Human HCC tissue samples were obtained from treatment-naive HCC patients who underwent surgical resection at the Department of Hepatobiliary Surgery, Affiliated Cancer Hospital of Guangxi Medical University (Nanning, China). All patients were informed about the procedure and signed an informed consent form. HCC tissues exhibiting characteristic macroscopic features were obtained from tumor nodules and subsequently confirmed through histological examination using hematoxylin and eosin (H&E) staining. Adjacent non-tumor tissues, free of histopathologically detectable tumor cells, were collected from regions located 5 cm away from the tumor margin.

The collection of all samples and medical data involving human subjects was approved by the Ethics Committees of the Affiliated Cancer Hospital of Guangxi Medical University (LW2024131) and conducted in accordance with the ethical principles of the Declaration of Helsinki. Upon admission, all patients gave written consent for the analysis and publication of their anonymized medical data for research purposes.

RNA sequencing library preparation was carried out according to the manufacturer’s protocols. Total RNA was extracted using the Trizol reagent RNA extraction kit (Invitrogen, USA), and rRNA depletion was performed to enrich both mRNA and non-coding RNAs. The enriched RNA was then fragmented into short segments (~200–700 bp). First-strand cDNA synthesis was performed using a random hexamer primer, followed by second-strand synthesis using DNA polymerase I. After purification, sequencing adapters were ligated to the cDNA fragments. The second cDNA strand was selectively degraded using Uracil-N-Glycosylase (UNG). The final cDNA library was generated by Polymerase Chain Reaction (PCR), followed by DNA isolation and purification, and quantified using an Agilent 2100 (Agilent Technologies, USA). The cDNA library was sequenced on Illumina platforms in accordance with commercial protocols.

Clean data was generated by removing low-quality reads and adapter sequences from raw sequencing reads using Trim Galore (v.0.6.10). The clean reads were aligned to the hg38 reference genome using Hisat2 (v. 2.2.1) with default parameters. A gene expression matrix was constructed using featureCounts (v.2.0.6). Normalization of gene expression and identification of differentially expressed genes (|Fold Change| > 2, p < 0.05) were performed using the R package ‘DESeq2’ (v.1.44.0). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted using the R package ‘clusterProfiler’ (v.4.12.0).

The DEGs between Cs ⁺ HCC tumors and Cs ^- HCC tumors were used to build protein–protein interactions and STRING (21) —a protein interactome database was used in this analysis. Finally, cytoscape (v.3.7.2) was used for visualization of the protein–protein interactions networks.

The up- and down-stream 3K sequences of up- and down-regulated genes transcriptional start sites (TSSs) were extracted, and motif enrichment analysis was performed based on the HOMER function findMotifsGenome.pl with the default option (p < 0.01).

The R package ‘CIBERSORT’ (v.0.1.0) was utilized to assess immune infiltration by applying the principle of linear support vector regression to deconvolute the expression matrix of 22 human immune cell subtypes. Finally, the Wilcoxon test was used to identify differential immune cells.

We selected the top 10 up-regulated and top 10 down-regulated genes between Cs ⁺ HCC tumors and Cs ^- HCC tumors based on fold change rankings for survival analysis using the GEPIA2 website (http://gepia2.cancer-pku.cn/#survival), utilizing the The Cancer Genome Atlas-Liver Hepatocellular Carcinoma (TCGA-LIHC) cohort with default parameters. A p-value of less than 0.05 was considered significant, indicating an association between the mRNA and the survival prognosis of HCC patients.

We meticulously gathered data from 1,461 HCC patients who were diagnosed at the Tumor Hospital of Guangxi Medical University between July 2013 and December 2022. All these patients underwent curative resection. The criteria for inclusion in this retrospective study were (1): HCC confirmed through postoperative pathological analysis, (2) patients consented to post-operative follow-up visits, (3) all patients underwent testing for Cs infection at the time of initial diagnosis. The exclusion criteria were: (1) patients who had received previous anti-tumor treatments such as radiotherapy or chemotherapy, (2) cases lacking a definitive pathological diagnosis, (3) patients with other tumor-related disease, (4) those with recurrent HCC, and (5) incomplete laboratory, follow-up data and died perioperatively.

