Sensing the Battlefield Environment
To accomplish their mission, soldiers must have safe air, food, and water. Unsafe air, water, and food can be avoided and remedied only if the threat is known. Therefore, sensing and detecting a threat must precede intervention to counter the threat. This chapter discusses biological sensors and their role on future Army battlefields.
Since the introduction of chemical agents in World War I, the Army has had to make special preparations to defend against chemical and biological warfare. Future soldiers must also be prepared for this possibility. Just as the Army now depends on electronic and thermal sensors to detect the presence of enemy combat systems, the future Army may come to depend on biological sensors, not just to sense chemical and biological warfare agents, but also to add to its intelligence collection capabilities and knowledge of the battlefield.
Sensing capabilities now packaged in small, lightweight forms that are easy to use and specific for some forms of biological and chemical warfare threats already exist. But these sensors are incapable of responding to all threats, or even to multiple threats; they are designed to perform isolated experiments to detect or verify the presence of known threat agents. Current detection systems and technologies, which may also be suitable for civilian defenses against biological weapons of mass destruction, will have to be adapted to the unique needs of the future soldier environment.
Less sophisticated, less expensive environmental sensors are now used routinely to monitor the environment on a continuing basis for the presence of known hazards. Health monitors, which sense changes in an individual’s blood sugar or blood pressure, are also biological sensors. With new developments in biology, and in the applications of biology through biotechnology, other types of sensors and sensing methods will certainly be developed. A challenge for the Army will be to determine the directions of biological research and identify avenues that would lead to sensing capabilities that would increase combat effectiveness.
Biological sensors, or biosensors, may be defined as devices that probe the environment for specific molecules or entities through chemical, biochemical, or biological assays. The targets can be airborne, in liquids, or in solid materials. Biosensors may involve any or all of the following functions: detection, capture, concentration, derivitization, and analysis of samples. Biosensor components may have microscale features but may not necessarily be small scale. Ranging from several square centimeters to the size of a computer chip, a small sensor that can perform all of the functions normally carried out at the laboratory bench is sometimes referred to as a “laboratory on a chip.”
In general, the response of a biosensor is based on an assay, an experiment or test to detect a target molecule. The test is designed based on the known interaction between the target, also known as the analyte, and a reagent or organism known to react in the presence of the analyte. Sensing systems generally consist of a recognition element (i.e., a detection element) and a transduction method to translate the recognition into an observable, quantifiable electronic or optical signal. One type of assay, known as an immunoassay, is commonly used to detect and identify biological agents, including bacteria, viruses, and proteins.
Different transduction methods based on systems such as mass spectrometry, microcantilevers, miniaturized flow cytometry, and electronic signatures have been demonstrated. Recently, biochemistries have been developed for chemical sensing. Optical methods based on changes in the refractive index have also opened the way for new detection capabilities. Any of these methods, however, will require
that the Army have an extensive database of potential threat agents and other “target threat molecules” and that these methods be validated for accuracy and robustness under battlefield conditions.
Small-scale biosensing devices that incorporate biologically derived molecules that selectively capture specific target molecules are called biochips. Biochips use a biorecognition element, such as an antibody (protein) or oligonucleotide, as the reagent to selectively capture and thereby identify target molecules. This process is further described in the section on assay formats later in this chapter.
Biochips may be used in external applications (e.g., to analyze a sample of biological fluids) or in internal applications (e.g., invasive assays, in which devices are temporarily implanted or placed inside the human body). A good example of a biochip is a sensor placed under the skin for detecting blood glucose (see Chapter 7 for a discussion of internal sensing applications and limitations).
Small biosensing devices could change the way soldiers “see” the battlefield. Miniaturized, postage-stamp-sized biosensors containing biochips for monitoring the battlefield environment might be worn like wristwatches. In sufficient quantities, these inexpensive, miniature biosensors could provide hundreds of monitoring points for sensing target molecules. Internal biosensors used to monitor physical reactions and other physical parameters could also provide monitoring points.
Most target threat molecules (e.g., chemical or biological warfare agents in liquid or aerosol form) are extremely difficult to detect. The concentrations of samples are very low, and the samples are likely to be “cluttered” with pollen, dust, and other natural biological constituents. The sensitivity of a biosensor also depends on the specific target molecule, some of which are more difficult to detect and assess than others.
