|
|
| Bruce Jacobson (left) and Carl Gehrs, manager of ORNL's Center for Biotechnology, examine a newly fabricated temperature control block for a genosensor chip designed by Mitch Doktycz. The genosensor chip can detect specific DNA sequ ences. |
acteria. Our invisible friends, our invisible enemies. Some aid our
digestion, others destroy our poisons. Still other "bugs" make us sick.
Living inside and outside our bodies, natural bacteria are a fact of life. We have learned to live with them and they with us.
Soldiers and sailors in action worry that even more dangerous bacteria are lurking in the
environment. These are "killer" bacteria—the kind used as weapons. Such biological warfare agents are viewed by many to be as threatening to human life as nuclear weapons. In 1993, R. J. Wooley, director of the U.S. Central Intelligence Agency,
proclaimed, "Proliferation (of nuclear, biological, and chemical weapons) poses one of the most complex challenges the intelligence community will face for the remainder of the century." General Colin Powell has stated that biological weapons worry him
more than anything else. He has good reason to worry. A recent U.S. Arms Control and Disarmament Agency report on worldwide arms control compliance states that both Syria and Egypt have offensive biological warfare programs. The key to protecting a military unit or community from dangerous bacteria is to detect them before they reach their intended victims. People can then be warned to leave an area or at least wear protective gear. Bacteria can be detected using "biosensors." A biosensor is a device that detects, records, and transmits information regarding a physiological change or the presence of various chemical or biological materials in the environment. More technically,
a biosensor is a probe that integrates a biological component, such as a whole bacterium or a biological product (e.g., an enzyme or antibody) with an electronic component to yield a measurable signal. Biosensors, which come in a large variety of sizes
and shapes, are used to monitor changes in environmental conditions. They can detect and measure concentrations of specific bacteria or hazardous chemicals; they can measure acidity levels (pH). In short, biosensors can use bacteria and detect them,
too.
Genetically engineered bacteria can also be useful because of their ability to "tattle" on the environment. Such commonly used bacteria have been designed in Oak Ridge and Knoxville to give off a detectable signal, such as light, in the presence of a specific pollutant they like to eat. They may glow in the presence of toluene, a hazardous compound found in gasoline and other petroleum products. They can indicate whether an underground fuel tank is leaking or whether the site of an oil spill has b een cleaned up effectively. These informer bacteria are called bioreporters. In 1990, when a bioreporter of naphthalene was developed and tested at the University of Tennessee at Knoxville (UTK), the ability of these b acteria to glow was demonstrated for President George Bush during his visit to UTK.
Oak Ridge National Laboratory has been developing biosensors and bioreporters for almost a decade. Carl Gehrs, director of ORNL's Center for Biotechnology, says that ORNL's program in biosensors is "a leader among DOE national labs, which is a best-kep t secret." He added that ORNL is proposing to develop biosensors that can detect the presence of biological and chemical warfare agents for military use and for determining the effectiveness of cleaning up waste sites.
It is clear that Oak Ridge can make some impressive, even revolutionary, contributions to military applications of biosensors. One of the rewards of developing biosensors for military use is that the resulting devices may eventually have applications in everyday life. As President Clinton recently told a group of military officials, "What you have done here is what I wish to do nationally: take some of the most talented people in the world who produce some of the most sophisticated military technol ogy and put that to work in the civilian economy."
In the mid-1980s, Tuan Vo-Dinh, Guy Griffin, and others in ORNL's Life Sciences Division were looking for a way to use light to detect cancer-causing agents in groundwater. So they attached to the end of an optical fi ber an antibody that reacts specifically with the carcinogen benzo(a)pyrene (BaP). The anti-BaP antibody on the end of the fiber was immersed in a sample of groundwater. The antibody was allowed to bind the BaP in the groun dwater sample. The antibody-BaP reaction product gives off light if illuminated by light of the right wavelength. So the right light was aimed through the fiber into the groundwater sample. After 5 to 10 minutes, the reaction product fluoresced , and the fluorescence was transmitted back up the fiber and measured. These successful results, reported in 1987 by Vo-Dinh and colleagues, brought the group a 1987 R&D 100 Award from R&D magazine and initiated the group's development of a seri es of fiber-optic-based biosensors.
Antibodies can be produced against bacteria, against complex carbohydrates, and even against smaller organic molecules that may cause cancer. To Vo-Dinh's group, the possibilities for applications of immunosensors seemed almost limitless. Indeed the po ssible uses for biosensors are limited only by our imagination. After all, there are many different ways to combine chemistry, physics, and biology with an electronic detector.
One type of biosensor has only five components: a biological sensing element, a transducer, a signal conditioner, a data processor, and a signal generator. The essential component must produce a signal that is related to the concentration of a specific chemical or biological substance in complex systems. This component takes advantage of the ability of a biomolecule, such as an antibody or enzyme, to specifically recognize the target substance.
Light emissions from microspheres and bacteria are
seen through a fluorescence
microscope. Shown are
red-fluorescing S. aureus bacteria bound to
6.5-µm spheres and one yellow orange-fluorescing
E. coli bound to the larger
10-µm sphere.
