Bio-agent sensing challenges photonics

It seems like it should be easy, but it�s tough for optics alone to meet the stringent requirements of sensing biological agents for security applications.

Nov 1st, 2004
Th 162949

It seems like it should be easy, but it’s tough for optics alone to meet the stringent requirements of sensing biological agents for security applications.

Optical techniques sound like a natural approach for detecting biological agents. Remote sensing systems have been developed for many other applications. They can probe the spectrum of an unknown substance with high sensitivity and resolution at a healthy distance from the material being studied. In principle, they can be quite specific for many materials, including chemical agents.

In practice though, detection and identification of bio­logical agents for military and security applications turns out to be a difficult and complex task. Biosensing systems must operate in environments far more complex than the controlled world of the laboratory. The optical properties of a deadly biological agent like anthrax can differ only subtly from those of innocuous materials that are common in the world at large. Screening programs must worry about false positives as well as false negatives. A false positive rate of one in a hundred samples can be disruptive if millions of samples are being screened and few are thought to be contaminated.

To overcome these problems, optical techniques can be combined with biological and biochemical techniques that offer higher specificity, such as antibodies that react to specific pathogens. In some cases, optical systems can be used as first-stage screening, with biological techniques used for detailed analysis of the selected samples. In other cases, optics can be used to read the results of biological analyses such as antibody tests.

Security sensing requirements

Sensing traditionally involves identification of the material being sought, measuring its concentration, and pinpointing its location. Several problems complicate the task of detecting hazardous biological agents.

One problem is the need to pick out the pathogens in an environment filled with a wide range of relatively innocuous organisms. Bacteria are everywhere, so looking for a common compound present in bacteria won’t provide the needed information. One has to find a marker specific to the particular pathogen being sought, which generally requires testing for a particular protein or nucleic acid using antibodies or nucleic acids that react with them in specific ways.

Another problem is that biological agents can be infectious in very small concentrations-as low as a few particles per cubic meter. This means that sensors must be very efficient in spotting particles and must sample large volumes of air so they can detect very low concentrations of particles.

Requirements for speed and accuracy differ widely among security applications. Troops on a battlefield need quick warning that biological agents may be deployed so they can don ­protective equipment, and they can tolerate a reasonable number of false alarms. First responders in other situations have a lower tolerance for false alarms, but can accept some to get a warning fast enough to try to control the situation. On the other hand, doctors treating people exposed to potential biological agents need to be sure what the agents were to ensure proper treatment is given. In practice, these requirements differ so much that ­separate families of sensors have been developed for particular applications.

Particle and fluorescence measurements

Laser techniques are particularly useful for measurements of particulate sizes, distribution, and fluorescence. Although these techniques generally are not ­specific enough to identify particular pathogens, they can yield vital information on the nature of potential hazards.

One approach is aerodynamic particle sizing by using a laser to measure the velocities of particles in a stream of air passing through the sensor. The airflow accelerates smaller particles more, so particle speed indicates size. The system also can count particles present in the air. The sizes and distribution of particles can indicate the presence of chemical or biological agents, although the technique can’t identify specific agents. To measure the particles, however, the instrument must be located at the site where the contamination is suspected.

Laser radars can measure the properties of aerosols from a distance, an important advantage when trying to warn soldiers of a potential biological attack. Short-wavelength IR lidar sensors can spot aerosol particles smaller than 20 µm that characterize a biological attack, as long as they have a clear line of sight. They can operate at distances well over 10 km (see Fig. 1, top). However, these lidar sensors don’t provide direct evidence that the particles are biological.

FIGURE 1. Infrared lidar (top) can spot clouds more than 10‑km away, and estimate particle size, but they can’t tell if microbes are present. Shorter-range UV lidar produces fluorescence that indicates presence of microbes (bottom), but can’t identify what type.
Click here to enlarge image

Fluorescence measurements go a step further. Short-wavelength lasers can excite fluorescence from some biomolecules. The fluorescence is a hallmark of life. While the emission is not specific to ­certain organisms, it is characteristic of life, indicating the possible presence of a biological-warfare agent. A variety of ­fluorescence sensors can spot signs of life in liquid or gaseous samples illuminated by a short-wavelength laser; flow cytometers measure flow as well as fluorescence. Ultraviolet lidars can spot biological fluorescence from a distance of about 3 km, depending on illumination and atmospheric conditions (see Fig. 1, bottom).

Identifying biological agents

Most living cells contain the same chemical building blocks, so identifying specific species requires more discriminating sensors. The two most effective approaches to species-specific identification are based on immunological reactions and the amplification of nucleic acids. A major practical challenge of both is to increase their speed to get quick results.

