By: Nigel G Halford, Rothamsted Research

Bakers across Europe will be familiar with acrylamide, a probably carcinogenic processing contaminant that forms during everyday cooking and processing of common, crop-derived raw materials, including grains. Acrylamide is found in fried, baked, roasted and toasted foods, and all cereal-derived products are affected.

The FAO/WHO Joint Expert Committee on Food Additives (JECFA) described the presence of acrylamide in food as a ‘concern’ as far back as 2005, and that opinion was reiterated by EFSA’s Expert Committee on Contaminants in the Food Chain (CONTAM) in 2015 and restated in 2022. The European Commission responded to CONTAM’s 2015 report by introducing Commission Regulation (EU) 2017/2158, which set Benchmark Levels for acrylamide in different food categories, as well as requiring food businesses to monitor the levels of acrylamide in their products and keep a record of the mitigation measures they apply. Note that the onus up to now has been on food businesses alone and not their supply chain.

Regulation (EU) 2017/2158 included an explicit threat to impose Maximum Levels for acrylamide in sectors where progress on reducing acrylamide levels were considered to be insufficient. That process started in 2020 when the European Commission brought draft legislation on Maximum Levels and lower Benchmark Levels for baby foods to the European Parliament.

The Parliament rejected the draft because the proposed levels were too high in the opinion of MEPs, and did not cover enough products. Since then, the European Commission has been struggling to find middle ground between the levels that MEPs are demanding and what food manufacturers would be able to comply with.

A document with draft Maximum Levels and revised Benchmark Levels was circulated amongst stakeholders in January this year, so food manufacturers should be preparing for the imminent imposition of Maximum Levels, even though we do not know exactly when that will happen. 

Acrylamide forms from an amino acid, asparagine, together with sugars such as glucose, fructose and maltose. This is the Maillard Reaction and when other amino acids participate it generates the colours, flavours and aromas that consumers demand in fried, baked, roasted and toasted foods, making the problem all the more difficult to solve.

In cereal products, it is the concentration of soluble (non-protein) asparagine in the grain that determines the amount of acrylamide that forms, together with the way that the raw material is processed. Manufacturers have developed many ways to reduce acrylamide formation in their products, from close control of cooking times and temperatures to quality control of the finished product, often using colour as an indicator. What they have been crying out for is low asparagine wheat that would enable them to comply with tightening regulations on acrylamide while retaining the characteristics that define their brands.

We have shown that asparagine concentrations in wheat grain can be massively reduced using the genome editing technique, CRISPR*, without affecting yield. This was achieved by targeting genes that are responsible for synthesising asparagine.

Wheat has five of these genes, but we targeted one or two that are most active in the grain, using CRISPR to introduce changes that rendered the genes dysfunctional (a so-called knockout). Knocking out a single gene reduced asparagine content by 59% over two years of field trials, while the double knockout reduced asparagine content by an astonishing 93%. Acrylamide in bread made from the double knockout line was below detection levels and even after toasting was a fraction of that in the unedited equivalent. A 50% reduction in asparagine was also achieved using old-fashioned chemical mutagenesis, but in that case there was a substantial yield penalty.

Chemical mutagenesis is completely random and plants produced from it will carry thousands of mutations alongside the ones we need in the target genes. It is likely that this genetic baggage was responsible for the yield drag. However, these plants have no specific regulatory issues associated with them, while that is not the case for genome edited (GE) plants.

The notion that a random technique for creating mutations in a plant is safer than a targeted one has no scientific basis and the situation arises simply due to the fact that chemical mutagenesis is an older, more established technique (traditional, even). This was the reasoning expressed in a European Court of Justice ruling in 2018 that flatly ignored the scientific consensus and stated that GE plants must undergo the same risk assessment process that has stymied the development of genetically modified crops in the EU for 30 years.

Eight years on from that ECJ ruling and the EU is moving, albeit at glacial speed, to introduce legislation that will allow for the commercialisation of some GE crops. The New Genomic Technologies regulation was approved by the European Council recently (21^st April 2026). However, it still needs to be formally adopted by the European Parliament and there is push-back from some Member States.

Even if it has a smooth passage, the new framework will not be in place until mid-2028. The UK is  further ahead: its equivalent legislation, the Genetic Technology (Precision Breeding) Act, gained royal assent in 2023 and the process for applications to market a precision bred organism (PBO) in England was opened in November 2025.

However the legislation has not been adopted by the devolved governments of Scotland, Wales or Northern Ireland and we are unlikely to see breeders getting involved in the genome editing of commodity crops like wheat while that is the case.

Further uncertainty has been generated by the UK and EU negotiating the possible realignment of food standards, with no clear indication of what might happen to the Genetic Technology (Precision Breeding) Act. Wheat breeders work on 5-10 year timelines and need clarity and certainty if they are to take the leap and adopt a new technology.

Breeders are also wary about the possible consumer reaction to GE crops, having got their fingers badly burned in the GM fiasco of the 1990s and 2000s. All I can say as someone who participated in the GM debate at that time is that there is nothing like the hostility to GE crops now as there was to GM crops then. Consumers do seem to understand the distinction between GM crops that contain an additional gene or genes, possibly from a different species and PBO/NGT crops that only have changes to their native genes.

We can also look to other countries where GE crops are already on the market. Japan, for example, has never allowed the cultivation of GM crops, but was the first to approve a GE variety (the Sicilian Rouge tomato) in 2021, with no apparent consumer backlash.

As of now, breeders are taking a wait and see approach, so any chance of very low asparagine CRISPR wheat being brought to market is some way off. Bakers and other food businesses also face uncertainty: what will Maximum Levels for acrylamide be set at, what products will be affected and will other jurisdictions follow the EU’s lead?

However, for now they will have to cope with evolving regulations on acrylamide without low asparagine grains.

Rothamsted Research, Harpenden AL5 2JQ, United Kingdom

Reference : Plant Biotechnology Journal. 2026. DOI: 10.1111/pbi.70661

*CRISPR - Clustered Regularly Interspaced Short Palindromic Repeats