TALENs targeting HBV: designer endonuclease therapies for viral infections.

Chronic viral infections have plagued humanity for millennia, causing lifelong and incurable diseases. For many viruses, persistence is caused by the presence of long-lived forms of viral DNA in infected cells. Current therapies can suppress viral replication, but they have little or no effect on long-lived DNA forms, and thus viral replication is able to resume as soon as therapy is stopped. This suggests that if long-lived DNA forms could be targeted therapeutically, cure might be possible (Figure 1). The recent development of designer targeted endonucleases such as homing endonucleases, zinc-finger nucleases (ZFNs), transcription activator–like effector nucleases (TALENs), and the CRISPR (clustered regularly interspaced short palindromic repeats) system has brought this idea closer to reality.1 The function of these enzymes is to specifically recognize and cleave selected DNA sequences, which results in gene disruption upon imprecise DNA repair. In this issue of Molecular Therapy, Bloom and colleagues investigate TALENs targeting the hepatitis B virus (HBV) genome and analyze their antiviral activity in several models of HBV infection, including an in vivo mouse model.2 Their detection of TALEN-induced mutations in the long-lived HBV covalently closed circular DNA (cccDNA) represents a substantial advance in the field and supports continued efforts to develop this approach for ultimate clinical use. The decision of Bloom et al. to target HBV is logical given the global burden of disease caused by this virus. More than 350 million people are chronically infected by HBV and therefore have a much higher risk of cirrhosis and hepatocellular carcinoma. The development of a definitive curative therapy, when coupled with the efficacious HBV vaccine, offers the possibility of substantially reducing the incidence of HBV-associated suffering and death. Similar curative therapy strategies could be envisioned for other persistent viral infections, including HIV, human papillomaviruses, and human herpesviruses. Genome editing as a therapy for viral disease is not a new concept. Indeed, HIV clinical trials are already under way (NCT01252641, NCT00842634, and NCT01044654) in which ZFNs are being used to disrupt the gene that encodes CCR5, an essential coreceptor for HIV entry.3,4,5 The goal of directly targeting long-lived viral DNA rather than a cellular receptor, however, is still in its infancy.1 The majority of work thus far has been performed in vitro, using artificial models that to varying degrees reflect the clinical situation. HBV DNA has been shown to be susceptible to gene disruption by ZFNs6 and transcriptional suppression by zinc-finger proteins.7 Similarly, ZFNs or homing endonucleases have also been used to successfully target HIV,8 human T-cell leukemia virus type 1,9 human papillomavirus,10 and herpes simplex virus type 1.11 In their new study, Bloom et al.2 have shown that the field continues to advance and that TALEN-based antiviral strategies can also be efficacious. They demonstrate TALEN-induced specific gene disruption and knockdown of the encoded viral proteins and, importantly, show gene disruption in hepatocytes harboring HBV DNA sequences in an in vivo mouse model. As a first step, the team designed TALENs for HBV sequences in the open reading frames for HBV core antigen (HBcAg), surface antigen (HBsAg), and polymerase protein (C-, S-, P1-, and P2-TALENs, respectively). They compared potential target sites across reference HBV genotypes to select sites with high conservation, and took care to avoid potential sites with homology to the human or murine genomes, which might lead to off-target cleavage. Initial in vitro experiments involved transfection of hepatocyte-derived cells with a plasmid that produces replication-competent HBV, along with TALEN-encoding plasmids. The researchers looked for targeted anti-HBV activity by measuring HBsAg production, which was significantly knocked down by S-TALEN and one of the P-TALENs. In subsequent experiments performed in HepG2.2.15 cells, an in vitro HBV model that contains cccDNA, the anti-HBV effect was most pronounced in cells transfected three times with the S-TALEN–encoding plasmid and cultured under mildly hypothermic conditions (30 °C), a treatment that allows gene disruption to be detected more easily.12 TALEN-mediated gene disruption was the likely cause of viral protein knockdown, and DNA mutations were found in approximately 31% of cccDNA molecules. This last finding is of particular interest, because the ability to specifically assay target sequence mutations in cccDNA is not trivial. In HepG2.2.15 cells, HBV DNA exists as integrated DNA, relaxed circular DNA, and cccDNA forms. The authors were able to convincingly