Biochemical Controller Made From DNA

Abstract

The potential of robots operating at a molecular or cellular scale is only limited by the imagination — for instance, nanorobots could navigate the bloodstream, identify a tumor and eliminate it cell by cell resulting in cancer treatment with minimal side effects. To perform such complex tasks, nanorobots need sensors for detecting their environment, actuators that allow them to move through their environment and embedded control circuits that convert sensor information to motor activity. In this thesis we focus on developing systematic design strategies for the de novo construction of embedded molecular controllers with DNA nanotechnology (introduced in Chapter 1). To systematically engineer DNA computing systems, we developed a new class of programmable DNA circuitry that, in principle, can implement any behavior captured by chemical reaction networks (CRNs) (Chapter 2). Although CRNs have been widely used as a framework for describing and modeling the time evolution of chemical systems, we were the first to show that CRNs can also serve as a prescriptive language for specifying many complex computations, including oscillations, memory, and distributed algorithms. We thus treated CRNs as a programming language, allowing us to design our DNA circuitry while abstracting away from the molecular details. To demonstrate our approach experimentally, we constructed DNA circuits to implement a CRN that embodies, at the molecular level, an algorithm used in distributed control systems for achieving consensus between multiple agents. Having made significant progress in the engineering of DNA computing systems in vitro, we next began to explore the design principles for adapting DNA circuitry to a much more complex environment, the mammalian cell (Chapter 3). Applying DNA circuitry in vivo, however, requires special considerations to minimize unwanted interference from host cellular activity. We showed that the use of modified RNA bases and backbones greatly enhances circuit performance in cells. Building on this breakthrough, we constructed nucleic acid-based AND and OR logic circuits and demonstrated that they function predictably and reliably within cells. Our work is a first step toward porting the rich toolbox of DNA nanotechnology into live cells.