Skip to : [Content] [Navigation]
 

IVD Technology

 
 
Email article link Print article

IVD Technology Magazine
IVDT Article Index

DNA-chip technologies

Part 3: What does the future hold?

Cliff Henke

The commercialization of DNA-chip technologies may depend as much on overcoming regulatory and marketplace hurdles as on advancing technology. Note: this is the third part in a series of articles.

Also in this article:

Despite the myriad applications for the DNA chips now being introduced to research markets, demand for this technology is still relatively small. Nevertheless, most observers agree that this situation will soon change, as DNA chips begin to show up in products designed for more-traditional clinical applications. The consensus of these experts is that clinical products involving DNA microarrays will begin to reach the marketplace sometime early in the next decade. To reach this point, however, several important regulatory and technical hurdles must be overcome.

This is the final installment of IVD Technology's three-part series on DNA-chip technologies. The first installment reviewed the theoretical underpinnings of the field and examined the market forces driving product development (IVDT, September 1998). The second part looked at the state of the art and the various competing technologies in this embryonic field (IVDT, November/December 1998). This installment covers the challenges to commercialization facing companies engaged in this new marketplace, as well as prospective near- and long-term applications for these technologies.

Current Market

A recent study by Front Line Strategic Management Consulting Inc. (Foster City, CA) estimates that the DNA-chip marketplace will grow at a compound annual rate of 77% between 1998 and 2001. Almost all of these chips are DNA microarrays now, but by 2001 10% of this market is expected to be in protein chips, and by 2005 18% is expected to be so-called "lab-on-a-chip" products. Several industry observers predict that this last category of products will make up the dominant share of DNA-chip industry output by 2010. The evolution of DNA chips toward products that integrate sample preparation with detection and analysis on a single chip is an important indicator of the growing maturity of this field. But when and how these new technologies will turn the corner to commercial success will be determined by a variety of factors.

Initially, DNA-chip developers have commercialized their new products by forming partnerships with drug manufacturers, which can afford to fund the enormous costs of bringing microarrays to market (one often used estimate currently puts such costs at $1000 for each element of a new DNA array chip). In addition, pharmaceutical companies can see an enormous payback on this funding if DNA chips realize their promise of significantly cutting the time it takes to bring a new drug to market, which currently averages 12 years.

Analysts have high hopes for future diagnostic applications of DNA chips, such as this standard 100-site microarray by Nanogen Inc. Photo Courtesy Nanogen Inc.

Many observers and industry officials expect that some products developed initially for drug-candidate screening will soon find wider commercial applications in clinical settings. Development of one such product is the goal of a partnership between Affymetrix (Santa Clara, CA) and Roche Molecular Systems (Branchburg, NJ), which is devoted to creating diagnostic kits for the detection of p53, a gene suspected in a variety of cancers, and HIV, to determine a patient's resistance to therapeutic regimens used in the treatment of AIDS. Initially, the venture will rely on sample preparation and analysis equipment developed for research applications by Hewlett-Packard. But according to Affymetrix president Stephen Fodor, Roche is expected to develop smaller, less-expensive analysis equipment more easily suited to IVD applications, presumably under the aegis of Roche's recent acquisition, Boehringer Mannheim (Indianapolis).

It is not a coincidence that virtually every one of these early DNA-chip IVD products is being developed for the detection and treatment of AIDS, cancer, and heart disease. Throughout the world, these diseases are among the most prolific killers, and thus also represent the most lucrative commercial potential for medical technology developers.

DNA-chip–based products, if they reach their commercial potential, will revolutionize the diagnosis and treatment of disease. "Imagine being able to detect a few cancer cells in the presence of huge numbers of healthy cells," says Theodore Christopolous, visiting professor of clinical chemistry at Harvard University. Or imagine many genes whose expression in response to various therapies one might like to study simultaneously. DNA-chip technologies, says Christopolous, are seeking to realize these goals.

"Everyone understands the advantages of DNA chips in terms of faster and better diagnosis and disease management," agrees Daniel Farkas, director of clinical diagnostics at Clinical Micro Sensors Inc. (CMS; Pasadena, CA). "What we in industry now need to do is to bring them to clinicians at the same cost as existing diagnostic tests, or, better still, at less cost." To achieve this goal, product developers are employing a variety of technology-based strategies.

Technological Trends

DNA-chip technologies are progressing rapidly on four fronts, some of which are interrelated. Moreover, progress has been made not only with regard to DNA chips specifically, but also in the broader field of molecular diagnostics.

