The Tumor Immune Estimation Resource (TIMER, version 2.0) (http://timer.cistrome.org/) “exploration” module was used to display the differential gene expression of APOBEC3B in various tumor tissues and corresponding normal tissues (19).

Tumoral RNA-seq data of cervical cancer patients as well as mRNA expression data of paired normal tissue samples were from TCGA database and were downloaded from the Genomic Data Commons (GDC) data portal. Other data of normal tissue samples were obtained from the normal cervix in the GTEx V8 release version (https://gtexportal.org/home/datasets). A complete description of information for each sample was described in the GTEx official annotation (20). The GSE67522 and GSE7803 datasets used are from the GEO database (https://www.ncbi.nlm.nih.gov/geo/), and the downloaded data format is MINIML. Analyses were performed after grouping the clinical phenotype. The box plot was implemented by the R software package ggplot2, and heatmap was displayed by the R software package pheatmap.

We used the Limma package (version: 3.40.2) of R software to study the differential expression of mRNAs in the cervical cancer database. The adjusted p-value was calculated to correct data in TCGA. The following criteria were used to screen for differential genes in TCGA dataset: “Adjusted p < 0.05 and Log (fold change) >2 or Log (fold change) <−2.” Based on the expression level of APOBEC3B, we divided TCGA samples into A3B^low (<25% of the quantile distribution) and A3B^high (>75% of the quantile distribution) expression groups. To further confirm the underlying function of potential targets, the data were analyzed by functional enrichment. Gene Ontology (GO) is a tool for annotating genes with functions, especially molecular functions, biological pathways, and cellular components. The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis is a practical resource for the analytical study of gene functions and associated high-level genome functional information. To better understand the carcinogenesis of mRNA, the ClusterProfiler package (version: 3.18.0) in R was employed (21). We searched “cervical cancer” in NCBI (https://www.ncbi.nlm.nih.gov/gene). GSE26511 samples were divided into A3B^high expression group (n = 20) and A3B^low expression group (n = 19) and were analyzed using the criterion “p < 0.05 and Log (fold change) >1.5 or Log (fold change) <−1.5” (21). The common upregulated genes were obtained to perform the functional analysis. KEGG analysis and GO analysis were conducted using the common upregulated genes with Metascape (https://metascape.org/).

For the Cancer Genome Atlas (TCGA) database, we downloaded tumor RNA-seq (FPKM) from the GDC. We used the one-class logistic regression (OCLR) score algorithm constructed by Malta et al. to calculate mRNAsi. We used the Spearman correlation (RNA expression data), and then subtracted the minimum value and divided it by the linear transformation of the maximum value, mapping the dryness index to the range. All the above analyses were performed using the R package by the R Foundation for Statistical Computing (2020), version 4.0.3.

Tumor tissues and normal adjacent tissues of five cervical cancer patients were obtained from the Tissue Bank of the Obstetrics and Gynecology Hospital of Fudan University.

Tissue microarrays (TMAs) of cervical cancer specimens were obtained from the Tissue Bank of the Obstetrics and Gynecology Hospital of Fudan University. TMAs were constructed following previous protocols (22, 23). After TMA construction, a hematoxylin and eosin (HE) section of the recipient block was reviewed to confirm that the cores contained the intended region. TMA cutting was performed, and finished slides were embedded in paraffin for preservation at 4°C before immunohistochemistry assays.

The de-paraffinized sections were incubated with 20% goat serum for 30 min to block nonspecific binding and were then incubated with primary antibodies against APOBEC3B (Abcam, Cambridge, UK, 1:100) at 4°C overnight followed by an anti-rabbit secondary antibody (1:100) for 1 h at 37°C. Bound antibody was then visualized using the EnVision™ Detection Systems (Dako, Glostrup, Denmark). The expression of APOBEC3B was examined using immunohistochemistry according to the previous protocol (22, 23). Disagreements were settled by consensus. The study protocol was approved by the institutional review board of the hospital. All patients provided written informed consent.

The human cervical cancer cell lines HeLa and SiHa were obtained from MD Anderson Cancer Center, and the original source is the American Type Culture Collection (ATCC; Manassas, VA, USA). HeLa cells were cultured in 1640 complete medium and DMEM/high-glucose medium supplemented with 10% FBS in 5% CO₂ at 37°C, respectively. SiHa cells were cultured in DMEM/high-glucose medium in the same condition.

