Ca2+/Calmodulin-Dependent Protein Kinase II Inhibits Hepatitis B Virus Replication from cccDNA via AMPK Activation and AKT/mTOR Suppression

Ca2+/calmodulin-dependent protein kinase II (CaMKII), which is involved in the calcium signaling pathway, is an important regulator of cancer cell proliferation, motility, growth, and metastasis. The effects of CaMKII on hepatitis B virus (HBV) replication have never been evaluated. Here, we found that phosphorylated, active CaMKII is reduced during HBV replication. Similar to other members of the AMPK/AKT/mTOR signaling pathway associated with HBV replication, CaMKII, which is associated with this pathway, was found to be a novel regulator of HBV replication. Overexpression of CaMKII reduced the expression of covalently closed circular DNA (cccDNA), HBV RNAs, and replicative intermediate (RI) DNAs while activating AMPK and inhibiting the AKT/mTOR signaling pathway. Findings in HBx-deficient mutant-transfected HepG2 cells showed that the CaMKII-mediated AMPK/AKT/mTOR signaling pathway was independent of HBx. Moreover, AMPK overexpression reduced HBV cccDNA, RNAs, and RI DNAs through CaMKII activation. Although AMPK acts downstream of CaMKII, AMPK overexpression altered CaMKII phosphorylation, suggesting that CaMKII and AMPK form a positive feedback loop. These results demonstrate that HBV replication suppresses CaMKII activity, and that CaMKII upregulation suppresses HBV replication from cccDNA via AMPK and the AKT/mTOR signaling pathway. Thus, activation or overexpression of CaMKII may be a new therapeutic target against HBV infection.


Introduction
Hepatitis B virus (HBV), a prototype virus of the Hepadnaviridae family, contains a small, partially double-stranded, relaxed 3.2 kb circular (RC) DNA [1]. Although an effective vaccine is available, HBV infection remains a public health concern. Chronic hepatitis B (CHB) infection, which has been estimated to affect 257 million individuals worldwide (WHO), is closely associated with cirrhosis and hepatocellular carcinoma (HCC) [2]. Upon HBV infection of a cell, its RC DNA is translocated to the nucleus and converted into a covalently closed circular DNA (cccDNA), which serves as the template for transcription

Transfection
Huh7 cells (1 × 10 6 ) were plated in 6 cm dishes and transfected 24 h after seeding with 6 µg of 1. To determine the effects of CaMKII α or AMPK α1 on 1.3 mer HBV X-null replication, 4 µg of 1.3 mer HBV WT (ayw) or HBx-deficient mutant constructs were co-transfected into 2 × 10 6 HepG2 cells with 4 µg of either HA-CaMKII α or Myc-AMPK α1. Culture medium was replaced by fresh medium 24 h after transfection, and the cells were lysed 72 h after transfection.

SDS-PAGE and Western Blotting
Cells were lysed in 0.2% NP-40 (IGEPAL)-TNE (10 mM Tris-HCl [pH 8.0], 50 mM NaCl, 1 mM EDTA) buffer with 1 × protease inhibitor cocktail (Calbiochem; San Diego, CA, USA; 535142) and 1 × phosphatase inhibitor cocktail (Sigma Aldrich; St. Louis, MO, USA; P0044). Tumor and non-tumor liver biopsy specimens from patients with HBV-associated HCC were obtained after surgery from Gangnam Severance Hospital, Yonsei University College of Medicine. All patients provided written informed consent, and the study protocol was approved by the institutional review board of Yonsei University College of Medicine (3-2019-0031). Paired tumor and non-tumor liver tissue specimens were chopped into small pieces and lysed in ice-cold M-PER mammalian protein extraction reagent (Thermo Fisher Scientific; Waltham, MA, USA; 78501) with 1 × protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific; Waltham, MA, USA; 78440). On the ice, the samples were homogenized with a Dounce homogenizer and centrifuged at 13,400× g for 15 min. The supernatants were collected, and their protein concentrations were determined using the Bradford assay [39]. Proteins were separated by sodium dodecyl sulfate-12% polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to PVDF membranes. The membranes were blocked with 4% skim milk, incubated with primary antibodies (1:1000) (Table S2), washed, and incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1:5000; Thermo Fisher Scientific; Waltham, MA, USA; 31460) or goat antimouse IgG (1:5000; KPL; New Delhi, India 474-1802) secondary antibodies. The bands were visualized using ECL (GE Healthcare Life Sciences; Piscataway, NJ, USA; RPN2106) and their relative intensities were measured using ImageJ 1.50b software.

