Send to printer »

Feature Articles : May 15, 2008 (Vol. 28, No. 10)

Mass Spec Looms Large in Discovery

Study Design and Sample Prep Must Be Carefully Carried Out for Benefits to be Realized
  • Nina Flanagan

Hardware and software advances have greatly improved the reliability of mass spectrometry. In addition, its versatility and movement into the clinic has expanded the number of users and applications. Also affecting the MS market are increased efforts to screen patients through identification of biomarkers and to develop more effective drugs based on structural properties of targets.

Researchers at the Fred Hutchinson Cancer Research Center (www.fhcrc.org) are working to discover cancer biomarkers using patient plasma samples. “The protein-concentration range in plasma is very large, 10 to 12 orders of magnitude. There’s no single analytical technique to handle so wide a dynamic range,” explains Hong Wang, Ph.D., director of the center’s mass spec facility.

The proteins that really reflect disease development are low-abundance proteins, which are below the nanogram/mL level. To detect these rare proteins, Dr. Wang’s group uses chromatogram-based protein fractionation first, followed by MS analysis.

There are four electron-spray ionization (ESI)-MS systems in Dr. Wang’s lab. They were chosen instead of LC-MALDI because of their high-sample throughput. “After 2-D HPLC intact-protein fractionation and trypsin digestion, I pool the digested fractions for LC-ESI-MS analysis.” Up to 720 samples can be analyzed in a single experiment. Dr. Wang adds that it is key to reduce carryover from run to run with the ESI system in order to increase the HPLC separation efficiency and to analyze protein isoforms.

Dr. Wang also uses a linear, ion-trap MS analyzer because it has a fast scanning reader. “If scanning of data acquisition is slow, you will miss peptide analysis.” Sensitivity is really not an issue for MS, it’s more important to have a dynamic range of three orders of magnitude, he adds. Since reproducibility depends on the HPLC system, Dr. Wang’s group uses a nanofluor HPLC and they pack their own columns to ensure stability.

Cerebral Spinal Fluid

Perhaps even more difficult to analyze than plasma, cerebral spinal fluid (CSF) poses different challenges. “We haven’t done the fundamental studies, like others have done with plasma, to know what the interfering compounds are,” explains Mark Hayward, Ph.D., associate director of analytical chemistry at Lundbeck Research USA (www.lundbeckresearch.lundbeck.com).

Triple quadrupole MS-MS is used to analyze CSF samples because it is the “tried-and-true best way to do quantitative analysis in complex matrices. These analyzers provide predictable performance, so you get a lot of dynamic range that’s not available with other types of MS systems.”

Dr. Hayward’s group is focusing on small peptides and neurotransmitters like dopamine and acetylcholine to discover what the best biomarkers are and to measure quantitative levels. “We’re able to get high-precision levels to three percent RSD (relative standard deviation), which is very precise for picogram/mL concentrations.” This is important because “it allows the information to be useful, so you can actually screen patients,” he adds.

Another hurdle has been to make nano-sized 2-D liquid chromatography reproducible in order to get the really low concentrations. “The chromatography is advanced enough for nanoscale, but getting it all to work is quite a technical challenge.” In addition, Dr. Hayward says, “as you move downstream and have greater impact, the reproducibility gains more importance.”

Detecting Protein Changes

Merck & Co’s(www.merck.com) proteomics research group is using high-resolution MS to identify and quantify changes in peptide and protein levels related to disease or therapeutic agents.

“Our goal is to facilitate Merck’s basic research and to impact clinical medicine by providing a scientific approach to address these questions without antibody reagents,” states Ronald Hendrickson, Ph.D., director of proteomics, molecular profiling, Merck Research Labs (www.merck.com/mrl). A main advantage to this approach is the ability to perform unbiased analyses in complex biological systems without having to prespecify the analyte being measured, adds Dr. Hendrickson.

Since MS can rapidly cross species boundaries, it enables quick movement from cell-based experiments to preclinical models and then to humans. Dr. Hendrickson’s lab utilizes a triple quadrupole MS-based assay that uses stable isotope-labeled internal standards to improve precision and provide absolute quantification in lieu of developing an ELISA assay, which is time limiting.

The group is currently focusing on peptide and protein markers of proximal target engagement. This class of markers helps to address the question “did treatment engage the desired biochemical target?” Knowing this information in both preclinical models and human trials is critical and can help inform early go or no-go decisions. Another interesting application involving this type of MS is looking at protein turnover rates, where an understanding of protein dynamics is especially important to some diseases.

