PhD defense of Laura TURCHI prepared at Laboratoire de Physiologie Cellulaire Végétale, LPCV, and TIMC MaGE, on Thursday, the 26 of october, at 2pm:

" Predicting and understanding transcriptional regulation of floral development by LEAFY "

Jury:

• François PARCY, Research Director, CNRS delegation Alpes, Supervisor
• Klaas VANDEPOELE, Full prefessor, Universiteit Gent, Reporter
• Marie-Laure MARTIN, Research Director, INRAE Ile-de-France - Versailles-Saclay, Reporter
• Cristel CARLES, University Professor, Université Grenoble-Alpes, Examiner
• Gabriel KROUK, Research Director, CNRS Délégation Occitane Est, Examiner

Thesis supervision:

• François PARCY, Research Director, CNRS délégation Alpes, Supervisor
• Antoine FRENOY, Associate Professor, Grenoble INP - Université Grenoble Alpes, Co-supervisor

  Keywords

Transcription factors, Computational biology, Gene regulation, Integrative analysis, Machine learning

  Abstract

Ensuring correct gene expression is crucial for living organisms, as its disruption can compromise survival. Transcription factors (TFs) regulate gene expression through the binding of specific DNA sequences called transcription factor binding sites (TFBSs). LEAFY (LFY) is a plant-specific TF with a crucial role in floral development, and it is highly conserved in sequence and binding specificity throughout flowering plant evolution.

LFY’s central role in flowering has been studied for decades, and yet it remains unclear why only a subset of the genomic regions bound by LFY are regulated. To elucidate this point, I present, in the first part of this manuscript, an approach to predict transcriptional regulation of LFY TFBSs in the model plant Arabidopsis thaliana. I used state-of-the-art LFY TFBS models and the genetically-encoded genomic context of LFY sites to successfully build a classifier that can distinguish functional LFY sites (i.e. TFBSs that are bound and have an effect on gene expression in vivo) from nonfunctional ones (i.e. TFBSs that are not bound and are not associated with gene expression changes in vivo). My results suggest that the presence of surrounding LFY TFBSs and, to a lesser extent, the level of non-LFY TFBS diversity around LFY sites, are important to distinguish functional and nonfunctional LFY sites.
Moreover, this approach reveals a number of co-occurring TFs that contribute to set apart LFYregulated sites from nonfunctional ones. Despite previous evidence of the functional importance of conserved regions in gene regulation, including conservation of LFY sites in our model did not improve predictions, and I discuss some possible reasons behind this result. Overall, this approach allowed me to further characterize LFY’s binding to DNA, and it can be used on new genomic sequences to predict transcriptional regulation of LFY sites, as well as with new TFs.

In addition to working on its own, LFY interacts with UNUSUAL FLORAL ORGANS (UFO), an F-box protein, to ensure correct petal and stamen development. While the LFY-UFO interaction and their implication in flower development were already known, the exact role of UFO in this process had yet to be determined. In the second part of this manuscript, I include a recently published article on the transcriptional role of the LFY-UFO complex in flower development that allows access to genomic regions distinct from those bound by LFY alone. Moreover, I present some additional results suggesting the implication of LFY and UFO in floral meristem establishment in the early stages of flower development, broadening their importance in this crucial developmental process.