HepAD38 cells were maintained in the Dulbecco’s modified Eagle’s medium F12 (DMEM/F12) complemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 100 U/mL penicillin, 100 mg/mL streptomycin, and 100 μg/mL G418. HepG2-NTCP cells were maintained in DMEM with 10% FBS, 2 mM L-glutamine, 100 U/mL penicillin, 100 mg/mL streptomycin supplemented with 2 ug/mL puromycine (Gbico) and 2.5% DMSO. All cells were cultured in a humidified atmosphere with 5% CO₂ at 37°C.

Branched PEI (MW 25 kDa, 1 ug/uL) and Dimethyl Sulfoxide (D2650) were purchased from the (Sigma Aldrich, USA). The X-tremeGENE 9 DNA Transfection Reagent (06365779001) was purchased from Roche (Mannheim, Germany). One-step DNA Transfecter (LD210-005) was used (Shanghai ENLIGHTEN Bioscience & Technology Co. Ltd., Shanghai, China). Oligodeoxynucleotides, including the fluorescein isothiocyanate (FITC)-labeled oligonucleotides (with different nt-length listed in Table 1), were synthesized (Shanghai Biosune Co. Ltd., Shanghai, China) and were uniformly resolved in nuclease-free water (Cell Signaling Technology, USA) at an initial concentration of 100 uM. The Opti-MEM™ I Reduced Serum Medium (#31985062, Gibco, USA) was a modification of the Eagle’s minimum essential medium and was used as an ideal medium for cationic polymer transfections. The plasmid pCMV-EGFP (pEGFP-C2, 4,735 bp) encoding enhanced green fluorescent protein (EGFP) under cytomegalovirus (CMV) promoter was previously obtained from the BD Biosciences Clontech (USA). Plasmid pHBV1.3 (#AF100309.1) was a generous gift from Dr. Zhongliang Shen (Fudan University, China). Polyclonal rabbit anti-Hepatitis B virus core antigen (HBc) antibody (B0586) was purchased from the Dako (Glostrup, Denmark), polyclonal rabbit anti-HBsAg antibody (NB100-62652) from Novus Biological (Littleton, CO, USA) and β-actin (AF0003) was purchased from the Beyotime Biotechnology (Shanghai, China). Anti-Rabbit IgG (H + L) cross-adsorbed secondary antibody (A-11010) was purchased from Invitrogen. The CCK-8 kit (#40203ES76), MTT Cell Proliferation assay Kit (#40206ES76) and Hoechst 33258 (#40730ES03) were purchased from Shanghai YiSheng biotech Co. Ltd. (Shanghai, China).

The cells were evenly seeded in a 12-well plate at a density of 5 × 10⁵ cells per well and incubated for 24 h before transfection. Different mass concentrations of plasmid DNA or oligonucleotides (100 uM) were diluted in 120 uL of opti-MEM and then mixed with 2 uL of PEI (1 ug/uL) according to the different ratios for 15 min at room temperature. These transfection mixtures were added with 380 uL of serum-free DMEM in wells and incubated for duration of 5 h. After the performance of transfection, the transfection mixture was removed and 1 mL of complete medium with 10% FBS was added per well. The transfection effect was detected after a period of incubation of 48 h. Post transfection of PEI/ONs for duration of 5 h, HepG2-NTCP cells were inoculated with HBV at 1,000 genome equivalents (GE) per cell in a complete medium containing 2.5% DMSO, 4% PEG8000 for 12 h at 37°C and 5% CO₂. Cell culture supernatants were harvested at 0, 4, and 8 days post-inoculation (dpi) for the measurement of HBsAg and HBeAg levels, which served as markers of infection.

The experiments were conducted using male specific pathogen-free BALB/c mice aged 6–8 weeks. For evaluating the therapeutic effects of PEI/ON in HBsAg, were hydrodynamiclly injected with pHBV1.3 plasmid through tail-vein at day 0. The 6-weeks-old BALB/c male mice (n = 6/group) were first injected with 15 μg of pHBV1.3 plasmid DNA on their tail veins through hydrodynamic injection (HDI), followed by injection of 300 uL PEI/ON complex solutions (12 uL PEI: 6 uL ON) which were administered every 5 days for a total of three doses. The mice were randomly divided into four groups and subjected to the injections of PBS, PEI, PEI/10 nt-AC complex and PEI/40 nt-AC complex, respectively. Serial serum samples and tissues were collected at the indicated time points or after sacrifice.