Sensitivity. The first area of technological advancement—improving the sensitivity of DNA-chip diagnostics—was discussed in the second installment of this series, but progress has been made even in the short time since that section was published. "There is increasing movement in product development toward detection of a single molecule or cell in an unamplified sample," says Larry Mimms, senior director of product development at Gen-Probe Inc. (San Diego). "This requires increasing sensitivity, which ultimately means either bumping up the signal or getting the background noise down."

A chief strategy employed to improve sensitivity has been to borrow better labeling technologies from the field of immunology, says Mimms. Currently, "fluorophore labels dominate all microformats in DNA testing," he says. But manufacturers are also experimenting with the use of time-resolved fluorescence, dendrimers, new uses of iodine radioisotopes, and chemical locks derived from improved washing techniques. In the very near future, adds Mimms, semiconductor nanocrystals as well as bioluminescent proteins will be developed to improve signal-to-noise ratios. However, virtually all of these techniques will require the laboratories that use them to perform additional, more-complicated sample preparation and detection procedures, he says.

If DNA-chip–based testing is to reach the clinical laboratory and point-of-care markets, a balance will need to be struck between the complexity of the testing procedures and the sensitivity of the tests. "In immunological applications, these processes are incorporated into a single device," Mimms acknowledges. "What's missing in chip-based systems is the ability to incorporate sample preparation with the detection and analysis procedures. Accomplishing this will require advances in microminiaturization."

But some observers insist there is another answer. "Target amplification strategies have problems because of the risk of contaminating the sample during the amplification steps," explains Christopolous. "With signal amplification, however, the target amount is constant." A variety of strategies are being employed to boost signal, ranging from better bioelectronic sensors to molecular strategies that bind probes and samples during the transcription and translation stages of DNA replication.

Microfabrication. Rapid advances are also being made in microfabrication, a technology essential for the manufacture of DNA chips. Companies are already fabricating chips with picoliter-sized wells and 10-microliter-sized chambers for sample preparation and detection.

There is evidence that manufacturers will be able to achieve such miniaturization on even smaller scales. Researchers are beginning to develop so-called nanoscale devices, the name implying a 1000-fold improvement over micrometer-scale achievements. But observers point out that a great deal of the planned development of nanoscale devices will rely heavily on the use of molecular self-assembled machines or engineered molecular structures—technologies that are themselves still quite young.

Density. Increasing the density of the arrays packed onto DNA chips is the third area in which manufacturers are making substantial progress. According to Mimms, current technologies enable manufacturers to fabricate single-chip arrays containing roughly 20,000 wells, with each well holding about 100 nl. One approach to attaining high densities is being studied by researchers at the National Institute of Standards and Technology, who have developed a chip made up of a single-molecule layer of DNA bound to a thin film of gold. The surface-tethered DNA then binds to a target strand and piezoelectric differences signal the level of hybridization. Tiny electrical charges, infrared spectroscopy, and scanning tunneling microscopy help to organize the ultrathin layers. Other researchers have displayed remarkable results with elastomer films. In fact, the progress in this area is so encouraging that Harvard University scientist George Whitesides has declared that "the days of silicon-based chips are numbered."

What is the theoretical limit of such arrays? That is a question routinely asked in this new field, as the scale of operations becomes ever smaller. One guess puts the theoretical boundary at 1 million elements on a 1-cm2 chip. Norman Nelson, senior research scientist at Gen-Probe, points out that "problems with hybridization worsen at higher densities. And there also is a point where the optics of detection reach their limits of resolution."

Integrated Chips. The final area of development is in the integration of sample preparation and analysis—the so-called lab-on-a-chip technology. "Today, all companies in this field recognize the importance of integration and are working on it," says Peter Wilding, director of clinical chemistry at the University of Pennsylvania. "To achieve this, these chips need to have a microfluidics platform."

"In order for chips to reach their full potential, they should evolve to include sample preparation, analysis, and signal acquisition," agrees Larry Kricka, also a professor in the pathology department at the University of Pennsylvania. "This form of a lab-on-a-chip will have many potential applications in both clinical and nonclinical areas."

Manufacturers are pursuing a variety of strategies to improve the microfluidics capabilities of chips. "Conventional capillary electrophoresis (CE) has been used to perform a diversity of analyses involving proteins, nucleic acid fragments, drugs, and so on. Such flexibility is a compelling reason for considering this technique for on-chip detection," explains Kricka. "CE and laser-induced fluorescence have been particularly popular for separation-detection combination. But other techniques have also been explored, including on-chip chemiluminescence, and electrochemiluminescence assays."

