Unveiling NUSAP1 as a common gene signature linking chronic HBV infection and HBV-related HCC

Six datasets were downloaded from the Gene Expression Omnibus (GEO) database (https://​www.​ncbi.​nlm.​nih.​gov/​geo/​), including the gene expression profile data and related annotation files. The obtained datasets were screened by the following criteria: (1) Profile information encompassed both experiment and control groups. (2) All samples were derived from liver tissues and had not undergone chemoradiotherapy prior to surgical resection. (3) All datasets provided raw data suitable for further analysis. From these datasets, GSE121248, GSE83148 and GSE55092 were chosen to screen differentially expressed genes (DEGs). Additionally, GSE84044, GSE136247 and GSE65359 were chosen for subsequent validation.

We used GEO2R (https://​www.​ncbi.​nlm.​nih.​gov/​geo/​geo2r/​) for preliminary comparison and downloaded the series matrix files of all the datasets from GEO datasets. The DEGs between experimental and control samples were identified using the “limma” package. Fold changes (FCs) were calculated for individual gene expression. DEGs were determined based on specific cutoff criteria of padj < 0.05 and |logFC|> 1.0. The DEGs was presented in volcano plot using “ggplot2” package, and in a heatmap using “pheatmap” package (Supplementary Fig. 1) respectively. The online Venn diagram tool jvenn (http://​jvenn.​toulouse.​inra.​fr/​) was used to identify the common DEGs [14].

The “ClusterProfiler” package was utilized for conducting Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses. The GO analysis was divided into three categories: cellular component (CC), molecular function (MF), and biological process (BP), allowing examination of physiological functions.

The common significant DEGs were mapped to the STRING database https://​cn.​string-db.​org/​to construct a protein–protein interaction (PPI) network. For the PPI network analysis, a confidence level greater than 0.4 was set, and the network was visualized using Cytoscape software. Significant modules and hub genes were identified using the MCODE and CytoHubba. To further elucidate the biological features and mechanism of interacting proteins of the DEGs, functional annotation analysis was performed using another online software Metascape (http://​metascape.​org/​).

The expression levels of the identified hub gene were validated in three GEO validation datasets. A comparison between the samples and controls was performed using the Wilcoxon test and a p value < 0.05 was considered significant.

Receiver operating characteristic (ROC) curves were generated to assess the predictive accuracy of the hub genes using the “pROC” package in the R language. To further validate the chosen gene, three online databases were utilized. The HPA database (https://​www.​proteinatlas.​org/​) provided information on tissue and cellular distribution of various human proteins using immunoassay technology. The Gene Expression Profiling Interactive Analysis (GEPIA) database (http://​gepia.​cancer-pku.​cn/​) was employed for validating gene expression. The HCCDB database (http://​lifeome.​net/​database/​hccdb/​) served as a specific resource for Hepatocellular Carcinoma Diagnostic and Survival Analysis [15].

Based on the median expression of NUSAP1, HBV-infected patients were divided into high- and low-expression groups. The WEB-based Gene Set Analysis Toolkit (http://​www.​webgestalt.​org/​) and the “cluster Profiler” package [16] were applied for GSEA analysis. GSEA analysis allowed us to examine functional categories, calculate enrichment scores of gene sets, and discover different functional phenotypes by comparing biological pathways between the two groups. Only KEGG pathways with false discovery rate (FDR) ≤ 0.05 were considered significant and included in the analysis.

The Tumor Immune Estimation Resource (TIMER 2.0) (http://​timer.​cistrome.​org/​) was used to explore the correlation between NUSAP1 expression and immune cell infiltration.

To evaluate the immune infiltration of NUSAP1 in CHB and HBV-HCC patients, the R packages of “GSVA” were utilized to quantify the relative abundance of 16 immune cells and 13 immune-related pathways. Subsequently, we compared the enrichment scores across the groups and visualized the results using the “ggpubr” package.

The human liver cancer cell lines HepG2 and HepG2.2.15 were provided from Wuxi Red Cross Blood Center with the original source being Cell Resource Center (IBMS, China). HepG2 and HepG2.2.15 cells were cultured in DMEM (Hyclone, (Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS; Biological Industries, Israel) and 1% penicillin–streptomycin (P/S; Hyclone) at 37 ℃ in 5% CO₂ environment.

