Abstract
Background/Aim: β-Catenin regulates cell-cell adhesion and gene transcription and acts as a master switch that controls proliferation in several types of cancer. ESE1 is an epithelium-restricted transcription factor and its multiple domain structure predicts its interaction with other proteins with diverse cellular functions. Here, for the first time, we report that endogenous β-catenin binds to and co-localizes with endogenous ESE1 in the cytoplasm. Materials and Methods: The binding sites were mapped to E26 transformation-specific (ETS) domain at carboxyl terminus of ESE1 and N-terminus of β-catenin. Results: We found that C-terminus of ESE1 also binds to α-catenin and that ESE1/β-catenin interaction was abrogated by knockdown of either β-catenin or α-catenin. Conclusion: The data suggest that interactions between ESE1 and β-/α-catenins might be a mechanism by which the ESE1 protein determines the β-catenin function and tumorigenesis.
The β-catenin signaling pathway is an important pathway in the development of tumorigenesis. The clinical significance of β-catenin nuclear translocation with respect to tumor progression, survival and differential diagnosis has been well-established. β-Catenin interacts with the cytoplasmic domain of cadherin and links cadherin to α-catenin, which, in turn, mediates anchorage of the cadherin complex to the cortical actin that is essential for the intercellular adhesion function of normal epithelia (1). Recent series of proteomic studies has revealed that nuclear proteins associated with the β-catenin-T-cell factor (TCF) complex may include feasible targets for molecular therapy against colorectal cancer (2). Thus, identification of nuclear protein interacting with β-catenin and the T-cell factor/lymphoid enhancer factor (TCF/LEF) complex may provide a milestone that could lead to development of new therapeutics with higher efficacy and specificity for colorectal cancer.
Epithelial-specific ETS1 (ESE1) is an E26 transformation-specific (ETS) family of transcription factors regulating terminal differentiation of the epidermis (3). ESE1 expression transforms MCF-12A human mammary epithelial cells and induces anchorage-independent growth (4, 5). ESE1 is overexpressed in human primary and metastatic tumors (6-8) and induces transformation phenotypes in vitro and in vivo in mammary epithelial cells (9-11). Recently, Wang et al. reported that ESE1 is associated with poor prognosis in colorectal cancer (12). However, the function of ESE1 appears to be complex. ESE1 has a multiple domain structure, including a N-terminus pointed domain (PNT) implicated in protein-protein interactions (13), and a C-terminus ETS domain. In addition, three other domains (TAD, SAR and AT hook) are located between the PNT and ETS domain in tandem (14). The presence of these motifs suggests that ESE1 may possess biological activities via interacting with other proteins rather than its nuclear transcriptional function. Identification of such interacting partners could enhance our understanding of the mechanism(s) through which ESE1 is involved in tumorigenesis.
Here we report that β-catenin is a novel ESE1 binding partner. We mapped their binging sites and provided responsible domains for protein-protein interaction.
Materials and Methods
Cell lines and antibodies. The human colon cancer cell lines (HCT116, LoVo, HCT15, SW480) and human embryonic kidney cell line (HEK293) were purchased from ATCC (Manassas, VA, USA). HEK293 cells were cultured in a DMEM medium (Hyclone, Logan, UT, USA) containing 10% FBS and the others were cultured in a DMEM/F12 medium (Hyclone) containing 10% FBS at 37°C in a humidified 5% CO2 atmosphere. The antibodies used in the current study were purchased from the following sources: anti-β-catenin antibody (Cell Signaling Technology, Danvers, MA, USA or Santa Cruz Biotechnologies, Santa Cruz, CA, USA), anti-ESE1 antibody (Sigma Aldrich, St. Louis, MO, USA or Abcam, Cambridge, MA, USA); anti-GST antibody, anti-actin antibody, anti-α-catenin, agarose-conjugated protein A/G beads (Santa Cruz Biotechnologies); and anti-V5 monoclonal antibody (Invitrogen, Carlsbad, CA, USA).
The sequences of the primers used for cloning of ESE1 and β-catenin clones.
Plasmid constructions and reverse transcription-polymerase chain reaction (RT-PCR). Full-length or truncated cDNA fragments encoding ESE1 with N-terminal V5 tag or β-catenin with GST tag were generated by standard PCR methods and subcloned into pcDNA3.1-n/V5 or pDEST™-27 vector (Invitrogen) according to the instruction, respectively. ESE1 ΔTAD mutant (128-160) was generated with pcDNA3.1/nV5/ESE1 using QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) according to manufacturer's instruction. All constructs were verified by DNA sequencing. The sequences of the primers used for PCR amplification are listed in Table I.
