) tumor samples were collected from eight clinically affected individuals with FAP from separate kindreds. Informed consent for tissue use was obtained from each patient as per local Institutional Review Board approval. Matched normal (six samples) and adenomatous tissues (eight samples) were dissected from fresh surgical colectomy specimens.

Mouse and human tissues were fixed in 4% paraformaldehyde or 10% buffered formalin, embedded in paraffin for sectioning, or frozen at –80°C in RNAlater (Ambion, Austin, TX), and subsequently processed for RNA extraction with TRI reagent (Sigma-Aldrich).

Real-time quantitative reverse transcription-PCR analysis. Reverse transcription was done with random hexamers on DNaseI-treated total RNA with Superscript II (Invitrogen). Real-time PCR was done using the ABI/Prism 7000 Sequence Detection System (Applied Biosystems) and the Brilliant SYBR Green QPCR master mix (Stratagene, La Jolla, CA). Primer pairs (Supplementary Table S3) designed using Primer Express (Applied Biosystems) were used at 100 nmol/L and were validated against hypoxanthine phosphoribosyltransferase (HPRT) or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) by comparative standard curve analysis. Relative quantitation was calculated by the {Delta}{Delta}Ct method normalized to HPRT or GAPDH.

Immunohistochemistry, in situ hybridization, and cancer profiling array hybridization. Paraffin-embedded tissue sections (7 µm) were prepared using standard protocols. Immunohistochemical staining was done using the horseradish peroxidase-AEC kit (R&D Systems, Minneapolis, MN) with mouse monoclonal anti–ß-catenin (1:1,000; BD Transduction Laboratories) or rabbit anti–phospho-Smad2 (1:500; Cell Signaling Technology) and counterstained with Hematoxylin QS (Vector Laboratories, Burlingame, CA). Digoxigenin-labeled RNA probes (Supplementary Table S3) for in situ hybridizations were synthesized from sequence-verified mouse cDNAs cloned from NMuMG into pBluescript(KS). Hybridizations were done using standard protocols. A pBluescript(KS)-InhibinßA human cDNA construct was used to synthesize a [32P]dATP-labeled DNA probe with the Rediprime II DNA labeling kit (Amersham Biosciences). The probe was hybridized to the Cancer Profiling Array II (Clontech, Mountain View, CA) according to the manufacturer's instructions, scanned with the STORM phosphorimager reader, and analyzed with ImageQuant (Molecular Dynamics, Sunnyvale, CA).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Identification and characterization of a Wnt3a- and TGF-ß–responsive cell line. To understand how the TGF-ß and Wnt signaling pathways cooperate to alter gene expression in mammalian cells, we first surveyed a variety of cell lines for their ability to respond to both pathways and thereby identified NMuMG epithelial cells. Stimulation of NMuMGs with Wnt3a and TGF-ß resulted in nuclear accumulation of ß-catenin and Smad4 (Fig. 1A ), and activated the TOPFLASH and 3TP-lux promoters, respectively (Fig. 1B, left). We previously showed that Smads interact with LEF1 to synergistically activate promoter constructs containing Smad and LEF/TCF binding sites (16). Consistent with this, costimulation with TGF-ß and Wnt3a synergistically activated the SBE-TOPFLASH reporter in which Smad-binding elements were introduced upstream of the LEF/TCF sites of TOPFLASH (Fig. 1B). Together, these results confirm that TGF-ß and Wnt3a induce nuclear accumulation of their respective transducers and cooperatively activate specific transcriptional responses in NMuMG cells.


Figure 1
View larger version (37K):
[in this window]
[in a new window]

  		Figure 1. Identification of TGF-ß and Wnt cooperatively induced genes in NMuMG cells. A, nuclear accumulation of endogenous ß-catenin in response to Wnt3a (left) or Smad4 in response to TGF-ß (right) was examined in NMuMG cells using a quantitative immunofluorescence assay. Points and columns, mean nuclear-to-cytoplasmic ratio of arbitrary fluorescence units in 100 cells from three wells; bars, SD. B, NMuMG epithelial cells were transiently transfected with a luciferase reporter plasmid responsive to TGF-ß (3TP-lux), Wnt3a (TOPFLASH), or both ligands (SBE-TOP). Transfected cells were incubated overnight in the presence of 100 pmol/L TGF-ß and/or Wnt3a conditioned medium. Luciferase activity was normalized using a cotransfected ß-galactosidase reporter. Columns, mean of quadruplicates; bars, SD. C, top 124 induced and repressed genes from each ligand treatment were sorted by decreasing fold change and organized in Venn groups according to responsiveness to Wnt3a, TGF-ß, and a combination of Wnt3a and TGF-ß. D, candidate Wnt3a+TGF-ß cooperative target genes obtained from the microarray analysis were validated by real-time quantitative reverse transcription-PCR using RNAs obtained from ligand-treated NMuMG cells for the indicated times.

 
Wnt3a and TGF-ß cooperate to induce expression of a subset of target genes. We next examined the effects of TGF-ß and Wnt3a on global gene expression patterns following 16-h treatment of NMuMG cells with either ligand alone or in combination. RNA was isolated and processed with control-treated samples for pairwise hybridizations to oligonucleotide microarrays composed of ~7,000 known genes. Lists of regulated genes for each condition were determined (Supplementary Table S1) as described in Materials and Methods.