An equal number of HepAD38 cells and HepG2-NTCP cells in 96-well plates were assayed after transfection. The cytotoxicity of PEI/oligonucleotide was measured by using the Cell Counting Kit-8 assay as per the manufacturer’s instructions (YiSheng biotech, Shanghai, China). Absorbance at 450 nm was measured using a multimode plate reader (Synergy 4, BioTek).

Culture supernatants and cell lysates were collected to detect the secretion levels and the intracellular levels of HBsAg and HBeAg, respectively. HBsAg and Hepatitis B virus e antigen (HBeAg) levels were measured quantitatively using the Abbott Architect immunoassay system (Abbott Laboratories).

HBV DNA was detected using the cell culture supernatant. Viral DNA analysis was then performed by using quantitative PCR (qPCR) kit and a 20 uL reaction volume (Tiangen Biotech, Beijing, China). Forward and reverse primers used were 5’-AATGCCCCTATCTTATCAACAC-3′ and 5′ -GAGATTGAGATCTTCTGCGACG-3′, respectively. Amplification was performed as follows: 95°C for 10 min, then 40 cycles of 95°C for 10 s, 55°C for 15 s, and 72°C for 20 s. Serial dilutions of a plasmid containing an HBV insert were used as quantification standards.

Total RNA purification was performed using the Trizol reagent according to the manufacturer’s instructions. 1ug of the total isolated RNA was subjected to treatment with 2.2 M formaldehyde, and subsequent electrophoresis using 1% agarose gel, after which bands were transferred onto a nylon membrane. The membrane-bound RNA was hybridized with a digoxigenin (DIG)-labeled HBV RNA probe specific for detection of genomic HBV RNA by using the DIG Northern starter kit (Roche Diagnostics), and the hybridization protocol was performed according to the manufacturer’s instructions. 1 μg of the purified RNA was subjected to electrophoresis on an agarose gel to detect 28S and 18S ribosomal RNAs (rRNAs) used as the loading control.

HBV core particles and HBV DNA replicative intermediates were extracted from the cytoplasmic fractions of cells and subjected to Southern blot analysis as per methods previously described (Lin et al., 2017). A DIG-labeled HBV RNA probe specific for detection of genomic HBV RNA was prepared using the DIG Northern starter kit (Roche Diagnostics), and the hybridization protocol was performed according to the manufacturer’s instructions. To calibrate the loading quantity for HBV replication, β-actin protein was determined by conducting sodium dodecylsulfate (SDS)-polyacrylamide gel electrophoresis and HBV core particles were resolved by performing native agarose gel electrophoresis using the same cellular lysates and were subjected to immunoblot analysis with antibodies to β-actin and HBcAg, respectively.

For the production of HBV particles, HepAD38cells were cultured as per methods described above (Cells and reagents, Cell culture). Cells were maintained in the DMEM/F12 medium supplemented with 5% fetal bovine serum and 2% DMSO. HBV particles were concentrated from the purified HepAD38 supernatant by conducting overnight precipitation with 10% PEG 8000 and centrifugation at 4°C for 1 h at 16,000 g. The pellet was resuspended in the Opti-MEM medium supplemented with 2.5% DMSO at a concentration of 10¹⁰ GE per milliliter. HBV inoculum concentration was quantified using real-time PCR by following the protocol of HBV DNA detection.

AD38 cells were seeded at a density of 1 × 10⁶ cells/well on circular glass coverslips placed in a 12-well plate and were grown overnight to obtain 70% confluence before subjection to transfection with FITC-labeled oligonucleotide or pEGPF-C2 for 5 h. For determination of transfection efficiency, green fluorescent protein (GFP) expression from the transfected plasmid pEGFP-C2 was monitored in living cells in culture using the Olympus CXC41 fluorescence microscope at a magnification of 100X. Cells were photographed at 48 h after transfection using bright-field and fluorescence microscopy.