The detection of chemical warfare agents is much less difficult than the detection of biological agents, and detectors for chemical agents most likely to be used in combat are under development by DOD. At the same time, for the foreseeable future, effective detection of biological pathogens will depend mostly on the timely reporting and correlation of evidence from the primary health care system, in other words, symptoms reported by affected individuals, rather than on small sensing devices.
For the future, the Army should think in terms of multiple biosensors acting together, rather than single biosensors. Consider that a single miniature sensor system will only be capable of digesting a small amount of air, which may or may not contain enough chemical or biological agent to be detectable. Only a network of biosensors acting in concert and over a period of time would have any chance of detecting a chemical or biological threat at extremely low concentrations.
A network of miniature biosensors carried by soldiers and vehicles deliberately placed throughout a likely battlefield could also have other Army applications. The presence of target threat molecules could provide advance knowledge of enemy presence, activities, or intentions. This biosensory intelligence could be combined with other sources of intelligence, thus providing commanders with new ways of seeing the battlefield and influencing the course of a battle. However, this desirable network of tiny biosensors cannot be implemented until critical development barriers are overcome related to the collection and handling of samples.
There are numerous ways of detecting and measuring bioanalytes and using biomolecules to detect other physical-chemical moieties. Typically, measurements of bioanalytes are measured by the identification of a specific aspect of the molecule, such as a marker on the surface of a cell, protein, or nucleic-acid sequence. Each recognition event or method is slightly different in terms of stability, sensitivity, and specificity. In general, cellular assays are fragile and unstable because the cell must be kept intact, if not alive.
Biologically based technologies for detecting pathogens include nucleic-acid testing and immunoassays. An immunoassay is a laboratory or clinical technique that uses the specific binding between an antigen and its homologous antibody to identify and quantify a substance in a sample. Faster, simpler pathogen detection would increase the usefulness of immunoassays.
The Army and DOD have significant programs in place for the development and testing of chemical and biological warfare (CBW) agent detection systems based on immunoassays and, more recently, on toxicogenomics. Toxicogenomics refers to methods that measure the expression of DNA in living organisms or cells that have been challenged with a toxin or pathogen (see Chapter 7 for a discussion of genomics applications). In toxicogenomic immunoassays, a predetermined antibody selectively binds a specific antigen in a system set up to indicate the binding event (e.g., by a change in color).
The primary markets for immunoassays are research laboratories and developers of in vitro diagnostics. Researchers use a wide range of immunoassay formats, ranging from ELISA (enzyme-linked immunosorbent assay) to western blots. Diagnostics, the largest commercial market for immunoassays, can be divided into two segments: the clinicallaboratory market and the consumer market.
The clinical-laboratory market uses instrumentation designed for high throughput (the ability to handle multiple analyses quickly) at minimal cost per test in a laboratory environment. The major suppliers to this market are Abbott Laboratories, Becton Dickinson, Roche, and Johnson & Johnson/Ortho. The consumer market is focused on single-use, “point-of-care,” disposable products (e.g., pregnancy test kits). In addition to the well-known major players listed above, Unilever, Biosite, and Carter-Wallace are also in this market.
In both markets, the majority of assays are single-analyte assays that rely on polyclonal or monoclonal antibodies derived from the immunization of animals or from cell cultures. In the future, the term immunoassay will probably be a misnomer because detection reagents are likely to be molecular-recognition reagents rather than full-length antibodies. Technologies (e.g., phage display) are currently being used to develop large libraries of antibody fragments that can be rapidly and reliably screened against a variety of target antigens (de Haard et al., 1999). Libraries of smaller protein domains and peptides have also been successfully screened against target molecules to identify recognition reagents with suitable affinities and specificities for use in immunoassay formats (Cannon et al., 1996). Ribosomal or messenger ribonucleic acid (mRNA) fusion-display techniques could construct even larger libraries 1 and provide more rapid screening (Roberts and Szostak, 1997).
In the next five years it will certainly be faster and cheaper to discover “capture-and-detection” reagents via these types of techniques than via conventional immunization techniques. The new reagent molecules will be less expensive to manufacture and may be more stable than antibodies. To allow for lengthy storage periods, the Army should ensure that preservation materials are also developed to extend the shelf life of the reagents.
