The replication and repair of DNA requires a vast array of different proteins. In order to understand what goes wrong when DNA is damaged or replicated incorrectly, we have been studying a number of systems to learn more about the details of these processes at the molecular level. The systems we work on are often bacterial as these model systems can help us to understand the basic mechanisms utilised by these enzymes but the proteins are often much more amenable to structural analysis than their human equivalents yet operate by similar mechanisms.
Mechanism of DNA helicases involved in repair of DNA
In bacteria, chromosome breaks are processed by the RecBCD enzyme complex. RecBCD acts by binding tightly to the blunt end and then, by using two helicase motor subunits of opposite polarity, it unwinds the DNA. The two separated strands are both digested by a single nuclease domain in the RecB protein. This process of DNA degradation continues until the enzyme encounters a specific eight base sequence known as a Chi site, at which point the activities of the enzyme change. The nuclease activity is attenuated on one strand but continues on the other to produce a 3' tail, onto which RecBCD loads the RecA protein. RecA then initiates a homologous recombination reaction to re-establish a replication fork and thus repair the break.
RecB and RecD helicase subunits have opposite polarities of unwinding. In order to understand the molecular determinants of helicase directionality we have determined the crystal structure of complexes of RecD with ssDNA in the presence and absence of an ATP analogue, ADPNP (Figure 1). The structures reveeal how the enzyme is able to walk along DNA in a 5'-3' direction in an ATP-dependent manner. In colaboration with Martin Webb's group at NIMR, London we have shown that the step size for translocation along DNA is one base per ATP hydrolysed, consistent with the mechanism suggested by the crystal structures.
Figure 1. Crystal structures of the complexes of D.radiodurans RecD2 with single stranded DNA in a) the absence and b) presence of an ATP analogue
Another area of active investigation is the interaction between the RecBCD complex and RecA in order for us to understand the mechanism by which RecA is loaded onto DNA as a prelude to homologous recombination. Recently, in collaboration with Ed Egelman at the University of Virginia, we have obtained EM data showing how the nuclease domain of RecB interacts with RecA filaments.
During DNA replication, the progression of the replication fork frequently stalls due to DNA damage that is detected by the polymerase during chain extension. Stalling of the replication fork is followed by disassembly of the replication apparatus. There appear to be several ways that stalled forks can be processed. One system that we have been studying involves reversal of the replication fork by a protein called RecG. This protein 'backs up' the fork to form a four-way 'Holliday' junction. This process allows the DNA polymerase to use a different DNA strand as the template for replication in a process called 'template switching'. Once the chain has been extended further, the four-way junction can be regressed past the original DNA lesion and replication can proceed. By utilising this mechanism, the DNA damage is by-passed but maintains the fidelity of replication. The DNA damage can then be repaired at a later stage.
In order to understand how RecG catalyses DNA replication fork reversal, we have determined the X-ray crystal structure of the enzyme bound to a synthetic DNA substrate that mimics a stalled replication fork. However, the substrate used previously was rather too short to extend across the motor domains and now we have crystallised the protein with longer DNA substrates that reach across to these domains. We have also crystallised the complex in the presence of an non-hydrolysable ATP analogue to determine the mechansm by which RecG walks along double-stranded DNA.
A molecular switch in AAA+ proteins
AAA+ proteins are involved in many processes in cells including DNA replication and repair. The ATPase activity of these proteins is regulated by ligand binding and also by association with other protein co-factors. However, the mechanism of this control was unclear. In a collaboration with Xiaodong Zhang at Imperial College London, we have identified a mechanism for this process. Analysis of structures in the protein database show that an active site glutamate residue adopts two different conformations in these proteins, one of which is competent to promote hydrolysis of ATP while the other is not. Ligand binding switches the glutamate residue between these two conformations and hence regulates ATPase activity. This switch was first observed in a transcriptional activator by Prof. Zhang's group. However, our work has extended these observations to look at different states of the proteins from different AAA+ families.
Structures of an archeal replication initiator protein bound to DNA that we reported last year show how this switch operates in these proteins to prevent ATP hydrolysis when the protein is bound to DNA. Similarly, our previous work on structures of the eukaryotic DNA polymerase processivity clamp (PCNA) loading complex (Replication Factor C) combined with biochemical data on mutant enzymes show that this regulation process works in the opposite direction with DNA binding being the cause of stimulation of the ATPase activity. Consequently, the switch operates across many different AAA+ families and reveals a regulatory mechanism that is widespread in AAA+ proteins.
For a list of refereed research papers, see Publications (in navigation on left).
