Figure 1: Screening tens of thousands of expression constructs of a target gene. Constructs are made as a random library and printed on membranes for soluble expression analysis by hybridisation of fluorescent antibodies.
Figure 2: A previously unsuspected domain from influenza polymerase, identified by HT expression screening of tens of thousands of random DNA constructs and structurally characterised by X-ray crystallography. A single mutation to lysine at residue 627 (A) can be responsible for the evolution of human influenza viruses from wild-type avian viruses that have a glutamic acid at this position (B). The mutation of residue 627 reinforces or disrupts a striking basic surface patch and we are seeking to understand how this affects polymerase function.
The Hart team develops new high-throughput molecular biology methods and uses them to study enzymes of biological and medical interest.
Previous and current research
Combinatorial methods (e.g. directed evolution, phage display) are used to address problems that are too complex for rational design approaches. Large random libraries of variants are constructed and screened to identify rare hits with the desired property. In our ESPRIT process, for example, all truncations of a target protein are generated and screened using advanced picking and arraying robotics. Consequently, we are able to study certain biological questions with advantages over classical approaches.
Influenza RNA polymerase: There is global concern that currently circulating avian influenza viruses will cross the species barrier and become highly pathogenic, human transmissible strains with pandemic potential. This could result from residue changes in several influenza proteins, either by point mutations or through shuffling of the segmented avian and mammalian viral genomes. We are now characterising the interactions of these mutants with host cell factors using both structural and biophysical methods with the aim of understanding mechanisms of influenza host specificity.
Human kinases: Protein kinases play a crucial role in cellular stress responses as mediators between the upstream receptor and downstream gene regulation and are key components in coping with changes in the intra-/extracellular environment. When these mechanisms malfunction, diseases such as excessive inflammation, autoimmune disorders and cancer can occur. Kinases therefore represent important pharmaceutical targets for drug design. The multi-domain nature of many kinases reflects the need to regulate catalytic activity. We are screening for stable constructs that extend beyond the conserved regions of the catalytic domain and well-expressed internal domains presumably implicated in complex formation or regulation.
Intrinsically unstructured proteins and their interactions: A large proportion of the proteome possesses little or no structure. These regions may serve as simple linkers, however sometimes they become structured upon binding partners (proteins, nucleic acids, small molecules). In collaboration with the Blackledge group (IBS, Grenoble), we have recently begun studying several such systems involved in viral replication.
Histone deacetylases (HDACs): Using our construct screening technology, we have identified well-expressing, catalytically active constructs of an HDAC involved in cholesterol homeostasis. Using these proteins, we are investigating how new inhibitors bind using X-ray crystallography and enzymatic inhibition assays. Secondly, using a library-format protein interaction screen, we are trying to identify HDAC-interacting domains of cellular proteins. If determined, disruption of such protein-protein interactions suggests a new route towards specific HDAC inhibition.
Future projects and goals
We will continue to develop expression methods to handle protein complexes, targets that require eukaryotic expression for correct folding, and possibly aspects of membrane proteins. Each project uses ‘real’ targets of interest and the aim is to take advantage of recent method advancements to yield previously unobtainable biological knowledge. For example, we are testing permutations of influenza-influenza and influenza-host proteins, with the aim of defining expressible, crystallisable protein complexes that should provide insights into virus-host cell interactions.