Bioinformatic analyses of the structural and functional complexity in chromosomal interactomes

Evolution requires information storage systems with different demands with respect to persistence. While the genome provides a mechanism for long term, static and accurate information storage, it is incapable of mediating adaptation to short term changes in the environment. Chromatin, however, constitutes a dynamic, reprogrammable memory with different levels of persistence. Moreover, chromatin states carry information not only in 2D, i.e. in the structure of the primary chromatin fibre, but also in the 3D organization of the genome in the nuclear space. The following thesis delves into the new bioinformatic and wet lab protocols developed to map, quantitative and functionally analyze the 3D architecture of chromatin. The chromatin insulator protein CTCF is a major factor underlying the 3D organization of the epigenome. We have uncovered, however, that CTCF binding sites within a regulatory region have multiple functions that are influenced by the chromatin environment and possibly the combinatorial usage of the 11 Zn-fingers of CTCF (Paper I). This observation exemplifies that understanding the function of dynamic and transient chromatin fibre interactions requires novel technology that enables the detection of 3D chromatin folding with high resolution in single cells and in small cell populations. We therefore set out to devise a novel method for the visualization of higher order chromatin structures by combining the strengths of both DNA Fluorescent In Situ Hybridization (FISH) and In Situ Proximity Ligation Assay (ISPLA) technologies (Paper II). The resulting Chromatin in Situ Proximity (ChrISP) assay thus takes advantage of the direct contact detection of ISPLA and the locus-specific nature of FISH and uncovered the existence of compact chromatin structures at the nuclear envelope with unprecedented resolution. To complement ChrISP with a high throughput method capable of quantitatively recovering chromatin fibre contacts in small cell populations, we furthermore innovated the Nodewalk assay (Paper III). The protocol builds on existing ligation based chromosome conformation capture methods, but features significant reduction in the random ligation event frequency, inclusion of negative and positive ligation controls, iterative template resampling, increased signal to noise ratio and improved sensitivity. Using this technique, we have uncovered a cancer cell-specific, productive chromatin fibre interactome connecting the promoter and enhancer of c-MYC to a network of enhancers and super-enhancers. Underpinning this new protocol, I have developed the Nodewalk Analysis Pipeline (NAP) (Paper IV). This suite of tools consists of preprocessing, analysis and post-processing modules designed specifically for the rapid and efficient analysis of Nodewalk datasets through an interactive and user-friendly web based interface. Overall the work described in this thesis advances our understanding of the role of CTCF in nuclear organization and provides innovative wet lab techniques along with specialized software tools. Moreover, this work is an example of an emerging trend where the challenge of understanding chromatin dynamics within the 3D nuclear architecture demands a close synergistic collaboration between the fields of biology, biotechnology and bioinformatics.

List of scientific papers

I. Guibert, S., Zhao, Z., Sjölinder, M., Göndör, A., Fernandez, A., Pant, V., Ohlsson, R. CTCF-binding sites within the H19 ICR differentially regulate local chromatin structures and cis-acting functions. Epigenetics. 2012;7(4):361-9.
https://doi.org/10.4161/epi.19487

II. Chen, X., Shi, C.,Yammine, S., Göndör, A., Rönnlund, D., Fernandez-Woodbridge, A., Sumida, N., Widengren, J., Ohlsson, R. Chromatin in situ proximity (ChrISP): single-cell analysis of chromatin proximities at a high resolution. Biotechniques. 2014;56(3):117-8, 120-4.
https://doi.org/10.2144/000114145

III. Sumida, N., Fernandez-Woodbridge, A., Göndör, A., Ohlsson, R. Nodewalk, an ultra-sensitive technique to quantitatively analyze stochastic chromatin interactomes, identifies long range-acting super-clusters of productive enhancers. [Manuscript]

IV. Fernandez-Woodbridge, A., Göndör, A., Sumida, N. Nodewalk Analysis Pipeline: a bioinformatic analysis platform for pre-processing, analysis and post-processing of Nodewalk libraries. [Manuscript]