This Special Advertising Feature is brought to you by AAAS OPMS

DOI: 10.1126/science.opms.p1200068

 For PDF version | New Products

LIFE SCIENCE TECHNOLOGIES
Panning the Proteome for Biomarker Gold

picture

Targeted proteomics is homing in on promising biomarkers to help screen for cancer and guide patient treatment, but much work still needs to be done to validate these biomarkers and develop technology capable of bringing them to the clinic.

By Anne Harding

Inclusion of companies in this article does not indicate endorsement by either AAAS or Science, nor is it meant to imply that their products or services are superior to those of other companies.

Testing breast tumors for certain cell surface receptors is now a routine part of cancer treatment. Patients with the HER2/neu receptor, for example, tend to have a more aggressive disease and will be given the targeted drug Herceptin. Tumors that carry estrogen receptors—about 70 percent of all breast cancers—will respond to tamoxifen and other selective estrogen receptor modulators.

But one day, if the promise of proteomics is realized, cancer treatments could be tailored to a patient by assessing the entire proteome of a tumor cell, rather than just one or two receptors. "We envision that a breast cancer patient could come to the doctor in the morning and get the proteomics analysis, and maybe get the results in the afternoon," says Matthias Mann, an expert in proteomics and mass spectrometry and director at the Max Planck Institute of Biochemistry in Martinsried, Germany. "The question is very open of whether this can be done or not. Whether or not it would work, it would be so important that it is incumbent on us that we should try."

To fulfill the promise of clinical proteomics, investigators must overcome multiple hurdles, including the difficulty of obtaining adequate samples, a dearth of antibodies—and funding—for validation studies, and limitations with the accuracy and throughput of available technologies, such as mass spectrometry. But bit by bit, some of these hurdles are being overcome as basic scientists develop innovative strategies for data analysis as ever more powerful and sensitive mass spectrometers hit the market.

SLOW PROGRESS

When investigators first began looking into clinical proteomics, technical limitations meant they could only measure a few hundred proteins at once, while flawed data analysis usually yielded "biomarkers" that didn't turn out to reveal anything at all. Although there have been over 10,000 publications on biomarker discovery with proteomics, the U.S. Food and Drug Administration (FDA) has only approved a single proteomics-based diagnostic test. Vermillion's OVA1, cleared by the FDA in 2009, measures five different proteins in the blood, and is not used to screen for ovarian cancer, but to help evaluate whether an ovarian mass is benign or malignant prior to surgery.

"What you've seen in the industry is a lot of great technological developments, but in terms of locking down and then validating there are very few of these tests that have really made it through that process and ultimately been commercialized and validated in clinical practice," says Paul Beresford, vice president of business development and strategic marketing at Biodesix, a molecular diagnostic company.

The classic approach to asking clinical questions with proteomics has been to look at tissue or body fluid samples from patients with a given disease, which is verified by pathology, and compare them to a control group. "One of the critical issues is, of course, that measuring only a couple of hundred samples, control and case, would need to be done at a high enough throughput, and these measurements must be reproducible," says Ruedi Aebersold, a professor of systems biology at the Swiss Federal Institute of Technology in Zurich. "The conventional mass spectrometry technologies have great difficulty achieving these goals."

And some say that testing hundreds of samples won't be enough. "There's a general conclusion coming out of the real diagnostics industry that if you can't look at 1,500 to 2,000 samples there simply isn't any way you could know whether a biomarker is relevant to clinical practice," says Leigh Anderson, CEO of SISCAPA Assay Technologies, which develops assays for analyte enrichment at the peptide level for use with multiple reaction monitoring (MRM) mass spectrometry. The company has 12 assays in production and 54 in development.

"Clinical proteomics has been done for 50 years; people have had their serum proteins measured to detect disease for decades," says Stephen Kron, a professor of molecular genetics and cell biology at The University of Chicago. "But the question is: Can you do better than the proteomics we have now? Rather than just incrementally better, can you do better than ELISA? Can you make systems that are robust and highly multiplexed? The key thing is, there's no one answer, and the current methods are inadequate despite all of us using them."

NEEDLES IN HAYSTACKS

The vast dynamic range of the human proteome—especially the human plasma proteome—has been one of the biggest challenges in using global proteomics for biomarker discovery. The concentration of the lowest abundant proteins found in plasma, among which many investigators believe biomarker "gold" is most likely to be found, is 10 to 12 orders of magnitude lower than the concentration of the most abundant plasma protein, albumin. "That exceeds the dynamic range of pretty much any instrument we're trying to use to measure proteins," says Steven Skates, an associate professor of medicine at Harvard Medical School and Massachusetts General Hospital (MGH) who is studying early detection of ovarian cancer.

