Feature story image (From left) Seung-Wuk Lee, Jin Huh, Justyn Jaworski, Digvijay Raorane, and Arun Majumdar used nature’s playbook to develop an extremely selective and sensitive receptor capable of detecting explosives.

Feature story imageThe nose knows: scientists have created peptides that mimic the sense of smell by binding selectively to molecules, in this case explosives. In this schematic, DNT-binding peptides are added to a gold surface, creating a detector that only snares gas-phase DNT molecules.

Berkeley Lab News Center Feature Story

May 2, 2008

Sniffing Out Trouble

Contact: Dan Krotz, (510) 486-4019

Berkeley Lab and UC Berkeley scientists have developed a way to detect explosives and other chemicals that promises to be as stubbornly effective as a bloodhound following a week-old scent.

The group engineered receptors that mimic the way people and other animals detect individual scents among a myriad of smells wafting around them. This ability to selectively sniff out a barbecue a few blocks away or predators in a vast forest is honed by millions of years of evolution, in which proteins in the nose become increasingly adept at binding with specific odiferous molecules — and nothing else.

Now, scientists have harnessed the same evolutionary process to fashion receptors that only bind with TNT and another explosive called DNT. They’ve embedded these receptors into a gel-based coating that emulates the sense of smell by detecting minute quantities of the explosive, even in the gas phase. Their work could lead to highly sensitive detection systems, to be used in airports or in the field, which don’t trigger time-consuming false alarms.

“Our approach mirrors nature,” said Arun Majumdar, an expert in sensor technology and a mechanical engineer and materials scientist with joint appointments at UC Berkeley and Berkeley Lab, where he’s the director of the Environmental Energy Technologies Division. “Much like how our sense of smell works, we’re developing receptors that fit lock-and-key to the compound we want to detect.”

Majumdar developed the technology with Seung-Wuk Lee, who holds a joint appointment with UC Berkeley and Berkeley Lab’s Physical Biosciences Division. Other UC Berkeley mechanical engineering and bioengineering graduate students involved in the research include Justyn Jaworski, Digvijay Raorane, and Jin Huh.

Their goal is to use the gel in a new class of portable devices capable of detecting a range of compounds — good and bad — such as explosives, pesticides, disease markers, even food aromas that indicate spoilage or allergens.

Other research groups have created chemical sensors that attempt to mimic the nose. But they have largely focused on fabricating receptors from synthetic polymers that bind to many compounds in addition to the target molecule. For example, a receptor made from synthetic polymers may bind with TNT as well as compounds that aren’t dangerous. A sensor that utilizes such a receptor could indicate it has detected TNT even if it binds with one of the harmless compounds.

“This leads to false positives,” said Majumdar. “You can have a sensor that is very sensitive, but it loses its effectiveness if it’s triggered by many things. We need both sensitivity and selectivity.”

To acquire this selectivity, the research team used nature as their guide. Many biological processes require extremely selective interactions in which a protein binds only with certain molecules. For example, antibodies only attack pathogens, leaving benevolent molecules alone. If they didn’t, life would go haywire. This highly refined selectivity is the gold standard, but it requires millions of years of natural selection to develop.

“And we don’t have time to wait for that,” said Lee, who is an expert in directing evolutionary processes in viruses. To expedite the development of a biologically inspired TNT receptor, Lee used a technology called phage display, which is commonly used by pharmaceutical companies to quickly identify the antibodies that bind with specific antigens. Instead of using the technology to pair antibody with antigen, however, they used it to pair a peptide with a TNT molecule. A peptide is a compound with two or more amino acids that plays a key role in driving biological interactions.

The team randomly synthesized several dozen DNA sequences, and introduced these sequences into the genomes of viruses. Next, they delivered these viruses into bacteria, which generated about 1 billion viruses, each with minor genetic variations in a peptide protruding from the viruses. 

“Then we play the survivor game, and determine which of these roughly one billion peptides binds to the surface of a TNT molecule,” said Lee.

About 1000 peptides passed the first hurdle, but the process is far from complete. They took this pool of “survivor” viruses and introduced them into bacteria, which generated about one million new viruses, each again with minor genetic variations in the peptides protruding from them. This step amplifies the peptides’ desirable TNT-binding traits.

“And then we run another survivor game in much harsher conditions. We want only the strongest TNT binder, and this won’t happen in the first generation,” said Lee.

The final product is a peptide that is fine tuned to bind with TNT, and only TNT. With this peptide in hand, the research team determined the sequence of amino acids that forms the peptide’s all-important binding pocket.

“This is like having a lock, and you create a billion keys, and throw the keys at the lock and determine which fits best,” said Majumdar. “It’s a materials discovery process. Peptides are a class of high-information materials, but the trick is finding the right one.”

In a final test, the team seeded a hydrogel coating with evolutionary screened DNT-detecting peptides. The coating is able to detect DNT even in the gas phase, which is a prerequisite of any sensor tasked with finding explosives hidden in luggage. Their work is part of a broader inquiry into determining the optimum way to detect chemicals of all kinds.

“We can detect chemicals today, but it requires expensive, lab-based machines. We want to take the lab outside, and one way to do this is to emulate the nose,” said Majumdar.

Their research, “Evolutionary Screening of Biomimetic Coatings for Selective Detection of Explosives,” was reported in the March 26, 2008 issue of the journal Langmuir. The research was funded by the Department of Energy, the National Science Foundation, and the Office of Naval Research.

Additional Information

Page URL: