Background: The expression of desmosomal genes in lung adenocarcinoma and lung squamous carcinoma is different. However, the regulatory mechanism of desmosomal gene expression in lung adenocarcinoma and lung squamous carcinoma remains unknown. Methods: The correlation between expression of desmosomal gene expression and SOX30 expression were analyzed by bioinformatics. The expression of SOX30, DSP, JUP and DSC3 were detected in lung cancer cell lines, lung tissues of mice and patients’ tissues by qPCR, WB, Immunofluorescence and Immunohistochemistry. A chromatin Immunoprecipitation assay was used to investigate the mechanisms of the SOX30 regulation on desmosomal gene expression. In vitro proliferation, migration and invasion assays, and an in vivo nude mice model were utilized to assess the important role of desmosomal genes on SOX30-induced tumor suppression. A WB assay and TOP/FOP flash reporter assay was used to investigate the downstream pathway regulated by the SOX30- desmosomal gene axis. A chemical carcinogenic model of SOX30-knockout mice was generated to confirm the role of the SOX30-desmosomal gene axis in tumorigenesis. Results: The expression of desmosomal genes were upregulated by SOX30 in lung adenocarcinoma but not in lung squamous carcinoma. Further mechanism studies showed that SOX30 acts as a key transcriptional regulator of desmosomal genes by directly binding to the ACAAT motif of desmosomal genes promoter region and activating their transcription in lung adenocarcinoma. Knockdown of the expression of related desmosomal genes by miRNA significantly attenuated the inhibitory effect of SOX30 on cell proliferation, migration and invasion in vitro and on tumor growth and metastasis in vivo. In addition, knockout of SOX30 promotes lung tumor development and loss the inhibition of desmosomal genes on downstream Wnt and ERK signal in urethane-induced lung carcinogenesis in SOX30-knockout mice. Conclusions: Overall, these findings demonstrate for the first time that SOX30 acts as a master switch of desmosomal genes, inhibits lung adenocarcinoma cell proliferation, migration and invasion by activating the transcription of desmosomal genes. This study provides novel insights on the regulatory mechanism of desmosomal genes in lung adenocarcinoma. Keywords: SOX30, Desmosome, Wnt, ERK, Lung adenocarcinoma * Correspondence: firstname.lastname@example.org Institute of Toxicology, College of Preventive Medicine, Third Military Medical University, 30 Gaotanyan Street, Shapingba District, Chongqing 400038, People’s Republic of China © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 2 of 15 Background In this study, our data demonstrates that the expres- Lung cancer is the leading cause of cancer-related inci- sions of desmosomal genes are regulated by SOX30 in dence and mortality throughout the world . Non-small ADC but not in SCC. Mechanistically, SOX30 activates cell lung cancer (NSCLC) is the major subtype of lung desmosomal gene expression through direct binding to cancer, which is typically divided into two histological sub- the promoter regions of the desmosomal genes in ADC types, lung adenocarcinoma (ADC) and lung squamous cells. Our findings elucidate the regulatory mechanisms carcinoma (SCC) . In addition to differences in morph- of desmosomal genes in ADC. ology, the underlying mechanisms, molecular profiling and therapeutic methods are quite different [3–5]. Therefore, Methods identifying differentially-expressed genes between ADC Cell lines and SCC is useful to better understand their pathogenesis. The lung cancer cell lines (A549, LTEP-a-2, H520 and In the process of cancer development, cell junction H226) were obtained from the Cell Bank of the Chinese molecules play a critical role in inhibiting tumor growth Academy of Science (Shanghai, China), cultured in and metastasis [6–10]. Previous evidence revealed that RPMI-1640 medium supplemented with 10% fetal bo- desmosomal proteins, as a member of cell junction vine serum (Gibco, CA). All the cells were maintained at molecules, are deregulated in various cancers, including 37 °C with 5% CO . lung cancer . In lung cancer, different members of desmosome genes have different roles in tumor progres- Plasmid construction and cell transfection sion. For example, plakophilin 1 (PKP1) can be useful in Construction of SOX30 expression vector was performed the diagnosis of patients with SCC [12, 13]. Plakophilin 3 as previously described [19, 22]. For knockdown, two pairs (PKP3) promotes tumor growth in lung cancer . of oligomeric singlestranded oligonucleotides and a pair of Decreased desmocollin 1 (DSC1) expression is associated negative oligomeric singlestranded oligonucleotides were with poor prognosis in human lung cancer . Desmo- synthesized, and then inserted into miRNA expression vec- glein 2 (DSG2) acts as an oncogene by accelerating tumor tor pcDNA6.2-GW/EmGFP-miR. Cells were transfected growth of non-small-cell lung carcinoma (NSCLC) . using Lipofectamine2000 Reagent (Invitrogen Preserva- Desmoglein 3 (DSG3) is underexpressed in lung cancer, tion, Carlsbad, CA, USA) according to the manufac- and a lower expression of DSG3 is significantly linked to turer’s instructions. The stably transfected cells were higher tumor grade . However, the downstream signal- screened under G418 (Calbiochem, La Jolla, CA, USA) ing pathway of these desmosomal genes in regulating or Blasticidin (Sigma). Cell clones were obtained by the tumor progression is not entirely clear, except for the limited diluted method. three canonical tumor suppressor desmoplakin (DSP), junction