Our work aims to elucidate how cells acquire and maintain their identity during development. This question is particularly critical in the context of cancer as, in tumors, cells have very frequently changed their identity. As in eukaryotes the “epigenome” predominantly determines cellular identity, we focus on the role of chromatin dynamics to maintain genome integrity and gene expression during development. Our model system, the Drosophila fruit fly, is ideally suited for classical and molecular genetic approaches, as it develops rapidly and is particularly well adapted for cutting-edge imaging. Our recent work focuses on the CAF-1 chromatin assembly factor and on the Bicoid transcription factor gradient.
The dynamics of chromatin during development
In order to better understand the maintenance of chromatin structure across cell division, we focus on CAF-1, which allows assembly of H3/H4 dimers on newly replicated/repaired DNA and contributes to the formation/maintenance of heterochromatin. Our model system permits in vivo investigations on the whole organism during development. We investigate CAF-1 function in specific cell types with unique features that lack appropriate in vitro culture models. These cell populations include i) meiotic cells, which divide according to a unique process leading to haploid gametes and ii) stem cells, which divide asymmetrically and require a direct interaction with somatic cells of the niche for proper development.
We characterized a novel evolutionary conserved HP1a interacting domain in the large subunit of Drosophila CAF-1. In contrast to complete loss of function, deletion of this domain does not perturb larval viability but reveals heterochromatin associated functions of CAF-1 in position effect variegation and homologous chromosome pairing during oocyte meiosis (Roelens et al., 2017). In addition, we demonstrated that female germline stem cells depleted of the CAF-1 large subunit still respond to niche signals but also express differentiation markers. In addition to this “identity crisis”, these cells exhibit elevated levels of replicative stress and de-repression of transposable elements. These DNA damage activate p53- and chk-2-dependent checkpoints leading to cell death and sterility. This work, performed in collaboration with the team of J.R. Huynh (Institut Curie, DEEP Labex), demonstrates that CAF-1-mediated chromatin dynamics plays a specific role to regulate stem cells identity and maintain their genome integrity (Clémot et al., 2018).
The Bicoid morphogenetic gradient
To better understand how cells acquire and maintain their identity during development, we concurrently focus on the transcriptional response downstream of the Bicoid morphogen gradient, a well characterized and simple paradigm patterning the embryonic antero-posterior axis. We combine integrative approaches from molecular genetics to quantitative imaging. In collaboration with physicists, we developed multi-scale imaging approaches, on fixed and living embryos, to understand the transcription process in a quantitative manner.
To understand how the Bicoid system rapidly provides a precise transcriptional response despite the challenge imposed by very frequent mitoses, we adapted to the fly embryo the MS2 approach that allows RNA labelling in living cells and provides access to the transcription process at a single locus resolution in real time (Lucas et al., 2013). This pioneering work created a path to the temporal (4th) dimension to understand how reproducible transcription patterns can robustly emerge from a smooth gradient given the inherent stochastic nature of transcription. We successfully built an MS2 reporter reproducing endogenous expression of hunchback, the main Bcd target, at the onset of zygotic transcription (see the movie).
Using this tool, we discovered that hunchback transcription per se is stochastic in early fruit fly embryos. However, advanced data analysis indicated that this stochasticity cannot be explained by random firing of the polymerase (Poisson model) only, pointing towards more complex activation models (Desponds et al., 2016). We found that the shape of the boundary separating hunchback expressing from non-expressing nuclei is extremely sharp and consistent with the snapshots obtained from RNA FISH. Thus despite high nucleus-to-nucleus variability in transcription kinetics, the hunchback promoter is able to establish a sharp expression boundary that separates the anterior from the posterior of the embryo. Surprisingly, despite the absence of transcription during the frequent mitoses at this stage of development, it only takes 3 min at each interphase for the system to measure subtle differences in Bicoid concentration and produce a complete sharp border (Lucas et al., 2018). This rapid responsiveness is fascinating because it is almost ten times faster than predicted by theoretical models which have to integrate how distinct read-outs result from the trade-offs between the short time to establish a boundary, the observed steepness of this boundary and the positional resolution of nuclei on both sides of the boundary (Tran et al., 2018). This work results from an interdisciplinary collaboration with biophysicists (M. Coppey at the Institut Curie/UMR168 & C. Fradin at McMaster University, Hamilton, Canada) and theoreticians (A. Walczak at the ENS, Paris).