Now that a new and supposedly more precise gene-editing technique has appeared, GMO promoters admit CRISPR is "clunky and prone to errors". Report: Claire Robinson

A new gene-editing technique has prompted excitement in pro-GMO circles, based on the premise that it could solve the problems of unintended effects and errors caused by the CRISPR gene-editing tool. But the excitement is premature, GMWatch has found, as the new technique is still prone to errors and there are doubts whether it will even be used to develop GM crops and foods.

The new technique, known as prime editing, was described in an article in the journal Nature (see abstract below). The authors describe prime editing as a "a versatile and precise genome editing method" that "substantially expands the scope and capabilities of genome editing, and in principle could correct about 89% of known pathogenic human genetic variants".

An article in the Guardian describes how David Liu and Andrew Anzalone at the Broad Institute of MIT and Harvard in Cambridge, Massachusetts, created prime editing by giving CRISPR gene-editing tool an overhaul. Instead of cutting the double strand of DNA, as in CRISPR, prime editing cuts only one strand, then introduces a new section of DNA into the specified position. The intention is to bypass the cell's own repair mechanisms, which, after the CRISPR gene-editing tool has made the initial double-strand cut in the DNA, can make unintended changes by inserting or deleting DNA sections at the cut sites.

While the Nature article focuses on applications of the technique for gene therapies, the Guardian quotes Robin Lovell-Badge, a promoter of deregulation of gene editing, as saying it could be applied to animal and plant genetic engineering.

Lovell-Badge doesn't waste the opportunity to use the new paper to support his lobbying message that gene-edited crops should be exempted from GMO regulations, stating his view that "Because the procedure can make changes that occur naturally in plants, at least some prime-edited crops should not be regulated as 'genetically modified'."

However, no one has ever proved that gene-editing can result in exactly the same changes, from the point of view of the genetic sequence, as natural breeding. So Lovell-Badge has departed from the evidence in making this statement.

New technique, old claims

What's fascinating about the hype over prime editing is that we've seen exactly the same claims about precision and safety from every GM technology in succession, followed by all-too-predictable realisations that it isn't so precise after all. In the 1990s it was transgenic genetic modification techniques that were supposedly more "precise" and "predictable" than natural breeding. But over the years, evidence built of the unpredictable effects of transgenesis. The first-generation GMOs foundered under a deluge of worrying findings in animal feeding studies, GM golden rice that grew stunted and malformed, herbicide-resistant superweeds, and pests that evolved resistance to GM insecticidal toxins.

Then when gene-editing techniques like CRISPR arrived, they were presented as the saviour of GM technology.

The GMO lobby and pro-GMO media argued that gene-editing techniques were precise compared with messy old transgenesis, that gene-edited crops and animals were the same as what could be found in nature, and that the techniques should be de-regulated. But even in the midst of these lobbying messages, a large body of evidence emerged showing that CRISPR is not precise after all, but results in unintended changes in the genome, both at off-target sites in the genome and at the intended editing site.

These problems are now being admitted to, even in GMO-boosting journals like Nature. An article published in the journal hyping the new study on prime editing begins, "For all the ease with which the wildly popular CRISPR–Cas9 gene-editing tool alters genomes, it’s still somewhat clunky and prone to errors and unintended effects. Now, an alternative offers greater control over genome edits — an advance that could be particularly important for developing gene therapies. The alternative method, called prime editing, improves researchers’ chances of getting only the edits they want, instead of a mix of changes that they can’t predict."

Only a matter of time

Of course, it's only a matter of time before problems will emerge with prime editing, too. Due to the fact that genes operate in networks and not as individual entities, you can't just tweak one or a few genes and expect there to be no side-effects.

Another technique, base editing, was previously touted as a way of introducing changes in genes while avoiding the unintended effects, such as large deletions or rearrangements, which can arise from DNA repair processes following the CRISPR-induced double-strand DNA break. But base editing turned out to introduce more unwanted mutations than expected in mouse embryos and rice plants.

Media hype

It's natural that the authors of the Nature article want to make their discovery appear in the best possible light. As the "competing interests" statement in the paper says, "Authors through the Broad Institute have filed patent applications on prime editing. D.R.L. [David Liu] is a consultant and co-founder of Prime Medicine, Beam Therapeutics, Pairwise Plants, and Editas Medicine, companies that use genome editing."

However, to their credit, they do not exaggerate their achievement in the new paper, nor do they downplay the problems with the technique. Molecular geneticist and gene therapy expert Dr Michael Antoniou commented:

"Prime editing is an elegant but complex technique involving multiple components and molecular manipulations. The authors emphasize that it needs a lot more work before it is ready for clinical use. There are still too many bugs and unknowns.

"They don't hype it, but allow the data to do the talking. The exaggeration is in the media coverage. The authors acknowledge current limitations and knowledge gaps. For example, they state that the technique produces both off-target and on-target undesired effects. They found a lower number of off-target effects with prime editing than with standard CRISPR tools, but they were not comparing it to the later generation higher-fidelity CRISPR systems that already have markedly reduced off-target effects. Instead they compared it to first-generation CRISPR-Cas9, which is highly prone to off-target cuts, and found that it was far superior. But whether this will still be the case when comparisons are made with more advanced higher fidelity CRISPR systems remains to be seen.

