PNNL scientists (from left) Gordon Anderson, Keqi Tang, Mikhail Belov and Dick Smith
Pulling Secrets from Blood
June / July 2008
Volume 6 Number 3
Haystack. Needle. Imagine having a machine into which you feed batches of hay. Each piece of hay zooms past a gadget that studies its characteristics—golden, flexible, musty—€”and identifies it as "hay." And then that odd piece hits the spotlight—silver, shiny, pointy.
"Needle."
An interdisciplinary group of researchers at Pacific Northwest National Laboratory wants to make searching for that needle in the haystack simple and fast. Their haystack is our blood or tissues. In them reside proteins that can serve as early signs of diseases. The challenge is to find the disease-related proteins among the thousands of normal ones. The team of physicists, biochemists, engineers and computer scientists is building instruments to find those signs.
"This is very much a collaborative process among many PNNL scientists," says computer scientist Gordon Anderson. Most of the team's work is conducted at the Environmental Molecular Science Laboratory, a DOE national scientific user facility located at PNNL.
Amassing patents along the way, the team is modifying mass spectrometers and ion mobility spectrometers to be able to run at least 100 samples a day quickly and accurately enough to make the technology useful in a commercial clinical setting.
"A sample from blood is hugely complex," says mass spectrometrist Keqi Tang. "It's a complex world in a small volume."
The researchers work in the field of proteomics, which explores the biology of organisms by studying the proteins they produce. Leading the instrument development team is laboratory fellow Dick Smith. With more than half a dozen R&D 100 awards and dozens of patents, Smith has pioneered the creation of instruments that can separate and identify proteins and metabolites with high sensitivity, accuracy and resolution for biological and biomedical applications.
For example, the end goal of a recently funded three-year project is a high-throughput system that can be commercialized for use by clinicians and hospitals to detect early signs of liver disease. The PNNL team also plans to make it adaptable to other diseases. To achieve that, the team members say they're going to have to make their current system a hundred times faster and more sensitive.
To separate and identify the thousands of proteins in a sample, the researchers fragment the proteins into smaller peptides and run the peptides through liquid chromatography, ion mobility spectrometry (IMS) and time-of-flight mass spectrometry (TOF-MS) instruments. The separation techniques work on the general principle that different molecules move through media at different speeds based on mass, shape, charge or chemical properties.
After liquid chromatography separates the peptides based on their chemical properties, they are ionized, usually via an electrospray ionization technique.
Ionization gives them an electrical charge that allows them to be both separated and identified.
Next in line is the IMS, where the peptides are sent through a drift tube at relatively high gas pressure. The drift speed is based on the peptides' shape and molecular weight. "A different structure with the same weight will have a different drift time," says Tang. "This separation helps reduce the mixture's complexity. Also, we can use this to look for protein misfolding, which may indicate disease."
The peptides pass from the IMS drift tube into the TOF-MS, which measures the ions' mass-to-charge ratio. Analysis of the peptides reveals the range of proteins found in the original blood sample.
Funnels and multiplexing
The researchers made breakthrough modifications at both ends of the IMS to improve the instrument's sensitivity. Before entering the instrument, the ions collect in an ion funnel trap, from which they are released into the drift tube in small packets. "Ten million ions can be stored in the trap," says mass spectrometrist Mikhail Belov. "Conventional traps keep ions at low pressure, and it's inefficient to release them into the higher pressure of the drift tube. We collect them at a pressure that closely matches that step." This greatly increases the efficiency of the system.
Unlike in the first-stage liquid chromatograph, which can take peptides up to 15 minutes to get through, packets of ionized peptides move through the IMS in tens of milliseconds. That might sound fast compared to the lumbering chromatograph, but the ions still pile up behind the starting line.
"The drawback of conventional IMS is that it takes a tenth of a millisecond to move an ion packet into the drift tube, but then we have to wait 10 to 100 milliseconds to get the ions through," says Belov. "When dealing with patients, we want analysis fast." Conventional LC/MS methods manage only six to eight samples a day.
To speed up the analysis, Belov developed clever technology to allow a trick called multiplexing. The multiplexing technology allows packets of ions to be released in waves in such a way that the ions can be identified by the wave in which they traveled. The sophisticated method is similar to keeping track of different groups of runners in a big-city marathon.
Multiplexing through the IMS drift tube offers a dramatic increase in speed and efficiency. "We're doing a 50-minute experiment in one minute," says Belov. That translates to a seven-fold improvement in sensitivity.
Tang, who won a Federal Laboratory Consortium Award in 2004 and was named PNNL 2007 Inventor of the Year, has helped the team tackle a problem in IMS that used to waste a high percentage of the sample when it reached the third stage of analysis, the TOF-MS. As anyone who's opened a plastic bottle at high altitude knows, a quick drop in pressure can spread the contents wide and far. The ion beam coming from the IMS drift tube is a plume of mixed gas with ions, and it has to fit into a tiny hole on the TOF-MS.
When transferring the peptides from the high pressure of the IMS to the low pressure of the TOF-MS, the vast majority of the peptides slam into the surface around the mass spectrometer entrance. So the team developed an ion funnel to position in front. The funnel consists of ringed electrodes of decreasing sizes that collect and focus the ions, herding in more than 95 percent of the ions that would otherwise be lost. For the high-throughput, high-sensitivity proteomics the PNNL researchers want to do, says Tang, "the ion funnel was the key element to gain optimum ion transfer for each stage of separation."
"The IMS was not very sensitive before we put the ion funnel into it. The funnel works so well, it has propagated through the IMS-MS community for a wide range of applications, but it's not yet available commercially," he says.
All that data
Of course, the fastest, most sensitive instruments in the world would be useless without the computer technology and algorithms to handle the data they create.
"The challenges associated with processing the data coming from the technology are equal to the challenges to develop it," says electrical engineer Gordon Anderson. "All of our current instruments generate 24 terabytes of data a year together, and this IMS system has the potential to produce 100 terabytes just by itself. Can you imagine how much data you would have to store, process and move around if you had a fleet of these instruments?"
The laboratory is developing new algorithms to take advantage of the speed and sensitivity of the IMS. That task requires masses of data on a wide range of ionized peptides, which the team is in the process of collecting.
"The instrument development team is already generating high quality data sets," Anderson says. "The drift time is based on a number of physical parameters such as mass, how many charges the peptide ion carries, the order of amino acids, which determines the shape of the molecule. That algorithm doesn't exist yet."
Belov says an additional improvement the team has made to the system set-up is a converter that allows more information about the ions to be collected, and for data to be collected, transferred and stored simultaneously.
The team hopes their improvements will give them a multitude of opportunities to work with the biomedical community to make better health a bright, pointed goal.
Mary Beckman, Ph.D., is a science writer at Pacific Northwest National Laboratory.
Copyright © 2012 | Innovation America