Synthetic biosensing systems

The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

Contents lists available at ScienceDirect

The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel

Review

Synthetic biosensing systems Mario Andrea Marchisio ∗ , Fabian Rudolf ∗∗ Department of Biosystems Science and Engineering, ETH Zurich, Mattenstrasse 26, 4058 Basel, Switzerland

a r t i c l e

i n f o

Article history: Received 4 August 2010 Received in revised form 12 November 2010 Accepted 16 November 2010 Available online 23 November 2010 Keywords: Synthetic biology Sensing systems Biosensors Signaling pathways Circuit design

a b s t r a c t An essential feature of synthetic biology devices is the conversion of signals from the exterior of the cell into specific cellular events such as the synthesis of a fluorescent protein. In the first synthetic gene circuits, signal transduction was accomplished via inducible or repressible transcription factors. Today, these rather simple transcription networks are the basis for the construction of more sophisticated devices that, for instance, couple artificial gene circuits with cellular pathways to create a biosensing moiety. In the future, completely artificial signaling pathways will give us the possibility to control cellular processes in a direct, precise and reliable way. At present, numerous pathway components such as receptors, adapters, scaffolds and their interactions with ligands and other signaling proteins have been already characterized and, in some cases, reengineered. In addition, important results have been obtained by rewiring pathways and building more complex gene networks—such as “cell phones” and ecosystems—based on synthetically induced cell–cell communication mechanisms. Furthermore, RNA-interference and lightdependent control of transcription factors have become new instruments to integrate different signals and better regulate protein synthesis. Overall, synthetic biology of sensing systems appears to be in continuous evolution. Nevertheless, rapid improvements on the available DNA-recombinant technology are essential to achieve, within few years, a full engineering of cell transduction pathways. © 2010 Elsevier Ltd. All rights reserved.

Contents 1. 2.

3. 4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transcriptional and translational biosensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Exogenous control—prokaryotic cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Exogenous control—eukaryotic cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Cell–cell signaling—prokaryotic cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4. Cell–cell signaling—eukaryotic cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5. Post-translational biosensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6. Signaling across the membrane—receptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7. Intracellular signal transduction—Notch receptors, adapters, signaling proteins and their docking interactions . . . . . . . . . . . . . . . . . . . . . . . . . Computational design of biosensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

311 311 311 313 313 314 314 314 315 316 317 317

Abbreviations: RBS, ribosome binding site; GPCR, G protein coupled receptor; RTK, receptor tyrosine kinase; MAPK, mitogen-activated protein kinase; GDP, guanosine diphosphate; GTP, guanosine triphosphate; GEF, guanine nucleotide exchange factor; GAP, GTPase activating protein; PoPS, polymerase per second; RiPS, ribosome per second; TetR, tetracycline repressor; aTC, anhydrotetracyclin; LacI, lactose repressor; IPTG, isopropyl ␤-d-1-thiogalactopyranoside; AraC, arabinose activator; AHL, Nacyl homoserine lactone; AI-1, N-3-oxohexanoyl-l-homoserine lactone; LuxR, luciferase activator; PIT, pristamycin-dependent transactivator; tTA, tetracycline-dependent transactivator; LOV, light-oxygen-voltage; RASSLs, receptors activated solely by synthetic ligands; SH2/3, Src (Sarcoma) homology 2/3; PTB, phosphotyrosine binding; Grb2, growth factor receptor-bound protein 2; FADD, FAS-associated death domain; Ste5/11, stellate5/11; PBS2, phosphate buffered saline 2; PDZ, post synaptic density protein (PSD95), Drosophila disc large tumor suppressor (DlgA), Zonula occludens-1 protein (zo-1); N-WASP, neuronal Wiskott-Aldrich syndrome protein; miRNA, micro RNA; siRNA, small interfering RNA. ∗ Corresponding author. Tel.: +41 61 387 32 02; fax: +41 61 387 39 94. ∗∗ Corresponding author. Tel.: +41 61 387 32 15; fax: +41 61 387 39 94. E-mail addresses: [email protected] (M.A. Marchisio), [email protected] (F. Rudolf). 1357-2725/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2010.11.012

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

1. Introduction Synthetic biology is a rather new discipline—commonly referred to as “life engineering”—that aims to design and construct new, biological systems characterized by specific and fully predictable outputs. Synthetically reengineered cells might target several important tasks from disease treatment (Ro et al., 2006) to biofuel production (Savage et al., 2008) and hazardous waste recognition and removal (de las Heras et al., 2008). Initial attempts to build synthetic circuits were mainly proof of principle studies based on the mechanisms that regulate DNA transcription. The first striking results were obtained a decade ago: the “Repressilator” (Elowitz and Leibler, 2000) and the “Toggle Switch” (Gardner et al., 2000). The Repressilator is a ring oscillating system made of three genes each synthesizing a different repressor. They are arranged in such a way that the first gene inhibits the second, the second represses the third, and the third closes the cycle by acting on the first one. The output of the circuit, a reporter protein coupled with one of the three genes, shows oscillations that can persist for hours. The “Toggle Switch” is slightly simpler because it contains only two mutually repressing genes. As a main feature, the circuit shows two stable steady states (bistability) where just one of the two genes is expressed. By deactivating the repressor with the corresponding inducer molecules, the system can be switched from one to the other stable state. Nature equipped cells with sensing systems that allow for constant monitoring of the surrounding environment. These cellular entities can be either simple transcriptional networks or more elaborate signaling pathways. However, the task they perform is identical: recognizing external signals and transducing them to the proper cell compartment via a series of chemical reactions that demand high regulation. According to the main point of signal intervention, cellular sensing systems can be divided into three classes, namely: transcriptional, translational (or posttranscriptional), and post-translational. Transcriptional sensing systems are gene networks where signals control promoter activities by binding to and modifying the structures of transcriptional regulator proteins with direct access to the promoter sequences. The two classic synthetic biology examples mentioned above have in common the usage of two chemically controllable transcriptional regulators: TetR and LacI. Despite the possibility of engineering new, synthetic transcription factors (AjoFranklin et al., 2007), the majority of the gene circuits constructed so far employs only a handful of natural repressor and activator proteins since they can be easily manipulated and directly controlled with chemicals to obtain synchronization and titration of their respective activities (see Fig. 1 for a representation of the “Toggle Switch”). Translational sensing systems exert their control by modifying the availability, the localization, the structure or the stability of mRNA molecules. Here, signals regulate protein synthesis directly by binding the mRNA at riboswitches and ribozymes (Serganov and Patel, 2007). They are RNA structures made of a single (Winkler et al., 2002; Winkler and Breaker, 2005) or a tandem (Sudarsan et al., 2006) aptamer, where the signals bind, and an actuator, which undergoes either conformational changes (riboswitches) or splicing (ribozymes) as a consequence of the signal’s arrival. Their intrinsic on/off behavior makes them particularly suitable for a biological implementation of Boolean (digital) gates (Win et al., 2009). These devices work with binary (0/1) values and convert several inputs into a single output after performing a logic operation (e.g. AND: logic multiplication and OR: logic addition). In biology, 0 and 1 are translated into low and high concentration, respectively. RNA-based gates use chemicals such as thiamine and theofilline as inputs and return, as an output, the translation of a (fluorescent) protein. As an example,

311

a simple NOT gate built on a tandem riboswitch is shown in Fig. 2. Finally, post-translational sensing systems are made of an ensemble of proteins that, upon the arrival of a stimulus, perform a specific function such as activating or repressing the expression of some genes. Examples of this kind of systems are given by endogenous signaling pathways. Here, an extra-cellular signal is “captured” by a membrane protein (receptor) and then transmitted to a cellular component such as the nucleus or the flagellar motors—as observed in bacterial chemotaxis (Koshland, 1979)—through a cascade of reversible chemical interactions (cycles) catalyzed by enzymes. A synthetic implementation of a post-translational sensor was realized in yeast (Bashor et al., 2008). By engineering positive and negative feedback loops onto the scaffold protein of the yeast mating pathway, both qualitative and quantitative behavior of the system were changed in a predictable manner (see Fig. 3 for the scaffold configuration that mimics an ultrasensitive switch.). In the following, we illustrate the synthetic biosensors assembled thus far on DNA/RNA, and signaling pathways. The former were, initially, small circuits—controlled by external cues—that, successively, became more complex and were based mainly on different cell–cell communication processes. The latter are still in their early age but encouraging results clearly indicate their suitability for the construction of complex networks with novel functionalities. In conclusion, we will trace some possible, future directions of synthetic biology.

