Exploring the operational characteristics of inference algorithms for transcriptional networks by means of synthetic data.

The development of structure-learning algorithms for gene regulatory networks depends heavily on the availability of synthetic data sets that contain both the original network and associated expression data. This article reports the application of SynTReN, an existing network generator that samples topologies from existing biological networks and uses Michaelis-Menten and Hill enzyme kinetics to simulate gene interactions. We illustrate the effects of different aspects of the expression data on the quality of the inferred network. The tested expression data parameters are network size, network topology, type and degree of noise, quantity of expression data, and interaction types between genes. This is done by applying three well-known inference algorithms to SynTReN data sets. The results show the power of synthetic data in revealing operational characteristics of inference algorithms that are unlikely to be discovered by means of biological microarray data only.

Molecular responses during cadmium-induced stress in Daphnia magna: integration of differential gene expression with higher-level effects.

DNA microarrays offer great potential in revealing insight into mechanistic toxicity of contaminants. The aim of the present study was (i) to gain insight in concentration- and time-dependent cadmium-induced molecular responses by using a customized Daphnia magna microarray, and (ii) to compare the gene expression profiles with effects at higher levels of biological organization (e.g. total energy budget and growth). Daphnids were exposed to three cadmium concentrations (nominal value of 10, 50, 100microg/l) for two time intervals (48 and 96h). In general, dynamic expression patterns were obtained with a clear increase of gene expression changes at higher concentrations and longer exposure duration. Microarray analysis revealed cadmium affected molecular pathways associated with processes such as digestion, oxygen transport, cuticula metabolism and embryo development. These effects were compared with higher-level effects (energy budgets and growth). For instance, next to reduced energy budgets due to a decline in lipid, carbohydrate and protein content, we found an up-regulated expression of genes related to digestive processes (e.g. alpha-esterase, cellulase, alpha-amylase). Furthermore, cadmium affected the expression of genes coding for proteins involved in molecular pathways associated with immune response, stress response, cell adhesion, visual perception and signal transduction in the present study.

Validating module network learning algorithms using simulated data.

BACKGROUND: In recent years, several authors have used probabilistic graphical models to learn expression modules and their regulatory programs from gene expression data. Despite the demonstrated success of such algorithms in uncovering biologically relevant regulatory relations, further developments in the area are hampered by a lack of tools to compare the performance of alternative module network learning strategies. Here, we demonstrate the use of the synthetic data generator SynTReN for the purpose of testing and comparing module network learning algorithms. We introduce a software package for learning module networks, called LeMoNe, which incorporates a novel strategy for learning regulatory programs. Novelties include the use of a bottom-up Bayesian hierarchical clustering to construct the regulatory programs, and the use of a conditional entropy measure to assign regulators to the regulation program nodes. Using SynTReN data, we test the performance of LeMoNe in a completely controlled situation and assess the effect of the methodological changes we made with respect to an existing software package, namely Genomica. Additionally, we assess the effect of various parameters, such as the size of the data set and the amount of noise, on the inference performance. RESULTS: Overall, application of Genomica and LeMoNe to simulated data sets gave comparable results. However, LeMoNe offers some advantages, one of them being that the learning process is considerably faster for larger data sets. Additionally, we show that the location of the regulators in the LeMoNe regulation programs and their conditional entropy may be used to prioritize regulators for functional validation, and that the combination of the bottom-up clustering strategy with the conditional entropy-based assignment of regulators improves the handling of missing or hidden regulators. CONCLUSION: We show that data simulators such as SynTReN are very well suited for the purpose of developing, testing and improving module network algorithms. We used SynTReN data to develop and test an alternative module network learning strategy, which is incorporated in the software package LeMoNe, and we provide evidence that this alternative strategy has several advantages with respect to existing methods.

Effluent impact assessment using microarray-based analysis in common carp: a systems toxicology approach.

Effluents are a main source of direct and continuous input of pollutants to the aquatic environment, and can cause ecotoxicological effects at different levels of biological organization. Since gene expression responses represent the primary interaction site between environmental contaminants and biota, they provide essential clues to understand how chemical exposure can affect organismal health. The aim of the present study was to investigate the applicability of a microarray approach for unraveling modes of action of whole effluent toxicity and impact assessment. A chronic toxicity test with common carp (Cyprinus carpio) was conducted where fish were exposed to a control and 100% effluent for 21 days under flow-through conditions. Microarray analysis revealed that effluent treatment mainly affected molecular pathways associated with the energy balance of the fish, including changes in carbohydrate and lipid metabolism, as well as digestive enzyme activity. These gene expression responses were in clear agreement with, and provided additional mechanistic information on various cellular and higher level effects observed for the same effluent. Our results demonstrate the benefit of toxicogenomic tools in a "systems toxicology" approach, involving the integration of adverse effects of chemicals and stressors across multiple levels of biological complexity.

