Semi-synthetic mammalian gene regulatory networks

In recent years gene network engineers have celebrated spectacular success: Genetic devices such as epigenetic toggle switches and oscillating networks have been engineered and pioneered a new ever-increasing scientific community known as synthetic biology. While synthetic biology was until recently restricted to network assembly and testing in prokaryotes, decisive advances have been achieved in eukaryotic systems based on current availability of different human-compatible transgene control technologies. Most prominent examples include the epigenetic gene network enabling metastable fully inheritable transgene expression states in mice, artificial regulatory cascades managing multi-level expression control and Boolean-type BioLogic gates supporting near-digital expression readout. The majority of transgene control networks available to date are fully synthetic and integrate artificial extracellular signals in a desired host metabolism-independent manner. Yet, in order to develop their full anticipated therapeutic potential, synthetic transgene control circuits need to be well interconnected with the host cell's regulatory networks in order to enable physiologic control of prosthetic molecular expression units. We have designed three semi-synthetic transcription control networks able to integrate physiologic oxygen levels and artificial antibiotic signals to produce expression readout with NOT IF or NOR-type Boolean logic or discrete multi-level control of several intracellular and secreted model product proteins. Subtle differences in the regulation performance of the endogenous oxygen-sensing system in CHO-K1 and human HT-1080 switched the semi-synthetic network's readout from a classic four-level (high, medium, low, basal) regulatory cascade to a network enabling six discrete transgene expression levels. These findings are in excellent correspondence with a mathematical model. Prosthetic networks, precisely embedded in host regulatory networks and co-fine-tuned by physiologic as well as pharmacologic input signals, will foster future advances in gene therapy and tissue engineering. (c) 2005 Elsevier Inc. All rights reserved.