The diagnostic criteria for infection with Cs are as follows (14, 22): (1) Preoperative imaging (MRI, CT, microscopy, or ultrasound) showing Cs eggs or adult worms in the intrahepatic bile ducts; (2) Intraoperative or postoperative pathological examination detecting adult Cs in the liver or gallbladder; (3) Preoperative fecal examination revealing Cs eggs.

Data collection was performed by two independent investigators, JXC and CBW, with validation by a third investigator, TJH. The data collection included the pathological characteristics of liver resection specimens from 1,461 HCC patients, focusing on indicators such as the presence of MVI and Ki67 positivity rate. Clinical characteristics of patients in the retrospective study can be found in Supplementary Table S1 .

All pathological specimens of postoperative liver cancer were fixed in formalin and subsequently embedded in paraffin wax, then sectioned into continuous slices of 4 μm thickness. Hematoxylin and eosin (HE) staining revealed clusters of tumor cells within the lumen of endothelial-lined blood vessels, primarily located in the peritumoral portal vein branches (including vessels within the capsule), confirming the presence of MVI. To further investigate cellular proliferation, immunohistochemical analysis was performed using Ki67 mouse anti-human monoclonal antibody (dilution 1:200; Maixin Biotech, China). After deparaffinization, the sections underwent high-pressure antigen retrieval in citrate buffer (pH 6.0), followed by treatment with 3% H₂O₂ to eliminate endogenous peroxidase activity. Normal non-immune sheep serum was then used to block the sections, reducing nonspecific binding. The sections were subsequently incubated overnight at 4℃ with the primary antibody. Finally, the DAKO EnVision detection system was used for visualization, with phosphate-buffered saline (PBS) serving as the negative control to ensure the reliability of the experimental results.

According to the microscopic observation of microvascular invasion (MVI) results, the presence of MVI is identified when there are more than 5 instances of MVI in postoperative liver cancer tissue, or when MVI occurs in the distant cancer-adjacent liver tissue region (> 1 cm).

Ki67 is primarily expressed in the cell nucleus, with positive results appearing as yellow or brownish-yellow staining. The scoring method for Ki67 is based on the Ki67 index, which refers to the percentage of tumor cells with stained nuclei relative to the total number of tumor cells observed.

All diagnoses of Ki67 and MVI were independently confirmed by two pathologists. In cases of discrepancy, slides were reviewed by a pathologist with a title of associate chief physician or higher, and the diagnosis of Ki67 and MVI was jointly confirmed by all three pathologists.

All patient follow-up data are carefully managed by healthcare professionals. Disease status and the date of metastasis are determined through telephone follow-up or outpatient monitoring systems. Tumor recurrence and metastasis are assessed based on radiological findings from CT or MRI scans, combined with pathological examination results. Using this information, the metastasis rates of Cs-infected and non-Cs-infected HCC patients were compared and analyzed.

Cs adult worms were washed several times with PBS (Gibco, USA) containing 1% Penicillin-Streptomycin solution (Solarbio, China) and then cultured in glass dishes using a phenol red-free 1640 medium (Solarbio, China). After 48 hours, the culture medium was collected and centrifuged at 4,000 rpm for 30 minutes at 4°C to remove eggs and cell debris. The resulting supernatant was then centrifuged at 12,000 rpm for 45 minutes at 4°C to obtain the supernatant. The supernatant was dialyzed against PBS to remove small molecules and then concentrated using sucrose or freeze-dried for further analysis. The molecular weight and concentration of the ESPs were measured, and the samples were stored at -80℃. Before use, the samples were sterilized via 0.22 μm filtration to ensure they were free from contaminants.

