BioSwitch: a tool for the detection of bistability and multi-steady state behaviour in signalling and gene regulatory networks.

MOTIVATION: Multi-steady state behaviour, and in particular multi-stability, provides biological systems with the capacity to take reliable decisions (such as cell fate determination). A problem arising frequently in systems biology is to elucidate whether a signal transduction mechanism or a gene regulatory network has the capacity for multi-steady state behaviour, and consequently for a switch-like response to stimuli. Bifurcation diagrams are a powerful instrument in non-linear analysis to study the qualitative and quantitative behaviour of equilibria including bifurcation into different equilibrium branches and bistability. However, in the context of signalling pathways, the inherent large parametric uncertainty hampers the (direct) use of standard bifurcation tools. RESULTS: We present BioSwitch, a toolbox to detect multi-steady state behaviour in signalling pathways and gene regulatory networks. The tool combines results from chemical reaction network theory with global optimization to efficiently detect whether a signalling pathway has the capacity to undergo a saddle node bifurcation, and in case of multi-stationarity, provides the exact coordinates of the bifurcation where to start a numerical continuation analysis with standard bifurcation tools, leading to two different branches of equilibria. Bistability detection in the G1/S transition pathway of Saccharomyces cerevisiae is included as an illustrative example. AVAILABILITY AND IMPLEMENTATION: BioSwitch runs under the popular MATLAB computational environment, and is available at https://sites.google.com/view/bioswitch.

SBML Level 3: an extensible format for the exchange and reuse of biological models.

Systems biology has experienced dramatic growth in the number, size, and complexity of computational models. To reproduce simulation results and reuse models, researchers must exchange unambiguous model descriptions. We review the latest edition of the Systems Biology Markup Language (SBML), a format designed for this purpose. A community of modelers and software authors developed SBML Level 3 over the past decade. Its modular form consists of a core suited to representing reaction-based models and packages that extend the core with features suited to other model types including constraint-based models, reaction-diffusion models, logical network models, and rule-based models. The format leverages two decades of SBML and a rich software ecosystem that transformed how systems biologists build and interact with models. More recently, the rise of multiscale models of whole cells and organs, and new data sources such as single-cell measurements and live imaging, has precipitated new ways of integrating data with models. We provide our perspectives on the challenges presented by these developments and how SBML Level 3 provides the foundation needed to support this evolution.

Chemical Reaction Network Theory elucidates sources of multistability in interferon signaling.

Bistability has important implications in signaling pathways, since it indicates a potential cell decision between alternative outcomes. We present two approaches developed in the framework of the Chemical Reaction Network Theory for easy and efficient search of multiple steady state behavior in signaling networks (both with and without mass conservation), and apply them to search for sources of bistability at different levels of the interferon signaling pathway. Different type I interferon subtypes and/or doses are known to elicit differential bioactivities (ranging from antiviral, antiproliferative to immunomodulatory activities). How different signaling outcomes can be generated through the same receptor and activating the same JAK/STAT pathway is still an open question. Here, we detect bistability at the level of early STAT signaling, showing how two different cell outcomes are achieved under or above a threshold in ligand dose or ligand-receptor affinity. This finding could contribute to explain the differential signaling (antiviral vs apoptotic) depending on interferon dose and subtype (α vs ß) observed in type I interferons.

Efficient Reconstruction of Predictive Consensus Metabolic Network Models.

Understanding cellular function requires accurate, comprehensive representations of metabolism. Genome-scale, constraint-based metabolic models (GSMs) provide such representations, but their usability is often hampered by inconsistencies at various levels, in particular for concurrent models. COMMGEN, our tool for COnsensus Metabolic Model GENeration, automatically identifies inconsistencies between concurrent models and semi-automatically resolves them, thereby contributing to consolidate knowledge of metabolic function. Tests of COMMGEN for four organisms showed that automatically generated consensus models were predictive and that they substantially increased coherence of knowledge representation. COMMGEN ought to be particularly useful for complex scenarios in which manual curation does not scale, such as for eukaryotic organisms, microbial communities, and host-pathogen interactions.

