Dr M Madan Babu of MRC LMB wins prize for insights into the machinery inside our cells
Cambridge computational biologist wins Blavatnik Award for work on key proteins in our bodies
If the breadth of our knowledge of human biology is incredible, then perhaps more astonishing is how much we have left to discover.
Dr M Madan Babu and the group he leads at the MRC Laboratory of Molecular Biology in Cambridge are on a mission to illuminate some particularly dark and disorderly gaps in our understanding.
With them, they are taking a toolkit of striking diversity: Computational biology, bioinformatics, genomics, machine learning, theoretical physics, big data analysis and more are being brought to bear on probing key proteins – the molecular machines that execute processes in our body.
The outcome will be a better understanding of human health and disease and could have significant implications for healthcare and the development of new medicines.
Dr Babu has been named Life Sciences Laureate at the 2018 Blavatnik Awards for Young Scientists in the United Kingdom for his work. On Wednesday March 14, 2018, he received the award at the Victoria and Albert museum in London.
For a decade, the Blavatnik Awards at the New York Academy of Sciences have honoured outstanding young scientists under the age of 42 in the United States. They have been extended to the UK and Israel for the first time this year.
Dr Babu’s citation acknowledges two key areas of his work – the first being his insight into G-protein coupled receptors, or GCPRs.
He explains: “GPCRs are a class of proteins that are a major drug target at the moment. They regulate virtually every aspect of human physiology, ranging from the functioning of the heart and the brain to the lungs.”
Implicated in numerous human disorders, these proteins sit on the membrane of our cells. They receive signals – which could be from the presence of light, nutrients or a small molecule – and they convey it inside the cell to ‘G proteins’, setting off a chain of other processes.
“All these GPCRs are an evolutionary-related family of proteins, which means they all have a similar shape and structure,” says Dr Babu.
GPCRs resemble threads that cross the cell membrane in seven loops. At either end are ‘cups’ for binding to a ligand – such as a drug.
“When it binds, the receptor changes its shape. Then it allows proteins inside the cell to bind the receptor and get activated, and that’s how it triggers a whole signalling cascade,” says Dr Babu. “Twenty-seven per cent of all drugs in the market target 108 of the 800 GPCRs in the human genome. Virtually every one is potentially a drug target. That’s why there’s a lot of interest from pharma and academia.”
Beta blockers, which lowers blood pressure by slowing down heart rate, are an example of a drug that bind to these receptors.
GPCRs are influential in the immune, nervous, reproductive and hormonal systems, as well as the sensory systems such as smell, vision and taste. Unsurprisingly, given their significance, they have been extensively studied – but Dr Babu and his group at the LMB on Cambridge Biomedical Campus are generating new understanding by bringing together global research into a common framework.
“We analyse large amounts of information about protein sequences, structures and genomic data,” he says. “Each one of these datasets has been independently generated by different labs or groups and what we try to do is put them all together to understand how these proteins function in ways that were not possible before.
“There is a lot of published information about the structure of GPCRs. At the moment there are about 216 structures describing 46 different GPCRs.
“We study these using new computational methods that we have developed to understand what is common between all these structures and what is unique. Then one can think about designing drugs unique to these receptors, or a molecule that can have an impact on more than one receptor.”
Developing an atomic level understanding of how these receptors function and using data from the genomes of 100,000 people that are now publicly available, the group seeks to understand the impact of mutations in the human population.
“An important question we wanted to address is what fraction of the population has a mutation in a GCPR and is therefore unlikely to respond effectively to drugs. And if you have a different type of mutation, you might over-react to the drug because it binds more strongly,” says Dr Babu.
“One of the key findings of our recent work is that at least 3-12 per cent of the population have mutations within a functional region of the GPCR drug target.
“We have provided a first view of the landscape of variability in drug targets, so we can start to assess whether it’s worth thinking about personalised medicine. I think there is a convincing case.”
Economic modelling by the group suggests the NHS could benefit from considering pharmacogenomics – which describes the variation in pharmacological response due to genetic differences.
“If there is variability, then we might have ineffective prescriptions in two ways: One, a drug doesn’t do anything so you are treating a patient for longer with no result and two, the individual might have an adverse reaction because they over-react. They could be hospitalised.
“It would mean sequencing an individual’s genome, finding out if they have mutations in the drug target and whether that is likely to affect drug binding,” says Dr Babu.
One of the group’s studies focused on opioid drugs given to patients after surgery or in acute pain.
“We found several polymorphisms – variations between individuals – near the drug-binding pocket,” says Dr Babu. “What’s interesting is that some individuals’ mutations render the drug ineffective. They need much higher doses. There’s another end of the spectrum where people have an adverse reaction. If the dose is very high, they could potentially die. This is called opioid-related death.
“It’s not common but it does happen. We have all this information so we can potentially prevent opioid-related death or provide better, more effective treatment.”
The group’s work could prove hugely valuable to researchers and pharmaceutical companies developing new drugs.
“Each mutation takes a very long time to test experimentally. Computational methods come in handy because you can now prioritise which of these 15,000 or so mutations we should look into. We do computational modelling and experimentally test our most promising hypotheses.
