Journal: Frontiers in Bioengineering and Biotechnology

Article Title: Experimental Control of Macrophage Pro-Inflammatory Dynamics Using Predictive Models

doi: 10.3389/fbioe.2020.00666

Figure Lengend Snippet: Conceptual diagram of modeling immune response in health and disease. (A) Immune response as dynamically regulated in health (left) and dysfunctional in chronic conditions (right). (B) Block diagram with macrophages as the “system” or “plant” that is being controlled. (C) Identification, validation, and prediction of inflammatory response as a three-step process consisting of (1) design of an engineering model structure and fit of model parameters, (2) comparison of predicted and experimental results, and (3) use of the predictive model to design input strategies to obtain a desired response.

Article Snippet: The Model Predictive Control toolbox in MATLAB (2018b) was used to create the controller and define manipulated input sequences for the MISO “global” model.

Techniques: Blocking Assay, Biomarker Discovery, Comparison

Journal: Frontiers in Bioengineering and Biotechnology

Article Title: Experimental Control of Macrophage Pro-Inflammatory Dynamics Using Predictive Models

doi: 10.3389/fbioe.2020.00666

Figure Lengend Snippet: Open-loop control of pro-inflammatory macrophage activity is experimentally achieved using a nested multiple regression. (A) RAW 264.7 macrophage temporal response to 1 μg/mL LPS and 100 ng/mL IFN-γ. (B) Model designed inputs u 1 and u 2 using hysteresis-free model, which reflects cells beginning in a naïve state. (C) Hysteresis-free model response to inputs defined in (B) . (D) Model designed inputs u 1 and u 2 using first generation model accounting for hysteresis, which reflects cells starting from a non-naïve 24 h IL-4 primed state. (E) Hysteretic model (red) and non-hysteretic model (blue) responses to inputs defined in (D) . (F) Experimental delivery of designed inputs in (D) reflects predicted control output (E) for both hysteretic IL-4 primed (red curve, mean ± SEM, N = 16; interpolated curve ± RMS CV error) and non-hysteretic (blue curve, mean ± SEM, N = 16; interpolated curve ± RMS CV error) RAW 264.7 macrophage cultures. (G) Representative images of iNOS staining in model predictive control experiments using the inputs in (D) . (H) Simulation of updated 2nd generation model with dynamic supra-additivity term in response to designed inputs (D) captures experimental RAW 264.7 iNOS expression for both hysteretic (red curve) and non-hysteretic (blue curve) systems. (I) Experimental validation of the second-generation global model. Delivery of inputs designed to maintain a constant unit output of iNOS in a hysteretic system using the new model (inputs shown in ) improves control output for both hysteretic IL-4 primed (red curve, mean ± SEM, N = 8; interpolated curve ± RMS CV error) and non-hysteretic (blue curve, mean ± SEM, N = 8; interpolated curve ± RMS CV error) macrophage cultures.

Article Snippet: The Model Predictive Control toolbox in MATLAB (2018b) was used to create the controller and define manipulated input sequences for the MISO “global” model.

Techniques: Control, Activity Assay, Staining, Expressing, Biomarker Discovery