Methods of discovering binding reagents more cheaply will become much more important as the number of interesting targets increases. Some companies, such as Millennium Predictive Medicine and Diadexus, are focusing on developing capabilities that can be used to define multiple targets of interest for many pathologies and physiologically relevant pathways (see discussion in Chapter 7).
Concurrently, format immunoassays will be moving toward decreasing the size of samples and increasing assay range and accuracy. To achieve this, many firms and academic laboratories are developing promising DNA-array and protein-array technologies. DNA-array technologies have already gained commercial acceptance in the transcription profiling or gene-expression monitoring fields (e.g., Affymetrix, Sequenom, Genomic Solutions); protein-array technologies that would apply to immunoassays are in earlier stages of development.
As a result of these developments, tools and capabilities for immunoassay-type products over a wide range of applications, including those that the Army may use, are likely to become far more capable than they are today. Although nucleic-acid testing techniques will likely remain superior in sensitivity for pathogen detection, immunoassays will be simpler and faster and will respond to a much wider range of challenges.
The issues of size, portability, and robustness for Army applications and field conditions will still have to be addressed. Although only minute (nanogram to microgram) quantities of key reagents, dissolved in buffer, will be necessary, the assays themselves will still require a laboratory bench. Systematic engineering will be necessary to integrate and package assays in easy-to-use kits to address Army needs.
Nucleic-acid assays use the specific sequences of DNA or RNA of an organism as an identification scheme. This requires isolating the nucleic acid from the background of other molecules and typically performing a DNA amplification step (RNA is usually transcribed back to DNA through a process referred to as reverse transcription) (see Box 3-1). The amplified sequence of the target organism is then detected. In most applications, the amplification step requires enzymatic synthesis of specific sequences of the targeted organism’s DNA. PCR (polymerase chain reaction), the most commonly used method of amplification, requires heat activation.
Detection can range from fluorescent detection of the amplified product via electrophoresis, via amplification of an additional nucleic-acid probe, or via the subsequent hybridization of the product to a known matching sequence on an immobilized probe. All of the variations of these methods use a hybridization probe or direct detection of the amplified DNA. Because these assays detect the basic building blocks of biomolecules (i.e., nucleic acids), which produce all of the materials in the biological entity, they are inherently very specific.
Currently, nucleic-acid assays and assay systems (instrumentation and devices) are being developed for pathogen detection, genetic screening, cancer diagnostics, food testing, and other applications. The degree of complexity of the assay is a direct function of the complexity of the sample type. For example, air or water samples may require less processing than tissue or soil samples from which the
1 In addition to these biologically based approaches, many molecular-diversity approaches use synthetic methodologies. Libraries constructed using synthetic methods include combinatorial chemistry libraries (Terret et al., 1995) and peptide and oligonucleotide aptamer libraries (Buettner et al., 1996; Ellington and Szostak, 1990). These libraries are generally less diverse than their biologic counterparts but have the advantage that selected compounds can, by definition, be synthesized by available chemistries. If synthetic compounds with desired selectivities and affinities are discovered, they could be very stable and readily producible for incorporation into detection systems.
Roles of DNA, mRNA, and Proteins
DNA is the informational basis from which living cells derive instructions for synthesizing proteins. Many of the resulting proteins are enzymes that catalyze biochemical reactions from which the cell derives energy or generates other molecules essential to its health and safety. The process normally occurs when the sequence of nucleotides in DNA is transcribed into a complementary, single strand of nucleotides known as messenger RNA, or mRNA. The mRNA provides the instructions by which other components in the cell synthesize proteins. Because not all genes are transcribed (or expressed) but all genes that are transcribed do so through mRNA, the presence of mRNA is an indicator that a gene from the cell’s DNA has been expressed.