For staining with anti-HBcAg antibodies after HBV infection, cells harvested at 8 days post injection were subjected to washing steps using PBS three times and fixation in 4% paraformaldehyde for 10 min at room temperature; thereafter, the cells were permeabilized with 0.2% Triton X-100 in PBS for 30 min. Detection of intracellular HBcAg levels was achieved after incubation with the anti-HBcAg antibody at 1:500 dilution overnight at 4°C. Coverslips were then washed twice with PBS and incubated with the secondary antibody for 1 h. Then coverslips were stained with 4′,6-diamidino-2-phenylindole (DAPI) and mounted on clean glass slides with a fluorescence-free glycerol-based medium and examined using a fluorescence microscope.

Liver sections were embedded in paraffin and were then subjected to staining with anti-HBsAg or hematoxylin and eosin (H&E) to visualize the pattern of HBsAg expression and the inflammation status, respectively. The histological features of the tissues were observed and imaged using a light microscope (ECLIPSE 80i. Nikon Tokyo, Japan).

All data have been presented as mean ± standard deviation (SD), based on data calculated from at least three independent repeats. Statistical significance was examined using the Student’s t-test. A p-value <0.05 was considered statistically significant for all analyses. The GraphPad Prism 6 software was used for plotting data and for statistical analysis.

The previous study demonstrated that NAPs with a length of 40 nucleotides (poly-AC repeat-based sequence, 40 nt-AC) conferred the best antiviral effect against HBV (Guillot et al., 2017). Additionally, the heteropolymeric adenosine-cytosine sequences are appropriate because of their lack of secondary structure, antisense activity, or immunostimulatory activity (Guillot et al., 2017). To investigate the antiviral effect of the oligonucleotide (40 nt-AC) transfected with PEI, we used the human hepatoma cell line HepAD38 which harbored a replicating HBV genome for investigation. The complexation of PEI with 40 nt-AC was first detected by agarose gel electrophoresis (Figure 1A). This gel retardation assay indicated that the positive charge on PEI could neutralize the negative charge of the oligonucleotide and could thereby hinder the migration of oligonucleotides in the gel. Complete complexations of PEI and oligonucleotides at different ratios were successfully achieved in our experiment.

Besides, as shown in Figure 1B, via fluorescence microscopy, the FITC-labeled oligonucleotide (40 nt-AC) could be efficiently transfected with PEI into the AD38 cells. Furthermore, the cytotoxicity of PEI/ON complexes was determined to be non-toxic using the CCK-8 assay (Figure 1C) and MTT assay (Supplementary Figure 1) by comparing the cytoxicity of the complexes with the control, i.e., the PEI without an oligonucleotide.

The PEI/ON complex significantly reduced the HBsAg secretion levels in the supernatant by approximately 59% of the control level in a dose-independent manner. Approximately 2 uL of PEI (1 ug/uL) transfected with 1 uL of 100 uM of 40 nt-AC yielded the best inhibitory effect on the HBsAg level (Figure 1D). However, evident changes were not observed in the extracellular HBeAg level (Figure 1E) and the viral DNA (Figure 1F) and in the accumulation of intracellular HBsAg and HBeAg (Figure 1G). Additionally, HBV replication intermediates and mRNAs were further examined using Southern blot and Northern blot analyses, respectively. As illustrated in Figure 1H, we ruled out the possibility that PEI/ONs had an influence on the levels of 3.2 kb relaxed circular DNA (rcDNA), and 3.5 kb and 2.4 kb/2.1 kb transcripts of HBV. Transfection of PEI/ON selectively reduced the HBsAg level but did not reduce the HBeAg level and the DNA level. These findings revealed a selective inhibition by the PEI/ON complex in terms of the secretion levels of HBsAg.

To exclude the possibility that free PEI or oligonucleotides alone in cells could influence the activities of HBV, HepAD38 cells were subjected to separate treatments with PEI or oligonucleotides alone in a dose-dependent manner. Results showed that there were no apparent changes in the levels of HBsAg and HBeAg in the culture supernatants subjected to treatment with either PEI (Figure 2A) or oligonucleotides (Figure 2B). These data emphasized the importance of the formation of the complex of PEI and oligonucleotide for the exhibition of inhibited activity.

