Browsing by Author "Buchler, Nicolas E"

Biological oscillators are ubiquitous in living systems. They allow cellular processes to anticipate and act in synchrony with regular events in the outside world (such as the day/night cycle), or they ensure that processes occur in a particular order. Living things typically contain multiple oscillators, which can often couple to each other and influence each other's timing and function. The purpose of this thesis has been to investigate the relationship between two coupled oscillators in \textit{Saccharomyces cerevisiae}: the yeast metabolic cycle and the cell division cycle. I have focused on two key questions: what is the biological significance of their coupling, and is one oscillator dominant in its interaction with the other?

First, I investigated the temporal relationship between the cell division cycle and metabolic shifts that occur during the metabolic cycle across diverse yeast strains. I showed that a particular cell cycle event (DNA replication) was consistently delayed relative to a metabolic event (entry into the high oxygen consumption phase). This suggested that an earlier cell cycle event (Start and commitment to the cell cycle) was tied to the onset of high oxygen consumption. Second, I used fluorescent probes to examine the relationship between the metabolic cycle and the commitment to cell cycle progression at single-cell resolution. This revealed that cells enter high oxygen consumption phase of the metabolic cycle before passing Start, supporting a model of metabolic cycle/cell division cycle coupling in which the shorter metabolic cycle controls cell cycle commitment, likely via modulation of cell size thresholds.

One of the main challenges that biological oscillators face at the cellular level is maintaining coherence in the presence of molecular noise. Mechanisms of noise resistance have been proposed, however the findings are sometimes contradictory and not universal. Another challenge faced by biological oscillators is the proper timing of cellular events and effective distribution of cellular resources when there is more than one oscillator in the same cell. Biological oscillators are often coupled, however, the mechanisms and extent of these couplings are poorly understood. In this thesis, I describe three separate yet interconnected projects in an attempt to understand these biophysical phenomena.

I show that slow DNA unbinding rates are important in titration-based oscillators and can mitigate molecular noise. Multiple DNA binding sites can also increase the coherence of the oscillations through protected states, where the DNA binding/unbinding between these states has little effect on gene expression. I then show that experimental titration-based oscillator in budding yeast is innately coupled to the cell cycle. The oscillator and the cell cycle show 1:1 and 2:1 phase locking similar to what has been observed in natural systems. Finally, by studying the relationship between the circadian redox rhythm and genetic circadian clock in plants I show how perturbation of one of the coupled oscillators can be transformed into a reinforcement signal for the other one via a balanced network architecture.

In the past decade, a challenge to the canonical model of cell cycle transcriptional control has been posed by a series of high-throughput gene expression studies in budding yeast. Using genetic methods to inhibit or lock the activity of the cyclin-CDK/APC oscillator, these population studies demonstrated that a significant proportion of cell cycle transcription persists in the absence of cyclin-CDK/APC oscillations. To account for these findings, a network of serially activating transcription factors with sources of negative feedback from transcriptional repressors (referred to as a \say{TF network}) was proposed to drive cyclin-CDK/APC independent gene expression.

However, population studies of cell cycle gene expression are limited due to loss of phase synchrony that limits the timescale of measurement of gene expression and due to expression averaging that limits assessment of heterogeneity of expression within the population. To circumvent these limitations I used a single-cell timelapse microscopy approach to assess transcriptional dynamics of cell cycle regulated genes during extended cell cycle arrests in both the Gl/S and early mitosis (metaphase) phases of the cell cycle.

During G1/S arrest, transcriptional dynamics of four cell cycle regulated genes was assessed and activation of out-of-phase cell cycle transcription was observed in two of these genes. Though budding oscillations were observed in G1/S arrested cells, robust transcriptional oscillations were not seen for any of the four genes and budding dynamics were uncoupled from transcriptional dynamics after the first bud emergence. During cell cycle arrest in early mitosis, transcriptional dynamics of ten cell cycle regulated genes was assessed and activation of out-of-phase transcription was observed for four genes. All four genes activated once with canonical ordering but robust oscillations were not observed during mitotic arrest. Together these studies demonstrate activation, but not oscillation, of cell cycle transcription in the absence of cyclin-CDK/APC oscillations.

How organisms adapt and survive in continuously fluctuating environments is a central question of evolutionary biology. Additionally, organisms have to deal with the inherent stochasticity in all cellular processes. The purpose of this thesis is to gain insights into how organisms can use epigenetics and the stochasticity of gene expression to deal with a fluctuating environment. To accomplish this, two cases at different temporal and structural scales were explored: (1) the early transcriptional response to an environmental stimulus in single cells, and (2) the evolutionary dynamics of a population adapting to a recurring fluctuating environment. Mathematical models of stochastic gene expression, population dynamics, and evolution were developed to explore these systems.