cccDNA Extraction
HBV cccDNA was extracted using the Hirt protein-free DNA extraction procedure, as described previously [45]. Briefly, 3 × 10 6 HepG2, HepG2-hNTCP-C9, and HepG2-hNTCP-C9-HA-CaMKII α, or -Myc-AMPK α1 cells in collagen-coated 10 cm dishes were infected with HBV as described above. Seven days after infection, the cells were lysed with 0.6% SDS-TE buffer (10 mM Tris-HCl [pH 7.5], 10 mM EDTA) for 30 min at room temperature. The NaCl concentrations of the lysates were adjusted to 1 M with 5 M NaCl, and the lysates were incubated at 4 • C for 16 h, and centrifuged at 14,500× g for 30 min. The supernatants containing cccDNA were extracted twice with phenol and once with phenol-chloroform, followed by ethanol precipitation, with cccDNA quantified by Southern blotting. To further validate the authenticity of HBV cccDNA, the Hirt DNA sample was heated to 85 • C for 5 min. Among three forms of HBV DNA from the Hirt protein-free DNA extraction sample, including cccDNA (2.1 Kbp), DL DNA (3.2 Kbp), and protein-free RC DNA (above 3.2 Kbp), RC and DL DNAs were denatured but cccDNA was not. Electrophoretic mobility of cccDNA remains unchanged. Then, heat-treated DNA sample was further digested with EcoR I to linearize cccDNA to a genome-length double-stranded DNA (3.2 Kbp).

Quantitative Real-Time RT-PCR and PCR
To analyze mRNA levels of CaMKII and AMPK in HBV replicating cells, 1.3 mer HBV WT (ayw) was transiently transfected in HepG2 and Huh7 cells and total RNAs were isolated as described above. Total RNAs were adjusted to 1 µg using a spectrophotometer (Thermo Fisher Scientific; ND-ONEC-W) and was reverse transcribed to cDNA using Su-perScript™ III Reverse Transcriptase (Invitrogen; Waltham, MA, USA; 18080093) according to the manufacturer's instructions. Quantitative real-time PCR (Applied Biosystems by Thermo Fisher Scientific, QuantStudio 3 Real-Time PCR; A28131) was performed using CaMKII-or AMPK-specific primers (Table S1). The relative mRNA levels were quantified by normalization to actin (loading control).
HBV cccDNA in HBV-infected cell was analyzed by quantitative real-time PCR. Briefly, HBV was infected in HepG2-hNTCP-C9, and HepG2-hNTCP-C9-HA-CaMKII α, and/or -Myc-AMPK α1 cells and then HBV cccDNA was extracted as described above. The DNA amount was adjusted to 50 ng and quantitative real-time PCR was performed using cccDNA-specific primers (Table S1) as described above. The cccDNA levels were quantified by normalization to actin (loading control).

Statistical Analysis
Data were reported as the means ± SD from at least three independent experiments. Mean values were compared using Student's t-test for most experiments, and GraphPad Prism version 5 for luciferase reporter assay. p-values < 0.05 were considered statistically.

Active, Phosphorylated CaMKII Is Decreased in HBV Replicating Cells
To replicate, many viruses utilize host kinases that are involved in signaling pathways [47]. Because CaMKII activation was reported to inhibit HCC metastasis [28,29], CaMKII phosphorylation levels were analyzed in cells in which HBV was undergoing replication ( Figure 1). Transient transfection of 1.3 mer WT HBV (ayw subtype) into Huh7 or HepG2 cells resulted in lower levels of activated CaMKII, as shown by T286 phosphorylation relative to control cells transfected with empty vector ( Figure 1A,B, top panel, lane 1 vs. 2). The phosphorylation of CaMKII was also analyzed in HBV replicating stable HepAD38 and HepG2.2.15 cells, in which HBV was undergoing replication ( Figure 1C,D). Removal of TC from HepAD38 cells, which was shown to induce HBV pgRNA transcription [36,37], also reduced the amount of active, phosphorylated CaMKII ( Figure 1C, top panel, lane 1 vs. 2). Similarly, the amount of active, phosphorylated CaMKII was lower in HepG2.2.15 than in control HepG2 cells ( Figure 1D, top panel, lane 1 vs. 2). These results indicate that HBV replication inhibits CaMKII activation. Because the level of total CaMKII was generally increased in HBV replicating cells ( Figure 1A,C,D), phosphorylated CaMKII level normalized to total CaMKII level (p-CaMKII/total CaMKII) showed an even greater reduction in HBV replicating cells than in control cells (Figure 1 and all the following results).