Considering that the delivery of short oligonucleotides using PEI inhibited HBsAg secretion in HepAD38 cells (Figure 1), we aimed to explore whether PEI/plasmid-DNA complexes might also exert its antiviral activity in HBV. To test this hypothesis, the plasmid pEGFP-C2 (4.7 kilo base pair-long) that encoded an enhanced green fluorescent protein (EGFP) under the influence of the cytomegalovirus promoter was transfected with PEI into the HepAD38 cells in a dose-dependent manner (Figure 2C). The gel retardation assay was used to comfirm complexation of PEI with pEGFP-C2 at different weight ratios as depicted in Figure 2D. Importantly, the levels of secreted HBsAg and HBeAg (Figure 2E) and HBV-DNA (Figure 2F) were not affected by the PEI/plasmid-DNA transfection. This suggested that the formation of the PEI/ON complex might demonstrate different physicochemical properties that play a specific role in its inhibited activity.

To assess the oligodeoxynucleotide size effects on inhibiting activity, four types of AC-repeated oligonucleotides with different lengths (10, 20, 30, and 40 nucleotides) were transfected using PEI into the HepAD38 cells. The results showed that the suppression of the secreted HBsAg was lost in the 10-nt oligonucleotide (Figure 3A), while the secretion of HBsAg decreased gradually in the 20-nt oligonucleotide in a dose-dependent manner (Figure 3B). Interestingly, the treatment of the 30-nt and the 40-nt oligonucleotides with PEI resulted in a strong inhibitory activity, with a 53% (Figure 3C) and a 62% reduction (Figure 3D) in the HBsAg levels in a dose-independent manner. Meanwhile, these four nucleotides with PEI exerted no effects on the HBeAg level as shown in Figure 3. These results indicated that a length of 40 nucleotides was optimal for the efficient inhibition of HBsAg secretion by PEI/ONs.

To detect whether other commercial transfection reagents could replace PEI to deliver the oligonucleotides that would result in an inhibited activity in HBsAg level, we confirmed that the successful internalization of the FITC-oligonucleotide (40 nt-length, poly AC repeat-based) were loaded with PEI (Figure 4A), OneStep-transfecter (Figure 4B) and X-tremeGENE (Figure 4C) in HepAD38 cells, respectively. The cell culture supernatant was collected for detection. Interestingly, results indicated that only PEI that undergo complexation with oligonucleotides could reduce the levels of secreted HBsAg without a decline in the HBeAg level (Figure 4D). However, oligonucleotides that were harbored by the other two commercial transfection reagents exerted no effects on HBsAg or HBeAg levels as shown in Figure 4E (OneStep-transfecter) and Figure 4F (X-tremeGENE). We further performed transfection of different length oligonucleotides (10 nt, 20 nt, 30 nt and 40 nt) with two commercial kits (OneStep-transfecter and X-tremeGENE) in HepAD38 cells. As shown in Supplementary Figure 2, no apparent changes were observed in the levels of HBsAg and HBeAg. These results demonstrated that PEI and oligonucleotides at a certain length were indispensable to the formation of functional complexes that inhibited HBsAg secretion.

The decoy oligonucleotide (decoy-ON, dON) is short, double-stranded DNA molecules that interfered with gene expression at the transcription level through the simulation of a binding ligand with their target transcriptional factor (Billioud et al., 2016). We also explored whether PEI complexes with double-stranded oligonucleotides could also inhibit HBsAg secretion. To this end, we employed a scramble decoy-oligonucleotide, which formed complexes with PEI to transfect HepAD38 cells in a dose-dependent manner. Similar to the results obtained for single-stranded oligonucleotides, PEI complexation with double-stranded oligonucleotides reduced HBsAg levels (Figure 5A). The levels of HBeAg (Figure 5A, right panel) and HBV DNA (Figure 5B) also demonstrated no alterations. Concurrently, Southern blot and Northern blot results showed that the expected changes in the levels of both viral rcDNA and mRNAs did not decrease upon PEI transfection of HepAD38 cell (Figure 5C).

To ascertain whether PEI/decoy-ON inhibited HBV gene expression and replication, five decoy-oligonucleotides for the Specificity protein 1 (Sp1), (Raney et al., 1992) hepatocyte nuclear factor 1α (HNF1α), (Raney et al., 1991) and C-AMP-response element binding protein (CREB), (Kim et al., 2008) which were key regulators for HBV replication and transfection, and scramble as a control group were included. All PEI/decoy-ONs transfected could reduce HBsAg levels in the supernatant, but the levels of HBeAg demonstrated no changes (Figure 5D). We further determined the effect of PEI/decoy-ON transfection on HBV DNA replication by performing Southern blot hybridization for the isolated DNA samples derived from core particles and ascertained HBV RNA levels by performing Northern blot detection. Treatment of cells with a set of decoy-ONs for HBV relative transcriptional factor did not reduce HBV replication and transcription compared to scramble-decoy-oligonucleotide (Figure 5E). In summary, decoy-ON as double strand oligonucleotides complexing with PEI showed remarkable reduction of HBsAg levels in a sequence-independent manner, which was similar to the PEI/ON complexes.

