Sample datasets and clinical profiles

Standardized fragments per kilobase of transcript per million fragments mapped (FPKM) gene expression data profiles, genetic alteration landscapes, and clinical data were obtained from The Cancer Genome Atlas (TCGA) by using the R package TCGAbiolinks (v.2.24.3) [21]. A total of 613 samples from ccRCC patients, comprising 541 tumor samples and 72 normal samples, were collected. Only 72 of the ccRCC patients had paired samples of tumor and normal tissue. In addition, the expression profiles of GSE53757 (72 tumor tissues and 72 normal tissues) and GSE36895 (29 tumor tissues and 23 normal tissues) were obtained from the Gene Expression Omnibus (GEO) repository (https://www.ncbi.nlm.nih.gov/).

Analysis of DBF4 expression patterns

Using the TCGA cohort and two datasets from GEO, we compared DBF4 expression levels between cancer and normal samples. Additionally, to further investigate the relationships between DBF4 expression levels and clinical characteristics, we evaluated differences in DBF4 expression across clinical stages genders, and age groups. Wilcoxon signed-rank tests were used for comparisons between two groups, whereas paired Wilcoxon signed-rank tests were applied for paired sample comparisons. The results were visualised via the R packages ggplot2 (v.3.4.3) [22] and ggpubr (v.0.6.0) [23].

Prognostic and diagnostic analysis

Using the TCGA cohort, we evaluated the impact of DBF4 gene (ENSG00000006634) expression on the overall survival (OS) of ccRCC patients in different subgroups. Patients were divided into high- and low-expression groups according to whether their DBF4 expression levels were above or below the mdian. Univariate Cox regression analyses were carried out to assess the significance of DBF4, gender, age, clinical stage and AJCC stage in predicting OS in patients with ccRCC. The significant factors were further examined by multivariate analysis in accordance with the results of univariate regression analysis. The p values and HRs of statistically significant factors are shown in the forest plot. Both survival analysis and Cox analyses were carried out with the R packages survival (v.3.3-1) [24] and survminer (v.0.4.9). The diagnostic value of DBF4 was assessed through receiver operating characteristic (ROC) curves plotted via the R package pROC (v.1.18.0).

Similar gene detection and enrichment analysis

Genes that exhibited an expression pattern comparable to that of DBF4 in ccRCC were identified via GEPIA2 (http://gepia2.cancer-pku.cn/#index) [25]. Genes similar to DBF4 were defined as the top 100 genes with the highest Pearson correlation coefficient. Gene Ontology (GO) and KEGG pathway enrichment analyses of genes similar to DBF4 were subsequently conducted with the R package clusterProfiler (v.4.4.4; parameters: Padj ≤ 0.05) [26]. The enrichment analysis results were visualised via the R package circlize (v.0.4.15) [27]. The protein‒protein interaction (PPI) network of DBF4 was constructed through the Search Tool for the Retrieval of Interacting Genes (STRING) online database (https://string-db.org/) [28]. Afterwards, the PPI network data were further processed and displayed via Cytoscape software version 3.9.1 [29].

Gene set enrichment analysis

To elucidate the potential mechanisms by which DBF4 impacts the progression of ccRCC, gene set enrichment analysis (GSEA) was conducted via GSEA software (v.4.3.0) (v.4.3.0) [30]. Within the TCGA cohort, 613 ccRCC samples were divided into two groups: samples with DBF4 expression levels above the median were defined as high-expression samples, whereas those with DBF4 expression levels below the median were defined as low-expression samples. The gene expression data and grouping information for all samples were then used as inputs for analysis based on the database of hallmark gene sets (h.all.v2023.1.Hs.symbols.gmt). Gene sets with |NES| > 1, NOM p < 0.05, and FDR q < 0.25 were considered significantly enriched. The significantly enriched gene sets were visualised via the R package pplot2 (v.3.4.3).

Relationship between DBF4 expression and immunity

The analysis of whether DBF4 is related to immunomodulators, including immunoinhibitors and immunostimulators, was carried out using the TISIDB database (http://cis.hku.hk/TISIDB/index.php) [31]. The top five immunoinhibitors and immunostimulators most highly correlated with DBF4 in ccRCC were identified, and a correlation bubble heatmap was subsequently generated via the R package ggplot2 (v.3.4.3).

