The term “genome-editing” has become synonymous with CRISPR-Cas. But there’s more than one way to edit a genome, and each has its pros and cons.
CRISPR-Cas uses a short RNA molecule complementary to a specific segment of DNA to target a DNA-cutting enzyme to that same region. The result is a so-called double-strand break, which the cell repairs via either non-homologous end joining (NHEJ) or homology-directed repair (HDR), either knocking out, repairing, or replacing a sequence in the process. (Zinc-finger nucleases and TALENs also create double-stranded breaks, only without the convenience of an RNA guide molecule.)
You don’t need me to tell you that approach has been wildly successful. But cells don’t particularly like double-strand breaks. And both their placement and correction are hard to control, leading sometimes to unintended changes. Thus, as a practical matter, there’s a limit to how many genomic modifications researchers can make at a time, making the technique less useful for large-scale genetic changes.
An alternative strategy, published by Harvard geneticist George Church and his team in Nature in 2009, is MAGE (multiplex automated genome engineering). Here, pools of oligonucleotides, each targeting a specific genomic locus, are introduced into a bacterial cell at once. These oligos direct genomic rewrites at defined locations during DNA replication with high efficiency, over 30%. But changes do not occur at every targeted locus at the same time. Thus, with each round of replication, only a subset of changes are made, allowing researchers to explore (by DNA sequencing) the fitness of the ‘genomic landscape’ over time.
The process supports considerably higher multiplexing than CRISPR-Cas, not to mention a degree of genetic exploration that is otherwise difficult to achieve. The 2009 study, for instance, used a pool of oligonucleotides targeting 24 genes to tune the production of lycopene. The top-performing strains contained combinations of gene tweaks and knockouts that might have been impossible to predict beforehand, notes Farren Isaacs, who was a postdoc in Church’s lab at the time and coauthored the MAGE study.
Now, in a study published 16 November in Cell, Isaacs and his team at Yale University have extended the process to the budding yeast, Saccharomyces cerevisiae. The efficiency of the new method, called eukaryotic MAGE (eMAGE), is about 40%.
“What we basically showed is that if you design your experiment appropriately, this should work in any yeast strain,”, says Isaacs. “There might be some engineering that you might have to do, but in principle it should work.”
And not just in yeast, Isaacs adds. “I believe this serves as a blueprint to implement in other eukaryotic systems, from other single-cell eukaryotes to multicellular eukaryotes.”
Much of the study, a ‘Herculean effort’ led by graduate student Edward Barbieri, involves working out the mechanism underlying eMAGE — untangling elements of DNA replication, homologous recombination, and DNA repair. (If you’re interested, the authors write, “This approach is rooted in biasing the annealing of synthetic [single-stranded oligonucleotides] during DNA replication rather than a Rad51-directed strand-invasion mechanism.”)
But as in 2009, the final portion of the study again targets a biosynthetic pathway, the production of beta-carotene (the orange color in carrots). Here, only four genes were targeted. But the team explored a richer set of possible mutations, using 74 oligos to target transcription factor binding sites, transcription start sites, polyadenylation sequences, and more — some 482 base positions in all.
The result is a spatter of colonies in colors from beige to dark orange. By sequencing 39 of the resulting strains and comparing them to the wild-type, the team was able to work out which genetic changes produced which phenotypes — information that should prove useful in future pathway engineering exercises, Isaacs says.
“The point here is really being able to demonstrate the very rapid and precise ability to generate lots of combinatorial genetic variants. And then be able to then study how each of those impact the expression of the different carotenoids, and tease out driver mutations that govern differences in the expression of those metabolites,” he says.
eMAGE is not yet commercially available, Isaacs says, though the reagents are easy to come by. But the article does summarize six key lessons for those who would like to try the method on their own. According to Isaacs, they can basically be reduced to two: Use an oligo that will flip a selectable marker near the replication fork, and make sure the remaining oligos target sequences within about 20-30 kb of that event.
eMAGE, Isaacs says, is not a replacement for CRISPR and its analogs, but rather their complement. Researchers could, for instance, use CRISPR-Cas and DNA synthesis technologies to insert a biosynthetic pathway into cells, and eMAGE (perhaps coupled with David Liu’s recently described base-editing technology) to tune its activity.
“I view all of those as being essential in really enabling the era of genome design and engineering.”