Katsuyuki KUNIDA

The optimization of bioprocess inputs using mathematical models is widely practiced. However, the mismatch between model prediction and the actual process [called process-model mismatch (PMM)] is problematic; when a large PMM exists, the process inputs optimized using the mathematical model in advance are no longer optimal for the actual process. In this study, we propose a hybrid control system that combines model-based optimization (in silico feedforward controller) and feedback controllers using synthetic genetic circuits integrated into cells (in-cell feedback controller) - which we named the hybrid in silico/in-cell controller (HISICC) - as a solution to this PMM issue. As a proof of concept for HISICC, we constructed a mathematical model of an engineered Escherichia coli strain for the isopropanol production process that was previously developed. This strain contains an in-cell feedback controller, and its combination with an in silico controller can be regarded as an example of HISICC. We demonstrated the robustness of HISICC against PMM by comparing the strain with another strain with no in-cell feedback controller in simulations assuming PMM of various magnitudes.

Limitations in simultaneously observing the activity of multiple molecules in live cells prevent researchers from elucidating how these molecules coordinate the dynamic regulation of cellular functions. Here, we propose the motion-triggered average (MTA) algorithm to characterize pseudo-simultaneous dynamic changes in arbitrary cellular deformation and molecular activities. Using MTA, we successfully extract a pseudo-simultaneous time series from individually observed activities of three Rho GTPases: Cdc42, Rac1, and RhoA. To verify that this time series encoded information on cell-edge movement, we use a mathematical regression model to predict the edge velocity from the activities of the three molecules. The model accurately predicts the unknown edge velocity, providing numerical evidence that these Rho GTPases regulate edge movement. Data preprocessing using MTA combined with mathematical regression provides an effective strategy for reusing numerous individual observations of molecular activities.

Cell-to-cell variability in signal transduction in biological systems is often considered noise. However, intercellular variation (i.e., cell-to-cell variability) has the potential to enable individual cells to encode different information. Here, we show that intercellular variation increases information transmission of skeletal muscle. We analyze the responses of multiple cultured myotubes or isolated skeletal muscle fibers as a multiple-cell channel composed of single-cell channels. We find that the multiple-cell channel, which incorporates intercellular variation as information, not noise, transmitted more information in the presence of intercellular variation than in the absence according to the "response diversity effect," increasing in the gradualness of dose response by summing the cell-to-cell variable dose responses. We quantify the information transmission of human facial muscle contraction during intraoperative neurophysiological monitoring and find that information transmission of muscle contraction is comparable to that of a multiple-cell channel. Thus, our data indicate that intercellular variation can increase the information capacity of tissues.

Background: ATP is the major energy source for myotube contraction, and is quickly produced to compensate ATP consumption and to maintain sufficient ATP level. ATP is consumed mainly in cytoplasm and produced in mitochondria during myotube contraction. To understand the mechanism of ATP homeostasis during myotube contraction, it is essential to monitor mitochondrial ATP at single-cell level, and examine how ATP is produced and consumed in mitochondria.Methods: We established C2C12 cell line stably expressing fluorescent probe of mitochondrial ATP, and induced differentiation into myotubes. We gave electric pulse stimulation to the differentiated myotubes, and measured mitochondrial ATP. We constructed mathematical model of mitochondrial ATP at single-cell level, and analyzed kinetic parameters of ATP production and consumption.Results: We performed hierarchical clustering analysis of time course of mitochondrial ATP, which resulted in two clusters. Cluster 1 showed strong transient increase, whereas cluster 2 showed weak transient increase. Mathematical modeling at single-cell level revealed that the ATP production rate of cluster 1 was larger than that of cluster 2, and that both regulatory pathways of ATP production and consumption of cluster 1 were faster than those of cluster 2. Cluster 1 showed larger mitochondrial mass than cluster 2, suggesting that cluster 1 shows the similar property of slow muscle fibers, and cluster 2 shows the similar property of fast muscle fibers.Conclusions: Cluster 1 showed the stronger mitochondrial ATP increase by larger ATP production rate, but not smaller consumption. Cluster 1 might reflect the larger oxidative capacity of slow muscle fiber.

