A Generalizable Platform for Adaptive Control of Complex Biological Systems

Researchers who study human diseases or test new drugs often use microfluidic devices that contain embedded cells that mimic the behavior of specific organs. The usual approach is to make a change in the cells’ environment and observe changes in the health of the cells. This project will expand that approach by finding ways to control the health of the cells as their environment changes. The project will create an “organ-on-a-controller” system that controls the health and function of human liver cells called hepatocytes. The system will integrate three components: 1) miniature sensors that monitor multiple vital signs of the hepatocytes in real-time, such as protein production and metabolite levels; 2) a computer model that learns how the cells respond to different drugs or nutrients; and 3) an intelligent control system that uses this knowledge to automatically adjust the input to the cells so that a particular cellular health state and function can be achieved. This approach will keep cells healthy and will guide unhealthy cells from a diseased state, such as fatty liver disease, back toward a healthy one. The technology will create a powerful tool that can accelerate the discovery of safer and more effective drugs, advance personalized medicine, reduce the need for animal testing, and provide a deeper understanding of complex chronic diseases. Results will help advance new concepts in biotechnology and advanced biomanufacturing. A fundamental gap exists in our ability to dynamically control complex biological systems. Current in vitro microphysiological systems (“organs-on-chips”) are largely open-loop, precluding the active regulation of cellular function based on real-time feedback. This project aims to address this knowledge gap by creating a first-of-its-kind “organ-on-a-controller” platform that integrates multiplexed biosensing, predictive modeling, and adaptive closed-loop control to actively steer cellular function. Using primary human hepatocytes as a biologically relevant model system, this project will design an integrated microfluidic platform for the simultaneous, real-time measurement of key secreted factors and intracellular reporters of transcription factor activity. Our approach will provide a continuous, multi-parameter view of the cellular state with high temporal accuracy. Further, a library of predictive mathematical models (transfer functions) will be developed that describe the dynamic input-output relationships of hepatocytes in response to metabolic and inflammatory stimuli. A sophisticated model predictive control will be implemented and validated to actively maintain hepatocyte homeostasis under inflammatory challenge and steer cells from a disease state toward a healthy phenotype. By closing the loop between sensing and actuation, the platform will be inherently adaptive, learning from cellular responses to account for biological variability and perturbation. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.