The Stress-Response Network in Animals: Proposals to Develop a Predictive Mathematical Model

Increasing evidence indicates that numerous genetic pathways responding to environmental stress in animals are regulated co-ordinately as well as independently. These stress-response systems should therefore be viewed in holistic terms as a network. As such, their behaviour is susceptible to mathematical modelling using a systems biology approach. This review outlines relevant evidence and describes a newly launched project to develop just such a model using stress- response data from multiple transgenic strains of C. elegans and D. melanogaster. We hope that our eventual model will be capable of predicting the effects of simple stressor mixtures with reasonable accuracy. To maximise the effectiveness and scope of this model, we appeal for help from colleagues to share reagents and data relevant to this project. We also present preliminary data where RNA interference has implicated the key transcription factor DAF-16 in an unexpected up- regulation of cyp-34A9 reporter expression by high cadmium. 1. STRESS RESPONSES AND MIXTURE TOXICITY In multicellular organisms, the defensive cellular re- sponses evoked by environmental stresses do not result from simple linear pathways, but rather from a network of inter- linked pathways with multiple outputs. This makes it diffi- cult to predict the biological effects of multiple stressors acting together, even though this is the normal situation for industrial pollution of soil or water, where several different contaminants are usually present together. There are few studies and no useful predictive models describing the mo- lecular responses of multicellular organisms to several toxi- cants acting in concert. This is essentially a systems biology problem, requiring integration of complex molecular and toxicological information. Under the auspices of a Major Award from the UK-India Education and Research Initiative (UK-IERI), we intend to develop an in silico model describ- ing the principal elements of a consensus stress response network (SRN) and its in vivo responses to single stressors, using data from two invertebrate model systems, the nema- tode Caenorhabditis elegans and the fruit fly Drosophila melanogaster. This model will be used to predict the likely SRN responses to stressor mixtures, and such predictions will then be tested experimentally in both species so that the model can be refined accordingly. Since the SRN core path- ways are highly conserved among animal taxa, general fea- tures of this model should find wider application in ecotoxi-