YCR annual science meeting

Tuesday 04 September 2007
Clarendon Lecture Theatre
and
Medical School Foyer
University of Leeds

Open Papers 2

L302 Role of the POZ Domain in Transcriptional Regulation by Miz-1
Mark A Stead, Gareth O Rosbrook, Thomas A Edwards and Stephanie C Wright
Institute of Molecular and Cellular Biology, University of Leeds, Leeds LS2 9JT

The BTB/POZ (poxvirus and zinc finger/bric à brac, tramtrack and broad complex) domain mediates protein-protein interactions in diverse biological processes, and is found in approximately 10% of zinc finger transcription factors. Most POZ-ZF factors play roles in cell proliferation and development, and many have also been implicated as oncogenes or tumour suppressors in human cancer. Targeting POZ domain interaction interfaces is a promising therapeutic strategy in cancer.

Miz-1 is a POZ-ZF factor that activates genes involved in cell cycle arrest, differentiation and DNA damage responses. The transcriptional properties of Miz-1 are modulated by binding of the c-myc proto-oncogene product to residues near its zinc-finger domain, leading to aberrant gene regulation in tumours. The N-terminal POZ domain of Miz-1 directs both self-association and the recruitment of non-POZ partners. Miz-1 is regulated during DNA damage responses by the recruitment of (topoisomeraseII binding protein), and its interaction with the ubiquitin ligase, HECTH9 (homologous to E6 AP carboxy terminus), modulates the ubiquitination and therefore transcriptional properties of c Myc. POZ-ZF factors are also regulated by hetero-oligomerisation, and the interaction of Miz-1 with BCL6 is associated with aberrant suppression of Miz-1 target genes in B cell lymphoma.

As an initial step towards the design of therapeutics that target interactions of the Miz-1 POZ domain, we are characterising interaction interfaces of Miz-1 with both POZ- and non-POZ partners.

Epigenetic priming of the GM-CSF locus in myeloid cells and in AML by GATA-1 and GATA-2
Fernando Calero-Nieto, Euan Baxter, Brett Johnson, Andrew Bert, and Peter Cockerill.
Experimental Haematology, Leeds Institute of Molecular Medicine, University of Leeds, Wellcome Trust Brenner Building, St.James's University Hospital, Leeds LS9 7TF.

We have defined the mechanisms that control the inducible tissue-specific regulation of the human GM-CSF locus. GM-CSF promotes the growth, differentiation and survival of myeloid precursors and their progeny within the granulocyte-macrophage (GM) lineage and also functions as an autocrine growth factor in AML. GM-CSF is expressed in a highly inducible fashion in a wide variety of cell types that include myeloid progenitors, differentiated myeloid cells such as mast cells and macrophages, and T cells. We have previously demonstrated that the GM-CSF gene is controlled by a highly inducible upstream enhancer. In T cells we found that the enhancer core comprises 2 essential composite NFAT/AP-1 elements which reside on 2 positioned nucleosomes (N1 and N2). When the enhancer is activated, NFAT directs the disruption of these 2 nucleosomes, resulting in the creation of an inducible DNase I hypersensitive site across this region.

We have now found that this same enhancer is activated by quite distinct mechanisms in myeloid cells that express GATA-1 or GATA-2. In these cells, one of the enhancer core AP-1 elements cooperates with upstream GATA elements linked to an additional AP-1 element to activate very high levels of inducible enhancer activity and GM-CSF gene expression. Furthermore, GATA-1 and GATA-2 can function to prime the GM-CSF locus in mast cells that express GATA-2 and in pro-erythroid leukaemic cell lines that express GATA-1. In these cells, GATA factors act to constitutively remodel the nucleosome architecture prior to activation, leading to the exclusion of nucleosome N0. As a consequence, nucleosomes occupy different fixed positions in T cells compared to myeloid cells. The region spanning the upstream GATA and AP-1 sites exists as a nucleosome-free region in non-activated myeloid cells, that can be detected as a weak DH site. Upon activation, the DH site intensifies in response to recruitment of AP-1 and further disruption of nucleosome architecture.

Significantly, the enhancer exists as a strong constitutive DH site in a high proportion of AMLs, suggesting that the locus is partially activated and primed for autocrine or paracrine expression of GM-CSF. Studies are in progress to define the mechanisms of locus priming in AML.

Repressosome formation and disruption regulates the Kaposi's sarcoma associated herpesvirus latent-lytic switch
Faye Gould and Adrian Whitehouse
Institute of Molecular and Cellular Biology, University of Leeds LS2 9JT.

Kaposi's sarcoma-associated herpesvirus (KSHV) is the most recently identified human tumour virus and the etiological agent of Kaposi's sarcoma (KS). Like all herpesviruses it has two distinct phases of infection; latent persistence and lytic replication. Lytic reactivation is critical for KS pathogenesis and spread of infection. KSHV replication and transcription activator (RTA) is the chief regulator of the latent-lytic switch. RTA interacts with various cellular transcription factors to activate viral lytic gene promoters, including its own. We have identified an interaction between RTA and Hey1, a cellular transcriptional repressor, and shown that it represses the Rta promoter by recruiting a repressosome complex, which helps maintain KSHV latency. This RTA-Hey1 interaction is particularly intriguing due to their opposing functions. However, RTA was recently shown to possess an E3 ubiquitin ligase activity toward IRF7, resulting in polyubiquitination and subsequent proteasomal degradation of IRF7. We have since demonstrated that RTA also targets Hey1 for proteasomal degradation via a mechanism dependent on its ubiquitin ligase activity. Degradation of Hey1 causes disruption of the repressosome, allowing upregulation of RTA expression and lytic reactivation.

Investigating the potential role of RecQ5 in the reinitiation of stalled replication forks
Rachel Blundred, Thomas Helleday, Alistair Goldman and Helen Bryant
Institute of Cancer Studies, University of Sheffield.

The DNA in our cells is constantly damaged. Multiple efficient repair systems have evolved to combat this damage and maintain the fidelity of the genome. It is becoming increasing apparent that despite the efforts of the repair systems some damage persists. Persistent damage is particularly significant in replicating cells. Here, if a replication fork collides with DNA damage or is perturbed in some other way the replication fork can stall or collapse. Damage must be repaired or bypassed at these replication forks in order that cell division can continue. In model organisms, one process which facilitates blockage resolution is regression of the replication fork away from the point of damage so that repair proteins can gain access to the DNA. Alternatively, regression can also facilitate template switching, which is a process that allows the accurate bypass of damage without the need for DNA lesion repair at the replication fork. The proteins required for this process in mammalian cells are not known. In E.coli regression is carried out by the helicase RecG. In vitro, the human protein RecQ5 has been shown to bind to similar forked DNA substrates as RecG and to possess the helicase and strand annealing activities proposed to be required for the regression of stalled forks. This project aims to discover if the protein RecQ5 plays a role in the regression of replication forks in mammalian cells. We have found that over-expressing RecQ5 in a wild type background helps to protect cells against thymidine treatment and we have seen that RecQ5 foci form in the presence of thymidine. Therefore we suspect that RecQ5 is regressing replication forks in vivo.