show that cccDNA molecules were mutated at the TALEN target sites. To do this, they utilized treatment with a DNase that degrades all DNA forms but cccDNA, and combined this with a well-designed cccDNA-specific polymerase chain reaction primer strategy. Although the TALEN-induced mutations detected in the cccDNA compartment are highly encouraging, it is important to note that the source of these cccDNA mutations could be from gene disruption of the integrated HBV genome, which is constitutively active in HepG2.2.15 cells and is the initial template for cccDNA production. Of course, these same mutations could have arisen from TALEN-mediated cleavage of cccDNA molecules themselves, and it is quite possible that these mechanisms occurred simultaneously. With an effect on the cccDNA sequences and a knockdown in viral products clearly shown, Bloom et al.2 next tested their TALENs in vivo. They used hydrodynamic injection to co-deliver TALEN-encoding plasmids and a plasmid containing HBV sequences to the mouse liver.13 They then monitored production of viral DNA and proteins, and checked for DNA mutations in livers and serum extracted from treated mice. The HBV TALENs were clearly specific, as S-TALEN–treated mice showed significant reductions in HBsAg, whereas mice receiving C-TALEN exhibited fewer cells positive for HBcAg. Perhaps most promising of all, mice receiving TALEN treatment exhibited no more signs of toxicity than did mock-injected mice. At the same time, there are limitations to the hydrodynamic injection model. Because it does not replicate HBV or produce cccDNA, it cannot be used to evaluate HBV replication or persistence. Given that in vitro cccDNA mutations were detected only upon triple transfection of cells treated at 30 °C, and that the in vivo model used here is at 37 °C and does not contain cccDNA, in the future it will be important to confirm that at 37 °C TALENs can cleave cccDNA directly. Given the promising efficacy data shown in this study, future attention needs to be given to delivery, which remains a significant hurdle to therapies of this kind. In this study, the authors efficiently delivered TALEN-expressing plasmids to the liver by hydrodynamic injection. In humans, however, intravenous hydrodynamic plasmid injection is not a realistic option, and therefore an alternative delivery method would be necessary. Numerous viral and nonviral gene delivery systems have been used for liver-directed gene therapies,14 but for TALEN delivery there are clear challenges for all the currently available methods. Most TALENs used for DNA editing are heterodimers made up of two DNA-recognizing domains, each bound to a catalytic FokI nuclease domain. Dimerization of the two TALEN halves is required for DNA cleavage. Therefore, two separate genes must be delivered to each HBV-infected cell to disrupt the HBV genome. Ideally a single vector would deliver both genes, because delivery via two separate vectors would require cotransduction of all HBV-infected hepatocytes to have an effect—a requirement that could pose a hurdle for an effective therapeutic outcome.15 Unfortunately, because of the highly repetitive repeat-variable diresidue domains that are inherent in TALENs, recombination between TALEN subunits or between separate domains on the same subunit may cause problems with delivering an effective enzyme. Both intramolecular and intermolecular repeat-variable diresidue domain recombination could produce a TALEN with altered DNA specificity, which could lead to off-target effects and therapy-associated toxicity. It is possible that enzyme platforms such as homing endonucleases, which are small and easily packaged and do not possess repeat regions or require dimerization, will ultimately prove to be more suitable for clinical use. Another limitation to TALEN delivery is their size, which is substantially larger than that of either ZFNs or homing endonucleases. Many gene delivery vectors have a limited packaging capacity, and this makes TALEN packaging challenging. For example, adeno-associated virus vectors have shown great promise for HBV-specific therapies in mice16 and liver-directed gene therapy in humans,17 but their payload capacity is, at most, 4.7 kilobases. For certain delivery systems, including adeno-associated virus, a single vector would not be able to deliver both TALEN subunits. Current efforts toward the development of new enzymes, such as the recently described chimeras of transcription activator-like effectors and homing endonucleases C-TALENs18 and MegaTALs19 may make packaging easier. Nevertheless, if TALENs can be packaged into gene transfer vectors and efficiently delivered to HBV-infected hepatocytes in a regulated manner, the work of Bloom et al.2 suggests that they may have great promise as a curative HBV therapy.