Inducible, tightly regulated and growth condition-independent transcription factor in Saccharomyces cerevisiae.

The precise control of gene expression is essential in basic biological research as well as in biotechnological applications. Most regulated systems available in yeast enable only the overexpression of the target gene, excluding the possibility of intermediate or weak expression. Moreover, these systems are frequently toxic or depend on growth conditions. We constructed a heterologous transcription factor that overcomes these limitations. Our system is a fusion of the bacterial LexA DNA-binding protein, the human estrogen receptor (ER) and an activation domain (AD). The activity of this chimera, called LexA-ER-AD, is tightly regulated by the hormone ß-estradiol. The selection of the AD proved to be crucial to avoid toxic effects and to define the range of activity that can be precisely tuned with ß-estradiol. As our system is based on a heterologous DNA-binding domain, induction in different metabolic contexts is possible. Additionally, by controlling the number of LexA-binding sites in the target promoter, one can scale the expression levels up or down. Overall, our LexA-ER-AD system is a valuable tool to precisely control gene expression in different experimental contexts without toxic side effects.

Effect of Immunosuppression on T-Helper 2 and B-Cell Responses to Influenza Vaccination.

BACKGROUND: Influenza vaccine immunogenicity is suboptimal in immunocompromised patients. However, there are limited data on the interplay of T- and B- cell responses to vaccination with simultaneous immunosuppression. METHODS: We collected peripheral blood mononuclear cells from transplant recipients before and 1 month after seasonal influenza vaccination. Before and after vaccination, H1N1-specific T- and B-cell activation were quantified with flow cytometry. We also developed a mathematical model using T- and B-cell markers and mycophenolate mofetil (MMF) dosage. RESULTS: In the 47 patients analyzed, seroconversion to H1N1 antigen was demonstrated in 34%. H1N1-specific interleukin 4 (IL-4)-producing CD4(+) T-cell frequencies increased significantly after vaccination in 53% of patients. Prevaccine expression of H1N1-induced HLA-DR and CD86 on B cells was high in patients who seroconverted. Seroconversion against H1N1 was strongly associated with HLA-DR expression on B cells, which was dependent on the increase between prevaccine and postvaccine H1N1-specific IL-4(+)CD4(+) T cells (R(2) = 0.35). High doses of MMF (≥ 2 g/d) led to lower seroconversion rates, smaller increase in H1N1-specific IL-4(+)CD4(+) T cells, and reduced HLA-DR expression on B cells. The mathematical model incorporating a MMF-inhibited positive feedback loop between H1N1-specific IL-4(+)CD4(+) T cells and HLA-DR expression on B cells captured seroconversion with high specificity. CONCLUSIONS: Seroconversion is associated with influenza-specific T-helper 2 and B-cell activation and seems to be modulated by MMF.

Accurate cell segmentation in microscopy images using membrane patterns.

MOTIVATION: Identifying cells in an image (cell segmentation) is essential for quantitative single-cell biology via optical microscopy. Although a plethora of segmentation methods exists, accurate segmentation is challenging and usually requires problem-specific tailoring of algorithms. In addition, most current segmentation algorithms rely on a few basic approaches that use the gradient field of the image to detect cell boundaries. However, many microscopy protocols can generate images with characteristic intensity profiles at the cell membrane. This has not yet been algorithmically exploited to establish more general segmentation methods. RESULTS: We present an automatic cell segmentation method that decodes the information across the cell membrane and guarantees optimal detection of the cell boundaries on a per-cell basis. Graph cuts account for the information of the cell boundaries through directional cross-correlations, and they automatically incorporate spatial constraints. The method accurately segments images of various cell types grown in dense cultures that are acquired with different microscopy techniques. In quantitative benchmarks and comparisons with established methods on synthetic and real images, we demonstrate significantly improved segmentation performance despite cell-shape irregularity, cell-to-cell variability and image noise. As a proof of concept, we monitor the internalization of green fluorescent protein-tagged plasma membrane transporters in single yeast cells. AVAILABILITY AND IMPLEMENTATION: Matlab code and examples are available at http://www.csb.ethz.ch/tools/cellSegmPackage.zip.

