Agent-based model predicts layered structure and 3d movement work synergistically to reduce bacterial load in 3d in vitro models of tuberculosis granulomas

Alexa Petrucciani

Purdue University

"Agent-based model predicts layered structure and 3d movement work synergistically to reduce bacterial load in 3d in vitro models of tuberculosis granulomas"

Tuberculosis (TB) caused over 1.6 million deaths in 2021. TB is associated with granulomas, organized structures of immune cells that contain the causative bacteria. These structures are three-dimensional with an inner core of macrophages and an outer cuff of T cells. Advanced 3D cell cultures have been applied to emulate these clinical structures in vitro. One in vitro approach showed that 3D spheroid models have improved bacterial control compared to traditional in vitro infection models. We use hybrid modeling to simulate these spheroid models and traditional counterparts, with an agent-based model of immune cell and bacteria rules coupled to a partial differential equation model of chemokine diffusion. We calibrate our model to experimental data while enforcing shared parameters between the spheroid and traditional setups, only changing the initial structure and movement rules to reflect the experimental setups. Lower bacterial load in spheroid simulations as compared to traditional culture simulations is predicted to be due to increased proportions of activated macrophage killing of bacteria, either in tandem with increased proportions of CD8+ T cell activation or not. The spatial distribution of cells was found to be an important factor in macrophage and CD8+ T cell activation in spheroid simulations, with more activation being associated with increased proximity. Next, an in silico experiment was performed, where the initial structure and movement rules were uncoupled to see if either of these independently lead to bacterial control. Neither of the uncoupled mechanisms reduced bacterial load on its own, rather they worked together synergistically. This work further emphasizes the impacts of spatial organization and dimension in biological processes, while highlighting the flexibility of in silico modeling and the perturbations it makes possible.
Additional authors: Alexis Hoerter, Weldon School of Biomedical Engineering, Purdue University; Leigh Kotze, DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medical and Health Sciences, Stellenbosch University, Cape Town, South Africa; Nelita Du Plessis, DSI-NRF Centre of Excellence for Biomedical Tuberculosis Research, South African Medical Research Council for Tuberculosis Research, Division of Molecular Biology and Human Genetics, Faculty of Medical and Health Sciences, Stellenbosch University, Cape Town, South Africa; Elsje Pienaar, Weldon School of Biomedical Engineering, Purdue University and Regenstrief Center for Healthcare Engineering, Purdue University

---