Human umbilical vascular endothelial cells (HUVECs) and human HCC cell lines MHCC-97H were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). MHCC-97H cells were cultured in high-glucose Dulbecco’s Modified Eagle Medium (DMEM) (Gibco, USA) supplemented with 10% fetal bovine serum (FBS) (Wisent, Canada) and 1% Penicillin-Streptomycin solution (Solarbio, China). HUVECs were cultured in endothelial cell medium (ECM) supplemented with endothelial cell growth supplement (ECGS) and 5% fetal bovine serum (FBS) (ScienCell Research Laboratories, USA), and utilized for experiments between passages 4 and 8. In the experimental group, the concentration of CsESPs was set at 50 μg/ml, while the control group received an equivalent volume of 1×PBS (Gibco, USA) solution. Cells were incubated at 37℃ in a humidified atmosphere containing 5% CO₂ and 95% air.

Cell proliferation was evaluated using the Cell Counting Kit-8 (CCK8) assay. Briefly, cells were seeded at a density of 2×10³ cells/100 μL per well in 96-well plates. In the experimental group, the concentration of CsESPs was set at 50 μg/mL, while the control group received an equivalent volume of 1×PBS. On Days 0, 1, 2, and3, the medium was replaced with 100 μL of serum-free medium in each well, followed by the addition of 10 μL of CCK8 reagent (UElandy, China). The plates were then incubated for 1 hour, and absorbance was measured at 450 nm using a microplate reader to assess relative cell proliferation.

Six-well plates were used to seed MHCC-97H cells, which were then incubated for 24 hours in DMEM containing 1% FBS. All cells were maintained at 37°C in a humidified incubator with 5% CO₂. Each well was inoculated with 5×10⁵ cells, ensuring thorough mixing during the seeding process. When the cells reached 80% to 90% confluence, the culture medium was removed, and the floating cells were carefully washed twice with 1×PBS (Gibco, USA). Using a 10 μL pipette tip, three horizontal scratches were made across each well. The wells were subsequently rinsed with 1×PBS to eliminate any remaining floating cells. Once the background was clean, the cells were cultured in medium containing 1% FBS and treated with different substances, with 2 ml of medium added to each well. In the experimental group, the concentration of CsESPs was set to 50 μg/ml, while the control group received an equal volume of 1×PBS solution. Imaging was performed using a light microscope (ZEISS Axio Vert.A1, Germany) immediately after adding the drug-containing medium (0 hours), followed by additional observations and imaging at 24-hour intervals until 72 hours.

The migration and invasion capabilities of MHCC-97H cells through filters were assessed using Transwell chambers (Costar, USA). Briefly, MHCC-97H cells co-cultured with CsESPs or PBS were serum-starved for 24 hours. Afterward, 1×10⁶ cells/well in 200 μL of serum-free DMEM were placed in the upper chamber at 37°C. For the matrix invasion assay, Matrigel (Corning, USA) was added to the upper chamber in a ratio of 1:8 with serum before the cells were introduced., while for the migration assay, Matrigel was not included. 600 μL of DMEM containing 20% FBS was added to the lower chamber. After 48 hours of incubation, the cells that migrated to the bottom surface of the membrane were washed with PBS, fixed with formaldehyde for 30 minutes, stained with 0.5% crystal violet (Solarbio, China), and counted. The migration and invasion cells located at the bottom of each chamber were photographed and quantified using a microscope (ZEISS Axio Vert.A1, Germany). Cell counts were randomly performed in three high-power fields.

In vitro capillary network formation was determined by performing a tube formation assay in Matrigel (Corning, USA). Human umbilical vein endothelial cells (HUVECs) co-cultured with CsESPs or PBS were plated in triplicate on 50 uL Matrigel -coated 96-well plate, with 1.5 × 10⁴ cells per well in 100 μL of medium. After 8h of incubation, tube formation was examined by microscopy (ZEISS Axio Vert.A1, Germany), and the tube numbers and tube length were quantified by randomly selecting three fields per well using the lmageJ program. Higher values of these parameters indicate greater quantity and quality of tube formation.