Not everyone agrees that lab-on-a-chip technologies are the coming wave, however. "I don't think the field will need to evolve to lab-on-a-chip," says Gen-Probe's Nelson. "Clinical laboratories will be willing to perform the preparation and analytical steps off-line, or by using a more extensive instrument." What then becomes critical, he and others add, is that DNA-chip technologies be compatible with existing equipment in such laboratory settings.

Trade-Offs. Of course, all these areas of technological progress are also interrelated. When detection sensitivity is improved, for example, less sample work is needed and therefore less miniaturization is required. Conversely, when greater miniaturization is achieved, sample preparation and analysis can be more easily integrated on a single chip, which can, in turn, improve sensitivity.

To help address these trade-offs quickly and reduce development times, a consortium composed of Caliper Technologies (Palo Alto, CA), Microcosm Technologies (Raleigh, NC) and the University of Washington has been awarded a $2 million Defense Advanced Research Projects Agency (DARPA) grant to develop computer-aided design (CAD) tools for microarray systems. The consortium hopes to develop a rapid-prototyping tool, called FlumeCAD, which will be similar to tools used in semiconductor manufacturing.

"FlumeCAD will enable this industry to create systems of hundreds to thousands of components and process steps, while meeting demanding time-to-market and reliability requirements," says John Gilbert, chief technology officer at Microcosm. The consortium's grant is part of DARPA's advanced CAD initiative, which has helped the private sector develop CAD technologies that could have both military and civilian applications. In this case, DARPA sees the need for industry to have a biochip-specific CAD tool system to help speed the development of advanced systems for the detection of biological and chemical warfare agents on the battlefield, notes Gilbert.

Commercial Hurdles

Much of the discussion thus far has focused on the future of the industry in terms of its technological developments. Whether DNA chips realize their potential, however, may depend even more on their progress in two areas quite remote from individual companies' product development breakthroughs.

Regulatory Requirements. The first area of concern is whether the FDA regulators of diagnostic products will be receptive to these strange new technologies. As DNA-chip products are developed for the clinical laboratory and even home-care settings, quality assurance in both design and manufacturing will become increasingly important issues, says Uwe Müller, director of advanced technology at Vysis Inc. (Downers Grove, IL). "Right now, FDA has no good way of judging what is an appropriate level of quality assurance in this field," he says. "What is needed is a good collaboration between the industry and the agency to consider these issues."

"At some point the whole industry will have to sit down at the table together to develop some basic standards and help FDA establish basic quality assurance requirements," agrees Deepak Thakkar, product manager for microarrays at Genometrix Inc. (Woodlands, TX). "Currently, each company in the field has its own corporate standards, which may differ from one another in significant details."

FDA legal expert Glen Freiberg, chief of regulatory affairs at Gen-Probe, also agrees. "As part of FDA's regulatory responsibility, developing good manufacturing practices requirements is an issue with any new technology," he says. "In this case, the challenge will be how to demonstrate reproducibility across the chip. There will have to be a lot of destructive testing to validate the manufacturing processes, because you can't verify these chips as you would computer chips."

Gen-Probe's Nelson suggests areas where both industry and FDA can begin. "The first level of quality assurance should be a physical inspection for material defects, patterned after the methods used in the computer chip industry. Then there will have to be a functional assay, where you use a standard set of samples for a given chip to determine an acceptable percentage base calling, or some other parameter. Perhaps there should also be a test of a certain number of the chips produced, using a randomized sampling."

Operational quality control could prove to be just as problematic as manufacturing quality assurance, but Kricka suggests a precedent that might become useful for both sides. "For devices with many test sites, a potential problem is the need to perform positive and negative controls for each and every test site. This issue has been addressed to some extent by the i-STAT analyzer that uses multitest cartridges." But Kricka also raises several other issues for FDA to resolve, including how isolated failures should be handled. Would the failure of a single element in a DNA array invalidate the entire chip device, or just the tests that failed?

To address some of these issues from an industry perspective, several large companies have launched a group called the Genetic Analysis Technology Consortium (GATC). This group intends to standardize array-based genetic analysis, paving the way for the more-affordable and productive development of diagnostic and therapeutic products. Affymetrix and Molecular Dynamics created the consortium to provide a unified technology platform capable of designing, processing, reading, and analyzing arrays of DNA. "As a result of the GATC, researchers will benefit as chips, readers, reagents, and software and database architectures that are GATC compliant will be compatible, eliminating the need for redundant equipment and software," said Edward Hurwitz of Affymetrix at the time the consortium was created in 1997.

