A fragment-based screening approach to rationalizing M. tuberculosis P450 molecular selectivity

The development of methods for the determination of entire DNA sequences of humans and other organisms has revealed huge amounts of new information in terms of identification of novel proteins and enzymes and unusual and unexpected chemical processes that take place within cells. From perspectives such as antibiotic therapy and biotechnological applications, there are obviously numerous opportunities once the substrates and reactions catalyzed by these novel enzymes are established. However, considerable efforts are often required to establish the function of a newly identified enzyme if there is no initial knowledge of the physiological role it plays. In addition, identification of specific inhibitors of its function can also be laborious and involve screening of enormous libraries of chemicals. However, recent years have seen the advent of a novel type of approach to defining molecules that bind to purified enzymes. This fragment based screening (FBS) technology involves analysis of binding of much more limited libraries of small molecules than would be used in typical compound screens as used, for example, in the Pharmaceutical sector. Initial 'hits' from the FBS approach may bind the target enzyme only quite weakly, but with the help of protein structural methods their position of binding can be determined accurately. Once a number of such hits are obtained, the combined information on their binding locations and the nature of the chemistry in these sites and the surrounding environment can provide a basis for 'elaborating' the structure of the molecules such that tighter binding is achieved, or even chemically 'merging' fragments bound at adjacent sites to enable construction of much tighter binding molecules as potent inhibitors of the target enzyme. This FBS method is now increasingly used in industry (in areas such as drug development) as a complementary method to the more typical 'high throughput screening' techniques that involved huge chemical libraries. In the current proposal, we will exploit and develop the FBS technology in studies of a class of enzyme called cytochromes P450 (P450s) from the TB-causing bacterium Mycobacterium tuberculosis (Mtb). The P450s are enzymes that bind oxygen to an iron atom in a heme group bound to the protein, and then 'activate' the oxygen to enable the insertion of an oxygen atom into their substrates, which are typically lipid molecules that bind in the active site of the enzyme, close to the heme group. A large number (20) of P450s are found in Mtb, and it is recognized that many of these are essential enzymes for bacterial survival and for their infection of the host. However, there is a paucity of information on the substrates and physiological functions of the majority of the P450s. This project will use FBS methods to define fragments that bind to selected Mtb P450s involved in cholesterol metabolism and in other unusual biochemistry, and then define their binding mode using structural biology methods. The identities and positions of binding of the fragments will then be used in further work to chemically join fragments and to add functionalities to fragments in order to create much tighter binding molecules as specific inhibitors of catalytic functions of the P450s, and as probes of their mechanism. Further, we will make new types of libraries that contain fragments of lipids, steroids and other molecules that are typical substrates for P450s, and use FBS in a novel way to identify substrate molecules for 'orphan' Mtb P450s with unknown function - again using a combination of chemical synthesis and structural biology to 'home in' on true substrates for these P450s, and so provide new knowledge on Mtb's complex lipid biochemistry. This application will thus provide important new information on the physiological chemistry of an important bacterium, and drive new applications of FBS in enzyme substrate identification through technical developments in the methodologies used.

Cytochrome P450 (P450) enzymes are a superfamily of oxygenases that perform an array of physiologically important reactions in organisms from bacteria through to man, including steroid and bioactive lipid syntheses, and xenobiotic transformations. With increasing numbers of P450 (CYP) genes identified from genome sequences, it is apparent that there is a large untapped resource of uncharacterized oxidase enzymes of unknown specificity. In Mycobacterium tuberculosis (Mtb) there are 20 P450s, with substrates definitively identified for ~6. The purpose of this proposal is to exploit and develop fragment based screening (FBS) technology to identify small molecule ligands that bind the active site of a panel of key Mtb P450 enzymes (using X-ray crystallography to define the binding modes), and then to (i) use a combination of chemical elaboration and 'merging' of fragments binding at adjacent positions, to iteratively improve their efficiency of binding and potency as inhibitors and as probes of structural/catalytic features of targeted P450s, and (ii) to generate novel libraries containing fragments representing both likely substrates classes for P450s (fatty acids, steroids, polyketides) and unusual lipids prevalent in and/or peculiar to Mtb, and then to exploit FBS with these libraries to generate/identify physiologically relevant substrate-like molecules, with reference to active site structure of the target P450s and knowledge of the metabolomics of Mtb. Studies on effects of inhibitory and substrate-like molecules on Mtb growth and transcriptional responses will also be done. The project will exploit FBS, structural biology, synthetic chemistry and various spectroscopic methods to generate novel inhibitors for Mtb P450s involved in cholesterol oxidation and secondary metabolite synthesis, and (in a development of the technological applications) to identify substrates for 'orphan' P450s to enhance knowledge of the biochemistry of a biomedically important bacterium.

