Source: Frontiers in Nutrition, by Liu, Wenhan; Lu, Yi; Xu, Na (September 15, 2026). AI-generated summary by biochip.com, published . Not independently reviewed.
Key takeaways
- The review maps organ-on-chip and multi-organ food-toxicology studies published through July 10, 2026.
- Gut chips showed flow- and route-dependent responses to mycotoxins, dioxins, metals, enterotoxins, and food-relevant particles.
- Most systems still lack authentic food matrices, chronic mixtures, microbiota, human benchmarks, and interlaboratory transfer.
Organ-on-chip systems could help food toxicology study how dietary chemicals move through the human body, rather than treating the gut as the end of the story. In a review published in Frontiers in Nutrition, Wenhan Liu, Yi Lu, and Na Xu examine living chip-based models and multi-organ microphysiological systems used to investigate contaminants, additives, toxins, pathogens, particles, and environmental chemicals relevant to food. These platforms combine human cells with controlled fluid flow, tissue interfaces, and sometimes mechanical forces to recreate selected functions of organs such as the intestine, liver, and placenta. The review argues that this approach may better reflect human exposure than high-dose animal studies or static cell cultures, particularly for low-level, long-term, and mixed dietary exposures. Gut chips have already revealed route- and flow-dependent responses to hazards including mycotoxins, dioxins, metals, enterotoxins, and food-relevant particles. Connected gut-liver systems also show how a substance that crosses the intestinal barrier can contribute to injury in a distant organ. Still, the field has not yet reproduced many real-world features of eating, including complex food matrices, digestion, microbiota, chronic mixtures, and well-established human exposure benchmarks. The review calls this emerging direction Green Food Toxicology and proposes that organ-on-chip systems should be qualified for specific regulatory uses, not presented as universal replacements for animal studies.
Why food toxicology needs better models
Food safety is difficult because people rarely encounter one purified chemical at one dose for one day. They may consume low concentrations of multiple substances over decades, while diet, geography, age, pregnancy, metabolism, and gut microbes all shape what reaches the body and what harm may follow.
Animal studies remain useful because they capture whole-body physiology, but animal and human intestines differ in their structure, transport proteins, metabolism, immune signaling, microbiota, and timing of exposure across life. Conventional cell cultures offer tighter experimental control, yet cells grown in a still dish generally lack flowing fluid, physical forces, tissue boundaries, circulating immune cells, and communication between organs.
What makes an organ-on-chip different
An organ-on-chip is not simply a small device with cells inside it. Think of it less like a sample tray and more like a working miniature environment, where cells experience carefully managed flow, geometry, chemical gradients, and sometimes stretching or other mechanical cues that influence how tissue behaves.
Liu, Lu, and Xu define an OoC as living tissue in a controlled microsystem that demonstrates at least one organ-relevant function. A microphysiological system, or MPS, is the broader category: it includes in vitro systems designed to reproduce selected features of tissue or organ physiology, whether they are microfluidic, larger perfused models, or another dynamically controlled format.
That distinction matters in food science, where “chip” can also describe an analytical sensor that detects a toxin. A microfluidic device that measures a toxin without living tissue is a lab-on-chip, not an organ-on-chip; likewise, a static organoid does not become an OoC unless it is integrated into a controlled system with defined exposure and functional measurements.
Following a dietary exposure beyond the gut
The intestine is the first major biological barrier for ingested hazards, but it is only the beginning of the exposure pathway. A food-associated substance may be released during digestion, changed by gut microbiota, pass through mucus and intestinal cells, enter circulation, and then be metabolized or activated by the liver.
From there, a chemical may circulate in free or protein-bound forms, be cleared by the kidney, or affect organs including the heart, brain, immune system, endocrine tissues, or the maternal-fetal interface. Multi-organ models are designed to follow this sequence by connecting the intestinal module to only the downstream tissue needed to answer a particular toxicology question.
This is why a gut-liver chip can be more informative than two isolated cultures. It can model a causal chain in which a substance crosses the gut barrier, reaches the liver, undergoes metabolism, and contributes to damage elsewhere. The review cautions that the goal should not be a decorative miniature human body, but a focused system that captures the relevant exposure-to-outcome pathway.
What the reviewed systems have shown
The review maps studies published through July 10, 2026, separating direct food-toxicology evidence from related route studies, transferable validation work, and governance evidence. The authors assess these studies using practical criteria: whether exposures resemble real conditions, whether organ connections are meaningful, how dose is handled, and whether a system is ready for regulatory use.
Gut chips have provided evidence that responses can depend on flow and the route of exposure. The systems reviewed examined hazards that include mycotoxins, toxic compounds made by certain fungi; dioxins; metals; enterotoxins; and particles relevant to food exposure.
Liver and placenta models add other pieces of the puzzle. Liver systems can help resolve effects shaped by metabolism, including cases where a chemical becomes more or less harmful after the body processes it, while placenta models can help examine life-stage-specific concerns during pregnancy.
Green Food Toxicology as a framework
The authors use the term Green Food Toxicology to describe a direction for food-safety science that is designed to be more human-relevant, less dependent on animals, guided by realistic exposure, conscious of resource use, and useful for regulation. These are not just broad aspirations: the review frames them as five auditable dimensions, meaning researchers should be able to show how a model meets them for a defined purpose.
That emphasis on context is important. A platform built to test whether a compound crosses the intestinal barrier may not be equipped to predict developmental effects in pregnancy. Similarly, a well-characterized liver module may be valuable for metabolism studies without being sufficient for a full dietary-risk assessment.
Why This Matters
Food toxicology increasingly needs tools that can connect exposure to biological effect in human-relevant ways. Organ-on-chip and multi-organ systems offer a practical route to study transport, metabolism, tissue injury, and organ-to-organ communication while retaining experimental control that is difficult to achieve in people or whole animals.
Yet the review's central message is measured rather than absolute. Most studies still rely on purified single agents, short experimental windows, transformed cell lines, and nominal concentrations, which are the concentrations added to an experiment rather than necessarily the amount that reaches cells. Authentic foods, digestion processes, microbiota, chronic mixed exposures, shared human benchmarks, and interlaboratory reproducibility remain uncommon.
The next stage will be to qualify individual OoC and MPS platforms for narrow, clearly stated contexts of use within an integrated weight-of-evidence assessment. If researchers can link realistic dietary exposure to reliable organ-specific outcomes, these living microsystems may become more useful tools for deciding which food-related hazards deserve closer scrutiny.