Mariko Hatakeyama

Inference of S-system models of genetic networks using a cooperative coevolutionary algorithm

MOTIVATION To resolve the high-dimensionality of the genetic network inference problem in the S-system model, a problem decomposition strategy has been proposed. While this strategy certainly shows promise, it cannot provide a model readily applicable to the computational simulation of the genetic network when the given time-series data contain measurement noise. This is a significant limitation of the problem decomposition, given that our analysis and understanding of the genetic network depend on the computational simulation. RESULTS We propose a new method for inferring S-system models of large-scale genetic networks. The proposed method is based on the problem decomposition strategy and a cooperative coevolutionary algorithm. As the subproblems divided by the problem decomposition strategy are solved simultaneously using the cooperative coevolutionary algorithm, the proposed method can be used to infer any S-system model ready for computational simulation. To verify the effectiveness of the proposed method, we apply it to two artificial genetic network inference problems. Finally, the proposed method is used to analyze the actual DNA microarray data.

Ligand‐dependent responses of the ErbB signaling network: experimental and modeling analyses

Deregulation of ErbB signaling plays a key role in the progression of multiple human cancers. To help understand ErbB signaling quantitatively, in this work we combine traditional experiments with computational modeling, building a model that describes how stimulation of all four ErbB receptors with epidermal growth factor (EGF) and heregulin (HRG) leads to activation of two critical downstream proteins, extracellular‐signal‐regulated kinase (ERK) and Akt. Model analysis and experimental validation show that (i) ErbB2 overexpression, which occurs in approximately 25% of all breast cancers, transforms transient EGF‐induced signaling into sustained signaling, (ii) HRG‐induced ERK activity is much more robust to the ERK cascade inhibitor U0126 than EGF‐induced ERK activity, and (iii) phosphoinositol‐3 kinase is a major regulator of post‐peak but not pre‐peak EGF‐induced ERK activity. Sensitivity analysis leads to the hypothesis that ERK activation is robust to parameter perturbation at high ligand doses, while Akt activation is not.

Function approximation approach to the inference of reduced NGnet models of genetic networks

BackgroundThe inference of a genetic network is a problem in which mutual interactions among genes are deduced using time-series of gene expression patterns. While a number of models have been proposed to describe genetic regulatory networks, this study focuses on a set of differential equations since it has the ability to model dynamic behavior of gene expression. When we use a set of differential equations to describe genetic networks, the inference problem can be defined as a function approximation problem. On the basis of this problem definition, we propose in this study a new method to infer reduced NGnet models of genetic networks.ResultsThrough numerical experiments on artificial genetic network inference problems, we demonstrated that our method has the ability to infer genetic networks correctly and it was faster than the other inference methods. We then applied the proposed method to actual expression data of the bacterial SOS DNA repair system, and succeeded in finding several reasonable regulations. When our method inferred the genetic network from the actual data, it required about 4.7 min on a single-CPU personal computer.ConclusionThe proposed method has an ability to obtain reasonable networks with a short computational time. As a high performance computer is not always available at every laboratory, the short computational time of our method is a preferable feature. There does not seem to be a perfect model for the inference of genetic networks yet. Therefore, in order to extract reliable information from the observed gene expression data, we should infer genetic networks using multiple inference methods based on different models. Our approach could be used as one of the promising inference methods.

Novel Mechanism of Interaction of p85 Subunit of Phosphatidylinositol 3-Kinase and ErbB3 Receptor-derived Phosphotyrosyl Peptides

