PDF, English Restricted access: Repository staff only until 6 May 2025. Login+Download (6MB) | Terms of use

Abstract

As byproducts of aerobic respiration, reactive oxygen species (ROS) play dual roles in many important biological processes by acting as both a potential source of damage as well as signal molecules in multiple signaling networks. The mechanisms whereby ROS direct transcriptional regulation have been documented, however, how evolution has shaped these mechanisms in diverse species which may experience very different levels of ROS is poorly understood. To explore this issue, I conducted a comparative study investigating the transcriptional response to ROS in cell lines derived from a range of vertebrate models, including zebrafish, mouse, turtle and frog. Additionally, I also examined a unique fish model, the Somalian cavefish Phreatichthys andruzzii, which has evolved in a completely dark environment for over 3 million years and where many light-dependent functions such as photoreactivation DNA repair and photoentrainment of the circadian clock have been lost. Using RNA-seq, I revealed diverse transcriptional profiles under H2O2 exposure among zebrafish, cavefish and mouse cell lines, notably involving the group of DNA repair genes. I then focused on the key nucleotide excision repair gene xpc which is transcriptionally induced by H2O2 in all the selected vertebrate cell lines with the exception of the cavefish. In the case of the zebrafish xpc gene, I identified a ROS inducible, D-box enhancer-driven expression mechanism. In contrast, in mouse cells no D-box was encountered in the xpc promoter and furthermore, transcription regulated by the D-box enhancer does not respond to ROS exposure, indicating a distinct regulatory mechanism targeting the mouse xpc promoter. Meanwhile, conserved D-box elements were identified in the cavefish xpc gene which still mediated ROS-induced transcription in zebrafish cells. However, ROS induced, D-box driven transcription was absent in the cavefish cells. I then demonstrated that the bZIP PAR factor TEF1 serves as a principal activator of D-box-mediated transcription in response to ROS. I identified several putative MAP Kinase phosphorylation sites in the N-terminal portion of the zebrafish TEF1 protein. 8 Mutations of these MAPK phosphorylation sites to match the corresponding sequences observed in the cavefish TEF1 protein led to much weaker activation and ROS sensitivity, which is consistent with the significant loss of ROS induced xpc gene expression in this cavefish species. Furthermore, I revealed that changes in intracellular ROS levels also affect the nuclear translocation of TEF1 proteins in zebrafish but not in cavefish cells. Differences in the MAPK-regulated nuclear localization of TEF1 may at least in part explain the species-specific differences in how the D-box responds to ROS. I have therefore revealed that vertebrates have evolved diverse ROS-induced DNA repair gene transcription mechanisms to adapt to varying aerobic environments. In addition, it is tempting to speculate that the loss of D-box mediated expression and differential PAR factor regulation observed in both the cavefish P. andruzzii and mouse may point to these species having shared similar selective pressures during their evolution. In summary, my results highlight plasticity in the ROS-induced transcriptional response over the course of vertebrate evolution