ATA was purchased from Sigma-Aldrich (St. Louis, MO, USA). ATA was dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich) as a 100-mM stock solution and stored at −20°C.

HepaRG cells purchased from Thermo Fisher (Waltham, MA, USA; catalog no. HPRGC10) were cultured in William’s medium E (Invitrogen, Carlsbad, CA, USA) supplemented with 2 mM of l-glutamine, 200 units/ml of penicillin, 200 μg/ml of streptomycin, 10% fetal bovine serum (FBS), 5 μg/ml of insulin, 20 ng/ml of epidermal growth factor, 50 μM of hydrocortisone (Sigma-Aldrich), and 2% DMSO. Huh-7 or HepG2.2.15 cells purchased from ATCC Cell Bank (Manassas, VA, USA) were cultured in Dulbecco’s modified Eagle medium with 10 mM of HEPES, 200 units/ml of penicillin, 200 μg/ml of streptomycin, and 10% FBS. Plasmids were transfected into Huh-7 or HepaRG cells with Fugene 6 (Roche, Shanghai, China) or Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions.

A 50-fold concentrated culture supernatant of HepG2.2.15 cells was used to prepare an infectious inoculum. The inoculum was prepared from freshly collected supernatants by precipitating viral particles in the presence of 6% polyethylene glycol 8000 (PEG_8k). The virion pellets were resuspended in phosphate-buffered saline containing 25% FBS. Differentiated HepaRG cells (with 1 × 10⁷ genome equivalent copies of HBV and 1 × 10⁵ HepaRG cells were inoculated) were incubated with the concentrated infectious inoculum as reported for primary human hepatocytes and HepaRG cells (19). The secreted HBeAg, HBV DNA, and RNA were determined on the indicated days after infection.

Native agarose gel electrophoresis (NAGE) assays were performed as described previously (20). In brief, intact nucleocapsid particles were separated from crude extracts of HBV-replicating cells using agarose gel electrophoresis. Nucleocapsid particles within the gel were then denatured under alkaline conditions and transferred onto nitrocellulose membranes (Roche). Nucleocapsid DNA and core proteins were detected using Southern and Western blotting with a double-stranded HBV DNA probe spanning the whole viral genome and an anti-core antibody, respectively.

The HBV replicon or RNase H-defective replicon plasmid was described previously (8). Probe labeling and the development of Southern blotting signals were performed using an AlkPhos direct labeling system (Amersham, Little Chalfont, UK).

Total RNA was isolated using TRIzol reagent (Invitrogen) and treated twice with RNase-free DNase I (20, 21). Total RNA (1 μg) was used for reverse transcription, and the viral transcripts were quantified as described previously (20, 21). The results were normalized against the amplification results for internal control [i.e., glyceraldehyde 3-phosphate dehydrogenase (GAPDH)]. The primers used to detect HBV RNA were forward 5′-CGGAAATATACATCGTTTCCAT-3′ and reverse 5′-AAGAGTCCTCTTATGTAAGACCTT-3′, and the GAPDH primers were forward 5′-TGCACCACCAACTGCTTAGC-3′ and reverse 5′-GGCATGGACTGTGGTCATGAG-3′. RcDNA purification from either secreted virion or cytoplasm nucleocapsids was performed as previously described (22). Total RNA was extracted from cells using TRIzol according to the manufacturer’s instructions. Purified RNA (1 μg) was treated with DNase I (Invitrogen), immediately primed with oligo(dT) or random primers, and reverse transcribed using Superscript III reverse transcriptase (Invitrogen). Real-time PCR was performed using the ΔΔCt method.

HBeAg ELISA was performed as previously described (23).

Statistical analysis was performed using Prism 6.0 (GraphPad software). Unpaired, two-tailed Student’s t-test was used for statistical comparison between groups, unless otherwise specifically mentioned.

To examine the possible inhibition of HBV replication by ATA, we first used HepaRG cells for HBV infection since these cells express high levels of HBV receptor NTCP (4), and we treated HBV-infected HepaRG cells with different concentrations of ATA and so measured the HBeAg content. In this experiment, HepaRG cells were infected with HBV virions for 14 days in the absence or the presence of ATA at 5 (low) or 10 (high) μM, and viral loads (secreted HBeAg) in the supernatant were measured by HBe ELISA at different time points. As a result (Figure 2A), ATA inhibited HBV replication in the HepaRG cells in a dose-dependent manner. ATA continuously suppressed the HBeAg content, and on Day 14, 5 and 10 μg/ml of ATA reduced the HBeAg level to 63.4% and 27.2% of that with no treatment, respectively. Therefore, ATA had an inhibitory effect on HBV replication in HepaRG cells.