Although they wonder whether the GATC effort will succeed, most familiar with the project praise its intentions. "It will be difficult to keep competing interests together and working on common goals, but these large companies are probably the only ones that can afford to drive this process forward," says Thakkar.

Reimbursement. The second major challenge facing DNA-chip manufacturers is perhaps also the greatest current problem for the industry as a whole: how to pay for the adoption of new technologies. "Many clinical labs just cannot do traditional [viral load] assays because reimbursement cutbacks have resulted in a lack of training and a lack of trained technologists," explains Richard Hodinka, associate professor of pediatrics and pathology at the University of Pennsylvania School of Medicine. And many laboratorians agree that Hodinka's comments about viral load tests could be applied to virtually any sophisticated laboratory diagnostic procedure.

One result of such pressures is that the need for automated systems is increasing—but not without regard for costs. To become accepted, new instruments and related technologies must make for less-expensive tests. "New diagnostic equipment must be cheaper, not necessarily in up-front costs but definitely when all factors, such as labor, are included," says CMS's Farkas. "When they are shown to be equivalent or cheaper, tests based on DNA chips will be on their way to wide adoption."

Making individual tests cheaper could be as much an information management challenge as it is a hardware issue, says Jonathan Pollack, a research fellow at Stanford University Medical Center. Part of Pollack's current work involves analyzing genomic information derived from microarray studies of genes implicated in cancer, using pattern-recognition software and other bioinformatics technologies. To reduce costs, he says, "we must find a better way to analyze the millions of points of data that microarrays can develop."

Conclusion

Because of the enormous challenges in meeting regulatory requirements and convincing clinicians and third-party payers of the cost-effectiveness of new DNA-chip–based testing, most observers do not expect these technologies to reach the clinical laboratory in significant numbers any time prior to 2003. Yet most are equally convinced that their migration from the research bench to the clinical setting is inevitable, and will happen soon after that year.

"By 2010—or sooner—I absolutely believe that we will see instrumentation using DNA chips in point-of-care applications," says Farkas. "This is a steam engine that no one group will be able to stop, because the advantages to everyone—in terms of better diagnosis and treatment and reduced costs to the whole system—are simply too great."

DNA chips will bring about a sea change in the way that some of humankind's most vexing diseases are diagnosed and treated. In some cases they will make it possible to move diagnostics out of the clinical lab, where traditional tests might take several days to perform, to the point of care, where results involving the detection of minute quantities of diseased cells will be provided within minutes.

The challenges facing manufacturers in bringing these technologies to the marketplace are enormous. However, virtually all observers agree that the development of DNA chips is accelerating at such a pace that those problems—even the problem of having a relatively high cost in a stingy payer environment—will soon be overcome.

Cliff Henke is a freelance writer based in Southern California.

Nanogen uses microelectronics for genetic analysis

Kieran Gallahue

Nanogen has established itself in the rapidly expanding biochip industry through its method of integrating the power of microelectronics with the promise of genetic analysis. This integration of electronics and genetics will enable researchers to accomplish all of the key requirements for a sample-to-result solution: speed, portability, integrated sample processing, genetic analysis, and single base-pair discrimination capabilities.

The company's long-term goal is to place the power of genetic analysis in the hands of doctors, pharmacists, and patients. Using the microelectronic chip as a base, a handheld sample-to-result device is planned for use by any of these groups. From a simple sample such as whole blood, the user would confirm patient identity, diagnose diseases from a panel of alternatives, determine therapeutic toxicity and resistance factors, and perform other genotyping analyses associated with specific health conditions. A wireless link of the device to a secured Internet database would allow evaluation of test results by physicians and other experts. The doctors would then prescribe therapies for a disease and later, repeating the above steps, monitor the effectiveness of treatment. It may even be possible for pharmaceutical companies to monitor outcomes and establish a process of continual product improvement and patient targeting.

A packaged 100-site microarray by Nanogen, with a built-on flow chamber. Photo Courtesy of Nanogen Inc.

Ultimately, fewer trips to doctors' offices, pharmacies, hospitals, and labs would be required, and prescriptions would be more accurate. Measured across a large population, the total cost of healthcare would be substantially reduced, as well as the time people spend away from work due to illness. Healthcare itself would improve, as a consequence of the convenience and availability of state-of-the-art diagnosis and testing.

