Shake-and-Make Proteins

Arthur Olson is shaking up the molecular world.

Bennett Daviss
Jun 5, 2005

Courtesy of Arthur Olson

Arthur Olson is shaking up the molecular world. A structural biologist at the Scripps Research Institute in La Jolla, Calif., Olson has been making models of proteins, DNA strands, and other biological objects for more than 25 years – first by hand, then with computers. These days, he uses rapid prototyping systems borrowed from industry, along with a small jar and a few magnets, to shake together biological models the way a bartender shakes together a good martini.

"If I'm talking with a colleague about a structure and I can hold it in my hand, that's a natural perceptual experience," more intuitive than using a mouse to rotate an image on a screen, he says. Using a modeling device that extrudes plastic parts, Olson crafted 12 identical pieces that went together in the spherical form of a viral capsid. "But the parts didn't hold together unless you put a rubber band around them," he recalls.

Olson's group had been using magnets to hold together other models, "so it was an obvious approach," he says. He realized that magnets could be placed along the edges of his virus model's plastic pieces, just like the electrostatic charges that connect the proteins of actual viruses. But he also saw something else: "If I had the right amount of magnetic force relative to the weight of the pieces, the parts could self-assemble" just as the real things do. "I put the pieces in a jar, shook them up, and it worked."

From experience, Olson knows that physical models can be more effective teachers than words or images, especially when explaining complex biological ideas such as self-assembly to nonscientists. "People often think that self-assembly is a linear or sequential process," he says. "The idea that you can get complex structures from the random motion of individual components can be hard for [people] to swallow if you just explain it. Now I can demonstrate it before their eyes."

As Olson shakes the capsid model's pieces in a small jar, he explains that the strength of the shaking mimics the effects of temperature. With light shaking, the equivalent of a cool environment, the pieces rattle loosely in the container; with violent shaking, the pieces come together but fly apart again. Under a moderate, persistent shaking (in other words, when the "temperature" is just right) during about 30 seconds of random collisions, the pebbles snap together into a complete, stable sphere.

Graduate student Shuye Pu at Toronto's Hospital for Sick Children has used Olson's shake-and-make models with children and their parents to illustrate concepts involved in disease and research. For youngsters beyond elementary-school age, "the models are helpful as an introductory tool" for the idea of self-assembly, he says.

Psychologists have learned that humans make constant use of an innate "body sense," a sort of physical intuition, to understand how the physical world works, Olson explains. For that reason, "There may be more useful information in physical models than computer images."