In collaboration with a team of investigators from MGH, the Broad Institute, and the Dana Farber Cancer Institute, Skates has been using several different strategies to expand the dynamic range of his experimental techniques. "One possibility is to use specimens that have a much more concentrated biomarker content," he explains. "It might be tumor tissue itself or fluid that arises from the tumor, but isn't diluted as it would be in the blood." He is currently looking at fluid from ovarian cysts, which potentially contain cancer-related proteins at a thousand-fold higher concentration than plasma.

Many investigators, including Skates, use fractionation to look at less abundant proteins; yet another strategy is to deplete higher abundance proteins from a sample.

"With a combination of depleting abundant proteins, high fractionation, and starting with a more concentrated biomarker source than blood, you've got about seven orders of magnitude that you're crossing, and then with three orders of magnitude from the instrument, that gives you 10 orders of magnitude," Skates adds. "We're just at the tip of where we think the biomarkers are."

TARGETED PROTEOMICS

Leaders in proteomics now agree that the "shotgun" or "brute force" approach to searching for biomarkers is an incomplete paradigm that falls short of the clinical goal. "I'm personally frustrated that we have been attempting to play with the technology over the past 15 years to find a shortcut," says Anderson. "But at the end of the day, you're going to have to commit to the details of specific hypotheses and make some serious measurements of the few proteins that make a difference."

That's where targeted mass spectrometry comes in. "Once we know what to look for, instead of trying to measure thousands of proteins, if we've got a specific target list we can develop assays that home in on these proteins," Skates says.

"One of the big rate limiting steps is going from those candidates we identify with global proteomics techniques to developing assays for accurate measurement of proteins in the blood," he adds. "Developing assays is very difficult because the blood is so complex that you can often get plasma or serum interference."

According to Sam Hanash of The University of Texas M.D. Anderson Cancer Center in Houston, it remains to be seen what technology will ultimately be used to bring these assays to the clinic. Hanash, who is studying blood-based markers for detecting early stage cancer, directs the McCombs Institute for Early Cancer Detection and Treatment.

Using fractionation and mass spec, Hanash and his colleagues have been able to find promising biomarkers at subnanogram per milliliter concentrations—concentrations too low for detection by ELISA. "The technology for research, for discovering clinical applications, are really far reaching at the present time, but we have not figured out how to achieve that level of sensitivity in a high throughput setting," he said. "It's clear that the discovery platform is too labor intensive and somewhat challenging to operate in a clinical setting...it doesn't mean that the situation is hopeless." Some promising approaches, he adds, include nanotechnology and more user-friendly mass spec. "There are a lot of other types of technology that have the potential to do the job."

But first come the "due diligence" validation studies that must be done to determine whether or not a biomarker is indeed clinically useful, and if so in which types of patients, Hanash says. "We have to dot the i's and cross the t's and see where have we seen this biomarker before," he explains. This work is done not just to reproduce the initial findings, he adds, but to find out how widely applicable a biomarker may be.

Biodesix has been spending the past five to six years doing validation work on its own test, VeriStrat, Beresford notes. "Although the discovery of algorithms and tests is important, validation is as important or more important to get those tests successfully launched and into doctors' hands to improve decision-making," he says.

The Boulder, Colorado-based company has created a technical platform, ProTS, to analyze matrix-assisted, laser desorption ionization (MALDI) mass spectra from biological samples. VeriStrat, which is based on this technology, analyzes blood from non-small cell lung cancer patients to help physicians determine whether they should be treated with the epidermal growth factor receptor (EGFR) inhibitor erlotinib. The company is also working with several pharmaceutical companies to develop companion diagnostics.

VeriStrat can distinguish between the two-thirds of advanced non-small cell lung cancer patients who will respond well to erlotinib and the remaining third who are poor responders, according to Beresford. "We've completed numerous validation studies in lung and other cancers where VeriStrat consistently identifies patients who are likely to have different outcomes following treatment with specific therapies," he says.

'STUNNING' PROGRESS IN MASS SPEC

Some proteomics experts say mass spec will ultimately be translatable to the clinic. Preparing for this possibility, Agilent registered its Infinity Series 1200 liquid chromatography systems and its 6000 Series mass spectrometry systems as Class I medical devices with the FDA this January, and registered its reagent manufacturing facility with the agency last June.

While this view is not universal, consensus in general is that advances in mass spec technology have been impressive. "Over the last five years, the pace of development and application of mass spectrometry has just been stunning," says Gilbert Omenn, a medical professor at the University of Michigan and chair of the Global Human Proteome Project.

