Background
Renal cell carcinoma (RCC) accounts for up to 85% to 90% of all kidney cancers [
1,
2]. According to the most recent pathological classification by the International Society of Urological Pathology (ISUP), RCC mainly includes clear cell (ccRCC), papillary (pRCC) and chromophobe (chRCC) subtypes [
3], with ccRCC being the most common subtype [
4]. The 5-year survival rate of patients with kidney cancer increased approximately 30% in recent years. This improvement is largely due to low-stage and low-grade tumors being incidentally detected by improved early-detection techniques [
2,
5,
6]. However, one third of patients with kidney cancer present with metastatic disease [
5], and 20%~ 30% of the patients who undergo curative surgery relapse in distant sites during follow-up [
7]. Although there has been considerable progress in the systemic treatment of metastatic renal cell carcinoma in the past 20 years [
2], metastatic RCC remains an incurable condition for the majority of patients.
Metastasis is the primary cause of death for RCC. The initial step in the process of metastasis is the loss of cell-to-cell adhesion in the neoplastic epithelium [
8]. Tight junctions (TJs) are the most apical components of the epithelial cell junctional complex and provide a form of cell–cell adhesion [
9]. Claudin-7 (CLDN7) is a major component of TJs in epithelial cells. In RCC, CLDN7 is a marker of renal tumors originating from the distal nephron marker, such as chRCC and oncocytoma [
10]. The expression of CLDN7 is significantly lower in ccRCC than non-ccRCC [
10‐
12]. However, the clinical significance and molecular mechanisms of downregulation of CLDN7 in ccRCC remain unknown.
In this study, we confirmed that downregulation of CLDN7 due to hypermethylation may help predict aggressive tumor status and poor prognosis in ccRCC patients. Using combined data from human patients and in vitro and in vivo analyses in ccRCC cells manipulated for CLDN7 overexpression, we previously demonstrated a tumor suppressing role of CLDN7 and that it induced epithelial features in ccRCC cells. Bioinformatics and RNA sequencing analysis showed that genes that were influenced by CLDN7 overexpression were mostly enriched in cancer- and EMT-related pathways, indicating that CLDN7 may have an important role in the development and progression of ccRCC.
Methods
Human tissues
For DNA and RNA extraction, a total of 120 patients who had been pathologically diagnosed with ccRCC were included, and detailed clinicopathological features of the samples are summarized in Additional file
1: Table S1. A total of 120 paired ccRCC tissues and adjacent normal kidney tissues were immediately snap-frozen in liquid nitrogen following surgical resection. For immunohistochemical (IHC) staining, 144 patients were selected who had RCC and had undergone radical or partial nephrectomy between May 1, 2012 and August 30, 2012 at Peking University First Hospital. All the RCC paraffin specimens were pathologically diagnosed as RCC, including 129 ccRCC, 7 chRCC, 7 pRCC and 1 collecting duct RCC. This study was approved by ethics committee of the Peking University First Hospital (Beijing, China). Written informed consent was also obtained from all patients.
We downloaded TCGA RCC RNA-Seq gene expression data, TCGA ccRCC Methylation 450 K data and clinical data from UCSC Xena (
http://xena.ucsc.edu/). A total of 129 normal kidney, 534 ccRCC, 66 chRCC and 291 pRCC tissues had CLDN7 mRNA expression, and 319 ccRCC tissues had CLDN7 promoter methylation data. All 534 ccRCC tumors had clinical data for clinical correlation and survival analyses. The heat map and the correlation between CLDN7 expression and methylation status were further analyzed in the same patient cohort and verified using UCSC Xena data. The relationship between CLDN7 mRNA expression and promoter methylation status was searched in the ccRCC cohort in TCGA database using cBioportal (
http://cbioportal.org). Additionally, gene-set enrichment analysis (GSEA) was performed to compare differences in molecular pathways in cell processes between the low CLDN7 and high CLDN7 groups on the data from the ccRCC dataset of TCGA.