Treatment of cells with TGF-ß alone resulted in the regulation of novel as well as known targets, such as Bmp1, Ctgf, Serpine1/plasminogen activator inhibitor-1 (PAI-1), and Edn1 (22–25). Similarly, genes regulated by Wnt3a included both novel (Hs3st1, SerpinB9, Ccna2, Sgk) and previously characterized targets such as Ahr, Axin2, Spp1, Ccnd1, and Thbs1 (26–30). These results confirm that NMuMG cells respond appropriately to TGF-ß and Wnt3a and show the validity of our approach. Interestingly, cells costimulated with TGF-ß and Wnt3a displayed an expression profile distinct from that observed with single ligand treatments (Fig. 1C). Thus, simultaneous receipt of TGF-ß and Wnt signals yields a gene expression profile that is not simply an additive effect of individual ligand treatments.

Validation of TGF-ß and Wnt3a-induced target genes by real-time quantitative reverse transcription-PCR. Given the limited sensitivity range typical of microarrays and the need for secondary validation, we next examined the expression of candidate targets using real-time quantitative PCR, focusing on induced genes. We selected the first 25 to 30 highest ranked genes from each treatment condition (Wnt3a, TGF-ß, and Wnt3a+TGF-ß) representing a total of 60 distinct genes. RNA was isolated from NMuMG cells treated with ligands for 2, 4, 8, or 16 h and was used to determine gene expression levels by quantitative PCR (Fig. 1D; Supplementary Table S2). Of 59 genes successfully amplified, 53 displayed ≥2-fold Wnt3a- and/or TGF-ß–dependent regulation, representing a 90% validation rate. Consistent with the use of a 16-h ligand treatment for our microarray screen, many of the regulated genes displayed maximal activation at the 8 or 16 h time points. However, a few genes were induced early such as known direct targets of TGF-ß (Ctgf) and Wnt (Axin2; Fig. 1D; Supplementary Table S2). Our quantitative PCR analysis of gene expression in cells treated simultaneously with TGF-ß and Wnt3a revealed patterns consistent with the microarray data. Of particular interest, a number of genes were additively or cooperatively induced when both TGF-ß and Wnt3a were used together (Fig. 1D; Supplementary Table S2). Of note, several of these genes have been previously implicated as modulators (Gpc1, Cyr61, Mmp14) or components (Axin2, Inhba) of TGF-ß or Wnt pathways.

Cooperative activation of TGF-ß and Wnt target genes in colorectal cancer cells. Components of the Wnt and TGF-ß pathways are frequently mutated in human colon cancers. Thus, we next investigated whether cooperative TGF-ß/Wnt targets identified in NMuMG epithelial cells might also be cooperatively enhanced in colorectal cancer cells. As most, if not all, established colorectal cell lines harbor activating mutations affecting the Wnt pathway, we focused on examining the effect of TGF-ß in the responsive human colorectal line, LS1034, which harbors a mutation in APC. Transfection of LS1034 cells with the TOPFLASH reporter revealed a high level of basal luciferase activity that was abrogated in the presence of overexpressed APC or by a mutant version of the reporter lacking LEF/TCF binding sites (FOPFLASH), thereby confirming constitutive Wnt signaling in this line (Fig. 2A, left ). Furthermore, TGF-ß–dependent activation of both the 3TP-lux and SBE-TOP reporters showed that LS1034 cells can respond to TGF-ß alone or in cooperation with the activated Wnt pathway to regulate target gene transcription (Fig. 2A, left). Consistent with this, a TGF-ß–induced increase in endogenous Smad7 and PAI-1 levels, two known TGF-ß target genes, was observed (Fig. 2A, right). Of 17 genes successfully amplified in LS1034 cells, incubation with TGF-ß increased transcript expression of 12 cooperative targets by ≥2-fold (Fig. 2A, right). Thus, TGF-ß can further stimulate the expression of 70% of cooperative target genes above the expression levels reached in the context of an activated Wnt/ß-catenin pathway. These results are consistent with the notion that TGF-ß and Wnt can cooperate to activate gene transcription in two epithelial cell systems, a mouse mammary epithelial line and a human colorectal cancer line.


Figure 2
View larger version (81K):
[in this window]
[in a new window]

  		Figure 2. A, a subset of Wnt3a/TGF-ß target genes is induced by TGF-ß in LS1034, a TGF-ß–responsive APC-mutated colorectal cell line. LS1034 cells transfected with Wnt3a [TOPLASH (TOP) and mutant control FOPFLASH (FOP)], Wnt3a+TGF-ß (SBE-TOP), or TGF-ß–responsive [3TP-lux (3TP)] reporters in the presence or absence of APC were treated with TGF-ß (left). Bars, SD of quadruplicate assays from a representative experiment. RNA from LS1034 cells treated overnight with 100 pmol/L TGF-ß was assayed for Wnt3a/TGF-ß target gene expression using real-time quantitative reverse transcription-PCR (right). Gene expression in each sample was normalized to HPRT as the internal control and expressed as fold induction by TGF-ß over control. Bars, SD for triplicate amplifications from a representative experiment. B to E, Wnt/ß-catenin and TGF-ß/Smad pathways are active in hyperplastic cells in Min intestinal adenomas and MMTV-Wnt1 tumors. B, cell lysates from HT29 and SW480 colon cancer cell lines treated overnight with 100 pmol/L TGF-ß were immunoblotted with anti–phospho-Smad2/3 (1:2,000) to show antibody specificity. C to E, paraffin-embedded sections of wild-type intestinal tissues (C), Min adenomas (D), wild-type mammary glands, and MMTV-Wnt1 mammary tumors (E) were stained for ß-catenin and phospho-Smad2/3, revealing extensive costaining in normal and tumor tissues.

