Targeting Hepatitis B Virus With CRISPR/Cas9

君看一叶舟

OPEN

Christoph Seeger1 and Ji A Sohn1

1Institute for Cancer Research, Fox Chase Cancer Center, Philadelphia, Pennsylvania, USA

Correspondence: Christoph Seeger, Institute for Cancer Research, Fox Chase Cancer Center, 333 Cottman Avenue, Philadelphia, Pennsylvania 19111, USA. E-mail: [email protected]

Received 29 September 2014; Accepted 3 November 2014

Top of page
Abstract
Hepatitis B virus persistence in infected hepatocytes is due to the presence of covalently closed circular DNA (cccDNA), the template for the transcription of viral RNAs. Antiviral therapies with nucleoside analogues inhibit replication of HBV DNA in capsids present in the cytoplasm of infected cells, but do not reduce or destroy nuclear cccDNA. To investigate whether cccDNA derived from infectious HBV could be directly targeted for destruction, we used the CRISPR/Cas9 system in HepG2 cells expressing the HBV receptor sodium taurocholate cotransporting polypeptide (NTCP). We tested different HBV-specific guide RNAs and demonstrated that they could inhibit HBV infections up to eightfold. Inhibition was due to mutations and deletions in cccDNA similar to those observed with chromosomal DNA cleaved by Cas9 and repaired by nonhomologous end joining (NHEJ). Interferon alpha (IFN-α) did not have a measurable effect on the antiviral activity of the CRISPR/Cas9 system, suggesting that Cas9 and NHEJ activities are not affected by induction of an innate immune response with the cytokine. Taken together, our results demonstrated that Cas9 can be recruited to cccDNA, opening the possibility for the development of future antiviral strategies aimed at targeting cccDNA for endonucleolytic cleavage with small molecules.

Keywords: antiviral therapy; chronic hepatitis B; covalently closed circular DNA; CRISPR/Cas9; hepatitis B virus; nonhomologous end joining

君看一叶舟

Introduction
Over 350 million people worldwide are infected with hepatitis B virus (HBV).1 About 10% of carriers suffer from chronic hepatitis B (CHB), a condition that can lead to cirrhosis and hepatocellular carcinoma. It is estimated that over 700 thousand HBV-infected patients succumb to the sequelae of infection each year.2 Antiviral therapies developed during the past 20 years are effective in suppressing, but not eliminating, HBV infections.3 The apparent lack of a sustained viral response following cessation of antiviral therapy is due to the stability of the covalently closed circular (ccc) DNA.4,5 CccDNA is formed from the relaxed circular (rc) genomic DNA by a DNA repair mechanism that is still not understood (Figure 1).6 Once established as a minichromosome, cccDNA functions as a template for the synthesis of four RNA transcripts, pregenomic (pg) RNA, preS, S, and X RNAs.4,7 DNA replication occurs in core particles present in the cytoplasm of infected cells by reverse transcription of pg RNA.8,9 DNA containing core particles assemble with envelope proteins, enter the secretory pathway, and are secreted into the blood stream. Nucleoside analogue-based antiviral therapies inhibit the viral reverse transcriptase, and hence the formation of progeny virus. However, they do not target nuclear cccDNA in infected hepatocytes. What makes matters worse is that infected hepatocytes harbor multiple copies of cccDNA as a consequence of an intracellular DNA amplification mechanism.10 CccDNA amplification occurs during the initial phase of an infection when levels of viral envelope proteins in hepatocytes are still low, favoring retrograde passage of core particles into the cell nucleus over assembly with envelope proteins.11

君看一叶舟

Life cycle and genomic organization of HBV. (a) The enveloped infectious HBV particle is uncoated after infection releasing the capsid (1) that migrates to the cell nucleus where it delivers the genomic rcDNA (2) which is converted into cccDNA (3), the template for transcription of the viral RNAs. (b) Physical map of HBV with the four open-reading frames and viral RNAs. Also indicated are the positions of enhancers I and II (EN1, EN2). For additional details, see ref. 5.

Full figure and legend (143K)

君看一叶舟

A key issue for antiviral therapy is the development of strategies that can either alone or in combination with nucleoside analogues lead to the reduction and eventual eradication of cccDNA from infected hepatocytes. Information about the stability of cccDNA has been forthcoming from studies with cell culture models for viral infection and replication as well as from the analysis of viral DNA extracted from liver biopsies of animals recovering from transient infections. Whereas in nondividing primary hepatocyte cultures cccDNA is stable,12,13 in dividing tissue culture cells, cccDNA is passed on to daughter cells suggesting that it can “survive” mitosis.14,15 Results from experimentally infected chimpanzees and woodchucks indicated that during resolution of acute infections loss of cccDNA occurs primarily by killing of infected heaptocytes by cytotoxic T cells and possibly by noncytolytic mechanisms induced by cytokines and other unknown factors that are part of the immune-mediated clearance phase.16,17,18,19,20 The exact level of contribution of each event to the decline of cccDNA levels observed during the resolution of an infection is still not resolved. A major gap in knowledge concerns the nature of the host factors that play a role, if any, during noncytolytic clearance of cccDNA. A recent report claimed that interferon alpha (IFN-α and lymphotoxin beta (LT-β) pathways might cause the destruction of cccDNA by certain endonucleases following deamination of the minus strand DNA by APOBEC3A (A3A) and A3B.21 However, whether these results bear any relevance for clearance of HBV infections during acute infections remains to be determined.22

