By David N. Leff

In the present post-Cold War era, terrorists have become private enemy number one.

At major airports, trained dogs sniff for pathogens, narcotics and explosives in suspicious-looking baggage. But since 1981, the U.S. Department of Defense (DOD) has been looking for a higher-tech means of detecting explosives than canines. One concept that DOD¿s Defense Advanced Research Projects Agency (DARPA) was scouting at the time read: ¿Exploit the highly specific binding properties of a monoclonal antibody as the key element in a microelectronic device, then greatly amplify the resulting signal to report the molecule¿s presence.¿ That was then.

Now, DARPA calls its program for phasing out Fido ¿Operation Dog¿s Nose.¿ One project that program supports is a prototype device reported in today¿s issue of Nature, dated April 22, 1999. Its title: ¿Stochastic sensing of organic analytes by a pore-forming protein containing a molecular adapter.¿

The paper¿s senior author, and principal designer of the hypersensitive sensor, is protein engineer Hagen Bayley, who heads the department of medical biochemistry and genetics at Texas A&M University, in College Station.

His instrument for detecting and identifying single molecules, Bayley told BioWorld Today, fits the ¿dog¿s nose¿ paradigm neatly. ¿It does have an analogy with olfaction,¿ he observed. ¿There are carrier proteins in the nasal passages that deliver organic hydrophobic molecules to the sensory neurons, so in a sense the receptor molecule that we¿re using can be seen as analogous to these olfactory binding proteins.¿

Bayley describes his system as an application of stochastic probability theory.

¿We¿ve got just a single molecule as a detector,¿ he explained, ¿so the analyte ¿ the thing we¿re measuring ¿ is going to be interacting in a randomized way. Thus, if you had tens of thousands of detector molecules, each with a certain binding affinity, when you expose them to the receptor, maybe 5,000 of them would be occupied at the binding site at any particular moment. But, when you have a single molecule to measure, you can actually see the analyte coming on and off the detector. So, in that sense it¿s a stochastic process.¿

This action plays out within a microscopic pore or channel drilled through a cell¿s double membrane. The drilling is done by a familiar human pathogen, Staphylococcus aureus. The drill this germ uses to punch holes in its victim¿s cells is a bacterial toxin called alpha-hemolysin. Those holes are not simple; they have an intricate architecture at the Engstrom level.

Pore formation begins when seven of these hemolysin molecules self-assemble to form a 15-Engstrom-diameter channel through the 50-Engstrom-thick membrane bilayer. But the resulting tunnel is twice as long, extending outward in a bulbous chamber surrounding its elongated internal pore.

Into this heptameric (seven-sided) aperture, Bayley and his co-authors snugly fitted a doughnut-shaped sugar molecule called beta-cyclodextrin. ¿Usually,¿ he pointed out, ¿people use cyclodextrins to solubilize molecules that don¿t dissolve in water. Look at a list of ingredients on cosmetics ¿ products like that ¿ you¿ll find they¿re used pretty heavily.¿

Bayley¿s device uses heptameric beta-cyclodextrins as receptors to bind molecules of interest on their way through a pore. To single out a molecule for detection and identification, he explained, ¿we measure the electric current that flows through one of these protein molecules. When the cyclodextrin binds this analyte, the current flow changes transiently, and little blips represent individual binding events.¿

Sorting Out Two Neurological Drugs

As a demonstrator model, the group chose two therapeutic drugs: the antipsychotic chlorpromazine, and a closely related antidepressant, imipramine. ¿They have very similar structures,¿ Bayley said, ¿and, as you can see in the Nature paper, they each gave quite different binary signals with this type of detection method.¿

To identify such molecules in a vast mixture, he explained, ¿the chlorpromazine binds for a much smaller time than the imipramine. So it gives a characteristic signature to its binding.¿

Going from lab to bedside, Bayley envisioned how the detection device might eventually serve in clinical practice. ¿For example, cancer chemotherapeutic agents are very toxic,¿ he said. ¿You have to have a good handle on the narrow range of concentrations in the blood plasma that¿s being delivered to the patient. And you would in principle be able to monitor that in real time, if you could use such a device.

¿There are lots of applications for sensing,¿ Bayley said, ¿in situations where such analyzing is not particularly complex ¿ like some narcotic drug that you¿ve confiscated at the airport, or [when you are] looking for explosives on a package. Also, pollutants in the environment, and toxins in living organisms. If we kept doing electric recording, as we¿re doing now, we¿d probably build a dedicated circuit just to recognize the signals that you were looking for out in the field ¿ at that airport, say. Whoever was operating the machine would like to have a kind of yes-no readout. And if it¿s yes,¿ they¿d look at things more carefully.¿

Shape Of Things To Come

But a hand-held device pressed against a suitcase is fairly far in the future.

¿Not at this point,¿ Bayley observed. ¿Now, our device is just a laboratory prototype which needs a lot more development to get into the outside world. It¿s right at the cutting edge of what can be done in the lab right now.¿

Last November, the university applied for a patent on his invention, claiming ¿stochastic sensing mediated by carrier molecules.¿

¿We¿re just basic scientists,¿ Bayley said, ¿so we¿re going to explore the range of different analytes that can be detected in this way. And then we may collaborate with other groups, and maybe small companies, to make a portable, rugged, inexpensive device.

¿We have a few firms that are interested,¿ Bayley concluded, ¿but we don¿t have any really solid arrangement with any of them, because this is really so new. I think it will take some time.¿ n