What we work on
We focus on epigenetic gene regulation by the Polycomb (PcG) and Trithorax (TrxG) groups of proteins.
Epigenetic memory is stable
The PcG and TrxG proteins are epigenetic regulators that work antagonistically on the same target genes, to maintain repressed (PcG) or active (TrxG) transcription states. These proteins bind to regulatory DNA elements called Polycomb/Trithorax response elements, or PRE/TREs. These fascinating regulatory elements are also known as “memory elements”. Depending on information received from the gene, a PRE/TRE can switch between active and silent states and can then maintain that state over many rounds of cell division, even after the initial transcription factors have disappeared. Thus, PRE/TREs can give epigenetic memory of gene expression states. In this role they are essential for correct development, and for the stable maintenance of cell identities.
Epigenetic memory elements ‘take' information from their associated gene, they ‘remember’ whether the gene was active or silent though replication and mitosis, and they ‘give back’ this information to the gene afterwards.
Epigenetic memory is flexible
However, there is another side to the story. Many target genes of the PcG and TrxG proteins are regulated in a dynamic way, switching several times during development or rapidly in response to external signals. If all transcriptional decisions were preserved forever by epigenetic memory, then life would come to a halt. So the PcG and TrxG system also has a built in flexibility, allowing some genes to escape memory when they need to, to forget their past, and switch more dynamically in response to new signals. Whether a PRE/TRE gives stable memory or allows flexibility depends on its interaction with promoters, on developmental context, genomic context and the nucleic acid sequence of the PRE/TRE itself.
Our goals
Fly PRE/TREs are about 1-2kb long and contain different combinations of short motifs (coloured bars).
Our goal is to understand how the PcG and TrxG proteins regulate genes, and why some PRE/TREs give memory, while others allow flexibility. Ultimately we want to understand the PRE/TRE code: how does the nucleic acid sequence of PRE/TREs quantitatively affect their function?
To address these questions we use a combination of molecular and developmental biology and quantitative live imaging in flies and in mouse cells, and mathematical and computational modelling.
Why fly and mouse?
One of the most interesting problems in the PcG/TrxG field can only be tackled by a combination of fly and mammalian experimental models. We call this problem the “mouse mystery”: Why are the PcG and TrxG proteins themselves and the genes they regulate so highly conserved between flies and mammals, while the DNA sequences that they bind (PRE/TREs) show no detectable similarity? We propose that there may be hidden similarities at some level, but to determine whether this is the case, we need a better understanding of the PRE/TRE code in both flies and mammals. Thus we work experimentally and computationally with both.
Why theory and experiment?
Theoretical models are immensely powerful but surprisingly rare in the field of epigenetics. Mechanistic mathematical models can explicitly capture dynamic processes that we cannot see, but can only infer from experiments. Computational modeling identifies hidden patterns in genomic sequence that are invisible to the eye. Both kinds of models identify unifying concepts and can predict system properties. Testing predictions by experimental perturbations tells us whether or not we have understood our system. Without such a coherent theoretical framework, we attempt to navigate in a sea of data without a compass.
Our research questions
What makes a PRE/TRE? Towards a PRE/TRE code
In a long running collaboration with the Rehmsmeier group, we combine computational approaches including machine learning (Bjørn, Paniz), with quantitative reporter assays in mouse (Natalia, Paniz) and fly model systems (Jeannette, Leonie) to understand the PRE/TRE code at the DNA sequence level.
DNA sequence motifs can be read directly by DNA binding proteins, or they may act at the level of non-coding RNAs, which are transcribed from many PRE/TREs (Jeannette). Other sequences may be more cryptic, acting to promote or prevent DNA or RNA structural changes (Vanessa, Leonie).
HOw does the system maintain memory?
In molecular terms, a “memory” of transcriptional activation or silencing has to survive both DNA replication and mitosis. These are very different molecular challenges and different mechanisms are likely to be required.
We have previously shown that the Trithorax group protein ASH1 binds robustly to mitotic chromatin in Drosophila. Current projects use single molecule tracking and Drosophila genetics, aiming to understand whether this binding is involved in bookmarking genes for reactivation after mitotic exit (Nancy, in collaboration with Davide Mazza).
We are also using mathematical modeling (Leonie) and reporter assays to investigate the inherent bistable properties of the PcG/TrxG system, and how these are modulated by developmental context and PRE/TRE identity.
Beyond memory: How does the system allow dynamic change?
Equally important to a full understanding of PcG/TrxG function is how the system allows a rapid response to signals, whether developmental or environmental. Several projects address this question from different angles, for example, are there different classes of PRE/TREs that are inherently designed for memory or flexible regulation during differentiation? (Paniz, Natalia). We have established mathematical models and experiments to address how epigenetic memory and dynamic regulation are modulated by developmental context and PRE/TRE identity (Leonie, Jeannette).