UC Berkeley

Due to their apparent passive behavior, plants have long been considered something in between inanimate objects and ”proper” living things, that is, animals. When an animal is incapable of moving or communicating we refer to it as being in a vegetative state. Research over the last three decades has shown that, when it comes to molecules, plants cells are bustling with activity. It was our task to just go and take a look.These dynamic molecular events operating at the cellular level include a myriad of signaling pathways that can sense environmental inputs such as light intensity and quality, temperature, osmotic pressure, the axis of gravity, mineral nutrients and several hormones to name but a few. The output of these signaling pathways often involves changes in the mRNA abundance of responsive genes. Plant growth and development depends on quantitative aspects of this regulation -when, where and how much of each mRNA species each cell makes in response to a specific stimulus. In turn, human societies ultimately depend on plants, which makes these problems not only fascinating from a basic science point of view, they are also pressing from a very practical perspective. In large part, we owe our current understanding of how this information transfer occurs to studies in other eukaryotic model organisms, in particular the fruit fly Drosophila melanogaster (Drosophila). From the first experimental description of morphogens to the recognition of transcriptional enhancers, foundational discoveries related to gene regulation in higher eukaryotes have been made using the fruit fly embryo as a model system. The fact that this inter-kingdom comparisons are possible and useful highlights one of the most important lessons of the molecular biology revolution, the notion that all life forms share fundamental principles because they are built largely of the same type of molecules. Jacques Monod captured this view when he famously claimed that ’what is true for E. coli is true for the Elephant’. While Monod’s quote nicely justifies the logic behind a specific research program, it also challenges us to question whether paradigms drawn in one model organism bias and limit our understanding when applied uncritically to other biological systems. Monod certainly was not expecting us to believe that the specific details of how bacterial genes are turned on and off in response to sugar sources are conserved all the way from bacteria to animals. However, the physical principles that nature exploits to bring about a bacterium and an elephant are ultimately the same. The genomic era has nothing but confirmed the underlying unity across all life kingdoms by revealing that the way these principles are encoded in DNA derives from one common ancestor. One of the challenges we face as biologists is finding what these shared principles are. In this effort, by distinguishing what is shared from what is different, drawing from different experimental systems can be illuminating. There are more practical reasons why a trans-kingdom perspective can be beneficial. Engaging at a deep level with the technical details of an experimental method developed in one model organism allows one to more easily transfer that knowledge to other contexts. A substantial part of this dissertation builds on this premise. It should be possible to express these biophysical principles as models that can be applied across biological contexts. Since the dawn of molecular biology, transcriptional regulation has had a central role as a test bed to combine theory and experiments to un- cover generalizable biological principles. A good example of this is the Monod-Wyman-Changeux (MWC) model of allostery, which captures such seemingly different behaviors as sugar regulation of gene expression in bacteria, oxygen transport in the blood by hemoglobin and the effect of drugs on cell surface receptors. During the summer of 2016 I had the privilege of attending the Physiology Course at the Marine Biological Laboratory where, under the guidance of Professor Rob Phillips, I used statistical mechanics to derive and test the MWC model in the context of sugar sensing in E. coli. This type of simple and intuitive mechanistic model is of a fundamentally different nature to the statistical regression-based models that are becoming increasingly popular in biology. Because they are conceived as arbitrarily complex black box ’fits’ these models can certainly ’recapitulate the data’, but, for that very same reason, cannot be used to test molecular or biophysical hypotheses. We should, however, recognize the relative failure of using theory and experiments to develop predictive models of transcription in eukaryotes. Compared to the resounding successes in phage λ and the E. coli Lac operon, we are far from achieving this type of understanding in plants or in any other higher eukaryote. Recent calls to advance a unified model of transcription in eukaryotes advocate for making an inventory of moving parts and building tools to measure their dynamics in single cells. Many of the results presented here align with the goal of closing technical gaps to make it possible to measure the phenomena under study. It may well be that we have reached a technical wall beyond which the dialog between theory and experiments will come naturally. For example, there is mounting evidence that the formation of biomolecular condensates of RNA Polymerase and transcription factors is a fundamental property of transcriptional regulation. The theoretical frame- work for these processes is an active area of research and it is thus still unclear what experiments could be designed to falsify these models . Theoretical explorations, even if frustrating, can sharpen our thinking of what experimental tools we are still missing. For example, as I argue in the following chapters, there is a still a need for live cell technologies to measure post-translational modifications of histones and other proteins at a specific locus. Beyond this quest for building unifying models based on physical principles, we should not forget that transcriptional regulation lies at the core of a myriad of diverse biological phenomena that we have only scratched the surface of. Simply taking a quantitative, single-cell view of gene regulation can offer unprecedented insights about the way a particular biological process works and make us revisit the way we think about it. As I have argued here, toggling between model systems offers several advantages, how- ever, it is does not come without risk. The practice of citing faulty research in animals models to buttress claims about plants goes back to the very origins of plant biology. Theophrastus, considered the father of botany, claimed that spontaneous generation in plants should not be taken with suspicion since it had already been demonstrated in animals: If some [plants] are generated in both ways, spontaneously as well as by seed, there is no absurdity: so some animals similarly come from two sources, both from other animals and from earth. It would be presumptuous for me to pretend that I have avoided the same mistake, but that did not stop me from trying. Chapter 1 consists on a comparative overview of novel insights about transcriptional regulation in Arabidopsis thaliana (Arabidopsis) and Drosophila acquired in the course of my research. In it, I try to find unifying themes and put results in the context of the field at large. Chapter 1 is also intended to give a relatively detailed overview of this dissertation and so many of its figures are found in chapters that deal with specific topics. The rest of the chapters are as follows. In chapter 2, I describe the motivations to establish optogenetic tools to manipulate transcription factors in Drosophila embryos and the results from our efforts. Chapter 3 contains a manuscript that is currently in preparation that describes the development of a minimal synthetic transcriptional system in Drosophila and the insights we have gained from it. Chapter 4 contains a manuscript currently in bioRxiv and under peer review that describes how transcription operates at the single cell level in plants. Chapter 5 builds upon the live imaging tools presented in Chapter 4 and describes experiments to measure the dynamics of light signaling during a dark to light transition in Arabidopsis seedlings. Finally, in Chapter 6, I outline experiments that I was unable to finish that point to potential future directions.