plakoglobin (JUP) and desmocollin 3 (DSC3). RT-PCR and qRT–PCR analysis The expression of DSP and JUP are reduced or absent in The RNA of cells was isolated using the Trizol reagent lung cancer and re-expression results in reduction of cell (Invitrogen, Life Technologies), and conversion of total growth and migration by suppressing the Wnt signaling RNA to cDNA was performed with the Reverse Tran- pathway in lung cancer [12–14]. DSC3 acts as a tumor scription System (Promega, Madison, WI, USA). All suppressor through inhibiting the EGFR/ERK pathway in qRT-PCR reactions was performed using the C1000 lung cancer [17, 18]. In previous studies, higher expression Real-Time Cycler (Bio-Rad Laboratories, Hercules, CA, of desmosomal genes were found in SCC tissues com- USA) and qRT-PCR Master mixes (Promega, Madison, pared with ADC tissues [2, 4, 17, 19, 20]. However, the WI, USA). Primers for amplification of the DSC1, DSG1, differentce in regulatory mechanisms of desmosomal DSC2, JUP, DSP, PKP1, DSC3, DSG3, PKP3 and ACTIN genes between ADC and SCC are still not clear. genes are listed in Additional file 1: Table S1. All experi- -ΔΔt SOX30 is a member of the Sox family of transcription ments were carried out in triplicate, and the 2 factors that has been isolated from Fugu, human, mouse method was used to determine expression of the genes and the Nile tilapia . A high SOX30 expression has a of interest. favorable and independent prognostic factor for ADC patients but not for SCC patients . Our in vitro ex- MTS periments show that SOX30 overexpression significantly For knockdown, A549 and LTEP-a-2 cells with SOX30 inhibits proliferation by inducing apoptosis in ADC but or empty vector stably expression were plated at 4 × 10 not in SCC [19, 22]. These studies indicate that SOX30 cells per well on 96-well plates, and transfected with may have different functional roles in distinct subtypes desmosomal gene miRNA or negative control. Cell pro- of lung cancer. Thus, the different downstream signal liferation was assessed using MTS Reagent (Promega, mechanism of SOX30 in these two histological subtypes Madison, WI, USA) on days 1, 2, 3, 4 and 5 after trans- of lung cancer needs to be further investigated. fection. The assays were performed in triplicate. Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 3 of 15 Colony formation assay Biotechnology; sc-717), MMP7 rabbit polyclonal antibody Stable transfected A549 and LTEP-a-2 cells (n = 1500) (1:100; Santa Cruz Biotechnology; sc-30,071), ERK1/2 were seeded into 100 mm cell culture dish and maintained rabbit polyclonal antibody (1:100; Santa Cruz Biotechnol- in media for 14 days. Surviving colonies were fixed with ogy; sc-93), p-ERK1/2 rabbit polyclonal antibody (1:100; 4% paraformaldehyde and stained with 0.1% crystal violet Santa Cruz Biotechnology; sc-101,761) and ACTIN mono- (Beyotime Biotechnology, China) for 15 min. Then, the clonal antibody (1:2000; Sigma; A5441). ImageJ software cell colonies were eluted by acetic acid and the absorbance was used to quantify the protein expression. of crystal violet were measured at 570 nm. The experi- ment was carried out in triplicate wells for three times. Immunofluorescence cell staining (IF) Lung cancer cell lines with SOX30 or vector control Boyden chamber migration/invasion assay stable expression grown on sterile glass coverslips, Transwell assays were performed by using transwell washed with PBS, fixed in 4% paraformaldehyde for plates with 8 μm pore (Corning; 3422). For the migra- 15 min and permeabilized by 0.5% Triton for 15 min at tion assay, 1 × 10 Cells were resuspended in serum-free room temperature. Following blocking with 1% bovine RPMI-1640 medium and plated onto the 24-well upper serum albumin for 30 min, rabbit polyclonal antibody chamber with the lower chamber filled with complete SOX30 (1:100; Santa Cruz Biotechnology; sc-20,104), medium. For the invasion assay, the 24 transwell plate mouse monoclonal antibody DSP (1:100; Santa Cruz with matrigel (BD biosciences) polymerized at the Biotechnology; sc-365,981), mouse monoclonal antibody 24-well upper chamber for 2 h at 37 °C as the interven- JUP (1:100; Santa Cruz Biotechnology; sc-33,634) and ing invasive barrier. After 24 h incubation at 37 °C, the mouse monoclonal antibody DSC3 (1:100; Santa Cruz cells on the upper chamber were removed and the num- Biotechnology; sc-81,806) were incubated overnight at ber of cells that migrated or invaded to the lower side 4 °C. After washing three times, the cells were incubated were fixed in 4% paraformaldehyde and stained with 1% with an appropriate fluorochrome-conjugated secondary crystal violet and counted at × 200 magnification in 10 antibody (1:300; Invitrogen; A-11010 and A-11003) for 1 h different fields of microscope (Leica; D-35578). The re- at 37 °C in the dark. After blue nuclear counterstaining sults were determined from three repeated experiments. with 4, 6-diamidino-2-phenylindole (Beyotime, Shanghai, China; C1006) for 10 min at room temperature, coverslips In vivo tumorigenicity assay were mounted and observed with a fluorescence micro- For the xenograft tumor growth assay, a total of 1 × 10 scope (Zeiss, Oberkochen, Germany; LSM800). ZEN soft- stable transfected A549 cells suspended in 150 μlPBS ware was used to quantify the protein expression. were injected subcutaneously into the right flanks of 5-week-old male Balb/c nude mice (n =4 mice per group), Patient sample and immunohistochemical (IHC) analysis respectively. Tumors size were measured every 3–5days Tissue microarrays contained a total of 30 lung cancer