Medical Oncology (2019) 36:92 https://doi.org/10.1007/s12032-019-1317-6 ORIGINAL PAPER SOX4 is activated by C‑MYC in prostate cancer Hongyan Dong1,2 · Jing Hu1 · Lin Wang3 · Mei Qi1 · Ning Lu1 · Xiao Tan2 · Muyi Yang1 · Xinnuo Bai4 · Xuemei Zhan2 · Bo Han1,5 Received: 26 July 2019 / Accepted: 12 September 2019 / Published online: 27 September 2019 © Springer Science+Business Media, LLC, part of Springer Nature 2019 Abstract Although MYC proto-oncogene (C-MYC) amplifcation has been consistently reported to be a potential marker for prostate cancer (PCa) progression and prognosis, the clinicopathological and prognostic signifcance of C-MYC protein expression remains controversial. Overexpression of SOX4 has been shown to play important roles in multiple cancers including PCa. However, the link between these two critical genetic aberrations is unclear. In the current study, we showed that C-MYC was overexpressed in 16.2% (17/105) of Chinese patients with localized PCa. Overexpression of C-MYC was signifcantly associated with high Gleason scores (P = 0.012) and high Ki67 labeling index (P = 0.005). C-MYC overexpression was cor- related with cancer-related mortality and suggested to be an unfavorable prognostic factor in Chinese PCa patients (P = 0.018). Overexpression of C-MYC is associated with SOX4 overexpression in PCa tissues. Notably, SOX4 is a direct target gene of C-MYC; C-MYC activates SOX4 expression via binding to its promoter. In addition, Co-IP analysis demonstrated a physi- cal interaction between C-MYC and SOX4 protein in PCa cells. Clinically, C-MYC+/SOX4+ characterized poor prognosis in a subset of PCa patients. In total, C-MYC overexpression may contribute to PCa progression by activating SOX4. Our fndings highlight an important role of C-MYC/SOX4 in PCa progression in a subset of PCa patients. Keywords Prostate cancer · C-MYC · SOX4 · Overexpression · Prognosis Introduction Prostate cancer (PCa) is a common, heterogeneous disease with marked variability in progression [1]. Currently, the established prognostic factors (Gleason score, pathological stage and serum PSA) cannot precisely distinguish indolent Electronic supplementary material The online version of this PCas from highly aggressive cancers [1, 2]. Thus, novel article (https ://doi.org/10.1007/s1203 2-019-1317-6) contains prognostic biomarkers are urgently needed for PCa diagno- supplementary material, which is available to authorized users. sis and management. * Bo Han C-MYC is a regulatory protein of transcription that regu- [email protected] lates multiple cellular processes including cell proliferation, cell-cycle progression, transcription, diferentiation, apop- 1 The Key Laboratory of Experimental Teratology, Ministry of Education and Department of Pathology, School of Basic tosis, and cellular motility [3]. C-MYC is overexpressed in Medical Sciences, Shandong University, Jinan 250012, large varieties of tumor types and plays an important role China in tumor initiation and progression [4]. In PCa, elevated 2 Department of Pathology, Linyi People’s Hospital, Linyi, C-MYC expression has been found in PCa tissues [5–7] China and is associated with biochemical recurrence after radical 3 Research Center for Medical Biotechnology, Shandong prostatectomy [7]. The overexpression of C-MYC, in many Academy of Medical Sciences, Jinan, China cases, is associated with somatic genetic alterations includ- 4 Department of Human Biology, University of Toronto, ing translocations and gene amplifcation [4]. C-MYC ampli- Toronto M5S3J6, Canada fcation has been consistently reported to be an independent 5 Department of Pathology, Shandong University Qilu potential marker for disease progression and prognosis in Hospital, Jinan, China Vol.:(0123456789)1 3 92 Page 2 of 11 Medical Oncology (2019) 36:92 PCa [8, 9]. However, the relationship of C-MYC overexpres- protocol was approved by the Institutional Review Board of sion and C-MYC amplifcation as well as its correlation with Medicine School of Shandong University. clinicopathological parameters, disease progression, and prognosis remains controversial [9–12]. Immunohistochemistry (IHC) The sex determining region Y-box 4 (SOX4) gene is a developmental regulatory protein of transcription involved Immunohistochemical staining was performed by the stand- in the regulation of many key cellular processes, including ardized labeled streptavidin biotin kit (Dako Cytomation, apoptosis, cell-cycle control, microRNA processing, dif- USA) according to the manufacturer’s instructions. The ferentiation, and growth factor signaling [13]. Large-scale slides were incubated with rabbit polyclonal anti-SOX4 gene expression studies have identifed the overexpression antibody (1:100 dilution, Abcam, UK), mouse monoclo- of the SOX4 gene in a variety of human cancers, including nal anti-C-MYC antibody (1:200 dilution, Abcam, UK), or prostate [14], breast [15], lung [16], and colon cancers [17]. mouse monoclonal anti-Ki67 (1:100 dilution, Dako, USA), SOX4 has been recognized as one of the 64 “cancer signa- and was blindedly evaluated by two