The DNA from which the mRNA is obtained is sometimes interspersed with oligonucleotide spacers that do not appear in the final mRNA. Because mRNA is used for the cell’s molecular machinery to generate the protein, the sequence of DNA (or gene) that corresponds to a protein can be obtained by reverse transcription of the mRNA. The DNA is generated from the mRNA using a test tube (in vitro) procedure in which enzymes and several reagents are added to a purified form of the mRNA, which is then used to form the oligonucleotide sequence, or DNA, from which the mRNA was originally generated. The opposite of transcription, this process is referred to as reverse transcription. The DNA obtained in this manner, called complementary DNA, or cDNA, provides a nucleotide that can be further amplified and used to carry out an analytical procedure to obtain the sequence of the cDNA.Source: Courtesy of Professor Michael Ladisch, Purdue University.
molecule of interest must be extracted. Sample preparation and purification are also critical factors. If the sample-processing challenge can be met, however, the assay will be extremely sensitive and specific.
The development of simple, automated, integrated sample-processing systems coupled with nucleic-acid assays is a critical barrier to their widespread adoption and implementation. Sample preparation is also a critical barrier to the sensing applications important to the Army because the materials to be interrogated by sensing devices are likely to be chemically complex and biologically dirty (see Barriers to the Development of Portable Sensors, below).
A variety of approaches to sample preparation apply. Whether applied on the nanoliter, microliter, or milliliter scales, the principles upon which initial fractionation of the sample is based are similar. The principles of separation that apply to biological molecules are referred to as bioseparations; the design of separation protocols is referred to as bioseparations engineering (Ladisch, 2001).
The selection of a particular detection method will affect the overall speed, efficiency, and accuracy of the sensing system. Miniaturization requires detection strategies compatible with smaller sample sizes to increase sensitivities and minimize background effects. The three primary methods of detection that have been used for screening programs are fluorescence, chemiluminescence, and mass spectrometry. Because neither fluorescent nor chemiluminescent methods require fluidic manipulations following an assay, the scale and format of their means of implementation can vary (e.g.,
plates or chips). Mass spectrometry, however, requires that a sample solution be transported from one location to another. On a microscale, this would require picoliter or nanoliter fluid management via capillaries or microfluidic chips. Therefore, the commercial development of chip-based mass spectrometry has been a high priority.
Two designs of microdevices for microanalysis by mass spectrometry have been developed. In both, electrospray is used for sample ionization and transfer of the analytes from the microdevice to the mass spectrometer. The first design incorporates a capillary-electrophoresis separation channel and a micromachined pneumatic nebulizer to generate a stable sample flow and electrospray. The second is designed for high-throughput infusion analysis in a format compatible with a standard microliter well plate. Samples are deposited into the wells and then analyzed in rapid sequence by consecutive application of the electrospray voltage and pressure to each well. During operation, the microdevice is positioned on a motorized translation stage to ensure that consecutive samples can be analyzed immediately after sufficient data have been collected from the previous sample.
A number of microenabled devices and techniques are emerging in the analytical chemistry industry to address requirements for increased sample throughput and decreased sample volumes. These components, methods, and materials are referred to in the commercial sensor industry as micro total analysis systems (MicroTAS) (see Box 3-2).
An optical sensor system also consists of a recognition (detection) element and a transduction method of translating the recognition event into an observable, quantifiable optical signal and, ultimately, an electronic signal. In optical sensor systems, the recognition step is generally based on (1) spectral interactions with the species to be detected (i.e., absorption, emission, and scattering); (2) interaction between the species to be measured (the analyte) and a reagent; and (3) for biological agents, interaction between the species of interest and a receptor.
In some systems, the recognition process involves a combination of steps, such as tags or labels to alter the fluorescent properties of the analyte. Optical transduction schemes typically rely on measurements of spectral intensity (e.g, direct-absorption measurements, fluorescence, FTIR [Fourier transform infrared], and scattering methods, including Raman and surface-enhanced Raman), interferometry (generally based on refractive-index measurements), mass-loading measurements, and stress/strain-induced deformation measurements.
With the emergence of multichannel sensing configurations, signal processing is becoming increasingly important. Fourier transform spectroscopy is an example of the power of signal processing. To meet the demand for field-portable sensor systems, attempts are being made to combine the recognition steps with optical transduction
Micro Total Analysis Systems (MicroTAS)
In the late 1980s and early 1990s micromachined fluid components (typically glass or silicon) seemed to be promising for chemical separation devices, reaction chambers, fluid-handling devices, sensors, and detectors. This field of research encompasses relevant advances in microfluidics and has become an important basis for microelectromechanical systems (MEMS) devices and technologies developed for chemical analysis systems. In recent years, the system-level technologies in chemical analysis systems have been increasingly referred to as micro total analysis systems (MicroTAS).
