The transcriptome data of 306 CESC and 3 normal samples were downloaded from The Cancer Genome Atlas (TCGA) database (https://tcga-data.nci.nih.gov/tcga/). Among the CESC patients, 24 cases had p53 mutation and 282 cases had no p53 mutation. Furthermore, the GSE27678 dataset, which includes 14 healthy and 30 squamous cell carcinomas of cervix (including two premalignant lesions and squamous cell carcinomas cell lines), was obtained from Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE27678). All data were publicly available and were downloaded for research purpose.

The expression of MIR205HG was analyzed by GEPIA2 (http://gepia2.cancer-pku.cn/#analysis) in 306 CESC and 13 normal samples (3 normal samples from TCGA and 10 normal samples from Genotype-Tissue Expression database). The differentially expressed lncRNAs were analyzed using R software version 3.6.3 (https://www.r-project.org/), and a |log2foldchange| >1 was used to determine significance. GSE27678 was analyzed by GEO2R. Pathway enrichment analysis was performed using DAVID (https://david.ncifcrf.gov/). The top 20 enriched pathways were selected (P<0.05). The bubble plots were designed using ggplot2 package of R (http://had.co.nz/ggplot2/). Survival and correlation analysis were performed using GEPIA2 (http://gepia2.cancer-pku.cn/#index). The network analysis was performed using Cytoscape V3.6.1 (https://cytoscape.org/).

Ca Ski and C-33 A cell lines were purchased from The Cell Bank of Type Culture Collection of the Chinese Academy of Sciences. Human cervical epithelial cell line (HCerEpiC) was obtained from Shanghai Zhongqiaoxinzhou Biotechnology Co., Ltd. (cat. no. 7060). Ca Ski was cultured in RPMI 1640 (cat. no. 10-040-CV; Corning, Inc.), C-33 A was cultured in MEM (cat. no. E600020; Sangon Biotech Co., Ltd.) and HCerEpiC was cultured in DMEM (cat. no. 10-013-CV; Corning, Inc.). All media were supplemented with 10% FBS (cat. no. 10099-141-FBS; Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin (cat. no. E607011; Sangon Biotech Co., Ltd.). All cells were placed at 37°C in a humidified incubator containing 5% CO₂.

Total RNA was extracted from cells using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.), and reverse transcription was performed using cDNA Synthesis kit (cat. no. K1622; Thermo Fisher Scientific, Inc.) according to the manufacturers' instructions. Quantitative PCR was carried out on an ABI Q6 system (Applied Biosystems; Thermo Fisher Scientific, Inc.). RT-qPCR reactions were performed as follows: 95°C for 10 min, 45 cycles of 95°C for 15 sec, 60°C for 60 sec and a final dissociation stage. The relative expression levels were normalized to endogenous control GAPDH and were expressed as 2^−ΔΔCq (15). The sequences of the primers used were as follows: GAPDH, forward 5′-AGAAGGCTGGGGCTCATT-3′, reverse 5′-TGCTAAGCAGTTGGTGGTG-3; MIR205HG, forward 5′-GTTTCACCATGTTGCCCAGACT-3′, reverse 5′-CCTGTGCGGAACAGAAATGACT-3′; fibroblast growth factor receptor 3 (FGFR3), forward 5′-GTGCTCAAGACGGCGGGC-3′, reverse 5′-GCCACGCAGAGTGATGAGAAAA-3′; thymidine phosphorylase (TYMP), forward 5′-GAGTCTATTCCTGGATTCAATGTCA-3′, reverse 5′-AGAATGGAGGCTGTGATGAGTG-3′; and GTPase HRas (HRAS), forward 5′-CTGAGGAGCGATGACGGAAT-3′ and reverse 5′-GGAATCCTCTATAGTGGGGTCGT-3′.

