College of American Pathologists
Printable Version



cap today

The analytic power of microarrays

August 2001
William Check, PhD

Back to Cover Story

Some scientists working in molecular technology believe that microarrays and expression microarrays will have an important place in clinical laboratories. As sequencing becomes more common, there will be a need to do it faster, a need that microarray platforms can meet—if other, better methods aren’t developed first.

Currently, the analytic power of microarrays can best be seen in research studies on, for example, the analysis of gene expression patterns in peripheral blood mononuclear cells (PBMCs) or macrophages from infected patients. Dr. David Persing, vice president of diagnostics development, Corixa Corp., and medical director of the Infectious Disease Research Institute, Seattle, is conducting such studies. Dr. Persing’s work centers on a recently identified family of proteins called Toll receptors, which are thought to protect against infections. Different Toll receptors respond to different components of microbes. If a person is infected with a gram-negative organism, for instance, one receptor recognizes bacterial DNA, another bacterial flagellin, another lipopolysaccharide. Other Toll receptors recognize glycoproteins, proteoglycans, and other cell wall components from mycobacteria and staphylococci.

Stimulation of Toll receptors triggers a cascade of signaling events that changes the transcriptional profile of the target cell. Mediators are produced, both interferons and cytokines such as TNF-alpha and IL-12. Transcriptional changes vary among cells and receptors and can be detected with expression microarrays soon after stimulation.

Response variability between receptors may determine a person’s susceptibility to infections, Dr. Persing suggests. "Ultimately, knowledge of this variation may make it possible to predict the likelihood for developing severe infectious sequelae in certain contexts," particularly immunosuppression, he says.

From the diagnostic viewpoint, a microarray that analyzes the expression profile of PBMCs or macrophages to see which Toll receptors were stimulated might make it possible to discriminate viral from bacterial syndromes: Expression of interferons might indicate a viral infection, while expression of certain cytokines might indicate a bacterial infection.

In animal experiments, Dr. Persing and colleagues at Corixa have characterized a possible protective drug that works by stimulating Toll receptors. A single intranasal treatment with this agent completely protected mice for several days from an influenza virus dose that typically leads to 70 percent mortality. This compound represents "a new class of drugs that can provide broad protection against infectious challenge by stimulating the body’s natural immune response," Dr. Persing says.

Based on Dr. Persing’s work so far, Corixa was just awarded a $3.5 million grant from DARPA, the biological warfare group, to develop a chip that monitors inflammatory responses and helps classify stimulation patterns of Toll receptors and to develop a class of drugs that stimulate Toll receptors and provide broad resistance to biological warfare agents, especially those transmitted by aerosol.

At Sequenom in San Diego, the search for new diagnostic sequencing tools has bypassed microarrays. "In diagnostics we would sometimes like to have sequence information to determine a bacterial species variant or a human genetic variant," says Charles Cantor, PhD, chief scientific officer. In this situation you usually know the sequence and only need to look at one or two sites on the DNA, "almost like a spellchecker," Dr. Cantor says. "One can use conventional tools, those used to determine the sequence in the first place," he says. "Right now that is used for want of fully mature technology, but it turns out not to be cost-effective."

An alternative approach when it is known where the variant can occur is to do a focused examination of a few bases. "For many viral and bacterial variations and many human diseases we know where to look, so that kind of very precise spellchecking is what we do," Dr. Cantor says. "We use mass spectrometry because it is essentially error-free. We literally weigh a fragment of DNA amplified from the target of interest." The fragment has a weight corresponding to one or the other variant, a 300-600 dalton difference in a target that is 6000 daltons in overall weight (a one- or two-base shift in a fragment that is typically 20 bases long). "Since our scale is accurate to one dalton, we will never make a mistake," Dr. Cantor says.

To make short fragments, the target DNA is copied with a DNA polymerase under conditions where the polymerase stop in a base-specific way corresponding to the altered nucleotide. Samples are spotted onto a microfabricated array; they are blasted off the chip with a laser and into the mass spectrometer where they are weighed. The spectrometer is fast—it reads one sample per second-but operates in serial mode. Presentation by an array is needed to allow the spectrometer to operate in high-throughput mode.

Clearly, microarrays are used only as carriers in this process. "I just don’t know if arrays will ever be accurate enough for sequence differences to make a mature diagnostic tool," Dr. Cantor says.

In the infectious disease arena, Sequenom has a collaboration with the New York Public Health Institute on M. tuberculosis, among others. And, even though clinical testing is not Sequenom’s major focus today, the company does supply Quest and Specialty Laboratories with chips, mass spectrometers, and design services. "We are very pleased," Dr. Cantor says, that clinical reference laboratories consider Sequenom’s techniques worth trying.