Researchers at the University of Illinois have developed techniques that are said to facilitate the rapid identification and characterisation of proteins. The techniques may prove invaluable in characterising the human proteome.
‘New analytical methodologies are needed to identify the form and function of the hundreds of thousands of proteins encoded by genes,’ said Neil Kelleher, a UI professor of chemistry. ‘Part of the problem is retrieving intact proteins from databases using high-resolution, tandem mass spectrometric data and correlating their predicted structures with those actually present in mature proteins.’
Contemporary approaches to protein identification using mass spectrometry have involved the measurement of peptide masses, but the direct fragmentation of protein ions can, according to Kelleher, be far more efficient than exhaustive peptide mapping. ‘This is a new strategy for proteome analysis,’ Kelleher said.
Kelleher’s instrumentation reportedly combines Fourier-Transform Mass Spectrometry with electrospray ionisation and separation methods. At the heart of the system is a liquid-helium cooled superconducting magnet. A vacuum system and mass spectrometer extend into the magnet’s centre.
‘This is a relatively new breed of magnet,’ said Kelleher. ‘Instead of using 12 tons of bulky steel, the magnet is actively shielded with a counter-propagating magnetic field. The fields cancel one another outside the magnet, but at the magnet’s centre the field strength is a hefty 9.4 tesla.’
Fractionated proteins are squirted into the vacuum system and then transported into the magnet, where they begin to spin. ‘The proteins spin at different frequencies, depending on their mass and charge,’ Kelleher said. ‘We gradually excite their orbits to higher and higher radii, and they eventually fly past sensitive detector plates in the mass spectrometer.’
Computers then analyse the data to identify and characterise the proteins. The entire system is becoming increasingly automated for ease and efficiency of operation.
For their initial studies, Kelleher and his students selected two representative life forms: Mycoplasma pneumoniae – a simple bacteria with a tiny genome – and Methanococcus jannaschii – an archaeon found in submarine hydrothermal vents.
First, the researchers showed that multiple proteins could be processed simultaneously. Then they tested a predictive model for database search specificity.
The model agreed with actual searches from a database of about 3,500 protein forms predicted from the genomic sequence of Methanococcus jannaschii. The method also should work for the millions of possible protein forms predicted from the human genome.
‘These conceptual and technical advances provide a powerful tool for protein characterisation in the post-genomic era,’ Kelleher said. ‘By better characterising proteins, we can improve our fundamental understanding of the blueprint of life.’