Analysis of Mechanistic Pathway Models in Drug Discovery: p38 Pathway

Finding disagreement pathway signatures and constructing an ensemble model for cancer classification

Cancer classification based on molecular level is a relatively routine research procedure with advances in high-throughput molecular profiling techniques. However, the number of genes typically far exceeds the number of the sample size in gene expression studies. The existing gene selection methods are almost based on statistics and machine learning, overlooking relevant biological principles or knowledge while working with biological data. Here, we propose a robust ensemble learning paradigm, which incorporates multiple pathways information, to predict cancer classification. We compare the proposed method with other methods, such as Elastic SCAD and PPDMF, and estimate the classification performance. The results show that the proposed method has the higher performances on most metrics and robust performance. We further investigate the biological mechanism of the ensemble feature genes. The results demonstrate that the ensemble feature genes are associated with drug targets/clinically-relevant cancer. In addition, some core biological pathways and biological process underlying clinically-relevant phenotypes are identified by function annotation. Overall, our research can provide a new perspective for the further study of molecular activities and manifestations of cancer.

IL-1β-induced and p38MAPK-dependent activation of the mitogen-activated protein kinase-activated protein kinase 2 (MK2) in hepatocytes: Signal transduction with robust and concentration-independent signal amplification

The IL-1β induced activation of the p38^MAPK/MAPK-activated protein kinase 2 (MK2) pathway in hepatocytes is important for control of the acute phase response and regulation of liver regeneration. Many aspects of the regulatory relevance of this pathway have been investigated in immune cells in the context of inflammation. However, very little is known about concentration-dependent activation kinetics and signal propagation in hepatocytes and the role of MK2. We established a mathematical model for IL-1β-induced activation of the p38^MAPK/MK2 pathway in hepatocytes that was calibrated to quantitative data on time- and IL-1β concentration-dependent phosphorylation of p38^MAPK and MK2 in primary mouse hepatocytes. This analysis showed that, in hepatocytes, signal transduction from IL-1β via p38^MAPK to MK2 is characterized by strong signal amplification. Quantification of p38^MAPK and MK2 revealed that, in hepatocytes, at maximum, 11.3% of p38^MAPK molecules and 36.5% of MK2 molecules are activated in response to IL-1β. The mathematical model was experimentally validated by employing phosphatase inhibitors and the p38^MAPK inhibitor SB203580. Model simulations predicted an IC₅₀ of 1-1.2 μm for SB203580 in hepatocytes. In silico analyses and experimental validation demonstrated that the kinase activity of p38^MAPK determines signal amplitude, whereas phosphatase activity affects both signal amplitude and duration. p38^MAPK and MK2 concentrations and responsiveness toward IL-1β were quantitatively compared between hepatocytes and macrophages. In macrophages, the absolute p38^MAPK and MK2 concentration was significantly higher. Finally, in line with experimental observations, the mathematical model predicted a significantly higher half-maximal effective concentration for IL-1β-induced pathway activation in macrophages compared with hepatocytes, underscoring the importance of cell type-specific differences in pathway regulation.

Actinomyosin contraction, phosphorylation of VE-cadherin, and actin remodeling enable melanoma-induced endothelial cell-cell junction disassembly

During melanoma cell extravasation through the vascular endothelium, melanoma cells interact with endothelial cells through secretion of cytokines and by adhesion between proteins displayed on opposing cell surfaces. How these tumor cell associated signals together regulate the dynamics of intracellular signaling pathways within endothelial cells leading to endothelial cell-cell junction disruption is not well understood. Here, we used a combination of experimental and computational approaches to examine the individual and combined effects of activation of the vascular cell adhesion molecule (VCAM)-1, interleukin (IL)-8, and IL-1β signaling pathways on the integrity of vascular junctions. Our simulations predict a multifaceted interplay of signaling resulting from individual activation of VCAM-1, IL-8 and IL-1β pathways that is neither synergistic nor additive compared to all inputs turned on simultaneously. Furthermore, we show that the levels of phosphorylated proteins associated with actinomyosin contractility and junction disassembly peak prior to those related to actin remodeling. The results of this work provide insight into the dynamics of tumor-mediated endothelial junction disassembly and suggest that targeting proteins downstream of several interaction pathways may be the most effective therapeutic approach to reduce melanoma extravasation.

