Chipping away

PROTEIN CHIP TECHNOLOGIES ARE HELPING TO ADVANCE
UNDERSTANDING OF DISEASE, AND OFFER NOVEL DIAGNOSTICS

By Deborah Komlos

Existing and emerging protein chip technologies are capturing a critical space in the health-care sphere. For investigators, they are providing effective tools to further basic research. For companies, they are being developed into promising applications for disease diagnostics and drug screening.
Fragile, complex and easily denatured are but a few of proteins’ many attributes that present challenges for the development of protein chips. Through its wholly owned subsidiary Sensium Technologies Inc. (Edmonton, AB), Helix BioPharma Corp. (Aurora, ON) has developed a proprietary biochip that factors in these limitations. The chip is based on Helix’s Heterodimer Protein Technology, which involves the formation of coiled coils: molecules comprising two small peptides that spontaneously associate in solution like a zipper.
With one of the peptides anchored tightly by chemical bonding to the chip surface, “what you do with the other peptide, (is) you add it to the protein of interest, and because it’s a peptide, we can use recombinant DNA techniques: you can make fusion proteins,” explains Helix vice-president of Technology, Heman Chao, PhD. “Now you have your protein tagged with this opposing coil — you add that to the other coil surface, the two will get together and you have your protein presented at an angle.” (See pg. 25)
This type of chip, called a capture biochip, prevents attached proteins from denaturing, Chao says, because the coiled coils cushion them from touching the solid chip surface.
Chao points out that while current recombinant protein expression techniques permit production of relatively simple proteins, many proteins are multimeric and require other proteins to assemble.
In the case of membrane-integrated proteins, protein chip researchers are having a “very, very difficult time,” he says, “not only about immobilization, not only about keeping them functional, but how to deal with perhaps a large school of proteins that are interesting.
“I think hurdle number one is (to be) able to present proteins, have them active and (be) able to do robust biochemistry on them, and we seem to be making some headway in that direction,” Chao says. “Now, the thinking is if you can put proteins down, if you can do biochemistry on them, perhaps then you can build complexes on them. That’s more into what they call the interaction chip. But that’s quite a significant leap.”
In addition to its unique coiled coils, Chao says Helix’s protein chip technology stands out because it allows for protein production right on the chip.
“All other protein chips probably need you to have the protein ready to go, or the company has in fact made the protein chips for you,” he says. “But in this particular case . . . you can add your cDNA, you add the appropriate in vitro translation mixture. The protein will be made, assuming that it doesn’t have the normal problems, and be captured and presented on a chip simultaneously.”
The firm has already worked with proteins that are of strong interest to drug-discovery research: those recognized by kinases and others recognized by proteases. Earlier this year, Helix licensed its biochip technology to Lumera Corp. (Bothell, WA), which intends to use the technology to advance the growing field of protein arrays.

Keeping it Together
At Toronto, Ont.-based Umedik Inc., innovation in protein chip technology has come in the form of an industrial microbe and human protein diagnostic platform called OnSite MicroDiagnostics™.
The platform consists of the DIAprep™ applicator, the BIOchip™ staging platform, and BACscan™, an opto-electronic reader, says Martin Gelb, Umedik’s director of Marketing.
“The reagent is freeze-dried in the applicator. The resulting mix is put onto the biochip, where the fluid, be it blood, is separated, and on the chip itself there’s the printed microarray,” he says. “This is put into the reader where it is moved about on a small reading tray. The fluorescent antibody is excited by the laser, and the subsequent data is read into the reader and presented.” (See pg. 10)
A feature that makes the system unique, Gelb says, is that the fluid to be tested is separated on the chip itself, which eliminates the need for centrifugation. Also innovative is the fact that calibration data is taken from the native sample. “The standards are actually derived from the patient-specific sample,” he adds.
ICEflo® — a 3-D dynamic filtration matrix — is also novel, providing on-chip separation of particles from fluid, says Umedik founder and CSO Peter Lea, PhD.
Lea says areas of concern regarding protein work are quite standard, and there are mechanisms to deal with them.
“There are ways to preserve proteins. There are certainly ways to make sure that they maintain their tertiary structure, and their activities. There are ways of ensuring that the epitopes of the binding sites are available or accessible to find,” he says. “So, 20 years of research have gone into this in the field, in general, by many people, and I think we’re all beginning to understand the real trouble spots.”
These inherent concerns need to be considered when developing an assay, Lea explains.
“Even if you are taking a known, existing assay from one format onto our chip format, the same questions will still come up every single time,” he says. “So you actually have to go through the development process and evaluate each step to make sure it works as you expect.”
While Umedik’s first revenue opportunity is in microbiological food-safety testing — the firm expects to launch a product in that area before year-end — its main focus is on protein microdiagnostics and microarrays, Gelb says.
A potential medical application of Umedik’s work involves cardiac markers.
“The National Academy of Clinical Biochemistry has a five-year strategic vision for improved standards in cardiac testing,” Gelb says, “and what they’re calling for is quantitative lab standard results in less than 30 minutes. We already meet that. In fact, we exceed it.
“We are working with a very well-known medical facility in the United States to create a novel cardiac prognostic, so that if we can identify a certain level of certain proteins within a patient’s blood, we can perhaps predict that that person is liable to have a cardiac event of some type or another.”

