Nearly 300 million people worldwide suffer from chronic hepatitis B virus (HBV) infections.1 Chronic HBV causes inflammation and is associated with an increased risk of extensive liver damage and can progress to hepatocellular carcinoma. The traditional therapeutic approach uses antiviral therapies, such as entecavir and tenofovir; while these therapeutics suppress HBV DNA levels they need to be continued long-term as they often do not lead to functional cure. Therefore, novel therapeutics are under development as monotherapies or co-therapy options to achieve functional cure.
Oligonucleotide therapeutic approaches (either using a monotherapy or co-therapy approach) have shown promising results in clinical trials, such as GSK’s antisense oligonucleotide, Bepirovirsen (Phase 3), and siRNAs developed by Vir Biotechnology/Alnylam Pharmaceuticals, Elebsiran co-therapy (Phase 2/3), and Janssen Pharmaceuticals, JNJ-73763989 (Phase 2). 2-4 However, even with the increased success of these therapeutics, there is still a subset of individuals who do not reach functional cure status. Therefore, it is important to understand why these therapeutics fail to clear HBV for all patients. Additionally, having other therapeutic approaches to add to the clinician’s repertoire will be key to curing HBV globally.
Another approach uses CRISPR/Cas9 nucleases targeting cccDNA. However, this approach creates double-stranded breaks (DSB) which can result in off-target effects including random insertion/deletions and genomic rearrangements.5 Due to this, gene editors are being developed to try to circumvent DSB-intermediates. Base editors, such as cytosine and adenine base editors, might offer the solution. Here we will introduce the models we offer at PhoenixBio (PXB-mice and PXB-cells) and discuss two papers published by Beam Therapeutics that utilized these models to investigate base editors as potential therapeutic avenues for chronic HBV.
The PXB-mouse®, a cDNA-uPA/SCID mouse transplanted with primary human hepatocytes, has high human hepatocyte engraftment, with up to 95% humanization. They have stable expression of human genes and proteins as well as human-specific metabolism and excretion pathways. Importantly, PXB-mice express human sodium taurocholate co-transporting polypeptide (NTCP), making them susceptible to HBV infection. However, unlike other preclinical models, PXB-mice can be infected with clinically relevant wild-type HBV variants and support the entire HBV life cycle. As such, relaxed circular DNA (rcDNA) is converted to covalently closed circular DNA (cccDNA) in the liver of PXB-mice.6-8 It is important to note that while PXB-mice lack an adaptive immune system and might not be suitable for some immune-mediated therapeutic development/discovery programs, the model excels at predicting therapeutic efficacy for other antiviral approaches.
PXB-cells® are primary human hepatocytes isolated from the PXB-mouse. They express high levels of phase I and II drug metabolizing enzymes, are stable in culture long-term, and are susceptible to a wide range of HBV inocula.9,10 Additionally, the HBV receptor NTCP is robustly expressed by PXB-cells long-term, demonstrating that these hepatocytes are susceptible to HBV and support investigating the complete viral life cycle.11 Taken together, these models are a good tool for anti-viral therapeutic development.
Targeted HBV gene editing aims to inactivate cccDNA and integrated HBV DNA to reach functional cure. With cytosine base editors (CBE), guide RNAs are used to direct CBE to relevant HBV targets and stop codons are implemented by converting cytosine into thymine to silence HBV. As such, the level of cccDNA should remain unchanged but viral gene expression should be effectively silenced. BE4 (CBE) was used with two guide RNAs targeting cccDNA (site g37 which targets HBs and site g40 which targets precore).12 In this recent study, BE4 treatment significantly reduced HBV DNA, HBsAg, and HBeAg levels in HBV-infected PXB-cells (Figure 1A-B).13 Moreover, HBV-targeted BE4 treated PXB-cells had similar levels of cccDNA as the control (PCSK9) CBE treated cells (Figure 2A).12 Cytosine to thymine changes were observed in both targeted regions of cccDNA, demonstrating that reductions in viral markers was due to an increase in stop codon conversions (Figure 2B).12
Figure 1: Cytosine base editing significantly reduces viral marker levels in HBV-infected PXB-cells. Extracellular HBV DNA levels were assessed in HBV-infected PXB-cells treated with either HBV-targeted BE4 (targeting site 37 and 40 of cccDNA), control BE4 (targeting PCSK9), or lamivudine (3TC) (A). HBsAg, HBeAg, and total HBV DNA levels were also assessed at 25 days post-infection (B). (modified from Smekalova EM, et al., 2023)
While cytosine base editors effectively reduce HBV markers in vitro and in vivo, they found that the editing efficiency may be improved by changing the base editing approach. With adenine base editors (ABE), adenine is converted to guanosine. Engineered adenine deaminases, such as TadA (tRNA-specific adenine deaminase), have been shown to have better editing efficacy with negligible off-target effects.13 In a recently published pre-print, Kumar and colleagues shared their results from their study that investigated the anti-HBV effects of ABE-encoding mRNA with two different guide RNAs (gS1 and gS2). These ABE target HBV S antigen open reading frames to generate missense mutations with the aim to inhibit HBV replication.14 HBV-infected PXB-cells had significantly less HBV surface antigen levels when treated with ABE (Figure 3B).14 One of the guide RNAs (gS1) also reduced HBV DNA levels as efficiently as the HBsAg (Figure 3C).14
Next, they tested whether these ABE reduce HBV infection in the PXB-mouse model. PXB-mice were infected with HBV and treated with ABE once viraemia stabilized (>4 weeks post-infection). HBV-targeted ABEs were encapsulated in lipid nanoparticles for delivery and were administered twice (Day 0 and ~Day 20 post-stable phase of infection) (Figure 4).14 Serum HBsAg and HBV DNA levels were reduced in PXB-mice treated with ABE (Figure 4B-D).14 While entecavir (ETV) treatment had a robust response, with a dramatic decrease in serum HBV DNA levels, at therapy cessation HBV DNA quickly rebounded (Figure 4D).14 Unlike ETV, reduced HBV DNA levels were maintained for the duration of the study for both LNP-delivered ABEs, with reductions between 1- to 2-fold compared to untreated HBV-infected PXB-mice at day 56 post-ABE administration (Figure 4D).14 Taken together, HBV-targeted base editors show promising results in both the PXB-cells and PXB-mouse models.
Figure 4: PXB-mice, a chimeric mouse transplanted with human hepatocytes, were infected with HBV for ~4 weeks prior to treatment with ABE-gS1, ABE-gS2, entecavir (ETV), or control ABE. Serum HBsAg (B) and HBV DNA levels (C) were assessed over 56 days after therapy administration (dpi: days post first LNP injection). Editing efficiency was assessed by measuring the percent of adenine to guanine conversion (47% for gS1 and 68% for gS2) (D). (modified from Kumar A, et al., 2026)
PXB-mice and cells express high-levels of human-specific genes important for drug metabolism. They are both susceptible to native HBV infections, including a wide range of HBV sources such as clinical strains. Making these models an invaluable tool for HBV research, including antiviral therapeutic development. Have a study in mind? Contact us today to discuss how PXB-mice and/or PXB-cells can enhance your research pipeline.