Jeffrey Perkel has been a scientific writer and editor since 2000. He holds a PhD in Cell and Molecular Biology from the University of Pennsylvania, and did postdoctoral work at the University of Pennsylvania and at Harvard Medical School.
Just a few short years since it burst on the scene, CRISPR’s star remains as bright as ever.
Over the past year, the technology has been profiled in The New Yorker, The New York Times Magazine and Wired. Researchers using CRISPR/Cas9 have advanced fields ranging from paleobiology to organ xenotransplantation. At the same time, ethical debates have swirled as reports emerged of edited human embryos and gene drive systems for controlling mosquito populations (for malaria disease and more recently the Zika virus outbreaks).
Yet CRISPR (“clustered regularly interspaced short palindromic repeats”)/Cas9 is not the first genome-editing technique; others include zinc finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs). Though they differ in the details, all three methods produce the same basic outcome: The introduction of a double-stranded DNA break at a defined location in living cells, the repair of which can either knock out or rewrite a gene sequence.
Here we round up some of the latest developments and tools for these systems.
CRISPR rising
How big is CRISPR/Cas9? More than 1,700 papers have been published on the technique since it was worked out in 2013, according to PubMed. But according to George Church, professor of genetics at Harvard Medical School and one of the technique’s pioneers, that likely underestimates its penetrance in the biological sciences. “It takes a while to do the experiments and get them published,” he notes.
A better indicator might be the popularity of CRISPR reagents from Addgene, the low-cost reagent repository. According to executive director Joanne Kamens, Addgene has hundreds of zinc finger reagents, nearly 2,000 TALENs and about 1,400 CRISPRs. Over the years, the company has distributed hundreds of ZFNs and TALENs, she says, but CRISPR dwarfs them both: “We have shipped 60,000 plasmids related to CRISPR technology,” Kamens estimates. Indeed, Addgene’s most popular plasmid of 2015 was lentiCRISPR V2, a lentiviral vector encoding both Cas9 and a gRNA, developed by Feng Zhang. (The firm’s most popular TALEN tool is the Golden Gate TALEN and TAL Effector Kit 2.0.)
There are several reasons for CRISPR’s popularity. One is simplicity. ZFNs, TALENs and Cas9 are all user-programmable DNA-cutting nucleases. But ZFNs and TALENs are targeted based on the selection and arrangement of DNA-binding protein domains—a process that can require considerable cloning and optimization time. In contrast, Cas9 needs only a short guide RNA (gRNA) complementary to the target of interest. That makes the technology fast, inexpensive and easy to use.
CRISPR/Cas9 also works across a wide range of organisms and can be directed toward multiple targets simultaneously simply by adding more than one guide RNA at a time.
Finally, it is easily configurable, with researchers exploring different Cas9 designs to improve target-site specificity, control gene expression and more.
When starting new experiments, there are “really not many reasons to use anything besides Cas9 now,” says Brett Robb, scientific director at New England Biolabs, which sells recombinant Cas9, a cloning-free guide RNA transcription kit and a mutation-detection kit for assessing editing efficiency.
Building a better Cas
That’s not to say the technology is static. In the past year, researchers have identified several Cas9 alternatives, including Cpf1 and NgAgo. Cpf1, says Eric Rhodes, chief technology officer at Horizon Discovery, a company that sells genome-edited cell lines and services, features a more AT-rich protospacer-adjacent motif (PAM) sequence than Cas9, making it potentially useful for cases in which the traditional Cas9 PAM (5’-NGG) won’t do. And NgAgo, guided by a 24-base DNA guide sequence (rather than RNA), exhibits certain properties that may make it less susceptible to off-targeting than Cas9.
Despite their differences, Cpf1 and NgAgo operate by the same mechanism as Cas9, TALENs and ZFNs. As a result, says Church, they suffer the same fundamental shortcoming: Because they produce a double-stranded DNA cut, genome editing becomes a race between competing DNA repair mechanisms, and multiple outcomes are possible. “What some call genome editing, I call genome vandalism,” he says. “You get a mess.”
One possibility is avoiding the double-strand break altogether. Homologous recombination using recombinant adeno-associated viruses (a technology commercialized by Horizon Discovery) works this way, as does a method developed in Church’s lab called MAGE (multiplex automated genome editing). In April, David Liu at Harvard University described another approach, fusing catalytically inactive (i.e., “dead”) Cas9 with cytidine deaminase to produce a variant capable of precisely rewriting C to T or G to A.
In a recent report describing the work, Church called Liu’s new design “arguably the most clever CRISPR gadget to date.” The plasmids encoding that design, like those for Cpf1, are available through Addgene, and NgAgo should be available shortly, says Kamens. “We are waiting for the material transfer agreement to be complete.”
