Advances in optical imaging techniques have allowed quantitative studies of many biological systems. This dissertation elaborates on our efforts in both developing novel imaging modalities based on detection of optical absorption and applying high-sensitivity fluorescence microscopy to the study of biology. / Chemistry and Chemical Biology

Stochasticity is a ubiquitous feature of cellular processes such as gene expression that can give rise to phenotypic differences for genetically identical cells. Understanding how the underlying biochemical reactions give rise to variations in mRNA/protein levels is thus of fundamental importance to diverse cellular processes. Recent technological developments have enabled single-cell measurements of cellular macromolecules which can shed new light on processes underlying gene expression. Correspondingly, there is a need for the development of theoretical tools to quantitatively model stochastic gene expression and its consequences for cellular processes. In this dissertation, we address this need by developing general stochastic models of gene expression. By mapping the system to models analyzed in queueing theory, we derive analytical expressions for the noise in steady-state protein distributions. Furthermore, given that the underlying processes are intrinsically stochastic, cellular regulation must be designed to control the`noise' in order to adapt and respond to changing environments. Another focus of this dissertation is to develop and analyze stochastic models of post-transcription regulation. The analytical solutions of the models proposed provide insight into the effects of different mechanisms of regulation and the role of small RNAs in fine-tunning the noise in gene expression. The results derived can serve as building blocks for future studies focusing on regulation of stochastic gene expression. / Ph. D.

The methods that fueled the microscale revolution (top-down design/fabrication, combined with application of forces large enough to overpower stochasticity) constitute an approach that will not scale down to nanoscale systems. In contrast, in nanotechnology, we strive to embrace nature’s quite different paradigms to create functional systems, such as self-assembly to create structures, exploiting stochasticity, rather than overwhelming it, in order to create deterministic, yet highly adaptable, behavior. Nature’s approach, through billions of years of evolutionary development, has achieved self-assembling, self-duplicating, self-healing, adaptive systems. Compared to microprocessors, nature’s approach has achieved eight orders of magnitude higher memory density and three orders of magnitude higher computing capacity while utilizing eight orders of magnitude less power. Perhaps the most complex of functions, homeostatis by a biological cell – i.e., the regulation of its internal environment to maintain stability and function – in a fluctuating and unpredictable environment, emerges from the interactions between perhaps 50M molecules of a few thousand different types. Many of these molecules (e.g. proteins, RNA) are produced in the stochastic processes of gene expression, and the resulting populations of these molecules are distributed across a range of values. So although homeostasis is maintained at the system (i.e. cell) level, there are considerable and unavoidable fluctuations at the component (protein, RNA) level. While on at least some level, we understand the variability in individual components, we have no understanding of how to integrate these fluctuating components together to achieve complex function at the system level. This thesis will explore the regulation and control of stochasticity in cells. In particular, the focus will be on (1) how genetic circuits use noise to generate more function in less space; (2) how stochastic and deterministic responses are co-regulated to enhance function at a system level; and (3) the development of high-throughput analytical techniques that enable a comprehensive view of the structure and distribution of noise on a whole organism level.

Transkription är stokastisk, där utbrottsmässiga episoder av RNA-transkription genererar RNA-molekyler. Trots att detta är en kärndel av eukaryotiskt liv, är lite känt om hur DNA-bindande transkriptionsfaktorer och transkriptionella kofaktorer formar gen-specifik transkriptionell utbrottskinetik. Syftet med detta examensarbete var att tyda rollerna hos kofaktorerna Med14 och P300/CBP inom transkriptionell utbrottskinetik. För detta ändamål användes Auxin inducible degron systemet för snabb nedbrytning av Med14 eller P300/CBP-proteiner i HCT116-celler, följt av Smart-seq3xpress single cell-RNA-sekvensering. Ett särskild fokus i denna avhandling var även att utvärdera förmågan att härleda direkta  genuttrycksförändringar genom analys av introniska reads – detta då introner ko-transkriptionellt splitsas och dess nyttjande skulle fånga effekter av mycket närliggande transkription. Resultaten visar en tidsberoende minskning av introniskt innehåll och en nedreglering av genuttryck för majoriteten av generna i de behandlade cellinjerna, medan opåverkade kontroller inte visar sådana trender. Utbrottskinetikresultaten indikerar att det inte finns någon korrelation mellan P300/CBP-pertuberade cellers geners ursprungliga utbrottsstorlek och några trender i genuttryckets relativa förändring, medan detsamma kan sägas för Med14-pertuberade cellers geners utbrottsfrekvens. Svaga trender från P300/CBP-påverkade cellers utbrottskinetik och uttrycksändring kan antyda att deras utbrottsfrekvens och inte utbrottsstorlek har påverkats. Resultaten antyder att perturbationen var framgångsrik och att P300/CBP inte påverkar utbrottsstorlek samt att Med14 kan reglera utbrottsfrekvensen för alla påverkade gener i lika hög grad. Vidare forskning behövs inom utbrottskinetikdata för att utöka vår förståelse av denna studies implikationer gällande Med14:s och P300/CBP:s reglerande roller på transkriptionella utbrott. / Transcription is stochastic with episodes of RNA transcription generating bursts of RNA molecules. Despite being a core part of eukaryotic life, little is known about how DNA-binding transcription factors and transcriptional co-factors shape gene-specific transcriptional bursting kinetics. The aim of this thesis was to decipher the roles of the co-factors Med14 and P300/CBP in transcriptional burst kinetics. To this end, the Auxin inducible degron system was used for rapid Med14 or P300/CBP protein degradation in HCT116 cells, followed by Smart-seq3xpress single-cell RNA-sequencing. A particular focus of this thesis was to evaluate the abilities to infer direct gene expression changes by analysis of intronic reads – since introns are co-transcriptionally spliced and would capture very recent transcription. Results show a time dependent decrease of intronic contents and a downregulation in gene expression for a majority of genes in the perturbed cell lines, while unperturbed controls show no such trends. Bursting kinetics results indicate that there is no correlation between P300/CBP perturbed cells’ gene’s original bursting size and any trends in gene expression fold change while the same can be said for Med14 perturbed cell’s gene’s burst frequency. Weak trends from P300/CBP perturbed cells’ bursting kinetics and expression fold change could imply that their bursting frequency and not bursting size has been affected. The results imply that the perturbation was successful and that P300/CBP does not affect bursting size as well as that Med14 could regulate bursting frequency for all affected genes to an equal degree. Further research is needed into the bursting kinetics data to expand our understanding of this study’s implications regarding regulatory roles of Med14 and P300/CBP on transcriptional bursting.