In another approach to the use of immunosensors, microspheres of different sizes are labeled with antibodies that bind to different bacteria; thus, microspheres of one size have one particular antibody and microspheres of another s ize have a different antibody. The sizes of the microspheres are identified by their "morphological resonances" (shape-based light emissions when excited by a laser), and the bacteria that become bound are detected by the color of fluorescent dye with which they are stained. In a 1995 paper, Bill Whitten of ORNL's Chemical and Analytical Sciences Division suggests that up to 100 different types of bacteria can be identified simultaneous ly because the stained bacteria all would fluoresce at one wavelength of light and the diameters of the spheres could be illuminated at another wavelength. Although all the spheres fluoresce when excited by one wavelength, the morphological resonances, which look like saw teeth superimposed on a fluorescence emission spectrum, can distinguish among diameters of many different-sized spheres. This approach satisfies one of today's challenges in biotechnology: multiplex biosensors to obtain more infor mation from one sample analysis.
| The goal is to develop an array of chips to collectively monitor bodily functions. |
This "medical telesensor" chip on a fingertip can measure and transmit body temperature.
A chip on your fingertip may someday measure and transmit data on your body temperature. An array of chips attached to your body may provide additional information on blood pressure, oxygen level, and pulse rate. This type of medical telesensor, which is being developed at ORNL for military troops in combat zones, will report measurements of vital functions to remote recorders. The goal is to develop an array of chips to collectively monitor bodily functions. These chips may be attached at various points on a soldier using a nonirritating adhesive like that used in waterproof band-aids. These medical telesensors would send physiological data by wireless transmission to an intelligent monitor on another soldier's helmet. The monitor could alert medics if the data showed that the soldier's condition fit one of five levels of trauma. The monitor also would receive and transmit global satellite positioning data to help medics locate the wounded soldier.
Development of medical telesensors at ORNL is supported by the Defense Sciences Office of the Advanced Research Projects Agency, but the development is expected to have civilian applications. Wireless monitors a ttached to the skin could provide valuable information on the physiological condition of intensive-care patients in hospitals, high-risk outpatients, babies at risk of suffering sudden infant death syndrome, and police and firefighting personnel in ha zardous situations.
In ORNL's Life Sciences Division, Tom Ferrell has shown that a 2 × 2-millimeter (mm) silicon chip attached to the skin can measure body temperature. The chip contains a temperature sensor in an integrated circuit, a lithium thin-film battery that supplies the very low level of power required by the circuit and signal processing and transmission electronics, and an antenna that sends the data by radio signals (radio-frequency transmission) to a monitor when the chip is queried. Ferrell calls thi s biosensor a "medical telesensor ASIC" because it uses an application-specific integrated circuit for telemetry—automatic measurement and transmission of data from remote sources to receivers for recording and analysis.
Ferrell also expects that a chip can be developed to measure blood oxygen level. As the blood oxygen level changes, the color of hemoglobin in blood is altered. Such a chip would have a light source and light detector that could measure changes in the color of hemoglobin transmitted when it is illuminated by light. The results are reported by wireless telemetry.
|
| ORNL has developed a sensitive detector for monitoring changes in the body's concentrations of calcium ions. It may be useful in diagnosing disease or exposure to chemical warfare agents. This biosensor consists of an optical fiber to which is attached a synthesized hybrid molecule. One half of the hybrid molecule binds calcium ions and the other half fluoresces when calcium ions are bound to the molecule. |
Blood pressure and pulse rate may be measured by chips designed to detect pressure changes. Indeed, Jeff Muhs and Steve Allison, both of ORNL's Engineering Technology Division, are working with optical fibers m ade of silicone to take advantage of the unique properties of this substance. Unlike a glass fiber, a silicone fiber is flexible—it can be squeezed or stretched, and the amount of compression or expansion can be measured by changes in light transm ission through the fiber. Thus, silicone fibers embedded in roads can be used to weigh trucks. If a silicone fiber on a chip can sense pressure at various positions in the body, it may be used for monitoring blood pressure, pulse rate, breathing (chest expansion), knee bending during physical rehabilitation, and foot pressure distribution.
Aging, diseases such as diabetes and Alzheimer's, and chemical warfare agents cause changes in metal ion concentrations in the body. If these changes could be detected and measured, the information could provide clues about changes in disease states a nd exposure to toxins. Tuan Vo-Dinh and his coworkers have developed a biosensor using a glass optical fiber and a hybrid molecule he synthesized. One half of the hybrid molecule binds calcium ions and the other half fluoresces when calcium ions are bo und to the molecule. By attaching this molecule to the end of a very small diameter optical fiber, Vo-Dinh measured the concentration of calcium ions in a solution. He plans to make a similar measurement within a single living cell!
Schematic of a microcantilever sensor, which can be adapted to detect physical, chemical, or biological activity.
An interesting alternative to the optical fiber is the microcantilever, which measures the presence of substances by nonoptical methods. It can act as a physical, chemical, or biological sensor by detecting changes in cantilever bending or vibrational frequency. Think of a diving board that wiggles up and down at a regular interval. This wiggling changes when someone steps on the board. Microcantilevers are a million times smaller but molecules adsorbed on a microcantilever cause vibrational freque ncy changes. Viscosity, density, and flow rate can also be measured by detecting the changes in vibrational frequency. Another way of detecting molecular adsorption is by measuring curling of the cantilever due to adsorption stress on just one side of the cantilever. Depending on the nature of chemical bonding of the molecule, the curling can be up or down. For example, if the microcantilever is bimetallic, just like the thermostat at home but a million times smaller, a temperature change as small as a millionth of a degree can be measured. There is much to learn about the basic mechanisms involved.
