Collection of data and tissue samples
We downloaded the RMS array sequencing data set from the GEO database (https://www.ncbi.nlm.nih.gov/geo/) and selected the gene expression profile of GSE92689. After deleting samples with incomplete information, 125 tumour samples from patients with RMS were obtained for subsequent analysis. In total, 33 paraffin-embedded RMS samples and 11 control striated muscle tissue specimens were selected from Shihezi University Hospital and the First Affiliated Hospital of Xinjiang Medical University, China. All subjects provided written informed consent. This research was performed in compliance with the ethical guidelines of the Helsinki Declaration, and was approved by the hospital ethics committee.
ESTIMATE analysis
The “ESTIMATE” and “limma” software packages in R (v4.0.3) were employed to measure the stromal and immune scores of RMS patients. The scores were presented in 3 types: stromal, immune, and ESTIMATE scores. The scores were utilized to measure the ratio of stromal or immune components in the TME. The lesser the scores, the lower the ratios of the two components. Finally, each sample was divided into low (<median) or high (>median) scoring group.
Stromal, immune and ESTIMATE scores with clinical stages
R language was employed to analyse the relationship between clinical staging data and stromal, immune and ESTIMATE scores. Kruskal-Wallis rank sum or Wilcoxon rank sum test was performed to compare the differences between two groups. A p-value of <0.05 was deemed statistically significant.
Generation of DEGs between the high and low scoring groups
The “limma Bioconductor” software package in R (v4.0.3) was employed to generate DEGs by comparing high- and low-scoring samples. The “pheatmap” software package in R (v4.0.3) was utilized to construct a DEG heatmap. Venn diagrams were created to compare upregulated and downregulated crossover genes associated with immune/matrix scores. The threshold conditions were as follows: |log2 fold change (log2FC) | >1.0, and false discovery rate (FDR) <0.05.
Functional enrichment analysis
Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontology (GO) analyses were conducted on 593 DEGs via the “clusterProfiler,” “enrichplot,” and “ggplot2” software packages in R (v4.0.3). Statistical significance was considered at p-value <0.05.
Protein-protein interaction (PPI) network analysis
A PPI network was constructed by the STRING database (https://string-db.org/). Then, Cytoscape (v3.7.2) was used for reconstruction. The CytoHubba Cytoscape plug-in was used to identify the core genes on the basis of PPI network, and the confidence of the interactive relationship between nodes in the network was larger than 0.95.
Relationship between the top 10 hub genes in PPI network and overall survival (OS)
The top 10 hub genes chosen from the PPI network were validated by Gene Expression Profiling Interactive Analysis (GEPIA, http://gepia.cancer-pku.cn/), which showed good performance for estimating 5-year OS. The expression levels of the hub genes were compared between sarcoma patients and control samples, and the OS Kaplan–Meier survival curve was established. Survival analysis was carried out using the survival package in R, and statistical significance was considered at p <0.05.
Gene expression level with clinical staging
Data on the clinicopathological features of RMS patients were retrieved from GSE92689. R language was employed to analyse the relationship between clinical staging data and gene expression. Kruskal-Wallis rank sum or Wilcoxon rank sum test was performed to compare the differences between two groups. P-values of <0.05 were deemed statistical significance.
Gene set enrichment analysis (GSEA)
A collection of C2 KEGG gene sets (v7.2) was retrieved from the Molecular Signatures database as the target set used by GSEA, and GSEA (v4.0.3) was conducted to elucidate the molecular mechanism of low-expression and high-expression populations. Upregulation and downregulation ways were obtained. Statistical significance was considered at FDR < 0.05.
Immune cell profile
The immune cell composition in tumour tissue was estimated using CIBERSORT (https://cibersort.stanford.edu/about.php) to assess the difference in the infiltration of 22 immune cells between high- and low-expression groups. Statistical significance was considered at p-value <0.05.
Culture and transfection of RMS cells
RMS cell lines (PLA802, RH30 and RD) were supplied by Biotechnology Co., Ltd. (Fu Xiang, Shanghai, China). The normal skeletal muscle cell line (HSKMC) was stored in our laboratory. All cells were cultured in DMEM (Gibco, USA) supplemented with 10% FBS (Gibco, USA) and 1% penicillin-streptomycin (Solarbio, China), and maintained at 37 ℃ with 5% CO2. Lipofectamine 2000 (Life Technologies, USA) was transiently transfected with RMS cells for 24 h in compliance with the manufacturer’s instructions. The siRNA sequences included si-h-MAD2L1, 5ʹ-GGGUCCAAAGUUGAGUGAGUCUUGA-3ʹ and si-h-CCNB2, 5ʹ-CAAGAATGTGGTGAAAGTA-3ʹ.
