Using Biotechnology to Detect and Counteract Chemical Weapons

Biocatalytic methods of detection and decontamination offer several advantages over existing methods.

The tragedy of September 11, 2001, and the ensuing anthrax attacks heightened public and governmental awareness of the need for reliable, cost-efficient, and deployable diagnostic and treatment systems for chemical weapons. Many of the present methods require cumbersome equipment and complex analytical techniques. In addition, many decontamination solutions are toxic, corrosive, and flammable, and therefore not appropriate for use over large areas or with personnel. Biocatalytic methods of detection and decontamination/demilitarization offer several advantages over existing methods.

Enzymes are environmentally benign, highly efficient biological catalysts on appropriate reactants, with hydrolysis rate enhancements exceeding a million fold over the uncatalyzed reactions. They may also be used in conjunction with other processes to limit the environmental impact of existing systems. Perhaps the most attractive feature of enzymes is that they can function effectively under ambient conditions, thus decreasing energy costs and increasing safety (especially when agents are stored in proximity to explosives). To date, enzymes are the most effective known catalysts for degrading nerve agents (organophosphates [OP] are the class of compounds commonly referred to as nerve agents). Enzyme concentrations in the micromolar range are sufficient to degrade nerve agents on contact, and just 10 milligrams of enzyme can degrade as much nerve agent as 1 kilogram of concentrated bleach (a conventional method of degrading nerve agent).

The study of OP biocatalysis dates back to the mid-1940s, when Abraham Mazur found that mammalian tissue hydrolyzed OP esters (Mazur, 1946). Douglas
Munnecke (1977) used cellular extract from a mixed bacterial culture as a catalyst to detoxify organophosphorous pesticide. The rate of enzymatic hydrolysis of parathion, an OP ester commonly used in pesticides, was found to be nearly 2,500 times the rate of hydrolysis in 0.1 Normal sodium hydroxide. Munnecke (1979) later attempted to immobilize the extract on porous glass and porous silica beads. It was reported that 2 to 5 percent of activity present in a cellular extract could be retained by the beads, and 50 percent of the immobilized activity could be maintained for a full day under ambient conditions.

The enzyme organophosphorus hydrolase (OPH) was later isolated and shown to be an effective catalyst for the degradation of a range of OP esters. Frank Raushel and colleagues have studied extensively the activity and stability of OPH from Pseudomonas diminuta (Dumas et al., 1989). At Texas A&M University, James Wild’s laboratory has even produced the enzyme in corn (personal communication). Squid diisopropylfluorophosphotase (DFPase), an enzyme capable of degrading nerve agents, such as soman, is now in commercial production. In expanding the library of such enzymes, Joseph DeFrank and colleagues from the U.S. Army have discovered and analyzed another nerve agent-degrading enzyme, organophosphorus acid anhydrolase (OPAA) (DeFrank and Cheng, 1991).

Many species produce enzymes that efficiently degrade nerve agents, although the natural function of these enzymes remains unknown. The class of OP compounds has only been exposed to nature for a few decades, certainly not long enough for natural systems to have evolved enzymes to degrade them. Thus, the ability of enzymes to degrade these compounds is still an unexplained quirk of fate. The only nerve agent-degrading enzyme with a known natural function is OPAA, which is a peptidase enzyme (an enzyme that catalyzes the hydrolysis of peptides into amino acids). Interestingly, this enzyme is far less active with its natural substrate than with soman.

Cells must constantly control their environments and, therefore, must be in a position to rapidly manipulate the concentrations of biocatalysts. Thus, many proteins, such as enzymes, last for only minutes or hours under ambient conditions, and enzymes have generally evolved to be unstable molecules. A number of research groups around the world have been working for many years to stabilize enzymes for numerous applications, including catalysts for the decontamination of nerve agents. These mostly military researchers are also part of a NATO project group (PG31) working to formulate "green" enzyme-based decontamination systems.