The microcantilever is ordinarily constructed of a silicon plank 100 micrometers (mm) long, 30 mm wide, and 3 to 4 mm thick (these dimensions are only approximate, and other geometries are sometimes used). When molecules are added to its surface, the extent to which the plank bends can be measured accurately by bouncing a light beam off the surface and measuring the extent to which the light beam is deflected. The vibrational frequency can be induced by piezoelectric transducers and measured with t he same laser beam that measured the deflection because it generates an alternating current in the detector.
Temperature is measured by coating the silicon surface with gold or aluminum, which expands at a different rate than silicon. Because the difference in heat expansion between silicon and gold affects bending of the microcantilever, temperature changes of a millionth of a degree can be measured. Chemical reactions generate heat, so this device can be used as a microcalorimeter to measure the heat of an enzyme-catalyzed reaction or a chemical reaction in a reaction volume of one microliter (1 µl).
Yet another mechanism of response was employed to measure proteins in solution. Antibodies were covalently attached to the silicon surface of a cantilever in such a way that the stresses induced in the antibody when it reacted with its antigen were det
ected. Detection of biological warfare agents or bacteria and viruses in the hospital laboratory should be expedited with this stressed antibody technique. Additional experiments are under way to demonstrate the usefulness of the microcantilever as a b
iosensor. Because of the small size and versatility of the microcantilever, arrays of sensors can be fabricated on a single chip to conceptually mimic the five sensory facilities: sight, hearing, smell, taste, and touch. ORNL researchers Thomas Thun
dat, Bruce Warmack, Eric Wachter, Patrick Oden, and Panos Datskos received a 1996 R&D 100 Award for development of the microcantilever.
Of these biosensors, the most publicized is the optical biopsy sensor developed by Tuan Vo-Dinh in collaboration with medical researchers at Thompson Cancer Survival Center in Knoxville. This sensor can tell whether a tumor in the esophagus is cancerou
s or benign. In the past, determining accurately whether a patient has cancer of the esophagus has required surgical biopsy. However, our laser-based fluorescence method has eliminated the need for biopsy, reducing pain and recovery time for patients.
Detecting Cancer
and Health Abnormalities
Another type of biosensor uses sophisticated technology to detect a specific trait or abnormality in a living organism. ORNL researchers have invented several biosensors of this type.
Tuan Vo-Dinh of ORNL (left) and Bergein Overholt and Masoud Panjehpour, both of Thompson Cancer Survival Center of Knoxville, have developed a new laser technique
for nonsurgically determining whether tumors in the esophagus are cancerous or benign.
| Laser fluorescence diagnosis is accurate in over 98% of the cases. |
Here's how it works. Laser light of the appropriate wavelength is directed to the inner surface of the esophagus by means of a fiber-optic device that is swallowed by the patient. The epithelial cells and tissue inside the esophagus fluoresce when exc ited by the laser light. When the esophagus interior is illuminated with blue light [410 nanometers (nm)], the normal tissue emits light at wavelengths different from those emitted by the cancer cells. Thanks to software developed by Vo-Dinh and colle agues, the spectral properties of the light at wavelengths ranging from 400 to 700 nm can be analyzed at various positions in the esophagus. Emissions from normal cells and cancer cells can be distinguished quite accurately; the difference is expressed as the differential normalized fluorescence index. Tests on more than 200 patients show that, compared with the results of surgical biopsies, laser fluorescence diagnosis is accurate in over 98% of the cases.
Another biosensor from Vo-Dinh's laboratory provides a way to monitor the status of diabetes without using blood samples. In this case, light is used to illuminate the eyeball and stimulate certain substances, including proteins, to emit fluorescent li ght. This fluorescence changes in intensity and wavelength when the distribution and status of proteins in the eye change. This truly noninvasive method depends on a relatively new development for selecting the wavelength of light for illumination. Ins tead of prisms or gratings to refract the light into different wavelengths, a device called the acousto-optic tunable filter (AOTF) is used. One AOTF selects the wavelength of light to shine on the eyeball and another selects the wavelength of fluores cent light emitted from the eyeball. Both AOTF wavelengths are scanned simultaneously using the synchronous luminescence technique developed previously for environmental screening. The AOTFs, which are manipulated with a radio-frequency signal, can scan the entire visible spectrum and portions of the ultraviolet and infrared spectra in milliseconds to select the appropriate wavelengths to use to illuminate the eyeball. They can also select the correct wavelengt h to use to read the fluorescence signal from the patient's eye instantaneously. In this way many spectral scans can be taken and averaged in a computer to obtain the accuracy required to measure the status and changes in the eye proteins of diabetics . Vo-Dinh's group is also studying ways to apply optical techniques to detect skin, cervical, and colon cancers.
Mike Simpson shows the newly developed "critters on a chip" in which bioluminescent bacteria signal the presence of pollutants.
Bob Burlage of the Environmental Science Division has been installing these proteins and working with Vo-Dinh to create an optical biosensor to analyze the emitted light. Also, Michael Simpson of the Instrumentation and Controls Division Division is collaborating with Gary Sayler of UTK to produce a "critters on a chip" technology in which light sensor s pick up and transmit information from chip bacteria that glow in the presence of trace levels of poisons, explosives, or pollutants. These types of biosensors are useful for monitoring efforts to clean up industrial spills because these light-emitt ing bacteria can "report" continuously the progress of biodegradation. Such bioreporters have proven successful using trichloroethylene, toluene, and various petroleum products in laboratory tests. They are now being tested on a much larger scale usi ng a lysimeter.