Network Modeling Approach to Predict Myofibroblast Differentiation

Fibrotic disease is a major cause of morbidity and mortality and is characterized by the transition of resident fibroblast cells into active myofibroblasts, identified by their expression of alpha smooth muscle actin. Myofibroblast differentiation is regulated by growth factor signaling and mechanical signals transduced through integrins, which converge at focal adhesion proteins (Src and FAK) and MAPK signaling, but lead to divergent outcomes. While details are known about individual pathways, little is known about their interactions. To this end, an ODE-based model of this cell signaling network was developed in parallel with in vitro experiments to analyze potential mechanisms of crosstalk and regulation of αSMA production. We found that cells lacking Src or FAK produce significantly less or more αSMA than wild type cells, respectively. Transforming growth factor beta 1 and fibroblast growth factor signal through ERK and MAPK p38 with different dynamic profiles to increase or decrease αSMA expression, respectively. Our model effectively recreated αSMA expression levels across a set of 22 experimental conditions and matched some features of transient phosphorylation of ERK and p38. These results support a potential mechanism for regulation of fibroblast differentiation: αSMA production is promoted by active p38 and Src and opposed by ERK.

Inferring cell-scale signalling networks via compressive sensing

Signalling network inference is a central problem in system biology. Previous studies investigate this problem by independently inferring local signalling networks and then linking them together via crosstalk. Since a cellular signalling system is in fact indivisible, this reductionistic approach may have an impact on the accuracy of the inference results. Preferably, a cell-scale signalling network should be inferred as a whole. However, the holistic approach suffers from three practical issues: scalability, measurement and overfitting. Here we make this approach feasible based on two key observations: 1) variations of concentrations are sparse due to separations of timescales; 2) several species can be measured together using cross-reactivity. We propose a method, CCELL, for cell-scale signalling network inference from time series generated by immunoprecipitation using Bayesian compressive sensing. A set of benchmark networks with varying numbers of time-variant species is used to demonstrate the effectiveness of our method. Instead of exhaustively measuring all individual species, high accuracy is achieved from relatively few measurements.

A systematic in silico mining of the mechanistic implications and therapeutic potentials of estrogen receptor (ER)-α in breast cancer

Estrogen receptor (ER)-α has long been a potential target in ER-α-positive breast cancer therapeutics. In this study, we integrated ER-α-related bioinformatic data at different levels to systematically explore the mechanistic and therapeutic implications of ER-α. Firstly, we identified ER-α-interacting proteins and target genes of ER-α-regulating microRNAs (miRNAs), and analyzed their functional gene ontology (GO) annotations of those ER-α-associated proteins. In addition, we predicted ten consensus miRNAs that could target ER-α, and screened candidate traditional Chinese medicine (TCM) compounds that might hit diverse conformations of ER-α ligand binding domain (LBD). These findings may help to uncover the mechanistic implications of ER-α in breast cancer at a systematic level, and provide clues of miRNAs- and small molecule modulators- based strategies for future ER-α-positive breast cancer therapeutics.

Biochemical fluctuations, optimisation and the linear noise approximation

BACKGROUND

Stochastic fluctuations in molecular numbers have been in many cases shown to be crucial for the understanding of biochemical systems. However, the systematic study of these fluctuations is severely hindered by the high computational demand of stochastic simulation algorithms. This is particularly problematic when, as is often the case, some or many model parameters are not well known. Here, we propose a solution to this problem, namely a combination of the linear noise approximation with optimisation methods. The linear noise approximation is used to efficiently estimate the covariances of particle numbers in the system. Combining it with optimisation methods in a closed-loop to find extrema of covariances within a possibly high-dimensional parameter space allows us to answer various questions. Examples are, what is the lowest amplitude of stochastic fluctuations possible within given parameter ranges? Or, which specific changes of parameter values lead to the increase of the correlation between certain chemical species? Unlike stochastic simulation methods, this has no requirement for small numbers of molecules and thus can be applied to cases where stochastic simulation is prohibitive.

