SOX4 interacts with plakoglobin in a Wnt3a-dependent manner in prostate cancer cells
© Lai et al; licensee BioMed Central Ltd. 2011
Received: 27 May 2011
Accepted: 19 November 2011
Published: 19 November 2011
Skip to main content
© Lai et al; licensee BioMed Central Ltd. 2011
Received: 27 May 2011
Accepted: 19 November 2011
Published: 19 November 2011
SOX4 is a developmental transcription factor that is required for differentiation and proliferation in multiple tissues. SOX4 is overexpressed in many human malignancies, but the precise role of SOX4 in cancer progression is still not well understood. Thus, the identification of additional SOX4 binding partners is essential for elucidating the mechanism of SOX4-mediated effects in cancer progression.
Here, we have adapted a one-step affinity purification method that enables rapid purification of SOX4 complexes via intracellular biotinylation of the amino-terminus of SOX4 to perform large-scale proteomics analysis. We have discovered that junction plakoglobin (JUP) interacts with SOX4 in both the cytosol and the nucleus and the interaction between SOX4 and plakoglobin is significantly increased when prostate and breast cancer cells are stimulated with WNT3A. Interactions between SOX4 and plakoglobin were further enhanced by the nuclear export inhibitor leptomycin B (LMB), suggesting that plakoglobin promotes nuclear export of SOX4. The SOX4-plakoglobin complex affected the expression of Wnt pathway target genes and SOX4 downstream targets, such as AXIN2, DICER1, and DHX9. In addition, SOX4 DNA binding activity to the promoters of DICER1, AXIN2, DHX9 and SOX4 itself was reduced by conditions that promote SOX4-plakoglobin complex formation. Conditions that enhanced SOX4-plakoglobin interactions resulted in reduced transcriptional activity of β-catenin luciferase reporters.
These data suggest that this newly identified interaction between SOX4 and plakoglobin is inhibitory and provides new insights into the role of SOX4 in key pathways in cell proliferation, development, and cancer progression.
The sex-determining region Y (SRY) box, or SOX, family consists of 20 highly conserved transcription factors in humans that play important roles in development . SOX4 is a 47-kDa protein that is encoded by a single exon gene and contains a conserved signature sequence in the high-mobility group (HMG) DNA-binding domain (DBD) related to the TCF/LEF family of transcription factors. The key effector of the canonical Wnt pathway, β-catenin, forms complexes with TCF/LEF HMG box factors and stimulates the transcription of Wnt downstream targets. Many studies have shown that β-catenin is regulated at several levels in cancer progression, but less is known about the regulation of the TCF/LEF transcription factor family. Structurally related to TCF/LEFs, several members of the SOX family, SOX17, SOX7, SOX9, and SOX4 have been implicated in regulating β-catenin activity [2–4]. Gain- and loss-of-function analyses have demonstrated that SOX17 and SOX7 proteins antagonize the Wnt pathway by competing with TCFs for β-catenin binding while SOX4 may function to stabilize β-catenin protein to help activate expression of target genes and promote cellular proliferation . Although these findings have implicated how SOX proteins regulate the transcriptional output of Wnt pathway, the precise role of SOX4 in the Wnt pathway remains unclear.
In humans, tight regulation of the levels of transcriptional factors is crucial to maintain tissue homoeostasis and many have been found to be oncogenic when their expression is deregulated or when their activity is functionally altered . For example, SOX4 is overexpressed in several cancers, such as bladder cancer, in which it is 5-fold upregulated compared with normal tissues . SOX4 is also upregulated at the mRNA and protein level in prostate cancer and this upregulation is correlated with higher Gleason score or tumor grade . In addition, SOX4 is overexpressed in leukemias, melanomas, glioblastomas, medulloblastomas , and lung cancer . SOX4 is also overexpressed in endometrial cancer through methylation-mediated silencing of miR-129-2 . Consistent with the concept that SOX4 is an oncogene, three independent studies searching for oncogenes have found SOX4 to be one of the most common retroviral integration sites, resulting in increased mRNA [12–14].