2. Transcriptional and translational biosensors 2.1. Exogenous control—prokaryotic cells Some transcription factors protein may be regarded as sensing devices since they are either activated or inactivated after binding an environmental signal (e.g. a chemical). This simple docking process finds diverse applications in nature such as the E. coli lac operon (Mueller-Hill, 1996) and the lysis/lysogeny control in phage lambda (Ptashne, 2004). Most of the synthetic gene circuits that were engineered in E. coli make use of only transcription factors together with their corresponding signals. For instance, the system TetR-aTC (Hillen and Berens, 1994) was exploited in multi-step cascade network (Hooshangi et al., 2005) to either switch on or off the production of a fluorescent protein depending on number of steps in the cascade. Furthermore, a bacterial tunable oscillator (Stricker et al., 2008) was recently implemented by coupling the LacI-IPTG system with an activator (AraC) induced by arabinose molecules. Here, the two chemicals were used to alter the period of the oscillations. As an important effect of transcription regulation, several synthetic and wild type promoters reproduce Boolean gates (Bintu et al., 2005; Silva-Rocha and de Lorenzo, 2008). Therefore, they can be exploited to build biological digital circuits that represent a means to develop new, efficient sensors for drug screening. In fact, thanks to the unequivocal input/output relation expressed into a truth table, synthetic gene digital circuits appear to be the most promising solution to properly integrate different input signals in a single readout. Remarkably, transcription factors can be regulated by inputs different from chemicals. Lou et al. (2010), for instance, implemented in E. coli a “push-on push-off” digital switch by controlling transcription with light. The output of this circuit is described as discrete (either 1 or 0) and can be switched between these two values by means of a luminous input signal. In another relevant application, E. coli cells were engineered to express the invasin gene (from Yersinia pseudotuberculosis) in response to a well-defined environ-

312

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

Fig. 1. Transcriptional gene circuits: the “Toggle Switch”. As constructed by Gardner et al. (2000), the core of the “Toggle Switch” contains two mutually repressing transcription units. Moreover, a third gene expressing GFP is under the control of one of the two repressors (not shown in figure). The “switch” functionality is conferred to the system by two chemicals (IPTG and aTC) that, when delivered to the system, bind and inactivate their corresponding repressor (LacI and TetR) such that the other one can be expressed. An example of hysteresis cycle—characteristic of bistable systems—is shown in the figure insert. Note that Standard Biological Parts are represented here with the icons from the MIT Registry whereas the two repressors are indicated with the transcription factor pool symbol.

mental signal (Anderson et al., 2006). Three possible variants were engineered by placing, each time, the inv gene under the control of a promoter activated by a different signal: arabinose, hypoxia (shortage of oxygen) and AI-1. Importantly, as invasin protein confers to bacteria the property of entering mammalian cells that express ␤1 integrins, E. coli reengineered to produce invasin in an inducible way can be exploited to sense and then invade tumor cells. In further work (Anderson et al., 2007), invasin protein synthesis was put under the control of one more input (Mg molecules) and the whole circuit became a bacterial AND gate. Hence, Boolean circuits

capable of responding to more inputs might achieve a very precise description of the tumor microenvironment and be used, in the future, for the treatment of certain kinds of cancer. The rational design of prokaryotic synthetic gene circuits, like the ones so far described, has been based on concepts and methods borrowed from electrical engineering such as parts standardization, abstraction hierarchy and part composability. DNA sequences have been classified—at the MIT Registry (http://partsregistry.org)—into Standard Biological Parts (such as promoters, ribosome binding sites (RBS), genes, small RNAs, and

Fig. 2. Translational Boolean gates: an RNA-based NOT gate. A possible genetic NOT gate implementation requires to place a tandem ribozyme on the mRNA in the proximity of the ribosome binding site. In the absence of the input signal (thiamine in this case), the ribozyme remains in an inactive state such that ribosomes can bind the mRNA and start translation. In contrast, after thiamine binding, the ribozyme gets activated and performs self-cleavage, which prevents ribosome binding and favors a rapid mRNA degradation.

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

313

Fig. 3. Post-translational networks: an ultrasensitive switch based on a synthetic scaffold. Ste5 scaffold (present in the yeast mating pathway) is modified with the insertion of a leucine zipper (Bashor et al., 2008) that serves as an anchor for a constitutively expressed repressor (red wedge in figure). Upon arrival of an external signal, the system stays initially in the “off” state since the repressor limits the signal propagation along the pathway. However, after a certain time enough activator protein (green wedge) is produced. Thanks to a higher affinity to the leucine zipper, the activator replaces the repressor on the scaffold and drastically enhances the signal propagation and, as a consequence, the fluorescent protein synthesis (“on” state). The switch is called ultrasensitive because of the sharp transition between the “on” and the “off” state.

terminators) according to their function in transcription and translation. Moreover, in order to join together different parts, Endy (2005) suggested to measure the input/output of each of them in terms of RNA polymerase per second (PoPS) and ribosome per second (RiPS), which represent shared biological currents. Therefore, they first permit to compose parts into more complex devices made of entire transcription units (e.g. inverters and logic gates) and then devices into circuits (abstraction hierarchy) in an electronic fashion. In addition to that, Marchisio and Stelling (2008) introduced the concept of pools of common signal carriers (RNA polymerases, ribosomes, transcription factors, small RNAs, and chemicals) to properly represent the interactions among devices and quantify the biological potential that determines the circuit performance. Within this framework, several tools for computational gene circuit design have been developed (Marchisio and Stelling, 2009) and a possible new protocol to describe parts and devices in terms of a set of specific parameters has also been proposed (Canton et al., 2008).

mammalian cells are represented by a transcriptional time-delay circuit (Weber et al., 2007b) based on the usage of biotin, and the replication of part of the Mycobacterium tuberculosis gene network that protects this pathogen from antibiotics (Weber et al., 2008). Remarkably, by screening a chemical library against the latter circuit, Weber and co-workers succeeded in finding a possible anti-tuberculosis compound. Finally, Boolean gates and more structured digital circuits have been implemented in eukaryotes as well. Here, in contrast to prokaryotes, various control mechanisms at the translational level have been exploited. In particular, RNA interference via siRNA (Rinaudo et al., 2007) and miRNA (Leisner et al., 2010) was adopted in mammalian cells to monitor mRNA abundance whereas antiswitches (Beisel and Smolke, 2009) and ribozymes (Win and Smolke, 2008) have been de novo engineered in yeast. Although they appear to be fairly efficient in reproducing Boolean logic, a proof of principle for coupling such biosensors to endogenous pathways is still missing.