Cell culture and transfection
The cell lines (HEK-293, Caki-2, Caki-1, OS-RC-2, 786-O, 769-P, ACHN and A498) were obtained from the American Type Culture Collection (Rockville, MD, USA). Cell lines were cultured according to conditions specified by the provider. The kidney cancer cell line KETR-3 was purchased from KeyGEN BioTECH Co., Ltd. (Jiangsu, China). For overexpression of CLDN7, recombinant pGC-LV-GV287-GFP vectors with the CLDN7 mRNA (NM_001185022.1) or with a scrambled control sequence (Control) were constructed by the Genechem Company (Genechem Co. Ltd., Shanghai, China). All the viral vectors contained GFP as a marker to track lentivirus-mediated target gene expression by fluorescence microscopy. Briefly, pGC-LV-GV287-GFP-CLDN7 or a control vector mixed with pHelper1.0 and pHelper2.0 were cotransfected into HEK-293 T cells with PEI (Sigma-Aldrich). Caki-1 and A498 cells were infected via lentiviruses according to the MOI value (the number of lentiviruses per number of cells) recommended by the manufacturer. Caki-1 or A498 CLDN7 and Caki-1 or A498 CTRL cells were sorted using a flow cytometer sorter (BD FACS AriaTM SORP, New Jersey, USA). The Luciferase-pcDNA3 was a gift from William Kaelin (Addgene plasmid # 18964) [
13]. Lentiviruses were produced using a three-vector system: Luciferase-pcDNA3: viral packaging (psPAX2): viral envelope (pMD2G) at a 4:3:1 ratio in HEK-293 T cells. Caki-1 CLDN7-Luc and Caki-1 CTRL-Luc cells were selected with neomycin (200 μg/mL). The transfection efficiency was validated (Additional file
2: Figure S1).
Quantitative real-time PCR (qRT-PCR) and reverse transcription PCR (RT-PCR)
Total RNA was extracted from the tissue samples or the transfected cells using the TRIzol reagent (Invitrogen; Thermo Fisher Scientific Inc.), according to the manufacturer’s instructions. cDNA was generated using reverse transcription (TansGEN, Beijing, China). qRT-PCR was performed using the ABI PRISM 7000 Fluorescent Quantitative PCR System (Applied Biosystems, Foster City, CA, USA), according to the manufacturer’s instructions, and normalized to GAPDH
. RT-PCR was performed by electrophoresis on a 1.5% agarose gel. All experiments were repeated at least three times. The detailed primer sequences included in this study are shown in Additional file
3: Table S2.
Immunohistochemistry (IHC) and Western blot analysis
The immunohistochemistry (IHC) and IHC scoring were carried out according to protocols that have been described previously [
14]. Protein lysates were prepared by homogenization in 1% NP40 containing 1 mM PMSF and 20 μg protein was separated by SDS-PAGE. The immunoreactive bands were visualized using an Immobilon™ Western Kit (Millipore, Billerica, MA) using the SYNGENE G: BOX imaging system (Frederick, USA). Antibodies specific to CLDN7 (ab27487), BCL-2 (ab32124), PARP1 (ab32064), Caspase-3 (ab13847), E-cadherin (CDH1, ab76055), N-cadherin (CDH2, ab18203), Vimentin (ab92547) and TGFB1 (ab25121) were purchased from Abcam (Hong Kong, China). GAPDH (TA309157) and Ki-67 (TA500265) were purchased from ZSGB-BIO, Beijing, China. Cleaved-Caspase3 (Asp175) (9661S) was purchased from Cell Signaling Technology, MA, USA. The antibodies for the IHC and Western blot assays were diluted with phosphate buffered solution (PBS) or PBS plus Tween-20 (PBST), according to the manufacturer’s instructions.
Methylation-specific PCR (MSP) and bisulfite genomic sequencing PCR (BGS)
Genomic DNA (1 μg) was denatured using 0.3 M NaOH for 10 min at 37 °C. The samples were then incubated at 50 °C for 16 h after adding hydroquinone (Sigma-Aldrich, St. Louis, Missouri, USA) and sodium bisulfate (Sigma-Aldrich). Genomic DNA was analyzed via MSP using primer sets located within a CpG-rich area in the CLDN7 promoter. PCR samples were then resolved by electrophoresis on a 1.5% agarose gel. For the BGS assay, DNA was purified, and a CpG-rich promoter region was amplified by PCR. The PCR products were purified and cloned into a PCR 2.1-TA cloning vector (Invitrogen). A minimum of six positive clones from each product were selected for sequencing. The detailed primer sequences are shown in Additional file
3: Table S2.
Demethylation analysis
Caki-1 and A498 cells were seeded in six-well plates at a concentration of 1 × 105 cells per well, grown for 24 h, and then treated with 5 μM 5-Aza-2′-deoxycytidine (5-Aza-dC, A, Sigma-Aldrich) for 4 days. Cells were cultured with or without 100 nM Trichostatin A (TSA, T, Sigma-Aldrich) for the final 24 h. RNA was isolated for RT-PCR analysis and DNA was extracted for CLDN7 MSP.
MTS assay
The metabolic activity of Caki-1 cells was assessed using CellTiter 96™ AQueous Nonradioactive Cell Proliferation Assay, according to the manufacturer’s instructions (Promega, Madison, WI, USA). The optical density of the wells was measured at 450 nm using a Multiscan microplate spectrophotometer (Thermo LabSystems, Milford, MA, USA).
