Qualitative modelling and formal verification of the FLR1 gene mancozeb response in Saccharomyces cerevisiae.

BACKGROUND: Qualitative models allow understanding the relation between the structure and the dynamics of gene regulatory networks. The dynamical properties of these models can be automatically analysed by means of formal verification methods, like model checking. This facilitates the model-validation process and the test of new hypotheses to reconcile model predictions with the experimental data. RESULTS: The authors report in this study the qualitative modelling and simulation of the transcriptional regulatory network controlling the response of the model eukaryote Saccharomyces cerevisiae to the agricultural fungicide mancozeb. The model allowed the analysis of the regulation level and activity of the components of the gene mancozeb-induced network controlling the transcriptional activation of the FLR1 gene, which is proposed to confer multidrug resistance through its putative role as a drug eflux pump. Formal verification analysis of the network allowed us to confront model predictions with the experimental data and to assess the model robustness to parameter ordering and gene deletion. CONCLUSIONS: This analysis enabled us to better understand the mechanisms regulating the FLR1 gene mancozeb response and confirmed the need of a new transcription factor for the full transcriptional activation of YAP1. The result is a computable model of the FLR1 gene response to mancozeb, permitting a quick and cost-effective test of hypotheses prior to experimental validation.

Refining current knowledge on the yeast FLR1 regulatory network by combined experimental and computational approaches.

Multidrug resistance is often the result of the activation of drug efflux pumps able to catalyze the extrusion of the toxic compound to the outer medium, this activation being frequently controlled at the transcriptional level. Transcriptional regulation in the model eukaryote S. cerevisiae is the result of the interaction and cross-talk between networks of transcription factors. This is the case of the transcriptional activation of the FLR1 gene occurring in response to stress induced by the agricultural fungicide mancozeb in yeast. FLR1 up-regulation depends on the integrated action of Yap1, a key regulator of oxidative stress response, Pdr3 and Yrr1, two of the transcription factors controlling multidrug resistance, and Rpn4, a regulator of proteasome gene expression, which interplay to produce the observed transcriptional up-shift. Based on the expression profiles of FLR1, YAP1, PDR3, YRR1 and RPN4 registered during yeast adaptation to stress induced by mancozeb and using a qualitative modeling approach, a model of the FLR1 regulatory network was built, and the response of S. cerevisiae to mancozeb stress was simulated. The use of a qualitative approach is especially useful to overcome the lack of enough quantitative data on kinetic parameters and molecular concentrations, permitting the immediate focus on the qualitative behavior of the system. This Systems Biology approach allowed the identification of essential features of the early yeast response to fungicide stress. The resulting model allowed the formulation of new hypotheses, in a quick and cost effective manner, on the qualitative behavior of the system following mancozeb challenge, some of which were validated experimentally. In particular, Pdr3 and Yrr1 were shown to directly control FLR1 up-regulation in mancozeb-challenged cells, based on the analysis of the effect of the inactivation of their putative binding sites in the FLR1 promoter. Furthermore, the inter-dependent role of Yap1 and Yrr1 in the regulation of PDR3 and RPN4 was brought to light, this joint activity possibly being extensible to eight other genes involved in multidrug resistance. The FLR1 network structure was revised, based on the comparison between simulated and experimental gene expression data in the double deletion mutant strains Δyrr1Δpdr3 and Δyrr1Δrpn4, and an additional, still unidentified, transcription factor was found to be required to fully explain the behavior of the network.