As a result of the developments in MEMS and MicroTAS, various microcomponents for biosensor applications have been demonstrated. However, very few of these components have left the academic setting, been transferred to the manufacturing/commercial environment, been successfully integrated with other key components, are actually for sale, or have been tested with real samples under field conditions.
Although system-level approaches to develop complete instruments are being attempted, the emphasis is still on component-level developments. For example, several groups have begun to explore using the centrifugal forces present in current compact disc players to actuate a pumping/valving mechanism in a molecular biological analysis system. This would be a clever way to take advantage of an inexpensive consumer product for fluid handling. Even if some simple fluid handling developments were achieved, other critical components of a complete system are still in their infancy. For meaningful analyses on real samples, the issues of repeatability, surface characteristics of the device (e.g., hydrophobicity and biocompatibility), reaction chambers, reagent storage, and detection must still be addressed.
Novel materials and methods for fabricating the key components of microsystems are being developed. An example is a hydrogel-based valve that swells/shrinks in the presence of certain salt/pH solutions. This valve has been shown to be very simple and biocompatible. Basic research on the development of microcomponents or nanocomponents have led to the discovery of fundamental principles, such as the very powerful miniature electrokinetic pumping mechanism, a phenomenon that works only in the microscale.
techniques. Improved optical manufacturing technologies have stimulated the development of miniaturized bulk-measurement systems. The development of fiber-optic and planar-optical wave-guide technologies and the emergence of silicon micromachining technologies have further stimulated the development of microsensor systems. Broadly defined, optical transducer technologies can be divided into the following categories:
miniaturized versions of bulk optical systems
evanescent wave techniques
Current sensor approaches rely on assays, require multiple steps, and use multiple reagents, all of which are difficult in field applications and require trained personnel. Therefore, considerable efforts have been made to miniaturize sensor systems that are based on fluorescence approaches, which have demonstrated detection sensitivity in laboratory tests and about which an extensive knowledge base exists.
Recently, direct-detection sensing methods based on evanescent (very short duration) wave and micro-optomechanical transduction techniques have been developed. Evanescent planar-wave sensors based on integrated optic interferometers are capable of detecting biological agents, either directly or indirectly through a two-step labeling process. Chemical sensing techniques using biological receptors to detect chemical species have also been demonstrated. All of these techniques depend on large quantities of analyte.
Unlike other devices that rely on mass change, microoptic devices detect refractive-index changes. The source of the change can be simple adsorption onto a surface film, the binding of a molecule to a wave-guide surface, or a reversible reaction on a wave-guide surface. The latter would enable active chemistries involving chemical reactions to be used for reversible in situ sensing. The electric field associated with a guided optical wave interacts with the target at the atomic or molecular level; thus reactions, even those involving small molecules, can be detected by the charge transfer caused by the reaction.
Some research has been focused on detecting, verifying the sequence, and measuring nucleic acids, such as DNA and RNA, in a “chip” format (i.e., on a planar surface). With this format, the chip, or surface material, would be inexpensive and easy to analyze. Hybridization is typically used as a source of specificity, and plastic or glass planar structures or beads are used as the chip. Detection of hybridization can be done by several methods, including fluorescent imaging or detection of labeled DNA probes (used by Affymetrix and Synteni), electrical signals, such as conduction, impedance, and electron transfer (used by Clinical Microsystems), refractive index, and others.
Solution-phase detection (used by Luminex) has also been demonstrated as a way to show if and where hybridization occurs relative to the position of the probes on the surface. Methods of making efficient, high-fidelity sequence probes range from the direct synthesis of nucleic acids on the chip (Affymetrix) or surface to the coupling of prefabricated sequences (known or unknown) with various attachment chemistries (Nanogen). Because the hybridization reaction to a surface in a microvolume can be slow, providing turbulence to extend the reaction time is a problem. Using electric fields to move the anionic DNA molecule, demonstrated by Nanogen, has overcome some diffusion issues; however, the electrodes in buffer solutions still undergo electrolysis, producing gases and plating reactions.
