The race to unlock the proteome

Scientists are ten years away from really understanding how proteins work Keystone Archive

Mapping man's complex network of proteins is the future of medicine. But how soon will proteomics-derived drugs be on the market?

The challenge facing proteomics is one of the topics being discussed at BioData 2002, a major symposium for the biotech industry in Geneva. The largest gathering of its kind in Europe, it has attracted representatives of companies large and small, academics and, crucially, investors.

Geneva is an ideal place to hold such a conference. It is home to some of the most important private and public biotech research facilities in the world, as well as the world’s first large-scale proteomics firm, Gene-Prot.

The fanfare that greeted last year’s publication of the complete human genome was tempered by the knowledge that, in itself, it did not hold the key to unlocking diseases. While a useful blueprint, it was little more than a gateway to an even more complex area: the proteins that actually make living cells work.

Bewildering variety

The task facing researchers is clear: there is only one, unchanging, genome, but a bewildering variety of proteins per individual.

“Today we don’t even know how many proteins there are in an organism,” says Friederich von Bohlen, boss of German biotech company LION Bioscience.

Proteomics involves isolating, identifying and explaining the behaviour of the proteins present in tissue samples. A given sample can contain at least 100,000 different protein molecules, and a person’s proteome – or protein profile – is changing constantly throughout their lifetime.

“We are very much at the beginning. It took 16 years to sequence the genome. It’s going to take us another five to ten years before we really understand how proteins work,” von Bohlen told swissinfo.

Integrated approach

He foresees a more integrated approach to drug discoveries, believing that proteomics will not dominate the field in the near future. But that does not mean that steps are not being taken along that path.

GeneProt has already identified six potentially therapeutic proteins that it has passed on to its partner company, Novartis. The Swiss pharmaceuticals giant has decided to carry out further research into three of these this year.

This is merely the tip of the iceberg. There are those who believe that proteomics will allow doctors to predict a patient’s susceptibility to certain illnesses and take preventative action. A simple blood test could be used to screen a person for hundreds of diseases.

Currently, our drugs are developed in a haphazard way, with thousands of chemicals being tested to see which interact best with our molecular make-up. With a map of the genome to hand, and naturally occurring proteins to work with, scientists should need less trial and error to come up with results.

“The failure rate should be lower, and it should take less time to get drugs onto the market,” says GeneProt’s Chief Executive Officer, Cédric Loiret-Bernal. Whether that will bring down the price of drugs remains to be seen.

Paradigm shift

There are already a number of protein-based drugs available, though these tend to involve the creation of anti-bodies. Proteomics is more concerned with seeking a functional understanding of diseases.

“It’s a paradigm shift in how drug discovery should evolve. Proteomics give us the chance to treat the causes of a disease, rather than its symptoms,” says Loiret-Bernal.

“If there are over 30,000 genes and ten times as many proteins, there must be more therapeutic proteins out there than EPO and insulin,” he told swissinfo.

As in the race to sequence the human genome, the mapping of the proteome is being conducted by an array of private companies and public institutions. The copyrighting of some proteins has already raised concerns about access to crucial data

Loiret-Bernal, whose company has partnerships with academic institutions as well as big multinationals, believes competition will speed up the search for vital proteins.

“If the public effort to find the genome had not been challenged by the private effort, we might have had to wait another five years for this genomic information, which is a prerequisite for proteomics,” he says.

Supercomputers

Besides scientific knowledge, two of the key tools in proteomics are mass-spectrometry – which can weigh with minute precision the peptides created from proteins – and a massive computer system. These are essential for separating tissue samples, identifying and selecting proteins and, ultimately, synthesising them.

It is difficult to overstate the importance of information technology. To a large extent, progress in the biotech field depends on advances in computing — such is the huge amount of data that has to be sifted.

GeneProt is recognised as the world’s first “industrial-scale” proteomics firm. Thanks to its collaboration with Compaq, it has the most powerful supercomputer in the Life Sciences sector.

Not wanting to be left out of this bioinformatics boom, IBM recently announced it was planning to build an even bigger facility for the Canadian biotech company, MDS.

“This information has to be processed very quickly, and then you have to test your assumptions. Only supercomputing can do that,” Loiret-Bernal explains.

And what follows the proteome? Scientists have already coined the phrase: the metabolome – how the proteins interact to create a living cell. If we understand that, surely one day we might be able to replicate it and create artificial life. That is certainly one for the future.

by Roy Probert