Clinical patient specimens

Seventy-five paired ccRCC and adjacent nontumor renal tissues were used to construct a tissue microarray (HKid-CRCC150CS-02; Shanghai Outdo Biotech Co. Ltd., China). The study was approved by the ethics committee of the Second Affiliated Hospital of Hainan Medical University, and informed consent was obtained from the patients for the use of their clinical information and specimens.

Cell lines and cell culture

The RCC cell lines A498, and 786-O were purchased from IMMOCELL (Xiamen, China) and were subjected to validation via short tandem repeat (STR) analysis. A498 cells were grown in MEM (Gibco, C12571500BT) supplemented with 10% fetal bovine serum (FBS) (Gibco, 10099141 C) and RPMI 1640 (Cytiva HyClone, SH30809.01) supplemented with 10% FBS was used for the culture of 786-O cells. All the cells were maintained at 37 °C, in a 5% CO2 incubator.

Specific knockdown of DBF4 expression

The shRNA targeting the human DBF4 lentiviral expression vector was constructed as described previously [32]. Specifically, three DNA templates were created to target various locations of the shRNAs (Table S1). These templates were annealed and subcloned and inserted into the sites between Age І and EcoR І to construct pLKO.1-DBF4-shRNA (named DBF4-sh1, DBF4-sh2, and DBF4-sh3). The constructed vectors were verified via PCR and direct DNA sequencing. The primers used are shown in Table S2. The constructed pLKO.1-DBF4-shRNA lentivirus was subsequently used to transfect A498 and 786-O ccRCC cells via Lipofectamine™ 2000 (Invitrogen) with polybrene (8 mg/mL). The infected cells were selected and harvested for further verification and experiments.

Cell apoptosis and cell cycle analysis

RCC 786-O or A498 cells seeded in 6-well plates were infected with pLKO.1-shDBF4 lentivirus containing 10 µg of polybrene per mL upon reaching 80% confluence. After an additional 48 h of culture, an annexin V-FITC/PI apoptosis detection kit (BD) was used to analyse the degree of cell apoptosis according to the manufacturer’s instructions. For cell cycle detection, the cells were fixed with 70% precooled alcohol at 4 °C overnight, washed and then resuspended in 0.2% Triton X-100 PBS supplemented with 10 µL of RNase A (100 mg/mL) for 30 min at 37 °C and then incubated with 7-AAD (Beyotime, C1053S) for 30 min at 4 °C. Cell apoptosis and the cell cycle were analysed via flow cytometry (BD).

Cell proliferation

The proliferation of ccRCC cells subjected to DBF4 knockdown was assessed via a cell counting kit 8 (CCK-8) (Monmouth Junction, NJ, USA). After DBF4 was knocked down with pLKO.1-DBF4-shRNA, 2,000 786-O or A498 RCC cells were seeded in 96-well plates and cultured in complete medium for 3 days. 10 µL of CCK-8 reagent was added at intervals of 24, 48, and 72 h, after which the mixture was incubated at 37 °C for 2‒3 h. The OD value at 450 nm was measured with a Multiskan (Thermo Fisher, USA).

Cell migration and invasion assays

Tumor cell migration ability was determined by scratch wound healing assays. Confluent cells were starved for 18 h, and then a sterile pipette tip was used to create a scratch in the monolayer. Wound closure was monitored by imaging each well every 6 h. Tumor migration ability was assessed by measuring the scratch closure percentages in phase-contrast images. For the Transwell invasion assay, 100 µL of serum-free medium containing 5 × 104 starved ccRCC cells was cultured in the upper chamber (Corning, USA) with Matrigel (BD Biosciences, USA). In the lower chamber, 500 µL of medium supplemented with 10% FBS was used as the attractant. Following a 24-h period of cultivation, the invading cells on the bottom were fixed with 4% paraformaldehyde and then stained with 0.1% crystal violet (Beyotime, China). Finally, the stained cells were counted under a microscope.

Colony formation assay

Colony formation assays were also conducted in vitro to assess tumor cell proliferation. Briefly, 1,000 cells in 500 µL of medium were seeded in the wells of a 6-well plate. After approximately 14 days of subculturing, the colonies were stained with 0.1% crystal violet for 10–30 min at room temperature. Finally, the colonies were photographed and counted. The experiment was carried out three times independently.

