For the first time, researchers have shown that a nonbiological molecule called a plastic antibody can work just like a natural antibody. In animal tests, the plastic particles bind to and neutralize a toxin found in bee stings; the toxin and antibody are then cleared to the liver, the same path taken by natural antibodies. Researchers are now developing plastic antibodies for a wider range of disease targets in hopes of broadening the availability of antibody therapies, which are currently very expensive.
Toxic target: The toxin melittin, labeled purple in these fluorescent images, spreads throughout the body of an untreated mouse, shown at bottom. The mouse at top has been injected with an artificial antibody, also fluorescently labeled, that binds to the toxin and takes it to the liver. The spread of the toxin throughout the treated mouse’s body is also more limited, which is why less of its body appears purple in this image.
For more than 20 years, biochemists have attempted to mimic antibodies' ability to zero in on their targets, as part of a strategy to make more effective and cheaper therapeutics and diagnostics. "Though antibodies are produced on an industrial scale today because they're so important, the cost is very, very high," says Kenneth Shea, professor of chemistry at the University of California, Irvine. That's because antibodies are grown in animals; they're complex molecules that can't be made in a test tube, or even by bacteria. And antibodies, like other proteins, are very fragile. Even under refrigeration, they last just months. The question Shea and others have asked for 20 years, he says, is "would it be possible to design them from inexpensive, abiotic starting materials?" Such plastic antibodies could be made cheaply and then sit on the shelf, in theory, for years.
In 2008, Shea's group, working with researchers from the Tokyo Institute of Technology, demonstrated for the first time that plastic antibodies made using a technique called molecular imprinting could bind to a target as strongly and specifically as natural antibodies. Molecular imprinting involves synthesizing a polymer in the presence of a target molecule. The polymer grows around the target, "imprinting" it with the target's shape. It's analogous to making a plaster cast of one's hand, says Shea. Looking to the properties of natural antibodies, Shea's group tailored the method for making polymers that more specifically target large proteins in biological solutions. Antibodies and their targets fit together like a key in a lock, or like a hand into a plaster cast. But they are also bound to their targets by chemistry and attracted by electrical interactions. Shea's methods involve looking to the properties of the target molecule and selecting starting materials that have an affinity for that target--in this case the protein melittin, the toxin in bee stings. At the same time, the method screens for starting materials that are not attracted to other, more common blood proteins. And the group took care to make the plastic antibody smaller than previous molecularly imprinted polymers, which were too big to be recognized by the body.
Shea's plastic antibody targeting melittin performed well in test tubes, but there was still some skepticism whether it would work in the complex environment of the body. This month in Journal of the American Chemical Society, the University of California researchers describe promising studies in mice. The researchers attached different fluorescent imaging probes to melittin and to the plastic antibody, injected them into the mice, and watched what happened in real time. Because the probes were two different colors, the researchers were able to watch as the polymer met its target in vivo, and as the two were then cleared to the liver. In mice given only the toxin and not the antidote, the mice's symptoms were much worse, and the toxin was more widely distributed throughout the body.
"They show that these materials are biocompatible and really act like antibodies--it's kind of surprising," says Ken Shimizu, professor of biochemistry at the University of South Carolina. Researchers had suspected that the body might not recognize the plastic particles as antibodies and thus they would be ineffective, or that they might get gummed up with other particles in the complex mixture that is the bloodstream.
Shea says that he's been contacted by several pharmaceutical companies that are interested in seeing how the work develops. David Spivak, professor of chemistry at Louisiana State University, agrees that the method is "a general strategy that will work again and again." "These particles have huge advantages in terms of stability and low cost," says Spivak. "I just hope this work is reproducible for many different targets."
The California researchers developed their imprinting methods using melittin because it's relatively inexpensive and easy to obtain, and it's a good representative of a class of small protein toxins, some of which are much more deadly. "Our next steps are to pursue more serious toxins," says Shea.
By Katherine Bourzac
From Technology Review