The molecular and cellular function of TAF8 in the embryonic brain

© 2016 Dr. Farrah El-SaafinTranscription of all class II genes (all protein coding genes and most ncRNA genes) requires the ordered assembly of a preinitiation complex (PIC). The first step in PIC assembly is the binding of the general transcription factor TFIID, to the promoter of the gene, which is followed by the binding of the other general transcription factors, then RNA pol II. TAF8 is part of at least a subset of TFIID complexes (Guermah et al., 2003). It is essential for the in vitro differentiation of adipocytes, but not myocytes (Guermah et al., 2003). Moreover, it is needed for the survival of the inner cell mass, but not the trophectoderm cells of the blastocyst (Voss et al., 2000). Cell-free studies suggest that TAF8 dramatically affects the overall structure of TFIID (Bieniossek et al., 2013). Embryos lacking Taf8 die at the blastocysts stage of embryonic development (Voss et al., 2000), which precluded analysis of the role of TAF8 at later time points in vivo. It was of interest to know the role of TAF8 at later time points in embryonic development, and I wanted to determine if loss of Taf8 resulted in selective apoptosis in cell types other than the inner cell mass. Prior to this study the molecular role of TAF8 in the structural stability of TFIID and the function of TFIID was highly contentious, and so I wanted to examine the requirements of TAF8 for TFIID function. Therefore the aims of my thesis were to determine the cellular function of TAF8 by characterising the consequences of lack of TAF8 in the developing mouse central nervous system using tissue-specific deletion of the Taf8 gene (Taf8–/loxNesCreT/+), and in doing so, investigate a potential role for TAF8 in the regulation of apoptosis. Another aim was to determine if loss of TAF8 affects the structural stability of TFIID, and general transcription in vivo. The major findings of my thesis were that loss of Taf8 in the embryonic central nervous system resulted in massive apoptosis of the E14.5 cerebral cortex, however brain areas other than the cortex were relatively unaffected. The apoptosis in the cerebral cortex of Taf8–/loxNesCreT/+ embryos was caused by a massive 23-fold elevation in p53 protein, an activation of pro-apoptotic p53 target gene Puma, and subsequent degradation of the pro-survival protein MCL1. Deletion of either p53 or Puma resulted in a complete or partial rescue of the apoptotic Taf8–/loxNesCreT/+ phenotype, respectively. However, the complete rescue of the apoptotic phenotype in the Taf8–/loxNesCreT/+p53-/- brains revealed another TAF8 phenotype, namely a failure of neuronal differentiation. I found that loss of TAF8 in the Taf8–/loxNesCreT/+p53-/- brain did not result in destabilization of the TFIID complex. Nevertheless, transcription was extensively effected in the absence of TAF8, albeit in a very surprising manner. Astonishingly, rather than observing a general failure or reduction in transcription, as one might expect when an essential member of TFIID is missing, a large number of genes were substantially upregulated in the absence of Taf8. Apart from upregulation of bona fide protein coding and ncRNA genes, 5' and 3' regions of genes, intergenic regions and predicted genes were up-regulated in the absence of Taf8, indicating that TAF8 normally represses transcription at least at these regions in the genome or plays a role in determining the boundaries of transcription. Finally, a potential role for TAF8 in the regulation of pre-mRNA splicing was uncovered, as a large number of genes were alternatively spliced, and the “sensor of defective splicing”, Mdm4-short, was activated in the absence of Taf8. In conclusion, I discovered that rather than having an essential role the activation of transcription, TAF8 appears to restrict transcription and facilitates splicing, suggesting that the aberrant transcription and splicing causes apoptotic cell death particularly in the cerebral cortex