Nanogen is not waiting for the handheld device to begin delivering solutions. The technology will be available in multiple products as the markets and applications develop. The company plans to deliver a benchtop instrument system in 1999, followed by a laptop PC–sized instrument, and then the handheld device.

Nanogen's process deposits DNA and proteins onto active microelectronic chips. The company is systematically addressing each step in the sample-to-result process on microelectronic chips, from sample processing to data transport. Papers have already been published demonstrating electronic cell separation, electronic sample transport, electronic hybridization, and electronic stringency. Hybridization occurs in seconds or minutes on Nanogen's chips; alternative biochip systems take hours. Work continues on integrated, electronically enabled DNA amplification with Nanogen's infectious-disease diagnostics joint venture partner, Becton Dickinson and Co.

Kieran Gallahue is vice president for strategic marketing at Nanogen Inc. (San Diego).

Microfluidics emerging at Aclara Biosciences

Travis Boone

Aclara Biosciences Inc. is a pioneer and leading developer in the advancement and application of electronically controlled microfluidics, often referred to as lab-on-a-chip technologies. The company's proprietary microfluidics technology enables the accurate measurement, dispensing, and mixing of liquid volumes as small as 10–10 L (one millionth the volume of a well in a standard 96-well microplate) within interconnected microcapillaries on its single-use, plastic microchips. The company's microchips are trademarked under the Labcard name.

Device Fabrication

To date, most microscale devices that employ electric fields for fluid transport have been fabricated from glass or silica. However, when single-use devices are desired, it is unlikely that glass fabrication will be cost-effective. Aclara has developed proprietary technology for the manufacture and use of polymeric devices in microfluidic applications, with the goal of producing single-use chips for a variety of applications. Aclara's plastic devices are fabricated by molding microchanneled substrates from micromachined masters and then sealing with a second polymer layer.

A variety of microfluidic formats have been developed, including discrete rigid cards (see Figure 1) and continuous flexible films (see Figure 2). Aclara's techniques for producing single-use plastic microfluidics are well suited for high-volume manufacturing (millions per year). The company expects that such plastic devices will be an order of magnitude more cost-effective than comparable glass devices.

Performance

Aclara has demonstrated on plastic devices the same valveless electrokinetic fluid control, variety of assays, and chip-to-chip reproducibility that is achievable with glass devices. The company believes that its technology platform will provide significant increases in throughput for a wide range of chemical and biochemical analyses, while also reducing the use of costly reagents. These improvements will be accompanied by an overall decrease in the operating costs and complexity associated with such analyses.

Figure 1. A single-use, rigid, plastic microfluidic card for multiplexed assays involving the incubation of microliters of reagents and the subsequent electrophoretic separation of nanoliters of product for optical detection.

Integration and Automation

Although advances in genomics, combinatorial chemistry, and assay technologies have created significant opportunities for discovering new therapeutics and understanding the roles and functions of genes in disease, they have also highlighted the limitations of existing technologies. New analytical systems have been developed in response to such limitations; however, these systems have offered only incremental improvements, including improved throughput and automation, over existing systems. In addition, all techniques for analyzing nucleic acids—including automated electrophoresis, mass spectrometry, and DNA microarrays (hybridization chips)—require a significant number of steps for sample preparation. As a result, existing systems—which make use of such manual and time-consuming laboratory processes as centrifugation, thermocycling, filtration, measuring, mixing and dispensing—are prone to human errors that compromise test data and result in increased costs.

Figure 2. Multiplexed plastic microfluidics in a flexible film format. Sample reservoirs register to the wells of standard microplates. Photos Courtesy Aclara Biosciences Inc.

To address these critical issues, Aclara is developing integrated, automated microfluidic systems in which a number of sample preparation and analytical steps—such as mixing, reaction, and separation—are performed on a single-use, plastic Labcard. The design of the Labcard is intended to increase laboratory productivity, reduce reagent and sample consumption, and decrease sample-preparation errors. For certain applications, integrated Labcard systems have the potential to replace many conventional laboratory systems, such as slab gel instrumentation, thermocyclers, and centrifuges. Aclara expects that its microfluidics technology will expand the market for chemical and biochemical information by making the acquisition of such information more readily attainable and reliable—and less expensive.

Commercialization Strategy

Initially, the company intends to apply its microfluidics technology in the areas of genetic analysis and drug screening, optimizing its technology through collaborations with partners that possess hardware- and software-development expertise, complementary intellectual properties, and global distribution capabilities. In April 1998, Aclara entered into a collaborative agreement with Perkin-Elmer to jointly develop genetic-analysis systems that combine Perkin-Elmer's fluorescent readers, reagents, and software with Aclara's Labcards and chemistries.

