Targeting the hepatitis B cccDNA with a sequence-specific ARCUS nuclease to eliminate hepatitis B virus in vivo

Abstract

Persistence of chronic hepatitis B (CHB) is attributed to maintenance of the intrahepatic pool of the viral covalently closed circular DNA (cccDNA), which serves as the transcriptional template for all viral gene products required for replication. Current nucleos(t)ide therapies for CHB prevent virus production and spread but have no direct impact on cccDNA or expression of viral genes. We describe a potential curative approach using a highly specific engineered ARCUS nuclease (ARCUS-POL) targeting the hepatitis B virus (HBV) genome. Transient ARCUS-POL expression in HBV-infected primary human hepatocytes produced substantial reductions in both cccDNA and hepatitis B surface antigen (HBsAg). To evaluate ARCUS-POL in vivo, we developed episomal adeno-associated virus (AAV) mouse and non-human primate (NHP) models containing a portion of the HBV genome serving as a surrogate for cccDNA. Clinically relevant delivery was achieved through systemic administration of lipid nanoparticles containing ARCUS-POL mRNA. In both mouse and NHP, we observed a significant decrease in total AAV copy number and high on-target indel frequency. In the case of the mouse model, which supports HBsAg expression, circulating surface antigen was durably reduced by 96%. Together, these data support a gene-editing approach for elimination of cccDNA toward an HBV cure.

Keywords: AAV; ARCUS; HBV; HBsAg; LNP; NHP; gene editing; gene therapy; hepatitis B; meganuclease.

Graphical abstract

Figure 1

Optimization of nucleases targeting the HBV polymerase gene (A) Nuclease activity as measured by indel frequency is shown for representative nucleases from each optimization cycle. On-target % indels quantified by droplet digital PCR (ddPCR) following nuclease treatment in Hep3B cells. (B) Semi-quantitative analysis of specificity measured by oligo-capture in HepG2-HBV cells following nuclease treatment. Individual genomic sites with incorporated probe are indicated as dots (blue represents off-target sites, orange represents the on-target site), with the darker blue dots representing sites with greater homology compared with the intended target site, and lighter-colored dots having more sequence variation. The read count indicates the number of times a site was recovered in the sequencing data, with higher read counts qualitatively correlating with higher frequency of editing.

Figure 2

Specificity of Gen 4 and Gen 5 ARCUS-POL nucleases (A) On-target indels in cccDNA and indels at three putative off-target sites identified in the oligo-capture assay were quantified by NGS. For on-target indels in cccDNA, total cellular DNA was treated with T5 exonuclease to enrich for cccDNA specifically prior to NGS. (B) Junctions between HBV and host DNA identified by targeted sequencing of PHHs mock treatment or treatment with Gen 4 or Gen 5 ARCUS nucleases. Junction points identified by oligo-capture analysis as potential off-target loci are shown as rhombi. Shape size reflects junction-site frequency after normalization with control host genes. Shape color indicates host chromosomes. Gray lines link junctions with at least 20 reads supporting a chromosomal translocation bridged by the integrated HBV. See also Figure S3.

Figure 3

Antiviral effect of Gen 5 ARCUS-POL nuclease in HBV-infected PHHs (A) Southern blot time-course analysis of HBV-infected PHHs following transfection of either Gen 5 ARCUS-POL or a non-HBV-targeting nuclease. (B) Densitometry was used to quantify cccDNA levels in PHH cells treated with Gen 5 ARCUS-POL or a non-targeting nuclease from the Southern blot. cccDNA levels were normalized to mtDNA and are shown as a percent of the normalized cccDNA levels of the non-targeting nuclease at day 3 post-HBV infection. (C) On-target indels in cccDNA were quantified by NGS. Prior to NGS, total cellular DNA was treated with T5 exonuclease to enrich for cccDNA. (D) Extracellular HBV DNA was quantified by a qPCR assay from PHH supernatant. (E) Extracellular HBsAg was quantified by CLIA from PHH supernatant. (F) Extracellular albumin was quantified by ELISA from PHH supernatant. Lines and error bars represent mean ± SD.

Figure 4

Targeting integrated viral DNA with Gen 5 ARCUS-POL nuclease HepG2-sAg cells were electroporated with Gen 5 ARCUS-POL nuclease mRNA or control mCherry mRNA at concentrations of 10 and 100 ng or were mock transfected. On days 3, 6, and 9, cells were harvested for (A) on-target indel frequency analysis via ddPCR, and (B) supernatant was collected for extracellular HBsAg analysis via CLIA. Lines and error bars represent mean ± SD of biological replicates.

Figure 5

Gen 5 ARCUS-POL nuclease evaluation in an episomal AAV mouse model Three weeks after AAV9-HBsAg administration, NSG mice were i.v. dosed with an LNP containing Gen 5 ARCUS-POL nuclease mRNA at 2 mg/kg. Blood draws were performed weekly, and animals were sacrificed 4 weeks post-LNP administration. At sacrifice, livers were harvested, and genomic DNA was extracted and assessed for (A) AAV copy number per diploid cell and (B) indel analysis at the target site present on the AAV9-HBsAg vector. (C) Serum was isolated from blood samples and analyzed for HBsAg via CLIA. (D) Immunohistochemistry for HBsAg detection was performed on liver tissue from necropsy (scale bars: 400 μm). Lines and error bars represent mean ± SD. Statistical analysis was performed using Mann-Whitney’s two-tailed test. ∗p ≤ 0.05. See also Figure S4.

Figure 6

Episomal HBV-AAV NHP study timeline All NHPs were given prednisolone at 1 mg/kg/day throughout the entire study. The AAV control and nuclease-treated groups were i.v. administered AAV8-HBsAg on day 8. At days 22 and 63, the LNP control and nuclease-treated groups were i.v. administered Gen 5 ARCUS-POL∗ LNP. Liver biopsies were taken for animals in all groups on days 36 and 77. Serum was collected throughout the study for HBsAg analysis, serum chemistry, and cytokine analysis. Schematic was created with BioRender.

Figure 7

ARCUS-POL∗ nuclease evaluation in an episomal HBV-AAV NHP model (A) AAV copy number was measured by ddPCR in liver tissue from biopsies and necropsy. (B) AAV was visualized by ISH in liver sections from biopsies (scale bars: 80 μm). (C) On-target indel analysis on the AAV8-HBsAg was measured via ddPCR. (D) Serum was analyzed for HBsAg levels. Lines and error bars represent mean ± SD. Statistical analysis was performed using two-tailed t test. ∗p ≤ 0.05; ∗∗p ≤ 0.01.