AI-designed molecular switch uses caffeine to control engineered cells
Researchers at the Texas A&M Health Institute of Biosciences and Technology have developed a caffeine-operated dissociation system (CODS). This AI-designed molecular switch uses caffeine to separate engineered proteins in living cells, providing a way to pause or reset gene and cell therapies, including a potential safety “OFF” switch for CAR T-cells.
How does the CODS molecular switch work?
The CODS platform acts as a molecular clasp. According to the research team, the system stays closed when caffeine is absent, holding engineered proteins together.
When caffeine is introduced, the clasp opens and the proteins separate. This triggers specific cellular responses on demand. Tianlu Wang, PhD, noted that while many molecular tools act as accelerators, CODS functions more like a “brake or pause button.”
Why is AI-driven protein design significant?
The team used AI to design synthetic mini-proteins with specific behaviors rather than relying solely on parts found in nature. This allowed them to create a small synthetic binder that recognizes a caffeine-responsive protein module.
This process required massive computational power provided by the Texas A&M High Performance Research Computing (HPRC) service. Yubin Zhou, MD, PhD, stated that this service helped the team move from a conceptual idea to a functional switch much faster.
How could caffeine control cancer therapies and gene activity?
The researchers demonstrated the system’s utility in three distinct ways. First, they used it to reduce gene activity by separating the proteins needed to keep an engineered gene circuit turned on.
Second, the team triggered pyroptosis, a form of inflammatory cell death, by rewiring a cell-death protein. This could help scientists study inflammation or design therapeutic cells that can be eliminated when necessary.
Finally, the team applied CODS to CAR T-cells, which are immune cells engineered to attack cancer. While CAR T-cell therapies are effective for some blood cancers, they can cause serious side effects. In laboratory tests, caffeine reduced CAR T-cell activation, suggesting it could serve as a practical safety switch for clinicians.
What happens next for programmable medicine?
Yubin Zhou emphasized that caffeine is not a cancer treatment itself, but a safe signal used to communicate with engineered cells. He noted that coffee will not replace medicine, but it helps imagine safer, more responsive treatments.
A possible next step involves testing the system in animal models and disease-relevant settings before any clinical use. Similar AI strategies may eventually be used to build switches controlled by other over-the-counter drugs or clinically approved medicines.
The study, published in the Journal of the American Chemical Society, provides a framework for therapies that can be adjusted after they are delivered to a patient.
Frequently Asked Questions
Is caffeine used as a direct treatment for cancer in this study?
No. According to Yubin Zhou, caffeine serves as a safe and familiar signal to communicate with specially engineered cells, not as a cancer treatment.
What is the primary purpose of the CODS platform?
The caffeine-operated dissociation system (CODS) is designed to rapidly separate engineered proteins inside living cells to trigger cellular responses on demand.
How does CODS improve CAR T-cell therapy?
It could provide a safety “OFF” switch, allowing clinicians to temporarily reduce the activity of engineered immune cells to prevent serious side effects without permanently destroying the cells.
How do you feel about the prospect of using common molecules like caffeine to control advanced medical therapies?