Biological systems are dynamic and multi-layered, characterized by internal structures with varying levels of complexity that are in constant interplay. For instance, the regulated interaction between gene expression machinery and the biochemical reactions among proteins is crucial for orchestrating cellular functions. This interplay is essential for maintaining life, even in seemingly "simpler" single-cell organisms, amidst an ever-changing external environment. These interactions create a complex web of connections between different biological organization levels, making studying such systems enormously challenging. However, the story does not end there; biological systems have the remarkable ability to evolve. Evolution is fundamental to all living beings on Earth, enabling the vast diversity of forms, shapes, lifestyles, and colors observed in nature. Anticipating evolutionary outcomes has long perplexed scientists. In some cases, evolution appears to follow reproducible trajectories, providing opportunities to investigate factors that may constrain evolutionary paths while controlling for external environmental in_uences. One such factor, discovered in the past century, is epistasis, which refers to the variable effect of a gene mutation depending on the presence or absence of mutations in other genes. This concept has played a pivotal role in shaping our understanding of evolutionary processes. In this thesis, we examine the effects of epistatic mutations on a genome-wide scale within a speci_c evolutionary trajectory...

Multivalent interactions are crucial mechanisms employed by cells to respond to their environment, often leading to superselectivity phenomena. Extensive experimental and theoretical efforts have been made to understand the variables controlling superselectivity, but challenges persist in achieving precise control over receptor and ligand numbers in biological systems. To address this, the Laan lab developed a model system using DNA origami, enabling precise manipulation of receptor and ligand numbers at the nanoscale. These nanoscopic (~15 nm) structures can mimic the receptors of a target surface, and the extracellular ligands are represented by a branched star-shaped DNA origami structure in solution. Both structures hold a fluorophore, allowing visualization using total internal reflection microscopy (TIRF) by measuring intensity values of the DNA nanostars absorbed into target surface. In this study, we investigated the effects of altering binding strength and flexibility in DNA nanostar structures on superselectivity. We found that experimental results for replicating previous experiments using the same DNA nanostars (Design A) exhibited minor variations within expected ranges, validating the reliability of the experimental protocol. However, sensitivity analysis highlighted the influence of data points on superselectivity interpretation, emphasizing the need for careful data analysis. Our study also evaluated the impact of introducing sequence mismatches on binding affinity (Design A*), revealing that modifications reducing binding affinity do not necessarily enhance superselective behavior for this system. Additionally, our investigation into the effects of increased flexibility (Design C) revealed unexpected behaviors in bound fraction and cluster formation, suggesting a potential relationship between cluster formation, intensity values registered, and the flexibility of the structure mediated by phase separation. These findings underscore the complexity of DNA nanostar behavior and stress the need for further research to elucidate the factors influencing superselectivity in DNA nanostars. By providing more understanding into how to develop highly selective particles, therapeutic molecules could sharply distinguish between healthy and corrupt cells, leading to customizable treatments with higher efficacy.

To bind two entities together, an attractive interaction is needed. In biological systems, such interactions are often between ligands and receptors. But this interaction constantly breaks and forms because it is (too) weak. To ensure a lasting bond, the system can form multiple weak bonds that form an overall strong bond – similar to velcro. An interesting feature of aweak multivalent systemis the sharp discrimination between surfaces based on receptor density. That means that when multivalent particles encounter surfaces with the specific receptor but different densities, they will most likely bind to the surface with the highest density, because it has the highest binding probability. This phenomenon is called superselectivity and emerges from the large entropic contribution in amultivalent system: The more ligands and receptors are involved in the binding, the more possibilities the system has to form a bond and hence a large entropy. In this thesis we investigate how the interaction strength and entropy influences superselective binding. In doing so, we study superselective binding of microparticleswith hundreds of interactions and, additionally, particles with only few interactions of the size of nanometers.

In budding yeast, a certain concentration of the GTPase Cdc42 is optimal for cell division. This optimal concentration depends on the phenotype and genetic background of the cells. Due to bet hedging, the Cdc42 concentration is different between cells in a population and it is expected that the concentration, measured for many cells, is gamma distributed. This study describes a method for determining the concentration of Cdc42 proteins depending on the phenotype of individual cells. Fluorescence microscopy images were analyzed using custom designed software for tracking single cells and detecting budding and polarization events, based on existing segmentation software. This allows for estimating the cell volume and determining the concentration distribution based on the fluorescence intensity. The estimated cell volume is larger than was expected which is proven to be caused by segmentation errors. The intensity is shown to scale linearly with the number of fluorescent sources and using this result, the distribution for the Cdc42 concentration and copy number can be determined. For calculating the absolute concentration and copy number values, a constant still needs to be determined. To assess the reliability of the obtained results, validation measurements should be performed, for example by using different galactose concentrations to control the Cdc42 production by means of a galactose promoter or by using different genetic backgrounds. Based on the results, this method shows a possible way of determining the budding and polarization events and measuring the protein concentration for individual cells.