In the late 1980s, a group at the Oak Ridge Y-12 Plant designed and built a set of 18 lysimeters. Each lysimeter is a cylinder 2.4 meters (8 feet) in diameter and 3 to 3.6 meters (10 to 12 feet) deep that is filled with soil and monitored for the prese nce of a variety of substances. The current experiment in which Sayler is collaborating with Burlage employs several of these cylinders so that they can observe the effectiveness of genetically engineered bacteria in degrading naphthalene.
John Hiller of the Oak Ridge Centers for Manufacturing Technology proposes to adapt a luminescence spectrometer, which was devised to measure uranium concentrations in groundwater, to detect bacteria and other organisms that may contaminate drinking water. Such organisms release small amounts of ATP (adensosine triphosphate, the universal bio-energy chemical), which, in the presence of oxygen, is converted into li ght energy by luciferase, which will be added to the assay system to detect the ATP. The luminescence spectrometer would detect and measure the intensity of light emitted by luciferase to determine the concentration of contaminants in water.
The site for photosynthesis in a green leaf contains a complex set of enzymes and proteins that capture light energy and convert carbon dioxide into compounds that help the plant grow. If a platinum salt in a certain oxidation state is supplied to one of two photosynthetic systems in plant chloroplasts, one photosynthetic reaction system will use light energy to provide electrons that will reduce platinum to the metal form. The metal is deposited on the photosystem complex to form a tiny platinum c enter that can be employed in sophisticated diode-based microelectronics for measurements at extremely high sensitivity, resolution, and speed. Such a biomolecular optoelectronic sensor has been demonstrated by Eli Greenbaum, James Lee, Ida Lee, and St eve Blankinship, all of ORNL's Chemical Technology Division.
The infrared microspectrometer developed at ORNL can be used for blood chemistry analysis, gasoline octane analysis, environmental monitoring, industrial process control,
aircraft corrosion monitoring, and detection of chemical warfare agents. Photo by Tom Cerniglio.
ORNL researchers have made dramatic progress in miniaturizating clinical and chemical laboratories—another aspect of our biosensor work. One useful miniature device is an infrared microspectrometer the size of a sugar cube, constructed by Slo Raj ic and Chuck Egert, both of ORNL's Engineering Technology Division. Carved out of a solid block of plastic, the device measures 1.5 centimeters on a side and has no moving parts. It can be used for blood chemistry analysis, gasoline octane analysis, environmental monitoring, industrial process control, aircraft corrosion monitoring, and detection of chemical warfare agents. The plastic device uses a light source to excite certain types of compounds in gases, liquids, and solids. These excited com pounds give off infrared light of various wavelengths. Because the microspectrometer is precisely manufactured using single-point diamond turning, it can gather up the emitted light and channel it into an optical fiber for analysis. The measured emiss ions wavelengths are fed into a microchip, which identifies and determines the concentrations of chemicals in a sample. The device is inexpensive and performs well in different customized uses.
ORNL's best known miniaturization feat is the "lab on a chip." In 1996 this miniaturized chemical laboratory received a Discover magazine innovation award and an R&D 100 Award from R&D magazine. The chip consists of 50- to 100-mm channel s that are etched into the surface of a microscope slide. When reagents are mixed together and forced to migrate down these channels by differences in electrical potential, reaction rates can be measured and chemicals can be separated. Laser-induced f luorescence monitors the reaction's progress along the channel and measures concentrations of the separated products.
Mike Ramsey, Steve Jacobson, and colleagues in ORNL's Chemical and Analytical Sciences Division have demonstrated that these miniaturized devices are often more accurate than standard laboratory procedures. They have separated mixtures by performing ch romatography in these channels, demonstrated the separation of products of digestion of DNA with restriction enzymes, and demonstrated enzyme-catalyzed reactions. They are adapting such miniaturized devices for many clinical and military applications.
| DNA can be used to identify organisms ranging from humans to bacteria and viruses. |
DNA can be used to identify organisms ranging from humans to bacteria and viruses. The identification consists of reading the sequence of the DNA letters (A, G, C, and T) that compose the alphabet used to describe the bases attached to the deoxyribose phosphate polymer that forms the backbone of the DNA helix. The bases join two strands of this polymer to form a spiral staircase, where the bases and the hydrogen bonds that join them are the steps on the stairs. Just as in the English alphabet, the letters are combined in various groupings and sequences to form words, sentences and paragraphs that organize information in a manner that can be read and utilized. Therefore, much effort has gone into ways to read the sequence of the four letters of t he DNA alphabet, and rather rudimentary methods have been developed that are reliable and useful. Using methods to form short sequences from the DNA of interest, DNA fragments are produced that have one of the letters at their end and that differ acco rding to their sizes. As a result, DNAs containing 400 to 600 letters can be sequenced accurately. However, many hours are required to prepare the fragments and separate them by size (using gel electrophoresis). Using this method, the sequences of mill ions of DNA letters have been determined, enabling the identification of the site of a genetic mutation that causes such diseases as sickle cell anemia, Huntington's disease, fragile X syndrome (a serious type of mental retardation), and several hundr ed other inherited diseases or traits. Furthermore, by combining sequences obtained ~400 at a time, the entire genome of a yeast was determined in 1996, a major accomplishment that led to the identification of an entire new group of genes that had not been recognized before. Many labs that were involved in the sequencing effort have now switched their attention to studying these new genes to learn of their function and significance. The Human Genome Project, which is sponsored by DOE and the Nation al Institutes of Health, plans to go well beyond yeast and determine the sequence of all DNA letters in the human genome. Undoubtedly, this information will lead to the discovery of many new human genes and a more sophisticated understanding of our he alth and disease states. More details may be found in the Primer for Molecular Genetics on the DOE World Wide Web site http://www.ornl.gov/TechResources/Human_Genome/pub licat/primer/intro.html.
