Primary liver cancer is the sixth most common malignancy globally and ranks as the third leading cause of cancer-related deaths, posing a persistent public health burden with limited effective treatment options.1 Hepatocellular carcinoma (HCC), the predominant histologic subtype, accounts for approximately 90% of primary liver cancer,2 with hepatitis B virus (HBV) infection being a major etiological factor. The poor prognosis of HCC is largely due to its late diagnosis and high metastatic propensity for intrahepatic or extrahepatic metastasis, which often leads to the loss of the optimal window for surgical intervention.3 Therefore, identifying key regulatory molecules that influence the development of HCC is crucial.
Fibronectin (FN1), a multifunctional extracellular matrix (ECM) glycoprotein, is involved in embryogenesis, wound healing, and tissue remodeling.4 Although FN1 is generally oncogenic in most solid tumors by remodeling the ECM and activating integrin-driven pathways, its deficiency paradoxically impairs collagen barriers and facilitates cell detachment, thereby promoting metastasis.5–10 Thus, this duality of FN1 is not contradictory but rather context-dependent, varying with tumor stage, tissue microenvironment, and the balance between mechanotransduction and immunomodulation. However, HBV DNA integration into FN1 occurs preferentially in adjacent non-cancerous tissues rather than in tumor tissues, a finding that appears to contradict the established pro-oncogenic function of FN1.11 Accordingly, clarifying the role of FN1 in HCC is of paramount importance.
Plasmid construction: The sgRNA/Cas9 dual-expression vector pSpCas9(BB)-2A-GFP (PX458) was sourced from Addgene (Cambridge, MA). The sgP53/Pten dual-cassette plasmid was constructed as described previously.12 The sgFn1 dual-cassette plasmids were generated following the same strategy. For clarity, all plasmids are referred to throughout this study as follows: sgP53/Pten, sgFn1(1+4), sgFn1(1+3), sgFn1(2+4), sgFn1(2+3), and vector. The following sequences are four sgRNAs targeting the Fn1 gene and primer sequences (5′-3′): sgRNA1: AACCAGGGCGTTGCCTAGGT; sgRNA2: GGGTTTTAACTGCGAGAGCA; sgRNA3: TGATCTGGGACTGTACCTGC; sgRNA4: GGGGGTCAGTCCTACAAGAT; F: AGGTAGTTCCCCTGCCCATA; R: CTCGCAGCACATTTCACTGG.
Culture of cells: The origin and specific culture methods for Hepa1-6 cells and those for H2.35 cells were as described previously.12,13
Cell transfection and monoclonal screening: Hepa1-6 cells were seeded at a density of 2 × 105 cells per 3.5-cm culture dish. When cells confluence reached approximately 70%, transfection was performed using the sgFn1(1+4) plasmid. After visible colonies formed in 96-well plates, GFP-positive clones were selected under a fluorescence microscope and expanded for further culture. Successful Fn1 knockout (Fn1-KO) was confirmed by Western blot analysis.
Functional assays: CCK-8, colony formation, cell migration, cell invasion, and cell adhesion assays were carried out as previously described.14,15 The collagen used was Collagen Type I Rat Tail (Corning, 354236), at a concentration of 5 µg/cm2.
Animal experiments: C57BL/6J mice and HBV transgenic mice on a C57BL/6 genetic background were sourced from Peking University Health Science Center. All animals were housed under specific pathogen-free conditions in individually ventilated cages within the institution’s Department of Laboratory Animal Science. All experimental procedures were approved by the Institutional Animal Care and Use Committee of Peking University Health Science Center (LA2021492). All animal experiments were conducted in accordance with institutional guidelines for the care and use of laboratory animals. The procedure for hydrodynamics-based tail vein injection of plasmids and the post-injection care of mice were performed as described previously.12,16
Subcutaneous tumor implantation: A 100 µL PBS suspension containing 1 × 106 Hepa1-6 cells (wild-type (WT), Fn1-KO cell lines #5 and #13) was inoculated into each mouse’s axillary region. Each group had five mice. The mice were not pregrouped but were randomly assigned to three groups during inoculation. Once palpable tumors were observed, tumor volume was monitored every two days. Measurements were taken six times before the mice were sacrificed.
Western blot: The procedures for Western blotting were conducted as described previously.12 Antibody details are listed in Supplementary Table 1. Flow cytometric analysis of tumor tissues: The procedures for flow cytometry were conducted as described previously.13 Antibody details are listed in Supplementary Table 1. Analysis of immune cell infiltration and RNA-seq: The TCGA-LIHC dataset was acquired from the Genomic Data Commons portal (https://portal.gdc.cancer.gov ). Following normalization of the raw expression data, immune cell infiltration analysis was conducted using the SangerBox platform.17 RNA sequencing was performed following established protocols. Transcriptomic data were analyzed for differentially expressed genes, followed by Gene Ontology enrichment and Gene Set Enrichment Analysis. Subsequent signaling pathway analyses focused on subcutaneous tumors derived from Fn1-KO and WT Hepa1-6 cells.