The pHBV1.3 plasmid containing 1.3 copies of the HBV genome was kindly provided from Wuxi Red Cross Blood Center. The plasmid was originally purchased from Fenghui Biological Co., Ltd (Changsha, China). HepG2 cells were transfected with plasmids containing the HBV whole genome, while HepG2.2.15 cells were transfected with NUSAP1 siRNA using Lipofectamine3000 reagent (Thermo-Fischer Scientific, MA, USA) according to the manufacturer’s instructions. To authenticate the siRNA, we carried out three transfections and validated the results with three repeating trials.

si-NUSAP1_1: CCUUGAAAGGCUACAUUAATT;

si-NUSAP1_2: GGAAGACUCUCUGUGGCUUTT;

si-NUSAP1_3: CCAAGACUCCAGCCAGAAATT.

Liver cancer cells were lysed using RIPA buffer (Cell Signal Technology, MA), centrifuged for the supernatant. The protein concentration was measured using bicinchoninic acid (BCA) assay (Cwbio, Beijing, China). The lysates were then diluted in loading buffer and denatured by heating at 100 ℃. Standard Western blot assay were performed using NUSAP1 antibody (Proteintech, 12024-1-AP) and GAPDH antibody (Abcam, ab77109) as the loading control.

Total RNA was extracted from the liver cancer cells using Trizol reagent (Invitrogen, USA) and cDNA was synthesized using the M-MLV Reverse Transcriptase Kit (Cwbio) following the manufacturer’s instructions. RT-PCR was performed using Real SYBR Mixture (Cwbio) on a Lightcycler 480 II instrument (Roche Applied Science, USA). GAPDH severed as the internal control.

NUSAP1 forward primer: 5ʹ-AGCCCATCAATAAGGGAGGG-3ʹ;

NUSAP1 reverse primer: 5ʹ-ACCTGACACCCGTTTTAGCTG-3ʹ.

GAPDH forward primer: 5ʹ-TGTTGCCATCAATGACCCCTT-3ʹ;

GAPDH reverse primer: 5ʹ-CTCCACGACGTACTCAGCG-3ʹ.

HepG2.2.15 cells transfected with NUSAP1 siRNA were seeded in 96-well plates at a density of 1 × 10³/well for the specified duration. In total, 10 μL of CCK-8 reagent (CCK8; Vazyme, Nanjing, China) was added to each well and incubated for 2 h. The optical density (OD) at 450 nm was measured using a microplate reader (Thermo Fisher). The experiment was performed three times with six replicates.

HepG2.2.15 cells were cultured for 24 h and then subjected to serum deprivation for 24 h to synchronize the cell cycle. Afterward, the cells were cultured in normal medium for an additional 24 h. Subsequently, the cells were harvested and fixed with 70% ethanol at − 20 ℃ for 24 h. For propidium iodide (PI) staining, the cells were incubated in the presence of 50 μg/mL propidium iodide (PI; Sigma-Aldrich, St Louis, MO) and 100 ng/mL RNase A (Boehringer Mannheim, Indianapolis, IN) at room temperature in the dark for 1 h. The cell cycle progression in each sample were measured via FACS Canto II (BD, Mountain View, CA). The percentages of cells in the G1, S, and G2 phases were calculated using ModFit LT software (RRID:SCR_016106, BD). Each experiment was repeated three times.

Cell apoptosis was detected by flow cytometry with Annexin V-FITC Apoptosis Detection Kit (KeyGen Biotech, China) according to the manufacturer’s instructions. Cells were incubated with 2 μM Doxorubicin (Dox, Slleck) for 48 h. The cells were then harvested, rinsed with ice cold PBS, and then resuspended in 200 μL of binding buffer. Next, 10 mL of AV-FITC stock solution was added to cell suspensions and incubated for 15 min at room temperature. Subsequently, the cells were rinsed with 200 μL of ice-cold PBS and immediately analyzed using a FACScan flow cytometry.