Because the preparation and purification of enzymes can be costly, enzymes must be immobilized to increase their reusability and stability. Effective immobilization requires that the enzyme-containing material be prepared so that the enzyme maintains most of its native activity, maintains a high operational stability in its working environment, and maintains a high storage stability. In conventional immobilization methods, either covalent or ionic interactions link the protein to a support material. Enzymes have been immobilized on a wide variety of support materials, including alumina pellets, trityl agarose, and glass/silica beads. Polymers also make excellent enzyme-support materials because of their structural flexibility and solvent resiliency. Numerous polymers, including nylons, acrylates, and several copolymer blends, have been used as effective supports.

Combining the power of biology with the sophistication of polymers presents considerable opportunities. Synthesis of enzyme-containing polymers involves the covalent immobilization of the enzyme directly into the polymer network via reactive functionalities on the enzyme surface, thus ensuring retention of the enzyme in the polymeric material. Bioplastics prepared in this way exhibit remarkable stability under normally deactivating conditions and, thus, are ideal for the development of the next generation of responsive, smart biomaterials. One of the most pressing needs at the beginning of the twenty-first century is for active, smart materials for chemical defense.

The threat of chemical weapons is now ubiquitous, and these poor man’s nuclear weapons can kill on contact. Existing chemical defense polymers are designed to provide an impenetrable barrier between us and our environment. By contrast, layered materials use polymers to bind toxic chemical agents irreversibly to the material. Degrading toxic chemicals requires catalysts that can be incorporated into the polymer in a stable and active form.

In protective clothing, a layer of polyurethane, foam-entrapped, activated charcoal is embedded between several layers of polyester fabric. Polyurethanes are effective substrates for OP adsorption, and reports have documented that polyurethane foam particles can be used as adsorbent materials for pesticide vapors in farming fields. If proteins could be incorporated into polyurethanes, some interesting materials might emerge.

Enzyme-containing polyurethane materials are ideal matrices because of their ease of preparation, the large range of polymer properties that can be prepared, and multipoint, covalent attachment of the enzyme to the polymer. Bioplastics are prepared by reacting a polyurethane prepolymer that contains multiple isocyanate functionalities with an aqueous solution containing the enzyme. The solubility of enzymes in this aqueous phase can be significant, enabling loadings up to 15 percent. Foams, gels, and coatings can be prepared depending on the reactivity of the isocyanate. Any enzyme that is present in the aqueous solution can participate in the polymer synthesis, effectively creating an enzyme-containing-polymer network with multipoint attachment.

The key question is how well such biopolyurethanes function. The answers are striking. Enzymes can be stabilized from days to years, and the enzymes in no way alter the physical properties of the polyurethane. The derived materials are so active that just 1 kilogram of enzyme immobilized in a multipoint covalent fashion is enough to degrade up to 30,000 tons of chemical agent in one year.

Detoxifying a chemical agent is only part of the technology we will need to combat the real and present danger posed by these materials. Once a toxic agent binds to a surface, it must be degraded and identified. Indeed, every time a chemical or biological sensor gives a false-positive result, tens of thousands of dollars are invested in responses that are not needed. In the Middle East and Africa, false positives could lead to war; and spurious, sensor-triggering events could change the course and nature of a war.

Chemical sensors in particular are generally unreliable, and engineering solutions to the problem will require overcoming some formidable technical challenges. We must first understand how chemical weapons work and use that knowledge to design biologically based sensing devices. The common feature of biosensors to date has been a lack of stability. An effective chemical-weapon sensor must not only be able to operate under ambient conditions, but must also be able to operate in extreme desert and arctic conditions.

A material that is inherently catalytic and can therefore be used to monitor continuously and detoxify detected molecules in real time represents the "holy grail" of this emerging discipline. Indeed, a surface that can both self-decontaminate and sense the progress of decontamination would represent a quantum improvement in protection for war fighters. The rapid detection of chemical agents is critical to inspections performed under the Chemical Weapons Convention, because inspectors are rarely able to control the environment at the inspection site. It is therefore essential that analytical tools vary in sensitivity, specificity, and simplicity.