Immunohistochemistry and multicolor immunohistochemistry: The procedures for immunohistochemistry and multicolor immunohistochemistry were conducted as described previously.13 Antibody details are listed in Supplementary Table 1.
Statistical analysis: All data in statistical graphs are presented as mean ± standard deviation, unless otherwise specified. Statistical analyses were performed using GraphPad Software (version 8.0) or R software (version 4.0.0). Unless otherwise indicated, an unpaired Student’s t-test or one-way ANOVA was used for group comparisons. Two-tailed P-values < 0.05 were considered statistically significant.
To clarify these conflicting observations, we designed four sgRNAs targeting the murine Fn1 gene and selected the sgFn1(1+4) dual-sgRNA combination for subsequent experiments (Supplementary Fig. 1A–E). Using monoclonal cell screening, we successfully established Fn1-KO Hepa1-6 cell strains (#5, #13, #18, and #19) (Fig. 1A). Functional experiments revealed that Fn1-KO cells showed markedly impaired proliferation, migration, invasion, and adhesion, with strains #5 and #13 exhibiting the most pronounced defects (Fig. 1B; Supplementary Fig. 1F–J). Notably, collagen supplementation failed to rescue the proliferation deficiency of Fn1-KO cells (Fig. 1C), although collagen alone could impair proliferation (Supplementary Fig. 1K). We then performed rescue experiments by overexpressing exogenous Fn1 in #5 and #13. CCK-8 proliferation assays showed that the proliferation rates were restored upon FN1 re-expression in both clones (Supplementary Fig. 1L–M). Moreover, xenograft tumors derived from Fn1-KO cells grew significantly more slowly in C57BL/6J mice than WT controls, with strain #5 displaying the strongest suppression (Fig. 1D–F). Further transcriptomic profiling (Supplementary Fig. 2A) and immunohistochemical analyses confirmed concomitant downregulation of types I and III collagen in Fn1-deficient tumors (Supplementary Fig. 2B), which might partly explain the observed functional impairments in these Fn1-KO cells.
Regarding the immunosuppressive microenvironment of HCC,18 bioinformatic interrogation of public datasets revealed that high FN1 expression in HCC correlated with increased infiltration of M2 macrophages and reduced recruitment of CD8+ T cells (Supplementary Fig. 2C–F). Consistent with this, flow cytometry demonstrated a decreased proportion of total and M2 macrophages (Fig. 1G; Supplementary Fig. 2G), along with a reduction in PD-1+ CD8+ T cells (Fig. 1H; Supplementary Fig. 2H), in the xenograft tumor tissues from the Fn1-KO group. In addition, Gene Set Enrichment Analysis identified significant activation of interferon-stimulated genes and interferon-β responses upon Fn1 deletion (Supplementary Fig. 2I).
To establish physiological relevance, we used HBV/S transgenic mice with P53/Pten deficiency.12Fn1-KO, accompanied by reduced ECM deposition, significantly attenuated hepatic tumor incidence and malignant progression arising from endogenous P53/Pten dual knockout (Fig. 1J–L; Supplementary Fig. 3A and B). Multicolor immunohistochemistry revealed a reduction in M2 macrophages and PD-1+ CD8+ T cells in the tumor tissues of the Fn1-KO group (Supplementary Fig. 3C). Intriguingly, despite the decreased incidence of primary tumors, the number of pulmonary metastatic tumors characterized by lower FN1 expression remained comparable between the P53/Pten/Fn1-KO group and the P53/Pten dual-KO group (Fig. 1M; Supplementary Fig. 3D).
As the organ with the highest FN1 expression, the liver presents a unique context for FN1 function.19 Our findings support a tumor-promoting role for FN1 in HBV-related HCC. Mechanistically, FN1 deficiency profoundly altered ECM composition in our models, with collagen reduction indicating microenvironmental remodeling. In addition, FN1 may facilitate HCC progression by increasing the infiltration of M2 macrophages and PD-1+ CD8+ T cells.
Although prior reports described FN1’s proliferative effects in vitro,3,20 its role in tumorigenesis remains undefined, particularly in HCC. Here, our results demonstrate that Fn1-KO significantly reduces hepatocarcinogenesis, supporting its tumor-promoting role. Mechanistically, FN1 promotes collagen deposition via the TGF-β/PI3K/Akt pathway,5,8,9 and mediates ECM-cytoskeleton signaling through integrin α5β1 to activate the FAK/PI3K/Akt proliferation axis.21,22 However, Fn1-KO permanently disrupts this integrin-mediated mechanotransduction, explaining why exogenous collagen supplementation failed to rescue the proliferative capacity of FN1-deficient cells.23,24 Paradoxically, we observed increased trend of pulmonary metastasis with diminished FN1 expression in metastatic cells. In line with a previous report that high FN1 expression can suppress metastasis,25 our in vitro study showed that FN1 deletion compromised adhesion and consequently facilitated lung metastasis of malignant tumor cells, possibly due to the unique alveolar architecture lacking vascular constraints. In clinical practice, lower expression of FN1 in primary HCC tissue indicates a higher risk of distant lung metastasis; therefore, closer monitoring with pulmonary imaging should be implemented to rule out early-stage lung metastasis in such patients.
