A computational modeling approach for predicting multicell patterns based on signaling-induced differential adhesion

Abstract

Dynamically activated differential adhesion within cell populations enables the emergence of unique patterns in heterogeneous multicellular systems. This process has previously been explored using synthetically engineered heterogenous multicell spheroid systems in which cell subpopulations engage in bidirectional intercellular signaling to regulate the expression of different cadherins. While engineered cell systems provide excellent experimental tools to observe pattern formation in cell populations, computational models may be leveraged to explore the key parameters that drive the emergence of different patterns more systematically. Here, we developed and validated two-and three-dimensional agent-based models (ABMs) of spheroid patterning for cells engineered with a bidirectional signaling circuit regulating N-and P-cadherin expression. The model was used to predict how varying initial cell seeding, cadherin induction probabilities, or homotypic adhesion strengths between cells leads to different spheroid patterns, and unsupervised machine learning techniques were used to map system parameters to unique spheroid patterns. The model was also used as a tool to design new synthetic cell signaling circuits based on a desired final multicell pattern.

Parameterized Computational Framework for the Description and Design of Genetic Circuits of Morphogenesis Based on Contact-Dependent Signaling and Changes in Cell–Cell Adhesion

“…Others have done parameterization in different ways: heuristic parametric tuning 42 or machine learning. 43 Our method was to design screenings of parameters in meaningful ranges, motivated by biology and by computational considerations. It allowed us flexibility and exploration of a meaningful parameter space and was able to provide parameters with qualitative and quantitative matching.…”

“…The combination of signaling and morphological effectors has been shown to be at the core of complex developmental transitions: tissues are a complex system where cell–cell signaling affects morphogenesis and then morphogenesis feeds back to influence signaling to robustly generate complex multicellular structures . In fact, the combination of signaling and morphological effectors has been incorporated in several computational models and shown to be able to replicate the complex morphogenesis of embryonic transitions. ,, Given that the synthetic morphogenesis described in ref is one of the first synthetic systems that couples chemical and mechanical signaling for synthetic developmental trajectories in mammalian cells, it has attracted other computational descriptions recently, , with different computational systems or objectives compared to ours. The system in ref describes a similar system in a cellular Potts model, with more focus on the underlying design principles and what can be learned about the logic of these kinds of networks.…”

“…The system in ref describes a similar system in a cellular Potts model, with more focus on the underlying design principles and what can be learned about the logic of these kinds of networks. The system in ref is developed with predictive capacity and with some novel calibration structures and relies on a new computational model for modeling cell movements that is distinct from cellular Potts. We think these efforts will collectively bring the field closer to the goal of rationally designing genetic networks for synthetic developmental trajectories.…”

“…Until recently, no example of a computational system for description at the level of genetic circuits of morphogenesis was available for assisting with design. The in vitro system in Toda seems perfectly placed to provide a case study as it has all of the features of a minimal “toy model” of synthetic circuit-guided morphogenesis (and it has attracted efforts from other groups as well , ): it uses mammalian cells (mouse fibroblast line L929), it has logically minimal circuits, shows the genotype-to-phenotype relationships, and has explored the phenotypic consequences of changes in either cell–cell adhesion and/or network topologies.…”

A parametrized computational framework for description and design of genetic circuits of morphogenesis based on contact-dependent signaling and changes in cell-cell adhesion

“…Until recently, no example of a computational system for description at the level of genetic circuits of morphogenesis was available for assisting with design. The in vitro system in Toda seems perfectly placed to provide a case study as it has all the features of a minimal "toy model" of synthetic circuit-guided morphogenesis (and it has attracted efforts from other groups as well 42,43 ): it has logically minimal circuits, shows the genotype-to-phenotype relationships, and has explored the phenotypic consequences of changes in either cell-cell adhesion and/or network topologies.…”