Automatic cell segmentation is a powerful method for quantifying signaling dynamics at single-cell resolution in live cell fluorescence imaging. Segmentation methods for mononuclear and round shape cells have been developed extensively. However, a segmentation method for elongated polynuclear cells, such as differentiated C2C12 myotubes, has yet to be developed. In addition, myotubes are surrounded by undifferentiated reserve cells, making it difficult to identify background regions and subsequent quantification. Here we developed an automatic quantitative segmentation method for myotubes using watershed segmentation of summed binary images and a two-component Gaussian mixture model. We used time-lapse fluorescence images of differentiated C2C12 cells stably expressing Eevee-S6K, a fluorescence resonance energy transfer (FRET) biosensor of S6 kinase (S6K). Summation of binary images enhanced the contrast between myotubes and reserve cells, permitting detection of a myotube and a myotube center. Using a myotube center instead of a nucleus, individual myotubes could be detected automatically by watershed segmentation. In addition, a background correction using the two-component Gaussian mixture model permitted automatic signal intensity quantification in individual myotubes. Thus, we provide an automatic quantitative segmentation method by combining automatic myotube detection and background correction. Furthermore, this method allowed us to quantify S6K activity in individual myotubes, demonstrating that some of the temporal properties of S6K activity such as peak time and half-life of adaptation show different dose-dependent changes of insulin between cell population and individuals.Key words: time lapse images, cell segmentation, fluorescence resonance energy transfer, C2C12, myotube.

This paper deals with an input/output signal recovery problem for nonlinear multiple-input single-output autoregressive exogenous (ARX) models with input nonlinearity, which are used in data-driven systems biology. A matrix rank minimization approach is applied, and a new signal recovery algorithm for nonlinear ARX models is provided. The proposed algorithm recovers output signals and nonlinear-transformed input signals on a linear subspace using some assumptions about nonlinear functions and does not require the exact knowledge of nonlinear functions. Numerical examples using experimental data of signal transduction of cellular systems show the efficiency of the proposed algorithm.

Cells decode information of signaling activation at a scale of tens of minutes by downstream gene expression with a scale of hours to days, leading to cell fate decisions such as cell differentiation. However, no system identification method with such different time scales exists. Here we used compressed sensing technology and developed a system identification method using data of different time scales by recovering signals of missing time points. We measured phosphorylation of ERK and CREB, immediate early gene expression products, and mRNAs of decoder genes for neurite elongation in PC12 cell differentiation and performed system identification, revealing the input +/- output relationships between signaling and gene expression with sensitivity such as graded or switch-like response and with time delay and gain, representing signal transfer efficiency. We predicted and validated the identified system using pharmacological perturbation. Thus, we provide a versatile method for system identification using data with different time scales.

Secretion of insulin transiently increases after eating, resulting in a high circulating concentration. Fasting limits insulin secretion, resulting in a low concentration of insulin in the circulation. We analyzed transcriptional responses to different temporal patterns and doses of insulin in the hepatoma FAO cells and identified 13 up-regulated and 16 down-regulated insulin-responsive genes (IRGs). The up-regulated IRGs responded more rapidly than did the down-regulated IRGs to transient stepwise or pulsatile increases in insulin concentration, whereas the down-regulated IRGs were repressed at lower concentrations of insulin than those required to stimulate the up-regulated IRGs. Mathematical modeling of the insulin response as two stages-(i) insulin signaling to transcription and (ii) transcription and mRNA stability-indicated that the first stage was the more rapid stage for the down-regulated IRGs, whereas the second stage of transcription was the more rapid stage for the up-regulated IRGs. A subset of the IRGs that were up-regulated or down-regulated in the FAO cells was similarly regulated in the livers of rats injected with a single dose of insulin. Thus, not only can cells respond to insulin but they can also interpret the intensity and pattern of signal to produce distinct transcriptional responses. These results provide insight that may be useful in treating obesity and type 2 diabetes associated with aberrant insulin production or tissue responsiveness.

Cellular signaling processes can exhibit pronounced cell-to-cell variability in genetically identical cells. This affects how individual cells respond differentially to the same environmental stimulus. However, the origins of cell-to-cell variability in cellular signaling systems remain poorly understood. Here, we measure the dynamics of phosphorylated MEK and ERK across cell populations and quantify the levels of population heterogeneity over time using highthroughput image cytometry. We use a statistical modeling framework to show that extrinsic noise, particularly that from upstream MEK, is the dominant factor causing cell-to-cell variability in ERK phosphorylation, rather than stochasticity in the phosphorylation/dephosphorylation of ERK. We furthermore show that without extrinsic noise in the core module, variable (including noisy) signals would be faithfully reproduced downstream, but the withinmodule extrinsic variability distorts these signals and leads to a drastic reduction in the mutual information between incoming signal and ERK activity.

The dynamic activity of the serine/threonine kinase Akt is crucial for the regulation of diverse cellular functions, but the precise spatiotemporal control of its activity remains a critical issue. Herein, we present a photo-activatable Akt (PA-Akt) system based on a light-inducible protein interaction module of Arabidopsis thaliana cryptochrome2 (CRY2) and CIB1. Akt fused to CRY2phr, which is a minimal light sensitive domain of CRY2 (CRY2-Akt), is reversibly activated by light illumination in several minutes within a physiological dynamic range and specifically regulates downstream molecules and inducible biological functions. We have generated a computational model of CRY2-Akt activation that allows us to use PA-Akt to control the activity quantitatively. The system provides evidence that the temporal patterns of Akt activity are crucial for generating one of the downstream functions of the Akt-FoxO pathway; the expression of a key gene involved in muscle atrophy (Atrogin-1). The use of an optical module with computational modeling represents a general framework for interrogating the temporal dynamics of biomolecules by predictive manipulation of optogenetic modules.

