PROTEOMICS FEATURE: The promise (and problem) of proteomics

Proteomics. It's often touted as being the next revolution in biotechnology, the key to understanding biological pathways and disease states, a force that is driving research forward with an explosion of new technologies. But is identifying and describing the proteins of the human body providing the answers scientists and drug companies are after?

There is no doubt that the technology is progressing. Protein separation, for example, now includes chromatographic or immunoaffinity techniques that can detect proteins present in low amounts and simultaneously separate them from abundant proteins. Mass spectrometry instruments can now undertake peptide analysis and peptide sequencing. Protein chip technology enables protein recovery in conjunction with mass spectrometry.

With suggestions that there may be thousands of proteins in human blood serum alone, not to mention variations in these proteins, researchers certainly have their work cut out for them.

Australia has a solid base for proteomics research -- perhaps not surprisingly, as the term was coined in Australia. So for a technology that has been around for about a decade, is it delivering on its promises?

"Not yet," says Dr Mark Raftery, senior research officer with the Bioanalytical Mass Spectrometry Facility (BMSF), based at the University of New South Wales in Sydney.

Raftery's team's work involves detecting proteins present in low levels as well as detecting post-translational modifications of protein, such as phosphorylation, glycosylation and different types of oxidation-type modifications. His team is involved in the pilot phase of the Human Plasma Proteome Project, and is finishing data analysis which it hopes will be of sufficient quality to give it a key role as the project moves forward.

"In terms of mass spectrometry we hope that [proteomics] will lead to new diagnostic tests, or new approaches to work out if somebody has a particular disease, by pointing at the right proteins to look at," he says.

One of the problems with proteomics is that it is very expensive technology. Raftery believes there is more room for improvement and new developments in proteomics technology, but concedes that the cost of this can be inhibitive.

"There are instruments called FTICR instruments which are capable of very high resolutions, much higher than are currently available within the Sydney area. So there are developments in mass spectrometry which may lower the detection limits, but that comes at quite a dollar cost."

Raftery also admits the huge amount of data generated by these experiments can be problematic.

"Data analysis is a problem for everybody, and that's where some developments could happen," he says. "For example, for each one of these runs you may get 5000 or 10,000 spectra, the raw MS data. That then has to be turned into peptide sequences and then those peptide sequences need to be turned into protein identities, and you want a protein ID, or many proteins IDs, at the end."

Still early days

"I think proteomics is now starting to deliver," says Ian Smith, head of the peptide group at the Baker Heart Research Institute in Melbourne. "I think it's fair to say it's still early days. I think we've probably spent too long in the developmental stage, fine-tuning, and not moved it quickly enough into the biological labs."

Smith believes one of the problems with proteomics, as with any technique, is that there is too much of a focus on pushing boundaries to improve or make the techniques more sensitive.

"There comes a point where they actually have to move from the technology to address specific questions, and now we're starting to see that happen and results are starting to come through," Smith says.

Thanks to a grant from the Ramaciotti Foundation, the Baker Institute has a high-throughput proteomics system in place for handling large numbers of samples. This system supplies proteomics analysis for researchers at the Baker Institute, plus a number of other sites including Monash University.

Through this set-up, Smith says, he is beginning to witness biologists -- who are not necessarily protein chemists -- starting to come to terms with what sorts of questions they can ask and what sorts of data they can expect to generate from a state-of-the-art proteomics platform.

One of the problems researchers come up against in proteomics is not so much in finding protein candidates, but working out which protein to choose to try to validate as important to the biological process of interest. The involvement of researchers with knowledge of the biological system under study is being acknowledged as necessary to help identify these protein targets.

Some of the results which Smith says are coming through involve functional proteomics, such as identifying which proteins are interacting with a specific drug, enabling researchers to get a handle on the pathways in which the drug may be operating. The Baker team is using proteomics to look at how different proteins circulate within a cell and what proteins are involved in this process. One application of proteomics is in the process of screening patients for diseases, such as heart disease or diabetes. By comparing disease blood or urine samples with control samples, changes in proteins can be looked that could be biomarkers that may then lead to a diagnostic.