CaMKII and AMPK Activities Tend to Be Lower in Tumor Tissues than Adjacent Non-Tumor Tissues from Patients with HBV-Associated HCC
Because the most significant risk factors for HCC development are infections with HBV and hepatitis C virus (HCV) [2], this study compared phosphorylated CaMKII levels in paired tumor and non-tumor liver biopsy specimens from twelve patients with HBVassociated HCC, who had serum HBV DNA concentrations of 6,400,000; 287,000; 939,000; 3990; 0; 48; 0; 85,800; 287; 41,200; 0; and 0 IU/mL, respectively. The clinical characteristics of HBV-associated HCC patients were presented in Table S3. Although HBV DNAs were not detected in the serum of four patients (Table S3) showed that, when HBc protein level was higher, phosphorylated CaMKII was lower in tumor tissue than in paired non-tumor tissue ( Figure 2, top panel, lanes 3-6 and 11-22). Interestingly, although HBc was undetectable and showed same level from patient No. 6 and No. 34, respectively, the level of phosphorylated CaMKII was lower in tumor than in paired non-tumor tissue ( Figure   The lysates were subjected to SDS-PAGE and immunoblotted with primary antibodies to p-CaMKII, CaMKII, HBc, and GAPDH. To detect the core particle, lysates were subjected to 1% native agarose gel electrophoresis (NAGE) and then incubated with anti-HBc antibody. HBV DNA synthesis was analyzed by Southern blotting. HBV replicative intermediate, partially double-stranded RC, and double-stranded linear DNAs are marked as HBV RI, RC, and DL, respectively. Relative expression was quantified by normalization to GAPDH (loading control) using ImageJ 1.50b software. The level of p-CaMKII was normalized to total CaMKII expression. * p < 0.05, ** p < 0.005 relative to control by Student's t-tests (n = 3). The bars represent means ± SD of three independent experiments.
Because activated AMPK restricts HBV replication via autophagic degradation [13], and CaMKII is an upstream regulator of AMPK [48,49], the level of phosphorylated AMPK was also determined ( Figure 2, panel 3, lanes 3, 4, 9-12, 15-20, 23 and 24). Similar to CaMKII phosphorylation, the level of active, phosphorylated AMPK at T172 was lower in seven biopsy specimens from patients Nos. 19, 10, 13, 20, 21, 26 and 34, indicating that HBV-associated HCCs tend to inhibit the activation of CaMKII and AMPK. Phosphorylation and expression of CaMKII and AMPK in patients with HBV-associated hepatocellular carcinoma. Paired tumor and non-tumor liver biopsy specimens from HBV-associated HCC patients were lysed in M-PER buffer with protease and phosphatase inhibitors [3]. The lysates were subjected to SDS-PAGE and immunoblotting with primary antibodies to HBc, p-CaMKII, CaMKII, p-AMPK, and AMPK. Levels of p-CaMKII and p-AMPK were normalized to total CaMKII and AMPK levels, respectively. N, non-tumor; T, tumor.