pa- with calipers after injection, and the tumor volume was tient tissue samples including 15 ADCs and 15 SCCs were calculated based on formula: 0.5 × (length × width ). After obtained from the collaboration (Shanghai Biochip Co Ltd., 42 days housing, the mice implanted tumors were sacri- Shanghai, People’s Republic of China). The rabbit polyclonal ficed, and liver tissues were dissected and subjected to antibodies used were SOX30 rabbit polyclonal antibody histological examination. Metastatic nodules were de- (1:100; Santa Cruz Biotechnology; sc-20,104), DSP rabbit tected by H&E staining and quantified by counting meta- polyclonal antibody (1:100; Santa Cruz Biotechnology; static lesions in each section. The images of the positive sc-33,586), JUP rabbit polyclonal antibody (1:100; Santa areas were taken. All experimental animal procedures Cruz Biotechnology; sc-7900), DSC3 rabbit polyclonal were approved by the Institutional Animal Care and Use antibody (1:100; Santa Cruz Biotechnology; sc-48,750). IHC Committee of Third Military Medical University, China. staining was performed as described previously . WB analysis Chromatin immunoprecipitation (ChIP) assay WB was performed as previously described . ACTIN ChIP analysis was performed using a ChIP Assay Kit was used as a loading control. The following primary (Pierce, Rockford, IL, USA). The immunoprecipitated and antibodies were used: SOX30 rabbit polyclonal antibody input DNA was used as a template for RT–PCR analysis (1:100; Santa Cruz Biotechnology; sc-20,104), DSP rabbit using the primers listed in Additional file 1: Table S1. polyclonal antibody (1:100; Santa Cruz Biotechnology; sc-33,586), JUP rabbit polyclonal antibody (1:100; Santa Chemoncogenic model Cruz Biotechnology; sc-7900), DSC3 rabbit polyclonal C57BL/6 J (SOX30 conditional knockout) mice were antibody (1:100; Santa Cruz Biotechnology; sc-48,750), obtained from the model animal research center of CCND1 rabbit polyclonal antibody (1:100; Santa Cruz Nanjing University (Nanjing, China) and maintained Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 4 of 15 in a controlled environment (12-h light/dark cycle, ad libi- Desmosomal gene expression positively correlates with tum access to food and water) at the laboratory animal fa- SOX30 expression in ADC cility of Third Military Medical University (Chongqing, Our previous research suggested that SOX30 is a tumor China). The W/W mice (n =5) and K/K mice (n =5) that suppressor in ADC but not in SCC [19, 22]. Thus, we were used were matched for age (14 weeks) and weight suspect that the expression of desmosomal genes are as- (16–19 g). These mice received a intraperitoneal injection sociated with SOX30 expression. Therefore, we analyzed of urethane (1 g/kg in 100 μl saline) once a week for 10 the correlation of SOX30 expression with the 8 desmo- consecutive weeks. Mice were sacrificed after 4 months. somal genes identified in the previous experiments be- Lung neoplastic lesions were observed by H&E staining. tween ADC and SCC by using TCGA data. The results All experiments on mice were approved by the Institu- revealed that the expressions of desmosomal genes were tional Animal Care and Use Committee of Third Military significantly positively correlated with SOX30 expression Medical University, China. in ADC but not in SCC (Fig. 2a and b). Previous studies revealed that desmosome family members play a crucial role in tumor suppression . Therefore, we hypothe- Analysis of publicly available datasets sized that SOX30 may inhibit tumor progression GSE43580 dataset were used to analyze differential through upregulation of the desmosomal gene expres- gene expression between ADC and SCC. The protein sion in ADC but not in SCC. To examine this hypoth- expression of desmosomal genes in human lung tumor esis, qRT-PCR analyses were performed to measure the tissues were determined from the human protein atlas expression of desmosomal genes in SOX30-transfected (www.proteinatlas.org). TCGA lung adenocarcinoma A549 and LTEP-a-2 cells (two human lung adenocarcin- RNAseq data (n = 571) and TCGA lung squamous car- oma cell lines), and H520 and H226 cells (two human cinoma RNAseq data (n =553) were used to compare lung squamous carcinoma cell lines). Consistently, the the expression of SOX30 with desmosomal genes. overexpression of SOX30 upregulated the expressions of the 8 desmosomal genes in the A549 and LTEP-a-2 cells but not in the H520 and H226 cells (Fig. 2c and d). Statistical analysis Among these 8 desmosomal genes, the most obvious Statistical analyses were performed with the SPSS 16.0 change in expression is three tumor suppressor genes software (SPSS, Inc., Chicago, IL, USA). Each experi- for lung cancer, DSP, JUP and DSC3. To identify the ment was performed at least three times. The data were functional target gene of SOX30, we determined the presented as the means ± SD. Results of expression protein expression of these desmosomal genes by WB analyses, cell proliferation, colony formation, migration analyses and immunofluorescence cell staining. The re- and invasion were analyzed using the two-tailed Stu- sults showed that these genes were upregulated in the dent’s t-test. Correlation analysis of gene expression SOX30 high expression group of A549 and LTEP-a-2 was performed using Spearman’s rank correlation coef- cells but not in the H520 and H226 cells (Fig. 2e and f, ficient analysis. A two-sided P-value<0.05 was taken as Additional file 2: Figure S1). To further determine the statistically significant. correction between SOX30 and these genes, we tested for