Promising Patterns
In the laboratory of Dr. Brian Ward, chief of McGill University’s (Montreal, QC) division of infectious diseases, the focus, he says, is “to try to identify biomarkers in the blood of individuals infected with latent or asymptomatic parasitic diseases with an eye to protecting the blood supply.”
Along with co-investigators, Ward has an agreement with Ciphergen Biosystems Inc. (Fremont, CA) under the University-Industry Partnership Program of the Canadian Institutes of Health Research (Ottawa, ON). By way of the agreement, his lab has received a fully automated Ciphergen ProteinChip® SELDI-ToF MS system, as well as access to any needed chips.
Working with sera from around the globe requires that they be well-defined, Ward says. That is, that they have a definitive diagnosis.
“For example, somebody in a malaria-endemic area, they can have malaria in their blood without actually being sick from malaria,” explains Ward, also the director of the National Reference Centre for Parasitology (Montreal, QC). “In individuals who live in Chagas disease areas, they may also be sick with Leishmania, or have had Leishmania in the past. Many of these tropical protozoan diseases are endemic in the same parts of the world. So, what you really need are people who have ideally just one of these infections, so that you can compare apples to apples.”
The ability to find proteins and protein fragments is valuable, particularly in the case of latent parasitic diseases, Ward says. Most of the biomarkers his group is identifying are fragments of host proteins, he says, and they appear to be uniquely fragmented in different diseases.
For his line of work, Ward says the traditional criticisms that have been levelled against SELDI, such as its relatively poor resolution, are of little consequence. “We’re not stopping at biomarker patterns.
“We’re going on to identify what those peaks are,” he says. “I think that diagnostic tests based on biomarker pattern alone have real potential weaknesses. But if you go beyond that to identify the biomarker identities, with each identification you can ask questions that are relevant to the biology of the underlying disease.”
Broadening the Scope
Ward says the presence of these identified uniquely truncated host proteins raises the possibility of generating antibody reagents, for instance, or other binding proteins directed against those unique products.
“If in fact we can succeed in doing that,” he says, “we can then put those antibodies onto a chip and utilize the exquisite sensitivity of (SELDI) that’s well beyond what’s achievable by current ELISA technologies, for example. But you know, this is sort of dreaming. It’s a year or two down the road.”
Professor Terry Spithill, PhD, director of McGill’s Institute of Parasitology, and a co-investigator on Ward’s research grant with Ciphergen, is using SELDI-ToF approaches in animal disease systems, with one project examining Fasciola (liver fluke) infections in sheep.
“We’re using the protein profiling approach to compare parasites of different virulence. We’re using it to look at markers in fecal material from infected animals as well as (in) sera. We’re also using the approach to try to identify in human malaria, (with) parasites grown in vitro — we’re looking at stage-specific expression profiles,” says Spithill, also the Canada Research Chair in Immunoparasitology.
The diagnostic pattern that’s detected from a mass spectrometer derives from both host and parasite molecules, Spithill explains. “So you’re not actually necessarily detecting the parasite,” he says. “You’re detecting the host response to the parasite, and that’s a paradigm shift, if you like, in terms of how we might look at diagnosing infections.”
Although researchers can also use 2-D gels to compare samples, this technique has issues of solubility, staining and particularly in detecting low-abundance proteins, Spithill says. Whereas, the discriminating power of the mass spectrometer allows the comparison of two proteins that are very similar in molecular weight.
“Both of these techniques are similar in concept,” Ward adds, “it’s just that you can’t robotize 2-D gels and you can robotize SELDI. However, we are clearly going in the direction of biomarker identification through screening with SELDI, followed by confirmation by 2-D gels and other approaches.”
Regarding the prospect for protein chips to automate protein screening, Chao points out that working with them takes longer than working with DNA chips, for which viewing 20,000 spots per day is not unusual. “High throughput to proteins for a yes or no answer is perhaps in the thousands. But when you talk about proteins, people usually want more than just a yes or no,” he says.
“They want to know who is it talking to, how is it communicating. So, when we say high throughput, we usually mean, OK, we have put down a thousand proteins and we want to screen them against a chemical library for drug discovery,” Chao says. “Then it’s closer to what the field means by high throughput.”
According to Lea, the future of protein chip technologies holds much promise.
“In terms of development, I think from my assessment with protein chips at this point, we are probably where Apple computers were 20 years ago,” he says. “I think we’re going to see some amazing things coming out of the technology. It’s an exciting place to be, especially when you make progress.”