Cracking the host functional network involved in hepatitis B virus cccDNA biology

Hepatitis B virus (HBV) causes chronic infection in at least 250 million people worldwide, resulting in approximately 850,000 deaths annually. Chronic carriers are at risk of developing severe liver disease including decompensated cirrhosis and hepatocellular carcinoma (HCC). HBV infection can be prevented with a highly effective prophylactic vaccine. Viremia can be suppressed with nucleos/tide analogues targeting the HBV polymerase, but standard-of-care therapy rarely leads to a cure leaving patients requiring lifelong therapy to prevent relapse of viral replication. Furthermore, even on treatment patients remain at a residual risk of developing HCC. Development of treatments resulting in a functional cure has been hampered by our incomplete understanding of HBV persistence and the difficulty of targeting pharmacologically essential steps in the viral replicative cycle [1].

HBV is a partially double-stranded DNA virus of the Hepadnaviridae family which has a complex replication cycle which is intricately intertwined with the host cell DNA repair machinery (Figure 1). Following receptor mediated endocytosis, hepatitis B virions carry a form of the genome into the host cell, the human hepatocyte, referred to as relaxed circular DNA (rcDNA). rcDNA is a lesion-bearing molecule which must be repaired to form covalently closed circular DNA (cccDNA), the major template for HBV gene transcription, which is central to HBV persistence. HBV’s 3.2 kb genome only encodes four gene products, none of which can convert rcDNA into cccDNA and thus the virus relies on numerous host factors for completing this process, some identified in targeted loss-of-function and biochemical screens (Figure 1). TDP2 plays a role in removing the covalently attached HBV polymerase [2]. Proliferating cell nuclear antigen (PCNA), flap-structure specific endonuclease 1 (FEN1), DNA ligase 1 (LIG1), polymerase delta (POLD) and replication factor C (RFC) other DNA polymerases [2] - crucial components of the cellular DNA lagging strand synthesis [3] - repair similar lesions present on the (+) and (−) strand of the rcDNA molecule, albeit through distinct mechanisms [4]. Repair of rcDNA occurs concomitantly with histone deposition to form a transcriptionally competent, chromatinized cccDNA [5]. However, it remains unclear whether other cellular factors serve redundant functions and/or might be involved modulating this process.

Figure 1: Putative model for the involvement of YBX1 in HBV cccDNA biogenesis.

A major limitation for unbiased screens aimed at identifying factors involved in cccDNA formation in the hepatocyte nucleus is the lack of a reliable, reproducible and sufficiently sensitive assays. This is mostly due to the very small number (<10) of cccDNA copies in HBV-infected cells, the lack of standardization of cccDNA quantification by cccDNA-selective qPCR, and the intrinsically insensitive nature of Southern Blotting, the only technique to unambiguously distinguish cccDNA from all other viral DNA forms.

In this issue of Gut, Verrier et al. [6] established an elegant cell-based HBV cccDNA reporter assay allowing assessment of cccDNA levels and transcription by a simple and robust ELISA readout. Conceptually, a very similar assay was previously described by Cai and colleagues for HBV [7], but Verrier and colleagues used an envelope-deficient duck hepatitis B virus (DHBV) genome to increase the number of cccDNA molecules produced per cell [8]. The authors established a stable hepatoma cell line able to produce DHBV pgRNA, but not precore (preC) RNA, in an inducible manner upon tetracycline withdrawal. Therefore, the N-terminally extended precore protein, the precursor of the processed, secretory e antigen, could only be synthesized from newly formed DHBV cccDNA. The introduction of an HA tag into the preC sequence upstream of the core open reading frame ensured the specific detection of secreted eAg by a HA-ELISA assay, hence correlating with cccDNA formation.

