Cortical Stabilization of B-Catenin Contributes to NHERF1/EBP50 Tumor Suppressor Function
Total Page:16
File Type:pdf, Size:1020Kb
Oncogene (2007) 26, 5290–5299 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc ORIGINAL ARTICLE Cortical stabilization of b-catenin contributes to NHERF1/EBP50 tumor suppressor function EL Kreimann1, FC Morales1, J de Orbeta-Cruz1, Y Takahashi1, H Adams2, T-J Liu1, PD McCrea3, and M-M Georgescu1,2 1Department of Neuro-Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA; 2Department of Molecular Genetics, The University of Texas MD Anderson Cancer Center, Houston, TX, USA and 3Department of Biochemistry and Molecular Biology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA Anchorage-independent growth is a hallmark of tumor extracellular matrix owing to alterations in cell surface growth and results from enhanced proliferation and altered molecules such as cadherins and integrins (Thiery, cell–cell and cell-matrix interactions. By using gene- 2002). Cadherins constitute a major family of trans- deficient mouse embryonic fibroblasts (MEFs), we showed membrane glycoproteins that mediate cell–cell adhesion. for the first time that NHERF1/EBP50 (Na/H exchanger Adjacent cells in tissues are linked by the interaction of regulator factor 1/ezrin-radixin-moesin binding phospho- cadherin extracellular domains, but the strength of the protein 50), an adapter protein with membrane localization binding is supported by cytoplasmic interactions, which under physiological conditions, inhibits cell motility and is include those with catenins and thereby the actin required to suppress anchorage-independent growth. Both cytoskeleton (Provost and Rimm, 1999). This intracel- NHERF1 PDZ domains are necessary for the tumor lular cluster of proteins is central to the assembly and suppressor effect. NHERF1 associates directly through the function of adherens junctions and the large complex is PDZ2 domain with b-catenin and is required for b-catenin composed of many components, of which b-catenin and localization at the cell–cell junctions in MEFs. Mechan- p120 catenin bind directly to the cytoplasmic tail of istically, the absence of NHERF1 selectively decreased cadherins (McCrea and Gumbiner, 1991; Reynolds the interaction of b-catenin with E-cadherin, but not with et al., 1994). b-catenin additionally acts as a transcrip- N-cadherin. The ensuing disorganization of E-cadherin- tional activator in the Wnt pathway (Nelson and Nusse, mediated adherens junctions as well as the observed 2004). Deregulation of Wnt signaling has been found in moderate increase in b-catenin transcriptional activity different types of cancer, in particular in colon cancer contributed most likely to the anchorage-independent where, most frequently, mutations in APC lead to growth of NHERF1-deficient MEFs. In vivo, NHERF1 is decreased protein turnover and cytoplasmic accumula- specifically localized at the apical brush-border membrane tion of b-catenin (Giles et al., 2003). Following in intestinal epithelial cells and is required to maintain translocation to the nucleus, b-catenin relieves T-cell a fraction of the cortical b-catenin at this level. Thus, factor/lymphoid enhancer factor (TCF/LEF)-mediated NHERF1 emerges as a cofactor essential for the inte- transcriptional repression to activate several oncogenes grity of epithelial tissues by maintaining the proper and other gene targets. Although the regulation of localization and complex assembly of b-catenin. b-catenin subcellular localization is a matter of extensive Oncogene (2007) 26, 5290–5299; doi:10.1038/sj.onc.1210336; investigation (Brembeck et al., 2004; Gottardi and published online 26 February 2007 Gumbiner, 2004), mechanisms of b-catenin shuttling between the membrane and cytoplasmic or nuclear Keywords: NHERF1/EBP50; b-catenin; E-cadherin; complexes are only beginning to be understood. transformation; colon cancer; mouse embryonic b-Catenin was shown to bind in vitro and in vivo fibroblasts (MEFs) through its C-terminal PDZ motif to Na/H exchanger regulator factor 1/ezrin-radixin-moesin (ERM)-binding phosphoprotein 50 (NHERF1/EBP50) (Shibata et al., 2003) that contains two PDZ domains and an ERM- binding region (Reczek et al., 1997; Weinman et al., Introduction 2003). NHERF1 is involved in the regulation of ion transporters and in the trafficking of many transmem- Transformed cells lack normal cell contact inhibition brane molecules (Shenolikar et al., 2004), but its role in and they grow in the absence of anchorage to the cancer is still controversial. NHERF1 gene mutations found with low frequency (3%) in breast cancer cell lines Correspondence: Dr M-M Georgescu, The University of Texas, and primary tumors were associated with loss of MD Anderson Cancer Center, 6767 Bertner Avenue, Houston, heterozygosity and increased aggressiveness (Dai et al., TX 77030, USA. E-mail: [email protected] 2004), suggesting a tumor suppressor role for NHERF1. Received 20 October 2006; revised 15 January 2007; accepted 15 January In contrast, Shibata et al. (2003) reported that the 2007; published online 26 February 2007 