Co-immunoprecipitation and immunoblotting. HCT116 cells were lysed in a NP-40 lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 5 mM EDTA, 0.5% NP-40) supplemented with protease inhibitor cocktail (Sigma-Aldrich) and cleared by centrifugation at 12,000 × g for 20 min at 4°C. Pre-cleared cell lysates (1,000 μg) were incubated with 1 μg of anti-ESE1 antibody (Abcam) or anti-β-catenin antibody (Cell signaling) at 4°C overnight. Immunocomplex was rotated with 30 μl of 50% of agarose-conjugated protein A/G beads (Santa Cruz) at 4°C for 4 h. The pellets were washed four times with a NP-40 lysis buffer, boiled in a protein loading buffer and then subjected to immunoblotting analyses using an anti-ESE1 (Sigma) or anti-β-catenin antibody (Santa Cruz).
In vitro glutathione S-transferase (GST) pull-down assay. HEK293 cells were co-transfected with a combination of plasmids expressing truncated β-catenin/GST fusion proteins and ESE1 272/371 for 48 hours. The cells were lysed and 500 μg of total cell lysate was incubated with 20 μl of glutathione agarose overnight at 4°C. The pellets were washed three times with NP-40 lysis buffer and the bound proteins were eluted in a protein loading buffer at 95°C for 5 min and then subjected to a immunoblotting analysis with anti-V5 (Invitrogen) or anti-GST antibody (Santa Cruz).
Confocal microscopy and immunofluroscence. Cells were grown on glass chambers at 37°C for 24 hours, fixed, permeabilized in ice-cold methanol and blocked with 1% BSA for 30 minutes at room temperature. Cells were stained with an anti-ESE1 (Abcam) or an anti-β-catenin monoclonal antibody (Cell signaling), followed by either an anti-rabbit Alexa 488- or an anti-mouse Alexa 568-conjugated secondary antibody, respectively (Molecular Probes, Inc., Eugene, OR, USA), and examined with a Zeiss LSM510 confocal microscopy (Carl Zeiss Inc, Thornwood, NY, USA).
Results
β-catenin is a novel binding partner of ESE1. A large number of studies have shown that both ETSs (transcription factors) and β-catenin (co-factor) bind to many other proteins and fine-tune the transcriptional activity of their target genes. Recently, Wang et al. reported that ESE1 suppresses transcription of β-catenin in human colon cancer cells (12). Therefore, we tested if there is an interaction between ESE1 and β-catenin in these cells. Firstly, we compared basal expression of ESE1 and β-catenin in a panel of colon cancer lines. ESE1 and β-catenin is highly expressed in HCT116 and LoVo cells out of the four colon cancer lines tested (Figure 1A). Since both ESE1 and β-catenin are abundant in HCT116 cells, we used this cell line for reciprocal immunoprecipitation to test if two nuclear proteins (ESE1 and β-catenin) interact physically. Whole-cell extracts (WCE) were pulled-down with the anti-ESE1 antibody and immunoblotted with anti-β-catenin antibody. As a result, we detected complex formation between ESE1 and β-catenin (Figure 1B). To confirm ESE1/β-catenin interaction, whole cell lysates were immunoprecipitated with an anti-β-catenin antibody and immunoblotted with anti-ESE1 antibody. The result confirmed that ESE1 forms an endogenous protein complex with β-catenin (Figure 1C).
ESE1 physically interacts and colocalizes with β-catenin. (A) Cell lysates extracted from different human colon cancer cell lines (HCT116, LoVo, HCT15 and SW490) were used for western blot analysis using anti-ESE1 and anti-β-catenin antibodies. (B) Cell lysates extracted from HCT116 cells were immunoprecipitated using anti-ESE1 antibody (lane 1) or IgG (lane 2) and immunoblotted with anti-β-catenin antibody (left panel) or anti-ESE1 antibody (right panel). Whole-cell lysates (lane 3) were used as control. (C) The same cell lysates were immunoprecipitated using anti-β-catenin antibody (lane 1) or IgG (lane 2) and immunoblotted with anti-β-catenin antibody (left panel) or anti-ESE1 antibody (right panel). Whole-cell lysates (lane 3) were used as control. (D) Co-localization of endogenous ESE1 with endogenous β-catenin in human HCT116 cells. (i) HCT116 cells were stained with anti-ESE1 or anti-β-catenin antibodies. (ii) HCT116 cells were incubated with BSA as a negative control. Arrow indicates one area of co-localization of β-catenin with ESE1 in the cytoplasm.
