CRISPR/Cas9 has generated much excitement as a genome editing technology that is simple enough for very widespread use.

In contrast to zinc-finger nucleases (ZFNs), which are made up entirely of proteins, CRISPR – which is a primitive immune system that recognizes and cuts foreign DNA in bacteria – contains only a single protein, along with an RNA guide sequence for targeting.

The technology, Stanford University's Stanley Qi told BioWorld Insight, "has spurred so much interest. . . . It's really like PCR technology back in the day."

The technology was named one of the top scientific advances of 2013 by both Science and Nature, after researchers showed that despite its bacterial origins the technology could be used to edit mammalian cells. (See BioWorld Today, Dec. 20, 2013.)

Start-ups working in the field are Swiss CRISPR Technologies AG and Editas Medicine, in Cambridge, Mass., which went through series A rounds in 2014 and 2013, respectively. (See BioWorld Today, April 23, 2013, and Nov. 25, 2013.)

A cynic might note that the technology is also like PCR in being quite error-prone – far more so than ZFNs. (See BioWorld Today, June 30, 2013.)

OFF-TARGET EFFECTS

Qi acknowledged that the method is prone to off-target effects, but also noted that scientists are steadily working on understanding the reason for and nature of those effects in order to make the method more precise.

Last week, his own team, along with that of colleague Sheng Ding at the University of California at San Francisco, reported progress on another front: making the technology more efficient at replacing genes, rather than merely knocking them out.

As in many areas, in genetic editing destruction is easier than coming up with a constructive alternative.

Its simplicity is what makes CRISPR attractive – but that same simplicity also means that "after cutting, you have to rely on the host cell."

The DNA targeting "can be programmed, and it can be controlled," he said. "But what happens afterwards is uncontrollable."

The host cell, in turn, deals with the DNA-breaks CRISPR induces in one of two ways. One is nonhomologous end joining, or NHEJ, which essentially knocks out genes. Nonhomologous end joining works with high efficiency. Under some experimental conditions, more than half of cells have the target gene knocked out.

The other repair mechanism, homology-directed repair, or HDR, is a more precise one. The target gene is not just inactivated, but is replaced with another variant of itself, for example a point mutation.

HDR allows a greater variety of uses for CRISPR than NHEJ. In particular, if the technology is ever to be used for gene therapy, it will need to be able to replace a defective gene with a functional one.

For research, too, going beyond simple knockout to editing extends the use of the technology. Recently, scientists used CRISPR to model the EML-ALK fusion gene that is the molecular driver of some lung cancers, and the target of ALK inhibitor Xalkori (crizotinib, Pfizer Inc.). Another group has used the technology to gain new insights into KRAS-driven cancers.

Such precision edits, however, occur far less frequently than knockouts. While the efficiency of gene knockout using CRISPR can reach as high as 60 percent of treated cells, gene replacement is a rare event, and, Qi and his colleagues wrote in their paper, "a long and tedious screening process via cell sorting or selection, expansion, and sequencing is often required to identify correctly edited cells."

In their experiments, which were published in the Feb. 5, 2015, issue of Cell Stem Cell, Qi and his team set out to answer the question, "How can we enhance this precise pathway?"

"Before this, we always relied 100 percent on the cell" – a reliance, in fact, which could explain a lot of cell to cell variation in the results of applying CRISPR.

Qi and his colleagues looked for small molecules that might enhance the HDR pathway through high-throughput screening of mouse embryonic stem cells.

Of the several thousand chemicals they screened, four affected the efficiency of HDR – two increasing it, and the other two decreasing it.

Using those drugs, the efficiency for generating point mutations in treated cells was increased almost 10-fold, and for larger insertions, by about threefold.

The authors discovered an unexpected reciprocal relationship between the two repair pathways. The small molecules that enhanced HDR also inhibited NHEJ. The converse was also true: The two molecules that decreased the efficiency of HDR increased that of NHEJ, leading Qi to conclude that "the NHEJ and HDR pathways are mutually exclusive."

Intriguingly, both of the molecules that decreased HDR efficiency were antiviral drugs – one was the anti-HIV medication AZT, and the other, trifluridine (TFT), was first identified as a herpes drug.

So far, the efficiency is still low, and the authors have not yet deciphered the mechanism by which the molecules they have identified affect HDR and NHEJ. But it provides proof of principle that small-molecule strategies can be used to direct the cell's response to CRISPR's cuts.