Discovering new drugs requires mix of experimental and computational methodologies

This according to Tom Blundell, Professor Emeritus at Cambridge University, one of the pioneers in the application of the fragment-based approach to discovering new drugs.

By Fábio de Castro

Agência FAPESP – Sir Tom Blundell, Professor Emeritus at Cambridge University’s Biochemistry Department, has been dedicated to structural biology and bioinformatics research with a focus on applications in medicine and discovering new drugs since 1970.

Over time, Blundell realized the growing financial difficulties that large companies in the pharmaceutical and agrochemicals areas faced when attempting to develop new products. The need to make discovering new molecules cheaper in the last two decades prompted Blundell to seek a new approach: “the discovery of new drugs based on fragments.”

The strategy consists of identifying small chemical fragments that link well-defined biological targets, utilizing computational methods to enlarge them and combine them, producing lead compounds with greater efficiency than methods based on screening large molecules.

In 1999, Blundell, his colleague Chris Abell and businessman Harren Jhoti founded Astex Therapeutics, a company that began to successfully apply the new approach to drug discovery.

The use of structural biology in diverse applications of the fragment-based approach to designing new drugs was the topic of the conference presented by Blundell on April 1st in Campinas (SP) at the opening of the course on Advanced Topics in Computational Biology – Agrochemical & Drug Design.

Coordinated by Goran Neschich of Embrapa Informática Agropecuária’s Computational Biology Research Group (GPBC), the event was held under the auspices of the São Paulo Advanced School of Sciences, a program funded by FAPESP.

The course brought together more that 80 Brazilian and foreign undergraduate and graduate students from the areas of computational biology, biochemistry, structural biology and biomedical sciences. During the event, Blundell, who is also a Royal Society Fellow, granted an interview to Agência FAPESP. The interview follows below.

Agência FAPESP – The approach known as “discovery of new drugs based on fragments” is considered an emerging paradigm for the sector. Why was this change needed?
Tom Blundell – We need many new drugs and agrochemicals. The problem is that, mainly in the last 20 years, the costs of making new medicine and agrochemicals have increased tremendously. This occurs due to the new genomic and chemical technologies, which are very expensive, but also because the regulatory demands have increased, impacting the cost of developing these new molecules.

Agência FAPESP – Could this fragment-based approach be the solution to this challenge?
Blundell – There are two ways of dealing with this problem. The first is improving the efficiency in manufacturing these new molecules in order to reduce costs. The other manner is to increase the degree of certainty involved in the initial stages of production so that failures during development do not increase costs. The new approach is focused on this second line of action.

Agência FAPESP – How do we increase this degree of certainty?
Blundell – In the last few years, we have been seeking new manners to apply structural and computational methods to identify molecules with more selectivity. For new drugs and agrochemicals, the methodologies are more or less identical. The advances in which I am involved – which had contributions from many others – are related to the identification of new targets to attempt to understand which are critical points for intervention. Until then, we had focused on large families of proteins, kinases or proteases, but we tended to lose selectivity. So we tried to use enzyme networks that we reached through different paths as the targets.

Agência FAPESP – Was the fundamental change identification of new targets?
Blundell – The identification of new targets was one step. Another step is the manners we discovered to explore the chemistry with the knowledge of these targets. What we have done, fundamentally, consists of thinking about new screening strategies to work with a smaller number of compounds and in how to use the knowledge we have about targets to elaborate these compounds. This tends to be a much more efficient manner of obtaining selective molecules.

Agência FAPESP – Is this the so-called fragment-based approach to new drug discovery?
Blundell – Yes, it is a methodology that we have been trying to develop in our company, Astex, since 1999. We use X-ray crystallography in different manners, either to analyze the structure of biological targets or to select molecule fragments similar to drugs that are compatible with the formats of the active areas. We also use the technique to screen fragments in order to select initial “batteries” that are later transformed into high-quality lead compounds with the application of knowledge from the computational and medicinal chemistry.

Agência FAPESP – There have been major and rapid technological advances in the last few years in areas such as genomics. Is technology more and more necessary to extract knowledge on the enormous quantity of data generated by this type of progress?
Blundell - We increasingly need new computational methods but also new experimental methods. In fact, we have a significant volume of accumulated information due to more enhanced computational techniques, but this should not be seen as part of the problem, rather part of the solution. Although we having plenty of information, it does not help to ignore the problem. There is precious data there and we have to discover how to focus on more selective targets. We must assume that we will have to deal with all this information. We can use it in a very constructive manner to identify targets and also to optimize the processes and find new chemical products. I see computation as useful both for managing data and seeking hypotheses. But I do see the need to combine experimental and computational methodology.

Agência FAPESP – Do you think developing new computational methods should be a priority?
Blundell – No. Computational methods are important but are far from being the most important factor. In addition to developing new computational approaches, there is a need for developing new experimental methods. The two must go hand and hand. We cannot understand the genome without experimental methods, and we cannot generate data without the computational methods. The same occurs with the transcriptome, the proteome and the metabolome.

Agência FAPESP – What’s most important, then, is to establish a balance between these two sides?
Blundell – The main thing is remembering that more and more science is essentially multidisciplinary. If you were to go to my laboratory, you would see that I work with researchers from the medical area (clinical and experimental), computational and biological areas. The objective is to, as a team, have the capacity to build new methods and from them new experiments.

Agência FAPESP – Do you consider it a challenge to work with scientists from other areas?
Blundell – It is much easier to work with people from other areas. When you bring together many people from the same field, the individuals become competitive and even argumentative. But if you have a researcher in medicine, another in computation and one in biology working together, everyone has something to learn from the others. In my laboratory, we don’t have any fighting. This is because everyone depends on the other person's knowledge and must have mutual respect. We have to understand each other in order to interact, to work together and to build a multidisciplinary approach.