Using steroid hormone action in breast cancer cells as a model, the group explores how eukaryotic cells respond to external cues; specifically, how signals are transduced to the nucleus and modulate chromatin structure and gene expression. Previously, we found that gene regulation by progestins involves the communication of membrane-attached and nuclear progesterone receptor (PR) associated with ERK and MSK1 kinases and involves two consecutive cycles of chromatin remodelling: a very rapid displacement of histone H1 mediated by NURF, ASCOM/MLL2, CDK2, PARP1 and KDM5, followed by a slower displacement of histones H2A/H2B mediated by PCAF and BAF. The Mb topological chromatin domains (TADs) conserved in the human genome behave as units of hormone response that undergo structural changes in response to hormone. Unexpectedly we found that the ATP required for the associated chromatin remodelling is generated in the nucleus from ADP-Ribose and PPi by the enzyme NUDIX5. Our final aim is to integrate the signalling network with the changes in the topological organization of chromatin and the transcriptional response to generate a multidimensional network that will reveal the logic of the hormonal control of cell proliferation.

Research projects

  • Role of genome architecture in cell identity and in the response to perturbations (Ferrari, Le Dily, Lioutas, Nacht, Quilez, Sharma, Vicent, Vidal, Wright). Progestin (Pg) addition to T47D breast cancer cells activates the Progesterone Receptor (PR) that binds to nucleosomaly organized promoter/enhancer regions of target genes along with kinases, histone modifying enzymes and chromatin remodeller, leading to coordinate activation or repression of genes clustered in topological associating domains (TADs) (Le Dily et al. Genes Dev 2014). In the context of the ERC Synergy Grant “4D Genome” and in collaboration with Thomas Graf, Marc Marti-Renom and Guillaume Filion and the 4D Genome Unit (Francois Le Dily, Enrique Vidal) we continue exploring the relevance of genome architecture for cell identity, the response to perturbations – such as inhibition of transcription, energy supply or signalling – and the epigenetic memory of various normal and cancer cells. We have generated a catalogue of lncRNAs expressed in the breast cancer cell line T47D and have identified a few that change expression in response to hormone. We want to explore their potential contribution to genome architecture using CRISPR-cas9 approaches.
  • ADP-Ribose derived nuclear of ATP generation in chromatin remodeling and gene regulation (Lioutas, Vicent, Wright). PARP1 mediated parylation is essential for hormonal gene regulation, but degradation of PAR by PARG is also important. Using ATP-reporters we found that in cells exposed to hormone the levels of ATP increase transiently in the cell nuclei, but not in mitochondria or in the cytoplasm. The nuclear ATP increase depends on PARP1 and PARG activity and cannot be blocked by blocking mitochondrial respiratory 10 min after hormone exposure. We identified NUDT5/NUDIX5 as a PAR-interacting enzyme that co-IP with PR and PARP1 and is needed for chromatin remodelling, gene regulation, and hormone-dependent cell proliferation. NUDIX5 forms a stable homodimer that hydrolyses ADPR to AMP and R-5-P. In response to hormone NUDIX5 is rapidly dephosphorylated at T45 leading to destabilization of the homodimer that facilities the reaction of ADPR with pyrophosphate to generate ATP and R-5-P. We are exploring the structure of the enzyme chain involved in nuclear ATP synthesis, including NMNAT1, PARP1, PARG and NUDIX5 and their role in the DNA damage response and transcriptional activation. NUDIX5 is overexpressed in breast cancer samples and we want to develop small molecular weight inhibitors of ATP synthesis by NUDIX5 that could be of use in cancer therapy, alone or in combination with PARP inhibitors.
  • Role of TFIIIC and Pol-III transcription in genome organization and gene regulation (Ferrari, Llobet). TFIIICcan act as an insulator that participates in 3D genome organization. We found an intimate relationship between TFIIIC and PR in breast cancer cells. Upon serum starvation TFIIIC is relocated from Pol-III genes to repetitive elements near Po-II gene, and upon hormone exposure PR binds nearby and activates transcription. We hypothesized that TFIIIC might be a sensor of growth signal that contributes to genomic rearrangements needed for rapid response to nutrient changes and we will explore whether its contribution to the cancer phenotype.
  • Role of PADI2-mediated citrullination of the Pol-II CTD in transcription elongation (Sharma, Vicent). Among the PADI family of ­protein arginine de-iminases, T47D cells only express PADI2 and PADI3. In response to hormone PADI2 is needed for hormonal gene regulation, interacts with PR and citrullinates R1810 in Pol-II CTD repeat 31. A Pol-II R1810K mutant Pol-II is defective in promoter release. We are exploring how hormones control PADI2 activity and genome location.