The human immune system has developed antibodies that react with specific materials alien to the body, including microbes. Immunological sensing is a two-stage process in which an antibody bonds to the targeted biological agent and the bonding is detected to show that the biological agent is present. Typically, bonding is detected by looking for a change in color or fluorescence. Field kits usually are set up to make it easy for the operator to read out the results visually from the test sample. Automated optical systems can read out the signal by looking for fluorescence.

Genetic amplification techniques are based on the high efficiency of reactions such as the polymerase chain reaction (PCR), which reproduces DNA and RNA in large quantities. Normally biological agents are hard to detect because they are present only in extremely small quantities. However, PCR can selectively amplify ­specific genetic sequences present only in the targeted biological agents. If the agents are present, enough copies of the gene ­segment are produced to detect using ­fluorescence or other techniques; if it isn’t present, there is no signal.

Optical readouts

Optical techniques allow automated readout of genetic or immunological test results. For example, antibodies can be coated on sampling surfaces that are probed optically (see Fig. 2). In this ­simplified system, air is blown over an antibody-coated disk that is probed with a laser beam. If a biological agent binds to the antibody, it fluoresces under laser light, producing light that can be detected to warn of the presence of the agent.

FIGURE 2. A biosensing system blows a sample across an antibody-coated disk, which is illuminated by a laser beam. Fluorescence indicates whether the antibody has reacted with a biological agent.
Click here to enlarge image

A variety of designs are possible. David Walt’s group at Tufts University (Medford, MA) has developed antibody biosensing microbeads that attach to the ends of fused fiber bundles. Each bead weighs only about one-trillionth of a gram. Each batch responds to one specific biological agent, and these batches can be combined to make composite sensors that respond to multiple agents.

The group assembles the sensing system by first etching the ends of the bundles, creating small pits at the sites of the fiber cores. Then microbeads are added from batches that respond to different bioagents and have different color-­coding. The beads attach to random locations on the fiber ends, and their locations can be calibrated according to their different color responses (see Fig. 3).

Click here to enlarge image

FIGURE 3: The fluorescent response of sensing beads at the ends of a fiber array can identify biological agents. The fluorescence intensity depends on concentration of fluorescing compounds in the beads, as shown for europium-doped beads used to calibrate locations of the beads (top). Two or more types of beads that respond to different biological agents can be distributed across the array. The fluorescence intensity of individual beads depends on the concentration of the agent they are responding to. Green spots (center) are beads responding to Interleukin-6, and yellow spots (bottom) are responding to Alexandrium tamarense.

Similarly, fluorescent dyes that bind to nucleic acids can be added to gene-­amplification systems. Optical systems can measure the fluorescence to determine if sequences from biological agents have been amplified.

Implementing systems

A major challenge is to integrate the variety of biosensing technologies into practical systems for military and homeland-security applications. A good example is the systems that the Joint Program Executive Office for Chemical and Biological Defense has developed to meet military requirements for both early warning and identification.

The Joint Biological Standoff Detection System (JBSDS) is intended to provide early warning of biological-warfare agents from a safe distance-a “standoff.” The first level is an IR lidar system that can detect and track aerosol clouds deployed by enemy shells at distances up to 15 km. Goals are to spot clouds at night against natural backgrounds, and to identify whether the cloud is natural, like a fog, or man-made. The second layer is a UV lidar system, capable of discriminating between biological and nonbiological particles at distances up to 3 km, but unable to identify specific agents. The standoff system can be operated from a fixed site or on mobile platforms that stop to make measurements.

The Joint Biological Point Detection System (JBPDS) includes an automated set of instruments that can be deployed in the field for quick identification of biological agents. The goal is to identify up to 10 biological-warfare agents simultaneously, providing results within 15 minutes. The system also collects samples that can be taken to a laboratory for further analysis.

An alternative is a portable field kit called the Joint Biological Agent Identification and Diagnostic System (JBAIDS), which can identify biological-warfare agents within 40 minutes after a sample is prepared. Like the point-detection system, it can simultaneously identify multiple biological agents. Designed for use in field medical units or fixed medical labs, it includes test kits for pathogens, an ­analytical platform, and a laptop computer ­programmed for quick analysis.

Looking to the future

We’re still a long way from ideal systems that can give more specific early warnings of the type of agents present. Developers are working on advanced sensing systems such as “electronic noses,” which collect a variety of information and ­analyze it to determine what potential agents are ­present. Such systems would analyze the ­varied responses from different sensing elements. Artificial-intelligence techniques could be used to train them to recognize known biological agents-and to recognize when unknown agents share important features with known pathogens. For now, the main goals are faster, better, cheaper, and more accurate versions of today’s laboratory biosensors.

More in Test & Measurement