PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 88 NR 2/2012 105 Branko ŠTER, Rok GABER, Monika AVBELJ, Roman JERALA, Andrej DOBNIKAR University of Ljubljana Design of information processing in cells using artificial gene repressors Abstract. The progress of synthetic biology allows one to design artificial repressors that inhibit selected genes. Combination of repressors enables construction of NOR logical gates that could form the foundation for information processing within cells. The theoretical potentials and limitations of constructing NOR gates were analyzed. They could be experimentally realized in bacterial cells. The number of required artificial repressors was analysed and temporal simulations of an example function were performed. Streszczenie. Postęp biologii syntetycznej pozwala zaprojektować sztuczne represory, które hamują rozwój wybranych genów. Połączenie represorów umożliwia budowę bramek logicznych NOR, które mogą stanowić podstawę do przetwarzania informacji w komórkach. Teoretyczny potencjał i ograniczenia budowy bramy NOR były analizowane. Mogą one być realizowane w komórkach bakteryjnych. Dokonano analizy liczby wymaganych represorów i wykonano symulacje funkcji przykładowych. (Projektowanie przetwarzania informacji w komórkach przy użyciu sztucznych represorów) Keywords: information processing, synthetic biology, artificial repressors, logical gates Słowa kluczowe: przetwarzanie informacji, syntetyczna biologia, sztuczne represory, bramki logiczne Introduction Synthetic biology and computer science are beginning to collaborate in designing logical functions within the living cells. The combination of several repressors allows the construction of NOR gates, which is the universal operator for building any logical function. The paper shows the original procedures that are relevant to design any possible logical function of three input variables with NOR gates implemented with repressors. Such implementation could be experimentally realized in bacterial cells. For the purpose of this paper, however, only simulation results of the proposed circuits are shown. The paper is organized as follows. The biological system for constructing a NOR gate with artificial repressors and for building any circuit of NOR gates is shown first, based on the scheme of biological processes and related equations. They are transformed into the differential equations, which are used for the simulation purposes. It is followed by the description of the genetic approach, for designing of all possible logical functions of three variables with NOR operators, together with some imposed limitations that are important for the experimental implementation, such as the maximal number of inputs to the NOR operators, maximal number of layers and operators and the minimal number of the zinc finger based repressors. Next, the results of genetic design are given for two runs, one with limitation of up to two inputs per NOR gates and the other with limitation of up to three inputs per NOR gates. In both cases, however, the minimal number of zinc finger repressors and the minimal number of inputs are searched for all possible functions of three variables. The simulation results of some optional function realized with genetically designed NOR circuits based on zinc finger repressors are finally given, followed by conclusion, where the recapitulation and the plans for future work are outlined. Construction of NOR based circuits with zinc finger repressors Regulation of the cellular response to the combination of input signals can be achieved at several levels, including protein and DNA level. While the response based on proteins is significantly faster then DNA-based response, the latter, which involves regulation of gene expression, allows using the same type of modules at several levels. The basic type of regulation involves transcriptional repression, where the constitutive gene expression is prevented by binding of a repressor to the appropriate region in front of the selected gene [1]. Cells contain many different repressors that bind to specific sequences and can be used to engineer modified response, however the number of well-characterised natural repressors limits the achievable complexity of synthetic information processing networks. Zinc fingers are protein domains that bind to a specific nucleotide sequence. Their main advantage is that they are modular, which allows us to design thousands of different variants that bind to selected sequence [2-4]. Therefore zinc fingers can be used as artificial repressors that limit the expression of selected genes, which are engineered to contain the zinc finger binding site [5]. This removes the limitation to naturally available repressors what is the main weakness of previously reported artificial bio- logic circuits [6-8]. Using zinc finger based transcriptional repressors is the foundation of complex artificial gene regulatory networks as information processing circuit. Introduction of two different zinc finger binding sites in the gene regulatory region creates a logical NOR gate with two inputs, which is open only in the absence of both zinc finger repressors, while any or both of them shut down gene transcription. If the gene in question is coding for another zinc finger repressor it can be used in the next level of information processing, allowing constructing, in principle, any level of logical information processing complexity. Biological processes that are relevant to function of the artificial gene regulatory networks include: 1. Binding of zinc fingers to the gene regulatory (operator) region, which is a reversible reaction. In case of NOR gate, different combinations of zinc fingers can be bound or released. 2. Transcription of the gene in the absence of bound zinc finger repressor, which involves binding of RNA polymerase and its processing and release of the mRNA 3. Translation of the mRNA into the protein (another zinc finger) 4. Proteolytic degradation (turnover) of the zinc finger The following figures present the bio-schemes of the 2-input NOR gate (Fig. 1) and the function f 19 (Fig. 2), shown also in standard form in Fig. 5. Each of the species, either being a single molecule or molecular complex can be formed or degraded through one or several pathways, which can be represented by differential equations. Rate constants and concentrations of each species can be estimated from the literature data or empirically determined. Experimental readout of the