Topological augmentation to infer hidden processes in biological systems.

MOTIVATION: A common problem in understanding a biochemical system is to infer its correct structure or topology. This topology consists of all relevant state variables-usually molecules and their interactions. Here we present a method called topological augmentation to infer this structure in a statistically rigorous and systematic way from prior knowledge and experimental data. RESULTS: Topological augmentation starts from a simple model that is unable to explain the experimental data and augments its topology by adding new terms that capture the experimental behavior. This process is guided by representing the uncertainty in the model topology through stochastic differential equations whose trajectories contain information about missing model parts. We first apply this semiautomatic procedure to a pharmacokinetic model. This example illustrates that a global sampling of the parameter space is critical for inferring a correct model structure. We also use our method to improve our understanding of glutamine transport in yeast. This analysis shows that transport dynamics is determined by glutamine permeases with two different kinds of kinetics. Topological augmentation can not only be applied to biochemical systems, but also to any system that can be described by ordinary differential equations. AVAILABILITY AND IMPLEMENTATION: Matlab code and examples are available at: http://www.csb.ethz.ch/tools/index

MetaNetX.org: a website and repository for accessing, analysing and manipulating metabolic networks.

MetaNetX.org is a website for accessing, analysing and manipulating genome-scale metabolic networks (GSMs) as well as biochemical pathways. It consistently integrates data from various public resources and makes the data accessible in a standardized format using a common namespace. Currently, it provides access to hundreds of GSMs and pathways that can be interactively compared (two or more), analysed (e.g. detection of dead-end metabolites and reactions, flux balance analysis or simulation of reaction and gene knockouts), manipulated and exported. Users can also upload their own metabolic models, choose to automatically map them into the common namespace and subsequently make use of the website's functionality.

Robustness and innovation in synthetic genotype networks.

Genotype networks are sets of genotypes connected by small mutational changes that share the same phenotype. They facilitate evolutionary innovation by enabling the exploration of different neighborhoods in genotype space. Genotype networks, first suggested by theoretical models, have been empirically confirmed for proteins and RNAs. Comparative studies also support their existence for gene regulatory networks (GRNs), but direct experimental evidence is lacking. Here, we report the construction of three interconnected genotype networks of synthetic GRNs producing three distinct phenotypes in Escherichia coli. Our synthetic GRNs contain three nodes regulating each other by CRISPR interference and governing the expression of fluorescent reporters. The genotype networks, composed of over twenty different synthetic GRNs, provide robustness in face of mutations while enabling transitions to innovative phenotypes. Through realistic mathematical modeling, we quantify robustness and evolvability for the complete genotype-phenotype map and link these features mechanistically to GRN motifs. Our work thereby exemplifies how GRN evolution along genotype networks might be driving evolutionary innovation.

A synthetic low-frequency mammalian oscillator.

Circadian clocks have long been known to be essential for the maintenance of physiological and behavioral processes in a variety of organisms ranging from plants to humans. Dysfunctions that subvert gene expression of oscillatory circadian-clock components may result in severe pathologies, including tumors and metabolic disorders. While the underlying molecular mechanisms and dynamics of complex gene behavior are not fully understood, synthetic approaches have provided substantial insight into the operation of complex control circuits, including that of oscillatory networks. Using iterative cycles of mathematical model-guided design and experimental analyses, we have developed a novel low-frequency mammalian oscillator. It incorporates intronically encoded siRNA-based silencing of the tetracycline-dependent transactivator to enable the autonomous and robust expression of a fluorescent transgene with periods of 26 h, a circadian clock-like oscillatory behavior. Using fluorescence-based time-lapse microscopy of engineered CHO-K1 cells, we profiled expression dynamics of a destabilized yellow fluorescent protein variant in single cells and real time. The novel oscillator design may enable further insights into the system dynamics of natural periodic processes as well as into siRNA-mediated transcription silencing. It may foster advances in design, analysis and application of complex synthetic systems in future gene therapy initiatives.