“Then we make a comprehensive resource available to the research and pharmaceutical community. For instance, if they are doing clinical trials, it can be important to know if a particular receptor is very polymorphic. They’ll need to stratify the patient population for the best response.”
While GPCRs are much studied, the second area cited in Dr Babu’s award has been somewhat mysterious to biologists until recently, despite being present in about 40 per cent of human proteins.
“Understanding an organism requires understanding all the proteins that are encoded by them,” says Dr Babu. “DNA is a blueprint and proteins are the molecular machinery that execute different processes.
“The classical view of how proteins function is that a sequence of amino acids needs to form a precise shape to carry out their function. This is exemplified by GPCRs. This is called the structure-function paradigm.
“The last few years there has been some very interesting research going on in our group and elsewhere looking at a huge group of proteins that don’t fit the paradigm, meaning the sequence of amino acids does not adopt a unique structure. It can adopt multiple different shapes and each can be functionally relevant. This is called the disorder-function paradigm.
“Our group, along with a few others, has been spearheading our understanding of this.”
Proteins with these ‘intrinsically disordered regions’ or IDRs are flexible, enabling them to interact with many other proteins. Unlike structured proteins, which are comprised of amino acids with an aversion to water, these ones like to interact with water.
“The major contribution from our group has been to understand the principles by which disordered regions contribute to function using computational methods,” says Dr Babu. “A large fraction of the human proteome has large disordered regions. In the past, people thought they were passive linkers and not that important. It’s becoming clear that there are functional bits. What we don’t know yet is which parts are functional and which are not.
“Through recent work we’ve been developing high-throughput approaches that try to examine thousands of disordered sequences simultaneously in a single experiment to see which ones have a function.
“Once we know that, we can apply machine learning to learn the rules or features that make certain sequences have a function.”
What is known is that IDRs are used to rewire protein networks.
“By rewiring you generate new combinations of proteins that come together that have distinct functions,” says Dr Babu.
These regions are not obvious drug targets because their flexibility means they lack the ‘pocket’ found on structured proteins like GPCRs that are used in binding.
“But it opens up completely new opportunities for developing drugs that don’t exploit the structure-function paradigm,” Dr Babu points out. “The current drugs have been worked out using that paradigm so we have a bias towards that. This presents a challenge and an opportunity.”
One technique being explored utilises the destruction machinery that is responsible for destroying proteins when they go bad.
“You can’t bind to the disordered region but these always occur in combination with some structured region. But the destruction machine typically recognises the disordered region. By developing molecules that can bring together the structured part of the protein and the destruction machinery together, one can label the disordered region for degradation.
“It’s a very different strategy for drug development,” says Dr Babu.
“People are also trying to develop drugs that lock the disordered region in a certain conformation or shape. Because flexibility is the most important aspect of its function, by making them rigid, you can also make them non-functional.”
The researchers are applying machine learning to uncover the rules that makes proteins in the disordered region function.
“Once we do it will be fantastic and we could synthesise new sequences that have never been seen in nature but that can have functions exquisitely tuned by the algorithm,” says Dr Babu. “We can also look at which parts of the disordered regions in the human genome are functional. Then we can ask if a mutation in the disordered region is likely to have an effect and cause disease.
“We’re hoping we can make a potentially big impact by understanding a part of the proteome that has been largely dismissed because they don’t adopt a defined structure.
“I find that personally extremely satisfying. There is an enormous untapped potential. What it is pointing us to is the other half of an unexplored world.
“A colleague of mine once told me: ‘We look for lost keys under the light because that’s where we can see’. It doesn’t mean there aren’t other places around it.”
Recognition for computational biology
Dr Babu said he was delighted to win the 2018 Blavatnik Awards – and particularly pleased to see the recognition for computational biology.
“It is so inter-disciplinary,” he said of the field. “It tries to integrate concepts and ideas from physics, chemistry, mathematics, computer science and different aspects of biology, like genetics, evolution and biochemistry, so we do have a very diverse group in our lab, some of whom are experimental and some computational scientists.
“Even among the computational researchers, we have diversity. Some are theoretical physicists, some are computer scientists by training, so it makes for a very lively group who learn from each other. It’s been very rewarding.”
Dr Babu says “I am very grateful and feel extremely fortunate to have had the opportunity to work with highly talented and motivated team members, collaborators, colleagues and inspiring mentors. Without the long-term support of the Medical Research Council, none of this work would have been possible.”
Dr Babu completed a BTech at Anna University in India before his PhD in computational biology at the MRC LMB in Cambridge under the advice of Dr Sarah Teichmann, who is currently working on the Human Cell Atlas.
As a postdoctoral scientist, he worked with Dr L Aravind at the NCBI, National Institutes of Health in the United States.
Meanwhile, the MRC LMB’s John Briggs is a Life Sciences finalist in the awards. He was recognised for ‘advances in high-resolution visualisation of viral particles including HIV, using cryo-electron tomography combined with fluorescence imaging’.