Data sources

The RNA-seq data for ovarian cancer patients was obtained from The Cancer Genome Atlas (TCGA), and downloaded from xena hubs https://tcga.xenahubs.net. Epithelial-mesenchymal transition (EMT) status was defined by Da Yang et al. [25]. Among 341 ovarian cancer patients in TCGA, 324 has EMT status, and patient characteristics was shown in Table 1.

Table 1 Clinical characteristics of study patients Identification of survival-related genes

The difference of SLFN5 expression between EMT status was detected by t test and the analysis of covariance, and the latter was used to adjust for several confounding factors, such as age, stage and grade in this study. The expression of SLFN5 was divided into high-expression and low-expression groups based on the first quartile (25th quantile). The overall survival (OS) curves of two groups were estimated using Kaplan–Meier method and compared by log-rank test [26]. Multivariate Cox proportional hazards model was used to explore the independent effect of SLFN5 expression on OS, which adjusted for confounding effects of age, stage and grade.

The co-expression genes of SLFN5 were identified using Spearman correlation, and false discovery rate (FDR) method was used to adjust for multiple comparisons [27]. Genes with absolute correlation coefficients≧0.4 and FDR q value < 0.001 were identified as co-expression genes of SLFN5. The biological functions of these co-expression genes were explored using Gene Ontology (GO) enrichment analysis [28] and significant GO items were selected based on FDR q value < 0.05.

Immune cells were inferred from the TCGA RNA-seq data using CIBERSORT algorithm [29], and 22 distinct immune cells were obtained. Patients with P value less than 0.05 were excluded from the study. Immune cells with the proportion of zero value larger than 0.5 were also excluded, and 13 immune cells were retained in the following analyses. The distribution differences of immune cells between high-expression and low-expression SLFN5 groups were detected using Wilcoxon rank sum test. All statistical analyses in this study were performed on R platform (Version: 3.6.0).

Tissue samples and ethical considerations. Tissue samples

A total of 27 human ovarian cancer tissues were obtained from patients who had surgery in the Affiliated Hangzhou First People’s Hospital of ZheJiang University School of Medicine from 2019 to 2022, and patients’ clinical data were obtained afterwards. All cases were included post review by pathologist and only where complete clinical and follow-up data was available. None of the 27 included patients underwent pre-operative local or systemic treatment. The study protocol was approved by the Institutional Review Board of the Hangzhou First People’s Hospital in China. Freshly harvested samples were immersed in RNAlater (Life Technologies, Shanghai, China) before snap freezing within 30 min post-surgery. All tissue samples were stored in liquid nitrogen until further use.

No patient had received chemotherapy before surgery. Four histological subtypes were included into the panel (serous (n = 14), endometrioid (n = 7), clear cell (n = 4), and mucinous (n = 2)). TNM classification (T = tumor, N = lymph nodes, M = metastasis) was performed according to the Union for International Cancer Control (UICC). Lymph node involvement (N0 (n = 11), N1 (n = 16) and distant metastasis M0 (n = 9), M1 (n = 18) was evaluated. FIGO stage was determined (I, II (n = 9), III,IV (n = 18)) according to the criteria of the International Federation of Gynecology and Obstetrics (FIGO). Median patients’ age was 62 ± 12 years with a range between 31 and 88 years. During the study 0 deaths have been observed (The data was shown in Table 2).

Table 2 Correlation between SLFN5 expression and pathological parameters in ovarian carcinoma patientsEthical considerations

A total of 27 human ovarian cancer tissues were obtained from patients who had surgery in the Affiliated Hangzhou First People’s Hospital of ZheJiang University School of Medicine from 2019 to 2021, and patients’ clinical data were obtained afterwards. The classification of clinical staging and histological grading of ovarian cancer were determined according to the FIGO 2014 system. Approval from the research ethics committee was obtained prior to the study. In addition, written informed consent from the patients were obtained before experiment for the use of their samples.

Cell culture and treatment

Four human ovarian cancer cell lines (SKOV3, A2780, OVCAR3, HO8910), normal epithelial ovarian cells (IOSE80) were purchased from the cell bank of China Academic of Science. The SKOV3 cells were cultured in McCoy’s 5A Media (modified with tricine) supplemented with 10% fetal bovine serum (FBS). The OVCAR3 and IOSE80 cell lines were maintained in 90% RPMI 1640 with 10% FBS, and the OVCAR3 cells were cultured in 80% RPMI 1640 with 20% FBS, sodium pyruvate, and 0.01 mg/ml bovine insulin. All the cell lines were cultured in an atmosphere of 5% CO2 and 95% air at 37 °C.

