For many of us, a hot cup of coffee is how we start our day. For Texas A&M Health researchers, it may also provide a new way to screen the modified cells in future drugs.
A team at the Texas A&M Health Institute of Biosciences and Technology has developed an artificial intelligence-engineered molecular switch that uses caffeine to quickly separate modified proteins inside living cells and trigger cellular responses on demand. The platform, called CODS, short for caffeine-activated cleavage system, could help scientists create safer and more controlled gene and cell therapies.
The study, published in Journal of the American Chemical Societyled by Yubin Zhou, MD, PhD, FAAAS, FAIMBE, FRSC, director of the Center for Translational Cancer Research at the Institute of Biosciences and Technology and professor at the Texas A&M Naresh K. Vashisht College of Medicine, along with Tianlu Wang, PhD, and colleagues. Graduate students Brendan McKee and Tatsuki Nonomura played central roles in the work, with McKee leading the AI-guided protein design and computational modeling efforts and Nonomura leading basic molecular engineering studies and live cell validation.
Artificial intelligence is changing the way we design biology. Instead of relying only on protein parts that already exist in nature, we can now design new mini-proteins with specific behaviors. Here, we used artificial intelligence to help turn caffeine into a precise trigger to control mechanical cells.”
Yubin Zhou, MD, Professor, Texas A&M Naresh K. Vashisht College of Medicine
AI as a molecular architect
The new project builds on Zhou’s previous caffeine-responsive technologies, but moves in a distinctly different direction.
Previous systems showed that caffeine could help integrate the manufactured proteins. However, CODS does the opposite: It uses caffeine to remove proteins. This difference matters because future therapies may need ways not only to activate the cells, but also to pause, silence or reset them when needed.
To make CODS, the team used AI-guided protein design to create a small synthetic binder that recognizes a caffeine-responsive protein unit. The binder holds the system together when caffeine is absent, and when caffeine is added, the proteins separate.
In this way, CODS acts like a molecular clasp. Without caffeine, the clasp remains closed. With caffeine, the button is opened.
“Many genetically encoded molecular tools act like accelerators,” Wang said. “CODS gives us something closer to a brake or pause button.”
High performance computers
The AI-driven design process required significant computing power. The team used protein design algorithms and molecular simulations to identify, evaluate and refine synthetic binders before testing the most promising candidates in living cells.
This work was enabled by the Texas A&M High Performance Research Computing (HPRC) service, which provided the computing power needed to run advanced AI-based protein design workflows at scale.
“High-performance computing was essential for this project,” Zhou said. “AI protein design is computationally demanding. The Texas A&M HPRC service helped us go from a conceptual idea to a functional molecular switch much faster.”
The resulting system responded to very low concentrations of caffeine, worked within minutes, and could be repeatedly reversed by adding or removing caffeine.
Control of genes, cell death and immune cells
The researchers demonstrated CODS in three main ways.
First, they used it to control gene activity. Without caffeine, a mechanical gene circuit remained active. When caffeine was added, CODS sequestered the target proteins needed to keep the gene activated, sharply reducing gene activity. Removing the caffeine allowed the system to recover.
Second, the team used CODS to control programmed cell death. By recombining a cell death protein with the caffeine-responsive switch, they created a system in which caffeine could trigger inflammatory cell death, known as pyroptosis. This could help scientists study inflammation and may one day support the design of therapeutic cells that can be eliminated when needed.
Finally, the most translational demonstration involved CAR-immune T cells engineered to recognize and attack cancer. CAR T-cell therapies have shown remarkable results in some blood cancers, but they can also cause serious side effects when immune system cells become too active. A caffeine-induced safety switch could give clinicians a way to temporarily reduce the activity of CAR T-cells without permanently destroying the therapeutic cells.
Using CODS, this Texas A&M team constructed a split CAR system that remains active when caffeine is absent, but remains passive when caffeine is added. In laboratory tests, caffeine significantly reduced CAR T-cell activation, suggesting that CODS could become a practical safety OFF switch for engineered immune cells.
Beyond coffee: Towards programmable medicine
Zhou emphasized that caffeine itself is not a cure for cancer. Instead, caffeine serves as a safe and familiar signal that can communicate with specially engineered cells.
“Coffee will not replace medicine,” Zhou said. “But caffeine can help us imagine drugs that are more controllable, more responsive and safer for patients.”
The broader advance is the use of artificial intelligence to design new proteins that behave in ways that nature does not readily provide. Similar strategies could eventually be used to construct switches controlled by other familiar molecules, over-the-counter or clinically approved drugs.
Before CODS can move toward clinical use, the system will need further testing in therapeutic cells, animal models, and disease-relevant environments. However, the study marks an important step towards programmable medicine by providing a framework for designing treatments that can be adjusted after they are delivered.
“Strong treatments need strong control,” Zhou said. “By combining AI-engineered proteins, high-performance computing, and familiar small molecules, we are building a new language for communicating with engineered cells.”