Effects of HBV Replication on CaMKII, AMPK, and AKT/mTOR Phosphorylation
HBx has been shown to increase mTOR/S6K1 phosphorylation, thereby enhancing the development of HBV-mediated HCC [50]. Considering our findings, that active, phosphorylated CaMKII was decreased in HBV replicating cells (Figure 1), and that phosphorylated CaMKII and AMPK levels tend to be lower in tumor than in non-tumor tissue of patients with HBV-mediated HCC (Figure 2), we examined the phosphorylation of mTOR, S6K1, and 4EBP1 in HBV replicating cells (Figure 3). Consistent with the reduction in active, phosphorylated CaMKII (Figures 1 and 3A-C, panels 1 and 2, and Figure S1) associated with HBV replication (bottom panels), the levels of both total and phosphorylated mTOR were increased ( Figure 3A-C, panels 8 and 9), as were the levels of total and phosphorylated S6K1 and 4EBP, which act downstream of mTOR, ( Figure 3A-C, panels 10-13), indicating that HBV replication enhances the mTOR/S6K1/4EBP1 signaling pathway. The levels of phosphorylated mTOR, S6K1, and 4EBP showed greater relative increases than their respective total protein levels in HBV replicating cells ( Figure 3A-C).
Since total CaMKII and AMPK protein levels were increased in HBV replicating cells, we hypothesized that expressions of CaMKII and AMPK might be increased through HBx because HBx regulates the transcription of several genes. Additionally, HBV replication itself may influence several gene expressions with various mechanisms [31]. In accordance with our hypothesis, real-time RT-PCR results revealed that mRNA levels of CaMKII and AMPK were significantly increased in HBV replicating cells ( Figure S2).  Figure 1, as were core particles and HBV DNA synthesis. The relative expression of each protein was quantified by normalization to GAPDH (loading control) using ImageJ 1.50b software. * p < 0.05, ** p < 0.005 relative to control by Student's t-tests (n = 3).

CaMKII Overexpression Inhibits the HBV Replication through AKT/mTOR Signaling
Because HBV replication was found to inhibit the activation of both CaMKII and AMPK (Figures 1 and 3  HepG2-hNTCP-C9 cells were mock-infected and incubated as above (lane 2). The phosphorylation and expression of the indicated proteins, core particle, and HBV DNA synthesis were analyzed as described in Figure 1. Relative expression was quantified by normalization to GAPDH (loading control) using ImageJ 1.50b software. ns, not significant; * p < 0.05, ** p < 0.005 relative to mocktransfected control by Student's t-tests (lane 1 vs. 2). # p < 0.05, ## p < 0.005 relative to corresponding control by Student's t-tests (lane 2 vs. 3). @ p < 0.05, @@ p < 0.005 relative to corresponding control by Student's t-tests (lane 3 vs. 4) (n = 3).
In accordance with above result and Figure 3, HBV transfection of Huh7 cells was found to reduce the levels of active, phosphorylated CaMKII and AMPK, and to increase the levels of active, phosphorylated AKT, mTOR, and S6K1, as well as inhibitory phosphorylated 4EBP1 ( Figure 4C, panels 2-14, lane 1 vs. 2, and Figure S3A). Transfection of Huh7 cells with CaMKII α and the 1.3 mer HBV WT reduced HBV DNA synthesis ( Figure 4C, bottom panel, lane 2 vs. 3). Accordingly, CaMKII and AMPK were activated; AKT, mTOR, and S6K1 were inhibited; and 4EBP1 was activated ( Figure 4C, panels 2-14, lane 2 vs. 3, and Figure S3A).
To verify the effect of CaMKII α on the HBV infection system, HepG2-hNTCP-C9 cells were established and CaMKII was overexpressed using a lentiviral vector. Briefly, CaMKIIoverexpressing HepG2-hNTCP-C9 stable cells were infected with 2 × 10 2 GEq of HBV ( Figure 4D), as described previously [31,33]. As expected, CaMKII and AMPK activities were reduced, and the AKT/mTOR signaling pathway increased in these HBV-infected cells ( Figure 4D, panels 3-15, lane 2 vs. 3, and Figure S3B). Upon CaMKII α overexpression, phosphorylation of AMPK was increased and phosphorylation of AKT/mTOR was decreased, resulting in the downregulation of HBV DNA synthesis ( Figure 4D, lane 3 vs. 4, and Figure S3B). Taken together, these results demonstrate that CaMKII α overexpression represses HBV replication by activating AMPK and inhibiting the AKT/mTOR signaling pathway, indicating that CaMKII α is an upstream protein in the AMPK/AKT/mTOR signaling pathway in HBV replicating cells.

