Gene Editing: From Roots to Riches

Advances in genetic manipulation have simplified the once daunting task of rewriting a gene.

Oct 1, 2016
Amanda B. Keener


In the mid-1980s, Oliver Smithies, then at the University of Wisconsin–Madison, and Mario Capecchi of the University of Utah independently used homologous recombination—a molecular process to repair broken DNA—to change specific regions of the genome in cultured mouse cells (Nature, 317:230-34, 1985; Cell, 44:419-28, 1986). The technique involved sandwiching an altered copy of a gene between two regions of code identical to those flanking the endogenous gene, which would be swapped out for its engineered counterpart.

But Capecchi and Smithies couldn’t introduce genetic changes into living animals until Martin Evans, now of Cardiff University in the U.K., established a method for culturing mouse embryonic stem cells (ESC). Only ESC or cancer cells could be kept in culture long enough to produce enough genetic material to confirm that homologous recombination had taken place. ESCs also provided a way to create animals harboring genetic modifications, allowing researchers to ask the question, “What does this do in something that runs and smells?” says Dirk Hockemeyer, a stem cell biologist at the University of California, Berkeley.

In 1987, Capecchi reported the targeted disruption of a wild-type gene and Smithies reported the targeted correction of a mutated gene in mouse ESCs. In a series of publications, they brought gene editing into mice, marking the first time anyone had bred a genetically edited animal. Smithies, Capecchi, and Evans shared the 2007 Nobel Prize in Physiology or Medicine for their work.

But the process wasn’t easy, says Thomas Doetschman, a molecular biologist at the University of Arizona in Tuscon who was a postdoc in Smithies’s lab when the group developed the technology. To obtain a physical copy of a gene of interest required creating a genomic library—a chopped-up mouse genome whose pieces were housed within thousands of bacteriophages, which were cultured inside E. coli on oversized petri dishes. “Once you made your library spread out on those plates, those were hot items,” Doetschman recalls. Researchers would save the plates in the freezer for years and probe them to find phage-infected bacteria carrying genes of interest. Then they would screen the library, grow up the appropriate bacterial clone, isolate the gene, and assemble a genetic construct—called the targeting vector—which would be transfected into mouse stem cells using electroporation. After confirming the event by PCR or Southern blot, researchers still had to inject ESCs into blastocysts and wait for the mice to cycle through a couple of generations to determine if the cells carrying the editing genome had been adopted.

1987: Emw/Wikimedia Commons; 1990:; 1991: NCBI; 2010: David Goodsell/Wikimedia Commons; 2012: molekuul_be/; 2014: Thomas Splettstoesser/Wikimedia Commons

Much of this grunt work became obsolete in the late 1980s and early 1990s, as the invention and automation of PCR converged with the sequencing and sharing of the entire mouse genome. But the editing process was still inefficient. Homologous recombination requires some sort of DNA synthesis or repair process to occur, making the technique imprecise and hard to direct. As a result, only 1 to 10 cells in a million would pick up the vector and swap out the DNA in the correct spot. In 1994, developmental biologist Maria Jasin and her team at the Sloan Kettering Institute in New York found a way to increase the rate of homologous recombination events 10- to several thousand-fold. The trick was to use an endonuclease that creates double-strand breaks in a unique sequence, usually 4–8 base pairs in length, called a restriction site (PNAS, 91:6064-68, 1994). Inducing cuts forces the cell to ramp up its repair through homologous recombination. Jasin’s team inserted known restriction sites at random spots in the genomes of mouse fibroblasts and used a yeast endonuclease to induce homologous recombination at those sites (Mol Cell Biol, 14:8096-106, 1994).

But to make double-strand breaks around a specific gene—to swap out that copy with another version—required that restriction sites flank that gene. Three years earlier, Nikola Pavletich and Carl Pabo, then at Johns Hopkins University, had already published the potential solution, visible in the crystal structure of DNA-binding proteins containing so-called zinc fingers. Pavletich and Pabo found that zinc finger domains each bind a particular sequence of three bases and can be mixed and matched to target desired stretches of DNA (Science, 252:809-17, 1991). In 2001, researchers successfully combined an engineered zinc finger protein and a nuclease to make targeted cuts in the DNA of Xenopus oocytes (Mol Cell Biol, 21:289-97, 2001). Combining this technique with Jasin’s work, researchers could not only make double-strand breaks, but make them anywhere in the genome. “Those two things together are what started gene editing,” says Charles Gersbach, a molecular biologist at Duke University.