Daphnia magna and ecotoxicogenomics: gene expression profiles of the anti-ecdysteroidal fungicide fenarimol using energy-, molting- and life stage-related cDNA libraries.

In the present study, the existing life stage-specific cDNA library was extended with energy- and molting-related genes using Suppression Subtractive Hybridization PCR and a microarray for the aquatic test organism Daphnia magna was created. A gene set of 2455 fragments was produced belonging to different pathways such as carbohydrate and lipid metabolism, O2 transport and heme metabolism, immune response, embryo development, cuticula metabolism and visual perception pathways. Using this custom microarray, gene expression profiles were generated from neonates exposed to three concentrations of the anti-ecdysteroidal fungicide fenarimol (0.5, 0.75, 1 microg/ml) during 48 h and 96 h. In total, 59 non-redundant genes were differentially expressed, of which more genes were down- than up-regulated. The gene expression data indicated a main effect on molting specific pathways. At the highest concentration, a set of proteolytic enzymes - including different serine proteases and carboxypeptidases - were induced whereas different cuticula proteins were down-regulated (48 h). Moreover, effects on embryo development were demonstrated at the gene expression as well as at the organismal level. The embryo development related gene vitellogenin was differentially expressed after 96 h of exposure together with a significant increase in embryo abnormalities in the offspring. This study suggests that this Daphnia magna microarray is of great further value for the elucidation of molecular mechanisms of toxicity and for the future development of specific biomarkers for hazard characterization.

Patterns of gene expression in carp liver after exposure to a mixture of waterborne and dietary cadmium using a custom-made microarray.

Gene expression changes in carp liver tissue were studied after acute (3 and 24h) and subchronic (7 and 28 days) exposure to a mixture of waterborne (9, 105 and 480 microg/l) and dietary (9.5, 122 and 144 microg/g) cadmium, using a custom-made microarray. Suppression subtractive hybridization-PCR (SSH-PCR) was applied to isolate a set of 643 liver genes, involved in multiple biological pathways, such as energy metabolism (e.g. glucokinase), immune response (e.g. complement C3) and stress and detoxification (e.g. cytochrome P450 2F2, glutathione-S-transferase pi). These genes were subsequently spotted on glass-slides for the construction of a custom-made microarray. Resulting microarray hybridizations indicated a highly dynamic response to cadmium exposure. At low exposure concentrations (9 microg/l through water and 9.5 microg/g dry weight through food) mostly energy-related genes (e.g. glucokinase, elastase) were influenced, while a general stress response was obvious through induction of several stress-related genes, including hemopexin and cytochrome P450 2F2, at high cadmium concentrations. In addition, fish exposed to the highest cadmium concentrations showed liver damage after 7 days of exposure, as measured by elevated alanine transaminase activity in plasma and increased liver water content (wet-to-dry weight ratio). Moreover, decreased hematocrit and growth were found at the end of the experiment. Altogether this study clearly demonstrated the importance of varying exposure conditions for the characterization of the molecular impact of cadmium and showed that microarray results can provide important information, required to unravel the molecular events and responses related to cadmium exposure.

Neuropharmaceuticals in the environment: mianserin-induced neuroendocrine disruption in zebrafish (Danio rerio) using cDNA microarrays.

Because of their environmental occurrence and high biological activity, human pharmaceuticals have received increasing attention from environmental and health agencies. A major bottleneck in their risk assessment is the lack of relevant and specific effect data. We developed an approach using gene expression analysis in quantifying adverse effects of neuroendocrine pharmaceuticals in the environment. We studied effects of mianserin on zebrafish (Danio rerio) gene expression using a brain-specific, custom microarray, with real-time polymerase chain reaction as confirmation. After exposure (0, 25, and 250 microg/L) for 2, 4, and 14 d, RNA was extracted from brain tissue and used for microarray hybridization. In parallel, we investigated the impact of exposure on egg production, fertilization, and hatching. After 2 d of exposure, microarray analysis showed a clear effect of mianserin on important neuroendocrine-related genes (e.g., aromatase and estrogen receptor), indicating that antidepressants can modulate neuroendocrine processes. This initial neuroendocrine effect was followed by a "late gene expression effect" on neuronal plasticity, supporting the current concept regarding the mode of action for antidepressants in mammals. Clear adverse effects on egg viability were seen after 14 d of exposure at the highest concentration tested. Based on the specific molecular impact and the effects on reproduction, we conclude that further investigation of the adverse effects on the brain-liver-gonad axis is needed for a correct ecological risk assessment of antidepressants.

