Incredible Shrinking Labs: Chipping Away at Analytical Costs
ORNL is developing chips ranging in size from a match tip to a match box for health and environmental applications.
Mike Ramsey shows an early version of the lab on a chip.
In 1987 ORNL’s Mike Ramsey proposed carving a winding tube in a glass chip, injecting a few drops of liquid into the tiny tube, and “pumping” the liquid through it using an applied electric field. His colleagues laughed at him. They didn’t realize that, thanks to micro-fabrication techniques used to make silicon chips for electronic devices, it was possible to construct a “lab on a chip”—a miniature device for separating molecules and measuring chemical reaction rates using very small liquid samples.
Such a postage-stamp-size chip based on ORNL designs is now being tested and commercialized for DNA and protein analysis by Caliper Technologies, Inc., a California company on whose scientific board of directors Ramsey sits. The chip can analyze DNA and small molecules for such uses as DNA sequencing, DNA fingerprinting of blood at crime scenes (see the Bytes Help Take the Bite out of Crime article), environmental monitoring, detecting chemical warfare agents, and developing new drugs.
Thanks to internal lab funding for his idea, in 1991 Ramsey and his colleagues in the Chemical and Analytical Sciences Division (CASD) began developing “microfluidic structures” and demonstrating that they can separate chemicals in very small volumes. He also gave scientific talks promoting the concept.
“Just as the electronic industry shifted from vacuum tubes to transistors to microelectronic chips,” Ramsey says, “I felt that the chemical analysis community would have a similar motivation to reduce analytical instrument size from a few cubic feet to several cubic centimeters. Such miniaturization offers low cost, low weight, high speed, high reliability, and operational simplicity. For chemical analysis it saves materials and labor and reduces discharges to the environment.”
In the lab chips, molecules rather than electrons flow through labyrinths of tiny channels and chambers outfitted with valves, filters, and pumps. In 1994 Ramsey and his colleagues published a paper in the journal Analytical Chemistry on how to use electric fields to make liquids glide quickly around the hairpin turns of a channel on a dime-size “serpentine chip.” The liquids carry a charge because, in a capillary, the liquid pulls protons off the glass wall. By applying the electric field to all chip ports simultaneously, the ORNL lab on a chip avoided the leakage problems of a similar chip designed by the Swiss group that first published a paper on the concept.
A lab on a chip can analyze a volume of liquid 10,000 times or more smaller than that used in a conventional analytical instrument. Smaller volumes permit faster mixing because molecules don’t have so far to travel. Chemicals in a liquid droplet can be mixed and separated very rapidly. In fact, in an “Oh wow” experiment, Ramsey and his ORNL colleague Steve Jacobson separated two species in less than a millisecond—100,000 times faster than conventional methods.
The pharmaceutical industry prizes analytical techniques that use small samples and quickly produce results. Generating large quantities of enzymes for drug design experiments is very difficult and expensive. Now, scientists no longer must synthesize and test compounds one at a time to see which will block an enzyme related to a disease state. Instead, they can make small quantities of many related substances and, using many lab-on-a-chip systems, screen them simultaneously for biochemical activity.
The ORNL group recently tested the ability of a lab on a chip to determine the effectiveness of various inhibitors in blocking enzyme reactions with a substrate. The test enzyme was reacted with a sugar, producing a chemical that emits light. A laser light was shone on the product, causing it to fluoresce. By determining the strength of the fluorescence, a computer linked to a light detector measured how fast the product formed. When an inhibitor was introduced into the mix, it blocked the reaction between the enzyme and sugar, reducing the rate at which the fluorescent product was generated. The inhibitor that slowed the reaction rate down the most might be the best candidate for an ingredient in a disease-fighting drug.
Parallel lab-on-a-chip systems also may offer the most economical way to rapidly analyze large DNA samples to determine the sequence of their chemical bases. Chips promise to carry out repetitive sequencing tasks for tens of thousands of genes more quickly and cheaply using smaller samples than other technologies.