Because the methods used to sequence the yeast DNA required dozens of labs in Europe and the United States to work together for several years to determine the total sequence, new and faster methods to read the DNA sequence would clearly be advantage ous. If the methods were faster and less expensive, DNA sequence information would be as common as your blood type and would be much more informative. You could determine whether you are a carrier of a disease gene known to exist in your family, wheth er blood at a crime scene was that of the arrested suspect, or whether a biowarfare bacterium was descending on a battlefield. In some of these uses, the need to know the DNA sequence may be urgent. Because the current methods are not fast enough to sa tisfy such needs, many laboratories, including ORNL, are developing faster methods. In addition to sequence information, methods are also being developed to identify the site where a DNA sequence has been modified.
When a chemical reacts with DNA in the cell nucleus, possibly causing damage, a chemical compound called an adduct forms from the addition of the two species. ORNL researchers led by Bob Hettich and Michelle Buchanan in ORNL's Chemical and Analytical S ciences Division have used mass spectrometry to identify chemicals that form adducts with DNA. They have determined both the chemical identity of the adduct and the specific DNA site at which it occurs.
Laser desorption mass spectrometry has been used at ORNL to detect double-stranded and single-stranded DNA and to determine the sequence of bases in single-stranded DNA.
Simply speaking, DNA is like two strings of beads. One string contains beads of four different colors—red, blue, green and yellow—arranged in a particular order. The order of beads in the other string is governed by this rule: red must be pai red with blue, and yellow must be paired with green. Thus, if the order of the first four beads in the first string is red, green, blue and yellow, the order of the matching four in the second string is blue, yellow, red, and green. The DNA sequence consists of a string of chemical bases, known as A, T, C, and G, and the DNA molecule is folded, looped and coiled in various ways. For double-stranded DNA, the rule is that A on one strand must pair with T on the other, and C must pair with G.
Over the past four years ORNL's mass spectrometry groups have shown that DNA fragments of about 50 to 75 nucleotides (bases) can be sequenced to obtain the order of the A, G, C, and T letters. Because mass spectrometry accomplishes the sequence determ ination in less than a second, this speed supersedes that of gel electrophoresis by orders of magnitude. However, the gel method can do 400 to 600 letters on 50 samples at a time and is not in danger of being replaced yet. Other ways of characterizing DNA consist of determining the sizes of fragments produced by restriction endonucleases, enzymes that cut DNA when they find a specific sequence of letters. A physicist friend of mine calls these enzymes magic scissors because they not only cut the DNA but do so reliably at a sequence site that is specific to each of the restriction enzymes, of which there are over 400. The pattern of sizes of the DNA fragments produced by these magic scissors can be unique to the DNA from a biological species o r even an individual.
Gel electrophoresis is the usual method of classifying the size pattern but mass spectrometry can do that very well, and again in much shorter time. This application of mass spectrometry to DNA characterization will come much sooner than sequencing; i ndeed, one of the ORNL labs held the size record in 1996 by characterizing a 500-base fragment of DNA using matrix-assisted laser desorption ionization mass spectrometry. An ORNL group is also pursuing use of an electrospray method of mass spectrometry that has advantages such as simplicity of sample preparation and analysis. Hettich, Buchanan, Winston Chen, Scott McLuckey, Greg Hurst, and Mitch Doktycz, along with Rick Woychik, have made ma jor contributions to the use of mass spectrometry for DNA analysis.
Besides mass spectrometry, other methods exploit the characteristic reaction between two strands of DNA. Again, recalling that A pairs with T and G with C in opposite strands, a short strand of DNA, called an oligonucleotide, with a sequence of 5'-AGCTTTAACC will bind to 5'-GGTTAAAGCT (read it backwards and match it to the previous letter series) but not to 5'-GGTTAGGACT. This simple direct method of selecting DNA of complementary sequences has been used by Bob Foote and Mitch Doktycz, both of the Life Sciences Division.
Immobilized oligonucleotide
arrays have been developed
at ORNL to analyze DNA sequences.
Above, the DNA sequence is amplified, labeled, and allowed to hybridize to the array of immobilized oligonucleotides. At
right, the DNA sequence is deciphered by overlapping
the hybridizing probes of similar sequence. The
technique has applications in medicine, forensics, agriculture, and environmental bioremediation.