2.3. Cell–cell signaling—prokaryotic cells 2.2. Exogenous control—eukaryotic cells Since the eukaryotic transcriptional machinery is much more complex than the bacterial one, the implementation of simple, transcriptional networks in organisms such as yeast and mammals is challenging. However, a eukaryotic version of some of the E. coli synthetic circuits has already been realized. For instance, Tigges et al. (2009) built a mammalian oscillator by using chemicalbased mechanisms. Here, two different kinds of small molecules are employed: tetracycline and pristinamicyn, which act on the activators tTA and PIT, respectively. Compared to a previously engineered bacterial oscillator (Atkinson et al., 2003), the circuit design still combines a positive and a negative feedback loop but, as a novelty, the latter targets tTA’s mRNA after stimulating the antisense transcription of the tTA gene. Other relevant results in

Cell–cell communication represents a valuable mechanism to engineer novel signaling constructs in bacteria. Conceptually, the easiest way to make cells “talk” is to exploit quorum sensing. This system, as observed in marine bacteria like Vibrio fischeri (Zavilgelsky and Manukhov, 2001), consists of sender cells producing AHL molecules that, after diffusing through the bacterial membranes, reach receiver cells where they bind the LuxR transcription activator and stimulate bioluminescence. The insertion of quorum sensing in E. coli was the key to engineer a “pulse generating network” (Basu et al., 2004) and a “band detector” (Basu et al., 2005). The latter, in particular, exploits a spatial gradient of AHL to show complex ring patterns (e.g. “French flag”) by expressing reporter proteins of different colors in concomitance of diverse AHL concentrations. Other significant results were achieved, in E. coli,

314

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

by You et al. (2004) and Danino et al. (2010). The former coupled quorum sensing with the synthesis of toxic proteins obtaining, in this way, a cell population control mechanism; the latter employed quorum sensing to construct a network of synchronized oscillators. An extension of these works on cell–cell communication is the generation of artificial ecosystems, where one cell produces molecules that the other needs. In the first application, quorum sensing was redesigned to establish a gene-metabolic network where bacterial cells exchange acetate molecules that are produced in proportion to cell growth (Bulter et al., 2004). Afterwards, Brenner et al. (2007) established a coordinated behavior between two bacterial populations (consortium) communicating via AHL. In this case, a particular gene is expressed only once both populations have reached a given density, which implies an AND gate with population concentrations as inputs. Furthermore, a synthetic predator–prey ecosystem was engineered in E. coli cells by using two different signals (Balagadde et al., 2008). Here, the predator cells emit a chemical that kills the prey by inducing the synthesis of a toxic protein; the prey, on the contrary, rescue predator cells by sending them the substance necessary to repress the production of the corresponding killer protein. Depending on the growth/death rate of the two groups, one of them will dominate or their concentrations will give rise to oscillations. 2.4. Cell–cell signaling—eukaryotic cells Engineering simple cell–cell communication in eukaryotes is a difficult task. Initial work on a cooperative system between two non-interacting yeast strains relied on a metabolic communication, where each strain produced the metabolites indispensable for the other’s survival (Shou et al., 2007). This approach is also feasible in higher eukaryotes. For instance, in mammalian cells Weber et al. (2009c) built what they called a “cell phone” controlled by a metabolite. The sender produces arginase that is not directly transmitted to the receiver but degrades the l-arginine molecules present in the medium that, in contrast, act on the receiver and stimulate the synthesis of a glycoprotein. Other examples of cell phones are provided by inter- and intra-kingdom signaling ecosystems employing volatile molecules to exchange information. For instance, a possible, new communication mechanism among mammalian cells, mediated by nitric oxide, was obtained in (Wang et al., 2008). Moreover, several kinds of coexistence patterns (e.g. symbiosis, parasitism, predator–prey oscillations) were engineered with mammalian cells either alone or mixed with other organisms such as bacteria and yeast (Weber et al., 2007a). Expanding the idea of the synthetic ecosystem, Weber et al. (2009a) constructed a biosensor able to monitor fruit quality. Here, the sender cell is replaced with fruit that, when aged, release ethylene. This chemical is oxidized to acetaldehyde to enter the receiver and trigger gene expression. In a more futuristic application, the sender was substituted with an artificial device, namely a microelectrolysis chamber, that directly produces acetaldehyde as a signal for the receiver. Consequently, output protein production inside the receiver can be tuned by varying frequency and amplitude of the electric current controlling the sender (Weber et al., 2009b). Overall, the synthetic gene circuits based on transcriptional and translational control represent a valid choice for construction of biosensors since their output behavior is, generally, quantitative predictable (Ellis et al., 2009; Guido et al., 2006). However, as a disadvantage, gene networks can show undesirable dynamics (e.g. oscillations) or may be coupled to the cellular physiology (e.g. quorum sensing). As an additional limitation to their employment as biosensors, gene circuits are characterized by a long response time that can go from several minutes to hours. Moreover, the usage of transcription factors different from the few ones currently adopted might be problematic since off-target effects cannot be easily excluded a priori. Finally, the

implementation of the mammalian synthetic oscillator has shown that stochastic noise may be high and crucial to determine the overall network performance mainly when the DNA and RNA species involved in the circuit are present in a low copy number. Hence, alternative architectures have to be considered for the design of reliable, complex biosensors. 2.5. Post-translational biosensors In nature, signal transduction is often implemented by post-translational sensing systems. Here, signals are sensed by transmembrane proteins (receptors) that represent the interface between the cell and the environment. A receptor, directly or indirectly, generally converts the external signal into a phospho-group that is propagated through the cytosol via a cascade of interacting proteins (adapters, scaffolds, and signaling proteins like kinases and phosphatases). Upon arrival in the nucleus, the signal elicits a transcriptional response. Receptor-dependent signal transduction is more prominent in eukaryotes than in prokaryotes and most of the work we are going to discuss refers to eukaryotic cells. However, as some receptor families are conserved among kingdoms, signaling pathway components can be reengineered in bacteria or plants and then brought into higher organisms. In addition, post-translational systems appear extremely versatile since their response time ranges from microseconds to hours (Legewie et al., 2008). Moreover, the actual signal transduction, along the cytoplasmic network of protein–protein interactions, is barely spoilt by stochastic noise (Yu et al., 2008) that becomes relevant again only once the signal has reached the nucleus (Colman-Lerner et al., 2005). For all these reasons, circuits based on post-translational signaling pathways appear to be the next frontier of synthetic biology. A complete transfer of a post-translational system from one species into another has already been accomplished (Chen and Weiss, 2005). In this work, the Arabidopsis thaliana cytokinin system was engineered in yeast in two variants. One was a classic cell–cell communication between a sender and a receiver cell where the production of the diffusible molecules (cytokinin) was regulated, in the sender, through an inducible promoter whilst the receiver had a constitutively expressed cytokinin receptor. Since the signal transducing moiety of the cytokinin receptor is homologous to the one of a yeast receptor, no additional changes were necessary to link the cytokinin receptor to an endogenous signaling pathway. In the other variant, an autokrine signaling system was implemented after integrating the sender into the receiver. Furthermore, cytokinin expression was regulated via a positive feedback loop that allowed for the overall pathway to be turned into a yeast quorum sensing system. Transmembrane receptors and most other proteins involved in signaling cascades show a modular structure. This feature is exploited to ‘locally’ design new properties and interaction modalities by, for instance, combining different domains, modifying single domains or grafting surface patches from a domain onto another one. In the following, we will discuss potential biosensor modules like receptors, adapters and signaling proteins (Pawson and Nash, 2003; Bhattacharyya et al., 2006; Takahashi and Pryciak, 2007) as well as their interaction domains and surfaces. For each of these categories, we will describe some significant work that appeared in the last few years. Finally, we will give a brief overview of the methods and the software currently available for the rational design of novel, synthetic signaling modules and pathways. 2.6. Signaling across the membrane—receptors Only some receptors are important for the scope of this review, namely: GPCR, RTK, and histidine kinase. Whilst the receptor

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

families are structurally diverse, they all show intracellular conformational changes upon binding of an extracellular ligand. Therefore, for synthetic biology applications one can either alter the sensing properties of a receptor or its coupling to the intracellular response, or both. First progress on the reengineering of receptor proteins came from bacterial works. For instance, E. coli histidine kinase was modified in the chemosensing domain, first, to bind trinitrotoluene (TNT) molecules (Looger et al., 2003) and, afterwards, to sense light instead of chemicals (Moeglich et al., 2009). The latter result was achieved by a replacement of the wild type histidine kinase chemosensor with a LOV (light-oxygen-voltage) domain, which allowed, as a final result, the control of gene expression via light signals. Furthermore, the histidine kinase receptor was also rewired to a new pathway after mutating only those few residues that completely determine the receptor-substrate specificity (Skerker et al., 2008). As the histidine kinase pathway is conserved among bacteria, yeast, and plants, the synthetic variants of histidine kinase obtained in E. coli should be transferable (Antunes et al., 2009; Ferreira and Kieber, 2005). For this reason, Bowen et al. (2008) suggested to insert the TNT sensing system described above into Arabidopsis thaliana where it might trigger a loss of chlorophyll, which is easily recognizable by the public. This could be a simple and cheap biosensor device for detecting toxic substances in the environment. In yeast, a variety of heterologous receptors have been engineered to react to the binding of different ligands. Apart from histidine kinase, which was employed in the cytokinin systems previously illustrated, GPCR—and its interactions with specific heterotrimeric G proteins—was also extensively studied. Initially, heterologous GPCRs were functionally expressed by coupling them to the signaling pathway normally downstream of the endogenous GPCRs. This was achieved by either coexpression of the corresponding heterologous alpha subunit or creation of a fusion protein between the foreign alpha subunit and the endogenous one (King et al., 1990; Pausch, 1997). This system is currently exploited in pharmaceutical companies to screen for molecules that modulate the activity of the heterologous GPCR. Starting from a hetereologous GPCR (Conklin et al., 2008) engineered a new class of receptors—named RASSLs (receptors activated solely by synthetic ligands)—that are insensitive to endogenous signals and can be faithfully controlled externally. GPCR belong to the family of seven transmembrane receptors and their prototypical member is the prokaryotic bacteriorhodpsins. These are light inducible pumps able to transport a wide variety of ions. Interestingly, the light sensing property of these proteins can be maintained if expressed in vertebrates. For instance, Kim et al. (2005) grafted the intracellular loops from a mammalian GPCR receptor onto bacteriorhodopsin and thereby created a synthetic, light-controllable receptor. Airan et al. (2009) extended this work by functionally expressing these heterologous receptors in mammalian cells and coupling each of them to different intracellular pathways. Since signaling pathways are also correlated with physical activities, the expression of a particular engineered bacteriorhodopsin in mouse-brain cells allowed for induction of a specific behavior (running) of the mice simply by shining green light on their head (optoXR). 2.7. Intracellular signal transduction—Notch receptors, adapters, signaling proteins and their docking interactions A different class of receptors is represented by Notch proteins. These particular membrane proteins upon activation undergo proteolytic cleavage and release their tail that encodes for a specific transcription factor. By replacing the wild type tails with different transcription factors sequences, the expression of a new set of

315

genes can be put under control of Notch ligand molecules (Struhl and Adachi, 1998). Barnea et al. (2008) obtained a similar result by engineering both GPCR and RTK. Here, a transcription factor domain was inserted into the cytoplasmic tail of both receptors whereas a protease was fused either to arrestin molecules, which are recruited in the cytoplasm by GPCR, or to SH2/PTB adapters that bind the RTK tail. Hence, upon arrival of an external cue, a transcription factor is freed by the protease action and moves to the nucleus where activates or represses the synthesis of a target gene. In a more classic way of cytoplasmic tail reenginering, Howard et al. (2003) rewired the tyrosine kinase pathway from cell proliferation to apoptosis by fusing the recognition domain of a Grb2 adapter—recruited by RTK—with the FADD effector domain that leads to cell death. Taken together, these works highlight the importance of the cytoplasmic moiety of a receptor since, by engineering this part, a signal can be rewired to a different output. Moreover, as a signal is transduced from the receptor tail to the inner cell by adapter molecules—such as scaffolds—or other intracellular components, modifications on the docking properties of such molecules can also drastically modify the wiring of a signaling pathway. Scaffolds, for instance, play an important role in the spatial and temporal regulation of signaling pathways. In yeast, the classic example of a signaling scaffold is the protein Ste5 and most of the work on scaffold-dependent signal rewiring and fine-tuning is carried out by using (and abusing) this protein. In a synthetic “diverter” scaffold (Park et al., 2003), Ste5 was engineered to redirect its natural stimulus to a neighboring pathway by exploiting a component shared by the two pathways. In an extension of this work the quantitative behavior of the signaling output was changed by integrating positive and negative feedback loops (Bashor et al., 2008). Here, the MAPK mating pathway controls the synthesis of modulator fusionproteins that can “come back” and bind to a synthetic Ste5 scaffold at a new recruitment site—a leucine zipper. By varying the scaffoldmodulator affinity and inducing competition between positive and negative feedback, this rewired pathway mimics circuits previously implemented in gene networks, like: pulse generators, ultrasensitive switches (as in Fig. 3) (response) accelerators, and delay devices. An alternative configuration of the positive feedback loop was engineered in (Ingolia and Murray, 2007). In this case, the pathway stimulates the production of the MAPKKK (Ste11) that, by binding to the scaffold, closes the loop and takes over its own synthesis. Therefore, this circuit behaves like a memory device that remembers an external pulse—of pheromones—even when the signal has ceased. A further option for reengineering signaling pathways demands to act on the interaction properties of MAP kinases, GTPases and other signaling proteins. These interaction surfaces or docking domains are indispensable for protein phosphorylation and have conserved functionality when moved to positions different from the original one (Grewal et al., 2006). Hence, by grafting them onto different proteins, they can be exploited to bring together partners that normally would not interact (Mody et al., 2009). According to this principle, Tatebayashi et al. (2003) achieved pathway hybridization simply by inserting a docking site into a “foreign” kinase. Signaling proteins have several other domains. Modifying them can lead to substantial changes (e.g. rewiring or a new output dynamics) in the transduction pathway (Mody et al., 2009). An example of an extensively engineered domain is the PDZ domain, which is common to prokaryotic and eukaryotic species. It was not only optimized to avoid cross-talk with similar domains of the same kind (Ernst et al., 2009), but also fused to a fibronectin domain—both bind the same peptide—to reach a higher specificity and affinity towards its ligand (Huang et al., 2008). Furthermore, Dueber et al. (2003) recombined the output domain of an N-WASP mammalian protein with an auto-inhibitory PDZ-peptide motif.

316

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

Table 1 URLs of computational tools of interest for synthetic biology. Pathway design and visualization CellDesigner (Funahashi et al., 2003) Virtual Cell (Loew and Schaff, 2001) CADLIVE (Kurata et al., 2003) ProMoT (Mirschel et al., 2009) Cytoscape (Shannon et al., 2003) SimBoolNet (Zheng et al., 2010) SNAVI (Ma’ayan et al., 2009) NAViGaTOR (Brown et al., 2009) BioNetCAD (Rialle et al., 2010) Formal languages BioNetGen (Blinov et al., 2004) Kappa language (Danos and Laneve, 2004) STOCHSIM (Le Novere and Shimizu, 2001) BIOCHAM (Fages et al., 2004) MDL (Ginkel et al., 2003) little b (Mallavarapu et al., 2009) ␲-Calculus (Regev et al., 2001) Bio-PEPA (Ciocchetta, 2008) Moleculizer (Lok and Brent, 2005) Protein design Rosetta (Simons et al., 1999) Raptor (Xu et al., 2003) HHpred (Soeding et al., 2005) Modeller (Eswar et al., 2007) EGAD (Pokala and Handel, 2005) PROTDES (Suarez et al., 2008) I-TASSER (Roy et al., 2010) Phyre (Kelley and Sternberg, 2009) PFP (Hawkins et al., 2009) HexServer (Macindoe et al., 2010) Autodock 4.2 (Morris et al., 2009) Biskit (Gruenberg et al., 2007)

