I aimed to be able to expand techniques that are available to distinguish sister chromatids from one another for chromosome conformation analysis. I chose to work with diploid immortalised human cell lines as my basis for understanding DNA conformation through the cell cycle. To complete this aim I decided to therefore modify and optimise a technique known as Sister-C in yeast for human cell application.
To distinguish two exactly identical chromatids from one another I chose to label the newly replicating DNA with a thymidine analogue. DNA normally comprises four different bases attached to the sugar phosphate backbone. These are Adenine, Guanine, Cytosine and Thymidine. In this project I needed to replace Thymidine with a different nucleoside analogue. I worked on trying different analogues to find one that would be the most suitable; incorporating efficiently and not causing toxicity to the cells. I finally chose to replace the Thymidine with a synthetic nucleoside analogue, also used in the yeast version of the technique, called 5-bromo-2′-deoxyuridine (BrdU) so that one half of the DNA duplex is labelled.
To be able to label DNA at an exact point in the cell cycle I chose to synchronise the human cells so that all cells in one experiment are going through the cell cycle together. Multiple cell lines and synchronization protocols were tested to find a method of synchronising as many cells as possible together. I synchronised the cells before the replication phase of the cell cycle which is the timepoint at which I added the BrdU. Therefore, as the cells all synchronously replicate, BrdU will be incorporated into the newly made DNA. Due to the semi conservative nature of DNA replication the half of the DNA strand which was used as a template will still contain the thymidine base and the newly synthesized other half of the duplex incorporates BrdU instead of T in most instances.
In order to study chromosome conformation I then optimised chromosome conformation capture techniques followed by next generation sequencing or long read sequencing in order to perform them on the sister chromatids. I crosslinked the cells followed by fragmenting DNA and ligating together pieces of DNA that were in close proximity to one another. Traditionally these can then be sequenced in order to build up a map of which pieces of DNA are close to one another in 3D space. However, for my project the BrdU containing DNA strands are first degraded by UV light to leave only the non-BrdU containing fragments. We sequence the remaining fragments and analyse the results to understand which sister chromatid each piece of DNA originally came from. If the remaining contacts are between the same strand of DNA (i.e. Watson-Watson) this will constitute a cis sister interaction and if the contacts are between different strands of DNA (i.e. Watson-Crick) this will come from a trans sister interaction.
I worked extensively with bioinformaticians to try to understand how best to analyse and visualise the results of the sequencing data. We built up new pipelines and computational methods for processing proximity ligated sister chromatid DNA data, to be able to simply show striking results in an easily understandable format.
This new human cell culture technique opens the world of chromatin conformation analysis to endless possibilities for further applications, including further study of human cell cycle dynamics and the role DNA conformation has in disease.