The formation of gas from electrolyses has long been an issue, but it can be resolved with proper electrode design (Keim and Ladisch, 2000). Similar principles are likely to be applicable to the microscale, thus providing a starting point for future research. Recent progress has made toward increasing the manufacturing rate and reducing the cost of DNA arrays (Corning, 2000).
The term DNA chips has been used to refer to miniature devices for analyzing molecules, such as nucleic-acids molecules and other biomolecules (e.g., peptides, proteins, carbohydrates), particularly the miniaturization of separation mechanisms (e.g., electrophoresis, isoelectric focusing of peptides and proteins, liquid and gas chromatography, surface plasmon resonance, electrochemical detection). PCR-based applications of miniature reaction chambers are often referred to as PCR chips.
Hybridization between DNA and DNA, RNA and RNA, or combinations including chemical analogues, is a simpler, more robust chemistry than using proteins or other labile biomolecules but is not foolproof. Chip surfaces must be nearly perfectly tuned to perform (or enhance) the recognition event that dictates specificity and sensitivity. DNA-based nucleic-acid assays have been enhanced via microsystems and microfluidics.
Although DNA chip technology is promising, it has many limitations, including reliance of the analytical portions on relatively large instruments (e.g., lasers, photomultipliers, microscopes). Simple correlations of DNA sequence and physiology (phenotype) are still far from adequate for simple, yet highly functional, chips. Proteins, which more closely represent the physiological action that results from life’s activities, should be investigated at the system level.
Protein sensors can be divided into sensors designed to detect, and perhaps quantify, molecules in a biological sample from a person, such as a potentially infected soldier, and sensors designed to detect molecules in samples taken from the environment, which could, perhaps, warn of an at-
tack by chemical or biological warfare agents. Because of market incentives for advances in biomedical applications, most commercial sensors are being developed for samples from people.
Sensors That Detect Proteinsin Biological Samples
A protein chip is a device that can detect the presence, and sometimes the amount, of specific proteins in a sample. By 2025, the technology for protein chips to be used for diagnostic purposes in the field might enable a chip to provide information about 20 to 100 proteins in a sample of, say, saliva. A protein chip used for genomic research in a laboratory might provide information on all of the proteins and posttranslational forms of proteins encoded by an entire genome. The development of protein chips will require the development of a number of critical technologies that are currently the subjects of basic and applied research.
Generation of Affinity (Capture) Reagents
DNA chips for monitoring gene expression depend on the fact that single-stranded nucleic acids can be isolated from a solution and unambiguously identified by hybridization. DNA chips are not absolutely perfect in DNA assays; however, analyses must be based on patterns derived from repeated experiments. Proteins consist of polymers of 20 different building blocks (amino acids) compared to only four for DNA (nucleic acids). Agents that can bind to, and thus capture, specific proteins from solution include antibodies, nucleic-acid aptamers, and peptide aptamers. However, no base-base recognition systems have been developed.
For antibodies to be generally useful, methods will have to be developed to produce them in large numbers in vitro rather than by immunizing mice or rabbits and collecting serum in a chemically homogeneous form (i.e., with no different glycosylated states). Cambridge Antibody Technology, with pharmaceutical partners, has developed a system for producing homogenous antibodies on a significant scale. Other companies (e.g., Morphosys, Dyax, and Bioinvent) are also active in antibody phage display research. Army applied research at Edgewood Arsenal is focused on analogous methods for chemical biological warfare defense applications.
Nucleic-acid aptamers, described by Ellington and Szostak (1990), are nucleic-acid reagents selected in vitro. An aptamer is an RNA or DNA molecule that assumes a particular shape and surface charge distribution that enables it to bind to a target. Aptamers can be isolated against proteins and produced in large quantities in vitro.
Peptide aptamers are proteins from combinatorial libraries that consist of fixed scaffolds and one or more variable regions encoded by random sequence DNA (Geyer and Brent, 2000). Peptide aptamers can be isolated in vivo and in vitro and can be produced in large quantities. Phylos, Inc., is the most significant commercial entity that can scale up production of peptide aptamers, but its interest is in pharmaceuticals. An effective protein chip to meet requirements for battlefield environmental sensors would require the systematic, scaled-up production of many, many different capture agents.