expression of SOX30, DSP, JUP and DSC3 in human Results lung cancer tissues and adjacent tissues. The expression Identification of differentially expressed genes between of DSP, JUP and DSC3 were associated with SOX30 ex- ADC and SCC pression in human ADC tissues but not in human SCC To analyze the differential gene expression between tissues (Fig. 2g and h). ADC and SCC, we first performed pathway analysis of one large-sample lung cancer datasets-GSE43580 and SOX30 upregulates the expression of DSP, JUP and DSC3 found that the expression of 17 cell junction genes has by directly binding to their promoter region in lung significant differences between ADC and SCC (Fig. 1a). adenocarcinoma cells Gene ontology (GO) analysis was conducted based on Emiko et al. reported that SOX30 was able to specifically the 17 genes identified in the pathway analysis revealing recognize the promoter of genes that have an ACAAT 8 desmosomal genes: DSC1, DSG1, DSC2, JUP, DSP, motif . Furthermore, SOX30 can transcriptionally PKP1, DSC3 and DSG3 (Fig. 1b). To further confirm the activate the target gene by directly binding to the pro- GO results, we analyzed the protein expression of these moter containing the ACAAT motif ref. Subsequently, 8 desmosomal genes in clinical specimens from the analyzing the promoter region of DSP, JUP and DSC3, human protein atlas (www.proteinatlas.org). We found and found their promoter regions all contain an ACAAT that all desmosomal genes were overexpressed in SCC, motif (Fig. 3a–c). Then, we determined whether SOX30 and underexpressed in ADC (Fig. 1c and Table 1). could activate DSP, JUP and DSC3 transcription by binding Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 5 of 15 Fig. 1 The expressions of desmosomal genes are different between ADC and SCC. a Identification of differential genes between ADC and SCC with GSE- 43580 dataset. b Analysis of desmosomal gene expression levels between ADC and SCC using GSE43580 dataset. c The expression of desmosomal genes in ADC specimens and SCC specimens. Images were taken from the Human Protein Atlas (http://www.proteinatlas.org) online database Table 1 Positive percentage of desmosomal gene expression to the ACAAT motif on their promoters. Chromatin im- between ADC and SCC munoprecipitation (ChIP) assay demonstrated that SOX30 Gene Desmosomal gene expression could directly bound to their promoters in A549 and name ADC PR(%) SCC PR(%) LTEP-a-2 cells but not in H520 and H226 cells (Fig. 3d). HIGH MID LOW ND HIGH MID LOW ND DSC1 0 0 2 10 16.7 0 4 3 2 77.8 DSP, JUP and DSC3 induction are critical for SOX30- DSG1 0 0 8 16 33.3 0 5 6 5 68.8 mediated cell proliferation, migration and invasion in DSC2 0 18 4 2 91.7 0 19 4 0 100 ADC cells in vitro JUP 9 4 4 7 70.8 9 5 2 1 94.1 To define whether DSP, JUP and DSC3 mediates the DSP 0 0 4 19 17.4 0 0 2 8 20.0 effect of SOX30 on cell proliferation, migration and invasion in ADC cells, we first constructed two PKP1 0 0 0 16 0.0 7 4 2 3 81.3 SOX30-overexpressing ADC cell lines, (A549-SOX30(+)) DSC3 0 0 4 18 18.2 2 6 3 4 73.3 and (LTEP-a-2-SOX30(+)), and their negative controls DSG3 0 0 2 20 9.1 0 4 0 16 20.0 (A549-Vector) and (LTEP-a-2-Vector). We then blocked HIGH High expression, MID Middle expression, LOW Low expression, ND Not detected, PR (%) positive rate either DSP, JUP or DSC3 expression in A549-SOX30(+) Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 6 of 15 Fig. 2 (See legend on next page.) Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 7 of 15 (See figure on previous page.) Fig. 2 SOX30 is positive correlated with desmosomal gene expression in ADC, but not in SCC. a Heatmaps for correlations between SOX30 and desmosomal genes in the TCGA lung adenocarcinoma RNAseq (IlluminaHiSeq; n = 571) data set. b Heatmaps for correlations between SOX30 and desmosomal genes in the TCGA lung squamous carcinoma RNAseq (IlluminaHiSeq; n = 553) data set. Correlation coefficient R and P-values were calculated by a Spearman correlation analysis. c qRT-PCR analysis of desmosomal gene expression in A549 and LTEP-a-2 cells transiently transfected with the vector control or SOX30 expression vector. d qRT-PCR analysis of desmosomal gene expression in H520 and H226 cells transiently transfected with the vector control or SOX30 expression vector. ACTIN was used as an internal control. e The protein levels of DSP, and JUP and DSC3 were monitored by WB after SOX30 overexpression in A549 and LTEP-a-2 cells. f The protein levels of DSP, and JUP and DSC3 were monitored by WB after SOX30 overexpression in H520 and H226 cells. g The protein levels of SOX30, DSP, JUP and DSC3 were further monitored by IHC in human ADC tissues. h The protein levels of SOX30, DSP, JUP and DSC3 were further monitored by IHC in human SCC tissues. Scale bar represents 50 mm Fig. 3 SOX30 directly binding to promoters of DSP, JUP and DSC3. a–c The ACAAT motif in promoter region of DSP, JUP and DSC3 were underlined. TSS, transcription start site. d ChIP-PCR was performed to identify DSP, JUP and DSC3 are direct binding targets of SOX30 Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 8 of 15 Fig. 4 (See legend on next page.) Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 9 of 15 (See figure on previous page.) Fig. 4 Desmosomal gene silencing diminished the effect of SOX30 overexpression on proliferation, migration and invasion of A549 and LTEP-a-2 cells. a SOX30, DSP, JUP and DSC3 expression were confirmed by WB in A549 and LTEP-a-2 cells. b MTS assays were performed to analyze cell proliferation of A549 and LTEP-a-2 cells cotransfected with desmosomal gene miRNA and SOX30 expression vector. *P < 0.05; **P < 0.01. c The effect of increasing SOX30 expression with decreasing desmosomal gene expression on cell growth was further confirmed by colony formation assay in A549 and LTEP-a-2 cells. d Assessment of the effects of the expression of SOX30 and desmosomal gene, respectively, on cell migration and invasion in A549 and LTEP-a-2 cells and (LTEP-a-2-SOX30(+)) cells by using miRNA; thus es- suppress the Wnt/β-catenin signaling pathway through tablishing ADC cell lines that are SOX30-overexpressing upregulation of DSP and JUP. To test this hypothesis, the with either DSP, JUP or DSC3 knockdown. The expression protein expression levels of SOX30, DSP, JUP, MMP7 and of SOX30, DSP, JUP and DSC3 were verified by WB ana- CCND1 in A549 stable cell lines and xenograft tumors lysis (Fig. 4a). The MTS, colony formation assay revealed were first examined by WB. Consistent with our theory, that the knockdown of DSP, JUP or DSC3 by miRNA the protein expression of MMP7 and CCND1 were significantly abrogated the effects on cell proliferation in- downregulated in a SOX30-overexpressing cell line duced by SOX30 overexpression in A549 and LTEP-a-2 and xenograft tumors; MMP7 and CCND1 expression cells (Fig. 4b–c). The transwell assay was performed to were increased significantly in SOX30-overexpressing determine whether desmosomal genes are involved in cells, or tumor tissue after DSP or JUP expression SOX30-mediated migration and invasion of ADC cells. was blocked (Fig. 6a). We then conducted a TOP/ The results showed that interference with the expression FOP FLASH reporter assay in SOX30-overexpressing of DSP, JUP or DSC3 reversed the inhibitory effect of A549 cells. As shown in Fig. 7a, overexpression of SOX30 on cell migration and invasion (Fig. 4d). SOX30 significantly inhibited the transcription activity of β-catenin/T-cell factor (TCF) compared with cells DSP, JUP and DSC3 are involved in SOX30-mediated transfected with the vector control (Fig. 6b). Previous tumor suppression of ADC in vivo studies have reported that DSC3 acts as a tumor sup- To further investigate whether upregulation of DSP, JUP pressor through inhibiting ERK signaling in human and DSC3 are necessary for SOX30-medicated tumorigen- lung cancer . WB assay were performed to test whether esis and metastasis in vivo, we utilized a xenograft tumor these changes affect ERK1/2-related signaling activity in model using these stable cell lines. As showed in Fig. 5a–c, SOX30-transfected A549 cells. The data showed that xenografted (A549-SOX30(+)/DSP(−)), (A549-SOX30(+)/ overexpression of SOX30 significantly reduced the phos- JUP(−)) and (A549-SOX30(+)/DSC3(−)) cells rapidly pro- phorylation of ERK1/2 whereas the knockdown of DSC3 liferated in mice compared with (A549-SOX30(+)) cells, significantly abrogated the SOX30 overexpression-induced and we found that tumor volume of (A549-SOX30(+)/ dephosphorylation of ERK1/2, suggesting that DSC3 is DSP(−)), (A549-SOX30(+)/JUP(−)) and (A549-SOX30(+)/ involved in the SOX30-regulated dephosphorylation of DSC3(−)) were significantly larger than those with ERK1/2 in vitro and in vivo (Fig. 6c). Therefore, these (A549-SOX30(+)) cells. Moreover, H&E staining revealed results indicated that SOX30 overexpression suppresses that knockdown of DSP, JUP or DSC3 dramatically in- Wnt/β-catenin signaling and ERK signaling by increasing creased the number of metastatic foci in the liver of nude the expression levels of desmosomal genes, and this may mice (Fig. 5d), and these results were further confirmed contribute to inhibition of cell growth, migration and by qRT-PCR using a human-specific β2-MG (beta-2-mi- invasion in ADC. croglobulin) with a mouse-specific β2-MG as an internal control (Fig. 5e). These results indicate that DSP, JUP The accelerating urethane-induced lung tumorigenesis of and DSC3 play critical roles in SOX30-mediated growth SOX30 loss is associated with desmosomal genes and metastasis of ADC. To further define the important role of the “SOX30-des- mosomal gene axis” axis in lung carcinogenesis, we first SOX30 expression impairs Wnt signaling and ERK analyzed the correlation of SOX30 expression with DSP, signaling through induction of desmosomal gene JUP and DSC3 in lung tissues of SOX30 knockout mice expression by qRT-PCR, WB and IHC analyses. Consistently, the Previous research has shown that a member of SOX expression levels of DSP, JUP and DSC3 were signifi- family mediates the Wnt signaling pathway [25–27]. cantly decreased in lung tissues of SOX30-knockout According to our observation presented above and con- mice. (Fig. 7a–b and Additional file 3: Figure S2). sidering the facts that DSP and JUP can suppress lung Next, we compared the lung carcinogenesis in W/W cancer progression by inhibition of the Wnt/β-catenin (wild type of SOX30) and K/K (homozygous deletion of signaling pathway, we hypothesized that SOX30 can SOX30) mice treated with urethane, an environmental Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 10 of 15 Fig. 5 (See legend on next page.) Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 11 of 15 (See figure on previous page.) Fig. 5 Knockdown of desmosomal gene expression attenuates the effects of SOX30 on impeding tumor formation and metastasis in vivo. a–c Evaluation of tumorigenesis in nude mice subcutaneously injected with (A549-Vector), (A549-SOX30(+)), (A549-SOX30(+)/DSP(−)), (A549-SOX30(+)/ JUP(−)) and (A549-SOX30(+)/DSC3(−)). To assess the effects of desmosomal gene on subcutaneous tumor growth, tumor volumes were recorded on the indicated days, and pictures of solid tumor tissues were taken after 42 days. Tumor weights in the indicated groups were determined. Error bars indicate s.d. (n = 4). d Liver metastases were observed by H&E staining in the indicated groups. Arrows indicate the metastatic loci. e Liver metastasis was further quantified using RT-qPCR. Human-specific β2-MG levels were used to quantify metastatic human cancer cells with the mouse-specific β2-MG level as an internal control. Error bars indicate s.d. (n = 4).