Applying Evolutionary Principles to Stabilize Genetic Circuits

A fundamental difference between an electric circuit and a genetic one is that while the former is stable, the latter is subject to spontaneous  mutations and ensuing instability caused by adaptation. This fragility is pronounced when the human-desired function that is engineered to the genetic circuit conflicts with fitness — the ability to reproduce — of the organisms that express the circuit. This problem is present in virtually all engineered biological systems. Works in synthetic biology from the recent decade have chartered the basic framework for the functionality of genetic circuits, such as logic gates, feedback loops, noise control; they were all done as proof of concept at the small scale in laboratory, thereby risking serious fragility. For instance, the synthetic gene oscillator, an early breakthrough in the field, could only oscillate for no more than a few cycles before breaking down. Hence, the robustness of synthetic systems is becoming a bottleneck to delivering discoveries from laboratories to more complex and demanding settings of industry and medicine. Here, we propose a novel approach to address robustness where knowledge about how organisms evolve and how competition occurs is integrated into the design of synthetic biological systems. Indeed, we have identified several evolution-aware principles of design that, according to our modeling, should improve robustness by several to effectively infinite folds, depending on the situation. Consider that a mutant carries two mutations and has a high fitness, and mutants that carry either of the two mutations have worse fitness than wildtype that carries no mutations. Evolution is short-sighted: only mutants with increased fitness are selected and capable of sweeping the population. Hence, assuming mutations occur one at a time, this short-sightedness dictates that the system is trapped in the local fitness optimum represented by the wildtype and that the double mutant, despite its large fitness, is unlikely to invade wildtype. Therefore, engineering a bacterial system that expresses a human-desired function to be a local fitness optimum should improve its stability. We believe our novel approach constitutes a missing link that is needed to potentiate the exciting findings emerging from synthetic biology to the real world.