They synthesized a number of different DNA sequences directly on a glass surface and showed that the complementary sequence of the free DNA could be obtained by determining to which of the sequences it would bind. The free DNA was labeled with phospho rus-32 so that the pattern could be observed by autoradiography; in other words the glass was placed against a photographic film for a time, after which it was developed to determine the locations of the radioactive spots. In this case the immobile DNAs contained only eight bases and thus could determine the sequence of only eight bases in the free DNA. Longer immobile DNAs would give longer reads in the free DNA. This technique is often called sequencing by hybridization (SBH).
Interestingly, Doktycz and his colleagues have shown that the rules "A binds to T" and "G binds to C" are not completely reliable; mismatches can and do occur in hybridization experiments. If the mismatch is at or near the end of the DNA strand, the r ules are more likely to be broken than if the mismatch is at an interior position. The ORNL researchers developed a number of rules that predict when a mismatch is likely to occur. If these rules are ignored in an interpretation of binding patterns, t he reader could be misled as to the true sequence of the DNAs of interest. ORNL is collaborating with three other organizations that are developing this method of sequencing by hybridization. Because hybridization methods are so much faster than gel el ectrophoresis for sequence determination, it is likely that hybridization may become the method of choice for many applications for DNA fingerprinting. Ken Beattie, who recently joined the Life Sciences Division, brings his international experience wi th SBH to ORNL. He has proposed use of a flow-through SBH chip for multiple applications—genome sequencing, assessing the effectiveness of bioremediation, and evaluating the quality of agricultural microorganisms, plants, and animals, for examples .
The work of Foote and Doktycz used phosphorus-32 to label the DNA fragments. Other ways of labeling DNA are also employed. Currently, the most used are fluorescent labels, but Bruce Jacobson and Heinrich Arlinghaus of Atom Sciences, Inc. have been expl oring the use of enriched stable isotopes of tin and rare earth elements to label DNA fragments for sequencing by hybridization.
When a tin-labeled DNA segment hybridizes to its complementary DNA sequence in an array of short DNAs on a glass or nylon surface, the segment's location is determined by a laser-induced resonance ionization microprobe, which uses a laser beam to dissociate all atoms within several hundred monolayers of the sample surface. Those at oms that appear as ions are discarded; however, the neutral atoms of a selected element (in this case, tin) are converted to ions by illuminating them with a resonance laser that is tuned to the quantum energy of that element's electrons, causing only that element's atoms to be ionized. The ions produced in this way are introduced into a mass spectrometer, which separates them into isotopes according to differences in mass and measures the concentration of each tin isotope.
By working in a clean room, researchers using this highly selective and extremely sensitive technique have located DNAs that were hybridized to the complementary sequence with little background from either environmental tin or tin from noncomplementary tin-labeled DNAs, detectable on noncomplementary sequences. Because 10 stable isotopes of tin exist, it is possible to use all 10 labels at one time to "multiplex" the process. Ways of labeling specific DNA strands or fragments 8 to 20 bases long wit h tin enriched in a specific isotope have been devised by Rick Sachleben, Gil Brown, and Fred Sloop, all of ORNL's Chemical and Analytical Sciences Division. They have also developed a way to label specific DNAs with individual enriched stable isotope s of several rare earths, enabling the use of 20 or more labels at one time. Use of mass spectrometric detection of the stable isotopes increases the multiplexing possibilities greatly, allowing much more information to be obtained than is possible us ing only four fluorescent labels at a time. It is also an improvement because it overcomes the problem of the general fluorescent background when using nylon. Because phosphorus-32 decays with a half-life of about 14 days, the tin labeling approach al so looks attractive. Stable isotope labels may someday play a significant role in future DNA analyses. Imagine the SBH chip being interrogated by 10 or 20 stable isotope-labeled DNAs at one time. The laser atomization and mass spectrometry would identi fy each label in one pass, making the process much more efficient in time and materials. In addition, the laser atomization method would be accomplished in a few minutes, again much faster than present gel electrophoresis.
Another hybridization method being developed at Oak Ridge uses a process called surface-enhanced Raman spectroscopy (SERS). Hybridized DNA is transferred from the nylon membrane to a glass strip coated with tiny silver spheres. The dye labels attached to the DNA bases have unique Raman infrared spectra, but the normally weak Raman lines are greatly enhanced by the presence of the silver spheres. This enhancement allows DNA bases to be detected at sufficient sensitivity to be useful for DNA sequenci ng studies. Earlier, Vo-Dinh, Mayo Uziel, and Alan Morrison, using SERS, detected a fluorescent carcinogen on DNA. Then, Dan Jacobson and Tom Ferrell demonstrated the power of SERS for DNA sequence analysis. Recently, Vo-Dinh, Kelly Houck, and David S tokes have developed it into a highly sensitive method that received an R&D 100 Award in 1996.
Yet another approach to DNA analysis taken by Vo-Dinh, in collaboration with members of the I&C Division, is to design an array of charge-coupled device (CCD) detectors and attach to their surface specific sequences of DNA. Free DNA strands are labeled with fluorescent molecules and hybridized to the bound DNA. When illuminated with laser light of the correct wavelength, the fluorescent tags of the hybridized DNA emit light signals that are detected by the individual CCD detectors behind each DNA site.
Because each of the CCD pixels has a different but known short DNA sequence bound to it, the sequence of the piece of the longer DNA strand that hybridizes with the short strand can be identified. By assembling all the information from the pixels, a l arger portion of the DNA sequence is obtained. Because bacteria have unique DNA sequences, these hybridization methods hold great promise as the basis for new techniques to rapidly analyze DNA, characterize its source, and identify bacteria.