pathologists (B.H. and ture” genes, suggesting that SOX4 plays a fundamental role H.Y.D.) independently. Sections with evaluation disagree- in tumorigenesis and tumor progression [18]. Previously, ment between the reviewers were re-reviewed until a con- we have reported that SOX4 may initiate a transcriptional sensus was reached. The scoring system to validate C-MYC, program that enables the epithelial–mesenchymal transi- SOX4, and Ki67 expression was described elsewhere [19]. tion (EMT) phenotype [19]. Tiwari et al. also reported that SOX4 is a master regulator of EMT by controlling EZH2 Fluorescence in situ hybridization (FISH) expression and epigenetic reprogramming [20, 21]. More importantly, SOX4 might contribute to metastasis by coop- FISH analysis of abnormalities in C-MYC gene copy was erating with other important oncogenes to promote PCa carried out with the C-MYC/CEP8 probe mix (Dako, metastasis [22, 23]. Most recently, we have demonstrated a Denmark) according to the manufacturer’s protocol, and link between SOX4 and ERG gene in the development of a the scoring system used was previously described [8]. A subset of PCa patients [22], and that SOX4 may serve as a minimum of 50 cancer cells under each visual feld were prognostic marker for Chinese PCa patients [19]. Therefore, recorded. Cases lacking tumor tissues in both two cores were these data highlight the role of SOX4 in PCa progression excluded, and cancer tissues with very weak or no signals and metastasis. were recorded as insufciently hybridized. Amplifcation of In this study, we show that SOX4 is a direct target of the C-MYC locus was defned as a C-MYC/CEP8 ratio of C-MYC. Clinically, the overexpression of C-MYC is sig- higher than 2.0. nifcantly associated with SOX4 expression and C-MYC+/ SOX4+ characterizes poor prognosis in a subset of PCa Cell lines and plasmids patients. Collectively, our data suggest a cooperative role of C-MYC and SOX4 in PCa. The HEK293T (CRL-3216) cell line and human PCa cell lines (LNCaP and VCaP) were obtained from American Type Culture Collection (Rockville, USA) and cultured according to the manufacturer’s recommendations. Materials and methods Cell transfection Patients and tissue microarray (TMA) construction The plasmids and siRNA were transfected into PCa A total of 126 PCa patients (Qilu cohort) diagnosed between cells by Lipofectamine 2000 (Invitrogen, Carlsbad, CA, 2007 and 2014 participated in our study. The frst cohort USA) according to the manufacturer’s instructions. The consisted of 105 men with localized PCas who have under- pcDNA3.3-C-MYC plasmid (Addgene plasmid #26818) gone radical prostatectomy as monotherapy. The follow- was obtained as a gift from Dr. Derrick Rossi at Harvard up data were available from 94 patients, ranging from 3 to Medical School. 115 months (mean 35 months). The second cohort included patients with CRPC treated by transurethral resection of RNA extraction and quantitative real‑time PCR the prostate to relieve symptomatic obstruction due to the (RT‑PCR) locally advanced disease (n = 21). The initial treatment for patients was either observation or surgery. Morphology was RNA extraction and RT-PCR were performed as previously validated by two pathologists (B. H. and H. Y. D.). Written described [19]. Total RNA was extracted with Trizol rea- informed consent was obtained from each patient. The study gents following the manufacturer’s instructions (Invitrogen). 1 3 Medical Oncology (2019) 36:92 Page 3 of 11 92 GAPDH was used as an internal loading control. The details CST, USA). Actin (1:1000, ZSGB-BIO, China) was used of the primers for each gene are listed in Supplementary as a loading control. Three independent experiments were Table 1. performed. Immunoprecipitations were performed with Protein A/G Western blot and co‑immunoprecipitation (Co‑IP) Agarose beads (Beyotime, China) to assess the interaction analysis between C-MYC and SOX4 according to manufacturer’s instructions. After harvesting the cells, the supernatants Western blot was carried out as previously described [19]. were incubated overnight at 4 °C with Protein G-Sepharose The membrane was incubated with primary antibodies for beads (Sigma) conjugated with rabbit anti-SOX4, anti-Flag, C-MYC (1:1000; CST, USA), SOX4 (1:1000; Abcam, UK), and anti-C-MYC antibody. The samples were then run E-cadherin (1:1000, CST, USA), and Vimentin (1:1000, through gradient SDS-polyacrylamide gels and transferred to membranes that were probed with anti-C-MYC or anti- SOX4 antibodies. Table 1 Association of clinicopathological variables and molecular biomarkers with C-MYC overexpression Chromatin immunoprecipitation assay (ChIP) Variables C-MYC protein expression P Negative and Moderate and ChIP assay was performed according to the manufacturer’s weak (/%) strong (/%) protocol (ChIP Assay Kit, Millipore, USA). The puri- fed chromatin was immunoprecipitated by 6 μg of mouse All cases 80 (84.2) 15 (15.8) monoclonal anti-C-MYC (Abcam, UK) or mouse anti-IgG Age (years) (Beyotime, China). The purifed chromatin was assessed by ≤ 65 14 (77.8) 4 (22.2) 0.474 RT-PCR.