Investigating mechanisms of genome folding by single-cell Hi-C

In recent years researchers have begun to reveal the hierarchy of mammalian DNA folding from the 10 nm fibre to the intact whole genome. One of the key approaches used is the single-cell Hi-C. My lab and I developed a single-cell Hi-C protocol that combines imaging and Hi-C processing on the same cell, and I tried to improve the sequencing library processing steps using a commercial Tn5 transposase. My work shows that without significant optimisation the commercial transposase is not compatible with our protocol. Using the protocol, we successfully processed haploid mouse embryonic stem cells (mESCs) and calculated 3D structures of their entire genomes. The structures show several genome-wide features that are highly conserved across all mESCs, including: the genome shape is an ellipsoid; the chromosomes arrange in a Rabl configuration; the A and B compartments form a bowl-like structure; and active enhancers, promoters along with gene expression cluster together in the A compartment. In contrast, relatively local features were found to significantly vary from cell to cell, including the folding of: individual chromosomes, topological-associated domains and loops. We also investigated mESCs in early stages of differentiation using the same single-cell Hi-C protocol. Preliminary analysis of chromosome and genome structures at 24 and 48 hour post differentiation reveals that certain features vary greatly. In particular, the genome shape of cells after 24 hour differentiation is relatively flat, whilst genomes after 48 hour differentiation are both ellipsoid and flat. Interestingly, the shapes of chromosomes from cells expressing the pluripotent marker Rex1 after 24 hour differentiation are similar to the shapes of chromosomes in ES cells; whereas chromosome shapes from cells both after 24 hour differentiation with low Rex1 expression and after 48 hour differentiation are even more varied