3.1 Selection and synthesis of HBx-specific sgRNAs and in vitro efficacy: To generate HBx knockdown in HBV-positive HCC cell lines, we first analyzed sequences in the viral genome (HBx ORF) to identify potential sgRNAs adjacent to CRISPR associated protein 9 (Cas9) target sequences plus proto-spacer adjacent motifs. Initially, 69 potential sgRNAs were identified. Out of these, sgRNA that showed maximum specificity to X gene sequences and had minimum off-target mismatches in the human genome was selected for target sequence knockdown (Figure S1A and S1B). This region was synthesized and cloned in non-viral eSpCas9-2A-Puro (PX459)V2.0 vector (HBx-CRISPR) (Fig S1C). HBx-CRISPR plasmid was isolated, digested with restriction enzymes and checked on agarose gel (Fig. 1A and S2A). Next, we evaluated the efficacy of sustained Cas9/sgRNA expression in inhibiting HBx using a model that more reliably recapitulated HBV life cycle components. For these studies, we used the HepG2.2.15 hepatoblastoma cell line, which harbors both a functional HBV integrated form and cccDNA. After 48 h of co-transfection (HBx-CRISPR and pcDNA3-EGFP plasmid) in HepG2.2.15 cells, transfection efficiency was around 60% (Fig. 1A and S2B). An optimum concentration of the plasmid (1.5 µg), which resulted in low amount of plasmid usage as well as effective knockdown of HBx and less cell death in culture was used for transfection in further experiments (Fig. 1a and S2C-S2D). We excised the full length, 465 bp HBx gene using our designed HBx-CRISPR that resulted in a significant knockdown of the HBx gene expression in the HepG2-2.15 cells after 96 hours (18-fold decrease, p < 0.01 Fig S3A). We also validated our HBx-CRISPR plasmid in another cell system, where we co-transfected HepG2 cells with whole HBV plasmid and then with the HBx-CRISPR plasmid. We observed a 4-fold decrease in HBx gene expression in these double transfected cells (p < 0.05 Fig. 1b).
3.2 CRISPR/Cas9-mediated disruption of HBx decreases HBV replication in HBV-integrated-hepatoma models: We examined whether knockdown of HBx affected hepatitis B surface antigen. HBsAg ELISA assays indicated that HBx-CRISPR transfection significantly decreased the levels of HBsAg in the HepG2.2.15 cells supernatants after 96 hours (2-fold, p < 0.05, Fig. 1c and Fig S3B) and in HepG2-HBV model, (4.5-fold, p < 0.01, Fig. 1c). HBx knockdown also significantly reduced HBV cccDNA expression in the HepG2-2.15 cells after 96 hours (2-fold, p < 0.01, Fig. 1d and Fig S3B). Simultaneously, in HepG2-HBV model, results showed significant reduction in the expression of HBV cccDNA (5-fold, p < 0.05, Fig. 1d). The two variables i,e. HBx and HBV-cccDNA showed a similar line of expression pattern, whereas significant reduction was clearly seen at 96 hrs in the HBx-CRISPR cells (Fig S3B).
3.3 Knockdown of HBx inhibits cell proliferation, migration and invasion in HBx-CRISPR-transfected cells: The HBx knockdown-HepG2.2.15 cells showed a marked reduction in cell proliferation in comparison to cells transfected with the control vector (2-fold p < 0.05 Fig. 2a-b) in the scratch assays. Colony formation assays showed that HBx knockdown clearly inhibited the cell proliferation in HBx-CRISPR transfected HepG2.2.15 as compared to cells transfected with the control vector (1.72-fold, p < 0.05 Fig. 2c-d). We observed a significant decrease in chemotactic migration and invasion in the cells treated with HBx-CRISPR as compared to cells treated with control vector (migration: 1.43-fold, p < 0.05 Fig. 2e-f and invasion: 1.75-fold, p < 0.05 Fig. 2g-h) indicating a reduced tumorigenic properties in HBx knockdown cells.
