Proteomics feature: The new biodiversity

What began as a trickle of new genes in the late 1970s has become a flood, as genomics projects deliver new genetic maps, huge catalogues of genes, and a ticker-tape blizzard of DNA sequences almost every other week.

Bioscience researchers around the world now have access to a wealth of genetic data on viruses, bacteria, parasites, plants and animals. Now comes the hard part: proteomics, a bold new science that seeks to understand how the protein products of all those genes conspire to assemble and operate Life as We Know It: animals, plants, fungi and microbes.

Proteomics is concerned with identifying the components of life's molecular machinery, determining how they are assembled -- the challenge is that the components, and the machinery, differ for every individual.

Geneticists believe that most of the differences between all six billion people on Earth stem from around 7000-odd single nucleotide polymorphisms (SNPs), scattered around the genome. Two variants of the same gene, differing by a single nucleotide -- one 'letter' of DNA code -- may function very differently. A change in the volume of a gene's protein product, or a subtle change that alters the structure and function of a protein, can have complex, ripple-down effects on dozens of other genes, that cannot be predicted from DNA code alone.

Proteomics deals with this complexity by ignoring the DNA source code, and focusing on the end-products of gene activity: the distinctive protein patterns generated by the unique genetic activity in each individual. Proteomics is a technology-intensive, multi-disciplinary business, a hunt for biological treasure buried in the genomes of humans, animals, plants and microbes. The treasure map is a small, square sheet of acrylamide gel featuring a pattern of protein spots.

Two-dimensional gel electrophoresis (2DE) sorts a complex cocktail of proteins from biological samples into a two-dimensional pattern, according to each protein's specific combination of charge and molecular weight. 2D gels are the gold standard for protein discovery -- the same proteins will consistently migrate to the same points on the gel. The skill lies in detecting anything unusual -- spots that are enlarged, reduced, absent, or which appear in unexpected locations.

When genome projects began to deliver vast amounts of genetic data, around four years ago, there was a rapid shift from largely manual methods to high-throughput automated systems. Computers and robots now prepare gels, run the protein extracts from tissue samples, produce digital images of the resulting protein patterns, excise proteins of interest, and prepare them for analysis in mass spectrometers.

Expensive science

Proteomics is an expensive business, according to Dr Gary Cobon, director of the Australian Proteome Analysis Facility (APAF) in Sydney. APAF has been providing proteomics services and training for the biotechnology and biomedical research industries for six years. Its 250 clients and collaborators throughout Australia come from all the major universities, teaching hospitals, CSIRO, eight of the 13 Cooperative Research Centres (CRCs) involved in biotechnology research, state departments of agriculture, and 30 private companies.

Thirty per cent of APAF's business comes from overseas. Since it was spun off from Macquarie University in 1996, APAF has built an international reputation. Cobon says it takes time to develop expertise, and APAF researchers have worked on organisms as diverse as cows, plants, fungi and spiders. It remains the only centre in Australia offering a full package of proteomics services. APAF's reputation attracts a steady stream of PhDs and postdoctoral researchers from overseas and they bring their expertise with them.

Companies that develop proteomics devices and products use APAF as a beta-testing centre -- "They like to involve us, and it gives us a good idea of what works and what doesn't, so we can be more efficient with our expenditure," says Cobon. The cost of staying ahead of the game is very high, says Cobon -- a new high pressure liquid chromatograph (HPLC) coupled mass spectrometer costs at least $1.5 million (HPLC sorts proteins while preserving their native activity).

APAF has just received a Major National Research Facility grant worth $16.25 million to establish three new interstate nodes for Australia's largest and best-equipped proteomics centre. The new nodes, at the University of NSW, Sydney University and TGR Biosciences in Adelaide, will contribute technologies complementing those offered by the Macquarie University node, widening the range of client services. Cobon says that much of APAF's work involves proteins that are typically present at around 1000 to 2000 copies in cells, but clients are increasingly interested in functional proteomics: protein-protein and protein-substrate interactions, or bioactivity of low-abundance proteins that switch genes on and off -- cells may contain as few as 10 to 100 copies of these proteins.