http://www.celldesigner.org http://www.nrcam.uchc.edu/ http://www.cadlive.jp/ http://www.mpi-magdeburg.mpg.de/projects/promot/ http://www.cytoscape.org http://www.ncbi.nlm.nih.gov/CBBresearch/Przytycka/SimBoolNet/ http://code.google.com/p/snavi/ http://ophid.utoronto.ca/navigator/. http://www.sysdiag.cnrs.fr/index.php?page=bionetcad http://bionetgen.org/index.php/BioNetGen Distributions http://www.kappalanguage.org/ http://www.ebi.ac.uk/∼lenov/stochsim.html http://contraintes.inria.fr/BIOCHAM/ see ProMoT http://www.littleb.org/ http://www.wisdom.weizmann.ac.il/∼biospi/index main.html http://homepages.inf.ed.ac.uk/jeh/Bio-PEPA/biopepa.html http://www.molsci.org/∼lok/moleculizer/ http://www.rosettacommons.org/main.html http://www.bioinformaticssolutions.com/products/raptor http://toolkit.lmb.uni-muenchen.de/hhpred http://salilab.org/modeller/ http://egad.ucsd.edu/EGAD manual/index.html http://soft.synth-bio.org/protdes.html http://zhanglab.ccmb.med.umich.edu/I-TASSER/ http://www.sbg.bio.ic.ac.uk/∼phyre/ http://dragon.bio.purdue.edu/pfp/ http://www.loria.fr/∼ritchied/hex server/ http://autodock.scripps.edu/ http://biskit.sf.net

In this way, the N-WASP protein becomes a switch that can be activated by an external signal sequestering the PDZ domain from the peptide. The addition of another auto-inhibitory domain (SH3peptide) has “evolved” the synthetic switch into an AND gate. In a more recent work, Dueber et al. (2007) also showed that by placing multiple copies of the only SH3-peptide motif on the N-WASP output domain, the activation response gets much faster and the switch shows ultrasensitivity. Finally, Yeh et al. (2007) applied the same construction principle to a GEF protein and rendered it activatable. To this aim, they used a modified PDZ-peptide motif, where the peptide had been mutated in order to be phosphorylated. Hence, this GEF variant can be activated by a protein kinase. As we have seen, signaling networks rely on protein–protein interactions and pathway rewiring requires synthetic modification of protein domains. Classic techniques of protein reengineering include (but are not limited to) random or saturating mutagenesis, domain swapping between proteins or grafting of functional patches within a domain family (Gruenberg and Serrano, 2010; Van der Sloot et al., 2009). Whilst the former are examples of an unbiased approach, the latter demand some prior knowledge that can be accessed online at several, large databases like Swissprot (http://www.expasy.ch/sprot/), EMBL-EBI (http://www.ebi.ac.uk/uniprot/), NCBI (http://www.expasy.ch/sprot/), and PDB (http://www.pdb.org/). Importantly, over the last decade the number of computational tools that analyze and use this information for the rational design of protein binding properties has considerably increased. Rosetta (Simons et al., 1999), for instance, is a package that performs ab initio computation of protein structures and permits simulation of complex phenomena like protein folding, protein-ligand docking, protein–DNA binding and unbinding and more general protein–protein interactions. In addition, several tools—like the web-servers I-TASSER (Roy et al., 2010) and Phyre (Kelley and Sternberg, 2009)—can be used for 3D-structure prediction (see Table 1). Moreover, PFP (Hawkins et al., 2009) estimates possible

protein functions whereas docking sites can be localized on the protein surface with the help of the HexServer (Macindoe et al., 2010) and docking events may be reproduced via Autodock 4.2 (Morris et al., 2009). On the whole, thanks to these tools pathways components can be precisely modified and highly improved. 3. Computational design of biosensors The computer-aided design of new signaling pathways appears challenging because of their very entangled interaction structure where, for instance, cross-talk plays a relevant—and yet poorly understood—role. Moreover, whereas the qualitative temporal behavior of gene circuits can be depicted by solving systems of ordinary differential equations (Kaern et al., 2003), the mathematical models for signaling systems may need to take into account spatial coordinates and diffusion phenomena since cellular localization is a fundamental feature (Kholodenko, 2006). Furthermore, protein synthetic biology still lacks a common accepted definition of modules and module composability and no computational tool has been developed ad hoc thus far for the design of synthetic biosensors. To this aim, however, one can make use of the software used, mainly, in the field of systems biology to study and analyze very complex biological systems. Visual design tools offer the most intuitive approach. For instance, programs like Celldesigner (Funahashi et al., 2003) and ProMoT (Mirschel et al., 2009) provide the users with a graphic interface where components can be chosen from a library, displayed on the computer screen and linked to each other to build complex networks (“drag-and-drop” design) that represent biological species (or more structured modules) together with their interactions. Celldesigner is also equipped with a simulation environment so that the system dynamics may be checked at different steps of the design procedure and the system can be modified accordingly. However, a signaling pathway represents a very entangled system that might be too difficult or even impossible to design by

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

hand since, when the number of biochemical species increases, the amount of interactions grows exponentially. A possible solution is to run stochastic simulations with Moleculizer (Lok and Brent, 2005). In this way, a full description of the system is no longer necessary since new system components (and their corresponding interactions) are taken into account only when they are effectively produced. An alternative approach, which has already been adopted by several tools (see Table 1), lies on the so called “rule-based” modeling. Here, system components and reactions do not have to be exhaustively defined but are generated from a set of abstract expressions (rules). Reaction rules, for instance, look like biochemical reaction templates where the interactions are fixed but reactants and products are variables. Otherwise, a different, abstract representation of a biological system can be achieved with a formal language. As an example, ProMoT has an internal formal language—MDL: Model Definition Language (Ginkel et al., 2003)—whose syntax was derived from a programming language (Lisp). MDL permits specification of each system module—in terms of its structure and interactions with other modules—into separate files. Thus, modules of different complexity can be constructed in an iterative way following a precise hierarchy (e.g. from domains to proteins and from proteins to sub-systems characterized by specific protein–protein interactions; for an application see (Saez-Rodriguez et al., 2008)). Once designed, a signaling pathway can be visualized, modified and analyzed with Cytoscape (Shannon et al., 2003). This tool has the important property to be easily extendible via plug-ins written by the users, such as SimBoolNet (Zheng et al., 2010), a kit for dynamic simulations. Finally, web-services dedicated to the study of the dynamics of complex biological systems are of particular interest since they allow running simulations of intricate pathways on powerful server machines (for a recent review, see (Lee et al., 2009)). 4. Future directions Signaling circuits engineered so far—from transcription networks made of few genes to small ecosystems and rewired transduction pathways—have shown an enormous potential to help elucidate the fundamental principles that regulate both temporal and spatial cell dynamics. We argue that both transcriptional and translational networks will become a relevant component for the next generation of synthetic biosensors thanks to the broad range of different operations (e.g. feedback loops and Boolean logic) that synthetic gene circuits can perform. Ideally, biosensors will be able to integrate a large number of diverse inputs i.e. they will switch on and off the synthesis of one or more proteins depending on the particular signals that will have stimulated the receptor. However, the genetic networks, which have been implemented in the wet-lab, show a rather low complexity i.e. they are made of no more than 5–6 genes whose intersection points are rather scarce as well and, whereas the content of the MIT Registry is constantly increasing, a rigorous, theoretical characterization of the parts’ intrinsic properties (e.g. I/O transfer function and noise) is basically missing. To be of any use for biosensors, future gene circuits will have to show higher complexity due not only to a simple increase in the total number of genes and parts, but also to the usage of more transcription factors combined, possibly, with different translational control molecules (like miRNAs, siRNAs, and antisense RNAs) and mRNA configurations (e.g. riboswitches and ribozymes). One of the bottlenecks in constructing new, powerful synthetic circuits into a cell is the assembly of the encoding DNA sequences in the wet-lab. This task should be carried out in a more efficient way than the ones currently available. In fact, as long as one relies on restriction enzymes and DNA-ligase-based cloning the number