For establishing HBV infection, the supernatant from HepAD38 cells was used as an inoculum. To assess the possible effect of the complex on the natural HBV infection, HepG2-NTCP cells were transfected with PEI/ON complexes in different ratios for 5 h. They were then inoculated with HBV at 1000 GE per cell for 12 h. HBV infection was assessed by measuring the HBsAg and HBeAg levels in supernatants at 4 and 8 dpi (Figure 6A). As shown in Figure 6B, PEI/ON transfection prior to HBV infection resulted in significant reductions of HBsAg and HBeAg secretion in a dose-dependent manner. At day 8 post-inoculation, cells were harvested and analyzed for the presence of HBcAg using fluorescence microscopy (Figure 6C). A dose-dependent inhibition of the HBV infection was observed under different concentrations of oligonucleotides with PEI, thus demonstrating that PEI/ON complexes were indeed active during viral infection (Figures 6B,C). Additionally, the count of HBcAg (+) cells (HBV-infected cells) enabled a more accurate assessment of the inhibition of PEI/ONs on viral entry in our experiments (Figure 6C, bottom). As shown in Supplementary Figure 3, PEI/ON transfection did not exert obvious cytotoxic effect by comparing with the blank group.

To further evaluate the anti-HBsAg efficacy of PEI/ON in vivo, we utilized BALB/c mice (n = 6/group) that were hydrodynamically injected with pHBV1.3 (15 ug/mouse) as the testing model. The experimental design is illustrated in Figure 7A. Approximately 300 uL of the PEI/ON complex solutions were injected through the tail vein and were administered every five days for a total of three doses after performing the pHBV1.3 injection at day 0. Sera were collected at the time points indicated for the analysis of HBsAg, HBeAg, anti-HBsAg, and anti-HBeAg levels. The results revealed that serum HBsAg and HBeAg levels were both significantly reduced by the injection of PEI/40 nt-AC, compared with PBS, PEI/PBS, and PEI/10 nt-AC groups at day 42 (Figure 7B). A total of five mice in the PEI/40 nt-AC group (n = 6) and 1 mouse in the PBS group (n = 6) presented with HBsAg and HBeAg losses in serum at day 42 (Figure 7B, right panels). However, only 1 mouse in the PEI/40 nt-AC group (n = 6) was anti-HBsAg-positive (Figure 7C, left panel), while anti-HBeAg-positive serum at day 42 (Figure 7C, right panel) was not detected in any mouse. Furthermore, the serum levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were measured to determine the degree of liver damage. As shown in Figure 7D, transient ALT elevations peaked at days 21 and 28 but reverted to normal levels at day 42. Additionally, liver tissues were obtained from each group of mice who were sacrificed at day 42 and subjected to immunohistochemistry and hematoxylin and eosin (HE) staining to detect HBsAg levels and liver damage. The results demonstrated that HBsAg-positive cells were not detected in mice injected with PEI/40 nt-AC (Figure 7E). There were no remarkable hepatocyte cellular changes observed in the liver section of the four groups (Figure 7F; Supplementary Figure 4). To further confirmed the specificity of the inhibition of PEI/ON complex in secreted HBsAg level, plasmid pNL-sNLuc which can express secreted form of Nano-Luciferase (sNLuc) was employed and delivered into BALB/c mice via hydrodianic injection (HDI). As shown in Supplementary Figure 5, the sNLuc level in mouse sera sample did not impaired by the transfection of PEI/40 nt-AC complex, which was different with the observation in pHBV1.3-mice and confirmed that PEI/40 nt-AC transfection had no impact on plasmid expression in mouse model and exerted its specific inhibiting in HBsAg secretion.

In summary, these data suggested that PEI/40 nt-AC induced HBsAg level decline and promoted HBV clearance independent of the anti-HBsAg development. However, these changes were associated with a transient elevation of ALT, which reflected liver inflammation.