Ligand-activated and tyrosine-phosphorylated ErbB3 receptor binds to the SH2 domain of the p85 subunit of phosphatidylinositol 3-kinase and initiates intracellular signaling. Here, we studied the interactions between the N- (N-SH2) and C- (C-SH2) terminal SH2 domains of the p85 subunit of the phosphatidylinositol 3-kinase and eight ErbB3 receptor-derived phosphotyrosyl peptides (P-peptides) by using molecular dynamics, free energy, and surface plasmon resonance (SPR) analyses. In SPR analysis, these P-peptides showed no binding to the C-SH2 domain, but P-peptides containing a phospho-YXXM or a non-phospho-YXXM motif did bind to the N-SH2 domain. The N-SH2 domain has two phosphotyrosine binding sites in its N- (N1) and C- (N2) terminal regions. Interestingly, we found that P-peptides of pY1180 and pY1241 favored to bind to the N2 site, although all other P-peptides showed favorable binding to the N1 site. Remarkably, two phosphotyrosines, pY1178 and pY1243, which are just 63 amino acids apart from the pY1241 and pY1180, respectively, showed favorable binding to the N1 site. These findings indicate a possibility that the pair of phosphotyrosines, pY1178-pY1241 or pY1243-pY1180, will fold into an appropriate configuration for binding to the N1 and N2 sites simultaneously. Our model structures of the cytoplasmic C-terminal domain of ErbB3 receptor also strongly supported the speculation. The calculated binding free energies between the N-SH2 domain and P-peptides showed excellent qualitative agreement with SPR data with a correlation coefficient of 0.91. The total electrostatic solvation energy between the N-SH2 domain and P-peptide was the dominant factor for its binding affinity.

Mutation of epidermal growth factor receptor is associated with MIG6 expression

Controlled activation of epidermal growth factor receptor (EGFR) is systematically guaranteed at the molecular level; however, aberrant activation of EGFR is frequently found in cancer. Transcription induced by EGFR activation often involves the coordinated expression of genes that positively and negatively regulate the original signaling pathway; therefore, alterations in EGFR kinase activity may reflect changes in gene expression associated with the pathway. In the present study, we investigated transcriptional changes after EGF stimulation with or without the EGFR kinase inhibitor Iressa in H1299 human non‐small‐cell lung cancer cells [parental H1299, H1299 cells that overexpress wild‐type EGFR (EGFR‐WT) and mutant H1299 cells that overexpress EGFR where Leu858 is substituted with Arg (L858R)]. The results obtained clearly demonstrate differences in transcriptional activity in the absence or presence of EGFR kinase activity, with genes sharing the same molecular functions showing distinct expression dynamics. The results show the particular enrichment of EGFR/ErbB signaling‐related genes in a differentially expressed gene set, and significant protein expression of MIG6/RALT(ERRFI1), an EGFR negative regulator, was confirmed in L858R. High MIG6 protein expression was correlated with basal EGFR phosphorylation and inversely correlated with EGF‐induced extracellular signal‐regulated protein kinase phosphorylation levels. Investigation of the NCI‐60 cell lines showed that ERRFI1 expression was correlated with EGFR expression, regardless of tissue type. These results suggest that cells accumulate MIG6 as an inherent negative regulator to suppress excess EGFR activity when basal EGFR kinase activity is considerably high. Taking all the above together, an EGFR mutation can cause transcriptional changes to accommodate the activation potency of the original signaling pathway at the cellular level.

Structural and Functional Basis of a Role for CRKL in a Fibroblast Growth Factor 8-Induced Feed-Forward Loop†

ABSTRACT The adapter protein CRKL is required for the normal development of multiple tissues that rely on fibroblast growth factor 8 (FGF8). The precise role of CRKL in receptor signaling has been unclear, however. To address this issue, we first modeled the three-dimensional structure of CRKL by molecular dynamics. By taking advantage of structural simulations, we performed in silico analysis of the interactions of the autophosphorylation sites of FGR receptor 1 (FGFR1) with the SH2 domain of CRKL or a highly related protein, CRK. As predicted by simulations, we confirm the specific physical interaction of phosphorylated Y463 (pY463) in FGFR1 with the CRKL SH2 domain at an affinity ∼30-fold stronger than that of CRK. We also provide evidence that interactions outside of the core YXXP motif have a significant impact on physical association, which is consistent with predictions from molecular-dynamics simulations. Furthermore, we identify CRKL as an essential component of an FGF8-induced feed-forward loop permissive for efficient activation of the mitogen-activated protein kinase Erk1/2, as well as FGF8-induced anchorage-independent cell growth, using Crkl-deficient cells or a pY463 synthetic peptide. Although many cells generally require cell-matrix adhesion, our results demonstrate that CRKL permits cells to bypass the strict need for adhesion in response to FGF8 through direct interaction with receptor.

Molecular dynamics simulations reveal that Tyr-317 phosphorylation reduces Shc binding affinity for phosphotyrosyl residues of epidermal growth factor receptor.

