TY - JOUR T1 - Transcription factor loads tend to decrease the robustness of stable gene transcription networks JF - bioRxiv DO - 10.1101/037564 SP - 037564 AU - Shaoshuai Mou AU - Domitilla Del Vecchio Y1 - 2016/01/01 UR - http://biorxiv.org/content/early/2016/01/21/037564.abstract N2 - Robustness of a system’s behavior to changes in parameter values is a remarkable property of natural systems and especially desirable when designing de novo synthetic gene circuits. Loads on transcription factors resulting from binding to target promoters have been shown to significantly affect the effective time constants of gene transcription networks and to thus alter the overall system’s behavior. Here, we employ models that explicitly account for load effects to investigate how these impact the robustness of a stable gene transcription network to parameter perturbations. By employing a combination of sampling-based methods and analytical tools from control theory, we demonstrate that the presence of loading shrinks the region of parameter space where a gene circuit performs the desired function. A number of multi-module synthetic gene circuits are analyzed to demonstrate this point, including an event detector and a molecular signature classifier. These results indicate that for designing genetic circuits that are robust to parameter uncertainty it is highly desirable to find ways to mitigate the effects of transcription factor loading.Author Summary Efforts to understand how loads affect gene transcription networks have been underway in the past decade. Here we perform a numerical investigation on three synthetic gene circuits to show that loads tend to decrease the robustness of stable gene transcriptional networks. We complement the numerical findings with analytical derivations that employ the stability radius to compare the robustness of different networks to parameter perturbations near an equilibrium point. Consistent with the numerical finding, the analytical results support that systems with substantial transcription factor loading have smaller stability radius (less robustness) than systems without loading. ER -