Based on the Protein Atlas (www.proteinatlas.org), the SiHa cell line has lower A3B expression relative to the HeLa cell line ( Supplementary Figure S1D ). Two cervical cancer cell lines (HeLa and SiHa) with different endogenous expressions of A3B were chosen. HeLa and SiHa cells were seeded in the 60-mm-diameter dishes at a density of 2 × 10⁵/well. SiHa cells were transfected with vector construct or APOBEC3B overexpression construct, while HeLa cells were transfected with APOBEC3B siRNA or scrambled siRNAs by Lipo2000 reagent according to the manufacturer’s instructions. The siRNA targeting APOBEC3B and scrambled siRNAs were synthesized and purified by RioBio Co. (Guangzhou, China). The APOBEC3B overexpression plasmid and vector constructs were obtained from Asia-Vector Biotechnology (Shanghai, China). Each experiment was repeated three times.

After 48 h of transfection, all transfected HeLa and SiHa cells were collected and lysed in 1× SDS lysis buffer (50 mM of Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, 1 mM of Na₃VO4, and 1 mM of PMSF); total protein was quantified by the BCA method. A total of 40 μg of protein per lane was loaded on an SDS-PAGE gel and transferred to a PVDF membrane (Millipore Corporation, USA), which was blocked with PBS containing 5% BSA and 0.05% Tween 20. The membrane was incubated with specific primary antibodies and followed by incubation with HRP-conjugated secondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA, USA). The labeled proteins were then visualized using fluorography using an enhanced chemiluminescence system (Thermo Scientific, Pierce Biotechnology, USA). Each experiment was repeated three times.

Relative cell viability was assayed by Cell Counting Kit-8 (CK04, Dojindo Laboratories) according to the manufacturer’s recommendation. HeLa and SiHa cells were grown in 96-well plates at a density of 3 × 10³/well and then transfected with APOBEC3B construct or siRNAs for the indicated time. In total, 20 µl of CCK-8 reagent was added to the cells and incubated for 1 h. The absorbance at 450 nm was measured by Multiskan Spectrum (Spectra Max190, Molecular Devices). Six replication experiments were performed.

The transwell chamber placed into the 24-well culture plate was called the upper chamber, and the culture plate was called the lower chamber. The cells were incubated in a serum-free medium and were seeded in the upper chamber; a complete medium containing 10% FBS is generally added to the lower chamber. The cells were fixed with 4% paraformaldehyde and then stained with 0.1% crystal violet (Solarbio, China) after 24 h of the migration experiment.

Cells were synchronized by serum starvation for 24 h and cultured in a complete medium for 24 h. Cells were then digested with 0.25% trypsin fixed in 75% ethanol. For PI staining, cells were incubated in the presence of 40 µg/ml PI (Molecular Probes) and 250 µg/ml RNase A (Roche Diagnostics Ltd.) at 37°C for 30 min. Progression into the different cell cycle phases was calculated by propidium iodide staining. The percentage of the positive cells and cell cycle progression in each sample were measured via flow cytometry. Each experiment was repeated three times.

In total, 3,000 cells (100 µl, cell suspension density: 3 * 10⁴/ml) were placed into each well of a 96-well plate. PBS solution was added around the 96-well plate to prevent loss of edge medium; 12 h after cell adhesion, the complete medium was replaced with cisplatin concentrations of 5, 10, 20, and 30 μM for 24, 48, and 96 h, respectively. Cell viability was then measured by the CCK8 assay as described above.

Triplicate assays were repeated in all molecular biology experiments. Experimental data are presented as the mean ± SD. The differences between the groups were analyzed via Student’s t-test and Mann–Whitney U test. A two-sided p < 0.05 was designated as statistically significant. All analyses were performed using GraphPad (version 8.4.3).

The expression of APOBEC3B was evaluated in different tumor types and adjacent normal tissues using the TIMER database. As shown in Figure 1A , the APOBEC3B expression level was significantly higher compared with adjacent normal tissues in a variety of tumors. In addition, the expression of APOBEC3B was significantly higher in HPV-positive HNSC compared with HPV-negative HNSC. However, in colon adenocarcinoma (COAD), rectum adenocarcinoma (READ), and thyroid carcinoma (THCA), the expression level of APOBEC3B was significantly lower than that in adjacent normal tissues.