Chips based on ORNL designs are already being used in desktop instruments linked to personal computers to provide a compact analytical instrument. The lab on a chip is small, but its potential is large.
Biochips for the Doctor’s Office
Tuan Vo-Dinh inspects an early version of the DNA biochip, now called the multifunctional biochip. It holds antibodies and proteins, as well as DNA, for disease detection in the doctor’s office someday.
You feel terribly ill, so you call your doctor. You are asked to come to the waiting room where the nurse takes a drop of blood from your finger. After about 40 minutes, you are escorted to the doctor’s office. The doctor says, “I have good news for you. According to this biochip, you don’t have tuberculosis, the AIDS virus, or any signs of genetic predisposition to cancer. You just have a bad case of the flu.”
Someday your doctor may be relying on a matchbox-size diagnostic device invented at ORNL. This “multifunctional biochip” has been developed by Tuan Vo-Dinh of ORNL’s Life Sciences Division (LSD), in collaboration with Alan Wintenberg and Nance Ericson, both of ORNL’s Instrumentation and Controls (I&C) Division. It will provide quick results using only a drop of blood. It will be highly selective and sensitive, able to distinguish between a bacterium and virus or between a chemical and biological organism.
“By integrating microelectronics, optics, and biological material in a single system,” Vo-Dinh says, “we have developed a second-generation device—a multifunctional biochip that will be able to simultaneously detect a variety of biomedical targets. One chip may hold specific DNA sequences, antibody probes, and protein receptors. It will process up to 100 samples in 30 minutes. It will be useful for DNA sequencing; gene identification and mapping; screening blood, vaccines, food, and water supplies for infectious agents; and diagnosis of diseases, including AIDS, hepatitis, genetic cancers, and Alzheimer’s disease.”
The multifunctional biochip has multiple components in a miniature format, including a sampling platform, excitation sources, and electro-optic sensor arrays. Here’s how it would work with DNA. On the sampling platform are attached short DNA fragments, each of which has a different sequence of chemical bases. The chip site and sequence of each fragment is known. The patient’s blood is processed for viral or bacterial material yielding DNA fragments of unknown sequences that are then tagged with a fluorescent dye. These blood fragments are introduced to the platform. Because a DNA sequence will link up with its mirror image, some of the tagged DNA fragments will pair, or “hybridize,” with the affixed fragments having the complementary sequence. The unattached fragments are then washed away.
A diode laser or light-emitting diode illuminates the array of DNA sites with light of one color, causing fluorescence at the sites of paired DNA fragments. Below each site is an array of tiny light detectors (photodiodes) with electronic circuitry that detects the fluorescence and sends an electrical signal. The “smart” analog and digital circuitry in the integrated electro-optic chip collects each signal, determines the total sequence of the captured DNA fragments, compares it with the known sequences of various bacteria and viruses, and issues a diagnosis.
Vo-Dinh’s group has demonstrated that the biochip works on synthetic DNA templates from a region of the AIDS virus, the tuberculosis bacillus, and a cancer gene. In 1998, the biochip was licensed to a private company. Because of its speed of diagnosis and because it does not rely on radioactive substances (whose disposal is expensive), the biochip is expected to reduce health care costs.
Flowthrough GenosensorKnowing the structures and functions of genes in the human genome is a holy grail of biology. Researchers in LSD have devised a three-dimensional approach to the problem of sequencing DNA bases. Instead of a checkerboard array of affixed DNA sequences with which tagged mirror-image sequences can pair (as in Vo-Dinh’s biochip), Ken Beattie and Mitch Doktycz have developed a “flowthrough genosensor” made of glass or silicon for DNA sequence analysis. This miniaturized chip is expected to increase the speed, economy, and throughput of genome mapping and sequencing for certain uses.
For example, by comparing particular DNA sequences, the genosensor can help biologists find genetic differences between individuals and determine the genetic basis for diseases. It can identify more quickly and accurately those microbes that “eat” a specific pollutant by detecting the signature gene producing the enzyme that metabolizes the target chemical. It can help determine which genes are active in various body organs and what their functions are.