Junction plakoglobin (JUP), also known as γ-catenin, is a major component of the submembrane of adherens junctions and desmosomes in mammalian cells . It is closely related to the Drosophila segment polarity gene armadillo, which has a role in the transduction of transmembrane signals that regulate cell fate [16, 17]. Plakoglobin shares more than 76% homology with β-catenin, contains a central armadillo repeat domain flanked by the carboxyl and amino terminal domains, and functions in cell-cell junctions, along with β-catenin and α-catenin when coupled with cadherins . While it is known that β-catenin is essential in the Wnt signaling cascade, plakoglobin also binds to TCF/LEF factors, and has lower TCF/LEF-dependent transcriptional activity compared to β-catenin when endogenous β-catenin is depleted [19, 20].
Mass spectrometry is a highly sensitive technique that enables the rapid identification of proteins and also protein-protein interactions from a variety of biological samples. When combined with affinity purification, whole or targeted protein interaction networks can be elucidated . To investigate the cellular functions of SOX4, we have developed a one-step affinity purification method that enables rapid purification of SOX4 complexes in LNCaP cells, a prostate cancer cell line. Here, in this study, we identified proteins that interact with SOX4 in LNCaP cells by liquid chromatography-tandem mass spectrometry (LC-MS/MS), including plakoglobin. This novel interaction between SOX4 and plakoglobin may provide insights into the role of SOX4 in key pathways in cell proliferation, development, and cancer progression.
LNCaP, PC3M, and MDA-MB-231 cells were cultured as described  by American Type Culture Collection except using T-medium (Invitrogen) for LNCaP cells. HA-tagged SOX4 was cloned into pHR-UBQ-IRES-eYFP-ΔU3 lentiviral vector (gift from Dr. Hihn Ly, Emory University), and stable cells were isolated, as previously described . Recombinant WNT3A was purchased from R & D Systems (5036-WNP) and reconstituted in 0.1% BSA/PBS prior to use. Nuclear export inhinbitor leptomycin b (LMB) was purchased from Sigma (L2913). Cells were treated for 24 hrs with 100 ng/ml WNT3A, 20 μM LMB, or both unless otherwise noted.
The pREP4-BLRPwt-IRES-BirA-XL9 plasmid was a gift from Dr. Jeremy Boss (Emory University). The pcDNA3.1-HisA-HASOX4 was constructed as described . To generate pREP4-BLRPwt-HASOX4-IRES-BirA-XL9, HA-tagged SOX4 from pcDNA3.1-HisA-HASOX4 was excised at KpnI and XbaI sites and overhangs were filled in with Klenow fragment. Filled-in HA-SOX4 was then blunt ligated into the filled-in NotI site of pREP4-BLRPwt-IRES-BirA-XL9.
The pREP4-BLRPwt-HASOX4-IRES-BirA-XL9 and control empty vector were transfected into two 90% confluent 100 mm dishes of LNCaP cells respectively. Forty-eight hours post-transfection, the plates were then placed on ice and the cells were washed twice with ice-cold PBS, and lysed in 1 ml/plate with ice-cold IP lysis buffer (0.137 M NaCl, 0.02 M Tris pH8.0, 10% glycerol, 1% NP-40) supplemented with protease inhibitors, and harvested by scraping. Biotinlyated HA-tagged SOX4 complexes were purified by incubating with 50 ul slurry of Dynabeads® M280 Streptavidin (Invitrogen) at 4°C for 2 hrs. The beads ware washed 3 times with IP lysis buffer and eluted by boiling in Laemmli sample buffer.
Five percent of the proteins resulting from the purification were subject to 4-15% gradient SDS-PAGE and silver staining to analyze sample purity. The rest of the purified proteins were subjected to 4-15% gradient SDS-PAGE and concentrated on a very short distance (~2 mm long), and the protein in those bands were excised for in-gel digestion and Liquid chromatography coupled with tandem mass spectrometry based on an optimized protocol .
Cells were washed twice with ice-cold PBS, lysed in 1 ml/plate with ice-cold IP lysis buffer supplemented with protease inhibitors, and harvested by scraping. The whole cell lysates were pre-cleared and then incubated with 25 ul slurry of Dynabeads® M280 Streptavidin or protein G (Invitrogen) at 4°C for 2 hrs. The beads were washed 3 times with IP lysis buffer and eluted by boiling in Laemmli sample buffer before running SDS-PAGE. Western blots were performed as previously described . Antibodies to plakoglobin (13-8500, Invitrogen) and hemagglutinin (HA12CA5) were used in immunoprecipitations (IPs). Antibodies to plakoglobin (610253, BD Biosciences), SOX4 (LS-B3520, LifeSpan Biosciences), β-actin (3700S, Cell Signaling), and HA 16B12 (AFC-101P-1000, Covance Research Products) were used in Western blot.