These advances have gone hand in hand with advances in proteomics, and some leading scientists in the field are collaborating closely with companies that make mass spec machines. For example, AB SCIEX and Aebersold are collaborating on SWATH Acquisition, a mass-spectrometry-based technique that creates complete ion maps of all the fragments and peptides in a sample by repeatedly cycling through 32 consecutive 25-Da precursor isolation windows, or swaths. Aebersold and his colleagues used a fast, high-resolution quadrupole-quadrupole time-of-flight (Qq-TOF) instrument to develop SWATH, and AB SCIEX is now enabling this functionality on its TRIPLE TOF 5600 system.

Aebersold compares SWATH to a satellite measuring the surface of the Earth by making several orbits and combining the information into a single image. "You could think of taking a digital photo of a crowd at a football match, and then also having photos of individuals who have been previously identified," he explains. You could then use the individual portraits to determine who was present in the group photo. "Since this all happens in the computer, we can also reexamine these maps later when new hypotheses have been generated," Aebersold adds.

Richard D. Smith, director of proteome research at Pacific Northwest National Laboratories (PNNL) in Richland, Washington, notes that mass spec can now be used to make targeted measurements with a sensitivity matching ELISA, with assays that can be constructed very quickly. "It's becoming even faster in its rate of improvement and growth, and there are still enormous gains that are going to be coming over the next few years," adds Smith, who directs the National Institutes of Health's Research Resource for Integrative Proteomics.

Smith and his colleague Karin Rodland, chief scientist for biomedical research at PNNL, are co-principal investigators on the National Cancer Institute Clinical Proteomics Tumor Analysis Consortium, which will undertake detailed proteomic characterization of a large number of genomically well-characterized ovarian tumor samples.

The project will involve looking at how faithfully genetic changes are translated into protein levels, and investigating different protein modification states and their role. "We understand increasingly the limitations of what genomics can do and how complex the biology really is, so proteomics is essential," Smith says. For example, he adds, no biomarker at the RNA level has been found that will predict whether or not an ovarian cancer patient will respond to platinum-based drugs. But studying posttranslational protein modifications may be much more revealing. "A lot of what has been done in proteomics to date has been very simplistic," he says. "We're becoming increasingly sophisticated in our ability to track posttranslational modifications of proteins."

"The project is going to really start to show in tremendous detail and depth of proteome coverage how these genomic changes play out at the proteome level," says Rodland. "We're going to learn a lot about the biology."


Featured Participants

Agilent
www.agilent.com

AB SCIEX
www.absciex.com

Biodesix
www.biodesix.com

Harvard Medical School
hms.harvard.edu

Human Proteome Project
www.hupo.org

Massachusetts General Hospital
www.massgeneral.org

Max Planck Institute of Biochemistry
www.biochem.mpg.de/en/index.html

Pacific Northwest National Laboratory
www.pnl.gov

SISCAPA Assay Technologies
www.siscapa.com

Swiss Federal Institute of Technology
www.ethz.ch/index_EN

The University of Chicago
www.uchicago.edu

University of Michigan
www.umich.edu

University of Texas M.D. Anderson Cancer Center
www.mdanderson.org


Note: Readers can find out more about the companies and organizations listed by accessing their sites on the World Wide Web (WWW). If the listed organization does not have a site on the WWW or if it is under construction, we have substituted its main telephone number. Every effort has been made to ensure the accuracy of this information. Inclusion of companies in his article does not indicate endorsement by either AAAS or Science, nor is it meant to imply that their products or services are superior to those of other companies.

Anne Harding is a freelance science writer based near New York City.

DOI: 10.1126/science.opms.p1200068



This article was published as a special advertising feature in the 31 August 2012 issue of Science magazine.


For PDF version

New Products: Panning the Proteome for Biomarker Gold

picture

CCD CAMERAS

The CoolSNAP MYO and the CoolSNAP KINO CCD cameras are designed to discern finer details in biological samples under lower light levels and enable scientists to achieve higher quality, higher resolution images than previous CCD technology. Packing twice as many pixels (2.8 megapixels) with a 15% improvement in peak quantum efficiency (75%), the MYO and KINO offer scientists the ability to visualize much finer details at much higher sensitivity than standard scientific CCD sensor technology. The MYO is capable of cooling to 0°C and features a fan that can be disabled, which provides flexibility to optimize cooling for longer exposures in low-light scenarios and still accommodate extremely vibration-sensitive measurements. The KINO is cooled to 20°C without a fan, making it suitable for sensitive applications such as AFM. The MYO and KINO are ideal camera technologies for immunofluorescence and fluorescent protein imaging, and also well suited for near-infrared DIC, electrophysiology, particle tracking, FRET, and FRAP imaging.

