Despite some disappointments, the market for protein arrays is following a healthy growth curve that will lead to an approximately $600 million market by 2008 if current trends continue and application-focused work accelerates, according to Steven Bodovitz, principal consultant for Select Biosciences, a firm that tracks life sciences market trends. “The demand is there — it’s just a matter of getting these technologies to work well,” he told ProteoMonitor. “If it works as advertised, there’s definitely a market for it.”
Bodovitz pointed to two companies — Protometrix and Molecular Staging — which are at different stages of applying technologies that he thinks look promising. Protometrix last year came out with a whole yeast proteome chip that the company and the technology’s originator, Yale’s Michael Snyder, claim can probe for a variety of interacting proteins with good specificity, and say has few problems associated with the denatured state of the immobilized proteins or with cross-reactivity events (see PM 11-7-03). The data “looks good, but people still just don’t believe it,” Bodovitz said. “The question is, can you really do biochemistry on a chip using immobilized proteins?” Bodovitz expressed cautious optimism for the company’s future, but noted that the proof in the pudding would be whether independent investigators can use the chips to get real data. Much hinges on the results. “Protometrix is the biggest wildcard in all these [economic] projections,” he said, explaining that the company could lead the way in making this type of array widespread, but only if the technology proves to be viable.
Still, several other companies are already producing results that they are delivering to pharma and government customers. One of Bodovitz’s top picks, Molecular Staging, announced a deal with Eli Lilly back in early 2003 (see PM 2-3-03), and both MSI and Lilly made presentations at the Peptalk protein array conference last week (see PM 1-16-04), discussing their initial results for studies involving the diagnosis of sepsis and the toxicology of sepsis drugs. MSI’s technology uses a signal amplification technique called “rolling circle amplification” to measure the responses of proteins that react with a variety of analytes — 30 to 40 on each chip. According to MSI chief operating officer Stephen Kingsmore, who presented the data at the conference, MSI currently has about 140 analytes available as content for testing six different diseases spread over five different biochips.
Kingsmore said MSI is currently in the process of completing its third study on sepsis markers for Lilly. In its first study using its sepsis analyte arrays, MSI found 12 analytes that showed a difference between patients with sepsis and those who were normal, and the reaction was followed over seven days as the patient improved and differences dropped to baseline. The second study repeated the process, only putting on the arrays the potential biomarkers found in study one.
The third study, which the companies are currently conducting, looks to identify additional biomarkers by examining a greater number of patient samples. By using biomarkers that showed the greatest differences between disease and non-disease states, Kingsmore said that MSI could diagnose sepsis with 96 percent positive predictive value and 99 percent specificity.
In a later lecture, Brian Edmonds, group leader of enabling technology at Lilly, said that using MSI’s chips enabled him to gather enough information from biomarkers to have predictive power for the disease. “There is no predictive power in individual analytes,” Edmonds said. “But when you multiplex, that’s when you get the home run.” One way in which Lilly hopes to ultimately use this predictive power is to identify people who are at risk of imminent death during a sepsis event, so that more doctors can be convinced to prescribe Lilly’s anti-sepsis drug, Xigris.
Biodefense, Cytokine Arrays Also Hot
Beyond the still-developing drug discovery realm, biodefense also featured among the Peptalk speakers as a major early application for protein chips. Frances Ligler, senior scientist of biosensors and biomaterials at the US Naval Research Labs, described a technology that she is developing to detect, right on the battlefield, pathogens that may be used in a bioterrorist attack. “The idea is that just about anybody with a little bit of sense can use it just about anywhere,” she said. To do this, Ligler essentially ignored all the stringent sample preparation and controls that most companies agonize over, to create a complete, shoebox-sized, fluid-based multiplexed antibody array system that can crudely screen for six different biological warfare agents in 10 minutes. Most iterations of the array use sandwich assays based on known antibodies to the pathogens, although Ligler said she is also testing less specific molecules — such as certain oligosaccharides and molecules called siderophores that can detect whole families of pathogens — for use in screening for pathogens for which there may be no readily available or known antibody. For the purposes of biodefense, Ligler said, exquisite specificity was less important than robustness. “We can analyze these materials in the presence of mud, blood, and guts,” she said.
As another example of how to use antibody arrays to screen for the presence of viruses, BioForce Nanosciences of Ames, Iowa — which did not present at the conference — described in a paper in Nanotechnology this week the use of a nanoscale silicon antibody chip called ViriChip for detecting viruses in combination with its atomic force microscope. BioForce already collaborates with Emanuel Petricoin and Lance Liotta on miniaturizing the NCI-FDA group’s reverse phase arrays for cancer biomarker discovery (see PM 5-13-02).
Phil Felgner, a professor at the Center for Virus Research at the University of California, Irvine, also described at Peptalk his work with pathogens to come up with a different kind of defense system: a protein-based vaccine. The idea here, he said in his talk, is to express every protein in the proteome of a particular pathogen, such as malaria, from starting material consisting of the pathogen’s genomic DNA. The proteins could be placed on arrays and then human serum could be added to screen for which proteins made by the disease react particularly strongly with immunoglobulins in the serum. Those reactive proteins, the thinking goes, would make good candidates for protein or peptide-based vaccines, since they are the ones that he would know the body reacts to. For large viruses like malaria, this is a more efficient system than just trying every protein, Felgner said.
Plain old cytokine detection arrays — along the lines of the one released by Zyomyx early last year (see PM 8-26-02, 6-13-03, 11-7-03), also figured prominently at Peptalk. Novagen, a subsidiary of EMD Biosciences, officially released its 16-well human cytokine array at the conference, and PerkinElmer said during a luncheon presentation that it has produced and is working to commercialize a 32-cytokine array.