Lorne Proteomics: Getting more out of proteomics

The Lorne Proteomics Symposium is geared towards assessing practical solutions for the technology.

As well as shedding light on the expression levels of the many proteins found within organisms, proteomics research is leading to the manifestation of synergies between other fields. Cell biologists and protein chemists, for example, are getting together to assess the proteome of cells and intracellular organelles.

But that's no simple task, because the further away from the DNA sequence of a gene the more complex things become. Detecting post-translational modifications, low abundant proteins and transmembranous proteins for example, means protein chemists have their work cut out for them.

The technologies used to analyse proteins are what makes the work possible and the Lorne Proteomics Symposium, hosted by the Australian Proteomics Society, focuses on developments in the core proteomics technologies as well as new emerging technologies in protein chemistry.

"The proteomics meeting is more of a 'what works down at the bench' type of meeting," says Dr Robert Moritz, who chairs the program committee. "Companies announce the latest versions of their instrumentation at this meeting, and because it is a small and isolated meeting they can gauge what sort of response the new instrument will get from the audience and not generate too much hype too soon."

The Australian Proteomics Society is the umbrella organisation for AOHUPO -- the Oceania-Asia arm of the Human Proteome Organisation (HUPO). This year the proteomics meeting is sharing its two overseas speakers with the Lorne Protein meeting -- "that's why we can afford to have two," says Moritz.

Prof Yaeta Endo, from Ehime University in Japan -- whose work Moritz describes as "proteins' answer to PCR" -- will be talking about large-scale, cell-free protein expression systems. "There is a big initiative to generate proteins from every gene in the human genome, and this is being pursued with different approaches to expressing proteins," says Prof Richard Simpson, from the Ludwig Institute in Melbourne, director of the HUPO finance committee vice-president of the organisation's new technologies and resources committee. "The cell-free expression system is one of these."

This work ties in with the HUPO antibody project, which aims to make antibodies to all the proteins in the human genome. Three groups worldwide, Sweden, Japan and Taiwan, are pursuing this independently and will form a consortium at some stage in the future and compare data.

Protein profiling

The other overseas speaker at Lorne Proteomics this year is Prof John Yates, of the Scripps Research Institute in the US, who will discuss multi-dimensional proteomic analysis by mass spectrometry.

"Yates is doing protein profiling at a high level," Simpson says. "He is using high-power equipment to profile proteins and the metrics are huge. Without this high-tech equipment it would take one person six months or more to determine one phophorylated site on a protein. Yates' team can look at all the phosphorylation sites on many proteins at once."

Yates' team performs rapid and large-scale proteomics analysis using multidimensional protein identification technology (MudPIT), which involves multidimensional liquid chromatography, tandem mass spectrometry, and database searching using the SEQUEST algorithm -- a sequencing program Yates' team devised for MS analysis that searches a sequence database by mass.

"We use this one general approach for proteomics," says Yates. "It is called shotgun proteomics, and the method minimises fractionation of proteins by analysing the proteins as a proteolytically digested mixture."

His team has applied MudPIT to the proteome of a strain of the yeast Saccharomyces cerevisiae, and has reportedly yielded the largest proteome analysis to date: the detection and identification of 1484 proteins, including low abundance proteins such as transcription factors and protein kinases, which are rarely seen in proteome analysis.

They also identified 131 proteins with three or more predicted transmembrane domains, enabling the soluble domains these integral membrane proteins to be mapped.

Post-translational modification of peptides and proteins is another area that Yates' team is exploring. "There are a couple of ways people do these types of analyses," says Yates. "We use a method that tries to achieve a high level of sequence coverage to ensure we collect a tandem mass spectrum for modified peptides if present. By this method, different modifications can be detected. Once a tandem mass spectrum is collected it's all data analysis."

Yates says the software or algorithms that his team uses to interpret their proteomics data are very accurate, as long as the data is good quality -- as data quality decreases the accuracy of an algorithm can also decrease. He says it is often advantageous to use several different algorithms, since different algorithms can have different sensitivities and selectivities.

"Since we use tandem mass spectra, we use algorithms that search databases using tandem mass spectra," says Yates. "They are unaffected by [post-translational] modifications, but you do have to specify the modification to most of the search algorithms. Algorithms are appearing that can identify peptides with unanticipated modifications."

Yates sees the future as bright for proteomics research, as technology continues to improve and provide benefits. "The hot item right now is the Fourier transform mass spectrometer," says Yates. "The benefit for peptide analysis is very high mass accuracy."

And he sees a place for this technology for a while to come, pointing out that protein chips are only as good as the available knowledge about proteins. "There will be protein chips, eventually, but creating affinity-based protein chips has some serious technical challenges to overcome," says Yates. "I don't think protein chips will completely replace the protein profiling since protein chips can only detect what you know or can predict."