Ethynyl-2-deoxyuridine (EdU) incorporation assay
Cell proliferation was also determined by Ethynyl-2-deoxyuridine incorporation assay using an EdU Apollo DNA in vitro kit (RIBOBIO, Guangzhou, China), according to the manufacturer’s instructions. Experiments were repeated at least three times.
Flow cytometry analysis assay
Cell apoptosis was assayed by staining with Annexin V-APC and PI (KeyGEN BioTECH) following manufacturer’s instructions and detected by a flow cytometer (FACSCalibur, Becton Dickinson, New Jersey, USA).
Wound healing assays
Cell migration was determined via a wound-healing assay. Briefly, approximately 3 × 105 cells were seeded in 12-well plates at equal densities and grown to 80%~ 90% confluency. Artificial gaps were generated by a 200 μL sterile pipette tip after transfection. Wounded areas were marked and photographed with a microscope (Leica DM IL, Leica Microsystems, Germany) equipped with a digital camera (Leica DFC300FX).
Transwell migratory and invasive assays
For the transwell migration assay, 1000 cells were plated into the upper chambers (24-well insert, pore size 8 μm, Corning) with 100 μL serum-free PRIM-1640. The lower chambers were filled with 500 μL PRIM-1640 containing 10% fetal bovine serum. Two days later, cells under the surface of the lower chamber were washed with PBS and stained with 0.5% crystal violet for 30 min.
For the invasion assay, 2000 cells were seeded on transwells (24-well insert, pore size 8 μm, Corning) coated with 60 μL Matrigel (1:3 dilution in PBS, Product #354234, Corning Inc., NY, USA). The culture conditions were the same as described for the transwell migration assay. After 72 h, adherent cells on the lower surface were stained with 0.5% crystal violet. The number of cells on the lower surface was photographed with a microscope (Leica DM IL, Leica Microsystems) equipped with a digital camera (Leica DFC300FX).
Mouse model experiments
Animal experiments were conducted in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals with the approval of the Review Board of Peking University First Hospital, Beijing. Mice were maintained under pathogen-free conditions with regulated temperature and humidity levels. Mice were randomly assigned to cages in groups of 5 and fed ad libitum under controlled light/dark cycles.
Twenty 5-week-old male BALB/c nude mice were purchased from Vitalriver, Beijing, China. Approximately 5 × 106 Caki-1 CLDN7 or Caki-1 CTRL cells suspended in 100 μL Hank’s Balanced Salt Solution (Thermo Fisher Scientific Inc., USA) were mixed with Matrigel (1:1, Product #354234, Corning Inc., NY, USA). Then, 200 μL tumor cells were subcutaneously implanted on the right flank of 6-week BALB/c nude mice using a 28-gauge needle (Thermo Fisher Scientific Inc., USA). Tumor size was measured every fourth day and calculated using the formula: (length × width2)/2.
For the metastasis experiment, eight 5-week-old male B-NDG mice (NOD- Prkdcscid IL2rgtm1/Bcgen) that lacked mature T cells, B cells, and natural killer (NK) cells, were purchased from BIOCYTOGEN, Beijing, China. Approximately 1 × 106 Caki-1 CTRL-Luc or Caki-1 CLDN7-Luc cells were suspended in 200 μL PBS and injected into the lateral tail veins of each unanesthetized B-NDG mouse at six-weeks-old. Twenty days after injection, mice were anesthetized with isoflurane (YIPIN Pharmaceutical CO., LTD, Hebei, China). Ten minutes after D-Luciferin, sodium salt (150 mg/kg) was injected intraperitoneally, and cancer cells were detected with an in vivo imaging system, Xenogen IVIS (PerkinElmer, MA, USA). The total flux in photons per second were calculated and reported for each mouse’s lung and liver region using Living Image 4.3.1 (PerkinElmer/Caliper).
High-throughput cDNA sequencing (RNA-Seq)
The RNA-Seq experiments were performed by Novogene (Beijing, China). The RNA-seq library was prepared for sequencing using standard Illumina protocols. Briefly, total RNAs from Caki-1 CTRL and CLDN7 cells were isolated using TRIzol reagent (Invitrogen) and treated with RNase-free DNase I (New England Biolabs, MA, USA), to remove any contaminating genomic DNA. RNA extraction was performed using Dynabeads oligo(dT) (Invitrogen Dynal). Double-stranded complementary DNAs were synthesized from 1 μg of total RNA using Superscript II reverse transcriptase (Invitrogen) and random hexamer primers. Escherichia coli RNase H (New England Biolabs) were added to remove RNA complementary to the cDNA. The cDNA was then fragmented by nebulization, and the standard Illumina protocol was followed thereafter to create the mRNA-seq library. The libraries were sequenced on an Illumina HiSeq 2000 platform. Sequencing reads were aligned to the human genome (hg19) using the TopHat program (v2.1.1) set to the default parameters. Total read counts for each protein-coding gene were extracted using HTSeq (version 0.6.0) and then loaded into R package DESeq2 to calculate the differentially expressed genes with a cut-off fold change of ≥1.5 and an FDR < 0.05. Gene expression was calculated using the RPKM (reads per kilobase transcriptome per million reads) method. Experiments were repeated three times.
