Courtesy of Caliper Technologies
Imagine visiting the doctor's office for a routine annual checkup. Instead of drawing several vials of blood for analysis by an outside diagnostics lab, the doctor collects a single drop. Using a breadbox-sized instrument, she runs 20 or so tests in a matter of minutes and discusses the results with you before you leave. Meanwhile, a large pharmaceutical company down the road is using a similar, albeit larger instrument to analyze the biochemical properties of a million potential lead drug compounds per hour, using picoliter samples.
Such is the promise of so-called "lab-on-a-chip" technology, which combines microfluidics (liquid handling on a nanoliter to femtoliter scale) with microfabrication techniques developed by the semiconductor industry. The resulting chips, some no larger than a computer microprocessor, perform biochemical reactions from many tiny samples in parallel, in theory reducing costs and conserving reagents while greatly improving speed and reproducibility for both diagnostics and drug discovery. And, they fit easily into automated workflows.
Yet the technology has largely failed to match its potential, says Colin Brenan, chief technology officer of Biotrove, a microfluidics company based in Woburn, Mass. "Over the past several years," he says, "microfluidic companies have over-promised and underdelivered." But Brenan remains optimistic about the technology's future. "Microfluidics ... will be beneficial in reducing healthcare costs, by decreasing the time and cost of pharmaceutical drug discovery and by delivering sophisticated point-of-care diagnostics at a reasonable price," he asserts.
Jeff Jensen, chief operating officer of Eksigent Technologies, Livermore, Calif., agrees. "The new crop of devices that are being introduced have all the right attributes: robustness, ease of use, fitting into existing process workflows. [But] progress will continue to be evolutionary rather than revolutionary."
According to Steve Edwards, an independent nanotechnology consultant, the slow pace of lab-on-a-chip adoption has both external and internal causes. Externally, microfluidics has fallen victim to the overall downturn in the pharmaceutical industry. "Pharma companies have been hit hard by drugs going off patent," he says, which has limited their ability to invest in new technologies. Internally, the problem is one of perspective. "These companies are started by scientists and engineers with an upside-down view of the world," says Edwards. Instead of finding a need and filling it, they have developed products based primarily on the capabilities of their technology.
Some companies, though, have turned their technological strengths to unmet application areas. Staying mindful of their customers' need to integrate microfluidics with preexisting technology, these firms are not merely surviving. They are thriving.
FINDING A NICHE
One such company is AVIVA Biosciences of San Diego, whose platform enables high-throughput screening of ion-channel drugs. The company's SealChip allows patch clamping of 16 cells at one time in a microfluidics platform. It is, says vice president for R&D Jia Xu, a "pretty unique technology." Working together with Axon Instruments of Union City, Calif., AVIVA has nearly 30 PatchXpress instruments installed at major pharmaceutical companies, including Pfizer, says Xu.
AVIVA is also pursuing a noninvasive alternative to amniocentesis, a diagnostic test used in high-risk pregnancies that has a low but real risk of miscarriage. Normally, fetal cells are isolated for genetic analysis by collecting amniotic fluid from the womb with a large needle. AVIVA's microfluidic system, currently in testing at Baylor School of Medicine in Houston, can instead isolate the rare fetal cells that make their way into the mother's bloodstream. Other potential applications include isolation of cancer cell metastasis detection and stem cell isolation.
Fluidigm of South San Francisco has found its own niche in protein crystallography, an application that is both "up and coming" and "uniquely enabled by our technology," says Jaime Bartel, public relations manager. Drug companies need three-dimensional protein structures for computer-aided drug design. But successful protein crystallization requires screening hundreds of conditions, a time-consuming and reagent-intensive task.
Using small amounts of protein, Fluidigm's TOPAZ 1.96 chip can screen 96 conditions at one time for their ability to promote crystal growth. A 384-condition variant is expected soon. The company also has a growth chip that can produce the larger crystals needed for X-ray diffraction. Among Fluidigm's collaborators is GlaxoSmithKline, which uses the system to crystallize its drug targets, Bartel adds.
The time and expense required to produce large amounts of biological material, such as proteins or enzymes for drug screening, is a major reason for the interest in microfluidics. By reducing the amount of material needed, companies can reduce the overhead costs of drug development, and potentially gain speed as well.
Michele Boudreau, director of corporate communications at Caliper Life Sciences, Mountain View, Calif., cites a customer who was able to perform a desired screen using $70,000 (US) worth of protein, instead of the $7 million worth that conventional technologies would have required. In this case the cost savings turned an impractical idea into a viable research avenue.