The research proposal aims to exploit and develop fragment-based screening (FBS) technology to enable both (i) the generation of tight binding ligands as inhibitors and mechanistic probes for individual cytochrome P450 (P450) enzymes from Mycobacterium tuberculosis (Mtb); and (ii) the pinpointing and confirmation of substrate and substrate-like molecules using specially designed fragment libraries composed of fragments representative of both 'typical' P450 substrates (fatty acids, steroids, polyketides etc) and of lipid molecules peculiar to the metabolism of Mtb and related mycobacteria. The work proposed is multidisciplinary, involving protein expression and isolation, enzyme-ligand binding analysis using NMR, ITC and other techniques, structural biology and synthetic chemistry, and the ultimate aim is the exploitation and development of FBS methodology to provide routes to both specific inhibitors and substrate identification for a group of key enzymes in Mtb, ultimately enhancing the knowledge base on the biochemistry of Mtb and on the roles of P450 enzymes in its physiology. Beneficiaries from the research include scientists in industrial/academic sectors working on Mtb from perspectives of both (i) understanding its unusual lipid biochemistry, physiology and infectivity, and (ii) developing therapeutic strategies to combat the pathogen. TB is a global health problem and huge numbers of scientists worldwide are engaged in fundamental and applied research in the area. This extends from Pharma with major programmes on TB therapeutics through to University labs exploring Mtb's biological chemistry, metabolomics and genetics. The proposal features a combination of academic and applied research that will be of interest to scientists researching fundamental Mtb biochemistry, to technologists able to apply further the novel approaches and applications of FBS methodologies detailed in this application, and to biochemists/industrialists who can capitalise on novel information forthcoming on the roles of P450 enzymes in Mtb physiology and on tight-binding inhibitory ligands that can be exploited for mechanistic studies and to inhibit activities of P450s critical to viability and cholesterol metabolism for energy generation. From an even wider perspective, TB affects not only an increasing number of the UK and world population (with drug resistant strains spreading), but is also an important disease of cattle, with bovine TB and badgers as a vector of infection currently being a controversial scientific area. In this respect, research in the TB area is of major interest worldwide, with the disease endemic in numerous third world countries and with several antibiotics redundant as a consequence of the proliferation of drug-resistant and MDR strains of Mtb and other related mycobacterial pathogens. Thus, research proposed here (with direct impact on understanding of Mtb physiology and enzymology) could be of great relevance to national/international agencies as well as influencing Pharma by advancing technologies in the area and by identifying new therapeutic targets and strategies for intervention. The timescale for such impact is difficult to predict, but we expect that the developments in FBS approaches and concomitant work to identify/validate specific P450 inhibitor and substrate-like molecules will yield numerous compounds and novel insights into Mtb P450 molecular selectivity in 2-3 years. According to outcomes from this work and allied studies validating specificity and potency of molecules generated, we expect the research to inspire similarly directed therapeutic studies and adoption of related screening approaches in academia/industry soon after relevant publications emerge. We are also confident that the range of approaches (spectroscopic, structural, chemical and microbiological) will provide the PDRA appointees with a range of professional skills suited to career development in both academic and industrial settings.