Award details

HMGB1: structural studies inter- and intra-molecular interactions and role in transcription factor binding and chromatin remodelling

Reference	BB/D002257/1
Principal Investigator / Supervisor	Professor Dame Jean Olwen Thomas
Co-Investigators / Co-Supervisors	Dr Katherine Stott
Institution	University of Cambridge
Department	Biochemistry
Funding type	Research
Value (£)	359,702
Status	Completed
Type	Research Grant
Start date	01/12/2005
End date	28/02/2010
Duration	51 months

Abstract

(1) To investigate how the long (30 residue) acidic tail of HMGB1 lowers the DNA binding activity of the HMG boxes, probably by interaction with one or both of the boxes, we will construct a model for the interaction of the tail with the body of the protein by making a series of constructs in which the tail is shortened in steps of 5 residues. 15N-HSQC experiments will be recorded to look for chemical shift changes corresponding to each deletion, thereby enabling specific tail sections to be mapped on to the body of the protein. An independent validation will be carried out using site-specific intramolecular cross-linking and/or cleavage reagents. We will take advantage of HMGB2a and 2b (shorter tails, approx. 20 residues) and the hybrids HMGB1/tail 2a or 2b, to gain further insights into tail length and composition (Glu:Asp ratio). We will also test for competition between DNA and the acidic tail for the HMG boxes. (2) Information about how the tandem HMG boxes of HMGB1 bind to DNA is lacking. To overcome the lack of sequence specificity of HMGB1, which precludes the formation of defined complexes of the AB didomain with DNA for NMR, we generated the hybrid SRY.B didomain in which the sequence-specific HMG box from SRY has replaced the A box of AB, and have very recently determined its structure. Although the binding of the individual boxes to the DNA is well defined, the structures lack long-range order. We now propose to use measurement of residual dipolar couplings to determine the overall bend angle for the DNA more precisely. Given the success with SRY.B, we will aim to determine also the structure of an A.SRY/DNA complex, to obtain information about the binding of the A box. This should lead to a model of the AB domain bound to DNA. (3) We have previously identified an interaction between the A box of HMGB1 and the N-terminal domain of p53 that seems likely to be involved in recruitment of HMGB1 to the p53 DNA binding site and facilitation of p53 binding. Wepropose now to characterise this interaction in detail by NMR to gain information about the protein-protein interface using chemical shift mapping. Unlabelled p53(1-93) will be titrated into a solution of 15N-labelled A domain and complex formation monitored by 15N-HSQC experiments (and vice versa for the labelling). We will aim to identify the minimal interacting region of p53 and if possible to determine the structure of a complex. Similar experiments will be carried out with full length HMGB1 which interacts more strongly with p53(1-93), perhaps through additional interactions. (4) Promoters with variant octamer sequences will be tested to investigate further the interactions involved in facilitation of the binding of Oct-1 to DNA by HMGB1. We will continue to look for evidence of the elusive ternary Oct-1/HMGB1/DNA complex and of protein-protein interactions. The role of DNA bending will be tested using HMGB1 mutants in which some or all of the three DNA-intercalating residues have been replaced by alanine. The ability of HMGB1 to facilitate Oct-1 binding to an occluded site in chromatin, and the requirement for the acidic tail, will also be tested, using a nucleosome reconstituted on to a natural promoter fragment. (5) The role of HMGB1 in increasing the accessibility of nucleosomal DNA will be investigated; this is probably how it enhances the action of chromatin remodelling machines. In particular we will examine the role of the long acidic tail and the effect of tail length and composition (Glu:Asp ratio). We will test the suggestion that HMGB1 tail/core histone contacts stabilise histone segments released by distortion of the DNA by the HMG boxes. (6) We will determine the effect of diacetylation of HMGB1, which occurs in vivo, on its ability to bind to various distorted DNA substrates and chromatin, and on the other properties mentioned above (4 and 5). Diacetylated HMGB1 will be produced enzymatically and/or by native peptide ligation.

Summary

DNA in the cell nucleus is packaged with a roughly equal mass of positive proteins called histones. The complex, chromatin, is organized as a repeating array of subunits (beads on a string) called nucleosomes, each containing a core of histones around which about 200 base pairs of DNA are wound. When particular genes are activated, the chromatin structure is partially disrupted over the promoter at the start of the gene to make the DNA more accessible to gene regulatory proteins and RNA polymerase, so that the information in the DNA can be transcribed into RNA. Unlike gene regulatory proteins, which are present in only a few copies in the nucleus, HMGB1 (high mobility group protein B1) and the related HMGB2a and 2b, are relatively abundant nuclear DNA-binding proteins, which do not have a strong preference for any particular DNA sequence. They exert their effects mainly through their ability to bend DNA and to recognise and stabilise distorted DNA. These properties probably account for the fact that HMGB1 and 2 have been implicated in many processes in which DNA is manipulated and kinks and distortions occur, for example, transcription, replication, DNA repair and recombination. The ability to bend DNA may also bring together gene regulatory proteins bound to sites on either side of the bend that need to interact in order to function. DNA binding is a property of two tandem HMG box domains of about 80 amino acid residues. They are attached through a positively charged linker to an unusual, long negatively charged tail consisting of 30 consecutive amino acids (aspartic acid or glutamic acid), whose side chains carry carboxyl groups. In the presence of the tail the HMG boxes bind DNA less well and there is some evidence that the tail interacts with one or both of the boxes. In the work that we plan to do over the next three years we will address several issues. First, we will find out more about how the acidic tail interacts with the HMG boxes and try to track its path. We will do this by progressively shortening the tail and then asking which regions of the HMG boxes have a changed signal using the technique of NMR spectroscopy; by also studying various tail mutants we will ask whether the relative amounts of the two acidic amino acids, or the particular amino acid sequence of the tail matters. Second, we will also use NMR spectroscopy to build up a picture of how the tandem HMG boxes of HMGB1 (A and B) bind and bend DNA. To help form unique stable complexes so that we can solve their structures, we will substitute the A or B box with an HMG box from a protein that binds to a unique DNA sequence. Third, we will try to describe in detail the nature of the interaction between HMGB1 and the important tumour suppressor protein, p53, which is mutated in 50 per cent of human cancers. We speculate that the interaction recruits HMGB1 to the p53 DNA binding site, thus facilitating p53 binding. Fourth, we will ask about the protein-DNA and protein-protein interactions involved in facilitation of binding of the transcription factor Oct-1 (important for mammalian development) to its target DNA, and whether DNA bending is important. We will study several different promoters to see how they behave and we will also ask whether HMGB1 can still help when the Oct-1 binding site is wrapped up in a nucleosome. Fifth, we will investigate the role of HMGB1 in increasing the accessibility of DNA wrapped up in a nucleosome, probably by distorting the DNA, which may help remodelling complexes to loosen the DNA even further, and whether the acidic tail plays a role by binding to positively charged histone regions that become exposed. Finally, we will ask what effect modification of two of the positively charged residues near the N-terminus of HMGB1 by acetylation (thus removing the positive charge) has on its DNA binding and other properties. Could this be a regulatory switch in the cell?

Committee	Closed Committee - Biomolecular Sciences (BMS)
Research Topics	Structural Biology
Research Priority	X – Research Priority information not available
Research Initiative	X - not in an Initiative
Funding Scheme	X – not Funded via a specific Funding Scheme

I accept the terms and conditions of use (opens in new window)

export PDF file

back to list new search