Exploiting Nature's Molecular Mechanisms to Combat Bioweapons

John Dileo and Elaine Mullen

lmost daily the media reports on a new medical or public health breakthrough made possible by advances in biotechnology.

Those same breakthroughs, however, increase the potential for bioterrorism as it becomes easier for individuals to generate significant quantities of deadly organisms that occur in nature. This could mean, for example, using anthrax in the subway system or adding other pathogens to U.S. water supplies. In addition, advances in synthetic biology are making it possible to design and produce hybrid or completely new organisms. To prevent the consequences of such possibilities, government agencies are evaluating the measures in place to deal with bio-related incidents. One area of emphasis is the monitoring of the environment for the intentional introduction of bioagents. While progress has been made in developing sensors for many agents, gaps remain. MITRE is lending a hand to close those gaps as quickly and fully as possible by working on new approaches to biosensor development.

There is also a significant need worldwide to monitor rivers, reservoirs, and other water supplies for a sudden increase in the concentration of organisms that could cause an outbreak of infectious disease. Through our own research program, we are attempting to develop such a capability by taking advantage of the molecular mechanisms involved in the process of infection. We're focusing on selectively capturing pathogens from water using inexpensive biological materials.

To cause infection, a virus or bacterial cell must cling to its host. Disease-causing organisms typically produce proteins called lectins that adhere to tissue-specific sugar sequences on the surfaces of our cells. These same sugars are often present in natural foods. For example, pigeon egg white glycoproteins bear sugars that bind to several human pathogens and prevent them from causing illness.

We are investigating the use of these sugars to capture waterborne pathogens. Working with scientists at Johns Hopkins Applied Physics Lab and GEOMET's Biodefense Lab, we are developing and testing "floating films" containing glycoproteins.

As we began to design films that would selectively capture human pathogens, we discovered we would need two resources: a database of sugar sequences bound to readily available glycoproteins, and a database of human pathogens and the sugar sequences to which they bind. Through a search of the GlycoSuiteDB, an online database of complex carbohydrate structures attached to proteins, we quickly identified readily available sources of specific sugar sequences.

The other resource, however, didn't exist, so to learn which cell-bound sugar tags are targeted by human pathogens we had to create a database from scratch. As part of this project, student intern Baddr Shakhsheer spent two summers conducting an exhaustive search of the literature published during the last 20 years in the fields of microbiology, glycobiology, and infectious disease. We used this information to populate a new database, which we now call SugarBindDB™ and offer to the public at http://SugarBindDB.mitre.org. This database will aid scientists designing carbohydrate-based therapeutics to prevent or ameliorate disease caused by microbes and biotoxins that bind to cell surface carbohydrates.

With the new SugarBind database, we could continue to investigate the development of biocapture (floating) films to detect certain pathogens. How does this work? As sugar molecules are released from complex carbohydrates by microbial enzymes, carbon dioxide gas is produced. Gas trapped under the film produces bubbles that potentially could be detected to indicate specific pathogens in water supplies. These films will contain two different carbohydrate structures: one to bind to pathogens, the other to signal the successful capture.

We expect the films to be used as a low-cost method to rapidly survey surface waters, facilitating the collection of water samples for analysis and providing early warning of biological contamination. The films are made from the fluids of plant and animal tissue, which contain glycoproteins. Mixed with oil, the fluids produce a film that floats on the water's surface with the glycoproteins facing toward the water so they can contact the waterborne pathogens. We are currently developing methods to test the efficiency of films that selectively capture harmful strains of E. coli that bind to different sugar sequences. In many cases, biotoxins produced by these organisms can be captured using the same films.

Potential Bioweapons

Another area of concern for the nation's defense is the sheer number of potential bioweapons. The gov-ernment lists more than 30 high-priority individual agents, as well as entire classes of pathogens such as "other pox viruses." To cope with such a diverse scope of organisms, a rapid biosensor development capability is needed.

One potential method for developing such a capability involves adapting molecule-binding proteins, such as antibodies and receptors, to serve as sensors. Just as these proteins can detect unwanted organisms in the body, they can be adapted to detect unwanted organisms in the environment. The current approach to adapting them, however, is labor intensive and time consuming. It requires constructing large libraries of diverse proteins, followed by multiple rounds of "directed evolution" in which modified proteins that can specifically and selectively bind the target molecule are identified. Recently, a group led by Duke University's Dr. Homme Hellinga devised an algorithm to redesign natural biosensor proteins to bind to whatever chemical the designers want. In this method, the "directed evolution" is done using computational methods instead of laboratory selection. Now, instead of taking weeks to months, a receptor can be developed in days.

The basis of this system is an E.coli sugar-binding protein equipped with two structural domains that clamp together when they encounter their specific sugar target (or ligand). Using Duke's computational design process, the protein is changed so that it binds to the agent of interest instead of its usual sugar. The process is initiated by placing a model of the new target in the binding site of a "virtual" receptor. The program then sequentially changes the amino acids involved in binding, searching for sequences that form a shape and chemistry complementary to the new target. Based on the number of amino acids involved in the binding, this step can create up to 10²³ possible sequences. After accounting for the various possible rotational conformations that might be adopted by those amino acids, the requirements expand the combinatorial problem to between 10⁵³ and 10⁷⁶ permutations.

To solve this problem, the system uses a modified "dead-end elimination" algorithm that eliminates all but a few of the most promising candidates. When this method was used to develop a sensor specific for the explosive TNT, the number of candidates was pared down to 17. These virtually designed receptors were then evaluated in the lab to find the best candidate. The selected protein is now being field tested in a system for locating sunken munitions. Proteins are also being developed using this method to detect such chemical and biologic agents as VX, Soman, and Sarin.

MITRE is making plans to advance this research by developing a standard software package for protein receptor/sensor design. Currently, the program that does the computational design is coded by hand on an as-needed basis.

We also plan to extend the technology to biowarfare (BW) agents, as all the current sensors designed by this method detect only chemical warfare agents, not BW agents, such as bacteria or viruses. Sensing a BW agent is much more complex, requiring the computational modeling process to be performed on two proteins.

Through projects such as these, our teams are working closely with government agencies and other research facilities to speed the development of biosensors.

For more information, please contact John Dileo or Elaine Mullen using the employee directory.

Solutions That Make a Difference.®
Copyright © 1997-2013, The MITRE Corporation. All rights reserved.
MITRE is a registered trademark of The MITRE Corporation.
Material on this site may be copied and distributed with permission only.

Privacy Policy | Contact Us