Reporter 421, 27 May 1998

Virtual chemistry on campus could prove key in fighting disease

For decades chemists have laboured over test tubes and flasks, mixing and matching molecules to come up with new compounds to help fight diseases such as malaria and tuberculosis. But new software being developed at the University can design chemical reactions for them and the process can be speeded up even more by a unique robot capable of creating drugs one hundred times faster than a human ever could.

There are tens of thousands of compounds for chemists to plough through when developing new drugs, but now software designed at the University of Leeds can predict results from a database of chemical information that will make the starting list a lot shorter. Time and money is saved as a result.

Professor Peter Johnson has been at the forefront of this work, which reduces the amount of time spent in developing and testing new compounds. “At the moment it’s a numbers game – we’re just trying to shorten the odds.”

This work is of great interest to pharmaceutical companies as there is enormous commercial pressure on them to develop drugs quickly. It can take ten to 13 years from the initial lead to having the drug on pharmacists’ shelves and, as the patent may last for only twenty years, there is little time to put the discovery to use.

In the past these large companies would have access to hundreds of thousands of compounds and when they came to develop a new drug may have had to test every one of them. But by using software to eliminate some chemicals from the start, the process can be speeded up dramatically.

The search for a software programme to help chemists design new compounds began in Harvard more than twenty years ago under Nobel prize-winner Professor E J Corey, pioneer of the current LHASA (Logic and Heuristics Applied to Synthetic Analysis) project. Professor Johnson worked with him in the 1970s and has collaborated with a team at Harvard ever since. Pharmaceutical companies showed a lot of interest and clubbed together to form LHASA, based at Leeds, which supports the fundamental scientific research – the basis of their own work.

In 1990 the team turned to the problem of designing drugs to target specific biological targets. One example is thrombin – a protein involved in the biochemical sequence which ultimately leads to a heart attack. Researchers have been working to develop a way of ‘inhibiting’ thrombin which would prevent potentially fatal blood clots from forming. A range of software can suggest possible inhibitors, indicate how easy it will be to construct them in the laboratory and how reliable they will be.

Automation in the laboratory is also helping to speed up testing. The School of Chemistry has a robot which can carry out research on an enormous scale – whereas one drug used to take a week to make manually, the robot can do 50 to 100. Leeds is the only UK university to have such a robot. But chemists needn’t fear for their jobs, says Professor Johnson. “These advances will change the nature of work that chemists do and remove some of the drudgery – and robots can’t do everything!”

In contrast to the rational method used to design drugs another more random method is also being developed by Professor Ron Grigg. Combinatorial chemistry involves making hundreds of thousands of compounds very quickly, coming up with several possibilities.

The researchers are collaborating with other teams across the University. The Anti-microbial Research Centre, headed by Professor Ian Chopra, is working to find inhibitors to bacterial proteins such as TB. “It’s rearing its ugly head again and a lot of traditional drugs are losing their effectiveness because of resistance,” says Professor Johnson. Work is also being carried out on anti-malarial drugs with the Department of Biology. The software may be able to predict if drugs will have any side-effects – a concern to many users.

Professor Johnson believes that the future of scientific advances lies in this kind of cross-departmental work: “Most of the major breakthroughs happen at the interface between disciplines but work within departments can also be fruitful.” The software is also applicable to the atmosphere and reactions within it and Professor Johnson is working with Professor Mike Pilling to build a model to predict the effects of releasing certain substances into the atmosphere.

A large part of the department’s software development is being carried out by LHASA, an independent company which works closely with the University and last year put back more than £200,000 into the School of Chemistry – one of Leeds’ largest incoming grants from research activities. Its collaborators include Glaxo Wellcome, Hoffman La-Roche and Smith-Kline Beecham. One of its programmes, DEREK (Deductive Estimation of Risk from Existing Knowledge), predicts whether new substances will be toxic or carcinogenic: “There is a big time and financial benefit to companies to minimise toxicity tests as well as there being ethical and political considerations,” says senior project officer Dr Jan Langowski. The software is currently used by government departments, pharmaceutical and agrochemical companies.

A wide range of experts including psychologists and external collaborators such as the Imperial Cancer Research Fund are involved in another LHASA programme designed to assess risk. The Standardised Argument Report project draws upon recent developments in mathematics and logic and human perception of risk and its initial aim is to build a programme for risk assessment of carcinogenicity. It will be useful to food industries and chemical manufacturers and could possibly be extended to other areas where risk is important such as financial planning.

“I see much less toxicity testing being done in labs in the future,” says Dr Langowski, whose team has recently embarked on a new project to predict the likely metabolic fate of chemical compounds. “In the majority of cases a chemical is not toxic in itself but causes a problem because the body converts it into something that is toxic.”

This software should be ready within three years, eliminating some of the more time-consuming tasks for chemists. The use of robots looks set to grow, so we will have to see who – or what – is shaking those test tubes in years to come. *More...

[Main news stories | In the news | Letters | News in brief | Events | Notice board]

HTML by Jeremy M. Harmer