*P < 0.05; **P < 0.01 carcinogen that induces ADC in C57BL/6 J mice (data regulatory mechanisms of desmosomal genes in ADC not shown) . H&E staining showed that the number and SCC, we analyzed the correlation between desmo- and size of lung tumors were enhanced in K/K mice somal gene expression and SOX30 expression in ADC compared with W/W mice (Fig. 7c). The results sug- and SCC using TCGA data. We found a positive correl- gested that SOX30 plays a vital role in chemical carcino- ation between desmosomal genes and SOX30 in ADC genesis. To further evaluate whether the promotion of but not in SCC. Further studies demonstrated that lung tumorigenesis in K/K mice was related to the SOX30 upregulated the expressions of desmosomal downregulation of the desmosomal gene expression with genes by directly binding to the ACAAT motif of their activation of the Wnt and ERK pathway, the protein promoter regions in ADC cells, but loss of regulatory level of SOX30, DSP, JUP, DSC3, CCND1, MMP7, function in SCC cells. Because SOX30 acts as a tran- ERK1/2 and p-ERK1/2 were determined by WB in the scription factor, we suspected that there is another lung tumor tissues. The protein expression of desmosomal molecule necessary for regulatory function of SOX30, genes were significantly decreased, and CCND1, MMP7, and this molecule interferes the function of SOX30 in ERK1/2 and p-ERK1/2 were increased in lung tumor SCC. Here, we found other desmosomal genes, including tissues of K/K mice compared to W/W mice (Fig. 7d). DSC1, DSG1, DSC2, PKP1 and DSG3, were also up-regulated by SOX30 (Fig. 2c). Moreover, according to Discussion previous reports, PKP3 is dysregulated in lung cancer Desmosomes are intercellular junctions that tether inter- . Therefore, we analyzed the correlation between mediate filaments to the plasma membrane and repre- SOX30 and PKP3 and found that there is no correlation sent the major adhesive cell junctions of epithelial cells between the two (Additional file 4:Figure S3a). The . It has been reported that desmosome family mem- qRT-PCR results confirmed that PKP3 expression is not bers play a crucial role in tumor growth and metastasis upregulated by SOX30 overexpression (Additional file 4: [21, 31, 32]. Many studies suggest that the expressions Figure S3b). We then analyzed the promoter regions of of desmosomal genes can be regulated by transcription these desmosomal genes and found all of them contain an factors. For instance, SIP1/ZEB2 induces EMT partly by ACAAT motif except PKP3 (data not shown), indicating repressing the expressions of DSP and PKP2 . DSG4 that SOX30 upregulated their expression possibly by bind- is repressed by HOXC13, LEF1 and FOXN1 in hair shaft ing to their promoter regions and activating transcription. differentiation . The p53 induced expression of Notably, this hypothesis should be empirically tested, and DSC3 implicated in human lung cancer . Stat3 regu- further studies are required to address this aim. lates DSG3 transcription in epithelial keratinocytes . To define whether desmosomal genes are involved in the KLF5 mediates the transcription of DSG2 in mouse co- antitumor effects of SOX30, we focused on the most lons . All these data indicated that transcription fac- differentially expressed desmosomal genes associated with tors may be important in the regulation of desmosomal SOX30 and chose the three prominent tumor suppressors genes. Nevertheless, prior to this study, the regulatory DSP, JUP and DSC3. These three desmosomal genes were mechanisms of desmosomal genes in tumorigenesis the main targets through which SOX30 suppresses the remained largely uninvestigated. tumorigenesis and metastasis of ADC. In vitro and in vivo In this study, we found differences in expression of assays demonstrated that desmosomal genes, DSP, JUP and desmosmal genes between ADC and SCC. These result DSC3 were required for SOX30-medicated cell proliferation indicate that the regulatory mechanisms of desmosomal and metastasis in ADC. Taken together, these results sug- genes between ADC and SCC are different. Previously, gested that the “SOX30-desmosomal genes axis” may act as we identified SOX30 as an epigenetically silenced tumor tumorsuppressorinADC.Toobtainfurther insightinto suppressor, inhibiting cell proliferation and inducing cell the downstream signaling pathway of SOX30-desmosomal apoptosis in ADC but not in SCC. Thus, we speculated genes axis in inhibiting tumor growth and metastasis of that the expressions of desmosomal genes might be ADC, we detected the change of Wnt and ERK signaling regulated by SOX30 in ADC. To further explore the pathway related molecules by WB assays and demonstrated Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 12 of 15 Fig. 6 SOX30 inhibits Wnt and ERK signaling by accelerating the expression of desmosomal gene. a The protein expression of SOX30, DSP, JUP, CCND1 and MMP7 were detected in A549-SOX30 cells and subcutaneous tumors of mice. b A549 cells with SOX30 or vector control stable expression were co-transfected with luciferase reporter constructs (TOPFlash or FOPFlash, respectively) and