| Abstract |
Epigenetics comprises chemical modifications of the DNA and the proteins that the DNA
is wrapped around them. These modifications play key roles in establishing and maintain-
ing cellular identity throughout development and adulthood. In recent years, it has become
increasingly clear that these actions are more dynamic than initially believed. The alter-
ation of cellular identities during regeneration, ageing, and the formation of tumors is closely
linked to systematic changes in epigenetic modifiers. The emergence of cutting-edge single-
cell sequencing technologies has enabled thorough explorations of biological processes with
high molecular precision. Nevertheless, the regulation of cellular behavior is intricately tied
to collective processes occurring in both spatial and temporal dimensions, operating on the
mesoscopic and macroscopic scales. However, these larger scales cannot be straightforwardly
deduced from microscopic measurements along the DNA sequence [1]. Consequently, the
findings obtained from sequencing experiments stay at the descriptive level until they are
coupled with methodologies capable of discerning collective degrees of freedom. Here, using
statistical physics tools and sequencing technologies, we study the collective processes un-
derlying epigenetic dynamics in cells that change their identity over time. Specifically, we
investigate collective epigenetic processes during ageing and the reprogramming of cells after
injury.
In the first part of this thesis, we study the mechanistic basis of epigenetic modifications
during ageing. Despite the accuracy of machine learning models in predicting the biological
age based on epigenetic DNA methylation marks, these tools do not inform about the mech-
anistic basis of epigenetic ageing. We show that epigenetic ageing is reflected in systematic
and collective changes in DNA methylation marks during ageing, which manifests in the
stereotypical behavior of two-point correlation functions. We devise a stochastic theory that
comprises competition of antagonistic enzymes at the boundaries of genomic regions with
atypically high content of cytosine-guanine pairs. We systematically coarse-grain this theory
to derive a macroscopic description in terms of a phase-field theory. This model predicts the
changes in two-point correlation functions during ageing and explains diverse observations
in the field of epigenetic ageing.
In the second part of this thesis, we study the collective epigenetic processes during the
regeneration of the liver after injury. In particular, we study the interplay between DNA
methylation and the accessibility of chromatin and show the necessity for emergent memory
of past injuries in the system. This memory is achieved by considering an effective projection
between different scales of epigenetic modifications.
In total, in this thesis, we derived theoretical descriptions of epigenetic processes that have so
far only been studied descriptively. We showed that both epigenetic alterations during ageing
and during reprogramming rely on an interplay between collective biochemical processes and
the geometry of the DNA. With this work, we show how linear DNA sequencing can inform
about collective epigenetic processes in space and time. |