For multicellular organisms to maintain distinct and specialized cell types, the genomes of a given lineage have to be partitioned such that unwanted developmental programs become heritably repressed. This partitioning is carried out by a nuclear ultrastructure called heterochromatin. We now understand that in stem cells, genome partitioning is highly dynamic, in that pre-existing small heterochromatin regions progressively expand (?spread?) in differentiation. This dynamic behavior contrasts with our prior understanding, which indicated that heterochromatin regions are delimited by hard-wired, DNA sequence encoded ?start? and ?stop? sites. Yet, this dynamic heterochromatin spreading is absolutely required for normal differentiation, and we do not understand how it is mechanistically accomplished or regulated by the cell. Here, we propose to uncover the biochemical and genetic basis of dynamic heterochromatin spreading in embryonic stem cell differentiation. The dynamic spreading reaction remains opaque because it has not been addressed at the appropriate length scale and timescale. Heterochromatin spreading occurs over stretches of the chromosome, 100s to 1000s of nucleosomes, and over a certain time window after S-phase. Previous studies have focused on documenting genome-wide effects, obtained in static snapshots. To capture the dynamics of the process at the correct length- and timescale, we will deploy two cutting edge, purpose built testbeds: 1) We will follow heterochromatin spreading in real time in single stem cells, following the induction of differentiation programs. This will be enabled by a multiple fluorescent reporter based heterochromatin spreading sensor we recently validated. The dynamics of long-range spreading have to date not been visualized. 2) We will biochemically reconstitute the heterochromatin spreading process using modular, barcoded chromatin reagents. We will determine the propagation kinetics of histone methylation, the spreading signal, and the biochemical activity of boundary elements in stopping this process. The spreading reaction has to date not been successfully reconstituted. Using these testbeds, we can ask several critical questions about dynamic, heterochromatin-driven genome partitioning: How does the cell regulate dynamic spreading? One hypothesis would be that the spreading reaction itself is tuned. An alternative hypothesis is that spreading is promoted by removal of a ?block?, i.e. via dismantling of spreading boundaries, or rearranging chromosome structural territories. Our testbeds will allow us to distinguish between these hypotheses. Additionally, we will begin to address the question whether spreading is a driver of differentiation, or a consequence of it, exploiting the exquisite resolution of our single cell system. Finally, we intend to use dynamic heterochromatin spreading as a means to identify and remedy inefficiencies in differentiating lineages from induced pluripotent cells. We will use the combination of our testbeds to devise tools to artificially tune the spreading reaction at will, in order to achieve genetically stable lineage decisions and provide highly predictable differentiation for regenerative medicine.
In regenerative medicine, the ability to drive lineage choices starting from stem cells is key to success, yet this process remains stochastic and not always efficient, which has been linked to differences in the epigenome. Different lineages are typified by divergent patterns of genome partitioning, and in differentiation from stem cells, this essential process occurs in a dynamic and stepwise fashion, subjecting increasing stretches of the chromosome to heritable repression. We propose to uncover the cellular, genetic and biochemical basis for dynamic genome partitioning in single embryonic stem cells, and aim to use this information to engineer and stabilize lineage choices by manipulating the genome partitioning process.
| Greenstein, R A; Jones, Stephen K; Spivey, Eric C et al. (2018) Noncoding RNA-nucleated heterochromatin spreading is intrinsically labile and requires accessory elements for epigenetic stability. Elife 7: |