NHERF1/b-catenin complex promotes Wnt signaling NHERF1 stabilizes b-catenin/E-cadherin complexes EL Kreimann et al 5291 and may cooperate in the development of liver cancer. polygonal and more epithelial-like than their parental, To clarify the role of NHERF1, we analysed the growth non-immortalized counterparts (not shown). However, properties of mouse embryonic fibroblasts (MEFs) although immortalized NHERF1( þ / þ ) MEFs grew as generated from NHERF1-deficient mice (Morales et al., a monolayer on both plastic or poly-D-lysine-coated 2004). We found that NHERF1(À/À) immortalized glass surfaces, NHERF1(À/À) MEFs retracted and MEFs, in contrast to their wild-type counterparts, presen- clustered in spheres on poly-D-lysine-coated glass cover- ted anchorage-independent growth. In NHERF1 (À/À) slips (Figure 1a). Compared to NHERF1( þ / þ ) cells, cells, b-catenin was delocalized from the plasma mem- (À/À) MEFs presented delayed spreading when plated brane to the cytoplasm and formed weaker complexes on poly-D-lysine but not on collagen matrix (Figure 1b), with E-cadherin. Our results point to a tumor sup- or on fibronectin (not shown). Unlike collagen or pressor role for NHERF1 via stabilization of b-catenin/ fibronectin, poly-D-lysine promotes adhesion by an E-cadherin complexes at the plasma membrane. integrin-independent mechanism based on electrostatic interactions with the cell membrane, suggesting integrin- independent membrane modifications of NHERF1 Results (À/À) cells. Importantly, immortalized NHERF1(À/À) cells were able to form colonies in soft agar (Figure 1c), Loss of NHERF1 confers anchorage-independent growth suggesting that the loss of NHERF1 promotes ancho- to immortalized cells rage-independent growth. The analysis of the prolifera- To assess the role in tumor development of NHERF1, tion of parental and immortalized MEFs showed no we immortalized NHERF1( þ / þ ) and (À/À) MEFs significant difference between the parental counterparts with SV40T antigens and characterized their growth and a reproducible moderate proliferation advantage for properties. At confluence, immortalized MEFs appeared NHERF1(À/À) immortalized cells relative to ( þ / þ ) Figure 1 Anchorage-independent growth and increased migration of NHERF1(À/À) MEFs. (a) Immortalized NHERF1(À/À) cells grown on poly-D-lysine-coated glass coverslips formed cell spheres (arrows), whereas NHERF1( þ / þ ) grew as monolayer (not shown). (b) Immortalized MEFs were plated in triplicate on poly-D-lysine or collagen-coated glass coverslips and allowed to attach and spread for the indicated periods of time. Means7s.e.m. for cells not spread or spread were calculated from equivalent numbers of cells (N ¼ 50–80) counted on five fields per plate and were expressed as percentages. Significant differences between NHERF1( þ / þ ) and (À/À) cells are marked with asterisk (Po0.05). (c) Immortalized MEFs were examined for anchorage-independent growth by soft-agar assay. Colonies were visible for NHERF1(À/À) MEFs after 4 weeks of incubation. Images ( Â 100) were taken with a Zeiss Axiovert 200 microscope. The graph represents the number of colonies per plate calculated as mean7s.d. from triplicate plates. (d) Proliferation assay showing growth of parental and immortalized MEFs in DMEM containing 0.5% FBS. Cells were counted after 48 h incubation and values are expressed as means7s.d. from triplicate measurements. Significant differences are marked with asterisk (Po0.005). (e) Scratch-assay showing migration of immortalized MEFs on plastic culture dishes. Images ( Â 40) were taken at time 0 (0 h) showing the initial scratch of the monolayer and after 13 h of migration, when the NHERF1(À/À) cells closed the scratch (T ¼ 13 h). All these experiments were repeated three times with similar results. Oncogene NHERF1 stabilizes b-catenin/E-cadherin complexes EL Kreimann et al 5292 Figure 2 Reconstitution of NHERF1 in NHERF1(À/À) cells suppresses colony formation. (a) Schematic organization of NHERF1 showing two PDZ domains and an EB region. NHERF1 FL and domain-deletion mutants are shown with amino-acid boundaries. (b–d) Immortalized NHERF1(À/À) MEFs were reconstituted by retroviral infection with NHERF1 FL and mutants and assayed for proliferation (b) and soft agar colony formation (c and d). The graph in (c) shows the average7s.d. of the number of colonies counted in triplicate plates. Significant differences, as compared to vector control, are marked with asterisk (Po0.005). Immunoblotting shows the levels of reconstituted proteins. These experiments were repeated three times with similar results. cells (Figure 1d). We also investigated the mobility NHERF1 controls the intracellular distribution of MEFs by wound-healing assay. NHERF1(À/À) of b-catenin cells migrated faster than their ( þ / þ ) counterparts, NHERF1 is known to bind multiple ligands through the filling the scratch in 13 versus 16 h respectively PDZ1 domain, to NHE3 through both PDZ domains, (Figure 1e), suggesting cytoskeletal changes in and to b-catenin and Yap 65 specifically through the NHERF1(À/À) cells. PDZ2domain (Voltz et al., 2001; Weinman et al., 2003).