The selectivity of enzymes that are active on, or inhibited by, chemical agents of interest, makes them attractive components for biosensor technology. Nerve agents exert their biological effect by inhibiting the hydrolytic enzyme, acetylcholinesterase. Thus, the inhibition of this enzyme makes it ideal for use in nerve agent-sensing systems. The most common kit available today for agent detection, the M256A1, uses enzymes and colored substrates to detect nerve agents. Unfortunately, current continuous enzyme-based sensing is limited by the instability of most enzymes and their sensitivity to changes in the environment. The M256A1 functions by reacting phenylacetate with eel acetylcholinesterase to produce a colored material. If an agent or other inhibitor is present, the color does not appear. The test takes 15 minutes to perform and is subject to interference from a number of sources (see Table 1 in full [pdf] version). Another drawback is the inability of the M256A1 to distinguish between different nerve agents. All of these techniques rely on the absence of a reaction to signal the presence of an agent. These sensors would be much more informative if they provided positive responses (e.g., undergoing a color change when the surface is either clean or contaminated). Nevertheless, enzyme-inhibition assays are commonly used to detect nerve agents.

In 2001, an enzyme-based sensor was developed and fielded that combines all of the desirable features and is resistant to almost all types of interference. The sensor is based on coupling the hydrolytic activity of acetylcholinesterase (an acid-producing reaction) with the biocatalytic hydrolysis of urea (a base-producing reaction). In an unbuffered solution containing substrate, the formation of acid by acetylcholinesterase-catalyzed hydrolysis would lead to a rapid drop in pH. Because changes in pH affect the ionization state of amino-acid residues in a protein’s structure, every enzyme has an optimal pH at which its catalytic activity is greatest (see Figure 1 in full [pdf] version). As the pH deviates from the optimal level, enzyme activity decreases. Therefore, the pH of the unbuffered solution ultimately reaches a point at which acetylcholinesterase is completely inactivated.

Now consider the idealized situation in which the unbuffered solution containing the acid-producing enzyme is supplemented with a second enzyme that catalyzes the formation of base (Figure 1). In this system, the base-producing enzyme (light line) has a pH optimum significantly lower than that of the acid-producing enzyme (dark line). Thus, the formation of base will counteract the production of acid, thereby maintaining pH by creating a dynamic pH equilibrium between the competing enzyme reactions (pH 7.5 in Figure 1). Because the generation of base is biocatalytic, base is produced only in response to a decrease in pH. If the catalytic activity of one of the enzymes is altered via inhibition, the dynamic equilibrium of the system is destroyed; the subsequent shift in pH caused by the inactivation of the dynamic equilibrium can then be used to induce a signal indicating the presence of a target enzyme inhibitor.

This is the basis for biocatalytic, dynamic-reaction equilibrium sensing. The advantages over traditional enzyme sensors include: rapid signal development; strong and intuitive responses; and high resistance to interference by temperature. An interferant-resistant nerve-agent sensor manufactured by Agentase, LLC, couples acetylcholine, acetylcholinesterase, urea, and urease in a polymer with a pH-sensitive dye.

In the rare event a weapon of mass destruction is used, such as another release of sarin in the Tokyo subway, no effective, environmentally benign decontamination systems are available. Even though enzymes provide an efficient, safe, and environmentally benign method of decontaminating nerve agents, several major hurdles still hinder their use. To decontaminate significant quantities of nerve agent, the system will require a good buffering system to handle the large amounts of acid produced from the enzyme-catalyzed hydrolysis. This problem is amplified in low-water environments, where the localized concentration of agent can easily reach 0.3 molar.

Currently, the preferred buffer for enzymatic decontamination systems under development by the U.S. Army is ammonium carbonate. The addition of solid ammonium carbonate to water results in a pH of 8.5 to 9.0 with no adjustment needed, and the ammonium ions are known to stimulate the activity of OPAA (Cheng and Calomiris, 1996). Ammonium carbonate, however, does not have a high buffering capacity, thus making it impractical for use in large-scale decontamination.