Mitch Doktycz checks the alignment of a robot’s dispensing probes as he prepares to construct a set of flowthrough genosensors. The dispensing probes transfer DNA sequences from laboratory plates to individual wells of the genosensors. At right, white spot on microscope’s computer screen shows where fluorescent DNA has paired with DNA affixed in one of the 100 wells in the genosensor (half an inch on a side).
“We built a porous glass chip with an array of tiny tubes, or channels, arranged parallel to each other and perpendicular to the array,” Beattie says. “Because these 10-micron channels penetrate the chip, a solution of target DNA will ‘flow through’ rather than ‘over’ the chip and will contact DNA probes attached to the channels’ inner surfaces.
“The porous structure leads to a greatly increased surface area, allowing a higher degree of miniaturization. Also, because the channels are so small, the molecules are brought together into close contact, speeding the hybridization.”
Each of these genosensor “chips” measures less than half an inch on a side and contains about 100,000 channels. The researchers’ recent development of silicon genosensors opens the possibility of using electronic components to manipulate fluids and detect hybridization.
“We have been developing a robotic spotting system to position DNA probes on the flowthrough genosensor,” Doktycz says. “The robot can dispense these probes just a few hundred microns apart to create an array of a few thousand different probes on a dime-size surface.”
The sequence and location of each immobilized DNA strand in the array are recorded on a computer. This information is critical to identifying and sequencing the target DNA.
The researchers also developed a fluorescence imaging system to locate target DNA sequences, which they tag with a fluorescent dye before flowing them through the chip. A solution with a low concentration of the target DNA is injected through a slender tube and into the genosensor, which is mounted onto the fluorescence microscope. The target DNA molecules pair with the immobilized probe molecules whose sequence is complementary. When light of the right wavelength from the microscope is shone on the target DNA fragments, their tags fluoresce, revealing the locations of hybridization events. The image of the bright and dark spots on the genosensor is magnified in the microscope and displayed on a computer monitor. The resulting image is then analyzed to determine sequences contained in the target DNA molecule or quantify the amount of specific target sequences.
The flowthrough genosensor appears to be a significant step in pursuit of biology’s holy grail.
Nose on a Chip
Kim Young examines the “nose on a chip,” which uses cantilevers coated with different chemicals to detect multiple gases.
When he arrives home from vacation, Clarence’s smart alarm system deluges him with messages. Natural gas is leaking from the furnace. Smoke is present in a guest bedroom. Food is spoiling in the refrigerator. In the garage, a paint can lid apparently has come loose, and a container of insecticide is leaking.
Such a warning capability may be possible someday, thanks to ORNL’s invention of a “nose on a chip.” This wireless electronic nose can simultaneously detect and measure a variety of vapors in the air and signal a receiver to sound an alarm or display a message. Already tests of this first battery-operated cantilever array sensor chip set have shown it can simultaneously sense various combinations of hydrogen, nitric oxide, mercury vapor, and alkane thiols in the air. Because the device is inexpensive and can provide instant results, it could soon be incorporated into household gas appliances to warn of hazardous leaks.
The electronic nose consists of an array of tiny fingerlike sensors sculpted along a small silicon chip, plus electronic signal processing and transmission capabilities on other integrated circuits. Hundreds of these springboard-like sensors, called cantilevers, can be carved in a 1-cm2 silicon chip using standard circuit manufacturing techniques. Selectively coating each cantilever sensor in an array with the right chemical can customize the chip to detect thousands of chemicals. This concept is an extension of an invention patented by ORNL’s Thomas Thundat and Eric Wachter.
A properly coated cantilever bends or changes its vibration ever so slightly when it interacts with a target chemical in the environment. For example, because mercury vapor is attracted to gold, a gold-coated cantilever will absorb airborne mercury and bend according to its mass, resulting in an electrical signal indicating the detection of mercury.
