Keywords

I would be more worried about pesticides and chemicals that are used than about a simple mutation, which if allowed enough time and enough resources might exist anyways.—Informant 18 (research scientist/NGO representative)

1 Introduction

The concept of editing a genome with the aid of a programmable nuclease emerged as part of advancements in functional genomics over 20 years ago (National Academies of Sciences & Medicine (NASEM), 2020). This field is focused on understanding the relationship between the information contained in an organism’s genome and its physical characteristics. All genome editing techniques rely on the single step of engineering an enzyme (i.e., the nuclease), that induces a Double-Strand Break (DSB) at a specific site of the DNA that is to be edited.Footnote 1 Unlike genetic modification, gene editing does not use foreign nucleotides to induce change in DNA. Instead, it harnesses natural repair processes found within the cell.

In almost every discussion of gene editing in agrifood, a description of various gene editing techniques is included. Gao’s (2018) piece in the scientific journal Cell compares gene-editing techniques in detail, as does the FAO’s comprehensive Gene Editing and Agrifood Systems document (FAO, 2022). Explaining the details of different techniques used to perform gene editing in language accessible to non-experts provides readers with accurate information about what gene editing is, what it is not and what it can and cannot do.Footnote 2 This chapter attempts to do just that. Briefly, explain gene editing techniques to better understand the dynamics of governance and regulation surrounding this groundbreaking technology as it is used in agrifood systems.

2 New Breeding Techniques

The most advanced New Breeding Techniques (NBTs) alter an organism’s genome by using gene editing and can be used in several different ways. They can be used to replace a disease-causing mutation sequence with a normal sequence. NBTs can also be used to disrupt an expressed gene by turning it off.Footnote 3 The genome is altered by targeting nucleotides via types of variants. This can be done by using one of two techniques: Oligo-Directed Mutagenesis (ODM) or Site-Directed Nucleases (SDN) (Lassoued et al., 2021). ODM is a rapid, precise, non-transgenic plant breeding alternative that uses synthetic oligonucleotidesFootnote 4 that are like DNA molecules with the target sequence (homologous) but for the nucleotide(s) to be modified. Oligonucleotides target the homologous sequence and create a mismatch at the base pair that is to be modified. This mismatch is recognized by the DNA repair machinery of the cell and the mismatch is repaired introducing the altered nucleotide.

2.1 Site Directed Nucleases

There are three types of Site-Directed Nucleases (SDN) (see Table 2.1). SDN uses DNA-cutting enzymes (nucleases) that are instructed to cut the DNA at a specific location by several binding systems. After the cut is made, the cell’s own DNA repair mechanisms recognize the break and repair it using one of two cell repair processes. In the case of the Non-Homologous End-Joining (NHEJ) pathway, there is no donor DNA. The cut in DNA is rejoined, however this may cause a few base pairs to be eaten away or added, resulting in random, small deletions or additions (a few base pairs) of nucleotides at the cut site (SDN1) (Entine et al., 2021: 559). The Homology-Directed Repair (HDR) pathway involves a donor DNA that carries the chosen change and has homologyFootnote 5 with the target site used to introduce the chosen change at the cut site. This allows for an introduction of specific intentional insertions, changes, or deletions. The SDN2 technique targets a gene for correction. The SDN3 technique inserts a gene into the DNA.

Table 2.1 Three main types of gene editing using Site-Directed Nucleases (SDN)
Full size table

In some cases, countries regulate the three main types of gene editing differently. Since the SDN3 technique sometimes uses transgenes to edit the genome (transgenics), its risk profile is treated differently from gene editing that uses SDN1 or SDN2 in some contexts. Some argue that mutations in a controlled environment regardless of the technology used may, in fact, be safer than those occurring in nature. As research scientist Informant 5 put it,

I find that technologies tend to be exceptionalized in the food system. And the reality is that technologies are probably the least risky part of how we produce and consume food.

We asked Informant 17, who is a research scientist, how they view gene editing compared to mutagenesis in terms of their safety profiles. When asked what they wish laypersons knew about gene editing, Informant 17 responded,

…that it is equivalent to traditional mutagenesis approach. For adapting gene function. And that it’s safer than those mutagenesis approaches. Because when you apply mutations, you get a plethora of mutants in every individual plant and then you screen them for something of interest related to your gene of interest. And then you have to back cross many, many times to get rid of all the background mutations.

Informant 16, also a research scientist, concurred with the above comments, stating,

I think the general idea would be that we are not doing anything new that nature hasn’t been done before. And the reason behind that is that the more media find out about genomes of the organisms around us, the more we discover that they have been moving pieces of their DNA all around throughout evolution history. We can find fragments of raw genome, we can find fragments of viral genome in the human genome.