Renilla luciferase constructs (normalizing transfection control). The reporter activity was measured 24 h after transfection. Error bars represent the s.d. of three independent experiments. ** SOX30 vs vector control, Student’s t-test, P < 0.01. c The protein expression of SOX30, DSC3, ERK1/2 and p-ERK1/2 were detected in A549-SOX30 cells and subcutaneous tumors of mice that SOX30 suppressed Wnt and ERK signal in a desmo- carcinogenesis model of SOX30-knockout mice was somal gene dependent manner. employed. We found that knockout of SOX30 promotes To further evaluate the role of the SOX30-desmosomal lung tumor formation in mice with urethane treatment and gene axis in lung tumor development, the chemical is associated with desmosomal gene loss. These results Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 13 of 15 Fig. 7 Inhibition of SOX30 promotes tumorigenesis of lung cancer with urethane treatment. a The mRNA expression of DSP, JUP and DSC3 in lung tissues of SOX30-knockout mice were measured by qRT-PCR. b The protein expression of DSP, JUP and DSC3 in lung tissues of SOX30- knockout mice were detected by WB. c Lung tissues were processed and stained with H&E for detection of tumor foci. d Tumor extracts were analyzed by WB. ACTIN was used as a loading control. e Schematic diagram of the mechanisms of SOX30 mediated suppression of ADC cell proliferation and metastasis based on our study suggested that SOX30-desmosomal gene axis might be in- desmosomal genes. SOX30 has pivotal roles in ADC volved in the occurrence and development of lung tumor. cell proliferation and metastasis in vitro and in vivo To evaluate this hypothesis, we plan to establish a SOX30 through the Wnt signal and ERK signal by directly conditional knockout mice to further clarify the important promoting transcriptional activating of desmosomal role of the “SOX30-desmosomal gene axis”. gene expression. These results provide potential novel mechanisms for the regulation of desmosome Conclusions genes and new mechanistic insight into the molecu- In conclusion, our results are the first to strongly indi- lar pathogenesis of SOX30-mediated ADC growth cate a key role of SOX30 in regulating most of the and metastasis. Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 14 of 15 Additional files Received: 11 January 2018 Accepted: 21 May 2018 Additional file 1: Table S1. Primers used in this study. Table S2. The statistical analysis between SOX30 and desmosomal genes. (DOCX 19 kb) References Additional file 2: Figure S1. SOX30 upregulate the expression of 1. Torre LA, Bray F, Siegel RL, Ferlay J, Lortettieulent J, Jemal A. Global cancer desmosomal gene in ADC cells but not in SCC cells. a The protein levels statistics, 2012. CA Cancer J Clin. 2015;65:87–108. of DSP, and JUP and DSC3 were monitored by IF after SOX30 2. Hou S, et al. Evidence, mechanism, and clinical relevance of the overexpression in A549 and LTEP-a-2 cells. b The protein levels of DSP, Transdifferentiation from lung adenocarcinoma to squamous cell and JUP and DSC3 were monitored by IF after SOX30 overexpression in carcinoma. Am J Pathol. 2017;187:954. H520 and H226 cells. Scale bar represents 30 mm. (TIF 1116 kb) 3. Wang T, Zhang L, Tian P, Tian S. Identification of differentially-expressed Additional file 3: Figure S2. The expression of DSP, JUP and DSC3 genes between early-stage adenocarcinoma and squamous cell carcinoma positively correlates with SOX30 expression in lung tissues of mice. lung cancer using meta-analysis methods. Oncol Lett. 2017;13:3314. (TIF 3837 kb) 4. Zhan C, Yan L, Wang L, et al. Identification of immunohistochemical markers for distinguishing lung adenocarcinoma from squamous cell carcinoma. Additional file 4: Figure S3. The expression of PKP3 is not not J Thorac Dis. 2015;7(8):1398–405. associate with SOX30 expression in ADC. a Heatmaps for correlations 5. Faruki H, Mayhew GM, Serody JS, Hayes DN, Perou CM, Lai-Goldman M. between SOX30 and PKP3 in the TCGA lung adenocarcinoma RNAseq Lung adenocarcinoma and squamous cell carcinoma gene expression (IlluminaHiSeq; n = 571) and TCGA lung squamous carcinoma RNAseq subtypes demonstrate significant differences in tumor immune landscape. (IlluminaHiSeq; n = 553) data set. Correlation coefficient R and P-values J Thorac Oncol Off. 2017;12(6):943–53. were calculated by a Spearman correlation analysis. b qRT-PCR analysis of 6. Mao XY, Li QQ, Gao YF, Zhou HH, Liu ZQ, Jin WL. Gap junction as an PKP3 expression in A549, LTEP-a-2, H520 and H226 cells transiently intercellular glue: emerging roles in cancer EMT and metastasis. Cancer Lett. transfected with the vector control or SOX30 expression vector. ACTIN was 2016;381(1):133–7. used as an internal control. (TIF 826 kb) 7. T A, M M, CC N, PD L. L. DW. Gap junctions and cancer: communicating for 50 years. Nat Rev Cancer. 2017;17(1):74. 8. Salvador E, Burek M, Förster CY. Tight junctions and the tumor Abbreviations microenvironment. Curr Pathobiol Rep. 2016;4(3):1–11. ADC: Lung adenocarcinoma; ChIP: Chromatin Immunoprecipitation; 9. Martin TA. The role of tight junctions in cancer metastasis. Semin Cell Dev DSC1: Desmocollin 1; DSC3: Desmocollin 3; DSG2: Desmoglein 2; Biol. 2014;36(1):224–31. DSG3: Desmoglein 3; DSP: Desmoplakin; IF: Immunofluorescence cell 10. Knights AJ, Funnell AP, Crossley M, Pearson RC. Holding tight: cell junctions staining; IHC: Immunohistochemistry; JUP: Junction plakoglobin; and cancer spread. Trends. Cancer. Res. 2012;8:61–9. NSCLC: Non-small cell lung cancer; PKP1: Plakophilin 1; PKP3: Plakophilin 11. Celentano A, Mignogna MD, Mccullough M, Cirillo N. Pathophysiology of 3; SCC: Lung squamous carcinoma; SOX30: SRY-related HMG-box 30; the Desmo-Adhesome. J Cell Physiol. 