AMPK Overexpression or Activation Inhibits HBV Replication
Since AMPK seems to act downstream of CaMKII ( Figure 4C,D) [48,49], the effect of AMPK overexpression was analyzed in HBV replicating cells. Similar to overexpression of CaMKII α in HBV replicating cells (Figure 4), AMPK α1 overexpression reduced HBV DNA synthesis in transiently co-transfected HepG2 and Huh7 cells ( Figure 5A Figure S3C). By contrast, HBc protein expression and core particle formation were not affected by AMPK overexpression ( Figure 5A-C, lane 2 vs. 3).
Because the AMPK activator metformin inhibits HBV replication [12] and AMPK overexpression inhibits HBV replication (  Figure S5, bottom panels) than on the levels of HBc protein and core particles ( Figure S5, panels 3 and 5). Taken together, these findings indicate that AMPK overexpression ( Figure 5) and activation ( Figure S5) have negative effects on HBV replication.  Figure 4D. The phosphorylation and expression of the indicated proteins, core particles, and HBV DNA synthesis were analyzed as described in Figure 1. Relative expression was quantified by normalization to GAPDH (loading control) using ImageJ 1.50b software. ns, not significant; * p < 0.05, ** p < 0.005 relative to mock-transfected control by Student's t-tests (lane 1 vs. 2). # p < 0.05, ## p < 0.005 relative to corresponding control by Student's t-tests (lane 2 vs. 3). @ p < 0.05, @@ p < 0.005 relative to corresponding control by Student's t-tests (lane 3 vs. 4) (n = 3).

Inhibition of CaMKII Enhances HBV Replication
Because overexpression of CaMKII upregulated AMPK activity and downregulated the AKT/mTOR signaling pathway in HBV replicating cells ( Figure 4C,D), the effects of the CaMKII inhibitor KN93 on the AMPK/AKT/mTOR signaling pathway were evaluated. However, because CaMKII activity is downregulated in HBV replicating cells (Figures 1, 3 and 4C,D), CaMKII was overexpressed before treatment with the CaMKII inhibitor. Although Huh7 cells transiently co-transfected with 1.3 mer HBV WT plus CaMKII showed increased AMPK activity, inhibition of the AKT/mTOR signaling pathway, and downregulation of HBV DNA synthesis (Figures 6 and S6, lane 2 vs. 3), treatment of these cells with 10 µM KN93 inhibited CaMKII and subsequent AMPK phosphorylation, while restoring the AKT/mTOR signaling pathway and increasing HBV DNA synthesis (Figures 6 and S6, lane 3 vs. 4). There was no significant cell viability change in 10µM KN93-treated Huh7 cells ( Figure S4C, top panel). The findings, that HBV replication inhibits CaMKII and AMPK activations and activates AKT/mTOR signaling, and that CaMKII overexpression inhibits HBV replication while inhibiting AKT/mTOR signaling, suggest that CaMKII may regulate the AMPK/AKT/mTOR signaling pathway in HBV replicating cells.

CaMKII and AMPK Form a Feedback Loop
Because metformin and AICAR repressed HBV replication ( Figure S5), the effect of metformin on the CaMKII/AMPK/AKT signaling pathway was evaluated in HBV replicating cells. Consistent with the above results, treatment of 1.3 mer HBV WT-transfected Huh7 cells with 2 mM metformin inhibited the AKT/mTOR signaling pathway by enhancing AMPK activation, while significantly reducing HBc protein expression, core particle formation, and HBV DNA synthesis ( Figures 7A and S7A, lane 2 vs. 3). 2mM metformin was not cytotoxic to Huh7 cell ( Figure S4C, middle panel). Although CaMKII acts upstream of AMPK ( Figure 4C,D) [48,49], CaMKII phosphorylation was also increased by metformin ( Figure 7A, top panel, lane 2 vs. 3). Because AMPK activation induces CaMKII activation ( Figure 7A), we hypothesized that CaMKII and AMKP may form a feedback loop, thereby affecting the AKT/mTOR signaling pathway during HBV replication. To further evaluate this hypothesis, Huh7 cells co-transfected with 1.3 mer HBV WT plus CaMKII ( Figures 7B and S7B, lanes 3 and 4) were treated with the AMPK inhibitor compound C (5 µM) (Figures 7B and S7B, lane 4), which reduced the activities of CaMKII and AMPK and enhanced AKT/mTOR activities, resulting in increased HBV DNA synthesis ( Figure 7B,  lane 3 vs. 4). HBc expression and core particle formation, however, were not changed significantly. Cell viability was not affected by 5µM compound C treatment to Huh7 cells ( Figure S4C, bottom panel). To further verify the hypothesis in AMPK-overexpressing cells, Huh7 cells co-transfected with 1.3 mer HBV WT plus AMPK were treated with compound C and the CaMKII/AMPK/AKT signaling pathway was evaluated (Figures 8A and S7C). AMPK overexpression was found to increase CaMKII activity in these cells, reducing HBV DNA synthesis by suppressing the AKT/mTOR signaling pathway (Figures 8A and S7C,  lane 2 vs. 3). The AMPK inhibitor also blocked CaMKII phosphorylation and activated the AKT/mTOR signaling pathway, increasing HBV DNA synthesis (Figures 8A and S7C,  lane 3 vs. 4).
Huh7 cells co-transfected with 1.3 mer HBV WT plus AMPK were treated with the CaMKII inhibitor KN93 (Figures 8B and S7D, lane 4) to examine the effect of KN93 on the CaMKII/AMPK/AKT/mTOR signaling pathway. AMPK-co-transfection was found to activate CaMKII/AMPK and suppress the AKT/mTOR signaling pathway, thereby reducing HBV DNA synthesis ( Figures 8B and S7D, lane 2 vs. 3). Similar to compound C, KN93 blocked CaMKII/AMPK activation, inhibited the AKT/mTOR signaling pathway, and increased HBV replication ( Figures 8B and S7D, lane 3 vs. 4). These results indicate that CaMKII and AMKP may form a positive feedback loop for HBV replication.  The phosphorylation and expression of the indicated proteins, core particles, and HBV DNA synthesis were analyzed as described in Figure 1. Relative expression was quantified by normalization to GAPDH (loading control) using ImageJ 1.50b software. ns, not significant; * p < 0.05, ** p < 0.005 relative to control by Student's t-tests (lane 1 vs. 2). # p < 0.05 relative to corresponding control by Student's t-tests (lane 2 vs. 3). @ p < 0.05, @@ p < 0.005 relative to corresponding control by Student's t-tests (lane 3 vs. 4) (n = 3). The bars represent means ± SD of three independent experiments.