3.4 Knockdown of HBx inhibits genes related to epithelial to mesenchymal transition and cancer stemness in HBV-integrated-hepatoma cells: We next studied the effects of HBx-CRISPR plasmid on the tumor attributes of hepatoma cells. We evaluated the expression of E-cadherin (CDH1) and Vimentin (VIM), both of which have been associated with EMT. The HBx knockdown in the HepG2.2.15 cells induced CDH1 protein expression in HBx-CRISPR transfected cells (Fig. 3a-b) and reduced VIM protein expression in HBx-CRISPR transfected cells compared to control-transfected HepG2.2.15 cells (Fig. 3d-e) as seen by the flow cytometry assays. Gene expression studies further validated the above observations i.e, significant reduction in VIM gene (2.14-fold p < 0.05, Fig. 3f) and an upregulation in CDH1 gene (1.6-fold p < 0.01, Fig. 3c) in the cells treated with HBx-CRISPR as compared to the control plasmid transfected cells. The expression of stemness gene, CD133 was also significantly reduced at protein (p < 0.01) as well as gene level (3.22-fold, p < 0.001) in HBx-CRISPR transfected cells compared to control-transfected HepG2.2.15 cells (Fig. 3g-i). EMT in hepatoma cells was further validated by studying the gene expression of mesenchymal genes. RT-PCR studies showed that after HBx knockdown, there was a significant decrease in the expression of mesenchymal genes, Thy1, CDH2, Fibronectin and alpha-SMA in HBx-CRISPR treated cells compared to the cells transfected with the control vector (CDH2; 6.5-fold p < 0.001, Thy-1; 1.84-fold p < 0.01, alpha-SMA; 1.49-fold p < 0.05 and fibronectin; 1.25-fold Fig. 3j-m). Furthermore, HBx knockdown conditions led to a significant decrease in the expression of β-catenin and TGF-β genes in HBx-CRISPR treated cells as compared to the cells transfected with the control vector (β-catenin; 5.07-fold p < 0.01 and TGF-β; 7.75-fold p < 0.001 Fig. 3n-o). Given a close relation between EMT with matrix metalloproteases (MMPs) and CSC properties of the tumor cells, we also evaluated the gene expression of some of the MMPs and CSC markers (CD24 and CD44) in HepG2.2.15 cells. When HepG2.2.15 cells were transfected with HBx-CRISPR, expression of CD24 and CD44 was significantly downregulated as compared to that observed in the control vector (CD24; 2.03-fold p < 0.05 and CD44; 2.05-fold p < 0.01 Fig. 3p-q). Knockdown of HBx also led to decrease in expression of MMPs (MMP2, MMP9 and MMP14) in HepG2.2.15 cells, which were transfected with HBx-CRISPR compared to the control vector (MMP2; 2.6-fold p < 0.01, MMP9; 4.35-fold p < 0.05 and MMP14; 1.8-fold p < 0.05 Fig. 3r-t).
3.5 HBx knockdown resulted in downregulation of genes significantly associated with cell proliferation and stemness: To identify the global effect of HBx knock-down on HBV-hepatoma cell transcriptomics, the RNA profile of HBx-CRISPR treated cells were compared with control-transfected cells. Out of 1159 differentially expressed genes (DEGs), 70 genes were upregulated while 1089 genes were downregulated in HBx-CRISPR transfected cells (p < 0.05) (Fig. 4a). Four DEGs; C1orf54, HLA-DQB1, RAC2, and PLAAT4 that were significantly downregulated in HBx-CRISPR have been earlier shown to be induced in the liver during hepatitis B (HBV) infection in chimpanzees (GSEA; M11620). ELK1, AKT3, and EGR2 that are known to be highly expressed in human HBV infection (Wikipathway; WP4666), were also significantly downregulated after HBx knockdown (Fig. 4b). Next, we performed a functional enrichment analysis of the whole genome repertoire of the significant DEGs. The downregulated genes were associated with the biological processes including cell population proliferation, regulation of cell differentiation, while less than 3 upregulated genes with FDR < 0.1 were linked to metabolic associated pathways regulation of amine transport and glutamate secretion (Fig. 4c). Pathway enrichment using Wikipathway (Fig. 4d) and Reactome (Fig S6A) showed that while the upregulated genes were not significantly involved in any pathway, the downregulated genes were enriched for cell proliferation- and stemness-related pathways including EGF, PI3K-AKT, NOTCH, WNT, Ras, and Hippo signaling pathways. In addition, genes associated with signaling by nucleus receptors and EMT were downregulated. (Further details of various genes, their log FC and p-value is in supplementary data as excel file.
