The obvious approach to assessing off-target effects is simply sequencing the modified genomes. That's precisely what Chad Cowan at the Harvard Stem Cell Institute, and colleagues did after using TALENs to generate mutations in human induced pluripotent stem cells. Sequencing six exomes, they identified two indels other than the expected ones, but lack of homology to targeted sequences suggests these sites are unrelated to the TALEN, the team says. The team also performed whole-genome sequencing on the same samples, using a computer algorithm to focus only on likely off-target sites. Of those 100,000 sites, none appeared mutated. “We conclude that off-target indels rarely occur,” the authors wrote.(4).
Such studies remain to be done with CRISPR/Cas9 systems, but a recent report suggests they are fairly precise, at least in bacteria. Working with Doudna and UCSF/HHMI investigators Wendell Lim and Jonathan Weissman, UCSF fellow Stanley Qi generated a form of Cas9 that can bind RNA and target DNA, but cannot cut it, acting in effect like a programmable version of the lac repressor. “It's a way of parking a big protein on a piece of DNA such that it essentially blocks the passage of RNA polymerase,” Doudna explains. Whole transcriptome sequencing of these cells failed to identify any genes, other than the specific target, whose expression was downregulated.
“At least in bacteria, this is an incredibly sequence-specific process,” Doudna says.
And it is one that will likely develop over the next few months, she adds, which should definitely help to define the system's potential and limitations. “It's a really great example of how fundamental basic research, which was not aimed at any particular target or goal or certainly a particular application, led to the discovery of a system that may turn out to be a really transformative technology for genome engineering.”
Besides Doudna and Lim's work on a Cas-based repressor, Joung has coupled TALE binding domains to transcriptional activation domains to create highly active “TALE activators.” Robert McKnight at the University of Utah is building TALE methyltransferase fusions (TALEMs) to manipulate the epigenome one CpG at a time, and imagines other applications to probe hydroxymethylation, histone modification, and more. “The potential is tremendous,” he says.
Fortunately, researchers who wish to tap into that potential needn't reinvent the wheel. Voytas, Joung, and Doudna, among others, have made their TALEN and CRISPR/Cas reagents available at AddGene. And Ekker's lab recently launched an online TALEN-design tool called Mojo Hand (www.talendesign.org). Looks like its time to clear some space in your toolbox.