Biocatalysis for novel drug synthesis: re-purposing haemoproteins for late stage heteroatom and isotope functionalisation of pharmaceuticals

This work aims to develop new enzymes to allow chemists to synthesise complex drug compounds in a more straightforward manner. Enzymes are biological catalysts, and they are increasingly being integrated into large scale chemical manufacture (such as for the pharmaceutical and flavour/fragrance markets). Biocatalytic routes hold a number of advantages over traditional chemical methods, particularly in their ability to carry out very specific reactions on complex molecules under environmentally benign conditions. The specific nature of enzymes allows the steps in a manufacturing process to be minimised, generating less waste and requiring less energy. However, enzymes have their limitations, particularly if the desired reactivity is not known to occur in the natural biological systems from which they originate. The goal of this work is to "engineer" the enzymes to enable them carry out new non-natural reactions, which are otherwise very difficult for chemists to achieve.

The target reaction for this work is the conversion of carbon-hydrogen (C-H) bonds to carbon-fluorine (C-F) or carbon-deuterium (C-D) bonds (deuterium is a safe isotope of hydrogen). Whilst these reactions sound simple, they are very hard to achieve in a controlled manner, and often require costly and wasteful reaction procedures. Yet the C-F and C-D bonds are very valuable to chemists designing new medicinal compounds, because they stabilise the drugs against enzymes in the body, such as the liver). The liver enzymes degrade the drug by attacking the C-H bond and converting it to a C-OH bond, marking it for removal from the body. Hence, swapping the C-H for C-F or C-D before the drug is taken can allow it to last longer in the body and therefore lower the dose required - which can very beneficial for the patient. The C-F strategy is so effective, that it was used in around 50 % of drugs approved in 2018. The C-D approach is newer, but has now been used in commercial drug compounds too. Unfortunately there are currently no widely available biocatalytic routes to converting C-H to C-F and C-D bonds.

Creating biocatalytic routes to C-F and C-D bonds would allow new drug molecules to be more easily prepared, allowing medicinal chemists to design and test more drugs more quickly and with less waste. In this work, enzymes like those found in the liver (and elsewhere in nature) are used as inspiration to design new strategies to form C-F and C-D bonds. The artificial enzymes will still be required to break the C-H bond, but then react in a different way to form C-F or C-D rather than C-OH. At this stage, even if the new enzymes for C-F or C-D bonds to only a very small degree, it will mark a major advance, and will enable further study to get the new behaviour to a level where it would be useful for drug development and manufacture.

The benefits of this research are therefore three-fold: (1) it will enable researchers working on drug discovery to make desired compounds for testing more quickly and easily. (2) it will enable manufacturers of high-value pharmaceuticals to establish more sustainable manufacturing practices, and (3) it will increase the understanding of the functioning of a very important class of enzymes, which will benefit other aspects of industrial biotechnology (such as flavour/fragrance manufacture). The work therefore falls into the BBSRC remit of Industrial Biotechnology and Bioenergy (the use of biological resources for producing and processing materials, chemicals (including pharmaceutical precursors and biopharmaceuticals) and energy). In the long term, it will make significant contributions to the economy of the UK-based pharmaceutical industry and public health.

This work aims to provide a new biocatalytic route to C-F and C-D bond formation by modified haemoproteins. Given the relevance of these reactions to the pharmaceutical industry, the work falls clearly into the BBSRC Industrial Biotechnology and Bioenergy remit. The overarching strategy is to utilise haemoproteins such as cytochromes P450 (CYPs) to activate selected C-H bonds, but then interrupt the typical next step (radical-rebound) and promote new pathways (hetero-atom rebound) to form C-F or C-D instead of the expected C-OH hydroxylation product.

A set of haemoproteins will be expressed and purified, and a library of mutants will be prepared by rational mutagenesis techniques. Then, haem-reconstitution will be carried out to generate enzymes based on Mn, Cr, and Ru, all displaying different rebound properties to the native Fe-protein. Following the development of suitable analytical procedures (GC-MS, NMR), the metalloenzyme library will be screened for C-H/C-F and C-H/C-D exchange, as well as conventional C-H/C-OH activity. At this stage, substrates will be simple (ethylbenzene, camphor, stearic acid).

Detection of C-H/C-F or C-H/C-D exchange activity will mark a major advance in haemoprotein chemistry, and characterisation will be carried out to attempt to rationalise the behaviour. Similarly, any change in the C-OH selectivity of the enzyme in question will be investigated further. EPR will be used to identify long-lived radical species, and computational methods will model the energy landscape of the various reaction pathways of the chosen candidate.

Finally, the scope of the reactivity of successful candidates will be explored to discern the substrate promiscuity and performance of the enzyme under different reaction conditions. A selected compound will be synthesised on a 100 mg scale, and transferred to medicinal chemistry collaborators. Finally, radiolabelling work with 18F (well established at Oxford) or T (tritium) will be explored.