To date, the central focus of systems biology has been on developing computational simulations of biological phenomena within single cells (eg. signaling pathways, metabolic processes), and this approach is gaining popularity and has already generated new understanding of how intracellular systems function that was not achievable using experimental approaches alone. However, equally important in biological processes, and especially central to how biological tissues grow and adapt, are the multi-cell interactions—cells interacting with other cells and cells responding to their dynamic and heterogeneous tissue environment. We have developed agent-based simulations to identify fundamental principles (‘modules’) of tissue patterning processes, including embryogenesis and angiogenesis, across different vertebrate tissues. Physiological and pathological tissue patterning responses are emergent phenomena resulting from an aggregate of individual cellular behaviors, such as migration, proliferation, differentiation, and apoptosis, which are ultimately controlled by molecular cues originating from the cells’ genomes and from the local heterogeneous tissue environment. Agent-based computational modeling is particularly well suited to the study of tissue patterning processes because it is premised on the concept that local interactions of autonomous members of a population, or agents (a.k.a. biological cells), give rise to global or emergent phenomena. The underlying philosophy of this “bottom-up” modeling technique is that relatively simple rules for agent interaction generate complex systems-level outcomes observed in the real world (eg. tissue lengthening, thickening, and branching). In the simulation, as in the biological system, each agent, or cell, is capable of sensing local environmental cues, integrating that information with their own state history, and making independent decisions from one another, all within a framework that computes the outcomes of these decisions in space and time. Moreover, agent-based modeling provides a tool for accessing information that is not easily obtained through experimental observation. For example, dynamic information about individual cells’ protein expression states or phenotypic lineages within a simulated whole tissue can be tracked and recorded with time. We have applied this modeling approach to study blastocoel roof thinning in the Xenopus embryo and to study capillary sprouting in adult mammalian tissues, and we have conducted independent experiments to validate the simulation predictions. Early indications are that agent-based modeling is a viable technique for assessing central mechanisms underlying tissue patterning, and challenges and caveats for this approach will be discussed, as well.
Audio (MP3 File, Podcast Ready)
Copyright. All Rights Reserved.