This point is not often included in contemporary discussions about the safety and efficacy of gene editing. In many jurisdictions, gene editing and transgenics are regulated differently than techniques used in conventional breeding like mutagenesis. We discuss how these techniques are regulated in Chap. 3.

Techniques used in plant breeding aim to achieve intentional and precise knockouts, or a (re)introduction of a desired trait. These techniques include Meganucleases, Zinc Finger Nuclease (ZFN), TAL Effector Nucleases (TALEN) and CRISPR-Cas9 (PRRI, 2023) (Table 2.2).

Table 2.2 Gene editing techniques
Full size table

Figure 2.1 shows the published discoveries and descriptions of gene editing techniques from 1985 until 2020. The following section discusses the four main techniques that continue to be used to this day.

Fig. 2.1
A timeline diagram. 1985, the discovery of Zinc finger proteins and meganucleases. 1987, C R I S P Rs described. 1996, Z D Ns created. 2009, T A L E described. 2010, T A L E Ns created. 2012, C R I S P E slash C a s 9 programmable nuclease. 2016, base editing. 2019, prime editing.

Timeline of gene editing techniques. (Source: Adapted from Tröder & Zevnik, 2022)

Full size image

2.2 Meganucleases

The study of meganucleases in 1985 first revealed the basic mechanisms of DNA cleavage and the DNA repair processes on which genome editing depends. Meganucleases are a special type of enzyme that binds to and cuts DNA at specific sequences that occur at a few sites in the genome. Meganucleases are single proteins that recognize a sequence in the DNA and break the target DNA, leaving a double-strand break that can be repaired through a natural repair process used to repair broken DNA (Silva et al., 2011). They naturally occur in bacteria, single celled organisms, plants, animals, and fungi. Meganucleases can be used in genome editing for both nonhomologous end joining and homology directed repair–mediated alterations. Meganucleases-mediated genome editing has been demonstrated in maize (Zea mays) and tobacco (Nicotiana spp.) (Baltes & Voytas, 2014). Production is difficult, and cost of production is costly. It can take months to conduct an experiment using meganucleases. It is difficult to change the target sequence specificity of meganucleases, so they are not widely used for genome editing (NASEM, 2016).

2.3 Zinc Finger Nuclease

One of the first reliable, successful methods of genome editing was reported in 2003, though it was discovered in 1985. Zinc Finger Nuclease (ZFN) interacts with three specific base pairs of DNA that cause double-strand breaks at a targeted site in the genome. It requires two proteins. ZFNs are used to introduce mutations via Non Homologous End Joining (NHEJ), which is a natural repair process used to join the two ends of a broken DNA strand back together. ZFNs have been used to modify plants, but the technique is not always accurate as it sometimes targets the wrong sequence (NASEM, 2016). An example of where this technique has been used is in the modification of the endogenous tobacco acetolactate synthase genes, which is the target enzyme for two types of herbicides. By using ZFN and a donor molecule, mutations were induced, thus generating plants which were herbicide resistant (see Novak, 2019). The production and cost of production of ZFNs is prohibitively expensive. It can take months to conduct an experiment using ZFNs for gene editing.

2.4 TALENs

Transcription activator-like effectors (TALEs) followed zinc finger nucleases and preceded CRISPR-Cas9 as effective genome editing tools. The discovery of TALEs was first published in 2009 in Science by scientists at Martin Luther University in Halle, Germany. TALEs are proteins with a unique DNA-binding domain that presents a predictable and programmable specificity. TALEs are made and used by plant pathogenic bacteria to control plant genes during infection. In nature, TALEs bind to plant DNA sequences and activate genes. The bacteria encode TALEs through a simple code that has been exploited to engineer proteins with custom site specificity in any target genome (Boch et al., 2009; Moscou & Bogdanove 2009; NASEM, 2016, 2017a). The ease of the design for specific target DNA sequences of the TALEs revolutionized genome editing.