Travis Boone, PhD, is a senior research engineer leading microfluidic design and rapid-prototyping efforts at Aclara Biosciences Inc. (Hayward, CA).

Clinical Micro Sensors developing point-of-care molecular diagnostics

Daniel H. Farkas

Clinical Micro Sensors Inc. (CMS) was organized in 1995 to capitalize on a bioelectronic DNA detection technology developed at the California Institute of Technology by the company's founder, Jon Faiz Kayyem, with Thomas Meade. The essence of the technology rests on specific detection of the electrical current generated by the reversible and continuous oxidation and reduction of labeled nucleic acid targets. Refinements of this system will have wide-ranging applications in clinical diagnostics, blood-safety monitoring, food safety, biomedical research, pharmacogenomics, and the detection of biowarfare agents.

The CMS technology exploits the electronic properties of nucleic acids while relying on the central principle of clinical molecular diagnostics, namely, hybridization. In the CMS technology, the solid support used as the original tool of molecular diagnostics—the nitrocellulose or nylon membrane of the Southern blot—is replaced by a DNA microchip. CMS microchips contain numerous electronically active "pads," each embedded with specific DNA capture probes. The number of pads and specificity of probes on a chip can be varied depending on the applications of interest. The chips also include redundancy for positive controls and nonsense sequences as negative controls.

To serve as an electron donor, a ferrocene organometallic complex is covalently attached to the single-stranded DNA capture probes, or to a signal probe that is complementary to a different region of the target of interest. The signal probes serve to label the target upon hybridization and are called AMBER probes (AMBER is an acronym for amperometric bioelectronic reporter). The capture probes are attached to a gold microelectrode through phenylacetylene "molecular wires" that maintain the desired electrical contact between the probes and the surface of the gold electrode. Except for these protruding molecular wires, however, the microelectrode surface is electrically insulated with a monolayer coating of alkane thiols. This coating prevents unwanted redox species in the sample chamber of the chip from interfering with the measurements of the test system. Generation of a signal, therefore, depends on specific probe-target interaction—that is, hybridization—which is transmitted through the molecular wire.

Figure 1. Working prototype of the CMS handheld reader. The reader applies microvoltage to a DNA microchip to initiate sensing (one such chip is shown inserted in the top of the reader and three others appear in the photo). Test results are generated in minutes and are displayed on the reader's built-in screen. The reader also includes an infrared port for data transmission, which appears as a small, dark oval to the right of the inserted chip. CMS's chips contain electronic pads embedded with DNA capture probes. Photo By Henry Blackham

When a slight voltage (in the millivolt range) is applied to the sample-containing microchip following hybridization, the ferrocene labels release electrons that rapidly travel through the double-stranded nucleic acid hybrid and molecular wires, yielding an electronic signal that can be detected through the microelectrode. CMS's proprietary handheld reader supplies the voltage that initiates these events, and has onboard software that interprets the electrical signal to both identify and quantitate, if appropriate, the target nucleic acid (see Figure 1).

Molecular diagnostics is a burgeoning field that will ultimately bring dramatic changes in the ways that laboratory medicine and clinical diagnostics are practiced. In 1999, however, molecular diagnostics represent only a tiny fraction of the total volume of diagnostic tests performed in hospitals, and an even smaller percentage of those performed in physician office laboratories. There are several reasons for such low volumes of molecular testing. The amplification procedures that define the gold standard of molecular diagnostics (e.g., polymerase chain reaction, ligase chain reaction, and so on) are complex, time-consuming, and expensive, and often require not only dedicated space but also the most highly trained, specialized medical technologists. Furthermore, enzymatically based amplification methods can be inhibited by certain components of the specimen under analysis. Relatively complicated and expensive specimen preparation technologies must often be used to prepare patient specimens for molecular diagnostic tests.

Extensive experimentation at CMS has demonstrated that the bioelectronic detection of DNA described above is not inhibited by the presence of serum, whole blood, saliva, or urine, and can even be performed on such samples as soil extracts and hamburger. The CMS technology, therefore, can be adapted to applications in molecular diagnostics and other markets without the need for extensive specimen preparation. Sample preparation—consisting of simple lysis of the cells, bacteria, or viruses of interest followed by denaturation and fragmentation—will not need to be augmented by significant purification. Ultimately, the company intends to integrate lysis of the specimen and presentation of the nucleic acid into the system. A complete test will then be performed by introduction of the specimen, a few moments of incubation for lysis and nucleic acid denaturation and fragmentation, several minutes of incubation for hybridization to occur, signal generation (initiated by pushing one button to generate the necessary voltage), and reading the result displayed on the device.

