UC Berkeley

Synthetic Biology seeks to apply engineering principles to the design of biological systems, such that new behaviors can be predictably and reliably designed. Increasingly, systems are being engineered with well-characterized, modular, independently tunable parts. This bottom-up approach allows us to test design principles as well as engineer systems that may be adapted for use in other contexts. Modular design, in which systems are connected through a small number of well-defined interactions, is an important engineering principle that aids the design of complex systems. As we work toward the design of sophisticated systems necessary to exist in complex environments, it will become important to utilize many layers of regulation, as is done ubiquitously in natural systems. Post-translational control is relatively under-utilized in Synthetic Biology, in part due to the often complex interactions distributed across protein-protein interfaces that determine connectivity. Eukaryotic signal transduction pathways, however, are generally organized around modular domains and scaffolding proteins that improve evolvability by allowing reuse of signaling parts and potential for pathway rewiring via recombination1. The power of modular pathway design has been demonstrated by recent successes in engineering eukaryotic scaffolds to serve as signaling hubs in order to create sophisticated behavior2. Using well-characterized parts, we seek to apply the modular strategies evolved in higher organisms to control signal transduction in prokaryotic systems.

We begin our discussion by detailing the characteristics of biological parts families that are ideal for engineering. We then explain how protein-protein interaction parts that display these characteristics can be composed to create scaffolds capable of spatially organizing enzymes. Systems organized around scaffolds appear to function efficiently at relatively low component concentrations. We found that at low expression levels, the eukaryotic parts we employ often contain internal translation initiation sites that are selected against in prokaryotes and can be removed with the introduction of mRNA secondary structure. Next we used simple synthetic scaffolds, built entirely from well-characterized protein-protein interaction motifs, to demonstrate that an increased local concentration effect, via tethering, is sufficient to direct signaling specificity. Scaffold-directed signaling was sensitive to expression levels of each component, thus additional regulation is necessary for robust behavior. We describe a strategy for introducing a peptide ligand into a polypeptide fold without perturbing activity that can be used for engineering an intra-molecular, autoinhibitory interaction. Robustness to varying component expression levels was gained by engineering an autoinhibitory interaction into the kinase protein such that the scaffold both colocalizes and activates components. Such domain-based allosteric regulation is frequently seen in natural eukaryotic signaling proteins. We also discuss a number of other strategies that used simple parts, which were able to improve some aspects of system behavior but not overall robustness. Interestingly, many natural systems that exhibit robust behavior make use of bifunctional components, and we found that designing a bifunctional scaffold (i.e. tethering and activation) improved robustness, while strategies that involved the expression of additional simple components failed to increase overall robustness. Taken together these results are among the first steps in designing highly modular, reliable signal transduction pathways in prokaryotes and demonstrate that principles governing pathway control of natural systems in higher organisms can be generalized and applied with well-characterized parts to prokaryotic systems.