We confirm that all methods were performed in accordance with the relevant guidelines and regulations.
Cell lines and reagent
The human cervical cancer cell lines HeLa, SiHa, and CaSki were purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). The human uterine cervical cancer cell lines TCS and QGU were obtained from the RIKEN BioResource Center (Ibaraki, Japan). Cells were cultured in RPMI1640 medium (cervical cancer cell lines) with 10% fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin at 37°C in a humidified 5% CO2 atmosphere.
Microarray data analysis
Two gene expression profiles, i.e., including GSE9750 (GPL96) and GSE6791 (GPL570) were selected from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/gds) database. Each profile, including cervical cancer and normal cervix tissues, was analyzed with R ver. 4.0.3 software; hundreds of differentially expressed genes (DEGs) were visible.
Prognostic analysis of DDIT4
DDIT4 was submitted to OncoLnc (www.oncolnc.org) to reveal the correlation between the expression of DDIT4 and cervical cancer patients' overall survival (OS).
Immunohistochemical (IHC) analysis
Primary tumors were fixed with SUPER FIX (Kurabo, Osaka, Japan) at 4°C overnight and then embedded in paraffin blocks. Tumors were sectioned serially at 5-µm width and mounted on slides. After initial deparaffinization/hydration of paraffin-embedded cervical cancer sections on slide, they were microwaved for 20 min in 0.01% citrate buffer, pH 6.0, to unmask the epitopes. Endogenous peroxidase activity was blocked by using 0.3% hydrogen peroxide-methanol for 20 min. The slides were then incubated overnight at 4°C with primary rabbit monoclonal antibody against DDIT4 (1:100 dilution; ab63059, Abcam, Cambridge, MA). Sections were then subjected to immunoperoxidase staining using the VECTASTAIN ABC Kit (Vector Laboratories, Burlingame, CA, USA) according to the standard protocol. Finally, all sections were counterstained with hematoxylin (FUJIFILM), dehydrated, and mounted. The slides were finally microscopically visualized and photographed [18].
The tissues were obtained by surgery and adjacent tissues were used as non-cancerous controls. The patients were all female that from Nagoya university hospital during 2004–2015, whose age range was 20–73. Our study was approved by the Ethics Committee at Nagoya University (No. 2017 − 0497). The IRB issued a waiver for written consent because of no additional sample collection.
The expressions of DDIT4 and HIF-1α were determined using the H-score method, as previously described: H-score = intensity score × positive percentage [19]. The intensity was classified and scored as negative: 0, weak: 1, moderate: 2, or strong: 3. The maximum score was 300, and the samples were grouped as those with higher expression (H-score ≥ 150) or lower expression (H-score < 150) of DDIT4 or HIF-1α.
siRNA transfection
For the specific knockdown of DDIT4, HeLa and CaSki cells were cultured in six-well plates and transfected with 12.5 nM target-specific siRNAs or silencer select negative control siRNA using Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA) for 72h,, and used for further analysis [20]. The sequences of the siRNAs used to suppress DDIT4 expression were 5'- GAUACUCACUGUUCAUGAATT-3' (DDIT4 siRNA#1) and 5'-ACGCAUGAAUGUAAGAGUATT-3' (DDIT4 siRNA#2), which were designed by Nobuhisa Yoshikawa. And the sequence of negative siRNA was 5'-UUC UCC GAA CGU GUC ACG UTT-3'.
Western blotting
Cellular extracts were prepared in a modified radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl pH 7.4, 1% NP-40, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM PMSF, and protease inhibitor cocktail). The protein concentrations of the cellular extracts were measured using a BCA assay kit (Thermo Fisher Scientific, Waltham, MA). Proteins were separated using 10% SDS polyacrylamide gels in an electrophoresis system (Bio-Rad, Hercules, CA, USA). The quantity of proteins loaded per lane was 50 µg for whole cell protein, 30 µg for cytoplasmic protein, and 2 µg for nuclear protein. Proteins were transferred to polyvinylidene difluoride (PVDF) membranes. After blocking for 1 h at room temperature in 5% milk, blots were probed with a primary antibody at a 1:1000 dilution and incubated overnight at 4°C. Subsequently, the blots were washed three times and incubated for 1 h at room temperature with a 1:2000 dilution of the secondary peroxidase-conjugated antibodies, following our previously published protocol [21]. After three washes, immunoreactive bands were detected by electrochemiluminescence.
The DDIT4 (ab63059) and HIF-1α (ab179483) primary antibody for immunoblotting were obtained from Abcam. Except for them, all the other primary antibodies were purchased from Cell Signaling Technology® (CST, Danvers, MA, USA): E-cadherin (#3195), N-cadherin (#13116), Vimentin (#5741), Snail (#3879), Slug (#9585), IKKα (#11930), IKKβ (#8943), p65 (#8242), p-p65 (#3033), p50 (#3035), IκBα (#4814), PARP (#9532) and α-Tubulin (#2144). β-Actin (010-27841,FUJIFILM WAKOChemicals) was used as a loading control. Anti-rabbit (#7074) and anti-mouse (#7076) secondary antibodies were obtained from Cell Signaling Technology®. The protein bands were visualized with ImageQuant LAS 4010 imager (Cytiva, Marlborough, MA).