 
TGF-ß and Wnt pathways are active in neoplastic intestinal and mammary gland epithelium. To investigate whether TGF-ß and Wnt can cooperate in vivo, we first assessed pathway activation in normal and neoplastic epithelium of two mouse models of Wnt-induced tumors. For gastrointestinal cancers, we used Min (APC+/–) mice, which harbor a mutation in APC and develop multiple small-bowel adenomas and colon microadenomas (31). Consistent with previous reports, ß-catenin staining was detected mainly at cell-to-cell junctions of wild-type intestinal epithelial cells (Fig. 2C) and occasionally in the nucleus of crypt cells, which is indicative of active Wnt signaling (data not shown). Staining for phospho-Smad2/3 in these sections, using a specific antibody (Fig. 2B), revealed extensive nuclear localization, demonstrating active TGF-ß signaling in small bowel and in colonic epithelial cells of both crypts and villi (Fig. 2C). In contrast, in intestinal polyps of Min mice, strong cytoplasmic ß-catenin staining was observed throughout areas displaying abnormal histology, whereas adjacent healthy epithelium showed only membrane staining (Fig. 2D, top). Extensive staining for nuclear phospho-Smad2/3 was also detected in adenomatous lesions (Fig. 2D, bottom). These results confirm that Wnt/ß-catenin signaling is strongly activated in Min intestinal adenomas, whereas normal epithelium only shows spatially limited activation within the crypt compartment. Surprisingly, however, the TGF-ß/Smad pathway seems to be widely activated in both normal and adenomatous epithelium of the small bowel and colon throughout the entire crypt-villus axis.

We next examined Wnt and TGF-ß pathway activation status in the mammary epithelium of MMTV-Wnt1 mice. These transgenic animals overexpress the Wnt1 oncogene in mammary epithelial cells leading to ductal hyperplasia during late embryogenesis and mammary adenocarcinomas in 50% of females by the age of 6 months (32). Staining of tissue sections from MMTV-Wnt1 mammary tumors revealed strong cytoplasmic ß-catenin staining in most ductal and alveolar epithelial cells, suggestive of active Wnt signaling (Fig. 2E). Extensive nuclear phospho-Smad2/3 staining was also evident in a large proportion of hyperplastic alveolar cells (Fig. 2E), indicative of an active TGF-ß/Smad pathway. Thus, these results suggest that both the Wnt/ß-catenin and TGF-ß/Smad pathways are fully activated in epithelial cells of Min intestinal adenomas and MMTV-Wnt1 mammary tumors, and may therefore cooperate to induce target genes in vivo.

Expression of TGF-ß and Wnt cooperatively induced genes in mouse models of Wnt/ß-catenin–induced tumors. As the TGF-ß and Wnt pathways are both active in Min and MMTV-Wnt1 tumors, we next examined the expression of our cooperative target genes in these tissues. For intestinal adenomas, we used macroscopically dissected samples from Min mice. Of 18 genes examined, 13 displayed a ≥2-fold (P < 0.05) increase in expression in adenomas versus normal samples. The average increase in expression ranged from 4.7- to 43-fold, with the top seven most highly overexpressed genes being IL11, Robo1, Inhba, Ankrd1, Hs6st2, Mmp14, Gpc1, and Axin2 (Fig. 3, left ). Overexpression of Axin2 and Ccnd1 in Min adenomas is consistent with previous reports (27, 33).


Figure 3
View larger version (30K):
[in this window]
[in a new window]

  		Figure 3. Wnt3a/TGF-ß target genes are overexpressed in adenomas from Min mice and MMTV-Wnt1 mammary tumors. The expression of Wnt3a/TGF-ß–regulated genes in total RNA extracted from small bowel and colon adenomas of Min mice (left) and MMTV-Wnt1 mammary tumors (right) was analyzed by real-time quantitative reverse transcription-PCR. Data are represented as log2-transformed mean ratio of expression in adenomas/tumors normalized to a matched healthy sample (for Min), or to the average of four normal mammary gland samples (for MMTV-Wnt1). Points, mean of triplicate amplification reactions. Vertical bars, average expression of each gene across all samples. *, genes significantly overexpressed in both tumor types.

 
The expression of cooperative targets was also examined in mammary tumors derived from MMTV-Wnt1 mice (Fig. 3, right). Of the 18 genes tested, 10 displayed a significant (P < 0.05) increase of ≥2-fold in tumors versus wild-type mammary glands. Ankrd1 and Gpc1 were the most highly overexpressed genes (14- and 7.5-fold, respectively). Nine genes, namely Ankrd1, Ccnd1, Ctgf, Gpc1, Hs6st2, IL11, Inhba, Mmp14, and Robo1, were significantly overexpressed in both Min and MMTV-Wnt1 tumor models. In summary, these results show that half of the genes identified as cooperatively induced genes in NMuMG cells show marked in vivo overexpression in Wnt/ß-catenin–induced intestinal and mammary tumors.

