CRISPR hits the mark

Or: A precautionary tale about the randomness of natural mutations

Translated and adapted from TransGen

Plants developed with CRISPR gene editing ‘scissors’ must be regulated just as strictly as transgenic plants. That’s at least what many environmental and consumer organizations, as well as many (German) Green and Social democratic politicians demand. They point to the precautionary principle for support. But they are applying double standards.

In contrast to traditional genetic engineering leading to transgenic plants, as well as to conventional breeding, the gene scissors CRISPR / Cas and other similar genome editing processes allow you to "rewrite" individual DNA building blocks. This happens not just anywhere in the huge genome of a plant, but at a predetermined location that the researchers know is crucial for a certain trait, such as susceptibility to a viral disease. This location is exactly where the DNA strand is cut and then repaired with minor deviations. This is the same process as every random mutation in nature.

CRISPR triggers a mutation, too. Not randomly somewhere in the vastness of the genome, but in the predetermined target sequence.

But as precise as the gene scissors may be, it cannot be ruled out that the DNA strand is not only cut at the respective target location, but also in other, unknown places. Such unintentional incorrect cuts are referred to as off-target effects.

We do not know exactly where these occur, how properties change as a result, and what negative consequences this could have for the environment and health. Therefore, some critics argue that a "strict interpretation of the precautionary principle" is required. Plants developed and bred using CRISPR and other genome editing methods would therefore have to be regulated just as strictly as genetically modified ones. They are prohibited unless the manufacturers can provide proof of their safety in the framework of an approval process.

But are off-target effects really as numerous and relevant to safety as the critics suggest?

A working group at the Julius Kühn Institute (Institute for the Safety of Biotechnological Processes in Plants, Quedlinburg, Germany) has systematically evaluated more than a thousand publications on genome editing applications in plants, mainly from basic research. 252 of these publications contained information on possible off-target effects; most (approx. 90 percent) used the CRISPR method.

The likelihood of off-target effects depends primarily on whether the molecular probe - for CRISPR the guide RNA - not only docks onto the matching target sequence, but also at other locations in the genome. If this is the case, the gene scissors will cut in the wrong place, which could negatively change the properties of the plant.

Bioinformatic programs can help identify sequences in the plant genome in which the risk of incorrect cuts and unintended mutations is high due to similarity to the target sequence. It is precisely these sensitive areas that can be more closely examined to determine whether this actually occurred (biased detection). This was the case in about three percent of the possible off-target areas, according to the JKI working group.

Some of the research projects examined not only potential off-target zones, but the entire genome of the edited plant (unbiased detection). No unintentional changes could be found. However, the number of these publications was too small to be able to derive reliable results from them.

Nevertheless, the longer the target sequence chosen, the fewer similarities it will have to other DNA sections and so the less likely it is that off-target effects will occur. By appropriately tailoring the respective target sequence and the corresponding design of the probe scientists can drastically reduce incorrect cuts.

However, unlike in transgenic genetic modification, off-target effects in genome editing are simply mutations. And these are nothing unusual in conventional plant breeding. On the contrary: from one generation to the next, a plant’s genetic material contains one mutation in 150 million base pairs (DNA building blocks). In potatoes there are about six such spontaneous, random changes per generation while wheat with its huge genome has up to 200.

In mutation breeding, which has been used in various crops for many years, these natural mutation rates are increased approximately 700-fold by chemicals and radiation.

The details of such "natural" mutations or mutations amplified by artificial stimuli, such as their locations in the genome or the effects that they cause, are usually not known. Unless the mutations are reflected in the physical appearance (phenotype) or are quantifiable, they are ignored. Countless mutations that do not have any recognizable disadvantage for the plant are thus included in conventional breeding processes.

This has never been a problem so far, even if properties - such as taste - are sometimes lost in the course of breeding or there is a possibility that a plant may suddenly develop allergens as a result of spontaneous mutations.  No one demands that the precautionary principle be applied to conventionally bred plants, or that there is complete certainty about all mutations that occur.

It's different with genome editing. Random mutations in conventional breeding, which have never been a subject of concern, are now called off-target effects and are a matter of great worry.

The molecular biological mechanism - cutting in the DNA strand and subsequent repair - is exactly the same. However, off-target events in genome editing are rare and largely avoidable, while mutations in breeding are not.