A variety of labs are harnessing a recently uncovered natural defense mechanism found in bacteria and archaea as a genome editing tool. The CRISPR-Cas system protects prokaryotes from viral attack by sequence-specific cleavage of foreign nucleic acids, a process similar to eukaryotic RNA interference. But a growing concern regarding the CRISPR-Cas system is the potential for non-specific targeting , which could lead to mutations elsewhere in the genome.
In a study published August 1st in Nature Biotechnology, George Church of Harvard Medical School and his colleagues compared the target specificity of the CRISPR-Cas system in human cells with that of another established genome-editing tool based on transcription activator–like (TAL) effectors. Importantly, Church’s team also proposed a new method to reduce potential off-target effects when using CRISPR-Cas.
The CRISPR-Cas system works by incorporating short sequences of foreign genetic material into the host genome. These sequences are transcribed and processed into small RNAs, which together with the Cas protein complexes can bind to foreign DNA or RNA that matches the small RNA sequence. This process results in the sequence-specific cleavage and destruction of invading nucleic acids.
In their new study, Church and his team found that while TAL effectors tolerated 1–2 base pair mismatches in the target sequence, complexes consisting of a single-guide RNA and the protein Cas9 tolerated 1–3 mismatches, depending on the single-guide they used. “This is not a huge problem, but it can lead to some off-target effects,” says Luciano Marraffini, an expert in CRISPR interference at Rockefeller University.
To address this issue, Church’s group generated a series of Cas9 mutants that, instead of creating double-strand breaks, introduced offset nicks, each affecting only a single strand of DNA. While this strategy in essence creates double-strand breaks, it does not usually result in the insertions or deletions caused by non-homologous end joining—a cellular mechanism for repair of double-strand breaks. As a result, this approach could potentially reduce off-target effects.
Genome re-engineering using nuclease systems is not entirely new. Researchers previously had to engineer another genome-editing tool based on zinc finger nucleases to create nicks, also for the purpose of reducing off-target effects. And earlier this year, Marraffini and his collaborators reported that Cas9 can be converted into a nicking enzyme to reduce off-target mutations associated with DNA repair mechanisms.
However, it remains to be proven that Church’s approach actually does reduce off-target effects. “Church takes a first step toward a potential fix to this problem,” says Philip Gregory, chief scientific officer at Sangamo BioSciences. “But the paper lacks any assessment of specificity for this nick-induced double-strand DNA cut idea. This specificity question will be very important to address.”
Sangamo is currently conducting clinical trials to evaluate the use of zinc finger nucleases for the treatment of HIV/AIDs. While Church points out that the CRISPR-Cas system is “one hundred to one thousand times easier to do” than zinc finger nucleases and TAL effectors, Gregory remains unconvinced. “The CRISPR-Cas9 system, at least in its current format, is highly prone to substantial off-target activity that can exceed the level of cutting at the desired target site,” he says.
Despite these concerns, Church’s study has generated cautious optimism within the scientific community. “With the poor specificity the CRISPR-Cas9 system for genome editing has shown to date, despite the high efficiency, it was headed for the dump,” says Adam Bogdanove, an expert in TAL effectors at Cornell University. “The double nickase approach pulls it off the garbage truck just in time.”
Mali P, Aach J, Stranges PB, Esvelt KM, Moosburner M, Kosuri S, Yang L, Church GM. 2013. CAS9 transcriptional activators for target specificity screening and paired nickases for cooperative genome engineering. Nat Biotechnol. doi: 10.1038/nbt.2675.