A Self-regulatory System of Interlinked Signaling Feedback Loops Controls Mouse Limb Patterning

Abstract

Developmental pathways need to be robust against environmental and genetic variation to enable reliable morphogenesis. Here, we take a systems biology approach to explain how robustness is achieved in the developing mouse limb, a classical model of organogenesis. By combining quantitative genetics with computational modeling we established a computational model of multiple interlocked feedback modules, involving sonic hedgehog (SHH) morphogen, fibroblast growth factor (FGFs) signaling, bone morphogenetic protein (BMP) and its antagonist GREM1. Earlier modeling work had emphasized the versatile kinetic characteristics of interlocked feedback loops operating at different time scales. Here we develop and then validate a similar computational model to show how BMP4 first initiates and SHH then propagates feedback in the network through differential transcriptional regulation of Grem1 to control digit specification. This switch occurs by linking a fast BMP4/GREM1 module to a slower SHH/GREM1/FGF feedback loop. Simulated gene expression profiles modeled normal limb development as well those of single-gene knockouts. Sensitivity analysis showed how the model was robust and insensitive to variability in parameters. A surprising prediction of the model was that an early Bmp4 signal is essential to kick-start Grem1 expression and the digit specification system. We experimentally validated the prediction using inducible alleles and showed that early, but not late, removal of Bmp4 dramatically disrupted limb development. Sensitivity analysis showed how robustness emerges from this circuitry. This study shows how modeling and computation can help us understand how self-regulatory signaling networks achieve robust regulation of limb development, by exploiting interconnectivity among the three signaling pathways. We expect that similar computational analyses will shed light on the origins of robustness in other developmental systems, and I will discuss some recent examples from our ongoing research on developmental patterning.