“Most existing chemical sensors can detect only one species and require large amounts of space and electrical power,” says Chuck Britton, one of the developers and a member of the I&C Division. “Our innovation is to use miniature arrays of low-power-consumption sensors and electronics on a single chip to detect many different species simultaneously.”
The chip electronics convert the continuous (analog) signals into digital data. An I&C team led by Steve Smith has equipped the wireless electronic nose with a transmitter that first turns the digital data into radiofrequency signals with unique identifier patterns and then sends them to a receiver. The receiver, which includes a computer to analyze the data, could be programmed to sound an alarm if a gas level gets dangerously high.
Copies of the wireless electronic nose can be produced inexpensively, using standard processes for semiconductor manufacturing. “Ultimately,” says Bruce Warmack, leader of the LSD group that developed the microcantilevers as sensors, “these chips could be made so inexpensively they could be thrown away after they are used.”
The project was supported by the internally funded Laboratory Directed Research and Development program. The key to the success of the electronic nose was the collaboration among three ORNL divisions. Expertise in integrated circuit design came from the I&C Division; in cantilever design, from LSD; and in selection and synthesis of coatings, from CASD. The multidisciplinary approach should improve the accuracy and versatility of the nose on a chip, giving new meaning to the phrase “the nose knows.”
Critters on a ChipA future scenario: Numerous silicon chips, each the size of a match tip, are sprinkled over contaminated soil. Bioluminescent bacteria intentionally placed on each chip feast on the pollutants and begin to glow. Their blue-green visible light is absorbed by the silicon, creating electrical charges that are fed into the chip’s circuitry. Signal-processing microelectronics measure the tiny electrical current. From some chips, signals are sent to a pollution engineer’s electronic receiver, which sounds an alarm.
The “critters on a chip,” which were genetically engineered to emit light as they eat and digest certain environmental pollutants, have detected naphthalene and toluene in the soil. The chip electronics reveal the concentration of each pollutant, which is related to the amount of electric current produced.
Mike Simpson shows the bioluminescent bioreporter chip, dubbed the critters on a chip. Genetically engineered bacteria on the chip emit blue-green light when they encounter a pollutant they like to eat. The light is detected by the chip electronics, and the resulting electrical signal reveals the identity and concentration of the pollutant.
The critters-on-a-chip technology developed by Mike Simpson of the I&C Division in 1996 could be used to map soil contamination. Other possible applications are to detect specific chemicals in soil or groundwater, including spilled fuel, toxic metals such as mercury, and explosives that may have leaked from land mines. Oil exploration companies might want to use the technology to detect hydrocarbons that indicate the presence of nearby oil and gas deposits.
ORNL researchers are now working with Perkin-Elmer Corporation, a large environmental instrument company, to commercialize the critters-on-a-chip technology. A growing demand for such wireless chips is anticipated because they can be deployed where other devices can’t—groundwater, industrial process vessels, and the battlefield.
Simpson and his colleagues developed a prototype device by coupling Pseudomonas fluorescens HK44, a novel naphthalene bioreporter microorganism developed by the University of Tennessee Center for Environmental Biotechnology (UT-CEB), to an optical application-specific integrated circuit (OASIC) developed at ORNL. A measured electrical signal was obtained when the OASIC chip was exposed to moth balls, which are made of naphthalene. A second prototype used the toluene-sensitive Pseudomonas Putida TVA8, also developed at UT-CEB.
Simpson is improving the chips to increase their longevity. “The key is to place the bacteria on a transparent silicon nitride film that protects the etched silicon chip from damage in the presence of hazardous chemicals,” Simpson says. “To increase the shelf life of the chip, the bacteria could be freeze dried, and a micromachine on the chip could activate its living sensors when needed by dumping water and nutrients on the dormant bacteria.”
The idea is to avoid having quitters on the chip.
[ Next article | Search | Mail | Contents | Review Home Page | ORNL Home Page ]