Data Acquisition and Processing:
The TCGA and ICGC databases provided the clinical samples and mRNA expression data for liver hepatocellular carcinoma (LIHC). In this work, 50 normal samples and 374 LIHC samples were chosen for gene expression profile analysis from The Cancer Genome Atlas (TCGA) database, while 202 normal samples and 243 LIHC samples were chosen from the International Cancer Genome Consortium (ICGC) database. Additionally, clinical data from 377 LIHC patients were collected for clinical-pathological feature analysis.
Weighted Gene Co-expression Network Analysis (WGCNA):
WGCNA is an R package that identifies a group of genes exhibiting synchronous changes [1]. Initially, the optimal soft-thresholding power (β) was set to 18 to follow a scale-free network distribution. Gene modules were determined using hierarchical clustering. Subsequently, key modules were selected based on the correlation analysis of clinical samples. The genetic connectivity, measuring the degree of correlation between one gene and another, was calculated. Core genes were those with high connectivity within the main module.
TIMER Database Analysis:
TIMER (https://cistrome.shinyapps.io/timer/), an integrated database, is used to analyze tumor immune infiltration [2]. The associations between immune cell infiltration, SCAF1 mRNA expression, and copy number changes in LIHC samples were predicted using the "Gene" and "SCNA" modules. Gene co-expression was displayed via the "Correlation" module. We also looked at the connection between different immune cell infiltrations in pan-cancer tissues and SCAF1.
Gene Set Enrichment Analysis (GSEA):
GSEA is a gene set enrichment analysis method based on a set of genes [3]. RNA sequencing data related to LIHC were obtained from TCGA. Using the following fundamental threshold values, we ranked the samples and obtained normalized enrichment scores (NES) by splitting them into two groups based on the expression of SCAF1 using the GSEA program (version 4.0.3): 0.05 FDR q-value and 0.05 P value.
TCMSP:
TCMSP (https://old.tcmsp-e.com/tcmsp.php, Traditional Chinese Medicine Systems Pharmacology) is built upon the framework of traditional Chinese medicine pharmacology [4].It contains 499 Chinese medications with 29,384 chemicals, 3,311 targets, and 837 associated disorders that are listed in the Chinese Pharmacopoeia. Based on this, we searched for herbal ingredients capable of treating LIHC by inhibiting angiogenesis: genistein, rutaecarpine, bifendate, hirsutasideB, and genistein.
PubChem:
PubChem (https://pubchem.ncbi.nlm.nih.gov/) gathers chemical information from hundreds of data integration sources (as of September 6, 2022, 872 sources). Most of the chemicals contained in PubChem are small molecules, and their chemical structures are provided. We obtained the chemical structures of these 5 herbal ingredients from PubChem for further analysis.
Protein Structure and Docking Analysis:
The RCSB Protein Data Bank (https://mcule.com/dashboard/) collected three-dimensional structural information about proteins. We obtained the molecular structure of SCAF1 protein from this database. Using the 1-Click Docking web server, we investigated the binding modes of SCAF1 with the molecules of 5 traditional Chinese medicine ingredients and used PYMOL to visualize, analyze, and map them.
TISIDB:
TISIDB (http://cis.hku.hk/TISIDB/) is a tumor immune analysis database that integrates data from five major types of tumor immune-related data [5]. Based on this, we investigated the association of SCAF1 with six immune system components (lymphocytes, immune modulators, convergent factors, etc.) in LIHC.
Statistical Analysis:
The R program (version 4.3.0) was used for statistical analysis. Differential expression of SCAF1 was detected by a run-sum test using the "limma" and "beeswarm" packages. The "survival" and "survminer" packages were used to conduct multivariate Cox regression analysis in order to identify significant predictors (p < 0.05). ROC curves were created using the SurvivalROC package (version 3.3.1) to assess the ability of SCAF1 expression levels to predict 1, 3, and 5-year survival rates. Survival scenarios for SCAF1 were created in prevalent cancers using the "Survival" and "SorvMiner" packages. The tumor mutation burden (TMB) of each tumor was calculated using the TMB function in the MAfTools package. The "pheatmap" package was used to investigated the relationship between SCAF1 expression and immune checkpoint genes. Unique states of tumor mutation burden were revealed by combining sample TMB and gene expression data.
Cell Culture:
Huh7, HCCLM3, and HepG2 LIHC cell lines were acquired from the National Collection of Authenticated Cell Cultures. Except for HCCLM3 cells, which were grown in RPMI-1640 media, all cell lines were grown in DMEM enhanced with 10% FBS and 1% penicillin-streptomycin. Every cell was grown in an incubator with 5% CO2 and a humidifier set at 37°C.
Western Blot Analysis and Antibodies:
Prepare the total protein extracts according to the method described previously. Using RIPA buffer (Beyotime, Shanghai, China) to extract protein, a buffer mixture containing protease and inhibitors (Termo Fisher Scientific, New York, USA). The monoclonal antibodies used in the experiment were: anti-SCAF1 (1:1000, Solarbio, K112765P); anti-VEGFA (1:1000, Abcam, ab46154); and anti-Tubulin (1:1000, Proteintech, 11224-1-AP).
quantitative real-time PCR (qRT-PCR):
To assess the relative RNA levels of genes, quantitative real-time PCR was used. Total RNA was extracted using the standard Trizol-based protocol (Invitrogen, USA). Reverse transcription of RNA was performed using the PrimeScript RT Reagent Kit (Invitrogen, USA), and qPCR was conducted using SYBR Premix Ex Taq (TaKaRa, China), in accordance with the manufacturer's instructions. Supplementary Table 1 contained details on the gene-specific primers.
Transwell migration assay:
Stable transfected cells were utilized for tests of invasion and migration after 48 hours of seeding. Serum-free media was used to plate 6 x 104 cells in the upper chamber for migration. 1 x 105 cells were plated in a Matrigel-coated chamber (BD Biosciences) in preparation for invasion. To get rid of unmigrated cells, the top surface of the membrane was gently wiped 24 hours (for migration) or 48 hours (for invasion) after seeding. Fixing and staining the migrating cells with 0.1% crystal violet was done. Five randomly chosen microscopic fields were used to count the cells using a light microscope equipped with a DP70 CCD system (Olympus Corp, Japan).
Tube Formation Assay:
Incubate HCCLM3 cells on six-well slides for 24 hours. The medium from the cells was collected and used as the medium for tube formation assay. Pipette tips and 96-well plates were pre-cooled. Apply 50 µl of Matrigel to the wells of the 96-well plate and incubate for 30 minutes. Inoculate the suspension of treated Human Umbilical Vein Endothelial Cells (HUVECs) into Matrigel-lined 96-well plates (5 x 104 cells/well). Following an 8-hour incubation period at 37°C, use the microscope to evaluate angiogenesis. Branching density and tube length were measured by randomly choosing three locations per well.
CCK8 assay:
Transfected cells were plated into each well of 96-well plates after being cultured for 48 hours in order to assess CCK8. 10 µl of CCK8 reagent was added to each well and incubated for 2 hours at 37°C after 24, 48, 72, 96, and 120 hours. With the use of a microplate reader (ELX-800; Bio-Tek, Winooski, VT, USA), the absorbance value was determined at 450 nm.