RESULTS

We implemented our strategy in the software COPASI and show its applicability on two different models of mitogen-activated kinases (MAPK) signalling -- one generic model of extracellular signal-regulated kinases (ERK) and one model of signalling via p38 MAPK. Using our method we were able to quickly find local maxima of covariances between particle numbers in the ERK model depending on the activities of phospho-MKKK and its corresponding phosphatase. With the p38 MAPK model our method was able to efficiently find conditions under which the coefficient of variation of the output of the signalling system, namely the particle number of Hsp27, could be minimised. We also investigated correlations between the two parallel signalling branches (MKK3 and MKK6) in this model.

CONCLUSIONS

Our strategy is a practical method for the efficient investigation of fluctuations in biochemical models even when some or many of the model parameters have not yet been fully characterised.

Early developing pig embryos mediate their own environment in the maternal tract

The maternal tract plays a critical role in the success of early embryonic development providing an optimal environment for establishment and maintenance of pregnancy. Preparation of this environment requires an intimate dialogue between the embryo and her mother. However, many intriguing aspects remain unknown in this unique communication system. To advance our understanding of the process by which a blastocyst is accepted by the endometrium and better address the clinical challenges of infertility and pregnancy failure, it is imperative to decipher this complex molecular dialogue. The objective of the present work is to define the local response of the maternal tract towards the embryo during the earliest stages of pregnancy. We used a novel in vivo experimental model that eliminated genetic variability and individual differences, followed by Affymetrix microarray to identify the signals involved in this embryo-maternal dialogue. Using laparoscopic insemination one oviduct of a sow was inseminated with spermatozoa and the contralateral oviduct was injected with diluent. This model allowed us to obtain samples from the oviduct and the tip of the uterine horn containing either embryos or oocytes from the same sow. Microarray analysis showed that most of the transcripts differentially expressed were down-regulated in the uterine horn in response to blastocysts when compared to oocytes. Many of the transcripts altered in response to the embryo in the uterine horn were related to the immune system. We used an in silico mathematical model to demonstrate the role of the embryo as a modulator of the immune system. This model revealed that relatively modest changes induced by the presence of the embryo could modulate the maternal immune response. These findings suggested that the presence of the embryo might regulate the immune system in the maternal tract to allow the refractory uterus to tolerate the embryo and support its development.

Assessment of hepatotoxicity potential of drug candidate molecules including kinase inhibitors by hepatocyte imaging assay technology and bile flux imaging assay technology

Kinases are members of a major protein family targeted for drug discovery and development. Given the ubiquitous nature of many kinases as well as the broad range of pathways controlled by these enzymes, early safety assessments of small molecule inhibitors of kinases are crucial in identifying new molecules with sufficient therapeutic window for clinical development. Failure or attrition of drug candidates in late-stage pipelines due to hepatotoxicity is a significant challenge in the drug development field. Herein we provide detailed methods for the hepatocyte imaging assay technology (HIAT) and the bile flux imaging assay technology (BIAT) to evaluate drug-induced liver injury (DILI) potentials for drug candidates. Optimized culturing methods for primary human hepatocytes, both freshly isolated and prequalified cryopreserved cells, are also presented. The applications of these high-content cellular imaging technologies in the evaluation of p38 and Her2 kinase inhibitors are highlighted to illustrate the usefulness of the research methodology in a compound screening as well as mechanistic investigative setting.