Epithelial and stromal overexpression of cooperative targets in Min adenomas. To confirm our quantitative PCR data, we did RNA in situ hybridization on a subset of genes in tumors from Min mice. As previously reported, Axin2 was highly expressed in Min adenomas (Fig. 4A ; ref. 27). Interestingly, Gpc1 and Robo1 showed similar expression patterns in adenomas, suggesting that these genes may be coregulated with Axin2 in adenomatous intestinal epithelial cells. Although we also identified Inhba as a TGF-ß/Wnt3a cooperative target in NMuMG epithelial cells, prominent Inhba expression was detected in stromal cells within and surrounding adenomas, but not in epithelial cells (Fig. 4B). Immunohistochemistry with an antibody specific to Activin A (Inhba homodimers) revealed that whereas Activin A marked the stromal compartment of normal small bowel and colon villi, this secreted factor was evident within both stromal and epithelial compartments of adenomas (Fig. 4B).


Figure 4
View larger version (113K):
[in this window]
[in a new window]

  		Figure 4. Wnt3a/TGF-ß target genes are overexpressed in the dysplastic cells of Min adenomas. A, paraffin-embedded sections of Min small bowel (top) and colon (bottom) adenomas were analyzed by anti–ß-catenin immunohistochemistry (left) and by RNA in situ hybridization (right) with the appropriate antisense digoxigenin-labeled RNA probes. Axin2, Gpc1, and Robo1 were strongly expressed in dysplastic epithelium with dysregulated ß-catenin (circled areas). B, in situ RNA hybridization for Inhba in small bowel and colon adenomas revealed strong stromal staining restricted to adenomas (left), whereas immunohistochemical detection of the active Inhba dimer (Activin A) showed stromal localization in normal tissues (center), but strong stromal and epithelial staining in adenomas (right).

 
Disruption of TGF-ß signaling in Wnt1-induced mammary tumors increases latency and alters expression of some Wnt/TGF-ß cooperative target genes. Overexpression of a dominant-negative TGF-ß type II receptor (DNIIR) driven by the MMTV enhancer attenuates TGF-ß signaling and results in the development of mammary carcinomas with a median latency of >2 years (21). To examine the contribution of the TGF-ß pathway to tumor latency in the context of activated Wnt signaling, we abrogated TGF-ß signaling in MMTV-Wnt1–induced tumors using the MMTV/DNIIR transgene. For this, we generated bigenic MMTV/Wnt1/DNIIR mice and analyzed the development of mammary tumors. As previously reported (13), 50% of MMTV-Wnt1 females develop tumors within 24 weeks. In mice expressing both transgenes, we observed an increase in tumor-free survival, with tumors appearing in 50% of mice only after 45 weeks (P < 0.0001; Fig. 5A ). These results indicate that endogenous TGF-ß signaling in the mammary epithelium promotes Wnt1-driven tumorigenesis.


Figure 5
View larger version (30K):
[in this window]
[in a new window]

  		Figure 5. Tumor latency is increased in bigenic Wnt1/DNIIR compared with MMTV-Wnt1 mammary tumors and is associated with down-regulation of a subset of Wnt3a/TGF-ß target genes. A, bigenic MMTV-WNT1/DNIIR mice remain tumor-free for ~20 wk longer than MMTV-Wnt1 transgenic mice (P < 0.0001). B, expression of Wnt1, Gpc1, Inhba, and Robo1 in wild-type mammary glands. MMTV-Wnt1 and MMTV-Wnt1/DNIIR mammary tumors were analyzed by real-time quantitative reverse transcription-PCR. Although Wnt1 levels are not significantly different in Wnt1 versus Wnt1/DNIIR tumors (P = 0.2), decreased expression of Gpc1 (P = 0.003), Inhba (P = 0.02), and Robo1 (P = 0.002) in bigenic Wnt1/DNIIR tumors was observed.

 
We next examined the effect of blocking TGF-ß signaling on the expression of cooperatively regulated TGF-ß/Wnt target genes by quantitative PCR. MMTV-Wnt1 and MMTV-Wnt1/DNIIR tumor samples with similar levels of Wnt1 expression were identified and further analyzed (P = 0.2; Fig. 5B, top left). Our previous analysis revealed that nine of the Wnt/TGF-ß cooperative target genes displayed increased expression in MMTV-Wnt1 tumors and Min adenomas (Fig. 3). Thus, we examined the expression of these genes to assess the effect of reduced TGF-ß signaling. Although Ankrd1, Ccnd1, Ctgf, IL11, Hs6st2, and Mmp14 were expressed at similar levels in both tumor types (data not shown), Gpc1, Inhba, and Robo1 displayed significantly reduced expression (P < 0.05) in Wnt1/DNIIR tumors (Fig. 5B). Of note, these three genes were among those showing the most dramatic TGF-ß/Wnt3a cooperativity in NMuMG cells.

Cooperative Wnt/TGF-ß targets in FAP adenomas and human cancers. Our previous data showed that Wnt/TGF-ß cooperative targets are overexpressed in intestinal adenomas of Min mice, a model for human FAP. We next investigated whether the same cooperative genes exhibited increased expression in eight human FAP samples using quantitative PCR. Consistent with previous studies (27), Axin2 was overexpressed in adenomas by an average 10.1-fold (P = 1.5 x 10–10). Of nine Wnt/TGF-ß targets whose expression was increased in Min and MMTV-Wnt1 tumors, Ccnd1, Hs6st2, and Inhba were also significantly overexpressed in FAP tumors by ≥2-fold and had mean expression ratios of 2.2-fold (P = 0.0001), 2.7-fold (P = 0.002), and 3.2-fold (P = 0.003), respectively. We also analyzed the Inhba status by hybridization of a radiolabeled probe to a human cancer profiling array. Extensive overexpression of Inhba in colon cancers (average 3.3-fold), as well as cancers of the small intestine (3.0-fold), rectum (4.1-fold), and, to a weaker extent, in breast malignancies (2.4-fold; Fig. 6B ), was observed. Although our sample size is limited, these findings indicate that increased expression of TGF-ß/Wnt cooperative targets may also be involved in human cancer progression.