Photometrics
For info: 800-874-9789
www.photometrics.com


IMAGE ANALYSIS SYSTEM

The PXi is a new high-resolution, multiapplication image analysis system. This powerful system is currently the best in its class for scientists who want a compact, one-click method for accurately imaging chemiluminescent and fluorescent blots as well as 1-D gels (up to 10 cm x 12 cm) stained with any type of fluorescent dye. The innovatively designed compact PXi, with its high-resolution 6.3 megapixel camera and large fixed aperture lens can quickly image even the faintest bands. It is easy to fit a range of lighting including infrared (IR) lighting and filter options inside the PXi system and using the intuitive GeneSys imaging software, the PXi can be rapidly set up to automatically select the best conditions. With just one-click, scientists can generate perfect images of chemiluminescent blots as well as 1-D gels (up to 10 cm x 12 cm) stained with any fluorescence and IR commercial dyes time after time.

Syngene
For info: +44-(0)-1223-727123
www.syngene.com


PROTEIN PRECIPITATION SAMPLE PREPARATION

The Combipack is an economic response to the growing use of protein precipitation separation techniques in chromatography labs. Comprising four Porvair p3 protein precipitation plates and four 1 mL deep well collection plates, the Combipack provides all the elements for protein precipitation separation, at a very affordable price. Based upon the industry standard MicroLute 96-well format, the p3 plate uses the CRASH method in which proteins are denatured with acetonitrile and the flocculent is filtered out. A novel dual filter matrix retains the samples in the wells of the p3 plate until required to release by vacuum or pressure. The use of a prefilter on top of an oleophobic filter eliminates blockages commonly found with other protein precipitation plates especially when handling high protein samples. The p3 has been proven in independent tests to eliminate the mess and complex sample preparation (no centrifuging, no vortexing) traditionally associated with protein precipitation techniques.

Porvair Sciences
For info: +44-(0)-1372-824290
www.porvairinformation.com/p3.htm


ATOMIC FORCE MICROSCOPE

The Dimension FastScan Bio Atomic Force Microscope (AFM) enables high-resolution microscopy research in biological dynamics. Breakthrough innovations in the design of the FastScan Bio system have resulted in a fast scanning AFM that allows temporal investigation under physiological operating environments in fluid while exceeding the diffraction limits of optical microscopy. The Dimension FastScan Bio system utilizes a revolutionary XYZ scanner designed to operate at high-speed rates while delivering extremely low drift and low noise, along with a small volume fluid sample cell and a user interface that removes operational complexity. Other features include an industry-proven fast scanner, automated laser and detector alignment, an easy sample engaging process, and the capability to investigate samples in a droplet of solution or its 60 μL sample cell. With the FastScan Bio, productive spatiotemporal nanometer scale research is a reality for every bio researcher.

Bruker Corporation
For info: 520-295-4373
www.bruker.com


NITROCELLULOSE MEMBRANES

Whatman FF High Performance (HP) membranes are a new diagnostic membrane that enables fast, razor-sharp line separation and highly reproducible results for the detection of target molecules in liquids such as water, urine, blood, and saliva. The target molecules may include drugs, hormones, proteins, antibodies, nucleic acids, whole bacteria, and viruses. The FF HP membrane is suitable for use in lateral flow assay manufacture as it provides improved reliability and rapid results. The new membrane is produced using an innovative casting method to yield a uniform, new powder-free surface that delivers a coefficient of variation in capillary rise of <_10. both="both" the="the" intra-="intra-" and="and" inter-lot="inter-lot" consistency="consistency" of="of" new="new" membrane="membrane" was="was" tested="tested" through="through" analysis="analysis" over="over" _1000="_1000" samples="samples" giving="giving" a="a" low="low" coefficient="coefficient" variation.="variation." three="three" ff="ff" hp="hp" membranes="membranes" are="are" available="available" with="with" different="different" capillary="capillary" rise="rise" times="times" wicking="wicking" rate="rate" allowing="allowing" researchers="researchers" to="to" choose="choose" most="most" suitable="suitable" for="for" their="their" assay.="assay." p="p">

GE Healthcare
For info: 800-526-3593
www.gelifesciences.com/dxcomponents

Electronically submit your new product description or product literature information! Go to www.sciencemag.org/products/newproducts.xhtml for more information.

Newly offered instrumentation, apparatus, and laboratory materials of interest to researchers in all disciplines in academic, industrial, and governmental organizations are featured in this space. Emphasis is given to purpose, chief characteristics, and availability of products and materials. Endorsement by Science or AAAS of any products or materials mentioned is not implied. Additional information may be obtained from the manufacturer or supplier.


Look for these Upcoming Articles

Genomics: Epigenomics — October 26
Tissue Engineering: 3-D/Scaffolding — December 7
Connectome — January 18