The CMS DNA microchips are inexpensive to manufacture, resulting in cost savings that will help to make molecular diagnostics more widely available. Progress on the sensitivity levels of the system has been dramatic. For many current applications, detection levels of 1000 to 100,000 targets is satisfactory. As CMS moves toward even more-sensitive detection levels, the requirement for enzymatic nucleic acid amplification may be eliminated or reduced. Coupling CMS's technology to microfluidics may also be useful in improving sensitivity by concentrating specimens.

The potential speed, sensitivity, ease of use, and low cost of CMS's bioelectronic DNA detection system are all factors that will contribute to its broad adoption as a new platform for nucleic acid detection. In medical applications, the system could enable physicians to administer antibiotics or antivirals rationally and specifically with point-of-care availability of results. CMS chips could comprise panels that cover the entire differential diagnosis associated with given symptom sets, such as respiratory infection panels, meningitis panels, sexually transmitted disease panels, and so on. The CMS platform could also be adapted to instruments designed for the central testing laboratory, enabling laboratorians to assay many patient specimens per batch.

Daniel H. Farkas, PhD, is director of clinical diagnostics at Clinical Micro Sensors Inc. (Pasadena, CA).

Incyte's gene expression platform

Melinda Baker

Differential expression data generated by microarray experiments represent a critical element in the race to develop gene-based diagnostics, drugs, and other therapeutics. By mining this data, companies gain the potential to shorten product development cycles and bring a higher percentage of successful products to market more quickly.

Incyte's GEM gene expression microarrays are part of a platform of products—including dedicated expression analysis software and reagent kits—which enable researchers to rapidly generate and explore huge quantities of gene expression data. Incyte microarrays, each of which contains either Incyte-proprietary or public domain clones, can help researchers:

* Identify and validate promising gene targets for drug
development.

* Identify previously unknown genes.

* Understand the behavior and interactions of genes in both diseased and healthy cells.

* Assess the efficacy and toxicity of compounds before beginning clinical trials.

Incyte released several new microarrays in 1998, and plans to release more throughout 1999.

10,000-Element Microarrays

Currently, up to 10,000 DNA samples, or clones, can be arrayed on each microarray. The actual number used depends on both the size of the genome and the number of high-quality clones available to represent a genome or particular area of study. DNA samples can include genes with known functions, giving researchers the ability to see how varying conditions affect expression levels. DNA samples can also include gene fragments or expressed sequence tags (ESTs) whose function is unknown, enabling researchers to identify new genes and gene functions.

Competitive Hybridization

Incyte's microarrays measure differential gene expression through two-color competitive hybridization, allowing researchers to quantify expression changes in healthy versus diseased samples, treated versus untreated samples, and in samples submitted to varying time-course experiments.

DNA samples with an average length of 1000 base pairs are generated by polymerase chain reaction (PCR) and arrayed on a glass surface before denaturing. Fluorescent-labeled probes with complementary sequences bind to the samples through competitive hybridization. Highly sensitive, these microarrays have a detection limit of 1 in 100,000 (approximately 1 to 10 copies per cell), and require as little as 200 ng of fluorescent-labeled sample per experiment.

To help ensure the highest degree of accuracy and reproducibility for each experiment, Incyte fabricates its microarrays and performs all experiments in state-of-the-art cleanroom facilities. Quality controls include bar coding of every sample, automated testing of all batches, and careful tracking of each sample through each stage of the experiment.

Associated Databases

Proprietary and public domain databases are important tools that can be used in tandem with microarray experiments, especially by researchers conducting in vitro experiments to further their understanding of particular genes. Databases can also be used to identify gene homologues, and to conduct a variety of database-wide analyses. Incyte databases include both proprietary and expertly edited public domain sequence information, relative expression levels in the tissue libraries from which the Incyte-proprietary sequences were generated, and annotations about putative and known gene functions.

Each Incyte microarray can contain up to 10,000 elements. Photo Courtesy of Incyte Pharmaceuticals

Incyte is developing additional microarrays that use clones represented in its extensive human, microbial, animal model, and plant databases. For example, microarrays associated with Incyte's Lifeseq human gene sequence and expression database (which includes more than 2.4 million proprietary sequences) include "encyclopedic" microarrays that represent all reagents available in a given set, and "theme" microarrays for the study of particular areas such as blood-inflammation or cancer-signal peptide mechanisms.