First, the information available in sparse single cell measurements was analyzed to better characterize the intrinsic stochasticity of gene expression regulation. A mathematical and statistical model was developed to characterize the kinetics of a single cell, single gene behavior in response to a single environmental stimulus. Bayesian inference approach was used to deduce the contribution of multiple gene promoter states on the experimentally measured cell-to-cell variability. The developed algorithm robustly estimated the kinetic parameters describing the early gene expression dynamics in response a stimulus in single neurons, even when the experimental samples were small and sparse. Additionally, this algorithm allowed testing and comparing different biological hypotheses, and can potentially be applied to a variety of systems.

Second, the evolutionary adaptation dynamics of epigenetic switches in a recurrent fluctuating environment were studied by observing the evolution of gene regulatory circuit in a population under multiple environmental cycles. The evolutionary advantage of using epigenetics to exploit the natural noise in gene expression was tested by competing this strategy against the classical genetic adaptation through mutations in a variety of evolutionary conditions. A trade-off between minimizing the adaptation time after each environmental transition and increasing the robustness of the phenotype during the constant environment between transitions was observed. Surviving lineages evolved bistable, epigenetic switching to adapt quickly in fast fluctuating environments, whereas genetic adaptation with high robustness was favored in slowly fluctuating environments.

The cell cycle is arguably one of the most conserved regulatory networks within Eukaryotes. Despite the animals and fungi are sibling “kingdoms” within the Opisthokont supergroup, the core transcription factors that control commitment to cell division (E2F and SBF, respectively) and their repressors (Rb and Whi5, respectively) do not appear to have a shared molecular origin. My thesis work has focused on understanding how the networks that regulate cell cycle decisions have changed and rewired through evolutionary time.

By using comparative genomics, I found that the main fungal regulator (SBF) was acquired very early in the evolution of fungi by horizontal gene transfer from a viral origin. I also showed that this viral-derived transcription factor still coexists with the ancestral E2F in the zooporic fungus Spizellomyces punctatus, forming a hybrid cell cycle control network. I hypothesize a viral-derived regulator (SBF) hijacked cell cycle control in the dawn of Fungi by binding the promoters regulated by the ancestral counterpart (E2F), pushing cells to proliferation. This requires the invading SBF to be able to bind regulatory regions controlled by E2F. Using a high-throughput analyses of the DNA-binding properties of the SBF and E2F-family across Eukaryotic lineages I found that E2F and SBF share binding preferences, but that these are not completely overlapping, which could permit the evolutionary conservation of the hybrid E2F/SBF network in Spizellomyces. I then proceeded to test the potential differences \textit{in vivo} in accessibility to E2F and SBF binding sites by coupling in vitro DNA-binding information with nucleosomal and TF-footprints generated from MNase-seq data.

Finally, I developed Agrobacterium-mediated transformation in Spizellomyces, allowing me to describe basic characteristics of its developmental program using live-cell and fluorescence microscopy. By following nuclear dynamics with a fluorescently tagged histone I found that mitosis only initiates after germination, and that nuclei divide synchronously during sporogenesis. Furthermore, by following actin dynamics with LifeAct I showed that zoospores use actin-filled pseudopods to crawl, much like amoeba or animal cells, and that sporangia rely on complex actin dynamics during the formation of zoospores that are reminiscent of animal cellularization processes. This work highlights the importance of non-model systems for finding new solutions to longstanding questions in biology. This is a first step towards establishing Spizellomyces as a model system to study the evolution of key animal and fungal traits, particularly cell cycle regulation and development.

Adapting to environmental changes (i.e. an increase in osmolarity) is critical for cell survival. How cells respond and adapt to osmotic stress has been well-studied in the model eukaryote Saccharomyces cerevisiae. Although the molecular and systems properties of osmo-adaptation have been well studied, few studies have focused on the effects of genetic variation. Understanding how genetic variation affects molecular pathways and their dynamics, which translates to variation in cellular and organismal phenotypes, is a key step towards understanding important phenomena such as complex gene by gene interactions and the mapping of genotype to phenotype. The challenge is to causally connect genetic differences with cellular function and differences in complex traits between individuals. As a first step towards addressing this challenge, my dissertation research investigates how natural genetic variation affects osmo-adaptation dynamics in budding yeast, Saccharomyces cerevisiae.

First, I characterized the natural variation in osmo-adaptation dynamics across S. cerevisiae. I showed that individual strains were highly variable in adaptation time and relative maximum growth rate after adapting to stress. Analysis of a broad set of genes involved in osmo-adaptation did not reveal any obvious genetic differences that could account for the observed variation. To identify alleles associated with the variation, I measured osmo-adaptation dynamics in progeny generated from a cross between two closely related lab strains. Identified alleles were outside the core signaling pathway and affected both adaptation time and relative maximum growth rate. Finally, I built a novel mapping panel and measured osmo-adaptation dynamics to obtain a more global, species-wide view. The panel showed an increased amount of variation in osmo-adaptation dynamics and a subset of progeny were phenotypically more extreme than their parents. Mapping the variation in this panel will generate a comprehensive list of alleles that affect osmo-adaptation. The strains in the mapping panel have a low number of mutations predicted to have strong effects in HOG pathway genes. Given our earlier results from the pairwise cross, I expect that many osmo-adaptation alleles discovered from the mapping panel will be outside the HOG pathway.