The cell line reporter assay was used to perform a loss-of-function screen targeting 239 genes encoding the human DNA damage response machinery. Among the top-scoring candidates was Y box binding protein 1 (YBX1), an DNA/RNA binding protein involved in several cellular processes, including DNA repair, regulation of transcription, mRNA maturation and translation, [9] localized in both the nucleus and cytoplasm. Cumulative evidence has given rise to the notion that YBX1 promotes the progression of multiple tumor types and serves as a potential tumor biomarker and therapeutic target. YBX1 acts by binding mispaired DNA and interacts with several factors involved in DNA repair, including Ku80, mutS homolog 2 (MSH2), Werner syndrome protein, PCNA and POLD [9] – the latter two being required for cccDNA formation [3] (Figure 1). YBX1 has also been reported to possess mismatch binding and repair activity in mitochondria [9]. Therefore, YBX1 is believed to participate in the DNA repair system and might be involved in responding to certain types of DNA damage. Notably, YBX1 has been shown to also influence the life-cycles of other viruses; e.g. YBX1 appears to interact with dengue virus nucleocapsid and mediates viral assembly and can enhance replication of adenovirus type 5. Verrier and colleagues validated the role of YBX1 in cccDNA formation in state-of-the-art infection models using human HBV, confirming its role in early steps of the HBV life cycle, after virus entry but prior to the establishment of the initial cccDNA pool in infected hepatocytes. Interestingly, no decrease in HBV infection was observed when YBX1 knockdown in HBV-infected HepG2-NTCP cells was achieved after virus infection when the initial cccDNA pool was already established.

However, the approach has a few caveats. Secreted eAg is only an indirect readout for cccDNA accumulation and thus the assay by Verrier et al. would detect not only varying efficiencies in DHBV cccDNA formation, but also in transcription of established cccDNA, and in translation of the precore protein and/or its processing and secretion as DHBeAg. This makes the assay versatile but needs an extensive functional validation of identified host factors or antiviral compounds tested to characterize their specific impact on the viral life cycle. Furthermore, any genes whose knock-down would compromise the viability of rapidly dividing hepatoma cells would drop out of the screen. Moreover, in the way the assay is developed, cccDNA is exclusively formed from DHBV rcDNA recycled into the nucleus, and differences from rcDNA-cccDNA conversion following de novo infection cannot be excluded [10]. Finally, although the biology of duck and human HBV cccDNA is similar, differences have been described, particularly during the recycling mechanism of rcDNA-bearing capsids to the nucleus and during rcDNA to cccDNA conversion [8]. Thus, it is conceivable that cccDNA biogenesis is affected by differences in the viral sequence – the DHBV and HBV share less than 40% sequence identity and/or by species-specific host-viral proteins interactions. Consequently, it cannot be excluded that discrepancies might arise between cellular factors identified in the DHBV-based screen and their involvement in human HBV cccDNA biology. Nonetheless, the identification and characterization of YBX1’s role in both duck and human cccDNA formation represents a strong proof-of-concept for the utility of the assay in deciphering functional interactions between host factors and viral life cycle, while identifying a targetable protein and a potential candidate for antiviral therapy. These contributions pave the way for future work to elucidate the precise mechanism at the basis of YBX1-dependent regulation of cccDNA biology and its potential effect on the steady-state levels of in vivo cccDNA.

Abbreviations:

cccDNA

covalently closed circular DNA

DHBV

duck HBV

DHBeAg

duck HBV e antigen

DNA

deoxyribonucleic acid

eAg

hepatitis B virus e antigen

ELISA

enzyme linked immunosorbent assay

FEN1

flap-structure specific endonuclease 1

HBV

Hepatitis B virus

HCC

Hepatocellular carcinoma

kb

kilo base

Ku80

Lupus Ku autoantigen protein p80

LIG1

DNA ligase 1

MSH2

mutS homolog 2

NTCP

sodium taurocholate co-transporting peptide

PCNA

Proliferating cell nuclear antigen

pgRNA

pre-genomic RNA

POLD

polymerase delta

preC

precore

rcDNA

relaxed circular DNA

RFC

replication factor C

RNA

ribonucleic acid

YBX1

Y box binding protein 1