Pathway enrichment analysis and literature search

Text mining was performed for the overlapping genes using Perl code. The published genes that were closely related with ovarian cancer were searched in the PubMed database. In addition, the overlapping genes were subjected to the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis using the Database for Annotation, Visualization and Integrated Discovery (DAVID) online tool. Pathways with p < 0.05 and counts ≥ 2 were considered significant.

RNA extraction and quantitative real time polymerase chain reaction(qRT-PCR)

Total RNA was extracted from cells using the RNA iso Plus (Trizol) reagent (TaKaRa, Japan) and cDNA was synthesized using the PrimeScript™RT Master Mix(Perfect Real Time) kit (TAKARA, Japan) according to the manufacturer’s instructions. Real-time PCR was performed to evaluate the expression levels of SLFN5,SNAIL,SLUG,E-cadherin,N-cadherin,GAPDH in tumor cells. A total of 8 μl of cDNA was used as template in a final 20 μl PCR volume containing 1 μl forward primer,1 μl reverse primer, and 10 μl SYBR Premix EX Taq (2x). PCRs were run as follows: 50.0 °C for 3 min, 95.0 °C for 3 min, followed by 40 cycles of 95.0 °C for 10 s and 60.0 °C for 30 s. Following PCR, a melting curve was obtained at temperatures from 60 °C to 95 °C, at increments of 0.5 °C for 10 s. Primer sequences are listed in Table 3.

Table 3 The primer sequences in PCR analysis

Transfection of siRNA SLFN5 siRNAs (Genechem, Shanghai, China) were used to downregulate SLFN5 expression. The two siRNA sequences were shown below:

SLFN5 siRNA-1: (forward)5'-GUGGUAUAUACUCCAGAGATT-3' and.

(reverse)5'-UUUCUGGAGUAUAUACCACTT-3';

SLFN5 siRNA-2: (forward)5'-GACUCAGACUCCAACGAAUTT-3'and.

(reverse)5'-AUUCGUUGGAGUCUGAGUCTT-3'.

SKOV3 cells、OVCAR3 cells、HO8910 cells、A2780 cells were transfected with SLFN5 siRNAs with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Four to six hours post transfection, the cell culture medium was changed. After 48 h of transfection, SKOV3 cells were used for further examination.

Cell Invasion Assay(transwell)

For invasion assays, before seeding cells, 60 μl of Matrigel (BD Biosciences, San Diego, CA, USA) was placed on the upper surface of the 24-well transwell chamber (Corning, New York, USA). Cells (104) in 100 μl of RPMI 1640 medium were seeded in the upper chamber, and the lower chamber was filled with 600 μl of medium with 20% FBS. Twenty-four hours after incubation, cells were remaining on the upper surface which were removed using a cotton swab, while the invaded cells were fixed, stained and photographed. Five random fields of cells were selected and counted for further calculation.

Wound healing assay

OC cells were transfected with SLFN5-siRNA or treated with TGF-β1 and wound healing assay was performed as described previously [30]. Briefly, highly confluent OC cells were serum starved and the wound was made through the cell monolayer using a 200 µL pipette tip. Following this, the cells were washed twice with 1X PBS to remove the non-adherent cells. The cells were then treated with TGF-β1 or transfected with SLFN5-siRNA (as mentioned) in a fresh serum-free medium. The cells were allowed to migrate for 48 h. Wound closure was monitored by visual examination and imaged every 24 h under microscope (EVOS, Invitrogen). The experiment was repeated twice.

Cell lysis and Western blot

Cells were lysed with RIPA buffer (Beyotime, Shanghai, China) to obtain total protein. Then, 30–50 μg of protein was separated in 10% SDS/PAGE gels and transferred to PVDF membranes, which were blocked with 5% fat-free milk. The membranes were then incubated overnight at 4℃ with a primary antibody and incubated at room temperature for one hour with a secondary antibody conjugated with horseradish peroxidase. In the end, the protein bands were examined with chemiluminescence assay.

Cell colony formation assay

Cells (600 cells/well)were seeded into 6-well plates with DMEM medium supplemented with 10% FBS and cultured for 14 days. Then, colonies were fixed with methanol at room temperature for 15 min and stained with 0.1% crystal violet for 15 min (Invitrogen, Carlsbad, CA), and the total number of visible colonies were counted.

Statistical analyses

SPSS 16.0 (IBM, USA) was used for the statistical analyses. Continuous data was expressed as the mean ± SD, and analysed by independent t-test between two groups. Among multiple groups, one-way ANOVA was applied, and Turkey test was applied as a post hoc test. The categorical data were compared via the Chi-squared or Fisher’s exact tests as appropriate. A p value < 0.05 was regarded as statistically significant.