We next explored ATA-mediated antiviral activity in HBV-transfected liver cells. Hepatocyte-derived cellular carcinoma Huh-7 cells were transfected with HBV replicon vectors for 24 h and further treated with or without ATA for 48 h before viral DNA in the supernatant containing the secreted virions was determined using qPCR (Figure 2B). The results showed that the treatment of ATA at 5 and 10 μM sustained the HBV DNA levels at 36.5% ± 4.2% and 23.4% ± 3.5% of that with no treatment, respectively, indicating a significant reduction of HBV production in ATA-treated hepatocytes (both p < 0.01). Moreover, in the cell lysates of different incubations as mentioned above, HBV nucleocapsids were estimated using NAGE assay, and total RNA was extracted to measure HBV transcripts using RT-qPCR. Levels of nucleocapsid-associated DNA (ncDNA) and cytoplasmic nucleocapsid core protein (i.e., HBcAg) were determined using HBV DNA probes and HBcAg antibody, respectively. In NAGE assays, the purified HBV particles were directly loaded onto a gel during electrophoresis, and thus ncDNA and HBV core (HBc) protein could be assessed using Southern and Western blotting, respectively, by HBV probes and an anti-HBV core protein antibody. As shown in Figure 2C, ATA treatments significantly lessened HBV DNA levels, whereas the HBcAg levels with or without ATA treatment were indistinguishable. This result revealed that the ATA had an inhibitory effect on viral DNAs but not on viral transcripts. This finding was confirmed when measuring HBV transcript levels using RT-qPCR. As shown in Figure 2D, HBV transcript from treatment with 5 or 10 μM of ATA displayed (1.03 ± 0.20)-fold or (0.92 ± 0.11)-fold of that with no treatment, showing no noticeable difference (p > 0.05). In addition, we also obtained the IC₅₀ values of Huh-7 cells treated with ATA at different periods, which read 39.8 and 31.6 mM at times 48 and 72 h, respectively (Figure 2E).

We further used HBV stably expressing hepatocytes (HepG2.2.15 cells) to reaffirm the above findings with regard to the inhibitory machinery of ATA on HBV replication. At first, we applied NAGE assays to determine ncDNA and HBcAg levels in the presence or absence of ATA, and the results are shown in Figure 2F. It was consistently observed that ATA had a dose-dependent inhibition on ncDNA synthesis but not on HBcAg expression in HepG2.2.15 cells. Moreover, HBV transcripts became (1.00 ± 0.14)-, (0.89 ± 0.08)-, and (0.95 ± 0.04)-fold in the presence of 2.5, 5, and 10 μM of ATA (Figure 2G), respectively, compared to that with no treatment (all p > 0.05), confirming that ATA treatment did not impair HBV transcripts.

Given the results insofar and the fact that ATA has been long regarded as a general inhibitor of nucleases, we postulated that ATA might inhibit HBV replication by blocking its Pol RNase H activity. To this end, a mutant HBV replicon plasmid was constructed where it expressed Pol protein with a point mutation D737V. This single amino acid replacement in the RNase H domain of the Pol protein causes the loss of its RNase H activity while maintaining DNA Pol activity, resulting in that HBV replication stalled at RNA/DNA hybrids in the nucleocapsids (8). We transfected Huh-7 cells with wild-type HBV or Pol RNase H dead mutant and further treated without or with ATA at different dosages. Results are shown in Figure 3A. While ATA had an inhibitory effect on wild-type HBV DNA levels but not on HBcAg levels, corroborating the results in Figure 1C, ATA treatment did not show any impact on HBV DNA levels in the RNase H dead mutant.

In parallel, total RNA extracted from cell lysates (wild-type HBV or D737V mutant transfected Huh-7 cells) without or with ATA treatment was measured for viral transcript levels (Figure 3B). In wild-type HBV transfected cells, treatment with 5 and 10 μM of ATA did not inhibit viral transcripts, as HBV transcript levels remain (0.98 ± 0.19)-fold (p > 0.05) and (0.96 ± 0.03)-fold (p > 0.05), respectively, compared to that without ATA treatment. In contrast, viral transcript in D737V HBV mutant transfected cells was reduced to (0.47 ± 0.03)-fold of that in wild-type HBV transfected counterpart (p < 0.001), verifying the halted viral replication in this mutant-transfected host. Nevertheless, 5 or 10 μM of ATA treatment did not deteriorate viral transcripts in D737V HBV mutant transfected cells, validating that ATA treatment did not impair HBV transcripts. Put together, no noticeable inhibitory effect on viral transcripts of wild-type or RNase H-defective mutant HBV was induced by ATA treatment. Therefore, ATA inhibits HBV replication by disrupting viral Pol RNase H activity and triggering HBV stalls at RNA/DNA hybrids, evincing a potential new reagent to treat HBV infection.

Since ATA showed a distinctive mechanism to inhibit HBV Pol RNase H activity, we next determined whether the lamivudine-resistant HBV could be inhibited by ATA. In the following experiments, we transfected YIDD, YVDD, and YMDD (wild-type) HBV into hepatocytes, and we treated the cells without or with ATA or lamivudine. The YMDD motif is highly conserved in the Pol of the retroviruses, where M552I and M552V mutations were found predominant in long-term lamivudine-treated HBV patients who developed drug resistance (24). Results are shown in Figure 4A. It demonstrated that while both ATA and lamivudine significantly inhibited wild-type HBV ncDNA levels, ATA (but not lamivudine) treatment substantially lowered viral DNA levels in YIDD and YVDD mutants. At the same time, ATA or lamivudine treatment did not alter the levels of HBcAg proteins. Furthermore, HBV transcripts were intact in the presence of ATA or lamivudine at the dosages applied (Figure 4B). Taken together, ATA could inhibit lamivudine-resistant HBV mutants, suggesting a potent reagent against HBV genotype mutations.