Our current laboratory research
Structure-function relationships in matrix metalloproteinases
Matrix metalloproteinases (MMPs) are a family of zinc-dependant endopeptidases that are important not only in normal physiological processes but also in conditions such as cancer, arthritis and cardiovascular disease. The MMPs are produced as inactive zymogens with an inhibitory pro-domain (PRO) blocking the active site of their catalytic (CAT) domains. The majority of MMPs also possess a hemopexin (HPX) domain which is multifunctional but in most cases allows the enzyme to discriminate between substrates (Figure 1). We are interested in answering two specific questions: (1) what is the series of events leading to MMP activation? and (2) what is the molecular mechanism of collagen breakdown (“collagenolysis”) by MMPs? The latter question is particularly intriguing as the active site cleft in the CAT domain is too narrow to accommodate the collagen triple helix. Thus, the enzyme must somehow unwind the helix prior to hydrolysing the polypeptide backbone.
We employ a variety of biophysical methods to perform multi-disciplinary research on the structure-function relationships in collagenolytic MMPs. This is illustrated in our recent article (Arnold et al., 2011) in which, using a combination of analytical ultracentrifugation (AUC), nuclear magnetic resonance (NMR) spectroscopy, surface plasmon resonance (SPR) and small-angle X-ray scattering (SAXS), we found that the interface between the CAT and HPX domains in MMP-1 conceals residues that are important for collagen recognition.
Biomolecular structure determination using swarm intelligence
The calculation of protein structures from NMR spectroscopy data is a laborious task. Typically, it involves the identification of hundreds to thousands of close-range interactions (nuclear Overhauser effects or NOEs) between pairs of hydrogen nuclei (protons) in the protein. These protons are identified by their chemical shifts and thus an NOE by a pair of chemical shifts. If multiple protons have the same (or very similar) chemical shifts then NOEs involving those protons are ambiguous, i.e. they cannot be assigned to a particular donor-acceptor proton pair without prior knowledge of the protein structure. However, the protein structure cannot be determined without assigning the ambiguous NOEs. This is known as “the NOE ambiguity problem” and solving it is usually the rate-limiting step in biomolecular structure determination from NMR data.
We have developed a novel method of solving the NOE ambiguity problem that is built upon the concept of swarm intelligence – the apparent cleverness that arises from the co-operation between simple beings such as social insects. In our swarm intelligence NMR method (Figure 2), a swarm of biomolecular ants explore their conformational space through molecular dynamics simulations, in an analogous fashion to spatial exploration by a real colony of ants. Any biomolecular ant that encounters a conformation which agrees well with the experimental data informs the other members of the swarm and encourages them to adopt that same structure. In so doing, the ants co-operate to find the solution structure of the protein without the need for the laborious explicit assignment of the ambiguous NOEs. This novel method has been patented and was used recently in the NMR structure determination of estrogen receptor ligands (Phillips et al., 2011).