Three protein fragments tell your immune system that corn, wheat, and soy are safe to eat

Every time you eat a cookie, your immune system performs an act of calculated restraint. The proteins in that cookie are foreign -- they are not you -- yet your body lets them pass without mounting an attack. This process, called oral tolerance, is so seamless that most people never think about it. But for the roughly 6% of young children and 3% to 4% of adults with food allergies, the system has broken down. Their immune cells treat harmless food proteins as threats.

Scientists have spent years mapping the proteins that trigger allergic reactions -- the "what goes wrong" side of the equation. The other side, the "what goes right" when tolerance works, has remained largely blank. A study published March 6, 2026, in Science Immunology begins to fill that gap.

Starting with a bowl of mouse chow

The research team, led by Jamie Blum (now at the Salk Institute) and Elizabeth Sattely at Stanford University, took an unusual approach. Rather than testing one food protein at a time, they started with regular mouse chow and asked a broad question: what are the regulatory T cells in these mice actually responding to?

Regulatory T cells are the immune system's peacekeepers. Prior research had established their role in suppressing inflammatory responses and maintaining tolerance. But no one had identified the specific food-derived protein fragments -- called epitopes -- that these cells recognize as safe.

By screening regulatory T cells from mice on a normal diet and mapping what the cells were attaching to, the team traced the responses back to three epitopes: one from corn, one from wheat, and one from soybean. All three come from seed storage proteins, which are among the most abundant proteins in plant-based foods.

The corn connection and the soy surprise

The relative abundance of regulatory T cells reactive to each epitope tracked with what immunologists would expect. The corn epitope recruited the most regulatory T cells, consistent with the fact that corn allergies are relatively uncommon. The immune system appears to have a robust tolerance mechanism for corn proteins.

The soybean epitope was more intriguing. Soy is one of the major allergens in humans, so finding a tolerance-associated epitope for soy suggests the system is more nuanced than a simple on-off switch. The same immune receptor that interacts with the soybean epitope also recognizes sesame proteins, which may explain cross-tolerance -- when tolerance to one food confers tolerance to a related one.

"Understanding how the immune system can normally see a protein as safe may lead to new therapies to promote tolerance in individuals with allergy," Blum said.

Where these cells live and what they do

With the epitopes identified, the team asked follow-up questions. Where in the body do these regulatory T cells reside? How do they behave differently in healthy versus inflamed tissue?

Using mice and cell culture models, the researchers found the regulatory T cells are primarily located in the gut -- which makes anatomical sense, since that is where food meets the immune system. Their behavior varied depending on context. In healthy tissue, the cells maintained a baseline of immune calm. In inflamed environments, they actively worked to reduce the inflammatory response.

This dual function matters for therapeutic design. A future therapy would need to work not just in calm conditions but in the inflammatory environment of an active allergic reaction.

From mice to potential therapies

The practical implications are significant but distant. Scientists have already been exploring regulatory T cells as a possible immunotherapy platform for severe food allergies. The idea: create regulatory T cells pre-programmed to tolerate specific food proteins, then introduce them into patients to dampen allergic responses.

The identification of specific tolerance epitopes provides something that approach has been missing -- a molecular target. If you want to engineer a regulatory T cell that tolerates peanut, you need to know which piece of peanut protein to aim at. This study demonstrates a workflow for finding those pieces.

"Diet is our most intimate interaction with our environment," Blum said. "Correctly recognizing foods as safe creates an anti-inflammatory environment to support nutrient acquisition and prevent allergy."

What remains unknown

The work was conducted in mice, and the identified epitopes are from mouse chow ingredients. Whether the same epitopes are relevant in humans is an open question that the team plans to address next. The reagent they developed to track these proteins has been made available to other researchers, which should accelerate follow-up studies.

The study also does not explain why tolerance fails in allergic individuals. Finding the proteins that healthy immune systems recognize as safe is a necessary step, but the mechanism by which that recognition breaks down remains to be mapped.

The research involved collaborators at Stanford, New York University, the Chan Zuckerberg Biohub, and Harvard University, and was funded by the National Institutes of Health, the National Science Foundation, and the Howard Hughes Medical Institute.