Oral medications must traverse the lining of the digestive tract through a process facilitated by proteins found in the cells lining the gastrointestinal tract. Researchers at MIT, Duke University, and Brigham and Women’s Hospital have developed a new strategy to identify these proteins (transporters) utilized by individual drugs. This knowledge could enhance patient treatment, as prescribing two drugs that depend on the same transporter could lead to unintended drug interactions.
Through a combination of tissue models and machine-learning algorithms, the team has already found that a popular blood thinner and antibiotic could potentially interfere with one another. Learning more about the transporters could also help drug developers increase the absorbability of future medications by identifying complimentary excipients that boost their interaction with transporters.
The researchers adapted a tissue model they had created the previous year to measure a drug’s absorbability. This model, using laboratory-grown pig intestinal tissue, was systematically exposed to various drug formulations to record their absorption levels. Additionally, siRNA strands were used to inhibit the expression of each transporter, allowing detailed investigation into how each transporter interacts with a variety of drugs.
The team tested 23 commonly used medicines, and then trained a machine-learning model using this data and additional data from multiple drug databases. The model was designed to predict which drugs would interact with specific transporters based on the similarities in the chemical structures of the drugs.
Using this model, the team analyzed 28 current and 1,595 experimental drugs. The screening resulted in nearly 2 million predictions of potential drug interactions, including previously unidentified interactions. These predictions were validated using data from roughly 50 patients who had been prescribed one of these drugs in tandem with the antibiotic doxycycline.
Such remarkable results indicate that this new approach could be beneficial for identifying interactions between already in-use drugs. It could also be leveraged by pharmaceutical developers to fine-tune new drugs’ formulation to prevent unwanted interactions and enhance absorbability. Notably, a Boston-based biotech firm, Vivtex, has already begun using this technology to develop improved oral drug delivery systems. The research was partially funded by the U.S. National Institutes of Health, MIT’s Department of Mechanical Engineering, and the Division of Gastroenterology at Brigham and Women’s Hospital.