GraphPad Prism 9.5.0 software was used for statistical analysis. Intergroup differences in categorical data presented as ratios were compared using the Chi-square test. Continuous variables were analyzed using the Student’s t-test and ANOVA. All experimental results are based on three independent experiments, presented as means ± standard deviation (SD), with p < 0.05 considered statistically significant.

To explore the tumor microenvironment and molecular features of Cs-infected HCC, an analytical workflow and experimental design was created ( Figure 1A ). Bulk RNA sequencing (RNA-seq) was conducted on samples collected from 10 Cs-positive HCC patients, comprising 10 cases of tumor tissue (Cs⁺ _T) and 9 cases of adjacent non-tumor tissue (Cs⁺ _P), with one adjacent non-tumor sample excluded due to unsatisfactory quality. Additionally, RNA-seq was conducted on samples collected from 10 Cs-negative HCC patients, including 10 pairs of tumor tissue (Cs ^-_T) and their corresponding adjacent non-tumor tissue (Cs ^-_P). All of which were obtained from treatment-naive HCC patients through surgical resection. Details for all patients are provided in Supplementary Table S2 . The gene expression levels of all samples are shown in Supplementary Table S3 .

To start with, we investigated the RNA expression levels of Cs⁺ HCC and identified the differentially expressed genes (DEGs) between tumors and their adjacent non-tumor tissues. Totally, 1,984 up-regulated genes and 2,010 down-regulated genes were found in Cs⁺ HCC tumors compared to adjacent non-tumor tissues ( Figure 1B ). Gene Ontology (GO) analysis revealed that up-regulated genes were enriched in biological processes such as nuclear division, cellular components like chromosomal region, and molecular functions such as ATP hydrolysis activity, etc ( Figure 1C ), and down-regulated genes were enriched in biological processes related to activation of immune response, cellular components such as collagen-containing extracellular matrix, and molecular functions like antigen binding, etc ( Figure 1D ). Additionally, the KEGG analysis revealed that up-regulated genes were significantly enriched in pathways such as cell cycle, DNA replication, and ECM-receptor interaction, etc ( Figure 1E ), and down-regulated genes were enriched in pathways including complement and coagulation cascades, chemokine signaling pathway, and cell adhesion molecules, etc ( Figure 1F ).

In addition, the DEGs between tumors and adjacent non-tumor tissues in Cs^- HCC were analyzed, and 1,804 up-regulated genes and 1,546 down-regulated genes were found in Cs^- HCC tumors compared to adjacent non-tumor tissues ( Figure 2A ). The GO analysis reveals that up-regulated genes are mainly enriched in nuclear division, chromosomal regions, and catalytic activity, acting on DNA, etc ( Figure 2B ), and down-regulated genes are primarily enriched in response to xenobiotic stimuli, collagen-containing extracellular matrix, and extracellular matrix structural constituents, etc ( Figure 2C ). The KEGG analysis revealed that up-regulated genes were significantly enriched in cell cycle, DNA replication, p53 signaling pathway, and motor proteins, etc ( Figure 2D ), and down-regulated genes were mainly enriched in cytokine-cytokine receptor interaction, complement and coagulation cascades and retinol metabolism, etc ( Figure 2E ). Finally, we also compared the up-regulated and down-regulated genes between tumors and adjacent tissues in Cs⁺ and Cs^- HCC. The results revealed a significant difference, suggesting that Cs infection remodels the gene expression landscape in HCC ( Figure 2F ).

