Abstract
The human cortex consists of upwards of 80-90 billion neurons. The vast majority of these neurons are born during early brain development during a short time frame in which progenitors rapidly proliferate, differentiate, and migrate to the appropriate cortical layer. Mouse models of cortical development have revealed that DNA damage occurs at high rates during this time period, and DNA damage repair is essential for normal development. One of the required repair pathways is non-homologous end joining (NHEJ) which repairs DNA double-strand breaks (DSBs) throughout the cell cycle. NHEJ deficient mice are embryonically lethal or develop small brains; however, concurrent p53 knockout rescues this phenotype, indicating that p53 kills neural progenitor cells (NPCs) harboring DNA DSBs during early cortical development. It was previously unclear how genomic stress activated p53 during neurogenesis and how p53 altered the fate of neural cells in human cortical development. With the discovery of the Yamanaka factors, human-induced pluripotent stem cells (hiPSCs) were quickly generated, providing an in vitro model of human neurodevelopment to study models of normal and pathological cortical development. Combined with single-cell approaches, including single-cell RNA-seq and imaging cytometry, the temporal dynamics of DNA DSBs and cell fate during neuronal differentiation can be quantified and better understood. The goal of this study was to determine the role of p53 in response to genomic stress in hiPSC-derived NPCs and their progeny.
NPCs are a heterogeneous cell population and require single-cell approaches to best examine DNA DSB dynamics during development. In Chapter II we developed a technique to assess genomic stress through the quantification of DNA DSB regions using Imaging Flow Cytometry. We characterized DNA DSBs, cell cycle stage, and cell fate (death, proliferation, differentiation) during hiPSC-derived neurogenesis and observed fluctuating levels of DNA DSBs. NPCs and neurons contained comparable percentages of cells with breaks, but cell death was increased in the early differentiating NPCs compared to the progenitors. This finding implicates a different response to transcription and replication stress in neural cells.
In Chapter III we examined the differences between transcription and replication stress in neural cells in more depth and discovered that p53 is activated in transcription but not replication stress. Furthermore, the activation of p53 during transcriptionally induced DNA DSBs often led to cell death. During hiPSC-derived neurogenesis, p53 eliminated NPCs with DNA DSBs by preventing them from proliferating and increasing cell death. Previous studies of DNA DSBs in NPCs have shown an enrichment in breaks in long neural genes due to a proposed collision between transcription and replication machinery. Using DNA DSB mapping, we located DNA DSBs in highly transcribed genes and observed no relationship in breaks between p53 and gene length. However, we observed DNA DSB enrichment at the transcription start sites of p53 deficient cells in high risk alleles for neurodevelopmental disorders.
We conducted additional experiments to examine how other isogenic cell types respond to genomic stress and assessed whether genomic stress led to neurons with neuroinflammatory phenotypes (Chapter IV, Appendix). hiPSC-derived astrocytes were more resistant to genomic stress; they exhibited fewer DNA DSBs, lower percentages of cells in cell cycle arrest, and less p53 activation compared to isogenic NPCs. Furthermore, genomic stress did not lead to neuroinflammatory phenotypes, but rather a deficiency in p53 increased levels of secreted VEGF and Fractalkine. We also observed different secretion patterns in astrocytes versus neurons and concluded that these differences in cytokine/chemokine secretion may be instrumental in future studies investigating the functional role of neuroinflammation during early brain development.
Overall, this work furthers our understanding of the role of p53 in human neurodevelopment by providing context for the parameters of activation of p53 in neural cells. p53 was previously known to induce cell death and cell cycle arrest in mammalian cells with DNA DSBs, and from the work presented here we show that p53 is activated in response to transcription and not replication stress in NPCs.
