The 2009 Barnett Lectures

Advances in Proteomics for Biomarker Discovery and Validation

Richard D. Smith

Richard D. Smith presented the 2009 Barnett Lecture on Sept 23’rd. Dr. Smith is a Battelle Fellow and Chief Scientist in the Biological Sciences Division, and Director of Proteomics Research, at Pacific Northwest National Labs (PNNL) in Richland, WA.

Dr. Smith has strong credentials in technology development, "but although there are a lot of interesting 'boutique' techniques that make good publications, many aren’t in practice useful or robust for clinical studies of hundreds of samples.” In practice, the sensitivity to do a “deep” biomarker search – detecting low-level proteins – is often compromised in high-throughput methods required in a broad study, assaying many patients, reproducibly.

Traditionally, proteins are identified not by mass alone but by partial sequence information from MS/MS spectra. Dick's group has furthered the development of "Accurate Mass and Time" tags – They first compiled a database of 8000 peptides, sequenced and assigned from MS/MS data acquired over many (50-100) LC-MS analyses of standard or control sera. In subsequent analyses of precious clinical samples, peptides can be rapidly and confidently assigned by reference to the accurate mass and elution time (AMT) of the well-characterized features, without acquiring or processing MS/MS spectra for each.

Attention then turns to handling the exceptions. Any resolved peaks without an AMT tag in the database can be prioritized for MS/MS, and his group has developed an approach of Unique Sequence (US) Tags to identify proteins and their modifications, with lower ambiguity, and support broad quantitative measurements using AMT. It is based on using high accuracy MS/MS spectra to identify amino acid sequence segments unique to proteins in genome.

The principal remaining problem is detecting unresolved LC features. A second separation dimension can be introduced within the mass spectrometer itself using ion mobility spectrometry (IMS). Ion mobility is not completely orthogonal to MS – the peak capacity is only 10 to 15 – but it is sufficient to resolve most of the coeluting compounds, and significantly reduce chemical noise. In a realistic test of standard peptides quantitatively spiked into human sera, LC-IMS-MS had more than an order of magnitude lower limits of detection than LC-LTQ-FTMS.

A useful technological contribution of his group is exceptionally bright ion sources, made by using ten or more nano-ESI tips in parallel, spraying through a ten-pinhole MS inlet into two ion funnels in series to capture the high ion flow. This provides a 50-fold increase in brightness, and 15-fold lower signal-to-noise in the resulting spectra. The additional signal strength greatly facilitates the most sensitive MS methods, of single-reaction monitoring and multiple reaction monitoring (SRM and MRM).

With these innovations, the bottleneck in clinical proteomics studies is no longer the time nor the expense of mass spectrometry. The ball is now in the court of sample preparation.

more pictures

by Roger Kautz, 23 Sept 2009

| Home | News | Members | Research Programs | Research Groups | Instrumentation |
| Publications | Jobs | Events | | Collaborations | Suggestions and Comments |