Single Cell Dynamic Boolean Network Modeling of T Cell Development Predicts Heterogeneous Transcriptional Trajectories
Langley, Matthew 1, 2 ; Shukla, Shreya 1, 2 ; Yachie-Kinoshita, Ayako 1, 2, 3 ; Zandstra, Peter W. 1, 2, 4
1. Institute of Biomaterials and Biomedical Engineering, University of Toronto; 2. Donnelly Centre for Cellular and Biomolecular Research, University of Toronto; 3. Systems Biology Institute, Tokyo, Japan; 4. Medicine By Design, A Canada First Research Excellence Program at the University of Toronto
The T cell development program drives hematopoietic progenitors toward T lineage commitment using dense networks of genes and proteins that respond to environmental signals. Although existing static models of these networks identify the major players and their pairwise interactions, they provide limited explanatory insight into the multi-stage dynamics of T cell specification. In this study, we use a computational approach based on Boolean networks (BNs) to simulate aspects of the T cell development program in silico and explore the dynamic response of progenitor cells under normal and perturbed conditions. A BN representation of the regulation of each gene in the network using AND/OR logic was constructed from microarray data and previous literature. By using asynchronous updates to simulate this BN, we mapped the transcriptional state space that developing T cells can traverse under various combinations of environmental inputs (such as Notch and IL-7 signaling). This state space is consistent with single cell qRT-PCR observations, and steady states within the space resemble known stages of T cell development. Notably, the BN model suggests the T cell development program permits multiple independent transcriptional trajectories toward T lineage commitment that involve different transient cell states. Furthermore, our analysis identifies genetic perturbations (ex. Bcl11b-/-) that may select for particular developmental trajectories or arrest development at specific stages. BN modeling presents a powerful advance over previous static models for exploring routes through the transcriptional space involved in T cell development and suggests new opportunities for improved in vitro T cell differentiation protocols.