Expression profiling of endocrine-disrupting compounds using a customized Cyprinus carpio cDNA microarray.

Exposure to a variety of anthropogenic compounds has been shown to interfere with normal development, physiology, and reproduction in a wide range of organisms, both in laboratory studies and wildlife. We have developed a Cyprinus carpio cDNA microarray consisting of endocrine-related genes. In the current study, we investigated the applicability of this microarray (1) to study the molecular effects induced by exposure to a variety of endocrine-disrupting compounds (EDCs) in fish and (2) to discriminate the specific transcriptional profiles associated with these compounds. To that purpose, gene expression profiles were generated in livers of juvenile carp exposed to 14 Organization of Economical Cooperation Development (OECD)-recommended reference EDCs (17beta-estradiol, 17alpha-ethinylestradiol, 4-nonylphenol, bisphenol A, tamoxifen, 17alpha-methyltestosterone, 11-ketotestosterone, dibutyl phthalate, flutamide, vinclozolin, hydrocortisone, CuCl(2), propylthiouracil, and a mixture of L-triiodothyronine and L-thyroxine). Our results show that, in addition to some expression similarities between analogous acting substances, each individual compound produced its own unique expression pattern on the array, distinct from the profiles generated by the other compounds. In addition, we were able to identify a minimal subset of genes, which also allowed to discriminate between the different compounds. Overall, our findings suggest that the developed cDNA array has great promise to screen new and existing chemicals on their endocrine-disruptive potential and to identify distinct classes of EDCs.

Effects of the antidepressant mianserin in zebrafish: molecular markers of endocrine disruption.

Due to their environmental occurrence and intrinsic biological activity, human pharmaceuticals have received increasing attention from environmental and health agencies. Of particular, ecotoxicological concern are drugs that affect nervous- and endocrine-systems. Zebrafish genome-wide oligo arrays are used to collect mechanistic information on mianserin-induced changes in gene expression in zebrafish. Gene expression analysis in brain and gonad tissue clearly demonstrated the estrogenic activity of mianserin and its potency to disrupt normal endocrine (estrogenic) signaling, based on induction of molecular biomarkers of estrogenicity (e.g., vitellogenin1 and zona pellucida proteins). The possible mechanism underlying this estrogenic activity of mianserin is disturbance of the Hypothalamo-Pituitary-Gonadal (HPG) axis by direct interference of mianserin with the serotonergic and adrenergic systems in the brain of zebrafish. Taking into account the importance of the HPG-axis, and considering the concept of 'critical window of exposure', our results reveal the importance for more elaborate testing of endocrine disruptive effects of aquatic antidepressants at different lifestages and during longer exposure periods (e.g., life cycle studies). Although there is a low concordance between the gene expression results in this study and previous cDNA microarray hybridizations, the global mechanistic expression patterns are similar in both platforms. This argues in favor of pathway-driven analysis of gene expression results compared to gene-per-gene analysis.

SynTReN: a generator of synthetic gene expression data for design and analysis of structure learning algorithms.

BACKGROUND: The development of algorithms to infer the structure of gene regulatory networks based on expression data is an important subject in bioinformatics research. Validation of these algorithms requires benchmark data sets for which the underlying network is known. Since experimental data sets of the appropriate size and design are usually not available, there is a clear need to generate well-characterized synthetic data sets that allow thorough testing of learning algorithms in a fast and reproducible manner. RESULTS: In this paper we describe a network generator that creates synthetic transcriptional regulatory networks and produces simulated gene expression data that approximates experimental data. Network topologies are generated by selecting subnetworks from previously described regulatory networks. Interaction kinetics are modeled by equations based on Michaelis-Menten and Hill kinetics. Our results show that the statistical properties of these topologies more closely approximate those of genuine biological networks than do those of different types of random graph models. Several user-definable parameters adjust the complexity of the resulting data set with respect to the structure learning algorithms. CONCLUSION: This network generation technique offers a valid alternative to existing methods. The topological characteristics of the generated networks more closely resemble the characteristics of real transcriptional networks. Simulation of the network scales well to large networks. The generator models different types of biological interactions and produces biologically plausible synthetic gene expression data.