Efforts to develop strategies that could induce cccDNA loss during CHB have been hampered by the lack of tissue culture systems that are permissive for HBV infection and at the same time are amenable to experimental manipulations, such as gene knockout technologies. However, identification of the HBV receptor, sodium taurocholate co-transporting polypeptide (NTCP), opened the possibility for establishment of HepG2 cell lines permissive for infections with HBV produced in tissue culture cells.23,24 In this report, we demonstrate the utility of this system for investigations on cccDNA formation and stability by combining it with the clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 platform for efficient gene knockout.25,26,27 Using this system, we investigated whether cccDNA formed from infectious virus can be targeted by a recombinant endonuclease and whether cellular enzymes can either repair or destroy cccDNA under selected conditions.

君看一叶舟

Results
Production of NTCP/Cas9 cells permissive for HBV infection
In a first step, we established HepG2 cells expressing NTCP from a lentivirus vector, NT-GFP, that also expresses GFP (Materials and Methods). HepG2/NTCP cells can be enriched by fluorescence activated cell sorting (FACS) for subpopulations that express high levels of the receptor. We found that permissiveness for HBV infections directly correlated with NTCP levels in these cells (Figure 2a). The fraction of HepG2/NTCP cells that could be infected with HBV derived from the supernatant of HepAD38 cells,28 reached up to 50–60%, primarily depending on NTCP cells expression levels (Figure 2b). HBV-infected HepG2/NTCP expressed the three major viral RNA transcripts, pregenomic RNA and preS and S RNAs (Figure 2c) as well as cytoplasmic HBV DNA that is characteristic of hepadnavirus replication (Figure 2d). As expected, accumulation of cytoplasmic DNA was reduced when cells were maintained in the presence of entecavir, an HBV reverse transcriptase inhibitor (Figure 2d). Southern blot analysis was not sensitive enough to detect cccDNA, suggesting that the copy number of cccDNA in infected cells was very low, most likely not exceeding one. However, cccDNA was detected in infected HepG2/NTCP cells by PCR using primers that favor amplification of cccDNA over the more abundant rcDNA (Figure 2e). To combine the HBV infection system with the CRISPR/Cas9 platform for efficient gene inactivation, we introduced in a second step the endonuclease Cas9 into HepG2/NTCP cells with a second lentivirus vector, pCW-cas9.27 Cas9 expression is under the control of the CMV-tetracycline (tet) promoter and can be induced with doxycycline (dox) (Figure 2f).
Figure 2.

君看一叶舟

HBV infection of NTCP/Cas9 cells. (a) Expression of NTCP in HepG2 cells expressing Cas9. Cells were sorted by FACS to enrich for populations with high NTCP levels. Lane 1, cell pool prior to FACS; lane 2, control HepG2 cells; lanes 3 and 4, 5 and 70% fraction of FACS sorted cells with the highest GFP expression levels. (b) Infection of NTCP/Cas9 cells with HBV derived from HepAD38 cells. Cells were processed for IF analysis with HBcAg-specific antibody C1-5 8 days after infection. (c) Northern blot analysis of total RNA extracted from HepG2/NTCP cells infected with HBV (lane 3) and HepAD38 cells as a control (lane1). Uninfected cells (lane 2) and ribosomal 28S RNA served as a control for the amount of RNA loaded onto each well. Lane 4 is a longer exposure of lane 3. (d) Analysis of core DNA extracted from HepG2/NTCP (H-NT) cells infected with HBV. DNA was extracted 14 days after HBV infection (lanes 1 and 2). Cells were cultured without (lane 1) and with entecavir (ETV, 10 µg/ml)) (lane 2). DNA extracted from HepAD38 (AD38) cells served as a control (lanes 3 and 4). Lane 3 was loaded with 1/10th of the amount of DNA used for lane 2. (e) PCR analysis of HBV DNA extracted from infected NTCP/Cas9 cells. DNA was extracted with the Hirt procedure (H) from total cell extracts or from the cytoplasm (C) of the HBV-infected cells. Primers used for the PCR reactions are described in Supplementary Table S1. Ccc, CCC DNA–specific primers overlapping the cohesive overlap in rcDNA (Figure 1a); mito, mitochondrial DNA-specific primers; rc, primers for amplification of rcDNA. (f) Conditional expression of Cas9 in two different cell clones. Cells were cultured with (lanes 2 and 4) and without (lanes 1 and 3) dox. dox, doxycycline tub; tubulin.