Clients are also interested in proteins that have been modified by processes such as phosphorylation and glycosylation. Such work requires a 10 to 100-fold increase in mass spectrometer sensitivity. Cobon nominates bioinformatics as a bottleneck zone -- proteomics generates huge volumes of data; the challenge is to devise systems for storing and retrieving data efficiently, and presenting it in forms that researchers can easily interpret. It's a formidable problem, says Dr Marc Wilkins, vice-president of bioinformatics with North Ryde, (Sydney)-based company Proteome Systems. The pattern of proteins on a gel combines positional and quantitative information -- the human eye and brain aren't up to the task of comparing 200 to 400 gels and extracting the significant differences between them.

"Graphic representations are the key -- we have to get away from using 2D gels as the primary means of representing information," he says. Another problem is to develop graphic representations of protein-protein interactions. A gel is a two-dimensional pattern of protein spots, but it takes a 3D diagram to represent the network of interactions between collaborating proteins. Wilkins calls it "an interesting problem in visualisation".

Bio-IT involvement

He says Proteome Systems established a global strategic alliance with IBM 12 months ago to develop an enterprise-level bioinformatics infrastructure combining a laboratory information management system and electronic lab book. This bioinformatics infrastructure is at the heart of a technology platform , called ProteomIQ, which is modular and scaleable, and can be installed at any greenfield site. Proteome Systems works with clients to establish a fully functioning proteomics lab, using automated tools to prepare and process samples, and identify and characterise proteins of interest, with minimal human intervention.

Researchers can manage the automated equipment via a local area network, linked to international gene and protein databases, allowing researchers to identify proteins and associate them with familiar or as-yet unidentified genes.

Two of Melbourne's major biomedical research agencies -- the Walter and Eliza Hall Institute (WEHI), and the Ludwig Institute for Cancer Research -- have pooled their resources and expertise in a shared proteomics research laboratory, the Joint Protein Research Facility (JPRF). The deputy head of the JPRF, Dr Rob Moritz, says the scale and diversity of research at WEHI and the Ludwig would make it too expensive to contract out proteomics work to a commercial operator like APAF. APAF's standard services might also be ill-fitted to the special requirements of both institutes.

Moritz says JPRF performs most of the basic proteomics work, from identifying protein spots on gels through to sophisticated protein quantitation using radioisotopes. "We also have a research program into targeted protein purification, using novel techniques that aren't yet available commercially," he says.

WEHI and the Ludwig have five mass spectrometers, capable of performing virtually every permutation of protein analysis, from identifying whole proteins, to protein sequencing. Melbourne's Baker Research Institute, Australia's leading research centre for cardiovascular disease, also operates its own in-house proteomics facility, but does contract work for biotech companies and other research agencies.

Academic clients and collaborators like the Macfarlane Burnet Institute, the Australian Centre for Blood Disorders, the Red Cross, and the Monash Institute of Reproduction and Development, are charged only for reagents, but private companies pay commercial rates, and the money goes back into keeping equipment up to date.

Robots and proteomics

Dr Ian Smith, associate director at the Baker Institute, says the institute reduced the capital cost of the facility by negotiating a deal with Amersham Biosciences to install a fully integrated, robotic proteomics system. "They wanted a representative in Australia," he says. "Under our strategic alliance, they upgrade our software and hardware, in return for the right to bring customers from elsewhere in Australia or Asia to inspect our instruments, and to watch us run samples. "Without the deal, it would have cost us $1.5 million to put the system in place.

"We have a very high-throughput capability," Smith says. "We can identify up to 1000 proteins in a 24-hour period, by loading up to a dozen 2D gels into the system. Robots cut out the protein spots, digest them out of the gels, and load the plated proteins into our mass spectrometer."

The mass spectrometer is used both to determine the absolute mass of proteins of interest, and as a prelude to deriving peptide sequences that can be matched against sequences in international protein and genomics databases to identify the source genes. Smith says the Baker's research is very much clinically driven; the institute has access to very large numbers of tissue samples, both of post-mortem and surgical origin.

