The role of an A-form DNA element in transcriptional regulation of the Xenopus gata2 gene

• Main Author: Llewellyn, Katrina
• Other Authors: Scarlett, Garry ; Guille, Matthew John
• Published: University of Portsmouth 2008
• Subjects: 572.865
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.494015

The development of a single celled embryo to a fully developed organism is tightly controlled by gene expression. Different genes are temporally and spatially regulated to form different cell types. Genes regulate each other to produce products that give rise to distinct cell fates and these interactions are referred to as gene regulatory networks (GRNs). Haematopoiesis is modulated by a complex GRN and one critical factor in this developmental pathway is GA TA2. Gata2 is a transcription factor and when it is inactivated in mice no erythrocytes develop. Xenopus laevis gata2 expression is itself regulated by another transcription factor, CCAA T box transcription factor (CBTF). CBTF is a multi-protein complex and one subunit in particular has been shown to be critical for gala2 activation, ilf3. Xenopus laevis ilf3 contains two double stranded RNA-binding domains (dsRBDs) that are necessary for its DNA binding and transcriptional activity. Binding to DNA is unusual for dsRBDs. It has been hypothesised that the ilf3 dsRBDs can bind the gata2 promoter and activate transcription only because binding is mediated by the partial A-form structure of the gata2 promoter, which mimics dsRNA. This hypothesis was tested by altering the degree of A-form structure in the promoter, then measuring binding of CBTF and transcriptional activation of gata2. Data presented in this thesis show that A-form structure is important for binding of CBTF and transcriptional activation at the gata2 promoter; however the presence of the inverted CCAA T box is an absolute requirement. In the absence of the inverted CCAA T box a strongly A-form sequence is not sufficient for activity. Secondly, the promoter region used in these experiments was shown to give correct temporal expression but not spatial. Suggesting a region(s) elsewhere directs correct spatial expression. Finally, the residues of ilf3 critical for dsRNA binding were examined to see if they were also critical for DNA binding. Those residues involved in binding dsRNA are almost all involved in binding DNA, although one mutant (K455A) reduced dsRNA binding but had no effect on DNA binding.