Sub-confluent LNCaP HASOX4 stable cells were grown on glass cover slips, and serum starved for 24 hrs with 0.5% FBS (fetal bovine serum) T-Medium before treating with 100 ng/ml WNT3A, 20 μM LMB, or both for another 24 hrs. Cells were washed three times with Hank's Balanced Salt Solution (HBSS+) and fixed with 3.7% paraformaldehyde for 20 min at room temperature. Cells were permeablized with 100% ethanol for 20 min at -20°C and blocked in 3% BSA in HBSS+ for 1hr. anti-HA 16B12 (1:500 dilution) and plakoglobin antibodies (1:50 dilution) (sc-H80, Santa Cruz Biotechnology) were used as primary antibodies and were diluted in blocking buffer and incubated for 1 hr. Then cells were washed in HBSS+ and incubated in fluorescence-labeled secondary antibodies for 1hr at room temperature. Cells were washed in HBSS+ and then stained with Topro (T3605, Invitrogen) for 5 min at room temperature. Labeled cells were then washed in HBSS+ and mounted in p-phenylene diamine antifade agent. Confocal fluorescence images were captured using a laser-scanning microscope.
One 100-mm dish of LNCaP HASOX4 stable cells was grown to 80-90% confluency and serum starved for 24 hrs with 0.5% FBS T-Medium before treating with 100 ng/ml WNT3A, 20 μM LMB, or both for another 24 hrs. Crude subcellular fractionation was performed as previously described using digitonin, NP40, and RIPA lysis methods .
Ninety-percent confluent cells were harvested using the RNeasy kit (Qiagen), and reverse transcription was performed using iScript cDNA Synthesis Kit (Bio-Rad Laboratories). Quantitative real-time PCR (qPCR) was performed using iQ SYBR Green Supermix (Bio-Rad Laboratories) on a Bio-Rad iCycler using 18s or β-actin as a control, and data were analyzed using the δCt method .
One 100-mm dish of LNCaP HASOX4 stable cells was grown to 80-90% confluency. Cells were fixed with 1% formaldehyde, then lysed, and sonicated as described . Sonicated chromatin was precleared and then immunoprecipitated with 4 μg of anti-HA 12CA5 ascites or mouse IgG overnight and immunoprecipitated by Dynabeads® Protein G (10004D, Invitrogen) for 2 hrs at 4°C. Beads were washed and eluted as described . ChIP DNA was purified and then subjected to PCR amplification.
Cells grown in twelve-well tissue culture plates were transfected with 0.5 μg of either TOP-flash or FOP-flash with 0.04 μg of TK-Renilla control vector (Promega). At 6 hrs post-transfection, cells were placed in 0.5% FBS medium for recovery and serum starvation. At 24 hrs post-transfection, cells were treated with WNT3A and LMB, or both. Reporter gene activity was measured in a TD-20/20 luminometer (Turner Design) with the DLR Luciferase Assay System (Promega) after 48 hrs and was normalized for transfection efficiency by measuring Renilla luciferase activity.
The siRNA sequence for plakoglobin AGTCGGCCATTGTGCATCT was targeted at the 5' end of the gene lacking of homology with other catenin members  (Dharmacon RNA Technologies). LNCaP HASOX4 cells were transfected using Lipofectamine 2000 (Invitrogen) with plakoglobin or scramble siRNAs at final concentration of 200 nM. At 6 hrs post-transfection, cells were placed in 0.5% FBS medium for recovery and serum starvation. At 24 hrs post-transfection, cells were treated with WNT3A, LMB, or both for Western blot analysis or chromatin immunoprecipitation assay.