Multistable and dynamic CRISPRi-based synthetic circuits.

Gene expression control based on CRISPRi (clustered regularly interspaced short palindromic repeats interference) has emerged as a powerful tool for creating synthetic gene circuits, both in prokaryotes and in eukaryotes; yet, its lack of cooperativity has been pointed out as a potential obstacle for dynamic or multistable synthetic circuit construction. Here we use CRISPRi to build a synthetic oscillator ("CRISPRlator"), bistable network (toggle switch) and stripe pattern-forming incoherent feed-forward loop (IFFL). Our circuit designs, conceived to feature high predictability and orthogonality, as well as low metabolic burden and context-dependency, allow us to achieve robust circuit behaviors in Escherichia coli populations. Mathematical modeling suggests that unspecific binding in CRISPRi is essential to establish multistability. Our work demonstrates the wide applicability of CRISPRi in synthetic circuits and paves the way for future efforts towards engineering more complex synthetic networks, boosted by the advantages of CRISPR technology.

A Simple and Flexible Computational Framework for Inferring Sources of Heterogeneity from Single-Cell Dynamics.

Single-cell time-lapse data provide the means for disentangling sources of cell-to-cell and intra-cellular variability, a key step for understanding heterogeneity in cell populations. However, single-cell analysis with dynamic models is a challenging open problem: current inference methods address only single-gene expression or neglect parameter correlations. We report on a simple, flexible, and scalable method for estimating cell-specific and population-average parameters of non-linear mixed-effects models of cellular networks, demonstrating its accuracy with a published model and dataset. We also propose sensitivity analysis for identifying which biological sub-processes quantitatively and dynamically contribute to cell-to-cell variability. Our application to endocytosis in yeast demonstrates that dynamic models of realistic size can be developed for the analysis of single-cell data and that shifting the focus from single reactions or parameters to nuanced and time-dependent contributions of sub-processes helps biological interpretation. Generality and simplicity of the approach will facilitate customized extensions for analyzing single-cell dynamics.

Predicting ligand-dependent tumors from multi-dimensional signaling features.

Targeted therapies have shown significant patient benefit in about 5-10% of solid tumors that are addicted to a single oncogene. Here, we explore the idea of ligand addiction as a driver of tumor growth. High ligand levels in tumors have been shown to be associated with impaired patient survival, but targeted therapies have not yet shown great benefit in unselected patient populations. Using an approach of applying Bagged Decision Trees (BDT) to high-dimensional signaling features derived from a computational model, we can predict ligand dependent proliferation across a set of 58 cell lines. This mechanistic, multi-pathway model that features receptor heterodimerization, was trained on seven cancer cell lines and can predict signaling across two independent cell lines by adjusting only the receptor expression levels for each cell line. Interestingly, for patient samples the predicted tumor growth response correlates with high growth factor expression in the tumor microenvironment, which argues for a co-evolution of both factors in vivo.

Comparison of network-based pathway analysis methods.

Network-based definitions of biochemical pathways have emerged in recent years. These pathway definitions insist on the balanced use of a whole network of biochemical reactions. Two such related definitions, elementary modes and extreme pathways, have generated novel hypotheses regarding biochemical network function. The relationship between these two approaches can be illustrated by comparing and contrasting the elementary modes and extreme pathways of previously published metabolic reconstructions of the human red blood cell (RBC) and the human pathogen Helicobacter pylori. Descriptions of network properties generated by using these two approaches in the analysis of realistic metabolic networks need careful interpretation.

A method for inverse bifurcation of biochemical switches: inferring parameters from dose response curves.