Full figure and legend (73K)

君看一叶舟

CRISPR/Cas9 targeting of HBV cccDNA
To investigate whether cccDNA derived from input HBV can be targeted for cleavage and destruction by Cas9, we designed several sgRNAs spanning the ENII/CP region and the PC regions on the HBV genome (Figure 1b and Table 1). We selected this region because it encodes HBx and the overlapping control region, both important for transcription of pregenomic RNA encoding the HBV core antigen (HBcAg), which can be detected in infected cells by immunofluorescence with available antibodies. We validated the cleavage efficiency of the sgRNAs in HepAD38/Cas9 cells carrying an integrated genome derived from HBV genotype ayw.28 Infection of HepAD38/Cas9 cells with lentiviruses expressing the relevant sgRNAs resulted in cleavage of integrated HBV DNA as determined with the Surveyor assay for the detection of mismatched DNA in PCR amplified DNA fragments29 (Supplementary Figure S2). The results showed that all five sgRNAs tested efficiently cleaved chromosomal HBV DNA.
Table 1 - Position and nucleotide sequences of Cas9 targets on HBV.
Table 1 - Position and nucleotide sequences of Cas9 targets on HBV - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact [email protected] or the author
Full table (20K)

君看一叶舟

To determine the potential antiviral activity of CRISPR/Cas9 against HBV, we infected NTCP/Cas9 cells first with lentiviruses expressing individual sgRNAs and then with HBV. HBV infected cells were maintained in medium containing dox to induce Cas9 expression. The fraction of HBV-infected cells was determined by immunofluorescence with an HBcAg antibody 7–10 days post HBV infection (Figure 3a,b). The results showed that expression of sg5 and sg6 guide RNAs inhibited HBcAg expression between 6- and 10-fold compared to controls that either were not infected with the respective lentivirus vector or were maintained in the absence of dox to prevent Cas9 expression. Sg7 exhibited a weaker antiviral effect reducing the fraction of HBcAg-positive cells between twofold and sixfold. In two experiments, we observed a slight reduction in the fraction of HBV-infected cells in lentivirus-infected cells maintained without dox compared with uninfected control cells, which could be attributed to leakiness of the tetracycline promoter resulting in low level expression of Cas9 even in the absence of dox (Figure 3b). Alternatively, we cannot rule out that small differences in expression of NTCP existed between control cells and cells expressing different sgRNAs. This would also explain higher infection levels observed with cells expressing sg5 compared with other cells in the experiment (Figure 3b).

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact [email protected] or the author
Inhibition of HBcAg expression with CRISPR/Cas9. (a) The results from an HBV infection on NTCP/cas9 cells expressing sg5 and sg7 guide RNAs. d, doxycycline. Immunofluorescence was performed with HBcAg-specific antibody C1-5 8 days after HBV infection. (b) Results from a quantitative analysis of HBcAg-positive cells from three different experiments performed with sgRNAs 5, 6, and 7, as described in a. Co, control; UI, uninfected. *P < 0.05; **P < 0.01; ***P < 0.001.

Full figure and legend (368K)

君看一叶舟

Analysis cccDNA in HBV-infected NTCP/Cas9 cells
So far, our results indicated that HBV infections can be inhibited with the CRISPR/Cas9 system, implying that newly formed cccDNA is cleaved and either repaired by NHEJ, like chromosomal DNA, or destroyed by cellular nucleases. To determine the fate of cccDNA, we extracted DNA from HBV-infected NTCP/Cas9 cells expressing sgRNAs 5, 6, and 10 under conditions described in the previous section. Sg10 RNA exhibited an antiviral activity similar to that of sg5 and sg6 (results not shown). Purified DNA was PCR amplified with cccDNA-specific primers (Supplementary Table S1, Figure 2e), and the PCR products were cloned and sequenced. As a control, we analyzed DNA from HBV-infected cells (maintained in dox-free medium) expressing sgRNAs in the absence of Cas9 expression. As expected, all (26) sequenced clones were identical with wild type (Table 2). In contrast, clones derived from cells expressing sgRNAs and Cas9 exhibited mutations characteristic of Cas9 cleavage (Tables 2 and 3). The most frequent type of mutation was a single-nucleotide insertion or deletion within the sequence motif corresponding to the respective sgRNA (Figure 4). In addition, we observed deletions spanning 2–30 nucleotides in length as well as larger deletions spanning up to 2.3 kb in length. In about two-thirds of the larger deletions, one of the two break points mapped to the Cas9 cleavage site near the PAM (protospacer adjacent region) sequence abutting the 3′ end of the sgRNA.

君看一叶舟

Nucleotide sequence analysis of cloned cccDNA. (a) The target sequence (plus strand) of sg5 guide RNA on the HBV genome and the position of five selected mutations identified in cloned PCR fragments derived from cccDNA of HBV-infected cells expressing sg5 are inidcated. Deletions are indicated with dots, insertions with a plus sign. Position 2,987 corresponds to the 3′ end of sgRNA 5 (sg5) abutting the PAM sequence (GTT). For the positions of the pol and X genes on the HBV genome, see Figure 1b.

Full figure and legend (50K)