Statistical analyses
All data were expressed as means ± SD from at least three separate experiments. All data were analyzed using SPSS 20.0 statistical software (IBM. Chicago, IL, USA). The comparison of immunostaining intensities between the two groups was analyzed using a Mann-Whitney U test. The clinicopathological correlation with gene expression of the patients was analyzed using a chi-square test. Survival curves for patients were plotted using the Kaplan-Meier method, with log-rank tests for statistical significance. Uni- and multivariable Cox regression analyses were used to test the prognostic relevance between clinicopathological and immunohistochemical data. In case of multiple tests, a one-way ANOVA followed by Bonferroni-Holm procedure was applied. Other statistically significant differences were determined using Student’s T-test. Statistical significance was defined as a two-tailed p < 0.05. *p < 0.05, **p < 0.01.
Discussion
Previous studies have reported that CLDN7 is dysregulated in various cancers [
10‐
12,
18‐
23]. During our study, we confirmed that there is a downregulation of CLDN7 in ccRCC. Bioinformatic Data Mining found that DNA hypermethylation in the promoter of CLDN7, and MSP and BGS results confirmed the CLDN7 promoter hypermethylation in ccRCC tissues. Analysis of follow up data, clinical features and CLDN7 expression and methylation data, demonstrated that the lower expression and higher methylation status of CLDN7 were significantly associated with tumor progression and poor prognosis. Interestingly, in vitro and in vivo assays both found that CLDN7 overexpression significantly inhibited cell proliferation and induced apoptosis. Cell migratory and invasive abilities were also suppressed by CLDN7. To explore the molecular mechanism of the tumor suppressive function of CLDN7 in ccRCC, GSEA was performed to evaluate the different gene expression profiles between low- and high-CLDN7 expression groups of ccRCC patients. We found that cancer pathways and EMT-related pathways both decreased in ccRCC patients with high CLDN7 expression [
23]. Furthermore, RNA-Seq found that different expressed genes influenced by CLDN7 overexpression were enriched in pathways related to cancer and EMT, which was confirmed by qRT-PCR, Western blot and IHC staining in vitro and in vivo. Taken together, it is suggested that CLDN7 may have a fundamental role in tumor progression in ccRCC, by downregulating genes in pathways relating to cancer and EMT at the transcriptional level.
Previous study has shown that TGFB1 exposure decreased expression of CLDN7 and diminished epithelial barrier function, however, CLDN7 overexpression resulted in protection from TGFB1-mediated barrier dysfunction [
24]. Many transcription repressors, including TWIST1, ZEB-1, Musashi-2, and Snail, promote EMT and can bind to E-box motifs, thus suppressing CLDN7 expression [
25‐
30]. CLDN7, as a member of tight junctions, is often considered to be downstream of EMT. Interestingly, a recent study reported that overexpression of CLDN7 in colon cancer induced epithelial features and suppressed EMT through upregulation of Rab25 then decreased expression of p-Src and mitogen-activated protein kinase–extracellular signal–regulated kinase 1/2 [
19]. However, the present study found that in CLDN7 overexpressed ccRCC cells, Rab25 did not increased significantly. Notably, we discovered that CLDN7 downregulated TGF-beta signaling pathway significantly. Additionally, we found that the EMT inducer, TGFB1, was decreased by CLDN7. This result was validated in vitro and in vivo. Therefore, we propose a possible vicious circle between the loss of CLDN7 and upregulation of TGFB1 in ccRCC carcinogenesis and development. Further investigation of this cycle is part of the ongoing studies in our laboratory.
Finally, given the suppressive role of CLDN7 during the process of EMT induced by TGFB1 and Musashi-2 [
24,
27], it is a novel strategy to inhibit tumor progression by increasing CLDN7 expression. Studies have demonstrated that CLDN7 overexpression inhibited human colon and lung cancer invasion though EMT and MAPK pathways [
19,
31]. In addition, the EMT process itself has been shown to influence cellular resistance to a number of drugs [
32]. And CLDN7 was revealed to increase chemosensitivity through the activation of caspase pathway in lung cancer [
33]. In this study, we found that CLDN7 downregulation was associated with poorer prognosis and CLDN7 overexpression inhibited EMT-related pathways in ccRCC. Therefore, we propose that CLDN7 may serve as a biomarker or even a therapeutic target for ccRCC.
Acknowledgements
We thank the staff at the Department of Urology, Peking University Hospital, Beijing, China, and the Institute of Urology, Peking University, Beijing 100034, China, for technical support.