Next, to test whether ESE1 and β-catenin protein localize in the same subcellular compartments, we carried out double immunofluorescence. The cells were stained with either the anti-ESE1 antibody or anti-β-catenin antibody and, then, stained with either Alexa 488 (green)- or Alexa 568 (orange red)-conjugated secondary antibodies, respectively. The two proteins extensively co-localized (yellow) in the membrane and cytoplasm regions (Figure 1D).
C-terminus of ESE1 binds to N-terminus of β-catenin. ESE1 contains multiple domains, including an amino-terminal pointed (PNT) domain, a TAD domain, a SAR domain, an A/T hook domain and a carboxy-terminal ETS domain (14). To define the binding region of ESE1 to β-catenin, we constructed several ESE1 truncated fragments with a V5 tag (Figure 2A, left) and β-catenin fragments with a GST tag (Figure 2B, left). Then, different ESE1 and β-catenin fragments were co-transfected to HEK293 cells and co-immunoprecipitation was performed. The results indicated that ESE1 binds to β-catenin through its carboxy terminus comprising amino acids 272-371 (Figure 2A, right). Interestingly, truncated ESE1 fragments with N-terminal deletion showed stronger binding affinity to β-catenin than that of full length ESE1 (Figure 2A, lanes 2, 6-8 versus lanes 1, 3) suggesting that N-terminal of ESE1 might interfere with binding to β-catenin. We further investigated what region of β-catenin is responsible for binding to ESE1. The deletion mapping for β-catenin showed that N-terminus of β-catenin (amino acids 1-271) is required to bind to ESE1 (Figure 2B, right).
ETS domain of ESE1 binds to N-terminus of β-catenin. (A) LEFT, illustration of the ESE1 structure and ESE1 fragments with V5 tag. RIGHT, HEK293 cells were transfected with full length ESE1 or ESE1 derivatives. After 48 h transfection, cell lysates were immunoprecipitated with anti-β-catenin antibody and immunopellets were subjected to western blot analysis with V5 antibody. Lane1, full length ESE1; lane2, ESE1 truncated fragment 60/371; lane3, ESE1 fragment lacking TAD (ΔTAD); lane4, ESE1 fragment 1/271; lane5, LacZ/V5 control; lane6, ESE1 fragment 128/371; lane7, ESE1 fragment 160/371; lane8, ESE1 fragment 272/371. (B) LEFT, illustration of β-catenin fragments with GST tag and ESE1 272/371 fragment with V5 tag. RIGHT, HEK293 cells were co-transfected with β-catenin fragments and ESE1 (272/371). After 48 h transfection, cell lysates were incubated with glutathione slurry and pellets were then analyzed with anti-GST or anti-V5 antibodies. Arrowheads indicate the positive bands and the numbers refer to the amino acids in which the sequences are located.
ESE1 binds to α-catenin. α- and β-catenin form a heterocomplex inside the cell (15) influencing transcriptional activity of β-catenin (16, 17). To determine whether ESE1 binds to α-catenin, we performed immunoprecipitation with both anti-α-catenin and anti-ESE1 antibodies. The results indicated that ESE1 also associates with α-catenin (Figure 3A). Next, to see whether the ESE1 ETS domain is responsible for the binding with α-catenin, we transfected HEK293 cells with ESE1 272/371 fragment tagged with V5 and pulled-down cell lysates with an anti α-catenin antibody and immonoblotted with anti-V5 antibody. The results indicated that ESE1 ETS domain (272/371) binds to α-catenin (Figure 3B, lane 2). We further confirmed protein interactions between ESE1 and α- or β-catenin after knocking-down α-catenin, β-catenin or both. HEK293 cells were co-transfected with V5-tagged ESE1 272/371 fragment and siRNAs against α-catenin, β-catenin or both, respectively. The immunoprecipitation result showed that the association of ESE1/β-catenin was abrogated with the suppression of either β-catenin or α-catenin in HEK293 cells by siRNA treatment (Figure 3C, lanes 2, 3 versus lane 4). Moreover, suppression of both α-catenin and β-catenin caused further decrease of ESE1/ β-catenin binding (Figure 3C, lane 1 versus lanes 2, 3). Taken together, these results demonstrate that ESE1 is a binding partner for both α-and β-catenin.