The Src homology 2 (SH2) and collagen domain protein Shc plays a pivotal role in signaling via tyrosine kinase receptors, including epidermal growth factor receptor (EGFR). Shc binding to phospho-tyrosine residues on activated receptors is mediated by the SH2 and phospho-tyrosine binding (PTB) domains. Subsequent phosphorylation on Tyr-317 within the Shc linker region induces Shc interactions with Grb2-Son of Sevenless that initiate Ras-mitogen-activated protein kinase signaling. We use molecular dynamics simulations of full-length Shc to examine how Tyr-317 phosphorylation controls Shc conformation and interactions with EGFR. Our simulations reveal that Shc tyrosine phosphorylation results in a significant rearrangement of the relative position of its domains, suggesting a key conformational change. Importantly, computational estimations of binding affinities show that EGFR-derived phosphotyrosyl peptides bind with significantly more strength to unphosphorylated than to phosphorylated Shc. Our results unveil what we believe is a novel structural phenomenon, i.e., tyrosine phosphorylation of Shc within its linker region regulates the binding affinity of SH2 and PTB domains for phosphorylated Shc partners, with important implications for signaling dynamics.

Topological analysis of MAPK cascade for kinetic ErbB signaling.

Ligand-induced homo- and hetero-dimer formation of ErbB receptors results in different biological outcomes irrespective of recruitment and activation of similar effector proteins. Earlier experimental research indicated that cells expressing both EGFR (epidermal growth factor receptor) and the ErbB4 receptor (E1/4 cells) induced E1/4 cell-specific B-Raf activation and higher extracellular signal-regulated kinase (ERK) activation, followed by cellular transformation, than cells solely expressing EGFR (E1 cells) in Chinese hamster ovary (CHO) cells. Since our experimental data revealed the presence of positive feedback by ERK on upstream pathways, it was estimated that the cross-talk/feedback pathway structure of the Raf-MEK-ERK cascade might affect ERK activation dynamics in our cell system. To uncover the regulatory mechanism concerning the ERK dynamics, we used topological models and performed parameter estimation for all candidate structures that possessed ERK-mediated positive feedback regulation of Raf. The structure that reliably reproduced a series of experimental data regarding signal amplitude and duration of the signaling molecules was selected as a solution. We found that the pathway structure is characterized by ERK-mediated positive feedback regulation of B-Raf and B-Raf-mediated negative regulation of Raf-1. Steady-state analysis of the estimated structure indicated that the amplitude of Ras activity might critically affect ERK activity through ERK-B-Raf positive feedback coordination with sustained B-Raf activation in E1/4 cells. However, Rap1 that positively regulates B-Raf activity might be less effective concerning ERK and B-Raf activity. Furthermore, we investigated how such Ras activity in E1/4 cells can be regulated by EGFR/ErbB4 heterodimer-mediated signaling. From a sensitivity analysis of the detailed upstream model for Ras activation, we concluded that Ras activation dynamics is dominated by heterodimer-mediated signaling coordination with a large initial speed of dimerization when the concentration of the ErbB4 receptor is considerably high. Such characteristics of the signaling cause the preferential binding of the Grb2-SOS complex to heterodimer-mediated signaling molecules.

Inference of S-system models of genetic networks using a genetic local search

In this paper, we propose a new method for the inference of S-system models of large-scale genetic networks. This method employs a technique to decompose the genetic network inference problem into several subproblems, and then applies a genetic local search to each of the subproblems. A local search method utilizing the feature of the S-system model is used as one of the search operators in this genetic local search. Finally, the effectiveness of the proposed method is verified through a genetic network inference problem.

The superstructure toward open bioinformatics grid

The grid design strongly depends on not only a network infrastructure but also a superstructure, that is, a social structure of virtual organizations where people trust each other, share resources and work together. Open Bioinformatics Grid (OBIGrid) is a grid aimed at building a cooperative bioinformatics environment for computer sicentists and biologists. In October 2003, OBIGrid consisted of 293 nodes with 492 CPUs provided by 27 sites at universities, laboratories and other enterprises, connected by a virtual private network over the Internet. So many organizations have participated because OBIGrid has been conscious of constructing a superstructure on a grid as well as a grid infrastructure. For the benefit of OBIGrid participants, we have developed a series of life science application services: an open bioinformatics environment (OBIEnv), a scalable genome database (OBISgd), a genome annotation system (OBITco), a biochemical network simulator (OBIYagns), and to name a few.