UC Berkeley

X-ray radiation is a damaging form of energy with the potential to disrupt virtually all cellular processes by inflicting broad, indiscriminate damage to the macromolecules that drive them. Despite its relatively non-specific effects, it has long been observed that different cell types, tissues, and organisms have differing responses to X-rays. Mounting evidence suggests that the source of this variability is diverse, and our understanding of its causes remains incomplete.

Some characteristics are generally predictive of the outcome of tissues exposed to X-rays, such as the state of proliferation, oxygenation, and cell cycle status of the cells that comprise them. This knowledge informs the treatment of cancers with radiotherapy which, for example, is thought to disproportionately kill cancers due in part to their abnormal proliferative state. However, these characteristics are inadequate to fully account for the variability of X-ray responses observed in different tissue targets, an observation made yet more complex by the fact that tissues frequently consist of heterogeneous cell populations which themselves may have variable X-ray responses.

Cumulative research, particularly in the field of cancer biology, has revealed that this variability is often associated with differences in the transcriptomes of irradiated cells. For example, some cancers that are resistant to killing by radiotherapy have been shown to have elevated expression of damage ameliorating genes involved in processes such as DNA damage repair. Moreover, tumors consisting of heterogeneous cell populations may grow resistant to irradiation during fractionated therapy through the unintended selection of cells with bolstered expression of radiation protective genes. Collectively, much work has been conducted on X-ray induced changes in gene expression in several systems ranging from human tumors to fruit fly embryos. However, the limitations of the sequencing methodologies employed in these studies prevented deep analysis of intratissue differences in X-ray response and the transcriptional states associated with them. The rise of single-cell RNA sequencing (scRNA-seq) technologies has made it possible to perform these analyses by allowing for the single-cell origin of transcripts to be retained during sequencing. In this work, we use the wing imaginal disc, a simple epithelial tissue found within the larvae of Drosophila melanogaster, to study the effects of X-rays on the transcriptomes of different cells within the same tissue.

In Chapter 1, I provide a summary of conserved molecular pathways that are integral to X-ray response, centered around DNA damage, and provide an overview of the factors that influence the sensitivity of cells to irradiation. I then place the work presented here within the context of past studies on the transcriptomic changes that occur after X-ray exposure in Drosophila and more recent scRNA-seq studies in mammalian tissues. Finally, I summarize the technology of scRNA-seq and its relevance to this work.

In Chapter 2, I present our primary findings. We use scRNA-seq to describe the major transcriptional changes induced by X-ray irradiation in different regions of the tissue and uncover two categories of transcriptional states present in the wing disc that are associated with variable X-ray induced gene expression. First, we find that transcriptional states associated with cell location in the proximodistal axis of the tissue are associated with the induction of different genes. Second, we find that transcriptional states defined by cell cycle genes are associated with varying levels of X-ray induced gene expression, with those cells likely invoking a G2/M cell-cycle arrest having the largest average X-ray gene induction. We also adapt the Herfindahl-Hirschman Index, a measure used in the field of economics to assess market concentration, to rank genes in terms of their expression homogeneity. Using this method, we find that genes involved in core damage response pathways are relatively homogeneous across cell clusters when compared to certain ligands and transcription factors.