LC-MS/MS analysis identified junction plakoglobin (JUP) as a SOX4 binding protein via two trypsinized fragments that perfectly matched to human plakoglobin sequences (Figure 1B). To confirm this interaction, we first repeated the transient transfection of the pREP4-BLRPwt-IRES-BirA-XL9-HASOX4 or vector control plasmid into LNCaP cells, and performed streptavidin-magnetic bead based purification followed by anti-plakoglobin (JUP) immunoblotting. As expected, the result of the immunoprecipitation (IP) and western blot validated the interaction between SOX4 and plakoglobin (Figure 1C). Furthermore, we performed a reverse-IP immunoprecipitating endogenous plakoglobin and probing for transfected HA-SOX4 (Figure 1D) and endogenous SOX4 in untransfected LNCaP cells (Figure 1E). To determine that the interactions were not cell-line specific or DNA-dependent, we repeated the immunoprecipitations in primary human keratinocytes, MDA-MB-231 breast cancer cells, and PC3M prostate cancer cells (Figure 1F-I) following DNase I digestion of whole cell lysates. Taken together, these data demonstrate that SOX4 directly interacts with plakoglobin and that plakoglobin is a novel SOX4 binding partner.
While SOX4 expression is elevated in many malignancies and is tightly correlated with prostate cancer tumor grade, little is known of the mechanism by which SOX4 affects the progression of prostate cancer. Using LC-MS/MS proteomic analysis, we identified a novel SOX4 binding protein, junction plakoglobin, in LNCaP prostate cancer cells. We observed a physical association between SOX4 and plakoglobin with both epitope-tagged and endogenous SOX4. Although the interactions of SOX4 with plakoglobin were enhanced by co-treatment of WNT3A and LMB, we could detect this interaction using four different untransfected and unstimulated cell types (Figure 1F-I) treated with DNase I. In addition, confocal microscopy and co-immunoprecipitation demonstrated co-localization of SOX4 and plakoglobin in the nucleus when Wnt signaling was induced. ChIP assays showed the SOX4-plakoglobin complex affected SOX4 DNA binding activity to the AXIN2, DICER1, and DHX9 promoters that are Wnt signaling downstream genes and SOX4-associated targets. In addition, mRNA expression changes were detected in AXIN2, DICER1, and DHX9 by realtime-PCR. These data suggest that the SOX4-plakoglobin complex may inhibit Wnt signaling. Indeed, conditions that induced the increased interaction between SOX4 and plakoglobin caused β-catenin TOP-FLASH transcriptional activity to be downregulated and reduced occupancy of the c-Myc promoter by β-catenin.
Our confocal data show that plakoglobin nuclear localization is strongly enhanced by LMB co-treatment, suggesting shuttling of plakoglobin into and out of the nucleus following WNT3A stimulation. Although LMB is an artificial stimulus, others have shown that plakoglobin overexpression can lead to nuclear localization [31, 32], and plakoglobin is overexpressed  and amplified  in several types of cancer. Moreover, plakoglobin nuclear localization was much stronger in cells expressing HASOX4, suggesting cytoplasmic SOX4 may facilitate plakoglobin nuclear import. Thus, SOX4 may induce nuclear import of plakoglobin in response to WNT3A while plakoglobin destabilizes SOX4 from DNA binding, facilitating nuclear export of SOX4. It is as yet unclear what signals or modifications might tip the balance between nuclear import and export of plakoglobin-SOX4 complexes, or whether shuttling might be constitutive.
The role of plakoglobin during cancer progression is still controversial. When plakoglobin is overexpressed, it induces cell migration and mobility in HCT116 cells, suggesting that plakoglobin may have some oncogenic effects . In contrast, several reports have demonstrated that plakoglobin has a tumor-suppressive effect that inhibits tumor cell growth [27, 35]. The tumor suppressive activity of plakoglobin may be via nuclear translocation to antogonize β-catenin binding to TCF/LEF proteins in keratinocytes , supporting our model that plakoglobin competes with SOX4-β-catenin interactions in the nucleus.
In summary, we have demonstrated that SOX4 interacts with plakoglobin in a Wnt-dependent manner in LNCaP cells and that this complex may function to inhibit Wnt signaling. Additional studies will be required to elucidate the detailed mechanisms by which SOX4-plakoglobin interactions may affect Wnt signaling, but the role of the SOX4-plakoglobin complex provides novel insights into the role of SOX4 in Wnt signaling and prostate cancer progression.
We thank Dr. Tamara Caspary for advice on experimental conditions for LMB treatments, Dr. Chris Scharer for advice on SOX4 binding site analysis, Dr. Cornel Phillip for advice with WNT3A treatment, and Dr. Andrew Kowalczyk for providing primary human keratinocytes. This work was supported by National Cancer Institute grants R01CA106826 (C.S.M.) and P30CA138292 (J.P.).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.