Detection of Captured Proteins
The most sensitive detectors of DNA (e.g., the devices made by Cepheid) depend on amplification of DNA by PCR, by which a single piece of DNA can be amplified 80 percent of the time. In most tests, 10 examples of the same sequence can be detected with near 100 percent reliability. Proteins, however, cannot be currently amplified by in vitro methods. Therefore, current devices for detecting proteins are less sensitive. These devices may be based on bioreceptors fixed to their surfaces to capture protein targets. Hence, methods of fixing bioreceptors (which are also proteins) must be addressed by future research (Bashir et al., in press).
Broadly speaking, proteins in a sample can be detected in four ways. One is competition, a technique in which proteins in a sample bind to a capture reagent, in the process bumping off a detectable molecule bound to the capture agent. Various techniques are then used to turn the detached molecule into an amplifiable signal. This technology is not common for protein chips.
The second detection method is to capture the protein and then detect it with a direct or indirect label. In direct labeling, the label (e.g., a radioactive atom or fluorescent dye) is attached to the protein by a chemical or covalent bond (e.g., a fluorescent molecule chemically attached to the cysteine, an amino acid). One could, for example, label all of the proteins to make them fluorescent and then detect their binding by fluorescence at the surface. In indirect labeling, the protein, before or after capture, is bound by another molecule that sticks tightly to it but is not covalently linked to it. This second molecule has some property that makes it detectable; it might be fluorescent, for example, or it might be linked to an enzyme that changes a colorless chemical to one whose color can be easily detected. With either direct or indirect labeling, it is difficult to label proteins uniformly so that the signal is proportionate to their abundance in the sample.
The third detection technique is to use a physical phenomenon that depends on binding of the native, underivatized protein to the affinity reagent. A number of approaches have been tried, such as detecting changes in mass at a vibrating surface, which alters the vibration of piezoelectric cantilevers or the transmission of sound at specific frequencies through surface acoustic-wave devices. Another technique, which involves optical phenomena, is based on changes in reflection of light from a surface caused by changes in the refractive index near that surface in the dis-
tance (also called surface plasmon-resonance phenomena). The main problem with these approaches is that a large amount of protein mass must be captured to cause an effect. Only the optical techniques appear to be feasible for the highly sensitive detectors that can be used to sense small (submicrogram) numbers of molecules.
The fourth means of detecting proteins is to capture them on a surface and then analyze the bound proteins by mass spectrometry. Insufficient sensitivity is a problem, but some instruments have subfemtomolar (below 10−15 mole) sensitivity, and, if a patterned affinity surface is used to capture different proteins, the mass of the bound protein expected to be captured in each spot could be determined.
Systems Engineering and Device Design
Some of the challenges to systems engineering and device design are common to protein detection devices: ruggedness, power requirements, and integration into an efficient system. Another problem, which also applies to other diagnostic devices, is preparing the sample. Proteins differ from each other chemically and have different properties in solution. They must be extracted from the biological sample in a form that protects the properties (e.g., their native structure) the device uses to detect them. Bioseparations engineering must be applied.
Another serious challenge is cost. Materials with the best defined and most manipulatable surface properties, such as silicon, are expensive to fabricate. A disposable or “few-use” device using these materials could be expensive based on current technology. Plastics might be used if they can be integrated with electronic and biological species. The surface chemistry of the plastic would have to be tailored to be compatible with biological species.
All of these challenges are likely to be met by commercial research and development. The Army should monitor developments and apply its resources to adapting commercial technologies to military uses.
Cells on a Chip
Like canaries in a coal mine, living cells have been proposed as sensors for environmental toxins or for pharmaceutical screening. Microelectronic arrays have been used for monitoring metabolism and cellular activity and measuring cellular responses to a variety of agonists. Because several types of cells can grow and thrive on surfaces, the cell/sensor electrode interface can be a transducer of both intrinsic and induced electrical activity. For example, packaged, miniaturized platinum electrode arrays have been used for monitoring and analyzing electrical activity from a spontaneously beating synctium of cardiac cells, providing direct measurement of cell responses to a variety of ion-channel-affecting pharmaceutical compounds (Borkholder, 1999). Impedance measurements of neurons and glial cells were targeted for measurements of cell membrane conductance, cellular adhesion, and motility.