Figure 6
View larger version (53K):
[in this window]
[in a new window]

  		Figure 6. Overexpression of Wnt3a/TGF-ß target genes in FAP polyps and other human cancers. A, RNA extracted from FAP polyps was analyzed by real-time quantitative RT-PCR for the indicated genes. Results are expressed as the log2-transformed ratio of transcript expression in polyps normalized to the matched normal samples, or, if unavailable, to the mean of all normals analyzed in parallel. Vertical bars, average expression of each gene across all samples. *, genes overexpressed in mouse tumor models (Fig. 3). Ankrd1, Dusp1, and IL11 are not shown due to poor or no amplification. B, the expression of Inhba in human cancers was examined by hybridization of antisense 32P-labeled cDNA probe to a cancer profiling array. Densitometric analysis of the autoradiogram (bottom). Tumor/normal (T/N) expression ratios are graphed using the indicated ranges.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Involvement of Wnt and TGF-ß signaling in embryonic development, tissue homeostasis, and pathologic conditions suggests that these pathways may function within a complex interdependent and mutually regulating network (1, 2). To gain mechanistic insights into cooperative signaling by Wnt and TGF-ß, we used a microarray-based approach to evaluate crosstalk effects on gene expression. We observed that costimulation of epithelial-derived tissue culture cells with both factors was not simply a sum of the individual responses, rather that the ligands both positively and negatively modified each other's gene responses. This highlights the complex, intertwined nature of the two signaling pathways. Interestingly, many of the identified cooperative genes encode for proteins involved in signaling processes and it is particularly intriguing that several of these genes are modulators or components of the TGF-ß and Wnt pathways themselves. For instance, Axin2, an identified cooperative target, functions redundantly with Axin within the ß-catenin destruction complex in the Wnt pathway (2) and can interact with Smad3 to facilitate its activation by the TGF-ß receptor complex (34). Another cooperative target, the heparan-sulfate proteoglycan, Gpc1, is known to modulate the activity of various growth factors, including TGF-ß and Wnts (35). Remodeling of the 6-O sulfation state of Gpc1 by QSulf1 can modulate Wnt signaling (36) and, interestingly, we identified 6-O-sulfotransferase 2 (Hs6st2), a potential modifier of Gpc1, as a cooperative target of Wnt and TGF-ß. Knockdown of Hs6st activity in zebrafish leads to myogenic defects and up-regulation of the Wnt target genes myoD and engrailed2 (37). Because both Wnt and TGF-ß–related pathways are strongly implicated in myogenesis, these results suggest that Hs6st activity may be a novel modulator of these pathways. Differential sulfation of Gpc1 also regulates its affinity for Slit ligands (38), which elicit their biological effects via binding to the Roundabout (Robo) family of transmembrane receptors (39). Of note, we also identified Robo1 as a cooperative target of Wnt and TGF-ß. Finally, Inhba, another cooperative target, is itself a member of the TGF-ß family of ligands. Our results, therefore, show that cooperative signaling by Wnt and TGF-ß up-regulates the expression of several known and potentially novel feedback regulators.

Molecular alterations affecting TGF-ß and Wnt pathway components in human cancers and mouse tumor models have been widely reported (6). Activation of Wnt signaling is a feature of several human malignancies with most colorectal tumors harboring mutations in APC or ß-catenin (40). In contrast, for the TGF-ß pathway, mutations or epigenetic modifications that target the receptors or Smads lead to loss or attenuation of TGF-ß–mediated growth inhibition. Although the selective loss of TGF-ß–mediated growth inhibition favors progression of epithelial tumors, maintenance of other aspects of TGF-ß signaling enhances tumor invasion and metastasis (11). This suggests that an intact TGF-ß pathway integrates two opposing actions: a dominant activity that prevents tumor progression and another activity that favors it. This also implies that crosstalk events between a fully functional TGF-ß pathway and other tumor-initiating oncogenic pathways, such as Wnt, may set up a gene expression program that facilitates early tumor progression. Consistent with this possibility, we observed that in two mouse models of Wnt-induced tumors, Min and MMTV-Wnt1 mice, the Wnt/ß-catenin and TGF-ß/Smad pathways are active in tumor epithelial cells. The molecular pathology of distinct cancer types is quite diverse; however, we were particularly interested in genes cooperatively regulated by TGF-ß and Wnt in multiple tumor types, and our expression analysis showed that nine cooperative targets identified in tissue culture cells were also significantly overexpressed in both Min adenomas and MMTV-Wnt1 tumors. Furthermore, preliminary analysis in a small sampling of FAP patients revealed that four of these genes (Axin2, Ccnd1, Hs6st2, and Inhba) also show increased expression. Examination of a larger number of samples will be required to determine the extent of overexpression of our set of cooperative targets in FAP adenomas and whether these genes might contribute to tumor progression in other human cancers.