Like ZFNs, TALEs can be fused with the nuclease domain of FokI (a restriction endonuclease—a ‘cleaver’ enzyme) and utilized to edit the genome, referred to as Transcription Activator-Like Effectors Nucleases (TALENs). TALENs are used in pairs like ZFNs to affect targeted mutations. TALENs have a higher efficiency than ZFNs and have been used to alter the genomes of a variety of different organisms. TALENs have been used to edit genomes in rice, maize, wheat, and soybean (Baltes & Voytas, 2014; NASEM, 2016). TALENs have also been used experimentally to correct mutations that cause human disease. Like ZFNs, TALENs experiments require two proteins. Production is relatively easy, and significantly more cost effective than meganucleases and ZFNs. Gene editing experiments using TALENs can take weeks, as opposed to months as with previous techniques. However, researchers continue to improve upon TALENs. In 2020, a rapid TALENs preparation protocol was developed. This has improved reproducibility and efficiency of this gene editing technique according to the International Service for the Acquisition of Agri-biotech Applications (ISAAA, 2023).Footnote 6

TALENs has been used in several plants to improve agronomic characteristics or nutritional profiles. For example, TALENs has been used to create soybeans with low levels of polyunsaturated fats. Oils with lower levels of these fats are considered healthier compared to oils that can be hydrogenated to produce trans-fatty acids (ISAAA, 2023). Soybeans that produce high oleic acid (the ‘healthier’ fat) emerged on the US market beginning in 2019 in the form of soybean oil. TALENs has also been used in rice to breed resistance to bacterial blight. Through gene editing, scientists were able to generate inheritable disease resistance. TALENs has also been used to reduce acrylamide in potatoes, maize, and wheat, as well as breed resistance to wheat powdery mildew. Further applications of TALENs include utilizing sugarcane and algae in the biofuels industry (ISAAA, 2023).

2.5 CRISPR-Cas9

The Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR-Cas9) system as a gene editing tool was reported in 2012 (Doudna & Charpentier, 2014). CRISPR is a naturally occurring mechanism found in bacteria. Bacteria harbour CRISPR as an innate defense mechanism against viruses and plasmidsFootnote 7 that uses RNA-guided nucleases to target the break or cut of foreign DNA sequences. The bacteria retain fragments of foreign DNA which provides it with some immunity to viruses. CRISPR-Cas systems can generate a range of DNA edits which are synonymous with those found in natural populations. Multiple genetic changes can be achieved in a single generation (Lyzenga et al., 2021). CRISPR-Cas9 has received the most academic attention, namely because it is more precise in targeting specific genes than other techniques and is more economical to use than other gene editing techniques mentioned above. This has improved the timeliness of experimenting with this technique, and reduced the costs associated with gene editing research. It is the most recently introduced of all gene editing platforms and appears to have the highest accuracy of all gene editing techniques discussed here.

CRISPR-Cas9 is the gene-editing platform in which RNA homologous with the targeted gene is combined with the Cas9 (DNA snipping enzyme). Cas9 gives CRISPR the ability to alter DNA sequences. Cas9 makes up part of the “toolkit” for the CRISPR-Cas9 system of genome editing. The other is a homing device that can be programmed to target the DNA sequence of interest, imparting precise control over the location of edits. Scientists have dissected the innate CRISPR-Cas9 system and re-engineered it in such a way that a single RNA, the guide RNA, is needed for Cas9-mediated cleavage of a target sequence in a genome (Alvarez, 2021; Sander & Joung, 2014).

Guide RNA (short segments of RNA used to direct the DNA-cutting enzyme to the target location in the genome) design requirements are limited to a unique sequence of about 20 nucleotides in the genome (to prevent off-target effects) and are restricted near the Protospacer Adjacent Motif sequence (PAM),Footnote 8 which is specific for the CRISPR-Cas9 system. Newer applications of CRISPR include the use of two guide RNAs with a modified nuclease that “nicks” one strand of the DNA, providing greater specificity for targeted deletions. The ease of design, the specificity of the guide RNA, and the simplicity of the CRISPR-Cas9 system have resulted in rapid demonstration of the utility of this method of editing genomes in plants and other organisms (Baltes & Voytas, 2014).

CRISPR-Cas9 is a versatile and robust gene editing tool for crop improvement. Because of its efficiency and accuracy, this method is rapidly becoming the most widely used approach for performing gene editing. For example, researchers from the Institute of Agricultural Technology of Argentina (INTA) used CRISPR-Cas9 to develop non-browning potatoes and lactose-free cow’s milk. CRISPR-Cas9 has also been used to increase alfalfa productivity and quality (Laaninen, 2020: 8). CRISPR-mediated gene knockout has been applied in rice, barley, soy, maize, wheat, tomato, potato, lettuce, citrus trees, mushrooms, cucumbers, grapes, watermelon, and others (Liang et al., 2014).