DNA analysis of a different type depends on magic scissor enzymes, as described previously. For example, the enzyme EcoRI cuts DNA when 5'-GAATTC is on one strand of the DNA and its complement 5'-CTTAAG is on the other; the cut separates the G from th e A in both strands. This sequence occurs rather frequently in DNA, so the enzyme produces a large number of fragments. The enzyme Not I requires eight nucleotides in the sequence 5'-GCGGCCGC and its complement 5'GCGGCCGC; such a sequence occurs infreq uently, so fewer DNA fragments are produced, and they are generally much longer than those fragments cut by EcoRI. Because of the availability of more than 400 restriction enzymes that have unique and well-known sequence requirements, the DNA fragment patterns produced by the various enzymes can be used to characterize an individual's DNA. This is the basis for DNA fingerprints obtained through separation by size of DNA fragments by gel electrophoresis. Such fingerprints are used widely for eviden ce in research and in judicial courts.
To determine which DNA fragments make up a fingerprint, they must be separated and identified. The "lab on a chip" developed by Mike Ramsey and colleagues has been adapted to perform such separations within a few minutes—much faster than standard gel procedures. Like microcircuits and computers operating in parallel, these chemical separations chips can be used in parallel for DNA analysis. In one application, liquids containing DNA and a restriction enzyme are injected into different chambers etched into the chip. Electric fields pump the liquids through a microscopic channel into a reaction chamber, where the enzyme cuts the DNA into pieces of different lengths. The DNA snippets are then electrically pumped to the separation channel, wher e they are tagged with fluorescent dyes for detection.
The DNA fragments of various sizes are sorted in a liquid containing fibrous strands of a polymer. The DNA fragments get tangled with the polymer strands, which slow them down as they pass through. Small chunks of DNA find their way through the tangled web faster than the larger ones, so separation results. As the fragments ar e separated, they are illuminated with a laser light, causing them to fluoresce. The detected light intensities are fed to a computer, which sorts through signals from separated fragments to provide a sample analysis.
|
| Conceptional design of DNA processing and analysis chip that Mike Ramsey and Bob Foote are developing. |
Automation and miniaturization are the two by-words in Ramsey's lab. To eliminate manual manipulation of DNA, Ramsey and his colleagues propose to build a more comprehensive chip that can prepare the DNA from a biological sample by lysing the cells, ex tracting the DNA, digesting it with restriction enzymes, amplifying it with the polymerase chain reaction, attaching fluorescent labels, and then obtaining the electrophoretic fingerprint. All steps would occur in a series of tiny chambers on the chip , and the results would be obtained in minutes rather than hours or days, as is the case now.
ORNL has developed a DNA mapping and sequencing technique using atomic force microscopy (AFM) to measure the length and width of double-stranded DNA molecules. The AFM is a sensitive device that traces the surface of a material, producing a three-dimen sional map. Certain viruses have circular DNA of known sizes, as characterized by various methods. Dave Allison, Doktycz, and Warmack, all in the Life Sciences Division, showed that the AFM can measure sizes of viral DNA circles.
 | This atomic force microscopy image shows two knots, or "protein bumps," formed on the DNA thread. Each knot results when a mutant enzyme binds to, rather than cleaves, the DNA strand. |
The magic scissors enzyme, or restriction endonuclease, cuts a DNA sequence each time it occurs in the strand. But the mutant form of this enzyme simply binds to, rather than cleaves, the DNA strand. The AFM can image the resulting knot, or protein bum p, formed on the DNA thread. ORNL researchers can identify where the enzyme binds to the DNA within 100 base pairs, which is a very high resolution. They've shown that the distance from one bump to the next, as imaged by the AFM, was exactly as predicted from enzyme cutting experiments analyzed by gel electropho resis.
Through genetic engineering, other enzymes and proteins could be developed that bind to, rather than cut, particular DNA sequences. By using the AFM to view a DNA strand treated with different engineered enzymes, it may be possible to determine a certa in sequence of its chemical bases, a certain map position, or a spacing of genes and other biochemical landmarks on chromosomes. Using this biosensor, ORNL researchers have shown that the AFM can characterize a DNA strand according to distances betwee n particular restriction binding sites, eliminating the need to enzymatically cut the DNA into fragments for separation by gel electrophoresis. Because the AFM method is faster than gel electrophoresis, the genome sequencing community is already start ing to employ this technique for DNA characterization and mapping
Another bacteria detection method being developed by Bill Whitten does not require prior extraction. He has shown that, by illuminating airborne bacteria with laser light to obtain a mass spectrum, different bacteria can be distinguished by their indiv idual spectra. U.S. military organizations are seeking methods for identifying bacteria within five minutes to provide soldiers with enough warning about the hazard to employ effective countermeasures.