Cellular homeostasis is regulated by signals through multiple molecular networks that include protein phosphorylation and metabolites. However, where and when the signal flows through a network and regulates homeostasis has not been explored. We have developed a reconstruction method for the signal flow based on time-course phosphoproteome and metabolome data, using multiple databases, and have applied it to acute action of insulin, an important hormone for metabolic homeostasis. An insulin signal flows through a network, through signaling pathways that involve 13 protein kinases, 26 phosphorylated metabolic enzymes, and 35 allosteric effectors, resulting in quantitative changes in 44 metabolites. Analysis of the network reveals that insulin induces phosphorylation and activation of liver-type phosphofructokinase 1, thereby controlling a key reaction in glycolysis. We thus provide a versatile method of reconstruction of signal flow through the network using phosphoproteome and metabolome data.

Cell migration plays an important role in many physiological processes. Rho GTPases (Rac1, Cdc42, RhoA) and phosphatidylinositols have been extensively studied in directional cell migration. However, it remains unclear how Rho GTPases and phosphatidylinositols regulate random cell migration in space and time. We have attempted to address this issue using fluorescence resonance energy transfer (FRET) imaging and statistical signal processing. First, we acquired time-lapse images of random migration of HT-1080 fibrosarcoma cells expressing FRET biosensors of Rho GTPases and phosphatidyl inositols. We developed an image-processing algorithm to extract FRET values and velocities at the leading edge of migrating cells. Auto-and cross-correlation analysis suggested the involvement of feedback regulations among Rac1, phosphatidyl inositols and membrane protrusions. To verify the feedback regulations, we employed an acute inhibition of the signaling pathway with pharmaceutical inhibitors. The inhibition of actin polymerization decreased Rac1 activity, indicating the presence of positive feedback from actin polymerization to Rac1. Furthermore, treatment with PI3-kinase inhibitor induced an adaptation of Rac1 activity, i.e. a transient reduction of Rac1 activity followed by recovery to the basal level. In silico modeling that reproduced the adaptation predicted the existence of a negative feedback loop from Rac1 to actin polymerization. Finally, we identified MLCK as the probable controlling factor in the negative feedback. These findings quantitatively demonstrate positive and negative feedback loops that involve actin, Rac1 and MLCK, and account for the ordered patterns of membrane dynamics observed in randomly migrating cells.

The mitogen-activated protein (MAP) kinase pathway is comprised of a three-tiered kinase cascade. The distributive kinetic mechanism of two-site MAP kinase phosphorylation inherently generates a nonlinear switch-like response. However, a linear graded response of MAP kinase has also been observed in mammalian cells, and its molecular mechanism remains unclear. To dissect these input-output behaviors, we quantitatively measured the kinetic parameters involved in the MEK (MAPK/ERK kinase)-ERK MAP kinase signaling module in HeLa cells. Using a numerical analysis based on experimentally determined parameters, we predicted in silico and validated in vivo that ERK is processively phosphorylated in HeLa cells. Finally, we identified molecular crowding as a critical factor that converts distributive phosphorylation into processive phosphorylation. We proposed the term quasi-processive phosphorylation to describe this mode of ERK phosphorylation that is operated under the physiological condition of molecular crowding. The generality of this phenomenon may provide a new paradigm for a diverse set of biochemical reactions including multiple posttranslational modifications.

The coordination of the several pathways involved in cell motility is poorly understood. Here, we identify SH3BP1, belonging to the RhoGAP family, as a partner of the exocyst complex and establish a physical and functional link between two motility-driving pathways, the RaI/exocyst and Rac signaling pathways. We show that SH3BP1 localizes together with the exocyst to the leading edge of motile cells and that SH3BP1 regulates cell migration via its GAP activity upon Rac1. SH3BP1 loss of function induces abnormally high Rac1 activity at the front, as visualized by in vivo biosensors, and disorganized and instable protrusions, as revealed by cell morphodynamics analysis. Consistently, constitutively active Rac1 mimics the phenotype of SH3BP1 depletion: slow migration and aberrant cell morphodynamics. Our finding that SH3BP1 downregulates Rac1 at the motile-cell front indicates that Rac1 inactivation in this location, as well as its activation by GEF proteins, is a fundamental requirement for cell motility.