Using a novel large-scale epigenetic analysis method to uncover the cis-regulatory potential of individual transposable elements

Transposable elements (TEs) are DNA sequences with the potential to move or replicate and insert themselves at new genomic locations. Over evolutionary time, many sequences derived from TEs have accumulated and account for 50% of the human genome. These mobile elements were initially considered mostly neutral and/or detrimental to the host. However, it is now clear that some individual elements can be essential for normal human development and healthy function. In particular, they can act as cis regulatory elements of nearby genes (i.e. enhancers or alternative promoters) in a cell type-specific manner. Epigenetic regulation of TEs is an important feature that our genome has evolved to restrict their transposition potential. One such mechanism is orchestrated by KRAB zinc finger proteins (KZFPs). These form the largest family of DNA binding factors in tetrapods, with more than 350 protein-coding members in humans. While little is known about their individual function, they have a recognised general role in targeting TEs and epigenetically silencing them. This is achieved by recruitment of TRIM28 and the deposition of the repressing H3K9me3 histone modification, which leads to formation of heterochromatin and restricts accessibility of transcription factors. Similarly to TEs, it was initially hypothesised that this silencing mechanism was exclusively aimed at restricting their ability to spread in our genome. However, since KZFPs’ expression patterns vary in a cell-type specific manner, it was recently hypothesised that they can act as epigenetic switches to control chromatin accessibility of regulatory platforms derived from old TEs. The experimental determination of cell types and contexts where TEs are either repressed by KZFPs or active and could be contributing to regulation of genes is a complex problem. The number of protein-coding KZFPs is large and each of them targets various TE subfamilies, each of which can contain thousands of individual sequences spread in the genome. Therefore, computational methods are needed to identify genomic locations with similar epigenetic characteristics. In this thesis, I present EpiDRAW (Epigenetic Dimensionality Reduction Analysis Workflow) – a novel multi-step method which facilitates the analysis of large-scale genome wide epigenetic datasets. It performs signal quantification, normalisation, dimensionality reduction and clustering of an arbitrarily large number of samples, and outputs tabular data and interactive graphical reports that can be used to further explore variability at selected subsets of genomic regions. I use EpiDRAW to analyse bulk chromatin accessibility (ATAC-seq) data from the ENCODE database, as well as a large public single-cell ATAC seq dataset from adult and fetal tissues. These analyses reveal numerous subsets of TEs which are differentially accessible in a cell type specific manner. I then identify TE subfamilies enriched in specific cellular contexts, as well as enrichment of binding sites of specific KZFPs. I showcase an example of TE activity in the differentiation pathway of oligodendrocytes by combining information from public datasets of transcription factor binding. Finally, I analyse a dataset generated in our research group which uses TRIM28 as a proxy for KZFP activity across multiple cell lines. This can be used to identify the cellular context in which the activity of specific KZFPs can be investigated further, to understand the importance of silencing their TE targets. Moreover, I show that transposable elements bound by KZFPs in the TRIM28 ChIP-exo are accessible in other cellular contexts, showcasing the differential regulatory potential of KZFPs and their TE targets. In conclusion, this work provides a powerful new method of finding TEs with potential cis regulatory activity and demonstrates how it can be used to generate future testable hypotheses for assessing their individual involvement in gene regulatory networks.

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

University of Cambridge

Year dc:date.issued

2024

Author dc:creator

• Voicu, Ioana Diana

Advisor dc:contributor.advisor

• Imbeault, Michael

• epigenetics
• bioinformatics
• transposable elements
• KZFPs

Licence

• https://apollo8-f-pro.lib.cam.ac.uk/bitstreams/22620453-36e1-4b98-99ed-cc839ab11ac4/download
• https://creativecommons.org/licenses/by/4.0/

DOI dc:identifier.doi

https://doi.org/10.17863/CAM.119630

OAI identifier oai:identifier

oai:www.repository.cam.ac.uk:1810/386403