Key characteristics to look for in a MS system, says Dr. Hendrickson, include resolution, mass accuracy, sensitivity, dynamic range, and compatibility with the chromatography timescale. “I look for instruments that can maintain mass accuracy of better than five parts per million over a three-week period without requiring recalibration and without an internal standard.”

Small Molecule Discovery

MS is also for small molecule discovery and quantitation. Chris Petucci, Ph.D., senior research scientist, analytical chemistry at Wyeth (www.wyeth.com), says his group uses MS to support medicinal chemistry and biological programs in drug discovery. MS is used to determine molecular weight, confirm the identity of compounds, and perform accurate mass analyses to determine molecular formulas of unknowns.

The Wyeth group uses tandem-MS to elucidate the structures of compounds and to quantitate trace levels of endogenous compounds in biological matrices such as steroids and fatty acids in serum and brain tissue. “This is an exciting area because the aim is to minimize compound attrition in the early stages of drug discovery,” states Dr. Petucci.

“Tandem-MS is the gold standard in the field of analytical chemistry for quantitating trace amounts of compounds down to the femtogram (10-15) level,” he explains. “It offers a degree of selectivity and specificity for quantitative studies because parent ions generated from a drug are fragmented into daughter ions. These are specific to the parent ion only and not to other molecules that could be present in the sample matrix.”

Important MS characteristics, says Dr. Petucci, include sensitivity, precision, and accuracy. “These are all important in quantitative studies, especially at the clinical stage, because there are FDA requirements for rugged and reproducible assays.” In addition, MS is versatile and rapid, providing the molecular weight for a compound in seconds, and when coupled with HPLC, it provides high levels of accuracy.

Protein Profiling of FFPE Tissue

According to Casey Eitner, CEO of Expression Pathology (www.expressionpathology.com until the Liquid Tissue® MS Protein Prep Kit was launched, there was no way to do MS analysis on formalin-fixed tissue. “This opens up a whole new avenue for protein analysis, biomarker discovery, and protein target discovery.”

This automated protocol solubilizes the entire tissue sample so there is a representation of the total protein content. “It has been optimized to work with 30,000 cells, which isn’t a lot, but it’s the amount you can typically microdissect. This allows users to harvest specific cellular features from a single tissue section and perform detailed proteomic analysis of those cellular features,” notes Eitner.

Liquid Tissue has two main MS applications—to identify differently expressed proteins associated with a specific condition or clinical outcome (e.g., which proteins change as cancer progresses) and once proteins of interest have been identified, using MS with single-reaction monitoring (SRM) or multiple-reaction monitoring (MRM) to quantitate the target protein(s) in the changing clinical circumstances.

“We’re just starting to offer the ability to specifically quantitate changes in protein expression in FFPE using SRM and MRM spectrometry,” says Eitner. “This can be done with labeled peptides or in an unlabeled manner.” Eitner thinks this technology will provide protein-expression analysis alternatives.”

Mass spectrometry is technically challenging, and the sample prep prior to running MS is especially important. “It’s not just about running things through the mass spec, it’s about integration and putting the correct thing into MS and interpreting results properly,” says Eustache Paramithiotis, Ph.D., director of cellular and molecular biology at Caprion Proteomics (www.caprion.com). “We’ve seamlessly integrated all the steps involving sample prep, analysis, and informatics to understand results and made it an industrialized process with QA and QC.”

The resultant CellCarta® proteomics platform allows the comprehensive detection of disease-relevant protein targets and biomarkers across a wider dynamic range of concentrations, according to Dr. Paramithiotis. Sample prep enables isolation of organelles. This provides enhanced proteomic comparisons of normal and diseased cells as well as the biological and functional context for identifying disease protein targets.

Caprion has 11 LC-MS systems capable of processing up to 250 biological samples per week. Samples are first analyzed via LC-MS to detect differentially expressed peptides and then reinjected in to LC-MS-MS to obtain sequence information for those peptides. Analytical tools transform raw LC-MS and LC-MS-MS data into reliable protein-expression information.

“With proteomics, you have to look at the overall picture from beginning to end, not just what MS does. You can have an absolutely precise instrument, but if your study design is poor and your sample prep is variable, then what good is it? Proteomics is not just about giving a list of proteins to the end user, it is also about giving some interpretation of what they mean,” says Dr. Paramithiotis.

Overall, mass spec is being adapted for more diverse applications. Sample prep continues to be a crucial step in MS that must be performed with great accuracy. The technology will continue to advance as the research requires. It remains to be seen, however, how the new data will be used. “There will always be a technological element, better machines, etc., but the technique’s penetration—how well the answers will be utilized—that’s the crux of it,” summarizes Dr. Paramithiotis.