CaMKII Overexpression Reduces HBV RNAs Due to the Decreased HBV cccDNA
Since HBV DNA synthesis was decreased in CaMKII-overexpressing cells (Figures 4, 6 and 7B), HBV transcription was evaluated in CaMKII-overexpressing HepG2 cells using luciferase reporter assay ( Figure 9A). Overexpression of CaMKII reduced transcriptional activities of enhancer II/core promoter (EnhII/Cp), preS1 promoter, preS2 promoter, and enhancer I/X promoter (EnhI/Xp) ( Figure 9A). Northern blotting showed that the levels of HBV pgRNA and subgenomic S mRNAs were significantly reduced in HepG2 cells co-transfected with 1.3 mer HBV WT plus CaMKII and in HBV-infected CAMKII-overexpressing HepG2-hNTCP-C9 cells ( Figure 9B, upper left and right). Subgenomic S mRNA was significantly reduced in CaMKII-transfected-HepG2.2.15 cells, but pgRNA level was not ( Figure 9B, lower left). These findings suggest that the downregulation of HBV transcription in CaMKIIoverexpressing cells reduced the levels of HBV RNAs. In addition, HBV cccDNA was significantly reduced in HBV-infected cells ( Figure 9C, lane 3 vs. 4, white arrowhead, and Figure S8A). Taken together, these results indicated that CaMKII overexpression downregulates HBV cccDNA, resulting in decreased syntheses of HBV RNAs and RI DNAs.  Figure 4D. Total RNA was harvested at 72 h after transfection. Northern blotting was performed as described [41]. (C) HBV cccDNA levels are decreased upon overexpression of CaMKII in HBV-infected cells. Cells were infected as described in Figure 4D. The white and black arrowheads indicate cccDNA and linearized cccDNA to a genome-length DL DNA, respectively. ns, not significant; * p < 0.05, ** p < 0.005 relative to control by Student's t-tests (n = 3). The bars represent means ± SD of three independent experiments.