α-catenin binds to ETS domain of ESE1. (A) Cell lysates extracted from HCT116 cells were immunoprecipitated using anti-ESE1 (left panel, Lane 1) or anti-α-catenin antibody (right panel, lane 1) and immunoblotted using anti-α-catenin or anti-ESE1 antibodies. Normal rabbit or mouse serum (lane 2 in both panels) was used as the immunoprecipitation (IP) control. Whole-cell lysates (lane 3 in both panels) were used as control. (B) HEK293 cells were transfected with V5-tagged control (lane 1) or V5-tagged ESE1 272/371 fragment (lane 2) and immunoprecipited with anti-α-catenin antibody and then immunoblotted by anti-α-catenin or anti-V5 antibodies. (C) HEK293 cells were co-transfected with siRNAs for α-catenin, β-catenin or both and V5-tagged ESE1 272/371 fragment in different combination as indicated. The cell lysates were immunoprecipitated with anti-β-catenin antibody and immunopellets were analyzed by V5 antibody. Right panel is western blot data from whole-cell extracts (WCE) and validate knockdown of α-catenin and β-catenin.
Discussion
In the current study, we identified β-catenin as a new binding partner of the ESE1 transcription factor and characterized the nature and consequences of their interaction. Endogenous ESE1 forms a stable complex with endogenous β-catenin and the two proteins extensively co-localize in the membrane and cytoplasm regions. This interaction occurs in the C-terminus ETS domain of ESE1 and N-terminus of β-catenin. In addition, α-catenin binds to the ETS domain of ESE1 and suppression of α-catenin impaired ESE1/β-catenin interaction, indicating a significant role of α-catenin in ESE1/β-catenin interaction.
The ESE1/β-catenin interaction was mediated through the ETS domain of ESE1 and N-terminus of β-catenin. Primary functions of ETS domain recognize and bind to purine-rich consensus sequence (GGAA/T) within the promoter/enhancer region of target genes and regulate their transcriptional activity. However, emerging evidences have shown that the ETS domain could also mediate protein-protein interactions and that such interactions determine fine-tuning of transcriptional activity. For example, ETS domain of ELK1 directly interacts with serum response factor (SRF) and increases binding affinity to enhancer DNA (18). In contrast, the ETS domain of prostate-derived ETS factor (PDEF) binds to prostate tumor suppressive factor NKX-3.1 and regulates transcription of prostate-specific antigen (PSA) without affecting DNA binding ability (19). On the other hand, the multiple domain structure of ESE1 raises the versatility of biological functions. ESE1 SAR domain is sufficient to mediate transformation phenotypes of non-transformed mammary epithelial cells via a cytoplasmic mechanism (9), possibly by interacting with PAK1, which selectively phosphorylates ESE1 at Serine 207 within the SAR domain (20). Correspondingly, ESE1 possesses both putative nuclear export signal (NES) and nuclear localization signal (NLS) motifs (21). The cytoplasmic-nuclear shuttling upon pathophysiologic stimuli plays a critical role in determining the exact biological functions of ESE1 in a specific context.
The ESE1/β-catenin interaction was also mediated through N-terminus of β-catenin. N-terminus of β-catenin contains multiple phosphorylation sites, which are important for β-catenin functions (22). Thus, it is worthy to further determine whether status of β-catenin phosphorylation in N-terminus influences interaction. Although the current analysis has been confined to the interaction between the ESE1 and β-catenin, we speculate that a similar interaction may occur between the other ESE family members, on the basis of the high homology of the ETS domain they share with ESE1.
In conclusion, our results demonstrate that the interactions between ESE1 and β-/α-catenins might be a mechanism by which the ESE1 protein determines the β-catenin function and tumorigenesis in colon cancer cells.
Acknowledgements
The Authors thank Dr. Xiaotian Ming (Center for Neuroscience Research, Children's Research Institute, Children's National Medical Center, Washington DC) for providing technical help on confocal microscopy. They also thank Dr. Eric Fearon (University of Michigan) for a kind provision of pcDNA3-S33Y β-catenin expression vector. This work was supported by #RSG-11-133-01-CCE (S-H Lee) from American Cancer Society.
Footnotes
This article is freely accessible online.
Conflicts of Interest
The Authors declare that we have no conflict of interest.
- Received March 15, 2016.
- Revision received April 18, 2016.
- Accepted April 20, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
