Deciphering the chromatin binding mechanisms of human HP1 variants at the single-molecule level

In eukaryotic cells, DNA is tightly packed in the form chromatin. The basic structure of chromatin is a nucleosome composed of 147 bp DNA wrapped around eight histone proteins; two copies of H2A, H2B, H3 and H4. These histone proteins are decorated with patterns of post-translational modifications leading to a direct change in chromatin structure or recruitment of effector proteins. Heterochromatin protein 1 (HP1) is an effector protein, that binds with low micromolar affinity to trimethylated lysine 9 on histone H3 (H3K9me3). It has been shown that HP1 localizes to specific domains where its binding is highly dynamic, however it is not well understood how HP1 is efficiently recruited to heterochromatin sites enacting stable gene silencing. The aim of this project was therefore to develop a single molecule microscopy method in total internal fluorescence (TIRF) to study the dynamic binding of HP1 towards chemically synthesized chromatin fibers. This would enable us to understand how HP1, which has low affinity towards H3K9me3, is able to increase its affinity and localization to chromatin. Initially, three models were studied; 1) The co-existence of many low-affinity binding sites in chromatin fibers allows rapid re-binding of HP1a after dissociation. 2) Stable and long-lived complexes are the result of HP1a oligomerization on the chromatin fiber. 3) Multivalent binding interactions of dimeric HP1a increases binding affinity towards chromatin. We found that HP1α residence time on chromatin depends on the density of H3K9me3, where the dissociated protein can rapidly rebind to a neighbouring site. Multivalency by HP1a dimerization leads to longer retention and accelerates the association rate and HP1a does not form oligomers but competes for free H3K9me3 sites. Following on from these results we aimed to improve our understanding of how HP1 dynamically samples the chromatin landscape, by studying the influence of chromatin states over HP1a binding, as well as finding out the binding differences between the HP1 isoforms (HP1a, b and g) and the influence of a phosphorylation mark on HP1a. We found that HP1a exhibits the longest residence times and fastest binding rates due to DNA interaction as well as H3K9me3 binding, confirmed by in vivo fluorescence recovery after photobleaching (FRAP) experiments. Interestingly, phosphorylated HP1a increases retention through strengthening of multivalency while reducing DNA binding. From the smTIRF results, a kinetic model was developed to dissect the detailed mechanism of HP1-chromatin binding, revealing a multivalent interaction network provided by multiple weak protein and DNA interactions. The final part of this thesis was focused towards improving our single molecule TIRF set up by developing a multiplexed read-out of nucleosomes. This will provide a way to measure the interaction of a chromatin binding protein towards a library of nucleosomes in one single molecule experiment. The technique relies on the nucleosomal DNA containing a short ssDNA site where the complementary labeled DNA piece can hybridize dynamically. The work presented in this thesis, was based on using a single molecule microscopy technique to shed light on the recruitment of an effector protein to chromatin and on the development of a new technique to further investigate more complicated chromatin-binding proteins.