AMPK Overexpression Reduces HBV RNAs Due to the Decreased HBV cccDNA
Since either overexpression or activation of AMPK had a negative effect on HBV DNA synthesis ( Figures 5, 7A, 8 and S5), HBV transcriptional activities in AMPK-overexpressing HepG2 cells were evaluated using luciferase reporter assay ( Figure 10A). Similar to overexpression of CaMKII ( Figure 9A), overexpression of AMPK reduced the transcriptional activities of enhancer II/core promoter (EnhII/Cp), preS1 promoter, preS2 promoter, and enhancer I/X promoter (EnhI/Xp) ( Figure 10A).
Northern blotting showed the levels of HBV pgRNA and subgenomic S mRNAs were reduced in HepG2 cells co-transfected with 1.3 mer HBV WT plus AMPK and in HBVinfected AMPK-overexpressing HepG2-hNTCP-C9 cells ( Figure 10B, upper left and right). Subgenomic S mRNA was significantly reduced in AMPK transfected-HepG2.2.15 cells, whereas pgRNA level was not ( Figure 10B, lower left). HBV cccDNA level was also significantly reduced in HBV-infected, AMPK-overexpressing HepG2-hNTCP-C9 cells, suggesting that HBV cccDNA was decreased by AMPK overexpression in HBV-infected cells ( Figure 10C, lane 3 vs. 4, white arrowhead, and Figure S8B). Similar to findings in CaMKIIoverexpressing cells, these results indicate that AMPK overexpression downregulates HBV cccDNA, reducing the synthesis of HBV RNAs and RI DNAs.  Figure 9B. (C) Decreased HBV cccDNA levels upon overexpression of AMPK in HBV-infected cells. ns, not significant; * p < 0.05, ** p < 0.005 relative to control by Student's t-tests (n = 3). The bars represent means ± SD of three independent experiments.

Reduced HBV Replication by Overexpression of CaMKII or AMPK Is HBx Independent
Because overexpression of CaMKII significantly reduced the transcriptional activity of enhancer I and HBx promoter ( Figure 9A, second panel), and because many HBVassociated signaling pathways depend on HBx [50], we assessed whether this reduction in HBV replication was affected by HBx. HBV replication was lower in HepG2 cells transiently transfected with HBx-deficient mutant HBV than with HBV WT [54] ( Figure 11A HBx has been shown to promote lipid accumulation and hepatic steatosis through a transcriptional activation and signaling pathway associated with metabolism [55,56]. Since HBx regulates intracellular ATP and the AMPK signaling pathway to facilitate persistent HBV replication [57], and sustains HCC cell survival [58], we evaluated whether repression of HBV replication through AMPK overexpression was dependent on HBx. HepG2 cells were therefore transiently co-transfected with AMPK α1 plus 1.3 mer HBV WT or HBxdeficient HBV mutant. As expected, HBV replication was suppressed in cells co-transfected with HBV WT plus AMPK α1 ( Figure 11B, panels 15