Cells-on-a-chip systems show promise as sensors of easily measured responses, such as beat rate (cardiac cells). Action potentials can be correlated with biologically anticipated responses to stimuli, such as pharmaceuticals and toxins. Issues to be overcome include cell-to-substrate adhesion, signal input/output interpretation, cell culture to cell culture consistency, cell culture chamber designs, increasing the variety of cell types, and keeping cultures alive and well in the field. Research is still at an early laboratory stage, and application of the technology to portable sensing devices is not likely in the near future.
BARRIERS TO THE DEVELOPMENT OF PORTABLE SENSORS
Because target proteins may be diluted in samples that also contain air (or an aerosol) and contaminants, such as spores, the sample sizes that can be obtained are often too small for detection. Enough material must be captured so that proteins (or other “fingerprints”) can be identified that are diagnostic for particular pathogens. Other immediate barriers to the development of portable, low-power sensor devices include:
the need for multiple reagents
complex systems that are not reliable enough for unattended operation over extended periods of time or operation by untrained operators
miniaturization without loss of sensitivity
weight reduction from the current 10 pounds to hand-held, wearable systems
improving specificity to reduce false positives and false negatives
Because there are few incentives for industry to develop devices that can measure proteins in a rugged battlefield environment, research and development will have to be supported by the government. Commercial developers of environmental sensors and many kinds of diagnostic sensors have paid little attention to capture reagents, active surface requirements, sample collection, or sample preparations. Most of the physical phenomena used to detect binding will not be sensitive enough to detect proteins in samples taken from the environment at large. However, current developers are addressing some systems engineering issues (e.g., power consumption).
Combined electronics and biological systems for detecting biohazards have been demonstrated in cultured cell preparations where normal cell activity is modified through exposure to the hazards of interest. Problems have arisen, not just in developing the electronic technology to interface with the biological system, but also with keeping prepara-
tions viable in the field and providing a stable response in the absence of toxic agents. Demonstrating these systems and exploring their potential use for guarding against specific pathogens will require a great deal more work.
Sample preparation remains one of the most challenging aspects of miniaturizing sensors for chemical and biomolecular analysis systems. Most real-world samples include background material, such as soil, tissue, biological materials, ions, and metals, all of which create differences in sample-processing methods. These processes range from simply diluting the sample to reduce the concentration of a material that inhibits the detection process or assay to elaborate procedures requiring large, sophisticated equipment. Methods used for different samples are rarely interchangeable. For example, the use of focused ultrasonic energy can easily break open a bacterial spore to release nucleic acids. However, too much energy can shear the same molecules and ultimately decrease the analytical sensitivity of the method.
Basic research will be necessary to study the fundamental physics and chemistries of sample purification processes and possibly to develop universal sample processing methods. Research in different types of detection mechanisms may also be useful to the Army. In addition to detection mechanisms with biological components, detection mechanisms might be based on infrared (thermal) signatures or piezo-electric, audio, or magnetic effects.
Miniaturized, biologically based sensing devices could increase battlefield intelligence and significantly counter unseen environmental threats. Timely sensing of biological, as opposed to chemical, agents will require a broad-based network of both internal and external sensing devices. These devices will require development of micro/nanotechnologies, as well as testing facilities to validate the resulting products. Many of the micro/nanotechnologies necessary for these devices will only be developed if the Army provides clearly defined requirements.
To influence the direction of commercial developments, the Army should immediately devise strategic and tactical concepts for the detection of target threat molecules on future battlefields. These concepts should identify Army-unique requirements for internal and external sensing, monitoring, and networking capabilities over and above those being developed for commercial applications and chemical-biological defense requirements.
The resulting concepts are likely to require significant miniaturization of sensors and sensor components that will not be pursued by commercial developers. For this reason, the Army should support basic research that will facilitate the miniaturization of biosensor capabilities for both internal and external applications (see Chapter 6 for a discussion of sensor miniaturization biotechnologies). Other Army research should address specific barriers to development of portable sensing devices outlined above.