To directly assess the effect of TGF-ß signaling on Wnt1-induced tumors, we generated double-transgenic mice by crossing MMTV-Wnt1 to MMTV/DNIIR animals. The MMTV/DNIIR transgene alone does not delay ductal development, although it causes lobulo-alveolar development in virgin animals (41). Thus, the observed delay in tumor formation in the bigenic versus MMTV-Wnt1 mice is consistent with the model that endogenous TGF-ß signaling in mammary epithelial cells contributes to Wnt1-induced tumorigenesis. Concomitant with the increase in tumor latency in bigenic animals, we observed that at least three cooperative target genes, namely Gpc1, Inhba, and Robo1, show reduced expression levels in Wnt1/DNIIR tumors compared with those from Wnt1 animals. Although a few gene targets may be primary contributors to the TGF-ß effect in this model, we believe that a combinatorial effect on overall gene expression patterns is more likely responsible for the increase in tumor latency. In contrast to our study, in two other mammary tumor models, MMTV-TGF-{alpha} and MMTV-polyoma middle T antigen–induced tumors, decreased TGF-ß signaling has been shown to reduce rather than augment tumor latency (21, 42). Interestingly, although targeted deletion of one Smad4 allele was shown to increase tumor size in Min mice (43), specific deletion of Smad4 in T cells, but not in epithelial cells, leads to spontaneous epithelial cancers of the gastrointestinal tract (44). Together with our data, these observations suggest that the molecular basis for the dichotomous role of TGF-ß in tumorigenesis may not stem from TGF-ß–only initiated signals per se, but rather from the combinatorial effects of TGF-ß on gene expression patterns that occur in the context of distinct oncogenic pathways.

In our study, we examined cooperativity between TGF-ß and Wnt. However, Activin A, which is composed of an Inhba dimer, is a TGF-ß family member that also signals through Smad2/3 to activate TGF-ß–like signals. Thus, our identification of Inhba as a cooperative target is of particular interest. We showed that Inhba expression is increased in MMTV-Wnt1 and Min tumors, decreased in MMTV-Wnt1/DNIIR tumors, and overexpressed in human intestinal and breast malignancies. Thus, Activin is a candidate for a TGF-ß superfamily factor that cooperates with activated Wnt/ß-catenin pathway in tumor epithelial cells. Although Inhba is a Wnt/TGF-ß cooperative target in epithelial cells, in the Min mouse-derived adenomas, mRNA expression was detected primarily in the stromal compartment. On the other hand, secreted Activin A protein was abundant throughout the tumor. We postulate that persistent stimulation of tumor cells by Activin A produced by the stroma may establish a positive autoregulatory loop that induces Inhba transcription in tumor epithelial cells in late tumorigenesis. Consistent with this notion, overexpression of Inhba leading to overproduction of Activin A protein by epithelial tumor cells has been reported in stage IV colorectal cancer (45). A more widespread role for Inhba in neoplasia is also suggested by the observation that Inhba is overexpressed in a range of human tumors (46–50).


   Acknowledgments
 
Grant support: National Cancer Institute of Canada with funds from the Cancer Research Society and a Canada Research Chair (L. Attisano), USPHS grants CA085492 and CA102162 (H.L. Moses), and a Canadian Institutes of Health Research Doctoral Research Award (E. Labbé). R. Gryfe is a Charles H. Hollenberg Clinician Scientist with funds from Eli-Lilly Canada, Cancer Care Ontario, and the Canadian Institutes of Health Research.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank C. Silvestri for insightful comments and C. Ash and K. Boras-Granic for expert technical advice.