To further explore the molecular differences between malignant cells in the two groups, the mRNA expression profiles of Cs⁺ HCC tumors and Cs^- HCC tumors were compared. 785 up-regulated genes and 675 down-regulated genes were found in Cs⁺ HCC tumors compared to Cs^- HCC tumors ( Figure 3A ). The GO analysis reveals that up-regulated genes are mainly enriched in extracellular matrix organization, collagen-containing extracellular matrix and extracellular matrix structural constituent, etc ( Figure 3B ), and down-regulated genes are primarily enriched in response to activation of immune response, external side of plasma membrane, and antigen binding, etc ( Figure 3C ). The KEGG analysis revealed that up-regulated genes were significantly enriched in Cytoskeleton in muscle cells, Focal Adhesion, ECM-receptor interaction, Focal adhesion, etc ( Figure 3D ), and down-regulated genes were mainly enriched in T cell receptor signaling pathway, natural killer cell mediated cytotoxicity, PD-L1 expression and PD-1 checkpoint pathway in cancer, etc ( Figure 3E ). This suggests that Cs infection may promote HCC progression involved ECM remodeling and immune responses pathway. Furthermore, to investigate the interactions among the DEGs, we constructed a PPI network ( Figures 3F, G ). 2034 and 2063 gene interactions were found in up- and down-regulated genes, respectively. Moreover, top 3 interaction genes were COL1A1, COL1A2 and COL3A1 in up-regulated genes, and top 3 interaction genes were CD2, CD4 and CD19 in down-regulated genes. Last but not least, motif analysis was performed on the regulatory regions of up-regulated and down-regulated genes. SOX17 and NFYC were enriched in the regulatory regions of up-regulated genes and HOXA11 and RORC were enriched in the regulatory regions of down-regulated genes, respectively ( Figure 3H ).

Considering the GO and KEGG results found that the DEGs between Cs⁺ HCC tumors and Cs^- HCC tumors were involved in tumor immune regulation, we next assessed the immune microenvironment in the Cs⁺ HCC tumors and Cs^- HCC tumors. The immune infiltration analysis showed that resting dendritic cells (DCs) were significantly more abundant in Cs⁺ HCC tumors compared to Cs^- HCC tumors (p < 0.05) ( Supplementary Figure S1A ). Additionally, the T-cell stimulant-related genes CD27, ICOS, and TNFRSF8 were significantly down-regulated in Cs⁺ HCC tumors ( Supplementary Figure S1B ). However, there were no significant differences in the HLA-related genes between Cs⁺ HCC tumors and Cs^- HCC tumors ( Supplementary Figure S1C ).

Increasing evidence suggests that the progression of HCC is closely linked to metabolic reprogramming (23, 24). Metabolic pathway events encompass the Warburg effect and an increased rate of oxidative phosphorylation (OXPHOS), which supply energy for the growth and invasion of cancer cells (25–27). These metabolic alterations not only support the rapid proliferation and survival of tumor cells but may also influence the aggressiveness of the tumor and its resistance to treatment (28). Therefore, we also compared the differential genes in this metabolic pathway between Cs⁺ tumors and Cs^- tumors. We found that genes associated with OXPHOS, such as complex I (MT-ND1, MT-ND2 NDUFA4L2 and NDUFS3), IV (COX7A2L), and V (ATP6V1E1 and ATP6V1H) of the electron transport chain were up-regulated in Cs⁺ tumors, while gene related to complex I (NDUFA8) was down-regulated in Cs⁺ tumors ( Supplementary Figure S2 ).

Given that lncRNAs play crucial regulatory roles in various biological events related to tumorigenesis and development, we also compared the differential gene expression of lncRNAs between Cs⁺ HCC tumors and Cs^- HCC tumors. 303 up-regulated genes and 286 down-regulated genes of lncRNA level were found in Cs⁺ HCC tumors compared to Cs^- HCC tumors ( Supplementary Figure S3 ).

To explore whether Cs infection also changed RNA expression profiles of adjacent non-tumor tissues, the relevant analyses were performed. 696 up-regulated and 663 down-regulated genes were found in Cs⁺ HCC adjacent non-tumor tissues compared to Cs^- HCC adjacent non-tumor tissues ( Supplementary Figure S4A ). The GO analysis reveals that up-regulated genes are mainly enriched in leukocyte migration, collagen-containing extracellular matrix, AGE-RAGE signaling pathway and receptor ligand activity, etc ( Supplementary Figure S4B ). Moreover, the KEGG analysis revealed that up-regulated genes were significantly enriched in Cytokine-cytokine receptor interaction, PI3K-Akt signaling pathway, and focal adhesion, etc ( Supplementary Figure S4C ).