SHRINKING CELL CULTURES
Cell-based assays in particular are time-consuming and reagent-hungry. This can be especially problematic for primary cells, which can be difficult to produce in large numbers. Caliper is one of a number of companies pursuing this need. According to Boudreau, Caliper's LabChip 3000 can screen drugs against such targets as G protein-coupled receptors, using cell-based calcium-ion flux assays.
Courtesy of Caliper Technologies
Caliper creates its microfluidic LabChip devices using a microfabrication process similar to that found in the semiconductor industry. Microfabrication makes it possible to create intricate designs of interconnected channels that are extremely small and precise.
Cells are driven through tiny channels and exposed to a library of compounds. Calcium influx is measured, one cell at a time, using fluorescence. The chip uses "one roller bottle of cells in place of the fifty" that would be needed to screen the same number of compounds using conventional microplates, says Boudreau. The same chips can also perform protease, kinase, and phosphatase screens.
BioProcessors of Woburn, Mass., also addresses the need for faster, cheaper cell cultures. Instead of using the cells to screen drugs, the company uses the cells to make them. BioProcessors has developed microfabricated bioreactors in which cells can be grown, harvested, and analyzed in numerous tiny, parallel cultures. These "SimCells" allow rapid testing of multiple growth conditions for their effects on drug production. And they can be adapted to grow different types of cells, including bacterial, fungal, and mammalian.
In the future, cell-based systems may become even more complex, integrating more functions onto a single chip. Pennsylvania State University associate professor Michael Pishko is working to develop chips containing "arrays of different phenotypes of cells." For example, a single chip might contain liver, immune, and vascular endothelial cells, he says. Compounds flow past the different types of cells, and fluorescence is used to detect their effects on various aspects of cell function, such as nitric oxide production. These chips could be used in cytotoxicity assays, Pishko says.
High-performance liquid chromatography (HPLC) is another tool that is making the transition to microfluidics. Companies such as Eksigent are using lab-on-a-chip technology to make HPLC "faster and multiplexible," says Jensen.
Eksigent is an anomaly in the field, Jensen says, in that it has "integrated microfluidics technology into a hybrid system" that does not involve an actual chip. He says the original focus on complete lab-on-a-chip strategies limited the potential benefits of the microfluidics technology. "Is there inherent benefit in the chip format? Does everything have to be on a chip?" he asks. His answer: "If the chip limits performance, don't use it. [Researchers] want the system that gives them results."
Eksigent's NanoLC system performs HPLC 10-times faster than do conventional systems, Jensen says, using currently available liquid-handling technologies. This acceleration should allow researchers to expand their use of HPLC to investigate the physical and chemical properties (such as solubility and pKa) of many more drugs before they enter the development pipeline, he adds. Eksigent has also developed the NanoLC-2D system for proteomics applications.
Another player in the HPLC arena is Nanostream of Pasadena, Calif., which officially launched its Veloce Micro Parallel Liquid Chromatography System this past February. The system can run 24 reverse-phase separations in parallel with real-time UV detection.
BEST OF BOTH WORLDS:
Courtesy of Network Biosystems
Network Biosystems' large-scale bioMEMS device combines the benefits of microfluidic miniaturization with the performance of traditional instrumentation.
One reason "microfluidics technology has failed to deliver is due to the cost of adopting proprietary interfaces" for the chips, says Biotrove's Brenan. Many drug companies are heavily invested in liquid-handling devices, robots, and instruments designed to use a standard microplate format. Biotrove's "Living Chip" performs microfluidics-based high-throughput screening while retaining this standard format. Biotrove has chosen PCR-based analysis of single nucleotide polymorphisms (SNPs) as its first commercial application. Brenan says technologies like this should make a "huge difference in researchers' access to genomic data" by greatly reducing per-SNP costs. The system is being tested by researchers screening for schizophrenia genes at Massachusetts General Hospital in Boston.
Another company adapting PCR to the microfluidics arena is industrial heavyweight STMicroelectronics, a large European microchip maker. The company has teamed up with Mobidiag, of Helsinki, Finland, to explore DNA-based diagnosis of infectious diseases. This technology could eventually lead to lab-on-a-chip diagnostic instruments for use in doctors' offices and hospital emergency rooms.
