Drug discovery is a complex business – particularly in a field like cancer, where new-style treatments are increasingly aimed at specific molecular features of tumours. Drug targets must be identified and validated, and prototype drugs designed, synthesised, tested and refined. It’s a process which used to take years, but at The Institute of Cancer Research, London, our scientists are now accelerating compounds into preclinical development through a sophisticated marriage of high-throughput chemistry and cutting-edge computational techniques. A virtual chemistry lab is helping scientists design drugs to target cancer-causing proteins more precisely and more quickly than ever before.
With over 20 years of experience in medicinal chemistry and drug discovery to his name, Professor Julian Blagg now leads the medicinal chemistry group, which is translating our knowledge of cancer genes into new drugs for patients as fast as possible. Leading a 50-strong team of chemists, Professor Blagg’s team designs, synthesises and optimises potential new therapies to combat cancer, collaborating with experts in clinical treatment, genetics, biology, structural biology and drug metabolism.
“Groundbreaking research has driven the tremendous progress in our basic understanding of cellular processes that lead to cancer,” Professor Blagg says. “For these exciting breakthroughs to translate into patient benefit, we need to discover therapeutics that selectively alter these cellular processes that go wrong in cancer. Many of these therapeutics are small molecules that are designed, synthesised and optimised by medicinal chemistry teams such as ours.”
Starting at the molecular level, the team homes in on certain proteins present in cancer cells, which are potential drug targets. “First of all, we need to identify the molecules, or ‘hits’ that bind to our target protein – a term we called ‘hit discovery’. High-throughput screening is used to test the activity of hundreds of thousands of potential hits against a particular target using a series of tests that can be carried out rapidly and on a huge scale. Not all of these hits will be relevant, so they are further refined through other rigorous tests before synthesis and testing. Once we know our effective hits, we use these as a basis to design and create new chemical compounds which are essentially prototype drugs, and usually only one or two hits will go on to become lead compounds for further development.”
One way to potentially speed the rate of drug discovery while reducing the need for costly laboratory work and clinical trials is using computational approaches – a term called in silico research. Dr Nathan Brown, who leads the in silico medicinal chemistry group, says: “As you can imagine, drug discovery activity is lengthy and expensive. But we can now design virtual small molecule compounds which modulate cancer-causing pathways using a suite of cutting-edge in silico technology.”
Computational tools used by Dr Brown include scaffold hopping and bioisosteric replacement. These are routinely used together in drug design to improve the properties of the drug and to remove unwanted side-effects such as toxicity. “Scaffold hopping allows us to identify other agents within a compound library that have a similar activity but differ in the core structure. Conversely, bioisosteric replacement is where we change the structure around the core. It’s a bit like creating structures out of building blocks and then replacing a bioactive part, say a blue rectangle block, with a red rectangle block that is similar in size and exhibits similar properties. But by changing the colour of the block, we might remove unwanted side-effects whilst maintaining the original bioactivity of the compound. As you can imagine, there are millions of potential new safe compounds and we call the area that they occupy ‘chemical space’. To give you an idea of the size of the space, if the surface area of the Earth represented all of theoretical chemical space, all of the compounds that have been made would cover less than the footprint of a London Routemaster bus.”
Professor Blagg’s team are always striving to create new in silico tools and have recently validated a new scaffold-focused virtual screening with a standard whole molecule similarity search. “Our method identified potential hit-to-lead compounds that were more structurally diverse from the query compound compared with the standard method,” Dr Brown reveals. Having a range of starting points provides more freedom in the drug discovery process as we are not limiting the early exploration of chemistry space. Once we sufficiently explore this space, we can home in and exploit the most promising regions.”
Computational chemists have a huge impact on every aspect of medicinal chemistry. “Using our in silico expertise, we can suggest and optimise compounds – this is extremely powerful technology at our fingertips,” Dr Brown says. “But to know that some of these drug discoveries we are working on will eventually make it to market, and will have the potential to transform patients lives, is truly inspiring.”