   Footnotes
 
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

L. Attisano is a Canada Research Chair.

5 X. Chytil, Y. Chen, and H.L. Moses, unpublished data. Back

6 http://www.mshri.on.ca/microarray/. Back

Received 7/11/06. Revised 10/ 7/06. Accepted 11/ 3/06.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Massague J. How cells read TGF-ß signals. Nat Rev Mol Cell Biol 2000;1:169–78.[CrossRef][Medline]
  2. Nusse R. Wnt signaling in disease and in development. Cell Res 2005;15:28–32.[CrossRef][Medline]
  3. Feng XH, Derynck R. Specificity and versatility in TGF-ß signaling through Smads. Annu Rev Cell Dev Biol 2005;21:659–93.[CrossRef][Medline]
  4. Xanthos JB, Kofron M, Tao Q, et al. The roles of three signaling pathways in the formation and function of the Spemann organizer. Development 2002;129:4027–43.[Abstract/Free Full Text]
  5. Willert J, Epping M, Pollack JR, et al. A transcriptional response to Wnt protein in human embryonic carcinoma cells. BMC Dev Biol 2002;2:8.[CrossRef][Medline]
  6. Elliott RL, Blobe GC. Role of transforming growth factor ß in human cancer. J Clin Oncol 2005;23:2078–93.[Abstract/Free Full Text]
  7. Lin SY, Xia W, Wang JC, et al. ß-Catenin, a novel prognostic marker for breast cancer: its roles in cyclin D1 expression and cancer progression. Proc Natl Acad Sci U S A 2000;97:4262–6.[Abstract/Free Full Text]
  8. Ugolini F, Charafe-Jauffret E, Bardou VJ, et al. WNT pathway and mammary carcinogenesis: loss of expression of candidate tumor suppressor gene SFRP1 in most invasive carcinomas except of the medullary type. Oncogene 2001;20:5810–7.[CrossRef][Medline]
  9. Wissmann C, Wild PJ, Kaiser S, et al. WIF1, a component of the Wnt pathway, is down-regulated in prostate, breast, lung, and bladder cancer. J Pathol 2003;201:204–12.[CrossRef][Medline]
  10. Howe LR, Brown AM. Wnt signaling and breast cancer. Cancer Biol Ther 2004;3:36–41.[Medline]
  11. Akhurst RJ, Derynck R. TGF-ß signaling in cancer—a double-edged sword. Trends Cell Biol 2001;11:S44–51.[Medline]
  12. Muraoka RS, Koh Y, Roebuck LR, et al. Increased malignancy of Neu-induced mammary tumors overexpressing active transforming growth factor ß1. Mol Cell Biol 2003;23:8691–703.[Abstract/Free Full Text]
  13. Bocchinfuso WP, Hively WP, Couse JF, et al. A mouse mammary tumor virus-Wnt-1 transgene induces mammary gland hyperplasia and tumorigenesis in mice lacking estrogen receptor-{alpha}. Cancer Res 1999;59:1869–76.[Abstract/Free Full Text]
  14. Kwan H, Pecenka V, Tsukamoto A, et al. Transgenes expressing the Wnt-1 and int-2 proto-oncogenes cooperate during mammary carcinogenesis in doubly transgenic mice. Mol Cell Biol 1992;12:147–54.[Abstract/Free Full Text]
  15. Labbe E, Attisano L. Integration of signalling pathways via Smad proteins. In: ten Dijke P, Heldin CH, editors. Smad signal transduction. Smads in proliferation, differentiation, and disease. Dordrecht, the Netherlands: Springer; 2006. p 296–316.
  16. Labbe E, Letamendia A, Attisano L. Association of Smads with lymphoid enhancer binding factor 1/T cell-specific factor mediates cooperative signaling by the transforming growth factor-ß and wnt pathways. Proc Natl Acad Sci U S A 2000;97:8358–63.[Abstract/Free Full Text]
  17. Hussein SM, Duff EK, Sirard C. Smad4 and ß-catenin co-activators functionally interact with lymphoid-enhancing factor to regulate graded expression of Msx2. J Biol Chem 2003;278:48805–14.[Abstract/Free Full Text]
  18. Lei S, Dubeykovskiy A, Chakladar A, et al. The murine gastrin promoter is synergistically activated by transforming growth factor-ß/Smad and Wnt signaling pathways. J Biol Chem 2004;279:42492–502.[Abstract/Free Full Text]
  19. Shibamoto S, Higano K, Takada R, et al. Cytoskeletal reorganization by soluble Wnt-3a protein signalling. Genes Cells 1998;3:659–70.[Abstract]
  20. Korinek V, Barker N, Morin PJ, et al. Constitutive transcriptional activation by a ß-catenin-Tcf complex in APC–/– colon carcinoma. Science 1997;275:1784–7.[Abstract/Free Full Text]
  21. Gorska AE, Jensen RA, Shyr Y, et al. Transgenic mice expressing a dominant-negative mutant type II transforming growth factor-ß receptor exhibit impaired mammary development and enhanced mammary tumor formation. Am J Pathol 2003;163:1539–49.[Abstract/Free Full Text]
  22. Lee S, Solow-Cordero DE, Kessler E, et al. Transforming growth factor-ß regulation of bone morphogenetic protein-1/procollagen C-proteinase and related proteins in fibrogenic cells and keratinocytes. J Biol Chem 1997;272:19059–66.[Abstract/Free Full Text]
  23. Igarashi A, Okochi H, Bradham DM, et al. Regulation of connective tissue growth factor gene expression in human skin fibroblasts and during wound repair. Mol Biol Cell 1993;4:637–45.[Abstract]
  24. Lund LR, Riccio A, Andreasen PA, et al. Transforming growth factor-ß is a strong and fast acting positive regulator of the level of type-1 plasminogen activator inhibitor mRNA in WI-38 human lung fibroblasts. EMBO J 1987;6:1281–6.[Medline]
  25. Hahn AW, Resink TJ, Scott-Burden T, et al. Stimulation of endothelin mRNA and secretion in rat vascular smooth muscle cells: a novel autocrine function. Cell Regul 1990;1:649–59.[Medline]
  26. Chesire DR, Dunn TA, Ewing CM, et al. Identification of aryl hydrocarbon receptor as a putative Wnt/ß-catenin pathway target gene in prostate cancer cells. Cancer Res 2004;64:2523–33.[Abstract/Free Full Text]