Advancements, accelerated

Zinc finger nucleases (ZFNs) did not catch on like wildfire, however. They were tricky to work with and required an advanced understanding of protein structure and a lot of trial and error, says Raj Chari, a postdoc in geneticist George Church’s lab at Harvard University. “Typically, just to get the reagent to even start the experiment, it would take months.”

Gersbach’s lab was two years into a project to repair a mutated form of the dystrophin gene using ZFNs when, in 2011, one of his graduate students asked to switch to the recently developed transcription activator-like effector nuclease (TALEN) system. Gersbach discouraged him, saying that it was probably not as easy as it looked and that it was wisest to see the zinc finger project through. But the student didn’t listen. “Two months later he came back in and said, ‘I’ve got it all working in TALENs now,’” Gersbach recalls.

Like zinc fingers, TALEs are DNA-binding proteins that can be attached to a nuclease (creating a TALEN), but each TALE subunit binds to only a single nucleotide, making them much easier to string together to target specific sequences. “Most of the ZFNs that we made didn’t work, and most of the TALENs did,” says Gersbach.

When CRISPR/Cas9 became available two years later, Gersbach’s students again wanted to switch to the newer technology, and again, he discouraged them. “Luckily, they don’t listen to me,” he says. In December 2015, his team was one of three that described CRISPR/Cas9-mediated editing of the gene for dystrophin in neonatal and adult mice (Science, doi:10.1126/science.aad5143, 2015).

CRISPR/Cas use has recently exploded. In 2012, just 126 publications indexed by Pubmed mentioned the technology; nearly 10 times as many came out in the first six months of this year. The new approach has spurred international discussion about the ethics of human gene editing and high-profile patent disputes between competitors and among collaborators illustrates just how crucial this gene-editing technology is to life science. Above all, it has changed the game of genome tinkering, having already demonstrated potential to edit DNA in cell lines, embryos, and even mice.

“When CRISPR came along it was pretty clear that was going to make life a lot simpler,” says Doetschman. Nowadays, when he wants to create a mouse line with an altered gene, in most cases he can find a guide RNA to target the CRISPR proteins to his gene of choice and inject plasmids encoding the RNA and a Cas nuclease directly into a fertilized mouse egg, skipping the in vitro ESC work. “It’s [nothing] short of miraculous. It’s hard to believe how efficient [CRISPR] is.

Read about how technological advances over the last three decades have enabled research in other life science fields in The Scientist’s special 30th anniversary feature.

July/August 2019

On Target

Researchers strive to make individualized medicine a reality


Sponsored Product Updates

DNASTAR® announced the release of Lasergene 16 Software
DNASTAR® announced the release of Lasergene 16 Software
DNASTAR® announced the release of Lasergene 16 today, which includes a broad range of improvements in for analysis of DNA, RNA and protein sequence data, as well as new advancements for predicting and analyzing protein structures. 
Arbor Biosciences Partners with Curio Genomics for Analysis of IWGSC Wheat Exome
Arbor Biosciences Partners with Curio Genomics for Analysis of IWGSC Wheat Exome
Arbor Biosciences, a division of Chiral Technologies, Inc and worldwide leader in next generation sequencing (NGS) target enrichment, announces a partnership with Curio Genomics for bioinformatics analysis of the wheat genome.
IDT and Washington University join forces to increase access to the latest NGS technologies
IDT and Washington University join forces to increase access to the latest NGS technologies
As part of its commitment to advocate for the genomics age, Integrated DNA Technologies (IDT) aims to lower the barriers to access the latest NGS technologies.
Bio-Rad Launches Bio-Plex Pro Human Immunotherapy Panel 20-plex Multiplex Assay, a targeted tool for researching signaling networks in Immunotherapy Research
Bio-Rad Launches Bio-Plex Pro Human Immunotherapy Panel 20-plex Multiplex Assay, a targeted tool for researching signaling networks in Immunotherapy Research
Bio-Rad Laboratories, Inc. (NYSE: BIO and BIOb) July 15, 2019 announced the launch of its Bio-Plex Pro Human Immunotherapy Panel 20-plex, a multiplex immunoassay that offers a targeted approach for Immunotherapy Research.