Summary: | <p>CREB (cAMP response element binding-protein) is a disordered essential mammalian transcription factor, activated by the cAMP signalling cascade. It is involved in a myriad of biological processes including memory and the circadian clock. Moreover, CREB is overexpressed in many cancers, such as leukaemia, sarcomas, melanoma. Thus, to understand both mammalian and cancer biology, it is important to characterise it well.</p>
<p>The C-terminal end of CREB contains a bZIP DNA-binding domain (DBD), and the N-terminal end of CREB contains three domains - two glutamine-rich domains surround its acidic activation domain KID (kinase inducible domain). Previous biophysical (and structural) studies have focused on understanding CREB’s behaviour using isolated KID and bZIP domains, but few have used full-length CREB or looked at the influence of the other domains on its biophysical properties. Recent work has challenged simplistic modular architectural understandings of transcription factors, for example: there is increasing evidence that these disordered sequences are essential for efficient target search. This Thesis aims to directly address this issue through comparative studies with full-length CREB and its isolated domains.</p>
<p>In this Thesis I developed a high yield methodology for production of highly pure full-length CREB protein and used the material to conduct thorough biophysical characterisation of CREB and its binding to DNA in-vitro. I found CREB’s non-DBDs lower DNA affinity around 5-fold compared to the isolated bZIP, but the homodimerization and target DNA affinity remain in the nanomolar range whereas affinity to non-target is in the micromolar range. Altogether, this makes the full-length CREB target search more efficient and less likely to be sequestered away by non-target DNA in a cellular context. Moreover, each of CREB’s non-bZIP domains impedes binding to DNA to a similar extent. Finally, current literature suggests bZIP transcription factors dimerise following target binding through a monomer search pathway. However, I find a high CREB homodimer stability which, together with unquantifiable binding of CREB monomers to DNA, favours a dimeric pathway instead.</p>
|