The Institute is interested in Ciphergen Biosystems' new ProteinChip system -- the technology recently used by US researchers to identify an elusive factor called "CAF" that prevents a small number of HIV-positive individuals progressing to full-blown AIDS (see Australian Biotechnology News, October 11). Smith describes ProteinChip as "very exciting technology", and an ideal accessory for a tandem mass spectrometer. He believes it would be particularly useful for identifying protein markers associated with cardiovascular disease.

It could also be used to profile patients and determine their responses to different medications. A particular pattern of serum proteins might indicate, for example, that a patient should be prescribed beta-blockers in preference to ACE inhibitors to reduce high blood pressure.

"If we're screening for early indications of congestive heart disease, we need to identify protein markers associated with the disorder," Smith says. "If we find an aberrant protein, we can look at the way it is expressed in different patients, and then go to animal models to see if we can switch the gene on or off, or inhibit its activity."

Enter the synchrotron

The biggest and brightest tool for proteomics research is beyond the financial reach of any private company or consortium of research institutions in Australia -- the capital cost of is a stretch even for governments. Construction of Australia's first synchrotron will begin on a site opposite Monash University in the eastern Melbourne suburb of Clayton next year. The $150 million, ring-shaped device will accelerate electrons to very high energies to generate brilliant X-ray beams to probe the molecular structure of a wide range of crystalline materials, including proteins.

The molecular shape of a protein is intimately associated with its function. Armed with information about how proteins work, chemists can design molecules that mimic or block the activity of target proteins or enzymes with exquisite selectivity. For decades, Australian bioscience researchers have been renting beamlines on synchrotrons in Japan and the US. The Victorian government's decision to fund construction of a synchrotron, in partnership with research institutions and industry, is expected to yield huge benefits for biomedical and biotechnology research in Australia and New Zealand.

Monash University's Professor of X-ray and Synchrotron Physics, Rob Lewis, this week described synchrotrons as "probably the most versatile research tool on the planet". Lewis says the Australian synchrotron will be world-class, and will have a major impact on "just about every area of science you care to name". Australian biomedical research will be a major beneficiary from the synchrotron project -- "The biotechnology revolution is happening because synchrotrons exist -- they make previously unimaginable research feasible," Lewis says.

NZ set-up makes its own way

Where Australia's APAF is subsidised by government grants, New Zealand's Proteomics Discovery Centre Ltd is a private company, serving New Zealand's biotechnology industry on a full cost-recovery basis. Scientific director Dr David Cottle says Proteomics Discovery Centre is a spin-off from the New Zealand Wool Research Corporation

The company recently purchased a $NZ1.6 million quadrupole mass spectrometer; it can operate in three different modes for extremely precise determinations of protein masses, as well as for deriving peptide sequence data for gene identification. "It can determine a protein molecule's mass to four decimal places, so it gives us a fantastic ability to differentiate between molecules that are very close together in mass," Cottle says.

"For example, it can distinguish between glycosylated and a phosporylated forms of the same protein, which gives us an idea about what a protein is doing." Cottle says Proteomics Discovery Centre emerged from research into wool fibre and protein chemistry and it maintains a strong research program in this field. It has recently used peptide-mass fingerprinting to identify some of the high-sulphur proteins from which the wool fibre is constructed. One of the proteins appears to be associated with crimp -- the wavy pattern in fleece that gives wool its 'springiness'.

Crimp is a desirable property in carpet wools because it enhances under-foot comfort, resilience and resistance to wear. It's also valuable in very fine merino wools, because it traps air, providing warmth without increasing weight. When the 'crimp' protein is linked to the source gene, breeders will be able identify animals carrying the gene with a DNA test.

As well as its own research, Proteomics Discovery Centre does contract research for its commercial clients. "We can't compete with APAF on a protein-by-protein basis, but companies in the wool and hair industry in the northern hemisphere recognise our specialised skills in wool research -- we have a major client in the cosmetics industry, whose interest is in the composition of human hair," Cottle says.