Bioinformatics analysis was carried out using R software (version.4.2.1). Statistical significance was defined as p < 0.05 and FC > 1.5. In vitro experiments were repeated at least 3 times. Data were analyzed using SPSS 20.0 software (SPSS Inc., Chicago, IL, USA). p < 0.05 indicated that differences were statistically significant.

A total of six GEO datasets including GSE83148, GSE55092, GSE121248, GSE84044, GSE136247 and GSE65359 were chosen for this study. Details regarding each dataset are summarized in Table 1. The discovery cohorts, consisting of GSE83148, GSE55092, and GSE121248, were selected for identifying common DEGs, while the remaining datasets were categorized as validation cohorts.

Using the “limma” package, we screened for DEGs between CHB patients and healthy controls in GSE83148. We found that 927 genes were upregulated and 113 genes were downregulated in this dataset (Fig. 1A). We further identified DEGs between HBV-related HCC tissues and adjacent normal tissues: 747 genes were upregulated and 1000 genes were downregulated in GSE55092; while 328 genes were upregulated and 625 genes were downregulated in GSE121248 (Fig. 1B, C). As illustrated in Fig. 1D, E, 80 common DEGs were identified across these three datasets, with 64 being upregulated and 16 being downregulated.

In order to understand the biological importance of 80 commonly observed DEGs, we conducted GO analysis and KEGG pathway enrichment. Our findings revealed that these genes were primarily associated with cell cycle processes in terms of biological functions. Additionally, they were found to be related to chromosome, microtubule cytoskeleton, and spindle in terms of cellular components. Furthermore, as for molecular function, the genes were mainly enriched in drug binding and protein kinase activity (Fig. 2A–C). The GO analysis of specific DEGs revealed that they were closely associated with cell cycle activities, including nuclear division, chromosome organization, and cell cycle phase transition (Fig. 2D). Additionally, the KEGG analysis indicated that the DEGs were mainly enriched in the cell cycle, virus infection, oocyte meiosis, and the p53 signaling pathway (Fig. 2E).

To explore the potential connections among the proteins encoded by the commonly observed DEGs and identify key hub genes, we employed STRING to construct a protein–protein interaction (PPI) network for the 80 DEGs. Subsequently, module analysis was conducted using the MCODE plug-in in Cytoscape to identify significant clustering modules. As depicted in Fig. 3A, B, the most notable module consisted of 20 nodes and 183 edges. 20 candidate hub genes were selected from this module, including CENPE, TOP2A, NCAPG, PRC1, DLGAP5, PTTG1, TPX2, MAD2L1, TTK, AURKA, NEK2, PBK, CCNB2, NUSAP1, KIF20A, BUB1B, ASPM, RRM2, CCNA2 and CDK1.

Afterwards, the most significant module underwent GO analysis and KEGG pathway enrichment to investigate the underlying biological processes (Fig. 3C, D). The pathway enrichment results revealed that the gene function within this module was linked to protein binding, kinase activity, and cell cycle regulation, aligning with the biological function of common DEGs. Heatmaps of 20 candidate hub genes between CHB liver tissues and healthy tissues (Fig. 3E), between HBV-HCC tissues and adjacent normal tissues (Fig. 3F, G) were also illustrated to display their expression levels.

In order to investigate the genes that may have a significant impact on the progression from CHB infection to HBV-HCC, we utilized CytoHubba to identify hub genes among the common DEGs. Given the heterogeneous nature of biological networks, we employed multiple topological analysis algorithms, namely MCC, MNC, Degree, and EPC, simultaneously to predict and explore the top ten crucial hub genes in the PPI networks. The convergence of these 10 genes from the four algorithms highlighted NUSAP1 as the most pivotal hub gene (Fig. 4). We downloaded GSE65359, GSE84044 and GSE136247 from the GEO database and analyzed NUSAP1 expression in different status of HBV-related liver diseases. The expression of NUSAP1 were elevated in immune clearance status, compared to inactive carriers and immune tolerant status. Moreover, NUSAP1 showed great accuracy for the diagnosis of HBV-related immune clearance status (Fig. 5A–D). In advanced CHB status with fibrosis and inflammation, NUSAP1 exhibited higher expression levels compared to CHB status without advanced symptoms and could differentiate between CHB with and without fibrosis and inflammation with AUC_NUSAP1 = 0.860 (Fig. 5E, F). The expression level of NUSAP1 was also higher in HBV-HCC tissues and effective in distinguishing HBV-HCC tissues from adjacent normal tissues with AUC_NUSAP1 = 0.915 (Fig. 5G, H).