317

of sequences that can be combined in a single step is not larger than three (classic multiway cloning). De novo synthesis of the samples for a new generation of both synthetic gene circuits and biosensors is possible but, probably, too cost intensive for nearly everyone. Therefore, it is necessary to expand the DNA manipulation repertoire, perhaps by further exploiting yeast homologous recombination (Ma et al., 1987; Gibson, 2009), enhancing classic cloning via type II/S based ligations (Engler et al., 2009) or following the technology adopted at the Shapiro (Linshiz et al., 2008) and Venter labs (Gibson et al., 2009). Alternatively, classic cloning based on restriction enzymes might be combined with in vitro or in vivo recombination techniques. Another bottleneck for the de novo construction of synthetic pathways is the absence of a coherent framework for in silico analysis and design. A non-negligible gap on the computational/modeling side is a catalogue of proteins, protein-protein interactions, and protein domains (possibly characterized with high specificity). Ideally, this would be the protein-syntheticbiology counterpart of the MIT Registry. However, it should be noted that, to this aim, an encouraging project is currently carried forward with the development of the BioNetCAD software (Rialle et al., 2010). Such a theoretical infrastructure would demand proofs of concept i.e. the implementation of small, synthetic pathways designed in compliance with the “new registry” content. Overall, we believe that the engineering of synthetic, posttranslational, biosensors is one of the next big steps in synthetic biology. Its achievement will rely considerably on past experience since, for instance, the necessary wet-lab techniques are known: domain shuffling, replacement or grafting of new sensory features on receptors, rewiring of endogenous pathways, and fine-tuning of signaling dynamics. Furthermore, the lessons learnt from the design of simpler biosensors, such as the synthetic probes used in live-cell microscopy, should guide the de novo construction of post-translational signaling pathways. In fact, these probes already need to be intensively engineered (through domain shuffling, linker length optimization and subcellular localization) to report events faithfully. Hence, to successfully (re)construct entire transduction pathways, an innovative wet-lab strategy, which allows permuting the individual components easily, will be a first, crucial result. References Airan RD, Thompson KR, Fenno LE, Bernstein H, Deisseroth K. Temporally precise in vivo control of intracellular signalling. Nature 2009;458:1025–9. Ajo-Franklin CM, Drubin DA, Eskin JA, Gee EPS, Landgraf D, Phillips I, et al. Rational design of memory in eukaryotic cells. Genes Dev 2007;21:2271–6. Anderson JC, Clarke EJ, Arkin AP, Voigt CA. Environmentally controlled invasion of cancer cells by engineered bacteria. J Mol Biol 2006;355:619–27. Anderson JC, Voigt CA, Arkin AP. Environmental signal integration by a modular AND gate. Mol Syst Biol 2007:3. Antunes MS, Morey KJ, Tewari-Singh N, Bowen TA, Smith JJ, Webb CT, et al. Engineering key components in a synthetic eukaryotic signal transduction pathway. Mol Syst Biol 2009:5. Atkinson MR, Savageau MA, Myers JT, Ninfa AJ. Development of genetic circuitry exhibiting toggle switch or oscillatory behavior in Escherichia coli. Cell 2003;113:597–607. Balagadde FK, Song H, Ozaki J, Collins CH, Barnet M, Arnold FH, et al. A synthetic Escherichia coli predator–prey ecosystem. Mol Syst Biol 2008:4. Barnea G, Strapps W, Herrada G, Berman Y, Ong J, Kloss B, et al. The genetic design of signaling cascades to record receptor activation. Proc Natl Acad Sci U S A 2008;105:64–9. Bashor CJ, Helman NC, Yan S, Lim WA. Using engineered scaffold interactions to reshape MAP kinase pathway signaling dynamics. Science 2008;319:1539–43. Basu S, Gerchman Y, Collins CH, Arnold FH, Weiss R. A synthetic multicellular system for programmed pattern formation. Nature 2005;434:1130–4. Basu S, Mehreja R, Thiberge S, Chen M-T, Weiss R. Spatiotemporal control of gene expression with pulse-generating networks. Proc Natl Acad Sci U S A 2004;101:6355–60. Beisel CL, Smolke CD. Design principles for riboswitch function. PLoS Comput Biol 2009:5. Bhattacharyya RP, Remenyi A, Yeh BJ, Lim WA. Domains, motifs, and scaffolds: the role of modular interactions in the evolution and wiring of cell signaling circuits. Annu Rev Biochem 2006;75:655–80.

