Human intestinal cell model enables precise detection of drug-induced barrier damage

Researchers have developed a human intestinal cell model that closely mimics the structure and function of the human gut, enabling more precise prediction of drug-induced gastrointestinal toxicity during the preclinical stage of drug development.

A research team led by Dr. Mi-Young Son at the National Agenda Research Institute of the Korea Research Institute of Bioscience and Biotechnology (KRIBB) has established a new evaluation platform that can more accurately assess gastrointestinal toxicity of drug candidates before they enter clinical trials.

Gastrointestinal toxicity refers to adverse effects on the digestive tract, such as vomiting, diarrhea, and mucositis, that may occur after drug administration. These side effects are a major cause of treatment discontinuation or dose reduction during clinical trials, often contributing to the failure of drug development programs.

In many cases, these adverse effects begin with the weakening of the intestinal barrier, followed by inflammation and tissue damage. Detecting these early changes is therefore critical for identifying potential toxicity at an early stage.

However, conventional evaluation methods have largely relied on colon cancer–derived cell lines such as Caco-2 or approaches that detect toxicity only after extensive cell death occurs. Such methods fail to fully reproduce the architecture of the normal human intestine and often miss early toxicity signals in which the intestinal barrier becomes compromised while the cells themselves remain alive.

To address these limitations, the KRIBB research team developed the hIEC (human intestinal epithelial cell) model, derived from human stem cells. The model contains multiple intestinal cell types—including nutrient-absorbing epithelial cells and mucus-secreting cells—allowing it to more closely replicate the cellular composition of the human intestine.

Importantly, the model demonstrated transepithelial electrical resistance (TEER) levels comparable to those observed in the human intestine, indicating that the platform can recreate a physiologically relevant intestinal barrier in laboratory conditions.

Because the hIEC cells can also be generated from three-dimensional intestinal organoids, the platform is expected to have broad applications in organoid-based disease modeling and drug testing.

To validate the system, the researchers tested 17 clinically used drugs, including anticancer agents, targeted therapies, and anti-inflammatory drugs, to evaluate toxicity prediction performance.

The results showed that the platform predicted gastrointestinal toxicity with 94% overall accuracy. Notably, it achieved 100% specificity, significantly reducing the risk of falsely identifying safe drugs as toxic.

Furthermore, the system detected early intestinal barrier damage with 92% sensitivity, including toxicity signals from anticancer drugs such as paclitaxel that could not be detected using conventional assays. This represents an important advance because the model directly measures functional changes in barrier integrity rather than simply assessing cell survival.

The research team also performed transcriptomic analysis to investigate the molecular mechanisms underlying drug-induced intestinal barrier impairment.

Their findings revealed that certain anticancer drugs strongly suppress genes associated with the cytoskeleton and cell-cell adhesion, which are essential for maintaining cellular structure and tight intercellular connections.

This indicates that intestinal barrier dysfunction can occur even before cells die, as weakening of the cytoskeletal framework destabilizes the epithelial barrier.

The study establishes a human-based evaluation platform capable of accurately predicting gastrointestinal toxicity of drug candidates prior to clinical trials. Such systems could reduce unnecessary clinical failures and potentially complement or replace some animal testing approaches.

“This study demonstrates that drug-induced intestinal damage can be predicted with high accuracy using a model that closely replicates real human intestinal function,” said Dr. Mi-Young Son, the lead investigator. “Our next goal is to expand this technology into a patient-specific organoid-based platform capable of precisely predicting intestinal toxicity in personalized medicine.”

Korea Research Institute of Bioscience and Biotechnology (KRIBB) is a leading national research institute in South Korea dedicated to cutting-edge research in biotechnology and life sciences. Established in 1985, KRIBB focuses on advancing scientific knowledge in areas such as molecular biology, genomics, bioinformatics, synthetic biology, and aging-related studies. As a government-funded institute, KRIBB plays a pivotal role in driving innovation, supporting national R&D strategies, and collaborating with academic and industrial partners both domestically and internationally.

This research was supported by the Major Research Programs of the Korea Research Institute of Bioscience and Biotechnology (KRIBB), the Ministry of Food and Drug Safety R&D Program, and the Bio and Medical Technology Development Program of the Ministry of Science and ICT.

The study was published online on February 12 in Experimental & Molecular Medicine (Impact Factor: 12.8), a leading international journal in basic and translational medicine.

The article is titled “Drug-induced gastrointestinal toxicity and barrier integrity: cytoskeleton-mediated impairment in a clinically relevant human intestinal epithelium model.”

The corresponding authors are Dr. Mi-Young Son, Dr. HaNa Lee, and Dr. Dae-Soo Kim of KRIBB, and the first author is Dr. Won Dong Yu of KRIBB.

This research was supported by the Major Research Programs of the Korea Research Institute of Bioscience and Biotechnology (KRIBB), the Ministry of Food and Drug Safety R&D Program, and the Bio and Medical Technology Development Program of the Ministry of Science and ICT.