3.6 Knockdown of HBx decreased HBV replication in Three-dimensional HBV-integrated-hepatoma models: Three dimensional (3D) models more closely resemble the in vivo tumor microenvironment by exhibiting complex phenotypic heterogeneity. We thus also analysed the functional effects of HBx-CRISPR on the HepG2.2.15 cells in 3D HBV-integrated-hepatoma tumor models. The expression of HBx was significantly reduced (2.57-fold p < 0.01) in HBx-CRISPR transfected cells compared to control-transfected HepG2.2.15 cells (Fig. 5a) in 3D cultures. In 3D tumor cultures, the expression of HBV cccDNA reduced significantly in HBx-CRISPR transfected HepG2.2.15 cells as compared to the control cells (1.6-fold, p < 0.01 Fig. 5b).
3.7 Knockdown of HBx reduced metastasis in 3D HBV-integrated-hepatoma in vitro models: We next probed, if transfection with HBx-CRISPR reduced the tumorigenicity in HBV-HCC 3D model. HBx-CRISPR-HepG2.2.15 and control hepatoma cells (8000 cells/9.5cm²) formed distinct spheroids within 14 days of suspension cultures (Fig. 5c). Cells transfected with HBx-CRISPR formed less number (p < 0.05 Fig. 5c and 5d) and smaller sizes (165.355µm w.r.t 251.866µm p < 0.001, Fig. 5c and 5e) of spheroids in culture as compared to cells transfected with control vector.
In 3D models, to assess the tumorigenic property, we also used non-adherent hanging drop cultures for spheroid formation where cellular aggregation is promoted based on gravity and there is complete absence of a substratum [15]. At day3, cell to cell adhesion and spheroid compaction were significantly reduced in HBx knockdown cell spheroids as compared to spheroids derived from control cells (Fig. 5f and 5g). Results showed that at day 3, the percentage of spheroid migration in HBx knockdown cells was 47.25% as compared to control vector cells which showed a percentage migration of 65.8% (Fig. 5f and 5h, p < 0.01).
3.8 Knockdown of HBx inhibit epithelial to mesenchymal transitions and reduce stemness potential in 3D HBV-integrated-hepatoma spheroids model: There was a significant increase in CDH1 expression in HBx-CRISPR transfected cells compared to control vector cells (CDH1; 8.28-fold p < 0.05 Fig. 6a) and significant decrease in the expression of mesenchymal genes, VIM, CDH2, Thy1 in HBx-CRISPR treated cells compared to the cells transfected with the control vector (VIM; 2.04-fold p < 0.05, CDH2; 2.72-fold p < 0.05, Thy-1; 2.34-fold p < 0.01 Fig. 6b-d) Next, we also evaluated the expression of MMP9 and CSC marker, CD133 in HepG2.2.15 cells spheroids. The expression of CD133 was significantly downregulated in HBx-CRISPR treated spheroid cells as compared to that observed in the control vector (CD133; 4.33-fold p < 0.01 Fig. 6e). In 3D cultures, we also observed that knockdown of HBx led to a decrease in expression of MMPs in HepG2.2.15 cells, which were transfected with HBx-CRISPR compared to the control vector (MMP9; 1.87-fold p < 0.05 Fig. 6f). Furthermore, HBx knockdown conditions led to a significant decrease in the expression of TGF-β and β-catenin genes in 3D HBx-CRISPR treated cells compared to the cells transfected with the control vector (TGF-β; 1.83-fold p < 0.05 and β-catenin; 5.12-fold p < 0.001 Fig. 6g-h).