An Experimental-Computational Framework to Characterize PBMCs Migration in the Glioblastoma Microenvironment
Please login to view abstract download link
One of the main physiological drivers of Glioblastoma (GB) progression is the interaction of the immune system with the tumor microenvironment. Tumor-secreted chemoattractants establish biochemical gradients that guide the migration and infiltration of the immune system, which in turn may affect the tumor dynamics. Characterizing how these signals, coupled with extracellular matrix (ECM) properties, modulate immune cells motility is essential for understanding immune activation and suppression [1]. This work presents an integrated pipeline that combines in vitro microfluidic experiments with agent-based modeling (ABM) to characterize immune cell dynamics and the role of ECM mechanical properties as moderator variables. Experimental data were obtained using a microfluidic device with a central chamber and two lateral channels. GB cells were seeded in one of the channels to establish a chemoattractant gradient, while Peripheral Blood Mononuclear Cells (PBMCs) were introduced into the opposite side of the chamber. Cell trajectories were extracted from experimental imaging using the Trackpy library [2] and an ABM incorporating both stochastic and biased migration was used to explain the results, developed with the software PhysiCell [3]. To bridge the gap between experimental observations and computational predictions, we developed a workflow to calculate statistical Quantities of Interest (QoI) and to calibrate the model parameters of an ABM. The methodology was validated first, using synthetic data, and then applied to real experiments. This framework allows the calibration of the migration parameters, enabling the model to reproduce the movement observed in vitro. The in silico tool provides a platform to test hypothesis, and for the prediction of PBMC behavior under varying cell densities, and initial configurations. Furthermore, it offers insights into the cell-level effects of tumor-secreted factors on immune motility. By providing a scalable method for parameter estimation from experimental trajectories, this framework is applicable to different cell-cell interaction scenarios, supporting the development of model-informed experimental designs.