MIR205HG-homo-474 was knocked down using small interfering (si) RNA, which was transfected into Ca Ski and C-33 A cell lines using Lipofectamine™ 2000 (Invitrogen; Thermo Fisher Scientific, Inc.). Ca Ski and C-33 A cells were seeded into a 6-well plate with 30×10⁴ cells/well 1 day before transfection. Cells were transfected with MIR205HG siRNA or control siRNA with a final concentration of 50 nM, and culture for 24 h. The transfection efficiency was confirmed by RT-qPCR. The siRNA was purchased from Suzhou GenePharma Co., Ltd. The MIR205HG siRNA sequence was 5′-GCUGAACUGGGUGCUUUAUTT-3′; 5′-GCUGAACUGGGUGCUUUAUTTAUAAAGCACCCAGUUCAGCTT-3′, and that of the siRNA control was 5′-UUCUCCGAACGUGUCACGUTT-3′ and 5′-ACGUGACACGUUCGGAGAAT-3′.

Cell proliferation was determined using CCK-8 assay (Beyotime Institute of Biotechnology). Cells were cultured at the density of 1,000 cells/well and culture for 24 h before transfection in a 6-well plate. Each sample was assessed in six duplicates. Subsequently, 10 µl CCK-8 was added to each well for 1 h, and absorbance was detected at 450 nm on a microplate reader (Infinite M1000; Tecan Group, Ltd.).

The migratory and invasive ability of cells was analyzed using Transwell assay. The 8.0-µm pore size membranes (cat. no. 353097; Falcon^®; BD Biosciences) were used for migration assay whereas the BioCoat™ Matrigel^® 0.8-µm pore size membranes (cat. no. 354480; Corning, Inc.) were used for invasion assay. The membranes were placed in a 24-well plate, and a total of 75,000 cells were seeded in the upper chamber containing serum-free medium. A volume of 700 µl medium containing 10% FBS was loaded into the lower chamber at the bottom of 24-well plate. The filters were stained with crystal violet (Sangon Biotech Co., Ltd.) after 24 h, at 20°C for 30 min. Cells were observed and counted under a light microscope at ×200 magnification (Nikon Corporation; SMZ1000). Three random fields were counted for each microscopic field.

Immunofluorescence staining was performed using the conditions suggested by the primary antibody suppliers. Briefly, coverslips were placed into the 24-well plate, and the digested cells were inoculated to the 24-well plate with a cell density of ~50,000 cells/well and 500 µl medium, which were then cultured at 37°C for 24 h. After the cell fusion rate was 70%, cells were transfected with MIR205HG or control siRNA (50 nM) and cultured for 24 h. Cells were washed with PBS, fixed with 4% paraformaldehyde for 15 min at 20–25°C, and permeabilized at 20°C using 0.1% Triton X-100 and 5% BSA (cat. no. A8020; Beijing Solarbio Science & Technology Co., Ltd.) in PBS for 5–15 min. After permeation, the cells were washed with PBS three times/5 min. Cells were incubated with 200 µl primary antibodies against Ki-67 (Cell Signaling Technology, Inc.; cat. no. 9449; 1:100; mouse mAb) and p16 (Beyotime Institute of Biotechnology; cat. no. AF1672; 1:300; rabbit mAb) at 4°C overnight. The cells were then incubated with 200 µl Alexa Fluor^® 488 labeled goat anti-mouse IgG secondary antibody (1:500; Abcam; cat. no. ab150113) and Cy3-labeled goat anti-rabbit IgG secondary antibodies (1:500; Institute of Biotechnology; cat. no. A0516) for 30 min at 20°C. The nuclei were counterstained with DAPI for 5 min. Images were obtained using fluorescence microscopy at ×400 magnification (Leitz Orthoplan; Leica Microsystems GmbH).

Comparison between two groups was performed using two-tailed Student's t-test and comparison between three groups was performed by one-way ANOVA followed by Tukey's post hoc test. Statistical analyses were made using SPSS package 17.0 (SPSS, Inc.). P<0.05 was considered to indicate a statistically significant difference.

