Habilitation à Diriger des Recherches

Mardi 3 novembre 2015
à 14h – amphi 2016, Campus de la Gaillarde


A Systems View of Nitrogen Signaling Interactions

Gabriel Krouk
BPMP, équipe Hormones, Nutriments et Développement


Composition du jury :
Loic Lepiniec
Francois Parcy
Susana Rivas
Sebastien Thomine
Anne Krapp
Laurent Laplaze


Résumé :
A drastic change in plant Nitrogen (N) nutrition results in systematic adaptations ranging from metabolic to growth changes. Interestingly, experimental evidences support the idea that it exists dedicated signaling pathways involved in the tuning of growth in response to nutritional status of the plant. On the other hand, growth can influence nutrition partly through hormones action. This constitutes a feed-forward loop that entangles nutrition and growth 1. This is our biological model.
We aim at getting deeper insights into such signaling interactions. To this purpose, two experimental approaches will be presented. First, genome wide investigations have been made to understand the effect of combinatorial interactions between nitrogen and hormone treatments in the control of i) gene expression and ii) root development. Multi- dimensional networks have been built and functional validations of the predicted roles for the genes belonging to these networks are currently made.
Second, by studying the genome wide effect of nitrate regulated transcription factors [technique named TARGET2], we yielded several insights into i) gene regulatory network complexity in Arabidopsis (unpublished), ii) transcription factor dynamics 3, iii) potential connections between nitrate and phosphate signaling in the control of root growth control4. All these aspects will be presented and discussed.
Finally, since we now have some cues about the Arabidopsis gene regulatory network topology, I will introduce FRANK: a Fast Randomizing Algorithm for Network Knowledge. FRANK generates in silico, very large GRNs having the known characteristics of Arabidopsis transcriptional network (and very likely to other eukaryotic genomes) and simulates gene expression (experiments) at a genome-wide scale. In our endeavour to develop stable GRNs (“stable” means: gene expression should be constant or in oscillation) we have defined basic mathematical rules that find echo in network biology. FRANK now helps to train supervised machine-learning algorithms in order to build GRNs on real transcriptomic data.

1 Krouk, G. et al. Trends Plant Sci,(2011) 16, 178-182
2 Bargmann, B. O. et al. Mol Plant,(2013) 6, 978-980
3 Para, A. et al. Proc Natl Acad Sci U S A,(2014) 111, 10371-10376
4 Medici, A. et al. Nature communications,(2015) 6, 6274