Discussion
Chronic HBV infection is one of the main risk factors for progression of HCC [9][10][11] and HCC is the most frequent type of primary liver cancer causing cancer-related deaths worldwide [9][10][11]. Several kinases have been found to play important roles in HCC progression [28,29] and to be associated with HBV replication [12][13][14][15]. Activated CaMKII suppresses the migration of liver cancer cells by increasing intracellular calcium, thus inhibiting metastasis [28,29]. Our finding, that CaMKII phosphorylation was decreased in HBV replicating cells (Figures 1, 3, 4C,D, 6-8 and 11), indicates that HBV replication could enhance HCC progression by suppressing CaMKII activity. Furthermore, the activities of CaMKII and AMPK tend to be lower in HBV-associated HCCs than in paired non-tumor tissues ( Figure 2).
Although HBV DNA synthesis was decreased in cells transfected with CaMKII ( Figures 4, 6, 7B and 11A) and AMPK ( Figures 5, 8 and 11B), HBc protein expression and core particle formation were not significantly decreased. The C-terminal domain (CTD) of HBc protein undergoes a dynamic phosphorylation-dephosphorylation cycle during HBV replication with overlapping consensus R-x-x-S/T, R-R-x-S/T, and S/T-P phosphorylation motifs [64,65]. The phosphorylated CTD of HBc contributes to its functions, including core particle stability, pgRNA encapsidation, minus-strand and plus-strand DNA synthesis, and RC DNA synthesis [66,67]. Hyperphosphorylation of the HBc CTD by kinases neutralizes its positive charge, reducing RNA binding capacity and resulting in empty capsids [68,69]. Therefore, we hypothesized that overexpression of CaMKII or AMPK might regulate the phosphorylation of HBc protein to induce the production of empty capsids, thus decreasing HBV DNA synthesis. The phosphorylation consensus motifs for CaMKII and AMPK are R-x-x-S/T [70,71] and B-Hyd-B-x-x-x-S/T-x-x-x-Hyd [72], respectively, where Hyd represent a hydrophobic, x any, and B a basic amino acid residue. CaMKII can participate in HBc CTD phosphorylation because HBc CTD has three R-x-x-S/T motifs. AMPK may participate in HBc CTD phosphorylation indirectly. Additional studies are required to confirm this hypothesis.
Our finding, that cccDNA was reduced in CaMKII-and AMPK-overexpressing HBV replicating cells (Figures 9C, 10C and S8), suggested that RNA transcription should also be decreased ( Figures 9A and 10A), resulting in reduced pgRNA and subgenomic S mRNA levels ( Figures 9B and 10B). However, the present study found that pgRNA levels were not significantly reduced in HepG2.2.15 cells transfected with CaMKII or AMPK ( Figures 9B and 10B, lower left). Given that HepG2.2.15 cells are HBV replicating stable cells [34,35], suggesting that HBV pgRNA levels may not be significantly reduced by CaMKII or AMPK overexpression.
Because a complete cure of HBV infection requires the elimination of cccDNA from the cell nucleus [3], understanding the molecular mechanisms responsible for cccDNA stability is necessary to completely eradicate HBV. Since CaMKII overexpression decreases cccDNA level (Figures 9C and S8A) and CaMKII α contains a nuclear localization signal (NLS) [73,74], further studies are needed to determine the intracellular localization of CaMKII α and its indirect association with cccDNA. We hypothesized that the decreased levels of HBV RNAs in cells overexpressing CaMKII ( Figure 9A,B) were due to decreased levels of cccDNA ( Figures 9C and S8A), the template for HBV RNA transcription [75]. Furthermore, CaMKII overexpression might affect the recruitment of host RNA polymerase II or transcriptional activators or repressors, suggesting the need to evaluate the transcriptional activities of cccDNA in CaMKII-overexpressing HBV replicating cells. Because AMPK overexpression decreases cccDNA ( Figures 10C and S8B); HBV infection induces the accumulation of ROS [76]; and increased ROS enhances the localization of AMPK to the nucleus [77], additional studies are needed to explore the cellular localization of AMPK and the recruitment of host RNA polymerase II or transcriptional activators or repressors in AMPK-overexpressing HBV replicating cells.
Since aged liver showed the increased persistent inflammation and activated NF-κB signaling, aging is one of the most significant risk factors for chronic liver diseases [78]. Accordingly, when aged people are infected with HBV or HCV, they are more prone to be chronically infected than young adults [79]. Since aged people have declined metabolism and functional capacities, such as decreased AMPK activity and impaired insulin sensitivity, AMPK activation may promote metabolic health and prevent age-related diseases [80]. Therefore, we hypothesize that AMPK overexpression or activation may have therapeutic effects in treating not only CHB but also metabolic diseases, including obesity and type 2 diabetes, suggesting multiple effects by modulation of AMPK signaling pathway. Further studies are needed to evaluate the relationship with HBV replication, age, and AMPK/AKT/mTOR signaling pathway.

Conclusions
The present study showed that HBV replication inhibits CaMKII activation and that CaMKII overexpression reduces HBV replication via the AMPK/AKT/mTOR signaling pathway. The findings, that overexpressed CaMKII or AMPK reduced cccDNA level, inhibiting the synthesis of HBV RNAs and RI DNA, suggested that activation or overexpression of CaMKII or AMPK might be a possible therapeutic option to cure HBV infection and HBV-associated hepatocarcinogenesis.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms10030498/s1, Figure S1: Relative levels of total and active CaMKII and AMPK in HBV replicating cells presented in Figure 3; Figure S2: Quantitative real-time RT-PCR results to show the levels of CaMKII and AMPK mRNAs in HBV replicating cells; Figure S3: Relative levels of total and active CaMKII and AMPK in HBV replicating cells presented in Figure 4C-D and Figure 5C-D; Figure S4: Cytotoxic effects of metformin, AICAR, KN93, and compound C in indicated cells; Figure S5: Activation of AMPK reduces HBV replication; Figure S6: AKT-mTOR/S6K1/4EBP1 signaling pathway presented in Figure 6; Figure S7: AKT-mTOR/S6K1/4EBP1 signaling pathway presented in Figure 7A Figure S18: Uncut scans of original western blotting images of Figure 11; Table S1: Primers for construction of CaMKII α and AMPK α1 expression plasmids, qPCR, and RT-qPCR; Table S2: Antibodies used for this study; Table S3: Clinical characteristics of HBV-associated HCC patients presented in Figure 2.