What seems to baffle some of the scientists we interviewed is the artificial distinctions arising in public debates over the safety and risks of gene editing—between genomic editing and mutations that take place in a controlled laboratory setting versus mutations that happen in nature. The distinction between manufactured risks and natural risks seems to be at the centre of the debate over the safety of gene editing in the agrifood system; though at a conceptual level, mutations or changes to genomes happen in nature. Informant 6 and Informant 7 (an NGO representative and a research scientist, respectively) had similar things to say about how the perception of the safety of gene editing compared to ‘conventional’ breeding techniques such as mutagenesis is distorted in the public dialogue on biotechnology in the agrifood system. Informant 7 notes,

…if you compare [gene editing] to a conventional breeding technique like a random mutagenesis which is considered conventional, it’s in the organic section of the grocery store, you get hundreds to thousands of random mutations per plant, using those techniques. And what? It’s fine. Those plants have been on the market and we’ve been eating them for a hundred years, and they’re fine. But that’s a lot of unintended mutations that we don’t even know about. Whereas with gene editing, if you get unintended mutations off-target first of all, we tend to know exactly where they would be, because they happen in regions that are very similar to the region that we’re targeting. We can scan those, and we do. We look at that, and we can tell if it’s been enough target mutation. …. And secondly, we’re certainly not getting thousands from the gene editing process itself.

Informant 7 had more to say about genetic mutations in agrifood plants. They continue,

they’re fine, I mean one rice plant. It produces a seed and that seed has around 40 mutations, random. We don’t know where they are. They’re all unintended. They’re all just…who knows? And it’s fine. No problem…Mutations happen, and that’s a good thing. That’s how things evolve. That’s how things adapt.

CRISPRoff/on

Another recent discovery in the CRISPR-Cas system is ‘CRISPRoff’ which is a technique using CRISPR-Cas9 to turn genes ‘on’, but which also has the potential to turn them ‘off’ (Nunez et al., 2021). CRISPRoff is a reversible tool for controlling gene expression that is specific, precise, and inheritable. It involves the addition of a chemical tag to the DNA making it inaccessible for reading and subsequent protein production. In the paper published in Cell, Nunez et al. (2021) describe how to modify CRISPR’s basic architecture to extend its reach beyond the genome and into what is known as the epigenome—proteins and small molecules that latch onto DNA and control when and where genes are switched on or off. The researchers show that once a gene is switched off, it remains inert in the cell’s descendants for hundreds of generations, unless it is switched back on with a complementary tool called CRISPRon.

Discussions have emerged regarding how to classify organisms that have had CRISPRoff applied to their genome. If the genome is not changed, but there is genetic manipulation, is it a GMO? Identification of organisms produced using CRISPRoff would be difficult to identify, as critics of gene editing in general have noted (Williams, 2021). These are advancements that may require new regulatory scrutiny in the coming years as their applications evolve, and the need for innovative ways to address climate change and food insecurity become even more pressing. However, many countries, such as Canada and Argentina, have designed their regulatory frameworks based on a risk assessment of products, not how the organism was developed. Therefore, as the National Academies of Sciences, Engineering and Medicine (NASEM) (2017a, p. 108) document states, “risk assessment endpoints for future biotechnology products are not new compared with those that have been identified for existing biotechnology products”. The benefit of having a regulatory system tooled to conduct risk assessments for previous and current biotechnologies in the agrifood system on a case-by-case basis is the built-in ability to assess future, unknown technologies for future, unknown risks. However, regulatory systems may have to continually update and adapt to assess yet unknown risks that accompany new breeding techniques. As the NAESM (2017a) report cautions—if regulatory systems are not prepared for the wave of new breeding techniques on the horizon, there will be stifling of innovation and a slowing in the development of products that can be useful tools in combating food insecurity and climate change.

Table 2.3 lists applications of gene editing in the agrifood plant breeding space, including the type of NBT used in the breeding process, the type of quality improvement (food & feed and/or agronomic), the trait that was the specific focus of the breeding initiative, country of origin and research organization(s) responsible for developing the application.

Table 2.3 Applications of gene editing in agrifood plant breeding
Full size table

3 Conclusion

Due to their accuracy, lower costs and application simplicity, genome editing tools can introduce valuable quantitative and qualitative traits into plants. Current research is focusing on improving agronomic traits including drought resistance, increased yield, pathogen resistance, and decreased time to ripening. CRISPR-Cas9 platforms are not perfect, and can result in off-target modifications, unwanted on-target modifications, and genomic rearrangements. ZFN and TALENs systems continue to evolve and are improving in terms of accuracy, cost and time requirements. A number of new techniques are on the horizon and are discussed in greater detail in Chap. 4.