Introduction

Hepatitis B virus (HBV) is a virus that infects human hepatocytes and causes acute or chronic infection of the liver (chronic hepatitis B virus infection, or CHB). Continuous replication of HBV in CHB results in progressing liver damage, leading to liver cirrhosis or hepatocellular carcinoma1. HBV has a complicated viral life cycle, starting from infection of human hepatocytes by HBV virions containing the partially double-stranded, relaxed circular DNA (rcDNA) viral genome. This is followed by nuclear import of rcDNA with the assistance of HBV core protein2 and its conversion into covalently closed circular DNA (cccDNA), a highly persistent genomic form that transcribes all viral RNA, including pregenomic RNA (pgRNA)3. HBV RNA is translated into viral proteins (X, core, surface proteins, and E-antigen), whereas pgRNA is reverse-transcribed by viral polymerase to produce rcDNA3. HBV cccDNA is not only very stable (persisting through the lifespan of resting hepatocytes; estimated half-life < 40 days to >9–26 months in patients undergoing nucleot(s)ide analog therapy)4–6, but is also replenished via intracellular amplification (re-import of generated rcDNA back to the nucleus)7–9 and re-infection10. Destroying HBV cccDNA or promoting its decay in hepatocytes can lead to a cure of HBV infection7,11. However, no medications are approved to target HBV cccDNA. While nucleot(s)ide analogs, interferons, and Bulevirtide (inhibitor of HBV entry) suppress HBV replication and reduce the risks of CHB outcomes, they do not cure the disease. Recent findings demonstrate that combining Bulevirtide with anti-HBV antisense oligonucleotides may result in sustained silencing of HBV cccDNA transcriptional activity12. HBV cccDNA can also be directly cleaved and destroyed by site-specific CRISPR/Cas9 nucleases7,13–16. Notably, some factors of innate immunity can limit HBV replication, while others, such as APOBEC/AID cytidine deaminases, can induce HBV cccDNA mutational inactivation and/or degradation17.
In humans, the APOBEC/AID includes 10 members: APOBEC1, APOBEC2, APOBEC3A (A3A), APOBEC3B (A3B), APOBEC3C (A3C), APOBEC3D (A3D), APOBEC3F (A3F), APOBEC3G (A3G), APOBEC3H (A3H), and AID18. A3A, A3B, A3C, A3D, A3F, A3G, A3H, and AID can directly deaminate cytidines in single-stranded DNA, introducing C to T and G to A mutations on the complementary strand; while APOBEC1, A3A, and A3G deaminate RNA, forming C to U mutations19. Recent findings indicate that elevated (but not endogenous) levels of A3B can also edit RNA at specific hotspots, causing lethality in a model of inducible A3B expression in mice20. APOBEC2 is unable to edit nucleic acids.
APOBEC/AID are famous mutators, implicated in the development of numerous cancers21, and important factors that restrict foreign RNA and DNA, such as viral nucleic acids or intracellular DNA leaks, e.g., from nucleus or mitochondria19,22. APOBEC/AID suppress replication of many viruses infecting humans, like human immunodeficiency virus, herpes simplex virus 1, Epstein–Barr virus, hepatitis C virus, human papilloma virus, and HBV19,23. Pioneering studies demonstrated HBV-editing activity of A3G, A3C, A3H, and A3F24–26. A3G was also shown to inhibit HBV without cytidine deamination by suppressing pgRNA packaging into viral capsids and inhibiting reverse transcription27. More recently, Chen et al . extensively evaluated APOBEC/AID cytidine deamination activity of HBV rcDNA, demonstrating that deaminating activity decreases in the order A3B >> A3G > A3H or A3C28, while the activity of A3A, A3D, A3H, and A3F was very low or undetectable. Several studies provided evidence that A3A, A3B, and AID can hyper-edit and destroy HBV cccDNA without off-site mutagenesis17,29,30. However, we recently described that transiently overexpressed A3A, A3B, and AID enzymes induce frequent mutations in cancer-related genes in the human genome in cells with low levels of HBV replication even upon transient activation, while A3G inflicts DNA double-stranded breaks 31.
Due to their potent anti-HBV activity, induction of APOBEC/AID is considered a promising antiviral strategy for potentially curing CHB patients. Intracellular expression of cytidine deaminases in the liver can be induced by different agonists, including interferon alpha (IFN-α)17, IFN-γ32, IFN-λ33, lymphotoxin-β receptor agonist17, and others. Otherwise, APOBEC/AID can be transcriptionally controlled by dCas-based molecular activators (CRISPR activation systems, CRISPRa)19,31,34. CRISPRa enables precise activation and tunable control of target gene expression and monogenic or simultaneous activation of several antiviral factors35. Typically, CRISPRa consists of a nucleolytically dead Cas9 protein (dCas) recruited to the regulatory regions of genes via single-guide RNA (sgRNA)36,37. Either dCas9 or sgRNA could also harbor units that mediate activation of transcription34.
HBV’s interactions with innate immunity are complex18; HBV is frequently regarded as a stealthy virus that evades immune recognition38, and a resistant one as it is only moderately sensitive to the IFN-α antiviral response39. Some evidence indicates active countering of innate immunity by the virus (via STAT1 signaling40, TLR2 recognition41, etc.). In this study, we demonstrate the existence of another type of mechanism: saturation of APOBEC/AID factors with their primary target, HBV rcDNA. This “background defense” is provided by the vast overabundance of HBV rcDNA compared to 0–50 intracellular copies of cccDNA and by the preference of APOBEC/AID for single-stranded DNA. We demonstrate that transcriptionally inactivated cccDNA is more efficiently deaminated by APOBEC/AID than the transcriptionally active form. Reducing HBV rcDNA with a reverse transcriptase inhibitor or HBV-specific siRNA markedly enhances cccDNA deamination by APOBEC/AID. This positions HBV rcDNA as an important player in viral suppression mechanisms. However, our results indicate that A3A and A3B, which were not shown to deaminate the host genome in some previous studies, deaminate the host genome when rcDNA is depleted. A3C, A3D and A3H demonstrated weak rcDNA deamination and did not exhibit cccDNA-deaminating properties, but A3C and A3H increased antiviral activity of anti-HBV siRNA. At the same time, we did not detect off-site deamination by A3C, A3D, or A3H in a limited set of cancer-related genes.