318

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319

Bintu L, Buchler NE, Garcia HG, Gerland U, Hwa T, Kondev J, et al. Transcriptional regulation by the numbers: applications. Curr Opin Genet Dev 2005;15:125–35. Blinov ML, Faeder JR, Goldstein B, Hlavacek WS. BioNetGen: software for rule-based modeling of signal transduction based on the interactions of molecular domains. Bioinformatics 2004;20:3289–91. Bowen TA, Zdunek JK, Medford JI. Cultivating plant synthetic biology from systems biology. New Phytol 2008;179:583–7. Brenner K, Karig DK, Weiss R, Arnold FH. Engineered bidirectional communication mediates a consensus in a microbial biofilm consortium. Proc Natl Acad Sci U S A 2007;104:17300–4. Brown KR, Otasek D, Ali M, McGuffin MJ, Xie W, Devani B, et al. NAViGaTOR: network analysis, visualization and graphing Toronto. Bioinformatics 2009;25:3327–9. Bulter T, Lee S-G, Wong WW, Fung E, Connor MR, Liao JC. Design of artificial cell-cell communication using gene and metabolic networks. Proc Natl Acad Sci U S A 2004;101:2299–304. Canton B, Labno A, Endy D. Refinement and standardization of synthetic biological parts and devices. Nat Biotechnol 2008;26:787–93. Chen M-T, Weiss R. Artificial cell–cell communication in yeast Saccharomyces cerevisiae using signaling elements from Arabidopsis thaliana. Nat Biotechnol 2005;23:1551–5. Ciocchetta FHJ. Bio-PEPA: an extension of the process algebra PEPA for biochemical networks. Electron Notes Theor Comput Sci 2008:194. Colman-Lerner A, Gordon A, Serra E, Chin T, Resnekov O, Endy D, et al. Regulated cell-to-cell variation in a cell-fate decision system. Nature 2005;437:699–706. Conklin BR, Hsiao EC, Claeysen S, Dumuis A, Srinivasan S, Forsayeth JR, et al. Engineering GPCR signaling pathways with RASSLs. Nat Methods 2008;5:673–8. Danino T, Mondragon-Palomino O, Tsimring L, Hasty J. A synchronized quorum of genetic clocks. Nature 2010;463:326–30. Danos V, Laneve C. Formal molecular biology. Theor Comp Sci 2004;325:69–110. de las Heras A, Carreno CA, de Lorenzo V. Stable implantation of orthogonal sensor circuits in Gram-negative bacteria for environmental release. Environ Microbiol 2008;10:3305–16. Dueber JE, Mirsky EA, Lim WA. Engineering synthetic signaling proteins with ultrasensitive input/output control. Nat Biotechnol 2007;25:660–2. Dueber JE, Yeh BJ, Chak K, Lim WA. Reprogramming control of an allosteric signaling switch through modular recombination. Science 2003;301:1904–8. Ellis T, Wang X, Collins JJ. Diversity-based, model-guided construction of synthetic gene networks with predicted functions. Nat Biotechnol 2009;27:465–71. Elowitz MB, Leibler S. A synthetic oscillatory network of transcriptional regulators. Nature 2000;403:335–8. Endy D. Foundations for engineering biology. Nature 2005;438:449–53. Engler C, Gruetzner R, Kandzia R, Marillonnet S. Golden gate shuffling: a one-pot DNA shuffling method based on type IIs restriction enzymes. PLoS One 2009;4:e5553. Ernst A, Sazinsky SL, Hui S, Currell B, Dharsee M, Seshagiri S, et al. Rapid evolution of functional complexity in a domain family. Sci Signal 2009:2. Eswar N, Webb B, Marti-Renom MA, Madhusudhan MS, Eramian D, Shen M-Y, et al. Curr protoc protein sci chapter 2. (2007). Comparative protein structure modeling using MODELLER. Curr Protoc Protein Sci 2007 [Chapter 2]. Fages F, Soliman S, Chabrier-Rivier N. Modelling and querying interaction networks in the biochemical abstract machine BIOCHAM. J Biol Phys Chem 2004:4. Ferreira FJ, Kieber JJ. Cytokinin signaling. Curr Opin Plant Biol 2005;8:518–25. Funahashi A, Tanimura N, Morohashi M, Kitano H. CellDesigner: a process diagram editor for gene-regulatory and biochemical networks. BIOSILICO 2003;1:159–62. Gardner TS, Cantor CR, Collins JJ. Construction of a genetic toggle switch in Escherichia coli. Nature 2000;403:339–42. Gibson DG. Synthesis of DNA fragments in yeast by one-step assembly of overlapping oligonucleotides. Nucleic Acids Res 2009;37:6984–90. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA, Smith HO. Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 2009;6:U341–3. Ginkel M, Kremling A, Nutsch T, Rehner R, Gilles ED. Modular modeling of cellular systems with ProMoT/Diva. Bioinformatics 2003;19:1169–76. Grewal S, Molina DM, Bardwell L. Mitogen-activated protein kinase (MAPK)-docking sites in MAPK kinases function as tethers that are crucial for MAPK regulation in vivo. Cell Signal 2006;18:123–34. Gruenberg R, Nilges M, Leckner J. Biskit—a software platform for structural bioinformatics. Bioinformatics 2007;23:769–70. Gruenberg R, Serrano L. Strategies for protein synthetic biology. Nucleic Acids Res 2010;38:2663–75. Guido NJ, Wang X, Adalsteinsson D, McMillen D, Hasty J, Cantor CR, et al. A bottom-up approach to gene regulation. Nature 2006;439:856–60. Hawkins T, Chitale M, Luban S, Kihara D. PFP: automated prediction of gene ontology functional annotations with confidence scores using protein sequence data. Proteins-Struct Funct Bioinformatics 2009;74:566–82. Hillen W, Berens C. Mechanisms underlying expression of Tn10 encoded tetracycline resistance. Annu Rev Microbiol 1994;48:345–69. Hooshangi S, Thiberge S, Weiss R. Ultrasensitivity and noise propagation in a synthetic transcriptional cascade. Proc Natl Acad Sci U S A 2005;102:3581–6. Howard PL, Chia MC, Rizzo SD, Liu F-F, Pawson T. Redirecting tyrosine kinase signaling to an apoptotic caspase pathway through chimeric adaptor proteins. Proc Natl Acad Sci U S A 2003;100:11267–72. Huang J, Koide A, Makabe K, Koide S. Design of protein function leaps by directed domain interface evolution. Proc Natl Acad Sci U S A 2008;105:6578–83. Ingolia NT, Murray AW. Positive-feedback loops as a flexible biological module. Curr Biol 2007;17:668–77.

Kaern M, Blake WJ, Collins JJ. The engineering of gene regulatory networks. Annu Rev Biomed Eng 2003;5:179–206. Kelley LA, Sternberg MJE. Protein structure prediction on the Web: a case study using the Phyre server. Nat Protoc 2009;4:363–71. Kholodenko BN. Cell-signalling dynamics in time and space. Nat Rev Mol Cell Biol 2006;7:165–76. Kim JM, Hwa J, Garriga P, Reeves PJ, RajBhandary UL, Khorana HG. Light-driven activation of beta 2-adrenergic receptor signaling by a chimeric rhodopsin containing the beta 2-adrenergic receptor cytoplasmic loops. Biochemistry 2005;44:2284–92. King K, Dohlman HG, Thorner J, Caron MG, Lefkowitz RJ. Control of yeast mating signal transduction by a mammalian beta 2-adrenergic receptor and Gs alpha subunit. Science 1990;250:121–3. Koshland DE. A model regulatory system: bacterial chemotaxis. Physiol Rev 1979:811–62. Kurata H, Matoba N, Shimizu N. CADLIVE for constructing a large-scale biochemical network based on a simulation-directed notation and its application to yeast cell cycle. Nucleic Acids Res 2003;31:4071–84. Le Novere N, Shimizu TS. STOCHSIM: modelling of stochastic biomolecular processes. Bioinformatics 2001;17:575–6. Lee D-Y, Saha R, Yusufi FNK, Park W, Karimi IA. Web-based applications for building, managing and analysing kinetic models of biological systems. Brief Bioinform 2009;10:65–74. Legewie S, Herzel H, Westerhoff HV, Bluthgen N. Recurrent design patterns in the feedback regulation of the mammalian signalling network. Mol Syst Biol 2008;4:190. Leisner M, Bleris L, Lohmueller J, Xie Z, Benenson Y. Rationally designed logic integration of regulatory signals in mammalian cells. Nat Nanotechnol 2010;5:666–70. Linshiz G, Ben Yehezkel T, Kaplan S, Gronau I, Ravid S, Adar R, et al. Recursive construction of perfect DNA molecules from imperfect oligonucleotides. Mol Syst Biol 2008;4. Loew LM, Schaff JC. The Virtual Cell: a software environment for computational cell biology. Trends Biotechnol 2001;19:401–6. Lok L, Brent R. Automatic generation of cellular reaction networks with Moleculizer 1.0. Nat Biotechnol 2005;23:131–6. Looger LL, Dwyer MA, Smith JJ, Hellinga HW. Computational design of receptor and sensor proteins with novel functions. Nature 2003;423:185–90. Lou CB, Liu XL, Ni M, Huang YQ, Huang QS, Huang LW, et al. Synthesizing a novel genetic sequential logic circuit: a push-on push-off switch. Mol Syst Biol 2010:6. Ma H, Kunes S, Schatz PJ, Botstein D. Plasmid construction by homologous recombination in yeast. Gene 1987;58:201–16. Ma’ayan A, Jenkins SL, Webb RL, Berger SI, Purushothaman SP, Abul-Husn NS, et al. SNAVI: desktop application for analysis and visualization of large-scale signaling networks. BMC Syst Biol 2009:3. Macindoe G, Mavridis L, Venkatraman V, Devignes MD, Ritchie DW. HexServer: an FFT-based protein docking server powered by graphics processors. Nucleic Acids Res 2010;38(Suppl.):W445–9. Mallavarapu A, Thomson M, Ullian B, Gunawardena J. Programming with models: modularity and abstraction provide powerful capabilities for systems biology. J R Soc Interface 2009;6:257–70. Marchisio MA, Stelling J. Computational design of synthetic gene circuits with composable parts. Bioinformatics 2008;24:1903–10. Marchisio MA, Stelling JR. Computational design tools for synthetic biology. Curr Opin Biotechnol 2009;20:479–85. Mirschel S, Steinmetz K, Rempel M, Ginkel M, Gilles ED. PROMOT: modular modeling for systems biology. Bioinformatics 2009;25:687–9. Mody A, Weiner J, Ramanathan S. Modularity of MAP kinases allows deformation of their signalling pathways. Nat Cell Biol 2009;11:484–91. Moeglich A, Ayers RA, Moffat K. Design and signaling mechanism of light-regulated histidine kinases. J Mol Biol 2009;385:1433–44. Morris GM, Huey R, Lindstrom W, Sanner MF, Belew RK, Goodsell DS, et al. AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility. J Comput Chem 2009;30:2785–91. Mueller-Hill B. The lac operon. Berlin: Editor ed; 1996. Park S-H, Zarrinpar A, Lim WA. Rewiring MAP kinase pathways using alternative scaffold assembly mechanisms. Science 2003;299:1061–4. Pausch MH. G-protein-coupled receptors in Saccharomyces cerevisiae: highthroughput screening assays for drug discovery. Trends Biotechnol 1997;15:487–94. Pawson T, Nash P. Assembly of cell regulatory systems through protein interaction domains. Science 2003;300:445–52. Pokala N, Handel TM. Energy functions for protein design: adjustment with protein–protein complex affinities, models for the unfolded state, and negative design of solubility and specificity. J Mol Biol 2005;347:203–27. Ptashne M. A genetic switch. Editor ed. John Inglis; 2004. Regev A, Silverman W, Shapiro E. Representation and simulation of biochemical processes using the pi-calculus process algebra. Pac Symp Biocomput 2001:459–70. Rialle S, Felicori L, Dias-Lopes C, Peres S, El Atia S, Thierry AR, et al. BioNetCAD: design, simulation and experimental validation of synthetic biochemical networks. Bioinformatics 2010;26:2298–304. Rinaudo K, Bleris L, Maddamsetti R, Subramanian S, Weiss R, Benenson Y. A universal RNAi-based logic evaluator that operates in mammalian cells. Nat Biotechnol 2007;25:795–801. Ro DK, Paradise EM, Ouellet M, Fisher KJ, Newman KL, Ndungu JM, et al. Production of the antimalarial drug precursor artemisinic acid in engineered yeast. Nature 2006;440:940–3.