Recent studies at Porton Down, U.K., demonstrated that in detergent and microemulsion systems containing 50 millimolar (mM) ammonium carbonate buffer, enzymatic degradation of high concentrations of soman (1.3 to 1.5 percent) resulted in a drop in pH from 9 to 6 in less than five minutes, at which point the enzyme was inactive. Complete decontamination of the agent was not achieved. Increasing the concentration of buffer two-fold decreased the rate of drop in pH; the pH fell to 6 in less than eight minutes. However, complete decontamination was still not achieved. The use of conventional biological buffers at concentrations sufficient to maintain pH in an optimum range for enzyme activity introduces additional obstacles, including potential enzyme inhibition and chelation of essential metals in the enzyme structure.

The biocatalytic, dynamic-reaction equilibrium approach previously described can overcome such obstacles in the large-scale decontamination of nerve agents. Base can be generated on demand in response to the decrease in pH resulting from agent degradation. Consider an unbuffered aqueous system that includes a nerve agent-degrading enzyme, urea, and urease. In the presence of a nerve agent, biocatalytic hydrolysis would cause a rapid decrease in pH. In a method analogous to biocatalytic, dynamic-reaction, equilibrium sensing, this decrease activates urease, which generates base via the conversion of urea. Our laboratory has successfully demonstrated using biocatalytic, dynamic-reaction, equilibrium sensing for the complete decontamination of paraoxon by OPH, using urease-catalyzed urea as a buffering agent (Russell et al., 2002). Such defense systems present a significant advancement in chemical weapons defense, giving military personnel and emergency response teams the ability to achieve complete decontamination using minimal amounts of buffering material.

In summary, chemical weapons exert their terrifying effects on human physiology by interrupting biological processes. The chemicals bind tightly to enzymes, but they can also be degraded by enzymes through fortuitous side reactions. The inhibition and activity of these enzymes can be part of our defensive arsenal. Thus, biotechnology is both the target and a potential weapon in the war on terrorism.

Acknowledgments
This work was partially funded by a research grant from the Army Research Office (DAAD19-02-1-0072) and by the U.S. Department of Defense Multidisciplinary University Research Initiative (MURI) Program administered by the Army Research Office (DAAD19-01-1-0619). The principal author has an equity stake in Agentase, LLC.

References
Cheng T.-c., and J.J. Calomiris. 1996. A cloned bacterial enzyme for nerve agent decontamination. Enzyme and Microbial Technology 18(8): 597-601.

DeFrank, J.J., and T.-c. Cheng. 1991. Purification and properties of an organophosphorus acid anhydrase from a halophilic bacterial isolate. Journal of Bacteriology 173(6): 1938-1943.

Dumas, D.P., S.R. Caldwell, J.R. Wild, and F.M. Raushel. 1989. Purification and properties of the phosphotriesterase from Pseudomonas diminuta. Journal of Biological Chemistry 264(33): 19659-19665.

Longworth, T.L., J.C. Cajigas, J.L. Barnhouse, K.Y. Ong, and S.A. Procell. 1999. Testing of Commercially Available Detectors against Chemical Warfare Agents: Summary Report. Aberdeen Proving Ground, Md.: Soldier and Biological Chemical Command, AMSSB-REN.

Mazur, A. 1946. An enzyme in animal tissue capable of hydrolyzing the phosphorus-fluorine bond of alkyl fluorophosphates. Journal of Biological Chemistry 164: 271-289.

Munnecke, D.M. 1977. Properties of an immobilized pesticide-hydrolyzing enzyme. Applied and Environmental Microbology 33: 503-507.

Munnecke, D.M. 1979. Hydrolysis of organophosphate insecticides by an immobilized-enzyme system. Biotechnology and Bioengineering 21(12): 2247-2261.

Russell, A.J., M. Erbeldinger, J.J. DeFrank, J. Kaar, and G. Drevon. 2002. Catalytic buffers enable positive-response inhibition-based sensing of nerve agents. Biotechnology and Bioengineering 77(3): 352-357.