by Mathias Uhlén Royal Institute of Technology Stockholm Why is protein analysis going to be so much more difficult and protracted than the 15 year genome analysis? One reason is simply a matter of numbers.
Proteomics is the art of analysing all of the proteins in a living cell. From a medical point of view the idea is to identify those which are important to a person's health and well-being, in the hope that medical science might then be able to rectify any aberrations.

Analysis of the human genome, now essentially complete, is the first step towards this much bigger, more complex exercise, which was given its name only in the mid 1990s, when the human genome project was well under way. If I can offer an analogy with the telephone directory, the genome exercise is providing us with a catalogue – the instruction book, as James Watson calls it. Proteomics will provide us with an understanding of the contents of that catalogue.

What we need to know is the precise function of the gene products our genome analysis is revealing. It’s simply a prelude to the real scientific challenge: what the myriad of proteins present are there for, and how they work. I predict that this problem will still be a challenge for scientists 100 years from now. But that said, the next decade will probably be the most interesting yet in all the history of the life sciences.

In the Department of Biotechnology at the Royal Institute of Technology in Stockholm we are at the heart of a worldwide quest for genome understanding. The Institute is the oldest and largest technical university in Scandinavia. I am part of a team of about 45 researchers tackling many disparate bioscience problems, such as the causes of atherosclerosis (narrowing of blood vessels) and the molecular biology of the tree. Can we grow better trees?

But 25 per cent of our effort is devoted to the development of instrumentation for analysis of the problems in molecular biology these research challenges throw up. The same technologies are needed for both.

As a chemical engineer, I have been involved in the development of tools for unravelling and analysing the components of living cells throughout my career, both in industry and academia. We now have good tools for genome analysis, capable of analysing rapidly and reliably and verifying their findings. You have probably seen pictures of the big laboratories engaged in the human genome project, with ranks of these instruments – robots, really – backed by huge computing power, much like a modern factory.

Why so difficult?
Why is protein analysis going to be so much more difficult and protracted than the 15 year genome analysis? One reason is simply a matter of numbers. The genome project is now discovering new genes at the rate of about 300 per day, and will continue to do so for the next three years to complete the catalogue. But most genes give rise to a multitude of proteins, any one of which might be the cause of a problem. A second reason is that separating proteins seems to be inherently more difficult than separating genes. There are sticky proteins, electrically-charged proteins and other characteristics like the way they are folded that are making them difficult to unravel.

A third reason is a membrane barrier in the living cell, enveloping about 20 per cent of the proteins present. It is incredibly difficult to penetrate this barrier without distorting its contents. And the contents are proteins of particular interest in drug development.

During the lifetime of the human genome project our rate of analysis accelerated at least a hundred-fold. We cannot yet see ways in which we will make this happen for proteomics. We still rely heavily on the skills of our researchers, rather than automation. We have no equivalent yet of PCR technology for amplifying the effect of proteins that appear in very small quantities.

Worldwide challenge
But we do not despair. For one thing, this is a challenge for science worldwide, occupying some of the best scientific and engineering talent. In Stockholm we have high hopes for our new technology involving affinity-enriched antibodies. For another, there are complementary technologies, which have progressed rapidly during the decade of genomics. One is bioinformatics, computer techniques which can provide the researcher with guidance on experimental conditions for unravelling, based on computer analyses of groups of proteins.

A novel challenge is also emerging with realisation that the proteome is a dynamic entity, responsive to its environment – for example, a pathogen to which it is exposed. Characterising our half-million or so proteins usefully for biomedical research and the eventual development of treatments will also have to take their environmental responses into account.