Through an extensive literature review, researchers mapped existing food and feed products made with genetically modified crops containing protein (e.g., corn, soybean, rapeseed, etc.) and novel foods (e.g., insect proteins, plant-based proteins, fermentation-derived proteins, etc.), as well as their processing steps.

The literature review revealed 40 products (and their processes) not included in the Organization for Economic Cooperation and Development (OECD) documents for compositional considerations of genetically modified crops. These OECD documents are published for the safety assessment of genetically modified foods of plant origin, and are considered in EFSA’s novel food risk assessments.

Available scientific evidence about the impact of processing on protein safety suggested that certain processes—including thermal treatments, fermentation, or enzymatic hydrolysis—can significantly enhance protein digestibility in various foods. Additionally, fermentation, ensiling, and extraction processes have been shown to improve nutritional properties in various products.

The researchers also found that heating processes can effectively reduce the activity of newly expressed proteins (i.e., proteins expressed in genetically modified crops with specific functions, conferring a new trait). Moreover, with some exceptions, the data showed that heating and enzymatic hydrolysis can reduce immunoglobulin E (IgE) reactivity for certain proteins and operational conditions. IgE is an antibody that plays a role in allergic reactions.

A Fit-for-Purpose Toxicity Testing Framework for Novel Proteins 

EU regulations mandate novel food applicants to submit toxicity studies for proteins, often requiring animal studies. At the same time, the safety assessment of proteins in food is currently based on strategies that have been developed for chemical risk assessment, even though toxicity testing for proteins presents unique challenges.

In this context, EFSA researchers developed an in vitro toxicity testing framework to meet the “urgent need” for fit-for-purpose safety assessment methods for novel proteins, which could be used alternatively to animal studies.

To do so, the researchers created a literature database of toxic outcomes associated with toxic proteins, and found that toxic proteins are produced by a relatively limited phylogenetic subset, including, but not limited to, bacteria, insects, serpents, mollusks, and fungi. Next, relevant toxicity tests to address and quantify these toxicological effects were identified and evaluated for their applicability in an in vitro-based toxicity testing strategy.

The researchers concluded that no single assay or model system is sufficient for an adequate evaluation of toxicity. Based on the data they collected, the researchers instead developed a three-step approach to in vitro novel protein toxicity testing, beginning with 1) an initial in silico toxicity prediction, followed by 2) in vitro assays to address the primary mechanisms of toxicity associated with toxic proteins, and if any concerns arise from these tests, 3) the use of relevant in vitro model systems to explore potential target organ toxicity.

The approach recommends mechanistic endpoints in in vitro assays instead of adverse outcomes observed in traditional animal testing. It considers diverse mechanisms of protein toxicity, and is meant to be a starting point for the development of in vitro protein toxicity testing strategies for novel food safety assessments.