Next, we screened the top 10 up-regulated and down-regulated genes in Cs ⁺ HCC tumor tissues compared to Cs ^- HCC tumor tissues based on our clinically collected cohort and further investigated the expression characteristics of these genes in the TCGA-LIHC dataset. The top 10 up-regulated genes are CLVS2, IBSP, GP2, CXCL5, LINC00348, SMYD1, BPIFB1, CLDN18, SULT1C3 and GABRG3. And the top 10 down-regulated genes are MYH4, GATA5, ODAM, POU6F2, ZIM2, GFRA3, IGKV3-11, IGLC7, COL2A1 and NPFFR1. As shown in Figures 4A, B , the grouped expression levels of the top 10 up-regulated and down-regulated genes show no significant differences across the three subtypes of LIHC. However, when analyzing the association between the signature scores of the top 10 up-regulated and down-regulated genes and overall survival (OS) in LIHC patients, we found a correlation between these scores and specific clinical outcomes in LIHC. High expression level of grouped top 10 up-regulated genes are associated with shorter OS in LIHC-TCGA cohort (p < 0.05) ( Figure 4C ). And low expression levels of the grouped top 10 down-regulated genes to be associated with shorter OS, although the differences are not statistically significant ( Figure 4D ). Consequently, our survival analysis results suggests that the DEGs between Cs⁺ HCC tumors and Cs^- HCC tumors have clinical relevance in LIHC-TCGA cohort.

The above analysis found that Cs⁺ HCC tumors are related to ECM-receptor interaction and focal adhesion ( Figure 3 ). It is well-known that ECM derived from tumor cells is crucial in cancer progression, as its remodeling through synthesis and degradation facilitates cell adhesion and the infiltration of endothelial and tumor cells, promoting angiogenesis and leading to metastasis (29). This suggests that Cs infection may promote HCC migration, invasion and angiogenesis. In order to confirm the above hypothesis, we conducted a retrospective analysis of 1,461 treatment-naive HCC patients who underwent surgical resection from January 2013 to December 2022, focusing on metastasis rates and MVI positivity rate. The results demonstrated that the metastasis rate in Cs ⁺ HCC patients was significantly higher than in Cs^- HCC patients (58.5% vs. 44.3%, p = 0.007) ( Figure 5A ). Likewise, the MVI positivity rate was also markedly higher in Cs ⁺ HCC patients compared to their Cs ^- counterparts (59.6% vs. 47.6%, p = 0.025) ( Figure 5B ). What’s more, we co-cultured Cs-derived excretory/secretory products (CsESPs) with MHCC-97H cells, and the control group was treated with PBS. It was found that the MHCC-97H cells treated with CsESPs had a greater migration ability than PBS ( Figure 5C ). The CsESPs co-culture group exhibited a significantly higher number of transmembrane cells, regardless of the presence of Matrigel, compared to the control group ( Figure 5D ). Finally, we further evaluated the effects of CsESPs on HUVECs tube formation. As expected, tube formation of HUVECs was significantly increased in the presence of CsESPs as demonstrated by the increase in total tube numbers and lengths ( Figure 5E ). Additionally, we investigated whether Cs affects tumor proliferation in HCC. The results showed no significant difference in the number of Ki67-positive cells between Cs ⁺ HCC patients and Cs ⁻ HCC patients ( Supplementary Figure S5A ). Furthermore, CsESPs had no significant proliferative effect on MHCC-97H cells ( Supplementary Figure S5B ). Collectively, our results suggest that Cs infection could promote HCC migration, invasion and angiogenesis.