23. December 2008 09:05
Computers team up with evolution to design novel enzymes
The dream of designer drugs highly specific in action with zero side effects has come closer by combining atomic level computer prediction with natural selection in the laboratory. Following the 2008’s first design of an artificial enzyme for catalysing a chemical reaction, there is increased collaboration among researchers to accelerate progress towards this goal. The stage for closer collaborative work between computer protein design specialists and experimental biochemists was set at a recent conference organised by the European Science Foundation (ESF).
The conference drew inspiration from the great achievement in the field during 2008 in designing an enzyme from scratch that speeded up a well known reaction called Kemp Elimination by up to 100,000 times. In Kemp Elimination, a proton is extracted from a carbon atom, in a reaction for which the enzyme is not actually found in Nature but involves the element upon which life is rooted. It sets the stage for the future design of enzymes that have applications in production of drugs and also non-medicinal branches of biotechnology, such as brewing and food processing, according to Jiri Damborsky, chair of the ESF conference, from Masaryk University in the Czech Republic.
“The way to move forward involves a mix of rational design and directed evolution,” said Damborsky. The process may start with computer models to generate a molecular design for a protein with structure and function as close as possible to a desired target. Although great progress has been made with such de novo design, current models do not take account of some of the finer physical aspects of molecules and their behaviour, such as dynamics and bonding with ions in solvents, and also the exact geometry of the structure. Differences in scale as small as a few picometers (that is 10-12 metres, or 1/100 the diameter of a hydrogen atom) can make a significant difference in the action of an enzyme.
So even though a computer can design a protein with atomic precision, it still has to be fine-tuned, which can only be achieved in the laboratory. This is why de novo design of proteins requires collaboration between experimentalists and theoreticians.
Next comes a two step process to refine the protein that has been designed by computer or other means. The first step is to mimic evolutionary mutation to create a diverse range of closely related molecules in the hope that one will have the right structure. This process, known as directed evolution, creates many variants of the original protein. The second step is then to apply selection or screening to this collection, in order to identify the molecule closest to the target. “Screens enable the researcher to identify and isolate high-performing mutants by hand, while selections automatically eliminate all non-functional mutants,” said Damborsky.
Whichever method is used, it is still unlikely that any of the mutants produced in the first round of directed evolution will yield a protein that closely matches the target, in other words that has all the desired properties. Therefore, this two-step round of directed evolution is repeated several times until no further improvement can be made. “In these experiments, the ‘winners’ of the previous round are diversified in the next round to create a new library,” said Damborsky.
Even after a number of rounds, the perfect molecule cannot usually be made at present, since further progress on various fronts is still required. Research needed to accomplish this was discussed at the ESF conference. One problem is that even when an apparently successful mutant has been created, it may not be stable. The result is that the protein does not fold properly, becoming stuck perhaps in some half way conformation and being unable to reach its final structure. Fortunately nature has had to solve the same problem during evolution, through molecules called chaperones that assist with the folding process, helping overcome kinetic obstacles on the way to reaching the final functional conformation. Damborsky pointed out that chaperones were being used in some experiments to stabilise mutant proteins and enable them to fold properly into their functional structure.
However progress is also required on other fronts identified at the ESF conference, including understanding the link between structure of a protein and the functions it has, and also how enzymes move around inside the cell. The first of these is needed to design proteins that have the right structure for a desired function, and the second to make drugs that reach their target efficiently. In general more needs to be learnt about how enzymes operate in cells, and the detailed mechanisms involved in biosynthesis of proteins and their subsequent movements or interactions.
Despite the recognition that much more research was needed, the conference sounded an upbeat note with hopes of further achievements in the immediate future. Indeed the agenda set for a follow up ESF conference included artificial life, creating novel organisms by mimicking selective processes of nature on a larger scale. This follow up conference will also pursue the goal of designing the ultimate artificial enzyme capable of catalysing biochemical reactions almost at will.
The ESF Conference Protein Design and Evolution for Biocatalysis was held in Sant Feliu de Guixols, Spain in October 2008.
For more information please click here.
Media contact:
- Mr. Thomas LauE-Mail
Science contact:
- Professor Jiri DamborskyE-Mail
