Noise and robustness downstream of a morphogen gradient: a quantitative approach by imaging transcription dynamics in living embryos – FIREFLY
The sparkling expression of genes
A major challenge in developmental biology is understanding how the right genes are activated in the right place and at the right time during development to produce and coordinate the many types of cells present in the adult organism. A multitude of transcription factors play an essential role in development, from the acquisition of cellular identity to terminal differentiation. In most cases, sufficient concentrations of a given transcription factor are necessary for its activity, and threshold concentration limits are essential for obtaining appropriate responses. In integrated cases, some of these factors are distributed as concentration gradients and can induce the expression of different sets of target genes at different concentrations. These factors, called morphogens, control molecular cascades that transform the information contained in their concentration gradient into spatial information, allowing the formation of discrete domains, each associated with a specific cellular fate. Morphogen gradients are used by various organisms to establish polarity along embryonic axes or within organs. The extreme robustness of these processes ensures the reproducibility of developmental patterns and the emergence of individuals with adequate proportions despite varying sizes and environmental conditions. Surprisingly, while most descriptions of developmental regulation assume that highly reproducible transcription programs directly control differentiation, when studied at the single-cell level, transcription appears as an extremely fluctuating process, difficult to reconcile with such precise control. Understanding how reproducible transcription patterns can robustly emerge from subtle differences in transcription factor concentration, given the inherent stochastic nature of transcription, is a major challenge and the overall goal of this research project.
The project relied on the expertise of N. Dostatni's team to develop the MS2 experimental approach, which allows direct observation of the transcription process in living Drosophila embryos (the sparkling expression of genes). To do this, the RNAs produced by the transcription process are labeled with fluorescence. The team was responsible for obtaining the biological material needed for the work, performing all the genetic manipulations on Drosophila flies required for imaging live embryos, and acquiring the imaging data in the form of films.
As part of the project, we analyzed synthetic reporters, which was very successful, but we also performed genome editing (CRISPR/Cas9) of Drosophila, which proved to be much more difficult than expected. L'imagerie des facteurs de transcription marqués par fluorescence a été réalisée sous la supervision de Cécile Fradin (Université Mc Master), experte en spectroscopie de corrélation de fluorescence (FCS), et de Mathieu Coppey (Institut Curie).
Data analysis and modeling were performed under the supervision of Aleksandra Walczak (LPENS).
This approach proved very productive and enabled us to better understand the role of the morphogen Bicoid and its partners Zelda and Hunchback in the transcriptional response to Bicoid. We were able to show that Bicoid is sufficient to provide most of the spatial characteristics to the expression of its main target gene, hunchback, while its known partners accelerate the process in different ways: the pioneer factor Zelda lowers the threshold concentration of Bicoid required for transcriptional activation, while Hunchback increases the polymerase initiation rate and reduces promoter bursting with fewer periods of inactivity. Halving the dose of Bicoid and quantifying the displacement of reporter gene boundaries dependent exclusively on Bicoid indicates that the Bicoid activity gradient has a lower exponential constant (stronger decay) than the Bicoid protein gradient. Our results also demonstrate quantitatively that Bicoid is indeed the main source of positional information for the early expression of its target gene, hunchback.
Using quantitative data obtained from these synthetic reporter genes and exploiting the structural differences between them, we were able to construct a model of transcriptional regulation equilibrium that is consistent with the reporter expression data placing the MS2 cassette under the control of the hunchback gene promoter. This regulatory model is highly informative. It is consistent with all of our experimental data and suggests that the transcription process involves several distinct mechanistic steps that we are currently seeking to characterize.
Our goal now is to determine how the different functional domains of the Bicoid protein contribute to the transcription process, and which domains of Bicoid are involved in its interactions with its two partners: Zelda and Hunchback. We have therefore begun a structure-function analysis of the Bicoid protein in vivo. Mutants are currently being obtained using Criper-Cas9 technology: our goal is to obtain deletion mutants at the endogenous bicoid locus. Analysis of the expression of synthetic reporter genes in these various mutants will enable us to determine at which mechanistic step of the transcription process each functional domain of Bicoid contributes.
Morphogen gradients are used by various organisms to establish polarity along embryonic axes or within organs. In these systems, it is assumed that positional information is provided by the concentration of the morphogen detected by each cell in the target tissue and responsible for the determination of cell identity. The extreme robustness of these processes ensures reproducibility of developmental patterns and emergence of properly proportioned individuals despite varying size and environmental conditions. Most descriptions of developmental regulation assume that highly reproducible transcription programs are directly controlling the mechanisms of differentiation. However, when studied at the single cell level, transcription is frequently observed to be an extremely noisy process, hardly suggestive of such precise control. Understanding how reproducible transcription patterns can robustly emerge from these smooth morphogen gradients given inherent stochastic transcription is an important challenge, and constitutes the general objective of this proposal.
Recently, methods to observe the kinetics of the transcription process directly in living cells have been developed. These methodes combine fluorescent labeling of nascent mRNA with live-cell imaging at high spatial and temporal resolution. We have recently adapted these approaches to one of the best characterized model organism, the fruit fly embryo. Our goal was to better understand how cell identity is controlled by the Bicoid (Bcd) morphogen system along the antero-posterior (AP) axis of the embryo. Focusing on the expression of the main Bcd target, hunchback (hb), at the onset of zygotic transcription, we have successfully built an MS2 reporter reproducing endogenous expression of the gene. Despite high nuclei-to-nuclei variability in transcription kinetics, the hb promoter was able to establish a sharp expression boundary along to separate anterior expressing from posterior non expressing nuclei. Surprisingly, it only takes ~3 min at each interphase for the system to measure subtle differences in Bicoid concentration and produce a complete sharp boundary.
If one assumes that the only driver for the hb transcription process is the Bcd gradient, and further that Bcd molecules reach their target sites by 3D diffusion, simple statistical mechanics considerations (combined with current estimates of Bcd concentration and mobility) show that a minimum of 25 min should be necessary for the system to produce the observed sharp boundary. The difference between predicted (25 min) and observed (~ 3 min) time-requirement is of 1 order of magnitude and calls for alternative explanations. We will explore the possibility of alternative mechanisms in an unbiased manner by combining genetics and optogenetics, live-imaging, advanced data analysis and modeling. Our goal is to identify mutant or natural variant contexts altering the dynamics of hb expression, characterize the contribution of spatial vs temporal correlations in the Bcd system to address the question of transcriptional memory, revisit the question of Bcd target search with the development of cutting-edge imaging to measure Bcd physical parameters (concentration/mobility) and use quantitative data driven modeling to shed light on the process and its robustness. The strength of our project, relies on a unique expertise in the MS2 approach combined with an established and productive collaboration between biologists, biophysicists and theoreticians.
Project coordination
Nathalie DOSTATNI LANTIERI (INSTITUT CURIE - SECT DE RECHERCHE)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
Partnership
IC INSTITUT CURIE - SECT DE RECHERCHE
LPENS Laboratoire de physique de l'ENS
McMaster University / Department of Physics and Astronomy
Help of the ANR 459,940 euros
Beginning and duration of the scientific project:
February 2020
- 36 Months