We further found that APOBEC3B expression was higher in HPV16-positive cervical cancers than in HPV16-positive nonmalignant tissues and HPV-negative histologically normal controls in GSE67522 ( Figure 1B ). Moreover, APOBEC3B was also highly expressed in cervical cancers than in high-grade squamous intraepithelial lesions and normal controls in GSE7803 ( Figure 1C ). TCGA database showed similar results ( Figure 1D ). In summary, all results showed that APOBEC3B is markedly elevated in cervical cancer tissues. This finding was further verified by the immunohistochemical results of our clinical samples ( Figure 1E ). In cervical cancers, APOBEC3B expression was strongly positive compared with the normal cervix.

Cervical cancer cases in TCGA were divided into two groups based on the high and low expression of APOBEC3B following previous methods. In total, 236 upregulated genes and 24 downregulated genes ( Figures 2A, B ) were identified in TCGA dataset. Studies have found that TFF1, one of the downregulated genes in the A3B^high group, was regulated by APOBEC3B in breast cancers (24). We then performed GO and KEGG analyses to evaluate the function of those altered genes. GO analysis revealed that the upregulated genes with APOBEC3B were significantly enriched in epidermis development ( Figure 2D ), and the downregulated genes were significantly enriched in cell import ( Figure 2F ). Moreover, KEGG analysis demonstrated that upregulated genes with APOBEC3B were mostly enriched in the Rap1 signaling pathway, estrogen signaling pathways, and cytokine–cytokine receptor interaction ( Figure 2C ). Moreover, downregulated genes for KEGG analysis also showed that estrogen signaling pathways were significantly enriched ( Figure 2E ).

Cervical cancer cases in GSE26511 were divided into two groups based on the high and low expression of APOBEC3B as described before. In total, 78 upregulated genes and 38 downregulated genes ( Figures 3A, B ) were identified in the GSE26511 dataset. We then performed GO and KEGG analyses to evaluate the function of those altered genes. GO analysis revealed that the genes, upregulated by APOBEC3B were significantly enriched in epidermis development and skin development ( Figure 3D ), and the downregulated genes with APOBEC3B were significantly enriched in myeloid leukocyte migration, leukocyte chemotaxis, as well as cell chemotaxis ( Figure 3F ). Moreover, KEGG analysis demonstrated that upregulated genes with APOBEC3B were mostly enriched in drug metabolism–cytochrome P450 ( Figure 3C ). Moreover, downregulated genes for KEGG analysis also showed that cytokine–cytokine receptor interactions were significantly enriched ( Figure 3E ).

Followed by the Venn diagram analysis, 39 genes were both upregulated from differential expression genes (DEGs) of A3B^low and A3B^high in TCGA dataset and DEGs of A3B^low and A3B^high in the GSE26511 dataset ( Supplementary Figure S1B ). Consistently, the gene enrichment analyses showed that upregulated genes were enriched in pathways related to cell fate commitment and regulation of extrinsic apoptotic signaling pathways ( Supplementary Figure S1C ). We further analyzed the relativity between one-class logistic regression (OCLR) and APOBEC3B gene expression, and our results revealed that tumor tissues with APOBEC3B high expression have higher stemness, which evaluates the proliferative activity and malignant potential of tumor cells ( Supplementary Figure S1A ). All suggest that A3B is likely to serve as an important factor in the development of cervical carcinogenesis in cervical cancer.

To investigate the effect of APOBEC3B on cervical cancer progression, transwell migration and immunofluorescence staining were conducted. SiHa cells were transfected with a plasmid expressing APOBEC3B or an empty vector (EV). HeLa cells were transfected with two independent siRNAs targeting human APOBEC3B (siA3B-1 and siA3B-2). As shown in Figure 4A, B , the expression of APOBEC3B protein decreased significantly after transfection with siRNA but increased remarkably after transfection with a plasmid expressing APOBEC3B, which confirmed the effectiveness of our gene knockdown and overexpression method. As shown in Figures 4E, F APOBEC3B overexpression demonstrated a positive effect on the migratory capacity of SiHa cells compared with the empty vector, while APOBEC3B knockdown inhibited the migratory capacity of HeLa cells compared with the negative control (NC). Subsequently, immunofluorescence staining of apoptosis-associated proteins, including Bcl-2 and caspase-3, as well as TUNEL staining were performed to explore the effect of APOBEC3B on cell apoptosis. As shown in Figure 5 , fewer TUNEL-positive cells were detected in the APOBEC3B overexpression group than in the EV group, while more TUNEL-positive cells were detected in the APOBEC3B knockdown group than in the NC group. Furthermore, fewer caspase-3 cells were positive in the APOBEC3B overexpression group than in the EV group, while more caspase-3-positive cells were detected in the APOBEC3B knockdown group than in the NC group. Similar results in immunofluorescence of Bcl-2 could also be observed in Figure 6A . Quantification of immunofluorescence staining is shown in Figure 6B . Cleaved caspase-3 was also investigated by Western blot ( Figure 4A ). A3B overexpression led to a significant reduction of cleaved caspase-3, which implied that A3B may act as a negative regulator of apoptosis. All these findings suggest that APOBEC3B was an important regulator of apoptosis in cervical cancer cells and was associated with the malignant phenotype of cervical cancer cells.