BACKGROUND: Within cells, stimuli are transduced into cell responses by complex networks of biochemical reactions. In many cell decision processes the underlying networks behave as bistable switches, converting graded stimuli or inputs into all or none cell responses. Observing how systems respond to different perturbations, insight can be gained into the underlying molecular mechanisms by developing mathematical models. Emergent properties of systems, like bistability, can be exploited to this purpose. One of the main challenges in modeling intracellular processes, from signaling pathways to gene regulatory networks, is to deal with high structural and parametric uncertainty, due to the complexity of the systems and the difficulty to obtain experimental measurements. Formal methods that exploit structural properties of networks for parameter estimation can help to overcome these problems. RESULTS: We here propose a novel method to infer the kinetic parameters of bistable biochemical network models. Bistable systems typically show hysteretic dose response curves, in which the so called bifurcation points can be located experimentally. We exploit the fact that, at the bifurcation points, a condition for multistationarity derived in the context of the Chemical Reaction Network Theory must be fulfilled. Chemical Reaction Network Theory has attracted attention from the (systems) biology community since it connects the structure of biochemical reaction networks to qualitative properties of the corresponding model of ordinary differential equations. The inverse bifurcation method developed here allows determining the parameters that produce the expected behavior of the dose response curves and, in particular, the observed location of the bifurcation points given by experimental data. CONCLUSIONS: Our inverse bifurcation method exploits inherent structural properties of bistable switches in order to estimate kinetic parameters of bistable biochemical networks, opening a promising route for developments in Chemical Reaction Network Theory towards kinetic model identification.

Automatic generation of predictive dynamic models reveals nuclear phosphorylation as the key Msn2 control mechanism.

Predictive dynamical models are critical for the analysis of complex biological systems. However, methods to systematically develop and discriminate among systems biology models are still lacking. We describe a computational method that incorporates all hypothetical mechanisms about the architecture of a biological system into a single model and automatically generates a set of simpler models compatible with observational data. As a proof of principle, we analyzed the dynamic control of the transcription factor Msn2 in Saccharomyces cerevisiae, specifically the short-term mechanisms mediating the cells' recovery after release from starvation stress. Our method determined that 12 of 192 possible models were compatible with available Msn2 localization data. Iterations between model predictions and rationally designed phosphoproteomics and imaging experiments identified a single-circuit topology with a relative probability of 99% among the 192 models. Model analysis revealed that the coupling of dynamic phenomena in Msn2 phosphorylation and transport could lead to efficient stress response signaling by establishing a rate-of-change sensor. Similar principles could apply to mammalian stress response pathways. Systematic construction of dynamic models may yield detailed insight into nonobvious molecular mechanisms.

Efficient characterization of high-dimensional parameter spaces for systems biology.

BACKGROUND: A biological system's robustness to mutations and its evolution are influenced by the structure of its viable space, the region of its space of biochemical parameters where it can exert its function. In systems with a large number of biochemical parameters, viable regions with potentially complex geometries fill a tiny fraction of the whole parameter space. This hampers explorations of the viable space based on "brute force" or Gaussian sampling. RESULTS: We here propose a novel algorithm to characterize viable spaces efficiently. The algorithm combines global and local explorations of a parameter space. The global exploration involves an out-of-equilibrium adaptive Metropolis Monte Carlo method aimed at identifying poorly connected viable regions. The local exploration then samples these regions in detail by a method we call multiple ellipsoid-based sampling. Our algorithm explores efficiently nonconvex and poorly connected viable regions of different test-problems. Most importantly, its computational effort scales linearly with the number of dimensions, in contrast to "brute force" sampling that shows an exponential dependence on the number of dimensions. We also apply this algorithm to a simplified model of a biochemical oscillator with positive and negative feedback loops. A detailed characterization of the model's viable space captures well known structural properties of circadian oscillators. Concretely, we find that model topologies with an essential negative feedback loop and a nonessential positive feedback loop provide the most robust fixed period oscillations. Moreover, the connectedness of the model's viable space suggests that biochemical oscillators with varying topologies can evolve from one another. CONCLUSIONS: Our algorithm permits an efficient analysis of high-dimensional, nonconvex, and poorly connected viable spaces characteristic of complex biological circuitry. It allows a systematic use of robustness as a tool for model discrimination.