Using the HPA websites, we predicted the intracellular localization of NUSAP1 proteins. Compared to normal liver tissues, the NUSAP1 proteins were found to be located in the cell nucleus with strong intensity in HCC tissues (Fig. 6A). In GEPIA, the expression level and prognosis value of NUSAP1 in HCC patients based on LIHC data from TCGA were analyzed. More NUSAP1 were expressed in LIHC tumor samples. From stage I to stage III, the expression of NUSAP1 was gradually increased (Fig. 6B, C). In Fig. 6D, the Kaplan–Meier survival analysis of three HCCDB databases displayed higher levels of NUSAP1 in HCC tissues associated with poor overall survival among HCC patients, while NUSAP1’s expression in adjacent normal tissues didn’t correlate with survival probability.

GSEA analysis was conducted to identify distinctively regulated pathways between the high and low expression groups of NUSAP1, as well as to determine the activated signaling pathways in liver diseases associated with HBV. Activated pathways including DNA replication and cell cycle were found in HBV-HCC patients with overexpressed NUSAP1 (Fig. 7A, C). In CHB patients with high NUSAP expression, the enriched pathways were different from HBV-HCC patients and were associated with cell adhesion molecules, antibody production and antigen presentation (Fig. 7D, F). Our findings suggested that inhibiting the activation of pathways that are distinct from classic tumor-associated signaling pathways may hold potential for reversing the progression of chronic hepatitis.

According to TIMER database, NUSAP1 expression was positively correlated with B cells, CD8⁺T cells, CD4⁺T cells, macrophages, neutrophils, and dendritic cells (Fig. 8A). Specifically, a higher number of somatic copy number variations of NUSAP1 was associated with increased dendritic cell infiltration (Fig. 8B). Furthermore, when HBV-HCC patients were categorized based on NUSAP1 expression, the high and low NUSAP1 expression groups showed significant differences in the ssGSEA (single sample Gene Set Enrichment Analysis, ssGSEA) scores of 29 immune pathways (Fig. 8C). Similarly, in CHB patients, dividing them based on NUSAP1 expression revealed significant differences in the activation of immune pathways, indicating heightened immune signaling during inflammatory infections (Fig. 8D). These findings suggest that more immune pathways are activated in the inflammatory period, while some pathways may be inhibited during the tumor progression.

We compared the expression of NUSAP1 at both mRNA and protein levels in HepG2.2.15, HepG2 and HepG2 transfected with plasmid containing HBV whole genome (HepG2/HBV). The results indicated that after transfection with HBV plasmids, there was no variation in the RNA level of NUSAP1, but the protein level increased (Fig. 9A, B). Similarly, we transfected the HBV plasmid in Huh7 cells and noticed no dissimilarity in the RNA level of NUSAP1, yet the protein level decreased (Supplementary Fig. 2A, B). We selected HepG2.2.15 for functional validation experiments, which showed a relatively high expression of NUSAP1. We knockdown NUSAP1 using three siRNAs, siNUSAP1-1 didn’t show any effect (Fig. 9C, D), thus we continue the study using the siNUSAP1-2 and siNUSAP1-3 site, the proliferation ability of HepG2.2.15 was significantly reduced (Fig. 9E), as well as a decrease in colony number was observed in the NUSAP1 knockdown cells (Supplementary Fig. 3), suggesting that NUSAP1 may promote cancer cell proliferation. Annexin V/PI staining and flow cytometry were used to detect apoptosis in NUSAP1 knockdown HepG2.2.15 cells, revealing a significant increase in the proportion of cells undergoing apoptosis in the NUSAP1 knockdown group (Fig. 9F). Flow cytometry analysis after PI staining revealed that NUSAP1 knockdown HepG2.2.15 cells were arrested in the S phase of the cell cycle (Fig. 9G). Our findings suggested that NUSAP1 may promote cancer cell proliferation, influence cell cycle related pathways, and inhibit cell apoptosis, thereby contributing the survival of tumor cells.