Incyte also created the Unigem V microarray, which contains sequence-verified, Incyte-proprietary cDNA clones mapped to the publicly available Unigene database at the National Center for Biotechnology Information. This single microarray makes it possible for researchers from a broad audience (including those from academic and small-industry environments) to conduct cost-effective expression analyses on the majority of known human genes in the public domain.

Expression Analysis Software

Without sophisticated computer-based analysis tools, exploration of the thousands or even millions of data points resulting from microarray experiments could virtually halt, rather than speed, research and testing. Expression analysis—as well as sequence analysis, data storage, project management, and other areas of biological research—fall into the larger field of bioinformatics—the use of computers to retrieve, process, analyze, and simulate biological information.

Currently, Incyte has two software programs dedicated to analyzing expression data from microarray experiments. With Incyte's Gemtools software, scientists can perform a variety of sophisticated analyses on data collected from microarray experiments. Incyte is also adding new capabilities to its Lifearray expression analysis software, which enables users to analyze data collected from a variety of microarray platforms.

Future Directions

Incyte will continue to develop new microarrays with its existing technology, the potential of which is still relatively untapped. Microarrays currently available and in development include human, microbial, plant, and animal model microarrays; microarrays designed from public domain human and animal model clones; and custom arrays designed for customers with particular research needs.

In addition, Incyte plans to sequence verify all clones on its microarrays to ensure that all elements on the array are properly identified. Sequence verification performed by Incyte will save researchers from having to perform time-consuming sequence verification themselves, adding further value to the Incyte platform.

Incyte also plans to pursue improvements to its technologies that could make it possible for a greater number of clones to be included on each microarray. As in all areas of genomics research, Incyte continues to explore new technologies for understanding the molecular basis of life.

Melinda Baker is promotions manager at Incyte Pharmaceuticals Inc. (Palo Alto, CA).

Sequencing technologies at Hyseq

Jose C. Carle-Urioste

The Hychip product, a universal microarray DNA-chip system, was developed by Hyseq in collaboration with Perkin-Elmer.

Figure 1. In the Hychip system, capture probes on the chip surface hybridize to target DNA and are covalently bound to labeled probes that also hybridize to the target DNA in an adjacent position. A wash sequence removes labeled probes not bound to the capture probes. Illustrations Courtesy of Hyseq Inc.

DNA sequencing takes place in a chip containing all 1024 possible 5-mer oligonucleotides (capture probes) attached to its surface by their ends. Target DNA is incubated with a set of labeled 5-mers (labeled probes) and the enzyme ligase. When a capture probe hybridizes to a complementary sequence in the target DNA and a labeled probe hybridizes to the immediate sequence of target DNA that is adjacent to the capture probe, ligation between the capture and labeled probe occurs (see Figure 1).

After ligation, the DNA target, as well as the labeled probes that did not bind, are washed under stringent conditions. Since the labeled and capture probes are covalently linked, the washing step can be performed at high temperatures. Signals only result where ligation has occurred. The final sequence is assembled through inherent multiple reads of every base via overlapped probes using proprietary software (see Figure 2).

Figure 2. Accuracy of sequence determination is improved by having 10 overlapping probes to hybridize with each base of the DNA target.

Flexible Application

Since it contains all possible 5-mer oligonucleotides, the Hychip product can be used to sequence any gene—only one chip is needed for a wide variety of applications and targets. Other technologies require previous knowledge of the sequence and of the mutations to design the chips, and at least one chip has to be made for each gene.

The use of probe arrays that include all 1024 pentamers combined with 1024 pentamer-labeled probes in solution has the potential to generate all possible 1,048,754 10-mer combinations without having to synthesize 1 million probes on a chip. All probes undergo stringent quality control procedures and result in a level of confidence that is not achieved by chemical synthesis of probes directly on a chip. The probes used in the production of the Hychip system are manufactured by Perkin-Elmer.

Early Access Program

Hyseq and Perkin-Elmer are now making the universal DNA-sequencing Hychip system available to selected pharmaceutical and academic institutions in an early access program for research and diagnostics.

Once an organization evaluates and is completely satisfied with the Hychip results, participation in the early access program will permit evaluation of the technology in a participant's facilities with prototype equipment.

Jose C. Carle-Urioste, PhD, is a business development analyst at Hyseq Inc. (Sunnyvale, CA).