Modelling the role of the Hsp70/Hsp90 system in the maintenance of protein homeostasis

Neurodegeneration is an age-related disorder which is characterised by the accumulation of aggregated protein and neuronal cell death. There are many different neurodegenerative diseases which are classified according to the specific proteins involved and the regions of the brain which are affected. Despite individual differences, there are common mechanisms at the sub-cellular level leading to loss of protein homeostasis. The two central systems in protein homeostasis are the chaperone system, which promotes correct protein folding, and the cellular proteolytic system, which degrades misfolded or damaged proteins. Since these systems and their interactions are very complex, we use mathematical modelling to aid understanding of the processes involved. The model developed in this study focuses on the role of Hsp70 (IPR00103) and Hsp90 (IPR001404) chaperones in preventing both protein aggregation and cell death. Simulations were performed under three different conditions: no stress; transient stress due to an increase in reactive oxygen species; and high stress due to sustained increases in reactive oxygen species. The model predicts that protein homeostasis can be maintained during short periods of stress. However, under long periods of stress, the chaperone system becomes overwhelmed and the probability of cell death pathways being activated increases. Simulations were also run in which cell death mediated by the JNK (P45983) and p38 (Q16539) pathways was inhibited. The model predicts that inhibiting either or both of these pathways may delay cell death but does not stop the aggregation process and that eventually cells die due to aggregated protein inhibiting proteasomal function. This problem can be overcome if the sequestration of aggregated protein into inclusion bodies is enhanced. This model predicts responses to reactive oxygen species-mediated stress that are consistent with currently available experimental data. The model can be used to assess specific interventions to reduce cell death due to impaired protein homeostasis.

Troubleshooting computational methods in drug discovery

Computational approaches for drug discovery such as ligand-based and structure-based methods, are increasingly seen as an efficient approach for lead discovery as well as providing insights on absorption, distribution, metabolism, excretion and toxicity (ADME/Tox). What is perhaps less well known and widely described are the limitations of the different technologies. We have therefore presented a troubleshooting approach to QSAR, homology modeling, docking as well as hybrid methods. If such computational or cheminformatics methods are to become more widely used by non-experts it is critical that such limitations are brought to the user's attention and addressed during their workflows. This could improve the quality of the models and results that are obtained.

A systems perspective of ras signaling in cancer

The development of cancer reflects the complex interactions and properties of many proteins functioning as part of large biochemical networks within the cancer cell. Although traditional experimental models have provided us with wonderful insights on the behavior of individual proteins within a cancer cell, they have been deficient in simultaneously keeping track of many proteins and their interactions in large networks. Computational models have emerged as a powerful tool for investigating biochemical networks due to their ability to meaningfully assimilate numerous network properties. Using the well-studied Ras oncogene as an example, we discuss the use of models to investigate pathologic Ras signaling and describe how these models could play a role in the development of new cancer drugs and the design of individualized treatment regimens.

Minocycline restores striatal tyrosine hydroxylase in GDNF heterozygous mice but not in methamphetamine-treated mice

Inflammation, phospho-p38 MAPK activation, and a reduction in glial cell line-derived neurotrophic factor (GDNF) occur in Parkinson's disease. Microglial activation in the substantia nigra and a tyrosine hydroxylase deficit in the striatum of 3-month-old GDNF heterozygous (GDNF(+/-)) mice were previously reported and both were exacerbated by a toxic methamphetamine binge. The current study assessed the effects of minocycline on these methamphetamine-induced effects. Minocycline (45 mg/kg, i.p.x 14 days post-methamphetamine or saline injections) reduced microglial activation and phospho-p38 MAPK in the substantia nigra of saline-treated GDNF(+/-) mice and in methamphetamine-treated wildtype and GDNF(+/-) mice. Although minocycline increased tyrosine hydroxylase-immunoreactivity in GDNF(+/-) mice, it did not attenuate the methamphetamine-induced reduction of tyrosine hydroxylase. The results suggest that neuroinflammation is deleterious to the dopamine system of GDNF(+/-) mice but is not the primary cause of methamphetamine-induced damage to the dopamine system in either GDNF(+/-) or wildtype mice.