Rapid proteome profiling using positional signature peptides

The proteome defines the entire complement of proteins expressed by a cell in a particular state. The 'protein world' is a challenging area to study, and if we are to conquer this world, currently we use a strategy based on 'divide and conquer', breaking up the proteome into smaller fragments to make them amenable to analysis. But, if we increase the intrinsic complexity of the protein world by fragmentation, aren't we making it even more difficult to analyse? At first glance, yes, but suppose we could capture just one, information rich fragment that was able to report on the parent protein - recovering information just became 50 times easier! This strategy for proteome simplification is exactly what we propose. We have worked out a way to reduce the complexity of a proteome (perhaps 500,000 fragments) to a highly information-rich subset (perhaps 5,000 fragments) using novel chemical tricks. We now need to develop and refine our approach, and build the matching informatics tools to make the most effective use of this limited set of fragments - there is a wealth of information buried in these. An approach such as this would be of great value to the community of scientists who study proteomes. The methodology is simple to deliver, cheap, and requires no sophisticated instrumentation over and above that which we would find in a typical proteomics laboratory. It has the potential to revolutionise the way in which we study proteomes

Global analysis of proteomes is a necessity for systems biology and for high throughput proteome screening. The most promising and readily available methods are based on peptide-level analysis, but the conversion of a proteome from protein space to peptide space increases the analyte complexity by a factor of 40-50 fold. This complexity continues to dog mass spectrometry-based approaches and presents a real bottleneck for truly genome-wide proteomics in a single experiment. We have recently published a novel approach to proteome simplification in peptide space, through selective isolation and recovery of N-terminal peptides (positional signature peptides, PSPs). This reduces analyte complexity to one peptide per protein, which greatly simplifies the protein identification problem (currently a bane to MudPIT style proteomics) since the unique peptides are usually diagnostic for the parent proteins. This also increases the depth to which a proteome can be characterised. Simplification of the proteome analytical challenge to the level of one peptide per protein brings within reach other aspects, including absolute quantification and determination of intracellular stability. However, it is now very clear that existing proteome software tools, predicated on the presence of more than one peptide for each protein, do not perform well with PSPs. Indeed, our preliminary studies suggest that many more true PSPs are buried in the spectra that are readily obtained, and improved informatic approaches can take this technology forward and deeper into the global proteome. Bespoke software and database searching strategies will be developed in this proposal in tandem with refined experimental procedures, making software available as open source tools. This will then be applied to demonstrator projects on the E.coli and serum proteomes to show how it may be exploited for rapid proteome profiling, making all data available via local databases and external repositories.