Quantitative real-time polymerase chain reaction (qRT‒PCR)

Total RNA was extracted with a total RNA kit II (Omega Bio-Tek, Norcross, GA, USA). The RNA was reverse transcribed to cDNA by RT Master mix for qPCR (MCE, HY-K0510). qRT‒PCR was performed on an ABI 7500 Real-Time PCR System with SYBR Green qPCR master mix (MCE, HY-K0501) to quantify DBF4 expression with the following primer sequences: DBF4 left primer, 5’-TGCAGTCCATTTGATGTAGACAAG-3’ and right primer, 5’-GAGGTTCCACCATACTTATCGCC-3’; and 18 S left primer, 5’- CGACGACCCATTCGAACGTCT-3’ and right primer, 5’- CTCTCCGGAATCGAACCCTGA-3’. After 40 cycles at 94 °C for 30 s and 58 °C for 60 s, the result was normalised to 18 S rRNA (ΔCt). The relative DBF4 expression level was calculated and is presented as the fold change (fold change = 2−ΔΔCt) compared with that of the control group.

Western blot

Western blotting was conducted following the protocol previously described by [33]. In brief, proteins were separated and transferred to PVDF membranes and then incubated with primary antibodies against DBF4 (abcam, ab124707, 1:10,000 dilution), CDC7 (Proteintech, 17980-1-AP, 1:3000 dilution), Cylclin D1 (Proteintech, 26939-1-AP, 1:4000 dilution), CDK4 (Proteintech, 11026-1-AP, 1:3000 dilution), phospho-MCM2(Ser139)(cell signaling, #12958, 1:1000 dilution)or beta actin (Proteintech, 20536-1-AP, 1:5,000 dilution) in TBST at 4 °C overnight. After being washed three times, the membranes were incubated with goat anti-rabbit IgG secondary antibodies (Proteintech, SA00001-2, 1:10000 dilution) at room temperature for 1 h. Finally, enhanced chemiluminescence reagents (ECL, Thermo Fisher, Rockford, IL, USA) were used to visualise the signals.

Immunohistochemical staining

The tumor and adjacent tissues were collected to confirm DBF4 expression in ccRCC. The protocol was approved by the ethics committee of the Second Affiliated Hospital of Hainan Medical University (LW2023058). The paraffin-embedded tissues were deparaffinised via the standard xylene-deparaffinisation procedure, stained with anti-DBF4 (BIOSS antibodies, bs-7895R) at 4 °C overnight, and then incubated with an HRP-conjugated secondary antibody (Cell Signalling Technology, #7074S, 1:1,000 dilution) for 1 h at 37 °C. Then, haematoxylin and eosin were used to counterstain the sections. The slides and the intensity of immunohistochemical staining were analyzed under light microscopy.

Animal experiments

The animal experiments were conducted in compliance with a protocol endorsed by the Ethics Committee of Hainan Medical University (HYLL-2022-410). Briefly, male BALB/c nude mice (6 weeks old) were procured from Charles River Laboratories (Beijing, China) and housed under specific pathogen-free conditions at a temperature ranging from 21 to 25 °C and humidity ranging from 50 to 60%. Six mice were housed in a single cage and provided free access to sufficient food and water in their enclosures. Twelve mice were then subcutaneously injected with 4 × 106 786-O/shNC cells or 786-O/shDBF4 cells (n = 6) in the lower back region. The total number of animals was calculated via the E value method, as reported [34]. The tumors were measured with callipers every 4 days, starting 21 days post inoculation. The tumor volume was calculated using the following formula: length × width2/2. On the 41st day after inoculation, the mice were euthanised through the administration of 100% CO2 gas, with a flow rate of 30–70% of the chamber volume per minute. Additionally, the tumors were weighed and photographed. For further analysis, the tumor tissue sections were stained for Ki-67 to enhance the validity of the vivo experiments.

Statistical analyses

Statistical analyses were conducted using GraphPad Prism version 8.0 and R version 4.3.1. Quantitative data are presented as means ± SDs. The Wilcoxon test was used to compare expression differences of DBF4 between tumor and normal samples, as well as among samples with varying clinical characteristics, based on external datasets. Differences between the DBF4 knockdown and control groups in ccRCC were assessed using Student’s t-test. P value of less than 0.05 was regarded as statistically significant.