It is suggested that APOBEC3B may play an important role in cancer progression (23); therefore, we observed whether APOBEC3B promoted cell viability of human cervical cancer cells. As shown in Figures 8A, B , APOBEC3B overexpression led to elevated cell viability in SiHa cells compared to the control groups, and A3B knockdown contributed to a decrease in cell viability in HeLa cells compared to the negative control groups ( Figures 8C, D ). It is speculated that the expression of A3B might influence the cell cycle and the apoptosis of cervical cancer (15). Western blot was performed to detect the protein expression levels of p53-mediated signaling pathways p53, p21, cyclin A, cyclin B1, cyclin E1, and NF-kB, which were relevant to the proliferation of tumor cells (25–27) ( Figures 4C, D ). As shown in Figure 4A , we found that silencing APOBEC3B would promote the expression of p53. At the same time, the expression of cyclin A, cyclin E1, and p21 was significantly upregulated, which were critical cell cycle proteins for cell proliferation (28–31). Considering the strong relationship of E6 and E7 proteins with A3B, we also tested the expression of E6 and E7 proteins ( Figures 4C, D ). A3B overexpression led to upregulation of E6 and E7 proteins, while A3B knockdown showed opposite results. Therefore, A3B may promote p53 degradation by upregulating the E6 protein. The elevated p21 is consistent with increased p53 protein upon A3B knockdown in the HeLa cell line. Conversely, both p53 and p21 proteins are reduced in A3B over expressing SiHa cell line.

The cell cycle protein-dependent kinase (CDK) inhibitor (p21) is an important negative regulator in cell cycle regulation. In response to DNA damage, it is stimulated by p53, which blocks the cell cycle (32). It is considered to be a tumor suppressor due to its ability to block cell cycle progression (33). Elevated levels of cell cycle proteins D1 and E have been observed in tumors, which can lead to uncontrolled cell proliferation. The active cell cycle protein/CDK complex can be regulated by binding to CDK inhibitors, thereby inhibiting cell cycle progression from G1- to S-phase (34). The result shows that overexpression of A3B in SiHa increased E7 protein and A3B siRNA reduced E7 protein in the HeLa cell line. It is known that E7 is responsible for G1-S transition by binding and degrading pRB pocket proteins. The active pRB normally downregulates the expression of cyclins and S-phase genes in the G1-phase of the cell cycle. Thus, the change in the protein level of cyclins and CDKs in response to A3B modulation may be due to altered E7 proteins. NfkB may have a supplementary role in regulating cyclins and CDKs (35, 36). However, reduced NfkB in the A3B knockdown HeLa cell line may have promoted the apoptotic pathway as indicated by elevated cleaved caspase-3. Flow cytometry (FCM) results demonstrated that upregulating A3B led to a significant S and G2 upregulation ( Figures 7A–C ), while A3B knockdown contributed to an increase in the G1-phase and a reduction of the G2-phase. Both results have indicated that APOBEC3B is related to the cell cycle thus promoting cell proliferation.

In order to explore the effect of APOBEC3B on cytotoxicity drugs, we then treated SiHa cells and HeLa cells with different concentrations of cisplatin. Cell proliferation was determined by the CCK8 assay at 24, 48, and 72 h. As shown in Figures 8E–H , cervical cancer cell lines with high expression of A3B have lower sensitivity to cisplatin in a time- and concentration-dependent manner. Therefore, A3B may be a promising therapeutic target for cervical cancer.