Researchers have used dead Cas9 to target other activities to the genome, as well. Several, for instance, have created programmable transcriptional activators. One, from Charles Gersbach at Duke University, couples the histone acetyltransferase p300 to dCas9 for programmable “epigenome editing”—a construct that has been commercialized by MilliporeSigma.
New tools
Researchers can deliver Cas9 to cells on plasmids, as mRNA or as purified protein complexed with gRNA. Of the three, “the protein/RNP complex is by far the most efficient,” says Jon Chestnut, a senior director at Thermo Fisher Scientific, which offers reagents in all three formats. It also appears to impact fewer off-target sites, he says, likely because the RNP complex doesn’t last very long in the cell. “You’re getting this transient delivery of nuclease, which should result in lower off-target cleavage, because the nuclease isn’t hanging around looking for something to do.” (Cas9 mRNA is also labile and therefore also offers this benefit.)
The RNP also appears to provide faster editing than DNA, says Chris Linnevers, product manager at Bio-Rad Laboratories, presumably because it’s ready to go—no transcription or translation is required. “Instead of 24-plus hours, you can get editing in a matter of hours,” he says.
Those researchers wanting the flexibility to deliver any type of molecule can try electroporation, for instance using Bio-Rad Laboratories’ Gene Pulser, Thermo Fisher Scientific’s Neon system or Lonza’s Nucleofector. According to Gregory Alberts, global subject matter expert at Lonza Pharma Bioscience Solutions, electroporation is largely agnostic with respect to molecular class, delivering DNA, RNA or protein with more or less equal efficiency. Thus, it can handle, for instance, the Cas9 ribonucleoprotein complex, plus a single-stranded DNA repair template. “When we open these pores, whatever material is there will passively diffuse into the cell,” he says.
One recent study used Bio-Rad’s Gene Pulser Xcell to deliver Cas9 ribonucleoprotein and a single-stranded repair template directly into mouse zygotes, a far simpler method for targeted genetic modification than the usual approach of microinjection. The method, which the authors dubbed CRISPR RNP Electroporation of Zygotes (CRISPR-EZ), yielded editing in 33 of 33 zygotes. “This was a fascinating finding, and it will speed up and make it easy to do this kind of workflow, to create transgenic models,” says Brandon Williams, a cell biology product manager at Bio-Rad.
Application-specific delivery tools also are available. Thermo Fisher Scientific’s new CRISPRMAX transfection reagent can deliver an intact Cas9-guide RNA ribonucleoprotein complex to cells, for instance. Similarly, MTI-GlobalStem will soon launch a line of transfection reagents called EditPro; the reagents will be able to deliver DNA, RNA or protein, or any combination thereof—enabling, for instance, delivery a Cas9/gRNA complex with a DNA donor template. (The first entry in that line, EditPro-Stem, is optimized for pluripotent stem cells and is set to launch in June, with a general formulation for adherent human primary cells and cancer cell lines to follow shortly thereafter.)
Also new in the genome-editing toolbox are prefabricated CRISPR libraries. MilliporeSigma and the Sanger Institute have created an arrayed whole-genome gRNA library, with two guides per gene for each gene in the human and mouse genomes. According to Shawn Shafer, functional genomics market segment manager at MilliporeSigma, this library—which he calls “the first arrayed CRISPR whole-genome product on the market”—could inject some much-needed standardization into the field, enabling researchers to compare results between experiments and across labs. “You know you’re targeting the exact same sequence,” he explains.
Thermo Fisher Scientific is developing both pooled and arrayed libraries of lentiviral CRISPR vectors, says Chestnut. “We have finished the human kinome [of] 840 genes, and we’ve done GPCRs as well,” he says. The company plans to complete a “druggable genome” CRISPR set by the end of 2016, and the full human genome by early 2017.
Making a decision
Since ZFNs, TALENs and CRISPRs all edit via the same mechanism, the question is, which to use? All else being equal, most agree that CRISPR is usually the best bet. And it’s certainly the easiest tool to try.
If off-targets are a concern, take a page from siRNA-based screens and use multiple gRNAs and test several independent clones to ensure that any observed phenotypes really are due to the targeted change. You also can blend editing techniques. For instance, Shafer at MilliporeSigma recommends using CRISPR/Cas9 for quick-and-dirty experiments and then following up with ZFNs or TALENs for validation and generating stable cell lines.
Or, you can simply outsource the work to companies like Thermo Fisher Scientific and Horizon Discovery. “We’re agnostic about the technologies,” says Horizon’s Rhodes. But as the technologies have matured, so too have the projects. Rarely will researchers ask for a simple knockout anymore, he says—they now can make those themselves. “What we are seeing more of is the sophisticated things,” like knock-ins and multisite modifications.
As genome editing continues to mature, expect such projects—and the tools to support them—to follow suit.