  27. Lustig B, Jerchow B, Sachs M, et al. Negative feedback loop of Wnt signaling through upregulation of conductin/axin2 in colorectal and liver tumors. Mol Cell Biol 2002;22:1184–93.[Abstract/Free Full Text]
  28. El-Tanani M, Platt-Higgins A, Rudland PS, et al. Ets gene PEA3 cooperates with ß-catenin-Lef-1 and c-Jun in regulation of osteopontin transcription. J Biol Chem 2004;279:20794–806.[Abstract/Free Full Text]
  29. Tetsu O, McCormick F. ß-Catenin regulates expression of cyclin D1 in colon carcinoma cells. Nature 1999;398:422–6.[CrossRef][Medline]
  30. Taneyhill L, Pennica D. Identification of Wnt responsive genes using a murine mammary epithelial cell line model system. BMC Dev Biol 2004;4:6.[CrossRef][Medline]
  31. Moser AR, Luongo C, Gould KA, et al. ApcMin: a mouse model for intestinal and mammary tumorigenesis. Eur J Cancer 1995;31A:1061–4.
  32. Li Y, Hively WP, Varmus HE. Use of MMTV-Wnt-1 transgenic mice for studying the genetic basis of breast cancer. Oncogene 2000;19:1002–9.[CrossRef][Medline]
  33. Sansom OJ, Reed KR, van de Wetering M, et al. Cyclin D1 is not an immediate target of ß-catenin following Apc loss in the intestine. J Biol Chem 2005;280:28463–7.[Abstract/Free Full Text]
  34. Furuhashi M, Yagi K, Yamamoto H, et al. Axin facilitates Smad3 activation in the transforming growth factor ß signaling pathway. Mol Cell Biol 2001;21:5132–41.[Abstract/Free Full Text]
  35. Lin X. Functions of heparan sulfate proteoglycans in cell signaling during development. Development 2004;131:6009–21.[Abstract/Free Full Text]
  36. Ai X, Do AT, Lozynska O, et al. QSulf1 remodels the 6-O sulfation states of cell surface heparan sulfate proteoglycans to promote Wnt signaling. J Cell Biol 2003;162:341–51.[Abstract/Free Full Text]
  37. Bink RJ, Habuchi H, Lele Z, et al. Heparan sulfate 6-O-sulfotransferase is essential for muscle development in zebrafish. J Biol Chem 2003;278:31118–27.[Abstract/Free Full Text]
  38. Ronca F, Andersen JS, Paech V, et al. Characterization of Slit protein interactions with glypican-1. J Biol Chem 2001;276:29141–7.[Abstract/Free Full Text]
  39. Couch J, Condron B. Axon guidance: Comm hither, Robo. Curr Biol 2002;12:R741–2.[CrossRef][Medline]
  40. Sparks AB, Morin PJ, Vogelstein B, et al. Mutational analysis of the APC/ß-catenin/Tcf pathway in colorectal cancer. Cancer Res 1998;58:1130–4.[Abstract/Free Full Text]
  41. Gorska AE, Joseph H, Derynck R, et al. Dominant-negative interference of the transforming growth factor ß type II receptor in mammary gland epithelium results in alveolar hyperplasia and differentiation in virgin mice. Cell Growth Differ 1998;9:229–38.[Abstract]
  42. Forrester E, Chytil A, Bierie B, et al. Effect of conditional knockout of the type II TGF-ß receptor gene in mammary epithelia on mammary gland development and polyomavirus middle T antigen induced tumor formation and metastasis. Cancer Res 2005;65:2296–302.[Abstract/Free Full Text]
  43. Takaku K, Oshima M, Miyoshi H, et al. Intestinal tumorigenesis in compound mutant mice of both Dpc4 (Smad4) and Apc genes. Cell 1998;92:645–56.[CrossRef][Medline]
  44. Kim BG, Li C, Qiao W, et al. Smad4 signalling in T cells is required for suppression of gastrointestinal cancer. Nature 2006;441:1015–9.[CrossRef][Medline]
  45. Wildi S, Kleeff J, Maruyama H, et al. Overexpression of activin A in stage IV colorectal cancer. Gut 2001;49:409–17.[Abstract/Free Full Text]
  46. Liang Y, Diehn M, Watson N, et al. Gene expression profiling reveals molecularly and clinically distinct subtypes of glioblastoma multiforme. Proc Natl Acad Sci U S A 2005;102:5814–9.[Abstract/Free Full Text]
  47. Zou TT, Selaru FM, Xu Y, et al. Application of cDNA microarrays to generate a molecular taxonomy capable of distinguishing between colon cancer and normal colon. Oncogene 2002;21:4855–62.[CrossRef][Medline]
  48. Sorlie T, Perou CM, Tibshirani R, et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A 2001;98:10869–74.[Abstract/Free Full Text]
  49. Iacobuzio-Donahue CA, Ashfaq R, Maitra A, et al. Highly expressed genes in pancreatic ductal adenocarcinomas: a comprehensive characterization and comparison of the transcription profiles obtained from three major technologies. Cancer Res 2003;63:8614–22.[Abstract/Free Full Text]
  50. Dhanasekaran SM, Dash A, Yu J, et al. Molecular profiling of human prostate tissues: insights into gene expression patterns of prostate development during puberty. FASEB J 2005;19:243–5.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Cancer Prevention ResearchHome page
N. L. Cho, M. Redston, A. G. Zauber, A. M. Carothers, J. Hornick, A. Wilton, S. Sontag, N. Nishioka, F. M. Giardiello, J. R. Saltzman, et al.
Aberrant Crypt Foci in the Adenoma Prevention with Celecoxib Trial
Cancer Prevention Research, June 1, 2008; 1(1): 21 - 31.
[Abstract] [Full Text] [PDF]

Home page
 Genes Dev.Home page
X. Guo, A. Ramirez, D. S. Waddell, Z. Li, X. Liu, and X.-F. Wang
Axin and GSK3- control Smad3 protein stability and modulate TGF- signaling
Genes & Dev., January 1, 2008; 22(1): 106 - 120.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Labbé, E.
Right arrow Articles by Attisano, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Labbé, E.
Right arrow Articles by Attisano, L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online