Wound-healing assay
Cells were seeded into six-well plates until 100% confluence. Linear wounds of equal width were then created by using pipette tips. The culture medium was changed to serum-free medium. Migrated cells were observed and photographed at 24 and 48 h, respectively, following our previously published protocol [21]. Three representative fields were selected, and the wound-healing ability was calculated as follows: migrated area/wounded area. Each area was measured using Image J software.
Transwell migration and invasion assays
Migration assays were conducted using Co-star® Transwell Permeable 5-mm inserts (8.0-µm polycarbonate membrane; Corning, Corning, NY). Invasion assays were performed similarly using Matrigel. The lower chamber was filled with medium containing 10% serum, and the top chamber contained 1 × 105 HeLa cells or 2 × 105 CaSki cells suspended in medium without serum. The plates were incubated at 37°C in 5% CO2 for 24 h (HeLa) or 48 h (CaSki). After migration, non-migrated cells remaining on the top of the inserts were removed with cotton swabs, and the cells that had migrated to the underside of the membrane were fixed with 4% paraformaldehyde and stained with May-Giemsa, following our previously published protocol [21]. Four representative fields of migrated or invaded cells were selected and counted using Image J.
Immunofluorescence
HeLa and CaSki were seeded onto four-well culture slides with a polystyrene vessel (Falcon, Tewksbury, MA). Cells were fixed with 4% paraformaldehyde for 20 min, permeabilized with 0.2% Triton X-100 for 5 min, and blocked with phosphate-buffered saline (PBS) containing 10% fetal bovine serum (FBS) for 1 h at room temperature. Cells were incubated with anti-E-cadherin antibody (Cell Signaling Technology®, #3195) and anti-vimentin antibody (Cell Signaling Technology®, #5741) by 1:200 dilution for 1 h, followed by incubation with an Alexa Fluor 488-labeled secondary antibody (Thermo Fisher, Alexa Fluor® 488 goat anti-rabbit IgG, #1966932) by 1:500 dilution for 1 h. Cells were stained with rhodamine-conjugated phalloidin (Thermo Fisher, #R415) by 1:200 dilution and DAPI (FUJIFILM Wako Pure Chemical Corporation, Cellstain® DAPI Solution, #D523) by 1:1000 dilution [22]. Then they were observed using a fluorescence microscope (Zeiss Elyra LSM880).
Nuclear protein extraction
Nuclear and cytoplasmic proteins were separated using a NE-PER Nuclear and Cytoplasmic Extraction Kit (#78833, Thermo Fisher Scientific) according to the manufacturer's instructions [23].
GSEA analysis
Gene set enrichment analysis (GSEA) of the DEGs were implemented by the clusterProfiler program developed by Yu et al. [24], which is an R package used for the analysis and visualization of functional profiles of genomic coordinates. A probability (p)-value < 0.05 was considered significant.
Construction of CaSki-shRNA-DDIT4 cells
To stably knock down the expression of DDIT4, we purchased pLV[shRNA]-EGFP: T2A cloning vectors from Vector Builder (Chicago, IL) to generate shRNA-DDIT4 (sequence: GTGTAGCATGTACCTTATTAT), which was designed by Nobuhisa Yoshikawa. And the sequence of negative shRNA was 5'- CCTAAGGTTAAGTCGCCCTCG − 3'.
The transfection of vectors used the Lenti-X Lentiviral Expression Systems (Takara) in accord with the manufacturer's user's manual.
In vivo model
All procedures involving mice and experimental protocols were approved by the Animal Experimental Committee of the Graduate School of Medicine, Nagoya University (permission no. M210418). We confirm that the study is reported in accordance with ARRIVE guidelines. The animal study was carried out in accordance with the Guidelines for Animal Experiments of the Nagoya University School of Medicine. Six five-week-old female BALB/C nude mice were purchased from Charles River (Yokohama, Japan). In the control group, a total of 1×107 sh-control CaSki cells was suspended in 200 µL of PBS and injected intraperitoneally (i.p.) into each mouse. The weight of each mouse was: 21.72g, 20.45g, and 22.37g. In the sh-DDIT4 groups, 1×107 sh-DDIT4 CaSki cells were also suspended in 200 µL of PBS for injection. The weight of each mouse was: 22.19g, 21.53g, and 20.86g. Tumor progression was monitored every week. The metastatic tumor-bearing burden was investigated 49 days after the injection with the use of the Xenogen IVIS 200 Imaging System (Caliper Life Sciences, Hopkinton, MA).
Statistical analyses
For normally distributed data, unpaired Student's t-test and one-way analysis of variance (ANOVA) was used to determine the statistical significance of differences. P-values < 0.05 were considered significant. All statistical tests were two-sided. Analyses of clinicopathological characteristics in Table 1 was performed by IBM-SPSS Statistics 25.0 software using chi-square test. P-values < 0.05 were considered significant. Non-normally distributed data were been made log2 normalization. The bioinformatic analysis was conducted in R ver. 4.0.3, and the other statistical analyses and creations of graphs were performed with GraphPad Prism 7 software (GraphPad Software, San Diego, CA).