Self-organisation and dynamics of 2D DNA-kinetoplast

Recent years have witnessed an ever-increasing attention on the properties of topologically interlinked (catenated) ring polymers. This class of promising topologically linked materials, includes both naturally occurring biological catenated polymers and synthetically produced DNA catenanes. The kinetoplast DNA in particular, is a unique mitochondrial structure that is common to some unicellular flagellar human pathogens. Constituting a complex structure formed by thousands of interlinked DNA-mini-rings, the kinetoplast DNA is uniquely suited as a model system for the study of 2D catenated polymers. Despite recent experimental progress, our fundamental knowledge of the theoretical physics of such two-dimensional catenated ring polymers has so far attracted limited interest. Our focus here is on understanding the structural properties of the kinetoplast under confinement or flow, with the aim of improving the fundamental understanding of the physical mechanisms that determine its self-organization. This work is computational in nature, proposing a two-level coarse-graining strategy for the DNA minicircles that constitute the building blocks of the kinetoplast. The first level involves coarse-graining from an all-atom representation to the nucleotide level ("effective monomer level"). Proceeding to a second level of coarse-graining, DNA minicircles are viewed as penetrable rings and all effective physical interactions are further reduced to a center-ofmass level, point-like description. By employing a proper two-level coarse graining model, thus retaining both base-resolution and coarse-grained resolution of the topologically linked DNA constituents, our primary aim is to quantify and analyze both the effects of the flow characteristics and their interplay with polymer size and architecture. The guiding hypothesis of this project is that through the variation of the geometry and topology of 2D catenated ring polymers (degree of topological linking and contour length of the DNA minicircle building blocks), and manipulation of the properties of imposed flow, we can eventually drive the manufacturing of two-dimensional polymers with desired Gaussian curvature. Complementary to our computational approach, a collaboration with the experimental group of Prof. Alex Klotz has been established, to provide a means of validation of our models. Recent work by the external collaborator of this project, Alexander Klotz, has experimentally established some fascinating facts about the mesoscopic structure of the kinetoplast, yet very little is known from the theoretical point of view. Our work will contribute to the understanding of the physical mechanisms that determine the self- organization and the elastic properties of the kinetoplast by employing accurate modeling across the scales.

Recent years have witnessed an ever-increasing attention on the properties of topologically interlinked (catenated) ring polymers, including both naturally occurring biological catenated polymers, and synthetically produced DNA catenanes. Despite experimental progress, our fundamental knowledge of the theoretical physics of catenated ring polymers still lags substantially behind that of other polymeric or interlocked systems (e.g., rotaxanes). The overarching goal of this project is to employ advanced computational tools and combine simulations across scales, in order to investigate the structural properties and self-organization of 2D catenated ring polymer systems, and provide realistic descriptions of their morphology and physical response (e.g., to mechanical stretching or fluid flow). The research activity has been focused on the derivation of bottom-up, coarse-grained (CG) models of a bead-spring type for the individual double stranded (ds) DNA minicircles that comprise catenated DNA systems. Two levels of coarse graining have been implemented: a) The first level of coarse-graining involves upscaling from an all-atom (AA atomistic) to a reduced, base-pair level (CG atomistic) description, and the derivation of effective potentials thereof has been performed by use of the Iterative Boltzmann Inversion method. b) The second level of coarse-graining involves upscaling from an all-atom (AA atomistic) description to an effective particle-like, center-of-mass minicircle description, and the respective effective potentials have been computed from umbrella sampling molecular dynamics simulations by employing the weighted histogram analysis method. Two types of effective interactions have been determined: a) interactions between directly bonded DNA minicircles and interactions between DNA minicircles that are indirectly bonded through their linking to another DNA minicircle. An important finding of this project is that the circular DNA topology can give rise to contrasting effects of divalent counterions on the effective pair potentials: while DNADNA repulsion in short center-of-mass distances decreases significantly in the presence of divalent counterion-ions (as compared to monovalent ions) for linear DNA fragments, the opposite effect occurs for moderately short DNA minicircles. This can be attributed to the fact that linear DNA fragments can easily adopt relative orientations that minimize electrostatic and steric repulsions by rotating relative to one another and by exhibiting more pronounced bending due to the presence of free ends. Another important finding of the project is the ion-specific modulation of the conformation and compactness of DNA oligo-catenanes: an interesting interplay arises depending on the contour length of the individual DNA minicircle constituents, and can give rise to a non-intuitive, non-monotonic variation of the overall size of DNA-catenanes upon increasing ionic strength of divalent counterions.