"If we can find out what these markers are, by using expensive proteomics, then we may be able to develop some cheap assays for a particular protein, a diagnostic," Smith says.

In the pipeline

Stuart Cordwell, senior research scientist at the Australian Proteome Analysis Facility (APAF) in Sydney, thinks proteomics technology is making a big contribution to research, although he says time will tell as to how big that contribution will ultimately be.

"I don't know at this stage whether we'll know the full implications of it for some years, because obviously proteomics is big in a number of drug development pipelines," Cordwell says. "And I think it's going to be some time until these things actually reach the market."

Cordwell sees proteomics as a tool that needs to be used alongside biological experiments so that relevant targets can be found amongst the data generated.

"Proteomics is something that needs to force further experimentation and validation," he says. "And I think that making that connection at the moment is where a lot of people are struggling. A lot of people are doing the 2D gels or the liquid chromatography and getting a list of proteins and while that's fantastic, then which protein do you work on next? How to go back biochemically, create knock-outs, mutagenise?"

Cordwell admits that, even with the new technological developments of which APAF has recently taken advantage (its Macquarie University node recently installed three new mass spectrometers), there is a long way to go.

"Knowledge of the biological system aids in moving to target validation," said Cordwell. "At the same time knowledge of the system may lead to passing up hypothetical proteins you don't know anything about. It's a long road."

Developing the technology

"I would say proteomics hasn't delivered as much as people would have expected," says Ed Nice, of the Ludwig Institute for Cancer Research in Melbourne. "But I think that's only because we are still developing the techniques and the technologies to their optimum, and I think we'll be able to do that."

Nice's lab is researching the processes associated with the signalling pathways of the adenomatous polyposis coli (APC) protein and how these become disregulated in cancers. The Ludwig's researchers know there are a large number of mutations in the APC gene which are strongly correlated with colon cancer, and are aiming to understand where the signalling is going wrong and why APC no longer acts as a tumour suppressor. To do this, they have pulled out the APC protein, plus the proteins it complexes with, and are looking for differences between the normal and tumour situation.

Nice places emphasis on improvements in protein separation as fundamental to driving things forward. "Already that is starting to happen, if you look at the publications that are coming out now," says Nice. "[Researchers] are starting to use multidimensional chromatography, either for looking at peptide mapping or for purifying the proteins. I think it's going to be important to get the proteins separated, otherwise you start losing a handle on the quantitation of proteins."

Even with the increasingly precise separation techniques on the market, large numbers of proteins are not detected -- groups of proteins do not run on gels and others get stuck on chromatography columns, making them invisible to the researcher.

But, Nice asks, is there a point in knowing exactly what every single protein in a cell is?

That question goes back to the relevance of knowing which proteins are undergoing changes in their concentrations, as these are the ones of interest. But homing in on these can be difficult.

"There is no doubt that with the amount of data one can generate there are still problems in the downstream analysis of that data," Nice says. "And even though we've moved forward, low abundance proteins can still get lost in the noise in the conventional kind of data handling process."

Protein science revisited

With some of the best protein experts in the world based here, Australia is well placed to be at the cutting edge of proteomics technology.

"The technology really is moving at a very rapid pace," Smith says. "In three years' time some of the equipment we are using will be out of date, and I believe that over the next three years we will start to see some delivery on the promise. And it's starting to come through now. We are starting to see journal publications using proteomics."

At present, whether proteomics can be considered to be delivering on its promise remains in the eye of the beholder. Results are emerging in both diagnostics and drug discovery, but the full extent of the promise -- which is racing against the cost and complexity of the technology required -- may not be disclosed for some time.

Others believe protein science has always been delivering. As the Ludwig's Ed Nice comments: "If you take a step back I think some people will say that proteomics is nothing new -- it's actually what we've been doing for years, protein purification."