Solving the structure of the mycobacterial chromosome condensing protein Lsr2 in complex with DNA

Abstract

Lsr2 is a DNA binding protein that is highly conserved in mycobacteria and related actinomycetes and it is thought to be essential in Mycobacterium tuberculosis. Previous studies have shown that Lsr2 is involved in down-regulating a range of genes involved in cell wall synthesis and metabolic functions and it is proposed that it does this by organisation of bacterial chromatin. We solved the structure of the N-terminal dimerisation domain of Lsr2 using crystallographic ab initio approaches1 whereas the C-terminal DNA binding domain structure was solved by others using NMR2. Electron microscopy shows that Lsr2 organises DNA into large helical structures involving several strands1. Whilst DNA binding by individual domains has been modelled based on the NMR structure, the exact mechanism of DNA binding and chromosome organisation by the entire protein is unknown. Lsr2 contains a long flexible loop between the two domains which may lead to the protein having a large range of movement, allowing it to bind DNA in a dynamic way. The Lsr2 dimer is also capable of forming linear oligomers through the interaction of overlapping N terminal residues and evidence of this has been shown in our crystal structure and using TEM1. We wish to solve the structure of Lsr2 bound to DNA. For the purposes of crystallisation of Lsr2 we have engineered the removal of important N-terminal residues involved in oligomersiation to prevent this process occurring in solution. The truncated form of Lsr2 provides a sub-population of purified Lsr2 that is DNA-free and we have utilised this population for binding specific dsDNA oligonucleotides for crystallisation. To date, we have determined the optimal length of DNA required and have refined the exact combination of nucleotides for ideal protein binding. We have progressed through the crystallisation of Lsr2 bound to a range of oligonucleotides. One oligonucleotide has yielded the most success to date with crystals that have diffracted to 2.7 Å resolution. Attempts to solve the structure using molecular replacement have thus far been unsuccessful. Our current focus is to reproduce crystals for improving resolution and for heavy atom soaking to assist with solving the structure.