In conclusion, this study demonstrates that FN1 deficiency suppresses hepatocyte malignant transformation and HCC progression but promotes pulmonary metastasis.
Supporting information
Supplementary Table 1
Information on the reagent.
(DOCX)
Supplementary Fig 1
Validation of Fn1 knockout and phenotypic characterization of Fn1-knockout HCC cells.
(A): Schematic representation of four sgRNAs target sites within the mouse Fn1 genome. (B-D): Validation of Fn1 knockout in Hepa1-6 cells transfected with plasmids expressing sgRNA combinations: sgFn1(1+3), sgFn1(1+4), sgFn1(2+3), sgFn1(2+4), sgFn1(3+4), or empty vector (PX458). Cells were harvested 48 hours post-transfection for genomic DNA and protein analysis. (B) PCR amplification of genomic DNA using Fn1-specific primers (F-R). (C) Western blot analysis of Fn1 expression in cells transfected with sgFn1(1+4), sgFn1(2+4), or PX458. (D) Sanger sequencing chromatogram of the PCR product. (E) Independent validation of Fn1 knockout in H2.35 cells transfected with sgFn1(1+4) via PCR, Sanger sequencing, and Western blot. (F) CCK-8 proliferation assay of four independent Fn1-KO Hepa1-6 cell clones. (G, H) Representative images of colony formation, migration, and invasion assays. (I) Quantitative analysis of colony formation ability.(J) Quantitative analysis of cell migration and invasion.(K) CCK-8 proliferation analysis of WT and Fn1-KO Hepa1-6 cells treated with collagen type I or PBS (n=4). (L, M) Fn1 overexpression in Fn1-knockout Hepa1-6 cells. (L) Western blot verification of exogenous Fn1 expression. (M) CCK-8 proliferation assays in wild-type (WT) Hepa1-6 cells, two independent Fn1-knockout clones (#5 and #13), and their corresponding Fn1-overexpressing rescue clones (#5 (Fn1-OE) and #13 (Fn1-OE) ), OE: over expression. Data are presented as mean ± SD. Statistical significance was determined by two-way ANOVA (F, K, M) or Student’s t-test (I, J).** P < 0.01, *** P < 0.001, **** P < 0.0001.
(DOCX)
Supplementary Fig 2
Characterization of immune cell populations in Fn1-knockout versus wild-type HCC.
(A) Gene Ontology (GO) enrichment analysis of Fn1-knockout versus wild-type tumor tissues. (B) Representative H&E and IHC staining of type I and type III collagen in subcutaneous tumor sections. (C-F) Immune infiltration profiling in the TCGA-LIHC cohort using four computational algorithms: (C) QUANTISEQ, (D) CIBERSORT, (E) TIMER, and (F) MCPcounter. (G-H) Flow cytometry analysis of immune cell populations in tumor tissues: (G) M1 macrophages, (H) CD8+ T cells and IFN-γ+ CD8+ T cells. (I) Gene Set Enrichment Analysis (GSEA) of tumor tissues based on FN1 expression. Data are presented as mean ± SD. Statistical significance was determined by Student’s t-test .* P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001 ns, no significance.
(DOCX)
Supplementary Fig 3
Representative image of a spontaneous HCC model in HBV transgenic mice.
(A) representative liver images. (B) Histological analysis of liver sections via H&E and IHC staining for FN1, α-SMA, type I, II, and III collagen. (C) Multi-color immunohistochemistry of liver tissue: left-PD-1+ CD8+ T cells (DAPI: blue; CD8: green; PD-1: yellow); right-M2 macrophages (DAPI: blue; F4/80: green; CD206: red). (D) Lung sections stained with H&E and FN1 IHC.
(DOCX)
Declarations
Ethical statement
All experimental procedures were approved by the Institutional Animal Care and Use Committee of Peking University Health Science Center (LA2021492). All animal experiments were conducted in accordance with institutional guidelines for the care and use of laboratory animals. All animals received human care. The use of public TCGA-LIHC data did not require additional ethical approval or informed consent because all data were publicly available and de-identified.
Data sharing statement
The RNA-seq data used to support the findings of this study are available from the corresponding author.
Funding
None to declare.
Conflict of interest
FML has been an Editorial Board Member of Journal of Clinical and Translational Hepatology since 2013. The other authors have no conflict of interests related to this publication.
Authors’ contributions
Conceptual design of the study (FML), data collection (MH, XL, JJG), statistical analysis, and writing of the original draft (MH, XL, FML). All authors reviewed and approved the final version and publication of the manuscript.