"In addition, although the frequency of off-target editing by prime editing was much lower than when they used CRISPR-Cas editing tools to bring about the same editing event, they nonetheless did see off-target edits from the prime editing tool in every situation they tested.  

"These off-target effects varied with the prime editing tool used and with the gene being targeted. In three cases they were low in number (0.1% of the known off-target sites or lower), but in one case, 2.5% plus or minus 5.2% (so up to 7.7%) of the known off-target sites were mutated.

"Also, to a variable degree, depending on the prime editing tool used and the gene being targeted, indels (unintended insertions or deletions of small stretches of DNA) were always found. The frequency of on-target indel formation was between 0.3% and 5% of the total edits made, where the authors attempted to correct a genetic mutation in the targeted gene.

"The authors acknowledge a major knowledge gap, in that they only looked at known off-target sites for the genes they tried to edit. They did not do whole genome analysis to identify the full spectrum of off-target outcomes.

“However, they achieved good efficiency of editing: in some cases over 50% of the targeted sites were edited in the desired manner. But because of the small yet significant degree of off-target mutations, as well as the on-target indel formation, prime editing is not ready for prime time gene therapy applications, as acknowledged by the authors. Its precision needs to be significantly improved before it can be considered for clinical use.

"Why? Let's say you are undertaking a gene therapy targeting bone marrow stem cells, e.g. for sickle cell disease. This will involve prime editing millions of cells and returning them to the patient. Even a low off-target frequency or on-target indel formation could result in thousands of cells being returned to the patient that are carrying an undesirable mutation. In a worst case scenario, if the function of an oncogene, which regulates normal cell growth, is affected, this could progress to a cancerous state.”

Some problems avoided but not all

Past articles on GMWatch have raised the issue of on-target unpredictable outcomes caused by the DNA repair process known as non-homologous end joining (NHEJ). NHEJ gets activated in order to repair the cut ends of the DNA molecule once the CRISPR has made its double-strand DNA break.

Dr Antoniou explained, “Prime editing does not make double-strand breaks and thus does not activate the NHEJ repair pathway. So some types of on-target undesirable outcomes, such as large deletions and rearrangements of DNA, are avoided. However, since prime editing in its current form still results in indel formation at the edited site, albeit at a low frequency, it can result in altering the sequence of the targeted gene in such a way that it produces new mRNA (messenger RNA) and proteins, with unknown consequences.

"If prime editing were applied to plants, in all likelihood the same undesirable outcomes at both on-target and off-target sites, as described in the new publication, will occur. These would be over and above the genome-wide mutations caused by the tissue culture and the GM transformation process, which would be used to introduce the prime editing tool into the plant cells.

"In conclusion, while prime editing's potential for gene therapy is great, it is questionable whether it offers anything to the plant biotechnology world. It solves some problems, but not all."

GMWatch would add that it is also questionable whether plant biotechnologists will bother to use this complex gene-editing technique. It seems likely that they will continue to use the less precise standard CRISPR editing tools involving the formation of double-strand DNA breaks – and that they will continue to simply discard those plants showing undesirable off-target mutational outcomes.

The problem with this approach is that plant genetic engineers are mostly concerned with ensuring that the plant grows well and looks healthy. They do not subject the plant to the necessary tests to ensure that the editing event has not led to the production of toxins or allergens. Neither do they ensure that the plant will be safe for wildlife during cultivation.

So it would be premature, as well as irresponsible, to use prime editing as a reason to support the lobbying push for de-regulation of gene editing in food plants and livestock animals.
---

The new study:
Search-and-replace genome editing without double-strand breaks or donor DNA
Andrew V. Anzalone, Peyton B. Randolph, Jessie R. Davis, Alexander A. Sousa, Luke W. Koblan, Jonathan M. Levy, Peter J. Chen, Christopher Wilson, Gregory A. Newby, Aditya Raguram & David R. Liu
Nature (2019)
https://www.nature.com/articles/s41586-019-1711-4

Abstract

Most genetic variants that contribute to disease1 are challenging to correct efficiently and without excess byproducts2–5. Here we describe prime editing, a versatile and precise genome editing method that directly writes new genetic information into a specified DNA site using a catalytically impaired Cas9 fused to an engineered reverse transcriptase, programmed with a prime editing guide RNA (pegRNA) that both specifies the target site and encodes the desired edit. We performed more than 175 edits in human cells including targeted insertions, deletions, and all 12 types of point mutation without requiring double-strand breaks or donor DNA templates. We applied prime editing in human cells to correct efficiently and with few byproducts the primary genetic causes of sickle cell disease (requiring a transversion in HBB) and Tay-Sachs disease (requiring a deletion in HEXA), to install a protective transversion in PRNP, and to insert various tags and epitopes precisely into target loci. Four human cell lines and primary post-mitotic mouse cortical neurons support prime editing with varying efficiencies. Prime editing offers efficiency and product purity advantages over homology-directed repair, complementary strengths and weaknesses compared to base editing, and much lower off-target editing than Cas9 nuclease at known Cas9 off-target sites. Prime editing substantially expands the scope and capabilities of genome editing, and in principle could correct about 89% of known pathogenic human genetic variants.