Immunohistochemistry (IHC) and IHC assessment
The IHC antibody is composed of anti-CCNB2 (Abcam, ab185622) and anti-MAD2L1 (Abcam, ab97777). Paraffin-embedded RMS tissue sections were taken, and IHC staining was performed. The sections were deparaffinized, rehydrated with xylene, and washed with graded alcohol and PBS. After heating for 15 min in citric acid buffer (pH 6.0), antigen retrieval was conducted. TBS/H2O2 was used to block endogenous peroxidase. After incubation with anti-human CCNB2 and MAD2L1 primary antibodies, the sections were exposed to goat anti-rabbit antibodies at 37 ℃ for 30 min.
IHC staining was assessed by two independent pathologists with no knowledge of patient characteristics. The staining results of MAD2L1 or CCNB2 were evaluated by staining intensity and degree. The scoring system was as follows: 0 (no staining), 1 (shallow yellow), 2 (brownish yellow), and 3 (dark brown). The proportions of positive staining cells were scored as follows: 0 (0%), 1 (<25%), 2 (25-75%), and 3 (>75%). The final scores of <4 and ≥4 was deemed as low and high expression, respectively[18, 19].
Western blot analysis
Approximately 48 h after transfection, RMS cell lysis was performed, and the total protein was isolated in RIPA buffer (Solarbio). After electrophoresis, the protein molecules on the gel were electrically transferred to PVDF membranes (Solarbio), immersed in blocking solution (5% non-fat milk/0.1% Tween-20) for 2 h, and then exposed to anti-CCNB2 (Abcam, ab185622) and anti-MAD2L1 (Abcam, ab97777) at 4 ℃ overnight. On the next day, the membrane was exposed the corresponding secondary antibodies at room temperature (RT) for 2 h.
Immunofluorescence (IF) procedure
The slides were fixed with 4% paraformaldehyde in the culture plate for 15 min, immersed three times in PBS, and permeated with 0.5% Triton X-100 (prepared in PBS) at RT for 20 min. Serum blocking was performed for 30 min at RT, and the blocking solution was absorbed with an absorbent paper. A sufficient volume of diluted primary antibody was added into each slide. After transferring into a humid box, the slide was incubated overnight at 4 ℃. On the next day, the slide was exposed to fluorescent secondary antibodies for 1 h. DAPI was added dropwise to the coverslips, and then incubated for 5 min in the dark. The specimens were stained with nuclei, and then examined using a fluorescence microscope.
CCK8 assays
Cell Counting Kit-8 (CCK8; Dojindo, Japan) analysis was conducted to measure cell toxicity and proliferation. Tumour cells (1 × 104 cells/well) were grown in a 96-well plate. After transfection or addition of inhibitors at approximately 0, 24, 48 and 72 h, the absorbance of the solution was recorded at 450 nm.
5-Ethynyl-2’-deoxyuridine (EdU) staining
The transfected tumour cells (1 × 105 cells/well) were grown in a 12-well plate. The cells were labeled with the EdU kit (KGA337, KeyGen BioTECH, China) and photographed under a fluorescence microscope.
Acridine orange staining
Acridine Orange (AO) and Evans Blue (EB) at 1 mg each were dissolved in 10 mL of pH 7.2 PBS to prepare a 100 µg/mL stock solution. The same amount was mixed before use and then set aside. A 100 µL aliquot of the pre-observed cell suspension that has been cultured and incubated with samples was added with 4 µL of AO/EB dye and mixed well. A glass slide was placed with a drop of the above mixture and then covered. The staining results were visualized using a fluorescence microscope.
TUNEL staining
TUNEL apoptosis detection kit was purchased from Shanghai Biyuntian Biotechnology. The transfected cells were fixed in 4% paraformaldehyde (Solarbio) for approximately 30 min, and then washed with PBS three times. Each well was added with TUNEL detection solution (50 µL) and then incubated at 37 ℃ for 1 h. Images were subsequently collected.
Flow cytometry of apoptosis
The transfected cells were collected in a six-well plate and washed three times with PBS. Then, Annexin V-FITC (5 µL) and PI staining solution (10 µL) were added to each test tube, and the cells were analysed using a flow cytometer (Partec, Germany).
Cell cycle analysis
At 24 h after siRNA interference, the cells were collected, washed with PBS, and then fixed with 75% alcohol in a refrigerator at 4 ℃ overnight. The cells lines were incubated with RNaseA for 30 min before detection. After staining with PI, the cells were analysed by flow cytometry.
Transwell assay
Matrigel gel was diluted with serum-free cell culture medium at a ratio of 1:8 at 4 °C. Then, 100 µL of this mixture was applied evenly on the surface of the polycarbonate membrane of the upper chamber and placed at 37 °C for 0.5–1 h. Cells in the log phase were harvested and washed with PBS. The density of cells was adjusted to 1 × 105 cells/well. Then, the upper and lower chambers were added with 100 µL of the cell suspension and 600 µL of 10% FBS, respectively. After being placed in the incubator for 24–48 h, the cells were subjected to fixing, staining and counting.
Statistics
SPSS v20.0 was employed for statistical analysis. Statistical significance was considered at p <0.05. All graphics were drawn using the GraphPad Prism v8.0.