(B-41) Model guided design of a biological controller for septic wound healing regulation.

Introduction:: Wound healing is a dynamic and multicellular process of highly coordinated intracellular networks. Within the wound environment, innate signaling molecules control immune cell dynamics and activity, influencing the healing phases and regenerative outcomes. The immune system reacts when pathogenic infiltrations or any of their components cross the barrier separating the external and internal environments. The immune response is essential for the healing of infected wounds. However, when this immune-mediated defense mechanism against invasive pathogenic bacteria becomes dysregulated, it leads to widespread inflammation and organ damage due to sepsis. The detrimental impact of sepsis is self-enhancing by causing an imbalance in signaling and cellular dynamics, stimulating pathogen invasion, and exacerbating tissue injury. Available treatment options to regulate inflammation for chronic wounds have focused on delivering chemokines and growth factors. Despite initial success, it has been challenging to introduce these treatments for septic wound resolution into the clinic because of delivery issues and safety concerns. An alternative approach will be to dynamically regulate the immune system to lower cytokine production to promote tissue repair and regeneration as a novel approach to maintain sustained therapeutic release over an extended period.


Materials and Methods:: The hyper- and hypo-inflammatory phases of sepsis wound healing were modeled in this work using an ODE-based approach. All the simulations were conducted in MATLAB_R2022a and solved with an ode23s solver. For the local sensitivity analysis, we independently perturbed each of the 22 kinetic parameters in our model to five discrete values spanning five orders of magnitude. We conducted a 22×5= 110 simulation and analyzed the results using the metrics: Endpoint Cell Counts, Maximum Cell Counts, and Time to Peak. For the global sensitivity analysis, we randomly perturbed the most impactful kinetic parameters 𝛾,𝛾,𝛾,𝛽, and 𝛿, within ±50% of their nominal values with a uniform distribution with MATLAB command rand and performed 1000 independent simulations to investigate the conditions that lead to chronic healing dynamics. Inferring from this, we designed a controller and evaluated the controller performance; 1000 simulations with the controller were conducted using the same kinetic parameters used for the global sensitivity analysis.


Results, Conclusions, and Discussions:: Our controller allowed us to design the best control approach for managing the immune response, thus encouraging tissue regeneration in our ODE model that captured the wound healing process. Our data demonstrate that abnormal macrophage polarization causes a concurrent overabundance of monocytes, M1 and M2 macrophage phenotypes, thus leading to immune paralysis. Once the model was verified, we added a controller that changes the dynamics of chronic healing in favor of a standard, homeostatic immune response. Our model also demonstrates how

Our systems biology model also demonstrates how dysregulated macrophage count, notably the simultaneous excess of M0, M1, and M2 macrophages, causes subpar wound healing, resulting in recurring wound infections. According to our local and global sensitivity analysis, the model describes the severe dysregulation of immune cell dynamics and points out the best monitoring and regulating point to restore immunological paralysis. By altering the pace at which neutrophils differentiate into M0 in response to M2 dynamics, the suggested feedback biological controller can also restore the septic inflammation condition to the healed state.

In conclusion, this work demonstrates a system’s biology approach to define the design principles of an immunotherapeutic intervention. Identifying biomarkers and modeling their interactions with immunes gives insight into therapeutic targets for regulation. Our approach proposes a novel system capable of adaptively reversing septic wound healing dynamics.






Acknowledgements (Optional): : L. G. and D. D. are supported by ONR grant #N00014-21-1-2624 and by the startup fund provided by the Weldon School of Biomedical Engineering at Purdue University. X. T. is supported by NSF grant #2223720 and the startup fund provided by the Cain Department of Chemical Engineering at Louisiana State University.


References (Optional): :