Kinetic negative-feedback-based barrier membrane attenuates pathological fluctuations and promotes bone regeneration in diabetes

The regeneration of diabetic bone defects remains challenging due to a mismatch between dynamically fluctuating pathological cues and passively responsive, open-loop drug release. Therefore, a negative-feedback model upon fluctuating pathology is needed to guide the design of drug-loaded biomaterials and to predict their closed-loop therapeutic performance. Here, we establish a negative biofeedback system where pathology acts as the input signal and therapeutic drugs serve as the feedback signal. A theoretical model is developed a priori by coupling exponential drug release kinetics with a closed-loop drug-pathology interaction. Simulations show that this design drives dynamic drug release in response to pathological fluctuations, and the released drugs targeting pathology in turn effectively attenuate and smooth pathological fluctuations. As a proof-of-concept, we fabricate a barrier membrane implementing the negative biofeedback system. The membrane dynamically adjusts its degradation rate and drug release in response to neutrophil extracellular traps (NETs). Model simulations parameterized by our experiments indicate that this negative-feedback behavior effectively attenuates NETs fluctuations. In vitro and in vivo experiments confirm that the membrane efficiently degrades fluctuating NETs and reverses NETs-mediated tissue damage, thereby enhancing bone regeneration in diabetic defects. In conclusion, this study provides a theoretical drug delivery framework and a negative-feedback model for material design and tissue regeneration under complex pathological conditions. STATEMENT OF SIGNIFICANCE: The repair of diabetic bone defects remains a significant clinical challenge. A key limitation is that most conventional biomaterials follow fixed drug-release kinetics and therefore cannot adapt to fluctuations in pathological cues. To address this limitation, we introduce a negative-feedback drug-delivery paradigm and establish a coupled mathematical model that integrates pathology-responsive release with feedback regulation to simulate closed-loop control. Guided by this framework, we fabricated a neutrophil extracellular traps (NETs)-responsive barrier membrane embedding a negative-feedback system. By integrating in silico simulations, in vitro experiments, and an in vivo diabetic bone-defect model, we demonstrate that this barrier membrane dampens NETs fluctuations via negative feedback and significantly promotes bone regeneration in diabetes. This study provides a generalizable framework for developing negative-feedback biomaterials capable of autonomously adapting to fluctuating and complex pathological microenvironments.