Expression data of 306 CESC samples and 13 normal samples were retrieved from the TCGA. The cut-off |log₂fold-change| >1 was used. A total of 28 upregulated and 175 downregulated lncRNAs in clinical cancer types were analyzed. Using the same standard, 1,542 upregulated and 2,726 downregulated mRNAs were identified. The expression of the top 20 differentially expressed lncRNAs in each sample are presented in the heat map of Fig. 1A. The red arrow corresponds to MIR205HG, which was the most commonly upregulated lncRNA. Each column represents a sample in the CESC data retrieved from the TCGA.

To predict the function of these lncRNAs, the top 400 coexpressed genes were selected by Spearman's correlation analysis (>0.2). Subsequently, overlaps of coexpressed and differentially expressed genes were selected. A pathway enrichment analysis was performed using DAVID. The results including MIR205HG, LINC00925 and EMX2OS are presented in Fig. 1B-D, respectively. MIR205HG-related genes were enriched in cancer-related pathways, such as ‘cell cycle’, ‘p53 signaling pathway’, ‘Ras signaling pathway’ and ‘bladder cancer’ (Fig. 1B). The LINC00958 coexpression genes were significantly enriched in pathways related to virus infection (Fig. 1C).

To investigate the clinical outcome of these lncRNAs, survival analysis was performed using GEPIA (Fig. 2). Nine lncRNAs, including six upregulated (Fig. 2A-F) and three downregulated (Fig. 2G-I) lncRNAs, were analyzed. Most of these lncRNAs have no significant association with overall survival (log rank P<0.05). EMX2OS, which is one of the most downregulated lncRNAs, had a significant association with the overall survival. In addition, the EMX2OS high expression group had an improved overall survival compared with the EMX2OS low expression group (Fig. 2H). The coexpressed genes of EMX2OS were enriched in the ‘cGMP-PKG signaling pathway’ (Fig. 1D).

To validate the previous results, the lncRNA MIR205HG was further studied. We analyzed the expression of MIR205HG in GSE27678 dataset, which contained 30 tumor samples and 3 normal samples. The results demonstrated that MIR205HG had higher expression in tumors samples compared with normal samples (Fig. 3B), which was in accordance with TCGA data (Fig. 3A and B). Subsequently, the expression of MIR205HG was detected in the two CESC cell lines Ca Ski and C-33 A. The results from RT-qPCR demonstrated that MIR205HG was overexpressed in Ca Ski and C-33 A cell lines compared with the normal cervix cell line HCerEpiC (Fig. 3C). Then, MIR205HG was knockdown in Ca Ski and C-33 A cells (Fig. 4A). C-33 A and Ca Ski cell proliferation was significantly decreased following MIR205HG knockdown (Fig. 4B). Furthermore, the migratory and invasive abilities were significantly inhibited following MIR205HG knockdown (Fig. 4C and D). The migratory and invasive abilities of Ca Ski cells appeared to be higher than that of C-33 A cells. Immunofluorescence staining of p16 and Ki-67 was then performed (Fig. 4E). Ki-67 staining, which is the marker of proliferation, was decreased in C-33 A following MIR205HG knockdown and was partially decreased in Ca Ski. In addition, p16 fluorescence intensity was lower in C-33 A cells after MIR205HG knockdown, whereas no change was observed in Ca Ski cells.

A network between MIR205HG and its coexpressed genes were analyzed by cytoscape3.6.1 (Fig. 5A). Genes in enriched pathways were selected. A total of 49 related genes are presented. Subsequently, the expression of three selected genes, FGFR3, TYMP and HRAS, was detected in MIR205HG knocked down CESC cells. In C-33 A cells, all these genes were significantly downregulated after MIR205HG knockdown (Fig. 5B-D), whereas HRAS showed no significant change (Fig. 5D). Furthermore, the expression of these three genes was positively correlated with MIR205HG expression in CESC data of TCGA (Fig. 5E-G).