2016;232(3):496–505. WB: Western Blot 12. Sanchezpalencia A, et al. Gene expression profiling reveals novel biomarkers in nonsmall cell lung cancer. Int J Cancer. 2011;129(2):355–64. 13. Gómez-Morales M, et al. Differential immunohistochemical localization of Funding desmosomal plaque-related proteins in non-small-cell lung cancer. This work was supported by the National Natural Science Foundation of Histopathology. 2013;63(1):103–13. China [No. 81773461 and 81573179]. 14. Furukawa C, et al. Plakophilin 3 oncogene as prognostic marker and therapeutic target for lung cancer. Cancer Res. 2005;65(16):7102–10. 15. Cui T, et al. Diagnostic and prognostic impact of desmocollins in human Availability of data and materials lung cancer. J Clin Pathol. 2012;65(12):1100–6. The datasets generated and/or analyzed during the current study are available 16. Feng C, Zhu Q, Miao Y, Shen S, Xin S, Yi S. Desmoglein-2 is overexpressed in the: https://www.ncbi.nlm.nih.gov/gds/, https://www.proteinatlas.org/and in non-small cell lung cancer tissues and its knockdown suppresses NSCLC http://xena.ucsc.edu/welcome-to-ucsc-xena/ hgHeatmap/. growth by regulation of p27 and CDK2. J Cancer Res Clin Oncol. 2017; 143(1):59–69. Authors’ contributions 17. Saaber F, Chen Y, Cui T, Yang L, Mireskandari M, Petersen I. Expression of HXL, HF, CJ, AL, LWB and LJY were responsible for the experimental desmogleins 1-3 and their clinical impacts on human lung cancer. Pathol design. HXL, MBJ and ZN contributed to the execution of experiments, Res Pract. 2015;211(3):208–13. data statistics, and manuscript composition. HF, ZN, CHQ, and LJY 18. Han F, et al. Epigenetic regulation of Sox30 is associated with testis participated in performing the experiment and in the manuscript development in mice. PLoS One. 2014;9(5):97203. mapping and submission. HXL, HF, MBJ, LWB, YL, JX and LJY 19. Han F, et al. SOX30, a novel epigenetic silenced tumor suppressor, participated in the discussion and interpretation of data. HF, YL, JX, CJ promotes tumor cell apoptosis by transcriptional activating p53 in lung and LJY conceived the study and revised the manuscript. HF, LWB, CJ cancer. Oncogene. 2015;34(33):4391–402. and LJY was responsible for the funding application, and the supervision 20. Lv J, et al. PCDH20 functions as a tumour-suppressor gene through and management of the project. All authors have contributed to and antagonizing the Wnt/β-catenin signalling pathway in hepatocellular approved the final manuscript. carcinoma. J Viral Hepat. 2015;22(2):201–11. 21. Dusek RL, Attardi LD. Desmosomes: new perpetrators in tumour suppression. Nat Rev Cancer. 2011;11(5):317. Ethics approval 22. Han F, et al. High expression of SOX30 is associated with favorable survival Animal research was approved by the Institutional Animal Care and Use in human lung adenocarcinoma. Sci Rep. 2015;5:13630. Committee of Third Military Medical University. 23. Osaki E, et al. Identification of a novel Sry-related gene and its germ cell- specific expression. Nucleic Acids Res. 1999;27(12):2503. 24. Lupberger J, Kreuzer KA, Baskaynak G, Peters UR, Le CP, Schmidt CA. Competing interests Quantitative analysis of beta-actin, beta-2-microglobulin and The authors declare that they have no competing interests. porphobilinogen deaminase mRNA and their comparison as control transcripts for RT-PCR. Mol Cell Probes. 2002;16(1):25–30. 25. Kormish JD, Sinner D, Zorn AM. Interactions between SOX factors and Wnt/ Publisher’sNote beta-catenin signaling in development and disease. Dev Dyn. 2010;239:56–68. Springer Nature remains neutral with regard to jurisdictional claims in 26. Xu YR, Yang WX, Yang WX. SOX-mediated molecular crosstalk during the published maps and institutional affiliations. progression of tumorigenesis. Semin Cell Dev Biol. 2016;63:23–34. Hao et al. Journal of Experimental & Clinical Cancer Research (2018) 37:111 Page 15 of 15 27. Xie SL, et al. SOX8 regulates cancer stem-like properties and cisplatin-induced EMT in tongue squamous cell carcinoma by acting on the Wnt/β-catenin pathway. Int J Cancer. 2017;142(6):1252–65. 28. Cui T, et al. The p53 target gene desmocollin 3 acts as a novel tumor suppressor through inhibiting EGFR/ERK pathway in human lung cancer. Carcinogenesis. 2012;33(12):2326–33. 29. Xu C, et al. Inflammation has a role in urethane-induced lung cancer in C57BL/6J mice. Mol Med Rep. 2016;14(4):3323. 30. Delva E, Tucker DK, Kowalczyk AP. The desmosome. Cold Spring Harb Perspect Biol. 2009;1:a002543. 31. Zhou G, et al. The role of desmosomes in carcinogenesis. OncoTargets Ther. 2017;10:4059–63. 32. Holthöfer B, Windoffer R, Troyanovsky S, Leube RE. Structure and function of desmosomes. Int Rev Cytol. 2007;264(3904):65. 33. Cindy V, et al. SIP1/ZEB2 induces EMT by repressing genes of different epithelial cell–cell junctions. Nucleic Acids Res. 2005;33(20):6566–78. 34. Bazzi H, Demehri S, Potter C. Desmoglein 4 is regulated by transcription factors implicated in hair shaft differentiation. Differentiation. 2009;78(5):292. 35. Mao X, Cho MJT, Ellebrecht CT, Mukherjee EM, Payne AS. Stat3 regulates desmoglein 3 transcription in epithelial keratinocytes. JCI Insight. 2017;2(9): e92253. 36. Liu Y, Chidgey M, Yang VW, Bialkowska AB. Krüppel-like factor 5 is essential for maintenance of barrier function in mouse colon. Am J Physiol Gastrointest Liver Physiol. 2017;313(5):G478–91.
Journal of Experimental & Clinical Cancer Research
– Springer Journals
Published: May 31, 2018