Applied Biosystems of Foster City, Calif., is also new to the microfluidics arena. Its TaqMan Low Density Array runs up to 384 real-time PCR reactions simultaneously for low-density gene-expression analysis. Martin Johnson, assistant professor of pharmacology and toxicology, University of Alabama, Birmingham, uses the system to screen for genes implicated in chemotherapeutic drug resistance. "We're very excited about the new technology," he says, noting it helped his team whittle a list of 470 candidate genes down to a "workable" II.
PCR also drives the PicoTiterPlate developed by 454 Life Sciences of New Haven, Conn. Containing nearly 300,000 75-picoliter wells, the plate enables massively parallel whole-genome sequencing, including sample preparation, amplification, sequencing, and bioinformatics in a single process. The company tested the chip by sequencing the 30-kilobase adenoviral genome in less than a day, from sample preparation through complete analysis.
This technology could eventually be applied to larger bacterial and human genomes. Rapid analysis of bacterial genomes will have applications in biodefense and in the diagnosis of infectious disease. Rapid analysis of individual human genomes could facilitate "designer medicine," in which a patient's treatment is tailored to his or her genetic background.
REALLY MINI GELS
Cell-based assays, HPLC, and PCR represent some of the more complex research techniques currently being developed in chip form. A number of companies also have successfully downsized the simpler technique of gel electrophoresis. Leo Kretzner, of the City of Hope National Medical Center, Los Angeles, has been using the Agilent 2100 Bioanalyzer from Agilent Technologies of Palo Alto, Calif., for several years to detect molecular markers for prostate cancer. "We use it for just about anything you'd run a gel out for," he says, adding that it is faster, uses less sample, and is more reproducible than traditional gels.
He does note several limitations. For one, there is no actual gel to use for further analysis by Northern, Southern, or Western blotting. Also, there is no way to recover the samples for use in downstream procedures such as subcloning. Overall, though, he finds the chip has "more strengths than limitations."
"Gel chips" have applications out side of academia, too. Fermentas of Vilnius, Lithuania, a supplier of restriction enzymes and gel markers, uses the Agilent 2100 Bioanalyzer in its quality control procedures. And Agilent has developed software for the Bioanalyzer that allows it to comply with regulations for the development and manufacture of protein-based therapeutics by pharmaceutical companies.
ECONOMIES OF SCALE
Kretzner's lab analyzes samples in quantities of tens and hundreds; but genomic and proteomic research require the analysis of thousands of samples. Several companies have stepped in with technologies aimed at separating biomolecules on this scale. Gyros, of Uppsala, Sweden, has developed a compact disc (CD)-based instrument, the Gyrolab, that can process 480 protein samples in parallel in about an hour, for use in peptide mapping or sequence analysis by mass spectrometry.
The CD format allows for a simple technology: Spinning the disc forces the samples through tiny channels by centrifugal force. Another disk, the Bioaffy, performs protein quantification by immunoassay. Researchers at George Washington University in Washington, DC, use the Bioaffy CD to identify molecular markers for brain damage in new born infants. Tecan, of Maennedorf, Switzerlands has developed a similar technology. In this case, the company has targeted the drug-discovery process known as ADMET, or absorption, distribution, metabolism, excretion, and toxicity assays of drug candidates.
Courtesy of Gyros
Another company in Woburn, Mass., Network Biosystems (formerly known as genoMEMs) has developed a technology that combines microfluidics-scale samples with a large-scale separation "chip," fabricated from glass. This large chip can separate 384 microscale sequencing reactions simultaneously, but uses a macroscale path length of 30–40 cm, comparable to that of traditional sequencing gels. The longer separation distance greatly increases the resolution and allows better integration with robotics and liquid-handling systems, according to the company.
At least part of the appeal of microfluidics systems is the idea of integration, performing all the tasks for a given procedure on a single chip. The companies above have met this challenge to varying degrees. But Cepheid, of Sunnyvale, Calif., concentrates on the sample preparation process itself. Cepheid is developing cartridges that use both macrofluidic and microfluidic technologies to extract DNA samples from real-world sources, such as blood, human tissue, soil, and food.
Such technology could well be the missing link between the macroworld of biological samples and the desired microworld of fast, cheap, and user-friendly labs-on-chips. Making this link in a way that is compatible with existing research infrastructures is likely to spell success for microfluidics companies, providing they pursue real-world applications. Says AVIVA's Xu, "Start with an unmet need and you can be pretty sure there will be some buyers."
Megan M. Stephan