The key to any gene therapy strategy is to ensure that modifications occur in the desired target sequence. Our analysis indicates that minimizing or eliminating the potential for off-target binding requires an endonuclease to target a sequence of at least 17 nucleotides in length. To that end, ZFNs, TALENs and HEs can all be designed to meet or exceed this threshold. Furthermore, the enzyme specificity must be retargeted to viral sequences. As noted above (see Our Tools), engineering of the three classes of rare-cutting endonuclease we utilize towards desired viral target sequences comes with varying degrees of difficulty. This fact has led us to productive collaborations with the Northwest Genome Engineering Consortium and groups both inside and outside Fred Hutch with expertise in TALEN and HE design.

Redirecting the specificity of rare-cutting endonucleases to essential viral genes will maximize our ability to disable latent viral genomes. An ideal viral sequence target will reside in an essential viral gene, be highly conserved across viral isolates and will incur a high fitness cost when disrupted. To this end, we are developing custom sequence analysis platforms that will facilitate our target sequence selection.

We are developing multiple methods for delivery of our enzymes to cell types that harbor latent viral reservoirs. One enzyme delivery method we are using is adeno-associated virus (AAV) vectors. AAV has a number of benefits as a gene therapy delivery vector as it is non pathogenic, is weakly immunogenicity, and has the ability to infect non-dividing cells – a prerequisite for targeting some latently infected cell populations. However, AAV has a relatively limited cloning capacity, which poses a challenge for packaging larger rare-cutting endonucleases. In addition, the utility of some AAV serotypes can be limited by neutralizing antibodies generated upon pre-exposure to wild type AAV. The development of novel AAV delivery vectors is an active area of development in the lab.

Another method under development is the use of aptamers as delivery vehicles. Aptamers are nucleic acid molecules that can be produced with high binding affinity and specificity to an array of desired targets. They can also be utilized to deliver nucleic acid payloads, and we believe aptamers may give us the opportunity to specifically target latently infected cell populations with rare-cutting endonucleases. We are currently collaborating with Dr. Geoffrey Baird at the University of Washington and Dr. John Rossi at the Beckman Research Institute of the City of Hope to develop this technology.

Rare-cutting endonuclease technology has emerged as a viable tool for gene therapy applications in large part due to their ability to create DNA double strand breaks (DSBs) upon cleavage of their target sequence. The generation of DNA DSBs is required to induce the cell’s homologous recombination and nonhomologous end-joining repair pathways. These pathways in turn may be exploited for gene editing and in our case, targeted gene mutation and disruption. 

As shown above (see Our Tools), not all rare-cutting endonucleases share similar cleavage or target binding profiles. ZFNs and TALENS are artificial enzymes that utilize DNA binding domains that operate independently of their cleavage domains. They both use the non-specific DNA cleavage domain of the FokI type II restriction enzyme for cleavage and operate as heterodimers that generate 5’ overhangs upon cleavage of their target sequence. Unlike ZFNs and TALENS, homing endonucleases possess intrinsic DNA binding and cleavage activities, and generate 3’ overhangs upon cleavage of their target sequence.

When a DNA DSB is generated within a the genomic sequence, the nonhomologous end-joining repair pathway will precisely religate the DNA strands to restore the original sequence. However, NHEJ is known to be error-prone and in the presence of a rare-cutting endonuclease, successive cycles of cleavage and precise repair will continue at a target sequence until a mutation is introduced that prevents subsequent binding and cleavage. Due to the highly specific nature of these enzymes, a single nucleotide change in the target sequence is usually sufficient to halt binding and cleavage.

The impact that a given mutation will have on downstream protein synthesis or function depends on the nature of the sequence change. It is reasonable to expect that large insertions and deletions, while possibly leading to a coding sequence frame shift, will result in more deleterious effects on protein function. Conversely, the possibility of generating an in-frame deletion or insertion in which the coding sequence is left intact still remains. Whether such in frame mutations effect protein function will depend on the specific target sequence and the length of insertion or deletion.

In order to increase the frequency of mutagenesis during rare-cutting endonuclease treatment, we are collaborating with Drs. Andy Scharenberg and David Rawlings (NGEC co-directors) at Seattle Children’s Research Institute. They have utilized the 3’-5’ exonucleases Trex2 and Artemis to enhance the mutagenic potential of rare-cutting endonucleases. When DNA DSBs are generated in the presence of these enzymes single stranded overhangs are chewed back before NHEJ is able to repair the DNA DSB. This increases the frequency of mutation for ZFNs, TALENs and HEs.

By introducing mutations into essential viral genes we aim to disrupt the viral coding sequence and disable viral replication. We previously demonstrated successful targeting and disruption of an integrated reporter lentivirus using an engineered homing endonuclease (Aubert et al, PLoS One 2011), and we have continued building on this success by developing ZFNs, TALENS and HEs targeting essential genes from HIV, HBV, HSV and HPV. See our Lab Projects page for more details on our efforts to target chronic and latent viral infections with these rare-cutting endonucleases.