Gene expression is primarily regulated by various genomic elements, namely enhancers. Enhancers are stretches of non-coding DNA that influence gene expression across space and time. Much of what we understand about enhancers comes from the study of spatially patterned gene expression and transcription factor localization. We know that enhancers contain clusters of transcription factor binding sites, but these sites alone do not define the identity of an enhancer and the question of how enhancer identity is specified remains unanswered. The work that I describe in this dissertation aims to explore other factors associated with enhancer activity in the early Drosophila melanogaster embryo and to improve existing technologies for the study of gene expression. In Chapter 1, I describe the many factors associated with enhancer activity as well as a new technology that may allow us to finally answer questions surrounding enhancer identity. Transcriptional co-factors, such as histone modifiers or insulator proteins, either directly or indirectly influence transcription factor binding and enhancer activity, which can drastically affect patterned gene expression. In Chapter 2, I screened for transcriptional co-factors with enhancer-specific defects by characterizing the effects of reduced transcriptional co-factor expression on enhancer activity. Given the importance of enhancer activity in regulating patterned gene expression, I assayed the expression of a classic patterned gene as a proxy for enhancer activity and the expression of a non-patterned gene as a transcriptional control. I found several candidates, all histone modifiers, that specifically affect expression of a patterned gene while the non-patterned gene expression remained unchanged. Ultimately, I found a wide and overlapping variability in the extent of knockdown of these co-factors and I pursued other methods as a result. Nonetheless these data suggest that a system outside of the canonical view of transcription factor binding exists to define enhancer activity and identity.
In early development, enhancer activity is primarily driven by maternally-deposited RNAs and proteins prior to zygotic genome activation. One of the largest hurdles in understanding the regulation of zygotic gene expression at this early point in development lies in our ability to distinguish between maternal and zygotic RNAs. Throughout the course of my PhD, single-cell RNA-sequencing became an increasingly popular way to study gene expression on a smaller scale than ever before. With that being said, zygotic genome activation occurs concurrently with cellularization in the early Drosophila melanogaster embryo, thus isolation of cells is not possible. In Chapter 3, I describe a set of experiments and analyses demonstrating the use of single-nucleus RNA-sequencing in the early Drosophila melanogaster embryo using an insulator protein as a case study. Under the assumption that the vast majority of RNAs present in the nucleus are zygotic transcripts, this allows us to assay zygotic gene expression outside of the context of cytoplasmic and maternal RNAs to really examine defects in transcription. I found that the nuclei retain spatial information, or where the nuclei were located in the embryo prior to dissociation. I also found examples of patterned genes that are differentially expressed upon the loss of the insulator protein in individual clusters, but not in bulk. These results highlight the importance of establishing the use of single-nucleus RNA-sequencing. From this, we now have the capacity to understand global changes in spatial gene expression across the embryo, giving us the ability to answer questions regarding the regulation of patterned gene expression.
As discussed above, single-nucleus RNA-sequencing is a powerful tool; however, current technologies are limited either by the number of nuclei captured or the accessibility of the technique in the first place. In an effort to improve barcoding of individual cells or nuclei, I began a collaboration with a group of researchers. In Chapter 4, I discuss testing a set of catalytic DNA hairpin oligos as a means of barcoding cDNA in individual cells or nuclei. I demonstrated that these hairpins are highly efficient and specific in barcoding single-stranded DNA molecules in vitro. I show additional work towards optimizing this reaction following reverse transcription; however, a completed method is outside of the scope of this dissertation and this work is ongoing.
Altogether, the work that I have completed throughout this dissertation has progressed our understanding of the regulation of gene expression by working outside of the canonical view of transcription, bringing new technologies to the table, and potentially drive the field forward with the development of new methods to improve existing technologies.