Our experience in analyzing lipids with mass spectrometry may have forensic applications. Lipids are the main components of human fingerprints. Recently, the Knoxville Police Department sought ORNL's assistance in determining why, under the same condit ions, fingerprints of young children disappear readily but adult fingerprints persist for weeks. Mass spectrometry of gas chromatographic analysis of lipids collected from fingers of children under the age of 10 shows that the spectra of children's li pids differ significantly from those of adult lipids. Lipids from children contain unesterified fatty acids, which disappear into the air from fingerprints exposed to the hot Tennessee sun. However, fatty acids in adult lipids are esterified with long- chain alcohols, making them much less volatile. Michelle Buchanan worked with Art Bohanan, a crime specialist in the Knoxville Police Department, to bring this finding at ORNL to the attention of the forensic community and to help stimulate support fo r forensics research at ORNL (see Pete Xiques' "ORNL's War on Crime, Technically Speaking"). Tuan Vo-Dinh and colleagues have developed computer software that can analyze the spectra of standard fingerprints very rapidly and enha nce spectra of fingerprint images too weak to be read visually, making possible identification of children's fingerprints
 |  |
| An ORNL technique using laser beams and mirrors can determine the shapes of human body parts. Its accuracy could facilitate the creation of clothes that fit. | |
Perhaps the most unusual of ORNL's biosensors is a new technique to measure human body surfaces. Such measurements, called anthropometry, are used by tailors, artists, and scientists. One of the finest minds in science to take a strong interest in anth ropometry was Leonardo da Vinci, who drew the famous Proportions of the Human Figure some 500 years ago. An ORNL scientist who has moved the field forward is Judson Jones of the Computer Science and Mathematic s Division. He has developed a technique using laser beams and mirrors to determine the shape of human body parts. He measures the topology of a solid surface, using amplitude-modulated laser radar, which measures arc lengths along complex and oft en inaccessible body contours. Because laser radar measures the phase and amplitude of a reflected, modulated laser beam, only one optical path is required between the sensor and the subject. Multiple images are combined by integrating information from different virtual viewpoints, any number of which can be created with strategically positioned mirrors. Arc lengths along arbitrary contours, surface areas, and volume estimates all become possible. The accuracy of the measurements is within 1 mm. Be cause creating "clothes that fit" may be accomplished using a single camera with several mirrors, a blue jeans manufacturer has shown interest in this methodology.
Oak Ridge scientists have been developing biosensors and bioreporters steadily for almost ten years. Now that the medical, military, and industrial communities are expressing more interest and providing more support, we can expect an accelerated pace i n new developments to use or evaluate living organisms, including bacteria, to gain valuable information about our bodies and the environment.
|
B I O G R A P H I C A L Sketch |
| K. Bruce Jacobson is deputy director of ORNL's Center for Biotechnology. He received his B.S. degree in chemistry from St. Bonaventure College. After serving two years in the U.S. Army, he earned his Ph.D. degree from Johns Hopkins University. Following a postdoctoral appointment at the California Institute of Technology, he became a staff member of the Biology Division at ORNL in 1958. He holds an additional appointment as adjunct professor in the University of Tennessee–Oak Ridge Gra duate School of Biomedical Sciences. His research interests include the development of new technologies for DNA sequencing, the mechanism of enzyme action, the structure-function relationships for transfer RNA, pteridine metabolism, and biochemical gen etics. Currently, he is a staff researcher in ORNL's Life Sciences Division. |
| Biosensor/Probe | Principal Investigator(s) |
|---|---|
| BaP detector using antibodies | Tuan Vo-Dinh |
| Biological threat detector using optical spectra | Bill Whitten |
| Medical telesensor ASIC | Tom Ferrell |
| Pressure sensor using silicone fiber | Jeff Muhs, Steve Allison |
| Calcium ion detector | Tuan Vo-Dinh |
| Microcantilevers using mass and vibrational frequency | Thomas Thundat, Bruce Warmack, Eric Wachter |
| Optical biopsy sensor for cancer detection | Tuan Vo-Dinh |
| Diabetes monitor using protein fluorescence | Tuan Vo-Dinh |
| Bacteria with luciferase that eat toluene | Robert Burlage, Larry Simpson, Tuan Vo-Dinh |
| Bacteria with green fluorescent protein that eat toluene | Robert Burlage, Larry Simpson, Tuan Vo-Dinh |
| Platinized chloroplasts | Eli Greenbaum |
| Microspectrometer | Slo Rajic, Chuck Egert |
| Lab on a chip | Mike Ramsey, Steve Jacobson |
| DNA analysis by mass spectrometry | Michelle Buchanan, Winston Chen, Mitch Doktycz, Greg Hurst, Scott McLuckey |
| Protein analysis by mass spectrometry | Michelle Buchanan, Greg Hurst, Scott McLuckey |
| Lipid signatures of bacteria | David White |
| Mass spectral signatures of bacteria | Bill Whitten |
| Lipid fingerprints of children | Michelle Buchanan |
| DNA analysis: in situ synthesis | Bob Foote, Mitch Doktycz Ken Beattie |
| DNA analysis: labeling with tin isotopes | Bruce Jacobson |
| DNA analysis: surface-enhanced Raman labels | Tuan Vo-Dinh, Kelly Houck, David Stokes |
| DNA analysis: CCD pixels and fluorescence | Tuan Vo-Dinh |
| Genome mapping with atomic force microscopy | Dave Allison, Mitch Doktycz, Bruce Warmack |
| Anthropometry | Judson Jones |
[ Next article | Search | Mail | Contents | Review Home Page | ORNL Home Page ]
ORNL's surface-enhanced Raman gene (SERG) probes can locate free DNA molecules that have hybridized to other DNAs fixed on a surface. The technique has use in medi
cine, forensics, agriculture, and environmental bioremediation.