M.A. Marchisio, F. Rudolf / The International Journal of Biochemistry & Cell Biology 43 (2011) 310–319 Roy A, Kucukural A, Zhang Y. I-TASSER: a unified platform for automated protein structure and function prediction. Nat Protoc 2010;5:725–38. Saez-Rodriguez J, Gayer S, Ginkel M, Gilles ED. Automatic decomposition of kinetic models of signaling networks minimizing the retroactivity among modules. Bioinformatics 2008;24:i213–9. Savage DF, Way J, Silver PA. Defossiling fuel: How synthetic biology can transform biofuel production. Acs Chem Biol 2008;3:13–6. Serganov A, Patel DJ. Ribozymes, riboswitches and beyond: regulation of gene expression without proteins. Nat Rev Genet 2007;8:776–90. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res 2003;13:2498–504. Shou W, Ram S, Vilar JMG. Synthetic cooperation in engineered yeast populations. Proc Natl Acad Sci U S A 2007;104:1877–82. Silva-Rocha R, de Lorenzo VC. Mining logic gates in prokaryotic transcriptional regulation networks. FEBS Lett 2008;582:1237–44. Simons KT, Bonneau R, Ruczinski I, Baker D. Ab initio protein structure prediction of CASP III targets using ROSETTA. Proteins-Struct Funct Genetics 1999: 171–6. Skerker JM, Perchuk BS, Siryaporn A, Lubin EA, Ashenberg O, Goulian M, et al. Rewiring the specificity of two-component signal transduction systems. Cell 2008;133:1043–54. Soeding J, Biegert A, Lupas AN. The HHpred interactive server for protein homology detection and structure prediction. Nucleic Acids Res 2005;33:W244–8. Stricker J, Cookson S, Bennett MR, Mather WH, Tsimring LS, Hasty J. A fast, robust and tunable synthetic gene oscillator. Nature 2008;456:516–9. Struhl G, Adachi A. Nuclear access and action of notch in vivo. Cell 1998;93: 649–60. Suarez M, Tortosa P, Jaramillo A. PROTDES: CHARMM toolbox for computational protein design. Syst Synth Biol 2008;2:105–13. Sudarsan N, Hammond MC, Block KF, Welz RD, Barrick JE, Roth A, et al. Tandem riboswitch architectures exhibit complex gene control functions. Science 2006;314:300–4. Takahashi S, Pryciak PM. Identification of novel membrane-binding domains in multiple yeast Cdc42 effectors. Mol Biol Cell 2007;18:4945–56. Tatebayashi K, Takekawa M, Saito H. A docking site determining specificity of Pbs2 MAPKK for Ssk2/Ssk22 MAPKKKs in the yeast HOG pathway. EMBO J 2003;22:3624–34. Tigges M, Marquez-Lago TT, Stelling J, Fussenegger M. A tunable synthetic mammalian oscillator. Nature 2009;457:309–12. Van der Sloot AM, Kiel C, Serrano L, Stricher F. Protein design in biological networks: from manipulating the input to modifying the output. Protein Eng Des Sel 2009;22:537–42.

319

Wang W-D, Chen Z-T, Kang B-G, Li R. Construction of an artificial intercellular communication network using the nitric oxide signaling elements in mammalian cells. Exp Cell Res 2008;314:699–706. Weber W, Baba MD-E, Fussenegger M. Synthetic ecosystems based on airborne inter- and intrakingdom communication. Proc Natl Acad Sci U S A 2007a;104:10435–40. Weber W, Luzi S, Karlsson M, Fussenegger M. A novel hybrid dual-channel catalytic-biological sensor system for assessment of fruit quality. J Biotechnol 2009a;139:314–7. Weber W, Luzi S, Karlsson M, Sanchez-Bustamante CD, Frey U, Hierlemann A, et al. A synthetic mammalian electro-genetic transcription circuit. Nucleic Acids Res 2009b:37. Weber W, Schoenmakers R, Keller B, Gitzinger M, Grau T, Baba MD-E, et al. A synthetic mammalian gene circuit reveals antituberculosis compounds. Proc Natl Acad Sci U S A 2008;105:9994–8. Weber W, Schuetz M, Denervaud N, Fussenegger M. A synthetic metabolite-based mammalian inter-cell signaling system. Mol Biosyst 2009c;5:757–63. Weber W, Stelling J, Rimann M, Keller B, Baba MD-E, Weber CC, et al. A synthetic time-delay circuit in mammalian cells and mice. Proc Natl Acad Sci U S A 2007b;104:2643–8. Win MN, Liang JC, Smolke CD. Frameworks for programming biological function through RNA parts and devices. Chem Biol 2009;16:298–310. Win MN, Smolke CD. Higher-order cellular information processing with synthetic RNA devices. Science 2008;322:456–60. Winkler W, Nahvi A, Breaker RR. Thiamine derivatives bind messenger RNAs directly to regulate bacterial gene expression. Nature 2002;419:952–6. Winkler WC, Breaker RR. Regulation of bacterial gene expression by riboswitches. Annu Rev Microbiol 2005;59:487–517. Xu J, Li M, Kim D, Xu Y. RAPTOR: optimal protein threading by linear programming. J Bioinform Comput Biol 2003;1:95–117. Yeh BJ, Rutigliano RJ, Deb A, Bar-Sagi D, Lim WA. Rewiring cellular morphology pathways with synthetic guanine nucleotide exchange factors. Nature 2007;447:596–600. You L, Cox RS, Weiss R, Arnold FH. Programmed population control by cell–cell communication and regulated killing. Nature 2004;428:868–71. Yu RC, Pesce CG, Colman-Lerner A, Lok L, Pincus D, Serra E, et al. Negative feedback that improves information transmission in yeast signalling. Nature 2008;456:755–61. Zavilgelsky GB, Manukhov IV. Quorum sensing, or how bacteria “Talk” to each other. Mol Biol 2001;35:224–32. Zheng J, Zhang D, Przytycki PF, Zielinski R, Capala J, Przytycka TM. SimBoolNet—a cytoscape plugin for dynamic simulation of signaling networks. Bioinformatics 2010;26:141–2.